diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,27321 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.965725806451613,
+  "eval_steps": 500,
+  "global_step": 39000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.1794965267181396,
+      "learning_rate": 9.997479838709678e-05,
+      "loss": 3.3587,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.1337946653366089,
+      "learning_rate": 9.994959677419355e-05,
+      "loss": 3.0324,
+      "step": 20
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.122248649597168,
+      "learning_rate": 9.992439516129033e-05,
+      "loss": 2.7894,
+      "step": 30
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.0954384803771973,
+      "learning_rate": 9.98991935483871e-05,
+      "loss": 2.8351,
+      "step": 40
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.8181716203689575,
+      "learning_rate": 9.987399193548388e-05,
+      "loss": 2.6836,
+      "step": 50
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.1688904762268066,
+      "learning_rate": 9.984879032258065e-05,
+      "loss": 2.4902,
+      "step": 60
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.0157718658447266,
+      "learning_rate": 9.982358870967743e-05,
+      "loss": 2.5195,
+      "step": 70
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.1250773668289185,
+      "learning_rate": 9.97983870967742e-05,
+      "loss": 2.5064,
+      "step": 80
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.891322135925293,
+      "learning_rate": 9.977318548387097e-05,
+      "loss": 2.2726,
+      "step": 90
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0769097805023193,
+      "learning_rate": 9.974798387096775e-05,
+      "loss": 2.4885,
+      "step": 100
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7138869166374207,
+      "learning_rate": 9.972278225806452e-05,
+      "loss": 2.3486,
+      "step": 110
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.695204257965088,
+      "learning_rate": 9.96975806451613e-05,
+      "loss": 2.1688,
+      "step": 120
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4924360513687134,
+      "learning_rate": 9.967237903225807e-05,
+      "loss": 2.2111,
+      "step": 130
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4759283065795898,
+      "learning_rate": 9.964717741935485e-05,
+      "loss": 2.3157,
+      "step": 140
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7681684494018555,
+      "learning_rate": 9.962197580645162e-05,
+      "loss": 2.2312,
+      "step": 150
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5629103183746338,
+      "learning_rate": 9.95967741935484e-05,
+      "loss": 2.2854,
+      "step": 160
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.588888168334961,
+      "learning_rate": 9.957157258064517e-05,
+      "loss": 2.4281,
+      "step": 170
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.181005835533142,
+      "learning_rate": 9.954637096774194e-05,
+      "loss": 2.2416,
+      "step": 180
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7629745006561279,
+      "learning_rate": 9.952116935483872e-05,
+      "loss": 2.3363,
+      "step": 190
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5010342597961426,
+      "learning_rate": 9.949596774193549e-05,
+      "loss": 2.2955,
+      "step": 200
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.1263885498046875,
+      "learning_rate": 9.947076612903227e-05,
+      "loss": 2.0889,
+      "step": 210
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9182795286178589,
+      "learning_rate": 9.944556451612904e-05,
+      "loss": 2.2283,
+      "step": 220
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.1709370613098145,
+      "learning_rate": 9.942036290322581e-05,
+      "loss": 2.3254,
+      "step": 230
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8916822075843811,
+      "learning_rate": 9.939516129032258e-05,
+      "loss": 2.1514,
+      "step": 240
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2090011835098267,
+      "learning_rate": 9.936995967741935e-05,
+      "loss": 2.1541,
+      "step": 250
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3280044794082642,
+      "learning_rate": 9.934475806451612e-05,
+      "loss": 2.1461,
+      "step": 260
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2679225206375122,
+      "learning_rate": 9.93195564516129e-05,
+      "loss": 2.3431,
+      "step": 270
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.1196717023849487,
+      "learning_rate": 9.929435483870969e-05,
+      "loss": 2.2497,
+      "step": 280
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.1959799528121948,
+      "learning_rate": 9.926915322580646e-05,
+      "loss": 2.0817,
+      "step": 290
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.291515588760376,
+      "learning_rate": 9.924395161290324e-05,
+      "loss": 2.0901,
+      "step": 300
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5644803047180176,
+      "learning_rate": 9.921875000000001e-05,
+      "loss": 2.1575,
+      "step": 310
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4184613227844238,
+      "learning_rate": 9.919354838709678e-05,
+      "loss": 1.9966,
+      "step": 320
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0326242446899414,
+      "learning_rate": 9.916834677419356e-05,
+      "loss": 2.249,
+      "step": 330
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0501596927642822,
+      "learning_rate": 9.914314516129033e-05,
+      "loss": 2.1836,
+      "step": 340
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1985058784484863,
+      "learning_rate": 9.911794354838711e-05,
+      "loss": 2.2435,
+      "step": 350
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2304984331130981,
+      "learning_rate": 9.909274193548388e-05,
+      "loss": 2.0871,
+      "step": 360
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9133213758468628,
+      "learning_rate": 9.906754032258066e-05,
+      "loss": 2.1476,
+      "step": 370
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2963943481445312,
+      "learning_rate": 9.904233870967743e-05,
+      "loss": 2.0921,
+      "step": 380
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0413427352905273,
+      "learning_rate": 9.901713709677419e-05,
+      "loss": 2.1459,
+      "step": 390
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.335339069366455,
+      "learning_rate": 9.899193548387097e-05,
+      "loss": 2.0931,
+      "step": 400
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7544111013412476,
+      "learning_rate": 9.896673387096774e-05,
+      "loss": 1.8092,
+      "step": 410
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2041220664978027,
+      "learning_rate": 9.894153225806451e-05,
+      "loss": 2.1038,
+      "step": 420
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1386065483093262,
+      "learning_rate": 9.891633064516129e-05,
+      "loss": 2.1112,
+      "step": 430
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0967787504196167,
+      "learning_rate": 9.889112903225806e-05,
+      "loss": 2.1748,
+      "step": 440
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1981559991836548,
+      "learning_rate": 9.886592741935484e-05,
+      "loss": 2.1014,
+      "step": 450
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.298099398612976,
+      "learning_rate": 9.884072580645161e-05,
+      "loss": 1.975,
+      "step": 460
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4126516580581665,
+      "learning_rate": 9.881552419354839e-05,
+      "loss": 1.9967,
+      "step": 470
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4300836324691772,
+      "learning_rate": 9.879032258064517e-05,
+      "loss": 2.1829,
+      "step": 480
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3544952869415283,
+      "learning_rate": 9.876512096774195e-05,
+      "loss": 2.0613,
+      "step": 490
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2396254539489746,
+      "learning_rate": 9.873991935483872e-05,
+      "loss": 2.205,
+      "step": 500
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7474645972251892,
+      "learning_rate": 9.87147177419355e-05,
+      "loss": 2.1699,
+      "step": 510
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1416542530059814,
+      "learning_rate": 9.868951612903227e-05,
+      "loss": 2.0208,
+      "step": 520
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3206595182418823,
+      "learning_rate": 9.866431451612905e-05,
+      "loss": 2.0859,
+      "step": 530
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7919610142707825,
+      "learning_rate": 9.863911290322582e-05,
+      "loss": 2.2325,
+      "step": 540
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3799082040786743,
+      "learning_rate": 9.861391129032258e-05,
+      "loss": 2.0705,
+      "step": 550
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2251194715499878,
+      "learning_rate": 9.858870967741935e-05,
+      "loss": 1.9726,
+      "step": 560
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8427766561508179,
+      "learning_rate": 9.856350806451613e-05,
+      "loss": 2.1134,
+      "step": 570
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3289265632629395,
+      "learning_rate": 9.85383064516129e-05,
+      "loss": 2.0659,
+      "step": 580
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0810966491699219,
+      "learning_rate": 9.851310483870968e-05,
+      "loss": 2.3646,
+      "step": 590
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3954964876174927,
+      "learning_rate": 9.848790322580645e-05,
+      "loss": 2.1128,
+      "step": 600
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.373162031173706,
+      "learning_rate": 9.846270161290323e-05,
+      "loss": 2.0707,
+      "step": 610
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9933974146842957,
+      "learning_rate": 9.84375e-05,
+      "loss": 2.1838,
+      "step": 620
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.936877965927124,
+      "learning_rate": 9.841229838709678e-05,
+      "loss": 2.0266,
+      "step": 630
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.805779218673706,
+      "learning_rate": 9.838709677419355e-05,
+      "loss": 2.0795,
+      "step": 640
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.67342209815979,
+      "learning_rate": 9.836189516129032e-05,
+      "loss": 2.1795,
+      "step": 650
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2284598350524902,
+      "learning_rate": 9.83366935483871e-05,
+      "loss": 2.2758,
+      "step": 660
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4808515310287476,
+      "learning_rate": 9.831149193548389e-05,
+      "loss": 2.0776,
+      "step": 670
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.025588870048523,
+      "learning_rate": 9.828629032258066e-05,
+      "loss": 2.1182,
+      "step": 680
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0305713415145874,
+      "learning_rate": 9.826108870967743e-05,
+      "loss": 2.1397,
+      "step": 690
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7589659690856934,
+      "learning_rate": 9.82358870967742e-05,
+      "loss": 2.1484,
+      "step": 700
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8690206408500671,
+      "learning_rate": 9.821068548387097e-05,
+      "loss": 1.8878,
+      "step": 710
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7498456239700317,
+      "learning_rate": 9.818548387096774e-05,
+      "loss": 2.1685,
+      "step": 720
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.124569892883301,
+      "learning_rate": 9.816028225806452e-05,
+      "loss": 2.0801,
+      "step": 730
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4434984922409058,
+      "learning_rate": 9.813508064516129e-05,
+      "loss": 2.1982,
+      "step": 740
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0860278606414795,
+      "learning_rate": 9.810987903225807e-05,
+      "loss": 2.0531,
+      "step": 750
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0510542392730713,
+      "learning_rate": 9.808467741935484e-05,
+      "loss": 2.0036,
+      "step": 760
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2764114141464233,
+      "learning_rate": 9.805947580645162e-05,
+      "loss": 1.9659,
+      "step": 770
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7157464027404785,
+      "learning_rate": 9.803427419354839e-05,
+      "loss": 2.1735,
+      "step": 780
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4821405410766602,
+      "learning_rate": 9.800907258064516e-05,
+      "loss": 2.0564,
+      "step": 790
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0212304592132568,
+      "learning_rate": 9.798387096774194e-05,
+      "loss": 2.0511,
+      "step": 800
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.289776086807251,
+      "learning_rate": 9.795866935483871e-05,
+      "loss": 2.2566,
+      "step": 810
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.856117844581604,
+      "learning_rate": 9.793346774193549e-05,
+      "loss": 2.0631,
+      "step": 820
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1942377090454102,
+      "learning_rate": 9.790826612903226e-05,
+      "loss": 1.9125,
+      "step": 830
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7468297481536865,
+      "learning_rate": 9.788306451612904e-05,
+      "loss": 2.0004,
+      "step": 840
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9262670278549194,
+      "learning_rate": 9.785786290322581e-05,
+      "loss": 2.1075,
+      "step": 850
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2341738939285278,
+      "learning_rate": 9.783266129032258e-05,
+      "loss": 2.0004,
+      "step": 860
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6439565420150757,
+      "learning_rate": 9.780745967741936e-05,
+      "loss": 1.9764,
+      "step": 870
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.2440450191497803,
+      "learning_rate": 9.778225806451613e-05,
+      "loss": 1.9226,
+      "step": 880
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.504655122756958,
+      "learning_rate": 9.775705645161291e-05,
+      "loss": 1.9701,
+      "step": 890
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.591951847076416,
+      "learning_rate": 9.773185483870968e-05,
+      "loss": 1.8211,
+      "step": 900
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0897972583770752,
+      "learning_rate": 9.770665322580646e-05,
+      "loss": 1.9552,
+      "step": 910
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8542840480804443,
+      "learning_rate": 9.768145161290323e-05,
+      "loss": 2.2803,
+      "step": 920
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1111481189727783,
+      "learning_rate": 9.765625e-05,
+      "loss": 1.9217,
+      "step": 930
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9833799600601196,
+      "learning_rate": 9.763104838709678e-05,
+      "loss": 2.1397,
+      "step": 940
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4057141542434692,
+      "learning_rate": 9.760584677419355e-05,
+      "loss": 2.0363,
+      "step": 950
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.162250518798828,
+      "learning_rate": 9.758064516129033e-05,
+      "loss": 1.8092,
+      "step": 960
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5231125354766846,
+      "learning_rate": 9.75554435483871e-05,
+      "loss": 1.964,
+      "step": 970
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2031828165054321,
+      "learning_rate": 9.753024193548388e-05,
+      "loss": 2.0403,
+      "step": 980
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.122621774673462,
+      "learning_rate": 9.750504032258065e-05,
+      "loss": 2.053,
+      "step": 990
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6828994750976562,
+      "learning_rate": 9.747983870967743e-05,
+      "loss": 1.9346,
+      "step": 1000
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6261448860168457,
+      "learning_rate": 9.745463709677419e-05,
+      "loss": 1.9946,
+      "step": 1010
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.051544427871704,
+      "learning_rate": 9.742943548387096e-05,
+      "loss": 2.039,
+      "step": 1020
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8216544389724731,
+      "learning_rate": 9.740423387096775e-05,
+      "loss": 2.0999,
+      "step": 1030
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5909929275512695,
+      "learning_rate": 9.737903225806452e-05,
+      "loss": 1.8428,
+      "step": 1040
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3343147039413452,
+      "learning_rate": 9.73538306451613e-05,
+      "loss": 1.937,
+      "step": 1050
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.081936001777649,
+      "learning_rate": 9.732862903225807e-05,
+      "loss": 2.0065,
+      "step": 1060
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.321244239807129,
+      "learning_rate": 9.730342741935485e-05,
+      "loss": 2.0257,
+      "step": 1070
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0805892944335938,
+      "learning_rate": 9.727822580645162e-05,
+      "loss": 2.0405,
+      "step": 1080
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.165697455406189,
+      "learning_rate": 9.72530241935484e-05,
+      "loss": 1.9575,
+      "step": 1090
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6714112758636475,
+      "learning_rate": 9.722782258064517e-05,
+      "loss": 1.8592,
+      "step": 1100
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8794459700584412,
+      "learning_rate": 9.720262096774194e-05,
+      "loss": 2.0264,
+      "step": 1110
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1143604516983032,
+      "learning_rate": 9.717741935483872e-05,
+      "loss": 2.0545,
+      "step": 1120
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5021623373031616,
+      "learning_rate": 9.715221774193549e-05,
+      "loss": 2.138,
+      "step": 1130
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2431867122650146,
+      "learning_rate": 9.712701612903227e-05,
+      "loss": 1.7612,
+      "step": 1140
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3302487134933472,
+      "learning_rate": 9.710181451612904e-05,
+      "loss": 1.9292,
+      "step": 1150
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5073152780532837,
+      "learning_rate": 9.707661290322582e-05,
+      "loss": 1.9894,
+      "step": 1160
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9150564074516296,
+      "learning_rate": 9.705141129032258e-05,
+      "loss": 2.1304,
+      "step": 1170
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7501333355903625,
+      "learning_rate": 9.702620967741935e-05,
+      "loss": 1.9473,
+      "step": 1180
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.636514663696289,
+      "learning_rate": 9.700100806451612e-05,
+      "loss": 1.915,
+      "step": 1190
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9736351370811462,
+      "learning_rate": 9.69758064516129e-05,
+      "loss": 1.9465,
+      "step": 1200
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4614986181259155,
+      "learning_rate": 9.695060483870967e-05,
+      "loss": 1.9349,
+      "step": 1210
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7856169939041138,
+      "learning_rate": 9.692540322580646e-05,
+      "loss": 2.1242,
+      "step": 1220
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0289955139160156,
+      "learning_rate": 9.690020161290324e-05,
+      "loss": 2.0543,
+      "step": 1230
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9941620826721191,
+      "learning_rate": 9.687500000000001e-05,
+      "loss": 1.7832,
+      "step": 1240
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5732817649841309,
+      "learning_rate": 9.684979838709678e-05,
+      "loss": 1.9947,
+      "step": 1250
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9455333948135376,
+      "learning_rate": 9.682459677419356e-05,
+      "loss": 2.1206,
+      "step": 1260
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.840183675289154,
+      "learning_rate": 9.679939516129033e-05,
+      "loss": 2.1025,
+      "step": 1270
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1012920141220093,
+      "learning_rate": 9.677419354838711e-05,
+      "loss": 2.1329,
+      "step": 1280
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.485025405883789,
+      "learning_rate": 9.674899193548388e-05,
+      "loss": 2.0445,
+      "step": 1290
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7725688219070435,
+      "learning_rate": 9.672379032258066e-05,
+      "loss": 1.9816,
+      "step": 1300
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7952070832252502,
+      "learning_rate": 9.669858870967743e-05,
+      "loss": 1.9669,
+      "step": 1310
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6139402389526367,
+      "learning_rate": 9.667338709677419e-05,
+      "loss": 1.8617,
+      "step": 1320
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7932485342025757,
+      "learning_rate": 9.664818548387097e-05,
+      "loss": 2.0453,
+      "step": 1330
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.641073226928711,
+      "learning_rate": 9.662298387096774e-05,
+      "loss": 2.2676,
+      "step": 1340
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7598496675491333,
+      "learning_rate": 9.659778225806451e-05,
+      "loss": 2.0448,
+      "step": 1350
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2583099603652954,
+      "learning_rate": 9.657258064516129e-05,
+      "loss": 2.1459,
+      "step": 1360
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2672468423843384,
+      "learning_rate": 9.654737903225806e-05,
+      "loss": 1.9308,
+      "step": 1370
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3083672523498535,
+      "learning_rate": 9.652217741935484e-05,
+      "loss": 1.8798,
+      "step": 1380
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.641868233680725,
+      "learning_rate": 9.649697580645161e-05,
+      "loss": 1.9355,
+      "step": 1390
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.881644070148468,
+      "learning_rate": 9.647177419354839e-05,
+      "loss": 1.99,
+      "step": 1400
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5086069107055664,
+      "learning_rate": 9.644657258064516e-05,
+      "loss": 1.843,
+      "step": 1410
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4167226552963257,
+      "learning_rate": 9.642137096774195e-05,
+      "loss": 2.1338,
+      "step": 1420
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4532346725463867,
+      "learning_rate": 9.639616935483872e-05,
+      "loss": 1.8731,
+      "step": 1430
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9580788016319275,
+      "learning_rate": 9.63709677419355e-05,
+      "loss": 1.8509,
+      "step": 1440
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3719030618667603,
+      "learning_rate": 9.634576612903227e-05,
+      "loss": 2.1377,
+      "step": 1450
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7830965518951416,
+      "learning_rate": 9.632056451612905e-05,
+      "loss": 2.2908,
+      "step": 1460
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.397725224494934,
+      "learning_rate": 9.629536290322582e-05,
+      "loss": 2.0565,
+      "step": 1470
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.081475853919983,
+      "learning_rate": 9.627016129032258e-05,
+      "loss": 2.0503,
+      "step": 1480
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9584048986434937,
+      "learning_rate": 9.624495967741936e-05,
+      "loss": 1.8166,
+      "step": 1490
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8202422857284546,
+      "learning_rate": 9.621975806451613e-05,
+      "loss": 1.9767,
+      "step": 1500
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7263189554214478,
+      "learning_rate": 9.61945564516129e-05,
+      "loss": 2.0861,
+      "step": 1510
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9306368827819824,
+      "learning_rate": 9.616935483870968e-05,
+      "loss": 1.9182,
+      "step": 1520
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2647794485092163,
+      "learning_rate": 9.614415322580645e-05,
+      "loss": 1.931,
+      "step": 1530
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5704960823059082,
+      "learning_rate": 9.611895161290323e-05,
+      "loss": 1.8239,
+      "step": 1540
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.09844172000885,
+      "learning_rate": 9.609375e-05,
+      "loss": 1.926,
+      "step": 1550
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8096204996109009,
+      "learning_rate": 9.606854838709678e-05,
+      "loss": 2.3091,
+      "step": 1560
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.494329810142517,
+      "learning_rate": 9.604334677419355e-05,
+      "loss": 2.0375,
+      "step": 1570
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7294707298278809,
+      "learning_rate": 9.601814516129032e-05,
+      "loss": 2.201,
+      "step": 1580
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0612577199935913,
+      "learning_rate": 9.59929435483871e-05,
+      "loss": 2.1384,
+      "step": 1590
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9931353330612183,
+      "learning_rate": 9.596774193548387e-05,
+      "loss": 1.9775,
+      "step": 1600
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3362724781036377,
+      "learning_rate": 9.594254032258065e-05,
+      "loss": 2.1661,
+      "step": 1610
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.3799688816070557,
+      "learning_rate": 9.591733870967744e-05,
+      "loss": 2.191,
+      "step": 1620
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9034447073936462,
+      "learning_rate": 9.58921370967742e-05,
+      "loss": 1.9704,
+      "step": 1630
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.2976465225219727,
+      "learning_rate": 9.586693548387097e-05,
+      "loss": 2.2311,
+      "step": 1640
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8832471370697021,
+      "learning_rate": 9.584173387096774e-05,
+      "loss": 2.0388,
+      "step": 1650
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4275864362716675,
+      "learning_rate": 9.581653225806452e-05,
+      "loss": 1.9591,
+      "step": 1660
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7437922954559326,
+      "learning_rate": 9.57913306451613e-05,
+      "loss": 1.8052,
+      "step": 1670
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3830605745315552,
+      "learning_rate": 9.576612903225807e-05,
+      "loss": 1.8488,
+      "step": 1680
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1911534070968628,
+      "learning_rate": 9.574092741935484e-05,
+      "loss": 2.1724,
+      "step": 1690
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3713386058807373,
+      "learning_rate": 9.571572580645162e-05,
+      "loss": 1.8695,
+      "step": 1700
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0680556297302246,
+      "learning_rate": 9.569052419354839e-05,
+      "loss": 2.0631,
+      "step": 1710
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.7318758964538574,
+      "learning_rate": 9.566532258064517e-05,
+      "loss": 1.9194,
+      "step": 1720
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5713481903076172,
+      "learning_rate": 9.564012096774194e-05,
+      "loss": 1.8589,
+      "step": 1730
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0396041870117188,
+      "learning_rate": 9.561491935483871e-05,
+      "loss": 2.2087,
+      "step": 1740
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9374070763587952,
+      "learning_rate": 9.558971774193549e-05,
+      "loss": 2.0106,
+      "step": 1750
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0840810537338257,
+      "learning_rate": 9.556451612903226e-05,
+      "loss": 1.9689,
+      "step": 1760
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6653531193733215,
+      "learning_rate": 9.553931451612904e-05,
+      "loss": 1.9084,
+      "step": 1770
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.270177125930786,
+      "learning_rate": 9.551411290322581e-05,
+      "loss": 2.0686,
+      "step": 1780
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3076845407485962,
+      "learning_rate": 9.548891129032259e-05,
+      "loss": 1.893,
+      "step": 1790
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3934273719787598,
+      "learning_rate": 9.546370967741936e-05,
+      "loss": 2.028,
+      "step": 1800
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0181070566177368,
+      "learning_rate": 9.543850806451613e-05,
+      "loss": 2.1116,
+      "step": 1810
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4238017797470093,
+      "learning_rate": 9.541330645161291e-05,
+      "loss": 2.0033,
+      "step": 1820
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2628740072250366,
+      "learning_rate": 9.538810483870968e-05,
+      "loss": 2.0932,
+      "step": 1830
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0557096004486084,
+      "learning_rate": 9.536290322580646e-05,
+      "loss": 1.7102,
+      "step": 1840
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.2740368843078613,
+      "learning_rate": 9.533770161290323e-05,
+      "loss": 1.9909,
+      "step": 1850
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.570090889930725,
+      "learning_rate": 9.53125e-05,
+      "loss": 1.9888,
+      "step": 1860
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2309668064117432,
+      "learning_rate": 9.528729838709678e-05,
+      "loss": 1.9996,
+      "step": 1870
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1102397441864014,
+      "learning_rate": 9.526209677419355e-05,
+      "loss": 2.0949,
+      "step": 1880
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.611424207687378,
+      "learning_rate": 9.523689516129033e-05,
+      "loss": 1.9973,
+      "step": 1890
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5840996503829956,
+      "learning_rate": 9.52116935483871e-05,
+      "loss": 1.9113,
+      "step": 1900
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8727772235870361,
+      "learning_rate": 9.518649193548388e-05,
+      "loss": 2.0398,
+      "step": 1910
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8769510984420776,
+      "learning_rate": 9.516129032258065e-05,
+      "loss": 1.9243,
+      "step": 1920
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0773354768753052,
+      "learning_rate": 9.513608870967743e-05,
+      "loss": 2.0529,
+      "step": 1930
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.721339225769043,
+      "learning_rate": 9.511088709677419e-05,
+      "loss": 1.8973,
+      "step": 1940
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.265521764755249,
+      "learning_rate": 9.508568548387096e-05,
+      "loss": 2.0024,
+      "step": 1950
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6861649751663208,
+      "learning_rate": 9.506048387096774e-05,
+      "loss": 2.0153,
+      "step": 1960
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5646659135818481,
+      "learning_rate": 9.503528225806452e-05,
+      "loss": 2.0346,
+      "step": 1970
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.5712592601776123,
+      "learning_rate": 9.50100806451613e-05,
+      "loss": 1.9415,
+      "step": 1980
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0997393131256104,
+      "learning_rate": 9.498487903225807e-05,
+      "loss": 1.8441,
+      "step": 1990
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5155184268951416,
+      "learning_rate": 9.495967741935485e-05,
+      "loss": 1.8641,
+      "step": 2000
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8395708799362183,
+      "learning_rate": 9.493447580645162e-05,
+      "loss": 1.7802,
+      "step": 2010
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4492772817611694,
+      "learning_rate": 9.49092741935484e-05,
+      "loss": 1.93,
+      "step": 2020
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.413654088973999,
+      "learning_rate": 9.488407258064517e-05,
+      "loss": 2.1289,
+      "step": 2030
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3992499113082886,
+      "learning_rate": 9.485887096774194e-05,
+      "loss": 1.8192,
+      "step": 2040
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0796810388565063,
+      "learning_rate": 9.483366935483872e-05,
+      "loss": 2.0179,
+      "step": 2050
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1251513957977295,
+      "learning_rate": 9.480846774193549e-05,
+      "loss": 1.83,
+      "step": 2060
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.487831473350525,
+      "learning_rate": 9.478326612903227e-05,
+      "loss": 2.0104,
+      "step": 2070
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7137647867202759,
+      "learning_rate": 9.475806451612904e-05,
+      "loss": 1.8731,
+      "step": 2080
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.010001540184021,
+      "learning_rate": 9.473286290322582e-05,
+      "loss": 2.0662,
+      "step": 2090
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.083350419998169,
+      "learning_rate": 9.470766129032258e-05,
+      "loss": 2.1006,
+      "step": 2100
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0546329021453857,
+      "learning_rate": 9.468245967741935e-05,
+      "loss": 2.0595,
+      "step": 2110
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.033692479133606,
+      "learning_rate": 9.465725806451613e-05,
+      "loss": 2.0402,
+      "step": 2120
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.135054588317871,
+      "learning_rate": 9.46320564516129e-05,
+      "loss": 1.9558,
+      "step": 2130
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8258410692214966,
+      "learning_rate": 9.460685483870967e-05,
+      "loss": 1.949,
+      "step": 2140
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6336489915847778,
+      "learning_rate": 9.458165322580645e-05,
+      "loss": 2.0865,
+      "step": 2150
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2403907775878906,
+      "learning_rate": 9.455645161290322e-05,
+      "loss": 2.0827,
+      "step": 2160
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1931679248809814,
+      "learning_rate": 9.453125000000001e-05,
+      "loss": 2.0402,
+      "step": 2170
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.8094074726104736,
+      "learning_rate": 9.450604838709679e-05,
+      "loss": 2.0069,
+      "step": 2180
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0455503463745117,
+      "learning_rate": 9.448084677419356e-05,
+      "loss": 1.9105,
+      "step": 2190
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9511374831199646,
+      "learning_rate": 9.445564516129033e-05,
+      "loss": 1.9017,
+      "step": 2200
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2926204204559326,
+      "learning_rate": 9.443044354838711e-05,
+      "loss": 1.8684,
+      "step": 2210
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4258010387420654,
+      "learning_rate": 9.440524193548388e-05,
+      "loss": 1.9425,
+      "step": 2220
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.978868842124939,
+      "learning_rate": 9.438004032258066e-05,
+      "loss": 1.9839,
+      "step": 2230
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0230255126953125,
+      "learning_rate": 9.435483870967743e-05,
+      "loss": 1.9045,
+      "step": 2240
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9253028035163879,
+      "learning_rate": 9.432963709677419e-05,
+      "loss": 1.9735,
+      "step": 2250
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2342355251312256,
+      "learning_rate": 9.430443548387097e-05,
+      "loss": 1.9939,
+      "step": 2260
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3178222179412842,
+      "learning_rate": 9.427923387096774e-05,
+      "loss": 1.9741,
+      "step": 2270
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4835388660430908,
+      "learning_rate": 9.425403225806452e-05,
+      "loss": 2.0463,
+      "step": 2280
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.4622998237609863,
+      "learning_rate": 9.422883064516129e-05,
+      "loss": 2.0309,
+      "step": 2290
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8435159921646118,
+      "learning_rate": 9.420362903225806e-05,
+      "loss": 1.8738,
+      "step": 2300
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0138598680496216,
+      "learning_rate": 9.417842741935484e-05,
+      "loss": 1.9334,
+      "step": 2310
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3110580444335938,
+      "learning_rate": 9.415322580645161e-05,
+      "loss": 1.7926,
+      "step": 2320
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8634729385375977,
+      "learning_rate": 9.412802419354839e-05,
+      "loss": 1.9185,
+      "step": 2330
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6818907260894775,
+      "learning_rate": 9.410282258064516e-05,
+      "loss": 2.0777,
+      "step": 2340
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5213080644607544,
+      "learning_rate": 9.407762096774194e-05,
+      "loss": 2.1321,
+      "step": 2350
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9287450313568115,
+      "learning_rate": 9.405241935483872e-05,
+      "loss": 1.9881,
+      "step": 2360
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5543783903121948,
+      "learning_rate": 9.40272177419355e-05,
+      "loss": 2.1611,
+      "step": 2370
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0999093055725098,
+      "learning_rate": 9.400201612903227e-05,
+      "loss": 1.838,
+      "step": 2380
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6901296377182007,
+      "learning_rate": 9.397681451612905e-05,
+      "loss": 1.9891,
+      "step": 2390
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5015904903411865,
+      "learning_rate": 9.395161290322582e-05,
+      "loss": 2.1533,
+      "step": 2400
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.668653130531311,
+      "learning_rate": 9.392641129032258e-05,
+      "loss": 1.9396,
+      "step": 2410
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3372501134872437,
+      "learning_rate": 9.390120967741936e-05,
+      "loss": 1.9453,
+      "step": 2420
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3987207412719727,
+      "learning_rate": 9.387600806451613e-05,
+      "loss": 1.9633,
+      "step": 2430
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.126643180847168,
+      "learning_rate": 9.38508064516129e-05,
+      "loss": 1.9502,
+      "step": 2440
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4671810865402222,
+      "learning_rate": 9.382560483870968e-05,
+      "loss": 1.9906,
+      "step": 2450
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9458364248275757,
+      "learning_rate": 9.380040322580645e-05,
+      "loss": 1.9323,
+      "step": 2460
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.135588526725769,
+      "learning_rate": 9.377520161290323e-05,
+      "loss": 2.0346,
+      "step": 2470
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0130577087402344,
+      "learning_rate": 9.375e-05,
+      "loss": 1.9115,
+      "step": 2480
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2623790502548218,
+      "learning_rate": 9.372479838709678e-05,
+      "loss": 2.0228,
+      "step": 2490
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.611956000328064,
+      "learning_rate": 9.369959677419355e-05,
+      "loss": 1.9709,
+      "step": 2500
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7934135794639587,
+      "learning_rate": 9.367439516129032e-05,
+      "loss": 2.0073,
+      "step": 2510
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9499655961990356,
+      "learning_rate": 9.36491935483871e-05,
+      "loss": 1.8493,
+      "step": 2520
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6667944192886353,
+      "learning_rate": 9.362399193548387e-05,
+      "loss": 1.8006,
+      "step": 2530
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0717980861663818,
+      "learning_rate": 9.359879032258065e-05,
+      "loss": 1.8716,
+      "step": 2540
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2375351190567017,
+      "learning_rate": 9.357358870967742e-05,
+      "loss": 1.9217,
+      "step": 2550
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.16543447971344,
+      "learning_rate": 9.35483870967742e-05,
+      "loss": 1.8075,
+      "step": 2560
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6690412759780884,
+      "learning_rate": 9.352318548387097e-05,
+      "loss": 1.7832,
+      "step": 2570
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.559948444366455,
+      "learning_rate": 9.349798387096775e-05,
+      "loss": 1.7735,
+      "step": 2580
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8290411233901978,
+      "learning_rate": 9.347278225806452e-05,
+      "loss": 2.0396,
+      "step": 2590
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3372671604156494,
+      "learning_rate": 9.34475806451613e-05,
+      "loss": 1.8925,
+      "step": 2600
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.141345500946045,
+      "learning_rate": 9.342237903225807e-05,
+      "loss": 2.0329,
+      "step": 2610
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8727923035621643,
+      "learning_rate": 9.339717741935484e-05,
+      "loss": 2.0305,
+      "step": 2620
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2888758182525635,
+      "learning_rate": 9.337197580645162e-05,
+      "loss": 2.0539,
+      "step": 2630
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.047920823097229,
+      "learning_rate": 9.334677419354839e-05,
+      "loss": 2.1404,
+      "step": 2640
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6245843172073364,
+      "learning_rate": 9.332157258064517e-05,
+      "loss": 1.8032,
+      "step": 2650
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7825778126716614,
+      "learning_rate": 9.329637096774194e-05,
+      "loss": 2.1102,
+      "step": 2660
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0130205154418945,
+      "learning_rate": 9.327116935483871e-05,
+      "loss": 2.0637,
+      "step": 2670
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.123182535171509,
+      "learning_rate": 9.324596774193549e-05,
+      "loss": 1.8922,
+      "step": 2680
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.6627249717712402,
+      "learning_rate": 9.322076612903226e-05,
+      "loss": 1.8472,
+      "step": 2690
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8047929406166077,
+      "learning_rate": 9.319556451612904e-05,
+      "loss": 1.7658,
+      "step": 2700
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2637733221054077,
+      "learning_rate": 9.317036290322581e-05,
+      "loss": 2.1191,
+      "step": 2710
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9870107173919678,
+      "learning_rate": 9.314516129032259e-05,
+      "loss": 1.701,
+      "step": 2720
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7081884145736694,
+      "learning_rate": 9.311995967741936e-05,
+      "loss": 1.881,
+      "step": 2730
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3320205211639404,
+      "learning_rate": 9.309475806451613e-05,
+      "loss": 1.7728,
+      "step": 2740
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2349406480789185,
+      "learning_rate": 9.306955645161291e-05,
+      "loss": 1.9862,
+      "step": 2750
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1466513872146606,
+      "learning_rate": 9.304435483870968e-05,
+      "loss": 2.0161,
+      "step": 2760
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2989059686660767,
+      "learning_rate": 9.301915322580646e-05,
+      "loss": 1.8571,
+      "step": 2770
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2982863187789917,
+      "learning_rate": 9.299395161290323e-05,
+      "loss": 2.062,
+      "step": 2780
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4981862306594849,
+      "learning_rate": 9.296875e-05,
+      "loss": 2.0004,
+      "step": 2790
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5772234201431274,
+      "learning_rate": 9.294354838709678e-05,
+      "loss": 1.7902,
+      "step": 2800
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.4681856632232666,
+      "learning_rate": 9.291834677419356e-05,
+      "loss": 2.064,
+      "step": 2810
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9098520874977112,
+      "learning_rate": 9.289314516129033e-05,
+      "loss": 1.9252,
+      "step": 2820
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2758041620254517,
+      "learning_rate": 9.28679435483871e-05,
+      "loss": 1.8384,
+      "step": 2830
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6394200325012207,
+      "learning_rate": 9.284274193548388e-05,
+      "loss": 1.9675,
+      "step": 2840
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6323776245117188,
+      "learning_rate": 9.281754032258065e-05,
+      "loss": 1.9647,
+      "step": 2850
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0769028663635254,
+      "learning_rate": 9.279233870967743e-05,
+      "loss": 1.9514,
+      "step": 2860
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0145025253295898,
+      "learning_rate": 9.276713709677419e-05,
+      "loss": 1.6596,
+      "step": 2870
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.160710334777832,
+      "learning_rate": 9.274193548387096e-05,
+      "loss": 2.0619,
+      "step": 2880
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5925042629241943,
+      "learning_rate": 9.271673387096774e-05,
+      "loss": 1.8917,
+      "step": 2890
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.759544849395752,
+      "learning_rate": 9.269153225806451e-05,
+      "loss": 2.1042,
+      "step": 2900
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1082566976547241,
+      "learning_rate": 9.26663306451613e-05,
+      "loss": 2.0623,
+      "step": 2910
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8014003038406372,
+      "learning_rate": 9.264112903225807e-05,
+      "loss": 1.9197,
+      "step": 2920
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7241004109382629,
+      "learning_rate": 9.261592741935485e-05,
+      "loss": 2.0253,
+      "step": 2930
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2464666366577148,
+      "learning_rate": 9.259072580645162e-05,
+      "loss": 2.044,
+      "step": 2940
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.278713583946228,
+      "learning_rate": 9.25655241935484e-05,
+      "loss": 1.99,
+      "step": 2950
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6077810525894165,
+      "learning_rate": 9.254032258064517e-05,
+      "loss": 1.8686,
+      "step": 2960
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1507235765457153,
+      "learning_rate": 9.251512096774194e-05,
+      "loss": 1.9236,
+      "step": 2970
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9427613615989685,
+      "learning_rate": 9.248991935483872e-05,
+      "loss": 2.0172,
+      "step": 2980
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3288710117340088,
+      "learning_rate": 9.24647177419355e-05,
+      "loss": 2.0103,
+      "step": 2990
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.723928689956665,
+      "learning_rate": 9.243951612903227e-05,
+      "loss": 1.8542,
+      "step": 3000
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6426094770431519,
+      "learning_rate": 9.241431451612904e-05,
+      "loss": 1.8609,
+      "step": 3010
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.363667368888855,
+      "learning_rate": 9.238911290322582e-05,
+      "loss": 1.6826,
+      "step": 3020
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1912142038345337,
+      "learning_rate": 9.236391129032258e-05,
+      "loss": 1.861,
+      "step": 3030
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8758140206336975,
+      "learning_rate": 9.233870967741935e-05,
+      "loss": 1.8161,
+      "step": 3040
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0439066886901855,
+      "learning_rate": 9.231350806451613e-05,
+      "loss": 1.7434,
+      "step": 3050
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9993200898170471,
+      "learning_rate": 9.22883064516129e-05,
+      "loss": 1.9037,
+      "step": 3060
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1631320714950562,
+      "learning_rate": 9.226310483870967e-05,
+      "loss": 1.9718,
+      "step": 3070
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3915998935699463,
+      "learning_rate": 9.223790322580645e-05,
+      "loss": 1.9373,
+      "step": 3080
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7717387676239014,
+      "learning_rate": 9.221270161290322e-05,
+      "loss": 1.8619,
+      "step": 3090
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.678252935409546,
+      "learning_rate": 9.21875e-05,
+      "loss": 2.0865,
+      "step": 3100
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8635882139205933,
+      "learning_rate": 9.216229838709679e-05,
+      "loss": 1.9945,
+      "step": 3110
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6390001773834229,
+      "learning_rate": 9.213709677419356e-05,
+      "loss": 2.092,
+      "step": 3120
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1088507175445557,
+      "learning_rate": 9.211189516129033e-05,
+      "loss": 1.9199,
+      "step": 3130
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7361661195755005,
+      "learning_rate": 9.208669354838711e-05,
+      "loss": 1.8794,
+      "step": 3140
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8918304443359375,
+      "learning_rate": 9.206149193548388e-05,
+      "loss": 1.9955,
+      "step": 3150
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1529821157455444,
+      "learning_rate": 9.203629032258066e-05,
+      "loss": 1.9322,
+      "step": 3160
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.419748067855835,
+      "learning_rate": 9.201108870967743e-05,
+      "loss": 1.9375,
+      "step": 3170
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8855389952659607,
+      "learning_rate": 9.198588709677419e-05,
+      "loss": 1.9504,
+      "step": 3180
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1205099821090698,
+      "learning_rate": 9.196068548387097e-05,
+      "loss": 1.8233,
+      "step": 3190
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9326832890510559,
+      "learning_rate": 9.193548387096774e-05,
+      "loss": 2.0541,
+      "step": 3200
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1365574598312378,
+      "learning_rate": 9.191028225806452e-05,
+      "loss": 1.9028,
+      "step": 3210
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6870282888412476,
+      "learning_rate": 9.188508064516129e-05,
+      "loss": 1.8629,
+      "step": 3220
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.057867407798767,
+      "learning_rate": 9.185987903225806e-05,
+      "loss": 1.966,
+      "step": 3230
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1084156036376953,
+      "learning_rate": 9.183467741935484e-05,
+      "loss": 1.825,
+      "step": 3240
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9650135040283203,
+      "learning_rate": 9.180947580645161e-05,
+      "loss": 2.0376,
+      "step": 3250
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5941529273986816,
+      "learning_rate": 9.178427419354839e-05,
+      "loss": 1.8725,
+      "step": 3260
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5933493375778198,
+      "learning_rate": 9.175907258064516e-05,
+      "loss": 1.978,
+      "step": 3270
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9752015471458435,
+      "learning_rate": 9.173387096774194e-05,
+      "loss": 1.9568,
+      "step": 3280
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9967694282531738,
+      "learning_rate": 9.170866935483871e-05,
+      "loss": 1.9482,
+      "step": 3290
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.017528772354126,
+      "learning_rate": 9.168346774193548e-05,
+      "loss": 2.0824,
+      "step": 3300
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.009690523147583,
+      "learning_rate": 9.165826612903227e-05,
+      "loss": 2.0292,
+      "step": 3310
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6556392312049866,
+      "learning_rate": 9.163306451612905e-05,
+      "loss": 1.9892,
+      "step": 3320
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.517575979232788,
+      "learning_rate": 9.160786290322582e-05,
+      "loss": 2.1448,
+      "step": 3330
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.575211763381958,
+      "learning_rate": 9.158266129032258e-05,
+      "loss": 1.8958,
+      "step": 3340
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8941549062728882,
+      "learning_rate": 9.155745967741936e-05,
+      "loss": 1.9171,
+      "step": 3350
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8913719654083252,
+      "learning_rate": 9.153225806451613e-05,
+      "loss": 1.9093,
+      "step": 3360
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3766728639602661,
+      "learning_rate": 9.15070564516129e-05,
+      "loss": 1.9553,
+      "step": 3370
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1442893743515015,
+      "learning_rate": 9.148185483870968e-05,
+      "loss": 2.0305,
+      "step": 3380
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.206863284111023,
+      "learning_rate": 9.145665322580645e-05,
+      "loss": 1.8773,
+      "step": 3390
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4691145420074463,
+      "learning_rate": 9.143145161290323e-05,
+      "loss": 1.7374,
+      "step": 3400
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.603576898574829,
+      "learning_rate": 9.140625e-05,
+      "loss": 1.922,
+      "step": 3410
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6233885288238525,
+      "learning_rate": 9.138104838709678e-05,
+      "loss": 1.8534,
+      "step": 3420
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1434792280197144,
+      "learning_rate": 9.135584677419355e-05,
+      "loss": 1.8824,
+      "step": 3430
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2980107069015503,
+      "learning_rate": 9.133064516129033e-05,
+      "loss": 2.0426,
+      "step": 3440
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.502358317375183,
+      "learning_rate": 9.13054435483871e-05,
+      "loss": 1.796,
+      "step": 3450
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3557907342910767,
+      "learning_rate": 9.128024193548387e-05,
+      "loss": 1.9192,
+      "step": 3460
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1002049446105957,
+      "learning_rate": 9.125504032258065e-05,
+      "loss": 1.8675,
+      "step": 3470
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9018680453300476,
+      "learning_rate": 9.122983870967742e-05,
+      "loss": 2.0888,
+      "step": 3480
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8970044255256653,
+      "learning_rate": 9.12046370967742e-05,
+      "loss": 1.9999,
+      "step": 3490
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9390895366668701,
+      "learning_rate": 9.117943548387097e-05,
+      "loss": 2.0021,
+      "step": 3500
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5169790983200073,
+      "learning_rate": 9.115423387096775e-05,
+      "loss": 1.8353,
+      "step": 3510
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0707337856292725,
+      "learning_rate": 9.112903225806452e-05,
+      "loss": 1.7763,
+      "step": 3520
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7476022243499756,
+      "learning_rate": 9.11038306451613e-05,
+      "loss": 1.9428,
+      "step": 3530
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6692254543304443,
+      "learning_rate": 9.107862903225807e-05,
+      "loss": 1.7884,
+      "step": 3540
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.096766471862793,
+      "learning_rate": 9.105342741935484e-05,
+      "loss": 1.857,
+      "step": 3550
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.963667094707489,
+      "learning_rate": 9.102822580645162e-05,
+      "loss": 1.9079,
+      "step": 3560
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7885515689849854,
+      "learning_rate": 9.100302419354839e-05,
+      "loss": 1.9154,
+      "step": 3570
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7611230611801147,
+      "learning_rate": 9.097782258064517e-05,
+      "loss": 1.9847,
+      "step": 3580
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8401360511779785,
+      "learning_rate": 9.095262096774194e-05,
+      "loss": 2.1077,
+      "step": 3590
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.13411283493042,
+      "learning_rate": 9.092741935483872e-05,
+      "loss": 1.8097,
+      "step": 3600
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1606708765029907,
+      "learning_rate": 9.090221774193549e-05,
+      "loss": 1.8499,
+      "step": 3610
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9232423305511475,
+      "learning_rate": 9.087701612903226e-05,
+      "loss": 1.9248,
+      "step": 3620
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.2549240589141846,
+      "learning_rate": 9.085181451612904e-05,
+      "loss": 1.9745,
+      "step": 3630
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.2512686252593994,
+      "learning_rate": 9.082661290322581e-05,
+      "loss": 1.8267,
+      "step": 3640
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9882325530052185,
+      "learning_rate": 9.080141129032257e-05,
+      "loss": 1.8371,
+      "step": 3650
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1430553197860718,
+      "learning_rate": 9.077620967741936e-05,
+      "loss": 2.2198,
+      "step": 3660
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.6589090824127197,
+      "learning_rate": 9.075100806451614e-05,
+      "loss": 1.76,
+      "step": 3670
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0968700647354126,
+      "learning_rate": 9.072580645161291e-05,
+      "loss": 1.9341,
+      "step": 3680
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2082455158233643,
+      "learning_rate": 9.070060483870968e-05,
+      "loss": 2.0859,
+      "step": 3690
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6459219455718994,
+      "learning_rate": 9.067540322580646e-05,
+      "loss": 1.8933,
+      "step": 3700
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2811520099639893,
+      "learning_rate": 9.065020161290323e-05,
+      "loss": 1.805,
+      "step": 3710
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.641817331314087,
+      "learning_rate": 9.062500000000001e-05,
+      "loss": 1.9314,
+      "step": 3720
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0053181648254395,
+      "learning_rate": 9.059979838709678e-05,
+      "loss": 2.0175,
+      "step": 3730
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6023327112197876,
+      "learning_rate": 9.057459677419356e-05,
+      "loss": 1.8416,
+      "step": 3740
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9793619513511658,
+      "learning_rate": 9.054939516129033e-05,
+      "loss": 2.0524,
+      "step": 3750
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3959987163543701,
+      "learning_rate": 9.05241935483871e-05,
+      "loss": 1.844,
+      "step": 3760
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0041372776031494,
+      "learning_rate": 9.049899193548388e-05,
+      "loss": 2.0173,
+      "step": 3770
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5536431074142456,
+      "learning_rate": 9.047379032258065e-05,
+      "loss": 1.8799,
+      "step": 3780
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2309901714324951,
+      "learning_rate": 9.044858870967743e-05,
+      "loss": 1.815,
+      "step": 3790
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.647364616394043,
+      "learning_rate": 9.042338709677419e-05,
+      "loss": 1.8838,
+      "step": 3800
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7798125743865967,
+      "learning_rate": 9.039818548387096e-05,
+      "loss": 1.969,
+      "step": 3810
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2773804664611816,
+      "learning_rate": 9.037298387096774e-05,
+      "loss": 1.9318,
+      "step": 3820
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8688080310821533,
+      "learning_rate": 9.034778225806451e-05,
+      "loss": 1.9387,
+      "step": 3830
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9782726764678955,
+      "learning_rate": 9.032258064516129e-05,
+      "loss": 1.9791,
+      "step": 3840
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9728130102157593,
+      "learning_rate": 9.029737903225807e-05,
+      "loss": 1.7457,
+      "step": 3850
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6487958431243896,
+      "learning_rate": 9.027217741935485e-05,
+      "loss": 1.9066,
+      "step": 3860
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.490662693977356,
+      "learning_rate": 9.024697580645162e-05,
+      "loss": 1.9811,
+      "step": 3870
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0628536939620972,
+      "learning_rate": 9.02217741935484e-05,
+      "loss": 1.8128,
+      "step": 3880
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.783426821231842,
+      "learning_rate": 9.019657258064517e-05,
+      "loss": 1.7071,
+      "step": 3890
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.6339993476867676,
+      "learning_rate": 9.017137096774195e-05,
+      "loss": 1.8273,
+      "step": 3900
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4142245054244995,
+      "learning_rate": 9.014616935483872e-05,
+      "loss": 1.9949,
+      "step": 3910
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6441869139671326,
+      "learning_rate": 9.01209677419355e-05,
+      "loss": 2.0643,
+      "step": 3920
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9136880040168762,
+      "learning_rate": 9.009576612903227e-05,
+      "loss": 1.8689,
+      "step": 3930
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3684062957763672,
+      "learning_rate": 9.007056451612904e-05,
+      "loss": 1.7314,
+      "step": 3940
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0993908643722534,
+      "learning_rate": 9.004536290322582e-05,
+      "loss": 2.1183,
+      "step": 3950
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5412412881851196,
+      "learning_rate": 9.002016129032258e-05,
+      "loss": 2.1238,
+      "step": 3960
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3627482652664185,
+      "learning_rate": 8.999495967741935e-05,
+      "loss": 1.8399,
+      "step": 3970
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.033787727355957,
+      "learning_rate": 8.996975806451613e-05,
+      "loss": 2.0027,
+      "step": 3980
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1848596334457397,
+      "learning_rate": 8.99445564516129e-05,
+      "loss": 1.9364,
+      "step": 3990
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.103233575820923,
+      "learning_rate": 8.991935483870968e-05,
+      "loss": 1.7368,
+      "step": 4000
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8747714757919312,
+      "learning_rate": 8.989415322580645e-05,
+      "loss": 1.9124,
+      "step": 4010
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8378656506538391,
+      "learning_rate": 8.986895161290322e-05,
+      "loss": 1.768,
+      "step": 4020
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.824456810951233,
+      "learning_rate": 8.984375e-05,
+      "loss": 1.8712,
+      "step": 4030
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1492997407913208,
+      "learning_rate": 8.981854838709677e-05,
+      "loss": 1.8558,
+      "step": 4040
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6846030950546265,
+      "learning_rate": 8.979334677419356e-05,
+      "loss": 1.9538,
+      "step": 4050
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0550131797790527,
+      "learning_rate": 8.976814516129034e-05,
+      "loss": 1.9989,
+      "step": 4060
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2373980283737183,
+      "learning_rate": 8.974294354838711e-05,
+      "loss": 1.798,
+      "step": 4070
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1064437627792358,
+      "learning_rate": 8.971774193548388e-05,
+      "loss": 1.8554,
+      "step": 4080
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9924767017364502,
+      "learning_rate": 8.969254032258066e-05,
+      "loss": 1.8857,
+      "step": 4090
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0961071252822876,
+      "learning_rate": 8.966733870967743e-05,
+      "loss": 2.0134,
+      "step": 4100
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9283311367034912,
+      "learning_rate": 8.96421370967742e-05,
+      "loss": 1.9583,
+      "step": 4110
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5703861713409424,
+      "learning_rate": 8.961693548387097e-05,
+      "loss": 2.04,
+      "step": 4120
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1366268396377563,
+      "learning_rate": 8.959173387096774e-05,
+      "loss": 1.8758,
+      "step": 4130
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9400813579559326,
+      "learning_rate": 8.956653225806452e-05,
+      "loss": 2.1048,
+      "step": 4140
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.271834373474121,
+      "learning_rate": 8.954133064516129e-05,
+      "loss": 1.747,
+      "step": 4150
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1409404277801514,
+      "learning_rate": 8.951612903225806e-05,
+      "loss": 1.9441,
+      "step": 4160
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2031481266021729,
+      "learning_rate": 8.949092741935484e-05,
+      "loss": 1.9132,
+      "step": 4170
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.392780303955078,
+      "learning_rate": 8.946572580645161e-05,
+      "loss": 1.9364,
+      "step": 4180
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7798130512237549,
+      "learning_rate": 8.944052419354839e-05,
+      "loss": 1.9783,
+      "step": 4190
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3840584754943848,
+      "learning_rate": 8.941532258064516e-05,
+      "loss": 1.6337,
+      "step": 4200
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.411050796508789,
+      "learning_rate": 8.939012096774194e-05,
+      "loss": 1.8962,
+      "step": 4210
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.483316421508789,
+      "learning_rate": 8.936491935483871e-05,
+      "loss": 1.8418,
+      "step": 4220
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.106916069984436,
+      "learning_rate": 8.933971774193549e-05,
+      "loss": 1.8029,
+      "step": 4230
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.306298851966858,
+      "learning_rate": 8.931451612903226e-05,
+      "loss": 2.0977,
+      "step": 4240
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5087617635726929,
+      "learning_rate": 8.928931451612905e-05,
+      "loss": 1.8838,
+      "step": 4250
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.2184507846832275,
+      "learning_rate": 8.926411290322582e-05,
+      "loss": 1.9696,
+      "step": 4260
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0591418743133545,
+      "learning_rate": 8.923891129032258e-05,
+      "loss": 2.1374,
+      "step": 4270
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7099411487579346,
+      "learning_rate": 8.921370967741936e-05,
+      "loss": 1.9586,
+      "step": 4280
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4200011491775513,
+      "learning_rate": 8.918850806451613e-05,
+      "loss": 2.059,
+      "step": 4290
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4279706478118896,
+      "learning_rate": 8.91633064516129e-05,
+      "loss": 1.7937,
+      "step": 4300
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.136945128440857,
+      "learning_rate": 8.913810483870968e-05,
+      "loss": 2.1144,
+      "step": 4310
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.29193115234375,
+      "learning_rate": 8.911290322580645e-05,
+      "loss": 2.0192,
+      "step": 4320
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8602392077445984,
+      "learning_rate": 8.908770161290323e-05,
+      "loss": 1.732,
+      "step": 4330
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8382965326309204,
+      "learning_rate": 8.90625e-05,
+      "loss": 1.8658,
+      "step": 4340
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.241531252861023,
+      "learning_rate": 8.903729838709678e-05,
+      "loss": 2.0569,
+      "step": 4350
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6508620977401733,
+      "learning_rate": 8.901209677419355e-05,
+      "loss": 1.8839,
+      "step": 4360
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.9538607597351074,
+      "learning_rate": 8.898689516129033e-05,
+      "loss": 2.0725,
+      "step": 4370
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8875327110290527,
+      "learning_rate": 8.89616935483871e-05,
+      "loss": 1.8143,
+      "step": 4380
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1023283004760742,
+      "learning_rate": 8.893649193548387e-05,
+      "loss": 1.9271,
+      "step": 4390
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7703896760940552,
+      "learning_rate": 8.891129032258065e-05,
+      "loss": 1.8474,
+      "step": 4400
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.043867588043213,
+      "learning_rate": 8.888608870967742e-05,
+      "loss": 2.1104,
+      "step": 4410
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3428051471710205,
+      "learning_rate": 8.88608870967742e-05,
+      "loss": 2.0628,
+      "step": 4420
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.047571063041687,
+      "learning_rate": 8.883568548387097e-05,
+      "loss": 1.9267,
+      "step": 4430
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0767362117767334,
+      "learning_rate": 8.881048387096775e-05,
+      "loss": 1.805,
+      "step": 4440
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6265778541564941,
+      "learning_rate": 8.878528225806452e-05,
+      "loss": 1.86,
+      "step": 4450
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7227888107299805,
+      "learning_rate": 8.87600806451613e-05,
+      "loss": 1.9457,
+      "step": 4460
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0098015069961548,
+      "learning_rate": 8.873487903225807e-05,
+      "loss": 1.8973,
+      "step": 4470
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0403642654418945,
+      "learning_rate": 8.870967741935484e-05,
+      "loss": 1.9185,
+      "step": 4480
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3939849138259888,
+      "learning_rate": 8.868447580645162e-05,
+      "loss": 2.0166,
+      "step": 4490
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0339646339416504,
+      "learning_rate": 8.865927419354839e-05,
+      "loss": 1.9144,
+      "step": 4500
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6131082773208618,
+      "learning_rate": 8.863407258064517e-05,
+      "loss": 1.8817,
+      "step": 4510
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3126741647720337,
+      "learning_rate": 8.860887096774194e-05,
+      "loss": 1.9422,
+      "step": 4520
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6145623922348022,
+      "learning_rate": 8.858366935483872e-05,
+      "loss": 1.8987,
+      "step": 4530
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6964072585105896,
+      "learning_rate": 8.855846774193549e-05,
+      "loss": 1.9742,
+      "step": 4540
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0766266584396362,
+      "learning_rate": 8.853326612903226e-05,
+      "loss": 1.7995,
+      "step": 4550
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2002376317977905,
+      "learning_rate": 8.850806451612904e-05,
+      "loss": 1.9922,
+      "step": 4560
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7042362689971924,
+      "learning_rate": 8.848286290322581e-05,
+      "loss": 2.0335,
+      "step": 4570
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0903449058532715,
+      "learning_rate": 8.845766129032257e-05,
+      "loss": 1.8939,
+      "step": 4580
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.303311824798584,
+      "learning_rate": 8.843245967741935e-05,
+      "loss": 1.8029,
+      "step": 4590
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.075009822845459,
+      "learning_rate": 8.840725806451614e-05,
+      "loss": 1.8439,
+      "step": 4600
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1369531154632568,
+      "learning_rate": 8.838205645161291e-05,
+      "loss": 2.1054,
+      "step": 4610
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4303449392318726,
+      "learning_rate": 8.835685483870968e-05,
+      "loss": 1.8727,
+      "step": 4620
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9171888828277588,
+      "learning_rate": 8.833165322580646e-05,
+      "loss": 1.9627,
+      "step": 4630
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.063285231590271,
+      "learning_rate": 8.830645161290323e-05,
+      "loss": 1.8355,
+      "step": 4640
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3678268194198608,
+      "learning_rate": 8.828125000000001e-05,
+      "loss": 1.864,
+      "step": 4650
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.822312831878662,
+      "learning_rate": 8.825604838709678e-05,
+      "loss": 1.8643,
+      "step": 4660
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9407198429107666,
+      "learning_rate": 8.823084677419356e-05,
+      "loss": 1.6807,
+      "step": 4670
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7111061811447144,
+      "learning_rate": 8.820564516129033e-05,
+      "loss": 1.8968,
+      "step": 4680
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2048020362854004,
+      "learning_rate": 8.81804435483871e-05,
+      "loss": 1.8271,
+      "step": 4690
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8648923635482788,
+      "learning_rate": 8.815524193548388e-05,
+      "loss": 1.9329,
+      "step": 4700
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.664719581604004,
+      "learning_rate": 8.813004032258065e-05,
+      "loss": 1.8718,
+      "step": 4710
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5428060293197632,
+      "learning_rate": 8.810483870967743e-05,
+      "loss": 1.9771,
+      "step": 4720
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7731084227561951,
+      "learning_rate": 8.807963709677419e-05,
+      "loss": 1.8408,
+      "step": 4730
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6582595109939575,
+      "learning_rate": 8.805443548387096e-05,
+      "loss": 2.0955,
+      "step": 4740
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5434423685073853,
+      "learning_rate": 8.802923387096774e-05,
+      "loss": 1.9711,
+      "step": 4750
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2580870389938354,
+      "learning_rate": 8.800403225806451e-05,
+      "loss": 1.8547,
+      "step": 4760
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.24542236328125,
+      "learning_rate": 8.797883064516129e-05,
+      "loss": 1.9327,
+      "step": 4770
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9300427436828613,
+      "learning_rate": 8.795362903225806e-05,
+      "loss": 1.9137,
+      "step": 4780
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.391340732574463,
+      "learning_rate": 8.792842741935484e-05,
+      "loss": 2.0411,
+      "step": 4790
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3969480991363525,
+      "learning_rate": 8.790322580645162e-05,
+      "loss": 1.803,
+      "step": 4800
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2221003770828247,
+      "learning_rate": 8.78780241935484e-05,
+      "loss": 1.9455,
+      "step": 4810
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.903885006904602,
+      "learning_rate": 8.785282258064517e-05,
+      "loss": 1.9479,
+      "step": 4820
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6164311766624451,
+      "learning_rate": 8.782762096774195e-05,
+      "loss": 1.7842,
+      "step": 4830
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.441127896308899,
+      "learning_rate": 8.780241935483872e-05,
+      "loss": 1.8396,
+      "step": 4840
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7177003026008606,
+      "learning_rate": 8.77772177419355e-05,
+      "loss": 1.8384,
+      "step": 4850
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9802618026733398,
+      "learning_rate": 8.775201612903227e-05,
+      "loss": 1.7718,
+      "step": 4860
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1556932926177979,
+      "learning_rate": 8.772681451612904e-05,
+      "loss": 1.8084,
+      "step": 4870
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8862715363502502,
+      "learning_rate": 8.770161290322582e-05,
+      "loss": 2.0358,
+      "step": 4880
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8875104188919067,
+      "learning_rate": 8.767641129032258e-05,
+      "loss": 1.8383,
+      "step": 4890
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8461484909057617,
+      "learning_rate": 8.765120967741935e-05,
+      "loss": 1.7445,
+      "step": 4900
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9814643859863281,
+      "learning_rate": 8.762600806451613e-05,
+      "loss": 1.8267,
+      "step": 4910
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2121065855026245,
+      "learning_rate": 8.76008064516129e-05,
+      "loss": 1.9651,
+      "step": 4920
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6541430354118347,
+      "learning_rate": 8.757560483870968e-05,
+      "loss": 1.843,
+      "step": 4930
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.1799519062042236,
+      "learning_rate": 8.755040322580645e-05,
+      "loss": 1.9878,
+      "step": 4940
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8590803742408752,
+      "learning_rate": 8.752520161290322e-05,
+      "loss": 1.6683,
+      "step": 4950
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.731145977973938,
+      "learning_rate": 8.75e-05,
+      "loss": 1.6298,
+      "step": 4960
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.364437222480774,
+      "learning_rate": 8.747479838709677e-05,
+      "loss": 2.0393,
+      "step": 4970
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8979908227920532,
+      "learning_rate": 8.744959677419355e-05,
+      "loss": 1.6629,
+      "step": 4980
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8326512575149536,
+      "learning_rate": 8.742439516129032e-05,
+      "loss": 1.9693,
+      "step": 4990
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7193617820739746,
+      "learning_rate": 8.739919354838711e-05,
+      "loss": 2.0095,
+      "step": 5000
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.282228469848633,
+      "learning_rate": 8.737399193548388e-05,
+      "loss": 1.983,
+      "step": 5010
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3098771572113037,
+      "learning_rate": 8.734879032258066e-05,
+      "loss": 2.0685,
+      "step": 5020
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0443278551101685,
+      "learning_rate": 8.732358870967743e-05,
+      "loss": 1.9895,
+      "step": 5030
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.1133742332458496,
+      "learning_rate": 8.72983870967742e-05,
+      "loss": 1.9317,
+      "step": 5040
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3825771808624268,
+      "learning_rate": 8.727318548387097e-05,
+      "loss": 1.9311,
+      "step": 5050
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6615091562271118,
+      "learning_rate": 8.724798387096774e-05,
+      "loss": 1.8057,
+      "step": 5060
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5071910619735718,
+      "learning_rate": 8.722278225806452e-05,
+      "loss": 1.8117,
+      "step": 5070
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9011846780776978,
+      "learning_rate": 8.719758064516129e-05,
+      "loss": 1.9962,
+      "step": 5080
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.374950647354126,
+      "learning_rate": 8.717237903225807e-05,
+      "loss": 1.9812,
+      "step": 5090
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.139669179916382,
+      "learning_rate": 8.714717741935484e-05,
+      "loss": 1.8483,
+      "step": 5100
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8491741418838501,
+      "learning_rate": 8.712197580645161e-05,
+      "loss": 1.5024,
+      "step": 5110
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0059155225753784,
+      "learning_rate": 8.709677419354839e-05,
+      "loss": 2.07,
+      "step": 5120
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7128082513809204,
+      "learning_rate": 8.707157258064516e-05,
+      "loss": 1.8806,
+      "step": 5130
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5000009536743164,
+      "learning_rate": 8.704637096774194e-05,
+      "loss": 1.9421,
+      "step": 5140
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8782912492752075,
+      "learning_rate": 8.702116935483871e-05,
+      "loss": 1.8933,
+      "step": 5150
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5199337005615234,
+      "learning_rate": 8.699596774193549e-05,
+      "loss": 1.9339,
+      "step": 5160
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1087020635604858,
+      "learning_rate": 8.697076612903226e-05,
+      "loss": 2.0181,
+      "step": 5170
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6854876279830933,
+      "learning_rate": 8.694556451612903e-05,
+      "loss": 2.117,
+      "step": 5180
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5723681449890137,
+      "learning_rate": 8.692036290322582e-05,
+      "loss": 1.975,
+      "step": 5190
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.671708106994629,
+      "learning_rate": 8.689516129032258e-05,
+      "loss": 1.6628,
+      "step": 5200
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9929522275924683,
+      "learning_rate": 8.686995967741936e-05,
+      "loss": 1.7337,
+      "step": 5210
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5153592824935913,
+      "learning_rate": 8.684475806451613e-05,
+      "loss": 1.7972,
+      "step": 5220
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.727735996246338,
+      "learning_rate": 8.68195564516129e-05,
+      "loss": 2.0128,
+      "step": 5230
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.175410509109497,
+      "learning_rate": 8.679435483870968e-05,
+      "loss": 1.753,
+      "step": 5240
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0349808931350708,
+      "learning_rate": 8.676915322580646e-05,
+      "loss": 1.9458,
+      "step": 5250
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2537401914596558,
+      "learning_rate": 8.674395161290323e-05,
+      "loss": 1.8411,
+      "step": 5260
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1625862121582031,
+      "learning_rate": 8.671875e-05,
+      "loss": 2.0096,
+      "step": 5270
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.919489324092865,
+      "learning_rate": 8.669354838709678e-05,
+      "loss": 1.7521,
+      "step": 5280
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.833990752696991,
+      "learning_rate": 8.666834677419355e-05,
+      "loss": 1.6381,
+      "step": 5290
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7840721607208252,
+      "learning_rate": 8.664314516129033e-05,
+      "loss": 2.0803,
+      "step": 5300
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7827556133270264,
+      "learning_rate": 8.66179435483871e-05,
+      "loss": 2.032,
+      "step": 5310
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8062973022460938,
+      "learning_rate": 8.659274193548388e-05,
+      "loss": 1.7939,
+      "step": 5320
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.325929045677185,
+      "learning_rate": 8.656754032258065e-05,
+      "loss": 1.8242,
+      "step": 5330
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.4101955890655518,
+      "learning_rate": 8.654233870967742e-05,
+      "loss": 1.8949,
+      "step": 5340
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1818186044692993,
+      "learning_rate": 8.65171370967742e-05,
+      "loss": 1.8858,
+      "step": 5350
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.805146336555481,
+      "learning_rate": 8.649193548387097e-05,
+      "loss": 2.0503,
+      "step": 5360
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.7839488983154297,
+      "learning_rate": 8.646673387096775e-05,
+      "loss": 1.9903,
+      "step": 5370
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1524195671081543,
+      "learning_rate": 8.644153225806452e-05,
+      "loss": 1.9228,
+      "step": 5380
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.9736742973327637,
+      "learning_rate": 8.64163306451613e-05,
+      "loss": 2.1143,
+      "step": 5390
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.17699134349823,
+      "learning_rate": 8.639112903225807e-05,
+      "loss": 2.006,
+      "step": 5400
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9413127899169922,
+      "learning_rate": 8.636592741935484e-05,
+      "loss": 1.9703,
+      "step": 5410
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.43418288230896,
+      "learning_rate": 8.634072580645162e-05,
+      "loss": 1.8997,
+      "step": 5420
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1531566381454468,
+      "learning_rate": 8.63155241935484e-05,
+      "loss": 1.956,
+      "step": 5430
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8893110752105713,
+      "learning_rate": 8.629032258064517e-05,
+      "loss": 1.9761,
+      "step": 5440
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5627915859222412,
+      "learning_rate": 8.626512096774194e-05,
+      "loss": 2.0248,
+      "step": 5450
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.333624243736267,
+      "learning_rate": 8.623991935483872e-05,
+      "loss": 1.9915,
+      "step": 5460
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9281079769134521,
+      "learning_rate": 8.621471774193549e-05,
+      "loss": 1.8807,
+      "step": 5470
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5395087003707886,
+      "learning_rate": 8.618951612903227e-05,
+      "loss": 1.7256,
+      "step": 5480
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4667574167251587,
+      "learning_rate": 8.616431451612904e-05,
+      "loss": 1.9669,
+      "step": 5490
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9357525706291199,
+      "learning_rate": 8.613911290322581e-05,
+      "loss": 1.8531,
+      "step": 5500
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.176626443862915,
+      "learning_rate": 8.611391129032257e-05,
+      "loss": 1.6385,
+      "step": 5510
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4255578517913818,
+      "learning_rate": 8.608870967741935e-05,
+      "loss": 2.0321,
+      "step": 5520
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.716092586517334,
+      "learning_rate": 8.606350806451612e-05,
+      "loss": 2.0086,
+      "step": 5530
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8229900598526001,
+      "learning_rate": 8.603830645161291e-05,
+      "loss": 1.8095,
+      "step": 5540
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7317543029785156,
+      "learning_rate": 8.601310483870969e-05,
+      "loss": 2.0443,
+      "step": 5550
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0730321407318115,
+      "learning_rate": 8.598790322580646e-05,
+      "loss": 1.9194,
+      "step": 5560
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0838953256607056,
+      "learning_rate": 8.596270161290323e-05,
+      "loss": 1.7994,
+      "step": 5570
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9576981663703918,
+      "learning_rate": 8.593750000000001e-05,
+      "loss": 1.8034,
+      "step": 5580
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0293493270874023,
+      "learning_rate": 8.591229838709678e-05,
+      "loss": 1.9562,
+      "step": 5590
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1848814487457275,
+      "learning_rate": 8.588709677419356e-05,
+      "loss": 1.6994,
+      "step": 5600
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3619037866592407,
+      "learning_rate": 8.586189516129033e-05,
+      "loss": 1.9618,
+      "step": 5610
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9497629404067993,
+      "learning_rate": 8.58366935483871e-05,
+      "loss": 2.0148,
+      "step": 5620
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9130837321281433,
+      "learning_rate": 8.581149193548388e-05,
+      "loss": 1.8043,
+      "step": 5630
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9334657192230225,
+      "learning_rate": 8.578629032258065e-05,
+      "loss": 1.9481,
+      "step": 5640
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.426999568939209,
+      "learning_rate": 8.576108870967743e-05,
+      "loss": 1.78,
+      "step": 5650
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7377471923828125,
+      "learning_rate": 8.573588709677419e-05,
+      "loss": 1.8993,
+      "step": 5660
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0883934497833252,
+      "learning_rate": 8.571068548387096e-05,
+      "loss": 2.0455,
+      "step": 5670
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.720836877822876,
+      "learning_rate": 8.568548387096774e-05,
+      "loss": 1.6953,
+      "step": 5680
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1682238578796387,
+      "learning_rate": 8.566028225806451e-05,
+      "loss": 1.8429,
+      "step": 5690
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4769766330718994,
+      "learning_rate": 8.563508064516129e-05,
+      "loss": 1.9304,
+      "step": 5700
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2279216051101685,
+      "learning_rate": 8.560987903225806e-05,
+      "loss": 1.8139,
+      "step": 5710
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3377622365951538,
+      "learning_rate": 8.558467741935484e-05,
+      "loss": 1.8805,
+      "step": 5720
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.436052918434143,
+      "learning_rate": 8.555947580645161e-05,
+      "loss": 1.9579,
+      "step": 5730
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.135115385055542,
+      "learning_rate": 8.55342741935484e-05,
+      "loss": 1.7901,
+      "step": 5740
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3363252878189087,
+      "learning_rate": 8.550907258064517e-05,
+      "loss": 1.8842,
+      "step": 5750
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.2244091033935547,
+      "learning_rate": 8.548387096774195e-05,
+      "loss": 1.8645,
+      "step": 5760
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.143428921699524,
+      "learning_rate": 8.545866935483872e-05,
+      "loss": 1.7294,
+      "step": 5770
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7365981936454773,
+      "learning_rate": 8.54334677419355e-05,
+      "loss": 1.9545,
+      "step": 5780
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3291981220245361,
+      "learning_rate": 8.540826612903227e-05,
+      "loss": 1.8968,
+      "step": 5790
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8393152952194214,
+      "learning_rate": 8.538306451612904e-05,
+      "loss": 1.8391,
+      "step": 5800
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7803762555122375,
+      "learning_rate": 8.535786290322582e-05,
+      "loss": 1.8045,
+      "step": 5810
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8063178658485413,
+      "learning_rate": 8.533266129032258e-05,
+      "loss": 2.0768,
+      "step": 5820
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.365981340408325,
+      "learning_rate": 8.530745967741935e-05,
+      "loss": 1.8877,
+      "step": 5830
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.048648715019226,
+      "learning_rate": 8.528225806451613e-05,
+      "loss": 1.9328,
+      "step": 5840
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4841969013214111,
+      "learning_rate": 8.52570564516129e-05,
+      "loss": 1.9008,
+      "step": 5850
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9464988112449646,
+      "learning_rate": 8.523185483870968e-05,
+      "loss": 1.6779,
+      "step": 5860
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.106590509414673,
+      "learning_rate": 8.520665322580645e-05,
+      "loss": 1.8273,
+      "step": 5870
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1759846210479736,
+      "learning_rate": 8.518145161290323e-05,
+      "loss": 1.8685,
+      "step": 5880
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3549154996871948,
+      "learning_rate": 8.515625e-05,
+      "loss": 1.907,
+      "step": 5890
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2738302946090698,
+      "learning_rate": 8.513104838709677e-05,
+      "loss": 2.0203,
+      "step": 5900
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0730969905853271,
+      "learning_rate": 8.510584677419355e-05,
+      "loss": 1.8515,
+      "step": 5910
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.062765121459961,
+      "learning_rate": 8.508064516129032e-05,
+      "loss": 2.0258,
+      "step": 5920
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0834860801696777,
+      "learning_rate": 8.50554435483871e-05,
+      "loss": 1.7408,
+      "step": 5930
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4372564554214478,
+      "learning_rate": 8.503024193548389e-05,
+      "loss": 1.9075,
+      "step": 5940
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6854885816574097,
+      "learning_rate": 8.500504032258066e-05,
+      "loss": 2.0121,
+      "step": 5950
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5405845642089844,
+      "learning_rate": 8.497983870967743e-05,
+      "loss": 1.9499,
+      "step": 5960
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2723169326782227,
+      "learning_rate": 8.49546370967742e-05,
+      "loss": 1.9098,
+      "step": 5970
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.799237072467804,
+      "learning_rate": 8.492943548387097e-05,
+      "loss": 1.8787,
+      "step": 5980
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.001237392425537,
+      "learning_rate": 8.490423387096774e-05,
+      "loss": 1.7082,
+      "step": 5990
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2957488298416138,
+      "learning_rate": 8.487903225806452e-05,
+      "loss": 1.7335,
+      "step": 6000
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.029874324798584,
+      "learning_rate": 8.485383064516129e-05,
+      "loss": 1.8511,
+      "step": 6010
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2507532835006714,
+      "learning_rate": 8.482862903225807e-05,
+      "loss": 1.9079,
+      "step": 6020
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5290732383728027,
+      "learning_rate": 8.480342741935484e-05,
+      "loss": 1.8067,
+      "step": 6030
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0571306943893433,
+      "learning_rate": 8.477822580645161e-05,
+      "loss": 1.9876,
+      "step": 6040
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.1160378456115723,
+      "learning_rate": 8.475302419354839e-05,
+      "loss": 1.9792,
+      "step": 6050
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2963894605636597,
+      "learning_rate": 8.472782258064516e-05,
+      "loss": 1.6413,
+      "step": 6060
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.334596633911133,
+      "learning_rate": 8.470262096774194e-05,
+      "loss": 2.1446,
+      "step": 6070
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9252187609672546,
+      "learning_rate": 8.467741935483871e-05,
+      "loss": 1.9064,
+      "step": 6080
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.648625135421753,
+      "learning_rate": 8.465221774193549e-05,
+      "loss": 1.8151,
+      "step": 6090
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.334426760673523,
+      "learning_rate": 8.462701612903226e-05,
+      "loss": 1.8826,
+      "step": 6100
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6939646005630493,
+      "learning_rate": 8.460181451612904e-05,
+      "loss": 1.7197,
+      "step": 6110
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.8127381801605225,
+      "learning_rate": 8.457661290322581e-05,
+      "loss": 2.0499,
+      "step": 6120
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2957074642181396,
+      "learning_rate": 8.455141129032258e-05,
+      "loss": 2.0699,
+      "step": 6130
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2246798276901245,
+      "learning_rate": 8.452620967741936e-05,
+      "loss": 1.9718,
+      "step": 6140
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.7389230728149414,
+      "learning_rate": 8.450100806451613e-05,
+      "loss": 1.6196,
+      "step": 6150
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.2866294384002686,
+      "learning_rate": 8.447580645161291e-05,
+      "loss": 1.7379,
+      "step": 6160
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7443833351135254,
+      "learning_rate": 8.445060483870968e-05,
+      "loss": 1.9459,
+      "step": 6170
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7993885278701782,
+      "learning_rate": 8.442540322580646e-05,
+      "loss": 1.8825,
+      "step": 6180
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8703120350837708,
+      "learning_rate": 8.440020161290323e-05,
+      "loss": 1.7958,
+      "step": 6190
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.635376214981079,
+      "learning_rate": 8.4375e-05,
+      "loss": 1.8153,
+      "step": 6200
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6382055282592773,
+      "learning_rate": 8.434979838709678e-05,
+      "loss": 2.0058,
+      "step": 6210
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3786938190460205,
+      "learning_rate": 8.432459677419355e-05,
+      "loss": 1.8156,
+      "step": 6220
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3826009035110474,
+      "learning_rate": 8.429939516129033e-05,
+      "loss": 1.8796,
+      "step": 6230
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6834018230438232,
+      "learning_rate": 8.42741935483871e-05,
+      "loss": 1.8373,
+      "step": 6240
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4164958000183105,
+      "learning_rate": 8.424899193548388e-05,
+      "loss": 1.8599,
+      "step": 6250
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9641794562339783,
+      "learning_rate": 8.422379032258065e-05,
+      "loss": 1.9156,
+      "step": 6260
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.788122296333313,
+      "learning_rate": 8.419858870967742e-05,
+      "loss": 1.9228,
+      "step": 6270
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7835673093795776,
+      "learning_rate": 8.417338709677419e-05,
+      "loss": 1.868,
+      "step": 6280
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8587355613708496,
+      "learning_rate": 8.414818548387097e-05,
+      "loss": 2.0361,
+      "step": 6290
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2186648845672607,
+      "learning_rate": 8.412298387096775e-05,
+      "loss": 1.827,
+      "step": 6300
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.0856986045837402,
+      "learning_rate": 8.409778225806452e-05,
+      "loss": 1.9216,
+      "step": 6310
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5245232582092285,
+      "learning_rate": 8.40725806451613e-05,
+      "loss": 1.9667,
+      "step": 6320
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2273426055908203,
+      "learning_rate": 8.404737903225807e-05,
+      "loss": 1.6675,
+      "step": 6330
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0996100902557373,
+      "learning_rate": 8.402217741935485e-05,
+      "loss": 1.7817,
+      "step": 6340
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.181156635284424,
+      "learning_rate": 8.399697580645162e-05,
+      "loss": 2.1452,
+      "step": 6350
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6251872777938843,
+      "learning_rate": 8.39717741935484e-05,
+      "loss": 1.8117,
+      "step": 6360
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8681102991104126,
+      "learning_rate": 8.394657258064517e-05,
+      "loss": 1.9417,
+      "step": 6370
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.047458529472351,
+      "learning_rate": 8.392137096774194e-05,
+      "loss": 1.814,
+      "step": 6380
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0189634561538696,
+      "learning_rate": 8.389616935483872e-05,
+      "loss": 1.9198,
+      "step": 6390
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.42051100730896,
+      "learning_rate": 8.387096774193549e-05,
+      "loss": 1.8041,
+      "step": 6400
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.0215952396392822,
+      "learning_rate": 8.384576612903227e-05,
+      "loss": 2.0573,
+      "step": 6410
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8030675649642944,
+      "learning_rate": 8.382056451612904e-05,
+      "loss": 1.7394,
+      "step": 6420
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.065163493156433,
+      "learning_rate": 8.379536290322581e-05,
+      "loss": 1.8705,
+      "step": 6430
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0185070037841797,
+      "learning_rate": 8.377016129032258e-05,
+      "loss": 1.9855,
+      "step": 6440
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1154260635375977,
+      "learning_rate": 8.374495967741935e-05,
+      "loss": 1.7484,
+      "step": 6450
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.243526577949524,
+      "learning_rate": 8.371975806451612e-05,
+      "loss": 1.8844,
+      "step": 6460
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8716336488723755,
+      "learning_rate": 8.36945564516129e-05,
+      "loss": 1.8898,
+      "step": 6470
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4541999101638794,
+      "learning_rate": 8.366935483870967e-05,
+      "loss": 1.971,
+      "step": 6480
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7359614372253418,
+      "learning_rate": 8.364415322580646e-05,
+      "loss": 1.8572,
+      "step": 6490
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9149277210235596,
+      "learning_rate": 8.361895161290323e-05,
+      "loss": 1.8839,
+      "step": 6500
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6605091094970703,
+      "learning_rate": 8.359375000000001e-05,
+      "loss": 1.9629,
+      "step": 6510
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1591055393218994,
+      "learning_rate": 8.356854838709678e-05,
+      "loss": 2.0612,
+      "step": 6520
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3572250604629517,
+      "learning_rate": 8.354334677419356e-05,
+      "loss": 1.7094,
+      "step": 6530
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3168613910675049,
+      "learning_rate": 8.351814516129033e-05,
+      "loss": 1.9499,
+      "step": 6540
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0262044668197632,
+      "learning_rate": 8.34929435483871e-05,
+      "loss": 1.8738,
+      "step": 6550
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0592466592788696,
+      "learning_rate": 8.346774193548388e-05,
+      "loss": 1.7674,
+      "step": 6560
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5036039352416992,
+      "learning_rate": 8.344254032258066e-05,
+      "loss": 1.8352,
+      "step": 6570
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.2911376953125,
+      "learning_rate": 8.341733870967743e-05,
+      "loss": 1.8652,
+      "step": 6580
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8810513019561768,
+      "learning_rate": 8.339213709677419e-05,
+      "loss": 1.7553,
+      "step": 6590
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9186859130859375,
+      "learning_rate": 8.336693548387096e-05,
+      "loss": 1.8224,
+      "step": 6600
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7884102463722229,
+      "learning_rate": 8.334173387096774e-05,
+      "loss": 1.793,
+      "step": 6610
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.38356876373291,
+      "learning_rate": 8.331653225806451e-05,
+      "loss": 1.8182,
+      "step": 6620
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.2125115394592285,
+      "learning_rate": 8.329133064516129e-05,
+      "loss": 2.0428,
+      "step": 6630
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9531257152557373,
+      "learning_rate": 8.326612903225806e-05,
+      "loss": 1.6741,
+      "step": 6640
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0909110307693481,
+      "learning_rate": 8.324092741935484e-05,
+      "loss": 2.0364,
+      "step": 6650
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8571873307228088,
+      "learning_rate": 8.321572580645161e-05,
+      "loss": 1.9539,
+      "step": 6660
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.624374508857727,
+      "learning_rate": 8.319052419354839e-05,
+      "loss": 1.7991,
+      "step": 6670
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2886717319488525,
+      "learning_rate": 8.316532258064517e-05,
+      "loss": 1.7955,
+      "step": 6680
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9813985824584961,
+      "learning_rate": 8.314012096774195e-05,
+      "loss": 1.7168,
+      "step": 6690
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0091089010238647,
+      "learning_rate": 8.311491935483872e-05,
+      "loss": 1.8159,
+      "step": 6700
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5593523979187012,
+      "learning_rate": 8.30897177419355e-05,
+      "loss": 2.1549,
+      "step": 6710
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1408164501190186,
+      "learning_rate": 8.306451612903227e-05,
+      "loss": 1.756,
+      "step": 6720
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0500702857971191,
+      "learning_rate": 8.303931451612904e-05,
+      "loss": 1.9957,
+      "step": 6730
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1897822618484497,
+      "learning_rate": 8.301411290322582e-05,
+      "loss": 1.9639,
+      "step": 6740
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6673760414123535,
+      "learning_rate": 8.298891129032258e-05,
+      "loss": 2.0512,
+      "step": 6750
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5629624128341675,
+      "learning_rate": 8.296370967741935e-05,
+      "loss": 1.736,
+      "step": 6760
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4218215942382812,
+      "learning_rate": 8.293850806451613e-05,
+      "loss": 1.9276,
+      "step": 6770
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.8032217025756836,
+      "learning_rate": 8.29133064516129e-05,
+      "loss": 1.8376,
+      "step": 6780
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.547479271888733,
+      "learning_rate": 8.288810483870968e-05,
+      "loss": 1.8687,
+      "step": 6790
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7930785417556763,
+      "learning_rate": 8.286290322580645e-05,
+      "loss": 1.8432,
+      "step": 6800
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.0666840076446533,
+      "learning_rate": 8.283770161290323e-05,
+      "loss": 2.0358,
+      "step": 6810
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6833534240722656,
+      "learning_rate": 8.28125e-05,
+      "loss": 1.8356,
+      "step": 6820
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5746026039123535,
+      "learning_rate": 8.278729838709677e-05,
+      "loss": 1.9213,
+      "step": 6830
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5233973264694214,
+      "learning_rate": 8.276209677419355e-05,
+      "loss": 1.7725,
+      "step": 6840
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7120718359947205,
+      "learning_rate": 8.273689516129032e-05,
+      "loss": 1.963,
+      "step": 6850
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2445943355560303,
+      "learning_rate": 8.27116935483871e-05,
+      "loss": 1.9837,
+      "step": 6860
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8460561037063599,
+      "learning_rate": 8.268649193548387e-05,
+      "loss": 1.9142,
+      "step": 6870
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9037141799926758,
+      "learning_rate": 8.266129032258066e-05,
+      "loss": 1.8152,
+      "step": 6880
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0428452491760254,
+      "learning_rate": 8.263608870967743e-05,
+      "loss": 1.7352,
+      "step": 6890
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1458503007888794,
+      "learning_rate": 8.26108870967742e-05,
+      "loss": 1.9077,
+      "step": 6900
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.194357991218567,
+      "learning_rate": 8.258568548387097e-05,
+      "loss": 1.9691,
+      "step": 6910
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.291100263595581,
+      "learning_rate": 8.256048387096774e-05,
+      "loss": 2.0121,
+      "step": 6920
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.081604242324829,
+      "learning_rate": 8.253528225806452e-05,
+      "loss": 2.0406,
+      "step": 6930
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7903962135314941,
+      "learning_rate": 8.251008064516129e-05,
+      "loss": 1.9512,
+      "step": 6940
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8859573006629944,
+      "learning_rate": 8.248487903225807e-05,
+      "loss": 1.7556,
+      "step": 6950
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.995909571647644,
+      "learning_rate": 8.245967741935484e-05,
+      "loss": 1.7344,
+      "step": 6960
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2884035110473633,
+      "learning_rate": 8.243447580645162e-05,
+      "loss": 1.9128,
+      "step": 6970
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4027931690216064,
+      "learning_rate": 8.240927419354839e-05,
+      "loss": 1.8633,
+      "step": 6980
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1019028425216675,
+      "learning_rate": 8.238407258064516e-05,
+      "loss": 1.9303,
+      "step": 6990
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9130522608757019,
+      "learning_rate": 8.235887096774194e-05,
+      "loss": 1.8654,
+      "step": 7000
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2252943515777588,
+      "learning_rate": 8.233366935483871e-05,
+      "loss": 1.7452,
+      "step": 7010
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5055691003799438,
+      "learning_rate": 8.230846774193549e-05,
+      "loss": 2.0861,
+      "step": 7020
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3446593284606934,
+      "learning_rate": 8.228326612903226e-05,
+      "loss": 1.9208,
+      "step": 7030
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9395591020584106,
+      "learning_rate": 8.225806451612904e-05,
+      "loss": 2.0028,
+      "step": 7040
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.120570421218872,
+      "learning_rate": 8.223286290322581e-05,
+      "loss": 1.8232,
+      "step": 7050
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9680400490760803,
+      "learning_rate": 8.220766129032258e-05,
+      "loss": 1.9092,
+      "step": 7060
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4184253215789795,
+      "learning_rate": 8.218245967741936e-05,
+      "loss": 1.8094,
+      "step": 7070
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2642617225646973,
+      "learning_rate": 8.215725806451613e-05,
+      "loss": 1.8672,
+      "step": 7080
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7254289388656616,
+      "learning_rate": 8.213205645161291e-05,
+      "loss": 1.9322,
+      "step": 7090
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0914796590805054,
+      "learning_rate": 8.210685483870968e-05,
+      "loss": 1.7684,
+      "step": 7100
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2319042682647705,
+      "learning_rate": 8.208165322580646e-05,
+      "loss": 1.8409,
+      "step": 7110
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8690718412399292,
+      "learning_rate": 8.205645161290323e-05,
+      "loss": 1.9377,
+      "step": 7120
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.410550594329834,
+      "learning_rate": 8.203125e-05,
+      "loss": 2.0528,
+      "step": 7130
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2189645767211914,
+      "learning_rate": 8.200604838709678e-05,
+      "loss": 1.8247,
+      "step": 7140
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.075155019760132,
+      "learning_rate": 8.198084677419355e-05,
+      "loss": 1.797,
+      "step": 7150
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.017909049987793,
+      "learning_rate": 8.195564516129033e-05,
+      "loss": 1.9098,
+      "step": 7160
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5938984155654907,
+      "learning_rate": 8.19304435483871e-05,
+      "loss": 1.9852,
+      "step": 7170
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5918670892715454,
+      "learning_rate": 8.190524193548388e-05,
+      "loss": 1.8899,
+      "step": 7180
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1991071701049805,
+      "learning_rate": 8.188004032258065e-05,
+      "loss": 1.9626,
+      "step": 7190
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6108486652374268,
+      "learning_rate": 8.185483870967743e-05,
+      "loss": 1.9042,
+      "step": 7200
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.746124029159546,
+      "learning_rate": 8.182963709677419e-05,
+      "loss": 1.716,
+      "step": 7210
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.436093330383301,
+      "learning_rate": 8.180443548387096e-05,
+      "loss": 1.8006,
+      "step": 7220
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5066144466400146,
+      "learning_rate": 8.177923387096775e-05,
+      "loss": 1.9305,
+      "step": 7230
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.3824262619018555,
+      "learning_rate": 8.175403225806452e-05,
+      "loss": 1.9572,
+      "step": 7240
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6272624731063843,
+      "learning_rate": 8.17288306451613e-05,
+      "loss": 1.7852,
+      "step": 7250
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2313551902770996,
+      "learning_rate": 8.170362903225807e-05,
+      "loss": 1.7508,
+      "step": 7260
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5643839836120605,
+      "learning_rate": 8.167842741935485e-05,
+      "loss": 2.0497,
+      "step": 7270
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.84746515750885,
+      "learning_rate": 8.165322580645162e-05,
+      "loss": 1.8911,
+      "step": 7280
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.144025444984436,
+      "learning_rate": 8.16280241935484e-05,
+      "loss": 1.7448,
+      "step": 7290
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5191675424575806,
+      "learning_rate": 8.160282258064517e-05,
+      "loss": 2.0394,
+      "step": 7300
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.087950587272644,
+      "learning_rate": 8.157762096774194e-05,
+      "loss": 2.0202,
+      "step": 7310
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9370905756950378,
+      "learning_rate": 8.155241935483872e-05,
+      "loss": 1.7991,
+      "step": 7320
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0464539527893066,
+      "learning_rate": 8.152721774193549e-05,
+      "loss": 1.8851,
+      "step": 7330
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3758870363235474,
+      "learning_rate": 8.150201612903227e-05,
+      "loss": 1.8497,
+      "step": 7340
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.1545984745025635,
+      "learning_rate": 8.147681451612904e-05,
+      "loss": 1.8334,
+      "step": 7350
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4549347162246704,
+      "learning_rate": 8.145161290322582e-05,
+      "loss": 1.8288,
+      "step": 7360
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0828857421875,
+      "learning_rate": 8.142641129032258e-05,
+      "loss": 1.9292,
+      "step": 7370
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.67167067527771,
+      "learning_rate": 8.140120967741935e-05,
+      "loss": 2.1122,
+      "step": 7380
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.424847960472107,
+      "learning_rate": 8.137600806451612e-05,
+      "loss": 1.8739,
+      "step": 7390
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5594614744186401,
+      "learning_rate": 8.13508064516129e-05,
+      "loss": 1.9554,
+      "step": 7400
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0690380334854126,
+      "learning_rate": 8.132560483870967e-05,
+      "loss": 2.0286,
+      "step": 7410
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.991178035736084,
+      "learning_rate": 8.130040322580645e-05,
+      "loss": 1.889,
+      "step": 7420
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2756601572036743,
+      "learning_rate": 8.127520161290324e-05,
+      "loss": 1.9739,
+      "step": 7430
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.316148281097412,
+      "learning_rate": 8.125000000000001e-05,
+      "loss": 2.0722,
+      "step": 7440
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8382697105407715,
+      "learning_rate": 8.122479838709678e-05,
+      "loss": 1.8983,
+      "step": 7450
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3752154111862183,
+      "learning_rate": 8.119959677419356e-05,
+      "loss": 2.0404,
+      "step": 7460
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2030837535858154,
+      "learning_rate": 8.117439516129033e-05,
+      "loss": 1.9638,
+      "step": 7470
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9619494676589966,
+      "learning_rate": 8.114919354838711e-05,
+      "loss": 1.8364,
+      "step": 7480
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8871836066246033,
+      "learning_rate": 8.112399193548388e-05,
+      "loss": 1.8316,
+      "step": 7490
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5749518871307373,
+      "learning_rate": 8.109879032258066e-05,
+      "loss": 1.8035,
+      "step": 7500
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5351563692092896,
+      "learning_rate": 8.107358870967743e-05,
+      "loss": 1.9723,
+      "step": 7510
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.914915919303894,
+      "learning_rate": 8.104838709677419e-05,
+      "loss": 1.8921,
+      "step": 7520
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3868452310562134,
+      "learning_rate": 8.102318548387097e-05,
+      "loss": 1.9946,
+      "step": 7530
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7173086404800415,
+      "learning_rate": 8.099798387096774e-05,
+      "loss": 1.9112,
+      "step": 7540
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5832910537719727,
+      "learning_rate": 8.097278225806451e-05,
+      "loss": 1.9238,
+      "step": 7550
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5022016763687134,
+      "learning_rate": 8.094758064516129e-05,
+      "loss": 2.0074,
+      "step": 7560
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9701142907142639,
+      "learning_rate": 8.092237903225806e-05,
+      "loss": 1.6734,
+      "step": 7570
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.760705828666687,
+      "learning_rate": 8.089717741935484e-05,
+      "loss": 1.8012,
+      "step": 7580
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.022233486175537,
+      "learning_rate": 8.087197580645161e-05,
+      "loss": 1.7767,
+      "step": 7590
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9845382571220398,
+      "learning_rate": 8.084677419354839e-05,
+      "loss": 1.9444,
+      "step": 7600
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9495793581008911,
+      "learning_rate": 8.082157258064516e-05,
+      "loss": 1.9873,
+      "step": 7610
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.257048487663269,
+      "learning_rate": 8.079637096774193e-05,
+      "loss": 1.7037,
+      "step": 7620
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3337973356246948,
+      "learning_rate": 8.077116935483872e-05,
+      "loss": 1.9499,
+      "step": 7630
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8914852142333984,
+      "learning_rate": 8.07459677419355e-05,
+      "loss": 2.0383,
+      "step": 7640
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.120595932006836,
+      "learning_rate": 8.072076612903227e-05,
+      "loss": 1.7335,
+      "step": 7650
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9817825555801392,
+      "learning_rate": 8.069556451612905e-05,
+      "loss": 1.8985,
+      "step": 7660
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.303882122039795,
+      "learning_rate": 8.067036290322582e-05,
+      "loss": 1.7517,
+      "step": 7670
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.232192039489746,
+      "learning_rate": 8.064516129032258e-05,
+      "loss": 1.6439,
+      "step": 7680
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6390267610549927,
+      "learning_rate": 8.061995967741935e-05,
+      "loss": 1.8787,
+      "step": 7690
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.661180853843689,
+      "learning_rate": 8.059475806451613e-05,
+      "loss": 1.9593,
+      "step": 7700
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7900233268737793,
+      "learning_rate": 8.05695564516129e-05,
+      "loss": 1.9274,
+      "step": 7710
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1088135242462158,
+      "learning_rate": 8.054435483870968e-05,
+      "loss": 1.8621,
+      "step": 7720
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1402857303619385,
+      "learning_rate": 8.051915322580645e-05,
+      "loss": 1.8159,
+      "step": 7730
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0286141633987427,
+      "learning_rate": 8.049395161290323e-05,
+      "loss": 2.0628,
+      "step": 7740
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.297602891921997,
+      "learning_rate": 8.046875e-05,
+      "loss": 1.831,
+      "step": 7750
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.877297282218933,
+      "learning_rate": 8.044354838709678e-05,
+      "loss": 2.0404,
+      "step": 7760
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0671412944793701,
+      "learning_rate": 8.041834677419355e-05,
+      "loss": 1.7863,
+      "step": 7770
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1550469398498535,
+      "learning_rate": 8.039314516129032e-05,
+      "loss": 1.9807,
+      "step": 7780
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6271408796310425,
+      "learning_rate": 8.03679435483871e-05,
+      "loss": 1.9338,
+      "step": 7790
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1950056552886963,
+      "learning_rate": 8.034274193548387e-05,
+      "loss": 1.9363,
+      "step": 7800
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9773239493370056,
+      "learning_rate": 8.031754032258065e-05,
+      "loss": 1.6658,
+      "step": 7810
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.143672227859497,
+      "learning_rate": 8.029233870967743e-05,
+      "loss": 1.9559,
+      "step": 7820
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9005433320999146,
+      "learning_rate": 8.02671370967742e-05,
+      "loss": 1.726,
+      "step": 7830
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0035150051116943,
+      "learning_rate": 8.024193548387097e-05,
+      "loss": 1.7031,
+      "step": 7840
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8358875513076782,
+      "learning_rate": 8.021673387096774e-05,
+      "loss": 1.9601,
+      "step": 7850
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8777676820755005,
+      "learning_rate": 8.019153225806452e-05,
+      "loss": 1.9493,
+      "step": 7860
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1832433938980103,
+      "learning_rate": 8.016633064516129e-05,
+      "loss": 1.8979,
+      "step": 7870
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0962862968444824,
+      "learning_rate": 8.014112903225807e-05,
+      "loss": 1.6644,
+      "step": 7880
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.038049578666687,
+      "learning_rate": 8.011592741935484e-05,
+      "loss": 1.6374,
+      "step": 7890
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7101986408233643,
+      "learning_rate": 8.009072580645162e-05,
+      "loss": 1.8789,
+      "step": 7900
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3773138523101807,
+      "learning_rate": 8.006552419354839e-05,
+      "loss": 1.8718,
+      "step": 7910
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8748659491539001,
+      "learning_rate": 8.004032258064516e-05,
+      "loss": 1.6778,
+      "step": 7920
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4577199220657349,
+      "learning_rate": 8.001512096774194e-05,
+      "loss": 1.8338,
+      "step": 7930
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1285208463668823,
+      "learning_rate": 7.998991935483871e-05,
+      "loss": 1.6887,
+      "step": 7940
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.224743127822876,
+      "learning_rate": 7.996471774193549e-05,
+      "loss": 1.7898,
+      "step": 7950
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.14567232131958,
+      "learning_rate": 7.993951612903226e-05,
+      "loss": 1.7032,
+      "step": 7960
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1482393741607666,
+      "learning_rate": 7.991431451612904e-05,
+      "loss": 1.7992,
+      "step": 7970
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9437839984893799,
+      "learning_rate": 7.988911290322581e-05,
+      "loss": 1.809,
+      "step": 7980
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.107179641723633,
+      "learning_rate": 7.986391129032259e-05,
+      "loss": 2.0047,
+      "step": 7990
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0202579498291016,
+      "learning_rate": 7.983870967741936e-05,
+      "loss": 1.9849,
+      "step": 8000
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3112866878509521,
+      "learning_rate": 7.981350806451613e-05,
+      "loss": 2.0294,
+      "step": 8010
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.826329231262207,
+      "learning_rate": 7.978830645161291e-05,
+      "loss": 1.9525,
+      "step": 8020
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.508089542388916,
+      "learning_rate": 7.976310483870968e-05,
+      "loss": 1.7501,
+      "step": 8030
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4878416061401367,
+      "learning_rate": 7.973790322580646e-05,
+      "loss": 1.7063,
+      "step": 8040
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3690828084945679,
+      "learning_rate": 7.971270161290323e-05,
+      "loss": 2.0004,
+      "step": 8050
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4743849039077759,
+      "learning_rate": 7.96875e-05,
+      "loss": 1.5971,
+      "step": 8060
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.670180082321167,
+      "learning_rate": 7.966229838709678e-05,
+      "loss": 1.8059,
+      "step": 8070
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5801221132278442,
+      "learning_rate": 7.963709677419355e-05,
+      "loss": 1.8238,
+      "step": 8080
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4766119718551636,
+      "learning_rate": 7.961189516129033e-05,
+      "loss": 1.9583,
+      "step": 8090
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.955820083618164,
+      "learning_rate": 7.95866935483871e-05,
+      "loss": 1.9322,
+      "step": 8100
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.271837592124939,
+      "learning_rate": 7.956149193548388e-05,
+      "loss": 1.9017,
+      "step": 8110
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8310510516166687,
+      "learning_rate": 7.953629032258065e-05,
+      "loss": 1.6684,
+      "step": 8120
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3256120681762695,
+      "learning_rate": 7.951108870967743e-05,
+      "loss": 2.0603,
+      "step": 8130
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.165334701538086,
+      "learning_rate": 7.948588709677419e-05,
+      "loss": 1.963,
+      "step": 8140
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2800134420394897,
+      "learning_rate": 7.946068548387096e-05,
+      "loss": 1.8275,
+      "step": 8150
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.322439193725586,
+      "learning_rate": 7.943548387096774e-05,
+      "loss": 1.7945,
+      "step": 8160
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9921194911003113,
+      "learning_rate": 7.941028225806451e-05,
+      "loss": 1.8585,
+      "step": 8170
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3062148094177246,
+      "learning_rate": 7.93850806451613e-05,
+      "loss": 1.9405,
+      "step": 8180
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.093291997909546,
+      "learning_rate": 7.935987903225807e-05,
+      "loss": 1.8114,
+      "step": 8190
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3523463010787964,
+      "learning_rate": 7.933467741935485e-05,
+      "loss": 1.8434,
+      "step": 8200
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.845554769039154,
+      "learning_rate": 7.930947580645162e-05,
+      "loss": 1.9854,
+      "step": 8210
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.168615460395813,
+      "learning_rate": 7.92842741935484e-05,
+      "loss": 2.0492,
+      "step": 8220
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.179154634475708,
+      "learning_rate": 7.925907258064517e-05,
+      "loss": 1.8716,
+      "step": 8230
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5088022947311401,
+      "learning_rate": 7.923387096774194e-05,
+      "loss": 2.0265,
+      "step": 8240
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.144437313079834,
+      "learning_rate": 7.920866935483872e-05,
+      "loss": 1.9418,
+      "step": 8250
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.4098360538482666,
+      "learning_rate": 7.918346774193549e-05,
+      "loss": 1.8638,
+      "step": 8260
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1946191787719727,
+      "learning_rate": 7.915826612903227e-05,
+      "loss": 1.9813,
+      "step": 8270
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5844733119010925,
+      "learning_rate": 7.913306451612904e-05,
+      "loss": 1.6616,
+      "step": 8280
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9644368290901184,
+      "learning_rate": 7.910786290322582e-05,
+      "loss": 1.6646,
+      "step": 8290
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.187357187271118,
+      "learning_rate": 7.908266129032258e-05,
+      "loss": 1.9972,
+      "step": 8300
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.656157970428467,
+      "learning_rate": 7.905745967741935e-05,
+      "loss": 1.7299,
+      "step": 8310
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5695854425430298,
+      "learning_rate": 7.903225806451613e-05,
+      "loss": 1.87,
+      "step": 8320
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0215736627578735,
+      "learning_rate": 7.90070564516129e-05,
+      "loss": 1.8338,
+      "step": 8330
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4404747486114502,
+      "learning_rate": 7.898185483870967e-05,
+      "loss": 1.9055,
+      "step": 8340
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.1423099040985107,
+      "learning_rate": 7.895665322580645e-05,
+      "loss": 2.0448,
+      "step": 8350
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3649463653564453,
+      "learning_rate": 7.893145161290322e-05,
+      "loss": 1.975,
+      "step": 8360
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1941394805908203,
+      "learning_rate": 7.890625000000001e-05,
+      "loss": 1.8033,
+      "step": 8370
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1206849813461304,
+      "learning_rate": 7.888104838709678e-05,
+      "loss": 1.736,
+      "step": 8380
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.3108301162719727,
+      "learning_rate": 7.885584677419356e-05,
+      "loss": 1.9586,
+      "step": 8390
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.737004518508911,
+      "learning_rate": 7.883064516129033e-05,
+      "loss": 1.9098,
+      "step": 8400
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4950056076049805,
+      "learning_rate": 7.880544354838711e-05,
+      "loss": 2.1578,
+      "step": 8410
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1863486766815186,
+      "learning_rate": 7.878024193548388e-05,
+      "loss": 1.9233,
+      "step": 8420
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2473092079162598,
+      "learning_rate": 7.875504032258066e-05,
+      "loss": 1.7847,
+      "step": 8430
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2708603143692017,
+      "learning_rate": 7.872983870967743e-05,
+      "loss": 1.9211,
+      "step": 8440
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5947640538215637,
+      "learning_rate": 7.870463709677419e-05,
+      "loss": 1.8237,
+      "step": 8450
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.923574686050415,
+      "learning_rate": 7.867943548387097e-05,
+      "loss": 1.8247,
+      "step": 8460
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8887185454368591,
+      "learning_rate": 7.865423387096774e-05,
+      "loss": 1.7379,
+      "step": 8470
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9528153538703918,
+      "learning_rate": 7.862903225806451e-05,
+      "loss": 1.8398,
+      "step": 8480
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6888993978500366,
+      "learning_rate": 7.860383064516129e-05,
+      "loss": 1.9063,
+      "step": 8490
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7411779165267944,
+      "learning_rate": 7.857862903225806e-05,
+      "loss": 2.1165,
+      "step": 8500
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0220638513565063,
+      "learning_rate": 7.855342741935484e-05,
+      "loss": 2.0086,
+      "step": 8510
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.656347632408142,
+      "learning_rate": 7.852822580645161e-05,
+      "loss": 1.6933,
+      "step": 8520
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0002708435058594,
+      "learning_rate": 7.850302419354839e-05,
+      "loss": 1.9501,
+      "step": 8530
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2823266983032227,
+      "learning_rate": 7.847782258064516e-05,
+      "loss": 1.7686,
+      "step": 8540
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0139843225479126,
+      "learning_rate": 7.845262096774194e-05,
+      "loss": 1.8658,
+      "step": 8550
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0108537673950195,
+      "learning_rate": 7.842741935483871e-05,
+      "loss": 1.8933,
+      "step": 8560
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.0528485774993896,
+      "learning_rate": 7.84022177419355e-05,
+      "loss": 1.9479,
+      "step": 8570
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5194003582000732,
+      "learning_rate": 7.837701612903227e-05,
+      "loss": 1.7248,
+      "step": 8580
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7669309377670288,
+      "learning_rate": 7.835181451612905e-05,
+      "loss": 1.8428,
+      "step": 8590
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8602083921432495,
+      "learning_rate": 7.832661290322582e-05,
+      "loss": 1.9566,
+      "step": 8600
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7124571800231934,
+      "learning_rate": 7.830141129032258e-05,
+      "loss": 2.0093,
+      "step": 8610
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0541378259658813,
+      "learning_rate": 7.827620967741936e-05,
+      "loss": 1.6307,
+      "step": 8620
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5085299015045166,
+      "learning_rate": 7.825100806451613e-05,
+      "loss": 1.9703,
+      "step": 8630
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.5353894233703613,
+      "learning_rate": 7.82258064516129e-05,
+      "loss": 2.0895,
+      "step": 8640
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7369152307510376,
+      "learning_rate": 7.820060483870968e-05,
+      "loss": 2.1117,
+      "step": 8650
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2966816425323486,
+      "learning_rate": 7.817540322580645e-05,
+      "loss": 2.0283,
+      "step": 8660
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0495866537094116,
+      "learning_rate": 7.815020161290323e-05,
+      "loss": 1.7492,
+      "step": 8670
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1473238468170166,
+      "learning_rate": 7.8125e-05,
+      "loss": 1.9477,
+      "step": 8680
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.502866268157959,
+      "learning_rate": 7.809979838709678e-05,
+      "loss": 1.8424,
+      "step": 8690
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8116331100463867,
+      "learning_rate": 7.807459677419355e-05,
+      "loss": 1.986,
+      "step": 8700
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6133512258529663,
+      "learning_rate": 7.804939516129032e-05,
+      "loss": 1.8742,
+      "step": 8710
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0546581745147705,
+      "learning_rate": 7.80241935483871e-05,
+      "loss": 1.9257,
+      "step": 8720
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.229851245880127,
+      "learning_rate": 7.799899193548387e-05,
+      "loss": 1.7816,
+      "step": 8730
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5405186414718628,
+      "learning_rate": 7.797379032258065e-05,
+      "loss": 1.9548,
+      "step": 8740
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.9373369216918945,
+      "learning_rate": 7.794858870967742e-05,
+      "loss": 1.8763,
+      "step": 8750
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9205381870269775,
+      "learning_rate": 7.79233870967742e-05,
+      "loss": 2.0134,
+      "step": 8760
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.1046676635742188,
+      "learning_rate": 7.789818548387097e-05,
+      "loss": 1.842,
+      "step": 8770
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7414477467536926,
+      "learning_rate": 7.787298387096775e-05,
+      "loss": 1.8628,
+      "step": 8780
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.005478620529175,
+      "learning_rate": 7.784778225806452e-05,
+      "loss": 1.8564,
+      "step": 8790
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5745129585266113,
+      "learning_rate": 7.78225806451613e-05,
+      "loss": 1.9171,
+      "step": 8800
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1416375637054443,
+      "learning_rate": 7.779737903225807e-05,
+      "loss": 1.9313,
+      "step": 8810
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9569885730743408,
+      "learning_rate": 7.777217741935484e-05,
+      "loss": 2.0178,
+      "step": 8820
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9609521627426147,
+      "learning_rate": 7.774697580645162e-05,
+      "loss": 1.9206,
+      "step": 8830
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4023690223693848,
+      "learning_rate": 7.772177419354839e-05,
+      "loss": 1.9582,
+      "step": 8840
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.209123373031616,
+      "learning_rate": 7.769657258064517e-05,
+      "loss": 1.9232,
+      "step": 8850
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.627531886100769,
+      "learning_rate": 7.767137096774194e-05,
+      "loss": 1.9669,
+      "step": 8860
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3856791257858276,
+      "learning_rate": 7.764616935483871e-05,
+      "loss": 1.7471,
+      "step": 8870
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.83492112159729,
+      "learning_rate": 7.762096774193549e-05,
+      "loss": 1.9921,
+      "step": 8880
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4671322107315063,
+      "learning_rate": 7.759576612903226e-05,
+      "loss": 1.9364,
+      "step": 8890
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3159751892089844,
+      "learning_rate": 7.757056451612904e-05,
+      "loss": 2.0538,
+      "step": 8900
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.393800973892212,
+      "learning_rate": 7.754536290322581e-05,
+      "loss": 1.8986,
+      "step": 8910
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5282772779464722,
+      "learning_rate": 7.752016129032259e-05,
+      "loss": 1.8958,
+      "step": 8920
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8281263113021851,
+      "learning_rate": 7.749495967741936e-05,
+      "loss": 1.7605,
+      "step": 8930
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8808881044387817,
+      "learning_rate": 7.746975806451613e-05,
+      "loss": 1.7783,
+      "step": 8940
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0158687829971313,
+      "learning_rate": 7.744455645161291e-05,
+      "loss": 1.9551,
+      "step": 8950
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8328729867935181,
+      "learning_rate": 7.741935483870968e-05,
+      "loss": 1.8599,
+      "step": 8960
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9793317317962646,
+      "learning_rate": 7.739415322580646e-05,
+      "loss": 1.9978,
+      "step": 8970
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6600232124328613,
+      "learning_rate": 7.736895161290323e-05,
+      "loss": 1.8967,
+      "step": 8980
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5028966665267944,
+      "learning_rate": 7.734375e-05,
+      "loss": 1.8919,
+      "step": 8990
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.215562343597412,
+      "learning_rate": 7.731854838709678e-05,
+      "loss": 2.1029,
+      "step": 9000
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.9349522590637207,
+      "learning_rate": 7.729334677419356e-05,
+      "loss": 2.0444,
+      "step": 9010
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0412317514419556,
+      "learning_rate": 7.726814516129033e-05,
+      "loss": 1.7747,
+      "step": 9020
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5292258262634277,
+      "learning_rate": 7.72429435483871e-05,
+      "loss": 1.8359,
+      "step": 9030
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1081246137619019,
+      "learning_rate": 7.721774193548388e-05,
+      "loss": 2.0114,
+      "step": 9040
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3633612394332886,
+      "learning_rate": 7.719254032258065e-05,
+      "loss": 1.9479,
+      "step": 9050
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7072218656539917,
+      "learning_rate": 7.716733870967743e-05,
+      "loss": 1.9179,
+      "step": 9060
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.195155382156372,
+      "learning_rate": 7.714213709677419e-05,
+      "loss": 1.9003,
+      "step": 9070
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4204546213150024,
+      "learning_rate": 7.711693548387096e-05,
+      "loss": 1.8334,
+      "step": 9080
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3352149724960327,
+      "learning_rate": 7.709173387096774e-05,
+      "loss": 1.7741,
+      "step": 9090
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.907107353210449,
+      "learning_rate": 7.706653225806451e-05,
+      "loss": 1.7911,
+      "step": 9100
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9949589967727661,
+      "learning_rate": 7.704133064516128e-05,
+      "loss": 1.8145,
+      "step": 9110
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7108676433563232,
+      "learning_rate": 7.701612903225807e-05,
+      "loss": 1.8915,
+      "step": 9120
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1672462224960327,
+      "learning_rate": 7.699092741935485e-05,
+      "loss": 1.9559,
+      "step": 9130
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.109964370727539,
+      "learning_rate": 7.696572580645162e-05,
+      "loss": 1.7455,
+      "step": 9140
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0365976095199585,
+      "learning_rate": 7.69405241935484e-05,
+      "loss": 1.8841,
+      "step": 9150
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.491157054901123,
+      "learning_rate": 7.691532258064517e-05,
+      "loss": 1.7611,
+      "step": 9160
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9735540151596069,
+      "learning_rate": 7.689012096774194e-05,
+      "loss": 1.7431,
+      "step": 9170
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6469701528549194,
+      "learning_rate": 7.686491935483872e-05,
+      "loss": 1.9401,
+      "step": 9180
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.16251802444458,
+      "learning_rate": 7.68397177419355e-05,
+      "loss": 1.8621,
+      "step": 9190
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.810158371925354,
+      "learning_rate": 7.681451612903227e-05,
+      "loss": 1.8433,
+      "step": 9200
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9513418078422546,
+      "learning_rate": 7.678931451612904e-05,
+      "loss": 1.8289,
+      "step": 9210
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.310531497001648,
+      "learning_rate": 7.676411290322582e-05,
+      "loss": 1.917,
+      "step": 9220
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.495252013206482,
+      "learning_rate": 7.673891129032258e-05,
+      "loss": 1.9194,
+      "step": 9230
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.417749285697937,
+      "learning_rate": 7.671370967741935e-05,
+      "loss": 1.8796,
+      "step": 9240
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1487481594085693,
+      "learning_rate": 7.668850806451613e-05,
+      "loss": 1.9273,
+      "step": 9250
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6958103775978088,
+      "learning_rate": 7.66633064516129e-05,
+      "loss": 1.7975,
+      "step": 9260
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4320131540298462,
+      "learning_rate": 7.663810483870967e-05,
+      "loss": 1.9748,
+      "step": 9270
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1957752704620361,
+      "learning_rate": 7.661290322580645e-05,
+      "loss": 1.838,
+      "step": 9280
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.125065326690674,
+      "learning_rate": 7.658770161290322e-05,
+      "loss": 1.7198,
+      "step": 9290
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2782047986984253,
+      "learning_rate": 7.65625e-05,
+      "loss": 1.9649,
+      "step": 9300
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.323254108428955,
+      "learning_rate": 7.653729838709677e-05,
+      "loss": 1.8265,
+      "step": 9310
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4309558868408203,
+      "learning_rate": 7.651209677419356e-05,
+      "loss": 1.7728,
+      "step": 9320
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8265936374664307,
+      "learning_rate": 7.648689516129033e-05,
+      "loss": 1.9186,
+      "step": 9330
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4147733449935913,
+      "learning_rate": 7.646169354838711e-05,
+      "loss": 1.942,
+      "step": 9340
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9708012342453003,
+      "learning_rate": 7.643649193548388e-05,
+      "loss": 1.6797,
+      "step": 9350
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.272778868675232,
+      "learning_rate": 7.641129032258066e-05,
+      "loss": 1.8293,
+      "step": 9360
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.466907024383545,
+      "learning_rate": 7.638608870967743e-05,
+      "loss": 1.7276,
+      "step": 9370
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8608651757240295,
+      "learning_rate": 7.636088709677419e-05,
+      "loss": 1.8674,
+      "step": 9380
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2072697877883911,
+      "learning_rate": 7.633568548387097e-05,
+      "loss": 1.8452,
+      "step": 9390
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7777898907661438,
+      "learning_rate": 7.631048387096774e-05,
+      "loss": 1.9995,
+      "step": 9400
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0407038927078247,
+      "learning_rate": 7.628528225806452e-05,
+      "loss": 1.8883,
+      "step": 9410
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1916221380233765,
+      "learning_rate": 7.626008064516129e-05,
+      "loss": 2.0669,
+      "step": 9420
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7974575161933899,
+      "learning_rate": 7.623487903225806e-05,
+      "loss": 1.7887,
+      "step": 9430
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3895624876022339,
+      "learning_rate": 7.620967741935484e-05,
+      "loss": 1.9486,
+      "step": 9440
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2316830158233643,
+      "learning_rate": 7.618447580645161e-05,
+      "loss": 1.7769,
+      "step": 9450
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3468798398971558,
+      "learning_rate": 7.615927419354839e-05,
+      "loss": 1.9718,
+      "step": 9460
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9262570142745972,
+      "learning_rate": 7.613407258064516e-05,
+      "loss": 1.941,
+      "step": 9470
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1528464555740356,
+      "learning_rate": 7.610887096774194e-05,
+      "loss": 2.0369,
+      "step": 9480
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5017763376235962,
+      "learning_rate": 7.608366935483871e-05,
+      "loss": 1.9855,
+      "step": 9490
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6908256411552429,
+      "learning_rate": 7.605846774193548e-05,
+      "loss": 1.7791,
+      "step": 9500
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7042288780212402,
+      "learning_rate": 7.603326612903227e-05,
+      "loss": 1.7616,
+      "step": 9510
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8132728338241577,
+      "learning_rate": 7.600806451612905e-05,
+      "loss": 1.8743,
+      "step": 9520
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.4682512283325195,
+      "learning_rate": 7.598286290322582e-05,
+      "loss": 1.8317,
+      "step": 9530
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.235108494758606,
+      "learning_rate": 7.595766129032258e-05,
+      "loss": 1.7362,
+      "step": 9540
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0129059553146362,
+      "learning_rate": 7.593245967741936e-05,
+      "loss": 1.9333,
+      "step": 9550
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6444118022918701,
+      "learning_rate": 7.590725806451613e-05,
+      "loss": 1.8786,
+      "step": 9560
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4819804430007935,
+      "learning_rate": 7.58820564516129e-05,
+      "loss": 1.823,
+      "step": 9570
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6458874940872192,
+      "learning_rate": 7.585685483870968e-05,
+      "loss": 1.9134,
+      "step": 9580
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7791017293930054,
+      "learning_rate": 7.583165322580645e-05,
+      "loss": 1.9333,
+      "step": 9590
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.28146231174469,
+      "learning_rate": 7.580645161290323e-05,
+      "loss": 1.8255,
+      "step": 9600
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.184740424156189,
+      "learning_rate": 7.578125e-05,
+      "loss": 1.7972,
+      "step": 9610
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7499341368675232,
+      "learning_rate": 7.575604838709678e-05,
+      "loss": 1.9361,
+      "step": 9620
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6286578178405762,
+      "learning_rate": 7.573084677419355e-05,
+      "loss": 1.777,
+      "step": 9630
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8273155689239502,
+      "learning_rate": 7.570564516129033e-05,
+      "loss": 1.6948,
+      "step": 9640
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.3549530506134033,
+      "learning_rate": 7.56804435483871e-05,
+      "loss": 2.0361,
+      "step": 9650
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.058880090713501,
+      "learning_rate": 7.565524193548387e-05,
+      "loss": 2.1033,
+      "step": 9660
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.1664986610412598,
+      "learning_rate": 7.563004032258065e-05,
+      "loss": 1.9674,
+      "step": 9670
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8700339794158936,
+      "learning_rate": 7.560483870967742e-05,
+      "loss": 2.1172,
+      "step": 9680
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2231051921844482,
+      "learning_rate": 7.55796370967742e-05,
+      "loss": 1.6987,
+      "step": 9690
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8680616021156311,
+      "learning_rate": 7.555443548387097e-05,
+      "loss": 1.9195,
+      "step": 9700
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7628701329231262,
+      "learning_rate": 7.552923387096775e-05,
+      "loss": 1.9841,
+      "step": 9710
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.3420493602752686,
+      "learning_rate": 7.550403225806452e-05,
+      "loss": 1.8748,
+      "step": 9720
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9162524342536926,
+      "learning_rate": 7.54788306451613e-05,
+      "loss": 2.0915,
+      "step": 9730
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7694321274757385,
+      "learning_rate": 7.545362903225807e-05,
+      "loss": 1.8399,
+      "step": 9740
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2106355428695679,
+      "learning_rate": 7.542842741935484e-05,
+      "loss": 1.9049,
+      "step": 9750
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0796549320220947,
+      "learning_rate": 7.540322580645162e-05,
+      "loss": 1.9402,
+      "step": 9760
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5596855878829956,
+      "learning_rate": 7.537802419354839e-05,
+      "loss": 1.9815,
+      "step": 9770
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.614392638206482,
+      "learning_rate": 7.535282258064517e-05,
+      "loss": 1.8042,
+      "step": 9780
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8946829438209534,
+      "learning_rate": 7.532762096774194e-05,
+      "loss": 1.9675,
+      "step": 9790
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0529166460037231,
+      "learning_rate": 7.530241935483871e-05,
+      "loss": 1.7348,
+      "step": 9800
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8235224485397339,
+      "learning_rate": 7.527721774193549e-05,
+      "loss": 1.9533,
+      "step": 9810
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.471670150756836,
+      "learning_rate": 7.525201612903226e-05,
+      "loss": 1.883,
+      "step": 9820
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8439244031906128,
+      "learning_rate": 7.522681451612904e-05,
+      "loss": 1.8373,
+      "step": 9830
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5398118495941162,
+      "learning_rate": 7.520161290322581e-05,
+      "loss": 1.8839,
+      "step": 9840
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.349800705909729,
+      "learning_rate": 7.517641129032257e-05,
+      "loss": 1.8407,
+      "step": 9850
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.775165319442749,
+      "learning_rate": 7.515120967741936e-05,
+      "loss": 1.7827,
+      "step": 9860
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.3521006107330322,
+      "learning_rate": 7.512600806451614e-05,
+      "loss": 1.7038,
+      "step": 9870
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.0576975345611572,
+      "learning_rate": 7.510080645161291e-05,
+      "loss": 1.8171,
+      "step": 9880
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0460199117660522,
+      "learning_rate": 7.507560483870968e-05,
+      "loss": 2.0841,
+      "step": 9890
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9272919297218323,
+      "learning_rate": 7.505040322580646e-05,
+      "loss": 1.8756,
+      "step": 9900
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.462572455406189,
+      "learning_rate": 7.502520161290323e-05,
+      "loss": 1.8104,
+      "step": 9910
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8769551515579224,
+      "learning_rate": 7.500000000000001e-05,
+      "loss": 1.9582,
+      "step": 9920
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.779557228088379,
+      "learning_rate": 7.497479838709678e-05,
+      "loss": 1.8365,
+      "step": 9930
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.211576461791992,
+      "learning_rate": 7.494959677419356e-05,
+      "loss": 1.9159,
+      "step": 9940
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8648312091827393,
+      "learning_rate": 7.492439516129033e-05,
+      "loss": 1.7639,
+      "step": 9950
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0310691595077515,
+      "learning_rate": 7.48991935483871e-05,
+      "loss": 1.8515,
+      "step": 9960
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.625552177429199,
+      "learning_rate": 7.487399193548388e-05,
+      "loss": 1.9363,
+      "step": 9970
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8627012372016907,
+      "learning_rate": 7.484879032258065e-05,
+      "loss": 1.6712,
+      "step": 9980
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0743408203125,
+      "learning_rate": 7.482358870967743e-05,
+      "loss": 1.8753,
+      "step": 9990
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2401208877563477,
+      "learning_rate": 7.479838709677419e-05,
+      "loss": 1.8753,
+      "step": 10000
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1524460315704346,
+      "learning_rate": 7.477318548387096e-05,
+      "loss": 1.7901,
+      "step": 10010
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4628137350082397,
+      "learning_rate": 7.474798387096774e-05,
+      "loss": 1.8959,
+      "step": 10020
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.521780014038086,
+      "learning_rate": 7.472278225806451e-05,
+      "loss": 1.6689,
+      "step": 10030
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6709047555923462,
+      "learning_rate": 7.469758064516129e-05,
+      "loss": 1.8882,
+      "step": 10040
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6719071865081787,
+      "learning_rate": 7.467237903225806e-05,
+      "loss": 1.9224,
+      "step": 10050
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2497707605361938,
+      "learning_rate": 7.464717741935485e-05,
+      "loss": 1.8939,
+      "step": 10060
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5659939050674438,
+      "learning_rate": 7.462197580645162e-05,
+      "loss": 1.6664,
+      "step": 10070
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6403392553329468,
+      "learning_rate": 7.45967741935484e-05,
+      "loss": 1.8292,
+      "step": 10080
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.2260873317718506,
+      "learning_rate": 7.457157258064517e-05,
+      "loss": 1.9228,
+      "step": 10090
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3463929891586304,
+      "learning_rate": 7.454637096774195e-05,
+      "loss": 1.8481,
+      "step": 10100
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1305840015411377,
+      "learning_rate": 7.452116935483872e-05,
+      "loss": 1.8977,
+      "step": 10110
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.895838737487793,
+      "learning_rate": 7.44959677419355e-05,
+      "loss": 1.991,
+      "step": 10120
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8507660031318665,
+      "learning_rate": 7.447076612903227e-05,
+      "loss": 1.8537,
+      "step": 10130
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.902208685874939,
+      "learning_rate": 7.444556451612904e-05,
+      "loss": 1.8146,
+      "step": 10140
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0858988761901855,
+      "learning_rate": 7.442036290322582e-05,
+      "loss": 1.7325,
+      "step": 10150
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6576449871063232,
+      "learning_rate": 7.439516129032258e-05,
+      "loss": 1.9955,
+      "step": 10160
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.199461817741394,
+      "learning_rate": 7.436995967741935e-05,
+      "loss": 1.7803,
+      "step": 10170
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1007906198501587,
+      "learning_rate": 7.434475806451613e-05,
+      "loss": 1.6321,
+      "step": 10180
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2700130939483643,
+      "learning_rate": 7.43195564516129e-05,
+      "loss": 1.7929,
+      "step": 10190
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2538598775863647,
+      "learning_rate": 7.429435483870968e-05,
+      "loss": 1.9231,
+      "step": 10200
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0397576093673706,
+      "learning_rate": 7.426915322580645e-05,
+      "loss": 1.6767,
+      "step": 10210
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7860151529312134,
+      "learning_rate": 7.424395161290322e-05,
+      "loss": 1.7777,
+      "step": 10220
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.527735948562622,
+      "learning_rate": 7.421875e-05,
+      "loss": 1.6971,
+      "step": 10230
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3440361022949219,
+      "learning_rate": 7.419354838709677e-05,
+      "loss": 2.0144,
+      "step": 10240
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6217674016952515,
+      "learning_rate": 7.416834677419355e-05,
+      "loss": 1.9456,
+      "step": 10250
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9862768054008484,
+      "learning_rate": 7.414314516129033e-05,
+      "loss": 1.8543,
+      "step": 10260
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2742429971694946,
+      "learning_rate": 7.411794354838711e-05,
+      "loss": 1.9964,
+      "step": 10270
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.3765487670898438,
+      "learning_rate": 7.409274193548388e-05,
+      "loss": 1.7896,
+      "step": 10280
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.359153151512146,
+      "learning_rate": 7.406754032258066e-05,
+      "loss": 2.1231,
+      "step": 10290
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.378043532371521,
+      "learning_rate": 7.404233870967743e-05,
+      "loss": 2.0332,
+      "step": 10300
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3861174583435059,
+      "learning_rate": 7.401713709677419e-05,
+      "loss": 1.9415,
+      "step": 10310
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4377895593643188,
+      "learning_rate": 7.399193548387097e-05,
+      "loss": 1.9255,
+      "step": 10320
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4208258390426636,
+      "learning_rate": 7.396673387096774e-05,
+      "loss": 1.8206,
+      "step": 10330
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7571712732315063,
+      "learning_rate": 7.394153225806452e-05,
+      "loss": 1.9901,
+      "step": 10340
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.085892677307129,
+      "learning_rate": 7.391633064516129e-05,
+      "loss": 1.8318,
+      "step": 10350
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1833686828613281,
+      "learning_rate": 7.389112903225806e-05,
+      "loss": 1.8716,
+      "step": 10360
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0435190200805664,
+      "learning_rate": 7.386592741935484e-05,
+      "loss": 1.9181,
+      "step": 10370
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.091964840888977,
+      "learning_rate": 7.384072580645161e-05,
+      "loss": 1.7336,
+      "step": 10380
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.3888638019561768,
+      "learning_rate": 7.381552419354839e-05,
+      "loss": 1.9969,
+      "step": 10390
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9394458532333374,
+      "learning_rate": 7.379032258064516e-05,
+      "loss": 1.8328,
+      "step": 10400
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2971819639205933,
+      "learning_rate": 7.376512096774194e-05,
+      "loss": 1.8367,
+      "step": 10410
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8689670562744141,
+      "learning_rate": 7.373991935483871e-05,
+      "loss": 1.7614,
+      "step": 10420
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.405235767364502,
+      "learning_rate": 7.371471774193549e-05,
+      "loss": 1.77,
+      "step": 10430
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2079178094863892,
+      "learning_rate": 7.368951612903226e-05,
+      "loss": 1.9153,
+      "step": 10440
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2022978067398071,
+      "learning_rate": 7.366431451612903e-05,
+      "loss": 1.7992,
+      "step": 10450
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.045955777168274,
+      "learning_rate": 7.363911290322582e-05,
+      "loss": 1.7642,
+      "step": 10460
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4850142002105713,
+      "learning_rate": 7.361391129032258e-05,
+      "loss": 1.8605,
+      "step": 10470
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2815362215042114,
+      "learning_rate": 7.358870967741936e-05,
+      "loss": 1.9579,
+      "step": 10480
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5780882835388184,
+      "learning_rate": 7.356350806451613e-05,
+      "loss": 1.8774,
+      "step": 10490
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.550321340560913,
+      "learning_rate": 7.35383064516129e-05,
+      "loss": 1.9,
+      "step": 10500
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2387356758117676,
+      "learning_rate": 7.351310483870968e-05,
+      "loss": 1.8111,
+      "step": 10510
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6553022861480713,
+      "learning_rate": 7.348790322580645e-05,
+      "loss": 1.7776,
+      "step": 10520
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9836485385894775,
+      "learning_rate": 7.346270161290323e-05,
+      "loss": 1.9516,
+      "step": 10530
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9291127920150757,
+      "learning_rate": 7.34375e-05,
+      "loss": 1.8927,
+      "step": 10540
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9961318969726562,
+      "learning_rate": 7.341229838709678e-05,
+      "loss": 1.8305,
+      "step": 10550
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8680959939956665,
+      "learning_rate": 7.338709677419355e-05,
+      "loss": 1.8369,
+      "step": 10560
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4202795028686523,
+      "learning_rate": 7.336189516129033e-05,
+      "loss": 2.0832,
+      "step": 10570
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.267809271812439,
+      "learning_rate": 7.33366935483871e-05,
+      "loss": 1.9724,
+      "step": 10580
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5115807056427002,
+      "learning_rate": 7.331149193548387e-05,
+      "loss": 1.9713,
+      "step": 10590
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6508578062057495,
+      "learning_rate": 7.328629032258065e-05,
+      "loss": 1.6883,
+      "step": 10600
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1911022663116455,
+      "learning_rate": 7.326108870967742e-05,
+      "loss": 1.9945,
+      "step": 10610
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2771891355514526,
+      "learning_rate": 7.32358870967742e-05,
+      "loss": 2.001,
+      "step": 10620
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.168134093284607,
+      "learning_rate": 7.321068548387097e-05,
+      "loss": 1.8031,
+      "step": 10630
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.6505117416381836,
+      "learning_rate": 7.318548387096775e-05,
+      "loss": 1.9711,
+      "step": 10640
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9227275252342224,
+      "learning_rate": 7.316028225806452e-05,
+      "loss": 2.0851,
+      "step": 10650
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2593010663986206,
+      "learning_rate": 7.31350806451613e-05,
+      "loss": 1.7902,
+      "step": 10660
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1032239198684692,
+      "learning_rate": 7.310987903225807e-05,
+      "loss": 1.8015,
+      "step": 10670
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.302917003631592,
+      "learning_rate": 7.308467741935484e-05,
+      "loss": 1.9125,
+      "step": 10680
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0144239664077759,
+      "learning_rate": 7.305947580645162e-05,
+      "loss": 1.9513,
+      "step": 10690
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9166842699050903,
+      "learning_rate": 7.303427419354839e-05,
+      "loss": 1.9659,
+      "step": 10700
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.637493371963501,
+      "learning_rate": 7.300907258064517e-05,
+      "loss": 1.7557,
+      "step": 10710
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.33362877368927,
+      "learning_rate": 7.298387096774194e-05,
+      "loss": 2.0417,
+      "step": 10720
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0469319820404053,
+      "learning_rate": 7.295866935483872e-05,
+      "loss": 1.8481,
+      "step": 10730
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.079763412475586,
+      "learning_rate": 7.293346774193549e-05,
+      "loss": 1.66,
+      "step": 10740
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7003172636032104,
+      "learning_rate": 7.290826612903226e-05,
+      "loss": 1.7483,
+      "step": 10750
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0826475620269775,
+      "learning_rate": 7.288306451612904e-05,
+      "loss": 1.6287,
+      "step": 10760
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0415453910827637,
+      "learning_rate": 7.285786290322581e-05,
+      "loss": 2.0058,
+      "step": 10770
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6321557760238647,
+      "learning_rate": 7.283266129032257e-05,
+      "loss": 1.7712,
+      "step": 10780
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6239466667175293,
+      "learning_rate": 7.280745967741935e-05,
+      "loss": 1.9716,
+      "step": 10790
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6424875259399414,
+      "learning_rate": 7.278225806451612e-05,
+      "loss": 1.963,
+      "step": 10800
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2557884454727173,
+      "learning_rate": 7.275705645161291e-05,
+      "loss": 1.7043,
+      "step": 10810
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9858161211013794,
+      "learning_rate": 7.273185483870968e-05,
+      "loss": 1.7141,
+      "step": 10820
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2142720222473145,
+      "learning_rate": 7.270665322580646e-05,
+      "loss": 1.7072,
+      "step": 10830
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4190329313278198,
+      "learning_rate": 7.268145161290323e-05,
+      "loss": 1.8724,
+      "step": 10840
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.236976385116577,
+      "learning_rate": 7.265625000000001e-05,
+      "loss": 1.8806,
+      "step": 10850
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.1303024291992188,
+      "learning_rate": 7.263104838709678e-05,
+      "loss": 1.7256,
+      "step": 10860
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6160517930984497,
+      "learning_rate": 7.260584677419356e-05,
+      "loss": 1.7868,
+      "step": 10870
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9327746629714966,
+      "learning_rate": 7.258064516129033e-05,
+      "loss": 2.0403,
+      "step": 10880
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.329038381576538,
+      "learning_rate": 7.25554435483871e-05,
+      "loss": 1.8094,
+      "step": 10890
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2968236207962036,
+      "learning_rate": 7.253024193548388e-05,
+      "loss": 1.8045,
+      "step": 10900
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0751770734786987,
+      "learning_rate": 7.250504032258065e-05,
+      "loss": 1.9987,
+      "step": 10910
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1865626573562622,
+      "learning_rate": 7.247983870967743e-05,
+      "loss": 1.6413,
+      "step": 10920
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1259900331497192,
+      "learning_rate": 7.245463709677419e-05,
+      "loss": 1.9766,
+      "step": 10930
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8146037459373474,
+      "learning_rate": 7.242943548387096e-05,
+      "loss": 1.8783,
+      "step": 10940
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6146631240844727,
+      "learning_rate": 7.240423387096774e-05,
+      "loss": 1.7746,
+      "step": 10950
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3488059043884277,
+      "learning_rate": 7.237903225806451e-05,
+      "loss": 1.9948,
+      "step": 10960
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5408573150634766,
+      "learning_rate": 7.235383064516129e-05,
+      "loss": 1.9362,
+      "step": 10970
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.737874150276184,
+      "learning_rate": 7.232862903225806e-05,
+      "loss": 1.6761,
+      "step": 10980
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9881741404533386,
+      "learning_rate": 7.230342741935483e-05,
+      "loss": 1.9686,
+      "step": 10990
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.297547459602356,
+      "learning_rate": 7.227822580645161e-05,
+      "loss": 1.8045,
+      "step": 11000
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3010973930358887,
+      "learning_rate": 7.22530241935484e-05,
+      "loss": 1.8826,
+      "step": 11010
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0601115226745605,
+      "learning_rate": 7.222782258064517e-05,
+      "loss": 1.8208,
+      "step": 11020
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6236106157302856,
+      "learning_rate": 7.220262096774195e-05,
+      "loss": 1.7262,
+      "step": 11030
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9448487758636475,
+      "learning_rate": 7.217741935483872e-05,
+      "loss": 1.8776,
+      "step": 11040
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1371604204177856,
+      "learning_rate": 7.21522177419355e-05,
+      "loss": 2.0896,
+      "step": 11050
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.397455096244812,
+      "learning_rate": 7.212701612903227e-05,
+      "loss": 1.7592,
+      "step": 11060
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8840553164482117,
+      "learning_rate": 7.210181451612904e-05,
+      "loss": 1.9038,
+      "step": 11070
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8205556869506836,
+      "learning_rate": 7.207661290322582e-05,
+      "loss": 1.7197,
+      "step": 11080
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8288129568099976,
+      "learning_rate": 7.205141129032258e-05,
+      "loss": 1.8571,
+      "step": 11090
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.995020627975464,
+      "learning_rate": 7.202620967741935e-05,
+      "loss": 1.9562,
+      "step": 11100
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2883529663085938,
+      "learning_rate": 7.200100806451613e-05,
+      "loss": 1.9829,
+      "step": 11110
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.923581600189209,
+      "learning_rate": 7.19758064516129e-05,
+      "loss": 1.729,
+      "step": 11120
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.674756646156311,
+      "learning_rate": 7.195060483870968e-05,
+      "loss": 1.7062,
+      "step": 11130
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9972915649414062,
+      "learning_rate": 7.192540322580645e-05,
+      "loss": 2.062,
+      "step": 11140
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0631355047225952,
+      "learning_rate": 7.190020161290322e-05,
+      "loss": 1.9824,
+      "step": 11150
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8068288564682007,
+      "learning_rate": 7.1875e-05,
+      "loss": 1.7409,
+      "step": 11160
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2534288167953491,
+      "learning_rate": 7.184979838709677e-05,
+      "loss": 1.6505,
+      "step": 11170
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3393688201904297,
+      "learning_rate": 7.182459677419355e-05,
+      "loss": 1.7121,
+      "step": 11180
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0412145853042603,
+      "learning_rate": 7.179939516129032e-05,
+      "loss": 1.6692,
+      "step": 11190
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0445919036865234,
+      "learning_rate": 7.177419354838711e-05,
+      "loss": 1.8185,
+      "step": 11200
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.768730401992798,
+      "learning_rate": 7.174899193548388e-05,
+      "loss": 1.8706,
+      "step": 11210
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5292152166366577,
+      "learning_rate": 7.172379032258066e-05,
+      "loss": 1.8638,
+      "step": 11220
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5459898710250854,
+      "learning_rate": 7.169858870967743e-05,
+      "loss": 1.9303,
+      "step": 11230
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.334681749343872,
+      "learning_rate": 7.16733870967742e-05,
+      "loss": 1.8338,
+      "step": 11240
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.973377525806427,
+      "learning_rate": 7.164818548387097e-05,
+      "loss": 1.8615,
+      "step": 11250
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3057364225387573,
+      "learning_rate": 7.162298387096774e-05,
+      "loss": 1.8389,
+      "step": 11260
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.080483317375183,
+      "learning_rate": 7.159778225806452e-05,
+      "loss": 1.8817,
+      "step": 11270
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.733896255493164,
+      "learning_rate": 7.157258064516129e-05,
+      "loss": 1.9784,
+      "step": 11280
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.615935206413269,
+      "learning_rate": 7.154737903225807e-05,
+      "loss": 1.716,
+      "step": 11290
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5532273054122925,
+      "learning_rate": 7.152217741935484e-05,
+      "loss": 1.7505,
+      "step": 11300
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4386087656021118,
+      "learning_rate": 7.149697580645161e-05,
+      "loss": 1.8308,
+      "step": 11310
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5767395496368408,
+      "learning_rate": 7.147177419354839e-05,
+      "loss": 1.7433,
+      "step": 11320
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0027694702148438,
+      "learning_rate": 7.144657258064516e-05,
+      "loss": 1.8423,
+      "step": 11330
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9200947284698486,
+      "learning_rate": 7.142137096774194e-05,
+      "loss": 1.8678,
+      "step": 11340
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.288964867591858,
+      "learning_rate": 7.139616935483871e-05,
+      "loss": 2.1504,
+      "step": 11350
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7829484939575195,
+      "learning_rate": 7.137096774193549e-05,
+      "loss": 2.0239,
+      "step": 11360
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9915928840637207,
+      "learning_rate": 7.134576612903226e-05,
+      "loss": 1.8495,
+      "step": 11370
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.070560336112976,
+      "learning_rate": 7.132056451612903e-05,
+      "loss": 1.6898,
+      "step": 11380
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.560875415802002,
+      "learning_rate": 7.129536290322581e-05,
+      "loss": 2.0235,
+      "step": 11390
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5667545795440674,
+      "learning_rate": 7.127016129032258e-05,
+      "loss": 1.9317,
+      "step": 11400
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.346531867980957,
+      "learning_rate": 7.124495967741936e-05,
+      "loss": 2.0561,
+      "step": 11410
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1890511512756348,
+      "learning_rate": 7.121975806451613e-05,
+      "loss": 2.0057,
+      "step": 11420
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7889693975448608,
+      "learning_rate": 7.11945564516129e-05,
+      "loss": 1.9552,
+      "step": 11430
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5332099199295044,
+      "learning_rate": 7.116935483870968e-05,
+      "loss": 1.8951,
+      "step": 11440
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1801767349243164,
+      "learning_rate": 7.114415322580645e-05,
+      "loss": 1.935,
+      "step": 11450
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0647538900375366,
+      "learning_rate": 7.111895161290323e-05,
+      "loss": 1.9159,
+      "step": 11460
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6137694120407104,
+      "learning_rate": 7.109375e-05,
+      "loss": 1.9235,
+      "step": 11470
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8797124624252319,
+      "learning_rate": 7.106854838709678e-05,
+      "loss": 1.9588,
+      "step": 11480
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0301696062088013,
+      "learning_rate": 7.104334677419355e-05,
+      "loss": 1.9717,
+      "step": 11490
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1188737154006958,
+      "learning_rate": 7.101814516129033e-05,
+      "loss": 1.7215,
+      "step": 11500
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4281500577926636,
+      "learning_rate": 7.09929435483871e-05,
+      "loss": 2.0179,
+      "step": 11510
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2050772905349731,
+      "learning_rate": 7.096774193548388e-05,
+      "loss": 1.8519,
+      "step": 11520
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1378766298294067,
+      "learning_rate": 7.094254032258065e-05,
+      "loss": 1.8466,
+      "step": 11530
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.244727611541748,
+      "learning_rate": 7.091733870967742e-05,
+      "loss": 1.7978,
+      "step": 11540
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2499024868011475,
+      "learning_rate": 7.08921370967742e-05,
+      "loss": 1.9069,
+      "step": 11550
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3092045783996582,
+      "learning_rate": 7.086693548387097e-05,
+      "loss": 1.647,
+      "step": 11560
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7527265548706055,
+      "learning_rate": 7.084173387096775e-05,
+      "loss": 1.9398,
+      "step": 11570
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1609185934066772,
+      "learning_rate": 7.081653225806452e-05,
+      "loss": 1.8538,
+      "step": 11580
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7768419981002808,
+      "learning_rate": 7.07913306451613e-05,
+      "loss": 1.8719,
+      "step": 11590
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9855479598045349,
+      "learning_rate": 7.076612903225807e-05,
+      "loss": 1.8025,
+      "step": 11600
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.085968017578125,
+      "learning_rate": 7.074092741935484e-05,
+      "loss": 1.9424,
+      "step": 11610
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5531903505325317,
+      "learning_rate": 7.071572580645162e-05,
+      "loss": 1.8508,
+      "step": 11620
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.055161714553833,
+      "learning_rate": 7.069052419354839e-05,
+      "loss": 1.9974,
+      "step": 11630
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8388280868530273,
+      "learning_rate": 7.066532258064517e-05,
+      "loss": 1.948,
+      "step": 11640
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9867286682128906,
+      "learning_rate": 7.064012096774194e-05,
+      "loss": 1.6704,
+      "step": 11650
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.80508553981781,
+      "learning_rate": 7.061491935483872e-05,
+      "loss": 1.9963,
+      "step": 11660
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4930161237716675,
+      "learning_rate": 7.058971774193549e-05,
+      "loss": 1.7649,
+      "step": 11670
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.109480381011963,
+      "learning_rate": 7.056451612903226e-05,
+      "loss": 1.9455,
+      "step": 11680
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5289627313613892,
+      "learning_rate": 7.053931451612904e-05,
+      "loss": 2.0494,
+      "step": 11690
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.329348087310791,
+      "learning_rate": 7.051411290322581e-05,
+      "loss": 1.908,
+      "step": 11700
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0013312101364136,
+      "learning_rate": 7.048891129032257e-05,
+      "loss": 1.8257,
+      "step": 11710
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9800733327865601,
+      "learning_rate": 7.046370967741935e-05,
+      "loss": 2.0271,
+      "step": 11720
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2686289548873901,
+      "learning_rate": 7.043850806451612e-05,
+      "loss": 1.9145,
+      "step": 11730
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5993425846099854,
+      "learning_rate": 7.04133064516129e-05,
+      "loss": 2.09,
+      "step": 11740
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8368712663650513,
+      "learning_rate": 7.038810483870969e-05,
+      "loss": 1.8404,
+      "step": 11750
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5898473262786865,
+      "learning_rate": 7.036290322580646e-05,
+      "loss": 1.783,
+      "step": 11760
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.294722318649292,
+      "learning_rate": 7.033770161290323e-05,
+      "loss": 1.709,
+      "step": 11770
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.1486189365386963,
+      "learning_rate": 7.031250000000001e-05,
+      "loss": 1.9242,
+      "step": 11780
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6221325397491455,
+      "learning_rate": 7.028729838709678e-05,
+      "loss": 1.7588,
+      "step": 11790
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5970109701156616,
+      "learning_rate": 7.026209677419356e-05,
+      "loss": 2.0379,
+      "step": 11800
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4521875381469727,
+      "learning_rate": 7.023689516129033e-05,
+      "loss": 1.6019,
+      "step": 11810
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1914833784103394,
+      "learning_rate": 7.02116935483871e-05,
+      "loss": 1.8968,
+      "step": 11820
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2371256351470947,
+      "learning_rate": 7.018649193548388e-05,
+      "loss": 2.0837,
+      "step": 11830
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7636529207229614,
+      "learning_rate": 7.016129032258065e-05,
+      "loss": 1.9857,
+      "step": 11840
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2903391122817993,
+      "learning_rate": 7.013608870967743e-05,
+      "loss": 1.8987,
+      "step": 11850
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.076730489730835,
+      "learning_rate": 7.011088709677419e-05,
+      "loss": 1.8818,
+      "step": 11860
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2003886699676514,
+      "learning_rate": 7.008568548387096e-05,
+      "loss": 1.8239,
+      "step": 11870
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2876148223876953,
+      "learning_rate": 7.006048387096774e-05,
+      "loss": 2.0276,
+      "step": 11880
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.746787428855896,
+      "learning_rate": 7.003528225806451e-05,
+      "loss": 1.7366,
+      "step": 11890
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.776820421218872,
+      "learning_rate": 7.001008064516129e-05,
+      "loss": 1.8957,
+      "step": 11900
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1764178276062012,
+      "learning_rate": 6.998487903225806e-05,
+      "loss": 2.0216,
+      "step": 11910
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4343034029006958,
+      "learning_rate": 6.995967741935484e-05,
+      "loss": 1.962,
+      "step": 11920
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.7375974655151367,
+      "learning_rate": 6.993447580645161e-05,
+      "loss": 1.9606,
+      "step": 11930
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9962373375892639,
+      "learning_rate": 6.990927419354838e-05,
+      "loss": 1.8545,
+      "step": 11940
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5224099159240723,
+      "learning_rate": 6.988407258064517e-05,
+      "loss": 1.974,
+      "step": 11950
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.767914056777954,
+      "learning_rate": 6.985887096774195e-05,
+      "loss": 1.8449,
+      "step": 11960
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.212012767791748,
+      "learning_rate": 6.983366935483872e-05,
+      "loss": 1.8281,
+      "step": 11970
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4698213338851929,
+      "learning_rate": 6.98084677419355e-05,
+      "loss": 1.7512,
+      "step": 11980
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9111374616622925,
+      "learning_rate": 6.978326612903227e-05,
+      "loss": 1.8466,
+      "step": 11990
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9588603377342224,
+      "learning_rate": 6.975806451612904e-05,
+      "loss": 1.7691,
+      "step": 12000
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.691219687461853,
+      "learning_rate": 6.973286290322582e-05,
+      "loss": 1.8041,
+      "step": 12010
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.093907117843628,
+      "learning_rate": 6.970766129032258e-05,
+      "loss": 1.8264,
+      "step": 12020
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9696716666221619,
+      "learning_rate": 6.968245967741935e-05,
+      "loss": 1.7886,
+      "step": 12030
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1794687509536743,
+      "learning_rate": 6.965725806451613e-05,
+      "loss": 1.7628,
+      "step": 12040
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.2805538177490234,
+      "learning_rate": 6.96320564516129e-05,
+      "loss": 1.6393,
+      "step": 12050
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3442268371582031,
+      "learning_rate": 6.960685483870968e-05,
+      "loss": 1.8498,
+      "step": 12060
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5337408781051636,
+      "learning_rate": 6.958165322580645e-05,
+      "loss": 1.5995,
+      "step": 12070
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8546984791755676,
+      "learning_rate": 6.955645161290323e-05,
+      "loss": 1.7379,
+      "step": 12080
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9505982995033264,
+      "learning_rate": 6.953125e-05,
+      "loss": 1.8567,
+      "step": 12090
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1850559711456299,
+      "learning_rate": 6.950604838709677e-05,
+      "loss": 1.8139,
+      "step": 12100
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.114724040031433,
+      "learning_rate": 6.948084677419355e-05,
+      "loss": 1.9434,
+      "step": 12110
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0990252494812012,
+      "learning_rate": 6.945564516129032e-05,
+      "loss": 1.7744,
+      "step": 12120
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7206604480743408,
+      "learning_rate": 6.94304435483871e-05,
+      "loss": 1.8367,
+      "step": 12130
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5847938060760498,
+      "learning_rate": 6.940524193548387e-05,
+      "loss": 1.8007,
+      "step": 12140
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9605498313903809,
+      "learning_rate": 6.938004032258066e-05,
+      "loss": 1.8338,
+      "step": 12150
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2003575563430786,
+      "learning_rate": 6.935483870967743e-05,
+      "loss": 1.8687,
+      "step": 12160
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8709436655044556,
+      "learning_rate": 6.93296370967742e-05,
+      "loss": 1.8517,
+      "step": 12170
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6599655151367188,
+      "learning_rate": 6.930443548387097e-05,
+      "loss": 1.7612,
+      "step": 12180
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.676730990409851,
+      "learning_rate": 6.927923387096774e-05,
+      "loss": 1.9961,
+      "step": 12190
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1340019702911377,
+      "learning_rate": 6.925403225806452e-05,
+      "loss": 1.8291,
+      "step": 12200
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.66368567943573,
+      "learning_rate": 6.922883064516129e-05,
+      "loss": 1.8655,
+      "step": 12210
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0888582468032837,
+      "learning_rate": 6.920362903225807e-05,
+      "loss": 1.7523,
+      "step": 12220
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.1903347969055176,
+      "learning_rate": 6.917842741935484e-05,
+      "loss": 1.8529,
+      "step": 12230
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3997174501419067,
+      "learning_rate": 6.915322580645161e-05,
+      "loss": 1.7472,
+      "step": 12240
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9005929231643677,
+      "learning_rate": 6.912802419354839e-05,
+      "loss": 1.8596,
+      "step": 12250
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.281358242034912,
+      "learning_rate": 6.910282258064516e-05,
+      "loss": 1.9771,
+      "step": 12260
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.96073317527771,
+      "learning_rate": 6.907762096774194e-05,
+      "loss": 1.9555,
+      "step": 12270
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1910552978515625,
+      "learning_rate": 6.905241935483871e-05,
+      "loss": 1.9824,
+      "step": 12280
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3333872556686401,
+      "learning_rate": 6.902721774193549e-05,
+      "loss": 1.7205,
+      "step": 12290
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1823211908340454,
+      "learning_rate": 6.900201612903226e-05,
+      "loss": 1.7519,
+      "step": 12300
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7884671688079834,
+      "learning_rate": 6.897681451612904e-05,
+      "loss": 1.8491,
+      "step": 12310
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2166478633880615,
+      "learning_rate": 6.895161290322581e-05,
+      "loss": 1.6734,
+      "step": 12320
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5723527669906616,
+      "learning_rate": 6.892641129032258e-05,
+      "loss": 1.9056,
+      "step": 12330
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8078614473342896,
+      "learning_rate": 6.890120967741936e-05,
+      "loss": 1.8571,
+      "step": 12340
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8835658431053162,
+      "learning_rate": 6.887600806451613e-05,
+      "loss": 1.6797,
+      "step": 12350
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7144461870193481,
+      "learning_rate": 6.88508064516129e-05,
+      "loss": 1.9635,
+      "step": 12360
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9593693614006042,
+      "learning_rate": 6.882560483870968e-05,
+      "loss": 1.6255,
+      "step": 12370
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3720459938049316,
+      "learning_rate": 6.880040322580646e-05,
+      "loss": 1.8776,
+      "step": 12380
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2752355337142944,
+      "learning_rate": 6.877520161290323e-05,
+      "loss": 1.8895,
+      "step": 12390
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.507026433944702,
+      "learning_rate": 6.875e-05,
+      "loss": 1.942,
+      "step": 12400
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0261242389678955,
+      "learning_rate": 6.872479838709678e-05,
+      "loss": 2.027,
+      "step": 12410
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9700976014137268,
+      "learning_rate": 6.869959677419355e-05,
+      "loss": 1.9705,
+      "step": 12420
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3159949779510498,
+      "learning_rate": 6.867439516129033e-05,
+      "loss": 1.6986,
+      "step": 12430
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.026977777481079,
+      "learning_rate": 6.86491935483871e-05,
+      "loss": 1.6948,
+      "step": 12440
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3463066816329956,
+      "learning_rate": 6.862399193548388e-05,
+      "loss": 1.8173,
+      "step": 12450
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9068715572357178,
+      "learning_rate": 6.859879032258065e-05,
+      "loss": 1.9265,
+      "step": 12460
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6502794027328491,
+      "learning_rate": 6.857358870967742e-05,
+      "loss": 1.8641,
+      "step": 12470
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3020834922790527,
+      "learning_rate": 6.854838709677419e-05,
+      "loss": 1.9931,
+      "step": 12480
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2233835458755493,
+      "learning_rate": 6.852318548387096e-05,
+      "loss": 1.8722,
+      "step": 12490
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.050653100013733,
+      "learning_rate": 6.849798387096775e-05,
+      "loss": 1.9127,
+      "step": 12500
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.278433084487915,
+      "learning_rate": 6.847278225806452e-05,
+      "loss": 1.9664,
+      "step": 12510
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6180709600448608,
+      "learning_rate": 6.84475806451613e-05,
+      "loss": 1.9928,
+      "step": 12520
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.902551293373108,
+      "learning_rate": 6.842237903225807e-05,
+      "loss": 1.779,
+      "step": 12530
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8400206565856934,
+      "learning_rate": 6.839717741935484e-05,
+      "loss": 1.8393,
+      "step": 12540
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.2027809619903564,
+      "learning_rate": 6.837197580645162e-05,
+      "loss": 1.8386,
+      "step": 12550
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2373783588409424,
+      "learning_rate": 6.83467741935484e-05,
+      "loss": 1.8471,
+      "step": 12560
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.1601834297180176,
+      "learning_rate": 6.832157258064517e-05,
+      "loss": 1.9204,
+      "step": 12570
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8733639717102051,
+      "learning_rate": 6.829637096774194e-05,
+      "loss": 1.7858,
+      "step": 12580
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0231199264526367,
+      "learning_rate": 6.827116935483872e-05,
+      "loss": 1.7718,
+      "step": 12590
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1885108947753906,
+      "learning_rate": 6.824596774193549e-05,
+      "loss": 1.8189,
+      "step": 12600
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7379491329193115,
+      "learning_rate": 6.822076612903227e-05,
+      "loss": 1.7813,
+      "step": 12610
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2483265399932861,
+      "learning_rate": 6.819556451612904e-05,
+      "loss": 1.8026,
+      "step": 12620
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0078257322311401,
+      "learning_rate": 6.817036290322581e-05,
+      "loss": 1.8871,
+      "step": 12630
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9199090003967285,
+      "learning_rate": 6.814516129032257e-05,
+      "loss": 1.6998,
+      "step": 12640
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2180472612380981,
+      "learning_rate": 6.811995967741935e-05,
+      "loss": 1.679,
+      "step": 12650
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8399699330329895,
+      "learning_rate": 6.809475806451612e-05,
+      "loss": 1.7735,
+      "step": 12660
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1830953359603882,
+      "learning_rate": 6.80695564516129e-05,
+      "loss": 1.8793,
+      "step": 12670
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.082911491394043,
+      "learning_rate": 6.804435483870967e-05,
+      "loss": 1.8529,
+      "step": 12680
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9185598492622375,
+      "learning_rate": 6.801915322580646e-05,
+      "loss": 1.8374,
+      "step": 12690
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3871444463729858,
+      "learning_rate": 6.799395161290323e-05,
+      "loss": 1.7868,
+      "step": 12700
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.522283673286438,
+      "learning_rate": 6.796875000000001e-05,
+      "loss": 1.9108,
+      "step": 12710
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3497226238250732,
+      "learning_rate": 6.794354838709678e-05,
+      "loss": 1.7762,
+      "step": 12720
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2732009887695312,
+      "learning_rate": 6.791834677419356e-05,
+      "loss": 1.8748,
+      "step": 12730
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5650458335876465,
+      "learning_rate": 6.789314516129033e-05,
+      "loss": 1.81,
+      "step": 12740
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8593194484710693,
+      "learning_rate": 6.78679435483871e-05,
+      "loss": 1.8983,
+      "step": 12750
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8577291965484619,
+      "learning_rate": 6.784274193548388e-05,
+      "loss": 1.9781,
+      "step": 12760
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3770978450775146,
+      "learning_rate": 6.781754032258065e-05,
+      "loss": 1.8032,
+      "step": 12770
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5335994958877563,
+      "learning_rate": 6.779233870967743e-05,
+      "loss": 1.9065,
+      "step": 12780
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.316230058670044,
+      "learning_rate": 6.776713709677419e-05,
+      "loss": 1.6921,
+      "step": 12790
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0534660816192627,
+      "learning_rate": 6.774193548387096e-05,
+      "loss": 1.8761,
+      "step": 12800
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4060367345809937,
+      "learning_rate": 6.771673387096774e-05,
+      "loss": 1.7022,
+      "step": 12810
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3199714422225952,
+      "learning_rate": 6.769153225806451e-05,
+      "loss": 1.8485,
+      "step": 12820
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.246673822402954,
+      "learning_rate": 6.766633064516129e-05,
+      "loss": 1.8348,
+      "step": 12830
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3308781385421753,
+      "learning_rate": 6.764112903225806e-05,
+      "loss": 1.8923,
+      "step": 12840
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9936067461967468,
+      "learning_rate": 6.761592741935484e-05,
+      "loss": 1.893,
+      "step": 12850
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6638327836990356,
+      "learning_rate": 6.759072580645161e-05,
+      "loss": 1.7983,
+      "step": 12860
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.409193754196167,
+      "learning_rate": 6.756552419354838e-05,
+      "loss": 1.9869,
+      "step": 12870
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2176718711853027,
+      "learning_rate": 6.754032258064516e-05,
+      "loss": 1.6067,
+      "step": 12880
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2995597124099731,
+      "learning_rate": 6.751512096774195e-05,
+      "loss": 1.8848,
+      "step": 12890
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6550852060317993,
+      "learning_rate": 6.748991935483872e-05,
+      "loss": 1.9092,
+      "step": 12900
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3017640113830566,
+      "learning_rate": 6.74647177419355e-05,
+      "loss": 1.7748,
+      "step": 12910
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2994149923324585,
+      "learning_rate": 6.743951612903227e-05,
+      "loss": 2.0919,
+      "step": 12920
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.355568528175354,
+      "learning_rate": 6.741431451612904e-05,
+      "loss": 1.8638,
+      "step": 12930
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4439775943756104,
+      "learning_rate": 6.738911290322582e-05,
+      "loss": 1.7455,
+      "step": 12940
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1351503133773804,
+      "learning_rate": 6.736391129032258e-05,
+      "loss": 1.9713,
+      "step": 12950
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9282092452049255,
+      "learning_rate": 6.733870967741935e-05,
+      "loss": 1.783,
+      "step": 12960
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8611167073249817,
+      "learning_rate": 6.731350806451613e-05,
+      "loss": 1.8605,
+      "step": 12970
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7482936978340149,
+      "learning_rate": 6.72883064516129e-05,
+      "loss": 1.6127,
+      "step": 12980
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8680581450462341,
+      "learning_rate": 6.726310483870968e-05,
+      "loss": 1.8125,
+      "step": 12990
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4086521863937378,
+      "learning_rate": 6.723790322580645e-05,
+      "loss": 1.6894,
+      "step": 13000
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.443799376487732,
+      "learning_rate": 6.721270161290323e-05,
+      "loss": 1.7619,
+      "step": 13010
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0582664012908936,
+      "learning_rate": 6.71875e-05,
+      "loss": 1.9003,
+      "step": 13020
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8580490350723267,
+      "learning_rate": 6.716229838709677e-05,
+      "loss": 1.9042,
+      "step": 13030
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9821445345878601,
+      "learning_rate": 6.713709677419355e-05,
+      "loss": 1.7866,
+      "step": 13040
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9881524443626404,
+      "learning_rate": 6.711189516129032e-05,
+      "loss": 1.9231,
+      "step": 13050
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8757095336914062,
+      "learning_rate": 6.70866935483871e-05,
+      "loss": 2.0071,
+      "step": 13060
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.109405517578125,
+      "learning_rate": 6.706149193548387e-05,
+      "loss": 1.7831,
+      "step": 13070
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.8313121795654297,
+      "learning_rate": 6.703629032258065e-05,
+      "loss": 1.7152,
+      "step": 13080
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8933016061782837,
+      "learning_rate": 6.701108870967743e-05,
+      "loss": 2.002,
+      "step": 13090
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6244829893112183,
+      "learning_rate": 6.69858870967742e-05,
+      "loss": 1.6092,
+      "step": 13100
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4628348350524902,
+      "learning_rate": 6.696068548387097e-05,
+      "loss": 1.9714,
+      "step": 13110
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8556376099586487,
+      "learning_rate": 6.693548387096774e-05,
+      "loss": 1.7762,
+      "step": 13120
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7355230450630188,
+      "learning_rate": 6.691028225806452e-05,
+      "loss": 1.8724,
+      "step": 13130
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.023439884185791,
+      "learning_rate": 6.688508064516129e-05,
+      "loss": 1.7648,
+      "step": 13140
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.800388216972351,
+      "learning_rate": 6.685987903225807e-05,
+      "loss": 1.9816,
+      "step": 13150
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8793922066688538,
+      "learning_rate": 6.683467741935484e-05,
+      "loss": 2.0328,
+      "step": 13160
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4768919944763184,
+      "learning_rate": 6.680947580645162e-05,
+      "loss": 1.8968,
+      "step": 13170
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7551766633987427,
+      "learning_rate": 6.678427419354839e-05,
+      "loss": 1.7265,
+      "step": 13180
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7383824586868286,
+      "learning_rate": 6.675907258064516e-05,
+      "loss": 1.6903,
+      "step": 13190
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4035708904266357,
+      "learning_rate": 6.673387096774194e-05,
+      "loss": 1.7766,
+      "step": 13200
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9028908610343933,
+      "learning_rate": 6.670866935483871e-05,
+      "loss": 1.7834,
+      "step": 13210
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8789945244789124,
+      "learning_rate": 6.668346774193549e-05,
+      "loss": 1.7522,
+      "step": 13220
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1126465797424316,
+      "learning_rate": 6.665826612903226e-05,
+      "loss": 1.8813,
+      "step": 13230
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.039380431175232,
+      "learning_rate": 6.663306451612904e-05,
+      "loss": 2.0424,
+      "step": 13240
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6819100975990295,
+      "learning_rate": 6.660786290322581e-05,
+      "loss": 1.7655,
+      "step": 13250
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5268179178237915,
+      "learning_rate": 6.658266129032258e-05,
+      "loss": 1.8315,
+      "step": 13260
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.207871913909912,
+      "learning_rate": 6.655745967741936e-05,
+      "loss": 1.8882,
+      "step": 13270
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3912250995635986,
+      "learning_rate": 6.653225806451613e-05,
+      "loss": 1.8463,
+      "step": 13280
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8487692475318909,
+      "learning_rate": 6.650705645161291e-05,
+      "loss": 1.7092,
+      "step": 13290
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2718256711959839,
+      "learning_rate": 6.648185483870968e-05,
+      "loss": 1.9987,
+      "step": 13300
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4826438426971436,
+      "learning_rate": 6.645665322580646e-05,
+      "loss": 1.7879,
+      "step": 13310
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8242260217666626,
+      "learning_rate": 6.643145161290323e-05,
+      "loss": 1.7688,
+      "step": 13320
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4583097696304321,
+      "learning_rate": 6.640625e-05,
+      "loss": 1.7358,
+      "step": 13330
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4439942836761475,
+      "learning_rate": 6.638104838709678e-05,
+      "loss": 2.0092,
+      "step": 13340
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1287939548492432,
+      "learning_rate": 6.635584677419355e-05,
+      "loss": 2.0871,
+      "step": 13350
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1803815364837646,
+      "learning_rate": 6.633064516129033e-05,
+      "loss": 1.8747,
+      "step": 13360
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9769756197929382,
+      "learning_rate": 6.63054435483871e-05,
+      "loss": 2.0734,
+      "step": 13370
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0983750820159912,
+      "learning_rate": 6.628024193548388e-05,
+      "loss": 2.0129,
+      "step": 13380
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.046190857887268,
+      "learning_rate": 6.625504032258065e-05,
+      "loss": 1.909,
+      "step": 13390
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5492640733718872,
+      "learning_rate": 6.622983870967743e-05,
+      "loss": 1.8576,
+      "step": 13400
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1529231071472168,
+      "learning_rate": 6.620463709677419e-05,
+      "loss": 1.957,
+      "step": 13410
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6886099576950073,
+      "learning_rate": 6.617943548387096e-05,
+      "loss": 1.8069,
+      "step": 13420
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.149436593055725,
+      "learning_rate": 6.615423387096773e-05,
+      "loss": 1.6817,
+      "step": 13430
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5723501443862915,
+      "learning_rate": 6.612903225806452e-05,
+      "loss": 1.9526,
+      "step": 13440
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1436060667037964,
+      "learning_rate": 6.61038306451613e-05,
+      "loss": 2.0108,
+      "step": 13450
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3986750841140747,
+      "learning_rate": 6.607862903225807e-05,
+      "loss": 1.6833,
+      "step": 13460
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7954114079475403,
+      "learning_rate": 6.605342741935485e-05,
+      "loss": 1.8428,
+      "step": 13470
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7256183624267578,
+      "learning_rate": 6.602822580645162e-05,
+      "loss": 1.9861,
+      "step": 13480
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.260638952255249,
+      "learning_rate": 6.60030241935484e-05,
+      "loss": 1.7831,
+      "step": 13490
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1471679210662842,
+      "learning_rate": 6.597782258064517e-05,
+      "loss": 1.8164,
+      "step": 13500
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8574875593185425,
+      "learning_rate": 6.595262096774194e-05,
+      "loss": 1.823,
+      "step": 13510
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.607849359512329,
+      "learning_rate": 6.592741935483872e-05,
+      "loss": 1.7206,
+      "step": 13520
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2598530054092407,
+      "learning_rate": 6.590221774193549e-05,
+      "loss": 1.6554,
+      "step": 13530
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9855432510375977,
+      "learning_rate": 6.587701612903227e-05,
+      "loss": 1.8353,
+      "step": 13540
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9563937187194824,
+      "learning_rate": 6.585181451612904e-05,
+      "loss": 1.9322,
+      "step": 13550
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8134301900863647,
+      "learning_rate": 6.582661290322581e-05,
+      "loss": 1.8483,
+      "step": 13560
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9947948455810547,
+      "learning_rate": 6.580141129032258e-05,
+      "loss": 1.8275,
+      "step": 13570
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8335902690887451,
+      "learning_rate": 6.577620967741935e-05,
+      "loss": 1.7695,
+      "step": 13580
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7899961471557617,
+      "learning_rate": 6.575100806451612e-05,
+      "loss": 1.8506,
+      "step": 13590
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.0352795124053955,
+      "learning_rate": 6.57258064516129e-05,
+      "loss": 2.0692,
+      "step": 13600
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5760895013809204,
+      "learning_rate": 6.570060483870967e-05,
+      "loss": 1.7338,
+      "step": 13610
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7633911371231079,
+      "learning_rate": 6.567540322580645e-05,
+      "loss": 1.7699,
+      "step": 13620
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7278303503990173,
+      "learning_rate": 6.565020161290322e-05,
+      "loss": 1.9069,
+      "step": 13630
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0519764423370361,
+      "learning_rate": 6.562500000000001e-05,
+      "loss": 1.829,
+      "step": 13640
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2277477979660034,
+      "learning_rate": 6.559979838709678e-05,
+      "loss": 1.7504,
+      "step": 13650
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6978176236152649,
+      "learning_rate": 6.557459677419356e-05,
+      "loss": 2.0109,
+      "step": 13660
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.111552119255066,
+      "learning_rate": 6.554939516129033e-05,
+      "loss": 1.8642,
+      "step": 13670
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6155098676681519,
+      "learning_rate": 6.552419354838711e-05,
+      "loss": 1.9757,
+      "step": 13680
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3474242687225342,
+      "learning_rate": 6.549899193548388e-05,
+      "loss": 1.8481,
+      "step": 13690
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1709973812103271,
+      "learning_rate": 6.547379032258066e-05,
+      "loss": 2.0334,
+      "step": 13700
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6990381479263306,
+      "learning_rate": 6.544858870967743e-05,
+      "loss": 1.7905,
+      "step": 13710
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.513744831085205,
+      "learning_rate": 6.542338709677419e-05,
+      "loss": 1.7868,
+      "step": 13720
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6507090330123901,
+      "learning_rate": 6.539818548387097e-05,
+      "loss": 1.6654,
+      "step": 13730
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4031784534454346,
+      "learning_rate": 6.537298387096774e-05,
+      "loss": 1.7592,
+      "step": 13740
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.313443899154663,
+      "learning_rate": 6.534778225806451e-05,
+      "loss": 1.852,
+      "step": 13750
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.868640661239624,
+      "learning_rate": 6.532258064516129e-05,
+      "loss": 1.8458,
+      "step": 13760
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.087459683418274,
+      "learning_rate": 6.529737903225806e-05,
+      "loss": 1.9059,
+      "step": 13770
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1789077520370483,
+      "learning_rate": 6.527217741935484e-05,
+      "loss": 1.9196,
+      "step": 13780
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0378360748291016,
+      "learning_rate": 6.524697580645161e-05,
+      "loss": 1.9901,
+      "step": 13790
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3074514865875244,
+      "learning_rate": 6.522177419354839e-05,
+      "loss": 1.7598,
+      "step": 13800
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6984933614730835,
+      "learning_rate": 6.519657258064516e-05,
+      "loss": 1.6557,
+      "step": 13810
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3897171020507812,
+      "learning_rate": 6.517137096774193e-05,
+      "loss": 1.8386,
+      "step": 13820
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5426668524742126,
+      "learning_rate": 6.514616935483872e-05,
+      "loss": 1.691,
+      "step": 13830
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9101991057395935,
+      "learning_rate": 6.51209677419355e-05,
+      "loss": 1.9934,
+      "step": 13840
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9987155199050903,
+      "learning_rate": 6.509576612903227e-05,
+      "loss": 1.8598,
+      "step": 13850
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0651723146438599,
+      "learning_rate": 6.507056451612905e-05,
+      "loss": 1.7962,
+      "step": 13860
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8485960960388184,
+      "learning_rate": 6.504536290322582e-05,
+      "loss": 1.714,
+      "step": 13870
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3432931900024414,
+      "learning_rate": 6.502016129032258e-05,
+      "loss": 1.8296,
+      "step": 13880
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0966196060180664,
+      "learning_rate": 6.499495967741935e-05,
+      "loss": 1.9811,
+      "step": 13890
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6620423793792725,
+      "learning_rate": 6.496975806451613e-05,
+      "loss": 1.7645,
+      "step": 13900
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4594101905822754,
+      "learning_rate": 6.49445564516129e-05,
+      "loss": 1.9823,
+      "step": 13910
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5605895519256592,
+      "learning_rate": 6.491935483870968e-05,
+      "loss": 1.9856,
+      "step": 13920
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5036343336105347,
+      "learning_rate": 6.489415322580645e-05,
+      "loss": 1.7541,
+      "step": 13930
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3929985761642456,
+      "learning_rate": 6.486895161290323e-05,
+      "loss": 1.9056,
+      "step": 13940
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.178691029548645,
+      "learning_rate": 6.484375e-05,
+      "loss": 1.7862,
+      "step": 13950
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2686207294464111,
+      "learning_rate": 6.481854838709678e-05,
+      "loss": 1.9383,
+      "step": 13960
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7212289571762085,
+      "learning_rate": 6.479334677419355e-05,
+      "loss": 1.8725,
+      "step": 13970
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5338246822357178,
+      "learning_rate": 6.476814516129032e-05,
+      "loss": 1.7798,
+      "step": 13980
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.591200351715088,
+      "learning_rate": 6.47429435483871e-05,
+      "loss": 1.7955,
+      "step": 13990
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9218335151672363,
+      "learning_rate": 6.471774193548387e-05,
+      "loss": 1.8884,
+      "step": 14000
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0533256530761719,
+      "learning_rate": 6.469254032258065e-05,
+      "loss": 1.8321,
+      "step": 14010
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2208746671676636,
+      "learning_rate": 6.466733870967742e-05,
+      "loss": 1.8725,
+      "step": 14020
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9557849168777466,
+      "learning_rate": 6.46421370967742e-05,
+      "loss": 1.8295,
+      "step": 14030
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6755704879760742,
+      "learning_rate": 6.461693548387097e-05,
+      "loss": 1.8012,
+      "step": 14040
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6014782786369324,
+      "learning_rate": 6.459173387096774e-05,
+      "loss": 1.7988,
+      "step": 14050
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.239531397819519,
+      "learning_rate": 6.456653225806452e-05,
+      "loss": 1.9833,
+      "step": 14060
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1874452829360962,
+      "learning_rate": 6.454133064516129e-05,
+      "loss": 1.8093,
+      "step": 14070
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4200916290283203,
+      "learning_rate": 6.451612903225807e-05,
+      "loss": 1.8718,
+      "step": 14080
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3773809671401978,
+      "learning_rate": 6.449092741935484e-05,
+      "loss": 1.5977,
+      "step": 14090
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7082947492599487,
+      "learning_rate": 6.446572580645162e-05,
+      "loss": 1.845,
+      "step": 14100
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8889748454093933,
+      "learning_rate": 6.444052419354839e-05,
+      "loss": 1.742,
+      "step": 14110
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6122761964797974,
+      "learning_rate": 6.441532258064516e-05,
+      "loss": 1.9149,
+      "step": 14120
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6642177104949951,
+      "learning_rate": 6.439012096774194e-05,
+      "loss": 1.8718,
+      "step": 14130
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5634392499923706,
+      "learning_rate": 6.436491935483871e-05,
+      "loss": 1.904,
+      "step": 14140
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9808370471000671,
+      "learning_rate": 6.433971774193549e-05,
+      "loss": 1.6781,
+      "step": 14150
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3594846725463867,
+      "learning_rate": 6.431451612903226e-05,
+      "loss": 1.8847,
+      "step": 14160
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.029052734375,
+      "learning_rate": 6.428931451612904e-05,
+      "loss": 1.7497,
+      "step": 14170
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8627665638923645,
+      "learning_rate": 6.426411290322581e-05,
+      "loss": 1.998,
+      "step": 14180
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.2316784858703613,
+      "learning_rate": 6.423891129032258e-05,
+      "loss": 1.8134,
+      "step": 14190
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8162317872047424,
+      "learning_rate": 6.421370967741936e-05,
+      "loss": 1.8269,
+      "step": 14200
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6994953155517578,
+      "learning_rate": 6.418850806451613e-05,
+      "loss": 1.7964,
+      "step": 14210
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3074219226837158,
+      "learning_rate": 6.416330645161291e-05,
+      "loss": 1.955,
+      "step": 14220
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3617773056030273,
+      "learning_rate": 6.413810483870968e-05,
+      "loss": 1.8955,
+      "step": 14230
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1213483810424805,
+      "learning_rate": 6.411290322580646e-05,
+      "loss": 1.661,
+      "step": 14240
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1667122840881348,
+      "learning_rate": 6.408770161290323e-05,
+      "loss": 1.9559,
+      "step": 14250
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8136681914329529,
+      "learning_rate": 6.40625e-05,
+      "loss": 1.8864,
+      "step": 14260
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0471352338790894,
+      "learning_rate": 6.403729838709678e-05,
+      "loss": 1.6617,
+      "step": 14270
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9656831622123718,
+      "learning_rate": 6.401209677419355e-05,
+      "loss": 1.8854,
+      "step": 14280
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4039716720581055,
+      "learning_rate": 6.398689516129033e-05,
+      "loss": 1.7456,
+      "step": 14290
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.313246965408325,
+      "learning_rate": 6.39616935483871e-05,
+      "loss": 2.1119,
+      "step": 14300
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5746735334396362,
+      "learning_rate": 6.393649193548388e-05,
+      "loss": 1.858,
+      "step": 14310
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.064955472946167,
+      "learning_rate": 6.391129032258065e-05,
+      "loss": 1.7551,
+      "step": 14320
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5821545124053955,
+      "learning_rate": 6.388608870967743e-05,
+      "loss": 1.8966,
+      "step": 14330
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8935563564300537,
+      "learning_rate": 6.386088709677419e-05,
+      "loss": 1.9443,
+      "step": 14340
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8073891401290894,
+      "learning_rate": 6.383568548387096e-05,
+      "loss": 1.9237,
+      "step": 14350
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.124988079071045,
+      "learning_rate": 6.381048387096774e-05,
+      "loss": 1.9013,
+      "step": 14360
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3875278234481812,
+      "learning_rate": 6.378528225806451e-05,
+      "loss": 2.015,
+      "step": 14370
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9059211611747742,
+      "learning_rate": 6.37600806451613e-05,
+      "loss": 2.0481,
+      "step": 14380
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.5179531574249268,
+      "learning_rate": 6.373487903225807e-05,
+      "loss": 1.909,
+      "step": 14390
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4911442995071411,
+      "learning_rate": 6.370967741935485e-05,
+      "loss": 1.9489,
+      "step": 14400
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.307727336883545,
+      "learning_rate": 6.368447580645162e-05,
+      "loss": 1.8612,
+      "step": 14410
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7466861009597778,
+      "learning_rate": 6.36592741935484e-05,
+      "loss": 1.6823,
+      "step": 14420
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7101143598556519,
+      "learning_rate": 6.363407258064517e-05,
+      "loss": 1.9004,
+      "step": 14430
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7482786178588867,
+      "learning_rate": 6.360887096774194e-05,
+      "loss": 1.6835,
+      "step": 14440
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2217681407928467,
+      "learning_rate": 6.358366935483872e-05,
+      "loss": 1.8762,
+      "step": 14450
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4669002294540405,
+      "learning_rate": 6.355846774193549e-05,
+      "loss": 1.9873,
+      "step": 14460
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8077692985534668,
+      "learning_rate": 6.353326612903227e-05,
+      "loss": 1.9449,
+      "step": 14470
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1962244510650635,
+      "learning_rate": 6.350806451612904e-05,
+      "loss": 1.807,
+      "step": 14480
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9862921237945557,
+      "learning_rate": 6.348286290322582e-05,
+      "loss": 1.9209,
+      "step": 14490
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.600436806678772,
+      "learning_rate": 6.345766129032258e-05,
+      "loss": 1.976,
+      "step": 14500
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6571576595306396,
+      "learning_rate": 6.343245967741935e-05,
+      "loss": 1.8258,
+      "step": 14510
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4544339179992676,
+      "learning_rate": 6.340725806451612e-05,
+      "loss": 1.6927,
+      "step": 14520
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3185324668884277,
+      "learning_rate": 6.33820564516129e-05,
+      "loss": 1.9287,
+      "step": 14530
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6504640579223633,
+      "learning_rate": 6.335685483870967e-05,
+      "loss": 1.9581,
+      "step": 14540
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1714420318603516,
+      "learning_rate": 6.333165322580645e-05,
+      "loss": 1.9196,
+      "step": 14550
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.7156760692596436,
+      "learning_rate": 6.330645161290322e-05,
+      "loss": 1.941,
+      "step": 14560
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9088534116744995,
+      "learning_rate": 6.328125e-05,
+      "loss": 1.8861,
+      "step": 14570
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.3305184841156006,
+      "learning_rate": 6.325604838709678e-05,
+      "loss": 1.7442,
+      "step": 14580
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1093741655349731,
+      "learning_rate": 6.323084677419356e-05,
+      "loss": 2.0244,
+      "step": 14590
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2397828102111816,
+      "learning_rate": 6.320564516129033e-05,
+      "loss": 1.9185,
+      "step": 14600
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9480420351028442,
+      "learning_rate": 6.318044354838711e-05,
+      "loss": 1.9029,
+      "step": 14610
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9392354488372803,
+      "learning_rate": 6.315524193548388e-05,
+      "loss": 2.0102,
+      "step": 14620
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5383937358856201,
+      "learning_rate": 6.313004032258066e-05,
+      "loss": 1.992,
+      "step": 14630
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0945026874542236,
+      "learning_rate": 6.310483870967743e-05,
+      "loss": 1.7185,
+      "step": 14640
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4119205474853516,
+      "learning_rate": 6.307963709677419e-05,
+      "loss": 1.6955,
+      "step": 14650
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.178303599357605,
+      "learning_rate": 6.305443548387097e-05,
+      "loss": 1.9582,
+      "step": 14660
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3106293678283691,
+      "learning_rate": 6.302923387096774e-05,
+      "loss": 1.9247,
+      "step": 14670
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5431506633758545,
+      "learning_rate": 6.300403225806451e-05,
+      "loss": 1.7891,
+      "step": 14680
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6490691900253296,
+      "learning_rate": 6.297883064516129e-05,
+      "loss": 1.86,
+      "step": 14690
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1694399118423462,
+      "learning_rate": 6.295362903225806e-05,
+      "loss": 1.7211,
+      "step": 14700
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.000266432762146,
+      "learning_rate": 6.292842741935484e-05,
+      "loss": 1.7362,
+      "step": 14710
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.908201277256012,
+      "learning_rate": 6.290322580645161e-05,
+      "loss": 1.7997,
+      "step": 14720
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2315220832824707,
+      "learning_rate": 6.287802419354839e-05,
+      "loss": 1.8203,
+      "step": 14730
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.514818549156189,
+      "learning_rate": 6.285282258064516e-05,
+      "loss": 1.8911,
+      "step": 14740
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7507808208465576,
+      "learning_rate": 6.282762096774193e-05,
+      "loss": 1.5809,
+      "step": 14750
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0080187320709229,
+      "learning_rate": 6.280241935483871e-05,
+      "loss": 1.7301,
+      "step": 14760
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6755834817886353,
+      "learning_rate": 6.277721774193548e-05,
+      "loss": 1.8899,
+      "step": 14770
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.6012861728668213,
+      "learning_rate": 6.275201612903227e-05,
+      "loss": 1.8381,
+      "step": 14780
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.00443434715271,
+      "learning_rate": 6.272681451612905e-05,
+      "loss": 1.9224,
+      "step": 14790
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.285569190979004,
+      "learning_rate": 6.270161290322582e-05,
+      "loss": 1.9958,
+      "step": 14800
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8543754816055298,
+      "learning_rate": 6.267641129032258e-05,
+      "loss": 1.9237,
+      "step": 14810
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9876225590705872,
+      "learning_rate": 6.265120967741936e-05,
+      "loss": 1.9763,
+      "step": 14820
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.1380581855773926,
+      "learning_rate": 6.262600806451613e-05,
+      "loss": 1.6844,
+      "step": 14830
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.305342674255371,
+      "learning_rate": 6.26008064516129e-05,
+      "loss": 1.7501,
+      "step": 14840
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8257023096084595,
+      "learning_rate": 6.257560483870968e-05,
+      "loss": 1.853,
+      "step": 14850
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5862764120101929,
+      "learning_rate": 6.255040322580645e-05,
+      "loss": 1.8213,
+      "step": 14860
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.735990047454834,
+      "learning_rate": 6.252520161290323e-05,
+      "loss": 1.8636,
+      "step": 14870
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.285215377807617,
+      "learning_rate": 6.25e-05,
+      "loss": 1.9311,
+      "step": 14880
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1204233169555664,
+      "learning_rate": 6.247479838709678e-05,
+      "loss": 1.9429,
+      "step": 14890
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.00249445438385,
+      "learning_rate": 6.244959677419355e-05,
+      "loss": 1.8789,
+      "step": 14900
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0878195762634277,
+      "learning_rate": 6.242439516129032e-05,
+      "loss": 1.9134,
+      "step": 14910
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.89115309715271,
+      "learning_rate": 6.23991935483871e-05,
+      "loss": 1.6996,
+      "step": 14920
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.27787446975708,
+      "learning_rate": 6.237399193548387e-05,
+      "loss": 1.7587,
+      "step": 14930
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2943001985549927,
+      "learning_rate": 6.234879032258065e-05,
+      "loss": 1.7093,
+      "step": 14940
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.8218297958374023,
+      "learning_rate": 6.232358870967742e-05,
+      "loss": 1.9149,
+      "step": 14950
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.287209987640381,
+      "learning_rate": 6.22983870967742e-05,
+      "loss": 1.7933,
+      "step": 14960
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0876914262771606,
+      "learning_rate": 6.227318548387097e-05,
+      "loss": 1.8345,
+      "step": 14970
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2807458639144897,
+      "learning_rate": 6.224798387096774e-05,
+      "loss": 1.7659,
+      "step": 14980
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1438300609588623,
+      "learning_rate": 6.222278225806452e-05,
+      "loss": 1.8939,
+      "step": 14990
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7212755084037781,
+      "learning_rate": 6.21975806451613e-05,
+      "loss": 1.6778,
+      "step": 15000
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2757841348648071,
+      "learning_rate": 6.217237903225807e-05,
+      "loss": 1.8884,
+      "step": 15010
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.725515365600586,
+      "learning_rate": 6.214717741935484e-05,
+      "loss": 2.1106,
+      "step": 15020
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6542632579803467,
+      "learning_rate": 6.212197580645162e-05,
+      "loss": 1.8605,
+      "step": 15030
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2230443954467773,
+      "learning_rate": 6.209677419354839e-05,
+      "loss": 1.8672,
+      "step": 15040
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4348273277282715,
+      "learning_rate": 6.207157258064517e-05,
+      "loss": 1.6847,
+      "step": 15050
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7463138103485107,
+      "learning_rate": 6.204637096774194e-05,
+      "loss": 1.896,
+      "step": 15060
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3837755918502808,
+      "learning_rate": 6.202116935483871e-05,
+      "loss": 1.9833,
+      "step": 15070
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1443698406219482,
+      "learning_rate": 6.199596774193549e-05,
+      "loss": 1.8443,
+      "step": 15080
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.018099308013916,
+      "learning_rate": 6.197076612903226e-05,
+      "loss": 1.7751,
+      "step": 15090
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.260897159576416,
+      "learning_rate": 6.194556451612904e-05,
+      "loss": 1.8926,
+      "step": 15100
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0969064235687256,
+      "learning_rate": 6.192036290322581e-05,
+      "loss": 1.8551,
+      "step": 15110
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.555351734161377,
+      "learning_rate": 6.189516129032257e-05,
+      "loss": 1.7883,
+      "step": 15120
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3094152212142944,
+      "learning_rate": 6.186995967741936e-05,
+      "loss": 1.8544,
+      "step": 15130
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.783996343612671,
+      "learning_rate": 6.184475806451613e-05,
+      "loss": 1.8963,
+      "step": 15140
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8172463178634644,
+      "learning_rate": 6.181955645161291e-05,
+      "loss": 1.807,
+      "step": 15150
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.015749931335449,
+      "learning_rate": 6.179435483870968e-05,
+      "loss": 1.9122,
+      "step": 15160
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.1991751194000244,
+      "learning_rate": 6.176915322580646e-05,
+      "loss": 1.768,
+      "step": 15170
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.0081417560577393,
+      "learning_rate": 6.174395161290323e-05,
+      "loss": 1.9196,
+      "step": 15180
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.122896432876587,
+      "learning_rate": 6.171875e-05,
+      "loss": 1.9894,
+      "step": 15190
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5572097301483154,
+      "learning_rate": 6.169354838709678e-05,
+      "loss": 1.8021,
+      "step": 15200
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3749744892120361,
+      "learning_rate": 6.166834677419355e-05,
+      "loss": 1.6866,
+      "step": 15210
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2325241565704346,
+      "learning_rate": 6.164314516129033e-05,
+      "loss": 1.9297,
+      "step": 15220
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.996680498123169,
+      "learning_rate": 6.16179435483871e-05,
+      "loss": 1.7837,
+      "step": 15230
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1807717084884644,
+      "learning_rate": 6.159274193548388e-05,
+      "loss": 2.0088,
+      "step": 15240
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0520846843719482,
+      "learning_rate": 6.156754032258065e-05,
+      "loss": 1.8264,
+      "step": 15250
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6649811267852783,
+      "learning_rate": 6.154233870967743e-05,
+      "loss": 1.9085,
+      "step": 15260
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9195472002029419,
+      "learning_rate": 6.151713709677419e-05,
+      "loss": 1.7049,
+      "step": 15270
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0356500148773193,
+      "learning_rate": 6.149193548387096e-05,
+      "loss": 1.8228,
+      "step": 15280
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.402671456336975,
+      "learning_rate": 6.146673387096774e-05,
+      "loss": 2.0769,
+      "step": 15290
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9310007095336914,
+      "learning_rate": 6.144153225806451e-05,
+      "loss": 1.9045,
+      "step": 15300
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.983824610710144,
+      "learning_rate": 6.141633064516128e-05,
+      "loss": 1.929,
+      "step": 15310
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1488289833068848,
+      "learning_rate": 6.139112903225806e-05,
+      "loss": 1.7599,
+      "step": 15320
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6099389791488647,
+      "learning_rate": 6.136592741935485e-05,
+      "loss": 1.8673,
+      "step": 15330
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9601927399635315,
+      "learning_rate": 6.134072580645162e-05,
+      "loss": 1.8434,
+      "step": 15340
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7734637260437012,
+      "learning_rate": 6.13155241935484e-05,
+      "loss": 1.7785,
+      "step": 15350
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.247165560722351,
+      "learning_rate": 6.129032258064517e-05,
+      "loss": 1.8413,
+      "step": 15360
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6431772708892822,
+      "learning_rate": 6.126512096774194e-05,
+      "loss": 1.7628,
+      "step": 15370
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9934298396110535,
+      "learning_rate": 6.123991935483872e-05,
+      "loss": 1.7276,
+      "step": 15380
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0093181133270264,
+      "learning_rate": 6.121471774193549e-05,
+      "loss": 1.6803,
+      "step": 15390
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4186115264892578,
+      "learning_rate": 6.118951612903227e-05,
+      "loss": 1.6528,
+      "step": 15400
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9310880899429321,
+      "learning_rate": 6.116431451612904e-05,
+      "loss": 1.9077,
+      "step": 15410
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.888521671295166,
+      "learning_rate": 6.113911290322582e-05,
+      "loss": 1.7922,
+      "step": 15420
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9873543977737427,
+      "learning_rate": 6.111391129032258e-05,
+      "loss": 1.7452,
+      "step": 15430
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6081303358078003,
+      "learning_rate": 6.108870967741935e-05,
+      "loss": 1.7698,
+      "step": 15440
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4444561004638672,
+      "learning_rate": 6.106350806451613e-05,
+      "loss": 1.6812,
+      "step": 15450
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5127744674682617,
+      "learning_rate": 6.10383064516129e-05,
+      "loss": 1.7853,
+      "step": 15460
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0388563871383667,
+      "learning_rate": 6.101310483870968e-05,
+      "loss": 1.6114,
+      "step": 15470
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0622758865356445,
+      "learning_rate": 6.0987903225806455e-05,
+      "loss": 1.9491,
+      "step": 15480
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6830378770828247,
+      "learning_rate": 6.096270161290323e-05,
+      "loss": 1.7194,
+      "step": 15490
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5108814239501953,
+      "learning_rate": 6.0937500000000004e-05,
+      "loss": 1.5899,
+      "step": 15500
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9548172950744629,
+      "learning_rate": 6.091229838709678e-05,
+      "loss": 1.9396,
+      "step": 15510
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3704253435134888,
+      "learning_rate": 6.088709677419355e-05,
+      "loss": 1.8028,
+      "step": 15520
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.914776623249054,
+      "learning_rate": 6.086189516129033e-05,
+      "loss": 1.8972,
+      "step": 15530
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.11908757686615,
+      "learning_rate": 6.08366935483871e-05,
+      "loss": 1.904,
+      "step": 15540
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2840378284454346,
+      "learning_rate": 6.0811491935483876e-05,
+      "loss": 1.7006,
+      "step": 15550
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6737542152404785,
+      "learning_rate": 6.078629032258065e-05,
+      "loss": 1.8208,
+      "step": 15560
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3364349603652954,
+      "learning_rate": 6.0761088709677424e-05,
+      "loss": 1.9473,
+      "step": 15570
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.009021520614624,
+      "learning_rate": 6.073588709677419e-05,
+      "loss": 1.9135,
+      "step": 15580
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2241828441619873,
+      "learning_rate": 6.0710685483870966e-05,
+      "loss": 1.7429,
+      "step": 15590
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8613426089286804,
+      "learning_rate": 6.068548387096774e-05,
+      "loss": 1.8668,
+      "step": 15600
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.848720908164978,
+      "learning_rate": 6.0660282258064515e-05,
+      "loss": 1.6329,
+      "step": 15610
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1922205686569214,
+      "learning_rate": 6.063508064516129e-05,
+      "loss": 1.7502,
+      "step": 15620
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4137698411941528,
+      "learning_rate": 6.0609879032258064e-05,
+      "loss": 1.7361,
+      "step": 15630
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4982850551605225,
+      "learning_rate": 6.058467741935484e-05,
+      "loss": 1.9175,
+      "step": 15640
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0946104526519775,
+      "learning_rate": 6.055947580645161e-05,
+      "loss": 1.8564,
+      "step": 15650
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1065142154693604,
+      "learning_rate": 6.0534274193548394e-05,
+      "loss": 1.6807,
+      "step": 15660
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8945730328559875,
+      "learning_rate": 6.050907258064517e-05,
+      "loss": 1.8786,
+      "step": 15670
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0110976696014404,
+      "learning_rate": 6.048387096774194e-05,
+      "loss": 1.9246,
+      "step": 15680
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.275103211402893,
+      "learning_rate": 6.0458669354838717e-05,
+      "loss": 1.8867,
+      "step": 15690
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8727425336837769,
+      "learning_rate": 6.043346774193549e-05,
+      "loss": 1.7834,
+      "step": 15700
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7498457431793213,
+      "learning_rate": 6.0408266129032265e-05,
+      "loss": 1.816,
+      "step": 15710
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.433864712715149,
+      "learning_rate": 6.038306451612904e-05,
+      "loss": 1.9366,
+      "step": 15720
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.985659122467041,
+      "learning_rate": 6.0357862903225814e-05,
+      "loss": 1.7787,
+      "step": 15730
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.0069832801818848,
+      "learning_rate": 6.033266129032258e-05,
+      "loss": 1.7909,
+      "step": 15740
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7948112487792969,
+      "learning_rate": 6.0307459677419356e-05,
+      "loss": 1.8095,
+      "step": 15750
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1293129920959473,
+      "learning_rate": 6.028225806451613e-05,
+      "loss": 1.755,
+      "step": 15760
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7022253274917603,
+      "learning_rate": 6.0257056451612905e-05,
+      "loss": 1.7221,
+      "step": 15770
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3486350774765015,
+      "learning_rate": 6.023185483870968e-05,
+      "loss": 1.6777,
+      "step": 15780
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8440837264060974,
+      "learning_rate": 6.020665322580645e-05,
+      "loss": 1.8387,
+      "step": 15790
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.167610764503479,
+      "learning_rate": 6.018145161290323e-05,
+      "loss": 1.7392,
+      "step": 15800
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2558748722076416,
+      "learning_rate": 6.015625e-05,
+      "loss": 1.843,
+      "step": 15810
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6610841751098633,
+      "learning_rate": 6.0131048387096776e-05,
+      "loss": 1.8634,
+      "step": 15820
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.560901403427124,
+      "learning_rate": 6.010584677419355e-05,
+      "loss": 1.8804,
+      "step": 15830
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.907558798789978,
+      "learning_rate": 6.0080645161290325e-05,
+      "loss": 1.786,
+      "step": 15840
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.721453905105591,
+      "learning_rate": 6.00554435483871e-05,
+      "loss": 1.7761,
+      "step": 15850
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2199782133102417,
+      "learning_rate": 6.003024193548388e-05,
+      "loss": 1.946,
+      "step": 15860
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8135555386543274,
+      "learning_rate": 6.0005040322580655e-05,
+      "loss": 1.9244,
+      "step": 15870
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.415285587310791,
+      "learning_rate": 5.997983870967743e-05,
+      "loss": 1.8777,
+      "step": 15880
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9285134077072144,
+      "learning_rate": 5.995463709677419e-05,
+      "loss": 1.886,
+      "step": 15890
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.373914122581482,
+      "learning_rate": 5.9929435483870964e-05,
+      "loss": 1.852,
+      "step": 15900
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.2505245208740234,
+      "learning_rate": 5.990423387096774e-05,
+      "loss": 1.9692,
+      "step": 15910
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1655501127243042,
+      "learning_rate": 5.987903225806451e-05,
+      "loss": 1.8368,
+      "step": 15920
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7870281338691711,
+      "learning_rate": 5.985383064516129e-05,
+      "loss": 1.8183,
+      "step": 15930
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2293602228164673,
+      "learning_rate": 5.982862903225807e-05,
+      "loss": 1.7622,
+      "step": 15940
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1006999015808105,
+      "learning_rate": 5.980342741935484e-05,
+      "loss": 1.5702,
+      "step": 15950
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1758613586425781,
+      "learning_rate": 5.977822580645162e-05,
+      "loss": 1.7183,
+      "step": 15960
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3495665788650513,
+      "learning_rate": 5.975302419354839e-05,
+      "loss": 1.8144,
+      "step": 15970
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.2604072093963623,
+      "learning_rate": 5.9727822580645166e-05,
+      "loss": 1.6883,
+      "step": 15980
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.0357437133789062,
+      "learning_rate": 5.970262096774194e-05,
+      "loss": 1.8138,
+      "step": 15990
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9550002813339233,
+      "learning_rate": 5.9677419354838715e-05,
+      "loss": 1.882,
+      "step": 16000
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.244924545288086,
+      "learning_rate": 5.965221774193549e-05,
+      "loss": 1.9652,
+      "step": 16010
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8491600155830383,
+      "learning_rate": 5.962701612903226e-05,
+      "loss": 1.6443,
+      "step": 16020
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.666952133178711,
+      "learning_rate": 5.960181451612904e-05,
+      "loss": 1.923,
+      "step": 16030
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2012766599655151,
+      "learning_rate": 5.957661290322581e-05,
+      "loss": 1.8874,
+      "step": 16040
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.1134257316589355,
+      "learning_rate": 5.955141129032258e-05,
+      "loss": 1.885,
+      "step": 16050
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.420135736465454,
+      "learning_rate": 5.9526209677419354e-05,
+      "loss": 1.9284,
+      "step": 16060
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3633850812911987,
+      "learning_rate": 5.950100806451613e-05,
+      "loss": 1.7035,
+      "step": 16070
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1123543977737427,
+      "learning_rate": 5.94758064516129e-05,
+      "loss": 1.9184,
+      "step": 16080
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5994455814361572,
+      "learning_rate": 5.945060483870968e-05,
+      "loss": 1.6381,
+      "step": 16090
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2592854499816895,
+      "learning_rate": 5.942540322580645e-05,
+      "loss": 1.86,
+      "step": 16100
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.331810474395752,
+      "learning_rate": 5.9400201612903226e-05,
+      "loss": 1.898,
+      "step": 16110
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.66259765625,
+      "learning_rate": 5.9375e-05,
+      "loss": 1.8632,
+      "step": 16120
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9403432607650757,
+      "learning_rate": 5.934979838709678e-05,
+      "loss": 1.7627,
+      "step": 16130
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7301186323165894,
+      "learning_rate": 5.9324596774193555e-05,
+      "loss": 1.9002,
+      "step": 16140
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.2293248176574707,
+      "learning_rate": 5.929939516129033e-05,
+      "loss": 1.8018,
+      "step": 16150
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.0337966680526733,
+      "learning_rate": 5.9274193548387104e-05,
+      "loss": 1.7484,
+      "step": 16160
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.0935182571411133,
+      "learning_rate": 5.924899193548388e-05,
+      "loss": 1.9137,
+      "step": 16170
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5941696166992188,
+      "learning_rate": 5.922379032258065e-05,
+      "loss": 1.9934,
+      "step": 16180
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.110309362411499,
+      "learning_rate": 5.919858870967743e-05,
+      "loss": 1.8563,
+      "step": 16190
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2092357873916626,
+      "learning_rate": 5.917338709677419e-05,
+      "loss": 1.781,
+      "step": 16200
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6538786888122559,
+      "learning_rate": 5.914818548387097e-05,
+      "loss": 1.7097,
+      "step": 16210
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9196608662605286,
+      "learning_rate": 5.912298387096774e-05,
+      "loss": 1.8848,
+      "step": 16220
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8155243396759033,
+      "learning_rate": 5.909778225806452e-05,
+      "loss": 1.6994,
+      "step": 16230
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2205146551132202,
+      "learning_rate": 5.907258064516129e-05,
+      "loss": 1.6979,
+      "step": 16240
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4621987342834473,
+      "learning_rate": 5.9047379032258066e-05,
+      "loss": 1.9102,
+      "step": 16250
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8238921165466309,
+      "learning_rate": 5.902217741935484e-05,
+      "loss": 1.643,
+      "step": 16260
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1681673526763916,
+      "learning_rate": 5.8996975806451615e-05,
+      "loss": 1.8157,
+      "step": 16270
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6401293873786926,
+      "learning_rate": 5.897177419354839e-05,
+      "loss": 1.7326,
+      "step": 16280
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5546239614486694,
+      "learning_rate": 5.8946572580645164e-05,
+      "loss": 1.7849,
+      "step": 16290
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3762476444244385,
+      "learning_rate": 5.892137096774194e-05,
+      "loss": 1.9053,
+      "step": 16300
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8129109144210815,
+      "learning_rate": 5.889616935483871e-05,
+      "loss": 1.9258,
+      "step": 16310
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9796985387802124,
+      "learning_rate": 5.887096774193549e-05,
+      "loss": 1.866,
+      "step": 16320
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3868162631988525,
+      "learning_rate": 5.884576612903227e-05,
+      "loss": 1.7098,
+      "step": 16330
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.79454505443573,
+      "learning_rate": 5.882056451612904e-05,
+      "loss": 1.9233,
+      "step": 16340
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7530170679092407,
+      "learning_rate": 5.879536290322582e-05,
+      "loss": 1.9342,
+      "step": 16350
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.60903799533844,
+      "learning_rate": 5.877016129032258e-05,
+      "loss": 1.8136,
+      "step": 16360
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6310821771621704,
+      "learning_rate": 5.874495967741935e-05,
+      "loss": 1.9156,
+      "step": 16370
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9356744289398193,
+      "learning_rate": 5.8719758064516126e-05,
+      "loss": 1.953,
+      "step": 16380
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8436790704727173,
+      "learning_rate": 5.86945564516129e-05,
+      "loss": 1.8232,
+      "step": 16390
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4925987720489502,
+      "learning_rate": 5.8669354838709675e-05,
+      "loss": 1.9892,
+      "step": 16400
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.897146224975586,
+      "learning_rate": 5.8644153225806456e-05,
+      "loss": 1.8664,
+      "step": 16410
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6222553253173828,
+      "learning_rate": 5.861895161290323e-05,
+      "loss": 1.6932,
+      "step": 16420
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0129344463348389,
+      "learning_rate": 5.8593750000000005e-05,
+      "loss": 1.8618,
+      "step": 16430
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2997008562088013,
+      "learning_rate": 5.856854838709678e-05,
+      "loss": 1.8996,
+      "step": 16440
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0054599046707153,
+      "learning_rate": 5.854334677419355e-05,
+      "loss": 1.9265,
+      "step": 16450
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.485474944114685,
+      "learning_rate": 5.851814516129033e-05,
+      "loss": 2.0074,
+      "step": 16460
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.487515926361084,
+      "learning_rate": 5.84929435483871e-05,
+      "loss": 1.7889,
+      "step": 16470
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2894760370254517,
+      "learning_rate": 5.8467741935483876e-05,
+      "loss": 1.6914,
+      "step": 16480
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.416994333267212,
+      "learning_rate": 5.844254032258065e-05,
+      "loss": 1.6385,
+      "step": 16490
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2708946466445923,
+      "learning_rate": 5.8417338709677425e-05,
+      "loss": 1.9223,
+      "step": 16500
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8931563496589661,
+      "learning_rate": 5.839213709677419e-05,
+      "loss": 1.8947,
+      "step": 16510
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2460523843765259,
+      "learning_rate": 5.836693548387097e-05,
+      "loss": 1.8721,
+      "step": 16520
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7285916209220886,
+      "learning_rate": 5.834173387096774e-05,
+      "loss": 1.9725,
+      "step": 16530
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8894128203392029,
+      "learning_rate": 5.8316532258064516e-05,
+      "loss": 1.6597,
+      "step": 16540
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2593612670898438,
+      "learning_rate": 5.829133064516129e-05,
+      "loss": 1.9174,
+      "step": 16550
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4841175079345703,
+      "learning_rate": 5.8266129032258064e-05,
+      "loss": 1.6818,
+      "step": 16560
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9330605268478394,
+      "learning_rate": 5.824092741935484e-05,
+      "loss": 1.8443,
+      "step": 16570
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1529515981674194,
+      "learning_rate": 5.821572580645161e-05,
+      "loss": 1.6952,
+      "step": 16580
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4964616298675537,
+      "learning_rate": 5.819052419354839e-05,
+      "loss": 1.7422,
+      "step": 16590
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4614460468292236,
+      "learning_rate": 5.816532258064517e-05,
+      "loss": 1.921,
+      "step": 16600
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9318482279777527,
+      "learning_rate": 5.814012096774194e-05,
+      "loss": 1.8387,
+      "step": 16610
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3643484115600586,
+      "learning_rate": 5.811491935483872e-05,
+      "loss": 1.9676,
+      "step": 16620
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.009186029434204,
+      "learning_rate": 5.808971774193549e-05,
+      "loss": 1.922,
+      "step": 16630
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2064968347549438,
+      "learning_rate": 5.8064516129032266e-05,
+      "loss": 1.8682,
+      "step": 16640
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5194593667984009,
+      "learning_rate": 5.803931451612904e-05,
+      "loss": 2.0479,
+      "step": 16650
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.311877965927124,
+      "learning_rate": 5.8014112903225815e-05,
+      "loss": 1.9047,
+      "step": 16660
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.464597225189209,
+      "learning_rate": 5.7988911290322575e-05,
+      "loss": 1.9115,
+      "step": 16670
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.696085810661316,
+      "learning_rate": 5.7963709677419356e-05,
+      "loss": 1.7917,
+      "step": 16680
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.4228858947753906,
+      "learning_rate": 5.793850806451613e-05,
+      "loss": 1.8009,
+      "step": 16690
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3683125972747803,
+      "learning_rate": 5.7913306451612905e-05,
+      "loss": 1.7094,
+      "step": 16700
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.438645601272583,
+      "learning_rate": 5.788810483870968e-05,
+      "loss": 1.885,
+      "step": 16710
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3746109008789062,
+      "learning_rate": 5.7862903225806454e-05,
+      "loss": 1.6643,
+      "step": 16720
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4415003061294556,
+      "learning_rate": 5.783770161290323e-05,
+      "loss": 1.856,
+      "step": 16730
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.47752046585083,
+      "learning_rate": 5.78125e-05,
+      "loss": 1.9574,
+      "step": 16740
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1161173582077026,
+      "learning_rate": 5.778729838709678e-05,
+      "loss": 1.9451,
+      "step": 16750
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4929100275039673,
+      "learning_rate": 5.776209677419355e-05,
+      "loss": 1.7514,
+      "step": 16760
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1349278688430786,
+      "learning_rate": 5.7736895161290326e-05,
+      "loss": 1.915,
+      "step": 16770
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6113014221191406,
+      "learning_rate": 5.77116935483871e-05,
+      "loss": 1.8151,
+      "step": 16780
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.824173629283905,
+      "learning_rate": 5.7686491935483874e-05,
+      "loss": 2.0428,
+      "step": 16790
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.228176236152649,
+      "learning_rate": 5.7661290322580655e-05,
+      "loss": 1.8683,
+      "step": 16800
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.266771912574768,
+      "learning_rate": 5.763608870967743e-05,
+      "loss": 2.0141,
+      "step": 16810
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4955586194992065,
+      "learning_rate": 5.761088709677419e-05,
+      "loss": 1.8589,
+      "step": 16820
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0474209785461426,
+      "learning_rate": 5.7585685483870965e-05,
+      "loss": 1.6532,
+      "step": 16830
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0398313999176025,
+      "learning_rate": 5.756048387096774e-05,
+      "loss": 1.7727,
+      "step": 16840
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1566379070281982,
+      "learning_rate": 5.7535282258064514e-05,
+      "loss": 1.8617,
+      "step": 16850
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0841624736785889,
+      "learning_rate": 5.751008064516129e-05,
+      "loss": 1.7486,
+      "step": 16860
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3228896856307983,
+      "learning_rate": 5.748487903225806e-05,
+      "loss": 2.1792,
+      "step": 16870
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7009594440460205,
+      "learning_rate": 5.7459677419354843e-05,
+      "loss": 1.8926,
+      "step": 16880
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8252977728843689,
+      "learning_rate": 5.743447580645162e-05,
+      "loss": 1.8756,
+      "step": 16890
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.4729483127593994,
+      "learning_rate": 5.740927419354839e-05,
+      "loss": 1.7816,
+      "step": 16900
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3692554235458374,
+      "learning_rate": 5.7384072580645166e-05,
+      "loss": 1.9446,
+      "step": 16910
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3653390407562256,
+      "learning_rate": 5.735887096774194e-05,
+      "loss": 1.8888,
+      "step": 16920
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2501511573791504,
+      "learning_rate": 5.7333669354838715e-05,
+      "loss": 2.0432,
+      "step": 16930
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2488518953323364,
+      "learning_rate": 5.730846774193549e-05,
+      "loss": 1.9069,
+      "step": 16940
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.744654655456543,
+      "learning_rate": 5.7283266129032264e-05,
+      "loss": 1.8313,
+      "step": 16950
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.334620475769043,
+      "learning_rate": 5.725806451612904e-05,
+      "loss": 1.8014,
+      "step": 16960
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.180234432220459,
+      "learning_rate": 5.723286290322581e-05,
+      "loss": 1.9129,
+      "step": 16970
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.499411940574646,
+      "learning_rate": 5.720766129032258e-05,
+      "loss": 1.7949,
+      "step": 16980
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5496389865875244,
+      "learning_rate": 5.7182459677419354e-05,
+      "loss": 1.9933,
+      "step": 16990
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5200172662734985,
+      "learning_rate": 5.715725806451613e-05,
+      "loss": 1.9621,
+      "step": 17000
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7246026992797852,
+      "learning_rate": 5.71320564516129e-05,
+      "loss": 1.8692,
+      "step": 17010
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5868839025497437,
+      "learning_rate": 5.710685483870968e-05,
+      "loss": 1.8918,
+      "step": 17020
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2328904867172241,
+      "learning_rate": 5.708165322580645e-05,
+      "loss": 1.808,
+      "step": 17030
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.506709337234497,
+      "learning_rate": 5.7056451612903226e-05,
+      "loss": 1.895,
+      "step": 17040
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.466723918914795,
+      "learning_rate": 5.703125e-05,
+      "loss": 1.6494,
+      "step": 17050
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.20452618598938,
+      "learning_rate": 5.7006048387096775e-05,
+      "loss": 1.7832,
+      "step": 17060
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7628130912780762,
+      "learning_rate": 5.698084677419355e-05,
+      "loss": 1.9409,
+      "step": 17070
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.413583755493164,
+      "learning_rate": 5.695564516129033e-05,
+      "loss": 2.0264,
+      "step": 17080
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.07462215423584,
+      "learning_rate": 5.6930443548387105e-05,
+      "loss": 1.9431,
+      "step": 17090
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9668834805488586,
+      "learning_rate": 5.690524193548388e-05,
+      "loss": 1.7639,
+      "step": 17100
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3588858842849731,
+      "learning_rate": 5.688004032258065e-05,
+      "loss": 1.9577,
+      "step": 17110
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2977176904678345,
+      "learning_rate": 5.685483870967743e-05,
+      "loss": 1.8903,
+      "step": 17120
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4815917015075684,
+      "learning_rate": 5.682963709677419e-05,
+      "loss": 2.0255,
+      "step": 17130
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7344715595245361,
+      "learning_rate": 5.680443548387096e-05,
+      "loss": 1.8168,
+      "step": 17140
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5629826784133911,
+      "learning_rate": 5.6779233870967744e-05,
+      "loss": 1.7715,
+      "step": 17150
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3727173805236816,
+      "learning_rate": 5.675403225806452e-05,
+      "loss": 1.7951,
+      "step": 17160
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.1836957931518555,
+      "learning_rate": 5.672883064516129e-05,
+      "loss": 1.8689,
+      "step": 17170
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.252232789993286,
+      "learning_rate": 5.670362903225807e-05,
+      "loss": 1.9807,
+      "step": 17180
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0217267274856567,
+      "learning_rate": 5.667842741935484e-05,
+      "loss": 1.8048,
+      "step": 17190
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2184579372406006,
+      "learning_rate": 5.6653225806451616e-05,
+      "loss": 1.8953,
+      "step": 17200
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2054996490478516,
+      "learning_rate": 5.662802419354839e-05,
+      "loss": 2.0564,
+      "step": 17210
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8358243703842163,
+      "learning_rate": 5.6602822580645164e-05,
+      "loss": 1.8307,
+      "step": 17220
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.38827383518219,
+      "learning_rate": 5.657762096774194e-05,
+      "loss": 1.841,
+      "step": 17230
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9021123647689819,
+      "learning_rate": 5.655241935483871e-05,
+      "loss": 1.6311,
+      "step": 17240
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8611418604850769,
+      "learning_rate": 5.652721774193549e-05,
+      "loss": 1.9572,
+      "step": 17250
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.566918969154358,
+      "learning_rate": 5.650201612903226e-05,
+      "loss": 1.7822,
+      "step": 17260
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.846107006072998,
+      "learning_rate": 5.647681451612904e-05,
+      "loss": 1.7204,
+      "step": 17270
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3673449754714966,
+      "learning_rate": 5.645161290322582e-05,
+      "loss": 1.7483,
+      "step": 17280
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7779127359390259,
+      "learning_rate": 5.642641129032258e-05,
+      "loss": 1.6993,
+      "step": 17290
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.38696551322937,
+      "learning_rate": 5.640120967741935e-05,
+      "loss": 1.8868,
+      "step": 17300
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9500272274017334,
+      "learning_rate": 5.637600806451613e-05,
+      "loss": 1.7881,
+      "step": 17310
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7991717457771301,
+      "learning_rate": 5.63508064516129e-05,
+      "loss": 1.732,
+      "step": 17320
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0568065643310547,
+      "learning_rate": 5.6325604838709675e-05,
+      "loss": 1.7424,
+      "step": 17330
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5661587715148926,
+      "learning_rate": 5.630040322580645e-05,
+      "loss": 1.7828,
+      "step": 17340
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8650426864624023,
+      "learning_rate": 5.627520161290323e-05,
+      "loss": 1.9262,
+      "step": 17350
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6337840557098389,
+      "learning_rate": 5.6250000000000005e-05,
+      "loss": 1.9365,
+      "step": 17360
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4116345643997192,
+      "learning_rate": 5.622479838709678e-05,
+      "loss": 1.7725,
+      "step": 17370
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.140702486038208,
+      "learning_rate": 5.6199596774193554e-05,
+      "loss": 1.7693,
+      "step": 17380
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.080690860748291,
+      "learning_rate": 5.617439516129033e-05,
+      "loss": 2.0142,
+      "step": 17390
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.96695476770401,
+      "learning_rate": 5.61491935483871e-05,
+      "loss": 1.7332,
+      "step": 17400
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2807554006576538,
+      "learning_rate": 5.612399193548388e-05,
+      "loss": 1.6596,
+      "step": 17410
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0806632041931152,
+      "learning_rate": 5.609879032258065e-05,
+      "loss": 1.6872,
+      "step": 17420
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5975972414016724,
+      "learning_rate": 5.6073588709677426e-05,
+      "loss": 1.7934,
+      "step": 17430
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5098443031311035,
+      "learning_rate": 5.604838709677419e-05,
+      "loss": 1.9321,
+      "step": 17440
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5198280811309814,
+      "learning_rate": 5.602318548387097e-05,
+      "loss": 1.8287,
+      "step": 17450
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3638832569122314,
+      "learning_rate": 5.599798387096774e-05,
+      "loss": 1.8455,
+      "step": 17460
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3368239402770996,
+      "learning_rate": 5.5972782258064516e-05,
+      "loss": 1.8999,
+      "step": 17470
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8732973337173462,
+      "learning_rate": 5.594758064516129e-05,
+      "loss": 1.914,
+      "step": 17480
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4149788618087769,
+      "learning_rate": 5.5922379032258065e-05,
+      "loss": 1.881,
+      "step": 17490
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2969204187393188,
+      "learning_rate": 5.589717741935484e-05,
+      "loss": 1.6883,
+      "step": 17500
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.1143152713775635,
+      "learning_rate": 5.5871975806451614e-05,
+      "loss": 1.6964,
+      "step": 17510
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9829666018486023,
+      "learning_rate": 5.584677419354839e-05,
+      "loss": 1.954,
+      "step": 17520
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.568890929222107,
+      "learning_rate": 5.582157258064516e-05,
+      "loss": 1.9049,
+      "step": 17530
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0054618120193481,
+      "learning_rate": 5.579637096774194e-05,
+      "loss": 1.8654,
+      "step": 17540
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8919579982757568,
+      "learning_rate": 5.577116935483872e-05,
+      "loss": 1.7599,
+      "step": 17550
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0216902494430542,
+      "learning_rate": 5.574596774193549e-05,
+      "loss": 1.7407,
+      "step": 17560
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2880454063415527,
+      "learning_rate": 5.5720766129032267e-05,
+      "loss": 1.9132,
+      "step": 17570
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1605669260025024,
+      "learning_rate": 5.569556451612904e-05,
+      "loss": 1.912,
+      "step": 17580
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.520478367805481,
+      "learning_rate": 5.5670362903225815e-05,
+      "loss": 1.953,
+      "step": 17590
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.770216941833496,
+      "learning_rate": 5.5645161290322576e-05,
+      "loss": 1.8359,
+      "step": 17600
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.587915062904358,
+      "learning_rate": 5.561995967741935e-05,
+      "loss": 1.917,
+      "step": 17610
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4603217840194702,
+      "learning_rate": 5.559475806451613e-05,
+      "loss": 1.8212,
+      "step": 17620
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6429760456085205,
+      "learning_rate": 5.5569556451612906e-05,
+      "loss": 1.984,
+      "step": 17630
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6397391557693481,
+      "learning_rate": 5.554435483870968e-05,
+      "loss": 1.8404,
+      "step": 17640
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8757877349853516,
+      "learning_rate": 5.5519153225806454e-05,
+      "loss": 1.7749,
+      "step": 17650
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.223899245262146,
+      "learning_rate": 5.549395161290323e-05,
+      "loss": 1.7249,
+      "step": 17660
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3431240320205688,
+      "learning_rate": 5.546875e-05,
+      "loss": 1.9459,
+      "step": 17670
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.048194408416748,
+      "learning_rate": 5.544354838709678e-05,
+      "loss": 1.8642,
+      "step": 17680
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.403993606567383,
+      "learning_rate": 5.541834677419355e-05,
+      "loss": 1.9106,
+      "step": 17690
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9815001487731934,
+      "learning_rate": 5.5393145161290326e-05,
+      "loss": 1.8345,
+      "step": 17700
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2842810153961182,
+      "learning_rate": 5.53679435483871e-05,
+      "loss": 1.8719,
+      "step": 17710
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3689652681350708,
+      "learning_rate": 5.5342741935483875e-05,
+      "loss": 1.7682,
+      "step": 17720
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.081344723701477,
+      "learning_rate": 5.531754032258065e-05,
+      "loss": 2.0215,
+      "step": 17730
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.865383505821228,
+      "learning_rate": 5.529233870967743e-05,
+      "loss": 2.0658,
+      "step": 17740
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4591317176818848,
+      "learning_rate": 5.526713709677419e-05,
+      "loss": 1.9184,
+      "step": 17750
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.2580301761627197,
+      "learning_rate": 5.5241935483870966e-05,
+      "loss": 1.8084,
+      "step": 17760
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3585402965545654,
+      "learning_rate": 5.521673387096774e-05,
+      "loss": 2.0553,
+      "step": 17770
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.129691243171692,
+      "learning_rate": 5.5191532258064514e-05,
+      "loss": 1.8801,
+      "step": 17780
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4069316387176514,
+      "learning_rate": 5.516633064516129e-05,
+      "loss": 1.9311,
+      "step": 17790
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5662918090820312,
+      "learning_rate": 5.514112903225806e-05,
+      "loss": 1.748,
+      "step": 17800
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.782707691192627,
+      "learning_rate": 5.511592741935484e-05,
+      "loss": 2.0217,
+      "step": 17810
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.18367862701416,
+      "learning_rate": 5.509072580645162e-05,
+      "loss": 1.9165,
+      "step": 17820
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2262167930603027,
+      "learning_rate": 5.506552419354839e-05,
+      "loss": 1.7873,
+      "step": 17830
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8448723554611206,
+      "learning_rate": 5.504032258064517e-05,
+      "loss": 1.8061,
+      "step": 17840
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6896344423294067,
+      "learning_rate": 5.501512096774194e-05,
+      "loss": 1.922,
+      "step": 17850
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4698189496994019,
+      "learning_rate": 5.4989919354838716e-05,
+      "loss": 1.8961,
+      "step": 17860
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5541282892227173,
+      "learning_rate": 5.496471774193549e-05,
+      "loss": 1.758,
+      "step": 17870
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.273051381111145,
+      "learning_rate": 5.4939516129032264e-05,
+      "loss": 1.8645,
+      "step": 17880
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.0488146543502808,
+      "learning_rate": 5.491431451612904e-05,
+      "loss": 1.868,
+      "step": 17890
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.057191014289856,
+      "learning_rate": 5.488911290322581e-05,
+      "loss": 1.7396,
+      "step": 17900
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.0166690349578857,
+      "learning_rate": 5.486391129032258e-05,
+      "loss": 1.7594,
+      "step": 17910
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.7844626903533936,
+      "learning_rate": 5.4838709677419355e-05,
+      "loss": 1.8623,
+      "step": 17920
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.157576084136963,
+      "learning_rate": 5.481350806451613e-05,
+      "loss": 1.8224,
+      "step": 17930
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9929746389389038,
+      "learning_rate": 5.4788306451612904e-05,
+      "loss": 1.93,
+      "step": 17940
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8016773462295532,
+      "learning_rate": 5.476310483870968e-05,
+      "loss": 1.8983,
+      "step": 17950
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.1163156032562256,
+      "learning_rate": 5.473790322580645e-05,
+      "loss": 1.6975,
+      "step": 17960
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1303445100784302,
+      "learning_rate": 5.471270161290323e-05,
+      "loss": 1.7908,
+      "step": 17970
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6643130779266357,
+      "learning_rate": 5.46875e-05,
+      "loss": 1.7863,
+      "step": 17980
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.329041600227356,
+      "learning_rate": 5.4662298387096775e-05,
+      "loss": 1.8281,
+      "step": 17990
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.770013451576233,
+      "learning_rate": 5.463709677419355e-05,
+      "loss": 1.9959,
+      "step": 18000
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4296618700027466,
+      "learning_rate": 5.4611895161290324e-05,
+      "loss": 1.9237,
+      "step": 18010
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.133723497390747,
+      "learning_rate": 5.4586693548387105e-05,
+      "loss": 2.1357,
+      "step": 18020
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.103461742401123,
+      "learning_rate": 5.456149193548388e-05,
+      "loss": 1.8335,
+      "step": 18030
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2370373010635376,
+      "learning_rate": 5.4536290322580654e-05,
+      "loss": 1.8338,
+      "step": 18040
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3647459745407104,
+      "learning_rate": 5.451108870967743e-05,
+      "loss": 1.831,
+      "step": 18050
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5328068733215332,
+      "learning_rate": 5.448588709677419e-05,
+      "loss": 1.7893,
+      "step": 18060
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5324825048446655,
+      "learning_rate": 5.4460685483870963e-05,
+      "loss": 1.947,
+      "step": 18070
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.0695600509643555,
+      "learning_rate": 5.443548387096774e-05,
+      "loss": 1.98,
+      "step": 18080
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9632848501205444,
+      "learning_rate": 5.441028225806451e-05,
+      "loss": 1.8572,
+      "step": 18090
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1290154457092285,
+      "learning_rate": 5.438508064516129e-05,
+      "loss": 1.7766,
+      "step": 18100
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3978049755096436,
+      "learning_rate": 5.435987903225807e-05,
+      "loss": 1.8699,
+      "step": 18110
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3822635412216187,
+      "learning_rate": 5.433467741935484e-05,
+      "loss": 1.8669,
+      "step": 18120
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3579354286193848,
+      "learning_rate": 5.4309475806451616e-05,
+      "loss": 1.561,
+      "step": 18130
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.2587788105010986,
+      "learning_rate": 5.428427419354839e-05,
+      "loss": 1.853,
+      "step": 18140
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.513980507850647,
+      "learning_rate": 5.4259072580645165e-05,
+      "loss": 1.8167,
+      "step": 18150
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.2227237224578857,
+      "learning_rate": 5.423387096774194e-05,
+      "loss": 2.1238,
+      "step": 18160
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4206455945968628,
+      "learning_rate": 5.4208669354838714e-05,
+      "loss": 1.7382,
+      "step": 18170
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.1453335285186768,
+      "learning_rate": 5.418346774193549e-05,
+      "loss": 1.7987,
+      "step": 18180
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6931467056274414,
+      "learning_rate": 5.415826612903226e-05,
+      "loss": 1.9128,
+      "step": 18190
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4835273027420044,
+      "learning_rate": 5.413306451612904e-05,
+      "loss": 1.8214,
+      "step": 18200
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7531219720840454,
+      "learning_rate": 5.410786290322581e-05,
+      "loss": 1.8155,
+      "step": 18210
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.1717031002044678,
+      "learning_rate": 5.408266129032258e-05,
+      "loss": 1.7643,
+      "step": 18220
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9261823892593384,
+      "learning_rate": 5.405745967741935e-05,
+      "loss": 1.9299,
+      "step": 18230
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7583166360855103,
+      "learning_rate": 5.403225806451613e-05,
+      "loss": 2.0311,
+      "step": 18240
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7322356700897217,
+      "learning_rate": 5.40070564516129e-05,
+      "loss": 1.9671,
+      "step": 18250
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.969856858253479,
+      "learning_rate": 5.3981854838709676e-05,
+      "loss": 1.9329,
+      "step": 18260
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3699142932891846,
+      "learning_rate": 5.395665322580645e-05,
+      "loss": 1.8472,
+      "step": 18270
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.268072485923767,
+      "learning_rate": 5.3931451612903225e-05,
+      "loss": 1.9328,
+      "step": 18280
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.1418943405151367,
+      "learning_rate": 5.3906250000000006e-05,
+      "loss": 1.8664,
+      "step": 18290
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3280118703842163,
+      "learning_rate": 5.388104838709678e-05,
+      "loss": 1.9908,
+      "step": 18300
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3657410144805908,
+      "learning_rate": 5.3855846774193555e-05,
+      "loss": 1.8468,
+      "step": 18310
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.1913851499557495,
+      "learning_rate": 5.383064516129033e-05,
+      "loss": 1.6061,
+      "step": 18320
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3878540992736816,
+      "learning_rate": 5.38054435483871e-05,
+      "loss": 1.795,
+      "step": 18330
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2987140417099,
+      "learning_rate": 5.378024193548388e-05,
+      "loss": 1.7585,
+      "step": 18340
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.324096918106079,
+      "learning_rate": 5.375504032258065e-05,
+      "loss": 1.9678,
+      "step": 18350
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2223808765411377,
+      "learning_rate": 5.3729838709677426e-05,
+      "loss": 1.7968,
+      "step": 18360
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4401593208312988,
+      "learning_rate": 5.3704637096774194e-05,
+      "loss": 1.8222,
+      "step": 18370
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1276520490646362,
+      "learning_rate": 5.367943548387097e-05,
+      "loss": 1.7743,
+      "step": 18380
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.133444309234619,
+      "learning_rate": 5.365423387096774e-05,
+      "loss": 1.9068,
+      "step": 18390
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.0602788925170898,
+      "learning_rate": 5.362903225806452e-05,
+      "loss": 1.7436,
+      "step": 18400
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.49454927444458,
+      "learning_rate": 5.360383064516129e-05,
+      "loss": 1.8816,
+      "step": 18410
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9956452250480652,
+      "learning_rate": 5.3578629032258066e-05,
+      "loss": 1.8518,
+      "step": 18420
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2502822875976562,
+      "learning_rate": 5.355342741935484e-05,
+      "loss": 1.5675,
+      "step": 18430
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9299494624137878,
+      "learning_rate": 5.3528225806451614e-05,
+      "loss": 1.9871,
+      "step": 18440
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6087390184402466,
+      "learning_rate": 5.350302419354839e-05,
+      "loss": 1.8948,
+      "step": 18450
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6991156339645386,
+      "learning_rate": 5.347782258064516e-05,
+      "loss": 1.7985,
+      "step": 18460
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9624131321907043,
+      "learning_rate": 5.345262096774194e-05,
+      "loss": 1.7797,
+      "step": 18470
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3196136951446533,
+      "learning_rate": 5.342741935483871e-05,
+      "loss": 1.7624,
+      "step": 18480
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.482833981513977,
+      "learning_rate": 5.340221774193549e-05,
+      "loss": 1.6777,
+      "step": 18490
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.25620174407959,
+      "learning_rate": 5.337701612903227e-05,
+      "loss": 1.7796,
+      "step": 18500
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1657735109329224,
+      "learning_rate": 5.335181451612904e-05,
+      "loss": 1.8489,
+      "step": 18510
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1427342891693115,
+      "learning_rate": 5.3326612903225816e-05,
+      "loss": 1.7334,
+      "step": 18520
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2763110399246216,
+      "learning_rate": 5.330141129032258e-05,
+      "loss": 1.8194,
+      "step": 18530
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.819624900817871,
+      "learning_rate": 5.327620967741935e-05,
+      "loss": 1.9471,
+      "step": 18540
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8640064001083374,
+      "learning_rate": 5.3251008064516125e-05,
+      "loss": 1.8322,
+      "step": 18550
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4828754663467407,
+      "learning_rate": 5.32258064516129e-05,
+      "loss": 1.7038,
+      "step": 18560
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.458784818649292,
+      "learning_rate": 5.320060483870968e-05,
+      "loss": 1.8501,
+      "step": 18570
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6425323486328125,
+      "learning_rate": 5.3175403225806455e-05,
+      "loss": 1.7597,
+      "step": 18580
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.102912425994873,
+      "learning_rate": 5.315020161290323e-05,
+      "loss": 1.9286,
+      "step": 18590
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2879949808120728,
+      "learning_rate": 5.3125000000000004e-05,
+      "loss": 1.9209,
+      "step": 18600
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2526116371154785,
+      "learning_rate": 5.309979838709678e-05,
+      "loss": 1.7622,
+      "step": 18610
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7939689755439758,
+      "learning_rate": 5.307459677419355e-05,
+      "loss": 1.7893,
+      "step": 18620
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3882848024368286,
+      "learning_rate": 5.304939516129033e-05,
+      "loss": 1.7712,
+      "step": 18630
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8248862028121948,
+      "learning_rate": 5.30241935483871e-05,
+      "loss": 1.8289,
+      "step": 18640
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8486753106117249,
+      "learning_rate": 5.2998991935483876e-05,
+      "loss": 1.7394,
+      "step": 18650
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.462324619293213,
+      "learning_rate": 5.297379032258065e-05,
+      "loss": 1.8318,
+      "step": 18660
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8602025508880615,
+      "learning_rate": 5.2948588709677424e-05,
+      "loss": 1.9197,
+      "step": 18670
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.989501953125,
+      "learning_rate": 5.292338709677419e-05,
+      "loss": 1.8832,
+      "step": 18680
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.458347797393799,
+      "learning_rate": 5.2898185483870966e-05,
+      "loss": 1.9924,
+      "step": 18690
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3084721565246582,
+      "learning_rate": 5.287298387096774e-05,
+      "loss": 1.9858,
+      "step": 18700
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6665098667144775,
+      "learning_rate": 5.2847782258064515e-05,
+      "loss": 1.7813,
+      "step": 18710
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7103375196456909,
+      "learning_rate": 5.282258064516129e-05,
+      "loss": 1.946,
+      "step": 18720
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.423071026802063,
+      "learning_rate": 5.2797379032258064e-05,
+      "loss": 1.6406,
+      "step": 18730
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6613318920135498,
+      "learning_rate": 5.277217741935484e-05,
+      "loss": 1.6687,
+      "step": 18740
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4874539375305176,
+      "learning_rate": 5.274697580645161e-05,
+      "loss": 1.8355,
+      "step": 18750
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6177746057510376,
+      "learning_rate": 5.272177419354839e-05,
+      "loss": 1.9195,
+      "step": 18760
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7322739362716675,
+      "learning_rate": 5.269657258064517e-05,
+      "loss": 1.6933,
+      "step": 18770
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2794102430343628,
+      "learning_rate": 5.267137096774194e-05,
+      "loss": 1.6754,
+      "step": 18780
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.525648832321167,
+      "learning_rate": 5.2646169354838716e-05,
+      "loss": 1.8116,
+      "step": 18790
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3353271484375,
+      "learning_rate": 5.262096774193549e-05,
+      "loss": 1.6464,
+      "step": 18800
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5984601974487305,
+      "learning_rate": 5.2595766129032265e-05,
+      "loss": 1.8554,
+      "step": 18810
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6255826950073242,
+      "learning_rate": 5.257056451612904e-05,
+      "loss": 1.8672,
+      "step": 18820
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2159287929534912,
+      "learning_rate": 5.2545362903225814e-05,
+      "loss": 1.6963,
+      "step": 18830
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.8018034100532532,
+      "learning_rate": 5.252016129032258e-05,
+      "loss": 1.784,
+      "step": 18840
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3325629234313965,
+      "learning_rate": 5.2494959677419356e-05,
+      "loss": 1.7163,
+      "step": 18850
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.993172287940979,
+      "learning_rate": 5.246975806451613e-05,
+      "loss": 2.0363,
+      "step": 18860
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1313930749893188,
+      "learning_rate": 5.2444556451612904e-05,
+      "loss": 1.6542,
+      "step": 18870
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2268941402435303,
+      "learning_rate": 5.241935483870968e-05,
+      "loss": 1.6912,
+      "step": 18880
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7621657252311707,
+      "learning_rate": 5.239415322580645e-05,
+      "loss": 1.8038,
+      "step": 18890
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3896924257278442,
+      "learning_rate": 5.236895161290323e-05,
+      "loss": 2.1321,
+      "step": 18900
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.962166965007782,
+      "learning_rate": 5.234375e-05,
+      "loss": 1.8044,
+      "step": 18910
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4755418300628662,
+      "learning_rate": 5.2318548387096776e-05,
+      "loss": 1.6811,
+      "step": 18920
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5500789880752563,
+      "learning_rate": 5.229334677419355e-05,
+      "loss": 1.6724,
+      "step": 18930
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.24942946434021,
+      "learning_rate": 5.2268145161290325e-05,
+      "loss": 1.9614,
+      "step": 18940
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.809902548789978,
+      "learning_rate": 5.22429435483871e-05,
+      "loss": 1.5988,
+      "step": 18950
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.0308127403259277,
+      "learning_rate": 5.221774193548388e-05,
+      "loss": 2.0024,
+      "step": 18960
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3597396612167358,
+      "learning_rate": 5.2192540322580655e-05,
+      "loss": 1.833,
+      "step": 18970
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9259357452392578,
+      "learning_rate": 5.216733870967743e-05,
+      "loss": 1.9051,
+      "step": 18980
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3146393299102783,
+      "learning_rate": 5.214213709677419e-05,
+      "loss": 1.6861,
+      "step": 18990
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.0335071086883545,
+      "learning_rate": 5.2116935483870964e-05,
+      "loss": 1.6602,
+      "step": 19000
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.443123698234558,
+      "learning_rate": 5.209173387096774e-05,
+      "loss": 1.9984,
+      "step": 19010
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6226738691329956,
+      "learning_rate": 5.206653225806451e-05,
+      "loss": 2.0702,
+      "step": 19020
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.308530807495117,
+      "learning_rate": 5.204133064516129e-05,
+      "loss": 1.9254,
+      "step": 19030
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3160682916641235,
+      "learning_rate": 5.201612903225807e-05,
+      "loss": 1.8873,
+      "step": 19040
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9209178686141968,
+      "learning_rate": 5.199092741935484e-05,
+      "loss": 1.6655,
+      "step": 19050
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3254390954971313,
+      "learning_rate": 5.196572580645162e-05,
+      "loss": 1.795,
+      "step": 19060
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3473490476608276,
+      "learning_rate": 5.194052419354839e-05,
+      "loss": 1.8162,
+      "step": 19070
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6390334367752075,
+      "learning_rate": 5.1915322580645166e-05,
+      "loss": 1.8294,
+      "step": 19080
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7239649295806885,
+      "learning_rate": 5.189012096774194e-05,
+      "loss": 1.7944,
+      "step": 19090
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5217039585113525,
+      "learning_rate": 5.1864919354838714e-05,
+      "loss": 1.816,
+      "step": 19100
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0940866470336914,
+      "learning_rate": 5.183971774193549e-05,
+      "loss": 1.8238,
+      "step": 19110
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9751317501068115,
+      "learning_rate": 5.181451612903226e-05,
+      "loss": 1.8495,
+      "step": 19120
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.8836818337440491,
+      "learning_rate": 5.178931451612904e-05,
+      "loss": 1.8068,
+      "step": 19130
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.516302466392517,
+      "learning_rate": 5.176411290322581e-05,
+      "loss": 1.7825,
+      "step": 19140
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4869805574417114,
+      "learning_rate": 5.173891129032258e-05,
+      "loss": 1.8385,
+      "step": 19150
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4858672618865967,
+      "learning_rate": 5.1713709677419354e-05,
+      "loss": 1.9179,
+      "step": 19160
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0536432266235352,
+      "learning_rate": 5.168850806451613e-05,
+      "loss": 1.786,
+      "step": 19170
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9299923181533813,
+      "learning_rate": 5.16633064516129e-05,
+      "loss": 1.9073,
+      "step": 19180
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9864319562911987,
+      "learning_rate": 5.163810483870968e-05,
+      "loss": 1.6013,
+      "step": 19190
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.245816469192505,
+      "learning_rate": 5.161290322580645e-05,
+      "loss": 1.8545,
+      "step": 19200
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6631780862808228,
+      "learning_rate": 5.1587701612903225e-05,
+      "loss": 1.9811,
+      "step": 19210
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5718727111816406,
+      "learning_rate": 5.15625e-05,
+      "loss": 1.7828,
+      "step": 19220
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1655147075653076,
+      "learning_rate": 5.1537298387096774e-05,
+      "loss": 1.7914,
+      "step": 19230
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8513822555541992,
+      "learning_rate": 5.1512096774193555e-05,
+      "loss": 1.9465,
+      "step": 19240
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5749272108078003,
+      "learning_rate": 5.148689516129033e-05,
+      "loss": 1.8115,
+      "step": 19250
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7210901975631714,
+      "learning_rate": 5.1461693548387104e-05,
+      "loss": 1.9288,
+      "step": 19260
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.532798409461975,
+      "learning_rate": 5.143649193548388e-05,
+      "loss": 1.8036,
+      "step": 19270
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4062786102294922,
+      "learning_rate": 5.141129032258065e-05,
+      "loss": 1.7625,
+      "step": 19280
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1032317876815796,
+      "learning_rate": 5.138608870967743e-05,
+      "loss": 1.8115,
+      "step": 19290
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1730623245239258,
+      "learning_rate": 5.136088709677419e-05,
+      "loss": 1.9383,
+      "step": 19300
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.132368803024292,
+      "learning_rate": 5.133568548387097e-05,
+      "loss": 1.8689,
+      "step": 19310
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8787751793861389,
+      "learning_rate": 5.131048387096774e-05,
+      "loss": 1.8728,
+      "step": 19320
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.37749183177948,
+      "learning_rate": 5.128528225806452e-05,
+      "loss": 1.8761,
+      "step": 19330
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0123610496520996,
+      "learning_rate": 5.126008064516129e-05,
+      "loss": 1.8296,
+      "step": 19340
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8673419952392578,
+      "learning_rate": 5.1234879032258066e-05,
+      "loss": 1.8222,
+      "step": 19350
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6806849241256714,
+      "learning_rate": 5.120967741935484e-05,
+      "loss": 1.83,
+      "step": 19360
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3686895370483398,
+      "learning_rate": 5.1184475806451615e-05,
+      "loss": 1.648,
+      "step": 19370
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0755271911621094,
+      "learning_rate": 5.115927419354839e-05,
+      "loss": 1.9313,
+      "step": 19380
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8655942678451538,
+      "learning_rate": 5.1134072580645164e-05,
+      "loss": 1.7216,
+      "step": 19390
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.170779824256897,
+      "learning_rate": 5.110887096774194e-05,
+      "loss": 1.872,
+      "step": 19400
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0157610177993774,
+      "learning_rate": 5.108366935483871e-05,
+      "loss": 1.8341,
+      "step": 19410
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9939554929733276,
+      "learning_rate": 5.105846774193549e-05,
+      "loss": 1.8721,
+      "step": 19420
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.701539397239685,
+      "learning_rate": 5.103326612903227e-05,
+      "loss": 1.7948,
+      "step": 19430
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.137074589729309,
+      "learning_rate": 5.100806451612904e-05,
+      "loss": 1.9275,
+      "step": 19440
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5196194648742676,
+      "learning_rate": 5.0982862903225816e-05,
+      "loss": 1.7816,
+      "step": 19450
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.617378830909729,
+      "learning_rate": 5.095766129032258e-05,
+      "loss": 1.845,
+      "step": 19460
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7246140241622925,
+      "learning_rate": 5.093245967741935e-05,
+      "loss": 1.927,
+      "step": 19470
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1106048822402954,
+      "learning_rate": 5.0907258064516126e-05,
+      "loss": 1.8966,
+      "step": 19480
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.4794956147670746,
+      "learning_rate": 5.08820564516129e-05,
+      "loss": 1.8426,
+      "step": 19490
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1508374214172363,
+      "learning_rate": 5.0856854838709675e-05,
+      "loss": 1.7698,
+      "step": 19500
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4683820009231567,
+      "learning_rate": 5.0831653225806456e-05,
+      "loss": 2.0415,
+      "step": 19510
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5710322856903076,
+      "learning_rate": 5.080645161290323e-05,
+      "loss": 1.7935,
+      "step": 19520
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9250189065933228,
+      "learning_rate": 5.0781250000000004e-05,
+      "loss": 1.9379,
+      "step": 19530
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4129551649093628,
+      "learning_rate": 5.075604838709678e-05,
+      "loss": 1.9801,
+      "step": 19540
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3108800649642944,
+      "learning_rate": 5.073084677419355e-05,
+      "loss": 1.8018,
+      "step": 19550
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9681835174560547,
+      "learning_rate": 5.070564516129033e-05,
+      "loss": 2.0189,
+      "step": 19560
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.774330735206604,
+      "learning_rate": 5.06804435483871e-05,
+      "loss": 1.706,
+      "step": 19570
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1112943887710571,
+      "learning_rate": 5.0655241935483876e-05,
+      "loss": 1.8852,
+      "step": 19580
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.0702364444732666,
+      "learning_rate": 5.063004032258065e-05,
+      "loss": 1.8965,
+      "step": 19590
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.388067603111267,
+      "learning_rate": 5.0604838709677425e-05,
+      "loss": 1.9892,
+      "step": 19600
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9357980489730835,
+      "learning_rate": 5.057963709677419e-05,
+      "loss": 1.7734,
+      "step": 19610
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.311848759651184,
+      "learning_rate": 5.055443548387097e-05,
+      "loss": 1.8539,
+      "step": 19620
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3943524360656738,
+      "learning_rate": 5.052923387096774e-05,
+      "loss": 1.721,
+      "step": 19630
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.8427196741104126,
+      "learning_rate": 5.0504032258064515e-05,
+      "loss": 1.7677,
+      "step": 19640
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6180109977722168,
+      "learning_rate": 5.047883064516129e-05,
+      "loss": 1.8426,
+      "step": 19650
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5102537870407104,
+      "learning_rate": 5.0453629032258064e-05,
+      "loss": 1.8107,
+      "step": 19660
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2933138608932495,
+      "learning_rate": 5.042842741935484e-05,
+      "loss": 1.7859,
+      "step": 19670
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.942417323589325,
+      "learning_rate": 5.040322580645161e-05,
+      "loss": 1.8346,
+      "step": 19680
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.332387089729309,
+      "learning_rate": 5.037802419354839e-05,
+      "loss": 1.691,
+      "step": 19690
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.519358515739441,
+      "learning_rate": 5.035282258064516e-05,
+      "loss": 1.7167,
+      "step": 19700
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.102912425994873,
+      "learning_rate": 5.032762096774194e-05,
+      "loss": 1.978,
+      "step": 19710
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5782140493392944,
+      "learning_rate": 5.030241935483872e-05,
+      "loss": 1.8994,
+      "step": 19720
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5720537900924683,
+      "learning_rate": 5.027721774193549e-05,
+      "loss": 2.0671,
+      "step": 19730
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.1773195266723633,
+      "learning_rate": 5.0252016129032266e-05,
+      "loss": 1.8026,
+      "step": 19740
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5043028593063354,
+      "learning_rate": 5.022681451612904e-05,
+      "loss": 1.8215,
+      "step": 19750
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5516098737716675,
+      "learning_rate": 5.0201612903225814e-05,
+      "loss": 1.973,
+      "step": 19760
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4751557111740112,
+      "learning_rate": 5.0176411290322575e-05,
+      "loss": 1.7985,
+      "step": 19770
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7080754041671753,
+      "learning_rate": 5.0151209677419356e-05,
+      "loss": 1.5807,
+      "step": 19780
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.19722318649292,
+      "learning_rate": 5.012600806451613e-05,
+      "loss": 2.0648,
+      "step": 19790
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.9836114048957825,
+      "learning_rate": 5.0100806451612905e-05,
+      "loss": 1.8196,
+      "step": 19800
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.804094910621643,
+      "learning_rate": 5.007560483870968e-05,
+      "loss": 1.7251,
+      "step": 19810
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3634778261184692,
+      "learning_rate": 5.0050403225806454e-05,
+      "loss": 1.9007,
+      "step": 19820
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.8419346213340759,
+      "learning_rate": 5.002520161290323e-05,
+      "loss": 1.8319,
+      "step": 19830
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3470367193222046,
+      "learning_rate": 5e-05,
+      "loss": 1.6577,
+      "step": 19840
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.725541830062866,
+      "learning_rate": 4.997479838709678e-05,
+      "loss": 1.7717,
+      "step": 19850
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.023478627204895,
+      "learning_rate": 4.994959677419355e-05,
+      "loss": 1.5686,
+      "step": 19860
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4028514623641968,
+      "learning_rate": 4.9924395161290325e-05,
+      "loss": 1.8774,
+      "step": 19870
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.0776605606079102,
+      "learning_rate": 4.98991935483871e-05,
+      "loss": 1.6802,
+      "step": 19880
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.2852592468261719,
+      "learning_rate": 4.9873991935483874e-05,
+      "loss": 1.8266,
+      "step": 19890
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7209049463272095,
+      "learning_rate": 4.984879032258065e-05,
+      "loss": 1.9337,
+      "step": 19900
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7746049165725708,
+      "learning_rate": 4.982358870967742e-05,
+      "loss": 1.9419,
+      "step": 19910
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.1448496580123901,
+      "learning_rate": 4.97983870967742e-05,
+      "loss": 1.7986,
+      "step": 19920
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.1173341274261475,
+      "learning_rate": 4.977318548387097e-05,
+      "loss": 1.8688,
+      "step": 19930
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.7739990949630737,
+      "learning_rate": 4.9747983870967746e-05,
+      "loss": 1.8403,
+      "step": 19940
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.9469005465507507,
+      "learning_rate": 4.972278225806452e-05,
+      "loss": 1.6883,
+      "step": 19950
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.2558064460754395,
+      "learning_rate": 4.969758064516129e-05,
+      "loss": 1.8532,
+      "step": 19960
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.8459283113479614,
+      "learning_rate": 4.967237903225806e-05,
+      "loss": 1.7941,
+      "step": 19970
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.8783189058303833,
+      "learning_rate": 4.964717741935484e-05,
+      "loss": 1.6409,
+      "step": 19980
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.5740159749984741,
+      "learning_rate": 4.962197580645162e-05,
+      "loss": 1.973,
+      "step": 19990
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.1903932094573975,
+      "learning_rate": 4.959677419354839e-05,
+      "loss": 1.8508,
+      "step": 20000
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.2174843549728394,
+      "learning_rate": 4.9571572580645166e-05,
+      "loss": 1.7811,
+      "step": 20010
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.9128254055976868,
+      "learning_rate": 4.954637096774194e-05,
+      "loss": 1.708,
+      "step": 20020
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.1337945461273193,
+      "learning_rate": 4.9521169354838715e-05,
+      "loss": 1.7889,
+      "step": 20030
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.667252779006958,
+      "learning_rate": 4.949596774193548e-05,
+      "loss": 1.8642,
+      "step": 20040
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.4509947299957275,
+      "learning_rate": 4.947076612903226e-05,
+      "loss": 1.9191,
+      "step": 20050
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.3955143690109253,
+      "learning_rate": 4.944556451612903e-05,
+      "loss": 1.7304,
+      "step": 20060
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.0530259609222412,
+      "learning_rate": 4.9420362903225806e-05,
+      "loss": 1.9202,
+      "step": 20070
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.0755207538604736,
+      "learning_rate": 4.939516129032259e-05,
+      "loss": 1.7418,
+      "step": 20080
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.4421571493148804,
+      "learning_rate": 4.936995967741936e-05,
+      "loss": 1.7708,
+      "step": 20090
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.7133270502090454,
+      "learning_rate": 4.9344758064516135e-05,
+      "loss": 1.7164,
+      "step": 20100
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.0778799057006836,
+      "learning_rate": 4.931955645161291e-05,
+      "loss": 1.7994,
+      "step": 20110
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.0143673419952393,
+      "learning_rate": 4.929435483870968e-05,
+      "loss": 1.8632,
+      "step": 20120
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.419112205505371,
+      "learning_rate": 4.926915322580645e-05,
+      "loss": 1.5366,
+      "step": 20130
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.977290153503418,
+      "learning_rate": 4.9243951612903226e-05,
+      "loss": 1.8607,
+      "step": 20140
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.8152599334716797,
+      "learning_rate": 4.921875e-05,
+      "loss": 1.709,
+      "step": 20150
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.8043677806854248,
+      "learning_rate": 4.9193548387096775e-05,
+      "loss": 1.718,
+      "step": 20160
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.7959333062171936,
+      "learning_rate": 4.916834677419355e-05,
+      "loss": 1.8921,
+      "step": 20170
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.3443049192428589,
+      "learning_rate": 4.914314516129033e-05,
+      "loss": 1.8663,
+      "step": 20180
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.6839659810066223,
+      "learning_rate": 4.91179435483871e-05,
+      "loss": 1.9141,
+      "step": 20190
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.2184699773788452,
+      "learning_rate": 4.909274193548387e-05,
+      "loss": 1.8425,
+      "step": 20200
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.7919439077377319,
+      "learning_rate": 4.9067540322580646e-05,
+      "loss": 1.7333,
+      "step": 20210
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.9974613189697266,
+      "learning_rate": 4.904233870967742e-05,
+      "loss": 1.898,
+      "step": 20220
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.1530650854110718,
+      "learning_rate": 4.9017137096774195e-05,
+      "loss": 1.8771,
+      "step": 20230
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.0380550622940063,
+      "learning_rate": 4.899193548387097e-05,
+      "loss": 1.6882,
+      "step": 20240
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.7635253071784973,
+      "learning_rate": 4.8966733870967744e-05,
+      "loss": 1.9189,
+      "step": 20250
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.0348327159881592,
+      "learning_rate": 4.894153225806452e-05,
+      "loss": 1.7888,
+      "step": 20260
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.907285451889038,
+      "learning_rate": 4.891633064516129e-05,
+      "loss": 1.8723,
+      "step": 20270
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.478313684463501,
+      "learning_rate": 4.889112903225807e-05,
+      "loss": 1.8904,
+      "step": 20280
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.230298399925232,
+      "learning_rate": 4.886592741935484e-05,
+      "loss": 1.8276,
+      "step": 20290
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.9019665718078613,
+      "learning_rate": 4.8840725806451616e-05,
+      "loss": 1.7706,
+      "step": 20300
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.1319636106491089,
+      "learning_rate": 4.881552419354839e-05,
+      "loss": 1.8501,
+      "step": 20310
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.4039336442947388,
+      "learning_rate": 4.8790322580645164e-05,
+      "loss": 2.0463,
+      "step": 20320
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.7572011947631836,
+      "learning_rate": 4.876512096774194e-05,
+      "loss": 1.7735,
+      "step": 20330
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.9889572858810425,
+      "learning_rate": 4.873991935483871e-05,
+      "loss": 1.9159,
+      "step": 20340
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.3456162214279175,
+      "learning_rate": 4.871471774193548e-05,
+      "loss": 1.776,
+      "step": 20350
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.9841784238815308,
+      "learning_rate": 4.868951612903226e-05,
+      "loss": 1.683,
+      "step": 20360
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.6519460678100586,
+      "learning_rate": 4.8664314516129036e-05,
+      "loss": 1.6882,
+      "step": 20370
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.7859209775924683,
+      "learning_rate": 4.863911290322581e-05,
+      "loss": 1.8952,
+      "step": 20380
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.345761775970459,
+      "learning_rate": 4.8613911290322585e-05,
+      "loss": 1.6275,
+      "step": 20390
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.2841273546218872,
+      "learning_rate": 4.858870967741936e-05,
+      "loss": 1.6482,
+      "step": 20400
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.9710187911987305,
+      "learning_rate": 4.856350806451613e-05,
+      "loss": 1.934,
+      "step": 20410
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.263417363166809,
+      "learning_rate": 4.853830645161291e-05,
+      "loss": 1.7911,
+      "step": 20420
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.367952585220337,
+      "learning_rate": 4.8513104838709675e-05,
+      "loss": 1.4893,
+      "step": 20430
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.5396358966827393,
+      "learning_rate": 4.848790322580645e-05,
+      "loss": 1.7513,
+      "step": 20440
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.428483009338379,
+      "learning_rate": 4.846270161290323e-05,
+      "loss": 1.8297,
+      "step": 20450
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.171694278717041,
+      "learning_rate": 4.8437500000000005e-05,
+      "loss": 1.7845,
+      "step": 20460
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.933369517326355,
+      "learning_rate": 4.841229838709678e-05,
+      "loss": 2.029,
+      "step": 20470
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.061107635498047,
+      "learning_rate": 4.8387096774193554e-05,
+      "loss": 1.8851,
+      "step": 20480
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.6314517259597778,
+      "learning_rate": 4.836189516129033e-05,
+      "loss": 1.7691,
+      "step": 20490
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.7762845158576965,
+      "learning_rate": 4.8336693548387096e-05,
+      "loss": 1.5701,
+      "step": 20500
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.4232535362243652,
+      "learning_rate": 4.831149193548387e-05,
+      "loss": 1.9238,
+      "step": 20510
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.862640619277954,
+      "learning_rate": 4.8286290322580644e-05,
+      "loss": 1.8826,
+      "step": 20520
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.5583014488220215,
+      "learning_rate": 4.826108870967742e-05,
+      "loss": 1.8834,
+      "step": 20530
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.4686211347579956,
+      "learning_rate": 4.823588709677419e-05,
+      "loss": 1.8422,
+      "step": 20540
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.8517003059387207,
+      "learning_rate": 4.8210685483870974e-05,
+      "loss": 1.8061,
+      "step": 20550
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.2043954133987427,
+      "learning_rate": 4.818548387096775e-05,
+      "loss": 1.7568,
+      "step": 20560
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.9422028660774231,
+      "learning_rate": 4.816028225806452e-05,
+      "loss": 1.809,
+      "step": 20570
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.637969970703125,
+      "learning_rate": 4.813508064516129e-05,
+      "loss": 1.6063,
+      "step": 20580
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.0520216226577759,
+      "learning_rate": 4.8109879032258065e-05,
+      "loss": 1.7206,
+      "step": 20590
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.4685229063034058,
+      "learning_rate": 4.808467741935484e-05,
+      "loss": 1.821,
+      "step": 20600
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.7676855325698853,
+      "learning_rate": 4.8059475806451613e-05,
+      "loss": 1.8755,
+      "step": 20610
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.1410266160964966,
+      "learning_rate": 4.803427419354839e-05,
+      "loss": 1.7418,
+      "step": 20620
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.0694966316223145,
+      "learning_rate": 4.800907258064516e-05,
+      "loss": 1.5678,
+      "step": 20630
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.301387071609497,
+      "learning_rate": 4.7983870967741937e-05,
+      "loss": 1.8454,
+      "step": 20640
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.0509235858917236,
+      "learning_rate": 4.795866935483872e-05,
+      "loss": 2.0156,
+      "step": 20650
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.190200924873352,
+      "learning_rate": 4.7933467741935485e-05,
+      "loss": 1.8406,
+      "step": 20660
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.099355697631836,
+      "learning_rate": 4.790826612903226e-05,
+      "loss": 2.0379,
+      "step": 20670
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.412194848060608,
+      "learning_rate": 4.7883064516129034e-05,
+      "loss": 1.7404,
+      "step": 20680
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 3.3393783569335938,
+      "learning_rate": 4.785786290322581e-05,
+      "loss": 1.7811,
+      "step": 20690
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.4856171607971191,
+      "learning_rate": 4.783266129032258e-05,
+      "loss": 1.8346,
+      "step": 20700
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.2283551692962646,
+      "learning_rate": 4.780745967741936e-05,
+      "loss": 1.7327,
+      "step": 20710
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.2089500427246094,
+      "learning_rate": 4.778225806451613e-05,
+      "loss": 1.7951,
+      "step": 20720
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.1747044324874878,
+      "learning_rate": 4.7757056451612906e-05,
+      "loss": 1.8379,
+      "step": 20730
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.024285078048706,
+      "learning_rate": 4.773185483870968e-05,
+      "loss": 1.8058,
+      "step": 20740
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.6714431047439575,
+      "learning_rate": 4.7706653225806454e-05,
+      "loss": 1.9061,
+      "step": 20750
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.465782642364502,
+      "learning_rate": 4.768145161290323e-05,
+      "loss": 1.7831,
+      "step": 20760
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.4173626899719238,
+      "learning_rate": 4.765625e-05,
+      "loss": 1.8091,
+      "step": 20770
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.1885173320770264,
+      "learning_rate": 4.763104838709678e-05,
+      "loss": 1.8758,
+      "step": 20780
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.7757688760757446,
+      "learning_rate": 4.760584677419355e-05,
+      "loss": 1.6916,
+      "step": 20790
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.3510843515396118,
+      "learning_rate": 4.7580645161290326e-05,
+      "loss": 1.9186,
+      "step": 20800
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.014058828353882,
+      "learning_rate": 4.7555443548387094e-05,
+      "loss": 1.9781,
+      "step": 20810
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.6426938772201538,
+      "learning_rate": 4.753024193548387e-05,
+      "loss": 1.906,
+      "step": 20820
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.68331778049469,
+      "learning_rate": 4.750504032258065e-05,
+      "loss": 1.7783,
+      "step": 20830
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.4496350288391113,
+      "learning_rate": 4.7479838709677423e-05,
+      "loss": 1.8296,
+      "step": 20840
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.602859377861023,
+      "learning_rate": 4.74546370967742e-05,
+      "loss": 1.7416,
+      "step": 20850
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.0266034603118896,
+      "learning_rate": 4.742943548387097e-05,
+      "loss": 2.0441,
+      "step": 20860
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.0481747388839722,
+      "learning_rate": 4.7404233870967746e-05,
+      "loss": 1.6258,
+      "step": 20870
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.7661733031272888,
+      "learning_rate": 4.737903225806452e-05,
+      "loss": 1.7524,
+      "step": 20880
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.6872485876083374,
+      "learning_rate": 4.735383064516129e-05,
+      "loss": 1.9215,
+      "step": 20890
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.00102698802948,
+      "learning_rate": 4.732862903225806e-05,
+      "loss": 1.9639,
+      "step": 20900
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.4659647941589355,
+      "learning_rate": 4.730342741935484e-05,
+      "loss": 1.5928,
+      "step": 20910
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.1227573156356812,
+      "learning_rate": 4.727822580645161e-05,
+      "loss": 1.8468,
+      "step": 20920
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.622872233390808,
+      "learning_rate": 4.725302419354839e-05,
+      "loss": 1.8619,
+      "step": 20930
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.4574980735778809,
+      "learning_rate": 4.722782258064517e-05,
+      "loss": 1.7405,
+      "step": 20940
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.864703893661499,
+      "learning_rate": 4.720262096774194e-05,
+      "loss": 1.8312,
+      "step": 20950
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.0741955041885376,
+      "learning_rate": 4.7177419354838716e-05,
+      "loss": 1.8663,
+      "step": 20960
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.170819640159607,
+      "learning_rate": 4.715221774193548e-05,
+      "loss": 1.6999,
+      "step": 20970
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.2674113512039185,
+      "learning_rate": 4.712701612903226e-05,
+      "loss": 1.7992,
+      "step": 20980
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.580033302307129,
+      "learning_rate": 4.710181451612903e-05,
+      "loss": 1.6812,
+      "step": 20990
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.2087595462799072,
+      "learning_rate": 4.7076612903225806e-05,
+      "loss": 1.8645,
+      "step": 21000
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.126739263534546,
+      "learning_rate": 4.705141129032258e-05,
+      "loss": 1.7014,
+      "step": 21010
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.275302529335022,
+      "learning_rate": 4.702620967741936e-05,
+      "loss": 2.067,
+      "step": 21020
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.8976848125457764,
+      "learning_rate": 4.7001008064516136e-05,
+      "loss": 1.8362,
+      "step": 21030
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.724313974380493,
+      "learning_rate": 4.697580645161291e-05,
+      "loss": 1.7909,
+      "step": 21040
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.8377379179000854,
+      "learning_rate": 4.695060483870968e-05,
+      "loss": 1.8332,
+      "step": 21050
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.5164469480514526,
+      "learning_rate": 4.692540322580645e-05,
+      "loss": 1.8299,
+      "step": 21060
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.6679975986480713,
+      "learning_rate": 4.690020161290323e-05,
+      "loss": 1.8407,
+      "step": 21070
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.5213364362716675,
+      "learning_rate": 4.6875e-05,
+      "loss": 1.7939,
+      "step": 21080
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.8200163841247559,
+      "learning_rate": 4.6849798387096775e-05,
+      "loss": 1.8107,
+      "step": 21090
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.9726907014846802,
+      "learning_rate": 4.682459677419355e-05,
+      "loss": 1.9984,
+      "step": 21100
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.9200849533081055,
+      "learning_rate": 4.6799395161290324e-05,
+      "loss": 1.8332,
+      "step": 21110
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.392740488052368,
+      "learning_rate": 4.67741935483871e-05,
+      "loss": 1.7785,
+      "step": 21120
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.6736910343170166,
+      "learning_rate": 4.674899193548387e-05,
+      "loss": 1.7601,
+      "step": 21130
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.3530454635620117,
+      "learning_rate": 4.672379032258065e-05,
+      "loss": 2.006,
+      "step": 21140
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.4480328559875488,
+      "learning_rate": 4.669858870967742e-05,
+      "loss": 1.8195,
+      "step": 21150
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.6921299695968628,
+      "learning_rate": 4.6673387096774196e-05,
+      "loss": 2.0207,
+      "step": 21160
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.9205206632614136,
+      "learning_rate": 4.664818548387097e-05,
+      "loss": 1.9677,
+      "step": 21170
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.8897685408592224,
+      "learning_rate": 4.6622983870967744e-05,
+      "loss": 1.8511,
+      "step": 21180
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.7205467224121094,
+      "learning_rate": 4.659778225806452e-05,
+      "loss": 1.8057,
+      "step": 21190
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.5909370183944702,
+      "learning_rate": 4.657258064516129e-05,
+      "loss": 1.9596,
+      "step": 21200
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.9639760255813599,
+      "learning_rate": 4.654737903225807e-05,
+      "loss": 1.7331,
+      "step": 21210
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.5253931283950806,
+      "learning_rate": 4.652217741935484e-05,
+      "loss": 1.6685,
+      "step": 21220
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.0157084465026855,
+      "learning_rate": 4.6496975806451616e-05,
+      "loss": 1.7226,
+      "step": 21230
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.9519565105438232,
+      "learning_rate": 4.647177419354839e-05,
+      "loss": 1.8887,
+      "step": 21240
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.134766697883606,
+      "learning_rate": 4.6446572580645165e-05,
+      "loss": 1.7253,
+      "step": 21250
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.4847817420959473,
+      "learning_rate": 4.642137096774194e-05,
+      "loss": 2.0112,
+      "step": 21260
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.5583120584487915,
+      "learning_rate": 4.6396169354838714e-05,
+      "loss": 1.7527,
+      "step": 21270
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.3329979181289673,
+      "learning_rate": 4.637096774193548e-05,
+      "loss": 1.6922,
+      "step": 21280
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.3046951293945312,
+      "learning_rate": 4.6345766129032255e-05,
+      "loss": 1.8032,
+      "step": 21290
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.082306146621704,
+      "learning_rate": 4.6320564516129037e-05,
+      "loss": 1.8739,
+      "step": 21300
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.5408998727798462,
+      "learning_rate": 4.629536290322581e-05,
+      "loss": 1.8378,
+      "step": 21310
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.2714152336120605,
+      "learning_rate": 4.6270161290322585e-05,
+      "loss": 1.8914,
+      "step": 21320
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.383149266242981,
+      "learning_rate": 4.624495967741936e-05,
+      "loss": 1.6475,
+      "step": 21330
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.7632426023483276,
+      "learning_rate": 4.6219758064516134e-05,
+      "loss": 2.0556,
+      "step": 21340
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.418745994567871,
+      "learning_rate": 4.619455645161291e-05,
+      "loss": 1.6534,
+      "step": 21350
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.2804380655288696,
+      "learning_rate": 4.6169354838709676e-05,
+      "loss": 1.7704,
+      "step": 21360
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.7900537252426147,
+      "learning_rate": 4.614415322580645e-05,
+      "loss": 1.7051,
+      "step": 21370
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.4457615613937378,
+      "learning_rate": 4.6118951612903225e-05,
+      "loss": 1.8063,
+      "step": 21380
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.211419701576233,
+      "learning_rate": 4.609375e-05,
+      "loss": 1.9384,
+      "step": 21390
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.5796228647232056,
+      "learning_rate": 4.606854838709678e-05,
+      "loss": 1.6798,
+      "step": 21400
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.1037603616714478,
+      "learning_rate": 4.6043346774193554e-05,
+      "loss": 1.7567,
+      "step": 21410
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.7267943620681763,
+      "learning_rate": 4.601814516129033e-05,
+      "loss": 1.9013,
+      "step": 21420
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.6035006642341614,
+      "learning_rate": 4.5992943548387096e-05,
+      "loss": 1.7493,
+      "step": 21430
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.9193801879882812,
+      "learning_rate": 4.596774193548387e-05,
+      "loss": 1.8118,
+      "step": 21440
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.1260793209075928,
+      "learning_rate": 4.5942540322580645e-05,
+      "loss": 1.7421,
+      "step": 21450
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.8077141046524048,
+      "learning_rate": 4.591733870967742e-05,
+      "loss": 1.7419,
+      "step": 21460
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.2104078531265259,
+      "learning_rate": 4.5892137096774194e-05,
+      "loss": 1.7541,
+      "step": 21470
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.97088623046875,
+      "learning_rate": 4.586693548387097e-05,
+      "loss": 1.8802,
+      "step": 21480
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.043222427368164,
+      "learning_rate": 4.584173387096774e-05,
+      "loss": 1.8359,
+      "step": 21490
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.1237989664077759,
+      "learning_rate": 4.5816532258064524e-05,
+      "loss": 1.6384,
+      "step": 21500
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.7070789337158203,
+      "learning_rate": 4.579133064516129e-05,
+      "loss": 1.6892,
+      "step": 21510
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.305580496788025,
+      "learning_rate": 4.5766129032258065e-05,
+      "loss": 1.917,
+      "step": 21520
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.9539462327957153,
+      "learning_rate": 4.574092741935484e-05,
+      "loss": 1.7458,
+      "step": 21530
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.4502270221710205,
+      "learning_rate": 4.5715725806451614e-05,
+      "loss": 1.7331,
+      "step": 21540
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.4336037635803223,
+      "learning_rate": 4.569052419354839e-05,
+      "loss": 1.6919,
+      "step": 21550
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.1060575246810913,
+      "learning_rate": 4.566532258064516e-05,
+      "loss": 1.4315,
+      "step": 21560
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.1218525171279907,
+      "learning_rate": 4.564012096774194e-05,
+      "loss": 1.884,
+      "step": 21570
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.8032273054122925,
+      "learning_rate": 4.561491935483871e-05,
+      "loss": 1.6739,
+      "step": 21580
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.903509259223938,
+      "learning_rate": 4.5589717741935486e-05,
+      "loss": 1.7718,
+      "step": 21590
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.7964780926704407,
+      "learning_rate": 4.556451612903226e-05,
+      "loss": 1.9384,
+      "step": 21600
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.8347128033638,
+      "learning_rate": 4.5539314516129035e-05,
+      "loss": 1.6281,
+      "step": 21610
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.8796850442886353,
+      "learning_rate": 4.551411290322581e-05,
+      "loss": 2.0124,
+      "step": 21620
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.7131448984146118,
+      "learning_rate": 4.548891129032258e-05,
+      "loss": 1.8427,
+      "step": 21630
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.6487430334091187,
+      "learning_rate": 4.546370967741936e-05,
+      "loss": 1.6811,
+      "step": 21640
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.299458384513855,
+      "learning_rate": 4.543850806451613e-05,
+      "loss": 1.7675,
+      "step": 21650
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.8929489850997925,
+      "learning_rate": 4.5413306451612906e-05,
+      "loss": 1.9274,
+      "step": 21660
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.508484959602356,
+      "learning_rate": 4.538810483870968e-05,
+      "loss": 1.7706,
+      "step": 21670
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.9787169694900513,
+      "learning_rate": 4.5362903225806455e-05,
+      "loss": 1.7806,
+      "step": 21680
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.498473882675171,
+      "learning_rate": 4.533770161290323e-05,
+      "loss": 2.0494,
+      "step": 21690
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.070332646369934,
+      "learning_rate": 4.5312500000000004e-05,
+      "loss": 1.9381,
+      "step": 21700
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.6289297938346863,
+      "learning_rate": 4.528729838709678e-05,
+      "loss": 1.8162,
+      "step": 21710
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.6247340440750122,
+      "learning_rate": 4.526209677419355e-05,
+      "loss": 2.0876,
+      "step": 21720
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.4340269565582275,
+      "learning_rate": 4.523689516129033e-05,
+      "loss": 1.9281,
+      "step": 21730
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.560603380203247,
+      "learning_rate": 4.5211693548387094e-05,
+      "loss": 1.7228,
+      "step": 21740
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.9338274002075195,
+      "learning_rate": 4.518649193548387e-05,
+      "loss": 1.819,
+      "step": 21750
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.7740694880485535,
+      "learning_rate": 4.516129032258064e-05,
+      "loss": 1.8086,
+      "step": 21760
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.3313980102539062,
+      "learning_rate": 4.5136088709677424e-05,
+      "loss": 1.6869,
+      "step": 21770
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.6811493635177612,
+      "learning_rate": 4.51108870967742e-05,
+      "loss": 1.9623,
+      "step": 21780
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.219314694404602,
+      "learning_rate": 4.508568548387097e-05,
+      "loss": 1.8245,
+      "step": 21790
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.3698571920394897,
+      "learning_rate": 4.506048387096775e-05,
+      "loss": 1.7172,
+      "step": 21800
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.6892366409301758,
+      "learning_rate": 4.503528225806452e-05,
+      "loss": 1.8523,
+      "step": 21810
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.6728068590164185,
+      "learning_rate": 4.501008064516129e-05,
+      "loss": 1.9605,
+      "step": 21820
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.9532008767127991,
+      "learning_rate": 4.498487903225806e-05,
+      "loss": 1.8404,
+      "step": 21830
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.392332673072815,
+      "learning_rate": 4.495967741935484e-05,
+      "loss": 1.721,
+      "step": 21840
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.0014393329620361,
+      "learning_rate": 4.493447580645161e-05,
+      "loss": 1.7508,
+      "step": 21850
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.415466070175171,
+      "learning_rate": 4.4909274193548386e-05,
+      "loss": 1.8335,
+      "step": 21860
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.9705161452293396,
+      "learning_rate": 4.488407258064517e-05,
+      "loss": 1.9015,
+      "step": 21870
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.6603256464004517,
+      "learning_rate": 4.485887096774194e-05,
+      "loss": 1.8758,
+      "step": 21880
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 3.424184799194336,
+      "learning_rate": 4.4833669354838716e-05,
+      "loss": 1.845,
+      "step": 21890
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.7502392530441284,
+      "learning_rate": 4.4808467741935484e-05,
+      "loss": 1.8327,
+      "step": 21900
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.5341150760650635,
+      "learning_rate": 4.478326612903226e-05,
+      "loss": 1.9614,
+      "step": 21910
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.0670925378799438,
+      "learning_rate": 4.475806451612903e-05,
+      "loss": 1.8042,
+      "step": 21920
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.3000555038452148,
+      "learning_rate": 4.473286290322581e-05,
+      "loss": 1.672,
+      "step": 21930
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.6566033363342285,
+      "learning_rate": 4.470766129032258e-05,
+      "loss": 2.0226,
+      "step": 21940
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.1589691638946533,
+      "learning_rate": 4.4682459677419356e-05,
+      "loss": 1.9072,
+      "step": 21950
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.366020679473877,
+      "learning_rate": 4.465725806451613e-05,
+      "loss": 1.8896,
+      "step": 21960
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.5499820709228516,
+      "learning_rate": 4.463205645161291e-05,
+      "loss": 1.9052,
+      "step": 21970
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.1683827638626099,
+      "learning_rate": 4.460685483870968e-05,
+      "loss": 1.6906,
+      "step": 21980
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.8128511905670166,
+      "learning_rate": 4.458165322580645e-05,
+      "loss": 1.6995,
+      "step": 21990
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.9281318187713623,
+      "learning_rate": 4.455645161290323e-05,
+      "loss": 1.723,
+      "step": 22000
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.327014207839966,
+      "learning_rate": 4.453125e-05,
+      "loss": 1.7253,
+      "step": 22010
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.2528247833251953,
+      "learning_rate": 4.4506048387096776e-05,
+      "loss": 1.7944,
+      "step": 22020
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.41196346282959,
+      "learning_rate": 4.448084677419355e-05,
+      "loss": 1.9758,
+      "step": 22030
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.0107593536376953,
+      "learning_rate": 4.4455645161290325e-05,
+      "loss": 1.7774,
+      "step": 22040
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.09134840965271,
+      "learning_rate": 4.44304435483871e-05,
+      "loss": 1.7649,
+      "step": 22050
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.9467769861221313,
+      "learning_rate": 4.440524193548387e-05,
+      "loss": 1.8118,
+      "step": 22060
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.578166663646698,
+      "learning_rate": 4.438004032258065e-05,
+      "loss": 1.7954,
+      "step": 22070
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.5660881996154785,
+      "learning_rate": 4.435483870967742e-05,
+      "loss": 1.7923,
+      "step": 22080
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.9080318212509155,
+      "learning_rate": 4.4329637096774196e-05,
+      "loss": 1.9655,
+      "step": 22090
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.0548131465911865,
+      "learning_rate": 4.430443548387097e-05,
+      "loss": 1.668,
+      "step": 22100
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.1673885583877563,
+      "learning_rate": 4.4279233870967745e-05,
+      "loss": 1.7101,
+      "step": 22110
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.140804648399353,
+      "learning_rate": 4.425403225806452e-05,
+      "loss": 1.8758,
+      "step": 22120
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.951873779296875,
+      "learning_rate": 4.422883064516129e-05,
+      "loss": 1.6802,
+      "step": 22130
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.142872929573059,
+      "learning_rate": 4.420362903225807e-05,
+      "loss": 1.8473,
+      "step": 22140
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.6770776510238647,
+      "learning_rate": 4.417842741935484e-05,
+      "loss": 1.7295,
+      "step": 22150
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.0318210124969482,
+      "learning_rate": 4.415322580645162e-05,
+      "loss": 1.8558,
+      "step": 22160
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.5260729789733887,
+      "learning_rate": 4.412802419354839e-05,
+      "loss": 1.882,
+      "step": 22170
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.5855631828308105,
+      "learning_rate": 4.4102822580645165e-05,
+      "loss": 1.7224,
+      "step": 22180
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.590476155281067,
+      "learning_rate": 4.407762096774194e-05,
+      "loss": 1.5406,
+      "step": 22190
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.8204014301300049,
+      "learning_rate": 4.4052419354838714e-05,
+      "loss": 1.7882,
+      "step": 22200
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.4912632405757904,
+      "learning_rate": 4.402721774193548e-05,
+      "loss": 1.6778,
+      "step": 22210
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.5133213996887207,
+      "learning_rate": 4.4002016129032256e-05,
+      "loss": 1.7411,
+      "step": 22220
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.4395631551742554,
+      "learning_rate": 4.397681451612903e-05,
+      "loss": 1.8163,
+      "step": 22230
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.4057024717330933,
+      "learning_rate": 4.395161290322581e-05,
+      "loss": 1.9085,
+      "step": 22240
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.1409965753555298,
+      "learning_rate": 4.3926411290322586e-05,
+      "loss": 1.8066,
+      "step": 22250
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.0066628456115723,
+      "learning_rate": 4.390120967741936e-05,
+      "loss": 1.6377,
+      "step": 22260
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.7223976850509644,
+      "learning_rate": 4.3876008064516135e-05,
+      "loss": 2.0277,
+      "step": 22270
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.5174152851104736,
+      "learning_rate": 4.385080645161291e-05,
+      "loss": 1.7989,
+      "step": 22280
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.218111515045166,
+      "learning_rate": 4.3825604838709677e-05,
+      "loss": 1.6248,
+      "step": 22290
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.9883317351341248,
+      "learning_rate": 4.380040322580645e-05,
+      "loss": 1.7194,
+      "step": 22300
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.8839454054832458,
+      "learning_rate": 4.3775201612903225e-05,
+      "loss": 1.6239,
+      "step": 22310
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.4721771478652954,
+      "learning_rate": 4.375e-05,
+      "loss": 1.8155,
+      "step": 22320
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.676626682281494,
+      "learning_rate": 4.3724798387096774e-05,
+      "loss": 1.9091,
+      "step": 22330
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.958523154258728,
+      "learning_rate": 4.3699596774193555e-05,
+      "loss": 1.9617,
+      "step": 22340
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.8137160539627075,
+      "learning_rate": 4.367439516129033e-05,
+      "loss": 1.706,
+      "step": 22350
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.0347504615783691,
+      "learning_rate": 4.36491935483871e-05,
+      "loss": 1.7809,
+      "step": 22360
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.49830952286720276,
+      "learning_rate": 4.362399193548387e-05,
+      "loss": 1.8725,
+      "step": 22370
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.3776698112487793,
+      "learning_rate": 4.3598790322580646e-05,
+      "loss": 1.8004,
+      "step": 22380
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.2521049976348877,
+      "learning_rate": 4.357358870967742e-05,
+      "loss": 1.846,
+      "step": 22390
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.3593287467956543,
+      "learning_rate": 4.3548387096774194e-05,
+      "loss": 1.7662,
+      "step": 22400
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.0529485940933228,
+      "learning_rate": 4.352318548387097e-05,
+      "loss": 1.983,
+      "step": 22410
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.3477754592895508,
+      "learning_rate": 4.349798387096774e-05,
+      "loss": 1.6783,
+      "step": 22420
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.9723754525184631,
+      "learning_rate": 4.347278225806452e-05,
+      "loss": 1.7956,
+      "step": 22430
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.0005680322647095,
+      "learning_rate": 4.344758064516129e-05,
+      "loss": 1.7853,
+      "step": 22440
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.536631166934967,
+      "learning_rate": 4.3422379032258066e-05,
+      "loss": 1.8798,
+      "step": 22450
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.4989845752716064,
+      "learning_rate": 4.339717741935484e-05,
+      "loss": 1.7105,
+      "step": 22460
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.6585626602172852,
+      "learning_rate": 4.3371975806451615e-05,
+      "loss": 1.8084,
+      "step": 22470
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.443318247795105,
+      "learning_rate": 4.334677419354839e-05,
+      "loss": 1.7944,
+      "step": 22480
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.383972406387329,
+      "learning_rate": 4.3321572580645163e-05,
+      "loss": 1.6808,
+      "step": 22490
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.0576753616333008,
+      "learning_rate": 4.329637096774194e-05,
+      "loss": 1.6241,
+      "step": 22500
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.0180659294128418,
+      "learning_rate": 4.327116935483871e-05,
+      "loss": 1.7572,
+      "step": 22510
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.0883879661560059,
+      "learning_rate": 4.3245967741935486e-05,
+      "loss": 1.7867,
+      "step": 22520
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.5569207668304443,
+      "learning_rate": 4.322076612903226e-05,
+      "loss": 1.7995,
+      "step": 22530
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.2687324285507202,
+      "learning_rate": 4.3195564516129035e-05,
+      "loss": 1.6935,
+      "step": 22540
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.5003366470336914,
+      "learning_rate": 4.317036290322581e-05,
+      "loss": 1.676,
+      "step": 22550
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.9432888031005859,
+      "learning_rate": 4.3145161290322584e-05,
+      "loss": 1.5568,
+      "step": 22560
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.111198902130127,
+      "learning_rate": 4.311995967741936e-05,
+      "loss": 1.7751,
+      "step": 22570
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.6922025680541992,
+      "learning_rate": 4.309475806451613e-05,
+      "loss": 1.7328,
+      "step": 22580
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.9782252311706543,
+      "learning_rate": 4.306955645161291e-05,
+      "loss": 1.6734,
+      "step": 22590
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.270436406135559,
+      "learning_rate": 4.3044354838709674e-05,
+      "loss": 1.8289,
+      "step": 22600
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.2441858053207397,
+      "learning_rate": 4.3019153225806456e-05,
+      "loss": 1.825,
+      "step": 22610
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.8007851243019104,
+      "learning_rate": 4.299395161290323e-05,
+      "loss": 1.6808,
+      "step": 22620
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.941435694694519,
+      "learning_rate": 4.2968750000000004e-05,
+      "loss": 1.947,
+      "step": 22630
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.0876191854476929,
+      "learning_rate": 4.294354838709678e-05,
+      "loss": 1.8023,
+      "step": 22640
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.1782952547073364,
+      "learning_rate": 4.291834677419355e-05,
+      "loss": 1.9309,
+      "step": 22650
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.4170387983322144,
+      "learning_rate": 4.289314516129033e-05,
+      "loss": 1.8493,
+      "step": 22660
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.017634391784668,
+      "learning_rate": 4.2867943548387095e-05,
+      "loss": 1.789,
+      "step": 22670
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.075516700744629,
+      "learning_rate": 4.284274193548387e-05,
+      "loss": 1.6101,
+      "step": 22680
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.8773365020751953,
+      "learning_rate": 4.2817540322580644e-05,
+      "loss": 1.7231,
+      "step": 22690
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.2277750968933105,
+      "learning_rate": 4.279233870967742e-05,
+      "loss": 1.6816,
+      "step": 22700
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.6882617473602295,
+      "learning_rate": 4.27671370967742e-05,
+      "loss": 1.804,
+      "step": 22710
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.8893221020698547,
+      "learning_rate": 4.2741935483870973e-05,
+      "loss": 1.6633,
+      "step": 22720
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.6867949366569519,
+      "learning_rate": 4.271673387096775e-05,
+      "loss": 1.9264,
+      "step": 22730
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.3229122161865234,
+      "learning_rate": 4.269153225806452e-05,
+      "loss": 1.9164,
+      "step": 22740
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 5.486172199249268,
+      "learning_rate": 4.266633064516129e-05,
+      "loss": 2.0158,
+      "step": 22750
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.484055995941162,
+      "learning_rate": 4.2641129032258064e-05,
+      "loss": 1.769,
+      "step": 22760
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.596452236175537,
+      "learning_rate": 4.261592741935484e-05,
+      "loss": 1.7498,
+      "step": 22770
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.5534584522247314,
+      "learning_rate": 4.259072580645161e-05,
+      "loss": 1.8015,
+      "step": 22780
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.713290214538574,
+      "learning_rate": 4.256552419354839e-05,
+      "loss": 1.6949,
+      "step": 22790
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.3921265602111816,
+      "learning_rate": 4.254032258064516e-05,
+      "loss": 1.794,
+      "step": 22800
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.998552143573761,
+      "learning_rate": 4.251512096774194e-05,
+      "loss": 1.8637,
+      "step": 22810
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.9259724617004395,
+      "learning_rate": 4.248991935483872e-05,
+      "loss": 1.6586,
+      "step": 22820
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.5849411487579346,
+      "learning_rate": 4.2464717741935484e-05,
+      "loss": 1.8347,
+      "step": 22830
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.9992247223854065,
+      "learning_rate": 4.243951612903226e-05,
+      "loss": 1.9174,
+      "step": 22840
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.5425471067428589,
+      "learning_rate": 4.241431451612903e-05,
+      "loss": 1.8849,
+      "step": 22850
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.9363353848457336,
+      "learning_rate": 4.238911290322581e-05,
+      "loss": 1.7462,
+      "step": 22860
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.4877500534057617,
+      "learning_rate": 4.236391129032258e-05,
+      "loss": 1.9854,
+      "step": 22870
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.139388084411621,
+      "learning_rate": 4.2338709677419356e-05,
+      "loss": 2.007,
+      "step": 22880
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.2652432918548584,
+      "learning_rate": 4.231350806451613e-05,
+      "loss": 1.9081,
+      "step": 22890
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.939110279083252,
+      "learning_rate": 4.2288306451612905e-05,
+      "loss": 1.7409,
+      "step": 22900
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.2444089651107788,
+      "learning_rate": 4.226310483870968e-05,
+      "loss": 1.7267,
+      "step": 22910
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.1097402572631836,
+      "learning_rate": 4.2237903225806454e-05,
+      "loss": 1.8303,
+      "step": 22920
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.5552946329116821,
+      "learning_rate": 4.221270161290323e-05,
+      "loss": 1.6968,
+      "step": 22930
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.155090093612671,
+      "learning_rate": 4.21875e-05,
+      "loss": 1.8279,
+      "step": 22940
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.242652416229248,
+      "learning_rate": 4.2162298387096777e-05,
+      "loss": 1.671,
+      "step": 22950
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.023042917251587,
+      "learning_rate": 4.213709677419355e-05,
+      "loss": 2.0066,
+      "step": 22960
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.245262622833252,
+      "learning_rate": 4.2111895161290325e-05,
+      "loss": 1.829,
+      "step": 22970
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.363004207611084,
+      "learning_rate": 4.208669354838709e-05,
+      "loss": 1.884,
+      "step": 22980
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.0047738552093506,
+      "learning_rate": 4.2061491935483874e-05,
+      "loss": 1.8701,
+      "step": 22990
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.1504148244857788,
+      "learning_rate": 4.203629032258065e-05,
+      "loss": 1.6426,
+      "step": 23000
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.6963468194007874,
+      "learning_rate": 4.201108870967742e-05,
+      "loss": 1.7768,
+      "step": 23010
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.2767564058303833,
+      "learning_rate": 4.19858870967742e-05,
+      "loss": 1.6908,
+      "step": 23020
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.0198200941085815,
+      "learning_rate": 4.196068548387097e-05,
+      "loss": 1.9024,
+      "step": 23030
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.0844767093658447,
+      "learning_rate": 4.1935483870967746e-05,
+      "loss": 1.5983,
+      "step": 23040
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.8455362915992737,
+      "learning_rate": 4.191028225806452e-05,
+      "loss": 1.8226,
+      "step": 23050
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.1943151950836182,
+      "learning_rate": 4.188508064516129e-05,
+      "loss": 1.6146,
+      "step": 23060
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.90919029712677,
+      "learning_rate": 4.185987903225806e-05,
+      "loss": 1.8447,
+      "step": 23070
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.129875898361206,
+      "learning_rate": 4.1834677419354836e-05,
+      "loss": 1.8466,
+      "step": 23080
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.1701180934906006,
+      "learning_rate": 4.180947580645162e-05,
+      "loss": 1.9352,
+      "step": 23090
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.1239447593688965,
+      "learning_rate": 4.178427419354839e-05,
+      "loss": 1.9629,
+      "step": 23100
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.53104829788208,
+      "learning_rate": 4.1759072580645166e-05,
+      "loss": 1.9147,
+      "step": 23110
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.2867636680603027,
+      "learning_rate": 4.173387096774194e-05,
+      "loss": 1.6571,
+      "step": 23120
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.9921642541885376,
+      "learning_rate": 4.1708669354838715e-05,
+      "loss": 1.7023,
+      "step": 23130
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.3662467002868652,
+      "learning_rate": 4.168346774193548e-05,
+      "loss": 1.6648,
+      "step": 23140
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.9485014081001282,
+      "learning_rate": 4.165826612903226e-05,
+      "loss": 1.5427,
+      "step": 23150
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.5467424392700195,
+      "learning_rate": 4.163306451612903e-05,
+      "loss": 1.7442,
+      "step": 23160
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.546236515045166,
+      "learning_rate": 4.1607862903225805e-05,
+      "loss": 1.8463,
+      "step": 23170
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.8755446672439575,
+      "learning_rate": 4.1582661290322587e-05,
+      "loss": 1.5454,
+      "step": 23180
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.7072744369506836,
+      "learning_rate": 4.155745967741936e-05,
+      "loss": 1.7245,
+      "step": 23190
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.0116344690322876,
+      "learning_rate": 4.1532258064516135e-05,
+      "loss": 1.6455,
+      "step": 23200
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.343191146850586,
+      "learning_rate": 4.150705645161291e-05,
+      "loss": 1.8714,
+      "step": 23210
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.32565176486969,
+      "learning_rate": 4.148185483870968e-05,
+      "loss": 1.8714,
+      "step": 23220
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.5232622623443604,
+      "learning_rate": 4.145665322580645e-05,
+      "loss": 1.6454,
+      "step": 23230
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.3971445560455322,
+      "learning_rate": 4.1431451612903226e-05,
+      "loss": 1.8705,
+      "step": 23240
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.1993556022644043,
+      "learning_rate": 4.140625e-05,
+      "loss": 1.8032,
+      "step": 23250
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.355189323425293,
+      "learning_rate": 4.1381048387096775e-05,
+      "loss": 1.7052,
+      "step": 23260
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.8116867542266846,
+      "learning_rate": 4.135584677419355e-05,
+      "loss": 1.7783,
+      "step": 23270
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.7716853618621826,
+      "learning_rate": 4.133064516129033e-05,
+      "loss": 1.7759,
+      "step": 23280
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.0657256841659546,
+      "learning_rate": 4.13054435483871e-05,
+      "loss": 1.7179,
+      "step": 23290
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.2592153549194336,
+      "learning_rate": 4.128024193548387e-05,
+      "loss": 1.7865,
+      "step": 23300
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.4432564973831177,
+      "learning_rate": 4.1255040322580646e-05,
+      "loss": 1.8571,
+      "step": 23310
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.6017426252365112,
+      "learning_rate": 4.122983870967742e-05,
+      "loss": 1.7473,
+      "step": 23320
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.6228058338165283,
+      "learning_rate": 4.1204637096774195e-05,
+      "loss": 1.5958,
+      "step": 23330
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.6755344867706299,
+      "learning_rate": 4.117943548387097e-05,
+      "loss": 1.8926,
+      "step": 23340
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.6232434511184692,
+      "learning_rate": 4.1154233870967744e-05,
+      "loss": 1.7103,
+      "step": 23350
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.3165283203125,
+      "learning_rate": 4.112903225806452e-05,
+      "loss": 1.7659,
+      "step": 23360
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.152485966682434,
+      "learning_rate": 4.110383064516129e-05,
+      "loss": 1.9166,
+      "step": 23370
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.277581810951233,
+      "learning_rate": 4.107862903225807e-05,
+      "loss": 1.7909,
+      "step": 23380
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.6585571765899658,
+      "learning_rate": 4.105342741935484e-05,
+      "loss": 1.8069,
+      "step": 23390
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.3355493545532227,
+      "learning_rate": 4.1028225806451615e-05,
+      "loss": 1.7085,
+      "step": 23400
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.8776298761367798,
+      "learning_rate": 4.100302419354839e-05,
+      "loss": 1.7305,
+      "step": 23410
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.4190964698791504,
+      "learning_rate": 4.0977822580645164e-05,
+      "loss": 1.8173,
+      "step": 23420
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.7752143144607544,
+      "learning_rate": 4.095262096774194e-05,
+      "loss": 1.7942,
+      "step": 23430
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.9198251366615295,
+      "learning_rate": 4.092741935483871e-05,
+      "loss": 1.8111,
+      "step": 23440
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.3498843908309937,
+      "learning_rate": 4.090221774193548e-05,
+      "loss": 1.8403,
+      "step": 23450
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.9643343687057495,
+      "learning_rate": 4.087701612903226e-05,
+      "loss": 1.794,
+      "step": 23460
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.5161173343658447,
+      "learning_rate": 4.0851814516129036e-05,
+      "loss": 1.8189,
+      "step": 23470
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.2990330457687378,
+      "learning_rate": 4.082661290322581e-05,
+      "loss": 1.752,
+      "step": 23480
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.0534782409667969,
+      "learning_rate": 4.0801411290322584e-05,
+      "loss": 1.8251,
+      "step": 23490
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.79275381565094,
+      "learning_rate": 4.077620967741936e-05,
+      "loss": 1.9289,
+      "step": 23500
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.1053781509399414,
+      "learning_rate": 4.075100806451613e-05,
+      "loss": 1.9189,
+      "step": 23510
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.1370339393615723,
+      "learning_rate": 4.072580645161291e-05,
+      "loss": 1.9778,
+      "step": 23520
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.1953496932983398,
+      "learning_rate": 4.0700604838709675e-05,
+      "loss": 1.8123,
+      "step": 23530
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.8414002656936646,
+      "learning_rate": 4.067540322580645e-05,
+      "loss": 2.097,
+      "step": 23540
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.0321738719940186,
+      "learning_rate": 4.0650201612903224e-05,
+      "loss": 1.7369,
+      "step": 23550
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.4152685403823853,
+      "learning_rate": 4.0625000000000005e-05,
+      "loss": 1.9675,
+      "step": 23560
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.2700377702713013,
+      "learning_rate": 4.059979838709678e-05,
+      "loss": 1.8188,
+      "step": 23570
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.0685404539108276,
+      "learning_rate": 4.0574596774193554e-05,
+      "loss": 1.7897,
+      "step": 23580
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.3726766109466553,
+      "learning_rate": 4.054939516129033e-05,
+      "loss": 1.8049,
+      "step": 23590
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.1236485242843628,
+      "learning_rate": 4.0524193548387096e-05,
+      "loss": 1.9355,
+      "step": 23600
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.614508032798767,
+      "learning_rate": 4.049899193548387e-05,
+      "loss": 1.7623,
+      "step": 23610
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.2259633541107178,
+      "learning_rate": 4.0473790322580644e-05,
+      "loss": 1.9119,
+      "step": 23620
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.3583998680114746,
+      "learning_rate": 4.044858870967742e-05,
+      "loss": 1.789,
+      "step": 23630
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.391863226890564,
+      "learning_rate": 4.042338709677419e-05,
+      "loss": 1.6528,
+      "step": 23640
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.2360681295394897,
+      "learning_rate": 4.039818548387097e-05,
+      "loss": 1.767,
+      "step": 23650
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.9706778526306152,
+      "learning_rate": 4.037298387096775e-05,
+      "loss": 1.7103,
+      "step": 23660
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.4039618968963623,
+      "learning_rate": 4.034778225806452e-05,
+      "loss": 1.6384,
+      "step": 23670
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.21925950050354,
+      "learning_rate": 4.032258064516129e-05,
+      "loss": 1.5527,
+      "step": 23680
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.6724603176116943,
+      "learning_rate": 4.0297379032258065e-05,
+      "loss": 1.8741,
+      "step": 23690
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.4752601385116577,
+      "learning_rate": 4.027217741935484e-05,
+      "loss": 1.875,
+      "step": 23700
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.5255006551742554,
+      "learning_rate": 4.024697580645161e-05,
+      "loss": 1.7745,
+      "step": 23710
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.9203068017959595,
+      "learning_rate": 4.022177419354839e-05,
+      "loss": 1.9588,
+      "step": 23720
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.9981523752212524,
+      "learning_rate": 4.019657258064516e-05,
+      "loss": 1.7997,
+      "step": 23730
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.159008264541626,
+      "learning_rate": 4.0171370967741936e-05,
+      "loss": 1.8854,
+      "step": 23740
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.9218286275863647,
+      "learning_rate": 4.014616935483872e-05,
+      "loss": 1.9407,
+      "step": 23750
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.5016955137252808,
+      "learning_rate": 4.0120967741935485e-05,
+      "loss": 1.8327,
+      "step": 23760
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.9382520914077759,
+      "learning_rate": 4.009576612903226e-05,
+      "loss": 1.9372,
+      "step": 23770
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.0661925077438354,
+      "learning_rate": 4.0070564516129034e-05,
+      "loss": 1.7812,
+      "step": 23780
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.9683375954627991,
+      "learning_rate": 4.004536290322581e-05,
+      "loss": 1.8012,
+      "step": 23790
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.1434125900268555,
+      "learning_rate": 4.002016129032258e-05,
+      "loss": 1.719,
+      "step": 23800
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.0790057182312012,
+      "learning_rate": 3.999495967741936e-05,
+      "loss": 2.0013,
+      "step": 23810
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.8281884789466858,
+      "learning_rate": 3.996975806451613e-05,
+      "loss": 1.8654,
+      "step": 23820
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.2483556270599365,
+      "learning_rate": 3.9944556451612905e-05,
+      "loss": 1.8435,
+      "step": 23830
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.9614571332931519,
+      "learning_rate": 3.991935483870968e-05,
+      "loss": 1.7879,
+      "step": 23840
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.502971887588501,
+      "learning_rate": 3.9894153225806454e-05,
+      "loss": 1.7301,
+      "step": 23850
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.4814743995666504,
+      "learning_rate": 3.986895161290323e-05,
+      "loss": 1.7412,
+      "step": 23860
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.6112923622131348,
+      "learning_rate": 3.984375e-05,
+      "loss": 1.7124,
+      "step": 23870
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.3875453472137451,
+      "learning_rate": 3.981854838709678e-05,
+      "loss": 1.6445,
+      "step": 23880
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.2181715965270996,
+      "learning_rate": 3.979334677419355e-05,
+      "loss": 1.7673,
+      "step": 23890
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.4868775606155396,
+      "learning_rate": 3.9768145161290326e-05,
+      "loss": 1.859,
+      "step": 23900
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.090408444404602,
+      "learning_rate": 3.9742943548387093e-05,
+      "loss": 1.6683,
+      "step": 23910
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.0764517784118652,
+      "learning_rate": 3.971774193548387e-05,
+      "loss": 1.8103,
+      "step": 23920
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.7064043283462524,
+      "learning_rate": 3.969254032258065e-05,
+      "loss": 1.7862,
+      "step": 23930
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.0479668378829956,
+      "learning_rate": 3.966733870967742e-05,
+      "loss": 1.7574,
+      "step": 23940
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.0704200267791748,
+      "learning_rate": 3.96421370967742e-05,
+      "loss": 1.8547,
+      "step": 23950
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.377324104309082,
+      "learning_rate": 3.961693548387097e-05,
+      "loss": 1.8704,
+      "step": 23960
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.83745276927948,
+      "learning_rate": 3.9591733870967746e-05,
+      "loss": 1.8923,
+      "step": 23970
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.3475624322891235,
+      "learning_rate": 3.956653225806452e-05,
+      "loss": 1.7506,
+      "step": 23980
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.1462305784225464,
+      "learning_rate": 3.954133064516129e-05,
+      "loss": 1.6918,
+      "step": 23990
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.224318504333496,
+      "learning_rate": 3.951612903225806e-05,
+      "loss": 1.8268,
+      "step": 24000
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.4191064834594727,
+      "learning_rate": 3.949092741935484e-05,
+      "loss": 1.8541,
+      "step": 24010
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.4490617513656616,
+      "learning_rate": 3.946572580645161e-05,
+      "loss": 1.813,
+      "step": 24020
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.7568625211715698,
+      "learning_rate": 3.944052419354839e-05,
+      "loss": 1.8139,
+      "step": 24030
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.3291369676589966,
+      "learning_rate": 3.941532258064517e-05,
+      "loss": 1.8764,
+      "step": 24040
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.384442687034607,
+      "learning_rate": 3.939012096774194e-05,
+      "loss": 1.8032,
+      "step": 24050
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.3003216981887817,
+      "learning_rate": 3.9364919354838715e-05,
+      "loss": 1.6497,
+      "step": 24060
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.0441828966140747,
+      "learning_rate": 3.933971774193548e-05,
+      "loss": 1.8188,
+      "step": 24070
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.949785053730011,
+      "learning_rate": 3.931451612903226e-05,
+      "loss": 1.9186,
+      "step": 24080
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.9812824130058289,
+      "learning_rate": 3.928931451612903e-05,
+      "loss": 1.8422,
+      "step": 24090
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.3022220134735107,
+      "learning_rate": 3.9264112903225806e-05,
+      "loss": 2.1605,
+      "step": 24100
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.0428085327148438,
+      "learning_rate": 3.923891129032258e-05,
+      "loss": 1.8271,
+      "step": 24110
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.5291430950164795,
+      "learning_rate": 3.9213709677419355e-05,
+      "loss": 1.8512,
+      "step": 24120
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.1014132499694824,
+      "learning_rate": 3.9188508064516136e-05,
+      "loss": 1.6838,
+      "step": 24130
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.2071014642715454,
+      "learning_rate": 3.916330645161291e-05,
+      "loss": 1.7847,
+      "step": 24140
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.315010666847229,
+      "learning_rate": 3.913810483870968e-05,
+      "loss": 1.6851,
+      "step": 24150
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.6689895391464233,
+      "learning_rate": 3.911290322580645e-05,
+      "loss": 1.8896,
+      "step": 24160
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.9586811065673828,
+      "learning_rate": 3.9087701612903226e-05,
+      "loss": 1.9418,
+      "step": 24170
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.3839447498321533,
+      "learning_rate": 3.90625e-05,
+      "loss": 1.7968,
+      "step": 24180
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.0015790462493896,
+      "learning_rate": 3.9037298387096775e-05,
+      "loss": 1.8997,
+      "step": 24190
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.6623173952102661,
+      "learning_rate": 3.901209677419355e-05,
+      "loss": 1.7886,
+      "step": 24200
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.7958637475967407,
+      "learning_rate": 3.8986895161290324e-05,
+      "loss": 1.7106,
+      "step": 24210
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.7138738036155701,
+      "learning_rate": 3.89616935483871e-05,
+      "loss": 1.499,
+      "step": 24220
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.6506472826004028,
+      "learning_rate": 3.893649193548387e-05,
+      "loss": 1.7246,
+      "step": 24230
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.6184237003326416,
+      "learning_rate": 3.891129032258065e-05,
+      "loss": 1.7272,
+      "step": 24240
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.450455904006958,
+      "learning_rate": 3.888608870967742e-05,
+      "loss": 2.0226,
+      "step": 24250
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.071236491203308,
+      "learning_rate": 3.8860887096774196e-05,
+      "loss": 1.8723,
+      "step": 24260
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.3186967372894287,
+      "learning_rate": 3.883568548387097e-05,
+      "loss": 1.8495,
+      "step": 24270
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.524312973022461,
+      "learning_rate": 3.8810483870967744e-05,
+      "loss": 1.8253,
+      "step": 24280
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.3066790103912354,
+      "learning_rate": 3.878528225806452e-05,
+      "loss": 1.6887,
+      "step": 24290
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.387058138847351,
+      "learning_rate": 3.876008064516129e-05,
+      "loss": 1.8848,
+      "step": 24300
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.2423301935195923,
+      "learning_rate": 3.873487903225807e-05,
+      "loss": 1.8359,
+      "step": 24310
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.163832187652588,
+      "learning_rate": 3.870967741935484e-05,
+      "loss": 1.7618,
+      "step": 24320
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.5879409313201904,
+      "learning_rate": 3.8684475806451616e-05,
+      "loss": 1.9942,
+      "step": 24330
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.092371940612793,
+      "learning_rate": 3.865927419354839e-05,
+      "loss": 1.9391,
+      "step": 24340
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.067067265510559,
+      "learning_rate": 3.8634072580645165e-05,
+      "loss": 1.895,
+      "step": 24350
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.1819530725479126,
+      "learning_rate": 3.860887096774194e-05,
+      "loss": 1.848,
+      "step": 24360
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.9437178373336792,
+      "learning_rate": 3.858366935483871e-05,
+      "loss": 1.9907,
+      "step": 24370
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.032775640487671,
+      "learning_rate": 3.855846774193548e-05,
+      "loss": 1.7751,
+      "step": 24380
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.629361867904663,
+      "learning_rate": 3.8533266129032255e-05,
+      "loss": 1.8843,
+      "step": 24390
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.1500575542449951,
+      "learning_rate": 3.8508064516129036e-05,
+      "loss": 1.713,
+      "step": 24400
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.778483510017395,
+      "learning_rate": 3.848286290322581e-05,
+      "loss": 1.71,
+      "step": 24410
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.1274770498275757,
+      "learning_rate": 3.8457661290322585e-05,
+      "loss": 1.5343,
+      "step": 24420
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.0447540283203125,
+      "learning_rate": 3.843245967741936e-05,
+      "loss": 1.7446,
+      "step": 24430
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.0831520557403564,
+      "learning_rate": 3.8407258064516134e-05,
+      "loss": 1.6866,
+      "step": 24440
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.0620754957199097,
+      "learning_rate": 3.838205645161291e-05,
+      "loss": 1.9021,
+      "step": 24450
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.0346617698669434,
+      "learning_rate": 3.8356854838709676e-05,
+      "loss": 1.7808,
+      "step": 24460
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.2563505172729492,
+      "learning_rate": 3.833165322580645e-05,
+      "loss": 1.8789,
+      "step": 24470
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.2310060262680054,
+      "learning_rate": 3.8306451612903224e-05,
+      "loss": 1.7811,
+      "step": 24480
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.1308537721633911,
+      "learning_rate": 3.828125e-05,
+      "loss": 1.8845,
+      "step": 24490
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.64974045753479,
+      "learning_rate": 3.825604838709678e-05,
+      "loss": 1.6972,
+      "step": 24500
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.2533867359161377,
+      "learning_rate": 3.8230846774193554e-05,
+      "loss": 1.8521,
+      "step": 24510
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.5066378116607666,
+      "learning_rate": 3.820564516129033e-05,
+      "loss": 1.6815,
+      "step": 24520
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.7755640745162964,
+      "learning_rate": 3.8180443548387096e-05,
+      "loss": 1.77,
+      "step": 24530
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.0661076307296753,
+      "learning_rate": 3.815524193548387e-05,
+      "loss": 1.7635,
+      "step": 24540
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.3134467601776123,
+      "learning_rate": 3.8130040322580645e-05,
+      "loss": 1.7834,
+      "step": 24550
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.4487885236740112,
+      "learning_rate": 3.810483870967742e-05,
+      "loss": 1.8639,
+      "step": 24560
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.0633201599121094,
+      "learning_rate": 3.8079637096774194e-05,
+      "loss": 1.727,
+      "step": 24570
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.6250972747802734,
+      "learning_rate": 3.805443548387097e-05,
+      "loss": 1.9077,
+      "step": 24580
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.1020195484161377,
+      "learning_rate": 3.802923387096774e-05,
+      "loss": 1.8161,
+      "step": 24590
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.1061317920684814,
+      "learning_rate": 3.800403225806452e-05,
+      "loss": 1.9629,
+      "step": 24600
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.471360445022583,
+      "learning_rate": 3.797883064516129e-05,
+      "loss": 1.8283,
+      "step": 24610
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.1406197547912598,
+      "learning_rate": 3.7953629032258065e-05,
+      "loss": 2.0909,
+      "step": 24620
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.8420767784118652,
+      "learning_rate": 3.792842741935484e-05,
+      "loss": 1.8162,
+      "step": 24630
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.1442750692367554,
+      "learning_rate": 3.7903225806451614e-05,
+      "loss": 1.8545,
+      "step": 24640
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.8323175311088562,
+      "learning_rate": 3.787802419354839e-05,
+      "loss": 1.7519,
+      "step": 24650
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.0051608085632324,
+      "learning_rate": 3.785282258064516e-05,
+      "loss": 1.7973,
+      "step": 24660
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.31570565700531,
+      "learning_rate": 3.782762096774194e-05,
+      "loss": 1.734,
+      "step": 24670
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.770530641078949,
+      "learning_rate": 3.780241935483871e-05,
+      "loss": 1.9565,
+      "step": 24680
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.135790467262268,
+      "learning_rate": 3.7777217741935486e-05,
+      "loss": 1.9171,
+      "step": 24690
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.6468251943588257,
+      "learning_rate": 3.775201612903226e-05,
+      "loss": 1.7058,
+      "step": 24700
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 3.16940975189209,
+      "learning_rate": 3.7726814516129034e-05,
+      "loss": 1.8907,
+      "step": 24710
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.628682255744934,
+      "learning_rate": 3.770161290322581e-05,
+      "loss": 1.9367,
+      "step": 24720
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.2755368947982788,
+      "learning_rate": 3.767641129032258e-05,
+      "loss": 1.9503,
+      "step": 24730
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.1169670820236206,
+      "learning_rate": 3.765120967741936e-05,
+      "loss": 1.8733,
+      "step": 24740
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.6209869384765625,
+      "learning_rate": 3.762600806451613e-05,
+      "loss": 1.7435,
+      "step": 24750
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.7603099346160889,
+      "learning_rate": 3.7600806451612906e-05,
+      "loss": 1.9338,
+      "step": 24760
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.9309702515602112,
+      "learning_rate": 3.757560483870968e-05,
+      "loss": 1.6184,
+      "step": 24770
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.0792899131774902,
+      "learning_rate": 3.7550403225806455e-05,
+      "loss": 1.8927,
+      "step": 24780
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.5750387907028198,
+      "learning_rate": 3.752520161290323e-05,
+      "loss": 2.0081,
+      "step": 24790
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.2335008382797241,
+      "learning_rate": 3.7500000000000003e-05,
+      "loss": 1.8987,
+      "step": 24800
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.3331036567687988,
+      "learning_rate": 3.747479838709678e-05,
+      "loss": 1.7553,
+      "step": 24810
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.0697460174560547,
+      "learning_rate": 3.744959677419355e-05,
+      "loss": 1.7737,
+      "step": 24820
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 3.1794989109039307,
+      "learning_rate": 3.7424395161290327e-05,
+      "loss": 1.8751,
+      "step": 24830
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.702677607536316,
+      "learning_rate": 3.7399193548387094e-05,
+      "loss": 1.8087,
+      "step": 24840
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.4494045972824097,
+      "learning_rate": 3.737399193548387e-05,
+      "loss": 1.6905,
+      "step": 24850
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.7570825815200806,
+      "learning_rate": 3.734879032258064e-05,
+      "loss": 1.555,
+      "step": 24860
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.2257930040359497,
+      "learning_rate": 3.7323588709677424e-05,
+      "loss": 1.7075,
+      "step": 24870
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.4381628036499023,
+      "learning_rate": 3.72983870967742e-05,
+      "loss": 1.893,
+      "step": 24880
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.0227274894714355,
+      "learning_rate": 3.727318548387097e-05,
+      "loss": 1.6779,
+      "step": 24890
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.2567925453186035,
+      "learning_rate": 3.724798387096775e-05,
+      "loss": 1.8427,
+      "step": 24900
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.597618818283081,
+      "learning_rate": 3.722278225806452e-05,
+      "loss": 1.786,
+      "step": 24910
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.8189289569854736,
+      "learning_rate": 3.719758064516129e-05,
+      "loss": 1.8703,
+      "step": 24920
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.5442767143249512,
+      "learning_rate": 3.717237903225806e-05,
+      "loss": 1.7285,
+      "step": 24930
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.3350988626480103,
+      "learning_rate": 3.714717741935484e-05,
+      "loss": 1.9491,
+      "step": 24940
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.3546431064605713,
+      "learning_rate": 3.712197580645161e-05,
+      "loss": 1.7294,
+      "step": 24950
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.195425033569336,
+      "learning_rate": 3.7096774193548386e-05,
+      "loss": 1.9264,
+      "step": 24960
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.775041937828064,
+      "learning_rate": 3.707157258064517e-05,
+      "loss": 1.953,
+      "step": 24970
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.1296374797821045,
+      "learning_rate": 3.704637096774194e-05,
+      "loss": 1.8273,
+      "step": 24980
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.7081122398376465,
+      "learning_rate": 3.7021169354838716e-05,
+      "loss": 1.7934,
+      "step": 24990
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.3087762594223022,
+      "learning_rate": 3.6995967741935484e-05,
+      "loss": 1.8773,
+      "step": 25000
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.5830187797546387,
+      "learning_rate": 3.697076612903226e-05,
+      "loss": 1.7605,
+      "step": 25010
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.9942324757575989,
+      "learning_rate": 3.694556451612903e-05,
+      "loss": 2.0084,
+      "step": 25020
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.378139853477478,
+      "learning_rate": 3.692036290322581e-05,
+      "loss": 1.8835,
+      "step": 25030
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.712257742881775,
+      "learning_rate": 3.689516129032258e-05,
+      "loss": 1.847,
+      "step": 25040
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.5576796531677246,
+      "learning_rate": 3.6869959677419355e-05,
+      "loss": 1.6343,
+      "step": 25050
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.269313097000122,
+      "learning_rate": 3.684475806451613e-05,
+      "loss": 1.9287,
+      "step": 25060
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.285069465637207,
+      "learning_rate": 3.681955645161291e-05,
+      "loss": 1.833,
+      "step": 25070
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.1856868267059326,
+      "learning_rate": 3.679435483870968e-05,
+      "loss": 1.6919,
+      "step": 25080
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.256518840789795,
+      "learning_rate": 3.676915322580645e-05,
+      "loss": 1.8291,
+      "step": 25090
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.947369933128357,
+      "learning_rate": 3.674395161290323e-05,
+      "loss": 1.8115,
+      "step": 25100
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.7294474840164185,
+      "learning_rate": 3.671875e-05,
+      "loss": 1.8249,
+      "step": 25110
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.418908953666687,
+      "learning_rate": 3.6693548387096776e-05,
+      "loss": 1.7751,
+      "step": 25120
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.3774021863937378,
+      "learning_rate": 3.666834677419355e-05,
+      "loss": 1.8365,
+      "step": 25130
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.2795604467391968,
+      "learning_rate": 3.6643145161290324e-05,
+      "loss": 1.8595,
+      "step": 25140
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.6327664852142334,
+      "learning_rate": 3.66179435483871e-05,
+      "loss": 1.8411,
+      "step": 25150
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.1739729642868042,
+      "learning_rate": 3.659274193548387e-05,
+      "loss": 1.8513,
+      "step": 25160
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.5523697137832642,
+      "learning_rate": 3.656754032258065e-05,
+      "loss": 1.7977,
+      "step": 25170
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.4053339958190918,
+      "learning_rate": 3.654233870967742e-05,
+      "loss": 1.8773,
+      "step": 25180
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.9556178450584412,
+      "learning_rate": 3.6517137096774196e-05,
+      "loss": 1.9466,
+      "step": 25190
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.9517956972122192,
+      "learning_rate": 3.649193548387097e-05,
+      "loss": 1.7659,
+      "step": 25200
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.5763143301010132,
+      "learning_rate": 3.6466733870967745e-05,
+      "loss": 1.8534,
+      "step": 25210
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.4103869199752808,
+      "learning_rate": 3.644153225806452e-05,
+      "loss": 1.8629,
+      "step": 25220
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.5512285232543945,
+      "learning_rate": 3.641633064516129e-05,
+      "loss": 1.7406,
+      "step": 25230
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.5158394575119019,
+      "learning_rate": 3.639112903225806e-05,
+      "loss": 1.8695,
+      "step": 25240
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.260918378829956,
+      "learning_rate": 3.636592741935484e-05,
+      "loss": 1.9032,
+      "step": 25250
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.2540242671966553,
+      "learning_rate": 3.634072580645162e-05,
+      "loss": 1.7099,
+      "step": 25260
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.1208174228668213,
+      "learning_rate": 3.631552419354839e-05,
+      "loss": 1.8374,
+      "step": 25270
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.0559520721435547,
+      "learning_rate": 3.6290322580645165e-05,
+      "loss": 1.7271,
+      "step": 25280
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.861675500869751,
+      "learning_rate": 3.626512096774194e-05,
+      "loss": 1.8563,
+      "step": 25290
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.2166972160339355,
+      "learning_rate": 3.6239919354838714e-05,
+      "loss": 1.7364,
+      "step": 25300
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.8407237529754639,
+      "learning_rate": 3.621471774193548e-05,
+      "loss": 1.8498,
+      "step": 25310
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.7857238054275513,
+      "learning_rate": 3.6189516129032256e-05,
+      "loss": 1.7596,
+      "step": 25320
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.04689621925354,
+      "learning_rate": 3.616431451612903e-05,
+      "loss": 1.5507,
+      "step": 25330
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.08146071434021,
+      "learning_rate": 3.6139112903225805e-05,
+      "loss": 1.8535,
+      "step": 25340
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.2259330749511719,
+      "learning_rate": 3.6113911290322586e-05,
+      "loss": 1.7325,
+      "step": 25350
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.812008261680603,
+      "learning_rate": 3.608870967741936e-05,
+      "loss": 1.8737,
+      "step": 25360
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.1980161666870117,
+      "learning_rate": 3.6063508064516134e-05,
+      "loss": 1.907,
+      "step": 25370
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.0286495685577393,
+      "learning_rate": 3.603830645161291e-05,
+      "loss": 1.9432,
+      "step": 25380
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.4331778287887573,
+      "learning_rate": 3.6013104838709676e-05,
+      "loss": 1.9502,
+      "step": 25390
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.4660382270812988,
+      "learning_rate": 3.598790322580645e-05,
+      "loss": 1.754,
+      "step": 25400
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.1045109033584595,
+      "learning_rate": 3.5962701612903225e-05,
+      "loss": 1.669,
+      "step": 25410
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.1277270317077637,
+      "learning_rate": 3.59375e-05,
+      "loss": 1.8886,
+      "step": 25420
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.0550090074539185,
+      "learning_rate": 3.5912298387096774e-05,
+      "loss": 1.6776,
+      "step": 25430
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.9288767576217651,
+      "learning_rate": 3.5887096774193555e-05,
+      "loss": 1.8463,
+      "step": 25440
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.6355587244033813,
+      "learning_rate": 3.586189516129033e-05,
+      "loss": 1.9285,
+      "step": 25450
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.7874634861946106,
+      "learning_rate": 3.58366935483871e-05,
+      "loss": 1.5618,
+      "step": 25460
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.282806634902954,
+      "learning_rate": 3.581149193548387e-05,
+      "loss": 1.674,
+      "step": 25470
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.7398942708969116,
+      "learning_rate": 3.5786290322580645e-05,
+      "loss": 1.8847,
+      "step": 25480
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.1322602033615112,
+      "learning_rate": 3.576108870967742e-05,
+      "loss": 1.9004,
+      "step": 25490
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.503607988357544,
+      "learning_rate": 3.5735887096774194e-05,
+      "loss": 1.817,
+      "step": 25500
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.9812738299369812,
+      "learning_rate": 3.571068548387097e-05,
+      "loss": 1.9299,
+      "step": 25510
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.604263424873352,
+      "learning_rate": 3.568548387096774e-05,
+      "loss": 1.8963,
+      "step": 25520
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.1754772663116455,
+      "learning_rate": 3.566028225806452e-05,
+      "loss": 1.6099,
+      "step": 25530
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.5675863027572632,
+      "learning_rate": 3.563508064516129e-05,
+      "loss": 1.7519,
+      "step": 25540
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.4346174001693726,
+      "learning_rate": 3.5609879032258066e-05,
+      "loss": 1.6807,
+      "step": 25550
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.6783623695373535,
+      "learning_rate": 3.558467741935484e-05,
+      "loss": 1.9179,
+      "step": 25560
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.9784950017929077,
+      "learning_rate": 3.5559475806451615e-05,
+      "loss": 1.8568,
+      "step": 25570
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.2225348949432373,
+      "learning_rate": 3.553427419354839e-05,
+      "loss": 1.6736,
+      "step": 25580
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.4139320850372314,
+      "learning_rate": 3.550907258064516e-05,
+      "loss": 1.9013,
+      "step": 25590
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.2433747053146362,
+      "learning_rate": 3.548387096774194e-05,
+      "loss": 1.8074,
+      "step": 25600
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.1760929822921753,
+      "learning_rate": 3.545866935483871e-05,
+      "loss": 1.5958,
+      "step": 25610
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.6994646787643433,
+      "learning_rate": 3.5433467741935486e-05,
+      "loss": 1.9234,
+      "step": 25620
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.421981692314148,
+      "learning_rate": 3.540826612903226e-05,
+      "loss": 1.8143,
+      "step": 25630
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.8699542880058289,
+      "learning_rate": 3.5383064516129035e-05,
+      "loss": 1.7852,
+      "step": 25640
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.5063714981079102,
+      "learning_rate": 3.535786290322581e-05,
+      "loss": 2.0113,
+      "step": 25650
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.9216351509094238,
+      "learning_rate": 3.5332661290322584e-05,
+      "loss": 1.7316,
+      "step": 25660
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.1318509578704834,
+      "learning_rate": 3.530745967741936e-05,
+      "loss": 2.0372,
+      "step": 25670
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.4636210203170776,
+      "learning_rate": 3.528225806451613e-05,
+      "loss": 1.8048,
+      "step": 25680
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.45465087890625,
+      "learning_rate": 3.525705645161291e-05,
+      "loss": 1.6585,
+      "step": 25690
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.9789825677871704,
+      "learning_rate": 3.5231854838709674e-05,
+      "loss": 1.7503,
+      "step": 25700
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.176859140396118,
+      "learning_rate": 3.520665322580645e-05,
+      "loss": 1.5485,
+      "step": 25710
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.4404244422912598,
+      "learning_rate": 3.518145161290323e-05,
+      "loss": 1.7962,
+      "step": 25720
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.6046651601791382,
+      "learning_rate": 3.5156250000000004e-05,
+      "loss": 1.7353,
+      "step": 25730
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.0327179431915283,
+      "learning_rate": 3.513104838709678e-05,
+      "loss": 1.8836,
+      "step": 25740
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.9276291728019714,
+      "learning_rate": 3.510584677419355e-05,
+      "loss": 1.5586,
+      "step": 25750
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.6838963031768799,
+      "learning_rate": 3.508064516129033e-05,
+      "loss": 1.7004,
+      "step": 25760
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.2518823146820068,
+      "learning_rate": 3.5055443548387095e-05,
+      "loss": 1.864,
+      "step": 25770
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.6771855354309082,
+      "learning_rate": 3.503024193548387e-05,
+      "loss": 1.7621,
+      "step": 25780
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.7889933586120605,
+      "learning_rate": 3.5005040322580643e-05,
+      "loss": 1.753,
+      "step": 25790
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.4459565877914429,
+      "learning_rate": 3.497983870967742e-05,
+      "loss": 1.6934,
+      "step": 25800
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.14077627658844,
+      "learning_rate": 3.495463709677419e-05,
+      "loss": 1.725,
+      "step": 25810
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.1523497104644775,
+      "learning_rate": 3.492943548387097e-05,
+      "loss": 1.7193,
+      "step": 25820
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.6188344955444336,
+      "learning_rate": 3.490423387096775e-05,
+      "loss": 1.7382,
+      "step": 25830
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.8791712522506714,
+      "learning_rate": 3.487903225806452e-05,
+      "loss": 1.597,
+      "step": 25840
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.0641663074493408,
+      "learning_rate": 3.485383064516129e-05,
+      "loss": 1.8671,
+      "step": 25850
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.6088355779647827,
+      "learning_rate": 3.4828629032258064e-05,
+      "loss": 1.853,
+      "step": 25860
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.1177278757095337,
+      "learning_rate": 3.480342741935484e-05,
+      "loss": 1.7546,
+      "step": 25870
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.2637892961502075,
+      "learning_rate": 3.477822580645161e-05,
+      "loss": 1.8247,
+      "step": 25880
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.1699740886688232,
+      "learning_rate": 3.475302419354839e-05,
+      "loss": 1.6737,
+      "step": 25890
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.025829792022705,
+      "learning_rate": 3.472782258064516e-05,
+      "loss": 1.9319,
+      "step": 25900
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.7259023189544678,
+      "learning_rate": 3.4702620967741936e-05,
+      "loss": 1.7061,
+      "step": 25910
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.683084487915039,
+      "learning_rate": 3.467741935483872e-05,
+      "loss": 1.8412,
+      "step": 25920
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.8769229650497437,
+      "learning_rate": 3.4652217741935484e-05,
+      "loss": 1.7608,
+      "step": 25930
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.296434164047241,
+      "learning_rate": 3.462701612903226e-05,
+      "loss": 1.9031,
+      "step": 25940
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.3907055854797363,
+      "learning_rate": 3.460181451612903e-05,
+      "loss": 1.8034,
+      "step": 25950
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.081751585006714,
+      "learning_rate": 3.457661290322581e-05,
+      "loss": 1.9616,
+      "step": 25960
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.4939967393875122,
+      "learning_rate": 3.455141129032258e-05,
+      "loss": 1.7541,
+      "step": 25970
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.9856767654418945,
+      "learning_rate": 3.4526209677419356e-05,
+      "loss": 1.8213,
+      "step": 25980
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.4664232730865479,
+      "learning_rate": 3.450100806451613e-05,
+      "loss": 1.8614,
+      "step": 25990
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.9345733523368835,
+      "learning_rate": 3.4475806451612905e-05,
+      "loss": 1.7999,
+      "step": 26000
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.4444332122802734,
+      "learning_rate": 3.445060483870968e-05,
+      "loss": 1.8209,
+      "step": 26010
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.468727707862854,
+      "learning_rate": 3.442540322580645e-05,
+      "loss": 1.6341,
+      "step": 26020
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.114827036857605,
+      "learning_rate": 3.440020161290323e-05,
+      "loss": 1.6651,
+      "step": 26030
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.8991426229476929,
+      "learning_rate": 3.4375e-05,
+      "loss": 1.8338,
+      "step": 26040
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.430335283279419,
+      "learning_rate": 3.4349798387096776e-05,
+      "loss": 1.7538,
+      "step": 26050
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.4808684587478638,
+      "learning_rate": 3.432459677419355e-05,
+      "loss": 1.7375,
+      "step": 26060
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.382254719734192,
+      "learning_rate": 3.4299395161290325e-05,
+      "loss": 1.9264,
+      "step": 26070
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.062035083770752,
+      "learning_rate": 3.427419354838709e-05,
+      "loss": 1.6609,
+      "step": 26080
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.433593511581421,
+      "learning_rate": 3.4248991935483874e-05,
+      "loss": 1.6754,
+      "step": 26090
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.868866205215454,
+      "learning_rate": 3.422379032258065e-05,
+      "loss": 1.5724,
+      "step": 26100
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.9414305686950684,
+      "learning_rate": 3.419858870967742e-05,
+      "loss": 1.8975,
+      "step": 26110
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.9743680357933044,
+      "learning_rate": 3.41733870967742e-05,
+      "loss": 1.7021,
+      "step": 26120
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.6589338779449463,
+      "learning_rate": 3.414818548387097e-05,
+      "loss": 1.9367,
+      "step": 26130
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.7130793333053589,
+      "learning_rate": 3.4122983870967746e-05,
+      "loss": 1.8759,
+      "step": 26140
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.6999446153640747,
+      "learning_rate": 3.409778225806452e-05,
+      "loss": 1.9071,
+      "step": 26150
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.3948981761932373,
+      "learning_rate": 3.407258064516129e-05,
+      "loss": 1.7141,
+      "step": 26160
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.7755945920944214,
+      "learning_rate": 3.404737903225806e-05,
+      "loss": 1.6974,
+      "step": 26170
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.2274476289749146,
+      "learning_rate": 3.4022177419354836e-05,
+      "loss": 1.7842,
+      "step": 26180
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.407785654067993,
+      "learning_rate": 3.399697580645162e-05,
+      "loss": 1.8878,
+      "step": 26190
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.0912842750549316,
+      "learning_rate": 3.397177419354839e-05,
+      "loss": 1.7906,
+      "step": 26200
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.1870064735412598,
+      "learning_rate": 3.3946572580645166e-05,
+      "loss": 1.6752,
+      "step": 26210
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.8598021268844604,
+      "learning_rate": 3.392137096774194e-05,
+      "loss": 2.0122,
+      "step": 26220
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.7128539085388184,
+      "learning_rate": 3.3896169354838715e-05,
+      "loss": 1.8609,
+      "step": 26230
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.5130163431167603,
+      "learning_rate": 3.387096774193548e-05,
+      "loss": 1.8362,
+      "step": 26240
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.9417235851287842,
+      "learning_rate": 3.3845766129032257e-05,
+      "loss": 1.8383,
+      "step": 26250
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.3714032173156738,
+      "learning_rate": 3.382056451612903e-05,
+      "loss": 2.0447,
+      "step": 26260
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.1071736812591553,
+      "learning_rate": 3.3795362903225805e-05,
+      "loss": 1.7661,
+      "step": 26270
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.0916730165481567,
+      "learning_rate": 3.377016129032258e-05,
+      "loss": 1.6815,
+      "step": 26280
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.6945453882217407,
+      "learning_rate": 3.374495967741936e-05,
+      "loss": 1.8322,
+      "step": 26290
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.9634777307510376,
+      "learning_rate": 3.3719758064516135e-05,
+      "loss": 1.803,
+      "step": 26300
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.6526271104812622,
+      "learning_rate": 3.369455645161291e-05,
+      "loss": 1.7503,
+      "step": 26310
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.206174373626709,
+      "learning_rate": 3.366935483870968e-05,
+      "loss": 1.8835,
+      "step": 26320
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.516771912574768,
+      "learning_rate": 3.364415322580645e-05,
+      "loss": 2.0233,
+      "step": 26330
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.3420190811157227,
+      "learning_rate": 3.3618951612903226e-05,
+      "loss": 1.7679,
+      "step": 26340
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.1771280765533447,
+      "learning_rate": 3.359375e-05,
+      "loss": 1.7168,
+      "step": 26350
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.9769381284713745,
+      "learning_rate": 3.3568548387096774e-05,
+      "loss": 1.7419,
+      "step": 26360
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.9492793083190918,
+      "learning_rate": 3.354334677419355e-05,
+      "loss": 1.8024,
+      "step": 26370
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.067551851272583,
+      "learning_rate": 3.351814516129032e-05,
+      "loss": 1.7563,
+      "step": 26380
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 4.010818958282471,
+      "learning_rate": 3.34929435483871e-05,
+      "loss": 1.8587,
+      "step": 26390
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.020967960357666,
+      "learning_rate": 3.346774193548387e-05,
+      "loss": 1.6669,
+      "step": 26400
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.60180401802063,
+      "learning_rate": 3.3442540322580646e-05,
+      "loss": 1.7106,
+      "step": 26410
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.6215604543685913,
+      "learning_rate": 3.341733870967742e-05,
+      "loss": 1.966,
+      "step": 26420
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 3.0315775871276855,
+      "learning_rate": 3.3392137096774195e-05,
+      "loss": 1.577,
+      "step": 26430
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.394742727279663,
+      "learning_rate": 3.336693548387097e-05,
+      "loss": 1.5626,
+      "step": 26440
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.0331069231033325,
+      "learning_rate": 3.3341733870967743e-05,
+      "loss": 1.8373,
+      "step": 26450
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.7054378986358643,
+      "learning_rate": 3.331653225806452e-05,
+      "loss": 1.8475,
+      "step": 26460
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.003691554069519,
+      "learning_rate": 3.329133064516129e-05,
+      "loss": 1.578,
+      "step": 26470
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.590470790863037,
+      "learning_rate": 3.3266129032258067e-05,
+      "loss": 1.6463,
+      "step": 26480
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.7741776704788208,
+      "learning_rate": 3.324092741935484e-05,
+      "loss": 1.8309,
+      "step": 26490
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.2487820386886597,
+      "learning_rate": 3.3215725806451615e-05,
+      "loss": 1.8492,
+      "step": 26500
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.97657310962677,
+      "learning_rate": 3.319052419354839e-05,
+      "loss": 1.8925,
+      "step": 26510
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.6466784477233887,
+      "learning_rate": 3.3165322580645164e-05,
+      "loss": 1.9144,
+      "step": 26520
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.0966997146606445,
+      "learning_rate": 3.314012096774194e-05,
+      "loss": 1.7435,
+      "step": 26530
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.9285196661949158,
+      "learning_rate": 3.311491935483871e-05,
+      "loss": 1.792,
+      "step": 26540
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.542268991470337,
+      "learning_rate": 3.308971774193548e-05,
+      "loss": 1.7757,
+      "step": 26550
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.177638053894043,
+      "learning_rate": 3.306451612903226e-05,
+      "loss": 1.7823,
+      "step": 26560
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.008866548538208,
+      "learning_rate": 3.3039314516129036e-05,
+      "loss": 1.7578,
+      "step": 26570
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.2122825384140015,
+      "learning_rate": 3.301411290322581e-05,
+      "loss": 2.0831,
+      "step": 26580
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.4399534463882446,
+      "learning_rate": 3.2988911290322584e-05,
+      "loss": 1.9805,
+      "step": 26590
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.3140069246292114,
+      "learning_rate": 3.296370967741936e-05,
+      "loss": 1.8427,
+      "step": 26600
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.4508424997329712,
+      "learning_rate": 3.293850806451613e-05,
+      "loss": 1.8833,
+      "step": 26610
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.4648374319076538,
+      "learning_rate": 3.291330645161291e-05,
+      "loss": 1.7278,
+      "step": 26620
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.1393319368362427,
+      "learning_rate": 3.2888104838709675e-05,
+      "loss": 1.7745,
+      "step": 26630
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 3.2786331176757812,
+      "learning_rate": 3.286290322580645e-05,
+      "loss": 2.07,
+      "step": 26640
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.4883265495300293,
+      "learning_rate": 3.2837701612903224e-05,
+      "loss": 1.7259,
+      "step": 26650
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.2573833465576172,
+      "learning_rate": 3.2812500000000005e-05,
+      "loss": 1.8785,
+      "step": 26660
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.2504042387008667,
+      "learning_rate": 3.278729838709678e-05,
+      "loss": 1.7183,
+      "step": 26670
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.4410721063613892,
+      "learning_rate": 3.2762096774193553e-05,
+      "loss": 1.9818,
+      "step": 26680
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.233393669128418,
+      "learning_rate": 3.273689516129033e-05,
+      "loss": 1.6729,
+      "step": 26690
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.3613576889038086,
+      "learning_rate": 3.2711693548387095e-05,
+      "loss": 1.6609,
+      "step": 26700
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.4628533124923706,
+      "learning_rate": 3.268649193548387e-05,
+      "loss": 1.9442,
+      "step": 26710
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.0166364908218384,
+      "learning_rate": 3.2661290322580644e-05,
+      "loss": 1.7353,
+      "step": 26720
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.9344767332077026,
+      "learning_rate": 3.263608870967742e-05,
+      "loss": 1.7214,
+      "step": 26730
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.2272502183914185,
+      "learning_rate": 3.261088709677419e-05,
+      "loss": 1.8464,
+      "step": 26740
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.7137508988380432,
+      "learning_rate": 3.258568548387097e-05,
+      "loss": 1.8877,
+      "step": 26750
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.2053495645523071,
+      "learning_rate": 3.256048387096775e-05,
+      "loss": 1.7503,
+      "step": 26760
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.9678647518157959,
+      "learning_rate": 3.253528225806452e-05,
+      "loss": 1.6913,
+      "step": 26770
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.208298921585083,
+      "learning_rate": 3.251008064516129e-05,
+      "loss": 1.6906,
+      "step": 26780
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.5452871322631836,
+      "learning_rate": 3.2484879032258064e-05,
+      "loss": 1.757,
+      "step": 26790
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.0387603044509888,
+      "learning_rate": 3.245967741935484e-05,
+      "loss": 1.6109,
+      "step": 26800
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.8556909561157227,
+      "learning_rate": 3.243447580645161e-05,
+      "loss": 1.7401,
+      "step": 26810
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.116698741912842,
+      "learning_rate": 3.240927419354839e-05,
+      "loss": 1.8576,
+      "step": 26820
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.537431240081787,
+      "learning_rate": 3.238407258064516e-05,
+      "loss": 1.7565,
+      "step": 26830
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.6010899543762207,
+      "learning_rate": 3.2358870967741936e-05,
+      "loss": 1.6418,
+      "step": 26840
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.2423553466796875,
+      "learning_rate": 3.233366935483871e-05,
+      "loss": 1.7455,
+      "step": 26850
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.5799401998519897,
+      "learning_rate": 3.2308467741935485e-05,
+      "loss": 1.6917,
+      "step": 26860
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.0207834243774414,
+      "learning_rate": 3.228326612903226e-05,
+      "loss": 1.7211,
+      "step": 26870
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.0866472721099854,
+      "learning_rate": 3.2258064516129034e-05,
+      "loss": 1.8088,
+      "step": 26880
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.1169679164886475,
+      "learning_rate": 3.223286290322581e-05,
+      "loss": 1.6233,
+      "step": 26890
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.0563406944274902,
+      "learning_rate": 3.220766129032258e-05,
+      "loss": 1.7667,
+      "step": 26900
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.9570907354354858,
+      "learning_rate": 3.218245967741936e-05,
+      "loss": 1.7048,
+      "step": 26910
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 3.431791305541992,
+      "learning_rate": 3.215725806451613e-05,
+      "loss": 1.9331,
+      "step": 26920
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.196765661239624,
+      "learning_rate": 3.2132056451612905e-05,
+      "loss": 1.8404,
+      "step": 26930
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.7952790260314941,
+      "learning_rate": 3.210685483870968e-05,
+      "loss": 1.7636,
+      "step": 26940
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.850143551826477,
+      "learning_rate": 3.2081653225806454e-05,
+      "loss": 1.653,
+      "step": 26950
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.0447331666946411,
+      "learning_rate": 3.205645161290323e-05,
+      "loss": 1.8554,
+      "step": 26960
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.2905287742614746,
+      "learning_rate": 3.203125e-05,
+      "loss": 1.8357,
+      "step": 26970
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.0360220670700073,
+      "learning_rate": 3.200604838709678e-05,
+      "loss": 1.8249,
+      "step": 26980
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.1173627376556396,
+      "learning_rate": 3.198084677419355e-05,
+      "loss": 1.9029,
+      "step": 26990
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.3011982440948486,
+      "learning_rate": 3.1955645161290326e-05,
+      "loss": 1.8391,
+      "step": 27000
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.8890591859817505,
+      "learning_rate": 3.193044354838709e-05,
+      "loss": 1.8986,
+      "step": 27010
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.770599365234375,
+      "learning_rate": 3.190524193548387e-05,
+      "loss": 1.7945,
+      "step": 27020
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.4505606889724731,
+      "learning_rate": 3.188004032258065e-05,
+      "loss": 1.6027,
+      "step": 27030
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.7588274478912354,
+      "learning_rate": 3.185483870967742e-05,
+      "loss": 1.7258,
+      "step": 27040
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.2658957242965698,
+      "learning_rate": 3.18296370967742e-05,
+      "loss": 1.707,
+      "step": 27050
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.206875205039978,
+      "learning_rate": 3.180443548387097e-05,
+      "loss": 1.6759,
+      "step": 27060
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.024965763092041,
+      "learning_rate": 3.1779233870967746e-05,
+      "loss": 1.7865,
+      "step": 27070
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.2907451391220093,
+      "learning_rate": 3.175403225806452e-05,
+      "loss": 1.8909,
+      "step": 27080
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.3301091194152832,
+      "learning_rate": 3.172883064516129e-05,
+      "loss": 1.751,
+      "step": 27090
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.7689526081085205,
+      "learning_rate": 3.170362903225806e-05,
+      "loss": 1.8704,
+      "step": 27100
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.5215550661087036,
+      "learning_rate": 3.167842741935484e-05,
+      "loss": 1.787,
+      "step": 27110
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.095498561859131,
+      "learning_rate": 3.165322580645161e-05,
+      "loss": 2.0035,
+      "step": 27120
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.9336097240447998,
+      "learning_rate": 3.162802419354839e-05,
+      "loss": 1.5863,
+      "step": 27130
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.7071185111999512,
+      "learning_rate": 3.1602822580645167e-05,
+      "loss": 1.7359,
+      "step": 27140
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.4316401481628418,
+      "learning_rate": 3.157762096774194e-05,
+      "loss": 1.7888,
+      "step": 27150
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.1057945489883423,
+      "learning_rate": 3.1552419354838715e-05,
+      "loss": 1.5107,
+      "step": 27160
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.1223514080047607,
+      "learning_rate": 3.152721774193548e-05,
+      "loss": 1.6722,
+      "step": 27170
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.1780390739440918,
+      "learning_rate": 3.150201612903226e-05,
+      "loss": 1.7603,
+      "step": 27180
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.1946700811386108,
+      "learning_rate": 3.147681451612903e-05,
+      "loss": 1.7712,
+      "step": 27190
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.8809229135513306,
+      "learning_rate": 3.1451612903225806e-05,
+      "loss": 1.7759,
+      "step": 27200
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.1302640438079834,
+      "learning_rate": 3.142641129032258e-05,
+      "loss": 1.7665,
+      "step": 27210
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.2791681289672852,
+      "learning_rate": 3.1401209677419355e-05,
+      "loss": 1.7273,
+      "step": 27220
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.4046205282211304,
+      "learning_rate": 3.1376008064516136e-05,
+      "loss": 2.0015,
+      "step": 27230
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.8611924648284912,
+      "learning_rate": 3.135080645161291e-05,
+      "loss": 1.9365,
+      "step": 27240
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.8917639851570129,
+      "learning_rate": 3.132560483870968e-05,
+      "loss": 1.6942,
+      "step": 27250
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.397609829902649,
+      "learning_rate": 3.130040322580645e-05,
+      "loss": 1.6579,
+      "step": 27260
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.9144561290740967,
+      "learning_rate": 3.1275201612903226e-05,
+      "loss": 1.7653,
+      "step": 27270
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.0120042562484741,
+      "learning_rate": 3.125e-05,
+      "loss": 1.846,
+      "step": 27280
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.978792130947113,
+      "learning_rate": 3.1224798387096775e-05,
+      "loss": 1.6844,
+      "step": 27290
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.2573938369750977,
+      "learning_rate": 3.119959677419355e-05,
+      "loss": 1.7835,
+      "step": 27300
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.9943901300430298,
+      "learning_rate": 3.1174395161290324e-05,
+      "loss": 2.0236,
+      "step": 27310
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.7839648723602295,
+      "learning_rate": 3.11491935483871e-05,
+      "loss": 1.8911,
+      "step": 27320
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.8152753114700317,
+      "learning_rate": 3.112399193548387e-05,
+      "loss": 1.7229,
+      "step": 27330
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.7408503293991089,
+      "learning_rate": 3.109879032258065e-05,
+      "loss": 1.8671,
+      "step": 27340
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.0653412342071533,
+      "learning_rate": 3.107358870967742e-05,
+      "loss": 1.7777,
+      "step": 27350
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.629546046257019,
+      "learning_rate": 3.1048387096774195e-05,
+      "loss": 1.7458,
+      "step": 27360
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.9189653396606445,
+      "learning_rate": 3.102318548387097e-05,
+      "loss": 1.7836,
+      "step": 27370
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.156329870223999,
+      "learning_rate": 3.0997983870967744e-05,
+      "loss": 1.7966,
+      "step": 27380
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.2499827146530151,
+      "learning_rate": 3.097278225806452e-05,
+      "loss": 1.7484,
+      "step": 27390
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.88551926612854,
+      "learning_rate": 3.0947580645161286e-05,
+      "loss": 1.9634,
+      "step": 27400
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.0629154443740845,
+      "learning_rate": 3.092237903225807e-05,
+      "loss": 1.7551,
+      "step": 27410
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.0696663856506348,
+      "learning_rate": 3.089717741935484e-05,
+      "loss": 1.823,
+      "step": 27420
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.3341882228851318,
+      "learning_rate": 3.0871975806451616e-05,
+      "loss": 1.6385,
+      "step": 27430
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.5551402568817139,
+      "learning_rate": 3.084677419354839e-05,
+      "loss": 1.7386,
+      "step": 27440
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.7226052284240723,
+      "learning_rate": 3.0821572580645165e-05,
+      "loss": 1.7819,
+      "step": 27450
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.0524976253509521,
+      "learning_rate": 3.079637096774194e-05,
+      "loss": 1.8717,
+      "step": 27460
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.0830774307250977,
+      "learning_rate": 3.077116935483871e-05,
+      "loss": 1.7027,
+      "step": 27470
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.8074485063552856,
+      "learning_rate": 3.074596774193548e-05,
+      "loss": 1.652,
+      "step": 27480
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.5679967403411865,
+      "learning_rate": 3.0720766129032255e-05,
+      "loss": 1.7445,
+      "step": 27490
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.1872891187667847,
+      "learning_rate": 3.069556451612903e-05,
+      "loss": 1.9603,
+      "step": 27500
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.5066791772842407,
+      "learning_rate": 3.067036290322581e-05,
+      "loss": 1.9608,
+      "step": 27510
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.071422815322876,
+      "learning_rate": 3.0645161290322585e-05,
+      "loss": 1.611,
+      "step": 27520
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.4530051946640015,
+      "learning_rate": 3.061995967741936e-05,
+      "loss": 1.6878,
+      "step": 27530
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.4470046758651733,
+      "learning_rate": 3.0594758064516134e-05,
+      "loss": 1.5986,
+      "step": 27540
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.3154253959655762,
+      "learning_rate": 3.056955645161291e-05,
+      "loss": 1.6822,
+      "step": 27550
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.791657567024231,
+      "learning_rate": 3.0544354838709676e-05,
+      "loss": 1.6595,
+      "step": 27560
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.596862554550171,
+      "learning_rate": 3.051915322580645e-05,
+      "loss": 1.599,
+      "step": 27570
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.4632937908172607,
+      "learning_rate": 3.0493951612903228e-05,
+      "loss": 1.6398,
+      "step": 27580
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.2962050437927246,
+      "learning_rate": 3.0468750000000002e-05,
+      "loss": 1.8606,
+      "step": 27590
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.9040840864181519,
+      "learning_rate": 3.0443548387096776e-05,
+      "loss": 1.5743,
+      "step": 27600
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.4428074359893799,
+      "learning_rate": 3.041834677419355e-05,
+      "loss": 1.8551,
+      "step": 27610
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.573683261871338,
+      "learning_rate": 3.0393145161290325e-05,
+      "loss": 1.8401,
+      "step": 27620
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.8869794607162476,
+      "learning_rate": 3.0367943548387096e-05,
+      "loss": 1.6857,
+      "step": 27630
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.9982945322990417,
+      "learning_rate": 3.034274193548387e-05,
+      "loss": 1.7623,
+      "step": 27640
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.1649187803268433,
+      "learning_rate": 3.0317540322580645e-05,
+      "loss": 1.811,
+      "step": 27650
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.2290154695510864,
+      "learning_rate": 3.029233870967742e-05,
+      "loss": 1.8482,
+      "step": 27660
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.5044114589691162,
+      "learning_rate": 3.0267137096774197e-05,
+      "loss": 1.7758,
+      "step": 27670
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.1660552024841309,
+      "learning_rate": 3.024193548387097e-05,
+      "loss": 1.8304,
+      "step": 27680
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.4135676622390747,
+      "learning_rate": 3.0216733870967745e-05,
+      "loss": 1.8168,
+      "step": 27690
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.405389666557312,
+      "learning_rate": 3.019153225806452e-05,
+      "loss": 1.7933,
+      "step": 27700
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.9517889618873596,
+      "learning_rate": 3.016633064516129e-05,
+      "loss": 1.7528,
+      "step": 27710
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.837257742881775,
+      "learning_rate": 3.0141129032258065e-05,
+      "loss": 1.8659,
+      "step": 27720
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.5208773612976074,
+      "learning_rate": 3.011592741935484e-05,
+      "loss": 1.8426,
+      "step": 27730
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.4117851257324219,
+      "learning_rate": 3.0090725806451614e-05,
+      "loss": 2.0161,
+      "step": 27740
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.3954097032546997,
+      "learning_rate": 3.0065524193548388e-05,
+      "loss": 1.6908,
+      "step": 27750
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.3990389108657837,
+      "learning_rate": 3.0040322580645162e-05,
+      "loss": 1.8598,
+      "step": 27760
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.394239902496338,
+      "learning_rate": 3.001512096774194e-05,
+      "loss": 1.7876,
+      "step": 27770
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.2093029022216797,
+      "learning_rate": 2.9989919354838715e-05,
+      "loss": 1.8345,
+      "step": 27780
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.1877237558364868,
+      "learning_rate": 2.9964717741935482e-05,
+      "loss": 1.9139,
+      "step": 27790
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.2244038581848145,
+      "learning_rate": 2.9939516129032256e-05,
+      "loss": 1.8883,
+      "step": 27800
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.4611999988555908,
+      "learning_rate": 2.9914314516129034e-05,
+      "loss": 1.8408,
+      "step": 27810
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.7915371656417847,
+      "learning_rate": 2.988911290322581e-05,
+      "loss": 1.7857,
+      "step": 27820
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.1456645727157593,
+      "learning_rate": 2.9863911290322583e-05,
+      "loss": 1.8177,
+      "step": 27830
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.533869981765747,
+      "learning_rate": 2.9838709677419357e-05,
+      "loss": 1.8572,
+      "step": 27840
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.3246248960494995,
+      "learning_rate": 2.981350806451613e-05,
+      "loss": 1.8138,
+      "step": 27850
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.0090954303741455,
+      "learning_rate": 2.9788306451612906e-05,
+      "loss": 1.8305,
+      "step": 27860
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.7615946531295776,
+      "learning_rate": 2.9763104838709677e-05,
+      "loss": 1.6368,
+      "step": 27870
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 3.3230857849121094,
+      "learning_rate": 2.973790322580645e-05,
+      "loss": 1.7077,
+      "step": 27880
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.1097389459609985,
+      "learning_rate": 2.9712701612903226e-05,
+      "loss": 1.913,
+      "step": 27890
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.1314055919647217,
+      "learning_rate": 2.96875e-05,
+      "loss": 1.7013,
+      "step": 27900
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.0251271724700928,
+      "learning_rate": 2.9662298387096778e-05,
+      "loss": 1.7482,
+      "step": 27910
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.4378145933151245,
+      "learning_rate": 2.9637096774193552e-05,
+      "loss": 1.8368,
+      "step": 27920
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.664634644985199,
+      "learning_rate": 2.9611895161290326e-05,
+      "loss": 1.7255,
+      "step": 27930
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.1108479499816895,
+      "learning_rate": 2.9586693548387094e-05,
+      "loss": 1.8203,
+      "step": 27940
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.8425639867782593,
+      "learning_rate": 2.956149193548387e-05,
+      "loss": 1.6541,
+      "step": 27950
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.9990367889404297,
+      "learning_rate": 2.9536290322580646e-05,
+      "loss": 1.7182,
+      "step": 27960
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.9387206435203552,
+      "learning_rate": 2.951108870967742e-05,
+      "loss": 1.7173,
+      "step": 27970
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.4695771932601929,
+      "learning_rate": 2.9485887096774195e-05,
+      "loss": 1.8964,
+      "step": 27980
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.391408085823059,
+      "learning_rate": 2.946068548387097e-05,
+      "loss": 1.7718,
+      "step": 27990
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.9452804327011108,
+      "learning_rate": 2.9435483870967743e-05,
+      "loss": 1.6544,
+      "step": 28000
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.8510650396347046,
+      "learning_rate": 2.941028225806452e-05,
+      "loss": 1.7816,
+      "step": 28010
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.028639554977417,
+      "learning_rate": 2.938508064516129e-05,
+      "loss": 1.8468,
+      "step": 28020
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.1311569213867188,
+      "learning_rate": 2.9359879032258063e-05,
+      "loss": 1.9056,
+      "step": 28030
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.8045054078102112,
+      "learning_rate": 2.9334677419354837e-05,
+      "loss": 1.7326,
+      "step": 28040
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.5639100074768066,
+      "learning_rate": 2.9309475806451615e-05,
+      "loss": 1.83,
+      "step": 28050
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.3804320096969604,
+      "learning_rate": 2.928427419354839e-05,
+      "loss": 1.7864,
+      "step": 28060
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.3837063312530518,
+      "learning_rate": 2.9259072580645164e-05,
+      "loss": 1.6268,
+      "step": 28070
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.7251735925674438,
+      "learning_rate": 2.9233870967741938e-05,
+      "loss": 2.0786,
+      "step": 28080
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.6921919584274292,
+      "learning_rate": 2.9208669354838713e-05,
+      "loss": 1.7602,
+      "step": 28090
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.7074424624443054,
+      "learning_rate": 2.9183467741935483e-05,
+      "loss": 1.7005,
+      "step": 28100
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.354042410850525,
+      "learning_rate": 2.9158266129032258e-05,
+      "loss": 1.7692,
+      "step": 28110
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.8737571835517883,
+      "learning_rate": 2.9133064516129032e-05,
+      "loss": 1.6822,
+      "step": 28120
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.1983879804611206,
+      "learning_rate": 2.9107862903225807e-05,
+      "loss": 1.7112,
+      "step": 28130
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.0532536506652832,
+      "learning_rate": 2.9082661290322584e-05,
+      "loss": 1.6649,
+      "step": 28140
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.080644965171814,
+      "learning_rate": 2.905745967741936e-05,
+      "loss": 1.9076,
+      "step": 28150
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.146431803703308,
+      "learning_rate": 2.9032258064516133e-05,
+      "loss": 1.8788,
+      "step": 28160
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.5423426628112793,
+      "learning_rate": 2.9007056451612907e-05,
+      "loss": 1.7877,
+      "step": 28170
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.295678734779358,
+      "learning_rate": 2.8981854838709678e-05,
+      "loss": 1.7269,
+      "step": 28180
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.0878565311431885,
+      "learning_rate": 2.8956653225806453e-05,
+      "loss": 1.5498,
+      "step": 28190
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.9327211380004883,
+      "learning_rate": 2.8931451612903227e-05,
+      "loss": 1.8481,
+      "step": 28200
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.3493373394012451,
+      "learning_rate": 2.890625e-05,
+      "loss": 1.7402,
+      "step": 28210
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.9438363313674927,
+      "learning_rate": 2.8881048387096776e-05,
+      "loss": 1.924,
+      "step": 28220
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.1412674188613892,
+      "learning_rate": 2.885584677419355e-05,
+      "loss": 1.863,
+      "step": 28230
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.9761775135993958,
+      "learning_rate": 2.8830645161290328e-05,
+      "loss": 1.6958,
+      "step": 28240
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.016577124595642,
+      "learning_rate": 2.8805443548387095e-05,
+      "loss": 1.631,
+      "step": 28250
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.4962575435638428,
+      "learning_rate": 2.878024193548387e-05,
+      "loss": 1.6391,
+      "step": 28260
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.8270187973976135,
+      "learning_rate": 2.8755040322580644e-05,
+      "loss": 1.8347,
+      "step": 28270
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 3.2643206119537354,
+      "learning_rate": 2.8729838709677422e-05,
+      "loss": 1.8022,
+      "step": 28280
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.293227195739746,
+      "learning_rate": 2.8704637096774196e-05,
+      "loss": 1.6105,
+      "step": 28290
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.888077437877655,
+      "learning_rate": 2.867943548387097e-05,
+      "loss": 1.866,
+      "step": 28300
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.7911893129348755,
+      "learning_rate": 2.8654233870967745e-05,
+      "loss": 1.7414,
+      "step": 28310
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.117730736732483,
+      "learning_rate": 2.862903225806452e-05,
+      "loss": 1.7338,
+      "step": 28320
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.3954064846038818,
+      "learning_rate": 2.860383064516129e-05,
+      "loss": 1.907,
+      "step": 28330
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.6644837856292725,
+      "learning_rate": 2.8578629032258064e-05,
+      "loss": 1.7824,
+      "step": 28340
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.6904391050338745,
+      "learning_rate": 2.855342741935484e-05,
+      "loss": 1.8102,
+      "step": 28350
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.2786898612976074,
+      "learning_rate": 2.8528225806451613e-05,
+      "loss": 1.7649,
+      "step": 28360
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 4.081326484680176,
+      "learning_rate": 2.8503024193548387e-05,
+      "loss": 1.7142,
+      "step": 28370
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.3490408658981323,
+      "learning_rate": 2.8477822580645165e-05,
+      "loss": 1.833,
+      "step": 28380
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.9102471470832825,
+      "learning_rate": 2.845262096774194e-05,
+      "loss": 1.8564,
+      "step": 28390
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.9845308065414429,
+      "learning_rate": 2.8427419354838714e-05,
+      "loss": 1.6139,
+      "step": 28400
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.8394770622253418,
+      "learning_rate": 2.840221774193548e-05,
+      "loss": 1.7017,
+      "step": 28410
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.6664412021636963,
+      "learning_rate": 2.837701612903226e-05,
+      "loss": 1.9039,
+      "step": 28420
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.3823177814483643,
+      "learning_rate": 2.8351814516129034e-05,
+      "loss": 1.7107,
+      "step": 28430
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.129533052444458,
+      "learning_rate": 2.8326612903225808e-05,
+      "loss": 1.9236,
+      "step": 28440
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.587670087814331,
+      "learning_rate": 2.8301411290322582e-05,
+      "loss": 1.8341,
+      "step": 28450
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.394681453704834,
+      "learning_rate": 2.8276209677419357e-05,
+      "loss": 1.822,
+      "step": 28460
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.8474830389022827,
+      "learning_rate": 2.825100806451613e-05,
+      "loss": 1.8145,
+      "step": 28470
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.936037302017212,
+      "learning_rate": 2.822580645161291e-05,
+      "loss": 1.6282,
+      "step": 28480
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.1051541566848755,
+      "learning_rate": 2.8200604838709676e-05,
+      "loss": 1.7859,
+      "step": 28490
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.7739005088806152,
+      "learning_rate": 2.817540322580645e-05,
+      "loss": 1.7823,
+      "step": 28500
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.4614070653915405,
+      "learning_rate": 2.8150201612903225e-05,
+      "loss": 1.7635,
+      "step": 28510
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.4348517656326294,
+      "learning_rate": 2.8125000000000003e-05,
+      "loss": 1.5945,
+      "step": 28520
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.8624076843261719,
+      "learning_rate": 2.8099798387096777e-05,
+      "loss": 1.4324,
+      "step": 28530
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.9341122508049011,
+      "learning_rate": 2.807459677419355e-05,
+      "loss": 1.7216,
+      "step": 28540
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.6537833213806152,
+      "learning_rate": 2.8049395161290326e-05,
+      "loss": 1.6303,
+      "step": 28550
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.9114584922790527,
+      "learning_rate": 2.8024193548387097e-05,
+      "loss": 1.7906,
+      "step": 28560
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.145323395729065,
+      "learning_rate": 2.799899193548387e-05,
+      "loss": 1.8472,
+      "step": 28570
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.2878048419952393,
+      "learning_rate": 2.7973790322580645e-05,
+      "loss": 1.877,
+      "step": 28580
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.4785714149475098,
+      "learning_rate": 2.794858870967742e-05,
+      "loss": 1.7035,
+      "step": 28590
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.6721456050872803,
+      "learning_rate": 2.7923387096774194e-05,
+      "loss": 1.6969,
+      "step": 28600
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.3887747526168823,
+      "learning_rate": 2.789818548387097e-05,
+      "loss": 1.6973,
+      "step": 28610
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.5721025466918945,
+      "learning_rate": 2.7872983870967746e-05,
+      "loss": 1.8714,
+      "step": 28620
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.0347737073898315,
+      "learning_rate": 2.784778225806452e-05,
+      "loss": 1.6997,
+      "step": 28630
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.8540313243865967,
+      "learning_rate": 2.7822580645161288e-05,
+      "loss": 1.7656,
+      "step": 28640
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 3.9794232845306396,
+      "learning_rate": 2.7797379032258066e-05,
+      "loss": 1.849,
+      "step": 28650
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.446833610534668,
+      "learning_rate": 2.777217741935484e-05,
+      "loss": 1.9183,
+      "step": 28660
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.9081591367721558,
+      "learning_rate": 2.7746975806451614e-05,
+      "loss": 1.829,
+      "step": 28670
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.4537127017974854,
+      "learning_rate": 2.772177419354839e-05,
+      "loss": 1.7112,
+      "step": 28680
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.5715779066085815,
+      "learning_rate": 2.7696572580645163e-05,
+      "loss": 1.7313,
+      "step": 28690
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.2513885498046875,
+      "learning_rate": 2.7671370967741937e-05,
+      "loss": 1.647,
+      "step": 28700
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.5850483179092407,
+      "learning_rate": 2.7646169354838715e-05,
+      "loss": 1.6819,
+      "step": 28710
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.6385022401809692,
+      "learning_rate": 2.7620967741935483e-05,
+      "loss": 1.6606,
+      "step": 28720
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.0912301540374756,
+      "learning_rate": 2.7595766129032257e-05,
+      "loss": 1.3948,
+      "step": 28730
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.202683448791504,
+      "learning_rate": 2.757056451612903e-05,
+      "loss": 1.9251,
+      "step": 28740
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.2097370624542236,
+      "learning_rate": 2.754536290322581e-05,
+      "loss": 1.8344,
+      "step": 28750
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.8121129274368286,
+      "learning_rate": 2.7520161290322584e-05,
+      "loss": 1.7369,
+      "step": 28760
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.7184251546859741,
+      "learning_rate": 2.7494959677419358e-05,
+      "loss": 1.8703,
+      "step": 28770
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.3328438997268677,
+      "learning_rate": 2.7469758064516132e-05,
+      "loss": 1.7116,
+      "step": 28780
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.2027907371520996,
+      "learning_rate": 2.7444556451612907e-05,
+      "loss": 1.7865,
+      "step": 28790
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.837236166000366,
+      "learning_rate": 2.7419354838709678e-05,
+      "loss": 1.8926,
+      "step": 28800
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.8856102824211121,
+      "learning_rate": 2.7394153225806452e-05,
+      "loss": 1.8618,
+      "step": 28810
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.1288758516311646,
+      "learning_rate": 2.7368951612903226e-05,
+      "loss": 1.7197,
+      "step": 28820
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.3666694164276123,
+      "learning_rate": 2.734375e-05,
+      "loss": 1.945,
+      "step": 28830
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.378069519996643,
+      "learning_rate": 2.7318548387096775e-05,
+      "loss": 1.8274,
+      "step": 28840
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.9071300625801086,
+      "learning_rate": 2.7293346774193553e-05,
+      "loss": 1.651,
+      "step": 28850
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.372689962387085,
+      "learning_rate": 2.7268145161290327e-05,
+      "loss": 1.84,
+      "step": 28860
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.4825412034988403,
+      "learning_rate": 2.7242943548387095e-05,
+      "loss": 1.9243,
+      "step": 28870
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.9763379096984863,
+      "learning_rate": 2.721774193548387e-05,
+      "loss": 1.8587,
+      "step": 28880
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.3694992065429688,
+      "learning_rate": 2.7192540322580647e-05,
+      "loss": 1.7068,
+      "step": 28890
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.12813401222229,
+      "learning_rate": 2.716733870967742e-05,
+      "loss": 1.7817,
+      "step": 28900
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.3228737115859985,
+      "learning_rate": 2.7142137096774195e-05,
+      "loss": 1.8723,
+      "step": 28910
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.2593836784362793,
+      "learning_rate": 2.711693548387097e-05,
+      "loss": 1.8646,
+      "step": 28920
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.7854911088943481,
+      "learning_rate": 2.7091733870967744e-05,
+      "loss": 1.6762,
+      "step": 28930
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.124541997909546,
+      "learning_rate": 2.706653225806452e-05,
+      "loss": 1.777,
+      "step": 28940
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.2461960315704346,
+      "learning_rate": 2.704133064516129e-05,
+      "loss": 1.7292,
+      "step": 28950
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.7227761149406433,
+      "learning_rate": 2.7016129032258064e-05,
+      "loss": 1.6879,
+      "step": 28960
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.5558578968048096,
+      "learning_rate": 2.6990927419354838e-05,
+      "loss": 1.8019,
+      "step": 28970
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.6402218341827393,
+      "learning_rate": 2.6965725806451612e-05,
+      "loss": 1.5802,
+      "step": 28980
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.6904274225234985,
+      "learning_rate": 2.694052419354839e-05,
+      "loss": 1.8264,
+      "step": 28990
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.5569896697998047,
+      "learning_rate": 2.6915322580645164e-05,
+      "loss": 1.9027,
+      "step": 29000
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.8624861836433411,
+      "learning_rate": 2.689012096774194e-05,
+      "loss": 1.797,
+      "step": 29010
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.1015255451202393,
+      "learning_rate": 2.6864919354838713e-05,
+      "loss": 1.8417,
+      "step": 29020
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.8417332172393799,
+      "learning_rate": 2.6839717741935484e-05,
+      "loss": 1.6658,
+      "step": 29030
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.463016152381897,
+      "learning_rate": 2.681451612903226e-05,
+      "loss": 1.8256,
+      "step": 29040
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.1427581310272217,
+      "learning_rate": 2.6789314516129033e-05,
+      "loss": 1.9808,
+      "step": 29050
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.692335605621338,
+      "learning_rate": 2.6764112903225807e-05,
+      "loss": 1.7964,
+      "step": 29060
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.1608142852783203,
+      "learning_rate": 2.673891129032258e-05,
+      "loss": 1.9109,
+      "step": 29070
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.7788482904434204,
+      "learning_rate": 2.6713709677419356e-05,
+      "loss": 1.8701,
+      "step": 29080
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 2.22286057472229,
+      "learning_rate": 2.6688508064516134e-05,
+      "loss": 1.8189,
+      "step": 29090
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.2713571786880493,
+      "learning_rate": 2.6663306451612908e-05,
+      "loss": 1.8134,
+      "step": 29100
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.7466511726379395,
+      "learning_rate": 2.6638104838709675e-05,
+      "loss": 1.5,
+      "step": 29110
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.3681955337524414,
+      "learning_rate": 2.661290322580645e-05,
+      "loss": 1.6816,
+      "step": 29120
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.978603482246399,
+      "learning_rate": 2.6587701612903228e-05,
+      "loss": 1.7352,
+      "step": 29130
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.2236359119415283,
+      "learning_rate": 2.6562500000000002e-05,
+      "loss": 1.7296,
+      "step": 29140
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.3428125381469727,
+      "learning_rate": 2.6537298387096776e-05,
+      "loss": 1.6437,
+      "step": 29150
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.2325800657272339,
+      "learning_rate": 2.651209677419355e-05,
+      "loss": 1.7732,
+      "step": 29160
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.7017361521720886,
+      "learning_rate": 2.6486895161290325e-05,
+      "loss": 1.6706,
+      "step": 29170
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.8391928672790527,
+      "learning_rate": 2.6461693548387096e-05,
+      "loss": 1.7147,
+      "step": 29180
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.0671457052230835,
+      "learning_rate": 2.643649193548387e-05,
+      "loss": 1.7134,
+      "step": 29190
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 2.0345959663391113,
+      "learning_rate": 2.6411290322580645e-05,
+      "loss": 1.8659,
+      "step": 29200
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.2739362716674805,
+      "learning_rate": 2.638608870967742e-05,
+      "loss": 1.7599,
+      "step": 29210
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.740612268447876,
+      "learning_rate": 2.6360887096774197e-05,
+      "loss": 1.6106,
+      "step": 29220
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.721281886100769,
+      "learning_rate": 2.633568548387097e-05,
+      "loss": 1.9356,
+      "step": 29230
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.93117356300354,
+      "learning_rate": 2.6310483870967745e-05,
+      "loss": 1.6965,
+      "step": 29240
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 3.5222253799438477,
+      "learning_rate": 2.628528225806452e-05,
+      "loss": 1.9432,
+      "step": 29250
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.9055311679840088,
+      "learning_rate": 2.626008064516129e-05,
+      "loss": 1.9207,
+      "step": 29260
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 1.6068949699401855,
+      "learning_rate": 2.6234879032258065e-05,
+      "loss": 1.7187,
+      "step": 29270
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 1.5461864471435547,
+      "learning_rate": 2.620967741935484e-05,
+      "loss": 1.8784,
+      "step": 29280
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 1.373337745666504,
+      "learning_rate": 2.6184475806451614e-05,
+      "loss": 1.9086,
+      "step": 29290
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 1.6942358016967773,
+      "learning_rate": 2.6159274193548388e-05,
+      "loss": 1.8365,
+      "step": 29300
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 1.2653268575668335,
+      "learning_rate": 2.6134072580645162e-05,
+      "loss": 1.6643,
+      "step": 29310
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 2.283572196960449,
+      "learning_rate": 2.610887096774194e-05,
+      "loss": 1.82,
+      "step": 29320
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 1.8657546043395996,
+      "learning_rate": 2.6083669354838714e-05,
+      "loss": 1.8042,
+      "step": 29330
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 2.9319570064544678,
+      "learning_rate": 2.6058467741935482e-05,
+      "loss": 1.8681,
+      "step": 29340
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 6.1617960929870605,
+      "learning_rate": 2.6033266129032256e-05,
+      "loss": 1.8889,
+      "step": 29350
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 3.039808750152588,
+      "learning_rate": 2.6008064516129034e-05,
+      "loss": 1.8449,
+      "step": 29360
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 2.03147554397583,
+      "learning_rate": 2.598286290322581e-05,
+      "loss": 1.8861,
+      "step": 29370
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 1.5385100841522217,
+      "learning_rate": 2.5957661290322583e-05,
+      "loss": 1.7036,
+      "step": 29380
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.89244544506073,
+      "learning_rate": 2.5932459677419357e-05,
+      "loss": 1.8674,
+      "step": 29390
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 1.1933832168579102,
+      "learning_rate": 2.590725806451613e-05,
+      "loss": 1.6838,
+      "step": 29400
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.7277722954750061,
+      "learning_rate": 2.5882056451612906e-05,
+      "loss": 1.9785,
+      "step": 29410
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 1.0044827461242676,
+      "learning_rate": 2.5856854838709677e-05,
+      "loss": 1.9674,
+      "step": 29420
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 1.2843390703201294,
+      "learning_rate": 2.583165322580645e-05,
+      "loss": 1.7596,
+      "step": 29430
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 1.2331796884536743,
+      "learning_rate": 2.5806451612903226e-05,
+      "loss": 1.8358,
+      "step": 29440
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 1.5178419351577759,
+      "learning_rate": 2.578125e-05,
+      "loss": 1.8067,
+      "step": 29450
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.8745360374450684,
+      "learning_rate": 2.5756048387096778e-05,
+      "loss": 1.8569,
+      "step": 29460
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 1.2188432216644287,
+      "learning_rate": 2.5730846774193552e-05,
+      "loss": 1.7892,
+      "step": 29470
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 2.5472302436828613,
+      "learning_rate": 2.5705645161290326e-05,
+      "loss": 1.7434,
+      "step": 29480
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 1.4197381734848022,
+      "learning_rate": 2.5680443548387094e-05,
+      "loss": 1.8394,
+      "step": 29490
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.8928569555282593,
+      "learning_rate": 2.565524193548387e-05,
+      "loss": 1.7175,
+      "step": 29500
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 1.2455778121948242,
+      "learning_rate": 2.5630040322580646e-05,
+      "loss": 1.5214,
+      "step": 29510
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 3.218074321746826,
+      "learning_rate": 2.560483870967742e-05,
+      "loss": 1.9473,
+      "step": 29520
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 1.041317343711853,
+      "learning_rate": 2.5579637096774195e-05,
+      "loss": 1.7605,
+      "step": 29530
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 1.3570218086242676,
+      "learning_rate": 2.555443548387097e-05,
+      "loss": 1.8391,
+      "step": 29540
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 1.4219915866851807,
+      "learning_rate": 2.5529233870967743e-05,
+      "loss": 1.6407,
+      "step": 29550
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 2.722759246826172,
+      "learning_rate": 2.550403225806452e-05,
+      "loss": 1.8412,
+      "step": 29560
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5528116822242737,
+      "learning_rate": 2.547883064516129e-05,
+      "loss": 1.8762,
+      "step": 29570
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 1.0652337074279785,
+      "learning_rate": 2.5453629032258063e-05,
+      "loss": 1.8231,
+      "step": 29580
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 1.544162392616272,
+      "learning_rate": 2.5428427419354837e-05,
+      "loss": 1.7724,
+      "step": 29590
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 1.5414986610412598,
+      "learning_rate": 2.5403225806451615e-05,
+      "loss": 1.8169,
+      "step": 29600
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 1.4007205963134766,
+      "learning_rate": 2.537802419354839e-05,
+      "loss": 1.8135,
+      "step": 29610
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 2.605116128921509,
+      "learning_rate": 2.5352822580645164e-05,
+      "loss": 1.896,
+      "step": 29620
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.9728342890739441,
+      "learning_rate": 2.5327620967741938e-05,
+      "loss": 1.7388,
+      "step": 29630
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 1.098065972328186,
+      "learning_rate": 2.5302419354838712e-05,
+      "loss": 1.6308,
+      "step": 29640
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.8889781832695007,
+      "learning_rate": 2.5277217741935483e-05,
+      "loss": 1.756,
+      "step": 29650
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 1.6777898073196411,
+      "learning_rate": 2.5252016129032258e-05,
+      "loss": 1.8072,
+      "step": 29660
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 1.236086368560791,
+      "learning_rate": 2.5226814516129032e-05,
+      "loss": 1.7114,
+      "step": 29670
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 1.1139485836029053,
+      "learning_rate": 2.5201612903225806e-05,
+      "loss": 1.5582,
+      "step": 29680
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 1.50374436378479,
+      "learning_rate": 2.517641129032258e-05,
+      "loss": 1.8207,
+      "step": 29690
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 1.2551511526107788,
+      "learning_rate": 2.515120967741936e-05,
+      "loss": 1.8361,
+      "step": 29700
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 1.2329485416412354,
+      "learning_rate": 2.5126008064516133e-05,
+      "loss": 1.8763,
+      "step": 29710
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 1.9638373851776123,
+      "learning_rate": 2.5100806451612907e-05,
+      "loss": 1.8775,
+      "step": 29720
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 1.7881748676300049,
+      "learning_rate": 2.5075604838709678e-05,
+      "loss": 1.6894,
+      "step": 29730
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.9923173785209656,
+      "learning_rate": 2.5050403225806453e-05,
+      "loss": 1.6645,
+      "step": 29740
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 1.5536866188049316,
+      "learning_rate": 2.5025201612903227e-05,
+      "loss": 1.7581,
+      "step": 29750
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 1.3249049186706543,
+      "learning_rate": 2.5e-05,
+      "loss": 1.819,
+      "step": 29760
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.8212242722511292,
+      "learning_rate": 2.4974798387096776e-05,
+      "loss": 1.7382,
+      "step": 29770
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 1.3295927047729492,
+      "learning_rate": 2.494959677419355e-05,
+      "loss": 1.7417,
+      "step": 29780
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.7498393058776855,
+      "learning_rate": 2.4924395161290324e-05,
+      "loss": 1.7406,
+      "step": 29790
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 2.3571298122406006,
+      "learning_rate": 2.48991935483871e-05,
+      "loss": 1.7756,
+      "step": 29800
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.7969825267791748,
+      "learning_rate": 2.4873991935483873e-05,
+      "loss": 1.7528,
+      "step": 29810
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.8778737187385559,
+      "learning_rate": 2.4848790322580644e-05,
+      "loss": 1.7791,
+      "step": 29820
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 1.5012801885604858,
+      "learning_rate": 2.482358870967742e-05,
+      "loss": 1.734,
+      "step": 29830
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 1.0221960544586182,
+      "learning_rate": 2.4798387096774196e-05,
+      "loss": 1.8489,
+      "step": 29840
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 1.9005531072616577,
+      "learning_rate": 2.477318548387097e-05,
+      "loss": 1.7799,
+      "step": 29850
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.9006466269493103,
+      "learning_rate": 2.474798387096774e-05,
+      "loss": 1.6815,
+      "step": 29860
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 1.7919285297393799,
+      "learning_rate": 2.4722782258064516e-05,
+      "loss": 1.8068,
+      "step": 29870
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 3.581774950027466,
+      "learning_rate": 2.4697580645161293e-05,
+      "loss": 1.686,
+      "step": 29880
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 1.8248499631881714,
+      "learning_rate": 2.4672379032258068e-05,
+      "loss": 1.8681,
+      "step": 29890
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 1.1161112785339355,
+      "learning_rate": 2.464717741935484e-05,
+      "loss": 1.6673,
+      "step": 29900
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 2.06722092628479,
+      "learning_rate": 2.4621975806451613e-05,
+      "loss": 1.8195,
+      "step": 29910
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 1.9002989530563354,
+      "learning_rate": 2.4596774193548387e-05,
+      "loss": 1.7882,
+      "step": 29920
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 1.2762290239334106,
+      "learning_rate": 2.4571572580645165e-05,
+      "loss": 1.7215,
+      "step": 29930
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 1.5339686870574951,
+      "learning_rate": 2.4546370967741936e-05,
+      "loss": 1.7406,
+      "step": 29940
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 2.76641583442688,
+      "learning_rate": 2.452116935483871e-05,
+      "loss": 2.0802,
+      "step": 29950
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 4.099884033203125,
+      "learning_rate": 2.4495967741935485e-05,
+      "loss": 1.9128,
+      "step": 29960
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 2.3815672397613525,
+      "learning_rate": 2.447076612903226e-05,
+      "loss": 1.9029,
+      "step": 29970
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 1.7372068166732788,
+      "learning_rate": 2.4445564516129033e-05,
+      "loss": 1.7885,
+      "step": 29980
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 1.4166446924209595,
+      "learning_rate": 2.4420362903225808e-05,
+      "loss": 1.6036,
+      "step": 29990
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 1.0003135204315186,
+      "learning_rate": 2.4395161290322582e-05,
+      "loss": 1.9594,
+      "step": 30000
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 2.332258939743042,
+      "learning_rate": 2.4369959677419356e-05,
+      "loss": 1.8064,
+      "step": 30010
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 1.1318824291229248,
+      "learning_rate": 2.434475806451613e-05,
+      "loss": 1.8352,
+      "step": 30020
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.8694331049919128,
+      "learning_rate": 2.4319556451612905e-05,
+      "loss": 1.9357,
+      "step": 30030
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.7871884107589722,
+      "learning_rate": 2.429435483870968e-05,
+      "loss": 1.7368,
+      "step": 30040
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 2.096379518508911,
+      "learning_rate": 2.4269153225806454e-05,
+      "loss": 1.8122,
+      "step": 30050
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 1.6266838312149048,
+      "learning_rate": 2.4243951612903225e-05,
+      "loss": 1.9008,
+      "step": 30060
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 1.5231765508651733,
+      "learning_rate": 2.4218750000000003e-05,
+      "loss": 2.1103,
+      "step": 30070
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 3.266094207763672,
+      "learning_rate": 2.4193548387096777e-05,
+      "loss": 1.8378,
+      "step": 30080
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 1.376300573348999,
+      "learning_rate": 2.4168346774193548e-05,
+      "loss": 1.7472,
+      "step": 30090
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 3.340886116027832,
+      "learning_rate": 2.4143145161290322e-05,
+      "loss": 1.9843,
+      "step": 30100
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 1.548095703125,
+      "learning_rate": 2.4117943548387097e-05,
+      "loss": 1.7241,
+      "step": 30110
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 1.1897659301757812,
+      "learning_rate": 2.4092741935483874e-05,
+      "loss": 1.9187,
+      "step": 30120
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 1.8396437168121338,
+      "learning_rate": 2.4067540322580645e-05,
+      "loss": 1.7097,
+      "step": 30130
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 1.6620134115219116,
+      "learning_rate": 2.404233870967742e-05,
+      "loss": 1.5022,
+      "step": 30140
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 1.1131492853164673,
+      "learning_rate": 2.4017137096774194e-05,
+      "loss": 1.7104,
+      "step": 30150
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 1.692568302154541,
+      "learning_rate": 2.3991935483870968e-05,
+      "loss": 1.7216,
+      "step": 30160
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 1.6983649730682373,
+      "learning_rate": 2.3966733870967743e-05,
+      "loss": 1.8791,
+      "step": 30170
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.9225351810455322,
+      "learning_rate": 2.3941532258064517e-05,
+      "loss": 1.8729,
+      "step": 30180
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 1.2060664892196655,
+      "learning_rate": 2.391633064516129e-05,
+      "loss": 1.8291,
+      "step": 30190
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 2.833542823791504,
+      "learning_rate": 2.3891129032258066e-05,
+      "loss": 1.8799,
+      "step": 30200
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 1.5487159490585327,
+      "learning_rate": 2.386592741935484e-05,
+      "loss": 1.8167,
+      "step": 30210
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 1.5011478662490845,
+      "learning_rate": 2.3840725806451614e-05,
+      "loss": 1.7902,
+      "step": 30220
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 2.1998558044433594,
+      "learning_rate": 2.381552419354839e-05,
+      "loss": 1.8713,
+      "step": 30230
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 1.0746917724609375,
+      "learning_rate": 2.3790322580645163e-05,
+      "loss": 1.6199,
+      "step": 30240
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 1.5168962478637695,
+      "learning_rate": 2.3765120967741934e-05,
+      "loss": 1.897,
+      "step": 30250
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 1.3452097177505493,
+      "learning_rate": 2.3739919354838712e-05,
+      "loss": 1.7195,
+      "step": 30260
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 1.3904005289077759,
+      "learning_rate": 2.3714717741935486e-05,
+      "loss": 1.9564,
+      "step": 30270
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 1.8370044231414795,
+      "learning_rate": 2.368951612903226e-05,
+      "loss": 1.6716,
+      "step": 30280
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 1.4109065532684326,
+      "learning_rate": 2.366431451612903e-05,
+      "loss": 1.9679,
+      "step": 30290
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 2.4946959018707275,
+      "learning_rate": 2.3639112903225806e-05,
+      "loss": 1.9217,
+      "step": 30300
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 1.2590285539627075,
+      "learning_rate": 2.3613911290322583e-05,
+      "loss": 1.6728,
+      "step": 30310
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 1.721440315246582,
+      "learning_rate": 2.3588709677419358e-05,
+      "loss": 1.9157,
+      "step": 30320
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.5939640402793884,
+      "learning_rate": 2.356350806451613e-05,
+      "loss": 1.7548,
+      "step": 30330
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 1.20307195186615,
+      "learning_rate": 2.3538306451612903e-05,
+      "loss": 1.8166,
+      "step": 30340
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 1.6340670585632324,
+      "learning_rate": 2.351310483870968e-05,
+      "loss": 1.8617,
+      "step": 30350
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 1.7891130447387695,
+      "learning_rate": 2.3487903225806455e-05,
+      "loss": 1.9574,
+      "step": 30360
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 1.4477919340133667,
+      "learning_rate": 2.3462701612903226e-05,
+      "loss": 1.8792,
+      "step": 30370
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 1.7139804363250732,
+      "learning_rate": 2.34375e-05,
+      "loss": 1.7648,
+      "step": 30380
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 1.415636658668518,
+      "learning_rate": 2.3412298387096775e-05,
+      "loss": 1.7742,
+      "step": 30390
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.7552508115768433,
+      "learning_rate": 2.338709677419355e-05,
+      "loss": 1.6127,
+      "step": 30400
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 2.569566011428833,
+      "learning_rate": 2.3361895161290324e-05,
+      "loss": 1.7821,
+      "step": 30410
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 1.8127129077911377,
+      "learning_rate": 2.3336693548387098e-05,
+      "loss": 1.7878,
+      "step": 30420
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 1.3728703260421753,
+      "learning_rate": 2.3311491935483872e-05,
+      "loss": 1.6734,
+      "step": 30430
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 1.1376951932907104,
+      "learning_rate": 2.3286290322580647e-05,
+      "loss": 1.6906,
+      "step": 30440
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 1.072453498840332,
+      "learning_rate": 2.326108870967742e-05,
+      "loss": 1.7552,
+      "step": 30450
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 1.8703382015228271,
+      "learning_rate": 2.3235887096774195e-05,
+      "loss": 1.9223,
+      "step": 30460
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.7798607349395752,
+      "learning_rate": 2.321068548387097e-05,
+      "loss": 1.5485,
+      "step": 30470
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 2.196448564529419,
+      "learning_rate": 2.318548387096774e-05,
+      "loss": 1.8569,
+      "step": 30480
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 1.715319275856018,
+      "learning_rate": 2.3160282258064518e-05,
+      "loss": 1.811,
+      "step": 30490
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.9453045725822449,
+      "learning_rate": 2.3135080645161293e-05,
+      "loss": 1.503,
+      "step": 30500
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 1.1946184635162354,
+      "learning_rate": 2.3109879032258067e-05,
+      "loss": 1.6343,
+      "step": 30510
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 1.3308545351028442,
+      "learning_rate": 2.3084677419354838e-05,
+      "loss": 1.7186,
+      "step": 30520
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 1.9617446660995483,
+      "learning_rate": 2.3059475806451612e-05,
+      "loss": 1.8491,
+      "step": 30530
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 1.6491388082504272,
+      "learning_rate": 2.303427419354839e-05,
+      "loss": 1.4906,
+      "step": 30540
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 1.2358758449554443,
+      "learning_rate": 2.3009072580645164e-05,
+      "loss": 1.8956,
+      "step": 30550
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 1.4955576658248901,
+      "learning_rate": 2.2983870967741935e-05,
+      "loss": 1.8398,
+      "step": 30560
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 1.7017693519592285,
+      "learning_rate": 2.295866935483871e-05,
+      "loss": 1.9784,
+      "step": 30570
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 1.9643428325653076,
+      "learning_rate": 2.2933467741935484e-05,
+      "loss": 1.8337,
+      "step": 30580
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 2.749197244644165,
+      "learning_rate": 2.2908266129032262e-05,
+      "loss": 1.6592,
+      "step": 30590
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 1.8363304138183594,
+      "learning_rate": 2.2883064516129033e-05,
+      "loss": 1.7072,
+      "step": 30600
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 1.457939624786377,
+      "learning_rate": 2.2857862903225807e-05,
+      "loss": 1.7543,
+      "step": 30610
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 1.171261191368103,
+      "learning_rate": 2.283266129032258e-05,
+      "loss": 1.8417,
+      "step": 30620
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 1.3251845836639404,
+      "learning_rate": 2.2807459677419356e-05,
+      "loss": 1.8415,
+      "step": 30630
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 1.5285133123397827,
+      "learning_rate": 2.278225806451613e-05,
+      "loss": 1.8103,
+      "step": 30640
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 1.9901608228683472,
+      "learning_rate": 2.2757056451612904e-05,
+      "loss": 1.7335,
+      "step": 30650
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 1.3826160430908203,
+      "learning_rate": 2.273185483870968e-05,
+      "loss": 1.7523,
+      "step": 30660
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 1.433363676071167,
+      "learning_rate": 2.2706653225806453e-05,
+      "loss": 1.9782,
+      "step": 30670
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 1.2009791135787964,
+      "learning_rate": 2.2681451612903227e-05,
+      "loss": 1.912,
+      "step": 30680
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 1.6721488237380981,
+      "learning_rate": 2.2656250000000002e-05,
+      "loss": 1.7639,
+      "step": 30690
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 1.1318165063858032,
+      "learning_rate": 2.2631048387096776e-05,
+      "loss": 2.0292,
+      "step": 30700
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 1.7000799179077148,
+      "learning_rate": 2.2605846774193547e-05,
+      "loss": 1.8603,
+      "step": 30710
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 1.3782899379730225,
+      "learning_rate": 2.258064516129032e-05,
+      "loss": 1.6332,
+      "step": 30720
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 1.4304676055908203,
+      "learning_rate": 2.25554435483871e-05,
+      "loss": 1.9543,
+      "step": 30730
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 1.3673580884933472,
+      "learning_rate": 2.2530241935483874e-05,
+      "loss": 1.9566,
+      "step": 30740
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 1.2038042545318604,
+      "learning_rate": 2.2505040322580645e-05,
+      "loss": 1.733,
+      "step": 30750
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 1.430306315422058,
+      "learning_rate": 2.247983870967742e-05,
+      "loss": 1.7361,
+      "step": 30760
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.5055277943611145,
+      "learning_rate": 2.2454637096774193e-05,
+      "loss": 1.8592,
+      "step": 30770
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 1.3457138538360596,
+      "learning_rate": 2.242943548387097e-05,
+      "loss": 1.588,
+      "step": 30780
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 1.5819923877716064,
+      "learning_rate": 2.2404233870967742e-05,
+      "loss": 1.8438,
+      "step": 30790
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 1.6003981828689575,
+      "learning_rate": 2.2379032258064516e-05,
+      "loss": 1.699,
+      "step": 30800
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 1.0775926113128662,
+      "learning_rate": 2.235383064516129e-05,
+      "loss": 1.767,
+      "step": 30810
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 1.4512295722961426,
+      "learning_rate": 2.2328629032258065e-05,
+      "loss": 1.9535,
+      "step": 30820
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 2.2924768924713135,
+      "learning_rate": 2.230342741935484e-05,
+      "loss": 1.8213,
+      "step": 30830
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 1.444682240486145,
+      "learning_rate": 2.2278225806451614e-05,
+      "loss": 1.6686,
+      "step": 30840
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 1.3644118309020996,
+      "learning_rate": 2.2253024193548388e-05,
+      "loss": 1.4173,
+      "step": 30850
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 1.6040750741958618,
+      "learning_rate": 2.2227822580645162e-05,
+      "loss": 1.8321,
+      "step": 30860
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 2.3805015087127686,
+      "learning_rate": 2.2202620967741937e-05,
+      "loss": 1.8,
+      "step": 30870
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 1.1117628812789917,
+      "learning_rate": 2.217741935483871e-05,
+      "loss": 1.7929,
+      "step": 30880
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 2.0806496143341064,
+      "learning_rate": 2.2152217741935485e-05,
+      "loss": 1.7234,
+      "step": 30890
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.9832504987716675,
+      "learning_rate": 2.212701612903226e-05,
+      "loss": 1.8111,
+      "step": 30900
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 1.1310473680496216,
+      "learning_rate": 2.2101814516129034e-05,
+      "loss": 1.6381,
+      "step": 30910
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 1.0470945835113525,
+      "learning_rate": 2.207661290322581e-05,
+      "loss": 1.8005,
+      "step": 30920
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 1.3812587261199951,
+      "learning_rate": 2.2051411290322583e-05,
+      "loss": 1.7918,
+      "step": 30930
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 1.4525903463363647,
+      "learning_rate": 2.2026209677419357e-05,
+      "loss": 1.6955,
+      "step": 30940
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.6666645407676697,
+      "learning_rate": 2.2001008064516128e-05,
+      "loss": 1.7723,
+      "step": 30950
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 2.1903719902038574,
+      "learning_rate": 2.1975806451612906e-05,
+      "loss": 1.9578,
+      "step": 30960
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 1.419280767440796,
+      "learning_rate": 2.195060483870968e-05,
+      "loss": 1.759,
+      "step": 30970
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 1.0745269060134888,
+      "learning_rate": 2.1925403225806454e-05,
+      "loss": 1.9227,
+      "step": 30980
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 2.435979127883911,
+      "learning_rate": 2.1900201612903225e-05,
+      "loss": 1.7905,
+      "step": 30990
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 1.8779035806655884,
+      "learning_rate": 2.1875e-05,
+      "loss": 1.7593,
+      "step": 31000
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 1.1014496088027954,
+      "learning_rate": 2.1849798387096778e-05,
+      "loss": 1.5619,
+      "step": 31010
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 1.3093678951263428,
+      "learning_rate": 2.182459677419355e-05,
+      "loss": 1.7517,
+      "step": 31020
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 1.12551748752594,
+      "learning_rate": 2.1799395161290323e-05,
+      "loss": 1.7599,
+      "step": 31030
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 1.2173627614974976,
+      "learning_rate": 2.1774193548387097e-05,
+      "loss": 1.8073,
+      "step": 31040
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 1.2952498197555542,
+      "learning_rate": 2.174899193548387e-05,
+      "loss": 1.6175,
+      "step": 31050
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 1.2082802057266235,
+      "learning_rate": 2.1723790322580646e-05,
+      "loss": 1.8723,
+      "step": 31060
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 1.3725805282592773,
+      "learning_rate": 2.169858870967742e-05,
+      "loss": 1.7328,
+      "step": 31070
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 1.0354795455932617,
+      "learning_rate": 2.1673387096774195e-05,
+      "loss": 1.8932,
+      "step": 31080
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 1.4092119932174683,
+      "learning_rate": 2.164818548387097e-05,
+      "loss": 1.8465,
+      "step": 31090
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 1.1847174167633057,
+      "learning_rate": 2.1622983870967743e-05,
+      "loss": 1.6287,
+      "step": 31100
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 1.0280671119689941,
+      "learning_rate": 2.1597782258064518e-05,
+      "loss": 1.8356,
+      "step": 31110
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 1.8567456007003784,
+      "learning_rate": 2.1572580645161292e-05,
+      "loss": 1.7949,
+      "step": 31120
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 1.7573846578598022,
+      "learning_rate": 2.1547379032258066e-05,
+      "loss": 1.6938,
+      "step": 31130
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 1.9399672746658325,
+      "learning_rate": 2.1522177419354837e-05,
+      "loss": 1.7699,
+      "step": 31140
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 1.6788302659988403,
+      "learning_rate": 2.1496975806451615e-05,
+      "loss": 1.8243,
+      "step": 31150
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 1.706986427307129,
+      "learning_rate": 2.147177419354839e-05,
+      "loss": 1.7696,
+      "step": 31160
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 1.5500811338424683,
+      "learning_rate": 2.1446572580645164e-05,
+      "loss": 1.557,
+      "step": 31170
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 1.2408126592636108,
+      "learning_rate": 2.1421370967741935e-05,
+      "loss": 1.8461,
+      "step": 31180
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.9615346789360046,
+      "learning_rate": 2.139616935483871e-05,
+      "loss": 1.7074,
+      "step": 31190
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 2.9064273834228516,
+      "learning_rate": 2.1370967741935487e-05,
+      "loss": 1.7598,
+      "step": 31200
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 2.331366777420044,
+      "learning_rate": 2.134576612903226e-05,
+      "loss": 1.8826,
+      "step": 31210
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 1.518226146697998,
+      "learning_rate": 2.1320564516129032e-05,
+      "loss": 1.7149,
+      "step": 31220
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 1.9157253503799438,
+      "learning_rate": 2.1295362903225806e-05,
+      "loss": 1.7738,
+      "step": 31230
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 2.158038854598999,
+      "learning_rate": 2.127016129032258e-05,
+      "loss": 1.8004,
+      "step": 31240
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 1.0206481218338013,
+      "learning_rate": 2.124495967741936e-05,
+      "loss": 1.6859,
+      "step": 31250
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 1.2547589540481567,
+      "learning_rate": 2.121975806451613e-05,
+      "loss": 1.8397,
+      "step": 31260
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.8262723088264465,
+      "learning_rate": 2.1194556451612904e-05,
+      "loss": 1.8662,
+      "step": 31270
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 1.4317469596862793,
+      "learning_rate": 2.1169354838709678e-05,
+      "loss": 1.7237,
+      "step": 31280
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 1.6417036056518555,
+      "learning_rate": 2.1144153225806452e-05,
+      "loss": 1.7875,
+      "step": 31290
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 1.2774657011032104,
+      "learning_rate": 2.1118951612903227e-05,
+      "loss": 1.8162,
+      "step": 31300
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 1.3275634050369263,
+      "learning_rate": 2.109375e-05,
+      "loss": 1.7998,
+      "step": 31310
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 1.4177782535552979,
+      "learning_rate": 2.1068548387096775e-05,
+      "loss": 1.6755,
+      "step": 31320
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 1.5082958936691284,
+      "learning_rate": 2.1043346774193546e-05,
+      "loss": 1.5703,
+      "step": 31330
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 1.2633808851242065,
+      "learning_rate": 2.1018145161290324e-05,
+      "loss": 1.91,
+      "step": 31340
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 4.409839153289795,
+      "learning_rate": 2.09929435483871e-05,
+      "loss": 1.8445,
+      "step": 31350
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 1.1917202472686768,
+      "learning_rate": 2.0967741935483873e-05,
+      "loss": 1.8184,
+      "step": 31360
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.9644743204116821,
+      "learning_rate": 2.0942540322580644e-05,
+      "loss": 1.7323,
+      "step": 31370
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 1.7170181274414062,
+      "learning_rate": 2.0917338709677418e-05,
+      "loss": 1.764,
+      "step": 31380
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 1.1171395778656006,
+      "learning_rate": 2.0892137096774196e-05,
+      "loss": 1.8261,
+      "step": 31390
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.9885112643241882,
+      "learning_rate": 2.086693548387097e-05,
+      "loss": 1.9858,
+      "step": 31400
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 2.2308037281036377,
+      "learning_rate": 2.084173387096774e-05,
+      "loss": 1.9395,
+      "step": 31410
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 1.2090790271759033,
+      "learning_rate": 2.0816532258064516e-05,
+      "loss": 1.7856,
+      "step": 31420
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 1.895058512687683,
+      "learning_rate": 2.0791330645161293e-05,
+      "loss": 1.8474,
+      "step": 31430
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 1.1808596849441528,
+      "learning_rate": 2.0766129032258068e-05,
+      "loss": 1.8278,
+      "step": 31440
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 1.4231454133987427,
+      "learning_rate": 2.074092741935484e-05,
+      "loss": 1.7645,
+      "step": 31450
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 3.842315912246704,
+      "learning_rate": 2.0715725806451613e-05,
+      "loss": 1.8937,
+      "step": 31460
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 1.483882188796997,
+      "learning_rate": 2.0690524193548387e-05,
+      "loss": 1.8636,
+      "step": 31470
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 1.7659419775009155,
+      "learning_rate": 2.0665322580645165e-05,
+      "loss": 1.7757,
+      "step": 31480
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 1.083162784576416,
+      "learning_rate": 2.0640120967741936e-05,
+      "loss": 1.5466,
+      "step": 31490
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 1.0764532089233398,
+      "learning_rate": 2.061491935483871e-05,
+      "loss": 1.7884,
+      "step": 31500
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 1.8837714195251465,
+      "learning_rate": 2.0589717741935485e-05,
+      "loss": 1.7462,
+      "step": 31510
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 1.0759587287902832,
+      "learning_rate": 2.056451612903226e-05,
+      "loss": 1.6921,
+      "step": 31520
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 1.8649309873580933,
+      "learning_rate": 2.0539314516129033e-05,
+      "loss": 1.6004,
+      "step": 31530
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 1.1448041200637817,
+      "learning_rate": 2.0514112903225808e-05,
+      "loss": 1.7218,
+      "step": 31540
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 1.272073745727539,
+      "learning_rate": 2.0488911290322582e-05,
+      "loss": 1.7143,
+      "step": 31550
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 1.3945937156677246,
+      "learning_rate": 2.0463709677419356e-05,
+      "loss": 1.5904,
+      "step": 31560
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 1.0962601900100708,
+      "learning_rate": 2.043850806451613e-05,
+      "loss": 1.7186,
+      "step": 31570
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 1.6760454177856445,
+      "learning_rate": 2.0413306451612905e-05,
+      "loss": 1.9838,
+      "step": 31580
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 1.4671212434768677,
+      "learning_rate": 2.038810483870968e-05,
+      "loss": 1.8059,
+      "step": 31590
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 1.2559648752212524,
+      "learning_rate": 2.0362903225806454e-05,
+      "loss": 1.8333,
+      "step": 31600
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 1.0465999841690063,
+      "learning_rate": 2.0337701612903225e-05,
+      "loss": 1.7258,
+      "step": 31610
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 1.3558374643325806,
+      "learning_rate": 2.0312500000000002e-05,
+      "loss": 1.8188,
+      "step": 31620
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 1.2438677549362183,
+      "learning_rate": 2.0287298387096777e-05,
+      "loss": 1.7728,
+      "step": 31630
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 1.2059234380722046,
+      "learning_rate": 2.0262096774193548e-05,
+      "loss": 1.7102,
+      "step": 31640
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 1.8832862377166748,
+      "learning_rate": 2.0236895161290322e-05,
+      "loss": 1.9838,
+      "step": 31650
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 1.257440209388733,
+      "learning_rate": 2.0211693548387096e-05,
+      "loss": 1.7067,
+      "step": 31660
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 1.78580904006958,
+      "learning_rate": 2.0186491935483874e-05,
+      "loss": 1.8802,
+      "step": 31670
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 2.6149418354034424,
+      "learning_rate": 2.0161290322580645e-05,
+      "loss": 1.5543,
+      "step": 31680
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 2.220379114151001,
+      "learning_rate": 2.013608870967742e-05,
+      "loss": 1.6943,
+      "step": 31690
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 1.345445156097412,
+      "learning_rate": 2.0110887096774194e-05,
+      "loss": 1.6023,
+      "step": 31700
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 1.3110320568084717,
+      "learning_rate": 2.0085685483870968e-05,
+      "loss": 1.8317,
+      "step": 31710
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 2.0120298862457275,
+      "learning_rate": 2.0060483870967743e-05,
+      "loss": 1.8218,
+      "step": 31720
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 1.2286854982376099,
+      "learning_rate": 2.0035282258064517e-05,
+      "loss": 1.6627,
+      "step": 31730
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.7436784505844116,
+      "learning_rate": 2.001008064516129e-05,
+      "loss": 1.9029,
+      "step": 31740
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 1.7187528610229492,
+      "learning_rate": 1.9984879032258066e-05,
+      "loss": 1.8028,
+      "step": 31750
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 2.493142604827881,
+      "learning_rate": 1.995967741935484e-05,
+      "loss": 1.9619,
+      "step": 31760
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 2.519582509994507,
+      "learning_rate": 1.9934475806451614e-05,
+      "loss": 1.8582,
+      "step": 31770
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 2.52058482170105,
+      "learning_rate": 1.990927419354839e-05,
+      "loss": 1.9071,
+      "step": 31780
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 1.4814963340759277,
+      "learning_rate": 1.9884072580645163e-05,
+      "loss": 1.778,
+      "step": 31790
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.9599142670631409,
+      "learning_rate": 1.9858870967741934e-05,
+      "loss": 1.6011,
+      "step": 31800
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 1.7839455604553223,
+      "learning_rate": 1.983366935483871e-05,
+      "loss": 1.6165,
+      "step": 31810
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 1.3308591842651367,
+      "learning_rate": 1.9808467741935486e-05,
+      "loss": 1.785,
+      "step": 31820
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 1.2342950105667114,
+      "learning_rate": 1.978326612903226e-05,
+      "loss": 1.777,
+      "step": 31830
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 2.272684335708618,
+      "learning_rate": 1.975806451612903e-05,
+      "loss": 1.7528,
+      "step": 31840
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 1.2012065649032593,
+      "learning_rate": 1.9732862903225806e-05,
+      "loss": 1.7534,
+      "step": 31850
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 1.949897050857544,
+      "learning_rate": 1.9707661290322583e-05,
+      "loss": 1.7154,
+      "step": 31860
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 1.8213434219360352,
+      "learning_rate": 1.9682459677419358e-05,
+      "loss": 1.8743,
+      "step": 31870
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.748674213886261,
+      "learning_rate": 1.965725806451613e-05,
+      "loss": 1.6488,
+      "step": 31880
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 1.0774415731430054,
+      "learning_rate": 1.9632056451612903e-05,
+      "loss": 1.9017,
+      "step": 31890
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 2.1822779178619385,
+      "learning_rate": 1.9606854838709677e-05,
+      "loss": 1.7144,
+      "step": 31900
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.8576963543891907,
+      "learning_rate": 1.9581653225806455e-05,
+      "loss": 1.7151,
+      "step": 31910
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 1.220291256904602,
+      "learning_rate": 1.9556451612903226e-05,
+      "loss": 1.666,
+      "step": 31920
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 1.4384976625442505,
+      "learning_rate": 1.953125e-05,
+      "loss": 1.8681,
+      "step": 31930
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 2.671752691268921,
+      "learning_rate": 1.9506048387096775e-05,
+      "loss": 1.803,
+      "step": 31940
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 1.1045801639556885,
+      "learning_rate": 1.948084677419355e-05,
+      "loss": 1.7885,
+      "step": 31950
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 1.302350401878357,
+      "learning_rate": 1.9455645161290323e-05,
+      "loss": 1.7979,
+      "step": 31960
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 1.2025892734527588,
+      "learning_rate": 1.9430443548387098e-05,
+      "loss": 1.7608,
+      "step": 31970
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 1.8169459104537964,
+      "learning_rate": 1.9405241935483872e-05,
+      "loss": 1.7991,
+      "step": 31980
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 1.41158127784729,
+      "learning_rate": 1.9380040322580646e-05,
+      "loss": 1.9011,
+      "step": 31990
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 1.000457525253296,
+      "learning_rate": 1.935483870967742e-05,
+      "loss": 1.4687,
+      "step": 32000
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 1.199607491493225,
+      "learning_rate": 1.9329637096774195e-05,
+      "loss": 1.8468,
+      "step": 32010
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 1.1331236362457275,
+      "learning_rate": 1.930443548387097e-05,
+      "loss": 1.8155,
+      "step": 32020
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 1.0199438333511353,
+      "learning_rate": 1.927923387096774e-05,
+      "loss": 1.841,
+      "step": 32030
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 1.303883671760559,
+      "learning_rate": 1.9254032258064518e-05,
+      "loss": 1.7554,
+      "step": 32040
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 1.5762603282928467,
+      "learning_rate": 1.9228830645161293e-05,
+      "loss": 1.7676,
+      "step": 32050
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.8675571084022522,
+      "learning_rate": 1.9203629032258067e-05,
+      "loss": 1.5396,
+      "step": 32060
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 1.6575021743774414,
+      "learning_rate": 1.9178427419354838e-05,
+      "loss": 1.7665,
+      "step": 32070
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 2.16495680809021,
+      "learning_rate": 1.9153225806451612e-05,
+      "loss": 1.8923,
+      "step": 32080
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 1.3414102792739868,
+      "learning_rate": 1.912802419354839e-05,
+      "loss": 1.5479,
+      "step": 32090
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 1.9730244874954224,
+      "learning_rate": 1.9102822580645164e-05,
+      "loss": 1.7722,
+      "step": 32100
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.9060013890266418,
+      "learning_rate": 1.9077620967741935e-05,
+      "loss": 1.7796,
+      "step": 32110
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 1.9963525533676147,
+      "learning_rate": 1.905241935483871e-05,
+      "loss": 1.7214,
+      "step": 32120
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 1.5456358194351196,
+      "learning_rate": 1.9027217741935484e-05,
+      "loss": 1.8274,
+      "step": 32130
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 4.275182247161865,
+      "learning_rate": 1.900201612903226e-05,
+      "loss": 1.7902,
+      "step": 32140
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 1.122870922088623,
+      "learning_rate": 1.8976814516129033e-05,
+      "loss": 1.741,
+      "step": 32150
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 1.0829421281814575,
+      "learning_rate": 1.8951612903225807e-05,
+      "loss": 1.6232,
+      "step": 32160
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 3.4876105785369873,
+      "learning_rate": 1.892641129032258e-05,
+      "loss": 2.0516,
+      "step": 32170
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 1.8298338651657104,
+      "learning_rate": 1.8901209677419356e-05,
+      "loss": 1.6939,
+      "step": 32180
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 1.169576644897461,
+      "learning_rate": 1.887600806451613e-05,
+      "loss": 1.8993,
+      "step": 32190
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 1.3131029605865479,
+      "learning_rate": 1.8850806451612904e-05,
+      "loss": 1.8844,
+      "step": 32200
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 3.2806437015533447,
+      "learning_rate": 1.882560483870968e-05,
+      "loss": 1.5973,
+      "step": 32210
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.9818363189697266,
+      "learning_rate": 1.8800403225806453e-05,
+      "loss": 1.8089,
+      "step": 32220
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 1.7929837703704834,
+      "learning_rate": 1.8775201612903227e-05,
+      "loss": 1.8584,
+      "step": 32230
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 1.4818141460418701,
+      "learning_rate": 1.8750000000000002e-05,
+      "loss": 1.6453,
+      "step": 32240
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 1.0312066078186035,
+      "learning_rate": 1.8724798387096776e-05,
+      "loss": 1.7941,
+      "step": 32250
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 1.5595736503601074,
+      "learning_rate": 1.8699596774193547e-05,
+      "loss": 1.6552,
+      "step": 32260
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 1.2300602197647095,
+      "learning_rate": 1.867439516129032e-05,
+      "loss": 1.855,
+      "step": 32270
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 1.4556666612625122,
+      "learning_rate": 1.86491935483871e-05,
+      "loss": 1.6445,
+      "step": 32280
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 1.2595055103302002,
+      "learning_rate": 1.8623991935483873e-05,
+      "loss": 1.8234,
+      "step": 32290
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 1.305450201034546,
+      "learning_rate": 1.8598790322580644e-05,
+      "loss": 1.7733,
+      "step": 32300
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 2.6044809818267822,
+      "learning_rate": 1.857358870967742e-05,
+      "loss": 1.7996,
+      "step": 32310
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 1.1965882778167725,
+      "learning_rate": 1.8548387096774193e-05,
+      "loss": 1.9566,
+      "step": 32320
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.9314250946044922,
+      "learning_rate": 1.852318548387097e-05,
+      "loss": 1.9846,
+      "step": 32330
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 1.4565799236297607,
+      "learning_rate": 1.8497983870967742e-05,
+      "loss": 1.7769,
+      "step": 32340
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 1.2986658811569214,
+      "learning_rate": 1.8472782258064516e-05,
+      "loss": 1.8634,
+      "step": 32350
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 1.1454496383666992,
+      "learning_rate": 1.844758064516129e-05,
+      "loss": 1.8168,
+      "step": 32360
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 1.2110981941223145,
+      "learning_rate": 1.8422379032258065e-05,
+      "loss": 1.8277,
+      "step": 32370
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 1.4428951740264893,
+      "learning_rate": 1.839717741935484e-05,
+      "loss": 1.6553,
+      "step": 32380
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 4.695744514465332,
+      "learning_rate": 1.8371975806451614e-05,
+      "loss": 1.9102,
+      "step": 32390
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 1.2316994667053223,
+      "learning_rate": 1.8346774193548388e-05,
+      "loss": 1.7246,
+      "step": 32400
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 1.615728497505188,
+      "learning_rate": 1.8321572580645162e-05,
+      "loss": 1.9895,
+      "step": 32410
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 1.0998812913894653,
+      "learning_rate": 1.8296370967741937e-05,
+      "loss": 1.7854,
+      "step": 32420
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 1.4411550760269165,
+      "learning_rate": 1.827116935483871e-05,
+      "loss": 1.9057,
+      "step": 32430
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 2.047656774520874,
+      "learning_rate": 1.8245967741935485e-05,
+      "loss": 2.0217,
+      "step": 32440
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 1.351747751235962,
+      "learning_rate": 1.822076612903226e-05,
+      "loss": 1.8423,
+      "step": 32450
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 1.711450457572937,
+      "learning_rate": 1.819556451612903e-05,
+      "loss": 1.7952,
+      "step": 32460
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 1.4096416234970093,
+      "learning_rate": 1.817036290322581e-05,
+      "loss": 1.8158,
+      "step": 32470
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 1.402268409729004,
+      "learning_rate": 1.8145161290322583e-05,
+      "loss": 1.7617,
+      "step": 32480
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 1.4947301149368286,
+      "learning_rate": 1.8119959677419357e-05,
+      "loss": 1.7964,
+      "step": 32490
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 1.4989169836044312,
+      "learning_rate": 1.8094758064516128e-05,
+      "loss": 1.7727,
+      "step": 32500
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 1.1168025732040405,
+      "learning_rate": 1.8069556451612902e-05,
+      "loss": 1.6757,
+      "step": 32510
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 2.1846399307250977,
+      "learning_rate": 1.804435483870968e-05,
+      "loss": 1.9175,
+      "step": 32520
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 1.593040108680725,
+      "learning_rate": 1.8019153225806454e-05,
+      "loss": 1.6026,
+      "step": 32530
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 2.0637805461883545,
+      "learning_rate": 1.7993951612903225e-05,
+      "loss": 1.9022,
+      "step": 32540
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 2.226268768310547,
+      "learning_rate": 1.796875e-05,
+      "loss": 1.6261,
+      "step": 32550
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 2.1268930435180664,
+      "learning_rate": 1.7943548387096777e-05,
+      "loss": 1.768,
+      "step": 32560
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 1.7395910024642944,
+      "learning_rate": 1.791834677419355e-05,
+      "loss": 1.9375,
+      "step": 32570
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 1.996485710144043,
+      "learning_rate": 1.7893145161290323e-05,
+      "loss": 1.7661,
+      "step": 32580
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 1.8850375413894653,
+      "learning_rate": 1.7867943548387097e-05,
+      "loss": 1.8991,
+      "step": 32590
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 1.4898598194122314,
+      "learning_rate": 1.784274193548387e-05,
+      "loss": 1.8085,
+      "step": 32600
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 1.8822178840637207,
+      "learning_rate": 1.7817540322580646e-05,
+      "loss": 1.8174,
+      "step": 32610
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 1.4489226341247559,
+      "learning_rate": 1.779233870967742e-05,
+      "loss": 1.7839,
+      "step": 32620
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 1.6543374061584473,
+      "learning_rate": 1.7767137096774194e-05,
+      "loss": 1.7365,
+      "step": 32630
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 1.6858607530593872,
+      "learning_rate": 1.774193548387097e-05,
+      "loss": 1.6271,
+      "step": 32640
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.9373005628585815,
+      "learning_rate": 1.7716733870967743e-05,
+      "loss": 1.834,
+      "step": 32650
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 1.104684829711914,
+      "learning_rate": 1.7691532258064518e-05,
+      "loss": 1.8794,
+      "step": 32660
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 1.7709009647369385,
+      "learning_rate": 1.7666330645161292e-05,
+      "loss": 1.9117,
+      "step": 32670
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 1.895287036895752,
+      "learning_rate": 1.7641129032258066e-05,
+      "loss": 1.7107,
+      "step": 32680
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 1.5934332609176636,
+      "learning_rate": 1.7615927419354837e-05,
+      "loss": 1.686,
+      "step": 32690
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 2.1799850463867188,
+      "learning_rate": 1.7590725806451615e-05,
+      "loss": 1.8343,
+      "step": 32700
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 1.398343801498413,
+      "learning_rate": 1.756552419354839e-05,
+      "loss": 1.8189,
+      "step": 32710
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 1.0081603527069092,
+      "learning_rate": 1.7540322580645164e-05,
+      "loss": 1.7513,
+      "step": 32720
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 1.759623408317566,
+      "learning_rate": 1.7515120967741935e-05,
+      "loss": 1.7456,
+      "step": 32730
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 1.5283091068267822,
+      "learning_rate": 1.748991935483871e-05,
+      "loss": 1.7808,
+      "step": 32740
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 1.68028724193573,
+      "learning_rate": 1.7464717741935487e-05,
+      "loss": 1.8845,
+      "step": 32750
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 1.0705845355987549,
+      "learning_rate": 1.743951612903226e-05,
+      "loss": 1.8818,
+      "step": 32760
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 1.064941644668579,
+      "learning_rate": 1.7414314516129032e-05,
+      "loss": 1.7066,
+      "step": 32770
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 1.1786919832229614,
+      "learning_rate": 1.7389112903225806e-05,
+      "loss": 2.0392,
+      "step": 32780
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 1.7205005884170532,
+      "learning_rate": 1.736391129032258e-05,
+      "loss": 1.9532,
+      "step": 32790
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 1.139602780342102,
+      "learning_rate": 1.733870967741936e-05,
+      "loss": 1.6982,
+      "step": 32800
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 1.6138654947280884,
+      "learning_rate": 1.731350806451613e-05,
+      "loss": 1.889,
+      "step": 32810
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 1.409814476966858,
+      "learning_rate": 1.7288306451612904e-05,
+      "loss": 1.6748,
+      "step": 32820
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 1.5347388982772827,
+      "learning_rate": 1.7263104838709678e-05,
+      "loss": 1.7264,
+      "step": 32830
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 1.1145960092544556,
+      "learning_rate": 1.7237903225806452e-05,
+      "loss": 1.8205,
+      "step": 32840
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 1.4902585744857788,
+      "learning_rate": 1.7212701612903227e-05,
+      "loss": 1.7532,
+      "step": 32850
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 1.4833812713623047,
+      "learning_rate": 1.71875e-05,
+      "loss": 1.7367,
+      "step": 32860
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.9034027457237244,
+      "learning_rate": 1.7162298387096775e-05,
+      "loss": 1.7264,
+      "step": 32870
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 1.7502334117889404,
+      "learning_rate": 1.7137096774193546e-05,
+      "loss": 1.8026,
+      "step": 32880
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.813915491104126,
+      "learning_rate": 1.7111895161290324e-05,
+      "loss": 1.6058,
+      "step": 32890
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 1.8752402067184448,
+      "learning_rate": 1.70866935483871e-05,
+      "loss": 1.753,
+      "step": 32900
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 1.4121441841125488,
+      "learning_rate": 1.7061491935483873e-05,
+      "loss": 1.77,
+      "step": 32910
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 2.35333514213562,
+      "learning_rate": 1.7036290322580644e-05,
+      "loss": 1.9866,
+      "step": 32920
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6013392806053162,
+      "learning_rate": 1.7011088709677418e-05,
+      "loss": 1.7968,
+      "step": 32930
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 1.045096516609192,
+      "learning_rate": 1.6985887096774196e-05,
+      "loss": 1.7103,
+      "step": 32940
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 1.603283405303955,
+      "learning_rate": 1.696068548387097e-05,
+      "loss": 1.6831,
+      "step": 32950
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 1.7540736198425293,
+      "learning_rate": 1.693548387096774e-05,
+      "loss": 1.8307,
+      "step": 32960
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 1.6996979713439941,
+      "learning_rate": 1.6910282258064515e-05,
+      "loss": 1.8741,
+      "step": 32970
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 1.4957307577133179,
+      "learning_rate": 1.688508064516129e-05,
+      "loss": 1.8146,
+      "step": 32980
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 1.3205597400665283,
+      "learning_rate": 1.6859879032258068e-05,
+      "loss": 1.7828,
+      "step": 32990
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 1.3557368516921997,
+      "learning_rate": 1.683467741935484e-05,
+      "loss": 1.6496,
+      "step": 33000
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 2.0203793048858643,
+      "learning_rate": 1.6809475806451613e-05,
+      "loss": 1.8738,
+      "step": 33010
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 1.5324968099594116,
+      "learning_rate": 1.6784274193548387e-05,
+      "loss": 1.6354,
+      "step": 33020
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.9115815758705139,
+      "learning_rate": 1.675907258064516e-05,
+      "loss": 1.9051,
+      "step": 33030
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 1.6889846324920654,
+      "learning_rate": 1.6733870967741936e-05,
+      "loss": 1.7198,
+      "step": 33040
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 1.1009942293167114,
+      "learning_rate": 1.670866935483871e-05,
+      "loss": 1.6905,
+      "step": 33050
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 1.499977946281433,
+      "learning_rate": 1.6683467741935485e-05,
+      "loss": 1.7792,
+      "step": 33060
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 1.1749621629714966,
+      "learning_rate": 1.665826612903226e-05,
+      "loss": 1.7453,
+      "step": 33070
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 1.1888600587844849,
+      "learning_rate": 1.6633064516129033e-05,
+      "loss": 1.8416,
+      "step": 33080
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 1.374924898147583,
+      "learning_rate": 1.6607862903225808e-05,
+      "loss": 1.7975,
+      "step": 33090
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 1.3252980709075928,
+      "learning_rate": 1.6582661290322582e-05,
+      "loss": 1.8975,
+      "step": 33100
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 1.404721736907959,
+      "learning_rate": 1.6557459677419356e-05,
+      "loss": 1.7038,
+      "step": 33110
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 2.5581963062286377,
+      "learning_rate": 1.653225806451613e-05,
+      "loss": 1.7802,
+      "step": 33120
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 2.1316795349121094,
+      "learning_rate": 1.6507056451612905e-05,
+      "loss": 1.7918,
+      "step": 33130
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 1.5719072818756104,
+      "learning_rate": 1.648185483870968e-05,
+      "loss": 1.9546,
+      "step": 33140
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 1.6193944215774536,
+      "learning_rate": 1.6456653225806454e-05,
+      "loss": 1.9248,
+      "step": 33150
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 1.9076110124588013,
+      "learning_rate": 1.6431451612903225e-05,
+      "loss": 1.8257,
+      "step": 33160
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 2.3797178268432617,
+      "learning_rate": 1.6406250000000002e-05,
+      "loss": 1.7955,
+      "step": 33170
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 1.0402649641036987,
+      "learning_rate": 1.6381048387096777e-05,
+      "loss": 1.9139,
+      "step": 33180
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 1.8948278427124023,
+      "learning_rate": 1.6355846774193548e-05,
+      "loss": 1.7499,
+      "step": 33190
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 1.1044493913650513,
+      "learning_rate": 1.6330645161290322e-05,
+      "loss": 1.817,
+      "step": 33200
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 1.1709611415863037,
+      "learning_rate": 1.6305443548387096e-05,
+      "loss": 1.9799,
+      "step": 33210
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 1.1573008298873901,
+      "learning_rate": 1.6280241935483874e-05,
+      "loss": 1.7077,
+      "step": 33220
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 1.0398807525634766,
+      "learning_rate": 1.6255040322580645e-05,
+      "loss": 1.7589,
+      "step": 33230
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 1.5626739263534546,
+      "learning_rate": 1.622983870967742e-05,
+      "loss": 1.7745,
+      "step": 33240
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 1.692694067955017,
+      "learning_rate": 1.6204637096774194e-05,
+      "loss": 1.7872,
+      "step": 33250
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 1.5916043519973755,
+      "learning_rate": 1.6179435483870968e-05,
+      "loss": 1.7512,
+      "step": 33260
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 1.8061860799789429,
+      "learning_rate": 1.6154233870967742e-05,
+      "loss": 1.7927,
+      "step": 33270
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.9044138789176941,
+      "learning_rate": 1.6129032258064517e-05,
+      "loss": 1.7368,
+      "step": 33280
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.9928666949272156,
+      "learning_rate": 1.610383064516129e-05,
+      "loss": 1.62,
+      "step": 33290
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 1.8242844343185425,
+      "learning_rate": 1.6078629032258065e-05,
+      "loss": 1.6481,
+      "step": 33300
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.9306523203849792,
+      "learning_rate": 1.605342741935484e-05,
+      "loss": 1.691,
+      "step": 33310
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 1.9004653692245483,
+      "learning_rate": 1.6028225806451614e-05,
+      "loss": 1.8754,
+      "step": 33320
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 1.7879743576049805,
+      "learning_rate": 1.600302419354839e-05,
+      "loss": 1.587,
+      "step": 33330
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 1.4189718961715698,
+      "learning_rate": 1.5977822580645163e-05,
+      "loss": 1.8249,
+      "step": 33340
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 1.4086508750915527,
+      "learning_rate": 1.5952620967741934e-05,
+      "loss": 1.7507,
+      "step": 33350
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 1.515584945678711,
+      "learning_rate": 1.592741935483871e-05,
+      "loss": 1.8526,
+      "step": 33360
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 1.1979918479919434,
+      "learning_rate": 1.5902217741935486e-05,
+      "loss": 1.6765,
+      "step": 33370
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 1.709639549255371,
+      "learning_rate": 1.587701612903226e-05,
+      "loss": 1.8336,
+      "step": 33380
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 2.3151185512542725,
+      "learning_rate": 1.585181451612903e-05,
+      "loss": 1.6459,
+      "step": 33390
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.871351420879364,
+      "learning_rate": 1.5826612903225806e-05,
+      "loss": 1.9548,
+      "step": 33400
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.9601405262947083,
+      "learning_rate": 1.5801411290322583e-05,
+      "loss": 1.655,
+      "step": 33410
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 1.6184051036834717,
+      "learning_rate": 1.5776209677419358e-05,
+      "loss": 1.8159,
+      "step": 33420
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 2.2493479251861572,
+      "learning_rate": 1.575100806451613e-05,
+      "loss": 1.7898,
+      "step": 33430
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 1.326961636543274,
+      "learning_rate": 1.5725806451612903e-05,
+      "loss": 1.8043,
+      "step": 33440
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.8613133430480957,
+      "learning_rate": 1.5700604838709677e-05,
+      "loss": 1.9111,
+      "step": 33450
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 1.4881073236465454,
+      "learning_rate": 1.5675403225806455e-05,
+      "loss": 1.626,
+      "step": 33460
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 2.398189067840576,
+      "learning_rate": 1.5650201612903226e-05,
+      "loss": 1.9592,
+      "step": 33470
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 2.559082269668579,
+      "learning_rate": 1.5625e-05,
+      "loss": 1.6281,
+      "step": 33480
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 1.2802702188491821,
+      "learning_rate": 1.5599798387096775e-05,
+      "loss": 1.9162,
+      "step": 33490
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 1.5401066541671753,
+      "learning_rate": 1.557459677419355e-05,
+      "loss": 1.8351,
+      "step": 33500
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 1.4671106338500977,
+      "learning_rate": 1.5549395161290323e-05,
+      "loss": 1.5919,
+      "step": 33510
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.8489421606063843,
+      "learning_rate": 1.5524193548387098e-05,
+      "loss": 1.8544,
+      "step": 33520
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 2.760457754135132,
+      "learning_rate": 1.5498991935483872e-05,
+      "loss": 1.7592,
+      "step": 33530
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 2.016407012939453,
+      "learning_rate": 1.5473790322580643e-05,
+      "loss": 1.8257,
+      "step": 33540
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 1.5972362756729126,
+      "learning_rate": 1.544858870967742e-05,
+      "loss": 1.8296,
+      "step": 33550
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 1.3354487419128418,
+      "learning_rate": 1.5423387096774195e-05,
+      "loss": 1.9342,
+      "step": 33560
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 1.3717316389083862,
+      "learning_rate": 1.539818548387097e-05,
+      "loss": 1.7686,
+      "step": 33570
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 2.1471335887908936,
+      "learning_rate": 1.537298387096774e-05,
+      "loss": 1.8047,
+      "step": 33580
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 1.6890807151794434,
+      "learning_rate": 1.5347782258064515e-05,
+      "loss": 1.8541,
+      "step": 33590
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.884679913520813,
+      "learning_rate": 1.5322580645161292e-05,
+      "loss": 1.8412,
+      "step": 33600
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 1.3990418910980225,
+      "learning_rate": 1.5297379032258067e-05,
+      "loss": 1.8125,
+      "step": 33610
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 1.5013306140899658,
+      "learning_rate": 1.5272177419354838e-05,
+      "loss": 1.72,
+      "step": 33620
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 1.718822717666626,
+      "learning_rate": 1.5246975806451614e-05,
+      "loss": 1.8041,
+      "step": 33630
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 1.1815522909164429,
+      "learning_rate": 1.5221774193548388e-05,
+      "loss": 1.8075,
+      "step": 33640
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 1.9119668006896973,
+      "learning_rate": 1.5196572580645163e-05,
+      "loss": 1.7769,
+      "step": 33650
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.924808144569397,
+      "learning_rate": 1.5171370967741935e-05,
+      "loss": 1.8332,
+      "step": 33660
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 1.7381653785705566,
+      "learning_rate": 1.514616935483871e-05,
+      "loss": 1.7352,
+      "step": 33670
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 1.951949119567871,
+      "learning_rate": 1.5120967741935486e-05,
+      "loss": 1.6059,
+      "step": 33680
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 1.2686573266983032,
+      "learning_rate": 1.509576612903226e-05,
+      "loss": 1.7353,
+      "step": 33690
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 1.1494786739349365,
+      "learning_rate": 1.5070564516129033e-05,
+      "loss": 1.7809,
+      "step": 33700
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 1.8090535402297974,
+      "learning_rate": 1.5045362903225807e-05,
+      "loss": 1.8105,
+      "step": 33710
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 1.1818082332611084,
+      "learning_rate": 1.5020161290322581e-05,
+      "loss": 1.7702,
+      "step": 33720
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 1.2743195295333862,
+      "learning_rate": 1.4994959677419357e-05,
+      "loss": 1.88,
+      "step": 33730
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.8523210287094116,
+      "learning_rate": 1.4969758064516128e-05,
+      "loss": 1.6924,
+      "step": 33740
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 1.140779733657837,
+      "learning_rate": 1.4944556451612904e-05,
+      "loss": 1.7851,
+      "step": 33750
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 1.5260815620422363,
+      "learning_rate": 1.4919354838709679e-05,
+      "loss": 1.6678,
+      "step": 33760
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 1.0565932989120483,
+      "learning_rate": 1.4894153225806453e-05,
+      "loss": 1.6436,
+      "step": 33770
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 1.4386216402053833,
+      "learning_rate": 1.4868951612903226e-05,
+      "loss": 1.8787,
+      "step": 33780
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 1.5400336980819702,
+      "learning_rate": 1.484375e-05,
+      "loss": 1.7629,
+      "step": 33790
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 1.8881813287734985,
+      "learning_rate": 1.4818548387096776e-05,
+      "loss": 1.8162,
+      "step": 33800
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.9996674060821533,
+      "learning_rate": 1.4793346774193547e-05,
+      "loss": 1.7656,
+      "step": 33810
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 1.805514931678772,
+      "learning_rate": 1.4768145161290323e-05,
+      "loss": 1.8308,
+      "step": 33820
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 1.6317267417907715,
+      "learning_rate": 1.4742943548387097e-05,
+      "loss": 1.9489,
+      "step": 33830
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 1.7620140314102173,
+      "learning_rate": 1.4717741935483872e-05,
+      "loss": 1.8978,
+      "step": 33840
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 1.0148876905441284,
+      "learning_rate": 1.4692540322580644e-05,
+      "loss": 1.8698,
+      "step": 33850
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 1.841996431350708,
+      "learning_rate": 1.4667338709677419e-05,
+      "loss": 1.9407,
+      "step": 33860
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 1.5575697422027588,
+      "learning_rate": 1.4642137096774195e-05,
+      "loss": 1.7171,
+      "step": 33870
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 1.407071828842163,
+      "learning_rate": 1.4616935483870969e-05,
+      "loss": 1.805,
+      "step": 33880
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 1.346091628074646,
+      "learning_rate": 1.4591733870967742e-05,
+      "loss": 1.9444,
+      "step": 33890
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 1.7372056245803833,
+      "learning_rate": 1.4566532258064516e-05,
+      "loss": 1.6038,
+      "step": 33900
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 1.3293724060058594,
+      "learning_rate": 1.4541330645161292e-05,
+      "loss": 1.7561,
+      "step": 33910
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.8848250508308411,
+      "learning_rate": 1.4516129032258066e-05,
+      "loss": 1.6733,
+      "step": 33920
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 1.477640986442566,
+      "learning_rate": 1.4490927419354839e-05,
+      "loss": 1.7804,
+      "step": 33930
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 1.964158058166504,
+      "learning_rate": 1.4465725806451613e-05,
+      "loss": 1.859,
+      "step": 33940
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 2.1119799613952637,
+      "learning_rate": 1.4440524193548388e-05,
+      "loss": 1.8221,
+      "step": 33950
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 1.932523250579834,
+      "learning_rate": 1.4415322580645164e-05,
+      "loss": 1.8323,
+      "step": 33960
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 1.66765558719635,
+      "learning_rate": 1.4390120967741935e-05,
+      "loss": 1.866,
+      "step": 33970
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 1.212243676185608,
+      "learning_rate": 1.4364919354838711e-05,
+      "loss": 2.04,
+      "step": 33980
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 1.7095532417297363,
+      "learning_rate": 1.4339717741935485e-05,
+      "loss": 1.8834,
+      "step": 33990
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 1.6822307109832764,
+      "learning_rate": 1.431451612903226e-05,
+      "loss": 1.9787,
+      "step": 34000
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 1.768472671508789,
+      "learning_rate": 1.4289314516129032e-05,
+      "loss": 1.8787,
+      "step": 34010
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 2.0790634155273438,
+      "learning_rate": 1.4264112903225807e-05,
+      "loss": 1.9104,
+      "step": 34020
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 1.056971788406372,
+      "learning_rate": 1.4238911290322583e-05,
+      "loss": 1.8723,
+      "step": 34030
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 2.526216745376587,
+      "learning_rate": 1.4213709677419357e-05,
+      "loss": 1.942,
+      "step": 34040
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 1.323618769645691,
+      "learning_rate": 1.418850806451613e-05,
+      "loss": 1.6168,
+      "step": 34050
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 1.2645102739334106,
+      "learning_rate": 1.4163306451612904e-05,
+      "loss": 1.7263,
+      "step": 34060
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 1.7429373264312744,
+      "learning_rate": 1.4138104838709678e-05,
+      "loss": 1.6767,
+      "step": 34070
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 1.2890541553497314,
+      "learning_rate": 1.4112903225806454e-05,
+      "loss": 1.7622,
+      "step": 34080
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 1.209720492362976,
+      "learning_rate": 1.4087701612903225e-05,
+      "loss": 1.8126,
+      "step": 34090
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 1.1765010356903076,
+      "learning_rate": 1.4062500000000001e-05,
+      "loss": 1.6831,
+      "step": 34100
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.9729856848716736,
+      "learning_rate": 1.4037298387096776e-05,
+      "loss": 1.8571,
+      "step": 34110
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 1.8464354276657104,
+      "learning_rate": 1.4012096774193548e-05,
+      "loss": 1.5579,
+      "step": 34120
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 1.627219319343567,
+      "learning_rate": 1.3986895161290323e-05,
+      "loss": 1.7762,
+      "step": 34130
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 1.6023107767105103,
+      "learning_rate": 1.3961693548387097e-05,
+      "loss": 1.8606,
+      "step": 34140
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 2.0483314990997314,
+      "learning_rate": 1.3936491935483873e-05,
+      "loss": 1.7277,
+      "step": 34150
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.9136579632759094,
+      "learning_rate": 1.3911290322580644e-05,
+      "loss": 1.6687,
+      "step": 34160
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 2.440531015396118,
+      "learning_rate": 1.388608870967742e-05,
+      "loss": 1.8108,
+      "step": 34170
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6542202234268188,
+      "learning_rate": 1.3860887096774194e-05,
+      "loss": 1.9224,
+      "step": 34180
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 1.6547138690948486,
+      "learning_rate": 1.3835685483870969e-05,
+      "loss": 1.784,
+      "step": 34190
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 1.388525366783142,
+      "learning_rate": 1.3810483870967741e-05,
+      "loss": 1.8268,
+      "step": 34200
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 1.3224115371704102,
+      "learning_rate": 1.3785282258064516e-05,
+      "loss": 1.7271,
+      "step": 34210
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 1.8573459386825562,
+      "learning_rate": 1.3760080645161292e-05,
+      "loss": 1.7749,
+      "step": 34220
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 1.1947194337844849,
+      "learning_rate": 1.3734879032258066e-05,
+      "loss": 1.5924,
+      "step": 34230
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6675629615783691,
+      "learning_rate": 1.3709677419354839e-05,
+      "loss": 1.963,
+      "step": 34240
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 1.3979710340499878,
+      "learning_rate": 1.3684475806451613e-05,
+      "loss": 1.8671,
+      "step": 34250
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 1.4006788730621338,
+      "learning_rate": 1.3659274193548387e-05,
+      "loss": 1.6309,
+      "step": 34260
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 1.2432501316070557,
+      "learning_rate": 1.3634072580645163e-05,
+      "loss": 1.7204,
+      "step": 34270
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 1.278660774230957,
+      "learning_rate": 1.3608870967741934e-05,
+      "loss": 1.7438,
+      "step": 34280
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 1.5867048501968384,
+      "learning_rate": 1.358366935483871e-05,
+      "loss": 1.6699,
+      "step": 34290
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 1.7261816263198853,
+      "learning_rate": 1.3558467741935485e-05,
+      "loss": 1.8329,
+      "step": 34300
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 1.2402093410491943,
+      "learning_rate": 1.353326612903226e-05,
+      "loss": 2.0245,
+      "step": 34310
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 1.694622278213501,
+      "learning_rate": 1.3508064516129032e-05,
+      "loss": 1.706,
+      "step": 34320
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.9541911482810974,
+      "learning_rate": 1.3482862903225806e-05,
+      "loss": 1.7958,
+      "step": 34330
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 1.792670488357544,
+      "learning_rate": 1.3457661290322582e-05,
+      "loss": 1.7987,
+      "step": 34340
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 1.051543116569519,
+      "learning_rate": 1.3432459677419357e-05,
+      "loss": 1.7722,
+      "step": 34350
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 1.6137208938598633,
+      "learning_rate": 1.340725806451613e-05,
+      "loss": 1.9025,
+      "step": 34360
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 1.1539995670318604,
+      "learning_rate": 1.3382056451612904e-05,
+      "loss": 1.8791,
+      "step": 34370
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 2.039015293121338,
+      "learning_rate": 1.3356854838709678e-05,
+      "loss": 1.9345,
+      "step": 34380
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 2.020725965499878,
+      "learning_rate": 1.3331653225806454e-05,
+      "loss": 1.9346,
+      "step": 34390
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.9070966839790344,
+      "learning_rate": 1.3306451612903225e-05,
+      "loss": 1.7433,
+      "step": 34400
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 1.266918659210205,
+      "learning_rate": 1.3281250000000001e-05,
+      "loss": 1.7346,
+      "step": 34410
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 2.1829030513763428,
+      "learning_rate": 1.3256048387096775e-05,
+      "loss": 1.8323,
+      "step": 34420
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 1.4869967699050903,
+      "learning_rate": 1.3230846774193548e-05,
+      "loss": 1.9121,
+      "step": 34430
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 1.4890785217285156,
+      "learning_rate": 1.3205645161290322e-05,
+      "loss": 1.8251,
+      "step": 34440
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 1.4250149726867676,
+      "learning_rate": 1.3180443548387098e-05,
+      "loss": 1.7099,
+      "step": 34450
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 2.262645721435547,
+      "learning_rate": 1.3155241935483873e-05,
+      "loss": 1.8723,
+      "step": 34460
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 1.4928416013717651,
+      "learning_rate": 1.3130040322580645e-05,
+      "loss": 1.6743,
+      "step": 34470
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 2.022503137588501,
+      "learning_rate": 1.310483870967742e-05,
+      "loss": 1.797,
+      "step": 34480
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 1.2006295919418335,
+      "learning_rate": 1.3079637096774194e-05,
+      "loss": 1.6578,
+      "step": 34490
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 1.3747526407241821,
+      "learning_rate": 1.305443548387097e-05,
+      "loss": 1.8286,
+      "step": 34500
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 1.0963401794433594,
+      "learning_rate": 1.3029233870967741e-05,
+      "loss": 1.983,
+      "step": 34510
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 1.7267804145812988,
+      "learning_rate": 1.3004032258064517e-05,
+      "loss": 1.8465,
+      "step": 34520
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 1.4655156135559082,
+      "learning_rate": 1.2978830645161291e-05,
+      "loss": 1.9144,
+      "step": 34530
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.9415842294692993,
+      "learning_rate": 1.2953629032258066e-05,
+      "loss": 1.7821,
+      "step": 34540
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 1.649336814880371,
+      "learning_rate": 1.2928427419354838e-05,
+      "loss": 1.7453,
+      "step": 34550
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 1.6380172967910767,
+      "learning_rate": 1.2903225806451613e-05,
+      "loss": 1.7667,
+      "step": 34560
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 1.2486231327056885,
+      "learning_rate": 1.2878024193548389e-05,
+      "loss": 1.7984,
+      "step": 34570
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 1.3570706844329834,
+      "learning_rate": 1.2852822580645163e-05,
+      "loss": 1.7744,
+      "step": 34580
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 1.354204773902893,
+      "learning_rate": 1.2827620967741936e-05,
+      "loss": 1.5923,
+      "step": 34590
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 1.50674569606781,
+      "learning_rate": 1.280241935483871e-05,
+      "loss": 1.8859,
+      "step": 34600
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 1.5789283514022827,
+      "learning_rate": 1.2777217741935484e-05,
+      "loss": 1.754,
+      "step": 34610
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 2.87563419342041,
+      "learning_rate": 1.275201612903226e-05,
+      "loss": 1.711,
+      "step": 34620
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.9886743426322937,
+      "learning_rate": 1.2726814516129031e-05,
+      "loss": 1.7657,
+      "step": 34630
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 1.477205514907837,
+      "learning_rate": 1.2701612903225808e-05,
+      "loss": 1.8464,
+      "step": 34640
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 2.7185544967651367,
+      "learning_rate": 1.2676411290322582e-05,
+      "loss": 1.8767,
+      "step": 34650
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 1.3323668241500854,
+      "learning_rate": 1.2651209677419356e-05,
+      "loss": 1.6711,
+      "step": 34660
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 1.2655704021453857,
+      "learning_rate": 1.2626008064516129e-05,
+      "loss": 1.9543,
+      "step": 34670
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 1.285814642906189,
+      "learning_rate": 1.2600806451612903e-05,
+      "loss": 2.0604,
+      "step": 34680
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 1.871640682220459,
+      "learning_rate": 1.257560483870968e-05,
+      "loss": 1.6995,
+      "step": 34690
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 1.5253214836120605,
+      "learning_rate": 1.2550403225806454e-05,
+      "loss": 1.9102,
+      "step": 34700
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 1.4711257219314575,
+      "learning_rate": 1.2525201612903226e-05,
+      "loss": 1.9719,
+      "step": 34710
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 1.4662418365478516,
+      "learning_rate": 1.25e-05,
+      "loss": 1.8894,
+      "step": 34720
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 1.1219366788864136,
+      "learning_rate": 1.2474798387096775e-05,
+      "loss": 1.764,
+      "step": 34730
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 1.1561824083328247,
+      "learning_rate": 1.244959677419355e-05,
+      "loss": 1.7682,
+      "step": 34740
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 1.392783522605896,
+      "learning_rate": 1.2424395161290322e-05,
+      "loss": 1.723,
+      "step": 34750
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 2.391662120819092,
+      "learning_rate": 1.2399193548387098e-05,
+      "loss": 1.7551,
+      "step": 34760
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 1.7713896036148071,
+      "learning_rate": 1.237399193548387e-05,
+      "loss": 1.6909,
+      "step": 34770
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.8974714875221252,
+      "learning_rate": 1.2348790322580647e-05,
+      "loss": 1.7758,
+      "step": 34780
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 1.642555594444275,
+      "learning_rate": 1.232358870967742e-05,
+      "loss": 1.8616,
+      "step": 34790
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.7924624681472778,
+      "learning_rate": 1.2298387096774194e-05,
+      "loss": 1.9176,
+      "step": 34800
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 1.6568251848220825,
+      "learning_rate": 1.2273185483870968e-05,
+      "loss": 2.0337,
+      "step": 34810
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.9826894998550415,
+      "learning_rate": 1.2247983870967742e-05,
+      "loss": 1.5715,
+      "step": 34820
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 1.3292245864868164,
+      "learning_rate": 1.2222782258064517e-05,
+      "loss": 1.7236,
+      "step": 34830
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 1.042962908744812,
+      "learning_rate": 1.2197580645161291e-05,
+      "loss": 1.7,
+      "step": 34840
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 1.5502090454101562,
+      "learning_rate": 1.2172379032258065e-05,
+      "loss": 1.8384,
+      "step": 34850
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 1.0668740272521973,
+      "learning_rate": 1.214717741935484e-05,
+      "loss": 1.5952,
+      "step": 34860
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 1.5362801551818848,
+      "learning_rate": 1.2121975806451612e-05,
+      "loss": 1.9045,
+      "step": 34870
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 1.1394953727722168,
+      "learning_rate": 1.2096774193548388e-05,
+      "loss": 1.7307,
+      "step": 34880
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 1.6300349235534668,
+      "learning_rate": 1.2071572580645161e-05,
+      "loss": 1.7957,
+      "step": 34890
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 2.473341941833496,
+      "learning_rate": 1.2046370967741937e-05,
+      "loss": 1.7127,
+      "step": 34900
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 1.380901575088501,
+      "learning_rate": 1.202116935483871e-05,
+      "loss": 1.7472,
+      "step": 34910
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 1.917965292930603,
+      "learning_rate": 1.1995967741935484e-05,
+      "loss": 1.7928,
+      "step": 34920
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 1.4509689807891846,
+      "learning_rate": 1.1970766129032258e-05,
+      "loss": 1.9078,
+      "step": 34930
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 1.96388840675354,
+      "learning_rate": 1.1945564516129033e-05,
+      "loss": 1.6619,
+      "step": 34940
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 2.632293939590454,
+      "learning_rate": 1.1920362903225807e-05,
+      "loss": 1.965,
+      "step": 34950
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 1.4416882991790771,
+      "learning_rate": 1.1895161290322582e-05,
+      "loss": 1.9333,
+      "step": 34960
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 1.0484460592269897,
+      "learning_rate": 1.1869959677419356e-05,
+      "loss": 1.7658,
+      "step": 34970
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 1.7141164541244507,
+      "learning_rate": 1.184475806451613e-05,
+      "loss": 1.7916,
+      "step": 34980
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 1.300503134727478,
+      "learning_rate": 1.1819556451612903e-05,
+      "loss": 1.6761,
+      "step": 34990
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 1.0069376230239868,
+      "learning_rate": 1.1794354838709679e-05,
+      "loss": 1.8153,
+      "step": 35000
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 1.23341703414917,
+      "learning_rate": 1.1769153225806452e-05,
+      "loss": 1.6692,
+      "step": 35010
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 1.1277674436569214,
+      "learning_rate": 1.1743951612903228e-05,
+      "loss": 1.6775,
+      "step": 35020
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.9882980585098267,
+      "learning_rate": 1.171875e-05,
+      "loss": 1.7925,
+      "step": 35030
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 2.4044816493988037,
+      "learning_rate": 1.1693548387096775e-05,
+      "loss": 1.9954,
+      "step": 35040
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 1.0479285717010498,
+      "learning_rate": 1.1668346774193549e-05,
+      "loss": 1.7009,
+      "step": 35050
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 1.5229885578155518,
+      "learning_rate": 1.1643145161290323e-05,
+      "loss": 1.8682,
+      "step": 35060
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 1.1176376342773438,
+      "learning_rate": 1.1617943548387098e-05,
+      "loss": 1.767,
+      "step": 35070
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 1.025222659111023,
+      "learning_rate": 1.159274193548387e-05,
+      "loss": 1.7779,
+      "step": 35080
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 2.402141809463501,
+      "learning_rate": 1.1567540322580646e-05,
+      "loss": 1.9209,
+      "step": 35090
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.8719015121459961,
+      "learning_rate": 1.1542338709677419e-05,
+      "loss": 1.639,
+      "step": 35100
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 2.1905722618103027,
+      "learning_rate": 1.1517137096774195e-05,
+      "loss": 1.77,
+      "step": 35110
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 1.4088919162750244,
+      "learning_rate": 1.1491935483870968e-05,
+      "loss": 1.7391,
+      "step": 35120
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 2.011066436767578,
+      "learning_rate": 1.1466733870967742e-05,
+      "loss": 1.805,
+      "step": 35130
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 1.9606380462646484,
+      "learning_rate": 1.1441532258064516e-05,
+      "loss": 2.0182,
+      "step": 35140
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 1.146959662437439,
+      "learning_rate": 1.141633064516129e-05,
+      "loss": 1.521,
+      "step": 35150
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 1.4316011667251587,
+      "learning_rate": 1.1391129032258065e-05,
+      "loss": 1.8566,
+      "step": 35160
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 1.3860236406326294,
+      "learning_rate": 1.136592741935484e-05,
+      "loss": 1.8368,
+      "step": 35170
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.9776244163513184,
+      "learning_rate": 1.1340725806451614e-05,
+      "loss": 1.8574,
+      "step": 35180
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 1.602031946182251,
+      "learning_rate": 1.1315524193548388e-05,
+      "loss": 1.8379,
+      "step": 35190
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 2.176476001739502,
+      "learning_rate": 1.129032258064516e-05,
+      "loss": 1.7631,
+      "step": 35200
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.7637023329734802,
+      "learning_rate": 1.1265120967741937e-05,
+      "loss": 1.7881,
+      "step": 35210
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 1.7604823112487793,
+      "learning_rate": 1.123991935483871e-05,
+      "loss": 1.805,
+      "step": 35220
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 1.5338783264160156,
+      "learning_rate": 1.1214717741935485e-05,
+      "loss": 1.7743,
+      "step": 35230
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 1.2608662843704224,
+      "learning_rate": 1.1189516129032258e-05,
+      "loss": 1.9841,
+      "step": 35240
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 1.5797710418701172,
+      "learning_rate": 1.1164314516129032e-05,
+      "loss": 1.7794,
+      "step": 35250
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 1.2201260328292847,
+      "learning_rate": 1.1139112903225807e-05,
+      "loss": 1.7557,
+      "step": 35260
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 1.751137375831604,
+      "learning_rate": 1.1113911290322581e-05,
+      "loss": 1.8449,
+      "step": 35270
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 1.701958179473877,
+      "learning_rate": 1.1088709677419356e-05,
+      "loss": 1.7663,
+      "step": 35280
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 1.553699016571045,
+      "learning_rate": 1.106350806451613e-05,
+      "loss": 1.8567,
+      "step": 35290
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 1.120373010635376,
+      "learning_rate": 1.1038306451612904e-05,
+      "loss": 1.7395,
+      "step": 35300
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 1.928885579109192,
+      "learning_rate": 1.1013104838709679e-05,
+      "loss": 1.7393,
+      "step": 35310
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 1.292385220527649,
+      "learning_rate": 1.0987903225806453e-05,
+      "loss": 1.8386,
+      "step": 35320
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 2.046734094619751,
+      "learning_rate": 1.0962701612903227e-05,
+      "loss": 1.8397,
+      "step": 35330
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.4910139739513397,
+      "learning_rate": 1.09375e-05,
+      "loss": 1.9763,
+      "step": 35340
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 1.0717662572860718,
+      "learning_rate": 1.0912298387096774e-05,
+      "loss": 1.9355,
+      "step": 35350
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 1.052740216255188,
+      "learning_rate": 1.0887096774193549e-05,
+      "loss": 1.6994,
+      "step": 35360
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 1.3854340314865112,
+      "learning_rate": 1.0861895161290323e-05,
+      "loss": 1.9131,
+      "step": 35370
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 1.9061628580093384,
+      "learning_rate": 1.0836693548387097e-05,
+      "loss": 1.8026,
+      "step": 35380
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 2.028916597366333,
+      "learning_rate": 1.0811491935483872e-05,
+      "loss": 1.5724,
+      "step": 35390
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 1.1304678916931152,
+      "learning_rate": 1.0786290322580646e-05,
+      "loss": 1.8497,
+      "step": 35400
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.9788485765457153,
+      "learning_rate": 1.0761088709677419e-05,
+      "loss": 1.7039,
+      "step": 35410
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 1.7547781467437744,
+      "learning_rate": 1.0735887096774195e-05,
+      "loss": 1.6254,
+      "step": 35420
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 1.2983570098876953,
+      "learning_rate": 1.0710685483870967e-05,
+      "loss": 1.8797,
+      "step": 35430
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 1.3385576009750366,
+      "learning_rate": 1.0685483870967743e-05,
+      "loss": 1.919,
+      "step": 35440
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 1.7279398441314697,
+      "learning_rate": 1.0660282258064516e-05,
+      "loss": 1.8694,
+      "step": 35450
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 1.1627771854400635,
+      "learning_rate": 1.063508064516129e-05,
+      "loss": 1.8525,
+      "step": 35460
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.8045688271522522,
+      "learning_rate": 1.0609879032258065e-05,
+      "loss": 1.8093,
+      "step": 35470
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 1.3966808319091797,
+      "learning_rate": 1.0584677419354839e-05,
+      "loss": 1.8899,
+      "step": 35480
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 1.4931981563568115,
+      "learning_rate": 1.0559475806451613e-05,
+      "loss": 1.8492,
+      "step": 35490
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 1.7919448614120483,
+      "learning_rate": 1.0534274193548388e-05,
+      "loss": 1.5859,
+      "step": 35500
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 1.227303147315979,
+      "learning_rate": 1.0509072580645162e-05,
+      "loss": 1.7608,
+      "step": 35510
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 1.2445677518844604,
+      "learning_rate": 1.0483870967741936e-05,
+      "loss": 1.9111,
+      "step": 35520
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 1.3919745683670044,
+      "learning_rate": 1.0458669354838709e-05,
+      "loss": 1.8621,
+      "step": 35530
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.8560446500778198,
+      "learning_rate": 1.0433467741935485e-05,
+      "loss": 1.6429,
+      "step": 35540
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 1.521107792854309,
+      "learning_rate": 1.0408266129032258e-05,
+      "loss": 1.7876,
+      "step": 35550
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.8255922198295593,
+      "learning_rate": 1.0383064516129034e-05,
+      "loss": 1.8757,
+      "step": 35560
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 1.1279582977294922,
+      "learning_rate": 1.0357862903225806e-05,
+      "loss": 1.937,
+      "step": 35570
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 1.3345603942871094,
+      "learning_rate": 1.0332661290322582e-05,
+      "loss": 1.873,
+      "step": 35580
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 1.6812241077423096,
+      "learning_rate": 1.0307459677419355e-05,
+      "loss": 1.8823,
+      "step": 35590
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 1.4878910779953003,
+      "learning_rate": 1.028225806451613e-05,
+      "loss": 1.6257,
+      "step": 35600
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 1.904634952545166,
+      "learning_rate": 1.0257056451612904e-05,
+      "loss": 1.785,
+      "step": 35610
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 1.2773956060409546,
+      "learning_rate": 1.0231854838709678e-05,
+      "loss": 1.5099,
+      "step": 35620
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 1.3781163692474365,
+      "learning_rate": 1.0206653225806453e-05,
+      "loss": 1.7838,
+      "step": 35630
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 1.7493046522140503,
+      "learning_rate": 1.0181451612903227e-05,
+      "loss": 1.7883,
+      "step": 35640
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 2.6882827281951904,
+      "learning_rate": 1.0156250000000001e-05,
+      "loss": 1.9082,
+      "step": 35650
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 1.072803258895874,
+      "learning_rate": 1.0131048387096774e-05,
+      "loss": 1.8787,
+      "step": 35660
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 1.3693630695343018,
+      "learning_rate": 1.0105846774193548e-05,
+      "loss": 1.6688,
+      "step": 35670
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 1.5721242427825928,
+      "learning_rate": 1.0080645161290323e-05,
+      "loss": 1.8172,
+      "step": 35680
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 1.0625810623168945,
+      "learning_rate": 1.0055443548387097e-05,
+      "loss": 1.6947,
+      "step": 35690
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 1.7398951053619385,
+      "learning_rate": 1.0030241935483871e-05,
+      "loss": 1.7415,
+      "step": 35700
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 1.3869012594223022,
+      "learning_rate": 1.0005040322580646e-05,
+      "loss": 1.8786,
+      "step": 35710
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 1.5553916692733765,
+      "learning_rate": 9.97983870967742e-06,
+      "loss": 1.6325,
+      "step": 35720
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.8565185070037842,
+      "learning_rate": 9.954637096774194e-06,
+      "loss": 1.7055,
+      "step": 35730
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 1.391282320022583,
+      "learning_rate": 9.929435483870967e-06,
+      "loss": 1.8739,
+      "step": 35740
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 2.4944636821746826,
+      "learning_rate": 9.904233870967743e-06,
+      "loss": 1.7608,
+      "step": 35750
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 1.0780380964279175,
+      "learning_rate": 9.879032258064516e-06,
+      "loss": 1.7559,
+      "step": 35760
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 1.793933629989624,
+      "learning_rate": 9.853830645161292e-06,
+      "loss": 1.8143,
+      "step": 35770
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 2.271881103515625,
+      "learning_rate": 9.828629032258064e-06,
+      "loss": 1.9301,
+      "step": 35780
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 2.2172186374664307,
+      "learning_rate": 9.803427419354839e-06,
+      "loss": 1.8776,
+      "step": 35790
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 1.9969838857650757,
+      "learning_rate": 9.778225806451613e-06,
+      "loss": 1.7482,
+      "step": 35800
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 2.097424030303955,
+      "learning_rate": 9.753024193548387e-06,
+      "loss": 1.6255,
+      "step": 35810
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 1.60562002658844,
+      "learning_rate": 9.727822580645162e-06,
+      "loss": 1.768,
+      "step": 35820
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 1.2569961547851562,
+      "learning_rate": 9.702620967741936e-06,
+      "loss": 1.5886,
+      "step": 35830
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 1.2189213037490845,
+      "learning_rate": 9.67741935483871e-06,
+      "loss": 1.8046,
+      "step": 35840
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 2.0332515239715576,
+      "learning_rate": 9.652217741935485e-06,
+      "loss": 1.8813,
+      "step": 35850
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 1.4702340364456177,
+      "learning_rate": 9.627016129032259e-06,
+      "loss": 1.7985,
+      "step": 35860
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 1.1818456649780273,
+      "learning_rate": 9.601814516129033e-06,
+      "loss": 1.7162,
+      "step": 35870
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 1.4927242994308472,
+      "learning_rate": 9.576612903225806e-06,
+      "loss": 1.6686,
+      "step": 35880
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 1.3903846740722656,
+      "learning_rate": 9.551411290322582e-06,
+      "loss": 1.6926,
+      "step": 35890
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 1.0837092399597168,
+      "learning_rate": 9.526209677419355e-06,
+      "loss": 1.7957,
+      "step": 35900
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 1.2898942232131958,
+      "learning_rate": 9.50100806451613e-06,
+      "loss": 1.8192,
+      "step": 35910
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 1.9477015733718872,
+      "learning_rate": 9.475806451612903e-06,
+      "loss": 1.9362,
+      "step": 35920
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 1.1268330812454224,
+      "learning_rate": 9.450604838709678e-06,
+      "loss": 1.7059,
+      "step": 35930
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.9600113034248352,
+      "learning_rate": 9.425403225806452e-06,
+      "loss": 1.8895,
+      "step": 35940
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 1.5135043859481812,
+      "learning_rate": 9.400201612903227e-06,
+      "loss": 1.7892,
+      "step": 35950
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 2.054722547531128,
+      "learning_rate": 9.375000000000001e-06,
+      "loss": 1.8189,
+      "step": 35960
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 1.1068767309188843,
+      "learning_rate": 9.349798387096774e-06,
+      "loss": 1.9214,
+      "step": 35970
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 1.252519965171814,
+      "learning_rate": 9.32459677419355e-06,
+      "loss": 1.7266,
+      "step": 35980
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 1.448464274406433,
+      "learning_rate": 9.299395161290322e-06,
+      "loss": 1.8223,
+      "step": 35990
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 1.2505929470062256,
+      "learning_rate": 9.274193548387097e-06,
+      "loss": 1.9827,
+      "step": 36000
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 1.4828120470046997,
+      "learning_rate": 9.248991935483871e-06,
+      "loss": 1.6435,
+      "step": 36010
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 1.4445147514343262,
+      "learning_rate": 9.223790322580645e-06,
+      "loss": 1.8303,
+      "step": 36020
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 1.918289303779602,
+      "learning_rate": 9.19858870967742e-06,
+      "loss": 1.8411,
+      "step": 36030
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 2.476227045059204,
+      "learning_rate": 9.173387096774194e-06,
+      "loss": 1.7027,
+      "step": 36040
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 1.9552397727966309,
+      "learning_rate": 9.148185483870968e-06,
+      "loss": 1.7505,
+      "step": 36050
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 1.6305761337280273,
+      "learning_rate": 9.122983870967743e-06,
+      "loss": 1.7567,
+      "step": 36060
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 1.5752007961273193,
+      "learning_rate": 9.097782258064515e-06,
+      "loss": 1.5789,
+      "step": 36070
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 1.945370078086853,
+      "learning_rate": 9.072580645161291e-06,
+      "loss": 1.7359,
+      "step": 36080
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 1.4586658477783203,
+      "learning_rate": 9.047379032258064e-06,
+      "loss": 1.6781,
+      "step": 36090
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.9555315971374512,
+      "learning_rate": 9.02217741935484e-06,
+      "loss": 1.6118,
+      "step": 36100
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 1.493424415588379,
+      "learning_rate": 8.996975806451613e-06,
+      "loss": 1.6622,
+      "step": 36110
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 1.9182578325271606,
+      "learning_rate": 8.971774193548389e-06,
+      "loss": 1.7485,
+      "step": 36120
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 1.3957390785217285,
+      "learning_rate": 8.946572580645161e-06,
+      "loss": 1.7619,
+      "step": 36130
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.7602506279945374,
+      "learning_rate": 8.921370967741936e-06,
+      "loss": 2.0142,
+      "step": 36140
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.9782031774520874,
+      "learning_rate": 8.89616935483871e-06,
+      "loss": 1.7006,
+      "step": 36150
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 2.190155267715454,
+      "learning_rate": 8.870967741935484e-06,
+      "loss": 1.6205,
+      "step": 36160
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 2.0813379287719727,
+      "learning_rate": 8.845766129032259e-06,
+      "loss": 1.8232,
+      "step": 36170
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 1.2089934349060059,
+      "learning_rate": 8.820564516129033e-06,
+      "loss": 1.831,
+      "step": 36180
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 2.6575751304626465,
+      "learning_rate": 8.795362903225807e-06,
+      "loss": 1.681,
+      "step": 36190
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 1.7982875108718872,
+      "learning_rate": 8.770161290322582e-06,
+      "loss": 1.7288,
+      "step": 36200
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 1.4474347829818726,
+      "learning_rate": 8.744959677419354e-06,
+      "loss": 1.8051,
+      "step": 36210
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 1.172349214553833,
+      "learning_rate": 8.71975806451613e-06,
+      "loss": 1.9331,
+      "step": 36220
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 1.9786067008972168,
+      "learning_rate": 8.694556451612903e-06,
+      "loss": 1.7795,
+      "step": 36230
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 2.8013007640838623,
+      "learning_rate": 8.66935483870968e-06,
+      "loss": 1.7418,
+      "step": 36240
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 1.266817569732666,
+      "learning_rate": 8.644153225806452e-06,
+      "loss": 1.8284,
+      "step": 36250
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 1.4410595893859863,
+      "learning_rate": 8.618951612903226e-06,
+      "loss": 1.7999,
+      "step": 36260
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 1.1499738693237305,
+      "learning_rate": 8.59375e-06,
+      "loss": 1.8118,
+      "step": 36270
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 2.7630200386047363,
+      "learning_rate": 8.568548387096773e-06,
+      "loss": 1.7198,
+      "step": 36280
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 1.106684923171997,
+      "learning_rate": 8.54334677419355e-06,
+      "loss": 1.8368,
+      "step": 36290
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 2.610393524169922,
+      "learning_rate": 8.518145161290322e-06,
+      "loss": 1.7734,
+      "step": 36300
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 1.5953667163848877,
+      "learning_rate": 8.492943548387098e-06,
+      "loss": 1.9071,
+      "step": 36310
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 1.6072356700897217,
+      "learning_rate": 8.46774193548387e-06,
+      "loss": 1.888,
+      "step": 36320
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 1.075134038925171,
+      "learning_rate": 8.442540322580645e-06,
+      "loss": 1.7695,
+      "step": 36330
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 1.769437313079834,
+      "learning_rate": 8.41733870967742e-06,
+      "loss": 1.6756,
+      "step": 36340
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.9160236716270447,
+      "learning_rate": 8.392137096774194e-06,
+      "loss": 1.8199,
+      "step": 36350
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.872322142124176,
+      "learning_rate": 8.366935483870968e-06,
+      "loss": 1.8031,
+      "step": 36360
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 1.4082344770431519,
+      "learning_rate": 8.341733870967742e-06,
+      "loss": 1.7093,
+      "step": 36370
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.8456238508224487,
+      "learning_rate": 8.316532258064517e-06,
+      "loss": 1.7884,
+      "step": 36380
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 1.2698601484298706,
+      "learning_rate": 8.291330645161291e-06,
+      "loss": 1.8131,
+      "step": 36390
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 1.5412002801895142,
+      "learning_rate": 8.266129032258065e-06,
+      "loss": 1.952,
+      "step": 36400
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 1.0240978002548218,
+      "learning_rate": 8.24092741935484e-06,
+      "loss": 1.8331,
+      "step": 36410
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 1.399126648902893,
+      "learning_rate": 8.215725806451612e-06,
+      "loss": 1.6693,
+      "step": 36420
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 2.017927885055542,
+      "learning_rate": 8.190524193548388e-06,
+      "loss": 1.7508,
+      "step": 36430
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 1.5599889755249023,
+      "learning_rate": 8.165322580645161e-06,
+      "loss": 1.7579,
+      "step": 36440
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 1.1271214485168457,
+      "learning_rate": 8.140120967741937e-06,
+      "loss": 1.8408,
+      "step": 36450
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 1.6658759117126465,
+      "learning_rate": 8.11491935483871e-06,
+      "loss": 1.8079,
+      "step": 36460
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6939863562583923,
+      "learning_rate": 8.089717741935484e-06,
+      "loss": 1.6854,
+      "step": 36470
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 1.2951539754867554,
+      "learning_rate": 8.064516129032258e-06,
+      "loss": 1.9531,
+      "step": 36480
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 1.5003094673156738,
+      "learning_rate": 8.039314516129033e-06,
+      "loss": 1.8732,
+      "step": 36490
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 1.4087048768997192,
+      "learning_rate": 8.014112903225807e-06,
+      "loss": 1.7745,
+      "step": 36500
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 1.7704477310180664,
+      "learning_rate": 7.988911290322581e-06,
+      "loss": 1.6562,
+      "step": 36510
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 1.504568338394165,
+      "learning_rate": 7.963709677419356e-06,
+      "loss": 1.8344,
+      "step": 36520
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 1.4398772716522217,
+      "learning_rate": 7.93850806451613e-06,
+      "loss": 1.8407,
+      "step": 36530
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 1.8025132417678833,
+      "learning_rate": 7.913306451612903e-06,
+      "loss": 1.8964,
+      "step": 36540
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 1.3510351181030273,
+      "learning_rate": 7.888104838709679e-06,
+      "loss": 1.8342,
+      "step": 36550
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 1.6046425104141235,
+      "learning_rate": 7.862903225806451e-06,
+      "loss": 1.8667,
+      "step": 36560
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 1.5016286373138428,
+      "learning_rate": 7.837701612903228e-06,
+      "loss": 2.0234,
+      "step": 36570
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 1.1071223020553589,
+      "learning_rate": 7.8125e-06,
+      "loss": 1.9305,
+      "step": 36580
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 1.8191975355148315,
+      "learning_rate": 7.787298387096775e-06,
+      "loss": 1.7867,
+      "step": 36590
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.8355802893638611,
+      "learning_rate": 7.762096774193549e-06,
+      "loss": 1.6907,
+      "step": 36600
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 1.0143616199493408,
+      "learning_rate": 7.736895161290322e-06,
+      "loss": 1.7817,
+      "step": 36610
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 2.000938653945923,
+      "learning_rate": 7.711693548387098e-06,
+      "loss": 1.7698,
+      "step": 36620
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 2.525449514389038,
+      "learning_rate": 7.68649193548387e-06,
+      "loss": 1.7364,
+      "step": 36630
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 1.5523021221160889,
+      "learning_rate": 7.661290322580646e-06,
+      "loss": 1.9159,
+      "step": 36640
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 1.0402398109436035,
+      "learning_rate": 7.636088709677419e-06,
+      "loss": 1.6745,
+      "step": 36650
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 1.3228315114974976,
+      "learning_rate": 7.610887096774194e-06,
+      "loss": 1.9929,
+      "step": 36660
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 1.0684540271759033,
+      "learning_rate": 7.585685483870968e-06,
+      "loss": 1.7958,
+      "step": 36670
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 1.8278487920761108,
+      "learning_rate": 7.560483870967743e-06,
+      "loss": 1.4693,
+      "step": 36680
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 2.108124256134033,
+      "learning_rate": 7.535282258064516e-06,
+      "loss": 1.7322,
+      "step": 36690
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 1.3348920345306396,
+      "learning_rate": 7.510080645161291e-06,
+      "loss": 1.7468,
+      "step": 36700
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 1.3507848978042603,
+      "learning_rate": 7.484879032258064e-06,
+      "loss": 2.003,
+      "step": 36710
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 1.3587337732315063,
+      "learning_rate": 7.459677419354839e-06,
+      "loss": 1.8606,
+      "step": 36720
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 1.4475183486938477,
+      "learning_rate": 7.434475806451613e-06,
+      "loss": 1.9709,
+      "step": 36730
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 1.194683313369751,
+      "learning_rate": 7.409274193548388e-06,
+      "loss": 1.8704,
+      "step": 36740
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 1.4805752038955688,
+      "learning_rate": 7.3840725806451615e-06,
+      "loss": 1.7091,
+      "step": 36750
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 1.7947769165039062,
+      "learning_rate": 7.358870967741936e-06,
+      "loss": 1.6898,
+      "step": 36760
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 1.6085330247879028,
+      "learning_rate": 7.333669354838709e-06,
+      "loss": 1.9532,
+      "step": 36770
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 2.163475513458252,
+      "learning_rate": 7.3084677419354845e-06,
+      "loss": 1.7973,
+      "step": 36780
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 1.1183041334152222,
+      "learning_rate": 7.283266129032258e-06,
+      "loss": 1.6807,
+      "step": 36790
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.9880658388137817,
+      "learning_rate": 7.258064516129033e-06,
+      "loss": 1.7639,
+      "step": 36800
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 1.0951911211013794,
+      "learning_rate": 7.232862903225807e-06,
+      "loss": 1.8557,
+      "step": 36810
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 2.3325161933898926,
+      "learning_rate": 7.207661290322582e-06,
+      "loss": 1.8977,
+      "step": 36820
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 1.5593726634979248,
+      "learning_rate": 7.1824596774193554e-06,
+      "loss": 1.66,
+      "step": 36830
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 1.6101003885269165,
+      "learning_rate": 7.15725806451613e-06,
+      "loss": 1.7928,
+      "step": 36840
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 1.2023204565048218,
+      "learning_rate": 7.132056451612903e-06,
+      "loss": 1.8065,
+      "step": 36850
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 2.0083374977111816,
+      "learning_rate": 7.1068548387096785e-06,
+      "loss": 1.7986,
+      "step": 36860
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.9798304438591003,
+      "learning_rate": 7.081653225806452e-06,
+      "loss": 1.6628,
+      "step": 36870
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 1.196536660194397,
+      "learning_rate": 7.056451612903227e-06,
+      "loss": 1.8686,
+      "step": 36880
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 1.310107707977295,
+      "learning_rate": 7.031250000000001e-06,
+      "loss": 1.8639,
+      "step": 36890
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 1.5220752954483032,
+      "learning_rate": 7.006048387096774e-06,
+      "loss": 1.6612,
+      "step": 36900
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 1.0612527132034302,
+      "learning_rate": 6.9808467741935485e-06,
+      "loss": 1.7246,
+      "step": 36910
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 2.5143191814422607,
+      "learning_rate": 6.955645161290322e-06,
+      "loss": 1.7949,
+      "step": 36920
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 1.4156107902526855,
+      "learning_rate": 6.930443548387097e-06,
+      "loss": 1.7527,
+      "step": 36930
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 1.790284514427185,
+      "learning_rate": 6.905241935483871e-06,
+      "loss": 1.768,
+      "step": 36940
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 1.3119839429855347,
+      "learning_rate": 6.880040322580646e-06,
+      "loss": 1.8595,
+      "step": 36950
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 1.4229341745376587,
+      "learning_rate": 6.854838709677419e-06,
+      "loss": 1.998,
+      "step": 36960
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 1.3340529203414917,
+      "learning_rate": 6.829637096774194e-06,
+      "loss": 1.8822,
+      "step": 36970
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 1.102401614189148,
+      "learning_rate": 6.804435483870967e-06,
+      "loss": 1.8588,
+      "step": 36980
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 1.8553776741027832,
+      "learning_rate": 6.779233870967742e-06,
+      "loss": 1.7541,
+      "step": 36990
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 1.1877295970916748,
+      "learning_rate": 6.754032258064516e-06,
+      "loss": 1.7535,
+      "step": 37000
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 2.1015360355377197,
+      "learning_rate": 6.728830645161291e-06,
+      "loss": 1.7017,
+      "step": 37010
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 2.259993076324463,
+      "learning_rate": 6.703629032258065e-06,
+      "loss": 1.5471,
+      "step": 37020
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 1.4479036331176758,
+      "learning_rate": 6.678427419354839e-06,
+      "loss": 1.6228,
+      "step": 37030
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 1.9566433429718018,
+      "learning_rate": 6.6532258064516125e-06,
+      "loss": 1.8098,
+      "step": 37040
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 1.791641116142273,
+      "learning_rate": 6.628024193548388e-06,
+      "loss": 1.6533,
+      "step": 37050
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 1.325700044631958,
+      "learning_rate": 6.602822580645161e-06,
+      "loss": 1.8519,
+      "step": 37060
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 1.2213832139968872,
+      "learning_rate": 6.577620967741936e-06,
+      "loss": 1.7053,
+      "step": 37070
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 1.1624809503555298,
+      "learning_rate": 6.55241935483871e-06,
+      "loss": 1.5668,
+      "step": 37080
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 1.437555193901062,
+      "learning_rate": 6.527217741935485e-06,
+      "loss": 1.8231,
+      "step": 37090
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 1.8645172119140625,
+      "learning_rate": 6.5020161290322585e-06,
+      "loss": 1.6101,
+      "step": 37100
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 1.8789284229278564,
+      "learning_rate": 6.476814516129033e-06,
+      "loss": 1.6905,
+      "step": 37110
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 1.3216228485107422,
+      "learning_rate": 6.451612903225806e-06,
+      "loss": 1.5535,
+      "step": 37120
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 1.1430431604385376,
+      "learning_rate": 6.4264112903225816e-06,
+      "loss": 1.9474,
+      "step": 37130
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 1.148832082748413,
+      "learning_rate": 6.401209677419355e-06,
+      "loss": 1.7387,
+      "step": 37140
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 1.4879958629608154,
+      "learning_rate": 6.37600806451613e-06,
+      "loss": 1.847,
+      "step": 37150
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 1.6767230033874512,
+      "learning_rate": 6.350806451612904e-06,
+      "loss": 1.74,
+      "step": 37160
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 1.463371992111206,
+      "learning_rate": 6.325604838709678e-06,
+      "loss": 1.8446,
+      "step": 37170
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 1.1493008136749268,
+      "learning_rate": 6.300403225806452e-06,
+      "loss": 1.877,
+      "step": 37180
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 1.4773237705230713,
+      "learning_rate": 6.275201612903227e-06,
+      "loss": 1.7639,
+      "step": 37190
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 2.9225573539733887,
+      "learning_rate": 6.25e-06,
+      "loss": 1.8418,
+      "step": 37200
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 1.6798039674758911,
+      "learning_rate": 6.224798387096775e-06,
+      "loss": 1.8091,
+      "step": 37210
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 1.2624444961547852,
+      "learning_rate": 6.199596774193549e-06,
+      "loss": 1.8965,
+      "step": 37220
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 1.175840139389038,
+      "learning_rate": 6.174395161290323e-06,
+      "loss": 1.694,
+      "step": 37230
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 1.8195207118988037,
+      "learning_rate": 6.149193548387097e-06,
+      "loss": 1.6902,
+      "step": 37240
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.8336354494094849,
+      "learning_rate": 6.123991935483871e-06,
+      "loss": 1.7952,
+      "step": 37250
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 1.3767187595367432,
+      "learning_rate": 6.0987903225806455e-06,
+      "loss": 1.9925,
+      "step": 37260
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 1.8563657999038696,
+      "learning_rate": 6.07358870967742e-06,
+      "loss": 2.0029,
+      "step": 37270
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 1.2369264364242554,
+      "learning_rate": 6.048387096774194e-06,
+      "loss": 1.919,
+      "step": 37280
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6002053618431091,
+      "learning_rate": 6.0231854838709686e-06,
+      "loss": 1.6513,
+      "step": 37290
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 1.3652684688568115,
+      "learning_rate": 5.997983870967742e-06,
+      "loss": 1.9505,
+      "step": 37300
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 2.4589905738830566,
+      "learning_rate": 5.972782258064516e-06,
+      "loss": 1.8757,
+      "step": 37310
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 1.2345190048217773,
+      "learning_rate": 5.947580645161291e-06,
+      "loss": 1.8905,
+      "step": 37320
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 1.2701138257980347,
+      "learning_rate": 5.922379032258065e-06,
+      "loss": 1.9293,
+      "step": 37330
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.7601884603500366,
+      "learning_rate": 5.8971774193548394e-06,
+      "loss": 1.6258,
+      "step": 37340
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 1.1152560710906982,
+      "learning_rate": 5.871975806451614e-06,
+      "loss": 1.8248,
+      "step": 37350
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 1.326200008392334,
+      "learning_rate": 5.846774193548387e-06,
+      "loss": 1.8016,
+      "step": 37360
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.8782685995101929,
+      "learning_rate": 5.821572580645162e-06,
+      "loss": 1.6993,
+      "step": 37370
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 1.4969420433044434,
+      "learning_rate": 5.796370967741935e-06,
+      "loss": 1.9912,
+      "step": 37380
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 1.578978180885315,
+      "learning_rate": 5.7711693548387095e-06,
+      "loss": 1.8695,
+      "step": 37390
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 2.348956346511841,
+      "learning_rate": 5.745967741935484e-06,
+      "loss": 1.71,
+      "step": 37400
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 1.2607457637786865,
+      "learning_rate": 5.720766129032258e-06,
+      "loss": 1.8087,
+      "step": 37410
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 1.4012140035629272,
+      "learning_rate": 5.6955645161290325e-06,
+      "loss": 1.7404,
+      "step": 37420
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 1.997958779335022,
+      "learning_rate": 5.670362903225807e-06,
+      "loss": 1.8238,
+      "step": 37430
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 1.1396044492721558,
+      "learning_rate": 5.64516129032258e-06,
+      "loss": 1.7785,
+      "step": 37440
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 1.2574045658111572,
+      "learning_rate": 5.619959677419355e-06,
+      "loss": 1.7567,
+      "step": 37450
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 1.3726853132247925,
+      "learning_rate": 5.594758064516129e-06,
+      "loss": 1.7628,
+      "step": 37460
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 1.1301125288009644,
+      "learning_rate": 5.569556451612903e-06,
+      "loss": 1.7471,
+      "step": 37470
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 2.332216739654541,
+      "learning_rate": 5.544354838709678e-06,
+      "loss": 1.9294,
+      "step": 37480
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 1.2760249376296997,
+      "learning_rate": 5.519153225806452e-06,
+      "loss": 1.8431,
+      "step": 37490
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 1.4540157318115234,
+      "learning_rate": 5.4939516129032264e-06,
+      "loss": 1.9109,
+      "step": 37500
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 1.5847976207733154,
+      "learning_rate": 5.46875e-06,
+      "loss": 1.7637,
+      "step": 37510
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 2.0198564529418945,
+      "learning_rate": 5.443548387096774e-06,
+      "loss": 1.5915,
+      "step": 37520
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 1.8017363548278809,
+      "learning_rate": 5.418346774193549e-06,
+      "loss": 1.6431,
+      "step": 37530
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 1.3348603248596191,
+      "learning_rate": 5.393145161290323e-06,
+      "loss": 1.8796,
+      "step": 37540
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 1.01182222366333,
+      "learning_rate": 5.367943548387097e-06,
+      "loss": 1.7648,
+      "step": 37550
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 1.5626134872436523,
+      "learning_rate": 5.342741935483872e-06,
+      "loss": 1.7978,
+      "step": 37560
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 1.8154191970825195,
+      "learning_rate": 5.317540322580645e-06,
+      "loss": 1.8132,
+      "step": 37570
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 1.1710180044174194,
+      "learning_rate": 5.2923387096774195e-06,
+      "loss": 1.8187,
+      "step": 37580
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 1.8511433601379395,
+      "learning_rate": 5.267137096774194e-06,
+      "loss": 1.5787,
+      "step": 37590
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.7542625069618225,
+      "learning_rate": 5.241935483870968e-06,
+      "loss": 1.6165,
+      "step": 37600
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 1.416358470916748,
+      "learning_rate": 5.2167338709677426e-06,
+      "loss": 1.7577,
+      "step": 37610
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 1.2438840866088867,
+      "learning_rate": 5.191532258064517e-06,
+      "loss": 1.8492,
+      "step": 37620
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.8957983255386353,
+      "learning_rate": 5.166330645161291e-06,
+      "loss": 1.889,
+      "step": 37630
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 1.3180512189865112,
+      "learning_rate": 5.141129032258065e-06,
+      "loss": 1.6732,
+      "step": 37640
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 2.0026447772979736,
+      "learning_rate": 5.115927419354839e-06,
+      "loss": 1.8613,
+      "step": 37650
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 1.6598964929580688,
+      "learning_rate": 5.0907258064516134e-06,
+      "loss": 1.9261,
+      "step": 37660
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 2.0712811946868896,
+      "learning_rate": 5.065524193548387e-06,
+      "loss": 1.8648,
+      "step": 37670
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 1.3434923887252808,
+      "learning_rate": 5.040322580645161e-06,
+      "loss": 1.785,
+      "step": 37680
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 1.466409683227539,
+      "learning_rate": 5.015120967741936e-06,
+      "loss": 1.8444,
+      "step": 37690
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 1.2131131887435913,
+      "learning_rate": 4.98991935483871e-06,
+      "loss": 1.7155,
+      "step": 37700
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 1.1753900051116943,
+      "learning_rate": 4.9647177419354835e-06,
+      "loss": 1.6853,
+      "step": 37710
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.8627914786338806,
+      "learning_rate": 4.939516129032258e-06,
+      "loss": 1.7129,
+      "step": 37720
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 1.4458900690078735,
+      "learning_rate": 4.914314516129032e-06,
+      "loss": 1.7731,
+      "step": 37730
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 1.7337548732757568,
+      "learning_rate": 4.8891129032258065e-06,
+      "loss": 1.778,
+      "step": 37740
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 2.060189962387085,
+      "learning_rate": 4.863911290322581e-06,
+      "loss": 1.865,
+      "step": 37750
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 1.0463095903396606,
+      "learning_rate": 4.838709677419355e-06,
+      "loss": 1.9132,
+      "step": 37760
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 1.209923267364502,
+      "learning_rate": 4.8135080645161296e-06,
+      "loss": 1.7437,
+      "step": 37770
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 1.4443707466125488,
+      "learning_rate": 4.788306451612903e-06,
+      "loss": 1.6754,
+      "step": 37780
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 1.6023110151290894,
+      "learning_rate": 4.763104838709677e-06,
+      "loss": 1.677,
+      "step": 37790
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 2.3413848876953125,
+      "learning_rate": 4.737903225806452e-06,
+      "loss": 1.8379,
+      "step": 37800
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 1.0863847732543945,
+      "learning_rate": 4.712701612903226e-06,
+      "loss": 1.7077,
+      "step": 37810
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 1.1071550846099854,
+      "learning_rate": 4.6875000000000004e-06,
+      "loss": 1.849,
+      "step": 37820
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 1.0773676633834839,
+      "learning_rate": 4.662298387096775e-06,
+      "loss": 1.6585,
+      "step": 37830
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 3.0338149070739746,
+      "learning_rate": 4.637096774193548e-06,
+      "loss": 1.78,
+      "step": 37840
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 1.8845102787017822,
+      "learning_rate": 4.611895161290323e-06,
+      "loss": 1.711,
+      "step": 37850
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 1.687903881072998,
+      "learning_rate": 4.586693548387097e-06,
+      "loss": 1.9214,
+      "step": 37860
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.8952406644821167,
+      "learning_rate": 4.561491935483871e-06,
+      "loss": 1.752,
+      "step": 37870
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 1.2971062660217285,
+      "learning_rate": 4.536290322580646e-06,
+      "loss": 1.6084,
+      "step": 37880
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 1.5500911474227905,
+      "learning_rate": 4.51108870967742e-06,
+      "loss": 1.8136,
+      "step": 37890
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 1.7014816999435425,
+      "learning_rate": 4.485887096774194e-06,
+      "loss": 1.8528,
+      "step": 37900
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 1.020188570022583,
+      "learning_rate": 4.460685483870968e-06,
+      "loss": 1.8505,
+      "step": 37910
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 1.702589750289917,
+      "learning_rate": 4.435483870967742e-06,
+      "loss": 1.8594,
+      "step": 37920
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 1.402048110961914,
+      "learning_rate": 4.4102822580645165e-06,
+      "loss": 2.0246,
+      "step": 37930
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 1.1861649751663208,
+      "learning_rate": 4.385080645161291e-06,
+      "loss": 1.6676,
+      "step": 37940
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.9101788997650146,
+      "learning_rate": 4.359879032258065e-06,
+      "loss": 1.7673,
+      "step": 37950
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 1.7470577955245972,
+      "learning_rate": 4.33467741935484e-06,
+      "loss": 1.9607,
+      "step": 37960
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 1.2408236265182495,
+      "learning_rate": 4.309475806451613e-06,
+      "loss": 1.8391,
+      "step": 37970
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 2.3454596996307373,
+      "learning_rate": 4.284274193548387e-06,
+      "loss": 1.8267,
+      "step": 37980
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.9975641369819641,
+      "learning_rate": 4.259072580645161e-06,
+      "loss": 1.6542,
+      "step": 37990
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.9366222023963928,
+      "learning_rate": 4.233870967741935e-06,
+      "loss": 1.8796,
+      "step": 38000
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 1.5389117002487183,
+      "learning_rate": 4.20866935483871e-06,
+      "loss": 1.9487,
+      "step": 38010
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 1.8198860883712769,
+      "learning_rate": 4.183467741935484e-06,
+      "loss": 1.8709,
+      "step": 38020
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 1.703210711479187,
+      "learning_rate": 4.158266129032258e-06,
+      "loss": 1.7685,
+      "step": 38030
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 1.4499989748001099,
+      "learning_rate": 4.133064516129033e-06,
+      "loss": 1.8361,
+      "step": 38040
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 2.043729782104492,
+      "learning_rate": 4.107862903225806e-06,
+      "loss": 1.8597,
+      "step": 38050
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 1.5115453004837036,
+      "learning_rate": 4.0826612903225805e-06,
+      "loss": 1.6677,
+      "step": 38060
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 1.227342128753662,
+      "learning_rate": 4.057459677419355e-06,
+      "loss": 1.7004,
+      "step": 38070
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.7584723830223083,
+      "learning_rate": 4.032258064516129e-06,
+      "loss": 1.8977,
+      "step": 38080
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 1.9289038181304932,
+      "learning_rate": 4.0070564516129035e-06,
+      "loss": 1.8065,
+      "step": 38090
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 1.7415162324905396,
+      "learning_rate": 3.981854838709678e-06,
+      "loss": 1.8769,
+      "step": 38100
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.8073474168777466,
+      "learning_rate": 3.956653225806451e-06,
+      "loss": 1.8978,
+      "step": 38110
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 1.5620120763778687,
+      "learning_rate": 3.931451612903226e-06,
+      "loss": 1.8361,
+      "step": 38120
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.8711804151535034,
+      "learning_rate": 3.90625e-06,
+      "loss": 1.8237,
+      "step": 38130
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 1.2821331024169922,
+      "learning_rate": 3.8810483870967744e-06,
+      "loss": 1.6756,
+      "step": 38140
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 1.9449238777160645,
+      "learning_rate": 3.855846774193549e-06,
+      "loss": 1.7867,
+      "step": 38150
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 1.1096111536026,
+      "learning_rate": 3.830645161290323e-06,
+      "loss": 1.6618,
+      "step": 38160
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 1.000383973121643,
+      "learning_rate": 3.805443548387097e-06,
+      "loss": 1.631,
+      "step": 38170
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 1.5528301000595093,
+      "learning_rate": 3.7802419354838714e-06,
+      "loss": 1.8168,
+      "step": 38180
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 1.3207968473434448,
+      "learning_rate": 3.7550403225806453e-06,
+      "loss": 1.8484,
+      "step": 38190
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 2.739356279373169,
+      "learning_rate": 3.7298387096774197e-06,
+      "loss": 1.7197,
+      "step": 38200
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 1.0143425464630127,
+      "learning_rate": 3.704637096774194e-06,
+      "loss": 1.652,
+      "step": 38210
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 1.5897810459136963,
+      "learning_rate": 3.679435483870968e-06,
+      "loss": 1.7126,
+      "step": 38220
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 1.3528326749801636,
+      "learning_rate": 3.6542338709677423e-06,
+      "loss": 1.6726,
+      "step": 38230
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.9339377880096436,
+      "learning_rate": 3.6290322580645166e-06,
+      "loss": 1.7668,
+      "step": 38240
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 2.340153694152832,
+      "learning_rate": 3.603830645161291e-06,
+      "loss": 1.8336,
+      "step": 38250
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 1.5839781761169434,
+      "learning_rate": 3.578629032258065e-06,
+      "loss": 1.8165,
+      "step": 38260
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 1.4641532897949219,
+      "learning_rate": 3.5534274193548392e-06,
+      "loss": 1.9005,
+      "step": 38270
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 2.190964937210083,
+      "learning_rate": 3.5282258064516136e-06,
+      "loss": 1.8835,
+      "step": 38280
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.9201270937919617,
+      "learning_rate": 3.503024193548387e-06,
+      "loss": 1.8417,
+      "step": 38290
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 1.4378013610839844,
+      "learning_rate": 3.477822580645161e-06,
+      "loss": 1.7118,
+      "step": 38300
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 2.6787805557250977,
+      "learning_rate": 3.4526209677419353e-06,
+      "loss": 1.8349,
+      "step": 38310
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 2.7822914123535156,
+      "learning_rate": 3.4274193548387097e-06,
+      "loss": 1.7964,
+      "step": 38320
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 1.4322879314422607,
+      "learning_rate": 3.4022177419354836e-06,
+      "loss": 1.5893,
+      "step": 38330
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 1.8684576749801636,
+      "learning_rate": 3.377016129032258e-06,
+      "loss": 1.8656,
+      "step": 38340
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 2.3487157821655273,
+      "learning_rate": 3.3518145161290323e-06,
+      "loss": 1.7828,
+      "step": 38350
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 1.944495439529419,
+      "learning_rate": 3.3266129032258062e-06,
+      "loss": 1.6986,
+      "step": 38360
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 1.067310094833374,
+      "learning_rate": 3.3014112903225806e-06,
+      "loss": 1.6193,
+      "step": 38370
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 2.7960333824157715,
+      "learning_rate": 3.276209677419355e-06,
+      "loss": 1.6732,
+      "step": 38380
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 1.2645069360733032,
+      "learning_rate": 3.2510080645161293e-06,
+      "loss": 1.7426,
+      "step": 38390
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 1.5527606010437012,
+      "learning_rate": 3.225806451612903e-06,
+      "loss": 1.6041,
+      "step": 38400
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 1.5937730073928833,
+      "learning_rate": 3.2006048387096775e-06,
+      "loss": 1.7343,
+      "step": 38410
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 1.7301063537597656,
+      "learning_rate": 3.175403225806452e-06,
+      "loss": 2.007,
+      "step": 38420
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 1.2938812971115112,
+      "learning_rate": 3.150201612903226e-06,
+      "loss": 1.6336,
+      "step": 38430
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 1.4524860382080078,
+      "learning_rate": 3.125e-06,
+      "loss": 1.6024,
+      "step": 38440
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 1.1089211702346802,
+      "learning_rate": 3.0997983870967745e-06,
+      "loss": 1.8917,
+      "step": 38450
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 1.6690031290054321,
+      "learning_rate": 3.0745967741935484e-06,
+      "loss": 1.6557,
+      "step": 38460
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 2.6736807823181152,
+      "learning_rate": 3.0493951612903228e-06,
+      "loss": 1.8295,
+      "step": 38470
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 1.427099585533142,
+      "learning_rate": 3.024193548387097e-06,
+      "loss": 1.7876,
+      "step": 38480
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.8637118935585022,
+      "learning_rate": 2.998991935483871e-06,
+      "loss": 1.7875,
+      "step": 38490
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 1.8728184700012207,
+      "learning_rate": 2.9737903225806454e-06,
+      "loss": 1.8437,
+      "step": 38500
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.9280319809913635,
+      "learning_rate": 2.9485887096774197e-06,
+      "loss": 1.6795,
+      "step": 38510
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 2.361722230911255,
+      "learning_rate": 2.9233870967741936e-06,
+      "loss": 1.8155,
+      "step": 38520
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 1.5140773057937622,
+      "learning_rate": 2.8981854838709676e-06,
+      "loss": 1.5892,
+      "step": 38530
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 1.9161895513534546,
+      "learning_rate": 2.872983870967742e-06,
+      "loss": 1.7801,
+      "step": 38540
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 1.5101937055587769,
+      "learning_rate": 2.8477822580645163e-06,
+      "loss": 1.8171,
+      "step": 38550
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 1.0115931034088135,
+      "learning_rate": 2.82258064516129e-06,
+      "loss": 1.7753,
+      "step": 38560
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 1.390844702720642,
+      "learning_rate": 2.7973790322580645e-06,
+      "loss": 1.9567,
+      "step": 38570
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 2.4053232669830322,
+      "learning_rate": 2.772177419354839e-06,
+      "loss": 1.9312,
+      "step": 38580
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.9232297539710999,
+      "learning_rate": 2.7469758064516132e-06,
+      "loss": 1.8046,
+      "step": 38590
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 1.0634303092956543,
+      "learning_rate": 2.721774193548387e-06,
+      "loss": 1.4786,
+      "step": 38600
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.8774827122688293,
+      "learning_rate": 2.6965725806451615e-06,
+      "loss": 1.7909,
+      "step": 38610
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 1.0526174306869507,
+      "learning_rate": 2.671370967741936e-06,
+      "loss": 1.7745,
+      "step": 38620
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 1.3172962665557861,
+      "learning_rate": 2.6461693548387098e-06,
+      "loss": 1.9997,
+      "step": 38630
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.9352841377258301,
+      "learning_rate": 2.620967741935484e-06,
+      "loss": 1.8178,
+      "step": 38640
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 2.200268268585205,
+      "learning_rate": 2.5957661290322585e-06,
+      "loss": 1.6865,
+      "step": 38650
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 2.4044673442840576,
+      "learning_rate": 2.5705645161290324e-06,
+      "loss": 1.8282,
+      "step": 38660
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 1.8035216331481934,
+      "learning_rate": 2.5453629032258067e-06,
+      "loss": 1.7758,
+      "step": 38670
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 1.3893768787384033,
+      "learning_rate": 2.5201612903225806e-06,
+      "loss": 1.8211,
+      "step": 38680
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 1.5243866443634033,
+      "learning_rate": 2.494959677419355e-06,
+      "loss": 1.5312,
+      "step": 38690
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 1.3379844427108765,
+      "learning_rate": 2.469758064516129e-06,
+      "loss": 1.9502,
+      "step": 38700
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 1.233444094657898,
+      "learning_rate": 2.4445564516129033e-06,
+      "loss": 1.5596,
+      "step": 38710
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 1.2969640493392944,
+      "learning_rate": 2.4193548387096776e-06,
+      "loss": 1.8371,
+      "step": 38720
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 1.3090646266937256,
+      "learning_rate": 2.3941532258064515e-06,
+      "loss": 1.8266,
+      "step": 38730
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 1.2111945152282715,
+      "learning_rate": 2.368951612903226e-06,
+      "loss": 1.8506,
+      "step": 38740
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 1.8361389636993408,
+      "learning_rate": 2.3437500000000002e-06,
+      "loss": 1.9847,
+      "step": 38750
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 1.050368309020996,
+      "learning_rate": 2.318548387096774e-06,
+      "loss": 1.6533,
+      "step": 38760
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 1.173164963722229,
+      "learning_rate": 2.2933467741935485e-06,
+      "loss": 1.6182,
+      "step": 38770
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 2.301607847213745,
+      "learning_rate": 2.268145161290323e-06,
+      "loss": 1.7152,
+      "step": 38780
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 1.632522463798523,
+      "learning_rate": 2.242943548387097e-06,
+      "loss": 1.7581,
+      "step": 38790
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 1.7555809020996094,
+      "learning_rate": 2.217741935483871e-06,
+      "loss": 1.9169,
+      "step": 38800
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.9660149216651917,
+      "learning_rate": 2.1925403225806454e-06,
+      "loss": 1.788,
+      "step": 38810
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 1.013667345046997,
+      "learning_rate": 2.16733870967742e-06,
+      "loss": 2.0077,
+      "step": 38820
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 1.3650189638137817,
+      "learning_rate": 2.1421370967741933e-06,
+      "loss": 1.7455,
+      "step": 38830
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.7961844205856323,
+      "learning_rate": 2.1169354838709676e-06,
+      "loss": 1.5673,
+      "step": 38840
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 1.554663062095642,
+      "learning_rate": 2.091733870967742e-06,
+      "loss": 1.794,
+      "step": 38850
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 1.5509215593338013,
+      "learning_rate": 2.0665322580645163e-06,
+      "loss": 1.6933,
+      "step": 38860
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.9071972370147705,
+      "learning_rate": 2.0413306451612903e-06,
+      "loss": 1.6966,
+      "step": 38870
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.9725357890129089,
+      "learning_rate": 2.0161290322580646e-06,
+      "loss": 1.8386,
+      "step": 38880
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 1.9880777597427368,
+      "learning_rate": 1.990927419354839e-06,
+      "loss": 1.8691,
+      "step": 38890
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 1.4876384735107422,
+      "learning_rate": 1.965725806451613e-06,
+      "loss": 1.7466,
+      "step": 38900
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 1.1476165056228638,
+      "learning_rate": 1.9405241935483872e-06,
+      "loss": 1.7666,
+      "step": 38910
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 1.8292884826660156,
+      "learning_rate": 1.9153225806451616e-06,
+      "loss": 1.9106,
+      "step": 38920
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 2.004047393798828,
+      "learning_rate": 1.8901209677419357e-06,
+      "loss": 1.7928,
+      "step": 38930
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 2.1043455600738525,
+      "learning_rate": 1.8649193548387098e-06,
+      "loss": 1.664,
+      "step": 38940
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 1.6563355922698975,
+      "learning_rate": 1.839717741935484e-06,
+      "loss": 1.8776,
+      "step": 38950
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.96150803565979,
+      "learning_rate": 1.8145161290322583e-06,
+      "loss": 1.861,
+      "step": 38960
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 2.1544246673583984,
+      "learning_rate": 1.7893145161290324e-06,
+      "loss": 1.7492,
+      "step": 38970
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 1.1623613834381104,
+      "learning_rate": 1.7641129032258068e-06,
+      "loss": 1.6624,
+      "step": 38980
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 2.3563108444213867,
+      "learning_rate": 1.7389112903225805e-06,
+      "loss": 1.7675,
+      "step": 38990
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 1.3422142267227173,
+      "learning_rate": 1.7137096774193548e-06,
+      "loss": 1.8913,
+      "step": 39000
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 39680,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 2,
+  "save_steps": 500,
+  "total_flos": 1.6887191274778476e+18,
+  "train_batch_size": 16,
+  "trial_name": null,
+  "trial_params": null
+}