diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,17451 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.0,
+  "eval_steps": 500,
+  "global_step": 2487,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0012062726176115801,
+      "grad_norm": 51.38860427847401,
+      "learning_rate": 2.0080321285140563e-07,
+      "loss": 11.2525,
+      "step": 1
+    },
+    {
+      "epoch": 0.0024125452352231603,
+      "grad_norm": 53.36026048366278,
+      "learning_rate": 4.0160642570281125e-07,
+      "loss": 11.2364,
+      "step": 2
+    },
+    {
+      "epoch": 0.0036188178528347406,
+      "grad_norm": 53.32175177566749,
+      "learning_rate": 6.024096385542169e-07,
+      "loss": 11.1112,
+      "step": 3
+    },
+    {
+      "epoch": 0.0048250904704463205,
+      "grad_norm": 54.3523155769173,
+      "learning_rate": 8.032128514056225e-07,
+      "loss": 11.017,
+      "step": 4
+    },
+    {
+      "epoch": 0.006031363088057901,
+      "grad_norm": 53.85262766532482,
+      "learning_rate": 1.0040160642570282e-06,
+      "loss": 11.1188,
+      "step": 5
+    },
+    {
+      "epoch": 0.007237635705669481,
+      "grad_norm": 54.13262202877557,
+      "learning_rate": 1.2048192771084338e-06,
+      "loss": 11.0221,
+      "step": 6
+    },
+    {
+      "epoch": 0.008443908323281062,
+      "grad_norm": 56.227774935885684,
+      "learning_rate": 1.4056224899598394e-06,
+      "loss": 10.8553,
+      "step": 7
+    },
+    {
+      "epoch": 0.009650180940892641,
+      "grad_norm": 59.35662182504017,
+      "learning_rate": 1.606425702811245e-06,
+      "loss": 10.7837,
+      "step": 8
+    },
+    {
+      "epoch": 0.010856453558504222,
+      "grad_norm": 58.81276787565515,
+      "learning_rate": 1.8072289156626506e-06,
+      "loss": 10.7875,
+      "step": 9
+    },
+    {
+      "epoch": 0.012062726176115802,
+      "grad_norm": 60.2995135454209,
+      "learning_rate": 2.0080321285140564e-06,
+      "loss": 10.7267,
+      "step": 10
+    },
+    {
+      "epoch": 0.013268998793727383,
+      "grad_norm": 80.16682488222419,
+      "learning_rate": 2.208835341365462e-06,
+      "loss": 9.5366,
+      "step": 11
+    },
+    {
+      "epoch": 0.014475271411338963,
+      "grad_norm": 80.01557833575397,
+      "learning_rate": 2.4096385542168676e-06,
+      "loss": 9.5007,
+      "step": 12
+    },
+    {
+      "epoch": 0.015681544028950542,
+      "grad_norm": 101.46955180189386,
+      "learning_rate": 2.610441767068273e-06,
+      "loss": 8.905,
+      "step": 13
+    },
+    {
+      "epoch": 0.016887816646562123,
+      "grad_norm": 81.57781816299472,
+      "learning_rate": 2.811244979919679e-06,
+      "loss": 4.339,
+      "step": 14
+    },
+    {
+      "epoch": 0.018094089264173704,
+      "grad_norm": 64.68209625020934,
+      "learning_rate": 3.0120481927710846e-06,
+      "loss": 3.6619,
+      "step": 15
+    },
+    {
+      "epoch": 0.019300361881785282,
+      "grad_norm": 56.85869610525475,
+      "learning_rate": 3.21285140562249e-06,
+      "loss": 3.4092,
+      "step": 16
+    },
+    {
+      "epoch": 0.020506634499396863,
+      "grad_norm": 40.39884604434922,
+      "learning_rate": 3.413654618473896e-06,
+      "loss": 2.7495,
+      "step": 17
+    },
+    {
+      "epoch": 0.021712907117008445,
+      "grad_norm": 33.48382966020048,
+      "learning_rate": 3.614457831325301e-06,
+      "loss": 2.5099,
+      "step": 18
+    },
+    {
+      "epoch": 0.022919179734620022,
+      "grad_norm": 9.068825320578902,
+      "learning_rate": 3.8152610441767074e-06,
+      "loss": 1.5497,
+      "step": 19
+    },
+    {
+      "epoch": 0.024125452352231604,
+      "grad_norm": 5.430810054078272,
+      "learning_rate": 4.016064257028113e-06,
+      "loss": 1.3625,
+      "step": 20
+    },
+    {
+      "epoch": 0.025331724969843185,
+      "grad_norm": 4.369877876698654,
+      "learning_rate": 4.216867469879518e-06,
+      "loss": 1.2633,
+      "step": 21
+    },
+    {
+      "epoch": 0.026537997587454766,
+      "grad_norm": 3.6966719303916666,
+      "learning_rate": 4.417670682730924e-06,
+      "loss": 1.2189,
+      "step": 22
+    },
+    {
+      "epoch": 0.027744270205066344,
+      "grad_norm": 2.7012172428180246,
+      "learning_rate": 4.618473895582329e-06,
+      "loss": 1.127,
+      "step": 23
+    },
+    {
+      "epoch": 0.028950542822677925,
+      "grad_norm": 2.1966628156071977,
+      "learning_rate": 4.819277108433735e-06,
+      "loss": 1.0832,
+      "step": 24
+    },
+    {
+      "epoch": 0.030156815440289506,
+      "grad_norm": 1.9505800224070988,
+      "learning_rate": 5.020080321285141e-06,
+      "loss": 1.0821,
+      "step": 25
+    },
+    {
+      "epoch": 0.031363088057901084,
+      "grad_norm": 1.4033827272745318,
+      "learning_rate": 5.220883534136546e-06,
+      "loss": 0.9829,
+      "step": 26
+    },
+    {
+      "epoch": 0.032569360675512665,
+      "grad_norm": 25.83482630979627,
+      "learning_rate": 5.421686746987952e-06,
+      "loss": 0.9409,
+      "step": 27
+    },
+    {
+      "epoch": 0.033775633293124246,
+      "grad_norm": 8.816041565642406,
+      "learning_rate": 5.622489959839358e-06,
+      "loss": 0.9217,
+      "step": 28
+    },
+    {
+      "epoch": 0.03498190591073583,
+      "grad_norm": 1.865924593494002,
+      "learning_rate": 5.823293172690764e-06,
+      "loss": 0.9477,
+      "step": 29
+    },
+    {
+      "epoch": 0.03618817852834741,
+      "grad_norm": 1.185561066111754,
+      "learning_rate": 6.024096385542169e-06,
+      "loss": 0.9124,
+      "step": 30
+    },
+    {
+      "epoch": 0.03739445114595899,
+      "grad_norm": 1.0072345862575207,
+      "learning_rate": 6.224899598393575e-06,
+      "loss": 0.8344,
+      "step": 31
+    },
+    {
+      "epoch": 0.038600723763570564,
+      "grad_norm": 0.8930661954864261,
+      "learning_rate": 6.42570281124498e-06,
+      "loss": 0.802,
+      "step": 32
+    },
+    {
+      "epoch": 0.039806996381182146,
+      "grad_norm": 0.8218534816604156,
+      "learning_rate": 6.626506024096386e-06,
+      "loss": 0.8519,
+      "step": 33
+    },
+    {
+      "epoch": 0.04101326899879373,
+      "grad_norm": 0.7409176841066717,
+      "learning_rate": 6.827309236947792e-06,
+      "loss": 0.8252,
+      "step": 34
+    },
+    {
+      "epoch": 0.04221954161640531,
+      "grad_norm": 0.694890954906707,
+      "learning_rate": 7.028112449799197e-06,
+      "loss": 0.7267,
+      "step": 35
+    },
+    {
+      "epoch": 0.04342581423401689,
+      "grad_norm": 1.0238217234534817,
+      "learning_rate": 7.228915662650602e-06,
+      "loss": 0.8264,
+      "step": 36
+    },
+    {
+      "epoch": 0.04463208685162847,
+      "grad_norm": 0.6077731225781826,
+      "learning_rate": 7.429718875502008e-06,
+      "loss": 0.7901,
+      "step": 37
+    },
+    {
+      "epoch": 0.045838359469240045,
+      "grad_norm": 0.6066967650950137,
+      "learning_rate": 7.630522088353415e-06,
+      "loss": 0.7383,
+      "step": 38
+    },
+    {
+      "epoch": 0.047044632086851626,
+      "grad_norm": 0.6209543801705237,
+      "learning_rate": 7.83132530120482e-06,
+      "loss": 0.7057,
+      "step": 39
+    },
+    {
+      "epoch": 0.04825090470446321,
+      "grad_norm": 0.5825557752671116,
+      "learning_rate": 8.032128514056226e-06,
+      "loss": 0.6934,
+      "step": 40
+    },
+    {
+      "epoch": 0.04945717732207479,
+      "grad_norm": 0.5769538262434953,
+      "learning_rate": 8.232931726907631e-06,
+      "loss": 0.7009,
+      "step": 41
+    },
+    {
+      "epoch": 0.05066344993968637,
+      "grad_norm": 0.5003790590362588,
+      "learning_rate": 8.433734939759036e-06,
+      "loss": 0.6833,
+      "step": 42
+    },
+    {
+      "epoch": 0.05186972255729795,
+      "grad_norm": 0.5156286849928355,
+      "learning_rate": 8.634538152610444e-06,
+      "loss": 0.7108,
+      "step": 43
+    },
+    {
+      "epoch": 0.05307599517490953,
+      "grad_norm": 0.5113020018822674,
+      "learning_rate": 8.835341365461847e-06,
+      "loss": 0.6978,
+      "step": 44
+    },
+    {
+      "epoch": 0.054282267792521106,
+      "grad_norm": 0.49956584530074893,
+      "learning_rate": 9.036144578313253e-06,
+      "loss": 0.6553,
+      "step": 45
+    },
+    {
+      "epoch": 0.05548854041013269,
+      "grad_norm": 0.488845369844212,
+      "learning_rate": 9.236947791164658e-06,
+      "loss": 0.6637,
+      "step": 46
+    },
+    {
+      "epoch": 0.05669481302774427,
+      "grad_norm": 0.4479235016367263,
+      "learning_rate": 9.437751004016063e-06,
+      "loss": 0.6509,
+      "step": 47
+    },
+    {
+      "epoch": 0.05790108564535585,
+      "grad_norm": 0.4407343454865804,
+      "learning_rate": 9.63855421686747e-06,
+      "loss": 0.6484,
+      "step": 48
+    },
+    {
+      "epoch": 0.05910735826296743,
+      "grad_norm": 0.4661813735492102,
+      "learning_rate": 9.839357429718876e-06,
+      "loss": 0.6106,
+      "step": 49
+    },
+    {
+      "epoch": 0.06031363088057901,
+      "grad_norm": 0.4688851051826549,
+      "learning_rate": 1.0040160642570281e-05,
+      "loss": 0.6536,
+      "step": 50
+    },
+    {
+      "epoch": 0.061519903498190594,
+      "grad_norm": 0.44874471772574015,
+      "learning_rate": 1.0240963855421687e-05,
+      "loss": 0.6356,
+      "step": 51
+    },
+    {
+      "epoch": 0.06272617611580217,
+      "grad_norm": 0.4205764980483692,
+      "learning_rate": 1.0441767068273092e-05,
+      "loss": 0.6395,
+      "step": 52
+    },
+    {
+      "epoch": 0.06393244873341375,
+      "grad_norm": 0.3773662466705985,
+      "learning_rate": 1.0642570281124499e-05,
+      "loss": 0.6105,
+      "step": 53
+    },
+    {
+      "epoch": 0.06513872135102533,
+      "grad_norm": 0.36862909221937795,
+      "learning_rate": 1.0843373493975904e-05,
+      "loss": 0.6087,
+      "step": 54
+    },
+    {
+      "epoch": 0.06634499396863691,
+      "grad_norm": 0.4535482567802974,
+      "learning_rate": 1.104417670682731e-05,
+      "loss": 0.6261,
+      "step": 55
+    },
+    {
+      "epoch": 0.06755126658624849,
+      "grad_norm": 0.35536767227242044,
+      "learning_rate": 1.1244979919678715e-05,
+      "loss": 0.6005,
+      "step": 56
+    },
+    {
+      "epoch": 0.06875753920386007,
+      "grad_norm": 0.3424177957048031,
+      "learning_rate": 1.144578313253012e-05,
+      "loss": 0.6432,
+      "step": 57
+    },
+    {
+      "epoch": 0.06996381182147166,
+      "grad_norm": 0.3701788632145171,
+      "learning_rate": 1.1646586345381528e-05,
+      "loss": 0.6008,
+      "step": 58
+    },
+    {
+      "epoch": 0.07117008443908324,
+      "grad_norm": 0.3101793503776111,
+      "learning_rate": 1.1847389558232933e-05,
+      "loss": 0.6143,
+      "step": 59
+    },
+    {
+      "epoch": 0.07237635705669482,
+      "grad_norm": 0.3001818267393327,
+      "learning_rate": 1.2048192771084338e-05,
+      "loss": 0.6043,
+      "step": 60
+    },
+    {
+      "epoch": 0.0735826296743064,
+      "grad_norm": 0.30001188823098984,
+      "learning_rate": 1.2248995983935744e-05,
+      "loss": 0.5812,
+      "step": 61
+    },
+    {
+      "epoch": 0.07478890229191798,
+      "grad_norm": 0.3287508814851007,
+      "learning_rate": 1.244979919678715e-05,
+      "loss": 0.6174,
+      "step": 62
+    },
+    {
+      "epoch": 0.07599517490952955,
+      "grad_norm": 0.31978551599047533,
+      "learning_rate": 1.2650602409638555e-05,
+      "loss": 0.6056,
+      "step": 63
+    },
+    {
+      "epoch": 0.07720144752714113,
+      "grad_norm": 0.3117661373695458,
+      "learning_rate": 1.285140562248996e-05,
+      "loss": 0.5736,
+      "step": 64
+    },
+    {
+      "epoch": 0.07840772014475271,
+      "grad_norm": 0.32162182570694564,
+      "learning_rate": 1.3052208835341367e-05,
+      "loss": 0.6024,
+      "step": 65
+    },
+    {
+      "epoch": 0.07961399276236429,
+      "grad_norm": 0.3119082159220764,
+      "learning_rate": 1.3253012048192772e-05,
+      "loss": 0.5884,
+      "step": 66
+    },
+    {
+      "epoch": 0.08082026537997587,
+      "grad_norm": 0.3754051455489471,
+      "learning_rate": 1.3453815261044178e-05,
+      "loss": 0.6004,
+      "step": 67
+    },
+    {
+      "epoch": 0.08202653799758745,
+      "grad_norm": 0.3109694023561302,
+      "learning_rate": 1.3654618473895583e-05,
+      "loss": 0.5617,
+      "step": 68
+    },
+    {
+      "epoch": 0.08323281061519903,
+      "grad_norm": 0.28664605601137993,
+      "learning_rate": 1.3855421686746989e-05,
+      "loss": 0.5829,
+      "step": 69
+    },
+    {
+      "epoch": 0.08443908323281062,
+      "grad_norm": 0.31651626321139076,
+      "learning_rate": 1.4056224899598394e-05,
+      "loss": 0.5846,
+      "step": 70
+    },
+    {
+      "epoch": 0.0856453558504222,
+      "grad_norm": 0.28296350697482814,
+      "learning_rate": 1.42570281124498e-05,
+      "loss": 0.5715,
+      "step": 71
+    },
+    {
+      "epoch": 0.08685162846803378,
+      "grad_norm": 0.3168399355296006,
+      "learning_rate": 1.4457831325301205e-05,
+      "loss": 0.5613,
+      "step": 72
+    },
+    {
+      "epoch": 0.08805790108564536,
+      "grad_norm": 0.2955763208938298,
+      "learning_rate": 1.465863453815261e-05,
+      "loss": 0.5761,
+      "step": 73
+    },
+    {
+      "epoch": 0.08926417370325694,
+      "grad_norm": 0.2660115535346986,
+      "learning_rate": 1.4859437751004016e-05,
+      "loss": 0.5814,
+      "step": 74
+    },
+    {
+      "epoch": 0.09047044632086852,
+      "grad_norm": 0.31092486179587725,
+      "learning_rate": 1.5060240963855424e-05,
+      "loss": 0.5901,
+      "step": 75
+    },
+    {
+      "epoch": 0.09167671893848009,
+      "grad_norm": 0.2672347270128979,
+      "learning_rate": 1.526104417670683e-05,
+      "loss": 0.5715,
+      "step": 76
+    },
+    {
+      "epoch": 0.09288299155609167,
+      "grad_norm": 0.31255025482262866,
+      "learning_rate": 1.5461847389558235e-05,
+      "loss": 0.5856,
+      "step": 77
+    },
+    {
+      "epoch": 0.09408926417370325,
+      "grad_norm": 0.2981092204177073,
+      "learning_rate": 1.566265060240964e-05,
+      "loss": 0.5938,
+      "step": 78
+    },
+    {
+      "epoch": 0.09529553679131483,
+      "grad_norm": 0.2901487206816295,
+      "learning_rate": 1.5863453815261046e-05,
+      "loss": 0.5699,
+      "step": 79
+    },
+    {
+      "epoch": 0.09650180940892641,
+      "grad_norm": 0.3242024566962533,
+      "learning_rate": 1.606425702811245e-05,
+      "loss": 0.5648,
+      "step": 80
+    },
+    {
+      "epoch": 0.097708082026538,
+      "grad_norm": 0.2901732345581997,
+      "learning_rate": 1.6265060240963857e-05,
+      "loss": 0.5567,
+      "step": 81
+    },
+    {
+      "epoch": 0.09891435464414958,
+      "grad_norm": 0.30430875830826826,
+      "learning_rate": 1.6465863453815262e-05,
+      "loss": 0.5482,
+      "step": 82
+    },
+    {
+      "epoch": 0.10012062726176116,
+      "grad_norm": 0.2693405288007236,
+      "learning_rate": 1.6666666666666667e-05,
+      "loss": 0.533,
+      "step": 83
+    },
+    {
+      "epoch": 0.10132689987937274,
+      "grad_norm": 0.3226079166873796,
+      "learning_rate": 1.6867469879518073e-05,
+      "loss": 0.5636,
+      "step": 84
+    },
+    {
+      "epoch": 0.10253317249698432,
+      "grad_norm": 0.28557229465998485,
+      "learning_rate": 1.706827309236948e-05,
+      "loss": 0.5521,
+      "step": 85
+    },
+    {
+      "epoch": 0.1037394451145959,
+      "grad_norm": 0.3137906150959221,
+      "learning_rate": 1.7269076305220887e-05,
+      "loss": 0.5475,
+      "step": 86
+    },
+    {
+      "epoch": 0.10494571773220748,
+      "grad_norm": 0.2814428939492701,
+      "learning_rate": 1.7469879518072292e-05,
+      "loss": 0.5297,
+      "step": 87
+    },
+    {
+      "epoch": 0.10615199034981906,
+      "grad_norm": 0.27927754569488406,
+      "learning_rate": 1.7670682730923694e-05,
+      "loss": 0.5571,
+      "step": 88
+    },
+    {
+      "epoch": 0.10735826296743065,
+      "grad_norm": 0.29699094559241995,
+      "learning_rate": 1.78714859437751e-05,
+      "loss": 0.5546,
+      "step": 89
+    },
+    {
+      "epoch": 0.10856453558504221,
+      "grad_norm": 0.30506480963814275,
+      "learning_rate": 1.8072289156626505e-05,
+      "loss": 0.5442,
+      "step": 90
+    },
+    {
+      "epoch": 0.1097708082026538,
+      "grad_norm": 0.29804871816448125,
+      "learning_rate": 1.827309236947791e-05,
+      "loss": 0.537,
+      "step": 91
+    },
+    {
+      "epoch": 0.11097708082026538,
+      "grad_norm": 0.3356618966578361,
+      "learning_rate": 1.8473895582329316e-05,
+      "loss": 0.5487,
+      "step": 92
+    },
+    {
+      "epoch": 0.11218335343787696,
+      "grad_norm": 0.29121306377344314,
+      "learning_rate": 1.867469879518072e-05,
+      "loss": 0.5313,
+      "step": 93
+    },
+    {
+      "epoch": 0.11338962605548854,
+      "grad_norm": 0.3004619917101442,
+      "learning_rate": 1.8875502008032127e-05,
+      "loss": 0.5113,
+      "step": 94
+    },
+    {
+      "epoch": 0.11459589867310012,
+      "grad_norm": 0.30776858982546973,
+      "learning_rate": 1.9076305220883535e-05,
+      "loss": 0.549,
+      "step": 95
+    },
+    {
+      "epoch": 0.1158021712907117,
+      "grad_norm": 0.2988781536907027,
+      "learning_rate": 1.927710843373494e-05,
+      "loss": 0.5426,
+      "step": 96
+    },
+    {
+      "epoch": 0.11700844390832328,
+      "grad_norm": 0.306483722050362,
+      "learning_rate": 1.9477911646586346e-05,
+      "loss": 0.5325,
+      "step": 97
+    },
+    {
+      "epoch": 0.11821471652593486,
+      "grad_norm": 0.3321830088641836,
+      "learning_rate": 1.967871485943775e-05,
+      "loss": 0.5573,
+      "step": 98
+    },
+    {
+      "epoch": 0.11942098914354644,
+      "grad_norm": 0.30241064255057876,
+      "learning_rate": 1.9879518072289157e-05,
+      "loss": 0.5362,
+      "step": 99
+    },
+    {
+      "epoch": 0.12062726176115803,
+      "grad_norm": 0.3058839687307177,
+      "learning_rate": 2.0080321285140562e-05,
+      "loss": 0.5259,
+      "step": 100
+    },
+    {
+      "epoch": 0.1218335343787696,
+      "grad_norm": 0.3495388835277465,
+      "learning_rate": 2.0281124497991968e-05,
+      "loss": 0.535,
+      "step": 101
+    },
+    {
+      "epoch": 0.12303980699638119,
+      "grad_norm": 0.2959606492457292,
+      "learning_rate": 2.0481927710843373e-05,
+      "loss": 0.5246,
+      "step": 102
+    },
+    {
+      "epoch": 0.12424607961399277,
+      "grad_norm": 0.39290570723266705,
+      "learning_rate": 2.068273092369478e-05,
+      "loss": 0.5455,
+      "step": 103
+    },
+    {
+      "epoch": 0.12545235223160434,
+      "grad_norm": 0.2926526545122298,
+      "learning_rate": 2.0883534136546184e-05,
+      "loss": 0.5248,
+      "step": 104
+    },
+    {
+      "epoch": 0.12665862484921592,
+      "grad_norm": 0.35016768106063734,
+      "learning_rate": 2.1084337349397593e-05,
+      "loss": 0.5269,
+      "step": 105
+    },
+    {
+      "epoch": 0.1278648974668275,
+      "grad_norm": 0.3699271965637879,
+      "learning_rate": 2.1285140562248998e-05,
+      "loss": 0.5615,
+      "step": 106
+    },
+    {
+      "epoch": 0.12907117008443908,
+      "grad_norm": 0.31020871716617304,
+      "learning_rate": 2.1485943775100404e-05,
+      "loss": 0.532,
+      "step": 107
+    },
+    {
+      "epoch": 0.13027744270205066,
+      "grad_norm": 0.2939373901082531,
+      "learning_rate": 2.168674698795181e-05,
+      "loss": 0.5288,
+      "step": 108
+    },
+    {
+      "epoch": 0.13148371531966224,
+      "grad_norm": 0.33537715712157723,
+      "learning_rate": 2.1887550200803214e-05,
+      "loss": 0.5386,
+      "step": 109
+    },
+    {
+      "epoch": 0.13268998793727382,
+      "grad_norm": 0.30785931040019476,
+      "learning_rate": 2.208835341365462e-05,
+      "loss": 0.5138,
+      "step": 110
+    },
+    {
+      "epoch": 0.1338962605548854,
+      "grad_norm": 0.3962283608203197,
+      "learning_rate": 2.2289156626506025e-05,
+      "loss": 0.5295,
+      "step": 111
+    },
+    {
+      "epoch": 0.13510253317249699,
+      "grad_norm": 0.3089933224294988,
+      "learning_rate": 2.248995983935743e-05,
+      "loss": 0.5313,
+      "step": 112
+    },
+    {
+      "epoch": 0.13630880579010857,
+      "grad_norm": 0.3401641479121955,
+      "learning_rate": 2.2690763052208836e-05,
+      "loss": 0.5454,
+      "step": 113
+    },
+    {
+      "epoch": 0.13751507840772015,
+      "grad_norm": 0.4083258381690134,
+      "learning_rate": 2.289156626506024e-05,
+      "loss": 0.5447,
+      "step": 114
+    },
+    {
+      "epoch": 0.13872135102533173,
+      "grad_norm": 0.3268362769014123,
+      "learning_rate": 2.309236947791165e-05,
+      "loss": 0.5139,
+      "step": 115
+    },
+    {
+      "epoch": 0.1399276236429433,
+      "grad_norm": 0.36409033931151674,
+      "learning_rate": 2.3293172690763055e-05,
+      "loss": 0.5043,
+      "step": 116
+    },
+    {
+      "epoch": 0.1411338962605549,
+      "grad_norm": 0.3333308518450483,
+      "learning_rate": 2.349397590361446e-05,
+      "loss": 0.5149,
+      "step": 117
+    },
+    {
+      "epoch": 0.14234016887816647,
+      "grad_norm": 0.31714621122727005,
+      "learning_rate": 2.3694779116465866e-05,
+      "loss": 0.5352,
+      "step": 118
+    },
+    {
+      "epoch": 0.14354644149577805,
+      "grad_norm": 0.38311242859384087,
+      "learning_rate": 2.389558232931727e-05,
+      "loss": 0.4967,
+      "step": 119
+    },
+    {
+      "epoch": 0.14475271411338964,
+      "grad_norm": 0.3379386762440851,
+      "learning_rate": 2.4096385542168677e-05,
+      "loss": 0.4944,
+      "step": 120
+    },
+    {
+      "epoch": 0.14595898673100122,
+      "grad_norm": 0.38722007899872013,
+      "learning_rate": 2.4297188755020082e-05,
+      "loss": 0.5149,
+      "step": 121
+    },
+    {
+      "epoch": 0.1471652593486128,
+      "grad_norm": 0.3791244486209898,
+      "learning_rate": 2.4497991967871488e-05,
+      "loss": 0.5064,
+      "step": 122
+    },
+    {
+      "epoch": 0.14837153196622438,
+      "grad_norm": 0.3936057186514771,
+      "learning_rate": 2.4698795180722893e-05,
+      "loss": 0.5461,
+      "step": 123
+    },
+    {
+      "epoch": 0.14957780458383596,
+      "grad_norm": 0.4127497762279216,
+      "learning_rate": 2.48995983935743e-05,
+      "loss": 0.5356,
+      "step": 124
+    },
+    {
+      "epoch": 0.15078407720144751,
+      "grad_norm": 0.37304035188820656,
+      "learning_rate": 2.5100401606425704e-05,
+      "loss": 0.5357,
+      "step": 125
+    },
+    {
+      "epoch": 0.1519903498190591,
+      "grad_norm": 0.35071128415816427,
+      "learning_rate": 2.530120481927711e-05,
+      "loss": 0.5145,
+      "step": 126
+    },
+    {
+      "epoch": 0.15319662243667068,
+      "grad_norm": 0.3309060548291426,
+      "learning_rate": 2.5502008032128515e-05,
+      "loss": 0.4809,
+      "step": 127
+    },
+    {
+      "epoch": 0.15440289505428226,
+      "grad_norm": 0.3584753468597298,
+      "learning_rate": 2.570281124497992e-05,
+      "loss": 0.5187,
+      "step": 128
+    },
+    {
+      "epoch": 0.15560916767189384,
+      "grad_norm": 0.3895056053670569,
+      "learning_rate": 2.5903614457831325e-05,
+      "loss": 0.5319,
+      "step": 129
+    },
+    {
+      "epoch": 0.15681544028950542,
+      "grad_norm": 0.30890760385437577,
+      "learning_rate": 2.6104417670682734e-05,
+      "loss": 0.5046,
+      "step": 130
+    },
+    {
+      "epoch": 0.158021712907117,
+      "grad_norm": 0.43680103931374326,
+      "learning_rate": 2.6305220883534136e-05,
+      "loss": 0.5127,
+      "step": 131
+    },
+    {
+      "epoch": 0.15922798552472858,
+      "grad_norm": 0.36126521977202153,
+      "learning_rate": 2.6506024096385545e-05,
+      "loss": 0.5057,
+      "step": 132
+    },
+    {
+      "epoch": 0.16043425814234016,
+      "grad_norm": 0.35586513761227634,
+      "learning_rate": 2.6706827309236947e-05,
+      "loss": 0.5197,
+      "step": 133
+    },
+    {
+      "epoch": 0.16164053075995174,
+      "grad_norm": 0.4173616426138242,
+      "learning_rate": 2.6907630522088356e-05,
+      "loss": 0.5262,
+      "step": 134
+    },
+    {
+      "epoch": 0.16284680337756333,
+      "grad_norm": 0.39991088553188797,
+      "learning_rate": 2.7108433734939758e-05,
+      "loss": 0.5202,
+      "step": 135
+    },
+    {
+      "epoch": 0.1640530759951749,
+      "grad_norm": 0.42015480982805903,
+      "learning_rate": 2.7309236947791167e-05,
+      "loss": 0.5122,
+      "step": 136
+    },
+    {
+      "epoch": 0.1652593486127865,
+      "grad_norm": 0.3943371071406799,
+      "learning_rate": 2.751004016064257e-05,
+      "loss": 0.5095,
+      "step": 137
+    },
+    {
+      "epoch": 0.16646562123039807,
+      "grad_norm": 0.35815925986601993,
+      "learning_rate": 2.7710843373493977e-05,
+      "loss": 0.4785,
+      "step": 138
+    },
+    {
+      "epoch": 0.16767189384800965,
+      "grad_norm": 0.3636720740754509,
+      "learning_rate": 2.791164658634538e-05,
+      "loss": 0.491,
+      "step": 139
+    },
+    {
+      "epoch": 0.16887816646562123,
+      "grad_norm": 0.36511051463449823,
+      "learning_rate": 2.8112449799196788e-05,
+      "loss": 0.4983,
+      "step": 140
+    },
+    {
+      "epoch": 0.1700844390832328,
+      "grad_norm": 0.39354315214120617,
+      "learning_rate": 2.8313253012048197e-05,
+      "loss": 0.499,
+      "step": 141
+    },
+    {
+      "epoch": 0.1712907117008444,
+      "grad_norm": 0.3101636125632076,
+      "learning_rate": 2.85140562248996e-05,
+      "loss": 0.4863,
+      "step": 142
+    },
+    {
+      "epoch": 0.17249698431845598,
+      "grad_norm": 0.41813155310935585,
+      "learning_rate": 2.8714859437751008e-05,
+      "loss": 0.5134,
+      "step": 143
+    },
+    {
+      "epoch": 0.17370325693606756,
+      "grad_norm": 0.437321744526369,
+      "learning_rate": 2.891566265060241e-05,
+      "loss": 0.5083,
+      "step": 144
+    },
+    {
+      "epoch": 0.17490952955367914,
+      "grad_norm": 0.32546457563741027,
+      "learning_rate": 2.911646586345382e-05,
+      "loss": 0.4988,
+      "step": 145
+    },
+    {
+      "epoch": 0.17611580217129072,
+      "grad_norm": 0.45406780502076305,
+      "learning_rate": 2.931726907630522e-05,
+      "loss": 0.5034,
+      "step": 146
+    },
+    {
+      "epoch": 0.1773220747889023,
+      "grad_norm": 0.4279546537128257,
+      "learning_rate": 2.951807228915663e-05,
+      "loss": 0.4901,
+      "step": 147
+    },
+    {
+      "epoch": 0.17852834740651388,
+      "grad_norm": 0.42676213380125827,
+      "learning_rate": 2.971887550200803e-05,
+      "loss": 0.5164,
+      "step": 148
+    },
+    {
+      "epoch": 0.17973462002412546,
+      "grad_norm": 0.5195748580254218,
+      "learning_rate": 2.991967871485944e-05,
+      "loss": 0.4891,
+      "step": 149
+    },
+    {
+      "epoch": 0.18094089264173704,
+      "grad_norm": 0.4208106819963618,
+      "learning_rate": 3.012048192771085e-05,
+      "loss": 0.4932,
+      "step": 150
+    },
+    {
+      "epoch": 0.18214716525934863,
+      "grad_norm": 0.3494159427966441,
+      "learning_rate": 3.032128514056225e-05,
+      "loss": 0.4823,
+      "step": 151
+    },
+    {
+      "epoch": 0.18335343787696018,
+      "grad_norm": 0.3823494518546758,
+      "learning_rate": 3.052208835341366e-05,
+      "loss": 0.4897,
+      "step": 152
+    },
+    {
+      "epoch": 0.18455971049457176,
+      "grad_norm": 0.4115543529305647,
+      "learning_rate": 3.072289156626506e-05,
+      "loss": 0.4991,
+      "step": 153
+    },
+    {
+      "epoch": 0.18576598311218334,
+      "grad_norm": 0.3958327481485804,
+      "learning_rate": 3.092369477911647e-05,
+      "loss": 0.5016,
+      "step": 154
+    },
+    {
+      "epoch": 0.18697225572979492,
+      "grad_norm": 0.39504172903294904,
+      "learning_rate": 3.112449799196787e-05,
+      "loss": 0.4892,
+      "step": 155
+    },
+    {
+      "epoch": 0.1881785283474065,
+      "grad_norm": 0.41868991663521743,
+      "learning_rate": 3.132530120481928e-05,
+      "loss": 0.5055,
+      "step": 156
+    },
+    {
+      "epoch": 0.18938480096501809,
+      "grad_norm": 0.3515111350200484,
+      "learning_rate": 3.152610441767068e-05,
+      "loss": 0.497,
+      "step": 157
+    },
+    {
+      "epoch": 0.19059107358262967,
+      "grad_norm": 0.3997733425198974,
+      "learning_rate": 3.172690763052209e-05,
+      "loss": 0.4958,
+      "step": 158
+    },
+    {
+      "epoch": 0.19179734620024125,
+      "grad_norm": 0.4525184693608031,
+      "learning_rate": 3.192771084337349e-05,
+      "loss": 0.5212,
+      "step": 159
+    },
+    {
+      "epoch": 0.19300361881785283,
+      "grad_norm": 0.3902705819333036,
+      "learning_rate": 3.21285140562249e-05,
+      "loss": 0.5178,
+      "step": 160
+    },
+    {
+      "epoch": 0.1942098914354644,
+      "grad_norm": 0.5030587307913869,
+      "learning_rate": 3.232931726907631e-05,
+      "loss": 0.5108,
+      "step": 161
+    },
+    {
+      "epoch": 0.195416164053076,
+      "grad_norm": 0.3624196375454655,
+      "learning_rate": 3.253012048192771e-05,
+      "loss": 0.5097,
+      "step": 162
+    },
+    {
+      "epoch": 0.19662243667068757,
+      "grad_norm": 0.5221595299517153,
+      "learning_rate": 3.273092369477912e-05,
+      "loss": 0.5065,
+      "step": 163
+    },
+    {
+      "epoch": 0.19782870928829915,
+      "grad_norm": 0.40528581225043786,
+      "learning_rate": 3.2931726907630524e-05,
+      "loss": 0.4954,
+      "step": 164
+    },
+    {
+      "epoch": 0.19903498190591074,
+      "grad_norm": 0.4115164956369205,
+      "learning_rate": 3.313253012048193e-05,
+      "loss": 0.4924,
+      "step": 165
+    },
+    {
+      "epoch": 0.20024125452352232,
+      "grad_norm": 0.48390541674694987,
+      "learning_rate": 3.3333333333333335e-05,
+      "loss": 0.511,
+      "step": 166
+    },
+    {
+      "epoch": 0.2014475271411339,
+      "grad_norm": 0.4912910650075616,
+      "learning_rate": 3.353413654618474e-05,
+      "loss": 0.5156,
+      "step": 167
+    },
+    {
+      "epoch": 0.20265379975874548,
+      "grad_norm": 0.45223965087989415,
+      "learning_rate": 3.3734939759036146e-05,
+      "loss": 0.4873,
+      "step": 168
+    },
+    {
+      "epoch": 0.20386007237635706,
+      "grad_norm": 0.4462359140749097,
+      "learning_rate": 3.393574297188755e-05,
+      "loss": 0.4808,
+      "step": 169
+    },
+    {
+      "epoch": 0.20506634499396864,
+      "grad_norm": 0.45946545348291873,
+      "learning_rate": 3.413654618473896e-05,
+      "loss": 0.5084,
+      "step": 170
+    },
+    {
+      "epoch": 0.20627261761158022,
+      "grad_norm": 0.4869113169396108,
+      "learning_rate": 3.433734939759036e-05,
+      "loss": 0.5226,
+      "step": 171
+    },
+    {
+      "epoch": 0.2074788902291918,
+      "grad_norm": 0.4529756289005346,
+      "learning_rate": 3.4538152610441774e-05,
+      "loss": 0.4973,
+      "step": 172
+    },
+    {
+      "epoch": 0.20868516284680338,
+      "grad_norm": 0.4143744137360863,
+      "learning_rate": 3.473895582329317e-05,
+      "loss": 0.508,
+      "step": 173
+    },
+    {
+      "epoch": 0.20989143546441497,
+      "grad_norm": 0.4351511961848573,
+      "learning_rate": 3.4939759036144585e-05,
+      "loss": 0.4734,
+      "step": 174
+    },
+    {
+      "epoch": 0.21109770808202655,
+      "grad_norm": 0.357771198622376,
+      "learning_rate": 3.5140562248995983e-05,
+      "loss": 0.5004,
+      "step": 175
+    },
+    {
+      "epoch": 0.21230398069963813,
+      "grad_norm": 0.40605221616071163,
+      "learning_rate": 3.534136546184739e-05,
+      "loss": 0.4883,
+      "step": 176
+    },
+    {
+      "epoch": 0.2135102533172497,
+      "grad_norm": 0.4667261994786403,
+      "learning_rate": 3.5542168674698794e-05,
+      "loss": 0.4899,
+      "step": 177
+    },
+    {
+      "epoch": 0.2147165259348613,
+      "grad_norm": 0.3282260272132664,
+      "learning_rate": 3.57429718875502e-05,
+      "loss": 0.4847,
+      "step": 178
+    },
+    {
+      "epoch": 0.21592279855247287,
+      "grad_norm": 0.45132796886106885,
+      "learning_rate": 3.5943775100401605e-05,
+      "loss": 0.4853,
+      "step": 179
+    },
+    {
+      "epoch": 0.21712907117008443,
+      "grad_norm": 0.37769190784059187,
+      "learning_rate": 3.614457831325301e-05,
+      "loss": 0.4793,
+      "step": 180
+    },
+    {
+      "epoch": 0.218335343787696,
+      "grad_norm": 0.4105355197529797,
+      "learning_rate": 3.634538152610442e-05,
+      "loss": 0.4969,
+      "step": 181
+    },
+    {
+      "epoch": 0.2195416164053076,
+      "grad_norm": 0.4886523984077889,
+      "learning_rate": 3.654618473895582e-05,
+      "loss": 0.5004,
+      "step": 182
+    },
+    {
+      "epoch": 0.22074788902291917,
+      "grad_norm": 0.34526155199172764,
+      "learning_rate": 3.674698795180723e-05,
+      "loss": 0.5041,
+      "step": 183
+    },
+    {
+      "epoch": 0.22195416164053075,
+      "grad_norm": 0.536078133653887,
+      "learning_rate": 3.694779116465863e-05,
+      "loss": 0.4884,
+      "step": 184
+    },
+    {
+      "epoch": 0.22316043425814233,
+      "grad_norm": 0.41213708136617055,
+      "learning_rate": 3.7148594377510044e-05,
+      "loss": 0.4844,
+      "step": 185
+    },
+    {
+      "epoch": 0.2243667068757539,
+      "grad_norm": 0.42042027820779376,
+      "learning_rate": 3.734939759036144e-05,
+      "loss": 0.4894,
+      "step": 186
+    },
+    {
+      "epoch": 0.2255729794933655,
+      "grad_norm": 0.4622195360104783,
+      "learning_rate": 3.7550200803212855e-05,
+      "loss": 0.5162,
+      "step": 187
+    },
+    {
+      "epoch": 0.22677925211097708,
+      "grad_norm": 0.4754012578138651,
+      "learning_rate": 3.7751004016064253e-05,
+      "loss": 0.4942,
+      "step": 188
+    },
+    {
+      "epoch": 0.22798552472858866,
+      "grad_norm": 0.5123600834395359,
+      "learning_rate": 3.7951807228915666e-05,
+      "loss": 0.5037,
+      "step": 189
+    },
+    {
+      "epoch": 0.22919179734620024,
+      "grad_norm": 0.4950795192470773,
+      "learning_rate": 3.815261044176707e-05,
+      "loss": 0.5062,
+      "step": 190
+    },
+    {
+      "epoch": 0.23039806996381182,
+      "grad_norm": 0.5675637905005346,
+      "learning_rate": 3.8353413654618476e-05,
+      "loss": 0.4962,
+      "step": 191
+    },
+    {
+      "epoch": 0.2316043425814234,
+      "grad_norm": 0.46625154792702506,
+      "learning_rate": 3.855421686746988e-05,
+      "loss": 0.4899,
+      "step": 192
+    },
+    {
+      "epoch": 0.23281061519903498,
+      "grad_norm": 0.5285632109360143,
+      "learning_rate": 3.875502008032129e-05,
+      "loss": 0.471,
+      "step": 193
+    },
+    {
+      "epoch": 0.23401688781664656,
+      "grad_norm": 0.7098580123670813,
+      "learning_rate": 3.895582329317269e-05,
+      "loss": 0.4902,
+      "step": 194
+    },
+    {
+      "epoch": 0.23522316043425814,
+      "grad_norm": 0.46566792189515693,
+      "learning_rate": 3.91566265060241e-05,
+      "loss": 0.4639,
+      "step": 195
+    },
+    {
+      "epoch": 0.23642943305186973,
+      "grad_norm": 0.48790846789133946,
+      "learning_rate": 3.93574297188755e-05,
+      "loss": 0.4687,
+      "step": 196
+    },
+    {
+      "epoch": 0.2376357056694813,
+      "grad_norm": 0.6616199623175774,
+      "learning_rate": 3.955823293172691e-05,
+      "loss": 0.4896,
+      "step": 197
+    },
+    {
+      "epoch": 0.2388419782870929,
+      "grad_norm": 0.6597217338301457,
+      "learning_rate": 3.9759036144578314e-05,
+      "loss": 0.5305,
+      "step": 198
+    },
+    {
+      "epoch": 0.24004825090470447,
+      "grad_norm": 0.40911035138284585,
+      "learning_rate": 3.995983935742972e-05,
+      "loss": 0.4845,
+      "step": 199
+    },
+    {
+      "epoch": 0.24125452352231605,
+      "grad_norm": 0.41051379447376024,
+      "learning_rate": 4.0160642570281125e-05,
+      "loss": 0.4943,
+      "step": 200
+    },
+    {
+      "epoch": 0.24246079613992763,
+      "grad_norm": 0.4856554636844061,
+      "learning_rate": 4.036144578313254e-05,
+      "loss": 0.4589,
+      "step": 201
+    },
+    {
+      "epoch": 0.2436670687575392,
+      "grad_norm": 0.46595255361079024,
+      "learning_rate": 4.0562248995983936e-05,
+      "loss": 0.4901,
+      "step": 202
+    },
+    {
+      "epoch": 0.2448733413751508,
+      "grad_norm": 0.4817145604840749,
+      "learning_rate": 4.076305220883535e-05,
+      "loss": 0.4838,
+      "step": 203
+    },
+    {
+      "epoch": 0.24607961399276237,
+      "grad_norm": 0.4386965552097178,
+      "learning_rate": 4.0963855421686746e-05,
+      "loss": 0.4851,
+      "step": 204
+    },
+    {
+      "epoch": 0.24728588661037396,
+      "grad_norm": 0.4520118344205907,
+      "learning_rate": 4.116465863453816e-05,
+      "loss": 0.4983,
+      "step": 205
+    },
+    {
+      "epoch": 0.24849215922798554,
+      "grad_norm": 0.3614677524759272,
+      "learning_rate": 4.136546184738956e-05,
+      "loss": 0.4679,
+      "step": 206
+    },
+    {
+      "epoch": 0.24969843184559712,
+      "grad_norm": 0.49375547406597686,
+      "learning_rate": 4.156626506024097e-05,
+      "loss": 0.4667,
+      "step": 207
+    },
+    {
+      "epoch": 0.25090470446320867,
+      "grad_norm": 0.38672197120064633,
+      "learning_rate": 4.176706827309237e-05,
+      "loss": 0.4725,
+      "step": 208
+    },
+    {
+      "epoch": 0.25211097708082025,
+      "grad_norm": 0.5147301761091354,
+      "learning_rate": 4.196787148594378e-05,
+      "loss": 0.4781,
+      "step": 209
+    },
+    {
+      "epoch": 0.25331724969843183,
+      "grad_norm": 0.3812166626819943,
+      "learning_rate": 4.2168674698795186e-05,
+      "loss": 0.4805,
+      "step": 210
+    },
+    {
+      "epoch": 0.2545235223160434,
+      "grad_norm": 0.4545562630855821,
+      "learning_rate": 4.236947791164659e-05,
+      "loss": 0.4762,
+      "step": 211
+    },
+    {
+      "epoch": 0.255729794933655,
+      "grad_norm": 0.436269716338551,
+      "learning_rate": 4.2570281124497996e-05,
+      "loss": 0.4797,
+      "step": 212
+    },
+    {
+      "epoch": 0.2569360675512666,
+      "grad_norm": 0.5384572866348979,
+      "learning_rate": 4.27710843373494e-05,
+      "loss": 0.4999,
+      "step": 213
+    },
+    {
+      "epoch": 0.25814234016887816,
+      "grad_norm": 0.5404751631725132,
+      "learning_rate": 4.297188755020081e-05,
+      "loss": 0.4798,
+      "step": 214
+    },
+    {
+      "epoch": 0.25934861278648974,
+      "grad_norm": 0.4429509321892677,
+      "learning_rate": 4.317269076305221e-05,
+      "loss": 0.5232,
+      "step": 215
+    },
+    {
+      "epoch": 0.2605548854041013,
+      "grad_norm": 0.5303855977581109,
+      "learning_rate": 4.337349397590362e-05,
+      "loss": 0.4884,
+      "step": 216
+    },
+    {
+      "epoch": 0.2617611580217129,
+      "grad_norm": 0.43125896662479185,
+      "learning_rate": 4.357429718875502e-05,
+      "loss": 0.4828,
+      "step": 217
+    },
+    {
+      "epoch": 0.2629674306393245,
+      "grad_norm": 0.438762700628616,
+      "learning_rate": 4.377510040160643e-05,
+      "loss": 0.4838,
+      "step": 218
+    },
+    {
+      "epoch": 0.26417370325693607,
+      "grad_norm": 0.4190780132169102,
+      "learning_rate": 4.3975903614457834e-05,
+      "loss": 0.4702,
+      "step": 219
+    },
+    {
+      "epoch": 0.26537997587454765,
+      "grad_norm": 0.3773440132348853,
+      "learning_rate": 4.417670682730924e-05,
+      "loss": 0.4942,
+      "step": 220
+    },
+    {
+      "epoch": 0.26658624849215923,
+      "grad_norm": 0.47799490507514975,
+      "learning_rate": 4.4377510040160645e-05,
+      "loss": 0.4698,
+      "step": 221
+    },
+    {
+      "epoch": 0.2677925211097708,
+      "grad_norm": 0.37758247336865564,
+      "learning_rate": 4.457831325301205e-05,
+      "loss": 0.4638,
+      "step": 222
+    },
+    {
+      "epoch": 0.2689987937273824,
+      "grad_norm": 0.4474146827235853,
+      "learning_rate": 4.4779116465863456e-05,
+      "loss": 0.4852,
+      "step": 223
+    },
+    {
+      "epoch": 0.27020506634499397,
+      "grad_norm": 0.47716587130345856,
+      "learning_rate": 4.497991967871486e-05,
+      "loss": 0.4587,
+      "step": 224
+    },
+    {
+      "epoch": 0.27141133896260555,
+      "grad_norm": 0.5083388261649718,
+      "learning_rate": 4.5180722891566266e-05,
+      "loss": 0.4734,
+      "step": 225
+    },
+    {
+      "epoch": 0.27261761158021713,
+      "grad_norm": 0.46930543099228017,
+      "learning_rate": 4.538152610441767e-05,
+      "loss": 0.4749,
+      "step": 226
+    },
+    {
+      "epoch": 0.2738238841978287,
+      "grad_norm": 0.44360824898966233,
+      "learning_rate": 4.558232931726908e-05,
+      "loss": 0.488,
+      "step": 227
+    },
+    {
+      "epoch": 0.2750301568154403,
+      "grad_norm": 0.39902429944785056,
+      "learning_rate": 4.578313253012048e-05,
+      "loss": 0.4561,
+      "step": 228
+    },
+    {
+      "epoch": 0.2762364294330519,
+      "grad_norm": 0.4597398820977258,
+      "learning_rate": 4.598393574297189e-05,
+      "loss": 0.4593,
+      "step": 229
+    },
+    {
+      "epoch": 0.27744270205066346,
+      "grad_norm": 0.5413711861567382,
+      "learning_rate": 4.61847389558233e-05,
+      "loss": 0.4749,
+      "step": 230
+    },
+    {
+      "epoch": 0.27864897466827504,
+      "grad_norm": 0.4631219743214568,
+      "learning_rate": 4.63855421686747e-05,
+      "loss": 0.4864,
+      "step": 231
+    },
+    {
+      "epoch": 0.2798552472858866,
+      "grad_norm": 0.4239404274562551,
+      "learning_rate": 4.658634538152611e-05,
+      "loss": 0.5028,
+      "step": 232
+    },
+    {
+      "epoch": 0.2810615199034982,
+      "grad_norm": 0.512533107003274,
+      "learning_rate": 4.678714859437751e-05,
+      "loss": 0.4529,
+      "step": 233
+    },
+    {
+      "epoch": 0.2822677925211098,
+      "grad_norm": 0.6384568999031756,
+      "learning_rate": 4.698795180722892e-05,
+      "loss": 0.4892,
+      "step": 234
+    },
+    {
+      "epoch": 0.28347406513872137,
+      "grad_norm": 0.7898314343362617,
+      "learning_rate": 4.718875502008032e-05,
+      "loss": 0.475,
+      "step": 235
+    },
+    {
+      "epoch": 0.28468033775633295,
+      "grad_norm": 0.5687679404684749,
+      "learning_rate": 4.738955823293173e-05,
+      "loss": 0.4751,
+      "step": 236
+    },
+    {
+      "epoch": 0.2858866103739445,
+      "grad_norm": 0.6318324876459296,
+      "learning_rate": 4.759036144578313e-05,
+      "loss": 0.4948,
+      "step": 237
+    },
+    {
+      "epoch": 0.2870928829915561,
+      "grad_norm": 0.6411579888771047,
+      "learning_rate": 4.779116465863454e-05,
+      "loss": 0.4862,
+      "step": 238
+    },
+    {
+      "epoch": 0.2882991556091677,
+      "grad_norm": 0.6548791719603922,
+      "learning_rate": 4.799196787148594e-05,
+      "loss": 0.4707,
+      "step": 239
+    },
+    {
+      "epoch": 0.28950542822677927,
+      "grad_norm": 0.48721678762737824,
+      "learning_rate": 4.8192771084337354e-05,
+      "loss": 0.4745,
+      "step": 240
+    },
+    {
+      "epoch": 0.29071170084439085,
+      "grad_norm": 0.585841744027512,
+      "learning_rate": 4.839357429718876e-05,
+      "loss": 0.4771,
+      "step": 241
+    },
+    {
+      "epoch": 0.29191797346200243,
+      "grad_norm": 0.7343233388019949,
+      "learning_rate": 4.8594377510040165e-05,
+      "loss": 0.4586,
+      "step": 242
+    },
+    {
+      "epoch": 0.293124246079614,
+      "grad_norm": 0.7306993109383978,
+      "learning_rate": 4.879518072289157e-05,
+      "loss": 0.5018,
+      "step": 243
+    },
+    {
+      "epoch": 0.2943305186972256,
+      "grad_norm": 0.5752152211307707,
+      "learning_rate": 4.8995983935742975e-05,
+      "loss": 0.5027,
+      "step": 244
+    },
+    {
+      "epoch": 0.2955367913148372,
+      "grad_norm": 0.6954347040340547,
+      "learning_rate": 4.919678714859438e-05,
+      "loss": 0.4779,
+      "step": 245
+    },
+    {
+      "epoch": 0.29674306393244876,
+      "grad_norm": 0.9605672172965798,
+      "learning_rate": 4.9397590361445786e-05,
+      "loss": 0.495,
+      "step": 246
+    },
+    {
+      "epoch": 0.29794933655006034,
+      "grad_norm": 0.48111198217867024,
+      "learning_rate": 4.959839357429719e-05,
+      "loss": 0.4774,
+      "step": 247
+    },
+    {
+      "epoch": 0.2991556091676719,
+      "grad_norm": 0.6959734943206007,
+      "learning_rate": 4.97991967871486e-05,
+      "loss": 0.4748,
+      "step": 248
+    },
+    {
+      "epoch": 0.30036188178528345,
+      "grad_norm": 0.6307753559437208,
+      "learning_rate": 5e-05,
+      "loss": 0.4556,
+      "step": 249
+    },
+    {
+      "epoch": 0.30156815440289503,
+      "grad_norm": 0.5271890748977731,
+      "learning_rate": 4.997765862377123e-05,
+      "loss": 0.4959,
+      "step": 250
+    },
+    {
+      "epoch": 0.3027744270205066,
+      "grad_norm": 0.4954669571709612,
+      "learning_rate": 4.995531724754245e-05,
+      "loss": 0.4636,
+      "step": 251
+    },
+    {
+      "epoch": 0.3039806996381182,
+      "grad_norm": 0.5362954459267386,
+      "learning_rate": 4.993297587131368e-05,
+      "loss": 0.4625,
+      "step": 252
+    },
+    {
+      "epoch": 0.30518697225572977,
+      "grad_norm": 0.5663830578871994,
+      "learning_rate": 4.99106344950849e-05,
+      "loss": 0.462,
+      "step": 253
+    },
+    {
+      "epoch": 0.30639324487334135,
+      "grad_norm": 0.5166937551194555,
+      "learning_rate": 4.988829311885612e-05,
+      "loss": 0.48,
+      "step": 254
+    },
+    {
+      "epoch": 0.30759951749095293,
+      "grad_norm": 0.45080582970761324,
+      "learning_rate": 4.986595174262735e-05,
+      "loss": 0.4765,
+      "step": 255
+    },
+    {
+      "epoch": 0.3088057901085645,
+      "grad_norm": 0.5488354412984369,
+      "learning_rate": 4.9843610366398574e-05,
+      "loss": 0.4824,
+      "step": 256
+    },
+    {
+      "epoch": 0.3100120627261761,
+      "grad_norm": 0.5669400193473989,
+      "learning_rate": 4.9821268990169795e-05,
+      "loss": 0.4503,
+      "step": 257
+    },
+    {
+      "epoch": 0.3112183353437877,
+      "grad_norm": 0.5115485943428845,
+      "learning_rate": 4.979892761394102e-05,
+      "loss": 0.4691,
+      "step": 258
+    },
+    {
+      "epoch": 0.31242460796139926,
+      "grad_norm": 0.5608966206639503,
+      "learning_rate": 4.977658623771225e-05,
+      "loss": 0.463,
+      "step": 259
+    },
+    {
+      "epoch": 0.31363088057901084,
+      "grad_norm": 0.4506008486154186,
+      "learning_rate": 4.975424486148347e-05,
+      "loss": 0.4882,
+      "step": 260
+    },
+    {
+      "epoch": 0.3148371531966224,
+      "grad_norm": 0.6109184182493657,
+      "learning_rate": 4.973190348525469e-05,
+      "loss": 0.4814,
+      "step": 261
+    },
+    {
+      "epoch": 0.316043425814234,
+      "grad_norm": 0.44011566912817257,
+      "learning_rate": 4.970956210902592e-05,
+      "loss": 0.4702,
+      "step": 262
+    },
+    {
+      "epoch": 0.3172496984318456,
+      "grad_norm": 0.5082547699158838,
+      "learning_rate": 4.968722073279714e-05,
+      "loss": 0.4676,
+      "step": 263
+    },
+    {
+      "epoch": 0.31845597104945716,
+      "grad_norm": 0.4634121406839268,
+      "learning_rate": 4.966487935656837e-05,
+      "loss": 0.4814,
+      "step": 264
+    },
+    {
+      "epoch": 0.31966224366706875,
+      "grad_norm": 0.4613999245360435,
+      "learning_rate": 4.9642537980339595e-05,
+      "loss": 0.4892,
+      "step": 265
+    },
+    {
+      "epoch": 0.3208685162846803,
+      "grad_norm": 0.5407571756739682,
+      "learning_rate": 4.9620196604110815e-05,
+      "loss": 0.4765,
+      "step": 266
+    },
+    {
+      "epoch": 0.3220747889022919,
+      "grad_norm": 0.5588948249098938,
+      "learning_rate": 4.959785522788204e-05,
+      "loss": 0.4434,
+      "step": 267
+    },
+    {
+      "epoch": 0.3232810615199035,
+      "grad_norm": 0.48459990169582196,
+      "learning_rate": 4.9575513851653263e-05,
+      "loss": 0.482,
+      "step": 268
+    },
+    {
+      "epoch": 0.32448733413751507,
+      "grad_norm": 0.5203680763697215,
+      "learning_rate": 4.9553172475424484e-05,
+      "loss": 0.4561,
+      "step": 269
+    },
+    {
+      "epoch": 0.32569360675512665,
+      "grad_norm": 0.49517970790477517,
+      "learning_rate": 4.953083109919571e-05,
+      "loss": 0.4463,
+      "step": 270
+    },
+    {
+      "epoch": 0.32689987937273823,
+      "grad_norm": 0.4496993668081543,
+      "learning_rate": 4.950848972296694e-05,
+      "loss": 0.4866,
+      "step": 271
+    },
+    {
+      "epoch": 0.3281061519903498,
+      "grad_norm": 0.5198314152754191,
+      "learning_rate": 4.948614834673816e-05,
+      "loss": 0.4736,
+      "step": 272
+    },
+    {
+      "epoch": 0.3293124246079614,
+      "grad_norm": 0.48165421751496224,
+      "learning_rate": 4.946380697050939e-05,
+      "loss": 0.4644,
+      "step": 273
+    },
+    {
+      "epoch": 0.330518697225573,
+      "grad_norm": 0.5017150374814877,
+      "learning_rate": 4.9441465594280615e-05,
+      "loss": 0.4714,
+      "step": 274
+    },
+    {
+      "epoch": 0.33172496984318456,
+      "grad_norm": 0.5125463646764701,
+      "learning_rate": 4.9419124218051835e-05,
+      "loss": 0.4783,
+      "step": 275
+    },
+    {
+      "epoch": 0.33293124246079614,
+      "grad_norm": 0.5118862848673149,
+      "learning_rate": 4.9396782841823056e-05,
+      "loss": 0.4506,
+      "step": 276
+    },
+    {
+      "epoch": 0.3341375150784077,
+      "grad_norm": 0.494848299345708,
+      "learning_rate": 4.9374441465594284e-05,
+      "loss": 0.4666,
+      "step": 277
+    },
+    {
+      "epoch": 0.3353437876960193,
+      "grad_norm": 0.574556205294029,
+      "learning_rate": 4.9352100089365504e-05,
+      "loss": 0.4715,
+      "step": 278
+    },
+    {
+      "epoch": 0.3365500603136309,
+      "grad_norm": 0.5013842930456912,
+      "learning_rate": 4.932975871313673e-05,
+      "loss": 0.4823,
+      "step": 279
+    },
+    {
+      "epoch": 0.33775633293124246,
+      "grad_norm": 0.575641680135763,
+      "learning_rate": 4.930741733690796e-05,
+      "loss": 0.4769,
+      "step": 280
+    },
+    {
+      "epoch": 0.33896260554885405,
+      "grad_norm": 0.5073551828383931,
+      "learning_rate": 4.928507596067918e-05,
+      "loss": 0.4687,
+      "step": 281
+    },
+    {
+      "epoch": 0.3401688781664656,
+      "grad_norm": 0.6213137019127235,
+      "learning_rate": 4.92627345844504e-05,
+      "loss": 0.4912,
+      "step": 282
+    },
+    {
+      "epoch": 0.3413751507840772,
+      "grad_norm": 0.4650384893742426,
+      "learning_rate": 4.924039320822163e-05,
+      "loss": 0.4686,
+      "step": 283
+    },
+    {
+      "epoch": 0.3425814234016888,
+      "grad_norm": 0.411216907390994,
+      "learning_rate": 4.921805183199285e-05,
+      "loss": 0.4543,
+      "step": 284
+    },
+    {
+      "epoch": 0.34378769601930037,
+      "grad_norm": 0.462814550454254,
+      "learning_rate": 4.9195710455764076e-05,
+      "loss": 0.4691,
+      "step": 285
+    },
+    {
+      "epoch": 0.34499396863691195,
+      "grad_norm": 0.45055083119935785,
+      "learning_rate": 4.9173369079535304e-05,
+      "loss": 0.4648,
+      "step": 286
+    },
+    {
+      "epoch": 0.34620024125452353,
+      "grad_norm": 0.4070711196003684,
+      "learning_rate": 4.9151027703306525e-05,
+      "loss": 0.4558,
+      "step": 287
+    },
+    {
+      "epoch": 0.3474065138721351,
+      "grad_norm": 0.38539662658940144,
+      "learning_rate": 4.912868632707775e-05,
+      "loss": 0.4802,
+      "step": 288
+    },
+    {
+      "epoch": 0.3486127864897467,
+      "grad_norm": 0.5046587298061578,
+      "learning_rate": 4.910634495084897e-05,
+      "loss": 0.4901,
+      "step": 289
+    },
+    {
+      "epoch": 0.3498190591073583,
+      "grad_norm": 0.45014404701699506,
+      "learning_rate": 4.9084003574620193e-05,
+      "loss": 0.439,
+      "step": 290
+    },
+    {
+      "epoch": 0.35102533172496986,
+      "grad_norm": 0.47834117803538234,
+      "learning_rate": 4.906166219839142e-05,
+      "loss": 0.4716,
+      "step": 291
+    },
+    {
+      "epoch": 0.35223160434258144,
+      "grad_norm": 0.6403594697917471,
+      "learning_rate": 4.903932082216265e-05,
+      "loss": 0.4449,
+      "step": 292
+    },
+    {
+      "epoch": 0.353437876960193,
+      "grad_norm": 0.5545894077440268,
+      "learning_rate": 4.901697944593387e-05,
+      "loss": 0.4991,
+      "step": 293
+    },
+    {
+      "epoch": 0.3546441495778046,
+      "grad_norm": 0.48207085217428985,
+      "learning_rate": 4.8994638069705097e-05,
+      "loss": 0.4424,
+      "step": 294
+    },
+    {
+      "epoch": 0.3558504221954162,
+      "grad_norm": 0.6531795184939722,
+      "learning_rate": 4.8972296693476324e-05,
+      "loss": 0.459,
+      "step": 295
+    },
+    {
+      "epoch": 0.35705669481302776,
+      "grad_norm": 0.658758404071505,
+      "learning_rate": 4.8949955317247545e-05,
+      "loss": 0.4812,
+      "step": 296
+    },
+    {
+      "epoch": 0.35826296743063935,
+      "grad_norm": 0.594767575961469,
+      "learning_rate": 4.8927613941018765e-05,
+      "loss": 0.4665,
+      "step": 297
+    },
+    {
+      "epoch": 0.3594692400482509,
+      "grad_norm": 0.5202270329599545,
+      "learning_rate": 4.890527256478999e-05,
+      "loss": 0.4507,
+      "step": 298
+    },
+    {
+      "epoch": 0.3606755126658625,
+      "grad_norm": 0.66410947267989,
+      "learning_rate": 4.8882931188561214e-05,
+      "loss": 0.4538,
+      "step": 299
+    },
+    {
+      "epoch": 0.3618817852834741,
+      "grad_norm": 0.688835455933309,
+      "learning_rate": 4.886058981233244e-05,
+      "loss": 0.4563,
+      "step": 300
+    },
+    {
+      "epoch": 0.36308805790108567,
+      "grad_norm": 0.44276594605737246,
+      "learning_rate": 4.883824843610367e-05,
+      "loss": 0.4669,
+      "step": 301
+    },
+    {
+      "epoch": 0.36429433051869725,
+      "grad_norm": 0.6244042925092166,
+      "learning_rate": 4.881590705987489e-05,
+      "loss": 0.4597,
+      "step": 302
+    },
+    {
+      "epoch": 0.36550060313630883,
+      "grad_norm": 0.4282473469132016,
+      "learning_rate": 4.879356568364612e-05,
+      "loss": 0.4631,
+      "step": 303
+    },
+    {
+      "epoch": 0.36670687575392036,
+      "grad_norm": 0.49393564380494653,
+      "learning_rate": 4.877122430741734e-05,
+      "loss": 0.4622,
+      "step": 304
+    },
+    {
+      "epoch": 0.36791314837153194,
+      "grad_norm": 0.5754977153666446,
+      "learning_rate": 4.874888293118856e-05,
+      "loss": 0.4503,
+      "step": 305
+    },
+    {
+      "epoch": 0.3691194209891435,
+      "grad_norm": 0.43417824463420274,
+      "learning_rate": 4.8726541554959786e-05,
+      "loss": 0.4685,
+      "step": 306
+    },
+    {
+      "epoch": 0.3703256936067551,
+      "grad_norm": 0.5493384400309728,
+      "learning_rate": 4.870420017873101e-05,
+      "loss": 0.4484,
+      "step": 307
+    },
+    {
+      "epoch": 0.3715319662243667,
+      "grad_norm": 0.563798258109028,
+      "learning_rate": 4.8681858802502234e-05,
+      "loss": 0.4866,
+      "step": 308
+    },
+    {
+      "epoch": 0.37273823884197826,
+      "grad_norm": 0.40236230684499746,
+      "learning_rate": 4.865951742627346e-05,
+      "loss": 0.4426,
+      "step": 309
+    },
+    {
+      "epoch": 0.37394451145958985,
+      "grad_norm": 0.4610270253104838,
+      "learning_rate": 4.863717605004469e-05,
+      "loss": 0.4381,
+      "step": 310
+    },
+    {
+      "epoch": 0.3751507840772014,
+      "grad_norm": 0.47654001944242114,
+      "learning_rate": 4.861483467381591e-05,
+      "loss": 0.4655,
+      "step": 311
+    },
+    {
+      "epoch": 0.376357056694813,
+      "grad_norm": 0.6151547050937779,
+      "learning_rate": 4.859249329758713e-05,
+      "loss": 0.472,
+      "step": 312
+    },
+    {
+      "epoch": 0.3775633293124246,
+      "grad_norm": 0.45232137136900996,
+      "learning_rate": 4.857015192135836e-05,
+      "loss": 0.46,
+      "step": 313
+    },
+    {
+      "epoch": 0.37876960193003617,
+      "grad_norm": 0.4506326312132236,
+      "learning_rate": 4.8547810545129585e-05,
+      "loss": 0.4573,
+      "step": 314
+    },
+    {
+      "epoch": 0.37997587454764775,
+      "grad_norm": 0.4983600614727651,
+      "learning_rate": 4.8525469168900806e-05,
+      "loss": 0.4748,
+      "step": 315
+    },
+    {
+      "epoch": 0.38118214716525933,
+      "grad_norm": 0.3879122279776259,
+      "learning_rate": 4.850312779267203e-05,
+      "loss": 0.4484,
+      "step": 316
+    },
+    {
+      "epoch": 0.3823884197828709,
+      "grad_norm": 0.4848285972697871,
+      "learning_rate": 4.848078641644326e-05,
+      "loss": 0.45,
+      "step": 317
+    },
+    {
+      "epoch": 0.3835946924004825,
+      "grad_norm": 0.4127959237985656,
+      "learning_rate": 4.845844504021448e-05,
+      "loss": 0.4388,
+      "step": 318
+    },
+    {
+      "epoch": 0.3848009650180941,
+      "grad_norm": 0.4474884654687781,
+      "learning_rate": 4.84361036639857e-05,
+      "loss": 0.5006,
+      "step": 319
+    },
+    {
+      "epoch": 0.38600723763570566,
+      "grad_norm": 0.4166281609224283,
+      "learning_rate": 4.841376228775693e-05,
+      "loss": 0.451,
+      "step": 320
+    },
+    {
+      "epoch": 0.38721351025331724,
+      "grad_norm": 0.6167629036074882,
+      "learning_rate": 4.839142091152815e-05,
+      "loss": 0.4808,
+      "step": 321
+    },
+    {
+      "epoch": 0.3884197828709288,
+      "grad_norm": 0.46452397726219896,
+      "learning_rate": 4.836907953529938e-05,
+      "loss": 0.4519,
+      "step": 322
+    },
+    {
+      "epoch": 0.3896260554885404,
+      "grad_norm": 0.5077639670260757,
+      "learning_rate": 4.8346738159070605e-05,
+      "loss": 0.453,
+      "step": 323
+    },
+    {
+      "epoch": 0.390832328106152,
+      "grad_norm": 0.4492633198717073,
+      "learning_rate": 4.8324396782841826e-05,
+      "loss": 0.448,
+      "step": 324
+    },
+    {
+      "epoch": 0.39203860072376356,
+      "grad_norm": 0.5119837268347922,
+      "learning_rate": 4.8302055406613053e-05,
+      "loss": 0.4703,
+      "step": 325
+    },
+    {
+      "epoch": 0.39324487334137515,
+      "grad_norm": 0.46324744252555106,
+      "learning_rate": 4.8279714030384274e-05,
+      "loss": 0.4407,
+      "step": 326
+    },
+    {
+      "epoch": 0.3944511459589867,
+      "grad_norm": 0.4894638654542045,
+      "learning_rate": 4.8257372654155495e-05,
+      "loss": 0.4733,
+      "step": 327
+    },
+    {
+      "epoch": 0.3956574185765983,
+      "grad_norm": 0.46894954218311713,
+      "learning_rate": 4.823503127792672e-05,
+      "loss": 0.4617,
+      "step": 328
+    },
+    {
+      "epoch": 0.3968636911942099,
+      "grad_norm": 0.45217642275344966,
+      "learning_rate": 4.821268990169795e-05,
+      "loss": 0.464,
+      "step": 329
+    },
+    {
+      "epoch": 0.39806996381182147,
+      "grad_norm": 0.4961460092686353,
+      "learning_rate": 4.819034852546917e-05,
+      "loss": 0.461,
+      "step": 330
+    },
+    {
+      "epoch": 0.39927623642943305,
+      "grad_norm": 0.46595217657545673,
+      "learning_rate": 4.81680071492404e-05,
+      "loss": 0.4704,
+      "step": 331
+    },
+    {
+      "epoch": 0.40048250904704463,
+      "grad_norm": 0.5559227542572431,
+      "learning_rate": 4.8145665773011625e-05,
+      "loss": 0.4903,
+      "step": 332
+    },
+    {
+      "epoch": 0.4016887816646562,
+      "grad_norm": 0.4948787429742588,
+      "learning_rate": 4.8123324396782846e-05,
+      "loss": 0.4715,
+      "step": 333
+    },
+    {
+      "epoch": 0.4028950542822678,
+      "grad_norm": 0.4970808248520605,
+      "learning_rate": 4.810098302055407e-05,
+      "loss": 0.4649,
+      "step": 334
+    },
+    {
+      "epoch": 0.4041013268998794,
+      "grad_norm": 0.4273747126680364,
+      "learning_rate": 4.8078641644325294e-05,
+      "loss": 0.4493,
+      "step": 335
+    },
+    {
+      "epoch": 0.40530759951749096,
+      "grad_norm": 0.50583259429466,
+      "learning_rate": 4.8056300268096515e-05,
+      "loss": 0.4623,
+      "step": 336
+    },
+    {
+      "epoch": 0.40651387213510254,
+      "grad_norm": 0.4297375922857019,
+      "learning_rate": 4.803395889186774e-05,
+      "loss": 0.4526,
+      "step": 337
+    },
+    {
+      "epoch": 0.4077201447527141,
+      "grad_norm": 0.35598323627722095,
+      "learning_rate": 4.801161751563897e-05,
+      "loss": 0.417,
+      "step": 338
+    },
+    {
+      "epoch": 0.4089264173703257,
+      "grad_norm": 0.39935840283688373,
+      "learning_rate": 4.798927613941019e-05,
+      "loss": 0.4326,
+      "step": 339
+    },
+    {
+      "epoch": 0.4101326899879373,
+      "grad_norm": 0.3622581622120651,
+      "learning_rate": 4.796693476318142e-05,
+      "loss": 0.4715,
+      "step": 340
+    },
+    {
+      "epoch": 0.41133896260554886,
+      "grad_norm": 0.39214159703489104,
+      "learning_rate": 4.794459338695264e-05,
+      "loss": 0.4603,
+      "step": 341
+    },
+    {
+      "epoch": 0.41254523522316044,
+      "grad_norm": 0.386458432467993,
+      "learning_rate": 4.792225201072386e-05,
+      "loss": 0.4555,
+      "step": 342
+    },
+    {
+      "epoch": 0.413751507840772,
+      "grad_norm": 0.41568373206095827,
+      "learning_rate": 4.789991063449509e-05,
+      "loss": 0.4512,
+      "step": 343
+    },
+    {
+      "epoch": 0.4149577804583836,
+      "grad_norm": 0.360509732388755,
+      "learning_rate": 4.7877569258266315e-05,
+      "loss": 0.469,
+      "step": 344
+    },
+    {
+      "epoch": 0.4161640530759952,
+      "grad_norm": 0.39515906701596504,
+      "learning_rate": 4.7855227882037535e-05,
+      "loss": 0.443,
+      "step": 345
+    },
+    {
+      "epoch": 0.41737032569360677,
+      "grad_norm": 0.34843667675457657,
+      "learning_rate": 4.783288650580876e-05,
+      "loss": 0.4512,
+      "step": 346
+    },
+    {
+      "epoch": 0.41857659831121835,
+      "grad_norm": 0.4195610422690541,
+      "learning_rate": 4.781054512957998e-05,
+      "loss": 0.457,
+      "step": 347
+    },
+    {
+      "epoch": 0.41978287092882993,
+      "grad_norm": 0.37324898276685664,
+      "learning_rate": 4.7788203753351204e-05,
+      "loss": 0.4547,
+      "step": 348
+    },
+    {
+      "epoch": 0.4209891435464415,
+      "grad_norm": 0.5536452542682623,
+      "learning_rate": 4.776586237712243e-05,
+      "loss": 0.467,
+      "step": 349
+    },
+    {
+      "epoch": 0.4221954161640531,
+      "grad_norm": 0.39285135804997573,
+      "learning_rate": 4.774352100089366e-05,
+      "loss": 0.4462,
+      "step": 350
+    },
+    {
+      "epoch": 0.4234016887816647,
+      "grad_norm": 0.48757905541108004,
+      "learning_rate": 4.772117962466488e-05,
+      "loss": 0.4537,
+      "step": 351
+    },
+    {
+      "epoch": 0.42460796139927626,
+      "grad_norm": 0.3968507691375903,
+      "learning_rate": 4.769883824843611e-05,
+      "loss": 0.4656,
+      "step": 352
+    },
+    {
+      "epoch": 0.42581423401688784,
+      "grad_norm": 0.521932683169365,
+      "learning_rate": 4.7676496872207335e-05,
+      "loss": 0.4346,
+      "step": 353
+    },
+    {
+      "epoch": 0.4270205066344994,
+      "grad_norm": 0.47502250284227127,
+      "learning_rate": 4.7654155495978555e-05,
+      "loss": 0.4586,
+      "step": 354
+    },
+    {
+      "epoch": 0.428226779252111,
+      "grad_norm": 0.4132516632500004,
+      "learning_rate": 4.7631814119749776e-05,
+      "loss": 0.4771,
+      "step": 355
+    },
+    {
+      "epoch": 0.4294330518697226,
+      "grad_norm": 0.4956716038670087,
+      "learning_rate": 4.7609472743521004e-05,
+      "loss": 0.5001,
+      "step": 356
+    },
+    {
+      "epoch": 0.43063932448733416,
+      "grad_norm": 0.43933494549593005,
+      "learning_rate": 4.7587131367292224e-05,
+      "loss": 0.4507,
+      "step": 357
+    },
+    {
+      "epoch": 0.43184559710494574,
+      "grad_norm": 0.4535428967390594,
+      "learning_rate": 4.756478999106345e-05,
+      "loss": 0.4487,
+      "step": 358
+    },
+    {
+      "epoch": 0.4330518697225573,
+      "grad_norm": 0.43344450609824314,
+      "learning_rate": 4.754244861483468e-05,
+      "loss": 0.4434,
+      "step": 359
+    },
+    {
+      "epoch": 0.43425814234016885,
+      "grad_norm": 0.39608164974537874,
+      "learning_rate": 4.75201072386059e-05,
+      "loss": 0.4463,
+      "step": 360
+    },
+    {
+      "epoch": 0.43546441495778043,
+      "grad_norm": 0.4534682933209474,
+      "learning_rate": 4.749776586237713e-05,
+      "loss": 0.4704,
+      "step": 361
+    },
+    {
+      "epoch": 0.436670687575392,
+      "grad_norm": 0.40734523179872223,
+      "learning_rate": 4.747542448614835e-05,
+      "loss": 0.473,
+      "step": 362
+    },
+    {
+      "epoch": 0.4378769601930036,
+      "grad_norm": 0.41613899892629674,
+      "learning_rate": 4.745308310991957e-05,
+      "loss": 0.4487,
+      "step": 363
+    },
+    {
+      "epoch": 0.4390832328106152,
+      "grad_norm": 0.4705681098407308,
+      "learning_rate": 4.7430741733690796e-05,
+      "loss": 0.4487,
+      "step": 364
+    },
+    {
+      "epoch": 0.44028950542822676,
+      "grad_norm": 0.44175385002893386,
+      "learning_rate": 4.7408400357462024e-05,
+      "loss": 0.4412,
+      "step": 365
+    },
+    {
+      "epoch": 0.44149577804583834,
+      "grad_norm": 0.40907300632327315,
+      "learning_rate": 4.7386058981233244e-05,
+      "loss": 0.4434,
+      "step": 366
+    },
+    {
+      "epoch": 0.4427020506634499,
+      "grad_norm": 0.4293644058129011,
+      "learning_rate": 4.736371760500447e-05,
+      "loss": 0.4541,
+      "step": 367
+    },
+    {
+      "epoch": 0.4439083232810615,
+      "grad_norm": 0.4311778069623195,
+      "learning_rate": 4.73413762287757e-05,
+      "loss": 0.4849,
+      "step": 368
+    },
+    {
+      "epoch": 0.4451145958986731,
+      "grad_norm": 0.39533927643970806,
+      "learning_rate": 4.731903485254692e-05,
+      "loss": 0.4711,
+      "step": 369
+    },
+    {
+      "epoch": 0.44632086851628466,
+      "grad_norm": 0.43025595785491744,
+      "learning_rate": 4.729669347631814e-05,
+      "loss": 0.4343,
+      "step": 370
+    },
+    {
+      "epoch": 0.44752714113389624,
+      "grad_norm": 0.4289663869921225,
+      "learning_rate": 4.727435210008937e-05,
+      "loss": 0.4721,
+      "step": 371
+    },
+    {
+      "epoch": 0.4487334137515078,
+      "grad_norm": 0.39948007331512414,
+      "learning_rate": 4.725201072386059e-05,
+      "loss": 0.4332,
+      "step": 372
+    },
+    {
+      "epoch": 0.4499396863691194,
+      "grad_norm": 0.4386452400368104,
+      "learning_rate": 4.7229669347631816e-05,
+      "loss": 0.4389,
+      "step": 373
+    },
+    {
+      "epoch": 0.451145958986731,
+      "grad_norm": 0.4262661814196975,
+      "learning_rate": 4.7207327971403044e-05,
+      "loss": 0.4458,
+      "step": 374
+    },
+    {
+      "epoch": 0.45235223160434257,
+      "grad_norm": 0.37931784449502054,
+      "learning_rate": 4.7184986595174265e-05,
+      "loss": 0.4495,
+      "step": 375
+    },
+    {
+      "epoch": 0.45355850422195415,
+      "grad_norm": 0.4524128475128064,
+      "learning_rate": 4.716264521894549e-05,
+      "loss": 0.4455,
+      "step": 376
+    },
+    {
+      "epoch": 0.45476477683956573,
+      "grad_norm": 0.3838605212579755,
+      "learning_rate": 4.714030384271671e-05,
+      "loss": 0.4432,
+      "step": 377
+    },
+    {
+      "epoch": 0.4559710494571773,
+      "grad_norm": 0.45065400120393345,
+      "learning_rate": 4.7117962466487934e-05,
+      "loss": 0.4746,
+      "step": 378
+    },
+    {
+      "epoch": 0.4571773220747889,
+      "grad_norm": 0.41737828415366907,
+      "learning_rate": 4.709562109025916e-05,
+      "loss": 0.4554,
+      "step": 379
+    },
+    {
+      "epoch": 0.4583835946924005,
+      "grad_norm": 0.4654886210274436,
+      "learning_rate": 4.707327971403039e-05,
+      "loss": 0.4628,
+      "step": 380
+    },
+    {
+      "epoch": 0.45958986731001206,
+      "grad_norm": 0.46471916366661375,
+      "learning_rate": 4.705093833780161e-05,
+      "loss": 0.4705,
+      "step": 381
+    },
+    {
+      "epoch": 0.46079613992762364,
+      "grad_norm": 0.4909372293107839,
+      "learning_rate": 4.702859696157284e-05,
+      "loss": 0.4428,
+      "step": 382
+    },
+    {
+      "epoch": 0.4620024125452352,
+      "grad_norm": 0.4809948968908599,
+      "learning_rate": 4.7006255585344064e-05,
+      "loss": 0.4729,
+      "step": 383
+    },
+    {
+      "epoch": 0.4632086851628468,
+      "grad_norm": 0.395516138642914,
+      "learning_rate": 4.6983914209115285e-05,
+      "loss": 0.453,
+      "step": 384
+    },
+    {
+      "epoch": 0.4644149577804584,
+      "grad_norm": 0.5130350022661608,
+      "learning_rate": 4.6961572832886506e-05,
+      "loss": 0.4491,
+      "step": 385
+    },
+    {
+      "epoch": 0.46562123039806996,
+      "grad_norm": 0.4210327177360951,
+      "learning_rate": 4.693923145665773e-05,
+      "loss": 0.4394,
+      "step": 386
+    },
+    {
+      "epoch": 0.46682750301568154,
+      "grad_norm": 0.43805602941227245,
+      "learning_rate": 4.6916890080428954e-05,
+      "loss": 0.4498,
+      "step": 387
+    },
+    {
+      "epoch": 0.4680337756332931,
+      "grad_norm": 0.4374100581941471,
+      "learning_rate": 4.689454870420018e-05,
+      "loss": 0.4429,
+      "step": 388
+    },
+    {
+      "epoch": 0.4692400482509047,
+      "grad_norm": 0.42690169548391277,
+      "learning_rate": 4.687220732797141e-05,
+      "loss": 0.4486,
+      "step": 389
+    },
+    {
+      "epoch": 0.4704463208685163,
+      "grad_norm": 0.566544160347167,
+      "learning_rate": 4.684986595174263e-05,
+      "loss": 0.4351,
+      "step": 390
+    },
+    {
+      "epoch": 0.47165259348612787,
+      "grad_norm": 0.5463676682693923,
+      "learning_rate": 4.682752457551386e-05,
+      "loss": 0.4475,
+      "step": 391
+    },
+    {
+      "epoch": 0.47285886610373945,
+      "grad_norm": 0.5744958747454104,
+      "learning_rate": 4.680518319928508e-05,
+      "loss": 0.4402,
+      "step": 392
+    },
+    {
+      "epoch": 0.47406513872135103,
+      "grad_norm": 0.35516581962807914,
+      "learning_rate": 4.67828418230563e-05,
+      "loss": 0.4536,
+      "step": 393
+    },
+    {
+      "epoch": 0.4752714113389626,
+      "grad_norm": 0.5946852579555668,
+      "learning_rate": 4.6760500446827526e-05,
+      "loss": 0.4592,
+      "step": 394
+    },
+    {
+      "epoch": 0.4764776839565742,
+      "grad_norm": 0.47812875942111466,
+      "learning_rate": 4.673815907059875e-05,
+      "loss": 0.4491,
+      "step": 395
+    },
+    {
+      "epoch": 0.4776839565741858,
+      "grad_norm": 0.44836535011830225,
+      "learning_rate": 4.6715817694369974e-05,
+      "loss": 0.456,
+      "step": 396
+    },
+    {
+      "epoch": 0.47889022919179736,
+      "grad_norm": 0.5656700294363519,
+      "learning_rate": 4.66934763181412e-05,
+      "loss": 0.4697,
+      "step": 397
+    },
+    {
+      "epoch": 0.48009650180940894,
+      "grad_norm": 0.4300488393072486,
+      "learning_rate": 4.667113494191243e-05,
+      "loss": 0.4495,
+      "step": 398
+    },
+    {
+      "epoch": 0.4813027744270205,
+      "grad_norm": 0.5038511492658745,
+      "learning_rate": 4.664879356568364e-05,
+      "loss": 0.4619,
+      "step": 399
+    },
+    {
+      "epoch": 0.4825090470446321,
+      "grad_norm": 0.3821023359414941,
+      "learning_rate": 4.662645218945487e-05,
+      "loss": 0.4335,
+      "step": 400
+    },
+    {
+      "epoch": 0.4837153196622437,
+      "grad_norm": 0.45116254617278473,
+      "learning_rate": 4.66041108132261e-05,
+      "loss": 0.4664,
+      "step": 401
+    },
+    {
+      "epoch": 0.48492159227985526,
+      "grad_norm": 0.4695210682061351,
+      "learning_rate": 4.658176943699732e-05,
+      "loss": 0.4458,
+      "step": 402
+    },
+    {
+      "epoch": 0.48612786489746684,
+      "grad_norm": 0.37016930218926347,
+      "learning_rate": 4.6559428060768546e-05,
+      "loss": 0.4485,
+      "step": 403
+    },
+    {
+      "epoch": 0.4873341375150784,
+      "grad_norm": 0.48562834405762084,
+      "learning_rate": 4.653708668453977e-05,
+      "loss": 0.4523,
+      "step": 404
+    },
+    {
+      "epoch": 0.48854041013269,
+      "grad_norm": 0.42419760007250784,
+      "learning_rate": 4.6514745308310994e-05,
+      "loss": 0.4512,
+      "step": 405
+    },
+    {
+      "epoch": 0.4897466827503016,
+      "grad_norm": 0.5247057241517089,
+      "learning_rate": 4.6492403932082215e-05,
+      "loss": 0.4755,
+      "step": 406
+    },
+    {
+      "epoch": 0.49095295536791317,
+      "grad_norm": 0.4906512258804201,
+      "learning_rate": 4.647006255585344e-05,
+      "loss": 0.4752,
+      "step": 407
+    },
+    {
+      "epoch": 0.49215922798552475,
+      "grad_norm": 0.376373380011517,
+      "learning_rate": 4.644772117962466e-05,
+      "loss": 0.431,
+      "step": 408
+    },
+    {
+      "epoch": 0.49336550060313633,
+      "grad_norm": 0.41098879886311784,
+      "learning_rate": 4.642537980339589e-05,
+      "loss": 0.445,
+      "step": 409
+    },
+    {
+      "epoch": 0.4945717732207479,
+      "grad_norm": 0.45803591010235595,
+      "learning_rate": 4.640303842716712e-05,
+      "loss": 0.4729,
+      "step": 410
+    },
+    {
+      "epoch": 0.4957780458383595,
+      "grad_norm": 0.41767185914642463,
+      "learning_rate": 4.638069705093834e-05,
+      "loss": 0.4545,
+      "step": 411
+    },
+    {
+      "epoch": 0.4969843184559711,
+      "grad_norm": 0.38004435364145217,
+      "learning_rate": 4.6358355674709566e-05,
+      "loss": 0.4604,
+      "step": 412
+    },
+    {
+      "epoch": 0.49819059107358266,
+      "grad_norm": 0.46482447063204524,
+      "learning_rate": 4.633601429848079e-05,
+      "loss": 0.4256,
+      "step": 413
+    },
+    {
+      "epoch": 0.49939686369119424,
+      "grad_norm": 0.35230303342753366,
+      "learning_rate": 4.631367292225201e-05,
+      "loss": 0.4576,
+      "step": 414
+    },
+    {
+      "epoch": 0.5006031363088058,
+      "grad_norm": 0.466911404514305,
+      "learning_rate": 4.6291331546023235e-05,
+      "loss": 0.4301,
+      "step": 415
+    },
+    {
+      "epoch": 0.5018094089264173,
+      "grad_norm": 0.44691606084067054,
+      "learning_rate": 4.626899016979446e-05,
+      "loss": 0.4506,
+      "step": 416
+    },
+    {
+      "epoch": 0.503015681544029,
+      "grad_norm": 0.39651239023018436,
+      "learning_rate": 4.624664879356568e-05,
+      "loss": 0.4423,
+      "step": 417
+    },
+    {
+      "epoch": 0.5042219541616405,
+      "grad_norm": 0.478399072958898,
+      "learning_rate": 4.622430741733691e-05,
+      "loss": 0.4547,
+      "step": 418
+    },
+    {
+      "epoch": 0.5054282267792521,
+      "grad_norm": 0.4641367903464798,
+      "learning_rate": 4.620196604110814e-05,
+      "loss": 0.4485,
+      "step": 419
+    },
+    {
+      "epoch": 0.5066344993968637,
+      "grad_norm": 0.3978320017768976,
+      "learning_rate": 4.617962466487936e-05,
+      "loss": 0.4495,
+      "step": 420
+    },
+    {
+      "epoch": 0.5078407720144753,
+      "grad_norm": 0.4867811541317465,
+      "learning_rate": 4.615728328865058e-05,
+      "loss": 0.4823,
+      "step": 421
+    },
+    {
+      "epoch": 0.5090470446320868,
+      "grad_norm": 0.37579542830012624,
+      "learning_rate": 4.613494191242181e-05,
+      "loss": 0.4562,
+      "step": 422
+    },
+    {
+      "epoch": 0.5102533172496985,
+      "grad_norm": 0.3933322015028353,
+      "learning_rate": 4.611260053619303e-05,
+      "loss": 0.4697,
+      "step": 423
+    },
+    {
+      "epoch": 0.51145958986731,
+      "grad_norm": 0.5140813767762246,
+      "learning_rate": 4.6090259159964255e-05,
+      "loss": 0.4539,
+      "step": 424
+    },
+    {
+      "epoch": 0.5126658624849216,
+      "grad_norm": 0.4293323108323189,
+      "learning_rate": 4.606791778373548e-05,
+      "loss": 0.4688,
+      "step": 425
+    },
+    {
+      "epoch": 0.5138721351025332,
+      "grad_norm": 0.4037976218720339,
+      "learning_rate": 4.60455764075067e-05,
+      "loss": 0.433,
+      "step": 426
+    },
+    {
+      "epoch": 0.5150784077201448,
+      "grad_norm": 0.49210382273054565,
+      "learning_rate": 4.602323503127793e-05,
+      "loss": 0.4721,
+      "step": 427
+    },
+    {
+      "epoch": 0.5162846803377563,
+      "grad_norm": 0.45360571122605636,
+      "learning_rate": 4.600089365504915e-05,
+      "loss": 0.4417,
+      "step": 428
+    },
+    {
+      "epoch": 0.517490952955368,
+      "grad_norm": 0.43167468125357994,
+      "learning_rate": 4.597855227882037e-05,
+      "loss": 0.4431,
+      "step": 429
+    },
+    {
+      "epoch": 0.5186972255729795,
+      "grad_norm": 0.4333047936227905,
+      "learning_rate": 4.59562109025916e-05,
+      "loss": 0.4513,
+      "step": 430
+    },
+    {
+      "epoch": 0.5199034981905911,
+      "grad_norm": 0.37541607482723177,
+      "learning_rate": 4.593386952636283e-05,
+      "loss": 0.4379,
+      "step": 431
+    },
+    {
+      "epoch": 0.5211097708082026,
+      "grad_norm": 0.43350699006432963,
+      "learning_rate": 4.591152815013405e-05,
+      "loss": 0.4465,
+      "step": 432
+    },
+    {
+      "epoch": 0.5223160434258143,
+      "grad_norm": 0.3396586628116001,
+      "learning_rate": 4.5889186773905275e-05,
+      "loss": 0.4658,
+      "step": 433
+    },
+    {
+      "epoch": 0.5235223160434258,
+      "grad_norm": 0.39404427630025723,
+      "learning_rate": 4.58668453976765e-05,
+      "loss": 0.4665,
+      "step": 434
+    },
+    {
+      "epoch": 0.5247285886610374,
+      "grad_norm": 0.66647772071702,
+      "learning_rate": 4.5844504021447723e-05,
+      "loss": 0.476,
+      "step": 435
+    },
+    {
+      "epoch": 0.525934861278649,
+      "grad_norm": 0.41590803227411977,
+      "learning_rate": 4.5822162645218944e-05,
+      "loss": 0.4628,
+      "step": 436
+    },
+    {
+      "epoch": 0.5271411338962606,
+      "grad_norm": 0.3535047122980892,
+      "learning_rate": 4.579982126899017e-05,
+      "loss": 0.463,
+      "step": 437
+    },
+    {
+      "epoch": 0.5283474065138721,
+      "grad_norm": 0.3984301763075602,
+      "learning_rate": 4.57774798927614e-05,
+      "loss": 0.4477,
+      "step": 438
+    },
+    {
+      "epoch": 0.5295536791314838,
+      "grad_norm": 0.41893046808056794,
+      "learning_rate": 4.575513851653262e-05,
+      "loss": 0.4408,
+      "step": 439
+    },
+    {
+      "epoch": 0.5307599517490953,
+      "grad_norm": 0.3462986350174594,
+      "learning_rate": 4.573279714030385e-05,
+      "loss": 0.4242,
+      "step": 440
+    },
+    {
+      "epoch": 0.5319662243667069,
+      "grad_norm": 0.46322150308522203,
+      "learning_rate": 4.5710455764075075e-05,
+      "loss": 0.449,
+      "step": 441
+    },
+    {
+      "epoch": 0.5331724969843185,
+      "grad_norm": 0.5067918359886772,
+      "learning_rate": 4.5688114387846295e-05,
+      "loss": 0.4825,
+      "step": 442
+    },
+    {
+      "epoch": 0.53437876960193,
+      "grad_norm": 0.45997864303535113,
+      "learning_rate": 4.5665773011617516e-05,
+      "loss": 0.4324,
+      "step": 443
+    },
+    {
+      "epoch": 0.5355850422195416,
+      "grad_norm": 0.5398561931583092,
+      "learning_rate": 4.5643431635388744e-05,
+      "loss": 0.4689,
+      "step": 444
+    },
+    {
+      "epoch": 0.5367913148371531,
+      "grad_norm": 0.3868465771958736,
+      "learning_rate": 4.5621090259159964e-05,
+      "loss": 0.4572,
+      "step": 445
+    },
+    {
+      "epoch": 0.5379975874547648,
+      "grad_norm": 0.4419588081800375,
+      "learning_rate": 4.559874888293119e-05,
+      "loss": 0.4345,
+      "step": 446
+    },
+    {
+      "epoch": 0.5392038600723763,
+      "grad_norm": 0.3578794337535823,
+      "learning_rate": 4.557640750670242e-05,
+      "loss": 0.4445,
+      "step": 447
+    },
+    {
+      "epoch": 0.5404101326899879,
+      "grad_norm": 0.47193397921160224,
+      "learning_rate": 4.555406613047364e-05,
+      "loss": 0.4675,
+      "step": 448
+    },
+    {
+      "epoch": 0.5416164053075995,
+      "grad_norm": 0.36285549390237526,
+      "learning_rate": 4.553172475424487e-05,
+      "loss": 0.4495,
+      "step": 449
+    },
+    {
+      "epoch": 0.5428226779252111,
+      "grad_norm": 0.3989213205809804,
+      "learning_rate": 4.550938337801609e-05,
+      "loss": 0.4361,
+      "step": 450
+    },
+    {
+      "epoch": 0.5440289505428226,
+      "grad_norm": 0.3664013991140493,
+      "learning_rate": 4.548704200178731e-05,
+      "loss": 0.4648,
+      "step": 451
+    },
+    {
+      "epoch": 0.5452352231604343,
+      "grad_norm": 0.3882957125804488,
+      "learning_rate": 4.5464700625558536e-05,
+      "loss": 0.4432,
+      "step": 452
+    },
+    {
+      "epoch": 0.5464414957780458,
+      "grad_norm": 0.41892351772052466,
+      "learning_rate": 4.5442359249329764e-05,
+      "loss": 0.4371,
+      "step": 453
+    },
+    {
+      "epoch": 0.5476477683956574,
+      "grad_norm": 0.39080990222858825,
+      "learning_rate": 4.5420017873100985e-05,
+      "loss": 0.4366,
+      "step": 454
+    },
+    {
+      "epoch": 0.548854041013269,
+      "grad_norm": 0.41778012912126616,
+      "learning_rate": 4.539767649687221e-05,
+      "loss": 0.4551,
+      "step": 455
+    },
+    {
+      "epoch": 0.5500603136308806,
+      "grad_norm": 0.36612593843186386,
+      "learning_rate": 4.537533512064344e-05,
+      "loss": 0.4333,
+      "step": 456
+    },
+    {
+      "epoch": 0.5512665862484921,
+      "grad_norm": 0.40433234185282513,
+      "learning_rate": 4.535299374441466e-05,
+      "loss": 0.4409,
+      "step": 457
+    },
+    {
+      "epoch": 0.5524728588661038,
+      "grad_norm": 0.7444650694896087,
+      "learning_rate": 4.533065236818588e-05,
+      "loss": 0.4561,
+      "step": 458
+    },
+    {
+      "epoch": 0.5536791314837153,
+      "grad_norm": 0.4116764564963248,
+      "learning_rate": 4.530831099195711e-05,
+      "loss": 0.4389,
+      "step": 459
+    },
+    {
+      "epoch": 0.5548854041013269,
+      "grad_norm": 0.38464112606556955,
+      "learning_rate": 4.528596961572833e-05,
+      "loss": 0.4512,
+      "step": 460
+    },
+    {
+      "epoch": 0.5560916767189384,
+      "grad_norm": 0.39310293733158486,
+      "learning_rate": 4.5263628239499557e-05,
+      "loss": 0.4466,
+      "step": 461
+    },
+    {
+      "epoch": 0.5572979493365501,
+      "grad_norm": 0.41483308376955597,
+      "learning_rate": 4.5241286863270784e-05,
+      "loss": 0.4452,
+      "step": 462
+    },
+    {
+      "epoch": 0.5585042219541616,
+      "grad_norm": 0.3809153812654458,
+      "learning_rate": 4.5218945487042005e-05,
+      "loss": 0.4531,
+      "step": 463
+    },
+    {
+      "epoch": 0.5597104945717732,
+      "grad_norm": 0.39580214313931394,
+      "learning_rate": 4.5196604110813225e-05,
+      "loss": 0.4635,
+      "step": 464
+    },
+    {
+      "epoch": 0.5609167671893848,
+      "grad_norm": 0.36413357758862114,
+      "learning_rate": 4.517426273458445e-05,
+      "loss": 0.4472,
+      "step": 465
+    },
+    {
+      "epoch": 0.5621230398069964,
+      "grad_norm": 0.3308446731299311,
+      "learning_rate": 4.5151921358355674e-05,
+      "loss": 0.425,
+      "step": 466
+    },
+    {
+      "epoch": 0.5633293124246079,
+      "grad_norm": 0.3281583052827164,
+      "learning_rate": 4.51295799821269e-05,
+      "loss": 0.43,
+      "step": 467
+    },
+    {
+      "epoch": 0.5645355850422196,
+      "grad_norm": 0.40183980071492237,
+      "learning_rate": 4.510723860589813e-05,
+      "loss": 0.4579,
+      "step": 468
+    },
+    {
+      "epoch": 0.5657418576598311,
+      "grad_norm": 0.42058205952983124,
+      "learning_rate": 4.508489722966935e-05,
+      "loss": 0.4631,
+      "step": 469
+    },
+    {
+      "epoch": 0.5669481302774427,
+      "grad_norm": 0.34646556614800056,
+      "learning_rate": 4.506255585344058e-05,
+      "loss": 0.4694,
+      "step": 470
+    },
+    {
+      "epoch": 0.5681544028950543,
+      "grad_norm": 0.45794848334831695,
+      "learning_rate": 4.50402144772118e-05,
+      "loss": 0.4425,
+      "step": 471
+    },
+    {
+      "epoch": 0.5693606755126659,
+      "grad_norm": 0.34739688810650715,
+      "learning_rate": 4.501787310098302e-05,
+      "loss": 0.4496,
+      "step": 472
+    },
+    {
+      "epoch": 0.5705669481302774,
+      "grad_norm": 0.5010209658550312,
+      "learning_rate": 4.4995531724754246e-05,
+      "loss": 0.4446,
+      "step": 473
+    },
+    {
+      "epoch": 0.571773220747889,
+      "grad_norm": 0.5164932286032147,
+      "learning_rate": 4.497319034852547e-05,
+      "loss": 0.4275,
+      "step": 474
+    },
+    {
+      "epoch": 0.5729794933655006,
+      "grad_norm": 0.3736542241126828,
+      "learning_rate": 4.4950848972296694e-05,
+      "loss": 0.4285,
+      "step": 475
+    },
+    {
+      "epoch": 0.5741857659831122,
+      "grad_norm": 0.4110262572490399,
+      "learning_rate": 4.492850759606792e-05,
+      "loss": 0.4561,
+      "step": 476
+    },
+    {
+      "epoch": 0.5753920386007237,
+      "grad_norm": 1.13426364723185,
+      "learning_rate": 4.490616621983915e-05,
+      "loss": 0.436,
+      "step": 477
+    },
+    {
+      "epoch": 0.5765983112183354,
+      "grad_norm": 0.3844019173292291,
+      "learning_rate": 4.488382484361037e-05,
+      "loss": 0.4502,
+      "step": 478
+    },
+    {
+      "epoch": 0.5778045838359469,
+      "grad_norm": 0.43306080304122824,
+      "learning_rate": 4.486148346738159e-05,
+      "loss": 0.4297,
+      "step": 479
+    },
+    {
+      "epoch": 0.5790108564535585,
+      "grad_norm": 0.3597160387122504,
+      "learning_rate": 4.483914209115282e-05,
+      "loss": 0.445,
+      "step": 480
+    },
+    {
+      "epoch": 0.5802171290711701,
+      "grad_norm": 0.3756419648637461,
+      "learning_rate": 4.481680071492404e-05,
+      "loss": 0.4273,
+      "step": 481
+    },
+    {
+      "epoch": 0.5814234016887817,
+      "grad_norm": 0.37035532088343315,
+      "learning_rate": 4.4794459338695266e-05,
+      "loss": 0.4654,
+      "step": 482
+    },
+    {
+      "epoch": 0.5826296743063932,
+      "grad_norm": 0.3807674777156649,
+      "learning_rate": 4.477211796246649e-05,
+      "loss": 0.4123,
+      "step": 483
+    },
+    {
+      "epoch": 0.5838359469240049,
+      "grad_norm": 0.4624416203917337,
+      "learning_rate": 4.4749776586237714e-05,
+      "loss": 0.4483,
+      "step": 484
+    },
+    {
+      "epoch": 0.5850422195416164,
+      "grad_norm": 0.40382677830929836,
+      "learning_rate": 4.472743521000894e-05,
+      "loss": 0.451,
+      "step": 485
+    },
+    {
+      "epoch": 0.586248492159228,
+      "grad_norm": 0.44548308481777593,
+      "learning_rate": 4.470509383378016e-05,
+      "loss": 0.4421,
+      "step": 486
+    },
+    {
+      "epoch": 0.5874547647768396,
+      "grad_norm": 0.38837454887772743,
+      "learning_rate": 4.468275245755138e-05,
+      "loss": 0.4473,
+      "step": 487
+    },
+    {
+      "epoch": 0.5886610373944512,
+      "grad_norm": 0.5034381554902535,
+      "learning_rate": 4.466041108132261e-05,
+      "loss": 0.4601,
+      "step": 488
+    },
+    {
+      "epoch": 0.5898673100120627,
+      "grad_norm": 0.5549913556398447,
+      "learning_rate": 4.463806970509384e-05,
+      "loss": 0.4501,
+      "step": 489
+    },
+    {
+      "epoch": 0.5910735826296744,
+      "grad_norm": 0.46407716031475194,
+      "learning_rate": 4.461572832886506e-05,
+      "loss": 0.4204,
+      "step": 490
+    },
+    {
+      "epoch": 0.5922798552472859,
+      "grad_norm": 0.4213356267800612,
+      "learning_rate": 4.4593386952636286e-05,
+      "loss": 0.4049,
+      "step": 491
+    },
+    {
+      "epoch": 0.5934861278648975,
+      "grad_norm": 0.5095914474336406,
+      "learning_rate": 4.4571045576407513e-05,
+      "loss": 0.4342,
+      "step": 492
+    },
+    {
+      "epoch": 0.594692400482509,
+      "grad_norm": 0.41864465345580903,
+      "learning_rate": 4.4548704200178734e-05,
+      "loss": 0.4376,
+      "step": 493
+    },
+    {
+      "epoch": 0.5958986731001207,
+      "grad_norm": 0.5958745960318322,
+      "learning_rate": 4.4526362823949955e-05,
+      "loss": 0.4447,
+      "step": 494
+    },
+    {
+      "epoch": 0.5971049457177322,
+      "grad_norm": 0.49437930231053906,
+      "learning_rate": 4.450402144772118e-05,
+      "loss": 0.4306,
+      "step": 495
+    },
+    {
+      "epoch": 0.5983112183353438,
+      "grad_norm": 0.4453966068547487,
+      "learning_rate": 4.44816800714924e-05,
+      "loss": 0.4574,
+      "step": 496
+    },
+    {
+      "epoch": 0.5995174909529554,
+      "grad_norm": 0.5804778387940887,
+      "learning_rate": 4.445933869526363e-05,
+      "loss": 0.4456,
+      "step": 497
+    },
+    {
+      "epoch": 0.6007237635705669,
+      "grad_norm": 0.513573961932933,
+      "learning_rate": 4.443699731903486e-05,
+      "loss": 0.43,
+      "step": 498
+    },
+    {
+      "epoch": 0.6019300361881785,
+      "grad_norm": 0.5536446946147545,
+      "learning_rate": 4.441465594280608e-05,
+      "loss": 0.4433,
+      "step": 499
+    },
+    {
+      "epoch": 0.6031363088057901,
+      "grad_norm": 0.422156053238679,
+      "learning_rate": 4.4392314566577306e-05,
+      "loss": 0.427,
+      "step": 500
+    },
+    {
+      "epoch": 0.6043425814234017,
+      "grad_norm": 0.3977480968975122,
+      "learning_rate": 4.436997319034853e-05,
+      "loss": 0.4286,
+      "step": 501
+    },
+    {
+      "epoch": 0.6055488540410132,
+      "grad_norm": 0.38578431608091396,
+      "learning_rate": 4.434763181411975e-05,
+      "loss": 0.4397,
+      "step": 502
+    },
+    {
+      "epoch": 0.6067551266586249,
+      "grad_norm": 0.4371318250540829,
+      "learning_rate": 4.4325290437890975e-05,
+      "loss": 0.4497,
+      "step": 503
+    },
+    {
+      "epoch": 0.6079613992762364,
+      "grad_norm": 0.38750843806678476,
+      "learning_rate": 4.43029490616622e-05,
+      "loss": 0.4515,
+      "step": 504
+    },
+    {
+      "epoch": 0.609167671893848,
+      "grad_norm": 0.3184771182337021,
+      "learning_rate": 4.428060768543342e-05,
+      "loss": 0.4226,
+      "step": 505
+    },
+    {
+      "epoch": 0.6103739445114595,
+      "grad_norm": 0.3349430507766991,
+      "learning_rate": 4.425826630920465e-05,
+      "loss": 0.4412,
+      "step": 506
+    },
+    {
+      "epoch": 0.6115802171290712,
+      "grad_norm": 0.460700030502988,
+      "learning_rate": 4.423592493297588e-05,
+      "loss": 0.4191,
+      "step": 507
+    },
+    {
+      "epoch": 0.6127864897466827,
+      "grad_norm": 0.3474525155837308,
+      "learning_rate": 4.42135835567471e-05,
+      "loss": 0.4433,
+      "step": 508
+    },
+    {
+      "epoch": 0.6139927623642943,
+      "grad_norm": 0.3890384095888057,
+      "learning_rate": 4.419124218051832e-05,
+      "loss": 0.4471,
+      "step": 509
+    },
+    {
+      "epoch": 0.6151990349819059,
+      "grad_norm": 0.3473241563303528,
+      "learning_rate": 4.416890080428955e-05,
+      "loss": 0.4333,
+      "step": 510
+    },
+    {
+      "epoch": 0.6164053075995175,
+      "grad_norm": 0.38649792348349676,
+      "learning_rate": 4.414655942806077e-05,
+      "loss": 0.4401,
+      "step": 511
+    },
+    {
+      "epoch": 0.617611580217129,
+      "grad_norm": 0.390326168737818,
+      "learning_rate": 4.4124218051831995e-05,
+      "loss": 0.4415,
+      "step": 512
+    },
+    {
+      "epoch": 0.6188178528347407,
+      "grad_norm": 0.36517910787407803,
+      "learning_rate": 4.410187667560322e-05,
+      "loss": 0.4161,
+      "step": 513
+    },
+    {
+      "epoch": 0.6200241254523522,
+      "grad_norm": 0.3871202882793126,
+      "learning_rate": 4.4079535299374443e-05,
+      "loss": 0.4388,
+      "step": 514
+    },
+    {
+      "epoch": 0.6212303980699638,
+      "grad_norm": 0.3606332726485659,
+      "learning_rate": 4.405719392314567e-05,
+      "loss": 0.4411,
+      "step": 515
+    },
+    {
+      "epoch": 0.6224366706875754,
+      "grad_norm": 0.437394776038846,
+      "learning_rate": 4.403485254691689e-05,
+      "loss": 0.4565,
+      "step": 516
+    },
+    {
+      "epoch": 0.623642943305187,
+      "grad_norm": 0.4130632114314605,
+      "learning_rate": 4.401251117068811e-05,
+      "loss": 0.4666,
+      "step": 517
+    },
+    {
+      "epoch": 0.6248492159227985,
+      "grad_norm": 0.6077447752036588,
+      "learning_rate": 4.399016979445934e-05,
+      "loss": 0.4219,
+      "step": 518
+    },
+    {
+      "epoch": 0.6260554885404102,
+      "grad_norm": 0.48865858289065855,
+      "learning_rate": 4.396782841823057e-05,
+      "loss": 0.4525,
+      "step": 519
+    },
+    {
+      "epoch": 0.6272617611580217,
+      "grad_norm": 0.37378858727980474,
+      "learning_rate": 4.394548704200179e-05,
+      "loss": 0.4154,
+      "step": 520
+    },
+    {
+      "epoch": 0.6284680337756333,
+      "grad_norm": 0.4731113479746855,
+      "learning_rate": 4.3923145665773015e-05,
+      "loss": 0.4294,
+      "step": 521
+    },
+    {
+      "epoch": 0.6296743063932448,
+      "grad_norm": 0.4090362259372264,
+      "learning_rate": 4.390080428954424e-05,
+      "loss": 0.4314,
+      "step": 522
+    },
+    {
+      "epoch": 0.6308805790108565,
+      "grad_norm": 0.3220703974636813,
+      "learning_rate": 4.387846291331546e-05,
+      "loss": 0.4306,
+      "step": 523
+    },
+    {
+      "epoch": 0.632086851628468,
+      "grad_norm": 0.42173924202764596,
+      "learning_rate": 4.3856121537086684e-05,
+      "loss": 0.4351,
+      "step": 524
+    },
+    {
+      "epoch": 0.6332931242460796,
+      "grad_norm": 0.3330928257528643,
+      "learning_rate": 4.383378016085791e-05,
+      "loss": 0.4303,
+      "step": 525
+    },
+    {
+      "epoch": 0.6344993968636912,
+      "grad_norm": 0.41161949162644296,
+      "learning_rate": 4.381143878462913e-05,
+      "loss": 0.4348,
+      "step": 526
+    },
+    {
+      "epoch": 0.6357056694813028,
+      "grad_norm": 0.3883545439321418,
+      "learning_rate": 4.378909740840036e-05,
+      "loss": 0.4369,
+      "step": 527
+    },
+    {
+      "epoch": 0.6369119420989143,
+      "grad_norm": 0.4325113358235531,
+      "learning_rate": 4.376675603217159e-05,
+      "loss": 0.4532,
+      "step": 528
+    },
+    {
+      "epoch": 0.638118214716526,
+      "grad_norm": 0.4380697726970751,
+      "learning_rate": 4.374441465594281e-05,
+      "loss": 0.4643,
+      "step": 529
+    },
+    {
+      "epoch": 0.6393244873341375,
+      "grad_norm": 0.35916390367525086,
+      "learning_rate": 4.372207327971403e-05,
+      "loss": 0.4171,
+      "step": 530
+    },
+    {
+      "epoch": 0.6405307599517491,
+      "grad_norm": 0.3486136621622154,
+      "learning_rate": 4.3699731903485256e-05,
+      "loss": 0.4488,
+      "step": 531
+    },
+    {
+      "epoch": 0.6417370325693607,
+      "grad_norm": 0.3312426151787181,
+      "learning_rate": 4.367739052725648e-05,
+      "loss": 0.4407,
+      "step": 532
+    },
+    {
+      "epoch": 0.6429433051869723,
+      "grad_norm": 0.4132145377973402,
+      "learning_rate": 4.3655049151027704e-05,
+      "loss": 0.4334,
+      "step": 533
+    },
+    {
+      "epoch": 0.6441495778045838,
+      "grad_norm": 0.34330718379196123,
+      "learning_rate": 4.363270777479893e-05,
+      "loss": 0.445,
+      "step": 534
+    },
+    {
+      "epoch": 0.6453558504221955,
+      "grad_norm": 0.45135766155768786,
+      "learning_rate": 4.361036639857015e-05,
+      "loss": 0.4507,
+      "step": 535
+    },
+    {
+      "epoch": 0.646562123039807,
+      "grad_norm": 0.35589892838561626,
+      "learning_rate": 4.358802502234138e-05,
+      "loss": 0.4202,
+      "step": 536
+    },
+    {
+      "epoch": 0.6477683956574186,
+      "grad_norm": 1.1341629551979147,
+      "learning_rate": 4.35656836461126e-05,
+      "loss": 0.4487,
+      "step": 537
+    },
+    {
+      "epoch": 0.6489746682750301,
+      "grad_norm": 0.3743916061859571,
+      "learning_rate": 4.354334226988382e-05,
+      "loss": 0.4502,
+      "step": 538
+    },
+    {
+      "epoch": 0.6501809408926418,
+      "grad_norm": 0.3701078540192121,
+      "learning_rate": 4.352100089365505e-05,
+      "loss": 0.4353,
+      "step": 539
+    },
+    {
+      "epoch": 0.6513872135102533,
+      "grad_norm": 0.37975867227218413,
+      "learning_rate": 4.3498659517426276e-05,
+      "loss": 0.4604,
+      "step": 540
+    },
+    {
+      "epoch": 0.6525934861278649,
+      "grad_norm": 0.45996926881333117,
+      "learning_rate": 4.34763181411975e-05,
+      "loss": 0.4381,
+      "step": 541
+    },
+    {
+      "epoch": 0.6537997587454765,
+      "grad_norm": 0.3529432931676143,
+      "learning_rate": 4.3453976764968725e-05,
+      "loss": 0.4469,
+      "step": 542
+    },
+    {
+      "epoch": 0.6550060313630881,
+      "grad_norm": 0.4637421891155913,
+      "learning_rate": 4.343163538873995e-05,
+      "loss": 0.4436,
+      "step": 543
+    },
+    {
+      "epoch": 0.6562123039806996,
+      "grad_norm": 0.44641761278318115,
+      "learning_rate": 4.340929401251117e-05,
+      "loss": 0.4237,
+      "step": 544
+    },
+    {
+      "epoch": 0.6574185765983113,
+      "grad_norm": 0.45548634034253993,
+      "learning_rate": 4.3386952636282394e-05,
+      "loss": 0.451,
+      "step": 545
+    },
+    {
+      "epoch": 0.6586248492159228,
+      "grad_norm": 0.6095224416117992,
+      "learning_rate": 4.336461126005362e-05,
+      "loss": 0.4234,
+      "step": 546
+    },
+    {
+      "epoch": 0.6598311218335344,
+      "grad_norm": 0.4794907978183171,
+      "learning_rate": 4.334226988382484e-05,
+      "loss": 0.4321,
+      "step": 547
+    },
+    {
+      "epoch": 0.661037394451146,
+      "grad_norm": 0.4372188988863073,
+      "learning_rate": 4.331992850759607e-05,
+      "loss": 0.4291,
+      "step": 548
+    },
+    {
+      "epoch": 0.6622436670687576,
+      "grad_norm": 0.41945601478116296,
+      "learning_rate": 4.32975871313673e-05,
+      "loss": 0.4205,
+      "step": 549
+    },
+    {
+      "epoch": 0.6634499396863691,
+      "grad_norm": 0.5243682305557889,
+      "learning_rate": 4.327524575513852e-05,
+      "loss": 0.4416,
+      "step": 550
+    },
+    {
+      "epoch": 0.6646562123039808,
+      "grad_norm": 0.4801155246465036,
+      "learning_rate": 4.3252904378909745e-05,
+      "loss": 0.4462,
+      "step": 551
+    },
+    {
+      "epoch": 0.6658624849215923,
+      "grad_norm": 0.42217526503844,
+      "learning_rate": 4.3230563002680966e-05,
+      "loss": 0.4252,
+      "step": 552
+    },
+    {
+      "epoch": 0.6670687575392038,
+      "grad_norm": 0.4769157452972345,
+      "learning_rate": 4.3208221626452186e-05,
+      "loss": 0.4277,
+      "step": 553
+    },
+    {
+      "epoch": 0.6682750301568154,
+      "grad_norm": 0.4446751167304439,
+      "learning_rate": 4.3185880250223414e-05,
+      "loss": 0.4029,
+      "step": 554
+    },
+    {
+      "epoch": 0.669481302774427,
+      "grad_norm": 0.44617281388579455,
+      "learning_rate": 4.316353887399464e-05,
+      "loss": 0.4358,
+      "step": 555
+    },
+    {
+      "epoch": 0.6706875753920386,
+      "grad_norm": 0.5947589007882237,
+      "learning_rate": 4.314119749776586e-05,
+      "loss": 0.4505,
+      "step": 556
+    },
+    {
+      "epoch": 0.6718938480096501,
+      "grad_norm": 0.3674033271102033,
+      "learning_rate": 4.311885612153709e-05,
+      "loss": 0.4437,
+      "step": 557
+    },
+    {
+      "epoch": 0.6731001206272618,
+      "grad_norm": 0.4400826256131888,
+      "learning_rate": 4.309651474530832e-05,
+      "loss": 0.4301,
+      "step": 558
+    },
+    {
+      "epoch": 0.6743063932448733,
+      "grad_norm": 0.43015064156020477,
+      "learning_rate": 4.307417336907954e-05,
+      "loss": 0.4334,
+      "step": 559
+    },
+    {
+      "epoch": 0.6755126658624849,
+      "grad_norm": 0.3782245475032066,
+      "learning_rate": 4.305183199285076e-05,
+      "loss": 0.419,
+      "step": 560
+    },
+    {
+      "epoch": 0.6767189384800965,
+      "grad_norm": 0.4066237562131077,
+      "learning_rate": 4.3029490616621986e-05,
+      "loss": 0.4356,
+      "step": 561
+    },
+    {
+      "epoch": 0.6779252110977081,
+      "grad_norm": 0.4187960770131329,
+      "learning_rate": 4.300714924039321e-05,
+      "loss": 0.4289,
+      "step": 562
+    },
+    {
+      "epoch": 0.6791314837153196,
+      "grad_norm": 0.35579617743327024,
+      "learning_rate": 4.2984807864164434e-05,
+      "loss": 0.4575,
+      "step": 563
+    },
+    {
+      "epoch": 0.6803377563329313,
+      "grad_norm": 0.4409858128600069,
+      "learning_rate": 4.296246648793566e-05,
+      "loss": 0.4633,
+      "step": 564
+    },
+    {
+      "epoch": 0.6815440289505428,
+      "grad_norm": 0.3354907808913072,
+      "learning_rate": 4.294012511170689e-05,
+      "loss": 0.4333,
+      "step": 565
+    },
+    {
+      "epoch": 0.6827503015681544,
+      "grad_norm": 0.42655397239204723,
+      "learning_rate": 4.291778373547811e-05,
+      "loss": 0.4428,
+      "step": 566
+    },
+    {
+      "epoch": 0.6839565741857659,
+      "grad_norm": 0.443037805064068,
+      "learning_rate": 4.289544235924933e-05,
+      "loss": 0.4451,
+      "step": 567
+    },
+    {
+      "epoch": 0.6851628468033776,
+      "grad_norm": 0.8914794928360468,
+      "learning_rate": 4.287310098302056e-05,
+      "loss": 0.4396,
+      "step": 568
+    },
+    {
+      "epoch": 0.6863691194209891,
+      "grad_norm": 0.5448067106591934,
+      "learning_rate": 4.285075960679178e-05,
+      "loss": 0.461,
+      "step": 569
+    },
+    {
+      "epoch": 0.6875753920386007,
+      "grad_norm": 0.42014703815134213,
+      "learning_rate": 4.2828418230563006e-05,
+      "loss": 0.4165,
+      "step": 570
+    },
+    {
+      "epoch": 0.6887816646562123,
+      "grad_norm": 0.5363032079607244,
+      "learning_rate": 4.280607685433423e-05,
+      "loss": 0.4481,
+      "step": 571
+    },
+    {
+      "epoch": 0.6899879372738239,
+      "grad_norm": 0.6788458284685935,
+      "learning_rate": 4.2783735478105454e-05,
+      "loss": 0.4332,
+      "step": 572
+    },
+    {
+      "epoch": 0.6911942098914354,
+      "grad_norm": 0.40379510471482416,
+      "learning_rate": 4.276139410187668e-05,
+      "loss": 0.4439,
+      "step": 573
+    },
+    {
+      "epoch": 0.6924004825090471,
+      "grad_norm": 0.6130371034513505,
+      "learning_rate": 4.27390527256479e-05,
+      "loss": 0.4377,
+      "step": 574
+    },
+    {
+      "epoch": 0.6936067551266586,
+      "grad_norm": 0.3627939971877631,
+      "learning_rate": 4.271671134941912e-05,
+      "loss": 0.4285,
+      "step": 575
+    },
+    {
+      "epoch": 0.6948130277442702,
+      "grad_norm": 0.5305583762077702,
+      "learning_rate": 4.269436997319035e-05,
+      "loss": 0.4201,
+      "step": 576
+    },
+    {
+      "epoch": 0.6960193003618818,
+      "grad_norm": 0.510353267102675,
+      "learning_rate": 4.267202859696158e-05,
+      "loss": 0.4132,
+      "step": 577
+    },
+    {
+      "epoch": 0.6972255729794934,
+      "grad_norm": 0.5113841119221939,
+      "learning_rate": 4.26496872207328e-05,
+      "loss": 0.4317,
+      "step": 578
+    },
+    {
+      "epoch": 0.6984318455971049,
+      "grad_norm": 0.6169504699258387,
+      "learning_rate": 4.2627345844504026e-05,
+      "loss": 0.4364,
+      "step": 579
+    },
+    {
+      "epoch": 0.6996381182147166,
+      "grad_norm": 0.5050461237759128,
+      "learning_rate": 4.2605004468275254e-05,
+      "loss": 0.4434,
+      "step": 580
+    },
+    {
+      "epoch": 0.7008443908323281,
+      "grad_norm": 0.595300192798241,
+      "learning_rate": 4.258266309204647e-05,
+      "loss": 0.4702,
+      "step": 581
+    },
+    {
+      "epoch": 0.7020506634499397,
+      "grad_norm": 0.38965873875054796,
+      "learning_rate": 4.2560321715817695e-05,
+      "loss": 0.4472,
+      "step": 582
+    },
+    {
+      "epoch": 0.7032569360675512,
+      "grad_norm": 0.5644938928003483,
+      "learning_rate": 4.253798033958892e-05,
+      "loss": 0.4268,
+      "step": 583
+    },
+    {
+      "epoch": 0.7044632086851629,
+      "grad_norm": 0.33911633262614643,
+      "learning_rate": 4.251563896336014e-05,
+      "loss": 0.4366,
+      "step": 584
+    },
+    {
+      "epoch": 0.7056694813027744,
+      "grad_norm": 0.5326007243448592,
+      "learning_rate": 4.249329758713137e-05,
+      "loss": 0.455,
+      "step": 585
+    },
+    {
+      "epoch": 0.706875753920386,
+      "grad_norm": 0.4721907283458342,
+      "learning_rate": 4.24709562109026e-05,
+      "loss": 0.4214,
+      "step": 586
+    },
+    {
+      "epoch": 0.7080820265379976,
+      "grad_norm": 0.4312507346368042,
+      "learning_rate": 4.244861483467382e-05,
+      "loss": 0.444,
+      "step": 587
+    },
+    {
+      "epoch": 0.7092882991556092,
+      "grad_norm": 0.4884922993522299,
+      "learning_rate": 4.242627345844504e-05,
+      "loss": 0.4205,
+      "step": 588
+    },
+    {
+      "epoch": 0.7104945717732207,
+      "grad_norm": 0.3242841502058886,
+      "learning_rate": 4.240393208221627e-05,
+      "loss": 0.4188,
+      "step": 589
+    },
+    {
+      "epoch": 0.7117008443908324,
+      "grad_norm": 0.529266738309419,
+      "learning_rate": 4.238159070598749e-05,
+      "loss": 0.4284,
+      "step": 590
+    },
+    {
+      "epoch": 0.7129071170084439,
+      "grad_norm": 0.31078521850095026,
+      "learning_rate": 4.2359249329758715e-05,
+      "loss": 0.4401,
+      "step": 591
+    },
+    {
+      "epoch": 0.7141133896260555,
+      "grad_norm": 0.42313591630009106,
+      "learning_rate": 4.233690795352994e-05,
+      "loss": 0.439,
+      "step": 592
+    },
+    {
+      "epoch": 0.715319662243667,
+      "grad_norm": 0.3458759202241377,
+      "learning_rate": 4.231456657730116e-05,
+      "loss": 0.432,
+      "step": 593
+    },
+    {
+      "epoch": 0.7165259348612787,
+      "grad_norm": 0.3856562354076321,
+      "learning_rate": 4.229222520107239e-05,
+      "loss": 0.4359,
+      "step": 594
+    },
+    {
+      "epoch": 0.7177322074788902,
+      "grad_norm": 0.4325118137990474,
+      "learning_rate": 4.226988382484361e-05,
+      "loss": 0.428,
+      "step": 595
+    },
+    {
+      "epoch": 0.7189384800965019,
+      "grad_norm": 0.36937544795868116,
+      "learning_rate": 4.224754244861483e-05,
+      "loss": 0.4378,
+      "step": 596
+    },
+    {
+      "epoch": 0.7201447527141134,
+      "grad_norm": 0.36827610206035255,
+      "learning_rate": 4.222520107238606e-05,
+      "loss": 0.4367,
+      "step": 597
+    },
+    {
+      "epoch": 0.721351025331725,
+      "grad_norm": 0.3412919037234244,
+      "learning_rate": 4.220285969615729e-05,
+      "loss": 0.4428,
+      "step": 598
+    },
+    {
+      "epoch": 0.7225572979493365,
+      "grad_norm": 0.4013821463345791,
+      "learning_rate": 4.218051831992851e-05,
+      "loss": 0.435,
+      "step": 599
+    },
+    {
+      "epoch": 0.7237635705669482,
+      "grad_norm": 0.35671033183339573,
+      "learning_rate": 4.2158176943699735e-05,
+      "loss": 0.4589,
+      "step": 600
+    },
+    {
+      "epoch": 0.7249698431845597,
+      "grad_norm": 0.41435093321906774,
+      "learning_rate": 4.213583556747096e-05,
+      "loss": 0.4463,
+      "step": 601
+    },
+    {
+      "epoch": 0.7261761158021713,
+      "grad_norm": 0.4102815510753044,
+      "learning_rate": 4.2113494191242184e-05,
+      "loss": 0.402,
+      "step": 602
+    },
+    {
+      "epoch": 0.7273823884197829,
+      "grad_norm": 0.40787860182142316,
+      "learning_rate": 4.2091152815013404e-05,
+      "loss": 0.4426,
+      "step": 603
+    },
+    {
+      "epoch": 0.7285886610373945,
+      "grad_norm": 0.35075907908441134,
+      "learning_rate": 4.206881143878463e-05,
+      "loss": 0.4432,
+      "step": 604
+    },
+    {
+      "epoch": 0.729794933655006,
+      "grad_norm": 0.3765858326922549,
+      "learning_rate": 4.204647006255585e-05,
+      "loss": 0.4256,
+      "step": 605
+    },
+    {
+      "epoch": 0.7310012062726177,
+      "grad_norm": 0.3662765996267215,
+      "learning_rate": 4.202412868632708e-05,
+      "loss": 0.4441,
+      "step": 606
+    },
+    {
+      "epoch": 0.7322074788902292,
+      "grad_norm": 0.3993830815115119,
+      "learning_rate": 4.200178731009831e-05,
+      "loss": 0.4585,
+      "step": 607
+    },
+    {
+      "epoch": 0.7334137515078407,
+      "grad_norm": 0.43628851383001205,
+      "learning_rate": 4.197944593386953e-05,
+      "loss": 0.4616,
+      "step": 608
+    },
+    {
+      "epoch": 0.7346200241254524,
+      "grad_norm": 0.31815655209223725,
+      "learning_rate": 4.1957104557640756e-05,
+      "loss": 0.4101,
+      "step": 609
+    },
+    {
+      "epoch": 0.7358262967430639,
+      "grad_norm": 0.3673715240625822,
+      "learning_rate": 4.1934763181411976e-05,
+      "loss": 0.4284,
+      "step": 610
+    },
+    {
+      "epoch": 0.7370325693606755,
+      "grad_norm": 0.3683279900141664,
+      "learning_rate": 4.19124218051832e-05,
+      "loss": 0.4392,
+      "step": 611
+    },
+    {
+      "epoch": 0.738238841978287,
+      "grad_norm": 0.3619292606667547,
+      "learning_rate": 4.1890080428954424e-05,
+      "loss": 0.4297,
+      "step": 612
+    },
+    {
+      "epoch": 0.7394451145958987,
+      "grad_norm": 0.35361671306430015,
+      "learning_rate": 4.186773905272565e-05,
+      "loss": 0.4434,
+      "step": 613
+    },
+    {
+      "epoch": 0.7406513872135102,
+      "grad_norm": 0.31710675852694437,
+      "learning_rate": 4.184539767649687e-05,
+      "loss": 0.4461,
+      "step": 614
+    },
+    {
+      "epoch": 0.7418576598311218,
+      "grad_norm": 0.35186947742683367,
+      "learning_rate": 4.18230563002681e-05,
+      "loss": 0.4279,
+      "step": 615
+    },
+    {
+      "epoch": 0.7430639324487334,
+      "grad_norm": 0.3445887427896891,
+      "learning_rate": 4.180071492403933e-05,
+      "loss": 0.4273,
+      "step": 616
+    },
+    {
+      "epoch": 0.744270205066345,
+      "grad_norm": 0.3681697377341119,
+      "learning_rate": 4.177837354781055e-05,
+      "loss": 0.4477,
+      "step": 617
+    },
+    {
+      "epoch": 0.7454764776839565,
+      "grad_norm": 0.38596304651196306,
+      "learning_rate": 4.175603217158177e-05,
+      "loss": 0.4548,
+      "step": 618
+    },
+    {
+      "epoch": 0.7466827503015682,
+      "grad_norm": 0.33378736186801616,
+      "learning_rate": 4.1733690795352996e-05,
+      "loss": 0.439,
+      "step": 619
+    },
+    {
+      "epoch": 0.7478890229191797,
+      "grad_norm": 0.41780476815690276,
+      "learning_rate": 4.171134941912422e-05,
+      "loss": 0.4344,
+      "step": 620
+    },
+    {
+      "epoch": 0.7490952955367913,
+      "grad_norm": 0.3404196374099104,
+      "learning_rate": 4.1689008042895445e-05,
+      "loss": 0.4204,
+      "step": 621
+    },
+    {
+      "epoch": 0.7503015681544029,
+      "grad_norm": 0.3631500695840166,
+      "learning_rate": 4.166666666666667e-05,
+      "loss": 0.4368,
+      "step": 622
+    },
+    {
+      "epoch": 0.7515078407720145,
+      "grad_norm": 0.33649523812124915,
+      "learning_rate": 4.164432529043789e-05,
+      "loss": 0.4411,
+      "step": 623
+    },
+    {
+      "epoch": 0.752714113389626,
+      "grad_norm": 0.4023001151383432,
+      "learning_rate": 4.162198391420912e-05,
+      "loss": 0.4318,
+      "step": 624
+    },
+    {
+      "epoch": 0.7539203860072377,
+      "grad_norm": 0.3827957756912793,
+      "learning_rate": 4.159964253798034e-05,
+      "loss": 0.4177,
+      "step": 625
+    },
+    {
+      "epoch": 0.7551266586248492,
+      "grad_norm": 0.39090196736454064,
+      "learning_rate": 4.157730116175156e-05,
+      "loss": 0.4113,
+      "step": 626
+    },
+    {
+      "epoch": 0.7563329312424608,
+      "grad_norm": 0.33944855823417563,
+      "learning_rate": 4.155495978552279e-05,
+      "loss": 0.4229,
+      "step": 627
+    },
+    {
+      "epoch": 0.7575392038600723,
+      "grad_norm": 0.4417958155266135,
+      "learning_rate": 4.1532618409294017e-05,
+      "loss": 0.4383,
+      "step": 628
+    },
+    {
+      "epoch": 0.758745476477684,
+      "grad_norm": 0.5006482946012983,
+      "learning_rate": 4.151027703306524e-05,
+      "loss": 0.426,
+      "step": 629
+    },
+    {
+      "epoch": 0.7599517490952955,
+      "grad_norm": 0.32717617902623447,
+      "learning_rate": 4.1487935656836465e-05,
+      "loss": 0.4379,
+      "step": 630
+    },
+    {
+      "epoch": 0.7611580217129071,
+      "grad_norm": 0.45592424124753594,
+      "learning_rate": 4.146559428060769e-05,
+      "loss": 0.4236,
+      "step": 631
+    },
+    {
+      "epoch": 0.7623642943305187,
+      "grad_norm": 0.347387273411557,
+      "learning_rate": 4.144325290437891e-05,
+      "loss": 0.4124,
+      "step": 632
+    },
+    {
+      "epoch": 0.7635705669481303,
+      "grad_norm": 0.3721402905982405,
+      "learning_rate": 4.1420911528150134e-05,
+      "loss": 0.411,
+      "step": 633
+    },
+    {
+      "epoch": 0.7647768395657418,
+      "grad_norm": 0.3970352649019345,
+      "learning_rate": 4.139857015192136e-05,
+      "loss": 0.4579,
+      "step": 634
+    },
+    {
+      "epoch": 0.7659831121833535,
+      "grad_norm": 0.43579423143308493,
+      "learning_rate": 4.137622877569258e-05,
+      "loss": 0.4296,
+      "step": 635
+    },
+    {
+      "epoch": 0.767189384800965,
+      "grad_norm": 0.39279465007605335,
+      "learning_rate": 4.135388739946381e-05,
+      "loss": 0.4231,
+      "step": 636
+    },
+    {
+      "epoch": 0.7683956574185766,
+      "grad_norm": 0.39264419792414523,
+      "learning_rate": 4.133154602323504e-05,
+      "loss": 0.4496,
+      "step": 637
+    },
+    {
+      "epoch": 0.7696019300361882,
+      "grad_norm": 0.37493841849128356,
+      "learning_rate": 4.130920464700626e-05,
+      "loss": 0.4291,
+      "step": 638
+    },
+    {
+      "epoch": 0.7708082026537998,
+      "grad_norm": 0.3116259448118478,
+      "learning_rate": 4.1286863270777485e-05,
+      "loss": 0.4065,
+      "step": 639
+    },
+    {
+      "epoch": 0.7720144752714113,
+      "grad_norm": 0.34983397918562,
+      "learning_rate": 4.1264521894548706e-05,
+      "loss": 0.4344,
+      "step": 640
+    },
+    {
+      "epoch": 0.773220747889023,
+      "grad_norm": 0.37624317148252306,
+      "learning_rate": 4.1242180518319926e-05,
+      "loss": 0.4294,
+      "step": 641
+    },
+    {
+      "epoch": 0.7744270205066345,
+      "grad_norm": 0.35520939634197213,
+      "learning_rate": 4.1219839142091154e-05,
+      "loss": 0.4148,
+      "step": 642
+    },
+    {
+      "epoch": 0.7756332931242461,
+      "grad_norm": 0.4209679722778236,
+      "learning_rate": 4.119749776586238e-05,
+      "loss": 0.4405,
+      "step": 643
+    },
+    {
+      "epoch": 0.7768395657418576,
+      "grad_norm": 0.37647621537087084,
+      "learning_rate": 4.11751563896336e-05,
+      "loss": 0.4428,
+      "step": 644
+    },
+    {
+      "epoch": 0.7780458383594693,
+      "grad_norm": 0.444753555144684,
+      "learning_rate": 4.115281501340483e-05,
+      "loss": 0.4147,
+      "step": 645
+    },
+    {
+      "epoch": 0.7792521109770808,
+      "grad_norm": 0.38638377205425406,
+      "learning_rate": 4.113047363717605e-05,
+      "loss": 0.4444,
+      "step": 646
+    },
+    {
+      "epoch": 0.7804583835946924,
+      "grad_norm": 0.3403369626116202,
+      "learning_rate": 4.110813226094727e-05,
+      "loss": 0.4094,
+      "step": 647
+    },
+    {
+      "epoch": 0.781664656212304,
+      "grad_norm": 0.4152796308986888,
+      "learning_rate": 4.10857908847185e-05,
+      "loss": 0.4184,
+      "step": 648
+    },
+    {
+      "epoch": 0.7828709288299156,
+      "grad_norm": 0.42659585360095775,
+      "learning_rate": 4.1063449508489726e-05,
+      "loss": 0.4376,
+      "step": 649
+    },
+    {
+      "epoch": 0.7840772014475271,
+      "grad_norm": 0.3878286833998237,
+      "learning_rate": 4.1041108132260947e-05,
+      "loss": 0.4424,
+      "step": 650
+    },
+    {
+      "epoch": 0.7852834740651388,
+      "grad_norm": 0.4279745826656998,
+      "learning_rate": 4.1018766756032174e-05,
+      "loss": 0.4171,
+      "step": 651
+    },
+    {
+      "epoch": 0.7864897466827503,
+      "grad_norm": 0.4037837710960183,
+      "learning_rate": 4.09964253798034e-05,
+      "loss": 0.4209,
+      "step": 652
+    },
+    {
+      "epoch": 0.7876960193003619,
+      "grad_norm": 0.328778160496694,
+      "learning_rate": 4.097408400357462e-05,
+      "loss": 0.4161,
+      "step": 653
+    },
+    {
+      "epoch": 0.7889022919179735,
+      "grad_norm": 0.5091749659268239,
+      "learning_rate": 4.095174262734584e-05,
+      "loss": 0.4059,
+      "step": 654
+    },
+    {
+      "epoch": 0.7901085645355851,
+      "grad_norm": 0.38575372225419235,
+      "learning_rate": 4.092940125111707e-05,
+      "loss": 0.4306,
+      "step": 655
+    },
+    {
+      "epoch": 0.7913148371531966,
+      "grad_norm": 0.3564858997238752,
+      "learning_rate": 4.090705987488829e-05,
+      "loss": 0.4229,
+      "step": 656
+    },
+    {
+      "epoch": 0.7925211097708083,
+      "grad_norm": 0.33263580029482265,
+      "learning_rate": 4.088471849865952e-05,
+      "loss": 0.4238,
+      "step": 657
+    },
+    {
+      "epoch": 0.7937273823884198,
+      "grad_norm": 0.3770972817311997,
+      "learning_rate": 4.0862377122430746e-05,
+      "loss": 0.4215,
+      "step": 658
+    },
+    {
+      "epoch": 0.7949336550060314,
+      "grad_norm": 0.34460138647484384,
+      "learning_rate": 4.084003574620197e-05,
+      "loss": 0.4624,
+      "step": 659
+    },
+    {
+      "epoch": 0.7961399276236429,
+      "grad_norm": 0.39795136433340583,
+      "learning_rate": 4.0817694369973194e-05,
+      "loss": 0.4717,
+      "step": 660
+    },
+    {
+      "epoch": 0.7973462002412546,
+      "grad_norm": 0.307065657764241,
+      "learning_rate": 4.0795352993744415e-05,
+      "loss": 0.4085,
+      "step": 661
+    },
+    {
+      "epoch": 0.7985524728588661,
+      "grad_norm": 0.36134507736793015,
+      "learning_rate": 4.0773011617515636e-05,
+      "loss": 0.4253,
+      "step": 662
+    },
+    {
+      "epoch": 0.7997587454764777,
+      "grad_norm": 0.36337250256856457,
+      "learning_rate": 4.075067024128686e-05,
+      "loss": 0.423,
+      "step": 663
+    },
+    {
+      "epoch": 0.8009650180940893,
+      "grad_norm": 0.34792061661231566,
+      "learning_rate": 4.072832886505809e-05,
+      "loss": 0.4099,
+      "step": 664
+    },
+    {
+      "epoch": 0.8021712907117008,
+      "grad_norm": 0.32356302620167704,
+      "learning_rate": 4.070598748882931e-05,
+      "loss": 0.4252,
+      "step": 665
+    },
+    {
+      "epoch": 0.8033775633293124,
+      "grad_norm": 0.32657888669961704,
+      "learning_rate": 4.068364611260054e-05,
+      "loss": 0.4373,
+      "step": 666
+    },
+    {
+      "epoch": 0.804583835946924,
+      "grad_norm": 0.37539157769448306,
+      "learning_rate": 4.0661304736371766e-05,
+      "loss": 0.4399,
+      "step": 667
+    },
+    {
+      "epoch": 0.8057901085645356,
+      "grad_norm": 0.3311339803296136,
+      "learning_rate": 4.063896336014299e-05,
+      "loss": 0.436,
+      "step": 668
+    },
+    {
+      "epoch": 0.8069963811821471,
+      "grad_norm": 0.36972556713265975,
+      "learning_rate": 4.061662198391421e-05,
+      "loss": 0.432,
+      "step": 669
+    },
+    {
+      "epoch": 0.8082026537997588,
+      "grad_norm": 0.37032736900349966,
+      "learning_rate": 4.0594280607685435e-05,
+      "loss": 0.4483,
+      "step": 670
+    },
+    {
+      "epoch": 0.8094089264173703,
+      "grad_norm": 0.35889492771597925,
+      "learning_rate": 4.0571939231456656e-05,
+      "loss": 0.4384,
+      "step": 671
+    },
+    {
+      "epoch": 0.8106151990349819,
+      "grad_norm": 0.3917827657498273,
+      "learning_rate": 4.054959785522788e-05,
+      "loss": 0.4586,
+      "step": 672
+    },
+    {
+      "epoch": 0.8118214716525934,
+      "grad_norm": 0.3449677493546546,
+      "learning_rate": 4.052725647899911e-05,
+      "loss": 0.4515,
+      "step": 673
+    },
+    {
+      "epoch": 0.8130277442702051,
+      "grad_norm": 0.3675384445750401,
+      "learning_rate": 4.050491510277033e-05,
+      "loss": 0.4085,
+      "step": 674
+    },
+    {
+      "epoch": 0.8142340168878166,
+      "grad_norm": 0.314056093770354,
+      "learning_rate": 4.048257372654156e-05,
+      "loss": 0.4145,
+      "step": 675
+    },
+    {
+      "epoch": 0.8154402895054282,
+      "grad_norm": 0.2884839297871985,
+      "learning_rate": 4.046023235031278e-05,
+      "loss": 0.4085,
+      "step": 676
+    },
+    {
+      "epoch": 0.8166465621230398,
+      "grad_norm": 0.3419757025514419,
+      "learning_rate": 4.0437890974084e-05,
+      "loss": 0.4372,
+      "step": 677
+    },
+    {
+      "epoch": 0.8178528347406514,
+      "grad_norm": 0.3350898764681184,
+      "learning_rate": 4.041554959785523e-05,
+      "loss": 0.4232,
+      "step": 678
+    },
+    {
+      "epoch": 0.8190591073582629,
+      "grad_norm": 0.3101794969989681,
+      "learning_rate": 4.0393208221626455e-05,
+      "loss": 0.4118,
+      "step": 679
+    },
+    {
+      "epoch": 0.8202653799758746,
+      "grad_norm": 0.34632856979523563,
+      "learning_rate": 4.0370866845397676e-05,
+      "loss": 0.3988,
+      "step": 680
+    },
+    {
+      "epoch": 0.8214716525934861,
+      "grad_norm": 0.3539539495668904,
+      "learning_rate": 4.0348525469168903e-05,
+      "loss": 0.4244,
+      "step": 681
+    },
+    {
+      "epoch": 0.8226779252110977,
+      "grad_norm": 0.31397960297819166,
+      "learning_rate": 4.032618409294013e-05,
+      "loss": 0.4191,
+      "step": 682
+    },
+    {
+      "epoch": 0.8238841978287093,
+      "grad_norm": 0.4450760042397345,
+      "learning_rate": 4.030384271671135e-05,
+      "loss": 0.4417,
+      "step": 683
+    },
+    {
+      "epoch": 0.8250904704463209,
+      "grad_norm": 0.3758257266264463,
+      "learning_rate": 4.028150134048257e-05,
+      "loss": 0.4326,
+      "step": 684
+    },
+    {
+      "epoch": 0.8262967430639324,
+      "grad_norm": 0.4347027419099405,
+      "learning_rate": 4.02591599642538e-05,
+      "loss": 0.432,
+      "step": 685
+    },
+    {
+      "epoch": 0.827503015681544,
+      "grad_norm": 0.33281647369934925,
+      "learning_rate": 4.023681858802503e-05,
+      "loss": 0.4443,
+      "step": 686
+    },
+    {
+      "epoch": 0.8287092882991556,
+      "grad_norm": 0.44210413889252337,
+      "learning_rate": 4.021447721179625e-05,
+      "loss": 0.4365,
+      "step": 687
+    },
+    {
+      "epoch": 0.8299155609167672,
+      "grad_norm": 0.36379774369825985,
+      "learning_rate": 4.0192135835567475e-05,
+      "loss": 0.4407,
+      "step": 688
+    },
+    {
+      "epoch": 0.8311218335343787,
+      "grad_norm": 0.3409375216448993,
+      "learning_rate": 4.01697944593387e-05,
+      "loss": 0.4163,
+      "step": 689
+    },
+    {
+      "epoch": 0.8323281061519904,
+      "grad_norm": 0.4407071439643426,
+      "learning_rate": 4.0147453083109924e-05,
+      "loss": 0.4289,
+      "step": 690
+    },
+    {
+      "epoch": 0.8335343787696019,
+      "grad_norm": 0.3440826546937714,
+      "learning_rate": 4.0125111706881144e-05,
+      "loss": 0.4262,
+      "step": 691
+    },
+    {
+      "epoch": 0.8347406513872135,
+      "grad_norm": 0.3415814685987661,
+      "learning_rate": 4.010277033065237e-05,
+      "loss": 0.4203,
+      "step": 692
+    },
+    {
+      "epoch": 0.8359469240048251,
+      "grad_norm": 0.3950448918022183,
+      "learning_rate": 4.008042895442359e-05,
+      "loss": 0.4552,
+      "step": 693
+    },
+    {
+      "epoch": 0.8371531966224367,
+      "grad_norm": 0.3682136713582369,
+      "learning_rate": 4.005808757819482e-05,
+      "loss": 0.4293,
+      "step": 694
+    },
+    {
+      "epoch": 0.8383594692400482,
+      "grad_norm": 0.3245379694872947,
+      "learning_rate": 4.003574620196605e-05,
+      "loss": 0.4466,
+      "step": 695
+    },
+    {
+      "epoch": 0.8395657418576599,
+      "grad_norm": 0.4043111926095754,
+      "learning_rate": 4.001340482573727e-05,
+      "loss": 0.4124,
+      "step": 696
+    },
+    {
+      "epoch": 0.8407720144752714,
+      "grad_norm": 0.3676847847505625,
+      "learning_rate": 3.9991063449508496e-05,
+      "loss": 0.4386,
+      "step": 697
+    },
+    {
+      "epoch": 0.841978287092883,
+      "grad_norm": 0.38345621376583905,
+      "learning_rate": 3.9968722073279716e-05,
+      "loss": 0.4206,
+      "step": 698
+    },
+    {
+      "epoch": 0.8431845597104946,
+      "grad_norm": 0.3973129732421497,
+      "learning_rate": 3.994638069705094e-05,
+      "loss": 0.444,
+      "step": 699
+    },
+    {
+      "epoch": 0.8443908323281062,
+      "grad_norm": 0.40507692158099656,
+      "learning_rate": 3.9924039320822164e-05,
+      "loss": 0.4292,
+      "step": 700
+    },
+    {
+      "epoch": 0.8455971049457177,
+      "grad_norm": 0.44304365309322413,
+      "learning_rate": 3.990169794459339e-05,
+      "loss": 0.4351,
+      "step": 701
+    },
+    {
+      "epoch": 0.8468033775633294,
+      "grad_norm": 0.34455473505953255,
+      "learning_rate": 3.987935656836461e-05,
+      "loss": 0.4377,
+      "step": 702
+    },
+    {
+      "epoch": 0.8480096501809409,
+      "grad_norm": 0.41711458301265714,
+      "learning_rate": 3.985701519213584e-05,
+      "loss": 0.432,
+      "step": 703
+    },
+    {
+      "epoch": 0.8492159227985525,
+      "grad_norm": 0.3930062754782553,
+      "learning_rate": 3.983467381590707e-05,
+      "loss": 0.4352,
+      "step": 704
+    },
+    {
+      "epoch": 0.850422195416164,
+      "grad_norm": 0.45827478708504854,
+      "learning_rate": 3.981233243967828e-05,
+      "loss": 0.4356,
+      "step": 705
+    },
+    {
+      "epoch": 0.8516284680337757,
+      "grad_norm": 0.4100191953270183,
+      "learning_rate": 3.978999106344951e-05,
+      "loss": 0.4113,
+      "step": 706
+    },
+    {
+      "epoch": 0.8528347406513872,
+      "grad_norm": 0.3674509170919512,
+      "learning_rate": 3.9767649687220737e-05,
+      "loss": 0.4135,
+      "step": 707
+    },
+    {
+      "epoch": 0.8540410132689988,
+      "grad_norm": 0.4479376512445783,
+      "learning_rate": 3.974530831099196e-05,
+      "loss": 0.4278,
+      "step": 708
+    },
+    {
+      "epoch": 0.8552472858866104,
+      "grad_norm": 0.38078667965255886,
+      "learning_rate": 3.9722966934763185e-05,
+      "loss": 0.4362,
+      "step": 709
+    },
+    {
+      "epoch": 0.856453558504222,
+      "grad_norm": 0.3795248071192456,
+      "learning_rate": 3.970062555853441e-05,
+      "loss": 0.4256,
+      "step": 710
+    },
+    {
+      "epoch": 0.8576598311218335,
+      "grad_norm": 0.4452777512730255,
+      "learning_rate": 3.967828418230563e-05,
+      "loss": 0.4048,
+      "step": 711
+    },
+    {
+      "epoch": 0.8588661037394452,
+      "grad_norm": 0.4039038032056285,
+      "learning_rate": 3.9655942806076854e-05,
+      "loss": 0.4305,
+      "step": 712
+    },
+    {
+      "epoch": 0.8600723763570567,
+      "grad_norm": 0.38705436215447936,
+      "learning_rate": 3.963360142984808e-05,
+      "loss": 0.4338,
+      "step": 713
+    },
+    {
+      "epoch": 0.8612786489746683,
+      "grad_norm": 0.37612795284145023,
+      "learning_rate": 3.96112600536193e-05,
+      "loss": 0.4356,
+      "step": 714
+    },
+    {
+      "epoch": 0.8624849215922799,
+      "grad_norm": 0.4027892381276643,
+      "learning_rate": 3.958891867739053e-05,
+      "loss": 0.4259,
+      "step": 715
+    },
+    {
+      "epoch": 0.8636911942098915,
+      "grad_norm": 0.4474291942673631,
+      "learning_rate": 3.956657730116176e-05,
+      "loss": 0.4191,
+      "step": 716
+    },
+    {
+      "epoch": 0.864897466827503,
+      "grad_norm": 0.37294989078633567,
+      "learning_rate": 3.954423592493298e-05,
+      "loss": 0.4421,
+      "step": 717
+    },
+    {
+      "epoch": 0.8661037394451147,
+      "grad_norm": 0.39823462537023213,
+      "learning_rate": 3.9521894548704205e-05,
+      "loss": 0.4038,
+      "step": 718
+    },
+    {
+      "epoch": 0.8673100120627262,
+      "grad_norm": 0.3165892994057179,
+      "learning_rate": 3.9499553172475426e-05,
+      "loss": 0.4276,
+      "step": 719
+    },
+    {
+      "epoch": 0.8685162846803377,
+      "grad_norm": 0.3217081991040394,
+      "learning_rate": 3.9477211796246646e-05,
+      "loss": 0.4421,
+      "step": 720
+    },
+    {
+      "epoch": 0.8697225572979493,
+      "grad_norm": 0.34234595334126144,
+      "learning_rate": 3.9454870420017874e-05,
+      "loss": 0.4153,
+      "step": 721
+    },
+    {
+      "epoch": 0.8709288299155609,
+      "grad_norm": 0.3012747513317574,
+      "learning_rate": 3.94325290437891e-05,
+      "loss": 0.431,
+      "step": 722
+    },
+    {
+      "epoch": 0.8721351025331725,
+      "grad_norm": 0.4426572348450103,
+      "learning_rate": 3.941018766756032e-05,
+      "loss": 0.4614,
+      "step": 723
+    },
+    {
+      "epoch": 0.873341375150784,
+      "grad_norm": 0.30865825529588037,
+      "learning_rate": 3.938784629133155e-05,
+      "loss": 0.4376,
+      "step": 724
+    },
+    {
+      "epoch": 0.8745476477683957,
+      "grad_norm": 0.3804165775048893,
+      "learning_rate": 3.936550491510278e-05,
+      "loss": 0.4242,
+      "step": 725
+    },
+    {
+      "epoch": 0.8757539203860072,
+      "grad_norm": 0.3590587327120939,
+      "learning_rate": 3.9343163538874e-05,
+      "loss": 0.4247,
+      "step": 726
+    },
+    {
+      "epoch": 0.8769601930036188,
+      "grad_norm": 0.38885826359929504,
+      "learning_rate": 3.932082216264522e-05,
+      "loss": 0.4548,
+      "step": 727
+    },
+    {
+      "epoch": 0.8781664656212304,
+      "grad_norm": 0.34482562337158273,
+      "learning_rate": 3.9298480786416446e-05,
+      "loss": 0.4311,
+      "step": 728
+    },
+    {
+      "epoch": 0.879372738238842,
+      "grad_norm": 0.4135444476948955,
+      "learning_rate": 3.9276139410187666e-05,
+      "loss": 0.4328,
+      "step": 729
+    },
+    {
+      "epoch": 0.8805790108564535,
+      "grad_norm": 0.3520471335612592,
+      "learning_rate": 3.9253798033958894e-05,
+      "loss": 0.4166,
+      "step": 730
+    },
+    {
+      "epoch": 0.8817852834740652,
+      "grad_norm": 0.4012911512212805,
+      "learning_rate": 3.923145665773012e-05,
+      "loss": 0.4189,
+      "step": 731
+    },
+    {
+      "epoch": 0.8829915560916767,
+      "grad_norm": 0.38055005376337286,
+      "learning_rate": 3.920911528150134e-05,
+      "loss": 0.4343,
+      "step": 732
+    },
+    {
+      "epoch": 0.8841978287092883,
+      "grad_norm": 0.3525565527349451,
+      "learning_rate": 3.918677390527257e-05,
+      "loss": 0.4211,
+      "step": 733
+    },
+    {
+      "epoch": 0.8854041013268998,
+      "grad_norm": 0.44952656074215147,
+      "learning_rate": 3.916443252904379e-05,
+      "loss": 0.4247,
+      "step": 734
+    },
+    {
+      "epoch": 0.8866103739445115,
+      "grad_norm": 0.38931178479387923,
+      "learning_rate": 3.914209115281501e-05,
+      "loss": 0.4091,
+      "step": 735
+    },
+    {
+      "epoch": 0.887816646562123,
+      "grad_norm": 0.4295833832365165,
+      "learning_rate": 3.911974977658624e-05,
+      "loss": 0.4273,
+      "step": 736
+    },
+    {
+      "epoch": 0.8890229191797346,
+      "grad_norm": 0.418788573846933,
+      "learning_rate": 3.9097408400357466e-05,
+      "loss": 0.4324,
+      "step": 737
+    },
+    {
+      "epoch": 0.8902291917973462,
+      "grad_norm": 0.4279678016240271,
+      "learning_rate": 3.907506702412869e-05,
+      "loss": 0.4369,
+      "step": 738
+    },
+    {
+      "epoch": 0.8914354644149578,
+      "grad_norm": 0.4706663367413349,
+      "learning_rate": 3.9052725647899914e-05,
+      "loss": 0.4425,
+      "step": 739
+    },
+    {
+      "epoch": 0.8926417370325693,
+      "grad_norm": 0.33644996594988985,
+      "learning_rate": 3.903038427167114e-05,
+      "loss": 0.4333,
+      "step": 740
+    },
+    {
+      "epoch": 0.893848009650181,
+      "grad_norm": 0.5219711361177203,
+      "learning_rate": 3.900804289544236e-05,
+      "loss": 0.4139,
+      "step": 741
+    },
+    {
+      "epoch": 0.8950542822677925,
+      "grad_norm": 0.3655070920705575,
+      "learning_rate": 3.898570151921358e-05,
+      "loss": 0.4215,
+      "step": 742
+    },
+    {
+      "epoch": 0.8962605548854041,
+      "grad_norm": 0.4620182845103301,
+      "learning_rate": 3.896336014298481e-05,
+      "loss": 0.4391,
+      "step": 743
+    },
+    {
+      "epoch": 0.8974668275030157,
+      "grad_norm": 0.446013501499081,
+      "learning_rate": 3.894101876675603e-05,
+      "loss": 0.4113,
+      "step": 744
+    },
+    {
+      "epoch": 0.8986731001206273,
+      "grad_norm": 0.34459110535041027,
+      "learning_rate": 3.891867739052726e-05,
+      "loss": 0.418,
+      "step": 745
+    },
+    {
+      "epoch": 0.8998793727382388,
+      "grad_norm": 0.39907797574459497,
+      "learning_rate": 3.8896336014298486e-05,
+      "loss": 0.4245,
+      "step": 746
+    },
+    {
+      "epoch": 0.9010856453558505,
+      "grad_norm": 0.3664286457363056,
+      "learning_rate": 3.887399463806971e-05,
+      "loss": 0.4253,
+      "step": 747
+    },
+    {
+      "epoch": 0.902291917973462,
+      "grad_norm": 0.41961271222361207,
+      "learning_rate": 3.8851653261840934e-05,
+      "loss": 0.422,
+      "step": 748
+    },
+    {
+      "epoch": 0.9034981905910736,
+      "grad_norm": 0.34696641892186714,
+      "learning_rate": 3.8829311885612155e-05,
+      "loss": 0.4262,
+      "step": 749
+    },
+    {
+      "epoch": 0.9047044632086851,
+      "grad_norm": 0.3805847464393101,
+      "learning_rate": 3.8806970509383376e-05,
+      "loss": 0.4222,
+      "step": 750
+    },
+    {
+      "epoch": 0.9059107358262968,
+      "grad_norm": 0.3628814065269647,
+      "learning_rate": 3.87846291331546e-05,
+      "loss": 0.415,
+      "step": 751
+    },
+    {
+      "epoch": 0.9071170084439083,
+      "grad_norm": 0.4155607336726821,
+      "learning_rate": 3.876228775692583e-05,
+      "loss": 0.4529,
+      "step": 752
+    },
+    {
+      "epoch": 0.9083232810615199,
+      "grad_norm": 0.3792089287803448,
+      "learning_rate": 3.873994638069705e-05,
+      "loss": 0.4427,
+      "step": 753
+    },
+    {
+      "epoch": 0.9095295536791315,
+      "grad_norm": 0.31130821301505,
+      "learning_rate": 3.871760500446828e-05,
+      "loss": 0.4065,
+      "step": 754
+    },
+    {
+      "epoch": 0.9107358262967431,
+      "grad_norm": 0.3129142676327276,
+      "learning_rate": 3.8695263628239506e-05,
+      "loss": 0.4089,
+      "step": 755
+    },
+    {
+      "epoch": 0.9119420989143546,
+      "grad_norm": 0.3903036615010061,
+      "learning_rate": 3.867292225201073e-05,
+      "loss": 0.4222,
+      "step": 756
+    },
+    {
+      "epoch": 0.9131483715319663,
+      "grad_norm": 0.3578101284227395,
+      "learning_rate": 3.865058087578195e-05,
+      "loss": 0.4048,
+      "step": 757
+    },
+    {
+      "epoch": 0.9143546441495778,
+      "grad_norm": 0.40198410528206696,
+      "learning_rate": 3.8628239499553175e-05,
+      "loss": 0.4247,
+      "step": 758
+    },
+    {
+      "epoch": 0.9155609167671894,
+      "grad_norm": 0.469693215846757,
+      "learning_rate": 3.8605898123324396e-05,
+      "loss": 0.4214,
+      "step": 759
+    },
+    {
+      "epoch": 0.916767189384801,
+      "grad_norm": 0.3485330736681551,
+      "learning_rate": 3.858355674709562e-05,
+      "loss": 0.4394,
+      "step": 760
+    },
+    {
+      "epoch": 0.9179734620024126,
+      "grad_norm": 0.5510670824668625,
+      "learning_rate": 3.856121537086685e-05,
+      "loss": 0.4265,
+      "step": 761
+    },
+    {
+      "epoch": 0.9191797346200241,
+      "grad_norm": 0.39837727033428083,
+      "learning_rate": 3.853887399463807e-05,
+      "loss": 0.418,
+      "step": 762
+    },
+    {
+      "epoch": 0.9203860072376358,
+      "grad_norm": 0.41864197863604735,
+      "learning_rate": 3.851653261840929e-05,
+      "loss": 0.4319,
+      "step": 763
+    },
+    {
+      "epoch": 0.9215922798552473,
+      "grad_norm": 0.4574033736259981,
+      "learning_rate": 3.849419124218052e-05,
+      "loss": 0.4292,
+      "step": 764
+    },
+    {
+      "epoch": 0.9227985524728589,
+      "grad_norm": 0.3213110702684741,
+      "learning_rate": 3.847184986595174e-05,
+      "loss": 0.4098,
+      "step": 765
+    },
+    {
+      "epoch": 0.9240048250904704,
+      "grad_norm": 0.4707116003780227,
+      "learning_rate": 3.844950848972297e-05,
+      "loss": 0.4244,
+      "step": 766
+    },
+    {
+      "epoch": 0.9252110977080821,
+      "grad_norm": 0.4830169788101642,
+      "learning_rate": 3.8427167113494195e-05,
+      "loss": 0.4157,
+      "step": 767
+    },
+    {
+      "epoch": 0.9264173703256936,
+      "grad_norm": 0.4560897134292833,
+      "learning_rate": 3.8404825737265416e-05,
+      "loss": 0.4393,
+      "step": 768
+    },
+    {
+      "epoch": 0.9276236429433052,
+      "grad_norm": 0.6482326714956513,
+      "learning_rate": 3.8382484361036644e-05,
+      "loss": 0.4185,
+      "step": 769
+    },
+    {
+      "epoch": 0.9288299155609168,
+      "grad_norm": 0.4137410345743272,
+      "learning_rate": 3.8360142984807864e-05,
+      "loss": 0.4452,
+      "step": 770
+    },
+    {
+      "epoch": 0.9300361881785284,
+      "grad_norm": 0.6822256387321779,
+      "learning_rate": 3.8337801608579085e-05,
+      "loss": 0.4293,
+      "step": 771
+    },
+    {
+      "epoch": 0.9312424607961399,
+      "grad_norm": 0.30603838243949,
+      "learning_rate": 3.831546023235031e-05,
+      "loss": 0.4346,
+      "step": 772
+    },
+    {
+      "epoch": 0.9324487334137516,
+      "grad_norm": 0.5898745156575074,
+      "learning_rate": 3.829311885612154e-05,
+      "loss": 0.4231,
+      "step": 773
+    },
+    {
+      "epoch": 0.9336550060313631,
+      "grad_norm": 0.33761968440146145,
+      "learning_rate": 3.827077747989276e-05,
+      "loss": 0.4153,
+      "step": 774
+    },
+    {
+      "epoch": 0.9348612786489746,
+      "grad_norm": 0.49516978453532084,
+      "learning_rate": 3.824843610366399e-05,
+      "loss": 0.4038,
+      "step": 775
+    },
+    {
+      "epoch": 0.9360675512665863,
+      "grad_norm": 0.4513687887601463,
+      "learning_rate": 3.8226094727435216e-05,
+      "loss": 0.4322,
+      "step": 776
+    },
+    {
+      "epoch": 0.9372738238841978,
+      "grad_norm": 0.4062297835365191,
+      "learning_rate": 3.8203753351206436e-05,
+      "loss": 0.4186,
+      "step": 777
+    },
+    {
+      "epoch": 0.9384800965018094,
+      "grad_norm": 0.48900573314200335,
+      "learning_rate": 3.818141197497766e-05,
+      "loss": 0.4166,
+      "step": 778
+    },
+    {
+      "epoch": 0.9396863691194209,
+      "grad_norm": 0.35020953986328074,
+      "learning_rate": 3.8159070598748884e-05,
+      "loss": 0.4244,
+      "step": 779
+    },
+    {
+      "epoch": 0.9408926417370326,
+      "grad_norm": 0.3670307965989599,
+      "learning_rate": 3.8136729222520105e-05,
+      "loss": 0.4629,
+      "step": 780
+    },
+    {
+      "epoch": 0.9420989143546441,
+      "grad_norm": 1.247256002324837,
+      "learning_rate": 3.811438784629133e-05,
+      "loss": 0.441,
+      "step": 781
+    },
+    {
+      "epoch": 0.9433051869722557,
+      "grad_norm": 0.3886395489272093,
+      "learning_rate": 3.809204647006256e-05,
+      "loss": 0.4312,
+      "step": 782
+    },
+    {
+      "epoch": 0.9445114595898673,
+      "grad_norm": 0.3762586952639499,
+      "learning_rate": 3.806970509383378e-05,
+      "loss": 0.429,
+      "step": 783
+    },
+    {
+      "epoch": 0.9457177322074789,
+      "grad_norm": 0.35565515533530456,
+      "learning_rate": 3.804736371760501e-05,
+      "loss": 0.4255,
+      "step": 784
+    },
+    {
+      "epoch": 0.9469240048250904,
+      "grad_norm": 0.33572714507490936,
+      "learning_rate": 3.802502234137623e-05,
+      "loss": 0.4103,
+      "step": 785
+    },
+    {
+      "epoch": 0.9481302774427021,
+      "grad_norm": 0.4421855761325749,
+      "learning_rate": 3.800268096514745e-05,
+      "loss": 0.4156,
+      "step": 786
+    },
+    {
+      "epoch": 0.9493365500603136,
+      "grad_norm": 0.3481572570009158,
+      "learning_rate": 3.798033958891868e-05,
+      "loss": 0.4246,
+      "step": 787
+    },
+    {
+      "epoch": 0.9505428226779252,
+      "grad_norm": 0.5149461446668178,
+      "learning_rate": 3.7957998212689905e-05,
+      "loss": 0.4447,
+      "step": 788
+    },
+    {
+      "epoch": 0.9517490952955368,
+      "grad_norm": 0.32657057689577534,
+      "learning_rate": 3.7935656836461125e-05,
+      "loss": 0.4223,
+      "step": 789
+    },
+    {
+      "epoch": 0.9529553679131484,
+      "grad_norm": 0.440663289717245,
+      "learning_rate": 3.791331546023235e-05,
+      "loss": 0.4075,
+      "step": 790
+    },
+    {
+      "epoch": 0.9541616405307599,
+      "grad_norm": 0.49488905419238877,
+      "learning_rate": 3.789097408400358e-05,
+      "loss": 0.4131,
+      "step": 791
+    },
+    {
+      "epoch": 0.9553679131483716,
+      "grad_norm": 0.3722447138794308,
+      "learning_rate": 3.78686327077748e-05,
+      "loss": 0.4196,
+      "step": 792
+    },
+    {
+      "epoch": 0.9565741857659831,
+      "grad_norm": 0.46844136877295967,
+      "learning_rate": 3.784629133154602e-05,
+      "loss": 0.3922,
+      "step": 793
+    },
+    {
+      "epoch": 0.9577804583835947,
+      "grad_norm": 0.3240774092465061,
+      "learning_rate": 3.782394995531725e-05,
+      "loss": 0.4453,
+      "step": 794
+    },
+    {
+      "epoch": 0.9589867310012062,
+      "grad_norm": 0.4267932674577182,
+      "learning_rate": 3.780160857908847e-05,
+      "loss": 0.4405,
+      "step": 795
+    },
+    {
+      "epoch": 0.9601930036188179,
+      "grad_norm": 0.3801824902130055,
+      "learning_rate": 3.77792672028597e-05,
+      "loss": 0.4295,
+      "step": 796
+    },
+    {
+      "epoch": 0.9613992762364294,
+      "grad_norm": 0.3892559150634845,
+      "learning_rate": 3.7756925826630925e-05,
+      "loss": 0.41,
+      "step": 797
+    },
+    {
+      "epoch": 0.962605548854041,
+      "grad_norm": 0.4667236925840342,
+      "learning_rate": 3.7734584450402145e-05,
+      "loss": 0.4374,
+      "step": 798
+    },
+    {
+      "epoch": 0.9638118214716526,
+      "grad_norm": 0.4352607128892435,
+      "learning_rate": 3.771224307417337e-05,
+      "loss": 0.4471,
+      "step": 799
+    },
+    {
+      "epoch": 0.9650180940892642,
+      "grad_norm": 0.3900629779626532,
+      "learning_rate": 3.7689901697944594e-05,
+      "loss": 0.4073,
+      "step": 800
+    },
+    {
+      "epoch": 0.9662243667068757,
+      "grad_norm": 0.43402977782983276,
+      "learning_rate": 3.7667560321715814e-05,
+      "loss": 0.4211,
+      "step": 801
+    },
+    {
+      "epoch": 0.9674306393244874,
+      "grad_norm": 0.3672704556853505,
+      "learning_rate": 3.764521894548704e-05,
+      "loss": 0.4289,
+      "step": 802
+    },
+    {
+      "epoch": 0.9686369119420989,
+      "grad_norm": 0.4648133025117742,
+      "learning_rate": 3.762287756925827e-05,
+      "loss": 0.4105,
+      "step": 803
+    },
+    {
+      "epoch": 0.9698431845597105,
+      "grad_norm": 0.43049540796677793,
+      "learning_rate": 3.760053619302949e-05,
+      "loss": 0.4462,
+      "step": 804
+    },
+    {
+      "epoch": 0.971049457177322,
+      "grad_norm": 0.4007241405451893,
+      "learning_rate": 3.757819481680072e-05,
+      "loss": 0.4373,
+      "step": 805
+    },
+    {
+      "epoch": 0.9722557297949337,
+      "grad_norm": 0.5109239018044767,
+      "learning_rate": 3.7555853440571945e-05,
+      "loss": 0.4517,
+      "step": 806
+    },
+    {
+      "epoch": 0.9734620024125452,
+      "grad_norm": 0.3488495344511992,
+      "learning_rate": 3.7533512064343166e-05,
+      "loss": 0.4318,
+      "step": 807
+    },
+    {
+      "epoch": 0.9746682750301568,
+      "grad_norm": 0.40087038818480636,
+      "learning_rate": 3.7511170688114386e-05,
+      "loss": 0.4323,
+      "step": 808
+    },
+    {
+      "epoch": 0.9758745476477684,
+      "grad_norm": 0.42484548812497636,
+      "learning_rate": 3.7488829311885614e-05,
+      "loss": 0.4263,
+      "step": 809
+    },
+    {
+      "epoch": 0.97708082026538,
+      "grad_norm": 0.4605312280800588,
+      "learning_rate": 3.746648793565684e-05,
+      "loss": 0.4176,
+      "step": 810
+    },
+    {
+      "epoch": 0.9782870928829915,
+      "grad_norm": 0.3833013635577786,
+      "learning_rate": 3.744414655942806e-05,
+      "loss": 0.4191,
+      "step": 811
+    },
+    {
+      "epoch": 0.9794933655006032,
+      "grad_norm": 0.5426098231658908,
+      "learning_rate": 3.742180518319929e-05,
+      "loss": 0.4177,
+      "step": 812
+    },
+    {
+      "epoch": 0.9806996381182147,
+      "grad_norm": 0.3580235511736533,
+      "learning_rate": 3.739946380697052e-05,
+      "loss": 0.419,
+      "step": 813
+    },
+    {
+      "epoch": 0.9819059107358263,
+      "grad_norm": 0.5434773850238956,
+      "learning_rate": 3.737712243074174e-05,
+      "loss": 0.4046,
+      "step": 814
+    },
+    {
+      "epoch": 0.9831121833534379,
+      "grad_norm": 0.42850270847902283,
+      "learning_rate": 3.735478105451296e-05,
+      "loss": 0.4463,
+      "step": 815
+    },
+    {
+      "epoch": 0.9843184559710495,
+      "grad_norm": 0.47719801180992827,
+      "learning_rate": 3.7332439678284186e-05,
+      "loss": 0.4071,
+      "step": 816
+    },
+    {
+      "epoch": 0.985524728588661,
+      "grad_norm": 0.5226031258479484,
+      "learning_rate": 3.7310098302055407e-05,
+      "loss": 0.439,
+      "step": 817
+    },
+    {
+      "epoch": 0.9867310012062727,
+      "grad_norm": 0.431981443960303,
+      "learning_rate": 3.7287756925826634e-05,
+      "loss": 0.4213,
+      "step": 818
+    },
+    {
+      "epoch": 0.9879372738238842,
+      "grad_norm": 0.494267544185392,
+      "learning_rate": 3.726541554959786e-05,
+      "loss": 0.4159,
+      "step": 819
+    },
+    {
+      "epoch": 0.9891435464414958,
+      "grad_norm": 0.4521890760596943,
+      "learning_rate": 3.724307417336908e-05,
+      "loss": 0.4219,
+      "step": 820
+    },
+    {
+      "epoch": 0.9903498190591074,
+      "grad_norm": 0.41115842360041144,
+      "learning_rate": 3.722073279714031e-05,
+      "loss": 0.4344,
+      "step": 821
+    },
+    {
+      "epoch": 0.991556091676719,
+      "grad_norm": 0.44776288282241156,
+      "learning_rate": 3.719839142091153e-05,
+      "loss": 0.4192,
+      "step": 822
+    },
+    {
+      "epoch": 0.9927623642943305,
+      "grad_norm": 0.3978722176491912,
+      "learning_rate": 3.717605004468275e-05,
+      "loss": 0.413,
+      "step": 823
+    },
+    {
+      "epoch": 0.9939686369119421,
+      "grad_norm": 0.5012546257517893,
+      "learning_rate": 3.715370866845398e-05,
+      "loss": 0.4323,
+      "step": 824
+    },
+    {
+      "epoch": 0.9951749095295537,
+      "grad_norm": 0.43828941736087085,
+      "learning_rate": 3.7131367292225206e-05,
+      "loss": 0.4219,
+      "step": 825
+    },
+    {
+      "epoch": 0.9963811821471653,
+      "grad_norm": 0.432905255507051,
+      "learning_rate": 3.710902591599643e-05,
+      "loss": 0.4263,
+      "step": 826
+    },
+    {
+      "epoch": 0.9975874547647768,
+      "grad_norm": 0.3960220006036349,
+      "learning_rate": 3.7086684539767654e-05,
+      "loss": 0.4311,
+      "step": 827
+    },
+    {
+      "epoch": 0.9987937273823885,
+      "grad_norm": 0.3711890664614277,
+      "learning_rate": 3.7064343163538875e-05,
+      "loss": 0.4081,
+      "step": 828
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.31977027269476055,
+      "learning_rate": 3.7042001787310096e-05,
+      "loss": 0.4128,
+      "step": 829
+    },
+    {
+      "epoch": 1.0012062726176116,
+      "grad_norm": 0.405378808755483,
+      "learning_rate": 3.701966041108132e-05,
+      "loss": 0.3679,
+      "step": 830
+    },
+    {
+      "epoch": 1.002412545235223,
+      "grad_norm": 0.316147075545926,
+      "learning_rate": 3.699731903485255e-05,
+      "loss": 0.3622,
+      "step": 831
+    },
+    {
+      "epoch": 1.0036188178528347,
+      "grad_norm": 0.33536369086770434,
+      "learning_rate": 3.697497765862377e-05,
+      "loss": 0.3609,
+      "step": 832
+    },
+    {
+      "epoch": 1.0048250904704463,
+      "grad_norm": 0.3543589357049451,
+      "learning_rate": 3.6952636282395e-05,
+      "loss": 0.3658,
+      "step": 833
+    },
+    {
+      "epoch": 1.006031363088058,
+      "grad_norm": 0.35093876986866807,
+      "learning_rate": 3.6930294906166226e-05,
+      "loss": 0.3741,
+      "step": 834
+    },
+    {
+      "epoch": 1.0072376357056694,
+      "grad_norm": 0.40714167335636686,
+      "learning_rate": 3.690795352993745e-05,
+      "loss": 0.3576,
+      "step": 835
+    },
+    {
+      "epoch": 1.008443908323281,
+      "grad_norm": 0.36075988738168036,
+      "learning_rate": 3.688561215370867e-05,
+      "loss": 0.3676,
+      "step": 836
+    },
+    {
+      "epoch": 1.0096501809408926,
+      "grad_norm": 0.45662381259214857,
+      "learning_rate": 3.6863270777479895e-05,
+      "loss": 0.367,
+      "step": 837
+    },
+    {
+      "epoch": 1.0108564535585043,
+      "grad_norm": 0.3389555200863269,
+      "learning_rate": 3.6840929401251116e-05,
+      "loss": 0.3571,
+      "step": 838
+    },
+    {
+      "epoch": 1.0120627261761157,
+      "grad_norm": 0.41115199270011005,
+      "learning_rate": 3.681858802502234e-05,
+      "loss": 0.3715,
+      "step": 839
+    },
+    {
+      "epoch": 1.0132689987937273,
+      "grad_norm": 0.3689682514541236,
+      "learning_rate": 3.679624664879357e-05,
+      "loss": 0.3679,
+      "step": 840
+    },
+    {
+      "epoch": 1.014475271411339,
+      "grad_norm": 0.4023860133335284,
+      "learning_rate": 3.677390527256479e-05,
+      "loss": 0.3477,
+      "step": 841
+    },
+    {
+      "epoch": 1.0156815440289506,
+      "grad_norm": 0.322524044360211,
+      "learning_rate": 3.675156389633602e-05,
+      "loss": 0.3385,
+      "step": 842
+    },
+    {
+      "epoch": 1.016887816646562,
+      "grad_norm": 0.45344836998974947,
+      "learning_rate": 3.672922252010724e-05,
+      "loss": 0.3761,
+      "step": 843
+    },
+    {
+      "epoch": 1.0180940892641737,
+      "grad_norm": 0.34647812053404026,
+      "learning_rate": 3.670688114387846e-05,
+      "loss": 0.3493,
+      "step": 844
+    },
+    {
+      "epoch": 1.0193003618817853,
+      "grad_norm": 0.38238665236951674,
+      "learning_rate": 3.668453976764969e-05,
+      "loss": 0.3842,
+      "step": 845
+    },
+    {
+      "epoch": 1.020506634499397,
+      "grad_norm": 0.39750349528942325,
+      "learning_rate": 3.6662198391420915e-05,
+      "loss": 0.3548,
+      "step": 846
+    },
+    {
+      "epoch": 1.0217129071170084,
+      "grad_norm": 0.352018046564739,
+      "learning_rate": 3.6639857015192136e-05,
+      "loss": 0.3615,
+      "step": 847
+    },
+    {
+      "epoch": 1.02291917973462,
+      "grad_norm": 0.399580936497818,
+      "learning_rate": 3.6617515638963363e-05,
+      "loss": 0.362,
+      "step": 848
+    },
+    {
+      "epoch": 1.0241254523522316,
+      "grad_norm": 0.33323256118835304,
+      "learning_rate": 3.659517426273459e-05,
+      "loss": 0.3675,
+      "step": 849
+    },
+    {
+      "epoch": 1.0253317249698433,
+      "grad_norm": 0.39773698978442645,
+      "learning_rate": 3.657283288650581e-05,
+      "loss": 0.3561,
+      "step": 850
+    },
+    {
+      "epoch": 1.0265379975874547,
+      "grad_norm": 0.43166150787321717,
+      "learning_rate": 3.655049151027703e-05,
+      "loss": 0.3526,
+      "step": 851
+    },
+    {
+      "epoch": 1.0277442702050663,
+      "grad_norm": 0.2904670204346489,
+      "learning_rate": 3.652815013404826e-05,
+      "loss": 0.3703,
+      "step": 852
+    },
+    {
+      "epoch": 1.028950542822678,
+      "grad_norm": 0.36983080433044474,
+      "learning_rate": 3.650580875781948e-05,
+      "loss": 0.3624,
+      "step": 853
+    },
+    {
+      "epoch": 1.0301568154402896,
+      "grad_norm": 0.34943281477018856,
+      "learning_rate": 3.648346738159071e-05,
+      "loss": 0.3705,
+      "step": 854
+    },
+    {
+      "epoch": 1.031363088057901,
+      "grad_norm": 0.3090520555256778,
+      "learning_rate": 3.6461126005361935e-05,
+      "loss": 0.3849,
+      "step": 855
+    },
+    {
+      "epoch": 1.0325693606755126,
+      "grad_norm": 0.3792763380912474,
+      "learning_rate": 3.6438784629133156e-05,
+      "loss": 0.343,
+      "step": 856
+    },
+    {
+      "epoch": 1.0337756332931243,
+      "grad_norm": 0.3358613723632487,
+      "learning_rate": 3.6416443252904384e-05,
+      "loss": 0.342,
+      "step": 857
+    },
+    {
+      "epoch": 1.034981905910736,
+      "grad_norm": 0.31597184846375864,
+      "learning_rate": 3.6394101876675604e-05,
+      "loss": 0.3928,
+      "step": 858
+    },
+    {
+      "epoch": 1.0361881785283473,
+      "grad_norm": 0.39698670274757913,
+      "learning_rate": 3.6371760500446825e-05,
+      "loss": 0.3701,
+      "step": 859
+    },
+    {
+      "epoch": 1.037394451145959,
+      "grad_norm": 0.3193347171316655,
+      "learning_rate": 3.634941912421805e-05,
+      "loss": 0.3583,
+      "step": 860
+    },
+    {
+      "epoch": 1.0386007237635706,
+      "grad_norm": 0.32500721830435475,
+      "learning_rate": 3.632707774798928e-05,
+      "loss": 0.3774,
+      "step": 861
+    },
+    {
+      "epoch": 1.0398069963811822,
+      "grad_norm": 0.3097308325536179,
+      "learning_rate": 3.63047363717605e-05,
+      "loss": 0.3527,
+      "step": 862
+    },
+    {
+      "epoch": 1.0410132689987937,
+      "grad_norm": 0.30675170250681827,
+      "learning_rate": 3.628239499553173e-05,
+      "loss": 0.3693,
+      "step": 863
+    },
+    {
+      "epoch": 1.0422195416164053,
+      "grad_norm": 0.2940557941988533,
+      "learning_rate": 3.6260053619302956e-05,
+      "loss": 0.3662,
+      "step": 864
+    },
+    {
+      "epoch": 1.043425814234017,
+      "grad_norm": 0.32259841695392805,
+      "learning_rate": 3.6237712243074176e-05,
+      "loss": 0.3791,
+      "step": 865
+    },
+    {
+      "epoch": 1.0446320868516286,
+      "grad_norm": 0.31457425401619904,
+      "learning_rate": 3.62153708668454e-05,
+      "loss": 0.3856,
+      "step": 866
+    },
+    {
+      "epoch": 1.04583835946924,
+      "grad_norm": 0.36463813044162835,
+      "learning_rate": 3.6193029490616625e-05,
+      "loss": 0.3528,
+      "step": 867
+    },
+    {
+      "epoch": 1.0470446320868516,
+      "grad_norm": 0.3370156737832902,
+      "learning_rate": 3.6170688114387845e-05,
+      "loss": 0.3598,
+      "step": 868
+    },
+    {
+      "epoch": 1.0482509047044632,
+      "grad_norm": 0.3761941351812865,
+      "learning_rate": 3.614834673815907e-05,
+      "loss": 0.3501,
+      "step": 869
+    },
+    {
+      "epoch": 1.0494571773220749,
+      "grad_norm": 0.3246343590333649,
+      "learning_rate": 3.61260053619303e-05,
+      "loss": 0.3817,
+      "step": 870
+    },
+    {
+      "epoch": 1.0506634499396863,
+      "grad_norm": 0.40612621160744156,
+      "learning_rate": 3.610366398570152e-05,
+      "loss": 0.3669,
+      "step": 871
+    },
+    {
+      "epoch": 1.051869722557298,
+      "grad_norm": 0.39855243669619894,
+      "learning_rate": 3.608132260947275e-05,
+      "loss": 0.3649,
+      "step": 872
+    },
+    {
+      "epoch": 1.0530759951749096,
+      "grad_norm": 0.32860941997340953,
+      "learning_rate": 3.605898123324397e-05,
+      "loss": 0.3615,
+      "step": 873
+    },
+    {
+      "epoch": 1.0542822677925212,
+      "grad_norm": 0.4135141126588348,
+      "learning_rate": 3.603663985701519e-05,
+      "loss": 0.3576,
+      "step": 874
+    },
+    {
+      "epoch": 1.0554885404101326,
+      "grad_norm": 0.3238167077263733,
+      "learning_rate": 3.601429848078642e-05,
+      "loss": 0.3528,
+      "step": 875
+    },
+    {
+      "epoch": 1.0566948130277443,
+      "grad_norm": 0.3660474075113215,
+      "learning_rate": 3.5991957104557645e-05,
+      "loss": 0.3613,
+      "step": 876
+    },
+    {
+      "epoch": 1.057901085645356,
+      "grad_norm": 0.33674061414084067,
+      "learning_rate": 3.5969615728328865e-05,
+      "loss": 0.3585,
+      "step": 877
+    },
+    {
+      "epoch": 1.0591073582629675,
+      "grad_norm": 0.3246101881612825,
+      "learning_rate": 3.594727435210009e-05,
+      "loss": 0.3692,
+      "step": 878
+    },
+    {
+      "epoch": 1.060313630880579,
+      "grad_norm": 0.3411005351843354,
+      "learning_rate": 3.592493297587132e-05,
+      "loss": 0.3615,
+      "step": 879
+    },
+    {
+      "epoch": 1.0615199034981906,
+      "grad_norm": 0.3108590519815204,
+      "learning_rate": 3.590259159964254e-05,
+      "loss": 0.3594,
+      "step": 880
+    },
+    {
+      "epoch": 1.0627261761158022,
+      "grad_norm": 0.3434174584609721,
+      "learning_rate": 3.588025022341376e-05,
+      "loss": 0.3568,
+      "step": 881
+    },
+    {
+      "epoch": 1.0639324487334139,
+      "grad_norm": 0.27495195172330483,
+      "learning_rate": 3.585790884718499e-05,
+      "loss": 0.3621,
+      "step": 882
+    },
+    {
+      "epoch": 1.0651387213510253,
+      "grad_norm": 0.3604762512563549,
+      "learning_rate": 3.583556747095621e-05,
+      "loss": 0.3718,
+      "step": 883
+    },
+    {
+      "epoch": 1.066344993968637,
+      "grad_norm": 0.3218209404107597,
+      "learning_rate": 3.581322609472744e-05,
+      "loss": 0.3493,
+      "step": 884
+    },
+    {
+      "epoch": 1.0675512665862485,
+      "grad_norm": 0.320576472870977,
+      "learning_rate": 3.5790884718498665e-05,
+      "loss": 0.3683,
+      "step": 885
+    },
+    {
+      "epoch": 1.06875753920386,
+      "grad_norm": 0.3073606422293175,
+      "learning_rate": 3.5768543342269886e-05,
+      "loss": 0.3788,
+      "step": 886
+    },
+    {
+      "epoch": 1.0699638118214716,
+      "grad_norm": 0.36492025212381535,
+      "learning_rate": 3.5746201966041106e-05,
+      "loss": 0.348,
+      "step": 887
+    },
+    {
+      "epoch": 1.0711700844390832,
+      "grad_norm": 0.3154824646598146,
+      "learning_rate": 3.5723860589812334e-05,
+      "loss": 0.3863,
+      "step": 888
+    },
+    {
+      "epoch": 1.0723763570566949,
+      "grad_norm": 0.38244729654560733,
+      "learning_rate": 3.5701519213583554e-05,
+      "loss": 0.3634,
+      "step": 889
+    },
+    {
+      "epoch": 1.0735826296743065,
+      "grad_norm": 0.29567778156519975,
+      "learning_rate": 3.567917783735478e-05,
+      "loss": 0.3703,
+      "step": 890
+    },
+    {
+      "epoch": 1.074788902291918,
+      "grad_norm": 0.8554106402979569,
+      "learning_rate": 3.565683646112601e-05,
+      "loss": 0.3794,
+      "step": 891
+    },
+    {
+      "epoch": 1.0759951749095296,
+      "grad_norm": 0.3030975645200047,
+      "learning_rate": 3.563449508489723e-05,
+      "loss": 0.3605,
+      "step": 892
+    },
+    {
+      "epoch": 1.0772014475271412,
+      "grad_norm": 0.2832483178708839,
+      "learning_rate": 3.561215370866846e-05,
+      "loss": 0.3473,
+      "step": 893
+    },
+    {
+      "epoch": 1.0784077201447526,
+      "grad_norm": 0.3279420805673933,
+      "learning_rate": 3.558981233243968e-05,
+      "loss": 0.3616,
+      "step": 894
+    },
+    {
+      "epoch": 1.0796139927623642,
+      "grad_norm": 0.3173532398101509,
+      "learning_rate": 3.55674709562109e-05,
+      "loss": 0.3578,
+      "step": 895
+    },
+    {
+      "epoch": 1.0808202653799759,
+      "grad_norm": 0.3225132430413757,
+      "learning_rate": 3.5545129579982126e-05,
+      "loss": 0.3717,
+      "step": 896
+    },
+    {
+      "epoch": 1.0820265379975875,
+      "grad_norm": 0.30488512630939474,
+      "learning_rate": 3.5522788203753354e-05,
+      "loss": 0.3559,
+      "step": 897
+    },
+    {
+      "epoch": 1.083232810615199,
+      "grad_norm": 0.3737259775563492,
+      "learning_rate": 3.5500446827524575e-05,
+      "loss": 0.3659,
+      "step": 898
+    },
+    {
+      "epoch": 1.0844390832328106,
+      "grad_norm": 0.29431467422973967,
+      "learning_rate": 3.54781054512958e-05,
+      "loss": 0.3626,
+      "step": 899
+    },
+    {
+      "epoch": 1.0856453558504222,
+      "grad_norm": 0.3056774050205318,
+      "learning_rate": 3.545576407506703e-05,
+      "loss": 0.3435,
+      "step": 900
+    },
+    {
+      "epoch": 1.0868516284680338,
+      "grad_norm": 0.2800691429045292,
+      "learning_rate": 3.543342269883825e-05,
+      "loss": 0.3554,
+      "step": 901
+    },
+    {
+      "epoch": 1.0880579010856453,
+      "grad_norm": 0.33666126556125986,
+      "learning_rate": 3.541108132260947e-05,
+      "loss": 0.3893,
+      "step": 902
+    },
+    {
+      "epoch": 1.089264173703257,
+      "grad_norm": 0.31181311277504536,
+      "learning_rate": 3.53887399463807e-05,
+      "loss": 0.3858,
+      "step": 903
+    },
+    {
+      "epoch": 1.0904704463208685,
+      "grad_norm": 0.33517625298580916,
+      "learning_rate": 3.536639857015192e-05,
+      "loss": 0.3512,
+      "step": 904
+    },
+    {
+      "epoch": 1.0916767189384802,
+      "grad_norm": 0.3426706216867623,
+      "learning_rate": 3.534405719392315e-05,
+      "loss": 0.3789,
+      "step": 905
+    },
+    {
+      "epoch": 1.0928829915560916,
+      "grad_norm": 0.3566325293955928,
+      "learning_rate": 3.5321715817694374e-05,
+      "loss": 0.3731,
+      "step": 906
+    },
+    {
+      "epoch": 1.0940892641737032,
+      "grad_norm": 0.3237953484166762,
+      "learning_rate": 3.5299374441465595e-05,
+      "loss": 0.3529,
+      "step": 907
+    },
+    {
+      "epoch": 1.0952955367913149,
+      "grad_norm": 0.4230160362167563,
+      "learning_rate": 3.527703306523682e-05,
+      "loss": 0.3637,
+      "step": 908
+    },
+    {
+      "epoch": 1.0965018094089265,
+      "grad_norm": 0.3287139538164379,
+      "learning_rate": 3.525469168900804e-05,
+      "loss": 0.3636,
+      "step": 909
+    },
+    {
+      "epoch": 1.097708082026538,
+      "grad_norm": 0.38658111001916495,
+      "learning_rate": 3.5232350312779264e-05,
+      "loss": 0.3742,
+      "step": 910
+    },
+    {
+      "epoch": 1.0989143546441495,
+      "grad_norm": 0.30256158696471586,
+      "learning_rate": 3.521000893655049e-05,
+      "loss": 0.3767,
+      "step": 911
+    },
+    {
+      "epoch": 1.1001206272617612,
+      "grad_norm": 0.4143128405160983,
+      "learning_rate": 3.518766756032172e-05,
+      "loss": 0.3865,
+      "step": 912
+    },
+    {
+      "epoch": 1.1013268998793728,
+      "grad_norm": 0.3179444203982378,
+      "learning_rate": 3.516532618409294e-05,
+      "loss": 0.3675,
+      "step": 913
+    },
+    {
+      "epoch": 1.1025331724969842,
+      "grad_norm": 0.36029708813831,
+      "learning_rate": 3.514298480786417e-05,
+      "loss": 0.3677,
+      "step": 914
+    },
+    {
+      "epoch": 1.1037394451145959,
+      "grad_norm": 0.29429182950704297,
+      "learning_rate": 3.5120643431635394e-05,
+      "loss": 0.3487,
+      "step": 915
+    },
+    {
+      "epoch": 1.1049457177322075,
+      "grad_norm": 0.37996208960282496,
+      "learning_rate": 3.5098302055406615e-05,
+      "loss": 0.3811,
+      "step": 916
+    },
+    {
+      "epoch": 1.1061519903498191,
+      "grad_norm": 0.7625080539920996,
+      "learning_rate": 3.5075960679177836e-05,
+      "loss": 0.4344,
+      "step": 917
+    },
+    {
+      "epoch": 1.1073582629674306,
+      "grad_norm": 0.3233237284317067,
+      "learning_rate": 3.505361930294906e-05,
+      "loss": 0.3478,
+      "step": 918
+    },
+    {
+      "epoch": 1.1085645355850422,
+      "grad_norm": 0.32634118122886424,
+      "learning_rate": 3.5031277926720284e-05,
+      "loss": 0.3608,
+      "step": 919
+    },
+    {
+      "epoch": 1.1097708082026538,
+      "grad_norm": 0.34423353361517317,
+      "learning_rate": 3.500893655049151e-05,
+      "loss": 0.3631,
+      "step": 920
+    },
+    {
+      "epoch": 1.1109770808202655,
+      "grad_norm": 0.3562353951470559,
+      "learning_rate": 3.498659517426274e-05,
+      "loss": 0.3673,
+      "step": 921
+    },
+    {
+      "epoch": 1.1121833534378769,
+      "grad_norm": 0.3702325265210759,
+      "learning_rate": 3.496425379803396e-05,
+      "loss": 0.3749,
+      "step": 922
+    },
+    {
+      "epoch": 1.1133896260554885,
+      "grad_norm": 0.4029117460745025,
+      "learning_rate": 3.494191242180519e-05,
+      "loss": 0.3601,
+      "step": 923
+    },
+    {
+      "epoch": 1.1145958986731002,
+      "grad_norm": 0.36246580149698215,
+      "learning_rate": 3.491957104557641e-05,
+      "loss": 0.3629,
+      "step": 924
+    },
+    {
+      "epoch": 1.1158021712907118,
+      "grad_norm": 0.38132206897614723,
+      "learning_rate": 3.489722966934763e-05,
+      "loss": 0.3814,
+      "step": 925
+    },
+    {
+      "epoch": 1.1170084439083232,
+      "grad_norm": 0.3387624425697889,
+      "learning_rate": 3.4874888293118856e-05,
+      "loss": 0.3629,
+      "step": 926
+    },
+    {
+      "epoch": 1.1182147165259348,
+      "grad_norm": 0.3678799590845831,
+      "learning_rate": 3.485254691689008e-05,
+      "loss": 0.3667,
+      "step": 927
+    },
+    {
+      "epoch": 1.1194209891435465,
+      "grad_norm": 0.3868471391215961,
+      "learning_rate": 3.4830205540661304e-05,
+      "loss": 0.3531,
+      "step": 928
+    },
+    {
+      "epoch": 1.1206272617611581,
+      "grad_norm": 0.35802571950959106,
+      "learning_rate": 3.480786416443253e-05,
+      "loss": 0.3728,
+      "step": 929
+    },
+    {
+      "epoch": 1.1218335343787695,
+      "grad_norm": 0.4527352418804462,
+      "learning_rate": 3.478552278820376e-05,
+      "loss": 0.3519,
+      "step": 930
+    },
+    {
+      "epoch": 1.1230398069963812,
+      "grad_norm": 0.38538894689912423,
+      "learning_rate": 3.476318141197498e-05,
+      "loss": 0.3654,
+      "step": 931
+    },
+    {
+      "epoch": 1.1242460796139928,
+      "grad_norm": 0.32599224994044007,
+      "learning_rate": 3.47408400357462e-05,
+      "loss": 0.3643,
+      "step": 932
+    },
+    {
+      "epoch": 1.1254523522316044,
+      "grad_norm": 0.4953992771024001,
+      "learning_rate": 3.471849865951743e-05,
+      "loss": 0.3627,
+      "step": 933
+    },
+    {
+      "epoch": 1.1266586248492159,
+      "grad_norm": 0.3186598726280072,
+      "learning_rate": 3.4696157283288655e-05,
+      "loss": 0.3565,
+      "step": 934
+    },
+    {
+      "epoch": 1.1278648974668275,
+      "grad_norm": 0.4038468704612751,
+      "learning_rate": 3.4673815907059876e-05,
+      "loss": 0.3623,
+      "step": 935
+    },
+    {
+      "epoch": 1.1290711700844391,
+      "grad_norm": 0.3481809289328933,
+      "learning_rate": 3.4651474530831104e-05,
+      "loss": 0.3954,
+      "step": 936
+    },
+    {
+      "epoch": 1.1302774427020505,
+      "grad_norm": 0.3683916961882533,
+      "learning_rate": 3.462913315460233e-05,
+      "loss": 0.3349,
+      "step": 937
+    },
+    {
+      "epoch": 1.1314837153196622,
+      "grad_norm": 0.32473875080902603,
+      "learning_rate": 3.460679177837355e-05,
+      "loss": 0.3743,
+      "step": 938
+    },
+    {
+      "epoch": 1.1326899879372738,
+      "grad_norm": 0.3928948622915465,
+      "learning_rate": 3.458445040214477e-05,
+      "loss": 0.3568,
+      "step": 939
+    },
+    {
+      "epoch": 1.1338962605548855,
+      "grad_norm": 0.34316792512475985,
+      "learning_rate": 3.4562109025916e-05,
+      "loss": 0.3595,
+      "step": 940
+    },
+    {
+      "epoch": 1.135102533172497,
+      "grad_norm": 0.38074589547250975,
+      "learning_rate": 3.453976764968722e-05,
+      "loss": 0.3717,
+      "step": 941
+    },
+    {
+      "epoch": 1.1363088057901085,
+      "grad_norm": 0.3381612708943419,
+      "learning_rate": 3.451742627345845e-05,
+      "loss": 0.3628,
+      "step": 942
+    },
+    {
+      "epoch": 1.1375150784077201,
+      "grad_norm": 0.3918788189329609,
+      "learning_rate": 3.4495084897229676e-05,
+      "loss": 0.3402,
+      "step": 943
+    },
+    {
+      "epoch": 1.1387213510253318,
+      "grad_norm": 0.3657275116322395,
+      "learning_rate": 3.4472743521000896e-05,
+      "loss": 0.349,
+      "step": 944
+    },
+    {
+      "epoch": 1.1399276236429432,
+      "grad_norm": 0.4007355044780056,
+      "learning_rate": 3.4450402144772124e-05,
+      "loss": 0.3688,
+      "step": 945
+    },
+    {
+      "epoch": 1.1411338962605548,
+      "grad_norm": 0.4138120642712477,
+      "learning_rate": 3.4428060768543344e-05,
+      "loss": 0.3738,
+      "step": 946
+    },
+    {
+      "epoch": 1.1423401688781665,
+      "grad_norm": 0.4289328018683492,
+      "learning_rate": 3.4405719392314565e-05,
+      "loss": 0.3658,
+      "step": 947
+    },
+    {
+      "epoch": 1.143546441495778,
+      "grad_norm": 0.34457376242795956,
+      "learning_rate": 3.438337801608579e-05,
+      "loss": 0.3648,
+      "step": 948
+    },
+    {
+      "epoch": 1.1447527141133897,
+      "grad_norm": 0.39016387340771286,
+      "learning_rate": 3.436103663985702e-05,
+      "loss": 0.3851,
+      "step": 949
+    },
+    {
+      "epoch": 1.1459589867310012,
+      "grad_norm": 0.3794539802442255,
+      "learning_rate": 3.433869526362824e-05,
+      "loss": 0.3703,
+      "step": 950
+    },
+    {
+      "epoch": 1.1471652593486128,
+      "grad_norm": 0.3400125800009884,
+      "learning_rate": 3.431635388739947e-05,
+      "loss": 0.3651,
+      "step": 951
+    },
+    {
+      "epoch": 1.1483715319662244,
+      "grad_norm": 0.3962308094664021,
+      "learning_rate": 3.429401251117069e-05,
+      "loss": 0.3711,
+      "step": 952
+    },
+    {
+      "epoch": 1.1495778045838358,
+      "grad_norm": 0.29475072361506954,
+      "learning_rate": 3.427167113494191e-05,
+      "loss": 0.3649,
+      "step": 953
+    },
+    {
+      "epoch": 1.1507840772014475,
+      "grad_norm": 0.3599623229725101,
+      "learning_rate": 3.424932975871314e-05,
+      "loss": 0.3755,
+      "step": 954
+    },
+    {
+      "epoch": 1.1519903498190591,
+      "grad_norm": 0.3234762124366387,
+      "learning_rate": 3.4226988382484365e-05,
+      "loss": 0.3419,
+      "step": 955
+    },
+    {
+      "epoch": 1.1531966224366708,
+      "grad_norm": 0.36396822545574137,
+      "learning_rate": 3.4204647006255585e-05,
+      "loss": 0.3589,
+      "step": 956
+    },
+    {
+      "epoch": 1.1544028950542822,
+      "grad_norm": 0.2965621293534876,
+      "learning_rate": 3.418230563002681e-05,
+      "loss": 0.3408,
+      "step": 957
+    },
+    {
+      "epoch": 1.1556091676718938,
+      "grad_norm": 0.40185597700007947,
+      "learning_rate": 3.415996425379804e-05,
+      "loss": 0.3767,
+      "step": 958
+    },
+    {
+      "epoch": 1.1568154402895054,
+      "grad_norm": 0.2814563352081147,
+      "learning_rate": 3.413762287756926e-05,
+      "loss": 0.3489,
+      "step": 959
+    },
+    {
+      "epoch": 1.158021712907117,
+      "grad_norm": 0.39116431668529056,
+      "learning_rate": 3.411528150134048e-05,
+      "loss": 0.3447,
+      "step": 960
+    },
+    {
+      "epoch": 1.1592279855247285,
+      "grad_norm": 0.37682364888520664,
+      "learning_rate": 3.409294012511171e-05,
+      "loss": 0.3699,
+      "step": 961
+    },
+    {
+      "epoch": 1.1604342581423401,
+      "grad_norm": 0.38866135079779807,
+      "learning_rate": 3.407059874888293e-05,
+      "loss": 0.379,
+      "step": 962
+    },
+    {
+      "epoch": 1.1616405307599518,
+      "grad_norm": 0.43349224212309945,
+      "learning_rate": 3.404825737265416e-05,
+      "loss": 0.3771,
+      "step": 963
+    },
+    {
+      "epoch": 1.1628468033775634,
+      "grad_norm": 0.3490798723729157,
+      "learning_rate": 3.4025915996425385e-05,
+      "loss": 0.366,
+      "step": 964
+    },
+    {
+      "epoch": 1.1640530759951748,
+      "grad_norm": 0.4291120472338438,
+      "learning_rate": 3.4003574620196606e-05,
+      "loss": 0.3735,
+      "step": 965
+    },
+    {
+      "epoch": 1.1652593486127865,
+      "grad_norm": 0.475166138749199,
+      "learning_rate": 3.398123324396783e-05,
+      "loss": 0.3856,
+      "step": 966
+    },
+    {
+      "epoch": 1.166465621230398,
+      "grad_norm": 0.44115919424080835,
+      "learning_rate": 3.3958891867739054e-05,
+      "loss": 0.3874,
+      "step": 967
+    },
+    {
+      "epoch": 1.1676718938480097,
+      "grad_norm": 0.4265603005387793,
+      "learning_rate": 3.3936550491510274e-05,
+      "loss": 0.3662,
+      "step": 968
+    },
+    {
+      "epoch": 1.1688781664656211,
+      "grad_norm": 0.3541937956792882,
+      "learning_rate": 3.39142091152815e-05,
+      "loss": 0.3678,
+      "step": 969
+    },
+    {
+      "epoch": 1.1700844390832328,
+      "grad_norm": 0.3983932739056756,
+      "learning_rate": 3.389186773905273e-05,
+      "loss": 0.3782,
+      "step": 970
+    },
+    {
+      "epoch": 1.1712907117008444,
+      "grad_norm": 0.2708040120469766,
+      "learning_rate": 3.386952636282395e-05,
+      "loss": 0.3508,
+      "step": 971
+    },
+    {
+      "epoch": 1.172496984318456,
+      "grad_norm": 0.38814511090940623,
+      "learning_rate": 3.384718498659518e-05,
+      "loss": 0.3674,
+      "step": 972
+    },
+    {
+      "epoch": 1.1737032569360675,
+      "grad_norm": 0.31952981884106635,
+      "learning_rate": 3.3824843610366405e-05,
+      "loss": 0.375,
+      "step": 973
+    },
+    {
+      "epoch": 1.174909529553679,
+      "grad_norm": 0.3405372128693083,
+      "learning_rate": 3.3802502234137626e-05,
+      "loss": 0.358,
+      "step": 974
+    },
+    {
+      "epoch": 1.1761158021712907,
+      "grad_norm": 0.325059651004217,
+      "learning_rate": 3.3780160857908846e-05,
+      "loss": 0.3643,
+      "step": 975
+    },
+    {
+      "epoch": 1.1773220747889024,
+      "grad_norm": 0.3125452451270937,
+      "learning_rate": 3.3757819481680074e-05,
+      "loss": 0.3586,
+      "step": 976
+    },
+    {
+      "epoch": 1.1785283474065138,
+      "grad_norm": 0.3352607281639852,
+      "learning_rate": 3.3735478105451295e-05,
+      "loss": 0.3688,
+      "step": 977
+    },
+    {
+      "epoch": 1.1797346200241254,
+      "grad_norm": 2.976439201363688,
+      "learning_rate": 3.371313672922252e-05,
+      "loss": 0.3982,
+      "step": 978
+    },
+    {
+      "epoch": 1.180940892641737,
+      "grad_norm": 0.4938721359348584,
+      "learning_rate": 3.369079535299375e-05,
+      "loss": 0.3763,
+      "step": 979
+    },
+    {
+      "epoch": 1.1821471652593487,
+      "grad_norm": 0.2726220782356802,
+      "learning_rate": 3.366845397676497e-05,
+      "loss": 0.3496,
+      "step": 980
+    },
+    {
+      "epoch": 1.1833534378769601,
+      "grad_norm": 0.3666194013950933,
+      "learning_rate": 3.36461126005362e-05,
+      "loss": 0.3567,
+      "step": 981
+    },
+    {
+      "epoch": 1.1845597104945718,
+      "grad_norm": 0.3989427724814577,
+      "learning_rate": 3.362377122430742e-05,
+      "loss": 0.3914,
+      "step": 982
+    },
+    {
+      "epoch": 1.1857659831121834,
+      "grad_norm": 0.35488518310984535,
+      "learning_rate": 3.360142984807864e-05,
+      "loss": 0.3816,
+      "step": 983
+    },
+    {
+      "epoch": 1.186972255729795,
+      "grad_norm": 0.4276280862353782,
+      "learning_rate": 3.3579088471849867e-05,
+      "loss": 0.3814,
+      "step": 984
+    },
+    {
+      "epoch": 1.1881785283474064,
+      "grad_norm": 0.3927987855722157,
+      "learning_rate": 3.3556747095621094e-05,
+      "loss": 0.3523,
+      "step": 985
+    },
+    {
+      "epoch": 1.189384800965018,
+      "grad_norm": 0.38737768232211683,
+      "learning_rate": 3.3534405719392315e-05,
+      "loss": 0.3823,
+      "step": 986
+    },
+    {
+      "epoch": 1.1905910735826297,
+      "grad_norm": 0.38347649772291015,
+      "learning_rate": 3.351206434316354e-05,
+      "loss": 0.3536,
+      "step": 987
+    },
+    {
+      "epoch": 1.1917973462002411,
+      "grad_norm": 0.30177462805239674,
+      "learning_rate": 3.348972296693477e-05,
+      "loss": 0.3736,
+      "step": 988
+    },
+    {
+      "epoch": 1.1930036188178528,
+      "grad_norm": 0.3551527270864664,
+      "learning_rate": 3.346738159070599e-05,
+      "loss": 0.3507,
+      "step": 989
+    },
+    {
+      "epoch": 1.1942098914354644,
+      "grad_norm": 0.3486555270352525,
+      "learning_rate": 3.344504021447721e-05,
+      "loss": 0.3971,
+      "step": 990
+    },
+    {
+      "epoch": 1.195416164053076,
+      "grad_norm": 0.3435059469091779,
+      "learning_rate": 3.342269883824844e-05,
+      "loss": 0.3561,
+      "step": 991
+    },
+    {
+      "epoch": 1.1966224366706877,
+      "grad_norm": 0.4189817765235362,
+      "learning_rate": 3.340035746201966e-05,
+      "loss": 0.3591,
+      "step": 992
+    },
+    {
+      "epoch": 1.197828709288299,
+      "grad_norm": 0.3122862791991749,
+      "learning_rate": 3.337801608579089e-05,
+      "loss": 0.3633,
+      "step": 993
+    },
+    {
+      "epoch": 1.1990349819059107,
+      "grad_norm": 0.34538308973829507,
+      "learning_rate": 3.3355674709562114e-05,
+      "loss": 0.3554,
+      "step": 994
+    },
+    {
+      "epoch": 1.2002412545235224,
+      "grad_norm": 0.3936966308431531,
+      "learning_rate": 3.3333333333333335e-05,
+      "loss": 0.3685,
+      "step": 995
+    },
+    {
+      "epoch": 1.2014475271411338,
+      "grad_norm": 0.2939043985854137,
+      "learning_rate": 3.331099195710456e-05,
+      "loss": 0.3707,
+      "step": 996
+    },
+    {
+      "epoch": 1.2026537997587454,
+      "grad_norm": 0.36151070128805723,
+      "learning_rate": 3.328865058087578e-05,
+      "loss": 0.3475,
+      "step": 997
+    },
+    {
+      "epoch": 1.203860072376357,
+      "grad_norm": 0.31584191649328586,
+      "learning_rate": 3.3266309204647004e-05,
+      "loss": 0.3614,
+      "step": 998
+    },
+    {
+      "epoch": 1.2050663449939687,
+      "grad_norm": 0.3065266896699921,
+      "learning_rate": 3.324396782841823e-05,
+      "loss": 0.3536,
+      "step": 999
+    },
+    {
+      "epoch": 1.2062726176115803,
+      "grad_norm": 0.3284469711990536,
+      "learning_rate": 3.322162645218946e-05,
+      "loss": 0.3709,
+      "step": 1000
+    },
+    {
+      "epoch": 1.2074788902291917,
+      "grad_norm": 0.28866547050936975,
+      "learning_rate": 3.319928507596068e-05,
+      "loss": 0.3685,
+      "step": 1001
+    },
+    {
+      "epoch": 1.2086851628468034,
+      "grad_norm": 0.33796420853850734,
+      "learning_rate": 3.317694369973191e-05,
+      "loss": 0.3652,
+      "step": 1002
+    },
+    {
+      "epoch": 1.209891435464415,
+      "grad_norm": 0.28342229294291926,
+      "learning_rate": 3.3154602323503134e-05,
+      "loss": 0.3337,
+      "step": 1003
+    },
+    {
+      "epoch": 1.2110977080820264,
+      "grad_norm": 0.333402960456291,
+      "learning_rate": 3.313226094727435e-05,
+      "loss": 0.3584,
+      "step": 1004
+    },
+    {
+      "epoch": 1.212303980699638,
+      "grad_norm": 0.3342504865352302,
+      "learning_rate": 3.3109919571045576e-05,
+      "loss": 0.3675,
+      "step": 1005
+    },
+    {
+      "epoch": 1.2135102533172497,
+      "grad_norm": 0.2913332665016377,
+      "learning_rate": 3.30875781948168e-05,
+      "loss": 0.3561,
+      "step": 1006
+    },
+    {
+      "epoch": 1.2147165259348613,
+      "grad_norm": 0.31253774645325566,
+      "learning_rate": 3.3065236818588024e-05,
+      "loss": 0.3625,
+      "step": 1007
+    },
+    {
+      "epoch": 1.215922798552473,
+      "grad_norm": 0.337334942984555,
+      "learning_rate": 3.304289544235925e-05,
+      "loss": 0.373,
+      "step": 1008
+    },
+    {
+      "epoch": 1.2171290711700844,
+      "grad_norm": 0.31936744219352725,
+      "learning_rate": 3.302055406613048e-05,
+      "loss": 0.3608,
+      "step": 1009
+    },
+    {
+      "epoch": 1.218335343787696,
+      "grad_norm": 0.36610894353405193,
+      "learning_rate": 3.29982126899017e-05,
+      "loss": 0.3673,
+      "step": 1010
+    },
+    {
+      "epoch": 1.2195416164053077,
+      "grad_norm": 0.3398885145604998,
+      "learning_rate": 3.297587131367292e-05,
+      "loss": 0.3524,
+      "step": 1011
+    },
+    {
+      "epoch": 1.220747889022919,
+      "grad_norm": 0.40102720999756397,
+      "learning_rate": 3.295352993744415e-05,
+      "loss": 0.3744,
+      "step": 1012
+    },
+    {
+      "epoch": 1.2219541616405307,
+      "grad_norm": 0.40168746016689283,
+      "learning_rate": 3.293118856121537e-05,
+      "loss": 0.369,
+      "step": 1013
+    },
+    {
+      "epoch": 1.2231604342581424,
+      "grad_norm": 0.3300948359331406,
+      "learning_rate": 3.2908847184986596e-05,
+      "loss": 0.3724,
+      "step": 1014
+    },
+    {
+      "epoch": 1.224366706875754,
+      "grad_norm": 0.38963810919167646,
+      "learning_rate": 3.2886505808757823e-05,
+      "loss": 0.3652,
+      "step": 1015
+    },
+    {
+      "epoch": 1.2255729794933654,
+      "grad_norm": 0.32140355256835956,
+      "learning_rate": 3.2864164432529044e-05,
+      "loss": 0.3715,
+      "step": 1016
+    },
+    {
+      "epoch": 1.226779252110977,
+      "grad_norm": 0.3483272281013881,
+      "learning_rate": 3.284182305630027e-05,
+      "loss": 0.3562,
+      "step": 1017
+    },
+    {
+      "epoch": 1.2279855247285887,
+      "grad_norm": 0.3309692180489672,
+      "learning_rate": 3.281948168007149e-05,
+      "loss": 0.3649,
+      "step": 1018
+    },
+    {
+      "epoch": 1.2291917973462003,
+      "grad_norm": 0.4077279870823795,
+      "learning_rate": 3.279714030384271e-05,
+      "loss": 0.3825,
+      "step": 1019
+    },
+    {
+      "epoch": 1.2303980699638117,
+      "grad_norm": 0.3150547142275848,
+      "learning_rate": 3.277479892761394e-05,
+      "loss": 0.3772,
+      "step": 1020
+    },
+    {
+      "epoch": 1.2316043425814234,
+      "grad_norm": 0.3798958280863762,
+      "learning_rate": 3.275245755138517e-05,
+      "loss": 0.3556,
+      "step": 1021
+    },
+    {
+      "epoch": 1.232810615199035,
+      "grad_norm": 0.31873298226111396,
+      "learning_rate": 3.273011617515639e-05,
+      "loss": 0.3635,
+      "step": 1022
+    },
+    {
+      "epoch": 1.2340168878166466,
+      "grad_norm": 0.336183080052793,
+      "learning_rate": 3.2707774798927616e-05,
+      "loss": 0.3714,
+      "step": 1023
+    },
+    {
+      "epoch": 1.235223160434258,
+      "grad_norm": 0.3799076103107832,
+      "learning_rate": 3.2685433422698844e-05,
+      "loss": 0.3736,
+      "step": 1024
+    },
+    {
+      "epoch": 1.2364294330518697,
+      "grad_norm": 0.28728541195117857,
+      "learning_rate": 3.2663092046470064e-05,
+      "loss": 0.3755,
+      "step": 1025
+    },
+    {
+      "epoch": 1.2376357056694813,
+      "grad_norm": 0.3232968103048298,
+      "learning_rate": 3.2640750670241285e-05,
+      "loss": 0.3717,
+      "step": 1026
+    },
+    {
+      "epoch": 1.238841978287093,
+      "grad_norm": 0.3301793860111746,
+      "learning_rate": 3.261840929401251e-05,
+      "loss": 0.3792,
+      "step": 1027
+    },
+    {
+      "epoch": 1.2400482509047044,
+      "grad_norm": 0.3143951692046986,
+      "learning_rate": 3.259606791778373e-05,
+      "loss": 0.3763,
+      "step": 1028
+    },
+    {
+      "epoch": 1.241254523522316,
+      "grad_norm": 0.32851664710301687,
+      "learning_rate": 3.257372654155496e-05,
+      "loss": 0.3652,
+      "step": 1029
+    },
+    {
+      "epoch": 1.2424607961399277,
+      "grad_norm": 0.3911048328835441,
+      "learning_rate": 3.255138516532619e-05,
+      "loss": 0.3616,
+      "step": 1030
+    },
+    {
+      "epoch": 1.2436670687575393,
+      "grad_norm": 0.27935714055304606,
+      "learning_rate": 3.252904378909741e-05,
+      "loss": 0.3596,
+      "step": 1031
+    },
+    {
+      "epoch": 1.2448733413751507,
+      "grad_norm": 0.3346861939949344,
+      "learning_rate": 3.2506702412868636e-05,
+      "loss": 0.3547,
+      "step": 1032
+    },
+    {
+      "epoch": 1.2460796139927623,
+      "grad_norm": 0.28091787367278925,
+      "learning_rate": 3.248436103663986e-05,
+      "loss": 0.3445,
+      "step": 1033
+    },
+    {
+      "epoch": 1.247285886610374,
+      "grad_norm": 0.29369101218699734,
+      "learning_rate": 3.246201966041108e-05,
+      "loss": 0.3597,
+      "step": 1034
+    },
+    {
+      "epoch": 1.2484921592279856,
+      "grad_norm": 0.29878313480140756,
+      "learning_rate": 3.2439678284182305e-05,
+      "loss": 0.3697,
+      "step": 1035
+    },
+    {
+      "epoch": 1.249698431845597,
+      "grad_norm": 0.32949020606504464,
+      "learning_rate": 3.241733690795353e-05,
+      "loss": 0.3685,
+      "step": 1036
+    },
+    {
+      "epoch": 1.2509047044632087,
+      "grad_norm": 0.2926475487110276,
+      "learning_rate": 3.2394995531724753e-05,
+      "loss": 0.3626,
+      "step": 1037
+    },
+    {
+      "epoch": 1.2521109770808203,
+      "grad_norm": 0.2859720388486372,
+      "learning_rate": 3.237265415549598e-05,
+      "loss": 0.3743,
+      "step": 1038
+    },
+    {
+      "epoch": 1.2533172496984317,
+      "grad_norm": 0.3402775775703186,
+      "learning_rate": 3.235031277926721e-05,
+      "loss": 0.3442,
+      "step": 1039
+    },
+    {
+      "epoch": 1.2545235223160434,
+      "grad_norm": 0.254391649837801,
+      "learning_rate": 3.232797140303843e-05,
+      "loss": 0.3594,
+      "step": 1040
+    },
+    {
+      "epoch": 1.255729794933655,
+      "grad_norm": 0.39919916051774196,
+      "learning_rate": 3.230563002680965e-05,
+      "loss": 0.3839,
+      "step": 1041
+    },
+    {
+      "epoch": 1.2569360675512666,
+      "grad_norm": 0.27024575975320314,
+      "learning_rate": 3.228328865058088e-05,
+      "loss": 0.3533,
+      "step": 1042
+    },
+    {
+      "epoch": 1.2581423401688783,
+      "grad_norm": 0.319622346903351,
+      "learning_rate": 3.22609472743521e-05,
+      "loss": 0.3794,
+      "step": 1043
+    },
+    {
+      "epoch": 1.2593486127864897,
+      "grad_norm": 0.389068060223662,
+      "learning_rate": 3.2238605898123325e-05,
+      "loss": 0.382,
+      "step": 1044
+    },
+    {
+      "epoch": 1.2605548854041013,
+      "grad_norm": 0.3268225256048267,
+      "learning_rate": 3.221626452189455e-05,
+      "loss": 0.3603,
+      "step": 1045
+    },
+    {
+      "epoch": 1.261761158021713,
+      "grad_norm": 0.3633092550933109,
+      "learning_rate": 3.2193923145665774e-05,
+      "loss": 0.3858,
+      "step": 1046
+    },
+    {
+      "epoch": 1.2629674306393244,
+      "grad_norm": 0.32431387176898796,
+      "learning_rate": 3.2171581769437e-05,
+      "loss": 0.3809,
+      "step": 1047
+    },
+    {
+      "epoch": 1.264173703256936,
+      "grad_norm": 0.3003671692231628,
+      "learning_rate": 3.214924039320822e-05,
+      "loss": 0.3654,
+      "step": 1048
+    },
+    {
+      "epoch": 1.2653799758745476,
+      "grad_norm": 0.3133031222068096,
+      "learning_rate": 3.212689901697944e-05,
+      "loss": 0.3542,
+      "step": 1049
+    },
+    {
+      "epoch": 1.2665862484921593,
+      "grad_norm": 0.27077930029731606,
+      "learning_rate": 3.210455764075067e-05,
+      "loss": 0.3732,
+      "step": 1050
+    },
+    {
+      "epoch": 1.267792521109771,
+      "grad_norm": 0.3054937382145299,
+      "learning_rate": 3.20822162645219e-05,
+      "loss": 0.3524,
+      "step": 1051
+    },
+    {
+      "epoch": 1.2689987937273823,
+      "grad_norm": 0.3178405295280176,
+      "learning_rate": 3.205987488829312e-05,
+      "loss": 0.3629,
+      "step": 1052
+    },
+    {
+      "epoch": 1.270205066344994,
+      "grad_norm": 0.31708936628519185,
+      "learning_rate": 3.2037533512064346e-05,
+      "loss": 0.381,
+      "step": 1053
+    },
+    {
+      "epoch": 1.2714113389626056,
+      "grad_norm": 0.24279020858927816,
+      "learning_rate": 3.201519213583557e-05,
+      "loss": 0.3411,
+      "step": 1054
+    },
+    {
+      "epoch": 1.272617611580217,
+      "grad_norm": 0.28116483364277983,
+      "learning_rate": 3.1992850759606794e-05,
+      "loss": 0.3676,
+      "step": 1055
+    },
+    {
+      "epoch": 1.2738238841978287,
+      "grad_norm": 0.3183015486291266,
+      "learning_rate": 3.1970509383378014e-05,
+      "loss": 0.3887,
+      "step": 1056
+    },
+    {
+      "epoch": 1.2750301568154403,
+      "grad_norm": 0.27640816805976137,
+      "learning_rate": 3.194816800714924e-05,
+      "loss": 0.3612,
+      "step": 1057
+    },
+    {
+      "epoch": 1.276236429433052,
+      "grad_norm": 0.2708010845965508,
+      "learning_rate": 3.192582663092047e-05,
+      "loss": 0.3549,
+      "step": 1058
+    },
+    {
+      "epoch": 1.2774427020506636,
+      "grad_norm": 0.3198350134586524,
+      "learning_rate": 3.190348525469169e-05,
+      "loss": 0.3726,
+      "step": 1059
+    },
+    {
+      "epoch": 1.278648974668275,
+      "grad_norm": 0.30699976358969727,
+      "learning_rate": 3.188114387846292e-05,
+      "loss": 0.3715,
+      "step": 1060
+    },
+    {
+      "epoch": 1.2798552472858866,
+      "grad_norm": 0.277809943346299,
+      "learning_rate": 3.1858802502234145e-05,
+      "loss": 0.3392,
+      "step": 1061
+    },
+    {
+      "epoch": 1.2810615199034983,
+      "grad_norm": 0.31996659071085015,
+      "learning_rate": 3.1836461126005366e-05,
+      "loss": 0.3803,
+      "step": 1062
+    },
+    {
+      "epoch": 1.2822677925211097,
+      "grad_norm": 0.33637724087127435,
+      "learning_rate": 3.1814119749776586e-05,
+      "loss": 0.3739,
+      "step": 1063
+    },
+    {
+      "epoch": 1.2834740651387213,
+      "grad_norm": 0.36483447767464733,
+      "learning_rate": 3.1791778373547814e-05,
+      "loss": 0.3475,
+      "step": 1064
+    },
+    {
+      "epoch": 1.284680337756333,
+      "grad_norm": 0.33797545906118115,
+      "learning_rate": 3.1769436997319035e-05,
+      "loss": 0.3559,
+      "step": 1065
+    },
+    {
+      "epoch": 1.2858866103739446,
+      "grad_norm": 0.3992655714858619,
+      "learning_rate": 3.174709562109026e-05,
+      "loss": 0.3496,
+      "step": 1066
+    },
+    {
+      "epoch": 1.2870928829915562,
+      "grad_norm": 0.3757105460842093,
+      "learning_rate": 3.172475424486149e-05,
+      "loss": 0.3574,
+      "step": 1067
+    },
+    {
+      "epoch": 1.2882991556091676,
+      "grad_norm": 0.3838514715771389,
+      "learning_rate": 3.170241286863271e-05,
+      "loss": 0.3714,
+      "step": 1068
+    },
+    {
+      "epoch": 1.2895054282267793,
+      "grad_norm": 0.3558726533103807,
+      "learning_rate": 3.168007149240393e-05,
+      "loss": 0.3607,
+      "step": 1069
+    },
+    {
+      "epoch": 1.290711700844391,
+      "grad_norm": 0.34928305006041077,
+      "learning_rate": 3.165773011617516e-05,
+      "loss": 0.3688,
+      "step": 1070
+    },
+    {
+      "epoch": 1.2919179734620023,
+      "grad_norm": 0.39750595875770633,
+      "learning_rate": 3.163538873994638e-05,
+      "loss": 0.3585,
+      "step": 1071
+    },
+    {
+      "epoch": 1.293124246079614,
+      "grad_norm": 0.31647638371635867,
+      "learning_rate": 3.161304736371761e-05,
+      "loss": 0.3327,
+      "step": 1072
+    },
+    {
+      "epoch": 1.2943305186972256,
+      "grad_norm": 0.370193912732274,
+      "learning_rate": 3.1590705987488834e-05,
+      "loss": 0.3827,
+      "step": 1073
+    },
+    {
+      "epoch": 1.2955367913148372,
+      "grad_norm": 0.362268565167923,
+      "learning_rate": 3.1568364611260055e-05,
+      "loss": 0.3817,
+      "step": 1074
+    },
+    {
+      "epoch": 1.2967430639324489,
+      "grad_norm": 0.3333040592999143,
+      "learning_rate": 3.154602323503128e-05,
+      "loss": 0.3859,
+      "step": 1075
+    },
+    {
+      "epoch": 1.2979493365500603,
+      "grad_norm": 0.37198883984977626,
+      "learning_rate": 3.15236818588025e-05,
+      "loss": 0.362,
+      "step": 1076
+    },
+    {
+      "epoch": 1.299155609167672,
+      "grad_norm": 0.32479600140192555,
+      "learning_rate": 3.1501340482573724e-05,
+      "loss": 0.361,
+      "step": 1077
+    },
+    {
+      "epoch": 1.3003618817852836,
+      "grad_norm": 0.3041287501118341,
+      "learning_rate": 3.147899910634495e-05,
+      "loss": 0.3551,
+      "step": 1078
+    },
+    {
+      "epoch": 1.301568154402895,
+      "grad_norm": 0.3508350111820672,
+      "learning_rate": 3.145665773011618e-05,
+      "loss": 0.3467,
+      "step": 1079
+    },
+    {
+      "epoch": 1.3027744270205066,
+      "grad_norm": 0.31289719028187885,
+      "learning_rate": 3.14343163538874e-05,
+      "loss": 0.3464,
+      "step": 1080
+    },
+    {
+      "epoch": 1.3039806996381182,
+      "grad_norm": 0.3525515286475814,
+      "learning_rate": 3.141197497765863e-05,
+      "loss": 0.3706,
+      "step": 1081
+    },
+    {
+      "epoch": 1.3051869722557297,
+      "grad_norm": 0.36739799881084373,
+      "learning_rate": 3.1389633601429854e-05,
+      "loss": 0.3826,
+      "step": 1082
+    },
+    {
+      "epoch": 1.3063932448733413,
+      "grad_norm": 0.3551639488288884,
+      "learning_rate": 3.1367292225201075e-05,
+      "loss": 0.3679,
+      "step": 1083
+    },
+    {
+      "epoch": 1.307599517490953,
+      "grad_norm": 0.5166672992275477,
+      "learning_rate": 3.1344950848972296e-05,
+      "loss": 0.3702,
+      "step": 1084
+    },
+    {
+      "epoch": 1.3088057901085646,
+      "grad_norm": 0.28002473093112323,
+      "learning_rate": 3.132260947274352e-05,
+      "loss": 0.365,
+      "step": 1085
+    },
+    {
+      "epoch": 1.3100120627261762,
+      "grad_norm": 0.28704719612423435,
+      "learning_rate": 3.1300268096514744e-05,
+      "loss": 0.3657,
+      "step": 1086
+    },
+    {
+      "epoch": 1.3112183353437876,
+      "grad_norm": 0.2885270894752884,
+      "learning_rate": 3.127792672028597e-05,
+      "loss": 0.3497,
+      "step": 1087
+    },
+    {
+      "epoch": 1.3124246079613993,
+      "grad_norm": 0.28563636025816486,
+      "learning_rate": 3.12555853440572e-05,
+      "loss": 0.3653,
+      "step": 1088
+    },
+    {
+      "epoch": 1.313630880579011,
+      "grad_norm": 0.3654990574163292,
+      "learning_rate": 3.123324396782842e-05,
+      "loss": 0.3598,
+      "step": 1089
+    },
+    {
+      "epoch": 1.3148371531966223,
+      "grad_norm": 0.31153048181894816,
+      "learning_rate": 3.121090259159965e-05,
+      "loss": 0.3688,
+      "step": 1090
+    },
+    {
+      "epoch": 1.316043425814234,
+      "grad_norm": 0.30372482261187445,
+      "learning_rate": 3.118856121537087e-05,
+      "loss": 0.372,
+      "step": 1091
+    },
+    {
+      "epoch": 1.3172496984318456,
+      "grad_norm": 0.3928628759286902,
+      "learning_rate": 3.116621983914209e-05,
+      "loss": 0.366,
+      "step": 1092
+    },
+    {
+      "epoch": 1.3184559710494572,
+      "grad_norm": 0.28051320063916957,
+      "learning_rate": 3.1143878462913316e-05,
+      "loss": 0.372,
+      "step": 1093
+    },
+    {
+      "epoch": 1.3196622436670689,
+      "grad_norm": 0.353437964831599,
+      "learning_rate": 3.1121537086684543e-05,
+      "loss": 0.3784,
+      "step": 1094
+    },
+    {
+      "epoch": 1.3208685162846803,
+      "grad_norm": 0.36270292248092095,
+      "learning_rate": 3.1099195710455764e-05,
+      "loss": 0.3784,
+      "step": 1095
+    },
+    {
+      "epoch": 1.322074788902292,
+      "grad_norm": 0.3100183493099864,
+      "learning_rate": 3.107685433422699e-05,
+      "loss": 0.366,
+      "step": 1096
+    },
+    {
+      "epoch": 1.3232810615199035,
+      "grad_norm": 0.33662286705267247,
+      "learning_rate": 3.105451295799822e-05,
+      "loss": 0.3688,
+      "step": 1097
+    },
+    {
+      "epoch": 1.324487334137515,
+      "grad_norm": 0.27058145874774586,
+      "learning_rate": 3.103217158176944e-05,
+      "loss": 0.3558,
+      "step": 1098
+    },
+    {
+      "epoch": 1.3256936067551266,
+      "grad_norm": 0.3814233812729274,
+      "learning_rate": 3.100983020554066e-05,
+      "loss": 0.3543,
+      "step": 1099
+    },
+    {
+      "epoch": 1.3268998793727382,
+      "grad_norm": 0.3440571607399901,
+      "learning_rate": 3.098748882931189e-05,
+      "loss": 0.3524,
+      "step": 1100
+    },
+    {
+      "epoch": 1.3281061519903499,
+      "grad_norm": 0.42548361963668074,
+      "learning_rate": 3.096514745308311e-05,
+      "loss": 0.3677,
+      "step": 1101
+    },
+    {
+      "epoch": 1.3293124246079615,
+      "grad_norm": 0.31583199094929404,
+      "learning_rate": 3.0942806076854336e-05,
+      "loss": 0.3821,
+      "step": 1102
+    },
+    {
+      "epoch": 1.330518697225573,
+      "grad_norm": 0.4079717432016741,
+      "learning_rate": 3.0920464700625564e-05,
+      "loss": 0.3686,
+      "step": 1103
+    },
+    {
+      "epoch": 1.3317249698431846,
+      "grad_norm": 0.3311395530178617,
+      "learning_rate": 3.0898123324396784e-05,
+      "loss": 0.3687,
+      "step": 1104
+    },
+    {
+      "epoch": 1.3329312424607962,
+      "grad_norm": 0.30119192202192563,
+      "learning_rate": 3.087578194816801e-05,
+      "loss": 0.3695,
+      "step": 1105
+    },
+    {
+      "epoch": 1.3341375150784076,
+      "grad_norm": 0.3024724334141376,
+      "learning_rate": 3.085344057193923e-05,
+      "loss": 0.3531,
+      "step": 1106
+    },
+    {
+      "epoch": 1.3353437876960192,
+      "grad_norm": 0.3020172197925826,
+      "learning_rate": 3.083109919571045e-05,
+      "loss": 0.3788,
+      "step": 1107
+    },
+    {
+      "epoch": 1.3365500603136309,
+      "grad_norm": 0.2823663382158882,
+      "learning_rate": 3.080875781948168e-05,
+      "loss": 0.3401,
+      "step": 1108
+    },
+    {
+      "epoch": 1.3377563329312425,
+      "grad_norm": 0.302427062615936,
+      "learning_rate": 3.078641644325291e-05,
+      "loss": 0.3642,
+      "step": 1109
+    },
+    {
+      "epoch": 1.3389626055488542,
+      "grad_norm": 0.31551900898503815,
+      "learning_rate": 3.076407506702413e-05,
+      "loss": 0.3555,
+      "step": 1110
+    },
+    {
+      "epoch": 1.3401688781664656,
+      "grad_norm": 0.3312984136735708,
+      "learning_rate": 3.0741733690795356e-05,
+      "loss": 0.3673,
+      "step": 1111
+    },
+    {
+      "epoch": 1.3413751507840772,
+      "grad_norm": 0.31739717512214,
+      "learning_rate": 3.0719392314566584e-05,
+      "loss": 0.357,
+      "step": 1112
+    },
+    {
+      "epoch": 1.3425814234016888,
+      "grad_norm": 0.2727046695858616,
+      "learning_rate": 3.0697050938337804e-05,
+      "loss": 0.3634,
+      "step": 1113
+    },
+    {
+      "epoch": 1.3437876960193003,
+      "grad_norm": 0.3776160250845144,
+      "learning_rate": 3.0674709562109025e-05,
+      "loss": 0.3812,
+      "step": 1114
+    },
+    {
+      "epoch": 1.344993968636912,
+      "grad_norm": 0.331153836222907,
+      "learning_rate": 3.065236818588025e-05,
+      "loss": 0.3706,
+      "step": 1115
+    },
+    {
+      "epoch": 1.3462002412545235,
+      "grad_norm": 0.3135798994454436,
+      "learning_rate": 3.063002680965147e-05,
+      "loss": 0.3695,
+      "step": 1116
+    },
+    {
+      "epoch": 1.3474065138721352,
+      "grad_norm": 0.3398371342476646,
+      "learning_rate": 3.06076854334227e-05,
+      "loss": 0.3648,
+      "step": 1117
+    },
+    {
+      "epoch": 1.3486127864897468,
+      "grad_norm": 0.3355446307451407,
+      "learning_rate": 3.058534405719393e-05,
+      "loss": 0.3689,
+      "step": 1118
+    },
+    {
+      "epoch": 1.3498190591073582,
+      "grad_norm": 0.32871921032944523,
+      "learning_rate": 3.056300268096515e-05,
+      "loss": 0.3823,
+      "step": 1119
+    },
+    {
+      "epoch": 1.3510253317249699,
+      "grad_norm": 0.3324992371527075,
+      "learning_rate": 3.0540661304736376e-05,
+      "loss": 0.3533,
+      "step": 1120
+    },
+    {
+      "epoch": 1.3522316043425815,
+      "grad_norm": 0.3898432118600308,
+      "learning_rate": 3.05183199285076e-05,
+      "loss": 0.3482,
+      "step": 1121
+    },
+    {
+      "epoch": 1.353437876960193,
+      "grad_norm": 0.27000069674814814,
+      "learning_rate": 3.049597855227882e-05,
+      "loss": 0.367,
+      "step": 1122
+    },
+    {
+      "epoch": 1.3546441495778045,
+      "grad_norm": 0.3560131101054812,
+      "learning_rate": 3.0473637176050045e-05,
+      "loss": 0.3725,
+      "step": 1123
+    },
+    {
+      "epoch": 1.3558504221954162,
+      "grad_norm": 0.3466814385369649,
+      "learning_rate": 3.0451295799821273e-05,
+      "loss": 0.3614,
+      "step": 1124
+    },
+    {
+      "epoch": 1.3570566948130278,
+      "grad_norm": 0.3147914304038281,
+      "learning_rate": 3.0428954423592494e-05,
+      "loss": 0.3672,
+      "step": 1125
+    },
+    {
+      "epoch": 1.3582629674306395,
+      "grad_norm": 0.3187982701455795,
+      "learning_rate": 3.0406613047363718e-05,
+      "loss": 0.3545,
+      "step": 1126
+    },
+    {
+      "epoch": 1.3594692400482509,
+      "grad_norm": 0.2971609813242021,
+      "learning_rate": 3.0384271671134945e-05,
+      "loss": 0.3646,
+      "step": 1127
+    },
+    {
+      "epoch": 1.3606755126658625,
+      "grad_norm": 0.28948184614925276,
+      "learning_rate": 3.0361930294906166e-05,
+      "loss": 0.3724,
+      "step": 1128
+    },
+    {
+      "epoch": 1.3618817852834741,
+      "grad_norm": 0.26100840084723725,
+      "learning_rate": 3.0339588918677393e-05,
+      "loss": 0.3589,
+      "step": 1129
+    },
+    {
+      "epoch": 1.3630880579010856,
+      "grad_norm": 0.26791285850652824,
+      "learning_rate": 3.0317247542448617e-05,
+      "loss": 0.3601,
+      "step": 1130
+    },
+    {
+      "epoch": 1.3642943305186972,
+      "grad_norm": 0.29962236845210755,
+      "learning_rate": 3.0294906166219838e-05,
+      "loss": 0.3475,
+      "step": 1131
+    },
+    {
+      "epoch": 1.3655006031363088,
+      "grad_norm": 0.3346919570043122,
+      "learning_rate": 3.0272564789991066e-05,
+      "loss": 0.3578,
+      "step": 1132
+    },
+    {
+      "epoch": 1.3667068757539202,
+      "grad_norm": 0.26181644687495764,
+      "learning_rate": 3.025022341376229e-05,
+      "loss": 0.3618,
+      "step": 1133
+    },
+    {
+      "epoch": 1.3679131483715319,
+      "grad_norm": 0.2724118620499691,
+      "learning_rate": 3.022788203753351e-05,
+      "loss": 0.3685,
+      "step": 1134
+    },
+    {
+      "epoch": 1.3691194209891435,
+      "grad_norm": 0.291344635313524,
+      "learning_rate": 3.0205540661304738e-05,
+      "loss": 0.3563,
+      "step": 1135
+    },
+    {
+      "epoch": 1.3703256936067552,
+      "grad_norm": 0.3091838446015483,
+      "learning_rate": 3.0183199285075965e-05,
+      "loss": 0.3297,
+      "step": 1136
+    },
+    {
+      "epoch": 1.3715319662243668,
+      "grad_norm": 0.2791543209149448,
+      "learning_rate": 3.0160857908847186e-05,
+      "loss": 0.3691,
+      "step": 1137
+    },
+    {
+      "epoch": 1.3727382388419782,
+      "grad_norm": 0.30051706948641876,
+      "learning_rate": 3.013851653261841e-05,
+      "loss": 0.3635,
+      "step": 1138
+    },
+    {
+      "epoch": 1.3739445114595898,
+      "grad_norm": 0.34203068886863713,
+      "learning_rate": 3.0116175156389638e-05,
+      "loss": 0.3786,
+      "step": 1139
+    },
+    {
+      "epoch": 1.3751507840772015,
+      "grad_norm": 0.2751714894785982,
+      "learning_rate": 3.0093833780160858e-05,
+      "loss": 0.3525,
+      "step": 1140
+    },
+    {
+      "epoch": 1.376357056694813,
+      "grad_norm": 0.3541112573734306,
+      "learning_rate": 3.0071492403932082e-05,
+      "loss": 0.3584,
+      "step": 1141
+    },
+    {
+      "epoch": 1.3775633293124245,
+      "grad_norm": 0.3310525285267929,
+      "learning_rate": 3.004915102770331e-05,
+      "loss": 0.3771,
+      "step": 1142
+    },
+    {
+      "epoch": 1.3787696019300362,
+      "grad_norm": 0.2870345778600548,
+      "learning_rate": 3.002680965147453e-05,
+      "loss": 0.3504,
+      "step": 1143
+    },
+    {
+      "epoch": 1.3799758745476478,
+      "grad_norm": 0.3248559870711437,
+      "learning_rate": 3.0004468275245755e-05,
+      "loss": 0.3377,
+      "step": 1144
+    },
+    {
+      "epoch": 1.3811821471652594,
+      "grad_norm": 0.36564049706754104,
+      "learning_rate": 2.9982126899016982e-05,
+      "loss": 0.3572,
+      "step": 1145
+    },
+    {
+      "epoch": 1.3823884197828709,
+      "grad_norm": 0.3125346957664997,
+      "learning_rate": 2.9959785522788203e-05,
+      "loss": 0.366,
+      "step": 1146
+    },
+    {
+      "epoch": 1.3835946924004825,
+      "grad_norm": 0.3716766984185386,
+      "learning_rate": 2.993744414655943e-05,
+      "loss": 0.3692,
+      "step": 1147
+    },
+    {
+      "epoch": 1.3848009650180941,
+      "grad_norm": 0.40406522080836904,
+      "learning_rate": 2.9915102770330654e-05,
+      "loss": 0.3847,
+      "step": 1148
+    },
+    {
+      "epoch": 1.3860072376357055,
+      "grad_norm": 0.3553874252899629,
+      "learning_rate": 2.9892761394101875e-05,
+      "loss": 0.3792,
+      "step": 1149
+    },
+    {
+      "epoch": 1.3872135102533172,
+      "grad_norm": 0.3654246145476641,
+      "learning_rate": 2.9870420017873103e-05,
+      "loss": 0.3523,
+      "step": 1150
+    },
+    {
+      "epoch": 1.3884197828709288,
+      "grad_norm": 0.372654778385698,
+      "learning_rate": 2.9848078641644327e-05,
+      "loss": 0.363,
+      "step": 1151
+    },
+    {
+      "epoch": 1.3896260554885405,
+      "grad_norm": 0.318400474242105,
+      "learning_rate": 2.9825737265415547e-05,
+      "loss": 0.3665,
+      "step": 1152
+    },
+    {
+      "epoch": 1.390832328106152,
+      "grad_norm": 0.4476226138153656,
+      "learning_rate": 2.9803395889186775e-05,
+      "loss": 0.3586,
+      "step": 1153
+    },
+    {
+      "epoch": 1.3920386007237635,
+      "grad_norm": 0.36090984552024646,
+      "learning_rate": 2.9781054512958002e-05,
+      "loss": 0.3878,
+      "step": 1154
+    },
+    {
+      "epoch": 1.3932448733413751,
+      "grad_norm": 0.34995570790330377,
+      "learning_rate": 2.9758713136729223e-05,
+      "loss": 0.3739,
+      "step": 1155
+    },
+    {
+      "epoch": 1.3944511459589868,
+      "grad_norm": 0.4036557498361607,
+      "learning_rate": 2.9736371760500447e-05,
+      "loss": 0.3656,
+      "step": 1156
+    },
+    {
+      "epoch": 1.3956574185765982,
+      "grad_norm": 0.25983094278157565,
+      "learning_rate": 2.9714030384271675e-05,
+      "loss": 0.3651,
+      "step": 1157
+    },
+    {
+      "epoch": 1.3968636911942098,
+      "grad_norm": 0.46585062876561006,
+      "learning_rate": 2.9691689008042895e-05,
+      "loss": 0.366,
+      "step": 1158
+    },
+    {
+      "epoch": 1.3980699638118215,
+      "grad_norm": 0.29308421171324583,
+      "learning_rate": 2.966934763181412e-05,
+      "loss": 0.3472,
+      "step": 1159
+    },
+    {
+      "epoch": 1.399276236429433,
+      "grad_norm": 0.4212438988181364,
+      "learning_rate": 2.9647006255585347e-05,
+      "loss": 0.3652,
+      "step": 1160
+    },
+    {
+      "epoch": 1.4004825090470447,
+      "grad_norm": 0.2845531602103912,
+      "learning_rate": 2.9624664879356567e-05,
+      "loss": 0.3736,
+      "step": 1161
+    },
+    {
+      "epoch": 1.4016887816646562,
+      "grad_norm": 0.4362233547700654,
+      "learning_rate": 2.9602323503127795e-05,
+      "loss": 0.3763,
+      "step": 1162
+    },
+    {
+      "epoch": 1.4028950542822678,
+      "grad_norm": 0.28669154667850383,
+      "learning_rate": 2.957998212689902e-05,
+      "loss": 0.3634,
+      "step": 1163
+    },
+    {
+      "epoch": 1.4041013268998794,
+      "grad_norm": 0.3752571742382678,
+      "learning_rate": 2.955764075067024e-05,
+      "loss": 0.3698,
+      "step": 1164
+    },
+    {
+      "epoch": 1.4053075995174908,
+      "grad_norm": 0.30884212641230485,
+      "learning_rate": 2.9535299374441467e-05,
+      "loss": 0.3504,
+      "step": 1165
+    },
+    {
+      "epoch": 1.4065138721351025,
+      "grad_norm": 0.3587145325155489,
+      "learning_rate": 2.951295799821269e-05,
+      "loss": 0.3646,
+      "step": 1166
+    },
+    {
+      "epoch": 1.4077201447527141,
+      "grad_norm": 0.29533411161129886,
+      "learning_rate": 2.9490616621983912e-05,
+      "loss": 0.3515,
+      "step": 1167
+    },
+    {
+      "epoch": 1.4089264173703258,
+      "grad_norm": 0.2615406154832229,
+      "learning_rate": 2.946827524575514e-05,
+      "loss": 0.3574,
+      "step": 1168
+    },
+    {
+      "epoch": 1.4101326899879374,
+      "grad_norm": 0.32893567819819586,
+      "learning_rate": 2.9445933869526367e-05,
+      "loss": 0.3603,
+      "step": 1169
+    },
+    {
+      "epoch": 1.4113389626055488,
+      "grad_norm": 0.29965434978983685,
+      "learning_rate": 2.9423592493297584e-05,
+      "loss": 0.3596,
+      "step": 1170
+    },
+    {
+      "epoch": 1.4125452352231604,
+      "grad_norm": 0.31822499295465917,
+      "learning_rate": 2.9401251117068812e-05,
+      "loss": 0.3675,
+      "step": 1171
+    },
+    {
+      "epoch": 1.413751507840772,
+      "grad_norm": 0.2753822644102095,
+      "learning_rate": 2.937890974084004e-05,
+      "loss": 0.3451,
+      "step": 1172
+    },
+    {
+      "epoch": 1.4149577804583835,
+      "grad_norm": 0.3298050786742377,
+      "learning_rate": 2.935656836461126e-05,
+      "loss": 0.3706,
+      "step": 1173
+    },
+    {
+      "epoch": 1.4161640530759951,
+      "grad_norm": 0.28445112216104007,
+      "learning_rate": 2.9334226988382484e-05,
+      "loss": 0.3507,
+      "step": 1174
+    },
+    {
+      "epoch": 1.4173703256936068,
+      "grad_norm": 0.3092808406057826,
+      "learning_rate": 2.931188561215371e-05,
+      "loss": 0.3958,
+      "step": 1175
+    },
+    {
+      "epoch": 1.4185765983112184,
+      "grad_norm": 0.2536950340386672,
+      "learning_rate": 2.9289544235924932e-05,
+      "loss": 0.3511,
+      "step": 1176
+    },
+    {
+      "epoch": 1.41978287092883,
+      "grad_norm": 0.35919383431613366,
+      "learning_rate": 2.9267202859696156e-05,
+      "loss": 0.3598,
+      "step": 1177
+    },
+    {
+      "epoch": 1.4209891435464415,
+      "grad_norm": 0.30220803684797587,
+      "learning_rate": 2.9244861483467384e-05,
+      "loss": 0.3671,
+      "step": 1178
+    },
+    {
+      "epoch": 1.422195416164053,
+      "grad_norm": 0.3015596602053583,
+      "learning_rate": 2.9222520107238604e-05,
+      "loss": 0.3627,
+      "step": 1179
+    },
+    {
+      "epoch": 1.4234016887816647,
+      "grad_norm": 0.28995228992941746,
+      "learning_rate": 2.9200178731009832e-05,
+      "loss": 0.35,
+      "step": 1180
+    },
+    {
+      "epoch": 1.4246079613992761,
+      "grad_norm": 0.29605766490997004,
+      "learning_rate": 2.9177837354781056e-05,
+      "loss": 0.3567,
+      "step": 1181
+    },
+    {
+      "epoch": 1.4258142340168878,
+      "grad_norm": 0.2976811007058239,
+      "learning_rate": 2.9155495978552283e-05,
+      "loss": 0.3436,
+      "step": 1182
+    },
+    {
+      "epoch": 1.4270205066344994,
+      "grad_norm": 0.2601540220761718,
+      "learning_rate": 2.9133154602323504e-05,
+      "loss": 0.3477,
+      "step": 1183
+    },
+    {
+      "epoch": 1.428226779252111,
+      "grad_norm": 0.2834282356161823,
+      "learning_rate": 2.9110813226094728e-05,
+      "loss": 0.3602,
+      "step": 1184
+    },
+    {
+      "epoch": 1.4294330518697227,
+      "grad_norm": 0.291754499437398,
+      "learning_rate": 2.9088471849865956e-05,
+      "loss": 0.3633,
+      "step": 1185
+    },
+    {
+      "epoch": 1.430639324487334,
+      "grad_norm": 0.2735533586976796,
+      "learning_rate": 2.9066130473637176e-05,
+      "loss": 0.3408,
+      "step": 1186
+    },
+    {
+      "epoch": 1.4318455971049457,
+      "grad_norm": 0.28797235804097754,
+      "learning_rate": 2.9043789097408404e-05,
+      "loss": 0.3665,
+      "step": 1187
+    },
+    {
+      "epoch": 1.4330518697225574,
+      "grad_norm": 0.27749621251874057,
+      "learning_rate": 2.9021447721179628e-05,
+      "loss": 0.364,
+      "step": 1188
+    },
+    {
+      "epoch": 1.4342581423401688,
+      "grad_norm": 0.32339392380392684,
+      "learning_rate": 2.899910634495085e-05,
+      "loss": 0.3514,
+      "step": 1189
+    },
+    {
+      "epoch": 1.4354644149577804,
+      "grad_norm": 0.2926720133950538,
+      "learning_rate": 2.8976764968722076e-05,
+      "loss": 0.3842,
+      "step": 1190
+    },
+    {
+      "epoch": 1.436670687575392,
+      "grad_norm": 0.28827795732382455,
+      "learning_rate": 2.89544235924933e-05,
+      "loss": 0.3606,
+      "step": 1191
+    },
+    {
+      "epoch": 1.4378769601930035,
+      "grad_norm": 0.2915271734549216,
+      "learning_rate": 2.893208221626452e-05,
+      "loss": 0.3536,
+      "step": 1192
+    },
+    {
+      "epoch": 1.4390832328106151,
+      "grad_norm": 0.40519254838528146,
+      "learning_rate": 2.890974084003575e-05,
+      "loss": 0.3751,
+      "step": 1193
+    },
+    {
+      "epoch": 1.4402895054282268,
+      "grad_norm": 0.29164522307451995,
+      "learning_rate": 2.8887399463806976e-05,
+      "loss": 0.3705,
+      "step": 1194
+    },
+    {
+      "epoch": 1.4414957780458384,
+      "grad_norm": 0.311020331010995,
+      "learning_rate": 2.8865058087578197e-05,
+      "loss": 0.3538,
+      "step": 1195
+    },
+    {
+      "epoch": 1.44270205066345,
+      "grad_norm": 0.35439846786285,
+      "learning_rate": 2.884271671134942e-05,
+      "loss": 0.3823,
+      "step": 1196
+    },
+    {
+      "epoch": 1.4439083232810614,
+      "grad_norm": 0.31436335956903844,
+      "learning_rate": 2.8820375335120648e-05,
+      "loss": 0.3714,
+      "step": 1197
+    },
+    {
+      "epoch": 1.445114595898673,
+      "grad_norm": 0.35023249437780235,
+      "learning_rate": 2.879803395889187e-05,
+      "loss": 0.3915,
+      "step": 1198
+    },
+    {
+      "epoch": 1.4463208685162847,
+      "grad_norm": 0.28045651639475067,
+      "learning_rate": 2.8775692582663093e-05,
+      "loss": 0.346,
+      "step": 1199
+    },
+    {
+      "epoch": 1.4475271411338961,
+      "grad_norm": 0.31926475304857066,
+      "learning_rate": 2.875335120643432e-05,
+      "loss": 0.3444,
+      "step": 1200
+    },
+    {
+      "epoch": 1.4487334137515078,
+      "grad_norm": 0.28708496856007876,
+      "learning_rate": 2.873100983020554e-05,
+      "loss": 0.3498,
+      "step": 1201
+    },
+    {
+      "epoch": 1.4499396863691194,
+      "grad_norm": 0.43262881864227015,
+      "learning_rate": 2.870866845397677e-05,
+      "loss": 0.3505,
+      "step": 1202
+    },
+    {
+      "epoch": 1.451145958986731,
+      "grad_norm": 0.2687357810356031,
+      "learning_rate": 2.8686327077747993e-05,
+      "loss": 0.3519,
+      "step": 1203
+    },
+    {
+      "epoch": 1.4523522316043427,
+      "grad_norm": 0.3607507222010132,
+      "learning_rate": 2.8663985701519213e-05,
+      "loss": 0.3723,
+      "step": 1204
+    },
+    {
+      "epoch": 1.453558504221954,
+      "grad_norm": 0.30338701749850056,
+      "learning_rate": 2.864164432529044e-05,
+      "loss": 0.3587,
+      "step": 1205
+    },
+    {
+      "epoch": 1.4547647768395657,
+      "grad_norm": 0.31603329875122715,
+      "learning_rate": 2.8619302949061665e-05,
+      "loss": 0.3757,
+      "step": 1206
+    },
+    {
+      "epoch": 1.4559710494571774,
+      "grad_norm": 0.33022323730816805,
+      "learning_rate": 2.8596961572832886e-05,
+      "loss": 0.3699,
+      "step": 1207
+    },
+    {
+      "epoch": 1.4571773220747888,
+      "grad_norm": 0.306872731375517,
+      "learning_rate": 2.8574620196604113e-05,
+      "loss": 0.3915,
+      "step": 1208
+    },
+    {
+      "epoch": 1.4583835946924004,
+      "grad_norm": 0.34818044786028685,
+      "learning_rate": 2.8552278820375337e-05,
+      "loss": 0.3686,
+      "step": 1209
+    },
+    {
+      "epoch": 1.459589867310012,
+      "grad_norm": 0.3180265730282909,
+      "learning_rate": 2.8529937444146558e-05,
+      "loss": 0.3433,
+      "step": 1210
+    },
+    {
+      "epoch": 1.4607961399276237,
+      "grad_norm": 0.28857128660801623,
+      "learning_rate": 2.8507596067917785e-05,
+      "loss": 0.3366,
+      "step": 1211
+    },
+    {
+      "epoch": 1.4620024125452353,
+      "grad_norm": 0.3482553573855266,
+      "learning_rate": 2.8485254691689013e-05,
+      "loss": 0.3604,
+      "step": 1212
+    },
+    {
+      "epoch": 1.4632086851628467,
+      "grad_norm": 0.30676083475386495,
+      "learning_rate": 2.8462913315460234e-05,
+      "loss": 0.3815,
+      "step": 1213
+    },
+    {
+      "epoch": 1.4644149577804584,
+      "grad_norm": 0.30367765388377194,
+      "learning_rate": 2.8440571939231458e-05,
+      "loss": 0.3726,
+      "step": 1214
+    },
+    {
+      "epoch": 1.46562123039807,
+      "grad_norm": 0.3446510768407655,
+      "learning_rate": 2.8418230563002685e-05,
+      "loss": 0.3677,
+      "step": 1215
+    },
+    {
+      "epoch": 1.4668275030156814,
+      "grad_norm": 0.3172797310351058,
+      "learning_rate": 2.8395889186773906e-05,
+      "loss": 0.363,
+      "step": 1216
+    },
+    {
+      "epoch": 1.468033775633293,
+      "grad_norm": 0.2953075162278567,
+      "learning_rate": 2.837354781054513e-05,
+      "loss": 0.3756,
+      "step": 1217
+    },
+    {
+      "epoch": 1.4692400482509047,
+      "grad_norm": 0.32783840098529415,
+      "learning_rate": 2.8351206434316357e-05,
+      "loss": 0.3642,
+      "step": 1218
+    },
+    {
+      "epoch": 1.4704463208685163,
+      "grad_norm": 0.280445028660401,
+      "learning_rate": 2.8328865058087578e-05,
+      "loss": 0.3513,
+      "step": 1219
+    },
+    {
+      "epoch": 1.471652593486128,
+      "grad_norm": 0.33558468097054406,
+      "learning_rate": 2.8306523681858806e-05,
+      "loss": 0.3651,
+      "step": 1220
+    },
+    {
+      "epoch": 1.4728588661037394,
+      "grad_norm": 0.336070959339012,
+      "learning_rate": 2.828418230563003e-05,
+      "loss": 0.3591,
+      "step": 1221
+    },
+    {
+      "epoch": 1.474065138721351,
+      "grad_norm": 0.2900358012573861,
+      "learning_rate": 2.826184092940125e-05,
+      "loss": 0.3744,
+      "step": 1222
+    },
+    {
+      "epoch": 1.4752714113389627,
+      "grad_norm": 0.2972532526663739,
+      "learning_rate": 2.8239499553172478e-05,
+      "loss": 0.3502,
+      "step": 1223
+    },
+    {
+      "epoch": 1.476477683956574,
+      "grad_norm": 0.3078587095137843,
+      "learning_rate": 2.8217158176943702e-05,
+      "loss": 0.3647,
+      "step": 1224
+    },
+    {
+      "epoch": 1.4776839565741857,
+      "grad_norm": 0.32293847798260233,
+      "learning_rate": 2.8194816800714923e-05,
+      "loss": 0.3621,
+      "step": 1225
+    },
+    {
+      "epoch": 1.4788902291917974,
+      "grad_norm": 0.2698381433035557,
+      "learning_rate": 2.817247542448615e-05,
+      "loss": 0.3625,
+      "step": 1226
+    },
+    {
+      "epoch": 1.480096501809409,
+      "grad_norm": 0.3092649084688229,
+      "learning_rate": 2.8150134048257378e-05,
+      "loss": 0.3588,
+      "step": 1227
+    },
+    {
+      "epoch": 1.4813027744270206,
+      "grad_norm": 0.32836804858453694,
+      "learning_rate": 2.81277926720286e-05,
+      "loss": 0.3726,
+      "step": 1228
+    },
+    {
+      "epoch": 1.482509047044632,
+      "grad_norm": 0.4687070308155192,
+      "learning_rate": 2.8105451295799822e-05,
+      "loss": 0.3517,
+      "step": 1229
+    },
+    {
+      "epoch": 1.4837153196622437,
+      "grad_norm": 0.33776920307320657,
+      "learning_rate": 2.808310991957105e-05,
+      "loss": 0.363,
+      "step": 1230
+    },
+    {
+      "epoch": 1.4849215922798553,
+      "grad_norm": 0.34831064110686916,
+      "learning_rate": 2.806076854334227e-05,
+      "loss": 0.3582,
+      "step": 1231
+    },
+    {
+      "epoch": 1.4861278648974667,
+      "grad_norm": 0.3530691789102461,
+      "learning_rate": 2.8038427167113495e-05,
+      "loss": 0.3749,
+      "step": 1232
+    },
+    {
+      "epoch": 1.4873341375150784,
+      "grad_norm": 0.31161336626579134,
+      "learning_rate": 2.8016085790884722e-05,
+      "loss": 0.3447,
+      "step": 1233
+    },
+    {
+      "epoch": 1.48854041013269,
+      "grad_norm": 0.2861464976049845,
+      "learning_rate": 2.7993744414655943e-05,
+      "loss": 0.3377,
+      "step": 1234
+    },
+    {
+      "epoch": 1.4897466827503016,
+      "grad_norm": 0.3600397490414202,
+      "learning_rate": 2.7971403038427167e-05,
+      "loss": 0.3677,
+      "step": 1235
+    },
+    {
+      "epoch": 1.4909529553679133,
+      "grad_norm": 0.28131622511858695,
+      "learning_rate": 2.7949061662198394e-05,
+      "loss": 0.3789,
+      "step": 1236
+    },
+    {
+      "epoch": 1.4921592279855247,
+      "grad_norm": 0.3516965064598944,
+      "learning_rate": 2.7926720285969615e-05,
+      "loss": 0.3852,
+      "step": 1237
+    },
+    {
+      "epoch": 1.4933655006031363,
+      "grad_norm": 0.302444809100267,
+      "learning_rate": 2.7904378909740843e-05,
+      "loss": 0.3666,
+      "step": 1238
+    },
+    {
+      "epoch": 1.494571773220748,
+      "grad_norm": 0.3409455236720351,
+      "learning_rate": 2.7882037533512067e-05,
+      "loss": 0.3713,
+      "step": 1239
+    },
+    {
+      "epoch": 1.4957780458383594,
+      "grad_norm": 0.27999659004938476,
+      "learning_rate": 2.7859696157283287e-05,
+      "loss": 0.3563,
+      "step": 1240
+    },
+    {
+      "epoch": 1.496984318455971,
+      "grad_norm": 0.2948450451811389,
+      "learning_rate": 2.7837354781054515e-05,
+      "loss": 0.3618,
+      "step": 1241
+    },
+    {
+      "epoch": 1.4981905910735827,
+      "grad_norm": 0.2986130381148085,
+      "learning_rate": 2.781501340482574e-05,
+      "loss": 0.3635,
+      "step": 1242
+    },
+    {
+      "epoch": 1.4993968636911943,
+      "grad_norm": 0.29615277744615615,
+      "learning_rate": 2.779267202859696e-05,
+      "loss": 0.3539,
+      "step": 1243
+    },
+    {
+      "epoch": 1.500603136308806,
+      "grad_norm": 0.3364174580644653,
+      "learning_rate": 2.7770330652368187e-05,
+      "loss": 0.3498,
+      "step": 1244
+    },
+    {
+      "epoch": 1.5018094089264173,
+      "grad_norm": 0.2763203714148372,
+      "learning_rate": 2.7747989276139415e-05,
+      "loss": 0.3558,
+      "step": 1245
+    },
+    {
+      "epoch": 1.503015681544029,
+      "grad_norm": 0.34787271866709774,
+      "learning_rate": 2.7725647899910635e-05,
+      "loss": 0.3673,
+      "step": 1246
+    },
+    {
+      "epoch": 1.5042219541616406,
+      "grad_norm": 0.30434396850748885,
+      "learning_rate": 2.770330652368186e-05,
+      "loss": 0.3638,
+      "step": 1247
+    },
+    {
+      "epoch": 1.505428226779252,
+      "grad_norm": 0.33166499767283425,
+      "learning_rate": 2.7680965147453087e-05,
+      "loss": 0.3515,
+      "step": 1248
+    },
+    {
+      "epoch": 1.5066344993968637,
+      "grad_norm": 0.29134837059756147,
+      "learning_rate": 2.7658623771224308e-05,
+      "loss": 0.3691,
+      "step": 1249
+    },
+    {
+      "epoch": 1.5078407720144753,
+      "grad_norm": 0.2628685279525233,
+      "learning_rate": 2.763628239499553e-05,
+      "loss": 0.362,
+      "step": 1250
+    },
+    {
+      "epoch": 1.5090470446320867,
+      "grad_norm": 0.32505828654754587,
+      "learning_rate": 2.761394101876676e-05,
+      "loss": 0.3617,
+      "step": 1251
+    },
+    {
+      "epoch": 1.5102533172496986,
+      "grad_norm": 0.29118585516826345,
+      "learning_rate": 2.759159964253798e-05,
+      "loss": 0.3559,
+      "step": 1252
+    },
+    {
+      "epoch": 1.51145958986731,
+      "grad_norm": 0.3142120436832583,
+      "learning_rate": 2.7569258266309207e-05,
+      "loss": 0.3615,
+      "step": 1253
+    },
+    {
+      "epoch": 1.5126658624849216,
+      "grad_norm": 0.2795160928845954,
+      "learning_rate": 2.754691689008043e-05,
+      "loss": 0.3386,
+      "step": 1254
+    },
+    {
+      "epoch": 1.5138721351025333,
+      "grad_norm": 0.35336962090838353,
+      "learning_rate": 2.7524575513851652e-05,
+      "loss": 0.364,
+      "step": 1255
+    },
+    {
+      "epoch": 1.5150784077201447,
+      "grad_norm": 0.27339107672471447,
+      "learning_rate": 2.750223413762288e-05,
+      "loss": 0.3586,
+      "step": 1256
+    },
+    {
+      "epoch": 1.5162846803377563,
+      "grad_norm": 0.349577261694894,
+      "learning_rate": 2.7479892761394104e-05,
+      "loss": 0.3643,
+      "step": 1257
+    },
+    {
+      "epoch": 1.517490952955368,
+      "grad_norm": 0.33322429078016025,
+      "learning_rate": 2.7457551385165324e-05,
+      "loss": 0.3607,
+      "step": 1258
+    },
+    {
+      "epoch": 1.5186972255729794,
+      "grad_norm": 0.3461037586939447,
+      "learning_rate": 2.7435210008936552e-05,
+      "loss": 0.3623,
+      "step": 1259
+    },
+    {
+      "epoch": 1.5199034981905912,
+      "grad_norm": 0.28408740662062104,
+      "learning_rate": 2.741286863270778e-05,
+      "loss": 0.3673,
+      "step": 1260
+    },
+    {
+      "epoch": 1.5211097708082026,
+      "grad_norm": 0.29583059946956347,
+      "learning_rate": 2.7390527256478997e-05,
+      "loss": 0.3416,
+      "step": 1261
+    },
+    {
+      "epoch": 1.5223160434258143,
+      "grad_norm": 0.3440235067952912,
+      "learning_rate": 2.7368185880250224e-05,
+      "loss": 0.3679,
+      "step": 1262
+    },
+    {
+      "epoch": 1.523522316043426,
+      "grad_norm": 0.3120646894770821,
+      "learning_rate": 2.734584450402145e-05,
+      "loss": 0.3489,
+      "step": 1263
+    },
+    {
+      "epoch": 1.5247285886610373,
+      "grad_norm": 0.31673685428487164,
+      "learning_rate": 2.7323503127792672e-05,
+      "loss": 0.3291,
+      "step": 1264
+    },
+    {
+      "epoch": 1.525934861278649,
+      "grad_norm": 0.3147048237479034,
+      "learning_rate": 2.7301161751563896e-05,
+      "loss": 0.3509,
+      "step": 1265
+    },
+    {
+      "epoch": 1.5271411338962606,
+      "grad_norm": 0.3914175398497735,
+      "learning_rate": 2.7278820375335124e-05,
+      "loss": 0.3815,
+      "step": 1266
+    },
+    {
+      "epoch": 1.528347406513872,
+      "grad_norm": 0.4317414622229666,
+      "learning_rate": 2.7256478999106345e-05,
+      "loss": 0.3656,
+      "step": 1267
+    },
+    {
+      "epoch": 1.5295536791314839,
+      "grad_norm": 0.3462625512487634,
+      "learning_rate": 2.723413762287757e-05,
+      "loss": 0.3548,
+      "step": 1268
+    },
+    {
+      "epoch": 1.5307599517490953,
+      "grad_norm": 0.3130762407373115,
+      "learning_rate": 2.7211796246648796e-05,
+      "loss": 0.3481,
+      "step": 1269
+    },
+    {
+      "epoch": 1.531966224366707,
+      "grad_norm": 0.3096037049836567,
+      "learning_rate": 2.7189454870420017e-05,
+      "loss": 0.3663,
+      "step": 1270
+    },
+    {
+      "epoch": 1.5331724969843186,
+      "grad_norm": 0.3567741337319747,
+      "learning_rate": 2.7167113494191244e-05,
+      "loss": 0.3654,
+      "step": 1271
+    },
+    {
+      "epoch": 1.53437876960193,
+      "grad_norm": 0.2695934120971819,
+      "learning_rate": 2.714477211796247e-05,
+      "loss": 0.3724,
+      "step": 1272
+    },
+    {
+      "epoch": 1.5355850422195416,
+      "grad_norm": 0.35011201758948296,
+      "learning_rate": 2.712243074173369e-05,
+      "loss": 0.3962,
+      "step": 1273
+    },
+    {
+      "epoch": 1.5367913148371533,
+      "grad_norm": 0.31227812163963997,
+      "learning_rate": 2.7100089365504917e-05,
+      "loss": 0.3743,
+      "step": 1274
+    },
+    {
+      "epoch": 1.5379975874547647,
+      "grad_norm": 0.24710433568638754,
+      "learning_rate": 2.707774798927614e-05,
+      "loss": 0.3548,
+      "step": 1275
+    },
+    {
+      "epoch": 1.5392038600723763,
+      "grad_norm": 0.32584533326819515,
+      "learning_rate": 2.705540661304736e-05,
+      "loss": 0.369,
+      "step": 1276
+    },
+    {
+      "epoch": 1.540410132689988,
+      "grad_norm": 0.3136506898477773,
+      "learning_rate": 2.703306523681859e-05,
+      "loss": 0.3483,
+      "step": 1277
+    },
+    {
+      "epoch": 1.5416164053075994,
+      "grad_norm": 0.2835482067624798,
+      "learning_rate": 2.7010723860589816e-05,
+      "loss": 0.3794,
+      "step": 1278
+    },
+    {
+      "epoch": 1.5428226779252112,
+      "grad_norm": 0.29765053016237486,
+      "learning_rate": 2.6988382484361037e-05,
+      "loss": 0.3619,
+      "step": 1279
+    },
+    {
+      "epoch": 1.5440289505428226,
+      "grad_norm": 0.2854721347083471,
+      "learning_rate": 2.696604110813226e-05,
+      "loss": 0.3574,
+      "step": 1280
+    },
+    {
+      "epoch": 1.5452352231604343,
+      "grad_norm": 0.27855645978313726,
+      "learning_rate": 2.694369973190349e-05,
+      "loss": 0.3576,
+      "step": 1281
+    },
+    {
+      "epoch": 1.546441495778046,
+      "grad_norm": 0.31553162770965193,
+      "learning_rate": 2.692135835567471e-05,
+      "loss": 0.3545,
+      "step": 1282
+    },
+    {
+      "epoch": 1.5476477683956573,
+      "grad_norm": 0.2860017791263232,
+      "learning_rate": 2.6899016979445933e-05,
+      "loss": 0.3788,
+      "step": 1283
+    },
+    {
+      "epoch": 1.548854041013269,
+      "grad_norm": 0.310042700334375,
+      "learning_rate": 2.687667560321716e-05,
+      "loss": 0.3392,
+      "step": 1284
+    },
+    {
+      "epoch": 1.5500603136308806,
+      "grad_norm": 0.3101101311732966,
+      "learning_rate": 2.685433422698838e-05,
+      "loss": 0.3831,
+      "step": 1285
+    },
+    {
+      "epoch": 1.551266586248492,
+      "grad_norm": 0.24397788929167546,
+      "learning_rate": 2.683199285075961e-05,
+      "loss": 0.3562,
+      "step": 1286
+    },
+    {
+      "epoch": 1.5524728588661039,
+      "grad_norm": 0.28475924698050303,
+      "learning_rate": 2.6809651474530833e-05,
+      "loss": 0.3524,
+      "step": 1287
+    },
+    {
+      "epoch": 1.5536791314837153,
+      "grad_norm": 0.2912846013587237,
+      "learning_rate": 2.6787310098302054e-05,
+      "loss": 0.3515,
+      "step": 1288
+    },
+    {
+      "epoch": 1.554885404101327,
+      "grad_norm": 0.31074089031425983,
+      "learning_rate": 2.676496872207328e-05,
+      "loss": 0.3566,
+      "step": 1289
+    },
+    {
+      "epoch": 1.5560916767189386,
+      "grad_norm": 0.28826636625919533,
+      "learning_rate": 2.6742627345844505e-05,
+      "loss": 0.3605,
+      "step": 1290
+    },
+    {
+      "epoch": 1.55729794933655,
+      "grad_norm": 0.33111363951512146,
+      "learning_rate": 2.6720285969615726e-05,
+      "loss": 0.3751,
+      "step": 1291
+    },
+    {
+      "epoch": 1.5585042219541616,
+      "grad_norm": 0.2677678491527853,
+      "learning_rate": 2.6697944593386954e-05,
+      "loss": 0.3632,
+      "step": 1292
+    },
+    {
+      "epoch": 1.5597104945717732,
+      "grad_norm": 0.31135416808124383,
+      "learning_rate": 2.667560321715818e-05,
+      "loss": 0.3744,
+      "step": 1293
+    },
+    {
+      "epoch": 1.5609167671893847,
+      "grad_norm": 0.30091299310513797,
+      "learning_rate": 2.66532618409294e-05,
+      "loss": 0.3855,
+      "step": 1294
+    },
+    {
+      "epoch": 1.5621230398069965,
+      "grad_norm": 0.23261273418362208,
+      "learning_rate": 2.6630920464700626e-05,
+      "loss": 0.3438,
+      "step": 1295
+    },
+    {
+      "epoch": 1.563329312424608,
+      "grad_norm": 0.32048535435397274,
+      "learning_rate": 2.6608579088471853e-05,
+      "loss": 0.3781,
+      "step": 1296
+    },
+    {
+      "epoch": 1.5645355850422196,
+      "grad_norm": 0.2628320985489263,
+      "learning_rate": 2.6586237712243074e-05,
+      "loss": 0.3387,
+      "step": 1297
+    },
+    {
+      "epoch": 1.5657418576598312,
+      "grad_norm": 0.27460750070719586,
+      "learning_rate": 2.6563896336014298e-05,
+      "loss": 0.3569,
+      "step": 1298
+    },
+    {
+      "epoch": 1.5669481302774426,
+      "grad_norm": 0.27441942018435256,
+      "learning_rate": 2.6541554959785526e-05,
+      "loss": 0.3647,
+      "step": 1299
+    },
+    {
+      "epoch": 1.5681544028950543,
+      "grad_norm": 0.2876053463336936,
+      "learning_rate": 2.6519213583556746e-05,
+      "loss": 0.3622,
+      "step": 1300
+    },
+    {
+      "epoch": 1.569360675512666,
+      "grad_norm": 0.2754619238920535,
+      "learning_rate": 2.649687220732797e-05,
+      "loss": 0.345,
+      "step": 1301
+    },
+    {
+      "epoch": 1.5705669481302773,
+      "grad_norm": 0.31328417122026686,
+      "learning_rate": 2.6474530831099198e-05,
+      "loss": 0.3875,
+      "step": 1302
+    },
+    {
+      "epoch": 1.5717732207478892,
+      "grad_norm": 0.2610769698506394,
+      "learning_rate": 2.645218945487042e-05,
+      "loss": 0.3473,
+      "step": 1303
+    },
+    {
+      "epoch": 1.5729794933655006,
+      "grad_norm": 0.31481781480005866,
+      "learning_rate": 2.6429848078641646e-05,
+      "loss": 0.3639,
+      "step": 1304
+    },
+    {
+      "epoch": 1.5741857659831122,
+      "grad_norm": 0.3016217566559489,
+      "learning_rate": 2.640750670241287e-05,
+      "loss": 0.3769,
+      "step": 1305
+    },
+    {
+      "epoch": 1.5753920386007239,
+      "grad_norm": 0.3065182968634683,
+      "learning_rate": 2.6385165326184098e-05,
+      "loss": 0.3576,
+      "step": 1306
+    },
+    {
+      "epoch": 1.5765983112183353,
+      "grad_norm": 0.3094495579248347,
+      "learning_rate": 2.6362823949955318e-05,
+      "loss": 0.3488,
+      "step": 1307
+    },
+    {
+      "epoch": 1.577804583835947,
+      "grad_norm": 0.37263099382228515,
+      "learning_rate": 2.6340482573726542e-05,
+      "loss": 0.365,
+      "step": 1308
+    },
+    {
+      "epoch": 1.5790108564535585,
+      "grad_norm": 0.2532097044448258,
+      "learning_rate": 2.631814119749777e-05,
+      "loss": 0.3459,
+      "step": 1309
+    },
+    {
+      "epoch": 1.58021712907117,
+      "grad_norm": 0.30034042957794177,
+      "learning_rate": 2.629579982126899e-05,
+      "loss": 0.3479,
+      "step": 1310
+    },
+    {
+      "epoch": 1.5814234016887818,
+      "grad_norm": 0.4315569077455076,
+      "learning_rate": 2.6273458445040218e-05,
+      "loss": 0.3785,
+      "step": 1311
+    },
+    {
+      "epoch": 1.5826296743063932,
+      "grad_norm": 0.2611868817203264,
+      "learning_rate": 2.6251117068811442e-05,
+      "loss": 0.3688,
+      "step": 1312
+    },
+    {
+      "epoch": 1.5838359469240049,
+      "grad_norm": 0.3436018569495356,
+      "learning_rate": 2.6228775692582663e-05,
+      "loss": 0.3661,
+      "step": 1313
+    },
+    {
+      "epoch": 1.5850422195416165,
+      "grad_norm": 0.3775983171861393,
+      "learning_rate": 2.620643431635389e-05,
+      "loss": 0.35,
+      "step": 1314
+    },
+    {
+      "epoch": 1.586248492159228,
+      "grad_norm": 0.260001030240192,
+      "learning_rate": 2.6184092940125114e-05,
+      "loss": 0.3607,
+      "step": 1315
+    },
+    {
+      "epoch": 1.5874547647768396,
+      "grad_norm": 0.3192806233423819,
+      "learning_rate": 2.6161751563896335e-05,
+      "loss": 0.3605,
+      "step": 1316
+    },
+    {
+      "epoch": 1.5886610373944512,
+      "grad_norm": 0.34906939857027053,
+      "learning_rate": 2.6139410187667563e-05,
+      "loss": 0.3392,
+      "step": 1317
+    },
+    {
+      "epoch": 1.5898673100120626,
+      "grad_norm": 0.3041401770631904,
+      "learning_rate": 2.611706881143879e-05,
+      "loss": 0.3579,
+      "step": 1318
+    },
+    {
+      "epoch": 1.5910735826296745,
+      "grad_norm": 0.30910377982596804,
+      "learning_rate": 2.609472743521001e-05,
+      "loss": 0.3506,
+      "step": 1319
+    },
+    {
+      "epoch": 1.5922798552472859,
+      "grad_norm": 0.3812561649468747,
+      "learning_rate": 2.6072386058981235e-05,
+      "loss": 0.3753,
+      "step": 1320
+    },
+    {
+      "epoch": 1.5934861278648975,
+      "grad_norm": 0.2945017335023286,
+      "learning_rate": 2.6050044682752462e-05,
+      "loss": 0.3779,
+      "step": 1321
+    },
+    {
+      "epoch": 1.5946924004825092,
+      "grad_norm": 0.3169444082111978,
+      "learning_rate": 2.6027703306523683e-05,
+      "loss": 0.3455,
+      "step": 1322
+    },
+    {
+      "epoch": 1.5958986731001206,
+      "grad_norm": 0.30538941387132723,
+      "learning_rate": 2.6005361930294907e-05,
+      "loss": 0.3619,
+      "step": 1323
+    },
+    {
+      "epoch": 1.5971049457177322,
+      "grad_norm": 0.2750984977884928,
+      "learning_rate": 2.5983020554066135e-05,
+      "loss": 0.3808,
+      "step": 1324
+    },
+    {
+      "epoch": 1.5983112183353438,
+      "grad_norm": 0.30285454154498515,
+      "learning_rate": 2.5960679177837355e-05,
+      "loss": 0.3519,
+      "step": 1325
+    },
+    {
+      "epoch": 1.5995174909529553,
+      "grad_norm": 0.33944179519883355,
+      "learning_rate": 2.593833780160858e-05,
+      "loss": 0.3616,
+      "step": 1326
+    },
+    {
+      "epoch": 1.600723763570567,
+      "grad_norm": 0.26318466057317974,
+      "learning_rate": 2.5915996425379807e-05,
+      "loss": 0.3706,
+      "step": 1327
+    },
+    {
+      "epoch": 1.6019300361881785,
+      "grad_norm": 0.29783411165362406,
+      "learning_rate": 2.5893655049151027e-05,
+      "loss": 0.3572,
+      "step": 1328
+    },
+    {
+      "epoch": 1.60313630880579,
+      "grad_norm": 0.35002375269009606,
+      "learning_rate": 2.5871313672922255e-05,
+      "loss": 0.3579,
+      "step": 1329
+    },
+    {
+      "epoch": 1.6043425814234018,
+      "grad_norm": 0.29418265664397103,
+      "learning_rate": 2.584897229669348e-05,
+      "loss": 0.372,
+      "step": 1330
+    },
+    {
+      "epoch": 1.6055488540410132,
+      "grad_norm": 0.3203570332410853,
+      "learning_rate": 2.58266309204647e-05,
+      "loss": 0.354,
+      "step": 1331
+    },
+    {
+      "epoch": 1.6067551266586249,
+      "grad_norm": 0.3258614955229903,
+      "learning_rate": 2.5804289544235927e-05,
+      "loss": 0.3599,
+      "step": 1332
+    },
+    {
+      "epoch": 1.6079613992762365,
+      "grad_norm": 0.27796800216478634,
+      "learning_rate": 2.578194816800715e-05,
+      "loss": 0.3685,
+      "step": 1333
+    },
+    {
+      "epoch": 1.609167671893848,
+      "grad_norm": 0.3939022371539636,
+      "learning_rate": 2.5759606791778372e-05,
+      "loss": 0.3825,
+      "step": 1334
+    },
+    {
+      "epoch": 1.6103739445114595,
+      "grad_norm": 0.2834697874561306,
+      "learning_rate": 2.57372654155496e-05,
+      "loss": 0.3795,
+      "step": 1335
+    },
+    {
+      "epoch": 1.6115802171290712,
+      "grad_norm": 0.29584363725941537,
+      "learning_rate": 2.5714924039320827e-05,
+      "loss": 0.3788,
+      "step": 1336
+    },
+    {
+      "epoch": 1.6127864897466826,
+      "grad_norm": 0.2986606089197964,
+      "learning_rate": 2.5692582663092048e-05,
+      "loss": 0.3662,
+      "step": 1337
+    },
+    {
+      "epoch": 1.6139927623642945,
+      "grad_norm": 0.33526376758814486,
+      "learning_rate": 2.5670241286863272e-05,
+      "loss": 0.3507,
+      "step": 1338
+    },
+    {
+      "epoch": 1.6151990349819059,
+      "grad_norm": 0.2741610353438533,
+      "learning_rate": 2.56478999106345e-05,
+      "loss": 0.3761,
+      "step": 1339
+    },
+    {
+      "epoch": 1.6164053075995175,
+      "grad_norm": 0.28973832017005374,
+      "learning_rate": 2.562555853440572e-05,
+      "loss": 0.3756,
+      "step": 1340
+    },
+    {
+      "epoch": 1.6176115802171291,
+      "grad_norm": 0.28594808720953097,
+      "learning_rate": 2.5603217158176944e-05,
+      "loss": 0.3501,
+      "step": 1341
+    },
+    {
+      "epoch": 1.6188178528347406,
+      "grad_norm": 0.27271839503423706,
+      "learning_rate": 2.558087578194817e-05,
+      "loss": 0.3552,
+      "step": 1342
+    },
+    {
+      "epoch": 1.6200241254523522,
+      "grad_norm": 0.2709700411591409,
+      "learning_rate": 2.5558534405719392e-05,
+      "loss": 0.3415,
+      "step": 1343
+    },
+    {
+      "epoch": 1.6212303980699638,
+      "grad_norm": 0.2934700703302023,
+      "learning_rate": 2.553619302949062e-05,
+      "loss": 0.3575,
+      "step": 1344
+    },
+    {
+      "epoch": 1.6224366706875752,
+      "grad_norm": 0.3093331948166226,
+      "learning_rate": 2.5513851653261844e-05,
+      "loss": 0.3487,
+      "step": 1345
+    },
+    {
+      "epoch": 1.623642943305187,
+      "grad_norm": 0.2587526200120511,
+      "learning_rate": 2.5491510277033064e-05,
+      "loss": 0.3523,
+      "step": 1346
+    },
+    {
+      "epoch": 1.6248492159227985,
+      "grad_norm": 0.30187549679671233,
+      "learning_rate": 2.5469168900804292e-05,
+      "loss": 0.3621,
+      "step": 1347
+    },
+    {
+      "epoch": 1.6260554885404102,
+      "grad_norm": 0.35929811837951886,
+      "learning_rate": 2.5446827524575516e-05,
+      "loss": 0.3724,
+      "step": 1348
+    },
+    {
+      "epoch": 1.6272617611580218,
+      "grad_norm": 0.2852663310752726,
+      "learning_rate": 2.5424486148346737e-05,
+      "loss": 0.3538,
+      "step": 1349
+    },
+    {
+      "epoch": 1.6284680337756332,
+      "grad_norm": 0.33317652752316707,
+      "learning_rate": 2.5402144772117964e-05,
+      "loss": 0.4013,
+      "step": 1350
+    },
+    {
+      "epoch": 1.6296743063932448,
+      "grad_norm": 0.33926976945090287,
+      "learning_rate": 2.5379803395889192e-05,
+      "loss": 0.3608,
+      "step": 1351
+    },
+    {
+      "epoch": 1.6308805790108565,
+      "grad_norm": 0.29336918116275684,
+      "learning_rate": 2.535746201966041e-05,
+      "loss": 0.3624,
+      "step": 1352
+    },
+    {
+      "epoch": 1.632086851628468,
+      "grad_norm": 0.32537359239341435,
+      "learning_rate": 2.5335120643431636e-05,
+      "loss": 0.3667,
+      "step": 1353
+    },
+    {
+      "epoch": 1.6332931242460798,
+      "grad_norm": 0.31694181375333264,
+      "learning_rate": 2.5312779267202864e-05,
+      "loss": 0.3572,
+      "step": 1354
+    },
+    {
+      "epoch": 1.6344993968636912,
+      "grad_norm": 0.31129200279927227,
+      "learning_rate": 2.5290437890974085e-05,
+      "loss": 0.3566,
+      "step": 1355
+    },
+    {
+      "epoch": 1.6357056694813028,
+      "grad_norm": 0.3026625971829585,
+      "learning_rate": 2.526809651474531e-05,
+      "loss": 0.3511,
+      "step": 1356
+    },
+    {
+      "epoch": 1.6369119420989144,
+      "grad_norm": 0.2778204933785298,
+      "learning_rate": 2.5245755138516536e-05,
+      "loss": 0.3515,
+      "step": 1357
+    },
+    {
+      "epoch": 1.6381182147165259,
+      "grad_norm": 0.28112074622730776,
+      "learning_rate": 2.5223413762287757e-05,
+      "loss": 0.3467,
+      "step": 1358
+    },
+    {
+      "epoch": 1.6393244873341375,
+      "grad_norm": 0.2760083290003972,
+      "learning_rate": 2.520107238605898e-05,
+      "loss": 0.3431,
+      "step": 1359
+    },
+    {
+      "epoch": 1.6405307599517491,
+      "grad_norm": 0.2840407711954704,
+      "learning_rate": 2.517873100983021e-05,
+      "loss": 0.337,
+      "step": 1360
+    },
+    {
+      "epoch": 1.6417370325693605,
+      "grad_norm": 0.2665258347081383,
+      "learning_rate": 2.515638963360143e-05,
+      "loss": 0.3506,
+      "step": 1361
+    },
+    {
+      "epoch": 1.6429433051869724,
+      "grad_norm": 0.32503179208618316,
+      "learning_rate": 2.5134048257372657e-05,
+      "loss": 0.3471,
+      "step": 1362
+    },
+    {
+      "epoch": 1.6441495778045838,
+      "grad_norm": 0.30286222825612846,
+      "learning_rate": 2.511170688114388e-05,
+      "loss": 0.363,
+      "step": 1363
+    },
+    {
+      "epoch": 1.6453558504221955,
+      "grad_norm": 0.29668163519295065,
+      "learning_rate": 2.50893655049151e-05,
+      "loss": 0.3752,
+      "step": 1364
+    },
+    {
+      "epoch": 1.646562123039807,
+      "grad_norm": 0.34957635877197785,
+      "learning_rate": 2.506702412868633e-05,
+      "loss": 0.3631,
+      "step": 1365
+    },
+    {
+      "epoch": 1.6477683956574185,
+      "grad_norm": 0.28633563148090796,
+      "learning_rate": 2.5044682752457553e-05,
+      "loss": 0.368,
+      "step": 1366
+    },
+    {
+      "epoch": 1.6489746682750301,
+      "grad_norm": 0.3607510749471786,
+      "learning_rate": 2.5022341376228774e-05,
+      "loss": 0.3544,
+      "step": 1367
+    },
+    {
+      "epoch": 1.6501809408926418,
+      "grad_norm": 0.3295890075011459,
+      "learning_rate": 2.5e-05,
+      "loss": 0.349,
+      "step": 1368
+    },
+    {
+      "epoch": 1.6513872135102532,
+      "grad_norm": 0.34302365121553463,
+      "learning_rate": 2.4977658623771225e-05,
+      "loss": 0.3565,
+      "step": 1369
+    },
+    {
+      "epoch": 1.652593486127865,
+      "grad_norm": 0.30784854721347316,
+      "learning_rate": 2.495531724754245e-05,
+      "loss": 0.3598,
+      "step": 1370
+    },
+    {
+      "epoch": 1.6537997587454765,
+      "grad_norm": 0.3580153689284883,
+      "learning_rate": 2.4932975871313673e-05,
+      "loss": 0.3552,
+      "step": 1371
+    },
+    {
+      "epoch": 1.655006031363088,
+      "grad_norm": 0.32893911340548043,
+      "learning_rate": 2.4910634495084898e-05,
+      "loss": 0.3796,
+      "step": 1372
+    },
+    {
+      "epoch": 1.6562123039806997,
+      "grad_norm": 0.34401166770895486,
+      "learning_rate": 2.4888293118856125e-05,
+      "loss": 0.3465,
+      "step": 1373
+    },
+    {
+      "epoch": 1.6574185765983112,
+      "grad_norm": 0.34733562136621865,
+      "learning_rate": 2.4865951742627346e-05,
+      "loss": 0.3728,
+      "step": 1374
+    },
+    {
+      "epoch": 1.6586248492159228,
+      "grad_norm": 0.3415210711237987,
+      "learning_rate": 2.484361036639857e-05,
+      "loss": 0.3738,
+      "step": 1375
+    },
+    {
+      "epoch": 1.6598311218335344,
+      "grad_norm": 0.3564407234532172,
+      "learning_rate": 2.4821268990169797e-05,
+      "loss": 0.3554,
+      "step": 1376
+    },
+    {
+      "epoch": 1.6610373944511458,
+      "grad_norm": 0.2733304045447679,
+      "learning_rate": 2.479892761394102e-05,
+      "loss": 0.3588,
+      "step": 1377
+    },
+    {
+      "epoch": 1.6622436670687577,
+      "grad_norm": 0.35404845473786883,
+      "learning_rate": 2.4776586237712242e-05,
+      "loss": 0.3456,
+      "step": 1378
+    },
+    {
+      "epoch": 1.6634499396863691,
+      "grad_norm": 0.26342823241932845,
+      "learning_rate": 2.475424486148347e-05,
+      "loss": 0.3445,
+      "step": 1379
+    },
+    {
+      "epoch": 1.6646562123039808,
+      "grad_norm": 0.3775927461538472,
+      "learning_rate": 2.4731903485254694e-05,
+      "loss": 0.3562,
+      "step": 1380
+    },
+    {
+      "epoch": 1.6658624849215924,
+      "grad_norm": 0.2903268980143278,
+      "learning_rate": 2.4709562109025918e-05,
+      "loss": 0.3657,
+      "step": 1381
+    },
+    {
+      "epoch": 1.6670687575392038,
+      "grad_norm": 0.31050102348781466,
+      "learning_rate": 2.4687220732797142e-05,
+      "loss": 0.3697,
+      "step": 1382
+    },
+    {
+      "epoch": 1.6682750301568154,
+      "grad_norm": 0.3099732969543504,
+      "learning_rate": 2.4664879356568366e-05,
+      "loss": 0.369,
+      "step": 1383
+    },
+    {
+      "epoch": 1.669481302774427,
+      "grad_norm": 0.30160506449369273,
+      "learning_rate": 2.464253798033959e-05,
+      "loss": 0.3599,
+      "step": 1384
+    },
+    {
+      "epoch": 1.6706875753920385,
+      "grad_norm": 0.3537825638200836,
+      "learning_rate": 2.4620196604110814e-05,
+      "loss": 0.3727,
+      "step": 1385
+    },
+    {
+      "epoch": 1.6718938480096501,
+      "grad_norm": 0.29825631027148963,
+      "learning_rate": 2.4597855227882038e-05,
+      "loss": 0.3608,
+      "step": 1386
+    },
+    {
+      "epoch": 1.6731001206272618,
+      "grad_norm": 0.3244351587542817,
+      "learning_rate": 2.4575513851653262e-05,
+      "loss": 0.3644,
+      "step": 1387
+    },
+    {
+      "epoch": 1.6743063932448732,
+      "grad_norm": 0.2692023706580779,
+      "learning_rate": 2.4553172475424486e-05,
+      "loss": 0.3862,
+      "step": 1388
+    },
+    {
+      "epoch": 1.675512665862485,
+      "grad_norm": 0.33882016385543945,
+      "learning_rate": 2.453083109919571e-05,
+      "loss": 0.3516,
+      "step": 1389
+    },
+    {
+      "epoch": 1.6767189384800965,
+      "grad_norm": 0.3351467173827849,
+      "learning_rate": 2.4508489722966935e-05,
+      "loss": 0.3576,
+      "step": 1390
+    },
+    {
+      "epoch": 1.677925211097708,
+      "grad_norm": 0.29184859188818785,
+      "learning_rate": 2.4486148346738162e-05,
+      "loss": 0.349,
+      "step": 1391
+    },
+    {
+      "epoch": 1.6791314837153197,
+      "grad_norm": 0.4019244815049262,
+      "learning_rate": 2.4463806970509383e-05,
+      "loss": 0.3743,
+      "step": 1392
+    },
+    {
+      "epoch": 1.6803377563329311,
+      "grad_norm": 0.3100696829967641,
+      "learning_rate": 2.4441465594280607e-05,
+      "loss": 0.3493,
+      "step": 1393
+    },
+    {
+      "epoch": 1.6815440289505428,
+      "grad_norm": 0.37990510032559094,
+      "learning_rate": 2.4419124218051834e-05,
+      "loss": 0.3404,
+      "step": 1394
+    },
+    {
+      "epoch": 1.6827503015681544,
+      "grad_norm": 0.28147957637419957,
+      "learning_rate": 2.439678284182306e-05,
+      "loss": 0.3683,
+      "step": 1395
+    },
+    {
+      "epoch": 1.6839565741857658,
+      "grad_norm": 0.2944742115661062,
+      "learning_rate": 2.437444146559428e-05,
+      "loss": 0.3448,
+      "step": 1396
+    },
+    {
+      "epoch": 1.6851628468033777,
+      "grad_norm": 0.29809478369119624,
+      "learning_rate": 2.4352100089365507e-05,
+      "loss": 0.3471,
+      "step": 1397
+    },
+    {
+      "epoch": 1.686369119420989,
+      "grad_norm": 0.26699037994675995,
+      "learning_rate": 2.432975871313673e-05,
+      "loss": 0.3473,
+      "step": 1398
+    },
+    {
+      "epoch": 1.6875753920386007,
+      "grad_norm": 0.33536542802489927,
+      "learning_rate": 2.4307417336907955e-05,
+      "loss": 0.3771,
+      "step": 1399
+    },
+    {
+      "epoch": 1.6887816646562124,
+      "grad_norm": 0.300867897752028,
+      "learning_rate": 2.428507596067918e-05,
+      "loss": 0.3517,
+      "step": 1400
+    },
+    {
+      "epoch": 1.6899879372738238,
+      "grad_norm": 0.32504296903202723,
+      "learning_rate": 2.4262734584450403e-05,
+      "loss": 0.3487,
+      "step": 1401
+    },
+    {
+      "epoch": 1.6911942098914354,
+      "grad_norm": 0.276555968975667,
+      "learning_rate": 2.424039320822163e-05,
+      "loss": 0.3643,
+      "step": 1402
+    },
+    {
+      "epoch": 1.692400482509047,
+      "grad_norm": 0.3017611551074008,
+      "learning_rate": 2.421805183199285e-05,
+      "loss": 0.365,
+      "step": 1403
+    },
+    {
+      "epoch": 1.6936067551266585,
+      "grad_norm": 0.2672773162267323,
+      "learning_rate": 2.4195710455764075e-05,
+      "loss": 0.3522,
+      "step": 1404
+    },
+    {
+      "epoch": 1.6948130277442703,
+      "grad_norm": 0.31404716787538994,
+      "learning_rate": 2.4173369079535303e-05,
+      "loss": 0.3486,
+      "step": 1405
+    },
+    {
+      "epoch": 1.6960193003618818,
+      "grad_norm": 0.29569604012112444,
+      "learning_rate": 2.4151027703306527e-05,
+      "loss": 0.3664,
+      "step": 1406
+    },
+    {
+      "epoch": 1.6972255729794934,
+      "grad_norm": 0.27535191548620136,
+      "learning_rate": 2.4128686327077747e-05,
+      "loss": 0.3702,
+      "step": 1407
+    },
+    {
+      "epoch": 1.698431845597105,
+      "grad_norm": 0.260506311606101,
+      "learning_rate": 2.4106344950848975e-05,
+      "loss": 0.3678,
+      "step": 1408
+    },
+    {
+      "epoch": 1.6996381182147164,
+      "grad_norm": 0.30167357587290894,
+      "learning_rate": 2.40840035746202e-05,
+      "loss": 0.3625,
+      "step": 1409
+    },
+    {
+      "epoch": 1.700844390832328,
+      "grad_norm": 0.29521679266675266,
+      "learning_rate": 2.4061662198391423e-05,
+      "loss": 0.3501,
+      "step": 1410
+    },
+    {
+      "epoch": 1.7020506634499397,
+      "grad_norm": 0.31021002864555397,
+      "learning_rate": 2.4039320822162647e-05,
+      "loss": 0.3836,
+      "step": 1411
+    },
+    {
+      "epoch": 1.7032569360675511,
+      "grad_norm": 0.282035427062887,
+      "learning_rate": 2.401697944593387e-05,
+      "loss": 0.3655,
+      "step": 1412
+    },
+    {
+      "epoch": 1.704463208685163,
+      "grad_norm": 0.33351450606256217,
+      "learning_rate": 2.3994638069705095e-05,
+      "loss": 0.3887,
+      "step": 1413
+    },
+    {
+      "epoch": 1.7056694813027744,
+      "grad_norm": 0.283269591150281,
+      "learning_rate": 2.397229669347632e-05,
+      "loss": 0.3631,
+      "step": 1414
+    },
+    {
+      "epoch": 1.706875753920386,
+      "grad_norm": 0.2946135788809987,
+      "learning_rate": 2.3949955317247544e-05,
+      "loss": 0.3646,
+      "step": 1415
+    },
+    {
+      "epoch": 1.7080820265379977,
+      "grad_norm": 0.31742421947897204,
+      "learning_rate": 2.3927613941018768e-05,
+      "loss": 0.3691,
+      "step": 1416
+    },
+    {
+      "epoch": 1.709288299155609,
+      "grad_norm": 0.33683394127880345,
+      "learning_rate": 2.390527256478999e-05,
+      "loss": 0.3524,
+      "step": 1417
+    },
+    {
+      "epoch": 1.7104945717732207,
+      "grad_norm": 0.2731456928477606,
+      "learning_rate": 2.3882931188561216e-05,
+      "loss": 0.3401,
+      "step": 1418
+    },
+    {
+      "epoch": 1.7117008443908324,
+      "grad_norm": 0.3471290083601201,
+      "learning_rate": 2.386058981233244e-05,
+      "loss": 0.3553,
+      "step": 1419
+    },
+    {
+      "epoch": 1.7129071170084438,
+      "grad_norm": 0.3528869496369736,
+      "learning_rate": 2.3838248436103667e-05,
+      "loss": 0.324,
+      "step": 1420
+    },
+    {
+      "epoch": 1.7141133896260556,
+      "grad_norm": 0.2925982551552773,
+      "learning_rate": 2.3815907059874888e-05,
+      "loss": 0.3751,
+      "step": 1421
+    },
+    {
+      "epoch": 1.715319662243667,
+      "grad_norm": 0.36851576055596835,
+      "learning_rate": 2.3793565683646112e-05,
+      "loss": 0.3601,
+      "step": 1422
+    },
+    {
+      "epoch": 1.7165259348612787,
+      "grad_norm": 0.30235285015745955,
+      "learning_rate": 2.377122430741734e-05,
+      "loss": 0.3464,
+      "step": 1423
+    },
+    {
+      "epoch": 1.7177322074788903,
+      "grad_norm": 0.29874368477246893,
+      "learning_rate": 2.3748882931188564e-05,
+      "loss": 0.3626,
+      "step": 1424
+    },
+    {
+      "epoch": 1.7189384800965017,
+      "grad_norm": 0.3260772749237896,
+      "learning_rate": 2.3726541554959784e-05,
+      "loss": 0.3763,
+      "step": 1425
+    },
+    {
+      "epoch": 1.7201447527141134,
+      "grad_norm": 0.3341146497077902,
+      "learning_rate": 2.3704200178731012e-05,
+      "loss": 0.3521,
+      "step": 1426
+    },
+    {
+      "epoch": 1.721351025331725,
+      "grad_norm": 0.28157924981156457,
+      "learning_rate": 2.3681858802502236e-05,
+      "loss": 0.3622,
+      "step": 1427
+    },
+    {
+      "epoch": 1.7225572979493364,
+      "grad_norm": 0.3739702781502209,
+      "learning_rate": 2.365951742627346e-05,
+      "loss": 0.4076,
+      "step": 1428
+    },
+    {
+      "epoch": 1.7237635705669483,
+      "grad_norm": 0.3665408279544029,
+      "learning_rate": 2.3637176050044684e-05,
+      "loss": 0.3794,
+      "step": 1429
+    },
+    {
+      "epoch": 1.7249698431845597,
+      "grad_norm": 0.2935452940038669,
+      "learning_rate": 2.3614834673815908e-05,
+      "loss": 0.372,
+      "step": 1430
+    },
+    {
+      "epoch": 1.7261761158021713,
+      "grad_norm": 0.3244721429577489,
+      "learning_rate": 2.3592493297587132e-05,
+      "loss": 0.3609,
+      "step": 1431
+    },
+    {
+      "epoch": 1.727382388419783,
+      "grad_norm": 0.2918770790325217,
+      "learning_rate": 2.3570151921358356e-05,
+      "loss": 0.3673,
+      "step": 1432
+    },
+    {
+      "epoch": 1.7285886610373944,
+      "grad_norm": 0.2992819875803547,
+      "learning_rate": 2.354781054512958e-05,
+      "loss": 0.3571,
+      "step": 1433
+    },
+    {
+      "epoch": 1.729794933655006,
+      "grad_norm": 0.29707565974563616,
+      "learning_rate": 2.3525469168900805e-05,
+      "loss": 0.3901,
+      "step": 1434
+    },
+    {
+      "epoch": 1.7310012062726177,
+      "grad_norm": 0.32560925039212163,
+      "learning_rate": 2.3503127792672032e-05,
+      "loss": 0.3416,
+      "step": 1435
+    },
+    {
+      "epoch": 1.732207478890229,
+      "grad_norm": 0.2690827871626142,
+      "learning_rate": 2.3480786416443253e-05,
+      "loss": 0.3683,
+      "step": 1436
+    },
+    {
+      "epoch": 1.7334137515078407,
+      "grad_norm": 0.3060993415094302,
+      "learning_rate": 2.3458445040214477e-05,
+      "loss": 0.3587,
+      "step": 1437
+    },
+    {
+      "epoch": 1.7346200241254524,
+      "grad_norm": 0.2505412183050965,
+      "learning_rate": 2.3436103663985704e-05,
+      "loss": 0.3674,
+      "step": 1438
+    },
+    {
+      "epoch": 1.7358262967430638,
+      "grad_norm": 0.2653457404524692,
+      "learning_rate": 2.341376228775693e-05,
+      "loss": 0.3576,
+      "step": 1439
+    },
+    {
+      "epoch": 1.7370325693606756,
+      "grad_norm": 0.2771853156246843,
+      "learning_rate": 2.339142091152815e-05,
+      "loss": 0.3812,
+      "step": 1440
+    },
+    {
+      "epoch": 1.738238841978287,
+      "grad_norm": 0.2596381263418958,
+      "learning_rate": 2.3369079535299377e-05,
+      "loss": 0.359,
+      "step": 1441
+    },
+    {
+      "epoch": 1.7394451145958987,
+      "grad_norm": 0.28891917561566355,
+      "learning_rate": 2.33467381590706e-05,
+      "loss": 0.3586,
+      "step": 1442
+    },
+    {
+      "epoch": 1.7406513872135103,
+      "grad_norm": 0.27706301721058363,
+      "learning_rate": 2.332439678284182e-05,
+      "loss": 0.3461,
+      "step": 1443
+    },
+    {
+      "epoch": 1.7418576598311217,
+      "grad_norm": 0.29260701520716376,
+      "learning_rate": 2.330205540661305e-05,
+      "loss": 0.3525,
+      "step": 1444
+    },
+    {
+      "epoch": 1.7430639324487334,
+      "grad_norm": 0.26347051526297266,
+      "learning_rate": 2.3279714030384273e-05,
+      "loss": 0.3432,
+      "step": 1445
+    },
+    {
+      "epoch": 1.744270205066345,
+      "grad_norm": 0.28261285996650526,
+      "learning_rate": 2.3257372654155497e-05,
+      "loss": 0.3542,
+      "step": 1446
+    },
+    {
+      "epoch": 1.7454764776839564,
+      "grad_norm": 0.2802265796653411,
+      "learning_rate": 2.323503127792672e-05,
+      "loss": 0.3862,
+      "step": 1447
+    },
+    {
+      "epoch": 1.7466827503015683,
+      "grad_norm": 0.2602101357811866,
+      "learning_rate": 2.3212689901697945e-05,
+      "loss": 0.3601,
+      "step": 1448
+    },
+    {
+      "epoch": 1.7478890229191797,
+      "grad_norm": 0.2512605438487134,
+      "learning_rate": 2.319034852546917e-05,
+      "loss": 0.3591,
+      "step": 1449
+    },
+    {
+      "epoch": 1.7490952955367913,
+      "grad_norm": 0.24761708762692664,
+      "learning_rate": 2.3168007149240393e-05,
+      "loss": 0.3555,
+      "step": 1450
+    },
+    {
+      "epoch": 1.750301568154403,
+      "grad_norm": 0.3120056641615667,
+      "learning_rate": 2.3145665773011617e-05,
+      "loss": 0.3604,
+      "step": 1451
+    },
+    {
+      "epoch": 1.7515078407720144,
+      "grad_norm": 0.28674400953294193,
+      "learning_rate": 2.312332439678284e-05,
+      "loss": 0.3379,
+      "step": 1452
+    },
+    {
+      "epoch": 1.752714113389626,
+      "grad_norm": 0.25622102965400334,
+      "learning_rate": 2.310098302055407e-05,
+      "loss": 0.3593,
+      "step": 1453
+    },
+    {
+      "epoch": 1.7539203860072377,
+      "grad_norm": 0.35149468504184234,
+      "learning_rate": 2.307864164432529e-05,
+      "loss": 0.3641,
+      "step": 1454
+    },
+    {
+      "epoch": 1.755126658624849,
+      "grad_norm": 0.2850804434851281,
+      "learning_rate": 2.3056300268096514e-05,
+      "loss": 0.354,
+      "step": 1455
+    },
+    {
+      "epoch": 1.756332931242461,
+      "grad_norm": 0.3256278208536396,
+      "learning_rate": 2.303395889186774e-05,
+      "loss": 0.3588,
+      "step": 1456
+    },
+    {
+      "epoch": 1.7575392038600723,
+      "grad_norm": 0.39924915699406693,
+      "learning_rate": 2.3011617515638965e-05,
+      "loss": 0.3772,
+      "step": 1457
+    },
+    {
+      "epoch": 1.758745476477684,
+      "grad_norm": 0.3071970298535891,
+      "learning_rate": 2.2989276139410186e-05,
+      "loss": 0.3565,
+      "step": 1458
+    },
+    {
+      "epoch": 1.7599517490952956,
+      "grad_norm": 0.33320432329737215,
+      "learning_rate": 2.2966934763181414e-05,
+      "loss": 0.3496,
+      "step": 1459
+    },
+    {
+      "epoch": 1.761158021712907,
+      "grad_norm": 0.31284583547199185,
+      "learning_rate": 2.2944593386952638e-05,
+      "loss": 0.3627,
+      "step": 1460
+    },
+    {
+      "epoch": 1.7623642943305187,
+      "grad_norm": 0.3720962834352751,
+      "learning_rate": 2.2922252010723862e-05,
+      "loss": 0.3764,
+      "step": 1461
+    },
+    {
+      "epoch": 1.7635705669481303,
+      "grad_norm": 0.28507295604604466,
+      "learning_rate": 2.2899910634495086e-05,
+      "loss": 0.3605,
+      "step": 1462
+    },
+    {
+      "epoch": 1.7647768395657417,
+      "grad_norm": 0.35564273103301347,
+      "learning_rate": 2.287756925826631e-05,
+      "loss": 0.3675,
+      "step": 1463
+    },
+    {
+      "epoch": 1.7659831121833536,
+      "grad_norm": 0.28933965320173666,
+      "learning_rate": 2.2855227882037537e-05,
+      "loss": 0.3406,
+      "step": 1464
+    },
+    {
+      "epoch": 1.767189384800965,
+      "grad_norm": 0.3116459317914364,
+      "learning_rate": 2.2832886505808758e-05,
+      "loss": 0.355,
+      "step": 1465
+    },
+    {
+      "epoch": 1.7683956574185766,
+      "grad_norm": 0.27245033306948696,
+      "learning_rate": 2.2810545129579982e-05,
+      "loss": 0.3462,
+      "step": 1466
+    },
+    {
+      "epoch": 1.7696019300361883,
+      "grad_norm": 0.2869215998169608,
+      "learning_rate": 2.278820375335121e-05,
+      "loss": 0.3733,
+      "step": 1467
+    },
+    {
+      "epoch": 1.7708082026537997,
+      "grad_norm": 0.3113608243209837,
+      "learning_rate": 2.2765862377122434e-05,
+      "loss": 0.3506,
+      "step": 1468
+    },
+    {
+      "epoch": 1.7720144752714113,
+      "grad_norm": 0.3207842949128961,
+      "learning_rate": 2.2743521000893654e-05,
+      "loss": 0.3631,
+      "step": 1469
+    },
+    {
+      "epoch": 1.773220747889023,
+      "grad_norm": 0.29624424369525815,
+      "learning_rate": 2.2721179624664882e-05,
+      "loss": 0.3649,
+      "step": 1470
+    },
+    {
+      "epoch": 1.7744270205066344,
+      "grad_norm": 0.3303763513390022,
+      "learning_rate": 2.2698838248436106e-05,
+      "loss": 0.3633,
+      "step": 1471
+    },
+    {
+      "epoch": 1.7756332931242462,
+      "grad_norm": 0.3227006129000047,
+      "learning_rate": 2.267649687220733e-05,
+      "loss": 0.3556,
+      "step": 1472
+    },
+    {
+      "epoch": 1.7768395657418576,
+      "grad_norm": 0.3393310341911691,
+      "learning_rate": 2.2654155495978554e-05,
+      "loss": 0.3631,
+      "step": 1473
+    },
+    {
+      "epoch": 1.7780458383594693,
+      "grad_norm": 0.3529552361183484,
+      "learning_rate": 2.2631814119749778e-05,
+      "loss": 0.3546,
+      "step": 1474
+    },
+    {
+      "epoch": 1.779252110977081,
+      "grad_norm": 0.31525744469559364,
+      "learning_rate": 2.2609472743521002e-05,
+      "loss": 0.343,
+      "step": 1475
+    },
+    {
+      "epoch": 1.7804583835946923,
+      "grad_norm": 0.3163718158924185,
+      "learning_rate": 2.2587131367292226e-05,
+      "loss": 0.3692,
+      "step": 1476
+    },
+    {
+      "epoch": 1.781664656212304,
+      "grad_norm": 0.32561762908782643,
+      "learning_rate": 2.256478999106345e-05,
+      "loss": 0.364,
+      "step": 1477
+    },
+    {
+      "epoch": 1.7828709288299156,
+      "grad_norm": 0.27131184421929944,
+      "learning_rate": 2.2542448614834675e-05,
+      "loss": 0.3518,
+      "step": 1478
+    },
+    {
+      "epoch": 1.784077201447527,
+      "grad_norm": 0.2811976684931462,
+      "learning_rate": 2.25201072386059e-05,
+      "loss": 0.3497,
+      "step": 1479
+    },
+    {
+      "epoch": 1.7852834740651389,
+      "grad_norm": 0.27610607298870293,
+      "learning_rate": 2.2497765862377123e-05,
+      "loss": 0.3659,
+      "step": 1480
+    },
+    {
+      "epoch": 1.7864897466827503,
+      "grad_norm": 0.3164778858830116,
+      "learning_rate": 2.2475424486148347e-05,
+      "loss": 0.3393,
+      "step": 1481
+    },
+    {
+      "epoch": 1.787696019300362,
+      "grad_norm": 0.3440205777362051,
+      "learning_rate": 2.2453083109919574e-05,
+      "loss": 0.3766,
+      "step": 1482
+    },
+    {
+      "epoch": 1.7889022919179736,
+      "grad_norm": 0.29128877422979677,
+      "learning_rate": 2.2430741733690795e-05,
+      "loss": 0.346,
+      "step": 1483
+    },
+    {
+      "epoch": 1.790108564535585,
+      "grad_norm": 0.2997901031965129,
+      "learning_rate": 2.240840035746202e-05,
+      "loss": 0.3439,
+      "step": 1484
+    },
+    {
+      "epoch": 1.7913148371531966,
+      "grad_norm": 0.31411546633151705,
+      "learning_rate": 2.2386058981233247e-05,
+      "loss": 0.3642,
+      "step": 1485
+    },
+    {
+      "epoch": 1.7925211097708083,
+      "grad_norm": 0.2954643791110668,
+      "learning_rate": 2.236371760500447e-05,
+      "loss": 0.3816,
+      "step": 1486
+    },
+    {
+      "epoch": 1.7937273823884197,
+      "grad_norm": 0.3080224579128579,
+      "learning_rate": 2.234137622877569e-05,
+      "loss": 0.3842,
+      "step": 1487
+    },
+    {
+      "epoch": 1.7949336550060315,
+      "grad_norm": 0.25606441087244014,
+      "learning_rate": 2.231903485254692e-05,
+      "loss": 0.3758,
+      "step": 1488
+    },
+    {
+      "epoch": 1.796139927623643,
+      "grad_norm": 0.3150542138350638,
+      "learning_rate": 2.2296693476318143e-05,
+      "loss": 0.3692,
+      "step": 1489
+    },
+    {
+      "epoch": 1.7973462002412546,
+      "grad_norm": 0.26939799555931804,
+      "learning_rate": 2.2274352100089367e-05,
+      "loss": 0.3304,
+      "step": 1490
+    },
+    {
+      "epoch": 1.7985524728588662,
+      "grad_norm": 0.2880747492641311,
+      "learning_rate": 2.225201072386059e-05,
+      "loss": 0.3635,
+      "step": 1491
+    },
+    {
+      "epoch": 1.7997587454764776,
+      "grad_norm": 0.27399275585186184,
+      "learning_rate": 2.2229669347631815e-05,
+      "loss": 0.3713,
+      "step": 1492
+    },
+    {
+      "epoch": 1.8009650180940893,
+      "grad_norm": 0.2457588382002816,
+      "learning_rate": 2.220732797140304e-05,
+      "loss": 0.3594,
+      "step": 1493
+    },
+    {
+      "epoch": 1.802171290711701,
+      "grad_norm": 0.2821381653581902,
+      "learning_rate": 2.2184986595174263e-05,
+      "loss": 0.3579,
+      "step": 1494
+    },
+    {
+      "epoch": 1.8033775633293123,
+      "grad_norm": 0.26906678345859164,
+      "learning_rate": 2.2162645218945488e-05,
+      "loss": 0.3704,
+      "step": 1495
+    },
+    {
+      "epoch": 1.804583835946924,
+      "grad_norm": 0.28860333792004156,
+      "learning_rate": 2.214030384271671e-05,
+      "loss": 0.3828,
+      "step": 1496
+    },
+    {
+      "epoch": 1.8057901085645356,
+      "grad_norm": 0.2941151374887319,
+      "learning_rate": 2.211796246648794e-05,
+      "loss": 0.3559,
+      "step": 1497
+    },
+    {
+      "epoch": 1.806996381182147,
+      "grad_norm": 0.33755789851920004,
+      "learning_rate": 2.209562109025916e-05,
+      "loss": 0.3444,
+      "step": 1498
+    },
+    {
+      "epoch": 1.8082026537997589,
+      "grad_norm": 0.3153818457382501,
+      "learning_rate": 2.2073279714030384e-05,
+      "loss": 0.3315,
+      "step": 1499
+    },
+    {
+      "epoch": 1.8094089264173703,
+      "grad_norm": 0.27122267879244044,
+      "learning_rate": 2.205093833780161e-05,
+      "loss": 0.3321,
+      "step": 1500
+    },
+    {
+      "epoch": 1.810615199034982,
+      "grad_norm": 0.36408451541394626,
+      "learning_rate": 2.2028596961572835e-05,
+      "loss": 0.3577,
+      "step": 1501
+    },
+    {
+      "epoch": 1.8118214716525936,
+      "grad_norm": 0.3048366478040056,
+      "learning_rate": 2.2006255585344056e-05,
+      "loss": 0.363,
+      "step": 1502
+    },
+    {
+      "epoch": 1.813027744270205,
+      "grad_norm": 0.289384365836586,
+      "learning_rate": 2.1983914209115284e-05,
+      "loss": 0.3476,
+      "step": 1503
+    },
+    {
+      "epoch": 1.8142340168878166,
+      "grad_norm": 0.3119269637358593,
+      "learning_rate": 2.1961572832886508e-05,
+      "loss": 0.3406,
+      "step": 1504
+    },
+    {
+      "epoch": 1.8154402895054282,
+      "grad_norm": 0.3357944937521883,
+      "learning_rate": 2.193923145665773e-05,
+      "loss": 0.3678,
+      "step": 1505
+    },
+    {
+      "epoch": 1.8166465621230397,
+      "grad_norm": 0.3277816414905101,
+      "learning_rate": 2.1916890080428956e-05,
+      "loss": 0.3711,
+      "step": 1506
+    },
+    {
+      "epoch": 1.8178528347406515,
+      "grad_norm": 0.29928477659654407,
+      "learning_rate": 2.189454870420018e-05,
+      "loss": 0.3408,
+      "step": 1507
+    },
+    {
+      "epoch": 1.819059107358263,
+      "grad_norm": 0.29004649198090976,
+      "learning_rate": 2.1872207327971404e-05,
+      "loss": 0.3461,
+      "step": 1508
+    },
+    {
+      "epoch": 1.8202653799758746,
+      "grad_norm": 0.25481916180153474,
+      "learning_rate": 2.1849865951742628e-05,
+      "loss": 0.3554,
+      "step": 1509
+    },
+    {
+      "epoch": 1.8214716525934862,
+      "grad_norm": 0.3206381774906527,
+      "learning_rate": 2.1827524575513852e-05,
+      "loss": 0.3654,
+      "step": 1510
+    },
+    {
+      "epoch": 1.8226779252110976,
+      "grad_norm": 0.27464477113928465,
+      "learning_rate": 2.1805183199285076e-05,
+      "loss": 0.3686,
+      "step": 1511
+    },
+    {
+      "epoch": 1.8238841978287093,
+      "grad_norm": 0.2978283923505883,
+      "learning_rate": 2.17828418230563e-05,
+      "loss": 0.3809,
+      "step": 1512
+    },
+    {
+      "epoch": 1.825090470446321,
+      "grad_norm": 0.3018397915218147,
+      "learning_rate": 2.1760500446827525e-05,
+      "loss": 0.3604,
+      "step": 1513
+    },
+    {
+      "epoch": 1.8262967430639323,
+      "grad_norm": 0.32003093006503486,
+      "learning_rate": 2.173815907059875e-05,
+      "loss": 0.3503,
+      "step": 1514
+    },
+    {
+      "epoch": 1.8275030156815442,
+      "grad_norm": 0.26271586653183376,
+      "learning_rate": 2.1715817694369976e-05,
+      "loss": 0.3615,
+      "step": 1515
+    },
+    {
+      "epoch": 1.8287092882991556,
+      "grad_norm": 0.34758559596839306,
+      "learning_rate": 2.1693476318141197e-05,
+      "loss": 0.3679,
+      "step": 1516
+    },
+    {
+      "epoch": 1.8299155609167672,
+      "grad_norm": 0.28304707959172154,
+      "learning_rate": 2.167113494191242e-05,
+      "loss": 0.3559,
+      "step": 1517
+    },
+    {
+      "epoch": 1.8311218335343789,
+      "grad_norm": 0.30716182879162085,
+      "learning_rate": 2.164879356568365e-05,
+      "loss": 0.3705,
+      "step": 1518
+    },
+    {
+      "epoch": 1.8323281061519903,
+      "grad_norm": 0.3272103551415873,
+      "learning_rate": 2.1626452189454872e-05,
+      "loss": 0.3557,
+      "step": 1519
+    },
+    {
+      "epoch": 1.833534378769602,
+      "grad_norm": 0.2856664835977563,
+      "learning_rate": 2.1604110813226093e-05,
+      "loss": 0.354,
+      "step": 1520
+    },
+    {
+      "epoch": 1.8347406513872135,
+      "grad_norm": 0.2995470532032037,
+      "learning_rate": 2.158176943699732e-05,
+      "loss": 0.3542,
+      "step": 1521
+    },
+    {
+      "epoch": 1.835946924004825,
+      "grad_norm": 0.3198503981887153,
+      "learning_rate": 2.1559428060768545e-05,
+      "loss": 0.3645,
+      "step": 1522
+    },
+    {
+      "epoch": 1.8371531966224368,
+      "grad_norm": 0.32754560915720893,
+      "learning_rate": 2.153708668453977e-05,
+      "loss": 0.3551,
+      "step": 1523
+    },
+    {
+      "epoch": 1.8383594692400482,
+      "grad_norm": 0.26948473809778084,
+      "learning_rate": 2.1514745308310993e-05,
+      "loss": 0.3604,
+      "step": 1524
+    },
+    {
+      "epoch": 1.8395657418576599,
+      "grad_norm": 0.2926288255208596,
+      "learning_rate": 2.1492403932082217e-05,
+      "loss": 0.3455,
+      "step": 1525
+    },
+    {
+      "epoch": 1.8407720144752715,
+      "grad_norm": 0.34964762071460354,
+      "learning_rate": 2.1470062555853444e-05,
+      "loss": 0.3513,
+      "step": 1526
+    },
+    {
+      "epoch": 1.841978287092883,
+      "grad_norm": 0.2592691589602463,
+      "learning_rate": 2.1447721179624665e-05,
+      "loss": 0.3591,
+      "step": 1527
+    },
+    {
+      "epoch": 1.8431845597104946,
+      "grad_norm": 0.309967263730365,
+      "learning_rate": 2.142537980339589e-05,
+      "loss": 0.3876,
+      "step": 1528
+    },
+    {
+      "epoch": 1.8443908323281062,
+      "grad_norm": 0.28029506099906043,
+      "learning_rate": 2.1403038427167117e-05,
+      "loss": 0.3583,
+      "step": 1529
+    },
+    {
+      "epoch": 1.8455971049457176,
+      "grad_norm": 0.26805452677117375,
+      "learning_rate": 2.138069705093834e-05,
+      "loss": 0.3466,
+      "step": 1530
+    },
+    {
+      "epoch": 1.8468033775633295,
+      "grad_norm": 0.2994299990919435,
+      "learning_rate": 2.135835567470956e-05,
+      "loss": 0.356,
+      "step": 1531
+    },
+    {
+      "epoch": 1.8480096501809409,
+      "grad_norm": 0.2731266768809756,
+      "learning_rate": 2.133601429848079e-05,
+      "loss": 0.363,
+      "step": 1532
+    },
+    {
+      "epoch": 1.8492159227985525,
+      "grad_norm": 0.32771389667540246,
+      "learning_rate": 2.1313672922252013e-05,
+      "loss": 0.3568,
+      "step": 1533
+    },
+    {
+      "epoch": 1.8504221954161642,
+      "grad_norm": 0.3573609136221111,
+      "learning_rate": 2.1291331546023234e-05,
+      "loss": 0.3957,
+      "step": 1534
+    },
+    {
+      "epoch": 1.8516284680337756,
+      "grad_norm": 0.3162680716287011,
+      "learning_rate": 2.126899016979446e-05,
+      "loss": 0.3535,
+      "step": 1535
+    },
+    {
+      "epoch": 1.8528347406513872,
+      "grad_norm": 0.295455790097431,
+      "learning_rate": 2.1246648793565685e-05,
+      "loss": 0.3545,
+      "step": 1536
+    },
+    {
+      "epoch": 1.8540410132689988,
+      "grad_norm": 0.26045858915660014,
+      "learning_rate": 2.122430741733691e-05,
+      "loss": 0.3456,
+      "step": 1537
+    },
+    {
+      "epoch": 1.8552472858866103,
+      "grad_norm": 0.31789979304930077,
+      "learning_rate": 2.1201966041108133e-05,
+      "loss": 0.3811,
+      "step": 1538
+    },
+    {
+      "epoch": 1.8564535585042221,
+      "grad_norm": 0.2458230363735759,
+      "learning_rate": 2.1179624664879358e-05,
+      "loss": 0.3498,
+      "step": 1539
+    },
+    {
+      "epoch": 1.8576598311218335,
+      "grad_norm": 0.31583664687268614,
+      "learning_rate": 2.115728328865058e-05,
+      "loss": 0.3609,
+      "step": 1540
+    },
+    {
+      "epoch": 1.8588661037394452,
+      "grad_norm": 0.2777923125909325,
+      "learning_rate": 2.1134941912421806e-05,
+      "loss": 0.3599,
+      "step": 1541
+    },
+    {
+      "epoch": 1.8600723763570568,
+      "grad_norm": 0.2996705803905586,
+      "learning_rate": 2.111260053619303e-05,
+      "loss": 0.3542,
+      "step": 1542
+    },
+    {
+      "epoch": 1.8612786489746682,
+      "grad_norm": 0.23278188821193466,
+      "learning_rate": 2.1090259159964254e-05,
+      "loss": 0.3408,
+      "step": 1543
+    },
+    {
+      "epoch": 1.8624849215922799,
+      "grad_norm": 0.32721942024267,
+      "learning_rate": 2.106791778373548e-05,
+      "loss": 0.3465,
+      "step": 1544
+    },
+    {
+      "epoch": 1.8636911942098915,
+      "grad_norm": 0.29620060616140187,
+      "learning_rate": 2.1045576407506702e-05,
+      "loss": 0.3483,
+      "step": 1545
+    },
+    {
+      "epoch": 1.864897466827503,
+      "grad_norm": 0.2833126537587724,
+      "learning_rate": 2.1023235031277926e-05,
+      "loss": 0.363,
+      "step": 1546
+    },
+    {
+      "epoch": 1.8661037394451148,
+      "grad_norm": 0.27966605358297153,
+      "learning_rate": 2.1000893655049154e-05,
+      "loss": 0.383,
+      "step": 1547
+    },
+    {
+      "epoch": 1.8673100120627262,
+      "grad_norm": 0.2953507697070746,
+      "learning_rate": 2.0978552278820378e-05,
+      "loss": 0.352,
+      "step": 1548
+    },
+    {
+      "epoch": 1.8685162846803376,
+      "grad_norm": 0.3118445272043699,
+      "learning_rate": 2.09562109025916e-05,
+      "loss": 0.361,
+      "step": 1549
+    },
+    {
+      "epoch": 1.8697225572979495,
+      "grad_norm": 0.27569695544577993,
+      "learning_rate": 2.0933869526362826e-05,
+      "loss": 0.3705,
+      "step": 1550
+    },
+    {
+      "epoch": 1.8709288299155609,
+      "grad_norm": 0.2931446668444685,
+      "learning_rate": 2.091152815013405e-05,
+      "loss": 0.3517,
+      "step": 1551
+    },
+    {
+      "epoch": 1.8721351025331725,
+      "grad_norm": 0.28741574092692435,
+      "learning_rate": 2.0889186773905274e-05,
+      "loss": 0.3482,
+      "step": 1552
+    },
+    {
+      "epoch": 1.8733413751507841,
+      "grad_norm": 0.33952287545856796,
+      "learning_rate": 2.0866845397676498e-05,
+      "loss": 0.3646,
+      "step": 1553
+    },
+    {
+      "epoch": 1.8745476477683956,
+      "grad_norm": 0.25889542403249843,
+      "learning_rate": 2.0844504021447722e-05,
+      "loss": 0.3625,
+      "step": 1554
+    },
+    {
+      "epoch": 1.8757539203860072,
+      "grad_norm": 0.2779529963292999,
+      "learning_rate": 2.0822162645218946e-05,
+      "loss": 0.3483,
+      "step": 1555
+    },
+    {
+      "epoch": 1.8769601930036188,
+      "grad_norm": 0.27128966596000303,
+      "learning_rate": 2.079982126899017e-05,
+      "loss": 0.352,
+      "step": 1556
+    },
+    {
+      "epoch": 1.8781664656212302,
+      "grad_norm": 0.2902910048214581,
+      "learning_rate": 2.0777479892761395e-05,
+      "loss": 0.3723,
+      "step": 1557
+    },
+    {
+      "epoch": 1.879372738238842,
+      "grad_norm": 0.27973862245194236,
+      "learning_rate": 2.075513851653262e-05,
+      "loss": 0.3743,
+      "step": 1558
+    },
+    {
+      "epoch": 1.8805790108564535,
+      "grad_norm": 0.30609968166421897,
+      "learning_rate": 2.0732797140303846e-05,
+      "loss": 0.3731,
+      "step": 1559
+    },
+    {
+      "epoch": 1.8817852834740652,
+      "grad_norm": 0.2728618671983489,
+      "learning_rate": 2.0710455764075067e-05,
+      "loss": 0.3617,
+      "step": 1560
+    },
+    {
+      "epoch": 1.8829915560916768,
+      "grad_norm": 0.2543254562059058,
+      "learning_rate": 2.068811438784629e-05,
+      "loss": 0.3436,
+      "step": 1561
+    },
+    {
+      "epoch": 1.8841978287092882,
+      "grad_norm": 0.24841468175025513,
+      "learning_rate": 2.066577301161752e-05,
+      "loss": 0.3522,
+      "step": 1562
+    },
+    {
+      "epoch": 1.8854041013268998,
+      "grad_norm": 0.31768402839745774,
+      "learning_rate": 2.0643431635388742e-05,
+      "loss": 0.3736,
+      "step": 1563
+    },
+    {
+      "epoch": 1.8866103739445115,
+      "grad_norm": 0.3191797777615365,
+      "learning_rate": 2.0621090259159963e-05,
+      "loss": 0.3766,
+      "step": 1564
+    },
+    {
+      "epoch": 1.887816646562123,
+      "grad_norm": 0.272658079681039,
+      "learning_rate": 2.059874888293119e-05,
+      "loss": 0.3611,
+      "step": 1565
+    },
+    {
+      "epoch": 1.8890229191797347,
+      "grad_norm": 0.2622171628236883,
+      "learning_rate": 2.0576407506702415e-05,
+      "loss": 0.3444,
+      "step": 1566
+    },
+    {
+      "epoch": 1.8902291917973462,
+      "grad_norm": 0.31135421295399496,
+      "learning_rate": 2.0554066130473635e-05,
+      "loss": 0.3534,
+      "step": 1567
+    },
+    {
+      "epoch": 1.8914354644149578,
+      "grad_norm": 0.2777552393153191,
+      "learning_rate": 2.0531724754244863e-05,
+      "loss": 0.3451,
+      "step": 1568
+    },
+    {
+      "epoch": 1.8926417370325694,
+      "grad_norm": 0.2808254408248436,
+      "learning_rate": 2.0509383378016087e-05,
+      "loss": 0.3711,
+      "step": 1569
+    },
+    {
+      "epoch": 1.8938480096501809,
+      "grad_norm": 0.2629707060553636,
+      "learning_rate": 2.048704200178731e-05,
+      "loss": 0.3536,
+      "step": 1570
+    },
+    {
+      "epoch": 1.8950542822677925,
+      "grad_norm": 0.284968406394402,
+      "learning_rate": 2.0464700625558535e-05,
+      "loss": 0.3478,
+      "step": 1571
+    },
+    {
+      "epoch": 1.8962605548854041,
+      "grad_norm": 0.277964729242244,
+      "learning_rate": 2.044235924932976e-05,
+      "loss": 0.3522,
+      "step": 1572
+    },
+    {
+      "epoch": 1.8974668275030155,
+      "grad_norm": 0.28027713264841336,
+      "learning_rate": 2.0420017873100983e-05,
+      "loss": 0.3589,
+      "step": 1573
+    },
+    {
+      "epoch": 1.8986731001206274,
+      "grad_norm": 0.3050543089008853,
+      "learning_rate": 2.0397676496872207e-05,
+      "loss": 0.3737,
+      "step": 1574
+    },
+    {
+      "epoch": 1.8998793727382388,
+      "grad_norm": 0.333573721833168,
+      "learning_rate": 2.037533512064343e-05,
+      "loss": 0.3723,
+      "step": 1575
+    },
+    {
+      "epoch": 1.9010856453558505,
+      "grad_norm": 0.27558632494530466,
+      "learning_rate": 2.0352993744414656e-05,
+      "loss": 0.347,
+      "step": 1576
+    },
+    {
+      "epoch": 1.902291917973462,
+      "grad_norm": 0.2749496186635863,
+      "learning_rate": 2.0330652368185883e-05,
+      "loss": 0.3584,
+      "step": 1577
+    },
+    {
+      "epoch": 1.9034981905910735,
+      "grad_norm": 0.27623036751877184,
+      "learning_rate": 2.0308310991957104e-05,
+      "loss": 0.3671,
+      "step": 1578
+    },
+    {
+      "epoch": 1.9047044632086851,
+      "grad_norm": 0.35924451163298154,
+      "learning_rate": 2.0285969615728328e-05,
+      "loss": 0.3551,
+      "step": 1579
+    },
+    {
+      "epoch": 1.9059107358262968,
+      "grad_norm": 0.2809181345145692,
+      "learning_rate": 2.0263628239499555e-05,
+      "loss": 0.3502,
+      "step": 1580
+    },
+    {
+      "epoch": 1.9071170084439082,
+      "grad_norm": 0.25879235314860666,
+      "learning_rate": 2.024128686327078e-05,
+      "loss": 0.3405,
+      "step": 1581
+    },
+    {
+      "epoch": 1.90832328106152,
+      "grad_norm": 0.2849966409585352,
+      "learning_rate": 2.0218945487042e-05,
+      "loss": 0.3378,
+      "step": 1582
+    },
+    {
+      "epoch": 1.9095295536791315,
+      "grad_norm": 0.3372371097579718,
+      "learning_rate": 2.0196604110813228e-05,
+      "loss": 0.3517,
+      "step": 1583
+    },
+    {
+      "epoch": 1.910735826296743,
+      "grad_norm": 0.25632078946719644,
+      "learning_rate": 2.0174262734584452e-05,
+      "loss": 0.3443,
+      "step": 1584
+    },
+    {
+      "epoch": 1.9119420989143547,
+      "grad_norm": 0.26956940258980094,
+      "learning_rate": 2.0151921358355676e-05,
+      "loss": 0.3401,
+      "step": 1585
+    },
+    {
+      "epoch": 1.9131483715319662,
+      "grad_norm": 0.26515149174118896,
+      "learning_rate": 2.01295799821269e-05,
+      "loss": 0.3523,
+      "step": 1586
+    },
+    {
+      "epoch": 1.9143546441495778,
+      "grad_norm": 0.2613492785001853,
+      "learning_rate": 2.0107238605898124e-05,
+      "loss": 0.3739,
+      "step": 1587
+    },
+    {
+      "epoch": 1.9155609167671894,
+      "grad_norm": 0.23950143457883746,
+      "learning_rate": 2.008489722966935e-05,
+      "loss": 0.3343,
+      "step": 1588
+    },
+    {
+      "epoch": 1.9167671893848008,
+      "grad_norm": 0.30117376449701433,
+      "learning_rate": 2.0062555853440572e-05,
+      "loss": 0.3709,
+      "step": 1589
+    },
+    {
+      "epoch": 1.9179734620024127,
+      "grad_norm": 0.2593161624873031,
+      "learning_rate": 2.0040214477211796e-05,
+      "loss": 0.3452,
+      "step": 1590
+    },
+    {
+      "epoch": 1.9191797346200241,
+      "grad_norm": 0.2407521061083514,
+      "learning_rate": 2.0017873100983024e-05,
+      "loss": 0.3735,
+      "step": 1591
+    },
+    {
+      "epoch": 1.9203860072376358,
+      "grad_norm": 0.26195291267250836,
+      "learning_rate": 1.9995531724754248e-05,
+      "loss": 0.3509,
+      "step": 1592
+    },
+    {
+      "epoch": 1.9215922798552474,
+      "grad_norm": 0.297432231962128,
+      "learning_rate": 1.997319034852547e-05,
+      "loss": 0.3442,
+      "step": 1593
+    },
+    {
+      "epoch": 1.9227985524728588,
+      "grad_norm": 0.2399537948274586,
+      "learning_rate": 1.9950848972296696e-05,
+      "loss": 0.3471,
+      "step": 1594
+    },
+    {
+      "epoch": 1.9240048250904704,
+      "grad_norm": 0.28295475054379304,
+      "learning_rate": 1.992850759606792e-05,
+      "loss": 0.3615,
+      "step": 1595
+    },
+    {
+      "epoch": 1.925211097708082,
+      "grad_norm": 0.2894150589928256,
+      "learning_rate": 1.990616621983914e-05,
+      "loss": 0.3744,
+      "step": 1596
+    },
+    {
+      "epoch": 1.9264173703256935,
+      "grad_norm": 0.27749447033260927,
+      "learning_rate": 1.9883824843610368e-05,
+      "loss": 0.363,
+      "step": 1597
+    },
+    {
+      "epoch": 1.9276236429433053,
+      "grad_norm": 0.28672944617778745,
+      "learning_rate": 1.9861483467381592e-05,
+      "loss": 0.3747,
+      "step": 1598
+    },
+    {
+      "epoch": 1.9288299155609168,
+      "grad_norm": 0.2847048913976895,
+      "learning_rate": 1.9839142091152816e-05,
+      "loss": 0.3405,
+      "step": 1599
+    },
+    {
+      "epoch": 1.9300361881785284,
+      "grad_norm": 0.28224450714587446,
+      "learning_rate": 1.981680071492404e-05,
+      "loss": 0.369,
+      "step": 1600
+    },
+    {
+      "epoch": 1.93124246079614,
+      "grad_norm": 0.251007009158062,
+      "learning_rate": 1.9794459338695265e-05,
+      "loss": 0.3583,
+      "step": 1601
+    },
+    {
+      "epoch": 1.9324487334137515,
+      "grad_norm": 0.2700503054912305,
+      "learning_rate": 1.977211796246649e-05,
+      "loss": 0.3487,
+      "step": 1602
+    },
+    {
+      "epoch": 1.933655006031363,
+      "grad_norm": 0.26601499415202334,
+      "learning_rate": 1.9749776586237713e-05,
+      "loss": 0.356,
+      "step": 1603
+    },
+    {
+      "epoch": 1.9348612786489747,
+      "grad_norm": 0.26517746425234484,
+      "learning_rate": 1.9727435210008937e-05,
+      "loss": 0.3751,
+      "step": 1604
+    },
+    {
+      "epoch": 1.9360675512665861,
+      "grad_norm": 0.3075644038798218,
+      "learning_rate": 1.970509383378016e-05,
+      "loss": 0.3865,
+      "step": 1605
+    },
+    {
+      "epoch": 1.9372738238841978,
+      "grad_norm": 0.2748549582323955,
+      "learning_rate": 1.968275245755139e-05,
+      "loss": 0.3537,
+      "step": 1606
+    },
+    {
+      "epoch": 1.9384800965018094,
+      "grad_norm": 0.2595112570755194,
+      "learning_rate": 1.966041108132261e-05,
+      "loss": 0.3576,
+      "step": 1607
+    },
+    {
+      "epoch": 1.9396863691194208,
+      "grad_norm": 0.2512201382268573,
+      "learning_rate": 1.9638069705093833e-05,
+      "loss": 0.3417,
+      "step": 1608
+    },
+    {
+      "epoch": 1.9408926417370327,
+      "grad_norm": 0.3057741641131344,
+      "learning_rate": 1.961572832886506e-05,
+      "loss": 0.3531,
+      "step": 1609
+    },
+    {
+      "epoch": 1.942098914354644,
+      "grad_norm": 0.2995428068053764,
+      "learning_rate": 1.9593386952636285e-05,
+      "loss": 0.3577,
+      "step": 1610
+    },
+    {
+      "epoch": 1.9433051869722557,
+      "grad_norm": 0.29519013006895595,
+      "learning_rate": 1.9571045576407505e-05,
+      "loss": 0.3645,
+      "step": 1611
+    },
+    {
+      "epoch": 1.9445114595898674,
+      "grad_norm": 0.31079328371619636,
+      "learning_rate": 1.9548704200178733e-05,
+      "loss": 0.3765,
+      "step": 1612
+    },
+    {
+      "epoch": 1.9457177322074788,
+      "grad_norm": 0.31234042894752406,
+      "learning_rate": 1.9526362823949957e-05,
+      "loss": 0.3538,
+      "step": 1613
+    },
+    {
+      "epoch": 1.9469240048250904,
+      "grad_norm": 0.28659000754418135,
+      "learning_rate": 1.950402144772118e-05,
+      "loss": 0.3731,
+      "step": 1614
+    },
+    {
+      "epoch": 1.948130277442702,
+      "grad_norm": 0.264814616626347,
+      "learning_rate": 1.9481680071492405e-05,
+      "loss": 0.3659,
+      "step": 1615
+    },
+    {
+      "epoch": 1.9493365500603135,
+      "grad_norm": 0.24366644264153936,
+      "learning_rate": 1.945933869526363e-05,
+      "loss": 0.3673,
+      "step": 1616
+    },
+    {
+      "epoch": 1.9505428226779253,
+      "grad_norm": 0.272103653808616,
+      "learning_rate": 1.9436997319034853e-05,
+      "loss": 0.3507,
+      "step": 1617
+    },
+    {
+      "epoch": 1.9517490952955368,
+      "grad_norm": 0.30326138219958865,
+      "learning_rate": 1.9414655942806077e-05,
+      "loss": 0.3673,
+      "step": 1618
+    },
+    {
+      "epoch": 1.9529553679131484,
+      "grad_norm": 0.2471852651757173,
+      "learning_rate": 1.93923145665773e-05,
+      "loss": 0.3429,
+      "step": 1619
+    },
+    {
+      "epoch": 1.95416164053076,
+      "grad_norm": 0.29823029929658096,
+      "learning_rate": 1.9369973190348526e-05,
+      "loss": 0.3498,
+      "step": 1620
+    },
+    {
+      "epoch": 1.9553679131483714,
+      "grad_norm": 0.25618786812647487,
+      "learning_rate": 1.9347631814119753e-05,
+      "loss": 0.3602,
+      "step": 1621
+    },
+    {
+      "epoch": 1.956574185765983,
+      "grad_norm": 0.2626954040000901,
+      "learning_rate": 1.9325290437890974e-05,
+      "loss": 0.3548,
+      "step": 1622
+    },
+    {
+      "epoch": 1.9577804583835947,
+      "grad_norm": 0.30278639551876474,
+      "learning_rate": 1.9302949061662198e-05,
+      "loss": 0.3602,
+      "step": 1623
+    },
+    {
+      "epoch": 1.9589867310012061,
+      "grad_norm": 0.32422885066788054,
+      "learning_rate": 1.9280607685433425e-05,
+      "loss": 0.3536,
+      "step": 1624
+    },
+    {
+      "epoch": 1.960193003618818,
+      "grad_norm": 0.25780893788890336,
+      "learning_rate": 1.9258266309204646e-05,
+      "loss": 0.3448,
+      "step": 1625
+    },
+    {
+      "epoch": 1.9613992762364294,
+      "grad_norm": 0.2701194515692144,
+      "learning_rate": 1.923592493297587e-05,
+      "loss": 0.3685,
+      "step": 1626
+    },
+    {
+      "epoch": 1.962605548854041,
+      "grad_norm": 0.2637933151314994,
+      "learning_rate": 1.9213583556747098e-05,
+      "loss": 0.3515,
+      "step": 1627
+    },
+    {
+      "epoch": 1.9638118214716527,
+      "grad_norm": 0.2934963853487863,
+      "learning_rate": 1.9191242180518322e-05,
+      "loss": 0.3742,
+      "step": 1628
+    },
+    {
+      "epoch": 1.965018094089264,
+      "grad_norm": 0.22910011332182156,
+      "learning_rate": 1.9168900804289542e-05,
+      "loss": 0.3647,
+      "step": 1629
+    },
+    {
+      "epoch": 1.9662243667068757,
+      "grad_norm": 0.3393223178431687,
+      "learning_rate": 1.914655942806077e-05,
+      "loss": 0.3479,
+      "step": 1630
+    },
+    {
+      "epoch": 1.9674306393244874,
+      "grad_norm": 0.27703033721697085,
+      "learning_rate": 1.9124218051831994e-05,
+      "loss": 0.3809,
+      "step": 1631
+    },
+    {
+      "epoch": 1.9686369119420988,
+      "grad_norm": 0.3208522052589231,
+      "learning_rate": 1.9101876675603218e-05,
+      "loss": 0.3681,
+      "step": 1632
+    },
+    {
+      "epoch": 1.9698431845597106,
+      "grad_norm": 0.27699992821466946,
+      "learning_rate": 1.9079535299374442e-05,
+      "loss": 0.3516,
+      "step": 1633
+    },
+    {
+      "epoch": 1.971049457177322,
+      "grad_norm": 0.2712403728322839,
+      "learning_rate": 1.9057193923145666e-05,
+      "loss": 0.3423,
+      "step": 1634
+    },
+    {
+      "epoch": 1.9722557297949337,
+      "grad_norm": 0.2855332321421069,
+      "learning_rate": 1.903485254691689e-05,
+      "loss": 0.3318,
+      "step": 1635
+    },
+    {
+      "epoch": 1.9734620024125453,
+      "grad_norm": 0.2815389249603608,
+      "learning_rate": 1.9012511170688114e-05,
+      "loss": 0.3459,
+      "step": 1636
+    },
+    {
+      "epoch": 1.9746682750301567,
+      "grad_norm": 0.28182396775800955,
+      "learning_rate": 1.899016979445934e-05,
+      "loss": 0.3384,
+      "step": 1637
+    },
+    {
+      "epoch": 1.9758745476477684,
+      "grad_norm": 0.2718840146128206,
+      "learning_rate": 1.8967828418230563e-05,
+      "loss": 0.3609,
+      "step": 1638
+    },
+    {
+      "epoch": 1.97708082026538,
+      "grad_norm": 0.2591947787566494,
+      "learning_rate": 1.894548704200179e-05,
+      "loss": 0.3428,
+      "step": 1639
+    },
+    {
+      "epoch": 1.9782870928829914,
+      "grad_norm": 0.32383105243939264,
+      "learning_rate": 1.892314566577301e-05,
+      "loss": 0.3647,
+      "step": 1640
+    },
+    {
+      "epoch": 1.9794933655006033,
+      "grad_norm": 0.2563990858898975,
+      "learning_rate": 1.8900804289544235e-05,
+      "loss": 0.3484,
+      "step": 1641
+    },
+    {
+      "epoch": 1.9806996381182147,
+      "grad_norm": 0.3105355894461767,
+      "learning_rate": 1.8878462913315462e-05,
+      "loss": 0.3959,
+      "step": 1642
+    },
+    {
+      "epoch": 1.9819059107358263,
+      "grad_norm": 0.301088060007025,
+      "learning_rate": 1.8856121537086686e-05,
+      "loss": 0.3333,
+      "step": 1643
+    },
+    {
+      "epoch": 1.983112183353438,
+      "grad_norm": 0.28554630545995097,
+      "learning_rate": 1.8833780160857907e-05,
+      "loss": 0.3487,
+      "step": 1644
+    },
+    {
+      "epoch": 1.9843184559710494,
+      "grad_norm": 0.2602079469425586,
+      "learning_rate": 1.8811438784629135e-05,
+      "loss": 0.349,
+      "step": 1645
+    },
+    {
+      "epoch": 1.985524728588661,
+      "grad_norm": 0.2923498070190884,
+      "learning_rate": 1.878909740840036e-05,
+      "loss": 0.374,
+      "step": 1646
+    },
+    {
+      "epoch": 1.9867310012062727,
+      "grad_norm": 0.3420070267678662,
+      "learning_rate": 1.8766756032171583e-05,
+      "loss": 0.363,
+      "step": 1647
+    },
+    {
+      "epoch": 1.987937273823884,
+      "grad_norm": 0.2706072187767052,
+      "learning_rate": 1.8744414655942807e-05,
+      "loss": 0.3674,
+      "step": 1648
+    },
+    {
+      "epoch": 1.989143546441496,
+      "grad_norm": 0.33498382557114115,
+      "learning_rate": 1.872207327971403e-05,
+      "loss": 0.3637,
+      "step": 1649
+    },
+    {
+      "epoch": 1.9903498190591074,
+      "grad_norm": 0.3450578881904668,
+      "learning_rate": 1.869973190348526e-05,
+      "loss": 0.3471,
+      "step": 1650
+    },
+    {
+      "epoch": 1.991556091676719,
+      "grad_norm": 0.28093156857550655,
+      "learning_rate": 1.867739052725648e-05,
+      "loss": 0.373,
+      "step": 1651
+    },
+    {
+      "epoch": 1.9927623642943306,
+      "grad_norm": 0.33731372039057506,
+      "learning_rate": 1.8655049151027703e-05,
+      "loss": 0.352,
+      "step": 1652
+    },
+    {
+      "epoch": 1.993968636911942,
+      "grad_norm": 0.32637042812980754,
+      "learning_rate": 1.863270777479893e-05,
+      "loss": 0.3553,
+      "step": 1653
+    },
+    {
+      "epoch": 1.9951749095295537,
+      "grad_norm": 0.2924533816211366,
+      "learning_rate": 1.8610366398570155e-05,
+      "loss": 0.3422,
+      "step": 1654
+    },
+    {
+      "epoch": 1.9963811821471653,
+      "grad_norm": 0.30816680687954695,
+      "learning_rate": 1.8588025022341376e-05,
+      "loss": 0.3514,
+      "step": 1655
+    },
+    {
+      "epoch": 1.9975874547647767,
+      "grad_norm": 0.37414553666596123,
+      "learning_rate": 1.8565683646112603e-05,
+      "loss": 0.3375,
+      "step": 1656
+    },
+    {
+      "epoch": 1.9987937273823886,
+      "grad_norm": 0.25177696486470447,
+      "learning_rate": 1.8543342269883827e-05,
+      "loss": 0.3447,
+      "step": 1657
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.2985722047195102,
+      "learning_rate": 1.8521000893655048e-05,
+      "loss": 0.3362,
+      "step": 1658
+    },
+    {
+      "epoch": 2.0012062726176114,
+      "grad_norm": 0.3582654343499216,
+      "learning_rate": 1.8498659517426275e-05,
+      "loss": 0.2966,
+      "step": 1659
+    },
+    {
+      "epoch": 2.0024125452352233,
+      "grad_norm": 0.2510003107801239,
+      "learning_rate": 1.84763181411975e-05,
+      "loss": 0.3027,
+      "step": 1660
+    },
+    {
+      "epoch": 2.0036188178528347,
+      "grad_norm": 0.31848482820403823,
+      "learning_rate": 1.8453976764968723e-05,
+      "loss": 0.3072,
+      "step": 1661
+    },
+    {
+      "epoch": 2.004825090470446,
+      "grad_norm": 0.28121589676814773,
+      "learning_rate": 1.8431635388739948e-05,
+      "loss": 0.2799,
+      "step": 1662
+    },
+    {
+      "epoch": 2.006031363088058,
+      "grad_norm": 0.2790990286631224,
+      "learning_rate": 1.840929401251117e-05,
+      "loss": 0.295,
+      "step": 1663
+    },
+    {
+      "epoch": 2.0072376357056694,
+      "grad_norm": 0.29716421480885014,
+      "learning_rate": 1.8386952636282396e-05,
+      "loss": 0.2905,
+      "step": 1664
+    },
+    {
+      "epoch": 2.0084439083232812,
+      "grad_norm": 0.2876586964710323,
+      "learning_rate": 1.836461126005362e-05,
+      "loss": 0.2738,
+      "step": 1665
+    },
+    {
+      "epoch": 2.0096501809408926,
+      "grad_norm": 0.2847467005141539,
+      "learning_rate": 1.8342269883824844e-05,
+      "loss": 0.2744,
+      "step": 1666
+    },
+    {
+      "epoch": 2.010856453558504,
+      "grad_norm": 0.2848323602535861,
+      "learning_rate": 1.8319928507596068e-05,
+      "loss": 0.2937,
+      "step": 1667
+    },
+    {
+      "epoch": 2.012062726176116,
+      "grad_norm": 0.30619210279161174,
+      "learning_rate": 1.8297587131367295e-05,
+      "loss": 0.2945,
+      "step": 1668
+    },
+    {
+      "epoch": 2.0132689987937273,
+      "grad_norm": 0.29414910099799263,
+      "learning_rate": 1.8275245755138516e-05,
+      "loss": 0.2946,
+      "step": 1669
+    },
+    {
+      "epoch": 2.0144752714113388,
+      "grad_norm": 0.289304762173054,
+      "learning_rate": 1.825290437890974e-05,
+      "loss": 0.284,
+      "step": 1670
+    },
+    {
+      "epoch": 2.0156815440289506,
+      "grad_norm": 0.28576033145670227,
+      "learning_rate": 1.8230563002680968e-05,
+      "loss": 0.2867,
+      "step": 1671
+    },
+    {
+      "epoch": 2.016887816646562,
+      "grad_norm": 0.32373169860293655,
+      "learning_rate": 1.8208221626452192e-05,
+      "loss": 0.2995,
+      "step": 1672
+    },
+    {
+      "epoch": 2.018094089264174,
+      "grad_norm": 0.23638681779954598,
+      "learning_rate": 1.8185880250223413e-05,
+      "loss": 0.2878,
+      "step": 1673
+    },
+    {
+      "epoch": 2.0193003618817853,
+      "grad_norm": 0.28453090920113555,
+      "learning_rate": 1.816353887399464e-05,
+      "loss": 0.2945,
+      "step": 1674
+    },
+    {
+      "epoch": 2.0205066344993967,
+      "grad_norm": 0.2549255381981737,
+      "learning_rate": 1.8141197497765864e-05,
+      "loss": 0.2966,
+      "step": 1675
+    },
+    {
+      "epoch": 2.0217129071170086,
+      "grad_norm": 0.29462585796542506,
+      "learning_rate": 1.8118856121537088e-05,
+      "loss": 0.2864,
+      "step": 1676
+    },
+    {
+      "epoch": 2.02291917973462,
+      "grad_norm": 0.26548129330247616,
+      "learning_rate": 1.8096514745308312e-05,
+      "loss": 0.3023,
+      "step": 1677
+    },
+    {
+      "epoch": 2.0241254523522314,
+      "grad_norm": 0.31332310597104845,
+      "learning_rate": 1.8074173369079536e-05,
+      "loss": 0.2946,
+      "step": 1678
+    },
+    {
+      "epoch": 2.0253317249698433,
+      "grad_norm": 0.27338445045211085,
+      "learning_rate": 1.805183199285076e-05,
+      "loss": 0.2804,
+      "step": 1679
+    },
+    {
+      "epoch": 2.0265379975874547,
+      "grad_norm": 0.3118425725530232,
+      "learning_rate": 1.8029490616621985e-05,
+      "loss": 0.3007,
+      "step": 1680
+    },
+    {
+      "epoch": 2.0277442702050665,
+      "grad_norm": 0.27714195106597606,
+      "learning_rate": 1.800714924039321e-05,
+      "loss": 0.2884,
+      "step": 1681
+    },
+    {
+      "epoch": 2.028950542822678,
+      "grad_norm": 0.31958347812909327,
+      "learning_rate": 1.7984807864164433e-05,
+      "loss": 0.2999,
+      "step": 1682
+    },
+    {
+      "epoch": 2.0301568154402894,
+      "grad_norm": 0.29186659149763977,
+      "learning_rate": 1.796246648793566e-05,
+      "loss": 0.2916,
+      "step": 1683
+    },
+    {
+      "epoch": 2.0313630880579012,
+      "grad_norm": 0.2733656352448462,
+      "learning_rate": 1.794012511170688e-05,
+      "loss": 0.2765,
+      "step": 1684
+    },
+    {
+      "epoch": 2.0325693606755126,
+      "grad_norm": 0.28574192979858715,
+      "learning_rate": 1.7917783735478105e-05,
+      "loss": 0.311,
+      "step": 1685
+    },
+    {
+      "epoch": 2.033775633293124,
+      "grad_norm": 0.2991351897521293,
+      "learning_rate": 1.7895442359249332e-05,
+      "loss": 0.2852,
+      "step": 1686
+    },
+    {
+      "epoch": 2.034981905910736,
+      "grad_norm": 0.24236890788237953,
+      "learning_rate": 1.7873100983020553e-05,
+      "loss": 0.2931,
+      "step": 1687
+    },
+    {
+      "epoch": 2.0361881785283473,
+      "grad_norm": 0.2551685310046532,
+      "learning_rate": 1.7850759606791777e-05,
+      "loss": 0.2908,
+      "step": 1688
+    },
+    {
+      "epoch": 2.037394451145959,
+      "grad_norm": 0.2549058353261533,
+      "learning_rate": 1.7828418230563005e-05,
+      "loss": 0.2823,
+      "step": 1689
+    },
+    {
+      "epoch": 2.0386007237635706,
+      "grad_norm": 0.24168430653360778,
+      "learning_rate": 1.780607685433423e-05,
+      "loss": 0.2804,
+      "step": 1690
+    },
+    {
+      "epoch": 2.039806996381182,
+      "grad_norm": 0.24712540293178345,
+      "learning_rate": 1.778373547810545e-05,
+      "loss": 0.2889,
+      "step": 1691
+    },
+    {
+      "epoch": 2.041013268998794,
+      "grad_norm": 0.2584950258915421,
+      "learning_rate": 1.7761394101876677e-05,
+      "loss": 0.2979,
+      "step": 1692
+    },
+    {
+      "epoch": 2.0422195416164053,
+      "grad_norm": 0.2547855415817537,
+      "learning_rate": 1.77390527256479e-05,
+      "loss": 0.2936,
+      "step": 1693
+    },
+    {
+      "epoch": 2.0434258142340167,
+      "grad_norm": 0.24820792443315048,
+      "learning_rate": 1.7716711349419125e-05,
+      "loss": 0.2929,
+      "step": 1694
+    },
+    {
+      "epoch": 2.0446320868516286,
+      "grad_norm": 0.2600712986700368,
+      "learning_rate": 1.769436997319035e-05,
+      "loss": 0.2912,
+      "step": 1695
+    },
+    {
+      "epoch": 2.04583835946924,
+      "grad_norm": 0.2369033927502002,
+      "learning_rate": 1.7672028596961573e-05,
+      "loss": 0.2935,
+      "step": 1696
+    },
+    {
+      "epoch": 2.047044632086852,
+      "grad_norm": 0.24572060560713774,
+      "learning_rate": 1.7649687220732797e-05,
+      "loss": 0.2728,
+      "step": 1697
+    },
+    {
+      "epoch": 2.0482509047044632,
+      "grad_norm": 0.31718716534120606,
+      "learning_rate": 1.762734584450402e-05,
+      "loss": 0.2952,
+      "step": 1698
+    },
+    {
+      "epoch": 2.0494571773220747,
+      "grad_norm": 0.23868855770487613,
+      "learning_rate": 1.7605004468275246e-05,
+      "loss": 0.2796,
+      "step": 1699
+    },
+    {
+      "epoch": 2.0506634499396865,
+      "grad_norm": 0.2810826971186264,
+      "learning_rate": 1.758266309204647e-05,
+      "loss": 0.2919,
+      "step": 1700
+    },
+    {
+      "epoch": 2.051869722557298,
+      "grad_norm": 0.27057773712748967,
+      "learning_rate": 1.7560321715817697e-05,
+      "loss": 0.2868,
+      "step": 1701
+    },
+    {
+      "epoch": 2.0530759951749094,
+      "grad_norm": 0.23055869199867401,
+      "learning_rate": 1.7537980339588918e-05,
+      "loss": 0.2811,
+      "step": 1702
+    },
+    {
+      "epoch": 2.054282267792521,
+      "grad_norm": 0.24322673852618046,
+      "learning_rate": 1.7515638963360142e-05,
+      "loss": 0.2859,
+      "step": 1703
+    },
+    {
+      "epoch": 2.0554885404101326,
+      "grad_norm": 0.2750690299702216,
+      "learning_rate": 1.749329758713137e-05,
+      "loss": 0.2966,
+      "step": 1704
+    },
+    {
+      "epoch": 2.0566948130277445,
+      "grad_norm": 0.6931475541633155,
+      "learning_rate": 1.7470956210902594e-05,
+      "loss": 0.3079,
+      "step": 1705
+    },
+    {
+      "epoch": 2.057901085645356,
+      "grad_norm": 0.21463239940499598,
+      "learning_rate": 1.7448614834673814e-05,
+      "loss": 0.283,
+      "step": 1706
+    },
+    {
+      "epoch": 2.0591073582629673,
+      "grad_norm": 0.2902830958637698,
+      "learning_rate": 1.742627345844504e-05,
+      "loss": 0.2766,
+      "step": 1707
+    },
+    {
+      "epoch": 2.060313630880579,
+      "grad_norm": 0.2605225165985984,
+      "learning_rate": 1.7403932082216266e-05,
+      "loss": 0.3049,
+      "step": 1708
+    },
+    {
+      "epoch": 2.0615199034981906,
+      "grad_norm": 0.24065535000122262,
+      "learning_rate": 1.738159070598749e-05,
+      "loss": 0.2764,
+      "step": 1709
+    },
+    {
+      "epoch": 2.062726176115802,
+      "grad_norm": 0.268617788452376,
+      "learning_rate": 1.7359249329758714e-05,
+      "loss": 0.3052,
+      "step": 1710
+    },
+    {
+      "epoch": 2.063932448733414,
+      "grad_norm": 0.24609202820968598,
+      "learning_rate": 1.7336907953529938e-05,
+      "loss": 0.3027,
+      "step": 1711
+    },
+    {
+      "epoch": 2.0651387213510253,
+      "grad_norm": 0.24966047333966127,
+      "learning_rate": 1.7314566577301166e-05,
+      "loss": 0.2841,
+      "step": 1712
+    },
+    {
+      "epoch": 2.066344993968637,
+      "grad_norm": 0.23861829254548647,
+      "learning_rate": 1.7292225201072386e-05,
+      "loss": 0.2844,
+      "step": 1713
+    },
+    {
+      "epoch": 2.0675512665862485,
+      "grad_norm": 0.21580214631270747,
+      "learning_rate": 1.726988382484361e-05,
+      "loss": 0.2854,
+      "step": 1714
+    },
+    {
+      "epoch": 2.06875753920386,
+      "grad_norm": 0.23851902544110398,
+      "learning_rate": 1.7247542448614838e-05,
+      "loss": 0.2991,
+      "step": 1715
+    },
+    {
+      "epoch": 2.069963811821472,
+      "grad_norm": 0.26704566565966975,
+      "learning_rate": 1.7225201072386062e-05,
+      "loss": 0.3057,
+      "step": 1716
+    },
+    {
+      "epoch": 2.0711700844390832,
+      "grad_norm": 0.2921229741945177,
+      "learning_rate": 1.7202859696157283e-05,
+      "loss": 0.2876,
+      "step": 1717
+    },
+    {
+      "epoch": 2.0723763570566947,
+      "grad_norm": 0.2704074358886806,
+      "learning_rate": 1.718051831992851e-05,
+      "loss": 0.2915,
+      "step": 1718
+    },
+    {
+      "epoch": 2.0735826296743065,
+      "grad_norm": 0.26102313901994467,
+      "learning_rate": 1.7158176943699734e-05,
+      "loss": 0.2728,
+      "step": 1719
+    },
+    {
+      "epoch": 2.074788902291918,
+      "grad_norm": 0.31354244476135346,
+      "learning_rate": 1.7135835567470955e-05,
+      "loss": 0.2925,
+      "step": 1720
+    },
+    {
+      "epoch": 2.0759951749095293,
+      "grad_norm": 0.3101725582423187,
+      "learning_rate": 1.7113494191242182e-05,
+      "loss": 0.2978,
+      "step": 1721
+    },
+    {
+      "epoch": 2.077201447527141,
+      "grad_norm": 0.2503609628190097,
+      "learning_rate": 1.7091152815013406e-05,
+      "loss": 0.2994,
+      "step": 1722
+    },
+    {
+      "epoch": 2.0784077201447526,
+      "grad_norm": 0.3010569194193657,
+      "learning_rate": 1.706881143878463e-05,
+      "loss": 0.2724,
+      "step": 1723
+    },
+    {
+      "epoch": 2.0796139927623645,
+      "grad_norm": 0.3554100305822572,
+      "learning_rate": 1.7046470062555855e-05,
+      "loss": 0.286,
+      "step": 1724
+    },
+    {
+      "epoch": 2.080820265379976,
+      "grad_norm": 0.28020867058791404,
+      "learning_rate": 1.702412868632708e-05,
+      "loss": 0.2924,
+      "step": 1725
+    },
+    {
+      "epoch": 2.0820265379975873,
+      "grad_norm": 0.27399523525779423,
+      "learning_rate": 1.7001787310098303e-05,
+      "loss": 0.2816,
+      "step": 1726
+    },
+    {
+      "epoch": 2.083232810615199,
+      "grad_norm": 0.30202972438037023,
+      "learning_rate": 1.6979445933869527e-05,
+      "loss": 0.2883,
+      "step": 1727
+    },
+    {
+      "epoch": 2.0844390832328106,
+      "grad_norm": 0.25985583955795966,
+      "learning_rate": 1.695710455764075e-05,
+      "loss": 0.2931,
+      "step": 1728
+    },
+    {
+      "epoch": 2.085645355850422,
+      "grad_norm": 0.23082414332906082,
+      "learning_rate": 1.6934763181411975e-05,
+      "loss": 0.2891,
+      "step": 1729
+    },
+    {
+      "epoch": 2.086851628468034,
+      "grad_norm": 0.2637973752139721,
+      "learning_rate": 1.6912421805183202e-05,
+      "loss": 0.291,
+      "step": 1730
+    },
+    {
+      "epoch": 2.0880579010856453,
+      "grad_norm": 0.31611307943527334,
+      "learning_rate": 1.6890080428954423e-05,
+      "loss": 0.2871,
+      "step": 1731
+    },
+    {
+      "epoch": 2.089264173703257,
+      "grad_norm": 0.27342534759657844,
+      "learning_rate": 1.6867739052725647e-05,
+      "loss": 0.2922,
+      "step": 1732
+    },
+    {
+      "epoch": 2.0904704463208685,
+      "grad_norm": 0.23859203460193187,
+      "learning_rate": 1.6845397676496875e-05,
+      "loss": 0.2916,
+      "step": 1733
+    },
+    {
+      "epoch": 2.09167671893848,
+      "grad_norm": 0.2894846792042188,
+      "learning_rate": 1.68230563002681e-05,
+      "loss": 0.2882,
+      "step": 1734
+    },
+    {
+      "epoch": 2.092882991556092,
+      "grad_norm": 0.2880969256927605,
+      "learning_rate": 1.680071492403932e-05,
+      "loss": 0.3016,
+      "step": 1735
+    },
+    {
+      "epoch": 2.0940892641737032,
+      "grad_norm": 0.24513385363687332,
+      "learning_rate": 1.6778373547810547e-05,
+      "loss": 0.2812,
+      "step": 1736
+    },
+    {
+      "epoch": 2.0952955367913146,
+      "grad_norm": 0.254098444251658,
+      "learning_rate": 1.675603217158177e-05,
+      "loss": 0.2919,
+      "step": 1737
+    },
+    {
+      "epoch": 2.0965018094089265,
+      "grad_norm": 0.2633220043801049,
+      "learning_rate": 1.6733690795352995e-05,
+      "loss": 0.2848,
+      "step": 1738
+    },
+    {
+      "epoch": 2.097708082026538,
+      "grad_norm": 0.2940832498604142,
+      "learning_rate": 1.671134941912422e-05,
+      "loss": 0.292,
+      "step": 1739
+    },
+    {
+      "epoch": 2.0989143546441498,
+      "grad_norm": 0.24954777585396712,
+      "learning_rate": 1.6689008042895443e-05,
+      "loss": 0.282,
+      "step": 1740
+    },
+    {
+      "epoch": 2.100120627261761,
+      "grad_norm": 0.248278540621884,
+      "learning_rate": 1.6666666666666667e-05,
+      "loss": 0.2894,
+      "step": 1741
+    },
+    {
+      "epoch": 2.1013268998793726,
+      "grad_norm": 0.2811884680572496,
+      "learning_rate": 1.664432529043789e-05,
+      "loss": 0.2865,
+      "step": 1742
+    },
+    {
+      "epoch": 2.1025331724969845,
+      "grad_norm": 0.2522636866338131,
+      "learning_rate": 1.6621983914209116e-05,
+      "loss": 0.2933,
+      "step": 1743
+    },
+    {
+      "epoch": 2.103739445114596,
+      "grad_norm": 0.2770783365680852,
+      "learning_rate": 1.659964253798034e-05,
+      "loss": 0.2738,
+      "step": 1744
+    },
+    {
+      "epoch": 2.1049457177322073,
+      "grad_norm": 0.25751293080579474,
+      "learning_rate": 1.6577301161751567e-05,
+      "loss": 0.2831,
+      "step": 1745
+    },
+    {
+      "epoch": 2.106151990349819,
+      "grad_norm": 0.22569893040934857,
+      "learning_rate": 1.6554959785522788e-05,
+      "loss": 0.2807,
+      "step": 1746
+    },
+    {
+      "epoch": 2.1073582629674306,
+      "grad_norm": 0.2676508071107855,
+      "learning_rate": 1.6532618409294012e-05,
+      "loss": 0.2827,
+      "step": 1747
+    },
+    {
+      "epoch": 2.1085645355850424,
+      "grad_norm": 0.22742787799565736,
+      "learning_rate": 1.651027703306524e-05,
+      "loss": 0.2729,
+      "step": 1748
+    },
+    {
+      "epoch": 2.109770808202654,
+      "grad_norm": 0.24790912426013842,
+      "learning_rate": 1.648793565683646e-05,
+      "loss": 0.2699,
+      "step": 1749
+    },
+    {
+      "epoch": 2.1109770808202653,
+      "grad_norm": 0.2916106304957534,
+      "learning_rate": 1.6465594280607684e-05,
+      "loss": 0.2862,
+      "step": 1750
+    },
+    {
+      "epoch": 2.112183353437877,
+      "grad_norm": 0.2401141131980366,
+      "learning_rate": 1.6443252904378912e-05,
+      "loss": 0.2824,
+      "step": 1751
+    },
+    {
+      "epoch": 2.1133896260554885,
+      "grad_norm": 0.26241788076180006,
+      "learning_rate": 1.6420911528150136e-05,
+      "loss": 0.2767,
+      "step": 1752
+    },
+    {
+      "epoch": 2.1145958986731,
+      "grad_norm": 0.2954786816300289,
+      "learning_rate": 1.6398570151921357e-05,
+      "loss": 0.2959,
+      "step": 1753
+    },
+    {
+      "epoch": 2.115802171290712,
+      "grad_norm": 0.22908967311620249,
+      "learning_rate": 1.6376228775692584e-05,
+      "loss": 0.2758,
+      "step": 1754
+    },
+    {
+      "epoch": 2.117008443908323,
+      "grad_norm": 0.2413217007927014,
+      "learning_rate": 1.6353887399463808e-05,
+      "loss": 0.2757,
+      "step": 1755
+    },
+    {
+      "epoch": 2.118214716525935,
+      "grad_norm": 0.2773658127485155,
+      "learning_rate": 1.6331546023235032e-05,
+      "loss": 0.2948,
+      "step": 1756
+    },
+    {
+      "epoch": 2.1194209891435465,
+      "grad_norm": 0.22417241060997734,
+      "learning_rate": 1.6309204647006256e-05,
+      "loss": 0.2968,
+      "step": 1757
+    },
+    {
+      "epoch": 2.120627261761158,
+      "grad_norm": 0.2553527879768229,
+      "learning_rate": 1.628686327077748e-05,
+      "loss": 0.2884,
+      "step": 1758
+    },
+    {
+      "epoch": 2.1218335343787698,
+      "grad_norm": 0.25633754052987345,
+      "learning_rate": 1.6264521894548704e-05,
+      "loss": 0.2814,
+      "step": 1759
+    },
+    {
+      "epoch": 2.123039806996381,
+      "grad_norm": 0.2425672162311776,
+      "learning_rate": 1.624218051831993e-05,
+      "loss": 0.3019,
+      "step": 1760
+    },
+    {
+      "epoch": 2.1242460796139926,
+      "grad_norm": 0.22861022137422327,
+      "learning_rate": 1.6219839142091153e-05,
+      "loss": 0.2925,
+      "step": 1761
+    },
+    {
+      "epoch": 2.1254523522316044,
+      "grad_norm": 0.2840288790695893,
+      "learning_rate": 1.6197497765862377e-05,
+      "loss": 0.3063,
+      "step": 1762
+    },
+    {
+      "epoch": 2.126658624849216,
+      "grad_norm": 0.2537508729451685,
+      "learning_rate": 1.6175156389633604e-05,
+      "loss": 0.2933,
+      "step": 1763
+    },
+    {
+      "epoch": 2.1278648974668277,
+      "grad_norm": 0.24195914928488352,
+      "learning_rate": 1.6152815013404825e-05,
+      "loss": 0.2993,
+      "step": 1764
+    },
+    {
+      "epoch": 2.129071170084439,
+      "grad_norm": 0.27873420100035506,
+      "learning_rate": 1.613047363717605e-05,
+      "loss": 0.3141,
+      "step": 1765
+    },
+    {
+      "epoch": 2.1302774427020505,
+      "grad_norm": 0.268235827640432,
+      "learning_rate": 1.6108132260947276e-05,
+      "loss": 0.2931,
+      "step": 1766
+    },
+    {
+      "epoch": 2.1314837153196624,
+      "grad_norm": 0.2191238701653833,
+      "learning_rate": 1.60857908847185e-05,
+      "loss": 0.2732,
+      "step": 1767
+    },
+    {
+      "epoch": 2.132689987937274,
+      "grad_norm": 0.25437125752546474,
+      "learning_rate": 1.606344950848972e-05,
+      "loss": 0.2967,
+      "step": 1768
+    },
+    {
+      "epoch": 2.1338962605548852,
+      "grad_norm": 0.26624909826119847,
+      "learning_rate": 1.604110813226095e-05,
+      "loss": 0.2803,
+      "step": 1769
+    },
+    {
+      "epoch": 2.135102533172497,
+      "grad_norm": 0.24857107954142815,
+      "learning_rate": 1.6018766756032173e-05,
+      "loss": 0.2844,
+      "step": 1770
+    },
+    {
+      "epoch": 2.1363088057901085,
+      "grad_norm": 0.25654303344242035,
+      "learning_rate": 1.5996425379803397e-05,
+      "loss": 0.2763,
+      "step": 1771
+    },
+    {
+      "epoch": 2.13751507840772,
+      "grad_norm": 0.2725659911600105,
+      "learning_rate": 1.597408400357462e-05,
+      "loss": 0.2796,
+      "step": 1772
+    },
+    {
+      "epoch": 2.138721351025332,
+      "grad_norm": 0.2657062977643569,
+      "learning_rate": 1.5951742627345845e-05,
+      "loss": 0.2998,
+      "step": 1773
+    },
+    {
+      "epoch": 2.139927623642943,
+      "grad_norm": 0.24246975530355397,
+      "learning_rate": 1.5929401251117073e-05,
+      "loss": 0.2967,
+      "step": 1774
+    },
+    {
+      "epoch": 2.141133896260555,
+      "grad_norm": 0.24199935788003604,
+      "learning_rate": 1.5907059874888293e-05,
+      "loss": 0.2833,
+      "step": 1775
+    },
+    {
+      "epoch": 2.1423401688781665,
+      "grad_norm": 0.2670641237942198,
+      "learning_rate": 1.5884718498659517e-05,
+      "loss": 0.2856,
+      "step": 1776
+    },
+    {
+      "epoch": 2.143546441495778,
+      "grad_norm": 0.23890355072402358,
+      "learning_rate": 1.5862377122430745e-05,
+      "loss": 0.2728,
+      "step": 1777
+    },
+    {
+      "epoch": 2.1447527141133897,
+      "grad_norm": 0.23673195464701202,
+      "learning_rate": 1.5840035746201966e-05,
+      "loss": 0.3005,
+      "step": 1778
+    },
+    {
+      "epoch": 2.145958986731001,
+      "grad_norm": 0.2865067160256694,
+      "learning_rate": 1.581769436997319e-05,
+      "loss": 0.2959,
+      "step": 1779
+    },
+    {
+      "epoch": 2.147165259348613,
+      "grad_norm": 0.27302383294132143,
+      "learning_rate": 1.5795352993744417e-05,
+      "loss": 0.3018,
+      "step": 1780
+    },
+    {
+      "epoch": 2.1483715319662244,
+      "grad_norm": 0.23492163997305548,
+      "learning_rate": 1.577301161751564e-05,
+      "loss": 0.2848,
+      "step": 1781
+    },
+    {
+      "epoch": 2.149577804583836,
+      "grad_norm": 0.27129742825552006,
+      "learning_rate": 1.5750670241286862e-05,
+      "loss": 0.2958,
+      "step": 1782
+    },
+    {
+      "epoch": 2.1507840772014477,
+      "grad_norm": 0.2684880578576593,
+      "learning_rate": 1.572832886505809e-05,
+      "loss": 0.2903,
+      "step": 1783
+    },
+    {
+      "epoch": 2.151990349819059,
+      "grad_norm": 0.27404258961414674,
+      "learning_rate": 1.5705987488829313e-05,
+      "loss": 0.2881,
+      "step": 1784
+    },
+    {
+      "epoch": 2.1531966224366705,
+      "grad_norm": 0.2632922456415148,
+      "learning_rate": 1.5683646112600538e-05,
+      "loss": 0.3016,
+      "step": 1785
+    },
+    {
+      "epoch": 2.1544028950542824,
+      "grad_norm": 0.26663256503427807,
+      "learning_rate": 1.566130473637176e-05,
+      "loss": 0.2963,
+      "step": 1786
+    },
+    {
+      "epoch": 2.155609167671894,
+      "grad_norm": 0.2508496833033383,
+      "learning_rate": 1.5638963360142986e-05,
+      "loss": 0.2801,
+      "step": 1787
+    },
+    {
+      "epoch": 2.1568154402895052,
+      "grad_norm": 0.22978346023328622,
+      "learning_rate": 1.561662198391421e-05,
+      "loss": 0.2754,
+      "step": 1788
+    },
+    {
+      "epoch": 2.158021712907117,
+      "grad_norm": 0.2679042839697001,
+      "learning_rate": 1.5594280607685434e-05,
+      "loss": 0.2995,
+      "step": 1789
+    },
+    {
+      "epoch": 2.1592279855247285,
+      "grad_norm": 0.2690040951898322,
+      "learning_rate": 1.5571939231456658e-05,
+      "loss": 0.3033,
+      "step": 1790
+    },
+    {
+      "epoch": 2.1604342581423404,
+      "grad_norm": 0.2319672822255795,
+      "learning_rate": 1.5549597855227882e-05,
+      "loss": 0.2825,
+      "step": 1791
+    },
+    {
+      "epoch": 2.1616405307599518,
+      "grad_norm": 0.2794357797508637,
+      "learning_rate": 1.552725647899911e-05,
+      "loss": 0.2911,
+      "step": 1792
+    },
+    {
+      "epoch": 2.162846803377563,
+      "grad_norm": 0.21909695211407065,
+      "learning_rate": 1.550491510277033e-05,
+      "loss": 0.2659,
+      "step": 1793
+    },
+    {
+      "epoch": 2.164053075995175,
+      "grad_norm": 0.29375511141007266,
+      "learning_rate": 1.5482573726541554e-05,
+      "loss": 0.3114,
+      "step": 1794
+    },
+    {
+      "epoch": 2.1652593486127865,
+      "grad_norm": 0.2529271568454275,
+      "learning_rate": 1.5460232350312782e-05,
+      "loss": 0.2793,
+      "step": 1795
+    },
+    {
+      "epoch": 2.166465621230398,
+      "grad_norm": 0.24510286216006683,
+      "learning_rate": 1.5437890974084006e-05,
+      "loss": 0.2938,
+      "step": 1796
+    },
+    {
+      "epoch": 2.1676718938480097,
+      "grad_norm": 0.24633271965834536,
+      "learning_rate": 1.5415549597855227e-05,
+      "loss": 0.2936,
+      "step": 1797
+    },
+    {
+      "epoch": 2.168878166465621,
+      "grad_norm": 0.2908065624489396,
+      "learning_rate": 1.5393208221626454e-05,
+      "loss": 0.2943,
+      "step": 1798
+    },
+    {
+      "epoch": 2.170084439083233,
+      "grad_norm": 0.2749611947114117,
+      "learning_rate": 1.5370866845397678e-05,
+      "loss": 0.2871,
+      "step": 1799
+    },
+    {
+      "epoch": 2.1712907117008444,
+      "grad_norm": 0.2231876084061707,
+      "learning_rate": 1.5348525469168902e-05,
+      "loss": 0.2904,
+      "step": 1800
+    },
+    {
+      "epoch": 2.172496984318456,
+      "grad_norm": 0.22679976460521523,
+      "learning_rate": 1.5326184092940126e-05,
+      "loss": 0.294,
+      "step": 1801
+    },
+    {
+      "epoch": 2.1737032569360677,
+      "grad_norm": 0.24525986813590006,
+      "learning_rate": 1.530384271671135e-05,
+      "loss": 0.3015,
+      "step": 1802
+    },
+    {
+      "epoch": 2.174909529553679,
+      "grad_norm": 0.22623155707285916,
+      "learning_rate": 1.5281501340482574e-05,
+      "loss": 0.2772,
+      "step": 1803
+    },
+    {
+      "epoch": 2.1761158021712905,
+      "grad_norm": 0.23398318830707168,
+      "learning_rate": 1.52591599642538e-05,
+      "loss": 0.284,
+      "step": 1804
+    },
+    {
+      "epoch": 2.1773220747889024,
+      "grad_norm": 0.23119064864869804,
+      "learning_rate": 1.5236818588025023e-05,
+      "loss": 0.281,
+      "step": 1805
+    },
+    {
+      "epoch": 2.178528347406514,
+      "grad_norm": 0.22917560469727433,
+      "learning_rate": 1.5214477211796247e-05,
+      "loss": 0.2786,
+      "step": 1806
+    },
+    {
+      "epoch": 2.1797346200241257,
+      "grad_norm": 0.23631284216344356,
+      "learning_rate": 1.5192135835567473e-05,
+      "loss": 0.2713,
+      "step": 1807
+    },
+    {
+      "epoch": 2.180940892641737,
+      "grad_norm": 0.2488903695574401,
+      "learning_rate": 1.5169794459338697e-05,
+      "loss": 0.2774,
+      "step": 1808
+    },
+    {
+      "epoch": 2.1821471652593485,
+      "grad_norm": 0.2155531209044899,
+      "learning_rate": 1.5147453083109919e-05,
+      "loss": 0.2966,
+      "step": 1809
+    },
+    {
+      "epoch": 2.1833534378769603,
+      "grad_norm": 0.23495355035903273,
+      "learning_rate": 1.5125111706881145e-05,
+      "loss": 0.3087,
+      "step": 1810
+    },
+    {
+      "epoch": 2.1845597104945718,
+      "grad_norm": 0.234213301291648,
+      "learning_rate": 1.5102770330652369e-05,
+      "loss": 0.2762,
+      "step": 1811
+    },
+    {
+      "epoch": 2.185765983112183,
+      "grad_norm": 0.2778933084532547,
+      "learning_rate": 1.5080428954423593e-05,
+      "loss": 0.2851,
+      "step": 1812
+    },
+    {
+      "epoch": 2.186972255729795,
+      "grad_norm": 0.24101135672543364,
+      "learning_rate": 1.5058087578194819e-05,
+      "loss": 0.2793,
+      "step": 1813
+    },
+    {
+      "epoch": 2.1881785283474064,
+      "grad_norm": 0.2510040494621872,
+      "learning_rate": 1.5035746201966041e-05,
+      "loss": 0.2994,
+      "step": 1814
+    },
+    {
+      "epoch": 2.1893848009650183,
+      "grad_norm": 0.2128063834439931,
+      "learning_rate": 1.5013404825737265e-05,
+      "loss": 0.2838,
+      "step": 1815
+    },
+    {
+      "epoch": 2.1905910735826297,
+      "grad_norm": 0.24754013889868678,
+      "learning_rate": 1.4991063449508491e-05,
+      "loss": 0.3053,
+      "step": 1816
+    },
+    {
+      "epoch": 2.191797346200241,
+      "grad_norm": 0.2581031353095379,
+      "learning_rate": 1.4968722073279715e-05,
+      "loss": 0.288,
+      "step": 1817
+    },
+    {
+      "epoch": 2.193003618817853,
+      "grad_norm": 0.23487250557642658,
+      "learning_rate": 1.4946380697050938e-05,
+      "loss": 0.2986,
+      "step": 1818
+    },
+    {
+      "epoch": 2.1942098914354644,
+      "grad_norm": 0.20696922559019662,
+      "learning_rate": 1.4924039320822163e-05,
+      "loss": 0.2813,
+      "step": 1819
+    },
+    {
+      "epoch": 2.195416164053076,
+      "grad_norm": 0.220731832587104,
+      "learning_rate": 1.4901697944593387e-05,
+      "loss": 0.2748,
+      "step": 1820
+    },
+    {
+      "epoch": 2.1966224366706877,
+      "grad_norm": 0.25166555395330575,
+      "learning_rate": 1.4879356568364611e-05,
+      "loss": 0.285,
+      "step": 1821
+    },
+    {
+      "epoch": 2.197828709288299,
+      "grad_norm": 0.23349864934060624,
+      "learning_rate": 1.4857015192135837e-05,
+      "loss": 0.2786,
+      "step": 1822
+    },
+    {
+      "epoch": 2.1990349819059105,
+      "grad_norm": 0.2352295275890298,
+      "learning_rate": 1.483467381590706e-05,
+      "loss": 0.2901,
+      "step": 1823
+    },
+    {
+      "epoch": 2.2002412545235224,
+      "grad_norm": 0.24218185360551212,
+      "learning_rate": 1.4812332439678284e-05,
+      "loss": 0.269,
+      "step": 1824
+    },
+    {
+      "epoch": 2.201447527141134,
+      "grad_norm": 0.23597928948583063,
+      "learning_rate": 1.478999106344951e-05,
+      "loss": 0.2852,
+      "step": 1825
+    },
+    {
+      "epoch": 2.2026537997587456,
+      "grad_norm": 0.20882073193696107,
+      "learning_rate": 1.4767649687220734e-05,
+      "loss": 0.2703,
+      "step": 1826
+    },
+    {
+      "epoch": 2.203860072376357,
+      "grad_norm": 0.24223129133976676,
+      "learning_rate": 1.4745308310991956e-05,
+      "loss": 0.2921,
+      "step": 1827
+    },
+    {
+      "epoch": 2.2050663449939685,
+      "grad_norm": 0.246628534539299,
+      "learning_rate": 1.4722966934763183e-05,
+      "loss": 0.2917,
+      "step": 1828
+    },
+    {
+      "epoch": 2.2062726176115803,
+      "grad_norm": 0.2302693025242607,
+      "learning_rate": 1.4700625558534406e-05,
+      "loss": 0.2709,
+      "step": 1829
+    },
+    {
+      "epoch": 2.2074788902291917,
+      "grad_norm": 0.224038143545482,
+      "learning_rate": 1.467828418230563e-05,
+      "loss": 0.2851,
+      "step": 1830
+    },
+    {
+      "epoch": 2.2086851628468036,
+      "grad_norm": 0.23867623831655044,
+      "learning_rate": 1.4655942806076856e-05,
+      "loss": 0.282,
+      "step": 1831
+    },
+    {
+      "epoch": 2.209891435464415,
+      "grad_norm": 0.23595428041685104,
+      "learning_rate": 1.4633601429848078e-05,
+      "loss": 0.279,
+      "step": 1832
+    },
+    {
+      "epoch": 2.2110977080820264,
+      "grad_norm": 0.2397339096346714,
+      "learning_rate": 1.4611260053619302e-05,
+      "loss": 0.2938,
+      "step": 1833
+    },
+    {
+      "epoch": 2.2123039806996383,
+      "grad_norm": 0.25086601174251194,
+      "learning_rate": 1.4588918677390528e-05,
+      "loss": 0.2881,
+      "step": 1834
+    },
+    {
+      "epoch": 2.2135102533172497,
+      "grad_norm": 0.2979190537928123,
+      "learning_rate": 1.4566577301161752e-05,
+      "loss": 0.3057,
+      "step": 1835
+    },
+    {
+      "epoch": 2.214716525934861,
+      "grad_norm": 0.23603851376001234,
+      "learning_rate": 1.4544235924932978e-05,
+      "loss": 0.2907,
+      "step": 1836
+    },
+    {
+      "epoch": 2.215922798552473,
+      "grad_norm": 0.22529168222023993,
+      "learning_rate": 1.4521894548704202e-05,
+      "loss": 0.2992,
+      "step": 1837
+    },
+    {
+      "epoch": 2.2171290711700844,
+      "grad_norm": 0.22362154450561503,
+      "learning_rate": 1.4499553172475424e-05,
+      "loss": 0.2694,
+      "step": 1838
+    },
+    {
+      "epoch": 2.218335343787696,
+      "grad_norm": 0.22302860615250908,
+      "learning_rate": 1.447721179624665e-05,
+      "loss": 0.284,
+      "step": 1839
+    },
+    {
+      "epoch": 2.2195416164053077,
+      "grad_norm": 0.22904183729570707,
+      "learning_rate": 1.4454870420017874e-05,
+      "loss": 0.2795,
+      "step": 1840
+    },
+    {
+      "epoch": 2.220747889022919,
+      "grad_norm": 0.26782453766381153,
+      "learning_rate": 1.4432529043789098e-05,
+      "loss": 0.2929,
+      "step": 1841
+    },
+    {
+      "epoch": 2.221954161640531,
+      "grad_norm": 0.21714316481031462,
+      "learning_rate": 1.4410187667560324e-05,
+      "loss": 0.2731,
+      "step": 1842
+    },
+    {
+      "epoch": 2.2231604342581424,
+      "grad_norm": 0.26186668193353185,
+      "learning_rate": 1.4387846291331546e-05,
+      "loss": 0.2839,
+      "step": 1843
+    },
+    {
+      "epoch": 2.2243667068757538,
+      "grad_norm": 0.2911995525000822,
+      "learning_rate": 1.436550491510277e-05,
+      "loss": 0.2972,
+      "step": 1844
+    },
+    {
+      "epoch": 2.2255729794933656,
+      "grad_norm": 0.24276375836112515,
+      "learning_rate": 1.4343163538873996e-05,
+      "loss": 0.2914,
+      "step": 1845
+    },
+    {
+      "epoch": 2.226779252110977,
+      "grad_norm": 0.24600706178312343,
+      "learning_rate": 1.432082216264522e-05,
+      "loss": 0.2945,
+      "step": 1846
+    },
+    {
+      "epoch": 2.2279855247285885,
+      "grad_norm": 0.22011966635882427,
+      "learning_rate": 1.4298480786416443e-05,
+      "loss": 0.2705,
+      "step": 1847
+    },
+    {
+      "epoch": 2.2291917973462003,
+      "grad_norm": 0.22274947815980448,
+      "learning_rate": 1.4276139410187669e-05,
+      "loss": 0.2858,
+      "step": 1848
+    },
+    {
+      "epoch": 2.2303980699638117,
+      "grad_norm": 0.24312414836648066,
+      "learning_rate": 1.4253798033958893e-05,
+      "loss": 0.275,
+      "step": 1849
+    },
+    {
+      "epoch": 2.2316043425814236,
+      "grad_norm": 0.25667716385389866,
+      "learning_rate": 1.4231456657730117e-05,
+      "loss": 0.2971,
+      "step": 1850
+    },
+    {
+      "epoch": 2.232810615199035,
+      "grad_norm": 0.23198803574642948,
+      "learning_rate": 1.4209115281501343e-05,
+      "loss": 0.2849,
+      "step": 1851
+    },
+    {
+      "epoch": 2.2340168878166464,
+      "grad_norm": 0.2401993446695254,
+      "learning_rate": 1.4186773905272565e-05,
+      "loss": 0.2916,
+      "step": 1852
+    },
+    {
+      "epoch": 2.2352231604342583,
+      "grad_norm": 0.2510433024863049,
+      "learning_rate": 1.4164432529043789e-05,
+      "loss": 0.2735,
+      "step": 1853
+    },
+    {
+      "epoch": 2.2364294330518697,
+      "grad_norm": 0.2404864962858689,
+      "learning_rate": 1.4142091152815015e-05,
+      "loss": 0.2892,
+      "step": 1854
+    },
+    {
+      "epoch": 2.237635705669481,
+      "grad_norm": 0.22016088446747195,
+      "learning_rate": 1.4119749776586239e-05,
+      "loss": 0.2798,
+      "step": 1855
+    },
+    {
+      "epoch": 2.238841978287093,
+      "grad_norm": 0.24403858808762297,
+      "learning_rate": 1.4097408400357461e-05,
+      "loss": 0.2967,
+      "step": 1856
+    },
+    {
+      "epoch": 2.2400482509047044,
+      "grad_norm": 0.23489950259808423,
+      "learning_rate": 1.4075067024128689e-05,
+      "loss": 0.2788,
+      "step": 1857
+    },
+    {
+      "epoch": 2.2412545235223162,
+      "grad_norm": 0.25532204034310485,
+      "learning_rate": 1.4052725647899911e-05,
+      "loss": 0.295,
+      "step": 1858
+    },
+    {
+      "epoch": 2.2424607961399277,
+      "grad_norm": 0.251973767975049,
+      "learning_rate": 1.4030384271671135e-05,
+      "loss": 0.3009,
+      "step": 1859
+    },
+    {
+      "epoch": 2.243667068757539,
+      "grad_norm": 0.217744929867661,
+      "learning_rate": 1.4008042895442361e-05,
+      "loss": 0.292,
+      "step": 1860
+    },
+    {
+      "epoch": 2.244873341375151,
+      "grad_norm": 0.20972929540640886,
+      "learning_rate": 1.3985701519213583e-05,
+      "loss": 0.2753,
+      "step": 1861
+    },
+    {
+      "epoch": 2.2460796139927623,
+      "grad_norm": 0.2593676734049886,
+      "learning_rate": 1.3963360142984808e-05,
+      "loss": 0.2945,
+      "step": 1862
+    },
+    {
+      "epoch": 2.2472858866103738,
+      "grad_norm": 0.2407026323892429,
+      "learning_rate": 1.3941018766756033e-05,
+      "loss": 0.2921,
+      "step": 1863
+    },
+    {
+      "epoch": 2.2484921592279856,
+      "grad_norm": 0.21792380789916738,
+      "learning_rate": 1.3918677390527257e-05,
+      "loss": 0.2968,
+      "step": 1864
+    },
+    {
+      "epoch": 2.249698431845597,
+      "grad_norm": 0.2294246384984617,
+      "learning_rate": 1.389633601429848e-05,
+      "loss": 0.2894,
+      "step": 1865
+    },
+    {
+      "epoch": 2.250904704463209,
+      "grad_norm": 0.2352525773133271,
+      "learning_rate": 1.3873994638069707e-05,
+      "loss": 0.2801,
+      "step": 1866
+    },
+    {
+      "epoch": 2.2521109770808203,
+      "grad_norm": 0.2261284819996082,
+      "learning_rate": 1.385165326184093e-05,
+      "loss": 0.2672,
+      "step": 1867
+    },
+    {
+      "epoch": 2.2533172496984317,
+      "grad_norm": 0.23872374706253088,
+      "learning_rate": 1.3829311885612154e-05,
+      "loss": 0.2956,
+      "step": 1868
+    },
+    {
+      "epoch": 2.2545235223160436,
+      "grad_norm": 0.23826064709676967,
+      "learning_rate": 1.380697050938338e-05,
+      "loss": 0.284,
+      "step": 1869
+    },
+    {
+      "epoch": 2.255729794933655,
+      "grad_norm": 0.2294043053362506,
+      "learning_rate": 1.3784629133154604e-05,
+      "loss": 0.2819,
+      "step": 1870
+    },
+    {
+      "epoch": 2.2569360675512664,
+      "grad_norm": 0.23258701478035285,
+      "learning_rate": 1.3762287756925826e-05,
+      "loss": 0.2697,
+      "step": 1871
+    },
+    {
+      "epoch": 2.2581423401688783,
+      "grad_norm": 0.204987775131546,
+      "learning_rate": 1.3739946380697052e-05,
+      "loss": 0.2875,
+      "step": 1872
+    },
+    {
+      "epoch": 2.2593486127864897,
+      "grad_norm": 0.24311420445354,
+      "learning_rate": 1.3717605004468276e-05,
+      "loss": 0.2797,
+      "step": 1873
+    },
+    {
+      "epoch": 2.260554885404101,
+      "grad_norm": 0.2706634799528978,
+      "learning_rate": 1.3695263628239498e-05,
+      "loss": 0.3022,
+      "step": 1874
+    },
+    {
+      "epoch": 2.261761158021713,
+      "grad_norm": 0.23301429893513645,
+      "learning_rate": 1.3672922252010726e-05,
+      "loss": 0.2736,
+      "step": 1875
+    },
+    {
+      "epoch": 2.2629674306393244,
+      "grad_norm": 0.2227015979945295,
+      "learning_rate": 1.3650580875781948e-05,
+      "loss": 0.2892,
+      "step": 1876
+    },
+    {
+      "epoch": 2.2641737032569362,
+      "grad_norm": 0.2518005952205528,
+      "learning_rate": 1.3628239499553172e-05,
+      "loss": 0.289,
+      "step": 1877
+    },
+    {
+      "epoch": 2.2653799758745476,
+      "grad_norm": 0.22770033941273815,
+      "learning_rate": 1.3605898123324398e-05,
+      "loss": 0.2707,
+      "step": 1878
+    },
+    {
+      "epoch": 2.266586248492159,
+      "grad_norm": 0.25592147280730076,
+      "learning_rate": 1.3583556747095622e-05,
+      "loss": 0.2984,
+      "step": 1879
+    },
+    {
+      "epoch": 2.267792521109771,
+      "grad_norm": 0.24678852398410292,
+      "learning_rate": 1.3561215370866845e-05,
+      "loss": 0.2893,
+      "step": 1880
+    },
+    {
+      "epoch": 2.2689987937273823,
+      "grad_norm": 0.21640104205658625,
+      "learning_rate": 1.353887399463807e-05,
+      "loss": 0.284,
+      "step": 1881
+    },
+    {
+      "epoch": 2.270205066344994,
+      "grad_norm": 0.22263483357092576,
+      "learning_rate": 1.3516532618409294e-05,
+      "loss": 0.2756,
+      "step": 1882
+    },
+    {
+      "epoch": 2.2714113389626056,
+      "grad_norm": 0.22549139263598314,
+      "learning_rate": 1.3494191242180519e-05,
+      "loss": 0.2804,
+      "step": 1883
+    },
+    {
+      "epoch": 2.272617611580217,
+      "grad_norm": 0.22848000714044425,
+      "learning_rate": 1.3471849865951744e-05,
+      "loss": 0.2943,
+      "step": 1884
+    },
+    {
+      "epoch": 2.273823884197829,
+      "grad_norm": 0.22965966451255318,
+      "learning_rate": 1.3449508489722967e-05,
+      "loss": 0.2939,
+      "step": 1885
+    },
+    {
+      "epoch": 2.2750301568154403,
+      "grad_norm": 0.22002800402535225,
+      "learning_rate": 1.342716711349419e-05,
+      "loss": 0.2708,
+      "step": 1886
+    },
+    {
+      "epoch": 2.2762364294330517,
+      "grad_norm": 0.22824703002205038,
+      "learning_rate": 1.3404825737265417e-05,
+      "loss": 0.2959,
+      "step": 1887
+    },
+    {
+      "epoch": 2.2774427020506636,
+      "grad_norm": 0.2175045283912371,
+      "learning_rate": 1.338248436103664e-05,
+      "loss": 0.2956,
+      "step": 1888
+    },
+    {
+      "epoch": 2.278648974668275,
+      "grad_norm": 0.2324410570763942,
+      "learning_rate": 1.3360142984807863e-05,
+      "loss": 0.2933,
+      "step": 1889
+    },
+    {
+      "epoch": 2.2798552472858864,
+      "grad_norm": 0.2291968726319126,
+      "learning_rate": 1.333780160857909e-05,
+      "loss": 0.289,
+      "step": 1890
+    },
+    {
+      "epoch": 2.2810615199034983,
+      "grad_norm": 0.23648168240852283,
+      "learning_rate": 1.3315460232350313e-05,
+      "loss": 0.3109,
+      "step": 1891
+    },
+    {
+      "epoch": 2.2822677925211097,
+      "grad_norm": 0.2231262318277235,
+      "learning_rate": 1.3293118856121537e-05,
+      "loss": 0.2835,
+      "step": 1892
+    },
+    {
+      "epoch": 2.2834740651387215,
+      "grad_norm": 0.23818818699053612,
+      "learning_rate": 1.3270777479892763e-05,
+      "loss": 0.2831,
+      "step": 1893
+    },
+    {
+      "epoch": 2.284680337756333,
+      "grad_norm": 0.2434904932138937,
+      "learning_rate": 1.3248436103663985e-05,
+      "loss": 0.294,
+      "step": 1894
+    },
+    {
+      "epoch": 2.2858866103739444,
+      "grad_norm": 0.21286990891188562,
+      "learning_rate": 1.322609472743521e-05,
+      "loss": 0.2756,
+      "step": 1895
+    },
+    {
+      "epoch": 2.287092882991556,
+      "grad_norm": 0.23610309357095516,
+      "learning_rate": 1.3203753351206435e-05,
+      "loss": 0.2905,
+      "step": 1896
+    },
+    {
+      "epoch": 2.2882991556091676,
+      "grad_norm": 0.240924967821922,
+      "learning_rate": 1.3181411974977659e-05,
+      "loss": 0.2721,
+      "step": 1897
+    },
+    {
+      "epoch": 2.2895054282267795,
+      "grad_norm": 0.2555889113252634,
+      "learning_rate": 1.3159070598748885e-05,
+      "loss": 0.3004,
+      "step": 1898
+    },
+    {
+      "epoch": 2.290711700844391,
+      "grad_norm": 0.2374596441116695,
+      "learning_rate": 1.3136729222520109e-05,
+      "loss": 0.276,
+      "step": 1899
+    },
+    {
+      "epoch": 2.2919179734620023,
+      "grad_norm": 0.22527696891452795,
+      "learning_rate": 1.3114387846291331e-05,
+      "loss": 0.2762,
+      "step": 1900
+    },
+    {
+      "epoch": 2.293124246079614,
+      "grad_norm": 0.25567802002526824,
+      "learning_rate": 1.3092046470062557e-05,
+      "loss": 0.2813,
+      "step": 1901
+    },
+    {
+      "epoch": 2.2943305186972256,
+      "grad_norm": 0.21906439644766362,
+      "learning_rate": 1.3069705093833781e-05,
+      "loss": 0.2875,
+      "step": 1902
+    },
+    {
+      "epoch": 2.295536791314837,
+      "grad_norm": 0.22064470665788954,
+      "learning_rate": 1.3047363717605005e-05,
+      "loss": 0.2857,
+      "step": 1903
+    },
+    {
+      "epoch": 2.296743063932449,
+      "grad_norm": 0.23151609094727654,
+      "learning_rate": 1.3025022341376231e-05,
+      "loss": 0.2842,
+      "step": 1904
+    },
+    {
+      "epoch": 2.2979493365500603,
+      "grad_norm": 0.2351690748078775,
+      "learning_rate": 1.3002680965147454e-05,
+      "loss": 0.2769,
+      "step": 1905
+    },
+    {
+      "epoch": 2.2991556091676717,
+      "grad_norm": 0.23004031471475894,
+      "learning_rate": 1.2980339588918678e-05,
+      "loss": 0.2915,
+      "step": 1906
+    },
+    {
+      "epoch": 2.3003618817852836,
+      "grad_norm": 0.21332903284760457,
+      "learning_rate": 1.2957998212689903e-05,
+      "loss": 0.2695,
+      "step": 1907
+    },
+    {
+      "epoch": 2.301568154402895,
+      "grad_norm": 0.2536337679173862,
+      "learning_rate": 1.2935656836461127e-05,
+      "loss": 0.2854,
+      "step": 1908
+    },
+    {
+      "epoch": 2.3027744270205064,
+      "grad_norm": 0.26978055805109497,
+      "learning_rate": 1.291331546023235e-05,
+      "loss": 0.2822,
+      "step": 1909
+    },
+    {
+      "epoch": 2.3039806996381182,
+      "grad_norm": 0.21833218367797522,
+      "learning_rate": 1.2890974084003576e-05,
+      "loss": 0.2751,
+      "step": 1910
+    },
+    {
+      "epoch": 2.3051869722557297,
+      "grad_norm": 0.213236596889352,
+      "learning_rate": 1.28686327077748e-05,
+      "loss": 0.2767,
+      "step": 1911
+    },
+    {
+      "epoch": 2.3063932448733415,
+      "grad_norm": 0.25023452493910114,
+      "learning_rate": 1.2846291331546024e-05,
+      "loss": 0.2727,
+      "step": 1912
+    },
+    {
+      "epoch": 2.307599517490953,
+      "grad_norm": 0.25357039532276837,
+      "learning_rate": 1.282394995531725e-05,
+      "loss": 0.2887,
+      "step": 1913
+    },
+    {
+      "epoch": 2.3088057901085643,
+      "grad_norm": 0.22829308034251444,
+      "learning_rate": 1.2801608579088472e-05,
+      "loss": 0.2766,
+      "step": 1914
+    },
+    {
+      "epoch": 2.310012062726176,
+      "grad_norm": 0.22565008404788817,
+      "learning_rate": 1.2779267202859696e-05,
+      "loss": 0.299,
+      "step": 1915
+    },
+    {
+      "epoch": 2.3112183353437876,
+      "grad_norm": 0.23456861896586176,
+      "learning_rate": 1.2756925826630922e-05,
+      "loss": 0.2747,
+      "step": 1916
+    },
+    {
+      "epoch": 2.3124246079613995,
+      "grad_norm": 0.23597661508111467,
+      "learning_rate": 1.2734584450402146e-05,
+      "loss": 0.2894,
+      "step": 1917
+    },
+    {
+      "epoch": 2.313630880579011,
+      "grad_norm": 0.23786461250041235,
+      "learning_rate": 1.2712243074173368e-05,
+      "loss": 0.2907,
+      "step": 1918
+    },
+    {
+      "epoch": 2.3148371531966223,
+      "grad_norm": 0.2534049640911676,
+      "learning_rate": 1.2689901697944596e-05,
+      "loss": 0.287,
+      "step": 1919
+    },
+    {
+      "epoch": 2.316043425814234,
+      "grad_norm": 0.22090587312904558,
+      "learning_rate": 1.2667560321715818e-05,
+      "loss": 0.2789,
+      "step": 1920
+    },
+    {
+      "epoch": 2.3172496984318456,
+      "grad_norm": 0.235549531483714,
+      "learning_rate": 1.2645218945487042e-05,
+      "loss": 0.286,
+      "step": 1921
+    },
+    {
+      "epoch": 2.318455971049457,
+      "grad_norm": 0.25142631802715215,
+      "learning_rate": 1.2622877569258268e-05,
+      "loss": 0.2846,
+      "step": 1922
+    },
+    {
+      "epoch": 2.319662243667069,
+      "grad_norm": 0.2524115813131248,
+      "learning_rate": 1.260053619302949e-05,
+      "loss": 0.2877,
+      "step": 1923
+    },
+    {
+      "epoch": 2.3208685162846803,
+      "grad_norm": 0.24100570731736226,
+      "learning_rate": 1.2578194816800715e-05,
+      "loss": 0.2905,
+      "step": 1924
+    },
+    {
+      "epoch": 2.3220747889022917,
+      "grad_norm": 0.23658962005375492,
+      "learning_rate": 1.255585344057194e-05,
+      "loss": 0.3116,
+      "step": 1925
+    },
+    {
+      "epoch": 2.3232810615199035,
+      "grad_norm": 0.2366915639739183,
+      "learning_rate": 1.2533512064343164e-05,
+      "loss": 0.2653,
+      "step": 1926
+    },
+    {
+      "epoch": 2.324487334137515,
+      "grad_norm": 0.24126309010219563,
+      "learning_rate": 1.2511170688114387e-05,
+      "loss": 0.2944,
+      "step": 1927
+    },
+    {
+      "epoch": 2.325693606755127,
+      "grad_norm": 0.256417259208843,
+      "learning_rate": 1.2488829311885613e-05,
+      "loss": 0.299,
+      "step": 1928
+    },
+    {
+      "epoch": 2.3268998793727382,
+      "grad_norm": 0.2533720151680546,
+      "learning_rate": 1.2466487935656837e-05,
+      "loss": 0.288,
+      "step": 1929
+    },
+    {
+      "epoch": 2.3281061519903496,
+      "grad_norm": 0.24913672554051416,
+      "learning_rate": 1.2444146559428063e-05,
+      "loss": 0.3067,
+      "step": 1930
+    },
+    {
+      "epoch": 2.3293124246079615,
+      "grad_norm": 0.24341114503544165,
+      "learning_rate": 1.2421805183199285e-05,
+      "loss": 0.2763,
+      "step": 1931
+    },
+    {
+      "epoch": 2.330518697225573,
+      "grad_norm": 0.2547834249597416,
+      "learning_rate": 1.239946380697051e-05,
+      "loss": 0.2936,
+      "step": 1932
+    },
+    {
+      "epoch": 2.331724969843185,
+      "grad_norm": 0.23300171268866948,
+      "learning_rate": 1.2377122430741735e-05,
+      "loss": 0.2934,
+      "step": 1933
+    },
+    {
+      "epoch": 2.332931242460796,
+      "grad_norm": 0.24474857033348835,
+      "learning_rate": 1.2354781054512959e-05,
+      "loss": 0.2887,
+      "step": 1934
+    },
+    {
+      "epoch": 2.3341375150784076,
+      "grad_norm": 0.2151206611437992,
+      "learning_rate": 1.2332439678284183e-05,
+      "loss": 0.2836,
+      "step": 1935
+    },
+    {
+      "epoch": 2.3353437876960195,
+      "grad_norm": 0.24359081217137948,
+      "learning_rate": 1.2310098302055407e-05,
+      "loss": 0.2987,
+      "step": 1936
+    },
+    {
+      "epoch": 2.336550060313631,
+      "grad_norm": 0.2558446677582328,
+      "learning_rate": 1.2287756925826631e-05,
+      "loss": 0.2736,
+      "step": 1937
+    },
+    {
+      "epoch": 2.3377563329312423,
+      "grad_norm": 0.23133494031808785,
+      "learning_rate": 1.2265415549597855e-05,
+      "loss": 0.2967,
+      "step": 1938
+    },
+    {
+      "epoch": 2.338962605548854,
+      "grad_norm": 0.22011413912360578,
+      "learning_rate": 1.2243074173369081e-05,
+      "loss": 0.2758,
+      "step": 1939
+    },
+    {
+      "epoch": 2.3401688781664656,
+      "grad_norm": 0.2465890608008465,
+      "learning_rate": 1.2220732797140303e-05,
+      "loss": 0.2862,
+      "step": 1940
+    },
+    {
+      "epoch": 2.341375150784077,
+      "grad_norm": 0.23400609607040407,
+      "learning_rate": 1.219839142091153e-05,
+      "loss": 0.2833,
+      "step": 1941
+    },
+    {
+      "epoch": 2.342581423401689,
+      "grad_norm": 0.2267979596723007,
+      "learning_rate": 1.2176050044682753e-05,
+      "loss": 0.29,
+      "step": 1942
+    },
+    {
+      "epoch": 2.3437876960193003,
+      "grad_norm": 0.24060657688093173,
+      "learning_rate": 1.2153708668453977e-05,
+      "loss": 0.295,
+      "step": 1943
+    },
+    {
+      "epoch": 2.344993968636912,
+      "grad_norm": 0.2152614594026681,
+      "learning_rate": 1.2131367292225201e-05,
+      "loss": 0.275,
+      "step": 1944
+    },
+    {
+      "epoch": 2.3462002412545235,
+      "grad_norm": 0.21991178211682616,
+      "learning_rate": 1.2109025915996426e-05,
+      "loss": 0.2734,
+      "step": 1945
+    },
+    {
+      "epoch": 2.347406513872135,
+      "grad_norm": 0.2421436523612531,
+      "learning_rate": 1.2086684539767651e-05,
+      "loss": 0.2944,
+      "step": 1946
+    },
+    {
+      "epoch": 2.348612786489747,
+      "grad_norm": 0.21785789529396196,
+      "learning_rate": 1.2064343163538874e-05,
+      "loss": 0.2836,
+      "step": 1947
+    },
+    {
+      "epoch": 2.349819059107358,
+      "grad_norm": 0.2223002921132739,
+      "learning_rate": 1.20420017873101e-05,
+      "loss": 0.2927,
+      "step": 1948
+    },
+    {
+      "epoch": 2.35102533172497,
+      "grad_norm": 0.23810589562126233,
+      "learning_rate": 1.2019660411081324e-05,
+      "loss": 0.2884,
+      "step": 1949
+    },
+    {
+      "epoch": 2.3522316043425815,
+      "grad_norm": 0.23259108876620938,
+      "learning_rate": 1.1997319034852548e-05,
+      "loss": 0.2826,
+      "step": 1950
+    },
+    {
+      "epoch": 2.353437876960193,
+      "grad_norm": 0.22456352439209676,
+      "learning_rate": 1.1974977658623772e-05,
+      "loss": 0.3101,
+      "step": 1951
+    },
+    {
+      "epoch": 2.3546441495778048,
+      "grad_norm": 0.24091153223400058,
+      "learning_rate": 1.1952636282394996e-05,
+      "loss": 0.3015,
+      "step": 1952
+    },
+    {
+      "epoch": 2.355850422195416,
+      "grad_norm": 0.23719969534608104,
+      "learning_rate": 1.193029490616622e-05,
+      "loss": 0.2764,
+      "step": 1953
+    },
+    {
+      "epoch": 2.3570566948130276,
+      "grad_norm": 0.2231954319848683,
+      "learning_rate": 1.1907953529937444e-05,
+      "loss": 0.2889,
+      "step": 1954
+    },
+    {
+      "epoch": 2.3582629674306395,
+      "grad_norm": 0.23040423833861387,
+      "learning_rate": 1.188561215370867e-05,
+      "loss": 0.2832,
+      "step": 1955
+    },
+    {
+      "epoch": 2.359469240048251,
+      "grad_norm": 0.22536151153551331,
+      "learning_rate": 1.1863270777479892e-05,
+      "loss": 0.284,
+      "step": 1956
+    },
+    {
+      "epoch": 2.3606755126658623,
+      "grad_norm": 0.21133558876391853,
+      "learning_rate": 1.1840929401251118e-05,
+      "loss": 0.2949,
+      "step": 1957
+    },
+    {
+      "epoch": 2.361881785283474,
+      "grad_norm": 0.22376185910186625,
+      "learning_rate": 1.1818588025022342e-05,
+      "loss": 0.2949,
+      "step": 1958
+    },
+    {
+      "epoch": 2.3630880579010856,
+      "grad_norm": 0.21550391712375538,
+      "learning_rate": 1.1796246648793566e-05,
+      "loss": 0.2767,
+      "step": 1959
+    },
+    {
+      "epoch": 2.3642943305186974,
+      "grad_norm": 0.21187486236001407,
+      "learning_rate": 1.177390527256479e-05,
+      "loss": 0.2948,
+      "step": 1960
+    },
+    {
+      "epoch": 2.365500603136309,
+      "grad_norm": 0.23091281680192807,
+      "learning_rate": 1.1751563896336016e-05,
+      "loss": 0.2961,
+      "step": 1961
+    },
+    {
+      "epoch": 2.3667068757539202,
+      "grad_norm": 0.23769114515813178,
+      "learning_rate": 1.1729222520107238e-05,
+      "loss": 0.2665,
+      "step": 1962
+    },
+    {
+      "epoch": 2.367913148371532,
+      "grad_norm": 0.22891068293328487,
+      "learning_rate": 1.1706881143878464e-05,
+      "loss": 0.2807,
+      "step": 1963
+    },
+    {
+      "epoch": 2.3691194209891435,
+      "grad_norm": 0.24591993183344646,
+      "learning_rate": 1.1684539767649688e-05,
+      "loss": 0.2877,
+      "step": 1964
+    },
+    {
+      "epoch": 2.370325693606755,
+      "grad_norm": 0.25582946324967376,
+      "learning_rate": 1.166219839142091e-05,
+      "loss": 0.2923,
+      "step": 1965
+    },
+    {
+      "epoch": 2.371531966224367,
+      "grad_norm": 0.22913321677896634,
+      "learning_rate": 1.1639857015192136e-05,
+      "loss": 0.2889,
+      "step": 1966
+    },
+    {
+      "epoch": 2.372738238841978,
+      "grad_norm": 0.246974008296304,
+      "learning_rate": 1.161751563896336e-05,
+      "loss": 0.2941,
+      "step": 1967
+    },
+    {
+      "epoch": 2.37394451145959,
+      "grad_norm": 0.2277084018677113,
+      "learning_rate": 1.1595174262734585e-05,
+      "loss": 0.2888,
+      "step": 1968
+    },
+    {
+      "epoch": 2.3751507840772015,
+      "grad_norm": 0.23156695668294586,
+      "learning_rate": 1.1572832886505809e-05,
+      "loss": 0.2833,
+      "step": 1969
+    },
+    {
+      "epoch": 2.376357056694813,
+      "grad_norm": 0.2525648596511969,
+      "learning_rate": 1.1550491510277035e-05,
+      "loss": 0.2798,
+      "step": 1970
+    },
+    {
+      "epoch": 2.3775633293124248,
+      "grad_norm": 0.2205192096748547,
+      "learning_rate": 1.1528150134048257e-05,
+      "loss": 0.2825,
+      "step": 1971
+    },
+    {
+      "epoch": 2.378769601930036,
+      "grad_norm": 0.23904424195465934,
+      "learning_rate": 1.1505808757819483e-05,
+      "loss": 0.2951,
+      "step": 1972
+    },
+    {
+      "epoch": 2.3799758745476476,
+      "grad_norm": 0.2362564226232104,
+      "learning_rate": 1.1483467381590707e-05,
+      "loss": 0.2992,
+      "step": 1973
+    },
+    {
+      "epoch": 2.3811821471652594,
+      "grad_norm": 0.23386421917567346,
+      "learning_rate": 1.1461126005361931e-05,
+      "loss": 0.2956,
+      "step": 1974
+    },
+    {
+      "epoch": 2.382388419782871,
+      "grad_norm": 0.26306584218072937,
+      "learning_rate": 1.1438784629133155e-05,
+      "loss": 0.2964,
+      "step": 1975
+    },
+    {
+      "epoch": 2.3835946924004823,
+      "grad_norm": 0.22880228551731152,
+      "learning_rate": 1.1416443252904379e-05,
+      "loss": 0.2986,
+      "step": 1976
+    },
+    {
+      "epoch": 2.384800965018094,
+      "grad_norm": 0.20846641089392354,
+      "learning_rate": 1.1394101876675605e-05,
+      "loss": 0.2926,
+      "step": 1977
+    },
+    {
+      "epoch": 2.3860072376357055,
+      "grad_norm": 0.23144685907710733,
+      "learning_rate": 1.1371760500446827e-05,
+      "loss": 0.2813,
+      "step": 1978
+    },
+    {
+      "epoch": 2.3872135102533174,
+      "grad_norm": 0.22403595755031833,
+      "learning_rate": 1.1349419124218053e-05,
+      "loss": 0.2779,
+      "step": 1979
+    },
+    {
+      "epoch": 2.388419782870929,
+      "grad_norm": 0.23705522152129233,
+      "learning_rate": 1.1327077747989277e-05,
+      "loss": 0.2913,
+      "step": 1980
+    },
+    {
+      "epoch": 2.3896260554885402,
+      "grad_norm": 0.20647613466526085,
+      "learning_rate": 1.1304736371760501e-05,
+      "loss": 0.2853,
+      "step": 1981
+    },
+    {
+      "epoch": 2.390832328106152,
+      "grad_norm": 0.2047928540204018,
+      "learning_rate": 1.1282394995531725e-05,
+      "loss": 0.2828,
+      "step": 1982
+    },
+    {
+      "epoch": 2.3920386007237635,
+      "grad_norm": 0.21929372924006985,
+      "learning_rate": 1.126005361930295e-05,
+      "loss": 0.2859,
+      "step": 1983
+    },
+    {
+      "epoch": 2.3932448733413754,
+      "grad_norm": 0.21794959265971092,
+      "learning_rate": 1.1237712243074173e-05,
+      "loss": 0.2765,
+      "step": 1984
+    },
+    {
+      "epoch": 2.394451145958987,
+      "grad_norm": 0.2316415929799842,
+      "learning_rate": 1.1215370866845398e-05,
+      "loss": 0.2775,
+      "step": 1985
+    },
+    {
+      "epoch": 2.395657418576598,
+      "grad_norm": 0.23705174878505497,
+      "learning_rate": 1.1193029490616623e-05,
+      "loss": 0.3013,
+      "step": 1986
+    },
+    {
+      "epoch": 2.39686369119421,
+      "grad_norm": 0.22799762544468544,
+      "learning_rate": 1.1170688114387846e-05,
+      "loss": 0.2877,
+      "step": 1987
+    },
+    {
+      "epoch": 2.3980699638118215,
+      "grad_norm": 0.21882815233428807,
+      "learning_rate": 1.1148346738159071e-05,
+      "loss": 0.2845,
+      "step": 1988
+    },
+    {
+      "epoch": 2.399276236429433,
+      "grad_norm": 0.20344632297160425,
+      "learning_rate": 1.1126005361930296e-05,
+      "loss": 0.2833,
+      "step": 1989
+    },
+    {
+      "epoch": 2.4004825090470447,
+      "grad_norm": 0.23503552618475457,
+      "learning_rate": 1.110366398570152e-05,
+      "loss": 0.2872,
+      "step": 1990
+    },
+    {
+      "epoch": 2.401688781664656,
+      "grad_norm": 0.22374098805502607,
+      "learning_rate": 1.1081322609472744e-05,
+      "loss": 0.2855,
+      "step": 1991
+    },
+    {
+      "epoch": 2.4028950542822676,
+      "grad_norm": 0.21020918399892033,
+      "learning_rate": 1.105898123324397e-05,
+      "loss": 0.2876,
+      "step": 1992
+    },
+    {
+      "epoch": 2.4041013268998794,
+      "grad_norm": 0.23373376474000448,
+      "learning_rate": 1.1036639857015192e-05,
+      "loss": 0.307,
+      "step": 1993
+    },
+    {
+      "epoch": 2.405307599517491,
+      "grad_norm": 0.2104550476556823,
+      "learning_rate": 1.1014298480786418e-05,
+      "loss": 0.2896,
+      "step": 1994
+    },
+    {
+      "epoch": 2.4065138721351027,
+      "grad_norm": 0.20137896629238947,
+      "learning_rate": 1.0991957104557642e-05,
+      "loss": 0.2981,
+      "step": 1995
+    },
+    {
+      "epoch": 2.407720144752714,
+      "grad_norm": 0.23012273495698185,
+      "learning_rate": 1.0969615728328864e-05,
+      "loss": 0.2843,
+      "step": 1996
+    },
+    {
+      "epoch": 2.4089264173703255,
+      "grad_norm": 0.20650123033912035,
+      "learning_rate": 1.094727435210009e-05,
+      "loss": 0.2853,
+      "step": 1997
+    },
+    {
+      "epoch": 2.4101326899879374,
+      "grad_norm": 0.21987946438527867,
+      "learning_rate": 1.0924932975871314e-05,
+      "loss": 0.2751,
+      "step": 1998
+    },
+    {
+      "epoch": 2.411338962605549,
+      "grad_norm": 0.23424456071416347,
+      "learning_rate": 1.0902591599642538e-05,
+      "loss": 0.3035,
+      "step": 1999
+    },
+    {
+      "epoch": 2.4125452352231607,
+      "grad_norm": 0.2149327854552427,
+      "learning_rate": 1.0880250223413762e-05,
+      "loss": 0.2925,
+      "step": 2000
+    },
+    {
+      "epoch": 2.413751507840772,
+      "grad_norm": 0.2084100213854339,
+      "learning_rate": 1.0857908847184988e-05,
+      "loss": 0.2747,
+      "step": 2001
+    },
+    {
+      "epoch": 2.4149577804583835,
+      "grad_norm": 0.23570811547273146,
+      "learning_rate": 1.083556747095621e-05,
+      "loss": 0.2906,
+      "step": 2002
+    },
+    {
+      "epoch": 2.4161640530759954,
+      "grad_norm": 0.2257613385126158,
+      "learning_rate": 1.0813226094727436e-05,
+      "loss": 0.3031,
+      "step": 2003
+    },
+    {
+      "epoch": 2.4173703256936068,
+      "grad_norm": 0.2365614982781767,
+      "learning_rate": 1.079088471849866e-05,
+      "loss": 0.2948,
+      "step": 2004
+    },
+    {
+      "epoch": 2.418576598311218,
+      "grad_norm": 0.22036504428958142,
+      "learning_rate": 1.0768543342269884e-05,
+      "loss": 0.2899,
+      "step": 2005
+    },
+    {
+      "epoch": 2.41978287092883,
+      "grad_norm": 0.22931384043977412,
+      "learning_rate": 1.0746201966041108e-05,
+      "loss": 0.2957,
+      "step": 2006
+    },
+    {
+      "epoch": 2.4209891435464415,
+      "grad_norm": 0.22648620005109288,
+      "learning_rate": 1.0723860589812333e-05,
+      "loss": 0.2921,
+      "step": 2007
+    },
+    {
+      "epoch": 2.422195416164053,
+      "grad_norm": 0.22020345396728686,
+      "learning_rate": 1.0701519213583558e-05,
+      "loss": 0.2889,
+      "step": 2008
+    },
+    {
+      "epoch": 2.4234016887816647,
+      "grad_norm": 0.23024407336074107,
+      "learning_rate": 1.067917783735478e-05,
+      "loss": 0.292,
+      "step": 2009
+    },
+    {
+      "epoch": 2.424607961399276,
+      "grad_norm": 0.21920021299966,
+      "learning_rate": 1.0656836461126007e-05,
+      "loss": 0.2792,
+      "step": 2010
+    },
+    {
+      "epoch": 2.425814234016888,
+      "grad_norm": 0.22348139341258205,
+      "learning_rate": 1.063449508489723e-05,
+      "loss": 0.2981,
+      "step": 2011
+    },
+    {
+      "epoch": 2.4270205066344994,
+      "grad_norm": 0.21145115022972605,
+      "learning_rate": 1.0612153708668455e-05,
+      "loss": 0.2807,
+      "step": 2012
+    },
+    {
+      "epoch": 2.428226779252111,
+      "grad_norm": 0.22074860399576737,
+      "learning_rate": 1.0589812332439679e-05,
+      "loss": 0.3048,
+      "step": 2013
+    },
+    {
+      "epoch": 2.4294330518697227,
+      "grad_norm": 0.2529900639870893,
+      "learning_rate": 1.0567470956210903e-05,
+      "loss": 0.2811,
+      "step": 2014
+    },
+    {
+      "epoch": 2.430639324487334,
+      "grad_norm": 0.2166067972140964,
+      "learning_rate": 1.0545129579982127e-05,
+      "loss": 0.2796,
+      "step": 2015
+    },
+    {
+      "epoch": 2.431845597104946,
+      "grad_norm": 0.21597450621839834,
+      "learning_rate": 1.0522788203753351e-05,
+      "loss": 0.2978,
+      "step": 2016
+    },
+    {
+      "epoch": 2.4330518697225574,
+      "grad_norm": 0.24619094911709458,
+      "learning_rate": 1.0500446827524577e-05,
+      "loss": 0.2933,
+      "step": 2017
+    },
+    {
+      "epoch": 2.434258142340169,
+      "grad_norm": 0.22294280116564494,
+      "learning_rate": 1.04781054512958e-05,
+      "loss": 0.2899,
+      "step": 2018
+    },
+    {
+      "epoch": 2.4354644149577807,
+      "grad_norm": 0.2218284276829594,
+      "learning_rate": 1.0455764075067025e-05,
+      "loss": 0.2946,
+      "step": 2019
+    },
+    {
+      "epoch": 2.436670687575392,
+      "grad_norm": 0.25621556450007477,
+      "learning_rate": 1.0433422698838249e-05,
+      "loss": 0.2868,
+      "step": 2020
+    },
+    {
+      "epoch": 2.4378769601930035,
+      "grad_norm": 0.22991632882380855,
+      "learning_rate": 1.0411081322609473e-05,
+      "loss": 0.2895,
+      "step": 2021
+    },
+    {
+      "epoch": 2.4390832328106153,
+      "grad_norm": 0.2260095543571356,
+      "learning_rate": 1.0388739946380697e-05,
+      "loss": 0.3088,
+      "step": 2022
+    },
+    {
+      "epoch": 2.4402895054282268,
+      "grad_norm": 0.22973240522262808,
+      "learning_rate": 1.0366398570151923e-05,
+      "loss": 0.2935,
+      "step": 2023
+    },
+    {
+      "epoch": 2.441495778045838,
+      "grad_norm": 0.23705590633564524,
+      "learning_rate": 1.0344057193923145e-05,
+      "loss": 0.2916,
+      "step": 2024
+    },
+    {
+      "epoch": 2.44270205066345,
+      "grad_norm": 0.2514158757053114,
+      "learning_rate": 1.0321715817694371e-05,
+      "loss": 0.2817,
+      "step": 2025
+    },
+    {
+      "epoch": 2.4439083232810614,
+      "grad_norm": 0.2198108837195466,
+      "learning_rate": 1.0299374441465595e-05,
+      "loss": 0.2884,
+      "step": 2026
+    },
+    {
+      "epoch": 2.445114595898673,
+      "grad_norm": 0.2504281500366312,
+      "learning_rate": 1.0277033065236818e-05,
+      "loss": 0.2989,
+      "step": 2027
+    },
+    {
+      "epoch": 2.4463208685162847,
+      "grad_norm": 0.2504968434548994,
+      "learning_rate": 1.0254691689008044e-05,
+      "loss": 0.2914,
+      "step": 2028
+    },
+    {
+      "epoch": 2.447527141133896,
+      "grad_norm": 0.21234515057169162,
+      "learning_rate": 1.0232350312779268e-05,
+      "loss": 0.2799,
+      "step": 2029
+    },
+    {
+      "epoch": 2.448733413751508,
+      "grad_norm": 0.21297856434093013,
+      "learning_rate": 1.0210008936550492e-05,
+      "loss": 0.2785,
+      "step": 2030
+    },
+    {
+      "epoch": 2.4499396863691194,
+      "grad_norm": 0.21905395061242522,
+      "learning_rate": 1.0187667560321716e-05,
+      "loss": 0.2738,
+      "step": 2031
+    },
+    {
+      "epoch": 2.451145958986731,
+      "grad_norm": 0.23870861408031346,
+      "learning_rate": 1.0165326184092942e-05,
+      "loss": 0.2829,
+      "step": 2032
+    },
+    {
+      "epoch": 2.4523522316043427,
+      "grad_norm": 0.2110603313911652,
+      "learning_rate": 1.0142984807864164e-05,
+      "loss": 0.2907,
+      "step": 2033
+    },
+    {
+      "epoch": 2.453558504221954,
+      "grad_norm": 0.2079993100221868,
+      "learning_rate": 1.012064343163539e-05,
+      "loss": 0.2797,
+      "step": 2034
+    },
+    {
+      "epoch": 2.454764776839566,
+      "grad_norm": 0.19952965390672966,
+      "learning_rate": 1.0098302055406614e-05,
+      "loss": 0.2821,
+      "step": 2035
+    },
+    {
+      "epoch": 2.4559710494571774,
+      "grad_norm": 0.3063957219989011,
+      "learning_rate": 1.0075960679177838e-05,
+      "loss": 0.2916,
+      "step": 2036
+    },
+    {
+      "epoch": 2.457177322074789,
+      "grad_norm": 0.2573913236576673,
+      "learning_rate": 1.0053619302949062e-05,
+      "loss": 0.2936,
+      "step": 2037
+    },
+    {
+      "epoch": 2.4583835946924006,
+      "grad_norm": 0.23628800661282728,
+      "learning_rate": 1.0031277926720286e-05,
+      "loss": 0.2997,
+      "step": 2038
+    },
+    {
+      "epoch": 2.459589867310012,
+      "grad_norm": 0.3707608999179138,
+      "learning_rate": 1.0008936550491512e-05,
+      "loss": 0.2921,
+      "step": 2039
+    },
+    {
+      "epoch": 2.4607961399276235,
+      "grad_norm": 0.2355724932937776,
+      "learning_rate": 9.986595174262734e-06,
+      "loss": 0.2842,
+      "step": 2040
+    },
+    {
+      "epoch": 2.4620024125452353,
+      "grad_norm": 0.20979675529449615,
+      "learning_rate": 9.96425379803396e-06,
+      "loss": 0.2745,
+      "step": 2041
+    },
+    {
+      "epoch": 2.4632086851628467,
+      "grad_norm": 0.23092304983801692,
+      "learning_rate": 9.941912421805184e-06,
+      "loss": 0.2991,
+      "step": 2042
+    },
+    {
+      "epoch": 2.464414957780458,
+      "grad_norm": 0.23268676022732815,
+      "learning_rate": 9.919571045576408e-06,
+      "loss": 0.2839,
+      "step": 2043
+    },
+    {
+      "epoch": 2.46562123039807,
+      "grad_norm": 0.22396777883699703,
+      "learning_rate": 9.897229669347632e-06,
+      "loss": 0.2968,
+      "step": 2044
+    },
+    {
+      "epoch": 2.4668275030156814,
+      "grad_norm": 0.26078276763958036,
+      "learning_rate": 9.874888293118856e-06,
+      "loss": 0.2971,
+      "step": 2045
+    },
+    {
+      "epoch": 2.4680337756332933,
+      "grad_norm": 0.24364965152903365,
+      "learning_rate": 9.85254691689008e-06,
+      "loss": 0.2895,
+      "step": 2046
+    },
+    {
+      "epoch": 2.4692400482509047,
+      "grad_norm": 0.24028537875953107,
+      "learning_rate": 9.830205540661305e-06,
+      "loss": 0.2878,
+      "step": 2047
+    },
+    {
+      "epoch": 2.470446320868516,
+      "grad_norm": 0.22503830564530097,
+      "learning_rate": 9.80786416443253e-06,
+      "loss": 0.2823,
+      "step": 2048
+    },
+    {
+      "epoch": 2.471652593486128,
+      "grad_norm": 0.23918310040715698,
+      "learning_rate": 9.785522788203753e-06,
+      "loss": 0.2976,
+      "step": 2049
+    },
+    {
+      "epoch": 2.4728588661037394,
+      "grad_norm": 0.23847496984436617,
+      "learning_rate": 9.763181411974979e-06,
+      "loss": 0.2841,
+      "step": 2050
+    },
+    {
+      "epoch": 2.4740651387213513,
+      "grad_norm": 0.21340888253047746,
+      "learning_rate": 9.740840035746203e-06,
+      "loss": 0.2751,
+      "step": 2051
+    },
+    {
+      "epoch": 2.4752714113389627,
+      "grad_norm": 0.27284485919627044,
+      "learning_rate": 9.718498659517427e-06,
+      "loss": 0.2983,
+      "step": 2052
+    },
+    {
+      "epoch": 2.476477683956574,
+      "grad_norm": 0.23262515940334016,
+      "learning_rate": 9.69615728328865e-06,
+      "loss": 0.2893,
+      "step": 2053
+    },
+    {
+      "epoch": 2.477683956574186,
+      "grad_norm": 0.21320522352667567,
+      "learning_rate": 9.673815907059877e-06,
+      "loss": 0.2904,
+      "step": 2054
+    },
+    {
+      "epoch": 2.4788902291917974,
+      "grad_norm": 0.21281048078611334,
+      "learning_rate": 9.651474530831099e-06,
+      "loss": 0.2783,
+      "step": 2055
+    },
+    {
+      "epoch": 2.4800965018094088,
+      "grad_norm": 0.24467691806482708,
+      "learning_rate": 9.629133154602323e-06,
+      "loss": 0.2854,
+      "step": 2056
+    },
+    {
+      "epoch": 2.4813027744270206,
+      "grad_norm": 0.253676597589621,
+      "learning_rate": 9.606791778373549e-06,
+      "loss": 0.3071,
+      "step": 2057
+    },
+    {
+      "epoch": 2.482509047044632,
+      "grad_norm": 0.23522323243807983,
+      "learning_rate": 9.584450402144771e-06,
+      "loss": 0.2883,
+      "step": 2058
+    },
+    {
+      "epoch": 2.4837153196622435,
+      "grad_norm": 0.21968405569335833,
+      "learning_rate": 9.562109025915997e-06,
+      "loss": 0.293,
+      "step": 2059
+    },
+    {
+      "epoch": 2.4849215922798553,
+      "grad_norm": 0.23336070313958515,
+      "learning_rate": 9.539767649687221e-06,
+      "loss": 0.2996,
+      "step": 2060
+    },
+    {
+      "epoch": 2.4861278648974667,
+      "grad_norm": 0.23574400321619537,
+      "learning_rate": 9.517426273458445e-06,
+      "loss": 0.319,
+      "step": 2061
+    },
+    {
+      "epoch": 2.4873341375150786,
+      "grad_norm": 0.25284110850110175,
+      "learning_rate": 9.49508489722967e-06,
+      "loss": 0.3039,
+      "step": 2062
+    },
+    {
+      "epoch": 2.48854041013269,
+      "grad_norm": 0.22609295713824532,
+      "learning_rate": 9.472743521000895e-06,
+      "loss": 0.2889,
+      "step": 2063
+    },
+    {
+      "epoch": 2.4897466827503014,
+      "grad_norm": 0.21782544942151286,
+      "learning_rate": 9.450402144772117e-06,
+      "loss": 0.2773,
+      "step": 2064
+    },
+    {
+      "epoch": 2.4909529553679133,
+      "grad_norm": 0.24395759916835988,
+      "learning_rate": 9.428060768543343e-06,
+      "loss": 0.2872,
+      "step": 2065
+    },
+    {
+      "epoch": 2.4921592279855247,
+      "grad_norm": 0.2100854112198129,
+      "learning_rate": 9.405719392314567e-06,
+      "loss": 0.2683,
+      "step": 2066
+    },
+    {
+      "epoch": 2.4933655006031366,
+      "grad_norm": 0.22018277045218249,
+      "learning_rate": 9.383378016085791e-06,
+      "loss": 0.2989,
+      "step": 2067
+    },
+    {
+      "epoch": 2.494571773220748,
+      "grad_norm": 0.22632341437009562,
+      "learning_rate": 9.361036639857016e-06,
+      "loss": 0.2764,
+      "step": 2068
+    },
+    {
+      "epoch": 2.4957780458383594,
+      "grad_norm": 0.23134887518421296,
+      "learning_rate": 9.33869526362824e-06,
+      "loss": 0.2621,
+      "step": 2069
+    },
+    {
+      "epoch": 2.4969843184559712,
+      "grad_norm": 0.2259347845947623,
+      "learning_rate": 9.316353887399465e-06,
+      "loss": 0.3099,
+      "step": 2070
+    },
+    {
+      "epoch": 2.4981905910735827,
+      "grad_norm": 0.21947311486952517,
+      "learning_rate": 9.294012511170688e-06,
+      "loss": 0.2836,
+      "step": 2071
+    },
+    {
+      "epoch": 2.499396863691194,
+      "grad_norm": 0.21153193775768228,
+      "learning_rate": 9.271671134941914e-06,
+      "loss": 0.2656,
+      "step": 2072
+    },
+    {
+      "epoch": 2.500603136308806,
+      "grad_norm": 0.22972463052223152,
+      "learning_rate": 9.249329758713138e-06,
+      "loss": 0.2958,
+      "step": 2073
+    },
+    {
+      "epoch": 2.5018094089264173,
+      "grad_norm": 0.22427297366530236,
+      "learning_rate": 9.226988382484362e-06,
+      "loss": 0.2846,
+      "step": 2074
+    },
+    {
+      "epoch": 2.5030156815440288,
+      "grad_norm": 0.2188475129476444,
+      "learning_rate": 9.204647006255586e-06,
+      "loss": 0.2822,
+      "step": 2075
+    },
+    {
+      "epoch": 2.5042219541616406,
+      "grad_norm": 0.22157141957950305,
+      "learning_rate": 9.18230563002681e-06,
+      "loss": 0.2859,
+      "step": 2076
+    },
+    {
+      "epoch": 2.505428226779252,
+      "grad_norm": 0.20369669117130704,
+      "learning_rate": 9.159964253798034e-06,
+      "loss": 0.2623,
+      "step": 2077
+    },
+    {
+      "epoch": 2.5066344993968634,
+      "grad_norm": 0.26069163206394774,
+      "learning_rate": 9.137622877569258e-06,
+      "loss": 0.3071,
+      "step": 2078
+    },
+    {
+      "epoch": 2.5078407720144753,
+      "grad_norm": 0.21427683157651178,
+      "learning_rate": 9.115281501340484e-06,
+      "loss": 0.2855,
+      "step": 2079
+    },
+    {
+      "epoch": 2.5090470446320867,
+      "grad_norm": 0.20008581879419765,
+      "learning_rate": 9.092940125111706e-06,
+      "loss": 0.2815,
+      "step": 2080
+    },
+    {
+      "epoch": 2.5102533172496986,
+      "grad_norm": 0.24743260230654526,
+      "learning_rate": 9.070598748882932e-06,
+      "loss": 0.2882,
+      "step": 2081
+    },
+    {
+      "epoch": 2.51145958986731,
+      "grad_norm": 0.234746962743461,
+      "learning_rate": 9.048257372654156e-06,
+      "loss": 0.2867,
+      "step": 2082
+    },
+    {
+      "epoch": 2.512665862484922,
+      "grad_norm": 0.23547248016350894,
+      "learning_rate": 9.02591599642538e-06,
+      "loss": 0.2846,
+      "step": 2083
+    },
+    {
+      "epoch": 2.5138721351025333,
+      "grad_norm": 0.22598658102916713,
+      "learning_rate": 9.003574620196604e-06,
+      "loss": 0.295,
+      "step": 2084
+    },
+    {
+      "epoch": 2.5150784077201447,
+      "grad_norm": 0.2464188752316304,
+      "learning_rate": 8.98123324396783e-06,
+      "loss": 0.2962,
+      "step": 2085
+    },
+    {
+      "epoch": 2.5162846803377565,
+      "grad_norm": 0.20823067381368154,
+      "learning_rate": 8.958891867739052e-06,
+      "loss": 0.2761,
+      "step": 2086
+    },
+    {
+      "epoch": 2.517490952955368,
+      "grad_norm": 0.2450510002186486,
+      "learning_rate": 8.936550491510277e-06,
+      "loss": 0.2865,
+      "step": 2087
+    },
+    {
+      "epoch": 2.5186972255729794,
+      "grad_norm": 0.24205249299984813,
+      "learning_rate": 8.914209115281502e-06,
+      "loss": 0.3093,
+      "step": 2088
+    },
+    {
+      "epoch": 2.5199034981905912,
+      "grad_norm": 0.24457280923901634,
+      "learning_rate": 8.891867739052725e-06,
+      "loss": 0.2896,
+      "step": 2089
+    },
+    {
+      "epoch": 2.5211097708082026,
+      "grad_norm": 0.21212091302235664,
+      "learning_rate": 8.86952636282395e-06,
+      "loss": 0.2833,
+      "step": 2090
+    },
+    {
+      "epoch": 2.522316043425814,
+      "grad_norm": 0.23580463237099886,
+      "learning_rate": 8.847184986595175e-06,
+      "loss": 0.2784,
+      "step": 2091
+    },
+    {
+      "epoch": 2.523522316043426,
+      "grad_norm": 0.2388276288521073,
+      "learning_rate": 8.824843610366399e-06,
+      "loss": 0.2906,
+      "step": 2092
+    },
+    {
+      "epoch": 2.5247285886610373,
+      "grad_norm": 0.22087481614885304,
+      "learning_rate": 8.802502234137623e-06,
+      "loss": 0.2763,
+      "step": 2093
+    },
+    {
+      "epoch": 2.5259348612786487,
+      "grad_norm": 0.2246523474210579,
+      "learning_rate": 8.780160857908849e-06,
+      "loss": 0.2842,
+      "step": 2094
+    },
+    {
+      "epoch": 2.5271411338962606,
+      "grad_norm": 0.23119978686478482,
+      "learning_rate": 8.757819481680071e-06,
+      "loss": 0.2892,
+      "step": 2095
+    },
+    {
+      "epoch": 2.528347406513872,
+      "grad_norm": 0.23233171332683422,
+      "learning_rate": 8.735478105451297e-06,
+      "loss": 0.2907,
+      "step": 2096
+    },
+    {
+      "epoch": 2.529553679131484,
+      "grad_norm": 0.21942382453851111,
+      "learning_rate": 8.71313672922252e-06,
+      "loss": 0.2872,
+      "step": 2097
+    },
+    {
+      "epoch": 2.5307599517490953,
+      "grad_norm": 0.2413498290860737,
+      "learning_rate": 8.690795352993745e-06,
+      "loss": 0.2875,
+      "step": 2098
+    },
+    {
+      "epoch": 2.531966224366707,
+      "grad_norm": 0.23083935116777204,
+      "learning_rate": 8.668453976764969e-06,
+      "loss": 0.2762,
+      "step": 2099
+    },
+    {
+      "epoch": 2.5331724969843186,
+      "grad_norm": 0.21390651009006564,
+      "learning_rate": 8.646112600536193e-06,
+      "loss": 0.2912,
+      "step": 2100
+    },
+    {
+      "epoch": 2.53437876960193,
+      "grad_norm": 0.2365999128822495,
+      "learning_rate": 8.623771224307419e-06,
+      "loss": 0.2918,
+      "step": 2101
+    },
+    {
+      "epoch": 2.535585042219542,
+      "grad_norm": 0.2322084809252393,
+      "learning_rate": 8.601429848078641e-06,
+      "loss": 0.2866,
+      "step": 2102
+    },
+    {
+      "epoch": 2.5367913148371533,
+      "grad_norm": 0.21403628042564693,
+      "learning_rate": 8.579088471849867e-06,
+      "loss": 0.2715,
+      "step": 2103
+    },
+    {
+      "epoch": 2.5379975874547647,
+      "grad_norm": 0.23633271492330205,
+      "learning_rate": 8.556747095621091e-06,
+      "loss": 0.3122,
+      "step": 2104
+    },
+    {
+      "epoch": 2.5392038600723765,
+      "grad_norm": 0.2171261190451419,
+      "learning_rate": 8.534405719392315e-06,
+      "loss": 0.2669,
+      "step": 2105
+    },
+    {
+      "epoch": 2.540410132689988,
+      "grad_norm": 0.20562237412666165,
+      "learning_rate": 8.51206434316354e-06,
+      "loss": 0.2838,
+      "step": 2106
+    },
+    {
+      "epoch": 2.5416164053075994,
+      "grad_norm": 0.22643156797827918,
+      "learning_rate": 8.489722966934763e-06,
+      "loss": 0.2767,
+      "step": 2107
+    },
+    {
+      "epoch": 2.542822677925211,
+      "grad_norm": 0.2482978387199731,
+      "learning_rate": 8.467381590705988e-06,
+      "loss": 0.2795,
+      "step": 2108
+    },
+    {
+      "epoch": 2.5440289505428226,
+      "grad_norm": 0.2221033509722392,
+      "learning_rate": 8.445040214477212e-06,
+      "loss": 0.2896,
+      "step": 2109
+    },
+    {
+      "epoch": 2.545235223160434,
+      "grad_norm": 0.21304030830619042,
+      "learning_rate": 8.422698838248437e-06,
+      "loss": 0.2899,
+      "step": 2110
+    },
+    {
+      "epoch": 2.546441495778046,
+      "grad_norm": 0.21298340995217513,
+      "learning_rate": 8.40035746201966e-06,
+      "loss": 0.2985,
+      "step": 2111
+    },
+    {
+      "epoch": 2.5476477683956573,
+      "grad_norm": 0.2268107049383085,
+      "learning_rate": 8.378016085790886e-06,
+      "loss": 0.2876,
+      "step": 2112
+    },
+    {
+      "epoch": 2.5488540410132687,
+      "grad_norm": 0.2283769777933946,
+      "learning_rate": 8.35567470956211e-06,
+      "loss": 0.2937,
+      "step": 2113
+    },
+    {
+      "epoch": 2.5500603136308806,
+      "grad_norm": 0.20201400435301958,
+      "learning_rate": 8.333333333333334e-06,
+      "loss": 0.2736,
+      "step": 2114
+    },
+    {
+      "epoch": 2.551266586248492,
+      "grad_norm": 0.21118648069579432,
+      "learning_rate": 8.310991957104558e-06,
+      "loss": 0.2801,
+      "step": 2115
+    },
+    {
+      "epoch": 2.552472858866104,
+      "grad_norm": 0.21380147734469018,
+      "learning_rate": 8.288650580875784e-06,
+      "loss": 0.2879,
+      "step": 2116
+    },
+    {
+      "epoch": 2.5536791314837153,
+      "grad_norm": 0.24427500368171587,
+      "learning_rate": 8.266309204647006e-06,
+      "loss": 0.2828,
+      "step": 2117
+    },
+    {
+      "epoch": 2.554885404101327,
+      "grad_norm": 0.2104444247720201,
+      "learning_rate": 8.24396782841823e-06,
+      "loss": 0.284,
+      "step": 2118
+    },
+    {
+      "epoch": 2.5560916767189386,
+      "grad_norm": 0.21710117554724775,
+      "learning_rate": 8.221626452189456e-06,
+      "loss": 0.2864,
+      "step": 2119
+    },
+    {
+      "epoch": 2.55729794933655,
+      "grad_norm": 0.22717244072496412,
+      "learning_rate": 8.199285075960678e-06,
+      "loss": 0.2781,
+      "step": 2120
+    },
+    {
+      "epoch": 2.558504221954162,
+      "grad_norm": 0.2254436961531605,
+      "learning_rate": 8.176943699731904e-06,
+      "loss": 0.2782,
+      "step": 2121
+    },
+    {
+      "epoch": 2.5597104945717732,
+      "grad_norm": 0.21519589525491942,
+      "learning_rate": 8.154602323503128e-06,
+      "loss": 0.297,
+      "step": 2122
+    },
+    {
+      "epoch": 2.5609167671893847,
+      "grad_norm": 0.2320770217495183,
+      "learning_rate": 8.132260947274352e-06,
+      "loss": 0.3072,
+      "step": 2123
+    },
+    {
+      "epoch": 2.5621230398069965,
+      "grad_norm": 0.213745103139696,
+      "learning_rate": 8.109919571045576e-06,
+      "loss": 0.2864,
+      "step": 2124
+    },
+    {
+      "epoch": 2.563329312424608,
+      "grad_norm": 0.21520826517792885,
+      "learning_rate": 8.087578194816802e-06,
+      "loss": 0.2772,
+      "step": 2125
+    },
+    {
+      "epoch": 2.5645355850422193,
+      "grad_norm": 0.23267029432546807,
+      "learning_rate": 8.065236818588024e-06,
+      "loss": 0.2934,
+      "step": 2126
+    },
+    {
+      "epoch": 2.565741857659831,
+      "grad_norm": 0.2346619111675145,
+      "learning_rate": 8.04289544235925e-06,
+      "loss": 0.2997,
+      "step": 2127
+    },
+    {
+      "epoch": 2.5669481302774426,
+      "grad_norm": 0.21089577135395482,
+      "learning_rate": 8.020554066130474e-06,
+      "loss": 0.291,
+      "step": 2128
+    },
+    {
+      "epoch": 2.568154402895054,
+      "grad_norm": 0.23237289510955644,
+      "learning_rate": 7.998212689901698e-06,
+      "loss": 0.2821,
+      "step": 2129
+    },
+    {
+      "epoch": 2.569360675512666,
+      "grad_norm": 0.22800080953514776,
+      "learning_rate": 7.975871313672923e-06,
+      "loss": 0.2965,
+      "step": 2130
+    },
+    {
+      "epoch": 2.5705669481302773,
+      "grad_norm": 0.20687732257395336,
+      "learning_rate": 7.953529937444147e-06,
+      "loss": 0.2843,
+      "step": 2131
+    },
+    {
+      "epoch": 2.571773220747889,
+      "grad_norm": 0.2252633109054214,
+      "learning_rate": 7.931188561215372e-06,
+      "loss": 0.2774,
+      "step": 2132
+    },
+    {
+      "epoch": 2.5729794933655006,
+      "grad_norm": 0.21503546781105598,
+      "learning_rate": 7.908847184986595e-06,
+      "loss": 0.2943,
+      "step": 2133
+    },
+    {
+      "epoch": 2.5741857659831124,
+      "grad_norm": 0.2103849220087177,
+      "learning_rate": 7.88650580875782e-06,
+      "loss": 0.2823,
+      "step": 2134
+    },
+    {
+      "epoch": 2.575392038600724,
+      "grad_norm": 0.2356664422191725,
+      "learning_rate": 7.864164432529045e-06,
+      "loss": 0.2865,
+      "step": 2135
+    },
+    {
+      "epoch": 2.5765983112183353,
+      "grad_norm": 0.21854984713818312,
+      "learning_rate": 7.841823056300269e-06,
+      "loss": 0.2851,
+      "step": 2136
+    },
+    {
+      "epoch": 2.577804583835947,
+      "grad_norm": 0.2081407007815824,
+      "learning_rate": 7.819481680071493e-06,
+      "loss": 0.2814,
+      "step": 2137
+    },
+    {
+      "epoch": 2.5790108564535585,
+      "grad_norm": 0.22624538817631823,
+      "learning_rate": 7.797140303842717e-06,
+      "loss": 0.2918,
+      "step": 2138
+    },
+    {
+      "epoch": 2.58021712907117,
+      "grad_norm": 0.20932221958486957,
+      "learning_rate": 7.774798927613941e-06,
+      "loss": 0.2874,
+      "step": 2139
+    },
+    {
+      "epoch": 2.581423401688782,
+      "grad_norm": 0.2162037323396819,
+      "learning_rate": 7.752457551385165e-06,
+      "loss": 0.292,
+      "step": 2140
+    },
+    {
+      "epoch": 2.5826296743063932,
+      "grad_norm": 0.2249937624133852,
+      "learning_rate": 7.730116175156391e-06,
+      "loss": 0.292,
+      "step": 2141
+    },
+    {
+      "epoch": 2.5838359469240046,
+      "grad_norm": 0.1994108304524604,
+      "learning_rate": 7.707774798927613e-06,
+      "loss": 0.278,
+      "step": 2142
+    },
+    {
+      "epoch": 2.5850422195416165,
+      "grad_norm": 0.19608262203469157,
+      "learning_rate": 7.685433422698839e-06,
+      "loss": 0.2789,
+      "step": 2143
+    },
+    {
+      "epoch": 2.586248492159228,
+      "grad_norm": 0.2086529068204614,
+      "learning_rate": 7.663092046470063e-06,
+      "loss": 0.2905,
+      "step": 2144
+    },
+    {
+      "epoch": 2.5874547647768393,
+      "grad_norm": 0.20938557655487647,
+      "learning_rate": 7.640750670241287e-06,
+      "loss": 0.2802,
+      "step": 2145
+    },
+    {
+      "epoch": 2.588661037394451,
+      "grad_norm": 0.20753877295099624,
+      "learning_rate": 7.618409294012511e-06,
+      "loss": 0.2819,
+      "step": 2146
+    },
+    {
+      "epoch": 2.5898673100120626,
+      "grad_norm": 0.20991545279572527,
+      "learning_rate": 7.596067917783736e-06,
+      "loss": 0.2841,
+      "step": 2147
+    },
+    {
+      "epoch": 2.5910735826296745,
+      "grad_norm": 0.22350098729610313,
+      "learning_rate": 7.5737265415549595e-06,
+      "loss": 0.2981,
+      "step": 2148
+    },
+    {
+      "epoch": 2.592279855247286,
+      "grad_norm": 0.2123357064554487,
+      "learning_rate": 7.5513851653261844e-06,
+      "loss": 0.2821,
+      "step": 2149
+    },
+    {
+      "epoch": 2.5934861278648977,
+      "grad_norm": 0.1991868250972996,
+      "learning_rate": 7.529043789097409e-06,
+      "loss": 0.2809,
+      "step": 2150
+    },
+    {
+      "epoch": 2.594692400482509,
+      "grad_norm": 0.21821718127029568,
+      "learning_rate": 7.506702412868633e-06,
+      "loss": 0.2985,
+      "step": 2151
+    },
+    {
+      "epoch": 2.5958986731001206,
+      "grad_norm": 0.21983928145571835,
+      "learning_rate": 7.4843610366398576e-06,
+      "loss": 0.2893,
+      "step": 2152
+    },
+    {
+      "epoch": 2.5971049457177324,
+      "grad_norm": 0.20937779766909467,
+      "learning_rate": 7.462019660411082e-06,
+      "loss": 0.2876,
+      "step": 2153
+    },
+    {
+      "epoch": 2.598311218335344,
+      "grad_norm": 0.24440569242279062,
+      "learning_rate": 7.439678284182306e-06,
+      "loss": 0.3091,
+      "step": 2154
+    },
+    {
+      "epoch": 2.5995174909529553,
+      "grad_norm": 0.2094410586767038,
+      "learning_rate": 7.41733690795353e-06,
+      "loss": 0.2932,
+      "step": 2155
+    },
+    {
+      "epoch": 2.600723763570567,
+      "grad_norm": 0.22993205214780774,
+      "learning_rate": 7.394995531724755e-06,
+      "loss": 0.2908,
+      "step": 2156
+    },
+    {
+      "epoch": 2.6019300361881785,
+      "grad_norm": 0.232233833116648,
+      "learning_rate": 7.372654155495978e-06,
+      "loss": 0.2849,
+      "step": 2157
+    },
+    {
+      "epoch": 2.60313630880579,
+      "grad_norm": 0.20920454232370092,
+      "learning_rate": 7.350312779267203e-06,
+      "loss": 0.2904,
+      "step": 2158
+    },
+    {
+      "epoch": 2.604342581423402,
+      "grad_norm": 0.22698018377587673,
+      "learning_rate": 7.327971403038428e-06,
+      "loss": 0.2888,
+      "step": 2159
+    },
+    {
+      "epoch": 2.605548854041013,
+      "grad_norm": 0.20161504934121768,
+      "learning_rate": 7.305630026809651e-06,
+      "loss": 0.2715,
+      "step": 2160
+    },
+    {
+      "epoch": 2.6067551266586246,
+      "grad_norm": 0.2060420861005284,
+      "learning_rate": 7.283288650580876e-06,
+      "loss": 0.2979,
+      "step": 2161
+    },
+    {
+      "epoch": 2.6079613992762365,
+      "grad_norm": 0.22127745692072953,
+      "learning_rate": 7.260947274352101e-06,
+      "loss": 0.2824,
+      "step": 2162
+    },
+    {
+      "epoch": 2.609167671893848,
+      "grad_norm": 0.23976066707695107,
+      "learning_rate": 7.238605898123325e-06,
+      "loss": 0.3018,
+      "step": 2163
+    },
+    {
+      "epoch": 2.6103739445114593,
+      "grad_norm": 0.21430343796962054,
+      "learning_rate": 7.216264521894549e-06,
+      "loss": 0.29,
+      "step": 2164
+    },
+    {
+      "epoch": 2.611580217129071,
+      "grad_norm": 0.22023668390455303,
+      "learning_rate": 7.193923145665773e-06,
+      "loss": 0.2942,
+      "step": 2165
+    },
+    {
+      "epoch": 2.6127864897466826,
+      "grad_norm": 0.21547955802919266,
+      "learning_rate": 7.171581769436998e-06,
+      "loss": 0.2808,
+      "step": 2166
+    },
+    {
+      "epoch": 2.6139927623642945,
+      "grad_norm": 0.23149386933699645,
+      "learning_rate": 7.1492403932082214e-06,
+      "loss": 0.3099,
+      "step": 2167
+    },
+    {
+      "epoch": 2.615199034981906,
+      "grad_norm": 0.23126252331722508,
+      "learning_rate": 7.126899016979446e-06,
+      "loss": 0.2752,
+      "step": 2168
+    },
+    {
+      "epoch": 2.6164053075995177,
+      "grad_norm": 0.19958085393309985,
+      "learning_rate": 7.104557640750671e-06,
+      "loss": 0.2746,
+      "step": 2169
+    },
+    {
+      "epoch": 2.617611580217129,
+      "grad_norm": 0.2238611225450866,
+      "learning_rate": 7.0822162645218945e-06,
+      "loss": 0.2851,
+      "step": 2170
+    },
+    {
+      "epoch": 2.6188178528347406,
+      "grad_norm": 0.2352732384970299,
+      "learning_rate": 7.0598748882931195e-06,
+      "loss": 0.2922,
+      "step": 2171
+    },
+    {
+      "epoch": 2.6200241254523524,
+      "grad_norm": 0.22415056452326848,
+      "learning_rate": 7.037533512064344e-06,
+      "loss": 0.2673,
+      "step": 2172
+    },
+    {
+      "epoch": 2.621230398069964,
+      "grad_norm": 0.21897180401823096,
+      "learning_rate": 7.015192135835568e-06,
+      "loss": 0.2812,
+      "step": 2173
+    },
+    {
+      "epoch": 2.6224366706875752,
+      "grad_norm": 0.23020781529953618,
+      "learning_rate": 6.992850759606792e-06,
+      "loss": 0.2853,
+      "step": 2174
+    },
+    {
+      "epoch": 2.623642943305187,
+      "grad_norm": 0.21875426785559132,
+      "learning_rate": 6.970509383378017e-06,
+      "loss": 0.2909,
+      "step": 2175
+    },
+    {
+      "epoch": 2.6248492159227985,
+      "grad_norm": 0.2816490674500168,
+      "learning_rate": 6.94816800714924e-06,
+      "loss": 0.2988,
+      "step": 2176
+    },
+    {
+      "epoch": 2.62605548854041,
+      "grad_norm": 0.21486929348764425,
+      "learning_rate": 6.925826630920465e-06,
+      "loss": 0.2792,
+      "step": 2177
+    },
+    {
+      "epoch": 2.627261761158022,
+      "grad_norm": 0.2301825259699349,
+      "learning_rate": 6.90348525469169e-06,
+      "loss": 0.2792,
+      "step": 2178
+    },
+    {
+      "epoch": 2.628468033775633,
+      "grad_norm": 0.2253863912294859,
+      "learning_rate": 6.881143878462913e-06,
+      "loss": 0.2953,
+      "step": 2179
+    },
+    {
+      "epoch": 2.6296743063932446,
+      "grad_norm": 0.2169639675897766,
+      "learning_rate": 6.858802502234138e-06,
+      "loss": 0.2891,
+      "step": 2180
+    },
+    {
+      "epoch": 2.6308805790108565,
+      "grad_norm": 0.21726317823189514,
+      "learning_rate": 6.836461126005363e-06,
+      "loss": 0.291,
+      "step": 2181
+    },
+    {
+      "epoch": 2.632086851628468,
+      "grad_norm": 0.19617546763397464,
+      "learning_rate": 6.814119749776586e-06,
+      "loss": 0.2647,
+      "step": 2182
+    },
+    {
+      "epoch": 2.6332931242460798,
+      "grad_norm": 0.21554251282986897,
+      "learning_rate": 6.791778373547811e-06,
+      "loss": 0.2843,
+      "step": 2183
+    },
+    {
+      "epoch": 2.634499396863691,
+      "grad_norm": 0.2459135828666879,
+      "learning_rate": 6.769436997319035e-06,
+      "loss": 0.3009,
+      "step": 2184
+    },
+    {
+      "epoch": 2.635705669481303,
+      "grad_norm": 0.21621503077120693,
+      "learning_rate": 6.747095621090259e-06,
+      "loss": 0.2952,
+      "step": 2185
+    },
+    {
+      "epoch": 2.6369119420989144,
+      "grad_norm": 0.20740297901076743,
+      "learning_rate": 6.724754244861483e-06,
+      "loss": 0.295,
+      "step": 2186
+    },
+    {
+      "epoch": 2.638118214716526,
+      "grad_norm": 0.21029148372921153,
+      "learning_rate": 6.702412868632708e-06,
+      "loss": 0.2795,
+      "step": 2187
+    },
+    {
+      "epoch": 2.6393244873341377,
+      "grad_norm": 0.22609754669207283,
+      "learning_rate": 6.6800714924039315e-06,
+      "loss": 0.2949,
+      "step": 2188
+    },
+    {
+      "epoch": 2.640530759951749,
+      "grad_norm": 0.22113015986707052,
+      "learning_rate": 6.6577301161751565e-06,
+      "loss": 0.283,
+      "step": 2189
+    },
+    {
+      "epoch": 2.6417370325693605,
+      "grad_norm": 0.21290296828182953,
+      "learning_rate": 6.635388739946381e-06,
+      "loss": 0.3058,
+      "step": 2190
+    },
+    {
+      "epoch": 2.6429433051869724,
+      "grad_norm": 0.20663250547689224,
+      "learning_rate": 6.613047363717605e-06,
+      "loss": 0.2704,
+      "step": 2191
+    },
+    {
+      "epoch": 2.644149577804584,
+      "grad_norm": 0.2194261553038497,
+      "learning_rate": 6.5907059874888296e-06,
+      "loss": 0.2822,
+      "step": 2192
+    },
+    {
+      "epoch": 2.6453558504221952,
+      "grad_norm": 0.2378506340578728,
+      "learning_rate": 6.5683646112600545e-06,
+      "loss": 0.2898,
+      "step": 2193
+    },
+    {
+      "epoch": 2.646562123039807,
+      "grad_norm": 0.2242127329287937,
+      "learning_rate": 6.546023235031279e-06,
+      "loss": 0.2779,
+      "step": 2194
+    },
+    {
+      "epoch": 2.6477683956574185,
+      "grad_norm": 0.21847059262012553,
+      "learning_rate": 6.523681858802503e-06,
+      "loss": 0.2939,
+      "step": 2195
+    },
+    {
+      "epoch": 2.64897466827503,
+      "grad_norm": 0.231399791978887,
+      "learning_rate": 6.501340482573727e-06,
+      "loss": 0.279,
+      "step": 2196
+    },
+    {
+      "epoch": 2.650180940892642,
+      "grad_norm": 0.2116623580528426,
+      "learning_rate": 6.478999106344952e-06,
+      "loss": 0.2859,
+      "step": 2197
+    },
+    {
+      "epoch": 2.651387213510253,
+      "grad_norm": 0.1980992165403409,
+      "learning_rate": 6.456657730116175e-06,
+      "loss": 0.2804,
+      "step": 2198
+    },
+    {
+      "epoch": 2.652593486127865,
+      "grad_norm": 0.24171516250636746,
+      "learning_rate": 6.4343163538874e-06,
+      "loss": 0.2951,
+      "step": 2199
+    },
+    {
+      "epoch": 2.6537997587454765,
+      "grad_norm": 0.21184774792981634,
+      "learning_rate": 6.411974977658625e-06,
+      "loss": 0.2931,
+      "step": 2200
+    },
+    {
+      "epoch": 2.6550060313630883,
+      "grad_norm": 0.21738233707648502,
+      "learning_rate": 6.389633601429848e-06,
+      "loss": 0.2785,
+      "step": 2201
+    },
+    {
+      "epoch": 2.6562123039806997,
+      "grad_norm": 0.2127604344024817,
+      "learning_rate": 6.367292225201073e-06,
+      "loss": 0.2868,
+      "step": 2202
+    },
+    {
+      "epoch": 2.657418576598311,
+      "grad_norm": 0.1931954397670524,
+      "learning_rate": 6.344950848972298e-06,
+      "loss": 0.2665,
+      "step": 2203
+    },
+    {
+      "epoch": 2.658624849215923,
+      "grad_norm": 0.20838444267534595,
+      "learning_rate": 6.322609472743521e-06,
+      "loss": 0.2807,
+      "step": 2204
+    },
+    {
+      "epoch": 2.6598311218335344,
+      "grad_norm": 0.2211018154320071,
+      "learning_rate": 6.300268096514745e-06,
+      "loss": 0.2813,
+      "step": 2205
+    },
+    {
+      "epoch": 2.661037394451146,
+      "grad_norm": 0.2168479274201597,
+      "learning_rate": 6.27792672028597e-06,
+      "loss": 0.2765,
+      "step": 2206
+    },
+    {
+      "epoch": 2.6622436670687577,
+      "grad_norm": 0.2213334733094897,
+      "learning_rate": 6.2555853440571934e-06,
+      "loss": 0.2798,
+      "step": 2207
+    },
+    {
+      "epoch": 2.663449939686369,
+      "grad_norm": 0.19954163195183472,
+      "learning_rate": 6.233243967828418e-06,
+      "loss": 0.2656,
+      "step": 2208
+    },
+    {
+      "epoch": 2.6646562123039805,
+      "grad_norm": 0.22078742024428197,
+      "learning_rate": 6.2109025915996425e-06,
+      "loss": 0.286,
+      "step": 2209
+    },
+    {
+      "epoch": 2.6658624849215924,
+      "grad_norm": 0.22161496613081794,
+      "learning_rate": 6.188561215370867e-06,
+      "loss": 0.2857,
+      "step": 2210
+    },
+    {
+      "epoch": 2.667068757539204,
+      "grad_norm": 0.21212036342000515,
+      "learning_rate": 6.1662198391420915e-06,
+      "loss": 0.286,
+      "step": 2211
+    },
+    {
+      "epoch": 2.668275030156815,
+      "grad_norm": 0.20187880736220248,
+      "learning_rate": 6.1438784629133156e-06,
+      "loss": 0.2815,
+      "step": 2212
+    },
+    {
+      "epoch": 2.669481302774427,
+      "grad_norm": 0.22474273562685815,
+      "learning_rate": 6.1215370866845405e-06,
+      "loss": 0.2982,
+      "step": 2213
+    },
+    {
+      "epoch": 2.6706875753920385,
+      "grad_norm": 0.20611518128991452,
+      "learning_rate": 6.099195710455765e-06,
+      "loss": 0.2976,
+      "step": 2214
+    },
+    {
+      "epoch": 2.67189384800965,
+      "grad_norm": 0.20695572664023934,
+      "learning_rate": 6.076854334226989e-06,
+      "loss": 0.2896,
+      "step": 2215
+    },
+    {
+      "epoch": 2.6731001206272618,
+      "grad_norm": 0.21701647489801484,
+      "learning_rate": 6.054512957998213e-06,
+      "loss": 0.306,
+      "step": 2216
+    },
+    {
+      "epoch": 2.674306393244873,
+      "grad_norm": 0.2390978357229728,
+      "learning_rate": 6.032171581769437e-06,
+      "loss": 0.2889,
+      "step": 2217
+    },
+    {
+      "epoch": 2.675512665862485,
+      "grad_norm": 0.2248591636170576,
+      "learning_rate": 6.009830205540662e-06,
+      "loss": 0.3062,
+      "step": 2218
+    },
+    {
+      "epoch": 2.6767189384800965,
+      "grad_norm": 0.20736477751728932,
+      "learning_rate": 5.987488829311886e-06,
+      "loss": 0.2959,
+      "step": 2219
+    },
+    {
+      "epoch": 2.6779252110977083,
+      "grad_norm": 0.20115538299047425,
+      "learning_rate": 5.96514745308311e-06,
+      "loss": 0.2721,
+      "step": 2220
+    },
+    {
+      "epoch": 2.6791314837153197,
+      "grad_norm": 0.20571833148964577,
+      "learning_rate": 5.942806076854335e-06,
+      "loss": 0.2871,
+      "step": 2221
+    },
+    {
+      "epoch": 2.680337756332931,
+      "grad_norm": 0.22354102926280847,
+      "learning_rate": 5.920464700625559e-06,
+      "loss": 0.2896,
+      "step": 2222
+    },
+    {
+      "epoch": 2.681544028950543,
+      "grad_norm": 0.20983791062606558,
+      "learning_rate": 5.898123324396783e-06,
+      "loss": 0.2887,
+      "step": 2223
+    },
+    {
+      "epoch": 2.6827503015681544,
+      "grad_norm": 0.2021459077857852,
+      "learning_rate": 5.875781948168008e-06,
+      "loss": 0.2832,
+      "step": 2224
+    },
+    {
+      "epoch": 2.683956574185766,
+      "grad_norm": 0.21740950707741213,
+      "learning_rate": 5.853440571939232e-06,
+      "loss": 0.2653,
+      "step": 2225
+    },
+    {
+      "epoch": 2.6851628468033777,
+      "grad_norm": 0.22039114740127552,
+      "learning_rate": 5.831099195710455e-06,
+      "loss": 0.2971,
+      "step": 2226
+    },
+    {
+      "epoch": 2.686369119420989,
+      "grad_norm": 0.22684638806753404,
+      "learning_rate": 5.80875781948168e-06,
+      "loss": 0.2937,
+      "step": 2227
+    },
+    {
+      "epoch": 2.6875753920386005,
+      "grad_norm": 0.19918080045760725,
+      "learning_rate": 5.786416443252904e-06,
+      "loss": 0.2867,
+      "step": 2228
+    },
+    {
+      "epoch": 2.6887816646562124,
+      "grad_norm": 0.227332586710256,
+      "learning_rate": 5.7640750670241285e-06,
+      "loss": 0.2726,
+      "step": 2229
+    },
+    {
+      "epoch": 2.689987937273824,
+      "grad_norm": 0.2210279769730908,
+      "learning_rate": 5.741733690795353e-06,
+      "loss": 0.2923,
+      "step": 2230
+    },
+    {
+      "epoch": 2.691194209891435,
+      "grad_norm": 0.21827221208506087,
+      "learning_rate": 5.7193923145665775e-06,
+      "loss": 0.2965,
+      "step": 2231
+    },
+    {
+      "epoch": 2.692400482509047,
+      "grad_norm": 0.2045014578759283,
+      "learning_rate": 5.697050938337802e-06,
+      "loss": 0.2914,
+      "step": 2232
+    },
+    {
+      "epoch": 2.6936067551266585,
+      "grad_norm": 0.6946714666447911,
+      "learning_rate": 5.6747095621090265e-06,
+      "loss": 0.3133,
+      "step": 2233
+    },
+    {
+      "epoch": 2.6948130277442703,
+      "grad_norm": 0.21370291737243427,
+      "learning_rate": 5.652368185880251e-06,
+      "loss": 0.2715,
+      "step": 2234
+    },
+    {
+      "epoch": 2.6960193003618818,
+      "grad_norm": 0.2044609100427451,
+      "learning_rate": 5.630026809651475e-06,
+      "loss": 0.2926,
+      "step": 2235
+    },
+    {
+      "epoch": 2.6972255729794936,
+      "grad_norm": 0.22829735674613444,
+      "learning_rate": 5.607685433422699e-06,
+      "loss": 0.2833,
+      "step": 2236
+    },
+    {
+      "epoch": 2.698431845597105,
+      "grad_norm": 0.26018311521323667,
+      "learning_rate": 5.585344057193923e-06,
+      "loss": 0.2839,
+      "step": 2237
+    },
+    {
+      "epoch": 2.6996381182147164,
+      "grad_norm": 0.2086058178570994,
+      "learning_rate": 5.563002680965148e-06,
+      "loss": 0.2895,
+      "step": 2238
+    },
+    {
+      "epoch": 2.7008443908323283,
+      "grad_norm": 0.2050257982382011,
+      "learning_rate": 5.540661304736372e-06,
+      "loss": 0.2828,
+      "step": 2239
+    },
+    {
+      "epoch": 2.7020506634499397,
+      "grad_norm": 0.2139514048129372,
+      "learning_rate": 5.518319928507596e-06,
+      "loss": 0.2867,
+      "step": 2240
+    },
+    {
+      "epoch": 2.703256936067551,
+      "grad_norm": 0.2165520506724918,
+      "learning_rate": 5.495978552278821e-06,
+      "loss": 0.2885,
+      "step": 2241
+    },
+    {
+      "epoch": 2.704463208685163,
+      "grad_norm": 0.21443900046194145,
+      "learning_rate": 5.473637176050045e-06,
+      "loss": 0.2856,
+      "step": 2242
+    },
+    {
+      "epoch": 2.7056694813027744,
+      "grad_norm": 0.21335757069783218,
+      "learning_rate": 5.451295799821269e-06,
+      "loss": 0.2975,
+      "step": 2243
+    },
+    {
+      "epoch": 2.706875753920386,
+      "grad_norm": 0.20145288057580185,
+      "learning_rate": 5.428954423592494e-06,
+      "loss": 0.2952,
+      "step": 2244
+    },
+    {
+      "epoch": 2.7080820265379977,
+      "grad_norm": 0.22761820406606897,
+      "learning_rate": 5.406613047363718e-06,
+      "loss": 0.2879,
+      "step": 2245
+    },
+    {
+      "epoch": 2.709288299155609,
+      "grad_norm": 0.2364015496890641,
+      "learning_rate": 5.384271671134942e-06,
+      "loss": 0.281,
+      "step": 2246
+    },
+    {
+      "epoch": 2.7104945717732205,
+      "grad_norm": 0.21753945940012162,
+      "learning_rate": 5.361930294906166e-06,
+      "loss": 0.2964,
+      "step": 2247
+    },
+    {
+      "epoch": 2.7117008443908324,
+      "grad_norm": 0.21327033716673088,
+      "learning_rate": 5.33958891867739e-06,
+      "loss": 0.296,
+      "step": 2248
+    },
+    {
+      "epoch": 2.712907117008444,
+      "grad_norm": 0.1951078699418731,
+      "learning_rate": 5.317247542448615e-06,
+      "loss": 0.2791,
+      "step": 2249
+    },
+    {
+      "epoch": 2.7141133896260556,
+      "grad_norm": 0.20240742137700055,
+      "learning_rate": 5.294906166219839e-06,
+      "loss": 0.2924,
+      "step": 2250
+    },
+    {
+      "epoch": 2.715319662243667,
+      "grad_norm": 0.22220578965053128,
+      "learning_rate": 5.2725647899910635e-06,
+      "loss": 0.3065,
+      "step": 2251
+    },
+    {
+      "epoch": 2.716525934861279,
+      "grad_norm": 0.20214399599477811,
+      "learning_rate": 5.250223413762288e-06,
+      "loss": 0.2744,
+      "step": 2252
+    },
+    {
+      "epoch": 2.7177322074788903,
+      "grad_norm": 0.21155302019530836,
+      "learning_rate": 5.2278820375335125e-06,
+      "loss": 0.2715,
+      "step": 2253
+    },
+    {
+      "epoch": 2.7189384800965017,
+      "grad_norm": 0.19608141646891808,
+      "learning_rate": 5.205540661304737e-06,
+      "loss": 0.2781,
+      "step": 2254
+    },
+    {
+      "epoch": 2.7201447527141136,
+      "grad_norm": 0.20753070114799987,
+      "learning_rate": 5.1831992850759615e-06,
+      "loss": 0.2807,
+      "step": 2255
+    },
+    {
+      "epoch": 2.721351025331725,
+      "grad_norm": 0.19788509046006095,
+      "learning_rate": 5.160857908847186e-06,
+      "loss": 0.2762,
+      "step": 2256
+    },
+    {
+      "epoch": 2.7225572979493364,
+      "grad_norm": 0.2066145277274399,
+      "learning_rate": 5.138516532618409e-06,
+      "loss": 0.3045,
+      "step": 2257
+    },
+    {
+      "epoch": 2.7237635705669483,
+      "grad_norm": 0.20139286285522892,
+      "learning_rate": 5.116175156389634e-06,
+      "loss": 0.2844,
+      "step": 2258
+    },
+    {
+      "epoch": 2.7249698431845597,
+      "grad_norm": 0.20798235148380942,
+      "learning_rate": 5.093833780160858e-06,
+      "loss": 0.2853,
+      "step": 2259
+    },
+    {
+      "epoch": 2.726176115802171,
+      "grad_norm": 0.2091974900829975,
+      "learning_rate": 5.071492403932082e-06,
+      "loss": 0.2756,
+      "step": 2260
+    },
+    {
+      "epoch": 2.727382388419783,
+      "grad_norm": 0.21745865410514648,
+      "learning_rate": 5.049151027703307e-06,
+      "loss": 0.2894,
+      "step": 2261
+    },
+    {
+      "epoch": 2.7285886610373944,
+      "grad_norm": 0.19485099560282385,
+      "learning_rate": 5.026809651474531e-06,
+      "loss": 0.2761,
+      "step": 2262
+    },
+    {
+      "epoch": 2.729794933655006,
+      "grad_norm": 0.19149290082897566,
+      "learning_rate": 5.004468275245756e-06,
+      "loss": 0.2888,
+      "step": 2263
+    },
+    {
+      "epoch": 2.7310012062726177,
+      "grad_norm": 0.20599673735385637,
+      "learning_rate": 4.98212689901698e-06,
+      "loss": 0.2884,
+      "step": 2264
+    },
+    {
+      "epoch": 2.732207478890229,
+      "grad_norm": 0.20713273074776942,
+      "learning_rate": 4.959785522788204e-06,
+      "loss": 0.2874,
+      "step": 2265
+    },
+    {
+      "epoch": 2.7334137515078405,
+      "grad_norm": 0.2081817873109111,
+      "learning_rate": 4.937444146559428e-06,
+      "loss": 0.2944,
+      "step": 2266
+    },
+    {
+      "epoch": 2.7346200241254524,
+      "grad_norm": 0.20553774261680427,
+      "learning_rate": 4.915102770330652e-06,
+      "loss": 0.2716,
+      "step": 2267
+    },
+    {
+      "epoch": 2.7358262967430638,
+      "grad_norm": 0.20501194493617642,
+      "learning_rate": 4.892761394101876e-06,
+      "loss": 0.3012,
+      "step": 2268
+    },
+    {
+      "epoch": 2.7370325693606756,
+      "grad_norm": 0.1990487984580353,
+      "learning_rate": 4.870420017873101e-06,
+      "loss": 0.2941,
+      "step": 2269
+    },
+    {
+      "epoch": 2.738238841978287,
+      "grad_norm": 0.1983334601951314,
+      "learning_rate": 4.848078641644325e-06,
+      "loss": 0.2753,
+      "step": 2270
+    },
+    {
+      "epoch": 2.739445114595899,
+      "grad_norm": 0.19403916283019812,
+      "learning_rate": 4.8257372654155495e-06,
+      "loss": 0.2862,
+      "step": 2271
+    },
+    {
+      "epoch": 2.7406513872135103,
+      "grad_norm": 0.21538723639291782,
+      "learning_rate": 4.803395889186774e-06,
+      "loss": 0.2943,
+      "step": 2272
+    },
+    {
+      "epoch": 2.7418576598311217,
+      "grad_norm": 0.2375574289392646,
+      "learning_rate": 4.7810545129579985e-06,
+      "loss": 0.2887,
+      "step": 2273
+    },
+    {
+      "epoch": 2.7430639324487336,
+      "grad_norm": 0.19945442614951428,
+      "learning_rate": 4.758713136729223e-06,
+      "loss": 0.3011,
+      "step": 2274
+    },
+    {
+      "epoch": 2.744270205066345,
+      "grad_norm": 0.20199363951824748,
+      "learning_rate": 4.7363717605004475e-06,
+      "loss": 0.2942,
+      "step": 2275
+    },
+    {
+      "epoch": 2.7454764776839564,
+      "grad_norm": 0.2059245247625709,
+      "learning_rate": 4.714030384271672e-06,
+      "loss": 0.2876,
+      "step": 2276
+    },
+    {
+      "epoch": 2.7466827503015683,
+      "grad_norm": 0.20187465821580058,
+      "learning_rate": 4.691689008042896e-06,
+      "loss": 0.2844,
+      "step": 2277
+    },
+    {
+      "epoch": 2.7478890229191797,
+      "grad_norm": 0.20636177253183094,
+      "learning_rate": 4.66934763181412e-06,
+      "loss": 0.2796,
+      "step": 2278
+    },
+    {
+      "epoch": 2.749095295536791,
+      "grad_norm": 0.20723033389487827,
+      "learning_rate": 4.647006255585344e-06,
+      "loss": 0.2767,
+      "step": 2279
+    },
+    {
+      "epoch": 2.750301568154403,
+      "grad_norm": 0.21174347393266027,
+      "learning_rate": 4.624664879356569e-06,
+      "loss": 0.3188,
+      "step": 2280
+    },
+    {
+      "epoch": 2.7515078407720144,
+      "grad_norm": 0.1978681060454767,
+      "learning_rate": 4.602323503127793e-06,
+      "loss": 0.2784,
+      "step": 2281
+    },
+    {
+      "epoch": 2.752714113389626,
+      "grad_norm": 0.18744202466276072,
+      "learning_rate": 4.579982126899017e-06,
+      "loss": 0.2682,
+      "step": 2282
+    },
+    {
+      "epoch": 2.7539203860072377,
+      "grad_norm": 0.21209688003204358,
+      "learning_rate": 4.557640750670242e-06,
+      "loss": 0.285,
+      "step": 2283
+    },
+    {
+      "epoch": 2.755126658624849,
+      "grad_norm": 0.19892738867053703,
+      "learning_rate": 4.535299374441466e-06,
+      "loss": 0.2714,
+      "step": 2284
+    },
+    {
+      "epoch": 2.756332931242461,
+      "grad_norm": 0.2041608394139942,
+      "learning_rate": 4.51295799821269e-06,
+      "loss": 0.2913,
+      "step": 2285
+    },
+    {
+      "epoch": 2.7575392038600723,
+      "grad_norm": 0.19915952593238026,
+      "learning_rate": 4.490616621983915e-06,
+      "loss": 0.274,
+      "step": 2286
+    },
+    {
+      "epoch": 2.758745476477684,
+      "grad_norm": 0.2044806484365547,
+      "learning_rate": 4.468275245755138e-06,
+      "loss": 0.2853,
+      "step": 2287
+    },
+    {
+      "epoch": 2.7599517490952956,
+      "grad_norm": 0.19573371237314433,
+      "learning_rate": 4.445933869526362e-06,
+      "loss": 0.2738,
+      "step": 2288
+    },
+    {
+      "epoch": 2.761158021712907,
+      "grad_norm": 0.2106092030662992,
+      "learning_rate": 4.423592493297587e-06,
+      "loss": 0.2972,
+      "step": 2289
+    },
+    {
+      "epoch": 2.762364294330519,
+      "grad_norm": 0.19573259291033523,
+      "learning_rate": 4.401251117068811e-06,
+      "loss": 0.2829,
+      "step": 2290
+    },
+    {
+      "epoch": 2.7635705669481303,
+      "grad_norm": 0.20564658398999378,
+      "learning_rate": 4.3789097408400355e-06,
+      "loss": 0.2976,
+      "step": 2291
+    },
+    {
+      "epoch": 2.7647768395657417,
+      "grad_norm": 0.1879596479464295,
+      "learning_rate": 4.35656836461126e-06,
+      "loss": 0.2705,
+      "step": 2292
+    },
+    {
+      "epoch": 2.7659831121833536,
+      "grad_norm": 0.20028084619699768,
+      "learning_rate": 4.3342269883824845e-06,
+      "loss": 0.2898,
+      "step": 2293
+    },
+    {
+      "epoch": 2.767189384800965,
+      "grad_norm": 0.204373079980006,
+      "learning_rate": 4.3118856121537094e-06,
+      "loss": 0.2854,
+      "step": 2294
+    },
+    {
+      "epoch": 2.7683956574185764,
+      "grad_norm": 0.19758717833265702,
+      "learning_rate": 4.2895442359249335e-06,
+      "loss": 0.2802,
+      "step": 2295
+    },
+    {
+      "epoch": 2.7696019300361883,
+      "grad_norm": 0.21023108969069418,
+      "learning_rate": 4.267202859696158e-06,
+      "loss": 0.284,
+      "step": 2296
+    },
+    {
+      "epoch": 2.7708082026537997,
+      "grad_norm": 0.19027759880066697,
+      "learning_rate": 4.244861483467382e-06,
+      "loss": 0.2821,
+      "step": 2297
+    },
+    {
+      "epoch": 2.772014475271411,
+      "grad_norm": 0.18871657043933537,
+      "learning_rate": 4.222520107238606e-06,
+      "loss": 0.2729,
+      "step": 2298
+    },
+    {
+      "epoch": 2.773220747889023,
+      "grad_norm": 0.20709511269571762,
+      "learning_rate": 4.20017873100983e-06,
+      "loss": 0.2768,
+      "step": 2299
+    },
+    {
+      "epoch": 2.7744270205066344,
+      "grad_norm": 0.20297218284984314,
+      "learning_rate": 4.177837354781055e-06,
+      "loss": 0.2869,
+      "step": 2300
+    },
+    {
+      "epoch": 2.7756332931242462,
+      "grad_norm": 0.19887378780810883,
+      "learning_rate": 4.155495978552279e-06,
+      "loss": 0.3008,
+      "step": 2301
+    },
+    {
+      "epoch": 2.7768395657418576,
+      "grad_norm": 0.2118503034579533,
+      "learning_rate": 4.133154602323503e-06,
+      "loss": 0.2828,
+      "step": 2302
+    },
+    {
+      "epoch": 2.7780458383594695,
+      "grad_norm": 0.2219706988384966,
+      "learning_rate": 4.110813226094728e-06,
+      "loss": 0.3062,
+      "step": 2303
+    },
+    {
+      "epoch": 2.779252110977081,
+      "grad_norm": 0.25701330993635146,
+      "learning_rate": 4.088471849865952e-06,
+      "loss": 0.2745,
+      "step": 2304
+    },
+    {
+      "epoch": 2.7804583835946923,
+      "grad_norm": 0.2234143696898838,
+      "learning_rate": 4.066130473637176e-06,
+      "loss": 0.291,
+      "step": 2305
+    },
+    {
+      "epoch": 2.781664656212304,
+      "grad_norm": 0.1941194837681158,
+      "learning_rate": 4.043789097408401e-06,
+      "loss": 0.2877,
+      "step": 2306
+    },
+    {
+      "epoch": 2.7828709288299156,
+      "grad_norm": 0.2064300988757331,
+      "learning_rate": 4.021447721179625e-06,
+      "loss": 0.2884,
+      "step": 2307
+    },
+    {
+      "epoch": 2.784077201447527,
+      "grad_norm": 0.21140547687570724,
+      "learning_rate": 3.999106344950849e-06,
+      "loss": 0.298,
+      "step": 2308
+    },
+    {
+      "epoch": 2.785283474065139,
+      "grad_norm": 0.19951947203291173,
+      "learning_rate": 3.976764968722073e-06,
+      "loss": 0.2699,
+      "step": 2309
+    },
+    {
+      "epoch": 2.7864897466827503,
+      "grad_norm": 0.20431837354268684,
+      "learning_rate": 3.954423592493297e-06,
+      "loss": 0.2865,
+      "step": 2310
+    },
+    {
+      "epoch": 2.7876960193003617,
+      "grad_norm": 0.19834889240552475,
+      "learning_rate": 3.932082216264522e-06,
+      "loss": 0.2905,
+      "step": 2311
+    },
+    {
+      "epoch": 2.7889022919179736,
+      "grad_norm": 0.21695363502888282,
+      "learning_rate": 3.909740840035746e-06,
+      "loss": 0.2763,
+      "step": 2312
+    },
+    {
+      "epoch": 2.790108564535585,
+      "grad_norm": 0.19273556462979177,
+      "learning_rate": 3.8873994638069705e-06,
+      "loss": 0.2774,
+      "step": 2313
+    },
+    {
+      "epoch": 2.7913148371531964,
+      "grad_norm": 0.19451733876401245,
+      "learning_rate": 3.8650580875781954e-06,
+      "loss": 0.29,
+      "step": 2314
+    },
+    {
+      "epoch": 2.7925211097708083,
+      "grad_norm": 0.1964077558155501,
+      "learning_rate": 3.8427167113494195e-06,
+      "loss": 0.2802,
+      "step": 2315
+    },
+    {
+      "epoch": 2.7937273823884197,
+      "grad_norm": 0.21179533332456585,
+      "learning_rate": 3.820375335120644e-06,
+      "loss": 0.3013,
+      "step": 2316
+    },
+    {
+      "epoch": 2.7949336550060315,
+      "grad_norm": 0.19720588063541614,
+      "learning_rate": 3.798033958891868e-06,
+      "loss": 0.2679,
+      "step": 2317
+    },
+    {
+      "epoch": 2.796139927623643,
+      "grad_norm": 0.20036859315386968,
+      "learning_rate": 3.7756925826630922e-06,
+      "loss": 0.2951,
+      "step": 2318
+    },
+    {
+      "epoch": 2.797346200241255,
+      "grad_norm": 0.205713796416112,
+      "learning_rate": 3.7533512064343163e-06,
+      "loss": 0.2753,
+      "step": 2319
+    },
+    {
+      "epoch": 2.798552472858866,
+      "grad_norm": 0.2086376396974652,
+      "learning_rate": 3.731009830205541e-06,
+      "loss": 0.2806,
+      "step": 2320
+    },
+    {
+      "epoch": 2.7997587454764776,
+      "grad_norm": 0.19776511558296467,
+      "learning_rate": 3.708668453976765e-06,
+      "loss": 0.2825,
+      "step": 2321
+    },
+    {
+      "epoch": 2.8009650180940895,
+      "grad_norm": 0.1979973932225097,
+      "learning_rate": 3.686327077747989e-06,
+      "loss": 0.2925,
+      "step": 2322
+    },
+    {
+      "epoch": 2.802171290711701,
+      "grad_norm": 0.20109286837062904,
+      "learning_rate": 3.663985701519214e-06,
+      "loss": 0.2994,
+      "step": 2323
+    },
+    {
+      "epoch": 2.8033775633293123,
+      "grad_norm": 0.19996893427115703,
+      "learning_rate": 3.641644325290438e-06,
+      "loss": 0.2603,
+      "step": 2324
+    },
+    {
+      "epoch": 2.804583835946924,
+      "grad_norm": 0.2156601549234228,
+      "learning_rate": 3.6193029490616625e-06,
+      "loss": 0.2892,
+      "step": 2325
+    },
+    {
+      "epoch": 2.8057901085645356,
+      "grad_norm": 0.19954283493636063,
+      "learning_rate": 3.5969615728328866e-06,
+      "loss": 0.2848,
+      "step": 2326
+    },
+    {
+      "epoch": 2.806996381182147,
+      "grad_norm": 0.20103437872062893,
+      "learning_rate": 3.5746201966041107e-06,
+      "loss": 0.2802,
+      "step": 2327
+    },
+    {
+      "epoch": 2.808202653799759,
+      "grad_norm": 0.19973565351335576,
+      "learning_rate": 3.5522788203753356e-06,
+      "loss": 0.284,
+      "step": 2328
+    },
+    {
+      "epoch": 2.8094089264173703,
+      "grad_norm": 0.20898595146924076,
+      "learning_rate": 3.5299374441465597e-06,
+      "loss": 0.2866,
+      "step": 2329
+    },
+    {
+      "epoch": 2.8106151990349817,
+      "grad_norm": 0.19573103856679328,
+      "learning_rate": 3.507596067917784e-06,
+      "loss": 0.2876,
+      "step": 2330
+    },
+    {
+      "epoch": 2.8118214716525936,
+      "grad_norm": 0.2077731765783218,
+      "learning_rate": 3.4852546916890083e-06,
+      "loss": 0.2956,
+      "step": 2331
+    },
+    {
+      "epoch": 2.813027744270205,
+      "grad_norm": 0.2016470923946435,
+      "learning_rate": 3.4629133154602324e-06,
+      "loss": 0.2948,
+      "step": 2332
+    },
+    {
+      "epoch": 2.8142340168878164,
+      "grad_norm": 0.2044886910285202,
+      "learning_rate": 3.4405719392314565e-06,
+      "loss": 0.3065,
+      "step": 2333
+    },
+    {
+      "epoch": 2.8154402895054282,
+      "grad_norm": 0.20278178377461592,
+      "learning_rate": 3.4182305630026814e-06,
+      "loss": 0.2805,
+      "step": 2334
+    },
+    {
+      "epoch": 2.8166465621230397,
+      "grad_norm": 0.20087995640692985,
+      "learning_rate": 3.3958891867739055e-06,
+      "loss": 0.2733,
+      "step": 2335
+    },
+    {
+      "epoch": 2.8178528347406515,
+      "grad_norm": 0.20454242639756476,
+      "learning_rate": 3.3735478105451296e-06,
+      "loss": 0.2947,
+      "step": 2336
+    },
+    {
+      "epoch": 2.819059107358263,
+      "grad_norm": 0.19624397685889447,
+      "learning_rate": 3.351206434316354e-06,
+      "loss": 0.2792,
+      "step": 2337
+    },
+    {
+      "epoch": 2.820265379975875,
+      "grad_norm": 0.1918852390942654,
+      "learning_rate": 3.3288650580875782e-06,
+      "loss": 0.2746,
+      "step": 2338
+    },
+    {
+      "epoch": 2.821471652593486,
+      "grad_norm": 0.19264796239668547,
+      "learning_rate": 3.3065236818588023e-06,
+      "loss": 0.2988,
+      "step": 2339
+    },
+    {
+      "epoch": 2.8226779252110976,
+      "grad_norm": 0.19970750561027092,
+      "learning_rate": 3.2841823056300272e-06,
+      "loss": 0.2843,
+      "step": 2340
+    },
+    {
+      "epoch": 2.8238841978287095,
+      "grad_norm": 0.19496241921024737,
+      "learning_rate": 3.2618409294012513e-06,
+      "loss": 0.2815,
+      "step": 2341
+    },
+    {
+      "epoch": 2.825090470446321,
+      "grad_norm": 0.21303286652759756,
+      "learning_rate": 3.239499553172476e-06,
+      "loss": 0.2979,
+      "step": 2342
+    },
+    {
+      "epoch": 2.8262967430639323,
+      "grad_norm": 0.1966256591701248,
+      "learning_rate": 3.2171581769437e-06,
+      "loss": 0.2733,
+      "step": 2343
+    },
+    {
+      "epoch": 2.827503015681544,
+      "grad_norm": 0.19157269512914996,
+      "learning_rate": 3.194816800714924e-06,
+      "loss": 0.284,
+      "step": 2344
+    },
+    {
+      "epoch": 2.8287092882991556,
+      "grad_norm": 0.20806116030817426,
+      "learning_rate": 3.172475424486149e-06,
+      "loss": 0.2771,
+      "step": 2345
+    },
+    {
+      "epoch": 2.829915560916767,
+      "grad_norm": 0.20393659922514393,
+      "learning_rate": 3.1501340482573726e-06,
+      "loss": 0.2842,
+      "step": 2346
+    },
+    {
+      "epoch": 2.831121833534379,
+      "grad_norm": 0.20545574063430777,
+      "learning_rate": 3.1277926720285967e-06,
+      "loss": 0.2999,
+      "step": 2347
+    },
+    {
+      "epoch": 2.8323281061519903,
+      "grad_norm": 0.20800833266552515,
+      "learning_rate": 3.1054512957998212e-06,
+      "loss": 0.3056,
+      "step": 2348
+    },
+    {
+      "epoch": 2.8335343787696017,
+      "grad_norm": 0.1920188003389075,
+      "learning_rate": 3.0831099195710457e-06,
+      "loss": 0.2801,
+      "step": 2349
+    },
+    {
+      "epoch": 2.8347406513872135,
+      "grad_norm": 0.1956810201413982,
+      "learning_rate": 3.0607685433422703e-06,
+      "loss": 0.3025,
+      "step": 2350
+    },
+    {
+      "epoch": 2.835946924004825,
+      "grad_norm": 0.20137028715015345,
+      "learning_rate": 3.0384271671134943e-06,
+      "loss": 0.3159,
+      "step": 2351
+    },
+    {
+      "epoch": 2.837153196622437,
+      "grad_norm": 0.21523120752022207,
+      "learning_rate": 3.0160857908847184e-06,
+      "loss": 0.3052,
+      "step": 2352
+    },
+    {
+      "epoch": 2.8383594692400482,
+      "grad_norm": 0.1855207216753987,
+      "learning_rate": 2.993744414655943e-06,
+      "loss": 0.2837,
+      "step": 2353
+    },
+    {
+      "epoch": 2.83956574185766,
+      "grad_norm": 0.2094086834658496,
+      "learning_rate": 2.9714030384271675e-06,
+      "loss": 0.2967,
+      "step": 2354
+    },
+    {
+      "epoch": 2.8407720144752715,
+      "grad_norm": 0.19794775999629896,
+      "learning_rate": 2.9490616621983915e-06,
+      "loss": 0.2842,
+      "step": 2355
+    },
+    {
+      "epoch": 2.841978287092883,
+      "grad_norm": 0.18988974466292338,
+      "learning_rate": 2.926720285969616e-06,
+      "loss": 0.2775,
+      "step": 2356
+    },
+    {
+      "epoch": 2.8431845597104948,
+      "grad_norm": 0.19913264082126989,
+      "learning_rate": 2.90437890974084e-06,
+      "loss": 0.2816,
+      "step": 2357
+    },
+    {
+      "epoch": 2.844390832328106,
+      "grad_norm": 0.19113590403389297,
+      "learning_rate": 2.8820375335120642e-06,
+      "loss": 0.2882,
+      "step": 2358
+    },
+    {
+      "epoch": 2.8455971049457176,
+      "grad_norm": 0.20391500027186676,
+      "learning_rate": 2.8596961572832887e-06,
+      "loss": 0.287,
+      "step": 2359
+    },
+    {
+      "epoch": 2.8468033775633295,
+      "grad_norm": 0.21140527601885406,
+      "learning_rate": 2.8373547810545133e-06,
+      "loss": 0.2966,
+      "step": 2360
+    },
+    {
+      "epoch": 2.848009650180941,
+      "grad_norm": 0.20533216153271688,
+      "learning_rate": 2.8150134048257373e-06,
+      "loss": 0.2924,
+      "step": 2361
+    },
+    {
+      "epoch": 2.8492159227985523,
+      "grad_norm": 0.20311908970852507,
+      "learning_rate": 2.7926720285969614e-06,
+      "loss": 0.3017,
+      "step": 2362
+    },
+    {
+      "epoch": 2.850422195416164,
+      "grad_norm": 0.19773289611540928,
+      "learning_rate": 2.770330652368186e-06,
+      "loss": 0.2833,
+      "step": 2363
+    },
+    {
+      "epoch": 2.8516284680337756,
+      "grad_norm": 0.2062842210490875,
+      "learning_rate": 2.7479892761394105e-06,
+      "loss": 0.293,
+      "step": 2364
+    },
+    {
+      "epoch": 2.852834740651387,
+      "grad_norm": 0.19703442031240742,
+      "learning_rate": 2.7256478999106345e-06,
+      "loss": 0.2901,
+      "step": 2365
+    },
+    {
+      "epoch": 2.854041013268999,
+      "grad_norm": 0.1979587715787965,
+      "learning_rate": 2.703306523681859e-06,
+      "loss": 0.2848,
+      "step": 2366
+    },
+    {
+      "epoch": 2.8552472858866103,
+      "grad_norm": 0.19236896429290468,
+      "learning_rate": 2.680965147453083e-06,
+      "loss": 0.2792,
+      "step": 2367
+    },
+    {
+      "epoch": 2.856453558504222,
+      "grad_norm": 0.19185791310903008,
+      "learning_rate": 2.6586237712243077e-06,
+      "loss": 0.2857,
+      "step": 2368
+    },
+    {
+      "epoch": 2.8576598311218335,
+      "grad_norm": 0.20097778464317417,
+      "learning_rate": 2.6362823949955317e-06,
+      "loss": 0.2912,
+      "step": 2369
+    },
+    {
+      "epoch": 2.8588661037394454,
+      "grad_norm": 0.2055714517459925,
+      "learning_rate": 2.6139410187667563e-06,
+      "loss": 0.2997,
+      "step": 2370
+    },
+    {
+      "epoch": 2.860072376357057,
+      "grad_norm": 0.21732786429246168,
+      "learning_rate": 2.5915996425379808e-06,
+      "loss": 0.3055,
+      "step": 2371
+    },
+    {
+      "epoch": 2.861278648974668,
+      "grad_norm": 0.2502005807261682,
+      "learning_rate": 2.5692582663092044e-06,
+      "loss": 0.2856,
+      "step": 2372
+    },
+    {
+      "epoch": 2.86248492159228,
+      "grad_norm": 0.19987073929777327,
+      "learning_rate": 2.546916890080429e-06,
+      "loss": 0.2814,
+      "step": 2373
+    },
+    {
+      "epoch": 2.8636911942098915,
+      "grad_norm": 0.21371971461929823,
+      "learning_rate": 2.5245755138516535e-06,
+      "loss": 0.2895,
+      "step": 2374
+    },
+    {
+      "epoch": 2.864897466827503,
+      "grad_norm": 0.1947082976853003,
+      "learning_rate": 2.502234137622878e-06,
+      "loss": 0.2793,
+      "step": 2375
+    },
+    {
+      "epoch": 2.8661037394451148,
+      "grad_norm": 0.2020722173392859,
+      "learning_rate": 2.479892761394102e-06,
+      "loss": 0.2902,
+      "step": 2376
+    },
+    {
+      "epoch": 2.867310012062726,
+      "grad_norm": 0.19209835857990548,
+      "learning_rate": 2.457551385165326e-06,
+      "loss": 0.2899,
+      "step": 2377
+    },
+    {
+      "epoch": 2.8685162846803376,
+      "grad_norm": 0.20529591382680887,
+      "learning_rate": 2.4352100089365507e-06,
+      "loss": 0.2825,
+      "step": 2378
+    },
+    {
+      "epoch": 2.8697225572979495,
+      "grad_norm": 0.20486394337649172,
+      "learning_rate": 2.4128686327077747e-06,
+      "loss": 0.2881,
+      "step": 2379
+    },
+    {
+      "epoch": 2.870928829915561,
+      "grad_norm": 0.20950607820875253,
+      "learning_rate": 2.3905272564789993e-06,
+      "loss": 0.2991,
+      "step": 2380
+    },
+    {
+      "epoch": 2.8721351025331723,
+      "grad_norm": 0.1908385872610009,
+      "learning_rate": 2.3681858802502238e-06,
+      "loss": 0.2861,
+      "step": 2381
+    },
+    {
+      "epoch": 2.873341375150784,
+      "grad_norm": 0.19922508299237268,
+      "learning_rate": 2.345844504021448e-06,
+      "loss": 0.2855,
+      "step": 2382
+    },
+    {
+      "epoch": 2.8745476477683956,
+      "grad_norm": 0.22166082316021662,
+      "learning_rate": 2.323503127792672e-06,
+      "loss": 0.2946,
+      "step": 2383
+    },
+    {
+      "epoch": 2.875753920386007,
+      "grad_norm": 0.19461242460581957,
+      "learning_rate": 2.3011617515638965e-06,
+      "loss": 0.2714,
+      "step": 2384
+    },
+    {
+      "epoch": 2.876960193003619,
+      "grad_norm": 0.19352881947289724,
+      "learning_rate": 2.278820375335121e-06,
+      "loss": 0.2721,
+      "step": 2385
+    },
+    {
+      "epoch": 2.8781664656212302,
+      "grad_norm": 0.22058385576253617,
+      "learning_rate": 2.256478999106345e-06,
+      "loss": 0.2794,
+      "step": 2386
+    },
+    {
+      "epoch": 2.879372738238842,
+      "grad_norm": 0.18948954109249133,
+      "learning_rate": 2.234137622877569e-06,
+      "loss": 0.2827,
+      "step": 2387
+    },
+    {
+      "epoch": 2.8805790108564535,
+      "grad_norm": 0.1901617667242354,
+      "learning_rate": 2.2117962466487937e-06,
+      "loss": 0.293,
+      "step": 2388
+    },
+    {
+      "epoch": 2.8817852834740654,
+      "grad_norm": 0.19227085609237324,
+      "learning_rate": 2.1894548704200177e-06,
+      "loss": 0.2796,
+      "step": 2389
+    },
+    {
+      "epoch": 2.882991556091677,
+      "grad_norm": 0.19896604225955727,
+      "learning_rate": 2.1671134941912423e-06,
+      "loss": 0.281,
+      "step": 2390
+    },
+    {
+      "epoch": 2.884197828709288,
+      "grad_norm": 0.1969715058172239,
+      "learning_rate": 2.1447721179624668e-06,
+      "loss": 0.2904,
+      "step": 2391
+    },
+    {
+      "epoch": 2.8854041013269,
+      "grad_norm": 0.2008657406945197,
+      "learning_rate": 2.122430741733691e-06,
+      "loss": 0.2803,
+      "step": 2392
+    },
+    {
+      "epoch": 2.8866103739445115,
+      "grad_norm": 0.2107471673459275,
+      "learning_rate": 2.100089365504915e-06,
+      "loss": 0.2879,
+      "step": 2393
+    },
+    {
+      "epoch": 2.887816646562123,
+      "grad_norm": 0.19150770147349974,
+      "learning_rate": 2.0777479892761395e-06,
+      "loss": 0.291,
+      "step": 2394
+    },
+    {
+      "epoch": 2.8890229191797347,
+      "grad_norm": 0.2041712230658877,
+      "learning_rate": 2.055406613047364e-06,
+      "loss": 0.2874,
+      "step": 2395
+    },
+    {
+      "epoch": 2.890229191797346,
+      "grad_norm": 0.19970618525293937,
+      "learning_rate": 2.033065236818588e-06,
+      "loss": 0.2895,
+      "step": 2396
+    },
+    {
+      "epoch": 2.8914354644149576,
+      "grad_norm": 0.21995699586382159,
+      "learning_rate": 2.0107238605898126e-06,
+      "loss": 0.2929,
+      "step": 2397
+    },
+    {
+      "epoch": 2.8926417370325694,
+      "grad_norm": 0.1973616517335882,
+      "learning_rate": 1.9883824843610367e-06,
+      "loss": 0.2936,
+      "step": 2398
+    },
+    {
+      "epoch": 2.893848009650181,
+      "grad_norm": 0.1989740318792929,
+      "learning_rate": 1.966041108132261e-06,
+      "loss": 0.2861,
+      "step": 2399
+    },
+    {
+      "epoch": 2.8950542822677923,
+      "grad_norm": 0.2112403964468139,
+      "learning_rate": 1.9436997319034853e-06,
+      "loss": 0.2825,
+      "step": 2400
+    },
+    {
+      "epoch": 2.896260554885404,
+      "grad_norm": 0.20828664919750745,
+      "learning_rate": 1.9213583556747098e-06,
+      "loss": 0.281,
+      "step": 2401
+    },
+    {
+      "epoch": 2.8974668275030155,
+      "grad_norm": 0.1955339081984292,
+      "learning_rate": 1.899016979445934e-06,
+      "loss": 0.2768,
+      "step": 2402
+    },
+    {
+      "epoch": 2.8986731001206274,
+      "grad_norm": 0.22204369188455533,
+      "learning_rate": 1.8766756032171582e-06,
+      "loss": 0.2874,
+      "step": 2403
+    },
+    {
+      "epoch": 2.899879372738239,
+      "grad_norm": 0.2077252151848404,
+      "learning_rate": 1.8543342269883825e-06,
+      "loss": 0.2991,
+      "step": 2404
+    },
+    {
+      "epoch": 2.9010856453558507,
+      "grad_norm": 0.21435787298444822,
+      "learning_rate": 1.831992850759607e-06,
+      "loss": 0.2943,
+      "step": 2405
+    },
+    {
+      "epoch": 2.902291917973462,
+      "grad_norm": 0.1911919947784039,
+      "learning_rate": 1.8096514745308313e-06,
+      "loss": 0.2815,
+      "step": 2406
+    },
+    {
+      "epoch": 2.9034981905910735,
+      "grad_norm": 0.18827809153347827,
+      "learning_rate": 1.7873100983020554e-06,
+      "loss": 0.2752,
+      "step": 2407
+    },
+    {
+      "epoch": 2.9047044632086854,
+      "grad_norm": 0.18931406811301763,
+      "learning_rate": 1.7649687220732799e-06,
+      "loss": 0.2853,
+      "step": 2408
+    },
+    {
+      "epoch": 2.9059107358262968,
+      "grad_norm": 0.22505141785294047,
+      "learning_rate": 1.7426273458445042e-06,
+      "loss": 0.2881,
+      "step": 2409
+    },
+    {
+      "epoch": 2.907117008443908,
+      "grad_norm": 0.21027595141026514,
+      "learning_rate": 1.7202859696157283e-06,
+      "loss": 0.2902,
+      "step": 2410
+    },
+    {
+      "epoch": 2.90832328106152,
+      "grad_norm": 0.20210659855206092,
+      "learning_rate": 1.6979445933869528e-06,
+      "loss": 0.2831,
+      "step": 2411
+    },
+    {
+      "epoch": 2.9095295536791315,
+      "grad_norm": 0.1842267253698121,
+      "learning_rate": 1.675603217158177e-06,
+      "loss": 0.2854,
+      "step": 2412
+    },
+    {
+      "epoch": 2.910735826296743,
+      "grad_norm": 0.2091748826161693,
+      "learning_rate": 1.6532618409294012e-06,
+      "loss": 0.2783,
+      "step": 2413
+    },
+    {
+      "epoch": 2.9119420989143547,
+      "grad_norm": 0.19331995813160882,
+      "learning_rate": 1.6309204647006257e-06,
+      "loss": 0.2742,
+      "step": 2414
+    },
+    {
+      "epoch": 2.913148371531966,
+      "grad_norm": 0.20789168735658,
+      "learning_rate": 1.60857908847185e-06,
+      "loss": 0.2834,
+      "step": 2415
+    },
+    {
+      "epoch": 2.9143546441495776,
+      "grad_norm": 0.22504179670692293,
+      "learning_rate": 1.5862377122430745e-06,
+      "loss": 0.2851,
+      "step": 2416
+    },
+    {
+      "epoch": 2.9155609167671894,
+      "grad_norm": 0.20044801055007996,
+      "learning_rate": 1.5638963360142984e-06,
+      "loss": 0.2793,
+      "step": 2417
+    },
+    {
+      "epoch": 2.916767189384801,
+      "grad_norm": 0.22060197936804501,
+      "learning_rate": 1.5415549597855229e-06,
+      "loss": 0.3086,
+      "step": 2418
+    },
+    {
+      "epoch": 2.9179734620024127,
+      "grad_norm": 0.21133137826378076,
+      "learning_rate": 1.5192135835567472e-06,
+      "loss": 0.2909,
+      "step": 2419
+    },
+    {
+      "epoch": 2.919179734620024,
+      "grad_norm": 0.19782064257474333,
+      "learning_rate": 1.4968722073279715e-06,
+      "loss": 0.278,
+      "step": 2420
+    },
+    {
+      "epoch": 2.920386007237636,
+      "grad_norm": 0.19534968132532227,
+      "learning_rate": 1.4745308310991958e-06,
+      "loss": 0.294,
+      "step": 2421
+    },
+    {
+      "epoch": 2.9215922798552474,
+      "grad_norm": 0.19077219973811435,
+      "learning_rate": 1.45218945487042e-06,
+      "loss": 0.2649,
+      "step": 2422
+    },
+    {
+      "epoch": 2.922798552472859,
+      "grad_norm": 0.18896064658647335,
+      "learning_rate": 1.4298480786416444e-06,
+      "loss": 0.2791,
+      "step": 2423
+    },
+    {
+      "epoch": 2.9240048250904707,
+      "grad_norm": 0.20355436076924752,
+      "learning_rate": 1.4075067024128687e-06,
+      "loss": 0.2912,
+      "step": 2424
+    },
+    {
+      "epoch": 2.925211097708082,
+      "grad_norm": 0.19105724867358181,
+      "learning_rate": 1.385165326184093e-06,
+      "loss": 0.281,
+      "step": 2425
+    },
+    {
+      "epoch": 2.9264173703256935,
+      "grad_norm": 0.1943840621294346,
+      "learning_rate": 1.3628239499553173e-06,
+      "loss": 0.2786,
+      "step": 2426
+    },
+    {
+      "epoch": 2.9276236429433053,
+      "grad_norm": 0.19061678832395212,
+      "learning_rate": 1.3404825737265416e-06,
+      "loss": 0.2776,
+      "step": 2427
+    },
+    {
+      "epoch": 2.9288299155609168,
+      "grad_norm": 0.20255117048830434,
+      "learning_rate": 1.3181411974977659e-06,
+      "loss": 0.2847,
+      "step": 2428
+    },
+    {
+      "epoch": 2.930036188178528,
+      "grad_norm": 0.20576627278355886,
+      "learning_rate": 1.2957998212689904e-06,
+      "loss": 0.3018,
+      "step": 2429
+    },
+    {
+      "epoch": 2.93124246079614,
+      "grad_norm": 0.20319570033218337,
+      "learning_rate": 1.2734584450402145e-06,
+      "loss": 0.2939,
+      "step": 2430
+    },
+    {
+      "epoch": 2.9324487334137515,
+      "grad_norm": 0.18540791897808953,
+      "learning_rate": 1.251117068811439e-06,
+      "loss": 0.2818,
+      "step": 2431
+    },
+    {
+      "epoch": 2.933655006031363,
+      "grad_norm": 0.18714821054278172,
+      "learning_rate": 1.228775692582663e-06,
+      "loss": 0.2899,
+      "step": 2432
+    },
+    {
+      "epoch": 2.9348612786489747,
+      "grad_norm": 0.19047448250427157,
+      "learning_rate": 1.2064343163538874e-06,
+      "loss": 0.2938,
+      "step": 2433
+    },
+    {
+      "epoch": 2.936067551266586,
+      "grad_norm": 0.1808164949151054,
+      "learning_rate": 1.1840929401251119e-06,
+      "loss": 0.273,
+      "step": 2434
+    },
+    {
+      "epoch": 2.9372738238841976,
+      "grad_norm": 0.18878761414697054,
+      "learning_rate": 1.161751563896336e-06,
+      "loss": 0.2898,
+      "step": 2435
+    },
+    {
+      "epoch": 2.9384800965018094,
+      "grad_norm": 0.20432487218307782,
+      "learning_rate": 1.1394101876675605e-06,
+      "loss": 0.2892,
+      "step": 2436
+    },
+    {
+      "epoch": 2.939686369119421,
+      "grad_norm": 0.18870479397965104,
+      "learning_rate": 1.1170688114387846e-06,
+      "loss": 0.2758,
+      "step": 2437
+    },
+    {
+      "epoch": 2.9408926417370327,
+      "grad_norm": 0.19391649412908152,
+      "learning_rate": 1.0947274352100089e-06,
+      "loss": 0.2812,
+      "step": 2438
+    },
+    {
+      "epoch": 2.942098914354644,
+      "grad_norm": 0.19901047358991764,
+      "learning_rate": 1.0723860589812334e-06,
+      "loss": 0.2915,
+      "step": 2439
+    },
+    {
+      "epoch": 2.943305186972256,
+      "grad_norm": 0.19947394557395554,
+      "learning_rate": 1.0500446827524575e-06,
+      "loss": 0.2731,
+      "step": 2440
+    },
+    {
+      "epoch": 2.9445114595898674,
+      "grad_norm": 0.1921348495014182,
+      "learning_rate": 1.027703306523682e-06,
+      "loss": 0.28,
+      "step": 2441
+    },
+    {
+      "epoch": 2.945717732207479,
+      "grad_norm": 0.19667583970583413,
+      "learning_rate": 1.0053619302949063e-06,
+      "loss": 0.2927,
+      "step": 2442
+    },
+    {
+      "epoch": 2.9469240048250906,
+      "grad_norm": 0.1875571885355997,
+      "learning_rate": 9.830205540661306e-07,
+      "loss": 0.2814,
+      "step": 2443
+    },
+    {
+      "epoch": 2.948130277442702,
+      "grad_norm": 0.19359347961380186,
+      "learning_rate": 9.606791778373549e-07,
+      "loss": 0.2836,
+      "step": 2444
+    },
+    {
+      "epoch": 2.9493365500603135,
+      "grad_norm": 0.1948502151304172,
+      "learning_rate": 9.383378016085791e-07,
+      "loss": 0.2765,
+      "step": 2445
+    },
+    {
+      "epoch": 2.9505428226779253,
+      "grad_norm": 0.1986095387977726,
+      "learning_rate": 9.159964253798035e-07,
+      "loss": 0.2939,
+      "step": 2446
+    },
+    {
+      "epoch": 2.9517490952955368,
+      "grad_norm": 0.20106803721894279,
+      "learning_rate": 8.936550491510277e-07,
+      "loss": 0.2807,
+      "step": 2447
+    },
+    {
+      "epoch": 2.952955367913148,
+      "grad_norm": 0.19573487567653272,
+      "learning_rate": 8.713136729222521e-07,
+      "loss": 0.2934,
+      "step": 2448
+    },
+    {
+      "epoch": 2.95416164053076,
+      "grad_norm": 0.19156764509284374,
+      "learning_rate": 8.489722966934764e-07,
+      "loss": 0.2672,
+      "step": 2449
+    },
+    {
+      "epoch": 2.9553679131483714,
+      "grad_norm": 0.19620055140220727,
+      "learning_rate": 8.266309204647006e-07,
+      "loss": 0.2925,
+      "step": 2450
+    },
+    {
+      "epoch": 2.956574185765983,
+      "grad_norm": 0.20150136995738124,
+      "learning_rate": 8.04289544235925e-07,
+      "loss": 0.2885,
+      "step": 2451
+    },
+    {
+      "epoch": 2.9577804583835947,
+      "grad_norm": 0.18808650789680123,
+      "learning_rate": 7.819481680071492e-07,
+      "loss": 0.2877,
+      "step": 2452
+    },
+    {
+      "epoch": 2.958986731001206,
+      "grad_norm": 0.20237101627697343,
+      "learning_rate": 7.596067917783736e-07,
+      "loss": 0.295,
+      "step": 2453
+    },
+    {
+      "epoch": 2.960193003618818,
+      "grad_norm": 0.1902312694516005,
+      "learning_rate": 7.372654155495979e-07,
+      "loss": 0.2761,
+      "step": 2454
+    },
+    {
+      "epoch": 2.9613992762364294,
+      "grad_norm": 0.1889315314894398,
+      "learning_rate": 7.149240393208222e-07,
+      "loss": 0.2949,
+      "step": 2455
+    },
+    {
+      "epoch": 2.9626055488540413,
+      "grad_norm": 0.19561298973269395,
+      "learning_rate": 6.925826630920465e-07,
+      "loss": 0.2881,
+      "step": 2456
+    },
+    {
+      "epoch": 2.9638118214716527,
+      "grad_norm": 0.19245025002510016,
+      "learning_rate": 6.702412868632708e-07,
+      "loss": 0.2836,
+      "step": 2457
+    },
+    {
+      "epoch": 2.965018094089264,
+      "grad_norm": 0.18665421617528258,
+      "learning_rate": 6.478999106344952e-07,
+      "loss": 0.2767,
+      "step": 2458
+    },
+    {
+      "epoch": 2.966224366706876,
+      "grad_norm": 0.21266300683887,
+      "learning_rate": 6.255585344057195e-07,
+      "loss": 0.2934,
+      "step": 2459
+    },
+    {
+      "epoch": 2.9674306393244874,
+      "grad_norm": 0.1933172786462093,
+      "learning_rate": 6.032171581769437e-07,
+      "loss": 0.2963,
+      "step": 2460
+    },
+    {
+      "epoch": 2.9686369119420988,
+      "grad_norm": 0.18850035175693095,
+      "learning_rate": 5.80875781948168e-07,
+      "loss": 0.2873,
+      "step": 2461
+    },
+    {
+      "epoch": 2.9698431845597106,
+      "grad_norm": 0.20078041695018908,
+      "learning_rate": 5.585344057193923e-07,
+      "loss": 0.2838,
+      "step": 2462
+    },
+    {
+      "epoch": 2.971049457177322,
+      "grad_norm": 0.21181100144588538,
+      "learning_rate": 5.361930294906167e-07,
+      "loss": 0.2913,
+      "step": 2463
+    },
+    {
+      "epoch": 2.9722557297949335,
+      "grad_norm": 0.207677897790048,
+      "learning_rate": 5.13851653261841e-07,
+      "loss": 0.2878,
+      "step": 2464
+    },
+    {
+      "epoch": 2.9734620024125453,
+      "grad_norm": 0.19564831186151077,
+      "learning_rate": 4.915102770330653e-07,
+      "loss": 0.2827,
+      "step": 2465
+    },
+    {
+      "epoch": 2.9746682750301567,
+      "grad_norm": 0.17970354493070714,
+      "learning_rate": 4.6916890080428954e-07,
+      "loss": 0.2792,
+      "step": 2466
+    },
+    {
+      "epoch": 2.975874547647768,
+      "grad_norm": 0.18873251167142202,
+      "learning_rate": 4.4682752457551384e-07,
+      "loss": 0.278,
+      "step": 2467
+    },
+    {
+      "epoch": 2.97708082026538,
+      "grad_norm": 0.18314280872829297,
+      "learning_rate": 4.244861483467382e-07,
+      "loss": 0.2812,
+      "step": 2468
+    },
+    {
+      "epoch": 2.9782870928829914,
+      "grad_norm": 0.1922397965492106,
+      "learning_rate": 4.021447721179625e-07,
+      "loss": 0.3008,
+      "step": 2469
+    },
+    {
+      "epoch": 2.9794933655006033,
+      "grad_norm": 0.20322955656849276,
+      "learning_rate": 3.798033958891868e-07,
+      "loss": 0.2938,
+      "step": 2470
+    },
+    {
+      "epoch": 2.9806996381182147,
+      "grad_norm": 0.18912287402642547,
+      "learning_rate": 3.574620196604111e-07,
+      "loss": 0.2856,
+      "step": 2471
+    },
+    {
+      "epoch": 2.9819059107358266,
+      "grad_norm": 0.18348812116265312,
+      "learning_rate": 3.351206434316354e-07,
+      "loss": 0.2734,
+      "step": 2472
+    },
+    {
+      "epoch": 2.983112183353438,
+      "grad_norm": 0.20205824365979516,
+      "learning_rate": 3.1277926720285975e-07,
+      "loss": 0.2856,
+      "step": 2473
+    },
+    {
+      "epoch": 2.9843184559710494,
+      "grad_norm": 0.18061090101631558,
+      "learning_rate": 2.90437890974084e-07,
+      "loss": 0.2889,
+      "step": 2474
+    },
+    {
+      "epoch": 2.9855247285886612,
+      "grad_norm": 0.19600190278992066,
+      "learning_rate": 2.6809651474530835e-07,
+      "loss": 0.2954,
+      "step": 2475
+    },
+    {
+      "epoch": 2.9867310012062727,
+      "grad_norm": 0.1851582944696054,
+      "learning_rate": 2.4575513851653265e-07,
+      "loss": 0.2702,
+      "step": 2476
+    },
+    {
+      "epoch": 2.987937273823884,
+      "grad_norm": 0.19177764461246363,
+      "learning_rate": 2.2341376228775692e-07,
+      "loss": 0.2818,
+      "step": 2477
+    },
+    {
+      "epoch": 2.989143546441496,
+      "grad_norm": 0.18832137389477646,
+      "learning_rate": 2.0107238605898125e-07,
+      "loss": 0.2744,
+      "step": 2478
+    },
+    {
+      "epoch": 2.9903498190591074,
+      "grad_norm": 0.1934068643268722,
+      "learning_rate": 1.7873100983020555e-07,
+      "loss": 0.2964,
+      "step": 2479
+    },
+    {
+      "epoch": 2.9915560916767188,
+      "grad_norm": 0.19723930708026718,
+      "learning_rate": 1.5638963360142987e-07,
+      "loss": 0.2963,
+      "step": 2480
+    },
+    {
+      "epoch": 2.9927623642943306,
+      "grad_norm": 0.18748029021039345,
+      "learning_rate": 1.3404825737265417e-07,
+      "loss": 0.2819,
+      "step": 2481
+    },
+    {
+      "epoch": 2.993968636911942,
+      "grad_norm": 0.19792207901814557,
+      "learning_rate": 1.1170688114387846e-07,
+      "loss": 0.2997,
+      "step": 2482
+    },
+    {
+      "epoch": 2.9951749095295535,
+      "grad_norm": 0.18747972922352799,
+      "learning_rate": 8.936550491510277e-08,
+      "loss": 0.2742,
+      "step": 2483
+    },
+    {
+      "epoch": 2.9963811821471653,
+      "grad_norm": 0.19061815955899578,
+      "learning_rate": 6.702412868632709e-08,
+      "loss": 0.2833,
+      "step": 2484
+    },
+    {
+      "epoch": 2.9975874547647767,
+      "grad_norm": 0.18986370796667845,
+      "learning_rate": 4.4682752457551387e-08,
+      "loss": 0.2966,
+      "step": 2485
+    },
+    {
+      "epoch": 2.9987937273823886,
+      "grad_norm": 0.1746287102520941,
+      "learning_rate": 2.2341376228775693e-08,
+      "loss": 0.2758,
+      "step": 2486
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.189086510210081,
+      "learning_rate": 0.0,
+      "loss": 0.2636,
+      "step": 2487
+    },
+    {
+      "epoch": 3.0,
+      "step": 2487,
+      "total_flos": 2.1271624359105004e+18,
+      "train_loss": 0.4341153975203184,
+      "train_runtime": 146020.7654,
+      "train_samples_per_second": 0.272,
+      "train_steps_per_second": 0.017
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 2487,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 2.1271624359105004e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}