diff --git "a/checkpoint-699/trainer_state.json" "b/checkpoint-699/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-699/trainer_state.json"
@@ -0,0 +1,4926 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.9971469329529246,
+  "eval_steps": 500,
+  "global_step": 699,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0042796005706134095,
+      "grad_norm": 34.48759460449219,
+      "learning_rate": 5.0000000000000004e-08,
+      "loss": 4.9483,
+      "step": 1
+    },
+    {
+      "epoch": 0.008559201141226819,
+      "grad_norm": 35.18870544433594,
+      "learning_rate": 1.0000000000000001e-07,
+      "loss": 5.1789,
+      "step": 2
+    },
+    {
+      "epoch": 0.012838801711840228,
+      "grad_norm": 34.3349494934082,
+      "learning_rate": 1.5000000000000002e-07,
+      "loss": 4.8881,
+      "step": 3
+    },
+    {
+      "epoch": 0.017118402282453638,
+      "grad_norm": 36.14989471435547,
+      "learning_rate": 2.0000000000000002e-07,
+      "loss": 5.0836,
+      "step": 4
+    },
+    {
+      "epoch": 0.021398002853067047,
+      "grad_norm": 33.93682861328125,
+      "learning_rate": 2.5000000000000004e-07,
+      "loss": 4.9619,
+      "step": 5
+    },
+    {
+      "epoch": 0.025677603423680456,
+      "grad_norm": 34.47760772705078,
+      "learning_rate": 3.0000000000000004e-07,
+      "loss": 4.9723,
+      "step": 6
+    },
+    {
+      "epoch": 0.029957203994293864,
+      "grad_norm": 34.52993392944336,
+      "learning_rate": 3.5000000000000004e-07,
+      "loss": 4.9278,
+      "step": 7
+    },
+    {
+      "epoch": 0.034236804564907276,
+      "grad_norm": 33.86513900756836,
+      "learning_rate": 4.0000000000000003e-07,
+      "loss": 4.9141,
+      "step": 8
+    },
+    {
+      "epoch": 0.03851640513552068,
+      "grad_norm": 32.446311950683594,
+      "learning_rate": 4.5000000000000003e-07,
+      "loss": 4.7811,
+      "step": 9
+    },
+    {
+      "epoch": 0.042796005706134094,
+      "grad_norm": 33.34199905395508,
+      "learning_rate": 5.000000000000001e-07,
+      "loss": 4.8566,
+      "step": 10
+    },
+    {
+      "epoch": 0.047075606276747506,
+      "grad_norm": 34.60908508300781,
+      "learning_rate": 5.5e-07,
+      "loss": 4.9406,
+      "step": 11
+    },
+    {
+      "epoch": 0.05135520684736091,
+      "grad_norm": 34.570987701416016,
+      "learning_rate": 6.000000000000001e-07,
+      "loss": 4.8281,
+      "step": 12
+    },
+    {
+      "epoch": 0.05563480741797432,
+      "grad_norm": 32.070953369140625,
+      "learning_rate": 6.5e-07,
+      "loss": 4.6165,
+      "step": 13
+    },
+    {
+      "epoch": 0.05991440798858773,
+      "grad_norm": 33.586509704589844,
+      "learning_rate": 7.000000000000001e-07,
+      "loss": 4.8012,
+      "step": 14
+    },
+    {
+      "epoch": 0.06419400855920114,
+      "grad_norm": 30.96600341796875,
+      "learning_rate": 7.5e-07,
+      "loss": 4.5845,
+      "step": 15
+    },
+    {
+      "epoch": 0.06847360912981455,
+      "grad_norm": 30.152454376220703,
+      "learning_rate": 8.000000000000001e-07,
+      "loss": 4.4356,
+      "step": 16
+    },
+    {
+      "epoch": 0.07275320970042796,
+      "grad_norm": 28.61970329284668,
+      "learning_rate": 8.500000000000001e-07,
+      "loss": 4.359,
+      "step": 17
+    },
+    {
+      "epoch": 0.07703281027104136,
+      "grad_norm": 27.507383346557617,
+      "learning_rate": 9.000000000000001e-07,
+      "loss": 4.4086,
+      "step": 18
+    },
+    {
+      "epoch": 0.08131241084165478,
+      "grad_norm": 25.110597610473633,
+      "learning_rate": 9.500000000000001e-07,
+      "loss": 4.1411,
+      "step": 19
+    },
+    {
+      "epoch": 0.08559201141226819,
+      "grad_norm": 23.737529754638672,
+      "learning_rate": 1.0000000000000002e-06,
+      "loss": 4.0639,
+      "step": 20
+    },
+    {
+      "epoch": 0.0898716119828816,
+      "grad_norm": 22.502748489379883,
+      "learning_rate": 1.0500000000000001e-06,
+      "loss": 3.9744,
+      "step": 21
+    },
+    {
+      "epoch": 0.09415121255349501,
+      "grad_norm": 21.957761764526367,
+      "learning_rate": 1.1e-06,
+      "loss": 3.8115,
+      "step": 22
+    },
+    {
+      "epoch": 0.09843081312410841,
+      "grad_norm": 20.552080154418945,
+      "learning_rate": 1.1500000000000002e-06,
+      "loss": 3.6059,
+      "step": 23
+    },
+    {
+      "epoch": 0.10271041369472182,
+      "grad_norm": 21.490442276000977,
+      "learning_rate": 1.2000000000000002e-06,
+      "loss": 3.5285,
+      "step": 24
+    },
+    {
+      "epoch": 0.10699001426533523,
+      "grad_norm": 20.63368034362793,
+      "learning_rate": 1.25e-06,
+      "loss": 3.3484,
+      "step": 25
+    },
+    {
+      "epoch": 0.11126961483594865,
+      "grad_norm": 20.170631408691406,
+      "learning_rate": 1.3e-06,
+      "loss": 3.0973,
+      "step": 26
+    },
+    {
+      "epoch": 0.11554921540656206,
+      "grad_norm": 20.217880249023438,
+      "learning_rate": 1.3500000000000002e-06,
+      "loss": 2.9758,
+      "step": 27
+    },
+    {
+      "epoch": 0.11982881597717546,
+      "grad_norm": 18.303476333618164,
+      "learning_rate": 1.4000000000000001e-06,
+      "loss": 2.8659,
+      "step": 28
+    },
+    {
+      "epoch": 0.12410841654778887,
+      "grad_norm": 15.922847747802734,
+      "learning_rate": 1.45e-06,
+      "loss": 2.6369,
+      "step": 29
+    },
+    {
+      "epoch": 0.12838801711840228,
+      "grad_norm": 14.529817581176758,
+      "learning_rate": 1.5e-06,
+      "loss": 2.5532,
+      "step": 30
+    },
+    {
+      "epoch": 0.1326676176890157,
+      "grad_norm": 14.177539825439453,
+      "learning_rate": 1.5500000000000002e-06,
+      "loss": 2.4497,
+      "step": 31
+    },
+    {
+      "epoch": 0.1369472182596291,
+      "grad_norm": 15.144468307495117,
+      "learning_rate": 1.6000000000000001e-06,
+      "loss": 2.3432,
+      "step": 32
+    },
+    {
+      "epoch": 0.14122681883024252,
+      "grad_norm": 14.920907974243164,
+      "learning_rate": 1.6500000000000003e-06,
+      "loss": 2.14,
+      "step": 33
+    },
+    {
+      "epoch": 0.14550641940085593,
+      "grad_norm": 15.700516700744629,
+      "learning_rate": 1.7000000000000002e-06,
+      "loss": 2.0247,
+      "step": 34
+    },
+    {
+      "epoch": 0.14978601997146934,
+      "grad_norm": 15.600655555725098,
+      "learning_rate": 1.75e-06,
+      "loss": 1.8893,
+      "step": 35
+    },
+    {
+      "epoch": 0.15406562054208273,
+      "grad_norm": 15.281895637512207,
+      "learning_rate": 1.8000000000000001e-06,
+      "loss": 1.7768,
+      "step": 36
+    },
+    {
+      "epoch": 0.15834522111269614,
+      "grad_norm": 13.99860668182373,
+      "learning_rate": 1.85e-06,
+      "loss": 1.5486,
+      "step": 37
+    },
+    {
+      "epoch": 0.16262482168330955,
+      "grad_norm": 13.295161247253418,
+      "learning_rate": 1.9000000000000002e-06,
+      "loss": 1.427,
+      "step": 38
+    },
+    {
+      "epoch": 0.16690442225392296,
+      "grad_norm": 12.667908668518066,
+      "learning_rate": 1.9500000000000004e-06,
+      "loss": 1.2251,
+      "step": 39
+    },
+    {
+      "epoch": 0.17118402282453637,
+      "grad_norm": 12.730255126953125,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 1.1402,
+      "step": 40
+    },
+    {
+      "epoch": 0.1754636233951498,
+      "grad_norm": 13.135544776916504,
+      "learning_rate": 2.05e-06,
+      "loss": 0.9867,
+      "step": 41
+    },
+    {
+      "epoch": 0.1797432239657632,
+      "grad_norm": 12.429509162902832,
+      "learning_rate": 2.1000000000000002e-06,
+      "loss": 0.8455,
+      "step": 42
+    },
+    {
+      "epoch": 0.1840228245363766,
+      "grad_norm": 11.61819076538086,
+      "learning_rate": 2.15e-06,
+      "loss": 0.6935,
+      "step": 43
+    },
+    {
+      "epoch": 0.18830242510699002,
+      "grad_norm": 10.11789608001709,
+      "learning_rate": 2.2e-06,
+      "loss": 0.5486,
+      "step": 44
+    },
+    {
+      "epoch": 0.19258202567760344,
+      "grad_norm": 8.705474853515625,
+      "learning_rate": 2.25e-06,
+      "loss": 0.4345,
+      "step": 45
+    },
+    {
+      "epoch": 0.19686162624821682,
+      "grad_norm": 7.407937526702881,
+      "learning_rate": 2.3000000000000004e-06,
+      "loss": 0.3223,
+      "step": 46
+    },
+    {
+      "epoch": 0.20114122681883023,
+      "grad_norm": 5.553786754608154,
+      "learning_rate": 2.35e-06,
+      "loss": 0.2376,
+      "step": 47
+    },
+    {
+      "epoch": 0.20542082738944364,
+      "grad_norm": 3.248605966567993,
+      "learning_rate": 2.4000000000000003e-06,
+      "loss": 0.198,
+      "step": 48
+    },
+    {
+      "epoch": 0.20970042796005706,
+      "grad_norm": 3.1356797218322754,
+      "learning_rate": 2.4500000000000003e-06,
+      "loss": 0.1706,
+      "step": 49
+    },
+    {
+      "epoch": 0.21398002853067047,
+      "grad_norm": 2.7913248538970947,
+      "learning_rate": 2.5e-06,
+      "loss": 0.1436,
+      "step": 50
+    },
+    {
+      "epoch": 0.21825962910128388,
+      "grad_norm": 2.2059645652770996,
+      "learning_rate": 2.55e-06,
+      "loss": 0.144,
+      "step": 51
+    },
+    {
+      "epoch": 0.2225392296718973,
+      "grad_norm": 1.5722575187683105,
+      "learning_rate": 2.6e-06,
+      "loss": 0.1203,
+      "step": 52
+    },
+    {
+      "epoch": 0.2268188302425107,
+      "grad_norm": 1.3645284175872803,
+      "learning_rate": 2.6500000000000005e-06,
+      "loss": 0.1309,
+      "step": 53
+    },
+    {
+      "epoch": 0.23109843081312412,
+      "grad_norm": 1.3658019304275513,
+      "learning_rate": 2.7000000000000004e-06,
+      "loss": 0.1122,
+      "step": 54
+    },
+    {
+      "epoch": 0.23537803138373753,
+      "grad_norm": 1.1719582080841064,
+      "learning_rate": 2.7500000000000004e-06,
+      "loss": 0.1239,
+      "step": 55
+    },
+    {
+      "epoch": 0.2396576319543509,
+      "grad_norm": 1.030941367149353,
+      "learning_rate": 2.8000000000000003e-06,
+      "loss": 0.1124,
+      "step": 56
+    },
+    {
+      "epoch": 0.24393723252496433,
+      "grad_norm": 0.7212232947349548,
+      "learning_rate": 2.85e-06,
+      "loss": 0.0848,
+      "step": 57
+    },
+    {
+      "epoch": 0.24821683309557774,
+      "grad_norm": 0.8292589783668518,
+      "learning_rate": 2.9e-06,
+      "loss": 0.0976,
+      "step": 58
+    },
+    {
+      "epoch": 0.2524964336661912,
+      "grad_norm": 1.0893069505691528,
+      "learning_rate": 2.95e-06,
+      "loss": 0.0954,
+      "step": 59
+    },
+    {
+      "epoch": 0.25677603423680456,
+      "grad_norm": 0.8184618353843689,
+      "learning_rate": 3e-06,
+      "loss": 0.1029,
+      "step": 60
+    },
+    {
+      "epoch": 0.26105563480741795,
+      "grad_norm": 0.551844596862793,
+      "learning_rate": 3.05e-06,
+      "loss": 0.0846,
+      "step": 61
+    },
+    {
+      "epoch": 0.2653352353780314,
+      "grad_norm": 0.6214303970336914,
+      "learning_rate": 3.1000000000000004e-06,
+      "loss": 0.0855,
+      "step": 62
+    },
+    {
+      "epoch": 0.26961483594864477,
+      "grad_norm": 0.6127054691314697,
+      "learning_rate": 3.1500000000000003e-06,
+      "loss": 0.0858,
+      "step": 63
+    },
+    {
+      "epoch": 0.2738944365192582,
+      "grad_norm": 0.6121392250061035,
+      "learning_rate": 3.2000000000000003e-06,
+      "loss": 0.0913,
+      "step": 64
+    },
+    {
+      "epoch": 0.2781740370898716,
+      "grad_norm": 0.5743774771690369,
+      "learning_rate": 3.2500000000000002e-06,
+      "loss": 0.0813,
+      "step": 65
+    },
+    {
+      "epoch": 0.28245363766048504,
+      "grad_norm": 0.6720462441444397,
+      "learning_rate": 3.3000000000000006e-06,
+      "loss": 0.0871,
+      "step": 66
+    },
+    {
+      "epoch": 0.2867332382310984,
+      "grad_norm": 0.5387166142463684,
+      "learning_rate": 3.3500000000000005e-06,
+      "loss": 0.0722,
+      "step": 67
+    },
+    {
+      "epoch": 0.29101283880171186,
+      "grad_norm": 0.4062807261943817,
+      "learning_rate": 3.4000000000000005e-06,
+      "loss": 0.0699,
+      "step": 68
+    },
+    {
+      "epoch": 0.29529243937232524,
+      "grad_norm": 0.43954649567604065,
+      "learning_rate": 3.45e-06,
+      "loss": 0.0808,
+      "step": 69
+    },
+    {
+      "epoch": 0.2995720399429387,
+      "grad_norm": 0.5136083960533142,
+      "learning_rate": 3.5e-06,
+      "loss": 0.0782,
+      "step": 70
+    },
+    {
+      "epoch": 0.30385164051355207,
+      "grad_norm": 0.43194347620010376,
+      "learning_rate": 3.5500000000000003e-06,
+      "loss": 0.0883,
+      "step": 71
+    },
+    {
+      "epoch": 0.30813124108416545,
+      "grad_norm": 0.42371511459350586,
+      "learning_rate": 3.6000000000000003e-06,
+      "loss": 0.0875,
+      "step": 72
+    },
+    {
+      "epoch": 0.3124108416547789,
+      "grad_norm": 0.4384445250034332,
+      "learning_rate": 3.65e-06,
+      "loss": 0.0842,
+      "step": 73
+    },
+    {
+      "epoch": 0.3166904422253923,
+      "grad_norm": 0.3909939229488373,
+      "learning_rate": 3.7e-06,
+      "loss": 0.0718,
+      "step": 74
+    },
+    {
+      "epoch": 0.3209700427960057,
+      "grad_norm": 0.33483919501304626,
+      "learning_rate": 3.7500000000000005e-06,
+      "loss": 0.0766,
+      "step": 75
+    },
+    {
+      "epoch": 0.3252496433666191,
+      "grad_norm": 0.6500667929649353,
+      "learning_rate": 3.8000000000000005e-06,
+      "loss": 0.0841,
+      "step": 76
+    },
+    {
+      "epoch": 0.32952924393723254,
+      "grad_norm": 0.42915841937065125,
+      "learning_rate": 3.85e-06,
+      "loss": 0.0743,
+      "step": 77
+    },
+    {
+      "epoch": 0.3338088445078459,
+      "grad_norm": 0.5390797853469849,
+      "learning_rate": 3.900000000000001e-06,
+      "loss": 0.0784,
+      "step": 78
+    },
+    {
+      "epoch": 0.33808844507845937,
+      "grad_norm": 0.4562499225139618,
+      "learning_rate": 3.95e-06,
+      "loss": 0.0768,
+      "step": 79
+    },
+    {
+      "epoch": 0.34236804564907275,
+      "grad_norm": 0.45602527260780334,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 0.076,
+      "step": 80
+    },
+    {
+      "epoch": 0.3466476462196862,
+      "grad_norm": 0.3802862763404846,
+      "learning_rate": 4.05e-06,
+      "loss": 0.0776,
+      "step": 81
+    },
+    {
+      "epoch": 0.3509272467902996,
+      "grad_norm": 0.3189656138420105,
+      "learning_rate": 4.1e-06,
+      "loss": 0.0695,
+      "step": 82
+    },
+    {
+      "epoch": 0.35520684736091296,
+      "grad_norm": 0.37464478611946106,
+      "learning_rate": 4.15e-06,
+      "loss": 0.0732,
+      "step": 83
+    },
+    {
+      "epoch": 0.3594864479315264,
+      "grad_norm": 0.5200878977775574,
+      "learning_rate": 4.2000000000000004e-06,
+      "loss": 0.0646,
+      "step": 84
+    },
+    {
+      "epoch": 0.3637660485021398,
+      "grad_norm": 0.3902634382247925,
+      "learning_rate": 4.25e-06,
+      "loss": 0.0751,
+      "step": 85
+    },
+    {
+      "epoch": 0.3680456490727532,
+      "grad_norm": 0.4341444969177246,
+      "learning_rate": 4.3e-06,
+      "loss": 0.0739,
+      "step": 86
+    },
+    {
+      "epoch": 0.3723252496433666,
+      "grad_norm": 0.3288861811161041,
+      "learning_rate": 4.350000000000001e-06,
+      "loss": 0.0631,
+      "step": 87
+    },
+    {
+      "epoch": 0.37660485021398005,
+      "grad_norm": 0.33849674463272095,
+      "learning_rate": 4.4e-06,
+      "loss": 0.0663,
+      "step": 88
+    },
+    {
+      "epoch": 0.38088445078459343,
+      "grad_norm": 0.36165380477905273,
+      "learning_rate": 4.450000000000001e-06,
+      "loss": 0.0618,
+      "step": 89
+    },
+    {
+      "epoch": 0.38516405135520687,
+      "grad_norm": 0.5453753471374512,
+      "learning_rate": 4.5e-06,
+      "loss": 0.0734,
+      "step": 90
+    },
+    {
+      "epoch": 0.38944365192582026,
+      "grad_norm": 0.4055081605911255,
+      "learning_rate": 4.5500000000000005e-06,
+      "loss": 0.0684,
+      "step": 91
+    },
+    {
+      "epoch": 0.39372325249643364,
+      "grad_norm": 0.4177473783493042,
+      "learning_rate": 4.600000000000001e-06,
+      "loss": 0.0605,
+      "step": 92
+    },
+    {
+      "epoch": 0.3980028530670471,
+      "grad_norm": 0.35140708088874817,
+      "learning_rate": 4.65e-06,
+      "loss": 0.0695,
+      "step": 93
+    },
+    {
+      "epoch": 0.40228245363766046,
+      "grad_norm": 0.5472511053085327,
+      "learning_rate": 4.7e-06,
+      "loss": 0.0743,
+      "step": 94
+    },
+    {
+      "epoch": 0.4065620542082739,
+      "grad_norm": 0.35946714878082275,
+      "learning_rate": 4.75e-06,
+      "loss": 0.0658,
+      "step": 95
+    },
+    {
+      "epoch": 0.4108416547788873,
+      "grad_norm": 0.3542482852935791,
+      "learning_rate": 4.800000000000001e-06,
+      "loss": 0.0707,
+      "step": 96
+    },
+    {
+      "epoch": 0.41512125534950073,
+      "grad_norm": 0.5081551671028137,
+      "learning_rate": 4.85e-06,
+      "loss": 0.0774,
+      "step": 97
+    },
+    {
+      "epoch": 0.4194008559201141,
+      "grad_norm": 0.3265087902545929,
+      "learning_rate": 4.9000000000000005e-06,
+      "loss": 0.0715,
+      "step": 98
+    },
+    {
+      "epoch": 0.42368045649072755,
+      "grad_norm": 0.4452495574951172,
+      "learning_rate": 4.95e-06,
+      "loss": 0.064,
+      "step": 99
+    },
+    {
+      "epoch": 0.42796005706134094,
+      "grad_norm": 0.33817920088768005,
+      "learning_rate": 5e-06,
+      "loss": 0.0674,
+      "step": 100
+    },
+    {
+      "epoch": 0.4322396576319544,
+      "grad_norm": 0.4477202296257019,
+      "learning_rate": 4.9999926774868305e-06,
+      "loss": 0.0701,
+      "step": 101
+    },
+    {
+      "epoch": 0.43651925820256776,
+      "grad_norm": 0.3426607847213745,
+      "learning_rate": 4.999970709990216e-06,
+      "loss": 0.0637,
+      "step": 102
+    },
+    {
+      "epoch": 0.44079885877318115,
+      "grad_norm": 0.4300517737865448,
+      "learning_rate": 4.999934097638842e-06,
+      "loss": 0.0716,
+      "step": 103
+    },
+    {
+      "epoch": 0.4450784593437946,
+      "grad_norm": 0.29154735803604126,
+      "learning_rate": 4.9998828406471855e-06,
+      "loss": 0.066,
+      "step": 104
+    },
+    {
+      "epoch": 0.44935805991440797,
+      "grad_norm": 0.38544416427612305,
+      "learning_rate": 4.99981693931551e-06,
+      "loss": 0.0735,
+      "step": 105
+    },
+    {
+      "epoch": 0.4536376604850214,
+      "grad_norm": 0.42879173159599304,
+      "learning_rate": 4.999736394029866e-06,
+      "loss": 0.0724,
+      "step": 106
+    },
+    {
+      "epoch": 0.4579172610556348,
+      "grad_norm": 0.45247069001197815,
+      "learning_rate": 4.999641205262088e-06,
+      "loss": 0.0624,
+      "step": 107
+    },
+    {
+      "epoch": 0.46219686162624823,
+      "grad_norm": 0.45127418637275696,
+      "learning_rate": 4.9995313735697935e-06,
+      "loss": 0.0635,
+      "step": 108
+    },
+    {
+      "epoch": 0.4664764621968616,
+      "grad_norm": 0.4134274125099182,
+      "learning_rate": 4.999406899596378e-06,
+      "loss": 0.0654,
+      "step": 109
+    },
+    {
+      "epoch": 0.47075606276747506,
+      "grad_norm": 0.5010129809379578,
+      "learning_rate": 4.99926778407101e-06,
+      "loss": 0.0678,
+      "step": 110
+    },
+    {
+      "epoch": 0.47503566333808844,
+      "grad_norm": 0.7583813667297363,
+      "learning_rate": 4.999114027808632e-06,
+      "loss": 0.0729,
+      "step": 111
+    },
+    {
+      "epoch": 0.4793152639087018,
+      "grad_norm": 0.39662906527519226,
+      "learning_rate": 4.998945631709948e-06,
+      "loss": 0.0688,
+      "step": 112
+    },
+    {
+      "epoch": 0.48359486447931527,
+      "grad_norm": 0.33845070004463196,
+      "learning_rate": 4.998762596761424e-06,
+      "loss": 0.0689,
+      "step": 113
+    },
+    {
+      "epoch": 0.48787446504992865,
+      "grad_norm": 0.6589358448982239,
+      "learning_rate": 4.998564924035282e-06,
+      "loss": 0.0664,
+      "step": 114
+    },
+    {
+      "epoch": 0.4921540656205421,
+      "grad_norm": 0.4025883078575134,
+      "learning_rate": 4.99835261468949e-06,
+      "loss": 0.0731,
+      "step": 115
+    },
+    {
+      "epoch": 0.4964336661911555,
+      "grad_norm": 0.32868894934654236,
+      "learning_rate": 4.998125669967758e-06,
+      "loss": 0.0657,
+      "step": 116
+    },
+    {
+      "epoch": 0.5007132667617689,
+      "grad_norm": 0.3705316483974457,
+      "learning_rate": 4.997884091199531e-06,
+      "loss": 0.0668,
+      "step": 117
+    },
+    {
+      "epoch": 0.5049928673323824,
+      "grad_norm": 0.32613426446914673,
+      "learning_rate": 4.997627879799981e-06,
+      "loss": 0.0654,
+      "step": 118
+    },
+    {
+      "epoch": 0.5092724679029957,
+      "grad_norm": 0.42335274815559387,
+      "learning_rate": 4.997357037269996e-06,
+      "loss": 0.0695,
+      "step": 119
+    },
+    {
+      "epoch": 0.5135520684736091,
+      "grad_norm": 0.39254000782966614,
+      "learning_rate": 4.997071565196175e-06,
+      "loss": 0.0663,
+      "step": 120
+    },
+    {
+      "epoch": 0.5178316690442225,
+      "grad_norm": 0.33641666173934937,
+      "learning_rate": 4.996771465250814e-06,
+      "loss": 0.0626,
+      "step": 121
+    },
+    {
+      "epoch": 0.5221112696148359,
+      "grad_norm": 0.38578540086746216,
+      "learning_rate": 4.996456739191905e-06,
+      "loss": 0.064,
+      "step": 122
+    },
+    {
+      "epoch": 0.5263908701854494,
+      "grad_norm": 0.31561926007270813,
+      "learning_rate": 4.996127388863116e-06,
+      "loss": 0.0626,
+      "step": 123
+    },
+    {
+      "epoch": 0.5306704707560628,
+      "grad_norm": 0.27469775080680847,
+      "learning_rate": 4.995783416193782e-06,
+      "loss": 0.0617,
+      "step": 124
+    },
+    {
+      "epoch": 0.5349500713266762,
+      "grad_norm": 0.2643572986125946,
+      "learning_rate": 4.9954248231989016e-06,
+      "loss": 0.0592,
+      "step": 125
+    },
+    {
+      "epoch": 0.5392296718972895,
+      "grad_norm": 0.32749131321907043,
+      "learning_rate": 4.995051611979115e-06,
+      "loss": 0.0597,
+      "step": 126
+    },
+    {
+      "epoch": 0.543509272467903,
+      "grad_norm": 0.4380977153778076,
+      "learning_rate": 4.994663784720698e-06,
+      "loss": 0.0663,
+      "step": 127
+    },
+    {
+      "epoch": 0.5477888730385164,
+      "grad_norm": 0.30976516008377075,
+      "learning_rate": 4.994261343695546e-06,
+      "loss": 0.0634,
+      "step": 128
+    },
+    {
+      "epoch": 0.5520684736091298,
+      "grad_norm": 0.4299394488334656,
+      "learning_rate": 4.9938442912611625e-06,
+      "loss": 0.065,
+      "step": 129
+    },
+    {
+      "epoch": 0.5563480741797432,
+      "grad_norm": 0.3520357012748718,
+      "learning_rate": 4.993412629860646e-06,
+      "loss": 0.059,
+      "step": 130
+    },
+    {
+      "epoch": 0.5606276747503567,
+      "grad_norm": 0.44977450370788574,
+      "learning_rate": 4.9929663620226734e-06,
+      "loss": 0.0639,
+      "step": 131
+    },
+    {
+      "epoch": 0.5649072753209701,
+      "grad_norm": 0.3896942436695099,
+      "learning_rate": 4.992505490361487e-06,
+      "loss": 0.0608,
+      "step": 132
+    },
+    {
+      "epoch": 0.5691868758915835,
+      "grad_norm": 0.3307473063468933,
+      "learning_rate": 4.992030017576876e-06,
+      "loss": 0.0649,
+      "step": 133
+    },
+    {
+      "epoch": 0.5734664764621968,
+      "grad_norm": 0.27266785502433777,
+      "learning_rate": 4.991539946454166e-06,
+      "loss": 0.058,
+      "step": 134
+    },
+    {
+      "epoch": 0.5777460770328102,
+      "grad_norm": 0.43257489800453186,
+      "learning_rate": 4.991035279864199e-06,
+      "loss": 0.0651,
+      "step": 135
+    },
+    {
+      "epoch": 0.5820256776034237,
+      "grad_norm": 0.33029407262802124,
+      "learning_rate": 4.990516020763317e-06,
+      "loss": 0.0601,
+      "step": 136
+    },
+    {
+      "epoch": 0.5863052781740371,
+      "grad_norm": 0.40888652205467224,
+      "learning_rate": 4.989982172193346e-06,
+      "loss": 0.0676,
+      "step": 137
+    },
+    {
+      "epoch": 0.5905848787446505,
+      "grad_norm": 0.2740240693092346,
+      "learning_rate": 4.989433737281576e-06,
+      "loss": 0.0599,
+      "step": 138
+    },
+    {
+      "epoch": 0.5948644793152639,
+      "grad_norm": 0.3733709156513214,
+      "learning_rate": 4.988870719240744e-06,
+      "loss": 0.0571,
+      "step": 139
+    },
+    {
+      "epoch": 0.5991440798858774,
+      "grad_norm": 0.33992213010787964,
+      "learning_rate": 4.988293121369016e-06,
+      "loss": 0.0615,
+      "step": 140
+    },
+    {
+      "epoch": 0.6034236804564908,
+      "grad_norm": 0.34187227487564087,
+      "learning_rate": 4.987700947049966e-06,
+      "loss": 0.0625,
+      "step": 141
+    },
+    {
+      "epoch": 0.6077032810271041,
+      "grad_norm": 0.35290244221687317,
+      "learning_rate": 4.987094199752558e-06,
+      "loss": 0.0606,
+      "step": 142
+    },
+    {
+      "epoch": 0.6119828815977175,
+      "grad_norm": 0.2595992982387543,
+      "learning_rate": 4.986472883031124e-06,
+      "loss": 0.0551,
+      "step": 143
+    },
+    {
+      "epoch": 0.6162624821683309,
+      "grad_norm": 0.43302714824676514,
+      "learning_rate": 4.9858370005253435e-06,
+      "loss": 0.0643,
+      "step": 144
+    },
+    {
+      "epoch": 0.6205420827389444,
+      "grad_norm": 0.3430291414260864,
+      "learning_rate": 4.985186555960223e-06,
+      "loss": 0.0576,
+      "step": 145
+    },
+    {
+      "epoch": 0.6248216833095578,
+      "grad_norm": 0.5924923419952393,
+      "learning_rate": 4.984521553146074e-06,
+      "loss": 0.0679,
+      "step": 146
+    },
+    {
+      "epoch": 0.6291012838801712,
+      "grad_norm": 0.3352043330669403,
+      "learning_rate": 4.98384199597849e-06,
+      "loss": 0.0636,
+      "step": 147
+    },
+    {
+      "epoch": 0.6333808844507846,
+      "grad_norm": 0.31009000539779663,
+      "learning_rate": 4.983147888438324e-06,
+      "loss": 0.0499,
+      "step": 148
+    },
+    {
+      "epoch": 0.637660485021398,
+      "grad_norm": 0.41210371255874634,
+      "learning_rate": 4.982439234591665e-06,
+      "loss": 0.0632,
+      "step": 149
+    },
+    {
+      "epoch": 0.6419400855920114,
+      "grad_norm": 0.30768534541130066,
+      "learning_rate": 4.9817160385898145e-06,
+      "loss": 0.0669,
+      "step": 150
+    },
+    {
+      "epoch": 0.6462196861626248,
+      "grad_norm": 0.4492046535015106,
+      "learning_rate": 4.980978304669263e-06,
+      "loss": 0.0603,
+      "step": 151
+    },
+    {
+      "epoch": 0.6504992867332382,
+      "grad_norm": 0.3134528696537018,
+      "learning_rate": 4.9802260371516635e-06,
+      "loss": 0.0553,
+      "step": 152
+    },
+    {
+      "epoch": 0.6547788873038516,
+      "grad_norm": 0.33358660340309143,
+      "learning_rate": 4.979459240443806e-06,
+      "loss": 0.061,
+      "step": 153
+    },
+    {
+      "epoch": 0.6590584878744651,
+      "grad_norm": 0.2705594301223755,
+      "learning_rate": 4.978677919037594e-06,
+      "loss": 0.0525,
+      "step": 154
+    },
+    {
+      "epoch": 0.6633380884450785,
+      "grad_norm": 0.4429916739463806,
+      "learning_rate": 4.977882077510018e-06,
+      "loss": 0.0595,
+      "step": 155
+    },
+    {
+      "epoch": 0.6676176890156919,
+      "grad_norm": 0.37663739919662476,
+      "learning_rate": 4.977071720523125e-06,
+      "loss": 0.0596,
+      "step": 156
+    },
+    {
+      "epoch": 0.6718972895863052,
+      "grad_norm": 0.5795411467552185,
+      "learning_rate": 4.9762468528239945e-06,
+      "loss": 0.0629,
+      "step": 157
+    },
+    {
+      "epoch": 0.6761768901569187,
+      "grad_norm": 0.2995544373989105,
+      "learning_rate": 4.975407479244711e-06,
+      "loss": 0.0577,
+      "step": 158
+    },
+    {
+      "epoch": 0.6804564907275321,
+      "grad_norm": 0.2688761055469513,
+      "learning_rate": 4.974553604702332e-06,
+      "loss": 0.0612,
+      "step": 159
+    },
+    {
+      "epoch": 0.6847360912981455,
+      "grad_norm": 0.32369598746299744,
+      "learning_rate": 4.9736852341988655e-06,
+      "loss": 0.054,
+      "step": 160
+    },
+    {
+      "epoch": 0.6890156918687589,
+      "grad_norm": 0.24816952645778656,
+      "learning_rate": 4.972802372821235e-06,
+      "loss": 0.0631,
+      "step": 161
+    },
+    {
+      "epoch": 0.6932952924393724,
+      "grad_norm": 0.33390095829963684,
+      "learning_rate": 4.97190502574125e-06,
+      "loss": 0.0629,
+      "step": 162
+    },
+    {
+      "epoch": 0.6975748930099858,
+      "grad_norm": 0.2538573443889618,
+      "learning_rate": 4.9709931982155805e-06,
+      "loss": 0.0585,
+      "step": 163
+    },
+    {
+      "epoch": 0.7018544935805991,
+      "grad_norm": 0.337729275226593,
+      "learning_rate": 4.970066895585722e-06,
+      "loss": 0.0534,
+      "step": 164
+    },
+    {
+      "epoch": 0.7061340941512125,
+      "grad_norm": 0.3311474621295929,
+      "learning_rate": 4.9691261232779644e-06,
+      "loss": 0.0592,
+      "step": 165
+    },
+    {
+      "epoch": 0.7104136947218259,
+      "grad_norm": 0.4538433849811554,
+      "learning_rate": 4.968170886803361e-06,
+      "loss": 0.0607,
+      "step": 166
+    },
+    {
+      "epoch": 0.7146932952924394,
+      "grad_norm": 0.3233712911605835,
+      "learning_rate": 4.9672011917577e-06,
+      "loss": 0.0557,
+      "step": 167
+    },
+    {
+      "epoch": 0.7189728958630528,
+      "grad_norm": 0.41257205605506897,
+      "learning_rate": 4.966217043821462e-06,
+      "loss": 0.0618,
+      "step": 168
+    },
+    {
+      "epoch": 0.7232524964336662,
+      "grad_norm": 0.36823365092277527,
+      "learning_rate": 4.965218448759797e-06,
+      "loss": 0.0619,
+      "step": 169
+    },
+    {
+      "epoch": 0.7275320970042796,
+      "grad_norm": 0.38311564922332764,
+      "learning_rate": 4.964205412422487e-06,
+      "loss": 0.0668,
+      "step": 170
+    },
+    {
+      "epoch": 0.7318116975748931,
+      "grad_norm": 0.31038492918014526,
+      "learning_rate": 4.963177940743908e-06,
+      "loss": 0.0573,
+      "step": 171
+    },
+    {
+      "epoch": 0.7360912981455064,
+      "grad_norm": 0.2835666835308075,
+      "learning_rate": 4.962136039742998e-06,
+      "loss": 0.0649,
+      "step": 172
+    },
+    {
+      "epoch": 0.7403708987161198,
+      "grad_norm": 0.4525282680988312,
+      "learning_rate": 4.961079715523228e-06,
+      "loss": 0.0593,
+      "step": 173
+    },
+    {
+      "epoch": 0.7446504992867332,
+      "grad_norm": 0.33394986391067505,
+      "learning_rate": 4.9600089742725545e-06,
+      "loss": 0.0601,
+      "step": 174
+    },
+    {
+      "epoch": 0.7489300998573466,
+      "grad_norm": 0.2642901837825775,
+      "learning_rate": 4.95892382226339e-06,
+      "loss": 0.0623,
+      "step": 175
+    },
+    {
+      "epoch": 0.7532097004279601,
+      "grad_norm": 0.34347841143608093,
+      "learning_rate": 4.95782426585257e-06,
+      "loss": 0.0568,
+      "step": 176
+    },
+    {
+      "epoch": 0.7574893009985735,
+      "grad_norm": 0.30288106203079224,
+      "learning_rate": 4.956710311481303e-06,
+      "loss": 0.0569,
+      "step": 177
+    },
+    {
+      "epoch": 0.7617689015691869,
+      "grad_norm": 0.285087913274765,
+      "learning_rate": 4.955581965675148e-06,
+      "loss": 0.0534,
+      "step": 178
+    },
+    {
+      "epoch": 0.7660485021398002,
+      "grad_norm": 0.2861518859863281,
+      "learning_rate": 4.954439235043966e-06,
+      "loss": 0.0536,
+      "step": 179
+    },
+    {
+      "epoch": 0.7703281027104137,
+      "grad_norm": 0.4016513526439667,
+      "learning_rate": 4.953282126281884e-06,
+      "loss": 0.0575,
+      "step": 180
+    },
+    {
+      "epoch": 0.7746077032810271,
+      "grad_norm": 0.4838986098766327,
+      "learning_rate": 4.9521106461672605e-06,
+      "loss": 0.0666,
+      "step": 181
+    },
+    {
+      "epoch": 0.7788873038516405,
+      "grad_norm": 0.270673006772995,
+      "learning_rate": 4.950924801562636e-06,
+      "loss": 0.0579,
+      "step": 182
+    },
+    {
+      "epoch": 0.7831669044222539,
+      "grad_norm": 0.31113702058792114,
+      "learning_rate": 4.9497245994147e-06,
+      "loss": 0.0546,
+      "step": 183
+    },
+    {
+      "epoch": 0.7874465049928673,
+      "grad_norm": 0.2960825264453888,
+      "learning_rate": 4.948510046754251e-06,
+      "loss": 0.0553,
+      "step": 184
+    },
+    {
+      "epoch": 0.7917261055634808,
+      "grad_norm": 0.3720844089984894,
+      "learning_rate": 4.947281150696151e-06,
+      "loss": 0.0607,
+      "step": 185
+    },
+    {
+      "epoch": 0.7960057061340942,
+      "grad_norm": 0.2569741904735565,
+      "learning_rate": 4.946037918439285e-06,
+      "loss": 0.0552,
+      "step": 186
+    },
+    {
+      "epoch": 0.8002853067047075,
+      "grad_norm": 0.3115193843841553,
+      "learning_rate": 4.944780357266522e-06,
+      "loss": 0.0544,
+      "step": 187
+    },
+    {
+      "epoch": 0.8045649072753209,
+      "grad_norm": 0.2828395366668701,
+      "learning_rate": 4.943508474544667e-06,
+      "loss": 0.0541,
+      "step": 188
+    },
+    {
+      "epoch": 0.8088445078459344,
+      "grad_norm": 0.29760900139808655,
+      "learning_rate": 4.942222277724424e-06,
+      "loss": 0.0576,
+      "step": 189
+    },
+    {
+      "epoch": 0.8131241084165478,
+      "grad_norm": 0.3851959705352783,
+      "learning_rate": 4.940921774340347e-06,
+      "loss": 0.059,
+      "step": 190
+    },
+    {
+      "epoch": 0.8174037089871612,
+      "grad_norm": 0.35299891233444214,
+      "learning_rate": 4.939606972010798e-06,
+      "loss": 0.0627,
+      "step": 191
+    },
+    {
+      "epoch": 0.8216833095577746,
+      "grad_norm": 0.30920645594596863,
+      "learning_rate": 4.9382778784379036e-06,
+      "loss": 0.053,
+      "step": 192
+    },
+    {
+      "epoch": 0.8259629101283881,
+      "grad_norm": 0.31357741355895996,
+      "learning_rate": 4.936934501407507e-06,
+      "loss": 0.0523,
+      "step": 193
+    },
+    {
+      "epoch": 0.8302425106990015,
+      "grad_norm": 0.28295785188674927,
+      "learning_rate": 4.935576848789127e-06,
+      "loss": 0.0521,
+      "step": 194
+    },
+    {
+      "epoch": 0.8345221112696148,
+      "grad_norm": 0.3303717076778412,
+      "learning_rate": 4.934204928535905e-06,
+      "loss": 0.0564,
+      "step": 195
+    },
+    {
+      "epoch": 0.8388017118402282,
+      "grad_norm": 0.31010597944259644,
+      "learning_rate": 4.932818748684565e-06,
+      "loss": 0.0495,
+      "step": 196
+    },
+    {
+      "epoch": 0.8430813124108416,
+      "grad_norm": 0.33654817938804626,
+      "learning_rate": 4.931418317355362e-06,
+      "loss": 0.0536,
+      "step": 197
+    },
+    {
+      "epoch": 0.8473609129814551,
+      "grad_norm": 0.35488635301589966,
+      "learning_rate": 4.9300036427520396e-06,
+      "loss": 0.054,
+      "step": 198
+    },
+    {
+      "epoch": 0.8516405135520685,
+      "grad_norm": 0.37233811616897583,
+      "learning_rate": 4.928574733161775e-06,
+      "loss": 0.0527,
+      "step": 199
+    },
+    {
+      "epoch": 0.8559201141226819,
+      "grad_norm": 0.3253094553947449,
+      "learning_rate": 4.927131596955137e-06,
+      "loss": 0.0545,
+      "step": 200
+    },
+    {
+      "epoch": 0.8601997146932953,
+      "grad_norm": 0.345102995634079,
+      "learning_rate": 4.92567424258603e-06,
+      "loss": 0.0606,
+      "step": 201
+    },
+    {
+      "epoch": 0.8644793152639088,
+      "grad_norm": 0.32971855998039246,
+      "learning_rate": 4.924202678591655e-06,
+      "loss": 0.0542,
+      "step": 202
+    },
+    {
+      "epoch": 0.8687589158345221,
+      "grad_norm": 0.30636993050575256,
+      "learning_rate": 4.922716913592447e-06,
+      "loss": 0.0509,
+      "step": 203
+    },
+    {
+      "epoch": 0.8730385164051355,
+      "grad_norm": 0.27384838461875916,
+      "learning_rate": 4.921216956292033e-06,
+      "loss": 0.0496,
+      "step": 204
+    },
+    {
+      "epoch": 0.8773181169757489,
+      "grad_norm": 0.30968862771987915,
+      "learning_rate": 4.919702815477179e-06,
+      "loss": 0.0608,
+      "step": 205
+    },
+    {
+      "epoch": 0.8815977175463623,
+      "grad_norm": 0.25974956154823303,
+      "learning_rate": 4.918174500017739e-06,
+      "loss": 0.0543,
+      "step": 206
+    },
+    {
+      "epoch": 0.8858773181169758,
+      "grad_norm": 0.30472415685653687,
+      "learning_rate": 4.916632018866599e-06,
+      "loss": 0.0567,
+      "step": 207
+    },
+    {
+      "epoch": 0.8901569186875892,
+      "grad_norm": 0.3091549277305603,
+      "learning_rate": 4.91507538105963e-06,
+      "loss": 0.0511,
+      "step": 208
+    },
+    {
+      "epoch": 0.8944365192582026,
+      "grad_norm": 0.26851367950439453,
+      "learning_rate": 4.9135045957156356e-06,
+      "loss": 0.0536,
+      "step": 209
+    },
+    {
+      "epoch": 0.8987161198288159,
+      "grad_norm": 0.25963643193244934,
+      "learning_rate": 4.911919672036291e-06,
+      "loss": 0.0509,
+      "step": 210
+    },
+    {
+      "epoch": 0.9029957203994294,
+      "grad_norm": 0.28279975056648254,
+      "learning_rate": 4.910320619306094e-06,
+      "loss": 0.0538,
+      "step": 211
+    },
+    {
+      "epoch": 0.9072753209700428,
+      "grad_norm": 0.26387840509414673,
+      "learning_rate": 4.908707446892315e-06,
+      "loss": 0.0504,
+      "step": 212
+    },
+    {
+      "epoch": 0.9115549215406562,
+      "grad_norm": 0.2996926009654999,
+      "learning_rate": 4.907080164244934e-06,
+      "loss": 0.0497,
+      "step": 213
+    },
+    {
+      "epoch": 0.9158345221112696,
+      "grad_norm": 0.4072028398513794,
+      "learning_rate": 4.905438780896589e-06,
+      "loss": 0.053,
+      "step": 214
+    },
+    {
+      "epoch": 0.920114122681883,
+      "grad_norm": 0.5946587324142456,
+      "learning_rate": 4.903783306462522e-06,
+      "loss": 0.0577,
+      "step": 215
+    },
+    {
+      "epoch": 0.9243937232524965,
+      "grad_norm": 0.34546515345573425,
+      "learning_rate": 4.9021137506405205e-06,
+      "loss": 0.0554,
+      "step": 216
+    },
+    {
+      "epoch": 0.9286733238231099,
+      "grad_norm": 0.3417337238788605,
+      "learning_rate": 4.9004301232108585e-06,
+      "loss": 0.0529,
+      "step": 217
+    },
+    {
+      "epoch": 0.9329529243937232,
+      "grad_norm": 0.31937581300735474,
+      "learning_rate": 4.8987324340362445e-06,
+      "loss": 0.0536,
+      "step": 218
+    },
+    {
+      "epoch": 0.9372325249643366,
+      "grad_norm": 0.28485214710235596,
+      "learning_rate": 4.897020693061758e-06,
+      "loss": 0.0488,
+      "step": 219
+    },
+    {
+      "epoch": 0.9415121255349501,
+      "grad_norm": 0.3833935558795929,
+      "learning_rate": 4.895294910314797e-06,
+      "loss": 0.0577,
+      "step": 220
+    },
+    {
+      "epoch": 0.9457917261055635,
+      "grad_norm": 0.2772054076194763,
+      "learning_rate": 4.8935550959050135e-06,
+      "loss": 0.0506,
+      "step": 221
+    },
+    {
+      "epoch": 0.9500713266761769,
+      "grad_norm": 0.35764405131340027,
+      "learning_rate": 4.891801260024262e-06,
+      "loss": 0.0518,
+      "step": 222
+    },
+    {
+      "epoch": 0.9543509272467903,
+      "grad_norm": 0.30019721388816833,
+      "learning_rate": 4.890033412946527e-06,
+      "loss": 0.0525,
+      "step": 223
+    },
+    {
+      "epoch": 0.9586305278174037,
+      "grad_norm": 0.3105527460575104,
+      "learning_rate": 4.888251565027879e-06,
+      "loss": 0.0552,
+      "step": 224
+    },
+    {
+      "epoch": 0.9629101283880172,
+      "grad_norm": 0.35943159461021423,
+      "learning_rate": 4.8864557267064e-06,
+      "loss": 0.0543,
+      "step": 225
+    },
+    {
+      "epoch": 0.9671897289586305,
+      "grad_norm": 0.3295605480670929,
+      "learning_rate": 4.88464590850213e-06,
+      "loss": 0.0565,
+      "step": 226
+    },
+    {
+      "epoch": 0.9714693295292439,
+      "grad_norm": 0.31729599833488464,
+      "learning_rate": 4.882822121017004e-06,
+      "loss": 0.0524,
+      "step": 227
+    },
+    {
+      "epoch": 0.9757489300998573,
+      "grad_norm": 0.4707350432872772,
+      "learning_rate": 4.880984374934788e-06,
+      "loss": 0.0589,
+      "step": 228
+    },
+    {
+      "epoch": 0.9800285306704708,
+      "grad_norm": 0.35285064578056335,
+      "learning_rate": 4.879132681021017e-06,
+      "loss": 0.0509,
+      "step": 229
+    },
+    {
+      "epoch": 0.9843081312410842,
+      "grad_norm": 0.32979413866996765,
+      "learning_rate": 4.877267050122934e-06,
+      "loss": 0.051,
+      "step": 230
+    },
+    {
+      "epoch": 0.9885877318116976,
+      "grad_norm": 0.3415399491786957,
+      "learning_rate": 4.8753874931694254e-06,
+      "loss": 0.0507,
+      "step": 231
+    },
+    {
+      "epoch": 0.992867332382311,
+      "grad_norm": 0.3240598738193512,
+      "learning_rate": 4.8734940211709535e-06,
+      "loss": 0.0524,
+      "step": 232
+    },
+    {
+      "epoch": 0.9971469329529244,
+      "grad_norm": 0.31305283308029175,
+      "learning_rate": 4.871586645219499e-06,
+      "loss": 0.0528,
+      "step": 233
+    },
+    {
+      "epoch": 1.0042796005706134,
+      "grad_norm": 0.5571584105491638,
+      "learning_rate": 4.869665376488491e-06,
+      "loss": 0.1007,
+      "step": 234
+    },
+    {
+      "epoch": 1.0085592011412268,
+      "grad_norm": 0.2822844386100769,
+      "learning_rate": 4.8677302262327386e-06,
+      "loss": 0.0419,
+      "step": 235
+    },
+    {
+      "epoch": 1.0128388017118402,
+      "grad_norm": 0.2870626747608185,
+      "learning_rate": 4.8657812057883755e-06,
+      "loss": 0.0487,
+      "step": 236
+    },
+    {
+      "epoch": 1.0171184022824535,
+      "grad_norm": 0.2784869968891144,
+      "learning_rate": 4.863818326572784e-06,
+      "loss": 0.0393,
+      "step": 237
+    },
+    {
+      "epoch": 1.0213980028530671,
+      "grad_norm": 0.26714813709259033,
+      "learning_rate": 4.8618416000845296e-06,
+      "loss": 0.0476,
+      "step": 238
+    },
+    {
+      "epoch": 1.0256776034236805,
+      "grad_norm": 0.30474796891212463,
+      "learning_rate": 4.859851037903297e-06,
+      "loss": 0.043,
+      "step": 239
+    },
+    {
+      "epoch": 1.029957203994294,
+      "grad_norm": 0.3593360483646393,
+      "learning_rate": 4.857846651689821e-06,
+      "loss": 0.0522,
+      "step": 240
+    },
+    {
+      "epoch": 1.0342368045649073,
+      "grad_norm": 0.359836608171463,
+      "learning_rate": 4.855828453185818e-06,
+      "loss": 0.0536,
+      "step": 241
+    },
+    {
+      "epoch": 1.0385164051355207,
+      "grad_norm": 0.3691597878932953,
+      "learning_rate": 4.853796454213915e-06,
+      "loss": 0.0432,
+      "step": 242
+    },
+    {
+      "epoch": 1.042796005706134,
+      "grad_norm": 0.38553348183631897,
+      "learning_rate": 4.851750666677583e-06,
+      "loss": 0.0444,
+      "step": 243
+    },
+    {
+      "epoch": 1.0470756062767475,
+      "grad_norm": 0.37977030873298645,
+      "learning_rate": 4.849691102561069e-06,
+      "loss": 0.0485,
+      "step": 244
+    },
+    {
+      "epoch": 1.0513552068473608,
+      "grad_norm": 0.3164876699447632,
+      "learning_rate": 4.847617773929319e-06,
+      "loss": 0.0413,
+      "step": 245
+    },
+    {
+      "epoch": 1.0556348074179742,
+      "grad_norm": 0.3686058223247528,
+      "learning_rate": 4.845530692927915e-06,
+      "loss": 0.048,
+      "step": 246
+    },
+    {
+      "epoch": 1.0599144079885878,
+      "grad_norm": 0.3639174699783325,
+      "learning_rate": 4.843429871782999e-06,
+      "loss": 0.0465,
+      "step": 247
+    },
+    {
+      "epoch": 1.0641940085592012,
+      "grad_norm": 0.33737680315971375,
+      "learning_rate": 4.841315322801204e-06,
+      "loss": 0.0417,
+      "step": 248
+    },
+    {
+      "epoch": 1.0684736091298146,
+      "grad_norm": 0.31087926030158997,
+      "learning_rate": 4.8391870583695796e-06,
+      "loss": 0.0449,
+      "step": 249
+    },
+    {
+      "epoch": 1.072753209700428,
+      "grad_norm": 0.3133867383003235,
+      "learning_rate": 4.837045090955522e-06,
+      "loss": 0.0453,
+      "step": 250
+    },
+    {
+      "epoch": 1.0770328102710414,
+      "grad_norm": 0.3388800621032715,
+      "learning_rate": 4.834889433106698e-06,
+      "loss": 0.0415,
+      "step": 251
+    },
+    {
+      "epoch": 1.0813124108416547,
+      "grad_norm": 0.3206201195716858,
+      "learning_rate": 4.8327200974509754e-06,
+      "loss": 0.048,
+      "step": 252
+    },
+    {
+      "epoch": 1.0855920114122681,
+      "grad_norm": 0.3376302719116211,
+      "learning_rate": 4.8305370966963446e-06,
+      "loss": 0.0435,
+      "step": 253
+    },
+    {
+      "epoch": 1.0898716119828815,
+      "grad_norm": 0.3998079001903534,
+      "learning_rate": 4.828340443630847e-06,
+      "loss": 0.0473,
+      "step": 254
+    },
+    {
+      "epoch": 1.094151212553495,
+      "grad_norm": 0.36078789830207825,
+      "learning_rate": 4.826130151122498e-06,
+      "loss": 0.0455,
+      "step": 255
+    },
+    {
+      "epoch": 1.0984308131241085,
+      "grad_norm": 0.3395013213157654,
+      "learning_rate": 4.823906232119217e-06,
+      "loss": 0.0435,
+      "step": 256
+    },
+    {
+      "epoch": 1.102710413694722,
+      "grad_norm": 0.31314757466316223,
+      "learning_rate": 4.8216686996487425e-06,
+      "loss": 0.0418,
+      "step": 257
+    },
+    {
+      "epoch": 1.1069900142653353,
+      "grad_norm": 0.3544161021709442,
+      "learning_rate": 4.8194175668185645e-06,
+      "loss": 0.046,
+      "step": 258
+    },
+    {
+      "epoch": 1.1112696148359487,
+      "grad_norm": 0.3597017228603363,
+      "learning_rate": 4.817152846815843e-06,
+      "loss": 0.0398,
+      "step": 259
+    },
+    {
+      "epoch": 1.115549215406562,
+      "grad_norm": 0.39875292778015137,
+      "learning_rate": 4.814874552907332e-06,
+      "loss": 0.0433,
+      "step": 260
+    },
+    {
+      "epoch": 1.1198288159771754,
+      "grad_norm": 0.42275699973106384,
+      "learning_rate": 4.812582698439299e-06,
+      "loss": 0.0476,
+      "step": 261
+    },
+    {
+      "epoch": 1.1241084165477888,
+      "grad_norm": 0.33560910820961,
+      "learning_rate": 4.810277296837455e-06,
+      "loss": 0.0383,
+      "step": 262
+    },
+    {
+      "epoch": 1.1283880171184022,
+      "grad_norm": 0.4616740047931671,
+      "learning_rate": 4.807958361606864e-06,
+      "loss": 0.0347,
+      "step": 263
+    },
+    {
+      "epoch": 1.1326676176890156,
+      "grad_norm": 0.3887166380882263,
+      "learning_rate": 4.805625906331874e-06,
+      "loss": 0.041,
+      "step": 264
+    },
+    {
+      "epoch": 1.1369472182596292,
+      "grad_norm": 0.4357439875602722,
+      "learning_rate": 4.8032799446760326e-06,
+      "loss": 0.0423,
+      "step": 265
+    },
+    {
+      "epoch": 1.1412268188302426,
+      "grad_norm": 0.3627154231071472,
+      "learning_rate": 4.800920490382008e-06,
+      "loss": 0.0403,
+      "step": 266
+    },
+    {
+      "epoch": 1.145506419400856,
+      "grad_norm": 0.40753647685050964,
+      "learning_rate": 4.798547557271509e-06,
+      "loss": 0.0386,
+      "step": 267
+    },
+    {
+      "epoch": 1.1497860199714693,
+      "grad_norm": 0.4159418046474457,
+      "learning_rate": 4.796161159245202e-06,
+      "loss": 0.0424,
+      "step": 268
+    },
+    {
+      "epoch": 1.1540656205420827,
+      "grad_norm": 0.3464265465736389,
+      "learning_rate": 4.793761310282631e-06,
+      "loss": 0.0365,
+      "step": 269
+    },
+    {
+      "epoch": 1.158345221112696,
+      "grad_norm": 0.517615556716919,
+      "learning_rate": 4.791348024442138e-06,
+      "loss": 0.0429,
+      "step": 270
+    },
+    {
+      "epoch": 1.1626248216833095,
+      "grad_norm": 0.369800329208374,
+      "learning_rate": 4.788921315860776e-06,
+      "loss": 0.0431,
+      "step": 271
+    },
+    {
+      "epoch": 1.1669044222539229,
+      "grad_norm": 0.4237928092479706,
+      "learning_rate": 4.78648119875423e-06,
+      "loss": 0.0374,
+      "step": 272
+    },
+    {
+      "epoch": 1.1711840228245363,
+      "grad_norm": 0.38458287715911865,
+      "learning_rate": 4.784027687416731e-06,
+      "loss": 0.0437,
+      "step": 273
+    },
+    {
+      "epoch": 1.1754636233951499,
+      "grad_norm": 0.4061521589756012,
+      "learning_rate": 4.7815607962209746e-06,
+      "loss": 0.0412,
+      "step": 274
+    },
+    {
+      "epoch": 1.1797432239657633,
+      "grad_norm": 0.3805032968521118,
+      "learning_rate": 4.779080539618036e-06,
+      "loss": 0.0399,
+      "step": 275
+    },
+    {
+      "epoch": 1.1840228245363766,
+      "grad_norm": 0.47470155358314514,
+      "learning_rate": 4.7765869321372835e-06,
+      "loss": 0.0453,
+      "step": 276
+    },
+    {
+      "epoch": 1.18830242510699,
+      "grad_norm": 0.3514023423194885,
+      "learning_rate": 4.7740799883862966e-06,
+      "loss": 0.0377,
+      "step": 277
+    },
+    {
+      "epoch": 1.1925820256776034,
+      "grad_norm": 0.37605416774749756,
+      "learning_rate": 4.771559723050778e-06,
+      "loss": 0.0414,
+      "step": 278
+    },
+    {
+      "epoch": 1.1968616262482168,
+      "grad_norm": 0.3633909821510315,
+      "learning_rate": 4.769026150894468e-06,
+      "loss": 0.0441,
+      "step": 279
+    },
+    {
+      "epoch": 1.2011412268188302,
+      "grad_norm": 0.38254550099372864,
+      "learning_rate": 4.766479286759061e-06,
+      "loss": 0.044,
+      "step": 280
+    },
+    {
+      "epoch": 1.2054208273894436,
+      "grad_norm": 0.4671253263950348,
+      "learning_rate": 4.763919145564111e-06,
+      "loss": 0.0365,
+      "step": 281
+    },
+    {
+      "epoch": 1.209700427960057,
+      "grad_norm": 0.3728236258029938,
+      "learning_rate": 4.761345742306955e-06,
+      "loss": 0.0393,
+      "step": 282
+    },
+    {
+      "epoch": 1.2139800285306706,
+      "grad_norm": 0.4013354480266571,
+      "learning_rate": 4.758759092062615e-06,
+      "loss": 0.0445,
+      "step": 283
+    },
+    {
+      "epoch": 1.218259629101284,
+      "grad_norm": 0.432792603969574,
+      "learning_rate": 4.756159209983714e-06,
+      "loss": 0.0311,
+      "step": 284
+    },
+    {
+      "epoch": 1.2225392296718973,
+      "grad_norm": 0.4125058352947235,
+      "learning_rate": 4.753546111300391e-06,
+      "loss": 0.0322,
+      "step": 285
+    },
+    {
+      "epoch": 1.2268188302425107,
+      "grad_norm": 0.6146723031997681,
+      "learning_rate": 4.750919811320205e-06,
+      "loss": 0.0463,
+      "step": 286
+    },
+    {
+      "epoch": 1.231098430813124,
+      "grad_norm": 0.6433658003807068,
+      "learning_rate": 4.7482803254280485e-06,
+      "loss": 0.0357,
+      "step": 287
+    },
+    {
+      "epoch": 1.2353780313837375,
+      "grad_norm": 0.5013871192932129,
+      "learning_rate": 4.745627669086058e-06,
+      "loss": 0.0365,
+      "step": 288
+    },
+    {
+      "epoch": 1.2396576319543509,
+      "grad_norm": 0.38044819235801697,
+      "learning_rate": 4.742961857833521e-06,
+      "loss": 0.041,
+      "step": 289
+    },
+    {
+      "epoch": 1.2439372325249642,
+      "grad_norm": 0.45295485854148865,
+      "learning_rate": 4.74028290728679e-06,
+      "loss": 0.0461,
+      "step": 290
+    },
+    {
+      "epoch": 1.2482168330955776,
+      "grad_norm": 0.4638918340206146,
+      "learning_rate": 4.737590833139185e-06,
+      "loss": 0.0434,
+      "step": 291
+    },
+    {
+      "epoch": 1.2524964336661912,
+      "grad_norm": 0.5618919730186462,
+      "learning_rate": 4.734885651160903e-06,
+      "loss": 0.0377,
+      "step": 292
+    },
+    {
+      "epoch": 1.2567760342368046,
+      "grad_norm": 0.411558598279953,
+      "learning_rate": 4.732167377198931e-06,
+      "loss": 0.0374,
+      "step": 293
+    },
+    {
+      "epoch": 1.261055634807418,
+      "grad_norm": 0.3951723277568817,
+      "learning_rate": 4.729436027176944e-06,
+      "loss": 0.0453,
+      "step": 294
+    },
+    {
+      "epoch": 1.2653352353780314,
+      "grad_norm": 0.3179916441440582,
+      "learning_rate": 4.726691617095221e-06,
+      "loss": 0.0403,
+      "step": 295
+    },
+    {
+      "epoch": 1.2696148359486448,
+      "grad_norm": 0.4432845413684845,
+      "learning_rate": 4.723934163030545e-06,
+      "loss": 0.0414,
+      "step": 296
+    },
+    {
+      "epoch": 1.2738944365192582,
+      "grad_norm": 0.446893572807312,
+      "learning_rate": 4.72116368113611e-06,
+      "loss": 0.0356,
+      "step": 297
+    },
+    {
+      "epoch": 1.2781740370898715,
+      "grad_norm": 0.3918814957141876,
+      "learning_rate": 4.718380187641429e-06,
+      "loss": 0.0334,
+      "step": 298
+    },
+    {
+      "epoch": 1.2824536376604851,
+      "grad_norm": 0.3894858956336975,
+      "learning_rate": 4.7155836988522364e-06,
+      "loss": 0.0407,
+      "step": 299
+    },
+    {
+      "epoch": 1.2867332382310983,
+      "grad_norm": 0.3865683972835541,
+      "learning_rate": 4.712774231150392e-06,
+      "loss": 0.0374,
+      "step": 300
+    },
+    {
+      "epoch": 1.291012838801712,
+      "grad_norm": 0.35576513409614563,
+      "learning_rate": 4.7099518009937885e-06,
+      "loss": 0.0349,
+      "step": 301
+    },
+    {
+      "epoch": 1.2952924393723253,
+      "grad_norm": 0.40566757321357727,
+      "learning_rate": 4.7071164249162504e-06,
+      "loss": 0.0433,
+      "step": 302
+    },
+    {
+      "epoch": 1.2995720399429387,
+      "grad_norm": 0.4917856454849243,
+      "learning_rate": 4.704268119527441e-06,
+      "loss": 0.0343,
+      "step": 303
+    },
+    {
+      "epoch": 1.303851640513552,
+      "grad_norm": 0.3986580967903137,
+      "learning_rate": 4.701406901512763e-06,
+      "loss": 0.0344,
+      "step": 304
+    },
+    {
+      "epoch": 1.3081312410841655,
+      "grad_norm": 0.43434515595436096,
+      "learning_rate": 4.698532787633262e-06,
+      "loss": 0.0431,
+      "step": 305
+    },
+    {
+      "epoch": 1.3124108416547788,
+      "grad_norm": 0.4825822710990906,
+      "learning_rate": 4.695645794725528e-06,
+      "loss": 0.0292,
+      "step": 306
+    },
+    {
+      "epoch": 1.3166904422253922,
+      "grad_norm": 0.3666658103466034,
+      "learning_rate": 4.692745939701595e-06,
+      "loss": 0.0383,
+      "step": 307
+    },
+    {
+      "epoch": 1.3209700427960058,
+      "grad_norm": 0.3598913848400116,
+      "learning_rate": 4.689833239548844e-06,
+      "loss": 0.0381,
+      "step": 308
+    },
+    {
+      "epoch": 1.325249643366619,
+      "grad_norm": 0.3856128752231598,
+      "learning_rate": 4.686907711329903e-06,
+      "loss": 0.0344,
+      "step": 309
+    },
+    {
+      "epoch": 1.3295292439372326,
+      "grad_norm": 0.422134667634964,
+      "learning_rate": 4.68396937218255e-06,
+      "loss": 0.0411,
+      "step": 310
+    },
+    {
+      "epoch": 1.333808844507846,
+      "grad_norm": 0.422688364982605,
+      "learning_rate": 4.681018239319603e-06,
+      "loss": 0.0349,
+      "step": 311
+    },
+    {
+      "epoch": 1.3380884450784594,
+      "grad_norm": 0.37932994961738586,
+      "learning_rate": 4.678054330028831e-06,
+      "loss": 0.0325,
+      "step": 312
+    },
+    {
+      "epoch": 1.3423680456490727,
+      "grad_norm": 0.3773266673088074,
+      "learning_rate": 4.675077661672845e-06,
+      "loss": 0.033,
+      "step": 313
+    },
+    {
+      "epoch": 1.3466476462196861,
+      "grad_norm": 0.39197105169296265,
+      "learning_rate": 4.6720882516890016e-06,
+      "loss": 0.0358,
+      "step": 314
+    },
+    {
+      "epoch": 1.3509272467902995,
+      "grad_norm": 0.4106651544570923,
+      "learning_rate": 4.669086117589295e-06,
+      "loss": 0.0322,
+      "step": 315
+    },
+    {
+      "epoch": 1.355206847360913,
+      "grad_norm": 0.4139370918273926,
+      "learning_rate": 4.666071276960256e-06,
+      "loss": 0.0306,
+      "step": 316
+    },
+    {
+      "epoch": 1.3594864479315265,
+      "grad_norm": 0.39681828022003174,
+      "learning_rate": 4.663043747462857e-06,
+      "loss": 0.0346,
+      "step": 317
+    },
+    {
+      "epoch": 1.3637660485021397,
+      "grad_norm": 0.39305955171585083,
+      "learning_rate": 4.660003546832395e-06,
+      "loss": 0.027,
+      "step": 318
+    },
+    {
+      "epoch": 1.3680456490727533,
+      "grad_norm": 0.41277015209198,
+      "learning_rate": 4.656950692878397e-06,
+      "loss": 0.0322,
+      "step": 319
+    },
+    {
+      "epoch": 1.3723252496433667,
+      "grad_norm": 0.4968262016773224,
+      "learning_rate": 4.653885203484516e-06,
+      "loss": 0.0315,
+      "step": 320
+    },
+    {
+      "epoch": 1.37660485021398,
+      "grad_norm": 0.5269052386283875,
+      "learning_rate": 4.650807096608419e-06,
+      "loss": 0.0365,
+      "step": 321
+    },
+    {
+      "epoch": 1.3808844507845934,
+      "grad_norm": 0.3427152633666992,
+      "learning_rate": 4.647716390281689e-06,
+      "loss": 0.0254,
+      "step": 322
+    },
+    {
+      "epoch": 1.3851640513552068,
+      "grad_norm": 0.42712923884391785,
+      "learning_rate": 4.644613102609716e-06,
+      "loss": 0.0337,
+      "step": 323
+    },
+    {
+      "epoch": 1.3894436519258202,
+      "grad_norm": 0.4049385190010071,
+      "learning_rate": 4.641497251771593e-06,
+      "loss": 0.0308,
+      "step": 324
+    },
+    {
+      "epoch": 1.3937232524964336,
+      "grad_norm": 0.36787325143814087,
+      "learning_rate": 4.638368856020006e-06,
+      "loss": 0.0353,
+      "step": 325
+    },
+    {
+      "epoch": 1.3980028530670472,
+      "grad_norm": 0.43888741731643677,
+      "learning_rate": 4.635227933681131e-06,
+      "loss": 0.0281,
+      "step": 326
+    },
+    {
+      "epoch": 1.4022824536376604,
+      "grad_norm": 0.40592697262763977,
+      "learning_rate": 4.632074503154524e-06,
+      "loss": 0.0294,
+      "step": 327
+    },
+    {
+      "epoch": 1.406562054208274,
+      "grad_norm": 0.3631339371204376,
+      "learning_rate": 4.628908582913013e-06,
+      "loss": 0.0298,
+      "step": 328
+    },
+    {
+      "epoch": 1.4108416547788873,
+      "grad_norm": 0.5272939205169678,
+      "learning_rate": 4.625730191502594e-06,
+      "loss": 0.0354,
+      "step": 329
+    },
+    {
+      "epoch": 1.4151212553495007,
+      "grad_norm": 0.47585079073905945,
+      "learning_rate": 4.622539347542316e-06,
+      "loss": 0.0341,
+      "step": 330
+    },
+    {
+      "epoch": 1.4194008559201141,
+      "grad_norm": 0.4558964967727661,
+      "learning_rate": 4.619336069724177e-06,
+      "loss": 0.0363,
+      "step": 331
+    },
+    {
+      "epoch": 1.4236804564907275,
+      "grad_norm": 0.43466460704803467,
+      "learning_rate": 4.6161203768130135e-06,
+      "loss": 0.0367,
+      "step": 332
+    },
+    {
+      "epoch": 1.4279600570613409,
+      "grad_norm": 0.4535815715789795,
+      "learning_rate": 4.612892287646386e-06,
+      "loss": 0.0331,
+      "step": 333
+    },
+    {
+      "epoch": 1.4322396576319543,
+      "grad_norm": 0.43253347277641296,
+      "learning_rate": 4.609651821134476e-06,
+      "loss": 0.028,
+      "step": 334
+    },
+    {
+      "epoch": 1.4365192582025679,
+      "grad_norm": 0.5027002096176147,
+      "learning_rate": 4.606398996259971e-06,
+      "loss": 0.0326,
+      "step": 335
+    },
+    {
+      "epoch": 1.440798858773181,
+      "grad_norm": 0.5224930644035339,
+      "learning_rate": 4.603133832077953e-06,
+      "loss": 0.0336,
+      "step": 336
+    },
+    {
+      "epoch": 1.4450784593437946,
+      "grad_norm": 0.4419247508049011,
+      "learning_rate": 4.5998563477157885e-06,
+      "loss": 0.0324,
+      "step": 337
+    },
+    {
+      "epoch": 1.449358059914408,
+      "grad_norm": 0.44416379928588867,
+      "learning_rate": 4.5965665623730145e-06,
+      "loss": 0.0383,
+      "step": 338
+    },
+    {
+      "epoch": 1.4536376604850214,
+      "grad_norm": 0.5683357119560242,
+      "learning_rate": 4.593264495321228e-06,
+      "loss": 0.0319,
+      "step": 339
+    },
+    {
+      "epoch": 1.4579172610556348,
+      "grad_norm": 0.4201441705226898,
+      "learning_rate": 4.589950165903974e-06,
+      "loss": 0.0343,
+      "step": 340
+    },
+    {
+      "epoch": 1.4621968616262482,
+      "grad_norm": 0.568689227104187,
+      "learning_rate": 4.586623593536629e-06,
+      "loss": 0.0322,
+      "step": 341
+    },
+    {
+      "epoch": 1.4664764621968616,
+      "grad_norm": 0.4224129617214203,
+      "learning_rate": 4.583284797706288e-06,
+      "loss": 0.0277,
+      "step": 342
+    },
+    {
+      "epoch": 1.470756062767475,
+      "grad_norm": 0.47905975580215454,
+      "learning_rate": 4.579933797971652e-06,
+      "loss": 0.037,
+      "step": 343
+    },
+    {
+      "epoch": 1.4750356633380886,
+      "grad_norm": 0.6059530377388,
+      "learning_rate": 4.576570613962914e-06,
+      "loss": 0.0412,
+      "step": 344
+    },
+    {
+      "epoch": 1.4793152639087017,
+      "grad_norm": 0.4478548765182495,
+      "learning_rate": 4.573195265381641e-06,
+      "loss": 0.0261,
+      "step": 345
+    },
+    {
+      "epoch": 1.4835948644793153,
+      "grad_norm": 0.4309764802455902,
+      "learning_rate": 4.5698077720006595e-06,
+      "loss": 0.0298,
+      "step": 346
+    },
+    {
+      "epoch": 1.4878744650499287,
+      "grad_norm": 0.4392475187778473,
+      "learning_rate": 4.566408153663943e-06,
+      "loss": 0.0222,
+      "step": 347
+    },
+    {
+      "epoch": 1.492154065620542,
+      "grad_norm": 0.39503395557403564,
+      "learning_rate": 4.5629964302864896e-06,
+      "loss": 0.0293,
+      "step": 348
+    },
+    {
+      "epoch": 1.4964336661911555,
+      "grad_norm": 0.46291348338127136,
+      "learning_rate": 4.5595726218542124e-06,
+      "loss": 0.029,
+      "step": 349
+    },
+    {
+      "epoch": 1.5007132667617689,
+      "grad_norm": 0.41241031885147095,
+      "learning_rate": 4.5561367484238165e-06,
+      "loss": 0.0332,
+      "step": 350
+    },
+    {
+      "epoch": 1.5049928673323825,
+      "grad_norm": 0.5383005142211914,
+      "learning_rate": 4.552688830122685e-06,
+      "loss": 0.0312,
+      "step": 351
+    },
+    {
+      "epoch": 1.5092724679029956,
+      "grad_norm": 0.5965086817741394,
+      "learning_rate": 4.54922888714876e-06,
+      "loss": 0.0333,
+      "step": 352
+    },
+    {
+      "epoch": 1.5135520684736092,
+      "grad_norm": 0.43624430894851685,
+      "learning_rate": 4.5457569397704226e-06,
+      "loss": 0.0254,
+      "step": 353
+    },
+    {
+      "epoch": 1.5178316690442224,
+      "grad_norm": 0.43263742327690125,
+      "learning_rate": 4.542273008326379e-06,
+      "loss": 0.023,
+      "step": 354
+    },
+    {
+      "epoch": 1.522111269614836,
+      "grad_norm": 0.4612899720668793,
+      "learning_rate": 4.538777113225535e-06,
+      "loss": 0.0237,
+      "step": 355
+    },
+    {
+      "epoch": 1.5263908701854494,
+      "grad_norm": 0.49297091364860535,
+      "learning_rate": 4.535269274946881e-06,
+      "loss": 0.0344,
+      "step": 356
+    },
+    {
+      "epoch": 1.5306704707560628,
+      "grad_norm": 0.5402700304985046,
+      "learning_rate": 4.531749514039371e-06,
+      "loss": 0.025,
+      "step": 357
+    },
+    {
+      "epoch": 1.5349500713266762,
+      "grad_norm": 0.4386838972568512,
+      "learning_rate": 4.5282178511218006e-06,
+      "loss": 0.026,
+      "step": 358
+    },
+    {
+      "epoch": 1.5392296718972895,
+      "grad_norm": 0.37339386343955994,
+      "learning_rate": 4.5246743068826895e-06,
+      "loss": 0.0232,
+      "step": 359
+    },
+    {
+      "epoch": 1.5435092724679031,
+      "grad_norm": 0.5581195950508118,
+      "learning_rate": 4.521118902080156e-06,
+      "loss": 0.0356,
+      "step": 360
+    },
+    {
+      "epoch": 1.5477888730385163,
+      "grad_norm": 0.5439732074737549,
+      "learning_rate": 4.517551657541801e-06,
+      "loss": 0.0313,
+      "step": 361
+    },
+    {
+      "epoch": 1.55206847360913,
+      "grad_norm": 0.4505188465118408,
+      "learning_rate": 4.513972594164578e-06,
+      "loss": 0.0154,
+      "step": 362
+    },
+    {
+      "epoch": 1.556348074179743,
+      "grad_norm": 0.44269970059394836,
+      "learning_rate": 4.510381732914681e-06,
+      "loss": 0.024,
+      "step": 363
+    },
+    {
+      "epoch": 1.5606276747503567,
+      "grad_norm": 1.0567827224731445,
+      "learning_rate": 4.506779094827409e-06,
+      "loss": 0.0238,
+      "step": 364
+    },
+    {
+      "epoch": 1.56490727532097,
+      "grad_norm": 0.47431910037994385,
+      "learning_rate": 4.503164701007058e-06,
+      "loss": 0.0343,
+      "step": 365
+    },
+    {
+      "epoch": 1.5691868758915835,
+      "grad_norm": 0.3802037835121155,
+      "learning_rate": 4.499538572626782e-06,
+      "loss": 0.0192,
+      "step": 366
+    },
+    {
+      "epoch": 1.5734664764621968,
+      "grad_norm": 0.37891197204589844,
+      "learning_rate": 4.495900730928481e-06,
+      "loss": 0.0243,
+      "step": 367
+    },
+    {
+      "epoch": 1.5777460770328102,
+      "grad_norm": 0.5435581207275391,
+      "learning_rate": 4.492251197222669e-06,
+      "loss": 0.0267,
+      "step": 368
+    },
+    {
+      "epoch": 1.5820256776034238,
+      "grad_norm": 0.5308198928833008,
+      "learning_rate": 4.4885899928883536e-06,
+      "loss": 0.0297,
+      "step": 369
+    },
+    {
+      "epoch": 1.586305278174037,
+      "grad_norm": 0.5913228988647461,
+      "learning_rate": 4.4849171393729084e-06,
+      "loss": 0.0279,
+      "step": 370
+    },
+    {
+      "epoch": 1.5905848787446506,
+      "grad_norm": 0.43857505917549133,
+      "learning_rate": 4.481232658191947e-06,
+      "loss": 0.0243,
+      "step": 371
+    },
+    {
+      "epoch": 1.5948644793152638,
+      "grad_norm": 0.41668838262557983,
+      "learning_rate": 4.4775365709292004e-06,
+      "loss": 0.0279,
+      "step": 372
+    },
+    {
+      "epoch": 1.5991440798858774,
+      "grad_norm": 0.3665734827518463,
+      "learning_rate": 4.473828899236385e-06,
+      "loss": 0.0257,
+      "step": 373
+    },
+    {
+      "epoch": 1.6034236804564908,
+      "grad_norm": 0.3707963526248932,
+      "learning_rate": 4.470109664833083e-06,
+      "loss": 0.02,
+      "step": 374
+    },
+    {
+      "epoch": 1.6077032810271041,
+      "grad_norm": 0.3789198100566864,
+      "learning_rate": 4.466378889506607e-06,
+      "loss": 0.022,
+      "step": 375
+    },
+    {
+      "epoch": 1.6119828815977175,
+      "grad_norm": 0.3439970910549164,
+      "learning_rate": 4.462636595111878e-06,
+      "loss": 0.0159,
+      "step": 376
+    },
+    {
+      "epoch": 1.616262482168331,
+      "grad_norm": 0.4878711700439453,
+      "learning_rate": 4.458882803571297e-06,
+      "loss": 0.0231,
+      "step": 377
+    },
+    {
+      "epoch": 1.6205420827389445,
+      "grad_norm": 0.4537769854068756,
+      "learning_rate": 4.455117536874613e-06,
+      "loss": 0.0262,
+      "step": 378
+    },
+    {
+      "epoch": 1.6248216833095577,
+      "grad_norm": 0.537686824798584,
+      "learning_rate": 4.4513408170788e-06,
+      "loss": 0.0224,
+      "step": 379
+    },
+    {
+      "epoch": 1.6291012838801713,
+      "grad_norm": 0.4968257248401642,
+      "learning_rate": 4.447552666307921e-06,
+      "loss": 0.0251,
+      "step": 380
+    },
+    {
+      "epoch": 1.6333808844507844,
+      "grad_norm": 0.5357342958450317,
+      "learning_rate": 4.443753106753003e-06,
+      "loss": 0.0219,
+      "step": 381
+    },
+    {
+      "epoch": 1.637660485021398,
+      "grad_norm": 0.4279462695121765,
+      "learning_rate": 4.4399421606719064e-06,
+      "loss": 0.0279,
+      "step": 382
+    },
+    {
+      "epoch": 1.6419400855920114,
+      "grad_norm": 0.8014388084411621,
+      "learning_rate": 4.436119850389193e-06,
+      "loss": 0.0237,
+      "step": 383
+    },
+    {
+      "epoch": 1.6462196861626248,
+      "grad_norm": 0.6127387881278992,
+      "learning_rate": 4.432286198295998e-06,
+      "loss": 0.0256,
+      "step": 384
+    },
+    {
+      "epoch": 1.6504992867332382,
+      "grad_norm": 0.3907901346683502,
+      "learning_rate": 4.428441226849894e-06,
+      "loss": 0.0221,
+      "step": 385
+    },
+    {
+      "epoch": 1.6547788873038516,
+      "grad_norm": 0.4546869695186615,
+      "learning_rate": 4.424584958574766e-06,
+      "loss": 0.0226,
+      "step": 386
+    },
+    {
+      "epoch": 1.6590584878744652,
+      "grad_norm": 0.4860124886035919,
+      "learning_rate": 4.4207174160606725e-06,
+      "loss": 0.023,
+      "step": 387
+    },
+    {
+      "epoch": 1.6633380884450784,
+      "grad_norm": 0.34656503796577454,
+      "learning_rate": 4.416838621963719e-06,
+      "loss": 0.0247,
+      "step": 388
+    },
+    {
+      "epoch": 1.667617689015692,
+      "grad_norm": 0.42502695322036743,
+      "learning_rate": 4.412948599005924e-06,
+      "loss": 0.0224,
+      "step": 389
+    },
+    {
+      "epoch": 1.6718972895863051,
+      "grad_norm": 0.3963342010974884,
+      "learning_rate": 4.409047369975079e-06,
+      "loss": 0.0338,
+      "step": 390
+    },
+    {
+      "epoch": 1.6761768901569187,
+      "grad_norm": 0.4617026150226593,
+      "learning_rate": 4.405134957724627e-06,
+      "loss": 0.0206,
+      "step": 391
+    },
+    {
+      "epoch": 1.6804564907275321,
+      "grad_norm": 0.36610788106918335,
+      "learning_rate": 4.401211385173522e-06,
+      "loss": 0.0276,
+      "step": 392
+    },
+    {
+      "epoch": 1.6847360912981455,
+      "grad_norm": 0.3438093066215515,
+      "learning_rate": 4.39727667530609e-06,
+      "loss": 0.0236,
+      "step": 393
+    },
+    {
+      "epoch": 1.6890156918687589,
+      "grad_norm": 0.48957353830337524,
+      "learning_rate": 4.3933308511719055e-06,
+      "loss": 0.0214,
+      "step": 394
+    },
+    {
+      "epoch": 1.6932952924393723,
+      "grad_norm": 0.31086310744285583,
+      "learning_rate": 4.3893739358856465e-06,
+      "loss": 0.0176,
+      "step": 395
+    },
+    {
+      "epoch": 1.6975748930099859,
+      "grad_norm": 0.4718981981277466,
+      "learning_rate": 4.3854059526269644e-06,
+      "loss": 0.0262,
+      "step": 396
+    },
+    {
+      "epoch": 1.701854493580599,
+      "grad_norm": 0.5084271430969238,
+      "learning_rate": 4.381426924640346e-06,
+      "loss": 0.0198,
+      "step": 397
+    },
+    {
+      "epoch": 1.7061340941512126,
+      "grad_norm": 0.39318305253982544,
+      "learning_rate": 4.377436875234982e-06,
+      "loss": 0.0203,
+      "step": 398
+    },
+    {
+      "epoch": 1.7104136947218258,
+      "grad_norm": 0.4365154802799225,
+      "learning_rate": 4.373435827784623e-06,
+      "loss": 0.0168,
+      "step": 399
+    },
+    {
+      "epoch": 1.7146932952924394,
+      "grad_norm": 0.5564635992050171,
+      "learning_rate": 4.369423805727444e-06,
+      "loss": 0.0227,
+      "step": 400
+    },
+    {
+      "epoch": 1.7189728958630528,
+      "grad_norm": 0.5515264868736267,
+      "learning_rate": 4.365400832565916e-06,
+      "loss": 0.0191,
+      "step": 401
+    },
+    {
+      "epoch": 1.7232524964336662,
+      "grad_norm": 0.8078024983406067,
+      "learning_rate": 4.361366931866657e-06,
+      "loss": 0.0284,
+      "step": 402
+    },
+    {
+      "epoch": 1.7275320970042796,
+      "grad_norm": 0.5032513737678528,
+      "learning_rate": 4.357322127260299e-06,
+      "loss": 0.0236,
+      "step": 403
+    },
+    {
+      "epoch": 1.731811697574893,
+      "grad_norm": 1.0713094472885132,
+      "learning_rate": 4.353266442441352e-06,
+      "loss": 0.0245,
+      "step": 404
+    },
+    {
+      "epoch": 1.7360912981455066,
+      "grad_norm": 0.5070701241493225,
+      "learning_rate": 4.349199901168058e-06,
+      "loss": 0.0215,
+      "step": 405
+    },
+    {
+      "epoch": 1.7403708987161197,
+      "grad_norm": 0.46457090973854065,
+      "learning_rate": 4.345122527262261e-06,
+      "loss": 0.018,
+      "step": 406
+    },
+    {
+      "epoch": 1.7446504992867333,
+      "grad_norm": 0.4174697995185852,
+      "learning_rate": 4.341034344609259e-06,
+      "loss": 0.0185,
+      "step": 407
+    },
+    {
+      "epoch": 1.7489300998573465,
+      "grad_norm": 0.4278315007686615,
+      "learning_rate": 4.336935377157668e-06,
+      "loss": 0.0213,
+      "step": 408
+    },
+    {
+      "epoch": 1.75320970042796,
+      "grad_norm": 0.7992134690284729,
+      "learning_rate": 4.332825648919286e-06,
+      "loss": 0.0185,
+      "step": 409
+    },
+    {
+      "epoch": 1.7574893009985735,
+      "grad_norm": 0.5143779516220093,
+      "learning_rate": 4.32870518396894e-06,
+      "loss": 0.0209,
+      "step": 410
+    },
+    {
+      "epoch": 1.7617689015691869,
+      "grad_norm": 0.4616805911064148,
+      "learning_rate": 4.3245740064443586e-06,
+      "loss": 0.02,
+      "step": 411
+    },
+    {
+      "epoch": 1.7660485021398002,
+      "grad_norm": 0.3600054979324341,
+      "learning_rate": 4.320432140546026e-06,
+      "loss": 0.019,
+      "step": 412
+    },
+    {
+      "epoch": 1.7703281027104136,
+      "grad_norm": 0.4483279585838318,
+      "learning_rate": 4.316279610537032e-06,
+      "loss": 0.0172,
+      "step": 413
+    },
+    {
+      "epoch": 1.7746077032810272,
+      "grad_norm": 0.5091086626052856,
+      "learning_rate": 4.312116440742944e-06,
+      "loss": 0.0271,
+      "step": 414
+    },
+    {
+      "epoch": 1.7788873038516404,
+      "grad_norm": 0.4678874611854553,
+      "learning_rate": 4.307942655551653e-06,
+      "loss": 0.0152,
+      "step": 415
+    },
+    {
+      "epoch": 1.783166904422254,
+      "grad_norm": 0.44703155755996704,
+      "learning_rate": 4.303758279413239e-06,
+      "loss": 0.0216,
+      "step": 416
+    },
+    {
+      "epoch": 1.7874465049928672,
+      "grad_norm": 0.44851160049438477,
+      "learning_rate": 4.299563336839818e-06,
+      "loss": 0.021,
+      "step": 417
+    },
+    {
+      "epoch": 1.7917261055634808,
+      "grad_norm": 0.4456520974636078,
+      "learning_rate": 4.2953578524054116e-06,
+      "loss": 0.0162,
+      "step": 418
+    },
+    {
+      "epoch": 1.7960057061340942,
+      "grad_norm": 0.5209571719169617,
+      "learning_rate": 4.291141850745788e-06,
+      "loss": 0.0193,
+      "step": 419
+    },
+    {
+      "epoch": 1.8002853067047075,
+      "grad_norm": 0.5911619067192078,
+      "learning_rate": 4.286915356558332e-06,
+      "loss": 0.019,
+      "step": 420
+    },
+    {
+      "epoch": 1.804564907275321,
+      "grad_norm": 0.5851296186447144,
+      "learning_rate": 4.2826783946018905e-06,
+      "loss": 0.0141,
+      "step": 421
+    },
+    {
+      "epoch": 1.8088445078459343,
+      "grad_norm": 0.3961900770664215,
+      "learning_rate": 4.278430989696631e-06,
+      "loss": 0.0255,
+      "step": 422
+    },
+    {
+      "epoch": 1.813124108416548,
+      "grad_norm": 0.7789033055305481,
+      "learning_rate": 4.274173166723897e-06,
+      "loss": 0.0124,
+      "step": 423
+    },
+    {
+      "epoch": 1.817403708987161,
+      "grad_norm": 0.4361071288585663,
+      "learning_rate": 4.269904950626059e-06,
+      "loss": 0.0157,
+      "step": 424
+    },
+    {
+      "epoch": 1.8216833095577747,
+      "grad_norm": 0.43065768480300903,
+      "learning_rate": 4.265626366406372e-06,
+      "loss": 0.0223,
+      "step": 425
+    },
+    {
+      "epoch": 1.825962910128388,
+      "grad_norm": 0.6288639307022095,
+      "learning_rate": 4.261337439128829e-06,
+      "loss": 0.0133,
+      "step": 426
+    },
+    {
+      "epoch": 1.8302425106990015,
+      "grad_norm": 0.4564136862754822,
+      "learning_rate": 4.257038193918011e-06,
+      "loss": 0.014,
+      "step": 427
+    },
+    {
+      "epoch": 1.8345221112696148,
+      "grad_norm": 0.6153598427772522,
+      "learning_rate": 4.252728655958939e-06,
+      "loss": 0.0158,
+      "step": 428
+    },
+    {
+      "epoch": 1.8388017118402282,
+      "grad_norm": 0.5261392593383789,
+      "learning_rate": 4.248408850496934e-06,
+      "loss": 0.0159,
+      "step": 429
+    },
+    {
+      "epoch": 1.8430813124108416,
+      "grad_norm": 0.5919463038444519,
+      "learning_rate": 4.244078802837462e-06,
+      "loss": 0.0201,
+      "step": 430
+    },
+    {
+      "epoch": 1.847360912981455,
+      "grad_norm": 0.4920569360256195,
+      "learning_rate": 4.239738538345988e-06,
+      "loss": 0.0169,
+      "step": 431
+    },
+    {
+      "epoch": 1.8516405135520686,
+      "grad_norm": 0.5132091641426086,
+      "learning_rate": 4.235388082447825e-06,
+      "loss": 0.0179,
+      "step": 432
+    },
+    {
+      "epoch": 1.8559201141226818,
+      "grad_norm": 0.49452340602874756,
+      "learning_rate": 4.231027460627992e-06,
+      "loss": 0.0166,
+      "step": 433
+    },
+    {
+      "epoch": 1.8601997146932954,
+      "grad_norm": 0.4602709412574768,
+      "learning_rate": 4.226656698431057e-06,
+      "loss": 0.0196,
+      "step": 434
+    },
+    {
+      "epoch": 1.8644793152639088,
+      "grad_norm": 0.3728005290031433,
+      "learning_rate": 4.22227582146099e-06,
+      "loss": 0.0167,
+      "step": 435
+    },
+    {
+      "epoch": 1.8687589158345221,
+      "grad_norm": 0.4013611674308777,
+      "learning_rate": 4.217884855381015e-06,
+      "loss": 0.019,
+      "step": 436
+    },
+    {
+      "epoch": 1.8730385164051355,
+      "grad_norm": 0.4389547109603882,
+      "learning_rate": 4.213483825913457e-06,
+      "loss": 0.0193,
+      "step": 437
+    },
+    {
+      "epoch": 1.877318116975749,
+      "grad_norm": 0.3908598721027374,
+      "learning_rate": 4.209072758839594e-06,
+      "loss": 0.012,
+      "step": 438
+    },
+    {
+      "epoch": 1.8815977175463623,
+      "grad_norm": 0.7128845453262329,
+      "learning_rate": 4.2046516799995034e-06,
+      "loss": 0.0169,
+      "step": 439
+    },
+    {
+      "epoch": 1.8858773181169757,
+      "grad_norm": 0.5817690491676331,
+      "learning_rate": 4.200220615291911e-06,
+      "loss": 0.0191,
+      "step": 440
+    },
+    {
+      "epoch": 1.8901569186875893,
+      "grad_norm": 0.7214891910552979,
+      "learning_rate": 4.1957795906740404e-06,
+      "loss": 0.0149,
+      "step": 441
+    },
+    {
+      "epoch": 1.8944365192582024,
+      "grad_norm": 0.5022482872009277,
+      "learning_rate": 4.191328632161462e-06,
+      "loss": 0.0179,
+      "step": 442
+    },
+    {
+      "epoch": 1.898716119828816,
+      "grad_norm": 0.4441869258880615,
+      "learning_rate": 4.186867765827935e-06,
+      "loss": 0.012,
+      "step": 443
+    },
+    {
+      "epoch": 1.9029957203994294,
+      "grad_norm": 0.493295282125473,
+      "learning_rate": 4.182397017805264e-06,
+      "loss": 0.0125,
+      "step": 444
+    },
+    {
+      "epoch": 1.9072753209700428,
+      "grad_norm": 0.4610918462276459,
+      "learning_rate": 4.177916414283137e-06,
+      "loss": 0.0099,
+      "step": 445
+    },
+    {
+      "epoch": 1.9115549215406562,
+      "grad_norm": 0.44065865874290466,
+      "learning_rate": 4.173425981508976e-06,
+      "loss": 0.0146,
+      "step": 446
+    },
+    {
+      "epoch": 1.9158345221112696,
+      "grad_norm": 0.415574312210083,
+      "learning_rate": 4.168925745787784e-06,
+      "loss": 0.013,
+      "step": 447
+    },
+    {
+      "epoch": 1.920114122681883,
+      "grad_norm": 0.5749611258506775,
+      "learning_rate": 4.16441573348199e-06,
+      "loss": 0.0124,
+      "step": 448
+    },
+    {
+      "epoch": 1.9243937232524964,
+      "grad_norm": 0.49864593148231506,
+      "learning_rate": 4.159895971011293e-06,
+      "loss": 0.016,
+      "step": 449
+    },
+    {
+      "epoch": 1.92867332382311,
+      "grad_norm": 0.7332733273506165,
+      "learning_rate": 4.155366484852509e-06,
+      "loss": 0.0129,
+      "step": 450
+    },
+    {
+      "epoch": 1.9329529243937231,
+      "grad_norm": 0.4229061007499695,
+      "learning_rate": 4.150827301539416e-06,
+      "loss": 0.0137,
+      "step": 451
+    },
+    {
+      "epoch": 1.9372325249643367,
+      "grad_norm": 0.4221555292606354,
+      "learning_rate": 4.146278447662597e-06,
+      "loss": 0.0124,
+      "step": 452
+    },
+    {
+      "epoch": 1.9415121255349501,
+      "grad_norm": 0.324710875749588,
+      "learning_rate": 4.141719949869287e-06,
+      "loss": 0.0109,
+      "step": 453
+    },
+    {
+      "epoch": 1.9457917261055635,
+      "grad_norm": 0.5981960296630859,
+      "learning_rate": 4.137151834863213e-06,
+      "loss": 0.0143,
+      "step": 454
+    },
+    {
+      "epoch": 1.9500713266761769,
+      "grad_norm": 0.9933749437332153,
+      "learning_rate": 4.132574129404442e-06,
+      "loss": 0.0157,
+      "step": 455
+    },
+    {
+      "epoch": 1.9543509272467903,
+      "grad_norm": 0.524457573890686,
+      "learning_rate": 4.127986860309219e-06,
+      "loss": 0.0133,
+      "step": 456
+    },
+    {
+      "epoch": 1.9586305278174037,
+      "grad_norm": 0.6651455760002136,
+      "learning_rate": 4.123390054449818e-06,
+      "loss": 0.0133,
+      "step": 457
+    },
+    {
+      "epoch": 1.962910128388017,
+      "grad_norm": 0.6450986266136169,
+      "learning_rate": 4.118783738754372e-06,
+      "loss": 0.0184,
+      "step": 458
+    },
+    {
+      "epoch": 1.9671897289586306,
+      "grad_norm": 0.5285755395889282,
+      "learning_rate": 4.114167940206729e-06,
+      "loss": 0.0095,
+      "step": 459
+    },
+    {
+      "epoch": 1.9714693295292438,
+      "grad_norm": 0.5533472895622253,
+      "learning_rate": 4.109542685846287e-06,
+      "loss": 0.018,
+      "step": 460
+    },
+    {
+      "epoch": 1.9757489300998574,
+      "grad_norm": 0.5881748199462891,
+      "learning_rate": 4.104908002767832e-06,
+      "loss": 0.0128,
+      "step": 461
+    },
+    {
+      "epoch": 1.9800285306704708,
+      "grad_norm": 0.637514054775238,
+      "learning_rate": 4.100263918121388e-06,
+      "loss": 0.0121,
+      "step": 462
+    },
+    {
+      "epoch": 1.9843081312410842,
+      "grad_norm": 1.390764594078064,
+      "learning_rate": 4.095610459112051e-06,
+      "loss": 0.0154,
+      "step": 463
+    },
+    {
+      "epoch": 1.9885877318116976,
+      "grad_norm": 0.6760232448577881,
+      "learning_rate": 4.090947652999832e-06,
+      "loss": 0.015,
+      "step": 464
+    },
+    {
+      "epoch": 1.992867332382311,
+      "grad_norm": 1.089862585067749,
+      "learning_rate": 4.086275527099501e-06,
+      "loss": 0.0114,
+      "step": 465
+    },
+    {
+      "epoch": 1.9971469329529246,
+      "grad_norm": 0.4834727346897125,
+      "learning_rate": 4.08159410878042e-06,
+      "loss": 0.0089,
+      "step": 466
+    },
+    {
+      "epoch": 2.0042796005706136,
+      "grad_norm": 1.0610970258712769,
+      "learning_rate": 4.0769034254663856e-06,
+      "loss": 0.0273,
+      "step": 467
+    },
+    {
+      "epoch": 2.0085592011412268,
+      "grad_norm": 0.3145875036716461,
+      "learning_rate": 4.072203504635469e-06,
+      "loss": 0.0078,
+      "step": 468
+    },
+    {
+      "epoch": 2.0128388017118404,
+      "grad_norm": 0.37535548210144043,
+      "learning_rate": 4.0674943738198595e-06,
+      "loss": 0.0126,
+      "step": 469
+    },
+    {
+      "epoch": 2.0171184022824535,
+      "grad_norm": 0.4032490849494934,
+      "learning_rate": 4.062776060605692e-06,
+      "loss": 0.0143,
+      "step": 470
+    },
+    {
+      "epoch": 2.021398002853067,
+      "grad_norm": 0.8238989114761353,
+      "learning_rate": 4.058048592632896e-06,
+      "loss": 0.0131,
+      "step": 471
+    },
+    {
+      "epoch": 2.0256776034236803,
+      "grad_norm": 0.5088585615158081,
+      "learning_rate": 4.053311997595029e-06,
+      "loss": 0.0094,
+      "step": 472
+    },
+    {
+      "epoch": 2.029957203994294,
+      "grad_norm": 0.3052437901496887,
+      "learning_rate": 4.048566303239115e-06,
+      "loss": 0.0065,
+      "step": 473
+    },
+    {
+      "epoch": 2.034236804564907,
+      "grad_norm": 0.3434632420539856,
+      "learning_rate": 4.04381153736548e-06,
+      "loss": 0.0121,
+      "step": 474
+    },
+    {
+      "epoch": 2.0385164051355207,
+      "grad_norm": 0.40094271302223206,
+      "learning_rate": 4.039047727827594e-06,
+      "loss": 0.009,
+      "step": 475
+    },
+    {
+      "epoch": 2.0427960057061343,
+      "grad_norm": 0.5058080554008484,
+      "learning_rate": 4.034274902531903e-06,
+      "loss": 0.0095,
+      "step": 476
+    },
+    {
+      "epoch": 2.0470756062767475,
+      "grad_norm": 0.33334746956825256,
+      "learning_rate": 4.029493089437669e-06,
+      "loss": 0.0092,
+      "step": 477
+    },
+    {
+      "epoch": 2.051355206847361,
+      "grad_norm": 0.3802107572555542,
+      "learning_rate": 4.0247023165568e-06,
+      "loss": 0.0064,
+      "step": 478
+    },
+    {
+      "epoch": 2.055634807417974,
+      "grad_norm": 0.36711373925209045,
+      "learning_rate": 4.019902611953698e-06,
+      "loss": 0.0085,
+      "step": 479
+    },
+    {
+      "epoch": 2.059914407988588,
+      "grad_norm": 0.4605230987071991,
+      "learning_rate": 4.015094003745082e-06,
+      "loss": 0.0069,
+      "step": 480
+    },
+    {
+      "epoch": 2.064194008559201,
+      "grad_norm": 0.5190231800079346,
+      "learning_rate": 4.010276520099827e-06,
+      "loss": 0.0112,
+      "step": 481
+    },
+    {
+      "epoch": 2.0684736091298146,
+      "grad_norm": 1.0071794986724854,
+      "learning_rate": 4.005450189238806e-06,
+      "loss": 0.0048,
+      "step": 482
+    },
+    {
+      "epoch": 2.0727532097004278,
+      "grad_norm": 0.4913085103034973,
+      "learning_rate": 4.000615039434715e-06,
+      "loss": 0.0099,
+      "step": 483
+    },
+    {
+      "epoch": 2.0770328102710414,
+      "grad_norm": 0.7372696995735168,
+      "learning_rate": 3.995771099011914e-06,
+      "loss": 0.0053,
+      "step": 484
+    },
+    {
+      "epoch": 2.081312410841655,
+      "grad_norm": 0.23670224845409393,
+      "learning_rate": 3.990918396346254e-06,
+      "loss": 0.0054,
+      "step": 485
+    },
+    {
+      "epoch": 2.085592011412268,
+      "grad_norm": 0.7463693022727966,
+      "learning_rate": 3.98605695986492e-06,
+      "loss": 0.0058,
+      "step": 486
+    },
+    {
+      "epoch": 2.0898716119828817,
+      "grad_norm": 0.5333644151687622,
+      "learning_rate": 3.9811868180462585e-06,
+      "loss": 0.0059,
+      "step": 487
+    },
+    {
+      "epoch": 2.094151212553495,
+      "grad_norm": 0.5247513055801392,
+      "learning_rate": 3.976307999419611e-06,
+      "loss": 0.008,
+      "step": 488
+    },
+    {
+      "epoch": 2.0984308131241085,
+      "grad_norm": 0.6092849373817444,
+      "learning_rate": 3.971420532565148e-06,
+      "loss": 0.004,
+      "step": 489
+    },
+    {
+      "epoch": 2.1027104136947217,
+      "grad_norm": 0.48461002111434937,
+      "learning_rate": 3.966524446113703e-06,
+      "loss": 0.0096,
+      "step": 490
+    },
+    {
+      "epoch": 2.1069900142653353,
+      "grad_norm": 0.6382566690444946,
+      "learning_rate": 3.961619768746602e-06,
+      "loss": 0.0083,
+      "step": 491
+    },
+    {
+      "epoch": 2.1112696148359484,
+      "grad_norm": 0.5699125528335571,
+      "learning_rate": 3.956706529195494e-06,
+      "loss": 0.0035,
+      "step": 492
+    },
+    {
+      "epoch": 2.115549215406562,
+      "grad_norm": 0.31096431612968445,
+      "learning_rate": 3.9517847562421915e-06,
+      "loss": 0.0035,
+      "step": 493
+    },
+    {
+      "epoch": 2.1198288159771757,
+      "grad_norm": 0.35952478647232056,
+      "learning_rate": 3.946854478718491e-06,
+      "loss": 0.0078,
+      "step": 494
+    },
+    {
+      "epoch": 2.124108416547789,
+      "grad_norm": 0.45269617438316345,
+      "learning_rate": 3.94191572550601e-06,
+      "loss": 0.0057,
+      "step": 495
+    },
+    {
+      "epoch": 2.1283880171184024,
+      "grad_norm": 0.3647656738758087,
+      "learning_rate": 3.936968525536018e-06,
+      "loss": 0.0068,
+      "step": 496
+    },
+    {
+      "epoch": 2.1326676176890156,
+      "grad_norm": 0.4928695559501648,
+      "learning_rate": 3.932012907789262e-06,
+      "loss": 0.0074,
+      "step": 497
+    },
+    {
+      "epoch": 2.136947218259629,
+      "grad_norm": 0.6422083973884583,
+      "learning_rate": 3.927048901295805e-06,
+      "loss": 0.0064,
+      "step": 498
+    },
+    {
+      "epoch": 2.1412268188302424,
+      "grad_norm": 0.32455766201019287,
+      "learning_rate": 3.922076535134849e-06,
+      "loss": 0.0037,
+      "step": 499
+    },
+    {
+      "epoch": 2.145506419400856,
+      "grad_norm": 0.4020403027534485,
+      "learning_rate": 3.917095838434567e-06,
+      "loss": 0.0056,
+      "step": 500
+    },
+    {
+      "epoch": 2.1497860199714696,
+      "grad_norm": 0.40988966822624207,
+      "learning_rate": 3.9121068403719335e-06,
+      "loss": 0.0034,
+      "step": 501
+    },
+    {
+      "epoch": 2.1540656205420827,
+      "grad_norm": 0.31292110681533813,
+      "learning_rate": 3.90710957017255e-06,
+      "loss": 0.007,
+      "step": 502
+    },
+    {
+      "epoch": 2.1583452211126963,
+      "grad_norm": 0.4979025721549988,
+      "learning_rate": 3.90210405711048e-06,
+      "loss": 0.005,
+      "step": 503
+    },
+    {
+      "epoch": 2.1626248216833095,
+      "grad_norm": 0.8948659896850586,
+      "learning_rate": 3.897090330508071e-06,
+      "loss": 0.0077,
+      "step": 504
+    },
+    {
+      "epoch": 2.166904422253923,
+      "grad_norm": 0.5055385231971741,
+      "learning_rate": 3.8920684197357865e-06,
+      "loss": 0.0072,
+      "step": 505
+    },
+    {
+      "epoch": 2.1711840228245363,
+      "grad_norm": 0.36552977561950684,
+      "learning_rate": 3.887038354212031e-06,
+      "loss": 0.0056,
+      "step": 506
+    },
+    {
+      "epoch": 2.17546362339515,
+      "grad_norm": 0.5923222303390503,
+      "learning_rate": 3.882000163402984e-06,
+      "loss": 0.0094,
+      "step": 507
+    },
+    {
+      "epoch": 2.179743223965763,
+      "grad_norm": 0.4993934631347656,
+      "learning_rate": 3.876953876822418e-06,
+      "loss": 0.0047,
+      "step": 508
+    },
+    {
+      "epoch": 2.1840228245363766,
+      "grad_norm": 0.3998502790927887,
+      "learning_rate": 3.871899524031535e-06,
+      "loss": 0.0103,
+      "step": 509
+    },
+    {
+      "epoch": 2.18830242510699,
+      "grad_norm": 0.4463786482810974,
+      "learning_rate": 3.866837134638786e-06,
+      "loss": 0.0036,
+      "step": 510
+    },
+    {
+      "epoch": 2.1925820256776034,
+      "grad_norm": 0.45763128995895386,
+      "learning_rate": 3.8617667382997e-06,
+      "loss": 0.0057,
+      "step": 511
+    },
+    {
+      "epoch": 2.196861626248217,
+      "grad_norm": 0.4340563118457794,
+      "learning_rate": 3.856688364716715e-06,
+      "loss": 0.0048,
+      "step": 512
+    },
+    {
+      "epoch": 2.20114122681883,
+      "grad_norm": 0.6045655608177185,
+      "learning_rate": 3.8516020436389945e-06,
+      "loss": 0.0064,
+      "step": 513
+    },
+    {
+      "epoch": 2.205420827389444,
+      "grad_norm": 0.42897504568099976,
+      "learning_rate": 3.846507804862262e-06,
+      "loss": 0.0079,
+      "step": 514
+    },
+    {
+      "epoch": 2.209700427960057,
+      "grad_norm": 0.557436466217041,
+      "learning_rate": 3.8414056782286225e-06,
+      "loss": 0.005,
+      "step": 515
+    },
+    {
+      "epoch": 2.2139800285306706,
+      "grad_norm": 0.3902797996997833,
+      "learning_rate": 3.836295693626386e-06,
+      "loss": 0.004,
+      "step": 516
+    },
+    {
+      "epoch": 2.2182596291012837,
+      "grad_norm": 0.4710390567779541,
+      "learning_rate": 3.831177880989897e-06,
+      "loss": 0.0076,
+      "step": 517
+    },
+    {
+      "epoch": 2.2225392296718973,
+      "grad_norm": 0.4178374707698822,
+      "learning_rate": 3.826052270299356e-06,
+      "loss": 0.0081,
+      "step": 518
+    },
+    {
+      "epoch": 2.226818830242511,
+      "grad_norm": 0.7133017778396606,
+      "learning_rate": 3.820918891580645e-06,
+      "loss": 0.0086,
+      "step": 519
+    },
+    {
+      "epoch": 2.231098430813124,
+      "grad_norm": 0.7762539386749268,
+      "learning_rate": 3.81577777490515e-06,
+      "loss": 0.0057,
+      "step": 520
+    },
+    {
+      "epoch": 2.2353780313837377,
+      "grad_norm": 0.6291279792785645,
+      "learning_rate": 3.8106289503895866e-06,
+      "loss": 0.0047,
+      "step": 521
+    },
+    {
+      "epoch": 2.239657631954351,
+      "grad_norm": 0.3010478913784027,
+      "learning_rate": 3.805472448195822e-06,
+      "loss": 0.0064,
+      "step": 522
+    },
+    {
+      "epoch": 2.2439372325249645,
+      "grad_norm": 0.4451744258403778,
+      "learning_rate": 3.8003082985307022e-06,
+      "loss": 0.0049,
+      "step": 523
+    },
+    {
+      "epoch": 2.2482168330955776,
+      "grad_norm": 0.444017231464386,
+      "learning_rate": 3.7951365316458698e-06,
+      "loss": 0.0077,
+      "step": 524
+    },
+    {
+      "epoch": 2.2524964336661912,
+      "grad_norm": 0.4439545273780823,
+      "learning_rate": 3.7899571778375886e-06,
+      "loss": 0.0068,
+      "step": 525
+    },
+    {
+      "epoch": 2.2567760342368044,
+      "grad_norm": 0.43541809916496277,
+      "learning_rate": 3.7847702674465693e-06,
+      "loss": 0.0079,
+      "step": 526
+    },
+    {
+      "epoch": 2.261055634807418,
+      "grad_norm": 0.3825772702693939,
+      "learning_rate": 3.779575830857787e-06,
+      "loss": 0.0082,
+      "step": 527
+    },
+    {
+      "epoch": 2.265335235378031,
+      "grad_norm": 0.24448992311954498,
+      "learning_rate": 3.7743738985003053e-06,
+      "loss": 0.0062,
+      "step": 528
+    },
+    {
+      "epoch": 2.2696148359486448,
+      "grad_norm": 0.8010409474372864,
+      "learning_rate": 3.7691645008471e-06,
+      "loss": 0.0056,
+      "step": 529
+    },
+    {
+      "epoch": 2.2738944365192584,
+      "grad_norm": 0.2623003125190735,
+      "learning_rate": 3.763947668414877e-06,
+      "loss": 0.0038,
+      "step": 530
+    },
+    {
+      "epoch": 2.2781740370898715,
+      "grad_norm": 0.25390884280204773,
+      "learning_rate": 3.7587234317638955e-06,
+      "loss": 0.007,
+      "step": 531
+    },
+    {
+      "epoch": 2.282453637660485,
+      "grad_norm": 0.5253174901008606,
+      "learning_rate": 3.753491821497788e-06,
+      "loss": 0.006,
+      "step": 532
+    },
+    {
+      "epoch": 2.2867332382310983,
+      "grad_norm": 0.37569546699523926,
+      "learning_rate": 3.7482528682633844e-06,
+      "loss": 0.0071,
+      "step": 533
+    },
+    {
+      "epoch": 2.291012838801712,
+      "grad_norm": 0.32741880416870117,
+      "learning_rate": 3.7430066027505267e-06,
+      "loss": 0.0045,
+      "step": 534
+    },
+    {
+      "epoch": 2.295292439372325,
+      "grad_norm": 0.6359190344810486,
+      "learning_rate": 3.7377530556918934e-06,
+      "loss": 0.0066,
+      "step": 535
+    },
+    {
+      "epoch": 2.2995720399429387,
+      "grad_norm": 0.5190404057502747,
+      "learning_rate": 3.732492257862819e-06,
+      "loss": 0.0054,
+      "step": 536
+    },
+    {
+      "epoch": 2.3038516405135523,
+      "grad_norm": 0.5946906805038452,
+      "learning_rate": 3.727224240081113e-06,
+      "loss": 0.011,
+      "step": 537
+    },
+    {
+      "epoch": 2.3081312410841655,
+      "grad_norm": 0.6893029808998108,
+      "learning_rate": 3.7219490332068774e-06,
+      "loss": 0.0063,
+      "step": 538
+    },
+    {
+      "epoch": 2.312410841654779,
+      "grad_norm": 0.43083977699279785,
+      "learning_rate": 3.7166666681423326e-06,
+      "loss": 0.0024,
+      "step": 539
+    },
+    {
+      "epoch": 2.316690442225392,
+      "grad_norm": 0.44732487201690674,
+      "learning_rate": 3.711377175831626e-06,
+      "loss": 0.0054,
+      "step": 540
+    },
+    {
+      "epoch": 2.320970042796006,
+      "grad_norm": 0.3756791055202484,
+      "learning_rate": 3.7060805872606604e-06,
+      "loss": 0.0057,
+      "step": 541
+    },
+    {
+      "epoch": 2.325249643366619,
+      "grad_norm": 0.5223040580749512,
+      "learning_rate": 3.7007769334569072e-06,
+      "loss": 0.0042,
+      "step": 542
+    },
+    {
+      "epoch": 2.3295292439372326,
+      "grad_norm": 0.30271726846694946,
+      "learning_rate": 3.6954662454892266e-06,
+      "loss": 0.0019,
+      "step": 543
+    },
+    {
+      "epoch": 2.3338088445078458,
+      "grad_norm": 0.23703891038894653,
+      "learning_rate": 3.6901485544676848e-06,
+      "loss": 0.002,
+      "step": 544
+    },
+    {
+      "epoch": 2.3380884450784594,
+      "grad_norm": 0.6809226274490356,
+      "learning_rate": 3.684823891543371e-06,
+      "loss": 0.0098,
+      "step": 545
+    },
+    {
+      "epoch": 2.3423680456490725,
+      "grad_norm": 1.772580862045288,
+      "learning_rate": 3.6794922879082185e-06,
+      "loss": 0.0053,
+      "step": 546
+    },
+    {
+      "epoch": 2.346647646219686,
+      "grad_norm": 0.5708749294281006,
+      "learning_rate": 3.6741537747948148e-06,
+      "loss": 0.0038,
+      "step": 547
+    },
+    {
+      "epoch": 2.3509272467902997,
+      "grad_norm": 0.43098023533821106,
+      "learning_rate": 3.6688083834762277e-06,
+      "loss": 0.0034,
+      "step": 548
+    },
+    {
+      "epoch": 2.355206847360913,
+      "grad_norm": 0.4059015214443207,
+      "learning_rate": 3.663456145265817e-06,
+      "loss": 0.0021,
+      "step": 549
+    },
+    {
+      "epoch": 2.3594864479315265,
+      "grad_norm": 0.2948651611804962,
+      "learning_rate": 3.658097091517048e-06,
+      "loss": 0.0034,
+      "step": 550
+    },
+    {
+      "epoch": 2.3637660485021397,
+      "grad_norm": 0.5517255067825317,
+      "learning_rate": 3.652731253623315e-06,
+      "loss": 0.0072,
+      "step": 551
+    },
+    {
+      "epoch": 2.3680456490727533,
+      "grad_norm": 0.4522925615310669,
+      "learning_rate": 3.647358663017754e-06,
+      "loss": 0.0076,
+      "step": 552
+    },
+    {
+      "epoch": 2.3723252496433664,
+      "grad_norm": 0.6154627203941345,
+      "learning_rate": 3.6419793511730554e-06,
+      "loss": 0.0072,
+      "step": 553
+    },
+    {
+      "epoch": 2.37660485021398,
+      "grad_norm": 0.5896870493888855,
+      "learning_rate": 3.636593349601286e-06,
+      "loss": 0.0073,
+      "step": 554
+    },
+    {
+      "epoch": 2.3808844507845937,
+      "grad_norm": 0.5177700519561768,
+      "learning_rate": 3.6312006898537007e-06,
+      "loss": 0.0057,
+      "step": 555
+    },
+    {
+      "epoch": 2.385164051355207,
+      "grad_norm": 0.3875308632850647,
+      "learning_rate": 3.6258014035205553e-06,
+      "loss": 0.0033,
+      "step": 556
+    },
+    {
+      "epoch": 2.3894436519258204,
+      "grad_norm": 0.4099850654602051,
+      "learning_rate": 3.6203955222309263e-06,
+      "loss": 0.0039,
+      "step": 557
+    },
+    {
+      "epoch": 2.3937232524964336,
+      "grad_norm": 0.3532715439796448,
+      "learning_rate": 3.6149830776525237e-06,
+      "loss": 0.0035,
+      "step": 558
+    },
+    {
+      "epoch": 2.398002853067047,
+      "grad_norm": 0.3508007526397705,
+      "learning_rate": 3.609564101491505e-06,
+      "loss": 0.0034,
+      "step": 559
+    },
+    {
+      "epoch": 2.4022824536376604,
+      "grad_norm": 0.5390321612358093,
+      "learning_rate": 3.6041386254922885e-06,
+      "loss": 0.0059,
+      "step": 560
+    },
+    {
+      "epoch": 2.406562054208274,
+      "grad_norm": 0.18945489823818207,
+      "learning_rate": 3.5987066814373718e-06,
+      "loss": 0.0055,
+      "step": 561
+    },
+    {
+      "epoch": 2.410841654778887,
+      "grad_norm": 0.3890082836151123,
+      "learning_rate": 3.593268301147139e-06,
+      "loss": 0.0026,
+      "step": 562
+    },
+    {
+      "epoch": 2.4151212553495007,
+      "grad_norm": 0.32187849283218384,
+      "learning_rate": 3.587823516479679e-06,
+      "loss": 0.0042,
+      "step": 563
+    },
+    {
+      "epoch": 2.419400855920114,
+      "grad_norm": 0.19139395654201508,
+      "learning_rate": 3.582372359330599e-06,
+      "loss": 0.0032,
+      "step": 564
+    },
+    {
+      "epoch": 2.4236804564907275,
+      "grad_norm": 0.8424011468887329,
+      "learning_rate": 3.576914861632833e-06,
+      "loss": 0.0062,
+      "step": 565
+    },
+    {
+      "epoch": 2.427960057061341,
+      "grad_norm": 0.3709321916103363,
+      "learning_rate": 3.5714510553564624e-06,
+      "loss": 0.0036,
+      "step": 566
+    },
+    {
+      "epoch": 2.4322396576319543,
+      "grad_norm": 0.8795949220657349,
+      "learning_rate": 3.5659809725085203e-06,
+      "loss": 0.0043,
+      "step": 567
+    },
+    {
+      "epoch": 2.436519258202568,
+      "grad_norm": 0.44260433316230774,
+      "learning_rate": 3.5605046451328095e-06,
+      "loss": 0.0034,
+      "step": 568
+    },
+    {
+      "epoch": 2.440798858773181,
+      "grad_norm": 0.36288049817085266,
+      "learning_rate": 3.555022105309715e-06,
+      "loss": 0.0042,
+      "step": 569
+    },
+    {
+      "epoch": 2.4450784593437946,
+      "grad_norm": 0.5439409613609314,
+      "learning_rate": 3.5495333851560115e-06,
+      "loss": 0.0021,
+      "step": 570
+    },
+    {
+      "epoch": 2.449358059914408,
+      "grad_norm": 0.4087218940258026,
+      "learning_rate": 3.5440385168246796e-06,
+      "loss": 0.0026,
+      "step": 571
+    },
+    {
+      "epoch": 2.4536376604850214,
+      "grad_norm": 0.2950667142868042,
+      "learning_rate": 3.5385375325047167e-06,
+      "loss": 0.0051,
+      "step": 572
+    },
+    {
+      "epoch": 2.457917261055635,
+      "grad_norm": 0.258934885263443,
+      "learning_rate": 3.5330304644209456e-06,
+      "loss": 0.0023,
+      "step": 573
+    },
+    {
+      "epoch": 2.462196861626248,
+      "grad_norm": 0.25606122612953186,
+      "learning_rate": 3.5275173448338306e-06,
+      "loss": 0.0044,
+      "step": 574
+    },
+    {
+      "epoch": 2.466476462196862,
+      "grad_norm": 0.39192715287208557,
+      "learning_rate": 3.5219982060392832e-06,
+      "loss": 0.0026,
+      "step": 575
+    },
+    {
+      "epoch": 2.470756062767475,
+      "grad_norm": 0.21482016146183014,
+      "learning_rate": 3.516473080368478e-06,
+      "loss": 0.0049,
+      "step": 576
+    },
+    {
+      "epoch": 2.4750356633380886,
+      "grad_norm": 0.3853050768375397,
+      "learning_rate": 3.5109420001876576e-06,
+      "loss": 0.0024,
+      "step": 577
+    },
+    {
+      "epoch": 2.4793152639087017,
+      "grad_norm": 0.46906372904777527,
+      "learning_rate": 3.5054049978979487e-06,
+      "loss": 0.0042,
+      "step": 578
+    },
+    {
+      "epoch": 2.4835948644793153,
+      "grad_norm": 0.40924930572509766,
+      "learning_rate": 3.4998621059351695e-06,
+      "loss": 0.0047,
+      "step": 579
+    },
+    {
+      "epoch": 2.4878744650499285,
+      "grad_norm": 0.7026729583740234,
+      "learning_rate": 3.49431335676964e-06,
+      "loss": 0.006,
+      "step": 580
+    },
+    {
+      "epoch": 2.492154065620542,
+      "grad_norm": 0.23798321187496185,
+      "learning_rate": 3.4887587829059887e-06,
+      "loss": 0.0023,
+      "step": 581
+    },
+    {
+      "epoch": 2.4964336661911553,
+      "grad_norm": 0.6722045540809631,
+      "learning_rate": 3.4831984168829706e-06,
+      "loss": 0.0035,
+      "step": 582
+    },
+    {
+      "epoch": 2.500713266761769,
+      "grad_norm": 0.429177463054657,
+      "learning_rate": 3.477632291273268e-06,
+      "loss": 0.0032,
+      "step": 583
+    },
+    {
+      "epoch": 2.5049928673323825,
+      "grad_norm": 0.20870012044906616,
+      "learning_rate": 3.4720604386833024e-06,
+      "loss": 0.002,
+      "step": 584
+    },
+    {
+      "epoch": 2.5092724679029956,
+      "grad_norm": 0.23582929372787476,
+      "learning_rate": 3.4664828917530445e-06,
+      "loss": 0.0021,
+      "step": 585
+    },
+    {
+      "epoch": 2.5135520684736092,
+      "grad_norm": 0.3133554458618164,
+      "learning_rate": 3.460899683155825e-06,
+      "loss": 0.0042,
+      "step": 586
+    },
+    {
+      "epoch": 2.5178316690442224,
+      "grad_norm": 0.2973891496658325,
+      "learning_rate": 3.455310845598137e-06,
+      "loss": 0.0024,
+      "step": 587
+    },
+    {
+      "epoch": 2.522111269614836,
+      "grad_norm": 0.29642683267593384,
+      "learning_rate": 3.4497164118194503e-06,
+      "loss": 0.0046,
+      "step": 588
+    },
+    {
+      "epoch": 2.526390870185449,
+      "grad_norm": 0.495253324508667,
+      "learning_rate": 3.4441164145920174e-06,
+      "loss": 0.0017,
+      "step": 589
+    },
+    {
+      "epoch": 2.5306704707560628,
+      "grad_norm": 0.20187705755233765,
+      "learning_rate": 3.4385108867206803e-06,
+      "loss": 0.0035,
+      "step": 590
+    },
+    {
+      "epoch": 2.5349500713266764,
+      "grad_norm": 0.46516257524490356,
+      "learning_rate": 3.4328998610426807e-06,
+      "loss": 0.0018,
+      "step": 591
+    },
+    {
+      "epoch": 2.5392296718972895,
+      "grad_norm": 0.3390999734401703,
+      "learning_rate": 3.427283370427466e-06,
+      "loss": 0.0073,
+      "step": 592
+    },
+    {
+      "epoch": 2.543509272467903,
+      "grad_norm": 0.34318435192108154,
+      "learning_rate": 3.4216614477764973e-06,
+      "loss": 0.0029,
+      "step": 593
+    },
+    {
+      "epoch": 2.5477888730385163,
+      "grad_norm": 0.22326043248176575,
+      "learning_rate": 3.416034126023058e-06,
+      "loss": 0.0038,
+      "step": 594
+    },
+    {
+      "epoch": 2.55206847360913,
+      "grad_norm": 0.2754570543766022,
+      "learning_rate": 3.410401438132056e-06,
+      "loss": 0.002,
+      "step": 595
+    },
+    {
+      "epoch": 2.556348074179743,
+      "grad_norm": 0.5922912955284119,
+      "learning_rate": 3.4047634170998374e-06,
+      "loss": 0.0029,
+      "step": 596
+    },
+    {
+      "epoch": 2.5606276747503567,
+      "grad_norm": 0.5190448760986328,
+      "learning_rate": 3.3991200959539894e-06,
+      "loss": 0.003,
+      "step": 597
+    },
+    {
+      "epoch": 2.5649072753209703,
+      "grad_norm": 1.955246925354004,
+      "learning_rate": 3.393471507753146e-06,
+      "loss": 0.0043,
+      "step": 598
+    },
+    {
+      "epoch": 2.5691868758915835,
+      "grad_norm": 0.2827969193458557,
+      "learning_rate": 3.3878176855867974e-06,
+      "loss": 0.0014,
+      "step": 599
+    },
+    {
+      "epoch": 2.5734664764621966,
+      "grad_norm": 0.13652905821800232,
+      "learning_rate": 3.3821586625750915e-06,
+      "loss": 0.0041,
+      "step": 600
+    },
+    {
+      "epoch": 2.5777460770328102,
+      "grad_norm": 1.3642975091934204,
+      "learning_rate": 3.3764944718686456e-06,
+      "loss": 0.0012,
+      "step": 601
+    },
+    {
+      "epoch": 2.582025677603424,
+      "grad_norm": 0.39340218901634216,
+      "learning_rate": 3.37082514664835e-06,
+      "loss": 0.0036,
+      "step": 602
+    },
+    {
+      "epoch": 2.586305278174037,
+      "grad_norm": 0.2879283130168915,
+      "learning_rate": 3.3651507201251685e-06,
+      "loss": 0.0015,
+      "step": 603
+    },
+    {
+      "epoch": 2.5905848787446506,
+      "grad_norm": 0.3106430768966675,
+      "learning_rate": 3.3594712255399554e-06,
+      "loss": 0.0033,
+      "step": 604
+    },
+    {
+      "epoch": 2.5948644793152638,
+      "grad_norm": 0.12472659349441528,
+      "learning_rate": 3.3537866961632465e-06,
+      "loss": 0.0007,
+      "step": 605
+    },
+    {
+      "epoch": 2.5991440798858774,
+      "grad_norm": 0.15480101108551025,
+      "learning_rate": 3.3480971652950757e-06,
+      "loss": 0.0018,
+      "step": 606
+    },
+    {
+      "epoch": 2.6034236804564905,
+      "grad_norm": 0.4321823716163635,
+      "learning_rate": 3.3424026662647763e-06,
+      "loss": 0.0024,
+      "step": 607
+    },
+    {
+      "epoch": 2.607703281027104,
+      "grad_norm": 0.20745764672756195,
+      "learning_rate": 3.3367032324307815e-06,
+      "loss": 0.0024,
+      "step": 608
+    },
+    {
+      "epoch": 2.6119828815977177,
+      "grad_norm": 0.5367287397384644,
+      "learning_rate": 3.330998897180436e-06,
+      "loss": 0.0026,
+      "step": 609
+    },
+    {
+      "epoch": 2.616262482168331,
+      "grad_norm": 0.33065301179885864,
+      "learning_rate": 3.325289693929795e-06,
+      "loss": 0.0031,
+      "step": 610
+    },
+    {
+      "epoch": 2.6205420827389445,
+      "grad_norm": 0.47890713810920715,
+      "learning_rate": 3.3195756561234315e-06,
+      "loss": 0.0017,
+      "step": 611
+    },
+    {
+      "epoch": 2.6248216833095577,
+      "grad_norm": 0.410844624042511,
+      "learning_rate": 3.31385681723424e-06,
+      "loss": 0.0012,
+      "step": 612
+    },
+    {
+      "epoch": 2.6291012838801713,
+      "grad_norm": 0.17063617706298828,
+      "learning_rate": 3.3081332107632386e-06,
+      "loss": 0.002,
+      "step": 613
+    },
+    {
+      "epoch": 2.6333808844507844,
+      "grad_norm": 0.24861586093902588,
+      "learning_rate": 3.302404870239374e-06,
+      "loss": 0.0013,
+      "step": 614
+    },
+    {
+      "epoch": 2.637660485021398,
+      "grad_norm": 0.22020435333251953,
+      "learning_rate": 3.2966718292193247e-06,
+      "loss": 0.0027,
+      "step": 615
+    },
+    {
+      "epoch": 2.6419400855920117,
+      "grad_norm": 0.30350565910339355,
+      "learning_rate": 3.2909341212873065e-06,
+      "loss": 0.003,
+      "step": 616
+    },
+    {
+      "epoch": 2.646219686162625,
+      "grad_norm": 0.39022594690322876,
+      "learning_rate": 3.2851917800548726e-06,
+      "loss": 0.0041,
+      "step": 617
+    },
+    {
+      "epoch": 2.650499286733238,
+      "grad_norm": 0.41128891706466675,
+      "learning_rate": 3.279444839160718e-06,
+      "loss": 0.0016,
+      "step": 618
+    },
+    {
+      "epoch": 2.6547788873038516,
+      "grad_norm": 0.2784186005592346,
+      "learning_rate": 3.273693332270484e-06,
+      "loss": 0.0015,
+      "step": 619
+    },
+    {
+      "epoch": 2.659058487874465,
+      "grad_norm": 0.3071463108062744,
+      "learning_rate": 3.2679372930765575e-06,
+      "loss": 0.0038,
+      "step": 620
+    },
+    {
+      "epoch": 2.6633380884450784,
+      "grad_norm": 1.0749564170837402,
+      "learning_rate": 3.262176755297877e-06,
+      "loss": 0.0021,
+      "step": 621
+    },
+    {
+      "epoch": 2.667617689015692,
+      "grad_norm": 0.3864515423774719,
+      "learning_rate": 3.2564117526797343e-06,
+      "loss": 0.0051,
+      "step": 622
+    },
+    {
+      "epoch": 2.671897289586305,
+      "grad_norm": 0.4214929938316345,
+      "learning_rate": 3.250642318993574e-06,
+      "loss": 0.003,
+      "step": 623
+    },
+    {
+      "epoch": 2.6761768901569187,
+      "grad_norm": 0.20947064459323883,
+      "learning_rate": 3.2448684880368016e-06,
+      "loss": 0.0021,
+      "step": 624
+    },
+    {
+      "epoch": 2.680456490727532,
+      "grad_norm": 0.41602402925491333,
+      "learning_rate": 3.2390902936325768e-06,
+      "loss": 0.0026,
+      "step": 625
+    },
+    {
+      "epoch": 2.6847360912981455,
+      "grad_norm": 0.36582034826278687,
+      "learning_rate": 3.233307769629626e-06,
+      "loss": 0.0043,
+      "step": 626
+    },
+    {
+      "epoch": 2.689015691868759,
+      "grad_norm": 0.2663152813911438,
+      "learning_rate": 3.2275209499020343e-06,
+      "loss": 0.0018,
+      "step": 627
+    },
+    {
+      "epoch": 2.6932952924393723,
+      "grad_norm": 0.3802527189254761,
+      "learning_rate": 3.221729868349053e-06,
+      "loss": 0.0015,
+      "step": 628
+    },
+    {
+      "epoch": 2.697574893009986,
+      "grad_norm": 0.2044801265001297,
+      "learning_rate": 3.2159345588948986e-06,
+      "loss": 0.0016,
+      "step": 629
+    },
+    {
+      "epoch": 2.701854493580599,
+      "grad_norm": 0.31439611315727234,
+      "learning_rate": 3.2101350554885552e-06,
+      "loss": 0.0012,
+      "step": 630
+    },
+    {
+      "epoch": 2.7061340941512126,
+      "grad_norm": 0.40683820843696594,
+      "learning_rate": 3.2043313921035747e-06,
+      "loss": 0.0044,
+      "step": 631
+    },
+    {
+      "epoch": 2.710413694721826,
+      "grad_norm": 0.3958558142185211,
+      "learning_rate": 3.1985236027378786e-06,
+      "loss": 0.0008,
+      "step": 632
+    },
+    {
+      "epoch": 2.7146932952924394,
+      "grad_norm": 0.5626212358474731,
+      "learning_rate": 3.192711721413558e-06,
+      "loss": 0.0025,
+      "step": 633
+    },
+    {
+      "epoch": 2.718972895863053,
+      "grad_norm": 0.4065411388874054,
+      "learning_rate": 3.1868957821766742e-06,
+      "loss": 0.0025,
+      "step": 634
+    },
+    {
+      "epoch": 2.723252496433666,
+      "grad_norm": 0.2741824686527252,
+      "learning_rate": 3.1810758190970623e-06,
+      "loss": 0.0009,
+      "step": 635
+    },
+    {
+      "epoch": 2.7275320970042793,
+      "grad_norm": 0.13432888686656952,
+      "learning_rate": 3.1752518662681257e-06,
+      "loss": 0.001,
+      "step": 636
+    },
+    {
+      "epoch": 2.731811697574893,
+      "grad_norm": 0.17365941405296326,
+      "learning_rate": 3.1694239578066427e-06,
+      "loss": 0.0039,
+      "step": 637
+    },
+    {
+      "epoch": 2.7360912981455066,
+      "grad_norm": 0.2563876509666443,
+      "learning_rate": 3.1635921278525615e-06,
+      "loss": 0.0015,
+      "step": 638
+    },
+    {
+      "epoch": 2.7403708987161197,
+      "grad_norm": 0.17576679587364197,
+      "learning_rate": 3.157756410568803e-06,
+      "loss": 0.0013,
+      "step": 639
+    },
+    {
+      "epoch": 2.7446504992867333,
+      "grad_norm": 0.3357720375061035,
+      "learning_rate": 3.1519168401410627e-06,
+      "loss": 0.0046,
+      "step": 640
+    },
+    {
+      "epoch": 2.7489300998573465,
+      "grad_norm": 0.34145796298980713,
+      "learning_rate": 3.1460734507776036e-06,
+      "loss": 0.0027,
+      "step": 641
+    },
+    {
+      "epoch": 2.75320970042796,
+      "grad_norm": 0.5672411918640137,
+      "learning_rate": 3.140226276709063e-06,
+      "loss": 0.0009,
+      "step": 642
+    },
+    {
+      "epoch": 2.7574893009985733,
+      "grad_norm": 0.039542071521282196,
+      "learning_rate": 3.134375352188247e-06,
+      "loss": 0.0029,
+      "step": 643
+    },
+    {
+      "epoch": 2.761768901569187,
+      "grad_norm": 0.3622141182422638,
+      "learning_rate": 3.1285207114899353e-06,
+      "loss": 0.0011,
+      "step": 644
+    },
+    {
+      "epoch": 2.7660485021398005,
+      "grad_norm": 0.12562383711338043,
+      "learning_rate": 3.122662388910673e-06,
+      "loss": 0.0022,
+      "step": 645
+    },
+    {
+      "epoch": 2.7703281027104136,
+      "grad_norm": 0.12533040344715118,
+      "learning_rate": 3.1168004187685765e-06,
+      "loss": 0.0023,
+      "step": 646
+    },
+    {
+      "epoch": 2.7746077032810272,
+      "grad_norm": 0.47782161831855774,
+      "learning_rate": 3.1109348354031284e-06,
+      "loss": 0.0026,
+      "step": 647
+    },
+    {
+      "epoch": 2.7788873038516404,
+      "grad_norm": 0.1668027937412262,
+      "learning_rate": 3.1050656731749767e-06,
+      "loss": 0.0006,
+      "step": 648
+    },
+    {
+      "epoch": 2.783166904422254,
+      "grad_norm": 0.23814761638641357,
+      "learning_rate": 3.0991929664657373e-06,
+      "loss": 0.0028,
+      "step": 649
+    },
+    {
+      "epoch": 2.787446504992867,
+      "grad_norm": 0.36936724185943604,
+      "learning_rate": 3.093316749677788e-06,
+      "loss": 0.0012,
+      "step": 650
+    },
+    {
+      "epoch": 2.7917261055634808,
+      "grad_norm": 0.26975494623184204,
+      "learning_rate": 3.087437057234066e-06,
+      "loss": 0.0014,
+      "step": 651
+    },
+    {
+      "epoch": 2.7960057061340944,
+      "grad_norm": 0.20466478168964386,
+      "learning_rate": 3.081553923577875e-06,
+      "loss": 0.0013,
+      "step": 652
+    },
+    {
+      "epoch": 2.8002853067047075,
+      "grad_norm": 0.17590539157390594,
+      "learning_rate": 3.0756673831726723e-06,
+      "loss": 0.0008,
+      "step": 653
+    },
+    {
+      "epoch": 2.8045649072753207,
+      "grad_norm": 0.16283346712589264,
+      "learning_rate": 3.0697774705018734e-06,
+      "loss": 0.0008,
+      "step": 654
+    },
+    {
+      "epoch": 2.8088445078459343,
+      "grad_norm": 0.17097610235214233,
+      "learning_rate": 3.063884220068649e-06,
+      "loss": 0.0011,
+      "step": 655
+    },
+    {
+      "epoch": 2.813124108416548,
+      "grad_norm": 0.1250288337469101,
+      "learning_rate": 3.057987666395723e-06,
+      "loss": 0.0041,
+      "step": 656
+    },
+    {
+      "epoch": 2.817403708987161,
+      "grad_norm": 0.2146853357553482,
+      "learning_rate": 3.052087844025168e-06,
+      "loss": 0.0007,
+      "step": 657
+    },
+    {
+      "epoch": 2.8216833095577747,
+      "grad_norm": 0.07225178927183151,
+      "learning_rate": 3.0461847875182056e-06,
+      "loss": 0.0004,
+      "step": 658
+    },
+    {
+      "epoch": 2.8259629101283883,
+      "grad_norm": 0.12505510449409485,
+      "learning_rate": 3.0402785314550037e-06,
+      "loss": 0.0013,
+      "step": 659
+    },
+    {
+      "epoch": 2.8302425106990015,
+      "grad_norm": 0.1191864013671875,
+      "learning_rate": 3.034369110434472e-06,
+      "loss": 0.0014,
+      "step": 660
+    },
+    {
+      "epoch": 2.8345221112696146,
+      "grad_norm": 0.49187931418418884,
+      "learning_rate": 3.028456559074061e-06,
+      "loss": 0.0003,
+      "step": 661
+    },
+    {
+      "epoch": 2.8388017118402282,
+      "grad_norm": 0.19233481585979462,
+      "learning_rate": 3.0225409120095597e-06,
+      "loss": 0.0015,
+      "step": 662
+    },
+    {
+      "epoch": 2.843081312410842,
+      "grad_norm": 0.13615825772285461,
+      "learning_rate": 3.01662220389489e-06,
+      "loss": 0.0047,
+      "step": 663
+    },
+    {
+      "epoch": 2.847360912981455,
+      "grad_norm": 0.33949339389801025,
+      "learning_rate": 3.0107004694019067e-06,
+      "loss": 0.0016,
+      "step": 664
+    },
+    {
+      "epoch": 2.8516405135520686,
+      "grad_norm": 0.22976908087730408,
+      "learning_rate": 3.0047757432201936e-06,
+      "loss": 0.0003,
+      "step": 665
+    },
+    {
+      "epoch": 2.8559201141226818,
+      "grad_norm": 0.06495361775159836,
+      "learning_rate": 2.9988480600568578e-06,
+      "loss": 0.0006,
+      "step": 666
+    },
+    {
+      "epoch": 2.8601997146932954,
+      "grad_norm": 0.19875776767730713,
+      "learning_rate": 2.9929174546363303e-06,
+      "loss": 0.0009,
+      "step": 667
+    },
+    {
+      "epoch": 2.8644793152639085,
+      "grad_norm": 0.08605135977268219,
+      "learning_rate": 2.9869839617001605e-06,
+      "loss": 0.0008,
+      "step": 668
+    },
+    {
+      "epoch": 2.868758915834522,
+      "grad_norm": 0.22537599503993988,
+      "learning_rate": 2.9810476160068113e-06,
+      "loss": 0.0012,
+      "step": 669
+    },
+    {
+      "epoch": 2.8730385164051357,
+      "grad_norm": 0.1354469507932663,
+      "learning_rate": 2.97510845233146e-06,
+      "loss": 0.001,
+      "step": 670
+    },
+    {
+      "epoch": 2.877318116975749,
+      "grad_norm": 0.28886279463768005,
+      "learning_rate": 2.969166505465789e-06,
+      "loss": 0.0011,
+      "step": 671
+    },
+    {
+      "epoch": 2.881597717546362,
+      "grad_norm": 0.35490158200263977,
+      "learning_rate": 2.963221810217786e-06,
+      "loss": 0.0002,
+      "step": 672
+    },
+    {
+      "epoch": 2.8858773181169757,
+      "grad_norm": 0.06331105530261993,
+      "learning_rate": 2.9572744014115384e-06,
+      "loss": 0.0012,
+      "step": 673
+    },
+    {
+      "epoch": 2.8901569186875893,
+      "grad_norm": 1.2145577669143677,
+      "learning_rate": 2.9513243138870294e-06,
+      "loss": 0.0013,
+      "step": 674
+    },
+    {
+      "epoch": 2.8944365192582024,
+      "grad_norm": 0.3513360917568207,
+      "learning_rate": 2.9453715824999348e-06,
+      "loss": 0.0026,
+      "step": 675
+    },
+    {
+      "epoch": 2.898716119828816,
+      "grad_norm": 0.33389878273010254,
+      "learning_rate": 2.9394162421214167e-06,
+      "loss": 0.0004,
+      "step": 676
+    },
+    {
+      "epoch": 2.9029957203994297,
+      "grad_norm": 0.10553620010614395,
+      "learning_rate": 2.933458327637923e-06,
+      "loss": 0.0009,
+      "step": 677
+    },
+    {
+      "epoch": 2.907275320970043,
+      "grad_norm": 0.296312540769577,
+      "learning_rate": 2.927497873950979e-06,
+      "loss": 0.0007,
+      "step": 678
+    },
+    {
+      "epoch": 2.911554921540656,
+      "grad_norm": 0.26307055354118347,
+      "learning_rate": 2.921534915976985e-06,
+      "loss": 0.0014,
+      "step": 679
+    },
+    {
+      "epoch": 2.9158345221112696,
+      "grad_norm": 0.05739770829677582,
+      "learning_rate": 2.9155694886470124e-06,
+      "loss": 0.0013,
+      "step": 680
+    },
+    {
+      "epoch": 2.920114122681883,
+      "grad_norm": 0.3625591993331909,
+      "learning_rate": 2.909601626906597e-06,
+      "loss": 0.0013,
+      "step": 681
+    },
+    {
+      "epoch": 2.9243937232524964,
+      "grad_norm": 0.40288034081459045,
+      "learning_rate": 2.9036313657155347e-06,
+      "loss": 0.001,
+      "step": 682
+    },
+    {
+      "epoch": 2.92867332382311,
+      "grad_norm": 0.9118512868881226,
+      "learning_rate": 2.8976587400476804e-06,
+      "loss": 0.0016,
+      "step": 683
+    },
+    {
+      "epoch": 2.932952924393723,
+      "grad_norm": 0.18625538051128387,
+      "learning_rate": 2.891683784890737e-06,
+      "loss": 0.0008,
+      "step": 684
+    },
+    {
+      "epoch": 2.9372325249643367,
+      "grad_norm": 0.10386449843645096,
+      "learning_rate": 2.885706535246055e-06,
+      "loss": 0.0005,
+      "step": 685
+    },
+    {
+      "epoch": 2.94151212553495,
+      "grad_norm": 0.10134569555521011,
+      "learning_rate": 2.8797270261284247e-06,
+      "loss": 0.0014,
+      "step": 686
+    },
+    {
+      "epoch": 2.9457917261055635,
+      "grad_norm": 0.2270444929599762,
+      "learning_rate": 2.8737452925658753e-06,
+      "loss": 0.0006,
+      "step": 687
+    },
+    {
+      "epoch": 2.950071326676177,
+      "grad_norm": 0.09928806871175766,
+      "learning_rate": 2.867761369599465e-06,
+      "loss": 0.0007,
+      "step": 688
+    },
+    {
+      "epoch": 2.9543509272467903,
+      "grad_norm": 0.4043658375740051,
+      "learning_rate": 2.8617752922830755e-06,
+      "loss": 0.001,
+      "step": 689
+    },
+    {
+      "epoch": 2.9586305278174034,
+      "grad_norm": 0.12267272174358368,
+      "learning_rate": 2.8557870956832135e-06,
+      "loss": 0.0011,
+      "step": 690
+    },
+    {
+      "epoch": 2.962910128388017,
+      "grad_norm": 0.29241859912872314,
+      "learning_rate": 2.8497968148787953e-06,
+      "loss": 0.0009,
+      "step": 691
+    },
+    {
+      "epoch": 2.9671897289586306,
+      "grad_norm": 0.38876116275787354,
+      "learning_rate": 2.8438044849609508e-06,
+      "loss": 0.0031,
+      "step": 692
+    },
+    {
+      "epoch": 2.971469329529244,
+      "grad_norm": 0.36466461420059204,
+      "learning_rate": 2.8378101410328107e-06,
+      "loss": 0.0011,
+      "step": 693
+    },
+    {
+      "epoch": 2.9757489300998574,
+      "grad_norm": 1.4299386739730835,
+      "learning_rate": 2.8318138182093053e-06,
+      "loss": 0.0023,
+      "step": 694
+    },
+    {
+      "epoch": 2.980028530670471,
+      "grad_norm": 0.07230447232723236,
+      "learning_rate": 2.825815551616957e-06,
+      "loss": 0.0024,
+      "step": 695
+    },
+    {
+      "epoch": 2.984308131241084,
+      "grad_norm": 0.268041729927063,
+      "learning_rate": 2.8198153763936746e-06,
+      "loss": 0.0023,
+      "step": 696
+    },
+    {
+      "epoch": 2.9885877318116973,
+      "grad_norm": 0.5011327266693115,
+      "learning_rate": 2.813813327688547e-06,
+      "loss": 0.001,
+      "step": 697
+    },
+    {
+      "epoch": 2.992867332382311,
+      "grad_norm": 3.3417763710021973,
+      "learning_rate": 2.8078094406616403e-06,
+      "loss": 0.0033,
+      "step": 698
+    },
+    {
+      "epoch": 2.9971469329529246,
+      "grad_norm": 0.11345068365335464,
+      "learning_rate": 2.801803750483787e-06,
+      "loss": 0.0037,
+      "step": 699
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 1398,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 6,
+  "save_steps": 233,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.567353388650201e+18,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}