diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,13093 @@
+{
+  "best_global_step": 88000,
+  "best_metric": 1.4495242834091187,
+  "best_model_checkpoint": "debertav3-ddp-8gpu-continue/checkpoint-88000",
+  "epoch": 15.0,
+  "eval_steps": 1000,
+  "global_step": 88200,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00017006802721088434,
+      "grad_norm": 3.9429101943969727,
+      "learning_rate": 0.0,
+      "loss": 10.4075,
+      "step": 1
+    },
+    {
+      "epoch": 0.008503401360544218,
+      "grad_norm": 1.971747636795044,
+      "learning_rate": 1.2250000000000001e-05,
+      "loss": 9.8035,
+      "step": 50
+    },
+    {
+      "epoch": 0.017006802721088437,
+      "grad_norm": 1.3740853071212769,
+      "learning_rate": 2.4750000000000002e-05,
+      "loss": 8.7689,
+      "step": 100
+    },
+    {
+      "epoch": 0.025510204081632654,
+      "grad_norm": 0.9082542061805725,
+      "learning_rate": 3.725e-05,
+      "loss": 7.832,
+      "step": 150
+    },
+    {
+      "epoch": 0.034013605442176874,
+      "grad_norm": 1.015873670578003,
+      "learning_rate": 4.975e-05,
+      "loss": 7.1484,
+      "step": 200
+    },
+    {
+      "epoch": 0.04251700680272109,
+      "grad_norm": 1.0678784847259521,
+      "learning_rate": 6.225e-05,
+      "loss": 6.6754,
+      "step": 250
+    },
+    {
+      "epoch": 0.05102040816326531,
+      "grad_norm": 1.0898432731628418,
+      "learning_rate": 7.475e-05,
+      "loss": 6.2519,
+      "step": 300
+    },
+    {
+      "epoch": 0.05952380952380952,
+      "grad_norm": 0.9323605298995972,
+      "learning_rate": 8.724999999999999e-05,
+      "loss": 5.8711,
+      "step": 350
+    },
+    {
+      "epoch": 0.06802721088435375,
+      "grad_norm": 0.7210457921028137,
+      "learning_rate": 9.975000000000001e-05,
+      "loss": 5.5604,
+      "step": 400
+    },
+    {
+      "epoch": 0.07653061224489796,
+      "grad_norm": 0.7947279214859009,
+      "learning_rate": 0.00011225,
+      "loss": 5.3079,
+      "step": 450
+    },
+    {
+      "epoch": 0.08503401360544217,
+      "grad_norm": 0.8039622902870178,
+      "learning_rate": 0.00012475,
+      "loss": 5.1044,
+      "step": 500
+    },
+    {
+      "epoch": 0.0935374149659864,
+      "grad_norm": 0.7226396799087524,
+      "learning_rate": 0.00013725,
+      "loss": 4.9271,
+      "step": 550
+    },
+    {
+      "epoch": 0.10204081632653061,
+      "grad_norm": 0.7884791493415833,
+      "learning_rate": 0.00014975,
+      "loss": 4.7654,
+      "step": 600
+    },
+    {
+      "epoch": 0.11054421768707483,
+      "grad_norm": 0.7507880330085754,
+      "learning_rate": 0.00016225000000000001,
+      "loss": 4.61,
+      "step": 650
+    },
+    {
+      "epoch": 0.11904761904761904,
+      "grad_norm": 0.6950189471244812,
+      "learning_rate": 0.00017475,
+      "loss": 4.474,
+      "step": 700
+    },
+    {
+      "epoch": 0.12755102040816327,
+      "grad_norm": 0.7212897539138794,
+      "learning_rate": 0.00018725,
+      "loss": 4.33,
+      "step": 750
+    },
+    {
+      "epoch": 0.1360544217687075,
+      "grad_norm": 0.8659721612930298,
+      "learning_rate": 0.00019975,
+      "loss": 4.177,
+      "step": 800
+    },
+    {
+      "epoch": 0.1445578231292517,
+      "grad_norm": 0.7730752229690552,
+      "learning_rate": 0.00021225,
+      "loss": 3.9974,
+      "step": 850
+    },
+    {
+      "epoch": 0.15306122448979592,
+      "grad_norm": 0.8463068604469299,
+      "learning_rate": 0.00022475000000000001,
+      "loss": 3.814,
+      "step": 900
+    },
+    {
+      "epoch": 0.16156462585034015,
+      "grad_norm": 0.8561194539070129,
+      "learning_rate": 0.00023725,
+      "loss": 3.6612,
+      "step": 950
+    },
+    {
+      "epoch": 0.17006802721088435,
+      "grad_norm": 0.7385916709899902,
+      "learning_rate": 0.00024975,
+      "loss": 3.5319,
+      "step": 1000
+    },
+    {
+      "epoch": 0.17006802721088435,
+      "eval_loss": 3.6021459102630615,
+      "eval_runtime": 75.3501,
+      "eval_samples_per_second": 1245.374,
+      "eval_steps_per_second": 4.871,
+      "step": 1000
+    },
+    {
+      "epoch": 0.17857142857142858,
+      "grad_norm": 0.7785052061080933,
+      "learning_rate": 0.00026225,
+      "loss": 3.4196,
+      "step": 1050
+    },
+    {
+      "epoch": 0.1870748299319728,
+      "grad_norm": 0.7699462175369263,
+      "learning_rate": 0.00027475,
+      "loss": 3.3248,
+      "step": 1100
+    },
+    {
+      "epoch": 0.195578231292517,
+      "grad_norm": 0.6943045258522034,
+      "learning_rate": 0.00028725,
+      "loss": 3.2327,
+      "step": 1150
+    },
+    {
+      "epoch": 0.20408163265306123,
+      "grad_norm": 0.7425329685211182,
+      "learning_rate": 0.00029975000000000005,
+      "loss": 3.1558,
+      "step": 1200
+    },
+    {
+      "epoch": 0.21258503401360543,
+      "grad_norm": 0.6596449613571167,
+      "learning_rate": 0.00031225000000000003,
+      "loss": 3.0896,
+      "step": 1250
+    },
+    {
+      "epoch": 0.22108843537414966,
+      "grad_norm": 0.6805656552314758,
+      "learning_rate": 0.00032475,
+      "loss": 3.0295,
+      "step": 1300
+    },
+    {
+      "epoch": 0.22959183673469388,
+      "grad_norm": 0.6616429686546326,
+      "learning_rate": 0.00033725,
+      "loss": 2.9716,
+      "step": 1350
+    },
+    {
+      "epoch": 0.23809523809523808,
+      "grad_norm": 0.699910581111908,
+      "learning_rate": 0.00034975,
+      "loss": 2.9154,
+      "step": 1400
+    },
+    {
+      "epoch": 0.2465986394557823,
+      "grad_norm": 0.6367819905281067,
+      "learning_rate": 0.00036225000000000005,
+      "loss": 2.88,
+      "step": 1450
+    },
+    {
+      "epoch": 0.25510204081632654,
+      "grad_norm": 0.633556067943573,
+      "learning_rate": 0.00037475000000000003,
+      "loss": 2.8288,
+      "step": 1500
+    },
+    {
+      "epoch": 0.26360544217687076,
+      "grad_norm": 0.6006012558937073,
+      "learning_rate": 0.00038725,
+      "loss": 2.8017,
+      "step": 1550
+    },
+    {
+      "epoch": 0.272108843537415,
+      "grad_norm": 0.6720077991485596,
+      "learning_rate": 0.00039975,
+      "loss": 2.762,
+      "step": 1600
+    },
+    {
+      "epoch": 0.28061224489795916,
+      "grad_norm": 0.6513054966926575,
+      "learning_rate": 0.00041225,
+      "loss": 2.7228,
+      "step": 1650
+    },
+    {
+      "epoch": 0.2891156462585034,
+      "grad_norm": 0.6120091676712036,
+      "learning_rate": 0.00042475000000000005,
+      "loss": 2.6978,
+      "step": 1700
+    },
+    {
+      "epoch": 0.2976190476190476,
+      "grad_norm": 0.6610727906227112,
+      "learning_rate": 0.00043725000000000003,
+      "loss": 2.6641,
+      "step": 1750
+    },
+    {
+      "epoch": 0.30612244897959184,
+      "grad_norm": 0.6299336552619934,
+      "learning_rate": 0.00044975,
+      "loss": 2.6343,
+      "step": 1800
+    },
+    {
+      "epoch": 0.31462585034013607,
+      "grad_norm": 0.6401428580284119,
+      "learning_rate": 0.00046225,
+      "loss": 2.6212,
+      "step": 1850
+    },
+    {
+      "epoch": 0.3231292517006803,
+      "grad_norm": 0.6018512845039368,
+      "learning_rate": 0.00047475,
+      "loss": 2.5919,
+      "step": 1900
+    },
+    {
+      "epoch": 0.33163265306122447,
+      "grad_norm": 0.63248610496521,
+      "learning_rate": 0.00048725000000000005,
+      "loss": 2.5739,
+      "step": 1950
+    },
+    {
+      "epoch": 0.3401360544217687,
+      "grad_norm": 0.646512508392334,
+      "learning_rate": 0.0004997500000000001,
+      "loss": 2.5493,
+      "step": 2000
+    },
+    {
+      "epoch": 0.3401360544217687,
+      "eval_loss": 2.755571126937866,
+      "eval_runtime": 75.4052,
+      "eval_samples_per_second": 1244.463,
+      "eval_steps_per_second": 4.867,
+      "step": 2000
+    },
+    {
+      "epoch": 0.3486394557823129,
+      "grad_norm": 0.6081250309944153,
+      "learning_rate": 0.000499568661971831,
+      "loss": 2.538,
+      "step": 2050
+    },
+    {
+      "epoch": 0.35714285714285715,
+      "grad_norm": 0.584749698638916,
+      "learning_rate": 0.0004991285211267605,
+      "loss": 2.5185,
+      "step": 2100
+    },
+    {
+      "epoch": 0.3656462585034014,
+      "grad_norm": 0.6018593907356262,
+      "learning_rate": 0.0004986883802816902,
+      "loss": 2.4893,
+      "step": 2150
+    },
+    {
+      "epoch": 0.3741496598639456,
+      "grad_norm": 0.5787435173988342,
+      "learning_rate": 0.0004982482394366197,
+      "loss": 2.4763,
+      "step": 2200
+    },
+    {
+      "epoch": 0.3826530612244898,
+      "grad_norm": 0.5765504240989685,
+      "learning_rate": 0.0004978080985915493,
+      "loss": 2.4471,
+      "step": 2250
+    },
+    {
+      "epoch": 0.391156462585034,
+      "grad_norm": 0.539054811000824,
+      "learning_rate": 0.0004973679577464789,
+      "loss": 2.4272,
+      "step": 2300
+    },
+    {
+      "epoch": 0.39965986394557823,
+      "grad_norm": 0.5542412400245667,
+      "learning_rate": 0.0004969278169014085,
+      "loss": 2.4178,
+      "step": 2350
+    },
+    {
+      "epoch": 0.40816326530612246,
+      "grad_norm": 0.5709657073020935,
+      "learning_rate": 0.0004964876760563381,
+      "loss": 2.4105,
+      "step": 2400
+    },
+    {
+      "epoch": 0.4166666666666667,
+      "grad_norm": 0.5530266165733337,
+      "learning_rate": 0.0004960475352112676,
+      "loss": 2.3812,
+      "step": 2450
+    },
+    {
+      "epoch": 0.42517006802721086,
+      "grad_norm": 0.5575669407844543,
+      "learning_rate": 0.0004956073943661972,
+      "loss": 2.3696,
+      "step": 2500
+    },
+    {
+      "epoch": 0.4336734693877551,
+      "grad_norm": 0.5466365814208984,
+      "learning_rate": 0.0004951672535211268,
+      "loss": 2.36,
+      "step": 2550
+    },
+    {
+      "epoch": 0.4421768707482993,
+      "grad_norm": 0.5788463354110718,
+      "learning_rate": 0.0004947271126760563,
+      "loss": 2.3409,
+      "step": 2600
+    },
+    {
+      "epoch": 0.45068027210884354,
+      "grad_norm": 0.5266286730766296,
+      "learning_rate": 0.0004942869718309859,
+      "loss": 2.3284,
+      "step": 2650
+    },
+    {
+      "epoch": 0.45918367346938777,
+      "grad_norm": 0.5419022440910339,
+      "learning_rate": 0.0004938468309859155,
+      "loss": 2.3175,
+      "step": 2700
+    },
+    {
+      "epoch": 0.467687074829932,
+      "grad_norm": 0.5173165202140808,
+      "learning_rate": 0.0004934066901408451,
+      "loss": 2.3021,
+      "step": 2750
+    },
+    {
+      "epoch": 0.47619047619047616,
+      "grad_norm": 0.5213769674301147,
+      "learning_rate": 0.0004929665492957746,
+      "loss": 2.2974,
+      "step": 2800
+    },
+    {
+      "epoch": 0.4846938775510204,
+      "grad_norm": 0.5393424034118652,
+      "learning_rate": 0.0004925264084507042,
+      "loss": 2.2834,
+      "step": 2850
+    },
+    {
+      "epoch": 0.4931972789115646,
+      "grad_norm": 0.5283843874931335,
+      "learning_rate": 0.0004920862676056338,
+      "loss": 2.2894,
+      "step": 2900
+    },
+    {
+      "epoch": 0.5017006802721088,
+      "grad_norm": 0.5366395711898804,
+      "learning_rate": 0.0004916461267605634,
+      "loss": 2.2643,
+      "step": 2950
+    },
+    {
+      "epoch": 0.5102040816326531,
+      "grad_norm": 0.5475978851318359,
+      "learning_rate": 0.000491205985915493,
+      "loss": 2.2525,
+      "step": 3000
+    },
+    {
+      "epoch": 0.5102040816326531,
+      "eval_loss": 2.454470634460449,
+      "eval_runtime": 75.3482,
+      "eval_samples_per_second": 1245.405,
+      "eval_steps_per_second": 4.871,
+      "step": 3000
+    },
+    {
+      "epoch": 0.5187074829931972,
+      "grad_norm": 0.5384633541107178,
+      "learning_rate": 0.0004907658450704226,
+      "loss": 2.2469,
+      "step": 3050
+    },
+    {
+      "epoch": 0.5272108843537415,
+      "grad_norm": 0.538545548915863,
+      "learning_rate": 0.0004903257042253521,
+      "loss": 2.2378,
+      "step": 3100
+    },
+    {
+      "epoch": 0.5357142857142857,
+      "grad_norm": 0.5223479270935059,
+      "learning_rate": 0.0004898855633802817,
+      "loss": 2.2343,
+      "step": 3150
+    },
+    {
+      "epoch": 0.54421768707483,
+      "grad_norm": 0.531447172164917,
+      "learning_rate": 0.0004894454225352113,
+      "loss": 2.2201,
+      "step": 3200
+    },
+    {
+      "epoch": 0.5527210884353742,
+      "grad_norm": 0.5139949321746826,
+      "learning_rate": 0.0004890052816901409,
+      "loss": 2.2098,
+      "step": 3250
+    },
+    {
+      "epoch": 0.5612244897959183,
+      "grad_norm": 0.5428385138511658,
+      "learning_rate": 0.0004885651408450705,
+      "loss": 2.2115,
+      "step": 3300
+    },
+    {
+      "epoch": 0.5697278911564626,
+      "grad_norm": 0.5274862051010132,
+      "learning_rate": 0.000488125,
+      "loss": 2.1987,
+      "step": 3350
+    },
+    {
+      "epoch": 0.5782312925170068,
+      "grad_norm": 0.5110312104225159,
+      "learning_rate": 0.0004876848591549296,
+      "loss": 2.1843,
+      "step": 3400
+    },
+    {
+      "epoch": 0.5867346938775511,
+      "grad_norm": 0.49753338098526,
+      "learning_rate": 0.00048724471830985914,
+      "loss": 2.1755,
+      "step": 3450
+    },
+    {
+      "epoch": 0.5952380952380952,
+      "grad_norm": 0.5169926881790161,
+      "learning_rate": 0.00048680457746478874,
+      "loss": 2.1779,
+      "step": 3500
+    },
+    {
+      "epoch": 0.6037414965986394,
+      "grad_norm": 0.5181500911712646,
+      "learning_rate": 0.0004863644366197183,
+      "loss": 2.1658,
+      "step": 3550
+    },
+    {
+      "epoch": 0.6122448979591837,
+      "grad_norm": 0.5171290636062622,
+      "learning_rate": 0.00048592429577464793,
+      "loss": 2.1668,
+      "step": 3600
+    },
+    {
+      "epoch": 0.6207482993197279,
+      "grad_norm": 0.4980362355709076,
+      "learning_rate": 0.00048548415492957747,
+      "loss": 2.1545,
+      "step": 3650
+    },
+    {
+      "epoch": 0.6292517006802721,
+      "grad_norm": 0.5099490284919739,
+      "learning_rate": 0.00048504401408450707,
+      "loss": 2.1415,
+      "step": 3700
+    },
+    {
+      "epoch": 0.6377551020408163,
+      "grad_norm": 0.5084192752838135,
+      "learning_rate": 0.0004846038732394366,
+      "loss": 2.1417,
+      "step": 3750
+    },
+    {
+      "epoch": 0.6462585034013606,
+      "grad_norm": 0.48829177021980286,
+      "learning_rate": 0.00048416373239436626,
+      "loss": 2.1359,
+      "step": 3800
+    },
+    {
+      "epoch": 0.6547619047619048,
+      "grad_norm": 0.5111795663833618,
+      "learning_rate": 0.0004837235915492958,
+      "loss": 2.1348,
+      "step": 3850
+    },
+    {
+      "epoch": 0.6632653061224489,
+      "grad_norm": 0.49867865443229675,
+      "learning_rate": 0.00048328345070422534,
+      "loss": 2.1199,
+      "step": 3900
+    },
+    {
+      "epoch": 0.6717687074829932,
+      "grad_norm": 0.4912058711051941,
+      "learning_rate": 0.00048284330985915493,
+      "loss": 2.1169,
+      "step": 3950
+    },
+    {
+      "epoch": 0.6802721088435374,
+      "grad_norm": 0.5115543603897095,
+      "learning_rate": 0.00048240316901408453,
+      "loss": 2.1093,
+      "step": 4000
+    },
+    {
+      "epoch": 0.6802721088435374,
+      "eval_loss": 2.320960521697998,
+      "eval_runtime": 75.9327,
+      "eval_samples_per_second": 1235.818,
+      "eval_steps_per_second": 4.833,
+      "step": 4000
+    },
+    {
+      "epoch": 0.6887755102040817,
+      "grad_norm": 0.5165016651153564,
+      "learning_rate": 0.0004819630281690141,
+      "loss": 2.1103,
+      "step": 4050
+    },
+    {
+      "epoch": 0.6972789115646258,
+      "grad_norm": 0.49575796723365784,
+      "learning_rate": 0.00048152288732394367,
+      "loss": 2.1027,
+      "step": 4100
+    },
+    {
+      "epoch": 0.70578231292517,
+      "grad_norm": 0.5010895133018494,
+      "learning_rate": 0.0004810827464788732,
+      "loss": 2.0939,
+      "step": 4150
+    },
+    {
+      "epoch": 0.7142857142857143,
+      "grad_norm": 0.49885183572769165,
+      "learning_rate": 0.00048064260563380286,
+      "loss": 2.0951,
+      "step": 4200
+    },
+    {
+      "epoch": 0.7227891156462585,
+      "grad_norm": 0.47758999466896057,
+      "learning_rate": 0.0004802024647887324,
+      "loss": 2.0834,
+      "step": 4250
+    },
+    {
+      "epoch": 0.7312925170068028,
+      "grad_norm": 0.48634713888168335,
+      "learning_rate": 0.000479762323943662,
+      "loss": 2.0833,
+      "step": 4300
+    },
+    {
+      "epoch": 0.7397959183673469,
+      "grad_norm": 0.5028336644172668,
+      "learning_rate": 0.00047932218309859153,
+      "loss": 2.0804,
+      "step": 4350
+    },
+    {
+      "epoch": 0.7482993197278912,
+      "grad_norm": 0.4951624870300293,
+      "learning_rate": 0.00047888204225352113,
+      "loss": 2.0697,
+      "step": 4400
+    },
+    {
+      "epoch": 0.7568027210884354,
+      "grad_norm": 0.5072620511054993,
+      "learning_rate": 0.0004784419014084507,
+      "loss": 2.0745,
+      "step": 4450
+    },
+    {
+      "epoch": 0.7653061224489796,
+      "grad_norm": 0.4807872474193573,
+      "learning_rate": 0.0004780017605633803,
+      "loss": 2.0618,
+      "step": 4500
+    },
+    {
+      "epoch": 0.7738095238095238,
+      "grad_norm": 0.49827975034713745,
+      "learning_rate": 0.00047756161971830986,
+      "loss": 2.0648,
+      "step": 4550
+    },
+    {
+      "epoch": 0.782312925170068,
+      "grad_norm": 0.48257699608802795,
+      "learning_rate": 0.0004771214788732394,
+      "loss": 2.0542,
+      "step": 4600
+    },
+    {
+      "epoch": 0.7908163265306123,
+      "grad_norm": 0.4961060583591461,
+      "learning_rate": 0.00047668133802816905,
+      "loss": 2.0531,
+      "step": 4650
+    },
+    {
+      "epoch": 0.7993197278911565,
+      "grad_norm": 0.48454368114471436,
+      "learning_rate": 0.0004762411971830986,
+      "loss": 2.0497,
+      "step": 4700
+    },
+    {
+      "epoch": 0.8078231292517006,
+      "grad_norm": 0.48959559202194214,
+      "learning_rate": 0.0004758010563380282,
+      "loss": 2.0539,
+      "step": 4750
+    },
+    {
+      "epoch": 0.8163265306122449,
+      "grad_norm": 0.4989263415336609,
+      "learning_rate": 0.00047536091549295773,
+      "loss": 2.0436,
+      "step": 4800
+    },
+    {
+      "epoch": 0.8248299319727891,
+      "grad_norm": 0.491557776927948,
+      "learning_rate": 0.0004749207746478874,
+      "loss": 2.0347,
+      "step": 4850
+    },
+    {
+      "epoch": 0.8333333333333334,
+      "grad_norm": 0.4861533045768738,
+      "learning_rate": 0.0004744806338028169,
+      "loss": 2.0269,
+      "step": 4900
+    },
+    {
+      "epoch": 0.8418367346938775,
+      "grad_norm": 0.5007792711257935,
+      "learning_rate": 0.00047404049295774646,
+      "loss": 2.0277,
+      "step": 4950
+    },
+    {
+      "epoch": 0.8503401360544217,
+      "grad_norm": 0.48347318172454834,
+      "learning_rate": 0.00047360035211267606,
+      "loss": 2.0278,
+      "step": 5000
+    },
+    {
+      "epoch": 0.8503401360544217,
+      "eval_loss": 2.221730947494507,
+      "eval_runtime": 75.3978,
+      "eval_samples_per_second": 1244.585,
+      "eval_steps_per_second": 4.868,
+      "step": 5000
+    },
+    {
+      "epoch": 0.858843537414966,
+      "grad_norm": 0.4758061468601227,
+      "learning_rate": 0.00047316021126760565,
+      "loss": 2.0207,
+      "step": 5050
+    },
+    {
+      "epoch": 0.8673469387755102,
+      "grad_norm": 0.5015333890914917,
+      "learning_rate": 0.00047272007042253525,
+      "loss": 2.0213,
+      "step": 5100
+    },
+    {
+      "epoch": 0.8758503401360545,
+      "grad_norm": 0.46741312742233276,
+      "learning_rate": 0.0004722799295774648,
+      "loss": 2.0115,
+      "step": 5150
+    },
+    {
+      "epoch": 0.8843537414965986,
+      "grad_norm": 0.4716244339942932,
+      "learning_rate": 0.0004718397887323944,
+      "loss": 2.0041,
+      "step": 5200
+    },
+    {
+      "epoch": 0.8928571428571429,
+      "grad_norm": 0.470093309879303,
+      "learning_rate": 0.000471399647887324,
+      "loss": 2.0101,
+      "step": 5250
+    },
+    {
+      "epoch": 0.9013605442176871,
+      "grad_norm": 0.46333491802215576,
+      "learning_rate": 0.0004709595070422535,
+      "loss": 1.9985,
+      "step": 5300
+    },
+    {
+      "epoch": 0.9098639455782312,
+      "grad_norm": 0.4772166609764099,
+      "learning_rate": 0.0004705193661971831,
+      "loss": 1.9956,
+      "step": 5350
+    },
+    {
+      "epoch": 0.9183673469387755,
+      "grad_norm": 0.49403470754623413,
+      "learning_rate": 0.00047007922535211266,
+      "loss": 1.991,
+      "step": 5400
+    },
+    {
+      "epoch": 0.9268707482993197,
+      "grad_norm": 0.4855027198791504,
+      "learning_rate": 0.00046963908450704225,
+      "loss": 1.9972,
+      "step": 5450
+    },
+    {
+      "epoch": 0.935374149659864,
+      "grad_norm": 0.4754594564437866,
+      "learning_rate": 0.00046919894366197185,
+      "loss": 1.9902,
+      "step": 5500
+    },
+    {
+      "epoch": 0.9438775510204082,
+      "grad_norm": 0.4977014660835266,
+      "learning_rate": 0.00046875880281690144,
+      "loss": 1.9904,
+      "step": 5550
+    },
+    {
+      "epoch": 0.9523809523809523,
+      "grad_norm": 0.48129725456237793,
+      "learning_rate": 0.000468318661971831,
+      "loss": 1.9855,
+      "step": 5600
+    },
+    {
+      "epoch": 0.9608843537414966,
+      "grad_norm": 0.484291672706604,
+      "learning_rate": 0.00046787852112676053,
+      "loss": 1.9781,
+      "step": 5650
+    },
+    {
+      "epoch": 0.9693877551020408,
+      "grad_norm": 0.4759155809879303,
+      "learning_rate": 0.0004674383802816902,
+      "loss": 1.9766,
+      "step": 5700
+    },
+    {
+      "epoch": 0.9778911564625851,
+      "grad_norm": 0.48138466477394104,
+      "learning_rate": 0.0004669982394366197,
+      "loss": 1.9678,
+      "step": 5750
+    },
+    {
+      "epoch": 0.9863945578231292,
+      "grad_norm": 0.48765894770622253,
+      "learning_rate": 0.0004665580985915493,
+      "loss": 1.9746,
+      "step": 5800
+    },
+    {
+      "epoch": 0.9948979591836735,
+      "grad_norm": 0.47016099095344543,
+      "learning_rate": 0.00046611795774647885,
+      "loss": 1.9745,
+      "step": 5850
+    },
+    {
+      "epoch": 1.0034013605442176,
+      "grad_norm": 0.4722880721092224,
+      "learning_rate": 0.0004656778169014085,
+      "loss": 1.961,
+      "step": 5900
+    },
+    {
+      "epoch": 1.0119047619047619,
+      "grad_norm": 0.47606250643730164,
+      "learning_rate": 0.00046523767605633805,
+      "loss": 1.9548,
+      "step": 5950
+    },
+    {
+      "epoch": 1.0204081632653061,
+      "grad_norm": 0.4867871403694153,
+      "learning_rate": 0.0004647975352112676,
+      "loss": 1.9531,
+      "step": 6000
+    },
+    {
+      "epoch": 1.0204081632653061,
+      "eval_loss": 2.1581504344940186,
+      "eval_runtime": 75.3496,
+      "eval_samples_per_second": 1245.381,
+      "eval_steps_per_second": 4.871,
+      "step": 6000
+    },
+    {
+      "epoch": 1.0289115646258504,
+      "grad_norm": 0.4676497280597687,
+      "learning_rate": 0.0004643573943661972,
+      "loss": 1.9505,
+      "step": 6050
+    },
+    {
+      "epoch": 1.0374149659863945,
+      "grad_norm": 0.47663334012031555,
+      "learning_rate": 0.0004639172535211268,
+      "loss": 1.9509,
+      "step": 6100
+    },
+    {
+      "epoch": 1.0459183673469388,
+      "grad_norm": 0.46838682889938354,
+      "learning_rate": 0.0004634771126760564,
+      "loss": 1.9522,
+      "step": 6150
+    },
+    {
+      "epoch": 1.054421768707483,
+      "grad_norm": 0.47476112842559814,
+      "learning_rate": 0.0004630369718309859,
+      "loss": 1.9575,
+      "step": 6200
+    },
+    {
+      "epoch": 1.0629251700680271,
+      "grad_norm": 0.4738709330558777,
+      "learning_rate": 0.0004625968309859155,
+      "loss": 1.9535,
+      "step": 6250
+    },
+    {
+      "epoch": 1.0714285714285714,
+      "grad_norm": 0.46784505248069763,
+      "learning_rate": 0.0004621566901408451,
+      "loss": 1.95,
+      "step": 6300
+    },
+    {
+      "epoch": 1.0799319727891157,
+      "grad_norm": 0.4729381501674652,
+      "learning_rate": 0.00046171654929577465,
+      "loss": 1.9382,
+      "step": 6350
+    },
+    {
+      "epoch": 1.08843537414966,
+      "grad_norm": 0.48842012882232666,
+      "learning_rate": 0.00046127640845070424,
+      "loss": 1.9383,
+      "step": 6400
+    },
+    {
+      "epoch": 1.096938775510204,
+      "grad_norm": 0.47705018520355225,
+      "learning_rate": 0.0004608362676056338,
+      "loss": 1.9298,
+      "step": 6450
+    },
+    {
+      "epoch": 1.1054421768707483,
+      "grad_norm": 0.47480863332748413,
+      "learning_rate": 0.00046039612676056343,
+      "loss": 1.9356,
+      "step": 6500
+    },
+    {
+      "epoch": 1.1139455782312926,
+      "grad_norm": 0.4836580455303192,
+      "learning_rate": 0.000459955985915493,
+      "loss": 1.936,
+      "step": 6550
+    },
+    {
+      "epoch": 1.1224489795918366,
+      "grad_norm": 0.47330114245414734,
+      "learning_rate": 0.00045951584507042257,
+      "loss": 1.9217,
+      "step": 6600
+    },
+    {
+      "epoch": 1.130952380952381,
+      "grad_norm": 0.45560622215270996,
+      "learning_rate": 0.0004590757042253521,
+      "loss": 1.9278,
+      "step": 6650
+    },
+    {
+      "epoch": 1.1394557823129252,
+      "grad_norm": 0.48356184363365173,
+      "learning_rate": 0.00045863556338028165,
+      "loss": 1.9204,
+      "step": 6700
+    },
+    {
+      "epoch": 1.1479591836734695,
+      "grad_norm": 0.48046737909317017,
+      "learning_rate": 0.0004581954225352113,
+      "loss": 1.9185,
+      "step": 6750
+    },
+    {
+      "epoch": 1.1564625850340136,
+      "grad_norm": 0.47801148891448975,
+      "learning_rate": 0.00045775528169014084,
+      "loss": 1.9084,
+      "step": 6800
+    },
+    {
+      "epoch": 1.1649659863945578,
+      "grad_norm": 0.46705126762390137,
+      "learning_rate": 0.00045731514084507044,
+      "loss": 1.9138,
+      "step": 6850
+    },
+    {
+      "epoch": 1.1734693877551021,
+      "grad_norm": 0.47566279768943787,
+      "learning_rate": 0.000456875,
+      "loss": 1.9139,
+      "step": 6900
+    },
+    {
+      "epoch": 1.1819727891156462,
+      "grad_norm": 0.4746539294719696,
+      "learning_rate": 0.00045643485915492963,
+      "loss": 1.916,
+      "step": 6950
+    },
+    {
+      "epoch": 1.1904761904761905,
+      "grad_norm": 0.49360892176628113,
+      "learning_rate": 0.00045599471830985917,
+      "loss": 1.9086,
+      "step": 7000
+    },
+    {
+      "epoch": 1.1904761904761905,
+      "eval_loss": 2.1105546951293945,
+      "eval_runtime": 75.3645,
+      "eval_samples_per_second": 1245.136,
+      "eval_steps_per_second": 4.87,
+      "step": 7000
+    },
+    {
+      "epoch": 1.1989795918367347,
+      "grad_norm": 0.49067452549934387,
+      "learning_rate": 0.0004555545774647887,
+      "loss": 1.9044,
+      "step": 7050
+    },
+    {
+      "epoch": 1.2074829931972788,
+      "grad_norm": 0.48024600744247437,
+      "learning_rate": 0.0004551144366197183,
+      "loss": 1.9143,
+      "step": 7100
+    },
+    {
+      "epoch": 1.215986394557823,
+      "grad_norm": 0.4815312922000885,
+      "learning_rate": 0.0004546742957746479,
+      "loss": 1.9038,
+      "step": 7150
+    },
+    {
+      "epoch": 1.2244897959183674,
+      "grad_norm": 0.47568100690841675,
+      "learning_rate": 0.0004542341549295775,
+      "loss": 1.9079,
+      "step": 7200
+    },
+    {
+      "epoch": 1.2329931972789117,
+      "grad_norm": 0.47568491101264954,
+      "learning_rate": 0.00045379401408450704,
+      "loss": 1.9024,
+      "step": 7250
+    },
+    {
+      "epoch": 1.2414965986394557,
+      "grad_norm": 0.46364548802375793,
+      "learning_rate": 0.00045335387323943663,
+      "loss": 1.9008,
+      "step": 7300
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.4672023355960846,
+      "learning_rate": 0.00045291373239436623,
+      "loss": 1.9022,
+      "step": 7350
+    },
+    {
+      "epoch": 1.2585034013605443,
+      "grad_norm": 0.4671889841556549,
+      "learning_rate": 0.00045247359154929577,
+      "loss": 1.894,
+      "step": 7400
+    },
+    {
+      "epoch": 1.2670068027210886,
+      "grad_norm": 0.4657513201236725,
+      "learning_rate": 0.00045203345070422537,
+      "loss": 1.8978,
+      "step": 7450
+    },
+    {
+      "epoch": 1.2755102040816326,
+      "grad_norm": 0.4698520004749298,
+      "learning_rate": 0.0004515933098591549,
+      "loss": 1.8945,
+      "step": 7500
+    },
+    {
+      "epoch": 1.284013605442177,
+      "grad_norm": 0.4804818332195282,
+      "learning_rate": 0.00045115316901408456,
+      "loss": 1.8921,
+      "step": 7550
+    },
+    {
+      "epoch": 1.2925170068027212,
+      "grad_norm": 0.4709530174732208,
+      "learning_rate": 0.0004507130281690141,
+      "loss": 1.889,
+      "step": 7600
+    },
+    {
+      "epoch": 1.3010204081632653,
+      "grad_norm": 0.47395414113998413,
+      "learning_rate": 0.0004502728873239437,
+      "loss": 1.8837,
+      "step": 7650
+    },
+    {
+      "epoch": 1.3095238095238095,
+      "grad_norm": 0.4659099876880646,
+      "learning_rate": 0.00044983274647887323,
+      "loss": 1.8836,
+      "step": 7700
+    },
+    {
+      "epoch": 1.3180272108843538,
+      "grad_norm": 0.4623669683933258,
+      "learning_rate": 0.00044939260563380283,
+      "loss": 1.8823,
+      "step": 7750
+    },
+    {
+      "epoch": 1.3265306122448979,
+      "grad_norm": 0.4751649498939514,
+      "learning_rate": 0.0004489524647887324,
+      "loss": 1.8703,
+      "step": 7800
+    },
+    {
+      "epoch": 1.3350340136054422,
+      "grad_norm": 0.4874952435493469,
+      "learning_rate": 0.00044851232394366197,
+      "loss": 1.8765,
+      "step": 7850
+    },
+    {
+      "epoch": 1.3435374149659864,
+      "grad_norm": 0.48088109493255615,
+      "learning_rate": 0.00044807218309859156,
+      "loss": 1.8787,
+      "step": 7900
+    },
+    {
+      "epoch": 1.3520408163265305,
+      "grad_norm": 0.48666709661483765,
+      "learning_rate": 0.0004476320422535211,
+      "loss": 1.8723,
+      "step": 7950
+    },
+    {
+      "epoch": 1.3605442176870748,
+      "grad_norm": 0.5058099031448364,
+      "learning_rate": 0.00044719190140845075,
+      "loss": 1.8682,
+      "step": 8000
+    },
+    {
+      "epoch": 1.3605442176870748,
+      "eval_loss": 2.066969156265259,
+      "eval_runtime": 75.2958,
+      "eval_samples_per_second": 1246.271,
+      "eval_steps_per_second": 4.874,
+      "step": 8000
+    },
+    {
+      "epoch": 1.369047619047619,
+      "grad_norm": 0.47028714418411255,
+      "learning_rate": 0.0004467517605633803,
+      "loss": 1.8619,
+      "step": 8050
+    },
+    {
+      "epoch": 1.3775510204081631,
+      "grad_norm": 0.468432754278183,
+      "learning_rate": 0.00044631161971830983,
+      "loss": 1.8669,
+      "step": 8100
+    },
+    {
+      "epoch": 1.3860544217687074,
+      "grad_norm": 0.46689048409461975,
+      "learning_rate": 0.00044587147887323943,
+      "loss": 1.8657,
+      "step": 8150
+    },
+    {
+      "epoch": 1.3945578231292517,
+      "grad_norm": 0.464141309261322,
+      "learning_rate": 0.000445431338028169,
+      "loss": 1.863,
+      "step": 8200
+    },
+    {
+      "epoch": 1.403061224489796,
+      "grad_norm": 0.46267494559288025,
+      "learning_rate": 0.0004449911971830986,
+      "loss": 1.8696,
+      "step": 8250
+    },
+    {
+      "epoch": 1.4115646258503403,
+      "grad_norm": 0.4653218984603882,
+      "learning_rate": 0.00044455105633802816,
+      "loss": 1.8554,
+      "step": 8300
+    },
+    {
+      "epoch": 1.4200680272108843,
+      "grad_norm": 0.4844844937324524,
+      "learning_rate": 0.00044411091549295776,
+      "loss": 1.8525,
+      "step": 8350
+    },
+    {
+      "epoch": 1.4285714285714286,
+      "grad_norm": 0.47869589924812317,
+      "learning_rate": 0.00044367077464788735,
+      "loss": 1.8584,
+      "step": 8400
+    },
+    {
+      "epoch": 1.4370748299319729,
+      "grad_norm": 0.4788654148578644,
+      "learning_rate": 0.0004432306338028169,
+      "loss": 1.8562,
+      "step": 8450
+    },
+    {
+      "epoch": 1.445578231292517,
+      "grad_norm": 0.4887782335281372,
+      "learning_rate": 0.0004427904929577465,
+      "loss": 1.8583,
+      "step": 8500
+    },
+    {
+      "epoch": 1.4540816326530612,
+      "grad_norm": 0.48299601674079895,
+      "learning_rate": 0.00044235035211267603,
+      "loss": 1.8557,
+      "step": 8550
+    },
+    {
+      "epoch": 1.4625850340136055,
+      "grad_norm": 0.4597221910953522,
+      "learning_rate": 0.0004419102112676057,
+      "loss": 1.847,
+      "step": 8600
+    },
+    {
+      "epoch": 1.4710884353741496,
+      "grad_norm": 0.4786689877510071,
+      "learning_rate": 0.0004414700704225352,
+      "loss": 1.8393,
+      "step": 8650
+    },
+    {
+      "epoch": 1.4795918367346939,
+      "grad_norm": 0.46319809556007385,
+      "learning_rate": 0.0004410299295774648,
+      "loss": 1.8468,
+      "step": 8700
+    },
+    {
+      "epoch": 1.4880952380952381,
+      "grad_norm": 0.46916770935058594,
+      "learning_rate": 0.00044058978873239436,
+      "loss": 1.8414,
+      "step": 8750
+    },
+    {
+      "epoch": 1.4965986394557822,
+      "grad_norm": 0.4812500774860382,
+      "learning_rate": 0.00044014964788732395,
+      "loss": 1.8496,
+      "step": 8800
+    },
+    {
+      "epoch": 1.5051020408163265,
+      "grad_norm": 0.47240397334098816,
+      "learning_rate": 0.00043970950704225355,
+      "loss": 1.8415,
+      "step": 8850
+    },
+    {
+      "epoch": 1.5136054421768708,
+      "grad_norm": 0.4675607979297638,
+      "learning_rate": 0.0004392693661971831,
+      "loss": 1.8395,
+      "step": 8900
+    },
+    {
+      "epoch": 1.5221088435374148,
+      "grad_norm": 0.4568563401699066,
+      "learning_rate": 0.0004388380281690141,
+      "loss": 1.8349,
+      "step": 8950
+    },
+    {
+      "epoch": 1.5306122448979593,
+      "grad_norm": 0.4698484539985657,
+      "learning_rate": 0.00043839788732394363,
+      "loss": 1.8388,
+      "step": 9000
+    },
+    {
+      "epoch": 1.5306122448979593,
+      "eval_loss": 2.031731128692627,
+      "eval_runtime": 75.3373,
+      "eval_samples_per_second": 1245.584,
+      "eval_steps_per_second": 4.871,
+      "step": 9000
+    },
+    {
+      "epoch": 1.5391156462585034,
+      "grad_norm": 0.4721793532371521,
+      "learning_rate": 0.0004379577464788733,
+      "loss": 1.8368,
+      "step": 9050
+    },
+    {
+      "epoch": 1.5476190476190477,
+      "grad_norm": 0.48082154989242554,
+      "learning_rate": 0.0004375176056338028,
+      "loss": 1.8367,
+      "step": 9100
+    },
+    {
+      "epoch": 1.556122448979592,
+      "grad_norm": 0.47875961661338806,
+      "learning_rate": 0.00043707746478873237,
+      "loss": 1.834,
+      "step": 9150
+    },
+    {
+      "epoch": 1.564625850340136,
+      "grad_norm": 0.4505126178264618,
+      "learning_rate": 0.00043663732394366196,
+      "loss": 1.8293,
+      "step": 9200
+    },
+    {
+      "epoch": 1.5731292517006803,
+      "grad_norm": 0.4831013083457947,
+      "learning_rate": 0.00043619718309859156,
+      "loss": 1.8325,
+      "step": 9250
+    },
+    {
+      "epoch": 1.5816326530612246,
+      "grad_norm": 0.46912670135498047,
+      "learning_rate": 0.00043575704225352115,
+      "loss": 1.8201,
+      "step": 9300
+    },
+    {
+      "epoch": 1.5901360544217686,
+      "grad_norm": 0.4775020480155945,
+      "learning_rate": 0.0004353169014084507,
+      "loss": 1.8324,
+      "step": 9350
+    },
+    {
+      "epoch": 1.598639455782313,
+      "grad_norm": 0.4624606668949127,
+      "learning_rate": 0.0004348767605633803,
+      "loss": 1.8328,
+      "step": 9400
+    },
+    {
+      "epoch": 1.6071428571428572,
+      "grad_norm": 0.47817811369895935,
+      "learning_rate": 0.0004344366197183099,
+      "loss": 1.8265,
+      "step": 9450
+    },
+    {
+      "epoch": 1.6156462585034013,
+      "grad_norm": 0.4682134985923767,
+      "learning_rate": 0.0004339964788732395,
+      "loss": 1.8193,
+      "step": 9500
+    },
+    {
+      "epoch": 1.6241496598639455,
+      "grad_norm": 0.4736348092556,
+      "learning_rate": 0.000433556338028169,
+      "loss": 1.8144,
+      "step": 9550
+    },
+    {
+      "epoch": 1.6326530612244898,
+      "grad_norm": 0.47171303629875183,
+      "learning_rate": 0.00043311619718309856,
+      "loss": 1.8156,
+      "step": 9600
+    },
+    {
+      "epoch": 1.641156462585034,
+      "grad_norm": 0.4707530736923218,
+      "learning_rate": 0.0004326760563380282,
+      "loss": 1.8146,
+      "step": 9650
+    },
+    {
+      "epoch": 1.6496598639455784,
+      "grad_norm": 0.47999903559684753,
+      "learning_rate": 0.00043223591549295775,
+      "loss": 1.822,
+      "step": 9700
+    },
+    {
+      "epoch": 1.6581632653061225,
+      "grad_norm": 0.4823855757713318,
+      "learning_rate": 0.00043179577464788735,
+      "loss": 1.8161,
+      "step": 9750
+    },
+    {
+      "epoch": 1.6666666666666665,
+      "grad_norm": 0.47148072719573975,
+      "learning_rate": 0.0004313556338028169,
+      "loss": 1.8117,
+      "step": 9800
+    },
+    {
+      "epoch": 1.675170068027211,
+      "grad_norm": 0.46029239892959595,
+      "learning_rate": 0.00043091549295774654,
+      "loss": 1.8132,
+      "step": 9850
+    },
+    {
+      "epoch": 1.683673469387755,
+      "grad_norm": 0.47228437662124634,
+      "learning_rate": 0.0004304753521126761,
+      "loss": 1.816,
+      "step": 9900
+    },
+    {
+      "epoch": 1.6921768707482994,
+      "grad_norm": 0.4653479754924774,
+      "learning_rate": 0.0004300352112676056,
+      "loss": 1.8084,
+      "step": 9950
+    },
+    {
+      "epoch": 1.7006802721088436,
+      "grad_norm": 0.4696131646633148,
+      "learning_rate": 0.0004295950704225352,
+      "loss": 1.8072,
+      "step": 10000
+    },
+    {
+      "epoch": 1.7006802721088436,
+      "eval_loss": 1.9987208843231201,
+      "eval_runtime": 75.2953,
+      "eval_samples_per_second": 1246.28,
+      "eval_steps_per_second": 4.874,
+      "step": 10000
+    },
+    {
+      "epoch": 1.7091836734693877,
+      "grad_norm": 0.4656902253627777,
+      "learning_rate": 0.00042915492957746476,
+      "loss": 1.8049,
+      "step": 10050
+    },
+    {
+      "epoch": 1.717687074829932,
+      "grad_norm": 0.47671329975128174,
+      "learning_rate": 0.0004287147887323944,
+      "loss": 1.7934,
+      "step": 10100
+    },
+    {
+      "epoch": 1.7261904761904763,
+      "grad_norm": 0.46021008491516113,
+      "learning_rate": 0.00042827464788732395,
+      "loss": 1.8059,
+      "step": 10150
+    },
+    {
+      "epoch": 1.7346938775510203,
+      "grad_norm": 0.46965909004211426,
+      "learning_rate": 0.00042783450704225354,
+      "loss": 1.7958,
+      "step": 10200
+    },
+    {
+      "epoch": 1.7431972789115646,
+      "grad_norm": 0.4655004143714905,
+      "learning_rate": 0.0004273943661971831,
+      "loss": 1.7989,
+      "step": 10250
+    },
+    {
+      "epoch": 1.751700680272109,
+      "grad_norm": 0.48593828082084656,
+      "learning_rate": 0.0004269542253521127,
+      "loss": 1.7904,
+      "step": 10300
+    },
+    {
+      "epoch": 1.760204081632653,
+      "grad_norm": 0.4698497951030731,
+      "learning_rate": 0.0004265140845070423,
+      "loss": 1.8005,
+      "step": 10350
+    },
+    {
+      "epoch": 1.7687074829931972,
+      "grad_norm": 0.4775938391685486,
+      "learning_rate": 0.0004260739436619718,
+      "loss": 1.7992,
+      "step": 10400
+    },
+    {
+      "epoch": 1.7772108843537415,
+      "grad_norm": 0.45465579628944397,
+      "learning_rate": 0.0004256338028169014,
+      "loss": 1.7973,
+      "step": 10450
+    },
+    {
+      "epoch": 1.7857142857142856,
+      "grad_norm": 0.4619791507720947,
+      "learning_rate": 0.000425193661971831,
+      "loss": 1.7913,
+      "step": 10500
+    },
+    {
+      "epoch": 1.79421768707483,
+      "grad_norm": 0.4738638699054718,
+      "learning_rate": 0.0004247535211267606,
+      "loss": 1.7922,
+      "step": 10550
+    },
+    {
+      "epoch": 1.8027210884353742,
+      "grad_norm": 0.46855342388153076,
+      "learning_rate": 0.00042431338028169015,
+      "loss": 1.7894,
+      "step": 10600
+    },
+    {
+      "epoch": 1.8112244897959182,
+      "grad_norm": 0.48508965969085693,
+      "learning_rate": 0.0004238732394366197,
+      "loss": 1.7815,
+      "step": 10650
+    },
+    {
+      "epoch": 1.8197278911564627,
+      "grad_norm": 0.47747498750686646,
+      "learning_rate": 0.00042343309859154934,
+      "loss": 1.7907,
+      "step": 10700
+    },
+    {
+      "epoch": 1.8282312925170068,
+      "grad_norm": 0.47017601132392883,
+      "learning_rate": 0.0004229929577464789,
+      "loss": 1.7855,
+      "step": 10750
+    },
+    {
+      "epoch": 1.836734693877551,
+      "grad_norm": 0.4648725092411041,
+      "learning_rate": 0.00042255281690140847,
+      "loss": 1.7878,
+      "step": 10800
+    },
+    {
+      "epoch": 1.8452380952380953,
+      "grad_norm": 0.4595974385738373,
+      "learning_rate": 0.000422112676056338,
+      "loss": 1.7808,
+      "step": 10850
+    },
+    {
+      "epoch": 1.8537414965986394,
+      "grad_norm": 0.5028242468833923,
+      "learning_rate": 0.00042167253521126766,
+      "loss": 1.7838,
+      "step": 10900
+    },
+    {
+      "epoch": 1.8622448979591837,
+      "grad_norm": 0.4698718190193176,
+      "learning_rate": 0.0004212323943661972,
+      "loss": 1.7838,
+      "step": 10950
+    },
+    {
+      "epoch": 1.870748299319728,
+      "grad_norm": 0.46658000349998474,
+      "learning_rate": 0.00042079225352112675,
+      "loss": 1.7884,
+      "step": 11000
+    },
+    {
+      "epoch": 1.870748299319728,
+      "eval_loss": 1.9652538299560547,
+      "eval_runtime": 75.3728,
+      "eval_samples_per_second": 1244.998,
+      "eval_steps_per_second": 4.869,
+      "step": 11000
+    },
+    {
+      "epoch": 1.879251700680272,
+      "grad_norm": 0.4587540924549103,
+      "learning_rate": 0.00042035211267605634,
+      "loss": 1.7795,
+      "step": 11050
+    },
+    {
+      "epoch": 1.8877551020408163,
+      "grad_norm": 0.48107513785362244,
+      "learning_rate": 0.00041991197183098594,
+      "loss": 1.7843,
+      "step": 11100
+    },
+    {
+      "epoch": 1.8962585034013606,
+      "grad_norm": 0.468262255191803,
+      "learning_rate": 0.00041947183098591553,
+      "loss": 1.7831,
+      "step": 11150
+    },
+    {
+      "epoch": 1.9047619047619047,
+      "grad_norm": 0.4687144458293915,
+      "learning_rate": 0.0004190316901408451,
+      "loss": 1.7761,
+      "step": 11200
+    },
+    {
+      "epoch": 1.913265306122449,
+      "grad_norm": 0.48110878467559814,
+      "learning_rate": 0.00041859154929577467,
+      "loss": 1.7658,
+      "step": 11250
+    },
+    {
+      "epoch": 1.9217687074829932,
+      "grad_norm": 0.47303488850593567,
+      "learning_rate": 0.0004181514084507042,
+      "loss": 1.7733,
+      "step": 11300
+    },
+    {
+      "epoch": 1.9302721088435373,
+      "grad_norm": 0.48464885354042053,
+      "learning_rate": 0.0004177112676056338,
+      "loss": 1.765,
+      "step": 11350
+    },
+    {
+      "epoch": 1.9387755102040818,
+      "grad_norm": 0.4908495247364044,
+      "learning_rate": 0.0004172799295774648,
+      "loss": 1.7662,
+      "step": 11400
+    },
+    {
+      "epoch": 1.9472789115646258,
+      "grad_norm": 0.47335630655288696,
+      "learning_rate": 0.00041683978873239435,
+      "loss": 1.7689,
+      "step": 11450
+    },
+    {
+      "epoch": 1.95578231292517,
+      "grad_norm": 0.4794527292251587,
+      "learning_rate": 0.00041639964788732395,
+      "loss": 1.7659,
+      "step": 11500
+    },
+    {
+      "epoch": 1.9642857142857144,
+      "grad_norm": 0.47013363242149353,
+      "learning_rate": 0.00041595950704225354,
+      "loss": 1.7683,
+      "step": 11550
+    },
+    {
+      "epoch": 1.9727891156462585,
+      "grad_norm": 0.4743632674217224,
+      "learning_rate": 0.00041551936619718314,
+      "loss": 1.7674,
+      "step": 11600
+    },
+    {
+      "epoch": 1.9812925170068028,
+      "grad_norm": 0.46521708369255066,
+      "learning_rate": 0.0004150792253521127,
+      "loss": 1.7686,
+      "step": 11650
+    },
+    {
+      "epoch": 1.989795918367347,
+      "grad_norm": 0.48240792751312256,
+      "learning_rate": 0.0004146390845070422,
+      "loss": 1.7688,
+      "step": 11700
+    },
+    {
+      "epoch": 1.998299319727891,
+      "grad_norm": 0.453975111246109,
+      "learning_rate": 0.00041419894366197187,
+      "loss": 1.7658,
+      "step": 11750
+    },
+    {
+      "epoch": 2.006802721088435,
+      "grad_norm": 0.45248886942863464,
+      "learning_rate": 0.0004137588028169014,
+      "loss": 1.7543,
+      "step": 11800
+    },
+    {
+      "epoch": 2.0153061224489797,
+      "grad_norm": 0.47196707129478455,
+      "learning_rate": 0.0004133274647887324,
+      "loss": 1.751,
+      "step": 11850
+    },
+    {
+      "epoch": 2.0238095238095237,
+      "grad_norm": 0.46333810687065125,
+      "learning_rate": 0.00041288732394366195,
+      "loss": 1.7544,
+      "step": 11900
+    },
+    {
+      "epoch": 2.0323129251700682,
+      "grad_norm": 0.46540361642837524,
+      "learning_rate": 0.0004124471830985916,
+      "loss": 1.7586,
+      "step": 11950
+    },
+    {
+      "epoch": 2.0408163265306123,
+      "grad_norm": 0.45927169919013977,
+      "learning_rate": 0.00041200704225352114,
+      "loss": 1.753,
+      "step": 12000
+    },
+    {
+      "epoch": 2.0408163265306123,
+      "eval_loss": 1.948580265045166,
+      "eval_runtime": 75.3406,
+      "eval_samples_per_second": 1245.53,
+      "eval_steps_per_second": 4.871,
+      "step": 12000
+    },
+    {
+      "epoch": 2.0493197278911564,
+      "grad_norm": 0.4919711947441101,
+      "learning_rate": 0.0004115669014084507,
+      "loss": 1.7501,
+      "step": 12050
+    },
+    {
+      "epoch": 2.057823129251701,
+      "grad_norm": 0.47275787591934204,
+      "learning_rate": 0.0004111267605633803,
+      "loss": 1.7579,
+      "step": 12100
+    },
+    {
+      "epoch": 2.066326530612245,
+      "grad_norm": 0.4734228551387787,
+      "learning_rate": 0.0004106866197183099,
+      "loss": 1.7445,
+      "step": 12150
+    },
+    {
+      "epoch": 2.074829931972789,
+      "grad_norm": 0.4777526259422302,
+      "learning_rate": 0.00041024647887323947,
+      "loss": 1.7558,
+      "step": 12200
+    },
+    {
+      "epoch": 2.0833333333333335,
+      "grad_norm": 0.4509522020816803,
+      "learning_rate": 0.000409806338028169,
+      "loss": 1.754,
+      "step": 12250
+    },
+    {
+      "epoch": 2.0918367346938775,
+      "grad_norm": 0.4808884859085083,
+      "learning_rate": 0.0004093661971830986,
+      "loss": 1.7487,
+      "step": 12300
+    },
+    {
+      "epoch": 2.1003401360544216,
+      "grad_norm": 0.46674636006355286,
+      "learning_rate": 0.00040892605633802815,
+      "loss": 1.7428,
+      "step": 12350
+    },
+    {
+      "epoch": 2.108843537414966,
+      "grad_norm": 0.4768598973751068,
+      "learning_rate": 0.00040848591549295775,
+      "loss": 1.7372,
+      "step": 12400
+    },
+    {
+      "epoch": 2.11734693877551,
+      "grad_norm": 0.4708414673805237,
+      "learning_rate": 0.00040804577464788734,
+      "loss": 1.7458,
+      "step": 12450
+    },
+    {
+      "epoch": 2.1258503401360542,
+      "grad_norm": 0.47017523646354675,
+      "learning_rate": 0.0004076056338028169,
+      "loss": 1.7497,
+      "step": 12500
+    },
+    {
+      "epoch": 2.1343537414965987,
+      "grad_norm": 0.4945257306098938,
+      "learning_rate": 0.0004071654929577465,
+      "loss": 1.7451,
+      "step": 12550
+    },
+    {
+      "epoch": 2.142857142857143,
+      "grad_norm": 0.4588756859302521,
+      "learning_rate": 0.00040672535211267607,
+      "loss": 1.7393,
+      "step": 12600
+    },
+    {
+      "epoch": 2.1513605442176873,
+      "grad_norm": 0.49352967739105225,
+      "learning_rate": 0.00040628521126760567,
+      "loss": 1.7366,
+      "step": 12650
+    },
+    {
+      "epoch": 2.1598639455782314,
+      "grad_norm": 0.47481125593185425,
+      "learning_rate": 0.0004058450704225352,
+      "loss": 1.7373,
+      "step": 12700
+    },
+    {
+      "epoch": 2.1683673469387754,
+      "grad_norm": 0.48250919580459595,
+      "learning_rate": 0.00040540492957746475,
+      "loss": 1.7409,
+      "step": 12750
+    },
+    {
+      "epoch": 2.17687074829932,
+      "grad_norm": 0.4601617157459259,
+      "learning_rate": 0.0004049647887323944,
+      "loss": 1.7413,
+      "step": 12800
+    },
+    {
+      "epoch": 2.185374149659864,
+      "grad_norm": 0.47254276275634766,
+      "learning_rate": 0.00040452464788732394,
+      "loss": 1.7495,
+      "step": 12850
+    },
+    {
+      "epoch": 2.193877551020408,
+      "grad_norm": 0.48082679510116577,
+      "learning_rate": 0.00040408450704225354,
+      "loss": 1.7429,
+      "step": 12900
+    },
+    {
+      "epoch": 2.2023809523809526,
+      "grad_norm": 0.46352913975715637,
+      "learning_rate": 0.0004036443661971831,
+      "loss": 1.7401,
+      "step": 12950
+    },
+    {
+      "epoch": 2.2108843537414966,
+      "grad_norm": 0.4566120207309723,
+      "learning_rate": 0.00040321302816901414,
+      "loss": 1.7337,
+      "step": 13000
+    },
+    {
+      "epoch": 2.2108843537414966,
+      "eval_loss": 1.921891689300537,
+      "eval_runtime": 75.3847,
+      "eval_samples_per_second": 1244.802,
+      "eval_steps_per_second": 4.868,
+      "step": 13000
+    },
+    {
+      "epoch": 2.2193877551020407,
+      "grad_norm": 0.5124084949493408,
+      "learning_rate": 0.0004027728873239437,
+      "loss": 1.741,
+      "step": 13050
+    },
+    {
+      "epoch": 2.227891156462585,
+      "grad_norm": 0.4604070782661438,
+      "learning_rate": 0.00040233274647887327,
+      "loss": 1.7372,
+      "step": 13100
+    },
+    {
+      "epoch": 2.2363945578231292,
+      "grad_norm": 0.49075254797935486,
+      "learning_rate": 0.0004019014084507042,
+      "loss": 1.7411,
+      "step": 13150
+    },
+    {
+      "epoch": 2.2448979591836733,
+      "grad_norm": 0.4861091077327728,
+      "learning_rate": 0.00040146126760563376,
+      "loss": 1.7418,
+      "step": 13200
+    },
+    {
+      "epoch": 2.253401360544218,
+      "grad_norm": 0.47723260521888733,
+      "learning_rate": 0.0004010211267605634,
+      "loss": 1.7375,
+      "step": 13250
+    },
+    {
+      "epoch": 2.261904761904762,
+      "grad_norm": 0.4777429401874542,
+      "learning_rate": 0.00040058098591549295,
+      "loss": 1.7305,
+      "step": 13300
+    },
+    {
+      "epoch": 2.270408163265306,
+      "grad_norm": 0.4744473993778229,
+      "learning_rate": 0.00040014084507042255,
+      "loss": 1.7291,
+      "step": 13350
+    },
+    {
+      "epoch": 2.2789115646258504,
+      "grad_norm": 0.46367913484573364,
+      "learning_rate": 0.0003997007042253521,
+      "loss": 1.7305,
+      "step": 13400
+    },
+    {
+      "epoch": 2.2874149659863945,
+      "grad_norm": 0.4830113649368286,
+      "learning_rate": 0.00039926056338028174,
+      "loss": 1.7298,
+      "step": 13450
+    },
+    {
+      "epoch": 2.295918367346939,
+      "grad_norm": 0.48178356885910034,
+      "learning_rate": 0.0003988204225352113,
+      "loss": 1.715,
+      "step": 13500
+    },
+    {
+      "epoch": 2.304421768707483,
+      "grad_norm": 0.46279290318489075,
+      "learning_rate": 0.0003983802816901408,
+      "loss": 1.7265,
+      "step": 13550
+    },
+    {
+      "epoch": 2.312925170068027,
+      "grad_norm": 0.4898684620857239,
+      "learning_rate": 0.0003979401408450704,
+      "loss": 1.7257,
+      "step": 13600
+    },
+    {
+      "epoch": 2.3214285714285716,
+      "grad_norm": 0.47215622663497925,
+      "learning_rate": 0.0003975,
+      "loss": 1.7291,
+      "step": 13650
+    },
+    {
+      "epoch": 2.3299319727891157,
+      "grad_norm": 0.4575498104095459,
+      "learning_rate": 0.0003970598591549296,
+      "loss": 1.7165,
+      "step": 13700
+    },
+    {
+      "epoch": 2.3384353741496597,
+      "grad_norm": 0.4656940996646881,
+      "learning_rate": 0.00039661971830985915,
+      "loss": 1.7183,
+      "step": 13750
+    },
+    {
+      "epoch": 2.3469387755102042,
+      "grad_norm": 0.4670725464820862,
+      "learning_rate": 0.00039617957746478874,
+      "loss": 1.7217,
+      "step": 13800
+    },
+    {
+      "epoch": 2.3554421768707483,
+      "grad_norm": 0.47170403599739075,
+      "learning_rate": 0.00039573943661971834,
+      "loss": 1.7255,
+      "step": 13850
+    },
+    {
+      "epoch": 2.3639455782312924,
+      "grad_norm": 0.49639126658439636,
+      "learning_rate": 0.0003952992957746479,
+      "loss": 1.7246,
+      "step": 13900
+    },
+    {
+      "epoch": 2.372448979591837,
+      "grad_norm": 0.4941042959690094,
+      "learning_rate": 0.0003948591549295775,
+      "loss": 1.7183,
+      "step": 13950
+    },
+    {
+      "epoch": 2.380952380952381,
+      "grad_norm": 0.4859507381916046,
+      "learning_rate": 0.000394419014084507,
+      "loss": 1.7239,
+      "step": 14000
+    },
+    {
+      "epoch": 2.380952380952381,
+      "eval_loss": 1.9032775163650513,
+      "eval_runtime": 75.3175,
+      "eval_samples_per_second": 1245.912,
+      "eval_steps_per_second": 4.873,
+      "step": 14000
+    },
+    {
+      "epoch": 2.389455782312925,
+      "grad_norm": 0.48601582646369934,
+      "learning_rate": 0.00039397887323943667,
+      "loss": 1.721,
+      "step": 14050
+    },
+    {
+      "epoch": 2.3979591836734695,
+      "grad_norm": 0.4685114920139313,
+      "learning_rate": 0.0003935387323943662,
+      "loss": 1.7155,
+      "step": 14100
+    },
+    {
+      "epoch": 2.4064625850340136,
+      "grad_norm": 0.44943785667419434,
+      "learning_rate": 0.0003930985915492958,
+      "loss": 1.7154,
+      "step": 14150
+    },
+    {
+      "epoch": 2.4149659863945576,
+      "grad_norm": 0.48964107036590576,
+      "learning_rate": 0.00039265845070422535,
+      "loss": 1.7123,
+      "step": 14200
+    },
+    {
+      "epoch": 2.423469387755102,
+      "grad_norm": 0.457769513130188,
+      "learning_rate": 0.00039221830985915494,
+      "loss": 1.7084,
+      "step": 14250
+    },
+    {
+      "epoch": 2.431972789115646,
+      "grad_norm": 0.5480780005455017,
+      "learning_rate": 0.00039177816901408454,
+      "loss": 1.7129,
+      "step": 14300
+    },
+    {
+      "epoch": 2.4404761904761907,
+      "grad_norm": 0.48546379804611206,
+      "learning_rate": 0.0003913380281690141,
+      "loss": 1.7078,
+      "step": 14350
+    },
+    {
+      "epoch": 2.4489795918367347,
+      "grad_norm": 0.46493494510650635,
+      "learning_rate": 0.00039089788732394367,
+      "loss": 1.7104,
+      "step": 14400
+    },
+    {
+      "epoch": 2.457482993197279,
+      "grad_norm": 0.4606137275695801,
+      "learning_rate": 0.0003904577464788732,
+      "loss": 1.7045,
+      "step": 14450
+    },
+    {
+      "epoch": 2.4659863945578233,
+      "grad_norm": 0.46123144030570984,
+      "learning_rate": 0.00039001760563380286,
+      "loss": 1.7047,
+      "step": 14500
+    },
+    {
+      "epoch": 2.4744897959183674,
+      "grad_norm": 0.4724046289920807,
+      "learning_rate": 0.0003895774647887324,
+      "loss": 1.711,
+      "step": 14550
+    },
+    {
+      "epoch": 2.4829931972789114,
+      "grad_norm": 0.47036615014076233,
+      "learning_rate": 0.00038913732394366195,
+      "loss": 1.7077,
+      "step": 14600
+    },
+    {
+      "epoch": 2.491496598639456,
+      "grad_norm": 0.465732216835022,
+      "learning_rate": 0.00038869718309859154,
+      "loss": 1.7041,
+      "step": 14650
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.4794946610927582,
+      "learning_rate": 0.00038825704225352114,
+      "loss": 1.7037,
+      "step": 14700
+    },
+    {
+      "epoch": 2.508503401360544,
+      "grad_norm": 0.47816696763038635,
+      "learning_rate": 0.00038781690140845073,
+      "loss": 1.7056,
+      "step": 14750
+    },
+    {
+      "epoch": 2.5170068027210886,
+      "grad_norm": 0.47812485694885254,
+      "learning_rate": 0.0003873767605633803,
+      "loss": 1.7063,
+      "step": 14800
+    },
+    {
+      "epoch": 2.5255102040816326,
+      "grad_norm": 0.4651651084423065,
+      "learning_rate": 0.00038693661971830987,
+      "loss": 1.693,
+      "step": 14850
+    },
+    {
+      "epoch": 2.534013605442177,
+      "grad_norm": 0.48417726159095764,
+      "learning_rate": 0.00038649647887323946,
+      "loss": 1.697,
+      "step": 14900
+    },
+    {
+      "epoch": 2.542517006802721,
+      "grad_norm": 0.4502563178539276,
+      "learning_rate": 0.000386056338028169,
+      "loss": 1.703,
+      "step": 14950
+    },
+    {
+      "epoch": 2.5510204081632653,
+      "grad_norm": 0.45644158124923706,
+      "learning_rate": 0.0003856161971830986,
+      "loss": 1.6959,
+      "step": 15000
+    },
+    {
+      "epoch": 2.5510204081632653,
+      "eval_loss": 1.8780722618103027,
+      "eval_runtime": 75.398,
+      "eval_samples_per_second": 1244.582,
+      "eval_steps_per_second": 4.868,
+      "step": 15000
+    },
+    {
+      "epoch": 2.5595238095238093,
+      "grad_norm": 0.4715486764907837,
+      "learning_rate": 0.00038517605633802814,
+      "loss": 1.7015,
+      "step": 15050
+    },
+    {
+      "epoch": 2.568027210884354,
+      "grad_norm": 0.44938182830810547,
+      "learning_rate": 0.0003847359154929578,
+      "loss": 1.6954,
+      "step": 15100
+    },
+    {
+      "epoch": 2.576530612244898,
+      "grad_norm": 0.4778023660182953,
+      "learning_rate": 0.00038429577464788733,
+      "loss": 1.6973,
+      "step": 15150
+    },
+    {
+      "epoch": 2.5850340136054424,
+      "grad_norm": 0.4602490961551666,
+      "learning_rate": 0.00038385563380281693,
+      "loss": 1.7041,
+      "step": 15200
+    },
+    {
+      "epoch": 2.5935374149659864,
+      "grad_norm": 0.4905562400817871,
+      "learning_rate": 0.0003834242957746479,
+      "loss": 1.6929,
+      "step": 15250
+    },
+    {
+      "epoch": 2.6020408163265305,
+      "grad_norm": 0.47374477982521057,
+      "learning_rate": 0.00038298415492957747,
+      "loss": 1.6963,
+      "step": 15300
+    },
+    {
+      "epoch": 2.6105442176870746,
+      "grad_norm": 0.45640283823013306,
+      "learning_rate": 0.00038254401408450707,
+      "loss": 1.6941,
+      "step": 15350
+    },
+    {
+      "epoch": 2.619047619047619,
+      "grad_norm": 0.46338459849357605,
+      "learning_rate": 0.0003821038732394366,
+      "loss": 1.6898,
+      "step": 15400
+    },
+    {
+      "epoch": 2.627551020408163,
+      "grad_norm": 0.458769291639328,
+      "learning_rate": 0.0003816637323943662,
+      "loss": 1.6897,
+      "step": 15450
+    },
+    {
+      "epoch": 2.6360544217687076,
+      "grad_norm": 0.44556647539138794,
+      "learning_rate": 0.00038122359154929575,
+      "loss": 1.6918,
+      "step": 15500
+    },
+    {
+      "epoch": 2.6445578231292517,
+      "grad_norm": 0.48705628514289856,
+      "learning_rate": 0.0003807834507042254,
+      "loss": 1.6948,
+      "step": 15550
+    },
+    {
+      "epoch": 2.6530612244897958,
+      "grad_norm": 0.46732062101364136,
+      "learning_rate": 0.00038034330985915494,
+      "loss": 1.6884,
+      "step": 15600
+    },
+    {
+      "epoch": 2.6615646258503403,
+      "grad_norm": 0.44862720370292664,
+      "learning_rate": 0.0003799031690140845,
+      "loss": 1.6921,
+      "step": 15650
+    },
+    {
+      "epoch": 2.6700680272108843,
+      "grad_norm": 0.4507545828819275,
+      "learning_rate": 0.0003794630281690141,
+      "loss": 1.6841,
+      "step": 15700
+    },
+    {
+      "epoch": 2.678571428571429,
+      "grad_norm": 0.4554910659790039,
+      "learning_rate": 0.00037902288732394367,
+      "loss": 1.6898,
+      "step": 15750
+    },
+    {
+      "epoch": 2.687074829931973,
+      "grad_norm": 0.47400516271591187,
+      "learning_rate": 0.00037858274647887326,
+      "loss": 1.6875,
+      "step": 15800
+    },
+    {
+      "epoch": 2.695578231292517,
+      "grad_norm": 0.46053117513656616,
+      "learning_rate": 0.0003781426056338028,
+      "loss": 1.6791,
+      "step": 15850
+    },
+    {
+      "epoch": 2.704081632653061,
+      "grad_norm": 0.4834235608577728,
+      "learning_rate": 0.0003777024647887324,
+      "loss": 1.6883,
+      "step": 15900
+    },
+    {
+      "epoch": 2.7125850340136055,
+      "grad_norm": 0.648454487323761,
+      "learning_rate": 0.000377262323943662,
+      "loss": 1.686,
+      "step": 15950
+    },
+    {
+      "epoch": 2.7210884353741496,
+      "grad_norm": 0.46427688002586365,
+      "learning_rate": 0.0003768221830985916,
+      "loss": 1.6771,
+      "step": 16000
+    },
+    {
+      "epoch": 2.7210884353741496,
+      "eval_loss": 1.8668512105941772,
+      "eval_runtime": 75.2617,
+      "eval_samples_per_second": 1246.836,
+      "eval_steps_per_second": 4.876,
+      "step": 16000
+    },
+    {
+      "epoch": 2.729591836734694,
+      "grad_norm": 0.4748222827911377,
+      "learning_rate": 0.00037638204225352113,
+      "loss": 1.6826,
+      "step": 16050
+    },
+    {
+      "epoch": 2.738095238095238,
+      "grad_norm": 0.4745228588581085,
+      "learning_rate": 0.0003759419014084507,
+      "loss": 1.6837,
+      "step": 16100
+    },
+    {
+      "epoch": 2.746598639455782,
+      "grad_norm": 0.4836510419845581,
+      "learning_rate": 0.0003755017605633803,
+      "loss": 1.6803,
+      "step": 16150
+    },
+    {
+      "epoch": 2.7551020408163263,
+      "grad_norm": 0.48129406571388245,
+      "learning_rate": 0.00037506161971830986,
+      "loss": 1.6804,
+      "step": 16200
+    },
+    {
+      "epoch": 2.7636054421768708,
+      "grad_norm": 0.45295995473861694,
+      "learning_rate": 0.00037462147887323946,
+      "loss": 1.6763,
+      "step": 16250
+    },
+    {
+      "epoch": 2.772108843537415,
+      "grad_norm": 0.45548656582832336,
+      "learning_rate": 0.000374181338028169,
+      "loss": 1.6793,
+      "step": 16300
+    },
+    {
+      "epoch": 2.7806122448979593,
+      "grad_norm": 0.46522030234336853,
+      "learning_rate": 0.00037374119718309865,
+      "loss": 1.6792,
+      "step": 16350
+    },
+    {
+      "epoch": 2.7891156462585034,
+      "grad_norm": 0.46659553050994873,
+      "learning_rate": 0.0003733010563380282,
+      "loss": 1.6705,
+      "step": 16400
+    },
+    {
+      "epoch": 2.7976190476190474,
+      "grad_norm": 0.4808714985847473,
+      "learning_rate": 0.00037286091549295773,
+      "loss": 1.6751,
+      "step": 16450
+    },
+    {
+      "epoch": 2.806122448979592,
+      "grad_norm": 0.46006572246551514,
+      "learning_rate": 0.00037242077464788733,
+      "loss": 1.6708,
+      "step": 16500
+    },
+    {
+      "epoch": 2.814625850340136,
+      "grad_norm": 0.4552723467350006,
+      "learning_rate": 0.0003719806338028169,
+      "loss": 1.6765,
+      "step": 16550
+    },
+    {
+      "epoch": 2.8231292517006805,
+      "grad_norm": 0.4788212180137634,
+      "learning_rate": 0.0003715404929577465,
+      "loss": 1.6774,
+      "step": 16600
+    },
+    {
+      "epoch": 2.8316326530612246,
+      "grad_norm": 0.4770471453666687,
+      "learning_rate": 0.00037110035211267606,
+      "loss": 1.6703,
+      "step": 16650
+    },
+    {
+      "epoch": 2.8401360544217686,
+      "grad_norm": 0.4665678143501282,
+      "learning_rate": 0.00037066021126760566,
+      "loss": 1.6754,
+      "step": 16700
+    },
+    {
+      "epoch": 2.8486394557823127,
+      "grad_norm": 0.46680372953414917,
+      "learning_rate": 0.0003702200704225352,
+      "loss": 1.67,
+      "step": 16750
+    },
+    {
+      "epoch": 2.857142857142857,
+      "grad_norm": 0.4522228240966797,
+      "learning_rate": 0.0003697799295774648,
+      "loss": 1.6718,
+      "step": 16800
+    },
+    {
+      "epoch": 2.8656462585034013,
+      "grad_norm": 0.4507662057876587,
+      "learning_rate": 0.0003693397887323944,
+      "loss": 1.6713,
+      "step": 16850
+    },
+    {
+      "epoch": 2.8741496598639458,
+      "grad_norm": 0.45078015327453613,
+      "learning_rate": 0.00036889964788732393,
+      "loss": 1.6641,
+      "step": 16900
+    },
+    {
+      "epoch": 2.88265306122449,
+      "grad_norm": 0.46813851594924927,
+      "learning_rate": 0.0003684595070422535,
+      "loss": 1.6684,
+      "step": 16950
+    },
+    {
+      "epoch": 2.891156462585034,
+      "grad_norm": 0.4676934480667114,
+      "learning_rate": 0.0003680193661971831,
+      "loss": 1.6678,
+      "step": 17000
+    },
+    {
+      "epoch": 2.891156462585034,
+      "eval_loss": 1.8514577150344849,
+      "eval_runtime": 75.2869,
+      "eval_samples_per_second": 1246.419,
+      "eval_steps_per_second": 4.875,
+      "step": 17000
+    },
+    {
+      "epoch": 2.8996598639455784,
+      "grad_norm": 0.4356513023376465,
+      "learning_rate": 0.0003675792253521127,
+      "loss": 1.6586,
+      "step": 17050
+    },
+    {
+      "epoch": 2.9081632653061225,
+      "grad_norm": 0.44688156247138977,
+      "learning_rate": 0.00036713908450704226,
+      "loss": 1.6596,
+      "step": 17100
+    },
+    {
+      "epoch": 2.9166666666666665,
+      "grad_norm": 0.46431487798690796,
+      "learning_rate": 0.0003666989436619718,
+      "loss": 1.6682,
+      "step": 17150
+    },
+    {
+      "epoch": 2.925170068027211,
+      "grad_norm": 0.44890233874320984,
+      "learning_rate": 0.00036625880281690145,
+      "loss": 1.6594,
+      "step": 17200
+    },
+    {
+      "epoch": 2.933673469387755,
+      "grad_norm": 0.45430788397789,
+      "learning_rate": 0.000365818661971831,
+      "loss": 1.6617,
+      "step": 17250
+    },
+    {
+      "epoch": 2.942176870748299,
+      "grad_norm": 0.4674071967601776,
+      "learning_rate": 0.0003653785211267606,
+      "loss": 1.6583,
+      "step": 17300
+    },
+    {
+      "epoch": 2.9506802721088436,
+      "grad_norm": 1.025103211402893,
+      "learning_rate": 0.0003649383802816901,
+      "loss": 1.6637,
+      "step": 17350
+    },
+    {
+      "epoch": 2.9591836734693877,
+      "grad_norm": 0.5043914318084717,
+      "learning_rate": 0.0003645070422535212,
+      "loss": 1.6673,
+      "step": 17400
+    },
+    {
+      "epoch": 2.967687074829932,
+      "grad_norm": 0.4789212942123413,
+      "learning_rate": 0.0003640669014084507,
+      "loss": 1.665,
+      "step": 17450
+    },
+    {
+      "epoch": 2.9761904761904763,
+      "grad_norm": 0.458146870136261,
+      "learning_rate": 0.00036362676056338027,
+      "loss": 1.6659,
+      "step": 17500
+    },
+    {
+      "epoch": 2.9846938775510203,
+      "grad_norm": 0.45948123931884766,
+      "learning_rate": 0.00036319542253521127,
+      "loss": 1.6559,
+      "step": 17550
+    },
+    {
+      "epoch": 2.9931972789115644,
+      "grad_norm": 0.44673070311546326,
+      "learning_rate": 0.0003627552816901408,
+      "loss": 1.6596,
+      "step": 17600
+    },
+    {
+      "epoch": 3.001700680272109,
+      "grad_norm": 0.46580860018730164,
+      "learning_rate": 0.00036231514084507046,
+      "loss": 1.6646,
+      "step": 17650
+    },
+    {
+      "epoch": 3.010204081632653,
+      "grad_norm": 0.48107218742370605,
+      "learning_rate": 0.000361875,
+      "loss": 1.6516,
+      "step": 17700
+    },
+    {
+      "epoch": 3.0187074829931975,
+      "grad_norm": 0.4705383777618408,
+      "learning_rate": 0.0003614348591549296,
+      "loss": 1.6506,
+      "step": 17750
+    },
+    {
+      "epoch": 3.0272108843537415,
+      "grad_norm": 0.4649271070957184,
+      "learning_rate": 0.00036099471830985914,
+      "loss": 1.6523,
+      "step": 17800
+    },
+    {
+      "epoch": 3.0357142857142856,
+      "grad_norm": 0.4697044789791107,
+      "learning_rate": 0.00036055457746478873,
+      "loss": 1.6556,
+      "step": 17850
+    },
+    {
+      "epoch": 3.04421768707483,
+      "grad_norm": 0.441441148519516,
+      "learning_rate": 0.00036011443661971833,
+      "loss": 1.6483,
+      "step": 17900
+    },
+    {
+      "epoch": 3.052721088435374,
+      "grad_norm": 0.4512791037559509,
+      "learning_rate": 0.00035967429577464787,
+      "loss": 1.655,
+      "step": 17950
+    },
+    {
+      "epoch": 3.061224489795918,
+      "grad_norm": 0.47064149379730225,
+      "learning_rate": 0.00035923415492957746,
+      "loss": 1.6493,
+      "step": 18000
+    },
+    {
+      "epoch": 3.061224489795918,
+      "eval_loss": 1.8354525566101074,
+      "eval_runtime": 75.3083,
+      "eval_samples_per_second": 1246.064,
+      "eval_steps_per_second": 4.873,
+      "step": 18000
+    },
+    {
+      "epoch": 3.0697278911564627,
+      "grad_norm": 0.4441589415073395,
+      "learning_rate": 0.00035879401408450706,
+      "loss": 1.6469,
+      "step": 18050
+    },
+    {
+      "epoch": 3.078231292517007,
+      "grad_norm": 0.46877768635749817,
+      "learning_rate": 0.00035835387323943666,
+      "loss": 1.6449,
+      "step": 18100
+    },
+    {
+      "epoch": 3.086734693877551,
+      "grad_norm": 0.9494704008102417,
+      "learning_rate": 0.0003579137323943662,
+      "loss": 1.6426,
+      "step": 18150
+    },
+    {
+      "epoch": 3.0952380952380953,
+      "grad_norm": 0.4520435631275177,
+      "learning_rate": 0.00035747359154929574,
+      "loss": 1.6521,
+      "step": 18200
+    },
+    {
+      "epoch": 3.1037414965986394,
+      "grad_norm": 0.46680590510368347,
+      "learning_rate": 0.0003570334507042254,
+      "loss": 1.655,
+      "step": 18250
+    },
+    {
+      "epoch": 3.1122448979591835,
+      "grad_norm": 0.47364234924316406,
+      "learning_rate": 0.00035659330985915493,
+      "loss": 1.6404,
+      "step": 18300
+    },
+    {
+      "epoch": 3.120748299319728,
+      "grad_norm": 0.47683462500572205,
+      "learning_rate": 0.0003561531690140845,
+      "loss": 1.6432,
+      "step": 18350
+    },
+    {
+      "epoch": 3.129251700680272,
+      "grad_norm": 0.4633096158504486,
+      "learning_rate": 0.00035571302816901407,
+      "loss": 1.6459,
+      "step": 18400
+    },
+    {
+      "epoch": 3.137755102040816,
+      "grad_norm": 0.4648244380950928,
+      "learning_rate": 0.0003552728873239437,
+      "loss": 1.6392,
+      "step": 18450
+    },
+    {
+      "epoch": 3.1462585034013606,
+      "grad_norm": 0.46435174345970154,
+      "learning_rate": 0.00035483274647887326,
+      "loss": 1.6396,
+      "step": 18500
+    },
+    {
+      "epoch": 3.1547619047619047,
+      "grad_norm": 0.45293596386909485,
+      "learning_rate": 0.0003543926056338028,
+      "loss": 1.6413,
+      "step": 18550
+    },
+    {
+      "epoch": 3.163265306122449,
+      "grad_norm": 0.46072807908058167,
+      "learning_rate": 0.0003539524647887324,
+      "loss": 1.6379,
+      "step": 18600
+    },
+    {
+      "epoch": 3.171768707482993,
+      "grad_norm": 0.45269516110420227,
+      "learning_rate": 0.000353512323943662,
+      "loss": 1.6409,
+      "step": 18650
+    },
+    {
+      "epoch": 3.1802721088435373,
+      "grad_norm": 0.4461529850959778,
+      "learning_rate": 0.0003530721830985916,
+      "loss": 1.6373,
+      "step": 18700
+    },
+    {
+      "epoch": 3.188775510204082,
+      "grad_norm": 0.45222413539886475,
+      "learning_rate": 0.0003526320422535211,
+      "loss": 1.6317,
+      "step": 18750
+    },
+    {
+      "epoch": 3.197278911564626,
+      "grad_norm": 0.44513842463493347,
+      "learning_rate": 0.0003521919014084507,
+      "loss": 1.6365,
+      "step": 18800
+    },
+    {
+      "epoch": 3.20578231292517,
+      "grad_norm": 0.4705728590488434,
+      "learning_rate": 0.00035175176056338026,
+      "loss": 1.6434,
+      "step": 18850
+    },
+    {
+      "epoch": 3.2142857142857144,
+      "grad_norm": 0.47750329971313477,
+      "learning_rate": 0.00035131161971830986,
+      "loss": 1.6337,
+      "step": 18900
+    },
+    {
+      "epoch": 3.2227891156462585,
+      "grad_norm": 0.4608128070831299,
+      "learning_rate": 0.00035087147887323945,
+      "loss": 1.642,
+      "step": 18950
+    },
+    {
+      "epoch": 3.2312925170068025,
+      "grad_norm": 0.46269285678863525,
+      "learning_rate": 0.000350431338028169,
+      "loss": 1.6326,
+      "step": 19000
+    },
+    {
+      "epoch": 3.2312925170068025,
+      "eval_loss": 1.8231135606765747,
+      "eval_runtime": 75.3579,
+      "eval_samples_per_second": 1245.245,
+      "eval_steps_per_second": 4.87,
+      "step": 19000
+    },
+    {
+      "epoch": 3.239795918367347,
+      "grad_norm": 0.44600149989128113,
+      "learning_rate": 0.0003499911971830986,
+      "loss": 1.6372,
+      "step": 19050
+    },
+    {
+      "epoch": 3.248299319727891,
+      "grad_norm": 0.45678427815437317,
+      "learning_rate": 0.0003495510563380282,
+      "loss": 1.6348,
+      "step": 19100
+    },
+    {
+      "epoch": 3.2568027210884356,
+      "grad_norm": 0.45453667640686035,
+      "learning_rate": 0.0003491109154929578,
+      "loss": 1.6422,
+      "step": 19150
+    },
+    {
+      "epoch": 3.2653061224489797,
+      "grad_norm": 0.4784705936908722,
+      "learning_rate": 0.0003486707746478873,
+      "loss": 1.6382,
+      "step": 19200
+    },
+    {
+      "epoch": 3.2738095238095237,
+      "grad_norm": 0.4635883569717407,
+      "learning_rate": 0.00034823063380281686,
+      "loss": 1.6388,
+      "step": 19250
+    },
+    {
+      "epoch": 3.282312925170068,
+      "grad_norm": 0.45983946323394775,
+      "learning_rate": 0.0003477904929577465,
+      "loss": 1.6408,
+      "step": 19300
+    },
+    {
+      "epoch": 3.2908163265306123,
+      "grad_norm": 0.4456377923488617,
+      "learning_rate": 0.00034735035211267605,
+      "loss": 1.6388,
+      "step": 19350
+    },
+    {
+      "epoch": 3.2993197278911564,
+      "grad_norm": 0.4579452872276306,
+      "learning_rate": 0.00034691021126760565,
+      "loss": 1.6283,
+      "step": 19400
+    },
+    {
+      "epoch": 3.307823129251701,
+      "grad_norm": 0.43990224599838257,
+      "learning_rate": 0.0003464700704225352,
+      "loss": 1.6327,
+      "step": 19450
+    },
+    {
+      "epoch": 3.316326530612245,
+      "grad_norm": 0.4531456530094147,
+      "learning_rate": 0.00034602992957746484,
+      "loss": 1.6302,
+      "step": 19500
+    },
+    {
+      "epoch": 3.324829931972789,
+      "grad_norm": 0.4910786747932434,
+      "learning_rate": 0.0003455897887323944,
+      "loss": 1.6275,
+      "step": 19550
+    },
+    {
+      "epoch": 3.3333333333333335,
+      "grad_norm": 0.46175333857536316,
+      "learning_rate": 0.0003451496478873239,
+      "loss": 1.6349,
+      "step": 19600
+    },
+    {
+      "epoch": 3.3418367346938775,
+      "grad_norm": 0.4583819508552551,
+      "learning_rate": 0.0003447095070422535,
+      "loss": 1.6279,
+      "step": 19650
+    },
+    {
+      "epoch": 3.3503401360544216,
+      "grad_norm": 0.4596096873283386,
+      "learning_rate": 0.0003442693661971831,
+      "loss": 1.6203,
+      "step": 19700
+    },
+    {
+      "epoch": 3.358843537414966,
+      "grad_norm": 0.476151704788208,
+      "learning_rate": 0.0003438292253521127,
+      "loss": 1.63,
+      "step": 19750
+    },
+    {
+      "epoch": 3.36734693877551,
+      "grad_norm": 0.45185205340385437,
+      "learning_rate": 0.00034338908450704225,
+      "loss": 1.6228,
+      "step": 19800
+    },
+    {
+      "epoch": 3.3758503401360542,
+      "grad_norm": 0.4486692249774933,
+      "learning_rate": 0.00034294894366197184,
+      "loss": 1.623,
+      "step": 19850
+    },
+    {
+      "epoch": 3.3843537414965987,
+      "grad_norm": 0.4323517084121704,
+      "learning_rate": 0.0003425088028169014,
+      "loss": 1.6203,
+      "step": 19900
+    },
+    {
+      "epoch": 3.392857142857143,
+      "grad_norm": 0.4673045575618744,
+      "learning_rate": 0.00034206866197183103,
+      "loss": 1.6229,
+      "step": 19950
+    },
+    {
+      "epoch": 3.4013605442176873,
+      "grad_norm": 0.5606664419174194,
+      "learning_rate": 0.0003416285211267606,
+      "loss": 1.6248,
+      "step": 20000
+    },
+    {
+      "epoch": 3.4013605442176873,
+      "eval_loss": 1.8098527193069458,
+      "eval_runtime": 75.3357,
+      "eval_samples_per_second": 1245.611,
+      "eval_steps_per_second": 4.872,
+      "step": 20000
+    },
+    {
+      "epoch": 3.4098639455782314,
+      "grad_norm": 0.4416612386703491,
+      "learning_rate": 0.0003411883802816901,
+      "loss": 1.6244,
+      "step": 20050
+    },
+    {
+      "epoch": 3.4183673469387754,
+      "grad_norm": 0.46314534544944763,
+      "learning_rate": 0.0003407482394366197,
+      "loss": 1.619,
+      "step": 20100
+    },
+    {
+      "epoch": 3.4268707482993195,
+      "grad_norm": 0.46182000637054443,
+      "learning_rate": 0.0003403080985915493,
+      "loss": 1.6198,
+      "step": 20150
+    },
+    {
+      "epoch": 3.435374149659864,
+      "grad_norm": 0.46425795555114746,
+      "learning_rate": 0.0003398679577464789,
+      "loss": 1.6198,
+      "step": 20200
+    },
+    {
+      "epoch": 3.443877551020408,
+      "grad_norm": 0.4472617208957672,
+      "learning_rate": 0.00033942781690140844,
+      "loss": 1.6208,
+      "step": 20250
+    },
+    {
+      "epoch": 3.4523809523809526,
+      "grad_norm": 0.4484054446220398,
+      "learning_rate": 0.000338987676056338,
+      "loss": 1.617,
+      "step": 20300
+    },
+    {
+      "epoch": 3.4608843537414966,
+      "grad_norm": 0.46471381187438965,
+      "learning_rate": 0.00033855633802816904,
+      "loss": 1.6235,
+      "step": 20350
+    },
+    {
+      "epoch": 3.4693877551020407,
+      "grad_norm": 0.471675306558609,
+      "learning_rate": 0.0003381161971830986,
+      "loss": 1.6211,
+      "step": 20400
+    },
+    {
+      "epoch": 3.477891156462585,
+      "grad_norm": 0.4538419246673584,
+      "learning_rate": 0.0003376760563380282,
+      "loss": 1.6159,
+      "step": 20450
+    },
+    {
+      "epoch": 3.4863945578231292,
+      "grad_norm": 0.4715994894504547,
+      "learning_rate": 0.0003372359154929577,
+      "loss": 1.6155,
+      "step": 20500
+    },
+    {
+      "epoch": 3.4948979591836733,
+      "grad_norm": 0.44585472345352173,
+      "learning_rate": 0.00033679577464788737,
+      "loss": 1.6169,
+      "step": 20550
+    },
+    {
+      "epoch": 3.503401360544218,
+      "grad_norm": 0.4732266962528229,
+      "learning_rate": 0.0003363556338028169,
+      "loss": 1.6175,
+      "step": 20600
+    },
+    {
+      "epoch": 3.511904761904762,
+      "grad_norm": 0.4438916742801666,
+      "learning_rate": 0.0003359154929577465,
+      "loss": 1.6179,
+      "step": 20650
+    },
+    {
+      "epoch": 3.520408163265306,
+      "grad_norm": 0.46406659483909607,
+      "learning_rate": 0.00033547535211267605,
+      "loss": 1.6169,
+      "step": 20700
+    },
+    {
+      "epoch": 3.5289115646258504,
+      "grad_norm": 0.438435435295105,
+      "learning_rate": 0.00033503521126760564,
+      "loss": 1.6155,
+      "step": 20750
+    },
+    {
+      "epoch": 3.5374149659863945,
+      "grad_norm": 0.4469165503978729,
+      "learning_rate": 0.00033459507042253524,
+      "loss": 1.6124,
+      "step": 20800
+    },
+    {
+      "epoch": 3.545918367346939,
+      "grad_norm": 0.4506942927837372,
+      "learning_rate": 0.0003341549295774648,
+      "loss": 1.6128,
+      "step": 20850
+    },
+    {
+      "epoch": 3.554421768707483,
+      "grad_norm": 0.4475231468677521,
+      "learning_rate": 0.0003337147887323944,
+      "loss": 1.6118,
+      "step": 20900
+    },
+    {
+      "epoch": 3.562925170068027,
+      "grad_norm": 0.45778709650039673,
+      "learning_rate": 0.0003332746478873239,
+      "loss": 1.6098,
+      "step": 20950
+    },
+    {
+      "epoch": 3.571428571428571,
+      "grad_norm": 0.44058629870414734,
+      "learning_rate": 0.00033283450704225357,
+      "loss": 1.6153,
+      "step": 21000
+    },
+    {
+      "epoch": 3.571428571428571,
+      "eval_loss": 1.791085124015808,
+      "eval_runtime": 75.3212,
+      "eval_samples_per_second": 1245.85,
+      "eval_steps_per_second": 4.872,
+      "step": 21000
+    },
+    {
+      "epoch": 3.5799319727891157,
+      "grad_norm": 0.44585561752319336,
+      "learning_rate": 0.0003323943661971831,
+      "loss": 1.6093,
+      "step": 21050
+    },
+    {
+      "epoch": 3.5884353741496597,
+      "grad_norm": 0.4633313715457916,
+      "learning_rate": 0.00033195422535211265,
+      "loss": 1.608,
+      "step": 21100
+    },
+    {
+      "epoch": 3.5969387755102042,
+      "grad_norm": 0.4513951539993286,
+      "learning_rate": 0.00033151408450704224,
+      "loss": 1.6066,
+      "step": 21150
+    },
+    {
+      "epoch": 3.6054421768707483,
+      "grad_norm": 0.44720685482025146,
+      "learning_rate": 0.00033107394366197184,
+      "loss": 1.6104,
+      "step": 21200
+    },
+    {
+      "epoch": 3.6139455782312924,
+      "grad_norm": 0.4755708873271942,
+      "learning_rate": 0.00033063380281690144,
+      "loss": 1.6138,
+      "step": 21250
+    },
+    {
+      "epoch": 3.622448979591837,
+      "grad_norm": 0.46023187041282654,
+      "learning_rate": 0.000330193661971831,
+      "loss": 1.6087,
+      "step": 21300
+    },
+    {
+      "epoch": 3.630952380952381,
+      "grad_norm": 0.45014435052871704,
+      "learning_rate": 0.00032975352112676057,
+      "loss": 1.606,
+      "step": 21350
+    },
+    {
+      "epoch": 3.6394557823129254,
+      "grad_norm": 0.440222829580307,
+      "learning_rate": 0.00032931338028169017,
+      "loss": 1.606,
+      "step": 21400
+    },
+    {
+      "epoch": 3.6479591836734695,
+      "grad_norm": 0.46397948265075684,
+      "learning_rate": 0.0003288732394366197,
+      "loss": 1.6085,
+      "step": 21450
+    },
+    {
+      "epoch": 3.6564625850340136,
+      "grad_norm": 0.5181043744087219,
+      "learning_rate": 0.0003284330985915493,
+      "loss": 1.6098,
+      "step": 21500
+    },
+    {
+      "epoch": 3.6649659863945576,
+      "grad_norm": 0.4351687431335449,
+      "learning_rate": 0.00032799295774647884,
+      "loss": 1.6087,
+      "step": 21550
+    },
+    {
+      "epoch": 3.673469387755102,
+      "grad_norm": 0.4317842721939087,
+      "learning_rate": 0.0003275528169014085,
+      "loss": 1.6068,
+      "step": 21600
+    },
+    {
+      "epoch": 3.681972789115646,
+      "grad_norm": 0.5115000009536743,
+      "learning_rate": 0.00032711267605633804,
+      "loss": 1.5992,
+      "step": 21650
+    },
+    {
+      "epoch": 3.6904761904761907,
+      "grad_norm": 0.4375057816505432,
+      "learning_rate": 0.00032667253521126763,
+      "loss": 1.6026,
+      "step": 21700
+    },
+    {
+      "epoch": 3.6989795918367347,
+      "grad_norm": 0.4488801956176758,
+      "learning_rate": 0.00032623239436619717,
+      "loss": 1.6057,
+      "step": 21750
+    },
+    {
+      "epoch": 3.707482993197279,
+      "grad_norm": 0.46254992485046387,
+      "learning_rate": 0.00032579225352112677,
+      "loss": 1.6031,
+      "step": 21800
+    },
+    {
+      "epoch": 3.715986394557823,
+      "grad_norm": 0.43931296467781067,
+      "learning_rate": 0.00032535211267605636,
+      "loss": 1.6035,
+      "step": 21850
+    },
+    {
+      "epoch": 3.7244897959183674,
+      "grad_norm": 0.4594497084617615,
+      "learning_rate": 0.0003249119718309859,
+      "loss": 1.6051,
+      "step": 21900
+    },
+    {
+      "epoch": 3.7329931972789114,
+      "grad_norm": 0.45316246151924133,
+      "learning_rate": 0.0003244718309859155,
+      "loss": 1.5937,
+      "step": 21950
+    },
+    {
+      "epoch": 3.741496598639456,
+      "grad_norm": 0.4681023359298706,
+      "learning_rate": 0.0003240316901408451,
+      "loss": 1.6015,
+      "step": 22000
+    },
+    {
+      "epoch": 3.741496598639456,
+      "eval_loss": 1.7835361957550049,
+      "eval_runtime": 75.3121,
+      "eval_samples_per_second": 1246.002,
+      "eval_steps_per_second": 4.873,
+      "step": 22000
+    },
+    {
+      "epoch": 3.75,
+      "grad_norm": 0.5095068216323853,
+      "learning_rate": 0.0003235915492957747,
+      "loss": 1.6031,
+      "step": 22050
+    },
+    {
+      "epoch": 3.758503401360544,
+      "grad_norm": 0.43521663546562195,
+      "learning_rate": 0.00032315140845070423,
+      "loss": 1.6038,
+      "step": 22100
+    },
+    {
+      "epoch": 3.7670068027210886,
+      "grad_norm": 0.4796842932701111,
+      "learning_rate": 0.00032271126760563377,
+      "loss": 1.6061,
+      "step": 22150
+    },
+    {
+      "epoch": 3.7755102040816326,
+      "grad_norm": 0.45405638217926025,
+      "learning_rate": 0.00032227112676056337,
+      "loss": 1.5938,
+      "step": 22200
+    },
+    {
+      "epoch": 3.784013605442177,
+      "grad_norm": 0.46786293387413025,
+      "learning_rate": 0.00032183098591549296,
+      "loss": 1.6031,
+      "step": 22250
+    },
+    {
+      "epoch": 3.792517006802721,
+      "grad_norm": 0.43882298469543457,
+      "learning_rate": 0.00032139084507042256,
+      "loss": 1.6018,
+      "step": 22300
+    },
+    {
+      "epoch": 3.8010204081632653,
+      "grad_norm": 0.45325416326522827,
+      "learning_rate": 0.0003209507042253521,
+      "loss": 1.5905,
+      "step": 22350
+    },
+    {
+      "epoch": 3.8095238095238093,
+      "grad_norm": 0.4605984687805176,
+      "learning_rate": 0.0003205105633802817,
+      "loss": 1.596,
+      "step": 22400
+    },
+    {
+      "epoch": 3.818027210884354,
+      "grad_norm": 0.5207810401916504,
+      "learning_rate": 0.0003200704225352113,
+      "loss": 1.6048,
+      "step": 22450
+    },
+    {
+      "epoch": 3.826530612244898,
+      "grad_norm": 0.45419174432754517,
+      "learning_rate": 0.00031963028169014083,
+      "loss": 1.5972,
+      "step": 22500
+    },
+    {
+      "epoch": 3.8350340136054424,
+      "grad_norm": 0.46196499466896057,
+      "learning_rate": 0.00031919014084507043,
+      "loss": 1.5952,
+      "step": 22550
+    },
+    {
+      "epoch": 3.8435374149659864,
+      "grad_norm": 0.48807084560394287,
+      "learning_rate": 0.00031874999999999997,
+      "loss": 1.6006,
+      "step": 22600
+    },
+    {
+      "epoch": 3.8520408163265305,
+      "grad_norm": 0.44178763031959534,
+      "learning_rate": 0.000318318661971831,
+      "loss": 1.6006,
+      "step": 22650
+    },
+    {
+      "epoch": 3.8605442176870746,
+      "grad_norm": 0.4443242847919464,
+      "learning_rate": 0.00031787852112676057,
+      "loss": 1.5875,
+      "step": 22700
+    },
+    {
+      "epoch": 3.869047619047619,
+      "grad_norm": 0.45137760043144226,
+      "learning_rate": 0.00031743838028169016,
+      "loss": 1.5874,
+      "step": 22750
+    },
+    {
+      "epoch": 3.877551020408163,
+      "grad_norm": 0.44187167286872864,
+      "learning_rate": 0.0003169982394366197,
+      "loss": 1.5935,
+      "step": 22800
+    },
+    {
+      "epoch": 3.8860544217687076,
+      "grad_norm": 0.46749913692474365,
+      "learning_rate": 0.0003165580985915493,
+      "loss": 1.5918,
+      "step": 22850
+    },
+    {
+      "epoch": 3.8945578231292517,
+      "grad_norm": 0.45157769322395325,
+      "learning_rate": 0.0003161179577464789,
+      "loss": 1.5965,
+      "step": 22900
+    },
+    {
+      "epoch": 3.9030612244897958,
+      "grad_norm": 0.6856850981712341,
+      "learning_rate": 0.00031567781690140844,
+      "loss": 1.5929,
+      "step": 22950
+    },
+    {
+      "epoch": 3.9115646258503403,
+      "grad_norm": 0.45894768834114075,
+      "learning_rate": 0.00031523767605633803,
+      "loss": 1.5915,
+      "step": 23000
+    },
+    {
+      "epoch": 3.9115646258503403,
+      "eval_loss": 1.7726092338562012,
+      "eval_runtime": 75.3291,
+      "eval_samples_per_second": 1245.72,
+      "eval_steps_per_second": 4.872,
+      "step": 23000
+    },
+    {
+      "epoch": 3.9200680272108843,
+      "grad_norm": 0.44938963651657104,
+      "learning_rate": 0.0003147975352112676,
+      "loss": 1.5913,
+      "step": 23050
+    },
+    {
+      "epoch": 3.928571428571429,
+      "grad_norm": 0.4493829309940338,
+      "learning_rate": 0.0003143573943661972,
+      "loss": 1.5999,
+      "step": 23100
+    },
+    {
+      "epoch": 3.937074829931973,
+      "grad_norm": 0.44146212935447693,
+      "learning_rate": 0.00031391725352112676,
+      "loss": 1.5937,
+      "step": 23150
+    },
+    {
+      "epoch": 3.945578231292517,
+      "grad_norm": 0.4436129927635193,
+      "learning_rate": 0.0003134771126760563,
+      "loss": 1.5937,
+      "step": 23200
+    },
+    {
+      "epoch": 3.954081632653061,
+      "grad_norm": 0.4842548668384552,
+      "learning_rate": 0.0003130369718309859,
+      "loss": 1.5921,
+      "step": 23250
+    },
+    {
+      "epoch": 3.9625850340136055,
+      "grad_norm": 0.4605764150619507,
+      "learning_rate": 0.0003125968309859155,
+      "loss": 1.585,
+      "step": 23300
+    },
+    {
+      "epoch": 3.9710884353741496,
+      "grad_norm": 0.446140319108963,
+      "learning_rate": 0.0003121566901408451,
+      "loss": 1.5773,
+      "step": 23350
+    },
+    {
+      "epoch": 3.979591836734694,
+      "grad_norm": 0.4490399658679962,
+      "learning_rate": 0.00031171654929577463,
+      "loss": 1.5862,
+      "step": 23400
+    },
+    {
+      "epoch": 3.988095238095238,
+      "grad_norm": 0.43997710943222046,
+      "learning_rate": 0.00031128521126760564,
+      "loss": 1.5873,
+      "step": 23450
+    },
+    {
+      "epoch": 3.996598639455782,
+      "grad_norm": 0.4559916853904724,
+      "learning_rate": 0.00031084507042253523,
+      "loss": 1.5936,
+      "step": 23500
+    },
+    {
+      "epoch": 4.005102040816326,
+      "grad_norm": 0.46398648619651794,
+      "learning_rate": 0.0003104049295774648,
+      "loss": 1.5861,
+      "step": 23550
+    },
+    {
+      "epoch": 4.01360544217687,
+      "grad_norm": 0.4367043972015381,
+      "learning_rate": 0.00030996478873239437,
+      "loss": 1.5814,
+      "step": 23600
+    },
+    {
+      "epoch": 4.022108843537415,
+      "grad_norm": 0.4262286126613617,
+      "learning_rate": 0.0003095246478873239,
+      "loss": 1.5831,
+      "step": 23650
+    },
+    {
+      "epoch": 4.030612244897959,
+      "grad_norm": 1.598874568939209,
+      "learning_rate": 0.00030908450704225356,
+      "loss": 1.5863,
+      "step": 23700
+    },
+    {
+      "epoch": 4.039115646258503,
+      "grad_norm": 0.44894129037857056,
+      "learning_rate": 0.0003086443661971831,
+      "loss": 1.5852,
+      "step": 23750
+    },
+    {
+      "epoch": 4.0476190476190474,
+      "grad_norm": 0.4517632722854614,
+      "learning_rate": 0.0003082042253521127,
+      "loss": 1.581,
+      "step": 23800
+    },
+    {
+      "epoch": 4.0561224489795915,
+      "grad_norm": 0.4445061683654785,
+      "learning_rate": 0.00030776408450704224,
+      "loss": 1.5777,
+      "step": 23850
+    },
+    {
+      "epoch": 4.0646258503401365,
+      "grad_norm": 0.4282425343990326,
+      "learning_rate": 0.0003073239436619719,
+      "loss": 1.579,
+      "step": 23900
+    },
+    {
+      "epoch": 4.0731292517006805,
+      "grad_norm": 0.48686137795448303,
+      "learning_rate": 0.0003068838028169014,
+      "loss": 1.5766,
+      "step": 23950
+    },
+    {
+      "epoch": 4.081632653061225,
+      "grad_norm": 0.4572753310203552,
+      "learning_rate": 0.00030644366197183097,
+      "loss": 1.5751,
+      "step": 24000
+    },
+    {
+      "epoch": 4.081632653061225,
+      "eval_loss": 1.7599811553955078,
+      "eval_runtime": 75.3428,
+      "eval_samples_per_second": 1245.494,
+      "eval_steps_per_second": 4.871,
+      "step": 24000
+    },
+    {
+      "epoch": 4.090136054421769,
+      "grad_norm": 0.44744428992271423,
+      "learning_rate": 0.00030600352112676056,
+      "loss": 1.573,
+      "step": 24050
+    },
+    {
+      "epoch": 4.098639455782313,
+      "grad_norm": 0.45439615845680237,
+      "learning_rate": 0.00030556338028169016,
+      "loss": 1.5717,
+      "step": 24100
+    },
+    {
+      "epoch": 4.107142857142857,
+      "grad_norm": 0.46508198976516724,
+      "learning_rate": 0.00030512323943661975,
+      "loss": 1.5716,
+      "step": 24150
+    },
+    {
+      "epoch": 4.115646258503402,
+      "grad_norm": 0.45242032408714294,
+      "learning_rate": 0.0003046830985915493,
+      "loss": 1.5692,
+      "step": 24200
+    },
+    {
+      "epoch": 4.124149659863946,
+      "grad_norm": 0.441568523645401,
+      "learning_rate": 0.0003042429577464789,
+      "loss": 1.569,
+      "step": 24250
+    },
+    {
+      "epoch": 4.13265306122449,
+      "grad_norm": 0.43780016899108887,
+      "learning_rate": 0.00030380281690140843,
+      "loss": 1.5751,
+      "step": 24300
+    },
+    {
+      "epoch": 4.141156462585034,
+      "grad_norm": 0.47256138920783997,
+      "learning_rate": 0.00030336267605633803,
+      "loss": 1.5711,
+      "step": 24350
+    },
+    {
+      "epoch": 4.149659863945578,
+      "grad_norm": 0.4236377477645874,
+      "learning_rate": 0.0003029225352112676,
+      "loss": 1.5737,
+      "step": 24400
+    },
+    {
+      "epoch": 4.158163265306122,
+      "grad_norm": 0.4356014132499695,
+      "learning_rate": 0.00030248239436619716,
+      "loss": 1.5683,
+      "step": 24450
+    },
+    {
+      "epoch": 4.166666666666667,
+      "grad_norm": 0.44677263498306274,
+      "learning_rate": 0.00030204225352112676,
+      "loss": 1.5661,
+      "step": 24500
+    },
+    {
+      "epoch": 4.175170068027211,
+      "grad_norm": 0.45358529686927795,
+      "learning_rate": 0.00030160211267605635,
+      "loss": 1.5718,
+      "step": 24550
+    },
+    {
+      "epoch": 4.183673469387755,
+      "grad_norm": 0.46313029527664185,
+      "learning_rate": 0.00030116197183098595,
+      "loss": 1.5726,
+      "step": 24600
+    },
+    {
+      "epoch": 4.192176870748299,
+      "grad_norm": 0.4524216055870056,
+      "learning_rate": 0.0003007218309859155,
+      "loss": 1.5659,
+      "step": 24650
+    },
+    {
+      "epoch": 4.200680272108843,
+      "grad_norm": 0.46442392468452454,
+      "learning_rate": 0.00030028169014084503,
+      "loss": 1.57,
+      "step": 24700
+    },
+    {
+      "epoch": 4.209183673469388,
+      "grad_norm": 0.43893301486968994,
+      "learning_rate": 0.0002998415492957747,
+      "loss": 1.5646,
+      "step": 24750
+    },
+    {
+      "epoch": 4.217687074829932,
+      "grad_norm": 0.4545600116252899,
+      "learning_rate": 0.0002994014084507042,
+      "loss": 1.5656,
+      "step": 24800
+    },
+    {
+      "epoch": 4.226190476190476,
+      "grad_norm": 0.43572577834129333,
+      "learning_rate": 0.0002989612676056338,
+      "loss": 1.5619,
+      "step": 24850
+    },
+    {
+      "epoch": 4.23469387755102,
+      "grad_norm": 0.4401288330554962,
+      "learning_rate": 0.00029852112676056336,
+      "loss": 1.5626,
+      "step": 24900
+    },
+    {
+      "epoch": 4.243197278911564,
+      "grad_norm": 0.4619799554347992,
+      "learning_rate": 0.000298080985915493,
+      "loss": 1.5679,
+      "step": 24950
+    },
+    {
+      "epoch": 4.2517006802721085,
+      "grad_norm": 0.43690764904022217,
+      "learning_rate": 0.00029764084507042255,
+      "loss": 1.5602,
+      "step": 25000
+    },
+    {
+      "epoch": 4.2517006802721085,
+      "eval_loss": 1.7416114807128906,
+      "eval_runtime": 75.3172,
+      "eval_samples_per_second": 1245.918,
+      "eval_steps_per_second": 4.873,
+      "step": 25000
+    },
+    {
+      "epoch": 4.260204081632653,
+      "grad_norm": 0.48442286252975464,
+      "learning_rate": 0.0002972007042253521,
+      "loss": 1.5592,
+      "step": 25050
+    },
+    {
+      "epoch": 4.2687074829931975,
+      "grad_norm": 0.44272610545158386,
+      "learning_rate": 0.0002967605633802817,
+      "loss": 1.5612,
+      "step": 25100
+    },
+    {
+      "epoch": 4.2772108843537415,
+      "grad_norm": 0.44438502192497253,
+      "learning_rate": 0.0002963204225352113,
+      "loss": 1.5622,
+      "step": 25150
+    },
+    {
+      "epoch": 4.285714285714286,
+      "grad_norm": 0.701813280582428,
+      "learning_rate": 0.0002958802816901409,
+      "loss": 1.5639,
+      "step": 25200
+    },
+    {
+      "epoch": 4.29421768707483,
+      "grad_norm": 0.4430041015148163,
+      "learning_rate": 0.0002954401408450704,
+      "loss": 1.5655,
+      "step": 25250
+    },
+    {
+      "epoch": 4.302721088435375,
+      "grad_norm": 0.4703254699707031,
+      "learning_rate": 0.000295,
+      "loss": 1.5585,
+      "step": 25300
+    },
+    {
+      "epoch": 4.311224489795919,
+      "grad_norm": 0.4456980228424072,
+      "learning_rate": 0.0002945598591549296,
+      "loss": 1.5602,
+      "step": 25350
+    },
+    {
+      "epoch": 4.319727891156463,
+      "grad_norm": 0.44341742992401123,
+      "learning_rate": 0.00029411971830985915,
+      "loss": 1.565,
+      "step": 25400
+    },
+    {
+      "epoch": 4.328231292517007,
+      "grad_norm": 0.6620036959648132,
+      "learning_rate": 0.00029367957746478875,
+      "loss": 1.563,
+      "step": 25450
+    },
+    {
+      "epoch": 4.336734693877551,
+      "grad_norm": 0.4276537597179413,
+      "learning_rate": 0.0002932482394366197,
+      "loss": 1.556,
+      "step": 25500
+    },
+    {
+      "epoch": 4.345238095238095,
+      "grad_norm": 0.5484892129898071,
+      "learning_rate": 0.0002928080985915493,
+      "loss": 1.5683,
+      "step": 25550
+    },
+    {
+      "epoch": 4.35374149659864,
+      "grad_norm": 0.4457222819328308,
+      "learning_rate": 0.0002923679577464789,
+      "loss": 1.5626,
+      "step": 25600
+    },
+    {
+      "epoch": 4.362244897959184,
+      "grad_norm": 0.4299587905406952,
+      "learning_rate": 0.0002919278169014085,
+      "loss": 1.5617,
+      "step": 25650
+    },
+    {
+      "epoch": 4.370748299319728,
+      "grad_norm": 0.6894689202308655,
+      "learning_rate": 0.000291487676056338,
+      "loss": 1.556,
+      "step": 25700
+    },
+    {
+      "epoch": 4.379251700680272,
+      "grad_norm": 0.4507542550563812,
+      "learning_rate": 0.00029104753521126756,
+      "loss": 1.5605,
+      "step": 25750
+    },
+    {
+      "epoch": 4.387755102040816,
+      "grad_norm": 0.45560920238494873,
+      "learning_rate": 0.0002906073943661972,
+      "loss": 1.5631,
+      "step": 25800
+    },
+    {
+      "epoch": 4.39625850340136,
+      "grad_norm": 0.4419881999492645,
+      "learning_rate": 0.00029016725352112676,
+      "loss": 1.558,
+      "step": 25850
+    },
+    {
+      "epoch": 4.404761904761905,
+      "grad_norm": 0.42712947726249695,
+      "learning_rate": 0.00028972711267605635,
+      "loss": 1.5571,
+      "step": 25900
+    },
+    {
+      "epoch": 4.413265306122449,
+      "grad_norm": 0.4276752769947052,
+      "learning_rate": 0.0002892869718309859,
+      "loss": 1.5517,
+      "step": 25950
+    },
+    {
+      "epoch": 4.421768707482993,
+      "grad_norm": 0.4421616494655609,
+      "learning_rate": 0.00028884683098591554,
+      "loss": 1.5549,
+      "step": 26000
+    },
+    {
+      "epoch": 4.421768707482993,
+      "eval_loss": 1.7275468111038208,
+      "eval_runtime": 75.3579,
+      "eval_samples_per_second": 1245.244,
+      "eval_steps_per_second": 4.87,
+      "step": 26000
+    },
+    {
+      "epoch": 4.430272108843537,
+      "grad_norm": 0.43331626057624817,
+      "learning_rate": 0.0002884066901408451,
+      "loss": 1.553,
+      "step": 26050
+    },
+    {
+      "epoch": 4.438775510204081,
+      "grad_norm": 0.42342087626457214,
+      "learning_rate": 0.0002879665492957746,
+      "loss": 1.5508,
+      "step": 26100
+    },
+    {
+      "epoch": 4.447278911564625,
+      "grad_norm": 0.4396851062774658,
+      "learning_rate": 0.0002875264084507042,
+      "loss": 1.5496,
+      "step": 26150
+    },
+    {
+      "epoch": 4.45578231292517,
+      "grad_norm": 0.4484754204750061,
+      "learning_rate": 0.0002870862676056338,
+      "loss": 1.559,
+      "step": 26200
+    },
+    {
+      "epoch": 4.464285714285714,
+      "grad_norm": 0.4413243532180786,
+      "learning_rate": 0.0002866461267605634,
+      "loss": 1.5497,
+      "step": 26250
+    },
+    {
+      "epoch": 4.4727891156462585,
+      "grad_norm": 0.509044885635376,
+      "learning_rate": 0.00028620598591549295,
+      "loss": 1.5507,
+      "step": 26300
+    },
+    {
+      "epoch": 4.4812925170068025,
+      "grad_norm": 0.4461870491504669,
+      "learning_rate": 0.00028576584507042255,
+      "loss": 1.5516,
+      "step": 26350
+    },
+    {
+      "epoch": 4.489795918367347,
+      "grad_norm": 0.48340821266174316,
+      "learning_rate": 0.00028532570422535214,
+      "loss": 1.5542,
+      "step": 26400
+    },
+    {
+      "epoch": 4.4982993197278915,
+      "grad_norm": 0.4417915344238281,
+      "learning_rate": 0.0002848855633802817,
+      "loss": 1.5539,
+      "step": 26450
+    },
+    {
+      "epoch": 4.506802721088436,
+      "grad_norm": 0.47127747535705566,
+      "learning_rate": 0.0002844454225352113,
+      "loss": 1.547,
+      "step": 26500
+    },
+    {
+      "epoch": 4.51530612244898,
+      "grad_norm": 0.7486183047294617,
+      "learning_rate": 0.0002840052816901408,
+      "loss": 1.5534,
+      "step": 26550
+    },
+    {
+      "epoch": 4.523809523809524,
+      "grad_norm": 0.4753170907497406,
+      "learning_rate": 0.0002835651408450704,
+      "loss": 1.5568,
+      "step": 26600
+    },
+    {
+      "epoch": 4.532312925170068,
+      "grad_norm": 0.43808746337890625,
+      "learning_rate": 0.000283125,
+      "loss": 1.5518,
+      "step": 26650
+    },
+    {
+      "epoch": 4.540816326530612,
+      "grad_norm": 0.4334624707698822,
+      "learning_rate": 0.0002826848591549296,
+      "loss": 1.5507,
+      "step": 26700
+    },
+    {
+      "epoch": 4.549319727891157,
+      "grad_norm": 0.4278333783149719,
+      "learning_rate": 0.00028224471830985915,
+      "loss": 1.5543,
+      "step": 26750
+    },
+    {
+      "epoch": 4.557823129251701,
+      "grad_norm": 0.44058939814567566,
+      "learning_rate": 0.0002818045774647887,
+      "loss": 1.5462,
+      "step": 26800
+    },
+    {
+      "epoch": 4.566326530612245,
+      "grad_norm": 0.42488718032836914,
+      "learning_rate": 0.00028136443661971834,
+      "loss": 1.5555,
+      "step": 26850
+    },
+    {
+      "epoch": 4.574829931972789,
+      "grad_norm": 0.4430455267429352,
+      "learning_rate": 0.0002809242957746479,
+      "loss": 1.5481,
+      "step": 26900
+    },
+    {
+      "epoch": 4.583333333333333,
+      "grad_norm": 1.010523796081543,
+      "learning_rate": 0.0002804841549295775,
+      "loss": 1.5446,
+      "step": 26950
+    },
+    {
+      "epoch": 4.591836734693878,
+      "grad_norm": 0.42729562520980835,
+      "learning_rate": 0.000280044014084507,
+      "loss": 1.5464,
+      "step": 27000
+    },
+    {
+      "epoch": 4.591836734693878,
+      "eval_loss": 1.7172296047210693,
+      "eval_runtime": 75.3539,
+      "eval_samples_per_second": 1245.31,
+      "eval_steps_per_second": 4.87,
+      "step": 27000
+    },
+    {
+      "epoch": 4.600340136054422,
+      "grad_norm": 0.4399702847003937,
+      "learning_rate": 0.00027960387323943667,
+      "loss": 1.5452,
+      "step": 27050
+    },
+    {
+      "epoch": 4.608843537414966,
+      "grad_norm": 0.4524286687374115,
+      "learning_rate": 0.0002791637323943662,
+      "loss": 1.5534,
+      "step": 27100
+    },
+    {
+      "epoch": 4.61734693877551,
+      "grad_norm": 0.43100252747535706,
+      "learning_rate": 0.00027872359154929575,
+      "loss": 1.5479,
+      "step": 27150
+    },
+    {
+      "epoch": 4.625850340136054,
+      "grad_norm": 0.43541115522384644,
+      "learning_rate": 0.00027828345070422534,
+      "loss": 1.542,
+      "step": 27200
+    },
+    {
+      "epoch": 4.634353741496598,
+      "grad_norm": 0.454149454832077,
+      "learning_rate": 0.00027784330985915494,
+      "loss": 1.5442,
+      "step": 27250
+    },
+    {
+      "epoch": 4.642857142857143,
+      "grad_norm": 0.4360577464103699,
+      "learning_rate": 0.00027740316901408453,
+      "loss": 1.541,
+      "step": 27300
+    },
+    {
+      "epoch": 4.651360544217687,
+      "grad_norm": 0.45018696784973145,
+      "learning_rate": 0.0002769630281690141,
+      "loss": 1.5522,
+      "step": 27350
+    },
+    {
+      "epoch": 4.659863945578231,
+      "grad_norm": 0.4641735851764679,
+      "learning_rate": 0.00027652288732394367,
+      "loss": 1.5409,
+      "step": 27400
+    },
+    {
+      "epoch": 4.668367346938775,
+      "grad_norm": 0.4704575836658478,
+      "learning_rate": 0.00027608274647887327,
+      "loss": 1.5386,
+      "step": 27450
+    },
+    {
+      "epoch": 4.6768707482993195,
+      "grad_norm": 0.4325176775455475,
+      "learning_rate": 0.0002756426056338028,
+      "loss": 1.5432,
+      "step": 27500
+    },
+    {
+      "epoch": 4.685374149659864,
+      "grad_norm": 0.5194480419158936,
+      "learning_rate": 0.0002752024647887324,
+      "loss": 1.5409,
+      "step": 27550
+    },
+    {
+      "epoch": 4.6938775510204085,
+      "grad_norm": 0.6043953895568848,
+      "learning_rate": 0.00027476232394366194,
+      "loss": 1.5412,
+      "step": 27600
+    },
+    {
+      "epoch": 4.7023809523809526,
+      "grad_norm": 0.49982741475105286,
+      "learning_rate": 0.0002743221830985916,
+      "loss": 1.544,
+      "step": 27650
+    },
+    {
+      "epoch": 4.710884353741497,
+      "grad_norm": 0.45368969440460205,
+      "learning_rate": 0.00027388204225352113,
+      "loss": 1.5452,
+      "step": 27700
+    },
+    {
+      "epoch": 4.719387755102041,
+      "grad_norm": 0.42949676513671875,
+      "learning_rate": 0.00027344190140845073,
+      "loss": 1.5431,
+      "step": 27750
+    },
+    {
+      "epoch": 4.727891156462585,
+      "grad_norm": 0.4460367262363434,
+      "learning_rate": 0.00027300176056338027,
+      "loss": 1.5379,
+      "step": 27800
+    },
+    {
+      "epoch": 4.736394557823129,
+      "grad_norm": 0.43941059708595276,
+      "learning_rate": 0.0002725616197183098,
+      "loss": 1.5438,
+      "step": 27850
+    },
+    {
+      "epoch": 4.744897959183674,
+      "grad_norm": 0.44306865334510803,
+      "learning_rate": 0.00027212147887323946,
+      "loss": 1.5403,
+      "step": 27900
+    },
+    {
+      "epoch": 4.753401360544218,
+      "grad_norm": 0.4779192805290222,
+      "learning_rate": 0.000271681338028169,
+      "loss": 1.5374,
+      "step": 27950
+    },
+    {
+      "epoch": 4.761904761904762,
+      "grad_norm": 0.44125354290008545,
+      "learning_rate": 0.0002712411971830986,
+      "loss": 1.5386,
+      "step": 28000
+    },
+    {
+      "epoch": 4.761904761904762,
+      "eval_loss": 1.7089167833328247,
+      "eval_runtime": 75.3103,
+      "eval_samples_per_second": 1246.032,
+      "eval_steps_per_second": 4.873,
+      "step": 28000
+    },
+    {
+      "epoch": 4.770408163265306,
+      "grad_norm": 0.47051918506622314,
+      "learning_rate": 0.00027080105633802814,
+      "loss": 1.5367,
+      "step": 28050
+    },
+    {
+      "epoch": 4.77891156462585,
+      "grad_norm": 0.44151777029037476,
+      "learning_rate": 0.0002703609154929578,
+      "loss": 1.5363,
+      "step": 28100
+    },
+    {
+      "epoch": 4.787414965986395,
+      "grad_norm": 0.43706709146499634,
+      "learning_rate": 0.00026992077464788733,
+      "loss": 1.5396,
+      "step": 28150
+    },
+    {
+      "epoch": 4.795918367346939,
+      "grad_norm": 0.4319562017917633,
+      "learning_rate": 0.0002694806338028169,
+      "loss": 1.5339,
+      "step": 28200
+    },
+    {
+      "epoch": 4.804421768707483,
+      "grad_norm": 0.4437416195869446,
+      "learning_rate": 0.00026904049295774647,
+      "loss": 1.5259,
+      "step": 28250
+    },
+    {
+      "epoch": 4.812925170068027,
+      "grad_norm": 0.45064008235931396,
+      "learning_rate": 0.00026860035211267606,
+      "loss": 1.5305,
+      "step": 28300
+    },
+    {
+      "epoch": 4.821428571428571,
+      "grad_norm": 0.4223947823047638,
+      "learning_rate": 0.00026816021126760566,
+      "loss": 1.5367,
+      "step": 28350
+    },
+    {
+      "epoch": 4.829931972789115,
+      "grad_norm": 0.41959166526794434,
+      "learning_rate": 0.0002677200704225352,
+      "loss": 1.5321,
+      "step": 28400
+    },
+    {
+      "epoch": 4.83843537414966,
+      "grad_norm": 0.4325518310070038,
+      "learning_rate": 0.0002672799295774648,
+      "loss": 1.5326,
+      "step": 28450
+    },
+    {
+      "epoch": 4.846938775510204,
+      "grad_norm": 0.43113574385643005,
+      "learning_rate": 0.0002668397887323944,
+      "loss": 1.5364,
+      "step": 28500
+    },
+    {
+      "epoch": 4.855442176870748,
+      "grad_norm": 0.4148525297641754,
+      "learning_rate": 0.000266399647887324,
+      "loss": 1.535,
+      "step": 28550
+    },
+    {
+      "epoch": 4.863945578231292,
+      "grad_norm": 0.4255242943763733,
+      "learning_rate": 0.0002659595070422535,
+      "loss": 1.5307,
+      "step": 28600
+    },
+    {
+      "epoch": 4.872448979591836,
+      "grad_norm": 0.4426940679550171,
+      "learning_rate": 0.00026551936619718307,
+      "loss": 1.5298,
+      "step": 28650
+    },
+    {
+      "epoch": 4.880952380952381,
+      "grad_norm": 0.4304163157939911,
+      "learning_rate": 0.0002650792253521127,
+      "loss": 1.5296,
+      "step": 28700
+    },
+    {
+      "epoch": 4.889455782312925,
+      "grad_norm": 0.43492448329925537,
+      "learning_rate": 0.00026463908450704226,
+      "loss": 1.5295,
+      "step": 28750
+    },
+    {
+      "epoch": 4.8979591836734695,
+      "grad_norm": 0.4191015660762787,
+      "learning_rate": 0.00026420774647887326,
+      "loss": 1.5275,
+      "step": 28800
+    },
+    {
+      "epoch": 4.906462585034014,
+      "grad_norm": 0.4442043602466583,
+      "learning_rate": 0.0002637676056338028,
+      "loss": 1.5241,
+      "step": 28850
+    },
+    {
+      "epoch": 4.914965986394558,
+      "grad_norm": 0.41631436347961426,
+      "learning_rate": 0.0002633274647887324,
+      "loss": 1.5296,
+      "step": 28900
+    },
+    {
+      "epoch": 4.923469387755102,
+      "grad_norm": 0.4311072528362274,
+      "learning_rate": 0.000262887323943662,
+      "loss": 1.5343,
+      "step": 28950
+    },
+    {
+      "epoch": 4.931972789115647,
+      "grad_norm": 0.4260921776294708,
+      "learning_rate": 0.00026244718309859154,
+      "loss": 1.528,
+      "step": 29000
+    },
+    {
+      "epoch": 4.931972789115647,
+      "eval_loss": 1.695530652999878,
+      "eval_runtime": 75.7527,
+      "eval_samples_per_second": 1238.755,
+      "eval_steps_per_second": 4.845,
+      "step": 29000
+    },
+    {
+      "epoch": 4.940476190476191,
+      "grad_norm": 0.43122732639312744,
+      "learning_rate": 0.00026200704225352113,
+      "loss": 1.5345,
+      "step": 29050
+    },
+    {
+      "epoch": 4.948979591836735,
+      "grad_norm": 0.44380810856819153,
+      "learning_rate": 0.00026156690140845067,
+      "loss": 1.5284,
+      "step": 29100
+    },
+    {
+      "epoch": 4.957482993197279,
+      "grad_norm": 0.43009868264198303,
+      "learning_rate": 0.0002611267605633803,
+      "loss": 1.5315,
+      "step": 29150
+    },
+    {
+      "epoch": 4.965986394557823,
+      "grad_norm": 0.4425157904624939,
+      "learning_rate": 0.00026068661971830986,
+      "loss": 1.5301,
+      "step": 29200
+    },
+    {
+      "epoch": 4.974489795918368,
+      "grad_norm": 0.41929659247398376,
+      "learning_rate": 0.00026024647887323946,
+      "loss": 1.533,
+      "step": 29250
+    },
+    {
+      "epoch": 4.982993197278912,
+      "grad_norm": 0.43218111991882324,
+      "learning_rate": 0.000259806338028169,
+      "loss": 1.5215,
+      "step": 29300
+    },
+    {
+      "epoch": 4.991496598639456,
+      "grad_norm": 0.4425402283668518,
+      "learning_rate": 0.0002593661971830986,
+      "loss": 1.5312,
+      "step": 29350
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 0.44686195254325867,
+      "learning_rate": 0.0002589260563380282,
+      "loss": 1.5262,
+      "step": 29400
+    },
+    {
+      "epoch": 5.008503401360544,
+      "grad_norm": 0.42469632625579834,
+      "learning_rate": 0.00025848591549295773,
+      "loss": 1.5216,
+      "step": 29450
+    },
+    {
+      "epoch": 5.017006802721088,
+      "grad_norm": 0.44917765259742737,
+      "learning_rate": 0.0002580457746478873,
+      "loss": 1.5148,
+      "step": 29500
+    },
+    {
+      "epoch": 5.025510204081633,
+      "grad_norm": 0.42792636156082153,
+      "learning_rate": 0.0002576056338028169,
+      "loss": 1.5236,
+      "step": 29550
+    },
+    {
+      "epoch": 5.034013605442177,
+      "grad_norm": 0.42785459756851196,
+      "learning_rate": 0.0002571654929577465,
+      "loss": 1.5185,
+      "step": 29600
+    },
+    {
+      "epoch": 5.042517006802721,
+      "grad_norm": 0.4389972388744354,
+      "learning_rate": 0.00025672535211267606,
+      "loss": 1.5152,
+      "step": 29650
+    },
+    {
+      "epoch": 5.051020408163265,
+      "grad_norm": 0.4294348657131195,
+      "learning_rate": 0.0002562852112676056,
+      "loss": 1.5164,
+      "step": 29700
+    },
+    {
+      "epoch": 5.059523809523809,
+      "grad_norm": 0.4204474985599518,
+      "learning_rate": 0.00025584507042253525,
+      "loss": 1.5126,
+      "step": 29750
+    },
+    {
+      "epoch": 5.068027210884353,
+      "grad_norm": 0.4308267831802368,
+      "learning_rate": 0.0002554049295774648,
+      "loss": 1.5138,
+      "step": 29800
+    },
+    {
+      "epoch": 5.076530612244898,
+      "grad_norm": 0.43260395526885986,
+      "learning_rate": 0.0002549647887323944,
+      "loss": 1.5211,
+      "step": 29850
+    },
+    {
+      "epoch": 5.085034013605442,
+      "grad_norm": 0.4192730486392975,
+      "learning_rate": 0.00025452464788732393,
+      "loss": 1.5165,
+      "step": 29900
+    },
+    {
+      "epoch": 5.093537414965986,
+      "grad_norm": 0.418124794960022,
+      "learning_rate": 0.0002540845070422535,
+      "loss": 1.5152,
+      "step": 29950
+    },
+    {
+      "epoch": 5.1020408163265305,
+      "grad_norm": 0.4320596158504486,
+      "learning_rate": 0.0002536443661971831,
+      "loss": 1.5071,
+      "step": 30000
+    },
+    {
+      "epoch": 5.1020408163265305,
+      "eval_loss": 1.6892961263656616,
+      "eval_runtime": 75.4228,
+      "eval_samples_per_second": 1244.173,
+      "eval_steps_per_second": 4.866,
+      "step": 30000
+    },
+    {
+      "epoch": 5.110544217687075,
+      "grad_norm": 0.43125247955322266,
+      "learning_rate": 0.00025320422535211266,
+      "loss": 1.5116,
+      "step": 30050
+    },
+    {
+      "epoch": 5.119047619047619,
+      "grad_norm": 0.4130070209503174,
+      "learning_rate": 0.00025276408450704225,
+      "loss": 1.5125,
+      "step": 30100
+    },
+    {
+      "epoch": 5.127551020408164,
+      "grad_norm": 0.42717909812927246,
+      "learning_rate": 0.0002523239436619718,
+      "loss": 1.5122,
+      "step": 30150
+    },
+    {
+      "epoch": 5.136054421768708,
+      "grad_norm": 0.4337432086467743,
+      "learning_rate": 0.00025188380281690145,
+      "loss": 1.5103,
+      "step": 30200
+    },
+    {
+      "epoch": 5.144557823129252,
+      "grad_norm": 0.4125542938709259,
+      "learning_rate": 0.000251443661971831,
+      "loss": 1.5218,
+      "step": 30250
+    },
+    {
+      "epoch": 5.153061224489796,
+      "grad_norm": 0.428830623626709,
+      "learning_rate": 0.0002510035211267606,
+      "loss": 1.5068,
+      "step": 30300
+    },
+    {
+      "epoch": 5.16156462585034,
+      "grad_norm": 0.41956421732902527,
+      "learning_rate": 0.0002505633802816901,
+      "loss": 1.5148,
+      "step": 30350
+    },
+    {
+      "epoch": 5.170068027210885,
+      "grad_norm": 0.43243858218193054,
+      "learning_rate": 0.0002501232394366197,
+      "loss": 1.5153,
+      "step": 30400
+    },
+    {
+      "epoch": 5.178571428571429,
+      "grad_norm": 0.41967883706092834,
+      "learning_rate": 0.0002496830985915493,
+      "loss": 1.517,
+      "step": 30450
+    },
+    {
+      "epoch": 5.187074829931973,
+      "grad_norm": 0.43599751591682434,
+      "learning_rate": 0.00024924295774647886,
+      "loss": 1.5097,
+      "step": 30500
+    },
+    {
+      "epoch": 5.195578231292517,
+      "grad_norm": 0.4145863950252533,
+      "learning_rate": 0.00024880281690140845,
+      "loss": 1.5054,
+      "step": 30550
+    },
+    {
+      "epoch": 5.204081632653061,
+      "grad_norm": 0.44736286997795105,
+      "learning_rate": 0.00024836267605633805,
+      "loss": 1.512,
+      "step": 30600
+    },
+    {
+      "epoch": 5.212585034013605,
+      "grad_norm": 0.4339098036289215,
+      "learning_rate": 0.00024792253521126764,
+      "loss": 1.5119,
+      "step": 30650
+    },
+    {
+      "epoch": 5.22108843537415,
+      "grad_norm": 0.4354366064071655,
+      "learning_rate": 0.0002474823943661972,
+      "loss": 1.5134,
+      "step": 30700
+    },
+    {
+      "epoch": 5.229591836734694,
+      "grad_norm": 0.4405953288078308,
+      "learning_rate": 0.0002470422535211268,
+      "loss": 1.5117,
+      "step": 30750
+    },
+    {
+      "epoch": 5.238095238095238,
+      "grad_norm": 0.4392901659011841,
+      "learning_rate": 0.0002466021126760563,
+      "loss": 1.5097,
+      "step": 30800
+    },
+    {
+      "epoch": 5.246598639455782,
+      "grad_norm": 0.4320152699947357,
+      "learning_rate": 0.0002461619718309859,
+      "loss": 1.5185,
+      "step": 30850
+    },
+    {
+      "epoch": 5.255102040816326,
+      "grad_norm": 0.4348084628582001,
+      "learning_rate": 0.0002457218309859155,
+      "loss": 1.5073,
+      "step": 30900
+    },
+    {
+      "epoch": 5.263605442176871,
+      "grad_norm": 0.4216504693031311,
+      "learning_rate": 0.0002452816901408451,
+      "loss": 1.5055,
+      "step": 30950
+    },
+    {
+      "epoch": 5.272108843537415,
+      "grad_norm": 0.417338103055954,
+      "learning_rate": 0.00024484154929577465,
+      "loss": 1.5102,
+      "step": 31000
+    },
+    {
+      "epoch": 5.272108843537415,
+      "eval_loss": 1.6826726198196411,
+      "eval_runtime": 75.3283,
+      "eval_samples_per_second": 1245.734,
+      "eval_steps_per_second": 4.872,
+      "step": 31000
+    },
+    {
+      "epoch": 5.280612244897959,
+      "grad_norm": 0.425647109746933,
+      "learning_rate": 0.00024440140845070424,
+      "loss": 1.51,
+      "step": 31050
+    },
+    {
+      "epoch": 5.289115646258503,
+      "grad_norm": 0.4205061197280884,
+      "learning_rate": 0.0002439612676056338,
+      "loss": 1.5071,
+      "step": 31100
+    },
+    {
+      "epoch": 5.2976190476190474,
+      "grad_norm": 0.4191541075706482,
+      "learning_rate": 0.00024352112676056338,
+      "loss": 1.5072,
+      "step": 31150
+    },
+    {
+      "epoch": 5.3061224489795915,
+      "grad_norm": 0.4266960024833679,
+      "learning_rate": 0.00024308098591549297,
+      "loss": 1.5024,
+      "step": 31200
+    },
+    {
+      "epoch": 5.3146258503401365,
+      "grad_norm": 0.43113160133361816,
+      "learning_rate": 0.00024264084507042252,
+      "loss": 1.5045,
+      "step": 31250
+    },
+    {
+      "epoch": 5.3231292517006805,
+      "grad_norm": 0.46872562170028687,
+      "learning_rate": 0.0002422007042253521,
+      "loss": 1.5043,
+      "step": 31300
+    },
+    {
+      "epoch": 5.331632653061225,
+      "grad_norm": 0.444063663482666,
+      "learning_rate": 0.00024176056338028168,
+      "loss": 1.503,
+      "step": 31350
+    },
+    {
+      "epoch": 5.340136054421769,
+      "grad_norm": 0.46224623918533325,
+      "learning_rate": 0.00024132042253521127,
+      "loss": 1.5001,
+      "step": 31400
+    },
+    {
+      "epoch": 5.348639455782313,
+      "grad_norm": 0.4229288697242737,
+      "learning_rate": 0.00024088028169014084,
+      "loss": 1.5051,
+      "step": 31450
+    },
+    {
+      "epoch": 5.357142857142857,
+      "grad_norm": 0.4342740476131439,
+      "learning_rate": 0.00024044014084507044,
+      "loss": 1.5043,
+      "step": 31500
+    },
+    {
+      "epoch": 5.365646258503402,
+      "grad_norm": 0.4987669885158539,
+      "learning_rate": 0.00024,
+      "loss": 1.5067,
+      "step": 31550
+    },
+    {
+      "epoch": 5.374149659863946,
+      "grad_norm": 0.4277132451534271,
+      "learning_rate": 0.0002395598591549296,
+      "loss": 1.5005,
+      "step": 31600
+    },
+    {
+      "epoch": 5.38265306122449,
+      "grad_norm": 0.4543018639087677,
+      "learning_rate": 0.00023911971830985914,
+      "loss": 1.5081,
+      "step": 31650
+    },
+    {
+      "epoch": 5.391156462585034,
+      "grad_norm": 0.4164925217628479,
+      "learning_rate": 0.00023867957746478874,
+      "loss": 1.4998,
+      "step": 31700
+    },
+    {
+      "epoch": 5.399659863945578,
+      "grad_norm": 0.4165375828742981,
+      "learning_rate": 0.0002382394366197183,
+      "loss": 1.5047,
+      "step": 31750
+    },
+    {
+      "epoch": 5.408163265306122,
+      "grad_norm": 0.43289613723754883,
+      "learning_rate": 0.0002377992957746479,
+      "loss": 1.499,
+      "step": 31800
+    },
+    {
+      "epoch": 5.416666666666667,
+      "grad_norm": 0.4113534688949585,
+      "learning_rate": 0.00023735915492957747,
+      "loss": 1.4992,
+      "step": 31850
+    },
+    {
+      "epoch": 5.425170068027211,
+      "grad_norm": 0.4257815480232239,
+      "learning_rate": 0.00023691901408450707,
+      "loss": 1.5033,
+      "step": 31900
+    },
+    {
+      "epoch": 5.433673469387755,
+      "grad_norm": 0.423606812953949,
+      "learning_rate": 0.00023647887323943663,
+      "loss": 1.4985,
+      "step": 31950
+    },
+    {
+      "epoch": 5.442176870748299,
+      "grad_norm": 0.5088992118835449,
+      "learning_rate": 0.0002360387323943662,
+      "loss": 1.5025,
+      "step": 32000
+    },
+    {
+      "epoch": 5.442176870748299,
+      "eval_loss": 1.6751487255096436,
+      "eval_runtime": 75.6944,
+      "eval_samples_per_second": 1239.708,
+      "eval_steps_per_second": 4.848,
+      "step": 32000
+    },
+    {
+      "epoch": 5.450680272108843,
+      "grad_norm": 0.42682725191116333,
+      "learning_rate": 0.00023559859154929577,
+      "loss": 1.5012,
+      "step": 32050
+    },
+    {
+      "epoch": 5.459183673469388,
+      "grad_norm": 0.4484409689903259,
+      "learning_rate": 0.00023515845070422537,
+      "loss": 1.5056,
+      "step": 32100
+    },
+    {
+      "epoch": 5.467687074829932,
+      "grad_norm": 0.4330119788646698,
+      "learning_rate": 0.00023471830985915493,
+      "loss": 1.4997,
+      "step": 32150
+    },
+    {
+      "epoch": 5.476190476190476,
+      "grad_norm": 0.4262082874774933,
+      "learning_rate": 0.00023427816901408453,
+      "loss": 1.4992,
+      "step": 32200
+    },
+    {
+      "epoch": 5.48469387755102,
+      "grad_norm": 0.4434804320335388,
+      "learning_rate": 0.0002338380281690141,
+      "loss": 1.4975,
+      "step": 32250
+    },
+    {
+      "epoch": 5.493197278911564,
+      "grad_norm": 0.42507055401802063,
+      "learning_rate": 0.00023340669014084507,
+      "loss": 1.4923,
+      "step": 32300
+    },
+    {
+      "epoch": 5.5017006802721085,
+      "grad_norm": 0.415923535823822,
+      "learning_rate": 0.00023296654929577464,
+      "loss": 1.4899,
+      "step": 32350
+    },
+    {
+      "epoch": 5.510204081632653,
+      "grad_norm": 0.43220096826553345,
+      "learning_rate": 0.0002325264084507042,
+      "loss": 1.4947,
+      "step": 32400
+    },
+    {
+      "epoch": 5.5187074829931975,
+      "grad_norm": 0.4283994436264038,
+      "learning_rate": 0.0002320862676056338,
+      "loss": 1.4921,
+      "step": 32450
+    },
+    {
+      "epoch": 5.5272108843537415,
+      "grad_norm": 0.439207524061203,
+      "learning_rate": 0.00023164612676056337,
+      "loss": 1.4859,
+      "step": 32500
+    },
+    {
+      "epoch": 5.535714285714286,
+      "grad_norm": 0.4383983612060547,
+      "learning_rate": 0.00023120598591549297,
+      "loss": 1.4933,
+      "step": 32550
+    },
+    {
+      "epoch": 5.54421768707483,
+      "grad_norm": 0.44482484459877014,
+      "learning_rate": 0.00023076584507042254,
+      "loss": 1.4957,
+      "step": 32600
+    },
+    {
+      "epoch": 5.552721088435375,
+      "grad_norm": 0.4300253987312317,
+      "learning_rate": 0.00023032570422535213,
+      "loss": 1.49,
+      "step": 32650
+    },
+    {
+      "epoch": 5.561224489795919,
+      "grad_norm": 0.46058133244514465,
+      "learning_rate": 0.00022988556338028167,
+      "loss": 1.4872,
+      "step": 32700
+    },
+    {
+      "epoch": 5.569727891156463,
+      "grad_norm": 0.4194552004337311,
+      "learning_rate": 0.00022944542253521127,
+      "loss": 1.4911,
+      "step": 32750
+    },
+    {
+      "epoch": 5.578231292517007,
+      "grad_norm": 0.42726796865463257,
+      "learning_rate": 0.00022900528169014084,
+      "loss": 1.4926,
+      "step": 32800
+    },
+    {
+      "epoch": 5.586734693877551,
+      "grad_norm": 0.42779192328453064,
+      "learning_rate": 0.00022856514084507043,
+      "loss": 1.4917,
+      "step": 32850
+    },
+    {
+      "epoch": 5.595238095238095,
+      "grad_norm": 0.42153412103652954,
+      "learning_rate": 0.000228125,
+      "loss": 1.4824,
+      "step": 32900
+    },
+    {
+      "epoch": 5.603741496598639,
+      "grad_norm": 0.4386555254459381,
+      "learning_rate": 0.0002276848591549296,
+      "loss": 1.4916,
+      "step": 32950
+    },
+    {
+      "epoch": 5.612244897959184,
+      "grad_norm": 0.4412122964859009,
+      "learning_rate": 0.00022724471830985917,
+      "loss": 1.4908,
+      "step": 33000
+    },
+    {
+      "epoch": 5.612244897959184,
+      "eval_loss": 1.6691830158233643,
+      "eval_runtime": 75.3357,
+      "eval_samples_per_second": 1245.611,
+      "eval_steps_per_second": 4.872,
+      "step": 33000
+    },
+    {
+      "epoch": 5.620748299319728,
+      "grad_norm": 0.42562374472618103,
+      "learning_rate": 0.00022680457746478873,
+      "loss": 1.4867,
+      "step": 33050
+    },
+    {
+      "epoch": 5.629251700680272,
+      "grad_norm": 0.4305470585823059,
+      "learning_rate": 0.0002263644366197183,
+      "loss": 1.4874,
+      "step": 33100
+    },
+    {
+      "epoch": 5.637755102040816,
+      "grad_norm": 0.42887523770332336,
+      "learning_rate": 0.0002259242957746479,
+      "loss": 1.4993,
+      "step": 33150
+    },
+    {
+      "epoch": 5.646258503401361,
+      "grad_norm": 0.5242183804512024,
+      "learning_rate": 0.00022548415492957747,
+      "loss": 1.4934,
+      "step": 33200
+    },
+    {
+      "epoch": 5.654761904761905,
+      "grad_norm": 0.4118496775627136,
+      "learning_rate": 0.00022504401408450706,
+      "loss": 1.4872,
+      "step": 33250
+    },
+    {
+      "epoch": 5.663265306122449,
+      "grad_norm": 0.4159320890903473,
+      "learning_rate": 0.00022460387323943663,
+      "loss": 1.4782,
+      "step": 33300
+    },
+    {
+      "epoch": 5.671768707482993,
+      "grad_norm": 0.4275069534778595,
+      "learning_rate": 0.00022416373239436623,
+      "loss": 1.4857,
+      "step": 33350
+    },
+    {
+      "epoch": 5.680272108843537,
+      "grad_norm": 0.41769281029701233,
+      "learning_rate": 0.00022372359154929577,
+      "loss": 1.479,
+      "step": 33400
+    },
+    {
+      "epoch": 5.688775510204081,
+      "grad_norm": 0.41575032472610474,
+      "learning_rate": 0.00022328345070422533,
+      "loss": 1.4849,
+      "step": 33450
+    },
+    {
+      "epoch": 5.697278911564625,
+      "grad_norm": 0.4337887763977051,
+      "learning_rate": 0.00022284330985915493,
+      "loss": 1.4877,
+      "step": 33500
+    },
+    {
+      "epoch": 5.70578231292517,
+      "grad_norm": 0.41992053389549255,
+      "learning_rate": 0.0002224031690140845,
+      "loss": 1.4809,
+      "step": 33550
+    },
+    {
+      "epoch": 5.714285714285714,
+      "grad_norm": 0.42422786355018616,
+      "learning_rate": 0.0002219630281690141,
+      "loss": 1.4877,
+      "step": 33600
+    },
+    {
+      "epoch": 5.7227891156462585,
+      "grad_norm": 0.415855348110199,
+      "learning_rate": 0.00022152288732394366,
+      "loss": 1.4863,
+      "step": 33650
+    },
+    {
+      "epoch": 5.7312925170068025,
+      "grad_norm": 0.4512421488761902,
+      "learning_rate": 0.00022108274647887326,
+      "loss": 1.4894,
+      "step": 33700
+    },
+    {
+      "epoch": 5.739795918367347,
+      "grad_norm": 0.42408186197280884,
+      "learning_rate": 0.0002206426056338028,
+      "loss": 1.4767,
+      "step": 33750
+    },
+    {
+      "epoch": 5.7482993197278915,
+      "grad_norm": 0.42864853143692017,
+      "learning_rate": 0.0002202024647887324,
+      "loss": 1.4825,
+      "step": 33800
+    },
+    {
+      "epoch": 5.756802721088436,
+      "grad_norm": 0.4149647355079651,
+      "learning_rate": 0.00021976232394366196,
+      "loss": 1.4819,
+      "step": 33850
+    },
+    {
+      "epoch": 5.76530612244898,
+      "grad_norm": 0.45217257738113403,
+      "learning_rate": 0.00021932218309859156,
+      "loss": 1.487,
+      "step": 33900
+    },
+    {
+      "epoch": 5.773809523809524,
+      "grad_norm": 0.4308101236820221,
+      "learning_rate": 0.00021888204225352113,
+      "loss": 1.4842,
+      "step": 33950
+    },
+    {
+      "epoch": 5.782312925170068,
+      "grad_norm": 0.41859814524650574,
+      "learning_rate": 0.00021844190140845072,
+      "loss": 1.4856,
+      "step": 34000
+    },
+    {
+      "epoch": 5.782312925170068,
+      "eval_loss": 1.6499308347702026,
+      "eval_runtime": 75.302,
+      "eval_samples_per_second": 1246.169,
+      "eval_steps_per_second": 4.874,
+      "step": 34000
+    },
+    {
+      "epoch": 5.790816326530612,
+      "grad_norm": 0.4088199734687805,
+      "learning_rate": 0.0002180017605633803,
+      "loss": 1.4838,
+      "step": 34050
+    },
+    {
+      "epoch": 5.799319727891157,
+      "grad_norm": 0.8137006163597107,
+      "learning_rate": 0.00021756161971830989,
+      "loss": 1.4826,
+      "step": 34100
+    },
+    {
+      "epoch": 5.807823129251701,
+      "grad_norm": 0.4124939441680908,
+      "learning_rate": 0.00021712147887323943,
+      "loss": 1.4916,
+      "step": 34150
+    },
+    {
+      "epoch": 5.816326530612245,
+      "grad_norm": 0.4147464334964752,
+      "learning_rate": 0.00021668133802816902,
+      "loss": 1.485,
+      "step": 34200
+    },
+    {
+      "epoch": 5.824829931972789,
+      "grad_norm": 0.42168349027633667,
+      "learning_rate": 0.0002162411971830986,
+      "loss": 1.4787,
+      "step": 34250
+    },
+    {
+      "epoch": 5.833333333333333,
+      "grad_norm": 0.43687647581100464,
+      "learning_rate": 0.00021580105633802819,
+      "loss": 1.4859,
+      "step": 34300
+    },
+    {
+      "epoch": 5.841836734693878,
+      "grad_norm": 0.4150829017162323,
+      "learning_rate": 0.00021536091549295775,
+      "loss": 1.4806,
+      "step": 34350
+    },
+    {
+      "epoch": 5.850340136054422,
+      "grad_norm": 0.42533501982688904,
+      "learning_rate": 0.00021492077464788735,
+      "loss": 1.4774,
+      "step": 34400
+    },
+    {
+      "epoch": 5.858843537414966,
+      "grad_norm": 0.42881202697753906,
+      "learning_rate": 0.00021448063380281692,
+      "loss": 1.4793,
+      "step": 34450
+    },
+    {
+      "epoch": 5.86734693877551,
+      "grad_norm": 0.41647574305534363,
+      "learning_rate": 0.00021404049295774649,
+      "loss": 1.4844,
+      "step": 34500
+    },
+    {
+      "epoch": 5.875850340136054,
+      "grad_norm": 0.45438751578330994,
+      "learning_rate": 0.00021360035211267605,
+      "loss": 1.476,
+      "step": 34550
+    },
+    {
+      "epoch": 5.884353741496598,
+      "grad_norm": 0.4157373309135437,
+      "learning_rate": 0.00021316021126760562,
+      "loss": 1.4783,
+      "step": 34600
+    },
+    {
+      "epoch": 5.892857142857143,
+      "grad_norm": 0.4178527593612671,
+      "learning_rate": 0.00021272007042253522,
+      "loss": 1.477,
+      "step": 34650
+    },
+    {
+      "epoch": 5.901360544217687,
+      "grad_norm": 0.4168199896812439,
+      "learning_rate": 0.00021227992957746479,
+      "loss": 1.4685,
+      "step": 34700
+    },
+    {
+      "epoch": 5.909863945578231,
+      "grad_norm": 0.4487851858139038,
+      "learning_rate": 0.00021183978873239438,
+      "loss": 1.4782,
+      "step": 34750
+    },
+    {
+      "epoch": 5.918367346938775,
+      "grad_norm": 0.43783414363861084,
+      "learning_rate": 0.00021139964788732395,
+      "loss": 1.4727,
+      "step": 34800
+    },
+    {
+      "epoch": 5.9268707482993195,
+      "grad_norm": 0.4123859107494354,
+      "learning_rate": 0.00021095950704225352,
+      "loss": 1.4711,
+      "step": 34850
+    },
+    {
+      "epoch": 5.935374149659864,
+      "grad_norm": 0.41514667868614197,
+      "learning_rate": 0.00021051936619718309,
+      "loss": 1.4776,
+      "step": 34900
+    },
+    {
+      "epoch": 5.9438775510204085,
+      "grad_norm": 0.41285404562950134,
+      "learning_rate": 0.00021007922535211268,
+      "loss": 1.4796,
+      "step": 34950
+    },
+    {
+      "epoch": 5.9523809523809526,
+      "grad_norm": 0.407366544008255,
+      "learning_rate": 0.00020963908450704225,
+      "loss": 1.4687,
+      "step": 35000
+    },
+    {
+      "epoch": 5.9523809523809526,
+      "eval_loss": 1.6382627487182617,
+      "eval_runtime": 75.3966,
+      "eval_samples_per_second": 1244.606,
+      "eval_steps_per_second": 4.868,
+      "step": 35000
+    },
+    {
+      "epoch": 5.960884353741497,
+      "grad_norm": 0.4154813885688782,
+      "learning_rate": 0.00020919894366197185,
+      "loss": 1.4731,
+      "step": 35050
+    },
+    {
+      "epoch": 5.969387755102041,
+      "grad_norm": 0.40901270508766174,
+      "learning_rate": 0.0002087588028169014,
+      "loss": 1.4767,
+      "step": 35100
+    },
+    {
+      "epoch": 5.977891156462585,
+      "grad_norm": 0.41301828622817993,
+      "learning_rate": 0.000208318661971831,
+      "loss": 1.4688,
+      "step": 35150
+    },
+    {
+      "epoch": 5.986394557823129,
+      "grad_norm": 0.43283048272132874,
+      "learning_rate": 0.00020787852112676055,
+      "loss": 1.4673,
+      "step": 35200
+    },
+    {
+      "epoch": 5.994897959183674,
+      "grad_norm": 0.4242665469646454,
+      "learning_rate": 0.00020743838028169015,
+      "loss": 1.4749,
+      "step": 35250
+    },
+    {
+      "epoch": 6.003401360544218,
+      "grad_norm": 0.4176988899707794,
+      "learning_rate": 0.00020699823943661971,
+      "loss": 1.4713,
+      "step": 35300
+    },
+    {
+      "epoch": 6.011904761904762,
+      "grad_norm": 0.4250063896179199,
+      "learning_rate": 0.0002065580985915493,
+      "loss": 1.4688,
+      "step": 35350
+    },
+    {
+      "epoch": 6.020408163265306,
+      "grad_norm": 0.4247675836086273,
+      "learning_rate": 0.00020611795774647888,
+      "loss": 1.4661,
+      "step": 35400
+    },
+    {
+      "epoch": 6.02891156462585,
+      "grad_norm": 0.47026437520980835,
+      "learning_rate": 0.00020567781690140847,
+      "loss": 1.4641,
+      "step": 35450
+    },
+    {
+      "epoch": 6.037414965986395,
+      "grad_norm": 0.41622287034988403,
+      "learning_rate": 0.00020523767605633804,
+      "loss": 1.4676,
+      "step": 35500
+    },
+    {
+      "epoch": 6.045918367346939,
+      "grad_norm": 0.4315873980522156,
+      "learning_rate": 0.0002047975352112676,
+      "loss": 1.4701,
+      "step": 35550
+    },
+    {
+      "epoch": 6.054421768707483,
+      "grad_norm": 0.4140373170375824,
+      "learning_rate": 0.00020435739436619718,
+      "loss": 1.4625,
+      "step": 35600
+    },
+    {
+      "epoch": 6.062925170068027,
+      "grad_norm": 0.41319021582603455,
+      "learning_rate": 0.00020391725352112677,
+      "loss": 1.4674,
+      "step": 35650
+    },
+    {
+      "epoch": 6.071428571428571,
+      "grad_norm": 0.4266904294490814,
+      "learning_rate": 0.00020347711267605634,
+      "loss": 1.4651,
+      "step": 35700
+    },
+    {
+      "epoch": 6.079931972789115,
+      "grad_norm": 0.422049880027771,
+      "learning_rate": 0.00020304577464788732,
+      "loss": 1.4602,
+      "step": 35750
+    },
+    {
+      "epoch": 6.08843537414966,
+      "grad_norm": 0.4264945089817047,
+      "learning_rate": 0.0002026056338028169,
+      "loss": 1.4629,
+      "step": 35800
+    },
+    {
+      "epoch": 6.096938775510204,
+      "grad_norm": 0.47240814566612244,
+      "learning_rate": 0.00020216549295774648,
+      "loss": 1.4675,
+      "step": 35850
+    },
+    {
+      "epoch": 6.105442176870748,
+      "grad_norm": 0.4204316735267639,
+      "learning_rate": 0.00020172535211267608,
+      "loss": 1.4627,
+      "step": 35900
+    },
+    {
+      "epoch": 6.113945578231292,
+      "grad_norm": 0.41577088832855225,
+      "learning_rate": 0.00020128521126760562,
+      "loss": 1.4605,
+      "step": 35950
+    },
+    {
+      "epoch": 6.122448979591836,
+      "grad_norm": 0.4297136962413788,
+      "learning_rate": 0.0002008450704225352,
+      "loss": 1.4685,
+      "step": 36000
+    },
+    {
+      "epoch": 6.122448979591836,
+      "eval_loss": 1.6419459581375122,
+      "eval_runtime": 75.3644,
+      "eval_samples_per_second": 1245.137,
+      "eval_steps_per_second": 4.87,
+      "step": 36000
+    },
+    {
+      "epoch": 6.130952380952381,
+      "grad_norm": 0.408674418926239,
+      "learning_rate": 0.00020040492957746478,
+      "loss": 1.4666,
+      "step": 36050
+    },
+    {
+      "epoch": 6.139455782312925,
+      "grad_norm": 0.41451412439346313,
+      "learning_rate": 0.00019996478873239438,
+      "loss": 1.4611,
+      "step": 36100
+    },
+    {
+      "epoch": 6.1479591836734695,
+      "grad_norm": 0.4128754734992981,
+      "learning_rate": 0.00019952464788732395,
+      "loss": 1.4589,
+      "step": 36150
+    },
+    {
+      "epoch": 6.156462585034014,
+      "grad_norm": 0.4213225543498993,
+      "learning_rate": 0.00019908450704225354,
+      "loss": 1.462,
+      "step": 36200
+    },
+    {
+      "epoch": 6.164965986394558,
+      "grad_norm": 0.4226750433444977,
+      "learning_rate": 0.0001986443661971831,
+      "loss": 1.4605,
+      "step": 36250
+    },
+    {
+      "epoch": 6.173469387755102,
+      "grad_norm": 0.4176504611968994,
+      "learning_rate": 0.00019820422535211268,
+      "loss": 1.459,
+      "step": 36300
+    },
+    {
+      "epoch": 6.181972789115647,
+      "grad_norm": 0.42486241459846497,
+      "learning_rate": 0.00019776408450704225,
+      "loss": 1.4656,
+      "step": 36350
+    },
+    {
+      "epoch": 6.190476190476191,
+      "grad_norm": 0.40891167521476746,
+      "learning_rate": 0.00019732394366197184,
+      "loss": 1.4601,
+      "step": 36400
+    },
+    {
+      "epoch": 6.198979591836735,
+      "grad_norm": 0.4230293035507202,
+      "learning_rate": 0.0001968838028169014,
+      "loss": 1.4537,
+      "step": 36450
+    },
+    {
+      "epoch": 6.207482993197279,
+      "grad_norm": 0.4342026710510254,
+      "learning_rate": 0.000196443661971831,
+      "loss": 1.4575,
+      "step": 36500
+    },
+    {
+      "epoch": 6.215986394557823,
+      "grad_norm": 0.417348176240921,
+      "learning_rate": 0.00019600352112676057,
+      "loss": 1.4618,
+      "step": 36550
+    },
+    {
+      "epoch": 6.224489795918367,
+      "grad_norm": 0.4206349551677704,
+      "learning_rate": 0.00019556338028169017,
+      "loss": 1.4596,
+      "step": 36600
+    },
+    {
+      "epoch": 6.232993197278912,
+      "grad_norm": 0.42500296235084534,
+      "learning_rate": 0.0001951232394366197,
+      "loss": 1.4534,
+      "step": 36650
+    },
+    {
+      "epoch": 6.241496598639456,
+      "grad_norm": 0.42607611417770386,
+      "learning_rate": 0.0001946830985915493,
+      "loss": 1.4621,
+      "step": 36700
+    },
+    {
+      "epoch": 6.25,
+      "grad_norm": 0.4912338852882385,
+      "learning_rate": 0.00019424295774647887,
+      "loss": 1.4629,
+      "step": 36750
+    },
+    {
+      "epoch": 6.258503401360544,
+      "grad_norm": 0.40748193860054016,
+      "learning_rate": 0.00019380281690140847,
+      "loss": 1.4639,
+      "step": 36800
+    },
+    {
+      "epoch": 6.267006802721088,
+      "grad_norm": 0.41081297397613525,
+      "learning_rate": 0.00019336267605633804,
+      "loss": 1.4505,
+      "step": 36850
+    },
+    {
+      "epoch": 6.275510204081632,
+      "grad_norm": 0.41694825887680054,
+      "learning_rate": 0.0001929225352112676,
+      "loss": 1.4573,
+      "step": 36900
+    },
+    {
+      "epoch": 6.284013605442177,
+      "grad_norm": 0.4185777008533478,
+      "learning_rate": 0.0001924823943661972,
+      "loss": 1.4621,
+      "step": 36950
+    },
+    {
+      "epoch": 6.292517006802721,
+      "grad_norm": 0.6816411018371582,
+      "learning_rate": 0.00019204225352112674,
+      "loss": 1.4526,
+      "step": 37000
+    },
+    {
+      "epoch": 6.292517006802721,
+      "eval_loss": 1.6315878629684448,
+      "eval_runtime": 75.3702,
+      "eval_samples_per_second": 1245.041,
+      "eval_steps_per_second": 4.869,
+      "step": 37000
+    },
+    {
+      "epoch": 6.301020408163265,
+      "grad_norm": 0.4173048436641693,
+      "learning_rate": 0.00019160211267605634,
+      "loss": 1.4542,
+      "step": 37050
+    },
+    {
+      "epoch": 6.309523809523809,
+      "grad_norm": 0.4043162167072296,
+      "learning_rate": 0.0001911619718309859,
+      "loss": 1.4541,
+      "step": 37100
+    },
+    {
+      "epoch": 6.318027210884353,
+      "grad_norm": 0.4198301136493683,
+      "learning_rate": 0.0001907218309859155,
+      "loss": 1.4526,
+      "step": 37150
+    },
+    {
+      "epoch": 6.326530612244898,
+      "grad_norm": 0.4946387708187103,
+      "learning_rate": 0.00019028169014084507,
+      "loss": 1.4631,
+      "step": 37200
+    },
+    {
+      "epoch": 6.335034013605442,
+      "grad_norm": 0.431738018989563,
+      "learning_rate": 0.00018984154929577466,
+      "loss": 1.449,
+      "step": 37250
+    },
+    {
+      "epoch": 6.343537414965986,
+      "grad_norm": 0.41441062092781067,
+      "learning_rate": 0.00018940140845070423,
+      "loss": 1.4501,
+      "step": 37300
+    },
+    {
+      "epoch": 6.3520408163265305,
+      "grad_norm": 0.41452255845069885,
+      "learning_rate": 0.0001889612676056338,
+      "loss": 1.445,
+      "step": 37350
+    },
+    {
+      "epoch": 6.360544217687075,
+      "grad_norm": 0.4315445125102997,
+      "learning_rate": 0.00018852112676056337,
+      "loss": 1.4556,
+      "step": 37400
+    },
+    {
+      "epoch": 6.369047619047619,
+      "grad_norm": 0.41262757778167725,
+      "learning_rate": 0.00018808098591549297,
+      "loss": 1.4531,
+      "step": 37450
+    },
+    {
+      "epoch": 6.377551020408164,
+      "grad_norm": 0.42860937118530273,
+      "learning_rate": 0.00018764084507042253,
+      "loss": 1.4612,
+      "step": 37500
+    },
+    {
+      "epoch": 6.386054421768708,
+      "grad_norm": 0.4128934442996979,
+      "learning_rate": 0.00018720070422535213,
+      "loss": 1.4467,
+      "step": 37550
+    },
+    {
+      "epoch": 6.394557823129252,
+      "grad_norm": 0.4122573137283325,
+      "learning_rate": 0.0001867605633802817,
+      "loss": 1.446,
+      "step": 37600
+    },
+    {
+      "epoch": 6.403061224489796,
+      "grad_norm": 0.41068488359451294,
+      "learning_rate": 0.0001863204225352113,
+      "loss": 1.4515,
+      "step": 37650
+    },
+    {
+      "epoch": 6.41156462585034,
+      "grad_norm": 0.4696698784828186,
+      "learning_rate": 0.00018588028169014083,
+      "loss": 1.4508,
+      "step": 37700
+    },
+    {
+      "epoch": 6.420068027210885,
+      "grad_norm": 0.41070887446403503,
+      "learning_rate": 0.00018544014084507043,
+      "loss": 1.4496,
+      "step": 37750
+    },
+    {
+      "epoch": 6.428571428571429,
+      "grad_norm": 0.4017719626426697,
+      "learning_rate": 0.000185,
+      "loss": 1.4457,
+      "step": 37800
+    },
+    {
+      "epoch": 6.437074829931973,
+      "grad_norm": 0.420674592256546,
+      "learning_rate": 0.0001845598591549296,
+      "loss": 1.4524,
+      "step": 37850
+    },
+    {
+      "epoch": 6.445578231292517,
+      "grad_norm": 0.4091247618198395,
+      "learning_rate": 0.00018411971830985916,
+      "loss": 1.4454,
+      "step": 37900
+    },
+    {
+      "epoch": 6.454081632653061,
+      "grad_norm": 0.45321521162986755,
+      "learning_rate": 0.00018367957746478876,
+      "loss": 1.4511,
+      "step": 37950
+    },
+    {
+      "epoch": 6.462585034013605,
+      "grad_norm": 0.4094868004322052,
+      "learning_rate": 0.00018323943661971832,
+      "loss": 1.4523,
+      "step": 38000
+    },
+    {
+      "epoch": 6.462585034013605,
+      "eval_loss": 1.6195621490478516,
+      "eval_runtime": 75.373,
+      "eval_samples_per_second": 1244.995,
+      "eval_steps_per_second": 4.869,
+      "step": 38000
+    },
+    {
+      "epoch": 6.47108843537415,
+      "grad_norm": 0.41573402285575867,
+      "learning_rate": 0.00018279929577464787,
+      "loss": 1.4493,
+      "step": 38050
+    },
+    {
+      "epoch": 6.479591836734694,
+      "grad_norm": 0.4076358675956726,
+      "learning_rate": 0.00018235915492957746,
+      "loss": 1.4479,
+      "step": 38100
+    },
+    {
+      "epoch": 6.488095238095238,
+      "grad_norm": 0.41350802779197693,
+      "learning_rate": 0.00018192781690140844,
+      "loss": 1.4474,
+      "step": 38150
+    },
+    {
+      "epoch": 6.496598639455782,
+      "grad_norm": 0.42011725902557373,
+      "learning_rate": 0.00018149647887323944,
+      "loss": 1.4579,
+      "step": 38200
+    },
+    {
+      "epoch": 6.505102040816326,
+      "grad_norm": 0.4202333688735962,
+      "learning_rate": 0.000181056338028169,
+      "loss": 1.448,
+      "step": 38250
+    },
+    {
+      "epoch": 6.513605442176871,
+      "grad_norm": 0.41736331582069397,
+      "learning_rate": 0.0001806161971830986,
+      "loss": 1.4421,
+      "step": 38300
+    },
+    {
+      "epoch": 6.522108843537415,
+      "grad_norm": 0.4162420332431793,
+      "learning_rate": 0.00018017605633802817,
+      "loss": 1.4401,
+      "step": 38350
+    },
+    {
+      "epoch": 6.530612244897959,
+      "grad_norm": 0.40687423944473267,
+      "learning_rate": 0.00017973591549295777,
+      "loss": 1.4439,
+      "step": 38400
+    },
+    {
+      "epoch": 6.539115646258503,
+      "grad_norm": 0.42472130060195923,
+      "learning_rate": 0.0001792957746478873,
+      "loss": 1.4468,
+      "step": 38450
+    },
+    {
+      "epoch": 6.5476190476190474,
+      "grad_norm": 0.4216206669807434,
+      "learning_rate": 0.0001788556338028169,
+      "loss": 1.4454,
+      "step": 38500
+    },
+    {
+      "epoch": 6.5561224489795915,
+      "grad_norm": 0.4373023211956024,
+      "learning_rate": 0.00017841549295774647,
+      "loss": 1.4441,
+      "step": 38550
+    },
+    {
+      "epoch": 6.564625850340136,
+      "grad_norm": 0.4680120050907135,
+      "learning_rate": 0.00017797535211267607,
+      "loss": 1.4383,
+      "step": 38600
+    },
+    {
+      "epoch": 6.5731292517006805,
+      "grad_norm": 0.4241814911365509,
+      "learning_rate": 0.00017753521126760564,
+      "loss": 1.4413,
+      "step": 38650
+    },
+    {
+      "epoch": 6.581632653061225,
+      "grad_norm": 0.449913889169693,
+      "learning_rate": 0.00017709507042253523,
+      "loss": 1.4427,
+      "step": 38700
+    },
+    {
+      "epoch": 6.590136054421769,
+      "grad_norm": 0.4140756130218506,
+      "learning_rate": 0.0001766549295774648,
+      "loss": 1.4429,
+      "step": 38750
+    },
+    {
+      "epoch": 6.598639455782313,
+      "grad_norm": 0.4235837161540985,
+      "learning_rate": 0.00017621478873239437,
+      "loss": 1.4425,
+      "step": 38800
+    },
+    {
+      "epoch": 6.607142857142857,
+      "grad_norm": 0.4102327227592468,
+      "learning_rate": 0.00017577464788732394,
+      "loss": 1.4428,
+      "step": 38850
+    },
+    {
+      "epoch": 6.615646258503402,
+      "grad_norm": 0.41676509380340576,
+      "learning_rate": 0.00017533450704225353,
+      "loss": 1.4515,
+      "step": 38900
+    },
+    {
+      "epoch": 6.624149659863946,
+      "grad_norm": 0.41074737906455994,
+      "learning_rate": 0.0001748943661971831,
+      "loss": 1.4376,
+      "step": 38950
+    },
+    {
+      "epoch": 6.63265306122449,
+      "grad_norm": 0.4003889262676239,
+      "learning_rate": 0.00017445422535211267,
+      "loss": 1.4425,
+      "step": 39000
+    },
+    {
+      "epoch": 6.63265306122449,
+      "eval_loss": 1.6100125312805176,
+      "eval_runtime": 75.3679,
+      "eval_samples_per_second": 1245.078,
+      "eval_steps_per_second": 4.869,
+      "step": 39000
+    },
+    {
+      "epoch": 6.641156462585034,
+      "grad_norm": 0.4165302813053131,
+      "learning_rate": 0.00017401408450704226,
+      "loss": 1.4366,
+      "step": 39050
+    },
+    {
+      "epoch": 6.649659863945578,
+      "grad_norm": 0.4187438488006592,
+      "learning_rate": 0.00017357394366197183,
+      "loss": 1.4403,
+      "step": 39100
+    },
+    {
+      "epoch": 6.658163265306122,
+      "grad_norm": 0.41520532965660095,
+      "learning_rate": 0.0001731338028169014,
+      "loss": 1.4465,
+      "step": 39150
+    },
+    {
+      "epoch": 6.666666666666667,
+      "grad_norm": 0.47344011068344116,
+      "learning_rate": 0.00017269366197183097,
+      "loss": 1.4451,
+      "step": 39200
+    },
+    {
+      "epoch": 6.675170068027211,
+      "grad_norm": 0.4409021735191345,
+      "learning_rate": 0.00017225352112676057,
+      "loss": 1.4367,
+      "step": 39250
+    },
+    {
+      "epoch": 6.683673469387755,
+      "grad_norm": 0.4071614742279053,
+      "learning_rate": 0.00017181338028169013,
+      "loss": 1.4343,
+      "step": 39300
+    },
+    {
+      "epoch": 6.692176870748299,
+      "grad_norm": 0.43377685546875,
+      "learning_rate": 0.00017137323943661973,
+      "loss": 1.4387,
+      "step": 39350
+    },
+    {
+      "epoch": 6.700680272108843,
+      "grad_norm": 0.41328275203704834,
+      "learning_rate": 0.0001709330985915493,
+      "loss": 1.4308,
+      "step": 39400
+    },
+    {
+      "epoch": 6.709183673469388,
+      "grad_norm": 0.4269335865974426,
+      "learning_rate": 0.0001704929577464789,
+      "loss": 1.4329,
+      "step": 39450
+    },
+    {
+      "epoch": 6.717687074829932,
+      "grad_norm": 0.4136207401752472,
+      "learning_rate": 0.00017005281690140843,
+      "loss": 1.4386,
+      "step": 39500
+    },
+    {
+      "epoch": 6.726190476190476,
+      "grad_norm": 0.41490909457206726,
+      "learning_rate": 0.00016961267605633803,
+      "loss": 1.4348,
+      "step": 39550
+    },
+    {
+      "epoch": 6.73469387755102,
+      "grad_norm": 0.41925111413002014,
+      "learning_rate": 0.0001691725352112676,
+      "loss": 1.4329,
+      "step": 39600
+    },
+    {
+      "epoch": 6.743197278911564,
+      "grad_norm": 0.4161663353443146,
+      "learning_rate": 0.0001687323943661972,
+      "loss": 1.4325,
+      "step": 39650
+    },
+    {
+      "epoch": 6.7517006802721085,
+      "grad_norm": 0.4175470769405365,
+      "learning_rate": 0.00016829225352112676,
+      "loss": 1.4331,
+      "step": 39700
+    },
+    {
+      "epoch": 6.760204081632653,
+      "grad_norm": 0.40980637073516846,
+      "learning_rate": 0.00016785211267605636,
+      "loss": 1.436,
+      "step": 39750
+    },
+    {
+      "epoch": 6.7687074829931975,
+      "grad_norm": 0.4164717197418213,
+      "learning_rate": 0.00016741197183098592,
+      "loss": 1.4354,
+      "step": 39800
+    },
+    {
+      "epoch": 6.7772108843537415,
+      "grad_norm": 0.41800665855407715,
+      "learning_rate": 0.00016697183098591552,
+      "loss": 1.4321,
+      "step": 39850
+    },
+    {
+      "epoch": 6.785714285714286,
+      "grad_norm": 0.4796483814716339,
+      "learning_rate": 0.00016653169014084506,
+      "loss": 1.428,
+      "step": 39900
+    },
+    {
+      "epoch": 6.79421768707483,
+      "grad_norm": 0.40372607111930847,
+      "learning_rate": 0.00016609154929577466,
+      "loss": 1.4335,
+      "step": 39950
+    },
+    {
+      "epoch": 6.802721088435375,
+      "grad_norm": 0.42244842648506165,
+      "learning_rate": 0.00016565140845070423,
+      "loss": 1.4391,
+      "step": 40000
+    },
+    {
+      "epoch": 6.802721088435375,
+      "eval_loss": 1.6078286170959473,
+      "eval_runtime": 75.3844,
+      "eval_samples_per_second": 1244.806,
+      "eval_steps_per_second": 4.868,
+      "step": 40000
+    },
+    {
+      "epoch": 6.811224489795919,
+      "grad_norm": 0.41039007902145386,
+      "learning_rate": 0.00016521126760563382,
+      "loss": 1.4388,
+      "step": 40050
+    },
+    {
+      "epoch": 6.819727891156463,
+      "grad_norm": 0.4182026982307434,
+      "learning_rate": 0.0001647711267605634,
+      "loss": 1.4335,
+      "step": 40100
+    },
+    {
+      "epoch": 6.828231292517007,
+      "grad_norm": 0.4472804665565491,
+      "learning_rate": 0.00016433098591549296,
+      "loss": 1.4357,
+      "step": 40150
+    },
+    {
+      "epoch": 6.836734693877551,
+      "grad_norm": 0.4294123649597168,
+      "learning_rate": 0.00016389084507042255,
+      "loss": 1.4355,
+      "step": 40200
+    },
+    {
+      "epoch": 6.845238095238095,
+      "grad_norm": 0.41815003752708435,
+      "learning_rate": 0.0001634507042253521,
+      "loss": 1.4324,
+      "step": 40250
+    },
+    {
+      "epoch": 6.853741496598639,
+      "grad_norm": 0.4064941704273224,
+      "learning_rate": 0.0001630105633802817,
+      "loss": 1.4366,
+      "step": 40300
+    },
+    {
+      "epoch": 6.862244897959184,
+      "grad_norm": 0.42009902000427246,
+      "learning_rate": 0.00016257922535211267,
+      "loss": 1.4325,
+      "step": 40350
+    },
+    {
+      "epoch": 6.870748299319728,
+      "grad_norm": 0.3989739716053009,
+      "learning_rate": 0.00016213908450704226,
+      "loss": 1.4238,
+      "step": 40400
+    },
+    {
+      "epoch": 6.879251700680272,
+      "grad_norm": 0.4133879840373993,
+      "learning_rate": 0.00016169894366197183,
+      "loss": 1.4281,
+      "step": 40450
+    },
+    {
+      "epoch": 6.887755102040816,
+      "grad_norm": 0.42547741532325745,
+      "learning_rate": 0.00016125880281690142,
+      "loss": 1.4279,
+      "step": 40500
+    },
+    {
+      "epoch": 6.896258503401361,
+      "grad_norm": 0.3915853202342987,
+      "learning_rate": 0.000160818661971831,
+      "loss": 1.4351,
+      "step": 40550
+    },
+    {
+      "epoch": 6.904761904761905,
+      "grad_norm": 0.4148283302783966,
+      "learning_rate": 0.00016037852112676056,
+      "loss": 1.4332,
+      "step": 40600
+    },
+    {
+      "epoch": 6.913265306122449,
+      "grad_norm": 0.4044345021247864,
+      "learning_rate": 0.00015993838028169013,
+      "loss": 1.4304,
+      "step": 40650
+    },
+    {
+      "epoch": 6.921768707482993,
+      "grad_norm": 0.4137207269668579,
+      "learning_rate": 0.00015949823943661972,
+      "loss": 1.4342,
+      "step": 40700
+    },
+    {
+      "epoch": 6.930272108843537,
+      "grad_norm": 0.40803512930870056,
+      "learning_rate": 0.0001590580985915493,
+      "loss": 1.4287,
+      "step": 40750
+    },
+    {
+      "epoch": 6.938775510204081,
+      "grad_norm": 0.4186759889125824,
+      "learning_rate": 0.0001586179577464789,
+      "loss": 1.4269,
+      "step": 40800
+    },
+    {
+      "epoch": 6.947278911564625,
+      "grad_norm": 0.5999934077262878,
+      "learning_rate": 0.00015817781690140846,
+      "loss": 1.4267,
+      "step": 40850
+    },
+    {
+      "epoch": 6.95578231292517,
+      "grad_norm": 3.3650548458099365,
+      "learning_rate": 0.00015773767605633805,
+      "loss": 1.4273,
+      "step": 40900
+    },
+    {
+      "epoch": 6.964285714285714,
+      "grad_norm": 0.4189580976963043,
+      "learning_rate": 0.0001572975352112676,
+      "loss": 1.4246,
+      "step": 40950
+    },
+    {
+      "epoch": 6.9727891156462585,
+      "grad_norm": 0.4199207127094269,
+      "learning_rate": 0.0001568573943661972,
+      "loss": 1.4264,
+      "step": 41000
+    },
+    {
+      "epoch": 6.9727891156462585,
+      "eval_loss": 1.6002153158187866,
+      "eval_runtime": 75.3647,
+      "eval_samples_per_second": 1245.132,
+      "eval_steps_per_second": 4.87,
+      "step": 41000
+    },
+    {
+      "epoch": 6.9812925170068025,
+      "grad_norm": 0.4282417595386505,
+      "learning_rate": 0.00015641725352112676,
+      "loss": 1.4302,
+      "step": 41050
+    },
+    {
+      "epoch": 6.989795918367347,
+      "grad_norm": 0.4040624797344208,
+      "learning_rate": 0.00015597711267605635,
+      "loss": 1.4289,
+      "step": 41100
+    },
+    {
+      "epoch": 6.9982993197278915,
+      "grad_norm": 0.4131171703338623,
+      "learning_rate": 0.00015553697183098592,
+      "loss": 1.4265,
+      "step": 41150
+    },
+    {
+      "epoch": 7.006802721088436,
+      "grad_norm": 0.41220322251319885,
+      "learning_rate": 0.0001550968309859155,
+      "loss": 1.421,
+      "step": 41200
+    },
+    {
+      "epoch": 7.01530612244898,
+      "grad_norm": 0.41320499777793884,
+      "learning_rate": 0.00015465669014084508,
+      "loss": 1.4126,
+      "step": 41250
+    },
+    {
+      "epoch": 7.023809523809524,
+      "grad_norm": 0.43131372332572937,
+      "learning_rate": 0.00015421654929577463,
+      "loss": 1.4185,
+      "step": 41300
+    },
+    {
+      "epoch": 7.032312925170068,
+      "grad_norm": 0.41249164938926697,
+      "learning_rate": 0.00015377640845070422,
+      "loss": 1.4225,
+      "step": 41350
+    },
+    {
+      "epoch": 7.040816326530612,
+      "grad_norm": 0.41785308718681335,
+      "learning_rate": 0.0001533362676056338,
+      "loss": 1.4193,
+      "step": 41400
+    },
+    {
+      "epoch": 7.049319727891157,
+      "grad_norm": 0.4036597013473511,
+      "learning_rate": 0.00015289612676056338,
+      "loss": 1.417,
+      "step": 41450
+    },
+    {
+      "epoch": 7.057823129251701,
+      "grad_norm": 1.470516562461853,
+      "learning_rate": 0.00015245598591549295,
+      "loss": 1.42,
+      "step": 41500
+    },
+    {
+      "epoch": 7.066326530612245,
+      "grad_norm": 0.4129942059516907,
+      "learning_rate": 0.00015201584507042255,
+      "loss": 1.4191,
+      "step": 41550
+    },
+    {
+      "epoch": 7.074829931972789,
+      "grad_norm": 0.41847845911979675,
+      "learning_rate": 0.00015157570422535212,
+      "loss": 1.4146,
+      "step": 41600
+    },
+    {
+      "epoch": 7.083333333333333,
+      "grad_norm": 0.4097050428390503,
+      "learning_rate": 0.0001511355633802817,
+      "loss": 1.4202,
+      "step": 41650
+    },
+    {
+      "epoch": 7.091836734693878,
+      "grad_norm": 0.42190149426460266,
+      "learning_rate": 0.00015069542253521125,
+      "loss": 1.4191,
+      "step": 41700
+    },
+    {
+      "epoch": 7.100340136054422,
+      "grad_norm": 0.4126221835613251,
+      "learning_rate": 0.00015025528169014085,
+      "loss": 1.4157,
+      "step": 41750
+    },
+    {
+      "epoch": 7.108843537414966,
+      "grad_norm": 0.42145752906799316,
+      "learning_rate": 0.00014981514084507042,
+      "loss": 1.4152,
+      "step": 41800
+    },
+    {
+      "epoch": 7.11734693877551,
+      "grad_norm": 0.41415074467658997,
+      "learning_rate": 0.000149375,
+      "loss": 1.4233,
+      "step": 41850
+    },
+    {
+      "epoch": 7.125850340136054,
+      "grad_norm": 0.4122840166091919,
+      "learning_rate": 0.00014893485915492958,
+      "loss": 1.4194,
+      "step": 41900
+    },
+    {
+      "epoch": 7.134353741496598,
+      "grad_norm": 0.4208846986293793,
+      "learning_rate": 0.00014849471830985918,
+      "loss": 1.417,
+      "step": 41950
+    },
+    {
+      "epoch": 7.142857142857143,
+      "grad_norm": 0.41025856137275696,
+      "learning_rate": 0.00014805457746478874,
+      "loss": 1.4138,
+      "step": 42000
+    },
+    {
+      "epoch": 7.142857142857143,
+      "eval_loss": 1.5856564044952393,
+      "eval_runtime": 75.3402,
+      "eval_samples_per_second": 1245.536,
+      "eval_steps_per_second": 4.871,
+      "step": 42000
+    },
+    {
+      "epoch": 7.151360544217687,
+      "grad_norm": 0.4345760941505432,
+      "learning_rate": 0.0001476144366197183,
+      "loss": 1.4162,
+      "step": 42050
+    },
+    {
+      "epoch": 7.159863945578231,
+      "grad_norm": 0.42329126596450806,
+      "learning_rate": 0.00014717429577464788,
+      "loss": 1.4183,
+      "step": 42100
+    },
+    {
+      "epoch": 7.168367346938775,
+      "grad_norm": 0.45433273911476135,
+      "learning_rate": 0.00014673415492957748,
+      "loss": 1.4209,
+      "step": 42150
+    },
+    {
+      "epoch": 7.1768707482993195,
+      "grad_norm": 0.401620477437973,
+      "learning_rate": 0.00014629401408450704,
+      "loss": 1.4157,
+      "step": 42200
+    },
+    {
+      "epoch": 7.1853741496598635,
+      "grad_norm": 0.4111977219581604,
+      "learning_rate": 0.00014585387323943664,
+      "loss": 1.4176,
+      "step": 42250
+    },
+    {
+      "epoch": 7.1938775510204085,
+      "grad_norm": 0.4081551134586334,
+      "learning_rate": 0.0001454137323943662,
+      "loss": 1.4175,
+      "step": 42300
+    },
+    {
+      "epoch": 7.2023809523809526,
+      "grad_norm": 0.41629016399383545,
+      "learning_rate": 0.0001449735915492958,
+      "loss": 1.4156,
+      "step": 42350
+    },
+    {
+      "epoch": 7.210884353741497,
+      "grad_norm": 0.4202435314655304,
+      "learning_rate": 0.00014453345070422534,
+      "loss": 1.4158,
+      "step": 42400
+    },
+    {
+      "epoch": 7.219387755102041,
+      "grad_norm": 0.4117106795310974,
+      "learning_rate": 0.0001440933098591549,
+      "loss": 1.4128,
+      "step": 42450
+    },
+    {
+      "epoch": 7.227891156462585,
+      "grad_norm": 0.40955278277397156,
+      "learning_rate": 0.0001436531690140845,
+      "loss": 1.4101,
+      "step": 42500
+    },
+    {
+      "epoch": 7.23639455782313,
+      "grad_norm": 0.4062648415565491,
+      "learning_rate": 0.00014321302816901408,
+      "loss": 1.4123,
+      "step": 42550
+    },
+    {
+      "epoch": 7.244897959183674,
+      "grad_norm": 0.39645469188690186,
+      "learning_rate": 0.00014277288732394367,
+      "loss": 1.4122,
+      "step": 42600
+    },
+    {
+      "epoch": 7.253401360544218,
+      "grad_norm": 0.4102307856082916,
+      "learning_rate": 0.00014233274647887324,
+      "loss": 1.4142,
+      "step": 42650
+    },
+    {
+      "epoch": 7.261904761904762,
+      "grad_norm": 0.39169833064079285,
+      "learning_rate": 0.00014189260563380284,
+      "loss": 1.4123,
+      "step": 42700
+    },
+    {
+      "epoch": 7.270408163265306,
+      "grad_norm": 0.40885522961616516,
+      "learning_rate": 0.00014145246478873238,
+      "loss": 1.4099,
+      "step": 42750
+    },
+    {
+      "epoch": 7.27891156462585,
+      "grad_norm": 0.40849608182907104,
+      "learning_rate": 0.00014101232394366197,
+      "loss": 1.4112,
+      "step": 42800
+    },
+    {
+      "epoch": 7.287414965986395,
+      "grad_norm": 0.416453093290329,
+      "learning_rate": 0.00014057218309859154,
+      "loss": 1.4139,
+      "step": 42850
+    },
+    {
+      "epoch": 7.295918367346939,
+      "grad_norm": 0.40005016326904297,
+      "learning_rate": 0.00014013204225352114,
+      "loss": 1.4122,
+      "step": 42900
+    },
+    {
+      "epoch": 7.304421768707483,
+      "grad_norm": 0.40970712900161743,
+      "learning_rate": 0.0001396919014084507,
+      "loss": 1.4071,
+      "step": 42950
+    },
+    {
+      "epoch": 7.312925170068027,
+      "grad_norm": 0.3983543813228607,
+      "learning_rate": 0.0001392517605633803,
+      "loss": 1.4095,
+      "step": 43000
+    },
+    {
+      "epoch": 7.312925170068027,
+      "eval_loss": 1.581950068473816,
+      "eval_runtime": 75.8081,
+      "eval_samples_per_second": 1237.849,
+      "eval_steps_per_second": 4.841,
+      "step": 43000
+    },
+    {
+      "epoch": 7.321428571428571,
+      "grad_norm": 0.42998814582824707,
+      "learning_rate": 0.00013881161971830987,
+      "loss": 1.4165,
+      "step": 43050
+    },
+    {
+      "epoch": 7.329931972789115,
+      "grad_norm": 0.40602609515190125,
+      "learning_rate": 0.00013837147887323944,
+      "loss": 1.4079,
+      "step": 43100
+    },
+    {
+      "epoch": 7.33843537414966,
+      "grad_norm": 0.4095648229122162,
+      "learning_rate": 0.000137931338028169,
+      "loss": 1.4155,
+      "step": 43150
+    },
+    {
+      "epoch": 7.346938775510204,
+      "grad_norm": 0.5845966339111328,
+      "learning_rate": 0.0001374911971830986,
+      "loss": 1.3978,
+      "step": 43200
+    },
+    {
+      "epoch": 7.355442176870748,
+      "grad_norm": 0.39307621121406555,
+      "learning_rate": 0.00013705105633802817,
+      "loss": 1.4063,
+      "step": 43250
+    },
+    {
+      "epoch": 7.363945578231292,
+      "grad_norm": 0.41059941053390503,
+      "learning_rate": 0.00013661091549295776,
+      "loss": 1.4111,
+      "step": 43300
+    },
+    {
+      "epoch": 7.372448979591836,
+      "grad_norm": 0.4473901689052582,
+      "learning_rate": 0.00013617077464788733,
+      "loss": 1.4107,
+      "step": 43350
+    },
+    {
+      "epoch": 7.380952380952381,
+      "grad_norm": 0.4176762104034424,
+      "learning_rate": 0.00013573063380281693,
+      "loss": 1.4074,
+      "step": 43400
+    },
+    {
+      "epoch": 7.389455782312925,
+      "grad_norm": 0.40306830406188965,
+      "learning_rate": 0.00013529049295774647,
+      "loss": 1.4018,
+      "step": 43450
+    },
+    {
+      "epoch": 7.3979591836734695,
+      "grad_norm": 0.40745970606803894,
+      "learning_rate": 0.00013485035211267606,
+      "loss": 1.4094,
+      "step": 43500
+    },
+    {
+      "epoch": 7.406462585034014,
+      "grad_norm": 0.4084283411502838,
+      "learning_rate": 0.00013441021126760563,
+      "loss": 1.4065,
+      "step": 43550
+    },
+    {
+      "epoch": 7.414965986394558,
+      "grad_norm": 0.3982578217983246,
+      "learning_rate": 0.0001339700704225352,
+      "loss": 1.4058,
+      "step": 43600
+    },
+    {
+      "epoch": 7.423469387755102,
+      "grad_norm": 0.40555986762046814,
+      "learning_rate": 0.0001335299295774648,
+      "loss": 1.4071,
+      "step": 43650
+    },
+    {
+      "epoch": 7.431972789115647,
+      "grad_norm": 0.41015100479125977,
+      "learning_rate": 0.00013308978873239436,
+      "loss": 1.4055,
+      "step": 43700
+    },
+    {
+      "epoch": 7.440476190476191,
+      "grad_norm": 0.39785054326057434,
+      "learning_rate": 0.00013264964788732396,
+      "loss": 1.4052,
+      "step": 43750
+    },
+    {
+      "epoch": 7.448979591836735,
+      "grad_norm": 0.4046226441860199,
+      "learning_rate": 0.0001322095070422535,
+      "loss": 1.4099,
+      "step": 43800
+    },
+    {
+      "epoch": 7.457482993197279,
+      "grad_norm": 0.4013233482837677,
+      "learning_rate": 0.0001317693661971831,
+      "loss": 1.4105,
+      "step": 43850
+    },
+    {
+      "epoch": 7.465986394557823,
+      "grad_norm": 0.42137226462364197,
+      "learning_rate": 0.00013132922535211266,
+      "loss": 1.4074,
+      "step": 43900
+    },
+    {
+      "epoch": 7.474489795918368,
+      "grad_norm": 0.4147741496562958,
+      "learning_rate": 0.00013088908450704226,
+      "loss": 1.4055,
+      "step": 43950
+    },
+    {
+      "epoch": 7.482993197278912,
+      "grad_norm": 0.43600887060165405,
+      "learning_rate": 0.00013044894366197183,
+      "loss": 1.4049,
+      "step": 44000
+    },
+    {
+      "epoch": 7.482993197278912,
+      "eval_loss": 1.575681209564209,
+      "eval_runtime": 75.3426,
+      "eval_samples_per_second": 1245.498,
+      "eval_steps_per_second": 4.871,
+      "step": 44000
+    },
+    {
+      "epoch": 7.491496598639456,
+      "grad_norm": 0.40329524874687195,
+      "learning_rate": 0.00013000880281690142,
+      "loss": 1.406,
+      "step": 44050
+    },
+    {
+      "epoch": 7.5,
+      "grad_norm": 0.39978840947151184,
+      "learning_rate": 0.000129568661971831,
+      "loss": 1.4051,
+      "step": 44100
+    },
+    {
+      "epoch": 7.508503401360544,
+      "grad_norm": 0.4256359934806824,
+      "learning_rate": 0.00012912852112676056,
+      "loss": 1.4096,
+      "step": 44150
+    },
+    {
+      "epoch": 7.517006802721088,
+      "grad_norm": 0.41504326462745667,
+      "learning_rate": 0.00012868838028169013,
+      "loss": 1.411,
+      "step": 44200
+    },
+    {
+      "epoch": 7.525510204081632,
+      "grad_norm": 0.41557520627975464,
+      "learning_rate": 0.00012824823943661972,
+      "loss": 1.3988,
+      "step": 44250
+    },
+    {
+      "epoch": 7.534013605442177,
+      "grad_norm": 0.654144287109375,
+      "learning_rate": 0.0001278080985915493,
+      "loss": 1.4002,
+      "step": 44300
+    },
+    {
+      "epoch": 7.542517006802721,
+      "grad_norm": 0.40340688824653625,
+      "learning_rate": 0.0001273679577464789,
+      "loss": 1.407,
+      "step": 44350
+    },
+    {
+      "epoch": 7.551020408163265,
+      "grad_norm": 0.4111216068267822,
+      "learning_rate": 0.00012692781690140846,
+      "loss": 1.4051,
+      "step": 44400
+    },
+    {
+      "epoch": 7.559523809523809,
+      "grad_norm": 0.40301740169525146,
+      "learning_rate": 0.00012648767605633805,
+      "loss": 1.4063,
+      "step": 44450
+    },
+    {
+      "epoch": 7.568027210884353,
+      "grad_norm": 0.40716397762298584,
+      "learning_rate": 0.0001260475352112676,
+      "loss": 1.4026,
+      "step": 44500
+    },
+    {
+      "epoch": 7.576530612244898,
+      "grad_norm": 0.40946149826049805,
+      "learning_rate": 0.0001256073943661972,
+      "loss": 1.3999,
+      "step": 44550
+    },
+    {
+      "epoch": 7.585034013605442,
+      "grad_norm": 0.4165254533290863,
+      "learning_rate": 0.00012516725352112676,
+      "loss": 1.4061,
+      "step": 44600
+    },
+    {
+      "epoch": 7.593537414965986,
+      "grad_norm": 0.4061264097690582,
+      "learning_rate": 0.00012472711267605635,
+      "loss": 1.3952,
+      "step": 44650
+    },
+    {
+      "epoch": 7.6020408163265305,
+      "grad_norm": 0.4414292573928833,
+      "learning_rate": 0.00012428697183098592,
+      "loss": 1.4061,
+      "step": 44700
+    },
+    {
+      "epoch": 7.610544217687075,
+      "grad_norm": 0.39658012986183167,
+      "learning_rate": 0.0001238468309859155,
+      "loss": 1.3985,
+      "step": 44750
+    },
+    {
+      "epoch": 7.619047619047619,
+      "grad_norm": 0.7822299599647522,
+      "learning_rate": 0.00012340669014084508,
+      "loss": 1.3993,
+      "step": 44800
+    },
+    {
+      "epoch": 7.627551020408164,
+      "grad_norm": 0.3982444703578949,
+      "learning_rate": 0.00012296654929577465,
+      "loss": 1.4011,
+      "step": 44850
+    },
+    {
+      "epoch": 7.636054421768708,
+      "grad_norm": 0.40807053446769714,
+      "learning_rate": 0.00012252640845070422,
+      "loss": 1.3984,
+      "step": 44900
+    },
+    {
+      "epoch": 7.644557823129252,
+      "grad_norm": 0.4086779057979584,
+      "learning_rate": 0.00012208626760563382,
+      "loss": 1.403,
+      "step": 44950
+    },
+    {
+      "epoch": 7.653061224489796,
+      "grad_norm": 0.4078817367553711,
+      "learning_rate": 0.00012164612676056338,
+      "loss": 1.3974,
+      "step": 45000
+    },
+    {
+      "epoch": 7.653061224489796,
+      "eval_loss": 1.564176082611084,
+      "eval_runtime": 75.8799,
+      "eval_samples_per_second": 1236.678,
+      "eval_steps_per_second": 4.837,
+      "step": 45000
+    },
+    {
+      "epoch": 7.66156462585034,
+      "grad_norm": 0.4033880829811096,
+      "learning_rate": 0.00012120598591549297,
+      "loss": 1.396,
+      "step": 45050
+    },
+    {
+      "epoch": 7.670068027210885,
+      "grad_norm": 0.41422680020332336,
+      "learning_rate": 0.00012076584507042255,
+      "loss": 1.3999,
+      "step": 45100
+    },
+    {
+      "epoch": 7.678571428571429,
+      "grad_norm": 0.41265836358070374,
+      "learning_rate": 0.00012032570422535212,
+      "loss": 1.3974,
+      "step": 45150
+    },
+    {
+      "epoch": 7.687074829931973,
+      "grad_norm": 0.4050711989402771,
+      "learning_rate": 0.0001198855633802817,
+      "loss": 1.3992,
+      "step": 45200
+    },
+    {
+      "epoch": 7.695578231292517,
+      "grad_norm": 0.4541753828525543,
+      "learning_rate": 0.00011944542253521128,
+      "loss": 1.3956,
+      "step": 45250
+    },
+    {
+      "epoch": 7.704081632653061,
+      "grad_norm": 0.469691663980484,
+      "learning_rate": 0.00011900528169014085,
+      "loss": 1.3941,
+      "step": 45300
+    },
+    {
+      "epoch": 7.712585034013605,
+      "grad_norm": 0.4149336516857147,
+      "learning_rate": 0.00011856514084507042,
+      "loss": 1.3926,
+      "step": 45350
+    },
+    {
+      "epoch": 7.72108843537415,
+      "grad_norm": 0.3968436121940613,
+      "learning_rate": 0.000118125,
+      "loss": 1.3945,
+      "step": 45400
+    },
+    {
+      "epoch": 7.729591836734694,
+      "grad_norm": 0.4118001461029053,
+      "learning_rate": 0.00011768485915492958,
+      "loss": 1.3945,
+      "step": 45450
+    },
+    {
+      "epoch": 7.738095238095238,
+      "grad_norm": 0.4041975438594818,
+      "learning_rate": 0.00011724471830985915,
+      "loss": 1.3958,
+      "step": 45500
+    },
+    {
+      "epoch": 7.746598639455782,
+      "grad_norm": 0.4147978723049164,
+      "learning_rate": 0.00011680457746478873,
+      "loss": 1.3885,
+      "step": 45550
+    },
+    {
+      "epoch": 7.755102040816326,
+      "grad_norm": 0.4063569903373718,
+      "learning_rate": 0.00011636443661971831,
+      "loss": 1.3955,
+      "step": 45600
+    },
+    {
+      "epoch": 7.763605442176871,
+      "grad_norm": 0.4077317416667938,
+      "learning_rate": 0.00011592429577464788,
+      "loss": 1.3969,
+      "step": 45650
+    },
+    {
+      "epoch": 7.772108843537415,
+      "grad_norm": 0.4094686508178711,
+      "learning_rate": 0.00011548415492957746,
+      "loss": 1.3956,
+      "step": 45700
+    },
+    {
+      "epoch": 7.780612244897959,
+      "grad_norm": 0.46225419640541077,
+      "learning_rate": 0.00011504401408450704,
+      "loss": 1.3964,
+      "step": 45750
+    },
+    {
+      "epoch": 7.789115646258503,
+      "grad_norm": 0.40823882818222046,
+      "learning_rate": 0.00011460387323943663,
+      "loss": 1.3865,
+      "step": 45800
+    },
+    {
+      "epoch": 7.7976190476190474,
+      "grad_norm": 0.4115474224090576,
+      "learning_rate": 0.0001141637323943662,
+      "loss": 1.3946,
+      "step": 45850
+    },
+    {
+      "epoch": 7.8061224489795915,
+      "grad_norm": 0.4270274043083191,
+      "learning_rate": 0.00011372359154929578,
+      "loss": 1.3961,
+      "step": 45900
+    },
+    {
+      "epoch": 7.814625850340136,
+      "grad_norm": 0.4096102714538574,
+      "learning_rate": 0.00011328345070422536,
+      "loss": 1.3909,
+      "step": 45950
+    },
+    {
+      "epoch": 7.8231292517006805,
+      "grad_norm": 0.41915324330329895,
+      "learning_rate": 0.00011284330985915494,
+      "loss": 1.3942,
+      "step": 46000
+    },
+    {
+      "epoch": 7.8231292517006805,
+      "eval_loss": 1.5574978590011597,
+      "eval_runtime": 75.3679,
+      "eval_samples_per_second": 1245.08,
+      "eval_steps_per_second": 4.869,
+      "step": 46000
+    },
+    {
+      "epoch": 7.831632653061225,
+      "grad_norm": 0.42200446128845215,
+      "learning_rate": 0.00011241197183098592,
+      "loss": 1.3898,
+      "step": 46050
+    },
+    {
+      "epoch": 7.840136054421769,
+      "grad_norm": 0.40634581446647644,
+      "learning_rate": 0.0001119718309859155,
+      "loss": 1.3872,
+      "step": 46100
+    },
+    {
+      "epoch": 7.848639455782313,
+      "grad_norm": 0.40481993556022644,
+      "learning_rate": 0.00011153169014084508,
+      "loss": 1.394,
+      "step": 46150
+    },
+    {
+      "epoch": 7.857142857142857,
+      "grad_norm": 0.8925232887268066,
+      "learning_rate": 0.00011109154929577466,
+      "loss": 1.3921,
+      "step": 46200
+    },
+    {
+      "epoch": 7.865646258503402,
+      "grad_norm": 0.5260804295539856,
+      "learning_rate": 0.00011065140845070423,
+      "loss": 1.3892,
+      "step": 46250
+    },
+    {
+      "epoch": 7.874149659863946,
+      "grad_norm": 0.4104487895965576,
+      "learning_rate": 0.00011021126760563381,
+      "loss": 1.3966,
+      "step": 46300
+    },
+    {
+      "epoch": 7.88265306122449,
+      "grad_norm": 0.4222128391265869,
+      "learning_rate": 0.0001097711267605634,
+      "loss": 1.3911,
+      "step": 46350
+    },
+    {
+      "epoch": 7.891156462585034,
+      "grad_norm": 0.4017203748226166,
+      "learning_rate": 0.00010933098591549295,
+      "loss": 1.3858,
+      "step": 46400
+    },
+    {
+      "epoch": 7.899659863945578,
+      "grad_norm": 0.41633373498916626,
+      "learning_rate": 0.00010889084507042253,
+      "loss": 1.3888,
+      "step": 46450
+    },
+    {
+      "epoch": 7.908163265306122,
+      "grad_norm": 0.41837745904922485,
+      "learning_rate": 0.00010845070422535211,
+      "loss": 1.3827,
+      "step": 46500
+    },
+    {
+      "epoch": 7.916666666666667,
+      "grad_norm": 0.4108099341392517,
+      "learning_rate": 0.0001080105633802817,
+      "loss": 1.3867,
+      "step": 46550
+    },
+    {
+      "epoch": 7.925170068027211,
+      "grad_norm": 0.4089981019496918,
+      "learning_rate": 0.00010757042253521126,
+      "loss": 1.3883,
+      "step": 46600
+    },
+    {
+      "epoch": 7.933673469387755,
+      "grad_norm": 0.40665385127067566,
+      "learning_rate": 0.00010713028169014084,
+      "loss": 1.3867,
+      "step": 46650
+    },
+    {
+      "epoch": 7.942176870748299,
+      "grad_norm": 0.4587544798851013,
+      "learning_rate": 0.00010669014084507043,
+      "loss": 1.3887,
+      "step": 46700
+    },
+    {
+      "epoch": 7.950680272108843,
+      "grad_norm": 0.41734760999679565,
+      "learning_rate": 0.00010625,
+      "loss": 1.3865,
+      "step": 46750
+    },
+    {
+      "epoch": 7.959183673469388,
+      "grad_norm": 0.3909563422203064,
+      "learning_rate": 0.00010580985915492958,
+      "loss": 1.3826,
+      "step": 46800
+    },
+    {
+      "epoch": 7.967687074829932,
+      "grad_norm": 0.4237678647041321,
+      "learning_rate": 0.00010536971830985916,
+      "loss": 1.3844,
+      "step": 46850
+    },
+    {
+      "epoch": 7.976190476190476,
+      "grad_norm": 0.3990273177623749,
+      "learning_rate": 0.00010492957746478874,
+      "loss": 1.3834,
+      "step": 46900
+    },
+    {
+      "epoch": 7.98469387755102,
+      "grad_norm": 0.4431270360946655,
+      "learning_rate": 0.00010448943661971831,
+      "loss": 1.3829,
+      "step": 46950
+    },
+    {
+      "epoch": 7.993197278911564,
+      "grad_norm": 0.423221230506897,
+      "learning_rate": 0.00010404929577464789,
+      "loss": 1.3861,
+      "step": 47000
+    },
+    {
+      "epoch": 7.993197278911564,
+      "eval_loss": 1.5540677309036255,
+      "eval_runtime": 75.3265,
+      "eval_samples_per_second": 1245.764,
+      "eval_steps_per_second": 4.872,
+      "step": 47000
+    },
+    {
+      "epoch": 8.001700680272108,
+      "grad_norm": 0.4330544173717499,
+      "learning_rate": 0.00010360915492957747,
+      "loss": 1.3834,
+      "step": 47050
+    },
+    {
+      "epoch": 8.010204081632653,
+      "grad_norm": 0.4099094867706299,
+      "learning_rate": 0.00010316901408450704,
+      "loss": 1.3826,
+      "step": 47100
+    },
+    {
+      "epoch": 8.018707482993197,
+      "grad_norm": 0.41134965419769287,
+      "learning_rate": 0.00010272887323943662,
+      "loss": 1.3846,
+      "step": 47150
+    },
+    {
+      "epoch": 8.02721088435374,
+      "grad_norm": 0.4183215796947479,
+      "learning_rate": 0.0001022887323943662,
+      "loss": 1.379,
+      "step": 47200
+    },
+    {
+      "epoch": 8.035714285714286,
+      "grad_norm": 0.39894890785217285,
+      "learning_rate": 0.00010184859154929579,
+      "loss": 1.3839,
+      "step": 47250
+    },
+    {
+      "epoch": 8.04421768707483,
+      "grad_norm": 0.4777671992778778,
+      "learning_rate": 0.00010140845070422535,
+      "loss": 1.3789,
+      "step": 47300
+    },
+    {
+      "epoch": 8.052721088435375,
+      "grad_norm": 0.41188567876815796,
+      "learning_rate": 0.00010096830985915494,
+      "loss": 1.3848,
+      "step": 47350
+    },
+    {
+      "epoch": 8.061224489795919,
+      "grad_norm": 0.4015198349952698,
+      "learning_rate": 0.00010052816901408452,
+      "loss": 1.3883,
+      "step": 47400
+    },
+    {
+      "epoch": 8.069727891156463,
+      "grad_norm": 0.401507169008255,
+      "learning_rate": 0.00010008802816901409,
+      "loss": 1.381,
+      "step": 47450
+    },
+    {
+      "epoch": 8.078231292517007,
+      "grad_norm": 0.40761667490005493,
+      "learning_rate": 9.964788732394367e-05,
+      "loss": 1.3714,
+      "step": 47500
+    },
+    {
+      "epoch": 8.08673469387755,
+      "grad_norm": 0.427236407995224,
+      "learning_rate": 9.920774647887325e-05,
+      "loss": 1.3799,
+      "step": 47550
+    },
+    {
+      "epoch": 8.095238095238095,
+      "grad_norm": 0.40780577063560486,
+      "learning_rate": 9.876760563380282e-05,
+      "loss": 1.3762,
+      "step": 47600
+    },
+    {
+      "epoch": 8.103741496598639,
+      "grad_norm": 0.4071747064590454,
+      "learning_rate": 9.832746478873239e-05,
+      "loss": 1.3815,
+      "step": 47650
+    },
+    {
+      "epoch": 8.112244897959183,
+      "grad_norm": 0.40746966004371643,
+      "learning_rate": 9.788732394366197e-05,
+      "loss": 1.3794,
+      "step": 47700
+    },
+    {
+      "epoch": 8.120748299319727,
+      "grad_norm": 0.39780792593955994,
+      "learning_rate": 9.744718309859155e-05,
+      "loss": 1.3804,
+      "step": 47750
+    },
+    {
+      "epoch": 8.129251700680273,
+      "grad_norm": 0.4328888952732086,
+      "learning_rate": 9.700704225352112e-05,
+      "loss": 1.3792,
+      "step": 47800
+    },
+    {
+      "epoch": 8.137755102040817,
+      "grad_norm": 0.40212351083755493,
+      "learning_rate": 9.65669014084507e-05,
+      "loss": 1.3702,
+      "step": 47850
+    },
+    {
+      "epoch": 8.146258503401361,
+      "grad_norm": 0.3966144919395447,
+      "learning_rate": 9.612676056338028e-05,
+      "loss": 1.3761,
+      "step": 47900
+    },
+    {
+      "epoch": 8.154761904761905,
+      "grad_norm": 0.41284072399139404,
+      "learning_rate": 9.568661971830986e-05,
+      "loss": 1.3774,
+      "step": 47950
+    },
+    {
+      "epoch": 8.16326530612245,
+      "grad_norm": 0.41598987579345703,
+      "learning_rate": 9.524647887323943e-05,
+      "loss": 1.3746,
+      "step": 48000
+    },
+    {
+      "epoch": 8.16326530612245,
+      "eval_loss": 1.544783115386963,
+      "eval_runtime": 75.374,
+      "eval_samples_per_second": 1244.979,
+      "eval_steps_per_second": 4.869,
+      "step": 48000
+    },
+    {
+      "epoch": 8.171768707482993,
+      "grad_norm": 0.7141363024711609,
+      "learning_rate": 9.480633802816901e-05,
+      "loss": 1.378,
+      "step": 48050
+    },
+    {
+      "epoch": 8.180272108843537,
+      "grad_norm": 0.400291383266449,
+      "learning_rate": 9.43661971830986e-05,
+      "loss": 1.3805,
+      "step": 48100
+    },
+    {
+      "epoch": 8.188775510204081,
+      "grad_norm": 0.40866729617118835,
+      "learning_rate": 9.392605633802818e-05,
+      "loss": 1.3765,
+      "step": 48150
+    },
+    {
+      "epoch": 8.197278911564625,
+      "grad_norm": 0.39690279960632324,
+      "learning_rate": 9.348591549295775e-05,
+      "loss": 1.3747,
+      "step": 48200
+    },
+    {
+      "epoch": 8.20578231292517,
+      "grad_norm": 0.41751721501350403,
+      "learning_rate": 9.304577464788733e-05,
+      "loss": 1.376,
+      "step": 48250
+    },
+    {
+      "epoch": 8.214285714285714,
+      "grad_norm": 0.4105045795440674,
+      "learning_rate": 9.260563380281691e-05,
+      "loss": 1.383,
+      "step": 48300
+    },
+    {
+      "epoch": 8.22278911564626,
+      "grad_norm": 0.40793538093566895,
+      "learning_rate": 9.216549295774648e-05,
+      "loss": 1.3716,
+      "step": 48350
+    },
+    {
+      "epoch": 8.231292517006803,
+      "grad_norm": 0.746793270111084,
+      "learning_rate": 9.172535211267606e-05,
+      "loss": 1.3736,
+      "step": 48400
+    },
+    {
+      "epoch": 8.239795918367347,
+      "grad_norm": 0.41186049580574036,
+      "learning_rate": 9.128521126760564e-05,
+      "loss": 1.3751,
+      "step": 48450
+    },
+    {
+      "epoch": 8.248299319727892,
+      "grad_norm": 0.4152040481567383,
+      "learning_rate": 9.084507042253522e-05,
+      "loss": 1.3716,
+      "step": 48500
+    },
+    {
+      "epoch": 8.256802721088436,
+      "grad_norm": 0.4200298488140106,
+      "learning_rate": 9.040492957746479e-05,
+      "loss": 1.3762,
+      "step": 48550
+    },
+    {
+      "epoch": 8.26530612244898,
+      "grad_norm": 0.40081730484962463,
+      "learning_rate": 8.996478873239437e-05,
+      "loss": 1.371,
+      "step": 48600
+    },
+    {
+      "epoch": 8.273809523809524,
+      "grad_norm": 0.4015451967716217,
+      "learning_rate": 8.952464788732396e-05,
+      "loss": 1.3717,
+      "step": 48650
+    },
+    {
+      "epoch": 8.282312925170068,
+      "grad_norm": 0.41303500533103943,
+      "learning_rate": 8.908450704225352e-05,
+      "loss": 1.3793,
+      "step": 48700
+    },
+    {
+      "epoch": 8.290816326530612,
+      "grad_norm": 0.41865503787994385,
+      "learning_rate": 8.86443661971831e-05,
+      "loss": 1.3733,
+      "step": 48750
+    },
+    {
+      "epoch": 8.299319727891156,
+      "grad_norm": 0.39655590057373047,
+      "learning_rate": 8.820422535211267e-05,
+      "loss": 1.371,
+      "step": 48800
+    },
+    {
+      "epoch": 8.3078231292517,
+      "grad_norm": 0.4100252389907837,
+      "learning_rate": 8.776408450704226e-05,
+      "loss": 1.3778,
+      "step": 48850
+    },
+    {
+      "epoch": 8.316326530612244,
+      "grad_norm": 0.4039982259273529,
+      "learning_rate": 8.732394366197182e-05,
+      "loss": 1.3737,
+      "step": 48900
+    },
+    {
+      "epoch": 8.32482993197279,
+      "grad_norm": 0.40856197476387024,
+      "learning_rate": 8.68838028169014e-05,
+      "loss": 1.3687,
+      "step": 48950
+    },
+    {
+      "epoch": 8.333333333333334,
+      "grad_norm": 0.40824389457702637,
+      "learning_rate": 8.644366197183099e-05,
+      "loss": 1.3726,
+      "step": 49000
+    },
+    {
+      "epoch": 8.333333333333334,
+      "eval_loss": 1.5451842546463013,
+      "eval_runtime": 75.3658,
+      "eval_samples_per_second": 1245.114,
+      "eval_steps_per_second": 4.87,
+      "step": 49000
+    },
+    {
+      "epoch": 8.341836734693878,
+      "grad_norm": 0.40559035539627075,
+      "learning_rate": 8.600352112676056e-05,
+      "loss": 1.3714,
+      "step": 49050
+    },
+    {
+      "epoch": 8.350340136054422,
+      "grad_norm": 0.40669360756874084,
+      "learning_rate": 8.556338028169014e-05,
+      "loss": 1.3716,
+      "step": 49100
+    },
+    {
+      "epoch": 8.358843537414966,
+      "grad_norm": 0.40370744466781616,
+      "learning_rate": 8.513204225352113e-05,
+      "loss": 1.37,
+      "step": 49150
+    },
+    {
+      "epoch": 8.36734693877551,
+      "grad_norm": 0.5396425127983093,
+      "learning_rate": 8.469190140845071e-05,
+      "loss": 1.3769,
+      "step": 49200
+    },
+    {
+      "epoch": 8.375850340136054,
+      "grad_norm": 0.401902973651886,
+      "learning_rate": 8.425176056338028e-05,
+      "loss": 1.364,
+      "step": 49250
+    },
+    {
+      "epoch": 8.384353741496598,
+      "grad_norm": 0.7124184966087341,
+      "learning_rate": 8.381161971830986e-05,
+      "loss": 1.3658,
+      "step": 49300
+    },
+    {
+      "epoch": 8.392857142857142,
+      "grad_norm": 0.4373975396156311,
+      "learning_rate": 8.337147887323944e-05,
+      "loss": 1.3658,
+      "step": 49350
+    },
+    {
+      "epoch": 8.401360544217686,
+      "grad_norm": 0.40361514687538147,
+      "learning_rate": 8.293133802816902e-05,
+      "loss": 1.366,
+      "step": 49400
+    },
+    {
+      "epoch": 8.40986394557823,
+      "grad_norm": 0.4027824103832245,
+      "learning_rate": 8.249119718309859e-05,
+      "loss": 1.3726,
+      "step": 49450
+    },
+    {
+      "epoch": 8.418367346938776,
+      "grad_norm": 0.41340282559394836,
+      "learning_rate": 8.205105633802817e-05,
+      "loss": 1.3691,
+      "step": 49500
+    },
+    {
+      "epoch": 8.42687074829932,
+      "grad_norm": 0.41044774651527405,
+      "learning_rate": 8.161091549295776e-05,
+      "loss": 1.3715,
+      "step": 49550
+    },
+    {
+      "epoch": 8.435374149659864,
+      "grad_norm": 0.7769300937652588,
+      "learning_rate": 8.117077464788732e-05,
+      "loss": 1.3604,
+      "step": 49600
+    },
+    {
+      "epoch": 8.443877551020408,
+      "grad_norm": 0.3973638713359833,
+      "learning_rate": 8.07306338028169e-05,
+      "loss": 1.3692,
+      "step": 49650
+    },
+    {
+      "epoch": 8.452380952380953,
+      "grad_norm": 0.3998255133628845,
+      "learning_rate": 8.029049295774649e-05,
+      "loss": 1.3758,
+      "step": 49700
+    },
+    {
+      "epoch": 8.460884353741497,
+      "grad_norm": 0.4052984416484833,
+      "learning_rate": 7.985035211267607e-05,
+      "loss": 1.3695,
+      "step": 49750
+    },
+    {
+      "epoch": 8.46938775510204,
+      "grad_norm": 0.40513521432876587,
+      "learning_rate": 7.941021126760564e-05,
+      "loss": 1.365,
+      "step": 49800
+    },
+    {
+      "epoch": 8.477891156462585,
+      "grad_norm": 0.41850411891937256,
+      "learning_rate": 7.89700704225352e-05,
+      "loss": 1.371,
+      "step": 49850
+    },
+    {
+      "epoch": 8.486394557823129,
+      "grad_norm": 0.40076538920402527,
+      "learning_rate": 7.852992957746479e-05,
+      "loss": 1.3583,
+      "step": 49900
+    },
+    {
+      "epoch": 8.494897959183673,
+      "grad_norm": 0.40688443183898926,
+      "learning_rate": 7.808978873239436e-05,
+      "loss": 1.368,
+      "step": 49950
+    },
+    {
+      "epoch": 8.503401360544217,
+      "grad_norm": 0.399404913187027,
+      "learning_rate": 7.764964788732394e-05,
+      "loss": 1.3683,
+      "step": 50000
+    },
+    {
+      "epoch": 8.503401360544217,
+      "eval_loss": 1.5378692150115967,
+      "eval_runtime": 75.368,
+      "eval_samples_per_second": 1245.078,
+      "eval_steps_per_second": 4.869,
+      "step": 50000
+    },
+    {
+      "epoch": 8.511904761904763,
+      "grad_norm": 0.40362656116485596,
+      "learning_rate": 7.720950704225352e-05,
+      "loss": 1.3642,
+      "step": 50050
+    },
+    {
+      "epoch": 8.520408163265307,
+      "grad_norm": 0.41159501671791077,
+      "learning_rate": 7.67693661971831e-05,
+      "loss": 1.3559,
+      "step": 50100
+    },
+    {
+      "epoch": 8.52891156462585,
+      "grad_norm": 0.4064503014087677,
+      "learning_rate": 7.632922535211267e-05,
+      "loss": 1.3689,
+      "step": 50150
+    },
+    {
+      "epoch": 8.537414965986395,
+      "grad_norm": 0.3967163562774658,
+      "learning_rate": 7.588908450704225e-05,
+      "loss": 1.3645,
+      "step": 50200
+    },
+    {
+      "epoch": 8.545918367346939,
+      "grad_norm": 0.40805837512016296,
+      "learning_rate": 7.544894366197183e-05,
+      "loss": 1.3617,
+      "step": 50250
+    },
+    {
+      "epoch": 8.554421768707483,
+      "grad_norm": 0.39720582962036133,
+      "learning_rate": 7.500880281690142e-05,
+      "loss": 1.3606,
+      "step": 50300
+    },
+    {
+      "epoch": 8.562925170068027,
+      "grad_norm": 0.3969172537326813,
+      "learning_rate": 7.456866197183098e-05,
+      "loss": 1.3661,
+      "step": 50350
+    },
+    {
+      "epoch": 8.571428571428571,
+      "grad_norm": 0.40754255652427673,
+      "learning_rate": 7.412852112676057e-05,
+      "loss": 1.3674,
+      "step": 50400
+    },
+    {
+      "epoch": 8.579931972789115,
+      "grad_norm": 0.40162739157676697,
+      "learning_rate": 7.368838028169015e-05,
+      "loss": 1.3654,
+      "step": 50450
+    },
+    {
+      "epoch": 8.58843537414966,
+      "grad_norm": 0.4004862606525421,
+      "learning_rate": 7.324823943661972e-05,
+      "loss": 1.3567,
+      "step": 50500
+    },
+    {
+      "epoch": 8.596938775510203,
+      "grad_norm": 0.4237268269062042,
+      "learning_rate": 7.28080985915493e-05,
+      "loss": 1.3617,
+      "step": 50550
+    },
+    {
+      "epoch": 8.60544217687075,
+      "grad_norm": 0.4031754434108734,
+      "learning_rate": 7.236795774647888e-05,
+      "loss": 1.3597,
+      "step": 50600
+    },
+    {
+      "epoch": 8.613945578231293,
+      "grad_norm": 0.39399534463882446,
+      "learning_rate": 7.192781690140846e-05,
+      "loss": 1.3633,
+      "step": 50650
+    },
+    {
+      "epoch": 8.622448979591837,
+      "grad_norm": 0.4071154296398163,
+      "learning_rate": 7.148767605633803e-05,
+      "loss": 1.3634,
+      "step": 50700
+    },
+    {
+      "epoch": 8.630952380952381,
+      "grad_norm": 0.40059253573417664,
+      "learning_rate": 7.104753521126761e-05,
+      "loss": 1.3674,
+      "step": 50750
+    },
+    {
+      "epoch": 8.639455782312925,
+      "grad_norm": 0.4087289571762085,
+      "learning_rate": 7.060739436619719e-05,
+      "loss": 1.3646,
+      "step": 50800
+    },
+    {
+      "epoch": 8.64795918367347,
+      "grad_norm": 0.42607581615448,
+      "learning_rate": 7.016725352112676e-05,
+      "loss": 1.3645,
+      "step": 50850
+    },
+    {
+      "epoch": 8.656462585034014,
+      "grad_norm": 0.40754997730255127,
+      "learning_rate": 6.972711267605634e-05,
+      "loss": 1.3582,
+      "step": 50900
+    },
+    {
+      "epoch": 8.664965986394558,
+      "grad_norm": 0.4103423058986664,
+      "learning_rate": 6.928697183098592e-05,
+      "loss": 1.3635,
+      "step": 50950
+    },
+    {
+      "epoch": 8.673469387755102,
+      "grad_norm": 0.4093490540981293,
+      "learning_rate": 6.88468309859155e-05,
+      "loss": 1.3593,
+      "step": 51000
+    },
+    {
+      "epoch": 8.673469387755102,
+      "eval_loss": 1.5301626920700073,
+      "eval_runtime": 75.3714,
+      "eval_samples_per_second": 1245.021,
+      "eval_steps_per_second": 4.869,
+      "step": 51000
+    },
+    {
+      "epoch": 8.681972789115646,
+      "grad_norm": 0.403463751077652,
+      "learning_rate": 6.840669014084506e-05,
+      "loss": 1.3596,
+      "step": 51050
+    },
+    {
+      "epoch": 8.69047619047619,
+      "grad_norm": 0.406222403049469,
+      "learning_rate": 6.796654929577464e-05,
+      "loss": 1.3592,
+      "step": 51100
+    },
+    {
+      "epoch": 8.698979591836736,
+      "grad_norm": 0.41207653284072876,
+      "learning_rate": 6.752640845070423e-05,
+      "loss": 1.3574,
+      "step": 51150
+    },
+    {
+      "epoch": 8.70748299319728,
+      "grad_norm": 0.4042452573776245,
+      "learning_rate": 6.70862676056338e-05,
+      "loss": 1.3714,
+      "step": 51200
+    },
+    {
+      "epoch": 8.715986394557824,
+      "grad_norm": 0.3971654176712036,
+      "learning_rate": 6.664612676056338e-05,
+      "loss": 1.3557,
+      "step": 51250
+    },
+    {
+      "epoch": 8.724489795918368,
+      "grad_norm": 0.40285438299179077,
+      "learning_rate": 6.620598591549296e-05,
+      "loss": 1.3587,
+      "step": 51300
+    },
+    {
+      "epoch": 8.732993197278912,
+      "grad_norm": 0.4234530031681061,
+      "learning_rate": 6.576584507042254e-05,
+      "loss": 1.3682,
+      "step": 51350
+    },
+    {
+      "epoch": 8.741496598639456,
+      "grad_norm": 0.3918672502040863,
+      "learning_rate": 6.532570422535211e-05,
+      "loss": 1.3545,
+      "step": 51400
+    },
+    {
+      "epoch": 8.75,
+      "grad_norm": 0.4000791013240814,
+      "learning_rate": 6.488556338028169e-05,
+      "loss": 1.361,
+      "step": 51450
+    },
+    {
+      "epoch": 8.758503401360544,
+      "grad_norm": 0.41339510679244995,
+      "learning_rate": 6.444542253521127e-05,
+      "loss": 1.3595,
+      "step": 51500
+    },
+    {
+      "epoch": 8.767006802721088,
+      "grad_norm": 0.4353317618370056,
+      "learning_rate": 6.400528169014084e-05,
+      "loss": 1.3601,
+      "step": 51550
+    },
+    {
+      "epoch": 8.775510204081632,
+      "grad_norm": 0.39730769395828247,
+      "learning_rate": 6.356514084507042e-05,
+      "loss": 1.3547,
+      "step": 51600
+    },
+    {
+      "epoch": 8.784013605442176,
+      "grad_norm": 0.41856640577316284,
+      "learning_rate": 6.3125e-05,
+      "loss": 1.354,
+      "step": 51650
+    },
+    {
+      "epoch": 8.79251700680272,
+      "grad_norm": 0.40725892782211304,
+      "learning_rate": 6.268485915492958e-05,
+      "loss": 1.3609,
+      "step": 51700
+    },
+    {
+      "epoch": 8.801020408163264,
+      "grad_norm": 0.4020816385746002,
+      "learning_rate": 6.224471830985915e-05,
+      "loss": 1.3507,
+      "step": 51750
+    },
+    {
+      "epoch": 8.80952380952381,
+      "grad_norm": 0.4009547233581543,
+      "learning_rate": 6.180457746478873e-05,
+      "loss": 1.3588,
+      "step": 51800
+    },
+    {
+      "epoch": 8.818027210884354,
+      "grad_norm": 0.42511284351348877,
+      "learning_rate": 6.136443661971832e-05,
+      "loss": 1.3567,
+      "step": 51850
+    },
+    {
+      "epoch": 8.826530612244898,
+      "grad_norm": 0.4135643541812897,
+      "learning_rate": 6.0924295774647885e-05,
+      "loss": 1.3536,
+      "step": 51900
+    },
+    {
+      "epoch": 8.835034013605442,
+      "grad_norm": 0.39955195784568787,
+      "learning_rate": 6.048415492957746e-05,
+      "loss": 1.3564,
+      "step": 51950
+    },
+    {
+      "epoch": 8.843537414965986,
+      "grad_norm": 0.4043436348438263,
+      "learning_rate": 6.004401408450704e-05,
+      "loss": 1.3527,
+      "step": 52000
+    },
+    {
+      "epoch": 8.843537414965986,
+      "eval_loss": 1.521606206893921,
+      "eval_runtime": 75.3981,
+      "eval_samples_per_second": 1244.58,
+      "eval_steps_per_second": 4.867,
+      "step": 52000
+    },
+    {
+      "epoch": 8.85204081632653,
+      "grad_norm": 0.40607118606567383,
+      "learning_rate": 5.960387323943662e-05,
+      "loss": 1.3567,
+      "step": 52050
+    },
+    {
+      "epoch": 8.860544217687075,
+      "grad_norm": 0.4122581481933594,
+      "learning_rate": 5.91637323943662e-05,
+      "loss": 1.355,
+      "step": 52100
+    },
+    {
+      "epoch": 8.869047619047619,
+      "grad_norm": 0.4980255961418152,
+      "learning_rate": 5.8723591549295774e-05,
+      "loss": 1.3523,
+      "step": 52150
+    },
+    {
+      "epoch": 8.877551020408163,
+      "grad_norm": 0.4053143262863159,
+      "learning_rate": 5.8283450704225356e-05,
+      "loss": 1.3556,
+      "step": 52200
+    },
+    {
+      "epoch": 8.886054421768707,
+      "grad_norm": 0.40937647223472595,
+      "learning_rate": 5.785211267605634e-05,
+      "loss": 1.3597,
+      "step": 52250
+    },
+    {
+      "epoch": 8.89455782312925,
+      "grad_norm": 0.4028383195400238,
+      "learning_rate": 5.741197183098592e-05,
+      "loss": 1.3548,
+      "step": 52300
+    },
+    {
+      "epoch": 8.903061224489797,
+      "grad_norm": 0.40696024894714355,
+      "learning_rate": 5.6971830985915496e-05,
+      "loss": 1.3553,
+      "step": 52350
+    },
+    {
+      "epoch": 8.91156462585034,
+      "grad_norm": 0.4004703164100647,
+      "learning_rate": 5.653169014084507e-05,
+      "loss": 1.3515,
+      "step": 52400
+    },
+    {
+      "epoch": 8.920068027210885,
+      "grad_norm": 0.4049835503101349,
+      "learning_rate": 5.6091549295774646e-05,
+      "loss": 1.3529,
+      "step": 52450
+    },
+    {
+      "epoch": 8.928571428571429,
+      "grad_norm": 0.4118352234363556,
+      "learning_rate": 5.565140845070423e-05,
+      "loss": 1.3606,
+      "step": 52500
+    },
+    {
+      "epoch": 8.937074829931973,
+      "grad_norm": 0.40466126799583435,
+      "learning_rate": 5.52112676056338e-05,
+      "loss": 1.3509,
+      "step": 52550
+    },
+    {
+      "epoch": 8.945578231292517,
+      "grad_norm": 0.39501574635505676,
+      "learning_rate": 5.477112676056338e-05,
+      "loss": 1.3515,
+      "step": 52600
+    },
+    {
+      "epoch": 8.954081632653061,
+      "grad_norm": 0.4046393036842346,
+      "learning_rate": 5.433098591549296e-05,
+      "loss": 1.3584,
+      "step": 52650
+    },
+    {
+      "epoch": 8.962585034013605,
+      "grad_norm": 0.4039493203163147,
+      "learning_rate": 5.3890845070422535e-05,
+      "loss": 1.3513,
+      "step": 52700
+    },
+    {
+      "epoch": 8.97108843537415,
+      "grad_norm": 0.4195407032966614,
+      "learning_rate": 5.345070422535212e-05,
+      "loss": 1.3536,
+      "step": 52750
+    },
+    {
+      "epoch": 8.979591836734693,
+      "grad_norm": 0.39963847398757935,
+      "learning_rate": 5.301056338028169e-05,
+      "loss": 1.3511,
+      "step": 52800
+    },
+    {
+      "epoch": 8.988095238095237,
+      "grad_norm": 0.41947850584983826,
+      "learning_rate": 5.2570422535211274e-05,
+      "loss": 1.3558,
+      "step": 52850
+    },
+    {
+      "epoch": 8.996598639455783,
+      "grad_norm": 0.4046323597431183,
+      "learning_rate": 5.213028169014085e-05,
+      "loss": 1.3511,
+      "step": 52900
+    },
+    {
+      "epoch": 9.005102040816327,
+      "grad_norm": 0.4054872393608093,
+      "learning_rate": 5.1690140845070424e-05,
+      "loss": 1.3428,
+      "step": 52950
+    },
+    {
+      "epoch": 9.013605442176871,
+      "grad_norm": 0.4009184241294861,
+      "learning_rate": 5.125e-05,
+      "loss": 1.3434,
+      "step": 53000
+    },
+    {
+      "epoch": 9.013605442176871,
+      "eval_loss": 1.5165001153945923,
+      "eval_runtime": 75.3865,
+      "eval_samples_per_second": 1244.772,
+      "eval_steps_per_second": 4.868,
+      "step": 53000
+    },
+    {
+      "epoch": 9.022108843537415,
+      "grad_norm": 0.4195316433906555,
+      "learning_rate": 5.0809859154929574e-05,
+      "loss": 1.358,
+      "step": 53050
+    },
+    {
+      "epoch": 9.03061224489796,
+      "grad_norm": 0.40940186381340027,
+      "learning_rate": 5.0369718309859156e-05,
+      "loss": 1.3453,
+      "step": 53100
+    },
+    {
+      "epoch": 9.039115646258503,
+      "grad_norm": 0.4012869596481323,
+      "learning_rate": 4.992957746478873e-05,
+      "loss": 1.3531,
+      "step": 53150
+    },
+    {
+      "epoch": 9.047619047619047,
+      "grad_norm": 0.4076648950576782,
+      "learning_rate": 4.948943661971831e-05,
+      "loss": 1.3433,
+      "step": 53200
+    },
+    {
+      "epoch": 9.056122448979592,
+      "grad_norm": 0.40203115344047546,
+      "learning_rate": 4.904929577464789e-05,
+      "loss": 1.3491,
+      "step": 53250
+    },
+    {
+      "epoch": 9.064625850340136,
+      "grad_norm": 0.40481826663017273,
+      "learning_rate": 4.860915492957747e-05,
+      "loss": 1.3509,
+      "step": 53300
+    },
+    {
+      "epoch": 9.07312925170068,
+      "grad_norm": 0.4018038213253021,
+      "learning_rate": 4.8169014084507045e-05,
+      "loss": 1.3485,
+      "step": 53350
+    },
+    {
+      "epoch": 9.081632653061224,
+      "grad_norm": 0.4100002646446228,
+      "learning_rate": 4.772887323943662e-05,
+      "loss": 1.3428,
+      "step": 53400
+    },
+    {
+      "epoch": 9.09013605442177,
+      "grad_norm": 1.7801686525344849,
+      "learning_rate": 4.72887323943662e-05,
+      "loss": 1.3445,
+      "step": 53450
+    },
+    {
+      "epoch": 9.098639455782314,
+      "grad_norm": 0.39769893884658813,
+      "learning_rate": 4.684859154929578e-05,
+      "loss": 1.3467,
+      "step": 53500
+    },
+    {
+      "epoch": 9.107142857142858,
+      "grad_norm": 0.4195043444633484,
+      "learning_rate": 4.640845070422536e-05,
+      "loss": 1.348,
+      "step": 53550
+    },
+    {
+      "epoch": 9.115646258503402,
+      "grad_norm": 0.4053761959075928,
+      "learning_rate": 4.596830985915493e-05,
+      "loss": 1.3465,
+      "step": 53600
+    },
+    {
+      "epoch": 9.124149659863946,
+      "grad_norm": 0.4017139673233032,
+      "learning_rate": 4.552816901408451e-05,
+      "loss": 1.3448,
+      "step": 53650
+    },
+    {
+      "epoch": 9.13265306122449,
+      "grad_norm": 0.4139016270637512,
+      "learning_rate": 4.5088028169014084e-05,
+      "loss": 1.349,
+      "step": 53700
+    },
+    {
+      "epoch": 9.141156462585034,
+      "grad_norm": 0.39910098910331726,
+      "learning_rate": 4.464788732394366e-05,
+      "loss": 1.3379,
+      "step": 53750
+    },
+    {
+      "epoch": 9.149659863945578,
+      "grad_norm": 0.39038217067718506,
+      "learning_rate": 4.420774647887324e-05,
+      "loss": 1.3422,
+      "step": 53800
+    },
+    {
+      "epoch": 9.158163265306122,
+      "grad_norm": 0.4174290895462036,
+      "learning_rate": 4.3767605633802816e-05,
+      "loss": 1.3417,
+      "step": 53850
+    },
+    {
+      "epoch": 9.166666666666666,
+      "grad_norm": 0.399420827627182,
+      "learning_rate": 4.33274647887324e-05,
+      "loss": 1.3414,
+      "step": 53900
+    },
+    {
+      "epoch": 9.17517006802721,
+      "grad_norm": 0.3991691470146179,
+      "learning_rate": 4.288732394366197e-05,
+      "loss": 1.3457,
+      "step": 53950
+    },
+    {
+      "epoch": 9.183673469387756,
+      "grad_norm": 0.4257224500179291,
+      "learning_rate": 4.2447183098591555e-05,
+      "loss": 1.3389,
+      "step": 54000
+    },
+    {
+      "epoch": 9.183673469387756,
+      "eval_loss": 1.514726996421814,
+      "eval_runtime": 75.3819,
+      "eval_samples_per_second": 1244.848,
+      "eval_steps_per_second": 4.869,
+      "step": 54000
+    },
+    {
+      "epoch": 9.1921768707483,
+      "grad_norm": 0.40039774775505066,
+      "learning_rate": 4.200704225352113e-05,
+      "loss": 1.3415,
+      "step": 54050
+    },
+    {
+      "epoch": 9.200680272108844,
+      "grad_norm": 0.3933996558189392,
+      "learning_rate": 4.1566901408450705e-05,
+      "loss": 1.3426,
+      "step": 54100
+    },
+    {
+      "epoch": 9.209183673469388,
+      "grad_norm": 0.4128095805644989,
+      "learning_rate": 4.112676056338028e-05,
+      "loss": 1.3471,
+      "step": 54150
+    },
+    {
+      "epoch": 9.217687074829932,
+      "grad_norm": 0.4024725556373596,
+      "learning_rate": 4.0686619718309855e-05,
+      "loss": 1.3454,
+      "step": 54200
+    },
+    {
+      "epoch": 9.226190476190476,
+      "grad_norm": 0.4105578064918518,
+      "learning_rate": 4.024647887323944e-05,
+      "loss": 1.3399,
+      "step": 54250
+    },
+    {
+      "epoch": 9.23469387755102,
+      "grad_norm": 0.4047715663909912,
+      "learning_rate": 3.980633802816901e-05,
+      "loss": 1.3442,
+      "step": 54300
+    },
+    {
+      "epoch": 9.243197278911564,
+      "grad_norm": 0.3958626687526703,
+      "learning_rate": 3.9366197183098594e-05,
+      "loss": 1.3441,
+      "step": 54350
+    },
+    {
+      "epoch": 9.251700680272108,
+      "grad_norm": 0.40880346298217773,
+      "learning_rate": 3.892605633802817e-05,
+      "loss": 1.3416,
+      "step": 54400
+    },
+    {
+      "epoch": 9.260204081632653,
+      "grad_norm": 0.41129016876220703,
+      "learning_rate": 3.848591549295775e-05,
+      "loss": 1.3428,
+      "step": 54450
+    },
+    {
+      "epoch": 9.268707482993197,
+      "grad_norm": 0.39472246170043945,
+      "learning_rate": 3.8045774647887326e-05,
+      "loss": 1.3396,
+      "step": 54500
+    },
+    {
+      "epoch": 9.27721088435374,
+      "grad_norm": 0.40167415142059326,
+      "learning_rate": 3.76056338028169e-05,
+      "loss": 1.3425,
+      "step": 54550
+    },
+    {
+      "epoch": 9.285714285714286,
+      "grad_norm": 0.40154701471328735,
+      "learning_rate": 3.716549295774648e-05,
+      "loss": 1.3423,
+      "step": 54600
+    },
+    {
+      "epoch": 9.29421768707483,
+      "grad_norm": 0.40191853046417236,
+      "learning_rate": 3.672535211267606e-05,
+      "loss": 1.3397,
+      "step": 54650
+    },
+    {
+      "epoch": 9.302721088435375,
+      "grad_norm": 0.4014011025428772,
+      "learning_rate": 3.628521126760564e-05,
+      "loss": 1.3431,
+      "step": 54700
+    },
+    {
+      "epoch": 9.311224489795919,
+      "grad_norm": 0.4313969910144806,
+      "learning_rate": 3.584507042253521e-05,
+      "loss": 1.3413,
+      "step": 54750
+    },
+    {
+      "epoch": 9.319727891156463,
+      "grad_norm": 0.4143257141113281,
+      "learning_rate": 3.540492957746479e-05,
+      "loss": 1.3424,
+      "step": 54800
+    },
+    {
+      "epoch": 9.328231292517007,
+      "grad_norm": 0.40343379974365234,
+      "learning_rate": 3.4964788732394365e-05,
+      "loss": 1.3377,
+      "step": 54850
+    },
+    {
+      "epoch": 9.33673469387755,
+      "grad_norm": 0.4093693792819977,
+      "learning_rate": 3.452464788732394e-05,
+      "loss": 1.3426,
+      "step": 54900
+    },
+    {
+      "epoch": 9.345238095238095,
+      "grad_norm": 0.40565499663352966,
+      "learning_rate": 3.408450704225352e-05,
+      "loss": 1.3347,
+      "step": 54950
+    },
+    {
+      "epoch": 9.353741496598639,
+      "grad_norm": 0.40517303347587585,
+      "learning_rate": 3.3644366197183097e-05,
+      "loss": 1.3411,
+      "step": 55000
+    },
+    {
+      "epoch": 9.353741496598639,
+      "eval_loss": 1.5079046487808228,
+      "eval_runtime": 75.3606,
+      "eval_samples_per_second": 1245.199,
+      "eval_steps_per_second": 4.87,
+      "step": 55000
+    },
+    {
+      "epoch": 9.362244897959183,
+      "grad_norm": 0.4189765155315399,
+      "learning_rate": 3.320422535211268e-05,
+      "loss": 1.3391,
+      "step": 55050
+    },
+    {
+      "epoch": 9.370748299319727,
+      "grad_norm": 0.41205212473869324,
+      "learning_rate": 3.2764084507042253e-05,
+      "loss": 1.3335,
+      "step": 55100
+    },
+    {
+      "epoch": 9.379251700680273,
+      "grad_norm": 0.4138495624065399,
+      "learning_rate": 3.2323943661971835e-05,
+      "loss": 1.3385,
+      "step": 55150
+    },
+    {
+      "epoch": 9.387755102040817,
+      "grad_norm": 0.40465956926345825,
+      "learning_rate": 3.188380281690141e-05,
+      "loss": 1.3366,
+      "step": 55200
+    },
+    {
+      "epoch": 9.396258503401361,
+      "grad_norm": 0.41961055994033813,
+      "learning_rate": 3.144366197183099e-05,
+      "loss": 1.3386,
+      "step": 55250
+    },
+    {
+      "epoch": 9.404761904761905,
+      "grad_norm": 0.4017610251903534,
+      "learning_rate": 3.100352112676056e-05,
+      "loss": 1.3389,
+      "step": 55300
+    },
+    {
+      "epoch": 9.41326530612245,
+      "grad_norm": 0.3947182893753052,
+      "learning_rate": 3.056338028169014e-05,
+      "loss": 1.3366,
+      "step": 55350
+    },
+    {
+      "epoch": 9.421768707482993,
+      "grad_norm": 0.41199058294296265,
+      "learning_rate": 3.012323943661972e-05,
+      "loss": 1.3319,
+      "step": 55400
+    },
+    {
+      "epoch": 9.430272108843537,
+      "grad_norm": 0.44556739926338196,
+      "learning_rate": 2.96830985915493e-05,
+      "loss": 1.3386,
+      "step": 55450
+    },
+    {
+      "epoch": 9.438775510204081,
+      "grad_norm": 0.4042491316795349,
+      "learning_rate": 2.924295774647887e-05,
+      "loss": 1.3352,
+      "step": 55500
+    },
+    {
+      "epoch": 9.447278911564625,
+      "grad_norm": 0.4033842384815216,
+      "learning_rate": 2.880281690140845e-05,
+      "loss": 1.3333,
+      "step": 55550
+    },
+    {
+      "epoch": 9.45578231292517,
+      "grad_norm": 0.409463495016098,
+      "learning_rate": 2.8362676056338028e-05,
+      "loss": 1.3359,
+      "step": 55600
+    },
+    {
+      "epoch": 9.464285714285714,
+      "grad_norm": 0.4049001634120941,
+      "learning_rate": 2.7922535211267606e-05,
+      "loss": 1.3362,
+      "step": 55650
+    },
+    {
+      "epoch": 9.47278911564626,
+      "grad_norm": 0.4021693766117096,
+      "learning_rate": 2.7482394366197185e-05,
+      "loss": 1.3334,
+      "step": 55700
+    },
+    {
+      "epoch": 9.481292517006803,
+      "grad_norm": 0.3943755030632019,
+      "learning_rate": 2.7042253521126763e-05,
+      "loss": 1.3356,
+      "step": 55750
+    },
+    {
+      "epoch": 9.489795918367347,
+      "grad_norm": 0.4090036153793335,
+      "learning_rate": 2.660211267605634e-05,
+      "loss": 1.3359,
+      "step": 55800
+    },
+    {
+      "epoch": 9.498299319727892,
+      "grad_norm": 0.3975585103034973,
+      "learning_rate": 2.6161971830985917e-05,
+      "loss": 1.3356,
+      "step": 55850
+    },
+    {
+      "epoch": 9.506802721088436,
+      "grad_norm": 0.3984331488609314,
+      "learning_rate": 2.5721830985915492e-05,
+      "loss": 1.3334,
+      "step": 55900
+    },
+    {
+      "epoch": 9.51530612244898,
+      "grad_norm": 0.3891252279281616,
+      "learning_rate": 2.528169014084507e-05,
+      "loss": 1.3297,
+      "step": 55950
+    },
+    {
+      "epoch": 9.523809523809524,
+      "grad_norm": 0.4065133333206177,
+      "learning_rate": 2.484154929577465e-05,
+      "loss": 1.3357,
+      "step": 56000
+    },
+    {
+      "epoch": 9.523809523809524,
+      "eval_loss": 1.4985688924789429,
+      "eval_runtime": 75.3846,
+      "eval_samples_per_second": 1244.803,
+      "eval_steps_per_second": 4.868,
+      "step": 56000
+    },
+    {
+      "epoch": 9.532312925170068,
+      "grad_norm": 0.403832346200943,
+      "learning_rate": 2.4401408450704227e-05,
+      "loss": 1.3389,
+      "step": 56050
+    },
+    {
+      "epoch": 9.540816326530612,
+      "grad_norm": 0.39873868227005005,
+      "learning_rate": 2.3961267605633802e-05,
+      "loss": 1.3344,
+      "step": 56100
+    },
+    {
+      "epoch": 9.549319727891156,
+      "grad_norm": 0.40277689695358276,
+      "learning_rate": 2.352112676056338e-05,
+      "loss": 1.3335,
+      "step": 56150
+    },
+    {
+      "epoch": 9.5578231292517,
+      "grad_norm": 0.39874860644340515,
+      "learning_rate": 2.308098591549296e-05,
+      "loss": 1.3377,
+      "step": 56200
+    },
+    {
+      "epoch": 9.566326530612244,
+      "grad_norm": 0.4036062955856323,
+      "learning_rate": 2.2640845070422538e-05,
+      "loss": 1.335,
+      "step": 56250
+    },
+    {
+      "epoch": 9.57482993197279,
+      "grad_norm": 0.41648924350738525,
+      "learning_rate": 2.2200704225352113e-05,
+      "loss": 1.3341,
+      "step": 56300
+    },
+    {
+      "epoch": 9.583333333333334,
+      "grad_norm": 0.39397549629211426,
+      "learning_rate": 2.176056338028169e-05,
+      "loss": 1.3366,
+      "step": 56350
+    },
+    {
+      "epoch": 9.591836734693878,
+      "grad_norm": 0.39437320828437805,
+      "learning_rate": 2.1320422535211266e-05,
+      "loss": 1.335,
+      "step": 56400
+    },
+    {
+      "epoch": 9.600340136054422,
+      "grad_norm": 0.39650458097457886,
+      "learning_rate": 2.0880281690140845e-05,
+      "loss": 1.3385,
+      "step": 56450
+    },
+    {
+      "epoch": 9.608843537414966,
+      "grad_norm": 0.4140496850013733,
+      "learning_rate": 2.0440140845070423e-05,
+      "loss": 1.3284,
+      "step": 56500
+    },
+    {
+      "epoch": 9.61734693877551,
+      "grad_norm": 0.39376911520957947,
+      "learning_rate": 2e-05,
+      "loss": 1.3307,
+      "step": 56550
+    },
+    {
+      "epoch": 9.625850340136054,
+      "grad_norm": 0.4171138405799866,
+      "learning_rate": 1.955985915492958e-05,
+      "loss": 1.333,
+      "step": 56600
+    },
+    {
+      "epoch": 9.634353741496598,
+      "grad_norm": 0.41457417607307434,
+      "learning_rate": 1.911971830985916e-05,
+      "loss": 1.3366,
+      "step": 56650
+    },
+    {
+      "epoch": 9.642857142857142,
+      "grad_norm": 0.3986314535140991,
+      "learning_rate": 1.867957746478873e-05,
+      "loss": 1.3382,
+      "step": 56700
+    },
+    {
+      "epoch": 9.651360544217686,
+      "grad_norm": 0.4179774820804596,
+      "learning_rate": 1.823943661971831e-05,
+      "loss": 1.3332,
+      "step": 56750
+    },
+    {
+      "epoch": 9.65986394557823,
+      "grad_norm": 0.4044124186038971,
+      "learning_rate": 1.7799295774647887e-05,
+      "loss": 1.3273,
+      "step": 56800
+    },
+    {
+      "epoch": 9.668367346938776,
+      "grad_norm": 0.4135982096195221,
+      "learning_rate": 1.7359154929577466e-05,
+      "loss": 1.3311,
+      "step": 56850
+    },
+    {
+      "epoch": 9.67687074829932,
+      "grad_norm": 0.39998510479927063,
+      "learning_rate": 1.6919014084507044e-05,
+      "loss": 1.3366,
+      "step": 56900
+    },
+    {
+      "epoch": 9.685374149659864,
+      "grad_norm": 0.40354451537132263,
+      "learning_rate": 1.6478873239436623e-05,
+      "loss": 1.3326,
+      "step": 56950
+    },
+    {
+      "epoch": 9.693877551020408,
+      "grad_norm": 0.4005592167377472,
+      "learning_rate": 1.6038732394366198e-05,
+      "loss": 1.3319,
+      "step": 57000
+    },
+    {
+      "epoch": 9.693877551020408,
+      "eval_loss": 1.5001171827316284,
+      "eval_runtime": 75.6646,
+      "eval_samples_per_second": 1240.196,
+      "eval_steps_per_second": 4.85,
+      "step": 57000
+    },
+    {
+      "epoch": 9.702380952380953,
+      "grad_norm": 0.3942902386188507,
+      "learning_rate": 1.5598591549295773e-05,
+      "loss": 1.3331,
+      "step": 57050
+    },
+    {
+      "epoch": 9.710884353741497,
+      "grad_norm": 0.4057478606700897,
+      "learning_rate": 1.5158450704225353e-05,
+      "loss": 1.3295,
+      "step": 57100
+    },
+    {
+      "epoch": 9.71938775510204,
+      "grad_norm": 0.40495890378952026,
+      "learning_rate": 1.471830985915493e-05,
+      "loss": 1.3419,
+      "step": 57150
+    },
+    {
+      "epoch": 9.727891156462585,
+      "grad_norm": 0.3950183093547821,
+      "learning_rate": 1.4278169014084506e-05,
+      "loss": 1.3316,
+      "step": 57200
+    },
+    {
+      "epoch": 9.736394557823129,
+      "grad_norm": 0.4021071195602417,
+      "learning_rate": 1.3838028169014085e-05,
+      "loss": 1.3297,
+      "step": 57250
+    },
+    {
+      "epoch": 9.744897959183673,
+      "grad_norm": 0.40653911232948303,
+      "learning_rate": 1.3397887323943663e-05,
+      "loss": 1.3329,
+      "step": 57300
+    },
+    {
+      "epoch": 9.753401360544217,
+      "grad_norm": 0.40104779601097107,
+      "learning_rate": 1.2957746478873238e-05,
+      "loss": 1.3234,
+      "step": 57350
+    },
+    {
+      "epoch": 9.761904761904763,
+      "grad_norm": 0.408331036567688,
+      "learning_rate": 1.2517605633802817e-05,
+      "loss": 1.3301,
+      "step": 57400
+    },
+    {
+      "epoch": 9.770408163265307,
+      "grad_norm": 0.4001865088939667,
+      "learning_rate": 1.2077464788732395e-05,
+      "loss": 1.3282,
+      "step": 57450
+    },
+    {
+      "epoch": 9.77891156462585,
+      "grad_norm": 0.4088568389415741,
+      "learning_rate": 1.1637323943661972e-05,
+      "loss": 1.3292,
+      "step": 57500
+    },
+    {
+      "epoch": 9.787414965986395,
+      "grad_norm": 0.42663103342056274,
+      "learning_rate": 1.1197183098591549e-05,
+      "loss": 1.3327,
+      "step": 57550
+    },
+    {
+      "epoch": 9.795918367346939,
+      "grad_norm": 0.3960348963737488,
+      "learning_rate": 1.0757042253521127e-05,
+      "loss": 1.3326,
+      "step": 57600
+    },
+    {
+      "epoch": 9.804421768707483,
+      "grad_norm": 0.39671510457992554,
+      "learning_rate": 1.0316901408450704e-05,
+      "loss": 1.3306,
+      "step": 57650
+    },
+    {
+      "epoch": 9.812925170068027,
+      "grad_norm": 0.39632678031921387,
+      "learning_rate": 9.876760563380282e-06,
+      "loss": 1.3311,
+      "step": 57700
+    },
+    {
+      "epoch": 9.821428571428571,
+      "grad_norm": 0.4076337516307831,
+      "learning_rate": 9.43661971830986e-06,
+      "loss": 1.3348,
+      "step": 57750
+    },
+    {
+      "epoch": 9.829931972789115,
+      "grad_norm": 0.40061214566230774,
+      "learning_rate": 8.996478873239436e-06,
+      "loss": 1.3323,
+      "step": 57800
+    },
+    {
+      "epoch": 9.83843537414966,
+      "grad_norm": 0.3868488371372223,
+      "learning_rate": 8.556338028169014e-06,
+      "loss": 1.3274,
+      "step": 57850
+    },
+    {
+      "epoch": 9.846938775510203,
+      "grad_norm": 0.43311014771461487,
+      "learning_rate": 8.116197183098593e-06,
+      "loss": 1.3316,
+      "step": 57900
+    },
+    {
+      "epoch": 9.85544217687075,
+      "grad_norm": 0.39683276414871216,
+      "learning_rate": 7.67605633802817e-06,
+      "loss": 1.3302,
+      "step": 57950
+    },
+    {
+      "epoch": 9.863945578231293,
+      "grad_norm": 0.3994409143924713,
+      "learning_rate": 7.235915492957746e-06,
+      "loss": 1.3311,
+      "step": 58000
+    },
+    {
+      "epoch": 9.863945578231293,
+      "eval_loss": 1.499453067779541,
+      "eval_runtime": 75.4095,
+      "eval_samples_per_second": 1244.392,
+      "eval_steps_per_second": 4.867,
+      "step": 58000
+    },
+    {
+      "epoch": 9.872448979591837,
+      "grad_norm": 0.40057429671287537,
+      "learning_rate": 6.795774647887324e-06,
+      "loss": 1.331,
+      "step": 58050
+    },
+    {
+      "epoch": 9.880952380952381,
+      "grad_norm": 0.39245638251304626,
+      "learning_rate": 6.355633802816902e-06,
+      "loss": 1.334,
+      "step": 58100
+    },
+    {
+      "epoch": 9.889455782312925,
+      "grad_norm": 0.3993644714355469,
+      "learning_rate": 5.915492957746479e-06,
+      "loss": 1.3267,
+      "step": 58150
+    },
+    {
+      "epoch": 9.89795918367347,
+      "grad_norm": 0.39710524678230286,
+      "learning_rate": 5.475352112676056e-06,
+      "loss": 1.3201,
+      "step": 58200
+    },
+    {
+      "epoch": 9.906462585034014,
+      "grad_norm": 0.3986225426197052,
+      "learning_rate": 5.0352112676056345e-06,
+      "loss": 1.3275,
+      "step": 58250
+    },
+    {
+      "epoch": 9.914965986394558,
+      "grad_norm": 0.4060732126235962,
+      "learning_rate": 4.595070422535211e-06,
+      "loss": 1.3274,
+      "step": 58300
+    },
+    {
+      "epoch": 9.923469387755102,
+      "grad_norm": 0.3890203833580017,
+      "learning_rate": 4.154929577464789e-06,
+      "loss": 1.3259,
+      "step": 58350
+    },
+    {
+      "epoch": 9.931972789115646,
+      "grad_norm": 0.39963120222091675,
+      "learning_rate": 3.7147887323943665e-06,
+      "loss": 1.3343,
+      "step": 58400
+    },
+    {
+      "epoch": 9.94047619047619,
+      "grad_norm": 0.4122091233730316,
+      "learning_rate": 3.274647887323944e-06,
+      "loss": 1.3313,
+      "step": 58450
+    },
+    {
+      "epoch": 9.948979591836736,
+      "grad_norm": 0.40625423192977905,
+      "learning_rate": 2.8433098591549298e-06,
+      "loss": 1.329,
+      "step": 58500
+    },
+    {
+      "epoch": 9.95748299319728,
+      "grad_norm": 0.39917826652526855,
+      "learning_rate": 2.4031690140845074e-06,
+      "loss": 1.329,
+      "step": 58550
+    },
+    {
+      "epoch": 9.965986394557824,
+      "grad_norm": 0.4005703330039978,
+      "learning_rate": 1.9630281690140846e-06,
+      "loss": 1.3286,
+      "step": 58600
+    },
+    {
+      "epoch": 9.974489795918368,
+      "grad_norm": 0.4052295386791229,
+      "learning_rate": 1.522887323943662e-06,
+      "loss": 1.3334,
+      "step": 58650
+    },
+    {
+      "epoch": 9.982993197278912,
+      "grad_norm": 0.3987277150154114,
+      "learning_rate": 1.0827464788732394e-06,
+      "loss": 1.3279,
+      "step": 58700
+    },
+    {
+      "epoch": 9.991496598639456,
+      "grad_norm": 0.39191877841949463,
+      "learning_rate": 6.426056338028169e-07,
+      "loss": 1.322,
+      "step": 58750
+    },
+    {
+      "epoch": 10.0,
+      "grad_norm": 0.39657750725746155,
+      "learning_rate": 2.0246478873239435e-07,
+      "loss": 1.3269,
+      "step": 58800
+    },
+    {
+      "epoch": 10.008503401360544,
+      "grad_norm": 0.4391802251338959,
+      "learning_rate": 0.000170377030162413,
+      "loss": 1.3375,
+      "step": 58850
+    },
+    {
+      "epoch": 10.017006802721088,
+      "grad_norm": 0.4108564257621765,
+      "learning_rate": 0.00017008700696055683,
+      "loss": 1.3625,
+      "step": 58900
+    },
+    {
+      "epoch": 10.025510204081632,
+      "grad_norm": 0.4225695729255676,
+      "learning_rate": 0.0001697969837587007,
+      "loss": 1.3704,
+      "step": 58950
+    },
+    {
+      "epoch": 10.034013605442176,
+      "grad_norm": 0.41939666867256165,
+      "learning_rate": 0.00016950696055684455,
+      "loss": 1.3693,
+      "step": 59000
+    },
+    {
+      "epoch": 10.034013605442176,
+      "eval_loss": 1.5424796342849731,
+      "eval_runtime": 75.2733,
+      "eval_samples_per_second": 1246.644,
+      "eval_steps_per_second": 4.876,
+      "step": 59000
+    },
+    {
+      "epoch": 10.04251700680272,
+      "grad_norm": 0.925662100315094,
+      "learning_rate": 0.00016921693735498842,
+      "loss": 1.3652,
+      "step": 59050
+    },
+    {
+      "epoch": 10.051020408163266,
+      "grad_norm": 0.42733487486839294,
+      "learning_rate": 0.00016892691415313225,
+      "loss": 1.3702,
+      "step": 59100
+    },
+    {
+      "epoch": 10.05952380952381,
+      "grad_norm": 0.43742504715919495,
+      "learning_rate": 0.0001686368909512761,
+      "loss": 1.3748,
+      "step": 59150
+    },
+    {
+      "epoch": 10.068027210884354,
+      "grad_norm": 0.5494563579559326,
+      "learning_rate": 0.00016834686774941994,
+      "loss": 1.3775,
+      "step": 59200
+    },
+    {
+      "epoch": 10.076530612244898,
+      "grad_norm": 0.43776175379753113,
+      "learning_rate": 0.0001680568445475638,
+      "loss": 1.3784,
+      "step": 59250
+    },
+    {
+      "epoch": 10.085034013605442,
+      "grad_norm": 0.4281368851661682,
+      "learning_rate": 0.00016776682134570767,
+      "loss": 1.3695,
+      "step": 59300
+    },
+    {
+      "epoch": 10.093537414965986,
+      "grad_norm": 0.5244280695915222,
+      "learning_rate": 0.0001674767981438515,
+      "loss": 1.3764,
+      "step": 59350
+    },
+    {
+      "epoch": 10.10204081632653,
+      "grad_norm": 0.40682801604270935,
+      "learning_rate": 0.00016718677494199536,
+      "loss": 1.3829,
+      "step": 59400
+    },
+    {
+      "epoch": 10.110544217687075,
+      "grad_norm": 0.42796218395233154,
+      "learning_rate": 0.00016690255220417631,
+      "loss": 1.376,
+      "step": 59450
+    },
+    {
+      "epoch": 10.119047619047619,
+      "grad_norm": 0.43184956908226013,
+      "learning_rate": 0.00016661252900232018,
+      "loss": 1.3794,
+      "step": 59500
+    },
+    {
+      "epoch": 10.127551020408163,
+      "grad_norm": 0.4699791371822357,
+      "learning_rate": 0.00016632250580046404,
+      "loss": 1.3849,
+      "step": 59550
+    },
+    {
+      "epoch": 10.136054421768707,
+      "grad_norm": 0.4018096923828125,
+      "learning_rate": 0.0001660324825986079,
+      "loss": 1.3765,
+      "step": 59600
+    },
+    {
+      "epoch": 10.14455782312925,
+      "grad_norm": 0.42693692445755005,
+      "learning_rate": 0.00016574245939675176,
+      "loss": 1.3726,
+      "step": 59650
+    },
+    {
+      "epoch": 10.153061224489797,
+      "grad_norm": 0.42560431361198425,
+      "learning_rate": 0.0001654524361948956,
+      "loss": 1.376,
+      "step": 59700
+    },
+    {
+      "epoch": 10.16156462585034,
+      "grad_norm": 0.42239609360694885,
+      "learning_rate": 0.00016516241299303943,
+      "loss": 1.3863,
+      "step": 59750
+    },
+    {
+      "epoch": 10.170068027210885,
+      "grad_norm": 0.41921770572662354,
+      "learning_rate": 0.0001648723897911833,
+      "loss": 1.3819,
+      "step": 59800
+    },
+    {
+      "epoch": 10.178571428571429,
+      "grad_norm": 0.41046616435050964,
+      "learning_rate": 0.00016458236658932715,
+      "loss": 1.3748,
+      "step": 59850
+    },
+    {
+      "epoch": 10.187074829931973,
+      "grad_norm": 0.41564711928367615,
+      "learning_rate": 0.000164292343387471,
+      "loss": 1.3802,
+      "step": 59900
+    },
+    {
+      "epoch": 10.195578231292517,
+      "grad_norm": 0.4275333881378174,
+      "learning_rate": 0.00016400232018561484,
+      "loss": 1.3757,
+      "step": 59950
+    },
+    {
+      "epoch": 10.204081632653061,
+      "grad_norm": 0.47480013966560364,
+      "learning_rate": 0.0001637122969837587,
+      "loss": 1.3858,
+      "step": 60000
+    },
+    {
+      "epoch": 10.204081632653061,
+      "eval_loss": 1.5496501922607422,
+      "eval_runtime": 75.245,
+      "eval_samples_per_second": 1247.113,
+      "eval_steps_per_second": 4.877,
+      "step": 60000
+    },
+    {
+      "epoch": 10.212585034013605,
+      "grad_norm": 2.0938565731048584,
+      "learning_rate": 0.00016342227378190257,
+      "loss": 1.3816,
+      "step": 60050
+    },
+    {
+      "epoch": 10.22108843537415,
+      "grad_norm": 0.4088470935821533,
+      "learning_rate": 0.0001631322505800464,
+      "loss": 1.3806,
+      "step": 60100
+    },
+    {
+      "epoch": 10.229591836734693,
+      "grad_norm": 0.43842536211013794,
+      "learning_rate": 0.00016284222737819026,
+      "loss": 1.3796,
+      "step": 60150
+    },
+    {
+      "epoch": 10.238095238095237,
+      "grad_norm": 0.41963255405426025,
+      "learning_rate": 0.00016255220417633412,
+      "loss": 1.3798,
+      "step": 60200
+    },
+    {
+      "epoch": 10.246598639455783,
+      "grad_norm": 0.4422262907028198,
+      "learning_rate": 0.00016226218097447796,
+      "loss": 1.3778,
+      "step": 60250
+    },
+    {
+      "epoch": 10.255102040816327,
+      "grad_norm": 0.43372678756713867,
+      "learning_rate": 0.00016197215777262182,
+      "loss": 1.3779,
+      "step": 60300
+    },
+    {
+      "epoch": 10.263605442176871,
+      "grad_norm": 0.42680031061172485,
+      "learning_rate": 0.00016168213457076568,
+      "loss": 1.3771,
+      "step": 60350
+    },
+    {
+      "epoch": 10.272108843537415,
+      "grad_norm": 0.4302733540534973,
+      "learning_rate": 0.0001613921113689095,
+      "loss": 1.3848,
+      "step": 60400
+    },
+    {
+      "epoch": 10.28061224489796,
+      "grad_norm": 0.4111240804195404,
+      "learning_rate": 0.00016110208816705337,
+      "loss": 1.3784,
+      "step": 60450
+    },
+    {
+      "epoch": 10.289115646258503,
+      "grad_norm": 0.43267524242401123,
+      "learning_rate": 0.0001608120649651972,
+      "loss": 1.3828,
+      "step": 60500
+    },
+    {
+      "epoch": 10.297619047619047,
+      "grad_norm": 0.4088628888130188,
+      "learning_rate": 0.00016052204176334107,
+      "loss": 1.3809,
+      "step": 60550
+    },
+    {
+      "epoch": 10.306122448979592,
+      "grad_norm": 0.4188777208328247,
+      "learning_rate": 0.00016023201856148493,
+      "loss": 1.3772,
+      "step": 60600
+    },
+    {
+      "epoch": 10.314625850340136,
+      "grad_norm": 0.4147690534591675,
+      "learning_rate": 0.0001599419953596288,
+      "loss": 1.3788,
+      "step": 60650
+    },
+    {
+      "epoch": 10.32312925170068,
+      "grad_norm": 0.48966458439826965,
+      "learning_rate": 0.00015965777262180974,
+      "loss": 1.378,
+      "step": 60700
+    },
+    {
+      "epoch": 10.331632653061224,
+      "grad_norm": 0.40553244948387146,
+      "learning_rate": 0.0001593677494199536,
+      "loss": 1.3879,
+      "step": 60750
+    },
+    {
+      "epoch": 10.34013605442177,
+      "grad_norm": 0.41595011949539185,
+      "learning_rate": 0.00015907772621809747,
+      "loss": 1.3801,
+      "step": 60800
+    },
+    {
+      "epoch": 10.348639455782314,
+      "grad_norm": 0.41190576553344727,
+      "learning_rate": 0.0001587877030162413,
+      "loss": 1.3838,
+      "step": 60850
+    },
+    {
+      "epoch": 10.357142857142858,
+      "grad_norm": 0.4079788625240326,
+      "learning_rate": 0.00015849767981438516,
+      "loss": 1.3885,
+      "step": 60900
+    },
+    {
+      "epoch": 10.365646258503402,
+      "grad_norm": 0.41605037450790405,
+      "learning_rate": 0.000158207656612529,
+      "loss": 1.3869,
+      "step": 60950
+    },
+    {
+      "epoch": 10.374149659863946,
+      "grad_norm": 0.41318464279174805,
+      "learning_rate": 0.00015791763341067286,
+      "loss": 1.3789,
+      "step": 61000
+    },
+    {
+      "epoch": 10.374149659863946,
+      "eval_loss": 1.549613356590271,
+      "eval_runtime": 75.299,
+      "eval_samples_per_second": 1246.219,
+      "eval_steps_per_second": 4.874,
+      "step": 61000
+    },
+    {
+      "epoch": 10.38265306122449,
+      "grad_norm": 0.42197248339653015,
+      "learning_rate": 0.00015762761020881672,
+      "loss": 1.3862,
+      "step": 61050
+    },
+    {
+      "epoch": 10.391156462585034,
+      "grad_norm": 0.423498272895813,
+      "learning_rate": 0.00015733758700696055,
+      "loss": 1.3831,
+      "step": 61100
+    },
+    {
+      "epoch": 10.399659863945578,
+      "grad_norm": 0.41805723309516907,
+      "learning_rate": 0.0001570475638051044,
+      "loss": 1.3807,
+      "step": 61150
+    },
+    {
+      "epoch": 10.408163265306122,
+      "grad_norm": 0.40790635347366333,
+      "learning_rate": 0.00015675754060324827,
+      "loss": 1.3818,
+      "step": 61200
+    },
+    {
+      "epoch": 10.416666666666666,
+      "grad_norm": 0.42006179690361023,
+      "learning_rate": 0.0001564675174013921,
+      "loss": 1.3775,
+      "step": 61250
+    },
+    {
+      "epoch": 10.42517006802721,
+      "grad_norm": 0.41415610909461975,
+      "learning_rate": 0.00015617749419953597,
+      "loss": 1.391,
+      "step": 61300
+    },
+    {
+      "epoch": 10.433673469387756,
+      "grad_norm": 0.44713294506073,
+      "learning_rate": 0.00015588747099767983,
+      "loss": 1.383,
+      "step": 61350
+    },
+    {
+      "epoch": 10.4421768707483,
+      "grad_norm": 0.44019782543182373,
+      "learning_rate": 0.00015559744779582366,
+      "loss": 1.3904,
+      "step": 61400
+    },
+    {
+      "epoch": 10.450680272108844,
+      "grad_norm": 0.4164921045303345,
+      "learning_rate": 0.00015530742459396752,
+      "loss": 1.3821,
+      "step": 61450
+    },
+    {
+      "epoch": 10.459183673469388,
+      "grad_norm": 0.4174480438232422,
+      "learning_rate": 0.00015501740139211139,
+      "loss": 1.3791,
+      "step": 61500
+    },
+    {
+      "epoch": 10.467687074829932,
+      "grad_norm": 0.49047085642814636,
+      "learning_rate": 0.00015472737819025522,
+      "loss": 1.3818,
+      "step": 61550
+    },
+    {
+      "epoch": 10.476190476190476,
+      "grad_norm": 0.405862957239151,
+      "learning_rate": 0.00015443735498839908,
+      "loss": 1.3817,
+      "step": 61600
+    },
+    {
+      "epoch": 10.48469387755102,
+      "grad_norm": 0.4351731538772583,
+      "learning_rate": 0.00015414733178654291,
+      "loss": 1.3806,
+      "step": 61650
+    },
+    {
+      "epoch": 10.493197278911564,
+      "grad_norm": 0.4245615303516388,
+      "learning_rate": 0.00015385730858468678,
+      "loss": 1.383,
+      "step": 61700
+    },
+    {
+      "epoch": 10.501700680272108,
+      "grad_norm": 0.39436227083206177,
+      "learning_rate": 0.00015356728538283064,
+      "loss": 1.3805,
+      "step": 61750
+    },
+    {
+      "epoch": 10.510204081632653,
+      "grad_norm": 0.41128745675086975,
+      "learning_rate": 0.0001532772621809745,
+      "loss": 1.386,
+      "step": 61800
+    },
+    {
+      "epoch": 10.518707482993197,
+      "grad_norm": 0.42046472430229187,
+      "learning_rate": 0.00015298723897911833,
+      "loss": 1.3897,
+      "step": 61850
+    },
+    {
+      "epoch": 10.527210884353742,
+      "grad_norm": 0.4486936926841736,
+      "learning_rate": 0.00015269721577726217,
+      "loss": 1.383,
+      "step": 61900
+    },
+    {
+      "epoch": 10.535714285714286,
+      "grad_norm": 0.4166390001773834,
+      "learning_rate": 0.00015240719257540603,
+      "loss": 1.3793,
+      "step": 61950
+    },
+    {
+      "epoch": 10.54421768707483,
+      "grad_norm": 0.42586439847946167,
+      "learning_rate": 0.0001521171693735499,
+      "loss": 1.3778,
+      "step": 62000
+    },
+    {
+      "epoch": 10.54421768707483,
+      "eval_loss": 1.5420804023742676,
+      "eval_runtime": 75.2999,
+      "eval_samples_per_second": 1246.204,
+      "eval_steps_per_second": 4.874,
+      "step": 62000
+    },
+    {
+      "epoch": 10.552721088435375,
+      "grad_norm": 0.39667069911956787,
+      "learning_rate": 0.00015182714617169375,
+      "loss": 1.3749,
+      "step": 62050
+    },
+    {
+      "epoch": 10.561224489795919,
+      "grad_norm": 0.40961819887161255,
+      "learning_rate": 0.0001515371229698376,
+      "loss": 1.3797,
+      "step": 62100
+    },
+    {
+      "epoch": 10.569727891156463,
+      "grad_norm": 0.4282205104827881,
+      "learning_rate": 0.00015124709976798144,
+      "loss": 1.3803,
+      "step": 62150
+    },
+    {
+      "epoch": 10.578231292517007,
+      "grad_norm": 0.42050084471702576,
+      "learning_rate": 0.00015095707656612528,
+      "loss": 1.3812,
+      "step": 62200
+    },
+    {
+      "epoch": 10.58673469387755,
+      "grad_norm": 0.4289666712284088,
+      "learning_rate": 0.00015066705336426914,
+      "loss": 1.3824,
+      "step": 62250
+    },
+    {
+      "epoch": 10.595238095238095,
+      "grad_norm": 0.4385221600532532,
+      "learning_rate": 0.000150377030162413,
+      "loss": 1.3733,
+      "step": 62300
+    },
+    {
+      "epoch": 10.603741496598639,
+      "grad_norm": 2.321608781814575,
+      "learning_rate": 0.00015008700696055686,
+      "loss": 1.3793,
+      "step": 62350
+    },
+    {
+      "epoch": 10.612244897959183,
+      "grad_norm": 0.4276087284088135,
+      "learning_rate": 0.00014979698375870072,
+      "loss": 1.382,
+      "step": 62400
+    },
+    {
+      "epoch": 10.620748299319727,
+      "grad_norm": 0.41472572088241577,
+      "learning_rate": 0.00014950696055684453,
+      "loss": 1.3772,
+      "step": 62450
+    },
+    {
+      "epoch": 10.629251700680273,
+      "grad_norm": 0.40703973174095154,
+      "learning_rate": 0.0001492169373549884,
+      "loss": 1.381,
+      "step": 62500
+    },
+    {
+      "epoch": 10.637755102040817,
+      "grad_norm": 0.41853588819503784,
+      "learning_rate": 0.00014892691415313225,
+      "loss": 1.377,
+      "step": 62550
+    },
+    {
+      "epoch": 10.646258503401361,
+      "grad_norm": 0.4194204807281494,
+      "learning_rate": 0.0001486368909512761,
+      "loss": 1.3803,
+      "step": 62600
+    },
+    {
+      "epoch": 10.654761904761905,
+      "grad_norm": 0.41111990809440613,
+      "learning_rate": 0.00014834686774941997,
+      "loss": 1.3778,
+      "step": 62650
+    },
+    {
+      "epoch": 10.66326530612245,
+      "grad_norm": 0.43283921480178833,
+      "learning_rate": 0.0001480568445475638,
+      "loss": 1.3759,
+      "step": 62700
+    },
+    {
+      "epoch": 10.671768707482993,
+      "grad_norm": 0.4069305956363678,
+      "learning_rate": 0.00014776682134570764,
+      "loss": 1.3779,
+      "step": 62750
+    },
+    {
+      "epoch": 10.680272108843537,
+      "grad_norm": 0.4265955984592438,
+      "learning_rate": 0.0001474767981438515,
+      "loss": 1.3818,
+      "step": 62800
+    },
+    {
+      "epoch": 10.688775510204081,
+      "grad_norm": 0.43886256217956543,
+      "learning_rate": 0.00014718677494199536,
+      "loss": 1.3739,
+      "step": 62850
+    },
+    {
+      "epoch": 10.697278911564625,
+      "grad_norm": 0.4144219756126404,
+      "learning_rate": 0.00014689675174013922,
+      "loss": 1.3748,
+      "step": 62900
+    },
+    {
+      "epoch": 10.70578231292517,
+      "grad_norm": 0.41580018401145935,
+      "learning_rate": 0.00014660672853828309,
+      "loss": 1.379,
+      "step": 62950
+    },
+    {
+      "epoch": 10.714285714285714,
+      "grad_norm": 0.41576939821243286,
+      "learning_rate": 0.00014631670533642692,
+      "loss": 1.3789,
+      "step": 63000
+    },
+    {
+      "epoch": 10.714285714285714,
+      "eval_loss": 1.5434892177581787,
+      "eval_runtime": 75.2563,
+      "eval_samples_per_second": 1246.925,
+      "eval_steps_per_second": 4.877,
+      "step": 63000
+    },
+    {
+      "epoch": 10.722789115646258,
+      "grad_norm": 0.44234171509742737,
+      "learning_rate": 0.00014602668213457075,
+      "loss": 1.3822,
+      "step": 63050
+    },
+    {
+      "epoch": 10.731292517006803,
+      "grad_norm": 0.4127776026725769,
+      "learning_rate": 0.00014573665893271461,
+      "loss": 1.3789,
+      "step": 63100
+    },
+    {
+      "epoch": 10.739795918367347,
+      "grad_norm": 0.4461323320865631,
+      "learning_rate": 0.00014544663573085848,
+      "loss": 1.3744,
+      "step": 63150
+    },
+    {
+      "epoch": 10.748299319727892,
+      "grad_norm": 0.5485215187072754,
+      "learning_rate": 0.00014515661252900234,
+      "loss": 1.3762,
+      "step": 63200
+    },
+    {
+      "epoch": 10.756802721088436,
+      "grad_norm": 0.4205337464809418,
+      "learning_rate": 0.00014486658932714617,
+      "loss": 1.3753,
+      "step": 63250
+    },
+    {
+      "epoch": 10.76530612244898,
+      "grad_norm": 0.4096565246582031,
+      "learning_rate": 0.00014457656612529003,
+      "loss": 1.3752,
+      "step": 63300
+    },
+    {
+      "epoch": 10.773809523809524,
+      "grad_norm": 0.44153106212615967,
+      "learning_rate": 0.00014428654292343387,
+      "loss": 1.3784,
+      "step": 63350
+    },
+    {
+      "epoch": 10.782312925170068,
+      "grad_norm": 0.41935479640960693,
+      "learning_rate": 0.00014399651972157773,
+      "loss": 1.3803,
+      "step": 63400
+    },
+    {
+      "epoch": 10.790816326530612,
+      "grad_norm": 0.40989023447036743,
+      "learning_rate": 0.0001437064965197216,
+      "loss": 1.3779,
+      "step": 63450
+    },
+    {
+      "epoch": 10.799319727891156,
+      "grad_norm": 0.41036033630371094,
+      "learning_rate": 0.00014341647331786542,
+      "loss": 1.38,
+      "step": 63500
+    },
+    {
+      "epoch": 10.8078231292517,
+      "grad_norm": 0.4159608781337738,
+      "learning_rate": 0.00014312645011600928,
+      "loss": 1.3775,
+      "step": 63550
+    },
+    {
+      "epoch": 10.816326530612244,
+      "grad_norm": 0.4197967052459717,
+      "learning_rate": 0.00014283642691415314,
+      "loss": 1.3763,
+      "step": 63600
+    },
+    {
+      "epoch": 10.82482993197279,
+      "grad_norm": 0.9038332104682922,
+      "learning_rate": 0.00014254640371229698,
+      "loss": 1.3774,
+      "step": 63650
+    },
+    {
+      "epoch": 10.833333333333334,
+      "grad_norm": 0.41402915120124817,
+      "learning_rate": 0.00014225638051044084,
+      "loss": 1.3699,
+      "step": 63700
+    },
+    {
+      "epoch": 10.841836734693878,
+      "grad_norm": 0.4222056567668915,
+      "learning_rate": 0.00014197215777262182,
+      "loss": 1.3716,
+      "step": 63750
+    },
+    {
+      "epoch": 10.850340136054422,
+      "grad_norm": 0.4178094267845154,
+      "learning_rate": 0.00014168213457076568,
+      "loss": 1.3776,
+      "step": 63800
+    },
+    {
+      "epoch": 10.858843537414966,
+      "grad_norm": 0.42892855405807495,
+      "learning_rate": 0.00014139211136890951,
+      "loss": 1.3747,
+      "step": 63850
+    },
+    {
+      "epoch": 10.86734693877551,
+      "grad_norm": 0.4110173285007477,
+      "learning_rate": 0.00014110208816705335,
+      "loss": 1.3807,
+      "step": 63900
+    },
+    {
+      "epoch": 10.875850340136054,
+      "grad_norm": 0.4244990050792694,
+      "learning_rate": 0.0001408120649651972,
+      "loss": 1.375,
+      "step": 63950
+    },
+    {
+      "epoch": 10.884353741496598,
+      "grad_norm": 0.5068066716194153,
+      "learning_rate": 0.00014052204176334107,
+      "loss": 1.3742,
+      "step": 64000
+    },
+    {
+      "epoch": 10.884353741496598,
+      "eval_loss": 1.5442039966583252,
+      "eval_runtime": 75.3144,
+      "eval_samples_per_second": 1245.964,
+      "eval_steps_per_second": 4.873,
+      "step": 64000
+    },
+    {
+      "epoch": 10.892857142857142,
+      "grad_norm": 0.4364668130874634,
+      "learning_rate": 0.00014023201856148493,
+      "loss": 1.3739,
+      "step": 64050
+    },
+    {
+      "epoch": 10.901360544217686,
+      "grad_norm": 0.4433183968067169,
+      "learning_rate": 0.00013994199535962877,
+      "loss": 1.3756,
+      "step": 64100
+    },
+    {
+      "epoch": 10.90986394557823,
+      "grad_norm": 0.41508781909942627,
+      "learning_rate": 0.00013965197215777263,
+      "loss": 1.3792,
+      "step": 64150
+    },
+    {
+      "epoch": 10.918367346938776,
+      "grad_norm": 0.4308226406574249,
+      "learning_rate": 0.00013936194895591646,
+      "loss": 1.3818,
+      "step": 64200
+    },
+    {
+      "epoch": 10.92687074829932,
+      "grad_norm": 0.43690988421440125,
+      "learning_rate": 0.00013907192575406032,
+      "loss": 1.3723,
+      "step": 64250
+    },
+    {
+      "epoch": 10.935374149659864,
+      "grad_norm": 0.9234839081764221,
+      "learning_rate": 0.00013878190255220418,
+      "loss": 1.3769,
+      "step": 64300
+    },
+    {
+      "epoch": 10.943877551020408,
+      "grad_norm": 0.4184640049934387,
+      "learning_rate": 0.00013849187935034804,
+      "loss": 1.3783,
+      "step": 64350
+    },
+    {
+      "epoch": 10.952380952380953,
+      "grad_norm": 0.418977826833725,
+      "learning_rate": 0.00013820185614849188,
+      "loss": 1.3708,
+      "step": 64400
+    },
+    {
+      "epoch": 10.960884353741497,
+      "grad_norm": 0.42863503098487854,
+      "learning_rate": 0.00013791183294663574,
+      "loss": 1.3643,
+      "step": 64450
+    },
+    {
+      "epoch": 10.96938775510204,
+      "grad_norm": 0.41124334931373596,
+      "learning_rate": 0.00013762180974477957,
+      "loss": 1.3788,
+      "step": 64500
+    },
+    {
+      "epoch": 10.977891156462585,
+      "grad_norm": 0.46177518367767334,
+      "learning_rate": 0.00013733178654292343,
+      "loss": 1.376,
+      "step": 64550
+    },
+    {
+      "epoch": 10.986394557823129,
+      "grad_norm": 0.4240172207355499,
+      "learning_rate": 0.0001370417633410673,
+      "loss": 1.3674,
+      "step": 64600
+    },
+    {
+      "epoch": 10.994897959183673,
+      "grad_norm": 0.4008079767227173,
+      "learning_rate": 0.00013675174013921113,
+      "loss": 1.3676,
+      "step": 64650
+    },
+    {
+      "epoch": 11.003401360544217,
+      "grad_norm": 0.40654611587524414,
+      "learning_rate": 0.000136461716937355,
+      "loss": 1.3716,
+      "step": 64700
+    },
+    {
+      "epoch": 11.011904761904763,
+      "grad_norm": 0.4086385667324066,
+      "learning_rate": 0.00013617169373549885,
+      "loss": 1.3692,
+      "step": 64750
+    },
+    {
+      "epoch": 11.020408163265307,
+      "grad_norm": 0.43329235911369324,
+      "learning_rate": 0.0001358816705336427,
+      "loss": 1.3694,
+      "step": 64800
+    },
+    {
+      "epoch": 11.02891156462585,
+      "grad_norm": 0.43725255131721497,
+      "learning_rate": 0.00013559164733178655,
+      "loss": 1.3648,
+      "step": 64850
+    },
+    {
+      "epoch": 11.037414965986395,
+      "grad_norm": 0.4894959628582001,
+      "learning_rate": 0.00013530162412993038,
+      "loss": 1.3681,
+      "step": 64900
+    },
+    {
+      "epoch": 11.045918367346939,
+      "grad_norm": 0.4180058538913727,
+      "learning_rate": 0.00013501160092807424,
+      "loss": 1.3637,
+      "step": 64950
+    },
+    {
+      "epoch": 11.054421768707483,
+      "grad_norm": 0.42746037244796753,
+      "learning_rate": 0.0001347215777262181,
+      "loss": 1.3674,
+      "step": 65000
+    },
+    {
+      "epoch": 11.054421768707483,
+      "eval_loss": 1.5379548072814941,
+      "eval_runtime": 75.1076,
+      "eval_samples_per_second": 1249.394,
+      "eval_steps_per_second": 4.886,
+      "step": 65000
+    },
+    {
+      "epoch": 11.062925170068027,
+      "grad_norm": 0.4063352644443512,
+      "learning_rate": 0.00013443155452436196,
+      "loss": 1.3644,
+      "step": 65050
+    },
+    {
+      "epoch": 11.071428571428571,
+      "grad_norm": 0.41466256976127625,
+      "learning_rate": 0.00013414153132250582,
+      "loss": 1.367,
+      "step": 65100
+    },
+    {
+      "epoch": 11.079931972789115,
+      "grad_norm": 0.40846332907676697,
+      "learning_rate": 0.00013385150812064966,
+      "loss": 1.3586,
+      "step": 65150
+    },
+    {
+      "epoch": 11.08843537414966,
+      "grad_norm": 0.4066697955131531,
+      "learning_rate": 0.0001335614849187935,
+      "loss": 1.3681,
+      "step": 65200
+    },
+    {
+      "epoch": 11.096938775510203,
+      "grad_norm": 0.4276025891304016,
+      "learning_rate": 0.00013327146171693735,
+      "loss": 1.3638,
+      "step": 65250
+    },
+    {
+      "epoch": 11.10544217687075,
+      "grad_norm": 0.4069554805755615,
+      "learning_rate": 0.00013298143851508121,
+      "loss": 1.3661,
+      "step": 65300
+    },
+    {
+      "epoch": 11.113945578231293,
+      "grad_norm": 0.5378205180168152,
+      "learning_rate": 0.00013269141531322508,
+      "loss": 1.365,
+      "step": 65350
+    },
+    {
+      "epoch": 11.122448979591837,
+      "grad_norm": 0.4328392744064331,
+      "learning_rate": 0.00013240139211136894,
+      "loss": 1.3681,
+      "step": 65400
+    },
+    {
+      "epoch": 11.130952380952381,
+      "grad_norm": 1.8447155952453613,
+      "learning_rate": 0.00013211136890951274,
+      "loss": 1.3643,
+      "step": 65450
+    },
+    {
+      "epoch": 11.139455782312925,
+      "grad_norm": 0.4180423319339752,
+      "learning_rate": 0.0001318213457076566,
+      "loss": 1.3644,
+      "step": 65500
+    },
+    {
+      "epoch": 11.14795918367347,
+      "grad_norm": 0.4122379720211029,
+      "learning_rate": 0.00013153132250580047,
+      "loss": 1.3626,
+      "step": 65550
+    },
+    {
+      "epoch": 11.156462585034014,
+      "grad_norm": 0.431257963180542,
+      "learning_rate": 0.00013124129930394433,
+      "loss": 1.3653,
+      "step": 65600
+    },
+    {
+      "epoch": 11.164965986394558,
+      "grad_norm": 0.39916712045669556,
+      "learning_rate": 0.0001309512761020882,
+      "loss": 1.3615,
+      "step": 65650
+    },
+    {
+      "epoch": 11.173469387755102,
+      "grad_norm": 0.4114948809146881,
+      "learning_rate": 0.00013066125290023202,
+      "loss": 1.37,
+      "step": 65700
+    },
+    {
+      "epoch": 11.181972789115646,
+      "grad_norm": 0.42501741647720337,
+      "learning_rate": 0.00013037122969837586,
+      "loss": 1.369,
+      "step": 65750
+    },
+    {
+      "epoch": 11.19047619047619,
+      "grad_norm": 0.4058555066585541,
+      "learning_rate": 0.00013008120649651972,
+      "loss": 1.3609,
+      "step": 65800
+    },
+    {
+      "epoch": 11.198979591836734,
+      "grad_norm": 0.40838614106178284,
+      "learning_rate": 0.00012979118329466358,
+      "loss": 1.3605,
+      "step": 65850
+    },
+    {
+      "epoch": 11.20748299319728,
+      "grad_norm": 0.4708266854286194,
+      "learning_rate": 0.00012950116009280744,
+      "loss": 1.3606,
+      "step": 65900
+    },
+    {
+      "epoch": 11.215986394557824,
+      "grad_norm": 0.40549665689468384,
+      "learning_rate": 0.0001292111368909513,
+      "loss": 1.3652,
+      "step": 65950
+    },
+    {
+      "epoch": 11.224489795918368,
+      "grad_norm": 0.4081163704395294,
+      "learning_rate": 0.00012892111368909513,
+      "loss": 1.3609,
+      "step": 66000
+    },
+    {
+      "epoch": 11.224489795918368,
+      "eval_loss": 1.5325791835784912,
+      "eval_runtime": 75.302,
+      "eval_samples_per_second": 1246.169,
+      "eval_steps_per_second": 4.874,
+      "step": 66000
+    },
+    {
+      "epoch": 11.232993197278912,
+      "grad_norm": 0.42392808198928833,
+      "learning_rate": 0.00012863109048723897,
+      "loss": 1.3617,
+      "step": 66050
+    },
+    {
+      "epoch": 11.241496598639456,
+      "grad_norm": 0.4850771725177765,
+      "learning_rate": 0.00012834106728538283,
+      "loss": 1.3733,
+      "step": 66100
+    },
+    {
+      "epoch": 11.25,
+      "grad_norm": 0.42814481258392334,
+      "learning_rate": 0.0001280510440835267,
+      "loss": 1.3652,
+      "step": 66150
+    },
+    {
+      "epoch": 11.258503401360544,
+      "grad_norm": 0.40000179409980774,
+      "learning_rate": 0.00012776102088167055,
+      "loss": 1.3717,
+      "step": 66200
+    },
+    {
+      "epoch": 11.267006802721088,
+      "grad_norm": 0.4144078493118286,
+      "learning_rate": 0.00012747099767981438,
+      "loss": 1.3645,
+      "step": 66250
+    },
+    {
+      "epoch": 11.275510204081632,
+      "grad_norm": 0.49152231216430664,
+      "learning_rate": 0.00012718097447795825,
+      "loss": 1.3661,
+      "step": 66300
+    },
+    {
+      "epoch": 11.284013605442176,
+      "grad_norm": 0.4072152078151703,
+      "learning_rate": 0.00012689095127610208,
+      "loss": 1.3671,
+      "step": 66350
+    },
+    {
+      "epoch": 11.29251700680272,
+      "grad_norm": 0.40859153866767883,
+      "learning_rate": 0.00012660092807424594,
+      "loss": 1.3667,
+      "step": 66400
+    },
+    {
+      "epoch": 11.301020408163266,
+      "grad_norm": 0.4389115869998932,
+      "learning_rate": 0.0001263109048723898,
+      "loss": 1.3648,
+      "step": 66450
+    },
+    {
+      "epoch": 11.30952380952381,
+      "grad_norm": 0.42225921154022217,
+      "learning_rate": 0.00012602088167053364,
+      "loss": 1.3584,
+      "step": 66500
+    },
+    {
+      "epoch": 11.318027210884354,
+      "grad_norm": 0.40919360518455505,
+      "learning_rate": 0.0001257308584686775,
+      "loss": 1.3613,
+      "step": 66550
+    },
+    {
+      "epoch": 11.326530612244898,
+      "grad_norm": 0.41595298051834106,
+      "learning_rate": 0.00012544083526682136,
+      "loss": 1.359,
+      "step": 66600
+    },
+    {
+      "epoch": 11.335034013605442,
+      "grad_norm": 0.405838280916214,
+      "learning_rate": 0.0001251508120649652,
+      "loss": 1.3581,
+      "step": 66650
+    },
+    {
+      "epoch": 11.343537414965986,
+      "grad_norm": 0.4283597767353058,
+      "learning_rate": 0.00012486078886310905,
+      "loss": 1.3653,
+      "step": 66700
+    },
+    {
+      "epoch": 11.35204081632653,
+      "grad_norm": 0.42228442430496216,
+      "learning_rate": 0.00012457076566125291,
+      "loss": 1.3666,
+      "step": 66750
+    },
+    {
+      "epoch": 11.360544217687075,
+      "grad_norm": 0.4238821566104889,
+      "learning_rate": 0.00012428074245939675,
+      "loss": 1.3621,
+      "step": 66800
+    },
+    {
+      "epoch": 11.369047619047619,
+      "grad_norm": 0.40990692377090454,
+      "learning_rate": 0.0001239907192575406,
+      "loss": 1.3663,
+      "step": 66850
+    },
+    {
+      "epoch": 11.377551020408163,
+      "grad_norm": 0.43895843625068665,
+      "learning_rate": 0.00012370069605568447,
+      "loss": 1.3581,
+      "step": 66900
+    },
+    {
+      "epoch": 11.386054421768707,
+      "grad_norm": 0.42183375358581543,
+      "learning_rate": 0.0001234106728538283,
+      "loss": 1.3655,
+      "step": 66950
+    },
+    {
+      "epoch": 11.39455782312925,
+      "grad_norm": 0.41603147983551025,
+      "learning_rate": 0.00012312064965197216,
+      "loss": 1.353,
+      "step": 67000
+    },
+    {
+      "epoch": 11.39455782312925,
+      "eval_loss": 1.5269132852554321,
+      "eval_runtime": 75.2762,
+      "eval_samples_per_second": 1246.596,
+      "eval_steps_per_second": 4.875,
+      "step": 67000
+    },
+    {
+      "epoch": 11.403061224489797,
+      "grad_norm": 0.42194870114326477,
+      "learning_rate": 0.00012283062645011603,
+      "loss": 1.3639,
+      "step": 67050
+    },
+    {
+      "epoch": 11.41156462585034,
+      "grad_norm": 0.42385604977607727,
+      "learning_rate": 0.00012254060324825986,
+      "loss": 1.3605,
+      "step": 67100
+    },
+    {
+      "epoch": 11.420068027210885,
+      "grad_norm": 0.4089812636375427,
+      "learning_rate": 0.00012225058004640372,
+      "loss": 1.3553,
+      "step": 67150
+    },
+    {
+      "epoch": 11.428571428571429,
+      "grad_norm": 0.402220219373703,
+      "learning_rate": 0.00012196055684454757,
+      "loss": 1.3643,
+      "step": 67200
+    },
+    {
+      "epoch": 11.437074829931973,
+      "grad_norm": 0.4062860310077667,
+      "learning_rate": 0.00012167053364269142,
+      "loss": 1.3615,
+      "step": 67250
+    },
+    {
+      "epoch": 11.445578231292517,
+      "grad_norm": 0.4270987808704376,
+      "learning_rate": 0.00012138051044083528,
+      "loss": 1.3586,
+      "step": 67300
+    },
+    {
+      "epoch": 11.454081632653061,
+      "grad_norm": 0.39737579226493835,
+      "learning_rate": 0.00012109048723897912,
+      "loss": 1.3566,
+      "step": 67350
+    },
+    {
+      "epoch": 11.462585034013605,
+      "grad_norm": 0.4278363287448883,
+      "learning_rate": 0.00012080046403712297,
+      "loss": 1.3597,
+      "step": 67400
+    },
+    {
+      "epoch": 11.47108843537415,
+      "grad_norm": 0.4197821319103241,
+      "learning_rate": 0.00012051044083526683,
+      "loss": 1.3627,
+      "step": 67450
+    },
+    {
+      "epoch": 11.479591836734693,
+      "grad_norm": 0.4141803979873657,
+      "learning_rate": 0.00012022041763341067,
+      "loss": 1.3576,
+      "step": 67500
+    },
+    {
+      "epoch": 11.488095238095237,
+      "grad_norm": 0.41713324189186096,
+      "learning_rate": 0.00011993039443155453,
+      "loss": 1.3544,
+      "step": 67550
+    },
+    {
+      "epoch": 11.496598639455783,
+      "grad_norm": 0.4070891737937927,
+      "learning_rate": 0.00011964037122969839,
+      "loss": 1.364,
+      "step": 67600
+    },
+    {
+      "epoch": 11.505102040816327,
+      "grad_norm": 0.4234033524990082,
+      "learning_rate": 0.00011935034802784222,
+      "loss": 1.3549,
+      "step": 67650
+    },
+    {
+      "epoch": 11.513605442176871,
+      "grad_norm": 0.4253610372543335,
+      "learning_rate": 0.00011906032482598608,
+      "loss": 1.36,
+      "step": 67700
+    },
+    {
+      "epoch": 11.522108843537415,
+      "grad_norm": 0.4099927246570587,
+      "learning_rate": 0.00011877030162412995,
+      "loss": 1.3639,
+      "step": 67750
+    },
+    {
+      "epoch": 11.53061224489796,
+      "grad_norm": 0.4126720726490021,
+      "learning_rate": 0.00011848607888631091,
+      "loss": 1.3629,
+      "step": 67800
+    },
+    {
+      "epoch": 11.539115646258503,
+      "grad_norm": 0.42233848571777344,
+      "learning_rate": 0.00011819605568445476,
+      "loss": 1.358,
+      "step": 67850
+    },
+    {
+      "epoch": 11.547619047619047,
+      "grad_norm": 0.42125800251960754,
+      "learning_rate": 0.00011790603248259861,
+      "loss": 1.3552,
+      "step": 67900
+    },
+    {
+      "epoch": 11.556122448979592,
+      "grad_norm": 0.4108964204788208,
+      "learning_rate": 0.00011761600928074247,
+      "loss": 1.3603,
+      "step": 67950
+    },
+    {
+      "epoch": 11.564625850340136,
+      "grad_norm": 0.4113200902938843,
+      "learning_rate": 0.00011732598607888632,
+      "loss": 1.3627,
+      "step": 68000
+    },
+    {
+      "epoch": 11.564625850340136,
+      "eval_loss": 1.5228930711746216,
+      "eval_runtime": 75.2543,
+      "eval_samples_per_second": 1246.958,
+      "eval_steps_per_second": 4.877,
+      "step": 68000
+    },
+    {
+      "epoch": 11.57312925170068,
+      "grad_norm": 0.43190208077430725,
+      "learning_rate": 0.00011703596287703016,
+      "loss": 1.354,
+      "step": 68050
+    },
+    {
+      "epoch": 11.581632653061224,
+      "grad_norm": 0.41988831758499146,
+      "learning_rate": 0.00011674593967517402,
+      "loss": 1.3598,
+      "step": 68100
+    },
+    {
+      "epoch": 11.59013605442177,
+      "grad_norm": 0.41497015953063965,
+      "learning_rate": 0.00011645591647331787,
+      "loss": 1.3432,
+      "step": 68150
+    },
+    {
+      "epoch": 11.598639455782314,
+      "grad_norm": 0.42652803659439087,
+      "learning_rate": 0.00011616589327146172,
+      "loss": 1.3565,
+      "step": 68200
+    },
+    {
+      "epoch": 11.607142857142858,
+      "grad_norm": 0.43378782272338867,
+      "learning_rate": 0.00011587587006960557,
+      "loss": 1.3538,
+      "step": 68250
+    },
+    {
+      "epoch": 11.615646258503402,
+      "grad_norm": 0.4106690287590027,
+      "learning_rate": 0.00011558584686774943,
+      "loss": 1.3514,
+      "step": 68300
+    },
+    {
+      "epoch": 11.624149659863946,
+      "grad_norm": 0.4098680913448334,
+      "learning_rate": 0.00011529582366589328,
+      "loss": 1.3628,
+      "step": 68350
+    },
+    {
+      "epoch": 11.63265306122449,
+      "grad_norm": 0.4068293869495392,
+      "learning_rate": 0.00011500580046403712,
+      "loss": 1.3561,
+      "step": 68400
+    },
+    {
+      "epoch": 11.641156462585034,
+      "grad_norm": 0.4472546875476837,
+      "learning_rate": 0.00011471577726218098,
+      "loss": 1.3565,
+      "step": 68450
+    },
+    {
+      "epoch": 11.649659863945578,
+      "grad_norm": 0.4033961594104767,
+      "learning_rate": 0.00011442575406032482,
+      "loss": 1.3545,
+      "step": 68500
+    },
+    {
+      "epoch": 11.658163265306122,
+      "grad_norm": 0.4254186749458313,
+      "learning_rate": 0.00011413573085846868,
+      "loss": 1.3571,
+      "step": 68550
+    },
+    {
+      "epoch": 11.666666666666666,
+      "grad_norm": 0.412675142288208,
+      "learning_rate": 0.00011384570765661254,
+      "loss": 1.3513,
+      "step": 68600
+    },
+    {
+      "epoch": 11.67517006802721,
+      "grad_norm": 0.4190445840358734,
+      "learning_rate": 0.00011355568445475637,
+      "loss": 1.3549,
+      "step": 68650
+    },
+    {
+      "epoch": 11.683673469387756,
+      "grad_norm": 0.4222116470336914,
+      "learning_rate": 0.00011326566125290024,
+      "loss": 1.3561,
+      "step": 68700
+    },
+    {
+      "epoch": 11.6921768707483,
+      "grad_norm": 0.41736775636672974,
+      "learning_rate": 0.0001129756380510441,
+      "loss": 1.3541,
+      "step": 68750
+    },
+    {
+      "epoch": 11.700680272108844,
+      "grad_norm": 0.40362343192100525,
+      "learning_rate": 0.00011268561484918793,
+      "loss": 1.3489,
+      "step": 68800
+    },
+    {
+      "epoch": 11.709183673469388,
+      "grad_norm": 0.40062522888183594,
+      "learning_rate": 0.00011239559164733179,
+      "loss": 1.355,
+      "step": 68850
+    },
+    {
+      "epoch": 11.717687074829932,
+      "grad_norm": 0.40170249342918396,
+      "learning_rate": 0.00011210556844547564,
+      "loss": 1.3549,
+      "step": 68900
+    },
+    {
+      "epoch": 11.726190476190476,
+      "grad_norm": 0.4284593164920807,
+      "learning_rate": 0.00011181554524361949,
+      "loss": 1.3502,
+      "step": 68950
+    },
+    {
+      "epoch": 11.73469387755102,
+      "grad_norm": 0.4220130145549774,
+      "learning_rate": 0.00011152552204176335,
+      "loss": 1.3508,
+      "step": 69000
+    },
+    {
+      "epoch": 11.73469387755102,
+      "eval_loss": 1.517913818359375,
+      "eval_runtime": 75.269,
+      "eval_samples_per_second": 1246.716,
+      "eval_steps_per_second": 4.876,
+      "step": 69000
+    },
+    {
+      "epoch": 11.743197278911564,
+      "grad_norm": 0.41656509041786194,
+      "learning_rate": 0.0001112354988399072,
+      "loss": 1.3573,
+      "step": 69050
+    },
+    {
+      "epoch": 11.751700680272108,
+      "grad_norm": 0.4296277165412903,
+      "learning_rate": 0.00011094547563805104,
+      "loss": 1.3586,
+      "step": 69100
+    },
+    {
+      "epoch": 11.760204081632653,
+      "grad_norm": 0.5169918537139893,
+      "learning_rate": 0.0001106554524361949,
+      "loss": 1.3502,
+      "step": 69150
+    },
+    {
+      "epoch": 11.768707482993197,
+      "grad_norm": 0.40869855880737305,
+      "learning_rate": 0.00011036542923433875,
+      "loss": 1.3586,
+      "step": 69200
+    },
+    {
+      "epoch": 11.777210884353742,
+      "grad_norm": 0.41961830854415894,
+      "learning_rate": 0.0001100754060324826,
+      "loss": 1.3563,
+      "step": 69250
+    },
+    {
+      "epoch": 11.785714285714286,
+      "grad_norm": 0.40917539596557617,
+      "learning_rate": 0.00010978538283062645,
+      "loss": 1.3554,
+      "step": 69300
+    },
+    {
+      "epoch": 11.79421768707483,
+      "grad_norm": 0.4206612706184387,
+      "learning_rate": 0.00010949535962877031,
+      "loss": 1.3568,
+      "step": 69350
+    },
+    {
+      "epoch": 11.802721088435375,
+      "grad_norm": 0.41846826672554016,
+      "learning_rate": 0.00010920533642691415,
+      "loss": 1.3518,
+      "step": 69400
+    },
+    {
+      "epoch": 11.811224489795919,
+      "grad_norm": 0.4007967412471771,
+      "learning_rate": 0.000108915313225058,
+      "loss": 1.3488,
+      "step": 69450
+    },
+    {
+      "epoch": 11.819727891156463,
+      "grad_norm": 0.45398712158203125,
+      "learning_rate": 0.00010862529002320186,
+      "loss": 1.3522,
+      "step": 69500
+    },
+    {
+      "epoch": 11.828231292517007,
+      "grad_norm": 0.4125469923019409,
+      "learning_rate": 0.00010833526682134571,
+      "loss": 1.3475,
+      "step": 69550
+    },
+    {
+      "epoch": 11.83673469387755,
+      "grad_norm": 0.4218948781490326,
+      "learning_rate": 0.00010804524361948956,
+      "loss": 1.3476,
+      "step": 69600
+    },
+    {
+      "epoch": 11.845238095238095,
+      "grad_norm": 0.4130956828594208,
+      "learning_rate": 0.00010775522041763342,
+      "loss": 1.3406,
+      "step": 69650
+    },
+    {
+      "epoch": 11.853741496598639,
+      "grad_norm": 0.4252954125404358,
+      "learning_rate": 0.00010746519721577725,
+      "loss": 1.3497,
+      "step": 69700
+    },
+    {
+      "epoch": 11.862244897959183,
+      "grad_norm": 0.4810144603252411,
+      "learning_rate": 0.00010717517401392111,
+      "loss": 1.3513,
+      "step": 69750
+    },
+    {
+      "epoch": 11.870748299319727,
+      "grad_norm": 0.4183041453361511,
+      "learning_rate": 0.00010688515081206498,
+      "loss": 1.3464,
+      "step": 69800
+    },
+    {
+      "epoch": 11.879251700680273,
+      "grad_norm": 0.4231345057487488,
+      "learning_rate": 0.00010659512761020881,
+      "loss": 1.348,
+      "step": 69850
+    },
+    {
+      "epoch": 11.887755102040817,
+      "grad_norm": 0.4193786084651947,
+      "learning_rate": 0.00010630510440835267,
+      "loss": 1.3494,
+      "step": 69900
+    },
+    {
+      "epoch": 11.896258503401361,
+      "grad_norm": 0.41639646887779236,
+      "learning_rate": 0.00010601508120649653,
+      "loss": 1.3489,
+      "step": 69950
+    },
+    {
+      "epoch": 11.904761904761905,
+      "grad_norm": 0.4062506854534149,
+      "learning_rate": 0.00010572505800464037,
+      "loss": 1.3478,
+      "step": 70000
+    },
+    {
+      "epoch": 11.904761904761905,
+      "eval_loss": 1.5139620304107666,
+      "eval_runtime": 75.2979,
+      "eval_samples_per_second": 1246.237,
+      "eval_steps_per_second": 4.874,
+      "step": 70000
+    },
+    {
+      "epoch": 11.91326530612245,
+      "grad_norm": 0.49928614497184753,
+      "learning_rate": 0.00010543503480278423,
+      "loss": 1.3525,
+      "step": 70050
+    },
+    {
+      "epoch": 11.921768707482993,
+      "grad_norm": 0.4152166247367859,
+      "learning_rate": 0.0001051508120649652,
+      "loss": 1.3526,
+      "step": 70100
+    },
+    {
+      "epoch": 11.930272108843537,
+      "grad_norm": 0.42939451336860657,
+      "learning_rate": 0.00010486078886310905,
+      "loss": 1.352,
+      "step": 70150
+    },
+    {
+      "epoch": 11.938775510204081,
+      "grad_norm": 0.40568697452545166,
+      "learning_rate": 0.0001045707656612529,
+      "loss": 1.3493,
+      "step": 70200
+    },
+    {
+      "epoch": 11.947278911564625,
+      "grad_norm": 0.43555957078933716,
+      "learning_rate": 0.00010428074245939675,
+      "loss": 1.3425,
+      "step": 70250
+    },
+    {
+      "epoch": 11.95578231292517,
+      "grad_norm": 0.4200745224952698,
+      "learning_rate": 0.00010399071925754061,
+      "loss": 1.3477,
+      "step": 70300
+    },
+    {
+      "epoch": 11.964285714285714,
+      "grad_norm": 0.43165090680122375,
+      "learning_rate": 0.00010370069605568446,
+      "loss": 1.3522,
+      "step": 70350
+    },
+    {
+      "epoch": 11.972789115646258,
+      "grad_norm": 0.41488534212112427,
+      "learning_rate": 0.0001034106728538283,
+      "loss": 1.3509,
+      "step": 70400
+    },
+    {
+      "epoch": 11.981292517006803,
+      "grad_norm": 0.4268278479576111,
+      "learning_rate": 0.00010312064965197215,
+      "loss": 1.3567,
+      "step": 70450
+    },
+    {
+      "epoch": 11.989795918367347,
+      "grad_norm": 0.4317072927951813,
+      "learning_rate": 0.00010283062645011601,
+      "loss": 1.3441,
+      "step": 70500
+    },
+    {
+      "epoch": 11.998299319727892,
+      "grad_norm": 0.39999693632125854,
+      "learning_rate": 0.00010254060324825986,
+      "loss": 1.3486,
+      "step": 70550
+    },
+    {
+      "epoch": 12.006802721088436,
+      "grad_norm": 0.39868417382240295,
+      "learning_rate": 0.00010225058004640371,
+      "loss": 1.3422,
+      "step": 70600
+    },
+    {
+      "epoch": 12.01530612244898,
+      "grad_norm": 0.4184889793395996,
+      "learning_rate": 0.00010196055684454757,
+      "loss": 1.342,
+      "step": 70650
+    },
+    {
+      "epoch": 12.023809523809524,
+      "grad_norm": 0.4223806858062744,
+      "learning_rate": 0.00010167053364269142,
+      "loss": 1.339,
+      "step": 70700
+    },
+    {
+      "epoch": 12.032312925170068,
+      "grad_norm": 0.413501113653183,
+      "learning_rate": 0.00010138051044083527,
+      "loss": 1.3458,
+      "step": 70750
+    },
+    {
+      "epoch": 12.040816326530612,
+      "grad_norm": 0.41243672370910645,
+      "learning_rate": 0.00010109048723897913,
+      "loss": 1.3497,
+      "step": 70800
+    },
+    {
+      "epoch": 12.049319727891156,
+      "grad_norm": 0.4220627546310425,
+      "learning_rate": 0.00010080046403712296,
+      "loss": 1.3465,
+      "step": 70850
+    },
+    {
+      "epoch": 12.0578231292517,
+      "grad_norm": 0.4198216199874878,
+      "learning_rate": 0.00010051044083526682,
+      "loss": 1.3477,
+      "step": 70900
+    },
+    {
+      "epoch": 12.066326530612244,
+      "grad_norm": 0.4057486951351166,
+      "learning_rate": 0.00010022041763341068,
+      "loss": 1.3406,
+      "step": 70950
+    },
+    {
+      "epoch": 12.07482993197279,
+      "grad_norm": 0.4336649179458618,
+      "learning_rate": 9.993039443155452e-05,
+      "loss": 1.342,
+      "step": 71000
+    },
+    {
+      "epoch": 12.07482993197279,
+      "eval_loss": 1.5144281387329102,
+      "eval_runtime": 75.2698,
+      "eval_samples_per_second": 1246.702,
+      "eval_steps_per_second": 4.876,
+      "step": 71000
+    },
+    {
+      "epoch": 12.083333333333334,
+      "grad_norm": 0.4182901680469513,
+      "learning_rate": 9.964037122969838e-05,
+      "loss": 1.3432,
+      "step": 71050
+    },
+    {
+      "epoch": 12.091836734693878,
+      "grad_norm": 0.39864903688430786,
+      "learning_rate": 9.935034802784224e-05,
+      "loss": 1.3367,
+      "step": 71100
+    },
+    {
+      "epoch": 12.100340136054422,
+      "grad_norm": 0.40879112482070923,
+      "learning_rate": 9.906032482598607e-05,
+      "loss": 1.3431,
+      "step": 71150
+    },
+    {
+      "epoch": 12.108843537414966,
+      "grad_norm": 0.412240594625473,
+      "learning_rate": 9.877030162412993e-05,
+      "loss": 1.3406,
+      "step": 71200
+    },
+    {
+      "epoch": 12.11734693877551,
+      "grad_norm": 0.4245721995830536,
+      "learning_rate": 9.848027842227378e-05,
+      "loss": 1.3453,
+      "step": 71250
+    },
+    {
+      "epoch": 12.125850340136054,
+      "grad_norm": 0.40284958481788635,
+      "learning_rate": 9.819025522041763e-05,
+      "loss": 1.3383,
+      "step": 71300
+    },
+    {
+      "epoch": 12.134353741496598,
+      "grad_norm": 0.57562255859375,
+      "learning_rate": 9.790023201856149e-05,
+      "loss": 1.3417,
+      "step": 71350
+    },
+    {
+      "epoch": 12.142857142857142,
+      "grad_norm": 0.4117988348007202,
+      "learning_rate": 9.761020881670534e-05,
+      "loss": 1.3371,
+      "step": 71400
+    },
+    {
+      "epoch": 12.151360544217686,
+      "grad_norm": 0.4112182557582855,
+      "learning_rate": 9.732018561484918e-05,
+      "loss": 1.3447,
+      "step": 71450
+    },
+    {
+      "epoch": 12.15986394557823,
+      "grad_norm": 0.41666939854621887,
+      "learning_rate": 9.703016241299305e-05,
+      "loss": 1.3408,
+      "step": 71500
+    },
+    {
+      "epoch": 12.168367346938776,
+      "grad_norm": 0.4010598063468933,
+      "learning_rate": 9.674013921113689e-05,
+      "loss": 1.3443,
+      "step": 71550
+    },
+    {
+      "epoch": 12.17687074829932,
+      "grad_norm": 0.41796889901161194,
+      "learning_rate": 9.645011600928074e-05,
+      "loss": 1.3343,
+      "step": 71600
+    },
+    {
+      "epoch": 12.185374149659864,
+      "grad_norm": 0.42123886942863464,
+      "learning_rate": 9.616009280742459e-05,
+      "loss": 1.3401,
+      "step": 71650
+    },
+    {
+      "epoch": 12.193877551020408,
+      "grad_norm": 0.40508341789245605,
+      "learning_rate": 9.587006960556845e-05,
+      "loss": 1.3437,
+      "step": 71700
+    },
+    {
+      "epoch": 12.202380952380953,
+      "grad_norm": 0.44286808371543884,
+      "learning_rate": 9.558004640371231e-05,
+      "loss": 1.3419,
+      "step": 71750
+    },
+    {
+      "epoch": 12.210884353741497,
+      "grad_norm": 0.40285545587539673,
+      "learning_rate": 9.529002320185614e-05,
+      "loss": 1.333,
+      "step": 71800
+    },
+    {
+      "epoch": 12.21938775510204,
+      "grad_norm": 0.39993083477020264,
+      "learning_rate": 9.5e-05,
+      "loss": 1.3402,
+      "step": 71850
+    },
+    {
+      "epoch": 12.227891156462585,
+      "grad_norm": 0.4155488610267639,
+      "learning_rate": 9.470997679814387e-05,
+      "loss": 1.3378,
+      "step": 71900
+    },
+    {
+      "epoch": 12.236394557823129,
+      "grad_norm": 0.5050736665725708,
+      "learning_rate": 9.44199535962877e-05,
+      "loss": 1.3354,
+      "step": 71950
+    },
+    {
+      "epoch": 12.244897959183673,
+      "grad_norm": 0.4185927212238312,
+      "learning_rate": 9.412993039443156e-05,
+      "loss": 1.3357,
+      "step": 72000
+    },
+    {
+      "epoch": 12.244897959183673,
+      "eval_loss": 1.5064618587493896,
+      "eval_runtime": 75.3162,
+      "eval_samples_per_second": 1245.933,
+      "eval_steps_per_second": 4.873,
+      "step": 72000
+    },
+    {
+      "epoch": 12.253401360544217,
+      "grad_norm": 0.41753119230270386,
+      "learning_rate": 9.383990719257541e-05,
+      "loss": 1.3352,
+      "step": 72050
+    },
+    {
+      "epoch": 12.261904761904763,
+      "grad_norm": 0.4204673171043396,
+      "learning_rate": 9.354988399071926e-05,
+      "loss": 1.3436,
+      "step": 72100
+    },
+    {
+      "epoch": 12.270408163265307,
+      "grad_norm": 0.4101797044277191,
+      "learning_rate": 9.325986078886312e-05,
+      "loss": 1.3385,
+      "step": 72150
+    },
+    {
+      "epoch": 12.27891156462585,
+      "grad_norm": 0.4880974292755127,
+      "learning_rate": 9.296983758700696e-05,
+      "loss": 1.3322,
+      "step": 72200
+    },
+    {
+      "epoch": 12.287414965986395,
+      "grad_norm": 0.415211945772171,
+      "learning_rate": 9.267981438515081e-05,
+      "loss": 1.3409,
+      "step": 72250
+    },
+    {
+      "epoch": 12.295918367346939,
+      "grad_norm": 0.41376402974128723,
+      "learning_rate": 9.238979118329467e-05,
+      "loss": 1.3392,
+      "step": 72300
+    },
+    {
+      "epoch": 12.304421768707483,
+      "grad_norm": 0.43394094705581665,
+      "learning_rate": 9.209976798143852e-05,
+      "loss": 1.3345,
+      "step": 72350
+    },
+    {
+      "epoch": 12.312925170068027,
+      "grad_norm": 0.43497174978256226,
+      "learning_rate": 9.180974477958237e-05,
+      "loss": 1.3381,
+      "step": 72400
+    },
+    {
+      "epoch": 12.321428571428571,
+      "grad_norm": 0.4110974371433258,
+      "learning_rate": 9.151972157772622e-05,
+      "loss": 1.3375,
+      "step": 72450
+    },
+    {
+      "epoch": 12.329931972789115,
+      "grad_norm": 0.4111982583999634,
+      "learning_rate": 9.12354988399072e-05,
+      "loss": 1.3352,
+      "step": 72500
+    },
+    {
+      "epoch": 12.33843537414966,
+      "grad_norm": 0.4248039126396179,
+      "learning_rate": 9.094547563805104e-05,
+      "loss": 1.3353,
+      "step": 72550
+    },
+    {
+      "epoch": 12.346938775510203,
+      "grad_norm": 0.4378061294555664,
+      "learning_rate": 9.06554524361949e-05,
+      "loss": 1.3397,
+      "step": 72600
+    },
+    {
+      "epoch": 12.35544217687075,
+      "grad_norm": 0.41765543818473816,
+      "learning_rate": 9.036542923433875e-05,
+      "loss": 1.3372,
+      "step": 72650
+    },
+    {
+      "epoch": 12.363945578231293,
+      "grad_norm": 0.4655171036720276,
+      "learning_rate": 9.00754060324826e-05,
+      "loss": 1.337,
+      "step": 72700
+    },
+    {
+      "epoch": 12.372448979591837,
+      "grad_norm": 0.4178127348423004,
+      "learning_rate": 8.978538283062646e-05,
+      "loss": 1.3362,
+      "step": 72750
+    },
+    {
+      "epoch": 12.380952380952381,
+      "grad_norm": 0.4332452118396759,
+      "learning_rate": 8.94953596287703e-05,
+      "loss": 1.3379,
+      "step": 72800
+    },
+    {
+      "epoch": 12.389455782312925,
+      "grad_norm": 0.44955676794052124,
+      "learning_rate": 8.920533642691416e-05,
+      "loss": 1.3384,
+      "step": 72850
+    },
+    {
+      "epoch": 12.39795918367347,
+      "grad_norm": 0.42252740263938904,
+      "learning_rate": 8.891531322505802e-05,
+      "loss": 1.3398,
+      "step": 72900
+    },
+    {
+      "epoch": 12.406462585034014,
+      "grad_norm": 0.3996742069721222,
+      "learning_rate": 8.862529002320185e-05,
+      "loss": 1.3294,
+      "step": 72950
+    },
+    {
+      "epoch": 12.414965986394558,
+      "grad_norm": 0.5868241786956787,
+      "learning_rate": 8.833526682134571e-05,
+      "loss": 1.3347,
+      "step": 73000
+    },
+    {
+      "epoch": 12.414965986394558,
+      "eval_loss": 1.4998681545257568,
+      "eval_runtime": 75.3074,
+      "eval_samples_per_second": 1246.079,
+      "eval_steps_per_second": 4.873,
+      "step": 73000
+    },
+    {
+      "epoch": 12.423469387755102,
+      "grad_norm": 0.4118203818798065,
+      "learning_rate": 8.804524361948957e-05,
+      "loss": 1.3422,
+      "step": 73050
+    },
+    {
+      "epoch": 12.431972789115646,
+      "grad_norm": 0.5292280316352844,
+      "learning_rate": 8.775522041763341e-05,
+      "loss": 1.3353,
+      "step": 73100
+    },
+    {
+      "epoch": 12.44047619047619,
+      "grad_norm": 0.42021283507347107,
+      "learning_rate": 8.747099767981439e-05,
+      "loss": 1.3379,
+      "step": 73150
+    },
+    {
+      "epoch": 12.448979591836734,
+      "grad_norm": 0.42471328377723694,
+      "learning_rate": 8.718097447795824e-05,
+      "loss": 1.3377,
+      "step": 73200
+    },
+    {
+      "epoch": 12.45748299319728,
+      "grad_norm": 0.4189905524253845,
+      "learning_rate": 8.68909512761021e-05,
+      "loss": 1.3338,
+      "step": 73250
+    },
+    {
+      "epoch": 12.465986394557824,
+      "grad_norm": 0.4137709438800812,
+      "learning_rate": 8.660092807424594e-05,
+      "loss": 1.3341,
+      "step": 73300
+    },
+    {
+      "epoch": 12.474489795918368,
+      "grad_norm": 0.41046592593193054,
+      "learning_rate": 8.631090487238979e-05,
+      "loss": 1.3325,
+      "step": 73350
+    },
+    {
+      "epoch": 12.482993197278912,
+      "grad_norm": 0.40613409876823425,
+      "learning_rate": 8.602088167053364e-05,
+      "loss": 1.334,
+      "step": 73400
+    },
+    {
+      "epoch": 12.491496598639456,
+      "grad_norm": 0.4053729474544525,
+      "learning_rate": 8.57308584686775e-05,
+      "loss": 1.3297,
+      "step": 73450
+    },
+    {
+      "epoch": 12.5,
+      "grad_norm": 0.4099692404270172,
+      "learning_rate": 8.544083526682135e-05,
+      "loss": 1.3303,
+      "step": 73500
+    },
+    {
+      "epoch": 12.508503401360544,
+      "grad_norm": 0.40552228689193726,
+      "learning_rate": 8.51508120649652e-05,
+      "loss": 1.3286,
+      "step": 73550
+    },
+    {
+      "epoch": 12.517006802721088,
+      "grad_norm": 0.41349565982818604,
+      "learning_rate": 8.486078886310906e-05,
+      "loss": 1.327,
+      "step": 73600
+    },
+    {
+      "epoch": 12.525510204081632,
+      "grad_norm": 0.40340515971183777,
+      "learning_rate": 8.45707656612529e-05,
+      "loss": 1.3282,
+      "step": 73650
+    },
+    {
+      "epoch": 12.534013605442176,
+      "grad_norm": 0.4043871760368347,
+      "learning_rate": 8.428074245939675e-05,
+      "loss": 1.3316,
+      "step": 73700
+    },
+    {
+      "epoch": 12.54251700680272,
+      "grad_norm": 0.41814517974853516,
+      "learning_rate": 8.399071925754061e-05,
+      "loss": 1.3278,
+      "step": 73750
+    },
+    {
+      "epoch": 12.551020408163264,
+      "grad_norm": 0.41396060585975647,
+      "learning_rate": 8.370069605568445e-05,
+      "loss": 1.3283,
+      "step": 73800
+    },
+    {
+      "epoch": 12.55952380952381,
+      "grad_norm": 0.4073675572872162,
+      "learning_rate": 8.341067285382831e-05,
+      "loss": 1.3274,
+      "step": 73850
+    },
+    {
+      "epoch": 12.568027210884354,
+      "grad_norm": 0.40986111760139465,
+      "learning_rate": 8.312064965197217e-05,
+      "loss": 1.3252,
+      "step": 73900
+    },
+    {
+      "epoch": 12.576530612244898,
+      "grad_norm": 0.3996681869029999,
+      "learning_rate": 8.2830626450116e-05,
+      "loss": 1.3308,
+      "step": 73950
+    },
+    {
+      "epoch": 12.585034013605442,
+      "grad_norm": 0.4171459376811981,
+      "learning_rate": 8.254060324825986e-05,
+      "loss": 1.3263,
+      "step": 74000
+    },
+    {
+      "epoch": 12.585034013605442,
+      "eval_loss": 1.4951106309890747,
+      "eval_runtime": 75.2691,
+      "eval_samples_per_second": 1246.714,
+      "eval_steps_per_second": 4.876,
+      "step": 74000
+    },
+    {
+      "epoch": 12.593537414965986,
+      "grad_norm": 0.4198111593723297,
+      "learning_rate": 8.225058004640372e-05,
+      "loss": 1.3306,
+      "step": 74050
+    },
+    {
+      "epoch": 12.60204081632653,
+      "grad_norm": 0.4122917056083679,
+      "learning_rate": 8.196055684454756e-05,
+      "loss": 1.3325,
+      "step": 74100
+    },
+    {
+      "epoch": 12.610544217687075,
+      "grad_norm": 0.43273040652275085,
+      "learning_rate": 8.167053364269142e-05,
+      "loss": 1.3385,
+      "step": 74150
+    },
+    {
+      "epoch": 12.619047619047619,
+      "grad_norm": 0.4447864294052124,
+      "learning_rate": 8.138051044083527e-05,
+      "loss": 1.3386,
+      "step": 74200
+    },
+    {
+      "epoch": 12.627551020408163,
+      "grad_norm": 0.41947802901268005,
+      "learning_rate": 8.109048723897911e-05,
+      "loss": 1.3313,
+      "step": 74250
+    },
+    {
+      "epoch": 12.636054421768707,
+      "grad_norm": 0.41405126452445984,
+      "learning_rate": 8.080046403712298e-05,
+      "loss": 1.3331,
+      "step": 74300
+    },
+    {
+      "epoch": 12.64455782312925,
+      "grad_norm": 0.407975971698761,
+      "learning_rate": 8.051044083526682e-05,
+      "loss": 1.3243,
+      "step": 74350
+    },
+    {
+      "epoch": 12.653061224489797,
+      "grad_norm": 0.3966875374317169,
+      "learning_rate": 8.022041763341067e-05,
+      "loss": 1.3276,
+      "step": 74400
+    },
+    {
+      "epoch": 12.66156462585034,
+      "grad_norm": 0.43828269839286804,
+      "learning_rate": 7.993619489559165e-05,
+      "loss": 1.335,
+      "step": 74450
+    },
+    {
+      "epoch": 12.670068027210885,
+      "grad_norm": 0.40009811520576477,
+      "learning_rate": 7.96461716937355e-05,
+      "loss": 1.3217,
+      "step": 74500
+    },
+    {
+      "epoch": 12.678571428571429,
+      "grad_norm": 0.40851446986198425,
+      "learning_rate": 7.935614849187935e-05,
+      "loss": 1.3338,
+      "step": 74550
+    },
+    {
+      "epoch": 12.687074829931973,
+      "grad_norm": 0.4100090265274048,
+      "learning_rate": 7.906612529002321e-05,
+      "loss": 1.334,
+      "step": 74600
+    },
+    {
+      "epoch": 12.695578231292517,
+      "grad_norm": 0.41323432326316833,
+      "learning_rate": 7.877610208816705e-05,
+      "loss": 1.3279,
+      "step": 74650
+    },
+    {
+      "epoch": 12.704081632653061,
+      "grad_norm": 0.4219471514225006,
+      "learning_rate": 7.84860788863109e-05,
+      "loss": 1.3265,
+      "step": 74700
+    },
+    {
+      "epoch": 12.712585034013605,
+      "grad_norm": 0.4054030478000641,
+      "learning_rate": 7.819605568445476e-05,
+      "loss": 1.3275,
+      "step": 74750
+    },
+    {
+      "epoch": 12.72108843537415,
+      "grad_norm": 0.4156668782234192,
+      "learning_rate": 7.790603248259861e-05,
+      "loss": 1.3246,
+      "step": 74800
+    },
+    {
+      "epoch": 12.729591836734693,
+      "grad_norm": 0.4128476083278656,
+      "learning_rate": 7.761600928074246e-05,
+      "loss": 1.325,
+      "step": 74850
+    },
+    {
+      "epoch": 12.738095238095237,
+      "grad_norm": 0.4204627275466919,
+      "learning_rate": 7.732598607888632e-05,
+      "loss": 1.3303,
+      "step": 74900
+    },
+    {
+      "epoch": 12.746598639455783,
+      "grad_norm": 0.4113794267177582,
+      "learning_rate": 7.703596287703015e-05,
+      "loss": 1.3297,
+      "step": 74950
+    },
+    {
+      "epoch": 12.755102040816327,
+      "grad_norm": 0.4143586754798889,
+      "learning_rate": 7.674593967517401e-05,
+      "loss": 1.323,
+      "step": 75000
+    },
+    {
+      "epoch": 12.755102040816327,
+      "eval_loss": 1.4961031675338745,
+      "eval_runtime": 75.2707,
+      "eval_samples_per_second": 1246.687,
+      "eval_steps_per_second": 4.876,
+      "step": 75000
+    },
+    {
+      "epoch": 12.763605442176871,
+      "grad_norm": 0.4443499743938446,
+      "learning_rate": 7.645591647331788e-05,
+      "loss": 1.3285,
+      "step": 75050
+    },
+    {
+      "epoch": 12.772108843537415,
+      "grad_norm": 0.4180935025215149,
+      "learning_rate": 7.616589327146171e-05,
+      "loss": 1.3299,
+      "step": 75100
+    },
+    {
+      "epoch": 12.78061224489796,
+      "grad_norm": 0.41669797897338867,
+      "learning_rate": 7.587587006960557e-05,
+      "loss": 1.3314,
+      "step": 75150
+    },
+    {
+      "epoch": 12.789115646258503,
+      "grad_norm": 0.40955501794815063,
+      "learning_rate": 7.558584686774943e-05,
+      "loss": 1.3285,
+      "step": 75200
+    },
+    {
+      "epoch": 12.797619047619047,
+      "grad_norm": 0.4140739142894745,
+      "learning_rate": 7.529582366589327e-05,
+      "loss": 1.331,
+      "step": 75250
+    },
+    {
+      "epoch": 12.806122448979592,
+      "grad_norm": 0.4092477858066559,
+      "learning_rate": 7.500580046403713e-05,
+      "loss": 1.3268,
+      "step": 75300
+    },
+    {
+      "epoch": 12.814625850340136,
+      "grad_norm": 0.412955105304718,
+      "learning_rate": 7.471577726218097e-05,
+      "loss": 1.324,
+      "step": 75350
+    },
+    {
+      "epoch": 12.82312925170068,
+      "grad_norm": 0.4366496205329895,
+      "learning_rate": 7.442575406032482e-05,
+      "loss": 1.3257,
+      "step": 75400
+    },
+    {
+      "epoch": 12.831632653061224,
+      "grad_norm": 0.4090063273906708,
+      "learning_rate": 7.413573085846868e-05,
+      "loss": 1.3261,
+      "step": 75450
+    },
+    {
+      "epoch": 12.84013605442177,
+      "grad_norm": 0.41486817598342896,
+      "learning_rate": 7.384570765661253e-05,
+      "loss": 1.3224,
+      "step": 75500
+    },
+    {
+      "epoch": 12.848639455782314,
+      "grad_norm": 0.40485501289367676,
+      "learning_rate": 7.355568445475638e-05,
+      "loss": 1.3285,
+      "step": 75550
+    },
+    {
+      "epoch": 12.857142857142858,
+      "grad_norm": 0.43988388776779175,
+      "learning_rate": 7.326566125290024e-05,
+      "loss": 1.3245,
+      "step": 75600
+    },
+    {
+      "epoch": 12.865646258503402,
+      "grad_norm": 0.43231526017189026,
+      "learning_rate": 7.297563805104409e-05,
+      "loss": 1.3266,
+      "step": 75650
+    },
+    {
+      "epoch": 12.874149659863946,
+      "grad_norm": 0.40446582436561584,
+      "learning_rate": 7.268561484918793e-05,
+      "loss": 1.3248,
+      "step": 75700
+    },
+    {
+      "epoch": 12.88265306122449,
+      "grad_norm": 0.40424802899360657,
+      "learning_rate": 7.239559164733178e-05,
+      "loss": 1.3187,
+      "step": 75750
+    },
+    {
+      "epoch": 12.891156462585034,
+      "grad_norm": 0.410151869058609,
+      "learning_rate": 7.210556844547564e-05,
+      "loss": 1.3212,
+      "step": 75800
+    },
+    {
+      "epoch": 12.899659863945578,
+      "grad_norm": 0.41777828335762024,
+      "learning_rate": 7.181554524361949e-05,
+      "loss": 1.3204,
+      "step": 75850
+    },
+    {
+      "epoch": 12.908163265306122,
+      "grad_norm": 0.4275606870651245,
+      "learning_rate": 7.152552204176334e-05,
+      "loss": 1.3239,
+      "step": 75900
+    },
+    {
+      "epoch": 12.916666666666666,
+      "grad_norm": 0.4166664481163025,
+      "learning_rate": 7.12354988399072e-05,
+      "loss": 1.3174,
+      "step": 75950
+    },
+    {
+      "epoch": 12.92517006802721,
+      "grad_norm": 0.3995276391506195,
+      "learning_rate": 7.094547563805105e-05,
+      "loss": 1.3231,
+      "step": 76000
+    },
+    {
+      "epoch": 12.92517006802721,
+      "eval_loss": 1.490061640739441,
+      "eval_runtime": 75.2857,
+      "eval_samples_per_second": 1246.439,
+      "eval_steps_per_second": 4.875,
+      "step": 76000
+    },
+    {
+      "epoch": 12.933673469387756,
+      "grad_norm": 0.4179072082042694,
+      "learning_rate": 7.06554524361949e-05,
+      "loss": 1.3217,
+      "step": 76050
+    },
+    {
+      "epoch": 12.9421768707483,
+      "grad_norm": 0.43183988332748413,
+      "learning_rate": 7.036542923433875e-05,
+      "loss": 1.3218,
+      "step": 76100
+    },
+    {
+      "epoch": 12.950680272108844,
+      "grad_norm": 0.4066503345966339,
+      "learning_rate": 7.007540603248259e-05,
+      "loss": 1.326,
+      "step": 76150
+    },
+    {
+      "epoch": 12.959183673469388,
+      "grad_norm": 0.4215805232524872,
+      "learning_rate": 6.978538283062645e-05,
+      "loss": 1.325,
+      "step": 76200
+    },
+    {
+      "epoch": 12.967687074829932,
+      "grad_norm": 0.4248071312904358,
+      "learning_rate": 6.949535962877031e-05,
+      "loss": 1.3179,
+      "step": 76250
+    },
+    {
+      "epoch": 12.976190476190476,
+      "grad_norm": 0.42110559344291687,
+      "learning_rate": 6.920533642691414e-05,
+      "loss": 1.3292,
+      "step": 76300
+    },
+    {
+      "epoch": 12.98469387755102,
+      "grad_norm": 0.4082868695259094,
+      "learning_rate": 6.8915313225058e-05,
+      "loss": 1.3254,
+      "step": 76350
+    },
+    {
+      "epoch": 12.993197278911564,
+      "grad_norm": 0.4124908447265625,
+      "learning_rate": 6.862529002320187e-05,
+      "loss": 1.3227,
+      "step": 76400
+    },
+    {
+      "epoch": 13.001700680272108,
+      "grad_norm": 0.5426220297813416,
+      "learning_rate": 6.83352668213457e-05,
+      "loss": 1.3198,
+      "step": 76450
+    },
+    {
+      "epoch": 13.010204081632653,
+      "grad_norm": 0.4062071442604065,
+      "learning_rate": 6.804524361948956e-05,
+      "loss": 1.3183,
+      "step": 76500
+    },
+    {
+      "epoch": 13.018707482993197,
+      "grad_norm": 0.40301281213760376,
+      "learning_rate": 6.775522041763341e-05,
+      "loss": 1.3232,
+      "step": 76550
+    },
+    {
+      "epoch": 13.02721088435374,
+      "grad_norm": 0.44012153148651123,
+      "learning_rate": 6.746519721577726e-05,
+      "loss": 1.318,
+      "step": 76600
+    },
+    {
+      "epoch": 13.035714285714286,
+      "grad_norm": 0.41154050827026367,
+      "learning_rate": 6.717517401392112e-05,
+      "loss": 1.3176,
+      "step": 76650
+    },
+    {
+      "epoch": 13.04421768707483,
+      "grad_norm": 0.42961668968200684,
+      "learning_rate": 6.688515081206497e-05,
+      "loss": 1.3188,
+      "step": 76700
+    },
+    {
+      "epoch": 13.052721088435375,
+      "grad_norm": 0.41780275106430054,
+      "learning_rate": 6.659512761020881e-05,
+      "loss": 1.3191,
+      "step": 76750
+    },
+    {
+      "epoch": 13.061224489795919,
+      "grad_norm": 0.4239339828491211,
+      "learning_rate": 6.630510440835267e-05,
+      "loss": 1.3225,
+      "step": 76800
+    },
+    {
+      "epoch": 13.069727891156463,
+      "grad_norm": 0.40867921710014343,
+      "learning_rate": 6.601508120649652e-05,
+      "loss": 1.3254,
+      "step": 76850
+    },
+    {
+      "epoch": 13.078231292517007,
+      "grad_norm": 0.4113767743110657,
+      "learning_rate": 6.572505800464038e-05,
+      "loss": 1.3151,
+      "step": 76900
+    },
+    {
+      "epoch": 13.08673469387755,
+      "grad_norm": 0.4053804874420166,
+      "learning_rate": 6.543503480278422e-05,
+      "loss": 1.3196,
+      "step": 76950
+    },
+    {
+      "epoch": 13.095238095238095,
+      "grad_norm": 0.40238648653030396,
+      "learning_rate": 6.514501160092808e-05,
+      "loss": 1.3119,
+      "step": 77000
+    },
+    {
+      "epoch": 13.095238095238095,
+      "eval_loss": 1.486020803451538,
+      "eval_runtime": 75.2657,
+      "eval_samples_per_second": 1246.769,
+      "eval_steps_per_second": 4.876,
+      "step": 77000
+    },
+    {
+      "epoch": 13.103741496598639,
+      "grad_norm": 0.4077170193195343,
+      "learning_rate": 6.485498839907194e-05,
+      "loss": 1.3185,
+      "step": 77050
+    },
+    {
+      "epoch": 13.112244897959183,
+      "grad_norm": 0.40546542406082153,
+      "learning_rate": 6.456496519721577e-05,
+      "loss": 1.3221,
+      "step": 77100
+    },
+    {
+      "epoch": 13.120748299319727,
+      "grad_norm": 0.4073767364025116,
+      "learning_rate": 6.427494199535963e-05,
+      "loss": 1.3168,
+      "step": 77150
+    },
+    {
+      "epoch": 13.129251700680273,
+      "grad_norm": 0.4032026529312134,
+      "learning_rate": 6.39849187935035e-05,
+      "loss": 1.3162,
+      "step": 77200
+    },
+    {
+      "epoch": 13.137755102040817,
+      "grad_norm": 0.4208144545555115,
+      "learning_rate": 6.369489559164733e-05,
+      "loss": 1.3218,
+      "step": 77250
+    },
+    {
+      "epoch": 13.146258503401361,
+      "grad_norm": 0.4274178743362427,
+      "learning_rate": 6.340487238979119e-05,
+      "loss": 1.3165,
+      "step": 77300
+    },
+    {
+      "epoch": 13.154761904761905,
+      "grad_norm": 0.41824012994766235,
+      "learning_rate": 6.311484918793504e-05,
+      "loss": 1.3185,
+      "step": 77350
+    },
+    {
+      "epoch": 13.16326530612245,
+      "grad_norm": 0.395487517118454,
+      "learning_rate": 6.282482598607888e-05,
+      "loss": 1.3115,
+      "step": 77400
+    },
+    {
+      "epoch": 13.171768707482993,
+      "grad_norm": 0.4073810577392578,
+      "learning_rate": 6.253480278422275e-05,
+      "loss": 1.3156,
+      "step": 77450
+    },
+    {
+      "epoch": 13.180272108843537,
+      "grad_norm": 0.4188859760761261,
+      "learning_rate": 6.224477958236659e-05,
+      "loss": 1.3164,
+      "step": 77500
+    },
+    {
+      "epoch": 13.188775510204081,
+      "grad_norm": 0.4151723086833954,
+      "learning_rate": 6.195475638051044e-05,
+      "loss": 1.316,
+      "step": 77550
+    },
+    {
+      "epoch": 13.197278911564625,
+      "grad_norm": 0.42075613141059875,
+      "learning_rate": 6.166473317865429e-05,
+      "loss": 1.3136,
+      "step": 77600
+    },
+    {
+      "epoch": 13.20578231292517,
+      "grad_norm": 0.41158735752105713,
+      "learning_rate": 6.137470997679815e-05,
+      "loss": 1.3157,
+      "step": 77650
+    },
+    {
+      "epoch": 13.214285714285714,
+      "grad_norm": 0.4071219563484192,
+      "learning_rate": 6.1084686774942e-05,
+      "loss": 1.3167,
+      "step": 77700
+    },
+    {
+      "epoch": 13.22278911564626,
+      "grad_norm": 0.430393248796463,
+      "learning_rate": 6.079466357308585e-05,
+      "loss": 1.3142,
+      "step": 77750
+    },
+    {
+      "epoch": 13.231292517006803,
+      "grad_norm": 0.402025043964386,
+      "learning_rate": 6.05046403712297e-05,
+      "loss": 1.3149,
+      "step": 77800
+    },
+    {
+      "epoch": 13.239795918367347,
+      "grad_norm": 0.39697858691215515,
+      "learning_rate": 6.0214617169373546e-05,
+      "loss": 1.3163,
+      "step": 77850
+    },
+    {
+      "epoch": 13.248299319727892,
+      "grad_norm": 0.45332589745521545,
+      "learning_rate": 5.992459396751741e-05,
+      "loss": 1.31,
+      "step": 77900
+    },
+    {
+      "epoch": 13.256802721088436,
+      "grad_norm": 0.42384395003318787,
+      "learning_rate": 5.9634570765661255e-05,
+      "loss": 1.3102,
+      "step": 77950
+    },
+    {
+      "epoch": 13.26530612244898,
+      "grad_norm": 0.42941874265670776,
+      "learning_rate": 5.934454756380511e-05,
+      "loss": 1.3184,
+      "step": 78000
+    },
+    {
+      "epoch": 13.26530612244898,
+      "eval_loss": 1.4811251163482666,
+      "eval_runtime": 75.2963,
+      "eval_samples_per_second": 1246.264,
+      "eval_steps_per_second": 4.874,
+      "step": 78000
+    },
+    {
+      "epoch": 13.273809523809524,
+      "grad_norm": 0.41413602232933044,
+      "learning_rate": 5.9054524361948956e-05,
+      "loss": 1.3109,
+      "step": 78050
+    },
+    {
+      "epoch": 13.282312925170068,
+      "grad_norm": 0.4183179438114166,
+      "learning_rate": 5.876450116009281e-05,
+      "loss": 1.3129,
+      "step": 78100
+    },
+    {
+      "epoch": 13.290816326530612,
+      "grad_norm": 0.41364216804504395,
+      "learning_rate": 5.8474477958236665e-05,
+      "loss": 1.3132,
+      "step": 78150
+    },
+    {
+      "epoch": 13.299319727891156,
+      "grad_norm": 0.41207000613212585,
+      "learning_rate": 5.818445475638051e-05,
+      "loss": 1.3238,
+      "step": 78200
+    },
+    {
+      "epoch": 13.3078231292517,
+      "grad_norm": 0.42450812458992004,
+      "learning_rate": 5.789443155452436e-05,
+      "loss": 1.3113,
+      "step": 78250
+    },
+    {
+      "epoch": 13.316326530612244,
+      "grad_norm": 0.4294716715812683,
+      "learning_rate": 5.760440835266822e-05,
+      "loss": 1.3195,
+      "step": 78300
+    },
+    {
+      "epoch": 13.32482993197279,
+      "grad_norm": 0.4036255478858948,
+      "learning_rate": 5.731438515081207e-05,
+      "loss": 1.3148,
+      "step": 78350
+    },
+    {
+      "epoch": 13.333333333333334,
+      "grad_norm": 0.3960818946361542,
+      "learning_rate": 5.7024361948955916e-05,
+      "loss": 1.3098,
+      "step": 78400
+    },
+    {
+      "epoch": 13.341836734693878,
+      "grad_norm": 0.40689873695373535,
+      "learning_rate": 5.673433874709977e-05,
+      "loss": 1.3136,
+      "step": 78450
+    },
+    {
+      "epoch": 13.350340136054422,
+      "grad_norm": 0.4078476130962372,
+      "learning_rate": 5.6444315545243625e-05,
+      "loss": 1.3111,
+      "step": 78500
+    },
+    {
+      "epoch": 13.358843537414966,
+      "grad_norm": 0.4066517651081085,
+      "learning_rate": 5.615429234338747e-05,
+      "loss": 1.3117,
+      "step": 78550
+    },
+    {
+      "epoch": 13.36734693877551,
+      "grad_norm": 0.4128108024597168,
+      "learning_rate": 5.5864269141531326e-05,
+      "loss": 1.3108,
+      "step": 78600
+    },
+    {
+      "epoch": 13.375850340136054,
+      "grad_norm": 0.40743228793144226,
+      "learning_rate": 5.5574245939675174e-05,
+      "loss": 1.3099,
+      "step": 78650
+    },
+    {
+      "epoch": 13.384353741496598,
+      "grad_norm": 0.43066033720970154,
+      "learning_rate": 5.528422273781902e-05,
+      "loss": 1.312,
+      "step": 78700
+    },
+    {
+      "epoch": 13.392857142857142,
+      "grad_norm": 0.3970576226711273,
+      "learning_rate": 5.499419953596288e-05,
+      "loss": 1.3128,
+      "step": 78750
+    },
+    {
+      "epoch": 13.401360544217686,
+      "grad_norm": 0.42503124475479126,
+      "learning_rate": 5.470417633410673e-05,
+      "loss": 1.3057,
+      "step": 78800
+    },
+    {
+      "epoch": 13.40986394557823,
+      "grad_norm": 0.4038301408290863,
+      "learning_rate": 5.441415313225058e-05,
+      "loss": 1.3093,
+      "step": 78850
+    },
+    {
+      "epoch": 13.418367346938776,
+      "grad_norm": 0.4250807762145996,
+      "learning_rate": 5.412412993039443e-05,
+      "loss": 1.3086,
+      "step": 78900
+    },
+    {
+      "epoch": 13.42687074829932,
+      "grad_norm": 0.4134768843650818,
+      "learning_rate": 5.3834106728538286e-05,
+      "loss": 1.3135,
+      "step": 78950
+    },
+    {
+      "epoch": 13.435374149659864,
+      "grad_norm": 0.39782798290252686,
+      "learning_rate": 5.3544083526682134e-05,
+      "loss": 1.305,
+      "step": 79000
+    },
+    {
+      "epoch": 13.435374149659864,
+      "eval_loss": 1.4755269289016724,
+      "eval_runtime": 75.3173,
+      "eval_samples_per_second": 1245.915,
+      "eval_steps_per_second": 4.873,
+      "step": 79000
+    },
+    {
+      "epoch": 13.443877551020408,
+      "grad_norm": 0.42068231105804443,
+      "learning_rate": 5.325406032482599e-05,
+      "loss": 1.3107,
+      "step": 79050
+    },
+    {
+      "epoch": 13.452380952380953,
+      "grad_norm": 0.40822988748550415,
+      "learning_rate": 5.2964037122969835e-05,
+      "loss": 1.3076,
+      "step": 79100
+    },
+    {
+      "epoch": 13.460884353741497,
+      "grad_norm": 0.42288413643836975,
+      "learning_rate": 5.267401392111369e-05,
+      "loss": 1.3128,
+      "step": 79150
+    },
+    {
+      "epoch": 13.46938775510204,
+      "grad_norm": 0.41481590270996094,
+      "learning_rate": 5.2383990719257544e-05,
+      "loss": 1.3083,
+      "step": 79200
+    },
+    {
+      "epoch": 13.477891156462585,
+      "grad_norm": 0.4217277467250824,
+      "learning_rate": 5.209396751740139e-05,
+      "loss": 1.2985,
+      "step": 79250
+    },
+    {
+      "epoch": 13.486394557823129,
+      "grad_norm": 0.4183300733566284,
+      "learning_rate": 5.180394431554524e-05,
+      "loss": 1.3067,
+      "step": 79300
+    },
+    {
+      "epoch": 13.494897959183673,
+      "grad_norm": 0.44343501329421997,
+      "learning_rate": 5.15139211136891e-05,
+      "loss": 1.3101,
+      "step": 79350
+    },
+    {
+      "epoch": 13.503401360544217,
+      "grad_norm": 0.47200411558151245,
+      "learning_rate": 5.122389791183295e-05,
+      "loss": 1.3113,
+      "step": 79400
+    },
+    {
+      "epoch": 13.511904761904763,
+      "grad_norm": 0.40812692046165466,
+      "learning_rate": 5.0933874709976795e-05,
+      "loss": 1.3085,
+      "step": 79450
+    },
+    {
+      "epoch": 13.520408163265307,
+      "grad_norm": 0.4125157594680786,
+      "learning_rate": 5.064385150812065e-05,
+      "loss": 1.3117,
+      "step": 79500
+    },
+    {
+      "epoch": 13.52891156462585,
+      "grad_norm": 0.42148134112358093,
+      "learning_rate": 5.0353828306264504e-05,
+      "loss": 1.3096,
+      "step": 79550
+    },
+    {
+      "epoch": 13.537414965986395,
+      "grad_norm": 0.4138471782207489,
+      "learning_rate": 5.006380510440835e-05,
+      "loss": 1.3099,
+      "step": 79600
+    },
+    {
+      "epoch": 13.545918367346939,
+      "grad_norm": 0.7425023317337036,
+      "learning_rate": 4.9773781902552205e-05,
+      "loss": 1.3098,
+      "step": 79650
+    },
+    {
+      "epoch": 13.554421768707483,
+      "grad_norm": 0.4228558838367462,
+      "learning_rate": 4.948375870069605e-05,
+      "loss": 1.3055,
+      "step": 79700
+    },
+    {
+      "epoch": 13.562925170068027,
+      "grad_norm": 0.39013656973838806,
+      "learning_rate": 4.919373549883991e-05,
+      "loss": 1.3109,
+      "step": 79750
+    },
+    {
+      "epoch": 13.571428571428571,
+      "grad_norm": 0.4059963822364807,
+      "learning_rate": 4.890371229698376e-05,
+      "loss": 1.3062,
+      "step": 79800
+    },
+    {
+      "epoch": 13.579931972789115,
+      "grad_norm": 0.41088271141052246,
+      "learning_rate": 4.861368909512761e-05,
+      "loss": 1.3104,
+      "step": 79850
+    },
+    {
+      "epoch": 13.58843537414966,
+      "grad_norm": 0.4267512857913971,
+      "learning_rate": 4.8323665893271456e-05,
+      "loss": 1.3094,
+      "step": 79900
+    },
+    {
+      "epoch": 13.596938775510203,
+      "grad_norm": 0.4166390001773834,
+      "learning_rate": 4.803364269141532e-05,
+      "loss": 1.3063,
+      "step": 79950
+    },
+    {
+      "epoch": 13.60544217687075,
+      "grad_norm": 0.43058088421821594,
+      "learning_rate": 4.7743619489559165e-05,
+      "loss": 1.3092,
+      "step": 80000
+    },
+    {
+      "epoch": 13.60544217687075,
+      "eval_loss": 1.4739412069320679,
+      "eval_runtime": 75.2973,
+      "eval_samples_per_second": 1246.247,
+      "eval_steps_per_second": 4.874,
+      "step": 80000
+    },
+    {
+      "epoch": 13.613945578231293,
+      "grad_norm": 0.4472084939479828,
+      "learning_rate": 4.745359628770301e-05,
+      "loss": 1.3092,
+      "step": 80050
+    },
+    {
+      "epoch": 13.622448979591837,
+      "grad_norm": 0.40930888056755066,
+      "learning_rate": 4.716357308584687e-05,
+      "loss": 1.2986,
+      "step": 80100
+    },
+    {
+      "epoch": 13.630952380952381,
+      "grad_norm": 0.4269891083240509,
+      "learning_rate": 4.687354988399072e-05,
+      "loss": 1.3085,
+      "step": 80150
+    },
+    {
+      "epoch": 13.639455782312925,
+      "grad_norm": 0.4400569498538971,
+      "learning_rate": 4.6583526682134575e-05,
+      "loss": 1.3091,
+      "step": 80200
+    },
+    {
+      "epoch": 13.64795918367347,
+      "grad_norm": 0.40618574619293213,
+      "learning_rate": 4.629350348027842e-05,
+      "loss": 1.3053,
+      "step": 80250
+    },
+    {
+      "epoch": 13.656462585034014,
+      "grad_norm": 0.40548551082611084,
+      "learning_rate": 4.600348027842227e-05,
+      "loss": 1.3097,
+      "step": 80300
+    },
+    {
+      "epoch": 13.664965986394558,
+      "grad_norm": 0.4029221534729004,
+      "learning_rate": 4.571345707656613e-05,
+      "loss": 1.3051,
+      "step": 80350
+    },
+    {
+      "epoch": 13.673469387755102,
+      "grad_norm": 0.4272826015949249,
+      "learning_rate": 4.542343387470998e-05,
+      "loss": 1.3078,
+      "step": 80400
+    },
+    {
+      "epoch": 13.681972789115646,
+      "grad_norm": 0.41242125630378723,
+      "learning_rate": 4.5133410672853826e-05,
+      "loss": 1.3068,
+      "step": 80450
+    },
+    {
+      "epoch": 13.69047619047619,
+      "grad_norm": 0.405517578125,
+      "learning_rate": 4.484338747099768e-05,
+      "loss": 1.3061,
+      "step": 80500
+    },
+    {
+      "epoch": 13.698979591836736,
+      "grad_norm": 0.4161379635334015,
+      "learning_rate": 4.4553364269141535e-05,
+      "loss": 1.3053,
+      "step": 80550
+    },
+    {
+      "epoch": 13.70748299319728,
+      "grad_norm": 0.4203460216522217,
+      "learning_rate": 4.426334106728538e-05,
+      "loss": 1.3043,
+      "step": 80600
+    },
+    {
+      "epoch": 13.715986394557824,
+      "grad_norm": 0.4214410185813904,
+      "learning_rate": 4.397331786542924e-05,
+      "loss": 1.3066,
+      "step": 80650
+    },
+    {
+      "epoch": 13.724489795918368,
+      "grad_norm": 0.42133694887161255,
+      "learning_rate": 4.3683294663573084e-05,
+      "loss": 1.3143,
+      "step": 80700
+    },
+    {
+      "epoch": 13.732993197278912,
+      "grad_norm": 0.39356616139411926,
+      "learning_rate": 4.339327146171694e-05,
+      "loss": 1.3078,
+      "step": 80750
+    },
+    {
+      "epoch": 13.741496598639456,
+      "grad_norm": 0.41399866342544556,
+      "learning_rate": 4.310324825986079e-05,
+      "loss": 1.3112,
+      "step": 80800
+    },
+    {
+      "epoch": 13.75,
+      "grad_norm": 0.41746941208839417,
+      "learning_rate": 4.281322505800464e-05,
+      "loss": 1.3019,
+      "step": 80850
+    },
+    {
+      "epoch": 13.758503401360544,
+      "grad_norm": 0.4164547324180603,
+      "learning_rate": 4.252320185614849e-05,
+      "loss": 1.309,
+      "step": 80900
+    },
+    {
+      "epoch": 13.767006802721088,
+      "grad_norm": 0.4242299795150757,
+      "learning_rate": 4.223317865429235e-05,
+      "loss": 1.3023,
+      "step": 80950
+    },
+    {
+      "epoch": 13.775510204081632,
+      "grad_norm": 0.4084872603416443,
+      "learning_rate": 4.1943155452436197e-05,
+      "loss": 1.3039,
+      "step": 81000
+    },
+    {
+      "epoch": 13.775510204081632,
+      "eval_loss": 1.4660990238189697,
+      "eval_runtime": 75.2655,
+      "eval_samples_per_second": 1246.773,
+      "eval_steps_per_second": 4.876,
+      "step": 81000
+    },
+    {
+      "epoch": 13.784013605442176,
+      "grad_norm": 0.4138684868812561,
+      "learning_rate": 4.1653132250580044e-05,
+      "loss": 1.3075,
+      "step": 81050
+    },
+    {
+      "epoch": 13.79251700680272,
+      "grad_norm": 0.40258848667144775,
+      "learning_rate": 4.13631090487239e-05,
+      "loss": 1.3068,
+      "step": 81100
+    },
+    {
+      "epoch": 13.801020408163264,
+      "grad_norm": 0.4167514145374298,
+      "learning_rate": 4.107308584686775e-05,
+      "loss": 1.305,
+      "step": 81150
+    },
+    {
+      "epoch": 13.80952380952381,
+      "grad_norm": 0.41513994336128235,
+      "learning_rate": 4.07830626450116e-05,
+      "loss": 1.3087,
+      "step": 81200
+    },
+    {
+      "epoch": 13.818027210884354,
+      "grad_norm": 0.42223405838012695,
+      "learning_rate": 4.0493039443155454e-05,
+      "loss": 1.3048,
+      "step": 81250
+    },
+    {
+      "epoch": 13.826530612244898,
+      "grad_norm": 0.3972738981246948,
+      "learning_rate": 4.02030162412993e-05,
+      "loss": 1.2986,
+      "step": 81300
+    },
+    {
+      "epoch": 13.835034013605442,
+      "grad_norm": 0.40687882900238037,
+      "learning_rate": 3.9912993039443156e-05,
+      "loss": 1.3106,
+      "step": 81350
+    },
+    {
+      "epoch": 13.843537414965986,
+      "grad_norm": 0.40394240617752075,
+      "learning_rate": 3.962296983758701e-05,
+      "loss": 1.309,
+      "step": 81400
+    },
+    {
+      "epoch": 13.85204081632653,
+      "grad_norm": 0.40160682797431946,
+      "learning_rate": 3.933294663573086e-05,
+      "loss": 1.3063,
+      "step": 81450
+    },
+    {
+      "epoch": 13.860544217687075,
+      "grad_norm": 0.401882529258728,
+      "learning_rate": 3.9042923433874705e-05,
+      "loss": 1.3085,
+      "step": 81500
+    },
+    {
+      "epoch": 13.869047619047619,
+      "grad_norm": 0.40280336141586304,
+      "learning_rate": 3.8758700696055686e-05,
+      "loss": 1.2966,
+      "step": 81550
+    },
+    {
+      "epoch": 13.877551020408163,
+      "grad_norm": 0.43757128715515137,
+      "learning_rate": 3.8468677494199534e-05,
+      "loss": 1.301,
+      "step": 81600
+    },
+    {
+      "epoch": 13.886054421768707,
+      "grad_norm": 0.40991008281707764,
+      "learning_rate": 3.817865429234339e-05,
+      "loss": 1.3037,
+      "step": 81650
+    },
+    {
+      "epoch": 13.89455782312925,
+      "grad_norm": 0.413152277469635,
+      "learning_rate": 3.788863109048724e-05,
+      "loss": 1.303,
+      "step": 81700
+    },
+    {
+      "epoch": 13.903061224489797,
+      "grad_norm": 0.40487709641456604,
+      "learning_rate": 3.759860788863109e-05,
+      "loss": 1.3071,
+      "step": 81750
+    },
+    {
+      "epoch": 13.91156462585034,
+      "grad_norm": 0.4063573181629181,
+      "learning_rate": 3.7308584686774944e-05,
+      "loss": 1.3037,
+      "step": 81800
+    },
+    {
+      "epoch": 13.920068027210885,
+      "grad_norm": 0.4249866306781769,
+      "learning_rate": 3.701856148491879e-05,
+      "loss": 1.3005,
+      "step": 81850
+    },
+    {
+      "epoch": 13.928571428571429,
+      "grad_norm": 0.41051676869392395,
+      "learning_rate": 3.6728538283062646e-05,
+      "loss": 1.3071,
+      "step": 81900
+    },
+    {
+      "epoch": 13.937074829931973,
+      "grad_norm": 0.39749205112457275,
+      "learning_rate": 3.64385150812065e-05,
+      "loss": 1.3049,
+      "step": 81950
+    },
+    {
+      "epoch": 13.945578231292517,
+      "grad_norm": 0.4107550382614136,
+      "learning_rate": 3.614849187935035e-05,
+      "loss": 1.3005,
+      "step": 82000
+    },
+    {
+      "epoch": 13.945578231292517,
+      "eval_loss": 1.4647165536880493,
+      "eval_runtime": 75.3191,
+      "eval_samples_per_second": 1245.886,
+      "eval_steps_per_second": 4.873,
+      "step": 82000
+    },
+    {
+      "epoch": 13.954081632653061,
+      "grad_norm": 0.4149150848388672,
+      "learning_rate": 3.5858468677494195e-05,
+      "loss": 1.3065,
+      "step": 82050
+    },
+    {
+      "epoch": 13.962585034013605,
+      "grad_norm": 0.4155055284500122,
+      "learning_rate": 3.5568445475638056e-05,
+      "loss": 1.2891,
+      "step": 82100
+    },
+    {
+      "epoch": 13.97108843537415,
+      "grad_norm": 0.40817248821258545,
+      "learning_rate": 3.5278422273781904e-05,
+      "loss": 1.303,
+      "step": 82150
+    },
+    {
+      "epoch": 13.979591836734693,
+      "grad_norm": 0.41482388973236084,
+      "learning_rate": 3.498839907192575e-05,
+      "loss": 1.3021,
+      "step": 82200
+    },
+    {
+      "epoch": 13.988095238095237,
+      "grad_norm": 0.41232186555862427,
+      "learning_rate": 3.4698375870069605e-05,
+      "loss": 1.2998,
+      "step": 82250
+    },
+    {
+      "epoch": 13.996598639455783,
+      "grad_norm": 0.40600040555000305,
+      "learning_rate": 3.440835266821346e-05,
+      "loss": 1.2983,
+      "step": 82300
+    },
+    {
+      "epoch": 14.005102040816327,
+      "grad_norm": 0.40963810682296753,
+      "learning_rate": 3.411832946635731e-05,
+      "loss": 1.2898,
+      "step": 82350
+    },
+    {
+      "epoch": 14.013605442176871,
+      "grad_norm": 0.41621676087379456,
+      "learning_rate": 3.382830626450116e-05,
+      "loss": 1.2984,
+      "step": 82400
+    },
+    {
+      "epoch": 14.022108843537415,
+      "grad_norm": 0.3934304118156433,
+      "learning_rate": 3.353828306264501e-05,
+      "loss": 1.2949,
+      "step": 82450
+    },
+    {
+      "epoch": 14.03061224489796,
+      "grad_norm": 0.4079570770263672,
+      "learning_rate": 3.324825986078887e-05,
+      "loss": 1.2966,
+      "step": 82500
+    },
+    {
+      "epoch": 14.039115646258503,
+      "grad_norm": 0.41960811614990234,
+      "learning_rate": 3.295823665893272e-05,
+      "loss": 1.3005,
+      "step": 82550
+    },
+    {
+      "epoch": 14.047619047619047,
+      "grad_norm": 0.40414050221443176,
+      "learning_rate": 3.2668213457076565e-05,
+      "loss": 1.299,
+      "step": 82600
+    },
+    {
+      "epoch": 14.056122448979592,
+      "grad_norm": 0.4004068970680237,
+      "learning_rate": 3.237819025522042e-05,
+      "loss": 1.2898,
+      "step": 82650
+    },
+    {
+      "epoch": 14.064625850340136,
+      "grad_norm": 0.41004568338394165,
+      "learning_rate": 3.2088167053364274e-05,
+      "loss": 1.2961,
+      "step": 82700
+    },
+    {
+      "epoch": 14.07312925170068,
+      "grad_norm": 0.4094617962837219,
+      "learning_rate": 3.179814385150812e-05,
+      "loss": 1.2972,
+      "step": 82750
+    },
+    {
+      "epoch": 14.081632653061224,
+      "grad_norm": 0.422753244638443,
+      "learning_rate": 3.1508120649651975e-05,
+      "loss": 1.3021,
+      "step": 82800
+    },
+    {
+      "epoch": 14.09013605442177,
+      "grad_norm": 0.41975682973861694,
+      "learning_rate": 3.121809744779582e-05,
+      "loss": 1.2996,
+      "step": 82850
+    },
+    {
+      "epoch": 14.098639455782314,
+      "grad_norm": 0.3985610008239746,
+      "learning_rate": 3.092807424593968e-05,
+      "loss": 1.3014,
+      "step": 82900
+    },
+    {
+      "epoch": 14.107142857142858,
+      "grad_norm": 0.4374600350856781,
+      "learning_rate": 3.063805104408353e-05,
+      "loss": 1.2963,
+      "step": 82950
+    },
+    {
+      "epoch": 14.115646258503402,
+      "grad_norm": 0.4096509516239166,
+      "learning_rate": 3.034802784222738e-05,
+      "loss": 1.2925,
+      "step": 83000
+    },
+    {
+      "epoch": 14.115646258503402,
+      "eval_loss": 1.4645365476608276,
+      "eval_runtime": 75.2852,
+      "eval_samples_per_second": 1246.447,
+      "eval_steps_per_second": 4.875,
+      "step": 83000
+    },
+    {
+      "epoch": 14.124149659863946,
+      "grad_norm": 0.6984773874282837,
+      "learning_rate": 3.0058004640371233e-05,
+      "loss": 1.2946,
+      "step": 83050
+    },
+    {
+      "epoch": 14.13265306122449,
+      "grad_norm": 0.4192226827144623,
+      "learning_rate": 2.976798143851508e-05,
+      "loss": 1.2906,
+      "step": 83100
+    },
+    {
+      "epoch": 14.141156462585034,
+      "grad_norm": 0.4226321280002594,
+      "learning_rate": 2.9477958236658935e-05,
+      "loss": 1.304,
+      "step": 83150
+    },
+    {
+      "epoch": 14.149659863945578,
+      "grad_norm": 0.4252030849456787,
+      "learning_rate": 2.9187935034802786e-05,
+      "loss": 1.2966,
+      "step": 83200
+    },
+    {
+      "epoch": 14.158163265306122,
+      "grad_norm": 0.3973104953765869,
+      "learning_rate": 2.8897911832946637e-05,
+      "loss": 1.2916,
+      "step": 83250
+    },
+    {
+      "epoch": 14.166666666666666,
+      "grad_norm": 0.42370930314064026,
+      "learning_rate": 2.8607888631090488e-05,
+      "loss": 1.2924,
+      "step": 83300
+    },
+    {
+      "epoch": 14.17517006802721,
+      "grad_norm": 1.85109543800354,
+      "learning_rate": 2.8317865429234342e-05,
+      "loss": 1.2927,
+      "step": 83350
+    },
+    {
+      "epoch": 14.183673469387756,
+      "grad_norm": 0.41108080744743347,
+      "learning_rate": 2.802784222737819e-05,
+      "loss": 1.2924,
+      "step": 83400
+    },
+    {
+      "epoch": 14.1921768707483,
+      "grad_norm": 0.45318904519081116,
+      "learning_rate": 2.7737819025522044e-05,
+      "loss": 1.2936,
+      "step": 83450
+    },
+    {
+      "epoch": 14.200680272108844,
+      "grad_norm": 0.402097225189209,
+      "learning_rate": 2.7447795823665895e-05,
+      "loss": 1.2966,
+      "step": 83500
+    },
+    {
+      "epoch": 14.209183673469388,
+      "grad_norm": 0.40355297923088074,
+      "learning_rate": 2.7157772621809746e-05,
+      "loss": 1.2926,
+      "step": 83550
+    },
+    {
+      "epoch": 14.217687074829932,
+      "grad_norm": 0.5472226738929749,
+      "learning_rate": 2.6873549883990723e-05,
+      "loss": 1.2938,
+      "step": 83600
+    },
+    {
+      "epoch": 14.226190476190476,
+      "grad_norm": 0.40354666113853455,
+      "learning_rate": 2.658352668213457e-05,
+      "loss": 1.2997,
+      "step": 83650
+    },
+    {
+      "epoch": 14.23469387755102,
+      "grad_norm": 0.399110347032547,
+      "learning_rate": 2.6293503480278425e-05,
+      "loss": 1.2916,
+      "step": 83700
+    },
+    {
+      "epoch": 14.243197278911564,
+      "grad_norm": 0.42456942796707153,
+      "learning_rate": 2.6003480278422276e-05,
+      "loss": 1.2987,
+      "step": 83750
+    },
+    {
+      "epoch": 14.251700680272108,
+      "grad_norm": 0.4085513949394226,
+      "learning_rate": 2.5713457076566123e-05,
+      "loss": 1.2922,
+      "step": 83800
+    },
+    {
+      "epoch": 14.260204081632653,
+      "grad_norm": 0.41955679655075073,
+      "learning_rate": 2.5423433874709977e-05,
+      "loss": 1.2942,
+      "step": 83850
+    },
+    {
+      "epoch": 14.268707482993197,
+      "grad_norm": 0.42755693197250366,
+      "learning_rate": 2.5133410672853828e-05,
+      "loss": 1.2927,
+      "step": 83900
+    },
+    {
+      "epoch": 14.27721088435374,
+      "grad_norm": 0.7680487036705017,
+      "learning_rate": 2.484338747099768e-05,
+      "loss": 1.2896,
+      "step": 83950
+    },
+    {
+      "epoch": 14.285714285714286,
+      "grad_norm": 0.41442444920539856,
+      "learning_rate": 2.455336426914153e-05,
+      "loss": 1.2993,
+      "step": 84000
+    },
+    {
+      "epoch": 14.285714285714286,
+      "eval_loss": 1.4609616994857788,
+      "eval_runtime": 75.2756,
+      "eval_samples_per_second": 1246.605,
+      "eval_steps_per_second": 4.875,
+      "step": 84000
+    },
+    {
+      "epoch": 14.29421768707483,
+      "grad_norm": 0.40402066707611084,
+      "learning_rate": 2.4263341067285384e-05,
+      "loss": 1.2965,
+      "step": 84050
+    },
+    {
+      "epoch": 14.302721088435375,
+      "grad_norm": 0.4172896146774292,
+      "learning_rate": 2.3973317865429232e-05,
+      "loss": 1.2912,
+      "step": 84100
+    },
+    {
+      "epoch": 14.311224489795919,
+      "grad_norm": 0.4188857078552246,
+      "learning_rate": 2.3683294663573086e-05,
+      "loss": 1.2941,
+      "step": 84150
+    },
+    {
+      "epoch": 14.319727891156463,
+      "grad_norm": 0.4177609384059906,
+      "learning_rate": 2.3393271461716937e-05,
+      "loss": 1.2956,
+      "step": 84200
+    },
+    {
+      "epoch": 14.328231292517007,
+      "grad_norm": 0.4155607223510742,
+      "learning_rate": 2.3103248259860788e-05,
+      "loss": 1.2909,
+      "step": 84250
+    },
+    {
+      "epoch": 14.33673469387755,
+      "grad_norm": 0.39458438754081726,
+      "learning_rate": 2.281322505800464e-05,
+      "loss": 1.2991,
+      "step": 84300
+    },
+    {
+      "epoch": 14.345238095238095,
+      "grad_norm": 0.3875535726547241,
+      "learning_rate": 2.2523201856148493e-05,
+      "loss": 1.2914,
+      "step": 84350
+    },
+    {
+      "epoch": 14.353741496598639,
+      "grad_norm": 0.39373454451560974,
+      "learning_rate": 2.223317865429234e-05,
+      "loss": 1.2921,
+      "step": 84400
+    },
+    {
+      "epoch": 14.362244897959183,
+      "grad_norm": 0.4142104685306549,
+      "learning_rate": 2.1943155452436195e-05,
+      "loss": 1.2924,
+      "step": 84450
+    },
+    {
+      "epoch": 14.370748299319727,
+      "grad_norm": 0.4059418737888336,
+      "learning_rate": 2.1653132250580046e-05,
+      "loss": 1.2927,
+      "step": 84500
+    },
+    {
+      "epoch": 14.379251700680273,
+      "grad_norm": 0.42010223865509033,
+      "learning_rate": 2.13631090487239e-05,
+      "loss": 1.2946,
+      "step": 84550
+    },
+    {
+      "epoch": 14.387755102040817,
+      "grad_norm": 0.4005703330039978,
+      "learning_rate": 2.1073085846867748e-05,
+      "loss": 1.2889,
+      "step": 84600
+    },
+    {
+      "epoch": 14.396258503401361,
+      "grad_norm": 0.434023916721344,
+      "learning_rate": 2.0783062645011602e-05,
+      "loss": 1.2925,
+      "step": 84650
+    },
+    {
+      "epoch": 14.404761904761905,
+      "grad_norm": 0.41903188824653625,
+      "learning_rate": 2.0493039443155453e-05,
+      "loss": 1.2982,
+      "step": 84700
+    },
+    {
+      "epoch": 14.41326530612245,
+      "grad_norm": 0.41402989625930786,
+      "learning_rate": 2.0203016241299304e-05,
+      "loss": 1.2972,
+      "step": 84750
+    },
+    {
+      "epoch": 14.421768707482993,
+      "grad_norm": 0.4167090952396393,
+      "learning_rate": 1.9912993039443155e-05,
+      "loss": 1.2975,
+      "step": 84800
+    },
+    {
+      "epoch": 14.430272108843537,
+      "grad_norm": 0.40198904275894165,
+      "learning_rate": 1.962296983758701e-05,
+      "loss": 1.2898,
+      "step": 84850
+    },
+    {
+      "epoch": 14.438775510204081,
+      "grad_norm": 0.412724107503891,
+      "learning_rate": 1.9332946635730856e-05,
+      "loss": 1.2955,
+      "step": 84900
+    },
+    {
+      "epoch": 14.447278911564625,
+      "grad_norm": 0.4073164165019989,
+      "learning_rate": 1.904292343387471e-05,
+      "loss": 1.2907,
+      "step": 84950
+    },
+    {
+      "epoch": 14.45578231292517,
+      "grad_norm": 0.4109826982021332,
+      "learning_rate": 1.875290023201856e-05,
+      "loss": 1.2935,
+      "step": 85000
+    },
+    {
+      "epoch": 14.45578231292517,
+      "eval_loss": 1.457992672920227,
+      "eval_runtime": 75.28,
+      "eval_samples_per_second": 1246.534,
+      "eval_steps_per_second": 4.875,
+      "step": 85000
+    },
+    {
+      "epoch": 14.464285714285714,
+      "grad_norm": 0.40559759736061096,
+      "learning_rate": 1.8462877030162412e-05,
+      "loss": 1.2888,
+      "step": 85050
+    },
+    {
+      "epoch": 14.47278911564626,
+      "grad_norm": 0.4154794216156006,
+      "learning_rate": 1.8172853828306263e-05,
+      "loss": 1.289,
+      "step": 85100
+    },
+    {
+      "epoch": 14.481292517006803,
+      "grad_norm": 0.4155610501766205,
+      "learning_rate": 1.7882830626450118e-05,
+      "loss": 1.2922,
+      "step": 85150
+    },
+    {
+      "epoch": 14.489795918367347,
+      "grad_norm": 0.4222477674484253,
+      "learning_rate": 1.7592807424593965e-05,
+      "loss": 1.2867,
+      "step": 85200
+    },
+    {
+      "epoch": 14.498299319727892,
+      "grad_norm": 0.4063816964626312,
+      "learning_rate": 1.730278422273782e-05,
+      "loss": 1.2888,
+      "step": 85250
+    },
+    {
+      "epoch": 14.506802721088436,
+      "grad_norm": 0.3986765742301941,
+      "learning_rate": 1.701276102088167e-05,
+      "loss": 1.285,
+      "step": 85300
+    },
+    {
+      "epoch": 14.51530612244898,
+      "grad_norm": 0.4038733243942261,
+      "learning_rate": 1.672273781902552e-05,
+      "loss": 1.2913,
+      "step": 85350
+    },
+    {
+      "epoch": 14.523809523809524,
+      "grad_norm": 0.40419521927833557,
+      "learning_rate": 1.6432714617169372e-05,
+      "loss": 1.2872,
+      "step": 85400
+    },
+    {
+      "epoch": 14.532312925170068,
+      "grad_norm": 0.41449958086013794,
+      "learning_rate": 1.6142691415313226e-05,
+      "loss": 1.2853,
+      "step": 85450
+    },
+    {
+      "epoch": 14.540816326530612,
+      "grad_norm": 0.4067400097846985,
+      "learning_rate": 1.5852668213457077e-05,
+      "loss": 1.2919,
+      "step": 85500
+    },
+    {
+      "epoch": 14.549319727891156,
+      "grad_norm": 0.6428622007369995,
+      "learning_rate": 1.5562645011600928e-05,
+      "loss": 1.2935,
+      "step": 85550
+    },
+    {
+      "epoch": 14.5578231292517,
+      "grad_norm": 0.4078386723995209,
+      "learning_rate": 1.527262180974478e-05,
+      "loss": 1.2892,
+      "step": 85600
+    },
+    {
+      "epoch": 14.566326530612244,
+      "grad_norm": 0.4054696261882782,
+      "learning_rate": 1.4982598607888632e-05,
+      "loss": 1.2871,
+      "step": 85650
+    },
+    {
+      "epoch": 14.57482993197279,
+      "grad_norm": 0.4058522880077362,
+      "learning_rate": 1.4692575406032483e-05,
+      "loss": 1.286,
+      "step": 85700
+    },
+    {
+      "epoch": 14.583333333333334,
+      "grad_norm": 0.40443381667137146,
+      "learning_rate": 1.4402552204176335e-05,
+      "loss": 1.2888,
+      "step": 85750
+    },
+    {
+      "epoch": 14.591836734693878,
+      "grad_norm": 0.4228183925151825,
+      "learning_rate": 1.4112529002320186e-05,
+      "loss": 1.2875,
+      "step": 85800
+    },
+    {
+      "epoch": 14.600340136054422,
+      "grad_norm": 0.4046574831008911,
+      "learning_rate": 1.3822505800464039e-05,
+      "loss": 1.2883,
+      "step": 85850
+    },
+    {
+      "epoch": 14.608843537414966,
+      "grad_norm": 0.4183114767074585,
+      "learning_rate": 1.353248259860789e-05,
+      "loss": 1.2929,
+      "step": 85900
+    },
+    {
+      "epoch": 14.61734693877551,
+      "grad_norm": 0.40642350912094116,
+      "learning_rate": 1.324245939675174e-05,
+      "loss": 1.2892,
+      "step": 85950
+    },
+    {
+      "epoch": 14.625850340136054,
+      "grad_norm": 0.40780848264694214,
+      "learning_rate": 1.2952436194895593e-05,
+      "loss": 1.2888,
+      "step": 86000
+    },
+    {
+      "epoch": 14.625850340136054,
+      "eval_loss": 1.4564687013626099,
+      "eval_runtime": 75.2583,
+      "eval_samples_per_second": 1246.893,
+      "eval_steps_per_second": 4.877,
+      "step": 86000
+    },
+    {
+      "epoch": 14.634353741496598,
+      "grad_norm": 0.399654746055603,
+      "learning_rate": 1.2662412993039444e-05,
+      "loss": 1.292,
+      "step": 86050
+    },
+    {
+      "epoch": 14.642857142857142,
+      "grad_norm": 0.4076179563999176,
+      "learning_rate": 1.2372389791183295e-05,
+      "loss": 1.2894,
+      "step": 86100
+    },
+    {
+      "epoch": 14.651360544217686,
+      "grad_norm": 0.4201255738735199,
+      "learning_rate": 1.2082366589327147e-05,
+      "loss": 1.289,
+      "step": 86150
+    },
+    {
+      "epoch": 14.65986394557823,
+      "grad_norm": 0.39747655391693115,
+      "learning_rate": 1.1792343387470998e-05,
+      "loss": 1.2915,
+      "step": 86200
+    },
+    {
+      "epoch": 14.668367346938776,
+      "grad_norm": 0.4104771316051483,
+      "learning_rate": 1.1508120649651974e-05,
+      "loss": 1.2864,
+      "step": 86250
+    },
+    {
+      "epoch": 14.67687074829932,
+      "grad_norm": 0.41255298256874084,
+      "learning_rate": 1.1218097447795825e-05,
+      "loss": 1.2916,
+      "step": 86300
+    },
+    {
+      "epoch": 14.685374149659864,
+      "grad_norm": 0.4036099910736084,
+      "learning_rate": 1.0928074245939676e-05,
+      "loss": 1.2921,
+      "step": 86350
+    },
+    {
+      "epoch": 14.693877551020408,
+      "grad_norm": 0.40650486946105957,
+      "learning_rate": 1.0638051044083526e-05,
+      "loss": 1.2896,
+      "step": 86400
+    },
+    {
+      "epoch": 14.702380952380953,
+      "grad_norm": 0.47484561800956726,
+      "learning_rate": 1.0348027842227377e-05,
+      "loss": 1.2917,
+      "step": 86450
+    },
+    {
+      "epoch": 14.710884353741497,
+      "grad_norm": 0.4110647737979889,
+      "learning_rate": 1.0058004640371228e-05,
+      "loss": 1.2878,
+      "step": 86500
+    },
+    {
+      "epoch": 14.71938775510204,
+      "grad_norm": 0.4205218553543091,
+      "learning_rate": 9.767981438515081e-06,
+      "loss": 1.2879,
+      "step": 86550
+    },
+    {
+      "epoch": 14.727891156462585,
+      "grad_norm": 0.4059794843196869,
+      "learning_rate": 9.477958236658932e-06,
+      "loss": 1.2871,
+      "step": 86600
+    },
+    {
+      "epoch": 14.736394557823129,
+      "grad_norm": 0.41226640343666077,
+      "learning_rate": 9.187935034802784e-06,
+      "loss": 1.2938,
+      "step": 86650
+    },
+    {
+      "epoch": 14.744897959183673,
+      "grad_norm": 0.41290587186813354,
+      "learning_rate": 8.897911832946635e-06,
+      "loss": 1.286,
+      "step": 86700
+    },
+    {
+      "epoch": 14.753401360544217,
+      "grad_norm": 0.4120457172393799,
+      "learning_rate": 8.607888631090486e-06,
+      "loss": 1.289,
+      "step": 86750
+    },
+    {
+      "epoch": 14.761904761904763,
+      "grad_norm": 0.4931107461452484,
+      "learning_rate": 8.317865429234339e-06,
+      "loss": 1.2896,
+      "step": 86800
+    },
+    {
+      "epoch": 14.770408163265307,
+      "grad_norm": 0.4076496660709381,
+      "learning_rate": 8.02784222737819e-06,
+      "loss": 1.2877,
+      "step": 86850
+    },
+    {
+      "epoch": 14.77891156462585,
+      "grad_norm": 0.41810768842697144,
+      "learning_rate": 7.737819025522042e-06,
+      "loss": 1.2873,
+      "step": 86900
+    },
+    {
+      "epoch": 14.787414965986395,
+      "grad_norm": 0.41356149315834045,
+      "learning_rate": 7.447795823665894e-06,
+      "loss": 1.2886,
+      "step": 86950
+    },
+    {
+      "epoch": 14.795918367346939,
+      "grad_norm": 0.43218472599983215,
+      "learning_rate": 7.157772621809745e-06,
+      "loss": 1.291,
+      "step": 87000
+    },
+    {
+      "epoch": 14.795918367346939,
+      "eval_loss": 1.4570873975753784,
+      "eval_runtime": 75.2933,
+      "eval_samples_per_second": 1246.313,
+      "eval_steps_per_second": 4.874,
+      "step": 87000
+    },
+    {
+      "epoch": 14.804421768707483,
+      "grad_norm": 0.4102339446544647,
+      "learning_rate": 6.867749419953597e-06,
+      "loss": 1.2848,
+      "step": 87050
+    },
+    {
+      "epoch": 14.812925170068027,
+      "grad_norm": 0.41269227862358093,
+      "learning_rate": 6.577726218097448e-06,
+      "loss": 1.2888,
+      "step": 87100
+    },
+    {
+      "epoch": 14.821428571428571,
+      "grad_norm": 0.4198719561100006,
+      "learning_rate": 6.287703016241299e-06,
+      "loss": 1.2904,
+      "step": 87150
+    },
+    {
+      "epoch": 14.829931972789115,
+      "grad_norm": 0.40524616837501526,
+      "learning_rate": 5.997679814385151e-06,
+      "loss": 1.2916,
+      "step": 87200
+    },
+    {
+      "epoch": 14.83843537414966,
+      "grad_norm": 0.40271249413490295,
+      "learning_rate": 5.707656612529003e-06,
+      "loss": 1.2821,
+      "step": 87250
+    },
+    {
+      "epoch": 14.846938775510203,
+      "grad_norm": 0.404710590839386,
+      "learning_rate": 5.4176334106728545e-06,
+      "loss": 1.2848,
+      "step": 87300
+    },
+    {
+      "epoch": 14.85544217687075,
+      "grad_norm": 0.40877318382263184,
+      "learning_rate": 5.127610208816705e-06,
+      "loss": 1.2824,
+      "step": 87350
+    },
+    {
+      "epoch": 14.863945578231293,
+      "grad_norm": 0.40743860602378845,
+      "learning_rate": 4.837587006960557e-06,
+      "loss": 1.2875,
+      "step": 87400
+    },
+    {
+      "epoch": 14.872448979591837,
+      "grad_norm": 0.40123429894447327,
+      "learning_rate": 4.547563805104409e-06,
+      "loss": 1.2879,
+      "step": 87450
+    },
+    {
+      "epoch": 14.880952380952381,
+      "grad_norm": 0.4120638072490692,
+      "learning_rate": 4.257540603248261e-06,
+      "loss": 1.2886,
+      "step": 87500
+    },
+    {
+      "epoch": 14.889455782312925,
+      "grad_norm": 0.4441770017147064,
+      "learning_rate": 3.9675174013921115e-06,
+      "loss": 1.2903,
+      "step": 87550
+    },
+    {
+      "epoch": 14.89795918367347,
+      "grad_norm": 0.39684155583381653,
+      "learning_rate": 3.677494199535963e-06,
+      "loss": 1.2894,
+      "step": 87600
+    },
+    {
+      "epoch": 14.906462585034014,
+      "grad_norm": 0.40048104524612427,
+      "learning_rate": 3.387470997679814e-06,
+      "loss": 1.2872,
+      "step": 87650
+    },
+    {
+      "epoch": 14.914965986394558,
+      "grad_norm": 0.40761691331863403,
+      "learning_rate": 3.097447795823666e-06,
+      "loss": 1.2856,
+      "step": 87700
+    },
+    {
+      "epoch": 14.923469387755102,
+      "grad_norm": 0.41914671659469604,
+      "learning_rate": 2.807424593967517e-06,
+      "loss": 1.2864,
+      "step": 87750
+    },
+    {
+      "epoch": 14.931972789115646,
+      "grad_norm": 0.41371914744377136,
+      "learning_rate": 2.517401392111369e-06,
+      "loss": 1.2807,
+      "step": 87800
+    },
+    {
+      "epoch": 14.94047619047619,
+      "grad_norm": 0.4007938802242279,
+      "learning_rate": 2.2273781902552203e-06,
+      "loss": 1.2869,
+      "step": 87850
+    },
+    {
+      "epoch": 14.948979591836736,
+      "grad_norm": 0.4132489264011383,
+      "learning_rate": 1.937354988399072e-06,
+      "loss": 1.2822,
+      "step": 87900
+    },
+    {
+      "epoch": 14.95748299319728,
+      "grad_norm": 0.4170679748058319,
+      "learning_rate": 1.6473317865429233e-06,
+      "loss": 1.2867,
+      "step": 87950
+    },
+    {
+      "epoch": 14.965986394557824,
+      "grad_norm": 0.4113543629646301,
+      "learning_rate": 1.3573085846867749e-06,
+      "loss": 1.2852,
+      "step": 88000
+    },
+    {
+      "epoch": 14.965986394557824,
+      "eval_loss": 1.4495242834091187,
+      "eval_runtime": 75.6807,
+      "eval_samples_per_second": 1239.933,
+      "eval_steps_per_second": 4.849,
+      "step": 88000
+    },
+    {
+      "epoch": 14.974489795918368,
+      "grad_norm": 0.4077957272529602,
+      "learning_rate": 1.0672853828306264e-06,
+      "loss": 1.2852,
+      "step": 88050
+    },
+    {
+      "epoch": 14.982993197278912,
+      "grad_norm": 0.403070330619812,
+      "learning_rate": 7.772621809744779e-07,
+      "loss": 1.2854,
+      "step": 88100
+    },
+    {
+      "epoch": 14.991496598639456,
+      "grad_norm": 0.41275766491889954,
+      "learning_rate": 4.872389791183295e-07,
+      "loss": 1.288,
+      "step": 88150
+    },
+    {
+      "epoch": 15.0,
+      "grad_norm": 0.4174801707267761,
+      "learning_rate": 1.97215777262181e-07,
+      "loss": 1.2938,
+      "step": 88200
+    }
+  ],
+  "logging_steps": 50,
+  "max_steps": 88200,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 15,
+  "save_steps": 1000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 2.377209771467368e+19,
+  "train_batch_size": 32,
+  "trial_name": null,
+  "trial_params": null
+}