diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,28569 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.999944821497545,
+  "eval_steps": 500,
+  "global_step": 40776,
+  "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
+    },
+    {
+      "epoch": 0.8835917526531664,
+      "grad_norm": 0.384056031703949,
+      "learning_rate": 8.225687499314201e-05,
+      "loss": 3.1511,
+      "step": 12010
+    },
+    {
+      "epoch": 0.884327466019239,
+      "grad_norm": 0.3878827393054962,
+      "learning_rate": 8.222669114080088e-05,
+      "loss": 3.1477,
+      "step": 12020
+    },
+    {
+      "epoch": 0.8850631793853115,
+      "grad_norm": 0.38082295656204224,
+      "learning_rate": 8.219648718488796e-05,
+      "loss": 3.1196,
+      "step": 12030
+    },
+    {
+      "epoch": 0.885798892751384,
+      "grad_norm": 0.4064112603664398,
+      "learning_rate": 8.216626314424496e-05,
+      "loss": 3.1508,
+      "step": 12040
+    },
+    {
+      "epoch": 0.8865346061174566,
+      "grad_norm": 0.36879783868789673,
+      "learning_rate": 8.213601903772619e-05,
+      "loss": 3.1914,
+      "step": 12050
+    },
+    {
+      "epoch": 0.8872703194835292,
+      "grad_norm": 0.39445754885673523,
+      "learning_rate": 8.210575488419842e-05,
+      "loss": 3.1541,
+      "step": 12060
+    },
+    {
+      "epoch": 0.8880060328496018,
+      "grad_norm": 0.37045836448669434,
+      "learning_rate": 8.2075470702541e-05,
+      "loss": 3.1005,
+      "step": 12070
+    },
+    {
+      "epoch": 0.8887417462156744,
+      "grad_norm": 0.3785856068134308,
+      "learning_rate": 8.204516651164569e-05,
+      "loss": 3.181,
+      "step": 12080
+    },
+    {
+      "epoch": 0.889477459581747,
+      "grad_norm": 0.3615811765193939,
+      "learning_rate": 8.20148423304168e-05,
+      "loss": 3.1494,
+      "step": 12090
+    },
+    {
+      "epoch": 0.8902131729478195,
+      "grad_norm": 0.3599592447280884,
+      "learning_rate": 8.198449817777107e-05,
+      "loss": 3.146,
+      "step": 12100
+    },
+    {
+      "epoch": 0.8909488863138921,
+      "grad_norm": 0.3932536840438843,
+      "learning_rate": 8.195413407263773e-05,
+      "loss": 3.1497,
+      "step": 12110
+    },
+    {
+      "epoch": 0.8916845996799647,
+      "grad_norm": 0.3764161467552185,
+      "learning_rate": 8.19237500339584e-05,
+      "loss": 3.1388,
+      "step": 12120
+    },
+    {
+      "epoch": 0.8924203130460373,
+      "grad_norm": 0.37524348497390747,
+      "learning_rate": 8.189334608068722e-05,
+      "loss": 3.1752,
+      "step": 12130
+    },
+    {
+      "epoch": 0.8931560264121099,
+      "grad_norm": 0.37849491834640503,
+      "learning_rate": 8.186292223179066e-05,
+      "loss": 3.1636,
+      "step": 12140
+    },
+    {
+      "epoch": 0.8938917397781824,
+      "grad_norm": 0.3842037320137024,
+      "learning_rate": 8.183247850624769e-05,
+      "loss": 3.1258,
+      "step": 12150
+    },
+    {
+      "epoch": 0.894627453144255,
+      "grad_norm": 0.3818165063858032,
+      "learning_rate": 8.180201492304961e-05,
+      "loss": 3.1679,
+      "step": 12160
+    },
+    {
+      "epoch": 0.8953631665103275,
+      "grad_norm": 0.3639425039291382,
+      "learning_rate": 8.177153150120015e-05,
+      "loss": 3.1179,
+      "step": 12170
+    },
+    {
+      "epoch": 0.8960988798764001,
+      "grad_norm": 0.3635939359664917,
+      "learning_rate": 8.174102825971537e-05,
+      "loss": 3.1823,
+      "step": 12180
+    },
+    {
+      "epoch": 0.8968345932424727,
+      "grad_norm": 0.36238768696784973,
+      "learning_rate": 8.171050521762377e-05,
+      "loss": 3.1742,
+      "step": 12190
+    },
+    {
+      "epoch": 0.8975703066085453,
+      "grad_norm": 0.38135668635368347,
+      "learning_rate": 8.167996239396612e-05,
+      "loss": 3.1725,
+      "step": 12200
+    },
+    {
+      "epoch": 0.8983060199746179,
+      "grad_norm": 0.3708590567111969,
+      "learning_rate": 8.164939980779558e-05,
+      "loss": 3.1419,
+      "step": 12210
+    },
+    {
+      "epoch": 0.8990417333406905,
+      "grad_norm": 0.36212167143821716,
+      "learning_rate": 8.161881747817762e-05,
+      "loss": 3.1576,
+      "step": 12220
+    },
+    {
+      "epoch": 0.8997774467067631,
+      "grad_norm": 0.3543083071708679,
+      "learning_rate": 8.158821542419002e-05,
+      "loss": 3.099,
+      "step": 12230
+    },
+    {
+      "epoch": 0.9005131600728357,
+      "grad_norm": 0.3975955843925476,
+      "learning_rate": 8.155759366492289e-05,
+      "loss": 3.0783,
+      "step": 12240
+    },
+    {
+      "epoch": 0.9012488734389082,
+      "grad_norm": 0.369435578584671,
+      "learning_rate": 8.152695221947862e-05,
+      "loss": 3.1026,
+      "step": 12250
+    },
+    {
+      "epoch": 0.9019845868049807,
+      "grad_norm": 0.35627511143684387,
+      "learning_rate": 8.149629110697187e-05,
+      "loss": 3.0996,
+      "step": 12260
+    },
+    {
+      "epoch": 0.9027203001710533,
+      "grad_norm": 0.36654290556907654,
+      "learning_rate": 8.146561034652958e-05,
+      "loss": 3.1544,
+      "step": 12270
+    },
+    {
+      "epoch": 0.9034560135371259,
+      "grad_norm": 0.37846630811691284,
+      "learning_rate": 8.143490995729095e-05,
+      "loss": 3.1311,
+      "step": 12280
+    },
+    {
+      "epoch": 0.9041917269031985,
+      "grad_norm": 0.3826802670955658,
+      "learning_rate": 8.14041899584074e-05,
+      "loss": 3.1778,
+      "step": 12290
+    },
+    {
+      "epoch": 0.9049274402692711,
+      "grad_norm": 0.4068825840950012,
+      "learning_rate": 8.13734503690426e-05,
+      "loss": 3.1459,
+      "step": 12300
+    },
+    {
+      "epoch": 0.9056631536353437,
+      "grad_norm": 0.3719536364078522,
+      "learning_rate": 8.134269120837246e-05,
+      "loss": 3.1201,
+      "step": 12310
+    },
+    {
+      "epoch": 0.9063988670014163,
+      "grad_norm": 0.37469327449798584,
+      "learning_rate": 8.131191249558508e-05,
+      "loss": 3.1282,
+      "step": 12320
+    },
+    {
+      "epoch": 0.9071345803674888,
+      "grad_norm": 0.3764030337333679,
+      "learning_rate": 8.128111424988075e-05,
+      "loss": 3.1359,
+      "step": 12330
+    },
+    {
+      "epoch": 0.9078702937335614,
+      "grad_norm": 0.3773621618747711,
+      "learning_rate": 8.125029649047194e-05,
+      "loss": 3.1376,
+      "step": 12340
+    },
+    {
+      "epoch": 0.908606007099634,
+      "grad_norm": 0.42097413539886475,
+      "learning_rate": 8.121945923658334e-05,
+      "loss": 3.1406,
+      "step": 12350
+    },
+    {
+      "epoch": 0.9093417204657066,
+      "grad_norm": 0.36304715275764465,
+      "learning_rate": 8.118860250745172e-05,
+      "loss": 3.109,
+      "step": 12360
+    },
+    {
+      "epoch": 0.9100774338317792,
+      "grad_norm": 0.37655285000801086,
+      "learning_rate": 8.11577263223261e-05,
+      "loss": 3.108,
+      "step": 12370
+    },
+    {
+      "epoch": 0.9108131471978517,
+      "grad_norm": 0.3614340126514435,
+      "learning_rate": 8.112683070046754e-05,
+      "loss": 3.158,
+      "step": 12380
+    },
+    {
+      "epoch": 0.9115488605639243,
+      "grad_norm": 0.4038553237915039,
+      "learning_rate": 8.109591566114927e-05,
+      "loss": 3.1277,
+      "step": 12390
+    },
+    {
+      "epoch": 0.9122845739299968,
+      "grad_norm": 0.3722144365310669,
+      "learning_rate": 8.106498122365667e-05,
+      "loss": 3.1167,
+      "step": 12400
+    },
+    {
+      "epoch": 0.9130202872960694,
+      "grad_norm": 0.37795814871788025,
+      "learning_rate": 8.103402740728714e-05,
+      "loss": 3.1157,
+      "step": 12410
+    },
+    {
+      "epoch": 0.913756000662142,
+      "grad_norm": 0.36336052417755127,
+      "learning_rate": 8.100305423135022e-05,
+      "loss": 3.1013,
+      "step": 12420
+    },
+    {
+      "epoch": 0.9144917140282146,
+      "grad_norm": 0.3927875757217407,
+      "learning_rate": 8.097206171516752e-05,
+      "loss": 3.1318,
+      "step": 12430
+    },
+    {
+      "epoch": 0.9152274273942872,
+      "grad_norm": 0.3939864933490753,
+      "learning_rate": 8.094104987807271e-05,
+      "loss": 3.1577,
+      "step": 12440
+    },
+    {
+      "epoch": 0.9159631407603598,
+      "grad_norm": 0.3720358610153198,
+      "learning_rate": 8.091001873941152e-05,
+      "loss": 3.1832,
+      "step": 12450
+    },
+    {
+      "epoch": 0.9166988541264324,
+      "grad_norm": 0.36232516169548035,
+      "learning_rate": 8.087896831854172e-05,
+      "loss": 3.1081,
+      "step": 12460
+    },
+    {
+      "epoch": 0.917434567492505,
+      "grad_norm": 0.36899566650390625,
+      "learning_rate": 8.084789863483308e-05,
+      "loss": 3.1659,
+      "step": 12470
+    },
+    {
+      "epoch": 0.9181702808585775,
+      "grad_norm": 0.36052051186561584,
+      "learning_rate": 8.081680970766744e-05,
+      "loss": 3.1422,
+      "step": 12480
+    },
+    {
+      "epoch": 0.91890599422465,
+      "grad_norm": 0.370313435792923,
+      "learning_rate": 8.078570155643861e-05,
+      "loss": 3.1042,
+      "step": 12490
+    },
+    {
+      "epoch": 0.9196417075907226,
+      "grad_norm": 0.37691476941108704,
+      "learning_rate": 8.075457420055237e-05,
+      "loss": 3.1193,
+      "step": 12500
+    },
+    {
+      "epoch": 0.9203774209567952,
+      "grad_norm": 0.37345367670059204,
+      "learning_rate": 8.072342765942654e-05,
+      "loss": 3.1489,
+      "step": 12510
+    },
+    {
+      "epoch": 0.9211131343228678,
+      "grad_norm": 0.3992007076740265,
+      "learning_rate": 8.069226195249086e-05,
+      "loss": 3.1026,
+      "step": 12520
+    },
+    {
+      "epoch": 0.9218488476889404,
+      "grad_norm": 0.36022311449050903,
+      "learning_rate": 8.066107709918704e-05,
+      "loss": 3.1271,
+      "step": 12530
+    },
+    {
+      "epoch": 0.922584561055013,
+      "grad_norm": 0.3909919857978821,
+      "learning_rate": 8.062987311896874e-05,
+      "loss": 3.1843,
+      "step": 12540
+    },
+    {
+      "epoch": 0.9233202744210856,
+      "grad_norm": 0.3629285395145416,
+      "learning_rate": 8.059865003130155e-05,
+      "loss": 3.1109,
+      "step": 12550
+    },
+    {
+      "epoch": 0.9240559877871581,
+      "grad_norm": 0.40764477849006653,
+      "learning_rate": 8.056740785566297e-05,
+      "loss": 3.1135,
+      "step": 12560
+    },
+    {
+      "epoch": 0.9247917011532307,
+      "grad_norm": 0.3628876209259033,
+      "learning_rate": 8.053614661154241e-05,
+      "loss": 3.1455,
+      "step": 12570
+    },
+    {
+      "epoch": 0.9255274145193033,
+      "grad_norm": 0.3637199103832245,
+      "learning_rate": 8.050486631844119e-05,
+      "loss": 3.1183,
+      "step": 12580
+    },
+    {
+      "epoch": 0.9262631278853759,
+      "grad_norm": 0.37777504324913025,
+      "learning_rate": 8.04735669958725e-05,
+      "loss": 3.1734,
+      "step": 12590
+    },
+    {
+      "epoch": 0.9269988412514485,
+      "grad_norm": 0.35676929354667664,
+      "learning_rate": 8.044224866336139e-05,
+      "loss": 3.1472,
+      "step": 12600
+    },
+    {
+      "epoch": 0.927734554617521,
+      "grad_norm": 0.3731575906276703,
+      "learning_rate": 8.041091134044481e-05,
+      "loss": 3.1575,
+      "step": 12610
+    },
+    {
+      "epoch": 0.9284702679835936,
+      "grad_norm": 0.36572179198265076,
+      "learning_rate": 8.037955504667149e-05,
+      "loss": 3.1557,
+      "step": 12620
+    },
+    {
+      "epoch": 0.9292059813496661,
+      "grad_norm": 0.35935401916503906,
+      "learning_rate": 8.034817980160206e-05,
+      "loss": 3.1428,
+      "step": 12630
+    },
+    {
+      "epoch": 0.9299416947157387,
+      "grad_norm": 0.3554179072380066,
+      "learning_rate": 8.031678562480895e-05,
+      "loss": 3.1074,
+      "step": 12640
+    },
+    {
+      "epoch": 0.9306774080818113,
+      "grad_norm": 0.375542551279068,
+      "learning_rate": 8.028537253587637e-05,
+      "loss": 3.116,
+      "step": 12650
+    },
+    {
+      "epoch": 0.9314131214478839,
+      "grad_norm": 0.4085438847541809,
+      "learning_rate": 8.025394055440036e-05,
+      "loss": 3.1518,
+      "step": 12660
+    },
+    {
+      "epoch": 0.9321488348139565,
+      "grad_norm": 0.3811233639717102,
+      "learning_rate": 8.022248969998876e-05,
+      "loss": 3.1088,
+      "step": 12670
+    },
+    {
+      "epoch": 0.9328845481800291,
+      "grad_norm": 0.38336530327796936,
+      "learning_rate": 8.019101999226113e-05,
+      "loss": 3.1159,
+      "step": 12680
+    },
+    {
+      "epoch": 0.9336202615461017,
+      "grad_norm": 0.4039757549762726,
+      "learning_rate": 8.015953145084885e-05,
+      "loss": 3.1084,
+      "step": 12690
+    },
+    {
+      "epoch": 0.9343559749121743,
+      "grad_norm": 0.34968018531799316,
+      "learning_rate": 8.012802409539502e-05,
+      "loss": 3.1568,
+      "step": 12700
+    },
+    {
+      "epoch": 0.9350916882782468,
+      "grad_norm": 0.3883592486381531,
+      "learning_rate": 8.009649794555443e-05,
+      "loss": 3.1653,
+      "step": 12710
+    },
+    {
+      "epoch": 0.9358274016443193,
+      "grad_norm": 0.36118948459625244,
+      "learning_rate": 8.00649530209937e-05,
+      "loss": 3.1278,
+      "step": 12720
+    },
+    {
+      "epoch": 0.9365631150103919,
+      "grad_norm": 0.37414583563804626,
+      "learning_rate": 8.003338934139109e-05,
+      "loss": 3.0814,
+      "step": 12730
+    },
+    {
+      "epoch": 0.9372988283764645,
+      "grad_norm": 0.37282320857048035,
+      "learning_rate": 8.000180692643655e-05,
+      "loss": 3.1265,
+      "step": 12740
+    },
+    {
+      "epoch": 0.9380345417425371,
+      "grad_norm": 0.3765084147453308,
+      "learning_rate": 7.997020579583176e-05,
+      "loss": 3.134,
+      "step": 12750
+    },
+    {
+      "epoch": 0.9387702551086097,
+      "grad_norm": 0.37491947412490845,
+      "learning_rate": 7.993858596929006e-05,
+      "loss": 3.1372,
+      "step": 12760
+    },
+    {
+      "epoch": 0.9395059684746823,
+      "grad_norm": 0.3822542428970337,
+      "learning_rate": 7.990694746653644e-05,
+      "loss": 3.1357,
+      "step": 12770
+    },
+    {
+      "epoch": 0.9402416818407549,
+      "grad_norm": 0.3770379424095154,
+      "learning_rate": 7.987529030730756e-05,
+      "loss": 3.1257,
+      "step": 12780
+    },
+    {
+      "epoch": 0.9409773952068274,
+      "grad_norm": 0.38460904359817505,
+      "learning_rate": 7.984361451135169e-05,
+      "loss": 3.1367,
+      "step": 12790
+    },
+    {
+      "epoch": 0.9417131085729,
+      "grad_norm": 0.3640877306461334,
+      "learning_rate": 7.981192009842875e-05,
+      "loss": 3.0863,
+      "step": 12800
+    },
+    {
+      "epoch": 0.9424488219389726,
+      "grad_norm": 0.3563246726989746,
+      "learning_rate": 7.978020708831027e-05,
+      "loss": 3.0937,
+      "step": 12810
+    },
+    {
+      "epoch": 0.9431845353050452,
+      "grad_norm": 0.36839959025382996,
+      "learning_rate": 7.97484755007794e-05,
+      "loss": 3.1342,
+      "step": 12820
+    },
+    {
+      "epoch": 0.9439202486711177,
+      "grad_norm": 0.389789342880249,
+      "learning_rate": 7.971672535563082e-05,
+      "loss": 3.1667,
+      "step": 12830
+    },
+    {
+      "epoch": 0.9446559620371903,
+      "grad_norm": 0.3732822835445404,
+      "learning_rate": 7.968495667267084e-05,
+      "loss": 3.0957,
+      "step": 12840
+    },
+    {
+      "epoch": 0.9453916754032629,
+      "grad_norm": 0.3937614858150482,
+      "learning_rate": 7.965316947171733e-05,
+      "loss": 3.1056,
+      "step": 12850
+    },
+    {
+      "epoch": 0.9461273887693354,
+      "grad_norm": 0.3977773189544678,
+      "learning_rate": 7.962136377259968e-05,
+      "loss": 3.1629,
+      "step": 12860
+    },
+    {
+      "epoch": 0.946863102135408,
+      "grad_norm": 0.3730141520500183,
+      "learning_rate": 7.958953959515885e-05,
+      "loss": 3.1369,
+      "step": 12870
+    },
+    {
+      "epoch": 0.9475988155014806,
+      "grad_norm": 0.3614790439605713,
+      "learning_rate": 7.955769695924731e-05,
+      "loss": 3.109,
+      "step": 12880
+    },
+    {
+      "epoch": 0.9483345288675532,
+      "grad_norm": 0.38277482986450195,
+      "learning_rate": 7.952583588472908e-05,
+      "loss": 3.145,
+      "step": 12890
+    },
+    {
+      "epoch": 0.9490702422336258,
+      "grad_norm": 0.37006932497024536,
+      "learning_rate": 7.949395639147961e-05,
+      "loss": 3.1408,
+      "step": 12900
+    },
+    {
+      "epoch": 0.9498059555996984,
+      "grad_norm": 0.35835763812065125,
+      "learning_rate": 7.946205849938591e-05,
+      "loss": 3.1179,
+      "step": 12910
+    },
+    {
+      "epoch": 0.950541668965771,
+      "grad_norm": 0.3633943200111389,
+      "learning_rate": 7.943014222834644e-05,
+      "loss": 3.1429,
+      "step": 12920
+    },
+    {
+      "epoch": 0.9512773823318436,
+      "grad_norm": 0.37867772579193115,
+      "learning_rate": 7.939820759827112e-05,
+      "loss": 3.0821,
+      "step": 12930
+    },
+    {
+      "epoch": 0.9520130956979161,
+      "grad_norm": 0.39402279257774353,
+      "learning_rate": 7.936625462908134e-05,
+      "loss": 3.1049,
+      "step": 12940
+    },
+    {
+      "epoch": 0.9527488090639886,
+      "grad_norm": 0.37046343088150024,
+      "learning_rate": 7.93342833407099e-05,
+      "loss": 3.0979,
+      "step": 12950
+    },
+    {
+      "epoch": 0.9534845224300612,
+      "grad_norm": 0.3563176691532135,
+      "learning_rate": 7.930229375310105e-05,
+      "loss": 3.0774,
+      "step": 12960
+    },
+    {
+      "epoch": 0.9542202357961338,
+      "grad_norm": 0.36553964018821716,
+      "learning_rate": 7.927028588621049e-05,
+      "loss": 3.0783,
+      "step": 12970
+    },
+    {
+      "epoch": 0.9549559491622064,
+      "grad_norm": 0.3654845952987671,
+      "learning_rate": 7.923825976000524e-05,
+      "loss": 3.1427,
+      "step": 12980
+    },
+    {
+      "epoch": 0.955691662528279,
+      "grad_norm": 0.3755556643009186,
+      "learning_rate": 7.920621539446378e-05,
+      "loss": 3.1866,
+      "step": 12990
+    },
+    {
+      "epoch": 0.9564273758943516,
+      "grad_norm": 0.36664122343063354,
+      "learning_rate": 7.917415280957592e-05,
+      "loss": 3.1482,
+      "step": 13000
+    },
+    {
+      "epoch": 0.9571630892604241,
+      "grad_norm": 0.36148959398269653,
+      "learning_rate": 7.914207202534288e-05,
+      "loss": 3.1013,
+      "step": 13010
+    },
+    {
+      "epoch": 0.9578988026264967,
+      "grad_norm": 0.36350470781326294,
+      "learning_rate": 7.910997306177721e-05,
+      "loss": 3.1396,
+      "step": 13020
+    },
+    {
+      "epoch": 0.9586345159925693,
+      "grad_norm": 0.3963392376899719,
+      "learning_rate": 7.90778559389028e-05,
+      "loss": 3.1136,
+      "step": 13030
+    },
+    {
+      "epoch": 0.9593702293586419,
+      "grad_norm": 0.3724636733531952,
+      "learning_rate": 7.904572067675487e-05,
+      "loss": 3.1085,
+      "step": 13040
+    },
+    {
+      "epoch": 0.9601059427247145,
+      "grad_norm": 0.37127721309661865,
+      "learning_rate": 7.901356729537997e-05,
+      "loss": 3.088,
+      "step": 13050
+    },
+    {
+      "epoch": 0.960841656090787,
+      "grad_norm": 0.37480801343917847,
+      "learning_rate": 7.898139581483593e-05,
+      "loss": 3.1121,
+      "step": 13060
+    },
+    {
+      "epoch": 0.9615773694568596,
+      "grad_norm": 0.37211763858795166,
+      "learning_rate": 7.894920625519186e-05,
+      "loss": 3.1662,
+      "step": 13070
+    },
+    {
+      "epoch": 0.9623130828229322,
+      "grad_norm": 0.3749638795852661,
+      "learning_rate": 7.891699863652822e-05,
+      "loss": 3.0937,
+      "step": 13080
+    },
+    {
+      "epoch": 0.9630487961890047,
+      "grad_norm": 0.3769051432609558,
+      "learning_rate": 7.888477297893662e-05,
+      "loss": 3.1387,
+      "step": 13090
+    },
+    {
+      "epoch": 0.9637845095550773,
+      "grad_norm": 0.3631693124771118,
+      "learning_rate": 7.885252930252002e-05,
+      "loss": 3.1002,
+      "step": 13100
+    },
+    {
+      "epoch": 0.9645202229211499,
+      "grad_norm": 0.3677249252796173,
+      "learning_rate": 7.882026762739261e-05,
+      "loss": 3.1057,
+      "step": 13110
+    },
+    {
+      "epoch": 0.9652559362872225,
+      "grad_norm": 0.38757720589637756,
+      "learning_rate": 7.878798797367976e-05,
+      "loss": 3.1532,
+      "step": 13120
+    },
+    {
+      "epoch": 0.9659916496532951,
+      "grad_norm": 0.36416444182395935,
+      "learning_rate": 7.875569036151806e-05,
+      "loss": 3.1152,
+      "step": 13130
+    },
+    {
+      "epoch": 0.9667273630193677,
+      "grad_norm": 0.3727430999279022,
+      "learning_rate": 7.872337481105536e-05,
+      "loss": 3.0831,
+      "step": 13140
+    },
+    {
+      "epoch": 0.9674630763854403,
+      "grad_norm": 0.35649845004081726,
+      "learning_rate": 7.869104134245067e-05,
+      "loss": 3.1092,
+      "step": 13150
+    },
+    {
+      "epoch": 0.9681987897515129,
+      "grad_norm": 0.36332109570503235,
+      "learning_rate": 7.865868997587413e-05,
+      "loss": 3.1554,
+      "step": 13160
+    },
+    {
+      "epoch": 0.9689345031175854,
+      "grad_norm": 0.3543636202812195,
+      "learning_rate": 7.862632073150711e-05,
+      "loss": 3.1036,
+      "step": 13170
+    },
+    {
+      "epoch": 0.9696702164836579,
+      "grad_norm": 0.3681279718875885,
+      "learning_rate": 7.859393362954211e-05,
+      "loss": 3.1262,
+      "step": 13180
+    },
+    {
+      "epoch": 0.9704059298497305,
+      "grad_norm": 0.36408326029777527,
+      "learning_rate": 7.856152869018276e-05,
+      "loss": 3.1728,
+      "step": 13190
+    },
+    {
+      "epoch": 0.9711416432158031,
+      "grad_norm": 0.37902170419692993,
+      "learning_rate": 7.852910593364385e-05,
+      "loss": 3.1296,
+      "step": 13200
+    },
+    {
+      "epoch": 0.9718773565818757,
+      "grad_norm": 0.3592076301574707,
+      "learning_rate": 7.84966653801512e-05,
+      "loss": 3.0681,
+      "step": 13210
+    },
+    {
+      "epoch": 0.9726130699479483,
+      "grad_norm": 0.36775195598602295,
+      "learning_rate": 7.846420704994189e-05,
+      "loss": 3.093,
+      "step": 13220
+    },
+    {
+      "epoch": 0.9733487833140209,
+      "grad_norm": 0.3821685314178467,
+      "learning_rate": 7.843173096326391e-05,
+      "loss": 3.1229,
+      "step": 13230
+    },
+    {
+      "epoch": 0.9740844966800934,
+      "grad_norm": 0.36848169565200806,
+      "learning_rate": 7.839923714037645e-05,
+      "loss": 3.1491,
+      "step": 13240
+    },
+    {
+      "epoch": 0.974820210046166,
+      "grad_norm": 0.37354278564453125,
+      "learning_rate": 7.83667256015497e-05,
+      "loss": 3.1242,
+      "step": 13250
+    },
+    {
+      "epoch": 0.9755559234122386,
+      "grad_norm": 0.38822486996650696,
+      "learning_rate": 7.833419636706494e-05,
+      "loss": 3.0887,
+      "step": 13260
+    },
+    {
+      "epoch": 0.9762916367783112,
+      "grad_norm": 0.3655138909816742,
+      "learning_rate": 7.830164945721447e-05,
+      "loss": 3.109,
+      "step": 13270
+    },
+    {
+      "epoch": 0.9770273501443838,
+      "grad_norm": 0.35819318890571594,
+      "learning_rate": 7.826908489230161e-05,
+      "loss": 3.1081,
+      "step": 13280
+    },
+    {
+      "epoch": 0.9777630635104563,
+      "grad_norm": 0.40228065848350525,
+      "learning_rate": 7.823650269264072e-05,
+      "loss": 3.0395,
+      "step": 13290
+    },
+    {
+      "epoch": 0.9784987768765289,
+      "grad_norm": 0.36907315254211426,
+      "learning_rate": 7.820390287855714e-05,
+      "loss": 3.1227,
+      "step": 13300
+    },
+    {
+      "epoch": 0.9792344902426015,
+      "grad_norm": 0.34603267908096313,
+      "learning_rate": 7.817128547038719e-05,
+      "loss": 3.0873,
+      "step": 13310
+    },
+    {
+      "epoch": 0.979970203608674,
+      "grad_norm": 0.3589996099472046,
+      "learning_rate": 7.813865048847819e-05,
+      "loss": 3.0838,
+      "step": 13320
+    },
+    {
+      "epoch": 0.9807059169747466,
+      "grad_norm": 0.3714243173599243,
+      "learning_rate": 7.810599795318841e-05,
+      "loss": 3.1352,
+      "step": 13330
+    },
+    {
+      "epoch": 0.9814416303408192,
+      "grad_norm": 0.3626115918159485,
+      "learning_rate": 7.807332788488708e-05,
+      "loss": 3.1093,
+      "step": 13340
+    },
+    {
+      "epoch": 0.9821773437068918,
+      "grad_norm": 0.36209580302238464,
+      "learning_rate": 7.804064030395433e-05,
+      "loss": 3.1061,
+      "step": 13350
+    },
+    {
+      "epoch": 0.9829130570729644,
+      "grad_norm": 0.3716304898262024,
+      "learning_rate": 7.800793523078127e-05,
+      "loss": 3.0897,
+      "step": 13360
+    },
+    {
+      "epoch": 0.983648770439037,
+      "grad_norm": 0.3670690953731537,
+      "learning_rate": 7.797521268576989e-05,
+      "loss": 3.1076,
+      "step": 13370
+    },
+    {
+      "epoch": 0.9843844838051096,
+      "grad_norm": 0.35003265738487244,
+      "learning_rate": 7.794247268933306e-05,
+      "loss": 3.1319,
+      "step": 13380
+    },
+    {
+      "epoch": 0.9851201971711822,
+      "grad_norm": 0.3642325699329376,
+      "learning_rate": 7.790971526189458e-05,
+      "loss": 3.0966,
+      "step": 13390
+    },
+    {
+      "epoch": 0.9858559105372546,
+      "grad_norm": 0.37202367186546326,
+      "learning_rate": 7.787694042388912e-05,
+      "loss": 3.1412,
+      "step": 13400
+    },
+    {
+      "epoch": 0.9865916239033272,
+      "grad_norm": 0.36938855051994324,
+      "learning_rate": 7.784414819576216e-05,
+      "loss": 3.1121,
+      "step": 13410
+    },
+    {
+      "epoch": 0.9873273372693998,
+      "grad_norm": 0.36945322155952454,
+      "learning_rate": 7.78113385979701e-05,
+      "loss": 3.0844,
+      "step": 13420
+    },
+    {
+      "epoch": 0.9880630506354724,
+      "grad_norm": 0.3800085186958313,
+      "learning_rate": 7.777851165098012e-05,
+      "loss": 3.0645,
+      "step": 13430
+    },
+    {
+      "epoch": 0.988798764001545,
+      "grad_norm": 0.3750995993614197,
+      "learning_rate": 7.774566737527024e-05,
+      "loss": 3.1037,
+      "step": 13440
+    },
+    {
+      "epoch": 0.9895344773676176,
+      "grad_norm": 0.3711072504520416,
+      "learning_rate": 7.77128057913293e-05,
+      "loss": 3.1099,
+      "step": 13450
+    },
+    {
+      "epoch": 0.9902701907336902,
+      "grad_norm": 0.3609575927257538,
+      "learning_rate": 7.767992691965691e-05,
+      "loss": 3.1381,
+      "step": 13460
+    },
+    {
+      "epoch": 0.9910059040997627,
+      "grad_norm": 0.36476248502731323,
+      "learning_rate": 7.76470307807635e-05,
+      "loss": 3.1068,
+      "step": 13470
+    },
+    {
+      "epoch": 0.9917416174658353,
+      "grad_norm": 0.38527876138687134,
+      "learning_rate": 7.761411739517026e-05,
+      "loss": 3.1427,
+      "step": 13480
+    },
+    {
+      "epoch": 0.9924773308319079,
+      "grad_norm": 0.3626437783241272,
+      "learning_rate": 7.758118678340913e-05,
+      "loss": 3.1093,
+      "step": 13490
+    },
+    {
+      "epoch": 0.9932130441979805,
+      "grad_norm": 0.35517948865890503,
+      "learning_rate": 7.754823896602279e-05,
+      "loss": 3.0807,
+      "step": 13500
+    },
+    {
+      "epoch": 0.9939487575640531,
+      "grad_norm": 0.37881192564964294,
+      "learning_rate": 7.751527396356468e-05,
+      "loss": 3.1387,
+      "step": 13510
+    },
+    {
+      "epoch": 0.9946844709301256,
+      "grad_norm": 0.3774098753929138,
+      "learning_rate": 7.748229179659893e-05,
+      "loss": 3.1136,
+      "step": 13520
+    },
+    {
+      "epoch": 0.9954201842961982,
+      "grad_norm": 0.34989410638809204,
+      "learning_rate": 7.744929248570038e-05,
+      "loss": 3.1537,
+      "step": 13530
+    },
+    {
+      "epoch": 0.9961558976622708,
+      "grad_norm": 0.3914855718612671,
+      "learning_rate": 7.74162760514546e-05,
+      "loss": 3.1459,
+      "step": 13540
+    },
+    {
+      "epoch": 0.9968916110283433,
+      "grad_norm": 0.36263689398765564,
+      "learning_rate": 7.738324251445782e-05,
+      "loss": 3.1451,
+      "step": 13550
+    },
+    {
+      "epoch": 0.9976273243944159,
+      "grad_norm": 0.37384799122810364,
+      "learning_rate": 7.73501918953169e-05,
+      "loss": 3.1215,
+      "step": 13560
+    },
+    {
+      "epoch": 0.9983630377604885,
+      "grad_norm": 0.36027422547340393,
+      "learning_rate": 7.731712421464942e-05,
+      "loss": 3.1713,
+      "step": 13570
+    },
+    {
+      "epoch": 0.9990987511265611,
+      "grad_norm": 0.38130810856819153,
+      "learning_rate": 7.728403949308355e-05,
+      "loss": 3.1429,
+      "step": 13580
+    },
+    {
+      "epoch": 0.9998344644926337,
+      "grad_norm": 0.36381796002388,
+      "learning_rate": 7.725093775125818e-05,
+      "loss": 3.0944,
+      "step": 13590
+    },
+    {
+      "epoch": 1.0005701778587062,
+      "grad_norm": 0.38200533390045166,
+      "learning_rate": 7.721781900982269e-05,
+      "loss": 3.0705,
+      "step": 13600
+    },
+    {
+      "epoch": 1.0013058912247788,
+      "grad_norm": 0.35553497076034546,
+      "learning_rate": 7.718468328943716e-05,
+      "loss": 3.0488,
+      "step": 13610
+    },
+    {
+      "epoch": 1.0020416045908513,
+      "grad_norm": 0.35915321111679077,
+      "learning_rate": 7.715153061077219e-05,
+      "loss": 3.023,
+      "step": 13620
+    },
+    {
+      "epoch": 1.002777317956924,
+      "grad_norm": 0.36823633313179016,
+      "learning_rate": 7.711836099450905e-05,
+      "loss": 3.0513,
+      "step": 13630
+    },
+    {
+      "epoch": 1.0035130313229965,
+      "grad_norm": 0.3569531738758087,
+      "learning_rate": 7.708517446133951e-05,
+      "loss": 3.0068,
+      "step": 13640
+    },
+    {
+      "epoch": 1.004248744689069,
+      "grad_norm": 0.36070168018341064,
+      "learning_rate": 7.705197103196589e-05,
+      "loss": 3.0576,
+      "step": 13650
+    },
+    {
+      "epoch": 1.0049844580551417,
+      "grad_norm": 0.35593706369400024,
+      "learning_rate": 7.701875072710108e-05,
+      "loss": 3.0042,
+      "step": 13660
+    },
+    {
+      "epoch": 1.0057201714212143,
+      "grad_norm": 0.3728356957435608,
+      "learning_rate": 7.698551356746848e-05,
+      "loss": 3.0511,
+      "step": 13670
+    },
+    {
+      "epoch": 1.0064558847872869,
+      "grad_norm": 0.3572307229042053,
+      "learning_rate": 7.695225957380203e-05,
+      "loss": 3.0363,
+      "step": 13680
+    },
+    {
+      "epoch": 1.0071915981533595,
+      "grad_norm": 0.3627204895019531,
+      "learning_rate": 7.691898876684612e-05,
+      "loss": 3.0021,
+      "step": 13690
+    },
+    {
+      "epoch": 1.007927311519432,
+      "grad_norm": 0.37532514333724976,
+      "learning_rate": 7.688570116735567e-05,
+      "loss": 3.0403,
+      "step": 13700
+    },
+    {
+      "epoch": 1.0086630248855046,
+      "grad_norm": 0.37163272500038147,
+      "learning_rate": 7.68523967960961e-05,
+      "loss": 3.0372,
+      "step": 13710
+    },
+    {
+      "epoch": 1.0093987382515772,
+      "grad_norm": 0.35406947135925293,
+      "learning_rate": 7.681907567384319e-05,
+      "loss": 3.0116,
+      "step": 13720
+    },
+    {
+      "epoch": 1.0101344516176498,
+      "grad_norm": 0.3697194755077362,
+      "learning_rate": 7.678573782138331e-05,
+      "loss": 3.0303,
+      "step": 13730
+    },
+    {
+      "epoch": 1.0108701649837224,
+      "grad_norm": 0.39772310853004456,
+      "learning_rate": 7.675238325951313e-05,
+      "loss": 3.0265,
+      "step": 13740
+    },
+    {
+      "epoch": 1.011605878349795,
+      "grad_norm": 0.36769720911979675,
+      "learning_rate": 7.671901200903986e-05,
+      "loss": 2.99,
+      "step": 13750
+    },
+    {
+      "epoch": 1.0123415917158676,
+      "grad_norm": 0.36903396248817444,
+      "learning_rate": 7.668562409078102e-05,
+      "loss": 3.0523,
+      "step": 13760
+    },
+    {
+      "epoch": 1.0130773050819402,
+      "grad_norm": 0.367825448513031,
+      "learning_rate": 7.665221952556461e-05,
+      "loss": 3.0198,
+      "step": 13770
+    },
+    {
+      "epoch": 1.0138130184480127,
+      "grad_norm": 0.35293030738830566,
+      "learning_rate": 7.661879833422896e-05,
+      "loss": 3.0526,
+      "step": 13780
+    },
+    {
+      "epoch": 1.0145487318140853,
+      "grad_norm": 0.3688339293003082,
+      "learning_rate": 7.658536053762281e-05,
+      "loss": 3.0749,
+      "step": 13790
+    },
+    {
+      "epoch": 1.0152844451801577,
+      "grad_norm": 0.3820160925388336,
+      "learning_rate": 7.655190615660523e-05,
+      "loss": 3.022,
+      "step": 13800
+    },
+    {
+      "epoch": 1.0160201585462303,
+      "grad_norm": 0.3556500971317291,
+      "learning_rate": 7.651843521204564e-05,
+      "loss": 3.037,
+      "step": 13810
+    },
+    {
+      "epoch": 1.0167558719123029,
+      "grad_norm": 0.35452452301979065,
+      "learning_rate": 7.648494772482383e-05,
+      "loss": 2.9885,
+      "step": 13820
+    },
+    {
+      "epoch": 1.0174915852783755,
+      "grad_norm": 0.35452884435653687,
+      "learning_rate": 7.645144371582982e-05,
+      "loss": 3.0852,
+      "step": 13830
+    },
+    {
+      "epoch": 1.018227298644448,
+      "grad_norm": 0.36470165848731995,
+      "learning_rate": 7.641792320596405e-05,
+      "loss": 3.0374,
+      "step": 13840
+    },
+    {
+      "epoch": 1.0189630120105206,
+      "grad_norm": 0.36535587906837463,
+      "learning_rate": 7.638438621613718e-05,
+      "loss": 3.0298,
+      "step": 13850
+    },
+    {
+      "epoch": 1.0196987253765932,
+      "grad_norm": 0.35986581444740295,
+      "learning_rate": 7.635083276727015e-05,
+      "loss": 3.0471,
+      "step": 13860
+    },
+    {
+      "epoch": 1.0204344387426658,
+      "grad_norm": 0.36018699407577515,
+      "learning_rate": 7.631726288029421e-05,
+      "loss": 3.029,
+      "step": 13870
+    },
+    {
+      "epoch": 1.0211701521087384,
+      "grad_norm": 0.3721524775028229,
+      "learning_rate": 7.628367657615085e-05,
+      "loss": 3.0347,
+      "step": 13880
+    },
+    {
+      "epoch": 1.021905865474811,
+      "grad_norm": 0.3706524968147278,
+      "learning_rate": 7.625007387579175e-05,
+      "loss": 3.047,
+      "step": 13890
+    },
+    {
+      "epoch": 1.0226415788408836,
+      "grad_norm": 0.37487688660621643,
+      "learning_rate": 7.621645480017891e-05,
+      "loss": 2.9755,
+      "step": 13900
+    },
+    {
+      "epoch": 1.0233772922069562,
+      "grad_norm": 0.3729473054409027,
+      "learning_rate": 7.618281937028447e-05,
+      "loss": 3.0473,
+      "step": 13910
+    },
+    {
+      "epoch": 1.0241130055730288,
+      "grad_norm": 0.3478502035140991,
+      "learning_rate": 7.614916760709078e-05,
+      "loss": 3.0307,
+      "step": 13920
+    },
+    {
+      "epoch": 1.0248487189391013,
+      "grad_norm": 0.36011430621147156,
+      "learning_rate": 7.611549953159045e-05,
+      "loss": 2.9693,
+      "step": 13930
+    },
+    {
+      "epoch": 1.025584432305174,
+      "grad_norm": 0.370433509349823,
+      "learning_rate": 7.608181516478616e-05,
+      "loss": 3.0424,
+      "step": 13940
+    },
+    {
+      "epoch": 1.0263201456712465,
+      "grad_norm": 0.3730204999446869,
+      "learning_rate": 7.604811452769081e-05,
+      "loss": 3.0138,
+      "step": 13950
+    },
+    {
+      "epoch": 1.027055859037319,
+      "grad_norm": 0.37531137466430664,
+      "learning_rate": 7.601439764132748e-05,
+      "loss": 3.0577,
+      "step": 13960
+    },
+    {
+      "epoch": 1.0277915724033917,
+      "grad_norm": 0.37572550773620605,
+      "learning_rate": 7.598066452672931e-05,
+      "loss": 3.0074,
+      "step": 13970
+    },
+    {
+      "epoch": 1.0285272857694643,
+      "grad_norm": 0.3774866759777069,
+      "learning_rate": 7.594691520493965e-05,
+      "loss": 3.0147,
+      "step": 13980
+    },
+    {
+      "epoch": 1.0292629991355369,
+      "grad_norm": 0.38011255860328674,
+      "learning_rate": 7.591314969701187e-05,
+      "loss": 3.0288,
+      "step": 13990
+    },
+    {
+      "epoch": 1.0299987125016095,
+      "grad_norm": 0.3587471544742584,
+      "learning_rate": 7.587936802400952e-05,
+      "loss": 3.0361,
+      "step": 14000
+    },
+    {
+      "epoch": 1.030734425867682,
+      "grad_norm": 0.38037359714508057,
+      "learning_rate": 7.584557020700618e-05,
+      "loss": 3.0553,
+      "step": 14010
+    },
+    {
+      "epoch": 1.0314701392337546,
+      "grad_norm": 0.3745957911014557,
+      "learning_rate": 7.581175626708552e-05,
+      "loss": 3.0321,
+      "step": 14020
+    },
+    {
+      "epoch": 1.032205852599827,
+      "grad_norm": 0.36815354228019714,
+      "learning_rate": 7.577792622534125e-05,
+      "loss": 3.0702,
+      "step": 14030
+    },
+    {
+      "epoch": 1.0329415659658996,
+      "grad_norm": 0.3649907112121582,
+      "learning_rate": 7.574408010287716e-05,
+      "loss": 3.0497,
+      "step": 14040
+    },
+    {
+      "epoch": 1.0336772793319722,
+      "grad_norm": 0.3590468168258667,
+      "learning_rate": 7.571021792080704e-05,
+      "loss": 3.0026,
+      "step": 14050
+    },
+    {
+      "epoch": 1.0344129926980448,
+      "grad_norm": 0.3700324594974518,
+      "learning_rate": 7.567633970025473e-05,
+      "loss": 3.0415,
+      "step": 14060
+    },
+    {
+      "epoch": 1.0351487060641174,
+      "grad_norm": 0.3878287374973297,
+      "learning_rate": 7.564244546235402e-05,
+      "loss": 3.0908,
+      "step": 14070
+    },
+    {
+      "epoch": 1.03588441943019,
+      "grad_norm": 0.3759715259075165,
+      "learning_rate": 7.560853522824875e-05,
+      "loss": 3.0188,
+      "step": 14080
+    },
+    {
+      "epoch": 1.0366201327962625,
+      "grad_norm": 0.35905465483665466,
+      "learning_rate": 7.557460901909273e-05,
+      "loss": 3.0391,
+      "step": 14090
+    },
+    {
+      "epoch": 1.0373558461623351,
+      "grad_norm": 0.38012492656707764,
+      "learning_rate": 7.554066685604967e-05,
+      "loss": 3.0243,
+      "step": 14100
+    },
+    {
+      "epoch": 1.0380915595284077,
+      "grad_norm": 0.37850454449653625,
+      "learning_rate": 7.550670876029333e-05,
+      "loss": 3.0917,
+      "step": 14110
+    },
+    {
+      "epoch": 1.0388272728944803,
+      "grad_norm": 0.35123690962791443,
+      "learning_rate": 7.547273475300734e-05,
+      "loss": 3.01,
+      "step": 14120
+    },
+    {
+      "epoch": 1.0395629862605529,
+      "grad_norm": 0.3715062141418457,
+      "learning_rate": 7.543874485538529e-05,
+      "loss": 3.0258,
+      "step": 14130
+    },
+    {
+      "epoch": 1.0402986996266255,
+      "grad_norm": 0.3719918131828308,
+      "learning_rate": 7.540473908863065e-05,
+      "loss": 2.9789,
+      "step": 14140
+    },
+    {
+      "epoch": 1.041034412992698,
+      "grad_norm": 0.37903308868408203,
+      "learning_rate": 7.537071747395687e-05,
+      "loss": 3.0532,
+      "step": 14150
+    },
+    {
+      "epoch": 1.0417701263587706,
+      "grad_norm": 0.3676205277442932,
+      "learning_rate": 7.533668003258714e-05,
+      "loss": 3.0311,
+      "step": 14160
+    },
+    {
+      "epoch": 1.0425058397248432,
+      "grad_norm": 0.3724045753479004,
+      "learning_rate": 7.530262678575466e-05,
+      "loss": 3.0186,
+      "step": 14170
+    },
+    {
+      "epoch": 1.0432415530909158,
+      "grad_norm": 0.4065383970737457,
+      "learning_rate": 7.526855775470245e-05,
+      "loss": 3.055,
+      "step": 14180
+    },
+    {
+      "epoch": 1.0439772664569884,
+      "grad_norm": 0.3902377784252167,
+      "learning_rate": 7.523447296068332e-05,
+      "loss": 3.0306,
+      "step": 14190
+    },
+    {
+      "epoch": 1.044712979823061,
+      "grad_norm": 0.3618945777416229,
+      "learning_rate": 7.520037242496e-05,
+      "loss": 3.0574,
+      "step": 14200
+    },
+    {
+      "epoch": 1.0454486931891336,
+      "grad_norm": 0.36113548278808594,
+      "learning_rate": 7.516625616880498e-05,
+      "loss": 3.0278,
+      "step": 14210
+    },
+    {
+      "epoch": 1.0461844065552062,
+      "grad_norm": 0.3701973259449005,
+      "learning_rate": 7.513212421350058e-05,
+      "loss": 3.0681,
+      "step": 14220
+    },
+    {
+      "epoch": 1.0469201199212788,
+      "grad_norm": 0.36408379673957825,
+      "learning_rate": 7.50979765803389e-05,
+      "loss": 3.0688,
+      "step": 14230
+    },
+    {
+      "epoch": 1.0476558332873513,
+      "grad_norm": 0.365655779838562,
+      "learning_rate": 7.506381329062184e-05,
+      "loss": 3.0378,
+      "step": 14240
+    },
+    {
+      "epoch": 1.048391546653424,
+      "grad_norm": 0.3665781021118164,
+      "learning_rate": 7.502963436566106e-05,
+      "loss": 2.9991,
+      "step": 14250
+    },
+    {
+      "epoch": 1.0491272600194963,
+      "grad_norm": 0.38295918703079224,
+      "learning_rate": 7.499543982677795e-05,
+      "loss": 3.1263,
+      "step": 14260
+    },
+    {
+      "epoch": 1.0498629733855689,
+      "grad_norm": 0.39406198263168335,
+      "learning_rate": 7.49612296953037e-05,
+      "loss": 3.0902,
+      "step": 14270
+    },
+    {
+      "epoch": 1.0505986867516415,
+      "grad_norm": 0.37145209312438965,
+      "learning_rate": 7.492700399257913e-05,
+      "loss": 3.0558,
+      "step": 14280
+    },
+    {
+      "epoch": 1.051334400117714,
+      "grad_norm": 0.44971105456352234,
+      "learning_rate": 7.489276273995487e-05,
+      "loss": 3.0603,
+      "step": 14290
+    },
+    {
+      "epoch": 1.0520701134837867,
+      "grad_norm": 0.3675541281700134,
+      "learning_rate": 7.485850595879117e-05,
+      "loss": 2.9996,
+      "step": 14300
+    },
+    {
+      "epoch": 1.0528058268498592,
+      "grad_norm": 0.36965155601501465,
+      "learning_rate": 7.482423367045808e-05,
+      "loss": 3.0562,
+      "step": 14310
+    },
+    {
+      "epoch": 1.0535415402159318,
+      "grad_norm": 0.37856119871139526,
+      "learning_rate": 7.478994589633518e-05,
+      "loss": 3.0229,
+      "step": 14320
+    },
+    {
+      "epoch": 1.0542772535820044,
+      "grad_norm": 0.3634575307369232,
+      "learning_rate": 7.475564265781183e-05,
+      "loss": 3.0108,
+      "step": 14330
+    },
+    {
+      "epoch": 1.055012966948077,
+      "grad_norm": 0.35185161232948303,
+      "learning_rate": 7.472132397628694e-05,
+      "loss": 3.0018,
+      "step": 14340
+    },
+    {
+      "epoch": 1.0557486803141496,
+      "grad_norm": 0.36539459228515625,
+      "learning_rate": 7.468698987316915e-05,
+      "loss": 3.079,
+      "step": 14350
+    },
+    {
+      "epoch": 1.0564843936802222,
+      "grad_norm": 0.36918023228645325,
+      "learning_rate": 7.465264036987662e-05,
+      "loss": 3.0428,
+      "step": 14360
+    },
+    {
+      "epoch": 1.0572201070462948,
+      "grad_norm": 0.37502649426460266,
+      "learning_rate": 7.461827548783724e-05,
+      "loss": 3.0479,
+      "step": 14370
+    },
+    {
+      "epoch": 1.0579558204123674,
+      "grad_norm": 0.3699197769165039,
+      "learning_rate": 7.458389524848835e-05,
+      "loss": 3.0427,
+      "step": 14380
+    },
+    {
+      "epoch": 1.05869153377844,
+      "grad_norm": 0.3672429025173187,
+      "learning_rate": 7.454949967327699e-05,
+      "loss": 3.1023,
+      "step": 14390
+    },
+    {
+      "epoch": 1.0594272471445125,
+      "grad_norm": 0.3570868968963623,
+      "learning_rate": 7.45150887836597e-05,
+      "loss": 3.0332,
+      "step": 14400
+    },
+    {
+      "epoch": 1.0601629605105851,
+      "grad_norm": 0.3757207691669464,
+      "learning_rate": 7.448066260110259e-05,
+      "loss": 3.0765,
+      "step": 14410
+    },
+    {
+      "epoch": 1.0608986738766577,
+      "grad_norm": 0.36892250180244446,
+      "learning_rate": 7.444622114708135e-05,
+      "loss": 3.06,
+      "step": 14420
+    },
+    {
+      "epoch": 1.0616343872427303,
+      "grad_norm": 0.36744457483291626,
+      "learning_rate": 7.44117644430811e-05,
+      "loss": 2.9722,
+      "step": 14430
+    },
+    {
+      "epoch": 1.0623701006088029,
+      "grad_norm": 0.3592512607574463,
+      "learning_rate": 7.437729251059656e-05,
+      "loss": 3.0678,
+      "step": 14440
+    },
+    {
+      "epoch": 1.0631058139748755,
+      "grad_norm": 0.3572598397731781,
+      "learning_rate": 7.434280537113194e-05,
+      "loss": 3.0471,
+      "step": 14450
+    },
+    {
+      "epoch": 1.063841527340948,
+      "grad_norm": 0.3860386908054352,
+      "learning_rate": 7.43083030462009e-05,
+      "loss": 3.0441,
+      "step": 14460
+    },
+    {
+      "epoch": 1.0645772407070206,
+      "grad_norm": 0.3672623634338379,
+      "learning_rate": 7.42737855573266e-05,
+      "loss": 3.027,
+      "step": 14470
+    },
+    {
+      "epoch": 1.065312954073093,
+      "grad_norm": 0.3730790615081787,
+      "learning_rate": 7.423925292604163e-05,
+      "loss": 3.0369,
+      "step": 14480
+    },
+    {
+      "epoch": 1.0660486674391656,
+      "grad_norm": 0.36241936683654785,
+      "learning_rate": 7.420470517388808e-05,
+      "loss": 3.0399,
+      "step": 14490
+    },
+    {
+      "epoch": 1.0667843808052382,
+      "grad_norm": 0.37162163853645325,
+      "learning_rate": 7.417014232241742e-05,
+      "loss": 3.0485,
+      "step": 14500
+    },
+    {
+      "epoch": 1.0675200941713108,
+      "grad_norm": 0.35315626859664917,
+      "learning_rate": 7.413556439319058e-05,
+      "loss": 3.0481,
+      "step": 14510
+    },
+    {
+      "epoch": 1.0682558075373834,
+      "grad_norm": 0.3689192533493042,
+      "learning_rate": 7.410097140777784e-05,
+      "loss": 3.0483,
+      "step": 14520
+    },
+    {
+      "epoch": 1.068991520903456,
+      "grad_norm": 0.36243700981140137,
+      "learning_rate": 7.406636338775894e-05,
+      "loss": 3.0528,
+      "step": 14530
+    },
+    {
+      "epoch": 1.0697272342695285,
+      "grad_norm": 0.3716164827346802,
+      "learning_rate": 7.403174035472295e-05,
+      "loss": 2.9902,
+      "step": 14540
+    },
+    {
+      "epoch": 1.0704629476356011,
+      "grad_norm": 0.3898087739944458,
+      "learning_rate": 7.399710233026831e-05,
+      "loss": 3.0672,
+      "step": 14550
+    },
+    {
+      "epoch": 1.0711986610016737,
+      "grad_norm": 0.3904206454753876,
+      "learning_rate": 7.396244933600285e-05,
+      "loss": 3.0234,
+      "step": 14560
+    },
+    {
+      "epoch": 1.0719343743677463,
+      "grad_norm": 0.35629862546920776,
+      "learning_rate": 7.392778139354369e-05,
+      "loss": 3.0476,
+      "step": 14570
+    },
+    {
+      "epoch": 1.0726700877338189,
+      "grad_norm": 0.37418025732040405,
+      "learning_rate": 7.389309852451733e-05,
+      "loss": 3.0902,
+      "step": 14580
+    },
+    {
+      "epoch": 1.0734058010998915,
+      "grad_norm": 0.3728621006011963,
+      "learning_rate": 7.38584007505595e-05,
+      "loss": 3.0105,
+      "step": 14590
+    },
+    {
+      "epoch": 1.074141514465964,
+      "grad_norm": 0.380960077047348,
+      "learning_rate": 7.382368809331532e-05,
+      "loss": 3.0307,
+      "step": 14600
+    },
+    {
+      "epoch": 1.0748772278320367,
+      "grad_norm": 0.36986520886421204,
+      "learning_rate": 7.378896057443912e-05,
+      "loss": 3.0468,
+      "step": 14610
+    },
+    {
+      "epoch": 1.0756129411981092,
+      "grad_norm": 0.36052417755126953,
+      "learning_rate": 7.375421821559455e-05,
+      "loss": 3.089,
+      "step": 14620
+    },
+    {
+      "epoch": 1.0763486545641818,
+      "grad_norm": 0.3690125644207001,
+      "learning_rate": 7.37194610384545e-05,
+      "loss": 3.0427,
+      "step": 14630
+    },
+    {
+      "epoch": 1.0770843679302544,
+      "grad_norm": 0.36218175292015076,
+      "learning_rate": 7.368468906470109e-05,
+      "loss": 3.0427,
+      "step": 14640
+    },
+    {
+      "epoch": 1.077820081296327,
+      "grad_norm": 0.373294472694397,
+      "learning_rate": 7.364990231602569e-05,
+      "loss": 2.9933,
+      "step": 14650
+    },
+    {
+      "epoch": 1.0785557946623996,
+      "grad_norm": 0.3562031090259552,
+      "learning_rate": 7.361510081412887e-05,
+      "loss": 3.0137,
+      "step": 14660
+    },
+    {
+      "epoch": 1.0792915080284722,
+      "grad_norm": 0.3716961443424225,
+      "learning_rate": 7.358028458072044e-05,
+      "loss": 3.0788,
+      "step": 14670
+    },
+    {
+      "epoch": 1.0800272213945448,
+      "grad_norm": 0.3744214177131653,
+      "learning_rate": 7.354545363751934e-05,
+      "loss": 3.0747,
+      "step": 14680
+    },
+    {
+      "epoch": 1.0807629347606174,
+      "grad_norm": 0.3793715834617615,
+      "learning_rate": 7.351060800625376e-05,
+      "loss": 3.0393,
+      "step": 14690
+    },
+    {
+      "epoch": 1.08149864812669,
+      "grad_norm": 0.3758842349052429,
+      "learning_rate": 7.347574770866099e-05,
+      "loss": 3.0626,
+      "step": 14700
+    },
+    {
+      "epoch": 1.0822343614927625,
+      "grad_norm": 0.3611127436161041,
+      "learning_rate": 7.344087276648747e-05,
+      "loss": 3.0152,
+      "step": 14710
+    },
+    {
+      "epoch": 1.082970074858835,
+      "grad_norm": 0.35528436303138733,
+      "learning_rate": 7.340598320148883e-05,
+      "loss": 3.0353,
+      "step": 14720
+    },
+    {
+      "epoch": 1.0837057882249075,
+      "grad_norm": 0.3713771402835846,
+      "learning_rate": 7.337107903542978e-05,
+      "loss": 3.0466,
+      "step": 14730
+    },
+    {
+      "epoch": 1.08444150159098,
+      "grad_norm": 0.3739469647407532,
+      "learning_rate": 7.333616029008415e-05,
+      "loss": 3.0327,
+      "step": 14740
+    },
+    {
+      "epoch": 1.0851772149570527,
+      "grad_norm": 0.3754708766937256,
+      "learning_rate": 7.330122698723484e-05,
+      "loss": 3.0366,
+      "step": 14750
+    },
+    {
+      "epoch": 1.0859129283231252,
+      "grad_norm": 0.3673410415649414,
+      "learning_rate": 7.326627914867388e-05,
+      "loss": 3.0699,
+      "step": 14760
+    },
+    {
+      "epoch": 1.0866486416891978,
+      "grad_norm": 0.3647342324256897,
+      "learning_rate": 7.323131679620233e-05,
+      "loss": 3.0106,
+      "step": 14770
+    },
+    {
+      "epoch": 1.0873843550552704,
+      "grad_norm": 0.3754418194293976,
+      "learning_rate": 7.31963399516303e-05,
+      "loss": 3.0504,
+      "step": 14780
+    },
+    {
+      "epoch": 1.088120068421343,
+      "grad_norm": 0.3695102632045746,
+      "learning_rate": 7.3161348636777e-05,
+      "loss": 3.0981,
+      "step": 14790
+    },
+    {
+      "epoch": 1.0888557817874156,
+      "grad_norm": 0.38330012559890747,
+      "learning_rate": 7.312634287347057e-05,
+      "loss": 3.0748,
+      "step": 14800
+    },
+    {
+      "epoch": 1.0895914951534882,
+      "grad_norm": 0.36199986934661865,
+      "learning_rate": 7.309132268354825e-05,
+      "loss": 3.0233,
+      "step": 14810
+    },
+    {
+      "epoch": 1.0903272085195608,
+      "grad_norm": 0.3592955470085144,
+      "learning_rate": 7.305628808885624e-05,
+      "loss": 2.9987,
+      "step": 14820
+    },
+    {
+      "epoch": 1.0910629218856334,
+      "grad_norm": 0.3516167998313904,
+      "learning_rate": 7.302123911124973e-05,
+      "loss": 3.0522,
+      "step": 14830
+    },
+    {
+      "epoch": 1.091798635251706,
+      "grad_norm": 0.3621143698692322,
+      "learning_rate": 7.298617577259286e-05,
+      "loss": 3.0629,
+      "step": 14840
+    },
+    {
+      "epoch": 1.0925343486177785,
+      "grad_norm": 0.37074902653694153,
+      "learning_rate": 7.295109809475877e-05,
+      "loss": 3.0556,
+      "step": 14850
+    },
+    {
+      "epoch": 1.0932700619838511,
+      "grad_norm": 0.3676816523075104,
+      "learning_rate": 7.291600609962954e-05,
+      "loss": 3.0331,
+      "step": 14860
+    },
+    {
+      "epoch": 1.0940057753499237,
+      "grad_norm": 0.36239302158355713,
+      "learning_rate": 7.288089980909616e-05,
+      "loss": 3.0383,
+      "step": 14870
+    },
+    {
+      "epoch": 1.0947414887159963,
+      "grad_norm": 0.36774101853370667,
+      "learning_rate": 7.284577924505853e-05,
+      "loss": 3.0739,
+      "step": 14880
+    },
+    {
+      "epoch": 1.095477202082069,
+      "grad_norm": 0.36842718720436096,
+      "learning_rate": 7.28106444294255e-05,
+      "loss": 3.043,
+      "step": 14890
+    },
+    {
+      "epoch": 1.0962129154481415,
+      "grad_norm": 0.34536731243133545,
+      "learning_rate": 7.277549538411474e-05,
+      "loss": 3.0069,
+      "step": 14900
+    },
+    {
+      "epoch": 1.096948628814214,
+      "grad_norm": 0.36958837509155273,
+      "learning_rate": 7.274033213105287e-05,
+      "loss": 3.0424,
+      "step": 14910
+    },
+    {
+      "epoch": 1.0976843421802867,
+      "grad_norm": 0.35424816608428955,
+      "learning_rate": 7.270515469217533e-05,
+      "loss": 3.0004,
+      "step": 14920
+    },
+    {
+      "epoch": 1.0984200555463592,
+      "grad_norm": 0.3760276734828949,
+      "learning_rate": 7.266996308942642e-05,
+      "loss": 3.0138,
+      "step": 14930
+    },
+    {
+      "epoch": 1.0991557689124316,
+      "grad_norm": 0.3770487308502197,
+      "learning_rate": 7.263475734475929e-05,
+      "loss": 3.0617,
+      "step": 14940
+    },
+    {
+      "epoch": 1.0998914822785042,
+      "grad_norm": 0.36666733026504517,
+      "learning_rate": 7.259953748013587e-05,
+      "loss": 2.9987,
+      "step": 14950
+    },
+    {
+      "epoch": 1.1006271956445768,
+      "grad_norm": 0.37574532628059387,
+      "learning_rate": 7.256430351752694e-05,
+      "loss": 3.0236,
+      "step": 14960
+    },
+    {
+      "epoch": 1.1013629090106494,
+      "grad_norm": 0.3686668276786804,
+      "learning_rate": 7.252905547891206e-05,
+      "loss": 2.9803,
+      "step": 14970
+    },
+    {
+      "epoch": 1.102098622376722,
+      "grad_norm": 0.3498009741306305,
+      "learning_rate": 7.249379338627956e-05,
+      "loss": 3.0089,
+      "step": 14980
+    },
+    {
+      "epoch": 1.1028343357427945,
+      "grad_norm": 0.3669489920139313,
+      "learning_rate": 7.245851726162657e-05,
+      "loss": 3.007,
+      "step": 14990
+    },
+    {
+      "epoch": 1.1035700491088671,
+      "grad_norm": 0.3778036832809448,
+      "learning_rate": 7.242322712695894e-05,
+      "loss": 3.0263,
+      "step": 15000
+    },
+    {
+      "epoch": 1.1043057624749397,
+      "grad_norm": 0.3754088878631592,
+      "learning_rate": 7.238792300429128e-05,
+      "loss": 3.1054,
+      "step": 15010
+    },
+    {
+      "epoch": 1.1050414758410123,
+      "grad_norm": 0.3669554591178894,
+      "learning_rate": 7.235260491564689e-05,
+      "loss": 3.0111,
+      "step": 15020
+    },
+    {
+      "epoch": 1.105777189207085,
+      "grad_norm": 0.3635033667087555,
+      "learning_rate": 7.231727288305783e-05,
+      "loss": 3.0376,
+      "step": 15030
+    },
+    {
+      "epoch": 1.1065129025731575,
+      "grad_norm": 0.3695336580276489,
+      "learning_rate": 7.228192692856484e-05,
+      "loss": 3.0426,
+      "step": 15040
+    },
+    {
+      "epoch": 1.10724861593923,
+      "grad_norm": 0.37608885765075684,
+      "learning_rate": 7.224656707421735e-05,
+      "loss": 3.0542,
+      "step": 15050
+    },
+    {
+      "epoch": 1.1079843293053027,
+      "grad_norm": 0.35501381754875183,
+      "learning_rate": 7.221119334207342e-05,
+      "loss": 3.0561,
+      "step": 15060
+    },
+    {
+      "epoch": 1.1087200426713752,
+      "grad_norm": 0.37130206823349,
+      "learning_rate": 7.217580575419983e-05,
+      "loss": 3.0336,
+      "step": 15070
+    },
+    {
+      "epoch": 1.1094557560374478,
+      "grad_norm": 0.3685947060585022,
+      "learning_rate": 7.214040433267198e-05,
+      "loss": 3.0071,
+      "step": 15080
+    },
+    {
+      "epoch": 1.1101914694035204,
+      "grad_norm": 0.3730693757534027,
+      "learning_rate": 7.210498909957388e-05,
+      "loss": 3.0587,
+      "step": 15090
+    },
+    {
+      "epoch": 1.110927182769593,
+      "grad_norm": 0.3776399493217468,
+      "learning_rate": 7.206956007699816e-05,
+      "loss": 3.0845,
+      "step": 15100
+    },
+    {
+      "epoch": 1.1116628961356656,
+      "grad_norm": 0.36200079321861267,
+      "learning_rate": 7.203411728704606e-05,
+      "loss": 3.007,
+      "step": 15110
+    },
+    {
+      "epoch": 1.1123986095017382,
+      "grad_norm": 0.3622397780418396,
+      "learning_rate": 7.199866075182745e-05,
+      "loss": 3.0645,
+      "step": 15120
+    },
+    {
+      "epoch": 1.1131343228678108,
+      "grad_norm": 0.3798138499259949,
+      "learning_rate": 7.19631904934607e-05,
+      "loss": 3.028,
+      "step": 15130
+    },
+    {
+      "epoch": 1.1138700362338834,
+      "grad_norm": 0.35861262679100037,
+      "learning_rate": 7.192770653407277e-05,
+      "loss": 3.036,
+      "step": 15140
+    },
+    {
+      "epoch": 1.114605749599956,
+      "grad_norm": 0.4072038531303406,
+      "learning_rate": 7.189220889579921e-05,
+      "loss": 3.0289,
+      "step": 15150
+    },
+    {
+      "epoch": 1.1153414629660285,
+      "grad_norm": 0.3686382472515106,
+      "learning_rate": 7.185669760078403e-05,
+      "loss": 3.0469,
+      "step": 15160
+    },
+    {
+      "epoch": 1.1160771763321011,
+      "grad_norm": 0.3779941499233246,
+      "learning_rate": 7.182117267117982e-05,
+      "loss": 3.0423,
+      "step": 15170
+    },
+    {
+      "epoch": 1.1168128896981735,
+      "grad_norm": 0.36711934208869934,
+      "learning_rate": 7.178563412914762e-05,
+      "loss": 2.9819,
+      "step": 15180
+    },
+    {
+      "epoch": 1.117548603064246,
+      "grad_norm": 0.36863258481025696,
+      "learning_rate": 7.175008199685703e-05,
+      "loss": 3.0924,
+      "step": 15190
+    },
+    {
+      "epoch": 1.1182843164303187,
+      "grad_norm": 0.36387646198272705,
+      "learning_rate": 7.171451629648604e-05,
+      "loss": 3.0725,
+      "step": 15200
+    },
+    {
+      "epoch": 1.1190200297963913,
+      "grad_norm": 0.3569769561290741,
+      "learning_rate": 7.16789370502212e-05,
+      "loss": 3.019,
+      "step": 15210
+    },
+    {
+      "epoch": 1.1197557431624638,
+      "grad_norm": 0.35440513491630554,
+      "learning_rate": 7.164334428025745e-05,
+      "loss": 2.9845,
+      "step": 15220
+    },
+    {
+      "epoch": 1.1204914565285364,
+      "grad_norm": 0.3555620610713959,
+      "learning_rate": 7.160773800879818e-05,
+      "loss": 3.0566,
+      "step": 15230
+    },
+    {
+      "epoch": 1.121227169894609,
+      "grad_norm": 0.3573719263076782,
+      "learning_rate": 7.157211825805519e-05,
+      "loss": 3.0455,
+      "step": 15240
+    },
+    {
+      "epoch": 1.1219628832606816,
+      "grad_norm": 0.3558180332183838,
+      "learning_rate": 7.153648505024872e-05,
+      "loss": 3.0277,
+      "step": 15250
+    },
+    {
+      "epoch": 1.1226985966267542,
+      "grad_norm": 0.35550549626350403,
+      "learning_rate": 7.150083840760738e-05,
+      "loss": 3.0278,
+      "step": 15260
+    },
+    {
+      "epoch": 1.1234343099928268,
+      "grad_norm": 0.3632723391056061,
+      "learning_rate": 7.146517835236817e-05,
+      "loss": 2.9995,
+      "step": 15270
+    },
+    {
+      "epoch": 1.1241700233588994,
+      "grad_norm": 0.3601170480251312,
+      "learning_rate": 7.142950490677645e-05,
+      "loss": 3.0567,
+      "step": 15280
+    },
+    {
+      "epoch": 1.124905736724972,
+      "grad_norm": 0.3749910295009613,
+      "learning_rate": 7.139381809308592e-05,
+      "loss": 3.0666,
+      "step": 15290
+    },
+    {
+      "epoch": 1.1256414500910445,
+      "grad_norm": 0.3659000098705292,
+      "learning_rate": 7.135811793355869e-05,
+      "loss": 3.0789,
+      "step": 15300
+    },
+    {
+      "epoch": 1.1263771634571171,
+      "grad_norm": 0.36209380626678467,
+      "learning_rate": 7.132240445046507e-05,
+      "loss": 3.0149,
+      "step": 15310
+    },
+    {
+      "epoch": 1.1271128768231897,
+      "grad_norm": 0.36656835675239563,
+      "learning_rate": 7.12866776660838e-05,
+      "loss": 3.1049,
+      "step": 15320
+    },
+    {
+      "epoch": 1.1278485901892623,
+      "grad_norm": 0.37641996145248413,
+      "learning_rate": 7.125093760270187e-05,
+      "loss": 3.0067,
+      "step": 15330
+    },
+    {
+      "epoch": 1.128584303555335,
+      "grad_norm": 0.35203486680984497,
+      "learning_rate": 7.121518428261456e-05,
+      "loss": 3.0463,
+      "step": 15340
+    },
+    {
+      "epoch": 1.1293200169214075,
+      "grad_norm": 0.3685363829135895,
+      "learning_rate": 7.117941772812537e-05,
+      "loss": 3.014,
+      "step": 15350
+    },
+    {
+      "epoch": 1.13005573028748,
+      "grad_norm": 0.35518285632133484,
+      "learning_rate": 7.114363796154615e-05,
+      "loss": 3.0378,
+      "step": 15360
+    },
+    {
+      "epoch": 1.1307914436535527,
+      "grad_norm": 0.3588918447494507,
+      "learning_rate": 7.110784500519691e-05,
+      "loss": 3.0122,
+      "step": 15370
+    },
+    {
+      "epoch": 1.131527157019625,
+      "grad_norm": 0.37091171741485596,
+      "learning_rate": 7.107203888140595e-05,
+      "loss": 3.0682,
+      "step": 15380
+    },
+    {
+      "epoch": 1.1322628703856976,
+      "grad_norm": 0.36039799451828003,
+      "learning_rate": 7.103621961250971e-05,
+      "loss": 3.0662,
+      "step": 15390
+    },
+    {
+      "epoch": 1.1329985837517702,
+      "grad_norm": 0.37948399782180786,
+      "learning_rate": 7.100038722085292e-05,
+      "loss": 3.0323,
+      "step": 15400
+    },
+    {
+      "epoch": 1.1337342971178428,
+      "grad_norm": 0.3666056990623474,
+      "learning_rate": 7.096454172878844e-05,
+      "loss": 3.0779,
+      "step": 15410
+    },
+    {
+      "epoch": 1.1344700104839154,
+      "grad_norm": 0.354842871427536,
+      "learning_rate": 7.092868315867729e-05,
+      "loss": 3.0105,
+      "step": 15420
+    },
+    {
+      "epoch": 1.135205723849988,
+      "grad_norm": 0.35736706852912903,
+      "learning_rate": 7.089281153288868e-05,
+      "loss": 3.0129,
+      "step": 15430
+    },
+    {
+      "epoch": 1.1359414372160606,
+      "grad_norm": 0.36379727721214294,
+      "learning_rate": 7.085692687379997e-05,
+      "loss": 3.0868,
+      "step": 15440
+    },
+    {
+      "epoch": 1.1366771505821331,
+      "grad_norm": 0.3604990541934967,
+      "learning_rate": 7.082102920379664e-05,
+      "loss": 3.0705,
+      "step": 15450
+    },
+    {
+      "epoch": 1.1374128639482057,
+      "grad_norm": 0.3602616488933563,
+      "learning_rate": 7.078511854527227e-05,
+      "loss": 3.0599,
+      "step": 15460
+    },
+    {
+      "epoch": 1.1381485773142783,
+      "grad_norm": 0.3518899083137512,
+      "learning_rate": 7.074919492062853e-05,
+      "loss": 3.0322,
+      "step": 15470
+    },
+    {
+      "epoch": 1.138884290680351,
+      "grad_norm": 0.3573099970817566,
+      "learning_rate": 7.071325835227525e-05,
+      "loss": 3.0194,
+      "step": 15480
+    },
+    {
+      "epoch": 1.1396200040464235,
+      "grad_norm": 0.368889182806015,
+      "learning_rate": 7.067730886263027e-05,
+      "loss": 3.0894,
+      "step": 15490
+    },
+    {
+      "epoch": 1.140355717412496,
+      "grad_norm": 0.360470175743103,
+      "learning_rate": 7.06413464741195e-05,
+      "loss": 2.9901,
+      "step": 15500
+    },
+    {
+      "epoch": 1.1410914307785687,
+      "grad_norm": 0.3495664894580841,
+      "learning_rate": 7.06053712091769e-05,
+      "loss": 3.0086,
+      "step": 15510
+    },
+    {
+      "epoch": 1.1418271441446413,
+      "grad_norm": 0.3800434172153473,
+      "learning_rate": 7.056938309024449e-05,
+      "loss": 3.0825,
+      "step": 15520
+    },
+    {
+      "epoch": 1.1425628575107138,
+      "grad_norm": 0.36110514402389526,
+      "learning_rate": 7.053338213977227e-05,
+      "loss": 3.0407,
+      "step": 15530
+    },
+    {
+      "epoch": 1.1432985708767864,
+      "grad_norm": 0.3506874144077301,
+      "learning_rate": 7.049736838021826e-05,
+      "loss": 3.0438,
+      "step": 15540
+    },
+    {
+      "epoch": 1.144034284242859,
+      "grad_norm": 0.3586377203464508,
+      "learning_rate": 7.046134183404848e-05,
+      "loss": 2.9803,
+      "step": 15550
+    },
+    {
+      "epoch": 1.1447699976089316,
+      "grad_norm": 0.39292410016059875,
+      "learning_rate": 7.04253025237369e-05,
+      "loss": 3.1149,
+      "step": 15560
+    },
+    {
+      "epoch": 1.1455057109750042,
+      "grad_norm": 0.36179158091545105,
+      "learning_rate": 7.03892504717655e-05,
+      "loss": 3.0369,
+      "step": 15570
+    },
+    {
+      "epoch": 1.1462414243410768,
+      "grad_norm": 0.3586963415145874,
+      "learning_rate": 7.035318570062416e-05,
+      "loss": 3.0941,
+      "step": 15580
+    },
+    {
+      "epoch": 1.1469771377071494,
+      "grad_norm": 0.37060362100601196,
+      "learning_rate": 7.031710823281072e-05,
+      "loss": 3.0252,
+      "step": 15590
+    },
+    {
+      "epoch": 1.147712851073222,
+      "grad_norm": 0.3765799105167389,
+      "learning_rate": 7.02810180908309e-05,
+      "loss": 3.0231,
+      "step": 15600
+    },
+    {
+      "epoch": 1.1484485644392946,
+      "grad_norm": 0.35655713081359863,
+      "learning_rate": 7.024491529719841e-05,
+      "loss": 3.0632,
+      "step": 15610
+    },
+    {
+      "epoch": 1.1491842778053671,
+      "grad_norm": 0.3502100706100464,
+      "learning_rate": 7.020879987443478e-05,
+      "loss": 3.0465,
+      "step": 15620
+    },
+    {
+      "epoch": 1.1499199911714397,
+      "grad_norm": 0.36901360750198364,
+      "learning_rate": 7.017267184506944e-05,
+      "loss": 3.0028,
+      "step": 15630
+    },
+    {
+      "epoch": 1.150655704537512,
+      "grad_norm": 0.36550623178482056,
+      "learning_rate": 7.013653123163969e-05,
+      "loss": 3.0151,
+      "step": 15640
+    },
+    {
+      "epoch": 1.1513914179035847,
+      "grad_norm": 0.3524268865585327,
+      "learning_rate": 7.010037805669066e-05,
+      "loss": 3.0887,
+      "step": 15650
+    },
+    {
+      "epoch": 1.1521271312696573,
+      "grad_norm": 0.36080747842788696,
+      "learning_rate": 7.006421234277535e-05,
+      "loss": 3.0073,
+      "step": 15660
+    },
+    {
+      "epoch": 1.1528628446357299,
+      "grad_norm": 0.35583287477493286,
+      "learning_rate": 7.002803411245453e-05,
+      "loss": 3.0403,
+      "step": 15670
+    },
+    {
+      "epoch": 1.1535985580018024,
+      "grad_norm": 0.35857105255126953,
+      "learning_rate": 6.999184338829685e-05,
+      "loss": 3.0916,
+      "step": 15680
+    },
+    {
+      "epoch": 1.154334271367875,
+      "grad_norm": 0.355186402797699,
+      "learning_rate": 6.995564019287871e-05,
+      "loss": 3.0287,
+      "step": 15690
+    },
+    {
+      "epoch": 1.1550699847339476,
+      "grad_norm": 0.35449060797691345,
+      "learning_rate": 6.991942454878429e-05,
+      "loss": 3.0322,
+      "step": 15700
+    },
+    {
+      "epoch": 1.1558056981000202,
+      "grad_norm": 0.36941638588905334,
+      "learning_rate": 6.98831964786055e-05,
+      "loss": 2.9872,
+      "step": 15710
+    },
+    {
+      "epoch": 1.1565414114660928,
+      "grad_norm": 0.3739472031593323,
+      "learning_rate": 6.98469560049421e-05,
+      "loss": 3.0461,
+      "step": 15720
+    },
+    {
+      "epoch": 1.1572771248321654,
+      "grad_norm": 0.36685386300086975,
+      "learning_rate": 6.981070315040151e-05,
+      "loss": 3.045,
+      "step": 15730
+    },
+    {
+      "epoch": 1.158012838198238,
+      "grad_norm": 0.35560208559036255,
+      "learning_rate": 6.977443793759886e-05,
+      "loss": 3.0352,
+      "step": 15740
+    },
+    {
+      "epoch": 1.1587485515643106,
+      "grad_norm": 0.365713894367218,
+      "learning_rate": 6.973816038915708e-05,
+      "loss": 3.0403,
+      "step": 15750
+    },
+    {
+      "epoch": 1.1594842649303831,
+      "grad_norm": 0.3721645176410675,
+      "learning_rate": 6.97018705277067e-05,
+      "loss": 3.0351,
+      "step": 15760
+    },
+    {
+      "epoch": 1.1602199782964557,
+      "grad_norm": 0.35145121812820435,
+      "learning_rate": 6.966556837588597e-05,
+      "loss": 3.0252,
+      "step": 15770
+    },
+    {
+      "epoch": 1.1609556916625283,
+      "grad_norm": 0.35836660861968994,
+      "learning_rate": 6.962925395634081e-05,
+      "loss": 3.0328,
+      "step": 15780
+    },
+    {
+      "epoch": 1.161691405028601,
+      "grad_norm": 0.35599228739738464,
+      "learning_rate": 6.959292729172479e-05,
+      "loss": 2.9948,
+      "step": 15790
+    },
+    {
+      "epoch": 1.1624271183946735,
+      "grad_norm": 0.36521443724632263,
+      "learning_rate": 6.955658840469912e-05,
+      "loss": 2.9992,
+      "step": 15800
+    },
+    {
+      "epoch": 1.163162831760746,
+      "grad_norm": 0.34881696105003357,
+      "learning_rate": 6.952023731793262e-05,
+      "loss": 3.0316,
+      "step": 15810
+    },
+    {
+      "epoch": 1.1638985451268187,
+      "grad_norm": 0.3558899760246277,
+      "learning_rate": 6.948387405410175e-05,
+      "loss": 3.0121,
+      "step": 15820
+    },
+    {
+      "epoch": 1.1646342584928913,
+      "grad_norm": 0.34750762581825256,
+      "learning_rate": 6.944749863589053e-05,
+      "loss": 3.0151,
+      "step": 15830
+    },
+    {
+      "epoch": 1.1653699718589636,
+      "grad_norm": 0.3648332953453064,
+      "learning_rate": 6.941111108599058e-05,
+      "loss": 3.1025,
+      "step": 15840
+    },
+    {
+      "epoch": 1.1661056852250362,
+      "grad_norm": 0.35662853717803955,
+      "learning_rate": 6.93747114271011e-05,
+      "loss": 3.0067,
+      "step": 15850
+    },
+    {
+      "epoch": 1.1668413985911088,
+      "grad_norm": 0.36678966879844666,
+      "learning_rate": 6.933829968192883e-05,
+      "loss": 3.0341,
+      "step": 15860
+    },
+    {
+      "epoch": 1.1675771119571814,
+      "grad_norm": 0.35157790780067444,
+      "learning_rate": 6.930187587318804e-05,
+      "loss": 3.0261,
+      "step": 15870
+    },
+    {
+      "epoch": 1.168312825323254,
+      "grad_norm": 0.37112855911254883,
+      "learning_rate": 6.926544002360055e-05,
+      "loss": 2.9995,
+      "step": 15880
+    },
+    {
+      "epoch": 1.1690485386893266,
+      "grad_norm": 0.36144548654556274,
+      "learning_rate": 6.922899215589568e-05,
+      "loss": 3.0327,
+      "step": 15890
+    },
+    {
+      "epoch": 1.1697842520553992,
+      "grad_norm": 0.3435918986797333,
+      "learning_rate": 6.919253229281024e-05,
+      "loss": 3.0582,
+      "step": 15900
+    },
+    {
+      "epoch": 1.1705199654214717,
+      "grad_norm": 0.36248284578323364,
+      "learning_rate": 6.915606045708851e-05,
+      "loss": 3.0589,
+      "step": 15910
+    },
+    {
+      "epoch": 1.1712556787875443,
+      "grad_norm": 0.3698318600654602,
+      "learning_rate": 6.91195766714823e-05,
+      "loss": 3.056,
+      "step": 15920
+    },
+    {
+      "epoch": 1.171991392153617,
+      "grad_norm": 0.35921046137809753,
+      "learning_rate": 6.908308095875077e-05,
+      "loss": 3.0095,
+      "step": 15930
+    },
+    {
+      "epoch": 1.1727271055196895,
+      "grad_norm": 0.34976983070373535,
+      "learning_rate": 6.904657334166064e-05,
+      "loss": 3.0407,
+      "step": 15940
+    },
+    {
+      "epoch": 1.173462818885762,
+      "grad_norm": 0.36711758375167847,
+      "learning_rate": 6.901005384298598e-05,
+      "loss": 3.0909,
+      "step": 15950
+    },
+    {
+      "epoch": 1.1741985322518347,
+      "grad_norm": 0.37532252073287964,
+      "learning_rate": 6.897352248550827e-05,
+      "loss": 3.0742,
+      "step": 15960
+    },
+    {
+      "epoch": 1.1749342456179073,
+      "grad_norm": 0.36899441480636597,
+      "learning_rate": 6.893697929201643e-05,
+      "loss": 2.9883,
+      "step": 15970
+    },
+    {
+      "epoch": 1.1756699589839799,
+      "grad_norm": 0.37338241934776306,
+      "learning_rate": 6.890042428530674e-05,
+      "loss": 3.0123,
+      "step": 15980
+    },
+    {
+      "epoch": 1.1764056723500524,
+      "grad_norm": 0.38112708926200867,
+      "learning_rate": 6.886385748818284e-05,
+      "loss": 3.0735,
+      "step": 15990
+    },
+    {
+      "epoch": 1.177141385716125,
+      "grad_norm": 0.36849597096443176,
+      "learning_rate": 6.882727892345575e-05,
+      "loss": 2.997,
+      "step": 16000
+    },
+    {
+      "epoch": 1.1778770990821976,
+      "grad_norm": 0.37463846802711487,
+      "learning_rate": 6.879068861394379e-05,
+      "loss": 3.0676,
+      "step": 16010
+    },
+    {
+      "epoch": 1.1786128124482702,
+      "grad_norm": 0.3590144217014313,
+      "learning_rate": 6.875408658247266e-05,
+      "loss": 3.0571,
+      "step": 16020
+    },
+    {
+      "epoch": 1.1793485258143428,
+      "grad_norm": 0.361030638217926,
+      "learning_rate": 6.871747285187533e-05,
+      "loss": 3.0522,
+      "step": 16030
+    },
+    {
+      "epoch": 1.1800842391804154,
+      "grad_norm": 0.3563062250614166,
+      "learning_rate": 6.86808474449921e-05,
+      "loss": 3.0107,
+      "step": 16040
+    },
+    {
+      "epoch": 1.180819952546488,
+      "grad_norm": 0.3560244143009186,
+      "learning_rate": 6.864421038467052e-05,
+      "loss": 3.0247,
+      "step": 16050
+    },
+    {
+      "epoch": 1.1815556659125606,
+      "grad_norm": 0.3522135019302368,
+      "learning_rate": 6.860756169376544e-05,
+      "loss": 3.0314,
+      "step": 16060
+    },
+    {
+      "epoch": 1.1822913792786331,
+      "grad_norm": 0.37250736355781555,
+      "learning_rate": 6.857090139513893e-05,
+      "loss": 3.0329,
+      "step": 16070
+    },
+    {
+      "epoch": 1.1830270926447057,
+      "grad_norm": 0.3519201874732971,
+      "learning_rate": 6.853422951166035e-05,
+      "loss": 3.0617,
+      "step": 16080
+    },
+    {
+      "epoch": 1.1837628060107783,
+      "grad_norm": 0.3638618588447571,
+      "learning_rate": 6.849754606620624e-05,
+      "loss": 2.9948,
+      "step": 16090
+    },
+    {
+      "epoch": 1.1844985193768507,
+      "grad_norm": 0.3536449074745178,
+      "learning_rate": 6.846085108166039e-05,
+      "loss": 3.0703,
+      "step": 16100
+    },
+    {
+      "epoch": 1.1852342327429233,
+      "grad_norm": 0.3603556156158447,
+      "learning_rate": 6.842414458091378e-05,
+      "loss": 3.0363,
+      "step": 16110
+    },
+    {
+      "epoch": 1.1859699461089959,
+      "grad_norm": 0.3605019748210907,
+      "learning_rate": 6.838742658686453e-05,
+      "loss": 3.0193,
+      "step": 16120
+    },
+    {
+      "epoch": 1.1867056594750685,
+      "grad_norm": 0.3593468964099884,
+      "learning_rate": 6.835069712241797e-05,
+      "loss": 3.0101,
+      "step": 16130
+    },
+    {
+      "epoch": 1.187441372841141,
+      "grad_norm": 0.3509270250797272,
+      "learning_rate": 6.831395621048661e-05,
+      "loss": 3.024,
+      "step": 16140
+    },
+    {
+      "epoch": 1.1881770862072136,
+      "grad_norm": 0.36826565861701965,
+      "learning_rate": 6.827720387399007e-05,
+      "loss": 3.0504,
+      "step": 16150
+    },
+    {
+      "epoch": 1.1889127995732862,
+      "grad_norm": 0.35772114992141724,
+      "learning_rate": 6.824044013585506e-05,
+      "loss": 3.0697,
+      "step": 16160
+    },
+    {
+      "epoch": 1.1896485129393588,
+      "grad_norm": 0.3500980734825134,
+      "learning_rate": 6.820366501901547e-05,
+      "loss": 3.0154,
+      "step": 16170
+    },
+    {
+      "epoch": 1.1903842263054314,
+      "grad_norm": 0.36379170417785645,
+      "learning_rate": 6.816687854641227e-05,
+      "loss": 2.9959,
+      "step": 16180
+    },
+    {
+      "epoch": 1.191119939671504,
+      "grad_norm": 0.3574915826320648,
+      "learning_rate": 6.813008074099347e-05,
+      "loss": 3.0467,
+      "step": 16190
+    },
+    {
+      "epoch": 1.1918556530375766,
+      "grad_norm": 0.3617134988307953,
+      "learning_rate": 6.809327162571422e-05,
+      "loss": 3.0401,
+      "step": 16200
+    },
+    {
+      "epoch": 1.1925913664036492,
+      "grad_norm": 0.3473548889160156,
+      "learning_rate": 6.805645122353663e-05,
+      "loss": 3.0316,
+      "step": 16210
+    },
+    {
+      "epoch": 1.1933270797697217,
+      "grad_norm": 0.37072911858558655,
+      "learning_rate": 6.801961955742996e-05,
+      "loss": 3.0271,
+      "step": 16220
+    },
+    {
+      "epoch": 1.1940627931357943,
+      "grad_norm": 0.37436825037002563,
+      "learning_rate": 6.798277665037045e-05,
+      "loss": 3.0106,
+      "step": 16230
+    },
+    {
+      "epoch": 1.194798506501867,
+      "grad_norm": 0.35086673498153687,
+      "learning_rate": 6.794592252534132e-05,
+      "loss": 3.0915,
+      "step": 16240
+    },
+    {
+      "epoch": 1.1955342198679395,
+      "grad_norm": 0.3697086572647095,
+      "learning_rate": 6.790905720533283e-05,
+      "loss": 3.0405,
+      "step": 16250
+    },
+    {
+      "epoch": 1.196269933234012,
+      "grad_norm": 0.36851561069488525,
+      "learning_rate": 6.787218071334218e-05,
+      "loss": 3.0237,
+      "step": 16260
+    },
+    {
+      "epoch": 1.1970056466000847,
+      "grad_norm": 0.3767666220664978,
+      "learning_rate": 6.783529307237362e-05,
+      "loss": 3.0016,
+      "step": 16270
+    },
+    {
+      "epoch": 1.1977413599661573,
+      "grad_norm": 0.3518880009651184,
+      "learning_rate": 6.779839430543828e-05,
+      "loss": 3.0412,
+      "step": 16280
+    },
+    {
+      "epoch": 1.1984770733322299,
+      "grad_norm": 0.3566107451915741,
+      "learning_rate": 6.776148443555423e-05,
+      "loss": 3.0465,
+      "step": 16290
+    },
+    {
+      "epoch": 1.1992127866983022,
+      "grad_norm": 0.3725287616252899,
+      "learning_rate": 6.772456348574651e-05,
+      "loss": 3.0468,
+      "step": 16300
+    },
+    {
+      "epoch": 1.1999485000643748,
+      "grad_norm": 0.3621078133583069,
+      "learning_rate": 6.768763147904706e-05,
+      "loss": 3.013,
+      "step": 16310
+    },
+    {
+      "epoch": 1.2006842134304474,
+      "grad_norm": 0.35452789068222046,
+      "learning_rate": 6.765068843849472e-05,
+      "loss": 2.984,
+      "step": 16320
+    },
+    {
+      "epoch": 1.20141992679652,
+      "grad_norm": 0.3671383261680603,
+      "learning_rate": 6.761373438713516e-05,
+      "loss": 3.0464,
+      "step": 16330
+    },
+    {
+      "epoch": 1.2021556401625926,
+      "grad_norm": 0.3627682030200958,
+      "learning_rate": 6.7576769348021e-05,
+      "loss": 3.0328,
+      "step": 16340
+    },
+    {
+      "epoch": 1.2028913535286652,
+      "grad_norm": 0.3729478418827057,
+      "learning_rate": 6.753979334421166e-05,
+      "loss": 3.0334,
+      "step": 16350
+    },
+    {
+      "epoch": 1.2036270668947378,
+      "grad_norm": 0.37562015652656555,
+      "learning_rate": 6.750280639877343e-05,
+      "loss": 3.0157,
+      "step": 16360
+    },
+    {
+      "epoch": 1.2043627802608103,
+      "grad_norm": 0.3583190441131592,
+      "learning_rate": 6.746580853477939e-05,
+      "loss": 3.0252,
+      "step": 16370
+    },
+    {
+      "epoch": 1.205098493626883,
+      "grad_norm": 0.356255441904068,
+      "learning_rate": 6.742879977530946e-05,
+      "loss": 3.0077,
+      "step": 16380
+    },
+    {
+      "epoch": 1.2058342069929555,
+      "grad_norm": 0.3496890068054199,
+      "learning_rate": 6.739178014345036e-05,
+      "loss": 3.0355,
+      "step": 16390
+    },
+    {
+      "epoch": 1.206569920359028,
+      "grad_norm": 0.3529980480670929,
+      "learning_rate": 6.735474966229558e-05,
+      "loss": 3.0039,
+      "step": 16400
+    },
+    {
+      "epoch": 1.2073056337251007,
+      "grad_norm": 0.35134485363960266,
+      "learning_rate": 6.731770835494536e-05,
+      "loss": 3.0311,
+      "step": 16410
+    },
+    {
+      "epoch": 1.2080413470911733,
+      "grad_norm": 0.35084834694862366,
+      "learning_rate": 6.728065624450674e-05,
+      "loss": 3.0439,
+      "step": 16420
+    },
+    {
+      "epoch": 1.2087770604572459,
+      "grad_norm": 0.36620283126831055,
+      "learning_rate": 6.724359335409345e-05,
+      "loss": 3.001,
+      "step": 16430
+    },
+    {
+      "epoch": 1.2095127738233185,
+      "grad_norm": 0.3619592785835266,
+      "learning_rate": 6.720651970682599e-05,
+      "loss": 3.0402,
+      "step": 16440
+    },
+    {
+      "epoch": 1.210248487189391,
+      "grad_norm": 0.3597840964794159,
+      "learning_rate": 6.716943532583153e-05,
+      "loss": 2.997,
+      "step": 16450
+    },
+    {
+      "epoch": 1.2109842005554636,
+      "grad_norm": 0.36596518754959106,
+      "learning_rate": 6.713234023424394e-05,
+      "loss": 3.0413,
+      "step": 16460
+    },
+    {
+      "epoch": 1.2117199139215362,
+      "grad_norm": 0.35466885566711426,
+      "learning_rate": 6.709523445520382e-05,
+      "loss": 3.0528,
+      "step": 16470
+    },
+    {
+      "epoch": 1.2124556272876088,
+      "grad_norm": 0.3720371127128601,
+      "learning_rate": 6.705811801185839e-05,
+      "loss": 3.0707,
+      "step": 16480
+    },
+    {
+      "epoch": 1.2131913406536814,
+      "grad_norm": 0.3623346984386444,
+      "learning_rate": 6.702099092736152e-05,
+      "loss": 3.012,
+      "step": 16490
+    },
+    {
+      "epoch": 1.213927054019754,
+      "grad_norm": 0.365092396736145,
+      "learning_rate": 6.698385322487375e-05,
+      "loss": 3.0477,
+      "step": 16500
+    },
+    {
+      "epoch": 1.2146627673858266,
+      "grad_norm": 0.36136770248413086,
+      "learning_rate": 6.694670492756221e-05,
+      "loss": 3.0287,
+      "step": 16510
+    },
+    {
+      "epoch": 1.2153984807518992,
+      "grad_norm": 0.35705703496932983,
+      "learning_rate": 6.690954605860069e-05,
+      "loss": 3.0773,
+      "step": 16520
+    },
+    {
+      "epoch": 1.2161341941179717,
+      "grad_norm": 0.36513903737068176,
+      "learning_rate": 6.687237664116947e-05,
+      "loss": 3.0481,
+      "step": 16530
+    },
+    {
+      "epoch": 1.2168699074840443,
+      "grad_norm": 0.3589208722114563,
+      "learning_rate": 6.683519669845554e-05,
+      "loss": 3.0447,
+      "step": 16540
+    },
+    {
+      "epoch": 1.2176056208501167,
+      "grad_norm": 0.3612947463989258,
+      "learning_rate": 6.67980062536524e-05,
+      "loss": 3.0984,
+      "step": 16550
+    },
+    {
+      "epoch": 1.2183413342161893,
+      "grad_norm": 0.3594346046447754,
+      "learning_rate": 6.676080532996007e-05,
+      "loss": 3.0515,
+      "step": 16560
+    },
+    {
+      "epoch": 1.2190770475822619,
+      "grad_norm": 0.3510363698005676,
+      "learning_rate": 6.672359395058514e-05,
+      "loss": 3.019,
+      "step": 16570
+    },
+    {
+      "epoch": 1.2198127609483345,
+      "grad_norm": 0.37483489513397217,
+      "learning_rate": 6.668637213874074e-05,
+      "loss": 3.0632,
+      "step": 16580
+    },
+    {
+      "epoch": 1.220548474314407,
+      "grad_norm": 0.352332204580307,
+      "learning_rate": 6.664913991764644e-05,
+      "loss": 2.9363,
+      "step": 16590
+    },
+    {
+      "epoch": 1.2212841876804796,
+      "grad_norm": 0.3624435067176819,
+      "learning_rate": 6.661189731052839e-05,
+      "loss": 3.0262,
+      "step": 16600
+    },
+    {
+      "epoch": 1.2220199010465522,
+      "grad_norm": 0.3623534142971039,
+      "learning_rate": 6.657464434061918e-05,
+      "loss": 3.0426,
+      "step": 16610
+    },
+    {
+      "epoch": 1.2227556144126248,
+      "grad_norm": 0.35897883772850037,
+      "learning_rate": 6.653738103115781e-05,
+      "loss": 3.0317,
+      "step": 16620
+    },
+    {
+      "epoch": 1.2234913277786974,
+      "grad_norm": 0.37491393089294434,
+      "learning_rate": 6.650010740538984e-05,
+      "loss": 3.0765,
+      "step": 16630
+    },
+    {
+      "epoch": 1.22422704114477,
+      "grad_norm": 0.34876549243927,
+      "learning_rate": 6.646282348656719e-05,
+      "loss": 3.0214,
+      "step": 16640
+    },
+    {
+      "epoch": 1.2249627545108426,
+      "grad_norm": 0.3542339503765106,
+      "learning_rate": 6.64255292979482e-05,
+      "loss": 3.0386,
+      "step": 16650
+    },
+    {
+      "epoch": 1.2256984678769152,
+      "grad_norm": 0.3774896562099457,
+      "learning_rate": 6.638822486279765e-05,
+      "loss": 3.0801,
+      "step": 16660
+    },
+    {
+      "epoch": 1.2264341812429878,
+      "grad_norm": 0.3710559606552124,
+      "learning_rate": 6.63509102043867e-05,
+      "loss": 3.0487,
+      "step": 16670
+    },
+    {
+      "epoch": 1.2271698946090603,
+      "grad_norm": 0.3601890504360199,
+      "learning_rate": 6.631358534599288e-05,
+      "loss": 2.9919,
+      "step": 16680
+    },
+    {
+      "epoch": 1.227905607975133,
+      "grad_norm": 0.36121076345443726,
+      "learning_rate": 6.627625031090009e-05,
+      "loss": 3.01,
+      "step": 16690
+    },
+    {
+      "epoch": 1.2286413213412055,
+      "grad_norm": 0.3575782775878906,
+      "learning_rate": 6.623890512239857e-05,
+      "loss": 3.0523,
+      "step": 16700
+    },
+    {
+      "epoch": 1.229377034707278,
+      "grad_norm": 0.35671988129615784,
+      "learning_rate": 6.62015498037849e-05,
+      "loss": 2.9874,
+      "step": 16710
+    },
+    {
+      "epoch": 1.2301127480733507,
+      "grad_norm": 0.3698766231536865,
+      "learning_rate": 6.6164184378362e-05,
+      "loss": 3.0466,
+      "step": 16720
+    },
+    {
+      "epoch": 1.2308484614394233,
+      "grad_norm": 0.35145121812820435,
+      "learning_rate": 6.612680886943902e-05,
+      "loss": 2.9757,
+      "step": 16730
+    },
+    {
+      "epoch": 1.2315841748054959,
+      "grad_norm": 0.3570511043071747,
+      "learning_rate": 6.608942330033154e-05,
+      "loss": 3.0152,
+      "step": 16740
+    },
+    {
+      "epoch": 1.2323198881715682,
+      "grad_norm": 0.36074942350387573,
+      "learning_rate": 6.605202769436123e-05,
+      "loss": 2.9993,
+      "step": 16750
+    },
+    {
+      "epoch": 1.2330556015376408,
+      "grad_norm": 0.37237951159477234,
+      "learning_rate": 6.601462207485621e-05,
+      "loss": 3.0012,
+      "step": 16760
+    },
+    {
+      "epoch": 1.2337913149037134,
+      "grad_norm": 0.3707306683063507,
+      "learning_rate": 6.59772064651507e-05,
+      "loss": 3.0392,
+      "step": 16770
+    },
+    {
+      "epoch": 1.234527028269786,
+      "grad_norm": 0.37803375720977783,
+      "learning_rate": 6.593978088858523e-05,
+      "loss": 2.9711,
+      "step": 16780
+    },
+    {
+      "epoch": 1.2352627416358586,
+      "grad_norm": 0.3524588942527771,
+      "learning_rate": 6.590234536850653e-05,
+      "loss": 3.0197,
+      "step": 16790
+    },
+    {
+      "epoch": 1.2359984550019312,
+      "grad_norm": 0.3548072576522827,
+      "learning_rate": 6.586489992826752e-05,
+      "loss": 3.0431,
+      "step": 16800
+    },
+    {
+      "epoch": 1.2367341683680038,
+      "grad_norm": 0.3497982621192932,
+      "learning_rate": 6.582744459122734e-05,
+      "loss": 3.0167,
+      "step": 16810
+    },
+    {
+      "epoch": 1.2374698817340763,
+      "grad_norm": 0.3573327958583832,
+      "learning_rate": 6.578997938075125e-05,
+      "loss": 2.9991,
+      "step": 16820
+    },
+    {
+      "epoch": 1.238205595100149,
+      "grad_norm": 0.3625450134277344,
+      "learning_rate": 6.575250432021073e-05,
+      "loss": 3.0483,
+      "step": 16830
+    },
+    {
+      "epoch": 1.2389413084662215,
+      "grad_norm": 0.3496323823928833,
+      "learning_rate": 6.571501943298334e-05,
+      "loss": 3.0114,
+      "step": 16840
+    },
+    {
+      "epoch": 1.2396770218322941,
+      "grad_norm": 0.39002302289009094,
+      "learning_rate": 6.567752474245286e-05,
+      "loss": 3.0095,
+      "step": 16850
+    },
+    {
+      "epoch": 1.2404127351983667,
+      "grad_norm": 0.374341756105423,
+      "learning_rate": 6.564002027200909e-05,
+      "loss": 3.0485,
+      "step": 16860
+    },
+    {
+      "epoch": 1.2411484485644393,
+      "grad_norm": 0.3576667606830597,
+      "learning_rate": 6.560250604504799e-05,
+      "loss": 3.0716,
+      "step": 16870
+    },
+    {
+      "epoch": 1.2418841619305119,
+      "grad_norm": 0.35809680819511414,
+      "learning_rate": 6.556498208497158e-05,
+      "loss": 3.02,
+      "step": 16880
+    },
+    {
+      "epoch": 1.2426198752965845,
+      "grad_norm": 0.3604182302951813,
+      "learning_rate": 6.552744841518797e-05,
+      "loss": 3.016,
+      "step": 16890
+    },
+    {
+      "epoch": 1.243355588662657,
+      "grad_norm": 0.3616543412208557,
+      "learning_rate": 6.548990505911131e-05,
+      "loss": 3.0102,
+      "step": 16900
+    },
+    {
+      "epoch": 1.2440913020287296,
+      "grad_norm": 0.3425278067588806,
+      "learning_rate": 6.545235204016177e-05,
+      "loss": 3.0163,
+      "step": 16910
+    },
+    {
+      "epoch": 1.2448270153948022,
+      "grad_norm": 0.3548402190208435,
+      "learning_rate": 6.541478938176566e-05,
+      "loss": 3.0125,
+      "step": 16920
+    },
+    {
+      "epoch": 1.2455627287608748,
+      "grad_norm": 0.377497136592865,
+      "learning_rate": 6.537721710735515e-05,
+      "loss": 2.9972,
+      "step": 16930
+    },
+    {
+      "epoch": 1.2462984421269474,
+      "grad_norm": 0.36204952001571655,
+      "learning_rate": 6.53396352403685e-05,
+      "loss": 3.056,
+      "step": 16940
+    },
+    {
+      "epoch": 1.24703415549302,
+      "grad_norm": 0.36146992444992065,
+      "learning_rate": 6.530204380424993e-05,
+      "loss": 3.0141,
+      "step": 16950
+    },
+    {
+      "epoch": 1.2477698688590926,
+      "grad_norm": 0.35990357398986816,
+      "learning_rate": 6.526444282244965e-05,
+      "loss": 3.022,
+      "step": 16960
+    },
+    {
+      "epoch": 1.2485055822251652,
+      "grad_norm": 0.37471073865890503,
+      "learning_rate": 6.522683231842379e-05,
+      "loss": 3.0191,
+      "step": 16970
+    },
+    {
+      "epoch": 1.2492412955912378,
+      "grad_norm": 0.3637163043022156,
+      "learning_rate": 6.518921231563445e-05,
+      "loss": 3.0286,
+      "step": 16980
+    },
+    {
+      "epoch": 1.2499770089573103,
+      "grad_norm": 0.37237876653671265,
+      "learning_rate": 6.515158283754964e-05,
+      "loss": 3.0073,
+      "step": 16990
+    },
+    {
+      "epoch": 1.250712722323383,
+      "grad_norm": 0.3546529710292816,
+      "learning_rate": 6.511394390764326e-05,
+      "loss": 3.0142,
+      "step": 17000
+    },
+    {
+      "epoch": 1.2514484356894555,
+      "grad_norm": 0.3584796190261841,
+      "learning_rate": 6.507629554939518e-05,
+      "loss": 3.0189,
+      "step": 17010
+    },
+    {
+      "epoch": 1.2521841490555279,
+      "grad_norm": 0.3632993698120117,
+      "learning_rate": 6.503863778629106e-05,
+      "loss": 3.0661,
+      "step": 17020
+    },
+    {
+      "epoch": 1.2529198624216005,
+      "grad_norm": 0.3670421540737152,
+      "learning_rate": 6.500097064182249e-05,
+      "loss": 2.9788,
+      "step": 17030
+    },
+    {
+      "epoch": 1.253655575787673,
+      "grad_norm": 0.3614768981933594,
+      "learning_rate": 6.496329413948688e-05,
+      "loss": 2.9962,
+      "step": 17040
+    },
+    {
+      "epoch": 1.2543912891537456,
+      "grad_norm": 0.35209009051322937,
+      "learning_rate": 6.49256083027875e-05,
+      "loss": 3.0549,
+      "step": 17050
+    },
+    {
+      "epoch": 1.2551270025198182,
+      "grad_norm": 0.3788132667541504,
+      "learning_rate": 6.488791315523343e-05,
+      "loss": 3.046,
+      "step": 17060
+    },
+    {
+      "epoch": 1.2558627158858908,
+      "grad_norm": 0.35586991906166077,
+      "learning_rate": 6.485020872033956e-05,
+      "loss": 3.0772,
+      "step": 17070
+    },
+    {
+      "epoch": 1.2565984292519634,
+      "grad_norm": 0.36091411113739014,
+      "learning_rate": 6.481249502162656e-05,
+      "loss": 3.0598,
+      "step": 17080
+    },
+    {
+      "epoch": 1.257334142618036,
+      "grad_norm": 0.37361618876457214,
+      "learning_rate": 6.477477208262091e-05,
+      "loss": 3.062,
+      "step": 17090
+    },
+    {
+      "epoch": 1.2580698559841086,
+      "grad_norm": 0.3713589012622833,
+      "learning_rate": 6.473703992685485e-05,
+      "loss": 3.0523,
+      "step": 17100
+    },
+    {
+      "epoch": 1.2588055693501812,
+      "grad_norm": 0.35614052414894104,
+      "learning_rate": 6.469929857786631e-05,
+      "loss": 3.0113,
+      "step": 17110
+    },
+    {
+      "epoch": 1.2595412827162538,
+      "grad_norm": 0.3531065285205841,
+      "learning_rate": 6.466154805919904e-05,
+      "loss": 3.0412,
+      "step": 17120
+    },
+    {
+      "epoch": 1.2602769960823264,
+      "grad_norm": 0.3637721836566925,
+      "learning_rate": 6.462378839440246e-05,
+      "loss": 3.0399,
+      "step": 17130
+    },
+    {
+      "epoch": 1.261012709448399,
+      "grad_norm": 0.346825510263443,
+      "learning_rate": 6.458601960703172e-05,
+      "loss": 2.9899,
+      "step": 17140
+    },
+    {
+      "epoch": 1.2617484228144715,
+      "grad_norm": 0.3579413890838623,
+      "learning_rate": 6.454824172064761e-05,
+      "loss": 3.015,
+      "step": 17150
+    },
+    {
+      "epoch": 1.2624841361805441,
+      "grad_norm": 0.37174171209335327,
+      "learning_rate": 6.451045475881668e-05,
+      "loss": 3.0334,
+      "step": 17160
+    },
+    {
+      "epoch": 1.2632198495466167,
+      "grad_norm": 0.34888362884521484,
+      "learning_rate": 6.447265874511109e-05,
+      "loss": 3.0245,
+      "step": 17170
+    },
+    {
+      "epoch": 1.2639555629126893,
+      "grad_norm": 0.36276373267173767,
+      "learning_rate": 6.443485370310861e-05,
+      "loss": 3.0383,
+      "step": 17180
+    },
+    {
+      "epoch": 1.2646912762787617,
+      "grad_norm": 0.3784925043582916,
+      "learning_rate": 6.439703965639273e-05,
+      "loss": 2.9974,
+      "step": 17190
+    },
+    {
+      "epoch": 1.2654269896448342,
+      "grad_norm": 0.35036545991897583,
+      "learning_rate": 6.435921662855248e-05,
+      "loss": 3.0081,
+      "step": 17200
+    },
+    {
+      "epoch": 1.2661627030109068,
+      "grad_norm": 0.34726816415786743,
+      "learning_rate": 6.432138464318256e-05,
+      "loss": 2.9948,
+      "step": 17210
+    },
+    {
+      "epoch": 1.2668984163769794,
+      "grad_norm": 0.3575078547000885,
+      "learning_rate": 6.428354372388319e-05,
+      "loss": 3.05,
+      "step": 17220
+    },
+    {
+      "epoch": 1.267634129743052,
+      "grad_norm": 0.3689034581184387,
+      "learning_rate": 6.424569389426021e-05,
+      "loss": 3.0087,
+      "step": 17230
+    },
+    {
+      "epoch": 1.2683698431091246,
+      "grad_norm": 0.3553621172904968,
+      "learning_rate": 6.420783517792499e-05,
+      "loss": 3.0008,
+      "step": 17240
+    },
+    {
+      "epoch": 1.2691055564751972,
+      "grad_norm": 0.37458479404449463,
+      "learning_rate": 6.416996759849448e-05,
+      "loss": 3.0603,
+      "step": 17250
+    },
+    {
+      "epoch": 1.2698412698412698,
+      "grad_norm": 0.36622488498687744,
+      "learning_rate": 6.413209117959111e-05,
+      "loss": 3.0536,
+      "step": 17260
+    },
+    {
+      "epoch": 1.2705769832073424,
+      "grad_norm": 0.36386966705322266,
+      "learning_rate": 6.409420594484287e-05,
+      "loss": 3.0224,
+      "step": 17270
+    },
+    {
+      "epoch": 1.271312696573415,
+      "grad_norm": 0.3479600250720978,
+      "learning_rate": 6.405631191788323e-05,
+      "loss": 3.0,
+      "step": 17280
+    },
+    {
+      "epoch": 1.2720484099394875,
+      "grad_norm": 0.364284485578537,
+      "learning_rate": 6.401840912235113e-05,
+      "loss": 3.0185,
+      "step": 17290
+    },
+    {
+      "epoch": 1.2727841233055601,
+      "grad_norm": 0.3535844385623932,
+      "learning_rate": 6.398049758189102e-05,
+      "loss": 2.9886,
+      "step": 17300
+    },
+    {
+      "epoch": 1.2735198366716327,
+      "grad_norm": 0.3747173547744751,
+      "learning_rate": 6.394257732015274e-05,
+      "loss": 3.0067,
+      "step": 17310
+    },
+    {
+      "epoch": 1.2742555500377053,
+      "grad_norm": 0.35885944962501526,
+      "learning_rate": 6.390464836079164e-05,
+      "loss": 3.027,
+      "step": 17320
+    },
+    {
+      "epoch": 1.2749912634037779,
+      "grad_norm": 0.3500259220600128,
+      "learning_rate": 6.386671072746848e-05,
+      "loss": 2.9958,
+      "step": 17330
+    },
+    {
+      "epoch": 1.2757269767698505,
+      "grad_norm": 0.349939227104187,
+      "learning_rate": 6.382876444384938e-05,
+      "loss": 2.9719,
+      "step": 17340
+    },
+    {
+      "epoch": 1.276462690135923,
+      "grad_norm": 0.3613039553165436,
+      "learning_rate": 6.37908095336059e-05,
+      "loss": 3.0172,
+      "step": 17350
+    },
+    {
+      "epoch": 1.2771984035019956,
+      "grad_norm": 0.3765120208263397,
+      "learning_rate": 6.3752846020415e-05,
+      "loss": 2.9957,
+      "step": 17360
+    },
+    {
+      "epoch": 1.2779341168680682,
+      "grad_norm": 0.36430907249450684,
+      "learning_rate": 6.371487392795896e-05,
+      "loss": 3.0297,
+      "step": 17370
+    },
+    {
+      "epoch": 1.2786698302341408,
+      "grad_norm": 0.3489398658275604,
+      "learning_rate": 6.367689327992546e-05,
+      "loss": 3.0563,
+      "step": 17380
+    },
+    {
+      "epoch": 1.2794055436002134,
+      "grad_norm": 0.3590885400772095,
+      "learning_rate": 6.363890410000745e-05,
+      "loss": 3.0204,
+      "step": 17390
+    },
+    {
+      "epoch": 1.280141256966286,
+      "grad_norm": 0.34929159283638,
+      "learning_rate": 6.360090641190326e-05,
+      "loss": 3.0594,
+      "step": 17400
+    },
+    {
+      "epoch": 1.2808769703323586,
+      "grad_norm": 0.3614228069782257,
+      "learning_rate": 6.356290023931652e-05,
+      "loss": 3.069,
+      "step": 17410
+    },
+    {
+      "epoch": 1.2816126836984312,
+      "grad_norm": 0.36280760169029236,
+      "learning_rate": 6.352488560595615e-05,
+      "loss": 2.9815,
+      "step": 17420
+    },
+    {
+      "epoch": 1.2823483970645038,
+      "grad_norm": 0.3524848222732544,
+      "learning_rate": 6.34868625355363e-05,
+      "loss": 3.0202,
+      "step": 17430
+    },
+    {
+      "epoch": 1.2830841104305764,
+      "grad_norm": 0.3480393886566162,
+      "learning_rate": 6.344883105177647e-05,
+      "loss": 3.0143,
+      "step": 17440
+    },
+    {
+      "epoch": 1.283819823796649,
+      "grad_norm": 0.3632698357105255,
+      "learning_rate": 6.341079117840134e-05,
+      "loss": 3.0257,
+      "step": 17450
+    },
+    {
+      "epoch": 1.2845555371627215,
+      "grad_norm": 0.35565683245658875,
+      "learning_rate": 6.337274293914086e-05,
+      "loss": 3.0023,
+      "step": 17460
+    },
+    {
+      "epoch": 1.2852912505287941,
+      "grad_norm": 0.3494541645050049,
+      "learning_rate": 6.333468635773017e-05,
+      "loss": 3.0188,
+      "step": 17470
+    },
+    {
+      "epoch": 1.2860269638948665,
+      "grad_norm": 0.3437279164791107,
+      "learning_rate": 6.329662145790964e-05,
+      "loss": 3.0154,
+      "step": 17480
+    },
+    {
+      "epoch": 1.286762677260939,
+      "grad_norm": 0.3516070544719696,
+      "learning_rate": 6.325854826342478e-05,
+      "loss": 2.9793,
+      "step": 17490
+    },
+    {
+      "epoch": 1.2874983906270117,
+      "grad_norm": 0.3656921982765198,
+      "learning_rate": 6.322046679802639e-05,
+      "loss": 3.0077,
+      "step": 17500
+    },
+    {
+      "epoch": 1.2882341039930842,
+      "grad_norm": 0.3535267114639282,
+      "learning_rate": 6.318237708547029e-05,
+      "loss": 2.9904,
+      "step": 17510
+    },
+    {
+      "epoch": 1.2889698173591568,
+      "grad_norm": 0.3563414514064789,
+      "learning_rate": 6.314427914951752e-05,
+      "loss": 2.9921,
+      "step": 17520
+    },
+    {
+      "epoch": 1.2897055307252294,
+      "grad_norm": 0.35284677147865295,
+      "learning_rate": 6.310617301393425e-05,
+      "loss": 3.0317,
+      "step": 17530
+    },
+    {
+      "epoch": 1.290441244091302,
+      "grad_norm": 0.36022713780403137,
+      "learning_rate": 6.306805870249173e-05,
+      "loss": 3.0238,
+      "step": 17540
+    },
+    {
+      "epoch": 1.2911769574573746,
+      "grad_norm": 0.34462112188339233,
+      "learning_rate": 6.302993623896638e-05,
+      "loss": 2.9958,
+      "step": 17550
+    },
+    {
+      "epoch": 1.2919126708234472,
+      "grad_norm": 0.3651984930038452,
+      "learning_rate": 6.299180564713958e-05,
+      "loss": 3.0187,
+      "step": 17560
+    },
+    {
+      "epoch": 1.2926483841895198,
+      "grad_norm": 0.3624741733074188,
+      "learning_rate": 6.295366695079791e-05,
+      "loss": 3.0233,
+      "step": 17570
+    },
+    {
+      "epoch": 1.2933840975555924,
+      "grad_norm": 0.35524824261665344,
+      "learning_rate": 6.291552017373296e-05,
+      "loss": 3.0153,
+      "step": 17580
+    },
+    {
+      "epoch": 1.294119810921665,
+      "grad_norm": 0.3574567139148712,
+      "learning_rate": 6.287736533974133e-05,
+      "loss": 2.9507,
+      "step": 17590
+    },
+    {
+      "epoch": 1.2948555242877375,
+      "grad_norm": 0.3478335738182068,
+      "learning_rate": 6.283920247262468e-05,
+      "loss": 3.0138,
+      "step": 17600
+    },
+    {
+      "epoch": 1.2955912376538101,
+      "grad_norm": 0.34411466121673584,
+      "learning_rate": 6.280103159618968e-05,
+      "loss": 3.0406,
+      "step": 17610
+    },
+    {
+      "epoch": 1.2963269510198827,
+      "grad_norm": 0.35705703496932983,
+      "learning_rate": 6.276285273424797e-05,
+      "loss": 2.9679,
+      "step": 17620
+    },
+    {
+      "epoch": 1.2970626643859553,
+      "grad_norm": 0.34910476207733154,
+      "learning_rate": 6.27246659106162e-05,
+      "loss": 3.0001,
+      "step": 17630
+    },
+    {
+      "epoch": 1.2977983777520279,
+      "grad_norm": 0.35742583870887756,
+      "learning_rate": 6.2686471149116e-05,
+      "loss": 2.972,
+      "step": 17640
+    },
+    {
+      "epoch": 1.2985340911181003,
+      "grad_norm": 0.36850234866142273,
+      "learning_rate": 6.264826847357384e-05,
+      "loss": 3.0271,
+      "step": 17650
+    },
+    {
+      "epoch": 1.2992698044841728,
+      "grad_norm": 0.3519894480705261,
+      "learning_rate": 6.261005790782133e-05,
+      "loss": 2.9785,
+      "step": 17660
+    },
+    {
+      "epoch": 1.3000055178502454,
+      "grad_norm": 0.3518581688404083,
+      "learning_rate": 6.257183947569483e-05,
+      "loss": 3.0102,
+      "step": 17670
+    },
+    {
+      "epoch": 1.300741231216318,
+      "grad_norm": 0.35914346575737,
+      "learning_rate": 6.253361320103566e-05,
+      "loss": 3.0141,
+      "step": 17680
+    },
+    {
+      "epoch": 1.3014769445823906,
+      "grad_norm": 0.35114917159080505,
+      "learning_rate": 6.249537910769006e-05,
+      "loss": 3.0322,
+      "step": 17690
+    },
+    {
+      "epoch": 1.3022126579484632,
+      "grad_norm": 0.3513551354408264,
+      "learning_rate": 6.24571372195091e-05,
+      "loss": 3.0288,
+      "step": 17700
+    },
+    {
+      "epoch": 1.3029483713145358,
+      "grad_norm": 0.3697724938392639,
+      "learning_rate": 6.241888756034876e-05,
+      "loss": 2.9914,
+      "step": 17710
+    },
+    {
+      "epoch": 1.3036840846806084,
+      "grad_norm": 0.35416221618652344,
+      "learning_rate": 6.238063015406982e-05,
+      "loss": 3.0071,
+      "step": 17720
+    },
+    {
+      "epoch": 1.304419798046681,
+      "grad_norm": 0.36622077226638794,
+      "learning_rate": 6.234236502453795e-05,
+      "loss": 2.991,
+      "step": 17730
+    },
+    {
+      "epoch": 1.3051555114127535,
+      "grad_norm": 0.3418654501438141,
+      "learning_rate": 6.230409219562354e-05,
+      "loss": 3.0445,
+      "step": 17740
+    },
+    {
+      "epoch": 1.3058912247788261,
+      "grad_norm": 0.35937049984931946,
+      "learning_rate": 6.226581169120194e-05,
+      "loss": 2.9644,
+      "step": 17750
+    },
+    {
+      "epoch": 1.3066269381448987,
+      "grad_norm": 0.36290431022644043,
+      "learning_rate": 6.222752353515311e-05,
+      "loss": 3.0205,
+      "step": 17760
+    },
+    {
+      "epoch": 1.3073626515109713,
+      "grad_norm": 0.3573070764541626,
+      "learning_rate": 6.218922775136193e-05,
+      "loss": 2.9902,
+      "step": 17770
+    },
+    {
+      "epoch": 1.308098364877044,
+      "grad_norm": 0.3501330614089966,
+      "learning_rate": 6.215092436371795e-05,
+      "loss": 3.046,
+      "step": 17780
+    },
+    {
+      "epoch": 1.3088340782431165,
+      "grad_norm": 0.3623714745044708,
+      "learning_rate": 6.21126133961155e-05,
+      "loss": 2.9997,
+      "step": 17790
+    },
+    {
+      "epoch": 1.309569791609189,
+      "grad_norm": 0.3452303111553192,
+      "learning_rate": 6.207429487245362e-05,
+      "loss": 3.0065,
+      "step": 17800
+    },
+    {
+      "epoch": 1.3103055049752617,
+      "grad_norm": 0.36598172783851624,
+      "learning_rate": 6.203596881663609e-05,
+      "loss": 3.0323,
+      "step": 17810
+    },
+    {
+      "epoch": 1.3110412183413342,
+      "grad_norm": 0.34711185097694397,
+      "learning_rate": 6.199763525257134e-05,
+      "loss": 3.0304,
+      "step": 17820
+    },
+    {
+      "epoch": 1.3117769317074068,
+      "grad_norm": 0.352658748626709,
+      "learning_rate": 6.195929420417255e-05,
+      "loss": 3.0356,
+      "step": 17830
+    },
+    {
+      "epoch": 1.3125126450734794,
+      "grad_norm": 0.36650845408439636,
+      "learning_rate": 6.192094569535753e-05,
+      "loss": 3.0562,
+      "step": 17840
+    },
+    {
+      "epoch": 1.313248358439552,
+      "grad_norm": 0.3513515293598175,
+      "learning_rate": 6.188258975004871e-05,
+      "loss": 3.0364,
+      "step": 17850
+    },
+    {
+      "epoch": 1.3139840718056246,
+      "grad_norm": 0.35336560010910034,
+      "learning_rate": 6.184422639217324e-05,
+      "loss": 3.0249,
+      "step": 17860
+    },
+    {
+      "epoch": 1.3147197851716972,
+      "grad_norm": 0.35664498805999756,
+      "learning_rate": 6.180585564566285e-05,
+      "loss": 3.0586,
+      "step": 17870
+    },
+    {
+      "epoch": 1.3154554985377698,
+      "grad_norm": 0.38727355003356934,
+      "learning_rate": 6.176747753445387e-05,
+      "loss": 3.0204,
+      "step": 17880
+    },
+    {
+      "epoch": 1.3161912119038424,
+      "grad_norm": 0.3572586178779602,
+      "learning_rate": 6.172909208248723e-05,
+      "loss": 3.0277,
+      "step": 17890
+    },
+    {
+      "epoch": 1.316926925269915,
+      "grad_norm": 0.3712838292121887,
+      "learning_rate": 6.169069931370842e-05,
+      "loss": 3.0146,
+      "step": 17900
+    },
+    {
+      "epoch": 1.3176626386359875,
+      "grad_norm": 0.36368274688720703,
+      "learning_rate": 6.165229925206758e-05,
+      "loss": 3.0368,
+      "step": 17910
+    },
+    {
+      "epoch": 1.3183983520020601,
+      "grad_norm": 0.36200493574142456,
+      "learning_rate": 6.16138919215193e-05,
+      "loss": 3.0147,
+      "step": 17920
+    },
+    {
+      "epoch": 1.3191340653681327,
+      "grad_norm": 0.3628910481929779,
+      "learning_rate": 6.157547734602274e-05,
+      "loss": 2.9766,
+      "step": 17930
+    },
+    {
+      "epoch": 1.319869778734205,
+      "grad_norm": 0.3657628297805786,
+      "learning_rate": 6.153705554954157e-05,
+      "loss": 3.0206,
+      "step": 17940
+    },
+    {
+      "epoch": 1.3206054921002777,
+      "grad_norm": 0.3507484495639801,
+      "learning_rate": 6.149862655604404e-05,
+      "loss": 3.0465,
+      "step": 17950
+    },
+    {
+      "epoch": 1.3213412054663503,
+      "grad_norm": 0.3542044162750244,
+      "learning_rate": 6.146019038950274e-05,
+      "loss": 3.0666,
+      "step": 17960
+    },
+    {
+      "epoch": 1.3220769188324228,
+      "grad_norm": 0.36235538125038147,
+      "learning_rate": 6.142174707389489e-05,
+      "loss": 3.0611,
+      "step": 17970
+    },
+    {
+      "epoch": 1.3228126321984954,
+      "grad_norm": 0.35866057872772217,
+      "learning_rate": 6.138329663320206e-05,
+      "loss": 3.0254,
+      "step": 17980
+    },
+    {
+      "epoch": 1.323548345564568,
+      "grad_norm": 0.34160804748535156,
+      "learning_rate": 6.134483909141033e-05,
+      "loss": 2.9913,
+      "step": 17990
+    },
+    {
+      "epoch": 1.3242840589306406,
+      "grad_norm": 0.3417339622974396,
+      "learning_rate": 6.130637447251017e-05,
+      "loss": 2.9801,
+      "step": 18000
+    },
+    {
+      "epoch": 1.3250197722967132,
+      "grad_norm": 0.3533668518066406,
+      "learning_rate": 6.126790280049646e-05,
+      "loss": 3.029,
+      "step": 18010
+    },
+    {
+      "epoch": 1.3257554856627858,
+      "grad_norm": 0.35247501730918884,
+      "learning_rate": 6.122942409936853e-05,
+      "loss": 3.0027,
+      "step": 18020
+    },
+    {
+      "epoch": 1.3264911990288584,
+      "grad_norm": 0.3525543510913849,
+      "learning_rate": 6.119093839313008e-05,
+      "loss": 3.0144,
+      "step": 18030
+    },
+    {
+      "epoch": 1.327226912394931,
+      "grad_norm": 0.3561651408672333,
+      "learning_rate": 6.115244570578913e-05,
+      "loss": 3.0141,
+      "step": 18040
+    },
+    {
+      "epoch": 1.3279626257610035,
+      "grad_norm": 0.3574586510658264,
+      "learning_rate": 6.11139460613581e-05,
+      "loss": 3.0231,
+      "step": 18050
+    },
+    {
+      "epoch": 1.3286983391270761,
+      "grad_norm": 0.36691224575042725,
+      "learning_rate": 6.107543948385376e-05,
+      "loss": 3.0312,
+      "step": 18060
+    },
+    {
+      "epoch": 1.3294340524931487,
+      "grad_norm": 0.3521580100059509,
+      "learning_rate": 6.103692599729713e-05,
+      "loss": 3.0204,
+      "step": 18070
+    },
+    {
+      "epoch": 1.3301697658592213,
+      "grad_norm": 0.35565710067749023,
+      "learning_rate": 6.099840562571366e-05,
+      "loss": 3.0325,
+      "step": 18080
+    },
+    {
+      "epoch": 1.330905479225294,
+      "grad_norm": 0.3587721884250641,
+      "learning_rate": 6.0959878393132994e-05,
+      "loss": 3.072,
+      "step": 18090
+    },
+    {
+      "epoch": 1.3316411925913665,
+      "grad_norm": 0.35350528359413147,
+      "learning_rate": 6.092134432358909e-05,
+      "loss": 3.0484,
+      "step": 18100
+    },
+    {
+      "epoch": 1.3323769059574389,
+      "grad_norm": 0.3417935073375702,
+      "learning_rate": 6.088280344112017e-05,
+      "loss": 3.0138,
+      "step": 18110
+    },
+    {
+      "epoch": 1.3331126193235114,
+      "grad_norm": 0.3528693914413452,
+      "learning_rate": 6.084425576976873e-05,
+      "loss": 2.9841,
+      "step": 18120
+    },
+    {
+      "epoch": 1.333848332689584,
+      "grad_norm": 0.35322049260139465,
+      "learning_rate": 6.080570133358148e-05,
+      "loss": 3.0011,
+      "step": 18130
+    },
+    {
+      "epoch": 1.3345840460556566,
+      "grad_norm": 0.3443946838378906,
+      "learning_rate": 6.0767140156609317e-05,
+      "loss": 3.0191,
+      "step": 18140
+    },
+    {
+      "epoch": 1.3353197594217292,
+      "grad_norm": 0.355679452419281,
+      "learning_rate": 6.072857226290739e-05,
+      "loss": 3.0335,
+      "step": 18150
+    },
+    {
+      "epoch": 1.3360554727878018,
+      "grad_norm": 0.358106404542923,
+      "learning_rate": 6.068999767653505e-05,
+      "loss": 3.0742,
+      "step": 18160
+    },
+    {
+      "epoch": 1.3367911861538744,
+      "grad_norm": 0.35108447074890137,
+      "learning_rate": 6.0651416421555766e-05,
+      "loss": 3.0156,
+      "step": 18170
+    },
+    {
+      "epoch": 1.337526899519947,
+      "grad_norm": 0.3535974323749542,
+      "learning_rate": 6.0612828522037214e-05,
+      "loss": 3.0441,
+      "step": 18180
+    },
+    {
+      "epoch": 1.3382626128860196,
+      "grad_norm": 0.35313886404037476,
+      "learning_rate": 6.057423400205117e-05,
+      "loss": 3.0615,
+      "step": 18190
+    },
+    {
+      "epoch": 1.3389983262520921,
+      "grad_norm": 0.36433276534080505,
+      "learning_rate": 6.053563288567361e-05,
+      "loss": 3.0108,
+      "step": 18200
+    },
+    {
+      "epoch": 1.3397340396181647,
+      "grad_norm": 0.3544667065143585,
+      "learning_rate": 6.049702519698456e-05,
+      "loss": 3.0285,
+      "step": 18210
+    },
+    {
+      "epoch": 1.3404697529842373,
+      "grad_norm": 0.3574714958667755,
+      "learning_rate": 6.045841096006817e-05,
+      "loss": 3.0221,
+      "step": 18220
+    },
+    {
+      "epoch": 1.34120546635031,
+      "grad_norm": 0.34618064761161804,
+      "learning_rate": 6.0419790199012674e-05,
+      "loss": 2.9851,
+      "step": 18230
+    },
+    {
+      "epoch": 1.3419411797163825,
+      "grad_norm": 0.3495144248008728,
+      "learning_rate": 6.038116293791039e-05,
+      "loss": 3.0857,
+      "step": 18240
+    },
+    {
+      "epoch": 1.342676893082455,
+      "grad_norm": 0.37529462575912476,
+      "learning_rate": 6.0342529200857654e-05,
+      "loss": 3.0099,
+      "step": 18250
+    },
+    {
+      "epoch": 1.3434126064485277,
+      "grad_norm": 0.37452372908592224,
+      "learning_rate": 6.0303889011954874e-05,
+      "loss": 3.0586,
+      "step": 18260
+    },
+    {
+      "epoch": 1.3441483198146003,
+      "grad_norm": 0.3606548607349396,
+      "learning_rate": 6.026524239530649e-05,
+      "loss": 3.0337,
+      "step": 18270
+    },
+    {
+      "epoch": 1.3448840331806728,
+      "grad_norm": 0.3565804064273834,
+      "learning_rate": 6.0226589375020904e-05,
+      "loss": 3.008,
+      "step": 18280
+    },
+    {
+      "epoch": 1.3456197465467454,
+      "grad_norm": 0.3474801182746887,
+      "learning_rate": 6.018792997521058e-05,
+      "loss": 3.0654,
+      "step": 18290
+    },
+    {
+      "epoch": 1.346355459912818,
+      "grad_norm": 0.3512808680534363,
+      "learning_rate": 6.01492642199919e-05,
+      "loss": 2.9957,
+      "step": 18300
+    },
+    {
+      "epoch": 1.3470911732788906,
+      "grad_norm": 0.3414289355278015,
+      "learning_rate": 6.0110592133485235e-05,
+      "loss": 3.0043,
+      "step": 18310
+    },
+    {
+      "epoch": 1.3478268866449632,
+      "grad_norm": 0.34323611855506897,
+      "learning_rate": 6.007191373981492e-05,
+      "loss": 3.0254,
+      "step": 18320
+    },
+    {
+      "epoch": 1.3485626000110358,
+      "grad_norm": 0.36324530839920044,
+      "learning_rate": 6.003322906310922e-05,
+      "loss": 3.0588,
+      "step": 18330
+    },
+    {
+      "epoch": 1.3492983133771084,
+      "grad_norm": 0.3647429347038269,
+      "learning_rate": 5.999453812750028e-05,
+      "loss": 2.9987,
+      "step": 18340
+    },
+    {
+      "epoch": 1.350034026743181,
+      "grad_norm": 0.3636254072189331,
+      "learning_rate": 5.99558409571242e-05,
+      "loss": 3.03,
+      "step": 18350
+    },
+    {
+      "epoch": 1.3507697401092535,
+      "grad_norm": 0.3647676110267639,
+      "learning_rate": 5.991713757612096e-05,
+      "loss": 3.0706,
+      "step": 18360
+    },
+    {
+      "epoch": 1.3515054534753261,
+      "grad_norm": 0.3620757460594177,
+      "learning_rate": 5.987842800863438e-05,
+      "loss": 3.0446,
+      "step": 18370
+    },
+    {
+      "epoch": 1.3522411668413987,
+      "grad_norm": 0.3800886571407318,
+      "learning_rate": 5.9839712278812186e-05,
+      "loss": 3.0526,
+      "step": 18380
+    },
+    {
+      "epoch": 1.352976880207471,
+      "grad_norm": 0.35337308049201965,
+      "learning_rate": 5.980099041080589e-05,
+      "loss": 3.0334,
+      "step": 18390
+    },
+    {
+      "epoch": 1.3537125935735437,
+      "grad_norm": 0.35593485832214355,
+      "learning_rate": 5.976226242877092e-05,
+      "loss": 3.031,
+      "step": 18400
+    },
+    {
+      "epoch": 1.3544483069396163,
+      "grad_norm": 0.3523681163787842,
+      "learning_rate": 5.972352835686642e-05,
+      "loss": 3.026,
+      "step": 18410
+    },
+    {
+      "epoch": 1.3551840203056889,
+      "grad_norm": 0.34817299246788025,
+      "learning_rate": 5.96847882192554e-05,
+      "loss": 2.9798,
+      "step": 18420
+    },
+    {
+      "epoch": 1.3559197336717614,
+      "grad_norm": 0.3531607985496521,
+      "learning_rate": 5.9646042040104645e-05,
+      "loss": 2.9892,
+      "step": 18430
+    },
+    {
+      "epoch": 1.356655447037834,
+      "grad_norm": 0.3451179265975952,
+      "learning_rate": 5.960728984358466e-05,
+      "loss": 2.9724,
+      "step": 18440
+    },
+    {
+      "epoch": 1.3573911604039066,
+      "grad_norm": 0.3463840186595917,
+      "learning_rate": 5.956853165386978e-05,
+      "loss": 3.038,
+      "step": 18450
+    },
+    {
+      "epoch": 1.3581268737699792,
+      "grad_norm": 0.35228395462036133,
+      "learning_rate": 5.9529767495138026e-05,
+      "loss": 2.9836,
+      "step": 18460
+    },
+    {
+      "epoch": 1.3588625871360518,
+      "grad_norm": 0.35736799240112305,
+      "learning_rate": 5.9490997391571157e-05,
+      "loss": 3.0272,
+      "step": 18470
+    },
+    {
+      "epoch": 1.3595983005021244,
+      "grad_norm": 0.3610997498035431,
+      "learning_rate": 5.945222136735461e-05,
+      "loss": 3.0357,
+      "step": 18480
+    },
+    {
+      "epoch": 1.360334013868197,
+      "grad_norm": 0.35050585865974426,
+      "learning_rate": 5.941343944667759e-05,
+      "loss": 2.9894,
+      "step": 18490
+    },
+    {
+      "epoch": 1.3610697272342696,
+      "grad_norm": 0.34473127126693726,
+      "learning_rate": 5.937465165373292e-05,
+      "loss": 3.0037,
+      "step": 18500
+    },
+    {
+      "epoch": 1.3618054406003421,
+      "grad_norm": 0.34127482771873474,
+      "learning_rate": 5.9335858012717096e-05,
+      "loss": 3.0001,
+      "step": 18510
+    },
+    {
+      "epoch": 1.3625411539664147,
+      "grad_norm": 0.39419862627983093,
+      "learning_rate": 5.929705854783025e-05,
+      "loss": 3.0537,
+      "step": 18520
+    },
+    {
+      "epoch": 1.3632768673324873,
+      "grad_norm": 0.3583771884441376,
+      "learning_rate": 5.925825328327621e-05,
+      "loss": 2.9943,
+      "step": 18530
+    },
+    {
+      "epoch": 1.36401258069856,
+      "grad_norm": 0.3565029799938202,
+      "learning_rate": 5.9219442243262346e-05,
+      "loss": 2.9848,
+      "step": 18540
+    },
+    {
+      "epoch": 1.3647482940646325,
+      "grad_norm": 0.36535730957984924,
+      "learning_rate": 5.918062545199965e-05,
+      "loss": 3.0398,
+      "step": 18550
+    },
+    {
+      "epoch": 1.3654840074307049,
+      "grad_norm": 0.3701373338699341,
+      "learning_rate": 5.9141802933702736e-05,
+      "loss": 2.9924,
+      "step": 18560
+    },
+    {
+      "epoch": 1.3662197207967774,
+      "grad_norm": 0.3723754286766052,
+      "learning_rate": 5.9102974712589766e-05,
+      "loss": 2.9895,
+      "step": 18570
+    },
+    {
+      "epoch": 1.36695543416285,
+      "grad_norm": 0.35265985131263733,
+      "learning_rate": 5.906414081288246e-05,
+      "loss": 2.9768,
+      "step": 18580
+    },
+    {
+      "epoch": 1.3676911475289226,
+      "grad_norm": 0.35446324944496155,
+      "learning_rate": 5.9025301258806074e-05,
+      "loss": 2.9474,
+      "step": 18590
+    },
+    {
+      "epoch": 1.3684268608949952,
+      "grad_norm": 0.35580405592918396,
+      "learning_rate": 5.8986456074589404e-05,
+      "loss": 2.9789,
+      "step": 18600
+    },
+    {
+      "epoch": 1.3691625742610678,
+      "grad_norm": 0.34461915493011475,
+      "learning_rate": 5.894760528446476e-05,
+      "loss": 3.0038,
+      "step": 18610
+    },
+    {
+      "epoch": 1.3698982876271404,
+      "grad_norm": 0.3618105947971344,
+      "learning_rate": 5.890874891266792e-05,
+      "loss": 3.0558,
+      "step": 18620
+    },
+    {
+      "epoch": 1.370634000993213,
+      "grad_norm": 0.3577014207839966,
+      "learning_rate": 5.88698869834382e-05,
+      "loss": 2.9967,
+      "step": 18630
+    },
+    {
+      "epoch": 1.3713697143592856,
+      "grad_norm": 0.3607986271381378,
+      "learning_rate": 5.8831019521018306e-05,
+      "loss": 3.0053,
+      "step": 18640
+    },
+    {
+      "epoch": 1.3721054277253582,
+      "grad_norm": 0.3699253797531128,
+      "learning_rate": 5.879214654965448e-05,
+      "loss": 3.0213,
+      "step": 18650
+    },
+    {
+      "epoch": 1.3728411410914307,
+      "grad_norm": 0.34234777092933655,
+      "learning_rate": 5.875326809359634e-05,
+      "loss": 3.0224,
+      "step": 18660
+    },
+    {
+      "epoch": 1.3735768544575033,
+      "grad_norm": 0.36211714148521423,
+      "learning_rate": 5.871438417709695e-05,
+      "loss": 3.0249,
+      "step": 18670
+    },
+    {
+      "epoch": 1.374312567823576,
+      "grad_norm": 0.3483603000640869,
+      "learning_rate": 5.867549482441276e-05,
+      "loss": 3.0245,
+      "step": 18680
+    },
+    {
+      "epoch": 1.3750482811896485,
+      "grad_norm": 0.36787933111190796,
+      "learning_rate": 5.863660005980364e-05,
+      "loss": 2.9914,
+      "step": 18690
+    },
+    {
+      "epoch": 1.375783994555721,
+      "grad_norm": 0.3375144302845001,
+      "learning_rate": 5.859769990753283e-05,
+      "loss": 3.0226,
+      "step": 18700
+    },
+    {
+      "epoch": 1.3765197079217937,
+      "grad_norm": 0.3520945906639099,
+      "learning_rate": 5.8558794391866914e-05,
+      "loss": 3.0239,
+      "step": 18710
+    },
+    {
+      "epoch": 1.3772554212878663,
+      "grad_norm": 0.3681041896343231,
+      "learning_rate": 5.851988353707582e-05,
+      "loss": 3.0096,
+      "step": 18720
+    },
+    {
+      "epoch": 1.3779911346539389,
+      "grad_norm": 0.37708455324172974,
+      "learning_rate": 5.8480967367432824e-05,
+      "loss": 2.9514,
+      "step": 18730
+    },
+    {
+      "epoch": 1.3787268480200114,
+      "grad_norm": 0.35953813791275024,
+      "learning_rate": 5.844204590721454e-05,
+      "loss": 3.0397,
+      "step": 18740
+    },
+    {
+      "epoch": 1.379462561386084,
+      "grad_norm": 0.3589445948600769,
+      "learning_rate": 5.840311918070081e-05,
+      "loss": 3.0377,
+      "step": 18750
+    },
+    {
+      "epoch": 1.3801982747521566,
+      "grad_norm": 0.3587503731250763,
+      "learning_rate": 5.8364187212174825e-05,
+      "loss": 3.0092,
+      "step": 18760
+    },
+    {
+      "epoch": 1.3809339881182292,
+      "grad_norm": 0.36021992564201355,
+      "learning_rate": 5.8325250025923015e-05,
+      "loss": 3.039,
+      "step": 18770
+    },
+    {
+      "epoch": 1.3816697014843018,
+      "grad_norm": 0.3543018400669098,
+      "learning_rate": 5.8286307646235104e-05,
+      "loss": 3.0169,
+      "step": 18780
+    },
+    {
+      "epoch": 1.3824054148503744,
+      "grad_norm": 0.3543086647987366,
+      "learning_rate": 5.8247360097404e-05,
+      "loss": 2.9712,
+      "step": 18790
+    },
+    {
+      "epoch": 1.383141128216447,
+      "grad_norm": 0.3551665246486664,
+      "learning_rate": 5.820840740372588e-05,
+      "loss": 2.995,
+      "step": 18800
+    },
+    {
+      "epoch": 1.3838768415825196,
+      "grad_norm": 0.3884110748767853,
+      "learning_rate": 5.8169449589500094e-05,
+      "loss": 2.9597,
+      "step": 18810
+    },
+    {
+      "epoch": 1.3846125549485921,
+      "grad_norm": 0.35295653343200684,
+      "learning_rate": 5.813048667902922e-05,
+      "loss": 3.0213,
+      "step": 18820
+    },
+    {
+      "epoch": 1.3853482683146647,
+      "grad_norm": 0.36051633954048157,
+      "learning_rate": 5.8091518696619005e-05,
+      "loss": 3.0247,
+      "step": 18830
+    },
+    {
+      "epoch": 1.3860839816807373,
+      "grad_norm": 0.3562413454055786,
+      "learning_rate": 5.805254566657834e-05,
+      "loss": 2.9649,
+      "step": 18840
+    },
+    {
+      "epoch": 1.3868196950468097,
+      "grad_norm": 0.36182910203933716,
+      "learning_rate": 5.8013567613219286e-05,
+      "loss": 3.025,
+      "step": 18850
+    },
+    {
+      "epoch": 1.3875554084128823,
+      "grad_norm": 0.35930660367012024,
+      "learning_rate": 5.797458456085704e-05,
+      "loss": 2.9987,
+      "step": 18860
+    },
+    {
+      "epoch": 1.3882911217789549,
+      "grad_norm": 0.35059815645217896,
+      "learning_rate": 5.793559653380991e-05,
+      "loss": 2.9675,
+      "step": 18870
+    },
+    {
+      "epoch": 1.3890268351450274,
+      "grad_norm": 0.3475961685180664,
+      "learning_rate": 5.78966035563993e-05,
+      "loss": 3.0161,
+      "step": 18880
+    },
+    {
+      "epoch": 1.3897625485111,
+      "grad_norm": 0.354216992855072,
+      "learning_rate": 5.78576056529497e-05,
+      "loss": 3.0053,
+      "step": 18890
+    },
+    {
+      "epoch": 1.3904982618771726,
+      "grad_norm": 0.3620551824569702,
+      "learning_rate": 5.7818602847788705e-05,
+      "loss": 3.0394,
+      "step": 18900
+    },
+    {
+      "epoch": 1.3912339752432452,
+      "grad_norm": 0.36621057987213135,
+      "learning_rate": 5.777959516524694e-05,
+      "loss": 3.0167,
+      "step": 18910
+    },
+    {
+      "epoch": 1.3919696886093178,
+      "grad_norm": 0.3596706688404083,
+      "learning_rate": 5.774058262965806e-05,
+      "loss": 2.9984,
+      "step": 18920
+    },
+    {
+      "epoch": 1.3927054019753904,
+      "grad_norm": 0.38292568922042847,
+      "learning_rate": 5.770156526535876e-05,
+      "loss": 2.9772,
+      "step": 18930
+    },
+    {
+      "epoch": 1.393441115341463,
+      "grad_norm": 0.3576861023902893,
+      "learning_rate": 5.7662543096688794e-05,
+      "loss": 3.0214,
+      "step": 18940
+    },
+    {
+      "epoch": 1.3941768287075356,
+      "grad_norm": 0.3552986681461334,
+      "learning_rate": 5.762351614799083e-05,
+      "loss": 3.0416,
+      "step": 18950
+    },
+    {
+      "epoch": 1.3949125420736082,
+      "grad_norm": 0.35676777362823486,
+      "learning_rate": 5.7584484443610596e-05,
+      "loss": 2.9983,
+      "step": 18960
+    },
+    {
+      "epoch": 1.3956482554396807,
+      "grad_norm": 0.3552328646183014,
+      "learning_rate": 5.754544800789671e-05,
+      "loss": 2.999,
+      "step": 18970
+    },
+    {
+      "epoch": 1.3963839688057533,
+      "grad_norm": 0.36102887988090515,
+      "learning_rate": 5.750640686520079e-05,
+      "loss": 3.0038,
+      "step": 18980
+    },
+    {
+      "epoch": 1.397119682171826,
+      "grad_norm": 0.34578076004981995,
+      "learning_rate": 5.746736103987741e-05,
+      "loss": 3.025,
+      "step": 18990
+    },
+    {
+      "epoch": 1.3978553955378985,
+      "grad_norm": 0.3563545346260071,
+      "learning_rate": 5.742831055628402e-05,
+      "loss": 3.0183,
+      "step": 19000
+    },
+    {
+      "epoch": 1.398591108903971,
+      "grad_norm": 0.3480781614780426,
+      "learning_rate": 5.738925543878096e-05,
+      "loss": 3.006,
+      "step": 19010
+    },
+    {
+      "epoch": 1.3993268222700435,
+      "grad_norm": 0.34552431106567383,
+      "learning_rate": 5.735019571173155e-05,
+      "loss": 3.022,
+      "step": 19020
+    },
+    {
+      "epoch": 1.400062535636116,
+      "grad_norm": 0.36714527010917664,
+      "learning_rate": 5.7311131399501896e-05,
+      "loss": 3.0075,
+      "step": 19030
+    },
+    {
+      "epoch": 1.4007982490021886,
+      "grad_norm": 0.353105753660202,
+      "learning_rate": 5.7272062526461e-05,
+      "loss": 2.9935,
+      "step": 19040
+    },
+    {
+      "epoch": 1.4015339623682612,
+      "grad_norm": 0.3681223392486572,
+      "learning_rate": 5.7232989116980694e-05,
+      "loss": 3.063,
+      "step": 19050
+    },
+    {
+      "epoch": 1.4022696757343338,
+      "grad_norm": 0.34944233298301697,
+      "learning_rate": 5.71939111954357e-05,
+      "loss": 3.0231,
+      "step": 19060
+    },
+    {
+      "epoch": 1.4030053891004064,
+      "grad_norm": 0.3491586446762085,
+      "learning_rate": 5.715482878620345e-05,
+      "loss": 2.9589,
+      "step": 19070
+    },
+    {
+      "epoch": 1.403741102466479,
+      "grad_norm": 0.36205020546913147,
+      "learning_rate": 5.7115741913664264e-05,
+      "loss": 3.0261,
+      "step": 19080
+    },
+    {
+      "epoch": 1.4044768158325516,
+      "grad_norm": 0.3585648536682129,
+      "learning_rate": 5.7076650602201196e-05,
+      "loss": 3.0739,
+      "step": 19090
+    },
+    {
+      "epoch": 1.4052125291986242,
+      "grad_norm": 0.3495137393474579,
+      "learning_rate": 5.703755487620012e-05,
+      "loss": 2.9956,
+      "step": 19100
+    },
+    {
+      "epoch": 1.4059482425646967,
+      "grad_norm": 0.3515518605709076,
+      "learning_rate": 5.699845476004959e-05,
+      "loss": 3.0148,
+      "step": 19110
+    },
+    {
+      "epoch": 1.4066839559307693,
+      "grad_norm": 0.3399387001991272,
+      "learning_rate": 5.695935027814098e-05,
+      "loss": 2.9858,
+      "step": 19120
+    },
+    {
+      "epoch": 1.407419669296842,
+      "grad_norm": 0.36767157912254333,
+      "learning_rate": 5.6920241454868304e-05,
+      "loss": 3.0431,
+      "step": 19130
+    },
+    {
+      "epoch": 1.4081553826629145,
+      "grad_norm": 0.34016692638397217,
+      "learning_rate": 5.6881128314628365e-05,
+      "loss": 2.9757,
+      "step": 19140
+    },
+    {
+      "epoch": 1.408891096028987,
+      "grad_norm": 0.33596450090408325,
+      "learning_rate": 5.684201088182061e-05,
+      "loss": 3.0168,
+      "step": 19150
+    },
+    {
+      "epoch": 1.4096268093950597,
+      "grad_norm": 0.3481621742248535,
+      "learning_rate": 5.680288918084716e-05,
+      "loss": 3.0254,
+      "step": 19160
+    },
+    {
+      "epoch": 1.4103625227611323,
+      "grad_norm": 0.37596067786216736,
+      "learning_rate": 5.676376323611285e-05,
+      "loss": 3.0812,
+      "step": 19170
+    },
+    {
+      "epoch": 1.4110982361272049,
+      "grad_norm": 0.35009533166885376,
+      "learning_rate": 5.672463307202508e-05,
+      "loss": 3.0074,
+      "step": 19180
+    },
+    {
+      "epoch": 1.4118339494932775,
+      "grad_norm": 0.3589169383049011,
+      "learning_rate": 5.668549871299398e-05,
+      "loss": 3.063,
+      "step": 19190
+    },
+    {
+      "epoch": 1.41256966285935,
+      "grad_norm": 0.36062905192375183,
+      "learning_rate": 5.664636018343219e-05,
+      "loss": 2.9785,
+      "step": 19200
+    },
+    {
+      "epoch": 1.4133053762254226,
+      "grad_norm": 0.3376767933368683,
+      "learning_rate": 5.660721750775506e-05,
+      "loss": 3.026,
+      "step": 19210
+    },
+    {
+      "epoch": 1.4140410895914952,
+      "grad_norm": 0.3527078330516815,
+      "learning_rate": 5.6568070710380427e-05,
+      "loss": 2.9954,
+      "step": 19220
+    },
+    {
+      "epoch": 1.4147768029575678,
+      "grad_norm": 0.350247323513031,
+      "learning_rate": 5.652891981572879e-05,
+      "loss": 2.9385,
+      "step": 19230
+    },
+    {
+      "epoch": 1.4155125163236404,
+      "grad_norm": 0.36022645235061646,
+      "learning_rate": 5.648976484822313e-05,
+      "loss": 3.0135,
+      "step": 19240
+    },
+    {
+      "epoch": 1.416248229689713,
+      "grad_norm": 0.36032214760780334,
+      "learning_rate": 5.6450605832289e-05,
+      "loss": 3.0151,
+      "step": 19250
+    },
+    {
+      "epoch": 1.4169839430557856,
+      "grad_norm": 0.3666633665561676,
+      "learning_rate": 5.64114427923545e-05,
+      "loss": 3.0383,
+      "step": 19260
+    },
+    {
+      "epoch": 1.4177196564218582,
+      "grad_norm": 0.3436683118343353,
+      "learning_rate": 5.637227575285018e-05,
+      "loss": 3.011,
+      "step": 19270
+    },
+    {
+      "epoch": 1.4184553697879307,
+      "grad_norm": 0.34924954175949097,
+      "learning_rate": 5.6333104738209165e-05,
+      "loss": 2.9734,
+      "step": 19280
+    },
+    {
+      "epoch": 1.4191910831540033,
+      "grad_norm": 0.3514811098575592,
+      "learning_rate": 5.629392977286697e-05,
+      "loss": 2.9753,
+      "step": 19290
+    },
+    {
+      "epoch": 1.4199267965200757,
+      "grad_norm": 0.3607385754585266,
+      "learning_rate": 5.6254750881261656e-05,
+      "loss": 2.9863,
+      "step": 19300
+    },
+    {
+      "epoch": 1.4206625098861483,
+      "grad_norm": 0.34886640310287476,
+      "learning_rate": 5.621556808783368e-05,
+      "loss": 3.0356,
+      "step": 19310
+    },
+    {
+      "epoch": 1.4213982232522209,
+      "grad_norm": 0.3406035900115967,
+      "learning_rate": 5.617638141702598e-05,
+      "loss": 3.0221,
+      "step": 19320
+    },
+    {
+      "epoch": 1.4221339366182935,
+      "grad_norm": 0.3472137749195099,
+      "learning_rate": 5.6137190893283854e-05,
+      "loss": 2.9742,
+      "step": 19330
+    },
+    {
+      "epoch": 1.422869649984366,
+      "grad_norm": 0.3494405150413513,
+      "learning_rate": 5.6097996541055034e-05,
+      "loss": 3.0011,
+      "step": 19340
+    },
+    {
+      "epoch": 1.4236053633504386,
+      "grad_norm": 0.36849454045295715,
+      "learning_rate": 5.605879838478967e-05,
+      "loss": 2.9841,
+      "step": 19350
+    },
+    {
+      "epoch": 1.4243410767165112,
+      "grad_norm": 0.3592820167541504,
+      "learning_rate": 5.6019596448940226e-05,
+      "loss": 3.0171,
+      "step": 19360
+    },
+    {
+      "epoch": 1.4250767900825838,
+      "grad_norm": 0.3619770109653473,
+      "learning_rate": 5.598039075796157e-05,
+      "loss": 3.0401,
+      "step": 19370
+    },
+    {
+      "epoch": 1.4258125034486564,
+      "grad_norm": 0.358020156621933,
+      "learning_rate": 5.594118133631087e-05,
+      "loss": 3.0309,
+      "step": 19380
+    },
+    {
+      "epoch": 1.426548216814729,
+      "grad_norm": 0.35207271575927734,
+      "learning_rate": 5.590196820844769e-05,
+      "loss": 2.9897,
+      "step": 19390
+    },
+    {
+      "epoch": 1.4272839301808016,
+      "grad_norm": 0.35930630564689636,
+      "learning_rate": 5.586275139883382e-05,
+      "loss": 2.9978,
+      "step": 19400
+    },
+    {
+      "epoch": 1.4280196435468742,
+      "grad_norm": 0.35888969898223877,
+      "learning_rate": 5.582353093193342e-05,
+      "loss": 3.0145,
+      "step": 19410
+    },
+    {
+      "epoch": 1.4287553569129468,
+      "grad_norm": 0.3499695956707001,
+      "learning_rate": 5.578430683221288e-05,
+      "loss": 2.9246,
+      "step": 19420
+    },
+    {
+      "epoch": 1.4294910702790193,
+      "grad_norm": 0.36832836270332336,
+      "learning_rate": 5.5745079124140897e-05,
+      "loss": 2.9753,
+      "step": 19430
+    },
+    {
+      "epoch": 1.430226783645092,
+      "grad_norm": 0.3586465120315552,
+      "learning_rate": 5.570584783218838e-05,
+      "loss": 3.011,
+      "step": 19440
+    },
+    {
+      "epoch": 1.4309624970111645,
+      "grad_norm": 0.35472095012664795,
+      "learning_rate": 5.5666612980828516e-05,
+      "loss": 3.003,
+      "step": 19450
+    },
+    {
+      "epoch": 1.431698210377237,
+      "grad_norm": 0.34835168719291687,
+      "learning_rate": 5.562737459453665e-05,
+      "loss": 3.0293,
+      "step": 19460
+    },
+    {
+      "epoch": 1.4324339237433097,
+      "grad_norm": 0.35374611616134644,
+      "learning_rate": 5.5588132697790416e-05,
+      "loss": 3.038,
+      "step": 19470
+    },
+    {
+      "epoch": 1.433169637109382,
+      "grad_norm": 0.36069589853286743,
+      "learning_rate": 5.5548887315069575e-05,
+      "loss": 3.0253,
+      "step": 19480
+    },
+    {
+      "epoch": 1.4339053504754546,
+      "grad_norm": 0.3513835668563843,
+      "learning_rate": 5.550963847085607e-05,
+      "loss": 3.017,
+      "step": 19490
+    },
+    {
+      "epoch": 1.4346410638415272,
+      "grad_norm": 0.3436732292175293,
+      "learning_rate": 5.547038618963404e-05,
+      "loss": 2.9984,
+      "step": 19500
+    },
+    {
+      "epoch": 1.4353767772075998,
+      "grad_norm": 0.35367876291275024,
+      "learning_rate": 5.543113049588971e-05,
+      "loss": 3.0072,
+      "step": 19510
+    },
+    {
+      "epoch": 1.4361124905736724,
+      "grad_norm": 0.35975176095962524,
+      "learning_rate": 5.539187141411148e-05,
+      "loss": 2.9966,
+      "step": 19520
+    },
+    {
+      "epoch": 1.436848203939745,
+      "grad_norm": 0.3589760959148407,
+      "learning_rate": 5.5352608968789865e-05,
+      "loss": 2.9629,
+      "step": 19530
+    },
+    {
+      "epoch": 1.4375839173058176,
+      "grad_norm": 0.3542312979698181,
+      "learning_rate": 5.531334318441745e-05,
+      "loss": 3.0035,
+      "step": 19540
+    },
+    {
+      "epoch": 1.4383196306718902,
+      "grad_norm": 0.3446711599826813,
+      "learning_rate": 5.52740740854889e-05,
+      "loss": 2.987,
+      "step": 19550
+    },
+    {
+      "epoch": 1.4390553440379628,
+      "grad_norm": 0.3555561900138855,
+      "learning_rate": 5.5234801696500994e-05,
+      "loss": 2.9774,
+      "step": 19560
+    },
+    {
+      "epoch": 1.4397910574040353,
+      "grad_norm": 0.35286226868629456,
+      "learning_rate": 5.519552604195253e-05,
+      "loss": 2.9931,
+      "step": 19570
+    },
+    {
+      "epoch": 1.440526770770108,
+      "grad_norm": 0.3550855815410614,
+      "learning_rate": 5.515624714634432e-05,
+      "loss": 2.9978,
+      "step": 19580
+    },
+    {
+      "epoch": 1.4412624841361805,
+      "grad_norm": 0.3527701497077942,
+      "learning_rate": 5.5116965034179235e-05,
+      "loss": 3.0296,
+      "step": 19590
+    },
+    {
+      "epoch": 1.441998197502253,
+      "grad_norm": 0.3383554220199585,
+      "learning_rate": 5.5077679729962164e-05,
+      "loss": 2.9859,
+      "step": 19600
+    },
+    {
+      "epoch": 1.4427339108683257,
+      "grad_norm": 0.34945011138916016,
+      "learning_rate": 5.503839125819993e-05,
+      "loss": 3.0195,
+      "step": 19610
+    },
+    {
+      "epoch": 1.4434696242343983,
+      "grad_norm": 0.3590049147605896,
+      "learning_rate": 5.49990996434014e-05,
+      "loss": 2.9736,
+      "step": 19620
+    },
+    {
+      "epoch": 1.4442053376004709,
+      "grad_norm": 0.3441041111946106,
+      "learning_rate": 5.495980491007732e-05,
+      "loss": 2.9636,
+      "step": 19630
+    },
+    {
+      "epoch": 1.4449410509665435,
+      "grad_norm": 0.3660474419593811,
+      "learning_rate": 5.492050708274049e-05,
+      "loss": 3.0069,
+      "step": 19640
+    },
+    {
+      "epoch": 1.445676764332616,
+      "grad_norm": 0.3662777543067932,
+      "learning_rate": 5.488120618590554e-05,
+      "loss": 3.0558,
+      "step": 19650
+    },
+    {
+      "epoch": 1.4464124776986886,
+      "grad_norm": 0.3546368479728699,
+      "learning_rate": 5.484190224408907e-05,
+      "loss": 3.0265,
+      "step": 19660
+    },
+    {
+      "epoch": 1.4471481910647612,
+      "grad_norm": 0.3501317799091339,
+      "learning_rate": 5.4802595281809545e-05,
+      "loss": 3.0249,
+      "step": 19670
+    },
+    {
+      "epoch": 1.4478839044308338,
+      "grad_norm": 0.35697367787361145,
+      "learning_rate": 5.4763285323587366e-05,
+      "loss": 3.0422,
+      "step": 19680
+    },
+    {
+      "epoch": 1.4486196177969064,
+      "grad_norm": 0.3489791452884674,
+      "learning_rate": 5.472397239394475e-05,
+      "loss": 2.9608,
+      "step": 19690
+    },
+    {
+      "epoch": 1.449355331162979,
+      "grad_norm": 0.36893218755722046,
+      "learning_rate": 5.468465651740579e-05,
+      "loss": 3.0123,
+      "step": 19700
+    },
+    {
+      "epoch": 1.4500910445290516,
+      "grad_norm": 0.3482329547405243,
+      "learning_rate": 5.46453377184964e-05,
+      "loss": 3.0066,
+      "step": 19710
+    },
+    {
+      "epoch": 1.4508267578951242,
+      "grad_norm": 0.3464795649051666,
+      "learning_rate": 5.460601602174437e-05,
+      "loss": 2.9805,
+      "step": 19720
+    },
+    {
+      "epoch": 1.4515624712611968,
+      "grad_norm": 0.348219633102417,
+      "learning_rate": 5.456669145167923e-05,
+      "loss": 2.9534,
+      "step": 19730
+    },
+    {
+      "epoch": 1.4522981846272693,
+      "grad_norm": 0.3509849011898041,
+      "learning_rate": 5.452736403283234e-05,
+      "loss": 3.0034,
+      "step": 19740
+    },
+    {
+      "epoch": 1.453033897993342,
+      "grad_norm": 0.35270920395851135,
+      "learning_rate": 5.448803378973683e-05,
+      "loss": 3.0058,
+      "step": 19750
+    },
+    {
+      "epoch": 1.4537696113594143,
+      "grad_norm": 0.36120450496673584,
+      "learning_rate": 5.4448700746927584e-05,
+      "loss": 3.0163,
+      "step": 19760
+    },
+    {
+      "epoch": 1.4545053247254869,
+      "grad_norm": 0.34886208176612854,
+      "learning_rate": 5.4409364928941266e-05,
+      "loss": 2.9977,
+      "step": 19770
+    },
+    {
+      "epoch": 1.4552410380915595,
+      "grad_norm": 0.36098363995552063,
+      "learning_rate": 5.437002636031621e-05,
+      "loss": 2.9641,
+      "step": 19780
+    },
+    {
+      "epoch": 1.455976751457632,
+      "grad_norm": 0.36371397972106934,
+      "learning_rate": 5.4330685065592514e-05,
+      "loss": 3.0371,
+      "step": 19790
+    },
+    {
+      "epoch": 1.4567124648237046,
+      "grad_norm": 0.35335397720336914,
+      "learning_rate": 5.4291341069311986e-05,
+      "loss": 3.002,
+      "step": 19800
+    },
+    {
+      "epoch": 1.4574481781897772,
+      "grad_norm": 0.35094645619392395,
+      "learning_rate": 5.4251994396018056e-05,
+      "loss": 3.0001,
+      "step": 19810
+    },
+    {
+      "epoch": 1.4581838915558498,
+      "grad_norm": 0.33477920293807983,
+      "learning_rate": 5.421264507025591e-05,
+      "loss": 2.952,
+      "step": 19820
+    },
+    {
+      "epoch": 1.4589196049219224,
+      "grad_norm": 0.3566073477268219,
+      "learning_rate": 5.417329311657231e-05,
+      "loss": 2.9642,
+      "step": 19830
+    },
+    {
+      "epoch": 1.459655318287995,
+      "grad_norm": 0.3429076373577118,
+      "learning_rate": 5.4133938559515695e-05,
+      "loss": 3.0022,
+      "step": 19840
+    },
+    {
+      "epoch": 1.4603910316540676,
+      "grad_norm": 0.3418024182319641,
+      "learning_rate": 5.4094581423636126e-05,
+      "loss": 3.0266,
+      "step": 19850
+    },
+    {
+      "epoch": 1.4611267450201402,
+      "grad_norm": 0.3650926351547241,
+      "learning_rate": 5.4055221733485295e-05,
+      "loss": 3.025,
+      "step": 19860
+    },
+    {
+      "epoch": 1.4618624583862128,
+      "grad_norm": 0.3517318367958069,
+      "learning_rate": 5.401585951361642e-05,
+      "loss": 3.026,
+      "step": 19870
+    },
+    {
+      "epoch": 1.4625981717522853,
+      "grad_norm": 0.3494042456150055,
+      "learning_rate": 5.397649478858438e-05,
+      "loss": 3.0254,
+      "step": 19880
+    },
+    {
+      "epoch": 1.463333885118358,
+      "grad_norm": 0.35053548216819763,
+      "learning_rate": 5.3937127582945546e-05,
+      "loss": 2.9691,
+      "step": 19890
+    },
+    {
+      "epoch": 1.4640695984844305,
+      "grad_norm": 0.3622230887413025,
+      "learning_rate": 5.389775792125789e-05,
+      "loss": 2.9889,
+      "step": 19900
+    },
+    {
+      "epoch": 1.4648053118505031,
+      "grad_norm": 0.3651433289051056,
+      "learning_rate": 5.385838582808088e-05,
+      "loss": 3.0342,
+      "step": 19910
+    },
+    {
+      "epoch": 1.4655410252165757,
+      "grad_norm": 0.34760138392448425,
+      "learning_rate": 5.381901132797551e-05,
+      "loss": 2.9634,
+      "step": 19920
+    },
+    {
+      "epoch": 1.466276738582648,
+      "grad_norm": 0.37648308277130127,
+      "learning_rate": 5.37796344455043e-05,
+      "loss": 3.0255,
+      "step": 19930
+    },
+    {
+      "epoch": 1.4670124519487207,
+      "grad_norm": 0.38236573338508606,
+      "learning_rate": 5.374025520523122e-05,
+      "loss": 2.9747,
+      "step": 19940
+    },
+    {
+      "epoch": 1.4677481653147932,
+      "grad_norm": 0.37035325169563293,
+      "learning_rate": 5.370087363172174e-05,
+      "loss": 3.0109,
+      "step": 19950
+    },
+    {
+      "epoch": 1.4684838786808658,
+      "grad_norm": 0.3538261950016022,
+      "learning_rate": 5.3661489749542746e-05,
+      "loss": 3.0177,
+      "step": 19960
+    },
+    {
+      "epoch": 1.4692195920469384,
+      "grad_norm": 0.3587804436683655,
+      "learning_rate": 5.362210358326261e-05,
+      "loss": 3.0038,
+      "step": 19970
+    },
+    {
+      "epoch": 1.469955305413011,
+      "grad_norm": 0.35976141691207886,
+      "learning_rate": 5.358271515745111e-05,
+      "loss": 2.9671,
+      "step": 19980
+    },
+    {
+      "epoch": 1.4706910187790836,
+      "grad_norm": 0.3535846769809723,
+      "learning_rate": 5.3543324496679424e-05,
+      "loss": 3.0011,
+      "step": 19990
+    },
+    {
+      "epoch": 1.4714267321451562,
+      "grad_norm": 0.3491998016834259,
+      "learning_rate": 5.350393162552014e-05,
+      "loss": 3.0375,
+      "step": 20000
+    },
+    {
+      "epoch": 1.4721624455112288,
+      "grad_norm": 0.35548117756843567,
+      "learning_rate": 5.346453656854722e-05,
+      "loss": 2.9712,
+      "step": 20010
+    },
+    {
+      "epoch": 1.4728981588773014,
+      "grad_norm": 0.35048940777778625,
+      "learning_rate": 5.342513935033598e-05,
+      "loss": 3.0057,
+      "step": 20020
+    },
+    {
+      "epoch": 1.473633872243374,
+      "grad_norm": 0.36360570788383484,
+      "learning_rate": 5.338573999546309e-05,
+      "loss": 2.9721,
+      "step": 20030
+    },
+    {
+      "epoch": 1.4743695856094465,
+      "grad_norm": 0.351386696100235,
+      "learning_rate": 5.3346338528506565e-05,
+      "loss": 3.0213,
+      "step": 20040
+    },
+    {
+      "epoch": 1.4751052989755191,
+      "grad_norm": 0.339138001203537,
+      "learning_rate": 5.330693497404573e-05,
+      "loss": 2.9765,
+      "step": 20050
+    },
+    {
+      "epoch": 1.4758410123415917,
+      "grad_norm": 0.3340703248977661,
+      "learning_rate": 5.3267529356661183e-05,
+      "loss": 3.0367,
+      "step": 20060
+    },
+    {
+      "epoch": 1.4765767257076643,
+      "grad_norm": 0.3471696078777313,
+      "learning_rate": 5.322812170093487e-05,
+      "loss": 3.0622,
+      "step": 20070
+    },
+    {
+      "epoch": 1.4773124390737369,
+      "grad_norm": 0.35119909048080444,
+      "learning_rate": 5.318871203144994e-05,
+      "loss": 3.0206,
+      "step": 20080
+    },
+    {
+      "epoch": 1.4780481524398095,
+      "grad_norm": 0.3459704518318176,
+      "learning_rate": 5.314930037279086e-05,
+      "loss": 3.0198,
+      "step": 20090
+    },
+    {
+      "epoch": 1.478783865805882,
+      "grad_norm": 0.3650210499763489,
+      "learning_rate": 5.310988674954329e-05,
+      "loss": 2.968,
+      "step": 20100
+    },
+    {
+      "epoch": 1.4795195791719546,
+      "grad_norm": 0.3578377962112427,
+      "learning_rate": 5.3070471186294157e-05,
+      "loss": 2.9759,
+      "step": 20110
+    },
+    {
+      "epoch": 1.4802552925380272,
+      "grad_norm": 0.35734254121780396,
+      "learning_rate": 5.303105370763154e-05,
+      "loss": 2.9862,
+      "step": 20120
+    },
+    {
+      "epoch": 1.4809910059040998,
+      "grad_norm": 0.3591006398200989,
+      "learning_rate": 5.299163433814479e-05,
+      "loss": 3.038,
+      "step": 20130
+    },
+    {
+      "epoch": 1.4817267192701724,
+      "grad_norm": 0.3443552553653717,
+      "learning_rate": 5.2952213102424376e-05,
+      "loss": 2.9784,
+      "step": 20140
+    },
+    {
+      "epoch": 1.482462432636245,
+      "grad_norm": 0.3936735689640045,
+      "learning_rate": 5.291279002506196e-05,
+      "loss": 2.9912,
+      "step": 20150
+    },
+    {
+      "epoch": 1.4831981460023176,
+      "grad_norm": 0.35305315256118774,
+      "learning_rate": 5.287336513065033e-05,
+      "loss": 2.9445,
+      "step": 20160
+    },
+    {
+      "epoch": 1.4839338593683902,
+      "grad_norm": 0.35259613394737244,
+      "learning_rate": 5.283393844378344e-05,
+      "loss": 2.9756,
+      "step": 20170
+    },
+    {
+      "epoch": 1.4846695727344628,
+      "grad_norm": 0.35446780920028687,
+      "learning_rate": 5.279450998905635e-05,
+      "loss": 3.0088,
+      "step": 20180
+    },
+    {
+      "epoch": 1.4854052861005353,
+      "grad_norm": 0.3534410297870636,
+      "learning_rate": 5.2755079791065186e-05,
+      "loss": 2.9781,
+      "step": 20190
+    },
+    {
+      "epoch": 1.486140999466608,
+      "grad_norm": 0.3572654724121094,
+      "learning_rate": 5.2715647874407225e-05,
+      "loss": 2.9677,
+      "step": 20200
+    },
+    {
+      "epoch": 1.4868767128326805,
+      "grad_norm": 0.3555501699447632,
+      "learning_rate": 5.2676214263680754e-05,
+      "loss": 2.9863,
+      "step": 20210
+    },
+    {
+      "epoch": 1.487612426198753,
+      "grad_norm": 0.3545670807361603,
+      "learning_rate": 5.263677898348519e-05,
+      "loss": 3.0322,
+      "step": 20220
+    },
+    {
+      "epoch": 1.4883481395648255,
+      "grad_norm": 0.3701029419898987,
+      "learning_rate": 5.2597342058420906e-05,
+      "loss": 2.991,
+      "step": 20230
+    },
+    {
+      "epoch": 1.489083852930898,
+      "grad_norm": 0.36001625657081604,
+      "learning_rate": 5.255790351308937e-05,
+      "loss": 2.9637,
+      "step": 20240
+    },
+    {
+      "epoch": 1.4898195662969707,
+      "grad_norm": 0.3514394760131836,
+      "learning_rate": 5.2518463372093016e-05,
+      "loss": 3.0048,
+      "step": 20250
+    },
+    {
+      "epoch": 1.4905552796630432,
+      "grad_norm": 0.34818950295448303,
+      "learning_rate": 5.24790216600353e-05,
+      "loss": 2.9999,
+      "step": 20260
+    },
+    {
+      "epoch": 1.4912909930291158,
+      "grad_norm": 0.3456975519657135,
+      "learning_rate": 5.243957840152066e-05,
+      "loss": 2.9657,
+      "step": 20270
+    },
+    {
+      "epoch": 1.4920267063951884,
+      "grad_norm": 0.3583052456378937,
+      "learning_rate": 5.2400133621154465e-05,
+      "loss": 2.975,
+      "step": 20280
+    },
+    {
+      "epoch": 1.492762419761261,
+      "grad_norm": 0.35923466086387634,
+      "learning_rate": 5.236068734354308e-05,
+      "loss": 2.984,
+      "step": 20290
+    },
+    {
+      "epoch": 1.4934981331273336,
+      "grad_norm": 0.3538628816604614,
+      "learning_rate": 5.2321239593293755e-05,
+      "loss": 2.989,
+      "step": 20300
+    },
+    {
+      "epoch": 1.4942338464934062,
+      "grad_norm": 0.3601844906806946,
+      "learning_rate": 5.228179039501473e-05,
+      "loss": 3.0116,
+      "step": 20310
+    },
+    {
+      "epoch": 1.4949695598594788,
+      "grad_norm": 0.3626857399940491,
+      "learning_rate": 5.224233977331505e-05,
+      "loss": 2.9883,
+      "step": 20320
+    },
+    {
+      "epoch": 1.4957052732255514,
+      "grad_norm": 0.3615439236164093,
+      "learning_rate": 5.220288775280473e-05,
+      "loss": 2.9804,
+      "step": 20330
+    },
+    {
+      "epoch": 1.496440986591624,
+      "grad_norm": 0.34773311018943787,
+      "learning_rate": 5.216343435809462e-05,
+      "loss": 2.9852,
+      "step": 20340
+    },
+    {
+      "epoch": 1.4971766999576965,
+      "grad_norm": 0.3425402343273163,
+      "learning_rate": 5.2123979613796435e-05,
+      "loss": 3.0156,
+      "step": 20350
+    },
+    {
+      "epoch": 1.4979124133237691,
+      "grad_norm": 0.34281396865844727,
+      "learning_rate": 5.208452354452274e-05,
+      "loss": 2.9815,
+      "step": 20360
+    },
+    {
+      "epoch": 1.4986481266898417,
+      "grad_norm": 0.347790002822876,
+      "learning_rate": 5.20450661748869e-05,
+      "loss": 3.0072,
+      "step": 20370
+    },
+    {
+      "epoch": 1.4993838400559143,
+      "grad_norm": 0.3499857187271118,
+      "learning_rate": 5.200560752950314e-05,
+      "loss": 2.993,
+      "step": 20380
+    },
+    {
+      "epoch": 1.5001195534219867,
+      "grad_norm": 0.3523205816745758,
+      "learning_rate": 5.1966147632986426e-05,
+      "loss": 2.9967,
+      "step": 20390
+    },
+    {
+      "epoch": 1.5008552667880593,
+      "grad_norm": 0.37040430307388306,
+      "learning_rate": 5.192668650995255e-05,
+      "loss": 2.9687,
+      "step": 20400
+    },
+    {
+      "epoch": 1.5015909801541318,
+      "grad_norm": 0.36892595887184143,
+      "learning_rate": 5.188722418501802e-05,
+      "loss": 3.0083,
+      "step": 20410
+    },
+    {
+      "epoch": 1.5023266935202044,
+      "grad_norm": 0.34257280826568604,
+      "learning_rate": 5.184776068280015e-05,
+      "loss": 2.9348,
+      "step": 20420
+    },
+    {
+      "epoch": 1.503062406886277,
+      "grad_norm": 0.34466466307640076,
+      "learning_rate": 5.1808296027916945e-05,
+      "loss": 3.0237,
+      "step": 20430
+    },
+    {
+      "epoch": 1.5037981202523496,
+      "grad_norm": 0.3632475435733795,
+      "learning_rate": 5.1768830244987157e-05,
+      "loss": 2.9647,
+      "step": 20440
+    },
+    {
+      "epoch": 1.5045338336184222,
+      "grad_norm": 0.37145963311195374,
+      "learning_rate": 5.17293633586302e-05,
+      "loss": 2.965,
+      "step": 20450
+    },
+    {
+      "epoch": 1.5052695469844948,
+      "grad_norm": 0.3636644780635834,
+      "learning_rate": 5.1689895393466246e-05,
+      "loss": 3.0613,
+      "step": 20460
+    },
+    {
+      "epoch": 1.5060052603505674,
+      "grad_norm": 0.36894574761390686,
+      "learning_rate": 5.165042637411609e-05,
+      "loss": 3.0295,
+      "step": 20470
+    },
+    {
+      "epoch": 1.50674097371664,
+      "grad_norm": 0.35078680515289307,
+      "learning_rate": 5.161095632520117e-05,
+      "loss": 3.0105,
+      "step": 20480
+    },
+    {
+      "epoch": 1.5074766870827125,
+      "grad_norm": 0.36884045600891113,
+      "learning_rate": 5.15714852713436e-05,
+      "loss": 3.048,
+      "step": 20490
+    },
+    {
+      "epoch": 1.5082124004487851,
+      "grad_norm": 0.35813096165657043,
+      "learning_rate": 5.1532013237166145e-05,
+      "loss": 2.9843,
+      "step": 20500
+    },
+    {
+      "epoch": 1.5089481138148577,
+      "grad_norm": 0.3699916899204254,
+      "learning_rate": 5.14925402472921e-05,
+      "loss": 3.0267,
+      "step": 20510
+    },
+    {
+      "epoch": 1.5096838271809303,
+      "grad_norm": 0.3491016626358032,
+      "learning_rate": 5.1453066326345434e-05,
+      "loss": 2.975,
+      "step": 20520
+    },
+    {
+      "epoch": 1.510419540547003,
+      "grad_norm": 0.35231319069862366,
+      "learning_rate": 5.141359149895064e-05,
+      "loss": 3.0061,
+      "step": 20530
+    },
+    {
+      "epoch": 1.5111552539130755,
+      "grad_norm": 0.33822643756866455,
+      "learning_rate": 5.137411578973286e-05,
+      "loss": 3.0101,
+      "step": 20540
+    },
+    {
+      "epoch": 1.511890967279148,
+      "grad_norm": 0.344269722700119,
+      "learning_rate": 5.133463922331766e-05,
+      "loss": 3.0558,
+      "step": 20550
+    },
+    {
+      "epoch": 1.5126266806452207,
+      "grad_norm": 0.3482740521430969,
+      "learning_rate": 5.129516182433126e-05,
+      "loss": 2.9624,
+      "step": 20560
+    },
+    {
+      "epoch": 1.5133623940112932,
+      "grad_norm": 0.3632541596889496,
+      "learning_rate": 5.125568361740033e-05,
+      "loss": 3.0418,
+      "step": 20570
+    },
+    {
+      "epoch": 1.5140981073773658,
+      "grad_norm": 0.34779149293899536,
+      "learning_rate": 5.1216204627152054e-05,
+      "loss": 3.0098,
+      "step": 20580
+    },
+    {
+      "epoch": 1.5148338207434384,
+      "grad_norm": 0.36604106426239014,
+      "learning_rate": 5.1176724878214134e-05,
+      "loss": 2.9949,
+      "step": 20590
+    },
+    {
+      "epoch": 1.515569534109511,
+      "grad_norm": 0.34307625889778137,
+      "learning_rate": 5.113724439521472e-05,
+      "loss": 2.9966,
+      "step": 20600
+    },
+    {
+      "epoch": 1.5163052474755836,
+      "grad_norm": 0.34036216139793396,
+      "learning_rate": 5.109776320278242e-05,
+      "loss": 2.9792,
+      "step": 20610
+    },
+    {
+      "epoch": 1.5170409608416562,
+      "grad_norm": 0.34819355607032776,
+      "learning_rate": 5.10582813255463e-05,
+      "loss": 2.9499,
+      "step": 20620
+    },
+    {
+      "epoch": 1.5177766742077288,
+      "grad_norm": 0.3420320153236389,
+      "learning_rate": 5.101879878813585e-05,
+      "loss": 3.0095,
+      "step": 20630
+    },
+    {
+      "epoch": 1.5185123875738014,
+      "grad_norm": 0.3523239195346832,
+      "learning_rate": 5.097931561518092e-05,
+      "loss": 2.998,
+      "step": 20640
+    },
+    {
+      "epoch": 1.519248100939874,
+      "grad_norm": 0.36634331941604614,
+      "learning_rate": 5.093983183131187e-05,
+      "loss": 3.0019,
+      "step": 20650
+    },
+    {
+      "epoch": 1.5199838143059465,
+      "grad_norm": 0.35214874148368835,
+      "learning_rate": 5.09003474611593e-05,
+      "loss": 2.9833,
+      "step": 20660
+    },
+    {
+      "epoch": 1.5207195276720191,
+      "grad_norm": 0.3431958854198456,
+      "learning_rate": 5.086086252935431e-05,
+      "loss": 3.0192,
+      "step": 20670
+    },
+    {
+      "epoch": 1.5214552410380917,
+      "grad_norm": 0.3640722930431366,
+      "learning_rate": 5.082137706052826e-05,
+      "loss": 3.0105,
+      "step": 20680
+    },
+    {
+      "epoch": 1.5221909544041643,
+      "grad_norm": 0.3478107750415802,
+      "learning_rate": 5.078189107931289e-05,
+      "loss": 3.002,
+      "step": 20690
+    },
+    {
+      "epoch": 1.5229266677702367,
+      "grad_norm": 0.3447904586791992,
+      "learning_rate": 5.0742404610340224e-05,
+      "loss": 2.9605,
+      "step": 20700
+    },
+    {
+      "epoch": 1.5236623811363093,
+      "grad_norm": 0.354278028011322,
+      "learning_rate": 5.070291767824261e-05,
+      "loss": 2.9541,
+      "step": 20710
+    },
+    {
+      "epoch": 1.5243980945023818,
+      "grad_norm": 0.3521896004676819,
+      "learning_rate": 5.06634303076527e-05,
+      "loss": 3.0014,
+      "step": 20720
+    },
+    {
+      "epoch": 1.5251338078684544,
+      "grad_norm": 0.3770923316478729,
+      "learning_rate": 5.062394252320339e-05,
+      "loss": 2.9514,
+      "step": 20730
+    },
+    {
+      "epoch": 1.525869521234527,
+      "grad_norm": 0.38161560893058777,
+      "learning_rate": 5.0584454349527865e-05,
+      "loss": 3.0046,
+      "step": 20740
+    },
+    {
+      "epoch": 1.5266052346005996,
+      "grad_norm": 0.3636596202850342,
+      "learning_rate": 5.0544965811259516e-05,
+      "loss": 3.0261,
+      "step": 20750
+    },
+    {
+      "epoch": 1.5273409479666722,
+      "grad_norm": 0.3606647551059723,
+      "learning_rate": 5.0505476933032006e-05,
+      "loss": 2.9421,
+      "step": 20760
+    },
+    {
+      "epoch": 1.5280766613327448,
+      "grad_norm": 0.3498188853263855,
+      "learning_rate": 5.046598773947916e-05,
+      "loss": 3.0272,
+      "step": 20770
+    },
+    {
+      "epoch": 1.5288123746988174,
+      "grad_norm": 0.3597339689731598,
+      "learning_rate": 5.042649825523503e-05,
+      "loss": 3.0024,
+      "step": 20780
+    },
+    {
+      "epoch": 1.52954808806489,
+      "grad_norm": 0.3477785885334015,
+      "learning_rate": 5.0387008504933876e-05,
+      "loss": 2.975,
+      "step": 20790
+    },
+    {
+      "epoch": 1.5302838014309625,
+      "grad_norm": 0.35603222250938416,
+      "learning_rate": 5.0347518513210054e-05,
+      "loss": 2.9693,
+      "step": 20800
+    },
+    {
+      "epoch": 1.5310195147970351,
+      "grad_norm": 0.3296222686767578,
+      "learning_rate": 5.030802830469814e-05,
+      "loss": 2.9487,
+      "step": 20810
+    },
+    {
+      "epoch": 1.5317552281631075,
+      "grad_norm": 0.34237420558929443,
+      "learning_rate": 5.0268537904032765e-05,
+      "loss": 2.9711,
+      "step": 20820
+    },
+    {
+      "epoch": 1.53249094152918,
+      "grad_norm": 0.3486603796482086,
+      "learning_rate": 5.02290473358488e-05,
+      "loss": 2.9815,
+      "step": 20830
+    },
+    {
+      "epoch": 1.5332266548952527,
+      "grad_norm": 0.353980153799057,
+      "learning_rate": 5.018955662478109e-05,
+      "loss": 2.9422,
+      "step": 20840
+    },
+    {
+      "epoch": 1.5339623682613253,
+      "grad_norm": 0.3858807384967804,
+      "learning_rate": 5.015006579546467e-05,
+      "loss": 3.0054,
+      "step": 20850
+    },
+    {
+      "epoch": 1.5346980816273978,
+      "grad_norm": 0.3703741431236267,
+      "learning_rate": 5.011057487253458e-05,
+      "loss": 2.9556,
+      "step": 20860
+    },
+    {
+      "epoch": 1.5354337949934704,
+      "grad_norm": 0.35528990626335144,
+      "learning_rate": 5.0071083880625945e-05,
+      "loss": 2.9671,
+      "step": 20870
+    },
+    {
+      "epoch": 1.536169508359543,
+      "grad_norm": 0.3363390266895294,
+      "learning_rate": 5.003159284437398e-05,
+      "loss": 2.9662,
+      "step": 20880
+    },
+    {
+      "epoch": 1.5369052217256156,
+      "grad_norm": 0.35573163628578186,
+      "learning_rate": 4.99921017884138e-05,
+      "loss": 2.9466,
+      "step": 20890
+    },
+    {
+      "epoch": 1.5376409350916882,
+      "grad_norm": 0.3478720486164093,
+      "learning_rate": 4.99526107373807e-05,
+      "loss": 3.0259,
+      "step": 20900
+    },
+    {
+      "epoch": 1.5383766484577608,
+      "grad_norm": 0.3386286497116089,
+      "learning_rate": 4.991311971590983e-05,
+      "loss": 3.0224,
+      "step": 20910
+    },
+    {
+      "epoch": 1.5391123618238334,
+      "grad_norm": 0.3773747384548187,
+      "learning_rate": 4.987362874863641e-05,
+      "loss": 2.9923,
+      "step": 20920
+    },
+    {
+      "epoch": 1.539848075189906,
+      "grad_norm": 0.3603582978248596,
+      "learning_rate": 4.983413786019557e-05,
+      "loss": 2.9741,
+      "step": 20930
+    },
+    {
+      "epoch": 1.5405837885559786,
+      "grad_norm": 0.3472493588924408,
+      "learning_rate": 4.9794647075222436e-05,
+      "loss": 3.0134,
+      "step": 20940
+    },
+    {
+      "epoch": 1.5413195019220511,
+      "grad_norm": 0.3513439893722534,
+      "learning_rate": 4.975515641835202e-05,
+      "loss": 2.9829,
+      "step": 20950
+    },
+    {
+      "epoch": 1.5420552152881237,
+      "grad_norm": 0.35026177763938904,
+      "learning_rate": 4.9715665914219316e-05,
+      "loss": 3.0236,
+      "step": 20960
+    },
+    {
+      "epoch": 1.5427909286541963,
+      "grad_norm": 0.35222265124320984,
+      "learning_rate": 4.9676175587459154e-05,
+      "loss": 3.0362,
+      "step": 20970
+    },
+    {
+      "epoch": 1.543526642020269,
+      "grad_norm": 0.3497045338153839,
+      "learning_rate": 4.963668546270631e-05,
+      "loss": 2.9832,
+      "step": 20980
+    },
+    {
+      "epoch": 1.5442623553863415,
+      "grad_norm": 0.35515788197517395,
+      "learning_rate": 4.959719556459542e-05,
+      "loss": 3.0581,
+      "step": 20990
+    },
+    {
+      "epoch": 1.544998068752414,
+      "grad_norm": 0.3672070801258087,
+      "learning_rate": 4.955770591776095e-05,
+      "loss": 3.0008,
+      "step": 21000
+    },
+    {
+      "epoch": 1.5457337821184867,
+      "grad_norm": 0.3461863696575165,
+      "learning_rate": 4.9518216546837254e-05,
+      "loss": 2.9732,
+      "step": 21010
+    },
+    {
+      "epoch": 1.5464694954845593,
+      "grad_norm": 0.3484613597393036,
+      "learning_rate": 4.9478727476458464e-05,
+      "loss": 3.0394,
+      "step": 21020
+    },
+    {
+      "epoch": 1.5472052088506318,
+      "grad_norm": 0.35905250906944275,
+      "learning_rate": 4.9439238731258576e-05,
+      "loss": 2.9394,
+      "step": 21030
+    },
+    {
+      "epoch": 1.5479409222167044,
+      "grad_norm": 0.3553624451160431,
+      "learning_rate": 4.939975033587132e-05,
+      "loss": 2.9815,
+      "step": 21040
+    },
+    {
+      "epoch": 1.548676635582777,
+      "grad_norm": 0.34723177552223206,
+      "learning_rate": 4.936026231493028e-05,
+      "loss": 2.9694,
+      "step": 21050
+    },
+    {
+      "epoch": 1.5494123489488496,
+      "grad_norm": 0.3437559902667999,
+      "learning_rate": 4.932077469306877e-05,
+      "loss": 2.9739,
+      "step": 21060
+    },
+    {
+      "epoch": 1.5501480623149222,
+      "grad_norm": 0.3376007676124573,
+      "learning_rate": 4.928128749491984e-05,
+      "loss": 2.9767,
+      "step": 21070
+    },
+    {
+      "epoch": 1.5508837756809948,
+      "grad_norm": 0.3432474434375763,
+      "learning_rate": 4.924180074511632e-05,
+      "loss": 2.9753,
+      "step": 21080
+    },
+    {
+      "epoch": 1.5516194890470674,
+      "grad_norm": 0.35019999742507935,
+      "learning_rate": 4.9202314468290696e-05,
+      "loss": 2.9549,
+      "step": 21090
+    },
+    {
+      "epoch": 1.55235520241314,
+      "grad_norm": 0.35744142532348633,
+      "learning_rate": 4.916282868907523e-05,
+      "loss": 2.9914,
+      "step": 21100
+    },
+    {
+      "epoch": 1.5530909157792125,
+      "grad_norm": 0.35550180077552795,
+      "learning_rate": 4.912334343210181e-05,
+      "loss": 3.0053,
+      "step": 21110
+    },
+    {
+      "epoch": 1.5538266291452851,
+      "grad_norm": 0.360637366771698,
+      "learning_rate": 4.908385872200205e-05,
+      "loss": 2.9974,
+      "step": 21120
+    },
+    {
+      "epoch": 1.5545623425113577,
+      "grad_norm": 0.35872945189476013,
+      "learning_rate": 4.904437458340718e-05,
+      "loss": 3.0009,
+      "step": 21130
+    },
+    {
+      "epoch": 1.5552980558774303,
+      "grad_norm": 0.3521031141281128,
+      "learning_rate": 4.900489104094811e-05,
+      "loss": 2.9762,
+      "step": 21140
+    },
+    {
+      "epoch": 1.556033769243503,
+      "grad_norm": 0.34354451298713684,
+      "learning_rate": 4.896540811925537e-05,
+      "loss": 2.9437,
+      "step": 21150
+    },
+    {
+      "epoch": 1.5567694826095753,
+      "grad_norm": 0.349997878074646,
+      "learning_rate": 4.892592584295907e-05,
+      "loss": 3.0339,
+      "step": 21160
+    },
+    {
+      "epoch": 1.5575051959756478,
+      "grad_norm": 0.3518412411212921,
+      "learning_rate": 4.888644423668897e-05,
+      "loss": 2.9668,
+      "step": 21170
+    },
+    {
+      "epoch": 1.5582409093417204,
+      "grad_norm": 0.35249418020248413,
+      "learning_rate": 4.8846963325074366e-05,
+      "loss": 2.9943,
+      "step": 21180
+    },
+    {
+      "epoch": 1.558976622707793,
+      "grad_norm": 0.3495465815067291,
+      "learning_rate": 4.880748313274415e-05,
+      "loss": 2.9501,
+      "step": 21190
+    },
+    {
+      "epoch": 1.5597123360738656,
+      "grad_norm": 0.3538327217102051,
+      "learning_rate": 4.8768003684326733e-05,
+      "loss": 2.9534,
+      "step": 21200
+    },
+    {
+      "epoch": 1.5604480494399382,
+      "grad_norm": 0.35595381259918213,
+      "learning_rate": 4.872852500445009e-05,
+      "loss": 3.0066,
+      "step": 21210
+    },
+    {
+      "epoch": 1.5611837628060108,
+      "grad_norm": 0.35275477170944214,
+      "learning_rate": 4.8689047117741736e-05,
+      "loss": 2.9718,
+      "step": 21220
+    },
+    {
+      "epoch": 1.5619194761720834,
+      "grad_norm": 0.3636651337146759,
+      "learning_rate": 4.864957004882864e-05,
+      "loss": 2.9624,
+      "step": 21230
+    },
+    {
+      "epoch": 1.562655189538156,
+      "grad_norm": 0.34316369891166687,
+      "learning_rate": 4.861009382233731e-05,
+      "loss": 3.0001,
+      "step": 21240
+    },
+    {
+      "epoch": 1.5633909029042286,
+      "grad_norm": 0.3631385862827301,
+      "learning_rate": 4.857061846289366e-05,
+      "loss": 3.021,
+      "step": 21250
+    },
+    {
+      "epoch": 1.5641266162703011,
+      "grad_norm": 0.3487800657749176,
+      "learning_rate": 4.8531143995123164e-05,
+      "loss": 2.9694,
+      "step": 21260
+    },
+    {
+      "epoch": 1.5648623296363735,
+      "grad_norm": 0.34994426369667053,
+      "learning_rate": 4.849167044365063e-05,
+      "loss": 2.9818,
+      "step": 21270
+    },
+    {
+      "epoch": 1.565598043002446,
+      "grad_norm": 0.37368515133857727,
+      "learning_rate": 4.845219783310039e-05,
+      "loss": 3.0362,
+      "step": 21280
+    },
+    {
+      "epoch": 1.5663337563685187,
+      "grad_norm": 0.3481435179710388,
+      "learning_rate": 4.841272618809611e-05,
+      "loss": 2.9585,
+      "step": 21290
+    },
+    {
+      "epoch": 1.5670694697345913,
+      "grad_norm": 0.3405168652534485,
+      "learning_rate": 4.8373255533260905e-05,
+      "loss": 2.9362,
+      "step": 21300
+    },
+    {
+      "epoch": 1.5678051831006639,
+      "grad_norm": 0.34773001074790955,
+      "learning_rate": 4.833378589321728e-05,
+      "loss": 3.0244,
+      "step": 21310
+    },
+    {
+      "epoch": 1.5685408964667364,
+      "grad_norm": 0.3556695878505707,
+      "learning_rate": 4.829431729258705e-05,
+      "loss": 3.0024,
+      "step": 21320
+    },
+    {
+      "epoch": 1.569276609832809,
+      "grad_norm": 0.36773231625556946,
+      "learning_rate": 4.825484975599145e-05,
+      "loss": 3.0397,
+      "step": 21330
+    },
+    {
+      "epoch": 1.5700123231988816,
+      "grad_norm": 0.3603079617023468,
+      "learning_rate": 4.821538330805098e-05,
+      "loss": 3.0018,
+      "step": 21340
+    },
+    {
+      "epoch": 1.5707480365649542,
+      "grad_norm": 0.3609039783477783,
+      "learning_rate": 4.817591797338554e-05,
+      "loss": 2.996,
+      "step": 21350
+    },
+    {
+      "epoch": 1.5714837499310268,
+      "grad_norm": 0.360826700925827,
+      "learning_rate": 4.8136453776614254e-05,
+      "loss": 3.0382,
+      "step": 21360
+    },
+    {
+      "epoch": 1.5722194632970994,
+      "grad_norm": 0.3730531334877014,
+      "learning_rate": 4.809699074235561e-05,
+      "loss": 2.9921,
+      "step": 21370
+    },
+    {
+      "epoch": 1.572955176663172,
+      "grad_norm": 0.362796425819397,
+      "learning_rate": 4.805752889522728e-05,
+      "loss": 2.9884,
+      "step": 21380
+    },
+    {
+      "epoch": 1.5736908900292446,
+      "grad_norm": 0.33355072140693665,
+      "learning_rate": 4.801806825984633e-05,
+      "loss": 2.9178,
+      "step": 21390
+    },
+    {
+      "epoch": 1.5744266033953171,
+      "grad_norm": 0.3495110869407654,
+      "learning_rate": 4.797860886082894e-05,
+      "loss": 2.9901,
+      "step": 21400
+    },
+    {
+      "epoch": 1.5751623167613897,
+      "grad_norm": 0.3577948212623596,
+      "learning_rate": 4.79391507227906e-05,
+      "loss": 2.9859,
+      "step": 21410
+    },
+    {
+      "epoch": 1.5758980301274623,
+      "grad_norm": 0.3669404983520508,
+      "learning_rate": 4.7899693870345976e-05,
+      "loss": 2.9855,
+      "step": 21420
+    },
+    {
+      "epoch": 1.576633743493535,
+      "grad_norm": 0.3482343256473541,
+      "learning_rate": 4.786023832810892e-05,
+      "loss": 2.9934,
+      "step": 21430
+    },
+    {
+      "epoch": 1.5773694568596075,
+      "grad_norm": 0.34475263953208923,
+      "learning_rate": 4.7820784120692516e-05,
+      "loss": 3.0186,
+      "step": 21440
+    },
+    {
+      "epoch": 1.57810517022568,
+      "grad_norm": 0.3497491478919983,
+      "learning_rate": 4.778133127270896e-05,
+      "loss": 2.9599,
+      "step": 21450
+    },
+    {
+      "epoch": 1.5788408835917527,
+      "grad_norm": 0.3471139967441559,
+      "learning_rate": 4.774187980876963e-05,
+      "loss": 3.031,
+      "step": 21460
+    },
+    {
+      "epoch": 1.5795765969578253,
+      "grad_norm": 0.33435630798339844,
+      "learning_rate": 4.770242975348505e-05,
+      "loss": 2.9824,
+      "step": 21470
+    },
+    {
+      "epoch": 1.5803123103238979,
+      "grad_norm": 0.33837267756462097,
+      "learning_rate": 4.766298113146487e-05,
+      "loss": 2.9987,
+      "step": 21480
+    },
+    {
+      "epoch": 1.5810480236899704,
+      "grad_norm": 0.3409595191478729,
+      "learning_rate": 4.762353396731779e-05,
+      "loss": 3.0215,
+      "step": 21490
+    },
+    {
+      "epoch": 1.581783737056043,
+      "grad_norm": 0.3423125147819519,
+      "learning_rate": 4.758408828565167e-05,
+      "loss": 2.954,
+      "step": 21500
+    },
+    {
+      "epoch": 1.5825194504221156,
+      "grad_norm": 0.3423803150653839,
+      "learning_rate": 4.754464411107338e-05,
+      "loss": 2.9987,
+      "step": 21510
+    },
+    {
+      "epoch": 1.5832551637881882,
+      "grad_norm": 0.34108760952949524,
+      "learning_rate": 4.750520146818893e-05,
+      "loss": 2.9239,
+      "step": 21520
+    },
+    {
+      "epoch": 1.5839908771542608,
+      "grad_norm": 0.3691301643848419,
+      "learning_rate": 4.746576038160328e-05,
+      "loss": 2.9826,
+      "step": 21530
+    },
+    {
+      "epoch": 1.5847265905203334,
+      "grad_norm": 0.36327221989631653,
+      "learning_rate": 4.7426320875920476e-05,
+      "loss": 2.9594,
+      "step": 21540
+    },
+    {
+      "epoch": 1.585462303886406,
+      "grad_norm": 0.35726088285446167,
+      "learning_rate": 4.7386882975743604e-05,
+      "loss": 2.9818,
+      "step": 21550
+    },
+    {
+      "epoch": 1.5861980172524786,
+      "grad_norm": 0.3579309582710266,
+      "learning_rate": 4.7347446705674694e-05,
+      "loss": 2.9887,
+      "step": 21560
+    },
+    {
+      "epoch": 1.5869337306185511,
+      "grad_norm": 0.3432201147079468,
+      "learning_rate": 4.730801209031478e-05,
+      "loss": 2.9476,
+      "step": 21570
+    },
+    {
+      "epoch": 1.5876694439846237,
+      "grad_norm": 0.34512314200401306,
+      "learning_rate": 4.726857915426385e-05,
+      "loss": 3.01,
+      "step": 21580
+    },
+    {
+      "epoch": 1.5884051573506963,
+      "grad_norm": 0.3597721457481384,
+      "learning_rate": 4.722914792212087e-05,
+      "loss": 3.0077,
+      "step": 21590
+    },
+    {
+      "epoch": 1.589140870716769,
+      "grad_norm": 0.3566022217273712,
+      "learning_rate": 4.718971841848373e-05,
+      "loss": 2.9798,
+      "step": 21600
+    },
+    {
+      "epoch": 1.5898765840828413,
+      "grad_norm": 0.3464104235172272,
+      "learning_rate": 4.7150290667949244e-05,
+      "loss": 3.015,
+      "step": 21610
+    },
+    {
+      "epoch": 1.5906122974489139,
+      "grad_norm": 0.3559989035129547,
+      "learning_rate": 4.7110864695113114e-05,
+      "loss": 3.0327,
+      "step": 21620
+    },
+    {
+      "epoch": 1.5913480108149864,
+      "grad_norm": 0.34521225094795227,
+      "learning_rate": 4.7071440524569944e-05,
+      "loss": 2.9844,
+      "step": 21630
+    },
+    {
+      "epoch": 1.592083724181059,
+      "grad_norm": 0.34717828035354614,
+      "learning_rate": 4.703201818091326e-05,
+      "loss": 2.9926,
+      "step": 21640
+    },
+    {
+      "epoch": 1.5928194375471316,
+      "grad_norm": 0.3550819456577301,
+      "learning_rate": 4.699259768873538e-05,
+      "loss": 2.9936,
+      "step": 21650
+    },
+    {
+      "epoch": 1.5935551509132042,
+      "grad_norm": 0.35908281803131104,
+      "learning_rate": 4.69531790726275e-05,
+      "loss": 2.9907,
+      "step": 21660
+    },
+    {
+      "epoch": 1.5942908642792768,
+      "grad_norm": 0.3484167754650116,
+      "learning_rate": 4.6913762357179625e-05,
+      "loss": 3.0188,
+      "step": 21670
+    },
+    {
+      "epoch": 1.5950265776453494,
+      "grad_norm": 0.35346874594688416,
+      "learning_rate": 4.6874347566980614e-05,
+      "loss": 2.979,
+      "step": 21680
+    },
+    {
+      "epoch": 1.595762291011422,
+      "grad_norm": 0.34139251708984375,
+      "learning_rate": 4.683493472661808e-05,
+      "loss": 2.9855,
+      "step": 21690
+    },
+    {
+      "epoch": 1.5964980043774946,
+      "grad_norm": 0.3595467507839203,
+      "learning_rate": 4.679552386067845e-05,
+      "loss": 2.959,
+      "step": 21700
+    },
+    {
+      "epoch": 1.5972337177435671,
+      "grad_norm": 0.35536545515060425,
+      "learning_rate": 4.675611499374689e-05,
+      "loss": 3.0138,
+      "step": 21710
+    },
+    {
+      "epoch": 1.5979694311096397,
+      "grad_norm": 0.3514389395713806,
+      "learning_rate": 4.671670815040737e-05,
+      "loss": 2.9857,
+      "step": 21720
+    },
+    {
+      "epoch": 1.598705144475712,
+      "grad_norm": 0.3507634401321411,
+      "learning_rate": 4.667730335524256e-05,
+      "loss": 2.9676,
+      "step": 21730
+    },
+    {
+      "epoch": 1.5994408578417847,
+      "grad_norm": 0.3487595021724701,
+      "learning_rate": 4.6637900632833835e-05,
+      "loss": 2.977,
+      "step": 21740
+    },
+    {
+      "epoch": 1.6001765712078573,
+      "grad_norm": 0.34801530838012695,
+      "learning_rate": 4.6598500007761334e-05,
+      "loss": 2.9605,
+      "step": 21750
+    },
+    {
+      "epoch": 1.6009122845739299,
+      "grad_norm": 0.3475155532360077,
+      "learning_rate": 4.6559101504603814e-05,
+      "loss": 2.9847,
+      "step": 21760
+    },
+    {
+      "epoch": 1.6016479979400025,
+      "grad_norm": 0.3437890112400055,
+      "learning_rate": 4.6519705147938785e-05,
+      "loss": 2.9668,
+      "step": 21770
+    },
+    {
+      "epoch": 1.602383711306075,
+      "grad_norm": 0.3718090057373047,
+      "learning_rate": 4.648031096234235e-05,
+      "loss": 2.9932,
+      "step": 21780
+    },
+    {
+      "epoch": 1.6031194246721476,
+      "grad_norm": 0.346178263425827,
+      "learning_rate": 4.64409189723893e-05,
+      "loss": 2.9782,
+      "step": 21790
+    },
+    {
+      "epoch": 1.6038551380382202,
+      "grad_norm": 0.34306132793426514,
+      "learning_rate": 4.640152920265306e-05,
+      "loss": 3.004,
+      "step": 21800
+    },
+    {
+      "epoch": 1.6045908514042928,
+      "grad_norm": 0.34141892194747925,
+      "learning_rate": 4.636214167770564e-05,
+      "loss": 3.0212,
+      "step": 21810
+    },
+    {
+      "epoch": 1.6053265647703654,
+      "grad_norm": 0.344409704208374,
+      "learning_rate": 4.632275642211768e-05,
+      "loss": 2.9696,
+      "step": 21820
+    },
+    {
+      "epoch": 1.606062278136438,
+      "grad_norm": 0.3617441654205322,
+      "learning_rate": 4.628337346045836e-05,
+      "loss": 2.9535,
+      "step": 21830
+    },
+    {
+      "epoch": 1.6067979915025106,
+      "grad_norm": 0.33017197251319885,
+      "learning_rate": 4.6243992817295495e-05,
+      "loss": 2.9502,
+      "step": 21840
+    },
+    {
+      "epoch": 1.6075337048685832,
+      "grad_norm": 0.34080588817596436,
+      "learning_rate": 4.620461451719539e-05,
+      "loss": 2.9639,
+      "step": 21850
+    },
+    {
+      "epoch": 1.6082694182346557,
+      "grad_norm": 0.35813194513320923,
+      "learning_rate": 4.6165238584722924e-05,
+      "loss": 2.9979,
+      "step": 21860
+    },
+    {
+      "epoch": 1.6090051316007283,
+      "grad_norm": 0.35272181034088135,
+      "learning_rate": 4.612586504444148e-05,
+      "loss": 2.98,
+      "step": 21870
+    },
+    {
+      "epoch": 1.609740844966801,
+      "grad_norm": 0.3480096161365509,
+      "learning_rate": 4.6086493920912964e-05,
+      "loss": 2.9838,
+      "step": 21880
+    },
+    {
+      "epoch": 1.6104765583328735,
+      "grad_norm": 0.3446296453475952,
+      "learning_rate": 4.604712523869779e-05,
+      "loss": 2.9942,
+      "step": 21890
+    },
+    {
+      "epoch": 1.611212271698946,
+      "grad_norm": 0.3708120286464691,
+      "learning_rate": 4.600775902235481e-05,
+      "loss": 2.9965,
+      "step": 21900
+    },
+    {
+      "epoch": 1.6119479850650187,
+      "grad_norm": 0.34935683012008667,
+      "learning_rate": 4.596839529644136e-05,
+      "loss": 2.9505,
+      "step": 21910
+    },
+    {
+      "epoch": 1.6126836984310913,
+      "grad_norm": 0.3391649127006531,
+      "learning_rate": 4.592903408551321e-05,
+      "loss": 3.0091,
+      "step": 21920
+    },
+    {
+      "epoch": 1.6134194117971639,
+      "grad_norm": 0.34947091341018677,
+      "learning_rate": 4.588967541412458e-05,
+      "loss": 2.9982,
+      "step": 21930
+    },
+    {
+      "epoch": 1.6141551251632364,
+      "grad_norm": 0.34114041924476624,
+      "learning_rate": 4.5850319306828086e-05,
+      "loss": 3.0034,
+      "step": 21940
+    },
+    {
+      "epoch": 1.614890838529309,
+      "grad_norm": 0.34920361638069153,
+      "learning_rate": 4.5810965788174745e-05,
+      "loss": 2.9828,
+      "step": 21950
+    },
+    {
+      "epoch": 1.6156265518953816,
+      "grad_norm": 0.3744545876979828,
+      "learning_rate": 4.5771614882714006e-05,
+      "loss": 2.9783,
+      "step": 21960
+    },
+    {
+      "epoch": 1.6163622652614542,
+      "grad_norm": 0.3450339436531067,
+      "learning_rate": 4.573226661499362e-05,
+      "loss": 3.0006,
+      "step": 21970
+    },
+    {
+      "epoch": 1.6170979786275268,
+      "grad_norm": 0.34535422921180725,
+      "learning_rate": 4.569292100955974e-05,
+      "loss": 2.9254,
+      "step": 21980
+    },
+    {
+      "epoch": 1.6178336919935994,
+      "grad_norm": 0.35398054122924805,
+      "learning_rate": 4.565357809095681e-05,
+      "loss": 3.0013,
+      "step": 21990
+    },
+    {
+      "epoch": 1.618569405359672,
+      "grad_norm": 0.36139732599258423,
+      "learning_rate": 4.561423788372767e-05,
+      "loss": 2.9816,
+      "step": 22000
+    },
+    {
+      "epoch": 1.6193051187257446,
+      "grad_norm": 0.34250226616859436,
+      "learning_rate": 4.5574900412413386e-05,
+      "loss": 3.0016,
+      "step": 22010
+    },
+    {
+      "epoch": 1.6200408320918172,
+      "grad_norm": 0.3612436056137085,
+      "learning_rate": 4.5535565701553394e-05,
+      "loss": 2.9571,
+      "step": 22020
+    },
+    {
+      "epoch": 1.6207765454578897,
+      "grad_norm": 0.36296573281288147,
+      "learning_rate": 4.5496233775685346e-05,
+      "loss": 2.9621,
+      "step": 22030
+    },
+    {
+      "epoch": 1.6215122588239623,
+      "grad_norm": 0.35303178429603577,
+      "learning_rate": 4.545690465934518e-05,
+      "loss": 3.0089,
+      "step": 22040
+    },
+    {
+      "epoch": 1.622247972190035,
+      "grad_norm": 0.3547990620136261,
+      "learning_rate": 4.5417578377067115e-05,
+      "loss": 3.0047,
+      "step": 22050
+    },
+    {
+      "epoch": 1.6229836855561075,
+      "grad_norm": 0.34736284613609314,
+      "learning_rate": 4.537825495338355e-05,
+      "loss": 2.9636,
+      "step": 22060
+    },
+    {
+      "epoch": 1.6237193989221799,
+      "grad_norm": 0.33305445313453674,
+      "learning_rate": 4.533893441282514e-05,
+      "loss": 2.9774,
+      "step": 22070
+    },
+    {
+      "epoch": 1.6244551122882525,
+      "grad_norm": 0.35458871722221375,
+      "learning_rate": 4.52996167799207e-05,
+      "loss": 3.0178,
+      "step": 22080
+    },
+    {
+      "epoch": 1.625190825654325,
+      "grad_norm": 0.36110827326774597,
+      "learning_rate": 4.526030207919727e-05,
+      "loss": 2.9932,
+      "step": 22090
+    },
+    {
+      "epoch": 1.6259265390203976,
+      "grad_norm": 0.3474743664264679,
+      "learning_rate": 4.5220990335180037e-05,
+      "loss": 2.9839,
+      "step": 22100
+    },
+    {
+      "epoch": 1.6266622523864702,
+      "grad_norm": 0.3576063811779022,
+      "learning_rate": 4.5181681572392356e-05,
+      "loss": 3.0135,
+      "step": 22110
+    },
+    {
+      "epoch": 1.6273979657525428,
+      "grad_norm": 0.3507232964038849,
+      "learning_rate": 4.5142375815355706e-05,
+      "loss": 2.9784,
+      "step": 22120
+    },
+    {
+      "epoch": 1.6281336791186154,
+      "grad_norm": 0.3477455675601959,
+      "learning_rate": 4.510307308858973e-05,
+      "loss": 2.9819,
+      "step": 22130
+    },
+    {
+      "epoch": 1.628869392484688,
+      "grad_norm": 0.3471401631832123,
+      "learning_rate": 4.5063773416612145e-05,
+      "loss": 2.9753,
+      "step": 22140
+    },
+    {
+      "epoch": 1.6296051058507606,
+      "grad_norm": 0.3493226170539856,
+      "learning_rate": 4.502447682393877e-05,
+      "loss": 2.9988,
+      "step": 22150
+    },
+    {
+      "epoch": 1.6303408192168332,
+      "grad_norm": 0.345857173204422,
+      "learning_rate": 4.49851833350835e-05,
+      "loss": 2.9461,
+      "step": 22160
+    },
+    {
+      "epoch": 1.6310765325829057,
+      "grad_norm": 0.34520477056503296,
+      "learning_rate": 4.4945892974558295e-05,
+      "loss": 2.9937,
+      "step": 22170
+    },
+    {
+      "epoch": 1.6318122459489781,
+      "grad_norm": 0.35884276032447815,
+      "learning_rate": 4.4906605766873186e-05,
+      "loss": 3.0123,
+      "step": 22180
+    },
+    {
+      "epoch": 1.6325479593150507,
+      "grad_norm": 0.3456844389438629,
+      "learning_rate": 4.4867321736536185e-05,
+      "loss": 2.9478,
+      "step": 22190
+    },
+    {
+      "epoch": 1.6332836726811233,
+      "grad_norm": 0.3489471971988678,
+      "learning_rate": 4.482804090805337e-05,
+      "loss": 2.9692,
+      "step": 22200
+    },
+    {
+      "epoch": 1.6340193860471959,
+      "grad_norm": 0.34100228548049927,
+      "learning_rate": 4.478876330592883e-05,
+      "loss": 2.9902,
+      "step": 22210
+    },
+    {
+      "epoch": 1.6347550994132685,
+      "grad_norm": 0.3468496799468994,
+      "learning_rate": 4.47494889546646e-05,
+      "loss": 2.9848,
+      "step": 22220
+    },
+    {
+      "epoch": 1.635490812779341,
+      "grad_norm": 0.3442668318748474,
+      "learning_rate": 4.4710217878760705e-05,
+      "loss": 2.9833,
+      "step": 22230
+    },
+    {
+      "epoch": 1.6362265261454136,
+      "grad_norm": 0.35678988695144653,
+      "learning_rate": 4.4670950102715115e-05,
+      "loss": 2.9509,
+      "step": 22240
+    },
+    {
+      "epoch": 1.6369622395114862,
+      "grad_norm": 0.3512960374355316,
+      "learning_rate": 4.463168565102378e-05,
+      "loss": 2.969,
+      "step": 22250
+    },
+    {
+      "epoch": 1.6376979528775588,
+      "grad_norm": 0.3556461036205292,
+      "learning_rate": 4.459242454818052e-05,
+      "loss": 2.9997,
+      "step": 22260
+    },
+    {
+      "epoch": 1.6384336662436314,
+      "grad_norm": 0.3333105146884918,
+      "learning_rate": 4.455316681867712e-05,
+      "loss": 2.9416,
+      "step": 22270
+    },
+    {
+      "epoch": 1.639169379609704,
+      "grad_norm": 0.35357415676116943,
+      "learning_rate": 4.451391248700319e-05,
+      "loss": 2.9546,
+      "step": 22280
+    },
+    {
+      "epoch": 1.6399050929757766,
+      "grad_norm": 0.3415786325931549,
+      "learning_rate": 4.4474661577646346e-05,
+      "loss": 2.9539,
+      "step": 22290
+    },
+    {
+      "epoch": 1.6406408063418492,
+      "grad_norm": 0.34051311016082764,
+      "learning_rate": 4.443541411509193e-05,
+      "loss": 2.9845,
+      "step": 22300
+    },
+    {
+      "epoch": 1.6413765197079218,
+      "grad_norm": 0.3502918481826782,
+      "learning_rate": 4.4396170123823225e-05,
+      "loss": 2.9436,
+      "step": 22310
+    },
+    {
+      "epoch": 1.6421122330739943,
+      "grad_norm": 0.35389769077301025,
+      "learning_rate": 4.435692962832129e-05,
+      "loss": 3.0212,
+      "step": 22320
+    },
+    {
+      "epoch": 1.642847946440067,
+      "grad_norm": 0.3504621684551239,
+      "learning_rate": 4.431769265306505e-05,
+      "loss": 3.0034,
+      "step": 22330
+    },
+    {
+      "epoch": 1.6435836598061395,
+      "grad_norm": 0.34921252727508545,
+      "learning_rate": 4.427845922253119e-05,
+      "loss": 3.0113,
+      "step": 22340
+    },
+    {
+      "epoch": 1.644319373172212,
+      "grad_norm": 0.3413257598876953,
+      "learning_rate": 4.423922936119424e-05,
+      "loss": 2.9603,
+      "step": 22350
+    },
+    {
+      "epoch": 1.6450550865382847,
+      "grad_norm": 0.34307077527046204,
+      "learning_rate": 4.4200003093526456e-05,
+      "loss": 2.9822,
+      "step": 22360
+    },
+    {
+      "epoch": 1.6457907999043573,
+      "grad_norm": 0.3567951023578644,
+      "learning_rate": 4.416078044399783e-05,
+      "loss": 2.9919,
+      "step": 22370
+    },
+    {
+      "epoch": 1.6465265132704299,
+      "grad_norm": 0.3547225296497345,
+      "learning_rate": 4.412156143707621e-05,
+      "loss": 3.0273,
+      "step": 22380
+    },
+    {
+      "epoch": 1.6472622266365025,
+      "grad_norm": 0.34059447050094604,
+      "learning_rate": 4.4082346097227045e-05,
+      "loss": 2.9689,
+      "step": 22390
+    },
+    {
+      "epoch": 1.647997940002575,
+      "grad_norm": 0.35707423090934753,
+      "learning_rate": 4.4043134448913576e-05,
+      "loss": 2.9757,
+      "step": 22400
+    },
+    {
+      "epoch": 1.6487336533686476,
+      "grad_norm": 0.35930317640304565,
+      "learning_rate": 4.4003926516596685e-05,
+      "loss": 2.9802,
+      "step": 22410
+    },
+    {
+      "epoch": 1.6494693667347202,
+      "grad_norm": 0.3571709990501404,
+      "learning_rate": 4.396472232473498e-05,
+      "loss": 3.0288,
+      "step": 22420
+    },
+    {
+      "epoch": 1.6502050801007928,
+      "grad_norm": 0.3489092290401459,
+      "learning_rate": 4.392552189778472e-05,
+      "loss": 2.9811,
+      "step": 22430
+    },
+    {
+      "epoch": 1.6509407934668654,
+      "grad_norm": 0.35287657380104065,
+      "learning_rate": 4.388632526019981e-05,
+      "loss": 2.9737,
+      "step": 22440
+    },
+    {
+      "epoch": 1.651676506832938,
+      "grad_norm": 0.3478105068206787,
+      "learning_rate": 4.384713243643178e-05,
+      "loss": 2.9967,
+      "step": 22450
+    },
+    {
+      "epoch": 1.6524122201990106,
+      "grad_norm": 0.3545510768890381,
+      "learning_rate": 4.3807943450929816e-05,
+      "loss": 2.9755,
+      "step": 22460
+    },
+    {
+      "epoch": 1.6531479335650832,
+      "grad_norm": 0.36810487508773804,
+      "learning_rate": 4.3768758328140705e-05,
+      "loss": 3.0072,
+      "step": 22470
+    },
+    {
+      "epoch": 1.6538836469311557,
+      "grad_norm": 0.3456381559371948,
+      "learning_rate": 4.3729577092508784e-05,
+      "loss": 2.9754,
+      "step": 22480
+    },
+    {
+      "epoch": 1.6546193602972283,
+      "grad_norm": 0.3420705199241638,
+      "learning_rate": 4.369039976847601e-05,
+      "loss": 2.9701,
+      "step": 22490
+    },
+    {
+      "epoch": 1.655355073663301,
+      "grad_norm": 0.35932889580726624,
+      "learning_rate": 4.365122638048186e-05,
+      "loss": 2.9542,
+      "step": 22500
+    },
+    {
+      "epoch": 1.6560907870293735,
+      "grad_norm": 0.35236239433288574,
+      "learning_rate": 4.36120569529634e-05,
+      "loss": 3.0244,
+      "step": 22510
+    },
+    {
+      "epoch": 1.656826500395446,
+      "grad_norm": 0.3526109457015991,
+      "learning_rate": 4.357289151035517e-05,
+      "loss": 3.0274,
+      "step": 22520
+    },
+    {
+      "epoch": 1.6575622137615185,
+      "grad_norm": 0.34922292828559875,
+      "learning_rate": 4.353373007708927e-05,
+      "loss": 2.9839,
+      "step": 22530
+    },
+    {
+      "epoch": 1.658297927127591,
+      "grad_norm": 0.3480394184589386,
+      "learning_rate": 4.349457267759531e-05,
+      "loss": 3.0185,
+      "step": 22540
+    },
+    {
+      "epoch": 1.6590336404936636,
+      "grad_norm": 0.3449772000312805,
+      "learning_rate": 4.3455419336300326e-05,
+      "loss": 3.0132,
+      "step": 22550
+    },
+    {
+      "epoch": 1.6597693538597362,
+      "grad_norm": 0.35980257391929626,
+      "learning_rate": 4.341627007762888e-05,
+      "loss": 3.0361,
+      "step": 22560
+    },
+    {
+      "epoch": 1.6605050672258088,
+      "grad_norm": 0.3546968698501587,
+      "learning_rate": 4.337712492600294e-05,
+      "loss": 3.01,
+      "step": 22570
+    },
+    {
+      "epoch": 1.6612407805918814,
+      "grad_norm": 0.3500770926475525,
+      "learning_rate": 4.333798390584195e-05,
+      "loss": 2.9582,
+      "step": 22580
+    },
+    {
+      "epoch": 1.661976493957954,
+      "grad_norm": 0.34252896904945374,
+      "learning_rate": 4.329884704156274e-05,
+      "loss": 2.9894,
+      "step": 22590
+    },
+    {
+      "epoch": 1.6627122073240266,
+      "grad_norm": 0.3549545705318451,
+      "learning_rate": 4.325971435757961e-05,
+      "loss": 2.9988,
+      "step": 22600
+    },
+    {
+      "epoch": 1.6634479206900992,
+      "grad_norm": 0.34121638536453247,
+      "learning_rate": 4.322058587830414e-05,
+      "loss": 2.9922,
+      "step": 22610
+    },
+    {
+      "epoch": 1.6641836340561718,
+      "grad_norm": 0.36031731963157654,
+      "learning_rate": 4.3181461628145424e-05,
+      "loss": 2.9425,
+      "step": 22620
+    },
+    {
+      "epoch": 1.6649193474222443,
+      "grad_norm": 0.35915088653564453,
+      "learning_rate": 4.314234163150984e-05,
+      "loss": 2.9541,
+      "step": 22630
+    },
+    {
+      "epoch": 1.6656550607883167,
+      "grad_norm": 0.34827104210853577,
+      "learning_rate": 4.31032259128011e-05,
+      "loss": 3.0126,
+      "step": 22640
+    },
+    {
+      "epoch": 1.6663907741543893,
+      "grad_norm": 0.36567044258117676,
+      "learning_rate": 4.3064114496420286e-05,
+      "loss": 2.9811,
+      "step": 22650
+    },
+    {
+      "epoch": 1.6671264875204619,
+      "grad_norm": 0.34746843576431274,
+      "learning_rate": 4.3025007406765775e-05,
+      "loss": 3.0543,
+      "step": 22660
+    },
+    {
+      "epoch": 1.6678622008865345,
+      "grad_norm": 0.3598044216632843,
+      "learning_rate": 4.2985904668233266e-05,
+      "loss": 2.9837,
+      "step": 22670
+    },
+    {
+      "epoch": 1.668597914252607,
+      "grad_norm": 0.3502768576145172,
+      "learning_rate": 4.294680630521571e-05,
+      "loss": 2.9526,
+      "step": 22680
+    },
+    {
+      "epoch": 1.6693336276186796,
+      "grad_norm": 0.3544299900531769,
+      "learning_rate": 4.290771234210336e-05,
+      "loss": 3.0055,
+      "step": 22690
+    },
+    {
+      "epoch": 1.6700693409847522,
+      "grad_norm": 0.3577900826931,
+      "learning_rate": 4.2868622803283694e-05,
+      "loss": 3.0031,
+      "step": 22700
+    },
+    {
+      "epoch": 1.6708050543508248,
+      "grad_norm": 0.34965163469314575,
+      "learning_rate": 4.282953771314147e-05,
+      "loss": 2.9509,
+      "step": 22710
+    },
+    {
+      "epoch": 1.6715407677168974,
+      "grad_norm": 0.3502456247806549,
+      "learning_rate": 4.2790457096058644e-05,
+      "loss": 2.9779,
+      "step": 22720
+    },
+    {
+      "epoch": 1.67227648108297,
+      "grad_norm": 0.3535136282444,
+      "learning_rate": 4.2751380976414374e-05,
+      "loss": 2.9417,
+      "step": 22730
+    },
+    {
+      "epoch": 1.6730121944490426,
+      "grad_norm": 0.33796995878219604,
+      "learning_rate": 4.271230937858505e-05,
+      "loss": 2.9642,
+      "step": 22740
+    },
+    {
+      "epoch": 1.6737479078151152,
+      "grad_norm": 0.35047245025634766,
+      "learning_rate": 4.267324232694418e-05,
+      "loss": 2.9671,
+      "step": 22750
+    },
+    {
+      "epoch": 1.6744836211811878,
+      "grad_norm": 0.3488644063472748,
+      "learning_rate": 4.263417984586251e-05,
+      "loss": 2.986,
+      "step": 22760
+    },
+    {
+      "epoch": 1.6752193345472604,
+      "grad_norm": 0.3577163517475128,
+      "learning_rate": 4.2595121959707864e-05,
+      "loss": 2.9893,
+      "step": 22770
+    },
+    {
+      "epoch": 1.675955047913333,
+      "grad_norm": 0.34487199783325195,
+      "learning_rate": 4.255606869284523e-05,
+      "loss": 3.0215,
+      "step": 22780
+    },
+    {
+      "epoch": 1.6766907612794055,
+      "grad_norm": 0.342141330242157,
+      "learning_rate": 4.251702006963676e-05,
+      "loss": 2.9638,
+      "step": 22790
+    },
+    {
+      "epoch": 1.6774264746454781,
+      "grad_norm": 0.3505381643772125,
+      "learning_rate": 4.247797611444163e-05,
+      "loss": 3.012,
+      "step": 22800
+    },
+    {
+      "epoch": 1.6781621880115507,
+      "grad_norm": 0.3432064652442932,
+      "learning_rate": 4.243893685161616e-05,
+      "loss": 2.995,
+      "step": 22810
+    },
+    {
+      "epoch": 1.6788979013776233,
+      "grad_norm": 0.34672215580940247,
+      "learning_rate": 4.23999023055137e-05,
+      "loss": 2.9605,
+      "step": 22820
+    },
+    {
+      "epoch": 1.6796336147436959,
+      "grad_norm": 0.3573468327522278,
+      "learning_rate": 4.23608725004847e-05,
+      "loss": 3.0133,
+      "step": 22830
+    },
+    {
+      "epoch": 1.6803693281097685,
+      "grad_norm": 0.3517932891845703,
+      "learning_rate": 4.232184746087661e-05,
+      "loss": 3.0316,
+      "step": 22840
+    },
+    {
+      "epoch": 1.681105041475841,
+      "grad_norm": 0.3501693904399872,
+      "learning_rate": 4.2282827211033964e-05,
+      "loss": 2.9997,
+      "step": 22850
+    },
+    {
+      "epoch": 1.6818407548419136,
+      "grad_norm": 0.3443460166454315,
+      "learning_rate": 4.224381177529823e-05,
+      "loss": 2.972,
+      "step": 22860
+    },
+    {
+      "epoch": 1.6825764682079862,
+      "grad_norm": 0.3522225022315979,
+      "learning_rate": 4.220480117800796e-05,
+      "loss": 2.9537,
+      "step": 22870
+    },
+    {
+      "epoch": 1.6833121815740588,
+      "grad_norm": 0.35135418176651,
+      "learning_rate": 4.2165795443498626e-05,
+      "loss": 2.9999,
+      "step": 22880
+    },
+    {
+      "epoch": 1.6840478949401314,
+      "grad_norm": 0.3464687764644623,
+      "learning_rate": 4.2126794596102684e-05,
+      "loss": 2.9953,
+      "step": 22890
+    },
+    {
+      "epoch": 1.684783608306204,
+      "grad_norm": 0.3541601896286011,
+      "learning_rate": 4.2087798660149556e-05,
+      "loss": 2.9919,
+      "step": 22900
+    },
+    {
+      "epoch": 1.6855193216722766,
+      "grad_norm": 0.36320218443870544,
+      "learning_rate": 4.204880765996557e-05,
+      "loss": 2.9876,
+      "step": 22910
+    },
+    {
+      "epoch": 1.6862550350383492,
+      "grad_norm": 0.3492906391620636,
+      "learning_rate": 4.2009821619874e-05,
+      "loss": 2.9532,
+      "step": 22920
+    },
+    {
+      "epoch": 1.6869907484044218,
+      "grad_norm": 0.36078473925590515,
+      "learning_rate": 4.197084056419501e-05,
+      "loss": 3.0063,
+      "step": 22930
+    },
+    {
+      "epoch": 1.6877264617704943,
+      "grad_norm": 0.34769847989082336,
+      "learning_rate": 4.193186451724565e-05,
+      "loss": 2.9932,
+      "step": 22940
+    },
+    {
+      "epoch": 1.688462175136567,
+      "grad_norm": 0.34324219822883606,
+      "learning_rate": 4.1892893503339904e-05,
+      "loss": 2.977,
+      "step": 22950
+    },
+    {
+      "epoch": 1.6891978885026395,
+      "grad_norm": 0.347002774477005,
+      "learning_rate": 4.1853927546788515e-05,
+      "loss": 2.9547,
+      "step": 22960
+    },
+    {
+      "epoch": 1.689933601868712,
+      "grad_norm": 0.341463178396225,
+      "learning_rate": 4.181496667189918e-05,
+      "loss": 2.9584,
+      "step": 22970
+    },
+    {
+      "epoch": 1.6906693152347845,
+      "grad_norm": 0.35894274711608887,
+      "learning_rate": 4.1776010902976324e-05,
+      "loss": 2.9821,
+      "step": 22980
+    },
+    {
+      "epoch": 1.691405028600857,
+      "grad_norm": 0.3457069396972656,
+      "learning_rate": 4.173706026432127e-05,
+      "loss": 2.9797,
+      "step": 22990
+    },
+    {
+      "epoch": 1.6921407419669297,
+      "grad_norm": 0.34433114528656006,
+      "learning_rate": 4.1698114780232085e-05,
+      "loss": 2.9445,
+      "step": 23000
+    },
+    {
+      "epoch": 1.6928764553330022,
+      "grad_norm": 0.3480866849422455,
+      "learning_rate": 4.165917447500365e-05,
+      "loss": 2.9757,
+      "step": 23010
+    },
+    {
+      "epoch": 1.6936121686990748,
+      "grad_norm": 0.3439785838127136,
+      "learning_rate": 4.162023937292759e-05,
+      "loss": 2.9255,
+      "step": 23020
+    },
+    {
+      "epoch": 1.6943478820651474,
+      "grad_norm": 0.36015868186950684,
+      "learning_rate": 4.158130949829231e-05,
+      "loss": 2.9636,
+      "step": 23030
+    },
+    {
+      "epoch": 1.69508359543122,
+      "grad_norm": 0.3638933300971985,
+      "learning_rate": 4.1542384875382964e-05,
+      "loss": 2.9306,
+      "step": 23040
+    },
+    {
+      "epoch": 1.6958193087972926,
+      "grad_norm": 0.35429543256759644,
+      "learning_rate": 4.1503465528481384e-05,
+      "loss": 2.9436,
+      "step": 23050
+    },
+    {
+      "epoch": 1.6965550221633652,
+      "grad_norm": 0.34403643012046814,
+      "learning_rate": 4.1464551481866156e-05,
+      "loss": 2.9548,
+      "step": 23060
+    },
+    {
+      "epoch": 1.6972907355294378,
+      "grad_norm": 0.3466806709766388,
+      "learning_rate": 4.142564275981253e-05,
+      "loss": 2.9584,
+      "step": 23070
+    },
+    {
+      "epoch": 1.6980264488955104,
+      "grad_norm": 0.3539373576641083,
+      "learning_rate": 4.138673938659245e-05,
+      "loss": 2.986,
+      "step": 23080
+    },
+    {
+      "epoch": 1.698762162261583,
+      "grad_norm": 0.3486233055591583,
+      "learning_rate": 4.1347841386474506e-05,
+      "loss": 2.953,
+      "step": 23090
+    },
+    {
+      "epoch": 1.6994978756276553,
+      "grad_norm": 0.35437101125717163,
+      "learning_rate": 4.1308948783723975e-05,
+      "loss": 2.9426,
+      "step": 23100
+    },
+    {
+      "epoch": 1.700233588993728,
+      "grad_norm": 0.3440972864627838,
+      "learning_rate": 4.1270061602602694e-05,
+      "loss": 2.9881,
+      "step": 23110
+    },
+    {
+      "epoch": 1.7009693023598005,
+      "grad_norm": 0.3660430908203125,
+      "learning_rate": 4.1231179867369214e-05,
+      "loss": 2.9667,
+      "step": 23120
+    },
+    {
+      "epoch": 1.701705015725873,
+      "grad_norm": 0.34180861711502075,
+      "learning_rate": 4.119230360227863e-05,
+      "loss": 2.9621,
+      "step": 23130
+    },
+    {
+      "epoch": 1.7024407290919457,
+      "grad_norm": 0.3553137481212616,
+      "learning_rate": 4.1153432831582604e-05,
+      "loss": 2.9704,
+      "step": 23140
+    },
+    {
+      "epoch": 1.7031764424580182,
+      "grad_norm": 0.3335139751434326,
+      "learning_rate": 4.111456757952944e-05,
+      "loss": 2.9404,
+      "step": 23150
+    },
+    {
+      "epoch": 1.7039121558240908,
+      "grad_norm": 0.3433009684085846,
+      "learning_rate": 4.107570787036393e-05,
+      "loss": 2.9604,
+      "step": 23160
+    },
+    {
+      "epoch": 1.7046478691901634,
+      "grad_norm": 0.3608010411262512,
+      "learning_rate": 4.1036853728327465e-05,
+      "loss": 2.9976,
+      "step": 23170
+    },
+    {
+      "epoch": 1.705383582556236,
+      "grad_norm": 0.3552555739879608,
+      "learning_rate": 4.099800517765789e-05,
+      "loss": 2.9976,
+      "step": 23180
+    },
+    {
+      "epoch": 1.7061192959223086,
+      "grad_norm": 0.35034385323524475,
+      "learning_rate": 4.095916224258964e-05,
+      "loss": 2.9855,
+      "step": 23190
+    },
+    {
+      "epoch": 1.7068550092883812,
+      "grad_norm": 0.3490162491798401,
+      "learning_rate": 4.0920324947353626e-05,
+      "loss": 2.9979,
+      "step": 23200
+    },
+    {
+      "epoch": 1.7075907226544538,
+      "grad_norm": 0.3468624949455261,
+      "learning_rate": 4.088149331617723e-05,
+      "loss": 2.9857,
+      "step": 23210
+    },
+    {
+      "epoch": 1.7083264360205264,
+      "grad_norm": 0.3372463881969452,
+      "learning_rate": 4.084266737328427e-05,
+      "loss": 2.9929,
+      "step": 23220
+    },
+    {
+      "epoch": 1.709062149386599,
+      "grad_norm": 0.3479689657688141,
+      "learning_rate": 4.080384714289508e-05,
+      "loss": 2.9565,
+      "step": 23230
+    },
+    {
+      "epoch": 1.7097978627526715,
+      "grad_norm": 0.34987664222717285,
+      "learning_rate": 4.076503264922637e-05,
+      "loss": 3.0344,
+      "step": 23240
+    },
+    {
+      "epoch": 1.7105335761187441,
+      "grad_norm": 0.3436524569988251,
+      "learning_rate": 4.0726223916491315e-05,
+      "loss": 2.9669,
+      "step": 23250
+    },
+    {
+      "epoch": 1.7112692894848167,
+      "grad_norm": 0.3446681499481201,
+      "learning_rate": 4.068742096889947e-05,
+      "loss": 2.9777,
+      "step": 23260
+    },
+    {
+      "epoch": 1.7120050028508893,
+      "grad_norm": 0.34612777829170227,
+      "learning_rate": 4.064862383065677e-05,
+      "loss": 2.9645,
+      "step": 23270
+    },
+    {
+      "epoch": 1.712740716216962,
+      "grad_norm": 0.3578917682170868,
+      "learning_rate": 4.0609832525965596e-05,
+      "loss": 2.9268,
+      "step": 23280
+    },
+    {
+      "epoch": 1.7134764295830345,
+      "grad_norm": 0.33856597542762756,
+      "learning_rate": 4.0571047079024607e-05,
+      "loss": 2.9709,
+      "step": 23290
+    },
+    {
+      "epoch": 1.714212142949107,
+      "grad_norm": 0.3613859713077545,
+      "learning_rate": 4.053226751402885e-05,
+      "loss": 2.951,
+      "step": 23300
+    },
+    {
+      "epoch": 1.7149478563151797,
+      "grad_norm": 0.3537937104701996,
+      "learning_rate": 4.049349385516968e-05,
+      "loss": 2.9518,
+      "step": 23310
+    },
+    {
+      "epoch": 1.7156835696812522,
+      "grad_norm": 0.35170820355415344,
+      "learning_rate": 4.045472612663481e-05,
+      "loss": 2.9916,
+      "step": 23320
+    },
+    {
+      "epoch": 1.7164192830473248,
+      "grad_norm": 0.3534245193004608,
+      "learning_rate": 4.041596435260819e-05,
+      "loss": 3.0214,
+      "step": 23330
+    },
+    {
+      "epoch": 1.7171549964133974,
+      "grad_norm": 0.34871846437454224,
+      "learning_rate": 4.037720855727013e-05,
+      "loss": 3.0077,
+      "step": 23340
+    },
+    {
+      "epoch": 1.71789070977947,
+      "grad_norm": 0.34917789697647095,
+      "learning_rate": 4.033845876479712e-05,
+      "loss": 2.9378,
+      "step": 23350
+    },
+    {
+      "epoch": 1.7186264231455426,
+      "grad_norm": 0.3826858103275299,
+      "learning_rate": 4.029971499936199e-05,
+      "loss": 2.9874,
+      "step": 23360
+    },
+    {
+      "epoch": 1.7193621365116152,
+      "grad_norm": 0.3708551526069641,
+      "learning_rate": 4.02609772851338e-05,
+      "loss": 3.0166,
+      "step": 23370
+    },
+    {
+      "epoch": 1.7200978498776878,
+      "grad_norm": 0.3503763973712921,
+      "learning_rate": 4.022224564627775e-05,
+      "loss": 3.0067,
+      "step": 23380
+    },
+    {
+      "epoch": 1.7208335632437604,
+      "grad_norm": 0.3445330858230591,
+      "learning_rate": 4.0183520106955366e-05,
+      "loss": 2.9624,
+      "step": 23390
+    },
+    {
+      "epoch": 1.721569276609833,
+      "grad_norm": 0.34792691469192505,
+      "learning_rate": 4.014480069132428e-05,
+      "loss": 2.9515,
+      "step": 23400
+    },
+    {
+      "epoch": 1.7223049899759055,
+      "grad_norm": 0.3410683274269104,
+      "learning_rate": 4.010608742353837e-05,
+      "loss": 2.977,
+      "step": 23410
+    },
+    {
+      "epoch": 1.7230407033419781,
+      "grad_norm": 0.35673004388809204,
+      "learning_rate": 4.006738032774761e-05,
+      "loss": 2.9438,
+      "step": 23420
+    },
+    {
+      "epoch": 1.7237764167080507,
+      "grad_norm": 0.3416217565536499,
+      "learning_rate": 4.002867942809819e-05,
+      "loss": 2.9769,
+      "step": 23430
+    },
+    {
+      "epoch": 1.724512130074123,
+      "grad_norm": 0.3230731785297394,
+      "learning_rate": 3.9989984748732386e-05,
+      "loss": 2.9659,
+      "step": 23440
+    },
+    {
+      "epoch": 1.7252478434401957,
+      "grad_norm": 0.3534250259399414,
+      "learning_rate": 3.995129631378862e-05,
+      "loss": 3.0056,
+      "step": 23450
+    },
+    {
+      "epoch": 1.7259835568062682,
+      "grad_norm": 0.33658337593078613,
+      "learning_rate": 3.991261414740144e-05,
+      "loss": 2.9946,
+      "step": 23460
+    },
+    {
+      "epoch": 1.7267192701723408,
+      "grad_norm": 0.3498584032058716,
+      "learning_rate": 3.987393827370142e-05,
+      "loss": 2.9692,
+      "step": 23470
+    },
+    {
+      "epoch": 1.7274549835384134,
+      "grad_norm": 0.34658220410346985,
+      "learning_rate": 3.983526871681525e-05,
+      "loss": 2.9457,
+      "step": 23480
+    },
+    {
+      "epoch": 1.728190696904486,
+      "grad_norm": 0.34107422828674316,
+      "learning_rate": 3.979660550086569e-05,
+      "loss": 2.925,
+      "step": 23490
+    },
+    {
+      "epoch": 1.7289264102705586,
+      "grad_norm": 0.3476036787033081,
+      "learning_rate": 3.9757948649971524e-05,
+      "loss": 2.9989,
+      "step": 23500
+    },
+    {
+      "epoch": 1.7296621236366312,
+      "grad_norm": 0.34889787435531616,
+      "learning_rate": 3.971929818824755e-05,
+      "loss": 3.0017,
+      "step": 23510
+    },
+    {
+      "epoch": 1.7303978370027038,
+      "grad_norm": 0.337389200925827,
+      "learning_rate": 3.96806541398046e-05,
+      "loss": 2.9607,
+      "step": 23520
+    },
+    {
+      "epoch": 1.7311335503687764,
+      "grad_norm": 0.33042651414871216,
+      "learning_rate": 3.964201652874955e-05,
+      "loss": 2.9435,
+      "step": 23530
+    },
+    {
+      "epoch": 1.731869263734849,
+      "grad_norm": 0.34432706236839294,
+      "learning_rate": 3.960338537918517e-05,
+      "loss": 2.9276,
+      "step": 23540
+    },
+    {
+      "epoch": 1.7326049771009213,
+      "grad_norm": 0.3430382013320923,
+      "learning_rate": 3.956476071521026e-05,
+      "loss": 2.9429,
+      "step": 23550
+    },
+    {
+      "epoch": 1.733340690466994,
+      "grad_norm": 0.3549818694591522,
+      "learning_rate": 3.952614256091956e-05,
+      "loss": 2.9832,
+      "step": 23560
+    },
+    {
+      "epoch": 1.7340764038330665,
+      "grad_norm": 0.36157238483428955,
+      "learning_rate": 3.9487530940403754e-05,
+      "loss": 3.0233,
+      "step": 23570
+    },
+    {
+      "epoch": 1.734812117199139,
+      "grad_norm": 0.3578517735004425,
+      "learning_rate": 3.944892587774942e-05,
+      "loss": 2.967,
+      "step": 23580
+    },
+    {
+      "epoch": 1.7355478305652117,
+      "grad_norm": 0.3492560386657715,
+      "learning_rate": 3.94103273970391e-05,
+      "loss": 2.9845,
+      "step": 23590
+    },
+    {
+      "epoch": 1.7362835439312843,
+      "grad_norm": 0.3477337956428528,
+      "learning_rate": 3.937173552235117e-05,
+      "loss": 3.0082,
+      "step": 23600
+    },
+    {
+      "epoch": 1.7370192572973568,
+      "grad_norm": 0.35257795453071594,
+      "learning_rate": 3.9333150277759926e-05,
+      "loss": 3.0123,
+      "step": 23610
+    },
+    {
+      "epoch": 1.7377549706634294,
+      "grad_norm": 0.35249119997024536,
+      "learning_rate": 3.929457168733554e-05,
+      "loss": 2.9737,
+      "step": 23620
+    },
+    {
+      "epoch": 1.738490684029502,
+      "grad_norm": 0.34828656911849976,
+      "learning_rate": 3.9255999775143976e-05,
+      "loss": 2.9526,
+      "step": 23630
+    },
+    {
+      "epoch": 1.7392263973955746,
+      "grad_norm": 0.3541480004787445,
+      "learning_rate": 3.921743456524708e-05,
+      "loss": 2.9709,
+      "step": 23640
+    },
+    {
+      "epoch": 1.7399621107616472,
+      "grad_norm": 0.3592761158943176,
+      "learning_rate": 3.91788760817025e-05,
+      "loss": 2.9991,
+      "step": 23650
+    },
+    {
+      "epoch": 1.7406978241277198,
+      "grad_norm": 0.3556549549102783,
+      "learning_rate": 3.9140324348563695e-05,
+      "loss": 2.9448,
+      "step": 23660
+    },
+    {
+      "epoch": 1.7414335374937924,
+      "grad_norm": 0.35440292954444885,
+      "learning_rate": 3.91017793898799e-05,
+      "loss": 3.0023,
+      "step": 23670
+    },
+    {
+      "epoch": 1.742169250859865,
+      "grad_norm": 0.35832077264785767,
+      "learning_rate": 3.906324122969612e-05,
+      "loss": 2.9662,
+      "step": 23680
+    },
+    {
+      "epoch": 1.7429049642259375,
+      "grad_norm": 0.3554212152957916,
+      "learning_rate": 3.9024709892053166e-05,
+      "loss": 2.9814,
+      "step": 23690
+    },
+    {
+      "epoch": 1.7436406775920101,
+      "grad_norm": 0.35513070225715637,
+      "learning_rate": 3.8986185400987534e-05,
+      "loss": 2.9738,
+      "step": 23700
+    },
+    {
+      "epoch": 1.7443763909580827,
+      "grad_norm": 0.3516767919063568,
+      "learning_rate": 3.8947667780531485e-05,
+      "loss": 2.9545,
+      "step": 23710
+    },
+    {
+      "epoch": 1.7451121043241553,
+      "grad_norm": 0.3510574400424957,
+      "learning_rate": 3.890915705471297e-05,
+      "loss": 2.9435,
+      "step": 23720
+    },
+    {
+      "epoch": 1.745847817690228,
+      "grad_norm": 0.3394900858402252,
+      "learning_rate": 3.8870653247555664e-05,
+      "loss": 2.9937,
+      "step": 23730
+    },
+    {
+      "epoch": 1.7465835310563005,
+      "grad_norm": 0.35016101598739624,
+      "learning_rate": 3.88321563830789e-05,
+      "loss": 3.0078,
+      "step": 23740
+    },
+    {
+      "epoch": 1.747319244422373,
+      "grad_norm": 0.3542245924472809,
+      "learning_rate": 3.879366648529771e-05,
+      "loss": 2.9808,
+      "step": 23750
+    },
+    {
+      "epoch": 1.7480549577884457,
+      "grad_norm": 0.3462967276573181,
+      "learning_rate": 3.8755183578222737e-05,
+      "loss": 2.9634,
+      "step": 23760
+    },
+    {
+      "epoch": 1.7487906711545183,
+      "grad_norm": 0.34834882616996765,
+      "learning_rate": 3.87167076858603e-05,
+      "loss": 2.9997,
+      "step": 23770
+    },
+    {
+      "epoch": 1.7495263845205908,
+      "grad_norm": 0.35272443294525146,
+      "learning_rate": 3.867823883221235e-05,
+      "loss": 2.9854,
+      "step": 23780
+    },
+    {
+      "epoch": 1.7502620978866634,
+      "grad_norm": 0.3505149781703949,
+      "learning_rate": 3.8639777041276416e-05,
+      "loss": 3.0201,
+      "step": 23790
+    },
+    {
+      "epoch": 1.750997811252736,
+      "grad_norm": 0.3406427800655365,
+      "learning_rate": 3.860132233704563e-05,
+      "loss": 2.9711,
+      "step": 23800
+    },
+    {
+      "epoch": 1.7517335246188086,
+      "grad_norm": 0.3990043103694916,
+      "learning_rate": 3.8562874743508715e-05,
+      "loss": 2.9963,
+      "step": 23810
+    },
+    {
+      "epoch": 1.7524692379848812,
+      "grad_norm": 0.340960294008255,
+      "learning_rate": 3.852443428464995e-05,
+      "loss": 2.9706,
+      "step": 23820
+    },
+    {
+      "epoch": 1.7532049513509538,
+      "grad_norm": 0.34657010436058044,
+      "learning_rate": 3.8486000984449146e-05,
+      "loss": 2.9975,
+      "step": 23830
+    },
+    {
+      "epoch": 1.7539406647170264,
+      "grad_norm": 0.36372703313827515,
+      "learning_rate": 3.8447574866881684e-05,
+      "loss": 2.9633,
+      "step": 23840
+    },
+    {
+      "epoch": 1.754676378083099,
+      "grad_norm": 0.34926316142082214,
+      "learning_rate": 3.840915595591843e-05,
+      "loss": 2.9595,
+      "step": 23850
+    },
+    {
+      "epoch": 1.7554120914491715,
+      "grad_norm": 0.34664928913116455,
+      "learning_rate": 3.8370744275525786e-05,
+      "loss": 2.9757,
+      "step": 23860
+    },
+    {
+      "epoch": 1.7561478048152441,
+      "grad_norm": 0.34869304299354553,
+      "learning_rate": 3.833233984966564e-05,
+      "loss": 2.9756,
+      "step": 23870
+    },
+    {
+      "epoch": 1.7568835181813167,
+      "grad_norm": 0.35055258870124817,
+      "learning_rate": 3.829394270229531e-05,
+      "loss": 2.9847,
+      "step": 23880
+    },
+    {
+      "epoch": 1.757619231547389,
+      "grad_norm": 0.3498460054397583,
+      "learning_rate": 3.825555285736763e-05,
+      "loss": 2.9178,
+      "step": 23890
+    },
+    {
+      "epoch": 1.7583549449134617,
+      "grad_norm": 0.3586640954017639,
+      "learning_rate": 3.8217170338830847e-05,
+      "loss": 2.9602,
+      "step": 23900
+    },
+    {
+      "epoch": 1.7590906582795343,
+      "grad_norm": 0.3469058871269226,
+      "learning_rate": 3.817879517062864e-05,
+      "loss": 2.9443,
+      "step": 23910
+    },
+    {
+      "epoch": 1.7598263716456068,
+      "grad_norm": 0.35669317841529846,
+      "learning_rate": 3.814042737670011e-05,
+      "loss": 2.9867,
+      "step": 23920
+    },
+    {
+      "epoch": 1.7605620850116794,
+      "grad_norm": 0.3473479151725769,
+      "learning_rate": 3.8102066980979745e-05,
+      "loss": 2.9713,
+      "step": 23930
+    },
+    {
+      "epoch": 1.761297798377752,
+      "grad_norm": 0.3396706283092499,
+      "learning_rate": 3.806371400739747e-05,
+      "loss": 2.9134,
+      "step": 23940
+    },
+    {
+      "epoch": 1.7620335117438246,
+      "grad_norm": 0.34082844853401184,
+      "learning_rate": 3.8025368479878496e-05,
+      "loss": 2.9357,
+      "step": 23950
+    },
+    {
+      "epoch": 1.7627692251098972,
+      "grad_norm": 0.35015684366226196,
+      "learning_rate": 3.798703042234345e-05,
+      "loss": 2.9433,
+      "step": 23960
+    },
+    {
+      "epoch": 1.7635049384759698,
+      "grad_norm": 0.3545287251472473,
+      "learning_rate": 3.794869985870827e-05,
+      "loss": 2.9417,
+      "step": 23970
+    },
+    {
+      "epoch": 1.7642406518420424,
+      "grad_norm": 0.3504725992679596,
+      "learning_rate": 3.7910376812884255e-05,
+      "loss": 2.9744,
+      "step": 23980
+    },
+    {
+      "epoch": 1.764976365208115,
+      "grad_norm": 0.3473058044910431,
+      "learning_rate": 3.787206130877795e-05,
+      "loss": 2.993,
+      "step": 23990
+    },
+    {
+      "epoch": 1.7657120785741875,
+      "grad_norm": 0.348460853099823,
+      "learning_rate": 3.7833753370291276e-05,
+      "loss": 2.9838,
+      "step": 24000
+    },
+    {
+      "epoch": 1.76644779194026,
+      "grad_norm": 0.34356898069381714,
+      "learning_rate": 3.779545302132135e-05,
+      "loss": 2.9832,
+      "step": 24010
+    },
+    {
+      "epoch": 1.7671835053063325,
+      "grad_norm": 0.3549615740776062,
+      "learning_rate": 3.775716028576062e-05,
+      "loss": 2.9813,
+      "step": 24020
+    },
+    {
+      "epoch": 1.767919218672405,
+      "grad_norm": 0.34388306736946106,
+      "learning_rate": 3.771887518749678e-05,
+      "loss": 2.9577,
+      "step": 24030
+    },
+    {
+      "epoch": 1.7686549320384777,
+      "grad_norm": 0.3426467776298523,
+      "learning_rate": 3.768059775041272e-05,
+      "loss": 2.9297,
+      "step": 24040
+    },
+    {
+      "epoch": 1.7693906454045503,
+      "grad_norm": 0.34296128153800964,
+      "learning_rate": 3.764232799838658e-05,
+      "loss": 2.9298,
+      "step": 24050
+    },
+    {
+      "epoch": 1.7701263587706229,
+      "grad_norm": 0.3429625630378723,
+      "learning_rate": 3.760406595529168e-05,
+      "loss": 2.9642,
+      "step": 24060
+    },
+    {
+      "epoch": 1.7708620721366954,
+      "grad_norm": 0.3455885052680969,
+      "learning_rate": 3.7565811644996584e-05,
+      "loss": 2.9667,
+      "step": 24070
+    },
+    {
+      "epoch": 1.771597785502768,
+      "grad_norm": 0.3480741083621979,
+      "learning_rate": 3.7527565091364965e-05,
+      "loss": 2.9619,
+      "step": 24080
+    },
+    {
+      "epoch": 1.7723334988688406,
+      "grad_norm": 0.3456726670265198,
+      "learning_rate": 3.7489326318255714e-05,
+      "loss": 2.9482,
+      "step": 24090
+    },
+    {
+      "epoch": 1.7730692122349132,
+      "grad_norm": 0.3482351005077362,
+      "learning_rate": 3.74510953495228e-05,
+      "loss": 2.9794,
+      "step": 24100
+    },
+    {
+      "epoch": 1.7738049256009858,
+      "grad_norm": 0.3563176393508911,
+      "learning_rate": 3.7412872209015435e-05,
+      "loss": 2.9476,
+      "step": 24110
+    },
+    {
+      "epoch": 1.7745406389670584,
+      "grad_norm": 0.35728830099105835,
+      "learning_rate": 3.7374656920577836e-05,
+      "loss": 2.9598,
+      "step": 24120
+    },
+    {
+      "epoch": 1.775276352333131,
+      "grad_norm": 0.3425745666027069,
+      "learning_rate": 3.7336449508049384e-05,
+      "loss": 2.9739,
+      "step": 24130
+    },
+    {
+      "epoch": 1.7760120656992036,
+      "grad_norm": 0.351065993309021,
+      "learning_rate": 3.7298249995264515e-05,
+      "loss": 2.9722,
+      "step": 24140
+    },
+    {
+      "epoch": 1.7767477790652761,
+      "grad_norm": 0.34705230593681335,
+      "learning_rate": 3.726005840605277e-05,
+      "loss": 2.9699,
+      "step": 24150
+    },
+    {
+      "epoch": 1.7774834924313487,
+      "grad_norm": 0.3461216986179352,
+      "learning_rate": 3.722187476423871e-05,
+      "loss": 2.9474,
+      "step": 24160
+    },
+    {
+      "epoch": 1.7782192057974213,
+      "grad_norm": 0.3438127934932709,
+      "learning_rate": 3.7183699093641964e-05,
+      "loss": 2.9371,
+      "step": 24170
+    },
+    {
+      "epoch": 1.778954919163494,
+      "grad_norm": 0.3386172652244568,
+      "learning_rate": 3.714553141807717e-05,
+      "loss": 2.9803,
+      "step": 24180
+    },
+    {
+      "epoch": 1.7796906325295665,
+      "grad_norm": 0.3443130850791931,
+      "learning_rate": 3.7107371761354e-05,
+      "loss": 2.982,
+      "step": 24190
+    },
+    {
+      "epoch": 1.780426345895639,
+      "grad_norm": 0.3543626666069031,
+      "learning_rate": 3.706922014727714e-05,
+      "loss": 3.0101,
+      "step": 24200
+    },
+    {
+      "epoch": 1.7811620592617117,
+      "grad_norm": 0.34130531549453735,
+      "learning_rate": 3.703107659964617e-05,
+      "loss": 2.9567,
+      "step": 24210
+    },
+    {
+      "epoch": 1.7818977726277843,
+      "grad_norm": 0.342997670173645,
+      "learning_rate": 3.699294114225576e-05,
+      "loss": 2.8935,
+      "step": 24220
+    },
+    {
+      "epoch": 1.7826334859938568,
+      "grad_norm": 0.3510555922985077,
+      "learning_rate": 3.695481379889545e-05,
+      "loss": 2.9783,
+      "step": 24230
+    },
+    {
+      "epoch": 1.7833691993599294,
+      "grad_norm": 0.3421156704425812,
+      "learning_rate": 3.691669459334973e-05,
+      "loss": 2.9294,
+      "step": 24240
+    },
+    {
+      "epoch": 1.784104912726002,
+      "grad_norm": 0.36113783717155457,
+      "learning_rate": 3.6878583549398034e-05,
+      "loss": 2.9689,
+      "step": 24250
+    },
+    {
+      "epoch": 1.7848406260920746,
+      "grad_norm": 0.34186890721321106,
+      "learning_rate": 3.684048069081468e-05,
+      "loss": 2.9767,
+      "step": 24260
+    },
+    {
+      "epoch": 1.7855763394581472,
+      "grad_norm": 0.3636676073074341,
+      "learning_rate": 3.680238604136894e-05,
+      "loss": 2.9878,
+      "step": 24270
+    },
+    {
+      "epoch": 1.7863120528242198,
+      "grad_norm": 0.34688520431518555,
+      "learning_rate": 3.676429962482488e-05,
+      "loss": 2.994,
+      "step": 24280
+    },
+    {
+      "epoch": 1.7870477661902924,
+      "grad_norm": 0.34583836793899536,
+      "learning_rate": 3.672622146494149e-05,
+      "loss": 2.9419,
+      "step": 24290
+    },
+    {
+      "epoch": 1.787783479556365,
+      "grad_norm": 0.3408557176589966,
+      "learning_rate": 3.668815158547257e-05,
+      "loss": 2.9456,
+      "step": 24300
+    },
+    {
+      "epoch": 1.7885191929224376,
+      "grad_norm": 0.3453543782234192,
+      "learning_rate": 3.665009001016681e-05,
+      "loss": 3.002,
+      "step": 24310
+    },
+    {
+      "epoch": 1.7892549062885101,
+      "grad_norm": 0.34788358211517334,
+      "learning_rate": 3.661203676276764e-05,
+      "loss": 2.9672,
+      "step": 24320
+    },
+    {
+      "epoch": 1.7899906196545827,
+      "grad_norm": 0.33514100313186646,
+      "learning_rate": 3.657399186701338e-05,
+      "loss": 3.02,
+      "step": 24330
+    },
+    {
+      "epoch": 1.7907263330206553,
+      "grad_norm": 0.3358774483203888,
+      "learning_rate": 3.653595534663706e-05,
+      "loss": 2.9783,
+      "step": 24340
+    },
+    {
+      "epoch": 1.7914620463867277,
+      "grad_norm": 0.35053980350494385,
+      "learning_rate": 3.649792722536655e-05,
+      "loss": 2.9737,
+      "step": 24350
+    },
+    {
+      "epoch": 1.7921977597528003,
+      "grad_norm": 0.34285247325897217,
+      "learning_rate": 3.6459907526924464e-05,
+      "loss": 2.9117,
+      "step": 24360
+    },
+    {
+      "epoch": 1.7929334731188729,
+      "grad_norm": 0.36496245861053467,
+      "learning_rate": 3.6421896275028135e-05,
+      "loss": 2.9638,
+      "step": 24370
+    },
+    {
+      "epoch": 1.7936691864849454,
+      "grad_norm": 0.346748948097229,
+      "learning_rate": 3.638389349338965e-05,
+      "loss": 2.9741,
+      "step": 24380
+    },
+    {
+      "epoch": 1.794404899851018,
+      "grad_norm": 0.3442789614200592,
+      "learning_rate": 3.634589920571579e-05,
+      "loss": 2.9925,
+      "step": 24390
+    },
+    {
+      "epoch": 1.7951406132170906,
+      "grad_norm": 0.33688220381736755,
+      "learning_rate": 3.63079134357081e-05,
+      "loss": 2.9681,
+      "step": 24400
+    },
+    {
+      "epoch": 1.7958763265831632,
+      "grad_norm": 0.3508323132991791,
+      "learning_rate": 3.626993620706271e-05,
+      "loss": 2.9764,
+      "step": 24410
+    },
+    {
+      "epoch": 1.7966120399492358,
+      "grad_norm": 0.358918160200119,
+      "learning_rate": 3.623196754347049e-05,
+      "loss": 2.9833,
+      "step": 24420
+    },
+    {
+      "epoch": 1.7973477533153084,
+      "grad_norm": 0.3648023307323456,
+      "learning_rate": 3.619400746861699e-05,
+      "loss": 2.9777,
+      "step": 24430
+    },
+    {
+      "epoch": 1.798083466681381,
+      "grad_norm": 0.3423284590244293,
+      "learning_rate": 3.6156056006182335e-05,
+      "loss": 3.0023,
+      "step": 24440
+    },
+    {
+      "epoch": 1.7988191800474536,
+      "grad_norm": 0.34254464507102966,
+      "learning_rate": 3.611811317984133e-05,
+      "loss": 2.9068,
+      "step": 24450
+    },
+    {
+      "epoch": 1.799554893413526,
+      "grad_norm": 0.34269359707832336,
+      "learning_rate": 3.6080179013263347e-05,
+      "loss": 3.0089,
+      "step": 24460
+    },
+    {
+      "epoch": 1.8002906067795985,
+      "grad_norm": 0.355373740196228,
+      "learning_rate": 3.60422535301124e-05,
+      "loss": 2.9376,
+      "step": 24470
+    },
+    {
+      "epoch": 1.801026320145671,
+      "grad_norm": 0.3487183153629303,
+      "learning_rate": 3.6004336754047055e-05,
+      "loss": 2.9683,
+      "step": 24480
+    },
+    {
+      "epoch": 1.8017620335117437,
+      "grad_norm": 0.34797635674476624,
+      "learning_rate": 3.5966428708720475e-05,
+      "loss": 2.954,
+      "step": 24490
+    },
+    {
+      "epoch": 1.8024977468778163,
+      "grad_norm": 0.36070752143859863,
+      "learning_rate": 3.592852941778033e-05,
+      "loss": 2.9629,
+      "step": 24500
+    },
+    {
+      "epoch": 1.8032334602438889,
+      "grad_norm": 0.34586814045906067,
+      "learning_rate": 3.589063890486887e-05,
+      "loss": 2.9293,
+      "step": 24510
+    },
+    {
+      "epoch": 1.8039691736099615,
+      "grad_norm": 0.34678804874420166,
+      "learning_rate": 3.5852757193622874e-05,
+      "loss": 2.9463,
+      "step": 24520
+    },
+    {
+      "epoch": 1.804704886976034,
+      "grad_norm": 0.34796300530433655,
+      "learning_rate": 3.58148843076736e-05,
+      "loss": 2.9639,
+      "step": 24530
+    },
+    {
+      "epoch": 1.8054406003421066,
+      "grad_norm": 0.35904189944267273,
+      "learning_rate": 3.577702027064683e-05,
+      "loss": 2.9795,
+      "step": 24540
+    },
+    {
+      "epoch": 1.8061763137081792,
+      "grad_norm": 0.356326162815094,
+      "learning_rate": 3.573916510616278e-05,
+      "loss": 2.9668,
+      "step": 24550
+    },
+    {
+      "epoch": 1.8069120270742518,
+      "grad_norm": 0.3478321433067322,
+      "learning_rate": 3.5701318837836195e-05,
+      "loss": 2.9557,
+      "step": 24560
+    },
+    {
+      "epoch": 1.8076477404403244,
+      "grad_norm": 0.3518775999546051,
+      "learning_rate": 3.566348148927621e-05,
+      "loss": 3.0034,
+      "step": 24570
+    },
+    {
+      "epoch": 1.808383453806397,
+      "grad_norm": 0.34748753905296326,
+      "learning_rate": 3.562565308408645e-05,
+      "loss": 2.9767,
+      "step": 24580
+    },
+    {
+      "epoch": 1.8091191671724696,
+      "grad_norm": 0.34587937593460083,
+      "learning_rate": 3.558783364586489e-05,
+      "loss": 2.9935,
+      "step": 24590
+    },
+    {
+      "epoch": 1.8098548805385422,
+      "grad_norm": 0.35243189334869385,
+      "learning_rate": 3.5550023198203995e-05,
+      "loss": 2.9643,
+      "step": 24600
+    },
+    {
+      "epoch": 1.8105905939046147,
+      "grad_norm": 0.3433395326137543,
+      "learning_rate": 3.551222176469058e-05,
+      "loss": 2.9449,
+      "step": 24610
+    },
+    {
+      "epoch": 1.8113263072706873,
+      "grad_norm": 0.3397443890571594,
+      "learning_rate": 3.547442936890582e-05,
+      "loss": 2.9973,
+      "step": 24620
+    },
+    {
+      "epoch": 1.81206202063676,
+      "grad_norm": 0.35141521692276,
+      "learning_rate": 3.5436646034425294e-05,
+      "loss": 2.9675,
+      "step": 24630
+    },
+    {
+      "epoch": 1.8127977340028325,
+      "grad_norm": 0.341878741979599,
+      "learning_rate": 3.539887178481886e-05,
+      "loss": 2.9707,
+      "step": 24640
+    },
+    {
+      "epoch": 1.813533447368905,
+      "grad_norm": 0.33795565366744995,
+      "learning_rate": 3.536110664365081e-05,
+      "loss": 2.9374,
+      "step": 24650
+    },
+    {
+      "epoch": 1.8142691607349777,
+      "grad_norm": 0.3453797399997711,
+      "learning_rate": 3.5323350634479656e-05,
+      "loss": 2.9813,
+      "step": 24660
+    },
+    {
+      "epoch": 1.8150048741010503,
+      "grad_norm": 0.35093796253204346,
+      "learning_rate": 3.5285603780858256e-05,
+      "loss": 2.9448,
+      "step": 24670
+    },
+    {
+      "epoch": 1.8157405874671229,
+      "grad_norm": 0.353671669960022,
+      "learning_rate": 3.524786610633379e-05,
+      "loss": 2.9704,
+      "step": 24680
+    },
+    {
+      "epoch": 1.8164763008331954,
+      "grad_norm": 0.34835854172706604,
+      "learning_rate": 3.5210137634447645e-05,
+      "loss": 3.0039,
+      "step": 24690
+    },
+    {
+      "epoch": 1.817212014199268,
+      "grad_norm": 0.34317857027053833,
+      "learning_rate": 3.5172418388735524e-05,
+      "loss": 2.9522,
+      "step": 24700
+    },
+    {
+      "epoch": 1.8179477275653406,
+      "grad_norm": 0.34638264775276184,
+      "learning_rate": 3.5134708392727323e-05,
+      "loss": 2.987,
+      "step": 24710
+    },
+    {
+      "epoch": 1.8186834409314132,
+      "grad_norm": 0.3505925238132477,
+      "learning_rate": 3.509700766994721e-05,
+      "loss": 3.0224,
+      "step": 24720
+    },
+    {
+      "epoch": 1.8194191542974858,
+      "grad_norm": 0.3413071632385254,
+      "learning_rate": 3.5059316243913554e-05,
+      "loss": 2.9431,
+      "step": 24730
+    },
+    {
+      "epoch": 1.8201548676635584,
+      "grad_norm": 0.34339675307273865,
+      "learning_rate": 3.5021634138138934e-05,
+      "loss": 2.9706,
+      "step": 24740
+    },
+    {
+      "epoch": 1.820890581029631,
+      "grad_norm": 0.3480238616466522,
+      "learning_rate": 3.498396137613006e-05,
+      "loss": 2.9947,
+      "step": 24750
+    },
+    {
+      "epoch": 1.8216262943957036,
+      "grad_norm": 0.3603719472885132,
+      "learning_rate": 3.494629798138791e-05,
+      "loss": 2.9959,
+      "step": 24760
+    },
+    {
+      "epoch": 1.8223620077617761,
+      "grad_norm": 0.3514951765537262,
+      "learning_rate": 3.490864397740755e-05,
+      "loss": 2.8946,
+      "step": 24770
+    },
+    {
+      "epoch": 1.8230977211278487,
+      "grad_norm": 0.34638676047325134,
+      "learning_rate": 3.487099938767819e-05,
+      "loss": 2.958,
+      "step": 24780
+    },
+    {
+      "epoch": 1.8238334344939213,
+      "grad_norm": 0.3388519585132599,
+      "learning_rate": 3.48333642356832e-05,
+      "loss": 2.9545,
+      "step": 24790
+    },
+    {
+      "epoch": 1.824569147859994,
+      "grad_norm": 0.34840163588523865,
+      "learning_rate": 3.479573854490002e-05,
+      "loss": 2.947,
+      "step": 24800
+    },
+    {
+      "epoch": 1.8253048612260663,
+      "grad_norm": 0.35466378927230835,
+      "learning_rate": 3.475812233880026e-05,
+      "loss": 2.9884,
+      "step": 24810
+    },
+    {
+      "epoch": 1.8260405745921389,
+      "grad_norm": 0.34805136919021606,
+      "learning_rate": 3.4720515640849504e-05,
+      "loss": 2.9565,
+      "step": 24820
+    },
+    {
+      "epoch": 1.8267762879582115,
+      "grad_norm": 0.3465035855770111,
+      "learning_rate": 3.468291847450751e-05,
+      "loss": 2.929,
+      "step": 24830
+    },
+    {
+      "epoch": 1.827512001324284,
+      "grad_norm": 0.3470146954059601,
+      "learning_rate": 3.464533086322801e-05,
+      "loss": 2.952,
+      "step": 24840
+    },
+    {
+      "epoch": 1.8282477146903566,
+      "grad_norm": 0.34149134159088135,
+      "learning_rate": 3.4607752830458836e-05,
+      "loss": 3.0038,
+      "step": 24850
+    },
+    {
+      "epoch": 1.8289834280564292,
+      "grad_norm": 0.35501939058303833,
+      "learning_rate": 3.457018439964183e-05,
+      "loss": 2.9526,
+      "step": 24860
+    },
+    {
+      "epoch": 1.8297191414225018,
+      "grad_norm": 0.35745781660079956,
+      "learning_rate": 3.4532625594212813e-05,
+      "loss": 2.9572,
+      "step": 24870
+    },
+    {
+      "epoch": 1.8304548547885744,
+      "grad_norm": 0.3456425666809082,
+      "learning_rate": 3.449507643760164e-05,
+      "loss": 2.8928,
+      "step": 24880
+    },
+    {
+      "epoch": 1.831190568154647,
+      "grad_norm": 0.34564852714538574,
+      "learning_rate": 3.445753695323211e-05,
+      "loss": 2.9903,
+      "step": 24890
+    },
+    {
+      "epoch": 1.8319262815207196,
+      "grad_norm": 0.3421176075935364,
+      "learning_rate": 3.442000716452202e-05,
+      "loss": 2.9358,
+      "step": 24900
+    },
+    {
+      "epoch": 1.8326619948867922,
+      "grad_norm": 0.3563895523548126,
+      "learning_rate": 3.4382487094883095e-05,
+      "loss": 2.9826,
+      "step": 24910
+    },
+    {
+      "epoch": 1.8333977082528645,
+      "grad_norm": 0.3328886926174164,
+      "learning_rate": 3.4344976767721016e-05,
+      "loss": 2.9894,
+      "step": 24920
+    },
+    {
+      "epoch": 1.834133421618937,
+      "grad_norm": 0.3458249568939209,
+      "learning_rate": 3.430747620643539e-05,
+      "loss": 3.0145,
+      "step": 24930
+    },
+    {
+      "epoch": 1.8348691349850097,
+      "grad_norm": 0.36163264513015747,
+      "learning_rate": 3.4269985434419707e-05,
+      "loss": 2.9601,
+      "step": 24940
+    },
+    {
+      "epoch": 1.8356048483510823,
+      "grad_norm": 0.35502487421035767,
+      "learning_rate": 3.4232504475061375e-05,
+      "loss": 2.9315,
+      "step": 24950
+    },
+    {
+      "epoch": 1.8363405617171549,
+      "grad_norm": 0.3413659930229187,
+      "learning_rate": 3.419503335174166e-05,
+      "loss": 2.9428,
+      "step": 24960
+    },
+    {
+      "epoch": 1.8370762750832275,
+      "grad_norm": 0.35091882944107056,
+      "learning_rate": 3.415757208783571e-05,
+      "loss": 2.95,
+      "step": 24970
+    },
+    {
+      "epoch": 1.8378119884493,
+      "grad_norm": 0.3446294963359833,
+      "learning_rate": 3.41201207067125e-05,
+      "loss": 2.9596,
+      "step": 24980
+    },
+    {
+      "epoch": 1.8385477018153726,
+      "grad_norm": 0.3647192418575287,
+      "learning_rate": 3.408267923173487e-05,
+      "loss": 2.953,
+      "step": 24990
+    },
+    {
+      "epoch": 1.8392834151814452,
+      "grad_norm": 0.34038805961608887,
+      "learning_rate": 3.404524768625943e-05,
+      "loss": 2.9845,
+      "step": 25000
+    },
+    {
+      "epoch": 1.8400191285475178,
+      "grad_norm": 0.3418276607990265,
+      "learning_rate": 3.40078260936367e-05,
+      "loss": 2.9723,
+      "step": 25010
+    },
+    {
+      "epoch": 1.8407548419135904,
+      "grad_norm": 0.3532450497150421,
+      "learning_rate": 3.397041447721087e-05,
+      "loss": 2.9969,
+      "step": 25020
+    },
+    {
+      "epoch": 1.841490555279663,
+      "grad_norm": 0.3488118648529053,
+      "learning_rate": 3.393301286031998e-05,
+      "loss": 2.9932,
+      "step": 25030
+    },
+    {
+      "epoch": 1.8422262686457356,
+      "grad_norm": 0.34950822591781616,
+      "learning_rate": 3.389562126629579e-05,
+      "loss": 2.9855,
+      "step": 25040
+    },
+    {
+      "epoch": 1.8429619820118082,
+      "grad_norm": 0.3663720488548279,
+      "learning_rate": 3.385823971846385e-05,
+      "loss": 2.9456,
+      "step": 25050
+    },
+    {
+      "epoch": 1.8436976953778808,
+      "grad_norm": 0.3517078757286072,
+      "learning_rate": 3.3820868240143394e-05,
+      "loss": 2.9459,
+      "step": 25060
+    },
+    {
+      "epoch": 1.8444334087439533,
+      "grad_norm": 0.3449481427669525,
+      "learning_rate": 3.378350685464744e-05,
+      "loss": 3.0274,
+      "step": 25070
+    },
+    {
+      "epoch": 1.845169122110026,
+      "grad_norm": 0.34110262989997864,
+      "learning_rate": 3.3746155585282625e-05,
+      "loss": 2.9136,
+      "step": 25080
+    },
+    {
+      "epoch": 1.8459048354760985,
+      "grad_norm": 0.34856221079826355,
+      "learning_rate": 3.370881445534936e-05,
+      "loss": 2.9657,
+      "step": 25090
+    },
+    {
+      "epoch": 1.846640548842171,
+      "grad_norm": 0.3476397395133972,
+      "learning_rate": 3.3671483488141684e-05,
+      "loss": 2.9729,
+      "step": 25100
+    },
+    {
+      "epoch": 1.8473762622082437,
+      "grad_norm": 0.35501065850257874,
+      "learning_rate": 3.363416270694729e-05,
+      "loss": 2.9672,
+      "step": 25110
+    },
+    {
+      "epoch": 1.8481119755743163,
+      "grad_norm": 0.34786567091941833,
+      "learning_rate": 3.3596852135047564e-05,
+      "loss": 2.9396,
+      "step": 25120
+    },
+    {
+      "epoch": 1.8488476889403889,
+      "grad_norm": 0.37141814827919006,
+      "learning_rate": 3.355955179571746e-05,
+      "loss": 2.956,
+      "step": 25130
+    },
+    {
+      "epoch": 1.8495834023064615,
+      "grad_norm": 0.3310892879962921,
+      "learning_rate": 3.35222617122256e-05,
+      "loss": 2.9603,
+      "step": 25140
+    },
+    {
+      "epoch": 1.850319115672534,
+      "grad_norm": 0.3526333272457123,
+      "learning_rate": 3.348498190783417e-05,
+      "loss": 2.9585,
+      "step": 25150
+    },
+    {
+      "epoch": 1.8510548290386066,
+      "grad_norm": 0.35179266333580017,
+      "learning_rate": 3.344771240579897e-05,
+      "loss": 2.9621,
+      "step": 25160
+    },
+    {
+      "epoch": 1.8517905424046792,
+      "grad_norm": 0.3478565514087677,
+      "learning_rate": 3.341045322936935e-05,
+      "loss": 2.9458,
+      "step": 25170
+    },
+    {
+      "epoch": 1.8525262557707518,
+      "grad_norm": 0.3463594317436218,
+      "learning_rate": 3.3373204401788275e-05,
+      "loss": 2.9944,
+      "step": 25180
+    },
+    {
+      "epoch": 1.8532619691368244,
+      "grad_norm": 0.34940028190612793,
+      "learning_rate": 3.333596594629218e-05,
+      "loss": 2.9671,
+      "step": 25190
+    },
+    {
+      "epoch": 1.853997682502897,
+      "grad_norm": 0.3455345630645752,
+      "learning_rate": 3.329873788611108e-05,
+      "loss": 2.9281,
+      "step": 25200
+    },
+    {
+      "epoch": 1.8547333958689696,
+      "grad_norm": 0.3563336730003357,
+      "learning_rate": 3.3261520244468475e-05,
+      "loss": 2.9705,
+      "step": 25210
+    },
+    {
+      "epoch": 1.8554691092350422,
+      "grad_norm": 0.33494168519973755,
+      "learning_rate": 3.322431304458138e-05,
+      "loss": 2.9701,
+      "step": 25220
+    },
+    {
+      "epoch": 1.8562048226011147,
+      "grad_norm": 0.3445129692554474,
+      "learning_rate": 3.318711630966031e-05,
+      "loss": 2.9766,
+      "step": 25230
+    },
+    {
+      "epoch": 1.8569405359671873,
+      "grad_norm": 0.3546302020549774,
+      "learning_rate": 3.31499300629092e-05,
+      "loss": 2.9487,
+      "step": 25240
+    },
+    {
+      "epoch": 1.85767624933326,
+      "grad_norm": 0.3464782238006592,
+      "learning_rate": 3.311275432752551e-05,
+      "loss": 2.9122,
+      "step": 25250
+    },
+    {
+      "epoch": 1.8584119626993323,
+      "grad_norm": 0.3499259352684021,
+      "learning_rate": 3.307558912670011e-05,
+      "loss": 2.9505,
+      "step": 25260
+    },
+    {
+      "epoch": 1.8591476760654049,
+      "grad_norm": 0.3440198004245758,
+      "learning_rate": 3.3038434483617295e-05,
+      "loss": 2.9845,
+      "step": 25270
+    },
+    {
+      "epoch": 1.8598833894314775,
+      "grad_norm": 0.3410769999027252,
+      "learning_rate": 3.300129042145479e-05,
+      "loss": 2.9392,
+      "step": 25280
+    },
+    {
+      "epoch": 1.86061910279755,
+      "grad_norm": 0.35861387848854065,
+      "learning_rate": 3.2964156963383695e-05,
+      "loss": 2.9545,
+      "step": 25290
+    },
+    {
+      "epoch": 1.8613548161636226,
+      "grad_norm": 0.3518628776073456,
+      "learning_rate": 3.2927034132568515e-05,
+      "loss": 2.9591,
+      "step": 25300
+    },
+    {
+      "epoch": 1.8620905295296952,
+      "grad_norm": 0.36279451847076416,
+      "learning_rate": 3.2889921952167106e-05,
+      "loss": 2.9984,
+      "step": 25310
+    },
+    {
+      "epoch": 1.8628262428957678,
+      "grad_norm": 0.35148653388023376,
+      "learning_rate": 3.2852820445330716e-05,
+      "loss": 2.9613,
+      "step": 25320
+    },
+    {
+      "epoch": 1.8635619562618404,
+      "grad_norm": 0.36404699087142944,
+      "learning_rate": 3.2815729635203876e-05,
+      "loss": 2.9983,
+      "step": 25330
+    },
+    {
+      "epoch": 1.864297669627913,
+      "grad_norm": 0.3519551753997803,
+      "learning_rate": 3.2778649544924524e-05,
+      "loss": 2.9496,
+      "step": 25340
+    },
+    {
+      "epoch": 1.8650333829939856,
+      "grad_norm": 0.3392998278141022,
+      "learning_rate": 3.274158019762384e-05,
+      "loss": 2.9619,
+      "step": 25350
+    },
+    {
+      "epoch": 1.8657690963600582,
+      "grad_norm": 0.3466440439224243,
+      "learning_rate": 3.270452161642634e-05,
+      "loss": 2.9492,
+      "step": 25360
+    },
+    {
+      "epoch": 1.8665048097261308,
+      "grad_norm": 0.3603527545928955,
+      "learning_rate": 3.2667473824449814e-05,
+      "loss": 2.9636,
+      "step": 25370
+    },
+    {
+      "epoch": 1.8672405230922031,
+      "grad_norm": 0.34616658091545105,
+      "learning_rate": 3.26304368448053e-05,
+      "loss": 2.9669,
+      "step": 25380
+    },
+    {
+      "epoch": 1.8679762364582757,
+      "grad_norm": 0.3375670313835144,
+      "learning_rate": 3.259341070059715e-05,
+      "loss": 2.9133,
+      "step": 25390
+    },
+    {
+      "epoch": 1.8687119498243483,
+      "grad_norm": 0.3607102036476135,
+      "learning_rate": 3.255639541492287e-05,
+      "loss": 2.9662,
+      "step": 25400
+    },
+    {
+      "epoch": 1.8694476631904209,
+      "grad_norm": 0.3704841732978821,
+      "learning_rate": 3.251939101087326e-05,
+      "loss": 2.986,
+      "step": 25410
+    },
+    {
+      "epoch": 1.8701833765564935,
+      "grad_norm": 0.3471236526966095,
+      "learning_rate": 3.248239751153235e-05,
+      "loss": 2.9211,
+      "step": 25420
+    },
+    {
+      "epoch": 1.870919089922566,
+      "grad_norm": 0.36135485768318176,
+      "learning_rate": 3.244541493997727e-05,
+      "loss": 2.9704,
+      "step": 25430
+    },
+    {
+      "epoch": 1.8716548032886386,
+      "grad_norm": 0.35255685448646545,
+      "learning_rate": 3.240844331927844e-05,
+      "loss": 2.9719,
+      "step": 25440
+    },
+    {
+      "epoch": 1.8723905166547112,
+      "grad_norm": 0.3549046218395233,
+      "learning_rate": 3.237148267249937e-05,
+      "loss": 2.9752,
+      "step": 25450
+    },
+    {
+      "epoch": 1.8731262300207838,
+      "grad_norm": 0.351749986410141,
+      "learning_rate": 3.233453302269678e-05,
+      "loss": 2.9975,
+      "step": 25460
+    },
+    {
+      "epoch": 1.8738619433868564,
+      "grad_norm": 0.33410558104515076,
+      "learning_rate": 3.229759439292049e-05,
+      "loss": 2.9567,
+      "step": 25470
+    },
+    {
+      "epoch": 1.874597656752929,
+      "grad_norm": 0.3621499538421631,
+      "learning_rate": 3.226066680621347e-05,
+      "loss": 2.9644,
+      "step": 25480
+    },
+    {
+      "epoch": 1.8753333701190016,
+      "grad_norm": 0.3556647002696991,
+      "learning_rate": 3.222375028561176e-05,
+      "loss": 2.9686,
+      "step": 25490
+    },
+    {
+      "epoch": 1.8760690834850742,
+      "grad_norm": 0.3528843820095062,
+      "learning_rate": 3.2186844854144584e-05,
+      "loss": 2.9756,
+      "step": 25500
+    },
+    {
+      "epoch": 1.8768047968511468,
+      "grad_norm": 0.3458603620529175,
+      "learning_rate": 3.2149950534834166e-05,
+      "loss": 2.924,
+      "step": 25510
+    },
+    {
+      "epoch": 1.8775405102172193,
+      "grad_norm": 0.3335363268852234,
+      "learning_rate": 3.211306735069582e-05,
+      "loss": 2.9506,
+      "step": 25520
+    },
+    {
+      "epoch": 1.878276223583292,
+      "grad_norm": 0.34703218936920166,
+      "learning_rate": 3.2076195324737933e-05,
+      "loss": 2.9222,
+      "step": 25530
+    },
+    {
+      "epoch": 1.8790119369493645,
+      "grad_norm": 0.3623386025428772,
+      "learning_rate": 3.2039334479961905e-05,
+      "loss": 2.9029,
+      "step": 25540
+    },
+    {
+      "epoch": 1.8797476503154371,
+      "grad_norm": 0.3553720712661743,
+      "learning_rate": 3.2002484839362176e-05,
+      "loss": 2.9415,
+      "step": 25550
+    },
+    {
+      "epoch": 1.8804833636815097,
+      "grad_norm": 0.35573941469192505,
+      "learning_rate": 3.1965646425926205e-05,
+      "loss": 2.9879,
+      "step": 25560
+    },
+    {
+      "epoch": 1.8812190770475823,
+      "grad_norm": 0.36144721508026123,
+      "learning_rate": 3.192881926263443e-05,
+      "loss": 3.0114,
+      "step": 25570
+    },
+    {
+      "epoch": 1.8819547904136549,
+      "grad_norm": 0.3427868187427521,
+      "learning_rate": 3.1892003372460255e-05,
+      "loss": 2.9676,
+      "step": 25580
+    },
+    {
+      "epoch": 1.8826905037797275,
+      "grad_norm": 0.34727516770362854,
+      "learning_rate": 3.185519877837011e-05,
+      "loss": 3.0012,
+      "step": 25590
+    },
+    {
+      "epoch": 1.8834262171458,
+      "grad_norm": 0.3481018543243408,
+      "learning_rate": 3.181840550332334e-05,
+      "loss": 2.9847,
+      "step": 25600
+    },
+    {
+      "epoch": 1.8841619305118726,
+      "grad_norm": 0.33993643522262573,
+      "learning_rate": 3.1781623570272206e-05,
+      "loss": 2.945,
+      "step": 25610
+    },
+    {
+      "epoch": 1.8848976438779452,
+      "grad_norm": 0.34721189737319946,
+      "learning_rate": 3.1744853002161946e-05,
+      "loss": 2.9409,
+      "step": 25620
+    },
+    {
+      "epoch": 1.8856333572440178,
+      "grad_norm": 0.339315265417099,
+      "learning_rate": 3.170809382193065e-05,
+      "loss": 2.9121,
+      "step": 25630
+    },
+    {
+      "epoch": 1.8863690706100904,
+      "grad_norm": 0.3456818163394928,
+      "learning_rate": 3.167134605250938e-05,
+      "loss": 2.9945,
+      "step": 25640
+    },
+    {
+      "epoch": 1.887104783976163,
+      "grad_norm": 0.34596115350723267,
+      "learning_rate": 3.163460971682198e-05,
+      "loss": 2.9669,
+      "step": 25650
+    },
+    {
+      "epoch": 1.8878404973422356,
+      "grad_norm": 0.34914010763168335,
+      "learning_rate": 3.1597884837785246e-05,
+      "loss": 2.9728,
+      "step": 25660
+    },
+    {
+      "epoch": 1.8885762107083082,
+      "grad_norm": 0.3455921709537506,
+      "learning_rate": 3.156117143830882e-05,
+      "loss": 2.9392,
+      "step": 25670
+    },
+    {
+      "epoch": 1.8893119240743808,
+      "grad_norm": 0.344281941652298,
+      "learning_rate": 3.152446954129512e-05,
+      "loss": 2.9636,
+      "step": 25680
+    },
+    {
+      "epoch": 1.8900476374404533,
+      "grad_norm": 0.3552781343460083,
+      "learning_rate": 3.1487779169639465e-05,
+      "loss": 2.9434,
+      "step": 25690
+    },
+    {
+      "epoch": 1.890783350806526,
+      "grad_norm": 0.34762638807296753,
+      "learning_rate": 3.145110034622992e-05,
+      "loss": 2.9824,
+      "step": 25700
+    },
+    {
+      "epoch": 1.8915190641725985,
+      "grad_norm": 0.3525771498680115,
+      "learning_rate": 3.141443309394739e-05,
+      "loss": 2.9739,
+      "step": 25710
+    },
+    {
+      "epoch": 1.8922547775386709,
+      "grad_norm": 0.3449614644050598,
+      "learning_rate": 3.137777743566554e-05,
+      "loss": 2.9553,
+      "step": 25720
+    },
+    {
+      "epoch": 1.8929904909047435,
+      "grad_norm": 0.3569880723953247,
+      "learning_rate": 3.134113339425082e-05,
+      "loss": 2.9731,
+      "step": 25730
+    },
+    {
+      "epoch": 1.893726204270816,
+      "grad_norm": 0.34848809242248535,
+      "learning_rate": 3.1304500992562394e-05,
+      "loss": 2.9869,
+      "step": 25740
+    },
+    {
+      "epoch": 1.8944619176368886,
+      "grad_norm": 0.36170753836631775,
+      "learning_rate": 3.126788025345221e-05,
+      "loss": 2.9676,
+      "step": 25750
+    },
+    {
+      "epoch": 1.8951976310029612,
+      "grad_norm": 0.3520166277885437,
+      "learning_rate": 3.123127119976494e-05,
+      "loss": 2.9214,
+      "step": 25760
+    },
+    {
+      "epoch": 1.8959333443690338,
+      "grad_norm": 0.3511207699775696,
+      "learning_rate": 3.1194673854337936e-05,
+      "loss": 2.9334,
+      "step": 25770
+    },
+    {
+      "epoch": 1.8966690577351064,
+      "grad_norm": 0.34764185547828674,
+      "learning_rate": 3.115808824000126e-05,
+      "loss": 2.9296,
+      "step": 25780
+    },
+    {
+      "epoch": 1.897404771101179,
+      "grad_norm": 0.34808072447776794,
+      "learning_rate": 3.112151437957765e-05,
+      "loss": 2.9677,
+      "step": 25790
+    },
+    {
+      "epoch": 1.8981404844672516,
+      "grad_norm": 0.3493986427783966,
+      "learning_rate": 3.108495229588253e-05,
+      "loss": 2.9442,
+      "step": 25800
+    },
+    {
+      "epoch": 1.8988761978333242,
+      "grad_norm": 0.3441488444805145,
+      "learning_rate": 3.104840201172395e-05,
+      "loss": 2.9969,
+      "step": 25810
+    },
+    {
+      "epoch": 1.8996119111993968,
+      "grad_norm": 0.351593554019928,
+      "learning_rate": 3.101186354990261e-05,
+      "loss": 2.9382,
+      "step": 25820
+    },
+    {
+      "epoch": 1.9003476245654691,
+      "grad_norm": 0.3503246307373047,
+      "learning_rate": 3.097533693321186e-05,
+      "loss": 2.9625,
+      "step": 25830
+    },
+    {
+      "epoch": 1.9010833379315417,
+      "grad_norm": 0.34619203209877014,
+      "learning_rate": 3.093882218443765e-05,
+      "loss": 2.9163,
+      "step": 25840
+    },
+    {
+      "epoch": 1.9018190512976143,
+      "grad_norm": 0.36059218645095825,
+      "learning_rate": 3.0902319326358484e-05,
+      "loss": 2.9944,
+      "step": 25850
+    },
+    {
+      "epoch": 1.902554764663687,
+      "grad_norm": 0.3493688106536865,
+      "learning_rate": 3.086582838174551e-05,
+      "loss": 2.9608,
+      "step": 25860
+    },
+    {
+      "epoch": 1.9032904780297595,
+      "grad_norm": 0.35749414563179016,
+      "learning_rate": 3.0829349373362406e-05,
+      "loss": 2.9186,
+      "step": 25870
+    },
+    {
+      "epoch": 1.904026191395832,
+      "grad_norm": 0.3555651009082794,
+      "learning_rate": 3.07928823239654e-05,
+      "loss": 2.9443,
+      "step": 25880
+    },
+    {
+      "epoch": 1.9047619047619047,
+      "grad_norm": 0.34943947196006775,
+      "learning_rate": 3.0756427256303296e-05,
+      "loss": 2.9469,
+      "step": 25890
+    },
+    {
+      "epoch": 1.9054976181279772,
+      "grad_norm": 0.3470650017261505,
+      "learning_rate": 3.0719984193117366e-05,
+      "loss": 2.9533,
+      "step": 25900
+    },
+    {
+      "epoch": 1.9062333314940498,
+      "grad_norm": 0.3506460189819336,
+      "learning_rate": 3.0683553157141454e-05,
+      "loss": 2.9815,
+      "step": 25910
+    },
+    {
+      "epoch": 1.9069690448601224,
+      "grad_norm": 0.34140074253082275,
+      "learning_rate": 3.0647134171101875e-05,
+      "loss": 2.961,
+      "step": 25920
+    },
+    {
+      "epoch": 1.907704758226195,
+      "grad_norm": 0.3473810851573944,
+      "learning_rate": 3.0610727257717433e-05,
+      "loss": 3.0333,
+      "step": 25930
+    },
+    {
+      "epoch": 1.9084404715922676,
+      "grad_norm": 0.3438740074634552,
+      "learning_rate": 3.0574332439699374e-05,
+      "loss": 2.9452,
+      "step": 25940
+    },
+    {
+      "epoch": 1.9091761849583402,
+      "grad_norm": 0.3498131334781647,
+      "learning_rate": 3.0537949739751446e-05,
+      "loss": 2.9719,
+      "step": 25950
+    },
+    {
+      "epoch": 1.9099118983244128,
+      "grad_norm": 0.3523942828178406,
+      "learning_rate": 3.0501579180569795e-05,
+      "loss": 3.0062,
+      "step": 25960
+    },
+    {
+      "epoch": 1.9106476116904854,
+      "grad_norm": 0.3532496988773346,
+      "learning_rate": 3.0465220784843018e-05,
+      "loss": 2.9332,
+      "step": 25970
+    },
+    {
+      "epoch": 1.911383325056558,
+      "grad_norm": 0.3434400260448456,
+      "learning_rate": 3.0428874575252097e-05,
+      "loss": 2.9425,
+      "step": 25980
+    },
+    {
+      "epoch": 1.9121190384226305,
+      "grad_norm": 0.3459074795246124,
+      "learning_rate": 3.039254057447044e-05,
+      "loss": 2.9751,
+      "step": 25990
+    },
+    {
+      "epoch": 1.9128547517887031,
+      "grad_norm": 0.3458581864833832,
+      "learning_rate": 3.035621880516386e-05,
+      "loss": 2.9651,
+      "step": 26000
+    },
+    {
+      "epoch": 1.9135904651547757,
+      "grad_norm": 0.3632730543613434,
+      "learning_rate": 3.0319909289990466e-05,
+      "loss": 3.0015,
+      "step": 26010
+    },
+    {
+      "epoch": 1.9143261785208483,
+      "grad_norm": 0.3605930507183075,
+      "learning_rate": 3.0283612051600805e-05,
+      "loss": 2.9496,
+      "step": 26020
+    },
+    {
+      "epoch": 1.9150618918869209,
+      "grad_norm": 0.34854891896247864,
+      "learning_rate": 3.024732711263769e-05,
+      "loss": 2.9191,
+      "step": 26030
+    },
+    {
+      "epoch": 1.9157976052529935,
+      "grad_norm": 0.34203699231147766,
+      "learning_rate": 3.0211054495736336e-05,
+      "loss": 2.9477,
+      "step": 26040
+    },
+    {
+      "epoch": 1.916533318619066,
+      "grad_norm": 0.3508712649345398,
+      "learning_rate": 3.0174794223524207e-05,
+      "loss": 2.9742,
+      "step": 26050
+    },
+    {
+      "epoch": 1.9172690319851386,
+      "grad_norm": 0.3500581979751587,
+      "learning_rate": 3.0138546318621098e-05,
+      "loss": 2.9272,
+      "step": 26060
+    },
+    {
+      "epoch": 1.9180047453512112,
+      "grad_norm": 0.349373459815979,
+      "learning_rate": 3.010231080363908e-05,
+      "loss": 2.9568,
+      "step": 26070
+    },
+    {
+      "epoch": 1.9187404587172838,
+      "grad_norm": 0.3544924259185791,
+      "learning_rate": 3.006608770118251e-05,
+      "loss": 2.8947,
+      "step": 26080
+    },
+    {
+      "epoch": 1.9194761720833564,
+      "grad_norm": 0.3337794542312622,
+      "learning_rate": 3.0029877033847998e-05,
+      "loss": 2.9187,
+      "step": 26090
+    },
+    {
+      "epoch": 1.920211885449429,
+      "grad_norm": 0.3485521376132965,
+      "learning_rate": 2.9993678824224363e-05,
+      "loss": 2.9424,
+      "step": 26100
+    },
+    {
+      "epoch": 1.9209475988155016,
+      "grad_norm": 0.35006311535835266,
+      "learning_rate": 2.9957493094892708e-05,
+      "loss": 2.9689,
+      "step": 26110
+    },
+    {
+      "epoch": 1.9216833121815742,
+      "grad_norm": 0.33827105164527893,
+      "learning_rate": 2.9921319868426302e-05,
+      "loss": 2.9324,
+      "step": 26120
+    },
+    {
+      "epoch": 1.9224190255476468,
+      "grad_norm": 0.34388259053230286,
+      "learning_rate": 2.9885159167390652e-05,
+      "loss": 2.9453,
+      "step": 26130
+    },
+    {
+      "epoch": 1.9231547389137194,
+      "grad_norm": 0.35847002267837524,
+      "learning_rate": 2.984901101434341e-05,
+      "loss": 3.0097,
+      "step": 26140
+    },
+    {
+      "epoch": 1.923890452279792,
+      "grad_norm": 0.3471504747867584,
+      "learning_rate": 2.981287543183443e-05,
+      "loss": 2.9196,
+      "step": 26150
+    },
+    {
+      "epoch": 1.9246261656458645,
+      "grad_norm": 0.34413838386535645,
+      "learning_rate": 2.9776752442405743e-05,
+      "loss": 2.9287,
+      "step": 26160
+    },
+    {
+      "epoch": 1.925361879011937,
+      "grad_norm": 0.3435196876525879,
+      "learning_rate": 2.9740642068591474e-05,
+      "loss": 2.9203,
+      "step": 26170
+    },
+    {
+      "epoch": 1.9260975923780095,
+      "grad_norm": 0.3407808542251587,
+      "learning_rate": 2.9704544332917915e-05,
+      "loss": 2.929,
+      "step": 26180
+    },
+    {
+      "epoch": 1.926833305744082,
+      "grad_norm": 0.3369019031524658,
+      "learning_rate": 2.9668459257903457e-05,
+      "loss": 2.9717,
+      "step": 26190
+    },
+    {
+      "epoch": 1.9275690191101547,
+      "grad_norm": 0.3564346432685852,
+      "learning_rate": 2.9632386866058608e-05,
+      "loss": 2.9418,
+      "step": 26200
+    },
+    {
+      "epoch": 1.9283047324762272,
+      "grad_norm": 0.3436613380908966,
+      "learning_rate": 2.9596327179885943e-05,
+      "loss": 2.913,
+      "step": 26210
+    },
+    {
+      "epoch": 1.9290404458422998,
+      "grad_norm": 0.3601857125759125,
+      "learning_rate": 2.9560280221880132e-05,
+      "loss": 2.9207,
+      "step": 26220
+    },
+    {
+      "epoch": 1.9297761592083724,
+      "grad_norm": 0.3765910863876343,
+      "learning_rate": 2.952424601452789e-05,
+      "loss": 2.9711,
+      "step": 26230
+    },
+    {
+      "epoch": 1.930511872574445,
+      "grad_norm": 0.35654014348983765,
+      "learning_rate": 2.9488224580307977e-05,
+      "loss": 2.9123,
+      "step": 26240
+    },
+    {
+      "epoch": 1.9312475859405176,
+      "grad_norm": 0.3340204358100891,
+      "learning_rate": 2.9452215941691214e-05,
+      "loss": 2.9223,
+      "step": 26250
+    },
+    {
+      "epoch": 1.9319832993065902,
+      "grad_norm": 0.35066112875938416,
+      "learning_rate": 2.94162201211404e-05,
+      "loss": 2.9657,
+      "step": 26260
+    },
+    {
+      "epoch": 1.9327190126726628,
+      "grad_norm": 0.34563148021698,
+      "learning_rate": 2.9380237141110368e-05,
+      "loss": 2.9729,
+      "step": 26270
+    },
+    {
+      "epoch": 1.9334547260387354,
+      "grad_norm": 0.3540903925895691,
+      "learning_rate": 2.9344267024047905e-05,
+      "loss": 2.9682,
+      "step": 26280
+    },
+    {
+      "epoch": 1.9341904394048077,
+      "grad_norm": 0.350699782371521,
+      "learning_rate": 2.930830979239183e-05,
+      "loss": 2.9117,
+      "step": 26290
+    },
+    {
+      "epoch": 1.9349261527708803,
+      "grad_norm": 0.3479897677898407,
+      "learning_rate": 2.927236546857286e-05,
+      "loss": 2.9613,
+      "step": 26300
+    },
+    {
+      "epoch": 1.935661866136953,
+      "grad_norm": 0.3468521535396576,
+      "learning_rate": 2.923643407501372e-05,
+      "loss": 2.9437,
+      "step": 26310
+    },
+    {
+      "epoch": 1.9363975795030255,
+      "grad_norm": 0.35198771953582764,
+      "learning_rate": 2.9200515634129e-05,
+      "loss": 2.9459,
+      "step": 26320
+    },
+    {
+      "epoch": 1.937133292869098,
+      "grad_norm": 0.35129573941230774,
+      "learning_rate": 2.916461016832529e-05,
+      "loss": 2.9946,
+      "step": 26330
+    },
+    {
+      "epoch": 1.9378690062351707,
+      "grad_norm": 0.34229612350463867,
+      "learning_rate": 2.912871770000102e-05,
+      "loss": 2.9377,
+      "step": 26340
+    },
+    {
+      "epoch": 1.9386047196012433,
+      "grad_norm": 0.39147892594337463,
+      "learning_rate": 2.9092838251546557e-05,
+      "loss": 2.985,
+      "step": 26350
+    },
+    {
+      "epoch": 1.9393404329673158,
+      "grad_norm": 0.34317928552627563,
+      "learning_rate": 2.9056971845344116e-05,
+      "loss": 2.94,
+      "step": 26360
+    },
+    {
+      "epoch": 1.9400761463333884,
+      "grad_norm": 0.33832457661628723,
+      "learning_rate": 2.9021118503767795e-05,
+      "loss": 2.933,
+      "step": 26370
+    },
+    {
+      "epoch": 1.940811859699461,
+      "grad_norm": 0.35166269540786743,
+      "learning_rate": 2.89852782491835e-05,
+      "loss": 2.9772,
+      "step": 26380
+    },
+    {
+      "epoch": 1.9415475730655336,
+      "grad_norm": 0.34960612654685974,
+      "learning_rate": 2.8949451103949056e-05,
+      "loss": 2.9699,
+      "step": 26390
+    },
+    {
+      "epoch": 1.9422832864316062,
+      "grad_norm": 0.34734126925468445,
+      "learning_rate": 2.8913637090414013e-05,
+      "loss": 2.9746,
+      "step": 26400
+    },
+    {
+      "epoch": 1.9430189997976788,
+      "grad_norm": 0.34398168325424194,
+      "learning_rate": 2.887783623091983e-05,
+      "loss": 2.9669,
+      "step": 26410
+    },
+    {
+      "epoch": 1.9437547131637514,
+      "grad_norm": 0.3545559048652649,
+      "learning_rate": 2.884204854779967e-05,
+      "loss": 2.9657,
+      "step": 26420
+    },
+    {
+      "epoch": 1.944490426529824,
+      "grad_norm": 0.35649919509887695,
+      "learning_rate": 2.8806274063378503e-05,
+      "loss": 2.9353,
+      "step": 26430
+    },
+    {
+      "epoch": 1.9452261398958965,
+      "grad_norm": 0.3456435799598694,
+      "learning_rate": 2.877051279997312e-05,
+      "loss": 2.9378,
+      "step": 26440
+    },
+    {
+      "epoch": 1.9459618532619691,
+      "grad_norm": 0.3514215052127838,
+      "learning_rate": 2.8734764779892e-05,
+      "loss": 2.9251,
+      "step": 26450
+    },
+    {
+      "epoch": 1.9466975666280417,
+      "grad_norm": 0.3462822139263153,
+      "learning_rate": 2.869903002543537e-05,
+      "loss": 2.9447,
+      "step": 26460
+    },
+    {
+      "epoch": 1.9474332799941143,
+      "grad_norm": 0.3503897190093994,
+      "learning_rate": 2.866330855889518e-05,
+      "loss": 2.9764,
+      "step": 26470
+    },
+    {
+      "epoch": 1.948168993360187,
+      "grad_norm": 0.3421068787574768,
+      "learning_rate": 2.8627600402555126e-05,
+      "loss": 2.9728,
+      "step": 26480
+    },
+    {
+      "epoch": 1.9489047067262595,
+      "grad_norm": 0.34611207246780396,
+      "learning_rate": 2.8591905578690592e-05,
+      "loss": 2.9585,
+      "step": 26490
+    },
+    {
+      "epoch": 1.949640420092332,
+      "grad_norm": 0.34353843331336975,
+      "learning_rate": 2.8556224109568623e-05,
+      "loss": 2.941,
+      "step": 26500
+    },
+    {
+      "epoch": 1.9503761334584047,
+      "grad_norm": 0.33854082226753235,
+      "learning_rate": 2.852055601744793e-05,
+      "loss": 2.9428,
+      "step": 26510
+    },
+    {
+      "epoch": 1.9511118468244772,
+      "grad_norm": 0.35856735706329346,
+      "learning_rate": 2.8484901324578883e-05,
+      "loss": 2.9436,
+      "step": 26520
+    },
+    {
+      "epoch": 1.9518475601905498,
+      "grad_norm": 0.3555583655834198,
+      "learning_rate": 2.8449260053203525e-05,
+      "loss": 2.979,
+      "step": 26530
+    },
+    {
+      "epoch": 1.9525832735566224,
+      "grad_norm": 0.3490956127643585,
+      "learning_rate": 2.841363222555549e-05,
+      "loss": 2.9564,
+      "step": 26540
+    },
+    {
+      "epoch": 1.953318986922695,
+      "grad_norm": 0.3460899293422699,
+      "learning_rate": 2.837801786386004e-05,
+      "loss": 2.9138,
+      "step": 26550
+    },
+    {
+      "epoch": 1.9540547002887676,
+      "grad_norm": 0.35729989409446716,
+      "learning_rate": 2.834241699033402e-05,
+      "loss": 2.9927,
+      "step": 26560
+    },
+    {
+      "epoch": 1.9547904136548402,
+      "grad_norm": 0.3588680028915405,
+      "learning_rate": 2.8306829627185882e-05,
+      "loss": 2.9646,
+      "step": 26570
+    },
+    {
+      "epoch": 1.9555261270209128,
+      "grad_norm": 0.3500853180885315,
+      "learning_rate": 2.8271255796615674e-05,
+      "loss": 2.9022,
+      "step": 26580
+    },
+    {
+      "epoch": 1.9562618403869854,
+      "grad_norm": 0.35143548250198364,
+      "learning_rate": 2.8235695520814953e-05,
+      "loss": 2.9453,
+      "step": 26590
+    },
+    {
+      "epoch": 1.956997553753058,
+      "grad_norm": 0.34655097126960754,
+      "learning_rate": 2.8200148821966826e-05,
+      "loss": 2.9427,
+      "step": 26600
+    },
+    {
+      "epoch": 1.9577332671191305,
+      "grad_norm": 0.34317687153816223,
+      "learning_rate": 2.8164615722245935e-05,
+      "loss": 2.9893,
+      "step": 26610
+    },
+    {
+      "epoch": 1.9584689804852031,
+      "grad_norm": 0.3458210825920105,
+      "learning_rate": 2.8129096243818485e-05,
+      "loss": 2.9446,
+      "step": 26620
+    },
+    {
+      "epoch": 1.9592046938512755,
+      "grad_norm": 0.34827637672424316,
+      "learning_rate": 2.809359040884213e-05,
+      "loss": 2.9388,
+      "step": 26630
+    },
+    {
+      "epoch": 1.959940407217348,
+      "grad_norm": 0.3444003462791443,
+      "learning_rate": 2.805809823946602e-05,
+      "loss": 2.9515,
+      "step": 26640
+    },
+    {
+      "epoch": 1.9606761205834207,
+      "grad_norm": 0.3513241708278656,
+      "learning_rate": 2.802261975783077e-05,
+      "loss": 2.9609,
+      "step": 26650
+    },
+    {
+      "epoch": 1.9614118339494933,
+      "grad_norm": 0.3505419194698334,
+      "learning_rate": 2.79871549860685e-05,
+      "loss": 2.9356,
+      "step": 26660
+    },
+    {
+      "epoch": 1.9621475473155658,
+      "grad_norm": 0.3449759781360626,
+      "learning_rate": 2.7951703946302776e-05,
+      "loss": 2.9057,
+      "step": 26670
+    },
+    {
+      "epoch": 1.9628832606816384,
+      "grad_norm": 0.3528783321380615,
+      "learning_rate": 2.7916266660648547e-05,
+      "loss": 2.9371,
+      "step": 26680
+    },
+    {
+      "epoch": 1.963618974047711,
+      "grad_norm": 0.3496946096420288,
+      "learning_rate": 2.788084315121221e-05,
+      "loss": 2.9835,
+      "step": 26690
+    },
+    {
+      "epoch": 1.9643546874137836,
+      "grad_norm": 0.3440243899822235,
+      "learning_rate": 2.7845433440091556e-05,
+      "loss": 2.9374,
+      "step": 26700
+    },
+    {
+      "epoch": 1.9650904007798562,
+      "grad_norm": 0.3409256339073181,
+      "learning_rate": 2.7810037549375813e-05,
+      "loss": 2.9201,
+      "step": 26710
+    },
+    {
+      "epoch": 1.9658261141459288,
+      "grad_norm": 0.34928569197654724,
+      "learning_rate": 2.7774655501145543e-05,
+      "loss": 2.9121,
+      "step": 26720
+    },
+    {
+      "epoch": 1.9665618275120014,
+      "grad_norm": 0.34791991114616394,
+      "learning_rate": 2.773928731747265e-05,
+      "loss": 2.9439,
+      "step": 26730
+    },
+    {
+      "epoch": 1.9672975408780737,
+      "grad_norm": 0.34656745195388794,
+      "learning_rate": 2.7703933020420454e-05,
+      "loss": 2.9596,
+      "step": 26740
+    },
+    {
+      "epoch": 1.9680332542441463,
+      "grad_norm": 0.35247519612312317,
+      "learning_rate": 2.76685926320436e-05,
+      "loss": 2.9524,
+      "step": 26750
+    },
+    {
+      "epoch": 1.968768967610219,
+      "grad_norm": 0.34850290417671204,
+      "learning_rate": 2.7633266174388016e-05,
+      "loss": 2.9864,
+      "step": 26760
+    },
+    {
+      "epoch": 1.9695046809762915,
+      "grad_norm": 0.341669499874115,
+      "learning_rate": 2.7597953669490966e-05,
+      "loss": 2.96,
+      "step": 26770
+    },
+    {
+      "epoch": 1.970240394342364,
+      "grad_norm": 0.3405725955963135,
+      "learning_rate": 2.756265513938099e-05,
+      "loss": 2.9671,
+      "step": 26780
+    },
+    {
+      "epoch": 1.9709761077084367,
+      "grad_norm": 0.3330259621143341,
+      "learning_rate": 2.752737060607795e-05,
+      "loss": 2.9776,
+      "step": 26790
+    },
+    {
+      "epoch": 1.9717118210745093,
+      "grad_norm": 0.3644200265407562,
+      "learning_rate": 2.7492100091592953e-05,
+      "loss": 2.9741,
+      "step": 26800
+    },
+    {
+      "epoch": 1.9724475344405819,
+      "grad_norm": 0.34592297673225403,
+      "learning_rate": 2.7456843617928335e-05,
+      "loss": 2.9605,
+      "step": 26810
+    },
+    {
+      "epoch": 1.9731832478066544,
+      "grad_norm": 0.34836962819099426,
+      "learning_rate": 2.742160120707773e-05,
+      "loss": 2.9454,
+      "step": 26820
+    },
+    {
+      "epoch": 1.973918961172727,
+      "grad_norm": 0.34214553236961365,
+      "learning_rate": 2.7386372881025934e-05,
+      "loss": 2.925,
+      "step": 26830
+    },
+    {
+      "epoch": 1.9746546745387996,
+      "grad_norm": 0.3517175316810608,
+      "learning_rate": 2.735115866174904e-05,
+      "loss": 2.9783,
+      "step": 26840
+    },
+    {
+      "epoch": 1.9753903879048722,
+      "grad_norm": 0.35157647728919983,
+      "learning_rate": 2.7315958571214255e-05,
+      "loss": 2.9268,
+      "step": 26850
+    },
+    {
+      "epoch": 1.9761261012709448,
+      "grad_norm": 0.35020992159843445,
+      "learning_rate": 2.7280772631380024e-05,
+      "loss": 2.9602,
+      "step": 26860
+    },
+    {
+      "epoch": 1.9768618146370174,
+      "grad_norm": 0.3508092761039734,
+      "learning_rate": 2.724560086419592e-05,
+      "loss": 2.9968,
+      "step": 26870
+    },
+    {
+      "epoch": 1.97759752800309,
+      "grad_norm": 0.35094162821769714,
+      "learning_rate": 2.7210443291602744e-05,
+      "loss": 2.9537,
+      "step": 26880
+    },
+    {
+      "epoch": 1.9783332413691626,
+      "grad_norm": 0.3514111638069153,
+      "learning_rate": 2.7175299935532374e-05,
+      "loss": 2.978,
+      "step": 26890
+    },
+    {
+      "epoch": 1.9790689547352351,
+      "grad_norm": 0.35783684253692627,
+      "learning_rate": 2.714017081790787e-05,
+      "loss": 2.9322,
+      "step": 26900
+    },
+    {
+      "epoch": 1.9798046681013077,
+      "grad_norm": 0.35521531105041504,
+      "learning_rate": 2.7105055960643395e-05,
+      "loss": 2.9036,
+      "step": 26910
+    },
+    {
+      "epoch": 1.9805403814673803,
+      "grad_norm": 0.3491211533546448,
+      "learning_rate": 2.7069955385644164e-05,
+      "loss": 2.9187,
+      "step": 26920
+    },
+    {
+      "epoch": 1.981276094833453,
+      "grad_norm": 0.3496791422367096,
+      "learning_rate": 2.703486911480658e-05,
+      "loss": 2.936,
+      "step": 26930
+    },
+    {
+      "epoch": 1.9820118081995255,
+      "grad_norm": 0.3428598642349243,
+      "learning_rate": 2.699979717001805e-05,
+      "loss": 2.9653,
+      "step": 26940
+    },
+    {
+      "epoch": 1.982747521565598,
+      "grad_norm": 0.3508605659008026,
+      "learning_rate": 2.696473957315705e-05,
+      "loss": 2.9738,
+      "step": 26950
+    },
+    {
+      "epoch": 1.9834832349316707,
+      "grad_norm": 0.3544585108757019,
+      "learning_rate": 2.6929696346093125e-05,
+      "loss": 2.9301,
+      "step": 26960
+    },
+    {
+      "epoch": 1.9842189482977433,
+      "grad_norm": 0.35597464442253113,
+      "learning_rate": 2.689466751068688e-05,
+      "loss": 2.9489,
+      "step": 26970
+    },
+    {
+      "epoch": 1.9849546616638158,
+      "grad_norm": 0.3475305438041687,
+      "learning_rate": 2.685965308878986e-05,
+      "loss": 2.9151,
+      "step": 26980
+    },
+    {
+      "epoch": 1.9856903750298884,
+      "grad_norm": 0.3658480942249298,
+      "learning_rate": 2.682465310224473e-05,
+      "loss": 2.956,
+      "step": 26990
+    },
+    {
+      "epoch": 1.986426088395961,
+      "grad_norm": 0.35342249274253845,
+      "learning_rate": 2.6789667572885067e-05,
+      "loss": 2.9577,
+      "step": 27000
+    },
+    {
+      "epoch": 1.9871618017620336,
+      "grad_norm": 0.34583503007888794,
+      "learning_rate": 2.6754696522535427e-05,
+      "loss": 2.9549,
+      "step": 27010
+    },
+    {
+      "epoch": 1.9878975151281062,
+      "grad_norm": 0.35147446393966675,
+      "learning_rate": 2.6719739973011417e-05,
+      "loss": 2.9778,
+      "step": 27020
+    },
+    {
+      "epoch": 1.9886332284941788,
+      "grad_norm": 0.3457833528518677,
+      "learning_rate": 2.66847979461195e-05,
+      "loss": 2.9259,
+      "step": 27030
+    },
+    {
+      "epoch": 1.9893689418602514,
+      "grad_norm": 0.3490820527076721,
+      "learning_rate": 2.6649870463657133e-05,
+      "loss": 2.9383,
+      "step": 27040
+    },
+    {
+      "epoch": 1.990104655226324,
+      "grad_norm": 0.3371882736682892,
+      "learning_rate": 2.6614957547412678e-05,
+      "loss": 2.9466,
+      "step": 27050
+    },
+    {
+      "epoch": 1.9908403685923965,
+      "grad_norm": 0.3562604784965515,
+      "learning_rate": 2.6580059219165426e-05,
+      "loss": 2.9688,
+      "step": 27060
+    },
+    {
+      "epoch": 1.9915760819584691,
+      "grad_norm": 0.3476478159427643,
+      "learning_rate": 2.6545175500685593e-05,
+      "loss": 2.9628,
+      "step": 27070
+    },
+    {
+      "epoch": 1.9923117953245417,
+      "grad_norm": 0.3531840145587921,
+      "learning_rate": 2.651030641373422e-05,
+      "loss": 2.9786,
+      "step": 27080
+    },
+    {
+      "epoch": 1.993047508690614,
+      "grad_norm": 0.3469148874282837,
+      "learning_rate": 2.647545198006326e-05,
+      "loss": 2.9257,
+      "step": 27090
+    },
+    {
+      "epoch": 1.9937832220566867,
+      "grad_norm": 0.3433736562728882,
+      "learning_rate": 2.6440612221415495e-05,
+      "loss": 2.9455,
+      "step": 27100
+    },
+    {
+      "epoch": 1.9945189354227593,
+      "grad_norm": 0.3566463589668274,
+      "learning_rate": 2.640578715952461e-05,
+      "loss": 2.9619,
+      "step": 27110
+    },
+    {
+      "epoch": 1.9952546487888319,
+      "grad_norm": 0.35686400532722473,
+      "learning_rate": 2.637097681611507e-05,
+      "loss": 2.9644,
+      "step": 27120
+    },
+    {
+      "epoch": 1.9959903621549044,
+      "grad_norm": 0.34093886613845825,
+      "learning_rate": 2.6336181212902173e-05,
+      "loss": 2.9464,
+      "step": 27130
+    },
+    {
+      "epoch": 1.996726075520977,
+      "grad_norm": 0.35102489590644836,
+      "learning_rate": 2.6301400371592e-05,
+      "loss": 2.9429,
+      "step": 27140
+    },
+    {
+      "epoch": 1.9974617888870496,
+      "grad_norm": 0.34986865520477295,
+      "learning_rate": 2.626663431388148e-05,
+      "loss": 2.9041,
+      "step": 27150
+    },
+    {
+      "epoch": 1.9981975022531222,
+      "grad_norm": 0.3497006297111511,
+      "learning_rate": 2.6231883061458284e-05,
+      "loss": 2.9682,
+      "step": 27160
+    },
+    {
+      "epoch": 1.9989332156191948,
+      "grad_norm": 0.3355732560157776,
+      "learning_rate": 2.6197146636000834e-05,
+      "loss": 2.9951,
+      "step": 27170
+    },
+    {
+      "epoch": 1.9996689289852674,
+      "grad_norm": 0.3498176336288452,
+      "learning_rate": 2.6162425059178337e-05,
+      "loss": 2.9575,
+      "step": 27180
+    },
+    {
+      "epoch": 2.0004046423513397,
+      "grad_norm": 0.3554339110851288,
+      "learning_rate": 2.6127718352650675e-05,
+      "loss": 2.9475,
+      "step": 27190
+    },
+    {
+      "epoch": 2.0011403557174123,
+      "grad_norm": 0.34532326459884644,
+      "learning_rate": 2.6093026538068556e-05,
+      "loss": 2.8679,
+      "step": 27200
+    },
+    {
+      "epoch": 2.001876069083485,
+      "grad_norm": 0.3400401771068573,
+      "learning_rate": 2.605834963707331e-05,
+      "loss": 2.9135,
+      "step": 27210
+    },
+    {
+      "epoch": 2.0026117824495575,
+      "grad_norm": 0.3416743874549866,
+      "learning_rate": 2.602368767129696e-05,
+      "loss": 2.9346,
+      "step": 27220
+    },
+    {
+      "epoch": 2.00334749581563,
+      "grad_norm": 0.36012691259384155,
+      "learning_rate": 2.5989040662362307e-05,
+      "loss": 2.8684,
+      "step": 27230
+    },
+    {
+      "epoch": 2.0040832091817027,
+      "grad_norm": 0.35780975222587585,
+      "learning_rate": 2.5954408631882698e-05,
+      "loss": 2.8962,
+      "step": 27240
+    },
+    {
+      "epoch": 2.0048189225477753,
+      "grad_norm": 0.3588829040527344,
+      "learning_rate": 2.591979160146226e-05,
+      "loss": 2.8424,
+      "step": 27250
+    },
+    {
+      "epoch": 2.005554635913848,
+      "grad_norm": 0.3540956377983093,
+      "learning_rate": 2.588518959269566e-05,
+      "loss": 2.8188,
+      "step": 27260
+    },
+    {
+      "epoch": 2.0062903492799204,
+      "grad_norm": 0.34116867184638977,
+      "learning_rate": 2.5850602627168248e-05,
+      "loss": 2.8811,
+      "step": 27270
+    },
+    {
+      "epoch": 2.007026062645993,
+      "grad_norm": 0.35695895552635193,
+      "learning_rate": 2.5816030726455947e-05,
+      "loss": 2.8536,
+      "step": 27280
+    },
+    {
+      "epoch": 2.0077617760120656,
+      "grad_norm": 0.3584601581096649,
+      "learning_rate": 2.5781473912125352e-05,
+      "loss": 2.8702,
+      "step": 27290
+    },
+    {
+      "epoch": 2.008497489378138,
+      "grad_norm": 0.36200013756752014,
+      "learning_rate": 2.57469322057336e-05,
+      "loss": 2.8855,
+      "step": 27300
+    },
+    {
+      "epoch": 2.009233202744211,
+      "grad_norm": 0.34632357954978943,
+      "learning_rate": 2.5712405628828375e-05,
+      "loss": 2.8412,
+      "step": 27310
+    },
+    {
+      "epoch": 2.0099689161102834,
+      "grad_norm": 0.3517228960990906,
+      "learning_rate": 2.5677894202947994e-05,
+      "loss": 2.8535,
+      "step": 27320
+    },
+    {
+      "epoch": 2.010704629476356,
+      "grad_norm": 0.3674863874912262,
+      "learning_rate": 2.5643397949621262e-05,
+      "loss": 2.8124,
+      "step": 27330
+    },
+    {
+      "epoch": 2.0114403428424286,
+      "grad_norm": 0.3513818383216858,
+      "learning_rate": 2.5608916890367568e-05,
+      "loss": 2.8612,
+      "step": 27340
+    },
+    {
+      "epoch": 2.012176056208501,
+      "grad_norm": 0.36085397005081177,
+      "learning_rate": 2.557445104669679e-05,
+      "loss": 2.8674,
+      "step": 27350
+    },
+    {
+      "epoch": 2.0129117695745737,
+      "grad_norm": 0.34026235342025757,
+      "learning_rate": 2.5540000440109313e-05,
+      "loss": 2.8493,
+      "step": 27360
+    },
+    {
+      "epoch": 2.0136474829406463,
+      "grad_norm": 0.3575206696987152,
+      "learning_rate": 2.5505565092096018e-05,
+      "loss": 2.8584,
+      "step": 27370
+    },
+    {
+      "epoch": 2.014383196306719,
+      "grad_norm": 0.35253220796585083,
+      "learning_rate": 2.5471145024138305e-05,
+      "loss": 2.8685,
+      "step": 27380
+    },
+    {
+      "epoch": 2.0151189096727915,
+      "grad_norm": 0.3441455364227295,
+      "learning_rate": 2.5436740257707974e-05,
+      "loss": 2.8563,
+      "step": 27390
+    },
+    {
+      "epoch": 2.015854623038864,
+      "grad_norm": 0.34542691707611084,
+      "learning_rate": 2.540235081426736e-05,
+      "loss": 2.9023,
+      "step": 27400
+    },
+    {
+      "epoch": 2.0165903364049367,
+      "grad_norm": 0.3514902889728546,
+      "learning_rate": 2.5367976715269176e-05,
+      "loss": 2.88,
+      "step": 27410
+    },
+    {
+      "epoch": 2.0173260497710093,
+      "grad_norm": 0.35899069905281067,
+      "learning_rate": 2.5333617982156567e-05,
+      "loss": 2.871,
+      "step": 27420
+    },
+    {
+      "epoch": 2.018061763137082,
+      "grad_norm": 0.3535335958003998,
+      "learning_rate": 2.5299274636363147e-05,
+      "loss": 2.8476,
+      "step": 27430
+    },
+    {
+      "epoch": 2.0187974765031544,
+      "grad_norm": 0.36238640546798706,
+      "learning_rate": 2.5264946699312873e-05,
+      "loss": 2.8601,
+      "step": 27440
+    },
+    {
+      "epoch": 2.019533189869227,
+      "grad_norm": 0.35086575150489807,
+      "learning_rate": 2.523063419242012e-05,
+      "loss": 2.7937,
+      "step": 27450
+    },
+    {
+      "epoch": 2.0202689032352996,
+      "grad_norm": 0.3572332561016083,
+      "learning_rate": 2.5196337137089598e-05,
+      "loss": 2.8727,
+      "step": 27460
+    },
+    {
+      "epoch": 2.021004616601372,
+      "grad_norm": 0.3625427484512329,
+      "learning_rate": 2.5162055554716423e-05,
+      "loss": 2.8354,
+      "step": 27470
+    },
+    {
+      "epoch": 2.021740329967445,
+      "grad_norm": 0.35718798637390137,
+      "learning_rate": 2.512778946668608e-05,
+      "loss": 2.8122,
+      "step": 27480
+    },
+    {
+      "epoch": 2.0224760433335174,
+      "grad_norm": 0.355476975440979,
+      "learning_rate": 2.509353889437432e-05,
+      "loss": 2.839,
+      "step": 27490
+    },
+    {
+      "epoch": 2.02321175669959,
+      "grad_norm": 0.35139593482017517,
+      "learning_rate": 2.5059303859147254e-05,
+      "loss": 2.892,
+      "step": 27500
+    },
+    {
+      "epoch": 2.0239474700656626,
+      "grad_norm": 0.35471948981285095,
+      "learning_rate": 2.5025084382361262e-05,
+      "loss": 2.9143,
+      "step": 27510
+    },
+    {
+      "epoch": 2.024683183431735,
+      "grad_norm": 0.35509562492370605,
+      "learning_rate": 2.499088048536311e-05,
+      "loss": 2.901,
+      "step": 27520
+    },
+    {
+      "epoch": 2.0254188967978077,
+      "grad_norm": 0.3525761365890503,
+      "learning_rate": 2.495669218948975e-05,
+      "loss": 2.8044,
+      "step": 27530
+    },
+    {
+      "epoch": 2.0261546101638803,
+      "grad_norm": 0.35157155990600586,
+      "learning_rate": 2.492251951606844e-05,
+      "loss": 2.9241,
+      "step": 27540
+    },
+    {
+      "epoch": 2.026890323529953,
+      "grad_norm": 0.3519863784313202,
+      "learning_rate": 2.4888362486416673e-05,
+      "loss": 2.8419,
+      "step": 27550
+    },
+    {
+      "epoch": 2.0276260368960255,
+      "grad_norm": 0.3537205755710602,
+      "learning_rate": 2.4854221121842214e-05,
+      "loss": 2.9088,
+      "step": 27560
+    },
+    {
+      "epoch": 2.028361750262098,
+      "grad_norm": 0.35056138038635254,
+      "learning_rate": 2.4820095443643067e-05,
+      "loss": 2.8446,
+      "step": 27570
+    },
+    {
+      "epoch": 2.0290974636281707,
+      "grad_norm": 0.3539247512817383,
+      "learning_rate": 2.4785985473107405e-05,
+      "loss": 2.8433,
+      "step": 27580
+    },
+    {
+      "epoch": 2.0298331769942433,
+      "grad_norm": 0.34905552864074707,
+      "learning_rate": 2.4751891231513618e-05,
+      "loss": 2.8733,
+      "step": 27590
+    },
+    {
+      "epoch": 2.0305688903603154,
+      "grad_norm": 0.35301047563552856,
+      "learning_rate": 2.471781274013027e-05,
+      "loss": 2.8688,
+      "step": 27600
+    },
+    {
+      "epoch": 2.031304603726388,
+      "grad_norm": 0.3552880883216858,
+      "learning_rate": 2.468375002021616e-05,
+      "loss": 2.8803,
+      "step": 27610
+    },
+    {
+      "epoch": 2.0320403170924606,
+      "grad_norm": 0.35055699944496155,
+      "learning_rate": 2.4649703093020186e-05,
+      "loss": 2.8629,
+      "step": 27620
+    },
+    {
+      "epoch": 2.032776030458533,
+      "grad_norm": 0.347362756729126,
+      "learning_rate": 2.461567197978139e-05,
+      "loss": 2.8068,
+      "step": 27630
+    },
+    {
+      "epoch": 2.0335117438246058,
+      "grad_norm": 0.36117297410964966,
+      "learning_rate": 2.4581656701729004e-05,
+      "loss": 2.8534,
+      "step": 27640
+    },
+    {
+      "epoch": 2.0342474571906783,
+      "grad_norm": 0.3489930033683777,
+      "learning_rate": 2.4547657280082365e-05,
+      "loss": 2.8101,
+      "step": 27650
+    },
+    {
+      "epoch": 2.034983170556751,
+      "grad_norm": 0.35757535696029663,
+      "learning_rate": 2.4513673736050884e-05,
+      "loss": 2.8871,
+      "step": 27660
+    },
+    {
+      "epoch": 2.0357188839228235,
+      "grad_norm": 0.3463372588157654,
+      "learning_rate": 2.4479706090834087e-05,
+      "loss": 2.8805,
+      "step": 27670
+    },
+    {
+      "epoch": 2.036454597288896,
+      "grad_norm": 0.36374610662460327,
+      "learning_rate": 2.444575436562157e-05,
+      "loss": 2.907,
+      "step": 27680
+    },
+    {
+      "epoch": 2.0371903106549687,
+      "grad_norm": 0.35738256573677063,
+      "learning_rate": 2.4411818581593044e-05,
+      "loss": 2.866,
+      "step": 27690
+    },
+    {
+      "epoch": 2.0379260240210413,
+      "grad_norm": 0.3600810468196869,
+      "learning_rate": 2.4377898759918228e-05,
+      "loss": 2.8688,
+      "step": 27700
+    },
+    {
+      "epoch": 2.038661737387114,
+      "grad_norm": 0.3569597005844116,
+      "learning_rate": 2.4343994921756896e-05,
+      "loss": 2.8783,
+      "step": 27710
+    },
+    {
+      "epoch": 2.0393974507531865,
+      "grad_norm": 0.3568631708621979,
+      "learning_rate": 2.4310107088258833e-05,
+      "loss": 2.8351,
+      "step": 27720
+    },
+    {
+      "epoch": 2.040133164119259,
+      "grad_norm": 0.34865808486938477,
+      "learning_rate": 2.4276235280563887e-05,
+      "loss": 2.8444,
+      "step": 27730
+    },
+    {
+      "epoch": 2.0408688774853316,
+      "grad_norm": 0.3385525643825531,
+      "learning_rate": 2.424237951980189e-05,
+      "loss": 2.8406,
+      "step": 27740
+    },
+    {
+      "epoch": 2.041604590851404,
+      "grad_norm": 0.35342299938201904,
+      "learning_rate": 2.420853982709264e-05,
+      "loss": 2.8263,
+      "step": 27750
+    },
+    {
+      "epoch": 2.042340304217477,
+      "grad_norm": 0.35292714834213257,
+      "learning_rate": 2.4174716223545928e-05,
+      "loss": 2.8756,
+      "step": 27760
+    },
+    {
+      "epoch": 2.0430760175835494,
+      "grad_norm": 0.34293684363365173,
+      "learning_rate": 2.4140908730261497e-05,
+      "loss": 2.8816,
+      "step": 27770
+    },
+    {
+      "epoch": 2.043811730949622,
+      "grad_norm": 0.3600517511367798,
+      "learning_rate": 2.410711736832908e-05,
+      "loss": 2.8382,
+      "step": 27780
+    },
+    {
+      "epoch": 2.0445474443156946,
+      "grad_norm": 0.3508221209049225,
+      "learning_rate": 2.4073342158828293e-05,
+      "loss": 2.8192,
+      "step": 27790
+    },
+    {
+      "epoch": 2.045283157681767,
+      "grad_norm": 0.35835230350494385,
+      "learning_rate": 2.4039583122828694e-05,
+      "loss": 2.8693,
+      "step": 27800
+    },
+    {
+      "epoch": 2.0460188710478397,
+      "grad_norm": 0.3622501492500305,
+      "learning_rate": 2.4005840281389796e-05,
+      "loss": 2.8721,
+      "step": 27810
+    },
+    {
+      "epoch": 2.0467545844139123,
+      "grad_norm": 0.35346221923828125,
+      "learning_rate": 2.3972113655560925e-05,
+      "loss": 2.876,
+      "step": 27820
+    },
+    {
+      "epoch": 2.047490297779985,
+      "grad_norm": 0.3535018861293793,
+      "learning_rate": 2.393840326638138e-05,
+      "loss": 2.8513,
+      "step": 27830
+    },
+    {
+      "epoch": 2.0482260111460575,
+      "grad_norm": 0.3495038151741028,
+      "learning_rate": 2.3904709134880273e-05,
+      "loss": 2.8968,
+      "step": 27840
+    },
+    {
+      "epoch": 2.04896172451213,
+      "grad_norm": 0.35158953070640564,
+      "learning_rate": 2.3871031282076585e-05,
+      "loss": 2.8576,
+      "step": 27850
+    },
+    {
+      "epoch": 2.0496974378782027,
+      "grad_norm": 0.36737874150276184,
+      "learning_rate": 2.3837369728979138e-05,
+      "loss": 2.8663,
+      "step": 27860
+    },
+    {
+      "epoch": 2.0504331512442753,
+      "grad_norm": 0.34086719155311584,
+      "learning_rate": 2.380372449658662e-05,
+      "loss": 2.8367,
+      "step": 27870
+    },
+    {
+      "epoch": 2.051168864610348,
+      "grad_norm": 0.34565573930740356,
+      "learning_rate": 2.377009560588749e-05,
+      "loss": 2.8575,
+      "step": 27880
+    },
+    {
+      "epoch": 2.0519045779764205,
+      "grad_norm": 0.34978368878364563,
+      "learning_rate": 2.373648307786006e-05,
+      "loss": 2.8737,
+      "step": 27890
+    },
+    {
+      "epoch": 2.052640291342493,
+      "grad_norm": 0.3677339255809784,
+      "learning_rate": 2.3702886933472407e-05,
+      "loss": 2.8886,
+      "step": 27900
+    },
+    {
+      "epoch": 2.0533760047085656,
+      "grad_norm": 0.34705767035484314,
+      "learning_rate": 2.3669307193682367e-05,
+      "loss": 2.8577,
+      "step": 27910
+    },
+    {
+      "epoch": 2.054111718074638,
+      "grad_norm": 0.3651942312717438,
+      "learning_rate": 2.36357438794376e-05,
+      "loss": 2.8958,
+      "step": 27920
+    },
+    {
+      "epoch": 2.054847431440711,
+      "grad_norm": 0.3560611307621002,
+      "learning_rate": 2.360219701167547e-05,
+      "loss": 2.8378,
+      "step": 27930
+    },
+    {
+      "epoch": 2.0555831448067834,
+      "grad_norm": 0.35784897208213806,
+      "learning_rate": 2.3568666611323103e-05,
+      "loss": 2.8671,
+      "step": 27940
+    },
+    {
+      "epoch": 2.056318858172856,
+      "grad_norm": 0.3546694219112396,
+      "learning_rate": 2.3535152699297326e-05,
+      "loss": 2.8417,
+      "step": 27950
+    },
+    {
+      "epoch": 2.0570545715389286,
+      "grad_norm": 0.3601793944835663,
+      "learning_rate": 2.3501655296504716e-05,
+      "loss": 2.8723,
+      "step": 27960
+    },
+    {
+      "epoch": 2.057790284905001,
+      "grad_norm": 0.351997435092926,
+      "learning_rate": 2.346817442384156e-05,
+      "loss": 2.8326,
+      "step": 27970
+    },
+    {
+      "epoch": 2.0585259982710737,
+      "grad_norm": 0.3527384400367737,
+      "learning_rate": 2.3434710102193785e-05,
+      "loss": 2.9111,
+      "step": 27980
+    },
+    {
+      "epoch": 2.0592617116371463,
+      "grad_norm": 0.35195666551589966,
+      "learning_rate": 2.340126235243703e-05,
+      "loss": 2.8473,
+      "step": 27990
+    },
+    {
+      "epoch": 2.059997425003219,
+      "grad_norm": 0.35623466968536377,
+      "learning_rate": 2.3367831195436564e-05,
+      "loss": 2.8449,
+      "step": 28000
+    },
+    {
+      "epoch": 2.0607331383692915,
+      "grad_norm": 0.36602258682250977,
+      "learning_rate": 2.3334416652047347e-05,
+      "loss": 2.9013,
+      "step": 28010
+    },
+    {
+      "epoch": 2.061468851735364,
+      "grad_norm": 0.3586275279521942,
+      "learning_rate": 2.3301018743113957e-05,
+      "loss": 2.8728,
+      "step": 28020
+    },
+    {
+      "epoch": 2.0622045651014367,
+      "grad_norm": 0.35837501287460327,
+      "learning_rate": 2.3267637489470574e-05,
+      "loss": 2.8375,
+      "step": 28030
+    },
+    {
+      "epoch": 2.0629402784675093,
+      "grad_norm": 0.3484499752521515,
+      "learning_rate": 2.3234272911940996e-05,
+      "loss": 2.89,
+      "step": 28040
+    },
+    {
+      "epoch": 2.0636759918335814,
+      "grad_norm": 0.36009708046913147,
+      "learning_rate": 2.320092503133865e-05,
+      "loss": 2.9036,
+      "step": 28050
+    },
+    {
+      "epoch": 2.064411705199654,
+      "grad_norm": 0.3654894530773163,
+      "learning_rate": 2.3167593868466536e-05,
+      "loss": 2.8679,
+      "step": 28060
+    },
+    {
+      "epoch": 2.0651474185657266,
+      "grad_norm": 0.3597780466079712,
+      "learning_rate": 2.31342794441172e-05,
+      "loss": 2.8705,
+      "step": 28070
+    },
+    {
+      "epoch": 2.065883131931799,
+      "grad_norm": 0.3477473258972168,
+      "learning_rate": 2.3100981779072762e-05,
+      "loss": 2.8709,
+      "step": 28080
+    },
+    {
+      "epoch": 2.0666188452978718,
+      "grad_norm": 0.3474847376346588,
+      "learning_rate": 2.3067700894104865e-05,
+      "loss": 2.8774,
+      "step": 28090
+    },
+    {
+      "epoch": 2.0673545586639444,
+      "grad_norm": 0.35553866624832153,
+      "learning_rate": 2.3034436809974747e-05,
+      "loss": 2.8536,
+      "step": 28100
+    },
+    {
+      "epoch": 2.068090272030017,
+      "grad_norm": 0.3494465947151184,
+      "learning_rate": 2.3001189547433093e-05,
+      "loss": 2.9079,
+      "step": 28110
+    },
+    {
+      "epoch": 2.0688259853960895,
+      "grad_norm": 0.3475459814071655,
+      "learning_rate": 2.296795912722014e-05,
+      "loss": 2.8387,
+      "step": 28120
+    },
+    {
+      "epoch": 2.069561698762162,
+      "grad_norm": 0.3552308976650238,
+      "learning_rate": 2.2934745570065573e-05,
+      "loss": 2.864,
+      "step": 28130
+    },
+    {
+      "epoch": 2.0702974121282347,
+      "grad_norm": 0.3538024127483368,
+      "learning_rate": 2.290154889668862e-05,
+      "loss": 2.8584,
+      "step": 28140
+    },
+    {
+      "epoch": 2.0710331254943073,
+      "grad_norm": 0.3731195032596588,
+      "learning_rate": 2.2868369127797944e-05,
+      "loss": 2.8601,
+      "step": 28150
+    },
+    {
+      "epoch": 2.07176883886038,
+      "grad_norm": 0.3621228039264679,
+      "learning_rate": 2.2835206284091663e-05,
+      "loss": 2.8642,
+      "step": 28160
+    },
+    {
+      "epoch": 2.0725045522264525,
+      "grad_norm": 0.3516838252544403,
+      "learning_rate": 2.2802060386257328e-05,
+      "loss": 2.8355,
+      "step": 28170
+    },
+    {
+      "epoch": 2.073240265592525,
+      "grad_norm": 0.34510084986686707,
+      "learning_rate": 2.2768931454971916e-05,
+      "loss": 2.8263,
+      "step": 28180
+    },
+    {
+      "epoch": 2.0739759789585976,
+      "grad_norm": 0.3533056676387787,
+      "learning_rate": 2.2735819510901863e-05,
+      "loss": 2.8664,
+      "step": 28190
+    },
+    {
+      "epoch": 2.0747116923246702,
+      "grad_norm": 0.35418930649757385,
+      "learning_rate": 2.2702724574702955e-05,
+      "loss": 2.886,
+      "step": 28200
+    },
+    {
+      "epoch": 2.075447405690743,
+      "grad_norm": 0.34745532274246216,
+      "learning_rate": 2.2669646667020377e-05,
+      "loss": 2.9093,
+      "step": 28210
+    },
+    {
+      "epoch": 2.0761831190568154,
+      "grad_norm": 0.34773001074790955,
+      "learning_rate": 2.263658580848875e-05,
+      "loss": 2.8508,
+      "step": 28220
+    },
+    {
+      "epoch": 2.076918832422888,
+      "grad_norm": 0.353232741355896,
+      "learning_rate": 2.260354201973196e-05,
+      "loss": 2.82,
+      "step": 28230
+    },
+    {
+      "epoch": 2.0776545457889606,
+      "grad_norm": 0.35694295167922974,
+      "learning_rate": 2.2570515321363346e-05,
+      "loss": 2.8076,
+      "step": 28240
+    },
+    {
+      "epoch": 2.078390259155033,
+      "grad_norm": 0.3461124897003174,
+      "learning_rate": 2.2537505733985514e-05,
+      "loss": 2.8789,
+      "step": 28250
+    },
+    {
+      "epoch": 2.0791259725211058,
+      "grad_norm": 0.3615930676460266,
+      "learning_rate": 2.250451327819043e-05,
+      "loss": 2.8912,
+      "step": 28260
+    },
+    {
+      "epoch": 2.0798616858871783,
+      "grad_norm": 0.36153489351272583,
+      "learning_rate": 2.2471537974559335e-05,
+      "loss": 2.8984,
+      "step": 28270
+    },
+    {
+      "epoch": 2.080597399253251,
+      "grad_norm": 0.3563705086708069,
+      "learning_rate": 2.243857984366284e-05,
+      "loss": 2.8808,
+      "step": 28280
+    },
+    {
+      "epoch": 2.0813331126193235,
+      "grad_norm": 0.3578133285045624,
+      "learning_rate": 2.240563890606076e-05,
+      "loss": 2.8639,
+      "step": 28290
+    },
+    {
+      "epoch": 2.082068825985396,
+      "grad_norm": 0.35942497849464417,
+      "learning_rate": 2.2372715182302267e-05,
+      "loss": 2.8481,
+      "step": 28300
+    },
+    {
+      "epoch": 2.0828045393514687,
+      "grad_norm": 0.3687622845172882,
+      "learning_rate": 2.233980869292574e-05,
+      "loss": 2.9058,
+      "step": 28310
+    },
+    {
+      "epoch": 2.0835402527175413,
+      "grad_norm": 0.34475964307785034,
+      "learning_rate": 2.2306919458458792e-05,
+      "loss": 2.8216,
+      "step": 28320
+    },
+    {
+      "epoch": 2.084275966083614,
+      "grad_norm": 0.35632210969924927,
+      "learning_rate": 2.227404749941835e-05,
+      "loss": 2.8501,
+      "step": 28330
+    },
+    {
+      "epoch": 2.0850116794496865,
+      "grad_norm": 0.35065969824790955,
+      "learning_rate": 2.224119283631048e-05,
+      "loss": 2.8638,
+      "step": 28340
+    },
+    {
+      "epoch": 2.085747392815759,
+      "grad_norm": 0.36123669147491455,
+      "learning_rate": 2.220835548963051e-05,
+      "loss": 2.905,
+      "step": 28350
+    },
+    {
+      "epoch": 2.0864831061818316,
+      "grad_norm": 0.35647714138031006,
+      "learning_rate": 2.2175535479862925e-05,
+      "loss": 2.8801,
+      "step": 28360
+    },
+    {
+      "epoch": 2.0872188195479042,
+      "grad_norm": 0.3591693639755249,
+      "learning_rate": 2.2142732827481456e-05,
+      "loss": 2.83,
+      "step": 28370
+    },
+    {
+      "epoch": 2.087954532913977,
+      "grad_norm": 0.35314521193504333,
+      "learning_rate": 2.2109947552948927e-05,
+      "loss": 2.8758,
+      "step": 28380
+    },
+    {
+      "epoch": 2.0886902462800494,
+      "grad_norm": 0.3649146258831024,
+      "learning_rate": 2.2077179676717412e-05,
+      "loss": 2.9166,
+      "step": 28390
+    },
+    {
+      "epoch": 2.089425959646122,
+      "grad_norm": 0.36447691917419434,
+      "learning_rate": 2.204442921922806e-05,
+      "loss": 2.8592,
+      "step": 28400
+    },
+    {
+      "epoch": 2.0901616730121946,
+      "grad_norm": 0.3466639220714569,
+      "learning_rate": 2.2011696200911154e-05,
+      "loss": 2.8297,
+      "step": 28410
+    },
+    {
+      "epoch": 2.090897386378267,
+      "grad_norm": 0.3450965881347656,
+      "learning_rate": 2.197898064218616e-05,
+      "loss": 2.8411,
+      "step": 28420
+    },
+    {
+      "epoch": 2.0916330997443398,
+      "grad_norm": 0.3541812598705292,
+      "learning_rate": 2.19462825634616e-05,
+      "loss": 2.8466,
+      "step": 28430
+    },
+    {
+      "epoch": 2.0923688131104123,
+      "grad_norm": 0.34512439370155334,
+      "learning_rate": 2.1913601985135106e-05,
+      "loss": 2.8612,
+      "step": 28440
+    },
+    {
+      "epoch": 2.093104526476485,
+      "grad_norm": 0.36206093430519104,
+      "learning_rate": 2.1880938927593377e-05,
+      "loss": 2.8765,
+      "step": 28450
+    },
+    {
+      "epoch": 2.0938402398425575,
+      "grad_norm": 0.35912156105041504,
+      "learning_rate": 2.18482934112122e-05,
+      "loss": 2.8622,
+      "step": 28460
+    },
+    {
+      "epoch": 2.09457595320863,
+      "grad_norm": 0.352763295173645,
+      "learning_rate": 2.1815665456356455e-05,
+      "loss": 2.8459,
+      "step": 28470
+    },
+    {
+      "epoch": 2.0953116665747027,
+      "grad_norm": 0.3398890495300293,
+      "learning_rate": 2.1783055083379994e-05,
+      "loss": 2.8699,
+      "step": 28480
+    },
+    {
+      "epoch": 2.0960473799407753,
+      "grad_norm": 0.34886324405670166,
+      "learning_rate": 2.1750462312625753e-05,
+      "loss": 2.8716,
+      "step": 28490
+    },
+    {
+      "epoch": 2.096783093306848,
+      "grad_norm": 0.3519967496395111,
+      "learning_rate": 2.1717887164425638e-05,
+      "loss": 2.8303,
+      "step": 28500
+    },
+    {
+      "epoch": 2.0975188066729205,
+      "grad_norm": 0.34490108489990234,
+      "learning_rate": 2.168532965910064e-05,
+      "loss": 2.8603,
+      "step": 28510
+    },
+    {
+      "epoch": 2.0982545200389926,
+      "grad_norm": 0.3449423015117645,
+      "learning_rate": 2.165278981696068e-05,
+      "loss": 2.899,
+      "step": 28520
+    },
+    {
+      "epoch": 2.098990233405065,
+      "grad_norm": 0.35227787494659424,
+      "learning_rate": 2.1620267658304672e-05,
+      "loss": 2.8301,
+      "step": 28530
+    },
+    {
+      "epoch": 2.0997259467711378,
+      "grad_norm": 0.3561704158782959,
+      "learning_rate": 2.1587763203420474e-05,
+      "loss": 2.8536,
+      "step": 28540
+    },
+    {
+      "epoch": 2.1004616601372104,
+      "grad_norm": 0.3587498068809509,
+      "learning_rate": 2.1555276472585002e-05,
+      "loss": 2.8609,
+      "step": 28550
+    },
+    {
+      "epoch": 2.101197373503283,
+      "grad_norm": 0.3579300343990326,
+      "learning_rate": 2.152280748606401e-05,
+      "loss": 2.8769,
+      "step": 28560
+    },
+    {
+      "epoch": 2.1019330868693555,
+      "grad_norm": 0.36251360177993774,
+      "learning_rate": 2.149035626411223e-05,
+      "loss": 2.9072,
+      "step": 28570
+    },
+    {
+      "epoch": 2.102668800235428,
+      "grad_norm": 0.3517322838306427,
+      "learning_rate": 2.1457922826973252e-05,
+      "loss": 2.8716,
+      "step": 28580
+    },
+    {
+      "epoch": 2.1034045136015007,
+      "grad_norm": 0.3522200584411621,
+      "learning_rate": 2.1425507194879685e-05,
+      "loss": 2.836,
+      "step": 28590
+    },
+    {
+      "epoch": 2.1041402269675733,
+      "grad_norm": 0.3648603558540344,
+      "learning_rate": 2.1393109388052934e-05,
+      "loss": 2.8464,
+      "step": 28600
+    },
+    {
+      "epoch": 2.104875940333646,
+      "grad_norm": 0.3534643054008484,
+      "learning_rate": 2.1360729426703323e-05,
+      "loss": 2.8606,
+      "step": 28610
+    },
+    {
+      "epoch": 2.1056116536997185,
+      "grad_norm": 0.3388735055923462,
+      "learning_rate": 2.1328367331030007e-05,
+      "loss": 2.8498,
+      "step": 28620
+    },
+    {
+      "epoch": 2.106347367065791,
+      "grad_norm": 0.3504191040992737,
+      "learning_rate": 2.1296023121221054e-05,
+      "loss": 2.8225,
+      "step": 28630
+    },
+    {
+      "epoch": 2.1070830804318637,
+      "grad_norm": 0.3467470109462738,
+      "learning_rate": 2.126369681745336e-05,
+      "loss": 2.8568,
+      "step": 28640
+    },
+    {
+      "epoch": 2.1078187937979362,
+      "grad_norm": 0.3596804738044739,
+      "learning_rate": 2.1231388439892618e-05,
+      "loss": 2.8834,
+      "step": 28650
+    },
+    {
+      "epoch": 2.108554507164009,
+      "grad_norm": 0.3548164963722229,
+      "learning_rate": 2.1199098008693356e-05,
+      "loss": 2.8829,
+      "step": 28660
+    },
+    {
+      "epoch": 2.1092902205300814,
+      "grad_norm": 0.3690046966075897,
+      "learning_rate": 2.1166825543998898e-05,
+      "loss": 2.8693,
+      "step": 28670
+    },
+    {
+      "epoch": 2.110025933896154,
+      "grad_norm": 0.3560093939304352,
+      "learning_rate": 2.1134571065941393e-05,
+      "loss": 2.8889,
+      "step": 28680
+    },
+    {
+      "epoch": 2.1107616472622266,
+      "grad_norm": 0.3471561074256897,
+      "learning_rate": 2.1102334594641743e-05,
+      "loss": 2.8518,
+      "step": 28690
+    },
+    {
+      "epoch": 2.111497360628299,
+      "grad_norm": 0.3563952147960663,
+      "learning_rate": 2.1070116150209594e-05,
+      "loss": 2.8527,
+      "step": 28700
+    },
+    {
+      "epoch": 2.1122330739943718,
+      "grad_norm": 0.3563307821750641,
+      "learning_rate": 2.103791575274341e-05,
+      "loss": 2.8263,
+      "step": 28710
+    },
+    {
+      "epoch": 2.1129687873604444,
+      "grad_norm": 0.3588174283504486,
+      "learning_rate": 2.1005733422330327e-05,
+      "loss": 2.8489,
+      "step": 28720
+    },
+    {
+      "epoch": 2.113704500726517,
+      "grad_norm": 0.34314316511154175,
+      "learning_rate": 2.0973569179046282e-05,
+      "loss": 2.8384,
+      "step": 28730
+    },
+    {
+      "epoch": 2.1144402140925895,
+      "grad_norm": 0.3531612455844879,
+      "learning_rate": 2.0941423042955865e-05,
+      "loss": 2.8951,
+      "step": 28740
+    },
+    {
+      "epoch": 2.115175927458662,
+      "grad_norm": 0.3679933547973633,
+      "learning_rate": 2.09092950341124e-05,
+      "loss": 2.8917,
+      "step": 28750
+    },
+    {
+      "epoch": 2.1159116408247347,
+      "grad_norm": 0.35472986102104187,
+      "learning_rate": 2.0877185172557884e-05,
+      "loss": 2.7987,
+      "step": 28760
+    },
+    {
+      "epoch": 2.1166473541908073,
+      "grad_norm": 0.3635730445384979,
+      "learning_rate": 2.084509347832304e-05,
+      "loss": 2.9029,
+      "step": 28770
+    },
+    {
+      "epoch": 2.11738306755688,
+      "grad_norm": 0.340679794549942,
+      "learning_rate": 2.0813019971427217e-05,
+      "loss": 2.8592,
+      "step": 28780
+    },
+    {
+      "epoch": 2.1181187809229525,
+      "grad_norm": 0.3596000373363495,
+      "learning_rate": 2.07809646718784e-05,
+      "loss": 2.9141,
+      "step": 28790
+    },
+    {
+      "epoch": 2.118854494289025,
+      "grad_norm": 0.3629891276359558,
+      "learning_rate": 2.074892759967328e-05,
+      "loss": 2.9065,
+      "step": 28800
+    },
+    {
+      "epoch": 2.1195902076550976,
+      "grad_norm": 0.35146433115005493,
+      "learning_rate": 2.071690877479711e-05,
+      "loss": 2.8603,
+      "step": 28810
+    },
+    {
+      "epoch": 2.1203259210211702,
+      "grad_norm": 0.3603712320327759,
+      "learning_rate": 2.0684908217223826e-05,
+      "loss": 2.8335,
+      "step": 28820
+    },
+    {
+      "epoch": 2.121061634387243,
+      "grad_norm": 0.3600578308105469,
+      "learning_rate": 2.0652925946915914e-05,
+      "loss": 2.8764,
+      "step": 28830
+    },
+    {
+      "epoch": 2.1217973477533154,
+      "grad_norm": 0.3532291054725647,
+      "learning_rate": 2.0620961983824477e-05,
+      "loss": 2.8874,
+      "step": 28840
+    },
+    {
+      "epoch": 2.122533061119388,
+      "grad_norm": 0.3531654477119446,
+      "learning_rate": 2.0589016347889167e-05,
+      "loss": 2.881,
+      "step": 28850
+    },
+    {
+      "epoch": 2.1232687744854606,
+      "grad_norm": 0.35075318813323975,
+      "learning_rate": 2.055708905903827e-05,
+      "loss": 2.831,
+      "step": 28860
+    },
+    {
+      "epoch": 2.124004487851533,
+      "grad_norm": 0.35174763202667236,
+      "learning_rate": 2.052518013718855e-05,
+      "loss": 2.8545,
+      "step": 28870
+    },
+    {
+      "epoch": 2.1247402012176058,
+      "grad_norm": 0.354565292596817,
+      "learning_rate": 2.0493289602245375e-05,
+      "loss": 2.8563,
+      "step": 28880
+    },
+    {
+      "epoch": 2.1254759145836783,
+      "grad_norm": 0.3590035140514374,
+      "learning_rate": 2.0461417474102613e-05,
+      "loss": 2.8748,
+      "step": 28890
+    },
+    {
+      "epoch": 2.126211627949751,
+      "grad_norm": 0.3570881187915802,
+      "learning_rate": 2.0429563772642628e-05,
+      "loss": 2.8241,
+      "step": 28900
+    },
+    {
+      "epoch": 2.1269473413158235,
+      "grad_norm": 0.35554957389831543,
+      "learning_rate": 2.039772851773635e-05,
+      "loss": 2.8722,
+      "step": 28910
+    },
+    {
+      "epoch": 2.127683054681896,
+      "grad_norm": 0.35786405205726624,
+      "learning_rate": 2.0365911729243147e-05,
+      "loss": 2.906,
+      "step": 28920
+    },
+    {
+      "epoch": 2.1284187680479687,
+      "grad_norm": 0.35172638297080994,
+      "learning_rate": 2.0334113427010892e-05,
+      "loss": 2.8627,
+      "step": 28930
+    },
+    {
+      "epoch": 2.1291544814140413,
+      "grad_norm": 0.34918928146362305,
+      "learning_rate": 2.0302333630875896e-05,
+      "loss": 2.8525,
+      "step": 28940
+    },
+    {
+      "epoch": 2.1298901947801134,
+      "grad_norm": 0.3627742528915405,
+      "learning_rate": 2.0270572360662972e-05,
+      "loss": 2.883,
+      "step": 28950
+    },
+    {
+      "epoch": 2.130625908146186,
+      "grad_norm": 0.3587284982204437,
+      "learning_rate": 2.023882963618537e-05,
+      "loss": 2.8795,
+      "step": 28960
+    },
+    {
+      "epoch": 2.1313616215122586,
+      "grad_norm": 0.3587230145931244,
+      "learning_rate": 2.0207105477244738e-05,
+      "loss": 2.8743,
+      "step": 28970
+    },
+    {
+      "epoch": 2.132097334878331,
+      "grad_norm": 0.35807812213897705,
+      "learning_rate": 2.017539990363116e-05,
+      "loss": 2.893,
+      "step": 28980
+    },
+    {
+      "epoch": 2.132833048244404,
+      "grad_norm": 0.3529198467731476,
+      "learning_rate": 2.0143712935123104e-05,
+      "loss": 2.8728,
+      "step": 28990
+    },
+    {
+      "epoch": 2.1335687616104764,
+      "grad_norm": 0.34928378462791443,
+      "learning_rate": 2.0112044591487495e-05,
+      "loss": 2.8243,
+      "step": 29000
+    },
+    {
+      "epoch": 2.134304474976549,
+      "grad_norm": 0.4021648168563843,
+      "learning_rate": 2.0080394892479577e-05,
+      "loss": 2.8428,
+      "step": 29010
+    },
+    {
+      "epoch": 2.1350401883426215,
+      "grad_norm": 0.3597836196422577,
+      "learning_rate": 2.0048763857842984e-05,
+      "loss": 2.8472,
+      "step": 29020
+    },
+    {
+      "epoch": 2.135775901708694,
+      "grad_norm": 0.3403218984603882,
+      "learning_rate": 2.0017151507309684e-05,
+      "loss": 2.8141,
+      "step": 29030
+    },
+    {
+      "epoch": 2.1365116150747667,
+      "grad_norm": 0.35421621799468994,
+      "learning_rate": 1.998555786060004e-05,
+      "loss": 2.9135,
+      "step": 29040
+    },
+    {
+      "epoch": 2.1372473284408393,
+      "grad_norm": 0.3545888662338257,
+      "learning_rate": 1.995398293742274e-05,
+      "loss": 2.8744,
+      "step": 29050
+    },
+    {
+      "epoch": 2.137983041806912,
+      "grad_norm": 0.35791316628456116,
+      "learning_rate": 1.992242675747474e-05,
+      "loss": 2.8529,
+      "step": 29060
+    },
+    {
+      "epoch": 2.1387187551729845,
+      "grad_norm": 0.351622998714447,
+      "learning_rate": 1.989088934044134e-05,
+      "loss": 2.861,
+      "step": 29070
+    },
+    {
+      "epoch": 2.139454468539057,
+      "grad_norm": 0.3443295657634735,
+      "learning_rate": 1.9859370705996117e-05,
+      "loss": 2.8545,
+      "step": 29080
+    },
+    {
+      "epoch": 2.1401901819051297,
+      "grad_norm": 0.35405850410461426,
+      "learning_rate": 1.982787087380098e-05,
+      "loss": 2.9025,
+      "step": 29090
+    },
+    {
+      "epoch": 2.1409258952712023,
+      "grad_norm": 0.35653066635131836,
+      "learning_rate": 1.979638986350605e-05,
+      "loss": 2.8962,
+      "step": 29100
+    },
+    {
+      "epoch": 2.141661608637275,
+      "grad_norm": 0.3546355068683624,
+      "learning_rate": 1.9764927694749734e-05,
+      "loss": 2.7963,
+      "step": 29110
+    },
+    {
+      "epoch": 2.1423973220033474,
+      "grad_norm": 0.3488227427005768,
+      "learning_rate": 1.9733484387158668e-05,
+      "loss": 2.8665,
+      "step": 29120
+    },
+    {
+      "epoch": 2.14313303536942,
+      "grad_norm": 0.35039350390434265,
+      "learning_rate": 1.9702059960347748e-05,
+      "loss": 2.8649,
+      "step": 29130
+    },
+    {
+      "epoch": 2.1438687487354926,
+      "grad_norm": 0.3531217575073242,
+      "learning_rate": 1.96706544339201e-05,
+      "loss": 2.9176,
+      "step": 29140
+    },
+    {
+      "epoch": 2.144604462101565,
+      "grad_norm": 0.3568272590637207,
+      "learning_rate": 1.963926782746701e-05,
+      "loss": 2.8411,
+      "step": 29150
+    },
+    {
+      "epoch": 2.1453401754676378,
+      "grad_norm": 0.35572025179862976,
+      "learning_rate": 1.960790016056801e-05,
+      "loss": 2.9292,
+      "step": 29160
+    },
+    {
+      "epoch": 2.1460758888337104,
+      "grad_norm": 0.35546526312828064,
+      "learning_rate": 1.9576551452790763e-05,
+      "loss": 2.8828,
+      "step": 29170
+    },
+    {
+      "epoch": 2.146811602199783,
+      "grad_norm": 0.34929758310317993,
+      "learning_rate": 1.954522172369119e-05,
+      "loss": 2.8523,
+      "step": 29180
+    },
+    {
+      "epoch": 2.1475473155658555,
+      "grad_norm": 0.3504849076271057,
+      "learning_rate": 1.9513910992813284e-05,
+      "loss": 2.8675,
+      "step": 29190
+    },
+    {
+      "epoch": 2.148283028931928,
+      "grad_norm": 0.3539334237575531,
+      "learning_rate": 1.9482619279689223e-05,
+      "loss": 2.8417,
+      "step": 29200
+    },
+    {
+      "epoch": 2.1490187422980007,
+      "grad_norm": 0.35210680961608887,
+      "learning_rate": 1.945134660383936e-05,
+      "loss": 2.847,
+      "step": 29210
+    },
+    {
+      "epoch": 2.1497544556640733,
+      "grad_norm": 0.3585560917854309,
+      "learning_rate": 1.9420092984772087e-05,
+      "loss": 2.9,
+      "step": 29220
+    },
+    {
+      "epoch": 2.150490169030146,
+      "grad_norm": 0.35338830947875977,
+      "learning_rate": 1.9388858441984003e-05,
+      "loss": 2.8633,
+      "step": 29230
+    },
+    {
+      "epoch": 2.1512258823962185,
+      "grad_norm": 0.34229418635368347,
+      "learning_rate": 1.935764299495974e-05,
+      "loss": 2.8552,
+      "step": 29240
+    },
+    {
+      "epoch": 2.151961595762291,
+      "grad_norm": 0.36570578813552856,
+      "learning_rate": 1.9326446663172036e-05,
+      "loss": 2.9031,
+      "step": 29250
+    },
+    {
+      "epoch": 2.1526973091283637,
+      "grad_norm": 0.3476269245147705,
+      "learning_rate": 1.9295269466081695e-05,
+      "loss": 2.8634,
+      "step": 29260
+    },
+    {
+      "epoch": 2.1534330224944362,
+      "grad_norm": 0.35977232456207275,
+      "learning_rate": 1.926411142313762e-05,
+      "loss": 2.8201,
+      "step": 29270
+    },
+    {
+      "epoch": 2.154168735860509,
+      "grad_norm": 0.3490533232688904,
+      "learning_rate": 1.9232972553776718e-05,
+      "loss": 2.8463,
+      "step": 29280
+    },
+    {
+      "epoch": 2.1549044492265814,
+      "grad_norm": 0.34933799505233765,
+      "learning_rate": 1.9201852877423987e-05,
+      "loss": 2.8841,
+      "step": 29290
+    },
+    {
+      "epoch": 2.155640162592654,
+      "grad_norm": 0.36150920391082764,
+      "learning_rate": 1.917075241349241e-05,
+      "loss": 2.914,
+      "step": 29300
+    },
+    {
+      "epoch": 2.1563758759587266,
+      "grad_norm": 0.36244997382164,
+      "learning_rate": 1.9139671181382978e-05,
+      "loss": 2.8232,
+      "step": 29310
+    },
+    {
+      "epoch": 2.157111589324799,
+      "grad_norm": 0.3603561818599701,
+      "learning_rate": 1.910860920048474e-05,
+      "loss": 2.8895,
+      "step": 29320
+    },
+    {
+      "epoch": 2.1578473026908718,
+      "grad_norm": 0.35820990800857544,
+      "learning_rate": 1.9077566490174686e-05,
+      "loss": 2.8629,
+      "step": 29330
+    },
+    {
+      "epoch": 2.1585830160569444,
+      "grad_norm": 0.3611311912536621,
+      "learning_rate": 1.9046543069817798e-05,
+      "loss": 2.8859,
+      "step": 29340
+    },
+    {
+      "epoch": 2.159318729423017,
+      "grad_norm": 0.35775139927864075,
+      "learning_rate": 1.901553895876701e-05,
+      "loss": 2.8436,
+      "step": 29350
+    },
+    {
+      "epoch": 2.1600544427890895,
+      "grad_norm": 0.3468707203865051,
+      "learning_rate": 1.8984554176363246e-05,
+      "loss": 2.86,
+      "step": 29360
+    },
+    {
+      "epoch": 2.160790156155162,
+      "grad_norm": 0.3513431251049042,
+      "learning_rate": 1.8953588741935374e-05,
+      "loss": 2.8699,
+      "step": 29370
+    },
+    {
+      "epoch": 2.1615258695212347,
+      "grad_norm": 0.3561619818210602,
+      "learning_rate": 1.8922642674800147e-05,
+      "loss": 2.8753,
+      "step": 29380
+    },
+    {
+      "epoch": 2.1622615828873073,
+      "grad_norm": 0.35381826758384705,
+      "learning_rate": 1.889171599426226e-05,
+      "loss": 2.8686,
+      "step": 29390
+    },
+    {
+      "epoch": 2.16299729625338,
+      "grad_norm": 0.35704073309898376,
+      "learning_rate": 1.886080871961431e-05,
+      "loss": 2.8702,
+      "step": 29400
+    },
+    {
+      "epoch": 2.1637330096194525,
+      "grad_norm": 0.34303492307662964,
+      "learning_rate": 1.8829920870136813e-05,
+      "loss": 2.8931,
+      "step": 29410
+    },
+    {
+      "epoch": 2.164468722985525,
+      "grad_norm": 0.3545786142349243,
+      "learning_rate": 1.8799052465098144e-05,
+      "loss": 2.8388,
+      "step": 29420
+    },
+    {
+      "epoch": 2.1652044363515976,
+      "grad_norm": 0.3528405725955963,
+      "learning_rate": 1.8768203523754545e-05,
+      "loss": 2.9245,
+      "step": 29430
+    },
+    {
+      "epoch": 2.16594014971767,
+      "grad_norm": 0.3620670735836029,
+      "learning_rate": 1.873737406535011e-05,
+      "loss": 2.9028,
+      "step": 29440
+    },
+    {
+      "epoch": 2.1666758630837424,
+      "grad_norm": 0.3597315847873688,
+      "learning_rate": 1.870656410911681e-05,
+      "loss": 2.8774,
+      "step": 29450
+    },
+    {
+      "epoch": 2.167411576449815,
+      "grad_norm": 0.3558579087257385,
+      "learning_rate": 1.867577367427445e-05,
+      "loss": 2.9009,
+      "step": 29460
+    },
+    {
+      "epoch": 2.1681472898158876,
+      "grad_norm": 0.35555997490882874,
+      "learning_rate": 1.8645002780030612e-05,
+      "loss": 2.8736,
+      "step": 29470
+    },
+    {
+      "epoch": 2.16888300318196,
+      "grad_norm": 0.35428252816200256,
+      "learning_rate": 1.8614251445580715e-05,
+      "loss": 2.8962,
+      "step": 29480
+    },
+    {
+      "epoch": 2.1696187165480327,
+      "grad_norm": 0.35141873359680176,
+      "learning_rate": 1.8583519690108003e-05,
+      "loss": 2.8594,
+      "step": 29490
+    },
+    {
+      "epoch": 2.1703544299141053,
+      "grad_norm": 0.34716492891311646,
+      "learning_rate": 1.855280753278347e-05,
+      "loss": 2.8773,
+      "step": 29500
+    },
+    {
+      "epoch": 2.171090143280178,
+      "grad_norm": 0.35277166962623596,
+      "learning_rate": 1.852211499276589e-05,
+      "loss": 2.8714,
+      "step": 29510
+    },
+    {
+      "epoch": 2.1718258566462505,
+      "grad_norm": 0.3487655520439148,
+      "learning_rate": 1.8491442089201787e-05,
+      "loss": 2.8617,
+      "step": 29520
+    },
+    {
+      "epoch": 2.172561570012323,
+      "grad_norm": 0.3530464768409729,
+      "learning_rate": 1.8460788841225467e-05,
+      "loss": 2.8831,
+      "step": 29530
+    },
+    {
+      "epoch": 2.1732972833783957,
+      "grad_norm": 0.3531593680381775,
+      "learning_rate": 1.8430155267958994e-05,
+      "loss": 2.872,
+      "step": 29540
+    },
+    {
+      "epoch": 2.1740329967444683,
+      "grad_norm": 0.35470977425575256,
+      "learning_rate": 1.8399541388512086e-05,
+      "loss": 2.8573,
+      "step": 29550
+    },
+    {
+      "epoch": 2.174768710110541,
+      "grad_norm": 0.34664246439933777,
+      "learning_rate": 1.836894722198223e-05,
+      "loss": 2.8264,
+      "step": 29560
+    },
+    {
+      "epoch": 2.1755044234766134,
+      "grad_norm": 0.3503643572330475,
+      "learning_rate": 1.8338372787454578e-05,
+      "loss": 2.8534,
+      "step": 29570
+    },
+    {
+      "epoch": 2.176240136842686,
+      "grad_norm": 0.3581112027168274,
+      "learning_rate": 1.8307818104002033e-05,
+      "loss": 2.8718,
+      "step": 29580
+    },
+    {
+      "epoch": 2.1769758502087586,
+      "grad_norm": 0.3483083248138428,
+      "learning_rate": 1.827728319068512e-05,
+      "loss": 2.8718,
+      "step": 29590
+    },
+    {
+      "epoch": 2.177711563574831,
+      "grad_norm": 0.3648553490638733,
+      "learning_rate": 1.8246768066552045e-05,
+      "loss": 2.8985,
+      "step": 29600
+    },
+    {
+      "epoch": 2.178447276940904,
+      "grad_norm": 0.34987473487854004,
+      "learning_rate": 1.8216272750638664e-05,
+      "loss": 2.8831,
+      "step": 29610
+    },
+    {
+      "epoch": 2.1791829903069764,
+      "grad_norm": 0.356497198343277,
+      "learning_rate": 1.81857972619685e-05,
+      "loss": 2.8433,
+      "step": 29620
+    },
+    {
+      "epoch": 2.179918703673049,
+      "grad_norm": 0.35180333256721497,
+      "learning_rate": 1.8155341619552713e-05,
+      "loss": 2.8282,
+      "step": 29630
+    },
+    {
+      "epoch": 2.1806544170391216,
+      "grad_norm": 0.34457477927207947,
+      "learning_rate": 1.8124905842390048e-05,
+      "loss": 2.8538,
+      "step": 29640
+    },
+    {
+      "epoch": 2.181390130405194,
+      "grad_norm": 0.3532300889492035,
+      "learning_rate": 1.8094489949466875e-05,
+      "loss": 2.8197,
+      "step": 29650
+    },
+    {
+      "epoch": 2.1821258437712667,
+      "grad_norm": 0.3724002242088318,
+      "learning_rate": 1.806409395975714e-05,
+      "loss": 2.8839,
+      "step": 29660
+    },
+    {
+      "epoch": 2.1828615571373393,
+      "grad_norm": 0.35598838329315186,
+      "learning_rate": 1.803371789222242e-05,
+      "loss": 2.8498,
+      "step": 29670
+    },
+    {
+      "epoch": 2.183597270503412,
+      "grad_norm": 0.3570966422557831,
+      "learning_rate": 1.8003361765811823e-05,
+      "loss": 2.9048,
+      "step": 29680
+    },
+    {
+      "epoch": 2.1843329838694845,
+      "grad_norm": 0.36287152767181396,
+      "learning_rate": 1.7973025599462012e-05,
+      "loss": 2.8625,
+      "step": 29690
+    },
+    {
+      "epoch": 2.185068697235557,
+      "grad_norm": 0.3662583827972412,
+      "learning_rate": 1.794270941209725e-05,
+      "loss": 2.8575,
+      "step": 29700
+    },
+    {
+      "epoch": 2.1858044106016297,
+      "grad_norm": 0.353614866733551,
+      "learning_rate": 1.791241322262927e-05,
+      "loss": 2.8278,
+      "step": 29710
+    },
+    {
+      "epoch": 2.1865401239677023,
+      "grad_norm": 0.35752639174461365,
+      "learning_rate": 1.7882137049957404e-05,
+      "loss": 2.8038,
+      "step": 29720
+    },
+    {
+      "epoch": 2.187275837333775,
+      "grad_norm": 0.3487575948238373,
+      "learning_rate": 1.7851880912968428e-05,
+      "loss": 2.8296,
+      "step": 29730
+    },
+    {
+      "epoch": 2.1880115506998474,
+      "grad_norm": 0.3605508506298065,
+      "learning_rate": 1.7821644830536648e-05,
+      "loss": 2.8892,
+      "step": 29740
+    },
+    {
+      "epoch": 2.18874726406592,
+      "grad_norm": 0.354921817779541,
+      "learning_rate": 1.7791428821523852e-05,
+      "loss": 2.8943,
+      "step": 29750
+    },
+    {
+      "epoch": 2.1894829774319926,
+      "grad_norm": 0.3598881959915161,
+      "learning_rate": 1.7761232904779347e-05,
+      "loss": 2.8262,
+      "step": 29760
+    },
+    {
+      "epoch": 2.190218690798065,
+      "grad_norm": 0.3541226089000702,
+      "learning_rate": 1.7731057099139842e-05,
+      "loss": 2.9035,
+      "step": 29770
+    },
+    {
+      "epoch": 2.190954404164138,
+      "grad_norm": 0.3546208441257477,
+      "learning_rate": 1.770090142342953e-05,
+      "loss": 2.8582,
+      "step": 29780
+    },
+    {
+      "epoch": 2.1916901175302104,
+      "grad_norm": 0.35933205485343933,
+      "learning_rate": 1.767076589646007e-05,
+      "loss": 2.8849,
+      "step": 29790
+    },
+    {
+      "epoch": 2.192425830896283,
+      "grad_norm": 0.350543737411499,
+      "learning_rate": 1.7640650537030495e-05,
+      "loss": 2.8795,
+      "step": 29800
+    },
+    {
+      "epoch": 2.1931615442623555,
+      "grad_norm": 0.3527929186820984,
+      "learning_rate": 1.7610555363927334e-05,
+      "loss": 2.8651,
+      "step": 29810
+    },
+    {
+      "epoch": 2.193897257628428,
+      "grad_norm": 0.3579067289829254,
+      "learning_rate": 1.7580480395924454e-05,
+      "loss": 2.864,
+      "step": 29820
+    },
+    {
+      "epoch": 2.1946329709945007,
+      "grad_norm": 0.3585156202316284,
+      "learning_rate": 1.7550425651783154e-05,
+      "loss": 2.8617,
+      "step": 29830
+    },
+    {
+      "epoch": 2.1953686843605733,
+      "grad_norm": 0.3578930199146271,
+      "learning_rate": 1.752039115025208e-05,
+      "loss": 2.8402,
+      "step": 29840
+    },
+    {
+      "epoch": 2.196104397726646,
+      "grad_norm": 0.35976335406303406,
+      "learning_rate": 1.7490376910067325e-05,
+      "loss": 2.8427,
+      "step": 29850
+    },
+    {
+      "epoch": 2.1968401110927185,
+      "grad_norm": 0.3478127419948578,
+      "learning_rate": 1.746038294995225e-05,
+      "loss": 2.8406,
+      "step": 29860
+    },
+    {
+      "epoch": 2.1975758244587906,
+      "grad_norm": 0.3557499349117279,
+      "learning_rate": 1.7430409288617644e-05,
+      "loss": 2.8875,
+      "step": 29870
+    },
+    {
+      "epoch": 2.198311537824863,
+      "grad_norm": 0.3486866056919098,
+      "learning_rate": 1.740045594476159e-05,
+      "loss": 2.8562,
+      "step": 29880
+    },
+    {
+      "epoch": 2.199047251190936,
+      "grad_norm": 0.36545079946517944,
+      "learning_rate": 1.7370522937069483e-05,
+      "loss": 2.8624,
+      "step": 29890
+    },
+    {
+      "epoch": 2.1997829645570084,
+      "grad_norm": 0.34855061769485474,
+      "learning_rate": 1.7340610284214098e-05,
+      "loss": 2.8573,
+      "step": 29900
+    },
+    {
+      "epoch": 2.200518677923081,
+      "grad_norm": 0.3478793501853943,
+      "learning_rate": 1.7310718004855443e-05,
+      "loss": 2.85,
+      "step": 29910
+    },
+    {
+      "epoch": 2.2012543912891536,
+      "grad_norm": 0.35258328914642334,
+      "learning_rate": 1.7280846117640843e-05,
+      "loss": 2.8546,
+      "step": 29920
+    },
+    {
+      "epoch": 2.201990104655226,
+      "grad_norm": 0.3589334785938263,
+      "learning_rate": 1.7250994641204888e-05,
+      "loss": 2.8975,
+      "step": 29930
+    },
+    {
+      "epoch": 2.2027258180212987,
+      "grad_norm": 0.355837881565094,
+      "learning_rate": 1.722116359416947e-05,
+      "loss": 2.8826,
+      "step": 29940
+    },
+    {
+      "epoch": 2.2034615313873713,
+      "grad_norm": 0.3589230477809906,
+      "learning_rate": 1.7191352995143727e-05,
+      "loss": 2.8654,
+      "step": 29950
+    },
+    {
+      "epoch": 2.204197244753444,
+      "grad_norm": 0.3549625873565674,
+      "learning_rate": 1.7161562862724017e-05,
+      "loss": 2.8968,
+      "step": 29960
+    },
+    {
+      "epoch": 2.2049329581195165,
+      "grad_norm": 0.34318357706069946,
+      "learning_rate": 1.7131793215493945e-05,
+      "loss": 2.8247,
+      "step": 29970
+    },
+    {
+      "epoch": 2.205668671485589,
+      "grad_norm": 0.358465313911438,
+      "learning_rate": 1.7102044072024316e-05,
+      "loss": 2.8689,
+      "step": 29980
+    },
+    {
+      "epoch": 2.2064043848516617,
+      "grad_norm": 0.3595186173915863,
+      "learning_rate": 1.70723154508732e-05,
+      "loss": 2.8731,
+      "step": 29990
+    },
+    {
+      "epoch": 2.2071400982177343,
+      "grad_norm": 0.3505420386791229,
+      "learning_rate": 1.7042607370585813e-05,
+      "loss": 2.8681,
+      "step": 30000
+    },
+    {
+      "epoch": 2.207875811583807,
+      "grad_norm": 0.3564614951610565,
+      "learning_rate": 1.701291984969457e-05,
+      "loss": 2.8844,
+      "step": 30010
+    },
+    {
+      "epoch": 2.2086115249498794,
+      "grad_norm": 0.3593564033508301,
+      "learning_rate": 1.698325290671905e-05,
+      "loss": 2.8769,
+      "step": 30020
+    },
+    {
+      "epoch": 2.209347238315952,
+      "grad_norm": 0.3504774272441864,
+      "learning_rate": 1.6953606560166025e-05,
+      "loss": 2.8699,
+      "step": 30030
+    },
+    {
+      "epoch": 2.2100829516820246,
+      "grad_norm": 0.35134217143058777,
+      "learning_rate": 1.6923980828529422e-05,
+      "loss": 2.8704,
+      "step": 30040
+    },
+    {
+      "epoch": 2.210818665048097,
+      "grad_norm": 0.3567639887332916,
+      "learning_rate": 1.6894375730290273e-05,
+      "loss": 2.875,
+      "step": 30050
+    },
+    {
+      "epoch": 2.21155437841417,
+      "grad_norm": 0.3586840033531189,
+      "learning_rate": 1.6864791283916747e-05,
+      "loss": 2.8604,
+      "step": 30060
+    },
+    {
+      "epoch": 2.2122900917802424,
+      "grad_norm": 0.35104134678840637,
+      "learning_rate": 1.6835227507864128e-05,
+      "loss": 2.8282,
+      "step": 30070
+    },
+    {
+      "epoch": 2.213025805146315,
+      "grad_norm": 0.3533231019973755,
+      "learning_rate": 1.6805684420574847e-05,
+      "loss": 2.8793,
+      "step": 30080
+    },
+    {
+      "epoch": 2.2137615185123876,
+      "grad_norm": 0.35919442772865295,
+      "learning_rate": 1.6776162040478376e-05,
+      "loss": 2.828,
+      "step": 30090
+    },
+    {
+      "epoch": 2.21449723187846,
+      "grad_norm": 0.3534218370914459,
+      "learning_rate": 1.674666038599128e-05,
+      "loss": 2.8496,
+      "step": 30100
+    },
+    {
+      "epoch": 2.2152329452445327,
+      "grad_norm": 0.3447113633155823,
+      "learning_rate": 1.6717179475517235e-05,
+      "loss": 2.8552,
+      "step": 30110
+    },
+    {
+      "epoch": 2.2159686586106053,
+      "grad_norm": 0.35506489872932434,
+      "learning_rate": 1.6687719327446916e-05,
+      "loss": 2.8858,
+      "step": 30120
+    },
+    {
+      "epoch": 2.216704371976678,
+      "grad_norm": 0.3585444688796997,
+      "learning_rate": 1.6658279960158106e-05,
+      "loss": 2.8516,
+      "step": 30130
+    },
+    {
+      "epoch": 2.2174400853427505,
+      "grad_norm": 0.35561603307724,
+      "learning_rate": 1.662886139201559e-05,
+      "loss": 2.9266,
+      "step": 30140
+    },
+    {
+      "epoch": 2.218175798708823,
+      "grad_norm": 0.3660274147987366,
+      "learning_rate": 1.6599463641371167e-05,
+      "loss": 2.835,
+      "step": 30150
+    },
+    {
+      "epoch": 2.2189115120748957,
+      "grad_norm": 0.3526948392391205,
+      "learning_rate": 1.657008672656366e-05,
+      "loss": 2.8174,
+      "step": 30160
+    },
+    {
+      "epoch": 2.2196472254409683,
+      "grad_norm": 0.34681206941604614,
+      "learning_rate": 1.6540730665918934e-05,
+      "loss": 2.8428,
+      "step": 30170
+    },
+    {
+      "epoch": 2.220382938807041,
+      "grad_norm": 0.348311185836792,
+      "learning_rate": 1.651139547774979e-05,
+      "loss": 2.8632,
+      "step": 30180
+    },
+    {
+      "epoch": 2.2211186521731134,
+      "grad_norm": 0.35942888259887695,
+      "learning_rate": 1.648208118035602e-05,
+      "loss": 2.8496,
+      "step": 30190
+    },
+    {
+      "epoch": 2.221854365539186,
+      "grad_norm": 0.3563474416732788,
+      "learning_rate": 1.6452787792024423e-05,
+      "loss": 2.8335,
+      "step": 30200
+    },
+    {
+      "epoch": 2.2225900789052586,
+      "grad_norm": 0.35734716057777405,
+      "learning_rate": 1.642351533102869e-05,
+      "loss": 2.8977,
+      "step": 30210
+    },
+    {
+      "epoch": 2.223325792271331,
+      "grad_norm": 0.35699233412742615,
+      "learning_rate": 1.6394263815629528e-05,
+      "loss": 2.8826,
+      "step": 30220
+    },
+    {
+      "epoch": 2.224061505637404,
+      "grad_norm": 0.35940930247306824,
+      "learning_rate": 1.636503326407453e-05,
+      "loss": 2.8761,
+      "step": 30230
+    },
+    {
+      "epoch": 2.2247972190034764,
+      "grad_norm": 0.35721567273139954,
+      "learning_rate": 1.6335823694598218e-05,
+      "loss": 2.8716,
+      "step": 30240
+    },
+    {
+      "epoch": 2.225532932369549,
+      "grad_norm": 0.3633698523044586,
+      "learning_rate": 1.6306635125422015e-05,
+      "loss": 2.9182,
+      "step": 30250
+    },
+    {
+      "epoch": 2.2262686457356216,
+      "grad_norm": 0.355669230222702,
+      "learning_rate": 1.6277467574754292e-05,
+      "loss": 2.8831,
+      "step": 30260
+    },
+    {
+      "epoch": 2.227004359101694,
+      "grad_norm": 0.34684884548187256,
+      "learning_rate": 1.6248321060790246e-05,
+      "loss": 2.8326,
+      "step": 30270
+    },
+    {
+      "epoch": 2.2277400724677667,
+      "grad_norm": 0.35859817266464233,
+      "learning_rate": 1.621919560171201e-05,
+      "loss": 2.8761,
+      "step": 30280
+    },
+    {
+      "epoch": 2.2284757858338393,
+      "grad_norm": 0.35695263743400574,
+      "learning_rate": 1.6190091215688525e-05,
+      "loss": 2.8862,
+      "step": 30290
+    },
+    {
+      "epoch": 2.229211499199912,
+      "grad_norm": 0.3560574948787689,
+      "learning_rate": 1.6161007920875642e-05,
+      "loss": 2.8982,
+      "step": 30300
+    },
+    {
+      "epoch": 2.2299472125659845,
+      "grad_norm": 0.36406657099723816,
+      "learning_rate": 1.613194573541601e-05,
+      "loss": 2.9263,
+      "step": 30310
+    },
+    {
+      "epoch": 2.230682925932057,
+      "grad_norm": 0.3575332760810852,
+      "learning_rate": 1.6102904677439135e-05,
+      "loss": 2.875,
+      "step": 30320
+    },
+    {
+      "epoch": 2.2314186392981297,
+      "grad_norm": 0.3546263873577118,
+      "learning_rate": 1.6073884765061325e-05,
+      "loss": 2.8542,
+      "step": 30330
+    },
+    {
+      "epoch": 2.2321543526642023,
+      "grad_norm": 0.34124618768692017,
+      "learning_rate": 1.60448860163857e-05,
+      "loss": 2.8586,
+      "step": 30340
+    },
+    {
+      "epoch": 2.2328900660302744,
+      "grad_norm": 0.3621087074279785,
+      "learning_rate": 1.60159084495022e-05,
+      "loss": 2.8666,
+      "step": 30350
+    },
+    {
+      "epoch": 2.233625779396347,
+      "grad_norm": 0.35944443941116333,
+      "learning_rate": 1.5986952082487548e-05,
+      "loss": 2.8637,
+      "step": 30360
+    },
+    {
+      "epoch": 2.2343614927624196,
+      "grad_norm": 0.36275267601013184,
+      "learning_rate": 1.5958016933405218e-05,
+      "loss": 2.861,
+      "step": 30370
+    },
+    {
+      "epoch": 2.235097206128492,
+      "grad_norm": 0.3626326024532318,
+      "learning_rate": 1.592910302030544e-05,
+      "loss": 2.8741,
+      "step": 30380
+    },
+    {
+      "epoch": 2.2358329194945648,
+      "grad_norm": 0.3577110767364502,
+      "learning_rate": 1.5900210361225255e-05,
+      "loss": 2.8993,
+      "step": 30390
+    },
+    {
+      "epoch": 2.2365686328606373,
+      "grad_norm": 0.35099855065345764,
+      "learning_rate": 1.5871338974188387e-05,
+      "loss": 2.7792,
+      "step": 30400
+    },
+    {
+      "epoch": 2.23730434622671,
+      "grad_norm": 0.36162930727005005,
+      "learning_rate": 1.5842488877205315e-05,
+      "loss": 2.832,
+      "step": 30410
+    },
+    {
+      "epoch": 2.2380400595927825,
+      "grad_norm": 0.3429602086544037,
+      "learning_rate": 1.5813660088273215e-05,
+      "loss": 2.8215,
+      "step": 30420
+    },
+    {
+      "epoch": 2.238775772958855,
+      "grad_norm": 0.3666570484638214,
+      "learning_rate": 1.5784852625375996e-05,
+      "loss": 2.8683,
+      "step": 30430
+    },
+    {
+      "epoch": 2.2395114863249277,
+      "grad_norm": 0.35528475046157837,
+      "learning_rate": 1.5756066506484286e-05,
+      "loss": 2.9026,
+      "step": 30440
+    },
+    {
+      "epoch": 2.2402471996910003,
+      "grad_norm": 0.34826335310935974,
+      "learning_rate": 1.572730174955534e-05,
+      "loss": 2.8335,
+      "step": 30450
+    },
+    {
+      "epoch": 2.240982913057073,
+      "grad_norm": 0.35766372084617615,
+      "learning_rate": 1.5698558372533124e-05,
+      "loss": 2.8696,
+      "step": 30460
+    },
+    {
+      "epoch": 2.2417186264231455,
+      "grad_norm": 0.35495641827583313,
+      "learning_rate": 1.5669836393348232e-05,
+      "loss": 2.8483,
+      "step": 30470
+    },
+    {
+      "epoch": 2.242454339789218,
+      "grad_norm": 0.35490477085113525,
+      "learning_rate": 1.5641135829917973e-05,
+      "loss": 2.8378,
+      "step": 30480
+    },
+    {
+      "epoch": 2.2431900531552906,
+      "grad_norm": 0.3528565764427185,
+      "learning_rate": 1.561245670014625e-05,
+      "loss": 2.8808,
+      "step": 30490
+    },
+    {
+      "epoch": 2.243925766521363,
+      "grad_norm": 0.3607218563556671,
+      "learning_rate": 1.558379902192359e-05,
+      "loss": 2.8563,
+      "step": 30500
+    },
+    {
+      "epoch": 2.244661479887436,
+      "grad_norm": 0.366131067276001,
+      "learning_rate": 1.5555162813127146e-05,
+      "loss": 2.848,
+      "step": 30510
+    },
+    {
+      "epoch": 2.2453971932535084,
+      "grad_norm": 0.3425903916358948,
+      "learning_rate": 1.5526548091620697e-05,
+      "loss": 2.8769,
+      "step": 30520
+    },
+    {
+      "epoch": 2.246132906619581,
+      "grad_norm": 0.3594939410686493,
+      "learning_rate": 1.5497954875254615e-05,
+      "loss": 2.8683,
+      "step": 30530
+    },
+    {
+      "epoch": 2.2468686199856536,
+      "grad_norm": 0.3551132082939148,
+      "learning_rate": 1.546938318186585e-05,
+      "loss": 2.8581,
+      "step": 30540
+    },
+    {
+      "epoch": 2.247604333351726,
+      "grad_norm": 0.3665505349636078,
+      "learning_rate": 1.54408330292779e-05,
+      "loss": 2.8643,
+      "step": 30550
+    },
+    {
+      "epoch": 2.2483400467177987,
+      "grad_norm": 0.35343581438064575,
+      "learning_rate": 1.5412304435300846e-05,
+      "loss": 2.885,
+      "step": 30560
+    },
+    {
+      "epoch": 2.2490757600838713,
+      "grad_norm": 0.36262479424476624,
+      "learning_rate": 1.538379741773135e-05,
+      "loss": 2.8316,
+      "step": 30570
+    },
+    {
+      "epoch": 2.249811473449944,
+      "grad_norm": 0.3485947251319885,
+      "learning_rate": 1.5355311994352572e-05,
+      "loss": 2.8349,
+      "step": 30580
+    },
+    {
+      "epoch": 2.2505471868160165,
+      "grad_norm": 0.3594449460506439,
+      "learning_rate": 1.5326848182934212e-05,
+      "loss": 2.8515,
+      "step": 30590
+    },
+    {
+      "epoch": 2.251282900182089,
+      "grad_norm": 0.36621734499931335,
+      "learning_rate": 1.529840600123248e-05,
+      "loss": 2.8596,
+      "step": 30600
+    },
+    {
+      "epoch": 2.2520186135481617,
+      "grad_norm": 0.3586372137069702,
+      "learning_rate": 1.5269985466990118e-05,
+      "loss": 2.8656,
+      "step": 30610
+    },
+    {
+      "epoch": 2.2527543269142343,
+      "grad_norm": 0.35911306738853455,
+      "learning_rate": 1.524158659793637e-05,
+      "loss": 2.8677,
+      "step": 30620
+    },
+    {
+      "epoch": 2.253490040280307,
+      "grad_norm": 0.35749679803848267,
+      "learning_rate": 1.5213209411786927e-05,
+      "loss": 2.8448,
+      "step": 30630
+    },
+    {
+      "epoch": 2.2542257536463794,
+      "grad_norm": 0.3654407858848572,
+      "learning_rate": 1.518485392624397e-05,
+      "loss": 2.8419,
+      "step": 30640
+    },
+    {
+      "epoch": 2.254961467012452,
+      "grad_norm": 0.35691043734550476,
+      "learning_rate": 1.5156520158996135e-05,
+      "loss": 2.8778,
+      "step": 30650
+    },
+    {
+      "epoch": 2.2556971803785246,
+      "grad_norm": 0.3702915608882904,
+      "learning_rate": 1.5128208127718552e-05,
+      "loss": 2.8954,
+      "step": 30660
+    },
+    {
+      "epoch": 2.256432893744597,
+      "grad_norm": 0.3591904044151306,
+      "learning_rate": 1.5099917850072737e-05,
+      "loss": 2.7922,
+      "step": 30670
+    },
+    {
+      "epoch": 2.25716860711067,
+      "grad_norm": 0.34788626432418823,
+      "learning_rate": 1.5071649343706651e-05,
+      "loss": 2.8443,
+      "step": 30680
+    },
+    {
+      "epoch": 2.2579043204767424,
+      "grad_norm": 0.35170507431030273,
+      "learning_rate": 1.5043402626254705e-05,
+      "loss": 2.8645,
+      "step": 30690
+    },
+    {
+      "epoch": 2.258640033842815,
+      "grad_norm": 0.3579574227333069,
+      "learning_rate": 1.501517771533767e-05,
+      "loss": 2.8691,
+      "step": 30700
+    },
+    {
+      "epoch": 2.2593757472088876,
+      "grad_norm": 0.3538725972175598,
+      "learning_rate": 1.4986974628562762e-05,
+      "loss": 2.8394,
+      "step": 30710
+    },
+    {
+      "epoch": 2.26011146057496,
+      "grad_norm": 0.3550536334514618,
+      "learning_rate": 1.4958793383523544e-05,
+      "loss": 2.8518,
+      "step": 30720
+    },
+    {
+      "epoch": 2.2608471739410327,
+      "grad_norm": 0.3522687554359436,
+      "learning_rate": 1.4930633997799975e-05,
+      "loss": 2.8698,
+      "step": 30730
+    },
+    {
+      "epoch": 2.2615828873071053,
+      "grad_norm": 0.35918912291526794,
+      "learning_rate": 1.4902496488958345e-05,
+      "loss": 2.8576,
+      "step": 30740
+    },
+    {
+      "epoch": 2.262318600673178,
+      "grad_norm": 0.35444319248199463,
+      "learning_rate": 1.4874380874551358e-05,
+      "loss": 2.8666,
+      "step": 30750
+    },
+    {
+      "epoch": 2.26305431403925,
+      "grad_norm": 0.35712435841560364,
+      "learning_rate": 1.4846287172117996e-05,
+      "loss": 2.8569,
+      "step": 30760
+    },
+    {
+      "epoch": 2.2637900274053226,
+      "grad_norm": 0.36596885323524475,
+      "learning_rate": 1.4818215399183627e-05,
+      "loss": 2.8356,
+      "step": 30770
+    },
+    {
+      "epoch": 2.2645257407713952,
+      "grad_norm": 0.3585096001625061,
+      "learning_rate": 1.4790165573259902e-05,
+      "loss": 2.8752,
+      "step": 30780
+    },
+    {
+      "epoch": 2.265261454137468,
+      "grad_norm": 0.3525094985961914,
+      "learning_rate": 1.476213771184477e-05,
+      "loss": 2.8196,
+      "step": 30790
+    },
+    {
+      "epoch": 2.2659971675035404,
+      "grad_norm": 0.3555854558944702,
+      "learning_rate": 1.4734131832422537e-05,
+      "loss": 2.8594,
+      "step": 30800
+    },
+    {
+      "epoch": 2.266732880869613,
+      "grad_norm": 0.35339590907096863,
+      "learning_rate": 1.470614795246374e-05,
+      "loss": 2.8824,
+      "step": 30810
+    },
+    {
+      "epoch": 2.2674685942356856,
+      "grad_norm": 0.3578505218029022,
+      "learning_rate": 1.4678186089425212e-05,
+      "loss": 2.8246,
+      "step": 30820
+    },
+    {
+      "epoch": 2.268204307601758,
+      "grad_norm": 0.356576144695282,
+      "learning_rate": 1.4650246260750033e-05,
+      "loss": 2.7958,
+      "step": 30830
+    },
+    {
+      "epoch": 2.2689400209678308,
+      "grad_norm": 0.3533620238304138,
+      "learning_rate": 1.4622328483867587e-05,
+      "loss": 2.8775,
+      "step": 30840
+    },
+    {
+      "epoch": 2.2696757343339033,
+      "grad_norm": 0.3509010374546051,
+      "learning_rate": 1.459443277619344e-05,
+      "loss": 2.8325,
+      "step": 30850
+    },
+    {
+      "epoch": 2.270411447699976,
+      "grad_norm": 0.36384648084640503,
+      "learning_rate": 1.456655915512945e-05,
+      "loss": 2.8792,
+      "step": 30860
+    },
+    {
+      "epoch": 2.2711471610660485,
+      "grad_norm": 0.36681991815567017,
+      "learning_rate": 1.4538707638063659e-05,
+      "loss": 2.8309,
+      "step": 30870
+    },
+    {
+      "epoch": 2.271882874432121,
+      "grad_norm": 0.35373207926750183,
+      "learning_rate": 1.4510878242370302e-05,
+      "loss": 2.849,
+      "step": 30880
+    },
+    {
+      "epoch": 2.2726185877981937,
+      "grad_norm": 0.36072272062301636,
+      "learning_rate": 1.4483070985409886e-05,
+      "loss": 2.8502,
+      "step": 30890
+    },
+    {
+      "epoch": 2.2733543011642663,
+      "grad_norm": 0.3501628339290619,
+      "learning_rate": 1.4455285884529035e-05,
+      "loss": 2.8645,
+      "step": 30900
+    },
+    {
+      "epoch": 2.274090014530339,
+      "grad_norm": 0.35633695125579834,
+      "learning_rate": 1.4427522957060585e-05,
+      "loss": 2.8745,
+      "step": 30910
+    },
+    {
+      "epoch": 2.2748257278964115,
+      "grad_norm": 0.35605162382125854,
+      "learning_rate": 1.4399782220323515e-05,
+      "loss": 2.8298,
+      "step": 30920
+    },
+    {
+      "epoch": 2.275561441262484,
+      "grad_norm": 0.35702913999557495,
+      "learning_rate": 1.4372063691623006e-05,
+      "loss": 2.8894,
+      "step": 30930
+    },
+    {
+      "epoch": 2.2762971546285566,
+      "grad_norm": 0.3572458326816559,
+      "learning_rate": 1.4344367388250363e-05,
+      "loss": 2.8562,
+      "step": 30940
+    },
+    {
+      "epoch": 2.2770328679946292,
+      "grad_norm": 0.35724306106567383,
+      "learning_rate": 1.4316693327483011e-05,
+      "loss": 2.9298,
+      "step": 30950
+    },
+    {
+      "epoch": 2.277768581360702,
+      "grad_norm": 0.355093777179718,
+      "learning_rate": 1.4289041526584512e-05,
+      "loss": 2.8795,
+      "step": 30960
+    },
+    {
+      "epoch": 2.2785042947267744,
+      "grad_norm": 0.35014182329177856,
+      "learning_rate": 1.4261412002804531e-05,
+      "loss": 2.8216,
+      "step": 30970
+    },
+    {
+      "epoch": 2.279240008092847,
+      "grad_norm": 0.35190922021865845,
+      "learning_rate": 1.4233804773378867e-05,
+      "loss": 2.8627,
+      "step": 30980
+    },
+    {
+      "epoch": 2.2799757214589196,
+      "grad_norm": 0.3628680109977722,
+      "learning_rate": 1.420621985552939e-05,
+      "loss": 2.8669,
+      "step": 30990
+    },
+    {
+      "epoch": 2.280711434824992,
+      "grad_norm": 0.3443011939525604,
+      "learning_rate": 1.417865726646404e-05,
+      "loss": 2.8277,
+      "step": 31000
+    },
+    {
+      "epoch": 2.2814471481910648,
+      "grad_norm": 0.34494051337242126,
+      "learning_rate": 1.4151117023376831e-05,
+      "loss": 2.8532,
+      "step": 31010
+    },
+    {
+      "epoch": 2.2821828615571373,
+      "grad_norm": 0.3549286723136902,
+      "learning_rate": 1.4123599143447857e-05,
+      "loss": 2.8903,
+      "step": 31020
+    },
+    {
+      "epoch": 2.28291857492321,
+      "grad_norm": 0.35300466418266296,
+      "learning_rate": 1.4096103643843277e-05,
+      "loss": 2.8749,
+      "step": 31030
+    },
+    {
+      "epoch": 2.2836542882892825,
+      "grad_norm": 0.3517322540283203,
+      "learning_rate": 1.4068630541715239e-05,
+      "loss": 2.8105,
+      "step": 31040
+    },
+    {
+      "epoch": 2.284390001655355,
+      "grad_norm": 0.3430500328540802,
+      "learning_rate": 1.4041179854201947e-05,
+      "loss": 2.8487,
+      "step": 31050
+    },
+    {
+      "epoch": 2.2851257150214277,
+      "grad_norm": 0.3613154888153076,
+      "learning_rate": 1.4013751598427599e-05,
+      "loss": 2.867,
+      "step": 31060
+    },
+    {
+      "epoch": 2.2858614283875003,
+      "grad_norm": 0.3647073805332184,
+      "learning_rate": 1.3986345791502453e-05,
+      "loss": 2.885,
+      "step": 31070
+    },
+    {
+      "epoch": 2.286597141753573,
+      "grad_norm": 0.362000048160553,
+      "learning_rate": 1.395896245052271e-05,
+      "loss": 2.8429,
+      "step": 31080
+    },
+    {
+      "epoch": 2.2873328551196455,
+      "grad_norm": 0.35595250129699707,
+      "learning_rate": 1.393160159257057e-05,
+      "loss": 2.878,
+      "step": 31090
+    },
+    {
+      "epoch": 2.288068568485718,
+      "grad_norm": 0.3460536003112793,
+      "learning_rate": 1.3904263234714238e-05,
+      "loss": 2.8596,
+      "step": 31100
+    },
+    {
+      "epoch": 2.2888042818517906,
+      "grad_norm": 0.3532830476760864,
+      "learning_rate": 1.3876947394007822e-05,
+      "loss": 2.7992,
+      "step": 31110
+    },
+    {
+      "epoch": 2.289539995217863,
+      "grad_norm": 0.34651488065719604,
+      "learning_rate": 1.384965408749146e-05,
+      "loss": 2.8254,
+      "step": 31120
+    },
+    {
+      "epoch": 2.290275708583936,
+      "grad_norm": 0.35325226187705994,
+      "learning_rate": 1.3822383332191169e-05,
+      "loss": 2.8104,
+      "step": 31130
+    },
+    {
+      "epoch": 2.2910114219500084,
+      "grad_norm": 0.3669968247413635,
+      "learning_rate": 1.3795135145118926e-05,
+      "loss": 2.8822,
+      "step": 31140
+    },
+    {
+      "epoch": 2.291747135316081,
+      "grad_norm": 0.3454671800136566,
+      "learning_rate": 1.3767909543272611e-05,
+      "loss": 2.8517,
+      "step": 31150
+    },
+    {
+      "epoch": 2.2924828486821536,
+      "grad_norm": 0.3456681966781616,
+      "learning_rate": 1.3740706543636056e-05,
+      "loss": 2.8891,
+      "step": 31160
+    },
+    {
+      "epoch": 2.293218562048226,
+      "grad_norm": 0.34885454177856445,
+      "learning_rate": 1.3713526163178935e-05,
+      "loss": 2.8738,
+      "step": 31170
+    },
+    {
+      "epoch": 2.2939542754142987,
+      "grad_norm": 0.35897067189216614,
+      "learning_rate": 1.3686368418856882e-05,
+      "loss": 2.8237,
+      "step": 31180
+    },
+    {
+      "epoch": 2.2946899887803713,
+      "grad_norm": 0.35465535521507263,
+      "learning_rate": 1.3659233327611332e-05,
+      "loss": 2.8605,
+      "step": 31190
+    },
+    {
+      "epoch": 2.295425702146444,
+      "grad_norm": 0.34609144926071167,
+      "learning_rate": 1.3632120906369661e-05,
+      "loss": 2.8759,
+      "step": 31200
+    },
+    {
+      "epoch": 2.2961614155125165,
+      "grad_norm": 0.3520955741405487,
+      "learning_rate": 1.3605031172045057e-05,
+      "loss": 2.8708,
+      "step": 31210
+    },
+    {
+      "epoch": 2.296897128878589,
+      "grad_norm": 0.3607901930809021,
+      "learning_rate": 1.357796414153657e-05,
+      "loss": 2.8548,
+      "step": 31220
+    },
+    {
+      "epoch": 2.2976328422446617,
+      "grad_norm": 0.35058385133743286,
+      "learning_rate": 1.3550919831729069e-05,
+      "loss": 2.8092,
+      "step": 31230
+    },
+    {
+      "epoch": 2.2983685556107343,
+      "grad_norm": 0.3589896261692047,
+      "learning_rate": 1.3523898259493289e-05,
+      "loss": 2.8975,
+      "step": 31240
+    },
+    {
+      "epoch": 2.299104268976807,
+      "grad_norm": 0.3654545843601227,
+      "learning_rate": 1.349689944168575e-05,
+      "loss": 2.8821,
+      "step": 31250
+    },
+    {
+      "epoch": 2.2998399823428795,
+      "grad_norm": 0.3606531322002411,
+      "learning_rate": 1.3469923395148775e-05,
+      "loss": 2.826,
+      "step": 31260
+    },
+    {
+      "epoch": 2.300575695708952,
+      "grad_norm": 0.35612982511520386,
+      "learning_rate": 1.3442970136710509e-05,
+      "loss": 2.8531,
+      "step": 31270
+    },
+    {
+      "epoch": 2.301311409075024,
+      "grad_norm": 0.35295265913009644,
+      "learning_rate": 1.3416039683184844e-05,
+      "loss": 2.8424,
+      "step": 31280
+    },
+    {
+      "epoch": 2.3020471224410968,
+      "grad_norm": 0.3507929742336273,
+      "learning_rate": 1.3389132051371494e-05,
+      "loss": 2.8833,
+      "step": 31290
+    },
+    {
+      "epoch": 2.3027828358071694,
+      "grad_norm": 0.3578413426876068,
+      "learning_rate": 1.3362247258055898e-05,
+      "loss": 2.8711,
+      "step": 31300
+    },
+    {
+      "epoch": 2.303518549173242,
+      "grad_norm": 0.35754716396331787,
+      "learning_rate": 1.3335385320009253e-05,
+      "loss": 2.8547,
+      "step": 31310
+    },
+    {
+      "epoch": 2.3042542625393145,
+      "grad_norm": 0.3636312484741211,
+      "learning_rate": 1.3308546253988491e-05,
+      "loss": 2.8445,
+      "step": 31320
+    },
+    {
+      "epoch": 2.304989975905387,
+      "grad_norm": 0.3605746328830719,
+      "learning_rate": 1.3281730076736315e-05,
+      "loss": 2.8745,
+      "step": 31330
+    },
+    {
+      "epoch": 2.3057256892714597,
+      "grad_norm": 0.35188964009284973,
+      "learning_rate": 1.3254936804981105e-05,
+      "loss": 2.833,
+      "step": 31340
+    },
+    {
+      "epoch": 2.3064614026375323,
+      "grad_norm": 0.34149134159088135,
+      "learning_rate": 1.3228166455436992e-05,
+      "loss": 2.8431,
+      "step": 31350
+    },
+    {
+      "epoch": 2.307197116003605,
+      "grad_norm": 0.355039119720459,
+      "learning_rate": 1.320141904480377e-05,
+      "loss": 2.8763,
+      "step": 31360
+    },
+    {
+      "epoch": 2.3079328293696775,
+      "grad_norm": 0.3662640452384949,
+      "learning_rate": 1.3174694589766933e-05,
+      "loss": 2.8419,
+      "step": 31370
+    },
+    {
+      "epoch": 2.30866854273575,
+      "grad_norm": 0.35953253507614136,
+      "learning_rate": 1.314799310699769e-05,
+      "loss": 2.869,
+      "step": 31380
+    },
+    {
+      "epoch": 2.3094042561018227,
+      "grad_norm": 0.3548811674118042,
+      "learning_rate": 1.3121314613152874e-05,
+      "loss": 2.915,
+      "step": 31390
+    },
+    {
+      "epoch": 2.3101399694678952,
+      "grad_norm": 0.34871944785118103,
+      "learning_rate": 1.3094659124874997e-05,
+      "loss": 2.8593,
+      "step": 31400
+    },
+    {
+      "epoch": 2.310875682833968,
+      "grad_norm": 0.35166510939598083,
+      "learning_rate": 1.3068026658792198e-05,
+      "loss": 2.8775,
+      "step": 31410
+    },
+    {
+      "epoch": 2.3116113962000404,
+      "grad_norm": 0.3533203601837158,
+      "learning_rate": 1.3041417231518294e-05,
+      "loss": 2.8286,
+      "step": 31420
+    },
+    {
+      "epoch": 2.312347109566113,
+      "grad_norm": 0.3497558534145355,
+      "learning_rate": 1.3014830859652722e-05,
+      "loss": 2.8532,
+      "step": 31430
+    },
+    {
+      "epoch": 2.3130828229321856,
+      "grad_norm": 0.3611317574977875,
+      "learning_rate": 1.2988267559780509e-05,
+      "loss": 2.889,
+      "step": 31440
+    },
+    {
+      "epoch": 2.313818536298258,
+      "grad_norm": 0.35179489850997925,
+      "learning_rate": 1.2961727348472308e-05,
+      "loss": 2.8728,
+      "step": 31450
+    },
+    {
+      "epoch": 2.3145542496643308,
+      "grad_norm": 0.36794549226760864,
+      "learning_rate": 1.293521024228434e-05,
+      "loss": 2.8251,
+      "step": 31460
+    },
+    {
+      "epoch": 2.3152899630304034,
+      "grad_norm": 0.34739014506340027,
+      "learning_rate": 1.2908716257758485e-05,
+      "loss": 2.8608,
+      "step": 31470
+    },
+    {
+      "epoch": 2.316025676396476,
+      "grad_norm": 0.3592948913574219,
+      "learning_rate": 1.2882245411422122e-05,
+      "loss": 2.8199,
+      "step": 31480
+    },
+    {
+      "epoch": 2.3167613897625485,
+      "grad_norm": 0.3445521295070648,
+      "learning_rate": 1.2855797719788238e-05,
+      "loss": 2.8507,
+      "step": 31490
+    },
+    {
+      "epoch": 2.317497103128621,
+      "grad_norm": 0.39729565382003784,
+      "learning_rate": 1.2829373199355337e-05,
+      "loss": 2.8973,
+      "step": 31500
+    },
+    {
+      "epoch": 2.3182328164946937,
+      "grad_norm": 0.3506152033805847,
+      "learning_rate": 1.2802971866607521e-05,
+      "loss": 2.8797,
+      "step": 31510
+    },
+    {
+      "epoch": 2.3189685298607663,
+      "grad_norm": 0.3569266200065613,
+      "learning_rate": 1.2776593738014415e-05,
+      "loss": 2.8614,
+      "step": 31520
+    },
+    {
+      "epoch": 2.319704243226839,
+      "grad_norm": 0.35598456859588623,
+      "learning_rate": 1.2750238830031135e-05,
+      "loss": 2.8515,
+      "step": 31530
+    },
+    {
+      "epoch": 2.3204399565929115,
+      "grad_norm": 0.36964595317840576,
+      "learning_rate": 1.2723907159098342e-05,
+      "loss": 2.8601,
+      "step": 31540
+    },
+    {
+      "epoch": 2.321175669958984,
+      "grad_norm": 0.35747453570365906,
+      "learning_rate": 1.269759874164217e-05,
+      "loss": 2.8215,
+      "step": 31550
+    },
+    {
+      "epoch": 2.3219113833250566,
+      "grad_norm": 0.3577366769313812,
+      "learning_rate": 1.2671313594074292e-05,
+      "loss": 2.8846,
+      "step": 31560
+    },
+    {
+      "epoch": 2.3226470966911292,
+      "grad_norm": 0.3563617467880249,
+      "learning_rate": 1.2645051732791834e-05,
+      "loss": 2.8537,
+      "step": 31570
+    },
+    {
+      "epoch": 2.323382810057202,
+      "grad_norm": 0.3528320789337158,
+      "learning_rate": 1.2618813174177407e-05,
+      "loss": 2.8283,
+      "step": 31580
+    },
+    {
+      "epoch": 2.3241185234232744,
+      "grad_norm": 0.35680529475212097,
+      "learning_rate": 1.259259793459906e-05,
+      "loss": 2.8665,
+      "step": 31590
+    },
+    {
+      "epoch": 2.324854236789347,
+      "grad_norm": 0.3555194139480591,
+      "learning_rate": 1.2566406030410333e-05,
+      "loss": 2.8269,
+      "step": 31600
+    },
+    {
+      "epoch": 2.3255899501554196,
+      "grad_norm": 0.3584590554237366,
+      "learning_rate": 1.2540237477950206e-05,
+      "loss": 2.8518,
+      "step": 31610
+    },
+    {
+      "epoch": 2.326325663521492,
+      "grad_norm": 0.35233354568481445,
+      "learning_rate": 1.2514092293543067e-05,
+      "loss": 2.8757,
+      "step": 31620
+    },
+    {
+      "epoch": 2.3270613768875648,
+      "grad_norm": 0.37261566519737244,
+      "learning_rate": 1.2487970493498735e-05,
+      "loss": 2.8579,
+      "step": 31630
+    },
+    {
+      "epoch": 2.3277970902536373,
+      "grad_norm": 0.3490120470523834,
+      "learning_rate": 1.2461872094112432e-05,
+      "loss": 2.8777,
+      "step": 31640
+    },
+    {
+      "epoch": 2.32853280361971,
+      "grad_norm": 0.3534788489341736,
+      "learning_rate": 1.2435797111664822e-05,
+      "loss": 2.864,
+      "step": 31650
+    },
+    {
+      "epoch": 2.3292685169857825,
+      "grad_norm": 0.3565320372581482,
+      "learning_rate": 1.2409745562421915e-05,
+      "loss": 2.8477,
+      "step": 31660
+    },
+    {
+      "epoch": 2.330004230351855,
+      "grad_norm": 0.35585370659828186,
+      "learning_rate": 1.2383717462635114e-05,
+      "loss": 2.8611,
+      "step": 31670
+    },
+    {
+      "epoch": 2.3307399437179273,
+      "grad_norm": 0.3535536527633667,
+      "learning_rate": 1.2357712828541223e-05,
+      "loss": 2.9021,
+      "step": 31680
+    },
+    {
+      "epoch": 2.331475657084,
+      "grad_norm": 0.36340251564979553,
+      "learning_rate": 1.2331731676362368e-05,
+      "loss": 2.8501,
+      "step": 31690
+    },
+    {
+      "epoch": 2.3322113704500724,
+      "grad_norm": 0.34816715121269226,
+      "learning_rate": 1.2305774022306066e-05,
+      "loss": 2.8326,
+      "step": 31700
+    },
+    {
+      "epoch": 2.332947083816145,
+      "grad_norm": 0.36916401982307434,
+      "learning_rate": 1.2279839882565141e-05,
+      "loss": 2.9293,
+      "step": 31710
+    },
+    {
+      "epoch": 2.3336827971822176,
+      "grad_norm": 0.35332873463630676,
+      "learning_rate": 1.2253929273317766e-05,
+      "loss": 2.8386,
+      "step": 31720
+    },
+    {
+      "epoch": 2.33441851054829,
+      "grad_norm": 0.3612922132015228,
+      "learning_rate": 1.2228042210727408e-05,
+      "loss": 2.8605,
+      "step": 31730
+    },
+    {
+      "epoch": 2.335154223914363,
+      "grad_norm": 0.3603113889694214,
+      "learning_rate": 1.2202178710942903e-05,
+      "loss": 2.8381,
+      "step": 31740
+    },
+    {
+      "epoch": 2.3358899372804354,
+      "grad_norm": 0.34851646423339844,
+      "learning_rate": 1.2176338790098318e-05,
+      "loss": 2.8084,
+      "step": 31750
+    },
+    {
+      "epoch": 2.336625650646508,
+      "grad_norm": 0.3553086817264557,
+      "learning_rate": 1.2150522464313086e-05,
+      "loss": 2.8742,
+      "step": 31760
+    },
+    {
+      "epoch": 2.3373613640125805,
+      "grad_norm": 0.3458890914916992,
+      "learning_rate": 1.2124729749691855e-05,
+      "loss": 2.8385,
+      "step": 31770
+    },
+    {
+      "epoch": 2.338097077378653,
+      "grad_norm": 0.34556642174720764,
+      "learning_rate": 1.209896066232456e-05,
+      "loss": 2.8627,
+      "step": 31780
+    },
+    {
+      "epoch": 2.3388327907447257,
+      "grad_norm": 0.3626575767993927,
+      "learning_rate": 1.2073215218286443e-05,
+      "loss": 2.8684,
+      "step": 31790
+    },
+    {
+      "epoch": 2.3395685041107983,
+      "grad_norm": 0.356800377368927,
+      "learning_rate": 1.2047493433637935e-05,
+      "loss": 2.915,
+      "step": 31800
+    },
+    {
+      "epoch": 2.340304217476871,
+      "grad_norm": 0.35971564054489136,
+      "learning_rate": 1.2021795324424739e-05,
+      "loss": 2.8886,
+      "step": 31810
+    },
+    {
+      "epoch": 2.3410399308429435,
+      "grad_norm": 0.36116185784339905,
+      "learning_rate": 1.1996120906677772e-05,
+      "loss": 2.8706,
+      "step": 31820
+    },
+    {
+      "epoch": 2.341775644209016,
+      "grad_norm": 0.347972571849823,
+      "learning_rate": 1.1970470196413191e-05,
+      "loss": 2.8452,
+      "step": 31830
+    },
+    {
+      "epoch": 2.3425113575750887,
+      "grad_norm": 0.34308820962905884,
+      "learning_rate": 1.1944843209632378e-05,
+      "loss": 2.8672,
+      "step": 31840
+    },
+    {
+      "epoch": 2.3432470709411612,
+      "grad_norm": 0.35902899503707886,
+      "learning_rate": 1.1919239962321876e-05,
+      "loss": 2.8309,
+      "step": 31850
+    },
+    {
+      "epoch": 2.343982784307234,
+      "grad_norm": 0.36421239376068115,
+      "learning_rate": 1.1893660470453432e-05,
+      "loss": 2.8337,
+      "step": 31860
+    },
+    {
+      "epoch": 2.3447184976733064,
+      "grad_norm": 0.356804221868515,
+      "learning_rate": 1.1868104749983976e-05,
+      "loss": 2.872,
+      "step": 31870
+    },
+    {
+      "epoch": 2.345454211039379,
+      "grad_norm": 0.35239648818969727,
+      "learning_rate": 1.1842572816855634e-05,
+      "loss": 2.8111,
+      "step": 31880
+    },
+    {
+      "epoch": 2.3461899244054516,
+      "grad_norm": 0.35379910469055176,
+      "learning_rate": 1.1817064686995654e-05,
+      "loss": 2.8801,
+      "step": 31890
+    },
+    {
+      "epoch": 2.346925637771524,
+      "grad_norm": 0.3546363413333893,
+      "learning_rate": 1.1791580376316458e-05,
+      "loss": 2.8714,
+      "step": 31900
+    },
+    {
+      "epoch": 2.3476613511375968,
+      "grad_norm": 0.3542933166027069,
+      "learning_rate": 1.176611990071559e-05,
+      "loss": 2.7759,
+      "step": 31910
+    },
+    {
+      "epoch": 2.3483970645036694,
+      "grad_norm": 0.35056272149086,
+      "learning_rate": 1.1740683276075742e-05,
+      "loss": 2.8708,
+      "step": 31920
+    },
+    {
+      "epoch": 2.349132777869742,
+      "grad_norm": 0.346612811088562,
+      "learning_rate": 1.1715270518264748e-05,
+      "loss": 2.8426,
+      "step": 31930
+    },
+    {
+      "epoch": 2.3498684912358145,
+      "grad_norm": 0.3584357500076294,
+      "learning_rate": 1.168988164313552e-05,
+      "loss": 2.8759,
+      "step": 31940
+    },
+    {
+      "epoch": 2.350604204601887,
+      "grad_norm": 0.37025776505470276,
+      "learning_rate": 1.1664516666526066e-05,
+      "loss": 2.9302,
+      "step": 31950
+    },
+    {
+      "epoch": 2.3513399179679597,
+      "grad_norm": 0.36508405208587646,
+      "learning_rate": 1.1639175604259495e-05,
+      "loss": 2.8462,
+      "step": 31960
+    },
+    {
+      "epoch": 2.3520756313340323,
+      "grad_norm": 0.35836321115493774,
+      "learning_rate": 1.1613858472144035e-05,
+      "loss": 2.873,
+      "step": 31970
+    },
+    {
+      "epoch": 2.352811344700105,
+      "grad_norm": 0.3531290292739868,
+      "learning_rate": 1.1588565285972935e-05,
+      "loss": 2.8542,
+      "step": 31980
+    },
+    {
+      "epoch": 2.3535470580661775,
+      "grad_norm": 0.36483678221702576,
+      "learning_rate": 1.1563296061524525e-05,
+      "loss": 2.8673,
+      "step": 31990
+    },
+    {
+      "epoch": 2.35428277143225,
+      "grad_norm": 0.36193570494651794,
+      "learning_rate": 1.1538050814562173e-05,
+      "loss": 2.8497,
+      "step": 32000
+    },
+    {
+      "epoch": 2.3550184847983227,
+      "grad_norm": 0.36617913842201233,
+      "learning_rate": 1.1512829560834326e-05,
+      "loss": 2.8467,
+      "step": 32010
+    },
+    {
+      "epoch": 2.3557541981643952,
+      "grad_norm": 0.3516647517681122,
+      "learning_rate": 1.1487632316074448e-05,
+      "loss": 2.847,
+      "step": 32020
+    },
+    {
+      "epoch": 2.356489911530468,
+      "grad_norm": 0.35806483030319214,
+      "learning_rate": 1.146245909600101e-05,
+      "loss": 2.8949,
+      "step": 32030
+    },
+    {
+      "epoch": 2.3572256248965404,
+      "grad_norm": 0.3554759919643402,
+      "learning_rate": 1.1437309916317507e-05,
+      "loss": 2.8349,
+      "step": 32040
+    },
+    {
+      "epoch": 2.357961338262613,
+      "grad_norm": 0.3525812327861786,
+      "learning_rate": 1.1412184792712411e-05,
+      "loss": 2.8651,
+      "step": 32050
+    },
+    {
+      "epoch": 2.3586970516286856,
+      "grad_norm": 0.35554200410842896,
+      "learning_rate": 1.1387083740859256e-05,
+      "loss": 2.841,
+      "step": 32060
+    },
+    {
+      "epoch": 2.359432764994758,
+      "grad_norm": 0.35038822889328003,
+      "learning_rate": 1.136200677641649e-05,
+      "loss": 2.8195,
+      "step": 32070
+    },
+    {
+      "epoch": 2.3601684783608308,
+      "grad_norm": 0.3603375554084778,
+      "learning_rate": 1.1336953915027548e-05,
+      "loss": 2.8649,
+      "step": 32080
+    },
+    {
+      "epoch": 2.3609041917269034,
+      "grad_norm": 0.3552504777908325,
+      "learning_rate": 1.1311925172320858e-05,
+      "loss": 2.8462,
+      "step": 32090
+    },
+    {
+      "epoch": 2.361639905092976,
+      "grad_norm": 0.3537072241306305,
+      "learning_rate": 1.1286920563909803e-05,
+      "loss": 2.8753,
+      "step": 32100
+    },
+    {
+      "epoch": 2.3623756184590485,
+      "grad_norm": 0.3583391308784485,
+      "learning_rate": 1.1261940105392677e-05,
+      "loss": 2.8666,
+      "step": 32110
+    },
+    {
+      "epoch": 2.363111331825121,
+      "grad_norm": 0.375865638256073,
+      "learning_rate": 1.1236983812352724e-05,
+      "loss": 2.8237,
+      "step": 32120
+    },
+    {
+      "epoch": 2.3638470451911937,
+      "grad_norm": 0.35605311393737793,
+      "learning_rate": 1.1212051700358101e-05,
+      "loss": 2.8202,
+      "step": 32130
+    },
+    {
+      "epoch": 2.3645827585572663,
+      "grad_norm": 0.3531211018562317,
+      "learning_rate": 1.1187143784961923e-05,
+      "loss": 2.9009,
+      "step": 32140
+    },
+    {
+      "epoch": 2.365318471923339,
+      "grad_norm": 0.345755398273468,
+      "learning_rate": 1.116226008170217e-05,
+      "loss": 2.8698,
+      "step": 32150
+    },
+    {
+      "epoch": 2.3660541852894115,
+      "grad_norm": 0.35066160559654236,
+      "learning_rate": 1.1137400606101716e-05,
+      "loss": 2.8652,
+      "step": 32160
+    },
+    {
+      "epoch": 2.366789898655484,
+      "grad_norm": 0.3563699424266815,
+      "learning_rate": 1.111256537366836e-05,
+      "loss": 2.842,
+      "step": 32170
+    },
+    {
+      "epoch": 2.3675256120215566,
+      "grad_norm": 0.34664997458457947,
+      "learning_rate": 1.1087754399894724e-05,
+      "loss": 2.8563,
+      "step": 32180
+    },
+    {
+      "epoch": 2.368261325387629,
+      "grad_norm": 0.35142841935157776,
+      "learning_rate": 1.1062967700258359e-05,
+      "loss": 2.8388,
+      "step": 32190
+    },
+    {
+      "epoch": 2.3689970387537014,
+      "grad_norm": 0.36184942722320557,
+      "learning_rate": 1.103820529022162e-05,
+      "loss": 2.8855,
+      "step": 32200
+    },
+    {
+      "epoch": 2.369732752119774,
+      "grad_norm": 0.3536902070045471,
+      "learning_rate": 1.101346718523174e-05,
+      "loss": 2.8367,
+      "step": 32210
+    },
+    {
+      "epoch": 2.3704684654858466,
+      "grad_norm": 0.35470911860466003,
+      "learning_rate": 1.0988753400720757e-05,
+      "loss": 2.8823,
+      "step": 32220
+    },
+    {
+      "epoch": 2.371204178851919,
+      "grad_norm": 0.35882291197776794,
+      "learning_rate": 1.0964063952105591e-05,
+      "loss": 2.8509,
+      "step": 32230
+    },
+    {
+      "epoch": 2.3719398922179917,
+      "grad_norm": 0.3614608645439148,
+      "learning_rate": 1.0939398854787924e-05,
+      "loss": 2.8539,
+      "step": 32240
+    },
+    {
+      "epoch": 2.3726756055840643,
+      "grad_norm": 0.35387125611305237,
+      "learning_rate": 1.09147581241543e-05,
+      "loss": 2.866,
+      "step": 32250
+    },
+    {
+      "epoch": 2.373411318950137,
+      "grad_norm": 0.35085031390190125,
+      "learning_rate": 1.089014177557603e-05,
+      "loss": 2.8716,
+      "step": 32260
+    },
+    {
+      "epoch": 2.3741470323162095,
+      "grad_norm": 0.3520703911781311,
+      "learning_rate": 1.0865549824409194e-05,
+      "loss": 2.8839,
+      "step": 32270
+    },
+    {
+      "epoch": 2.374882745682282,
+      "grad_norm": 0.34246954321861267,
+      "learning_rate": 1.0840982285994716e-05,
+      "loss": 2.8218,
+      "step": 32280
+    },
+    {
+      "epoch": 2.3756184590483547,
+      "grad_norm": 0.3553103506565094,
+      "learning_rate": 1.0816439175658244e-05,
+      "loss": 2.8339,
+      "step": 32290
+    },
+    {
+      "epoch": 2.3763541724144273,
+      "grad_norm": 0.3537861406803131,
+      "learning_rate": 1.079192050871019e-05,
+      "loss": 2.8486,
+      "step": 32300
+    },
+    {
+      "epoch": 2.3770898857805,
+      "grad_norm": 0.35975727438926697,
+      "learning_rate": 1.0767426300445716e-05,
+      "loss": 2.7862,
+      "step": 32310
+    },
+    {
+      "epoch": 2.3778255991465724,
+      "grad_norm": 0.3591471314430237,
+      "learning_rate": 1.0742956566144764e-05,
+      "loss": 2.8412,
+      "step": 32320
+    },
+    {
+      "epoch": 2.378561312512645,
+      "grad_norm": 0.35566937923431396,
+      "learning_rate": 1.071851132107195e-05,
+      "loss": 2.8619,
+      "step": 32330
+    },
+    {
+      "epoch": 2.3792970258787176,
+      "grad_norm": 0.34755295515060425,
+      "learning_rate": 1.0694090580476679e-05,
+      "loss": 2.8574,
+      "step": 32340
+    },
+    {
+      "epoch": 2.38003273924479,
+      "grad_norm": 0.3440442681312561,
+      "learning_rate": 1.0669694359593014e-05,
+      "loss": 2.8472,
+      "step": 32350
+    },
+    {
+      "epoch": 2.380768452610863,
+      "grad_norm": 0.3477231562137604,
+      "learning_rate": 1.0645322673639734e-05,
+      "loss": 2.7994,
+      "step": 32360
+    },
+    {
+      "epoch": 2.3815041659769354,
+      "grad_norm": 0.3632752001285553,
+      "learning_rate": 1.062097553782035e-05,
+      "loss": 2.8685,
+      "step": 32370
+    },
+    {
+      "epoch": 2.382239879343008,
+      "grad_norm": 0.3406107723712921,
+      "learning_rate": 1.0596652967323007e-05,
+      "loss": 2.8558,
+      "step": 32380
+    },
+    {
+      "epoch": 2.3829755927090805,
+      "grad_norm": 0.3581578731536865,
+      "learning_rate": 1.0572354977320553e-05,
+      "loss": 2.8381,
+      "step": 32390
+    },
+    {
+      "epoch": 2.383711306075153,
+      "grad_norm": 0.3552815914154053,
+      "learning_rate": 1.0548081582970493e-05,
+      "loss": 2.8523,
+      "step": 32400
+    },
+    {
+      "epoch": 2.3844470194412257,
+      "grad_norm": 0.35062411427497864,
+      "learning_rate": 1.0523832799414996e-05,
+      "loss": 2.854,
+      "step": 32410
+    },
+    {
+      "epoch": 2.3851827328072983,
+      "grad_norm": 0.34366872906684875,
+      "learning_rate": 1.0499608641780895e-05,
+      "loss": 2.8922,
+      "step": 32420
+    },
+    {
+      "epoch": 2.385918446173371,
+      "grad_norm": 0.3527258634567261,
+      "learning_rate": 1.0475409125179626e-05,
+      "loss": 2.8668,
+      "step": 32430
+    },
+    {
+      "epoch": 2.3866541595394435,
+      "grad_norm": 0.35069820284843445,
+      "learning_rate": 1.0451234264707271e-05,
+      "loss": 2.8283,
+      "step": 32440
+    },
+    {
+      "epoch": 2.387389872905516,
+      "grad_norm": 0.36639371514320374,
+      "learning_rate": 1.0427084075444515e-05,
+      "loss": 2.8471,
+      "step": 32450
+    },
+    {
+      "epoch": 2.3881255862715887,
+      "grad_norm": 0.3698582649230957,
+      "learning_rate": 1.0402958572456694e-05,
+      "loss": 2.8419,
+      "step": 32460
+    },
+    {
+      "epoch": 2.3888612996376613,
+      "grad_norm": 0.3598490059375763,
+      "learning_rate": 1.0378857770793704e-05,
+      "loss": 2.9155,
+      "step": 32470
+    },
+    {
+      "epoch": 2.389597013003734,
+      "grad_norm": 0.351211279630661,
+      "learning_rate": 1.0354781685490045e-05,
+      "loss": 2.8367,
+      "step": 32480
+    },
+    {
+      "epoch": 2.3903327263698064,
+      "grad_norm": 0.3545929491519928,
+      "learning_rate": 1.0330730331564786e-05,
+      "loss": 2.9028,
+      "step": 32490
+    },
+    {
+      "epoch": 2.391068439735879,
+      "grad_norm": 0.35224366188049316,
+      "learning_rate": 1.0306703724021594e-05,
+      "loss": 2.8005,
+      "step": 32500
+    },
+    {
+      "epoch": 2.3918041531019516,
+      "grad_norm": 0.35452401638031006,
+      "learning_rate": 1.0282701877848706e-05,
+      "loss": 2.9039,
+      "step": 32510
+    },
+    {
+      "epoch": 2.392539866468024,
+      "grad_norm": 0.3507433533668518,
+      "learning_rate": 1.0258724808018865e-05,
+      "loss": 2.8424,
+      "step": 32520
+    },
+    {
+      "epoch": 2.3932755798340968,
+      "grad_norm": 0.36394402384757996,
+      "learning_rate": 1.0234772529489395e-05,
+      "loss": 2.8268,
+      "step": 32530
+    },
+    {
+      "epoch": 2.3940112932001694,
+      "grad_norm": 0.35344988107681274,
+      "learning_rate": 1.0210845057202122e-05,
+      "loss": 2.8915,
+      "step": 32540
+    },
+    {
+      "epoch": 2.394747006566242,
+      "grad_norm": 0.36208784580230713,
+      "learning_rate": 1.018694240608346e-05,
+      "loss": 2.8898,
+      "step": 32550
+    },
+    {
+      "epoch": 2.3954827199323145,
+      "grad_norm": 0.3682962954044342,
+      "learning_rate": 1.016306459104428e-05,
+      "loss": 2.8561,
+      "step": 32560
+    },
+    {
+      "epoch": 2.396218433298387,
+      "grad_norm": 0.34901919960975647,
+      "learning_rate": 1.0139211626979967e-05,
+      "loss": 2.8612,
+      "step": 32570
+    },
+    {
+      "epoch": 2.3969541466644597,
+      "grad_norm": 0.3575538396835327,
+      "learning_rate": 1.0115383528770444e-05,
+      "loss": 2.8571,
+      "step": 32580
+    },
+    {
+      "epoch": 2.397689860030532,
+      "grad_norm": 0.3519527316093445,
+      "learning_rate": 1.0091580311280063e-05,
+      "loss": 2.8687,
+      "step": 32590
+    },
+    {
+      "epoch": 2.3984255733966044,
+      "grad_norm": 0.355743944644928,
+      "learning_rate": 1.0067801989357711e-05,
+      "loss": 2.8644,
+      "step": 32600
+    },
+    {
+      "epoch": 2.399161286762677,
+      "grad_norm": 0.3427495062351227,
+      "learning_rate": 1.0044048577836712e-05,
+      "loss": 2.8374,
+      "step": 32610
+    },
+    {
+      "epoch": 2.3998970001287496,
+      "grad_norm": 0.36114078760147095,
+      "learning_rate": 1.0020320091534862e-05,
+      "loss": 2.8578,
+      "step": 32620
+    },
+    {
+      "epoch": 2.400632713494822,
+      "grad_norm": 0.35111626982688904,
+      "learning_rate": 9.996616545254379e-06,
+      "loss": 2.8604,
+      "step": 32630
+    },
+    {
+      "epoch": 2.401368426860895,
+      "grad_norm": 0.34994420409202576,
+      "learning_rate": 9.972937953781986e-06,
+      "loss": 2.835,
+      "step": 32640
+    },
+    {
+      "epoch": 2.4021041402269674,
+      "grad_norm": 0.35318952798843384,
+      "learning_rate": 9.94928433188878e-06,
+      "loss": 2.8574,
+      "step": 32650
+    },
+    {
+      "epoch": 2.40283985359304,
+      "grad_norm": 0.3529825806617737,
+      "learning_rate": 9.925655694330299e-06,
+      "loss": 2.8515,
+      "step": 32660
+    },
+    {
+      "epoch": 2.4035755669591126,
+      "grad_norm": 0.36459651589393616,
+      "learning_rate": 9.902052055846517e-06,
+      "loss": 2.8627,
+      "step": 32670
+    },
+    {
+      "epoch": 2.404311280325185,
+      "grad_norm": 0.3571184575557709,
+      "learning_rate": 9.878473431161767e-06,
+      "loss": 2.8535,
+      "step": 32680
+    },
+    {
+      "epoch": 2.4050469936912577,
+      "grad_norm": 0.35616788268089294,
+      "learning_rate": 9.854919834984844e-06,
+      "loss": 2.8716,
+      "step": 32690
+    },
+    {
+      "epoch": 2.4057827070573303,
+      "grad_norm": 0.358401358127594,
+      "learning_rate": 9.831391282008867e-06,
+      "loss": 2.8668,
+      "step": 32700
+    },
+    {
+      "epoch": 2.406518420423403,
+      "grad_norm": 0.3607894480228424,
+      "learning_rate": 9.807887786911368e-06,
+      "loss": 2.8685,
+      "step": 32710
+    },
+    {
+      "epoch": 2.4072541337894755,
+      "grad_norm": 0.3571551442146301,
+      "learning_rate": 9.784409364354213e-06,
+      "loss": 2.8476,
+      "step": 32720
+    },
+    {
+      "epoch": 2.407989847155548,
+      "grad_norm": 0.3572337329387665,
+      "learning_rate": 9.760956028983687e-06,
+      "loss": 2.833,
+      "step": 32730
+    },
+    {
+      "epoch": 2.4087255605216207,
+      "grad_norm": 0.3588632047176361,
+      "learning_rate": 9.737527795430357e-06,
+      "loss": 2.8182,
+      "step": 32740
+    },
+    {
+      "epoch": 2.4094612738876933,
+      "grad_norm": 0.3576171398162842,
+      "learning_rate": 9.714124678309195e-06,
+      "loss": 2.8154,
+      "step": 32750
+    },
+    {
+      "epoch": 2.410196987253766,
+      "grad_norm": 0.34731629490852356,
+      "learning_rate": 9.690746692219454e-06,
+      "loss": 2.8511,
+      "step": 32760
+    },
+    {
+      "epoch": 2.4109327006198384,
+      "grad_norm": 0.35650861263275146,
+      "learning_rate": 9.667393851744721e-06,
+      "loss": 2.8151,
+      "step": 32770
+    },
+    {
+      "epoch": 2.411668413985911,
+      "grad_norm": 0.36247745156288147,
+      "learning_rate": 9.644066171452931e-06,
+      "loss": 2.892,
+      "step": 32780
+    },
+    {
+      "epoch": 2.4124041273519836,
+      "grad_norm": 0.35526689887046814,
+      "learning_rate": 9.620763665896281e-06,
+      "loss": 2.8604,
+      "step": 32790
+    },
+    {
+      "epoch": 2.413139840718056,
+      "grad_norm": 0.36496421694755554,
+      "learning_rate": 9.597486349611285e-06,
+      "loss": 2.874,
+      "step": 32800
+    },
+    {
+      "epoch": 2.413875554084129,
+      "grad_norm": 0.351098895072937,
+      "learning_rate": 9.574234237118724e-06,
+      "loss": 2.8588,
+      "step": 32810
+    },
+    {
+      "epoch": 2.4146112674502014,
+      "grad_norm": 0.36249589920043945,
+      "learning_rate": 9.551007342923679e-06,
+      "loss": 2.8873,
+      "step": 32820
+    },
+    {
+      "epoch": 2.415346980816274,
+      "grad_norm": 0.3584132492542267,
+      "learning_rate": 9.527805681515517e-06,
+      "loss": 2.8621,
+      "step": 32830
+    },
+    {
+      "epoch": 2.4160826941823466,
+      "grad_norm": 0.35482916235923767,
+      "learning_rate": 9.50462926736781e-06,
+      "loss": 2.8578,
+      "step": 32840
+    },
+    {
+      "epoch": 2.416818407548419,
+      "grad_norm": 0.3543655574321747,
+      "learning_rate": 9.481478114938424e-06,
+      "loss": 2.8409,
+      "step": 32850
+    },
+    {
+      "epoch": 2.4175541209144917,
+      "grad_norm": 0.34888774156570435,
+      "learning_rate": 9.458352238669427e-06,
+      "loss": 2.8362,
+      "step": 32860
+    },
+    {
+      "epoch": 2.4182898342805643,
+      "grad_norm": 0.35147392749786377,
+      "learning_rate": 9.435251652987181e-06,
+      "loss": 2.8596,
+      "step": 32870
+    },
+    {
+      "epoch": 2.419025547646637,
+      "grad_norm": 0.34171101450920105,
+      "learning_rate": 9.412176372302211e-06,
+      "loss": 2.8407,
+      "step": 32880
+    },
+    {
+      "epoch": 2.4197612610127095,
+      "grad_norm": 0.35053345561027527,
+      "learning_rate": 9.389126411009286e-06,
+      "loss": 2.8547,
+      "step": 32890
+    },
+    {
+      "epoch": 2.420496974378782,
+      "grad_norm": 0.35593441128730774,
+      "learning_rate": 9.366101783487352e-06,
+      "loss": 2.8727,
+      "step": 32900
+    },
+    {
+      "epoch": 2.4212326877448547,
+      "grad_norm": 0.35566771030426025,
+      "learning_rate": 9.3431025040996e-06,
+      "loss": 2.8141,
+      "step": 32910
+    },
+    {
+      "epoch": 2.4219684011109273,
+      "grad_norm": 0.3623896539211273,
+      "learning_rate": 9.320128587193382e-06,
+      "loss": 2.8695,
+      "step": 32920
+    },
+    {
+      "epoch": 2.422704114477,
+      "grad_norm": 0.3481222689151764,
+      "learning_rate": 9.297180047100217e-06,
+      "loss": 2.8549,
+      "step": 32930
+    },
+    {
+      "epoch": 2.4234398278430724,
+      "grad_norm": 0.35534611344337463,
+      "learning_rate": 9.274256898135814e-06,
+      "loss": 2.8369,
+      "step": 32940
+    },
+    {
+      "epoch": 2.424175541209145,
+      "grad_norm": 0.35732102394104004,
+      "learning_rate": 9.251359154600008e-06,
+      "loss": 2.8331,
+      "step": 32950
+    },
+    {
+      "epoch": 2.4249112545752176,
+      "grad_norm": 0.3531363010406494,
+      "learning_rate": 9.228486830776845e-06,
+      "loss": 2.8125,
+      "step": 32960
+    },
+    {
+      "epoch": 2.42564696794129,
+      "grad_norm": 0.35451528429985046,
+      "learning_rate": 9.205639940934463e-06,
+      "loss": 2.7992,
+      "step": 32970
+    },
+    {
+      "epoch": 2.426382681307363,
+      "grad_norm": 0.35614630579948425,
+      "learning_rate": 9.182818499325147e-06,
+      "loss": 2.822,
+      "step": 32980
+    },
+    {
+      "epoch": 2.4271183946734354,
+      "grad_norm": 0.34831297397613525,
+      "learning_rate": 9.16002252018529e-06,
+      "loss": 2.903,
+      "step": 32990
+    },
+    {
+      "epoch": 2.427854108039508,
+      "grad_norm": 0.3459022045135498,
+      "learning_rate": 9.13725201773547e-06,
+      "loss": 2.8503,
+      "step": 33000
+    },
+    {
+      "epoch": 2.4285898214055806,
+      "grad_norm": 0.3472837507724762,
+      "learning_rate": 9.114507006180294e-06,
+      "loss": 2.8468,
+      "step": 33010
+    },
+    {
+      "epoch": 2.429325534771653,
+      "grad_norm": 0.3544243276119232,
+      "learning_rate": 9.09178749970851e-06,
+      "loss": 2.8844,
+      "step": 33020
+    },
+    {
+      "epoch": 2.4300612481377257,
+      "grad_norm": 0.36265090107917786,
+      "learning_rate": 9.069093512492916e-06,
+      "loss": 2.883,
+      "step": 33030
+    },
+    {
+      "epoch": 2.4307969615037983,
+      "grad_norm": 0.3590310513973236,
+      "learning_rate": 9.046425058690444e-06,
+      "loss": 2.871,
+      "step": 33040
+    },
+    {
+      "epoch": 2.431532674869871,
+      "grad_norm": 0.35629740357398987,
+      "learning_rate": 9.023782152442062e-06,
+      "loss": 2.838,
+      "step": 33050
+    },
+    {
+      "epoch": 2.4322683882359435,
+      "grad_norm": 0.3514953851699829,
+      "learning_rate": 9.001164807872808e-06,
+      "loss": 2.8299,
+      "step": 33060
+    },
+    {
+      "epoch": 2.433004101602016,
+      "grad_norm": 0.3529304563999176,
+      "learning_rate": 8.97857303909177e-06,
+      "loss": 2.8446,
+      "step": 33070
+    },
+    {
+      "epoch": 2.4337398149680887,
+      "grad_norm": 0.35669639706611633,
+      "learning_rate": 8.956006860192084e-06,
+      "loss": 2.876,
+      "step": 33080
+    },
+    {
+      "epoch": 2.4344755283341613,
+      "grad_norm": 0.35620713233947754,
+      "learning_rate": 8.933466285250958e-06,
+      "loss": 2.871,
+      "step": 33090
+    },
+    {
+      "epoch": 2.4352112417002334,
+      "grad_norm": 0.36282792687416077,
+      "learning_rate": 8.910951328329565e-06,
+      "loss": 2.8866,
+      "step": 33100
+    },
+    {
+      "epoch": 2.435946955066306,
+      "grad_norm": 0.35579872131347656,
+      "learning_rate": 8.88846200347314e-06,
+      "loss": 2.8492,
+      "step": 33110
+    },
+    {
+      "epoch": 2.4366826684323786,
+      "grad_norm": 0.3675694167613983,
+      "learning_rate": 8.865998324710901e-06,
+      "loss": 2.8654,
+      "step": 33120
+    },
+    {
+      "epoch": 2.437418381798451,
+      "grad_norm": 0.35112717747688293,
+      "learning_rate": 8.843560306056104e-06,
+      "loss": 2.8037,
+      "step": 33130
+    },
+    {
+      "epoch": 2.4381540951645237,
+      "grad_norm": 0.351921945810318,
+      "learning_rate": 8.82114796150596e-06,
+      "loss": 2.8866,
+      "step": 33140
+    },
+    {
+      "epoch": 2.4388898085305963,
+      "grad_norm": 0.35680723190307617,
+      "learning_rate": 8.79876130504168e-06,
+      "loss": 2.859,
+      "step": 33150
+    },
+    {
+      "epoch": 2.439625521896669,
+      "grad_norm": 0.35821032524108887,
+      "learning_rate": 8.776400350628466e-06,
+      "loss": 2.8817,
+      "step": 33160
+    },
+    {
+      "epoch": 2.4403612352627415,
+      "grad_norm": 0.3592459261417389,
+      "learning_rate": 8.754065112215454e-06,
+      "loss": 2.8238,
+      "step": 33170
+    },
+    {
+      "epoch": 2.441096948628814,
+      "grad_norm": 0.35926517844200134,
+      "learning_rate": 8.73175560373577e-06,
+      "loss": 2.8722,
+      "step": 33180
+    },
+    {
+      "epoch": 2.4418326619948867,
+      "grad_norm": 0.35706499218940735,
+      "learning_rate": 8.70947183910647e-06,
+      "loss": 2.8793,
+      "step": 33190
+    },
+    {
+      "epoch": 2.4425683753609593,
+      "grad_norm": 0.35748881101608276,
+      "learning_rate": 8.687213832228547e-06,
+      "loss": 2.879,
+      "step": 33200
+    },
+    {
+      "epoch": 2.443304088727032,
+      "grad_norm": 0.3548639118671417,
+      "learning_rate": 8.664981596986931e-06,
+      "loss": 2.8863,
+      "step": 33210
+    },
+    {
+      "epoch": 2.4440398020931045,
+      "grad_norm": 0.36017340421676636,
+      "learning_rate": 8.642775147250498e-06,
+      "loss": 2.8667,
+      "step": 33220
+    },
+    {
+      "epoch": 2.444775515459177,
+      "grad_norm": 0.3464897572994232,
+      "learning_rate": 8.620594496871992e-06,
+      "loss": 2.8502,
+      "step": 33230
+    },
+    {
+      "epoch": 2.4455112288252496,
+      "grad_norm": 0.36990272998809814,
+      "learning_rate": 8.598439659688118e-06,
+      "loss": 2.8242,
+      "step": 33240
+    },
+    {
+      "epoch": 2.446246942191322,
+      "grad_norm": 0.3470519185066223,
+      "learning_rate": 8.576310649519437e-06,
+      "loss": 2.8103,
+      "step": 33250
+    },
+    {
+      "epoch": 2.446982655557395,
+      "grad_norm": 0.36009058356285095,
+      "learning_rate": 8.554207480170395e-06,
+      "loss": 2.8495,
+      "step": 33260
+    },
+    {
+      "epoch": 2.4477183689234674,
+      "grad_norm": 0.35218051075935364,
+      "learning_rate": 8.532130165429364e-06,
+      "loss": 2.8592,
+      "step": 33270
+    },
+    {
+      "epoch": 2.44845408228954,
+      "grad_norm": 0.35970771312713623,
+      "learning_rate": 8.510078719068537e-06,
+      "loss": 2.8323,
+      "step": 33280
+    },
+    {
+      "epoch": 2.4491897956556126,
+      "grad_norm": 0.3691481649875641,
+      "learning_rate": 8.488053154844005e-06,
+      "loss": 2.8425,
+      "step": 33290
+    },
+    {
+      "epoch": 2.449925509021685,
+      "grad_norm": 0.3542400896549225,
+      "learning_rate": 8.466053486495674e-06,
+      "loss": 2.853,
+      "step": 33300
+    },
+    {
+      "epoch": 2.4506612223877577,
+      "grad_norm": 0.34340354800224304,
+      "learning_rate": 8.444079727747339e-06,
+      "loss": 2.86,
+      "step": 33310
+    },
+    {
+      "epoch": 2.4513969357538303,
+      "grad_norm": 0.354559987783432,
+      "learning_rate": 8.422131892306623e-06,
+      "loss": 2.8225,
+      "step": 33320
+    },
+    {
+      "epoch": 2.452132649119903,
+      "grad_norm": 0.3545278012752533,
+      "learning_rate": 8.400209993864961e-06,
+      "loss": 2.838,
+      "step": 33330
+    },
+    {
+      "epoch": 2.4528683624859755,
+      "grad_norm": 0.35428038239479065,
+      "learning_rate": 8.378314046097607e-06,
+      "loss": 2.8114,
+      "step": 33340
+    },
+    {
+      "epoch": 2.453604075852048,
+      "grad_norm": 0.3553619086742401,
+      "learning_rate": 8.356444062663632e-06,
+      "loss": 2.8235,
+      "step": 33350
+    },
+    {
+      "epoch": 2.4543397892181207,
+      "grad_norm": 0.426922470331192,
+      "learning_rate": 8.334600057205932e-06,
+      "loss": 2.8661,
+      "step": 33360
+    },
+    {
+      "epoch": 2.4550755025841933,
+      "grad_norm": 0.35251080989837646,
+      "learning_rate": 8.31278204335117e-06,
+      "loss": 2.8537,
+      "step": 33370
+    },
+    {
+      "epoch": 2.455811215950266,
+      "grad_norm": 0.3673775792121887,
+      "learning_rate": 8.290990034709801e-06,
+      "loss": 2.8479,
+      "step": 33380
+    },
+    {
+      "epoch": 2.4565469293163384,
+      "grad_norm": 0.350271075963974,
+      "learning_rate": 8.269224044876045e-06,
+      "loss": 2.8467,
+      "step": 33390
+    },
+    {
+      "epoch": 2.457282642682411,
+      "grad_norm": 0.354034960269928,
+      "learning_rate": 8.24748408742792e-06,
+      "loss": 2.8755,
+      "step": 33400
+    },
+    {
+      "epoch": 2.4580183560484836,
+      "grad_norm": 0.3478994369506836,
+      "learning_rate": 8.225770175927205e-06,
+      "loss": 2.8179,
+      "step": 33410
+    },
+    {
+      "epoch": 2.458754069414556,
+      "grad_norm": 0.34503114223480225,
+      "learning_rate": 8.204082323919405e-06,
+      "loss": 2.8029,
+      "step": 33420
+    },
+    {
+      "epoch": 2.459489782780629,
+      "grad_norm": 0.36057549715042114,
+      "learning_rate": 8.182420544933777e-06,
+      "loss": 2.8476,
+      "step": 33430
+    },
+    {
+      "epoch": 2.4602254961467014,
+      "grad_norm": 0.36730721592903137,
+      "learning_rate": 8.160784852483306e-06,
+      "loss": 2.8358,
+      "step": 33440
+    },
+    {
+      "epoch": 2.460961209512774,
+      "grad_norm": 0.3461676239967346,
+      "learning_rate": 8.139175260064736e-06,
+      "loss": 2.8647,
+      "step": 33450
+    },
+    {
+      "epoch": 2.4616969228788466,
+      "grad_norm": 0.35185202956199646,
+      "learning_rate": 8.1175917811585e-06,
+      "loss": 2.8541,
+      "step": 33460
+    },
+    {
+      "epoch": 2.462432636244919,
+      "grad_norm": 0.3578980267047882,
+      "learning_rate": 8.096034429228743e-06,
+      "loss": 2.8747,
+      "step": 33470
+    },
+    {
+      "epoch": 2.4631683496109917,
+      "grad_norm": 0.35509902238845825,
+      "learning_rate": 8.07450321772331e-06,
+      "loss": 2.8328,
+      "step": 33480
+    },
+    {
+      "epoch": 2.4639040629770643,
+      "grad_norm": 0.3586413860321045,
+      "learning_rate": 8.05299816007376e-06,
+      "loss": 2.8558,
+      "step": 33490
+    },
+    {
+      "epoch": 2.4646397763431365,
+      "grad_norm": 0.35781750082969666,
+      "learning_rate": 8.031519269695326e-06,
+      "loss": 2.889,
+      "step": 33500
+    },
+    {
+      "epoch": 2.465375489709209,
+      "grad_norm": 0.3559013307094574,
+      "learning_rate": 8.010066559986907e-06,
+      "loss": 2.8148,
+      "step": 33510
+    },
+    {
+      "epoch": 2.4661112030752816,
+      "grad_norm": 0.3545481562614441,
+      "learning_rate": 7.98864004433108e-06,
+      "loss": 2.8276,
+      "step": 33520
+    },
+    {
+      "epoch": 2.4668469164413542,
+      "grad_norm": 0.3552219867706299,
+      "learning_rate": 7.967239736094063e-06,
+      "loss": 2.8752,
+      "step": 33530
+    },
+    {
+      "epoch": 2.467582629807427,
+      "grad_norm": 0.35747528076171875,
+      "learning_rate": 7.945865648625766e-06,
+      "loss": 2.814,
+      "step": 33540
+    },
+    {
+      "epoch": 2.4683183431734994,
+      "grad_norm": 0.35849204659461975,
+      "learning_rate": 7.924517795259712e-06,
+      "loss": 2.8512,
+      "step": 33550
+    },
+    {
+      "epoch": 2.469054056539572,
+      "grad_norm": 0.3469422161579132,
+      "learning_rate": 7.90319618931304e-06,
+      "loss": 2.8642,
+      "step": 33560
+    },
+    {
+      "epoch": 2.4697897699056446,
+      "grad_norm": 0.35371389985084534,
+      "learning_rate": 7.88190084408657e-06,
+      "loss": 2.8414,
+      "step": 33570
+    },
+    {
+      "epoch": 2.470525483271717,
+      "grad_norm": 0.3586072325706482,
+      "learning_rate": 7.860631772864691e-06,
+      "loss": 2.8152,
+      "step": 33580
+    },
+    {
+      "epoch": 2.4712611966377898,
+      "grad_norm": 0.349292129278183,
+      "learning_rate": 7.839388988915435e-06,
+      "loss": 2.8546,
+      "step": 33590
+    },
+    {
+      "epoch": 2.4719969100038623,
+      "grad_norm": 0.35836923122406006,
+      "learning_rate": 7.818172505490417e-06,
+      "loss": 2.8444,
+      "step": 33600
+    },
+    {
+      "epoch": 2.472732623369935,
+      "grad_norm": 0.3604229986667633,
+      "learning_rate": 7.796982335824842e-06,
+      "loss": 2.8742,
+      "step": 33610
+    },
+    {
+      "epoch": 2.4734683367360075,
+      "grad_norm": 0.346037894487381,
+      "learning_rate": 7.775818493137498e-06,
+      "loss": 2.8481,
+      "step": 33620
+    },
+    {
+      "epoch": 2.47420405010208,
+      "grad_norm": 0.35516247153282166,
+      "learning_rate": 7.754680990630791e-06,
+      "loss": 2.8911,
+      "step": 33630
+    },
+    {
+      "epoch": 2.4749397634681527,
+      "grad_norm": 0.35092487931251526,
+      "learning_rate": 7.73356984149063e-06,
+      "loss": 2.8367,
+      "step": 33640
+    },
+    {
+      "epoch": 2.4756754768342253,
+      "grad_norm": 0.35850825905799866,
+      "learning_rate": 7.712485058886537e-06,
+      "loss": 2.8775,
+      "step": 33650
+    },
+    {
+      "epoch": 2.476411190200298,
+      "grad_norm": 0.35583341121673584,
+      "learning_rate": 7.691426655971568e-06,
+      "loss": 2.8496,
+      "step": 33660
+    },
+    {
+      "epoch": 2.4771469035663705,
+      "grad_norm": 0.35149475932121277,
+      "learning_rate": 7.670394645882306e-06,
+      "loss": 2.8427,
+      "step": 33670
+    },
+    {
+      "epoch": 2.477882616932443,
+      "grad_norm": 0.3571380376815796,
+      "learning_rate": 7.649389041738902e-06,
+      "loss": 2.8168,
+      "step": 33680
+    },
+    {
+      "epoch": 2.4786183302985156,
+      "grad_norm": 0.3581167757511139,
+      "learning_rate": 7.628409856645019e-06,
+      "loss": 2.8937,
+      "step": 33690
+    },
+    {
+      "epoch": 2.4793540436645882,
+      "grad_norm": 0.35807880759239197,
+      "learning_rate": 7.607457103687826e-06,
+      "loss": 2.7906,
+      "step": 33700
+    },
+    {
+      "epoch": 2.480089757030661,
+      "grad_norm": 0.3539064824581146,
+      "learning_rate": 7.586530795938013e-06,
+      "loss": 2.8437,
+      "step": 33710
+    },
+    {
+      "epoch": 2.4808254703967334,
+      "grad_norm": 0.35453566908836365,
+      "learning_rate": 7.565630946449792e-06,
+      "loss": 2.8483,
+      "step": 33720
+    },
+    {
+      "epoch": 2.481561183762806,
+      "grad_norm": 0.3582206666469574,
+      "learning_rate": 7.544757568260835e-06,
+      "loss": 2.8379,
+      "step": 33730
+    },
+    {
+      "epoch": 2.4822968971288786,
+      "grad_norm": 0.3597036898136139,
+      "learning_rate": 7.523910674392343e-06,
+      "loss": 2.8668,
+      "step": 33740
+    },
+    {
+      "epoch": 2.483032610494951,
+      "grad_norm": 0.3531782627105713,
+      "learning_rate": 7.503090277848956e-06,
+      "loss": 2.8478,
+      "step": 33750
+    },
+    {
+      "epoch": 2.4837683238610238,
+      "grad_norm": 0.3603537678718567,
+      "learning_rate": 7.482296391618793e-06,
+      "loss": 2.9041,
+      "step": 33760
+    },
+    {
+      "epoch": 2.4845040372270963,
+      "grad_norm": 0.3523320257663727,
+      "learning_rate": 7.461529028673464e-06,
+      "loss": 2.833,
+      "step": 33770
+    },
+    {
+      "epoch": 2.485239750593169,
+      "grad_norm": 0.3563089668750763,
+      "learning_rate": 7.440788201968002e-06,
+      "loss": 2.8711,
+      "step": 33780
+    },
+    {
+      "epoch": 2.4859754639592415,
+      "grad_norm": 0.3472355008125305,
+      "learning_rate": 7.42007392444089e-06,
+      "loss": 2.8081,
+      "step": 33790
+    },
+    {
+      "epoch": 2.486711177325314,
+      "grad_norm": 0.35880446434020996,
+      "learning_rate": 7.399386209014053e-06,
+      "loss": 2.8468,
+      "step": 33800
+    },
+    {
+      "epoch": 2.4874468906913867,
+      "grad_norm": 0.3559292256832123,
+      "learning_rate": 7.378725068592851e-06,
+      "loss": 2.8526,
+      "step": 33810
+    },
+    {
+      "epoch": 2.4881826040574593,
+      "grad_norm": 0.3569248914718628,
+      "learning_rate": 7.358090516066085e-06,
+      "loss": 2.8613,
+      "step": 33820
+    },
+    {
+      "epoch": 2.488918317423532,
+      "grad_norm": 0.34992220997810364,
+      "learning_rate": 7.337482564305931e-06,
+      "loss": 2.8157,
+      "step": 33830
+    },
+    {
+      "epoch": 2.4896540307896045,
+      "grad_norm": 0.3485108017921448,
+      "learning_rate": 7.316901226167988e-06,
+      "loss": 2.8956,
+      "step": 33840
+    },
+    {
+      "epoch": 2.490389744155677,
+      "grad_norm": 0.36055782437324524,
+      "learning_rate": 7.296346514491248e-06,
+      "loss": 2.889,
+      "step": 33850
+    },
+    {
+      "epoch": 2.4911254575217496,
+      "grad_norm": 0.35142526030540466,
+      "learning_rate": 7.275818442098115e-06,
+      "loss": 2.8769,
+      "step": 33860
+    },
+    {
+      "epoch": 2.491861170887822,
+      "grad_norm": 0.35482335090637207,
+      "learning_rate": 7.25531702179435e-06,
+      "loss": 2.8316,
+      "step": 33870
+    },
+    {
+      "epoch": 2.492596884253895,
+      "grad_norm": 0.3710637390613556,
+      "learning_rate": 7.234842266369091e-06,
+      "loss": 2.8599,
+      "step": 33880
+    },
+    {
+      "epoch": 2.4933325976199674,
+      "grad_norm": 0.3554149568080902,
+      "learning_rate": 7.214394188594836e-06,
+      "loss": 2.8338,
+      "step": 33890
+    },
+    {
+      "epoch": 2.49406831098604,
+      "grad_norm": 0.34676870703697205,
+      "learning_rate": 7.1939728012274885e-06,
+      "loss": 2.8312,
+      "step": 33900
+    },
+    {
+      "epoch": 2.4948040243521126,
+      "grad_norm": 0.34753406047821045,
+      "learning_rate": 7.173578117006246e-06,
+      "loss": 2.8115,
+      "step": 33910
+    },
+    {
+      "epoch": 2.495539737718185,
+      "grad_norm": 0.3590540587902069,
+      "learning_rate": 7.153210148653672e-06,
+      "loss": 2.8689,
+      "step": 33920
+    },
+    {
+      "epoch": 2.4962754510842577,
+      "grad_norm": 0.3534415066242218,
+      "learning_rate": 7.132868908875639e-06,
+      "loss": 2.8431,
+      "step": 33930
+    },
+    {
+      "epoch": 2.4970111644503303,
+      "grad_norm": 0.34907054901123047,
+      "learning_rate": 7.112554410361399e-06,
+      "loss": 2.8463,
+      "step": 33940
+    },
+    {
+      "epoch": 2.497746877816403,
+      "grad_norm": 0.34873369336128235,
+      "learning_rate": 7.092266665783475e-06,
+      "loss": 2.8442,
+      "step": 33950
+    },
+    {
+      "epoch": 2.4984825911824755,
+      "grad_norm": 0.35083550214767456,
+      "learning_rate": 7.072005687797722e-06,
+      "loss": 2.8308,
+      "step": 33960
+    },
+    {
+      "epoch": 2.499218304548548,
+      "grad_norm": 0.3604968786239624,
+      "learning_rate": 7.051771489043268e-06,
+      "loss": 2.8345,
+      "step": 33970
+    },
+    {
+      "epoch": 2.4999540179146207,
+      "grad_norm": 0.3432743549346924,
+      "learning_rate": 7.031564082142583e-06,
+      "loss": 2.8563,
+      "step": 33980
+    },
+    {
+      "epoch": 2.5006897312806933,
+      "grad_norm": 0.35237976908683777,
+      "learning_rate": 7.0113834797014e-06,
+      "loss": 2.8729,
+      "step": 33990
+    },
+    {
+      "epoch": 2.501425444646766,
+      "grad_norm": 0.35592156648635864,
+      "learning_rate": 6.991229694308726e-06,
+      "loss": 2.8646,
+      "step": 34000
+    },
+    {
+      "epoch": 2.5021611580128384,
+      "grad_norm": 0.353193461894989,
+      "learning_rate": 6.971102738536844e-06,
+      "loss": 2.8168,
+      "step": 34010
+    },
+    {
+      "epoch": 2.502896871378911,
+      "grad_norm": 0.34972190856933594,
+      "learning_rate": 6.951002624941283e-06,
+      "loss": 2.8141,
+      "step": 34020
+    },
+    {
+      "epoch": 2.5036325847449836,
+      "grad_norm": 0.3629728853702545,
+      "learning_rate": 6.930929366060873e-06,
+      "loss": 2.8965,
+      "step": 34030
+    },
+    {
+      "epoch": 2.5043682981110558,
+      "grad_norm": 0.36246904730796814,
+      "learning_rate": 6.910882974417654e-06,
+      "loss": 2.8283,
+      "step": 34040
+    },
+    {
+      "epoch": 2.5051040114771284,
+      "grad_norm": 0.36388489603996277,
+      "learning_rate": 6.8908634625169054e-06,
+      "loss": 2.8347,
+      "step": 34050
+    },
+    {
+      "epoch": 2.505839724843201,
+      "grad_norm": 0.346941202878952,
+      "learning_rate": 6.870870842847138e-06,
+      "loss": 2.8397,
+      "step": 34060
+    },
+    {
+      "epoch": 2.5065754382092735,
+      "grad_norm": 0.3567986786365509,
+      "learning_rate": 6.850905127880114e-06,
+      "loss": 2.8835,
+      "step": 34070
+    },
+    {
+      "epoch": 2.507311151575346,
+      "grad_norm": 0.34916162490844727,
+      "learning_rate": 6.8309663300708034e-06,
+      "loss": 2.8595,
+      "step": 34080
+    },
+    {
+      "epoch": 2.5080468649414187,
+      "grad_norm": 0.35004734992980957,
+      "learning_rate": 6.811054461857363e-06,
+      "loss": 2.818,
+      "step": 34090
+    },
+    {
+      "epoch": 2.5087825783074913,
+      "grad_norm": 0.35771462321281433,
+      "learning_rate": 6.791169535661163e-06,
+      "loss": 2.8237,
+      "step": 34100
+    },
+    {
+      "epoch": 2.509518291673564,
+      "grad_norm": 0.3517802357673645,
+      "learning_rate": 6.771311563886762e-06,
+      "loss": 2.8335,
+      "step": 34110
+    },
+    {
+      "epoch": 2.5102540050396365,
+      "grad_norm": 0.35630613565444946,
+      "learning_rate": 6.751480558921919e-06,
+      "loss": 2.8567,
+      "step": 34120
+    },
+    {
+      "epoch": 2.510989718405709,
+      "grad_norm": 0.35152190923690796,
+      "learning_rate": 6.73167653313756e-06,
+      "loss": 2.8411,
+      "step": 34130
+    },
+    {
+      "epoch": 2.5117254317717816,
+      "grad_norm": 0.35704168677330017,
+      "learning_rate": 6.711899498887769e-06,
+      "loss": 2.8455,
+      "step": 34140
+    },
+    {
+      "epoch": 2.5124611451378542,
+      "grad_norm": 0.3544277846813202,
+      "learning_rate": 6.6921494685098284e-06,
+      "loss": 2.8358,
+      "step": 34150
+    },
+    {
+      "epoch": 2.513196858503927,
+      "grad_norm": 0.34802743792533875,
+      "learning_rate": 6.672426454324121e-06,
+      "loss": 2.8644,
+      "step": 34160
+    },
+    {
+      "epoch": 2.5139325718699994,
+      "grad_norm": 0.357770174741745,
+      "learning_rate": 6.652730468634233e-06,
+      "loss": 2.8581,
+      "step": 34170
+    },
+    {
+      "epoch": 2.514668285236072,
+      "grad_norm": 0.34689316153526306,
+      "learning_rate": 6.633061523726858e-06,
+      "loss": 2.8216,
+      "step": 34180
+    },
+    {
+      "epoch": 2.5154039986021446,
+      "grad_norm": 0.3538762331008911,
+      "learning_rate": 6.61341963187182e-06,
+      "loss": 2.8535,
+      "step": 34190
+    },
+    {
+      "epoch": 2.516139711968217,
+      "grad_norm": 0.354240357875824,
+      "learning_rate": 6.593804805322068e-06,
+      "loss": 2.8544,
+      "step": 34200
+    },
+    {
+      "epoch": 2.5168754253342898,
+      "grad_norm": 0.3550654351711273,
+      "learning_rate": 6.574217056313692e-06,
+      "loss": 2.8655,
+      "step": 34210
+    },
+    {
+      "epoch": 2.5176111387003623,
+      "grad_norm": 0.36292973160743713,
+      "learning_rate": 6.55465639706585e-06,
+      "loss": 2.8883,
+      "step": 34220
+    },
+    {
+      "epoch": 2.518346852066435,
+      "grad_norm": 0.3598898947238922,
+      "learning_rate": 6.535122839780844e-06,
+      "loss": 2.8427,
+      "step": 34230
+    },
+    {
+      "epoch": 2.5190825654325075,
+      "grad_norm": 0.35678303241729736,
+      "learning_rate": 6.515616396644031e-06,
+      "loss": 2.8614,
+      "step": 34240
+    },
+    {
+      "epoch": 2.51981827879858,
+      "grad_norm": 0.3484988808631897,
+      "learning_rate": 6.4961370798238716e-06,
+      "loss": 2.8178,
+      "step": 34250
+    },
+    {
+      "epoch": 2.5205539921646527,
+      "grad_norm": 0.3532414138317108,
+      "learning_rate": 6.476684901471913e-06,
+      "loss": 2.8743,
+      "step": 34260
+    },
+    {
+      "epoch": 2.5212897055307253,
+      "grad_norm": 0.35128718614578247,
+      "learning_rate": 6.457259873722765e-06,
+      "loss": 2.8435,
+      "step": 34270
+    },
+    {
+      "epoch": 2.522025418896798,
+      "grad_norm": 0.3568994402885437,
+      "learning_rate": 6.437862008694084e-06,
+      "loss": 2.8576,
+      "step": 34280
+    },
+    {
+      "epoch": 2.5227611322628705,
+      "grad_norm": 0.35171395540237427,
+      "learning_rate": 6.418491318486597e-06,
+      "loss": 2.8779,
+      "step": 34290
+    },
+    {
+      "epoch": 2.523496845628943,
+      "grad_norm": 0.3578436076641083,
+      "learning_rate": 6.399147815184081e-06,
+      "loss": 2.8838,
+      "step": 34300
+    },
+    {
+      "epoch": 2.5242325589950156,
+      "grad_norm": 0.356166809797287,
+      "learning_rate": 6.3798315108533675e-06,
+      "loss": 2.8561,
+      "step": 34310
+    },
+    {
+      "epoch": 2.5249682723610882,
+      "grad_norm": 0.35286077857017517,
+      "learning_rate": 6.360542417544291e-06,
+      "loss": 2.8507,
+      "step": 34320
+    },
+    {
+      "epoch": 2.525703985727161,
+      "grad_norm": 0.35248029232025146,
+      "learning_rate": 6.341280547289724e-06,
+      "loss": 2.8506,
+      "step": 34330
+    },
+    {
+      "epoch": 2.5264396990932334,
+      "grad_norm": 0.3513822853565216,
+      "learning_rate": 6.322045912105546e-06,
+      "loss": 2.7811,
+      "step": 34340
+    },
+    {
+      "epoch": 2.527175412459306,
+      "grad_norm": 0.34720301628112793,
+      "learning_rate": 6.302838523990684e-06,
+      "loss": 2.8369,
+      "step": 34350
+    },
+    {
+      "epoch": 2.5279111258253786,
+      "grad_norm": 0.3542155623435974,
+      "learning_rate": 6.283658394927023e-06,
+      "loss": 2.8555,
+      "step": 34360
+    },
+    {
+      "epoch": 2.528646839191451,
+      "grad_norm": 0.35349902510643005,
+      "learning_rate": 6.264505536879461e-06,
+      "loss": 2.8163,
+      "step": 34370
+    },
+    {
+      "epoch": 2.5293825525575233,
+      "grad_norm": 0.3546314239501953,
+      "learning_rate": 6.245379961795888e-06,
+      "loss": 2.822,
+      "step": 34380
+    },
+    {
+      "epoch": 2.530118265923596,
+      "grad_norm": 0.3481293022632599,
+      "learning_rate": 6.226281681607166e-06,
+      "loss": 2.879,
+      "step": 34390
+    },
+    {
+      "epoch": 2.5308539792896685,
+      "grad_norm": 0.36185672879219055,
+      "learning_rate": 6.207210708227151e-06,
+      "loss": 2.8467,
+      "step": 34400
+    },
+    {
+      "epoch": 2.531589692655741,
+      "grad_norm": 0.35591915249824524,
+      "learning_rate": 6.188167053552646e-06,
+      "loss": 2.846,
+      "step": 34410
+    },
+    {
+      "epoch": 2.5323254060218137,
+      "grad_norm": 0.35439935326576233,
+      "learning_rate": 6.169150729463402e-06,
+      "loss": 2.7981,
+      "step": 34420
+    },
+    {
+      "epoch": 2.5330611193878863,
+      "grad_norm": 0.3485656678676605,
+      "learning_rate": 6.1501617478221304e-06,
+      "loss": 2.8391,
+      "step": 34430
+    },
+    {
+      "epoch": 2.533796832753959,
+      "grad_norm": 0.3532913625240326,
+      "learning_rate": 6.1312001204745115e-06,
+      "loss": 2.8392,
+      "step": 34440
+    },
+    {
+      "epoch": 2.5345325461200314,
+      "grad_norm": 0.3537358343601227,
+      "learning_rate": 6.1122658592491275e-06,
+      "loss": 2.8468,
+      "step": 34450
+    },
+    {
+      "epoch": 2.535268259486104,
+      "grad_norm": 0.3611201047897339,
+      "learning_rate": 6.093358975957497e-06,
+      "loss": 2.8461,
+      "step": 34460
+    },
+    {
+      "epoch": 2.5360039728521766,
+      "grad_norm": 0.35566937923431396,
+      "learning_rate": 6.074479482394052e-06,
+      "loss": 2.8197,
+      "step": 34470
+    },
+    {
+      "epoch": 2.536739686218249,
+      "grad_norm": 0.3534085750579834,
+      "learning_rate": 6.055627390336166e-06,
+      "loss": 2.8578,
+      "step": 34480
+    },
+    {
+      "epoch": 2.537475399584322,
+      "grad_norm": 0.34801584482192993,
+      "learning_rate": 6.036802711544109e-06,
+      "loss": 2.8191,
+      "step": 34490
+    },
+    {
+      "epoch": 2.5382111129503944,
+      "grad_norm": 0.3492884039878845,
+      "learning_rate": 6.018005457761033e-06,
+      "loss": 2.8369,
+      "step": 34500
+    },
+    {
+      "epoch": 2.538946826316467,
+      "grad_norm": 0.35288798809051514,
+      "learning_rate": 5.9992356407129965e-06,
+      "loss": 2.8572,
+      "step": 34510
+    },
+    {
+      "epoch": 2.5396825396825395,
+      "grad_norm": 0.3575350344181061,
+      "learning_rate": 5.980493272108928e-06,
+      "loss": 2.8842,
+      "step": 34520
+    },
+    {
+      "epoch": 2.540418253048612,
+      "grad_norm": 0.35145509243011475,
+      "learning_rate": 5.961778363640663e-06,
+      "loss": 2.8809,
+      "step": 34530
+    },
+    {
+      "epoch": 2.5411539664146847,
+      "grad_norm": 0.3523208796977997,
+      "learning_rate": 5.943090926982875e-06,
+      "loss": 2.8671,
+      "step": 34540
+    },
+    {
+      "epoch": 2.5418896797807573,
+      "grad_norm": 0.3611590564250946,
+      "learning_rate": 5.924430973793105e-06,
+      "loss": 2.8315,
+      "step": 34550
+    },
+    {
+      "epoch": 2.54262539314683,
+      "grad_norm": 0.36057910323143005,
+      "learning_rate": 5.9057985157117805e-06,
+      "loss": 2.8757,
+      "step": 34560
+    },
+    {
+      "epoch": 2.5433611065129025,
+      "grad_norm": 0.3599465489387512,
+      "learning_rate": 5.887193564362125e-06,
+      "loss": 2.8861,
+      "step": 34570
+    },
+    {
+      "epoch": 2.544096819878975,
+      "grad_norm": 0.35537534952163696,
+      "learning_rate": 5.868616131350258e-06,
+      "loss": 2.8919,
+      "step": 34580
+    },
+    {
+      "epoch": 2.5448325332450477,
+      "grad_norm": 0.3627186119556427,
+      "learning_rate": 5.8500662282651e-06,
+      "loss": 2.8546,
+      "step": 34590
+    },
+    {
+      "epoch": 2.5455682466111202,
+      "grad_norm": 0.3607963025569916,
+      "learning_rate": 5.831543866678391e-06,
+      "loss": 2.9156,
+      "step": 34600
+    },
+    {
+      "epoch": 2.546303959977193,
+      "grad_norm": 0.3559984564781189,
+      "learning_rate": 5.81304905814471e-06,
+      "loss": 2.8388,
+      "step": 34610
+    },
+    {
+      "epoch": 2.5470396733432654,
+      "grad_norm": 0.3525046408176422,
+      "learning_rate": 5.794581814201444e-06,
+      "loss": 2.8785,
+      "step": 34620
+    },
+    {
+      "epoch": 2.547775386709338,
+      "grad_norm": 0.35862019658088684,
+      "learning_rate": 5.7761421463687734e-06,
+      "loss": 2.8265,
+      "step": 34630
+    },
+    {
+      "epoch": 2.5485111000754106,
+      "grad_norm": 0.364872008562088,
+      "learning_rate": 5.7577300661497024e-06,
+      "loss": 2.843,
+      "step": 34640
+    },
+    {
+      "epoch": 2.549246813441483,
+      "grad_norm": 0.35501405596733093,
+      "learning_rate": 5.7393455850299925e-06,
+      "loss": 2.8176,
+      "step": 34650
+    },
+    {
+      "epoch": 2.5499825268075558,
+      "grad_norm": 0.34877753257751465,
+      "learning_rate": 5.720988714478198e-06,
+      "loss": 2.9149,
+      "step": 34660
+    },
+    {
+      "epoch": 2.5507182401736284,
+      "grad_norm": 0.35799792408943176,
+      "learning_rate": 5.702659465945675e-06,
+      "loss": 2.8544,
+      "step": 34670
+    },
+    {
+      "epoch": 2.551453953539701,
+      "grad_norm": 0.36026081442832947,
+      "learning_rate": 5.684357850866518e-06,
+      "loss": 2.8801,
+      "step": 34680
+    },
+    {
+      "epoch": 2.5521896669057735,
+      "grad_norm": 0.3508109748363495,
+      "learning_rate": 5.666083880657591e-06,
+      "loss": 2.8262,
+      "step": 34690
+    },
+    {
+      "epoch": 2.552925380271846,
+      "grad_norm": 0.34675076603889465,
+      "learning_rate": 5.6478375667185055e-06,
+      "loss": 2.832,
+      "step": 34700
+    },
+    {
+      "epoch": 2.5536610936379187,
+      "grad_norm": 0.36108505725860596,
+      "learning_rate": 5.629618920431645e-06,
+      "loss": 2.8735,
+      "step": 34710
+    },
+    {
+      "epoch": 2.5543968070039913,
+      "grad_norm": 0.3439852297306061,
+      "learning_rate": 5.611427953162129e-06,
+      "loss": 2.8248,
+      "step": 34720
+    },
+    {
+      "epoch": 2.555132520370064,
+      "grad_norm": 0.3634498715400696,
+      "learning_rate": 5.593264676257781e-06,
+      "loss": 2.8226,
+      "step": 34730
+    },
+    {
+      "epoch": 2.5558682337361365,
+      "grad_norm": 0.35187816619873047,
+      "learning_rate": 5.57512910104917e-06,
+      "loss": 2.8044,
+      "step": 34740
+    },
+    {
+      "epoch": 2.556603947102209,
+      "grad_norm": 0.346326619386673,
+      "learning_rate": 5.5570212388495966e-06,
+      "loss": 2.832,
+      "step": 34750
+    },
+    {
+      "epoch": 2.5573396604682817,
+      "grad_norm": 0.35494983196258545,
+      "learning_rate": 5.5389411009550526e-06,
+      "loss": 2.8146,
+      "step": 34760
+    },
+    {
+      "epoch": 2.5580753738343542,
+      "grad_norm": 0.35779476165771484,
+      "learning_rate": 5.520888698644244e-06,
+      "loss": 2.8967,
+      "step": 34770
+    },
+    {
+      "epoch": 2.558811087200427,
+      "grad_norm": 0.3533632159233093,
+      "learning_rate": 5.5028640431785795e-06,
+      "loss": 2.8346,
+      "step": 34780
+    },
+    {
+      "epoch": 2.5595468005664994,
+      "grad_norm": 0.3545354902744293,
+      "learning_rate": 5.484867145802136e-06,
+      "loss": 2.8558,
+      "step": 34790
+    },
+    {
+      "epoch": 2.560282513932572,
+      "grad_norm": 0.3570384681224823,
+      "learning_rate": 5.466898017741701e-06,
+      "loss": 2.8193,
+      "step": 34800
+    },
+    {
+      "epoch": 2.5610182272986446,
+      "grad_norm": 0.35139229893684387,
+      "learning_rate": 5.448956670206751e-06,
+      "loss": 2.8775,
+      "step": 34810
+    },
+    {
+      "epoch": 2.561753940664717,
+      "grad_norm": 0.3476870357990265,
+      "learning_rate": 5.431043114389389e-06,
+      "loss": 2.837,
+      "step": 34820
+    },
+    {
+      "epoch": 2.5624896540307898,
+      "grad_norm": 0.3604823350906372,
+      "learning_rate": 5.413157361464394e-06,
+      "loss": 2.8169,
+      "step": 34830
+    },
+    {
+      "epoch": 2.5632253673968624,
+      "grad_norm": 0.36214837431907654,
+      "learning_rate": 5.39529942258924e-06,
+      "loss": 2.8312,
+      "step": 34840
+    },
+    {
+      "epoch": 2.563961080762935,
+      "grad_norm": 0.361138254404068,
+      "learning_rate": 5.377469308904004e-06,
+      "loss": 2.8747,
+      "step": 34850
+    },
+    {
+      "epoch": 2.5646967941290075,
+      "grad_norm": 0.34585121273994446,
+      "learning_rate": 5.3596670315314234e-06,
+      "loss": 2.8398,
+      "step": 34860
+    },
+    {
+      "epoch": 2.56543250749508,
+      "grad_norm": 0.3549595773220062,
+      "learning_rate": 5.341892601576859e-06,
+      "loss": 2.8673,
+      "step": 34870
+    },
+    {
+      "epoch": 2.5661682208611527,
+      "grad_norm": 0.3526514172554016,
+      "learning_rate": 5.3241460301283245e-06,
+      "loss": 2.8279,
+      "step": 34880
+    },
+    {
+      "epoch": 2.5669039342272253,
+      "grad_norm": 0.3528748154640198,
+      "learning_rate": 5.3064273282564415e-06,
+      "loss": 2.8065,
+      "step": 34890
+    },
+    {
+      "epoch": 2.567639647593298,
+      "grad_norm": 0.35635823011398315,
+      "learning_rate": 5.288736507014435e-06,
+      "loss": 2.8199,
+      "step": 34900
+    },
+    {
+      "epoch": 2.5683753609593705,
+      "grad_norm": 0.3525192439556122,
+      "learning_rate": 5.271073577438163e-06,
+      "loss": 2.8309,
+      "step": 34910
+    },
+    {
+      "epoch": 2.569111074325443,
+      "grad_norm": 0.3572302460670471,
+      "learning_rate": 5.253438550546041e-06,
+      "loss": 2.8257,
+      "step": 34920
+    },
+    {
+      "epoch": 2.5698467876915156,
+      "grad_norm": 0.35250580310821533,
+      "learning_rate": 5.235831437339134e-06,
+      "loss": 2.8665,
+      "step": 34930
+    },
+    {
+      "epoch": 2.5705825010575882,
+      "grad_norm": 0.35535407066345215,
+      "learning_rate": 5.218252248801059e-06,
+      "loss": 2.8036,
+      "step": 34940
+    },
+    {
+      "epoch": 2.5713182144236604,
+      "grad_norm": 0.35798943042755127,
+      "learning_rate": 5.200700995898017e-06,
+      "loss": 2.8775,
+      "step": 34950
+    },
+    {
+      "epoch": 2.572053927789733,
+      "grad_norm": 0.35482558608055115,
+      "learning_rate": 5.183177689578772e-06,
+      "loss": 2.8847,
+      "step": 34960
+    },
+    {
+      "epoch": 2.5727896411558056,
+      "grad_norm": 0.3632205128669739,
+      "learning_rate": 5.1656823407746856e-06,
+      "loss": 2.8622,
+      "step": 34970
+    },
+    {
+      "epoch": 2.573525354521878,
+      "grad_norm": 0.34744560718536377,
+      "learning_rate": 5.148214960399667e-06,
+      "loss": 2.7721,
+      "step": 34980
+    },
+    {
+      "epoch": 2.5742610678879507,
+      "grad_norm": 0.3517625331878662,
+      "learning_rate": 5.130775559350159e-06,
+      "loss": 2.852,
+      "step": 34990
+    },
+    {
+      "epoch": 2.5749967812540233,
+      "grad_norm": 0.3417135775089264,
+      "learning_rate": 5.113364148505173e-06,
+      "loss": 2.8451,
+      "step": 35000
+    },
+    {
+      "epoch": 2.575732494620096,
+      "grad_norm": 0.358468621969223,
+      "learning_rate": 5.095980738726236e-06,
+      "loss": 2.833,
+      "step": 35010
+    },
+    {
+      "epoch": 2.5764682079861685,
+      "grad_norm": 0.35115501284599304,
+      "learning_rate": 5.07862534085744e-06,
+      "loss": 2.8457,
+      "step": 35020
+    },
+    {
+      "epoch": 2.577203921352241,
+      "grad_norm": 0.3552725911140442,
+      "learning_rate": 5.061297965725376e-06,
+      "loss": 2.7974,
+      "step": 35030
+    },
+    {
+      "epoch": 2.5779396347183137,
+      "grad_norm": 0.34344372153282166,
+      "learning_rate": 5.043998624139157e-06,
+      "loss": 2.8752,
+      "step": 35040
+    },
+    {
+      "epoch": 2.5786753480843863,
+      "grad_norm": 0.3491573631763458,
+      "learning_rate": 5.026727326890429e-06,
+      "loss": 2.877,
+      "step": 35050
+    },
+    {
+      "epoch": 2.579411061450459,
+      "grad_norm": 0.3492386043071747,
+      "learning_rate": 5.009484084753307e-06,
+      "loss": 2.8435,
+      "step": 35060
+    },
+    {
+      "epoch": 2.5801467748165314,
+      "grad_norm": 0.3548882007598877,
+      "learning_rate": 4.9922689084844545e-06,
+      "loss": 2.8692,
+      "step": 35070
+    },
+    {
+      "epoch": 2.580882488182604,
+      "grad_norm": 0.3555954098701477,
+      "learning_rate": 4.975081808822979e-06,
+      "loss": 2.8763,
+      "step": 35080
+    },
+    {
+      "epoch": 2.5816182015486766,
+      "grad_norm": 0.3637039363384247,
+      "learning_rate": 4.957922796490494e-06,
+      "loss": 2.8614,
+      "step": 35090
+    },
+    {
+      "epoch": 2.582353914914749,
+      "grad_norm": 0.3520558476448059,
+      "learning_rate": 4.9407918821910874e-06,
+      "loss": 2.839,
+      "step": 35100
+    },
+    {
+      "epoch": 2.583089628280822,
+      "grad_norm": 0.3606053590774536,
+      "learning_rate": 4.923689076611337e-06,
+      "loss": 2.8614,
+      "step": 35110
+    },
+    {
+      "epoch": 2.5838253416468944,
+      "grad_norm": 0.3474246859550476,
+      "learning_rate": 4.906614390420267e-06,
+      "loss": 2.8078,
+      "step": 35120
+    },
+    {
+      "epoch": 2.584561055012967,
+      "grad_norm": 0.35836145281791687,
+      "learning_rate": 4.889567834269349e-06,
+      "loss": 2.843,
+      "step": 35130
+    },
+    {
+      "epoch": 2.5852967683790395,
+      "grad_norm": 0.3623798191547394,
+      "learning_rate": 4.872549418792538e-06,
+      "loss": 2.8318,
+      "step": 35140
+    },
+    {
+      "epoch": 2.586032481745112,
+      "grad_norm": 0.3576414883136749,
+      "learning_rate": 4.855559154606204e-06,
+      "loss": 2.8373,
+      "step": 35150
+    },
+    {
+      "epoch": 2.5867681951111847,
+      "grad_norm": 0.35372403264045715,
+      "learning_rate": 4.83859705230919e-06,
+      "loss": 2.8639,
+      "step": 35160
+    },
+    {
+      "epoch": 2.5875039084772573,
+      "grad_norm": 0.3555085361003876,
+      "learning_rate": 4.821663122482739e-06,
+      "loss": 2.8514,
+      "step": 35170
+    },
+    {
+      "epoch": 2.58823962184333,
+      "grad_norm": 0.35871753096580505,
+      "learning_rate": 4.804757375690533e-06,
+      "loss": 2.7789,
+      "step": 35180
+    },
+    {
+      "epoch": 2.5889753352094025,
+      "grad_norm": 0.34866079688072205,
+      "learning_rate": 4.787879822478664e-06,
+      "loss": 2.7963,
+      "step": 35190
+    },
+    {
+      "epoch": 2.589711048575475,
+      "grad_norm": 0.3430629372596741,
+      "learning_rate": 4.77103047337566e-06,
+      "loss": 2.7994,
+      "step": 35200
+    },
+    {
+      "epoch": 2.5904467619415477,
+      "grad_norm": 0.3444986641407013,
+      "learning_rate": 4.754209338892424e-06,
+      "loss": 2.8459,
+      "step": 35210
+    },
+    {
+      "epoch": 2.5911824753076202,
+      "grad_norm": 0.3562524914741516,
+      "learning_rate": 4.737416429522285e-06,
+      "loss": 2.8553,
+      "step": 35220
+    },
+    {
+      "epoch": 2.591918188673693,
+      "grad_norm": 0.3516131639480591,
+      "learning_rate": 4.7206517557409454e-06,
+      "loss": 2.844,
+      "step": 35230
+    },
+    {
+      "epoch": 2.5926539020397654,
+      "grad_norm": 0.3560907244682312,
+      "learning_rate": 4.703915328006492e-06,
+      "loss": 2.8665,
+      "step": 35240
+    },
+    {
+      "epoch": 2.593389615405838,
+      "grad_norm": 0.35479164123535156,
+      "learning_rate": 4.687207156759427e-06,
+      "loss": 2.8271,
+      "step": 35250
+    },
+    {
+      "epoch": 2.5941253287719106,
+      "grad_norm": 0.3530963659286499,
+      "learning_rate": 4.670527252422574e-06,
+      "loss": 2.8678,
+      "step": 35260
+    },
+    {
+      "epoch": 2.594861042137983,
+      "grad_norm": 0.35626524686813354,
+      "learning_rate": 4.65387562540116e-06,
+      "loss": 2.8581,
+      "step": 35270
+    },
+    {
+      "epoch": 2.5955967555040558,
+      "grad_norm": 0.3517557680606842,
+      "learning_rate": 4.63725228608275e-06,
+      "loss": 2.812,
+      "step": 35280
+    },
+    {
+      "epoch": 2.5963324688701284,
+      "grad_norm": 0.3578193783760071,
+      "learning_rate": 4.6206572448372785e-06,
+      "loss": 2.8614,
+      "step": 35290
+    },
+    {
+      "epoch": 2.5970681822362005,
+      "grad_norm": 0.35045626759529114,
+      "learning_rate": 4.604090512017034e-06,
+      "loss": 2.8998,
+      "step": 35300
+    },
+    {
+      "epoch": 2.597803895602273,
+      "grad_norm": 0.3596821129322052,
+      "learning_rate": 4.587552097956621e-06,
+      "loss": 2.8448,
+      "step": 35310
+    },
+    {
+      "epoch": 2.5985396089683457,
+      "grad_norm": 0.34645193815231323,
+      "learning_rate": 4.571042012972992e-06,
+      "loss": 2.821,
+      "step": 35320
+    },
+    {
+      "epoch": 2.5992753223344183,
+      "grad_norm": 0.35674211382865906,
+      "learning_rate": 4.5545602673654155e-06,
+      "loss": 2.8092,
+      "step": 35330
+    },
+    {
+      "epoch": 2.600011035700491,
+      "grad_norm": 0.357374370098114,
+      "learning_rate": 4.538106871415509e-06,
+      "loss": 2.8373,
+      "step": 35340
+    },
+    {
+      "epoch": 2.6007467490665634,
+      "grad_norm": 0.35389024019241333,
+      "learning_rate": 4.521681835387187e-06,
+      "loss": 2.8352,
+      "step": 35350
+    },
+    {
+      "epoch": 2.601482462432636,
+      "grad_norm": 0.3545044958591461,
+      "learning_rate": 4.505285169526663e-06,
+      "loss": 2.8816,
+      "step": 35360
+    },
+    {
+      "epoch": 2.6022181757987086,
+      "grad_norm": 0.3558023273944855,
+      "learning_rate": 4.488916884062455e-06,
+      "loss": 2.8968,
+      "step": 35370
+    },
+    {
+      "epoch": 2.602953889164781,
+      "grad_norm": 0.36237186193466187,
+      "learning_rate": 4.472576989205402e-06,
+      "loss": 2.8749,
+      "step": 35380
+    },
+    {
+      "epoch": 2.603689602530854,
+      "grad_norm": 0.3502497375011444,
+      "learning_rate": 4.456265495148615e-06,
+      "loss": 2.8761,
+      "step": 35390
+    },
+    {
+      "epoch": 2.6044253158969264,
+      "grad_norm": 0.3535205125808716,
+      "learning_rate": 4.439982412067489e-06,
+      "loss": 2.8614,
+      "step": 35400
+    },
+    {
+      "epoch": 2.605161029262999,
+      "grad_norm": 0.35819587111473083,
+      "learning_rate": 4.423727750119683e-06,
+      "loss": 2.8328,
+      "step": 35410
+    },
+    {
+      "epoch": 2.6058967426290716,
+      "grad_norm": 0.351230263710022,
+      "learning_rate": 4.407501519445145e-06,
+      "loss": 2.8323,
+      "step": 35420
+    },
+    {
+      "epoch": 2.606632455995144,
+      "grad_norm": 0.358737051486969,
+      "learning_rate": 4.391303730166085e-06,
+      "loss": 2.7975,
+      "step": 35430
+    },
+    {
+      "epoch": 2.6073681693612167,
+      "grad_norm": 0.3439580798149109,
+      "learning_rate": 4.3751343923869595e-06,
+      "loss": 2.8063,
+      "step": 35440
+    },
+    {
+      "epoch": 2.6081038827272893,
+      "grad_norm": 0.3464832305908203,
+      "learning_rate": 4.358993516194487e-06,
+      "loss": 2.8401,
+      "step": 35450
+    },
+    {
+      "epoch": 2.608839596093362,
+      "grad_norm": 0.35420164465904236,
+      "learning_rate": 4.3428811116576204e-06,
+      "loss": 2.8635,
+      "step": 35460
+    },
+    {
+      "epoch": 2.6095753094594345,
+      "grad_norm": 0.3531213700771332,
+      "learning_rate": 4.326797188827564e-06,
+      "loss": 2.8931,
+      "step": 35470
+    },
+    {
+      "epoch": 2.610311022825507,
+      "grad_norm": 0.36589857935905457,
+      "learning_rate": 4.310741757737763e-06,
+      "loss": 2.8523,
+      "step": 35480
+    },
+    {
+      "epoch": 2.6110467361915797,
+      "grad_norm": 0.350562185049057,
+      "learning_rate": 4.2947148284038566e-06,
+      "loss": 2.886,
+      "step": 35490
+    },
+    {
+      "epoch": 2.6117824495576523,
+      "grad_norm": 0.3622821569442749,
+      "learning_rate": 4.27871641082373e-06,
+      "loss": 2.8409,
+      "step": 35500
+    },
+    {
+      "epoch": 2.612518162923725,
+      "grad_norm": 0.3557232916355133,
+      "learning_rate": 4.2627465149774614e-06,
+      "loss": 2.8678,
+      "step": 35510
+    },
+    {
+      "epoch": 2.6132538762897974,
+      "grad_norm": 0.3483336865901947,
+      "learning_rate": 4.246805150827377e-06,
+      "loss": 2.8218,
+      "step": 35520
+    },
+    {
+      "epoch": 2.61398958965587,
+      "grad_norm": 0.3544950485229492,
+      "learning_rate": 4.230892328317959e-06,
+      "loss": 2.855,
+      "step": 35530
+    },
+    {
+      "epoch": 2.6147253030219426,
+      "grad_norm": 0.34842449426651,
+      "learning_rate": 4.215008057375902e-06,
+      "loss": 2.8981,
+      "step": 35540
+    },
+    {
+      "epoch": 2.615461016388015,
+      "grad_norm": 0.34743309020996094,
+      "learning_rate": 4.199152347910102e-06,
+      "loss": 2.829,
+      "step": 35550
+    },
+    {
+      "epoch": 2.616196729754088,
+      "grad_norm": 0.354101300239563,
+      "learning_rate": 4.183325209811617e-06,
+      "loss": 2.8575,
+      "step": 35560
+    },
+    {
+      "epoch": 2.6169324431201604,
+      "grad_norm": 0.35682976245880127,
+      "learning_rate": 4.1675266529537015e-06,
+      "loss": 2.8556,
+      "step": 35570
+    },
+    {
+      "epoch": 2.617668156486233,
+      "grad_norm": 0.35930177569389343,
+      "learning_rate": 4.151756687191766e-06,
+      "loss": 2.8714,
+      "step": 35580
+    },
+    {
+      "epoch": 2.6184038698523056,
+      "grad_norm": 0.35298365354537964,
+      "learning_rate": 4.136015322363396e-06,
+      "loss": 2.8466,
+      "step": 35590
+    },
+    {
+      "epoch": 2.619139583218378,
+      "grad_norm": 0.34881919622421265,
+      "learning_rate": 4.120302568288309e-06,
+      "loss": 2.8784,
+      "step": 35600
+    },
+    {
+      "epoch": 2.6198752965844507,
+      "grad_norm": 0.35236120223999023,
+      "learning_rate": 4.104618434768415e-06,
+      "loss": 2.8747,
+      "step": 35610
+    },
+    {
+      "epoch": 2.6206110099505233,
+      "grad_norm": 0.3512903153896332,
+      "learning_rate": 4.088962931587742e-06,
+      "loss": 2.8614,
+      "step": 35620
+    },
+    {
+      "epoch": 2.621346723316596,
+      "grad_norm": 0.35775357484817505,
+      "learning_rate": 4.073336068512468e-06,
+      "loss": 2.8582,
+      "step": 35630
+    },
+    {
+      "epoch": 2.6220824366826685,
+      "grad_norm": 0.3560003936290741,
+      "learning_rate": 4.0577378552909e-06,
+      "loss": 2.8818,
+      "step": 35640
+    },
+    {
+      "epoch": 2.622818150048741,
+      "grad_norm": 0.355671226978302,
+      "learning_rate": 4.042168301653487e-06,
+      "loss": 2.8766,
+      "step": 35650
+    },
+    {
+      "epoch": 2.6235538634148137,
+      "grad_norm": 0.3599035441875458,
+      "learning_rate": 4.026627417312778e-06,
+      "loss": 2.9171,
+      "step": 35660
+    },
+    {
+      "epoch": 2.6242895767808863,
+      "grad_norm": 0.3468172252178192,
+      "learning_rate": 4.011115211963446e-06,
+      "loss": 2.8357,
+      "step": 35670
+    },
+    {
+      "epoch": 2.625025290146959,
+      "grad_norm": 0.3642551898956299,
+      "learning_rate": 3.99563169528227e-06,
+      "loss": 2.8445,
+      "step": 35680
+    },
+    {
+      "epoch": 2.6257610035130314,
+      "grad_norm": 0.3535473346710205,
+      "learning_rate": 3.980176876928154e-06,
+      "loss": 2.8697,
+      "step": 35690
+    },
+    {
+      "epoch": 2.626496716879104,
+      "grad_norm": 0.3580264747142792,
+      "learning_rate": 3.964750766542069e-06,
+      "loss": 2.8397,
+      "step": 35700
+    },
+    {
+      "epoch": 2.6272324302451766,
+      "grad_norm": 0.35768619179725647,
+      "learning_rate": 3.949353373747106e-06,
+      "loss": 2.8568,
+      "step": 35710
+    },
+    {
+      "epoch": 2.627968143611249,
+      "grad_norm": 0.3536481559276581,
+      "learning_rate": 3.933984708148414e-06,
+      "loss": 2.8408,
+      "step": 35720
+    },
+    {
+      "epoch": 2.628703856977322,
+      "grad_norm": 0.3568023145198822,
+      "learning_rate": 3.91864477933323e-06,
+      "loss": 2.8745,
+      "step": 35730
+    },
+    {
+      "epoch": 2.6294395703433944,
+      "grad_norm": 0.35367071628570557,
+      "learning_rate": 3.9033335968708895e-06,
+      "loss": 2.8068,
+      "step": 35740
+    },
+    {
+      "epoch": 2.630175283709467,
+      "grad_norm": 0.35229575634002686,
+      "learning_rate": 3.888051170312762e-06,
+      "loss": 2.8139,
+      "step": 35750
+    },
+    {
+      "epoch": 2.6309109970755395,
+      "grad_norm": 0.3524012863636017,
+      "learning_rate": 3.872797509192289e-06,
+      "loss": 2.8434,
+      "step": 35760
+    },
+    {
+      "epoch": 2.631646710441612,
+      "grad_norm": 0.35714295506477356,
+      "learning_rate": 3.857572623024974e-06,
+      "loss": 2.8407,
+      "step": 35770
+    },
+    {
+      "epoch": 2.6323824238076847,
+      "grad_norm": 0.35321491956710815,
+      "learning_rate": 3.84237652130836e-06,
+      "loss": 2.8223,
+      "step": 35780
+    },
+    {
+      "epoch": 2.6331181371737573,
+      "grad_norm": 0.3564974367618561,
+      "learning_rate": 3.827209213522059e-06,
+      "loss": 2.8576,
+      "step": 35790
+    },
+    {
+      "epoch": 2.63385385053983,
+      "grad_norm": 0.34589964151382446,
+      "learning_rate": 3.8120707091276854e-06,
+      "loss": 2.8443,
+      "step": 35800
+    },
+    {
+      "epoch": 2.6345895639059025,
+      "grad_norm": 0.3574235141277313,
+      "learning_rate": 3.796961017568912e-06,
+      "loss": 2.8506,
+      "step": 35810
+    },
+    {
+      "epoch": 2.635325277271975,
+      "grad_norm": 0.35355496406555176,
+      "learning_rate": 3.7818801482714107e-06,
+      "loss": 2.8618,
+      "step": 35820
+    },
+    {
+      "epoch": 2.6360609906380477,
+      "grad_norm": 0.3555285930633545,
+      "learning_rate": 3.7668281106429113e-06,
+      "loss": 2.8636,
+      "step": 35830
+    },
+    {
+      "epoch": 2.6367967040041203,
+      "grad_norm": 0.34864774346351624,
+      "learning_rate": 3.7518049140731316e-06,
+      "loss": 2.8142,
+      "step": 35840
+    },
+    {
+      "epoch": 2.637532417370193,
+      "grad_norm": 0.3653092384338379,
+      "learning_rate": 3.7368105679338016e-06,
+      "loss": 2.9029,
+      "step": 35850
+    },
+    {
+      "epoch": 2.6382681307362654,
+      "grad_norm": 0.3527710735797882,
+      "learning_rate": 3.72184508157864e-06,
+      "loss": 2.8922,
+      "step": 35860
+    },
+    {
+      "epoch": 2.6390038441023376,
+      "grad_norm": 0.34910181164741516,
+      "learning_rate": 3.706908464343395e-06,
+      "loss": 2.8646,
+      "step": 35870
+    },
+    {
+      "epoch": 2.63973955746841,
+      "grad_norm": 0.35044485330581665,
+      "learning_rate": 3.6920007255457966e-06,
+      "loss": 2.8337,
+      "step": 35880
+    },
+    {
+      "epoch": 2.6404752708344827,
+      "grad_norm": 0.3586481511592865,
+      "learning_rate": 3.677121874485545e-06,
+      "loss": 2.9033,
+      "step": 35890
+    },
+    {
+      "epoch": 2.6412109842005553,
+      "grad_norm": 0.35384953022003174,
+      "learning_rate": 3.662271920444321e-06,
+      "loss": 2.8329,
+      "step": 35900
+    },
+    {
+      "epoch": 2.641946697566628,
+      "grad_norm": 0.3543202877044678,
+      "learning_rate": 3.6474508726857815e-06,
+      "loss": 2.8738,
+      "step": 35910
+    },
+    {
+      "epoch": 2.6426824109327005,
+      "grad_norm": 0.34686681628227234,
+      "learning_rate": 3.632658740455569e-06,
+      "loss": 2.8361,
+      "step": 35920
+    },
+    {
+      "epoch": 2.643418124298773,
+      "grad_norm": 0.34684446454048157,
+      "learning_rate": 3.6178955329812635e-06,
+      "loss": 2.8057,
+      "step": 35930
+    },
+    {
+      "epoch": 2.6441538376648457,
+      "grad_norm": 0.3559458553791046,
+      "learning_rate": 3.603161259472415e-06,
+      "loss": 2.8312,
+      "step": 35940
+    },
+    {
+      "epoch": 2.6448895510309183,
+      "grad_norm": 0.34946638345718384,
+      "learning_rate": 3.58845592912051e-06,
+      "loss": 2.8376,
+      "step": 35950
+    },
+    {
+      "epoch": 2.645625264396991,
+      "grad_norm": 0.3577750623226166,
+      "learning_rate": 3.573779551098999e-06,
+      "loss": 2.8301,
+      "step": 35960
+    },
+    {
+      "epoch": 2.6463609777630634,
+      "grad_norm": 0.3560861349105835,
+      "learning_rate": 3.55913213456327e-06,
+      "loss": 2.8479,
+      "step": 35970
+    },
+    {
+      "epoch": 2.647096691129136,
+      "grad_norm": 0.3636762499809265,
+      "learning_rate": 3.544513688650625e-06,
+      "loss": 2.8544,
+      "step": 35980
+    },
+    {
+      "epoch": 2.6478324044952086,
+      "grad_norm": 0.36199668049812317,
+      "learning_rate": 3.5299242224803133e-06,
+      "loss": 2.8986,
+      "step": 35990
+    },
+    {
+      "epoch": 2.648568117861281,
+      "grad_norm": 0.35867244005203247,
+      "learning_rate": 3.5153637451534882e-06,
+      "loss": 2.8633,
+      "step": 36000
+    },
+    {
+      "epoch": 2.649303831227354,
+      "grad_norm": 0.35806673765182495,
+      "learning_rate": 3.5008322657532398e-06,
+      "loss": 2.8917,
+      "step": 36010
+    },
+    {
+      "epoch": 2.6500395445934264,
+      "grad_norm": 0.3565541207790375,
+      "learning_rate": 3.486329793344556e-06,
+      "loss": 2.8395,
+      "step": 36020
+    },
+    {
+      "epoch": 2.650775257959499,
+      "grad_norm": 0.3679090142250061,
+      "learning_rate": 3.471856336974322e-06,
+      "loss": 2.7876,
+      "step": 36030
+    },
+    {
+      "epoch": 2.6515109713255716,
+      "grad_norm": 0.34930306673049927,
+      "learning_rate": 3.4574119056713427e-06,
+      "loss": 2.8042,
+      "step": 36040
+    },
+    {
+      "epoch": 2.652246684691644,
+      "grad_norm": 0.3561093807220459,
+      "learning_rate": 3.4429965084463e-06,
+      "loss": 2.8431,
+      "step": 36050
+    },
+    {
+      "epoch": 2.6529823980577167,
+      "grad_norm": 0.3582862317562103,
+      "learning_rate": 3.4286101542917837e-06,
+      "loss": 2.824,
+      "step": 36060
+    },
+    {
+      "epoch": 2.6537181114237893,
+      "grad_norm": 0.3488673269748688,
+      "learning_rate": 3.414252852182237e-06,
+      "loss": 2.8264,
+      "step": 36070
+    },
+    {
+      "epoch": 2.654453824789862,
+      "grad_norm": 0.365986168384552,
+      "learning_rate": 3.399924611074007e-06,
+      "loss": 2.8375,
+      "step": 36080
+    },
+    {
+      "epoch": 2.6551895381559345,
+      "grad_norm": 0.35745906829833984,
+      "learning_rate": 3.3856254399052824e-06,
+      "loss": 2.9134,
+      "step": 36090
+    },
+    {
+      "epoch": 2.655925251522007,
+      "grad_norm": 0.3522699773311615,
+      "learning_rate": 3.37135534759615e-06,
+      "loss": 2.8123,
+      "step": 36100
+    },
+    {
+      "epoch": 2.6566609648880797,
+      "grad_norm": 0.3483119606971741,
+      "learning_rate": 3.3571143430485397e-06,
+      "loss": 2.8029,
+      "step": 36110
+    },
+    {
+      "epoch": 2.6573966782541523,
+      "grad_norm": 0.36627569794654846,
+      "learning_rate": 3.3429024351462392e-06,
+      "loss": 2.9125,
+      "step": 36120
+    },
+    {
+      "epoch": 2.658132391620225,
+      "grad_norm": 0.34490320086479187,
+      "learning_rate": 3.3287196327548843e-06,
+      "loss": 2.8362,
+      "step": 36130
+    },
+    {
+      "epoch": 2.6588681049862974,
+      "grad_norm": 0.36275714635849,
+      "learning_rate": 3.314565944721948e-06,
+      "loss": 2.8878,
+      "step": 36140
+    },
+    {
+      "epoch": 2.65960381835237,
+      "grad_norm": 0.3495403528213501,
+      "learning_rate": 3.3004413798767617e-06,
+      "loss": 2.8933,
+      "step": 36150
+    },
+    {
+      "epoch": 2.6603395317184426,
+      "grad_norm": 0.35380616784095764,
+      "learning_rate": 3.2863459470304615e-06,
+      "loss": 2.8299,
+      "step": 36160
+    },
+    {
+      "epoch": 2.661075245084515,
+      "grad_norm": 0.3434477150440216,
+      "learning_rate": 3.272279654976035e-06,
+      "loss": 2.8201,
+      "step": 36170
+    },
+    {
+      "epoch": 2.661810958450588,
+      "grad_norm": 0.3668873608112335,
+      "learning_rate": 3.2582425124882633e-06,
+      "loss": 2.8853,
+      "step": 36180
+    },
+    {
+      "epoch": 2.6625466718166604,
+      "grad_norm": 0.3458184599876404,
+      "learning_rate": 3.244234528323775e-06,
+      "loss": 2.8177,
+      "step": 36190
+    },
+    {
+      "epoch": 2.663282385182733,
+      "grad_norm": 0.35831740498542786,
+      "learning_rate": 3.230255711220992e-06,
+      "loss": 2.9123,
+      "step": 36200
+    },
+    {
+      "epoch": 2.664018098548805,
+      "grad_norm": 0.3599366545677185,
+      "learning_rate": 3.2163060699001435e-06,
+      "loss": 2.8391,
+      "step": 36210
+    },
+    {
+      "epoch": 2.6647538119148777,
+      "grad_norm": 0.3606792688369751,
+      "learning_rate": 3.2023856130632635e-06,
+      "loss": 2.8596,
+      "step": 36220
+    },
+    {
+      "epoch": 2.6654895252809503,
+      "grad_norm": 0.3445892930030823,
+      "learning_rate": 3.1884943493941666e-06,
+      "loss": 2.8166,
+      "step": 36230
+    },
+    {
+      "epoch": 2.666225238647023,
+      "grad_norm": 0.34717074036598206,
+      "learning_rate": 3.1746322875584768e-06,
+      "loss": 2.8439,
+      "step": 36240
+    },
+    {
+      "epoch": 2.6669609520130955,
+      "grad_norm": 0.35464805364608765,
+      "learning_rate": 3.160799436203582e-06,
+      "loss": 2.8817,
+      "step": 36250
+    },
+    {
+      "epoch": 2.667696665379168,
+      "grad_norm": 0.35289961099624634,
+      "learning_rate": 3.1469958039586522e-06,
+      "loss": 2.8663,
+      "step": 36260
+    },
+    {
+      "epoch": 2.6684323787452406,
+      "grad_norm": 0.3482816815376282,
+      "learning_rate": 3.133221399434638e-06,
+      "loss": 2.8435,
+      "step": 36270
+    },
+    {
+      "epoch": 2.6691680921113132,
+      "grad_norm": 0.35229161381721497,
+      "learning_rate": 3.119476231224255e-06,
+      "loss": 2.8614,
+      "step": 36280
+    },
+    {
+      "epoch": 2.669903805477386,
+      "grad_norm": 0.35111215710639954,
+      "learning_rate": 3.105760307901978e-06,
+      "loss": 2.8159,
+      "step": 36290
+    },
+    {
+      "epoch": 2.6706395188434584,
+      "grad_norm": 0.36267319321632385,
+      "learning_rate": 3.0920736380240455e-06,
+      "loss": 2.9111,
+      "step": 36300
+    },
+    {
+      "epoch": 2.671375232209531,
+      "grad_norm": 0.3615294396877289,
+      "learning_rate": 3.0784162301284337e-06,
+      "loss": 2.8237,
+      "step": 36310
+    },
+    {
+      "epoch": 2.6721109455756036,
+      "grad_norm": 0.3596484661102295,
+      "learning_rate": 3.0647880927348602e-06,
+      "loss": 2.88,
+      "step": 36320
+    },
+    {
+      "epoch": 2.672846658941676,
+      "grad_norm": 0.3571300804615021,
+      "learning_rate": 3.05118923434482e-06,
+      "loss": 2.8646,
+      "step": 36330
+    },
+    {
+      "epoch": 2.6735823723077488,
+      "grad_norm": 0.3523973226547241,
+      "learning_rate": 3.0376196634415043e-06,
+      "loss": 2.8313,
+      "step": 36340
+    },
+    {
+      "epoch": 2.6743180856738213,
+      "grad_norm": 0.3510914742946625,
+      "learning_rate": 3.024079388489848e-06,
+      "loss": 2.9127,
+      "step": 36350
+    },
+    {
+      "epoch": 2.675053799039894,
+      "grad_norm": 0.3617897927761078,
+      "learning_rate": 3.010568417936499e-06,
+      "loss": 2.8838,
+      "step": 36360
+    },
+    {
+      "epoch": 2.6757895124059665,
+      "grad_norm": 0.35268446803092957,
+      "learning_rate": 2.9970867602098553e-06,
+      "loss": 2.8024,
+      "step": 36370
+    },
+    {
+      "epoch": 2.676525225772039,
+      "grad_norm": 0.3574046194553375,
+      "learning_rate": 2.9836344237200053e-06,
+      "loss": 2.8592,
+      "step": 36380
+    },
+    {
+      "epoch": 2.6772609391381117,
+      "grad_norm": 0.3499765694141388,
+      "learning_rate": 2.9702114168587536e-06,
+      "loss": 2.8796,
+      "step": 36390
+    },
+    {
+      "epoch": 2.6779966525041843,
+      "grad_norm": 0.3450222313404083,
+      "learning_rate": 2.9568177479996017e-06,
+      "loss": 2.8435,
+      "step": 36400
+    },
+    {
+      "epoch": 2.678732365870257,
+      "grad_norm": 0.3479706943035126,
+      "learning_rate": 2.9434534254977443e-06,
+      "loss": 2.8396,
+      "step": 36410
+    },
+    {
+      "epoch": 2.6794680792363295,
+      "grad_norm": 0.3678218722343445,
+      "learning_rate": 2.930118457690095e-06,
+      "loss": 2.8371,
+      "step": 36420
+    },
+    {
+      "epoch": 2.680203792602402,
+      "grad_norm": 0.3589637577533722,
+      "learning_rate": 2.9168128528952344e-06,
+      "loss": 2.8823,
+      "step": 36430
+    },
+    {
+      "epoch": 2.6809395059684746,
+      "grad_norm": 0.34771761298179626,
+      "learning_rate": 2.9035366194134162e-06,
+      "loss": 2.8297,
+      "step": 36440
+    },
+    {
+      "epoch": 2.6816752193345472,
+      "grad_norm": 0.3574524223804474,
+      "learning_rate": 2.8902897655266103e-06,
+      "loss": 2.8354,
+      "step": 36450
+    },
+    {
+      "epoch": 2.68241093270062,
+      "grad_norm": 0.358864426612854,
+      "learning_rate": 2.877072299498418e-06,
+      "loss": 2.8779,
+      "step": 36460
+    },
+    {
+      "epoch": 2.6831466460666924,
+      "grad_norm": 0.34265199303627014,
+      "learning_rate": 2.863884229574132e-06,
+      "loss": 2.828,
+      "step": 36470
+    },
+    {
+      "epoch": 2.683882359432765,
+      "grad_norm": 0.35038596391677856,
+      "learning_rate": 2.850725563980705e-06,
+      "loss": 2.8341,
+      "step": 36480
+    },
+    {
+      "epoch": 2.6846180727988376,
+      "grad_norm": 0.3619864284992218,
+      "learning_rate": 2.837596310926732e-06,
+      "loss": 2.835,
+      "step": 36490
+    },
+    {
+      "epoch": 2.68535378616491,
+      "grad_norm": 0.34702086448669434,
+      "learning_rate": 2.8244964786024674e-06,
+      "loss": 2.8741,
+      "step": 36500
+    },
+    {
+      "epoch": 2.6860894995309827,
+      "grad_norm": 0.35705944895744324,
+      "learning_rate": 2.8114260751798295e-06,
+      "loss": 2.8911,
+      "step": 36510
+    },
+    {
+      "epoch": 2.6868252128970553,
+      "grad_norm": 0.35176917910575867,
+      "learning_rate": 2.798385108812357e-06,
+      "loss": 2.8066,
+      "step": 36520
+    },
+    {
+      "epoch": 2.687560926263128,
+      "grad_norm": 0.36152660846710205,
+      "learning_rate": 2.7853735876352206e-06,
+      "loss": 2.8132,
+      "step": 36530
+    },
+    {
+      "epoch": 2.6882966396292005,
+      "grad_norm": 0.35037901997566223,
+      "learning_rate": 2.772391519765244e-06,
+      "loss": 2.838,
+      "step": 36540
+    },
+    {
+      "epoch": 2.689032352995273,
+      "grad_norm": 0.35159310698509216,
+      "learning_rate": 2.759438913300877e-06,
+      "loss": 2.7816,
+      "step": 36550
+    },
+    {
+      "epoch": 2.6897680663613457,
+      "grad_norm": 0.3596172332763672,
+      "learning_rate": 2.7465157763221727e-06,
+      "loss": 2.9097,
+      "step": 36560
+    },
+    {
+      "epoch": 2.6905037797274183,
+      "grad_norm": 0.3756032884120941,
+      "learning_rate": 2.7336221168908103e-06,
+      "loss": 2.877,
+      "step": 36570
+    },
+    {
+      "epoch": 2.691239493093491,
+      "grad_norm": 0.3487653434276581,
+      "learning_rate": 2.7207579430500676e-06,
+      "loss": 2.7998,
+      "step": 36580
+    },
+    {
+      "epoch": 2.6919752064595635,
+      "grad_norm": 0.34851446747779846,
+      "learning_rate": 2.7079232628248575e-06,
+      "loss": 2.8313,
+      "step": 36590
+    },
+    {
+      "epoch": 2.692710919825636,
+      "grad_norm": 0.35487809777259827,
+      "learning_rate": 2.6951180842216704e-06,
+      "loss": 2.8489,
+      "step": 36600
+    },
+    {
+      "epoch": 2.6934466331917086,
+      "grad_norm": 0.3485107719898224,
+      "learning_rate": 2.6823424152285947e-06,
+      "loss": 2.8431,
+      "step": 36610
+    },
+    {
+      "epoch": 2.694182346557781,
+      "grad_norm": 0.34937962889671326,
+      "learning_rate": 2.669596263815327e-06,
+      "loss": 2.8441,
+      "step": 36620
+    },
+    {
+      "epoch": 2.694918059923854,
+      "grad_norm": 0.3529322147369385,
+      "learning_rate": 2.6568796379331297e-06,
+      "loss": 2.8396,
+      "step": 36630
+    },
+    {
+      "epoch": 2.6956537732899264,
+      "grad_norm": 0.34904178977012634,
+      "learning_rate": 2.6441925455148618e-06,
+      "loss": 2.8182,
+      "step": 36640
+    },
+    {
+      "epoch": 2.696389486655999,
+      "grad_norm": 0.3651334047317505,
+      "learning_rate": 2.631534994474949e-06,
+      "loss": 2.8049,
+      "step": 36650
+    },
+    {
+      "epoch": 2.6971252000220716,
+      "grad_norm": 0.3608277142047882,
+      "learning_rate": 2.6189069927093913e-06,
+      "loss": 2.8677,
+      "step": 36660
+    },
+    {
+      "epoch": 2.697860913388144,
+      "grad_norm": 0.35738688707351685,
+      "learning_rate": 2.6063085480957537e-06,
+      "loss": 2.8454,
+      "step": 36670
+    },
+    {
+      "epoch": 2.6985966267542167,
+      "grad_norm": 0.3562159240245819,
+      "learning_rate": 2.593739668493178e-06,
+      "loss": 2.8705,
+      "step": 36680
+    },
+    {
+      "epoch": 2.6993323401202893,
+      "grad_norm": 0.34853512048721313,
+      "learning_rate": 2.581200361742336e-06,
+      "loss": 2.843,
+      "step": 36690
+    },
+    {
+      "epoch": 2.700068053486362,
+      "grad_norm": 0.3620928227901459,
+      "learning_rate": 2.568690635665477e-06,
+      "loss": 2.8277,
+      "step": 36700
+    },
+    {
+      "epoch": 2.7008037668524345,
+      "grad_norm": 0.351656973361969,
+      "learning_rate": 2.5562104980663857e-06,
+      "loss": 2.8519,
+      "step": 36710
+    },
+    {
+      "epoch": 2.701539480218507,
+      "grad_norm": 0.3516268730163574,
+      "learning_rate": 2.5437599567303738e-06,
+      "loss": 2.8491,
+      "step": 36720
+    },
+    {
+      "epoch": 2.7022751935845797,
+      "grad_norm": 0.35083866119384766,
+      "learning_rate": 2.531339019424328e-06,
+      "loss": 2.8326,
+      "step": 36730
+    },
+    {
+      "epoch": 2.7030109069506523,
+      "grad_norm": 0.354348748922348,
+      "learning_rate": 2.5189476938966338e-06,
+      "loss": 2.8535,
+      "step": 36740
+    },
+    {
+      "epoch": 2.703746620316725,
+      "grad_norm": 0.348808616399765,
+      "learning_rate": 2.5065859878772135e-06,
+      "loss": 2.8832,
+      "step": 36750
+    },
+    {
+      "epoch": 2.7044823336827974,
+      "grad_norm": 0.3641868233680725,
+      "learning_rate": 2.4942539090775153e-06,
+      "loss": 2.8774,
+      "step": 36760
+    },
+    {
+      "epoch": 2.70521804704887,
+      "grad_norm": 0.3460129201412201,
+      "learning_rate": 2.481951465190507e-06,
+      "loss": 2.805,
+      "step": 36770
+    },
+    {
+      "epoch": 2.705953760414942,
+      "grad_norm": 0.35824331641197205,
+      "learning_rate": 2.469678663890679e-06,
+      "loss": 2.8092,
+      "step": 36780
+    },
+    {
+      "epoch": 2.7066894737810148,
+      "grad_norm": 0.3646627366542816,
+      "learning_rate": 2.457435512834011e-06,
+      "loss": 2.8691,
+      "step": 36790
+    },
+    {
+      "epoch": 2.7074251871470874,
+      "grad_norm": 0.3422071039676666,
+      "learning_rate": 2.4452220196579835e-06,
+      "loss": 2.8525,
+      "step": 36800
+    },
+    {
+      "epoch": 2.70816090051316,
+      "grad_norm": 0.3469374477863312,
+      "learning_rate": 2.433038191981596e-06,
+      "loss": 2.8321,
+      "step": 36810
+    },
+    {
+      "epoch": 2.7088966138792325,
+      "grad_norm": 0.34474003314971924,
+      "learning_rate": 2.4208840374053353e-06,
+      "loss": 2.7606,
+      "step": 36820
+    },
+    {
+      "epoch": 2.709632327245305,
+      "grad_norm": 0.34081652760505676,
+      "learning_rate": 2.4087595635111705e-06,
+      "loss": 2.8718,
+      "step": 36830
+    },
+    {
+      "epoch": 2.7103680406113777,
+      "grad_norm": 0.35089510679244995,
+      "learning_rate": 2.396664777862562e-06,
+      "loss": 2.8547,
+      "step": 36840
+    },
+    {
+      "epoch": 2.7111037539774503,
+      "grad_norm": 0.35770079493522644,
+      "learning_rate": 2.3845996880044407e-06,
+      "loss": 2.8569,
+      "step": 36850
+    },
+    {
+      "epoch": 2.711839467343523,
+      "grad_norm": 0.35327571630477905,
+      "learning_rate": 2.3725643014632202e-06,
+      "loss": 2.7858,
+      "step": 36860
+    },
+    {
+      "epoch": 2.7125751807095955,
+      "grad_norm": 0.3552086353302002,
+      "learning_rate": 2.3605586257468006e-06,
+      "loss": 2.8265,
+      "step": 36870
+    },
+    {
+      "epoch": 2.713310894075668,
+      "grad_norm": 0.34446415305137634,
+      "learning_rate": 2.3485826683445235e-06,
+      "loss": 2.8353,
+      "step": 36880
+    },
+    {
+      "epoch": 2.7140466074417406,
+      "grad_norm": 0.3575133681297302,
+      "learning_rate": 2.3366364367271967e-06,
+      "loss": 2.8814,
+      "step": 36890
+    },
+    {
+      "epoch": 2.7147823208078132,
+      "grad_norm": 0.3534136414527893,
+      "learning_rate": 2.3247199383470807e-06,
+      "loss": 2.8542,
+      "step": 36900
+    },
+    {
+      "epoch": 2.715518034173886,
+      "grad_norm": 0.3531450033187866,
+      "learning_rate": 2.312833180637908e-06,
+      "loss": 2.8536,
+      "step": 36910
+    },
+    {
+      "epoch": 2.7162537475399584,
+      "grad_norm": 0.35042592883110046,
+      "learning_rate": 2.30097617101484e-06,
+      "loss": 2.8543,
+      "step": 36920
+    },
+    {
+      "epoch": 2.716989460906031,
+      "grad_norm": 0.3475674092769623,
+      "learning_rate": 2.2891489168744893e-06,
+      "loss": 2.8243,
+      "step": 36930
+    },
+    {
+      "epoch": 2.7177251742721036,
+      "grad_norm": 0.34900960326194763,
+      "learning_rate": 2.2773514255948923e-06,
+      "loss": 2.8456,
+      "step": 36940
+    },
+    {
+      "epoch": 2.718460887638176,
+      "grad_norm": 0.35013091564178467,
+      "learning_rate": 2.26558370453554e-06,
+      "loss": 2.8731,
+      "step": 36950
+    },
+    {
+      "epoch": 2.7191966010042488,
+      "grad_norm": 0.34886035323143005,
+      "learning_rate": 2.253845761037343e-06,
+      "loss": 2.7862,
+      "step": 36960
+    },
+    {
+      "epoch": 2.7199323143703213,
+      "grad_norm": 0.35427263379096985,
+      "learning_rate": 2.2421376024226338e-06,
+      "loss": 2.8076,
+      "step": 36970
+    },
+    {
+      "epoch": 2.720668027736394,
+      "grad_norm": 0.34409666061401367,
+      "learning_rate": 2.230459235995164e-06,
+      "loss": 2.8576,
+      "step": 36980
+    },
+    {
+      "epoch": 2.7214037411024665,
+      "grad_norm": 0.35431531071662903,
+      "learning_rate": 2.2188106690400946e-06,
+      "loss": 2.8373,
+      "step": 36990
+    },
+    {
+      "epoch": 2.722139454468539,
+      "grad_norm": 0.3577077090740204,
+      "learning_rate": 2.2071919088240187e-06,
+      "loss": 2.8419,
+      "step": 37000
+    },
+    {
+      "epoch": 2.7228751678346117,
+      "grad_norm": 0.35210809111595154,
+      "learning_rate": 2.1956029625949116e-06,
+      "loss": 2.8649,
+      "step": 37010
+    },
+    {
+      "epoch": 2.7236108812006843,
+      "grad_norm": 0.36822712421417236,
+      "learning_rate": 2.1840438375821614e-06,
+      "loss": 2.9043,
+      "step": 37020
+    },
+    {
+      "epoch": 2.724346594566757,
+      "grad_norm": 0.3567112684249878,
+      "learning_rate": 2.1725145409965565e-06,
+      "loss": 2.8802,
+      "step": 37030
+    },
+    {
+      "epoch": 2.7250823079328295,
+      "grad_norm": 0.35724371671676636,
+      "learning_rate": 2.161015080030271e-06,
+      "loss": 2.8964,
+      "step": 37040
+    },
+    {
+      "epoch": 2.725818021298902,
+      "grad_norm": 0.3512875437736511,
+      "learning_rate": 2.1495454618568666e-06,
+      "loss": 2.8415,
+      "step": 37050
+    },
+    {
+      "epoch": 2.7265537346649746,
+      "grad_norm": 0.3504386246204376,
+      "learning_rate": 2.138105693631298e-06,
+      "loss": 2.8569,
+      "step": 37060
+    },
+    {
+      "epoch": 2.7272894480310472,
+      "grad_norm": 0.3547613322734833,
+      "learning_rate": 2.1266957824898905e-06,
+      "loss": 2.8046,
+      "step": 37070
+    },
+    {
+      "epoch": 2.72802516139712,
+      "grad_norm": 0.35498207807540894,
+      "learning_rate": 2.1153157355503274e-06,
+      "loss": 2.834,
+      "step": 37080
+    },
+    {
+      "epoch": 2.7287608747631924,
+      "grad_norm": 0.3593672811985016,
+      "learning_rate": 2.103965559911708e-06,
+      "loss": 2.8452,
+      "step": 37090
+    },
+    {
+      "epoch": 2.729496588129265,
+      "grad_norm": 0.3431476056575775,
+      "learning_rate": 2.0926452626544466e-06,
+      "loss": 2.8091,
+      "step": 37100
+    },
+    {
+      "epoch": 2.7302323014953376,
+      "grad_norm": 0.35229626297950745,
+      "learning_rate": 2.0813548508403613e-06,
+      "loss": 2.8307,
+      "step": 37110
+    },
+    {
+      "epoch": 2.7309680148614097,
+      "grad_norm": 0.3566063344478607,
+      "learning_rate": 2.070094331512601e-06,
+      "loss": 2.7962,
+      "step": 37120
+    },
+    {
+      "epoch": 2.7317037282274823,
+      "grad_norm": 0.3534442186355591,
+      "learning_rate": 2.058863711695663e-06,
+      "loss": 2.8785,
+      "step": 37130
+    },
+    {
+      "epoch": 2.732439441593555,
+      "grad_norm": 0.3649783134460449,
+      "learning_rate": 2.0476629983954222e-06,
+      "loss": 2.8529,
+      "step": 37140
+    },
+    {
+      "epoch": 2.7331751549596275,
+      "grad_norm": 0.3442041575908661,
+      "learning_rate": 2.036492198599066e-06,
+      "loss": 2.8175,
+      "step": 37150
+    },
+    {
+      "epoch": 2.7339108683257,
+      "grad_norm": 0.3532116413116455,
+      "learning_rate": 2.0253513192751373e-06,
+      "loss": 2.8317,
+      "step": 37160
+    },
+    {
+      "epoch": 2.7346465816917727,
+      "grad_norm": 0.35195299983024597,
+      "learning_rate": 2.0142403673735044e-06,
+      "loss": 2.8346,
+      "step": 37170
+    },
+    {
+      "epoch": 2.7353822950578452,
+      "grad_norm": 0.351925253868103,
+      "learning_rate": 2.0031593498253785e-06,
+      "loss": 2.8344,
+      "step": 37180
+    },
+    {
+      "epoch": 2.736118008423918,
+      "grad_norm": 0.34304651618003845,
+      "learning_rate": 1.992108273543303e-06,
+      "loss": 2.8128,
+      "step": 37190
+    },
+    {
+      "epoch": 2.7368537217899904,
+      "grad_norm": 0.367616206407547,
+      "learning_rate": 1.981087145421118e-06,
+      "loss": 2.8867,
+      "step": 37200
+    },
+    {
+      "epoch": 2.737589435156063,
+      "grad_norm": 0.34709399938583374,
+      "learning_rate": 1.970095972334002e-06,
+      "loss": 2.811,
+      "step": 37210
+    },
+    {
+      "epoch": 2.7383251485221356,
+      "grad_norm": 0.34985437989234924,
+      "learning_rate": 1.9591347611384257e-06,
+      "loss": 2.8257,
+      "step": 37220
+    },
+    {
+      "epoch": 2.739060861888208,
+      "grad_norm": 0.35936376452445984,
+      "learning_rate": 1.9482035186722093e-06,
+      "loss": 2.8729,
+      "step": 37230
+    },
+    {
+      "epoch": 2.7397965752542808,
+      "grad_norm": 0.34850063920021057,
+      "learning_rate": 1.9373022517544305e-06,
+      "loss": 2.8689,
+      "step": 37240
+    },
+    {
+      "epoch": 2.7405322886203534,
+      "grad_norm": 0.3589111566543579,
+      "learning_rate": 1.9264309671855054e-06,
+      "loss": 2.8499,
+      "step": 37250
+    },
+    {
+      "epoch": 2.741268001986426,
+      "grad_norm": 0.3594447672367096,
+      "learning_rate": 1.9155896717471087e-06,
+      "loss": 2.8236,
+      "step": 37260
+    },
+    {
+      "epoch": 2.7420037153524985,
+      "grad_norm": 0.35577619075775146,
+      "learning_rate": 1.9047783722022482e-06,
+      "loss": 2.8377,
+      "step": 37270
+    },
+    {
+      "epoch": 2.742739428718571,
+      "grad_norm": 0.3540429174900055,
+      "learning_rate": 1.8939970752952063e-06,
+      "loss": 2.8693,
+      "step": 37280
+    },
+    {
+      "epoch": 2.7434751420846437,
+      "grad_norm": 0.3595455288887024,
+      "learning_rate": 1.8832457877515308e-06,
+      "loss": 2.8388,
+      "step": 37290
+    },
+    {
+      "epoch": 2.7442108554507163,
+      "grad_norm": 0.35307249426841736,
+      "learning_rate": 1.8725245162780626e-06,
+      "loss": 2.8058,
+      "step": 37300
+    },
+    {
+      "epoch": 2.744946568816789,
+      "grad_norm": 0.35251784324645996,
+      "learning_rate": 1.861833267562918e-06,
+      "loss": 2.8702,
+      "step": 37310
+    },
+    {
+      "epoch": 2.7456822821828615,
+      "grad_norm": 0.3533673882484436,
+      "learning_rate": 1.8511720482754958e-06,
+      "loss": 2.8032,
+      "step": 37320
+    },
+    {
+      "epoch": 2.746417995548934,
+      "grad_norm": 0.352582722902298,
+      "learning_rate": 1.8405408650664424e-06,
+      "loss": 2.8443,
+      "step": 37330
+    },
+    {
+      "epoch": 2.7471537089150067,
+      "grad_norm": 0.3679851293563843,
+      "learning_rate": 1.8299397245676698e-06,
+      "loss": 2.8806,
+      "step": 37340
+    },
+    {
+      "epoch": 2.7478894222810792,
+      "grad_norm": 0.3475781977176666,
+      "learning_rate": 1.8193686333923488e-06,
+      "loss": 2.7878,
+      "step": 37350
+    },
+    {
+      "epoch": 2.748625135647152,
+      "grad_norm": 0.3561195731163025,
+      "learning_rate": 1.8088275981349377e-06,
+      "loss": 2.8474,
+      "step": 37360
+    },
+    {
+      "epoch": 2.7493608490132244,
+      "grad_norm": 0.3596433103084564,
+      "learning_rate": 1.7983166253710992e-06,
+      "loss": 2.855,
+      "step": 37370
+    },
+    {
+      "epoch": 2.750096562379297,
+      "grad_norm": 0.3506121039390564,
+      "learning_rate": 1.7878357216577602e-06,
+      "loss": 2.8132,
+      "step": 37380
+    },
+    {
+      "epoch": 2.7508322757453696,
+      "grad_norm": 0.34753671288490295,
+      "learning_rate": 1.7773848935330961e-06,
+      "loss": 2.8796,
+      "step": 37390
+    },
+    {
+      "epoch": 2.751567989111442,
+      "grad_norm": 0.3480505645275116,
+      "learning_rate": 1.7669641475165144e-06,
+      "loss": 2.825,
+      "step": 37400
+    },
+    {
+      "epoch": 2.7523037024775148,
+      "grad_norm": 0.35291972756385803,
+      "learning_rate": 1.7565734901086538e-06,
+      "loss": 2.8564,
+      "step": 37410
+    },
+    {
+      "epoch": 2.7530394158435874,
+      "grad_norm": 0.3545636534690857,
+      "learning_rate": 1.7462129277914018e-06,
+      "loss": 2.8019,
+      "step": 37420
+    },
+    {
+      "epoch": 2.75377512920966,
+      "grad_norm": 0.35266727209091187,
+      "learning_rate": 1.7358824670278329e-06,
+      "loss": 2.8758,
+      "step": 37430
+    },
+    {
+      "epoch": 2.7545108425757325,
+      "grad_norm": 0.36245593428611755,
+      "learning_rate": 1.725582114262292e-06,
+      "loss": 2.7908,
+      "step": 37440
+    },
+    {
+      "epoch": 2.755246555941805,
+      "grad_norm": 0.351978600025177,
+      "learning_rate": 1.7153118759203114e-06,
+      "loss": 2.8381,
+      "step": 37450
+    },
+    {
+      "epoch": 2.7559822693078777,
+      "grad_norm": 0.35723474621772766,
+      "learning_rate": 1.7050717584086384e-06,
+      "loss": 2.8514,
+      "step": 37460
+    },
+    {
+      "epoch": 2.7567179826739503,
+      "grad_norm": 0.36322957277297974,
+      "learning_rate": 1.6948617681152523e-06,
+      "loss": 2.8277,
+      "step": 37470
+    },
+    {
+      "epoch": 2.757453696040023,
+      "grad_norm": 0.35327285528182983,
+      "learning_rate": 1.6846819114093027e-06,
+      "loss": 2.8517,
+      "step": 37480
+    },
+    {
+      "epoch": 2.7581894094060955,
+      "grad_norm": 0.3513939380645752,
+      "learning_rate": 1.6745321946411763e-06,
+      "loss": 2.8935,
+      "step": 37490
+    },
+    {
+      "epoch": 2.758925122772168,
+      "grad_norm": 0.3616991639137268,
+      "learning_rate": 1.6644126241424362e-06,
+      "loss": 2.8663,
+      "step": 37500
+    },
+    {
+      "epoch": 2.7596608361382406,
+      "grad_norm": 0.36036038398742676,
+      "learning_rate": 1.6543232062258496e-06,
+      "loss": 2.8621,
+      "step": 37510
+    },
+    {
+      "epoch": 2.7603965495043132,
+      "grad_norm": 0.34972357749938965,
+      "learning_rate": 1.644263947185376e-06,
+      "loss": 2.873,
+      "step": 37520
+    },
+    {
+      "epoch": 2.761132262870386,
+      "grad_norm": 0.35491055250167847,
+      "learning_rate": 1.634234853296146e-06,
+      "loss": 2.8356,
+      "step": 37530
+    },
+    {
+      "epoch": 2.7618679762364584,
+      "grad_norm": 0.3507627248764038,
+      "learning_rate": 1.624235930814494e-06,
+      "loss": 2.9103,
+      "step": 37540
+    },
+    {
+      "epoch": 2.762603689602531,
+      "grad_norm": 0.3702545166015625,
+      "learning_rate": 1.6142671859779135e-06,
+      "loss": 2.9006,
+      "step": 37550
+    },
+    {
+      "epoch": 2.7633394029686036,
+      "grad_norm": 0.3555087149143219,
+      "learning_rate": 1.6043286250050915e-06,
+      "loss": 2.8574,
+      "step": 37560
+    },
+    {
+      "epoch": 2.764075116334676,
+      "grad_norm": 0.35074692964553833,
+      "learning_rate": 1.5944202540958576e-06,
+      "loss": 2.874,
+      "step": 37570
+    },
+    {
+      "epoch": 2.7648108297007488,
+      "grad_norm": 0.34493979811668396,
+      "learning_rate": 1.5845420794312504e-06,
+      "loss": 2.8482,
+      "step": 37580
+    },
+    {
+      "epoch": 2.7655465430668214,
+      "grad_norm": 0.3631422817707062,
+      "learning_rate": 1.57469410717343e-06,
+      "loss": 2.9044,
+      "step": 37590
+    },
+    {
+      "epoch": 2.766282256432894,
+      "grad_norm": 0.35426369309425354,
+      "learning_rate": 1.5648763434657376e-06,
+      "loss": 2.8597,
+      "step": 37600
+    },
+    {
+      "epoch": 2.7670179697989665,
+      "grad_norm": 0.3503166139125824,
+      "learning_rate": 1.5550887944326686e-06,
+      "loss": 2.8629,
+      "step": 37610
+    },
+    {
+      "epoch": 2.767753683165039,
+      "grad_norm": 0.34839946031570435,
+      "learning_rate": 1.5453314661798556e-06,
+      "loss": 2.8455,
+      "step": 37620
+    },
+    {
+      "epoch": 2.7684893965311117,
+      "grad_norm": 0.34570321440696716,
+      "learning_rate": 1.5356043647941077e-06,
+      "loss": 2.8436,
+      "step": 37630
+    },
+    {
+      "epoch": 2.7692251098971843,
+      "grad_norm": 0.3614666163921356,
+      "learning_rate": 1.5259074963433539e-06,
+      "loss": 2.8749,
+      "step": 37640
+    },
+    {
+      "epoch": 2.769960823263257,
+      "grad_norm": 0.3657136857509613,
+      "learning_rate": 1.5162408668766615e-06,
+      "loss": 2.8243,
+      "step": 37650
+    },
+    {
+      "epoch": 2.7706965366293295,
+      "grad_norm": 0.35381099581718445,
+      "learning_rate": 1.5066044824242454e-06,
+      "loss": 2.797,
+      "step": 37660
+    },
+    {
+      "epoch": 2.771432249995402,
+      "grad_norm": 0.36116260290145874,
+      "learning_rate": 1.4969983489974581e-06,
+      "loss": 2.8618,
+      "step": 37670
+    },
+    {
+      "epoch": 2.7721679633614746,
+      "grad_norm": 0.3469048738479614,
+      "learning_rate": 1.4874224725887674e-06,
+      "loss": 2.8262,
+      "step": 37680
+    },
+    {
+      "epoch": 2.772903676727547,
+      "grad_norm": 0.3522867262363434,
+      "learning_rate": 1.4778768591717729e-06,
+      "loss": 2.8674,
+      "step": 37690
+    },
+    {
+      "epoch": 2.7736393900936194,
+      "grad_norm": 0.3638545274734497,
+      "learning_rate": 1.4683615147011941e-06,
+      "loss": 2.8605,
+      "step": 37700
+    },
+    {
+      "epoch": 2.774375103459692,
+      "grad_norm": 0.3495312035083771,
+      "learning_rate": 1.4588764451128666e-06,
+      "loss": 2.8361,
+      "step": 37710
+    },
+    {
+      "epoch": 2.7751108168257645,
+      "grad_norm": 0.35454076528549194,
+      "learning_rate": 1.4494216563237462e-06,
+      "loss": 2.8798,
+      "step": 37720
+    },
+    {
+      "epoch": 2.775846530191837,
+      "grad_norm": 0.3586793541908264,
+      "learning_rate": 1.439997154231898e-06,
+      "loss": 2.8243,
+      "step": 37730
+    },
+    {
+      "epoch": 2.7765822435579097,
+      "grad_norm": 0.35758402943611145,
+      "learning_rate": 1.430602944716486e-06,
+      "loss": 2.8466,
+      "step": 37740
+    },
+    {
+      "epoch": 2.7773179569239823,
+      "grad_norm": 0.3513910472393036,
+      "learning_rate": 1.4212390336377723e-06,
+      "loss": 2.8542,
+      "step": 37750
+    },
+    {
+      "epoch": 2.778053670290055,
+      "grad_norm": 0.353851318359375,
+      "learning_rate": 1.4119054268371457e-06,
+      "loss": 2.8536,
+      "step": 37760
+    },
+    {
+      "epoch": 2.7787893836561275,
+      "grad_norm": 0.35213640332221985,
+      "learning_rate": 1.4026021301370652e-06,
+      "loss": 2.7574,
+      "step": 37770
+    },
+    {
+      "epoch": 2.7795250970222,
+      "grad_norm": 0.3485030233860016,
+      "learning_rate": 1.3933291493410883e-06,
+      "loss": 2.8724,
+      "step": 37780
+    },
+    {
+      "epoch": 2.7802608103882727,
+      "grad_norm": 0.36303088068962097,
+      "learning_rate": 1.3840864902338602e-06,
+      "loss": 2.8609,
+      "step": 37790
+    },
+    {
+      "epoch": 2.7809965237543453,
+      "grad_norm": 0.34345126152038574,
+      "learning_rate": 1.3748741585811131e-06,
+      "loss": 2.8052,
+      "step": 37800
+    },
+    {
+      "epoch": 2.781732237120418,
+      "grad_norm": 0.35132405161857605,
+      "learning_rate": 1.3656921601296723e-06,
+      "loss": 2.8459,
+      "step": 37810
+    },
+    {
+      "epoch": 2.7824679504864904,
+      "grad_norm": 0.3470771312713623,
+      "learning_rate": 1.3565405006074116e-06,
+      "loss": 2.8537,
+      "step": 37820
+    },
+    {
+      "epoch": 2.783203663852563,
+      "grad_norm": 0.34596723318099976,
+      "learning_rate": 1.3474191857233088e-06,
+      "loss": 2.8601,
+      "step": 37830
+    },
+    {
+      "epoch": 2.7839393772186356,
+      "grad_norm": 0.3524584174156189,
+      "learning_rate": 1.3383282211673842e-06,
+      "loss": 2.8222,
+      "step": 37840
+    },
+    {
+      "epoch": 2.784675090584708,
+      "grad_norm": 0.3530162274837494,
+      "learning_rate": 1.3292676126107518e-06,
+      "loss": 2.8563,
+      "step": 37850
+    },
+    {
+      "epoch": 2.785410803950781,
+      "grad_norm": 0.34915411472320557,
+      "learning_rate": 1.3202373657055789e-06,
+      "loss": 2.8449,
+      "step": 37860
+    },
+    {
+      "epoch": 2.7861465173168534,
+      "grad_norm": 0.3498871922492981,
+      "learning_rate": 1.3112374860850873e-06,
+      "loss": 2.8436,
+      "step": 37870
+    },
+    {
+      "epoch": 2.786882230682926,
+      "grad_norm": 0.36818572878837585,
+      "learning_rate": 1.3022679793635584e-06,
+      "loss": 2.829,
+      "step": 37880
+    },
+    {
+      "epoch": 2.7876179440489985,
+      "grad_norm": 0.35415494441986084,
+      "learning_rate": 1.293328851136316e-06,
+      "loss": 2.8648,
+      "step": 37890
+    },
+    {
+      "epoch": 2.788353657415071,
+      "grad_norm": 0.3540479242801666,
+      "learning_rate": 1.2844201069797668e-06,
+      "loss": 2.8332,
+      "step": 37900
+    },
+    {
+      "epoch": 2.7890893707811437,
+      "grad_norm": 0.36057254672050476,
+      "learning_rate": 1.2755417524513258e-06,
+      "loss": 2.8798,
+      "step": 37910
+    },
+    {
+      "epoch": 2.7898250841472163,
+      "grad_norm": 0.3550379276275635,
+      "learning_rate": 1.266693793089463e-06,
+      "loss": 2.8352,
+      "step": 37920
+    },
+    {
+      "epoch": 2.790560797513289,
+      "grad_norm": 0.35827556252479553,
+      "learning_rate": 1.257876234413702e-06,
+      "loss": 2.8348,
+      "step": 37930
+    },
+    {
+      "epoch": 2.7912965108793615,
+      "grad_norm": 0.35371822118759155,
+      "learning_rate": 1.249089081924576e-06,
+      "loss": 2.8675,
+      "step": 37940
+    },
+    {
+      "epoch": 2.792032224245434,
+      "grad_norm": 0.3532661199569702,
+      "learning_rate": 1.2403323411036727e-06,
+      "loss": 2.8226,
+      "step": 37950
+    },
+    {
+      "epoch": 2.7927679376115067,
+      "grad_norm": 0.35149693489074707,
+      "learning_rate": 1.2316060174136002e-06,
+      "loss": 2.8661,
+      "step": 37960
+    },
+    {
+      "epoch": 2.7935036509775792,
+      "grad_norm": 0.3574874997138977,
+      "learning_rate": 1.2229101162979873e-06,
+      "loss": 2.8635,
+      "step": 37970
+    },
+    {
+      "epoch": 2.794239364343652,
+      "grad_norm": 0.3545931279659271,
+      "learning_rate": 1.214244643181489e-06,
+      "loss": 2.9128,
+      "step": 37980
+    },
+    {
+      "epoch": 2.7949750777097244,
+      "grad_norm": 0.3634801208972931,
+      "learning_rate": 1.2056096034697818e-06,
+      "loss": 2.8765,
+      "step": 37990
+    },
+    {
+      "epoch": 2.795710791075797,
+      "grad_norm": 0.35699209570884705,
+      "learning_rate": 1.1970050025495506e-06,
+      "loss": 2.8357,
+      "step": 38000
+    },
+    {
+      "epoch": 2.7964465044418696,
+      "grad_norm": 0.3671589195728302,
+      "learning_rate": 1.1884308457884908e-06,
+      "loss": 2.868,
+      "step": 38010
+    },
+    {
+      "epoch": 2.797182217807942,
+      "grad_norm": 0.3535459041595459,
+      "learning_rate": 1.1798871385353238e-06,
+      "loss": 2.8428,
+      "step": 38020
+    },
+    {
+      "epoch": 2.7979179311740143,
+      "grad_norm": 0.3631500005722046,
+      "learning_rate": 1.1713738861197531e-06,
+      "loss": 2.8092,
+      "step": 38030
+    },
+    {
+      "epoch": 2.798653644540087,
+      "grad_norm": 0.35819336771965027,
+      "learning_rate": 1.1628910938525028e-06,
+      "loss": 2.8299,
+      "step": 38040
+    },
+    {
+      "epoch": 2.7993893579061595,
+      "grad_norm": 0.3481830656528473,
+      "learning_rate": 1.1544387670252787e-06,
+      "loss": 2.8546,
+      "step": 38050
+    },
+    {
+      "epoch": 2.800125071272232,
+      "grad_norm": 0.3491680324077606,
+      "learning_rate": 1.146016910910791e-06,
+      "loss": 2.8112,
+      "step": 38060
+    },
+    {
+      "epoch": 2.8008607846383047,
+      "grad_norm": 0.35505613684654236,
+      "learning_rate": 1.137625530762737e-06,
+      "loss": 2.8269,
+      "step": 38070
+    },
+    {
+      "epoch": 2.8015964980043773,
+      "grad_norm": 0.3539564311504364,
+      "learning_rate": 1.1292646318158129e-06,
+      "loss": 2.8171,
+      "step": 38080
+    },
+    {
+      "epoch": 2.80233221137045,
+      "grad_norm": 0.3565945625305176,
+      "learning_rate": 1.1209342192856852e-06,
+      "loss": 2.8673,
+      "step": 38090
+    },
+    {
+      "epoch": 2.8030679247365224,
+      "grad_norm": 0.3503715991973877,
+      "learning_rate": 1.1126342983690252e-06,
+      "loss": 2.7728,
+      "step": 38100
+    },
+    {
+      "epoch": 2.803803638102595,
+      "grad_norm": 0.3498539328575134,
+      "learning_rate": 1.1043648742434576e-06,
+      "loss": 2.8908,
+      "step": 38110
+    },
+    {
+      "epoch": 2.8045393514686676,
+      "grad_norm": 0.3556746542453766,
+      "learning_rate": 1.0961259520675893e-06,
+      "loss": 2.8453,
+      "step": 38120
+    },
+    {
+      "epoch": 2.80527506483474,
+      "grad_norm": 0.35865160822868347,
+      "learning_rate": 1.087917536981009e-06,
+      "loss": 2.8408,
+      "step": 38130
+    },
+    {
+      "epoch": 2.806010778200813,
+      "grad_norm": 0.35179319977760315,
+      "learning_rate": 1.0797396341042765e-06,
+      "loss": 2.8286,
+      "step": 38140
+    },
+    {
+      "epoch": 2.8067464915668854,
+      "grad_norm": 0.34916627407073975,
+      "learning_rate": 1.0715922485388996e-06,
+      "loss": 2.7915,
+      "step": 38150
+    },
+    {
+      "epoch": 2.807482204932958,
+      "grad_norm": 0.3624122440814972,
+      "learning_rate": 1.0634753853673574e-06,
+      "loss": 2.8417,
+      "step": 38160
+    },
+    {
+      "epoch": 2.8082179182990306,
+      "grad_norm": 0.3545166850090027,
+      "learning_rate": 1.0553890496530938e-06,
+      "loss": 2.7966,
+      "step": 38170
+    },
+    {
+      "epoch": 2.808953631665103,
+      "grad_norm": 0.35965290665626526,
+      "learning_rate": 1.0473332464405183e-06,
+      "loss": 2.8361,
+      "step": 38180
+    },
+    {
+      "epoch": 2.8096893450311757,
+      "grad_norm": 0.349408894777298,
+      "learning_rate": 1.0393079807549667e-06,
+      "loss": 2.8462,
+      "step": 38190
+    },
+    {
+      "epoch": 2.8104250583972483,
+      "grad_norm": 0.3517548739910126,
+      "learning_rate": 1.0313132576027451e-06,
+      "loss": 2.7889,
+      "step": 38200
+    },
+    {
+      "epoch": 2.811160771763321,
+      "grad_norm": 0.35395148396492004,
+      "learning_rate": 1.0233490819710922e-06,
+      "loss": 2.8348,
+      "step": 38210
+    },
+    {
+      "epoch": 2.8118964851293935,
+      "grad_norm": 0.35165607929229736,
+      "learning_rate": 1.0154154588282172e-06,
+      "loss": 2.8439,
+      "step": 38220
+    },
+    {
+      "epoch": 2.812632198495466,
+      "grad_norm": 0.35307514667510986,
+      "learning_rate": 1.0075123931232444e-06,
+      "loss": 2.859,
+      "step": 38230
+    },
+    {
+      "epoch": 2.8133679118615387,
+      "grad_norm": 0.35489621758461,
+      "learning_rate": 9.996398897862358e-07,
+      "loss": 2.8215,
+      "step": 38240
+    },
+    {
+      "epoch": 2.8141036252276113,
+      "grad_norm": 0.35394665598869324,
+      "learning_rate": 9.91797953728202e-07,
+      "loss": 2.8703,
+      "step": 38250
+    },
+    {
+      "epoch": 2.814839338593684,
+      "grad_norm": 0.35619476437568665,
+      "learning_rate": 9.839865898410848e-07,
+      "loss": 2.8225,
+      "step": 38260
+    },
+    {
+      "epoch": 2.8155750519597564,
+      "grad_norm": 0.36469021439552307,
+      "learning_rate": 9.762058029977427e-07,
+      "loss": 2.8025,
+      "step": 38270
+    },
+    {
+      "epoch": 2.816310765325829,
+      "grad_norm": 0.3537813425064087,
+      "learning_rate": 9.684555980519706e-07,
+      "loss": 2.8601,
+      "step": 38280
+    },
+    {
+      "epoch": 2.8170464786919016,
+      "grad_norm": 0.3518713414669037,
+      "learning_rate": 9.607359798384785e-07,
+      "loss": 2.8474,
+      "step": 38290
+    },
+    {
+      "epoch": 2.817782192057974,
+      "grad_norm": 0.35041460394859314,
+      "learning_rate": 9.53046953172898e-07,
+      "loss": 2.8262,
+      "step": 38300
+    },
+    {
+      "epoch": 2.818517905424047,
+      "grad_norm": 0.3520907759666443,
+      "learning_rate": 9.453885228517812e-07,
+      "loss": 2.8327,
+      "step": 38310
+    },
+    {
+      "epoch": 2.8192536187901194,
+      "grad_norm": 0.3585839569568634,
+      "learning_rate": 9.377606936525895e-07,
+      "loss": 2.8728,
+      "step": 38320
+    },
+    {
+      "epoch": 2.819989332156192,
+      "grad_norm": 0.35138970613479614,
+      "learning_rate": 9.301634703336948e-07,
+      "loss": 2.8774,
+      "step": 38330
+    },
+    {
+      "epoch": 2.8207250455222646,
+      "grad_norm": 0.3506159484386444,
+      "learning_rate": 9.225968576343724e-07,
+      "loss": 2.8463,
+      "step": 38340
+    },
+    {
+      "epoch": 2.821460758888337,
+      "grad_norm": 0.35272055864334106,
+      "learning_rate": 9.150608602748189e-07,
+      "loss": 2.8586,
+      "step": 38350
+    },
+    {
+      "epoch": 2.8221964722544097,
+      "grad_norm": 0.3567464351654053,
+      "learning_rate": 9.075554829561128e-07,
+      "loss": 2.7923,
+      "step": 38360
+    },
+    {
+      "epoch": 2.8229321856204823,
+      "grad_norm": 0.3551969826221466,
+      "learning_rate": 9.000807303602421e-07,
+      "loss": 2.8759,
+      "step": 38370
+    },
+    {
+      "epoch": 2.823667898986555,
+      "grad_norm": 0.35370033979415894,
+      "learning_rate": 8.926366071500824e-07,
+      "loss": 2.8418,
+      "step": 38380
+    },
+    {
+      "epoch": 2.8244036123526275,
+      "grad_norm": 0.35376280546188354,
+      "learning_rate": 8.852231179694137e-07,
+      "loss": 2.8926,
+      "step": 38390
+    },
+    {
+      "epoch": 2.8251393257187,
+      "grad_norm": 0.36275237798690796,
+      "learning_rate": 8.778402674429032e-07,
+      "loss": 2.8476,
+      "step": 38400
+    },
+    {
+      "epoch": 2.8258750390847727,
+      "grad_norm": 0.3525364100933075,
+      "learning_rate": 8.70488060176089e-07,
+      "loss": 2.8483,
+      "step": 38410
+    },
+    {
+      "epoch": 2.8266107524508453,
+      "grad_norm": 0.3549235165119171,
+      "learning_rate": 8.631665007554135e-07,
+      "loss": 2.8744,
+      "step": 38420
+    },
+    {
+      "epoch": 2.827346465816918,
+      "grad_norm": 0.3512353301048279,
+      "learning_rate": 8.558755937481954e-07,
+      "loss": 2.8504,
+      "step": 38430
+    },
+    {
+      "epoch": 2.8280821791829904,
+      "grad_norm": 0.36531877517700195,
+      "learning_rate": 8.486153437026301e-07,
+      "loss": 2.8703,
+      "step": 38440
+    },
+    {
+      "epoch": 2.828817892549063,
+      "grad_norm": 0.354703426361084,
+      "learning_rate": 8.413857551477888e-07,
+      "loss": 2.8316,
+      "step": 38450
+    },
+    {
+      "epoch": 2.8295536059151356,
+      "grad_norm": 0.36544790863990784,
+      "learning_rate": 8.341868325936086e-07,
+      "loss": 2.8486,
+      "step": 38460
+    },
+    {
+      "epoch": 2.830289319281208,
+      "grad_norm": 0.35509875416755676,
+      "learning_rate": 8.270185805308972e-07,
+      "loss": 2.8235,
+      "step": 38470
+    },
+    {
+      "epoch": 2.831025032647281,
+      "grad_norm": 0.3595902621746063,
+      "learning_rate": 8.198810034313553e-07,
+      "loss": 2.8281,
+      "step": 38480
+    },
+    {
+      "epoch": 2.8317607460133534,
+      "grad_norm": 0.35145944356918335,
+      "learning_rate": 8.127741057475157e-07,
+      "loss": 2.8321,
+      "step": 38490
+    },
+    {
+      "epoch": 2.832496459379426,
+      "grad_norm": 0.34478214383125305,
+      "learning_rate": 8.056978919127766e-07,
+      "loss": 2.8791,
+      "step": 38500
+    },
+    {
+      "epoch": 2.8332321727454985,
+      "grad_norm": 0.3576992452144623,
+      "learning_rate": 7.986523663414236e-07,
+      "loss": 2.8724,
+      "step": 38510
+    },
+    {
+      "epoch": 2.833967886111571,
+      "grad_norm": 0.3584095239639282,
+      "learning_rate": 7.916375334285631e-07,
+      "loss": 2.8297,
+      "step": 38520
+    },
+    {
+      "epoch": 2.8347035994776437,
+      "grad_norm": 0.3555443584918976,
+      "learning_rate": 7.846533975501724e-07,
+      "loss": 2.8311,
+      "step": 38530
+    },
+    {
+      "epoch": 2.8354393128437163,
+      "grad_norm": 0.34698283672332764,
+      "learning_rate": 7.776999630630833e-07,
+      "loss": 2.8396,
+      "step": 38540
+    },
+    {
+      "epoch": 2.836175026209789,
+      "grad_norm": 0.34753572940826416,
+      "learning_rate": 7.707772343049702e-07,
+      "loss": 2.8457,
+      "step": 38550
+    },
+    {
+      "epoch": 2.8369107395758615,
+      "grad_norm": 0.3490901589393616,
+      "learning_rate": 7.638852155943399e-07,
+      "loss": 2.85,
+      "step": 38560
+    },
+    {
+      "epoch": 2.837646452941934,
+      "grad_norm": 0.3558933734893799,
+      "learning_rate": 7.570239112305644e-07,
+      "loss": 2.8624,
+      "step": 38570
+    },
+    {
+      "epoch": 2.8383821663080067,
+      "grad_norm": 0.3518975079059601,
+      "learning_rate": 7.50193325493842e-07,
+      "loss": 2.8541,
+      "step": 38580
+    },
+    {
+      "epoch": 2.8391178796740792,
+      "grad_norm": 0.35525035858154297,
+      "learning_rate": 7.433934626452144e-07,
+      "loss": 2.8382,
+      "step": 38590
+    },
+    {
+      "epoch": 2.8398535930401514,
+      "grad_norm": 0.3507790267467499,
+      "learning_rate": 7.366243269265493e-07,
+      "loss": 2.7892,
+      "step": 38600
+    },
+    {
+      "epoch": 2.840589306406224,
+      "grad_norm": 0.3581407368183136,
+      "learning_rate": 7.298859225605526e-07,
+      "loss": 2.8945,
+      "step": 38610
+    },
+    {
+      "epoch": 2.8413250197722966,
+      "grad_norm": 0.35862547159194946,
+      "learning_rate": 7.231782537507614e-07,
+      "loss": 2.8035,
+      "step": 38620
+    },
+    {
+      "epoch": 2.842060733138369,
+      "grad_norm": 0.3510931134223938,
+      "learning_rate": 7.165013246815344e-07,
+      "loss": 2.8397,
+      "step": 38630
+    },
+    {
+      "epoch": 2.8427964465044417,
+      "grad_norm": 0.35147207975387573,
+      "learning_rate": 7.098551395180563e-07,
+      "loss": 2.8732,
+      "step": 38640
+    },
+    {
+      "epoch": 2.8435321598705143,
+      "grad_norm": 0.3580421805381775,
+      "learning_rate": 7.032397024063331e-07,
+      "loss": 2.8764,
+      "step": 38650
+    },
+    {
+      "epoch": 2.844267873236587,
+      "grad_norm": 0.3496814966201782,
+      "learning_rate": 6.966550174731912e-07,
+      "loss": 2.7946,
+      "step": 38660
+    },
+    {
+      "epoch": 2.8450035866026595,
+      "grad_norm": 0.35197022557258606,
+      "learning_rate": 6.901010888262671e-07,
+      "loss": 2.8215,
+      "step": 38670
+    },
+    {
+      "epoch": 2.845739299968732,
+      "grad_norm": 0.3523489832878113,
+      "learning_rate": 6.835779205540238e-07,
+      "loss": 2.8093,
+      "step": 38680
+    },
+    {
+      "epoch": 2.8464750133348047,
+      "grad_norm": 0.3464307188987732,
+      "learning_rate": 6.770855167257284e-07,
+      "loss": 2.8886,
+      "step": 38690
+    },
+    {
+      "epoch": 2.8472107267008773,
+      "grad_norm": 0.3425907492637634,
+      "learning_rate": 6.706238813914411e-07,
+      "loss": 2.841,
+      "step": 38700
+    },
+    {
+      "epoch": 2.84794644006695,
+      "grad_norm": 0.3567936420440674,
+      "learning_rate": 6.6419301858206e-07,
+      "loss": 2.8333,
+      "step": 38710
+    },
+    {
+      "epoch": 2.8486821534330224,
+      "grad_norm": 0.3508964776992798,
+      "learning_rate": 6.577929323092591e-07,
+      "loss": 2.8675,
+      "step": 38720
+    },
+    {
+      "epoch": 2.849417866799095,
+      "grad_norm": 0.3553392291069031,
+      "learning_rate": 6.514236265655227e-07,
+      "loss": 2.8257,
+      "step": 38730
+    },
+    {
+      "epoch": 2.8501535801651676,
+      "grad_norm": 0.35673439502716064,
+      "learning_rate": 6.450851053241391e-07,
+      "loss": 2.8007,
+      "step": 38740
+    },
+    {
+      "epoch": 2.85088929353124,
+      "grad_norm": 0.34796643257141113,
+      "learning_rate": 6.387773725391899e-07,
+      "loss": 2.8637,
+      "step": 38750
+    },
+    {
+      "epoch": 2.851625006897313,
+      "grad_norm": 0.3523545563220978,
+      "learning_rate": 6.325004321455441e-07,
+      "loss": 2.8764,
+      "step": 38760
+    },
+    {
+      "epoch": 2.8523607202633854,
+      "grad_norm": 0.35014915466308594,
+      "learning_rate": 6.262542880588751e-07,
+      "loss": 2.8449,
+      "step": 38770
+    },
+    {
+      "epoch": 2.853096433629458,
+      "grad_norm": 0.3441470265388489,
+      "learning_rate": 6.200389441756271e-07,
+      "loss": 2.8459,
+      "step": 38780
+    },
+    {
+      "epoch": 2.8538321469955306,
+      "grad_norm": 0.35210317373275757,
+      "learning_rate": 6.138544043730377e-07,
+      "loss": 2.8147,
+      "step": 38790
+    },
+    {
+      "epoch": 2.854567860361603,
+      "grad_norm": 0.3533152937889099,
+      "learning_rate": 6.077006725091427e-07,
+      "loss": 2.8468,
+      "step": 38800
+    },
+    {
+      "epoch": 2.8553035737276757,
+      "grad_norm": 0.36064669489860535,
+      "learning_rate": 6.015777524227384e-07,
+      "loss": 2.8763,
+      "step": 38810
+    },
+    {
+      "epoch": 2.8560392870937483,
+      "grad_norm": 0.3516472578048706,
+      "learning_rate": 5.954856479334137e-07,
+      "loss": 2.8836,
+      "step": 38820
+    },
+    {
+      "epoch": 2.856775000459821,
+      "grad_norm": 0.3468547463417053,
+      "learning_rate": 5.89424362841523e-07,
+      "loss": 2.8543,
+      "step": 38830
+    },
+    {
+      "epoch": 2.8575107138258935,
+      "grad_norm": 0.3474918603897095,
+      "learning_rate": 5.833939009282086e-07,
+      "loss": 2.829,
+      "step": 38840
+    },
+    {
+      "epoch": 2.858246427191966,
+      "grad_norm": 0.3458395302295685,
+      "learning_rate": 5.773942659553777e-07,
+      "loss": 2.8051,
+      "step": 38850
+    },
+    {
+      "epoch": 2.8589821405580387,
+      "grad_norm": 0.35283133387565613,
+      "learning_rate": 5.714254616657033e-07,
+      "loss": 2.902,
+      "step": 38860
+    },
+    {
+      "epoch": 2.8597178539241113,
+      "grad_norm": 0.3498219847679138,
+      "learning_rate": 5.654874917826292e-07,
+      "loss": 2.7938,
+      "step": 38870
+    },
+    {
+      "epoch": 2.860453567290184,
+      "grad_norm": 0.35753101110458374,
+      "learning_rate": 5.595803600103699e-07,
+      "loss": 2.8852,
+      "step": 38880
+    },
+    {
+      "epoch": 2.8611892806562564,
+      "grad_norm": 0.35117393732070923,
+      "learning_rate": 5.537040700338891e-07,
+      "loss": 2.8492,
+      "step": 38890
+    },
+    {
+      "epoch": 2.861924994022329,
+      "grad_norm": 0.35485750436782837,
+      "learning_rate": 5.478586255189266e-07,
+      "loss": 2.8477,
+      "step": 38900
+    },
+    {
+      "epoch": 2.8626607073884016,
+      "grad_norm": 0.3494305908679962,
+      "learning_rate": 5.420440301119656e-07,
+      "loss": 2.8455,
+      "step": 38910
+    },
+    {
+      "epoch": 2.863396420754474,
+      "grad_norm": 0.3479386568069458,
+      "learning_rate": 5.3626028744026e-07,
+      "loss": 2.8111,
+      "step": 38920
+    },
+    {
+      "epoch": 2.864132134120547,
+      "grad_norm": 0.3553057312965393,
+      "learning_rate": 5.305074011118072e-07,
+      "loss": 2.8335,
+      "step": 38930
+    },
+    {
+      "epoch": 2.8648678474866194,
+      "grad_norm": 0.35610923171043396,
+      "learning_rate": 5.24785374715353e-07,
+      "loss": 2.8233,
+      "step": 38940
+    },
+    {
+      "epoch": 2.8656035608526915,
+      "grad_norm": 0.35471856594085693,
+      "learning_rate": 5.190942118203979e-07,
+      "loss": 2.8949,
+      "step": 38950
+    },
+    {
+      "epoch": 2.866339274218764,
+      "grad_norm": 0.35570481419563293,
+      "learning_rate": 5.134339159771961e-07,
+      "loss": 2.9046,
+      "step": 38960
+    },
+    {
+      "epoch": 2.8670749875848367,
+      "grad_norm": 0.3420692980289459,
+      "learning_rate": 5.07804490716729e-07,
+      "loss": 2.8506,
+      "step": 38970
+    },
+    {
+      "epoch": 2.8678107009509093,
+      "grad_norm": 0.3660537004470825,
+      "learning_rate": 5.022059395507372e-07,
+      "loss": 2.8872,
+      "step": 38980
+    },
+    {
+      "epoch": 2.868546414316982,
+      "grad_norm": 0.3597838878631592,
+      "learning_rate": 4.966382659716939e-07,
+      "loss": 2.8929,
+      "step": 38990
+    },
+    {
+      "epoch": 2.8692821276830545,
+      "grad_norm": 0.3510751724243164,
+      "learning_rate": 4.911014734528096e-07,
+      "loss": 2.8589,
+      "step": 39000
+    },
+    {
+      "epoch": 2.870017841049127,
+      "grad_norm": 0.35269594192504883,
+      "learning_rate": 4.855955654480271e-07,
+      "loss": 2.8047,
+      "step": 39010
+    },
+    {
+      "epoch": 2.8707535544151996,
+      "grad_norm": 0.35643744468688965,
+      "learning_rate": 4.801205453920376e-07,
+      "loss": 2.8908,
+      "step": 39020
+    },
+    {
+      "epoch": 2.8714892677812722,
+      "grad_norm": 0.35601240396499634,
+      "learning_rate": 4.746764167002482e-07,
+      "loss": 2.8376,
+      "step": 39030
+    },
+    {
+      "epoch": 2.872224981147345,
+      "grad_norm": 0.35460686683654785,
+      "learning_rate": 4.6926318276879766e-07,
+      "loss": 2.8776,
+      "step": 39040
+    },
+    {
+      "epoch": 2.8729606945134174,
+      "grad_norm": 0.347586989402771,
+      "learning_rate": 4.6388084697456815e-07,
+      "loss": 2.8175,
+      "step": 39050
+    },
+    {
+      "epoch": 2.87369640787949,
+      "grad_norm": 0.3580293655395508,
+      "learning_rate": 4.585294126751349e-07,
+      "loss": 2.849,
+      "step": 39060
+    },
+    {
+      "epoch": 2.8744321212455626,
+      "grad_norm": 0.34969639778137207,
+      "learning_rate": 4.532088832088388e-07,
+      "loss": 2.8542,
+      "step": 39070
+    },
+    {
+      "epoch": 2.875167834611635,
+      "grad_norm": 0.35449376702308655,
+      "learning_rate": 4.47919261894697e-07,
+      "loss": 2.8449,
+      "step": 39080
+    },
+    {
+      "epoch": 2.8759035479777078,
+      "grad_norm": 0.35622820258140564,
+      "learning_rate": 4.426605520324867e-07,
+      "loss": 2.8648,
+      "step": 39090
+    },
+    {
+      "epoch": 2.8766392613437803,
+      "grad_norm": 0.3581145107746124,
+      "learning_rate": 4.374327569026726e-07,
+      "loss": 2.8235,
+      "step": 39100
+    },
+    {
+      "epoch": 2.877374974709853,
+      "grad_norm": 0.36063528060913086,
+      "learning_rate": 4.322358797664405e-07,
+      "loss": 2.8551,
+      "step": 39110
+    },
+    {
+      "epoch": 2.8781106880759255,
+      "grad_norm": 0.3666221797466278,
+      "learning_rate": 4.2706992386570834e-07,
+      "loss": 2.917,
+      "step": 39120
+    },
+    {
+      "epoch": 2.878846401441998,
+      "grad_norm": 0.34821319580078125,
+      "learning_rate": 4.219348924230704e-07,
+      "loss": 2.8171,
+      "step": 39130
+    },
+    {
+      "epoch": 2.8795821148080707,
+      "grad_norm": 0.3506257236003876,
+      "learning_rate": 4.168307886418643e-07,
+      "loss": 2.8255,
+      "step": 39140
+    },
+    {
+      "epoch": 2.8803178281741433,
+      "grad_norm": 0.3522619903087616,
+      "learning_rate": 4.117576157061098e-07,
+      "loss": 2.8204,
+      "step": 39150
+    },
+    {
+      "epoch": 2.881053541540216,
+      "grad_norm": 0.3574064373970032,
+      "learning_rate": 4.0671537678054207e-07,
+      "loss": 2.8652,
+      "step": 39160
+    },
+    {
+      "epoch": 2.8817892549062885,
+      "grad_norm": 0.3573540449142456,
+      "learning_rate": 4.017040750106005e-07,
+      "loss": 2.8102,
+      "step": 39170
+    },
+    {
+      "epoch": 2.882524968272361,
+      "grad_norm": 0.3595264256000519,
+      "learning_rate": 3.9672371352242334e-07,
+      "loss": 2.8323,
+      "step": 39180
+    },
+    {
+      "epoch": 2.8832606816384336,
+      "grad_norm": 0.364810973405838,
+      "learning_rate": 3.9177429542284206e-07,
+      "loss": 2.8264,
+      "step": 39190
+    },
+    {
+      "epoch": 2.883996395004506,
+      "grad_norm": 0.3523925840854645,
+      "learning_rate": 3.86855823799398e-07,
+      "loss": 2.8437,
+      "step": 39200
+    },
+    {
+      "epoch": 2.884732108370579,
+      "grad_norm": 0.36235326528549194,
+      "learning_rate": 3.8196830172030904e-07,
+      "loss": 2.9125,
+      "step": 39210
+    },
+    {
+      "epoch": 2.8854678217366514,
+      "grad_norm": 0.35653936862945557,
+      "learning_rate": 3.7711173223450303e-07,
+      "loss": 2.8476,
+      "step": 39220
+    },
+    {
+      "epoch": 2.886203535102724,
+      "grad_norm": 0.35905417799949646,
+      "learning_rate": 3.722861183715953e-07,
+      "loss": 2.8871,
+      "step": 39230
+    },
+    {
+      "epoch": 2.8869392484687966,
+      "grad_norm": 0.34990420937538147,
+      "learning_rate": 3.6749146314187244e-07,
+      "loss": 2.8383,
+      "step": 39240
+    },
+    {
+      "epoch": 2.887674961834869,
+      "grad_norm": 0.3606944978237152,
+      "learning_rate": 3.627277695363474e-07,
+      "loss": 2.8516,
+      "step": 39250
+    },
+    {
+      "epoch": 2.8884106752009417,
+      "grad_norm": 0.35960420966148376,
+      "learning_rate": 3.5799504052668187e-07,
+      "loss": 2.8123,
+      "step": 39260
+    },
+    {
+      "epoch": 2.8891463885670143,
+      "grad_norm": 0.35482460260391235,
+      "learning_rate": 3.5329327906523656e-07,
+      "loss": 2.8181,
+      "step": 39270
+    },
+    {
+      "epoch": 2.889882101933087,
+      "grad_norm": 0.352066308259964,
+      "learning_rate": 3.4862248808504304e-07,
+      "loss": 2.8055,
+      "step": 39280
+    },
+    {
+      "epoch": 2.8906178152991595,
+      "grad_norm": 0.34272414445877075,
+      "learning_rate": 3.439826704998428e-07,
+      "loss": 2.8383,
+      "step": 39290
+    },
+    {
+      "epoch": 2.891353528665232,
+      "grad_norm": 0.35531747341156006,
+      "learning_rate": 3.393738292040205e-07,
+      "loss": 2.8043,
+      "step": 39300
+    },
+    {
+      "epoch": 2.8920892420313047,
+      "grad_norm": 0.3573128581047058,
+      "learning_rate": 3.347959670726486e-07,
+      "loss": 2.8592,
+      "step": 39310
+    },
+    {
+      "epoch": 2.8928249553973773,
+      "grad_norm": 0.35831108689308167,
+      "learning_rate": 3.3024908696148716e-07,
+      "loss": 2.8953,
+      "step": 39320
+    },
+    {
+      "epoch": 2.89356066876345,
+      "grad_norm": 0.3587474226951599,
+      "learning_rate": 3.2573319170695617e-07,
+      "loss": 2.8696,
+      "step": 39330
+    },
+    {
+      "epoch": 2.8942963821295224,
+      "grad_norm": 0.35401877760887146,
+      "learning_rate": 3.212482841261466e-07,
+      "loss": 2.8271,
+      "step": 39340
+    },
+    {
+      "epoch": 2.895032095495595,
+      "grad_norm": 0.35091838240623474,
+      "learning_rate": 3.1679436701683164e-07,
+      "loss": 2.8625,
+      "step": 39350
+    },
+    {
+      "epoch": 2.8957678088616676,
+      "grad_norm": 0.34690558910369873,
+      "learning_rate": 3.123714431574276e-07,
+      "loss": 2.8423,
+      "step": 39360
+    },
+    {
+      "epoch": 2.89650352222774,
+      "grad_norm": 0.35382506251335144,
+      "learning_rate": 3.079795153070386e-07,
+      "loss": 2.8448,
+      "step": 39370
+    },
+    {
+      "epoch": 2.897239235593813,
+      "grad_norm": 0.3465462327003479,
+      "learning_rate": 3.0361858620543415e-07,
+      "loss": 2.8492,
+      "step": 39380
+    },
+    {
+      "epoch": 2.8979749489598854,
+      "grad_norm": 0.35538744926452637,
+      "learning_rate": 2.9928865857302147e-07,
+      "loss": 2.793,
+      "step": 39390
+    },
+    {
+      "epoch": 2.898710662325958,
+      "grad_norm": 0.350984662771225,
+      "learning_rate": 2.94989735110901e-07,
+      "loss": 2.8552,
+      "step": 39400
+    },
+    {
+      "epoch": 2.8994463756920306,
+      "grad_norm": 0.34965720772743225,
+      "learning_rate": 2.9072181850079984e-07,
+      "loss": 2.8402,
+      "step": 39410
+    },
+    {
+      "epoch": 2.900182089058103,
+      "grad_norm": 0.36526912450790405,
+      "learning_rate": 2.8648491140513266e-07,
+      "loss": 2.8454,
+      "step": 39420
+    },
+    {
+      "epoch": 2.9009178024241757,
+      "grad_norm": 0.3539934456348419,
+      "learning_rate": 2.82279016466952e-07,
+      "loss": 2.8569,
+      "step": 39430
+    },
+    {
+      "epoch": 2.9016535157902483,
+      "grad_norm": 0.3533295691013336,
+      "learning_rate": 2.7810413630996456e-07,
+      "loss": 2.8818,
+      "step": 39440
+    },
+    {
+      "epoch": 2.902389229156321,
+      "grad_norm": 0.353777676820755,
+      "learning_rate": 2.739602735385427e-07,
+      "loss": 2.8356,
+      "step": 39450
+    },
+    {
+      "epoch": 2.9031249425223935,
+      "grad_norm": 0.35035479068756104,
+      "learning_rate": 2.698474307376853e-07,
+      "loss": 2.8345,
+      "step": 39460
+    },
+    {
+      "epoch": 2.903860655888466,
+      "grad_norm": 0.3570537567138672,
+      "learning_rate": 2.657656104730677e-07,
+      "loss": 2.8308,
+      "step": 39470
+    },
+    {
+      "epoch": 2.9045963692545387,
+      "grad_norm": 0.3505394458770752,
+      "learning_rate": 2.617148152909976e-07,
+      "loss": 2.8348,
+      "step": 39480
+    },
+    {
+      "epoch": 2.9053320826206113,
+      "grad_norm": 0.35672879219055176,
+      "learning_rate": 2.576950477184259e-07,
+      "loss": 2.8989,
+      "step": 39490
+    },
+    {
+      "epoch": 2.906067795986684,
+      "grad_norm": 0.3504009246826172,
+      "learning_rate": 2.537063102629633e-07,
+      "loss": 2.7988,
+      "step": 39500
+    },
+    {
+      "epoch": 2.9068035093527564,
+      "grad_norm": 0.3520059287548065,
+      "learning_rate": 2.497486054128473e-07,
+      "loss": 2.8817,
+      "step": 39510
+    },
+    {
+      "epoch": 2.9075392227188286,
+      "grad_norm": 0.35659059882164,
+      "learning_rate": 2.4582193563696956e-07,
+      "loss": 2.7797,
+      "step": 39520
+    },
+    {
+      "epoch": 2.908274936084901,
+      "grad_norm": 0.3549395501613617,
+      "learning_rate": 2.419263033848429e-07,
+      "loss": 2.8305,
+      "step": 39530
+    },
+    {
+      "epoch": 2.9090106494509738,
+      "grad_norm": 0.3504616320133209,
+      "learning_rate": 2.3806171108664566e-07,
+      "loss": 2.8622,
+      "step": 39540
+    },
+    {
+      "epoch": 2.9097463628170464,
+      "grad_norm": 0.35603848099708557,
+      "learning_rate": 2.3422816115316603e-07,
+      "loss": 2.8385,
+      "step": 39550
+    },
+    {
+      "epoch": 2.910482076183119,
+      "grad_norm": 0.3515443205833435,
+      "learning_rate": 2.3042565597584108e-07,
+      "loss": 2.8218,
+      "step": 39560
+    },
+    {
+      "epoch": 2.9112177895491915,
+      "grad_norm": 0.35419803857803345,
+      "learning_rate": 2.266541979267456e-07,
+      "loss": 2.8352,
+      "step": 39570
+    },
+    {
+      "epoch": 2.911953502915264,
+      "grad_norm": 0.36014023423194885,
+      "learning_rate": 2.2291378935858087e-07,
+      "loss": 2.8434,
+      "step": 39580
+    },
+    {
+      "epoch": 2.9126892162813367,
+      "grad_norm": 0.3540354073047638,
+      "learning_rate": 2.1920443260466939e-07,
+      "loss": 2.8345,
+      "step": 39590
+    },
+    {
+      "epoch": 2.9134249296474093,
+      "grad_norm": 0.3401762843132019,
+      "learning_rate": 2.1552612997898235e-07,
+      "loss": 2.8594,
+      "step": 39600
+    },
+    {
+      "epoch": 2.914160643013482,
+      "grad_norm": 0.35671165585517883,
+      "learning_rate": 2.1187888377610655e-07,
+      "loss": 2.8254,
+      "step": 39610
+    },
+    {
+      "epoch": 2.9148963563795545,
+      "grad_norm": 0.34946754574775696,
+      "learning_rate": 2.0826269627126082e-07,
+      "loss": 2.8562,
+      "step": 39620
+    },
+    {
+      "epoch": 2.915632069745627,
+      "grad_norm": 0.35555514693260193,
+      "learning_rate": 2.0467756972027407e-07,
+      "loss": 2.8628,
+      "step": 39630
+    },
+    {
+      "epoch": 2.9163677831116996,
+      "grad_norm": 0.35324206948280334,
+      "learning_rate": 2.0112350635962396e-07,
+      "loss": 2.8384,
+      "step": 39640
+    },
+    {
+      "epoch": 2.9171034964777722,
+      "grad_norm": 0.348349928855896,
+      "learning_rate": 1.97600508406387e-07,
+      "loss": 2.8154,
+      "step": 39650
+    },
+    {
+      "epoch": 2.917839209843845,
+      "grad_norm": 0.3587605059146881,
+      "learning_rate": 1.9410857805828297e-07,
+      "loss": 2.8491,
+      "step": 39660
+    },
+    {
+      "epoch": 2.9185749232099174,
+      "grad_norm": 0.3533768355846405,
+      "learning_rate": 1.9064771749362497e-07,
+      "loss": 2.8414,
+      "step": 39670
+    },
+    {
+      "epoch": 2.91931063657599,
+      "grad_norm": 0.3447302281856537,
+      "learning_rate": 1.8721792887136935e-07,
+      "loss": 2.8584,
+      "step": 39680
+    },
+    {
+      "epoch": 2.9200463499420626,
+      "grad_norm": 0.3582588732242584,
+      "learning_rate": 1.838192143310713e-07,
+      "loss": 2.8671,
+      "step": 39690
+    },
+    {
+      "epoch": 2.920782063308135,
+      "grad_norm": 0.35392409563064575,
+      "learning_rate": 1.8045157599291261e-07,
+      "loss": 2.8308,
+      "step": 39700
+    },
+    {
+      "epoch": 2.9215177766742078,
+      "grad_norm": 0.34092825651168823,
+      "learning_rate": 1.771150159576851e-07,
+      "loss": 2.8555,
+      "step": 39710
+    },
+    {
+      "epoch": 2.9222534900402803,
+      "grad_norm": 0.34853097796440125,
+      "learning_rate": 1.7380953630678488e-07,
+      "loss": 2.8566,
+      "step": 39720
+    },
+    {
+      "epoch": 2.922989203406353,
+      "grad_norm": 0.3468574285507202,
+      "learning_rate": 1.7053513910223472e-07,
+      "loss": 2.8674,
+      "step": 39730
+    },
+    {
+      "epoch": 2.9237249167724255,
+      "grad_norm": 0.35896000266075134,
+      "learning_rate": 1.6729182638666185e-07,
+      "loss": 2.8809,
+      "step": 39740
+    },
+    {
+      "epoch": 2.924460630138498,
+      "grad_norm": 0.3523907959461212,
+      "learning_rate": 1.6407960018329783e-07,
+      "loss": 2.8452,
+      "step": 39750
+    },
+    {
+      "epoch": 2.9251963435045707,
+      "grad_norm": 0.3549048602581024,
+      "learning_rate": 1.6089846249598972e-07,
+      "loss": 2.8206,
+      "step": 39760
+    },
+    {
+      "epoch": 2.9259320568706433,
+      "grad_norm": 0.35478678345680237,
+      "learning_rate": 1.5774841530917795e-07,
+      "loss": 2.8355,
+      "step": 39770
+    },
+    {
+      "epoch": 2.926667770236716,
+      "grad_norm": 0.3501017391681671,
+      "learning_rate": 1.5462946058792393e-07,
+      "loss": 2.8333,
+      "step": 39780
+    },
+    {
+      "epoch": 2.9274034836027885,
+      "grad_norm": 0.3514855206012726,
+      "learning_rate": 1.51541600277888e-07,
+      "loss": 2.8972,
+      "step": 39790
+    },
+    {
+      "epoch": 2.928139196968861,
+      "grad_norm": 0.3615357577800751,
+      "learning_rate": 1.4848483630532373e-07,
+      "loss": 2.8969,
+      "step": 39800
+    },
+    {
+      "epoch": 2.9288749103349336,
+      "grad_norm": 0.35782158374786377,
+      "learning_rate": 1.4545917057710024e-07,
+      "loss": 2.8159,
+      "step": 39810
+    },
+    {
+      "epoch": 2.9296106237010062,
+      "grad_norm": 0.3593333065509796,
+      "learning_rate": 1.4246460498067991e-07,
+      "loss": 2.9038,
+      "step": 39820
+    },
+    {
+      "epoch": 2.930346337067079,
+      "grad_norm": 0.3555089831352234,
+      "learning_rate": 1.3950114138411852e-07,
+      "loss": 2.8451,
+      "step": 39830
+    },
+    {
+      "epoch": 2.9310820504331514,
+      "grad_norm": 0.3489868938922882,
+      "learning_rate": 1.3656878163608168e-07,
+      "loss": 2.8836,
+      "step": 39840
+    },
+    {
+      "epoch": 2.931817763799224,
+      "grad_norm": 0.35483691096305847,
+      "learning_rate": 1.3366752756582834e-07,
+      "loss": 2.8286,
+      "step": 39850
+    },
+    {
+      "epoch": 2.932553477165296,
+      "grad_norm": 0.3441714942455292,
+      "learning_rate": 1.307973809832108e-07,
+      "loss": 2.7443,
+      "step": 39860
+    },
+    {
+      "epoch": 2.9332891905313687,
+      "grad_norm": 0.3490469753742218,
+      "learning_rate": 1.279583436786691e-07,
+      "loss": 2.8256,
+      "step": 39870
+    },
+    {
+      "epoch": 2.9340249038974413,
+      "grad_norm": 0.3525741398334503,
+      "learning_rate": 1.2515041742325316e-07,
+      "loss": 2.8317,
+      "step": 39880
+    },
+    {
+      "epoch": 2.934760617263514,
+      "grad_norm": 0.3526895344257355,
+      "learning_rate": 1.2237360396858966e-07,
+      "loss": 2.8599,
+      "step": 39890
+    },
+    {
+      "epoch": 2.9354963306295865,
+      "grad_norm": 0.3478415310382843,
+      "learning_rate": 1.1962790504690958e-07,
+      "loss": 2.8683,
+      "step": 39900
+    },
+    {
+      "epoch": 2.936232043995659,
+      "grad_norm": 0.3627457022666931,
+      "learning_rate": 1.1691332237102615e-07,
+      "loss": 2.8701,
+      "step": 39910
+    },
+    {
+      "epoch": 2.9369677573617317,
+      "grad_norm": 0.3554377555847168,
+      "learning_rate": 1.142298576343348e-07,
+      "loss": 2.8361,
+      "step": 39920
+    },
+    {
+      "epoch": 2.9377034707278042,
+      "grad_norm": 0.35778379440307617,
+      "learning_rate": 1.1157751251084092e-07,
+      "loss": 2.8511,
+      "step": 39930
+    },
+    {
+      "epoch": 2.938439184093877,
+      "grad_norm": 0.35034841299057007,
+      "learning_rate": 1.0895628865512097e-07,
+      "loss": 2.8484,
+      "step": 39940
+    },
+    {
+      "epoch": 2.9391748974599494,
+      "grad_norm": 0.3462885320186615,
+      "learning_rate": 1.0636618770232809e-07,
+      "loss": 2.8193,
+      "step": 39950
+    },
+    {
+      "epoch": 2.939910610826022,
+      "grad_norm": 0.3555814027786255,
+      "learning_rate": 1.0380721126822534e-07,
+      "loss": 2.8553,
+      "step": 39960
+    },
+    {
+      "epoch": 2.9406463241920946,
+      "grad_norm": 0.34871914982795715,
+      "learning_rate": 1.0127936094914136e-07,
+      "loss": 2.7886,
+      "step": 39970
+    },
+    {
+      "epoch": 2.941382037558167,
+      "grad_norm": 0.3572273254394531,
+      "learning_rate": 9.878263832199252e-08,
+      "loss": 2.8278,
+      "step": 39980
+    },
+    {
+      "epoch": 2.9421177509242398,
+      "grad_norm": 0.3569453954696655,
+      "learning_rate": 9.631704494428295e-08,
+      "loss": 2.8778,
+      "step": 39990
+    },
+    {
+      "epoch": 2.9428534642903124,
+      "grad_norm": 0.3541705906391144,
+      "learning_rate": 9.388258235409342e-08,
+      "loss": 2.8256,
+      "step": 40000
+    },
+    {
+      "epoch": 2.943589177656385,
+      "grad_norm": 0.35455265641212463,
+      "learning_rate": 9.147925207007024e-08,
+      "loss": 2.8327,
+      "step": 40010
+    },
+    {
+      "epoch": 2.9443248910224575,
+      "grad_norm": 0.34613925218582153,
+      "learning_rate": 8.91070555914697e-08,
+      "loss": 2.8643,
+      "step": 40020
+    },
+    {
+      "epoch": 2.94506060438853,
+      "grad_norm": 0.3521136939525604,
+      "learning_rate": 8.676599439810251e-08,
+      "loss": 2.8541,
+      "step": 40030
+    },
+    {
+      "epoch": 2.9457963177546027,
+      "grad_norm": 0.3618794083595276,
+      "learning_rate": 8.44560699503616e-08,
+      "loss": 2.8981,
+      "step": 40040
+    },
+    {
+      "epoch": 2.9465320311206753,
+      "grad_norm": 0.35057854652404785,
+      "learning_rate": 8.217728368921651e-08,
+      "loss": 2.8548,
+      "step": 40050
+    },
+    {
+      "epoch": 2.947267744486748,
+      "grad_norm": 0.3577747344970703,
+      "learning_rate": 7.992963703621903e-08,
+      "loss": 2.7709,
+      "step": 40060
+    },
+    {
+      "epoch": 2.9480034578528205,
+      "grad_norm": 0.35435280203819275,
+      "learning_rate": 7.771313139349202e-08,
+      "loss": 2.8209,
+      "step": 40070
+    },
+    {
+      "epoch": 2.948739171218893,
+      "grad_norm": 0.3446013927459717,
+      "learning_rate": 7.552776814372387e-08,
+      "loss": 2.8219,
+      "step": 40080
+    },
+    {
+      "epoch": 2.9494748845849657,
+      "grad_norm": 0.352539986371994,
+      "learning_rate": 7.33735486501852e-08,
+      "loss": 2.8298,
+      "step": 40090
+    },
+    {
+      "epoch": 2.9502105979510382,
+      "grad_norm": 0.3450770378112793,
+      "learning_rate": 7.125047425671772e-08,
+      "loss": 2.8437,
+      "step": 40100
+    },
+    {
+      "epoch": 2.950946311317111,
+      "grad_norm": 0.34738028049468994,
+      "learning_rate": 6.915854628772866e-08,
+      "loss": 2.8432,
+      "step": 40110
+    },
+    {
+      "epoch": 2.9516820246831834,
+      "grad_norm": 0.3545306324958801,
+      "learning_rate": 6.709776604820195e-08,
+      "loss": 2.8129,
+      "step": 40120
+    },
+    {
+      "epoch": 2.952417738049256,
+      "grad_norm": 0.3510771691799164,
+      "learning_rate": 6.5068134823687e-08,
+      "loss": 2.8158,
+      "step": 40130
+    },
+    {
+      "epoch": 2.9531534514153286,
+      "grad_norm": 0.35913416743278503,
+      "learning_rate": 6.30696538803044e-08,
+      "loss": 2.8488,
+      "step": 40140
+    },
+    {
+      "epoch": 2.953889164781401,
+      "grad_norm": 0.35155612230300903,
+      "learning_rate": 6.110232446473463e-08,
+      "loss": 2.8322,
+      "step": 40150
+    },
+    {
+      "epoch": 2.9546248781474738,
+      "grad_norm": 0.35944974422454834,
+      "learning_rate": 5.916614780424046e-08,
+      "loss": 2.8559,
+      "step": 40160
+    },
+    {
+      "epoch": 2.9553605915135464,
+      "grad_norm": 0.3524097502231598,
+      "learning_rate": 5.726112510664461e-08,
+      "loss": 2.8003,
+      "step": 40170
+    },
+    {
+      "epoch": 2.956096304879619,
+      "grad_norm": 0.36307844519615173,
+      "learning_rate": 5.5387257560324256e-08,
+      "loss": 2.853,
+      "step": 40180
+    },
+    {
+      "epoch": 2.9568320182456915,
+      "grad_norm": 0.35420405864715576,
+      "learning_rate": 5.354454633423323e-08,
+      "loss": 2.8245,
+      "step": 40190
+    },
+    {
+      "epoch": 2.957567731611764,
+      "grad_norm": 0.3434230387210846,
+      "learning_rate": 5.1732992577890883e-08,
+      "loss": 2.8354,
+      "step": 40200
+    },
+    {
+      "epoch": 2.9583034449778367,
+      "grad_norm": 0.3682567775249481,
+      "learning_rate": 4.995259742137659e-08,
+      "loss": 2.8774,
+      "step": 40210
+    },
+    {
+      "epoch": 2.9590391583439093,
+      "grad_norm": 0.3482959270477295,
+      "learning_rate": 4.820336197532416e-08,
+      "loss": 2.8518,
+      "step": 40220
+    },
+    {
+      "epoch": 2.959774871709982,
+      "grad_norm": 0.3611307442188263,
+      "learning_rate": 4.6485287330944036e-08,
+      "loss": 2.9252,
+      "step": 40230
+    },
+    {
+      "epoch": 2.9605105850760545,
+      "grad_norm": 0.35166075825691223,
+      "learning_rate": 4.479837456000113e-08,
+      "loss": 2.8205,
+      "step": 40240
+    },
+    {
+      "epoch": 2.961246298442127,
+      "grad_norm": 0.3581395745277405,
+      "learning_rate": 4.314262471482033e-08,
+      "loss": 2.842,
+      "step": 40250
+    },
+    {
+      "epoch": 2.9619820118081996,
+      "grad_norm": 0.3495548963546753,
+      "learning_rate": 4.151803882828098e-08,
+      "loss": 2.8481,
+      "step": 40260
+    },
+    {
+      "epoch": 2.9627177251742722,
+      "grad_norm": 0.34471070766448975,
+      "learning_rate": 3.992461791383906e-08,
+      "loss": 2.8226,
+      "step": 40270
+    },
+    {
+      "epoch": 2.963453438540345,
+      "grad_norm": 0.35881149768829346,
+      "learning_rate": 3.836236296548834e-08,
+      "loss": 2.8442,
+      "step": 40280
+    },
+    {
+      "epoch": 2.9641891519064174,
+      "grad_norm": 0.3574000895023346,
+      "learning_rate": 3.6831274957793704e-08,
+      "loss": 2.8211,
+      "step": 40290
+    },
+    {
+      "epoch": 2.96492486527249,
+      "grad_norm": 0.3563789129257202,
+      "learning_rate": 3.533135484588002e-08,
+      "loss": 2.8433,
+      "step": 40300
+    },
+    {
+      "epoch": 2.9656605786385626,
+      "grad_norm": 0.35546740889549255,
+      "learning_rate": 3.386260356541548e-08,
+      "loss": 2.8443,
+      "step": 40310
+    },
+    {
+      "epoch": 2.966396292004635,
+      "grad_norm": 0.35332944989204407,
+      "learning_rate": 3.2425022032639417e-08,
+      "loss": 2.8325,
+      "step": 40320
+    },
+    {
+      "epoch": 2.9671320053707078,
+      "grad_norm": 0.35158273577690125,
+      "learning_rate": 3.101861114434001e-08,
+      "loss": 2.8652,
+      "step": 40330
+    },
+    {
+      "epoch": 2.9678677187367803,
+      "grad_norm": 0.34854280948638916,
+      "learning_rate": 2.964337177785437e-08,
+      "loss": 2.8905,
+      "step": 40340
+    },
+    {
+      "epoch": 2.968603432102853,
+      "grad_norm": 0.35338836908340454,
+      "learning_rate": 2.8299304791085113e-08,
+      "loss": 2.8273,
+      "step": 40350
+    },
+    {
+      "epoch": 2.9693391454689255,
+      "grad_norm": 0.35744231939315796,
+      "learning_rate": 2.698641102248378e-08,
+      "loss": 2.8299,
+      "step": 40360
+    },
+    {
+      "epoch": 2.970074858834998,
+      "grad_norm": 0.35066330432891846,
+      "learning_rate": 2.5704691291056348e-08,
+      "loss": 2.8162,
+      "step": 40370
+    },
+    {
+      "epoch": 2.9708105722010707,
+      "grad_norm": 0.3585476577281952,
+      "learning_rate": 2.445414639636323e-08,
+      "loss": 2.8496,
+      "step": 40380
+    },
+    {
+      "epoch": 2.9715462855671433,
+      "grad_norm": 0.3533668518066406,
+      "learning_rate": 2.323477711851374e-08,
+      "loss": 2.8764,
+      "step": 40390
+    },
+    {
+      "epoch": 2.972281998933216,
+      "grad_norm": 0.35567134618759155,
+      "learning_rate": 2.204658421817718e-08,
+      "loss": 2.8303,
+      "step": 40400
+    },
+    {
+      "epoch": 2.9730177122992885,
+      "grad_norm": 0.35332998633384705,
+      "learning_rate": 2.0889568436560647e-08,
+      "loss": 2.8891,
+      "step": 40410
+    },
+    {
+      "epoch": 2.973753425665361,
+      "grad_norm": 0.3475884199142456,
+      "learning_rate": 1.9763730495436785e-08,
+      "loss": 2.8221,
+      "step": 40420
+    },
+    {
+      "epoch": 2.974489139031433,
+      "grad_norm": 0.36162737011909485,
+      "learning_rate": 1.8669071097127123e-08,
+      "loss": 2.8441,
+      "step": 40430
+    },
+    {
+      "epoch": 2.975224852397506,
+      "grad_norm": 0.3608178198337555,
+      "learning_rate": 1.760559092448544e-08,
+      "loss": 2.8288,
+      "step": 40440
+    },
+    {
+      "epoch": 2.9759605657635784,
+      "grad_norm": 0.370004802942276,
+      "learning_rate": 1.657329064094215e-08,
+      "loss": 2.8283,
+      "step": 40450
+    },
+    {
+      "epoch": 2.976696279129651,
+      "grad_norm": 0.35101181268692017,
+      "learning_rate": 1.5572170890454374e-08,
+      "loss": 2.8192,
+      "step": 40460
+    },
+    {
+      "epoch": 2.9774319924957235,
+      "grad_norm": 0.36110401153564453,
+      "learning_rate": 1.4602232297550311e-08,
+      "loss": 2.8587,
+      "step": 40470
+    },
+    {
+      "epoch": 2.978167705861796,
+      "grad_norm": 0.3538599908351898,
+      "learning_rate": 1.3663475467279308e-08,
+      "loss": 2.8515,
+      "step": 40480
+    },
+    {
+      "epoch": 2.9789034192278687,
+      "grad_norm": 0.3524283766746521,
+      "learning_rate": 1.2755900985267354e-08,
+      "loss": 2.8568,
+      "step": 40490
+    },
+    {
+      "epoch": 2.9796391325939413,
+      "grad_norm": 0.3458719551563263,
+      "learning_rate": 1.1879509417667134e-08,
+      "loss": 2.8446,
+      "step": 40500
+    },
+    {
+      "epoch": 2.980374845960014,
+      "grad_norm": 0.3463917672634125,
+      "learning_rate": 1.103430131119687e-08,
+      "loss": 2.8651,
+      "step": 40510
+    },
+    {
+      "epoch": 2.9811105593260865,
+      "grad_norm": 0.3477877378463745,
+      "learning_rate": 1.0220277193095928e-08,
+      "loss": 2.7935,
+      "step": 40520
+    },
+    {
+      "epoch": 2.981846272692159,
+      "grad_norm": 0.35533004999160767,
+      "learning_rate": 9.437437571180319e-09,
+      "loss": 2.8301,
+      "step": 40530
+    },
+    {
+      "epoch": 2.9825819860582317,
+      "grad_norm": 0.36103564500808716,
+      "learning_rate": 8.685782933798292e-09,
+      "loss": 2.8386,
+      "step": 40540
+    },
+    {
+      "epoch": 2.9833176994243042,
+      "grad_norm": 0.3614270091056824,
+      "learning_rate": 7.96531374983589e-09,
+      "loss": 2.8439,
+      "step": 40550
+    },
+    {
+      "epoch": 2.984053412790377,
+      "grad_norm": 0.35036417841911316,
+      "learning_rate": 7.276030468739148e-09,
+      "loss": 2.8635,
+      "step": 40560
+    },
+    {
+      "epoch": 2.9847891261564494,
+      "grad_norm": 0.35889285802841187,
+      "learning_rate": 6.6179335205029946e-09,
+      "loss": 2.8642,
+      "step": 40570
+    },
+    {
+      "epoch": 2.985524839522522,
+      "grad_norm": 0.35606804490089417,
+      "learning_rate": 5.9910233156434955e-09,
+      "loss": 2.8263,
+      "step": 40580
+    },
+    {
+      "epoch": 2.9862605528885946,
+      "grad_norm": 0.35697445273399353,
+      "learning_rate": 5.395300245253365e-09,
+      "loss": 2.8533,
+      "step": 40590
+    },
+    {
+      "epoch": 2.986996266254667,
+      "grad_norm": 0.35635432600975037,
+      "learning_rate": 4.8307646809464536e-09,
+      "loss": 2.8151,
+      "step": 40600
+    },
+    {
+      "epoch": 2.9877319796207398,
+      "grad_norm": 0.3581090569496155,
+      "learning_rate": 4.297416974896606e-09,
+      "loss": 2.8294,
+      "step": 40610
+    },
+    {
+      "epoch": 2.9884676929868124,
+      "grad_norm": 0.3472132980823517,
+      "learning_rate": 3.795257459809909e-09,
+      "loss": 2.8353,
+      "step": 40620
+    },
+    {
+      "epoch": 2.989203406352885,
+      "grad_norm": 0.3511686325073242,
+      "learning_rate": 3.3242864489413383e-09,
+      "loss": 2.8434,
+      "step": 40630
+    },
+    {
+      "epoch": 2.9899391197189575,
+      "grad_norm": 0.36015400290489197,
+      "learning_rate": 2.884504236094765e-09,
+      "loss": 2.8419,
+      "step": 40640
+    },
+    {
+      "epoch": 2.99067483308503,
+      "grad_norm": 0.35950011014938354,
+      "learning_rate": 2.475911095606298e-09,
+      "loss": 2.898,
+      "step": 40650
+    },
+    {
+      "epoch": 2.9914105464511027,
+      "grad_norm": 0.3610963523387909,
+      "learning_rate": 2.0985072823775933e-09,
+      "loss": 2.8128,
+      "step": 40660
+    },
+    {
+      "epoch": 2.9921462598171753,
+      "grad_norm": 0.34512999653816223,
+      "learning_rate": 1.7522930318258912e-09,
+      "loss": 2.8107,
+      "step": 40670
+    },
+    {
+      "epoch": 2.992881973183248,
+      "grad_norm": 0.3686096966266632,
+      "learning_rate": 1.4372685599339798e-09,
+      "loss": 2.8317,
+      "step": 40680
+    },
+    {
+      "epoch": 2.9936176865493205,
+      "grad_norm": 0.36063432693481445,
+      "learning_rate": 1.153434063216885e-09,
+      "loss": 2.8631,
+      "step": 40690
+    },
+    {
+      "epoch": 2.994353399915393,
+      "grad_norm": 0.3461795747280121,
+      "learning_rate": 9.007897187385261e-10,
+      "loss": 2.7775,
+      "step": 40700
+    },
+    {
+      "epoch": 2.9950891132814657,
+      "grad_norm": 0.35914844274520874,
+      "learning_rate": 6.793356840950616e-10,
+      "loss": 2.8302,
+      "step": 40710
+    },
+    {
+      "epoch": 2.9958248266475382,
+      "grad_norm": 0.35153597593307495,
+      "learning_rate": 4.890720974426444e-10,
+      "loss": 2.837,
+      "step": 40720
+    },
+    {
+      "epoch": 2.996560540013611,
+      "grad_norm": 0.3526630401611328,
+      "learning_rate": 3.2999907746411595e-10,
+      "loss": 2.837,
+      "step": 40730
+    },
+    {
+      "epoch": 2.9972962533796834,
+      "grad_norm": 0.3597138226032257,
+      "learning_rate": 2.0211672339121023e-10,
+      "loss": 2.8531,
+      "step": 40740
+    },
+    {
+      "epoch": 2.998031966745756,
+      "grad_norm": 0.3528171181678772,
+      "learning_rate": 1.0542511501010489e-10,
+      "loss": 2.9138,
+      "step": 40750
+    },
+    {
+      "epoch": 2.9987676801118286,
+      "grad_norm": 0.3507966995239258,
+      "learning_rate": 3.992431262811458e-11,
+      "loss": 2.8545,
+      "step": 40760
+    },
+    {
+      "epoch": 2.9995033934779007,
+      "grad_norm": 0.35401690006256104,
+      "learning_rate": 5.6143571069977584e-12,
+      "loss": 2.8371,
+      "step": 40770
+    },
+    {
+      "epoch": 2.999944821497545,
+      "step": 40776,
+      "total_flos": 3.6968537365004943e+18,
+      "train_loss": 3.121767396105489,
+      "train_runtime": 856366.0469,
+      "train_samples_per_second": 3.047,
+      "train_steps_per_second": 0.048
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 40776,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 4000,
+  "total_flos": 3.6968537365004943e+18,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}