diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,21042 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 177.5147928994083,
+  "eval_steps": 500,
+  "global_step": 30000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.05917159763313609,
+      "grad_norm": 174.9995880126953,
+      "learning_rate": 0.0001,
+      "loss": 21.1983,
+      "step": 10
+    },
+    {
+      "epoch": 0.11834319526627218,
+      "grad_norm": 88.79548645019531,
+      "learning_rate": 0.0001,
+      "loss": 5.4202,
+      "step": 20
+    },
+    {
+      "epoch": 0.17751479289940827,
+      "grad_norm": 60.9506721496582,
+      "learning_rate": 0.0001,
+      "loss": 3.9479,
+      "step": 30
+    },
+    {
+      "epoch": 0.23668639053254437,
+      "grad_norm": 54.834285736083984,
+      "learning_rate": 0.0001,
+      "loss": 3.4888,
+      "step": 40
+    },
+    {
+      "epoch": 0.2958579881656805,
+      "grad_norm": 57.539466857910156,
+      "learning_rate": 0.0001,
+      "loss": 3.1011,
+      "step": 50
+    },
+    {
+      "epoch": 0.35502958579881655,
+      "grad_norm": 51.58094787597656,
+      "learning_rate": 0.0001,
+      "loss": 2.7793,
+      "step": 60
+    },
+    {
+      "epoch": 0.41420118343195267,
+      "grad_norm": 46.063045501708984,
+      "learning_rate": 0.0001,
+      "loss": 2.5763,
+      "step": 70
+    },
+    {
+      "epoch": 0.47337278106508873,
+      "grad_norm": 47.30936813354492,
+      "learning_rate": 0.0001,
+      "loss": 2.3486,
+      "step": 80
+    },
+    {
+      "epoch": 0.5325443786982249,
+      "grad_norm": 58.13508224487305,
+      "learning_rate": 0.0001,
+      "loss": 2.2598,
+      "step": 90
+    },
+    {
+      "epoch": 0.591715976331361,
+      "grad_norm": 47.5648307800293,
+      "learning_rate": 0.0001,
+      "loss": 2.1153,
+      "step": 100
+    },
+    {
+      "epoch": 0.650887573964497,
+      "grad_norm": 44.50985336303711,
+      "learning_rate": 0.0001,
+      "loss": 1.9304,
+      "step": 110
+    },
+    {
+      "epoch": 0.7100591715976331,
+      "grad_norm": 40.581214904785156,
+      "learning_rate": 0.0001,
+      "loss": 1.8131,
+      "step": 120
+    },
+    {
+      "epoch": 0.7692307692307693,
+      "grad_norm": 42.132205963134766,
+      "learning_rate": 0.0001,
+      "loss": 1.718,
+      "step": 130
+    },
+    {
+      "epoch": 0.8284023668639053,
+      "grad_norm": 37.27217483520508,
+      "learning_rate": 0.0001,
+      "loss": 1.6276,
+      "step": 140
+    },
+    {
+      "epoch": 0.8875739644970414,
+      "grad_norm": 42.04931640625,
+      "learning_rate": 0.0001,
+      "loss": 1.5218,
+      "step": 150
+    },
+    {
+      "epoch": 0.9467455621301775,
+      "grad_norm": 45.046295166015625,
+      "learning_rate": 0.0001,
+      "loss": 1.4777,
+      "step": 160
+    },
+    {
+      "epoch": 1.0059171597633136,
+      "grad_norm": 40.04572677612305,
+      "learning_rate": 0.0001,
+      "loss": 1.4081,
+      "step": 170
+    },
+    {
+      "epoch": 1.0650887573964498,
+      "grad_norm": 38.59673309326172,
+      "learning_rate": 0.0001,
+      "loss": 1.3217,
+      "step": 180
+    },
+    {
+      "epoch": 1.1242603550295858,
+      "grad_norm": 31.95562744140625,
+      "learning_rate": 0.0001,
+      "loss": 1.227,
+      "step": 190
+    },
+    {
+      "epoch": 1.183431952662722,
+      "grad_norm": 32.33133316040039,
+      "learning_rate": 0.0001,
+      "loss": 1.1588,
+      "step": 200
+    },
+    {
+      "epoch": 1.242603550295858,
+      "grad_norm": 36.7282600402832,
+      "learning_rate": 0.0001,
+      "loss": 1.1121,
+      "step": 210
+    },
+    {
+      "epoch": 1.301775147928994,
+      "grad_norm": 35.95631790161133,
+      "learning_rate": 0.0001,
+      "loss": 1.0593,
+      "step": 220
+    },
+    {
+      "epoch": 1.3609467455621302,
+      "grad_norm": 36.83339309692383,
+      "learning_rate": 0.0001,
+      "loss": 1.0002,
+      "step": 230
+    },
+    {
+      "epoch": 1.4201183431952662,
+      "grad_norm": 32.869117736816406,
+      "learning_rate": 0.0001,
+      "loss": 0.9396,
+      "step": 240
+    },
+    {
+      "epoch": 1.4792899408284024,
+      "grad_norm": 32.54338073730469,
+      "learning_rate": 0.0001,
+      "loss": 0.8936,
+      "step": 250
+    },
+    {
+      "epoch": 1.5384615384615383,
+      "grad_norm": 28.00010871887207,
+      "learning_rate": 0.0001,
+      "loss": 0.8271,
+      "step": 260
+    },
+    {
+      "epoch": 1.5976331360946747,
+      "grad_norm": 33.478515625,
+      "learning_rate": 0.0001,
+      "loss": 0.807,
+      "step": 270
+    },
+    {
+      "epoch": 1.6568047337278107,
+      "grad_norm": 29.047765731811523,
+      "learning_rate": 0.0001,
+      "loss": 0.7601,
+      "step": 280
+    },
+    {
+      "epoch": 1.7159763313609466,
+      "grad_norm": 28.845853805541992,
+      "learning_rate": 0.0001,
+      "loss": 0.7073,
+      "step": 290
+    },
+    {
+      "epoch": 1.7751479289940828,
+      "grad_norm": 25.670988082885742,
+      "learning_rate": 0.0001,
+      "loss": 0.6772,
+      "step": 300
+    },
+    {
+      "epoch": 1.834319526627219,
+      "grad_norm": 27.993547439575195,
+      "learning_rate": 0.0001,
+      "loss": 0.6433,
+      "step": 310
+    },
+    {
+      "epoch": 1.893491124260355,
+      "grad_norm": 25.02227210998535,
+      "learning_rate": 0.0001,
+      "loss": 0.6143,
+      "step": 320
+    },
+    {
+      "epoch": 1.952662721893491,
+      "grad_norm": 24.85578155517578,
+      "learning_rate": 0.0001,
+      "loss": 0.5882,
+      "step": 330
+    },
+    {
+      "epoch": 2.0118343195266273,
+      "grad_norm": 23.767911911010742,
+      "learning_rate": 0.0001,
+      "loss": 0.5557,
+      "step": 340
+    },
+    {
+      "epoch": 2.0710059171597632,
+      "grad_norm": 22.744504928588867,
+      "learning_rate": 0.0001,
+      "loss": 0.535,
+      "step": 350
+    },
+    {
+      "epoch": 2.1301775147928996,
+      "grad_norm": 24.207067489624023,
+      "learning_rate": 0.0001,
+      "loss": 0.524,
+      "step": 360
+    },
+    {
+      "epoch": 2.1893491124260356,
+      "grad_norm": 22.372703552246094,
+      "learning_rate": 0.0001,
+      "loss": 0.499,
+      "step": 370
+    },
+    {
+      "epoch": 2.2485207100591715,
+      "grad_norm": 22.381425857543945,
+      "learning_rate": 0.0001,
+      "loss": 0.4852,
+      "step": 380
+    },
+    {
+      "epoch": 2.3076923076923075,
+      "grad_norm": 22.285106658935547,
+      "learning_rate": 0.0001,
+      "loss": 0.4623,
+      "step": 390
+    },
+    {
+      "epoch": 2.366863905325444,
+      "grad_norm": 20.64701271057129,
+      "learning_rate": 0.0001,
+      "loss": 0.4435,
+      "step": 400
+    },
+    {
+      "epoch": 2.42603550295858,
+      "grad_norm": 22.761093139648438,
+      "learning_rate": 0.0001,
+      "loss": 0.4268,
+      "step": 410
+    },
+    {
+      "epoch": 2.485207100591716,
+      "grad_norm": 22.484844207763672,
+      "learning_rate": 0.0001,
+      "loss": 0.4205,
+      "step": 420
+    },
+    {
+      "epoch": 2.544378698224852,
+      "grad_norm": 22.757444381713867,
+      "learning_rate": 0.0001,
+      "loss": 0.4121,
+      "step": 430
+    },
+    {
+      "epoch": 2.603550295857988,
+      "grad_norm": 20.99678611755371,
+      "learning_rate": 0.0001,
+      "loss": 0.3852,
+      "step": 440
+    },
+    {
+      "epoch": 2.662721893491124,
+      "grad_norm": 19.638227462768555,
+      "learning_rate": 0.0001,
+      "loss": 0.3687,
+      "step": 450
+    },
+    {
+      "epoch": 2.7218934911242605,
+      "grad_norm": 18.762407302856445,
+      "learning_rate": 0.0001,
+      "loss": 0.3518,
+      "step": 460
+    },
+    {
+      "epoch": 2.7810650887573964,
+      "grad_norm": 18.946866989135742,
+      "learning_rate": 0.0001,
+      "loss": 0.3573,
+      "step": 470
+    },
+    {
+      "epoch": 2.8402366863905324,
+      "grad_norm": 17.21717643737793,
+      "learning_rate": 0.0001,
+      "loss": 0.335,
+      "step": 480
+    },
+    {
+      "epoch": 2.899408284023669,
+      "grad_norm": 20.277074813842773,
+      "learning_rate": 0.0001,
+      "loss": 0.3327,
+      "step": 490
+    },
+    {
+      "epoch": 2.9585798816568047,
+      "grad_norm": 19.717561721801758,
+      "learning_rate": 0.0001,
+      "loss": 0.3278,
+      "step": 500
+    },
+    {
+      "epoch": 3.0177514792899407,
+      "grad_norm": 17.66796112060547,
+      "learning_rate": 0.0001,
+      "loss": 0.3183,
+      "step": 510
+    },
+    {
+      "epoch": 3.076923076923077,
+      "grad_norm": 17.042224884033203,
+      "learning_rate": 0.0001,
+      "loss": 0.3045,
+      "step": 520
+    },
+    {
+      "epoch": 3.136094674556213,
+      "grad_norm": 15.651921272277832,
+      "learning_rate": 0.0001,
+      "loss": 0.303,
+      "step": 530
+    },
+    {
+      "epoch": 3.195266272189349,
+      "grad_norm": 15.386090278625488,
+      "learning_rate": 0.0001,
+      "loss": 0.2888,
+      "step": 540
+    },
+    {
+      "epoch": 3.2544378698224854,
+      "grad_norm": 17.88136863708496,
+      "learning_rate": 0.0001,
+      "loss": 0.2866,
+      "step": 550
+    },
+    {
+      "epoch": 3.3136094674556213,
+      "grad_norm": 21.284189224243164,
+      "learning_rate": 0.0001,
+      "loss": 0.2821,
+      "step": 560
+    },
+    {
+      "epoch": 3.3727810650887573,
+      "grad_norm": 16.279726028442383,
+      "learning_rate": 0.0001,
+      "loss": 0.2785,
+      "step": 570
+    },
+    {
+      "epoch": 3.4319526627218933,
+      "grad_norm": 17.655969619750977,
+      "learning_rate": 0.0001,
+      "loss": 0.2718,
+      "step": 580
+    },
+    {
+      "epoch": 3.4911242603550297,
+      "grad_norm": 14.347151756286621,
+      "learning_rate": 0.0001,
+      "loss": 0.2605,
+      "step": 590
+    },
+    {
+      "epoch": 3.5502958579881656,
+      "grad_norm": 15.892441749572754,
+      "learning_rate": 0.0001,
+      "loss": 0.2543,
+      "step": 600
+    },
+    {
+      "epoch": 3.609467455621302,
+      "grad_norm": 15.025949478149414,
+      "learning_rate": 0.0001,
+      "loss": 0.2507,
+      "step": 610
+    },
+    {
+      "epoch": 3.668639053254438,
+      "grad_norm": 14.651044845581055,
+      "learning_rate": 0.0001,
+      "loss": 0.2521,
+      "step": 620
+    },
+    {
+      "epoch": 3.727810650887574,
+      "grad_norm": 14.25905704498291,
+      "learning_rate": 0.0001,
+      "loss": 0.244,
+      "step": 630
+    },
+    {
+      "epoch": 3.78698224852071,
+      "grad_norm": 13.280366897583008,
+      "learning_rate": 0.0001,
+      "loss": 0.2456,
+      "step": 640
+    },
+    {
+      "epoch": 3.8461538461538463,
+      "grad_norm": 14.474982261657715,
+      "learning_rate": 0.0001,
+      "loss": 0.2464,
+      "step": 650
+    },
+    {
+      "epoch": 3.905325443786982,
+      "grad_norm": 12.5128755569458,
+      "learning_rate": 0.0001,
+      "loss": 0.2392,
+      "step": 660
+    },
+    {
+      "epoch": 3.9644970414201186,
+      "grad_norm": 13.463188171386719,
+      "learning_rate": 0.0001,
+      "loss": 0.2345,
+      "step": 670
+    },
+    {
+      "epoch": 4.023668639053255,
+      "grad_norm": 14.782414436340332,
+      "learning_rate": 0.0001,
+      "loss": 0.2255,
+      "step": 680
+    },
+    {
+      "epoch": 4.0828402366863905,
+      "grad_norm": 10.73136043548584,
+      "learning_rate": 0.0001,
+      "loss": 0.2227,
+      "step": 690
+    },
+    {
+      "epoch": 4.1420118343195265,
+      "grad_norm": 13.535078048706055,
+      "learning_rate": 0.0001,
+      "loss": 0.2259,
+      "step": 700
+    },
+    {
+      "epoch": 4.201183431952662,
+      "grad_norm": 13.905795097351074,
+      "learning_rate": 0.0001,
+      "loss": 0.2277,
+      "step": 710
+    },
+    {
+      "epoch": 4.260355029585799,
+      "grad_norm": 13.936788558959961,
+      "learning_rate": 0.0001,
+      "loss": 0.2206,
+      "step": 720
+    },
+    {
+      "epoch": 4.319526627218935,
+      "grad_norm": 13.85198974609375,
+      "learning_rate": 0.0001,
+      "loss": 0.2118,
+      "step": 730
+    },
+    {
+      "epoch": 4.378698224852071,
+      "grad_norm": 15.073653221130371,
+      "learning_rate": 0.0001,
+      "loss": 0.213,
+      "step": 740
+    },
+    {
+      "epoch": 4.437869822485207,
+      "grad_norm": 14.68563461303711,
+      "learning_rate": 0.0001,
+      "loss": 0.2109,
+      "step": 750
+    },
+    {
+      "epoch": 4.497041420118343,
+      "grad_norm": 12.35783863067627,
+      "learning_rate": 0.0001,
+      "loss": 0.2144,
+      "step": 760
+    },
+    {
+      "epoch": 4.556213017751479,
+      "grad_norm": 13.271652221679688,
+      "learning_rate": 0.0001,
+      "loss": 0.2087,
+      "step": 770
+    },
+    {
+      "epoch": 4.615384615384615,
+      "grad_norm": 12.783738136291504,
+      "learning_rate": 0.0001,
+      "loss": 0.2046,
+      "step": 780
+    },
+    {
+      "epoch": 4.674556213017752,
+      "grad_norm": 13.170055389404297,
+      "learning_rate": 0.0001,
+      "loss": 0.2092,
+      "step": 790
+    },
+    {
+      "epoch": 4.733727810650888,
+      "grad_norm": 14.701375961303711,
+      "learning_rate": 0.0001,
+      "loss": 0.2007,
+      "step": 800
+    },
+    {
+      "epoch": 4.792899408284024,
+      "grad_norm": 11.574836730957031,
+      "learning_rate": 0.0001,
+      "loss": 0.1973,
+      "step": 810
+    },
+    {
+      "epoch": 4.85207100591716,
+      "grad_norm": 12.130415916442871,
+      "learning_rate": 0.0001,
+      "loss": 0.1957,
+      "step": 820
+    },
+    {
+      "epoch": 4.911242603550296,
+      "grad_norm": 12.733306884765625,
+      "learning_rate": 0.0001,
+      "loss": 0.1862,
+      "step": 830
+    },
+    {
+      "epoch": 4.970414201183432,
+      "grad_norm": 12.252595901489258,
+      "learning_rate": 0.0001,
+      "loss": 0.192,
+      "step": 840
+    },
+    {
+      "epoch": 5.029585798816568,
+      "grad_norm": 12.803046226501465,
+      "learning_rate": 0.0001,
+      "loss": 0.1893,
+      "step": 850
+    },
+    {
+      "epoch": 5.088757396449704,
+      "grad_norm": 11.160612106323242,
+      "learning_rate": 0.0001,
+      "loss": 0.1875,
+      "step": 860
+    },
+    {
+      "epoch": 5.14792899408284,
+      "grad_norm": 11.698083877563477,
+      "learning_rate": 0.0001,
+      "loss": 0.184,
+      "step": 870
+    },
+    {
+      "epoch": 5.207100591715976,
+      "grad_norm": 10.715482711791992,
+      "learning_rate": 0.0001,
+      "loss": 0.1829,
+      "step": 880
+    },
+    {
+      "epoch": 5.266272189349112,
+      "grad_norm": 12.175360679626465,
+      "learning_rate": 0.0001,
+      "loss": 0.1824,
+      "step": 890
+    },
+    {
+      "epoch": 5.325443786982248,
+      "grad_norm": 13.77058219909668,
+      "learning_rate": 0.0001,
+      "loss": 0.1808,
+      "step": 900
+    },
+    {
+      "epoch": 5.384615384615385,
+      "grad_norm": 12.61340045928955,
+      "learning_rate": 0.0001,
+      "loss": 0.1801,
+      "step": 910
+    },
+    {
+      "epoch": 5.443786982248521,
+      "grad_norm": 11.543051719665527,
+      "learning_rate": 0.0001,
+      "loss": 0.1836,
+      "step": 920
+    },
+    {
+      "epoch": 5.502958579881657,
+      "grad_norm": 14.268769264221191,
+      "learning_rate": 0.0001,
+      "loss": 0.1836,
+      "step": 930
+    },
+    {
+      "epoch": 5.562130177514793,
+      "grad_norm": 12.61422348022461,
+      "learning_rate": 0.0001,
+      "loss": 0.1799,
+      "step": 940
+    },
+    {
+      "epoch": 5.621301775147929,
+      "grad_norm": 11.867342948913574,
+      "learning_rate": 0.0001,
+      "loss": 0.1759,
+      "step": 950
+    },
+    {
+      "epoch": 5.680473372781065,
+      "grad_norm": 10.003403663635254,
+      "learning_rate": 0.0001,
+      "loss": 0.1677,
+      "step": 960
+    },
+    {
+      "epoch": 5.739644970414201,
+      "grad_norm": 10.707908630371094,
+      "learning_rate": 0.0001,
+      "loss": 0.166,
+      "step": 970
+    },
+    {
+      "epoch": 5.798816568047338,
+      "grad_norm": 9.132226943969727,
+      "learning_rate": 0.0001,
+      "loss": 0.1599,
+      "step": 980
+    },
+    {
+      "epoch": 5.8579881656804735,
+      "grad_norm": 10.596758842468262,
+      "learning_rate": 0.0001,
+      "loss": 0.1551,
+      "step": 990
+    },
+    {
+      "epoch": 5.9171597633136095,
+      "grad_norm": 10.703545570373535,
+      "learning_rate": 0.0001,
+      "loss": 0.1643,
+      "step": 1000
+    },
+    {
+      "epoch": 5.976331360946745,
+      "grad_norm": 10.75903606414795,
+      "learning_rate": 0.0001,
+      "loss": 0.158,
+      "step": 1010
+    },
+    {
+      "epoch": 6.035502958579881,
+      "grad_norm": 11.10091781616211,
+      "learning_rate": 0.0001,
+      "loss": 0.1543,
+      "step": 1020
+    },
+    {
+      "epoch": 6.094674556213017,
+      "grad_norm": 10.262393951416016,
+      "learning_rate": 0.0001,
+      "loss": 0.1566,
+      "step": 1030
+    },
+    {
+      "epoch": 6.153846153846154,
+      "grad_norm": 9.474786758422852,
+      "learning_rate": 0.0001,
+      "loss": 0.1526,
+      "step": 1040
+    },
+    {
+      "epoch": 6.21301775147929,
+      "grad_norm": 7.61234712600708,
+      "learning_rate": 0.0001,
+      "loss": 0.1546,
+      "step": 1050
+    },
+    {
+      "epoch": 6.272189349112426,
+      "grad_norm": 11.92391586303711,
+      "learning_rate": 0.0001,
+      "loss": 0.157,
+      "step": 1060
+    },
+    {
+      "epoch": 6.331360946745562,
+      "grad_norm": 9.999370574951172,
+      "learning_rate": 0.0001,
+      "loss": 0.1553,
+      "step": 1070
+    },
+    {
+      "epoch": 6.390532544378698,
+      "grad_norm": 10.439435005187988,
+      "learning_rate": 0.0001,
+      "loss": 0.1592,
+      "step": 1080
+    },
+    {
+      "epoch": 6.449704142011834,
+      "grad_norm": 9.323165893554688,
+      "learning_rate": 0.0001,
+      "loss": 0.1501,
+      "step": 1090
+    },
+    {
+      "epoch": 6.508875739644971,
+      "grad_norm": 8.45301342010498,
+      "learning_rate": 0.0001,
+      "loss": 0.1487,
+      "step": 1100
+    },
+    {
+      "epoch": 6.568047337278107,
+      "grad_norm": 11.393036842346191,
+      "learning_rate": 0.0001,
+      "loss": 0.1488,
+      "step": 1110
+    },
+    {
+      "epoch": 6.627218934911243,
+      "grad_norm": 10.938389778137207,
+      "learning_rate": 0.0001,
+      "loss": 0.151,
+      "step": 1120
+    },
+    {
+      "epoch": 6.686390532544379,
+      "grad_norm": 10.78990364074707,
+      "learning_rate": 0.0001,
+      "loss": 0.1452,
+      "step": 1130
+    },
+    {
+      "epoch": 6.745562130177515,
+      "grad_norm": 8.883673667907715,
+      "learning_rate": 0.0001,
+      "loss": 0.1442,
+      "step": 1140
+    },
+    {
+      "epoch": 6.804733727810651,
+      "grad_norm": 9.506607055664062,
+      "learning_rate": 0.0001,
+      "loss": 0.1412,
+      "step": 1150
+    },
+    {
+      "epoch": 6.8639053254437865,
+      "grad_norm": 10.19692611694336,
+      "learning_rate": 0.0001,
+      "loss": 0.1415,
+      "step": 1160
+    },
+    {
+      "epoch": 6.923076923076923,
+      "grad_norm": 9.091435432434082,
+      "learning_rate": 0.0001,
+      "loss": 0.1393,
+      "step": 1170
+    },
+    {
+      "epoch": 6.982248520710059,
+      "grad_norm": 9.56710147857666,
+      "learning_rate": 0.0001,
+      "loss": 0.1408,
+      "step": 1180
+    },
+    {
+      "epoch": 7.041420118343195,
+      "grad_norm": 8.897675514221191,
+      "learning_rate": 0.0001,
+      "loss": 0.1406,
+      "step": 1190
+    },
+    {
+      "epoch": 7.100591715976331,
+      "grad_norm": 8.53128719329834,
+      "learning_rate": 0.0001,
+      "loss": 0.1435,
+      "step": 1200
+    },
+    {
+      "epoch": 7.159763313609467,
+      "grad_norm": 7.786635398864746,
+      "learning_rate": 0.0001,
+      "loss": 0.1378,
+      "step": 1210
+    },
+    {
+      "epoch": 7.218934911242604,
+      "grad_norm": 8.461904525756836,
+      "learning_rate": 0.0001,
+      "loss": 0.1385,
+      "step": 1220
+    },
+    {
+      "epoch": 7.27810650887574,
+      "grad_norm": 9.95323657989502,
+      "learning_rate": 0.0001,
+      "loss": 0.137,
+      "step": 1230
+    },
+    {
+      "epoch": 7.337278106508876,
+      "grad_norm": 8.362894058227539,
+      "learning_rate": 0.0001,
+      "loss": 0.1334,
+      "step": 1240
+    },
+    {
+      "epoch": 7.396449704142012,
+      "grad_norm": 8.03234577178955,
+      "learning_rate": 0.0001,
+      "loss": 0.1268,
+      "step": 1250
+    },
+    {
+      "epoch": 7.455621301775148,
+      "grad_norm": 10.620657920837402,
+      "learning_rate": 0.0001,
+      "loss": 0.1277,
+      "step": 1260
+    },
+    {
+      "epoch": 7.514792899408284,
+      "grad_norm": 9.013126373291016,
+      "learning_rate": 0.0001,
+      "loss": 0.1314,
+      "step": 1270
+    },
+    {
+      "epoch": 7.57396449704142,
+      "grad_norm": 8.546844482421875,
+      "learning_rate": 0.0001,
+      "loss": 0.1245,
+      "step": 1280
+    },
+    {
+      "epoch": 7.633136094674557,
+      "grad_norm": 9.392996788024902,
+      "learning_rate": 0.0001,
+      "loss": 0.1306,
+      "step": 1290
+    },
+    {
+      "epoch": 7.6923076923076925,
+      "grad_norm": 8.910809516906738,
+      "learning_rate": 0.0001,
+      "loss": 0.133,
+      "step": 1300
+    },
+    {
+      "epoch": 7.7514792899408285,
+      "grad_norm": 9.005481719970703,
+      "learning_rate": 0.0001,
+      "loss": 0.1311,
+      "step": 1310
+    },
+    {
+      "epoch": 7.810650887573964,
+      "grad_norm": 7.805796146392822,
+      "learning_rate": 0.0001,
+      "loss": 0.127,
+      "step": 1320
+    },
+    {
+      "epoch": 7.8698224852071,
+      "grad_norm": 9.207897186279297,
+      "learning_rate": 0.0001,
+      "loss": 0.1257,
+      "step": 1330
+    },
+    {
+      "epoch": 7.928994082840236,
+      "grad_norm": 9.249608039855957,
+      "learning_rate": 0.0001,
+      "loss": 0.1253,
+      "step": 1340
+    },
+    {
+      "epoch": 7.988165680473373,
+      "grad_norm": 8.21968936920166,
+      "learning_rate": 0.0001,
+      "loss": 0.1218,
+      "step": 1350
+    },
+    {
+      "epoch": 8.04733727810651,
+      "grad_norm": 8.336978912353516,
+      "learning_rate": 0.0001,
+      "loss": 0.1236,
+      "step": 1360
+    },
+    {
+      "epoch": 8.106508875739644,
+      "grad_norm": 10.590923309326172,
+      "learning_rate": 0.0001,
+      "loss": 0.1227,
+      "step": 1370
+    },
+    {
+      "epoch": 8.165680473372781,
+      "grad_norm": 9.839675903320312,
+      "learning_rate": 0.0001,
+      "loss": 0.1215,
+      "step": 1380
+    },
+    {
+      "epoch": 8.224852071005918,
+      "grad_norm": 9.125936508178711,
+      "learning_rate": 0.0001,
+      "loss": 0.1212,
+      "step": 1390
+    },
+    {
+      "epoch": 8.284023668639053,
+      "grad_norm": 8.98107624053955,
+      "learning_rate": 0.0001,
+      "loss": 0.1229,
+      "step": 1400
+    },
+    {
+      "epoch": 8.34319526627219,
+      "grad_norm": 9.424227714538574,
+      "learning_rate": 0.0001,
+      "loss": 0.1214,
+      "step": 1410
+    },
+    {
+      "epoch": 8.402366863905325,
+      "grad_norm": 9.905462265014648,
+      "learning_rate": 0.0001,
+      "loss": 0.117,
+      "step": 1420
+    },
+    {
+      "epoch": 8.461538461538462,
+      "grad_norm": 8.423462867736816,
+      "learning_rate": 0.0001,
+      "loss": 0.1156,
+      "step": 1430
+    },
+    {
+      "epoch": 8.520710059171599,
+      "grad_norm": 7.6935906410217285,
+      "learning_rate": 0.0001,
+      "loss": 0.1185,
+      "step": 1440
+    },
+    {
+      "epoch": 8.579881656804734,
+      "grad_norm": 7.752745151519775,
+      "learning_rate": 0.0001,
+      "loss": 0.1177,
+      "step": 1450
+    },
+    {
+      "epoch": 8.63905325443787,
+      "grad_norm": 7.3799028396606445,
+      "learning_rate": 0.0001,
+      "loss": 0.1151,
+      "step": 1460
+    },
+    {
+      "epoch": 8.698224852071005,
+      "grad_norm": 8.428478240966797,
+      "learning_rate": 0.0001,
+      "loss": 0.1132,
+      "step": 1470
+    },
+    {
+      "epoch": 8.757396449704142,
+      "grad_norm": 8.008804321289062,
+      "learning_rate": 0.0001,
+      "loss": 0.1138,
+      "step": 1480
+    },
+    {
+      "epoch": 8.816568047337277,
+      "grad_norm": 7.186567783355713,
+      "learning_rate": 0.0001,
+      "loss": 0.1106,
+      "step": 1490
+    },
+    {
+      "epoch": 8.875739644970414,
+      "grad_norm": 7.9275102615356445,
+      "learning_rate": 0.0001,
+      "loss": 0.1152,
+      "step": 1500
+    },
+    {
+      "epoch": 8.934911242603551,
+      "grad_norm": 9.903050422668457,
+      "learning_rate": 0.0001,
+      "loss": 0.115,
+      "step": 1510
+    },
+    {
+      "epoch": 8.994082840236686,
+      "grad_norm": 8.086341857910156,
+      "learning_rate": 0.0001,
+      "loss": 0.1147,
+      "step": 1520
+    },
+    {
+      "epoch": 9.053254437869823,
+      "grad_norm": 7.353363037109375,
+      "learning_rate": 0.0001,
+      "loss": 0.1062,
+      "step": 1530
+    },
+    {
+      "epoch": 9.112426035502958,
+      "grad_norm": 9.058284759521484,
+      "learning_rate": 0.0001,
+      "loss": 0.1087,
+      "step": 1540
+    },
+    {
+      "epoch": 9.171597633136095,
+      "grad_norm": 7.744143009185791,
+      "learning_rate": 0.0001,
+      "loss": 0.111,
+      "step": 1550
+    },
+    {
+      "epoch": 9.23076923076923,
+      "grad_norm": 7.199446678161621,
+      "learning_rate": 0.0001,
+      "loss": 0.1102,
+      "step": 1560
+    },
+    {
+      "epoch": 9.289940828402367,
+      "grad_norm": 8.082643508911133,
+      "learning_rate": 0.0001,
+      "loss": 0.1079,
+      "step": 1570
+    },
+    {
+      "epoch": 9.349112426035504,
+      "grad_norm": 7.800445556640625,
+      "learning_rate": 0.0001,
+      "loss": 0.1088,
+      "step": 1580
+    },
+    {
+      "epoch": 9.408284023668639,
+      "grad_norm": 7.523202419281006,
+      "learning_rate": 0.0001,
+      "loss": 0.1092,
+      "step": 1590
+    },
+    {
+      "epoch": 9.467455621301776,
+      "grad_norm": 6.986664295196533,
+      "learning_rate": 0.0001,
+      "loss": 0.1101,
+      "step": 1600
+    },
+    {
+      "epoch": 9.52662721893491,
+      "grad_norm": 6.420450687408447,
+      "learning_rate": 0.0001,
+      "loss": 0.1029,
+      "step": 1610
+    },
+    {
+      "epoch": 9.585798816568047,
+      "grad_norm": 6.502248287200928,
+      "learning_rate": 0.0001,
+      "loss": 0.1053,
+      "step": 1620
+    },
+    {
+      "epoch": 9.644970414201183,
+      "grad_norm": 6.941205024719238,
+      "learning_rate": 0.0001,
+      "loss": 0.1029,
+      "step": 1630
+    },
+    {
+      "epoch": 9.70414201183432,
+      "grad_norm": 7.186707496643066,
+      "learning_rate": 0.0001,
+      "loss": 0.1039,
+      "step": 1640
+    },
+    {
+      "epoch": 9.763313609467456,
+      "grad_norm": 7.567976951599121,
+      "learning_rate": 0.0001,
+      "loss": 0.1088,
+      "step": 1650
+    },
+    {
+      "epoch": 9.822485207100591,
+      "grad_norm": 8.716438293457031,
+      "learning_rate": 0.0001,
+      "loss": 0.1031,
+      "step": 1660
+    },
+    {
+      "epoch": 9.881656804733728,
+      "grad_norm": 7.470892429351807,
+      "learning_rate": 0.0001,
+      "loss": 0.105,
+      "step": 1670
+    },
+    {
+      "epoch": 9.940828402366863,
+      "grad_norm": 6.363335609436035,
+      "learning_rate": 0.0001,
+      "loss": 0.1049,
+      "step": 1680
+    },
+    {
+      "epoch": 10.0,
+      "grad_norm": 7.282110691070557,
+      "learning_rate": 0.0001,
+      "loss": 0.1026,
+      "step": 1690
+    },
+    {
+      "epoch": 10.059171597633137,
+      "grad_norm": 7.072840213775635,
+      "learning_rate": 0.0001,
+      "loss": 0.1039,
+      "step": 1700
+    },
+    {
+      "epoch": 10.118343195266272,
+      "grad_norm": 8.16348934173584,
+      "learning_rate": 0.0001,
+      "loss": 0.105,
+      "step": 1710
+    },
+    {
+      "epoch": 10.177514792899409,
+      "grad_norm": 7.419730186462402,
+      "learning_rate": 0.0001,
+      "loss": 0.1016,
+      "step": 1720
+    },
+    {
+      "epoch": 10.236686390532544,
+      "grad_norm": 7.181439399719238,
+      "learning_rate": 0.0001,
+      "loss": 0.1049,
+      "step": 1730
+    },
+    {
+      "epoch": 10.29585798816568,
+      "grad_norm": 6.2606730461120605,
+      "learning_rate": 0.0001,
+      "loss": 0.1028,
+      "step": 1740
+    },
+    {
+      "epoch": 10.355029585798816,
+      "grad_norm": 7.7327799797058105,
+      "learning_rate": 0.0001,
+      "loss": 0.0984,
+      "step": 1750
+    },
+    {
+      "epoch": 10.414201183431953,
+      "grad_norm": 6.06395149230957,
+      "learning_rate": 0.0001,
+      "loss": 0.0992,
+      "step": 1760
+    },
+    {
+      "epoch": 10.47337278106509,
+      "grad_norm": 7.358921051025391,
+      "learning_rate": 0.0001,
+      "loss": 0.0989,
+      "step": 1770
+    },
+    {
+      "epoch": 10.532544378698224,
+      "grad_norm": 8.369234085083008,
+      "learning_rate": 0.0001,
+      "loss": 0.1019,
+      "step": 1780
+    },
+    {
+      "epoch": 10.591715976331361,
+      "grad_norm": 7.5672760009765625,
+      "learning_rate": 0.0001,
+      "loss": 0.0942,
+      "step": 1790
+    },
+    {
+      "epoch": 10.650887573964496,
+      "grad_norm": 7.100854396820068,
+      "learning_rate": 0.0001,
+      "loss": 0.0945,
+      "step": 1800
+    },
+    {
+      "epoch": 10.710059171597633,
+      "grad_norm": 7.470778465270996,
+      "learning_rate": 0.0001,
+      "loss": 0.0948,
+      "step": 1810
+    },
+    {
+      "epoch": 10.76923076923077,
+      "grad_norm": 7.235325336456299,
+      "learning_rate": 0.0001,
+      "loss": 0.0944,
+      "step": 1820
+    },
+    {
+      "epoch": 10.828402366863905,
+      "grad_norm": 7.238568305969238,
+      "learning_rate": 0.0001,
+      "loss": 0.0956,
+      "step": 1830
+    },
+    {
+      "epoch": 10.887573964497042,
+      "grad_norm": 8.64164924621582,
+      "learning_rate": 0.0001,
+      "loss": 0.0987,
+      "step": 1840
+    },
+    {
+      "epoch": 10.946745562130177,
+      "grad_norm": 6.154812335968018,
+      "learning_rate": 0.0001,
+      "loss": 0.0986,
+      "step": 1850
+    },
+    {
+      "epoch": 11.005917159763314,
+      "grad_norm": 6.677303314208984,
+      "learning_rate": 0.0001,
+      "loss": 0.0947,
+      "step": 1860
+    },
+    {
+      "epoch": 11.065088757396449,
+      "grad_norm": 6.930010795593262,
+      "learning_rate": 0.0001,
+      "loss": 0.0926,
+      "step": 1870
+    },
+    {
+      "epoch": 11.124260355029586,
+      "grad_norm": 6.618977069854736,
+      "learning_rate": 0.0001,
+      "loss": 0.0934,
+      "step": 1880
+    },
+    {
+      "epoch": 11.183431952662723,
+      "grad_norm": 6.593055248260498,
+      "learning_rate": 0.0001,
+      "loss": 0.0919,
+      "step": 1890
+    },
+    {
+      "epoch": 11.242603550295858,
+      "grad_norm": 6.386303424835205,
+      "learning_rate": 0.0001,
+      "loss": 0.091,
+      "step": 1900
+    },
+    {
+      "epoch": 11.301775147928995,
+      "grad_norm": 7.087122440338135,
+      "learning_rate": 0.0001,
+      "loss": 0.0888,
+      "step": 1910
+    },
+    {
+      "epoch": 11.36094674556213,
+      "grad_norm": 5.8347978591918945,
+      "learning_rate": 0.0001,
+      "loss": 0.0899,
+      "step": 1920
+    },
+    {
+      "epoch": 11.420118343195266,
+      "grad_norm": 7.013113975524902,
+      "learning_rate": 0.0001,
+      "loss": 0.0944,
+      "step": 1930
+    },
+    {
+      "epoch": 11.479289940828401,
+      "grad_norm": 5.794093132019043,
+      "learning_rate": 0.0001,
+      "loss": 0.0945,
+      "step": 1940
+    },
+    {
+      "epoch": 11.538461538461538,
+      "grad_norm": 7.019704818725586,
+      "learning_rate": 0.0001,
+      "loss": 0.0928,
+      "step": 1950
+    },
+    {
+      "epoch": 11.597633136094675,
+      "grad_norm": 6.52449893951416,
+      "learning_rate": 0.0001,
+      "loss": 0.0912,
+      "step": 1960
+    },
+    {
+      "epoch": 11.65680473372781,
+      "grad_norm": 5.8095316886901855,
+      "learning_rate": 0.0001,
+      "loss": 0.0955,
+      "step": 1970
+    },
+    {
+      "epoch": 11.715976331360947,
+      "grad_norm": 5.927523136138916,
+      "learning_rate": 0.0001,
+      "loss": 0.0947,
+      "step": 1980
+    },
+    {
+      "epoch": 11.775147928994082,
+      "grad_norm": 6.514799118041992,
+      "learning_rate": 0.0001,
+      "loss": 0.0866,
+      "step": 1990
+    },
+    {
+      "epoch": 11.834319526627219,
+      "grad_norm": 6.5933966636657715,
+      "learning_rate": 0.0001,
+      "loss": 0.0886,
+      "step": 2000
+    },
+    {
+      "epoch": 11.893491124260356,
+      "grad_norm": 6.423705577850342,
+      "learning_rate": 0.0001,
+      "loss": 0.0898,
+      "step": 2010
+    },
+    {
+      "epoch": 11.95266272189349,
+      "grad_norm": 7.474347114562988,
+      "learning_rate": 0.0001,
+      "loss": 0.0885,
+      "step": 2020
+    },
+    {
+      "epoch": 12.011834319526628,
+      "grad_norm": 6.661672115325928,
+      "learning_rate": 0.0001,
+      "loss": 0.087,
+      "step": 2030
+    },
+    {
+      "epoch": 12.071005917159763,
+      "grad_norm": 6.7026567459106445,
+      "learning_rate": 0.0001,
+      "loss": 0.0854,
+      "step": 2040
+    },
+    {
+      "epoch": 12.1301775147929,
+      "grad_norm": 6.045496940612793,
+      "learning_rate": 0.0001,
+      "loss": 0.0847,
+      "step": 2050
+    },
+    {
+      "epoch": 12.189349112426035,
+      "grad_norm": 6.913471221923828,
+      "learning_rate": 0.0001,
+      "loss": 0.0845,
+      "step": 2060
+    },
+    {
+      "epoch": 12.248520710059172,
+      "grad_norm": 5.359554290771484,
+      "learning_rate": 0.0001,
+      "loss": 0.0841,
+      "step": 2070
+    },
+    {
+      "epoch": 12.307692307692308,
+      "grad_norm": 6.624461650848389,
+      "learning_rate": 0.0001,
+      "loss": 0.0837,
+      "step": 2080
+    },
+    {
+      "epoch": 12.366863905325443,
+      "grad_norm": 6.005741119384766,
+      "learning_rate": 0.0001,
+      "loss": 0.0841,
+      "step": 2090
+    },
+    {
+      "epoch": 12.42603550295858,
+      "grad_norm": 6.167571544647217,
+      "learning_rate": 0.0001,
+      "loss": 0.0839,
+      "step": 2100
+    },
+    {
+      "epoch": 12.485207100591715,
+      "grad_norm": 5.949735641479492,
+      "learning_rate": 0.0001,
+      "loss": 0.0869,
+      "step": 2110
+    },
+    {
+      "epoch": 12.544378698224852,
+      "grad_norm": 5.554780960083008,
+      "learning_rate": 0.0001,
+      "loss": 0.0803,
+      "step": 2120
+    },
+    {
+      "epoch": 12.603550295857989,
+      "grad_norm": 6.141355037689209,
+      "learning_rate": 0.0001,
+      "loss": 0.0818,
+      "step": 2130
+    },
+    {
+      "epoch": 12.662721893491124,
+      "grad_norm": 6.558176040649414,
+      "learning_rate": 0.0001,
+      "loss": 0.0839,
+      "step": 2140
+    },
+    {
+      "epoch": 12.721893491124261,
+      "grad_norm": 5.761529445648193,
+      "learning_rate": 0.0001,
+      "loss": 0.0836,
+      "step": 2150
+    },
+    {
+      "epoch": 12.781065088757396,
+      "grad_norm": 5.793097496032715,
+      "learning_rate": 0.0001,
+      "loss": 0.0831,
+      "step": 2160
+    },
+    {
+      "epoch": 12.840236686390533,
+      "grad_norm": 4.7468976974487305,
+      "learning_rate": 0.0001,
+      "loss": 0.0783,
+      "step": 2170
+    },
+    {
+      "epoch": 12.899408284023668,
+      "grad_norm": 4.700244903564453,
+      "learning_rate": 0.0001,
+      "loss": 0.0816,
+      "step": 2180
+    },
+    {
+      "epoch": 12.958579881656805,
+      "grad_norm": 6.6420698165893555,
+      "learning_rate": 0.0001,
+      "loss": 0.0834,
+      "step": 2190
+    },
+    {
+      "epoch": 13.017751479289942,
+      "grad_norm": 5.871669292449951,
+      "learning_rate": 0.0001,
+      "loss": 0.0825,
+      "step": 2200
+    },
+    {
+      "epoch": 13.076923076923077,
+      "grad_norm": 5.816864013671875,
+      "learning_rate": 0.0001,
+      "loss": 0.0821,
+      "step": 2210
+    },
+    {
+      "epoch": 13.136094674556213,
+      "grad_norm": 6.098887920379639,
+      "learning_rate": 0.0001,
+      "loss": 0.0798,
+      "step": 2220
+    },
+    {
+      "epoch": 13.195266272189349,
+      "grad_norm": 5.793552398681641,
+      "learning_rate": 0.0001,
+      "loss": 0.0787,
+      "step": 2230
+    },
+    {
+      "epoch": 13.254437869822485,
+      "grad_norm": 6.018148422241211,
+      "learning_rate": 0.0001,
+      "loss": 0.0769,
+      "step": 2240
+    },
+    {
+      "epoch": 13.31360946745562,
+      "grad_norm": 5.787800312042236,
+      "learning_rate": 0.0001,
+      "loss": 0.0782,
+      "step": 2250
+    },
+    {
+      "epoch": 13.372781065088757,
+      "grad_norm": 5.947436332702637,
+      "learning_rate": 0.0001,
+      "loss": 0.0762,
+      "step": 2260
+    },
+    {
+      "epoch": 13.431952662721894,
+      "grad_norm": 6.586573600769043,
+      "learning_rate": 0.0001,
+      "loss": 0.0729,
+      "step": 2270
+    },
+    {
+      "epoch": 13.49112426035503,
+      "grad_norm": 6.115123748779297,
+      "learning_rate": 0.0001,
+      "loss": 0.0751,
+      "step": 2280
+    },
+    {
+      "epoch": 13.550295857988166,
+      "grad_norm": 5.407589912414551,
+      "learning_rate": 0.0001,
+      "loss": 0.0731,
+      "step": 2290
+    },
+    {
+      "epoch": 13.609467455621301,
+      "grad_norm": 5.538122653961182,
+      "learning_rate": 0.0001,
+      "loss": 0.076,
+      "step": 2300
+    },
+    {
+      "epoch": 13.668639053254438,
+      "grad_norm": 5.655011177062988,
+      "learning_rate": 0.0001,
+      "loss": 0.0734,
+      "step": 2310
+    },
+    {
+      "epoch": 13.727810650887575,
+      "grad_norm": 5.045595169067383,
+      "learning_rate": 0.0001,
+      "loss": 0.0727,
+      "step": 2320
+    },
+    {
+      "epoch": 13.78698224852071,
+      "grad_norm": 6.3624396324157715,
+      "learning_rate": 0.0001,
+      "loss": 0.0705,
+      "step": 2330
+    },
+    {
+      "epoch": 13.846153846153847,
+      "grad_norm": 5.380771636962891,
+      "learning_rate": 0.0001,
+      "loss": 0.0724,
+      "step": 2340
+    },
+    {
+      "epoch": 13.905325443786982,
+      "grad_norm": 4.899779319763184,
+      "learning_rate": 0.0001,
+      "loss": 0.0711,
+      "step": 2350
+    },
+    {
+      "epoch": 13.964497041420119,
+      "grad_norm": 4.387864589691162,
+      "learning_rate": 0.0001,
+      "loss": 0.0749,
+      "step": 2360
+    },
+    {
+      "epoch": 14.023668639053254,
+      "grad_norm": 5.456191539764404,
+      "learning_rate": 0.0001,
+      "loss": 0.0755,
+      "step": 2370
+    },
+    {
+      "epoch": 14.08284023668639,
+      "grad_norm": 4.455077648162842,
+      "learning_rate": 0.0001,
+      "loss": 0.075,
+      "step": 2380
+    },
+    {
+      "epoch": 14.142011834319527,
+      "grad_norm": 4.902334690093994,
+      "learning_rate": 0.0001,
+      "loss": 0.0738,
+      "step": 2390
+    },
+    {
+      "epoch": 14.201183431952662,
+      "grad_norm": 5.609550952911377,
+      "learning_rate": 0.0001,
+      "loss": 0.0768,
+      "step": 2400
+    },
+    {
+      "epoch": 14.2603550295858,
+      "grad_norm": 5.279762268066406,
+      "learning_rate": 0.0001,
+      "loss": 0.0753,
+      "step": 2410
+    },
+    {
+      "epoch": 14.319526627218934,
+      "grad_norm": 5.009883880615234,
+      "learning_rate": 0.0001,
+      "loss": 0.0734,
+      "step": 2420
+    },
+    {
+      "epoch": 14.378698224852071,
+      "grad_norm": 5.5598039627075195,
+      "learning_rate": 0.0001,
+      "loss": 0.0713,
+      "step": 2430
+    },
+    {
+      "epoch": 14.437869822485208,
+      "grad_norm": 4.762701988220215,
+      "learning_rate": 0.0001,
+      "loss": 0.0715,
+      "step": 2440
+    },
+    {
+      "epoch": 14.497041420118343,
+      "grad_norm": 5.417370319366455,
+      "learning_rate": 0.0001,
+      "loss": 0.0721,
+      "step": 2450
+    },
+    {
+      "epoch": 14.55621301775148,
+      "grad_norm": 6.082530498504639,
+      "learning_rate": 0.0001,
+      "loss": 0.0724,
+      "step": 2460
+    },
+    {
+      "epoch": 14.615384615384615,
+      "grad_norm": 5.353112697601318,
+      "learning_rate": 0.0001,
+      "loss": 0.0707,
+      "step": 2470
+    },
+    {
+      "epoch": 14.674556213017752,
+      "grad_norm": 5.059446811676025,
+      "learning_rate": 0.0001,
+      "loss": 0.0716,
+      "step": 2480
+    },
+    {
+      "epoch": 14.733727810650887,
+      "grad_norm": 5.622556686401367,
+      "learning_rate": 0.0001,
+      "loss": 0.0705,
+      "step": 2490
+    },
+    {
+      "epoch": 14.792899408284024,
+      "grad_norm": 5.557983875274658,
+      "learning_rate": 0.0001,
+      "loss": 0.0698,
+      "step": 2500
+    },
+    {
+      "epoch": 14.85207100591716,
+      "grad_norm": 4.384890079498291,
+      "learning_rate": 0.0001,
+      "loss": 0.0712,
+      "step": 2510
+    },
+    {
+      "epoch": 14.911242603550296,
+      "grad_norm": 4.381173133850098,
+      "learning_rate": 0.0001,
+      "loss": 0.0676,
+      "step": 2520
+    },
+    {
+      "epoch": 14.970414201183432,
+      "grad_norm": 4.982142448425293,
+      "learning_rate": 0.0001,
+      "loss": 0.068,
+      "step": 2530
+    },
+    {
+      "epoch": 15.029585798816568,
+      "grad_norm": 5.8004469871521,
+      "learning_rate": 0.0001,
+      "loss": 0.0689,
+      "step": 2540
+    },
+    {
+      "epoch": 15.088757396449704,
+      "grad_norm": 5.157318592071533,
+      "learning_rate": 0.0001,
+      "loss": 0.0685,
+      "step": 2550
+    },
+    {
+      "epoch": 15.14792899408284,
+      "grad_norm": 6.35864782333374,
+      "learning_rate": 0.0001,
+      "loss": 0.0697,
+      "step": 2560
+    },
+    {
+      "epoch": 15.207100591715976,
+      "grad_norm": 4.64348840713501,
+      "learning_rate": 0.0001,
+      "loss": 0.0637,
+      "step": 2570
+    },
+    {
+      "epoch": 15.266272189349113,
+      "grad_norm": 5.300350189208984,
+      "learning_rate": 0.0001,
+      "loss": 0.0674,
+      "step": 2580
+    },
+    {
+      "epoch": 15.325443786982248,
+      "grad_norm": 5.179243087768555,
+      "learning_rate": 0.0001,
+      "loss": 0.0662,
+      "step": 2590
+    },
+    {
+      "epoch": 15.384615384615385,
+      "grad_norm": 5.599083423614502,
+      "learning_rate": 0.0001,
+      "loss": 0.0675,
+      "step": 2600
+    },
+    {
+      "epoch": 15.44378698224852,
+      "grad_norm": 5.323422908782959,
+      "learning_rate": 0.0001,
+      "loss": 0.068,
+      "step": 2610
+    },
+    {
+      "epoch": 15.502958579881657,
+      "grad_norm": 4.894012928009033,
+      "learning_rate": 0.0001,
+      "loss": 0.0664,
+      "step": 2620
+    },
+    {
+      "epoch": 15.562130177514792,
+      "grad_norm": 5.003327369689941,
+      "learning_rate": 0.0001,
+      "loss": 0.0667,
+      "step": 2630
+    },
+    {
+      "epoch": 15.621301775147929,
+      "grad_norm": 4.582026481628418,
+      "learning_rate": 0.0001,
+      "loss": 0.067,
+      "step": 2640
+    },
+    {
+      "epoch": 15.680473372781066,
+      "grad_norm": 4.6689605712890625,
+      "learning_rate": 0.0001,
+      "loss": 0.0661,
+      "step": 2650
+    },
+    {
+      "epoch": 15.7396449704142,
+      "grad_norm": 4.672581195831299,
+      "learning_rate": 0.0001,
+      "loss": 0.0651,
+      "step": 2660
+    },
+    {
+      "epoch": 15.798816568047338,
+      "grad_norm": 4.754590034484863,
+      "learning_rate": 0.0001,
+      "loss": 0.0653,
+      "step": 2670
+    },
+    {
+      "epoch": 15.857988165680473,
+      "grad_norm": 5.777609825134277,
+      "learning_rate": 0.0001,
+      "loss": 0.066,
+      "step": 2680
+    },
+    {
+      "epoch": 15.91715976331361,
+      "grad_norm": 4.492793560028076,
+      "learning_rate": 0.0001,
+      "loss": 0.0657,
+      "step": 2690
+    },
+    {
+      "epoch": 15.976331360946746,
+      "grad_norm": 4.068188667297363,
+      "learning_rate": 0.0001,
+      "loss": 0.0635,
+      "step": 2700
+    },
+    {
+      "epoch": 16.035502958579883,
+      "grad_norm": 5.356994152069092,
+      "learning_rate": 0.0001,
+      "loss": 0.0638,
+      "step": 2710
+    },
+    {
+      "epoch": 16.09467455621302,
+      "grad_norm": 4.007376670837402,
+      "learning_rate": 0.0001,
+      "loss": 0.0643,
+      "step": 2720
+    },
+    {
+      "epoch": 16.153846153846153,
+      "grad_norm": 4.8432464599609375,
+      "learning_rate": 0.0001,
+      "loss": 0.0651,
+      "step": 2730
+    },
+    {
+      "epoch": 16.21301775147929,
+      "grad_norm": 5.0063629150390625,
+      "learning_rate": 0.0001,
+      "loss": 0.0666,
+      "step": 2740
+    },
+    {
+      "epoch": 16.272189349112427,
+      "grad_norm": 5.3229193687438965,
+      "learning_rate": 0.0001,
+      "loss": 0.0652,
+      "step": 2750
+    },
+    {
+      "epoch": 16.331360946745562,
+      "grad_norm": 4.688714981079102,
+      "learning_rate": 0.0001,
+      "loss": 0.065,
+      "step": 2760
+    },
+    {
+      "epoch": 16.390532544378697,
+      "grad_norm": 5.109912872314453,
+      "learning_rate": 0.0001,
+      "loss": 0.0668,
+      "step": 2770
+    },
+    {
+      "epoch": 16.449704142011836,
+      "grad_norm": 4.174536228179932,
+      "learning_rate": 0.0001,
+      "loss": 0.0641,
+      "step": 2780
+    },
+    {
+      "epoch": 16.50887573964497,
+      "grad_norm": 5.154987335205078,
+      "learning_rate": 0.0001,
+      "loss": 0.0651,
+      "step": 2790
+    },
+    {
+      "epoch": 16.568047337278106,
+      "grad_norm": 5.064062595367432,
+      "learning_rate": 0.0001,
+      "loss": 0.0652,
+      "step": 2800
+    },
+    {
+      "epoch": 16.62721893491124,
+      "grad_norm": 4.491312503814697,
+      "learning_rate": 0.0001,
+      "loss": 0.0629,
+      "step": 2810
+    },
+    {
+      "epoch": 16.68639053254438,
+      "grad_norm": 4.9111528396606445,
+      "learning_rate": 0.0001,
+      "loss": 0.0664,
+      "step": 2820
+    },
+    {
+      "epoch": 16.745562130177515,
+      "grad_norm": 4.732112884521484,
+      "learning_rate": 0.0001,
+      "loss": 0.065,
+      "step": 2830
+    },
+    {
+      "epoch": 16.80473372781065,
+      "grad_norm": 5.044336795806885,
+      "learning_rate": 0.0001,
+      "loss": 0.0627,
+      "step": 2840
+    },
+    {
+      "epoch": 16.86390532544379,
+      "grad_norm": 5.345180988311768,
+      "learning_rate": 0.0001,
+      "loss": 0.0661,
+      "step": 2850
+    },
+    {
+      "epoch": 16.923076923076923,
+      "grad_norm": 5.118403434753418,
+      "learning_rate": 0.0001,
+      "loss": 0.0646,
+      "step": 2860
+    },
+    {
+      "epoch": 16.98224852071006,
+      "grad_norm": 5.003620147705078,
+      "learning_rate": 0.0001,
+      "loss": 0.0659,
+      "step": 2870
+    },
+    {
+      "epoch": 17.041420118343197,
+      "grad_norm": 5.433116912841797,
+      "learning_rate": 0.0001,
+      "loss": 0.066,
+      "step": 2880
+    },
+    {
+      "epoch": 17.100591715976332,
+      "grad_norm": 5.044699668884277,
+      "learning_rate": 0.0001,
+      "loss": 0.0646,
+      "step": 2890
+    },
+    {
+      "epoch": 17.159763313609467,
+      "grad_norm": 4.479295253753662,
+      "learning_rate": 0.0001,
+      "loss": 0.0647,
+      "step": 2900
+    },
+    {
+      "epoch": 17.218934911242602,
+      "grad_norm": 4.760400295257568,
+      "learning_rate": 0.0001,
+      "loss": 0.0649,
+      "step": 2910
+    },
+    {
+      "epoch": 17.27810650887574,
+      "grad_norm": 4.305962085723877,
+      "learning_rate": 0.0001,
+      "loss": 0.067,
+      "step": 2920
+    },
+    {
+      "epoch": 17.337278106508876,
+      "grad_norm": 4.982907295227051,
+      "learning_rate": 0.0001,
+      "loss": 0.0623,
+      "step": 2930
+    },
+    {
+      "epoch": 17.39644970414201,
+      "grad_norm": 5.5369181632995605,
+      "learning_rate": 0.0001,
+      "loss": 0.0645,
+      "step": 2940
+    },
+    {
+      "epoch": 17.45562130177515,
+      "grad_norm": 4.570610046386719,
+      "learning_rate": 0.0001,
+      "loss": 0.0662,
+      "step": 2950
+    },
+    {
+      "epoch": 17.514792899408285,
+      "grad_norm": 4.763125896453857,
+      "learning_rate": 0.0001,
+      "loss": 0.0627,
+      "step": 2960
+    },
+    {
+      "epoch": 17.57396449704142,
+      "grad_norm": 5.046164035797119,
+      "learning_rate": 0.0001,
+      "loss": 0.0613,
+      "step": 2970
+    },
+    {
+      "epoch": 17.633136094674555,
+      "grad_norm": 3.3583176136016846,
+      "learning_rate": 0.0001,
+      "loss": 0.0636,
+      "step": 2980
+    },
+    {
+      "epoch": 17.692307692307693,
+      "grad_norm": 4.258029937744141,
+      "learning_rate": 0.0001,
+      "loss": 0.0617,
+      "step": 2990
+    },
+    {
+      "epoch": 17.75147928994083,
+      "grad_norm": 5.450995922088623,
+      "learning_rate": 0.0001,
+      "loss": 0.0618,
+      "step": 3000
+    },
+    {
+      "epoch": 17.810650887573964,
+      "grad_norm": 4.161924362182617,
+      "learning_rate": 0.0001,
+      "loss": 0.0619,
+      "step": 3010
+    },
+    {
+      "epoch": 17.869822485207102,
+      "grad_norm": 3.873508930206299,
+      "learning_rate": 0.0001,
+      "loss": 0.0612,
+      "step": 3020
+    },
+    {
+      "epoch": 17.928994082840237,
+      "grad_norm": 4.159547805786133,
+      "learning_rate": 0.0001,
+      "loss": 0.0586,
+      "step": 3030
+    },
+    {
+      "epoch": 17.988165680473372,
+      "grad_norm": 4.17963981628418,
+      "learning_rate": 0.0001,
+      "loss": 0.0599,
+      "step": 3040
+    },
+    {
+      "epoch": 18.047337278106507,
+      "grad_norm": 5.066028118133545,
+      "learning_rate": 0.0001,
+      "loss": 0.0604,
+      "step": 3050
+    },
+    {
+      "epoch": 18.106508875739646,
+      "grad_norm": 4.3584465980529785,
+      "learning_rate": 0.0001,
+      "loss": 0.0598,
+      "step": 3060
+    },
+    {
+      "epoch": 18.16568047337278,
+      "grad_norm": 4.816352367401123,
+      "learning_rate": 0.0001,
+      "loss": 0.0586,
+      "step": 3070
+    },
+    {
+      "epoch": 18.224852071005916,
+      "grad_norm": 4.523645877838135,
+      "learning_rate": 0.0001,
+      "loss": 0.059,
+      "step": 3080
+    },
+    {
+      "epoch": 18.284023668639055,
+      "grad_norm": 3.818697690963745,
+      "learning_rate": 0.0001,
+      "loss": 0.0577,
+      "step": 3090
+    },
+    {
+      "epoch": 18.34319526627219,
+      "grad_norm": 3.892241954803467,
+      "learning_rate": 0.0001,
+      "loss": 0.0578,
+      "step": 3100
+    },
+    {
+      "epoch": 18.402366863905325,
+      "grad_norm": 4.067519187927246,
+      "learning_rate": 0.0001,
+      "loss": 0.0561,
+      "step": 3110
+    },
+    {
+      "epoch": 18.46153846153846,
+      "grad_norm": 4.37803840637207,
+      "learning_rate": 0.0001,
+      "loss": 0.0557,
+      "step": 3120
+    },
+    {
+      "epoch": 18.5207100591716,
+      "grad_norm": 4.3098015785217285,
+      "learning_rate": 0.0001,
+      "loss": 0.0585,
+      "step": 3130
+    },
+    {
+      "epoch": 18.579881656804734,
+      "grad_norm": 4.395653247833252,
+      "learning_rate": 0.0001,
+      "loss": 0.0575,
+      "step": 3140
+    },
+    {
+      "epoch": 18.63905325443787,
+      "grad_norm": 4.343656063079834,
+      "learning_rate": 0.0001,
+      "loss": 0.0586,
+      "step": 3150
+    },
+    {
+      "epoch": 18.698224852071007,
+      "grad_norm": 4.551352024078369,
+      "learning_rate": 0.0001,
+      "loss": 0.0589,
+      "step": 3160
+    },
+    {
+      "epoch": 18.757396449704142,
+      "grad_norm": 4.751080513000488,
+      "learning_rate": 0.0001,
+      "loss": 0.0584,
+      "step": 3170
+    },
+    {
+      "epoch": 18.816568047337277,
+      "grad_norm": 4.0395588874816895,
+      "learning_rate": 0.0001,
+      "loss": 0.0585,
+      "step": 3180
+    },
+    {
+      "epoch": 18.875739644970416,
+      "grad_norm": 4.047962188720703,
+      "learning_rate": 0.0001,
+      "loss": 0.0575,
+      "step": 3190
+    },
+    {
+      "epoch": 18.93491124260355,
+      "grad_norm": 3.8828914165496826,
+      "learning_rate": 0.0001,
+      "loss": 0.0582,
+      "step": 3200
+    },
+    {
+      "epoch": 18.994082840236686,
+      "grad_norm": 3.688521146774292,
+      "learning_rate": 0.0001,
+      "loss": 0.057,
+      "step": 3210
+    },
+    {
+      "epoch": 19.05325443786982,
+      "grad_norm": 3.2845399379730225,
+      "learning_rate": 0.0001,
+      "loss": 0.0566,
+      "step": 3220
+    },
+    {
+      "epoch": 19.11242603550296,
+      "grad_norm": 4.365805149078369,
+      "learning_rate": 0.0001,
+      "loss": 0.0567,
+      "step": 3230
+    },
+    {
+      "epoch": 19.171597633136095,
+      "grad_norm": 4.187885761260986,
+      "learning_rate": 0.0001,
+      "loss": 0.0609,
+      "step": 3240
+    },
+    {
+      "epoch": 19.23076923076923,
+      "grad_norm": 4.105882167816162,
+      "learning_rate": 0.0001,
+      "loss": 0.0593,
+      "step": 3250
+    },
+    {
+      "epoch": 19.28994082840237,
+      "grad_norm": 4.762232303619385,
+      "learning_rate": 0.0001,
+      "loss": 0.0611,
+      "step": 3260
+    },
+    {
+      "epoch": 19.349112426035504,
+      "grad_norm": 4.138009071350098,
+      "learning_rate": 0.0001,
+      "loss": 0.0564,
+      "step": 3270
+    },
+    {
+      "epoch": 19.40828402366864,
+      "grad_norm": 4.4852447509765625,
+      "learning_rate": 0.0001,
+      "loss": 0.0573,
+      "step": 3280
+    },
+    {
+      "epoch": 19.467455621301774,
+      "grad_norm": 4.1162238121032715,
+      "learning_rate": 0.0001,
+      "loss": 0.0557,
+      "step": 3290
+    },
+    {
+      "epoch": 19.526627218934912,
+      "grad_norm": 4.385313510894775,
+      "learning_rate": 0.0001,
+      "loss": 0.0575,
+      "step": 3300
+    },
+    {
+      "epoch": 19.585798816568047,
+      "grad_norm": 3.7840652465820312,
+      "learning_rate": 0.0001,
+      "loss": 0.057,
+      "step": 3310
+    },
+    {
+      "epoch": 19.644970414201183,
+      "grad_norm": 4.175228118896484,
+      "learning_rate": 0.0001,
+      "loss": 0.0562,
+      "step": 3320
+    },
+    {
+      "epoch": 19.70414201183432,
+      "grad_norm": 4.3501877784729,
+      "learning_rate": 0.0001,
+      "loss": 0.0553,
+      "step": 3330
+    },
+    {
+      "epoch": 19.763313609467456,
+      "grad_norm": 3.9982478618621826,
+      "learning_rate": 0.0001,
+      "loss": 0.0556,
+      "step": 3340
+    },
+    {
+      "epoch": 19.82248520710059,
+      "grad_norm": 3.5812315940856934,
+      "learning_rate": 0.0001,
+      "loss": 0.0541,
+      "step": 3350
+    },
+    {
+      "epoch": 19.881656804733726,
+      "grad_norm": 4.104846477508545,
+      "learning_rate": 0.0001,
+      "loss": 0.0536,
+      "step": 3360
+    },
+    {
+      "epoch": 19.940828402366865,
+      "grad_norm": 4.363539218902588,
+      "learning_rate": 0.0001,
+      "loss": 0.056,
+      "step": 3370
+    },
+    {
+      "epoch": 20.0,
+      "grad_norm": 4.858595371246338,
+      "learning_rate": 0.0001,
+      "loss": 0.0574,
+      "step": 3380
+    },
+    {
+      "epoch": 20.059171597633135,
+      "grad_norm": 4.302316188812256,
+      "learning_rate": 0.0001,
+      "loss": 0.0541,
+      "step": 3390
+    },
+    {
+      "epoch": 20.118343195266274,
+      "grad_norm": 4.605861663818359,
+      "learning_rate": 0.0001,
+      "loss": 0.0548,
+      "step": 3400
+    },
+    {
+      "epoch": 20.17751479289941,
+      "grad_norm": 4.09200382232666,
+      "learning_rate": 0.0001,
+      "loss": 0.0534,
+      "step": 3410
+    },
+    {
+      "epoch": 20.236686390532544,
+      "grad_norm": 4.432004928588867,
+      "learning_rate": 0.0001,
+      "loss": 0.0536,
+      "step": 3420
+    },
+    {
+      "epoch": 20.29585798816568,
+      "grad_norm": 4.275376796722412,
+      "learning_rate": 0.0001,
+      "loss": 0.0534,
+      "step": 3430
+    },
+    {
+      "epoch": 20.355029585798817,
+      "grad_norm": 3.1744744777679443,
+      "learning_rate": 0.0001,
+      "loss": 0.0523,
+      "step": 3440
+    },
+    {
+      "epoch": 20.414201183431953,
+      "grad_norm": 3.813413143157959,
+      "learning_rate": 0.0001,
+      "loss": 0.0533,
+      "step": 3450
+    },
+    {
+      "epoch": 20.473372781065088,
+      "grad_norm": 3.993971586227417,
+      "learning_rate": 0.0001,
+      "loss": 0.0527,
+      "step": 3460
+    },
+    {
+      "epoch": 20.532544378698226,
+      "grad_norm": 4.192675590515137,
+      "learning_rate": 0.0001,
+      "loss": 0.0512,
+      "step": 3470
+    },
+    {
+      "epoch": 20.59171597633136,
+      "grad_norm": 3.731384515762329,
+      "learning_rate": 0.0001,
+      "loss": 0.0515,
+      "step": 3480
+    },
+    {
+      "epoch": 20.650887573964496,
+      "grad_norm": 3.9884278774261475,
+      "learning_rate": 0.0001,
+      "loss": 0.0524,
+      "step": 3490
+    },
+    {
+      "epoch": 20.71005917159763,
+      "grad_norm": 3.9658291339874268,
+      "learning_rate": 0.0001,
+      "loss": 0.0519,
+      "step": 3500
+    },
+    {
+      "epoch": 20.76923076923077,
+      "grad_norm": 4.243136405944824,
+      "learning_rate": 0.0001,
+      "loss": 0.0536,
+      "step": 3510
+    },
+    {
+      "epoch": 20.828402366863905,
+      "grad_norm": 5.130043029785156,
+      "learning_rate": 0.0001,
+      "loss": 0.0542,
+      "step": 3520
+    },
+    {
+      "epoch": 20.88757396449704,
+      "grad_norm": 4.002963542938232,
+      "learning_rate": 0.0001,
+      "loss": 0.0558,
+      "step": 3530
+    },
+    {
+      "epoch": 20.94674556213018,
+      "grad_norm": 3.8863823413848877,
+      "learning_rate": 0.0001,
+      "loss": 0.0558,
+      "step": 3540
+    },
+    {
+      "epoch": 21.005917159763314,
+      "grad_norm": 4.567497730255127,
+      "learning_rate": 0.0001,
+      "loss": 0.0557,
+      "step": 3550
+    },
+    {
+      "epoch": 21.06508875739645,
+      "grad_norm": 3.500450372695923,
+      "learning_rate": 0.0001,
+      "loss": 0.0538,
+      "step": 3560
+    },
+    {
+      "epoch": 21.124260355029588,
+      "grad_norm": 4.081122398376465,
+      "learning_rate": 0.0001,
+      "loss": 0.0537,
+      "step": 3570
+    },
+    {
+      "epoch": 21.183431952662723,
+      "grad_norm": 3.907001495361328,
+      "learning_rate": 0.0001,
+      "loss": 0.054,
+      "step": 3580
+    },
+    {
+      "epoch": 21.242603550295858,
+      "grad_norm": 4.0510125160217285,
+      "learning_rate": 0.0001,
+      "loss": 0.0528,
+      "step": 3590
+    },
+    {
+      "epoch": 21.301775147928993,
+      "grad_norm": 4.6179585456848145,
+      "learning_rate": 0.0001,
+      "loss": 0.0529,
+      "step": 3600
+    },
+    {
+      "epoch": 21.36094674556213,
+      "grad_norm": 4.210280418395996,
+      "learning_rate": 0.0001,
+      "loss": 0.0542,
+      "step": 3610
+    },
+    {
+      "epoch": 21.420118343195266,
+      "grad_norm": 4.838978290557861,
+      "learning_rate": 0.0001,
+      "loss": 0.0539,
+      "step": 3620
+    },
+    {
+      "epoch": 21.4792899408284,
+      "grad_norm": 4.554750919342041,
+      "learning_rate": 0.0001,
+      "loss": 0.0516,
+      "step": 3630
+    },
+    {
+      "epoch": 21.53846153846154,
+      "grad_norm": 3.2339529991149902,
+      "learning_rate": 0.0001,
+      "loss": 0.0503,
+      "step": 3640
+    },
+    {
+      "epoch": 21.597633136094675,
+      "grad_norm": 3.993428945541382,
+      "learning_rate": 0.0001,
+      "loss": 0.0517,
+      "step": 3650
+    },
+    {
+      "epoch": 21.65680473372781,
+      "grad_norm": 3.662785053253174,
+      "learning_rate": 0.0001,
+      "loss": 0.0508,
+      "step": 3660
+    },
+    {
+      "epoch": 21.715976331360945,
+      "grad_norm": 3.156325578689575,
+      "learning_rate": 0.0001,
+      "loss": 0.0497,
+      "step": 3670
+    },
+    {
+      "epoch": 21.775147928994084,
+      "grad_norm": 3.6561360359191895,
+      "learning_rate": 0.0001,
+      "loss": 0.0505,
+      "step": 3680
+    },
+    {
+      "epoch": 21.83431952662722,
+      "grad_norm": 3.4016590118408203,
+      "learning_rate": 0.0001,
+      "loss": 0.0537,
+      "step": 3690
+    },
+    {
+      "epoch": 21.893491124260354,
+      "grad_norm": 3.8727705478668213,
+      "learning_rate": 0.0001,
+      "loss": 0.0537,
+      "step": 3700
+    },
+    {
+      "epoch": 21.952662721893493,
+      "grad_norm": 4.083427429199219,
+      "learning_rate": 0.0001,
+      "loss": 0.052,
+      "step": 3710
+    },
+    {
+      "epoch": 22.011834319526628,
+      "grad_norm": 3.379897117614746,
+      "learning_rate": 0.0001,
+      "loss": 0.0529,
+      "step": 3720
+    },
+    {
+      "epoch": 22.071005917159763,
+      "grad_norm": 3.5911831855773926,
+      "learning_rate": 0.0001,
+      "loss": 0.0492,
+      "step": 3730
+    },
+    {
+      "epoch": 22.130177514792898,
+      "grad_norm": 3.665997266769409,
+      "learning_rate": 0.0001,
+      "loss": 0.0517,
+      "step": 3740
+    },
+    {
+      "epoch": 22.189349112426036,
+      "grad_norm": 3.501633644104004,
+      "learning_rate": 0.0001,
+      "loss": 0.0517,
+      "step": 3750
+    },
+    {
+      "epoch": 22.24852071005917,
+      "grad_norm": 3.775315999984741,
+      "learning_rate": 0.0001,
+      "loss": 0.0544,
+      "step": 3760
+    },
+    {
+      "epoch": 22.307692307692307,
+      "grad_norm": 3.627082586288452,
+      "learning_rate": 0.0001,
+      "loss": 0.0529,
+      "step": 3770
+    },
+    {
+      "epoch": 22.366863905325445,
+      "grad_norm": 3.722388744354248,
+      "learning_rate": 0.0001,
+      "loss": 0.0516,
+      "step": 3780
+    },
+    {
+      "epoch": 22.42603550295858,
+      "grad_norm": 3.439772367477417,
+      "learning_rate": 0.0001,
+      "loss": 0.0504,
+      "step": 3790
+    },
+    {
+      "epoch": 22.485207100591715,
+      "grad_norm": 3.85457444190979,
+      "learning_rate": 0.0001,
+      "loss": 0.0515,
+      "step": 3800
+    },
+    {
+      "epoch": 22.54437869822485,
+      "grad_norm": 3.7671821117401123,
+      "learning_rate": 0.0001,
+      "loss": 0.052,
+      "step": 3810
+    },
+    {
+      "epoch": 22.60355029585799,
+      "grad_norm": 3.4532511234283447,
+      "learning_rate": 0.0001,
+      "loss": 0.0505,
+      "step": 3820
+    },
+    {
+      "epoch": 22.662721893491124,
+      "grad_norm": 3.736163377761841,
+      "learning_rate": 0.0001,
+      "loss": 0.051,
+      "step": 3830
+    },
+    {
+      "epoch": 22.72189349112426,
+      "grad_norm": 3.307004928588867,
+      "learning_rate": 0.0001,
+      "loss": 0.0505,
+      "step": 3840
+    },
+    {
+      "epoch": 22.781065088757398,
+      "grad_norm": 4.597588539123535,
+      "learning_rate": 0.0001,
+      "loss": 0.0516,
+      "step": 3850
+    },
+    {
+      "epoch": 22.840236686390533,
+      "grad_norm": 3.2445716857910156,
+      "learning_rate": 0.0001,
+      "loss": 0.051,
+      "step": 3860
+    },
+    {
+      "epoch": 22.899408284023668,
+      "grad_norm": 3.028111457824707,
+      "learning_rate": 0.0001,
+      "loss": 0.0512,
+      "step": 3870
+    },
+    {
+      "epoch": 22.958579881656803,
+      "grad_norm": 3.5591249465942383,
+      "learning_rate": 0.0001,
+      "loss": 0.0487,
+      "step": 3880
+    },
+    {
+      "epoch": 23.01775147928994,
+      "grad_norm": 3.628044843673706,
+      "learning_rate": 0.0001,
+      "loss": 0.0495,
+      "step": 3890
+    },
+    {
+      "epoch": 23.076923076923077,
+      "grad_norm": 3.6083438396453857,
+      "learning_rate": 0.0001,
+      "loss": 0.053,
+      "step": 3900
+    },
+    {
+      "epoch": 23.13609467455621,
+      "grad_norm": 3.8031418323516846,
+      "learning_rate": 0.0001,
+      "loss": 0.051,
+      "step": 3910
+    },
+    {
+      "epoch": 23.19526627218935,
+      "grad_norm": 3.0712578296661377,
+      "learning_rate": 0.0001,
+      "loss": 0.0485,
+      "step": 3920
+    },
+    {
+      "epoch": 23.254437869822485,
+      "grad_norm": 3.4879884719848633,
+      "learning_rate": 0.0001,
+      "loss": 0.0497,
+      "step": 3930
+    },
+    {
+      "epoch": 23.31360946745562,
+      "grad_norm": 3.487415075302124,
+      "learning_rate": 0.0001,
+      "loss": 0.0514,
+      "step": 3940
+    },
+    {
+      "epoch": 23.37278106508876,
+      "grad_norm": 3.4723920822143555,
+      "learning_rate": 0.0001,
+      "loss": 0.0512,
+      "step": 3950
+    },
+    {
+      "epoch": 23.431952662721894,
+      "grad_norm": 4.087646961212158,
+      "learning_rate": 0.0001,
+      "loss": 0.0499,
+      "step": 3960
+    },
+    {
+      "epoch": 23.49112426035503,
+      "grad_norm": 4.027114391326904,
+      "learning_rate": 0.0001,
+      "loss": 0.0518,
+      "step": 3970
+    },
+    {
+      "epoch": 23.550295857988164,
+      "grad_norm": 3.307061195373535,
+      "learning_rate": 0.0001,
+      "loss": 0.0505,
+      "step": 3980
+    },
+    {
+      "epoch": 23.609467455621303,
+      "grad_norm": 3.3092315196990967,
+      "learning_rate": 0.0001,
+      "loss": 0.0509,
+      "step": 3990
+    },
+    {
+      "epoch": 23.668639053254438,
+      "grad_norm": 3.17655348777771,
+      "learning_rate": 0.0001,
+      "loss": 0.0494,
+      "step": 4000
+    },
+    {
+      "epoch": 23.727810650887573,
+      "grad_norm": 2.8691954612731934,
+      "learning_rate": 0.0001,
+      "loss": 0.0487,
+      "step": 4010
+    },
+    {
+      "epoch": 23.78698224852071,
+      "grad_norm": 2.9267141819000244,
+      "learning_rate": 0.0001,
+      "loss": 0.048,
+      "step": 4020
+    },
+    {
+      "epoch": 23.846153846153847,
+      "grad_norm": 3.8273580074310303,
+      "learning_rate": 0.0001,
+      "loss": 0.0464,
+      "step": 4030
+    },
+    {
+      "epoch": 23.90532544378698,
+      "grad_norm": 4.048878192901611,
+      "learning_rate": 0.0001,
+      "loss": 0.0481,
+      "step": 4040
+    },
+    {
+      "epoch": 23.964497041420117,
+      "grad_norm": 3.2853825092315674,
+      "learning_rate": 0.0001,
+      "loss": 0.0478,
+      "step": 4050
+    },
+    {
+      "epoch": 24.023668639053255,
+      "grad_norm": 3.2402687072753906,
+      "learning_rate": 0.0001,
+      "loss": 0.0494,
+      "step": 4060
+    },
+    {
+      "epoch": 24.08284023668639,
+      "grad_norm": 2.8202438354492188,
+      "learning_rate": 0.0001,
+      "loss": 0.0493,
+      "step": 4070
+    },
+    {
+      "epoch": 24.142011834319526,
+      "grad_norm": 3.909874439239502,
+      "learning_rate": 0.0001,
+      "loss": 0.0477,
+      "step": 4080
+    },
+    {
+      "epoch": 24.201183431952664,
+      "grad_norm": 3.88059663772583,
+      "learning_rate": 0.0001,
+      "loss": 0.0477,
+      "step": 4090
+    },
+    {
+      "epoch": 24.2603550295858,
+      "grad_norm": 3.1255383491516113,
+      "learning_rate": 0.0001,
+      "loss": 0.0483,
+      "step": 4100
+    },
+    {
+      "epoch": 24.319526627218934,
+      "grad_norm": 4.056564807891846,
+      "learning_rate": 0.0001,
+      "loss": 0.0502,
+      "step": 4110
+    },
+    {
+      "epoch": 24.37869822485207,
+      "grad_norm": 3.2773141860961914,
+      "learning_rate": 0.0001,
+      "loss": 0.0487,
+      "step": 4120
+    },
+    {
+      "epoch": 24.437869822485208,
+      "grad_norm": 3.534104585647583,
+      "learning_rate": 0.0001,
+      "loss": 0.0477,
+      "step": 4130
+    },
+    {
+      "epoch": 24.497041420118343,
+      "grad_norm": 3.571101665496826,
+      "learning_rate": 0.0001,
+      "loss": 0.0477,
+      "step": 4140
+    },
+    {
+      "epoch": 24.556213017751478,
+      "grad_norm": 3.6277639865875244,
+      "learning_rate": 0.0001,
+      "loss": 0.0475,
+      "step": 4150
+    },
+    {
+      "epoch": 24.615384615384617,
+      "grad_norm": 3.156118154525757,
+      "learning_rate": 0.0001,
+      "loss": 0.0474,
+      "step": 4160
+    },
+    {
+      "epoch": 24.674556213017752,
+      "grad_norm": 3.648056983947754,
+      "learning_rate": 0.0001,
+      "loss": 0.048,
+      "step": 4170
+    },
+    {
+      "epoch": 24.733727810650887,
+      "grad_norm": 3.0412302017211914,
+      "learning_rate": 0.0001,
+      "loss": 0.0469,
+      "step": 4180
+    },
+    {
+      "epoch": 24.792899408284022,
+      "grad_norm": 3.4855148792266846,
+      "learning_rate": 0.0001,
+      "loss": 0.047,
+      "step": 4190
+    },
+    {
+      "epoch": 24.85207100591716,
+      "grad_norm": 3.117156744003296,
+      "learning_rate": 0.0001,
+      "loss": 0.0479,
+      "step": 4200
+    },
+    {
+      "epoch": 24.911242603550296,
+      "grad_norm": 3.245882034301758,
+      "learning_rate": 0.0001,
+      "loss": 0.0453,
+      "step": 4210
+    },
+    {
+      "epoch": 24.97041420118343,
+      "grad_norm": 3.722379446029663,
+      "learning_rate": 0.0001,
+      "loss": 0.0456,
+      "step": 4220
+    },
+    {
+      "epoch": 25.02958579881657,
+      "grad_norm": 4.014776229858398,
+      "learning_rate": 0.0001,
+      "loss": 0.046,
+      "step": 4230
+    },
+    {
+      "epoch": 25.088757396449704,
+      "grad_norm": 3.0845611095428467,
+      "learning_rate": 0.0001,
+      "loss": 0.0475,
+      "step": 4240
+    },
+    {
+      "epoch": 25.14792899408284,
+      "grad_norm": 3.1267402172088623,
+      "learning_rate": 0.0001,
+      "loss": 0.0481,
+      "step": 4250
+    },
+    {
+      "epoch": 25.207100591715978,
+      "grad_norm": 3.380932331085205,
+      "learning_rate": 0.0001,
+      "loss": 0.0481,
+      "step": 4260
+    },
+    {
+      "epoch": 25.266272189349113,
+      "grad_norm": 2.927638292312622,
+      "learning_rate": 0.0001,
+      "loss": 0.0488,
+      "step": 4270
+    },
+    {
+      "epoch": 25.325443786982248,
+      "grad_norm": 3.5392262935638428,
+      "learning_rate": 0.0001,
+      "loss": 0.0491,
+      "step": 4280
+    },
+    {
+      "epoch": 25.384615384615383,
+      "grad_norm": 3.3068196773529053,
+      "learning_rate": 0.0001,
+      "loss": 0.0463,
+      "step": 4290
+    },
+    {
+      "epoch": 25.443786982248522,
+      "grad_norm": 2.740900993347168,
+      "learning_rate": 0.0001,
+      "loss": 0.0465,
+      "step": 4300
+    },
+    {
+      "epoch": 25.502958579881657,
+      "grad_norm": 2.9603111743927,
+      "learning_rate": 0.0001,
+      "loss": 0.0463,
+      "step": 4310
+    },
+    {
+      "epoch": 25.562130177514792,
+      "grad_norm": 3.3971924781799316,
+      "learning_rate": 0.0001,
+      "loss": 0.0465,
+      "step": 4320
+    },
+    {
+      "epoch": 25.62130177514793,
+      "grad_norm": 2.8862221240997314,
+      "learning_rate": 0.0001,
+      "loss": 0.0477,
+      "step": 4330
+    },
+    {
+      "epoch": 25.680473372781066,
+      "grad_norm": 2.9598562717437744,
+      "learning_rate": 0.0001,
+      "loss": 0.047,
+      "step": 4340
+    },
+    {
+      "epoch": 25.7396449704142,
+      "grad_norm": 3.5965535640716553,
+      "learning_rate": 0.0001,
+      "loss": 0.0452,
+      "step": 4350
+    },
+    {
+      "epoch": 25.798816568047336,
+      "grad_norm": 3.184624433517456,
+      "learning_rate": 0.0001,
+      "loss": 0.0474,
+      "step": 4360
+    },
+    {
+      "epoch": 25.857988165680474,
+      "grad_norm": 2.9978439807891846,
+      "learning_rate": 0.0001,
+      "loss": 0.0458,
+      "step": 4370
+    },
+    {
+      "epoch": 25.91715976331361,
+      "grad_norm": 3.5438551902770996,
+      "learning_rate": 0.0001,
+      "loss": 0.0466,
+      "step": 4380
+    },
+    {
+      "epoch": 25.976331360946745,
+      "grad_norm": 3.1312928199768066,
+      "learning_rate": 0.0001,
+      "loss": 0.0482,
+      "step": 4390
+    },
+    {
+      "epoch": 26.035502958579883,
+      "grad_norm": 3.053823947906494,
+      "learning_rate": 0.0001,
+      "loss": 0.0461,
+      "step": 4400
+    },
+    {
+      "epoch": 26.09467455621302,
+      "grad_norm": 3.127730369567871,
+      "learning_rate": 0.0001,
+      "loss": 0.0486,
+      "step": 4410
+    },
+    {
+      "epoch": 26.153846153846153,
+      "grad_norm": 3.188199758529663,
+      "learning_rate": 0.0001,
+      "loss": 0.0461,
+      "step": 4420
+    },
+    {
+      "epoch": 26.21301775147929,
+      "grad_norm": 3.4058473110198975,
+      "learning_rate": 0.0001,
+      "loss": 0.0468,
+      "step": 4430
+    },
+    {
+      "epoch": 26.272189349112427,
+      "grad_norm": 3.788417339324951,
+      "learning_rate": 0.0001,
+      "loss": 0.0499,
+      "step": 4440
+    },
+    {
+      "epoch": 26.331360946745562,
+      "grad_norm": 3.7460741996765137,
+      "learning_rate": 0.0001,
+      "loss": 0.052,
+      "step": 4450
+    },
+    {
+      "epoch": 26.390532544378697,
+      "grad_norm": 3.156660795211792,
+      "learning_rate": 0.0001,
+      "loss": 0.0494,
+      "step": 4460
+    },
+    {
+      "epoch": 26.449704142011836,
+      "grad_norm": 3.4617912769317627,
+      "learning_rate": 0.0001,
+      "loss": 0.0492,
+      "step": 4470
+    },
+    {
+      "epoch": 26.50887573964497,
+      "grad_norm": 2.7324185371398926,
+      "learning_rate": 0.0001,
+      "loss": 0.0475,
+      "step": 4480
+    },
+    {
+      "epoch": 26.568047337278106,
+      "grad_norm": 2.9896185398101807,
+      "learning_rate": 0.0001,
+      "loss": 0.046,
+      "step": 4490
+    },
+    {
+      "epoch": 26.62721893491124,
+      "grad_norm": 3.1339163780212402,
+      "learning_rate": 0.0001,
+      "loss": 0.0466,
+      "step": 4500
+    },
+    {
+      "epoch": 26.68639053254438,
+      "grad_norm": 2.9264252185821533,
+      "learning_rate": 0.0001,
+      "loss": 0.0469,
+      "step": 4510
+    },
+    {
+      "epoch": 26.745562130177515,
+      "grad_norm": 2.5732967853546143,
+      "learning_rate": 0.0001,
+      "loss": 0.0459,
+      "step": 4520
+    },
+    {
+      "epoch": 26.80473372781065,
+      "grad_norm": 2.8136539459228516,
+      "learning_rate": 0.0001,
+      "loss": 0.0443,
+      "step": 4530
+    },
+    {
+      "epoch": 26.86390532544379,
+      "grad_norm": 2.868886947631836,
+      "learning_rate": 0.0001,
+      "loss": 0.0447,
+      "step": 4540
+    },
+    {
+      "epoch": 26.923076923076923,
+      "grad_norm": 3.123053789138794,
+      "learning_rate": 0.0001,
+      "loss": 0.0443,
+      "step": 4550
+    },
+    {
+      "epoch": 26.98224852071006,
+      "grad_norm": 3.6158499717712402,
+      "learning_rate": 0.0001,
+      "loss": 0.0453,
+      "step": 4560
+    },
+    {
+      "epoch": 27.041420118343197,
+      "grad_norm": 2.8775112628936768,
+      "learning_rate": 0.0001,
+      "loss": 0.0482,
+      "step": 4570
+    },
+    {
+      "epoch": 27.100591715976332,
+      "grad_norm": 3.395939350128174,
+      "learning_rate": 0.0001,
+      "loss": 0.0484,
+      "step": 4580
+    },
+    {
+      "epoch": 27.159763313609467,
+      "grad_norm": 3.336581230163574,
+      "learning_rate": 0.0001,
+      "loss": 0.0487,
+      "step": 4590
+    },
+    {
+      "epoch": 27.218934911242602,
+      "grad_norm": 3.2861766815185547,
+      "learning_rate": 0.0001,
+      "loss": 0.0477,
+      "step": 4600
+    },
+    {
+      "epoch": 27.27810650887574,
+      "grad_norm": 3.3157200813293457,
+      "learning_rate": 0.0001,
+      "loss": 0.0498,
+      "step": 4610
+    },
+    {
+      "epoch": 27.337278106508876,
+      "grad_norm": 3.4261231422424316,
+      "learning_rate": 0.0001,
+      "loss": 0.047,
+      "step": 4620
+    },
+    {
+      "epoch": 27.39644970414201,
+      "grad_norm": 3.2643227577209473,
+      "learning_rate": 0.0001,
+      "loss": 0.0491,
+      "step": 4630
+    },
+    {
+      "epoch": 27.45562130177515,
+      "grad_norm": 2.9594922065734863,
+      "learning_rate": 0.0001,
+      "loss": 0.0468,
+      "step": 4640
+    },
+    {
+      "epoch": 27.514792899408285,
+      "grad_norm": 2.9212262630462646,
+      "learning_rate": 0.0001,
+      "loss": 0.0467,
+      "step": 4650
+    },
+    {
+      "epoch": 27.57396449704142,
+      "grad_norm": 2.940035820007324,
+      "learning_rate": 0.0001,
+      "loss": 0.0468,
+      "step": 4660
+    },
+    {
+      "epoch": 27.633136094674555,
+      "grad_norm": 3.4176833629608154,
+      "learning_rate": 0.0001,
+      "loss": 0.0443,
+      "step": 4670
+    },
+    {
+      "epoch": 27.692307692307693,
+      "grad_norm": 3.0427374839782715,
+      "learning_rate": 0.0001,
+      "loss": 0.0448,
+      "step": 4680
+    },
+    {
+      "epoch": 27.75147928994083,
+      "grad_norm": 3.1793110370635986,
+      "learning_rate": 0.0001,
+      "loss": 0.0428,
+      "step": 4690
+    },
+    {
+      "epoch": 27.810650887573964,
+      "grad_norm": 3.4074504375457764,
+      "learning_rate": 0.0001,
+      "loss": 0.0436,
+      "step": 4700
+    },
+    {
+      "epoch": 27.869822485207102,
+      "grad_norm": 3.3215200901031494,
+      "learning_rate": 0.0001,
+      "loss": 0.0425,
+      "step": 4710
+    },
+    {
+      "epoch": 27.928994082840237,
+      "grad_norm": 3.4873764514923096,
+      "learning_rate": 0.0001,
+      "loss": 0.0429,
+      "step": 4720
+    },
+    {
+      "epoch": 27.988165680473372,
+      "grad_norm": 3.3128554821014404,
+      "learning_rate": 0.0001,
+      "loss": 0.043,
+      "step": 4730
+    },
+    {
+      "epoch": 28.047337278106507,
+      "grad_norm": 3.442042589187622,
+      "learning_rate": 0.0001,
+      "loss": 0.0411,
+      "step": 4740
+    },
+    {
+      "epoch": 28.106508875739646,
+      "grad_norm": 3.236996650695801,
+      "learning_rate": 0.0001,
+      "loss": 0.0418,
+      "step": 4750
+    },
+    {
+      "epoch": 28.16568047337278,
+      "grad_norm": 2.7661309242248535,
+      "learning_rate": 0.0001,
+      "loss": 0.042,
+      "step": 4760
+    },
+    {
+      "epoch": 28.224852071005916,
+      "grad_norm": 2.731294870376587,
+      "learning_rate": 0.0001,
+      "loss": 0.0427,
+      "step": 4770
+    },
+    {
+      "epoch": 28.284023668639055,
+      "grad_norm": 2.9517154693603516,
+      "learning_rate": 0.0001,
+      "loss": 0.0412,
+      "step": 4780
+    },
+    {
+      "epoch": 28.34319526627219,
+      "grad_norm": 2.730461359024048,
+      "learning_rate": 0.0001,
+      "loss": 0.0437,
+      "step": 4790
+    },
+    {
+      "epoch": 28.402366863905325,
+      "grad_norm": 2.598295211791992,
+      "learning_rate": 0.0001,
+      "loss": 0.044,
+      "step": 4800
+    },
+    {
+      "epoch": 28.46153846153846,
+      "grad_norm": 2.4978463649749756,
+      "learning_rate": 0.0001,
+      "loss": 0.0426,
+      "step": 4810
+    },
+    {
+      "epoch": 28.5207100591716,
+      "grad_norm": 2.4381778240203857,
+      "learning_rate": 0.0001,
+      "loss": 0.0434,
+      "step": 4820
+    },
+    {
+      "epoch": 28.579881656804734,
+      "grad_norm": 2.8496360778808594,
+      "learning_rate": 0.0001,
+      "loss": 0.0444,
+      "step": 4830
+    },
+    {
+      "epoch": 28.63905325443787,
+      "grad_norm": 2.8395731449127197,
+      "learning_rate": 0.0001,
+      "loss": 0.0445,
+      "step": 4840
+    },
+    {
+      "epoch": 28.698224852071007,
+      "grad_norm": 3.2911291122436523,
+      "learning_rate": 0.0001,
+      "loss": 0.0463,
+      "step": 4850
+    },
+    {
+      "epoch": 28.757396449704142,
+      "grad_norm": 3.9360129833221436,
+      "learning_rate": 0.0001,
+      "loss": 0.0484,
+      "step": 4860
+    },
+    {
+      "epoch": 28.816568047337277,
+      "grad_norm": 2.9939589500427246,
+      "learning_rate": 0.0001,
+      "loss": 0.0435,
+      "step": 4870
+    },
+    {
+      "epoch": 28.875739644970416,
+      "grad_norm": 2.3708269596099854,
+      "learning_rate": 0.0001,
+      "loss": 0.0432,
+      "step": 4880
+    },
+    {
+      "epoch": 28.93491124260355,
+      "grad_norm": 2.766211986541748,
+      "learning_rate": 0.0001,
+      "loss": 0.0448,
+      "step": 4890
+    },
+    {
+      "epoch": 28.994082840236686,
+      "grad_norm": 3.365046262741089,
+      "learning_rate": 0.0001,
+      "loss": 0.0432,
+      "step": 4900
+    },
+    {
+      "epoch": 29.05325443786982,
+      "grad_norm": 2.4397871494293213,
+      "learning_rate": 0.0001,
+      "loss": 0.0443,
+      "step": 4910
+    },
+    {
+      "epoch": 29.11242603550296,
+      "grad_norm": 2.8324618339538574,
+      "learning_rate": 0.0001,
+      "loss": 0.0436,
+      "step": 4920
+    },
+    {
+      "epoch": 29.171597633136095,
+      "grad_norm": 2.846784830093384,
+      "learning_rate": 0.0001,
+      "loss": 0.0436,
+      "step": 4930
+    },
+    {
+      "epoch": 29.23076923076923,
+      "grad_norm": 2.5215604305267334,
+      "learning_rate": 0.0001,
+      "loss": 0.0451,
+      "step": 4940
+    },
+    {
+      "epoch": 29.28994082840237,
+      "grad_norm": 2.446906805038452,
+      "learning_rate": 0.0001,
+      "loss": 0.0439,
+      "step": 4950
+    },
+    {
+      "epoch": 29.349112426035504,
+      "grad_norm": 3.016557455062866,
+      "learning_rate": 0.0001,
+      "loss": 0.0461,
+      "step": 4960
+    },
+    {
+      "epoch": 29.40828402366864,
+      "grad_norm": 3.321094274520874,
+      "learning_rate": 0.0001,
+      "loss": 0.0453,
+      "step": 4970
+    },
+    {
+      "epoch": 29.467455621301774,
+      "grad_norm": 3.3248119354248047,
+      "learning_rate": 0.0001,
+      "loss": 0.0435,
+      "step": 4980
+    },
+    {
+      "epoch": 29.526627218934912,
+      "grad_norm": 2.902519941329956,
+      "learning_rate": 0.0001,
+      "loss": 0.0434,
+      "step": 4990
+    },
+    {
+      "epoch": 29.585798816568047,
+      "grad_norm": 2.787186622619629,
+      "learning_rate": 0.0001,
+      "loss": 0.0419,
+      "step": 5000
+    },
+    {
+      "epoch": 29.644970414201183,
+      "grad_norm": 3.0807485580444336,
+      "learning_rate": 0.0001,
+      "loss": 0.0421,
+      "step": 5010
+    },
+    {
+      "epoch": 29.70414201183432,
+      "grad_norm": 2.63411283493042,
+      "learning_rate": 0.0001,
+      "loss": 0.0436,
+      "step": 5020
+    },
+    {
+      "epoch": 29.763313609467456,
+      "grad_norm": 2.9151620864868164,
+      "learning_rate": 0.0001,
+      "loss": 0.0421,
+      "step": 5030
+    },
+    {
+      "epoch": 29.82248520710059,
+      "grad_norm": 2.728679895401001,
+      "learning_rate": 0.0001,
+      "loss": 0.0431,
+      "step": 5040
+    },
+    {
+      "epoch": 29.881656804733726,
+      "grad_norm": 2.570870876312256,
+      "learning_rate": 0.0001,
+      "loss": 0.0447,
+      "step": 5050
+    },
+    {
+      "epoch": 29.940828402366865,
+      "grad_norm": 2.6758365631103516,
+      "learning_rate": 0.0001,
+      "loss": 0.0415,
+      "step": 5060
+    },
+    {
+      "epoch": 30.0,
+      "grad_norm": 2.620326280593872,
+      "learning_rate": 0.0001,
+      "loss": 0.0426,
+      "step": 5070
+    },
+    {
+      "epoch": 30.059171597633135,
+      "grad_norm": 2.598512887954712,
+      "learning_rate": 0.0001,
+      "loss": 0.0426,
+      "step": 5080
+    },
+    {
+      "epoch": 30.118343195266274,
+      "grad_norm": 2.908621072769165,
+      "learning_rate": 0.0001,
+      "loss": 0.0412,
+      "step": 5090
+    },
+    {
+      "epoch": 30.17751479289941,
+      "grad_norm": 2.841421604156494,
+      "learning_rate": 0.0001,
+      "loss": 0.0431,
+      "step": 5100
+    },
+    {
+      "epoch": 30.236686390532544,
+      "grad_norm": 2.5824601650238037,
+      "learning_rate": 0.0001,
+      "loss": 0.0434,
+      "step": 5110
+    },
+    {
+      "epoch": 30.29585798816568,
+      "grad_norm": 2.8675408363342285,
+      "learning_rate": 0.0001,
+      "loss": 0.0435,
+      "step": 5120
+    },
+    {
+      "epoch": 30.355029585798817,
+      "grad_norm": 2.213034152984619,
+      "learning_rate": 0.0001,
+      "loss": 0.0434,
+      "step": 5130
+    },
+    {
+      "epoch": 30.414201183431953,
+      "grad_norm": 3.357454776763916,
+      "learning_rate": 0.0001,
+      "loss": 0.0436,
+      "step": 5140
+    },
+    {
+      "epoch": 30.473372781065088,
+      "grad_norm": 3.113046169281006,
+      "learning_rate": 0.0001,
+      "loss": 0.0473,
+      "step": 5150
+    },
+    {
+      "epoch": 30.532544378698226,
+      "grad_norm": 2.682926654815674,
+      "learning_rate": 0.0001,
+      "loss": 0.0445,
+      "step": 5160
+    },
+    {
+      "epoch": 30.59171597633136,
+      "grad_norm": 2.8853464126586914,
+      "learning_rate": 0.0001,
+      "loss": 0.0427,
+      "step": 5170
+    },
+    {
+      "epoch": 30.650887573964496,
+      "grad_norm": 3.0199553966522217,
+      "learning_rate": 0.0001,
+      "loss": 0.0421,
+      "step": 5180
+    },
+    {
+      "epoch": 30.71005917159763,
+      "grad_norm": 3.734431743621826,
+      "learning_rate": 0.0001,
+      "loss": 0.0438,
+      "step": 5190
+    },
+    {
+      "epoch": 30.76923076923077,
+      "grad_norm": 3.1161088943481445,
+      "learning_rate": 0.0001,
+      "loss": 0.0408,
+      "step": 5200
+    },
+    {
+      "epoch": 30.828402366863905,
+      "grad_norm": 2.9835052490234375,
+      "learning_rate": 0.0001,
+      "loss": 0.0412,
+      "step": 5210
+    },
+    {
+      "epoch": 30.88757396449704,
+      "grad_norm": 2.7090277671813965,
+      "learning_rate": 0.0001,
+      "loss": 0.0406,
+      "step": 5220
+    },
+    {
+      "epoch": 30.94674556213018,
+      "grad_norm": 2.43406343460083,
+      "learning_rate": 0.0001,
+      "loss": 0.0393,
+      "step": 5230
+    },
+    {
+      "epoch": 31.005917159763314,
+      "grad_norm": 2.120563268661499,
+      "learning_rate": 0.0001,
+      "loss": 0.0421,
+      "step": 5240
+    },
+    {
+      "epoch": 31.06508875739645,
+      "grad_norm": 2.9324002265930176,
+      "learning_rate": 0.0001,
+      "loss": 0.0421,
+      "step": 5250
+    },
+    {
+      "epoch": 31.124260355029588,
+      "grad_norm": 2.5056211948394775,
+      "learning_rate": 0.0001,
+      "loss": 0.0422,
+      "step": 5260
+    },
+    {
+      "epoch": 31.183431952662723,
+      "grad_norm": 2.4987547397613525,
+      "learning_rate": 0.0001,
+      "loss": 0.0417,
+      "step": 5270
+    },
+    {
+      "epoch": 31.242603550295858,
+      "grad_norm": 2.581291913986206,
+      "learning_rate": 0.0001,
+      "loss": 0.0404,
+      "step": 5280
+    },
+    {
+      "epoch": 31.301775147928993,
+      "grad_norm": 2.3173515796661377,
+      "learning_rate": 0.0001,
+      "loss": 0.0422,
+      "step": 5290
+    },
+    {
+      "epoch": 31.36094674556213,
+      "grad_norm": 2.9121341705322266,
+      "learning_rate": 0.0001,
+      "loss": 0.0418,
+      "step": 5300
+    },
+    {
+      "epoch": 31.420118343195266,
+      "grad_norm": 2.373626470565796,
+      "learning_rate": 0.0001,
+      "loss": 0.0415,
+      "step": 5310
+    },
+    {
+      "epoch": 31.4792899408284,
+      "grad_norm": 2.786344051361084,
+      "learning_rate": 0.0001,
+      "loss": 0.0415,
+      "step": 5320
+    },
+    {
+      "epoch": 31.53846153846154,
+      "grad_norm": 2.3212780952453613,
+      "learning_rate": 0.0001,
+      "loss": 0.0417,
+      "step": 5330
+    },
+    {
+      "epoch": 31.597633136094675,
+      "grad_norm": 3.000011444091797,
+      "learning_rate": 0.0001,
+      "loss": 0.0414,
+      "step": 5340
+    },
+    {
+      "epoch": 31.65680473372781,
+      "grad_norm": 2.939387321472168,
+      "learning_rate": 0.0001,
+      "loss": 0.0427,
+      "step": 5350
+    },
+    {
+      "epoch": 31.715976331360945,
+      "grad_norm": 2.7995383739471436,
+      "learning_rate": 0.0001,
+      "loss": 0.0443,
+      "step": 5360
+    },
+    {
+      "epoch": 31.775147928994084,
+      "grad_norm": 2.3119311332702637,
+      "learning_rate": 0.0001,
+      "loss": 0.0415,
+      "step": 5370
+    },
+    {
+      "epoch": 31.83431952662722,
+      "grad_norm": 2.3990094661712646,
+      "learning_rate": 0.0001,
+      "loss": 0.0407,
+      "step": 5380
+    },
+    {
+      "epoch": 31.893491124260354,
+      "grad_norm": 2.8436527252197266,
+      "learning_rate": 0.0001,
+      "loss": 0.0399,
+      "step": 5390
+    },
+    {
+      "epoch": 31.952662721893493,
+      "grad_norm": 2.735164165496826,
+      "learning_rate": 0.0001,
+      "loss": 0.0412,
+      "step": 5400
+    },
+    {
+      "epoch": 32.01183431952663,
+      "grad_norm": 2.670602798461914,
+      "learning_rate": 0.0001,
+      "loss": 0.0392,
+      "step": 5410
+    },
+    {
+      "epoch": 32.071005917159766,
+      "grad_norm": 2.7820117473602295,
+      "learning_rate": 0.0001,
+      "loss": 0.0411,
+      "step": 5420
+    },
+    {
+      "epoch": 32.1301775147929,
+      "grad_norm": 2.9033374786376953,
+      "learning_rate": 0.0001,
+      "loss": 0.04,
+      "step": 5430
+    },
+    {
+      "epoch": 32.18934911242604,
+      "grad_norm": 2.726259469985962,
+      "learning_rate": 0.0001,
+      "loss": 0.0398,
+      "step": 5440
+    },
+    {
+      "epoch": 32.248520710059175,
+      "grad_norm": 3.0624237060546875,
+      "learning_rate": 0.0001,
+      "loss": 0.0393,
+      "step": 5450
+    },
+    {
+      "epoch": 32.30769230769231,
+      "grad_norm": 2.4605252742767334,
+      "learning_rate": 0.0001,
+      "loss": 0.0399,
+      "step": 5460
+    },
+    {
+      "epoch": 32.366863905325445,
+      "grad_norm": 3.3417742252349854,
+      "learning_rate": 0.0001,
+      "loss": 0.041,
+      "step": 5470
+    },
+    {
+      "epoch": 32.42603550295858,
+      "grad_norm": 2.160215139389038,
+      "learning_rate": 0.0001,
+      "loss": 0.0398,
+      "step": 5480
+    },
+    {
+      "epoch": 32.485207100591715,
+      "grad_norm": 2.390230655670166,
+      "learning_rate": 0.0001,
+      "loss": 0.0405,
+      "step": 5490
+    },
+    {
+      "epoch": 32.544378698224854,
+      "grad_norm": 2.6840054988861084,
+      "learning_rate": 0.0001,
+      "loss": 0.0401,
+      "step": 5500
+    },
+    {
+      "epoch": 32.603550295857985,
+      "grad_norm": 2.510371208190918,
+      "learning_rate": 0.0001,
+      "loss": 0.0401,
+      "step": 5510
+    },
+    {
+      "epoch": 32.662721893491124,
+      "grad_norm": 2.244293212890625,
+      "learning_rate": 0.0001,
+      "loss": 0.0395,
+      "step": 5520
+    },
+    {
+      "epoch": 32.72189349112426,
+      "grad_norm": 2.3354482650756836,
+      "learning_rate": 0.0001,
+      "loss": 0.0402,
+      "step": 5530
+    },
+    {
+      "epoch": 32.781065088757394,
+      "grad_norm": 2.974591016769409,
+      "learning_rate": 0.0001,
+      "loss": 0.0407,
+      "step": 5540
+    },
+    {
+      "epoch": 32.84023668639053,
+      "grad_norm": 2.2013814449310303,
+      "learning_rate": 0.0001,
+      "loss": 0.0413,
+      "step": 5550
+    },
+    {
+      "epoch": 32.89940828402367,
+      "grad_norm": 2.4692649841308594,
+      "learning_rate": 0.0001,
+      "loss": 0.0394,
+      "step": 5560
+    },
+    {
+      "epoch": 32.9585798816568,
+      "grad_norm": 2.59855580329895,
+      "learning_rate": 0.0001,
+      "loss": 0.0411,
+      "step": 5570
+    },
+    {
+      "epoch": 33.01775147928994,
+      "grad_norm": 3.0601751804351807,
+      "learning_rate": 0.0001,
+      "loss": 0.041,
+      "step": 5580
+    },
+    {
+      "epoch": 33.07692307692308,
+      "grad_norm": 2.8953089714050293,
+      "learning_rate": 0.0001,
+      "loss": 0.0429,
+      "step": 5590
+    },
+    {
+      "epoch": 33.13609467455621,
+      "grad_norm": 2.76509428024292,
+      "learning_rate": 0.0001,
+      "loss": 0.0424,
+      "step": 5600
+    },
+    {
+      "epoch": 33.19526627218935,
+      "grad_norm": 2.8845596313476562,
+      "learning_rate": 0.0001,
+      "loss": 0.0431,
+      "step": 5610
+    },
+    {
+      "epoch": 33.25443786982248,
+      "grad_norm": 2.391890048980713,
+      "learning_rate": 0.0001,
+      "loss": 0.0434,
+      "step": 5620
+    },
+    {
+      "epoch": 33.31360946745562,
+      "grad_norm": 2.6841378211975098,
+      "learning_rate": 0.0001,
+      "loss": 0.0418,
+      "step": 5630
+    },
+    {
+      "epoch": 33.37278106508876,
+      "grad_norm": 2.655158042907715,
+      "learning_rate": 0.0001,
+      "loss": 0.0412,
+      "step": 5640
+    },
+    {
+      "epoch": 33.43195266272189,
+      "grad_norm": 2.6897428035736084,
+      "learning_rate": 0.0001,
+      "loss": 0.0413,
+      "step": 5650
+    },
+    {
+      "epoch": 33.49112426035503,
+      "grad_norm": 2.4828686714172363,
+      "learning_rate": 0.0001,
+      "loss": 0.0417,
+      "step": 5660
+    },
+    {
+      "epoch": 33.55029585798817,
+      "grad_norm": 2.096879005432129,
+      "learning_rate": 0.0001,
+      "loss": 0.0402,
+      "step": 5670
+    },
+    {
+      "epoch": 33.6094674556213,
+      "grad_norm": 2.2518105506896973,
+      "learning_rate": 0.0001,
+      "loss": 0.0403,
+      "step": 5680
+    },
+    {
+      "epoch": 33.66863905325444,
+      "grad_norm": 2.8636226654052734,
+      "learning_rate": 0.0001,
+      "loss": 0.0398,
+      "step": 5690
+    },
+    {
+      "epoch": 33.72781065088758,
+      "grad_norm": 2.4722647666931152,
+      "learning_rate": 0.0001,
+      "loss": 0.0401,
+      "step": 5700
+    },
+    {
+      "epoch": 33.78698224852071,
+      "grad_norm": 2.5116543769836426,
+      "learning_rate": 0.0001,
+      "loss": 0.0392,
+      "step": 5710
+    },
+    {
+      "epoch": 33.84615384615385,
+      "grad_norm": 2.650794506072998,
+      "learning_rate": 0.0001,
+      "loss": 0.041,
+      "step": 5720
+    },
+    {
+      "epoch": 33.905325443786985,
+      "grad_norm": 2.461094379425049,
+      "learning_rate": 0.0001,
+      "loss": 0.04,
+      "step": 5730
+    },
+    {
+      "epoch": 33.96449704142012,
+      "grad_norm": 2.849574327468872,
+      "learning_rate": 0.0001,
+      "loss": 0.0412,
+      "step": 5740
+    },
+    {
+      "epoch": 34.023668639053255,
+      "grad_norm": 2.9352269172668457,
+      "learning_rate": 0.0001,
+      "loss": 0.0382,
+      "step": 5750
+    },
+    {
+      "epoch": 34.082840236686394,
+      "grad_norm": 3.3457493782043457,
+      "learning_rate": 0.0001,
+      "loss": 0.0394,
+      "step": 5760
+    },
+    {
+      "epoch": 34.142011834319526,
+      "grad_norm": 3.5685365200042725,
+      "learning_rate": 0.0001,
+      "loss": 0.039,
+      "step": 5770
+    },
+    {
+      "epoch": 34.201183431952664,
+      "grad_norm": 2.4509737491607666,
+      "learning_rate": 0.0001,
+      "loss": 0.0362,
+      "step": 5780
+    },
+    {
+      "epoch": 34.260355029585796,
+      "grad_norm": 2.49623966217041,
+      "learning_rate": 0.0001,
+      "loss": 0.0375,
+      "step": 5790
+    },
+    {
+      "epoch": 34.319526627218934,
+      "grad_norm": 2.8502838611602783,
+      "learning_rate": 0.0001,
+      "loss": 0.0384,
+      "step": 5800
+    },
+    {
+      "epoch": 34.37869822485207,
+      "grad_norm": 3.2586593627929688,
+      "learning_rate": 0.0001,
+      "loss": 0.0384,
+      "step": 5810
+    },
+    {
+      "epoch": 34.437869822485204,
+      "grad_norm": 3.1439669132232666,
+      "learning_rate": 0.0001,
+      "loss": 0.0389,
+      "step": 5820
+    },
+    {
+      "epoch": 34.49704142011834,
+      "grad_norm": 2.6732826232910156,
+      "learning_rate": 0.0001,
+      "loss": 0.0389,
+      "step": 5830
+    },
+    {
+      "epoch": 34.55621301775148,
+      "grad_norm": 2.663051128387451,
+      "learning_rate": 0.0001,
+      "loss": 0.039,
+      "step": 5840
+    },
+    {
+      "epoch": 34.61538461538461,
+      "grad_norm": 2.174128293991089,
+      "learning_rate": 0.0001,
+      "loss": 0.0378,
+      "step": 5850
+    },
+    {
+      "epoch": 34.67455621301775,
+      "grad_norm": 2.4969770908355713,
+      "learning_rate": 0.0001,
+      "loss": 0.0371,
+      "step": 5860
+    },
+    {
+      "epoch": 34.73372781065089,
+      "grad_norm": 2.620683431625366,
+      "learning_rate": 0.0001,
+      "loss": 0.0376,
+      "step": 5870
+    },
+    {
+      "epoch": 34.79289940828402,
+      "grad_norm": 2.3133041858673096,
+      "learning_rate": 0.0001,
+      "loss": 0.0377,
+      "step": 5880
+    },
+    {
+      "epoch": 34.85207100591716,
+      "grad_norm": 2.375354528427124,
+      "learning_rate": 0.0001,
+      "loss": 0.0385,
+      "step": 5890
+    },
+    {
+      "epoch": 34.9112426035503,
+      "grad_norm": 2.96539568901062,
+      "learning_rate": 0.0001,
+      "loss": 0.0394,
+      "step": 5900
+    },
+    {
+      "epoch": 34.97041420118343,
+      "grad_norm": 2.421586036682129,
+      "learning_rate": 0.0001,
+      "loss": 0.0393,
+      "step": 5910
+    },
+    {
+      "epoch": 35.02958579881657,
+      "grad_norm": 2.7406249046325684,
+      "learning_rate": 0.0001,
+      "loss": 0.0374,
+      "step": 5920
+    },
+    {
+      "epoch": 35.0887573964497,
+      "grad_norm": 2.2939305305480957,
+      "learning_rate": 0.0001,
+      "loss": 0.0382,
+      "step": 5930
+    },
+    {
+      "epoch": 35.14792899408284,
+      "grad_norm": 2.5900344848632812,
+      "learning_rate": 0.0001,
+      "loss": 0.0386,
+      "step": 5940
+    },
+    {
+      "epoch": 35.20710059171598,
+      "grad_norm": 2.5751914978027344,
+      "learning_rate": 0.0001,
+      "loss": 0.0388,
+      "step": 5950
+    },
+    {
+      "epoch": 35.26627218934911,
+      "grad_norm": 2.7437937259674072,
+      "learning_rate": 0.0001,
+      "loss": 0.0393,
+      "step": 5960
+    },
+    {
+      "epoch": 35.32544378698225,
+      "grad_norm": 2.5376670360565186,
+      "learning_rate": 0.0001,
+      "loss": 0.0392,
+      "step": 5970
+    },
+    {
+      "epoch": 35.38461538461539,
+      "grad_norm": 2.3735177516937256,
+      "learning_rate": 0.0001,
+      "loss": 0.0394,
+      "step": 5980
+    },
+    {
+      "epoch": 35.44378698224852,
+      "grad_norm": 3.5083789825439453,
+      "learning_rate": 0.0001,
+      "loss": 0.0402,
+      "step": 5990
+    },
+    {
+      "epoch": 35.50295857988166,
+      "grad_norm": 2.499983549118042,
+      "learning_rate": 0.0001,
+      "loss": 0.0428,
+      "step": 6000
+    },
+    {
+      "epoch": 35.562130177514796,
+      "grad_norm": 2.214320182800293,
+      "learning_rate": 0.0001,
+      "loss": 0.0416,
+      "step": 6010
+    },
+    {
+      "epoch": 35.62130177514793,
+      "grad_norm": 2.7791621685028076,
+      "learning_rate": 0.0001,
+      "loss": 0.042,
+      "step": 6020
+    },
+    {
+      "epoch": 35.680473372781066,
+      "grad_norm": 2.353794813156128,
+      "learning_rate": 0.0001,
+      "loss": 0.0411,
+      "step": 6030
+    },
+    {
+      "epoch": 35.739644970414204,
+      "grad_norm": 2.4961941242218018,
+      "learning_rate": 0.0001,
+      "loss": 0.0402,
+      "step": 6040
+    },
+    {
+      "epoch": 35.798816568047336,
+      "grad_norm": 2.641179323196411,
+      "learning_rate": 0.0001,
+      "loss": 0.0402,
+      "step": 6050
+    },
+    {
+      "epoch": 35.857988165680474,
+      "grad_norm": 1.8833891153335571,
+      "learning_rate": 0.0001,
+      "loss": 0.0401,
+      "step": 6060
+    },
+    {
+      "epoch": 35.917159763313606,
+      "grad_norm": 2.4077365398406982,
+      "learning_rate": 0.0001,
+      "loss": 0.0387,
+      "step": 6070
+    },
+    {
+      "epoch": 35.976331360946745,
+      "grad_norm": 2.20727276802063,
+      "learning_rate": 0.0001,
+      "loss": 0.04,
+      "step": 6080
+    },
+    {
+      "epoch": 36.03550295857988,
+      "grad_norm": 2.805233955383301,
+      "learning_rate": 0.0001,
+      "loss": 0.0403,
+      "step": 6090
+    },
+    {
+      "epoch": 36.094674556213015,
+      "grad_norm": 1.8451623916625977,
+      "learning_rate": 0.0001,
+      "loss": 0.0405,
+      "step": 6100
+    },
+    {
+      "epoch": 36.15384615384615,
+      "grad_norm": 2.280503273010254,
+      "learning_rate": 0.0001,
+      "loss": 0.0394,
+      "step": 6110
+    },
+    {
+      "epoch": 36.21301775147929,
+      "grad_norm": 2.8806192874908447,
+      "learning_rate": 0.0001,
+      "loss": 0.0396,
+      "step": 6120
+    },
+    {
+      "epoch": 36.27218934911242,
+      "grad_norm": 2.749778985977173,
+      "learning_rate": 0.0001,
+      "loss": 0.0395,
+      "step": 6130
+    },
+    {
+      "epoch": 36.33136094674556,
+      "grad_norm": 1.9586924314498901,
+      "learning_rate": 0.0001,
+      "loss": 0.0376,
+      "step": 6140
+    },
+    {
+      "epoch": 36.3905325443787,
+      "grad_norm": 2.3840696811676025,
+      "learning_rate": 0.0001,
+      "loss": 0.037,
+      "step": 6150
+    },
+    {
+      "epoch": 36.44970414201183,
+      "grad_norm": 2.3572278022766113,
+      "learning_rate": 0.0001,
+      "loss": 0.037,
+      "step": 6160
+    },
+    {
+      "epoch": 36.50887573964497,
+      "grad_norm": 2.1688976287841797,
+      "learning_rate": 0.0001,
+      "loss": 0.0372,
+      "step": 6170
+    },
+    {
+      "epoch": 36.56804733727811,
+      "grad_norm": 1.8140296936035156,
+      "learning_rate": 0.0001,
+      "loss": 0.0381,
+      "step": 6180
+    },
+    {
+      "epoch": 36.62721893491124,
+      "grad_norm": 1.9743294715881348,
+      "learning_rate": 0.0001,
+      "loss": 0.0366,
+      "step": 6190
+    },
+    {
+      "epoch": 36.68639053254438,
+      "grad_norm": 2.789410352706909,
+      "learning_rate": 0.0001,
+      "loss": 0.0374,
+      "step": 6200
+    },
+    {
+      "epoch": 36.74556213017752,
+      "grad_norm": 2.2342898845672607,
+      "learning_rate": 0.0001,
+      "loss": 0.0388,
+      "step": 6210
+    },
+    {
+      "epoch": 36.80473372781065,
+      "grad_norm": 2.190018653869629,
+      "learning_rate": 0.0001,
+      "loss": 0.0376,
+      "step": 6220
+    },
+    {
+      "epoch": 36.86390532544379,
+      "grad_norm": 2.492464065551758,
+      "learning_rate": 0.0001,
+      "loss": 0.0396,
+      "step": 6230
+    },
+    {
+      "epoch": 36.92307692307692,
+      "grad_norm": 2.3696422576904297,
+      "learning_rate": 0.0001,
+      "loss": 0.0384,
+      "step": 6240
+    },
+    {
+      "epoch": 36.98224852071006,
+      "grad_norm": 2.038090467453003,
+      "learning_rate": 0.0001,
+      "loss": 0.0382,
+      "step": 6250
+    },
+    {
+      "epoch": 37.0414201183432,
+      "grad_norm": 2.501413345336914,
+      "learning_rate": 0.0001,
+      "loss": 0.0388,
+      "step": 6260
+    },
+    {
+      "epoch": 37.10059171597633,
+      "grad_norm": 2.343127727508545,
+      "learning_rate": 0.0001,
+      "loss": 0.0386,
+      "step": 6270
+    },
+    {
+      "epoch": 37.15976331360947,
+      "grad_norm": 2.752690315246582,
+      "learning_rate": 0.0001,
+      "loss": 0.0378,
+      "step": 6280
+    },
+    {
+      "epoch": 37.218934911242606,
+      "grad_norm": 2.4334535598754883,
+      "learning_rate": 0.0001,
+      "loss": 0.0384,
+      "step": 6290
+    },
+    {
+      "epoch": 37.27810650887574,
+      "grad_norm": 2.3313214778900146,
+      "learning_rate": 0.0001,
+      "loss": 0.0377,
+      "step": 6300
+    },
+    {
+      "epoch": 37.337278106508876,
+      "grad_norm": 2.554887294769287,
+      "learning_rate": 0.0001,
+      "loss": 0.0376,
+      "step": 6310
+    },
+    {
+      "epoch": 37.396449704142015,
+      "grad_norm": 2.1963515281677246,
+      "learning_rate": 0.0001,
+      "loss": 0.0369,
+      "step": 6320
+    },
+    {
+      "epoch": 37.455621301775146,
+      "grad_norm": 2.4000205993652344,
+      "learning_rate": 0.0001,
+      "loss": 0.0366,
+      "step": 6330
+    },
+    {
+      "epoch": 37.514792899408285,
+      "grad_norm": 2.5008344650268555,
+      "learning_rate": 0.0001,
+      "loss": 0.0359,
+      "step": 6340
+    },
+    {
+      "epoch": 37.57396449704142,
+      "grad_norm": 2.2653043270111084,
+      "learning_rate": 0.0001,
+      "loss": 0.0364,
+      "step": 6350
+    },
+    {
+      "epoch": 37.633136094674555,
+      "grad_norm": 2.347411870956421,
+      "learning_rate": 0.0001,
+      "loss": 0.0373,
+      "step": 6360
+    },
+    {
+      "epoch": 37.69230769230769,
+      "grad_norm": 2.5278966426849365,
+      "learning_rate": 0.0001,
+      "loss": 0.0374,
+      "step": 6370
+    },
+    {
+      "epoch": 37.751479289940825,
+      "grad_norm": 2.3004722595214844,
+      "learning_rate": 0.0001,
+      "loss": 0.039,
+      "step": 6380
+    },
+    {
+      "epoch": 37.81065088757396,
+      "grad_norm": 2.4736993312835693,
+      "learning_rate": 0.0001,
+      "loss": 0.037,
+      "step": 6390
+    },
+    {
+      "epoch": 37.8698224852071,
+      "grad_norm": 2.2091798782348633,
+      "learning_rate": 0.0001,
+      "loss": 0.0355,
+      "step": 6400
+    },
+    {
+      "epoch": 37.928994082840234,
+      "grad_norm": 2.859865665435791,
+      "learning_rate": 0.0001,
+      "loss": 0.0374,
+      "step": 6410
+    },
+    {
+      "epoch": 37.98816568047337,
+      "grad_norm": 2.0931196212768555,
+      "learning_rate": 0.0001,
+      "loss": 0.0396,
+      "step": 6420
+    },
+    {
+      "epoch": 38.04733727810651,
+      "grad_norm": 2.1994707584381104,
+      "learning_rate": 0.0001,
+      "loss": 0.0369,
+      "step": 6430
+    },
+    {
+      "epoch": 38.10650887573964,
+      "grad_norm": 2.5779335498809814,
+      "learning_rate": 0.0001,
+      "loss": 0.0376,
+      "step": 6440
+    },
+    {
+      "epoch": 38.16568047337278,
+      "grad_norm": 2.168508768081665,
+      "learning_rate": 0.0001,
+      "loss": 0.0365,
+      "step": 6450
+    },
+    {
+      "epoch": 38.22485207100592,
+      "grad_norm": 2.1566107273101807,
+      "learning_rate": 0.0001,
+      "loss": 0.0376,
+      "step": 6460
+    },
+    {
+      "epoch": 38.28402366863905,
+      "grad_norm": 2.031437397003174,
+      "learning_rate": 0.0001,
+      "loss": 0.0369,
+      "step": 6470
+    },
+    {
+      "epoch": 38.34319526627219,
+      "grad_norm": 2.342960834503174,
+      "learning_rate": 0.0001,
+      "loss": 0.0386,
+      "step": 6480
+    },
+    {
+      "epoch": 38.40236686390533,
+      "grad_norm": 1.9592527151107788,
+      "learning_rate": 0.0001,
+      "loss": 0.0386,
+      "step": 6490
+    },
+    {
+      "epoch": 38.46153846153846,
+      "grad_norm": 2.1036033630371094,
+      "learning_rate": 0.0001,
+      "loss": 0.0374,
+      "step": 6500
+    },
+    {
+      "epoch": 38.5207100591716,
+      "grad_norm": 2.107426404953003,
+      "learning_rate": 0.0001,
+      "loss": 0.0369,
+      "step": 6510
+    },
+    {
+      "epoch": 38.57988165680474,
+      "grad_norm": 2.5449817180633545,
+      "learning_rate": 0.0001,
+      "loss": 0.0369,
+      "step": 6520
+    },
+    {
+      "epoch": 38.63905325443787,
+      "grad_norm": 2.0007851123809814,
+      "learning_rate": 0.0001,
+      "loss": 0.0386,
+      "step": 6530
+    },
+    {
+      "epoch": 38.69822485207101,
+      "grad_norm": 2.2029869556427,
+      "learning_rate": 0.0001,
+      "loss": 0.0365,
+      "step": 6540
+    },
+    {
+      "epoch": 38.75739644970414,
+      "grad_norm": 2.2885212898254395,
+      "learning_rate": 0.0001,
+      "loss": 0.0365,
+      "step": 6550
+    },
+    {
+      "epoch": 38.81656804733728,
+      "grad_norm": 3.17850399017334,
+      "learning_rate": 0.0001,
+      "loss": 0.0375,
+      "step": 6560
+    },
+    {
+      "epoch": 38.875739644970416,
+      "grad_norm": 2.116330146789551,
+      "learning_rate": 0.0001,
+      "loss": 0.0376,
+      "step": 6570
+    },
+    {
+      "epoch": 38.93491124260355,
+      "grad_norm": 2.061427116394043,
+      "learning_rate": 0.0001,
+      "loss": 0.0365,
+      "step": 6580
+    },
+    {
+      "epoch": 38.994082840236686,
+      "grad_norm": 2.645265817642212,
+      "learning_rate": 0.0001,
+      "loss": 0.0383,
+      "step": 6590
+    },
+    {
+      "epoch": 39.053254437869825,
+      "grad_norm": 1.9446637630462646,
+      "learning_rate": 0.0001,
+      "loss": 0.037,
+      "step": 6600
+    },
+    {
+      "epoch": 39.112426035502956,
+      "grad_norm": 2.2229831218719482,
+      "learning_rate": 0.0001,
+      "loss": 0.0371,
+      "step": 6610
+    },
+    {
+      "epoch": 39.171597633136095,
+      "grad_norm": 1.785588264465332,
+      "learning_rate": 0.0001,
+      "loss": 0.0378,
+      "step": 6620
+    },
+    {
+      "epoch": 39.23076923076923,
+      "grad_norm": 2.5389487743377686,
+      "learning_rate": 0.0001,
+      "loss": 0.0377,
+      "step": 6630
+    },
+    {
+      "epoch": 39.289940828402365,
+      "grad_norm": 2.5257928371429443,
+      "learning_rate": 0.0001,
+      "loss": 0.0353,
+      "step": 6640
+    },
+    {
+      "epoch": 39.349112426035504,
+      "grad_norm": 2.3250818252563477,
+      "learning_rate": 0.0001,
+      "loss": 0.037,
+      "step": 6650
+    },
+    {
+      "epoch": 39.40828402366864,
+      "grad_norm": 2.37992000579834,
+      "learning_rate": 0.0001,
+      "loss": 0.0368,
+      "step": 6660
+    },
+    {
+      "epoch": 39.467455621301774,
+      "grad_norm": 2.038193702697754,
+      "learning_rate": 0.0001,
+      "loss": 0.0354,
+      "step": 6670
+    },
+    {
+      "epoch": 39.52662721893491,
+      "grad_norm": 2.3759708404541016,
+      "learning_rate": 0.0001,
+      "loss": 0.0361,
+      "step": 6680
+    },
+    {
+      "epoch": 39.585798816568044,
+      "grad_norm": 2.296311140060425,
+      "learning_rate": 0.0001,
+      "loss": 0.0364,
+      "step": 6690
+    },
+    {
+      "epoch": 39.64497041420118,
+      "grad_norm": 2.9102301597595215,
+      "learning_rate": 0.0001,
+      "loss": 0.0363,
+      "step": 6700
+    },
+    {
+      "epoch": 39.70414201183432,
+      "grad_norm": 2.576260566711426,
+      "learning_rate": 0.0001,
+      "loss": 0.0381,
+      "step": 6710
+    },
+    {
+      "epoch": 39.76331360946745,
+      "grad_norm": 2.1134250164031982,
+      "learning_rate": 0.0001,
+      "loss": 0.0377,
+      "step": 6720
+    },
+    {
+      "epoch": 39.82248520710059,
+      "grad_norm": 1.9798874855041504,
+      "learning_rate": 0.0001,
+      "loss": 0.0382,
+      "step": 6730
+    },
+    {
+      "epoch": 39.88165680473373,
+      "grad_norm": 2.2950212955474854,
+      "learning_rate": 0.0001,
+      "loss": 0.0365,
+      "step": 6740
+    },
+    {
+      "epoch": 39.94082840236686,
+      "grad_norm": 2.6816320419311523,
+      "learning_rate": 0.0001,
+      "loss": 0.0361,
+      "step": 6750
+    },
+    {
+      "epoch": 40.0,
+      "grad_norm": 2.4200894832611084,
+      "learning_rate": 0.0001,
+      "loss": 0.0364,
+      "step": 6760
+    },
+    {
+      "epoch": 40.05917159763314,
+      "grad_norm": 2.193113088607788,
+      "learning_rate": 0.0001,
+      "loss": 0.0353,
+      "step": 6770
+    },
+    {
+      "epoch": 40.11834319526627,
+      "grad_norm": 2.565507650375366,
+      "learning_rate": 0.0001,
+      "loss": 0.0366,
+      "step": 6780
+    },
+    {
+      "epoch": 40.17751479289941,
+      "grad_norm": 2.2668135166168213,
+      "learning_rate": 0.0001,
+      "loss": 0.0361,
+      "step": 6790
+    },
+    {
+      "epoch": 40.23668639053255,
+      "grad_norm": 1.708004117012024,
+      "learning_rate": 0.0001,
+      "loss": 0.0355,
+      "step": 6800
+    },
+    {
+      "epoch": 40.29585798816568,
+      "grad_norm": 1.8339115381240845,
+      "learning_rate": 0.0001,
+      "loss": 0.0359,
+      "step": 6810
+    },
+    {
+      "epoch": 40.35502958579882,
+      "grad_norm": 2.0544984340667725,
+      "learning_rate": 0.0001,
+      "loss": 0.0366,
+      "step": 6820
+    },
+    {
+      "epoch": 40.414201183431956,
+      "grad_norm": 1.8792948722839355,
+      "learning_rate": 0.0001,
+      "loss": 0.0356,
+      "step": 6830
+    },
+    {
+      "epoch": 40.47337278106509,
+      "grad_norm": 2.1796622276306152,
+      "learning_rate": 0.0001,
+      "loss": 0.035,
+      "step": 6840
+    },
+    {
+      "epoch": 40.532544378698226,
+      "grad_norm": 2.5065059661865234,
+      "learning_rate": 0.0001,
+      "loss": 0.0346,
+      "step": 6850
+    },
+    {
+      "epoch": 40.59171597633136,
+      "grad_norm": 2.312072992324829,
+      "learning_rate": 0.0001,
+      "loss": 0.0355,
+      "step": 6860
+    },
+    {
+      "epoch": 40.650887573964496,
+      "grad_norm": 2.4892995357513428,
+      "learning_rate": 0.0001,
+      "loss": 0.0345,
+      "step": 6870
+    },
+    {
+      "epoch": 40.710059171597635,
+      "grad_norm": 2.9363138675689697,
+      "learning_rate": 0.0001,
+      "loss": 0.0352,
+      "step": 6880
+    },
+    {
+      "epoch": 40.76923076923077,
+      "grad_norm": 2.7155702114105225,
+      "learning_rate": 0.0001,
+      "loss": 0.0353,
+      "step": 6890
+    },
+    {
+      "epoch": 40.828402366863905,
+      "grad_norm": 2.88881254196167,
+      "learning_rate": 0.0001,
+      "loss": 0.035,
+      "step": 6900
+    },
+    {
+      "epoch": 40.887573964497044,
+      "grad_norm": 2.321302652359009,
+      "learning_rate": 0.0001,
+      "loss": 0.0359,
+      "step": 6910
+    },
+    {
+      "epoch": 40.946745562130175,
+      "grad_norm": 2.143500804901123,
+      "learning_rate": 0.0001,
+      "loss": 0.0352,
+      "step": 6920
+    },
+    {
+      "epoch": 41.005917159763314,
+      "grad_norm": 2.242567539215088,
+      "learning_rate": 0.0001,
+      "loss": 0.0357,
+      "step": 6930
+    },
+    {
+      "epoch": 41.06508875739645,
+      "grad_norm": 2.470912456512451,
+      "learning_rate": 0.0001,
+      "loss": 0.0341,
+      "step": 6940
+    },
+    {
+      "epoch": 41.124260355029584,
+      "grad_norm": 2.708672523498535,
+      "learning_rate": 0.0001,
+      "loss": 0.0374,
+      "step": 6950
+    },
+    {
+      "epoch": 41.18343195266272,
+      "grad_norm": 2.8577306270599365,
+      "learning_rate": 0.0001,
+      "loss": 0.0363,
+      "step": 6960
+    },
+    {
+      "epoch": 41.24260355029586,
+      "grad_norm": 2.4244048595428467,
+      "learning_rate": 0.0001,
+      "loss": 0.0361,
+      "step": 6970
+    },
+    {
+      "epoch": 41.30177514792899,
+      "grad_norm": 2.445434808731079,
+      "learning_rate": 0.0001,
+      "loss": 0.035,
+      "step": 6980
+    },
+    {
+      "epoch": 41.36094674556213,
+      "grad_norm": 2.789565086364746,
+      "learning_rate": 0.0001,
+      "loss": 0.0351,
+      "step": 6990
+    },
+    {
+      "epoch": 41.42011834319526,
+      "grad_norm": 2.238605260848999,
+      "learning_rate": 0.0001,
+      "loss": 0.0327,
+      "step": 7000
+    },
+    {
+      "epoch": 41.4792899408284,
+      "grad_norm": 2.6819660663604736,
+      "learning_rate": 0.0001,
+      "loss": 0.035,
+      "step": 7010
+    },
+    {
+      "epoch": 41.53846153846154,
+      "grad_norm": 2.951162099838257,
+      "learning_rate": 0.0001,
+      "loss": 0.0371,
+      "step": 7020
+    },
+    {
+      "epoch": 41.59763313609467,
+      "grad_norm": 3.6508190631866455,
+      "learning_rate": 0.0001,
+      "loss": 0.0368,
+      "step": 7030
+    },
+    {
+      "epoch": 41.65680473372781,
+      "grad_norm": 2.5877511501312256,
+      "learning_rate": 0.0001,
+      "loss": 0.0366,
+      "step": 7040
+    },
+    {
+      "epoch": 41.71597633136095,
+      "grad_norm": 2.8352103233337402,
+      "learning_rate": 0.0001,
+      "loss": 0.0335,
+      "step": 7050
+    },
+    {
+      "epoch": 41.77514792899408,
+      "grad_norm": 2.5562212467193604,
+      "learning_rate": 0.0001,
+      "loss": 0.0354,
+      "step": 7060
+    },
+    {
+      "epoch": 41.83431952662722,
+      "grad_norm": 2.2208566665649414,
+      "learning_rate": 0.0001,
+      "loss": 0.0349,
+      "step": 7070
+    },
+    {
+      "epoch": 41.89349112426036,
+      "grad_norm": 2.3077497482299805,
+      "learning_rate": 0.0001,
+      "loss": 0.0339,
+      "step": 7080
+    },
+    {
+      "epoch": 41.95266272189349,
+      "grad_norm": 1.9440656900405884,
+      "learning_rate": 0.0001,
+      "loss": 0.0339,
+      "step": 7090
+    },
+    {
+      "epoch": 42.01183431952663,
+      "grad_norm": 1.8663362264633179,
+      "learning_rate": 0.0001,
+      "loss": 0.0354,
+      "step": 7100
+    },
+    {
+      "epoch": 42.071005917159766,
+      "grad_norm": 2.544034242630005,
+      "learning_rate": 0.0001,
+      "loss": 0.0345,
+      "step": 7110
+    },
+    {
+      "epoch": 42.1301775147929,
+      "grad_norm": 1.9617440700531006,
+      "learning_rate": 0.0001,
+      "loss": 0.0357,
+      "step": 7120
+    },
+    {
+      "epoch": 42.18934911242604,
+      "grad_norm": 2.762183666229248,
+      "learning_rate": 0.0001,
+      "loss": 0.0355,
+      "step": 7130
+    },
+    {
+      "epoch": 42.248520710059175,
+      "grad_norm": 2.119473934173584,
+      "learning_rate": 0.0001,
+      "loss": 0.0362,
+      "step": 7140
+    },
+    {
+      "epoch": 42.30769230769231,
+      "grad_norm": 3.2559993267059326,
+      "learning_rate": 0.0001,
+      "loss": 0.0364,
+      "step": 7150
+    },
+    {
+      "epoch": 42.366863905325445,
+      "grad_norm": 2.6083145141601562,
+      "learning_rate": 0.0001,
+      "loss": 0.0377,
+      "step": 7160
+    },
+    {
+      "epoch": 42.42603550295858,
+      "grad_norm": 2.054844617843628,
+      "learning_rate": 0.0001,
+      "loss": 0.0371,
+      "step": 7170
+    },
+    {
+      "epoch": 42.485207100591715,
+      "grad_norm": 2.824201822280884,
+      "learning_rate": 0.0001,
+      "loss": 0.0359,
+      "step": 7180
+    },
+    {
+      "epoch": 42.544378698224854,
+      "grad_norm": 2.8800153732299805,
+      "learning_rate": 0.0001,
+      "loss": 0.0366,
+      "step": 7190
+    },
+    {
+      "epoch": 42.603550295857985,
+      "grad_norm": 2.4122092723846436,
+      "learning_rate": 0.0001,
+      "loss": 0.0356,
+      "step": 7200
+    },
+    {
+      "epoch": 42.662721893491124,
+      "grad_norm": 2.501124858856201,
+      "learning_rate": 0.0001,
+      "loss": 0.0363,
+      "step": 7210
+    },
+    {
+      "epoch": 42.72189349112426,
+      "grad_norm": 2.4320292472839355,
+      "learning_rate": 0.0001,
+      "loss": 0.0346,
+      "step": 7220
+    },
+    {
+      "epoch": 42.781065088757394,
+      "grad_norm": 1.7300277948379517,
+      "learning_rate": 0.0001,
+      "loss": 0.0361,
+      "step": 7230
+    },
+    {
+      "epoch": 42.84023668639053,
+      "grad_norm": 2.2419075965881348,
+      "learning_rate": 0.0001,
+      "loss": 0.0353,
+      "step": 7240
+    },
+    {
+      "epoch": 42.89940828402367,
+      "grad_norm": 2.2318737506866455,
+      "learning_rate": 0.0001,
+      "loss": 0.0363,
+      "step": 7250
+    },
+    {
+      "epoch": 42.9585798816568,
+      "grad_norm": 3.175114870071411,
+      "learning_rate": 0.0001,
+      "loss": 0.0374,
+      "step": 7260
+    },
+    {
+      "epoch": 43.01775147928994,
+      "grad_norm": 1.964353322982788,
+      "learning_rate": 0.0001,
+      "loss": 0.0346,
+      "step": 7270
+    },
+    {
+      "epoch": 43.07692307692308,
+      "grad_norm": 1.7617719173431396,
+      "learning_rate": 0.0001,
+      "loss": 0.035,
+      "step": 7280
+    },
+    {
+      "epoch": 43.13609467455621,
+      "grad_norm": 1.6404685974121094,
+      "learning_rate": 0.0001,
+      "loss": 0.0342,
+      "step": 7290
+    },
+    {
+      "epoch": 43.19526627218935,
+      "grad_norm": 1.7819008827209473,
+      "learning_rate": 0.0001,
+      "loss": 0.0333,
+      "step": 7300
+    },
+    {
+      "epoch": 43.25443786982248,
+      "grad_norm": 1.7775909900665283,
+      "learning_rate": 0.0001,
+      "loss": 0.0353,
+      "step": 7310
+    },
+    {
+      "epoch": 43.31360946745562,
+      "grad_norm": 1.803158164024353,
+      "learning_rate": 0.0001,
+      "loss": 0.0344,
+      "step": 7320
+    },
+    {
+      "epoch": 43.37278106508876,
+      "grad_norm": 1.838541030883789,
+      "learning_rate": 0.0001,
+      "loss": 0.0352,
+      "step": 7330
+    },
+    {
+      "epoch": 43.43195266272189,
+      "grad_norm": 2.348422050476074,
+      "learning_rate": 0.0001,
+      "loss": 0.0349,
+      "step": 7340
+    },
+    {
+      "epoch": 43.49112426035503,
+      "grad_norm": 1.9703973531723022,
+      "learning_rate": 0.0001,
+      "loss": 0.0343,
+      "step": 7350
+    },
+    {
+      "epoch": 43.55029585798817,
+      "grad_norm": 2.0023550987243652,
+      "learning_rate": 0.0001,
+      "loss": 0.0333,
+      "step": 7360
+    },
+    {
+      "epoch": 43.6094674556213,
+      "grad_norm": 2.6823432445526123,
+      "learning_rate": 0.0001,
+      "loss": 0.0341,
+      "step": 7370
+    },
+    {
+      "epoch": 43.66863905325444,
+      "grad_norm": 2.0128867626190186,
+      "learning_rate": 0.0001,
+      "loss": 0.0361,
+      "step": 7380
+    },
+    {
+      "epoch": 43.72781065088758,
+      "grad_norm": 1.992146372795105,
+      "learning_rate": 0.0001,
+      "loss": 0.0344,
+      "step": 7390
+    },
+    {
+      "epoch": 43.78698224852071,
+      "grad_norm": 1.5589733123779297,
+      "learning_rate": 0.0001,
+      "loss": 0.0337,
+      "step": 7400
+    },
+    {
+      "epoch": 43.84615384615385,
+      "grad_norm": 2.0462069511413574,
+      "learning_rate": 0.0001,
+      "loss": 0.034,
+      "step": 7410
+    },
+    {
+      "epoch": 43.905325443786985,
+      "grad_norm": 1.8134146928787231,
+      "learning_rate": 0.0001,
+      "loss": 0.035,
+      "step": 7420
+    },
+    {
+      "epoch": 43.96449704142012,
+      "grad_norm": 2.8621485233306885,
+      "learning_rate": 0.0001,
+      "loss": 0.035,
+      "step": 7430
+    },
+    {
+      "epoch": 44.023668639053255,
+      "grad_norm": 2.185274839401245,
+      "learning_rate": 0.0001,
+      "loss": 0.0359,
+      "step": 7440
+    },
+    {
+      "epoch": 44.082840236686394,
+      "grad_norm": 2.0238847732543945,
+      "learning_rate": 0.0001,
+      "loss": 0.0363,
+      "step": 7450
+    },
+    {
+      "epoch": 44.142011834319526,
+      "grad_norm": 2.4574801921844482,
+      "learning_rate": 0.0001,
+      "loss": 0.0341,
+      "step": 7460
+    },
+    {
+      "epoch": 44.201183431952664,
+      "grad_norm": 1.9655004739761353,
+      "learning_rate": 0.0001,
+      "loss": 0.0343,
+      "step": 7470
+    },
+    {
+      "epoch": 44.260355029585796,
+      "grad_norm": 2.605940103530884,
+      "learning_rate": 0.0001,
+      "loss": 0.0348,
+      "step": 7480
+    },
+    {
+      "epoch": 44.319526627218934,
+      "grad_norm": 2.158921241760254,
+      "learning_rate": 0.0001,
+      "loss": 0.0356,
+      "step": 7490
+    },
+    {
+      "epoch": 44.37869822485207,
+      "grad_norm": 2.360276937484741,
+      "learning_rate": 0.0001,
+      "loss": 0.0334,
+      "step": 7500
+    },
+    {
+      "epoch": 44.437869822485204,
+      "grad_norm": 2.6869707107543945,
+      "learning_rate": 0.0001,
+      "loss": 0.0354,
+      "step": 7510
+    },
+    {
+      "epoch": 44.49704142011834,
+      "grad_norm": 2.9994823932647705,
+      "learning_rate": 0.0001,
+      "loss": 0.0354,
+      "step": 7520
+    },
+    {
+      "epoch": 44.55621301775148,
+      "grad_norm": 3.0287437438964844,
+      "learning_rate": 0.0001,
+      "loss": 0.0362,
+      "step": 7530
+    },
+    {
+      "epoch": 44.61538461538461,
+      "grad_norm": 2.484825372695923,
+      "learning_rate": 0.0001,
+      "loss": 0.0353,
+      "step": 7540
+    },
+    {
+      "epoch": 44.67455621301775,
+      "grad_norm": 3.133854866027832,
+      "learning_rate": 0.0001,
+      "loss": 0.0356,
+      "step": 7550
+    },
+    {
+      "epoch": 44.73372781065089,
+      "grad_norm": 2.2783255577087402,
+      "learning_rate": 0.0001,
+      "loss": 0.0363,
+      "step": 7560
+    },
+    {
+      "epoch": 44.79289940828402,
+      "grad_norm": 2.5281548500061035,
+      "learning_rate": 0.0001,
+      "loss": 0.0339,
+      "step": 7570
+    },
+    {
+      "epoch": 44.85207100591716,
+      "grad_norm": 2.4707069396972656,
+      "learning_rate": 0.0001,
+      "loss": 0.034,
+      "step": 7580
+    },
+    {
+      "epoch": 44.9112426035503,
+      "grad_norm": 2.155191421508789,
+      "learning_rate": 0.0001,
+      "loss": 0.036,
+      "step": 7590
+    },
+    {
+      "epoch": 44.97041420118343,
+      "grad_norm": 2.5939745903015137,
+      "learning_rate": 0.0001,
+      "loss": 0.0354,
+      "step": 7600
+    },
+    {
+      "epoch": 45.02958579881657,
+      "grad_norm": 1.7218459844589233,
+      "learning_rate": 0.0001,
+      "loss": 0.0339,
+      "step": 7610
+    },
+    {
+      "epoch": 45.0887573964497,
+      "grad_norm": 2.3446643352508545,
+      "learning_rate": 0.0001,
+      "loss": 0.0332,
+      "step": 7620
+    },
+    {
+      "epoch": 45.14792899408284,
+      "grad_norm": 1.6789230108261108,
+      "learning_rate": 0.0001,
+      "loss": 0.0324,
+      "step": 7630
+    },
+    {
+      "epoch": 45.20710059171598,
+      "grad_norm": 2.2627315521240234,
+      "learning_rate": 0.0001,
+      "loss": 0.0342,
+      "step": 7640
+    },
+    {
+      "epoch": 45.26627218934911,
+      "grad_norm": 2.318279504776001,
+      "learning_rate": 0.0001,
+      "loss": 0.0354,
+      "step": 7650
+    },
+    {
+      "epoch": 45.32544378698225,
+      "grad_norm": 1.9918235540390015,
+      "learning_rate": 0.0001,
+      "loss": 0.0357,
+      "step": 7660
+    },
+    {
+      "epoch": 45.38461538461539,
+      "grad_norm": 2.218508005142212,
+      "learning_rate": 0.0001,
+      "loss": 0.0349,
+      "step": 7670
+    },
+    {
+      "epoch": 45.44378698224852,
+      "grad_norm": 2.494114637374878,
+      "learning_rate": 0.0001,
+      "loss": 0.0352,
+      "step": 7680
+    },
+    {
+      "epoch": 45.50295857988166,
+      "grad_norm": 1.8003493547439575,
+      "learning_rate": 0.0001,
+      "loss": 0.0348,
+      "step": 7690
+    },
+    {
+      "epoch": 45.562130177514796,
+      "grad_norm": 2.0301296710968018,
+      "learning_rate": 0.0001,
+      "loss": 0.0333,
+      "step": 7700
+    },
+    {
+      "epoch": 45.62130177514793,
+      "grad_norm": 1.9897499084472656,
+      "learning_rate": 0.0001,
+      "loss": 0.0339,
+      "step": 7710
+    },
+    {
+      "epoch": 45.680473372781066,
+      "grad_norm": 1.806282639503479,
+      "learning_rate": 0.0001,
+      "loss": 0.0325,
+      "step": 7720
+    },
+    {
+      "epoch": 45.739644970414204,
+      "grad_norm": 2.1193735599517822,
+      "learning_rate": 0.0001,
+      "loss": 0.0317,
+      "step": 7730
+    },
+    {
+      "epoch": 45.798816568047336,
+      "grad_norm": 1.6028085947036743,
+      "learning_rate": 0.0001,
+      "loss": 0.034,
+      "step": 7740
+    },
+    {
+      "epoch": 45.857988165680474,
+      "grad_norm": 2.2742269039154053,
+      "learning_rate": 0.0001,
+      "loss": 0.0347,
+      "step": 7750
+    },
+    {
+      "epoch": 45.917159763313606,
+      "grad_norm": 2.205491542816162,
+      "learning_rate": 0.0001,
+      "loss": 0.0354,
+      "step": 7760
+    },
+    {
+      "epoch": 45.976331360946745,
+      "grad_norm": 2.260340452194214,
+      "learning_rate": 0.0001,
+      "loss": 0.0332,
+      "step": 7770
+    },
+    {
+      "epoch": 46.03550295857988,
+      "grad_norm": 2.0995397567749023,
+      "learning_rate": 0.0001,
+      "loss": 0.0342,
+      "step": 7780
+    },
+    {
+      "epoch": 46.094674556213015,
+      "grad_norm": 2.1535277366638184,
+      "learning_rate": 0.0001,
+      "loss": 0.0333,
+      "step": 7790
+    },
+    {
+      "epoch": 46.15384615384615,
+      "grad_norm": 1.7552653551101685,
+      "learning_rate": 0.0001,
+      "loss": 0.0332,
+      "step": 7800
+    },
+    {
+      "epoch": 46.21301775147929,
+      "grad_norm": 2.2739157676696777,
+      "learning_rate": 0.0001,
+      "loss": 0.0334,
+      "step": 7810
+    },
+    {
+      "epoch": 46.27218934911242,
+      "grad_norm": 2.3574724197387695,
+      "learning_rate": 0.0001,
+      "loss": 0.0317,
+      "step": 7820
+    },
+    {
+      "epoch": 46.33136094674556,
+      "grad_norm": 2.08369517326355,
+      "learning_rate": 0.0001,
+      "loss": 0.0335,
+      "step": 7830
+    },
+    {
+      "epoch": 46.3905325443787,
+      "grad_norm": 2.3034181594848633,
+      "learning_rate": 0.0001,
+      "loss": 0.0331,
+      "step": 7840
+    },
+    {
+      "epoch": 46.44970414201183,
+      "grad_norm": 1.6334459781646729,
+      "learning_rate": 0.0001,
+      "loss": 0.0333,
+      "step": 7850
+    },
+    {
+      "epoch": 46.50887573964497,
+      "grad_norm": 3.0524678230285645,
+      "learning_rate": 0.0001,
+      "loss": 0.0335,
+      "step": 7860
+    },
+    {
+      "epoch": 46.56804733727811,
+      "grad_norm": 1.9907009601593018,
+      "learning_rate": 0.0001,
+      "loss": 0.0347,
+      "step": 7870
+    },
+    {
+      "epoch": 46.62721893491124,
+      "grad_norm": 2.152864694595337,
+      "learning_rate": 0.0001,
+      "loss": 0.0351,
+      "step": 7880
+    },
+    {
+      "epoch": 46.68639053254438,
+      "grad_norm": 2.657325267791748,
+      "learning_rate": 0.0001,
+      "loss": 0.0353,
+      "step": 7890
+    },
+    {
+      "epoch": 46.74556213017752,
+      "grad_norm": 2.6477668285369873,
+      "learning_rate": 0.0001,
+      "loss": 0.0349,
+      "step": 7900
+    },
+    {
+      "epoch": 46.80473372781065,
+      "grad_norm": 2.3875138759613037,
+      "learning_rate": 0.0001,
+      "loss": 0.0355,
+      "step": 7910
+    },
+    {
+      "epoch": 46.86390532544379,
+      "grad_norm": 2.425374746322632,
+      "learning_rate": 0.0001,
+      "loss": 0.035,
+      "step": 7920
+    },
+    {
+      "epoch": 46.92307692307692,
+      "grad_norm": 2.205251932144165,
+      "learning_rate": 0.0001,
+      "loss": 0.0349,
+      "step": 7930
+    },
+    {
+      "epoch": 46.98224852071006,
+      "grad_norm": 2.3108153343200684,
+      "learning_rate": 0.0001,
+      "loss": 0.033,
+      "step": 7940
+    },
+    {
+      "epoch": 47.0414201183432,
+      "grad_norm": 2.0401222705841064,
+      "learning_rate": 0.0001,
+      "loss": 0.0331,
+      "step": 7950
+    },
+    {
+      "epoch": 47.10059171597633,
+      "grad_norm": 1.8593013286590576,
+      "learning_rate": 0.0001,
+      "loss": 0.0335,
+      "step": 7960
+    },
+    {
+      "epoch": 47.15976331360947,
+      "grad_norm": 1.5773780345916748,
+      "learning_rate": 0.0001,
+      "loss": 0.0335,
+      "step": 7970
+    },
+    {
+      "epoch": 47.218934911242606,
+      "grad_norm": 2.183591842651367,
+      "learning_rate": 0.0001,
+      "loss": 0.0338,
+      "step": 7980
+    },
+    {
+      "epoch": 47.27810650887574,
+      "grad_norm": 1.762555480003357,
+      "learning_rate": 0.0001,
+      "loss": 0.0348,
+      "step": 7990
+    },
+    {
+      "epoch": 47.337278106508876,
+      "grad_norm": 2.106579303741455,
+      "learning_rate": 0.0001,
+      "loss": 0.0359,
+      "step": 8000
+    },
+    {
+      "epoch": 47.396449704142015,
+      "grad_norm": 1.7366595268249512,
+      "learning_rate": 0.0001,
+      "loss": 0.0357,
+      "step": 8010
+    },
+    {
+      "epoch": 47.455621301775146,
+      "grad_norm": 1.805619478225708,
+      "learning_rate": 0.0001,
+      "loss": 0.0352,
+      "step": 8020
+    },
+    {
+      "epoch": 47.514792899408285,
+      "grad_norm": 1.7611534595489502,
+      "learning_rate": 0.0001,
+      "loss": 0.0347,
+      "step": 8030
+    },
+    {
+      "epoch": 47.57396449704142,
+      "grad_norm": 2.209580898284912,
+      "learning_rate": 0.0001,
+      "loss": 0.0358,
+      "step": 8040
+    },
+    {
+      "epoch": 47.633136094674555,
+      "grad_norm": 1.5984245538711548,
+      "learning_rate": 0.0001,
+      "loss": 0.0358,
+      "step": 8050
+    },
+    {
+      "epoch": 47.69230769230769,
+      "grad_norm": 1.733137607574463,
+      "learning_rate": 0.0001,
+      "loss": 0.0349,
+      "step": 8060
+    },
+    {
+      "epoch": 47.751479289940825,
+      "grad_norm": 2.3158180713653564,
+      "learning_rate": 0.0001,
+      "loss": 0.0354,
+      "step": 8070
+    },
+    {
+      "epoch": 47.81065088757396,
+      "grad_norm": 1.8525919914245605,
+      "learning_rate": 0.0001,
+      "loss": 0.035,
+      "step": 8080
+    },
+    {
+      "epoch": 47.8698224852071,
+      "grad_norm": 1.7492717504501343,
+      "learning_rate": 0.0001,
+      "loss": 0.036,
+      "step": 8090
+    },
+    {
+      "epoch": 47.928994082840234,
+      "grad_norm": 2.016191244125366,
+      "learning_rate": 0.0001,
+      "loss": 0.0363,
+      "step": 8100
+    },
+    {
+      "epoch": 47.98816568047337,
+      "grad_norm": 2.8384389877319336,
+      "learning_rate": 0.0001,
+      "loss": 0.0385,
+      "step": 8110
+    },
+    {
+      "epoch": 48.04733727810651,
+      "grad_norm": 2.2267377376556396,
+      "learning_rate": 0.0001,
+      "loss": 0.035,
+      "step": 8120
+    },
+    {
+      "epoch": 48.10650887573964,
+      "grad_norm": 1.9384708404541016,
+      "learning_rate": 0.0001,
+      "loss": 0.0348,
+      "step": 8130
+    },
+    {
+      "epoch": 48.16568047337278,
+      "grad_norm": 1.9673793315887451,
+      "learning_rate": 0.0001,
+      "loss": 0.0343,
+      "step": 8140
+    },
+    {
+      "epoch": 48.22485207100592,
+      "grad_norm": 1.963778018951416,
+      "learning_rate": 0.0001,
+      "loss": 0.0354,
+      "step": 8150
+    },
+    {
+      "epoch": 48.28402366863905,
+      "grad_norm": 1.833004117012024,
+      "learning_rate": 0.0001,
+      "loss": 0.0349,
+      "step": 8160
+    },
+    {
+      "epoch": 48.34319526627219,
+      "grad_norm": 1.4917781352996826,
+      "learning_rate": 0.0001,
+      "loss": 0.0372,
+      "step": 8170
+    },
+    {
+      "epoch": 48.40236686390533,
+      "grad_norm": 2.036400079727173,
+      "learning_rate": 0.0001,
+      "loss": 0.0338,
+      "step": 8180
+    },
+    {
+      "epoch": 48.46153846153846,
+      "grad_norm": 2.428159475326538,
+      "learning_rate": 0.0001,
+      "loss": 0.0342,
+      "step": 8190
+    },
+    {
+      "epoch": 48.5207100591716,
+      "grad_norm": 2.092939615249634,
+      "learning_rate": 0.0001,
+      "loss": 0.0342,
+      "step": 8200
+    },
+    {
+      "epoch": 48.57988165680474,
+      "grad_norm": 2.055976390838623,
+      "learning_rate": 0.0001,
+      "loss": 0.0353,
+      "step": 8210
+    },
+    {
+      "epoch": 48.63905325443787,
+      "grad_norm": 2.0701112747192383,
+      "learning_rate": 0.0001,
+      "loss": 0.0346,
+      "step": 8220
+    },
+    {
+      "epoch": 48.69822485207101,
+      "grad_norm": 2.2323639392852783,
+      "learning_rate": 0.0001,
+      "loss": 0.0351,
+      "step": 8230
+    },
+    {
+      "epoch": 48.75739644970414,
+      "grad_norm": 1.9810377359390259,
+      "learning_rate": 0.0001,
+      "loss": 0.0337,
+      "step": 8240
+    },
+    {
+      "epoch": 48.81656804733728,
+      "grad_norm": 2.136165142059326,
+      "learning_rate": 0.0001,
+      "loss": 0.0332,
+      "step": 8250
+    },
+    {
+      "epoch": 48.875739644970416,
+      "grad_norm": 1.5917547941207886,
+      "learning_rate": 0.0001,
+      "loss": 0.0329,
+      "step": 8260
+    },
+    {
+      "epoch": 48.93491124260355,
+      "grad_norm": 2.1646876335144043,
+      "learning_rate": 0.0001,
+      "loss": 0.0321,
+      "step": 8270
+    },
+    {
+      "epoch": 48.994082840236686,
+      "grad_norm": 1.7858912944793701,
+      "learning_rate": 0.0001,
+      "loss": 0.033,
+      "step": 8280
+    },
+    {
+      "epoch": 49.053254437869825,
+      "grad_norm": 1.750210165977478,
+      "learning_rate": 0.0001,
+      "loss": 0.0333,
+      "step": 8290
+    },
+    {
+      "epoch": 49.112426035502956,
+      "grad_norm": 1.8941763639450073,
+      "learning_rate": 0.0001,
+      "loss": 0.0325,
+      "step": 8300
+    },
+    {
+      "epoch": 49.171597633136095,
+      "grad_norm": 1.9790654182434082,
+      "learning_rate": 0.0001,
+      "loss": 0.0324,
+      "step": 8310
+    },
+    {
+      "epoch": 49.23076923076923,
+      "grad_norm": 2.2026004791259766,
+      "learning_rate": 0.0001,
+      "loss": 0.0327,
+      "step": 8320
+    },
+    {
+      "epoch": 49.289940828402365,
+      "grad_norm": 2.336383581161499,
+      "learning_rate": 0.0001,
+      "loss": 0.0332,
+      "step": 8330
+    },
+    {
+      "epoch": 49.349112426035504,
+      "grad_norm": 2.1846563816070557,
+      "learning_rate": 0.0001,
+      "loss": 0.0329,
+      "step": 8340
+    },
+    {
+      "epoch": 49.40828402366864,
+      "grad_norm": 2.0261788368225098,
+      "learning_rate": 0.0001,
+      "loss": 0.0333,
+      "step": 8350
+    },
+    {
+      "epoch": 49.467455621301774,
+      "grad_norm": 2.0120623111724854,
+      "learning_rate": 0.0001,
+      "loss": 0.0323,
+      "step": 8360
+    },
+    {
+      "epoch": 49.52662721893491,
+      "grad_norm": 1.5230213403701782,
+      "learning_rate": 0.0001,
+      "loss": 0.0322,
+      "step": 8370
+    },
+    {
+      "epoch": 49.585798816568044,
+      "grad_norm": 2.6539719104766846,
+      "learning_rate": 0.0001,
+      "loss": 0.0321,
+      "step": 8380
+    },
+    {
+      "epoch": 49.64497041420118,
+      "grad_norm": 2.20536470413208,
+      "learning_rate": 0.0001,
+      "loss": 0.0347,
+      "step": 8390
+    },
+    {
+      "epoch": 49.70414201183432,
+      "grad_norm": 2.005849838256836,
+      "learning_rate": 0.0001,
+      "loss": 0.0328,
+      "step": 8400
+    },
+    {
+      "epoch": 49.76331360946745,
+      "grad_norm": 2.293684720993042,
+      "learning_rate": 0.0001,
+      "loss": 0.0335,
+      "step": 8410
+    },
+    {
+      "epoch": 49.82248520710059,
+      "grad_norm": 2.1393322944641113,
+      "learning_rate": 0.0001,
+      "loss": 0.0342,
+      "step": 8420
+    },
+    {
+      "epoch": 49.88165680473373,
+      "grad_norm": 2.6882436275482178,
+      "learning_rate": 0.0001,
+      "loss": 0.0342,
+      "step": 8430
+    },
+    {
+      "epoch": 49.94082840236686,
+      "grad_norm": 2.57011342048645,
+      "learning_rate": 0.0001,
+      "loss": 0.0343,
+      "step": 8440
+    },
+    {
+      "epoch": 50.0,
+      "grad_norm": 2.610537052154541,
+      "learning_rate": 0.0001,
+      "loss": 0.0344,
+      "step": 8450
+    },
+    {
+      "epoch": 50.05917159763314,
+      "grad_norm": 2.353839635848999,
+      "learning_rate": 0.0001,
+      "loss": 0.0328,
+      "step": 8460
+    },
+    {
+      "epoch": 50.11834319526627,
+      "grad_norm": 2.325312852859497,
+      "learning_rate": 0.0001,
+      "loss": 0.0351,
+      "step": 8470
+    },
+    {
+      "epoch": 50.17751479289941,
+      "grad_norm": 1.697034478187561,
+      "learning_rate": 0.0001,
+      "loss": 0.0331,
+      "step": 8480
+    },
+    {
+      "epoch": 50.23668639053255,
+      "grad_norm": 2.1400222778320312,
+      "learning_rate": 0.0001,
+      "loss": 0.0338,
+      "step": 8490
+    },
+    {
+      "epoch": 50.29585798816568,
+      "grad_norm": 2.2705936431884766,
+      "learning_rate": 0.0001,
+      "loss": 0.0351,
+      "step": 8500
+    },
+    {
+      "epoch": 50.35502958579882,
+      "grad_norm": 1.5698002576828003,
+      "learning_rate": 0.0001,
+      "loss": 0.0337,
+      "step": 8510
+    },
+    {
+      "epoch": 50.414201183431956,
+      "grad_norm": 2.112440824508667,
+      "learning_rate": 0.0001,
+      "loss": 0.0323,
+      "step": 8520
+    },
+    {
+      "epoch": 50.47337278106509,
+      "grad_norm": 1.8457298278808594,
+      "learning_rate": 0.0001,
+      "loss": 0.0319,
+      "step": 8530
+    },
+    {
+      "epoch": 50.532544378698226,
+      "grad_norm": 2.1696298122406006,
+      "learning_rate": 0.0001,
+      "loss": 0.0336,
+      "step": 8540
+    },
+    {
+      "epoch": 50.59171597633136,
+      "grad_norm": 2.0588157176971436,
+      "learning_rate": 0.0001,
+      "loss": 0.0319,
+      "step": 8550
+    },
+    {
+      "epoch": 50.650887573964496,
+      "grad_norm": 2.0903031826019287,
+      "learning_rate": 0.0001,
+      "loss": 0.0329,
+      "step": 8560
+    },
+    {
+      "epoch": 50.710059171597635,
+      "grad_norm": 2.194716215133667,
+      "learning_rate": 0.0001,
+      "loss": 0.0316,
+      "step": 8570
+    },
+    {
+      "epoch": 50.76923076923077,
+      "grad_norm": 1.5895949602127075,
+      "learning_rate": 0.0001,
+      "loss": 0.0325,
+      "step": 8580
+    },
+    {
+      "epoch": 50.828402366863905,
+      "grad_norm": 2.1726391315460205,
+      "learning_rate": 0.0001,
+      "loss": 0.0336,
+      "step": 8590
+    },
+    {
+      "epoch": 50.887573964497044,
+      "grad_norm": 1.830047845840454,
+      "learning_rate": 0.0001,
+      "loss": 0.0331,
+      "step": 8600
+    },
+    {
+      "epoch": 50.946745562130175,
+      "grad_norm": 2.1310925483703613,
+      "learning_rate": 0.0001,
+      "loss": 0.0307,
+      "step": 8610
+    },
+    {
+      "epoch": 51.005917159763314,
+      "grad_norm": 2.0378634929656982,
+      "learning_rate": 0.0001,
+      "loss": 0.0332,
+      "step": 8620
+    },
+    {
+      "epoch": 51.06508875739645,
+      "grad_norm": 1.8991228342056274,
+      "learning_rate": 0.0001,
+      "loss": 0.0339,
+      "step": 8630
+    },
+    {
+      "epoch": 51.124260355029584,
+      "grad_norm": 2.16483736038208,
+      "learning_rate": 0.0001,
+      "loss": 0.0339,
+      "step": 8640
+    },
+    {
+      "epoch": 51.18343195266272,
+      "grad_norm": 2.067754030227661,
+      "learning_rate": 0.0001,
+      "loss": 0.0331,
+      "step": 8650
+    },
+    {
+      "epoch": 51.24260355029586,
+      "grad_norm": 2.11576509475708,
+      "learning_rate": 0.0001,
+      "loss": 0.0322,
+      "step": 8660
+    },
+    {
+      "epoch": 51.30177514792899,
+      "grad_norm": 1.921156406402588,
+      "learning_rate": 0.0001,
+      "loss": 0.0328,
+      "step": 8670
+    },
+    {
+      "epoch": 51.36094674556213,
+      "grad_norm": 1.7753546237945557,
+      "learning_rate": 0.0001,
+      "loss": 0.03,
+      "step": 8680
+    },
+    {
+      "epoch": 51.42011834319526,
+      "grad_norm": 1.7234447002410889,
+      "learning_rate": 0.0001,
+      "loss": 0.0314,
+      "step": 8690
+    },
+    {
+      "epoch": 51.4792899408284,
+      "grad_norm": 1.6457452774047852,
+      "learning_rate": 0.0001,
+      "loss": 0.0322,
+      "step": 8700
+    },
+    {
+      "epoch": 51.53846153846154,
+      "grad_norm": 2.2582039833068848,
+      "learning_rate": 0.0001,
+      "loss": 0.0332,
+      "step": 8710
+    },
+    {
+      "epoch": 51.59763313609467,
+      "grad_norm": 2.5045700073242188,
+      "learning_rate": 0.0001,
+      "loss": 0.0334,
+      "step": 8720
+    },
+    {
+      "epoch": 51.65680473372781,
+      "grad_norm": 1.9544016122817993,
+      "learning_rate": 0.0001,
+      "loss": 0.0329,
+      "step": 8730
+    },
+    {
+      "epoch": 51.71597633136095,
+      "grad_norm": 2.2486937046051025,
+      "learning_rate": 0.0001,
+      "loss": 0.0322,
+      "step": 8740
+    },
+    {
+      "epoch": 51.77514792899408,
+      "grad_norm": 2.0340559482574463,
+      "learning_rate": 0.0001,
+      "loss": 0.032,
+      "step": 8750
+    },
+    {
+      "epoch": 51.83431952662722,
+      "grad_norm": 2.8832316398620605,
+      "learning_rate": 0.0001,
+      "loss": 0.0329,
+      "step": 8760
+    },
+    {
+      "epoch": 51.89349112426036,
+      "grad_norm": 2.1337101459503174,
+      "learning_rate": 0.0001,
+      "loss": 0.0326,
+      "step": 8770
+    },
+    {
+      "epoch": 51.95266272189349,
+      "grad_norm": 2.8878862857818604,
+      "learning_rate": 0.0001,
+      "loss": 0.0321,
+      "step": 8780
+    },
+    {
+      "epoch": 52.01183431952663,
+      "grad_norm": 2.022778034210205,
+      "learning_rate": 0.0001,
+      "loss": 0.0329,
+      "step": 8790
+    },
+    {
+      "epoch": 52.071005917159766,
+      "grad_norm": 1.9821972846984863,
+      "learning_rate": 0.0001,
+      "loss": 0.0307,
+      "step": 8800
+    },
+    {
+      "epoch": 52.1301775147929,
+      "grad_norm": 2.0924289226531982,
+      "learning_rate": 0.0001,
+      "loss": 0.0308,
+      "step": 8810
+    },
+    {
+      "epoch": 52.18934911242604,
+      "grad_norm": 2.3329174518585205,
+      "learning_rate": 0.0001,
+      "loss": 0.031,
+      "step": 8820
+    },
+    {
+      "epoch": 52.248520710059175,
+      "grad_norm": 1.8705933094024658,
+      "learning_rate": 0.0001,
+      "loss": 0.0322,
+      "step": 8830
+    },
+    {
+      "epoch": 52.30769230769231,
+      "grad_norm": 2.2881994247436523,
+      "learning_rate": 0.0001,
+      "loss": 0.0324,
+      "step": 8840
+    },
+    {
+      "epoch": 52.366863905325445,
+      "grad_norm": 1.7713308334350586,
+      "learning_rate": 0.0001,
+      "loss": 0.0318,
+      "step": 8850
+    },
+    {
+      "epoch": 52.42603550295858,
+      "grad_norm": 2.0944693088531494,
+      "learning_rate": 0.0001,
+      "loss": 0.0315,
+      "step": 8860
+    },
+    {
+      "epoch": 52.485207100591715,
+      "grad_norm": 1.867463231086731,
+      "learning_rate": 0.0001,
+      "loss": 0.0313,
+      "step": 8870
+    },
+    {
+      "epoch": 52.544378698224854,
+      "grad_norm": 1.8982528448104858,
+      "learning_rate": 0.0001,
+      "loss": 0.0332,
+      "step": 8880
+    },
+    {
+      "epoch": 52.603550295857985,
+      "grad_norm": 1.8469318151474,
+      "learning_rate": 0.0001,
+      "loss": 0.0319,
+      "step": 8890
+    },
+    {
+      "epoch": 52.662721893491124,
+      "grad_norm": 1.6908539533615112,
+      "learning_rate": 0.0001,
+      "loss": 0.0314,
+      "step": 8900
+    },
+    {
+      "epoch": 52.72189349112426,
+      "grad_norm": 2.048063278198242,
+      "learning_rate": 0.0001,
+      "loss": 0.0319,
+      "step": 8910
+    },
+    {
+      "epoch": 52.781065088757394,
+      "grad_norm": 2.18879771232605,
+      "learning_rate": 0.0001,
+      "loss": 0.0321,
+      "step": 8920
+    },
+    {
+      "epoch": 52.84023668639053,
+      "grad_norm": 1.7338558435440063,
+      "learning_rate": 0.0001,
+      "loss": 0.0314,
+      "step": 8930
+    },
+    {
+      "epoch": 52.89940828402367,
+      "grad_norm": 1.8839168548583984,
+      "learning_rate": 0.0001,
+      "loss": 0.031,
+      "step": 8940
+    },
+    {
+      "epoch": 52.9585798816568,
+      "grad_norm": 1.5505837202072144,
+      "learning_rate": 0.0001,
+      "loss": 0.032,
+      "step": 8950
+    },
+    {
+      "epoch": 53.01775147928994,
+      "grad_norm": 1.5521057844161987,
+      "learning_rate": 0.0001,
+      "loss": 0.0325,
+      "step": 8960
+    },
+    {
+      "epoch": 53.07692307692308,
+      "grad_norm": 2.180845260620117,
+      "learning_rate": 0.0001,
+      "loss": 0.0328,
+      "step": 8970
+    },
+    {
+      "epoch": 53.13609467455621,
+      "grad_norm": 2.2614099979400635,
+      "learning_rate": 0.0001,
+      "loss": 0.0338,
+      "step": 8980
+    },
+    {
+      "epoch": 53.19526627218935,
+      "grad_norm": 2.6195003986358643,
+      "learning_rate": 0.0001,
+      "loss": 0.0336,
+      "step": 8990
+    },
+    {
+      "epoch": 53.25443786982248,
+      "grad_norm": 1.7065534591674805,
+      "learning_rate": 0.0001,
+      "loss": 0.0316,
+      "step": 9000
+    },
+    {
+      "epoch": 53.31360946745562,
+      "grad_norm": 1.9091246128082275,
+      "learning_rate": 0.0001,
+      "loss": 0.0323,
+      "step": 9010
+    },
+    {
+      "epoch": 53.37278106508876,
+      "grad_norm": 1.7090070247650146,
+      "learning_rate": 0.0001,
+      "loss": 0.0324,
+      "step": 9020
+    },
+    {
+      "epoch": 53.43195266272189,
+      "grad_norm": 1.6685603857040405,
+      "learning_rate": 0.0001,
+      "loss": 0.0321,
+      "step": 9030
+    },
+    {
+      "epoch": 53.49112426035503,
+      "grad_norm": 2.881209135055542,
+      "learning_rate": 0.0001,
+      "loss": 0.0322,
+      "step": 9040
+    },
+    {
+      "epoch": 53.55029585798817,
+      "grad_norm": 2.0528886318206787,
+      "learning_rate": 0.0001,
+      "loss": 0.0336,
+      "step": 9050
+    },
+    {
+      "epoch": 53.6094674556213,
+      "grad_norm": 2.2183735370635986,
+      "learning_rate": 0.0001,
+      "loss": 0.033,
+      "step": 9060
+    },
+    {
+      "epoch": 53.66863905325444,
+      "grad_norm": 2.2254414558410645,
+      "learning_rate": 0.0001,
+      "loss": 0.0313,
+      "step": 9070
+    },
+    {
+      "epoch": 53.72781065088758,
+      "grad_norm": 2.3155035972595215,
+      "learning_rate": 0.0001,
+      "loss": 0.0324,
+      "step": 9080
+    },
+    {
+      "epoch": 53.78698224852071,
+      "grad_norm": 2.069934844970703,
+      "learning_rate": 0.0001,
+      "loss": 0.0316,
+      "step": 9090
+    },
+    {
+      "epoch": 53.84615384615385,
+      "grad_norm": 2.18259334564209,
+      "learning_rate": 0.0001,
+      "loss": 0.031,
+      "step": 9100
+    },
+    {
+      "epoch": 53.905325443786985,
+      "grad_norm": 2.5839242935180664,
+      "learning_rate": 0.0001,
+      "loss": 0.0319,
+      "step": 9110
+    },
+    {
+      "epoch": 53.96449704142012,
+      "grad_norm": 2.640685558319092,
+      "learning_rate": 0.0001,
+      "loss": 0.0312,
+      "step": 9120
+    },
+    {
+      "epoch": 54.023668639053255,
+      "grad_norm": 1.726183295249939,
+      "learning_rate": 0.0001,
+      "loss": 0.0313,
+      "step": 9130
+    },
+    {
+      "epoch": 54.082840236686394,
+      "grad_norm": 2.129166841506958,
+      "learning_rate": 0.0001,
+      "loss": 0.0299,
+      "step": 9140
+    },
+    {
+      "epoch": 54.142011834319526,
+      "grad_norm": 1.935209035873413,
+      "learning_rate": 0.0001,
+      "loss": 0.0293,
+      "step": 9150
+    },
+    {
+      "epoch": 54.201183431952664,
+      "grad_norm": 1.8166563510894775,
+      "learning_rate": 0.0001,
+      "loss": 0.0306,
+      "step": 9160
+    },
+    {
+      "epoch": 54.260355029585796,
+      "grad_norm": 1.5967921018600464,
+      "learning_rate": 0.0001,
+      "loss": 0.0303,
+      "step": 9170
+    },
+    {
+      "epoch": 54.319526627218934,
+      "grad_norm": 1.6898949146270752,
+      "learning_rate": 0.0001,
+      "loss": 0.0314,
+      "step": 9180
+    },
+    {
+      "epoch": 54.37869822485207,
+      "grad_norm": 1.7409635782241821,
+      "learning_rate": 0.0001,
+      "loss": 0.0318,
+      "step": 9190
+    },
+    {
+      "epoch": 54.437869822485204,
+      "grad_norm": 1.6817439794540405,
+      "learning_rate": 0.0001,
+      "loss": 0.0318,
+      "step": 9200
+    },
+    {
+      "epoch": 54.49704142011834,
+      "grad_norm": 1.8987990617752075,
+      "learning_rate": 0.0001,
+      "loss": 0.0313,
+      "step": 9210
+    },
+    {
+      "epoch": 54.55621301775148,
+      "grad_norm": 1.9725265502929688,
+      "learning_rate": 0.0001,
+      "loss": 0.0324,
+      "step": 9220
+    },
+    {
+      "epoch": 54.61538461538461,
+      "grad_norm": 1.6091896295547485,
+      "learning_rate": 0.0001,
+      "loss": 0.0318,
+      "step": 9230
+    },
+    {
+      "epoch": 54.67455621301775,
+      "grad_norm": 1.8824845552444458,
+      "learning_rate": 0.0001,
+      "loss": 0.0313,
+      "step": 9240
+    },
+    {
+      "epoch": 54.73372781065089,
+      "grad_norm": 2.031568765640259,
+      "learning_rate": 0.0001,
+      "loss": 0.0314,
+      "step": 9250
+    },
+    {
+      "epoch": 54.79289940828402,
+      "grad_norm": 2.086686134338379,
+      "learning_rate": 0.0001,
+      "loss": 0.0323,
+      "step": 9260
+    },
+    {
+      "epoch": 54.85207100591716,
+      "grad_norm": 1.8941878080368042,
+      "learning_rate": 0.0001,
+      "loss": 0.0313,
+      "step": 9270
+    },
+    {
+      "epoch": 54.9112426035503,
+      "grad_norm": 2.147495985031128,
+      "learning_rate": 0.0001,
+      "loss": 0.0324,
+      "step": 9280
+    },
+    {
+      "epoch": 54.97041420118343,
+      "grad_norm": 1.473995327949524,
+      "learning_rate": 0.0001,
+      "loss": 0.0318,
+      "step": 9290
+    },
+    {
+      "epoch": 55.02958579881657,
+      "grad_norm": 1.961800217628479,
+      "learning_rate": 0.0001,
+      "loss": 0.0315,
+      "step": 9300
+    },
+    {
+      "epoch": 55.0887573964497,
+      "grad_norm": 2.208514928817749,
+      "learning_rate": 0.0001,
+      "loss": 0.0336,
+      "step": 9310
+    },
+    {
+      "epoch": 55.14792899408284,
+      "grad_norm": 2.0775299072265625,
+      "learning_rate": 0.0001,
+      "loss": 0.0323,
+      "step": 9320
+    },
+    {
+      "epoch": 55.20710059171598,
+      "grad_norm": 2.0815985202789307,
+      "learning_rate": 0.0001,
+      "loss": 0.0328,
+      "step": 9330
+    },
+    {
+      "epoch": 55.26627218934911,
+      "grad_norm": 1.3869959115982056,
+      "learning_rate": 0.0001,
+      "loss": 0.032,
+      "step": 9340
+    },
+    {
+      "epoch": 55.32544378698225,
+      "grad_norm": 2.2878143787384033,
+      "learning_rate": 0.0001,
+      "loss": 0.0321,
+      "step": 9350
+    },
+    {
+      "epoch": 55.38461538461539,
+      "grad_norm": 1.55245840549469,
+      "learning_rate": 0.0001,
+      "loss": 0.0328,
+      "step": 9360
+    },
+    {
+      "epoch": 55.44378698224852,
+      "grad_norm": 1.696345329284668,
+      "learning_rate": 0.0001,
+      "loss": 0.0316,
+      "step": 9370
+    },
+    {
+      "epoch": 55.50295857988166,
+      "grad_norm": 1.4600911140441895,
+      "learning_rate": 0.0001,
+      "loss": 0.0318,
+      "step": 9380
+    },
+    {
+      "epoch": 55.562130177514796,
+      "grad_norm": 1.8653849363327026,
+      "learning_rate": 0.0001,
+      "loss": 0.0309,
+      "step": 9390
+    },
+    {
+      "epoch": 55.62130177514793,
+      "grad_norm": 2.01830792427063,
+      "learning_rate": 0.0001,
+      "loss": 0.0317,
+      "step": 9400
+    },
+    {
+      "epoch": 55.680473372781066,
+      "grad_norm": 2.103524923324585,
+      "learning_rate": 0.0001,
+      "loss": 0.0317,
+      "step": 9410
+    },
+    {
+      "epoch": 55.739644970414204,
+      "grad_norm": 1.9455193281173706,
+      "learning_rate": 0.0001,
+      "loss": 0.0303,
+      "step": 9420
+    },
+    {
+      "epoch": 55.798816568047336,
+      "grad_norm": 1.6023354530334473,
+      "learning_rate": 0.0001,
+      "loss": 0.0306,
+      "step": 9430
+    },
+    {
+      "epoch": 55.857988165680474,
+      "grad_norm": 1.3738919496536255,
+      "learning_rate": 0.0001,
+      "loss": 0.03,
+      "step": 9440
+    },
+    {
+      "epoch": 55.917159763313606,
+      "grad_norm": 1.9328325986862183,
+      "learning_rate": 0.0001,
+      "loss": 0.0306,
+      "step": 9450
+    },
+    {
+      "epoch": 55.976331360946745,
+      "grad_norm": 2.0509064197540283,
+      "learning_rate": 0.0001,
+      "loss": 0.0321,
+      "step": 9460
+    },
+    {
+      "epoch": 56.03550295857988,
+      "grad_norm": 2.0007171630859375,
+      "learning_rate": 0.0001,
+      "loss": 0.0336,
+      "step": 9470
+    },
+    {
+      "epoch": 56.094674556213015,
+      "grad_norm": 1.585517168045044,
+      "learning_rate": 0.0001,
+      "loss": 0.0316,
+      "step": 9480
+    },
+    {
+      "epoch": 56.15384615384615,
+      "grad_norm": 1.7611768245697021,
+      "learning_rate": 0.0001,
+      "loss": 0.031,
+      "step": 9490
+    },
+    {
+      "epoch": 56.21301775147929,
+      "grad_norm": 2.2337145805358887,
+      "learning_rate": 0.0001,
+      "loss": 0.0326,
+      "step": 9500
+    },
+    {
+      "epoch": 56.27218934911242,
+      "grad_norm": 1.5489686727523804,
+      "learning_rate": 0.0001,
+      "loss": 0.0324,
+      "step": 9510
+    },
+    {
+      "epoch": 56.33136094674556,
+      "grad_norm": 2.257706880569458,
+      "learning_rate": 0.0001,
+      "loss": 0.0322,
+      "step": 9520
+    },
+    {
+      "epoch": 56.3905325443787,
+      "grad_norm": 1.9078609943389893,
+      "learning_rate": 0.0001,
+      "loss": 0.0327,
+      "step": 9530
+    },
+    {
+      "epoch": 56.44970414201183,
+      "grad_norm": 2.397015333175659,
+      "learning_rate": 0.0001,
+      "loss": 0.0324,
+      "step": 9540
+    },
+    {
+      "epoch": 56.50887573964497,
+      "grad_norm": 1.8677520751953125,
+      "learning_rate": 0.0001,
+      "loss": 0.0327,
+      "step": 9550
+    },
+    {
+      "epoch": 56.56804733727811,
+      "grad_norm": 1.9998503923416138,
+      "learning_rate": 0.0001,
+      "loss": 0.0325,
+      "step": 9560
+    },
+    {
+      "epoch": 56.62721893491124,
+      "grad_norm": 2.2100415229797363,
+      "learning_rate": 0.0001,
+      "loss": 0.0331,
+      "step": 9570
+    },
+    {
+      "epoch": 56.68639053254438,
+      "grad_norm": 2.163853883743286,
+      "learning_rate": 0.0001,
+      "loss": 0.034,
+      "step": 9580
+    },
+    {
+      "epoch": 56.74556213017752,
+      "grad_norm": 2.271080732345581,
+      "learning_rate": 0.0001,
+      "loss": 0.0331,
+      "step": 9590
+    },
+    {
+      "epoch": 56.80473372781065,
+      "grad_norm": 1.584336757659912,
+      "learning_rate": 0.0001,
+      "loss": 0.0306,
+      "step": 9600
+    },
+    {
+      "epoch": 56.86390532544379,
+      "grad_norm": 1.7379599809646606,
+      "learning_rate": 0.0001,
+      "loss": 0.0303,
+      "step": 9610
+    },
+    {
+      "epoch": 56.92307692307692,
+      "grad_norm": 2.0200870037078857,
+      "learning_rate": 0.0001,
+      "loss": 0.0325,
+      "step": 9620
+    },
+    {
+      "epoch": 56.98224852071006,
+      "grad_norm": 2.2698328495025635,
+      "learning_rate": 0.0001,
+      "loss": 0.0326,
+      "step": 9630
+    },
+    {
+      "epoch": 57.0414201183432,
+      "grad_norm": 1.7476683855056763,
+      "learning_rate": 0.0001,
+      "loss": 0.031,
+      "step": 9640
+    },
+    {
+      "epoch": 57.10059171597633,
+      "grad_norm": 1.7359435558319092,
+      "learning_rate": 0.0001,
+      "loss": 0.0304,
+      "step": 9650
+    },
+    {
+      "epoch": 57.15976331360947,
+      "grad_norm": 1.9918187856674194,
+      "learning_rate": 0.0001,
+      "loss": 0.0304,
+      "step": 9660
+    },
+    {
+      "epoch": 57.218934911242606,
+      "grad_norm": 2.379845380783081,
+      "learning_rate": 0.0001,
+      "loss": 0.0316,
+      "step": 9670
+    },
+    {
+      "epoch": 57.27810650887574,
+      "grad_norm": 1.6139721870422363,
+      "learning_rate": 0.0001,
+      "loss": 0.0307,
+      "step": 9680
+    },
+    {
+      "epoch": 57.337278106508876,
+      "grad_norm": 2.1552584171295166,
+      "learning_rate": 0.0001,
+      "loss": 0.0327,
+      "step": 9690
+    },
+    {
+      "epoch": 57.396449704142015,
+      "grad_norm": 1.8134112358093262,
+      "learning_rate": 0.0001,
+      "loss": 0.0308,
+      "step": 9700
+    },
+    {
+      "epoch": 57.455621301775146,
+      "grad_norm": 2.241772413253784,
+      "learning_rate": 0.0001,
+      "loss": 0.0303,
+      "step": 9710
+    },
+    {
+      "epoch": 57.514792899408285,
+      "grad_norm": 1.8378987312316895,
+      "learning_rate": 0.0001,
+      "loss": 0.0299,
+      "step": 9720
+    },
+    {
+      "epoch": 57.57396449704142,
+      "grad_norm": 1.5637216567993164,
+      "learning_rate": 0.0001,
+      "loss": 0.0292,
+      "step": 9730
+    },
+    {
+      "epoch": 57.633136094674555,
+      "grad_norm": 1.909513235092163,
+      "learning_rate": 0.0001,
+      "loss": 0.0285,
+      "step": 9740
+    },
+    {
+      "epoch": 57.69230769230769,
+      "grad_norm": 1.9131604433059692,
+      "learning_rate": 0.0001,
+      "loss": 0.0303,
+      "step": 9750
+    },
+    {
+      "epoch": 57.751479289940825,
+      "grad_norm": 2.2477493286132812,
+      "learning_rate": 0.0001,
+      "loss": 0.0317,
+      "step": 9760
+    },
+    {
+      "epoch": 57.81065088757396,
+      "grad_norm": 1.8605237007141113,
+      "learning_rate": 0.0001,
+      "loss": 0.0312,
+      "step": 9770
+    },
+    {
+      "epoch": 57.8698224852071,
+      "grad_norm": 3.132593870162964,
+      "learning_rate": 0.0001,
+      "loss": 0.0322,
+      "step": 9780
+    },
+    {
+      "epoch": 57.928994082840234,
+      "grad_norm": 2.1441218852996826,
+      "learning_rate": 0.0001,
+      "loss": 0.0324,
+      "step": 9790
+    },
+    {
+      "epoch": 57.98816568047337,
+      "grad_norm": 1.7330306768417358,
+      "learning_rate": 0.0001,
+      "loss": 0.0321,
+      "step": 9800
+    },
+    {
+      "epoch": 58.04733727810651,
+      "grad_norm": 2.128164768218994,
+      "learning_rate": 0.0001,
+      "loss": 0.0317,
+      "step": 9810
+    },
+    {
+      "epoch": 58.10650887573964,
+      "grad_norm": 1.8707060813903809,
+      "learning_rate": 0.0001,
+      "loss": 0.0311,
+      "step": 9820
+    },
+    {
+      "epoch": 58.16568047337278,
+      "grad_norm": 1.5044137239456177,
+      "learning_rate": 0.0001,
+      "loss": 0.0301,
+      "step": 9830
+    },
+    {
+      "epoch": 58.22485207100592,
+      "grad_norm": 1.730989694595337,
+      "learning_rate": 0.0001,
+      "loss": 0.0303,
+      "step": 9840
+    },
+    {
+      "epoch": 58.28402366863905,
+      "grad_norm": 2.167001962661743,
+      "learning_rate": 0.0001,
+      "loss": 0.0321,
+      "step": 9850
+    },
+    {
+      "epoch": 58.34319526627219,
+      "grad_norm": 1.8830881118774414,
+      "learning_rate": 0.0001,
+      "loss": 0.0304,
+      "step": 9860
+    },
+    {
+      "epoch": 58.40236686390533,
+      "grad_norm": 1.767619013786316,
+      "learning_rate": 0.0001,
+      "loss": 0.0303,
+      "step": 9870
+    },
+    {
+      "epoch": 58.46153846153846,
+      "grad_norm": 2.2543301582336426,
+      "learning_rate": 0.0001,
+      "loss": 0.03,
+      "step": 9880
+    },
+    {
+      "epoch": 58.5207100591716,
+      "grad_norm": 1.6075273752212524,
+      "learning_rate": 0.0001,
+      "loss": 0.0302,
+      "step": 9890
+    },
+    {
+      "epoch": 58.57988165680474,
+      "grad_norm": 1.7080843448638916,
+      "learning_rate": 0.0001,
+      "loss": 0.0307,
+      "step": 9900
+    },
+    {
+      "epoch": 58.63905325443787,
+      "grad_norm": 2.2916605472564697,
+      "learning_rate": 0.0001,
+      "loss": 0.0322,
+      "step": 9910
+    },
+    {
+      "epoch": 58.69822485207101,
+      "grad_norm": 1.7076942920684814,
+      "learning_rate": 0.0001,
+      "loss": 0.0321,
+      "step": 9920
+    },
+    {
+      "epoch": 58.75739644970414,
+      "grad_norm": 1.5956099033355713,
+      "learning_rate": 0.0001,
+      "loss": 0.0311,
+      "step": 9930
+    },
+    {
+      "epoch": 58.81656804733728,
+      "grad_norm": 1.531049370765686,
+      "learning_rate": 0.0001,
+      "loss": 0.0307,
+      "step": 9940
+    },
+    {
+      "epoch": 58.875739644970416,
+      "grad_norm": 1.7461658716201782,
+      "learning_rate": 0.0001,
+      "loss": 0.0303,
+      "step": 9950
+    },
+    {
+      "epoch": 58.93491124260355,
+      "grad_norm": 1.395787239074707,
+      "learning_rate": 0.0001,
+      "loss": 0.0316,
+      "step": 9960
+    },
+    {
+      "epoch": 58.994082840236686,
+      "grad_norm": 2.1642537117004395,
+      "learning_rate": 0.0001,
+      "loss": 0.0321,
+      "step": 9970
+    },
+    {
+      "epoch": 59.053254437869825,
+      "grad_norm": 1.8260043859481812,
+      "learning_rate": 0.0001,
+      "loss": 0.031,
+      "step": 9980
+    },
+    {
+      "epoch": 59.112426035502956,
+      "grad_norm": 1.517285943031311,
+      "learning_rate": 0.0001,
+      "loss": 0.0317,
+      "step": 9990
+    },
+    {
+      "epoch": 59.171597633136095,
+      "grad_norm": 2.1005969047546387,
+      "learning_rate": 0.0001,
+      "loss": 0.0303,
+      "step": 10000
+    },
+    {
+      "epoch": 59.23076923076923,
+      "grad_norm": 2.2040884494781494,
+      "learning_rate": 0.0001,
+      "loss": 0.0317,
+      "step": 10010
+    },
+    {
+      "epoch": 59.289940828402365,
+      "grad_norm": 2.3694634437561035,
+      "learning_rate": 0.0001,
+      "loss": 0.0322,
+      "step": 10020
+    },
+    {
+      "epoch": 59.349112426035504,
+      "grad_norm": 1.7825241088867188,
+      "learning_rate": 0.0001,
+      "loss": 0.03,
+      "step": 10030
+    },
+    {
+      "epoch": 59.40828402366864,
+      "grad_norm": 1.8941017389297485,
+      "learning_rate": 0.0001,
+      "loss": 0.0314,
+      "step": 10040
+    },
+    {
+      "epoch": 59.467455621301774,
+      "grad_norm": 1.853210210800171,
+      "learning_rate": 0.0001,
+      "loss": 0.0301,
+      "step": 10050
+    },
+    {
+      "epoch": 59.52662721893491,
+      "grad_norm": 1.4500467777252197,
+      "learning_rate": 0.0001,
+      "loss": 0.0305,
+      "step": 10060
+    },
+    {
+      "epoch": 59.585798816568044,
+      "grad_norm": 1.8807293176651,
+      "learning_rate": 0.0001,
+      "loss": 0.0301,
+      "step": 10070
+    },
+    {
+      "epoch": 59.64497041420118,
+      "grad_norm": 1.8200446367263794,
+      "learning_rate": 0.0001,
+      "loss": 0.0294,
+      "step": 10080
+    },
+    {
+      "epoch": 59.70414201183432,
+      "grad_norm": 1.7710888385772705,
+      "learning_rate": 0.0001,
+      "loss": 0.0297,
+      "step": 10090
+    },
+    {
+      "epoch": 59.76331360946745,
+      "grad_norm": 2.0216760635375977,
+      "learning_rate": 0.0001,
+      "loss": 0.03,
+      "step": 10100
+    },
+    {
+      "epoch": 59.82248520710059,
+      "grad_norm": 1.6908553838729858,
+      "learning_rate": 0.0001,
+      "loss": 0.0309,
+      "step": 10110
+    },
+    {
+      "epoch": 59.88165680473373,
+      "grad_norm": 1.617326021194458,
+      "learning_rate": 0.0001,
+      "loss": 0.0296,
+      "step": 10120
+    },
+    {
+      "epoch": 59.94082840236686,
+      "grad_norm": 1.7559802532196045,
+      "learning_rate": 0.0001,
+      "loss": 0.0303,
+      "step": 10130
+    },
+    {
+      "epoch": 60.0,
+      "grad_norm": 1.937020182609558,
+      "learning_rate": 0.0001,
+      "loss": 0.031,
+      "step": 10140
+    },
+    {
+      "epoch": 60.05917159763314,
+      "grad_norm": 1.6620839834213257,
+      "learning_rate": 0.0001,
+      "loss": 0.0305,
+      "step": 10150
+    },
+    {
+      "epoch": 60.11834319526627,
+      "grad_norm": 1.5040069818496704,
+      "learning_rate": 0.0001,
+      "loss": 0.0292,
+      "step": 10160
+    },
+    {
+      "epoch": 60.17751479289941,
+      "grad_norm": 2.0442278385162354,
+      "learning_rate": 0.0001,
+      "loss": 0.0298,
+      "step": 10170
+    },
+    {
+      "epoch": 60.23668639053255,
+      "grad_norm": 1.837834119796753,
+      "learning_rate": 0.0001,
+      "loss": 0.0286,
+      "step": 10180
+    },
+    {
+      "epoch": 60.29585798816568,
+      "grad_norm": 1.5686897039413452,
+      "learning_rate": 0.0001,
+      "loss": 0.0303,
+      "step": 10190
+    },
+    {
+      "epoch": 60.35502958579882,
+      "grad_norm": 1.7981812953948975,
+      "learning_rate": 0.0001,
+      "loss": 0.0298,
+      "step": 10200
+    },
+    {
+      "epoch": 60.414201183431956,
+      "grad_norm": 2.020423412322998,
+      "learning_rate": 0.0001,
+      "loss": 0.0311,
+      "step": 10210
+    },
+    {
+      "epoch": 60.47337278106509,
+      "grad_norm": 1.9064350128173828,
+      "learning_rate": 0.0001,
+      "loss": 0.0313,
+      "step": 10220
+    },
+    {
+      "epoch": 60.532544378698226,
+      "grad_norm": 1.4197514057159424,
+      "learning_rate": 0.0001,
+      "loss": 0.0314,
+      "step": 10230
+    },
+    {
+      "epoch": 60.59171597633136,
+      "grad_norm": 1.9174798727035522,
+      "learning_rate": 0.0001,
+      "loss": 0.0313,
+      "step": 10240
+    },
+    {
+      "epoch": 60.650887573964496,
+      "grad_norm": 1.8817318677902222,
+      "learning_rate": 0.0001,
+      "loss": 0.0311,
+      "step": 10250
+    },
+    {
+      "epoch": 60.710059171597635,
+      "grad_norm": 1.7899407148361206,
+      "learning_rate": 0.0001,
+      "loss": 0.0309,
+      "step": 10260
+    },
+    {
+      "epoch": 60.76923076923077,
+      "grad_norm": 2.3800339698791504,
+      "learning_rate": 0.0001,
+      "loss": 0.0318,
+      "step": 10270
+    },
+    {
+      "epoch": 60.828402366863905,
+      "grad_norm": 1.796679973602295,
+      "learning_rate": 0.0001,
+      "loss": 0.0321,
+      "step": 10280
+    },
+    {
+      "epoch": 60.887573964497044,
+      "grad_norm": 2.0754497051239014,
+      "learning_rate": 0.0001,
+      "loss": 0.0307,
+      "step": 10290
+    },
+    {
+      "epoch": 60.946745562130175,
+      "grad_norm": 2.204084634780884,
+      "learning_rate": 0.0001,
+      "loss": 0.0312,
+      "step": 10300
+    },
+    {
+      "epoch": 61.005917159763314,
+      "grad_norm": 2.0353293418884277,
+      "learning_rate": 0.0001,
+      "loss": 0.0306,
+      "step": 10310
+    },
+    {
+      "epoch": 61.06508875739645,
+      "grad_norm": 2.3918139934539795,
+      "learning_rate": 0.0001,
+      "loss": 0.0306,
+      "step": 10320
+    },
+    {
+      "epoch": 61.124260355029584,
+      "grad_norm": 2.322150230407715,
+      "learning_rate": 0.0001,
+      "loss": 0.0315,
+      "step": 10330
+    },
+    {
+      "epoch": 61.18343195266272,
+      "grad_norm": 1.9441698789596558,
+      "learning_rate": 0.0001,
+      "loss": 0.0307,
+      "step": 10340
+    },
+    {
+      "epoch": 61.24260355029586,
+      "grad_norm": 1.91065514087677,
+      "learning_rate": 0.0001,
+      "loss": 0.03,
+      "step": 10350
+    },
+    {
+      "epoch": 61.30177514792899,
+      "grad_norm": 1.8158997297286987,
+      "learning_rate": 0.0001,
+      "loss": 0.0294,
+      "step": 10360
+    },
+    {
+      "epoch": 61.36094674556213,
+      "grad_norm": 1.760204792022705,
+      "learning_rate": 0.0001,
+      "loss": 0.0297,
+      "step": 10370
+    },
+    {
+      "epoch": 61.42011834319526,
+      "grad_norm": 1.5322269201278687,
+      "learning_rate": 0.0001,
+      "loss": 0.0285,
+      "step": 10380
+    },
+    {
+      "epoch": 61.4792899408284,
+      "grad_norm": 1.5393898487091064,
+      "learning_rate": 0.0001,
+      "loss": 0.0294,
+      "step": 10390
+    },
+    {
+      "epoch": 61.53846153846154,
+      "grad_norm": 1.9503962993621826,
+      "learning_rate": 0.0001,
+      "loss": 0.0298,
+      "step": 10400
+    },
+    {
+      "epoch": 61.59763313609467,
+      "grad_norm": 1.5803238153457642,
+      "learning_rate": 0.0001,
+      "loss": 0.0291,
+      "step": 10410
+    },
+    {
+      "epoch": 61.65680473372781,
+      "grad_norm": 2.3723838329315186,
+      "learning_rate": 0.0001,
+      "loss": 0.0291,
+      "step": 10420
+    },
+    {
+      "epoch": 61.71597633136095,
+      "grad_norm": 1.6206157207489014,
+      "learning_rate": 0.0001,
+      "loss": 0.0299,
+      "step": 10430
+    },
+    {
+      "epoch": 61.77514792899408,
+      "grad_norm": 1.570833683013916,
+      "learning_rate": 0.0001,
+      "loss": 0.0292,
+      "step": 10440
+    },
+    {
+      "epoch": 61.83431952662722,
+      "grad_norm": 1.5050595998764038,
+      "learning_rate": 0.0001,
+      "loss": 0.03,
+      "step": 10450
+    },
+    {
+      "epoch": 61.89349112426036,
+      "grad_norm": 1.9845263957977295,
+      "learning_rate": 0.0001,
+      "loss": 0.0301,
+      "step": 10460
+    },
+    {
+      "epoch": 61.95266272189349,
+      "grad_norm": 1.530484676361084,
+      "learning_rate": 0.0001,
+      "loss": 0.0296,
+      "step": 10470
+    },
+    {
+      "epoch": 62.01183431952663,
+      "grad_norm": 1.745964527130127,
+      "learning_rate": 0.0001,
+      "loss": 0.0302,
+      "step": 10480
+    },
+    {
+      "epoch": 62.071005917159766,
+      "grad_norm": 2.2849202156066895,
+      "learning_rate": 0.0001,
+      "loss": 0.0305,
+      "step": 10490
+    },
+    {
+      "epoch": 62.1301775147929,
+      "grad_norm": 1.5362452268600464,
+      "learning_rate": 0.0001,
+      "loss": 0.0306,
+      "step": 10500
+    },
+    {
+      "epoch": 62.18934911242604,
+      "grad_norm": 2.142918825149536,
+      "learning_rate": 0.0001,
+      "loss": 0.0304,
+      "step": 10510
+    },
+    {
+      "epoch": 62.248520710059175,
+      "grad_norm": 2.536658525466919,
+      "learning_rate": 0.0001,
+      "loss": 0.0315,
+      "step": 10520
+    },
+    {
+      "epoch": 62.30769230769231,
+      "grad_norm": 2.1485326290130615,
+      "learning_rate": 0.0001,
+      "loss": 0.0298,
+      "step": 10530
+    },
+    {
+      "epoch": 62.366863905325445,
+      "grad_norm": 1.751765489578247,
+      "learning_rate": 0.0001,
+      "loss": 0.0279,
+      "step": 10540
+    },
+    {
+      "epoch": 62.42603550295858,
+      "grad_norm": 1.6838475465774536,
+      "learning_rate": 0.0001,
+      "loss": 0.028,
+      "step": 10550
+    },
+    {
+      "epoch": 62.485207100591715,
+      "grad_norm": 2.1659936904907227,
+      "learning_rate": 0.0001,
+      "loss": 0.0292,
+      "step": 10560
+    },
+    {
+      "epoch": 62.544378698224854,
+      "grad_norm": 1.8107887506484985,
+      "learning_rate": 0.0001,
+      "loss": 0.0298,
+      "step": 10570
+    },
+    {
+      "epoch": 62.603550295857985,
+      "grad_norm": 1.564025640487671,
+      "learning_rate": 0.0001,
+      "loss": 0.0287,
+      "step": 10580
+    },
+    {
+      "epoch": 62.662721893491124,
+      "grad_norm": 1.7257368564605713,
+      "learning_rate": 0.0001,
+      "loss": 0.029,
+      "step": 10590
+    },
+    {
+      "epoch": 62.72189349112426,
+      "grad_norm": 1.7080910205841064,
+      "learning_rate": 0.0001,
+      "loss": 0.029,
+      "step": 10600
+    },
+    {
+      "epoch": 62.781065088757394,
+      "grad_norm": 2.3297119140625,
+      "learning_rate": 0.0001,
+      "loss": 0.0301,
+      "step": 10610
+    },
+    {
+      "epoch": 62.84023668639053,
+      "grad_norm": 1.6523752212524414,
+      "learning_rate": 0.0001,
+      "loss": 0.029,
+      "step": 10620
+    },
+    {
+      "epoch": 62.89940828402367,
+      "grad_norm": 1.492505669593811,
+      "learning_rate": 0.0001,
+      "loss": 0.0298,
+      "step": 10630
+    },
+    {
+      "epoch": 62.9585798816568,
+      "grad_norm": 1.9790211915969849,
+      "learning_rate": 0.0001,
+      "loss": 0.0297,
+      "step": 10640
+    },
+    {
+      "epoch": 63.01775147928994,
+      "grad_norm": 2.2688260078430176,
+      "learning_rate": 0.0001,
+      "loss": 0.0297,
+      "step": 10650
+    },
+    {
+      "epoch": 63.07692307692308,
+      "grad_norm": 1.4216831922531128,
+      "learning_rate": 0.0001,
+      "loss": 0.0301,
+      "step": 10660
+    },
+    {
+      "epoch": 63.13609467455621,
+      "grad_norm": 1.888412356376648,
+      "learning_rate": 0.0001,
+      "loss": 0.0293,
+      "step": 10670
+    },
+    {
+      "epoch": 63.19526627218935,
+      "grad_norm": 1.9707571268081665,
+      "learning_rate": 0.0001,
+      "loss": 0.0287,
+      "step": 10680
+    },
+    {
+      "epoch": 63.25443786982248,
+      "grad_norm": 2.1282002925872803,
+      "learning_rate": 0.0001,
+      "loss": 0.0296,
+      "step": 10690
+    },
+    {
+      "epoch": 63.31360946745562,
+      "grad_norm": 1.989397406578064,
+      "learning_rate": 0.0001,
+      "loss": 0.0297,
+      "step": 10700
+    },
+    {
+      "epoch": 63.37278106508876,
+      "grad_norm": 1.7101837396621704,
+      "learning_rate": 0.0001,
+      "loss": 0.03,
+      "step": 10710
+    },
+    {
+      "epoch": 63.43195266272189,
+      "grad_norm": 1.3152800798416138,
+      "learning_rate": 0.0001,
+      "loss": 0.0294,
+      "step": 10720
+    },
+    {
+      "epoch": 63.49112426035503,
+      "grad_norm": 1.9535712003707886,
+      "learning_rate": 0.0001,
+      "loss": 0.0315,
+      "step": 10730
+    },
+    {
+      "epoch": 63.55029585798817,
+      "grad_norm": 2.093728542327881,
+      "learning_rate": 0.0001,
+      "loss": 0.029,
+      "step": 10740
+    },
+    {
+      "epoch": 63.6094674556213,
+      "grad_norm": 1.6733689308166504,
+      "learning_rate": 0.0001,
+      "loss": 0.028,
+      "step": 10750
+    },
+    {
+      "epoch": 63.66863905325444,
+      "grad_norm": 1.26914381980896,
+      "learning_rate": 0.0001,
+      "loss": 0.0291,
+      "step": 10760
+    },
+    {
+      "epoch": 63.72781065088758,
+      "grad_norm": 2.2817811965942383,
+      "learning_rate": 0.0001,
+      "loss": 0.0296,
+      "step": 10770
+    },
+    {
+      "epoch": 63.78698224852071,
+      "grad_norm": 2.209808349609375,
+      "learning_rate": 0.0001,
+      "loss": 0.0294,
+      "step": 10780
+    },
+    {
+      "epoch": 63.84615384615385,
+      "grad_norm": 2.3054471015930176,
+      "learning_rate": 0.0001,
+      "loss": 0.0307,
+      "step": 10790
+    },
+    {
+      "epoch": 63.905325443786985,
+      "grad_norm": 2.111701250076294,
+      "learning_rate": 0.0001,
+      "loss": 0.029,
+      "step": 10800
+    },
+    {
+      "epoch": 63.96449704142012,
+      "grad_norm": 1.9286569356918335,
+      "learning_rate": 0.0001,
+      "loss": 0.0299,
+      "step": 10810
+    },
+    {
+      "epoch": 64.02366863905326,
+      "grad_norm": 2.6353023052215576,
+      "learning_rate": 0.0001,
+      "loss": 0.0305,
+      "step": 10820
+    },
+    {
+      "epoch": 64.0828402366864,
+      "grad_norm": 2.0067996978759766,
+      "learning_rate": 0.0001,
+      "loss": 0.0293,
+      "step": 10830
+    },
+    {
+      "epoch": 64.14201183431953,
+      "grad_norm": 2.065363883972168,
+      "learning_rate": 0.0001,
+      "loss": 0.0281,
+      "step": 10840
+    },
+    {
+      "epoch": 64.20118343195266,
+      "grad_norm": 1.8264496326446533,
+      "learning_rate": 0.0001,
+      "loss": 0.0274,
+      "step": 10850
+    },
+    {
+      "epoch": 64.2603550295858,
+      "grad_norm": 1.8948506116867065,
+      "learning_rate": 0.0001,
+      "loss": 0.0275,
+      "step": 10860
+    },
+    {
+      "epoch": 64.31952662721893,
+      "grad_norm": 1.6936414241790771,
+      "learning_rate": 0.0001,
+      "loss": 0.0288,
+      "step": 10870
+    },
+    {
+      "epoch": 64.37869822485207,
+      "grad_norm": 2.1033096313476562,
+      "learning_rate": 0.0001,
+      "loss": 0.0277,
+      "step": 10880
+    },
+    {
+      "epoch": 64.43786982248521,
+      "grad_norm": 1.9560198783874512,
+      "learning_rate": 0.0001,
+      "loss": 0.028,
+      "step": 10890
+    },
+    {
+      "epoch": 64.49704142011835,
+      "grad_norm": 1.9866431951522827,
+      "learning_rate": 0.0001,
+      "loss": 0.0294,
+      "step": 10900
+    },
+    {
+      "epoch": 64.55621301775147,
+      "grad_norm": 1.7387144565582275,
+      "learning_rate": 0.0001,
+      "loss": 0.0298,
+      "step": 10910
+    },
+    {
+      "epoch": 64.61538461538461,
+      "grad_norm": 2.0249991416931152,
+      "learning_rate": 0.0001,
+      "loss": 0.0293,
+      "step": 10920
+    },
+    {
+      "epoch": 64.67455621301775,
+      "grad_norm": 1.9182769060134888,
+      "learning_rate": 0.0001,
+      "loss": 0.0284,
+      "step": 10930
+    },
+    {
+      "epoch": 64.73372781065089,
+      "grad_norm": 1.7350016832351685,
+      "learning_rate": 0.0001,
+      "loss": 0.029,
+      "step": 10940
+    },
+    {
+      "epoch": 64.79289940828403,
+      "grad_norm": 1.5628937482833862,
+      "learning_rate": 0.0001,
+      "loss": 0.0287,
+      "step": 10950
+    },
+    {
+      "epoch": 64.85207100591715,
+      "grad_norm": 1.7418267726898193,
+      "learning_rate": 0.0001,
+      "loss": 0.0289,
+      "step": 10960
+    },
+    {
+      "epoch": 64.91124260355029,
+      "grad_norm": 1.9856709241867065,
+      "learning_rate": 0.0001,
+      "loss": 0.0288,
+      "step": 10970
+    },
+    {
+      "epoch": 64.97041420118343,
+      "grad_norm": 2.2123005390167236,
+      "learning_rate": 0.0001,
+      "loss": 0.0301,
+      "step": 10980
+    },
+    {
+      "epoch": 65.02958579881657,
+      "grad_norm": 1.843113899230957,
+      "learning_rate": 0.0001,
+      "loss": 0.0295,
+      "step": 10990
+    },
+    {
+      "epoch": 65.08875739644971,
+      "grad_norm": 1.688407063484192,
+      "learning_rate": 0.0001,
+      "loss": 0.0283,
+      "step": 11000
+    },
+    {
+      "epoch": 65.14792899408285,
+      "grad_norm": 1.744457483291626,
+      "learning_rate": 0.0001,
+      "loss": 0.0283,
+      "step": 11010
+    },
+    {
+      "epoch": 65.20710059171597,
+      "grad_norm": 1.9946364164352417,
+      "learning_rate": 0.0001,
+      "loss": 0.0287,
+      "step": 11020
+    },
+    {
+      "epoch": 65.26627218934911,
+      "grad_norm": 1.726738452911377,
+      "learning_rate": 0.0001,
+      "loss": 0.0282,
+      "step": 11030
+    },
+    {
+      "epoch": 65.32544378698225,
+      "grad_norm": 2.118312358856201,
+      "learning_rate": 0.0001,
+      "loss": 0.0291,
+      "step": 11040
+    },
+    {
+      "epoch": 65.38461538461539,
+      "grad_norm": 2.1961419582366943,
+      "learning_rate": 0.0001,
+      "loss": 0.0295,
+      "step": 11050
+    },
+    {
+      "epoch": 65.44378698224853,
+      "grad_norm": 1.7672499418258667,
+      "learning_rate": 0.0001,
+      "loss": 0.0287,
+      "step": 11060
+    },
+    {
+      "epoch": 65.50295857988165,
+      "grad_norm": 1.9085171222686768,
+      "learning_rate": 0.0001,
+      "loss": 0.0274,
+      "step": 11070
+    },
+    {
+      "epoch": 65.56213017751479,
+      "grad_norm": 2.597341775894165,
+      "learning_rate": 0.0001,
+      "loss": 0.0282,
+      "step": 11080
+    },
+    {
+      "epoch": 65.62130177514793,
+      "grad_norm": 2.0319433212280273,
+      "learning_rate": 0.0001,
+      "loss": 0.0282,
+      "step": 11090
+    },
+    {
+      "epoch": 65.68047337278107,
+      "grad_norm": 1.604514718055725,
+      "learning_rate": 0.0001,
+      "loss": 0.0282,
+      "step": 11100
+    },
+    {
+      "epoch": 65.7396449704142,
+      "grad_norm": 2.253392457962036,
+      "learning_rate": 0.0001,
+      "loss": 0.0276,
+      "step": 11110
+    },
+    {
+      "epoch": 65.79881656804734,
+      "grad_norm": 1.8604778051376343,
+      "learning_rate": 0.0001,
+      "loss": 0.0294,
+      "step": 11120
+    },
+    {
+      "epoch": 65.85798816568047,
+      "grad_norm": 1.6478239297866821,
+      "learning_rate": 0.0001,
+      "loss": 0.0276,
+      "step": 11130
+    },
+    {
+      "epoch": 65.9171597633136,
+      "grad_norm": 1.869320273399353,
+      "learning_rate": 0.0001,
+      "loss": 0.028,
+      "step": 11140
+    },
+    {
+      "epoch": 65.97633136094674,
+      "grad_norm": 2.4263553619384766,
+      "learning_rate": 0.0001,
+      "loss": 0.0293,
+      "step": 11150
+    },
+    {
+      "epoch": 66.03550295857988,
+      "grad_norm": 2.1131489276885986,
+      "learning_rate": 0.0001,
+      "loss": 0.0287,
+      "step": 11160
+    },
+    {
+      "epoch": 66.09467455621302,
+      "grad_norm": 2.0195040702819824,
+      "learning_rate": 0.0001,
+      "loss": 0.0282,
+      "step": 11170
+    },
+    {
+      "epoch": 66.15384615384616,
+      "grad_norm": 1.8548766374588013,
+      "learning_rate": 0.0001,
+      "loss": 0.028,
+      "step": 11180
+    },
+    {
+      "epoch": 66.21301775147928,
+      "grad_norm": 1.9463905096054077,
+      "learning_rate": 0.0001,
+      "loss": 0.0283,
+      "step": 11190
+    },
+    {
+      "epoch": 66.27218934911242,
+      "grad_norm": 1.7676604986190796,
+      "learning_rate": 0.0001,
+      "loss": 0.0274,
+      "step": 11200
+    },
+    {
+      "epoch": 66.33136094674556,
+      "grad_norm": 1.5998543500900269,
+      "learning_rate": 0.0001,
+      "loss": 0.028,
+      "step": 11210
+    },
+    {
+      "epoch": 66.3905325443787,
+      "grad_norm": 2.0161960124969482,
+      "learning_rate": 0.0001,
+      "loss": 0.0285,
+      "step": 11220
+    },
+    {
+      "epoch": 66.44970414201184,
+      "grad_norm": 1.6452523469924927,
+      "learning_rate": 0.0001,
+      "loss": 0.0278,
+      "step": 11230
+    },
+    {
+      "epoch": 66.50887573964496,
+      "grad_norm": 1.8008826971054077,
+      "learning_rate": 0.0001,
+      "loss": 0.0289,
+      "step": 11240
+    },
+    {
+      "epoch": 66.5680473372781,
+      "grad_norm": 2.5364344120025635,
+      "learning_rate": 0.0001,
+      "loss": 0.0297,
+      "step": 11250
+    },
+    {
+      "epoch": 66.62721893491124,
+      "grad_norm": 2.1963210105895996,
+      "learning_rate": 0.0001,
+      "loss": 0.0292,
+      "step": 11260
+    },
+    {
+      "epoch": 66.68639053254438,
+      "grad_norm": 2.3213887214660645,
+      "learning_rate": 0.0001,
+      "loss": 0.028,
+      "step": 11270
+    },
+    {
+      "epoch": 66.74556213017752,
+      "grad_norm": 1.68406343460083,
+      "learning_rate": 0.0001,
+      "loss": 0.0287,
+      "step": 11280
+    },
+    {
+      "epoch": 66.80473372781066,
+      "grad_norm": 1.7630653381347656,
+      "learning_rate": 0.0001,
+      "loss": 0.0267,
+      "step": 11290
+    },
+    {
+      "epoch": 66.86390532544378,
+      "grad_norm": 1.6980582475662231,
+      "learning_rate": 0.0001,
+      "loss": 0.0277,
+      "step": 11300
+    },
+    {
+      "epoch": 66.92307692307692,
+      "grad_norm": 1.7737339735031128,
+      "learning_rate": 0.0001,
+      "loss": 0.0277,
+      "step": 11310
+    },
+    {
+      "epoch": 66.98224852071006,
+      "grad_norm": 1.6511731147766113,
+      "learning_rate": 0.0001,
+      "loss": 0.0268,
+      "step": 11320
+    },
+    {
+      "epoch": 67.0414201183432,
+      "grad_norm": 1.6685658693313599,
+      "learning_rate": 0.0001,
+      "loss": 0.0285,
+      "step": 11330
+    },
+    {
+      "epoch": 67.10059171597634,
+      "grad_norm": 2.1383092403411865,
+      "learning_rate": 0.0001,
+      "loss": 0.0297,
+      "step": 11340
+    },
+    {
+      "epoch": 67.15976331360947,
+      "grad_norm": 1.8411238193511963,
+      "learning_rate": 0.0001,
+      "loss": 0.0283,
+      "step": 11350
+    },
+    {
+      "epoch": 67.2189349112426,
+      "grad_norm": 2.367124080657959,
+      "learning_rate": 0.0001,
+      "loss": 0.028,
+      "step": 11360
+    },
+    {
+      "epoch": 67.27810650887574,
+      "grad_norm": 2.333242893218994,
+      "learning_rate": 0.0001,
+      "loss": 0.0303,
+      "step": 11370
+    },
+    {
+      "epoch": 67.33727810650888,
+      "grad_norm": 2.2461860179901123,
+      "learning_rate": 0.0001,
+      "loss": 0.0294,
+      "step": 11380
+    },
+    {
+      "epoch": 67.39644970414201,
+      "grad_norm": 2.34466814994812,
+      "learning_rate": 0.0001,
+      "loss": 0.0292,
+      "step": 11390
+    },
+    {
+      "epoch": 67.45562130177515,
+      "grad_norm": 2.1762726306915283,
+      "learning_rate": 0.0001,
+      "loss": 0.0291,
+      "step": 11400
+    },
+    {
+      "epoch": 67.51479289940828,
+      "grad_norm": 2.630364418029785,
+      "learning_rate": 0.0001,
+      "loss": 0.0307,
+      "step": 11410
+    },
+    {
+      "epoch": 67.57396449704142,
+      "grad_norm": 2.02215576171875,
+      "learning_rate": 0.0001,
+      "loss": 0.0286,
+      "step": 11420
+    },
+    {
+      "epoch": 67.63313609467455,
+      "grad_norm": 1.8779592514038086,
+      "learning_rate": 0.0001,
+      "loss": 0.0278,
+      "step": 11430
+    },
+    {
+      "epoch": 67.6923076923077,
+      "grad_norm": 2.0115387439727783,
+      "learning_rate": 0.0001,
+      "loss": 0.0282,
+      "step": 11440
+    },
+    {
+      "epoch": 67.75147928994083,
+      "grad_norm": 1.8664910793304443,
+      "learning_rate": 0.0001,
+      "loss": 0.0278,
+      "step": 11450
+    },
+    {
+      "epoch": 67.81065088757397,
+      "grad_norm": 1.7702056169509888,
+      "learning_rate": 0.0001,
+      "loss": 0.0288,
+      "step": 11460
+    },
+    {
+      "epoch": 67.8698224852071,
+      "grad_norm": 1.6739057302474976,
+      "learning_rate": 0.0001,
+      "loss": 0.0283,
+      "step": 11470
+    },
+    {
+      "epoch": 67.92899408284023,
+      "grad_norm": 1.966712474822998,
+      "learning_rate": 0.0001,
+      "loss": 0.0283,
+      "step": 11480
+    },
+    {
+      "epoch": 67.98816568047337,
+      "grad_norm": 1.8273341655731201,
+      "learning_rate": 0.0001,
+      "loss": 0.0289,
+      "step": 11490
+    },
+    {
+      "epoch": 68.04733727810651,
+      "grad_norm": 1.8917275667190552,
+      "learning_rate": 0.0001,
+      "loss": 0.0273,
+      "step": 11500
+    },
+    {
+      "epoch": 68.10650887573965,
+      "grad_norm": 1.7940802574157715,
+      "learning_rate": 0.0001,
+      "loss": 0.0283,
+      "step": 11510
+    },
+    {
+      "epoch": 68.16568047337279,
+      "grad_norm": 1.8613991737365723,
+      "learning_rate": 0.0001,
+      "loss": 0.0292,
+      "step": 11520
+    },
+    {
+      "epoch": 68.22485207100591,
+      "grad_norm": 1.955140233039856,
+      "learning_rate": 0.0001,
+      "loss": 0.0289,
+      "step": 11530
+    },
+    {
+      "epoch": 68.28402366863905,
+      "grad_norm": 1.8209688663482666,
+      "learning_rate": 0.0001,
+      "loss": 0.0284,
+      "step": 11540
+    },
+    {
+      "epoch": 68.34319526627219,
+      "grad_norm": 2.5242557525634766,
+      "learning_rate": 0.0001,
+      "loss": 0.0296,
+      "step": 11550
+    },
+    {
+      "epoch": 68.40236686390533,
+      "grad_norm": 1.6232807636260986,
+      "learning_rate": 0.0001,
+      "loss": 0.0294,
+      "step": 11560
+    },
+    {
+      "epoch": 68.46153846153847,
+      "grad_norm": 2.3849024772644043,
+      "learning_rate": 0.0001,
+      "loss": 0.0295,
+      "step": 11570
+    },
+    {
+      "epoch": 68.52071005917159,
+      "grad_norm": 1.785730004310608,
+      "learning_rate": 0.0001,
+      "loss": 0.0297,
+      "step": 11580
+    },
+    {
+      "epoch": 68.57988165680473,
+      "grad_norm": 1.8025121688842773,
+      "learning_rate": 0.0001,
+      "loss": 0.0274,
+      "step": 11590
+    },
+    {
+      "epoch": 68.63905325443787,
+      "grad_norm": 1.651186466217041,
+      "learning_rate": 0.0001,
+      "loss": 0.028,
+      "step": 11600
+    },
+    {
+      "epoch": 68.69822485207101,
+      "grad_norm": 2.1935863494873047,
+      "learning_rate": 0.0001,
+      "loss": 0.0269,
+      "step": 11610
+    },
+    {
+      "epoch": 68.75739644970415,
+      "grad_norm": 1.6546251773834229,
+      "learning_rate": 0.0001,
+      "loss": 0.0288,
+      "step": 11620
+    },
+    {
+      "epoch": 68.81656804733728,
+      "grad_norm": 2.152904748916626,
+      "learning_rate": 0.0001,
+      "loss": 0.0275,
+      "step": 11630
+    },
+    {
+      "epoch": 68.87573964497041,
+      "grad_norm": 1.8369593620300293,
+      "learning_rate": 0.0001,
+      "loss": 0.0286,
+      "step": 11640
+    },
+    {
+      "epoch": 68.93491124260355,
+      "grad_norm": 1.7649458646774292,
+      "learning_rate": 0.0001,
+      "loss": 0.0291,
+      "step": 11650
+    },
+    {
+      "epoch": 68.99408284023669,
+      "grad_norm": 1.5516008138656616,
+      "learning_rate": 0.0001,
+      "loss": 0.0282,
+      "step": 11660
+    },
+    {
+      "epoch": 69.05325443786982,
+      "grad_norm": 1.7939095497131348,
+      "learning_rate": 0.0001,
+      "loss": 0.0278,
+      "step": 11670
+    },
+    {
+      "epoch": 69.11242603550296,
+      "grad_norm": 1.5844287872314453,
+      "learning_rate": 0.0001,
+      "loss": 0.0275,
+      "step": 11680
+    },
+    {
+      "epoch": 69.17159763313609,
+      "grad_norm": 2.040956974029541,
+      "learning_rate": 0.0001,
+      "loss": 0.0278,
+      "step": 11690
+    },
+    {
+      "epoch": 69.23076923076923,
+      "grad_norm": 1.7223420143127441,
+      "learning_rate": 0.0001,
+      "loss": 0.0281,
+      "step": 11700
+    },
+    {
+      "epoch": 69.28994082840237,
+      "grad_norm": 1.6375782489776611,
+      "learning_rate": 0.0001,
+      "loss": 0.0284,
+      "step": 11710
+    },
+    {
+      "epoch": 69.3491124260355,
+      "grad_norm": 2.252098560333252,
+      "learning_rate": 0.0001,
+      "loss": 0.0281,
+      "step": 11720
+    },
+    {
+      "epoch": 69.40828402366864,
+      "grad_norm": 1.7104015350341797,
+      "learning_rate": 0.0001,
+      "loss": 0.0277,
+      "step": 11730
+    },
+    {
+      "epoch": 69.46745562130178,
+      "grad_norm": 1.1876802444458008,
+      "learning_rate": 0.0001,
+      "loss": 0.0277,
+      "step": 11740
+    },
+    {
+      "epoch": 69.5266272189349,
+      "grad_norm": 1.42478346824646,
+      "learning_rate": 0.0001,
+      "loss": 0.0271,
+      "step": 11750
+    },
+    {
+      "epoch": 69.58579881656804,
+      "grad_norm": 1.5924077033996582,
+      "learning_rate": 0.0001,
+      "loss": 0.0276,
+      "step": 11760
+    },
+    {
+      "epoch": 69.64497041420118,
+      "grad_norm": 1.853326678276062,
+      "learning_rate": 0.0001,
+      "loss": 0.0288,
+      "step": 11770
+    },
+    {
+      "epoch": 69.70414201183432,
+      "grad_norm": 1.8595331907272339,
+      "learning_rate": 0.0001,
+      "loss": 0.0283,
+      "step": 11780
+    },
+    {
+      "epoch": 69.76331360946746,
+      "grad_norm": 1.6751657724380493,
+      "learning_rate": 0.0001,
+      "loss": 0.0282,
+      "step": 11790
+    },
+    {
+      "epoch": 69.8224852071006,
+      "grad_norm": 2.517115831375122,
+      "learning_rate": 0.0001,
+      "loss": 0.0307,
+      "step": 11800
+    },
+    {
+      "epoch": 69.88165680473372,
+      "grad_norm": 1.4701077938079834,
+      "learning_rate": 0.0001,
+      "loss": 0.0309,
+      "step": 11810
+    },
+    {
+      "epoch": 69.94082840236686,
+      "grad_norm": 2.1840176582336426,
+      "learning_rate": 0.0001,
+      "loss": 0.0316,
+      "step": 11820
+    },
+    {
+      "epoch": 70.0,
+      "grad_norm": 2.0750036239624023,
+      "learning_rate": 0.0001,
+      "loss": 0.0297,
+      "step": 11830
+    },
+    {
+      "epoch": 70.05917159763314,
+      "grad_norm": 1.7510558366775513,
+      "learning_rate": 0.0001,
+      "loss": 0.0301,
+      "step": 11840
+    },
+    {
+      "epoch": 70.11834319526628,
+      "grad_norm": 2.690769672393799,
+      "learning_rate": 0.0001,
+      "loss": 0.0293,
+      "step": 11850
+    },
+    {
+      "epoch": 70.1775147928994,
+      "grad_norm": 1.9595177173614502,
+      "learning_rate": 0.0001,
+      "loss": 0.0287,
+      "step": 11860
+    },
+    {
+      "epoch": 70.23668639053254,
+      "grad_norm": 1.6515610218048096,
+      "learning_rate": 0.0001,
+      "loss": 0.0281,
+      "step": 11870
+    },
+    {
+      "epoch": 70.29585798816568,
+      "grad_norm": 2.319464921951294,
+      "learning_rate": 0.0001,
+      "loss": 0.028,
+      "step": 11880
+    },
+    {
+      "epoch": 70.35502958579882,
+      "grad_norm": 2.295499324798584,
+      "learning_rate": 0.0001,
+      "loss": 0.0276,
+      "step": 11890
+    },
+    {
+      "epoch": 70.41420118343196,
+      "grad_norm": 2.1146726608276367,
+      "learning_rate": 0.0001,
+      "loss": 0.0278,
+      "step": 11900
+    },
+    {
+      "epoch": 70.4733727810651,
+      "grad_norm": 1.6517279148101807,
+      "learning_rate": 0.0001,
+      "loss": 0.0277,
+      "step": 11910
+    },
+    {
+      "epoch": 70.53254437869822,
+      "grad_norm": 1.474465250968933,
+      "learning_rate": 0.0001,
+      "loss": 0.027,
+      "step": 11920
+    },
+    {
+      "epoch": 70.59171597633136,
+      "grad_norm": 2.1878929138183594,
+      "learning_rate": 0.0001,
+      "loss": 0.0297,
+      "step": 11930
+    },
+    {
+      "epoch": 70.6508875739645,
+      "grad_norm": 1.8438724279403687,
+      "learning_rate": 0.0001,
+      "loss": 0.0289,
+      "step": 11940
+    },
+    {
+      "epoch": 70.71005917159763,
+      "grad_norm": 1.6197317838668823,
+      "learning_rate": 0.0001,
+      "loss": 0.0277,
+      "step": 11950
+    },
+    {
+      "epoch": 70.76923076923077,
+      "grad_norm": 1.9095301628112793,
+      "learning_rate": 0.0001,
+      "loss": 0.0289,
+      "step": 11960
+    },
+    {
+      "epoch": 70.82840236686391,
+      "grad_norm": 1.5303864479064941,
+      "learning_rate": 0.0001,
+      "loss": 0.0287,
+      "step": 11970
+    },
+    {
+      "epoch": 70.88757396449704,
+      "grad_norm": 2.262409210205078,
+      "learning_rate": 0.0001,
+      "loss": 0.0279,
+      "step": 11980
+    },
+    {
+      "epoch": 70.94674556213018,
+      "grad_norm": 1.2372634410858154,
+      "learning_rate": 0.0001,
+      "loss": 0.0265,
+      "step": 11990
+    },
+    {
+      "epoch": 71.00591715976331,
+      "grad_norm": 1.7873915433883667,
+      "learning_rate": 0.0001,
+      "loss": 0.0271,
+      "step": 12000
+    },
+    {
+      "epoch": 71.06508875739645,
+      "grad_norm": 2.1135268211364746,
+      "learning_rate": 0.0001,
+      "loss": 0.0269,
+      "step": 12010
+    },
+    {
+      "epoch": 71.12426035502959,
+      "grad_norm": 1.395949363708496,
+      "learning_rate": 0.0001,
+      "loss": 0.0267,
+      "step": 12020
+    },
+    {
+      "epoch": 71.18343195266272,
+      "grad_norm": 1.285654067993164,
+      "learning_rate": 0.0001,
+      "loss": 0.0265,
+      "step": 12030
+    },
+    {
+      "epoch": 71.24260355029585,
+      "grad_norm": 1.4060099124908447,
+      "learning_rate": 0.0001,
+      "loss": 0.0273,
+      "step": 12040
+    },
+    {
+      "epoch": 71.30177514792899,
+      "grad_norm": 1.928000807762146,
+      "learning_rate": 0.0001,
+      "loss": 0.0265,
+      "step": 12050
+    },
+    {
+      "epoch": 71.36094674556213,
+      "grad_norm": 1.4371389150619507,
+      "learning_rate": 0.0001,
+      "loss": 0.028,
+      "step": 12060
+    },
+    {
+      "epoch": 71.42011834319527,
+      "grad_norm": 1.9911222457885742,
+      "learning_rate": 0.0001,
+      "loss": 0.0279,
+      "step": 12070
+    },
+    {
+      "epoch": 71.47928994082841,
+      "grad_norm": 2.0492377281188965,
+      "learning_rate": 0.0001,
+      "loss": 0.0288,
+      "step": 12080
+    },
+    {
+      "epoch": 71.53846153846153,
+      "grad_norm": 2.630584239959717,
+      "learning_rate": 0.0001,
+      "loss": 0.0294,
+      "step": 12090
+    },
+    {
+      "epoch": 71.59763313609467,
+      "grad_norm": 1.8017398118972778,
+      "learning_rate": 0.0001,
+      "loss": 0.0284,
+      "step": 12100
+    },
+    {
+      "epoch": 71.65680473372781,
+      "grad_norm": 2.668877124786377,
+      "learning_rate": 0.0001,
+      "loss": 0.0309,
+      "step": 12110
+    },
+    {
+      "epoch": 71.71597633136095,
+      "grad_norm": 2.213547468185425,
+      "learning_rate": 0.0001,
+      "loss": 0.0298,
+      "step": 12120
+    },
+    {
+      "epoch": 71.77514792899409,
+      "grad_norm": 2.248497247695923,
+      "learning_rate": 0.0001,
+      "loss": 0.0295,
+      "step": 12130
+    },
+    {
+      "epoch": 71.83431952662721,
+      "grad_norm": 2.9122376441955566,
+      "learning_rate": 0.0001,
+      "loss": 0.0288,
+      "step": 12140
+    },
+    {
+      "epoch": 71.89349112426035,
+      "grad_norm": 2.0023772716522217,
+      "learning_rate": 0.0001,
+      "loss": 0.0289,
+      "step": 12150
+    },
+    {
+      "epoch": 71.95266272189349,
+      "grad_norm": 2.234588623046875,
+      "learning_rate": 0.0001,
+      "loss": 0.0273,
+      "step": 12160
+    },
+    {
+      "epoch": 72.01183431952663,
+      "grad_norm": 2.086965322494507,
+      "learning_rate": 0.0001,
+      "loss": 0.027,
+      "step": 12170
+    },
+    {
+      "epoch": 72.07100591715977,
+      "grad_norm": 1.553324580192566,
+      "learning_rate": 0.0001,
+      "loss": 0.0267,
+      "step": 12180
+    },
+    {
+      "epoch": 72.1301775147929,
+      "grad_norm": 1.992963433265686,
+      "learning_rate": 0.0001,
+      "loss": 0.0265,
+      "step": 12190
+    },
+    {
+      "epoch": 72.18934911242603,
+      "grad_norm": 2.2860047817230225,
+      "learning_rate": 0.0001,
+      "loss": 0.0285,
+      "step": 12200
+    },
+    {
+      "epoch": 72.24852071005917,
+      "grad_norm": 1.741243600845337,
+      "learning_rate": 0.0001,
+      "loss": 0.0285,
+      "step": 12210
+    },
+    {
+      "epoch": 72.3076923076923,
+      "grad_norm": 1.7974306344985962,
+      "learning_rate": 0.0001,
+      "loss": 0.0284,
+      "step": 12220
+    },
+    {
+      "epoch": 72.36686390532545,
+      "grad_norm": 1.4983789920806885,
+      "learning_rate": 0.0001,
+      "loss": 0.0288,
+      "step": 12230
+    },
+    {
+      "epoch": 72.42603550295858,
+      "grad_norm": 3.466529130935669,
+      "learning_rate": 0.0001,
+      "loss": 0.0346,
+      "step": 12240
+    },
+    {
+      "epoch": 72.48520710059172,
+      "grad_norm": 3.0857725143432617,
+      "learning_rate": 0.0001,
+      "loss": 0.0324,
+      "step": 12250
+    },
+    {
+      "epoch": 72.54437869822485,
+      "grad_norm": 2.9473135471343994,
+      "learning_rate": 0.0001,
+      "loss": 0.032,
+      "step": 12260
+    },
+    {
+      "epoch": 72.60355029585799,
+      "grad_norm": 2.927501916885376,
+      "learning_rate": 0.0001,
+      "loss": 0.0294,
+      "step": 12270
+    },
+    {
+      "epoch": 72.66272189349112,
+      "grad_norm": 2.461545944213867,
+      "learning_rate": 0.0001,
+      "loss": 0.0292,
+      "step": 12280
+    },
+    {
+      "epoch": 72.72189349112426,
+      "grad_norm": 2.317444324493408,
+      "learning_rate": 0.0001,
+      "loss": 0.0279,
+      "step": 12290
+    },
+    {
+      "epoch": 72.7810650887574,
+      "grad_norm": 2.0168285369873047,
+      "learning_rate": 0.0001,
+      "loss": 0.0283,
+      "step": 12300
+    },
+    {
+      "epoch": 72.84023668639053,
+      "grad_norm": 1.8050506114959717,
+      "learning_rate": 0.0001,
+      "loss": 0.0271,
+      "step": 12310
+    },
+    {
+      "epoch": 72.89940828402366,
+      "grad_norm": 1.7260138988494873,
+      "learning_rate": 0.0001,
+      "loss": 0.0272,
+      "step": 12320
+    },
+    {
+      "epoch": 72.9585798816568,
+      "grad_norm": 2.2498533725738525,
+      "learning_rate": 0.0001,
+      "loss": 0.0256,
+      "step": 12330
+    },
+    {
+      "epoch": 73.01775147928994,
+      "grad_norm": 2.13832426071167,
+      "learning_rate": 0.0001,
+      "loss": 0.0273,
+      "step": 12340
+    },
+    {
+      "epoch": 73.07692307692308,
+      "grad_norm": 2.128333568572998,
+      "learning_rate": 0.0001,
+      "loss": 0.0268,
+      "step": 12350
+    },
+    {
+      "epoch": 73.13609467455622,
+      "grad_norm": 2.197119951248169,
+      "learning_rate": 0.0001,
+      "loss": 0.0285,
+      "step": 12360
+    },
+    {
+      "epoch": 73.19526627218934,
+      "grad_norm": 1.7648060321807861,
+      "learning_rate": 0.0001,
+      "loss": 0.0279,
+      "step": 12370
+    },
+    {
+      "epoch": 73.25443786982248,
+      "grad_norm": 1.596906304359436,
+      "learning_rate": 0.0001,
+      "loss": 0.0276,
+      "step": 12380
+    },
+    {
+      "epoch": 73.31360946745562,
+      "grad_norm": 2.564260244369507,
+      "learning_rate": 0.0001,
+      "loss": 0.0275,
+      "step": 12390
+    },
+    {
+      "epoch": 73.37278106508876,
+      "grad_norm": 2.4574925899505615,
+      "learning_rate": 0.0001,
+      "loss": 0.0282,
+      "step": 12400
+    },
+    {
+      "epoch": 73.4319526627219,
+      "grad_norm": 2.321512460708618,
+      "learning_rate": 0.0001,
+      "loss": 0.0287,
+      "step": 12410
+    },
+    {
+      "epoch": 73.49112426035504,
+      "grad_norm": 2.505284547805786,
+      "learning_rate": 0.0001,
+      "loss": 0.0278,
+      "step": 12420
+    },
+    {
+      "epoch": 73.55029585798816,
+      "grad_norm": 1.624753475189209,
+      "learning_rate": 0.0001,
+      "loss": 0.0274,
+      "step": 12430
+    },
+    {
+      "epoch": 73.6094674556213,
+      "grad_norm": 2.0071310997009277,
+      "learning_rate": 0.0001,
+      "loss": 0.0263,
+      "step": 12440
+    },
+    {
+      "epoch": 73.66863905325444,
+      "grad_norm": 1.7024009227752686,
+      "learning_rate": 0.0001,
+      "loss": 0.0265,
+      "step": 12450
+    },
+    {
+      "epoch": 73.72781065088758,
+      "grad_norm": 1.5280365943908691,
+      "learning_rate": 0.0001,
+      "loss": 0.0282,
+      "step": 12460
+    },
+    {
+      "epoch": 73.78698224852072,
+      "grad_norm": 1.3942376375198364,
+      "learning_rate": 0.0001,
+      "loss": 0.0275,
+      "step": 12470
+    },
+    {
+      "epoch": 73.84615384615384,
+      "grad_norm": 1.6773260831832886,
+      "learning_rate": 0.0001,
+      "loss": 0.0284,
+      "step": 12480
+    },
+    {
+      "epoch": 73.90532544378698,
+      "grad_norm": 1.9438140392303467,
+      "learning_rate": 0.0001,
+      "loss": 0.0286,
+      "step": 12490
+    },
+    {
+      "epoch": 73.96449704142012,
+      "grad_norm": 1.8076319694519043,
+      "learning_rate": 0.0001,
+      "loss": 0.0279,
+      "step": 12500
+    },
+    {
+      "epoch": 74.02366863905326,
+      "grad_norm": 2.1900219917297363,
+      "learning_rate": 0.0001,
+      "loss": 0.0289,
+      "step": 12510
+    },
+    {
+      "epoch": 74.0828402366864,
+      "grad_norm": 2.338517665863037,
+      "learning_rate": 0.0001,
+      "loss": 0.0301,
+      "step": 12520
+    },
+    {
+      "epoch": 74.14201183431953,
+      "grad_norm": 1.4687228202819824,
+      "learning_rate": 0.0001,
+      "loss": 0.0287,
+      "step": 12530
+    },
+    {
+      "epoch": 74.20118343195266,
+      "grad_norm": 2.9790847301483154,
+      "learning_rate": 0.0001,
+      "loss": 0.0308,
+      "step": 12540
+    },
+    {
+      "epoch": 74.2603550295858,
+      "grad_norm": 3.187764883041382,
+      "learning_rate": 0.0001,
+      "loss": 0.0303,
+      "step": 12550
+    },
+    {
+      "epoch": 74.31952662721893,
+      "grad_norm": 2.765580892562866,
+      "learning_rate": 0.0001,
+      "loss": 0.0289,
+      "step": 12560
+    },
+    {
+      "epoch": 74.37869822485207,
+      "grad_norm": 2.2064602375030518,
+      "learning_rate": 0.0001,
+      "loss": 0.0289,
+      "step": 12570
+    },
+    {
+      "epoch": 74.43786982248521,
+      "grad_norm": 1.9740427732467651,
+      "learning_rate": 0.0001,
+      "loss": 0.0292,
+      "step": 12580
+    },
+    {
+      "epoch": 74.49704142011835,
+      "grad_norm": 2.205927610397339,
+      "learning_rate": 0.0001,
+      "loss": 0.0279,
+      "step": 12590
+    },
+    {
+      "epoch": 74.55621301775147,
+      "grad_norm": 1.9258025884628296,
+      "learning_rate": 0.0001,
+      "loss": 0.0276,
+      "step": 12600
+    },
+    {
+      "epoch": 74.61538461538461,
+      "grad_norm": 1.5550994873046875,
+      "learning_rate": 0.0001,
+      "loss": 0.0277,
+      "step": 12610
+    },
+    {
+      "epoch": 74.67455621301775,
+      "grad_norm": 1.7193535566329956,
+      "learning_rate": 0.0001,
+      "loss": 0.0266,
+      "step": 12620
+    },
+    {
+      "epoch": 74.73372781065089,
+      "grad_norm": 1.2494144439697266,
+      "learning_rate": 0.0001,
+      "loss": 0.0276,
+      "step": 12630
+    },
+    {
+      "epoch": 74.79289940828403,
+      "grad_norm": 1.399194359779358,
+      "learning_rate": 0.0001,
+      "loss": 0.0273,
+      "step": 12640
+    },
+    {
+      "epoch": 74.85207100591715,
+      "grad_norm": 1.922747254371643,
+      "learning_rate": 0.0001,
+      "loss": 0.0287,
+      "step": 12650
+    },
+    {
+      "epoch": 74.91124260355029,
+      "grad_norm": 1.9999606609344482,
+      "learning_rate": 0.0001,
+      "loss": 0.0278,
+      "step": 12660
+    },
+    {
+      "epoch": 74.97041420118343,
+      "grad_norm": 1.538883924484253,
+      "learning_rate": 0.0001,
+      "loss": 0.0282,
+      "step": 12670
+    },
+    {
+      "epoch": 75.02958579881657,
+      "grad_norm": 1.8625574111938477,
+      "learning_rate": 0.0001,
+      "loss": 0.0267,
+      "step": 12680
+    },
+    {
+      "epoch": 75.08875739644971,
+      "grad_norm": 1.6682173013687134,
+      "learning_rate": 0.0001,
+      "loss": 0.0293,
+      "step": 12690
+    },
+    {
+      "epoch": 75.14792899408285,
+      "grad_norm": 1.5894958972930908,
+      "learning_rate": 0.0001,
+      "loss": 0.0288,
+      "step": 12700
+    },
+    {
+      "epoch": 75.20710059171597,
+      "grad_norm": 1.5210741758346558,
+      "learning_rate": 0.0001,
+      "loss": 0.029,
+      "step": 12710
+    },
+    {
+      "epoch": 75.26627218934911,
+      "grad_norm": 2.3625807762145996,
+      "learning_rate": 0.0001,
+      "loss": 0.028,
+      "step": 12720
+    },
+    {
+      "epoch": 75.32544378698225,
+      "grad_norm": 1.474217176437378,
+      "learning_rate": 0.0001,
+      "loss": 0.0283,
+      "step": 12730
+    },
+    {
+      "epoch": 75.38461538461539,
+      "grad_norm": 1.9648138284683228,
+      "learning_rate": 0.0001,
+      "loss": 0.0275,
+      "step": 12740
+    },
+    {
+      "epoch": 75.44378698224853,
+      "grad_norm": 2.1769723892211914,
+      "learning_rate": 0.0001,
+      "loss": 0.0285,
+      "step": 12750
+    },
+    {
+      "epoch": 75.50295857988165,
+      "grad_norm": 1.8122273683547974,
+      "learning_rate": 0.0001,
+      "loss": 0.0286,
+      "step": 12760
+    },
+    {
+      "epoch": 75.56213017751479,
+      "grad_norm": 1.4470155239105225,
+      "learning_rate": 0.0001,
+      "loss": 0.0295,
+      "step": 12770
+    },
+    {
+      "epoch": 75.62130177514793,
+      "grad_norm": 2.2424378395080566,
+      "learning_rate": 0.0001,
+      "loss": 0.0312,
+      "step": 12780
+    },
+    {
+      "epoch": 75.68047337278107,
+      "grad_norm": 1.7034374475479126,
+      "learning_rate": 0.0001,
+      "loss": 0.0291,
+      "step": 12790
+    },
+    {
+      "epoch": 75.7396449704142,
+      "grad_norm": 1.4852824211120605,
+      "learning_rate": 0.0001,
+      "loss": 0.0295,
+      "step": 12800
+    },
+    {
+      "epoch": 75.79881656804734,
+      "grad_norm": 1.9763224124908447,
+      "learning_rate": 0.0001,
+      "loss": 0.029,
+      "step": 12810
+    },
+    {
+      "epoch": 75.85798816568047,
+      "grad_norm": 1.5732263326644897,
+      "learning_rate": 0.0001,
+      "loss": 0.0298,
+      "step": 12820
+    },
+    {
+      "epoch": 75.9171597633136,
+      "grad_norm": 1.7625895738601685,
+      "learning_rate": 0.0001,
+      "loss": 0.0311,
+      "step": 12830
+    },
+    {
+      "epoch": 75.97633136094674,
+      "grad_norm": 2.319737434387207,
+      "learning_rate": 0.0001,
+      "loss": 0.0297,
+      "step": 12840
+    },
+    {
+      "epoch": 76.03550295857988,
+      "grad_norm": 1.7892645597457886,
+      "learning_rate": 0.0001,
+      "loss": 0.0299,
+      "step": 12850
+    },
+    {
+      "epoch": 76.09467455621302,
+      "grad_norm": 1.8642587661743164,
+      "learning_rate": 0.0001,
+      "loss": 0.0294,
+      "step": 12860
+    },
+    {
+      "epoch": 76.15384615384616,
+      "grad_norm": 2.071037530899048,
+      "learning_rate": 0.0001,
+      "loss": 0.0282,
+      "step": 12870
+    },
+    {
+      "epoch": 76.21301775147928,
+      "grad_norm": 1.838343620300293,
+      "learning_rate": 0.0001,
+      "loss": 0.0293,
+      "step": 12880
+    },
+    {
+      "epoch": 76.27218934911242,
+      "grad_norm": 1.5505099296569824,
+      "learning_rate": 0.0001,
+      "loss": 0.0284,
+      "step": 12890
+    },
+    {
+      "epoch": 76.33136094674556,
+      "grad_norm": 1.6460497379302979,
+      "learning_rate": 0.0001,
+      "loss": 0.0284,
+      "step": 12900
+    },
+    {
+      "epoch": 76.3905325443787,
+      "grad_norm": 2.0226938724517822,
+      "learning_rate": 0.0001,
+      "loss": 0.0276,
+      "step": 12910
+    },
+    {
+      "epoch": 76.44970414201184,
+      "grad_norm": 1.4064687490463257,
+      "learning_rate": 0.0001,
+      "loss": 0.0281,
+      "step": 12920
+    },
+    {
+      "epoch": 76.50887573964496,
+      "grad_norm": 1.3913443088531494,
+      "learning_rate": 0.0001,
+      "loss": 0.0275,
+      "step": 12930
+    },
+    {
+      "epoch": 76.5680473372781,
+      "grad_norm": 1.6919381618499756,
+      "learning_rate": 0.0001,
+      "loss": 0.0292,
+      "step": 12940
+    },
+    {
+      "epoch": 76.62721893491124,
+      "grad_norm": 1.458065152168274,
+      "learning_rate": 0.0001,
+      "loss": 0.0291,
+      "step": 12950
+    },
+    {
+      "epoch": 76.68639053254438,
+      "grad_norm": 1.9224157333374023,
+      "learning_rate": 0.0001,
+      "loss": 0.0305,
+      "step": 12960
+    },
+    {
+      "epoch": 76.74556213017752,
+      "grad_norm": 1.611428141593933,
+      "learning_rate": 0.0001,
+      "loss": 0.0305,
+      "step": 12970
+    },
+    {
+      "epoch": 76.80473372781066,
+      "grad_norm": 1.830028772354126,
+      "learning_rate": 0.0001,
+      "loss": 0.0287,
+      "step": 12980
+    },
+    {
+      "epoch": 76.86390532544378,
+      "grad_norm": 2.147338628768921,
+      "learning_rate": 0.0001,
+      "loss": 0.0283,
+      "step": 12990
+    },
+    {
+      "epoch": 76.92307692307692,
+      "grad_norm": 1.8643895387649536,
+      "learning_rate": 0.0001,
+      "loss": 0.0281,
+      "step": 13000
+    },
+    {
+      "epoch": 76.98224852071006,
+      "grad_norm": 1.755447506904602,
+      "learning_rate": 0.0001,
+      "loss": 0.0274,
+      "step": 13010
+    },
+    {
+      "epoch": 77.0414201183432,
+      "grad_norm": 1.5869688987731934,
+      "learning_rate": 0.0001,
+      "loss": 0.0269,
+      "step": 13020
+    },
+    {
+      "epoch": 77.10059171597634,
+      "grad_norm": 1.8581516742706299,
+      "learning_rate": 0.0001,
+      "loss": 0.0283,
+      "step": 13030
+    },
+    {
+      "epoch": 77.15976331360947,
+      "grad_norm": 1.6732298135757446,
+      "learning_rate": 0.0001,
+      "loss": 0.0274,
+      "step": 13040
+    },
+    {
+      "epoch": 77.2189349112426,
+      "grad_norm": 1.822118878364563,
+      "learning_rate": 0.0001,
+      "loss": 0.0276,
+      "step": 13050
+    },
+    {
+      "epoch": 77.27810650887574,
+      "grad_norm": 1.9395040273666382,
+      "learning_rate": 0.0001,
+      "loss": 0.0278,
+      "step": 13060
+    },
+    {
+      "epoch": 77.33727810650888,
+      "grad_norm": 1.5575655698776245,
+      "learning_rate": 0.0001,
+      "loss": 0.0282,
+      "step": 13070
+    },
+    {
+      "epoch": 77.39644970414201,
+      "grad_norm": 1.541513204574585,
+      "learning_rate": 0.0001,
+      "loss": 0.0269,
+      "step": 13080
+    },
+    {
+      "epoch": 77.45562130177515,
+      "grad_norm": 1.8625190258026123,
+      "learning_rate": 0.0001,
+      "loss": 0.0286,
+      "step": 13090
+    },
+    {
+      "epoch": 77.51479289940828,
+      "grad_norm": 1.588661551475525,
+      "learning_rate": 0.0001,
+      "loss": 0.0279,
+      "step": 13100
+    },
+    {
+      "epoch": 77.57396449704142,
+      "grad_norm": 2.398642063140869,
+      "learning_rate": 0.0001,
+      "loss": 0.0267,
+      "step": 13110
+    },
+    {
+      "epoch": 77.63313609467455,
+      "grad_norm": 1.3619097471237183,
+      "learning_rate": 0.0001,
+      "loss": 0.0278,
+      "step": 13120
+    },
+    {
+      "epoch": 77.6923076923077,
+      "grad_norm": 1.858024001121521,
+      "learning_rate": 0.0001,
+      "loss": 0.0275,
+      "step": 13130
+    },
+    {
+      "epoch": 77.75147928994083,
+      "grad_norm": 1.2112067937850952,
+      "learning_rate": 0.0001,
+      "loss": 0.0278,
+      "step": 13140
+    },
+    {
+      "epoch": 77.81065088757397,
+      "grad_norm": 1.4806443452835083,
+      "learning_rate": 0.0001,
+      "loss": 0.0283,
+      "step": 13150
+    },
+    {
+      "epoch": 77.8698224852071,
+      "grad_norm": 1.6459596157073975,
+      "learning_rate": 0.0001,
+      "loss": 0.0271,
+      "step": 13160
+    },
+    {
+      "epoch": 77.92899408284023,
+      "grad_norm": 1.6955714225769043,
+      "learning_rate": 0.0001,
+      "loss": 0.0294,
+      "step": 13170
+    },
+    {
+      "epoch": 77.98816568047337,
+      "grad_norm": 1.5850615501403809,
+      "learning_rate": 0.0001,
+      "loss": 0.0295,
+      "step": 13180
+    },
+    {
+      "epoch": 78.04733727810651,
+      "grad_norm": 1.4026334285736084,
+      "learning_rate": 0.0001,
+      "loss": 0.0284,
+      "step": 13190
+    },
+    {
+      "epoch": 78.10650887573965,
+      "grad_norm": 1.9902260303497314,
+      "learning_rate": 0.0001,
+      "loss": 0.0282,
+      "step": 13200
+    },
+    {
+      "epoch": 78.16568047337279,
+      "grad_norm": 1.9664236307144165,
+      "learning_rate": 0.0001,
+      "loss": 0.029,
+      "step": 13210
+    },
+    {
+      "epoch": 78.22485207100591,
+      "grad_norm": 1.820135235786438,
+      "learning_rate": 0.0001,
+      "loss": 0.028,
+      "step": 13220
+    },
+    {
+      "epoch": 78.28402366863905,
+      "grad_norm": 1.6468032598495483,
+      "learning_rate": 0.0001,
+      "loss": 0.0276,
+      "step": 13230
+    },
+    {
+      "epoch": 78.34319526627219,
+      "grad_norm": 1.3373830318450928,
+      "learning_rate": 0.0001,
+      "loss": 0.0278,
+      "step": 13240
+    },
+    {
+      "epoch": 78.40236686390533,
+      "grad_norm": 1.835564374923706,
+      "learning_rate": 0.0001,
+      "loss": 0.0282,
+      "step": 13250
+    },
+    {
+      "epoch": 78.46153846153847,
+      "grad_norm": 1.6432665586471558,
+      "learning_rate": 0.0001,
+      "loss": 0.0279,
+      "step": 13260
+    },
+    {
+      "epoch": 78.52071005917159,
+      "grad_norm": 1.6171239614486694,
+      "learning_rate": 0.0001,
+      "loss": 0.0282,
+      "step": 13270
+    },
+    {
+      "epoch": 78.57988165680473,
+      "grad_norm": 1.7857800722122192,
+      "learning_rate": 0.0001,
+      "loss": 0.0285,
+      "step": 13280
+    },
+    {
+      "epoch": 78.63905325443787,
+      "grad_norm": 1.6145315170288086,
+      "learning_rate": 0.0001,
+      "loss": 0.0288,
+      "step": 13290
+    },
+    {
+      "epoch": 78.69822485207101,
+      "grad_norm": 1.3754020929336548,
+      "learning_rate": 0.0001,
+      "loss": 0.028,
+      "step": 13300
+    },
+    {
+      "epoch": 78.75739644970415,
+      "grad_norm": 2.504894256591797,
+      "learning_rate": 0.0001,
+      "loss": 0.0277,
+      "step": 13310
+    },
+    {
+      "epoch": 78.81656804733728,
+      "grad_norm": 1.7269688844680786,
+      "learning_rate": 0.0001,
+      "loss": 0.0279,
+      "step": 13320
+    },
+    {
+      "epoch": 78.87573964497041,
+      "grad_norm": 1.0000234842300415,
+      "learning_rate": 0.0001,
+      "loss": 0.027,
+      "step": 13330
+    },
+    {
+      "epoch": 78.93491124260355,
+      "grad_norm": 1.391007661819458,
+      "learning_rate": 0.0001,
+      "loss": 0.027,
+      "step": 13340
+    },
+    {
+      "epoch": 78.99408284023669,
+      "grad_norm": 1.2920475006103516,
+      "learning_rate": 0.0001,
+      "loss": 0.0268,
+      "step": 13350
+    },
+    {
+      "epoch": 79.05325443786982,
+      "grad_norm": 1.4586950540542603,
+      "learning_rate": 0.0001,
+      "loss": 0.0265,
+      "step": 13360
+    },
+    {
+      "epoch": 79.11242603550296,
+      "grad_norm": 1.5049959421157837,
+      "learning_rate": 0.0001,
+      "loss": 0.0267,
+      "step": 13370
+    },
+    {
+      "epoch": 79.17159763313609,
+      "grad_norm": 1.3885337114334106,
+      "learning_rate": 0.0001,
+      "loss": 0.0268,
+      "step": 13380
+    },
+    {
+      "epoch": 79.23076923076923,
+      "grad_norm": 1.6495060920715332,
+      "learning_rate": 0.0001,
+      "loss": 0.0274,
+      "step": 13390
+    },
+    {
+      "epoch": 79.28994082840237,
+      "grad_norm": 2.0278308391571045,
+      "learning_rate": 0.0001,
+      "loss": 0.0272,
+      "step": 13400
+    },
+    {
+      "epoch": 79.3491124260355,
+      "grad_norm": 1.6315144300460815,
+      "learning_rate": 0.0001,
+      "loss": 0.0275,
+      "step": 13410
+    },
+    {
+      "epoch": 79.40828402366864,
+      "grad_norm": 1.6109946966171265,
+      "learning_rate": 0.0001,
+      "loss": 0.0273,
+      "step": 13420
+    },
+    {
+      "epoch": 79.46745562130178,
+      "grad_norm": 1.9447684288024902,
+      "learning_rate": 0.0001,
+      "loss": 0.0272,
+      "step": 13430
+    },
+    {
+      "epoch": 79.5266272189349,
+      "grad_norm": 1.2351568937301636,
+      "learning_rate": 0.0001,
+      "loss": 0.0271,
+      "step": 13440
+    },
+    {
+      "epoch": 79.58579881656804,
+      "grad_norm": 1.62055504322052,
+      "learning_rate": 0.0001,
+      "loss": 0.0267,
+      "step": 13450
+    },
+    {
+      "epoch": 79.64497041420118,
+      "grad_norm": 1.7545634508132935,
+      "learning_rate": 0.0001,
+      "loss": 0.0266,
+      "step": 13460
+    },
+    {
+      "epoch": 79.70414201183432,
+      "grad_norm": 2.3588948249816895,
+      "learning_rate": 0.0001,
+      "loss": 0.0259,
+      "step": 13470
+    },
+    {
+      "epoch": 79.76331360946746,
+      "grad_norm": 1.7819347381591797,
+      "learning_rate": 0.0001,
+      "loss": 0.0257,
+      "step": 13480
+    },
+    {
+      "epoch": 79.8224852071006,
+      "grad_norm": 1.7651571035385132,
+      "learning_rate": 0.0001,
+      "loss": 0.0262,
+      "step": 13490
+    },
+    {
+      "epoch": 79.88165680473372,
+      "grad_norm": 2.098773717880249,
+      "learning_rate": 0.0001,
+      "loss": 0.0256,
+      "step": 13500
+    },
+    {
+      "epoch": 79.94082840236686,
+      "grad_norm": 2.0386404991149902,
+      "learning_rate": 0.0001,
+      "loss": 0.0256,
+      "step": 13510
+    },
+    {
+      "epoch": 80.0,
+      "grad_norm": 2.1002614498138428,
+      "learning_rate": 0.0001,
+      "loss": 0.0264,
+      "step": 13520
+    },
+    {
+      "epoch": 80.05917159763314,
+      "grad_norm": 2.677924394607544,
+      "learning_rate": 0.0001,
+      "loss": 0.0326,
+      "step": 13530
+    },
+    {
+      "epoch": 80.11834319526628,
+      "grad_norm": 2.323535203933716,
+      "learning_rate": 0.0001,
+      "loss": 0.0278,
+      "step": 13540
+    },
+    {
+      "epoch": 80.1775147928994,
+      "grad_norm": 2.4471118450164795,
+      "learning_rate": 0.0001,
+      "loss": 0.0275,
+      "step": 13550
+    },
+    {
+      "epoch": 80.23668639053254,
+      "grad_norm": 2.4190471172332764,
+      "learning_rate": 0.0001,
+      "loss": 0.0264,
+      "step": 13560
+    },
+    {
+      "epoch": 80.29585798816568,
+      "grad_norm": 2.126497268676758,
+      "learning_rate": 0.0001,
+      "loss": 0.0251,
+      "step": 13570
+    },
+    {
+      "epoch": 80.35502958579882,
+      "grad_norm": 1.9908725023269653,
+      "learning_rate": 0.0001,
+      "loss": 0.0253,
+      "step": 13580
+    },
+    {
+      "epoch": 80.41420118343196,
+      "grad_norm": 1.7533633708953857,
+      "learning_rate": 0.0001,
+      "loss": 0.0255,
+      "step": 13590
+    },
+    {
+      "epoch": 80.4733727810651,
+      "grad_norm": 1.8987650871276855,
+      "learning_rate": 0.0001,
+      "loss": 0.026,
+      "step": 13600
+    },
+    {
+      "epoch": 80.53254437869822,
+      "grad_norm": 2.1181752681732178,
+      "learning_rate": 0.0001,
+      "loss": 0.0247,
+      "step": 13610
+    },
+    {
+      "epoch": 80.59171597633136,
+      "grad_norm": 2.9005842208862305,
+      "learning_rate": 0.0001,
+      "loss": 0.0269,
+      "step": 13620
+    },
+    {
+      "epoch": 80.6508875739645,
+      "grad_norm": 1.7335445880889893,
+      "learning_rate": 0.0001,
+      "loss": 0.0271,
+      "step": 13630
+    },
+    {
+      "epoch": 80.71005917159763,
+      "grad_norm": 2.436126708984375,
+      "learning_rate": 0.0001,
+      "loss": 0.0277,
+      "step": 13640
+    },
+    {
+      "epoch": 80.76923076923077,
+      "grad_norm": 1.8038551807403564,
+      "learning_rate": 0.0001,
+      "loss": 0.0271,
+      "step": 13650
+    },
+    {
+      "epoch": 80.82840236686391,
+      "grad_norm": 1.873564600944519,
+      "learning_rate": 0.0001,
+      "loss": 0.0257,
+      "step": 13660
+    },
+    {
+      "epoch": 80.88757396449704,
+      "grad_norm": 1.7780539989471436,
+      "learning_rate": 0.0001,
+      "loss": 0.0282,
+      "step": 13670
+    },
+    {
+      "epoch": 80.94674556213018,
+      "grad_norm": 1.9945662021636963,
+      "learning_rate": 0.0001,
+      "loss": 0.0258,
+      "step": 13680
+    },
+    {
+      "epoch": 81.00591715976331,
+      "grad_norm": 1.6491867303848267,
+      "learning_rate": 0.0001,
+      "loss": 0.0264,
+      "step": 13690
+    },
+    {
+      "epoch": 81.06508875739645,
+      "grad_norm": 1.9175587892532349,
+      "learning_rate": 0.0001,
+      "loss": 0.0261,
+      "step": 13700
+    },
+    {
+      "epoch": 81.12426035502959,
+      "grad_norm": 2.297849178314209,
+      "learning_rate": 0.0001,
+      "loss": 0.0262,
+      "step": 13710
+    },
+    {
+      "epoch": 81.18343195266272,
+      "grad_norm": 2.206448793411255,
+      "learning_rate": 0.0001,
+      "loss": 0.0267,
+      "step": 13720
+    },
+    {
+      "epoch": 81.24260355029585,
+      "grad_norm": 1.6405094861984253,
+      "learning_rate": 0.0001,
+      "loss": 0.026,
+      "step": 13730
+    },
+    {
+      "epoch": 81.30177514792899,
+      "grad_norm": 1.737815260887146,
+      "learning_rate": 0.0001,
+      "loss": 0.0261,
+      "step": 13740
+    },
+    {
+      "epoch": 81.36094674556213,
+      "grad_norm": 2.051853895187378,
+      "learning_rate": 0.0001,
+      "loss": 0.0275,
+      "step": 13750
+    },
+    {
+      "epoch": 81.42011834319527,
+      "grad_norm": 1.711469292640686,
+      "learning_rate": 0.0001,
+      "loss": 0.0268,
+      "step": 13760
+    },
+    {
+      "epoch": 81.47928994082841,
+      "grad_norm": 1.4849634170532227,
+      "learning_rate": 0.0001,
+      "loss": 0.0262,
+      "step": 13770
+    },
+    {
+      "epoch": 81.53846153846153,
+      "grad_norm": 1.7072783708572388,
+      "learning_rate": 0.0001,
+      "loss": 0.0269,
+      "step": 13780
+    },
+    {
+      "epoch": 81.59763313609467,
+      "grad_norm": 1.5725935697555542,
+      "learning_rate": 0.0001,
+      "loss": 0.0279,
+      "step": 13790
+    },
+    {
+      "epoch": 81.65680473372781,
+      "grad_norm": 1.8924936056137085,
+      "learning_rate": 0.0001,
+      "loss": 0.0277,
+      "step": 13800
+    },
+    {
+      "epoch": 81.71597633136095,
+      "grad_norm": 1.6488643884658813,
+      "learning_rate": 0.0001,
+      "loss": 0.0277,
+      "step": 13810
+    },
+    {
+      "epoch": 81.77514792899409,
+      "grad_norm": 1.625576376914978,
+      "learning_rate": 0.0001,
+      "loss": 0.0271,
+      "step": 13820
+    },
+    {
+      "epoch": 81.83431952662721,
+      "grad_norm": 2.2656078338623047,
+      "learning_rate": 0.0001,
+      "loss": 0.0276,
+      "step": 13830
+    },
+    {
+      "epoch": 81.89349112426035,
+      "grad_norm": 1.5859264135360718,
+      "learning_rate": 0.0001,
+      "loss": 0.026,
+      "step": 13840
+    },
+    {
+      "epoch": 81.95266272189349,
+      "grad_norm": 1.759376049041748,
+      "learning_rate": 0.0001,
+      "loss": 0.0268,
+      "step": 13850
+    },
+    {
+      "epoch": 82.01183431952663,
+      "grad_norm": 1.5925393104553223,
+      "learning_rate": 0.0001,
+      "loss": 0.0271,
+      "step": 13860
+    },
+    {
+      "epoch": 82.07100591715977,
+      "grad_norm": 1.6294904947280884,
+      "learning_rate": 0.0001,
+      "loss": 0.0263,
+      "step": 13870
+    },
+    {
+      "epoch": 82.1301775147929,
+      "grad_norm": 2.1179983615875244,
+      "learning_rate": 0.0001,
+      "loss": 0.0262,
+      "step": 13880
+    },
+    {
+      "epoch": 82.18934911242603,
+      "grad_norm": 1.9530357122421265,
+      "learning_rate": 0.0001,
+      "loss": 0.027,
+      "step": 13890
+    },
+    {
+      "epoch": 82.24852071005917,
+      "grad_norm": 2.098775625228882,
+      "learning_rate": 0.0001,
+      "loss": 0.0269,
+      "step": 13900
+    },
+    {
+      "epoch": 82.3076923076923,
+      "grad_norm": 1.9112647771835327,
+      "learning_rate": 0.0001,
+      "loss": 0.027,
+      "step": 13910
+    },
+    {
+      "epoch": 82.36686390532545,
+      "grad_norm": 2.509934425354004,
+      "learning_rate": 0.0001,
+      "loss": 0.0279,
+      "step": 13920
+    },
+    {
+      "epoch": 82.42603550295858,
+      "grad_norm": 2.1358492374420166,
+      "learning_rate": 0.0001,
+      "loss": 0.0274,
+      "step": 13930
+    },
+    {
+      "epoch": 82.48520710059172,
+      "grad_norm": 1.9020432233810425,
+      "learning_rate": 0.0001,
+      "loss": 0.0293,
+      "step": 13940
+    },
+    {
+      "epoch": 82.54437869822485,
+      "grad_norm": 1.5463789701461792,
+      "learning_rate": 0.0001,
+      "loss": 0.0271,
+      "step": 13950
+    },
+    {
+      "epoch": 82.60355029585799,
+      "grad_norm": 1.2848080396652222,
+      "learning_rate": 0.0001,
+      "loss": 0.0265,
+      "step": 13960
+    },
+    {
+      "epoch": 82.66272189349112,
+      "grad_norm": 1.5767860412597656,
+      "learning_rate": 0.0001,
+      "loss": 0.0248,
+      "step": 13970
+    },
+    {
+      "epoch": 82.72189349112426,
+      "grad_norm": 1.8156096935272217,
+      "learning_rate": 0.0001,
+      "loss": 0.0269,
+      "step": 13980
+    },
+    {
+      "epoch": 82.7810650887574,
+      "grad_norm": 1.4919930696487427,
+      "learning_rate": 0.0001,
+      "loss": 0.0263,
+      "step": 13990
+    },
+    {
+      "epoch": 82.84023668639053,
+      "grad_norm": 1.5417284965515137,
+      "learning_rate": 0.0001,
+      "loss": 0.026,
+      "step": 14000
+    },
+    {
+      "epoch": 82.89940828402366,
+      "grad_norm": 2.150217294692993,
+      "learning_rate": 0.0001,
+      "loss": 0.0273,
+      "step": 14010
+    },
+    {
+      "epoch": 82.9585798816568,
+      "grad_norm": 1.802996277809143,
+      "learning_rate": 0.0001,
+      "loss": 0.0284,
+      "step": 14020
+    },
+    {
+      "epoch": 83.01775147928994,
+      "grad_norm": 1.4453362226486206,
+      "learning_rate": 0.0001,
+      "loss": 0.0272,
+      "step": 14030
+    },
+    {
+      "epoch": 83.07692307692308,
+      "grad_norm": 1.6498535871505737,
+      "learning_rate": 0.0001,
+      "loss": 0.0266,
+      "step": 14040
+    },
+    {
+      "epoch": 83.13609467455622,
+      "grad_norm": 1.8970208168029785,
+      "learning_rate": 0.0001,
+      "loss": 0.0275,
+      "step": 14050
+    },
+    {
+      "epoch": 83.19526627218934,
+      "grad_norm": 1.2003551721572876,
+      "learning_rate": 0.0001,
+      "loss": 0.0278,
+      "step": 14060
+    },
+    {
+      "epoch": 83.25443786982248,
+      "grad_norm": 1.6260120868682861,
+      "learning_rate": 0.0001,
+      "loss": 0.0266,
+      "step": 14070
+    },
+    {
+      "epoch": 83.31360946745562,
+      "grad_norm": 1.3744405508041382,
+      "learning_rate": 0.0001,
+      "loss": 0.0265,
+      "step": 14080
+    },
+    {
+      "epoch": 83.37278106508876,
+      "grad_norm": 1.5436270236968994,
+      "learning_rate": 0.0001,
+      "loss": 0.0271,
+      "step": 14090
+    },
+    {
+      "epoch": 83.4319526627219,
+      "grad_norm": 1.2730234861373901,
+      "learning_rate": 0.0001,
+      "loss": 0.0275,
+      "step": 14100
+    },
+    {
+      "epoch": 83.49112426035504,
+      "grad_norm": 1.467305302619934,
+      "learning_rate": 0.0001,
+      "loss": 0.0261,
+      "step": 14110
+    },
+    {
+      "epoch": 83.55029585798816,
+      "grad_norm": 2.2118680477142334,
+      "learning_rate": 0.0001,
+      "loss": 0.027,
+      "step": 14120
+    },
+    {
+      "epoch": 83.6094674556213,
+      "grad_norm": 1.4269821643829346,
+      "learning_rate": 0.0001,
+      "loss": 0.0268,
+      "step": 14130
+    },
+    {
+      "epoch": 83.66863905325444,
+      "grad_norm": 1.488857626914978,
+      "learning_rate": 0.0001,
+      "loss": 0.0277,
+      "step": 14140
+    },
+    {
+      "epoch": 83.72781065088758,
+      "grad_norm": 2.206981658935547,
+      "learning_rate": 0.0001,
+      "loss": 0.0277,
+      "step": 14150
+    },
+    {
+      "epoch": 83.78698224852072,
+      "grad_norm": 1.5116686820983887,
+      "learning_rate": 0.0001,
+      "loss": 0.0278,
+      "step": 14160
+    },
+    {
+      "epoch": 83.84615384615384,
+      "grad_norm": 1.3844295740127563,
+      "learning_rate": 0.0001,
+      "loss": 0.0275,
+      "step": 14170
+    },
+    {
+      "epoch": 83.90532544378698,
+      "grad_norm": 1.5175310373306274,
+      "learning_rate": 0.0001,
+      "loss": 0.0274,
+      "step": 14180
+    },
+    {
+      "epoch": 83.96449704142012,
+      "grad_norm": 1.6370232105255127,
+      "learning_rate": 0.0001,
+      "loss": 0.0273,
+      "step": 14190
+    },
+    {
+      "epoch": 84.02366863905326,
+      "grad_norm": 1.391525387763977,
+      "learning_rate": 0.0001,
+      "loss": 0.0275,
+      "step": 14200
+    },
+    {
+      "epoch": 84.0828402366864,
+      "grad_norm": 1.245666742324829,
+      "learning_rate": 0.0001,
+      "loss": 0.0288,
+      "step": 14210
+    },
+    {
+      "epoch": 84.14201183431953,
+      "grad_norm": 1.706787109375,
+      "learning_rate": 0.0001,
+      "loss": 0.0294,
+      "step": 14220
+    },
+    {
+      "epoch": 84.20118343195266,
+      "grad_norm": 2.095344305038452,
+      "learning_rate": 0.0001,
+      "loss": 0.0289,
+      "step": 14230
+    },
+    {
+      "epoch": 84.2603550295858,
+      "grad_norm": 1.3527956008911133,
+      "learning_rate": 0.0001,
+      "loss": 0.0268,
+      "step": 14240
+    },
+    {
+      "epoch": 84.31952662721893,
+      "grad_norm": 2.124955177307129,
+      "learning_rate": 0.0001,
+      "loss": 0.0269,
+      "step": 14250
+    },
+    {
+      "epoch": 84.37869822485207,
+      "grad_norm": 1.9385125637054443,
+      "learning_rate": 0.0001,
+      "loss": 0.0269,
+      "step": 14260
+    },
+    {
+      "epoch": 84.43786982248521,
+      "grad_norm": 1.4271492958068848,
+      "learning_rate": 0.0001,
+      "loss": 0.0269,
+      "step": 14270
+    },
+    {
+      "epoch": 84.49704142011835,
+      "grad_norm": 1.7068454027175903,
+      "learning_rate": 0.0001,
+      "loss": 0.026,
+      "step": 14280
+    },
+    {
+      "epoch": 84.55621301775147,
+      "grad_norm": 2.200274705886841,
+      "learning_rate": 0.0001,
+      "loss": 0.0269,
+      "step": 14290
+    },
+    {
+      "epoch": 84.61538461538461,
+      "grad_norm": 1.7167702913284302,
+      "learning_rate": 0.0001,
+      "loss": 0.0267,
+      "step": 14300
+    },
+    {
+      "epoch": 84.67455621301775,
+      "grad_norm": 1.483392357826233,
+      "learning_rate": 0.0001,
+      "loss": 0.0267,
+      "step": 14310
+    },
+    {
+      "epoch": 84.73372781065089,
+      "grad_norm": 1.6335539817810059,
+      "learning_rate": 0.0001,
+      "loss": 0.0269,
+      "step": 14320
+    },
+    {
+      "epoch": 84.79289940828403,
+      "grad_norm": 2.1830525398254395,
+      "learning_rate": 0.0001,
+      "loss": 0.0268,
+      "step": 14330
+    },
+    {
+      "epoch": 84.85207100591715,
+      "grad_norm": 2.410512924194336,
+      "learning_rate": 0.0001,
+      "loss": 0.0281,
+      "step": 14340
+    },
+    {
+      "epoch": 84.91124260355029,
+      "grad_norm": 2.7541041374206543,
+      "learning_rate": 0.0001,
+      "loss": 0.0284,
+      "step": 14350
+    },
+    {
+      "epoch": 84.97041420118343,
+      "grad_norm": 1.7611759901046753,
+      "learning_rate": 0.0001,
+      "loss": 0.0259,
+      "step": 14360
+    },
+    {
+      "epoch": 85.02958579881657,
+      "grad_norm": 1.6813422441482544,
+      "learning_rate": 0.0001,
+      "loss": 0.0273,
+      "step": 14370
+    },
+    {
+      "epoch": 85.08875739644971,
+      "grad_norm": 1.5463656187057495,
+      "learning_rate": 0.0001,
+      "loss": 0.0262,
+      "step": 14380
+    },
+    {
+      "epoch": 85.14792899408285,
+      "grad_norm": 1.8891743421554565,
+      "learning_rate": 0.0001,
+      "loss": 0.0278,
+      "step": 14390
+    },
+    {
+      "epoch": 85.20710059171597,
+      "grad_norm": 1.7381260395050049,
+      "learning_rate": 0.0001,
+      "loss": 0.0271,
+      "step": 14400
+    },
+    {
+      "epoch": 85.26627218934911,
+      "grad_norm": 1.766524076461792,
+      "learning_rate": 0.0001,
+      "loss": 0.027,
+      "step": 14410
+    },
+    {
+      "epoch": 85.32544378698225,
+      "grad_norm": 1.41001558303833,
+      "learning_rate": 0.0001,
+      "loss": 0.0262,
+      "step": 14420
+    },
+    {
+      "epoch": 85.38461538461539,
+      "grad_norm": 1.8273169994354248,
+      "learning_rate": 0.0001,
+      "loss": 0.0259,
+      "step": 14430
+    },
+    {
+      "epoch": 85.44378698224853,
+      "grad_norm": 1.8059457540512085,
+      "learning_rate": 0.0001,
+      "loss": 0.0268,
+      "step": 14440
+    },
+    {
+      "epoch": 85.50295857988165,
+      "grad_norm": 1.894208550453186,
+      "learning_rate": 0.0001,
+      "loss": 0.0275,
+      "step": 14450
+    },
+    {
+      "epoch": 85.56213017751479,
+      "grad_norm": 1.280652403831482,
+      "learning_rate": 0.0001,
+      "loss": 0.0261,
+      "step": 14460
+    },
+    {
+      "epoch": 85.62130177514793,
+      "grad_norm": 1.6881978511810303,
+      "learning_rate": 0.0001,
+      "loss": 0.0263,
+      "step": 14470
+    },
+    {
+      "epoch": 85.68047337278107,
+      "grad_norm": 1.6599359512329102,
+      "learning_rate": 0.0001,
+      "loss": 0.0267,
+      "step": 14480
+    },
+    {
+      "epoch": 85.7396449704142,
+      "grad_norm": 1.451037883758545,
+      "learning_rate": 0.0001,
+      "loss": 0.0255,
+      "step": 14490
+    },
+    {
+      "epoch": 85.79881656804734,
+      "grad_norm": 1.9234201908111572,
+      "learning_rate": 0.0001,
+      "loss": 0.0258,
+      "step": 14500
+    },
+    {
+      "epoch": 85.85798816568047,
+      "grad_norm": 1.408051609992981,
+      "learning_rate": 0.0001,
+      "loss": 0.0253,
+      "step": 14510
+    },
+    {
+      "epoch": 85.9171597633136,
+      "grad_norm": 1.5418953895568848,
+      "learning_rate": 0.0001,
+      "loss": 0.0251,
+      "step": 14520
+    },
+    {
+      "epoch": 85.97633136094674,
+      "grad_norm": 1.61346435546875,
+      "learning_rate": 0.0001,
+      "loss": 0.0261,
+      "step": 14530
+    },
+    {
+      "epoch": 86.03550295857988,
+      "grad_norm": 2.014265537261963,
+      "learning_rate": 0.0001,
+      "loss": 0.0258,
+      "step": 14540
+    },
+    {
+      "epoch": 86.09467455621302,
+      "grad_norm": 2.0684974193573,
+      "learning_rate": 0.0001,
+      "loss": 0.0277,
+      "step": 14550
+    },
+    {
+      "epoch": 86.15384615384616,
+      "grad_norm": 1.5834964513778687,
+      "learning_rate": 0.0001,
+      "loss": 0.0259,
+      "step": 14560
+    },
+    {
+      "epoch": 86.21301775147928,
+      "grad_norm": 1.7940055131912231,
+      "learning_rate": 0.0001,
+      "loss": 0.0263,
+      "step": 14570
+    },
+    {
+      "epoch": 86.27218934911242,
+      "grad_norm": 1.4121571779251099,
+      "learning_rate": 0.0001,
+      "loss": 0.027,
+      "step": 14580
+    },
+    {
+      "epoch": 86.33136094674556,
+      "grad_norm": 1.7081290483474731,
+      "learning_rate": 0.0001,
+      "loss": 0.0282,
+      "step": 14590
+    },
+    {
+      "epoch": 86.3905325443787,
+      "grad_norm": 2.3601887226104736,
+      "learning_rate": 0.0001,
+      "loss": 0.0276,
+      "step": 14600
+    },
+    {
+      "epoch": 86.44970414201184,
+      "grad_norm": 2.329240560531616,
+      "learning_rate": 0.0001,
+      "loss": 0.0272,
+      "step": 14610
+    },
+    {
+      "epoch": 86.50887573964496,
+      "grad_norm": 1.8769017457962036,
+      "learning_rate": 0.0001,
+      "loss": 0.027,
+      "step": 14620
+    },
+    {
+      "epoch": 86.5680473372781,
+      "grad_norm": 1.986582636833191,
+      "learning_rate": 0.0001,
+      "loss": 0.0278,
+      "step": 14630
+    },
+    {
+      "epoch": 86.62721893491124,
+      "grad_norm": 1.5818064212799072,
+      "learning_rate": 0.0001,
+      "loss": 0.0262,
+      "step": 14640
+    },
+    {
+      "epoch": 86.68639053254438,
+      "grad_norm": 1.9717246294021606,
+      "learning_rate": 0.0001,
+      "loss": 0.025,
+      "step": 14650
+    },
+    {
+      "epoch": 86.74556213017752,
+      "grad_norm": 2.16619873046875,
+      "learning_rate": 0.0001,
+      "loss": 0.026,
+      "step": 14660
+    },
+    {
+      "epoch": 86.80473372781066,
+      "grad_norm": 1.9293133020401,
+      "learning_rate": 0.0001,
+      "loss": 0.0252,
+      "step": 14670
+    },
+    {
+      "epoch": 86.86390532544378,
+      "grad_norm": 1.632278561592102,
+      "learning_rate": 0.0001,
+      "loss": 0.0253,
+      "step": 14680
+    },
+    {
+      "epoch": 86.92307692307692,
+      "grad_norm": 2.0721118450164795,
+      "learning_rate": 0.0001,
+      "loss": 0.0251,
+      "step": 14690
+    },
+    {
+      "epoch": 86.98224852071006,
+      "grad_norm": 1.5310566425323486,
+      "learning_rate": 0.0001,
+      "loss": 0.0244,
+      "step": 14700
+    },
+    {
+      "epoch": 87.0414201183432,
+      "grad_norm": 1.7277806997299194,
+      "learning_rate": 0.0001,
+      "loss": 0.0249,
+      "step": 14710
+    },
+    {
+      "epoch": 87.10059171597634,
+      "grad_norm": 1.5306246280670166,
+      "learning_rate": 0.0001,
+      "loss": 0.0254,
+      "step": 14720
+    },
+    {
+      "epoch": 87.15976331360947,
+      "grad_norm": 1.6462931632995605,
+      "learning_rate": 0.0001,
+      "loss": 0.0241,
+      "step": 14730
+    },
+    {
+      "epoch": 87.2189349112426,
+      "grad_norm": 1.497218132019043,
+      "learning_rate": 0.0001,
+      "loss": 0.0245,
+      "step": 14740
+    },
+    {
+      "epoch": 87.27810650887574,
+      "grad_norm": 1.5881305932998657,
+      "learning_rate": 0.0001,
+      "loss": 0.0261,
+      "step": 14750
+    },
+    {
+      "epoch": 87.33727810650888,
+      "grad_norm": 1.7341984510421753,
+      "learning_rate": 0.0001,
+      "loss": 0.0255,
+      "step": 14760
+    },
+    {
+      "epoch": 87.39644970414201,
+      "grad_norm": 1.8004612922668457,
+      "learning_rate": 0.0001,
+      "loss": 0.0262,
+      "step": 14770
+    },
+    {
+      "epoch": 87.45562130177515,
+      "grad_norm": 1.628067970275879,
+      "learning_rate": 0.0001,
+      "loss": 0.0243,
+      "step": 14780
+    },
+    {
+      "epoch": 87.51479289940828,
+      "grad_norm": 2.2272181510925293,
+      "learning_rate": 0.0001,
+      "loss": 0.025,
+      "step": 14790
+    },
+    {
+      "epoch": 87.57396449704142,
+      "grad_norm": 1.926284670829773,
+      "learning_rate": 0.0001,
+      "loss": 0.0256,
+      "step": 14800
+    },
+    {
+      "epoch": 87.63313609467455,
+      "grad_norm": 1.612046718597412,
+      "learning_rate": 0.0001,
+      "loss": 0.0252,
+      "step": 14810
+    },
+    {
+      "epoch": 87.6923076923077,
+      "grad_norm": 1.3398481607437134,
+      "learning_rate": 0.0001,
+      "loss": 0.0254,
+      "step": 14820
+    },
+    {
+      "epoch": 87.75147928994083,
+      "grad_norm": 1.1910611391067505,
+      "learning_rate": 0.0001,
+      "loss": 0.0242,
+      "step": 14830
+    },
+    {
+      "epoch": 87.81065088757397,
+      "grad_norm": 1.7203288078308105,
+      "learning_rate": 0.0001,
+      "loss": 0.0259,
+      "step": 14840
+    },
+    {
+      "epoch": 87.8698224852071,
+      "grad_norm": 1.2025917768478394,
+      "learning_rate": 0.0001,
+      "loss": 0.0261,
+      "step": 14850
+    },
+    {
+      "epoch": 87.92899408284023,
+      "grad_norm": 1.6855958700180054,
+      "learning_rate": 0.0001,
+      "loss": 0.0262,
+      "step": 14860
+    },
+    {
+      "epoch": 87.98816568047337,
+      "grad_norm": 1.486555814743042,
+      "learning_rate": 0.0001,
+      "loss": 0.027,
+      "step": 14870
+    },
+    {
+      "epoch": 88.04733727810651,
+      "grad_norm": 1.8529131412506104,
+      "learning_rate": 0.0001,
+      "loss": 0.0253,
+      "step": 14880
+    },
+    {
+      "epoch": 88.10650887573965,
+      "grad_norm": 1.785558819770813,
+      "learning_rate": 0.0001,
+      "loss": 0.0268,
+      "step": 14890
+    },
+    {
+      "epoch": 88.16568047337279,
+      "grad_norm": 2.4225378036499023,
+      "learning_rate": 0.0001,
+      "loss": 0.0274,
+      "step": 14900
+    },
+    {
+      "epoch": 88.22485207100591,
+      "grad_norm": 2.1440465450286865,
+      "learning_rate": 0.0001,
+      "loss": 0.0286,
+      "step": 14910
+    },
+    {
+      "epoch": 88.28402366863905,
+      "grad_norm": 1.602925181388855,
+      "learning_rate": 0.0001,
+      "loss": 0.0287,
+      "step": 14920
+    },
+    {
+      "epoch": 88.34319526627219,
+      "grad_norm": 1.6938937902450562,
+      "learning_rate": 0.0001,
+      "loss": 0.0291,
+      "step": 14930
+    },
+    {
+      "epoch": 88.40236686390533,
+      "grad_norm": 1.405253529548645,
+      "learning_rate": 0.0001,
+      "loss": 0.0279,
+      "step": 14940
+    },
+    {
+      "epoch": 88.46153846153847,
+      "grad_norm": 1.9767032861709595,
+      "learning_rate": 0.0001,
+      "loss": 0.0269,
+      "step": 14950
+    },
+    {
+      "epoch": 88.52071005917159,
+      "grad_norm": 2.1344337463378906,
+      "learning_rate": 0.0001,
+      "loss": 0.0263,
+      "step": 14960
+    },
+    {
+      "epoch": 88.57988165680473,
+      "grad_norm": 1.893290638923645,
+      "learning_rate": 0.0001,
+      "loss": 0.0272,
+      "step": 14970
+    },
+    {
+      "epoch": 88.63905325443787,
+      "grad_norm": 1.3925150632858276,
+      "learning_rate": 0.0001,
+      "loss": 0.0255,
+      "step": 14980
+    },
+    {
+      "epoch": 88.69822485207101,
+      "grad_norm": 1.7590032815933228,
+      "learning_rate": 0.0001,
+      "loss": 0.0256,
+      "step": 14990
+    },
+    {
+      "epoch": 88.75739644970415,
+      "grad_norm": 1.1528652906417847,
+      "learning_rate": 0.0001,
+      "loss": 0.0267,
+      "step": 15000
+    },
+    {
+      "epoch": 88.81656804733728,
+      "grad_norm": 1.2265241146087646,
+      "learning_rate": 0.0001,
+      "loss": 0.0258,
+      "step": 15010
+    },
+    {
+      "epoch": 88.87573964497041,
+      "grad_norm": 1.762597680091858,
+      "learning_rate": 0.0001,
+      "loss": 0.0264,
+      "step": 15020
+    },
+    {
+      "epoch": 88.93491124260355,
+      "grad_norm": 1.8456777334213257,
+      "learning_rate": 0.0001,
+      "loss": 0.0274,
+      "step": 15030
+    },
+    {
+      "epoch": 88.99408284023669,
+      "grad_norm": 1.490128755569458,
+      "learning_rate": 0.0001,
+      "loss": 0.028,
+      "step": 15040
+    },
+    {
+      "epoch": 89.05325443786982,
+      "grad_norm": 1.5456863641738892,
+      "learning_rate": 0.0001,
+      "loss": 0.0271,
+      "step": 15050
+    },
+    {
+      "epoch": 89.11242603550296,
+      "grad_norm": 1.4380412101745605,
+      "learning_rate": 0.0001,
+      "loss": 0.0276,
+      "step": 15060
+    },
+    {
+      "epoch": 89.17159763313609,
+      "grad_norm": 2.0881528854370117,
+      "learning_rate": 0.0001,
+      "loss": 0.0286,
+      "step": 15070
+    },
+    {
+      "epoch": 89.23076923076923,
+      "grad_norm": 1.465851068496704,
+      "learning_rate": 0.0001,
+      "loss": 0.0266,
+      "step": 15080
+    },
+    {
+      "epoch": 89.28994082840237,
+      "grad_norm": 1.3556605577468872,
+      "learning_rate": 0.0001,
+      "loss": 0.0252,
+      "step": 15090
+    },
+    {
+      "epoch": 89.3491124260355,
+      "grad_norm": 1.8683713674545288,
+      "learning_rate": 0.0001,
+      "loss": 0.0256,
+      "step": 15100
+    },
+    {
+      "epoch": 89.40828402366864,
+      "grad_norm": 1.5283063650131226,
+      "learning_rate": 0.0001,
+      "loss": 0.0255,
+      "step": 15110
+    },
+    {
+      "epoch": 89.46745562130178,
+      "grad_norm": 1.5006719827651978,
+      "learning_rate": 0.0001,
+      "loss": 0.0253,
+      "step": 15120
+    },
+    {
+      "epoch": 89.5266272189349,
+      "grad_norm": 1.5189605951309204,
+      "learning_rate": 0.0001,
+      "loss": 0.0253,
+      "step": 15130
+    },
+    {
+      "epoch": 89.58579881656804,
+      "grad_norm": 1.356449007987976,
+      "learning_rate": 0.0001,
+      "loss": 0.026,
+      "step": 15140
+    },
+    {
+      "epoch": 89.64497041420118,
+      "grad_norm": 1.3553647994995117,
+      "learning_rate": 0.0001,
+      "loss": 0.0267,
+      "step": 15150
+    },
+    {
+      "epoch": 89.70414201183432,
+      "grad_norm": 1.8295836448669434,
+      "learning_rate": 0.0001,
+      "loss": 0.0268,
+      "step": 15160
+    },
+    {
+      "epoch": 89.76331360946746,
+      "grad_norm": 2.240229368209839,
+      "learning_rate": 0.0001,
+      "loss": 0.0281,
+      "step": 15170
+    },
+    {
+      "epoch": 89.8224852071006,
+      "grad_norm": 2.229602575302124,
+      "learning_rate": 0.0001,
+      "loss": 0.0267,
+      "step": 15180
+    },
+    {
+      "epoch": 89.88165680473372,
+      "grad_norm": 1.74336576461792,
+      "learning_rate": 0.0001,
+      "loss": 0.0263,
+      "step": 15190
+    },
+    {
+      "epoch": 89.94082840236686,
+      "grad_norm": 1.2296243906021118,
+      "learning_rate": 0.0001,
+      "loss": 0.026,
+      "step": 15200
+    },
+    {
+      "epoch": 90.0,
+      "grad_norm": 1.0360831022262573,
+      "learning_rate": 0.0001,
+      "loss": 0.0254,
+      "step": 15210
+    },
+    {
+      "epoch": 90.05917159763314,
+      "grad_norm": 1.720502495765686,
+      "learning_rate": 0.0001,
+      "loss": 0.025,
+      "step": 15220
+    },
+    {
+      "epoch": 90.11834319526628,
+      "grad_norm": 1.4228390455245972,
+      "learning_rate": 0.0001,
+      "loss": 0.025,
+      "step": 15230
+    },
+    {
+      "epoch": 90.1775147928994,
+      "grad_norm": 1.5123862028121948,
+      "learning_rate": 0.0001,
+      "loss": 0.0259,
+      "step": 15240
+    },
+    {
+      "epoch": 90.23668639053254,
+      "grad_norm": 1.8659727573394775,
+      "learning_rate": 0.0001,
+      "loss": 0.0259,
+      "step": 15250
+    },
+    {
+      "epoch": 90.29585798816568,
+      "grad_norm": 1.8817659616470337,
+      "learning_rate": 0.0001,
+      "loss": 0.0272,
+      "step": 15260
+    },
+    {
+      "epoch": 90.35502958579882,
+      "grad_norm": 0.9984346032142639,
+      "learning_rate": 0.0001,
+      "loss": 0.0259,
+      "step": 15270
+    },
+    {
+      "epoch": 90.41420118343196,
+      "grad_norm": 1.743518590927124,
+      "learning_rate": 0.0001,
+      "loss": 0.0265,
+      "step": 15280
+    },
+    {
+      "epoch": 90.4733727810651,
+      "grad_norm": 1.4117923974990845,
+      "learning_rate": 0.0001,
+      "loss": 0.0263,
+      "step": 15290
+    },
+    {
+      "epoch": 90.53254437869822,
+      "grad_norm": 2.041005849838257,
+      "learning_rate": 0.0001,
+      "loss": 0.0266,
+      "step": 15300
+    },
+    {
+      "epoch": 90.59171597633136,
+      "grad_norm": 2.0724732875823975,
+      "learning_rate": 0.0001,
+      "loss": 0.0271,
+      "step": 15310
+    },
+    {
+      "epoch": 90.6508875739645,
+      "grad_norm": 1.918570876121521,
+      "learning_rate": 0.0001,
+      "loss": 0.027,
+      "step": 15320
+    },
+    {
+      "epoch": 90.71005917159763,
+      "grad_norm": 1.442788004875183,
+      "learning_rate": 0.0001,
+      "loss": 0.026,
+      "step": 15330
+    },
+    {
+      "epoch": 90.76923076923077,
+      "grad_norm": 1.277442216873169,
+      "learning_rate": 0.0001,
+      "loss": 0.0252,
+      "step": 15340
+    },
+    {
+      "epoch": 90.82840236686391,
+      "grad_norm": 1.577277421951294,
+      "learning_rate": 0.0001,
+      "loss": 0.0255,
+      "step": 15350
+    },
+    {
+      "epoch": 90.88757396449704,
+      "grad_norm": 1.7573376893997192,
+      "learning_rate": 0.0001,
+      "loss": 0.0253,
+      "step": 15360
+    },
+    {
+      "epoch": 90.94674556213018,
+      "grad_norm": 1.322043538093567,
+      "learning_rate": 0.0001,
+      "loss": 0.0246,
+      "step": 15370
+    },
+    {
+      "epoch": 91.00591715976331,
+      "grad_norm": 2.215698719024658,
+      "learning_rate": 0.0001,
+      "loss": 0.0245,
+      "step": 15380
+    },
+    {
+      "epoch": 91.06508875739645,
+      "grad_norm": 1.3488421440124512,
+      "learning_rate": 0.0001,
+      "loss": 0.026,
+      "step": 15390
+    },
+    {
+      "epoch": 91.12426035502959,
+      "grad_norm": 1.55535089969635,
+      "learning_rate": 0.0001,
+      "loss": 0.0256,
+      "step": 15400
+    },
+    {
+      "epoch": 91.18343195266272,
+      "grad_norm": 1.513114094734192,
+      "learning_rate": 0.0001,
+      "loss": 0.0251,
+      "step": 15410
+    },
+    {
+      "epoch": 91.24260355029585,
+      "grad_norm": 1.2715460062026978,
+      "learning_rate": 0.0001,
+      "loss": 0.0257,
+      "step": 15420
+    },
+    {
+      "epoch": 91.30177514792899,
+      "grad_norm": 1.6839754581451416,
+      "learning_rate": 0.0001,
+      "loss": 0.0256,
+      "step": 15430
+    },
+    {
+      "epoch": 91.36094674556213,
+      "grad_norm": 1.6173697710037231,
+      "learning_rate": 0.0001,
+      "loss": 0.0266,
+      "step": 15440
+    },
+    {
+      "epoch": 91.42011834319527,
+      "grad_norm": 2.6134190559387207,
+      "learning_rate": 0.0001,
+      "loss": 0.0257,
+      "step": 15450
+    },
+    {
+      "epoch": 91.47928994082841,
+      "grad_norm": 2.4632530212402344,
+      "learning_rate": 0.0001,
+      "loss": 0.0256,
+      "step": 15460
+    },
+    {
+      "epoch": 91.53846153846153,
+      "grad_norm": 1.8501228094100952,
+      "learning_rate": 0.0001,
+      "loss": 0.0256,
+      "step": 15470
+    },
+    {
+      "epoch": 91.59763313609467,
+      "grad_norm": 1.1356680393218994,
+      "learning_rate": 0.0001,
+      "loss": 0.0261,
+      "step": 15480
+    },
+    {
+      "epoch": 91.65680473372781,
+      "grad_norm": 1.7547487020492554,
+      "learning_rate": 0.0001,
+      "loss": 0.0258,
+      "step": 15490
+    },
+    {
+      "epoch": 91.71597633136095,
+      "grad_norm": 1.3695405721664429,
+      "learning_rate": 0.0001,
+      "loss": 0.0264,
+      "step": 15500
+    },
+    {
+      "epoch": 91.77514792899409,
+      "grad_norm": 1.331587553024292,
+      "learning_rate": 0.0001,
+      "loss": 0.0264,
+      "step": 15510
+    },
+    {
+      "epoch": 91.83431952662721,
+      "grad_norm": 1.3338955640792847,
+      "learning_rate": 0.0001,
+      "loss": 0.0256,
+      "step": 15520
+    },
+    {
+      "epoch": 91.89349112426035,
+      "grad_norm": 1.4382702112197876,
+      "learning_rate": 0.0001,
+      "loss": 0.0254,
+      "step": 15530
+    },
+    {
+      "epoch": 91.95266272189349,
+      "grad_norm": 1.5058112144470215,
+      "learning_rate": 0.0001,
+      "loss": 0.0267,
+      "step": 15540
+    },
+    {
+      "epoch": 92.01183431952663,
+      "grad_norm": 1.6571437120437622,
+      "learning_rate": 0.0001,
+      "loss": 0.0261,
+      "step": 15550
+    },
+    {
+      "epoch": 92.07100591715977,
+      "grad_norm": 1.544205665588379,
+      "learning_rate": 0.0001,
+      "loss": 0.0264,
+      "step": 15560
+    },
+    {
+      "epoch": 92.1301775147929,
+      "grad_norm": 1.3698023557662964,
+      "learning_rate": 0.0001,
+      "loss": 0.0261,
+      "step": 15570
+    },
+    {
+      "epoch": 92.18934911242603,
+      "grad_norm": 1.9873988628387451,
+      "learning_rate": 0.0001,
+      "loss": 0.0269,
+      "step": 15580
+    },
+    {
+      "epoch": 92.24852071005917,
+      "grad_norm": 1.8890694379806519,
+      "learning_rate": 0.0001,
+      "loss": 0.0262,
+      "step": 15590
+    },
+    {
+      "epoch": 92.3076923076923,
+      "grad_norm": 1.4275931119918823,
+      "learning_rate": 0.0001,
+      "loss": 0.0253,
+      "step": 15600
+    },
+    {
+      "epoch": 92.36686390532545,
+      "grad_norm": 1.0295602083206177,
+      "learning_rate": 0.0001,
+      "loss": 0.0259,
+      "step": 15610
+    },
+    {
+      "epoch": 92.42603550295858,
+      "grad_norm": 1.6218546628952026,
+      "learning_rate": 0.0001,
+      "loss": 0.0243,
+      "step": 15620
+    },
+    {
+      "epoch": 92.48520710059172,
+      "grad_norm": 1.7823566198349,
+      "learning_rate": 0.0001,
+      "loss": 0.0255,
+      "step": 15630
+    },
+    {
+      "epoch": 92.54437869822485,
+      "grad_norm": 1.863084077835083,
+      "learning_rate": 0.0001,
+      "loss": 0.0255,
+      "step": 15640
+    },
+    {
+      "epoch": 92.60355029585799,
+      "grad_norm": 1.3952381610870361,
+      "learning_rate": 0.0001,
+      "loss": 0.0249,
+      "step": 15650
+    },
+    {
+      "epoch": 92.66272189349112,
+      "grad_norm": 1.633943796157837,
+      "learning_rate": 0.0001,
+      "loss": 0.0255,
+      "step": 15660
+    },
+    {
+      "epoch": 92.72189349112426,
+      "grad_norm": 1.773621678352356,
+      "learning_rate": 0.0001,
+      "loss": 0.0251,
+      "step": 15670
+    },
+    {
+      "epoch": 92.7810650887574,
+      "grad_norm": 1.6607987880706787,
+      "learning_rate": 0.0001,
+      "loss": 0.0265,
+      "step": 15680
+    },
+    {
+      "epoch": 92.84023668639053,
+      "grad_norm": 1.545490026473999,
+      "learning_rate": 0.0001,
+      "loss": 0.0248,
+      "step": 15690
+    },
+    {
+      "epoch": 92.89940828402366,
+      "grad_norm": 1.5495250225067139,
+      "learning_rate": 0.0001,
+      "loss": 0.0255,
+      "step": 15700
+    },
+    {
+      "epoch": 92.9585798816568,
+      "grad_norm": 1.665964961051941,
+      "learning_rate": 0.0001,
+      "loss": 0.026,
+      "step": 15710
+    },
+    {
+      "epoch": 93.01775147928994,
+      "grad_norm": 1.737145185470581,
+      "learning_rate": 0.0001,
+      "loss": 0.0256,
+      "step": 15720
+    },
+    {
+      "epoch": 93.07692307692308,
+      "grad_norm": 1.269252896308899,
+      "learning_rate": 0.0001,
+      "loss": 0.0249,
+      "step": 15730
+    },
+    {
+      "epoch": 93.13609467455622,
+      "grad_norm": 1.8183435201644897,
+      "learning_rate": 0.0001,
+      "loss": 0.0257,
+      "step": 15740
+    },
+    {
+      "epoch": 93.19526627218934,
+      "grad_norm": 1.5960513353347778,
+      "learning_rate": 0.0001,
+      "loss": 0.0263,
+      "step": 15750
+    },
+    {
+      "epoch": 93.25443786982248,
+      "grad_norm": 1.920259952545166,
+      "learning_rate": 0.0001,
+      "loss": 0.0269,
+      "step": 15760
+    },
+    {
+      "epoch": 93.31360946745562,
+      "grad_norm": 1.5223828554153442,
+      "learning_rate": 0.0001,
+      "loss": 0.0262,
+      "step": 15770
+    },
+    {
+      "epoch": 93.37278106508876,
+      "grad_norm": 1.6080313920974731,
+      "learning_rate": 0.0001,
+      "loss": 0.0256,
+      "step": 15780
+    },
+    {
+      "epoch": 93.4319526627219,
+      "grad_norm": 1.2926435470581055,
+      "learning_rate": 0.0001,
+      "loss": 0.0259,
+      "step": 15790
+    },
+    {
+      "epoch": 93.49112426035504,
+      "grad_norm": 1.9077039957046509,
+      "learning_rate": 0.0001,
+      "loss": 0.0272,
+      "step": 15800
+    },
+    {
+      "epoch": 93.55029585798816,
+      "grad_norm": 1.4424465894699097,
+      "learning_rate": 0.0001,
+      "loss": 0.0265,
+      "step": 15810
+    },
+    {
+      "epoch": 93.6094674556213,
+      "grad_norm": 1.639133095741272,
+      "learning_rate": 0.0001,
+      "loss": 0.0254,
+      "step": 15820
+    },
+    {
+      "epoch": 93.66863905325444,
+      "grad_norm": 1.536895751953125,
+      "learning_rate": 0.0001,
+      "loss": 0.0261,
+      "step": 15830
+    },
+    {
+      "epoch": 93.72781065088758,
+      "grad_norm": 1.759664535522461,
+      "learning_rate": 0.0001,
+      "loss": 0.0259,
+      "step": 15840
+    },
+    {
+      "epoch": 93.78698224852072,
+      "grad_norm": 1.968985676765442,
+      "learning_rate": 0.0001,
+      "loss": 0.0256,
+      "step": 15850
+    },
+    {
+      "epoch": 93.84615384615384,
+      "grad_norm": 1.962324857711792,
+      "learning_rate": 0.0001,
+      "loss": 0.0252,
+      "step": 15860
+    },
+    {
+      "epoch": 93.90532544378698,
+      "grad_norm": 1.8022102117538452,
+      "learning_rate": 0.0001,
+      "loss": 0.0253,
+      "step": 15870
+    },
+    {
+      "epoch": 93.96449704142012,
+      "grad_norm": 1.3529881238937378,
+      "learning_rate": 0.0001,
+      "loss": 0.0247,
+      "step": 15880
+    },
+    {
+      "epoch": 94.02366863905326,
+      "grad_norm": 2.125988245010376,
+      "learning_rate": 0.0001,
+      "loss": 0.0259,
+      "step": 15890
+    },
+    {
+      "epoch": 94.0828402366864,
+      "grad_norm": 1.6225851774215698,
+      "learning_rate": 0.0001,
+      "loss": 0.0251,
+      "step": 15900
+    },
+    {
+      "epoch": 94.14201183431953,
+      "grad_norm": 1.2988731861114502,
+      "learning_rate": 0.0001,
+      "loss": 0.0242,
+      "step": 15910
+    },
+    {
+      "epoch": 94.20118343195266,
+      "grad_norm": 1.2784271240234375,
+      "learning_rate": 0.0001,
+      "loss": 0.0248,
+      "step": 15920
+    },
+    {
+      "epoch": 94.2603550295858,
+      "grad_norm": 1.8139183521270752,
+      "learning_rate": 0.0001,
+      "loss": 0.0248,
+      "step": 15930
+    },
+    {
+      "epoch": 94.31952662721893,
+      "grad_norm": 1.68779718875885,
+      "learning_rate": 0.0001,
+      "loss": 0.0248,
+      "step": 15940
+    },
+    {
+      "epoch": 94.37869822485207,
+      "grad_norm": 1.6746162176132202,
+      "learning_rate": 0.0001,
+      "loss": 0.0263,
+      "step": 15950
+    },
+    {
+      "epoch": 94.43786982248521,
+      "grad_norm": 1.7047228813171387,
+      "learning_rate": 0.0001,
+      "loss": 0.0254,
+      "step": 15960
+    },
+    {
+      "epoch": 94.49704142011835,
+      "grad_norm": 1.4197998046875,
+      "learning_rate": 0.0001,
+      "loss": 0.0252,
+      "step": 15970
+    },
+    {
+      "epoch": 94.55621301775147,
+      "grad_norm": 1.6975795030593872,
+      "learning_rate": 0.0001,
+      "loss": 0.0249,
+      "step": 15980
+    },
+    {
+      "epoch": 94.61538461538461,
+      "grad_norm": 2.049863338470459,
+      "learning_rate": 0.0001,
+      "loss": 0.0257,
+      "step": 15990
+    },
+    {
+      "epoch": 94.67455621301775,
+      "grad_norm": 1.931198239326477,
+      "learning_rate": 0.0001,
+      "loss": 0.026,
+      "step": 16000
+    },
+    {
+      "epoch": 94.73372781065089,
+      "grad_norm": 2.00770902633667,
+      "learning_rate": 0.0001,
+      "loss": 0.0248,
+      "step": 16010
+    },
+    {
+      "epoch": 94.79289940828403,
+      "grad_norm": 1.6842414140701294,
+      "learning_rate": 0.0001,
+      "loss": 0.0258,
+      "step": 16020
+    },
+    {
+      "epoch": 94.85207100591715,
+      "grad_norm": 1.166480302810669,
+      "learning_rate": 0.0001,
+      "loss": 0.0247,
+      "step": 16030
+    },
+    {
+      "epoch": 94.91124260355029,
+      "grad_norm": 1.2624388933181763,
+      "learning_rate": 0.0001,
+      "loss": 0.0256,
+      "step": 16040
+    },
+    {
+      "epoch": 94.97041420118343,
+      "grad_norm": 1.4022231101989746,
+      "learning_rate": 0.0001,
+      "loss": 0.0261,
+      "step": 16050
+    },
+    {
+      "epoch": 95.02958579881657,
+      "grad_norm": 1.722373366355896,
+      "learning_rate": 0.0001,
+      "loss": 0.0267,
+      "step": 16060
+    },
+    {
+      "epoch": 95.08875739644971,
+      "grad_norm": 1.4325075149536133,
+      "learning_rate": 0.0001,
+      "loss": 0.027,
+      "step": 16070
+    },
+    {
+      "epoch": 95.14792899408285,
+      "grad_norm": 2.596135139465332,
+      "learning_rate": 0.0001,
+      "loss": 0.0261,
+      "step": 16080
+    },
+    {
+      "epoch": 95.20710059171597,
+      "grad_norm": 2.249368190765381,
+      "learning_rate": 0.0001,
+      "loss": 0.0252,
+      "step": 16090
+    },
+    {
+      "epoch": 95.26627218934911,
+      "grad_norm": 1.7726277112960815,
+      "learning_rate": 0.0001,
+      "loss": 0.027,
+      "step": 16100
+    },
+    {
+      "epoch": 95.32544378698225,
+      "grad_norm": 1.5229803323745728,
+      "learning_rate": 0.0001,
+      "loss": 0.0247,
+      "step": 16110
+    },
+    {
+      "epoch": 95.38461538461539,
+      "grad_norm": 1.19990873336792,
+      "learning_rate": 0.0001,
+      "loss": 0.0242,
+      "step": 16120
+    },
+    {
+      "epoch": 95.44378698224853,
+      "grad_norm": 1.5664896965026855,
+      "learning_rate": 0.0001,
+      "loss": 0.0247,
+      "step": 16130
+    },
+    {
+      "epoch": 95.50295857988165,
+      "grad_norm": 0.9906611442565918,
+      "learning_rate": 0.0001,
+      "loss": 0.0247,
+      "step": 16140
+    },
+    {
+      "epoch": 95.56213017751479,
+      "grad_norm": 1.9654383659362793,
+      "learning_rate": 0.0001,
+      "loss": 0.0253,
+      "step": 16150
+    },
+    {
+      "epoch": 95.62130177514793,
+      "grad_norm": 0.9442912936210632,
+      "learning_rate": 0.0001,
+      "loss": 0.0255,
+      "step": 16160
+    },
+    {
+      "epoch": 95.68047337278107,
+      "grad_norm": 1.7974833250045776,
+      "learning_rate": 0.0001,
+      "loss": 0.0253,
+      "step": 16170
+    },
+    {
+      "epoch": 95.7396449704142,
+      "grad_norm": 1.3414286375045776,
+      "learning_rate": 0.0001,
+      "loss": 0.0261,
+      "step": 16180
+    },
+    {
+      "epoch": 95.79881656804734,
+      "grad_norm": 1.3341009616851807,
+      "learning_rate": 0.0001,
+      "loss": 0.0241,
+      "step": 16190
+    },
+    {
+      "epoch": 95.85798816568047,
+      "grad_norm": 1.3767280578613281,
+      "learning_rate": 0.0001,
+      "loss": 0.0246,
+      "step": 16200
+    },
+    {
+      "epoch": 95.9171597633136,
+      "grad_norm": 1.2226334810256958,
+      "learning_rate": 0.0001,
+      "loss": 0.0247,
+      "step": 16210
+    },
+    {
+      "epoch": 95.97633136094674,
+      "grad_norm": 1.475182294845581,
+      "learning_rate": 0.0001,
+      "loss": 0.0247,
+      "step": 16220
+    },
+    {
+      "epoch": 96.03550295857988,
+      "grad_norm": 1.322830080986023,
+      "learning_rate": 0.0001,
+      "loss": 0.0249,
+      "step": 16230
+    },
+    {
+      "epoch": 96.09467455621302,
+      "grad_norm": 1.3341760635375977,
+      "learning_rate": 0.0001,
+      "loss": 0.0256,
+      "step": 16240
+    },
+    {
+      "epoch": 96.15384615384616,
+      "grad_norm": 1.1770628690719604,
+      "learning_rate": 0.0001,
+      "loss": 0.0255,
+      "step": 16250
+    },
+    {
+      "epoch": 96.21301775147928,
+      "grad_norm": 1.6754480600357056,
+      "learning_rate": 0.0001,
+      "loss": 0.0245,
+      "step": 16260
+    },
+    {
+      "epoch": 96.27218934911242,
+      "grad_norm": 1.9781720638275146,
+      "learning_rate": 0.0001,
+      "loss": 0.0247,
+      "step": 16270
+    },
+    {
+      "epoch": 96.33136094674556,
+      "grad_norm": 1.9889814853668213,
+      "learning_rate": 0.0001,
+      "loss": 0.0248,
+      "step": 16280
+    },
+    {
+      "epoch": 96.3905325443787,
+      "grad_norm": 1.958816409111023,
+      "learning_rate": 0.0001,
+      "loss": 0.025,
+      "step": 16290
+    },
+    {
+      "epoch": 96.44970414201184,
+      "grad_norm": 1.782660961151123,
+      "learning_rate": 0.0001,
+      "loss": 0.0248,
+      "step": 16300
+    },
+    {
+      "epoch": 96.50887573964496,
+      "grad_norm": 1.448146939277649,
+      "learning_rate": 0.0001,
+      "loss": 0.0247,
+      "step": 16310
+    },
+    {
+      "epoch": 96.5680473372781,
+      "grad_norm": 1.855676531791687,
+      "learning_rate": 0.0001,
+      "loss": 0.0249,
+      "step": 16320
+    },
+    {
+      "epoch": 96.62721893491124,
+      "grad_norm": 1.9052342176437378,
+      "learning_rate": 0.0001,
+      "loss": 0.0252,
+      "step": 16330
+    },
+    {
+      "epoch": 96.68639053254438,
+      "grad_norm": 1.545432448387146,
+      "learning_rate": 0.0001,
+      "loss": 0.025,
+      "step": 16340
+    },
+    {
+      "epoch": 96.74556213017752,
+      "grad_norm": 1.9901163578033447,
+      "learning_rate": 0.0001,
+      "loss": 0.0246,
+      "step": 16350
+    },
+    {
+      "epoch": 96.80473372781066,
+      "grad_norm": 2.1851253509521484,
+      "learning_rate": 0.0001,
+      "loss": 0.0245,
+      "step": 16360
+    },
+    {
+      "epoch": 96.86390532544378,
+      "grad_norm": 1.7916219234466553,
+      "learning_rate": 0.0001,
+      "loss": 0.0247,
+      "step": 16370
+    },
+    {
+      "epoch": 96.92307692307692,
+      "grad_norm": 2.010885238647461,
+      "learning_rate": 0.0001,
+      "loss": 0.0259,
+      "step": 16380
+    },
+    {
+      "epoch": 96.98224852071006,
+      "grad_norm": 1.5019755363464355,
+      "learning_rate": 0.0001,
+      "loss": 0.0247,
+      "step": 16390
+    },
+    {
+      "epoch": 97.0414201183432,
+      "grad_norm": 1.4399560689926147,
+      "learning_rate": 0.0001,
+      "loss": 0.0251,
+      "step": 16400
+    },
+    {
+      "epoch": 97.10059171597634,
+      "grad_norm": 1.951032280921936,
+      "learning_rate": 0.0001,
+      "loss": 0.0236,
+      "step": 16410
+    },
+    {
+      "epoch": 97.15976331360947,
+      "grad_norm": 2.2124853134155273,
+      "learning_rate": 0.0001,
+      "loss": 0.024,
+      "step": 16420
+    },
+    {
+      "epoch": 97.2189349112426,
+      "grad_norm": 1.9628318548202515,
+      "learning_rate": 0.0001,
+      "loss": 0.0245,
+      "step": 16430
+    },
+    {
+      "epoch": 97.27810650887574,
+      "grad_norm": 1.538129210472107,
+      "learning_rate": 0.0001,
+      "loss": 0.0241,
+      "step": 16440
+    },
+    {
+      "epoch": 97.33727810650888,
+      "grad_norm": 1.3873004913330078,
+      "learning_rate": 0.0001,
+      "loss": 0.0247,
+      "step": 16450
+    },
+    {
+      "epoch": 97.39644970414201,
+      "grad_norm": 1.9455159902572632,
+      "learning_rate": 0.0001,
+      "loss": 0.0247,
+      "step": 16460
+    },
+    {
+      "epoch": 97.45562130177515,
+      "grad_norm": 1.9581302404403687,
+      "learning_rate": 0.0001,
+      "loss": 0.0253,
+      "step": 16470
+    },
+    {
+      "epoch": 97.51479289940828,
+      "grad_norm": 1.9844659566879272,
+      "learning_rate": 0.0001,
+      "loss": 0.0249,
+      "step": 16480
+    },
+    {
+      "epoch": 97.57396449704142,
+      "grad_norm": 1.956838607788086,
+      "learning_rate": 0.0001,
+      "loss": 0.024,
+      "step": 16490
+    },
+    {
+      "epoch": 97.63313609467455,
+      "grad_norm": 2.2970056533813477,
+      "learning_rate": 0.0001,
+      "loss": 0.025,
+      "step": 16500
+    },
+    {
+      "epoch": 97.6923076923077,
+      "grad_norm": 1.9392861127853394,
+      "learning_rate": 0.0001,
+      "loss": 0.0241,
+      "step": 16510
+    },
+    {
+      "epoch": 97.75147928994083,
+      "grad_norm": 1.5597891807556152,
+      "learning_rate": 0.0001,
+      "loss": 0.0241,
+      "step": 16520
+    },
+    {
+      "epoch": 97.81065088757397,
+      "grad_norm": 2.028132915496826,
+      "learning_rate": 0.0001,
+      "loss": 0.0241,
+      "step": 16530
+    },
+    {
+      "epoch": 97.8698224852071,
+      "grad_norm": 1.5536037683486938,
+      "learning_rate": 0.0001,
+      "loss": 0.0246,
+      "step": 16540
+    },
+    {
+      "epoch": 97.92899408284023,
+      "grad_norm": 1.576088786125183,
+      "learning_rate": 0.0001,
+      "loss": 0.0236,
+      "step": 16550
+    },
+    {
+      "epoch": 97.98816568047337,
+      "grad_norm": 1.6795191764831543,
+      "learning_rate": 0.0001,
+      "loss": 0.0251,
+      "step": 16560
+    },
+    {
+      "epoch": 98.04733727810651,
+      "grad_norm": 1.9462701082229614,
+      "learning_rate": 0.0001,
+      "loss": 0.0237,
+      "step": 16570
+    },
+    {
+      "epoch": 98.10650887573965,
+      "grad_norm": 1.2989611625671387,
+      "learning_rate": 0.0001,
+      "loss": 0.0234,
+      "step": 16580
+    },
+    {
+      "epoch": 98.16568047337279,
+      "grad_norm": 1.315066933631897,
+      "learning_rate": 0.0001,
+      "loss": 0.0239,
+      "step": 16590
+    },
+    {
+      "epoch": 98.22485207100591,
+      "grad_norm": 1.5262898206710815,
+      "learning_rate": 0.0001,
+      "loss": 0.0236,
+      "step": 16600
+    },
+    {
+      "epoch": 98.28402366863905,
+      "grad_norm": 1.2227145433425903,
+      "learning_rate": 0.0001,
+      "loss": 0.0247,
+      "step": 16610
+    },
+    {
+      "epoch": 98.34319526627219,
+      "grad_norm": 1.39451265335083,
+      "learning_rate": 0.0001,
+      "loss": 0.0237,
+      "step": 16620
+    },
+    {
+      "epoch": 98.40236686390533,
+      "grad_norm": 1.239441990852356,
+      "learning_rate": 0.0001,
+      "loss": 0.024,
+      "step": 16630
+    },
+    {
+      "epoch": 98.46153846153847,
+      "grad_norm": 2.6769330501556396,
+      "learning_rate": 0.0001,
+      "loss": 0.0247,
+      "step": 16640
+    },
+    {
+      "epoch": 98.52071005917159,
+      "grad_norm": 2.274594783782959,
+      "learning_rate": 0.0001,
+      "loss": 0.0247,
+      "step": 16650
+    },
+    {
+      "epoch": 98.57988165680473,
+      "grad_norm": 3.101330041885376,
+      "learning_rate": 0.0001,
+      "loss": 0.0252,
+      "step": 16660
+    },
+    {
+      "epoch": 98.63905325443787,
+      "grad_norm": 2.1785764694213867,
+      "learning_rate": 0.0001,
+      "loss": 0.0249,
+      "step": 16670
+    },
+    {
+      "epoch": 98.69822485207101,
+      "grad_norm": 1.7452174425125122,
+      "learning_rate": 0.0001,
+      "loss": 0.0242,
+      "step": 16680
+    },
+    {
+      "epoch": 98.75739644970415,
+      "grad_norm": 1.8460540771484375,
+      "learning_rate": 0.0001,
+      "loss": 0.0236,
+      "step": 16690
+    },
+    {
+      "epoch": 98.81656804733728,
+      "grad_norm": 1.8243061304092407,
+      "learning_rate": 0.0001,
+      "loss": 0.024,
+      "step": 16700
+    },
+    {
+      "epoch": 98.87573964497041,
+      "grad_norm": 1.7397054433822632,
+      "learning_rate": 0.0001,
+      "loss": 0.0237,
+      "step": 16710
+    },
+    {
+      "epoch": 98.93491124260355,
+      "grad_norm": 1.4976600408554077,
+      "learning_rate": 0.0001,
+      "loss": 0.0234,
+      "step": 16720
+    },
+    {
+      "epoch": 98.99408284023669,
+      "grad_norm": 1.6783545017242432,
+      "learning_rate": 0.0001,
+      "loss": 0.0249,
+      "step": 16730
+    },
+    {
+      "epoch": 99.05325443786982,
+      "grad_norm": 1.7942402362823486,
+      "learning_rate": 0.0001,
+      "loss": 0.0255,
+      "step": 16740
+    },
+    {
+      "epoch": 99.11242603550296,
+      "grad_norm": 1.6930099725723267,
+      "learning_rate": 0.0001,
+      "loss": 0.0241,
+      "step": 16750
+    },
+    {
+      "epoch": 99.17159763313609,
+      "grad_norm": 1.3360161781311035,
+      "learning_rate": 0.0001,
+      "loss": 0.0254,
+      "step": 16760
+    },
+    {
+      "epoch": 99.23076923076923,
+      "grad_norm": 1.0786783695220947,
+      "learning_rate": 0.0001,
+      "loss": 0.024,
+      "step": 16770
+    },
+    {
+      "epoch": 99.28994082840237,
+      "grad_norm": 1.542889952659607,
+      "learning_rate": 0.0001,
+      "loss": 0.0253,
+      "step": 16780
+    },
+    {
+      "epoch": 99.3491124260355,
+      "grad_norm": 1.4609732627868652,
+      "learning_rate": 0.0001,
+      "loss": 0.0241,
+      "step": 16790
+    },
+    {
+      "epoch": 99.40828402366864,
+      "grad_norm": 1.2191812992095947,
+      "learning_rate": 0.0001,
+      "loss": 0.0241,
+      "step": 16800
+    },
+    {
+      "epoch": 99.46745562130178,
+      "grad_norm": 1.2641983032226562,
+      "learning_rate": 0.0001,
+      "loss": 0.0239,
+      "step": 16810
+    },
+    {
+      "epoch": 99.5266272189349,
+      "grad_norm": 1.5858970880508423,
+      "learning_rate": 0.0001,
+      "loss": 0.0253,
+      "step": 16820
+    },
+    {
+      "epoch": 99.58579881656804,
+      "grad_norm": 1.58848237991333,
+      "learning_rate": 0.0001,
+      "loss": 0.0259,
+      "step": 16830
+    },
+    {
+      "epoch": 99.64497041420118,
+      "grad_norm": 1.9430514574050903,
+      "learning_rate": 0.0001,
+      "loss": 0.0249,
+      "step": 16840
+    },
+    {
+      "epoch": 99.70414201183432,
+      "grad_norm": 2.0699682235717773,
+      "learning_rate": 0.0001,
+      "loss": 0.025,
+      "step": 16850
+    },
+    {
+      "epoch": 99.76331360946746,
+      "grad_norm": 1.904311180114746,
+      "learning_rate": 0.0001,
+      "loss": 0.0251,
+      "step": 16860
+    },
+    {
+      "epoch": 99.8224852071006,
+      "grad_norm": 1.8055840730667114,
+      "learning_rate": 0.0001,
+      "loss": 0.0239,
+      "step": 16870
+    },
+    {
+      "epoch": 99.88165680473372,
+      "grad_norm": 1.378669261932373,
+      "learning_rate": 0.0001,
+      "loss": 0.0238,
+      "step": 16880
+    },
+    {
+      "epoch": 99.94082840236686,
+      "grad_norm": 2.3051464557647705,
+      "learning_rate": 0.0001,
+      "loss": 0.0239,
+      "step": 16890
+    },
+    {
+      "epoch": 100.0,
+      "grad_norm": 1.7632306814193726,
+      "learning_rate": 0.0001,
+      "loss": 0.0244,
+      "step": 16900
+    },
+    {
+      "epoch": 100.05917159763314,
+      "grad_norm": 2.151419162750244,
+      "learning_rate": 0.0001,
+      "loss": 0.0247,
+      "step": 16910
+    },
+    {
+      "epoch": 100.11834319526628,
+      "grad_norm": 1.9590014219284058,
+      "learning_rate": 0.0001,
+      "loss": 0.026,
+      "step": 16920
+    },
+    {
+      "epoch": 100.1775147928994,
+      "grad_norm": 2.343923330307007,
+      "learning_rate": 0.0001,
+      "loss": 0.026,
+      "step": 16930
+    },
+    {
+      "epoch": 100.23668639053254,
+      "grad_norm": 1.9434462785720825,
+      "learning_rate": 0.0001,
+      "loss": 0.0252,
+      "step": 16940
+    },
+    {
+      "epoch": 100.29585798816568,
+      "grad_norm": 1.5466896295547485,
+      "learning_rate": 0.0001,
+      "loss": 0.0244,
+      "step": 16950
+    },
+    {
+      "epoch": 100.35502958579882,
+      "grad_norm": 1.4434455633163452,
+      "learning_rate": 0.0001,
+      "loss": 0.0249,
+      "step": 16960
+    },
+    {
+      "epoch": 100.41420118343196,
+      "grad_norm": 1.9462908506393433,
+      "learning_rate": 0.0001,
+      "loss": 0.0244,
+      "step": 16970
+    },
+    {
+      "epoch": 100.4733727810651,
+      "grad_norm": 1.6192795038223267,
+      "learning_rate": 0.0001,
+      "loss": 0.024,
+      "step": 16980
+    },
+    {
+      "epoch": 100.53254437869822,
+      "grad_norm": 1.4578166007995605,
+      "learning_rate": 0.0001,
+      "loss": 0.0238,
+      "step": 16990
+    },
+    {
+      "epoch": 100.59171597633136,
+      "grad_norm": 1.9701905250549316,
+      "learning_rate": 0.0001,
+      "loss": 0.0241,
+      "step": 17000
+    },
+    {
+      "epoch": 100.6508875739645,
+      "grad_norm": 1.1881053447723389,
+      "learning_rate": 0.0001,
+      "loss": 0.0242,
+      "step": 17010
+    },
+    {
+      "epoch": 100.71005917159763,
+      "grad_norm": 1.590164065361023,
+      "learning_rate": 0.0001,
+      "loss": 0.0247,
+      "step": 17020
+    },
+    {
+      "epoch": 100.76923076923077,
+      "grad_norm": 1.7912896871566772,
+      "learning_rate": 0.0001,
+      "loss": 0.0245,
+      "step": 17030
+    },
+    {
+      "epoch": 100.82840236686391,
+      "grad_norm": 1.0785051584243774,
+      "learning_rate": 0.0001,
+      "loss": 0.025,
+      "step": 17040
+    },
+    {
+      "epoch": 100.88757396449704,
+      "grad_norm": 1.1836121082305908,
+      "learning_rate": 0.0001,
+      "loss": 0.0256,
+      "step": 17050
+    },
+    {
+      "epoch": 100.94674556213018,
+      "grad_norm": 1.4531973600387573,
+      "learning_rate": 0.0001,
+      "loss": 0.0255,
+      "step": 17060
+    },
+    {
+      "epoch": 101.00591715976331,
+      "grad_norm": 1.3616410493850708,
+      "learning_rate": 0.0001,
+      "loss": 0.0264,
+      "step": 17070
+    },
+    {
+      "epoch": 101.06508875739645,
+      "grad_norm": 1.3781055212020874,
+      "learning_rate": 0.0001,
+      "loss": 0.0264,
+      "step": 17080
+    },
+    {
+      "epoch": 101.12426035502959,
+      "grad_norm": 1.6313138008117676,
+      "learning_rate": 0.0001,
+      "loss": 0.0261,
+      "step": 17090
+    },
+    {
+      "epoch": 101.18343195266272,
+      "grad_norm": 1.5720473527908325,
+      "learning_rate": 0.0001,
+      "loss": 0.0256,
+      "step": 17100
+    },
+    {
+      "epoch": 101.24260355029585,
+      "grad_norm": 1.9254413843154907,
+      "learning_rate": 0.0001,
+      "loss": 0.0257,
+      "step": 17110
+    },
+    {
+      "epoch": 101.30177514792899,
+      "grad_norm": 1.1741353273391724,
+      "learning_rate": 0.0001,
+      "loss": 0.0248,
+      "step": 17120
+    },
+    {
+      "epoch": 101.36094674556213,
+      "grad_norm": 1.80202054977417,
+      "learning_rate": 0.0001,
+      "loss": 0.0256,
+      "step": 17130
+    },
+    {
+      "epoch": 101.42011834319527,
+      "grad_norm": 1.4390785694122314,
+      "learning_rate": 0.0001,
+      "loss": 0.0243,
+      "step": 17140
+    },
+    {
+      "epoch": 101.47928994082841,
+      "grad_norm": 1.337611198425293,
+      "learning_rate": 0.0001,
+      "loss": 0.0249,
+      "step": 17150
+    },
+    {
+      "epoch": 101.53846153846153,
+      "grad_norm": 1.8224354982376099,
+      "learning_rate": 0.0001,
+      "loss": 0.0245,
+      "step": 17160
+    },
+    {
+      "epoch": 101.59763313609467,
+      "grad_norm": 1.7253559827804565,
+      "learning_rate": 0.0001,
+      "loss": 0.0248,
+      "step": 17170
+    },
+    {
+      "epoch": 101.65680473372781,
+      "grad_norm": 1.1495214700698853,
+      "learning_rate": 0.0001,
+      "loss": 0.0247,
+      "step": 17180
+    },
+    {
+      "epoch": 101.71597633136095,
+      "grad_norm": 1.2924764156341553,
+      "learning_rate": 0.0001,
+      "loss": 0.0238,
+      "step": 17190
+    },
+    {
+      "epoch": 101.77514792899409,
+      "grad_norm": 1.7413723468780518,
+      "learning_rate": 0.0001,
+      "loss": 0.0239,
+      "step": 17200
+    },
+    {
+      "epoch": 101.83431952662721,
+      "grad_norm": 1.7160892486572266,
+      "learning_rate": 0.0001,
+      "loss": 0.0245,
+      "step": 17210
+    },
+    {
+      "epoch": 101.89349112426035,
+      "grad_norm": 2.499390125274658,
+      "learning_rate": 0.0001,
+      "loss": 0.0237,
+      "step": 17220
+    },
+    {
+      "epoch": 101.95266272189349,
+      "grad_norm": 1.987691879272461,
+      "learning_rate": 0.0001,
+      "loss": 0.0243,
+      "step": 17230
+    },
+    {
+      "epoch": 102.01183431952663,
+      "grad_norm": 1.4141303300857544,
+      "learning_rate": 0.0001,
+      "loss": 0.0238,
+      "step": 17240
+    },
+    {
+      "epoch": 102.07100591715977,
+      "grad_norm": 1.4520175457000732,
+      "learning_rate": 0.0001,
+      "loss": 0.0242,
+      "step": 17250
+    },
+    {
+      "epoch": 102.1301775147929,
+      "grad_norm": 1.1867307424545288,
+      "learning_rate": 0.0001,
+      "loss": 0.024,
+      "step": 17260
+    },
+    {
+      "epoch": 102.18934911242603,
+      "grad_norm": 1.7476365566253662,
+      "learning_rate": 0.0001,
+      "loss": 0.0248,
+      "step": 17270
+    },
+    {
+      "epoch": 102.24852071005917,
+      "grad_norm": 1.8855823278427124,
+      "learning_rate": 0.0001,
+      "loss": 0.0244,
+      "step": 17280
+    },
+    {
+      "epoch": 102.3076923076923,
+      "grad_norm": 1.1343092918395996,
+      "learning_rate": 0.0001,
+      "loss": 0.0237,
+      "step": 17290
+    },
+    {
+      "epoch": 102.36686390532545,
+      "grad_norm": 1.4090570211410522,
+      "learning_rate": 0.0001,
+      "loss": 0.0244,
+      "step": 17300
+    },
+    {
+      "epoch": 102.42603550295858,
+      "grad_norm": 1.2537813186645508,
+      "learning_rate": 0.0001,
+      "loss": 0.0248,
+      "step": 17310
+    },
+    {
+      "epoch": 102.48520710059172,
+      "grad_norm": 1.4019060134887695,
+      "learning_rate": 0.0001,
+      "loss": 0.0244,
+      "step": 17320
+    },
+    {
+      "epoch": 102.54437869822485,
+      "grad_norm": 1.2202293872833252,
+      "learning_rate": 0.0001,
+      "loss": 0.0255,
+      "step": 17330
+    },
+    {
+      "epoch": 102.60355029585799,
+      "grad_norm": 1.862802505493164,
+      "learning_rate": 0.0001,
+      "loss": 0.0255,
+      "step": 17340
+    },
+    {
+      "epoch": 102.66272189349112,
+      "grad_norm": 1.9269821643829346,
+      "learning_rate": 0.0001,
+      "loss": 0.0253,
+      "step": 17350
+    },
+    {
+      "epoch": 102.72189349112426,
+      "grad_norm": 1.783424735069275,
+      "learning_rate": 0.0001,
+      "loss": 0.0257,
+      "step": 17360
+    },
+    {
+      "epoch": 102.7810650887574,
+      "grad_norm": 1.1931513547897339,
+      "learning_rate": 0.0001,
+      "loss": 0.0247,
+      "step": 17370
+    },
+    {
+      "epoch": 102.84023668639053,
+      "grad_norm": 1.3637630939483643,
+      "learning_rate": 0.0001,
+      "loss": 0.0233,
+      "step": 17380
+    },
+    {
+      "epoch": 102.89940828402366,
+      "grad_norm": 1.4374241828918457,
+      "learning_rate": 0.0001,
+      "loss": 0.0242,
+      "step": 17390
+    },
+    {
+      "epoch": 102.9585798816568,
+      "grad_norm": 1.3271995782852173,
+      "learning_rate": 0.0001,
+      "loss": 0.0253,
+      "step": 17400
+    },
+    {
+      "epoch": 103.01775147928994,
+      "grad_norm": 1.310217022895813,
+      "learning_rate": 0.0001,
+      "loss": 0.025,
+      "step": 17410
+    },
+    {
+      "epoch": 103.07692307692308,
+      "grad_norm": 1.865014672279358,
+      "learning_rate": 0.0001,
+      "loss": 0.0257,
+      "step": 17420
+    },
+    {
+      "epoch": 103.13609467455622,
+      "grad_norm": 1.1619547605514526,
+      "learning_rate": 0.0001,
+      "loss": 0.0252,
+      "step": 17430
+    },
+    {
+      "epoch": 103.19526627218934,
+      "grad_norm": 1.4556515216827393,
+      "learning_rate": 0.0001,
+      "loss": 0.0248,
+      "step": 17440
+    },
+    {
+      "epoch": 103.25443786982248,
+      "grad_norm": 1.465700387954712,
+      "learning_rate": 0.0001,
+      "loss": 0.0248,
+      "step": 17450
+    },
+    {
+      "epoch": 103.31360946745562,
+      "grad_norm": 1.0786173343658447,
+      "learning_rate": 0.0001,
+      "loss": 0.0247,
+      "step": 17460
+    },
+    {
+      "epoch": 103.37278106508876,
+      "grad_norm": 1.276053547859192,
+      "learning_rate": 0.0001,
+      "loss": 0.0255,
+      "step": 17470
+    },
+    {
+      "epoch": 103.4319526627219,
+      "grad_norm": 2.069765329360962,
+      "learning_rate": 0.0001,
+      "loss": 0.0248,
+      "step": 17480
+    },
+    {
+      "epoch": 103.49112426035504,
+      "grad_norm": 1.7892131805419922,
+      "learning_rate": 0.0001,
+      "loss": 0.0253,
+      "step": 17490
+    },
+    {
+      "epoch": 103.55029585798816,
+      "grad_norm": 1.5292075872421265,
+      "learning_rate": 0.0001,
+      "loss": 0.0246,
+      "step": 17500
+    },
+    {
+      "epoch": 103.6094674556213,
+      "grad_norm": 1.5193679332733154,
+      "learning_rate": 0.0001,
+      "loss": 0.026,
+      "step": 17510
+    },
+    {
+      "epoch": 103.66863905325444,
+      "grad_norm": 1.5338925123214722,
+      "learning_rate": 0.0001,
+      "loss": 0.025,
+      "step": 17520
+    },
+    {
+      "epoch": 103.72781065088758,
+      "grad_norm": 1.9597268104553223,
+      "learning_rate": 0.0001,
+      "loss": 0.0248,
+      "step": 17530
+    },
+    {
+      "epoch": 103.78698224852072,
+      "grad_norm": 1.5174331665039062,
+      "learning_rate": 0.0001,
+      "loss": 0.0249,
+      "step": 17540
+    },
+    {
+      "epoch": 103.84615384615384,
+      "grad_norm": 1.5722726583480835,
+      "learning_rate": 0.0001,
+      "loss": 0.0256,
+      "step": 17550
+    },
+    {
+      "epoch": 103.90532544378698,
+      "grad_norm": 1.6332615613937378,
+      "learning_rate": 0.0001,
+      "loss": 0.0245,
+      "step": 17560
+    },
+    {
+      "epoch": 103.96449704142012,
+      "grad_norm": 1.5341461896896362,
+      "learning_rate": 0.0001,
+      "loss": 0.0241,
+      "step": 17570
+    },
+    {
+      "epoch": 104.02366863905326,
+      "grad_norm": 1.4219595193862915,
+      "learning_rate": 0.0001,
+      "loss": 0.0249,
+      "step": 17580
+    },
+    {
+      "epoch": 104.0828402366864,
+      "grad_norm": 2.477701187133789,
+      "learning_rate": 0.0001,
+      "loss": 0.0265,
+      "step": 17590
+    },
+    {
+      "epoch": 104.14201183431953,
+      "grad_norm": 1.8523873090744019,
+      "learning_rate": 0.0001,
+      "loss": 0.0251,
+      "step": 17600
+    },
+    {
+      "epoch": 104.20118343195266,
+      "grad_norm": 1.6446479558944702,
+      "learning_rate": 0.0001,
+      "loss": 0.0253,
+      "step": 17610
+    },
+    {
+      "epoch": 104.2603550295858,
+      "grad_norm": 1.8975907564163208,
+      "learning_rate": 0.0001,
+      "loss": 0.0245,
+      "step": 17620
+    },
+    {
+      "epoch": 104.31952662721893,
+      "grad_norm": 1.4146679639816284,
+      "learning_rate": 0.0001,
+      "loss": 0.0235,
+      "step": 17630
+    },
+    {
+      "epoch": 104.37869822485207,
+      "grad_norm": 1.7020050287246704,
+      "learning_rate": 0.0001,
+      "loss": 0.0241,
+      "step": 17640
+    },
+    {
+      "epoch": 104.43786982248521,
+      "grad_norm": 1.4540374279022217,
+      "learning_rate": 0.0001,
+      "loss": 0.0246,
+      "step": 17650
+    },
+    {
+      "epoch": 104.49704142011835,
+      "grad_norm": 1.7307575941085815,
+      "learning_rate": 0.0001,
+      "loss": 0.0245,
+      "step": 17660
+    },
+    {
+      "epoch": 104.55621301775147,
+      "grad_norm": 1.8799189329147339,
+      "learning_rate": 0.0001,
+      "loss": 0.0245,
+      "step": 17670
+    },
+    {
+      "epoch": 104.61538461538461,
+      "grad_norm": 1.320449948310852,
+      "learning_rate": 0.0001,
+      "loss": 0.0241,
+      "step": 17680
+    },
+    {
+      "epoch": 104.67455621301775,
+      "grad_norm": 1.8316500186920166,
+      "learning_rate": 0.0001,
+      "loss": 0.0236,
+      "step": 17690
+    },
+    {
+      "epoch": 104.73372781065089,
+      "grad_norm": 2.194802761077881,
+      "learning_rate": 0.0001,
+      "loss": 0.025,
+      "step": 17700
+    },
+    {
+      "epoch": 104.79289940828403,
+      "grad_norm": 1.5404913425445557,
+      "learning_rate": 0.0001,
+      "loss": 0.0248,
+      "step": 17710
+    },
+    {
+      "epoch": 104.85207100591715,
+      "grad_norm": 1.1562561988830566,
+      "learning_rate": 0.0001,
+      "loss": 0.0239,
+      "step": 17720
+    },
+    {
+      "epoch": 104.91124260355029,
+      "grad_norm": 1.422286033630371,
+      "learning_rate": 0.0001,
+      "loss": 0.0247,
+      "step": 17730
+    },
+    {
+      "epoch": 104.97041420118343,
+      "grad_norm": 2.047193765640259,
+      "learning_rate": 0.0001,
+      "loss": 0.024,
+      "step": 17740
+    },
+    {
+      "epoch": 105.02958579881657,
+      "grad_norm": 1.0452032089233398,
+      "learning_rate": 0.0001,
+      "loss": 0.0225,
+      "step": 17750
+    },
+    {
+      "epoch": 105.08875739644971,
+      "grad_norm": 1.74553382396698,
+      "learning_rate": 0.0001,
+      "loss": 0.0235,
+      "step": 17760
+    },
+    {
+      "epoch": 105.14792899408285,
+      "grad_norm": 1.6669174432754517,
+      "learning_rate": 0.0001,
+      "loss": 0.0238,
+      "step": 17770
+    },
+    {
+      "epoch": 105.20710059171597,
+      "grad_norm": 1.1358529329299927,
+      "learning_rate": 0.0001,
+      "loss": 0.0239,
+      "step": 17780
+    },
+    {
+      "epoch": 105.26627218934911,
+      "grad_norm": 1.6233428716659546,
+      "learning_rate": 0.0001,
+      "loss": 0.0234,
+      "step": 17790
+    },
+    {
+      "epoch": 105.32544378698225,
+      "grad_norm": 1.325111746788025,
+      "learning_rate": 0.0001,
+      "loss": 0.0242,
+      "step": 17800
+    },
+    {
+      "epoch": 105.38461538461539,
+      "grad_norm": 1.0074485540390015,
+      "learning_rate": 0.0001,
+      "loss": 0.0236,
+      "step": 17810
+    },
+    {
+      "epoch": 105.44378698224853,
+      "grad_norm": 1.584591269493103,
+      "learning_rate": 0.0001,
+      "loss": 0.0247,
+      "step": 17820
+    },
+    {
+      "epoch": 105.50295857988165,
+      "grad_norm": 1.600882649421692,
+      "learning_rate": 0.0001,
+      "loss": 0.0242,
+      "step": 17830
+    },
+    {
+      "epoch": 105.56213017751479,
+      "grad_norm": 1.3158701658248901,
+      "learning_rate": 0.0001,
+      "loss": 0.0236,
+      "step": 17840
+    },
+    {
+      "epoch": 105.62130177514793,
+      "grad_norm": 1.3540611267089844,
+      "learning_rate": 0.0001,
+      "loss": 0.0244,
+      "step": 17850
+    },
+    {
+      "epoch": 105.68047337278107,
+      "grad_norm": 1.5947779417037964,
+      "learning_rate": 0.0001,
+      "loss": 0.024,
+      "step": 17860
+    },
+    {
+      "epoch": 105.7396449704142,
+      "grad_norm": 1.1042803525924683,
+      "learning_rate": 0.0001,
+      "loss": 0.0242,
+      "step": 17870
+    },
+    {
+      "epoch": 105.79881656804734,
+      "grad_norm": 1.5443651676177979,
+      "learning_rate": 0.0001,
+      "loss": 0.0254,
+      "step": 17880
+    },
+    {
+      "epoch": 105.85798816568047,
+      "grad_norm": 1.3998723030090332,
+      "learning_rate": 0.0001,
+      "loss": 0.0255,
+      "step": 17890
+    },
+    {
+      "epoch": 105.9171597633136,
+      "grad_norm": 1.7016702890396118,
+      "learning_rate": 0.0001,
+      "loss": 0.0247,
+      "step": 17900
+    },
+    {
+      "epoch": 105.97633136094674,
+      "grad_norm": 1.555227518081665,
+      "learning_rate": 0.0001,
+      "loss": 0.0244,
+      "step": 17910
+    },
+    {
+      "epoch": 106.03550295857988,
+      "grad_norm": 2.048102378845215,
+      "learning_rate": 0.0001,
+      "loss": 0.0243,
+      "step": 17920
+    },
+    {
+      "epoch": 106.09467455621302,
+      "grad_norm": 1.141298770904541,
+      "learning_rate": 0.0001,
+      "loss": 0.0232,
+      "step": 17930
+    },
+    {
+      "epoch": 106.15384615384616,
+      "grad_norm": 1.4130231142044067,
+      "learning_rate": 0.0001,
+      "loss": 0.023,
+      "step": 17940
+    },
+    {
+      "epoch": 106.21301775147928,
+      "grad_norm": 2.1230781078338623,
+      "learning_rate": 0.0001,
+      "loss": 0.0234,
+      "step": 17950
+    },
+    {
+      "epoch": 106.27218934911242,
+      "grad_norm": 1.8795009851455688,
+      "learning_rate": 0.0001,
+      "loss": 0.0247,
+      "step": 17960
+    },
+    {
+      "epoch": 106.33136094674556,
+      "grad_norm": 1.315991759300232,
+      "learning_rate": 0.0001,
+      "loss": 0.0237,
+      "step": 17970
+    },
+    {
+      "epoch": 106.3905325443787,
+      "grad_norm": 1.6347544193267822,
+      "learning_rate": 0.0001,
+      "loss": 0.0225,
+      "step": 17980
+    },
+    {
+      "epoch": 106.44970414201184,
+      "grad_norm": 1.765291690826416,
+      "learning_rate": 0.0001,
+      "loss": 0.0223,
+      "step": 17990
+    },
+    {
+      "epoch": 106.50887573964496,
+      "grad_norm": 1.670806884765625,
+      "learning_rate": 0.0001,
+      "loss": 0.0239,
+      "step": 18000
+    },
+    {
+      "epoch": 106.5680473372781,
+      "grad_norm": 1.9420770406723022,
+      "learning_rate": 0.0001,
+      "loss": 0.024,
+      "step": 18010
+    },
+    {
+      "epoch": 106.62721893491124,
+      "grad_norm": 1.4127461910247803,
+      "learning_rate": 0.0001,
+      "loss": 0.024,
+      "step": 18020
+    },
+    {
+      "epoch": 106.68639053254438,
+      "grad_norm": 1.4923428297042847,
+      "learning_rate": 0.0001,
+      "loss": 0.024,
+      "step": 18030
+    },
+    {
+      "epoch": 106.74556213017752,
+      "grad_norm": 1.3911137580871582,
+      "learning_rate": 0.0001,
+      "loss": 0.0233,
+      "step": 18040
+    },
+    {
+      "epoch": 106.80473372781066,
+      "grad_norm": 1.0588150024414062,
+      "learning_rate": 0.0001,
+      "loss": 0.024,
+      "step": 18050
+    },
+    {
+      "epoch": 106.86390532544378,
+      "grad_norm": 1.9845844507217407,
+      "learning_rate": 0.0001,
+      "loss": 0.026,
+      "step": 18060
+    },
+    {
+      "epoch": 106.92307692307692,
+      "grad_norm": 1.8139684200286865,
+      "learning_rate": 0.0001,
+      "loss": 0.0238,
+      "step": 18070
+    },
+    {
+      "epoch": 106.98224852071006,
+      "grad_norm": 1.4871219396591187,
+      "learning_rate": 0.0001,
+      "loss": 0.0227,
+      "step": 18080
+    },
+    {
+      "epoch": 107.0414201183432,
+      "grad_norm": 1.9399596452713013,
+      "learning_rate": 0.0001,
+      "loss": 0.0239,
+      "step": 18090
+    },
+    {
+      "epoch": 107.10059171597634,
+      "grad_norm": 1.6418794393539429,
+      "learning_rate": 0.0001,
+      "loss": 0.0228,
+      "step": 18100
+    },
+    {
+      "epoch": 107.15976331360947,
+      "grad_norm": 1.5365206003189087,
+      "learning_rate": 0.0001,
+      "loss": 0.0234,
+      "step": 18110
+    },
+    {
+      "epoch": 107.2189349112426,
+      "grad_norm": 1.9159715175628662,
+      "learning_rate": 0.0001,
+      "loss": 0.0241,
+      "step": 18120
+    },
+    {
+      "epoch": 107.27810650887574,
+      "grad_norm": 1.5416733026504517,
+      "learning_rate": 0.0001,
+      "loss": 0.0226,
+      "step": 18130
+    },
+    {
+      "epoch": 107.33727810650888,
+      "grad_norm": 1.790571928024292,
+      "learning_rate": 0.0001,
+      "loss": 0.0224,
+      "step": 18140
+    },
+    {
+      "epoch": 107.39644970414201,
+      "grad_norm": 1.4058839082717896,
+      "learning_rate": 0.0001,
+      "loss": 0.023,
+      "step": 18150
+    },
+    {
+      "epoch": 107.45562130177515,
+      "grad_norm": 2.026355743408203,
+      "learning_rate": 0.0001,
+      "loss": 0.025,
+      "step": 18160
+    },
+    {
+      "epoch": 107.51479289940828,
+      "grad_norm": 1.9854003190994263,
+      "learning_rate": 0.0001,
+      "loss": 0.0235,
+      "step": 18170
+    },
+    {
+      "epoch": 107.57396449704142,
+      "grad_norm": 2.0986621379852295,
+      "learning_rate": 0.0001,
+      "loss": 0.0239,
+      "step": 18180
+    },
+    {
+      "epoch": 107.63313609467455,
+      "grad_norm": 2.4363503456115723,
+      "learning_rate": 0.0001,
+      "loss": 0.0229,
+      "step": 18190
+    },
+    {
+      "epoch": 107.6923076923077,
+      "grad_norm": 1.718054175376892,
+      "learning_rate": 0.0001,
+      "loss": 0.0228,
+      "step": 18200
+    },
+    {
+      "epoch": 107.75147928994083,
+      "grad_norm": 1.5067126750946045,
+      "learning_rate": 0.0001,
+      "loss": 0.0227,
+      "step": 18210
+    },
+    {
+      "epoch": 107.81065088757397,
+      "grad_norm": 1.4755473136901855,
+      "learning_rate": 0.0001,
+      "loss": 0.0218,
+      "step": 18220
+    },
+    {
+      "epoch": 107.8698224852071,
+      "grad_norm": 1.6640297174453735,
+      "learning_rate": 0.0001,
+      "loss": 0.0223,
+      "step": 18230
+    },
+    {
+      "epoch": 107.92899408284023,
+      "grad_norm": 1.4255433082580566,
+      "learning_rate": 0.0001,
+      "loss": 0.0227,
+      "step": 18240
+    },
+    {
+      "epoch": 107.98816568047337,
+      "grad_norm": 1.2273614406585693,
+      "learning_rate": 0.0001,
+      "loss": 0.0249,
+      "step": 18250
+    },
+    {
+      "epoch": 108.04733727810651,
+      "grad_norm": 2.3989014625549316,
+      "learning_rate": 0.0001,
+      "loss": 0.0244,
+      "step": 18260
+    },
+    {
+      "epoch": 108.10650887573965,
+      "grad_norm": 1.9016281366348267,
+      "learning_rate": 0.0001,
+      "loss": 0.0238,
+      "step": 18270
+    },
+    {
+      "epoch": 108.16568047337279,
+      "grad_norm": 1.5704691410064697,
+      "learning_rate": 0.0001,
+      "loss": 0.0231,
+      "step": 18280
+    },
+    {
+      "epoch": 108.22485207100591,
+      "grad_norm": 2.4850356578826904,
+      "learning_rate": 0.0001,
+      "loss": 0.0242,
+      "step": 18290
+    },
+    {
+      "epoch": 108.28402366863905,
+      "grad_norm": 1.9066758155822754,
+      "learning_rate": 0.0001,
+      "loss": 0.0243,
+      "step": 18300
+    },
+    {
+      "epoch": 108.34319526627219,
+      "grad_norm": 2.046584367752075,
+      "learning_rate": 0.0001,
+      "loss": 0.0228,
+      "step": 18310
+    },
+    {
+      "epoch": 108.40236686390533,
+      "grad_norm": 1.6976393461227417,
+      "learning_rate": 0.0001,
+      "loss": 0.0235,
+      "step": 18320
+    },
+    {
+      "epoch": 108.46153846153847,
+      "grad_norm": 1.76020348072052,
+      "learning_rate": 0.0001,
+      "loss": 0.0237,
+      "step": 18330
+    },
+    {
+      "epoch": 108.52071005917159,
+      "grad_norm": 1.6783928871154785,
+      "learning_rate": 0.0001,
+      "loss": 0.0221,
+      "step": 18340
+    },
+    {
+      "epoch": 108.57988165680473,
+      "grad_norm": 1.2821146249771118,
+      "learning_rate": 0.0001,
+      "loss": 0.0237,
+      "step": 18350
+    },
+    {
+      "epoch": 108.63905325443787,
+      "grad_norm": 2.0235345363616943,
+      "learning_rate": 0.0001,
+      "loss": 0.023,
+      "step": 18360
+    },
+    {
+      "epoch": 108.69822485207101,
+      "grad_norm": 1.801581859588623,
+      "learning_rate": 0.0001,
+      "loss": 0.0237,
+      "step": 18370
+    },
+    {
+      "epoch": 108.75739644970415,
+      "grad_norm": 1.2788413763046265,
+      "learning_rate": 0.0001,
+      "loss": 0.0239,
+      "step": 18380
+    },
+    {
+      "epoch": 108.81656804733728,
+      "grad_norm": 2.101212501525879,
+      "learning_rate": 0.0001,
+      "loss": 0.0235,
+      "step": 18390
+    },
+    {
+      "epoch": 108.87573964497041,
+      "grad_norm": 1.5364439487457275,
+      "learning_rate": 0.0001,
+      "loss": 0.0248,
+      "step": 18400
+    },
+    {
+      "epoch": 108.93491124260355,
+      "grad_norm": 1.5246155261993408,
+      "learning_rate": 0.0001,
+      "loss": 0.0239,
+      "step": 18410
+    },
+    {
+      "epoch": 108.99408284023669,
+      "grad_norm": 1.7899212837219238,
+      "learning_rate": 0.0001,
+      "loss": 0.0237,
+      "step": 18420
+    },
+    {
+      "epoch": 109.05325443786982,
+      "grad_norm": 1.3979517221450806,
+      "learning_rate": 0.0001,
+      "loss": 0.0238,
+      "step": 18430
+    },
+    {
+      "epoch": 109.11242603550296,
+      "grad_norm": 1.5425375699996948,
+      "learning_rate": 0.0001,
+      "loss": 0.024,
+      "step": 18440
+    },
+    {
+      "epoch": 109.17159763313609,
+      "grad_norm": 1.9369560480117798,
+      "learning_rate": 0.0001,
+      "loss": 0.0248,
+      "step": 18450
+    },
+    {
+      "epoch": 109.23076923076923,
+      "grad_norm": 1.2704060077667236,
+      "learning_rate": 0.0001,
+      "loss": 0.0238,
+      "step": 18460
+    },
+    {
+      "epoch": 109.28994082840237,
+      "grad_norm": 1.1898707151412964,
+      "learning_rate": 0.0001,
+      "loss": 0.0227,
+      "step": 18470
+    },
+    {
+      "epoch": 109.3491124260355,
+      "grad_norm": 1.3073447942733765,
+      "learning_rate": 0.0001,
+      "loss": 0.0238,
+      "step": 18480
+    },
+    {
+      "epoch": 109.40828402366864,
+      "grad_norm": 1.3756870031356812,
+      "learning_rate": 0.0001,
+      "loss": 0.0243,
+      "step": 18490
+    },
+    {
+      "epoch": 109.46745562130178,
+      "grad_norm": 1.0849601030349731,
+      "learning_rate": 0.0001,
+      "loss": 0.024,
+      "step": 18500
+    },
+    {
+      "epoch": 109.5266272189349,
+      "grad_norm": 1.7405685186386108,
+      "learning_rate": 0.0001,
+      "loss": 0.024,
+      "step": 18510
+    },
+    {
+      "epoch": 109.58579881656804,
+      "grad_norm": 1.936854600906372,
+      "learning_rate": 0.0001,
+      "loss": 0.0239,
+      "step": 18520
+    },
+    {
+      "epoch": 109.64497041420118,
+      "grad_norm": 1.27566397190094,
+      "learning_rate": 0.0001,
+      "loss": 0.0239,
+      "step": 18530
+    },
+    {
+      "epoch": 109.70414201183432,
+      "grad_norm": 1.3689104318618774,
+      "learning_rate": 0.0001,
+      "loss": 0.0247,
+      "step": 18540
+    },
+    {
+      "epoch": 109.76331360946746,
+      "grad_norm": 2.1343297958374023,
+      "learning_rate": 0.0001,
+      "loss": 0.0248,
+      "step": 18550
+    },
+    {
+      "epoch": 109.8224852071006,
+      "grad_norm": 1.6312379837036133,
+      "learning_rate": 0.0001,
+      "loss": 0.0244,
+      "step": 18560
+    },
+    {
+      "epoch": 109.88165680473372,
+      "grad_norm": 1.2922239303588867,
+      "learning_rate": 0.0001,
+      "loss": 0.0238,
+      "step": 18570
+    },
+    {
+      "epoch": 109.94082840236686,
+      "grad_norm": 1.462161660194397,
+      "learning_rate": 0.0001,
+      "loss": 0.0246,
+      "step": 18580
+    },
+    {
+      "epoch": 110.0,
+      "grad_norm": 1.6690082550048828,
+      "learning_rate": 0.0001,
+      "loss": 0.0249,
+      "step": 18590
+    },
+    {
+      "epoch": 110.05917159763314,
+      "grad_norm": 1.4380525350570679,
+      "learning_rate": 0.0001,
+      "loss": 0.027,
+      "step": 18600
+    },
+    {
+      "epoch": 110.11834319526628,
+      "grad_norm": 1.7187892198562622,
+      "learning_rate": 0.0001,
+      "loss": 0.0259,
+      "step": 18610
+    },
+    {
+      "epoch": 110.1775147928994,
+      "grad_norm": 1.4591492414474487,
+      "learning_rate": 0.0001,
+      "loss": 0.025,
+      "step": 18620
+    },
+    {
+      "epoch": 110.23668639053254,
+      "grad_norm": 1.8756537437438965,
+      "learning_rate": 0.0001,
+      "loss": 0.025,
+      "step": 18630
+    },
+    {
+      "epoch": 110.29585798816568,
+      "grad_norm": 1.5523681640625,
+      "learning_rate": 0.0001,
+      "loss": 0.0239,
+      "step": 18640
+    },
+    {
+      "epoch": 110.35502958579882,
+      "grad_norm": 1.7149626016616821,
+      "learning_rate": 0.0001,
+      "loss": 0.0231,
+      "step": 18650
+    },
+    {
+      "epoch": 110.41420118343196,
+      "grad_norm": 1.826210379600525,
+      "learning_rate": 0.0001,
+      "loss": 0.0232,
+      "step": 18660
+    },
+    {
+      "epoch": 110.4733727810651,
+      "grad_norm": 1.9161325693130493,
+      "learning_rate": 0.0001,
+      "loss": 0.0239,
+      "step": 18670
+    },
+    {
+      "epoch": 110.53254437869822,
+      "grad_norm": 1.450586199760437,
+      "learning_rate": 0.0001,
+      "loss": 0.0231,
+      "step": 18680
+    },
+    {
+      "epoch": 110.59171597633136,
+      "grad_norm": 1.6246062517166138,
+      "learning_rate": 0.0001,
+      "loss": 0.0238,
+      "step": 18690
+    },
+    {
+      "epoch": 110.6508875739645,
+      "grad_norm": 1.3721132278442383,
+      "learning_rate": 0.0001,
+      "loss": 0.0233,
+      "step": 18700
+    },
+    {
+      "epoch": 110.71005917159763,
+      "grad_norm": 1.2404536008834839,
+      "learning_rate": 0.0001,
+      "loss": 0.0229,
+      "step": 18710
+    },
+    {
+      "epoch": 110.76923076923077,
+      "grad_norm": 1.4290307760238647,
+      "learning_rate": 0.0001,
+      "loss": 0.0239,
+      "step": 18720
+    },
+    {
+      "epoch": 110.82840236686391,
+      "grad_norm": 1.4093533754348755,
+      "learning_rate": 0.0001,
+      "loss": 0.024,
+      "step": 18730
+    },
+    {
+      "epoch": 110.88757396449704,
+      "grad_norm": 2.1454269886016846,
+      "learning_rate": 0.0001,
+      "loss": 0.0234,
+      "step": 18740
+    },
+    {
+      "epoch": 110.94674556213018,
+      "grad_norm": 2.4416592121124268,
+      "learning_rate": 0.0001,
+      "loss": 0.0255,
+      "step": 18750
+    },
+    {
+      "epoch": 111.00591715976331,
+      "grad_norm": 2.4591033458709717,
+      "learning_rate": 0.0001,
+      "loss": 0.0251,
+      "step": 18760
+    },
+    {
+      "epoch": 111.06508875739645,
+      "grad_norm": 1.6192554235458374,
+      "learning_rate": 0.0001,
+      "loss": 0.0247,
+      "step": 18770
+    },
+    {
+      "epoch": 111.12426035502959,
+      "grad_norm": 1.5049679279327393,
+      "learning_rate": 0.0001,
+      "loss": 0.0239,
+      "step": 18780
+    },
+    {
+      "epoch": 111.18343195266272,
+      "grad_norm": 0.9855606555938721,
+      "learning_rate": 0.0001,
+      "loss": 0.0237,
+      "step": 18790
+    },
+    {
+      "epoch": 111.24260355029585,
+      "grad_norm": 2.100961923599243,
+      "learning_rate": 0.0001,
+      "loss": 0.0233,
+      "step": 18800
+    },
+    {
+      "epoch": 111.30177514792899,
+      "grad_norm": 1.8579765558242798,
+      "learning_rate": 0.0001,
+      "loss": 0.0246,
+      "step": 18810
+    },
+    {
+      "epoch": 111.36094674556213,
+      "grad_norm": 1.6942815780639648,
+      "learning_rate": 0.0001,
+      "loss": 0.0244,
+      "step": 18820
+    },
+    {
+      "epoch": 111.42011834319527,
+      "grad_norm": 1.5085196495056152,
+      "learning_rate": 0.0001,
+      "loss": 0.0243,
+      "step": 18830
+    },
+    {
+      "epoch": 111.47928994082841,
+      "grad_norm": 1.3338913917541504,
+      "learning_rate": 0.0001,
+      "loss": 0.0243,
+      "step": 18840
+    },
+    {
+      "epoch": 111.53846153846153,
+      "grad_norm": 1.4442957639694214,
+      "learning_rate": 0.0001,
+      "loss": 0.0243,
+      "step": 18850
+    },
+    {
+      "epoch": 111.59763313609467,
+      "grad_norm": 1.6123669147491455,
+      "learning_rate": 0.0001,
+      "loss": 0.0243,
+      "step": 18860
+    },
+    {
+      "epoch": 111.65680473372781,
+      "grad_norm": 1.4258482456207275,
+      "learning_rate": 0.0001,
+      "loss": 0.0234,
+      "step": 18870
+    },
+    {
+      "epoch": 111.71597633136095,
+      "grad_norm": 1.5295238494873047,
+      "learning_rate": 0.0001,
+      "loss": 0.0237,
+      "step": 18880
+    },
+    {
+      "epoch": 111.77514792899409,
+      "grad_norm": 1.6922048330307007,
+      "learning_rate": 0.0001,
+      "loss": 0.0226,
+      "step": 18890
+    },
+    {
+      "epoch": 111.83431952662721,
+      "grad_norm": 1.9325705766677856,
+      "learning_rate": 0.0001,
+      "loss": 0.0239,
+      "step": 18900
+    },
+    {
+      "epoch": 111.89349112426035,
+      "grad_norm": 1.5474536418914795,
+      "learning_rate": 0.0001,
+      "loss": 0.0233,
+      "step": 18910
+    },
+    {
+      "epoch": 111.95266272189349,
+      "grad_norm": 1.712049961090088,
+      "learning_rate": 0.0001,
+      "loss": 0.024,
+      "step": 18920
+    },
+    {
+      "epoch": 112.01183431952663,
+      "grad_norm": 1.5668643712997437,
+      "learning_rate": 0.0001,
+      "loss": 0.0237,
+      "step": 18930
+    },
+    {
+      "epoch": 112.07100591715977,
+      "grad_norm": 1.3157771825790405,
+      "learning_rate": 0.0001,
+      "loss": 0.0234,
+      "step": 18940
+    },
+    {
+      "epoch": 112.1301775147929,
+      "grad_norm": 1.5523183345794678,
+      "learning_rate": 0.0001,
+      "loss": 0.0229,
+      "step": 18950
+    },
+    {
+      "epoch": 112.18934911242603,
+      "grad_norm": 1.2441097497940063,
+      "learning_rate": 0.0001,
+      "loss": 0.0235,
+      "step": 18960
+    },
+    {
+      "epoch": 112.24852071005917,
+      "grad_norm": 1.9158892631530762,
+      "learning_rate": 0.0001,
+      "loss": 0.0233,
+      "step": 18970
+    },
+    {
+      "epoch": 112.3076923076923,
+      "grad_norm": 1.5876926183700562,
+      "learning_rate": 0.0001,
+      "loss": 0.0235,
+      "step": 18980
+    },
+    {
+      "epoch": 112.36686390532545,
+      "grad_norm": 1.2936400175094604,
+      "learning_rate": 0.0001,
+      "loss": 0.0229,
+      "step": 18990
+    },
+    {
+      "epoch": 112.42603550295858,
+      "grad_norm": 1.702101469039917,
+      "learning_rate": 0.0001,
+      "loss": 0.0234,
+      "step": 19000
+    },
+    {
+      "epoch": 112.48520710059172,
+      "grad_norm": 1.8644615411758423,
+      "learning_rate": 0.0001,
+      "loss": 0.0236,
+      "step": 19010
+    },
+    {
+      "epoch": 112.54437869822485,
+      "grad_norm": 1.465781569480896,
+      "learning_rate": 0.0001,
+      "loss": 0.0235,
+      "step": 19020
+    },
+    {
+      "epoch": 112.60355029585799,
+      "grad_norm": 1.5026826858520508,
+      "learning_rate": 0.0001,
+      "loss": 0.0233,
+      "step": 19030
+    },
+    {
+      "epoch": 112.66272189349112,
+      "grad_norm": 1.4677021503448486,
+      "learning_rate": 0.0001,
+      "loss": 0.0245,
+      "step": 19040
+    },
+    {
+      "epoch": 112.72189349112426,
+      "grad_norm": 1.5909640789031982,
+      "learning_rate": 0.0001,
+      "loss": 0.0244,
+      "step": 19050
+    },
+    {
+      "epoch": 112.7810650887574,
+      "grad_norm": 1.5885133743286133,
+      "learning_rate": 0.0001,
+      "loss": 0.0245,
+      "step": 19060
+    },
+    {
+      "epoch": 112.84023668639053,
+      "grad_norm": 1.5809088945388794,
+      "learning_rate": 0.0001,
+      "loss": 0.0254,
+      "step": 19070
+    },
+    {
+      "epoch": 112.89940828402366,
+      "grad_norm": 2.2121214866638184,
+      "learning_rate": 0.0001,
+      "loss": 0.0252,
+      "step": 19080
+    },
+    {
+      "epoch": 112.9585798816568,
+      "grad_norm": 1.4797953367233276,
+      "learning_rate": 0.0001,
+      "loss": 0.0254,
+      "step": 19090
+    },
+    {
+      "epoch": 113.01775147928994,
+      "grad_norm": 1.4344440698623657,
+      "learning_rate": 0.0001,
+      "loss": 0.0239,
+      "step": 19100
+    },
+    {
+      "epoch": 113.07692307692308,
+      "grad_norm": 1.5104477405548096,
+      "learning_rate": 0.0001,
+      "loss": 0.0234,
+      "step": 19110
+    },
+    {
+      "epoch": 113.13609467455622,
+      "grad_norm": 1.344158411026001,
+      "learning_rate": 0.0001,
+      "loss": 0.0239,
+      "step": 19120
+    },
+    {
+      "epoch": 113.19526627218934,
+      "grad_norm": 1.4073817729949951,
+      "learning_rate": 0.0001,
+      "loss": 0.025,
+      "step": 19130
+    },
+    {
+      "epoch": 113.25443786982248,
+      "grad_norm": 1.532403588294983,
+      "learning_rate": 0.0001,
+      "loss": 0.0243,
+      "step": 19140
+    },
+    {
+      "epoch": 113.31360946745562,
+      "grad_norm": 1.5756255388259888,
+      "learning_rate": 0.0001,
+      "loss": 0.0252,
+      "step": 19150
+    },
+    {
+      "epoch": 113.37278106508876,
+      "grad_norm": 1.1864091157913208,
+      "learning_rate": 0.0001,
+      "loss": 0.0233,
+      "step": 19160
+    },
+    {
+      "epoch": 113.4319526627219,
+      "grad_norm": 1.6608409881591797,
+      "learning_rate": 0.0001,
+      "loss": 0.0246,
+      "step": 19170
+    },
+    {
+      "epoch": 113.49112426035504,
+      "grad_norm": 1.7308393716812134,
+      "learning_rate": 0.0001,
+      "loss": 0.0243,
+      "step": 19180
+    },
+    {
+      "epoch": 113.55029585798816,
+      "grad_norm": 2.024353504180908,
+      "learning_rate": 0.0001,
+      "loss": 0.0237,
+      "step": 19190
+    },
+    {
+      "epoch": 113.6094674556213,
+      "grad_norm": 1.7123843431472778,
+      "learning_rate": 0.0001,
+      "loss": 0.0248,
+      "step": 19200
+    },
+    {
+      "epoch": 113.66863905325444,
+      "grad_norm": 1.938563585281372,
+      "learning_rate": 0.0001,
+      "loss": 0.0244,
+      "step": 19210
+    },
+    {
+      "epoch": 113.72781065088758,
+      "grad_norm": 1.477941870689392,
+      "learning_rate": 0.0001,
+      "loss": 0.0245,
+      "step": 19220
+    },
+    {
+      "epoch": 113.78698224852072,
+      "grad_norm": 1.3929296731948853,
+      "learning_rate": 0.0001,
+      "loss": 0.0227,
+      "step": 19230
+    },
+    {
+      "epoch": 113.84615384615384,
+      "grad_norm": 1.9145386219024658,
+      "learning_rate": 0.0001,
+      "loss": 0.0237,
+      "step": 19240
+    },
+    {
+      "epoch": 113.90532544378698,
+      "grad_norm": 1.389234185218811,
+      "learning_rate": 0.0001,
+      "loss": 0.0237,
+      "step": 19250
+    },
+    {
+      "epoch": 113.96449704142012,
+      "grad_norm": 2.6218371391296387,
+      "learning_rate": 0.0001,
+      "loss": 0.0246,
+      "step": 19260
+    },
+    {
+      "epoch": 114.02366863905326,
+      "grad_norm": 1.8044966459274292,
+      "learning_rate": 0.0001,
+      "loss": 0.0251,
+      "step": 19270
+    },
+    {
+      "epoch": 114.0828402366864,
+      "grad_norm": 2.1463310718536377,
+      "learning_rate": 0.0001,
+      "loss": 0.0235,
+      "step": 19280
+    },
+    {
+      "epoch": 114.14201183431953,
+      "grad_norm": 1.6011149883270264,
+      "learning_rate": 0.0001,
+      "loss": 0.0226,
+      "step": 19290
+    },
+    {
+      "epoch": 114.20118343195266,
+      "grad_norm": 1.3946635723114014,
+      "learning_rate": 0.0001,
+      "loss": 0.0237,
+      "step": 19300
+    },
+    {
+      "epoch": 114.2603550295858,
+      "grad_norm": 1.8248989582061768,
+      "learning_rate": 0.0001,
+      "loss": 0.0221,
+      "step": 19310
+    },
+    {
+      "epoch": 114.31952662721893,
+      "grad_norm": 1.1237101554870605,
+      "learning_rate": 0.0001,
+      "loss": 0.023,
+      "step": 19320
+    },
+    {
+      "epoch": 114.37869822485207,
+      "grad_norm": 1.4547003507614136,
+      "learning_rate": 0.0001,
+      "loss": 0.0235,
+      "step": 19330
+    },
+    {
+      "epoch": 114.43786982248521,
+      "grad_norm": 1.5586135387420654,
+      "learning_rate": 0.0001,
+      "loss": 0.0238,
+      "step": 19340
+    },
+    {
+      "epoch": 114.49704142011835,
+      "grad_norm": 1.17821204662323,
+      "learning_rate": 0.0001,
+      "loss": 0.0226,
+      "step": 19350
+    },
+    {
+      "epoch": 114.55621301775147,
+      "grad_norm": 1.206130862236023,
+      "learning_rate": 0.0001,
+      "loss": 0.0239,
+      "step": 19360
+    },
+    {
+      "epoch": 114.61538461538461,
+      "grad_norm": 1.7322198152542114,
+      "learning_rate": 0.0001,
+      "loss": 0.0238,
+      "step": 19370
+    },
+    {
+      "epoch": 114.67455621301775,
+      "grad_norm": 1.1390666961669922,
+      "learning_rate": 0.0001,
+      "loss": 0.0232,
+      "step": 19380
+    },
+    {
+      "epoch": 114.73372781065089,
+      "grad_norm": 0.9414839744567871,
+      "learning_rate": 0.0001,
+      "loss": 0.0237,
+      "step": 19390
+    },
+    {
+      "epoch": 114.79289940828403,
+      "grad_norm": 1.452406883239746,
+      "learning_rate": 0.0001,
+      "loss": 0.0224,
+      "step": 19400
+    },
+    {
+      "epoch": 114.85207100591715,
+      "grad_norm": 1.378127098083496,
+      "learning_rate": 0.0001,
+      "loss": 0.023,
+      "step": 19410
+    },
+    {
+      "epoch": 114.91124260355029,
+      "grad_norm": 1.7707573175430298,
+      "learning_rate": 0.0001,
+      "loss": 0.0244,
+      "step": 19420
+    },
+    {
+      "epoch": 114.97041420118343,
+      "grad_norm": 1.306472659111023,
+      "learning_rate": 0.0001,
+      "loss": 0.0242,
+      "step": 19430
+    },
+    {
+      "epoch": 115.02958579881657,
+      "grad_norm": 1.4937834739685059,
+      "learning_rate": 0.0001,
+      "loss": 0.0243,
+      "step": 19440
+    },
+    {
+      "epoch": 115.08875739644971,
+      "grad_norm": 1.6576955318450928,
+      "learning_rate": 0.0001,
+      "loss": 0.0236,
+      "step": 19450
+    },
+    {
+      "epoch": 115.14792899408285,
+      "grad_norm": 1.2817796468734741,
+      "learning_rate": 0.0001,
+      "loss": 0.0247,
+      "step": 19460
+    },
+    {
+      "epoch": 115.20710059171597,
+      "grad_norm": 1.5545413494110107,
+      "learning_rate": 0.0001,
+      "loss": 0.024,
+      "step": 19470
+    },
+    {
+      "epoch": 115.26627218934911,
+      "grad_norm": 1.7495691776275635,
+      "learning_rate": 0.0001,
+      "loss": 0.0231,
+      "step": 19480
+    },
+    {
+      "epoch": 115.32544378698225,
+      "grad_norm": 1.3021104335784912,
+      "learning_rate": 0.0001,
+      "loss": 0.0247,
+      "step": 19490
+    },
+    {
+      "epoch": 115.38461538461539,
+      "grad_norm": 1.4788888692855835,
+      "learning_rate": 0.0001,
+      "loss": 0.0247,
+      "step": 19500
+    },
+    {
+      "epoch": 115.44378698224853,
+      "grad_norm": 1.2812113761901855,
+      "learning_rate": 0.0001,
+      "loss": 0.0236,
+      "step": 19510
+    },
+    {
+      "epoch": 115.50295857988165,
+      "grad_norm": 1.6180239915847778,
+      "learning_rate": 0.0001,
+      "loss": 0.0249,
+      "step": 19520
+    },
+    {
+      "epoch": 115.56213017751479,
+      "grad_norm": 1.7366526126861572,
+      "learning_rate": 0.0001,
+      "loss": 0.0253,
+      "step": 19530
+    },
+    {
+      "epoch": 115.62130177514793,
+      "grad_norm": 1.547866702079773,
+      "learning_rate": 0.0001,
+      "loss": 0.0244,
+      "step": 19540
+    },
+    {
+      "epoch": 115.68047337278107,
+      "grad_norm": 1.565651535987854,
+      "learning_rate": 0.0001,
+      "loss": 0.0247,
+      "step": 19550
+    },
+    {
+      "epoch": 115.7396449704142,
+      "grad_norm": 1.6102259159088135,
+      "learning_rate": 0.0001,
+      "loss": 0.0245,
+      "step": 19560
+    },
+    {
+      "epoch": 115.79881656804734,
+      "grad_norm": 1.8250877857208252,
+      "learning_rate": 0.0001,
+      "loss": 0.0249,
+      "step": 19570
+    },
+    {
+      "epoch": 115.85798816568047,
+      "grad_norm": 1.663439154624939,
+      "learning_rate": 0.0001,
+      "loss": 0.0231,
+      "step": 19580
+    },
+    {
+      "epoch": 115.9171597633136,
+      "grad_norm": 2.092439651489258,
+      "learning_rate": 0.0001,
+      "loss": 0.0232,
+      "step": 19590
+    },
+    {
+      "epoch": 115.97633136094674,
+      "grad_norm": 1.4379382133483887,
+      "learning_rate": 0.0001,
+      "loss": 0.0224,
+      "step": 19600
+    },
+    {
+      "epoch": 116.03550295857988,
+      "grad_norm": 1.820536732673645,
+      "learning_rate": 0.0001,
+      "loss": 0.0226,
+      "step": 19610
+    },
+    {
+      "epoch": 116.09467455621302,
+      "grad_norm": 1.4892946481704712,
+      "learning_rate": 0.0001,
+      "loss": 0.0222,
+      "step": 19620
+    },
+    {
+      "epoch": 116.15384615384616,
+      "grad_norm": 2.302943229675293,
+      "learning_rate": 0.0001,
+      "loss": 0.0226,
+      "step": 19630
+    },
+    {
+      "epoch": 116.21301775147928,
+      "grad_norm": 1.3892040252685547,
+      "learning_rate": 0.0001,
+      "loss": 0.0223,
+      "step": 19640
+    },
+    {
+      "epoch": 116.27218934911242,
+      "grad_norm": 1.520744800567627,
+      "learning_rate": 0.0001,
+      "loss": 0.022,
+      "step": 19650
+    },
+    {
+      "epoch": 116.33136094674556,
+      "grad_norm": 1.4067323207855225,
+      "learning_rate": 0.0001,
+      "loss": 0.0225,
+      "step": 19660
+    },
+    {
+      "epoch": 116.3905325443787,
+      "grad_norm": 1.3158884048461914,
+      "learning_rate": 0.0001,
+      "loss": 0.0233,
+      "step": 19670
+    },
+    {
+      "epoch": 116.44970414201184,
+      "grad_norm": 1.683178424835205,
+      "learning_rate": 0.0001,
+      "loss": 0.0243,
+      "step": 19680
+    },
+    {
+      "epoch": 116.50887573964496,
+      "grad_norm": 1.6277329921722412,
+      "learning_rate": 0.0001,
+      "loss": 0.0239,
+      "step": 19690
+    },
+    {
+      "epoch": 116.5680473372781,
+      "grad_norm": 1.6767256259918213,
+      "learning_rate": 0.0001,
+      "loss": 0.0258,
+      "step": 19700
+    },
+    {
+      "epoch": 116.62721893491124,
+      "grad_norm": 1.5590250492095947,
+      "learning_rate": 0.0001,
+      "loss": 0.0247,
+      "step": 19710
+    },
+    {
+      "epoch": 116.68639053254438,
+      "grad_norm": 1.2871882915496826,
+      "learning_rate": 0.0001,
+      "loss": 0.0236,
+      "step": 19720
+    },
+    {
+      "epoch": 116.74556213017752,
+      "grad_norm": 1.2820372581481934,
+      "learning_rate": 0.0001,
+      "loss": 0.0238,
+      "step": 19730
+    },
+    {
+      "epoch": 116.80473372781066,
+      "grad_norm": 1.1759978532791138,
+      "learning_rate": 0.0001,
+      "loss": 0.0231,
+      "step": 19740
+    },
+    {
+      "epoch": 116.86390532544378,
+      "grad_norm": 1.9329886436462402,
+      "learning_rate": 0.0001,
+      "loss": 0.0242,
+      "step": 19750
+    },
+    {
+      "epoch": 116.92307692307692,
+      "grad_norm": 2.220757484436035,
+      "learning_rate": 0.0001,
+      "loss": 0.0232,
+      "step": 19760
+    },
+    {
+      "epoch": 116.98224852071006,
+      "grad_norm": 1.6667869091033936,
+      "learning_rate": 0.0001,
+      "loss": 0.0226,
+      "step": 19770
+    },
+    {
+      "epoch": 117.0414201183432,
+      "grad_norm": 1.814763069152832,
+      "learning_rate": 0.0001,
+      "loss": 0.0231,
+      "step": 19780
+    },
+    {
+      "epoch": 117.10059171597634,
+      "grad_norm": 2.053699254989624,
+      "learning_rate": 0.0001,
+      "loss": 0.0229,
+      "step": 19790
+    },
+    {
+      "epoch": 117.15976331360947,
+      "grad_norm": 1.8936330080032349,
+      "learning_rate": 0.0001,
+      "loss": 0.0228,
+      "step": 19800
+    },
+    {
+      "epoch": 117.2189349112426,
+      "grad_norm": 1.390587329864502,
+      "learning_rate": 0.0001,
+      "loss": 0.0222,
+      "step": 19810
+    },
+    {
+      "epoch": 117.27810650887574,
+      "grad_norm": 1.3521565198898315,
+      "learning_rate": 0.0001,
+      "loss": 0.0226,
+      "step": 19820
+    },
+    {
+      "epoch": 117.33727810650888,
+      "grad_norm": 1.3511319160461426,
+      "learning_rate": 0.0001,
+      "loss": 0.0232,
+      "step": 19830
+    },
+    {
+      "epoch": 117.39644970414201,
+      "grad_norm": 1.6990842819213867,
+      "learning_rate": 0.0001,
+      "loss": 0.0227,
+      "step": 19840
+    },
+    {
+      "epoch": 117.45562130177515,
+      "grad_norm": 1.705824613571167,
+      "learning_rate": 0.0001,
+      "loss": 0.0236,
+      "step": 19850
+    },
+    {
+      "epoch": 117.51479289940828,
+      "grad_norm": 1.3330501317977905,
+      "learning_rate": 0.0001,
+      "loss": 0.0225,
+      "step": 19860
+    },
+    {
+      "epoch": 117.57396449704142,
+      "grad_norm": 2.111391544342041,
+      "learning_rate": 0.0001,
+      "loss": 0.023,
+      "step": 19870
+    },
+    {
+      "epoch": 117.63313609467455,
+      "grad_norm": 1.487842321395874,
+      "learning_rate": 0.0001,
+      "loss": 0.0233,
+      "step": 19880
+    },
+    {
+      "epoch": 117.6923076923077,
+      "grad_norm": 1.8104013204574585,
+      "learning_rate": 0.0001,
+      "loss": 0.0233,
+      "step": 19890
+    },
+    {
+      "epoch": 117.75147928994083,
+      "grad_norm": 1.6497268676757812,
+      "learning_rate": 0.0001,
+      "loss": 0.0236,
+      "step": 19900
+    },
+    {
+      "epoch": 117.81065088757397,
+      "grad_norm": 1.799215316772461,
+      "learning_rate": 0.0001,
+      "loss": 0.0237,
+      "step": 19910
+    },
+    {
+      "epoch": 117.8698224852071,
+      "grad_norm": 1.6056774854660034,
+      "learning_rate": 0.0001,
+      "loss": 0.0246,
+      "step": 19920
+    },
+    {
+      "epoch": 117.92899408284023,
+      "grad_norm": 2.169076681137085,
+      "learning_rate": 0.0001,
+      "loss": 0.0246,
+      "step": 19930
+    },
+    {
+      "epoch": 117.98816568047337,
+      "grad_norm": 2.0190317630767822,
+      "learning_rate": 0.0001,
+      "loss": 0.0238,
+      "step": 19940
+    },
+    {
+      "epoch": 118.04733727810651,
+      "grad_norm": 1.7783856391906738,
+      "learning_rate": 0.0001,
+      "loss": 0.0231,
+      "step": 19950
+    },
+    {
+      "epoch": 118.10650887573965,
+      "grad_norm": 1.3581403493881226,
+      "learning_rate": 0.0001,
+      "loss": 0.0233,
+      "step": 19960
+    },
+    {
+      "epoch": 118.16568047337279,
+      "grad_norm": 1.4854947328567505,
+      "learning_rate": 0.0001,
+      "loss": 0.0222,
+      "step": 19970
+    },
+    {
+      "epoch": 118.22485207100591,
+      "grad_norm": 1.3405940532684326,
+      "learning_rate": 0.0001,
+      "loss": 0.0228,
+      "step": 19980
+    },
+    {
+      "epoch": 118.28402366863905,
+      "grad_norm": 1.3686249256134033,
+      "learning_rate": 0.0001,
+      "loss": 0.0226,
+      "step": 19990
+    },
+    {
+      "epoch": 118.34319526627219,
+      "grad_norm": 1.4328416585922241,
+      "learning_rate": 0.0001,
+      "loss": 0.0232,
+      "step": 20000
+    },
+    {
+      "epoch": 118.40236686390533,
+      "grad_norm": 1.314095139503479,
+      "learning_rate": 0.0001,
+      "loss": 0.0225,
+      "step": 20010
+    },
+    {
+      "epoch": 118.46153846153847,
+      "grad_norm": 1.3998072147369385,
+      "learning_rate": 0.0001,
+      "loss": 0.0243,
+      "step": 20020
+    },
+    {
+      "epoch": 118.52071005917159,
+      "grad_norm": 2.2941207885742188,
+      "learning_rate": 0.0001,
+      "loss": 0.0242,
+      "step": 20030
+    },
+    {
+      "epoch": 118.57988165680473,
+      "grad_norm": 1.7800618410110474,
+      "learning_rate": 0.0001,
+      "loss": 0.0244,
+      "step": 20040
+    },
+    {
+      "epoch": 118.63905325443787,
+      "grad_norm": 1.347954511642456,
+      "learning_rate": 0.0001,
+      "loss": 0.0236,
+      "step": 20050
+    },
+    {
+      "epoch": 118.69822485207101,
+      "grad_norm": 1.3598318099975586,
+      "learning_rate": 0.0001,
+      "loss": 0.0235,
+      "step": 20060
+    },
+    {
+      "epoch": 118.75739644970415,
+      "grad_norm": 1.0675314664840698,
+      "learning_rate": 0.0001,
+      "loss": 0.0232,
+      "step": 20070
+    },
+    {
+      "epoch": 118.81656804733728,
+      "grad_norm": 1.6646771430969238,
+      "learning_rate": 0.0001,
+      "loss": 0.0232,
+      "step": 20080
+    },
+    {
+      "epoch": 118.87573964497041,
+      "grad_norm": 1.6578128337860107,
+      "learning_rate": 0.0001,
+      "loss": 0.0235,
+      "step": 20090
+    },
+    {
+      "epoch": 118.93491124260355,
+      "grad_norm": 1.8425971269607544,
+      "learning_rate": 0.0001,
+      "loss": 0.0239,
+      "step": 20100
+    },
+    {
+      "epoch": 118.99408284023669,
+      "grad_norm": 1.454912781715393,
+      "learning_rate": 0.0001,
+      "loss": 0.0237,
+      "step": 20110
+    },
+    {
+      "epoch": 119.05325443786982,
+      "grad_norm": 1.6903294324874878,
+      "learning_rate": 0.0001,
+      "loss": 0.0234,
+      "step": 20120
+    },
+    {
+      "epoch": 119.11242603550296,
+      "grad_norm": 1.2494791746139526,
+      "learning_rate": 0.0001,
+      "loss": 0.0227,
+      "step": 20130
+    },
+    {
+      "epoch": 119.17159763313609,
+      "grad_norm": 1.222266435623169,
+      "learning_rate": 0.0001,
+      "loss": 0.0225,
+      "step": 20140
+    },
+    {
+      "epoch": 119.23076923076923,
+      "grad_norm": 1.3532673120498657,
+      "learning_rate": 0.0001,
+      "loss": 0.0226,
+      "step": 20150
+    },
+    {
+      "epoch": 119.28994082840237,
+      "grad_norm": 1.6829416751861572,
+      "learning_rate": 0.0001,
+      "loss": 0.0233,
+      "step": 20160
+    },
+    {
+      "epoch": 119.3491124260355,
+      "grad_norm": 1.5614653825759888,
+      "learning_rate": 0.0001,
+      "loss": 0.0237,
+      "step": 20170
+    },
+    {
+      "epoch": 119.40828402366864,
+      "grad_norm": 1.512175440788269,
+      "learning_rate": 0.0001,
+      "loss": 0.024,
+      "step": 20180
+    },
+    {
+      "epoch": 119.46745562130178,
+      "grad_norm": 1.5109405517578125,
+      "learning_rate": 0.0001,
+      "loss": 0.0229,
+      "step": 20190
+    },
+    {
+      "epoch": 119.5266272189349,
+      "grad_norm": 1.5895750522613525,
+      "learning_rate": 0.0001,
+      "loss": 0.0227,
+      "step": 20200
+    },
+    {
+      "epoch": 119.58579881656804,
+      "grad_norm": 1.7043721675872803,
+      "learning_rate": 0.0001,
+      "loss": 0.0229,
+      "step": 20210
+    },
+    {
+      "epoch": 119.64497041420118,
+      "grad_norm": 1.3565478324890137,
+      "learning_rate": 0.0001,
+      "loss": 0.0229,
+      "step": 20220
+    },
+    {
+      "epoch": 119.70414201183432,
+      "grad_norm": 1.3484808206558228,
+      "learning_rate": 0.0001,
+      "loss": 0.0232,
+      "step": 20230
+    },
+    {
+      "epoch": 119.76331360946746,
+      "grad_norm": 1.7464561462402344,
+      "learning_rate": 0.0001,
+      "loss": 0.024,
+      "step": 20240
+    },
+    {
+      "epoch": 119.8224852071006,
+      "grad_norm": 1.6434247493743896,
+      "learning_rate": 0.0001,
+      "loss": 0.0238,
+      "step": 20250
+    },
+    {
+      "epoch": 119.88165680473372,
+      "grad_norm": 0.994000256061554,
+      "learning_rate": 0.0001,
+      "loss": 0.0231,
+      "step": 20260
+    },
+    {
+      "epoch": 119.94082840236686,
+      "grad_norm": 1.0915237665176392,
+      "learning_rate": 0.0001,
+      "loss": 0.0226,
+      "step": 20270
+    },
+    {
+      "epoch": 120.0,
+      "grad_norm": 0.9594990611076355,
+      "learning_rate": 0.0001,
+      "loss": 0.0241,
+      "step": 20280
+    },
+    {
+      "epoch": 120.05917159763314,
+      "grad_norm": 1.2761406898498535,
+      "learning_rate": 0.0001,
+      "loss": 0.0232,
+      "step": 20290
+    },
+    {
+      "epoch": 120.11834319526628,
+      "grad_norm": 1.5102019309997559,
+      "learning_rate": 0.0001,
+      "loss": 0.0245,
+      "step": 20300
+    },
+    {
+      "epoch": 120.1775147928994,
+      "grad_norm": 1.9019640684127808,
+      "learning_rate": 0.0001,
+      "loss": 0.0251,
+      "step": 20310
+    },
+    {
+      "epoch": 120.23668639053254,
+      "grad_norm": 1.1323012113571167,
+      "learning_rate": 0.0001,
+      "loss": 0.0247,
+      "step": 20320
+    },
+    {
+      "epoch": 120.29585798816568,
+      "grad_norm": 1.9056973457336426,
+      "learning_rate": 0.0001,
+      "loss": 0.0244,
+      "step": 20330
+    },
+    {
+      "epoch": 120.35502958579882,
+      "grad_norm": 2.193131923675537,
+      "learning_rate": 0.0001,
+      "loss": 0.0234,
+      "step": 20340
+    },
+    {
+      "epoch": 120.41420118343196,
+      "grad_norm": 1.6800955533981323,
+      "learning_rate": 0.0001,
+      "loss": 0.0242,
+      "step": 20350
+    },
+    {
+      "epoch": 120.4733727810651,
+      "grad_norm": 1.83359956741333,
+      "learning_rate": 0.0001,
+      "loss": 0.0234,
+      "step": 20360
+    },
+    {
+      "epoch": 120.53254437869822,
+      "grad_norm": 1.1426379680633545,
+      "learning_rate": 0.0001,
+      "loss": 0.0242,
+      "step": 20370
+    },
+    {
+      "epoch": 120.59171597633136,
+      "grad_norm": 1.532046914100647,
+      "learning_rate": 0.0001,
+      "loss": 0.0249,
+      "step": 20380
+    },
+    {
+      "epoch": 120.6508875739645,
+      "grad_norm": 1.6982733011245728,
+      "learning_rate": 0.0001,
+      "loss": 0.0243,
+      "step": 20390
+    },
+    {
+      "epoch": 120.71005917159763,
+      "grad_norm": 1.6299041509628296,
+      "learning_rate": 0.0001,
+      "loss": 0.0253,
+      "step": 20400
+    },
+    {
+      "epoch": 120.76923076923077,
+      "grad_norm": 1.2626619338989258,
+      "learning_rate": 0.0001,
+      "loss": 0.0246,
+      "step": 20410
+    },
+    {
+      "epoch": 120.82840236686391,
+      "grad_norm": 1.4834842681884766,
+      "learning_rate": 0.0001,
+      "loss": 0.0251,
+      "step": 20420
+    },
+    {
+      "epoch": 120.88757396449704,
+      "grad_norm": 1.638554573059082,
+      "learning_rate": 0.0001,
+      "loss": 0.024,
+      "step": 20430
+    },
+    {
+      "epoch": 120.94674556213018,
+      "grad_norm": 1.366302728652954,
+      "learning_rate": 0.0001,
+      "loss": 0.025,
+      "step": 20440
+    },
+    {
+      "epoch": 121.00591715976331,
+      "grad_norm": 1.6834704875946045,
+      "learning_rate": 0.0001,
+      "loss": 0.0229,
+      "step": 20450
+    },
+    {
+      "epoch": 121.06508875739645,
+      "grad_norm": 1.072150468826294,
+      "learning_rate": 0.0001,
+      "loss": 0.0228,
+      "step": 20460
+    },
+    {
+      "epoch": 121.12426035502959,
+      "grad_norm": 1.7651511430740356,
+      "learning_rate": 0.0001,
+      "loss": 0.0226,
+      "step": 20470
+    },
+    {
+      "epoch": 121.18343195266272,
+      "grad_norm": 1.0291211605072021,
+      "learning_rate": 0.0001,
+      "loss": 0.023,
+      "step": 20480
+    },
+    {
+      "epoch": 121.24260355029585,
+      "grad_norm": 1.672871470451355,
+      "learning_rate": 0.0001,
+      "loss": 0.0236,
+      "step": 20490
+    },
+    {
+      "epoch": 121.30177514792899,
+      "grad_norm": 1.8796555995941162,
+      "learning_rate": 0.0001,
+      "loss": 0.0233,
+      "step": 20500
+    },
+    {
+      "epoch": 121.36094674556213,
+      "grad_norm": 1.6090160608291626,
+      "learning_rate": 0.0001,
+      "loss": 0.0233,
+      "step": 20510
+    },
+    {
+      "epoch": 121.42011834319527,
+      "grad_norm": 1.3602937459945679,
+      "learning_rate": 0.0001,
+      "loss": 0.0235,
+      "step": 20520
+    },
+    {
+      "epoch": 121.47928994082841,
+      "grad_norm": 1.8083723783493042,
+      "learning_rate": 0.0001,
+      "loss": 0.0231,
+      "step": 20530
+    },
+    {
+      "epoch": 121.53846153846153,
+      "grad_norm": 1.489169955253601,
+      "learning_rate": 0.0001,
+      "loss": 0.0234,
+      "step": 20540
+    },
+    {
+      "epoch": 121.59763313609467,
+      "grad_norm": 1.798827886581421,
+      "learning_rate": 0.0001,
+      "loss": 0.023,
+      "step": 20550
+    },
+    {
+      "epoch": 121.65680473372781,
+      "grad_norm": 1.256948471069336,
+      "learning_rate": 0.0001,
+      "loss": 0.023,
+      "step": 20560
+    },
+    {
+      "epoch": 121.71597633136095,
+      "grad_norm": 1.37706458568573,
+      "learning_rate": 0.0001,
+      "loss": 0.0228,
+      "step": 20570
+    },
+    {
+      "epoch": 121.77514792899409,
+      "grad_norm": 1.3038328886032104,
+      "learning_rate": 0.0001,
+      "loss": 0.023,
+      "step": 20580
+    },
+    {
+      "epoch": 121.83431952662721,
+      "grad_norm": 1.4907702207565308,
+      "learning_rate": 0.0001,
+      "loss": 0.024,
+      "step": 20590
+    },
+    {
+      "epoch": 121.89349112426035,
+      "grad_norm": 1.7627522945404053,
+      "learning_rate": 0.0001,
+      "loss": 0.0249,
+      "step": 20600
+    },
+    {
+      "epoch": 121.95266272189349,
+      "grad_norm": 1.2299731969833374,
+      "learning_rate": 0.0001,
+      "loss": 0.0246,
+      "step": 20610
+    },
+    {
+      "epoch": 122.01183431952663,
+      "grad_norm": 1.2428611516952515,
+      "learning_rate": 0.0001,
+      "loss": 0.0238,
+      "step": 20620
+    },
+    {
+      "epoch": 122.07100591715977,
+      "grad_norm": 1.9673820734024048,
+      "learning_rate": 0.0001,
+      "loss": 0.024,
+      "step": 20630
+    },
+    {
+      "epoch": 122.1301775147929,
+      "grad_norm": 1.8363322019577026,
+      "learning_rate": 0.0001,
+      "loss": 0.0242,
+      "step": 20640
+    },
+    {
+      "epoch": 122.18934911242603,
+      "grad_norm": 1.1473056077957153,
+      "learning_rate": 0.0001,
+      "loss": 0.0227,
+      "step": 20650
+    },
+    {
+      "epoch": 122.24852071005917,
+      "grad_norm": 1.2694402933120728,
+      "learning_rate": 0.0001,
+      "loss": 0.0223,
+      "step": 20660
+    },
+    {
+      "epoch": 122.3076923076923,
+      "grad_norm": 1.137391448020935,
+      "learning_rate": 0.0001,
+      "loss": 0.0227,
+      "step": 20670
+    },
+    {
+      "epoch": 122.36686390532545,
+      "grad_norm": 1.5363447666168213,
+      "learning_rate": 0.0001,
+      "loss": 0.0228,
+      "step": 20680
+    },
+    {
+      "epoch": 122.42603550295858,
+      "grad_norm": 1.447498083114624,
+      "learning_rate": 0.0001,
+      "loss": 0.0234,
+      "step": 20690
+    },
+    {
+      "epoch": 122.48520710059172,
+      "grad_norm": 1.4120616912841797,
+      "learning_rate": 0.0001,
+      "loss": 0.0235,
+      "step": 20700
+    },
+    {
+      "epoch": 122.54437869822485,
+      "grad_norm": 1.7621294260025024,
+      "learning_rate": 0.0001,
+      "loss": 0.0238,
+      "step": 20710
+    },
+    {
+      "epoch": 122.60355029585799,
+      "grad_norm": 1.0842241048812866,
+      "learning_rate": 0.0001,
+      "loss": 0.023,
+      "step": 20720
+    },
+    {
+      "epoch": 122.66272189349112,
+      "grad_norm": 1.9327573776245117,
+      "learning_rate": 0.0001,
+      "loss": 0.0238,
+      "step": 20730
+    },
+    {
+      "epoch": 122.72189349112426,
+      "grad_norm": 2.1728267669677734,
+      "learning_rate": 0.0001,
+      "loss": 0.0243,
+      "step": 20740
+    },
+    {
+      "epoch": 122.7810650887574,
+      "grad_norm": 1.165481686592102,
+      "learning_rate": 0.0001,
+      "loss": 0.0232,
+      "step": 20750
+    },
+    {
+      "epoch": 122.84023668639053,
+      "grad_norm": 1.48213529586792,
+      "learning_rate": 0.0001,
+      "loss": 0.0234,
+      "step": 20760
+    },
+    {
+      "epoch": 122.89940828402366,
+      "grad_norm": 1.5164821147918701,
+      "learning_rate": 0.0001,
+      "loss": 0.0231,
+      "step": 20770
+    },
+    {
+      "epoch": 122.9585798816568,
+      "grad_norm": 1.3426398038864136,
+      "learning_rate": 0.0001,
+      "loss": 0.0229,
+      "step": 20780
+    },
+    {
+      "epoch": 123.01775147928994,
+      "grad_norm": 1.1179726123809814,
+      "learning_rate": 0.0001,
+      "loss": 0.0239,
+      "step": 20790
+    },
+    {
+      "epoch": 123.07692307692308,
+      "grad_norm": 0.9754278063774109,
+      "learning_rate": 0.0001,
+      "loss": 0.0229,
+      "step": 20800
+    },
+    {
+      "epoch": 123.13609467455622,
+      "grad_norm": 1.6338154077529907,
+      "learning_rate": 0.0001,
+      "loss": 0.0227,
+      "step": 20810
+    },
+    {
+      "epoch": 123.19526627218934,
+      "grad_norm": 1.5298885107040405,
+      "learning_rate": 0.0001,
+      "loss": 0.023,
+      "step": 20820
+    },
+    {
+      "epoch": 123.25443786982248,
+      "grad_norm": 1.642648696899414,
+      "learning_rate": 0.0001,
+      "loss": 0.0223,
+      "step": 20830
+    },
+    {
+      "epoch": 123.31360946745562,
+      "grad_norm": 1.746422290802002,
+      "learning_rate": 0.0001,
+      "loss": 0.0235,
+      "step": 20840
+    },
+    {
+      "epoch": 123.37278106508876,
+      "grad_norm": 1.5254266262054443,
+      "learning_rate": 0.0001,
+      "loss": 0.0226,
+      "step": 20850
+    },
+    {
+      "epoch": 123.4319526627219,
+      "grad_norm": 1.9833745956420898,
+      "learning_rate": 0.0001,
+      "loss": 0.023,
+      "step": 20860
+    },
+    {
+      "epoch": 123.49112426035504,
+      "grad_norm": 1.63018000125885,
+      "learning_rate": 0.0001,
+      "loss": 0.024,
+      "step": 20870
+    },
+    {
+      "epoch": 123.55029585798816,
+      "grad_norm": 1.9519351720809937,
+      "learning_rate": 0.0001,
+      "loss": 0.0238,
+      "step": 20880
+    },
+    {
+      "epoch": 123.6094674556213,
+      "grad_norm": 1.1955246925354004,
+      "learning_rate": 0.0001,
+      "loss": 0.0228,
+      "step": 20890
+    },
+    {
+      "epoch": 123.66863905325444,
+      "grad_norm": 2.4289801120758057,
+      "learning_rate": 0.0001,
+      "loss": 0.0238,
+      "step": 20900
+    },
+    {
+      "epoch": 123.72781065088758,
+      "grad_norm": 1.9642705917358398,
+      "learning_rate": 0.0001,
+      "loss": 0.0233,
+      "step": 20910
+    },
+    {
+      "epoch": 123.78698224852072,
+      "grad_norm": 1.5356130599975586,
+      "learning_rate": 0.0001,
+      "loss": 0.0222,
+      "step": 20920
+    },
+    {
+      "epoch": 123.84615384615384,
+      "grad_norm": 1.0688326358795166,
+      "learning_rate": 0.0001,
+      "loss": 0.0214,
+      "step": 20930
+    },
+    {
+      "epoch": 123.90532544378698,
+      "grad_norm": 1.2926667928695679,
+      "learning_rate": 0.0001,
+      "loss": 0.0222,
+      "step": 20940
+    },
+    {
+      "epoch": 123.96449704142012,
+      "grad_norm": 1.3277307748794556,
+      "learning_rate": 0.0001,
+      "loss": 0.0227,
+      "step": 20950
+    },
+    {
+      "epoch": 124.02366863905326,
+      "grad_norm": 1.9115105867385864,
+      "learning_rate": 0.0001,
+      "loss": 0.0224,
+      "step": 20960
+    },
+    {
+      "epoch": 124.0828402366864,
+      "grad_norm": 1.7300077676773071,
+      "learning_rate": 0.0001,
+      "loss": 0.0221,
+      "step": 20970
+    },
+    {
+      "epoch": 124.14201183431953,
+      "grad_norm": 1.7841664552688599,
+      "learning_rate": 0.0001,
+      "loss": 0.022,
+      "step": 20980
+    },
+    {
+      "epoch": 124.20118343195266,
+      "grad_norm": 1.4203828573226929,
+      "learning_rate": 0.0001,
+      "loss": 0.0227,
+      "step": 20990
+    },
+    {
+      "epoch": 124.2603550295858,
+      "grad_norm": 1.2407503128051758,
+      "learning_rate": 0.0001,
+      "loss": 0.0229,
+      "step": 21000
+    },
+    {
+      "epoch": 124.31952662721893,
+      "grad_norm": 1.4121196269989014,
+      "learning_rate": 0.0001,
+      "loss": 0.0236,
+      "step": 21010
+    },
+    {
+      "epoch": 124.37869822485207,
+      "grad_norm": 1.346779227256775,
+      "learning_rate": 0.0001,
+      "loss": 0.0234,
+      "step": 21020
+    },
+    {
+      "epoch": 124.43786982248521,
+      "grad_norm": 2.013970375061035,
+      "learning_rate": 0.0001,
+      "loss": 0.0239,
+      "step": 21030
+    },
+    {
+      "epoch": 124.49704142011835,
+      "grad_norm": 1.6955645084381104,
+      "learning_rate": 0.0001,
+      "loss": 0.0231,
+      "step": 21040
+    },
+    {
+      "epoch": 124.55621301775147,
+      "grad_norm": 1.5064244270324707,
+      "learning_rate": 0.0001,
+      "loss": 0.0224,
+      "step": 21050
+    },
+    {
+      "epoch": 124.61538461538461,
+      "grad_norm": 1.6921324729919434,
+      "learning_rate": 0.0001,
+      "loss": 0.0226,
+      "step": 21060
+    },
+    {
+      "epoch": 124.67455621301775,
+      "grad_norm": 1.4081296920776367,
+      "learning_rate": 0.0001,
+      "loss": 0.0223,
+      "step": 21070
+    },
+    {
+      "epoch": 124.73372781065089,
+      "grad_norm": 1.5960279703140259,
+      "learning_rate": 0.0001,
+      "loss": 0.0235,
+      "step": 21080
+    },
+    {
+      "epoch": 124.79289940828403,
+      "grad_norm": 2.1927590370178223,
+      "learning_rate": 0.0001,
+      "loss": 0.0231,
+      "step": 21090
+    },
+    {
+      "epoch": 124.85207100591715,
+      "grad_norm": 1.8707386255264282,
+      "learning_rate": 0.0001,
+      "loss": 0.0238,
+      "step": 21100
+    },
+    {
+      "epoch": 124.91124260355029,
+      "grad_norm": 1.9166415929794312,
+      "learning_rate": 0.0001,
+      "loss": 0.0233,
+      "step": 21110
+    },
+    {
+      "epoch": 124.97041420118343,
+      "grad_norm": 2.0922372341156006,
+      "learning_rate": 0.0001,
+      "loss": 0.0221,
+      "step": 21120
+    },
+    {
+      "epoch": 125.02958579881657,
+      "grad_norm": 1.6090058088302612,
+      "learning_rate": 0.0001,
+      "loss": 0.0233,
+      "step": 21130
+    },
+    {
+      "epoch": 125.08875739644971,
+      "grad_norm": 1.3405508995056152,
+      "learning_rate": 0.0001,
+      "loss": 0.0224,
+      "step": 21140
+    },
+    {
+      "epoch": 125.14792899408285,
+      "grad_norm": 1.5459500551223755,
+      "learning_rate": 0.0001,
+      "loss": 0.024,
+      "step": 21150
+    },
+    {
+      "epoch": 125.20710059171597,
+      "grad_norm": 1.0749977827072144,
+      "learning_rate": 0.0001,
+      "loss": 0.0224,
+      "step": 21160
+    },
+    {
+      "epoch": 125.26627218934911,
+      "grad_norm": 1.0855412483215332,
+      "learning_rate": 0.0001,
+      "loss": 0.023,
+      "step": 21170
+    },
+    {
+      "epoch": 125.32544378698225,
+      "grad_norm": 1.435312032699585,
+      "learning_rate": 0.0001,
+      "loss": 0.0242,
+      "step": 21180
+    },
+    {
+      "epoch": 125.38461538461539,
+      "grad_norm": 1.1181412935256958,
+      "learning_rate": 0.0001,
+      "loss": 0.0242,
+      "step": 21190
+    },
+    {
+      "epoch": 125.44378698224853,
+      "grad_norm": 1.7906783819198608,
+      "learning_rate": 0.0001,
+      "loss": 0.025,
+      "step": 21200
+    },
+    {
+      "epoch": 125.50295857988165,
+      "grad_norm": 1.5742876529693604,
+      "learning_rate": 0.0001,
+      "loss": 0.0243,
+      "step": 21210
+    },
+    {
+      "epoch": 125.56213017751479,
+      "grad_norm": 1.35334312915802,
+      "learning_rate": 0.0001,
+      "loss": 0.0231,
+      "step": 21220
+    },
+    {
+      "epoch": 125.62130177514793,
+      "grad_norm": 1.3369613885879517,
+      "learning_rate": 0.0001,
+      "loss": 0.0238,
+      "step": 21230
+    },
+    {
+      "epoch": 125.68047337278107,
+      "grad_norm": 1.3288614749908447,
+      "learning_rate": 0.0001,
+      "loss": 0.0231,
+      "step": 21240
+    },
+    {
+      "epoch": 125.7396449704142,
+      "grad_norm": 2.3308634757995605,
+      "learning_rate": 0.0001,
+      "loss": 0.0242,
+      "step": 21250
+    },
+    {
+      "epoch": 125.79881656804734,
+      "grad_norm": 1.5735810995101929,
+      "learning_rate": 0.0001,
+      "loss": 0.024,
+      "step": 21260
+    },
+    {
+      "epoch": 125.85798816568047,
+      "grad_norm": 2.2582550048828125,
+      "learning_rate": 0.0001,
+      "loss": 0.0229,
+      "step": 21270
+    },
+    {
+      "epoch": 125.9171597633136,
+      "grad_norm": 1.4708681106567383,
+      "learning_rate": 0.0001,
+      "loss": 0.0225,
+      "step": 21280
+    },
+    {
+      "epoch": 125.97633136094674,
+      "grad_norm": 1.294925332069397,
+      "learning_rate": 0.0001,
+      "loss": 0.0221,
+      "step": 21290
+    },
+    {
+      "epoch": 126.03550295857988,
+      "grad_norm": 1.8016443252563477,
+      "learning_rate": 0.0001,
+      "loss": 0.0215,
+      "step": 21300
+    },
+    {
+      "epoch": 126.09467455621302,
+      "grad_norm": 1.0653190612792969,
+      "learning_rate": 0.0001,
+      "loss": 0.022,
+      "step": 21310
+    },
+    {
+      "epoch": 126.15384615384616,
+      "grad_norm": 1.3720836639404297,
+      "learning_rate": 0.0001,
+      "loss": 0.0222,
+      "step": 21320
+    },
+    {
+      "epoch": 126.21301775147928,
+      "grad_norm": 1.4729589223861694,
+      "learning_rate": 0.0001,
+      "loss": 0.0227,
+      "step": 21330
+    },
+    {
+      "epoch": 126.27218934911242,
+      "grad_norm": 1.5231032371520996,
+      "learning_rate": 0.0001,
+      "loss": 0.0231,
+      "step": 21340
+    },
+    {
+      "epoch": 126.33136094674556,
+      "grad_norm": 1.4617881774902344,
+      "learning_rate": 0.0001,
+      "loss": 0.0224,
+      "step": 21350
+    },
+    {
+      "epoch": 126.3905325443787,
+      "grad_norm": 1.6344490051269531,
+      "learning_rate": 0.0001,
+      "loss": 0.0218,
+      "step": 21360
+    },
+    {
+      "epoch": 126.44970414201184,
+      "grad_norm": 1.6493488550186157,
+      "learning_rate": 0.0001,
+      "loss": 0.023,
+      "step": 21370
+    },
+    {
+      "epoch": 126.50887573964496,
+      "grad_norm": 1.3114792108535767,
+      "learning_rate": 0.0001,
+      "loss": 0.022,
+      "step": 21380
+    },
+    {
+      "epoch": 126.5680473372781,
+      "grad_norm": 1.7020283937454224,
+      "learning_rate": 0.0001,
+      "loss": 0.0225,
+      "step": 21390
+    },
+    {
+      "epoch": 126.62721893491124,
+      "grad_norm": 1.9729636907577515,
+      "learning_rate": 0.0001,
+      "loss": 0.023,
+      "step": 21400
+    },
+    {
+      "epoch": 126.68639053254438,
+      "grad_norm": 1.2255303859710693,
+      "learning_rate": 0.0001,
+      "loss": 0.0222,
+      "step": 21410
+    },
+    {
+      "epoch": 126.74556213017752,
+      "grad_norm": 1.3636138439178467,
+      "learning_rate": 0.0001,
+      "loss": 0.0222,
+      "step": 21420
+    },
+    {
+      "epoch": 126.80473372781066,
+      "grad_norm": 1.6777435541152954,
+      "learning_rate": 0.0001,
+      "loss": 0.0219,
+      "step": 21430
+    },
+    {
+      "epoch": 126.86390532544378,
+      "grad_norm": 1.359948992729187,
+      "learning_rate": 0.0001,
+      "loss": 0.0219,
+      "step": 21440
+    },
+    {
+      "epoch": 126.92307692307692,
+      "grad_norm": 1.4733692407608032,
+      "learning_rate": 0.0001,
+      "loss": 0.022,
+      "step": 21450
+    },
+    {
+      "epoch": 126.98224852071006,
+      "grad_norm": 1.5066161155700684,
+      "learning_rate": 0.0001,
+      "loss": 0.0233,
+      "step": 21460
+    },
+    {
+      "epoch": 127.0414201183432,
+      "grad_norm": 1.973374843597412,
+      "learning_rate": 0.0001,
+      "loss": 0.0224,
+      "step": 21470
+    },
+    {
+      "epoch": 127.10059171597634,
+      "grad_norm": 1.7602729797363281,
+      "learning_rate": 0.0001,
+      "loss": 0.0224,
+      "step": 21480
+    },
+    {
+      "epoch": 127.15976331360947,
+      "grad_norm": 1.9331430196762085,
+      "learning_rate": 0.0001,
+      "loss": 0.0232,
+      "step": 21490
+    },
+    {
+      "epoch": 127.2189349112426,
+      "grad_norm": 1.842774748802185,
+      "learning_rate": 0.0001,
+      "loss": 0.0231,
+      "step": 21500
+    },
+    {
+      "epoch": 127.27810650887574,
+      "grad_norm": 1.651479959487915,
+      "learning_rate": 0.0001,
+      "loss": 0.0222,
+      "step": 21510
+    },
+    {
+      "epoch": 127.33727810650888,
+      "grad_norm": 1.938618779182434,
+      "learning_rate": 0.0001,
+      "loss": 0.0231,
+      "step": 21520
+    },
+    {
+      "epoch": 127.39644970414201,
+      "grad_norm": 1.906001091003418,
+      "learning_rate": 0.0001,
+      "loss": 0.0231,
+      "step": 21530
+    },
+    {
+      "epoch": 127.45562130177515,
+      "grad_norm": 1.4963901042938232,
+      "learning_rate": 0.0001,
+      "loss": 0.0226,
+      "step": 21540
+    },
+    {
+      "epoch": 127.51479289940828,
+      "grad_norm": 1.8701282739639282,
+      "learning_rate": 0.0001,
+      "loss": 0.0214,
+      "step": 21550
+    },
+    {
+      "epoch": 127.57396449704142,
+      "grad_norm": 1.5418201684951782,
+      "learning_rate": 0.0001,
+      "loss": 0.0212,
+      "step": 21560
+    },
+    {
+      "epoch": 127.63313609467455,
+      "grad_norm": 1.050010085105896,
+      "learning_rate": 0.0001,
+      "loss": 0.0219,
+      "step": 21570
+    },
+    {
+      "epoch": 127.6923076923077,
+      "grad_norm": 1.5892976522445679,
+      "learning_rate": 0.0001,
+      "loss": 0.0215,
+      "step": 21580
+    },
+    {
+      "epoch": 127.75147928994083,
+      "grad_norm": 1.359131932258606,
+      "learning_rate": 0.0001,
+      "loss": 0.0217,
+      "step": 21590
+    },
+    {
+      "epoch": 127.81065088757397,
+      "grad_norm": 1.1952059268951416,
+      "learning_rate": 0.0001,
+      "loss": 0.0219,
+      "step": 21600
+    },
+    {
+      "epoch": 127.8698224852071,
+      "grad_norm": 0.9953412413597107,
+      "learning_rate": 0.0001,
+      "loss": 0.0218,
+      "step": 21610
+    },
+    {
+      "epoch": 127.92899408284023,
+      "grad_norm": 1.0903805494308472,
+      "learning_rate": 0.0001,
+      "loss": 0.023,
+      "step": 21620
+    },
+    {
+      "epoch": 127.98816568047337,
+      "grad_norm": 1.532383680343628,
+      "learning_rate": 0.0001,
+      "loss": 0.022,
+      "step": 21630
+    },
+    {
+      "epoch": 128.0473372781065,
+      "grad_norm": 1.1569442749023438,
+      "learning_rate": 0.0001,
+      "loss": 0.0226,
+      "step": 21640
+    },
+    {
+      "epoch": 128.10650887573965,
+      "grad_norm": 1.4853934049606323,
+      "learning_rate": 0.0001,
+      "loss": 0.0226,
+      "step": 21650
+    },
+    {
+      "epoch": 128.1656804733728,
+      "grad_norm": 0.9103670716285706,
+      "learning_rate": 0.0001,
+      "loss": 0.0227,
+      "step": 21660
+    },
+    {
+      "epoch": 128.22485207100593,
+      "grad_norm": 1.7763510942459106,
+      "learning_rate": 0.0001,
+      "loss": 0.0232,
+      "step": 21670
+    },
+    {
+      "epoch": 128.28402366863907,
+      "grad_norm": 1.306823968887329,
+      "learning_rate": 0.0001,
+      "loss": 0.0231,
+      "step": 21680
+    },
+    {
+      "epoch": 128.34319526627218,
+      "grad_norm": 1.4585990905761719,
+      "learning_rate": 0.0001,
+      "loss": 0.0222,
+      "step": 21690
+    },
+    {
+      "epoch": 128.4023668639053,
+      "grad_norm": 1.4222877025604248,
+      "learning_rate": 0.0001,
+      "loss": 0.0221,
+      "step": 21700
+    },
+    {
+      "epoch": 128.46153846153845,
+      "grad_norm": 1.5088201761245728,
+      "learning_rate": 0.0001,
+      "loss": 0.0226,
+      "step": 21710
+    },
+    {
+      "epoch": 128.5207100591716,
+      "grad_norm": 1.127785563468933,
+      "learning_rate": 0.0001,
+      "loss": 0.0218,
+      "step": 21720
+    },
+    {
+      "epoch": 128.57988165680473,
+      "grad_norm": 1.4617713689804077,
+      "learning_rate": 0.0001,
+      "loss": 0.0227,
+      "step": 21730
+    },
+    {
+      "epoch": 128.63905325443787,
+      "grad_norm": 1.5193133354187012,
+      "learning_rate": 0.0001,
+      "loss": 0.0231,
+      "step": 21740
+    },
+    {
+      "epoch": 128.698224852071,
+      "grad_norm": 1.211240291595459,
+      "learning_rate": 0.0001,
+      "loss": 0.023,
+      "step": 21750
+    },
+    {
+      "epoch": 128.75739644970415,
+      "grad_norm": 1.210755705833435,
+      "learning_rate": 0.0001,
+      "loss": 0.0237,
+      "step": 21760
+    },
+    {
+      "epoch": 128.81656804733728,
+      "grad_norm": 1.4008526802062988,
+      "learning_rate": 0.0001,
+      "loss": 0.0224,
+      "step": 21770
+    },
+    {
+      "epoch": 128.87573964497042,
+      "grad_norm": 1.793505311012268,
+      "learning_rate": 0.0001,
+      "loss": 0.0238,
+      "step": 21780
+    },
+    {
+      "epoch": 128.93491124260356,
+      "grad_norm": 1.8642908334732056,
+      "learning_rate": 0.0001,
+      "loss": 0.0237,
+      "step": 21790
+    },
+    {
+      "epoch": 128.9940828402367,
+      "grad_norm": 1.6427247524261475,
+      "learning_rate": 0.0001,
+      "loss": 0.0234,
+      "step": 21800
+    },
+    {
+      "epoch": 129.0532544378698,
+      "grad_norm": 1.2126848697662354,
+      "learning_rate": 0.0001,
+      "loss": 0.0242,
+      "step": 21810
+    },
+    {
+      "epoch": 129.11242603550295,
+      "grad_norm": 1.613160490989685,
+      "learning_rate": 0.0001,
+      "loss": 0.0236,
+      "step": 21820
+    },
+    {
+      "epoch": 129.1715976331361,
+      "grad_norm": 2.080007791519165,
+      "learning_rate": 0.0001,
+      "loss": 0.0234,
+      "step": 21830
+    },
+    {
+      "epoch": 129.23076923076923,
+      "grad_norm": 1.3556817770004272,
+      "learning_rate": 0.0001,
+      "loss": 0.0222,
+      "step": 21840
+    },
+    {
+      "epoch": 129.28994082840237,
+      "grad_norm": 1.4367899894714355,
+      "learning_rate": 0.0001,
+      "loss": 0.0219,
+      "step": 21850
+    },
+    {
+      "epoch": 129.3491124260355,
+      "grad_norm": 1.0004467964172363,
+      "learning_rate": 0.0001,
+      "loss": 0.0216,
+      "step": 21860
+    },
+    {
+      "epoch": 129.40828402366864,
+      "grad_norm": 1.3104599714279175,
+      "learning_rate": 0.0001,
+      "loss": 0.0214,
+      "step": 21870
+    },
+    {
+      "epoch": 129.46745562130178,
+      "grad_norm": 1.2123992443084717,
+      "learning_rate": 0.0001,
+      "loss": 0.0218,
+      "step": 21880
+    },
+    {
+      "epoch": 129.52662721893492,
+      "grad_norm": 1.4873669147491455,
+      "learning_rate": 0.0001,
+      "loss": 0.0211,
+      "step": 21890
+    },
+    {
+      "epoch": 129.58579881656806,
+      "grad_norm": 1.3683115243911743,
+      "learning_rate": 0.0001,
+      "loss": 0.0223,
+      "step": 21900
+    },
+    {
+      "epoch": 129.6449704142012,
+      "grad_norm": 1.1770163774490356,
+      "learning_rate": 0.0001,
+      "loss": 0.0222,
+      "step": 21910
+    },
+    {
+      "epoch": 129.7041420118343,
+      "grad_norm": 1.2093645334243774,
+      "learning_rate": 0.0001,
+      "loss": 0.0225,
+      "step": 21920
+    },
+    {
+      "epoch": 129.76331360946745,
+      "grad_norm": 0.9644898176193237,
+      "learning_rate": 0.0001,
+      "loss": 0.0226,
+      "step": 21930
+    },
+    {
+      "epoch": 129.82248520710058,
+      "grad_norm": 1.8431044816970825,
+      "learning_rate": 0.0001,
+      "loss": 0.0226,
+      "step": 21940
+    },
+    {
+      "epoch": 129.88165680473372,
+      "grad_norm": 1.694710612297058,
+      "learning_rate": 0.0001,
+      "loss": 0.0227,
+      "step": 21950
+    },
+    {
+      "epoch": 129.94082840236686,
+      "grad_norm": 1.308655023574829,
+      "learning_rate": 0.0001,
+      "loss": 0.0237,
+      "step": 21960
+    },
+    {
+      "epoch": 130.0,
+      "grad_norm": 1.6879324913024902,
+      "learning_rate": 0.0001,
+      "loss": 0.0226,
+      "step": 21970
+    },
+    {
+      "epoch": 130.05917159763314,
+      "grad_norm": 1.483576774597168,
+      "learning_rate": 0.0001,
+      "loss": 0.0238,
+      "step": 21980
+    },
+    {
+      "epoch": 130.11834319526628,
+      "grad_norm": 1.2481461763381958,
+      "learning_rate": 0.0001,
+      "loss": 0.0228,
+      "step": 21990
+    },
+    {
+      "epoch": 130.17751479289942,
+      "grad_norm": 1.274997591972351,
+      "learning_rate": 0.0001,
+      "loss": 0.0229,
+      "step": 22000
+    },
+    {
+      "epoch": 130.23668639053255,
+      "grad_norm": 0.8794621825218201,
+      "learning_rate": 0.0001,
+      "loss": 0.0226,
+      "step": 22010
+    },
+    {
+      "epoch": 130.2958579881657,
+      "grad_norm": 1.488114833831787,
+      "learning_rate": 0.0001,
+      "loss": 0.0231,
+      "step": 22020
+    },
+    {
+      "epoch": 130.3550295857988,
+      "grad_norm": 1.3086566925048828,
+      "learning_rate": 0.0001,
+      "loss": 0.0228,
+      "step": 22030
+    },
+    {
+      "epoch": 130.41420118343194,
+      "grad_norm": 1.4385002851486206,
+      "learning_rate": 0.0001,
+      "loss": 0.0226,
+      "step": 22040
+    },
+    {
+      "epoch": 130.47337278106508,
+      "grad_norm": 1.3582595586776733,
+      "learning_rate": 0.0001,
+      "loss": 0.0235,
+      "step": 22050
+    },
+    {
+      "epoch": 130.53254437869822,
+      "grad_norm": 1.4870405197143555,
+      "learning_rate": 0.0001,
+      "loss": 0.023,
+      "step": 22060
+    },
+    {
+      "epoch": 130.59171597633136,
+      "grad_norm": 1.2826147079467773,
+      "learning_rate": 0.0001,
+      "loss": 0.0226,
+      "step": 22070
+    },
+    {
+      "epoch": 130.6508875739645,
+      "grad_norm": 1.5948021411895752,
+      "learning_rate": 0.0001,
+      "loss": 0.0231,
+      "step": 22080
+    },
+    {
+      "epoch": 130.71005917159763,
+      "grad_norm": 1.0095163583755493,
+      "learning_rate": 0.0001,
+      "loss": 0.0232,
+      "step": 22090
+    },
+    {
+      "epoch": 130.76923076923077,
+      "grad_norm": 1.589220643043518,
+      "learning_rate": 0.0001,
+      "loss": 0.0228,
+      "step": 22100
+    },
+    {
+      "epoch": 130.8284023668639,
+      "grad_norm": 2.3085834980010986,
+      "learning_rate": 0.0001,
+      "loss": 0.0234,
+      "step": 22110
+    },
+    {
+      "epoch": 130.88757396449705,
+      "grad_norm": 1.5226207971572876,
+      "learning_rate": 0.0001,
+      "loss": 0.0228,
+      "step": 22120
+    },
+    {
+      "epoch": 130.9467455621302,
+      "grad_norm": 1.734958529472351,
+      "learning_rate": 0.0001,
+      "loss": 0.0222,
+      "step": 22130
+    },
+    {
+      "epoch": 131.0059171597633,
+      "grad_norm": 1.4683297872543335,
+      "learning_rate": 0.0001,
+      "loss": 0.0221,
+      "step": 22140
+    },
+    {
+      "epoch": 131.06508875739644,
+      "grad_norm": 1.2361044883728027,
+      "learning_rate": 0.0001,
+      "loss": 0.0219,
+      "step": 22150
+    },
+    {
+      "epoch": 131.12426035502958,
+      "grad_norm": 1.1856882572174072,
+      "learning_rate": 0.0001,
+      "loss": 0.0222,
+      "step": 22160
+    },
+    {
+      "epoch": 131.18343195266272,
+      "grad_norm": 1.1886342763900757,
+      "learning_rate": 0.0001,
+      "loss": 0.0215,
+      "step": 22170
+    },
+    {
+      "epoch": 131.24260355029585,
+      "grad_norm": 0.9051979184150696,
+      "learning_rate": 0.0001,
+      "loss": 0.0223,
+      "step": 22180
+    },
+    {
+      "epoch": 131.301775147929,
+      "grad_norm": 1.4024182558059692,
+      "learning_rate": 0.0001,
+      "loss": 0.0219,
+      "step": 22190
+    },
+    {
+      "epoch": 131.36094674556213,
+      "grad_norm": 1.469608187675476,
+      "learning_rate": 0.0001,
+      "loss": 0.0224,
+      "step": 22200
+    },
+    {
+      "epoch": 131.42011834319527,
+      "grad_norm": 1.8806906938552856,
+      "learning_rate": 0.0001,
+      "loss": 0.0229,
+      "step": 22210
+    },
+    {
+      "epoch": 131.4792899408284,
+      "grad_norm": 1.6373777389526367,
+      "learning_rate": 0.0001,
+      "loss": 0.0219,
+      "step": 22220
+    },
+    {
+      "epoch": 131.53846153846155,
+      "grad_norm": 1.7214469909667969,
+      "learning_rate": 0.0001,
+      "loss": 0.0222,
+      "step": 22230
+    },
+    {
+      "epoch": 131.5976331360947,
+      "grad_norm": 1.2678046226501465,
+      "learning_rate": 0.0001,
+      "loss": 0.0226,
+      "step": 22240
+    },
+    {
+      "epoch": 131.65680473372782,
+      "grad_norm": 1.1631290912628174,
+      "learning_rate": 0.0001,
+      "loss": 0.0233,
+      "step": 22250
+    },
+    {
+      "epoch": 131.71597633136093,
+      "grad_norm": 1.1317272186279297,
+      "learning_rate": 0.0001,
+      "loss": 0.0228,
+      "step": 22260
+    },
+    {
+      "epoch": 131.77514792899407,
+      "grad_norm": 1.3560905456542969,
+      "learning_rate": 0.0001,
+      "loss": 0.0226,
+      "step": 22270
+    },
+    {
+      "epoch": 131.8343195266272,
+      "grad_norm": 1.9673725366592407,
+      "learning_rate": 0.0001,
+      "loss": 0.0233,
+      "step": 22280
+    },
+    {
+      "epoch": 131.89349112426035,
+      "grad_norm": 1.679540753364563,
+      "learning_rate": 0.0001,
+      "loss": 0.0233,
+      "step": 22290
+    },
+    {
+      "epoch": 131.9526627218935,
+      "grad_norm": 1.3510504961013794,
+      "learning_rate": 0.0001,
+      "loss": 0.0228,
+      "step": 22300
+    },
+    {
+      "epoch": 132.01183431952663,
+      "grad_norm": 1.5680084228515625,
+      "learning_rate": 0.0001,
+      "loss": 0.0221,
+      "step": 22310
+    },
+    {
+      "epoch": 132.07100591715977,
+      "grad_norm": 1.4863905906677246,
+      "learning_rate": 0.0001,
+      "loss": 0.0218,
+      "step": 22320
+    },
+    {
+      "epoch": 132.1301775147929,
+      "grad_norm": 1.3025755882263184,
+      "learning_rate": 0.0001,
+      "loss": 0.0215,
+      "step": 22330
+    },
+    {
+      "epoch": 132.18934911242604,
+      "grad_norm": 1.6381012201309204,
+      "learning_rate": 0.0001,
+      "loss": 0.0222,
+      "step": 22340
+    },
+    {
+      "epoch": 132.24852071005918,
+      "grad_norm": 1.404417872428894,
+      "learning_rate": 0.0001,
+      "loss": 0.0221,
+      "step": 22350
+    },
+    {
+      "epoch": 132.30769230769232,
+      "grad_norm": 1.5084037780761719,
+      "learning_rate": 0.0001,
+      "loss": 0.0233,
+      "step": 22360
+    },
+    {
+      "epoch": 132.36686390532543,
+      "grad_norm": 1.794124960899353,
+      "learning_rate": 0.0001,
+      "loss": 0.0226,
+      "step": 22370
+    },
+    {
+      "epoch": 132.42603550295857,
+      "grad_norm": 1.1071964502334595,
+      "learning_rate": 0.0001,
+      "loss": 0.0215,
+      "step": 22380
+    },
+    {
+      "epoch": 132.4852071005917,
+      "grad_norm": 1.6524299383163452,
+      "learning_rate": 0.0001,
+      "loss": 0.0214,
+      "step": 22390
+    },
+    {
+      "epoch": 132.54437869822485,
+      "grad_norm": 1.3592467308044434,
+      "learning_rate": 0.0001,
+      "loss": 0.0215,
+      "step": 22400
+    },
+    {
+      "epoch": 132.60355029585799,
+      "grad_norm": 1.6825467348098755,
+      "learning_rate": 0.0001,
+      "loss": 0.0215,
+      "step": 22410
+    },
+    {
+      "epoch": 132.66272189349112,
+      "grad_norm": 1.6526931524276733,
+      "learning_rate": 0.0001,
+      "loss": 0.0215,
+      "step": 22420
+    },
+    {
+      "epoch": 132.72189349112426,
+      "grad_norm": 1.713154911994934,
+      "learning_rate": 0.0001,
+      "loss": 0.0222,
+      "step": 22430
+    },
+    {
+      "epoch": 132.7810650887574,
+      "grad_norm": 1.3124300241470337,
+      "learning_rate": 0.0001,
+      "loss": 0.0218,
+      "step": 22440
+    },
+    {
+      "epoch": 132.84023668639054,
+      "grad_norm": 1.491317629814148,
+      "learning_rate": 0.0001,
+      "loss": 0.022,
+      "step": 22450
+    },
+    {
+      "epoch": 132.89940828402368,
+      "grad_norm": 1.1711320877075195,
+      "learning_rate": 0.0001,
+      "loss": 0.0207,
+      "step": 22460
+    },
+    {
+      "epoch": 132.95857988165682,
+      "grad_norm": 1.4871643781661987,
+      "learning_rate": 0.0001,
+      "loss": 0.022,
+      "step": 22470
+    },
+    {
+      "epoch": 133.01775147928993,
+      "grad_norm": 1.5804264545440674,
+      "learning_rate": 0.0001,
+      "loss": 0.0226,
+      "step": 22480
+    },
+    {
+      "epoch": 133.07692307692307,
+      "grad_norm": 1.646530270576477,
+      "learning_rate": 0.0001,
+      "loss": 0.021,
+      "step": 22490
+    },
+    {
+      "epoch": 133.1360946745562,
+      "grad_norm": 1.4927196502685547,
+      "learning_rate": 0.0001,
+      "loss": 0.0222,
+      "step": 22500
+    },
+    {
+      "epoch": 133.19526627218934,
+      "grad_norm": 1.4430549144744873,
+      "learning_rate": 0.0001,
+      "loss": 0.0222,
+      "step": 22510
+    },
+    {
+      "epoch": 133.25443786982248,
+      "grad_norm": 1.8244553804397583,
+      "learning_rate": 0.0001,
+      "loss": 0.0224,
+      "step": 22520
+    },
+    {
+      "epoch": 133.31360946745562,
+      "grad_norm": 1.3352702856063843,
+      "learning_rate": 0.0001,
+      "loss": 0.0218,
+      "step": 22530
+    },
+    {
+      "epoch": 133.37278106508876,
+      "grad_norm": 1.654445767402649,
+      "learning_rate": 0.0001,
+      "loss": 0.0211,
+      "step": 22540
+    },
+    {
+      "epoch": 133.4319526627219,
+      "grad_norm": 1.6577790975570679,
+      "learning_rate": 0.0001,
+      "loss": 0.0224,
+      "step": 22550
+    },
+    {
+      "epoch": 133.49112426035504,
+      "grad_norm": 1.322371482849121,
+      "learning_rate": 0.0001,
+      "loss": 0.0214,
+      "step": 22560
+    },
+    {
+      "epoch": 133.55029585798817,
+      "grad_norm": 1.1540875434875488,
+      "learning_rate": 0.0001,
+      "loss": 0.021,
+      "step": 22570
+    },
+    {
+      "epoch": 133.6094674556213,
+      "grad_norm": 2.0015499591827393,
+      "learning_rate": 0.0001,
+      "loss": 0.0218,
+      "step": 22580
+    },
+    {
+      "epoch": 133.66863905325442,
+      "grad_norm": 1.6411784887313843,
+      "learning_rate": 0.0001,
+      "loss": 0.0231,
+      "step": 22590
+    },
+    {
+      "epoch": 133.72781065088756,
+      "grad_norm": 1.4953075647354126,
+      "learning_rate": 0.0001,
+      "loss": 0.0224,
+      "step": 22600
+    },
+    {
+      "epoch": 133.7869822485207,
+      "grad_norm": 1.6028869152069092,
+      "learning_rate": 0.0001,
+      "loss": 0.0215,
+      "step": 22610
+    },
+    {
+      "epoch": 133.84615384615384,
+      "grad_norm": 1.2864543199539185,
+      "learning_rate": 0.0001,
+      "loss": 0.0214,
+      "step": 22620
+    },
+    {
+      "epoch": 133.90532544378698,
+      "grad_norm": 1.1994976997375488,
+      "learning_rate": 0.0001,
+      "loss": 0.022,
+      "step": 22630
+    },
+    {
+      "epoch": 133.96449704142012,
+      "grad_norm": 1.1467247009277344,
+      "learning_rate": 0.0001,
+      "loss": 0.022,
+      "step": 22640
+    },
+    {
+      "epoch": 134.02366863905326,
+      "grad_norm": 1.3722200393676758,
+      "learning_rate": 0.0001,
+      "loss": 0.0217,
+      "step": 22650
+    },
+    {
+      "epoch": 134.0828402366864,
+      "grad_norm": 1.555859088897705,
+      "learning_rate": 0.0001,
+      "loss": 0.0221,
+      "step": 22660
+    },
+    {
+      "epoch": 134.14201183431953,
+      "grad_norm": 1.442474603652954,
+      "learning_rate": 0.0001,
+      "loss": 0.0224,
+      "step": 22670
+    },
+    {
+      "epoch": 134.20118343195267,
+      "grad_norm": 1.7671356201171875,
+      "learning_rate": 0.0001,
+      "loss": 0.0224,
+      "step": 22680
+    },
+    {
+      "epoch": 134.2603550295858,
+      "grad_norm": 1.4419420957565308,
+      "learning_rate": 0.0001,
+      "loss": 0.0222,
+      "step": 22690
+    },
+    {
+      "epoch": 134.31952662721895,
+      "grad_norm": 1.283589482307434,
+      "learning_rate": 0.0001,
+      "loss": 0.0229,
+      "step": 22700
+    },
+    {
+      "epoch": 134.37869822485206,
+      "grad_norm": 1.648884892463684,
+      "learning_rate": 0.0001,
+      "loss": 0.0219,
+      "step": 22710
+    },
+    {
+      "epoch": 134.4378698224852,
+      "grad_norm": 1.148618221282959,
+      "learning_rate": 0.0001,
+      "loss": 0.0219,
+      "step": 22720
+    },
+    {
+      "epoch": 134.49704142011834,
+      "grad_norm": 1.5184396505355835,
+      "learning_rate": 0.0001,
+      "loss": 0.0216,
+      "step": 22730
+    },
+    {
+      "epoch": 134.55621301775147,
+      "grad_norm": 1.3241599798202515,
+      "learning_rate": 0.0001,
+      "loss": 0.0214,
+      "step": 22740
+    },
+    {
+      "epoch": 134.6153846153846,
+      "grad_norm": 1.879340648651123,
+      "learning_rate": 0.0001,
+      "loss": 0.0215,
+      "step": 22750
+    },
+    {
+      "epoch": 134.67455621301775,
+      "grad_norm": 1.4771236181259155,
+      "learning_rate": 0.0001,
+      "loss": 0.021,
+      "step": 22760
+    },
+    {
+      "epoch": 134.7337278106509,
+      "grad_norm": 1.5463896989822388,
+      "learning_rate": 0.0001,
+      "loss": 0.0223,
+      "step": 22770
+    },
+    {
+      "epoch": 134.79289940828403,
+      "grad_norm": 1.4774996042251587,
+      "learning_rate": 0.0001,
+      "loss": 0.0217,
+      "step": 22780
+    },
+    {
+      "epoch": 134.85207100591717,
+      "grad_norm": 1.616721510887146,
+      "learning_rate": 0.0001,
+      "loss": 0.0222,
+      "step": 22790
+    },
+    {
+      "epoch": 134.9112426035503,
+      "grad_norm": 1.6535022258758545,
+      "learning_rate": 0.0001,
+      "loss": 0.0216,
+      "step": 22800
+    },
+    {
+      "epoch": 134.97041420118344,
+      "grad_norm": 1.499223232269287,
+      "learning_rate": 0.0001,
+      "loss": 0.0217,
+      "step": 22810
+    },
+    {
+      "epoch": 135.02958579881656,
+      "grad_norm": 2.269160270690918,
+      "learning_rate": 0.0001,
+      "loss": 0.0219,
+      "step": 22820
+    },
+    {
+      "epoch": 135.0887573964497,
+      "grad_norm": 1.46591055393219,
+      "learning_rate": 0.0001,
+      "loss": 0.021,
+      "step": 22830
+    },
+    {
+      "epoch": 135.14792899408283,
+      "grad_norm": 1.51723051071167,
+      "learning_rate": 0.0001,
+      "loss": 0.0216,
+      "step": 22840
+    },
+    {
+      "epoch": 135.20710059171597,
+      "grad_norm": 1.260359287261963,
+      "learning_rate": 0.0001,
+      "loss": 0.0208,
+      "step": 22850
+    },
+    {
+      "epoch": 135.2662721893491,
+      "grad_norm": 1.6945911645889282,
+      "learning_rate": 0.0001,
+      "loss": 0.0215,
+      "step": 22860
+    },
+    {
+      "epoch": 135.32544378698225,
+      "grad_norm": 1.5379806756973267,
+      "learning_rate": 0.0001,
+      "loss": 0.0218,
+      "step": 22870
+    },
+    {
+      "epoch": 135.3846153846154,
+      "grad_norm": 1.5574581623077393,
+      "learning_rate": 0.0001,
+      "loss": 0.0215,
+      "step": 22880
+    },
+    {
+      "epoch": 135.44378698224853,
+      "grad_norm": 1.3897563219070435,
+      "learning_rate": 0.0001,
+      "loss": 0.021,
+      "step": 22890
+    },
+    {
+      "epoch": 135.50295857988166,
+      "grad_norm": 1.3326464891433716,
+      "learning_rate": 0.0001,
+      "loss": 0.0208,
+      "step": 22900
+    },
+    {
+      "epoch": 135.5621301775148,
+      "grad_norm": 1.67226243019104,
+      "learning_rate": 0.0001,
+      "loss": 0.022,
+      "step": 22910
+    },
+    {
+      "epoch": 135.62130177514794,
+      "grad_norm": 1.466518521308899,
+      "learning_rate": 0.0001,
+      "loss": 0.0213,
+      "step": 22920
+    },
+    {
+      "epoch": 135.68047337278105,
+      "grad_norm": 1.5972672700881958,
+      "learning_rate": 0.0001,
+      "loss": 0.0207,
+      "step": 22930
+    },
+    {
+      "epoch": 135.7396449704142,
+      "grad_norm": 1.6881394386291504,
+      "learning_rate": 0.0001,
+      "loss": 0.0217,
+      "step": 22940
+    },
+    {
+      "epoch": 135.79881656804733,
+      "grad_norm": 1.3864045143127441,
+      "learning_rate": 0.0001,
+      "loss": 0.0216,
+      "step": 22950
+    },
+    {
+      "epoch": 135.85798816568047,
+      "grad_norm": 1.542222499847412,
+      "learning_rate": 0.0001,
+      "loss": 0.0215,
+      "step": 22960
+    },
+    {
+      "epoch": 135.9171597633136,
+      "grad_norm": 1.0768643617630005,
+      "learning_rate": 0.0001,
+      "loss": 0.0214,
+      "step": 22970
+    },
+    {
+      "epoch": 135.97633136094674,
+      "grad_norm": 1.405646562576294,
+      "learning_rate": 0.0001,
+      "loss": 0.0212,
+      "step": 22980
+    },
+    {
+      "epoch": 136.03550295857988,
+      "grad_norm": 1.311967372894287,
+      "learning_rate": 0.0001,
+      "loss": 0.0214,
+      "step": 22990
+    },
+    {
+      "epoch": 136.09467455621302,
+      "grad_norm": 1.4330273866653442,
+      "learning_rate": 0.0001,
+      "loss": 0.0212,
+      "step": 23000
+    },
+    {
+      "epoch": 136.15384615384616,
+      "grad_norm": 1.1867833137512207,
+      "learning_rate": 0.0001,
+      "loss": 0.0212,
+      "step": 23010
+    },
+    {
+      "epoch": 136.2130177514793,
+      "grad_norm": 1.6134796142578125,
+      "learning_rate": 0.0001,
+      "loss": 0.0215,
+      "step": 23020
+    },
+    {
+      "epoch": 136.27218934911244,
+      "grad_norm": 1.201897144317627,
+      "learning_rate": 0.0001,
+      "loss": 0.0216,
+      "step": 23030
+    },
+    {
+      "epoch": 136.33136094674558,
+      "grad_norm": 1.7867860794067383,
+      "learning_rate": 0.0001,
+      "loss": 0.0221,
+      "step": 23040
+    },
+    {
+      "epoch": 136.3905325443787,
+      "grad_norm": 1.5361649990081787,
+      "learning_rate": 0.0001,
+      "loss": 0.0225,
+      "step": 23050
+    },
+    {
+      "epoch": 136.44970414201183,
+      "grad_norm": 1.5910617113113403,
+      "learning_rate": 0.0001,
+      "loss": 0.0216,
+      "step": 23060
+    },
+    {
+      "epoch": 136.50887573964496,
+      "grad_norm": 1.629176139831543,
+      "learning_rate": 0.0001,
+      "loss": 0.0221,
+      "step": 23070
+    },
+    {
+      "epoch": 136.5680473372781,
+      "grad_norm": 1.4280478954315186,
+      "learning_rate": 0.0001,
+      "loss": 0.0219,
+      "step": 23080
+    },
+    {
+      "epoch": 136.62721893491124,
+      "grad_norm": 1.1592539548873901,
+      "learning_rate": 0.0001,
+      "loss": 0.0228,
+      "step": 23090
+    },
+    {
+      "epoch": 136.68639053254438,
+      "grad_norm": 1.5471359491348267,
+      "learning_rate": 0.0001,
+      "loss": 0.0222,
+      "step": 23100
+    },
+    {
+      "epoch": 136.74556213017752,
+      "grad_norm": 1.9148070812225342,
+      "learning_rate": 0.0001,
+      "loss": 0.0224,
+      "step": 23110
+    },
+    {
+      "epoch": 136.80473372781066,
+      "grad_norm": 2.065307378768921,
+      "learning_rate": 0.0001,
+      "loss": 0.0244,
+      "step": 23120
+    },
+    {
+      "epoch": 136.8639053254438,
+      "grad_norm": 1.731217622756958,
+      "learning_rate": 0.0001,
+      "loss": 0.0239,
+      "step": 23130
+    },
+    {
+      "epoch": 136.92307692307693,
+      "grad_norm": 1.2442898750305176,
+      "learning_rate": 0.0001,
+      "loss": 0.0227,
+      "step": 23140
+    },
+    {
+      "epoch": 136.98224852071007,
+      "grad_norm": 1.4652217626571655,
+      "learning_rate": 0.0001,
+      "loss": 0.0231,
+      "step": 23150
+    },
+    {
+      "epoch": 137.04142011834318,
+      "grad_norm": 1.2781460285186768,
+      "learning_rate": 0.0001,
+      "loss": 0.0218,
+      "step": 23160
+    },
+    {
+      "epoch": 137.10059171597632,
+      "grad_norm": 1.0592002868652344,
+      "learning_rate": 0.0001,
+      "loss": 0.0218,
+      "step": 23170
+    },
+    {
+      "epoch": 137.15976331360946,
+      "grad_norm": 1.1652051210403442,
+      "learning_rate": 0.0001,
+      "loss": 0.0217,
+      "step": 23180
+    },
+    {
+      "epoch": 137.2189349112426,
+      "grad_norm": 1.453547477722168,
+      "learning_rate": 0.0001,
+      "loss": 0.0224,
+      "step": 23190
+    },
+    {
+      "epoch": 137.27810650887574,
+      "grad_norm": 1.7044692039489746,
+      "learning_rate": 0.0001,
+      "loss": 0.0225,
+      "step": 23200
+    },
+    {
+      "epoch": 137.33727810650888,
+      "grad_norm": 1.4545385837554932,
+      "learning_rate": 0.0001,
+      "loss": 0.0222,
+      "step": 23210
+    },
+    {
+      "epoch": 137.39644970414201,
+      "grad_norm": 1.1384074687957764,
+      "learning_rate": 0.0001,
+      "loss": 0.022,
+      "step": 23220
+    },
+    {
+      "epoch": 137.45562130177515,
+      "grad_norm": 1.153648018836975,
+      "learning_rate": 0.0001,
+      "loss": 0.0218,
+      "step": 23230
+    },
+    {
+      "epoch": 137.5147928994083,
+      "grad_norm": 1.747144103050232,
+      "learning_rate": 0.0001,
+      "loss": 0.0225,
+      "step": 23240
+    },
+    {
+      "epoch": 137.57396449704143,
+      "grad_norm": 1.5110656023025513,
+      "learning_rate": 0.0001,
+      "loss": 0.0221,
+      "step": 23250
+    },
+    {
+      "epoch": 137.63313609467457,
+      "grad_norm": 1.9595192670822144,
+      "learning_rate": 0.0001,
+      "loss": 0.0215,
+      "step": 23260
+    },
+    {
+      "epoch": 137.69230769230768,
+      "grad_norm": 1.539832353591919,
+      "learning_rate": 0.0001,
+      "loss": 0.0212,
+      "step": 23270
+    },
+    {
+      "epoch": 137.75147928994082,
+      "grad_norm": 1.688449501991272,
+      "learning_rate": 0.0001,
+      "loss": 0.0221,
+      "step": 23280
+    },
+    {
+      "epoch": 137.81065088757396,
+      "grad_norm": 1.3839391469955444,
+      "learning_rate": 0.0001,
+      "loss": 0.0214,
+      "step": 23290
+    },
+    {
+      "epoch": 137.8698224852071,
+      "grad_norm": 1.841422438621521,
+      "learning_rate": 0.0001,
+      "loss": 0.022,
+      "step": 23300
+    },
+    {
+      "epoch": 137.92899408284023,
+      "grad_norm": 1.579600214958191,
+      "learning_rate": 0.0001,
+      "loss": 0.0213,
+      "step": 23310
+    },
+    {
+      "epoch": 137.98816568047337,
+      "grad_norm": 1.2686787843704224,
+      "learning_rate": 0.0001,
+      "loss": 0.022,
+      "step": 23320
+    },
+    {
+      "epoch": 138.0473372781065,
+      "grad_norm": 1.6413787603378296,
+      "learning_rate": 0.0001,
+      "loss": 0.0207,
+      "step": 23330
+    },
+    {
+      "epoch": 138.10650887573965,
+      "grad_norm": 1.2481971979141235,
+      "learning_rate": 0.0001,
+      "loss": 0.0211,
+      "step": 23340
+    },
+    {
+      "epoch": 138.1656804733728,
+      "grad_norm": 1.008709192276001,
+      "learning_rate": 0.0001,
+      "loss": 0.0212,
+      "step": 23350
+    },
+    {
+      "epoch": 138.22485207100593,
+      "grad_norm": 1.814927339553833,
+      "learning_rate": 0.0001,
+      "loss": 0.0219,
+      "step": 23360
+    },
+    {
+      "epoch": 138.28402366863907,
+      "grad_norm": 1.3550914525985718,
+      "learning_rate": 0.0001,
+      "loss": 0.0229,
+      "step": 23370
+    },
+    {
+      "epoch": 138.34319526627218,
+      "grad_norm": 1.811688780784607,
+      "learning_rate": 0.0001,
+      "loss": 0.0223,
+      "step": 23380
+    },
+    {
+      "epoch": 138.4023668639053,
+      "grad_norm": 2.0709118843078613,
+      "learning_rate": 0.0001,
+      "loss": 0.0227,
+      "step": 23390
+    },
+    {
+      "epoch": 138.46153846153845,
+      "grad_norm": 1.3581500053405762,
+      "learning_rate": 0.0001,
+      "loss": 0.0223,
+      "step": 23400
+    },
+    {
+      "epoch": 138.5207100591716,
+      "grad_norm": 2.0751867294311523,
+      "learning_rate": 0.0001,
+      "loss": 0.0219,
+      "step": 23410
+    },
+    {
+      "epoch": 138.57988165680473,
+      "grad_norm": 1.16642427444458,
+      "learning_rate": 0.0001,
+      "loss": 0.0219,
+      "step": 23420
+    },
+    {
+      "epoch": 138.63905325443787,
+      "grad_norm": 1.10236656665802,
+      "learning_rate": 0.0001,
+      "loss": 0.021,
+      "step": 23430
+    },
+    {
+      "epoch": 138.698224852071,
+      "grad_norm": 1.3413803577423096,
+      "learning_rate": 0.0001,
+      "loss": 0.021,
+      "step": 23440
+    },
+    {
+      "epoch": 138.75739644970415,
+      "grad_norm": 1.6522318124771118,
+      "learning_rate": 0.0001,
+      "loss": 0.0213,
+      "step": 23450
+    },
+    {
+      "epoch": 138.81656804733728,
+      "grad_norm": 1.3619824647903442,
+      "learning_rate": 0.0001,
+      "loss": 0.0229,
+      "step": 23460
+    },
+    {
+      "epoch": 138.87573964497042,
+      "grad_norm": 1.182519793510437,
+      "learning_rate": 0.0001,
+      "loss": 0.023,
+      "step": 23470
+    },
+    {
+      "epoch": 138.93491124260356,
+      "grad_norm": 1.5175858736038208,
+      "learning_rate": 0.0001,
+      "loss": 0.0216,
+      "step": 23480
+    },
+    {
+      "epoch": 138.9940828402367,
+      "grad_norm": 1.4886835813522339,
+      "learning_rate": 0.0001,
+      "loss": 0.0222,
+      "step": 23490
+    },
+    {
+      "epoch": 139.0532544378698,
+      "grad_norm": 1.4079160690307617,
+      "learning_rate": 0.0001,
+      "loss": 0.0223,
+      "step": 23500
+    },
+    {
+      "epoch": 139.11242603550295,
+      "grad_norm": 1.3034294843673706,
+      "learning_rate": 0.0001,
+      "loss": 0.0217,
+      "step": 23510
+    },
+    {
+      "epoch": 139.1715976331361,
+      "grad_norm": 0.935799777507782,
+      "learning_rate": 0.0001,
+      "loss": 0.0222,
+      "step": 23520
+    },
+    {
+      "epoch": 139.23076923076923,
+      "grad_norm": 1.1833924055099487,
+      "learning_rate": 0.0001,
+      "loss": 0.022,
+      "step": 23530
+    },
+    {
+      "epoch": 139.28994082840237,
+      "grad_norm": 1.563151240348816,
+      "learning_rate": 0.0001,
+      "loss": 0.0226,
+      "step": 23540
+    },
+    {
+      "epoch": 139.3491124260355,
+      "grad_norm": 0.9998111724853516,
+      "learning_rate": 0.0001,
+      "loss": 0.0227,
+      "step": 23550
+    },
+    {
+      "epoch": 139.40828402366864,
+      "grad_norm": 1.5661993026733398,
+      "learning_rate": 0.0001,
+      "loss": 0.0225,
+      "step": 23560
+    },
+    {
+      "epoch": 139.46745562130178,
+      "grad_norm": 1.6827553510665894,
+      "learning_rate": 0.0001,
+      "loss": 0.0219,
+      "step": 23570
+    },
+    {
+      "epoch": 139.52662721893492,
+      "grad_norm": 1.396440029144287,
+      "learning_rate": 0.0001,
+      "loss": 0.022,
+      "step": 23580
+    },
+    {
+      "epoch": 139.58579881656806,
+      "grad_norm": 1.6347497701644897,
+      "learning_rate": 0.0001,
+      "loss": 0.0221,
+      "step": 23590
+    },
+    {
+      "epoch": 139.6449704142012,
+      "grad_norm": 1.4759371280670166,
+      "learning_rate": 0.0001,
+      "loss": 0.0217,
+      "step": 23600
+    },
+    {
+      "epoch": 139.7041420118343,
+      "grad_norm": 1.5390127897262573,
+      "learning_rate": 0.0001,
+      "loss": 0.0207,
+      "step": 23610
+    },
+    {
+      "epoch": 139.76331360946745,
+      "grad_norm": 1.5270627737045288,
+      "learning_rate": 0.0001,
+      "loss": 0.0216,
+      "step": 23620
+    },
+    {
+      "epoch": 139.82248520710058,
+      "grad_norm": 1.5148276090621948,
+      "learning_rate": 0.0001,
+      "loss": 0.021,
+      "step": 23630
+    },
+    {
+      "epoch": 139.88165680473372,
+      "grad_norm": 1.007659912109375,
+      "learning_rate": 0.0001,
+      "loss": 0.0224,
+      "step": 23640
+    },
+    {
+      "epoch": 139.94082840236686,
+      "grad_norm": 1.4162933826446533,
+      "learning_rate": 0.0001,
+      "loss": 0.0213,
+      "step": 23650
+    },
+    {
+      "epoch": 140.0,
+      "grad_norm": 1.3758448362350464,
+      "learning_rate": 0.0001,
+      "loss": 0.0212,
+      "step": 23660
+    },
+    {
+      "epoch": 140.05917159763314,
+      "grad_norm": 1.3619537353515625,
+      "learning_rate": 0.0001,
+      "loss": 0.0209,
+      "step": 23670
+    },
+    {
+      "epoch": 140.11834319526628,
+      "grad_norm": 1.2015057802200317,
+      "learning_rate": 0.0001,
+      "loss": 0.0209,
+      "step": 23680
+    },
+    {
+      "epoch": 140.17751479289942,
+      "grad_norm": 1.1587117910385132,
+      "learning_rate": 0.0001,
+      "loss": 0.0214,
+      "step": 23690
+    },
+    {
+      "epoch": 140.23668639053255,
+      "grad_norm": 1.6333088874816895,
+      "learning_rate": 0.0001,
+      "loss": 0.0219,
+      "step": 23700
+    },
+    {
+      "epoch": 140.2958579881657,
+      "grad_norm": 1.4158580303192139,
+      "learning_rate": 0.0001,
+      "loss": 0.0219,
+      "step": 23710
+    },
+    {
+      "epoch": 140.3550295857988,
+      "grad_norm": 1.5921796560287476,
+      "learning_rate": 0.0001,
+      "loss": 0.0223,
+      "step": 23720
+    },
+    {
+      "epoch": 140.41420118343194,
+      "grad_norm": 1.8549199104309082,
+      "learning_rate": 0.0001,
+      "loss": 0.0221,
+      "step": 23730
+    },
+    {
+      "epoch": 140.47337278106508,
+      "grad_norm": 1.7742773294448853,
+      "learning_rate": 0.0001,
+      "loss": 0.0227,
+      "step": 23740
+    },
+    {
+      "epoch": 140.53254437869822,
+      "grad_norm": 1.8545560836791992,
+      "learning_rate": 0.0001,
+      "loss": 0.0216,
+      "step": 23750
+    },
+    {
+      "epoch": 140.59171597633136,
+      "grad_norm": 1.511832356452942,
+      "learning_rate": 0.0001,
+      "loss": 0.0218,
+      "step": 23760
+    },
+    {
+      "epoch": 140.6508875739645,
+      "grad_norm": 1.07363760471344,
+      "learning_rate": 0.0001,
+      "loss": 0.0218,
+      "step": 23770
+    },
+    {
+      "epoch": 140.71005917159763,
+      "grad_norm": 1.7059550285339355,
+      "learning_rate": 0.0001,
+      "loss": 0.022,
+      "step": 23780
+    },
+    {
+      "epoch": 140.76923076923077,
+      "grad_norm": 1.6168529987335205,
+      "learning_rate": 0.0001,
+      "loss": 0.022,
+      "step": 23790
+    },
+    {
+      "epoch": 140.8284023668639,
+      "grad_norm": 1.435173749923706,
+      "learning_rate": 0.0001,
+      "loss": 0.0217,
+      "step": 23800
+    },
+    {
+      "epoch": 140.88757396449705,
+      "grad_norm": 1.2082135677337646,
+      "learning_rate": 0.0001,
+      "loss": 0.021,
+      "step": 23810
+    },
+    {
+      "epoch": 140.9467455621302,
+      "grad_norm": 1.5283962488174438,
+      "learning_rate": 0.0001,
+      "loss": 0.0223,
+      "step": 23820
+    },
+    {
+      "epoch": 141.0059171597633,
+      "grad_norm": 1.3792208433151245,
+      "learning_rate": 0.0001,
+      "loss": 0.0223,
+      "step": 23830
+    },
+    {
+      "epoch": 141.06508875739644,
+      "grad_norm": 1.3916105031967163,
+      "learning_rate": 0.0001,
+      "loss": 0.0222,
+      "step": 23840
+    },
+    {
+      "epoch": 141.12426035502958,
+      "grad_norm": 1.782760739326477,
+      "learning_rate": 0.0001,
+      "loss": 0.0227,
+      "step": 23850
+    },
+    {
+      "epoch": 141.18343195266272,
+      "grad_norm": 1.7008661031723022,
+      "learning_rate": 0.0001,
+      "loss": 0.0221,
+      "step": 23860
+    },
+    {
+      "epoch": 141.24260355029585,
+      "grad_norm": 1.1160820722579956,
+      "learning_rate": 0.0001,
+      "loss": 0.0222,
+      "step": 23870
+    },
+    {
+      "epoch": 141.301775147929,
+      "grad_norm": 1.8799108266830444,
+      "learning_rate": 0.0001,
+      "loss": 0.0224,
+      "step": 23880
+    },
+    {
+      "epoch": 141.36094674556213,
+      "grad_norm": 1.125301480293274,
+      "learning_rate": 0.0001,
+      "loss": 0.0223,
+      "step": 23890
+    },
+    {
+      "epoch": 141.42011834319527,
+      "grad_norm": 1.601943016052246,
+      "learning_rate": 0.0001,
+      "loss": 0.0227,
+      "step": 23900
+    },
+    {
+      "epoch": 141.4792899408284,
+      "grad_norm": 1.4895862340927124,
+      "learning_rate": 0.0001,
+      "loss": 0.0222,
+      "step": 23910
+    },
+    {
+      "epoch": 141.53846153846155,
+      "grad_norm": 1.436660647392273,
+      "learning_rate": 0.0001,
+      "loss": 0.0227,
+      "step": 23920
+    },
+    {
+      "epoch": 141.5976331360947,
+      "grad_norm": 1.3656383752822876,
+      "learning_rate": 0.0001,
+      "loss": 0.0227,
+      "step": 23930
+    },
+    {
+      "epoch": 141.65680473372782,
+      "grad_norm": 1.7161632776260376,
+      "learning_rate": 0.0001,
+      "loss": 0.0214,
+      "step": 23940
+    },
+    {
+      "epoch": 141.71597633136093,
+      "grad_norm": 1.5660655498504639,
+      "learning_rate": 0.0001,
+      "loss": 0.0218,
+      "step": 23950
+    },
+    {
+      "epoch": 141.77514792899407,
+      "grad_norm": 1.415414810180664,
+      "learning_rate": 0.0001,
+      "loss": 0.022,
+      "step": 23960
+    },
+    {
+      "epoch": 141.8343195266272,
+      "grad_norm": 2.3683743476867676,
+      "learning_rate": 0.0001,
+      "loss": 0.0218,
+      "step": 23970
+    },
+    {
+      "epoch": 141.89349112426035,
+      "grad_norm": 1.4694912433624268,
+      "learning_rate": 0.0001,
+      "loss": 0.021,
+      "step": 23980
+    },
+    {
+      "epoch": 141.9526627218935,
+      "grad_norm": 1.4214612245559692,
+      "learning_rate": 0.0001,
+      "loss": 0.0215,
+      "step": 23990
+    },
+    {
+      "epoch": 142.01183431952663,
+      "grad_norm": 2.6020617485046387,
+      "learning_rate": 0.0001,
+      "loss": 0.0223,
+      "step": 24000
+    },
+    {
+      "epoch": 142.07100591715977,
+      "grad_norm": 1.8711864948272705,
+      "learning_rate": 0.0001,
+      "loss": 0.0216,
+      "step": 24010
+    },
+    {
+      "epoch": 142.1301775147929,
+      "grad_norm": 1.758337378501892,
+      "learning_rate": 0.0001,
+      "loss": 0.0218,
+      "step": 24020
+    },
+    {
+      "epoch": 142.18934911242604,
+      "grad_norm": 1.6721073389053345,
+      "learning_rate": 0.0001,
+      "loss": 0.0213,
+      "step": 24030
+    },
+    {
+      "epoch": 142.24852071005918,
+      "grad_norm": 1.481138825416565,
+      "learning_rate": 0.0001,
+      "loss": 0.0202,
+      "step": 24040
+    },
+    {
+      "epoch": 142.30769230769232,
+      "grad_norm": 1.41562020778656,
+      "learning_rate": 0.0001,
+      "loss": 0.0204,
+      "step": 24050
+    },
+    {
+      "epoch": 142.36686390532543,
+      "grad_norm": 2.1933305263519287,
+      "learning_rate": 0.0001,
+      "loss": 0.0213,
+      "step": 24060
+    },
+    {
+      "epoch": 142.42603550295857,
+      "grad_norm": 1.2705397605895996,
+      "learning_rate": 0.0001,
+      "loss": 0.0213,
+      "step": 24070
+    },
+    {
+      "epoch": 142.4852071005917,
+      "grad_norm": 1.5867044925689697,
+      "learning_rate": 0.0001,
+      "loss": 0.0211,
+      "step": 24080
+    },
+    {
+      "epoch": 142.54437869822485,
+      "grad_norm": 1.8735051155090332,
+      "learning_rate": 0.0001,
+      "loss": 0.0203,
+      "step": 24090
+    },
+    {
+      "epoch": 142.60355029585799,
+      "grad_norm": 1.678731083869934,
+      "learning_rate": 0.0001,
+      "loss": 0.021,
+      "step": 24100
+    },
+    {
+      "epoch": 142.66272189349112,
+      "grad_norm": 1.6542314291000366,
+      "learning_rate": 0.0001,
+      "loss": 0.0204,
+      "step": 24110
+    },
+    {
+      "epoch": 142.72189349112426,
+      "grad_norm": 1.6461701393127441,
+      "learning_rate": 0.0001,
+      "loss": 0.0206,
+      "step": 24120
+    },
+    {
+      "epoch": 142.7810650887574,
+      "grad_norm": 1.6911569833755493,
+      "learning_rate": 0.0001,
+      "loss": 0.0205,
+      "step": 24130
+    },
+    {
+      "epoch": 142.84023668639054,
+      "grad_norm": 1.5840963125228882,
+      "learning_rate": 0.0001,
+      "loss": 0.0207,
+      "step": 24140
+    },
+    {
+      "epoch": 142.89940828402368,
+      "grad_norm": 1.4853278398513794,
+      "learning_rate": 0.0001,
+      "loss": 0.0205,
+      "step": 24150
+    },
+    {
+      "epoch": 142.95857988165682,
+      "grad_norm": 1.7572276592254639,
+      "learning_rate": 0.0001,
+      "loss": 0.0208,
+      "step": 24160
+    },
+    {
+      "epoch": 143.01775147928993,
+      "grad_norm": 1.743883490562439,
+      "learning_rate": 0.0001,
+      "loss": 0.0203,
+      "step": 24170
+    },
+    {
+      "epoch": 143.07692307692307,
+      "grad_norm": 1.805227279663086,
+      "learning_rate": 0.0001,
+      "loss": 0.0208,
+      "step": 24180
+    },
+    {
+      "epoch": 143.1360946745562,
+      "grad_norm": 1.429710865020752,
+      "learning_rate": 0.0001,
+      "loss": 0.0207,
+      "step": 24190
+    },
+    {
+      "epoch": 143.19526627218934,
+      "grad_norm": 1.7176575660705566,
+      "learning_rate": 0.0001,
+      "loss": 0.0218,
+      "step": 24200
+    },
+    {
+      "epoch": 143.25443786982248,
+      "grad_norm": 1.8963769674301147,
+      "learning_rate": 0.0001,
+      "loss": 0.0214,
+      "step": 24210
+    },
+    {
+      "epoch": 143.31360946745562,
+      "grad_norm": 1.1980042457580566,
+      "learning_rate": 0.0001,
+      "loss": 0.0211,
+      "step": 24220
+    },
+    {
+      "epoch": 143.37278106508876,
+      "grad_norm": 1.2232005596160889,
+      "learning_rate": 0.0001,
+      "loss": 0.0211,
+      "step": 24230
+    },
+    {
+      "epoch": 143.4319526627219,
+      "grad_norm": 1.5614715814590454,
+      "learning_rate": 0.0001,
+      "loss": 0.0214,
+      "step": 24240
+    },
+    {
+      "epoch": 143.49112426035504,
+      "grad_norm": 1.6883203983306885,
+      "learning_rate": 0.0001,
+      "loss": 0.0211,
+      "step": 24250
+    },
+    {
+      "epoch": 143.55029585798817,
+      "grad_norm": 1.889915943145752,
+      "learning_rate": 0.0001,
+      "loss": 0.0214,
+      "step": 24260
+    },
+    {
+      "epoch": 143.6094674556213,
+      "grad_norm": 1.3784376382827759,
+      "learning_rate": 0.0001,
+      "loss": 0.0217,
+      "step": 24270
+    },
+    {
+      "epoch": 143.66863905325442,
+      "grad_norm": 1.116148591041565,
+      "learning_rate": 0.0001,
+      "loss": 0.0215,
+      "step": 24280
+    },
+    {
+      "epoch": 143.72781065088756,
+      "grad_norm": 1.688830852508545,
+      "learning_rate": 0.0001,
+      "loss": 0.0229,
+      "step": 24290
+    },
+    {
+      "epoch": 143.7869822485207,
+      "grad_norm": 1.826689600944519,
+      "learning_rate": 0.0001,
+      "loss": 0.0233,
+      "step": 24300
+    },
+    {
+      "epoch": 143.84615384615384,
+      "grad_norm": 1.074652075767517,
+      "learning_rate": 0.0001,
+      "loss": 0.0226,
+      "step": 24310
+    },
+    {
+      "epoch": 143.90532544378698,
+      "grad_norm": 1.2936787605285645,
+      "learning_rate": 0.0001,
+      "loss": 0.0209,
+      "step": 24320
+    },
+    {
+      "epoch": 143.96449704142012,
+      "grad_norm": 1.3530282974243164,
+      "learning_rate": 0.0001,
+      "loss": 0.0212,
+      "step": 24330
+    },
+    {
+      "epoch": 144.02366863905326,
+      "grad_norm": 1.3657077550888062,
+      "learning_rate": 0.0001,
+      "loss": 0.0217,
+      "step": 24340
+    },
+    {
+      "epoch": 144.0828402366864,
+      "grad_norm": 0.9302562475204468,
+      "learning_rate": 0.0001,
+      "loss": 0.0211,
+      "step": 24350
+    },
+    {
+      "epoch": 144.14201183431953,
+      "grad_norm": 1.9408231973648071,
+      "learning_rate": 0.0001,
+      "loss": 0.0221,
+      "step": 24360
+    },
+    {
+      "epoch": 144.20118343195267,
+      "grad_norm": 1.7521504163742065,
+      "learning_rate": 0.0001,
+      "loss": 0.0216,
+      "step": 24370
+    },
+    {
+      "epoch": 144.2603550295858,
+      "grad_norm": 1.2671380043029785,
+      "learning_rate": 0.0001,
+      "loss": 0.0211,
+      "step": 24380
+    },
+    {
+      "epoch": 144.31952662721895,
+      "grad_norm": 1.4806045293807983,
+      "learning_rate": 0.0001,
+      "loss": 0.0211,
+      "step": 24390
+    },
+    {
+      "epoch": 144.37869822485206,
+      "grad_norm": 1.5571852922439575,
+      "learning_rate": 0.0001,
+      "loss": 0.0206,
+      "step": 24400
+    },
+    {
+      "epoch": 144.4378698224852,
+      "grad_norm": 1.4486463069915771,
+      "learning_rate": 0.0001,
+      "loss": 0.0209,
+      "step": 24410
+    },
+    {
+      "epoch": 144.49704142011834,
+      "grad_norm": 1.1941167116165161,
+      "learning_rate": 0.0001,
+      "loss": 0.0218,
+      "step": 24420
+    },
+    {
+      "epoch": 144.55621301775147,
+      "grad_norm": 1.624249815940857,
+      "learning_rate": 0.0001,
+      "loss": 0.0217,
+      "step": 24430
+    },
+    {
+      "epoch": 144.6153846153846,
+      "grad_norm": 1.8006073236465454,
+      "learning_rate": 0.0001,
+      "loss": 0.0217,
+      "step": 24440
+    },
+    {
+      "epoch": 144.67455621301775,
+      "grad_norm": 1.230064034461975,
+      "learning_rate": 0.0001,
+      "loss": 0.0205,
+      "step": 24450
+    },
+    {
+      "epoch": 144.7337278106509,
+      "grad_norm": 2.3492772579193115,
+      "learning_rate": 0.0001,
+      "loss": 0.0216,
+      "step": 24460
+    },
+    {
+      "epoch": 144.79289940828403,
+      "grad_norm": 1.5187469720840454,
+      "learning_rate": 0.0001,
+      "loss": 0.0212,
+      "step": 24470
+    },
+    {
+      "epoch": 144.85207100591717,
+      "grad_norm": 0.9291148781776428,
+      "learning_rate": 0.0001,
+      "loss": 0.0221,
+      "step": 24480
+    },
+    {
+      "epoch": 144.9112426035503,
+      "grad_norm": 0.94749915599823,
+      "learning_rate": 0.0001,
+      "loss": 0.0213,
+      "step": 24490
+    },
+    {
+      "epoch": 144.97041420118344,
+      "grad_norm": 1.611016869544983,
+      "learning_rate": 0.0001,
+      "loss": 0.0219,
+      "step": 24500
+    },
+    {
+      "epoch": 145.02958579881656,
+      "grad_norm": 1.105819582939148,
+      "learning_rate": 0.0001,
+      "loss": 0.0228,
+      "step": 24510
+    },
+    {
+      "epoch": 145.0887573964497,
+      "grad_norm": 1.6253284215927124,
+      "learning_rate": 0.0001,
+      "loss": 0.0226,
+      "step": 24520
+    },
+    {
+      "epoch": 145.14792899408283,
+      "grad_norm": 1.6301974058151245,
+      "learning_rate": 0.0001,
+      "loss": 0.0227,
+      "step": 24530
+    },
+    {
+      "epoch": 145.20710059171597,
+      "grad_norm": 1.7667920589447021,
+      "learning_rate": 0.0001,
+      "loss": 0.0228,
+      "step": 24540
+    },
+    {
+      "epoch": 145.2662721893491,
+      "grad_norm": 1.4631075859069824,
+      "learning_rate": 0.0001,
+      "loss": 0.0227,
+      "step": 24550
+    },
+    {
+      "epoch": 145.32544378698225,
+      "grad_norm": 1.3137433528900146,
+      "learning_rate": 0.0001,
+      "loss": 0.0224,
+      "step": 24560
+    },
+    {
+      "epoch": 145.3846153846154,
+      "grad_norm": 1.1705330610275269,
+      "learning_rate": 0.0001,
+      "loss": 0.0223,
+      "step": 24570
+    },
+    {
+      "epoch": 145.44378698224853,
+      "grad_norm": 1.5640184879302979,
+      "learning_rate": 0.0001,
+      "loss": 0.022,
+      "step": 24580
+    },
+    {
+      "epoch": 145.50295857988166,
+      "grad_norm": 1.889053463935852,
+      "learning_rate": 0.0001,
+      "loss": 0.0213,
+      "step": 24590
+    },
+    {
+      "epoch": 145.5621301775148,
+      "grad_norm": 1.2883977890014648,
+      "learning_rate": 0.0001,
+      "loss": 0.0217,
+      "step": 24600
+    },
+    {
+      "epoch": 145.62130177514794,
+      "grad_norm": 1.278820276260376,
+      "learning_rate": 0.0001,
+      "loss": 0.0214,
+      "step": 24610
+    },
+    {
+      "epoch": 145.68047337278105,
+      "grad_norm": 1.546832799911499,
+      "learning_rate": 0.0001,
+      "loss": 0.0217,
+      "step": 24620
+    },
+    {
+      "epoch": 145.7396449704142,
+      "grad_norm": 1.908720850944519,
+      "learning_rate": 0.0001,
+      "loss": 0.0217,
+      "step": 24630
+    },
+    {
+      "epoch": 145.79881656804733,
+      "grad_norm": 1.6055552959442139,
+      "learning_rate": 0.0001,
+      "loss": 0.021,
+      "step": 24640
+    },
+    {
+      "epoch": 145.85798816568047,
+      "grad_norm": 1.6154195070266724,
+      "learning_rate": 0.0001,
+      "loss": 0.0206,
+      "step": 24650
+    },
+    {
+      "epoch": 145.9171597633136,
+      "grad_norm": 1.3109469413757324,
+      "learning_rate": 0.0001,
+      "loss": 0.0215,
+      "step": 24660
+    },
+    {
+      "epoch": 145.97633136094674,
+      "grad_norm": 1.1860404014587402,
+      "learning_rate": 0.0001,
+      "loss": 0.0216,
+      "step": 24670
+    },
+    {
+      "epoch": 146.03550295857988,
+      "grad_norm": 1.9275978803634644,
+      "learning_rate": 0.0001,
+      "loss": 0.023,
+      "step": 24680
+    },
+    {
+      "epoch": 146.09467455621302,
+      "grad_norm": 1.7662742137908936,
+      "learning_rate": 0.0001,
+      "loss": 0.0227,
+      "step": 24690
+    },
+    {
+      "epoch": 146.15384615384616,
+      "grad_norm": 1.5083646774291992,
+      "learning_rate": 0.0001,
+      "loss": 0.0236,
+      "step": 24700
+    },
+    {
+      "epoch": 146.2130177514793,
+      "grad_norm": 1.7511041164398193,
+      "learning_rate": 0.0001,
+      "loss": 0.0222,
+      "step": 24710
+    },
+    {
+      "epoch": 146.27218934911244,
+      "grad_norm": 1.3487932682037354,
+      "learning_rate": 0.0001,
+      "loss": 0.0218,
+      "step": 24720
+    },
+    {
+      "epoch": 146.33136094674558,
+      "grad_norm": 1.3227574825286865,
+      "learning_rate": 0.0001,
+      "loss": 0.022,
+      "step": 24730
+    },
+    {
+      "epoch": 146.3905325443787,
+      "grad_norm": 1.513459324836731,
+      "learning_rate": 0.0001,
+      "loss": 0.0219,
+      "step": 24740
+    },
+    {
+      "epoch": 146.44970414201183,
+      "grad_norm": 1.2468619346618652,
+      "learning_rate": 0.0001,
+      "loss": 0.0221,
+      "step": 24750
+    },
+    {
+      "epoch": 146.50887573964496,
+      "grad_norm": 1.0894838571548462,
+      "learning_rate": 0.0001,
+      "loss": 0.0218,
+      "step": 24760
+    },
+    {
+      "epoch": 146.5680473372781,
+      "grad_norm": 1.008520245552063,
+      "learning_rate": 0.0001,
+      "loss": 0.0217,
+      "step": 24770
+    },
+    {
+      "epoch": 146.62721893491124,
+      "grad_norm": 1.2074817419052124,
+      "learning_rate": 0.0001,
+      "loss": 0.0215,
+      "step": 24780
+    },
+    {
+      "epoch": 146.68639053254438,
+      "grad_norm": 1.362784743309021,
+      "learning_rate": 0.0001,
+      "loss": 0.022,
+      "step": 24790
+    },
+    {
+      "epoch": 146.74556213017752,
+      "grad_norm": 1.3360635042190552,
+      "learning_rate": 0.0001,
+      "loss": 0.0224,
+      "step": 24800
+    },
+    {
+      "epoch": 146.80473372781066,
+      "grad_norm": 1.8148034811019897,
+      "learning_rate": 0.0001,
+      "loss": 0.0234,
+      "step": 24810
+    },
+    {
+      "epoch": 146.8639053254438,
+      "grad_norm": 1.7307058572769165,
+      "learning_rate": 0.0001,
+      "loss": 0.0232,
+      "step": 24820
+    },
+    {
+      "epoch": 146.92307692307693,
+      "grad_norm": 1.4149826765060425,
+      "learning_rate": 0.0001,
+      "loss": 0.0227,
+      "step": 24830
+    },
+    {
+      "epoch": 146.98224852071007,
+      "grad_norm": 1.487953543663025,
+      "learning_rate": 0.0001,
+      "loss": 0.0223,
+      "step": 24840
+    },
+    {
+      "epoch": 147.04142011834318,
+      "grad_norm": 1.5610328912734985,
+      "learning_rate": 0.0001,
+      "loss": 0.0213,
+      "step": 24850
+    },
+    {
+      "epoch": 147.10059171597632,
+      "grad_norm": 1.533774971961975,
+      "learning_rate": 0.0001,
+      "loss": 0.0214,
+      "step": 24860
+    },
+    {
+      "epoch": 147.15976331360946,
+      "grad_norm": 1.522325038909912,
+      "learning_rate": 0.0001,
+      "loss": 0.0211,
+      "step": 24870
+    },
+    {
+      "epoch": 147.2189349112426,
+      "grad_norm": 1.4288784265518188,
+      "learning_rate": 0.0001,
+      "loss": 0.0215,
+      "step": 24880
+    },
+    {
+      "epoch": 147.27810650887574,
+      "grad_norm": 1.3875890970230103,
+      "learning_rate": 0.0001,
+      "loss": 0.0215,
+      "step": 24890
+    },
+    {
+      "epoch": 147.33727810650888,
+      "grad_norm": 1.1390424966812134,
+      "learning_rate": 0.0001,
+      "loss": 0.0207,
+      "step": 24900
+    },
+    {
+      "epoch": 147.39644970414201,
+      "grad_norm": 1.6168129444122314,
+      "learning_rate": 0.0001,
+      "loss": 0.0223,
+      "step": 24910
+    },
+    {
+      "epoch": 147.45562130177515,
+      "grad_norm": 0.9985970258712769,
+      "learning_rate": 0.0001,
+      "loss": 0.0215,
+      "step": 24920
+    },
+    {
+      "epoch": 147.5147928994083,
+      "grad_norm": 1.6727039813995361,
+      "learning_rate": 0.0001,
+      "loss": 0.0211,
+      "step": 24930
+    },
+    {
+      "epoch": 147.57396449704143,
+      "grad_norm": 1.1086598634719849,
+      "learning_rate": 0.0001,
+      "loss": 0.0222,
+      "step": 24940
+    },
+    {
+      "epoch": 147.63313609467457,
+      "grad_norm": 0.9935755133628845,
+      "learning_rate": 0.0001,
+      "loss": 0.0211,
+      "step": 24950
+    },
+    {
+      "epoch": 147.69230769230768,
+      "grad_norm": 1.432590365409851,
+      "learning_rate": 0.0001,
+      "loss": 0.0213,
+      "step": 24960
+    },
+    {
+      "epoch": 147.75147928994082,
+      "grad_norm": 1.3184270858764648,
+      "learning_rate": 0.0001,
+      "loss": 0.022,
+      "step": 24970
+    },
+    {
+      "epoch": 147.81065088757396,
+      "grad_norm": 1.6467958688735962,
+      "learning_rate": 0.0001,
+      "loss": 0.0216,
+      "step": 24980
+    },
+    {
+      "epoch": 147.8698224852071,
+      "grad_norm": 1.9754306077957153,
+      "learning_rate": 0.0001,
+      "loss": 0.0221,
+      "step": 24990
+    },
+    {
+      "epoch": 147.92899408284023,
+      "grad_norm": 2.3045294284820557,
+      "learning_rate": 0.0001,
+      "loss": 0.0225,
+      "step": 25000
+    },
+    {
+      "epoch": 147.98816568047337,
+      "grad_norm": 1.9099152088165283,
+      "learning_rate": 0.0001,
+      "loss": 0.0231,
+      "step": 25010
+    },
+    {
+      "epoch": 148.0473372781065,
+      "grad_norm": 1.9629545211791992,
+      "learning_rate": 0.0001,
+      "loss": 0.0231,
+      "step": 25020
+    },
+    {
+      "epoch": 148.10650887573965,
+      "grad_norm": 1.7272634506225586,
+      "learning_rate": 0.0001,
+      "loss": 0.0228,
+      "step": 25030
+    },
+    {
+      "epoch": 148.1656804733728,
+      "grad_norm": 1.6915799379348755,
+      "learning_rate": 0.0001,
+      "loss": 0.0225,
+      "step": 25040
+    },
+    {
+      "epoch": 148.22485207100593,
+      "grad_norm": 1.8784633874893188,
+      "learning_rate": 0.0001,
+      "loss": 0.0222,
+      "step": 25050
+    },
+    {
+      "epoch": 148.28402366863907,
+      "grad_norm": 1.711879014968872,
+      "learning_rate": 0.0001,
+      "loss": 0.0223,
+      "step": 25060
+    },
+    {
+      "epoch": 148.34319526627218,
+      "grad_norm": 2.3307113647460938,
+      "learning_rate": 0.0001,
+      "loss": 0.0222,
+      "step": 25070
+    },
+    {
+      "epoch": 148.4023668639053,
+      "grad_norm": 2.070801258087158,
+      "learning_rate": 0.0001,
+      "loss": 0.0223,
+      "step": 25080
+    },
+    {
+      "epoch": 148.46153846153845,
+      "grad_norm": 1.773373007774353,
+      "learning_rate": 0.0001,
+      "loss": 0.0207,
+      "step": 25090
+    },
+    {
+      "epoch": 148.5207100591716,
+      "grad_norm": 0.9848353862762451,
+      "learning_rate": 0.0001,
+      "loss": 0.0203,
+      "step": 25100
+    },
+    {
+      "epoch": 148.57988165680473,
+      "grad_norm": 1.2250289916992188,
+      "learning_rate": 0.0001,
+      "loss": 0.0197,
+      "step": 25110
+    },
+    {
+      "epoch": 148.63905325443787,
+      "grad_norm": 0.9202752709388733,
+      "learning_rate": 0.0001,
+      "loss": 0.0205,
+      "step": 25120
+    },
+    {
+      "epoch": 148.698224852071,
+      "grad_norm": 1.4927376508712769,
+      "learning_rate": 0.0001,
+      "loss": 0.0209,
+      "step": 25130
+    },
+    {
+      "epoch": 148.75739644970415,
+      "grad_norm": 1.3392972946166992,
+      "learning_rate": 0.0001,
+      "loss": 0.0213,
+      "step": 25140
+    },
+    {
+      "epoch": 148.81656804733728,
+      "grad_norm": 1.3352611064910889,
+      "learning_rate": 0.0001,
+      "loss": 0.0213,
+      "step": 25150
+    },
+    {
+      "epoch": 148.87573964497042,
+      "grad_norm": 2.0409295558929443,
+      "learning_rate": 0.0001,
+      "loss": 0.0206,
+      "step": 25160
+    },
+    {
+      "epoch": 148.93491124260356,
+      "grad_norm": 1.6947757005691528,
+      "learning_rate": 0.0001,
+      "loss": 0.0225,
+      "step": 25170
+    },
+    {
+      "epoch": 148.9940828402367,
+      "grad_norm": 1.250657081604004,
+      "learning_rate": 0.0001,
+      "loss": 0.0209,
+      "step": 25180
+    },
+    {
+      "epoch": 149.0532544378698,
+      "grad_norm": 1.7786442041397095,
+      "learning_rate": 0.0001,
+      "loss": 0.0199,
+      "step": 25190
+    },
+    {
+      "epoch": 149.11242603550295,
+      "grad_norm": 2.175553321838379,
+      "learning_rate": 0.0001,
+      "loss": 0.0207,
+      "step": 25200
+    },
+    {
+      "epoch": 149.1715976331361,
+      "grad_norm": 1.5531606674194336,
+      "learning_rate": 0.0001,
+      "loss": 0.021,
+      "step": 25210
+    },
+    {
+      "epoch": 149.23076923076923,
+      "grad_norm": 2.3008904457092285,
+      "learning_rate": 0.0001,
+      "loss": 0.0215,
+      "step": 25220
+    },
+    {
+      "epoch": 149.28994082840237,
+      "grad_norm": 1.2973908185958862,
+      "learning_rate": 0.0001,
+      "loss": 0.0215,
+      "step": 25230
+    },
+    {
+      "epoch": 149.3491124260355,
+      "grad_norm": 1.5577019453048706,
+      "learning_rate": 0.0001,
+      "loss": 0.0202,
+      "step": 25240
+    },
+    {
+      "epoch": 149.40828402366864,
+      "grad_norm": 1.6389931440353394,
+      "learning_rate": 0.0001,
+      "loss": 0.0207,
+      "step": 25250
+    },
+    {
+      "epoch": 149.46745562130178,
+      "grad_norm": 1.3772011995315552,
+      "learning_rate": 0.0001,
+      "loss": 0.0205,
+      "step": 25260
+    },
+    {
+      "epoch": 149.52662721893492,
+      "grad_norm": 1.6344187259674072,
+      "learning_rate": 0.0001,
+      "loss": 0.0209,
+      "step": 25270
+    },
+    {
+      "epoch": 149.58579881656806,
+      "grad_norm": 1.1910266876220703,
+      "learning_rate": 0.0001,
+      "loss": 0.0201,
+      "step": 25280
+    },
+    {
+      "epoch": 149.6449704142012,
+      "grad_norm": 1.5954008102416992,
+      "learning_rate": 0.0001,
+      "loss": 0.0204,
+      "step": 25290
+    },
+    {
+      "epoch": 149.7041420118343,
+      "grad_norm": 1.8754186630249023,
+      "learning_rate": 0.0001,
+      "loss": 0.0215,
+      "step": 25300
+    },
+    {
+      "epoch": 149.76331360946745,
+      "grad_norm": 1.6102299690246582,
+      "learning_rate": 0.0001,
+      "loss": 0.0213,
+      "step": 25310
+    },
+    {
+      "epoch": 149.82248520710058,
+      "grad_norm": 1.325474739074707,
+      "learning_rate": 0.0001,
+      "loss": 0.0208,
+      "step": 25320
+    },
+    {
+      "epoch": 149.88165680473372,
+      "grad_norm": 1.4368053674697876,
+      "learning_rate": 0.0001,
+      "loss": 0.0207,
+      "step": 25330
+    },
+    {
+      "epoch": 149.94082840236686,
+      "grad_norm": 1.6274086236953735,
+      "learning_rate": 0.0001,
+      "loss": 0.0209,
+      "step": 25340
+    },
+    {
+      "epoch": 150.0,
+      "grad_norm": 1.0471465587615967,
+      "learning_rate": 0.0001,
+      "loss": 0.0213,
+      "step": 25350
+    },
+    {
+      "epoch": 150.05917159763314,
+      "grad_norm": 0.9887391924858093,
+      "learning_rate": 0.0001,
+      "loss": 0.0204,
+      "step": 25360
+    },
+    {
+      "epoch": 150.11834319526628,
+      "grad_norm": 1.1600033044815063,
+      "learning_rate": 0.0001,
+      "loss": 0.0218,
+      "step": 25370
+    },
+    {
+      "epoch": 150.17751479289942,
+      "grad_norm": 1.4537696838378906,
+      "learning_rate": 0.0001,
+      "loss": 0.0218,
+      "step": 25380
+    },
+    {
+      "epoch": 150.23668639053255,
+      "grad_norm": 1.3565564155578613,
+      "learning_rate": 0.0001,
+      "loss": 0.0211,
+      "step": 25390
+    },
+    {
+      "epoch": 150.2958579881657,
+      "grad_norm": 1.3504096269607544,
+      "learning_rate": 0.0001,
+      "loss": 0.0216,
+      "step": 25400
+    },
+    {
+      "epoch": 150.3550295857988,
+      "grad_norm": 2.078139305114746,
+      "learning_rate": 0.0001,
+      "loss": 0.0215,
+      "step": 25410
+    },
+    {
+      "epoch": 150.41420118343194,
+      "grad_norm": 1.60502028465271,
+      "learning_rate": 0.0001,
+      "loss": 0.0218,
+      "step": 25420
+    },
+    {
+      "epoch": 150.47337278106508,
+      "grad_norm": 1.3359532356262207,
+      "learning_rate": 0.0001,
+      "loss": 0.0209,
+      "step": 25430
+    },
+    {
+      "epoch": 150.53254437869822,
+      "grad_norm": 1.5543134212493896,
+      "learning_rate": 0.0001,
+      "loss": 0.0211,
+      "step": 25440
+    },
+    {
+      "epoch": 150.59171597633136,
+      "grad_norm": 1.7949843406677246,
+      "learning_rate": 0.0001,
+      "loss": 0.0218,
+      "step": 25450
+    },
+    {
+      "epoch": 150.6508875739645,
+      "grad_norm": 1.4030319452285767,
+      "learning_rate": 0.0001,
+      "loss": 0.0222,
+      "step": 25460
+    },
+    {
+      "epoch": 150.71005917159763,
+      "grad_norm": 2.6521875858306885,
+      "learning_rate": 0.0001,
+      "loss": 0.0214,
+      "step": 25470
+    },
+    {
+      "epoch": 150.76923076923077,
+      "grad_norm": 2.5452377796173096,
+      "learning_rate": 0.0001,
+      "loss": 0.0227,
+      "step": 25480
+    },
+    {
+      "epoch": 150.8284023668639,
+      "grad_norm": 2.113701581954956,
+      "learning_rate": 0.0001,
+      "loss": 0.0218,
+      "step": 25490
+    },
+    {
+      "epoch": 150.88757396449705,
+      "grad_norm": 1.4770405292510986,
+      "learning_rate": 0.0001,
+      "loss": 0.0216,
+      "step": 25500
+    },
+    {
+      "epoch": 150.9467455621302,
+      "grad_norm": 1.2545053958892822,
+      "learning_rate": 0.0001,
+      "loss": 0.0208,
+      "step": 25510
+    },
+    {
+      "epoch": 151.0059171597633,
+      "grad_norm": 1.341094970703125,
+      "learning_rate": 0.0001,
+      "loss": 0.0208,
+      "step": 25520
+    },
+    {
+      "epoch": 151.06508875739644,
+      "grad_norm": 1.4963527917861938,
+      "learning_rate": 0.0001,
+      "loss": 0.0209,
+      "step": 25530
+    },
+    {
+      "epoch": 151.12426035502958,
+      "grad_norm": 1.3286073207855225,
+      "learning_rate": 0.0001,
+      "loss": 0.021,
+      "step": 25540
+    },
+    {
+      "epoch": 151.18343195266272,
+      "grad_norm": 1.729671835899353,
+      "learning_rate": 0.0001,
+      "loss": 0.0202,
+      "step": 25550
+    },
+    {
+      "epoch": 151.24260355029585,
+      "grad_norm": 1.242275595664978,
+      "learning_rate": 0.0001,
+      "loss": 0.0203,
+      "step": 25560
+    },
+    {
+      "epoch": 151.301775147929,
+      "grad_norm": 1.6160680055618286,
+      "learning_rate": 0.0001,
+      "loss": 0.0206,
+      "step": 25570
+    },
+    {
+      "epoch": 151.36094674556213,
+      "grad_norm": 1.0544663667678833,
+      "learning_rate": 0.0001,
+      "loss": 0.0211,
+      "step": 25580
+    },
+    {
+      "epoch": 151.42011834319527,
+      "grad_norm": 1.151103138923645,
+      "learning_rate": 0.0001,
+      "loss": 0.0223,
+      "step": 25590
+    },
+    {
+      "epoch": 151.4792899408284,
+      "grad_norm": 1.7656123638153076,
+      "learning_rate": 0.0001,
+      "loss": 0.0207,
+      "step": 25600
+    },
+    {
+      "epoch": 151.53846153846155,
+      "grad_norm": 1.5466721057891846,
+      "learning_rate": 0.0001,
+      "loss": 0.0215,
+      "step": 25610
+    },
+    {
+      "epoch": 151.5976331360947,
+      "grad_norm": 1.741058588027954,
+      "learning_rate": 0.0001,
+      "loss": 0.0222,
+      "step": 25620
+    },
+    {
+      "epoch": 151.65680473372782,
+      "grad_norm": 1.116859793663025,
+      "learning_rate": 0.0001,
+      "loss": 0.0202,
+      "step": 25630
+    },
+    {
+      "epoch": 151.71597633136093,
+      "grad_norm": 1.7700176239013672,
+      "learning_rate": 0.0001,
+      "loss": 0.0221,
+      "step": 25640
+    },
+    {
+      "epoch": 151.77514792899407,
+      "grad_norm": 1.5682357549667358,
+      "learning_rate": 0.0001,
+      "loss": 0.0212,
+      "step": 25650
+    },
+    {
+      "epoch": 151.8343195266272,
+      "grad_norm": 2.0216481685638428,
+      "learning_rate": 0.0001,
+      "loss": 0.0216,
+      "step": 25660
+    },
+    {
+      "epoch": 151.89349112426035,
+      "grad_norm": 1.5635340213775635,
+      "learning_rate": 0.0001,
+      "loss": 0.0208,
+      "step": 25670
+    },
+    {
+      "epoch": 151.9526627218935,
+      "grad_norm": 1.8286322355270386,
+      "learning_rate": 0.0001,
+      "loss": 0.0211,
+      "step": 25680
+    },
+    {
+      "epoch": 152.01183431952663,
+      "grad_norm": 1.25504732131958,
+      "learning_rate": 0.0001,
+      "loss": 0.0211,
+      "step": 25690
+    },
+    {
+      "epoch": 152.07100591715977,
+      "grad_norm": 1.6869962215423584,
+      "learning_rate": 0.0001,
+      "loss": 0.0213,
+      "step": 25700
+    },
+    {
+      "epoch": 152.1301775147929,
+      "grad_norm": 1.4163267612457275,
+      "learning_rate": 0.0001,
+      "loss": 0.0233,
+      "step": 25710
+    },
+    {
+      "epoch": 152.18934911242604,
+      "grad_norm": 1.7080047130584717,
+      "learning_rate": 0.0001,
+      "loss": 0.0227,
+      "step": 25720
+    },
+    {
+      "epoch": 152.24852071005918,
+      "grad_norm": 1.5715261697769165,
+      "learning_rate": 0.0001,
+      "loss": 0.0219,
+      "step": 25730
+    },
+    {
+      "epoch": 152.30769230769232,
+      "grad_norm": 2.0011422634124756,
+      "learning_rate": 0.0001,
+      "loss": 0.0218,
+      "step": 25740
+    },
+    {
+      "epoch": 152.36686390532543,
+      "grad_norm": 1.8720799684524536,
+      "learning_rate": 0.0001,
+      "loss": 0.0209,
+      "step": 25750
+    },
+    {
+      "epoch": 152.42603550295857,
+      "grad_norm": 1.1511582136154175,
+      "learning_rate": 0.0001,
+      "loss": 0.021,
+      "step": 25760
+    },
+    {
+      "epoch": 152.4852071005917,
+      "grad_norm": 1.9970887899398804,
+      "learning_rate": 0.0001,
+      "loss": 0.0207,
+      "step": 25770
+    },
+    {
+      "epoch": 152.54437869822485,
+      "grad_norm": 1.0882607698440552,
+      "learning_rate": 0.0001,
+      "loss": 0.0216,
+      "step": 25780
+    },
+    {
+      "epoch": 152.60355029585799,
+      "grad_norm": 1.4984229803085327,
+      "learning_rate": 0.0001,
+      "loss": 0.0228,
+      "step": 25790
+    },
+    {
+      "epoch": 152.66272189349112,
+      "grad_norm": 1.712168574333191,
+      "learning_rate": 0.0001,
+      "loss": 0.0229,
+      "step": 25800
+    },
+    {
+      "epoch": 152.72189349112426,
+      "grad_norm": 1.172399878501892,
+      "learning_rate": 0.0001,
+      "loss": 0.022,
+      "step": 25810
+    },
+    {
+      "epoch": 152.7810650887574,
+      "grad_norm": 1.064012885093689,
+      "learning_rate": 0.0001,
+      "loss": 0.0224,
+      "step": 25820
+    },
+    {
+      "epoch": 152.84023668639054,
+      "grad_norm": 1.487667202949524,
+      "learning_rate": 0.0001,
+      "loss": 0.0227,
+      "step": 25830
+    },
+    {
+      "epoch": 152.89940828402368,
+      "grad_norm": 1.2239387035369873,
+      "learning_rate": 0.0001,
+      "loss": 0.022,
+      "step": 25840
+    },
+    {
+      "epoch": 152.95857988165682,
+      "grad_norm": 1.015724778175354,
+      "learning_rate": 0.0001,
+      "loss": 0.0224,
+      "step": 25850
+    },
+    {
+      "epoch": 153.01775147928993,
+      "grad_norm": 1.7180930376052856,
+      "learning_rate": 0.0001,
+      "loss": 0.0219,
+      "step": 25860
+    },
+    {
+      "epoch": 153.07692307692307,
+      "grad_norm": 1.6737507581710815,
+      "learning_rate": 0.0001,
+      "loss": 0.0218,
+      "step": 25870
+    },
+    {
+      "epoch": 153.1360946745562,
+      "grad_norm": 1.4114919900894165,
+      "learning_rate": 0.0001,
+      "loss": 0.0222,
+      "step": 25880
+    },
+    {
+      "epoch": 153.19526627218934,
+      "grad_norm": 1.52801513671875,
+      "learning_rate": 0.0001,
+      "loss": 0.0226,
+      "step": 25890
+    },
+    {
+      "epoch": 153.25443786982248,
+      "grad_norm": 1.3589574098587036,
+      "learning_rate": 0.0001,
+      "loss": 0.0219,
+      "step": 25900
+    },
+    {
+      "epoch": 153.31360946745562,
+      "grad_norm": 1.4642192125320435,
+      "learning_rate": 0.0001,
+      "loss": 0.022,
+      "step": 25910
+    },
+    {
+      "epoch": 153.37278106508876,
+      "grad_norm": 1.9014003276824951,
+      "learning_rate": 0.0001,
+      "loss": 0.0214,
+      "step": 25920
+    },
+    {
+      "epoch": 153.4319526627219,
+      "grad_norm": 1.5836914777755737,
+      "learning_rate": 0.0001,
+      "loss": 0.0213,
+      "step": 25930
+    },
+    {
+      "epoch": 153.49112426035504,
+      "grad_norm": 1.092085599899292,
+      "learning_rate": 0.0001,
+      "loss": 0.0212,
+      "step": 25940
+    },
+    {
+      "epoch": 153.55029585798817,
+      "grad_norm": 1.2248846292495728,
+      "learning_rate": 0.0001,
+      "loss": 0.021,
+      "step": 25950
+    },
+    {
+      "epoch": 153.6094674556213,
+      "grad_norm": 0.8705076575279236,
+      "learning_rate": 0.0001,
+      "loss": 0.0209,
+      "step": 25960
+    },
+    {
+      "epoch": 153.66863905325442,
+      "grad_norm": 1.1099029779434204,
+      "learning_rate": 0.0001,
+      "loss": 0.0211,
+      "step": 25970
+    },
+    {
+      "epoch": 153.72781065088756,
+      "grad_norm": 1.5218979120254517,
+      "learning_rate": 0.0001,
+      "loss": 0.0214,
+      "step": 25980
+    },
+    {
+      "epoch": 153.7869822485207,
+      "grad_norm": 1.447679877281189,
+      "learning_rate": 0.0001,
+      "loss": 0.0221,
+      "step": 25990
+    },
+    {
+      "epoch": 153.84615384615384,
+      "grad_norm": 1.5342786312103271,
+      "learning_rate": 0.0001,
+      "loss": 0.021,
+      "step": 26000
+    },
+    {
+      "epoch": 153.90532544378698,
+      "grad_norm": 2.100250482559204,
+      "learning_rate": 0.0001,
+      "loss": 0.0215,
+      "step": 26010
+    },
+    {
+      "epoch": 153.96449704142012,
+      "grad_norm": 1.4730910062789917,
+      "learning_rate": 0.0001,
+      "loss": 0.0214,
+      "step": 26020
+    },
+    {
+      "epoch": 154.02366863905326,
+      "grad_norm": 1.2550325393676758,
+      "learning_rate": 0.0001,
+      "loss": 0.0216,
+      "step": 26030
+    },
+    {
+      "epoch": 154.0828402366864,
+      "grad_norm": 1.4883971214294434,
+      "learning_rate": 0.0001,
+      "loss": 0.0208,
+      "step": 26040
+    },
+    {
+      "epoch": 154.14201183431953,
+      "grad_norm": 1.4996434450149536,
+      "learning_rate": 0.0001,
+      "loss": 0.0205,
+      "step": 26050
+    },
+    {
+      "epoch": 154.20118343195267,
+      "grad_norm": 1.385128140449524,
+      "learning_rate": 0.0001,
+      "loss": 0.0208,
+      "step": 26060
+    },
+    {
+      "epoch": 154.2603550295858,
+      "grad_norm": 1.6819508075714111,
+      "learning_rate": 0.0001,
+      "loss": 0.0206,
+      "step": 26070
+    },
+    {
+      "epoch": 154.31952662721895,
+      "grad_norm": 1.4793570041656494,
+      "learning_rate": 0.0001,
+      "loss": 0.0207,
+      "step": 26080
+    },
+    {
+      "epoch": 154.37869822485206,
+      "grad_norm": 1.4416894912719727,
+      "learning_rate": 0.0001,
+      "loss": 0.0214,
+      "step": 26090
+    },
+    {
+      "epoch": 154.4378698224852,
+      "grad_norm": 1.96162748336792,
+      "learning_rate": 0.0001,
+      "loss": 0.021,
+      "step": 26100
+    },
+    {
+      "epoch": 154.49704142011834,
+      "grad_norm": 1.7055914402008057,
+      "learning_rate": 0.0001,
+      "loss": 0.0205,
+      "step": 26110
+    },
+    {
+      "epoch": 154.55621301775147,
+      "grad_norm": 1.4808541536331177,
+      "learning_rate": 0.0001,
+      "loss": 0.0205,
+      "step": 26120
+    },
+    {
+      "epoch": 154.6153846153846,
+      "grad_norm": 1.5020105838775635,
+      "learning_rate": 0.0001,
+      "loss": 0.0201,
+      "step": 26130
+    },
+    {
+      "epoch": 154.67455621301775,
+      "grad_norm": 2.016833543777466,
+      "learning_rate": 0.0001,
+      "loss": 0.0205,
+      "step": 26140
+    },
+    {
+      "epoch": 154.7337278106509,
+      "grad_norm": 1.9550371170043945,
+      "learning_rate": 0.0001,
+      "loss": 0.0209,
+      "step": 26150
+    },
+    {
+      "epoch": 154.79289940828403,
+      "grad_norm": 1.3006995916366577,
+      "learning_rate": 0.0001,
+      "loss": 0.021,
+      "step": 26160
+    },
+    {
+      "epoch": 154.85207100591717,
+      "grad_norm": 1.5109091997146606,
+      "learning_rate": 0.0001,
+      "loss": 0.0206,
+      "step": 26170
+    },
+    {
+      "epoch": 154.9112426035503,
+      "grad_norm": 1.4291115999221802,
+      "learning_rate": 0.0001,
+      "loss": 0.0201,
+      "step": 26180
+    },
+    {
+      "epoch": 154.97041420118344,
+      "grad_norm": 1.2600953578948975,
+      "learning_rate": 0.0001,
+      "loss": 0.021,
+      "step": 26190
+    },
+    {
+      "epoch": 155.02958579881656,
+      "grad_norm": 1.65549635887146,
+      "learning_rate": 0.0001,
+      "loss": 0.021,
+      "step": 26200
+    },
+    {
+      "epoch": 155.0887573964497,
+      "grad_norm": 1.3921526670455933,
+      "learning_rate": 0.0001,
+      "loss": 0.021,
+      "step": 26210
+    },
+    {
+      "epoch": 155.14792899408283,
+      "grad_norm": 1.0890072584152222,
+      "learning_rate": 0.0001,
+      "loss": 0.02,
+      "step": 26220
+    },
+    {
+      "epoch": 155.20710059171597,
+      "grad_norm": 1.1429088115692139,
+      "learning_rate": 0.0001,
+      "loss": 0.0201,
+      "step": 26230
+    },
+    {
+      "epoch": 155.2662721893491,
+      "grad_norm": 1.3489137887954712,
+      "learning_rate": 0.0001,
+      "loss": 0.0196,
+      "step": 26240
+    },
+    {
+      "epoch": 155.32544378698225,
+      "grad_norm": 1.41928231716156,
+      "learning_rate": 0.0001,
+      "loss": 0.02,
+      "step": 26250
+    },
+    {
+      "epoch": 155.3846153846154,
+      "grad_norm": 1.2024985551834106,
+      "learning_rate": 0.0001,
+      "loss": 0.0207,
+      "step": 26260
+    },
+    {
+      "epoch": 155.44378698224853,
+      "grad_norm": 1.3420028686523438,
+      "learning_rate": 0.0001,
+      "loss": 0.0204,
+      "step": 26270
+    },
+    {
+      "epoch": 155.50295857988166,
+      "grad_norm": 1.2282062768936157,
+      "learning_rate": 0.0001,
+      "loss": 0.0198,
+      "step": 26280
+    },
+    {
+      "epoch": 155.5621301775148,
+      "grad_norm": 1.6317216157913208,
+      "learning_rate": 0.0001,
+      "loss": 0.0207,
+      "step": 26290
+    },
+    {
+      "epoch": 155.62130177514794,
+      "grad_norm": 1.862438440322876,
+      "learning_rate": 0.0001,
+      "loss": 0.021,
+      "step": 26300
+    },
+    {
+      "epoch": 155.68047337278105,
+      "grad_norm": 2.055371046066284,
+      "learning_rate": 0.0001,
+      "loss": 0.0213,
+      "step": 26310
+    },
+    {
+      "epoch": 155.7396449704142,
+      "grad_norm": 1.564012050628662,
+      "learning_rate": 0.0001,
+      "loss": 0.0205,
+      "step": 26320
+    },
+    {
+      "epoch": 155.79881656804733,
+      "grad_norm": 1.2697653770446777,
+      "learning_rate": 0.0001,
+      "loss": 0.0208,
+      "step": 26330
+    },
+    {
+      "epoch": 155.85798816568047,
+      "grad_norm": 1.5213861465454102,
+      "learning_rate": 0.0001,
+      "loss": 0.0214,
+      "step": 26340
+    },
+    {
+      "epoch": 155.9171597633136,
+      "grad_norm": 1.6746435165405273,
+      "learning_rate": 0.0001,
+      "loss": 0.0208,
+      "step": 26350
+    },
+    {
+      "epoch": 155.97633136094674,
+      "grad_norm": 1.078543782234192,
+      "learning_rate": 0.0001,
+      "loss": 0.021,
+      "step": 26360
+    },
+    {
+      "epoch": 156.03550295857988,
+      "grad_norm": 2.201373338699341,
+      "learning_rate": 0.0001,
+      "loss": 0.0229,
+      "step": 26370
+    },
+    {
+      "epoch": 156.09467455621302,
+      "grad_norm": 1.699474573135376,
+      "learning_rate": 0.0001,
+      "loss": 0.0213,
+      "step": 26380
+    },
+    {
+      "epoch": 156.15384615384616,
+      "grad_norm": 1.9506713151931763,
+      "learning_rate": 0.0001,
+      "loss": 0.0221,
+      "step": 26390
+    },
+    {
+      "epoch": 156.2130177514793,
+      "grad_norm": 1.9464579820632935,
+      "learning_rate": 0.0001,
+      "loss": 0.0206,
+      "step": 26400
+    },
+    {
+      "epoch": 156.27218934911244,
+      "grad_norm": 1.676459789276123,
+      "learning_rate": 0.0001,
+      "loss": 0.021,
+      "step": 26410
+    },
+    {
+      "epoch": 156.33136094674558,
+      "grad_norm": 2.439486026763916,
+      "learning_rate": 0.0001,
+      "loss": 0.0198,
+      "step": 26420
+    },
+    {
+      "epoch": 156.3905325443787,
+      "grad_norm": 2.1242942810058594,
+      "learning_rate": 0.0001,
+      "loss": 0.0206,
+      "step": 26430
+    },
+    {
+      "epoch": 156.44970414201183,
+      "grad_norm": 1.802720546722412,
+      "learning_rate": 0.0001,
+      "loss": 0.0206,
+      "step": 26440
+    },
+    {
+      "epoch": 156.50887573964496,
+      "grad_norm": 1.4209725856781006,
+      "learning_rate": 0.0001,
+      "loss": 0.02,
+      "step": 26450
+    },
+    {
+      "epoch": 156.5680473372781,
+      "grad_norm": 1.9678995609283447,
+      "learning_rate": 0.0001,
+      "loss": 0.0198,
+      "step": 26460
+    },
+    {
+      "epoch": 156.62721893491124,
+      "grad_norm": 1.741742491722107,
+      "learning_rate": 0.0001,
+      "loss": 0.0192,
+      "step": 26470
+    },
+    {
+      "epoch": 156.68639053254438,
+      "grad_norm": 1.5673197507858276,
+      "learning_rate": 0.0001,
+      "loss": 0.0195,
+      "step": 26480
+    },
+    {
+      "epoch": 156.74556213017752,
+      "grad_norm": 1.4484319686889648,
+      "learning_rate": 0.0001,
+      "loss": 0.0199,
+      "step": 26490
+    },
+    {
+      "epoch": 156.80473372781066,
+      "grad_norm": 1.6195127964019775,
+      "learning_rate": 0.0001,
+      "loss": 0.0193,
+      "step": 26500
+    },
+    {
+      "epoch": 156.8639053254438,
+      "grad_norm": 1.3534082174301147,
+      "learning_rate": 0.0001,
+      "loss": 0.0196,
+      "step": 26510
+    },
+    {
+      "epoch": 156.92307692307693,
+      "grad_norm": 1.1817723512649536,
+      "learning_rate": 0.0001,
+      "loss": 0.0198,
+      "step": 26520
+    },
+    {
+      "epoch": 156.98224852071007,
+      "grad_norm": 1.522890329360962,
+      "learning_rate": 0.0001,
+      "loss": 0.0209,
+      "step": 26530
+    },
+    {
+      "epoch": 157.04142011834318,
+      "grad_norm": 1.787109375,
+      "learning_rate": 0.0001,
+      "loss": 0.0209,
+      "step": 26540
+    },
+    {
+      "epoch": 157.10059171597632,
+      "grad_norm": 1.3914752006530762,
+      "learning_rate": 0.0001,
+      "loss": 0.0207,
+      "step": 26550
+    },
+    {
+      "epoch": 157.15976331360946,
+      "grad_norm": 1.3127388954162598,
+      "learning_rate": 0.0001,
+      "loss": 0.0205,
+      "step": 26560
+    },
+    {
+      "epoch": 157.2189349112426,
+      "grad_norm": 1.3728588819503784,
+      "learning_rate": 0.0001,
+      "loss": 0.0205,
+      "step": 26570
+    },
+    {
+      "epoch": 157.27810650887574,
+      "grad_norm": 1.6442830562591553,
+      "learning_rate": 0.0001,
+      "loss": 0.0208,
+      "step": 26580
+    },
+    {
+      "epoch": 157.33727810650888,
+      "grad_norm": 1.4659403562545776,
+      "learning_rate": 0.0001,
+      "loss": 0.0206,
+      "step": 26590
+    },
+    {
+      "epoch": 157.39644970414201,
+      "grad_norm": 1.404071569442749,
+      "learning_rate": 0.0001,
+      "loss": 0.0204,
+      "step": 26600
+    },
+    {
+      "epoch": 157.45562130177515,
+      "grad_norm": 1.196016788482666,
+      "learning_rate": 0.0001,
+      "loss": 0.0203,
+      "step": 26610
+    },
+    {
+      "epoch": 157.5147928994083,
+      "grad_norm": 1.1300283670425415,
+      "learning_rate": 0.0001,
+      "loss": 0.0212,
+      "step": 26620
+    },
+    {
+      "epoch": 157.57396449704143,
+      "grad_norm": 1.113511562347412,
+      "learning_rate": 0.0001,
+      "loss": 0.0209,
+      "step": 26630
+    },
+    {
+      "epoch": 157.63313609467457,
+      "grad_norm": 1.2861113548278809,
+      "learning_rate": 0.0001,
+      "loss": 0.0211,
+      "step": 26640
+    },
+    {
+      "epoch": 157.69230769230768,
+      "grad_norm": 1.501885175704956,
+      "learning_rate": 0.0001,
+      "loss": 0.0216,
+      "step": 26650
+    },
+    {
+      "epoch": 157.75147928994082,
+      "grad_norm": 1.8719037771224976,
+      "learning_rate": 0.0001,
+      "loss": 0.021,
+      "step": 26660
+    },
+    {
+      "epoch": 157.81065088757396,
+      "grad_norm": 1.2902911901474,
+      "learning_rate": 0.0001,
+      "loss": 0.0207,
+      "step": 26670
+    },
+    {
+      "epoch": 157.8698224852071,
+      "grad_norm": 1.1591600179672241,
+      "learning_rate": 0.0001,
+      "loss": 0.0212,
+      "step": 26680
+    },
+    {
+      "epoch": 157.92899408284023,
+      "grad_norm": 1.1679754257202148,
+      "learning_rate": 0.0001,
+      "loss": 0.0214,
+      "step": 26690
+    },
+    {
+      "epoch": 157.98816568047337,
+      "grad_norm": 1.8898550271987915,
+      "learning_rate": 0.0001,
+      "loss": 0.0213,
+      "step": 26700
+    },
+    {
+      "epoch": 158.0473372781065,
+      "grad_norm": 1.5150840282440186,
+      "learning_rate": 0.0001,
+      "loss": 0.021,
+      "step": 26710
+    },
+    {
+      "epoch": 158.10650887573965,
+      "grad_norm": 1.6401528120040894,
+      "learning_rate": 0.0001,
+      "loss": 0.0204,
+      "step": 26720
+    },
+    {
+      "epoch": 158.1656804733728,
+      "grad_norm": 1.2955777645111084,
+      "learning_rate": 0.0001,
+      "loss": 0.0208,
+      "step": 26730
+    },
+    {
+      "epoch": 158.22485207100593,
+      "grad_norm": 1.658198356628418,
+      "learning_rate": 0.0001,
+      "loss": 0.0212,
+      "step": 26740
+    },
+    {
+      "epoch": 158.28402366863907,
+      "grad_norm": 1.905434012413025,
+      "learning_rate": 0.0001,
+      "loss": 0.0203,
+      "step": 26750
+    },
+    {
+      "epoch": 158.34319526627218,
+      "grad_norm": 1.6666035652160645,
+      "learning_rate": 0.0001,
+      "loss": 0.0213,
+      "step": 26760
+    },
+    {
+      "epoch": 158.4023668639053,
+      "grad_norm": 1.1749194860458374,
+      "learning_rate": 0.0001,
+      "loss": 0.0211,
+      "step": 26770
+    },
+    {
+      "epoch": 158.46153846153845,
+      "grad_norm": 1.2984646558761597,
+      "learning_rate": 0.0001,
+      "loss": 0.0212,
+      "step": 26780
+    },
+    {
+      "epoch": 158.5207100591716,
+      "grad_norm": 1.0474600791931152,
+      "learning_rate": 0.0001,
+      "loss": 0.0202,
+      "step": 26790
+    },
+    {
+      "epoch": 158.57988165680473,
+      "grad_norm": 1.2343430519104004,
+      "learning_rate": 0.0001,
+      "loss": 0.0206,
+      "step": 26800
+    },
+    {
+      "epoch": 158.63905325443787,
+      "grad_norm": 1.2120416164398193,
+      "learning_rate": 0.0001,
+      "loss": 0.0209,
+      "step": 26810
+    },
+    {
+      "epoch": 158.698224852071,
+      "grad_norm": 1.4323725700378418,
+      "learning_rate": 0.0001,
+      "loss": 0.0211,
+      "step": 26820
+    },
+    {
+      "epoch": 158.75739644970415,
+      "grad_norm": 2.242377758026123,
+      "learning_rate": 0.0001,
+      "loss": 0.0217,
+      "step": 26830
+    },
+    {
+      "epoch": 158.81656804733728,
+      "grad_norm": 1.4264330863952637,
+      "learning_rate": 0.0001,
+      "loss": 0.0224,
+      "step": 26840
+    },
+    {
+      "epoch": 158.87573964497042,
+      "grad_norm": 1.6665781736373901,
+      "learning_rate": 0.0001,
+      "loss": 0.0211,
+      "step": 26850
+    },
+    {
+      "epoch": 158.93491124260356,
+      "grad_norm": 1.364776611328125,
+      "learning_rate": 0.0001,
+      "loss": 0.0206,
+      "step": 26860
+    },
+    {
+      "epoch": 158.9940828402367,
+      "grad_norm": 1.1067391633987427,
+      "learning_rate": 0.0001,
+      "loss": 0.0211,
+      "step": 26870
+    },
+    {
+      "epoch": 159.0532544378698,
+      "grad_norm": 1.7306609153747559,
+      "learning_rate": 0.0001,
+      "loss": 0.0203,
+      "step": 26880
+    },
+    {
+      "epoch": 159.11242603550295,
+      "grad_norm": 1.3120485544204712,
+      "learning_rate": 0.0001,
+      "loss": 0.0205,
+      "step": 26890
+    },
+    {
+      "epoch": 159.1715976331361,
+      "grad_norm": 1.4558175802230835,
+      "learning_rate": 0.0001,
+      "loss": 0.021,
+      "step": 26900
+    },
+    {
+      "epoch": 159.23076923076923,
+      "grad_norm": 1.2709959745407104,
+      "learning_rate": 0.0001,
+      "loss": 0.0212,
+      "step": 26910
+    },
+    {
+      "epoch": 159.28994082840237,
+      "grad_norm": 1.1686744689941406,
+      "learning_rate": 0.0001,
+      "loss": 0.0211,
+      "step": 26920
+    },
+    {
+      "epoch": 159.3491124260355,
+      "grad_norm": 1.2401418685913086,
+      "learning_rate": 0.0001,
+      "loss": 0.021,
+      "step": 26930
+    },
+    {
+      "epoch": 159.40828402366864,
+      "grad_norm": 1.770949363708496,
+      "learning_rate": 0.0001,
+      "loss": 0.0213,
+      "step": 26940
+    },
+    {
+      "epoch": 159.46745562130178,
+      "grad_norm": 1.180551528930664,
+      "learning_rate": 0.0001,
+      "loss": 0.0212,
+      "step": 26950
+    },
+    {
+      "epoch": 159.52662721893492,
+      "grad_norm": 1.0209028720855713,
+      "learning_rate": 0.0001,
+      "loss": 0.0209,
+      "step": 26960
+    },
+    {
+      "epoch": 159.58579881656806,
+      "grad_norm": 1.9734721183776855,
+      "learning_rate": 0.0001,
+      "loss": 0.0219,
+      "step": 26970
+    },
+    {
+      "epoch": 159.6449704142012,
+      "grad_norm": 1.378026008605957,
+      "learning_rate": 0.0001,
+      "loss": 0.0212,
+      "step": 26980
+    },
+    {
+      "epoch": 159.7041420118343,
+      "grad_norm": 1.9911059141159058,
+      "learning_rate": 0.0001,
+      "loss": 0.0216,
+      "step": 26990
+    },
+    {
+      "epoch": 159.76331360946745,
+      "grad_norm": 1.9442814588546753,
+      "learning_rate": 0.0001,
+      "loss": 0.0216,
+      "step": 27000
+    },
+    {
+      "epoch": 159.82248520710058,
+      "grad_norm": 1.8272459506988525,
+      "learning_rate": 0.0001,
+      "loss": 0.0208,
+      "step": 27010
+    },
+    {
+      "epoch": 159.88165680473372,
+      "grad_norm": 1.8973618745803833,
+      "learning_rate": 0.0001,
+      "loss": 0.0208,
+      "step": 27020
+    },
+    {
+      "epoch": 159.94082840236686,
+      "grad_norm": 1.102816104888916,
+      "learning_rate": 0.0001,
+      "loss": 0.0202,
+      "step": 27030
+    },
+    {
+      "epoch": 160.0,
+      "grad_norm": 1.4078744649887085,
+      "learning_rate": 0.0001,
+      "loss": 0.0205,
+      "step": 27040
+    },
+    {
+      "epoch": 160.05917159763314,
+      "grad_norm": 1.2238883972167969,
+      "learning_rate": 0.0001,
+      "loss": 0.0192,
+      "step": 27050
+    },
+    {
+      "epoch": 160.11834319526628,
+      "grad_norm": 1.3749524354934692,
+      "learning_rate": 0.0001,
+      "loss": 0.02,
+      "step": 27060
+    },
+    {
+      "epoch": 160.17751479289942,
+      "grad_norm": 2.104325294494629,
+      "learning_rate": 0.0001,
+      "loss": 0.0203,
+      "step": 27070
+    },
+    {
+      "epoch": 160.23668639053255,
+      "grad_norm": 1.5650626420974731,
+      "learning_rate": 0.0001,
+      "loss": 0.0205,
+      "step": 27080
+    },
+    {
+      "epoch": 160.2958579881657,
+      "grad_norm": 1.0444995164871216,
+      "learning_rate": 0.0001,
+      "loss": 0.0208,
+      "step": 27090
+    },
+    {
+      "epoch": 160.3550295857988,
+      "grad_norm": 1.1349681615829468,
+      "learning_rate": 0.0001,
+      "loss": 0.021,
+      "step": 27100
+    },
+    {
+      "epoch": 160.41420118343194,
+      "grad_norm": 0.8105968236923218,
+      "learning_rate": 0.0001,
+      "loss": 0.0205,
+      "step": 27110
+    },
+    {
+      "epoch": 160.47337278106508,
+      "grad_norm": 1.3990508317947388,
+      "learning_rate": 0.0001,
+      "loss": 0.0209,
+      "step": 27120
+    },
+    {
+      "epoch": 160.53254437869822,
+      "grad_norm": 1.226505160331726,
+      "learning_rate": 0.0001,
+      "loss": 0.0212,
+      "step": 27130
+    },
+    {
+      "epoch": 160.59171597633136,
+      "grad_norm": 1.6955674886703491,
+      "learning_rate": 0.0001,
+      "loss": 0.0217,
+      "step": 27140
+    },
+    {
+      "epoch": 160.6508875739645,
+      "grad_norm": 1.7311537265777588,
+      "learning_rate": 0.0001,
+      "loss": 0.0216,
+      "step": 27150
+    },
+    {
+      "epoch": 160.71005917159763,
+      "grad_norm": 1.2058486938476562,
+      "learning_rate": 0.0001,
+      "loss": 0.0218,
+      "step": 27160
+    },
+    {
+      "epoch": 160.76923076923077,
+      "grad_norm": 2.4075024127960205,
+      "learning_rate": 0.0001,
+      "loss": 0.0212,
+      "step": 27170
+    },
+    {
+      "epoch": 160.8284023668639,
+      "grad_norm": 2.5634541511535645,
+      "learning_rate": 0.0001,
+      "loss": 0.022,
+      "step": 27180
+    },
+    {
+      "epoch": 160.88757396449705,
+      "grad_norm": 2.1461312770843506,
+      "learning_rate": 0.0001,
+      "loss": 0.0222,
+      "step": 27190
+    },
+    {
+      "epoch": 160.9467455621302,
+      "grad_norm": 2.182626724243164,
+      "learning_rate": 0.0001,
+      "loss": 0.0215,
+      "step": 27200
+    },
+    {
+      "epoch": 161.0059171597633,
+      "grad_norm": 1.8480207920074463,
+      "learning_rate": 0.0001,
+      "loss": 0.0212,
+      "step": 27210
+    },
+    {
+      "epoch": 161.06508875739644,
+      "grad_norm": 1.048386812210083,
+      "learning_rate": 0.0001,
+      "loss": 0.0194,
+      "step": 27220
+    },
+    {
+      "epoch": 161.12426035502958,
+      "grad_norm": 1.7133028507232666,
+      "learning_rate": 0.0001,
+      "loss": 0.0199,
+      "step": 27230
+    },
+    {
+      "epoch": 161.18343195266272,
+      "grad_norm": 1.2933645248413086,
+      "learning_rate": 0.0001,
+      "loss": 0.0198,
+      "step": 27240
+    },
+    {
+      "epoch": 161.24260355029585,
+      "grad_norm": 1.5108689069747925,
+      "learning_rate": 0.0001,
+      "loss": 0.0206,
+      "step": 27250
+    },
+    {
+      "epoch": 161.301775147929,
+      "grad_norm": 1.7369829416275024,
+      "learning_rate": 0.0001,
+      "loss": 0.021,
+      "step": 27260
+    },
+    {
+      "epoch": 161.36094674556213,
+      "grad_norm": 1.5660901069641113,
+      "learning_rate": 0.0001,
+      "loss": 0.0203,
+      "step": 27270
+    },
+    {
+      "epoch": 161.42011834319527,
+      "grad_norm": 1.5899075269699097,
+      "learning_rate": 0.0001,
+      "loss": 0.0211,
+      "step": 27280
+    },
+    {
+      "epoch": 161.4792899408284,
+      "grad_norm": 1.7049801349639893,
+      "learning_rate": 0.0001,
+      "loss": 0.0214,
+      "step": 27290
+    },
+    {
+      "epoch": 161.53846153846155,
+      "grad_norm": 1.536257266998291,
+      "learning_rate": 0.0001,
+      "loss": 0.0209,
+      "step": 27300
+    },
+    {
+      "epoch": 161.5976331360947,
+      "grad_norm": 1.2624473571777344,
+      "learning_rate": 0.0001,
+      "loss": 0.0203,
+      "step": 27310
+    },
+    {
+      "epoch": 161.65680473372782,
+      "grad_norm": 1.1442803144454956,
+      "learning_rate": 0.0001,
+      "loss": 0.0203,
+      "step": 27320
+    },
+    {
+      "epoch": 161.71597633136093,
+      "grad_norm": 1.2201980352401733,
+      "learning_rate": 0.0001,
+      "loss": 0.0208,
+      "step": 27330
+    },
+    {
+      "epoch": 161.77514792899407,
+      "grad_norm": 1.0795060396194458,
+      "learning_rate": 0.0001,
+      "loss": 0.0204,
+      "step": 27340
+    },
+    {
+      "epoch": 161.8343195266272,
+      "grad_norm": 1.709027886390686,
+      "learning_rate": 0.0001,
+      "loss": 0.022,
+      "step": 27350
+    },
+    {
+      "epoch": 161.89349112426035,
+      "grad_norm": 1.2025103569030762,
+      "learning_rate": 0.0001,
+      "loss": 0.0218,
+      "step": 27360
+    },
+    {
+      "epoch": 161.9526627218935,
+      "grad_norm": 1.0459328889846802,
+      "learning_rate": 0.0001,
+      "loss": 0.0219,
+      "step": 27370
+    },
+    {
+      "epoch": 162.01183431952663,
+      "grad_norm": 1.9738409519195557,
+      "learning_rate": 0.0001,
+      "loss": 0.0218,
+      "step": 27380
+    },
+    {
+      "epoch": 162.07100591715977,
+      "grad_norm": 1.4781568050384521,
+      "learning_rate": 0.0001,
+      "loss": 0.0208,
+      "step": 27390
+    },
+    {
+      "epoch": 162.1301775147929,
+      "grad_norm": 0.9911723732948303,
+      "learning_rate": 0.0001,
+      "loss": 0.0217,
+      "step": 27400
+    },
+    {
+      "epoch": 162.18934911242604,
+      "grad_norm": 1.1431188583374023,
+      "learning_rate": 0.0001,
+      "loss": 0.0224,
+      "step": 27410
+    },
+    {
+      "epoch": 162.24852071005918,
+      "grad_norm": 1.4991638660430908,
+      "learning_rate": 0.0001,
+      "loss": 0.0222,
+      "step": 27420
+    },
+    {
+      "epoch": 162.30769230769232,
+      "grad_norm": 1.4117403030395508,
+      "learning_rate": 0.0001,
+      "loss": 0.0218,
+      "step": 27430
+    },
+    {
+      "epoch": 162.36686390532543,
+      "grad_norm": 0.9494815468788147,
+      "learning_rate": 0.0001,
+      "loss": 0.0209,
+      "step": 27440
+    },
+    {
+      "epoch": 162.42603550295857,
+      "grad_norm": 1.3298752307891846,
+      "learning_rate": 0.0001,
+      "loss": 0.0205,
+      "step": 27450
+    },
+    {
+      "epoch": 162.4852071005917,
+      "grad_norm": 1.180720329284668,
+      "learning_rate": 0.0001,
+      "loss": 0.0206,
+      "step": 27460
+    },
+    {
+      "epoch": 162.54437869822485,
+      "grad_norm": 1.5091681480407715,
+      "learning_rate": 0.0001,
+      "loss": 0.0204,
+      "step": 27470
+    },
+    {
+      "epoch": 162.60355029585799,
+      "grad_norm": 1.3787050247192383,
+      "learning_rate": 0.0001,
+      "loss": 0.0215,
+      "step": 27480
+    },
+    {
+      "epoch": 162.66272189349112,
+      "grad_norm": 1.1194329261779785,
+      "learning_rate": 0.0001,
+      "loss": 0.0208,
+      "step": 27490
+    },
+    {
+      "epoch": 162.72189349112426,
+      "grad_norm": 1.220835566520691,
+      "learning_rate": 0.0001,
+      "loss": 0.0202,
+      "step": 27500
+    },
+    {
+      "epoch": 162.7810650887574,
+      "grad_norm": 1.5954328775405884,
+      "learning_rate": 0.0001,
+      "loss": 0.0212,
+      "step": 27510
+    },
+    {
+      "epoch": 162.84023668639054,
+      "grad_norm": 1.031205654144287,
+      "learning_rate": 0.0001,
+      "loss": 0.0205,
+      "step": 27520
+    },
+    {
+      "epoch": 162.89940828402368,
+      "grad_norm": 1.498602271080017,
+      "learning_rate": 0.0001,
+      "loss": 0.0205,
+      "step": 27530
+    },
+    {
+      "epoch": 162.95857988165682,
+      "grad_norm": 1.4933857917785645,
+      "learning_rate": 0.0001,
+      "loss": 0.0205,
+      "step": 27540
+    },
+    {
+      "epoch": 163.01775147928993,
+      "grad_norm": 1.5343776941299438,
+      "learning_rate": 0.0001,
+      "loss": 0.0208,
+      "step": 27550
+    },
+    {
+      "epoch": 163.07692307692307,
+      "grad_norm": 1.2958348989486694,
+      "learning_rate": 0.0001,
+      "loss": 0.0209,
+      "step": 27560
+    },
+    {
+      "epoch": 163.1360946745562,
+      "grad_norm": 1.1238911151885986,
+      "learning_rate": 0.0001,
+      "loss": 0.021,
+      "step": 27570
+    },
+    {
+      "epoch": 163.19526627218934,
+      "grad_norm": 1.4696681499481201,
+      "learning_rate": 0.0001,
+      "loss": 0.0209,
+      "step": 27580
+    },
+    {
+      "epoch": 163.25443786982248,
+      "grad_norm": 1.2284115552902222,
+      "learning_rate": 0.0001,
+      "loss": 0.0209,
+      "step": 27590
+    },
+    {
+      "epoch": 163.31360946745562,
+      "grad_norm": 1.4012279510498047,
+      "learning_rate": 0.0001,
+      "loss": 0.0205,
+      "step": 27600
+    },
+    {
+      "epoch": 163.37278106508876,
+      "grad_norm": 1.7274731397628784,
+      "learning_rate": 0.0001,
+      "loss": 0.0216,
+      "step": 27610
+    },
+    {
+      "epoch": 163.4319526627219,
+      "grad_norm": 1.2599066495895386,
+      "learning_rate": 0.0001,
+      "loss": 0.0209,
+      "step": 27620
+    },
+    {
+      "epoch": 163.49112426035504,
+      "grad_norm": 1.6199090480804443,
+      "learning_rate": 0.0001,
+      "loss": 0.0201,
+      "step": 27630
+    },
+    {
+      "epoch": 163.55029585798817,
+      "grad_norm": 1.6784976720809937,
+      "learning_rate": 0.0001,
+      "loss": 0.0203,
+      "step": 27640
+    },
+    {
+      "epoch": 163.6094674556213,
+      "grad_norm": 1.0221621990203857,
+      "learning_rate": 0.0001,
+      "loss": 0.0197,
+      "step": 27650
+    },
+    {
+      "epoch": 163.66863905325442,
+      "grad_norm": 1.2185795307159424,
+      "learning_rate": 0.0001,
+      "loss": 0.0204,
+      "step": 27660
+    },
+    {
+      "epoch": 163.72781065088756,
+      "grad_norm": 1.4099150896072388,
+      "learning_rate": 0.0001,
+      "loss": 0.021,
+      "step": 27670
+    },
+    {
+      "epoch": 163.7869822485207,
+      "grad_norm": 1.5567409992218018,
+      "learning_rate": 0.0001,
+      "loss": 0.0206,
+      "step": 27680
+    },
+    {
+      "epoch": 163.84615384615384,
+      "grad_norm": 1.5958704948425293,
+      "learning_rate": 0.0001,
+      "loss": 0.0206,
+      "step": 27690
+    },
+    {
+      "epoch": 163.90532544378698,
+      "grad_norm": 1.362967848777771,
+      "learning_rate": 0.0001,
+      "loss": 0.0209,
+      "step": 27700
+    },
+    {
+      "epoch": 163.96449704142012,
+      "grad_norm": 1.5655937194824219,
+      "learning_rate": 0.0001,
+      "loss": 0.0207,
+      "step": 27710
+    },
+    {
+      "epoch": 164.02366863905326,
+      "grad_norm": 2.046501874923706,
+      "learning_rate": 0.0001,
+      "loss": 0.0208,
+      "step": 27720
+    },
+    {
+      "epoch": 164.0828402366864,
+      "grad_norm": 1.525864839553833,
+      "learning_rate": 0.0001,
+      "loss": 0.0218,
+      "step": 27730
+    },
+    {
+      "epoch": 164.14201183431953,
+      "grad_norm": 1.4534013271331787,
+      "learning_rate": 0.0001,
+      "loss": 0.0212,
+      "step": 27740
+    },
+    {
+      "epoch": 164.20118343195267,
+      "grad_norm": 1.642037272453308,
+      "learning_rate": 0.0001,
+      "loss": 0.0204,
+      "step": 27750
+    },
+    {
+      "epoch": 164.2603550295858,
+      "grad_norm": 1.1394238471984863,
+      "learning_rate": 0.0001,
+      "loss": 0.0202,
+      "step": 27760
+    },
+    {
+      "epoch": 164.31952662721895,
+      "grad_norm": 1.769667148590088,
+      "learning_rate": 0.0001,
+      "loss": 0.0201,
+      "step": 27770
+    },
+    {
+      "epoch": 164.37869822485206,
+      "grad_norm": 1.9288551807403564,
+      "learning_rate": 0.0001,
+      "loss": 0.021,
+      "step": 27780
+    },
+    {
+      "epoch": 164.4378698224852,
+      "grad_norm": 1.4136570692062378,
+      "learning_rate": 0.0001,
+      "loss": 0.02,
+      "step": 27790
+    },
+    {
+      "epoch": 164.49704142011834,
+      "grad_norm": 1.3508259057998657,
+      "learning_rate": 0.0001,
+      "loss": 0.0208,
+      "step": 27800
+    },
+    {
+      "epoch": 164.55621301775147,
+      "grad_norm": 2.1387526988983154,
+      "learning_rate": 0.0001,
+      "loss": 0.0194,
+      "step": 27810
+    },
+    {
+      "epoch": 164.6153846153846,
+      "grad_norm": 1.5893616676330566,
+      "learning_rate": 0.0001,
+      "loss": 0.0205,
+      "step": 27820
+    },
+    {
+      "epoch": 164.67455621301775,
+      "grad_norm": 1.3508079051971436,
+      "learning_rate": 0.0001,
+      "loss": 0.0204,
+      "step": 27830
+    },
+    {
+      "epoch": 164.7337278106509,
+      "grad_norm": 1.213654637336731,
+      "learning_rate": 0.0001,
+      "loss": 0.0202,
+      "step": 27840
+    },
+    {
+      "epoch": 164.79289940828403,
+      "grad_norm": 1.1652841567993164,
+      "learning_rate": 0.0001,
+      "loss": 0.0193,
+      "step": 27850
+    },
+    {
+      "epoch": 164.85207100591717,
+      "grad_norm": 1.4941798448562622,
+      "learning_rate": 0.0001,
+      "loss": 0.0197,
+      "step": 27860
+    },
+    {
+      "epoch": 164.9112426035503,
+      "grad_norm": 1.5487040281295776,
+      "learning_rate": 0.0001,
+      "loss": 0.0199,
+      "step": 27870
+    },
+    {
+      "epoch": 164.97041420118344,
+      "grad_norm": 1.3027018308639526,
+      "learning_rate": 0.0001,
+      "loss": 0.0203,
+      "step": 27880
+    },
+    {
+      "epoch": 165.02958579881656,
+      "grad_norm": 1.9115771055221558,
+      "learning_rate": 0.0001,
+      "loss": 0.0212,
+      "step": 27890
+    },
+    {
+      "epoch": 165.0887573964497,
+      "grad_norm": 2.1398773193359375,
+      "learning_rate": 0.0001,
+      "loss": 0.0206,
+      "step": 27900
+    },
+    {
+      "epoch": 165.14792899408283,
+      "grad_norm": 2.067063331604004,
+      "learning_rate": 0.0001,
+      "loss": 0.0212,
+      "step": 27910
+    },
+    {
+      "epoch": 165.20710059171597,
+      "grad_norm": 2.0062592029571533,
+      "learning_rate": 0.0001,
+      "loss": 0.0201,
+      "step": 27920
+    },
+    {
+      "epoch": 165.2662721893491,
+      "grad_norm": 1.900478720664978,
+      "learning_rate": 0.0001,
+      "loss": 0.021,
+      "step": 27930
+    },
+    {
+      "epoch": 165.32544378698225,
+      "grad_norm": 1.6567600965499878,
+      "learning_rate": 0.0001,
+      "loss": 0.0205,
+      "step": 27940
+    },
+    {
+      "epoch": 165.3846153846154,
+      "grad_norm": 1.256301999092102,
+      "learning_rate": 0.0001,
+      "loss": 0.0193,
+      "step": 27950
+    },
+    {
+      "epoch": 165.44378698224853,
+      "grad_norm": 1.362735629081726,
+      "learning_rate": 0.0001,
+      "loss": 0.0194,
+      "step": 27960
+    },
+    {
+      "epoch": 165.50295857988166,
+      "grad_norm": 1.129341959953308,
+      "learning_rate": 0.0001,
+      "loss": 0.0189,
+      "step": 27970
+    },
+    {
+      "epoch": 165.5621301775148,
+      "grad_norm": 1.402551531791687,
+      "learning_rate": 0.0001,
+      "loss": 0.0198,
+      "step": 27980
+    },
+    {
+      "epoch": 165.62130177514794,
+      "grad_norm": 1.0141092538833618,
+      "learning_rate": 0.0001,
+      "loss": 0.0204,
+      "step": 27990
+    },
+    {
+      "epoch": 165.68047337278105,
+      "grad_norm": 1.6125699281692505,
+      "learning_rate": 0.0001,
+      "loss": 0.0206,
+      "step": 28000
+    },
+    {
+      "epoch": 165.7396449704142,
+      "grad_norm": 1.5297123193740845,
+      "learning_rate": 0.0001,
+      "loss": 0.0211,
+      "step": 28010
+    },
+    {
+      "epoch": 165.79881656804733,
+      "grad_norm": 1.2079156637191772,
+      "learning_rate": 0.0001,
+      "loss": 0.0207,
+      "step": 28020
+    },
+    {
+      "epoch": 165.85798816568047,
+      "grad_norm": 1.598610758781433,
+      "learning_rate": 0.0001,
+      "loss": 0.0209,
+      "step": 28030
+    },
+    {
+      "epoch": 165.9171597633136,
+      "grad_norm": 1.3283164501190186,
+      "learning_rate": 0.0001,
+      "loss": 0.0214,
+      "step": 28040
+    },
+    {
+      "epoch": 165.97633136094674,
+      "grad_norm": 1.4421133995056152,
+      "learning_rate": 0.0001,
+      "loss": 0.0206,
+      "step": 28050
+    },
+    {
+      "epoch": 166.03550295857988,
+      "grad_norm": 1.7973997592926025,
+      "learning_rate": 0.0001,
+      "loss": 0.022,
+      "step": 28060
+    },
+    {
+      "epoch": 166.09467455621302,
+      "grad_norm": 1.5220794677734375,
+      "learning_rate": 0.0001,
+      "loss": 0.0212,
+      "step": 28070
+    },
+    {
+      "epoch": 166.15384615384616,
+      "grad_norm": 2.1048288345336914,
+      "learning_rate": 0.0001,
+      "loss": 0.0211,
+      "step": 28080
+    },
+    {
+      "epoch": 166.2130177514793,
+      "grad_norm": 1.98041570186615,
+      "learning_rate": 0.0001,
+      "loss": 0.02,
+      "step": 28090
+    },
+    {
+      "epoch": 166.27218934911244,
+      "grad_norm": 1.1499509811401367,
+      "learning_rate": 0.0001,
+      "loss": 0.02,
+      "step": 28100
+    },
+    {
+      "epoch": 166.33136094674558,
+      "grad_norm": 1.602230191230774,
+      "learning_rate": 0.0001,
+      "loss": 0.0197,
+      "step": 28110
+    },
+    {
+      "epoch": 166.3905325443787,
+      "grad_norm": 1.6328662633895874,
+      "learning_rate": 0.0001,
+      "loss": 0.0198,
+      "step": 28120
+    },
+    {
+      "epoch": 166.44970414201183,
+      "grad_norm": 1.5837839841842651,
+      "learning_rate": 0.0001,
+      "loss": 0.0202,
+      "step": 28130
+    },
+    {
+      "epoch": 166.50887573964496,
+      "grad_norm": 1.112396478652954,
+      "learning_rate": 0.0001,
+      "loss": 0.0204,
+      "step": 28140
+    },
+    {
+      "epoch": 166.5680473372781,
+      "grad_norm": 1.3772350549697876,
+      "learning_rate": 0.0001,
+      "loss": 0.0195,
+      "step": 28150
+    },
+    {
+      "epoch": 166.62721893491124,
+      "grad_norm": 1.2232868671417236,
+      "learning_rate": 0.0001,
+      "loss": 0.0208,
+      "step": 28160
+    },
+    {
+      "epoch": 166.68639053254438,
+      "grad_norm": 1.5135313272476196,
+      "learning_rate": 0.0001,
+      "loss": 0.0206,
+      "step": 28170
+    },
+    {
+      "epoch": 166.74556213017752,
+      "grad_norm": 1.9510449171066284,
+      "learning_rate": 0.0001,
+      "loss": 0.0219,
+      "step": 28180
+    },
+    {
+      "epoch": 166.80473372781066,
+      "grad_norm": 1.631888508796692,
+      "learning_rate": 0.0001,
+      "loss": 0.0209,
+      "step": 28190
+    },
+    {
+      "epoch": 166.8639053254438,
+      "grad_norm": 0.998680591583252,
+      "learning_rate": 0.0001,
+      "loss": 0.0202,
+      "step": 28200
+    },
+    {
+      "epoch": 166.92307692307693,
+      "grad_norm": 1.169971227645874,
+      "learning_rate": 0.0001,
+      "loss": 0.021,
+      "step": 28210
+    },
+    {
+      "epoch": 166.98224852071007,
+      "grad_norm": 1.02693510055542,
+      "learning_rate": 0.0001,
+      "loss": 0.021,
+      "step": 28220
+    },
+    {
+      "epoch": 167.04142011834318,
+      "grad_norm": 1.1905282735824585,
+      "learning_rate": 0.0001,
+      "loss": 0.0207,
+      "step": 28230
+    },
+    {
+      "epoch": 167.10059171597632,
+      "grad_norm": 1.2627322673797607,
+      "learning_rate": 0.0001,
+      "loss": 0.0212,
+      "step": 28240
+    },
+    {
+      "epoch": 167.15976331360946,
+      "grad_norm": 1.3498303890228271,
+      "learning_rate": 0.0001,
+      "loss": 0.0219,
+      "step": 28250
+    },
+    {
+      "epoch": 167.2189349112426,
+      "grad_norm": 1.7049435377120972,
+      "learning_rate": 0.0001,
+      "loss": 0.0228,
+      "step": 28260
+    },
+    {
+      "epoch": 167.27810650887574,
+      "grad_norm": 1.4110229015350342,
+      "learning_rate": 0.0001,
+      "loss": 0.0224,
+      "step": 28270
+    },
+    {
+      "epoch": 167.33727810650888,
+      "grad_norm": 1.2246208190917969,
+      "learning_rate": 0.0001,
+      "loss": 0.0212,
+      "step": 28280
+    },
+    {
+      "epoch": 167.39644970414201,
+      "grad_norm": 1.0329488515853882,
+      "learning_rate": 0.0001,
+      "loss": 0.0208,
+      "step": 28290
+    },
+    {
+      "epoch": 167.45562130177515,
+      "grad_norm": 1.6445579528808594,
+      "learning_rate": 0.0001,
+      "loss": 0.0215,
+      "step": 28300
+    },
+    {
+      "epoch": 167.5147928994083,
+      "grad_norm": 1.343801498413086,
+      "learning_rate": 0.0001,
+      "loss": 0.0215,
+      "step": 28310
+    },
+    {
+      "epoch": 167.57396449704143,
+      "grad_norm": 1.3765617609024048,
+      "learning_rate": 0.0001,
+      "loss": 0.0209,
+      "step": 28320
+    },
+    {
+      "epoch": 167.63313609467457,
+      "grad_norm": 1.5076580047607422,
+      "learning_rate": 0.0001,
+      "loss": 0.0217,
+      "step": 28330
+    },
+    {
+      "epoch": 167.69230769230768,
+      "grad_norm": 1.040397047996521,
+      "learning_rate": 0.0001,
+      "loss": 0.0203,
+      "step": 28340
+    },
+    {
+      "epoch": 167.75147928994082,
+      "grad_norm": 1.3323358297348022,
+      "learning_rate": 0.0001,
+      "loss": 0.0207,
+      "step": 28350
+    },
+    {
+      "epoch": 167.81065088757396,
+      "grad_norm": 1.6750715970993042,
+      "learning_rate": 0.0001,
+      "loss": 0.0203,
+      "step": 28360
+    },
+    {
+      "epoch": 167.8698224852071,
+      "grad_norm": 1.3826377391815186,
+      "learning_rate": 0.0001,
+      "loss": 0.0208,
+      "step": 28370
+    },
+    {
+      "epoch": 167.92899408284023,
+      "grad_norm": 1.2411855459213257,
+      "learning_rate": 0.0001,
+      "loss": 0.0212,
+      "step": 28380
+    },
+    {
+      "epoch": 167.98816568047337,
+      "grad_norm": 1.6303590536117554,
+      "learning_rate": 0.0001,
+      "loss": 0.0206,
+      "step": 28390
+    },
+    {
+      "epoch": 168.0473372781065,
+      "grad_norm": 1.4624512195587158,
+      "learning_rate": 0.0001,
+      "loss": 0.0207,
+      "step": 28400
+    },
+    {
+      "epoch": 168.10650887573965,
+      "grad_norm": 1.8371764421463013,
+      "learning_rate": 0.0001,
+      "loss": 0.0207,
+      "step": 28410
+    },
+    {
+      "epoch": 168.1656804733728,
+      "grad_norm": 1.3704978227615356,
+      "learning_rate": 0.0001,
+      "loss": 0.0207,
+      "step": 28420
+    },
+    {
+      "epoch": 168.22485207100593,
+      "grad_norm": 1.2947602272033691,
+      "learning_rate": 0.0001,
+      "loss": 0.0201,
+      "step": 28430
+    },
+    {
+      "epoch": 168.28402366863907,
+      "grad_norm": 0.8694601655006409,
+      "learning_rate": 0.0001,
+      "loss": 0.0198,
+      "step": 28440
+    },
+    {
+      "epoch": 168.34319526627218,
+      "grad_norm": 1.1108675003051758,
+      "learning_rate": 0.0001,
+      "loss": 0.0202,
+      "step": 28450
+    },
+    {
+      "epoch": 168.4023668639053,
+      "grad_norm": 1.426690936088562,
+      "learning_rate": 0.0001,
+      "loss": 0.0202,
+      "step": 28460
+    },
+    {
+      "epoch": 168.46153846153845,
+      "grad_norm": 1.339786171913147,
+      "learning_rate": 0.0001,
+      "loss": 0.0211,
+      "step": 28470
+    },
+    {
+      "epoch": 168.5207100591716,
+      "grad_norm": 1.308549165725708,
+      "learning_rate": 0.0001,
+      "loss": 0.0218,
+      "step": 28480
+    },
+    {
+      "epoch": 168.57988165680473,
+      "grad_norm": 1.0546276569366455,
+      "learning_rate": 0.0001,
+      "loss": 0.0214,
+      "step": 28490
+    },
+    {
+      "epoch": 168.63905325443787,
+      "grad_norm": 1.1581183671951294,
+      "learning_rate": 0.0001,
+      "loss": 0.0211,
+      "step": 28500
+    },
+    {
+      "epoch": 168.698224852071,
+      "grad_norm": 1.21005117893219,
+      "learning_rate": 0.0001,
+      "loss": 0.0211,
+      "step": 28510
+    },
+    {
+      "epoch": 168.75739644970415,
+      "grad_norm": 1.1822372674942017,
+      "learning_rate": 0.0001,
+      "loss": 0.0206,
+      "step": 28520
+    },
+    {
+      "epoch": 168.81656804733728,
+      "grad_norm": 1.6504062414169312,
+      "learning_rate": 0.0001,
+      "loss": 0.0208,
+      "step": 28530
+    },
+    {
+      "epoch": 168.87573964497042,
+      "grad_norm": 1.3964111804962158,
+      "learning_rate": 0.0001,
+      "loss": 0.0212,
+      "step": 28540
+    },
+    {
+      "epoch": 168.93491124260356,
+      "grad_norm": 1.2515379190444946,
+      "learning_rate": 0.0001,
+      "loss": 0.0215,
+      "step": 28550
+    },
+    {
+      "epoch": 168.9940828402367,
+      "grad_norm": 1.3420222997665405,
+      "learning_rate": 0.0001,
+      "loss": 0.0206,
+      "step": 28560
+    },
+    {
+      "epoch": 169.0532544378698,
+      "grad_norm": 1.9098690748214722,
+      "learning_rate": 0.0001,
+      "loss": 0.0214,
+      "step": 28570
+    },
+    {
+      "epoch": 169.11242603550295,
+      "grad_norm": 1.1409969329833984,
+      "learning_rate": 0.0001,
+      "loss": 0.0195,
+      "step": 28580
+    },
+    {
+      "epoch": 169.1715976331361,
+      "grad_norm": 1.4341984987258911,
+      "learning_rate": 0.0001,
+      "loss": 0.0199,
+      "step": 28590
+    },
+    {
+      "epoch": 169.23076923076923,
+      "grad_norm": 1.3402860164642334,
+      "learning_rate": 0.0001,
+      "loss": 0.0201,
+      "step": 28600
+    },
+    {
+      "epoch": 169.28994082840237,
+      "grad_norm": 1.084418773651123,
+      "learning_rate": 0.0001,
+      "loss": 0.0202,
+      "step": 28610
+    },
+    {
+      "epoch": 169.3491124260355,
+      "grad_norm": 1.647748589515686,
+      "learning_rate": 0.0001,
+      "loss": 0.0206,
+      "step": 28620
+    },
+    {
+      "epoch": 169.40828402366864,
+      "grad_norm": 1.5535495281219482,
+      "learning_rate": 0.0001,
+      "loss": 0.0207,
+      "step": 28630
+    },
+    {
+      "epoch": 169.46745562130178,
+      "grad_norm": 1.8607282638549805,
+      "learning_rate": 0.0001,
+      "loss": 0.0209,
+      "step": 28640
+    },
+    {
+      "epoch": 169.52662721893492,
+      "grad_norm": 1.7550369501113892,
+      "learning_rate": 0.0001,
+      "loss": 0.0212,
+      "step": 28650
+    },
+    {
+      "epoch": 169.58579881656806,
+      "grad_norm": 1.263322353363037,
+      "learning_rate": 0.0001,
+      "loss": 0.0206,
+      "step": 28660
+    },
+    {
+      "epoch": 169.6449704142012,
+      "grad_norm": 1.3744840621948242,
+      "learning_rate": 0.0001,
+      "loss": 0.0201,
+      "step": 28670
+    },
+    {
+      "epoch": 169.7041420118343,
+      "grad_norm": 1.6408401727676392,
+      "learning_rate": 0.0001,
+      "loss": 0.0199,
+      "step": 28680
+    },
+    {
+      "epoch": 169.76331360946745,
+      "grad_norm": 1.4143186807632446,
+      "learning_rate": 0.0001,
+      "loss": 0.0203,
+      "step": 28690
+    },
+    {
+      "epoch": 169.82248520710058,
+      "grad_norm": 1.4227958917617798,
+      "learning_rate": 0.0001,
+      "loss": 0.0205,
+      "step": 28700
+    },
+    {
+      "epoch": 169.88165680473372,
+      "grad_norm": 1.1799561977386475,
+      "learning_rate": 0.0001,
+      "loss": 0.0203,
+      "step": 28710
+    },
+    {
+      "epoch": 169.94082840236686,
+      "grad_norm": 1.5431208610534668,
+      "learning_rate": 0.0001,
+      "loss": 0.021,
+      "step": 28720
+    },
+    {
+      "epoch": 170.0,
+      "grad_norm": 1.1486120223999023,
+      "learning_rate": 0.0001,
+      "loss": 0.0208,
+      "step": 28730
+    },
+    {
+      "epoch": 170.05917159763314,
+      "grad_norm": 1.3249890804290771,
+      "learning_rate": 0.0001,
+      "loss": 0.0202,
+      "step": 28740
+    },
+    {
+      "epoch": 170.11834319526628,
+      "grad_norm": 1.0075592994689941,
+      "learning_rate": 0.0001,
+      "loss": 0.0205,
+      "step": 28750
+    },
+    {
+      "epoch": 170.17751479289942,
+      "grad_norm": 1.2552599906921387,
+      "learning_rate": 0.0001,
+      "loss": 0.0201,
+      "step": 28760
+    },
+    {
+      "epoch": 170.23668639053255,
+      "grad_norm": 1.3920470476150513,
+      "learning_rate": 0.0001,
+      "loss": 0.0202,
+      "step": 28770
+    },
+    {
+      "epoch": 170.2958579881657,
+      "grad_norm": 1.213984489440918,
+      "learning_rate": 0.0001,
+      "loss": 0.0201,
+      "step": 28780
+    },
+    {
+      "epoch": 170.3550295857988,
+      "grad_norm": 1.1243146657943726,
+      "learning_rate": 0.0001,
+      "loss": 0.0213,
+      "step": 28790
+    },
+    {
+      "epoch": 170.41420118343194,
+      "grad_norm": 1.6383343935012817,
+      "learning_rate": 0.0001,
+      "loss": 0.0215,
+      "step": 28800
+    },
+    {
+      "epoch": 170.47337278106508,
+      "grad_norm": 1.5189059972763062,
+      "learning_rate": 0.0001,
+      "loss": 0.0213,
+      "step": 28810
+    },
+    {
+      "epoch": 170.53254437869822,
+      "grad_norm": 1.2604376077651978,
+      "learning_rate": 0.0001,
+      "loss": 0.0217,
+      "step": 28820
+    },
+    {
+      "epoch": 170.59171597633136,
+      "grad_norm": 1.3695119619369507,
+      "learning_rate": 0.0001,
+      "loss": 0.0207,
+      "step": 28830
+    },
+    {
+      "epoch": 170.6508875739645,
+      "grad_norm": 1.5070149898529053,
+      "learning_rate": 0.0001,
+      "loss": 0.021,
+      "step": 28840
+    },
+    {
+      "epoch": 170.71005917159763,
+      "grad_norm": 1.1843678951263428,
+      "learning_rate": 0.0001,
+      "loss": 0.0201,
+      "step": 28850
+    },
+    {
+      "epoch": 170.76923076923077,
+      "grad_norm": 1.320669412612915,
+      "learning_rate": 0.0001,
+      "loss": 0.0204,
+      "step": 28860
+    },
+    {
+      "epoch": 170.8284023668639,
+      "grad_norm": 1.3693448305130005,
+      "learning_rate": 0.0001,
+      "loss": 0.0205,
+      "step": 28870
+    },
+    {
+      "epoch": 170.88757396449705,
+      "grad_norm": 1.0337870121002197,
+      "learning_rate": 0.0001,
+      "loss": 0.0203,
+      "step": 28880
+    },
+    {
+      "epoch": 170.9467455621302,
+      "grad_norm": 1.3718266487121582,
+      "learning_rate": 0.0001,
+      "loss": 0.0207,
+      "step": 28890
+    },
+    {
+      "epoch": 171.0059171597633,
+      "grad_norm": 1.3123154640197754,
+      "learning_rate": 0.0001,
+      "loss": 0.0202,
+      "step": 28900
+    },
+    {
+      "epoch": 171.06508875739644,
+      "grad_norm": 1.2668001651763916,
+      "learning_rate": 0.0001,
+      "loss": 0.0199,
+      "step": 28910
+    },
+    {
+      "epoch": 171.12426035502958,
+      "grad_norm": 1.158882975578308,
+      "learning_rate": 0.0001,
+      "loss": 0.0198,
+      "step": 28920
+    },
+    {
+      "epoch": 171.18343195266272,
+      "grad_norm": 1.5298535823822021,
+      "learning_rate": 0.0001,
+      "loss": 0.0201,
+      "step": 28930
+    },
+    {
+      "epoch": 171.24260355029585,
+      "grad_norm": 1.679901361465454,
+      "learning_rate": 0.0001,
+      "loss": 0.0216,
+      "step": 28940
+    },
+    {
+      "epoch": 171.301775147929,
+      "grad_norm": 1.382987380027771,
+      "learning_rate": 0.0001,
+      "loss": 0.0208,
+      "step": 28950
+    },
+    {
+      "epoch": 171.36094674556213,
+      "grad_norm": 1.252353549003601,
+      "learning_rate": 0.0001,
+      "loss": 0.021,
+      "step": 28960
+    },
+    {
+      "epoch": 171.42011834319527,
+      "grad_norm": 1.1847411394119263,
+      "learning_rate": 0.0001,
+      "loss": 0.0208,
+      "step": 28970
+    },
+    {
+      "epoch": 171.4792899408284,
+      "grad_norm": 1.5485897064208984,
+      "learning_rate": 0.0001,
+      "loss": 0.0204,
+      "step": 28980
+    },
+    {
+      "epoch": 171.53846153846155,
+      "grad_norm": 1.0680222511291504,
+      "learning_rate": 0.0001,
+      "loss": 0.022,
+      "step": 28990
+    },
+    {
+      "epoch": 171.5976331360947,
+      "grad_norm": 1.4391332864761353,
+      "learning_rate": 0.0001,
+      "loss": 0.0212,
+      "step": 29000
+    },
+    {
+      "epoch": 171.65680473372782,
+      "grad_norm": 1.5344312191009521,
+      "learning_rate": 0.0001,
+      "loss": 0.0215,
+      "step": 29010
+    },
+    {
+      "epoch": 171.71597633136093,
+      "grad_norm": 1.7186328172683716,
+      "learning_rate": 0.0001,
+      "loss": 0.0218,
+      "step": 29020
+    },
+    {
+      "epoch": 171.77514792899407,
+      "grad_norm": 1.4204037189483643,
+      "learning_rate": 0.0001,
+      "loss": 0.0215,
+      "step": 29030
+    },
+    {
+      "epoch": 171.8343195266272,
+      "grad_norm": 1.4369336366653442,
+      "learning_rate": 0.0001,
+      "loss": 0.021,
+      "step": 29040
+    },
+    {
+      "epoch": 171.89349112426035,
+      "grad_norm": 1.0658568143844604,
+      "learning_rate": 0.0001,
+      "loss": 0.0206,
+      "step": 29050
+    },
+    {
+      "epoch": 171.9526627218935,
+      "grad_norm": 1.1952731609344482,
+      "learning_rate": 0.0001,
+      "loss": 0.02,
+      "step": 29060
+    },
+    {
+      "epoch": 172.01183431952663,
+      "grad_norm": 1.6892961263656616,
+      "learning_rate": 0.0001,
+      "loss": 0.0199,
+      "step": 29070
+    },
+    {
+      "epoch": 172.07100591715977,
+      "grad_norm": 1.3147574663162231,
+      "learning_rate": 0.0001,
+      "loss": 0.02,
+      "step": 29080
+    },
+    {
+      "epoch": 172.1301775147929,
+      "grad_norm": 1.232138752937317,
+      "learning_rate": 0.0001,
+      "loss": 0.02,
+      "step": 29090
+    },
+    {
+      "epoch": 172.18934911242604,
+      "grad_norm": 1.2529348134994507,
+      "learning_rate": 0.0001,
+      "loss": 0.0195,
+      "step": 29100
+    },
+    {
+      "epoch": 172.24852071005918,
+      "grad_norm": 1.0703256130218506,
+      "learning_rate": 0.0001,
+      "loss": 0.0199,
+      "step": 29110
+    },
+    {
+      "epoch": 172.30769230769232,
+      "grad_norm": 1.1079317331314087,
+      "learning_rate": 0.0001,
+      "loss": 0.02,
+      "step": 29120
+    },
+    {
+      "epoch": 172.36686390532543,
+      "grad_norm": 1.4073727130889893,
+      "learning_rate": 0.0001,
+      "loss": 0.0213,
+      "step": 29130
+    },
+    {
+      "epoch": 172.42603550295857,
+      "grad_norm": 1.081437349319458,
+      "learning_rate": 0.0001,
+      "loss": 0.0198,
+      "step": 29140
+    },
+    {
+      "epoch": 172.4852071005917,
+      "grad_norm": 1.0226389169692993,
+      "learning_rate": 0.0001,
+      "loss": 0.0197,
+      "step": 29150
+    },
+    {
+      "epoch": 172.54437869822485,
+      "grad_norm": 1.392824649810791,
+      "learning_rate": 0.0001,
+      "loss": 0.0208,
+      "step": 29160
+    },
+    {
+      "epoch": 172.60355029585799,
+      "grad_norm": 1.2494075298309326,
+      "learning_rate": 0.0001,
+      "loss": 0.0207,
+      "step": 29170
+    },
+    {
+      "epoch": 172.66272189349112,
+      "grad_norm": 1.3780487775802612,
+      "learning_rate": 0.0001,
+      "loss": 0.0203,
+      "step": 29180
+    },
+    {
+      "epoch": 172.72189349112426,
+      "grad_norm": 1.5273252725601196,
+      "learning_rate": 0.0001,
+      "loss": 0.0204,
+      "step": 29190
+    },
+    {
+      "epoch": 172.7810650887574,
+      "grad_norm": 1.6077675819396973,
+      "learning_rate": 0.0001,
+      "loss": 0.0207,
+      "step": 29200
+    },
+    {
+      "epoch": 172.84023668639054,
+      "grad_norm": 1.3842045068740845,
+      "learning_rate": 0.0001,
+      "loss": 0.0205,
+      "step": 29210
+    },
+    {
+      "epoch": 172.89940828402368,
+      "grad_norm": 1.388234257698059,
+      "learning_rate": 0.0001,
+      "loss": 0.02,
+      "step": 29220
+    },
+    {
+      "epoch": 172.95857988165682,
+      "grad_norm": 1.3920035362243652,
+      "learning_rate": 0.0001,
+      "loss": 0.0198,
+      "step": 29230
+    },
+    {
+      "epoch": 173.01775147928993,
+      "grad_norm": 1.481914758682251,
+      "learning_rate": 0.0001,
+      "loss": 0.0197,
+      "step": 29240
+    },
+    {
+      "epoch": 173.07692307692307,
+      "grad_norm": 1.9563484191894531,
+      "learning_rate": 0.0001,
+      "loss": 0.0205,
+      "step": 29250
+    },
+    {
+      "epoch": 173.1360946745562,
+      "grad_norm": 1.6201084852218628,
+      "learning_rate": 0.0001,
+      "loss": 0.0204,
+      "step": 29260
+    },
+    {
+      "epoch": 173.19526627218934,
+      "grad_norm": 1.8054239749908447,
+      "learning_rate": 0.0001,
+      "loss": 0.0195,
+      "step": 29270
+    },
+    {
+      "epoch": 173.25443786982248,
+      "grad_norm": 1.5917344093322754,
+      "learning_rate": 0.0001,
+      "loss": 0.0205,
+      "step": 29280
+    },
+    {
+      "epoch": 173.31360946745562,
+      "grad_norm": 1.4886672496795654,
+      "learning_rate": 0.0001,
+      "loss": 0.0195,
+      "step": 29290
+    },
+    {
+      "epoch": 173.37278106508876,
+      "grad_norm": 1.471673846244812,
+      "learning_rate": 0.0001,
+      "loss": 0.0205,
+      "step": 29300
+    },
+    {
+      "epoch": 173.4319526627219,
+      "grad_norm": 1.7491230964660645,
+      "learning_rate": 0.0001,
+      "loss": 0.0209,
+      "step": 29310
+    },
+    {
+      "epoch": 173.49112426035504,
+      "grad_norm": 1.7799426317214966,
+      "learning_rate": 0.0001,
+      "loss": 0.0202,
+      "step": 29320
+    },
+    {
+      "epoch": 173.55029585798817,
+      "grad_norm": 1.5339809656143188,
+      "learning_rate": 0.0001,
+      "loss": 0.0199,
+      "step": 29330
+    },
+    {
+      "epoch": 173.6094674556213,
+      "grad_norm": 1.555969476699829,
+      "learning_rate": 0.0001,
+      "loss": 0.0195,
+      "step": 29340
+    },
+    {
+      "epoch": 173.66863905325442,
+      "grad_norm": 1.247040867805481,
+      "learning_rate": 0.0001,
+      "loss": 0.0201,
+      "step": 29350
+    },
+    {
+      "epoch": 173.72781065088756,
+      "grad_norm": 1.0060384273529053,
+      "learning_rate": 0.0001,
+      "loss": 0.0204,
+      "step": 29360
+    },
+    {
+      "epoch": 173.7869822485207,
+      "grad_norm": 1.3172825574874878,
+      "learning_rate": 0.0001,
+      "loss": 0.0206,
+      "step": 29370
+    },
+    {
+      "epoch": 173.84615384615384,
+      "grad_norm": 1.3392817974090576,
+      "learning_rate": 0.0001,
+      "loss": 0.0211,
+      "step": 29380
+    },
+    {
+      "epoch": 173.90532544378698,
+      "grad_norm": 1.2696956396102905,
+      "learning_rate": 0.0001,
+      "loss": 0.0197,
+      "step": 29390
+    },
+    {
+      "epoch": 173.96449704142012,
+      "grad_norm": 1.515369176864624,
+      "learning_rate": 0.0001,
+      "loss": 0.0196,
+      "step": 29400
+    },
+    {
+      "epoch": 174.02366863905326,
+      "grad_norm": 1.594237208366394,
+      "learning_rate": 0.0001,
+      "loss": 0.0198,
+      "step": 29410
+    },
+    {
+      "epoch": 174.0828402366864,
+      "grad_norm": 1.0530644655227661,
+      "learning_rate": 0.0001,
+      "loss": 0.0202,
+      "step": 29420
+    },
+    {
+      "epoch": 174.14201183431953,
+      "grad_norm": 1.2468581199645996,
+      "learning_rate": 0.0001,
+      "loss": 0.0197,
+      "step": 29430
+    },
+    {
+      "epoch": 174.20118343195267,
+      "grad_norm": 1.633934736251831,
+      "learning_rate": 0.0001,
+      "loss": 0.0203,
+      "step": 29440
+    },
+    {
+      "epoch": 174.2603550295858,
+      "grad_norm": 1.3016362190246582,
+      "learning_rate": 0.0001,
+      "loss": 0.0206,
+      "step": 29450
+    },
+    {
+      "epoch": 174.31952662721895,
+      "grad_norm": 1.9396010637283325,
+      "learning_rate": 0.0001,
+      "loss": 0.021,
+      "step": 29460
+    },
+    {
+      "epoch": 174.37869822485206,
+      "grad_norm": 1.6446977853775024,
+      "learning_rate": 0.0001,
+      "loss": 0.0206,
+      "step": 29470
+    },
+    {
+      "epoch": 174.4378698224852,
+      "grad_norm": 1.350695252418518,
+      "learning_rate": 0.0001,
+      "loss": 0.0206,
+      "step": 29480
+    },
+    {
+      "epoch": 174.49704142011834,
+      "grad_norm": 1.6368964910507202,
+      "learning_rate": 0.0001,
+      "loss": 0.0203,
+      "step": 29490
+    },
+    {
+      "epoch": 174.55621301775147,
+      "grad_norm": 1.0427780151367188,
+      "learning_rate": 0.0001,
+      "loss": 0.0198,
+      "step": 29500
+    },
+    {
+      "epoch": 174.6153846153846,
+      "grad_norm": 1.2636420726776123,
+      "learning_rate": 0.0001,
+      "loss": 0.0192,
+      "step": 29510
+    },
+    {
+      "epoch": 174.67455621301775,
+      "grad_norm": 1.108978271484375,
+      "learning_rate": 0.0001,
+      "loss": 0.0197,
+      "step": 29520
+    },
+    {
+      "epoch": 174.7337278106509,
+      "grad_norm": 1.4367252588272095,
+      "learning_rate": 0.0001,
+      "loss": 0.0202,
+      "step": 29530
+    },
+    {
+      "epoch": 174.79289940828403,
+      "grad_norm": 1.2309788465499878,
+      "learning_rate": 0.0001,
+      "loss": 0.0198,
+      "step": 29540
+    },
+    {
+      "epoch": 174.85207100591717,
+      "grad_norm": 1.6016262769699097,
+      "learning_rate": 0.0001,
+      "loss": 0.0205,
+      "step": 29550
+    },
+    {
+      "epoch": 174.9112426035503,
+      "grad_norm": 1.1001559495925903,
+      "learning_rate": 0.0001,
+      "loss": 0.0211,
+      "step": 29560
+    },
+    {
+      "epoch": 174.97041420118344,
+      "grad_norm": 1.0402005910873413,
+      "learning_rate": 0.0001,
+      "loss": 0.0205,
+      "step": 29570
+    },
+    {
+      "epoch": 175.02958579881656,
+      "grad_norm": 1.2219501733779907,
+      "learning_rate": 0.0001,
+      "loss": 0.021,
+      "step": 29580
+    },
+    {
+      "epoch": 175.0887573964497,
+      "grad_norm": 1.9616849422454834,
+      "learning_rate": 0.0001,
+      "loss": 0.0213,
+      "step": 29590
+    },
+    {
+      "epoch": 175.14792899408283,
+      "grad_norm": 1.4221283197402954,
+      "learning_rate": 0.0001,
+      "loss": 0.0215,
+      "step": 29600
+    },
+    {
+      "epoch": 175.20710059171597,
+      "grad_norm": 1.669772982597351,
+      "learning_rate": 0.0001,
+      "loss": 0.0211,
+      "step": 29610
+    },
+    {
+      "epoch": 175.2662721893491,
+      "grad_norm": 1.7154561281204224,
+      "learning_rate": 0.0001,
+      "loss": 0.021,
+      "step": 29620
+    },
+    {
+      "epoch": 175.32544378698225,
+      "grad_norm": 2.0588152408599854,
+      "learning_rate": 0.0001,
+      "loss": 0.0212,
+      "step": 29630
+    },
+    {
+      "epoch": 175.3846153846154,
+      "grad_norm": 1.3357089757919312,
+      "learning_rate": 0.0001,
+      "loss": 0.0217,
+      "step": 29640
+    },
+    {
+      "epoch": 175.44378698224853,
+      "grad_norm": 1.5747579336166382,
+      "learning_rate": 0.0001,
+      "loss": 0.0202,
+      "step": 29650
+    },
+    {
+      "epoch": 175.50295857988166,
+      "grad_norm": 1.3923267126083374,
+      "learning_rate": 0.0001,
+      "loss": 0.021,
+      "step": 29660
+    },
+    {
+      "epoch": 175.5621301775148,
+      "grad_norm": 0.9493241906166077,
+      "learning_rate": 0.0001,
+      "loss": 0.0209,
+      "step": 29670
+    },
+    {
+      "epoch": 175.62130177514794,
+      "grad_norm": 1.2377265691757202,
+      "learning_rate": 0.0001,
+      "loss": 0.02,
+      "step": 29680
+    },
+    {
+      "epoch": 175.68047337278105,
+      "grad_norm": 1.642826795578003,
+      "learning_rate": 0.0001,
+      "loss": 0.0204,
+      "step": 29690
+    },
+    {
+      "epoch": 175.7396449704142,
+      "grad_norm": 1.214561939239502,
+      "learning_rate": 0.0001,
+      "loss": 0.0205,
+      "step": 29700
+    },
+    {
+      "epoch": 175.79881656804733,
+      "grad_norm": 1.1034890413284302,
+      "learning_rate": 0.0001,
+      "loss": 0.0213,
+      "step": 29710
+    },
+    {
+      "epoch": 175.85798816568047,
+      "grad_norm": 1.6714036464691162,
+      "learning_rate": 0.0001,
+      "loss": 0.0204,
+      "step": 29720
+    },
+    {
+      "epoch": 175.9171597633136,
+      "grad_norm": 1.4150586128234863,
+      "learning_rate": 0.0001,
+      "loss": 0.0195,
+      "step": 29730
+    },
+    {
+      "epoch": 175.97633136094674,
+      "grad_norm": 0.9823798537254333,
+      "learning_rate": 0.0001,
+      "loss": 0.0205,
+      "step": 29740
+    },
+    {
+      "epoch": 176.03550295857988,
+      "grad_norm": 1.3715885877609253,
+      "learning_rate": 0.0001,
+      "loss": 0.0204,
+      "step": 29750
+    },
+    {
+      "epoch": 176.09467455621302,
+      "grad_norm": 1.2801471948623657,
+      "learning_rate": 0.0001,
+      "loss": 0.0208,
+      "step": 29760
+    },
+    {
+      "epoch": 176.15384615384616,
+      "grad_norm": 1.676878571510315,
+      "learning_rate": 0.0001,
+      "loss": 0.0198,
+      "step": 29770
+    },
+    {
+      "epoch": 176.2130177514793,
+      "grad_norm": 1.4663437604904175,
+      "learning_rate": 0.0001,
+      "loss": 0.0196,
+      "step": 29780
+    },
+    {
+      "epoch": 176.27218934911244,
+      "grad_norm": 1.6433913707733154,
+      "learning_rate": 0.0001,
+      "loss": 0.021,
+      "step": 29790
+    },
+    {
+      "epoch": 176.33136094674558,
+      "grad_norm": 1.424636960029602,
+      "learning_rate": 0.0001,
+      "loss": 0.0203,
+      "step": 29800
+    },
+    {
+      "epoch": 176.3905325443787,
+      "grad_norm": 1.5283465385437012,
+      "learning_rate": 0.0001,
+      "loss": 0.0205,
+      "step": 29810
+    },
+    {
+      "epoch": 176.44970414201183,
+      "grad_norm": 1.5467497110366821,
+      "learning_rate": 0.0001,
+      "loss": 0.0198,
+      "step": 29820
+    },
+    {
+      "epoch": 176.50887573964496,
+      "grad_norm": 1.709425687789917,
+      "learning_rate": 0.0001,
+      "loss": 0.0203,
+      "step": 29830
+    },
+    {
+      "epoch": 176.5680473372781,
+      "grad_norm": 1.3818328380584717,
+      "learning_rate": 0.0001,
+      "loss": 0.0206,
+      "step": 29840
+    },
+    {
+      "epoch": 176.62721893491124,
+      "grad_norm": 1.6466234922409058,
+      "learning_rate": 0.0001,
+      "loss": 0.0189,
+      "step": 29850
+    },
+    {
+      "epoch": 176.68639053254438,
+      "grad_norm": 1.3043073415756226,
+      "learning_rate": 0.0001,
+      "loss": 0.0195,
+      "step": 29860
+    },
+    {
+      "epoch": 176.74556213017752,
+      "grad_norm": 1.2633230686187744,
+      "learning_rate": 0.0001,
+      "loss": 0.0195,
+      "step": 29870
+    },
+    {
+      "epoch": 176.80473372781066,
+      "grad_norm": 1.1208112239837646,
+      "learning_rate": 0.0001,
+      "loss": 0.0187,
+      "step": 29880
+    },
+    {
+      "epoch": 176.8639053254438,
+      "grad_norm": 1.8045263290405273,
+      "learning_rate": 0.0001,
+      "loss": 0.0203,
+      "step": 29890
+    },
+    {
+      "epoch": 176.92307692307693,
+      "grad_norm": 1.2053465843200684,
+      "learning_rate": 0.0001,
+      "loss": 0.0203,
+      "step": 29900
+    },
+    {
+      "epoch": 176.98224852071007,
+      "grad_norm": 1.2236818075180054,
+      "learning_rate": 0.0001,
+      "loss": 0.0197,
+      "step": 29910
+    },
+    {
+      "epoch": 177.04142011834318,
+      "grad_norm": 0.8661774396896362,
+      "learning_rate": 0.0001,
+      "loss": 0.0189,
+      "step": 29920
+    },
+    {
+      "epoch": 177.10059171597632,
+      "grad_norm": 1.1514365673065186,
+      "learning_rate": 0.0001,
+      "loss": 0.0194,
+      "step": 29930
+    },
+    {
+      "epoch": 177.15976331360946,
+      "grad_norm": 1.3736686706542969,
+      "learning_rate": 0.0001,
+      "loss": 0.0193,
+      "step": 29940
+    },
+    {
+      "epoch": 177.2189349112426,
+      "grad_norm": 1.5175044536590576,
+      "learning_rate": 0.0001,
+      "loss": 0.0193,
+      "step": 29950
+    },
+    {
+      "epoch": 177.27810650887574,
+      "grad_norm": 1.2698074579238892,
+      "learning_rate": 0.0001,
+      "loss": 0.0203,
+      "step": 29960
+    },
+    {
+      "epoch": 177.33727810650888,
+      "grad_norm": 1.6316462755203247,
+      "learning_rate": 0.0001,
+      "loss": 0.0198,
+      "step": 29970
+    },
+    {
+      "epoch": 177.39644970414201,
+      "grad_norm": 1.8149237632751465,
+      "learning_rate": 0.0001,
+      "loss": 0.0197,
+      "step": 29980
+    },
+    {
+      "epoch": 177.45562130177515,
+      "grad_norm": 1.2934119701385498,
+      "learning_rate": 0.0001,
+      "loss": 0.0197,
+      "step": 29990
+    },
+    {
+      "epoch": 177.5147928994083,
+      "grad_norm": 1.1632301807403564,
+      "learning_rate": 0.0001,
+      "loss": 0.0197,
+      "step": 30000
+    },
+    {
+      "epoch": 177.5147928994083,
+      "step": 30000,
+      "total_flos": 0.0,
+      "train_loss": 0.06260814528216918,
+      "train_runtime": 8082.3039,
+      "train_samples_per_second": 237.556,
+      "train_steps_per_second": 3.712
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 30000,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 178,
+  "save_steps": 5000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 0.0,
+  "train_batch_size": 64,
+  "trial_name": null,
+  "trial_params": null
+}