diff --git "a/checkpoint-3500/trainer_state.json" "b/checkpoint-3500/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-3500/trainer_state.json"
@@ -0,0 +1,24533 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.33489618218352313,
+  "eval_steps": 500,
+  "global_step": 3500,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00033478406427854036,
+      "grad_norm": 2.5951156616210938,
+      "learning_rate": 5e-06,
+      "loss": 2.0872,
+      "step": 1
+    },
+    {
+      "epoch": 0.0006695681285570807,
+      "grad_norm": 2.011127471923828,
+      "learning_rate": 1e-05,
+      "loss": 2.0068,
+      "step": 2
+    },
+    {
+      "epoch": 0.001004352192835621,
+      "grad_norm": 2.058666467666626,
+      "learning_rate": 1.5e-05,
+      "loss": 2.0258,
+      "step": 3
+    },
+    {
+      "epoch": 0.0013391362571141614,
+      "grad_norm": 2.2120566368103027,
+      "learning_rate": 2e-05,
+      "loss": 2.0142,
+      "step": 4
+    },
+    {
+      "epoch": 0.0016739203213927017,
+      "grad_norm": 2.370628595352173,
+      "learning_rate": 2.5e-05,
+      "loss": 2.0344,
+      "step": 5
+    },
+    {
+      "epoch": 0.002008704385671242,
+      "grad_norm": 1.437334418296814,
+      "learning_rate": 3e-05,
+      "loss": 1.9203,
+      "step": 6
+    },
+    {
+      "epoch": 0.002343488449949782,
+      "grad_norm": 1.1889039278030396,
+      "learning_rate": 3.5e-05,
+      "loss": 1.9264,
+      "step": 7
+    },
+    {
+      "epoch": 0.002678272514228323,
+      "grad_norm": 1.0925624370574951,
+      "learning_rate": 4e-05,
+      "loss": 1.9148,
+      "step": 8
+    },
+    {
+      "epoch": 0.003013056578506863,
+      "grad_norm": 0.5106806755065918,
+      "learning_rate": 4.5e-05,
+      "loss": 1.9098,
+      "step": 9
+    },
+    {
+      "epoch": 0.0033478406427854034,
+      "grad_norm": 0.506732702255249,
+      "learning_rate": 5e-05,
+      "loss": 1.832,
+      "step": 10
+    },
+    {
+      "epoch": 0.0036826247070639436,
+      "grad_norm": 0.47460949420928955,
+      "learning_rate": 5.500000000000001e-05,
+      "loss": 1.879,
+      "step": 11
+    },
+    {
+      "epoch": 0.004017408771342484,
+      "grad_norm": 0.5833293199539185,
+      "learning_rate": 6e-05,
+      "loss": 1.8777,
+      "step": 12
+    },
+    {
+      "epoch": 0.004352192835621024,
+      "grad_norm": 0.4383687376976013,
+      "learning_rate": 6.500000000000001e-05,
+      "loss": 1.8559,
+      "step": 13
+    },
+    {
+      "epoch": 0.004686976899899564,
+      "grad_norm": 0.35056746006011963,
+      "learning_rate": 7e-05,
+      "loss": 1.7573,
+      "step": 14
+    },
+    {
+      "epoch": 0.0050217609641781055,
+      "grad_norm": 0.5545064210891724,
+      "learning_rate": 7.500000000000001e-05,
+      "loss": 1.7541,
+      "step": 15
+    },
+    {
+      "epoch": 0.005356545028456646,
+      "grad_norm": 0.3440060019493103,
+      "learning_rate": 8e-05,
+      "loss": 1.6876,
+      "step": 16
+    },
+    {
+      "epoch": 0.005691329092735186,
+      "grad_norm": 0.36561861634254456,
+      "learning_rate": 8.5e-05,
+      "loss": 1.7454,
+      "step": 17
+    },
+    {
+      "epoch": 0.006026113157013726,
+      "grad_norm": 0.282402902841568,
+      "learning_rate": 9e-05,
+      "loss": 1.8184,
+      "step": 18
+    },
+    {
+      "epoch": 0.0063608972212922665,
+      "grad_norm": 0.1981375813484192,
+      "learning_rate": 9.5e-05,
+      "loss": 1.7448,
+      "step": 19
+    },
+    {
+      "epoch": 0.006695681285570807,
+      "grad_norm": 0.16754242777824402,
+      "learning_rate": 0.0001,
+      "loss": 1.7555,
+      "step": 20
+    },
+    {
+      "epoch": 0.007030465349849347,
+      "grad_norm": 0.17915141582489014,
+      "learning_rate": 0.0001,
+      "loss": 1.7533,
+      "step": 21
+    },
+    {
+      "epoch": 0.007365249414127887,
+      "grad_norm": 0.1990516483783722,
+      "learning_rate": 0.0001,
+      "loss": 1.6819,
+      "step": 22
+    },
+    {
+      "epoch": 0.0077000334784064275,
+      "grad_norm": 0.20808538794517517,
+      "learning_rate": 0.0001,
+      "loss": 1.7345,
+      "step": 23
+    },
+    {
+      "epoch": 0.008034817542684968,
+      "grad_norm": 0.2500799894332886,
+      "learning_rate": 0.0001,
+      "loss": 1.7636,
+      "step": 24
+    },
+    {
+      "epoch": 0.008369601606963508,
+      "grad_norm": 0.2141977846622467,
+      "learning_rate": 0.0001,
+      "loss": 1.7475,
+      "step": 25
+    },
+    {
+      "epoch": 0.008704385671242048,
+      "grad_norm": 0.2018044888973236,
+      "learning_rate": 0.0001,
+      "loss": 1.6445,
+      "step": 26
+    },
+    {
+      "epoch": 0.009039169735520589,
+      "grad_norm": 0.19822722673416138,
+      "learning_rate": 0.0001,
+      "loss": 1.7604,
+      "step": 27
+    },
+    {
+      "epoch": 0.009373953799799129,
+      "grad_norm": 0.18675795197486877,
+      "learning_rate": 0.0001,
+      "loss": 1.7743,
+      "step": 28
+    },
+    {
+      "epoch": 0.009708737864077669,
+      "grad_norm": 0.16032469272613525,
+      "learning_rate": 0.0001,
+      "loss": 1.7221,
+      "step": 29
+    },
+    {
+      "epoch": 0.010043521928356211,
+      "grad_norm": 0.17107701301574707,
+      "learning_rate": 0.0001,
+      "loss": 1.7514,
+      "step": 30
+    },
+    {
+      "epoch": 0.010378305992634751,
+      "grad_norm": 0.1589154601097107,
+      "learning_rate": 0.0001,
+      "loss": 1.6738,
+      "step": 31
+    },
+    {
+      "epoch": 0.010713090056913292,
+      "grad_norm": 0.13346004486083984,
+      "learning_rate": 0.0001,
+      "loss": 1.6011,
+      "step": 32
+    },
+    {
+      "epoch": 0.011047874121191832,
+      "grad_norm": 0.1687479168176651,
+      "learning_rate": 0.0001,
+      "loss": 1.7694,
+      "step": 33
+    },
+    {
+      "epoch": 0.011382658185470372,
+      "grad_norm": 0.14785747230052948,
+      "learning_rate": 0.0001,
+      "loss": 1.6836,
+      "step": 34
+    },
+    {
+      "epoch": 0.011717442249748912,
+      "grad_norm": 0.13441652059555054,
+      "learning_rate": 0.0001,
+      "loss": 1.7087,
+      "step": 35
+    },
+    {
+      "epoch": 0.012052226314027453,
+      "grad_norm": 0.13479024171829224,
+      "learning_rate": 0.0001,
+      "loss": 1.6456,
+      "step": 36
+    },
+    {
+      "epoch": 0.012387010378305993,
+      "grad_norm": 0.15816231071949005,
+      "learning_rate": 0.0001,
+      "loss": 1.6643,
+      "step": 37
+    },
+    {
+      "epoch": 0.012721794442584533,
+      "grad_norm": 0.12814071774482727,
+      "learning_rate": 0.0001,
+      "loss": 1.6382,
+      "step": 38
+    },
+    {
+      "epoch": 0.013056578506863073,
+      "grad_norm": 0.129450261592865,
+      "learning_rate": 0.0001,
+      "loss": 1.7623,
+      "step": 39
+    },
+    {
+      "epoch": 0.013391362571141614,
+      "grad_norm": 0.13946504890918732,
+      "learning_rate": 0.0001,
+      "loss": 1.8067,
+      "step": 40
+    },
+    {
+      "epoch": 0.013726146635420154,
+      "grad_norm": 0.1161293238401413,
+      "learning_rate": 0.0001,
+      "loss": 1.688,
+      "step": 41
+    },
+    {
+      "epoch": 0.014060930699698694,
+      "grad_norm": 0.11296379566192627,
+      "learning_rate": 0.0001,
+      "loss": 1.6035,
+      "step": 42
+    },
+    {
+      "epoch": 0.014395714763977234,
+      "grad_norm": 0.12507247924804688,
+      "learning_rate": 0.0001,
+      "loss": 1.7287,
+      "step": 43
+    },
+    {
+      "epoch": 0.014730498828255775,
+      "grad_norm": 0.11496929079294205,
+      "learning_rate": 0.0001,
+      "loss": 1.626,
+      "step": 44
+    },
+    {
+      "epoch": 0.015065282892534315,
+      "grad_norm": 0.13881774246692657,
+      "learning_rate": 0.0001,
+      "loss": 1.7501,
+      "step": 45
+    },
+    {
+      "epoch": 0.015400066956812855,
+      "grad_norm": 0.1255090981721878,
+      "learning_rate": 0.0001,
+      "loss": 1.6952,
+      "step": 46
+    },
+    {
+      "epoch": 0.015734851021091397,
+      "grad_norm": 0.11783197522163391,
+      "learning_rate": 0.0001,
+      "loss": 1.6256,
+      "step": 47
+    },
+    {
+      "epoch": 0.016069635085369936,
+      "grad_norm": 0.12152993679046631,
+      "learning_rate": 0.0001,
+      "loss": 1.6443,
+      "step": 48
+    },
+    {
+      "epoch": 0.016404419149648478,
+      "grad_norm": 0.12172088027000427,
+      "learning_rate": 0.0001,
+      "loss": 1.6927,
+      "step": 49
+    },
+    {
+      "epoch": 0.016739203213927016,
+      "grad_norm": 0.13490882515907288,
+      "learning_rate": 0.0001,
+      "loss": 1.7372,
+      "step": 50
+    },
+    {
+      "epoch": 0.017073987278205558,
+      "grad_norm": 0.1124483197927475,
+      "learning_rate": 0.0001,
+      "loss": 1.6206,
+      "step": 51
+    },
+    {
+      "epoch": 0.017408771342484097,
+      "grad_norm": 0.11569201201200485,
+      "learning_rate": 0.0001,
+      "loss": 1.7156,
+      "step": 52
+    },
+    {
+      "epoch": 0.01774355540676264,
+      "grad_norm": 0.12394021451473236,
+      "learning_rate": 0.0001,
+      "loss": 1.6132,
+      "step": 53
+    },
+    {
+      "epoch": 0.018078339471041177,
+      "grad_norm": 0.11930014938116074,
+      "learning_rate": 0.0001,
+      "loss": 1.6552,
+      "step": 54
+    },
+    {
+      "epoch": 0.01841312353531972,
+      "grad_norm": 0.1183612123131752,
+      "learning_rate": 0.0001,
+      "loss": 1.6953,
+      "step": 55
+    },
+    {
+      "epoch": 0.018747907599598258,
+      "grad_norm": 0.11677711457014084,
+      "learning_rate": 0.0001,
+      "loss": 1.6936,
+      "step": 56
+    },
+    {
+      "epoch": 0.0190826916638768,
+      "grad_norm": 0.12049452215433121,
+      "learning_rate": 0.0001,
+      "loss": 1.6381,
+      "step": 57
+    },
+    {
+      "epoch": 0.019417475728155338,
+      "grad_norm": 0.11653623729944229,
+      "learning_rate": 0.0001,
+      "loss": 1.7704,
+      "step": 58
+    },
+    {
+      "epoch": 0.01975225979243388,
+      "grad_norm": 0.12089766561985016,
+      "learning_rate": 0.0001,
+      "loss": 1.6819,
+      "step": 59
+    },
+    {
+      "epoch": 0.020087043856712422,
+      "grad_norm": 0.12823008000850677,
+      "learning_rate": 0.0001,
+      "loss": 1.7584,
+      "step": 60
+    },
+    {
+      "epoch": 0.02042182792099096,
+      "grad_norm": 0.12439601868391037,
+      "learning_rate": 0.0001,
+      "loss": 1.6761,
+      "step": 61
+    },
+    {
+      "epoch": 0.020756611985269503,
+      "grad_norm": 0.12000609189271927,
+      "learning_rate": 0.0001,
+      "loss": 1.7014,
+      "step": 62
+    },
+    {
+      "epoch": 0.02109139604954804,
+      "grad_norm": 0.12034812569618225,
+      "learning_rate": 0.0001,
+      "loss": 1.7128,
+      "step": 63
+    },
+    {
+      "epoch": 0.021426180113826583,
+      "grad_norm": 0.11534720659255981,
+      "learning_rate": 0.0001,
+      "loss": 1.694,
+      "step": 64
+    },
+    {
+      "epoch": 0.02176096417810512,
+      "grad_norm": 0.11633310467004776,
+      "learning_rate": 0.0001,
+      "loss": 1.6718,
+      "step": 65
+    },
+    {
+      "epoch": 0.022095748242383664,
+      "grad_norm": 0.13419900834560394,
+      "learning_rate": 0.0001,
+      "loss": 1.707,
+      "step": 66
+    },
+    {
+      "epoch": 0.022430532306662202,
+      "grad_norm": 0.11928509920835495,
+      "learning_rate": 0.0001,
+      "loss": 1.6935,
+      "step": 67
+    },
+    {
+      "epoch": 0.022765316370940744,
+      "grad_norm": 0.11948949843645096,
+      "learning_rate": 0.0001,
+      "loss": 1.6304,
+      "step": 68
+    },
+    {
+      "epoch": 0.023100100435219283,
+      "grad_norm": 0.12679244577884674,
+      "learning_rate": 0.0001,
+      "loss": 1.6605,
+      "step": 69
+    },
+    {
+      "epoch": 0.023434884499497825,
+      "grad_norm": 0.10675504058599472,
+      "learning_rate": 0.0001,
+      "loss": 1.6785,
+      "step": 70
+    },
+    {
+      "epoch": 0.023769668563776363,
+      "grad_norm": 0.12108162045478821,
+      "learning_rate": 0.0001,
+      "loss": 1.6695,
+      "step": 71
+    },
+    {
+      "epoch": 0.024104452628054905,
+      "grad_norm": 0.11032188683748245,
+      "learning_rate": 0.0001,
+      "loss": 1.7293,
+      "step": 72
+    },
+    {
+      "epoch": 0.024439236692333444,
+      "grad_norm": 0.11592775583267212,
+      "learning_rate": 0.0001,
+      "loss": 1.6726,
+      "step": 73
+    },
+    {
+      "epoch": 0.024774020756611986,
+      "grad_norm": 0.11566442996263504,
+      "learning_rate": 0.0001,
+      "loss": 1.6396,
+      "step": 74
+    },
+    {
+      "epoch": 0.025108804820890524,
+      "grad_norm": 0.11673177778720856,
+      "learning_rate": 0.0001,
+      "loss": 1.6223,
+      "step": 75
+    },
+    {
+      "epoch": 0.025443588885169066,
+      "grad_norm": 0.1140669733285904,
+      "learning_rate": 0.0001,
+      "loss": 1.6886,
+      "step": 76
+    },
+    {
+      "epoch": 0.025778372949447605,
+      "grad_norm": 0.11448585987091064,
+      "learning_rate": 0.0001,
+      "loss": 1.6765,
+      "step": 77
+    },
+    {
+      "epoch": 0.026113157013726147,
+      "grad_norm": 0.11363522708415985,
+      "learning_rate": 0.0001,
+      "loss": 1.6241,
+      "step": 78
+    },
+    {
+      "epoch": 0.02644794107800469,
+      "grad_norm": 0.10882357507944107,
+      "learning_rate": 0.0001,
+      "loss": 1.6495,
+      "step": 79
+    },
+    {
+      "epoch": 0.026782725142283227,
+      "grad_norm": 0.11577261239290237,
+      "learning_rate": 0.0001,
+      "loss": 1.6941,
+      "step": 80
+    },
+    {
+      "epoch": 0.02711750920656177,
+      "grad_norm": 0.12674297392368317,
+      "learning_rate": 0.0001,
+      "loss": 1.7615,
+      "step": 81
+    },
+    {
+      "epoch": 0.027452293270840308,
+      "grad_norm": 0.11801646649837494,
+      "learning_rate": 0.0001,
+      "loss": 1.6414,
+      "step": 82
+    },
+    {
+      "epoch": 0.02778707733511885,
+      "grad_norm": 0.11615725606679916,
+      "learning_rate": 0.0001,
+      "loss": 1.6586,
+      "step": 83
+    },
+    {
+      "epoch": 0.028121861399397388,
+      "grad_norm": 0.1159651130437851,
+      "learning_rate": 0.0001,
+      "loss": 1.6371,
+      "step": 84
+    },
+    {
+      "epoch": 0.02845664546367593,
+      "grad_norm": 0.12539416551589966,
+      "learning_rate": 0.0001,
+      "loss": 1.7152,
+      "step": 85
+    },
+    {
+      "epoch": 0.02879142952795447,
+      "grad_norm": 0.10691766440868378,
+      "learning_rate": 0.0001,
+      "loss": 1.552,
+      "step": 86
+    },
+    {
+      "epoch": 0.02912621359223301,
+      "grad_norm": 0.11859432607889175,
+      "learning_rate": 0.0001,
+      "loss": 1.6516,
+      "step": 87
+    },
+    {
+      "epoch": 0.02946099765651155,
+      "grad_norm": 0.12362800538539886,
+      "learning_rate": 0.0001,
+      "loss": 1.6944,
+      "step": 88
+    },
+    {
+      "epoch": 0.02979578172079009,
+      "grad_norm": 0.12135861068964005,
+      "learning_rate": 0.0001,
+      "loss": 1.6703,
+      "step": 89
+    },
+    {
+      "epoch": 0.03013056578506863,
+      "grad_norm": 0.15077495574951172,
+      "learning_rate": 0.0001,
+      "loss": 1.7522,
+      "step": 90
+    },
+    {
+      "epoch": 0.03046534984934717,
+      "grad_norm": 0.1137770563364029,
+      "learning_rate": 0.0001,
+      "loss": 1.6263,
+      "step": 91
+    },
+    {
+      "epoch": 0.03080013391362571,
+      "grad_norm": 0.11616989970207214,
+      "learning_rate": 0.0001,
+      "loss": 1.7166,
+      "step": 92
+    },
+    {
+      "epoch": 0.031134917977904252,
+      "grad_norm": 0.14210130274295807,
+      "learning_rate": 0.0001,
+      "loss": 1.7889,
+      "step": 93
+    },
+    {
+      "epoch": 0.031469702042182794,
+      "grad_norm": 0.1261507272720337,
+      "learning_rate": 0.0001,
+      "loss": 1.6593,
+      "step": 94
+    },
+    {
+      "epoch": 0.03180448610646133,
+      "grad_norm": 0.13197694718837738,
+      "learning_rate": 0.0001,
+      "loss": 1.6182,
+      "step": 95
+    },
+    {
+      "epoch": 0.03213927017073987,
+      "grad_norm": 0.11830636113882065,
+      "learning_rate": 0.0001,
+      "loss": 1.6373,
+      "step": 96
+    },
+    {
+      "epoch": 0.03247405423501841,
+      "grad_norm": 0.12643662095069885,
+      "learning_rate": 0.0001,
+      "loss": 1.6601,
+      "step": 97
+    },
+    {
+      "epoch": 0.032808838299296955,
+      "grad_norm": 0.13787776231765747,
+      "learning_rate": 0.0001,
+      "loss": 1.7496,
+      "step": 98
+    },
+    {
+      "epoch": 0.033143622363575494,
+      "grad_norm": 0.1096898540854454,
+      "learning_rate": 0.0001,
+      "loss": 1.5582,
+      "step": 99
+    },
+    {
+      "epoch": 0.03347840642785403,
+      "grad_norm": 0.13948234915733337,
+      "learning_rate": 0.0001,
+      "loss": 1.6281,
+      "step": 100
+    },
+    {
+      "epoch": 0.03381319049213258,
+      "grad_norm": 0.11294490098953247,
+      "learning_rate": 0.0001,
+      "loss": 1.6703,
+      "step": 101
+    },
+    {
+      "epoch": 0.034147974556411116,
+      "grad_norm": 0.12141433358192444,
+      "learning_rate": 0.0001,
+      "loss": 1.6553,
+      "step": 102
+    },
+    {
+      "epoch": 0.034482758620689655,
+      "grad_norm": 0.13332489132881165,
+      "learning_rate": 0.0001,
+      "loss": 1.6761,
+      "step": 103
+    },
+    {
+      "epoch": 0.03481754268496819,
+      "grad_norm": 0.12173039466142654,
+      "learning_rate": 0.0001,
+      "loss": 1.6304,
+      "step": 104
+    },
+    {
+      "epoch": 0.03515232674924674,
+      "grad_norm": 0.12168910354375839,
+      "learning_rate": 0.0001,
+      "loss": 1.6396,
+      "step": 105
+    },
+    {
+      "epoch": 0.03548711081352528,
+      "grad_norm": 0.1244431585073471,
+      "learning_rate": 0.0001,
+      "loss": 1.6463,
+      "step": 106
+    },
+    {
+      "epoch": 0.035821894877803816,
+      "grad_norm": 0.12028734385967255,
+      "learning_rate": 0.0001,
+      "loss": 1.684,
+      "step": 107
+    },
+    {
+      "epoch": 0.036156678942082354,
+      "grad_norm": 0.12029126286506653,
+      "learning_rate": 0.0001,
+      "loss": 1.6799,
+      "step": 108
+    },
+    {
+      "epoch": 0.0364914630063609,
+      "grad_norm": 0.11806860566139221,
+      "learning_rate": 0.0001,
+      "loss": 1.7245,
+      "step": 109
+    },
+    {
+      "epoch": 0.03682624707063944,
+      "grad_norm": 0.12406452000141144,
+      "learning_rate": 0.0001,
+      "loss": 1.6881,
+      "step": 110
+    },
+    {
+      "epoch": 0.03716103113491798,
+      "grad_norm": 0.118985615670681,
+      "learning_rate": 0.0001,
+      "loss": 1.6675,
+      "step": 111
+    },
+    {
+      "epoch": 0.037495815199196515,
+      "grad_norm": 0.12949040532112122,
+      "learning_rate": 0.0001,
+      "loss": 1.6871,
+      "step": 112
+    },
+    {
+      "epoch": 0.03783059926347506,
+      "grad_norm": 0.12375173717737198,
+      "learning_rate": 0.0001,
+      "loss": 1.6234,
+      "step": 113
+    },
+    {
+      "epoch": 0.0381653833277536,
+      "grad_norm": 0.11779066920280457,
+      "learning_rate": 0.0001,
+      "loss": 1.7399,
+      "step": 114
+    },
+    {
+      "epoch": 0.03850016739203214,
+      "grad_norm": 0.1195269301533699,
+      "learning_rate": 0.0001,
+      "loss": 1.65,
+      "step": 115
+    },
+    {
+      "epoch": 0.038834951456310676,
+      "grad_norm": 0.11929327249526978,
+      "learning_rate": 0.0001,
+      "loss": 1.6214,
+      "step": 116
+    },
+    {
+      "epoch": 0.03916973552058922,
+      "grad_norm": 0.11532218009233475,
+      "learning_rate": 0.0001,
+      "loss": 1.6395,
+      "step": 117
+    },
+    {
+      "epoch": 0.03950451958486776,
+      "grad_norm": 0.11126700043678284,
+      "learning_rate": 0.0001,
+      "loss": 1.622,
+      "step": 118
+    },
+    {
+      "epoch": 0.0398393036491463,
+      "grad_norm": 0.1309433877468109,
+      "learning_rate": 0.0001,
+      "loss": 1.5791,
+      "step": 119
+    },
+    {
+      "epoch": 0.040174087713424844,
+      "grad_norm": 0.12015924602746964,
+      "learning_rate": 0.0001,
+      "loss": 1.655,
+      "step": 120
+    },
+    {
+      "epoch": 0.04050887177770338,
+      "grad_norm": 0.12615351378917694,
+      "learning_rate": 0.0001,
+      "loss": 1.6215,
+      "step": 121
+    },
+    {
+      "epoch": 0.04084365584198192,
+      "grad_norm": 0.1387631893157959,
+      "learning_rate": 0.0001,
+      "loss": 1.7451,
+      "step": 122
+    },
+    {
+      "epoch": 0.04117843990626046,
+      "grad_norm": 0.1166117936372757,
+      "learning_rate": 0.0001,
+      "loss": 1.6537,
+      "step": 123
+    },
+    {
+      "epoch": 0.041513223970539005,
+      "grad_norm": 0.1521015763282776,
+      "learning_rate": 0.0001,
+      "loss": 1.6545,
+      "step": 124
+    },
+    {
+      "epoch": 0.041848008034817544,
+      "grad_norm": 0.1296280473470688,
+      "learning_rate": 0.0001,
+      "loss": 1.6355,
+      "step": 125
+    },
+    {
+      "epoch": 0.04218279209909608,
+      "grad_norm": 0.13189557194709778,
+      "learning_rate": 0.0001,
+      "loss": 1.5868,
+      "step": 126
+    },
+    {
+      "epoch": 0.04251757616337462,
+      "grad_norm": 0.1445418745279312,
+      "learning_rate": 0.0001,
+      "loss": 1.7444,
+      "step": 127
+    },
+    {
+      "epoch": 0.042852360227653166,
+      "grad_norm": 0.11560577899217606,
+      "learning_rate": 0.0001,
+      "loss": 1.6468,
+      "step": 128
+    },
+    {
+      "epoch": 0.043187144291931705,
+      "grad_norm": 0.16312864422798157,
+      "learning_rate": 0.0001,
+      "loss": 1.6734,
+      "step": 129
+    },
+    {
+      "epoch": 0.04352192835621024,
+      "grad_norm": 0.1284494251012802,
+      "learning_rate": 0.0001,
+      "loss": 1.6643,
+      "step": 130
+    },
+    {
+      "epoch": 0.04385671242048878,
+      "grad_norm": 0.11743518710136414,
+      "learning_rate": 0.0001,
+      "loss": 1.6273,
+      "step": 131
+    },
+    {
+      "epoch": 0.04419149648476733,
+      "grad_norm": 0.17127898335456848,
+      "learning_rate": 0.0001,
+      "loss": 1.5955,
+      "step": 132
+    },
+    {
+      "epoch": 0.044526280549045866,
+      "grad_norm": 0.1554144024848938,
+      "learning_rate": 0.0001,
+      "loss": 1.7738,
+      "step": 133
+    },
+    {
+      "epoch": 0.044861064613324404,
+      "grad_norm": 0.13085848093032837,
+      "learning_rate": 0.0001,
+      "loss": 1.5957,
+      "step": 134
+    },
+    {
+      "epoch": 0.04519584867760294,
+      "grad_norm": 0.1883288025856018,
+      "learning_rate": 0.0001,
+      "loss": 1.6159,
+      "step": 135
+    },
+    {
+      "epoch": 0.04553063274188149,
+      "grad_norm": 0.11826716363430023,
+      "learning_rate": 0.0001,
+      "loss": 1.6284,
+      "step": 136
+    },
+    {
+      "epoch": 0.04586541680616003,
+      "grad_norm": 0.15767724812030792,
+      "learning_rate": 0.0001,
+      "loss": 1.682,
+      "step": 137
+    },
+    {
+      "epoch": 0.046200200870438565,
+      "grad_norm": 0.14300817251205444,
+      "learning_rate": 0.0001,
+      "loss": 1.6152,
+      "step": 138
+    },
+    {
+      "epoch": 0.04653498493471711,
+      "grad_norm": 0.11646521836519241,
+      "learning_rate": 0.0001,
+      "loss": 1.6343,
+      "step": 139
+    },
+    {
+      "epoch": 0.04686976899899565,
+      "grad_norm": 0.12624727189540863,
+      "learning_rate": 0.0001,
+      "loss": 1.6128,
+      "step": 140
+    },
+    {
+      "epoch": 0.04720455306327419,
+      "grad_norm": 0.14111122488975525,
+      "learning_rate": 0.0001,
+      "loss": 1.618,
+      "step": 141
+    },
+    {
+      "epoch": 0.047539337127552726,
+      "grad_norm": 0.1404058188199997,
+      "learning_rate": 0.0001,
+      "loss": 1.66,
+      "step": 142
+    },
+    {
+      "epoch": 0.04787412119183127,
+      "grad_norm": 0.12555940449237823,
+      "learning_rate": 0.0001,
+      "loss": 1.666,
+      "step": 143
+    },
+    {
+      "epoch": 0.04820890525610981,
+      "grad_norm": 0.14494475722312927,
+      "learning_rate": 0.0001,
+      "loss": 1.6147,
+      "step": 144
+    },
+    {
+      "epoch": 0.04854368932038835,
+      "grad_norm": 0.12508632242679596,
+      "learning_rate": 0.0001,
+      "loss": 1.5765,
+      "step": 145
+    },
+    {
+      "epoch": 0.04887847338466689,
+      "grad_norm": 0.11790450662374496,
+      "learning_rate": 0.0001,
+      "loss": 1.7342,
+      "step": 146
+    },
+    {
+      "epoch": 0.04921325744894543,
+      "grad_norm": 0.1416400671005249,
+      "learning_rate": 0.0001,
+      "loss": 1.6673,
+      "step": 147
+    },
+    {
+      "epoch": 0.04954804151322397,
+      "grad_norm": 0.13537850975990295,
+      "learning_rate": 0.0001,
+      "loss": 1.6328,
+      "step": 148
+    },
+    {
+      "epoch": 0.04988282557750251,
+      "grad_norm": 0.12219058722257614,
+      "learning_rate": 0.0001,
+      "loss": 1.6677,
+      "step": 149
+    },
+    {
+      "epoch": 0.05021760964178105,
+      "grad_norm": 0.1398639678955078,
+      "learning_rate": 0.0001,
+      "loss": 1.6454,
+      "step": 150
+    },
+    {
+      "epoch": 0.050552393706059594,
+      "grad_norm": 0.14572647213935852,
+      "learning_rate": 0.0001,
+      "loss": 1.6094,
+      "step": 151
+    },
+    {
+      "epoch": 0.05088717777033813,
+      "grad_norm": 0.10937194526195526,
+      "learning_rate": 0.0001,
+      "loss": 1.5776,
+      "step": 152
+    },
+    {
+      "epoch": 0.05122196183461667,
+      "grad_norm": 0.1404120773077011,
+      "learning_rate": 0.0001,
+      "loss": 1.6112,
+      "step": 153
+    },
+    {
+      "epoch": 0.05155674589889521,
+      "grad_norm": 0.1480460911989212,
+      "learning_rate": 0.0001,
+      "loss": 1.6196,
+      "step": 154
+    },
+    {
+      "epoch": 0.051891529963173755,
+      "grad_norm": 0.10971348732709885,
+      "learning_rate": 0.0001,
+      "loss": 1.5744,
+      "step": 155
+    },
+    {
+      "epoch": 0.05222631402745229,
+      "grad_norm": 0.1468382179737091,
+      "learning_rate": 0.0001,
+      "loss": 1.7518,
+      "step": 156
+    },
+    {
+      "epoch": 0.05256109809173083,
+      "grad_norm": 0.13429516553878784,
+      "learning_rate": 0.0001,
+      "loss": 1.5812,
+      "step": 157
+    },
+    {
+      "epoch": 0.05289588215600938,
+      "grad_norm": 0.11399335414171219,
+      "learning_rate": 0.0001,
+      "loss": 1.6812,
+      "step": 158
+    },
+    {
+      "epoch": 0.053230666220287916,
+      "grad_norm": 0.13944409787654877,
+      "learning_rate": 0.0001,
+      "loss": 1.6789,
+      "step": 159
+    },
+    {
+      "epoch": 0.053565450284566454,
+      "grad_norm": 0.1390630453824997,
+      "learning_rate": 0.0001,
+      "loss": 1.6368,
+      "step": 160
+    },
+    {
+      "epoch": 0.05390023434884499,
+      "grad_norm": 0.1098702922463417,
+      "learning_rate": 0.0001,
+      "loss": 1.5462,
+      "step": 161
+    },
+    {
+      "epoch": 0.05423501841312354,
+      "grad_norm": 0.13710471987724304,
+      "learning_rate": 0.0001,
+      "loss": 1.7208,
+      "step": 162
+    },
+    {
+      "epoch": 0.05456980247740208,
+      "grad_norm": 0.1283336579799652,
+      "learning_rate": 0.0001,
+      "loss": 1.6648,
+      "step": 163
+    },
+    {
+      "epoch": 0.054904586541680615,
+      "grad_norm": 0.11550601571798325,
+      "learning_rate": 0.0001,
+      "loss": 1.7409,
+      "step": 164
+    },
+    {
+      "epoch": 0.055239370605959154,
+      "grad_norm": 0.12028289586305618,
+      "learning_rate": 0.0001,
+      "loss": 1.6685,
+      "step": 165
+    },
+    {
+      "epoch": 0.0555741546702377,
+      "grad_norm": 0.13237926363945007,
+      "learning_rate": 0.0001,
+      "loss": 1.6639,
+      "step": 166
+    },
+    {
+      "epoch": 0.05590893873451624,
+      "grad_norm": 0.11385014653205872,
+      "learning_rate": 0.0001,
+      "loss": 1.6742,
+      "step": 167
+    },
+    {
+      "epoch": 0.056243722798794776,
+      "grad_norm": 0.13613030314445496,
+      "learning_rate": 0.0001,
+      "loss": 1.6898,
+      "step": 168
+    },
+    {
+      "epoch": 0.056578506863073315,
+      "grad_norm": 0.12617048621177673,
+      "learning_rate": 0.0001,
+      "loss": 1.6239,
+      "step": 169
+    },
+    {
+      "epoch": 0.05691329092735186,
+      "grad_norm": 0.11637625098228455,
+      "learning_rate": 0.0001,
+      "loss": 1.6362,
+      "step": 170
+    },
+    {
+      "epoch": 0.0572480749916304,
+      "grad_norm": 0.13217699527740479,
+      "learning_rate": 0.0001,
+      "loss": 1.6319,
+      "step": 171
+    },
+    {
+      "epoch": 0.05758285905590894,
+      "grad_norm": 0.12088079750537872,
+      "learning_rate": 0.0001,
+      "loss": 1.4997,
+      "step": 172
+    },
+    {
+      "epoch": 0.057917643120187476,
+      "grad_norm": 0.11359237879514694,
+      "learning_rate": 0.0001,
+      "loss": 1.564,
+      "step": 173
+    },
+    {
+      "epoch": 0.05825242718446602,
+      "grad_norm": 0.12509793043136597,
+      "learning_rate": 0.0001,
+      "loss": 1.6855,
+      "step": 174
+    },
+    {
+      "epoch": 0.05858721124874456,
+      "grad_norm": 0.1233699694275856,
+      "learning_rate": 0.0001,
+      "loss": 1.665,
+      "step": 175
+    },
+    {
+      "epoch": 0.0589219953130231,
+      "grad_norm": 0.11172114312648773,
+      "learning_rate": 0.0001,
+      "loss": 1.6242,
+      "step": 176
+    },
+    {
+      "epoch": 0.059256779377301644,
+      "grad_norm": 0.12242110818624496,
+      "learning_rate": 0.0001,
+      "loss": 1.6736,
+      "step": 177
+    },
+    {
+      "epoch": 0.05959156344158018,
+      "grad_norm": 0.12275474518537521,
+      "learning_rate": 0.0001,
+      "loss": 1.6373,
+      "step": 178
+    },
+    {
+      "epoch": 0.05992634750585872,
+      "grad_norm": 0.11666038632392883,
+      "learning_rate": 0.0001,
+      "loss": 1.6957,
+      "step": 179
+    },
+    {
+      "epoch": 0.06026113157013726,
+      "grad_norm": 0.1209944486618042,
+      "learning_rate": 0.0001,
+      "loss": 1.618,
+      "step": 180
+    },
+    {
+      "epoch": 0.060595915634415805,
+      "grad_norm": 0.12028312683105469,
+      "learning_rate": 0.0001,
+      "loss": 1.6738,
+      "step": 181
+    },
+    {
+      "epoch": 0.06093069969869434,
+      "grad_norm": 0.11835712194442749,
+      "learning_rate": 0.0001,
+      "loss": 1.6348,
+      "step": 182
+    },
+    {
+      "epoch": 0.06126548376297288,
+      "grad_norm": 0.13166043162345886,
+      "learning_rate": 0.0001,
+      "loss": 1.6064,
+      "step": 183
+    },
+    {
+      "epoch": 0.06160026782725142,
+      "grad_norm": 0.1366170346736908,
+      "learning_rate": 0.0001,
+      "loss": 1.674,
+      "step": 184
+    },
+    {
+      "epoch": 0.061935051891529966,
+      "grad_norm": 0.12185468524694443,
+      "learning_rate": 0.0001,
+      "loss": 1.5695,
+      "step": 185
+    },
+    {
+      "epoch": 0.062269835955808504,
+      "grad_norm": 0.12310407310724258,
+      "learning_rate": 0.0001,
+      "loss": 1.6799,
+      "step": 186
+    },
+    {
+      "epoch": 0.06260462002008704,
+      "grad_norm": 0.14412462711334229,
+      "learning_rate": 0.0001,
+      "loss": 1.5855,
+      "step": 187
+    },
+    {
+      "epoch": 0.06293940408436559,
+      "grad_norm": 0.11908841878175735,
+      "learning_rate": 0.0001,
+      "loss": 1.5752,
+      "step": 188
+    },
+    {
+      "epoch": 0.06327418814864412,
+      "grad_norm": 0.12137061357498169,
+      "learning_rate": 0.0001,
+      "loss": 1.6018,
+      "step": 189
+    },
+    {
+      "epoch": 0.06360897221292267,
+      "grad_norm": 0.128020778298378,
+      "learning_rate": 0.0001,
+      "loss": 1.5894,
+      "step": 190
+    },
+    {
+      "epoch": 0.06394375627720121,
+      "grad_norm": 0.13447493314743042,
+      "learning_rate": 0.0001,
+      "loss": 1.5884,
+      "step": 191
+    },
+    {
+      "epoch": 0.06427854034147974,
+      "grad_norm": 0.11885492503643036,
+      "learning_rate": 0.0001,
+      "loss": 1.6245,
+      "step": 192
+    },
+    {
+      "epoch": 0.06461332440575829,
+      "grad_norm": 0.13066913187503815,
+      "learning_rate": 0.0001,
+      "loss": 1.6807,
+      "step": 193
+    },
+    {
+      "epoch": 0.06494810847003682,
+      "grad_norm": 0.12650778889656067,
+      "learning_rate": 0.0001,
+      "loss": 1.6498,
+      "step": 194
+    },
+    {
+      "epoch": 0.06528289253431536,
+      "grad_norm": 0.116504967212677,
+      "learning_rate": 0.0001,
+      "loss": 1.6037,
+      "step": 195
+    },
+    {
+      "epoch": 0.06561767659859391,
+      "grad_norm": 0.12200898677110672,
+      "learning_rate": 0.0001,
+      "loss": 1.5816,
+      "step": 196
+    },
+    {
+      "epoch": 0.06595246066287244,
+      "grad_norm": 0.13350239396095276,
+      "learning_rate": 0.0001,
+      "loss": 1.6281,
+      "step": 197
+    },
+    {
+      "epoch": 0.06628724472715099,
+      "grad_norm": 0.12119137495756149,
+      "learning_rate": 0.0001,
+      "loss": 1.5747,
+      "step": 198
+    },
+    {
+      "epoch": 0.06662202879142953,
+      "grad_norm": 0.12292595952749252,
+      "learning_rate": 0.0001,
+      "loss": 1.6294,
+      "step": 199
+    },
+    {
+      "epoch": 0.06695681285570806,
+      "grad_norm": 0.14958657324314117,
+      "learning_rate": 0.0001,
+      "loss": 1.7248,
+      "step": 200
+    },
+    {
+      "epoch": 0.06729159691998661,
+      "grad_norm": 0.1206580251455307,
+      "learning_rate": 0.0001,
+      "loss": 1.647,
+      "step": 201
+    },
+    {
+      "epoch": 0.06762638098426516,
+      "grad_norm": 0.13404549658298492,
+      "learning_rate": 0.0001,
+      "loss": 1.6827,
+      "step": 202
+    },
+    {
+      "epoch": 0.06796116504854369,
+      "grad_norm": 0.11746184527873993,
+      "learning_rate": 0.0001,
+      "loss": 1.5827,
+      "step": 203
+    },
+    {
+      "epoch": 0.06829594911282223,
+      "grad_norm": 0.1220933049917221,
+      "learning_rate": 0.0001,
+      "loss": 1.6209,
+      "step": 204
+    },
+    {
+      "epoch": 0.06863073317710076,
+      "grad_norm": 0.1395500898361206,
+      "learning_rate": 0.0001,
+      "loss": 1.6691,
+      "step": 205
+    },
+    {
+      "epoch": 0.06896551724137931,
+      "grad_norm": 0.12085775285959244,
+      "learning_rate": 0.0001,
+      "loss": 1.6186,
+      "step": 206
+    },
+    {
+      "epoch": 0.06930030130565785,
+      "grad_norm": 0.139579176902771,
+      "learning_rate": 0.0001,
+      "loss": 1.6357,
+      "step": 207
+    },
+    {
+      "epoch": 0.06963508536993639,
+      "grad_norm": 0.12011922895908356,
+      "learning_rate": 0.0001,
+      "loss": 1.5418,
+      "step": 208
+    },
+    {
+      "epoch": 0.06996986943421493,
+      "grad_norm": 0.11939892917871475,
+      "learning_rate": 0.0001,
+      "loss": 1.5816,
+      "step": 209
+    },
+    {
+      "epoch": 0.07030465349849348,
+      "grad_norm": 0.12651924788951874,
+      "learning_rate": 0.0001,
+      "loss": 1.5286,
+      "step": 210
+    },
+    {
+      "epoch": 0.07063943756277201,
+      "grad_norm": 0.13420534133911133,
+      "learning_rate": 0.0001,
+      "loss": 1.6213,
+      "step": 211
+    },
+    {
+      "epoch": 0.07097422162705055,
+      "grad_norm": 0.11868797987699509,
+      "learning_rate": 0.0001,
+      "loss": 1.6367,
+      "step": 212
+    },
+    {
+      "epoch": 0.07130900569132909,
+      "grad_norm": 0.11338218301534653,
+      "learning_rate": 0.0001,
+      "loss": 1.517,
+      "step": 213
+    },
+    {
+      "epoch": 0.07164378975560763,
+      "grad_norm": 0.14230981469154358,
+      "learning_rate": 0.0001,
+      "loss": 1.6773,
+      "step": 214
+    },
+    {
+      "epoch": 0.07197857381988618,
+      "grad_norm": 0.11315491795539856,
+      "learning_rate": 0.0001,
+      "loss": 1.5564,
+      "step": 215
+    },
+    {
+      "epoch": 0.07231335788416471,
+      "grad_norm": 0.12009023874998093,
+      "learning_rate": 0.0001,
+      "loss": 1.6317,
+      "step": 216
+    },
+    {
+      "epoch": 0.07264814194844325,
+      "grad_norm": 0.1332681030035019,
+      "learning_rate": 0.0001,
+      "loss": 1.6393,
+      "step": 217
+    },
+    {
+      "epoch": 0.0729829260127218,
+      "grad_norm": 0.12581905722618103,
+      "learning_rate": 0.0001,
+      "loss": 1.7155,
+      "step": 218
+    },
+    {
+      "epoch": 0.07331771007700033,
+      "grad_norm": 0.12259216606616974,
+      "learning_rate": 0.0001,
+      "loss": 1.661,
+      "step": 219
+    },
+    {
+      "epoch": 0.07365249414127888,
+      "grad_norm": 0.13090763986110687,
+      "learning_rate": 0.0001,
+      "loss": 1.6692,
+      "step": 220
+    },
+    {
+      "epoch": 0.07398727820555742,
+      "grad_norm": 0.11311494559049606,
+      "learning_rate": 0.0001,
+      "loss": 1.6653,
+      "step": 221
+    },
+    {
+      "epoch": 0.07432206226983595,
+      "grad_norm": 0.1307578831911087,
+      "learning_rate": 0.0001,
+      "loss": 1.5978,
+      "step": 222
+    },
+    {
+      "epoch": 0.0746568463341145,
+      "grad_norm": 0.12622885406017303,
+      "learning_rate": 0.0001,
+      "loss": 1.7782,
+      "step": 223
+    },
+    {
+      "epoch": 0.07499163039839303,
+      "grad_norm": 0.11902297288179398,
+      "learning_rate": 0.0001,
+      "loss": 1.5689,
+      "step": 224
+    },
+    {
+      "epoch": 0.07532641446267158,
+      "grad_norm": 0.11696305125951767,
+      "learning_rate": 0.0001,
+      "loss": 1.6077,
+      "step": 225
+    },
+    {
+      "epoch": 0.07566119852695012,
+      "grad_norm": 0.11666855216026306,
+      "learning_rate": 0.0001,
+      "loss": 1.5568,
+      "step": 226
+    },
+    {
+      "epoch": 0.07599598259122865,
+      "grad_norm": 0.12056950479745865,
+      "learning_rate": 0.0001,
+      "loss": 1.6829,
+      "step": 227
+    },
+    {
+      "epoch": 0.0763307666555072,
+      "grad_norm": 0.11957021802663803,
+      "learning_rate": 0.0001,
+      "loss": 1.7184,
+      "step": 228
+    },
+    {
+      "epoch": 0.07666555071978574,
+      "grad_norm": 0.11590487509965897,
+      "learning_rate": 0.0001,
+      "loss": 1.6775,
+      "step": 229
+    },
+    {
+      "epoch": 0.07700033478406428,
+      "grad_norm": 0.11034328490495682,
+      "learning_rate": 0.0001,
+      "loss": 1.5773,
+      "step": 230
+    },
+    {
+      "epoch": 0.07733511884834282,
+      "grad_norm": 0.12097325176000595,
+      "learning_rate": 0.0001,
+      "loss": 1.5552,
+      "step": 231
+    },
+    {
+      "epoch": 0.07766990291262135,
+      "grad_norm": 0.11697199940681458,
+      "learning_rate": 0.0001,
+      "loss": 1.6762,
+      "step": 232
+    },
+    {
+      "epoch": 0.0780046869768999,
+      "grad_norm": 0.11488549411296844,
+      "learning_rate": 0.0001,
+      "loss": 1.6219,
+      "step": 233
+    },
+    {
+      "epoch": 0.07833947104117844,
+      "grad_norm": 0.12868645787239075,
+      "learning_rate": 0.0001,
+      "loss": 1.6596,
+      "step": 234
+    },
+    {
+      "epoch": 0.07867425510545697,
+      "grad_norm": 0.11428504437208176,
+      "learning_rate": 0.0001,
+      "loss": 1.5926,
+      "step": 235
+    },
+    {
+      "epoch": 0.07900903916973552,
+      "grad_norm": 0.14550745487213135,
+      "learning_rate": 0.0001,
+      "loss": 1.6773,
+      "step": 236
+    },
+    {
+      "epoch": 0.07934382323401407,
+      "grad_norm": 0.11800127476453781,
+      "learning_rate": 0.0001,
+      "loss": 1.7403,
+      "step": 237
+    },
+    {
+      "epoch": 0.0796786072982926,
+      "grad_norm": 0.12732075154781342,
+      "learning_rate": 0.0001,
+      "loss": 1.6886,
+      "step": 238
+    },
+    {
+      "epoch": 0.08001339136257114,
+      "grad_norm": 0.1188284233212471,
+      "learning_rate": 0.0001,
+      "loss": 1.6552,
+      "step": 239
+    },
+    {
+      "epoch": 0.08034817542684969,
+      "grad_norm": 0.12447573244571686,
+      "learning_rate": 0.0001,
+      "loss": 1.668,
+      "step": 240
+    },
+    {
+      "epoch": 0.08068295949112822,
+      "grad_norm": 0.129620760679245,
+      "learning_rate": 0.0001,
+      "loss": 1.6134,
+      "step": 241
+    },
+    {
+      "epoch": 0.08101774355540677,
+      "grad_norm": 0.12539665400981903,
+      "learning_rate": 0.0001,
+      "loss": 1.7069,
+      "step": 242
+    },
+    {
+      "epoch": 0.0813525276196853,
+      "grad_norm": 0.13554492592811584,
+      "learning_rate": 0.0001,
+      "loss": 1.6704,
+      "step": 243
+    },
+    {
+      "epoch": 0.08168731168396384,
+      "grad_norm": 0.11758473515510559,
+      "learning_rate": 0.0001,
+      "loss": 1.6329,
+      "step": 244
+    },
+    {
+      "epoch": 0.08202209574824239,
+      "grad_norm": 0.11309672147035599,
+      "learning_rate": 0.0001,
+      "loss": 1.5836,
+      "step": 245
+    },
+    {
+      "epoch": 0.08235687981252092,
+      "grad_norm": 0.12910054624080658,
+      "learning_rate": 0.0001,
+      "loss": 1.6104,
+      "step": 246
+    },
+    {
+      "epoch": 0.08269166387679946,
+      "grad_norm": 0.12267620116472244,
+      "learning_rate": 0.0001,
+      "loss": 1.6505,
+      "step": 247
+    },
+    {
+      "epoch": 0.08302644794107801,
+      "grad_norm": 0.12700802087783813,
+      "learning_rate": 0.0001,
+      "loss": 1.6474,
+      "step": 248
+    },
+    {
+      "epoch": 0.08336123200535654,
+      "grad_norm": 0.13106848299503326,
+      "learning_rate": 0.0001,
+      "loss": 1.7076,
+      "step": 249
+    },
+    {
+      "epoch": 0.08369601606963509,
+      "grad_norm": 0.12598051130771637,
+      "learning_rate": 0.0001,
+      "loss": 1.6463,
+      "step": 250
+    },
+    {
+      "epoch": 0.08403080013391362,
+      "grad_norm": 0.1270611584186554,
+      "learning_rate": 0.0001,
+      "loss": 1.6407,
+      "step": 251
+    },
+    {
+      "epoch": 0.08436558419819216,
+      "grad_norm": 0.1215846911072731,
+      "learning_rate": 0.0001,
+      "loss": 1.7082,
+      "step": 252
+    },
+    {
+      "epoch": 0.08470036826247071,
+      "grad_norm": 0.11944068968296051,
+      "learning_rate": 0.0001,
+      "loss": 1.6046,
+      "step": 253
+    },
+    {
+      "epoch": 0.08503515232674924,
+      "grad_norm": 0.12395983189344406,
+      "learning_rate": 0.0001,
+      "loss": 1.6444,
+      "step": 254
+    },
+    {
+      "epoch": 0.08536993639102779,
+      "grad_norm": 0.11616060882806778,
+      "learning_rate": 0.0001,
+      "loss": 1.6514,
+      "step": 255
+    },
+    {
+      "epoch": 0.08570472045530633,
+      "grad_norm": 0.1274399757385254,
+      "learning_rate": 0.0001,
+      "loss": 1.6023,
+      "step": 256
+    },
+    {
+      "epoch": 0.08603950451958486,
+      "grad_norm": 0.11419884115457535,
+      "learning_rate": 0.0001,
+      "loss": 1.6053,
+      "step": 257
+    },
+    {
+      "epoch": 0.08637428858386341,
+      "grad_norm": 0.11922091245651245,
+      "learning_rate": 0.0001,
+      "loss": 1.6771,
+      "step": 258
+    },
+    {
+      "epoch": 0.08670907264814195,
+      "grad_norm": 0.12727287411689758,
+      "learning_rate": 0.0001,
+      "loss": 1.5332,
+      "step": 259
+    },
+    {
+      "epoch": 0.08704385671242049,
+      "grad_norm": 0.12368068844079971,
+      "learning_rate": 0.0001,
+      "loss": 1.6962,
+      "step": 260
+    },
+    {
+      "epoch": 0.08737864077669903,
+      "grad_norm": 0.11546538770198822,
+      "learning_rate": 0.0001,
+      "loss": 1.6239,
+      "step": 261
+    },
+    {
+      "epoch": 0.08771342484097756,
+      "grad_norm": 0.13736455142498016,
+      "learning_rate": 0.0001,
+      "loss": 1.7133,
+      "step": 262
+    },
+    {
+      "epoch": 0.08804820890525611,
+      "grad_norm": 0.12773726880550385,
+      "learning_rate": 0.0001,
+      "loss": 1.6127,
+      "step": 263
+    },
+    {
+      "epoch": 0.08838299296953465,
+      "grad_norm": 0.12833422422409058,
+      "learning_rate": 0.0001,
+      "loss": 1.5803,
+      "step": 264
+    },
+    {
+      "epoch": 0.08871777703381319,
+      "grad_norm": 0.13427826762199402,
+      "learning_rate": 0.0001,
+      "loss": 1.5815,
+      "step": 265
+    },
+    {
+      "epoch": 0.08905256109809173,
+      "grad_norm": 0.1173439621925354,
+      "learning_rate": 0.0001,
+      "loss": 1.5457,
+      "step": 266
+    },
+    {
+      "epoch": 0.08938734516237028,
+      "grad_norm": 0.12156970053911209,
+      "learning_rate": 0.0001,
+      "loss": 1.5969,
+      "step": 267
+    },
+    {
+      "epoch": 0.08972212922664881,
+      "grad_norm": 0.15133506059646606,
+      "learning_rate": 0.0001,
+      "loss": 1.6223,
+      "step": 268
+    },
+    {
+      "epoch": 0.09005691329092735,
+      "grad_norm": 0.13353589177131653,
+      "learning_rate": 0.0001,
+      "loss": 1.545,
+      "step": 269
+    },
+    {
+      "epoch": 0.09039169735520589,
+      "grad_norm": 0.12940257787704468,
+      "learning_rate": 0.0001,
+      "loss": 1.6135,
+      "step": 270
+    },
+    {
+      "epoch": 0.09072648141948443,
+      "grad_norm": 0.12897267937660217,
+      "learning_rate": 0.0001,
+      "loss": 1.6413,
+      "step": 271
+    },
+    {
+      "epoch": 0.09106126548376298,
+      "grad_norm": 0.12336087226867676,
+      "learning_rate": 0.0001,
+      "loss": 1.702,
+      "step": 272
+    },
+    {
+      "epoch": 0.09139604954804151,
+      "grad_norm": 0.11277737468481064,
+      "learning_rate": 0.0001,
+      "loss": 1.5743,
+      "step": 273
+    },
+    {
+      "epoch": 0.09173083361232005,
+      "grad_norm": 0.11659134924411774,
+      "learning_rate": 0.0001,
+      "loss": 1.6456,
+      "step": 274
+    },
+    {
+      "epoch": 0.0920656176765986,
+      "grad_norm": 0.11736118793487549,
+      "learning_rate": 0.0001,
+      "loss": 1.655,
+      "step": 275
+    },
+    {
+      "epoch": 0.09240040174087713,
+      "grad_norm": 0.12133463472127914,
+      "learning_rate": 0.0001,
+      "loss": 1.6771,
+      "step": 276
+    },
+    {
+      "epoch": 0.09273518580515568,
+      "grad_norm": 0.11516664177179337,
+      "learning_rate": 0.0001,
+      "loss": 1.5545,
+      "step": 277
+    },
+    {
+      "epoch": 0.09306996986943422,
+      "grad_norm": 0.10916180163621902,
+      "learning_rate": 0.0001,
+      "loss": 1.5301,
+      "step": 278
+    },
+    {
+      "epoch": 0.09340475393371275,
+      "grad_norm": 0.11232040077447891,
+      "learning_rate": 0.0001,
+      "loss": 1.5489,
+      "step": 279
+    },
+    {
+      "epoch": 0.0937395379979913,
+      "grad_norm": 0.12515543401241302,
+      "learning_rate": 0.0001,
+      "loss": 1.6817,
+      "step": 280
+    },
+    {
+      "epoch": 0.09407432206226983,
+      "grad_norm": 0.11998307704925537,
+      "learning_rate": 0.0001,
+      "loss": 1.563,
+      "step": 281
+    },
+    {
+      "epoch": 0.09440910612654838,
+      "grad_norm": 0.12774354219436646,
+      "learning_rate": 0.0001,
+      "loss": 1.622,
+      "step": 282
+    },
+    {
+      "epoch": 0.09474389019082692,
+      "grad_norm": 0.12023581564426422,
+      "learning_rate": 0.0001,
+      "loss": 1.5367,
+      "step": 283
+    },
+    {
+      "epoch": 0.09507867425510545,
+      "grad_norm": 0.12877605855464935,
+      "learning_rate": 0.0001,
+      "loss": 1.5806,
+      "step": 284
+    },
+    {
+      "epoch": 0.095413458319384,
+      "grad_norm": 0.11994509398937225,
+      "learning_rate": 0.0001,
+      "loss": 1.6017,
+      "step": 285
+    },
+    {
+      "epoch": 0.09574824238366254,
+      "grad_norm": 0.12522728741168976,
+      "learning_rate": 0.0001,
+      "loss": 1.6213,
+      "step": 286
+    },
+    {
+      "epoch": 0.09608302644794108,
+      "grad_norm": 0.13130401074886322,
+      "learning_rate": 0.0001,
+      "loss": 1.6211,
+      "step": 287
+    },
+    {
+      "epoch": 0.09641781051221962,
+      "grad_norm": 0.1242026537656784,
+      "learning_rate": 0.0001,
+      "loss": 1.6428,
+      "step": 288
+    },
+    {
+      "epoch": 0.09675259457649815,
+      "grad_norm": 0.12561045587062836,
+      "learning_rate": 0.0001,
+      "loss": 1.7275,
+      "step": 289
+    },
+    {
+      "epoch": 0.0970873786407767,
+      "grad_norm": 0.11756443232297897,
+      "learning_rate": 0.0001,
+      "loss": 1.5905,
+      "step": 290
+    },
+    {
+      "epoch": 0.09742216270505524,
+      "grad_norm": 0.11787443608045578,
+      "learning_rate": 0.0001,
+      "loss": 1.5809,
+      "step": 291
+    },
+    {
+      "epoch": 0.09775694676933377,
+      "grad_norm": 0.11708027869462967,
+      "learning_rate": 0.0001,
+      "loss": 1.6205,
+      "step": 292
+    },
+    {
+      "epoch": 0.09809173083361232,
+      "grad_norm": 0.12011709064245224,
+      "learning_rate": 0.0001,
+      "loss": 1.6327,
+      "step": 293
+    },
+    {
+      "epoch": 0.09842651489789087,
+      "grad_norm": 0.12868238985538483,
+      "learning_rate": 0.0001,
+      "loss": 1.7539,
+      "step": 294
+    },
+    {
+      "epoch": 0.0987612989621694,
+      "grad_norm": 0.11626073718070984,
+      "learning_rate": 0.0001,
+      "loss": 1.6877,
+      "step": 295
+    },
+    {
+      "epoch": 0.09909608302644794,
+      "grad_norm": 0.1279468834400177,
+      "learning_rate": 0.0001,
+      "loss": 1.635,
+      "step": 296
+    },
+    {
+      "epoch": 0.09943086709072649,
+      "grad_norm": 0.12956663966178894,
+      "learning_rate": 0.0001,
+      "loss": 1.5736,
+      "step": 297
+    },
+    {
+      "epoch": 0.09976565115500502,
+      "grad_norm": 0.11931903660297394,
+      "learning_rate": 0.0001,
+      "loss": 1.6534,
+      "step": 298
+    },
+    {
+      "epoch": 0.10010043521928357,
+      "grad_norm": 0.12837816774845123,
+      "learning_rate": 0.0001,
+      "loss": 1.5923,
+      "step": 299
+    },
+    {
+      "epoch": 0.1004352192835621,
+      "grad_norm": 0.12146858870983124,
+      "learning_rate": 0.0001,
+      "loss": 1.6206,
+      "step": 300
+    },
+    {
+      "epoch": 0.10077000334784064,
+      "grad_norm": 0.11455334722995758,
+      "learning_rate": 0.0001,
+      "loss": 1.5292,
+      "step": 301
+    },
+    {
+      "epoch": 0.10110478741211919,
+      "grad_norm": 0.12035822868347168,
+      "learning_rate": 0.0001,
+      "loss": 1.576,
+      "step": 302
+    },
+    {
+      "epoch": 0.10143957147639772,
+      "grad_norm": 0.12373282760381699,
+      "learning_rate": 0.0001,
+      "loss": 1.6688,
+      "step": 303
+    },
+    {
+      "epoch": 0.10177435554067626,
+      "grad_norm": 0.13985779881477356,
+      "learning_rate": 0.0001,
+      "loss": 1.667,
+      "step": 304
+    },
+    {
+      "epoch": 0.10210913960495481,
+      "grad_norm": 0.11246056109666824,
+      "learning_rate": 0.0001,
+      "loss": 1.6014,
+      "step": 305
+    },
+    {
+      "epoch": 0.10244392366923334,
+      "grad_norm": 0.13154080510139465,
+      "learning_rate": 0.0001,
+      "loss": 1.5909,
+      "step": 306
+    },
+    {
+      "epoch": 0.10277870773351189,
+      "grad_norm": 0.13235047459602356,
+      "learning_rate": 0.0001,
+      "loss": 1.6888,
+      "step": 307
+    },
+    {
+      "epoch": 0.10311349179779042,
+      "grad_norm": 0.13294562697410583,
+      "learning_rate": 0.0001,
+      "loss": 1.6534,
+      "step": 308
+    },
+    {
+      "epoch": 0.10344827586206896,
+      "grad_norm": 0.1274106800556183,
+      "learning_rate": 0.0001,
+      "loss": 1.7178,
+      "step": 309
+    },
+    {
+      "epoch": 0.10378305992634751,
+      "grad_norm": 0.11676975339651108,
+      "learning_rate": 0.0001,
+      "loss": 1.5587,
+      "step": 310
+    },
+    {
+      "epoch": 0.10411784399062604,
+      "grad_norm": 0.1180170550942421,
+      "learning_rate": 0.0001,
+      "loss": 1.5579,
+      "step": 311
+    },
+    {
+      "epoch": 0.10445262805490459,
+      "grad_norm": 0.1267906278371811,
+      "learning_rate": 0.0001,
+      "loss": 1.5994,
+      "step": 312
+    },
+    {
+      "epoch": 0.10478741211918313,
+      "grad_norm": 0.12398704141378403,
+      "learning_rate": 0.0001,
+      "loss": 1.5459,
+      "step": 313
+    },
+    {
+      "epoch": 0.10512219618346166,
+      "grad_norm": 0.12039758265018463,
+      "learning_rate": 0.0001,
+      "loss": 1.5995,
+      "step": 314
+    },
+    {
+      "epoch": 0.10545698024774021,
+      "grad_norm": 0.12191271781921387,
+      "learning_rate": 0.0001,
+      "loss": 1.5639,
+      "step": 315
+    },
+    {
+      "epoch": 0.10579176431201875,
+      "grad_norm": 0.1351427584886551,
+      "learning_rate": 0.0001,
+      "loss": 1.6553,
+      "step": 316
+    },
+    {
+      "epoch": 0.10612654837629729,
+      "grad_norm": 0.13542529940605164,
+      "learning_rate": 0.0001,
+      "loss": 1.5455,
+      "step": 317
+    },
+    {
+      "epoch": 0.10646133244057583,
+      "grad_norm": 0.13739462196826935,
+      "learning_rate": 0.0001,
+      "loss": 1.6414,
+      "step": 318
+    },
+    {
+      "epoch": 0.10679611650485436,
+      "grad_norm": 0.11810696870088577,
+      "learning_rate": 0.0001,
+      "loss": 1.7078,
+      "step": 319
+    },
+    {
+      "epoch": 0.10713090056913291,
+      "grad_norm": 0.13632580637931824,
+      "learning_rate": 0.0001,
+      "loss": 1.6044,
+      "step": 320
+    },
+    {
+      "epoch": 0.10746568463341145,
+      "grad_norm": 0.12454043328762054,
+      "learning_rate": 0.0001,
+      "loss": 1.6654,
+      "step": 321
+    },
+    {
+      "epoch": 0.10780046869768999,
+      "grad_norm": 0.11818061023950577,
+      "learning_rate": 0.0001,
+      "loss": 1.5693,
+      "step": 322
+    },
+    {
+      "epoch": 0.10813525276196853,
+      "grad_norm": 0.12229089438915253,
+      "learning_rate": 0.0001,
+      "loss": 1.6248,
+      "step": 323
+    },
+    {
+      "epoch": 0.10847003682624708,
+      "grad_norm": 0.11546499282121658,
+      "learning_rate": 0.0001,
+      "loss": 1.5091,
+      "step": 324
+    },
+    {
+      "epoch": 0.10880482089052561,
+      "grad_norm": 0.12005545943975449,
+      "learning_rate": 0.0001,
+      "loss": 1.5801,
+      "step": 325
+    },
+    {
+      "epoch": 0.10913960495480415,
+      "grad_norm": 0.12114623188972473,
+      "learning_rate": 0.0001,
+      "loss": 1.6552,
+      "step": 326
+    },
+    {
+      "epoch": 0.10947438901908269,
+      "grad_norm": 0.11608844995498657,
+      "learning_rate": 0.0001,
+      "loss": 1.5183,
+      "step": 327
+    },
+    {
+      "epoch": 0.10980917308336123,
+      "grad_norm": 0.11119306832551956,
+      "learning_rate": 0.0001,
+      "loss": 1.5515,
+      "step": 328
+    },
+    {
+      "epoch": 0.11014395714763978,
+      "grad_norm": 0.12586964666843414,
+      "learning_rate": 0.0001,
+      "loss": 1.6353,
+      "step": 329
+    },
+    {
+      "epoch": 0.11047874121191831,
+      "grad_norm": 0.127826526761055,
+      "learning_rate": 0.0001,
+      "loss": 1.7205,
+      "step": 330
+    },
+    {
+      "epoch": 0.11081352527619685,
+      "grad_norm": 0.11828092485666275,
+      "learning_rate": 0.0001,
+      "loss": 1.6711,
+      "step": 331
+    },
+    {
+      "epoch": 0.1111483093404754,
+      "grad_norm": 0.13583530485630035,
+      "learning_rate": 0.0001,
+      "loss": 1.6455,
+      "step": 332
+    },
+    {
+      "epoch": 0.11148309340475393,
+      "grad_norm": 0.11893647909164429,
+      "learning_rate": 0.0001,
+      "loss": 1.5707,
+      "step": 333
+    },
+    {
+      "epoch": 0.11181787746903248,
+      "grad_norm": 0.13151027262210846,
+      "learning_rate": 0.0001,
+      "loss": 1.6576,
+      "step": 334
+    },
+    {
+      "epoch": 0.11215266153331102,
+      "grad_norm": 0.11656352877616882,
+      "learning_rate": 0.0001,
+      "loss": 1.6456,
+      "step": 335
+    },
+    {
+      "epoch": 0.11248744559758955,
+      "grad_norm": 0.1267959475517273,
+      "learning_rate": 0.0001,
+      "loss": 1.5069,
+      "step": 336
+    },
+    {
+      "epoch": 0.1128222296618681,
+      "grad_norm": 0.12403184920549393,
+      "learning_rate": 0.0001,
+      "loss": 1.5273,
+      "step": 337
+    },
+    {
+      "epoch": 0.11315701372614663,
+      "grad_norm": 0.12692154943943024,
+      "learning_rate": 0.0001,
+      "loss": 1.6647,
+      "step": 338
+    },
+    {
+      "epoch": 0.11349179779042518,
+      "grad_norm": 0.11919606477022171,
+      "learning_rate": 0.0001,
+      "loss": 1.6833,
+      "step": 339
+    },
+    {
+      "epoch": 0.11382658185470372,
+      "grad_norm": 0.11304503679275513,
+      "learning_rate": 0.0001,
+      "loss": 1.5757,
+      "step": 340
+    },
+    {
+      "epoch": 0.11416136591898225,
+      "grad_norm": 0.11996794492006302,
+      "learning_rate": 0.0001,
+      "loss": 1.6102,
+      "step": 341
+    },
+    {
+      "epoch": 0.1144961499832608,
+      "grad_norm": 0.12606146931648254,
+      "learning_rate": 0.0001,
+      "loss": 1.59,
+      "step": 342
+    },
+    {
+      "epoch": 0.11483093404753934,
+      "grad_norm": 0.12146681547164917,
+      "learning_rate": 0.0001,
+      "loss": 1.4989,
+      "step": 343
+    },
+    {
+      "epoch": 0.11516571811181787,
+      "grad_norm": 0.13275377452373505,
+      "learning_rate": 0.0001,
+      "loss": 1.6152,
+      "step": 344
+    },
+    {
+      "epoch": 0.11550050217609642,
+      "grad_norm": 0.12684765458106995,
+      "learning_rate": 0.0001,
+      "loss": 1.542,
+      "step": 345
+    },
+    {
+      "epoch": 0.11583528624037495,
+      "grad_norm": 0.1186991035938263,
+      "learning_rate": 0.0001,
+      "loss": 1.573,
+      "step": 346
+    },
+    {
+      "epoch": 0.1161700703046535,
+      "grad_norm": 0.12221034616231918,
+      "learning_rate": 0.0001,
+      "loss": 1.6418,
+      "step": 347
+    },
+    {
+      "epoch": 0.11650485436893204,
+      "grad_norm": 0.11776617169380188,
+      "learning_rate": 0.0001,
+      "loss": 1.5821,
+      "step": 348
+    },
+    {
+      "epoch": 0.11683963843321057,
+      "grad_norm": 0.13464072346687317,
+      "learning_rate": 0.0001,
+      "loss": 1.6188,
+      "step": 349
+    },
+    {
+      "epoch": 0.11717442249748912,
+      "grad_norm": 0.13101482391357422,
+      "learning_rate": 0.0001,
+      "loss": 1.5194,
+      "step": 350
+    },
+    {
+      "epoch": 0.11750920656176767,
+      "grad_norm": 0.11970439553260803,
+      "learning_rate": 0.0001,
+      "loss": 1.5891,
+      "step": 351
+    },
+    {
+      "epoch": 0.1178439906260462,
+      "grad_norm": 0.11731956154108047,
+      "learning_rate": 0.0001,
+      "loss": 1.6441,
+      "step": 352
+    },
+    {
+      "epoch": 0.11817877469032474,
+      "grad_norm": 0.1163954809308052,
+      "learning_rate": 0.0001,
+      "loss": 1.5739,
+      "step": 353
+    },
+    {
+      "epoch": 0.11851355875460329,
+      "grad_norm": 0.13119016587734222,
+      "learning_rate": 0.0001,
+      "loss": 1.6667,
+      "step": 354
+    },
+    {
+      "epoch": 0.11884834281888182,
+      "grad_norm": 0.11406403034925461,
+      "learning_rate": 0.0001,
+      "loss": 1.5391,
+      "step": 355
+    },
+    {
+      "epoch": 0.11918312688316036,
+      "grad_norm": 0.12543243169784546,
+      "learning_rate": 0.0001,
+      "loss": 1.6413,
+      "step": 356
+    },
+    {
+      "epoch": 0.1195179109474389,
+      "grad_norm": 0.11639681458473206,
+      "learning_rate": 0.0001,
+      "loss": 1.5946,
+      "step": 357
+    },
+    {
+      "epoch": 0.11985269501171744,
+      "grad_norm": 0.11582693457603455,
+      "learning_rate": 0.0001,
+      "loss": 1.5797,
+      "step": 358
+    },
+    {
+      "epoch": 0.12018747907599599,
+      "grad_norm": 0.12131619453430176,
+      "learning_rate": 0.0001,
+      "loss": 1.5762,
+      "step": 359
+    },
+    {
+      "epoch": 0.12052226314027452,
+      "grad_norm": 0.1220826804637909,
+      "learning_rate": 0.0001,
+      "loss": 1.4938,
+      "step": 360
+    },
+    {
+      "epoch": 0.12085704720455306,
+      "grad_norm": 0.12737631797790527,
+      "learning_rate": 0.0001,
+      "loss": 1.5622,
+      "step": 361
+    },
+    {
+      "epoch": 0.12119183126883161,
+      "grad_norm": 0.12794937193393707,
+      "learning_rate": 0.0001,
+      "loss": 1.5852,
+      "step": 362
+    },
+    {
+      "epoch": 0.12152661533311014,
+      "grad_norm": 0.11786255985498428,
+      "learning_rate": 0.0001,
+      "loss": 1.6532,
+      "step": 363
+    },
+    {
+      "epoch": 0.12186139939738869,
+      "grad_norm": 0.12443582713603973,
+      "learning_rate": 0.0001,
+      "loss": 1.5664,
+      "step": 364
+    },
+    {
+      "epoch": 0.12219618346166722,
+      "grad_norm": 0.124130979180336,
+      "learning_rate": 0.0001,
+      "loss": 1.5809,
+      "step": 365
+    },
+    {
+      "epoch": 0.12253096752594576,
+      "grad_norm": 0.11969106644392014,
+      "learning_rate": 0.0001,
+      "loss": 1.5073,
+      "step": 366
+    },
+    {
+      "epoch": 0.12286575159022431,
+      "grad_norm": 0.12146104872226715,
+      "learning_rate": 0.0001,
+      "loss": 1.6322,
+      "step": 367
+    },
+    {
+      "epoch": 0.12320053565450284,
+      "grad_norm": 0.11919710785150528,
+      "learning_rate": 0.0001,
+      "loss": 1.6405,
+      "step": 368
+    },
+    {
+      "epoch": 0.12353531971878139,
+      "grad_norm": 0.12359990924596786,
+      "learning_rate": 0.0001,
+      "loss": 1.6564,
+      "step": 369
+    },
+    {
+      "epoch": 0.12387010378305993,
+      "grad_norm": 0.12216739356517792,
+      "learning_rate": 0.0001,
+      "loss": 1.658,
+      "step": 370
+    },
+    {
+      "epoch": 0.12420488784733846,
+      "grad_norm": 0.12388269603252411,
+      "learning_rate": 0.0001,
+      "loss": 1.6542,
+      "step": 371
+    },
+    {
+      "epoch": 0.12453967191161701,
+      "grad_norm": 0.12631452083587646,
+      "learning_rate": 0.0001,
+      "loss": 1.5741,
+      "step": 372
+    },
+    {
+      "epoch": 0.12487445597589555,
+      "grad_norm": 0.11718172580003738,
+      "learning_rate": 0.0001,
+      "loss": 1.5247,
+      "step": 373
+    },
+    {
+      "epoch": 0.12520924004017409,
+      "grad_norm": 0.11787404865026474,
+      "learning_rate": 0.0001,
+      "loss": 1.604,
+      "step": 374
+    },
+    {
+      "epoch": 0.12554402410445262,
+      "grad_norm": 0.1190713569521904,
+      "learning_rate": 0.0001,
+      "loss": 1.5771,
+      "step": 375
+    },
+    {
+      "epoch": 0.12587880816873118,
+      "grad_norm": 0.11780121177434921,
+      "learning_rate": 0.0001,
+      "loss": 1.6445,
+      "step": 376
+    },
+    {
+      "epoch": 0.1262135922330097,
+      "grad_norm": 0.11370184272527695,
+      "learning_rate": 0.0001,
+      "loss": 1.4544,
+      "step": 377
+    },
+    {
+      "epoch": 0.12654837629728824,
+      "grad_norm": 0.12931419909000397,
+      "learning_rate": 0.0001,
+      "loss": 1.5261,
+      "step": 378
+    },
+    {
+      "epoch": 0.1268831603615668,
+      "grad_norm": 0.11074584722518921,
+      "learning_rate": 0.0001,
+      "loss": 1.5329,
+      "step": 379
+    },
+    {
+      "epoch": 0.12721794442584533,
+      "grad_norm": 0.1251228302717209,
+      "learning_rate": 0.0001,
+      "loss": 1.6181,
+      "step": 380
+    },
+    {
+      "epoch": 0.12755272849012386,
+      "grad_norm": 0.11304245889186859,
+      "learning_rate": 0.0001,
+      "loss": 1.5198,
+      "step": 381
+    },
+    {
+      "epoch": 0.12788751255440242,
+      "grad_norm": 0.11219135671854019,
+      "learning_rate": 0.0001,
+      "loss": 1.494,
+      "step": 382
+    },
+    {
+      "epoch": 0.12822229661868095,
+      "grad_norm": 0.13162165880203247,
+      "learning_rate": 0.0001,
+      "loss": 1.6073,
+      "step": 383
+    },
+    {
+      "epoch": 0.12855708068295948,
+      "grad_norm": 0.11944107711315155,
+      "learning_rate": 0.0001,
+      "loss": 1.6021,
+      "step": 384
+    },
+    {
+      "epoch": 0.12889186474723804,
+      "grad_norm": 0.11878252029418945,
+      "learning_rate": 0.0001,
+      "loss": 1.6051,
+      "step": 385
+    },
+    {
+      "epoch": 0.12922664881151658,
+      "grad_norm": 0.1224270910024643,
+      "learning_rate": 0.0001,
+      "loss": 1.596,
+      "step": 386
+    },
+    {
+      "epoch": 0.1295614328757951,
+      "grad_norm": 0.12815283238887787,
+      "learning_rate": 0.0001,
+      "loss": 1.6652,
+      "step": 387
+    },
+    {
+      "epoch": 0.12989621694007364,
+      "grad_norm": 0.11265059560537338,
+      "learning_rate": 0.0001,
+      "loss": 1.5478,
+      "step": 388
+    },
+    {
+      "epoch": 0.1302310010043522,
+      "grad_norm": 0.12850640714168549,
+      "learning_rate": 0.0001,
+      "loss": 1.6242,
+      "step": 389
+    },
+    {
+      "epoch": 0.13056578506863073,
+      "grad_norm": 0.11487656831741333,
+      "learning_rate": 0.0001,
+      "loss": 1.6089,
+      "step": 390
+    },
+    {
+      "epoch": 0.13090056913290926,
+      "grad_norm": 0.1160978302359581,
+      "learning_rate": 0.0001,
+      "loss": 1.6248,
+      "step": 391
+    },
+    {
+      "epoch": 0.13123535319718782,
+      "grad_norm": 0.12001185864210129,
+      "learning_rate": 0.0001,
+      "loss": 1.5911,
+      "step": 392
+    },
+    {
+      "epoch": 0.13157013726146635,
+      "grad_norm": 0.11623065918684006,
+      "learning_rate": 0.0001,
+      "loss": 1.6194,
+      "step": 393
+    },
+    {
+      "epoch": 0.13190492132574488,
+      "grad_norm": 0.11913128942251205,
+      "learning_rate": 0.0001,
+      "loss": 1.6233,
+      "step": 394
+    },
+    {
+      "epoch": 0.13223970539002344,
+      "grad_norm": 0.11658355593681335,
+      "learning_rate": 0.0001,
+      "loss": 1.629,
+      "step": 395
+    },
+    {
+      "epoch": 0.13257448945430197,
+      "grad_norm": 0.11986858397722244,
+      "learning_rate": 0.0001,
+      "loss": 1.7414,
+      "step": 396
+    },
+    {
+      "epoch": 0.1329092735185805,
+      "grad_norm": 0.12659533321857452,
+      "learning_rate": 0.0001,
+      "loss": 1.6037,
+      "step": 397
+    },
+    {
+      "epoch": 0.13324405758285907,
+      "grad_norm": 0.11471698433160782,
+      "learning_rate": 0.0001,
+      "loss": 1.5939,
+      "step": 398
+    },
+    {
+      "epoch": 0.1335788416471376,
+      "grad_norm": 0.12152232974767685,
+      "learning_rate": 0.0001,
+      "loss": 1.5663,
+      "step": 399
+    },
+    {
+      "epoch": 0.13391362571141613,
+      "grad_norm": 0.12228668481111526,
+      "learning_rate": 0.0001,
+      "loss": 1.6717,
+      "step": 400
+    },
+    {
+      "epoch": 0.1342484097756947,
+      "grad_norm": 0.11998744308948517,
+      "learning_rate": 0.0001,
+      "loss": 1.6532,
+      "step": 401
+    },
+    {
+      "epoch": 0.13458319383997322,
+      "grad_norm": 0.12556074559688568,
+      "learning_rate": 0.0001,
+      "loss": 1.6477,
+      "step": 402
+    },
+    {
+      "epoch": 0.13491797790425175,
+      "grad_norm": 0.12216352671384811,
+      "learning_rate": 0.0001,
+      "loss": 1.6084,
+      "step": 403
+    },
+    {
+      "epoch": 0.1352527619685303,
+      "grad_norm": 0.1290225237607956,
+      "learning_rate": 0.0001,
+      "loss": 1.6369,
+      "step": 404
+    },
+    {
+      "epoch": 0.13558754603280884,
+      "grad_norm": 0.11453018337488174,
+      "learning_rate": 0.0001,
+      "loss": 1.5886,
+      "step": 405
+    },
+    {
+      "epoch": 0.13592233009708737,
+      "grad_norm": 0.12192509323358536,
+      "learning_rate": 0.0001,
+      "loss": 1.5788,
+      "step": 406
+    },
+    {
+      "epoch": 0.1362571141613659,
+      "grad_norm": 0.11374159157276154,
+      "learning_rate": 0.0001,
+      "loss": 1.5429,
+      "step": 407
+    },
+    {
+      "epoch": 0.13659189822564446,
+      "grad_norm": 0.11875942349433899,
+      "learning_rate": 0.0001,
+      "loss": 1.6524,
+      "step": 408
+    },
+    {
+      "epoch": 0.136926682289923,
+      "grad_norm": 0.12176533043384552,
+      "learning_rate": 0.0001,
+      "loss": 1.6572,
+      "step": 409
+    },
+    {
+      "epoch": 0.13726146635420153,
+      "grad_norm": 0.12200423330068588,
+      "learning_rate": 0.0001,
+      "loss": 1.7139,
+      "step": 410
+    },
+    {
+      "epoch": 0.1375962504184801,
+      "grad_norm": 0.11800340563058853,
+      "learning_rate": 0.0001,
+      "loss": 1.6276,
+      "step": 411
+    },
+    {
+      "epoch": 0.13793103448275862,
+      "grad_norm": 0.12321179360151291,
+      "learning_rate": 0.0001,
+      "loss": 1.6849,
+      "step": 412
+    },
+    {
+      "epoch": 0.13826581854703715,
+      "grad_norm": 0.12165375053882599,
+      "learning_rate": 0.0001,
+      "loss": 1.5823,
+      "step": 413
+    },
+    {
+      "epoch": 0.1386006026113157,
+      "grad_norm": 0.12587733566761017,
+      "learning_rate": 0.0001,
+      "loss": 1.5712,
+      "step": 414
+    },
+    {
+      "epoch": 0.13893538667559424,
+      "grad_norm": 0.11877655982971191,
+      "learning_rate": 0.0001,
+      "loss": 1.606,
+      "step": 415
+    },
+    {
+      "epoch": 0.13927017073987277,
+      "grad_norm": 0.11970411241054535,
+      "learning_rate": 0.0001,
+      "loss": 1.4995,
+      "step": 416
+    },
+    {
+      "epoch": 0.13960495480415133,
+      "grad_norm": 0.14770293235778809,
+      "learning_rate": 0.0001,
+      "loss": 1.7334,
+      "step": 417
+    },
+    {
+      "epoch": 0.13993973886842986,
+      "grad_norm": 0.11904104053974152,
+      "learning_rate": 0.0001,
+      "loss": 1.6258,
+      "step": 418
+    },
+    {
+      "epoch": 0.1402745229327084,
+      "grad_norm": 0.13043157756328583,
+      "learning_rate": 0.0001,
+      "loss": 1.5564,
+      "step": 419
+    },
+    {
+      "epoch": 0.14060930699698695,
+      "grad_norm": 0.1354888528585434,
+      "learning_rate": 0.0001,
+      "loss": 1.6391,
+      "step": 420
+    },
+    {
+      "epoch": 0.1409440910612655,
+      "grad_norm": 0.11834760010242462,
+      "learning_rate": 0.0001,
+      "loss": 1.5345,
+      "step": 421
+    },
+    {
+      "epoch": 0.14127887512554402,
+      "grad_norm": 0.13029152154922485,
+      "learning_rate": 0.0001,
+      "loss": 1.5007,
+      "step": 422
+    },
+    {
+      "epoch": 0.14161365918982258,
+      "grad_norm": 0.1352154165506363,
+      "learning_rate": 0.0001,
+      "loss": 1.5925,
+      "step": 423
+    },
+    {
+      "epoch": 0.1419484432541011,
+      "grad_norm": 0.13768818974494934,
+      "learning_rate": 0.0001,
+      "loss": 1.6513,
+      "step": 424
+    },
+    {
+      "epoch": 0.14228322731837964,
+      "grad_norm": 0.1345231682062149,
+      "learning_rate": 0.0001,
+      "loss": 1.6524,
+      "step": 425
+    },
+    {
+      "epoch": 0.14261801138265817,
+      "grad_norm": 0.11808541417121887,
+      "learning_rate": 0.0001,
+      "loss": 1.6038,
+      "step": 426
+    },
+    {
+      "epoch": 0.14295279544693673,
+      "grad_norm": 0.1403636336326599,
+      "learning_rate": 0.0001,
+      "loss": 1.5559,
+      "step": 427
+    },
+    {
+      "epoch": 0.14328757951121526,
+      "grad_norm": 0.13042065501213074,
+      "learning_rate": 0.0001,
+      "loss": 1.6516,
+      "step": 428
+    },
+    {
+      "epoch": 0.1436223635754938,
+      "grad_norm": 0.12809261679649353,
+      "learning_rate": 0.0001,
+      "loss": 1.5913,
+      "step": 429
+    },
+    {
+      "epoch": 0.14395714763977235,
+      "grad_norm": 0.13735899329185486,
+      "learning_rate": 0.0001,
+      "loss": 1.5814,
+      "step": 430
+    },
+    {
+      "epoch": 0.14429193170405089,
+      "grad_norm": 0.12458304315805435,
+      "learning_rate": 0.0001,
+      "loss": 1.6909,
+      "step": 431
+    },
+    {
+      "epoch": 0.14462671576832942,
+      "grad_norm": 0.11777736246585846,
+      "learning_rate": 0.0001,
+      "loss": 1.599,
+      "step": 432
+    },
+    {
+      "epoch": 0.14496149983260798,
+      "grad_norm": 0.11958497762680054,
+      "learning_rate": 0.0001,
+      "loss": 1.6224,
+      "step": 433
+    },
+    {
+      "epoch": 0.1452962838968865,
+      "grad_norm": 0.11626480519771576,
+      "learning_rate": 0.0001,
+      "loss": 1.6192,
+      "step": 434
+    },
+    {
+      "epoch": 0.14563106796116504,
+      "grad_norm": 0.12103210389614105,
+      "learning_rate": 0.0001,
+      "loss": 1.5581,
+      "step": 435
+    },
+    {
+      "epoch": 0.1459658520254436,
+      "grad_norm": 0.1175006702542305,
+      "learning_rate": 0.0001,
+      "loss": 1.6147,
+      "step": 436
+    },
+    {
+      "epoch": 0.14630063608972213,
+      "grad_norm": 0.1194823831319809,
+      "learning_rate": 0.0001,
+      "loss": 1.4559,
+      "step": 437
+    },
+    {
+      "epoch": 0.14663542015400066,
+      "grad_norm": 0.12060422450304031,
+      "learning_rate": 0.0001,
+      "loss": 1.706,
+      "step": 438
+    },
+    {
+      "epoch": 0.14697020421827922,
+      "grad_norm": 0.12133188545703888,
+      "learning_rate": 0.0001,
+      "loss": 1.6583,
+      "step": 439
+    },
+    {
+      "epoch": 0.14730498828255775,
+      "grad_norm": 0.11069684475660324,
+      "learning_rate": 0.0001,
+      "loss": 1.5626,
+      "step": 440
+    },
+    {
+      "epoch": 0.14763977234683628,
+      "grad_norm": 0.11735668778419495,
+      "learning_rate": 0.0001,
+      "loss": 1.5014,
+      "step": 441
+    },
+    {
+      "epoch": 0.14797455641111484,
+      "grad_norm": 0.11778223514556885,
+      "learning_rate": 0.0001,
+      "loss": 1.6483,
+      "step": 442
+    },
+    {
+      "epoch": 0.14830934047539338,
+      "grad_norm": 0.11628784239292145,
+      "learning_rate": 0.0001,
+      "loss": 1.5629,
+      "step": 443
+    },
+    {
+      "epoch": 0.1486441245396719,
+      "grad_norm": 0.12314952164888382,
+      "learning_rate": 0.0001,
+      "loss": 1.6362,
+      "step": 444
+    },
+    {
+      "epoch": 0.14897890860395044,
+      "grad_norm": 0.11853016167879105,
+      "learning_rate": 0.0001,
+      "loss": 1.6355,
+      "step": 445
+    },
+    {
+      "epoch": 0.149313692668229,
+      "grad_norm": 0.1322093904018402,
+      "learning_rate": 0.0001,
+      "loss": 1.7655,
+      "step": 446
+    },
+    {
+      "epoch": 0.14964847673250753,
+      "grad_norm": 0.11611328274011612,
+      "learning_rate": 0.0001,
+      "loss": 1.5878,
+      "step": 447
+    },
+    {
+      "epoch": 0.14998326079678606,
+      "grad_norm": 0.11989305913448334,
+      "learning_rate": 0.0001,
+      "loss": 1.5576,
+      "step": 448
+    },
+    {
+      "epoch": 0.15031804486106462,
+      "grad_norm": 0.11867792159318924,
+      "learning_rate": 0.0001,
+      "loss": 1.5453,
+      "step": 449
+    },
+    {
+      "epoch": 0.15065282892534315,
+      "grad_norm": 0.11955395340919495,
+      "learning_rate": 0.0001,
+      "loss": 1.6089,
+      "step": 450
+    },
+    {
+      "epoch": 0.15098761298962168,
+      "grad_norm": 0.13159644603729248,
+      "learning_rate": 0.0001,
+      "loss": 1.6053,
+      "step": 451
+    },
+    {
+      "epoch": 0.15132239705390024,
+      "grad_norm": 0.12264451384544373,
+      "learning_rate": 0.0001,
+      "loss": 1.6277,
+      "step": 452
+    },
+    {
+      "epoch": 0.15165718111817877,
+      "grad_norm": 0.1267840564250946,
+      "learning_rate": 0.0001,
+      "loss": 1.6047,
+      "step": 453
+    },
+    {
+      "epoch": 0.1519919651824573,
+      "grad_norm": 0.1316317319869995,
+      "learning_rate": 0.0001,
+      "loss": 1.5497,
+      "step": 454
+    },
+    {
+      "epoch": 0.15232674924673587,
+      "grad_norm": 0.12278051674365997,
+      "learning_rate": 0.0001,
+      "loss": 1.665,
+      "step": 455
+    },
+    {
+      "epoch": 0.1526615333110144,
+      "grad_norm": 0.13153740763664246,
+      "learning_rate": 0.0001,
+      "loss": 1.6262,
+      "step": 456
+    },
+    {
+      "epoch": 0.15299631737529293,
+      "grad_norm": 0.12118583172559738,
+      "learning_rate": 0.0001,
+      "loss": 1.5897,
+      "step": 457
+    },
+    {
+      "epoch": 0.1533311014395715,
+      "grad_norm": 0.12203945219516754,
+      "learning_rate": 0.0001,
+      "loss": 1.5709,
+      "step": 458
+    },
+    {
+      "epoch": 0.15366588550385002,
+      "grad_norm": 0.13483074307441711,
+      "learning_rate": 0.0001,
+      "loss": 1.662,
+      "step": 459
+    },
+    {
+      "epoch": 0.15400066956812855,
+      "grad_norm": 0.12122450023889542,
+      "learning_rate": 0.0001,
+      "loss": 1.6289,
+      "step": 460
+    },
+    {
+      "epoch": 0.1543354536324071,
+      "grad_norm": 0.1384558528661728,
+      "learning_rate": 0.0001,
+      "loss": 1.6274,
+      "step": 461
+    },
+    {
+      "epoch": 0.15467023769668564,
+      "grad_norm": 0.1436455249786377,
+      "learning_rate": 0.0001,
+      "loss": 1.6007,
+      "step": 462
+    },
+    {
+      "epoch": 0.15500502176096417,
+      "grad_norm": 0.12359965592622757,
+      "learning_rate": 0.0001,
+      "loss": 1.6757,
+      "step": 463
+    },
+    {
+      "epoch": 0.1553398058252427,
+      "grad_norm": 0.13497023284435272,
+      "learning_rate": 0.0001,
+      "loss": 1.6328,
+      "step": 464
+    },
+    {
+      "epoch": 0.15567458988952126,
+      "grad_norm": 0.12588655948638916,
+      "learning_rate": 0.0001,
+      "loss": 1.6066,
+      "step": 465
+    },
+    {
+      "epoch": 0.1560093739537998,
+      "grad_norm": 0.11950384825468063,
+      "learning_rate": 0.0001,
+      "loss": 1.6388,
+      "step": 466
+    },
+    {
+      "epoch": 0.15634415801807833,
+      "grad_norm": 0.13280175626277924,
+      "learning_rate": 0.0001,
+      "loss": 1.6097,
+      "step": 467
+    },
+    {
+      "epoch": 0.1566789420823569,
+      "grad_norm": 0.11717383563518524,
+      "learning_rate": 0.0001,
+      "loss": 1.6519,
+      "step": 468
+    },
+    {
+      "epoch": 0.15701372614663542,
+      "grad_norm": 0.12387187778949738,
+      "learning_rate": 0.0001,
+      "loss": 1.5661,
+      "step": 469
+    },
+    {
+      "epoch": 0.15734851021091395,
+      "grad_norm": 0.12535057961940765,
+      "learning_rate": 0.0001,
+      "loss": 1.601,
+      "step": 470
+    },
+    {
+      "epoch": 0.1576832942751925,
+      "grad_norm": 0.12057804316282272,
+      "learning_rate": 0.0001,
+      "loss": 1.6463,
+      "step": 471
+    },
+    {
+      "epoch": 0.15801807833947104,
+      "grad_norm": 0.1360681802034378,
+      "learning_rate": 0.0001,
+      "loss": 1.7093,
+      "step": 472
+    },
+    {
+      "epoch": 0.15835286240374957,
+      "grad_norm": 0.11986411362886429,
+      "learning_rate": 0.0001,
+      "loss": 1.5864,
+      "step": 473
+    },
+    {
+      "epoch": 0.15868764646802813,
+      "grad_norm": 0.11335694789886475,
+      "learning_rate": 0.0001,
+      "loss": 1.5495,
+      "step": 474
+    },
+    {
+      "epoch": 0.15902243053230666,
+      "grad_norm": 0.11684451997280121,
+      "learning_rate": 0.0001,
+      "loss": 1.5295,
+      "step": 475
+    },
+    {
+      "epoch": 0.1593572145965852,
+      "grad_norm": 0.12882184982299805,
+      "learning_rate": 0.0001,
+      "loss": 1.6903,
+      "step": 476
+    },
+    {
+      "epoch": 0.15969199866086375,
+      "grad_norm": 0.12175029516220093,
+      "learning_rate": 0.0001,
+      "loss": 1.6421,
+      "step": 477
+    },
+    {
+      "epoch": 0.16002678272514229,
+      "grad_norm": 0.1330244094133377,
+      "learning_rate": 0.0001,
+      "loss": 1.5691,
+      "step": 478
+    },
+    {
+      "epoch": 0.16036156678942082,
+      "grad_norm": 0.12204015254974365,
+      "learning_rate": 0.0001,
+      "loss": 1.6557,
+      "step": 479
+    },
+    {
+      "epoch": 0.16069635085369938,
+      "grad_norm": 0.1265457272529602,
+      "learning_rate": 0.0001,
+      "loss": 1.6319,
+      "step": 480
+    },
+    {
+      "epoch": 0.1610311349179779,
+      "grad_norm": 0.13419146835803986,
+      "learning_rate": 0.0001,
+      "loss": 1.5694,
+      "step": 481
+    },
+    {
+      "epoch": 0.16136591898225644,
+      "grad_norm": 0.12663477659225464,
+      "learning_rate": 0.0001,
+      "loss": 1.49,
+      "step": 482
+    },
+    {
+      "epoch": 0.16170070304653497,
+      "grad_norm": 0.1386338621377945,
+      "learning_rate": 0.0001,
+      "loss": 1.5774,
+      "step": 483
+    },
+    {
+      "epoch": 0.16203548711081353,
+      "grad_norm": 0.1266423612833023,
+      "learning_rate": 0.0001,
+      "loss": 1.568,
+      "step": 484
+    },
+    {
+      "epoch": 0.16237027117509206,
+      "grad_norm": 0.11795584112405777,
+      "learning_rate": 0.0001,
+      "loss": 1.5952,
+      "step": 485
+    },
+    {
+      "epoch": 0.1627050552393706,
+      "grad_norm": 0.13247069716453552,
+      "learning_rate": 0.0001,
+      "loss": 1.5486,
+      "step": 486
+    },
+    {
+      "epoch": 0.16303983930364915,
+      "grad_norm": 0.12367638945579529,
+      "learning_rate": 0.0001,
+      "loss": 1.6618,
+      "step": 487
+    },
+    {
+      "epoch": 0.16337462336792768,
+      "grad_norm": 0.11988285183906555,
+      "learning_rate": 0.0001,
+      "loss": 1.6338,
+      "step": 488
+    },
+    {
+      "epoch": 0.16370940743220622,
+      "grad_norm": 0.12422308325767517,
+      "learning_rate": 0.0001,
+      "loss": 1.5753,
+      "step": 489
+    },
+    {
+      "epoch": 0.16404419149648478,
+      "grad_norm": 0.12060552090406418,
+      "learning_rate": 0.0001,
+      "loss": 1.6158,
+      "step": 490
+    },
+    {
+      "epoch": 0.1643789755607633,
+      "grad_norm": 0.1219470277428627,
+      "learning_rate": 0.0001,
+      "loss": 1.5057,
+      "step": 491
+    },
+    {
+      "epoch": 0.16471375962504184,
+      "grad_norm": 0.12771841883659363,
+      "learning_rate": 0.0001,
+      "loss": 1.6627,
+      "step": 492
+    },
+    {
+      "epoch": 0.1650485436893204,
+      "grad_norm": 0.11713176220655441,
+      "learning_rate": 0.0001,
+      "loss": 1.5697,
+      "step": 493
+    },
+    {
+      "epoch": 0.16538332775359893,
+      "grad_norm": 0.1419348567724228,
+      "learning_rate": 0.0001,
+      "loss": 1.7253,
+      "step": 494
+    },
+    {
+      "epoch": 0.16571811181787746,
+      "grad_norm": 0.1297536939382553,
+      "learning_rate": 0.0001,
+      "loss": 1.666,
+      "step": 495
+    },
+    {
+      "epoch": 0.16605289588215602,
+      "grad_norm": 0.12997077405452728,
+      "learning_rate": 0.0001,
+      "loss": 1.5825,
+      "step": 496
+    },
+    {
+      "epoch": 0.16638767994643455,
+      "grad_norm": 0.14354097843170166,
+      "learning_rate": 0.0001,
+      "loss": 1.628,
+      "step": 497
+    },
+    {
+      "epoch": 0.16672246401071308,
+      "grad_norm": 0.12498887628316879,
+      "learning_rate": 0.0001,
+      "loss": 1.7003,
+      "step": 498
+    },
+    {
+      "epoch": 0.16705724807499164,
+      "grad_norm": 0.13219912350177765,
+      "learning_rate": 0.0001,
+      "loss": 1.6218,
+      "step": 499
+    },
+    {
+      "epoch": 0.16739203213927017,
+      "grad_norm": 0.13144424557685852,
+      "learning_rate": 0.0001,
+      "loss": 1.5874,
+      "step": 500
+    },
+    {
+      "epoch": 0.1677268162035487,
+      "grad_norm": 0.12147901952266693,
+      "learning_rate": 0.0001,
+      "loss": 1.6308,
+      "step": 501
+    },
+    {
+      "epoch": 0.16806160026782724,
+      "grad_norm": 0.13109005987644196,
+      "learning_rate": 0.0001,
+      "loss": 1.7168,
+      "step": 502
+    },
+    {
+      "epoch": 0.1683963843321058,
+      "grad_norm": 0.1306311935186386,
+      "learning_rate": 0.0001,
+      "loss": 1.5859,
+      "step": 503
+    },
+    {
+      "epoch": 0.16873116839638433,
+      "grad_norm": 0.115351103246212,
+      "learning_rate": 0.0001,
+      "loss": 1.6124,
+      "step": 504
+    },
+    {
+      "epoch": 0.16906595246066286,
+      "grad_norm": 0.12713004648685455,
+      "learning_rate": 0.0001,
+      "loss": 1.5558,
+      "step": 505
+    },
+    {
+      "epoch": 0.16940073652494142,
+      "grad_norm": 0.1304563283920288,
+      "learning_rate": 0.0001,
+      "loss": 1.676,
+      "step": 506
+    },
+    {
+      "epoch": 0.16973552058921995,
+      "grad_norm": 0.12284432351589203,
+      "learning_rate": 0.0001,
+      "loss": 1.5585,
+      "step": 507
+    },
+    {
+      "epoch": 0.17007030465349848,
+      "grad_norm": 0.12343181669712067,
+      "learning_rate": 0.0001,
+      "loss": 1.5869,
+      "step": 508
+    },
+    {
+      "epoch": 0.17040508871777704,
+      "grad_norm": 0.11459839344024658,
+      "learning_rate": 0.0001,
+      "loss": 1.6051,
+      "step": 509
+    },
+    {
+      "epoch": 0.17073987278205557,
+      "grad_norm": 0.11883780360221863,
+      "learning_rate": 0.0001,
+      "loss": 1.5064,
+      "step": 510
+    },
+    {
+      "epoch": 0.1710746568463341,
+      "grad_norm": 0.12307373434305191,
+      "learning_rate": 0.0001,
+      "loss": 1.5257,
+      "step": 511
+    },
+    {
+      "epoch": 0.17140944091061266,
+      "grad_norm": 0.11666516959667206,
+      "learning_rate": 0.0001,
+      "loss": 1.5842,
+      "step": 512
+    },
+    {
+      "epoch": 0.1717442249748912,
+      "grad_norm": 0.11493846029043198,
+      "learning_rate": 0.0001,
+      "loss": 1.6215,
+      "step": 513
+    },
+    {
+      "epoch": 0.17207900903916973,
+      "grad_norm": 0.1198093444108963,
+      "learning_rate": 0.0001,
+      "loss": 1.5875,
+      "step": 514
+    },
+    {
+      "epoch": 0.1724137931034483,
+      "grad_norm": 0.11997364461421967,
+      "learning_rate": 0.0001,
+      "loss": 1.5819,
+      "step": 515
+    },
+    {
+      "epoch": 0.17274857716772682,
+      "grad_norm": 0.12003917992115021,
+      "learning_rate": 0.0001,
+      "loss": 1.7019,
+      "step": 516
+    },
+    {
+      "epoch": 0.17308336123200535,
+      "grad_norm": 0.11761089414358139,
+      "learning_rate": 0.0001,
+      "loss": 1.5742,
+      "step": 517
+    },
+    {
+      "epoch": 0.1734181452962839,
+      "grad_norm": 0.12004124373197556,
+      "learning_rate": 0.0001,
+      "loss": 1.5947,
+      "step": 518
+    },
+    {
+      "epoch": 0.17375292936056244,
+      "grad_norm": 0.12139872461557388,
+      "learning_rate": 0.0001,
+      "loss": 1.4861,
+      "step": 519
+    },
+    {
+      "epoch": 0.17408771342484097,
+      "grad_norm": 0.12214326858520508,
+      "learning_rate": 0.0001,
+      "loss": 1.6953,
+      "step": 520
+    },
+    {
+      "epoch": 0.1744224974891195,
+      "grad_norm": 0.12239626795053482,
+      "learning_rate": 0.0001,
+      "loss": 1.5529,
+      "step": 521
+    },
+    {
+      "epoch": 0.17475728155339806,
+      "grad_norm": 0.11888886988162994,
+      "learning_rate": 0.0001,
+      "loss": 1.5099,
+      "step": 522
+    },
+    {
+      "epoch": 0.1750920656176766,
+      "grad_norm": 0.11585521697998047,
+      "learning_rate": 0.0001,
+      "loss": 1.5392,
+      "step": 523
+    },
+    {
+      "epoch": 0.17542684968195513,
+      "grad_norm": 0.1300823837518692,
+      "learning_rate": 0.0001,
+      "loss": 1.6598,
+      "step": 524
+    },
+    {
+      "epoch": 0.1757616337462337,
+      "grad_norm": 0.12741157412528992,
+      "learning_rate": 0.0001,
+      "loss": 1.5798,
+      "step": 525
+    },
+    {
+      "epoch": 0.17609641781051222,
+      "grad_norm": 0.11614137142896652,
+      "learning_rate": 0.0001,
+      "loss": 1.5343,
+      "step": 526
+    },
+    {
+      "epoch": 0.17643120187479075,
+      "grad_norm": 0.12221526354551315,
+      "learning_rate": 0.0001,
+      "loss": 1.552,
+      "step": 527
+    },
+    {
+      "epoch": 0.1767659859390693,
+      "grad_norm": 0.13221661746501923,
+      "learning_rate": 0.0001,
+      "loss": 1.6213,
+      "step": 528
+    },
+    {
+      "epoch": 0.17710077000334784,
+      "grad_norm": 0.12069322913885117,
+      "learning_rate": 0.0001,
+      "loss": 1.6148,
+      "step": 529
+    },
+    {
+      "epoch": 0.17743555406762637,
+      "grad_norm": 0.11254309117794037,
+      "learning_rate": 0.0001,
+      "loss": 1.5917,
+      "step": 530
+    },
+    {
+      "epoch": 0.17777033813190493,
+      "grad_norm": 0.11715224385261536,
+      "learning_rate": 0.0001,
+      "loss": 1.6343,
+      "step": 531
+    },
+    {
+      "epoch": 0.17810512219618346,
+      "grad_norm": 0.1183256059885025,
+      "learning_rate": 0.0001,
+      "loss": 1.4889,
+      "step": 532
+    },
+    {
+      "epoch": 0.178439906260462,
+      "grad_norm": 0.12182603031396866,
+      "learning_rate": 0.0001,
+      "loss": 1.5487,
+      "step": 533
+    },
+    {
+      "epoch": 0.17877469032474055,
+      "grad_norm": 0.1232253909111023,
+      "learning_rate": 0.0001,
+      "loss": 1.6754,
+      "step": 534
+    },
+    {
+      "epoch": 0.17910947438901909,
+      "grad_norm": 0.11796277016401291,
+      "learning_rate": 0.0001,
+      "loss": 1.6396,
+      "step": 535
+    },
+    {
+      "epoch": 0.17944425845329762,
+      "grad_norm": 0.13181637227535248,
+      "learning_rate": 0.0001,
+      "loss": 1.6505,
+      "step": 536
+    },
+    {
+      "epoch": 0.17977904251757618,
+      "grad_norm": 0.11481553316116333,
+      "learning_rate": 0.0001,
+      "loss": 1.492,
+      "step": 537
+    },
+    {
+      "epoch": 0.1801138265818547,
+      "grad_norm": 0.12842705845832825,
+      "learning_rate": 0.0001,
+      "loss": 1.734,
+      "step": 538
+    },
+    {
+      "epoch": 0.18044861064613324,
+      "grad_norm": 0.1235375851392746,
+      "learning_rate": 0.0001,
+      "loss": 1.6496,
+      "step": 539
+    },
+    {
+      "epoch": 0.18078339471041177,
+      "grad_norm": 0.12111697345972061,
+      "learning_rate": 0.0001,
+      "loss": 1.5044,
+      "step": 540
+    },
+    {
+      "epoch": 0.18111817877469033,
+      "grad_norm": 0.12484171241521835,
+      "learning_rate": 0.0001,
+      "loss": 1.6643,
+      "step": 541
+    },
+    {
+      "epoch": 0.18145296283896886,
+      "grad_norm": 0.12675760686397552,
+      "learning_rate": 0.0001,
+      "loss": 1.6188,
+      "step": 542
+    },
+    {
+      "epoch": 0.1817877469032474,
+      "grad_norm": 0.12203079462051392,
+      "learning_rate": 0.0001,
+      "loss": 1.507,
+      "step": 543
+    },
+    {
+      "epoch": 0.18212253096752595,
+      "grad_norm": 0.12013613432645798,
+      "learning_rate": 0.0001,
+      "loss": 1.6247,
+      "step": 544
+    },
+    {
+      "epoch": 0.18245731503180448,
+      "grad_norm": 0.12438444793224335,
+      "learning_rate": 0.0001,
+      "loss": 1.5849,
+      "step": 545
+    },
+    {
+      "epoch": 0.18279209909608302,
+      "grad_norm": 0.13607415556907654,
+      "learning_rate": 0.0001,
+      "loss": 1.6562,
+      "step": 546
+    },
+    {
+      "epoch": 0.18312688316036158,
+      "grad_norm": 0.1240532174706459,
+      "learning_rate": 0.0001,
+      "loss": 1.5205,
+      "step": 547
+    },
+    {
+      "epoch": 0.1834616672246401,
+      "grad_norm": 0.1510075032711029,
+      "learning_rate": 0.0001,
+      "loss": 1.6608,
+      "step": 548
+    },
+    {
+      "epoch": 0.18379645128891864,
+      "grad_norm": 0.11965179443359375,
+      "learning_rate": 0.0001,
+      "loss": 1.6391,
+      "step": 549
+    },
+    {
+      "epoch": 0.1841312353531972,
+      "grad_norm": 0.14874660968780518,
+      "learning_rate": 0.0001,
+      "loss": 1.6156,
+      "step": 550
+    },
+    {
+      "epoch": 0.18446601941747573,
+      "grad_norm": 0.1273370385169983,
+      "learning_rate": 0.0001,
+      "loss": 1.5117,
+      "step": 551
+    },
+    {
+      "epoch": 0.18480080348175426,
+      "grad_norm": 0.1213572546839714,
+      "learning_rate": 0.0001,
+      "loss": 1.5124,
+      "step": 552
+    },
+    {
+      "epoch": 0.18513558754603282,
+      "grad_norm": 0.1602640151977539,
+      "learning_rate": 0.0001,
+      "loss": 1.6318,
+      "step": 553
+    },
+    {
+      "epoch": 0.18547037161031135,
+      "grad_norm": 0.12859167158603668,
+      "learning_rate": 0.0001,
+      "loss": 1.6562,
+      "step": 554
+    },
+    {
+      "epoch": 0.18580515567458988,
+      "grad_norm": 0.13728216290473938,
+      "learning_rate": 0.0001,
+      "loss": 1.5873,
+      "step": 555
+    },
+    {
+      "epoch": 0.18613993973886844,
+      "grad_norm": 0.12880103290081024,
+      "learning_rate": 0.0001,
+      "loss": 1.5121,
+      "step": 556
+    },
+    {
+      "epoch": 0.18647472380314697,
+      "grad_norm": 0.1293378323316574,
+      "learning_rate": 0.0001,
+      "loss": 1.6275,
+      "step": 557
+    },
+    {
+      "epoch": 0.1868095078674255,
+      "grad_norm": 0.1387391984462738,
+      "learning_rate": 0.0001,
+      "loss": 1.6486,
+      "step": 558
+    },
+    {
+      "epoch": 0.18714429193170404,
+      "grad_norm": 0.14882785081863403,
+      "learning_rate": 0.0001,
+      "loss": 1.6422,
+      "step": 559
+    },
+    {
+      "epoch": 0.1874790759959826,
+      "grad_norm": 0.11521956324577332,
+      "learning_rate": 0.0001,
+      "loss": 1.5032,
+      "step": 560
+    },
+    {
+      "epoch": 0.18781386006026113,
+      "grad_norm": 0.12418463081121445,
+      "learning_rate": 0.0001,
+      "loss": 1.5422,
+      "step": 561
+    },
+    {
+      "epoch": 0.18814864412453966,
+      "grad_norm": 0.13123475015163422,
+      "learning_rate": 0.0001,
+      "loss": 1.6459,
+      "step": 562
+    },
+    {
+      "epoch": 0.18848342818881822,
+      "grad_norm": 0.12267505377531052,
+      "learning_rate": 0.0001,
+      "loss": 1.61,
+      "step": 563
+    },
+    {
+      "epoch": 0.18881821225309675,
+      "grad_norm": 0.12172992527484894,
+      "learning_rate": 0.0001,
+      "loss": 1.551,
+      "step": 564
+    },
+    {
+      "epoch": 0.18915299631737528,
+      "grad_norm": 0.12027712911367416,
+      "learning_rate": 0.0001,
+      "loss": 1.6178,
+      "step": 565
+    },
+    {
+      "epoch": 0.18948778038165384,
+      "grad_norm": 0.11598297208547592,
+      "learning_rate": 0.0001,
+      "loss": 1.5959,
+      "step": 566
+    },
+    {
+      "epoch": 0.18982256444593237,
+      "grad_norm": 0.11541326344013214,
+      "learning_rate": 0.0001,
+      "loss": 1.5936,
+      "step": 567
+    },
+    {
+      "epoch": 0.1901573485102109,
+      "grad_norm": 0.12343809008598328,
+      "learning_rate": 0.0001,
+      "loss": 1.6091,
+      "step": 568
+    },
+    {
+      "epoch": 0.19049213257448946,
+      "grad_norm": 0.11451027542352676,
+      "learning_rate": 0.0001,
+      "loss": 1.6203,
+      "step": 569
+    },
+    {
+      "epoch": 0.190826916638768,
+      "grad_norm": 0.1260651797056198,
+      "learning_rate": 0.0001,
+      "loss": 1.6105,
+      "step": 570
+    },
+    {
+      "epoch": 0.19116170070304653,
+      "grad_norm": 0.1183401346206665,
+      "learning_rate": 0.0001,
+      "loss": 1.583,
+      "step": 571
+    },
+    {
+      "epoch": 0.1914964847673251,
+      "grad_norm": 0.11767153441905975,
+      "learning_rate": 0.0001,
+      "loss": 1.5717,
+      "step": 572
+    },
+    {
+      "epoch": 0.19183126883160362,
+      "grad_norm": 0.11693871766328812,
+      "learning_rate": 0.0001,
+      "loss": 1.5783,
+      "step": 573
+    },
+    {
+      "epoch": 0.19216605289588215,
+      "grad_norm": 0.1267687827348709,
+      "learning_rate": 0.0001,
+      "loss": 1.5803,
+      "step": 574
+    },
+    {
+      "epoch": 0.1925008369601607,
+      "grad_norm": 0.11946652829647064,
+      "learning_rate": 0.0001,
+      "loss": 1.5575,
+      "step": 575
+    },
+    {
+      "epoch": 0.19283562102443924,
+      "grad_norm": 0.12602412700653076,
+      "learning_rate": 0.0001,
+      "loss": 1.7297,
+      "step": 576
+    },
+    {
+      "epoch": 0.19317040508871777,
+      "grad_norm": 0.12529441714286804,
+      "learning_rate": 0.0001,
+      "loss": 1.6877,
+      "step": 577
+    },
+    {
+      "epoch": 0.1935051891529963,
+      "grad_norm": 0.12578092515468597,
+      "learning_rate": 0.0001,
+      "loss": 1.5397,
+      "step": 578
+    },
+    {
+      "epoch": 0.19383997321727486,
+      "grad_norm": 0.12697197496891022,
+      "learning_rate": 0.0001,
+      "loss": 1.5541,
+      "step": 579
+    },
+    {
+      "epoch": 0.1941747572815534,
+      "grad_norm": 0.12927542626857758,
+      "learning_rate": 0.0001,
+      "loss": 1.6155,
+      "step": 580
+    },
+    {
+      "epoch": 0.19450954134583193,
+      "grad_norm": 0.1361040472984314,
+      "learning_rate": 0.0001,
+      "loss": 1.5857,
+      "step": 581
+    },
+    {
+      "epoch": 0.19484432541011049,
+      "grad_norm": 0.11877462267875671,
+      "learning_rate": 0.0001,
+      "loss": 1.5558,
+      "step": 582
+    },
+    {
+      "epoch": 0.19517910947438902,
+      "grad_norm": 0.14642973244190216,
+      "learning_rate": 0.0001,
+      "loss": 1.6476,
+      "step": 583
+    },
+    {
+      "epoch": 0.19551389353866755,
+      "grad_norm": 0.13428737223148346,
+      "learning_rate": 0.0001,
+      "loss": 1.5862,
+      "step": 584
+    },
+    {
+      "epoch": 0.1958486776029461,
+      "grad_norm": 0.1275390088558197,
+      "learning_rate": 0.0001,
+      "loss": 1.5418,
+      "step": 585
+    },
+    {
+      "epoch": 0.19618346166722464,
+      "grad_norm": 0.1398482322692871,
+      "learning_rate": 0.0001,
+      "loss": 1.4985,
+      "step": 586
+    },
+    {
+      "epoch": 0.19651824573150317,
+      "grad_norm": 0.12443619966506958,
+      "learning_rate": 0.0001,
+      "loss": 1.5726,
+      "step": 587
+    },
+    {
+      "epoch": 0.19685302979578173,
+      "grad_norm": 0.12923243641853333,
+      "learning_rate": 0.0001,
+      "loss": 1.5596,
+      "step": 588
+    },
+    {
+      "epoch": 0.19718781386006026,
+      "grad_norm": 0.14045698940753937,
+      "learning_rate": 0.0001,
+      "loss": 1.5475,
+      "step": 589
+    },
+    {
+      "epoch": 0.1975225979243388,
+      "grad_norm": 0.12687772512435913,
+      "learning_rate": 0.0001,
+      "loss": 1.7041,
+      "step": 590
+    },
+    {
+      "epoch": 0.19785738198861735,
+      "grad_norm": 0.14536388218402863,
+      "learning_rate": 0.0001,
+      "loss": 1.5724,
+      "step": 591
+    },
+    {
+      "epoch": 0.19819216605289589,
+      "grad_norm": 0.1331462413072586,
+      "learning_rate": 0.0001,
+      "loss": 1.6991,
+      "step": 592
+    },
+    {
+      "epoch": 0.19852695011717442,
+      "grad_norm": 0.13363464176654816,
+      "learning_rate": 0.0001,
+      "loss": 1.6665,
+      "step": 593
+    },
+    {
+      "epoch": 0.19886173418145298,
+      "grad_norm": 0.13291539251804352,
+      "learning_rate": 0.0001,
+      "loss": 1.6278,
+      "step": 594
+    },
+    {
+      "epoch": 0.1991965182457315,
+      "grad_norm": 0.1261158436536789,
+      "learning_rate": 0.0001,
+      "loss": 1.6129,
+      "step": 595
+    },
+    {
+      "epoch": 0.19953130231001004,
+      "grad_norm": 0.12324585020542145,
+      "learning_rate": 0.0001,
+      "loss": 1.6509,
+      "step": 596
+    },
+    {
+      "epoch": 0.19986608637428857,
+      "grad_norm": 0.11849376559257507,
+      "learning_rate": 0.0001,
+      "loss": 1.6226,
+      "step": 597
+    },
+    {
+      "epoch": 0.20020087043856713,
+      "grad_norm": 0.1167241707444191,
+      "learning_rate": 0.0001,
+      "loss": 1.5539,
+      "step": 598
+    },
+    {
+      "epoch": 0.20053565450284566,
+      "grad_norm": 0.11860879510641098,
+      "learning_rate": 0.0001,
+      "loss": 1.5962,
+      "step": 599
+    },
+    {
+      "epoch": 0.2008704385671242,
+      "grad_norm": 0.12385833263397217,
+      "learning_rate": 0.0001,
+      "loss": 1.593,
+      "step": 600
+    },
+    {
+      "epoch": 0.20120522263140275,
+      "grad_norm": 0.12093829363584518,
+      "learning_rate": 0.0001,
+      "loss": 1.6914,
+      "step": 601
+    },
+    {
+      "epoch": 0.20154000669568128,
+      "grad_norm": 0.11839880049228668,
+      "learning_rate": 0.0001,
+      "loss": 1.5645,
+      "step": 602
+    },
+    {
+      "epoch": 0.20187479075995982,
+      "grad_norm": 0.11958955973386765,
+      "learning_rate": 0.0001,
+      "loss": 1.6964,
+      "step": 603
+    },
+    {
+      "epoch": 0.20220957482423838,
+      "grad_norm": 0.12148015946149826,
+      "learning_rate": 0.0001,
+      "loss": 1.6201,
+      "step": 604
+    },
+    {
+      "epoch": 0.2025443588885169,
+      "grad_norm": 0.11879414319992065,
+      "learning_rate": 0.0001,
+      "loss": 1.5696,
+      "step": 605
+    },
+    {
+      "epoch": 0.20287914295279544,
+      "grad_norm": 0.11815709620714188,
+      "learning_rate": 0.0001,
+      "loss": 1.5771,
+      "step": 606
+    },
+    {
+      "epoch": 0.203213927017074,
+      "grad_norm": 0.12391653656959534,
+      "learning_rate": 0.0001,
+      "loss": 1.4984,
+      "step": 607
+    },
+    {
+      "epoch": 0.20354871108135253,
+      "grad_norm": 0.12949740886688232,
+      "learning_rate": 0.0001,
+      "loss": 1.6746,
+      "step": 608
+    },
+    {
+      "epoch": 0.20388349514563106,
+      "grad_norm": 0.12630179524421692,
+      "learning_rate": 0.0001,
+      "loss": 1.5984,
+      "step": 609
+    },
+    {
+      "epoch": 0.20421827920990962,
+      "grad_norm": 0.13836237788200378,
+      "learning_rate": 0.0001,
+      "loss": 1.6562,
+      "step": 610
+    },
+    {
+      "epoch": 0.20455306327418815,
+      "grad_norm": 0.12105460464954376,
+      "learning_rate": 0.0001,
+      "loss": 1.628,
+      "step": 611
+    },
+    {
+      "epoch": 0.20488784733846668,
+      "grad_norm": 0.13807529211044312,
+      "learning_rate": 0.0001,
+      "loss": 1.5858,
+      "step": 612
+    },
+    {
+      "epoch": 0.20522263140274524,
+      "grad_norm": 0.12660756707191467,
+      "learning_rate": 0.0001,
+      "loss": 1.5819,
+      "step": 613
+    },
+    {
+      "epoch": 0.20555741546702377,
+      "grad_norm": 0.11513250321149826,
+      "learning_rate": 0.0001,
+      "loss": 1.5572,
+      "step": 614
+    },
+    {
+      "epoch": 0.2058921995313023,
+      "grad_norm": 0.12499019503593445,
+      "learning_rate": 0.0001,
+      "loss": 1.5902,
+      "step": 615
+    },
+    {
+      "epoch": 0.20622698359558084,
+      "grad_norm": 0.13060630857944489,
+      "learning_rate": 0.0001,
+      "loss": 1.6933,
+      "step": 616
+    },
+    {
+      "epoch": 0.2065617676598594,
+      "grad_norm": 0.11751000583171844,
+      "learning_rate": 0.0001,
+      "loss": 1.6165,
+      "step": 617
+    },
+    {
+      "epoch": 0.20689655172413793,
+      "grad_norm": 0.12362553179264069,
+      "learning_rate": 0.0001,
+      "loss": 1.6214,
+      "step": 618
+    },
+    {
+      "epoch": 0.20723133578841646,
+      "grad_norm": 0.11933618783950806,
+      "learning_rate": 0.0001,
+      "loss": 1.6041,
+      "step": 619
+    },
+    {
+      "epoch": 0.20756611985269502,
+      "grad_norm": 0.12560446560382843,
+      "learning_rate": 0.0001,
+      "loss": 1.689,
+      "step": 620
+    },
+    {
+      "epoch": 0.20790090391697355,
+      "grad_norm": 0.12433163821697235,
+      "learning_rate": 0.0001,
+      "loss": 1.6717,
+      "step": 621
+    },
+    {
+      "epoch": 0.20823568798125208,
+      "grad_norm": 0.12220048159360886,
+      "learning_rate": 0.0001,
+      "loss": 1.6216,
+      "step": 622
+    },
+    {
+      "epoch": 0.20857047204553064,
+      "grad_norm": 0.11404889076948166,
+      "learning_rate": 0.0001,
+      "loss": 1.5362,
+      "step": 623
+    },
+    {
+      "epoch": 0.20890525610980917,
+      "grad_norm": 0.11990871280431747,
+      "learning_rate": 0.0001,
+      "loss": 1.5971,
+      "step": 624
+    },
+    {
+      "epoch": 0.2092400401740877,
+      "grad_norm": 0.11785005033016205,
+      "learning_rate": 0.0001,
+      "loss": 1.5641,
+      "step": 625
+    },
+    {
+      "epoch": 0.20957482423836626,
+      "grad_norm": 0.12312883138656616,
+      "learning_rate": 0.0001,
+      "loss": 1.617,
+      "step": 626
+    },
+    {
+      "epoch": 0.2099096083026448,
+      "grad_norm": 0.11449938267469406,
+      "learning_rate": 0.0001,
+      "loss": 1.5396,
+      "step": 627
+    },
+    {
+      "epoch": 0.21024439236692333,
+      "grad_norm": 0.1219322681427002,
+      "learning_rate": 0.0001,
+      "loss": 1.5951,
+      "step": 628
+    },
+    {
+      "epoch": 0.2105791764312019,
+      "grad_norm": 0.12152589112520218,
+      "learning_rate": 0.0001,
+      "loss": 1.6017,
+      "step": 629
+    },
+    {
+      "epoch": 0.21091396049548042,
+      "grad_norm": 0.11546038091182709,
+      "learning_rate": 0.0001,
+      "loss": 1.5969,
+      "step": 630
+    },
+    {
+      "epoch": 0.21124874455975895,
+      "grad_norm": 0.1294824779033661,
+      "learning_rate": 0.0001,
+      "loss": 1.5983,
+      "step": 631
+    },
+    {
+      "epoch": 0.2115835286240375,
+      "grad_norm": 0.12606552243232727,
+      "learning_rate": 0.0001,
+      "loss": 1.6026,
+      "step": 632
+    },
+    {
+      "epoch": 0.21191831268831604,
+      "grad_norm": 0.12761344015598297,
+      "learning_rate": 0.0001,
+      "loss": 1.6561,
+      "step": 633
+    },
+    {
+      "epoch": 0.21225309675259457,
+      "grad_norm": 0.11588580161333084,
+      "learning_rate": 0.0001,
+      "loss": 1.5967,
+      "step": 634
+    },
+    {
+      "epoch": 0.2125878808168731,
+      "grad_norm": 0.11629272252321243,
+      "learning_rate": 0.0001,
+      "loss": 1.5894,
+      "step": 635
+    },
+    {
+      "epoch": 0.21292266488115166,
+      "grad_norm": 0.1237213984131813,
+      "learning_rate": 0.0001,
+      "loss": 1.6113,
+      "step": 636
+    },
+    {
+      "epoch": 0.2132574489454302,
+      "grad_norm": 0.12293344736099243,
+      "learning_rate": 0.0001,
+      "loss": 1.5972,
+      "step": 637
+    },
+    {
+      "epoch": 0.21359223300970873,
+      "grad_norm": 0.1172887459397316,
+      "learning_rate": 0.0001,
+      "loss": 1.5765,
+      "step": 638
+    },
+    {
+      "epoch": 0.21392701707398729,
+      "grad_norm": 0.12403010576963425,
+      "learning_rate": 0.0001,
+      "loss": 1.5639,
+      "step": 639
+    },
+    {
+      "epoch": 0.21426180113826582,
+      "grad_norm": 0.12683235108852386,
+      "learning_rate": 0.0001,
+      "loss": 1.5197,
+      "step": 640
+    },
+    {
+      "epoch": 0.21459658520254435,
+      "grad_norm": 0.11593903601169586,
+      "learning_rate": 0.0001,
+      "loss": 1.5158,
+      "step": 641
+    },
+    {
+      "epoch": 0.2149313692668229,
+      "grad_norm": 0.1251828819513321,
+      "learning_rate": 0.0001,
+      "loss": 1.6396,
+      "step": 642
+    },
+    {
+      "epoch": 0.21526615333110144,
+      "grad_norm": 0.12358346581459045,
+      "learning_rate": 0.0001,
+      "loss": 1.6012,
+      "step": 643
+    },
+    {
+      "epoch": 0.21560093739537997,
+      "grad_norm": 0.11473721265792847,
+      "learning_rate": 0.0001,
+      "loss": 1.5365,
+      "step": 644
+    },
+    {
+      "epoch": 0.21593572145965853,
+      "grad_norm": 0.1184060201048851,
+      "learning_rate": 0.0001,
+      "loss": 1.4507,
+      "step": 645
+    },
+    {
+      "epoch": 0.21627050552393706,
+      "grad_norm": 0.12540043890476227,
+      "learning_rate": 0.0001,
+      "loss": 1.5854,
+      "step": 646
+    },
+    {
+      "epoch": 0.2166052895882156,
+      "grad_norm": 0.12070447206497192,
+      "learning_rate": 0.0001,
+      "loss": 1.6097,
+      "step": 647
+    },
+    {
+      "epoch": 0.21694007365249415,
+      "grad_norm": 0.11351459473371506,
+      "learning_rate": 0.0001,
+      "loss": 1.5937,
+      "step": 648
+    },
+    {
+      "epoch": 0.21727485771677268,
+      "grad_norm": 0.1242094486951828,
+      "learning_rate": 0.0001,
+      "loss": 1.5222,
+      "step": 649
+    },
+    {
+      "epoch": 0.21760964178105122,
+      "grad_norm": 0.12054958194494247,
+      "learning_rate": 0.0001,
+      "loss": 1.5285,
+      "step": 650
+    },
+    {
+      "epoch": 0.21794442584532978,
+      "grad_norm": 0.12539923191070557,
+      "learning_rate": 0.0001,
+      "loss": 1.5001,
+      "step": 651
+    },
+    {
+      "epoch": 0.2182792099096083,
+      "grad_norm": 0.12270530313253403,
+      "learning_rate": 0.0001,
+      "loss": 1.6769,
+      "step": 652
+    },
+    {
+      "epoch": 0.21861399397388684,
+      "grad_norm": 0.12920905649662018,
+      "learning_rate": 0.0001,
+      "loss": 1.5999,
+      "step": 653
+    },
+    {
+      "epoch": 0.21894877803816537,
+      "grad_norm": 0.13267312943935394,
+      "learning_rate": 0.0001,
+      "loss": 1.5382,
+      "step": 654
+    },
+    {
+      "epoch": 0.21928356210244393,
+      "grad_norm": 0.11984428763389587,
+      "learning_rate": 0.0001,
+      "loss": 1.631,
+      "step": 655
+    },
+    {
+      "epoch": 0.21961834616672246,
+      "grad_norm": 0.1474982053041458,
+      "learning_rate": 0.0001,
+      "loss": 1.6709,
+      "step": 656
+    },
+    {
+      "epoch": 0.219953130231001,
+      "grad_norm": 0.13542193174362183,
+      "learning_rate": 0.0001,
+      "loss": 1.6415,
+      "step": 657
+    },
+    {
+      "epoch": 0.22028791429527955,
+      "grad_norm": 0.13832658529281616,
+      "learning_rate": 0.0001,
+      "loss": 1.6118,
+      "step": 658
+    },
+    {
+      "epoch": 0.22062269835955808,
+      "grad_norm": 0.15140588581562042,
+      "learning_rate": 0.0001,
+      "loss": 1.6314,
+      "step": 659
+    },
+    {
+      "epoch": 0.22095748242383662,
+      "grad_norm": 0.12110920995473862,
+      "learning_rate": 0.0001,
+      "loss": 1.5884,
+      "step": 660
+    },
+    {
+      "epoch": 0.22129226648811517,
+      "grad_norm": 0.14811581373214722,
+      "learning_rate": 0.0001,
+      "loss": 1.6642,
+      "step": 661
+    },
+    {
+      "epoch": 0.2216270505523937,
+      "grad_norm": 0.12733857333660126,
+      "learning_rate": 0.0001,
+      "loss": 1.5512,
+      "step": 662
+    },
+    {
+      "epoch": 0.22196183461667224,
+      "grad_norm": 0.13028332591056824,
+      "learning_rate": 0.0001,
+      "loss": 1.5613,
+      "step": 663
+    },
+    {
+      "epoch": 0.2222966186809508,
+      "grad_norm": 0.1242808997631073,
+      "learning_rate": 0.0001,
+      "loss": 1.5869,
+      "step": 664
+    },
+    {
+      "epoch": 0.22263140274522933,
+      "grad_norm": 0.12380847334861755,
+      "learning_rate": 0.0001,
+      "loss": 1.5926,
+      "step": 665
+    },
+    {
+      "epoch": 0.22296618680950786,
+      "grad_norm": 0.12564754486083984,
+      "learning_rate": 0.0001,
+      "loss": 1.5811,
+      "step": 666
+    },
+    {
+      "epoch": 0.22330097087378642,
+      "grad_norm": 0.1509399712085724,
+      "learning_rate": 0.0001,
+      "loss": 1.7172,
+      "step": 667
+    },
+    {
+      "epoch": 0.22363575493806495,
+      "grad_norm": 0.12397512793540955,
+      "learning_rate": 0.0001,
+      "loss": 1.642,
+      "step": 668
+    },
+    {
+      "epoch": 0.22397053900234348,
+      "grad_norm": 0.13826021552085876,
+      "learning_rate": 0.0001,
+      "loss": 1.6395,
+      "step": 669
+    },
+    {
+      "epoch": 0.22430532306662204,
+      "grad_norm": 0.1417902112007141,
+      "learning_rate": 0.0001,
+      "loss": 1.6169,
+      "step": 670
+    },
+    {
+      "epoch": 0.22464010713090057,
+      "grad_norm": 0.12220132350921631,
+      "learning_rate": 0.0001,
+      "loss": 1.5686,
+      "step": 671
+    },
+    {
+      "epoch": 0.2249748911951791,
+      "grad_norm": 0.13563144207000732,
+      "learning_rate": 0.0001,
+      "loss": 1.6556,
+      "step": 672
+    },
+    {
+      "epoch": 0.22530967525945764,
+      "grad_norm": 0.13794521987438202,
+      "learning_rate": 0.0001,
+      "loss": 1.5187,
+      "step": 673
+    },
+    {
+      "epoch": 0.2256444593237362,
+      "grad_norm": 0.12060145288705826,
+      "learning_rate": 0.0001,
+      "loss": 1.5901,
+      "step": 674
+    },
+    {
+      "epoch": 0.22597924338801473,
+      "grad_norm": 0.13909369707107544,
+      "learning_rate": 0.0001,
+      "loss": 1.5101,
+      "step": 675
+    },
+    {
+      "epoch": 0.22631402745229326,
+      "grad_norm": 0.13746792078018188,
+      "learning_rate": 0.0001,
+      "loss": 1.6084,
+      "step": 676
+    },
+    {
+      "epoch": 0.22664881151657182,
+      "grad_norm": 0.11612525582313538,
+      "learning_rate": 0.0001,
+      "loss": 1.606,
+      "step": 677
+    },
+    {
+      "epoch": 0.22698359558085035,
+      "grad_norm": 0.13988125324249268,
+      "learning_rate": 0.0001,
+      "loss": 1.6123,
+      "step": 678
+    },
+    {
+      "epoch": 0.22731837964512888,
+      "grad_norm": 0.13023462891578674,
+      "learning_rate": 0.0001,
+      "loss": 1.6202,
+      "step": 679
+    },
+    {
+      "epoch": 0.22765316370940744,
+      "grad_norm": 0.11764882504940033,
+      "learning_rate": 0.0001,
+      "loss": 1.5744,
+      "step": 680
+    },
+    {
+      "epoch": 0.22798794777368597,
+      "grad_norm": 0.12987253069877625,
+      "learning_rate": 0.0001,
+      "loss": 1.6287,
+      "step": 681
+    },
+    {
+      "epoch": 0.2283227318379645,
+      "grad_norm": 0.12687528133392334,
+      "learning_rate": 0.0001,
+      "loss": 1.6177,
+      "step": 682
+    },
+    {
+      "epoch": 0.22865751590224306,
+      "grad_norm": 0.117088183760643,
+      "learning_rate": 0.0001,
+      "loss": 1.5704,
+      "step": 683
+    },
+    {
+      "epoch": 0.2289922999665216,
+      "grad_norm": 0.13380305469036102,
+      "learning_rate": 0.0001,
+      "loss": 1.5013,
+      "step": 684
+    },
+    {
+      "epoch": 0.22932708403080013,
+      "grad_norm": 0.13155803084373474,
+      "learning_rate": 0.0001,
+      "loss": 1.6627,
+      "step": 685
+    },
+    {
+      "epoch": 0.2296618680950787,
+      "grad_norm": 0.12210634350776672,
+      "learning_rate": 0.0001,
+      "loss": 1.491,
+      "step": 686
+    },
+    {
+      "epoch": 0.22999665215935722,
+      "grad_norm": 0.12427474558353424,
+      "learning_rate": 0.0001,
+      "loss": 1.6381,
+      "step": 687
+    },
+    {
+      "epoch": 0.23033143622363575,
+      "grad_norm": 0.12354297190904617,
+      "learning_rate": 0.0001,
+      "loss": 1.5804,
+      "step": 688
+    },
+    {
+      "epoch": 0.2306662202879143,
+      "grad_norm": 0.11402271687984467,
+      "learning_rate": 0.0001,
+      "loss": 1.5562,
+      "step": 689
+    },
+    {
+      "epoch": 0.23100100435219284,
+      "grad_norm": 0.12571346759796143,
+      "learning_rate": 0.0001,
+      "loss": 1.6974,
+      "step": 690
+    },
+    {
+      "epoch": 0.23133578841647137,
+      "grad_norm": 0.12201119214296341,
+      "learning_rate": 0.0001,
+      "loss": 1.5866,
+      "step": 691
+    },
+    {
+      "epoch": 0.2316705724807499,
+      "grad_norm": 0.13017117977142334,
+      "learning_rate": 0.0001,
+      "loss": 1.6493,
+      "step": 692
+    },
+    {
+      "epoch": 0.23200535654502846,
+      "grad_norm": 0.11595404893159866,
+      "learning_rate": 0.0001,
+      "loss": 1.5047,
+      "step": 693
+    },
+    {
+      "epoch": 0.232340140609307,
+      "grad_norm": 0.11953503638505936,
+      "learning_rate": 0.0001,
+      "loss": 1.4952,
+      "step": 694
+    },
+    {
+      "epoch": 0.23267492467358553,
+      "grad_norm": 0.11844140291213989,
+      "learning_rate": 0.0001,
+      "loss": 1.6223,
+      "step": 695
+    },
+    {
+      "epoch": 0.23300970873786409,
+      "grad_norm": 0.12358598411083221,
+      "learning_rate": 0.0001,
+      "loss": 1.6303,
+      "step": 696
+    },
+    {
+      "epoch": 0.23334449280214262,
+      "grad_norm": 0.12384648621082306,
+      "learning_rate": 0.0001,
+      "loss": 1.6594,
+      "step": 697
+    },
+    {
+      "epoch": 0.23367927686642115,
+      "grad_norm": 0.11835581809282303,
+      "learning_rate": 0.0001,
+      "loss": 1.6098,
+      "step": 698
+    },
+    {
+      "epoch": 0.2340140609306997,
+      "grad_norm": 0.1138228103518486,
+      "learning_rate": 0.0001,
+      "loss": 1.4118,
+      "step": 699
+    },
+    {
+      "epoch": 0.23434884499497824,
+      "grad_norm": 0.11459102481603622,
+      "learning_rate": 0.0001,
+      "loss": 1.5633,
+      "step": 700
+    },
+    {
+      "epoch": 0.23468362905925677,
+      "grad_norm": 0.11587528139352798,
+      "learning_rate": 0.0001,
+      "loss": 1.6973,
+      "step": 701
+    },
+    {
+      "epoch": 0.23501841312353533,
+      "grad_norm": 0.13280251622200012,
+      "learning_rate": 0.0001,
+      "loss": 1.5161,
+      "step": 702
+    },
+    {
+      "epoch": 0.23535319718781386,
+      "grad_norm": 0.12264399230480194,
+      "learning_rate": 0.0001,
+      "loss": 1.656,
+      "step": 703
+    },
+    {
+      "epoch": 0.2356879812520924,
+      "grad_norm": 0.11608457565307617,
+      "learning_rate": 0.0001,
+      "loss": 1.5099,
+      "step": 704
+    },
+    {
+      "epoch": 0.23602276531637095,
+      "grad_norm": 0.12152610719203949,
+      "learning_rate": 0.0001,
+      "loss": 1.5169,
+      "step": 705
+    },
+    {
+      "epoch": 0.23635754938064948,
+      "grad_norm": 0.12914855778217316,
+      "learning_rate": 0.0001,
+      "loss": 1.5904,
+      "step": 706
+    },
+    {
+      "epoch": 0.23669233344492802,
+      "grad_norm": 0.13277898728847504,
+      "learning_rate": 0.0001,
+      "loss": 1.656,
+      "step": 707
+    },
+    {
+      "epoch": 0.23702711750920658,
+      "grad_norm": 0.12540487945079803,
+      "learning_rate": 0.0001,
+      "loss": 1.6178,
+      "step": 708
+    },
+    {
+      "epoch": 0.2373619015734851,
+      "grad_norm": 0.11845158785581589,
+      "learning_rate": 0.0001,
+      "loss": 1.5014,
+      "step": 709
+    },
+    {
+      "epoch": 0.23769668563776364,
+      "grad_norm": 0.11418534815311432,
+      "learning_rate": 0.0001,
+      "loss": 1.5292,
+      "step": 710
+    },
+    {
+      "epoch": 0.23803146970204217,
+      "grad_norm": 0.1384686678647995,
+      "learning_rate": 0.0001,
+      "loss": 1.6188,
+      "step": 711
+    },
+    {
+      "epoch": 0.23836625376632073,
+      "grad_norm": 0.12325987964868546,
+      "learning_rate": 0.0001,
+      "loss": 1.5636,
+      "step": 712
+    },
+    {
+      "epoch": 0.23870103783059926,
+      "grad_norm": 0.11931071430444717,
+      "learning_rate": 0.0001,
+      "loss": 1.5655,
+      "step": 713
+    },
+    {
+      "epoch": 0.2390358218948778,
+      "grad_norm": 0.12119931727647781,
+      "learning_rate": 0.0001,
+      "loss": 1.5289,
+      "step": 714
+    },
+    {
+      "epoch": 0.23937060595915635,
+      "grad_norm": 0.12172186374664307,
+      "learning_rate": 0.0001,
+      "loss": 1.6467,
+      "step": 715
+    },
+    {
+      "epoch": 0.23970539002343488,
+      "grad_norm": 0.12344299256801605,
+      "learning_rate": 0.0001,
+      "loss": 1.5616,
+      "step": 716
+    },
+    {
+      "epoch": 0.24004017408771341,
+      "grad_norm": 0.12173335254192352,
+      "learning_rate": 0.0001,
+      "loss": 1.6135,
+      "step": 717
+    },
+    {
+      "epoch": 0.24037495815199197,
+      "grad_norm": 0.1223810538649559,
+      "learning_rate": 0.0001,
+      "loss": 1.6239,
+      "step": 718
+    },
+    {
+      "epoch": 0.2407097422162705,
+      "grad_norm": 0.11744136363267899,
+      "learning_rate": 0.0001,
+      "loss": 1.5704,
+      "step": 719
+    },
+    {
+      "epoch": 0.24104452628054904,
+      "grad_norm": 0.12341196089982986,
+      "learning_rate": 0.0001,
+      "loss": 1.6704,
+      "step": 720
+    },
+    {
+      "epoch": 0.2413793103448276,
+      "grad_norm": 0.12578146159648895,
+      "learning_rate": 0.0001,
+      "loss": 1.604,
+      "step": 721
+    },
+    {
+      "epoch": 0.24171409440910613,
+      "grad_norm": 0.12708286941051483,
+      "learning_rate": 0.0001,
+      "loss": 1.5583,
+      "step": 722
+    },
+    {
+      "epoch": 0.24204887847338466,
+      "grad_norm": 0.11757246404886246,
+      "learning_rate": 0.0001,
+      "loss": 1.4911,
+      "step": 723
+    },
+    {
+      "epoch": 0.24238366253766322,
+      "grad_norm": 0.1309349238872528,
+      "learning_rate": 0.0001,
+      "loss": 1.6648,
+      "step": 724
+    },
+    {
+      "epoch": 0.24271844660194175,
+      "grad_norm": 0.13289286196231842,
+      "learning_rate": 0.0001,
+      "loss": 1.6547,
+      "step": 725
+    },
+    {
+      "epoch": 0.24305323066622028,
+      "grad_norm": 0.12044942378997803,
+      "learning_rate": 0.0001,
+      "loss": 1.661,
+      "step": 726
+    },
+    {
+      "epoch": 0.24338801473049884,
+      "grad_norm": 0.12810328602790833,
+      "learning_rate": 0.0001,
+      "loss": 1.6775,
+      "step": 727
+    },
+    {
+      "epoch": 0.24372279879477737,
+      "grad_norm": 0.12643273174762726,
+      "learning_rate": 0.0001,
+      "loss": 1.4938,
+      "step": 728
+    },
+    {
+      "epoch": 0.2440575828590559,
+      "grad_norm": 0.1253504455089569,
+      "learning_rate": 0.0001,
+      "loss": 1.6482,
+      "step": 729
+    },
+    {
+      "epoch": 0.24439236692333444,
+      "grad_norm": 0.12725912034511566,
+      "learning_rate": 0.0001,
+      "loss": 1.4911,
+      "step": 730
+    },
+    {
+      "epoch": 0.244727150987613,
+      "grad_norm": 0.13506008684635162,
+      "learning_rate": 0.0001,
+      "loss": 1.5739,
+      "step": 731
+    },
+    {
+      "epoch": 0.24506193505189153,
+      "grad_norm": 0.12034797668457031,
+      "learning_rate": 0.0001,
+      "loss": 1.6477,
+      "step": 732
+    },
+    {
+      "epoch": 0.24539671911617006,
+      "grad_norm": 0.12169791758060455,
+      "learning_rate": 0.0001,
+      "loss": 1.6398,
+      "step": 733
+    },
+    {
+      "epoch": 0.24573150318044862,
+      "grad_norm": 0.1253383755683899,
+      "learning_rate": 0.0001,
+      "loss": 1.5921,
+      "step": 734
+    },
+    {
+      "epoch": 0.24606628724472715,
+      "grad_norm": 0.11854001134634018,
+      "learning_rate": 0.0001,
+      "loss": 1.598,
+      "step": 735
+    },
+    {
+      "epoch": 0.24640107130900568,
+      "grad_norm": 0.13825742900371552,
+      "learning_rate": 0.0001,
+      "loss": 1.6588,
+      "step": 736
+    },
+    {
+      "epoch": 0.24673585537328424,
+      "grad_norm": 0.1235450729727745,
+      "learning_rate": 0.0001,
+      "loss": 1.5872,
+      "step": 737
+    },
+    {
+      "epoch": 0.24707063943756277,
+      "grad_norm": 0.12598398327827454,
+      "learning_rate": 0.0001,
+      "loss": 1.6038,
+      "step": 738
+    },
+    {
+      "epoch": 0.2474054235018413,
+      "grad_norm": 0.14527225494384766,
+      "learning_rate": 0.0001,
+      "loss": 1.6419,
+      "step": 739
+    },
+    {
+      "epoch": 0.24774020756611986,
+      "grad_norm": 0.11842803657054901,
+      "learning_rate": 0.0001,
+      "loss": 1.5628,
+      "step": 740
+    },
+    {
+      "epoch": 0.2480749916303984,
+      "grad_norm": 0.12376052141189575,
+      "learning_rate": 0.0001,
+      "loss": 1.5271,
+      "step": 741
+    },
+    {
+      "epoch": 0.24840977569467693,
+      "grad_norm": 0.13634417951107025,
+      "learning_rate": 0.0001,
+      "loss": 1.7012,
+      "step": 742
+    },
+    {
+      "epoch": 0.24874455975895549,
+      "grad_norm": 0.12457748502492905,
+      "learning_rate": 0.0001,
+      "loss": 1.5623,
+      "step": 743
+    },
+    {
+      "epoch": 0.24907934382323402,
+      "grad_norm": 0.11860496550798416,
+      "learning_rate": 0.0001,
+      "loss": 1.6049,
+      "step": 744
+    },
+    {
+      "epoch": 0.24941412788751255,
+      "grad_norm": 0.12447136640548706,
+      "learning_rate": 0.0001,
+      "loss": 1.6967,
+      "step": 745
+    },
+    {
+      "epoch": 0.2497489119517911,
+      "grad_norm": 0.12220341712236404,
+      "learning_rate": 0.0001,
+      "loss": 1.5819,
+      "step": 746
+    },
+    {
+      "epoch": 0.2500836960160696,
+      "grad_norm": 0.11865612119436264,
+      "learning_rate": 0.0001,
+      "loss": 1.5519,
+      "step": 747
+    },
+    {
+      "epoch": 0.25041848008034817,
+      "grad_norm": 0.11847954988479614,
+      "learning_rate": 0.0001,
+      "loss": 1.5087,
+      "step": 748
+    },
+    {
+      "epoch": 0.25075326414462673,
+      "grad_norm": 0.12107084691524506,
+      "learning_rate": 0.0001,
+      "loss": 1.5995,
+      "step": 749
+    },
+    {
+      "epoch": 0.25108804820890523,
+      "grad_norm": 0.12188322097063065,
+      "learning_rate": 0.0001,
+      "loss": 1.6439,
+      "step": 750
+    },
+    {
+      "epoch": 0.2514228322731838,
+      "grad_norm": 0.12144109606742859,
+      "learning_rate": 0.0001,
+      "loss": 1.5613,
+      "step": 751
+    },
+    {
+      "epoch": 0.25175761633746235,
+      "grad_norm": 0.12133816629648209,
+      "learning_rate": 0.0001,
+      "loss": 1.5364,
+      "step": 752
+    },
+    {
+      "epoch": 0.25209240040174086,
+      "grad_norm": 0.11708073318004608,
+      "learning_rate": 0.0001,
+      "loss": 1.5221,
+      "step": 753
+    },
+    {
+      "epoch": 0.2524271844660194,
+      "grad_norm": 0.1203671544790268,
+      "learning_rate": 0.0001,
+      "loss": 1.5736,
+      "step": 754
+    },
+    {
+      "epoch": 0.252761968530298,
+      "grad_norm": 0.12079092115163803,
+      "learning_rate": 0.0001,
+      "loss": 1.5842,
+      "step": 755
+    },
+    {
+      "epoch": 0.2530967525945765,
+      "grad_norm": 0.1294735223054886,
+      "learning_rate": 0.0001,
+      "loss": 1.5994,
+      "step": 756
+    },
+    {
+      "epoch": 0.25343153665885504,
+      "grad_norm": 0.1251528263092041,
+      "learning_rate": 0.0001,
+      "loss": 1.6391,
+      "step": 757
+    },
+    {
+      "epoch": 0.2537663207231336,
+      "grad_norm": 0.12093610316514969,
+      "learning_rate": 0.0001,
+      "loss": 1.6275,
+      "step": 758
+    },
+    {
+      "epoch": 0.2541011047874121,
+      "grad_norm": 0.1214980036020279,
+      "learning_rate": 0.0001,
+      "loss": 1.5887,
+      "step": 759
+    },
+    {
+      "epoch": 0.25443588885169066,
+      "grad_norm": 0.12011279165744781,
+      "learning_rate": 0.0001,
+      "loss": 1.5973,
+      "step": 760
+    },
+    {
+      "epoch": 0.2547706729159692,
+      "grad_norm": 0.12630945444107056,
+      "learning_rate": 0.0001,
+      "loss": 1.6184,
+      "step": 761
+    },
+    {
+      "epoch": 0.2551054569802477,
+      "grad_norm": 0.12001120299100876,
+      "learning_rate": 0.0001,
+      "loss": 1.5298,
+      "step": 762
+    },
+    {
+      "epoch": 0.2554402410445263,
+      "grad_norm": 0.1369365155696869,
+      "learning_rate": 0.0001,
+      "loss": 1.5718,
+      "step": 763
+    },
+    {
+      "epoch": 0.25577502510880484,
+      "grad_norm": 0.1201329231262207,
+      "learning_rate": 0.0001,
+      "loss": 1.5354,
+      "step": 764
+    },
+    {
+      "epoch": 0.25610980917308335,
+      "grad_norm": 0.12741532921791077,
+      "learning_rate": 0.0001,
+      "loss": 1.6193,
+      "step": 765
+    },
+    {
+      "epoch": 0.2564445932373619,
+      "grad_norm": 0.12349703162908554,
+      "learning_rate": 0.0001,
+      "loss": 1.6143,
+      "step": 766
+    },
+    {
+      "epoch": 0.25677937730164047,
+      "grad_norm": 0.11855439841747284,
+      "learning_rate": 0.0001,
+      "loss": 1.6037,
+      "step": 767
+    },
+    {
+      "epoch": 0.25711416136591897,
+      "grad_norm": 0.12034845352172852,
+      "learning_rate": 0.0001,
+      "loss": 1.5317,
+      "step": 768
+    },
+    {
+      "epoch": 0.25744894543019753,
+      "grad_norm": 0.11987943202257156,
+      "learning_rate": 0.0001,
+      "loss": 1.535,
+      "step": 769
+    },
+    {
+      "epoch": 0.2577837294944761,
+      "grad_norm": 0.12118515372276306,
+      "learning_rate": 0.0001,
+      "loss": 1.5974,
+      "step": 770
+    },
+    {
+      "epoch": 0.2581185135587546,
+      "grad_norm": 0.12842996418476105,
+      "learning_rate": 0.0001,
+      "loss": 1.609,
+      "step": 771
+    },
+    {
+      "epoch": 0.25845329762303315,
+      "grad_norm": 0.12420446425676346,
+      "learning_rate": 0.0001,
+      "loss": 1.6093,
+      "step": 772
+    },
+    {
+      "epoch": 0.2587880816873117,
+      "grad_norm": 0.12443120032548904,
+      "learning_rate": 0.0001,
+      "loss": 1.6122,
+      "step": 773
+    },
+    {
+      "epoch": 0.2591228657515902,
+      "grad_norm": 0.11912049353122711,
+      "learning_rate": 0.0001,
+      "loss": 1.5209,
+      "step": 774
+    },
+    {
+      "epoch": 0.2594576498158688,
+      "grad_norm": 0.1273064911365509,
+      "learning_rate": 0.0001,
+      "loss": 1.608,
+      "step": 775
+    },
+    {
+      "epoch": 0.2597924338801473,
+      "grad_norm": 0.11585114896297455,
+      "learning_rate": 0.0001,
+      "loss": 1.3888,
+      "step": 776
+    },
+    {
+      "epoch": 0.26012721794442584,
+      "grad_norm": 0.12005290389060974,
+      "learning_rate": 0.0001,
+      "loss": 1.4666,
+      "step": 777
+    },
+    {
+      "epoch": 0.2604620020087044,
+      "grad_norm": 0.11954595148563385,
+      "learning_rate": 0.0001,
+      "loss": 1.5558,
+      "step": 778
+    },
+    {
+      "epoch": 0.2607967860729829,
+      "grad_norm": 0.1307271122932434,
+      "learning_rate": 0.0001,
+      "loss": 1.6063,
+      "step": 779
+    },
+    {
+      "epoch": 0.26113157013726146,
+      "grad_norm": 0.113981693983078,
+      "learning_rate": 0.0001,
+      "loss": 1.4857,
+      "step": 780
+    },
+    {
+      "epoch": 0.26146635420154,
+      "grad_norm": 0.1225418671965599,
+      "learning_rate": 0.0001,
+      "loss": 1.5508,
+      "step": 781
+    },
+    {
+      "epoch": 0.2618011382658185,
+      "grad_norm": 0.12919741868972778,
+      "learning_rate": 0.0001,
+      "loss": 1.6255,
+      "step": 782
+    },
+    {
+      "epoch": 0.2621359223300971,
+      "grad_norm": 0.11552941054105759,
+      "learning_rate": 0.0001,
+      "loss": 1.6183,
+      "step": 783
+    },
+    {
+      "epoch": 0.26247070639437564,
+      "grad_norm": 0.13457614183425903,
+      "learning_rate": 0.0001,
+      "loss": 1.6461,
+      "step": 784
+    },
+    {
+      "epoch": 0.26280549045865415,
+      "grad_norm": 0.11841408908367157,
+      "learning_rate": 0.0001,
+      "loss": 1.5481,
+      "step": 785
+    },
+    {
+      "epoch": 0.2631402745229327,
+      "grad_norm": 0.11701938509941101,
+      "learning_rate": 0.0001,
+      "loss": 1.5883,
+      "step": 786
+    },
+    {
+      "epoch": 0.26347505858721126,
+      "grad_norm": 0.14221838116645813,
+      "learning_rate": 0.0001,
+      "loss": 1.5904,
+      "step": 787
+    },
+    {
+      "epoch": 0.26380984265148977,
+      "grad_norm": 0.11813905090093613,
+      "learning_rate": 0.0001,
+      "loss": 1.5653,
+      "step": 788
+    },
+    {
+      "epoch": 0.2641446267157683,
+      "grad_norm": 0.1315639317035675,
+      "learning_rate": 0.0001,
+      "loss": 1.5811,
+      "step": 789
+    },
+    {
+      "epoch": 0.2644794107800469,
+      "grad_norm": 0.13400433957576752,
+      "learning_rate": 0.0001,
+      "loss": 1.5363,
+      "step": 790
+    },
+    {
+      "epoch": 0.2648141948443254,
+      "grad_norm": 0.12116281688213348,
+      "learning_rate": 0.0001,
+      "loss": 1.6353,
+      "step": 791
+    },
+    {
+      "epoch": 0.26514897890860395,
+      "grad_norm": 0.1382567137479782,
+      "learning_rate": 0.0001,
+      "loss": 1.592,
+      "step": 792
+    },
+    {
+      "epoch": 0.2654837629728825,
+      "grad_norm": 0.14005912840366364,
+      "learning_rate": 0.0001,
+      "loss": 1.6114,
+      "step": 793
+    },
+    {
+      "epoch": 0.265818547037161,
+      "grad_norm": 0.13382911682128906,
+      "learning_rate": 0.0001,
+      "loss": 1.5942,
+      "step": 794
+    },
+    {
+      "epoch": 0.26615333110143957,
+      "grad_norm": 0.12423510104417801,
+      "learning_rate": 0.0001,
+      "loss": 1.5378,
+      "step": 795
+    },
+    {
+      "epoch": 0.26648811516571813,
+      "grad_norm": 0.12228628993034363,
+      "learning_rate": 0.0001,
+      "loss": 1.5704,
+      "step": 796
+    },
+    {
+      "epoch": 0.26682289922999664,
+      "grad_norm": 0.1286916881799698,
+      "learning_rate": 0.0001,
+      "loss": 1.6037,
+      "step": 797
+    },
+    {
+      "epoch": 0.2671576832942752,
+      "grad_norm": 0.12864018976688385,
+      "learning_rate": 0.0001,
+      "loss": 1.6522,
+      "step": 798
+    },
+    {
+      "epoch": 0.26749246735855375,
+      "grad_norm": 0.12012400478124619,
+      "learning_rate": 0.0001,
+      "loss": 1.5275,
+      "step": 799
+    },
+    {
+      "epoch": 0.26782725142283226,
+      "grad_norm": 0.12273643165826797,
+      "learning_rate": 0.0001,
+      "loss": 1.5848,
+      "step": 800
+    },
+    {
+      "epoch": 0.2681620354871108,
+      "grad_norm": 0.13991284370422363,
+      "learning_rate": 0.0001,
+      "loss": 1.6271,
+      "step": 801
+    },
+    {
+      "epoch": 0.2684968195513894,
+      "grad_norm": 0.1236526146531105,
+      "learning_rate": 0.0001,
+      "loss": 1.57,
+      "step": 802
+    },
+    {
+      "epoch": 0.2688316036156679,
+      "grad_norm": 0.1302153319120407,
+      "learning_rate": 0.0001,
+      "loss": 1.5638,
+      "step": 803
+    },
+    {
+      "epoch": 0.26916638767994644,
+      "grad_norm": 0.11963735520839691,
+      "learning_rate": 0.0001,
+      "loss": 1.6089,
+      "step": 804
+    },
+    {
+      "epoch": 0.269501171744225,
+      "grad_norm": 0.13298673927783966,
+      "learning_rate": 0.0001,
+      "loss": 1.6313,
+      "step": 805
+    },
+    {
+      "epoch": 0.2698359558085035,
+      "grad_norm": 0.13616934418678284,
+      "learning_rate": 0.0001,
+      "loss": 1.653,
+      "step": 806
+    },
+    {
+      "epoch": 0.27017073987278206,
+      "grad_norm": 0.12497668713331223,
+      "learning_rate": 0.0001,
+      "loss": 1.5514,
+      "step": 807
+    },
+    {
+      "epoch": 0.2705055239370606,
+      "grad_norm": 0.11764683574438095,
+      "learning_rate": 0.0001,
+      "loss": 1.5878,
+      "step": 808
+    },
+    {
+      "epoch": 0.2708403080013391,
+      "grad_norm": 0.12114263325929642,
+      "learning_rate": 0.0001,
+      "loss": 1.5628,
+      "step": 809
+    },
+    {
+      "epoch": 0.2711750920656177,
+      "grad_norm": 0.1347784847021103,
+      "learning_rate": 0.0001,
+      "loss": 1.7159,
+      "step": 810
+    },
+    {
+      "epoch": 0.27150987612989624,
+      "grad_norm": 0.12009880691766739,
+      "learning_rate": 0.0001,
+      "loss": 1.6043,
+      "step": 811
+    },
+    {
+      "epoch": 0.27184466019417475,
+      "grad_norm": 0.1278241127729416,
+      "learning_rate": 0.0001,
+      "loss": 1.6309,
+      "step": 812
+    },
+    {
+      "epoch": 0.2721794442584533,
+      "grad_norm": 0.1216406300663948,
+      "learning_rate": 0.0001,
+      "loss": 1.5867,
+      "step": 813
+    },
+    {
+      "epoch": 0.2725142283227318,
+      "grad_norm": 0.11623333394527435,
+      "learning_rate": 0.0001,
+      "loss": 1.5272,
+      "step": 814
+    },
+    {
+      "epoch": 0.27284901238701037,
+      "grad_norm": 0.11762827634811401,
+      "learning_rate": 0.0001,
+      "loss": 1.4148,
+      "step": 815
+    },
+    {
+      "epoch": 0.27318379645128893,
+      "grad_norm": 0.12679798901081085,
+      "learning_rate": 0.0001,
+      "loss": 1.678,
+      "step": 816
+    },
+    {
+      "epoch": 0.27351858051556743,
+      "grad_norm": 0.12463215738534927,
+      "learning_rate": 0.0001,
+      "loss": 1.6383,
+      "step": 817
+    },
+    {
+      "epoch": 0.273853364579846,
+      "grad_norm": 0.12248417735099792,
+      "learning_rate": 0.0001,
+      "loss": 1.5937,
+      "step": 818
+    },
+    {
+      "epoch": 0.27418814864412455,
+      "grad_norm": 0.11953899264335632,
+      "learning_rate": 0.0001,
+      "loss": 1.5704,
+      "step": 819
+    },
+    {
+      "epoch": 0.27452293270840306,
+      "grad_norm": 0.12919148802757263,
+      "learning_rate": 0.0001,
+      "loss": 1.6948,
+      "step": 820
+    },
+    {
+      "epoch": 0.2748577167726816,
+      "grad_norm": 0.11798353493213654,
+      "learning_rate": 0.0001,
+      "loss": 1.4814,
+      "step": 821
+    },
+    {
+      "epoch": 0.2751925008369602,
+      "grad_norm": 0.13017946481704712,
+      "learning_rate": 0.0001,
+      "loss": 1.5837,
+      "step": 822
+    },
+    {
+      "epoch": 0.2755272849012387,
+      "grad_norm": 0.1253434419631958,
+      "learning_rate": 0.0001,
+      "loss": 1.5418,
+      "step": 823
+    },
+    {
+      "epoch": 0.27586206896551724,
+      "grad_norm": 0.11546894907951355,
+      "learning_rate": 0.0001,
+      "loss": 1.5851,
+      "step": 824
+    },
+    {
+      "epoch": 0.2761968530297958,
+      "grad_norm": 0.12117631733417511,
+      "learning_rate": 0.0001,
+      "loss": 1.6335,
+      "step": 825
+    },
+    {
+      "epoch": 0.2765316370940743,
+      "grad_norm": 0.12088704109191895,
+      "learning_rate": 0.0001,
+      "loss": 1.571,
+      "step": 826
+    },
+    {
+      "epoch": 0.27686642115835286,
+      "grad_norm": 0.12261457741260529,
+      "learning_rate": 0.0001,
+      "loss": 1.5143,
+      "step": 827
+    },
+    {
+      "epoch": 0.2772012052226314,
+      "grad_norm": 0.12313897162675858,
+      "learning_rate": 0.0001,
+      "loss": 1.621,
+      "step": 828
+    },
+    {
+      "epoch": 0.2775359892869099,
+      "grad_norm": 0.12563903629779816,
+      "learning_rate": 0.0001,
+      "loss": 1.657,
+      "step": 829
+    },
+    {
+      "epoch": 0.2778707733511885,
+      "grad_norm": 0.1187531128525734,
+      "learning_rate": 0.0001,
+      "loss": 1.5346,
+      "step": 830
+    },
+    {
+      "epoch": 0.27820555741546704,
+      "grad_norm": 0.12233595550060272,
+      "learning_rate": 0.0001,
+      "loss": 1.5835,
+      "step": 831
+    },
+    {
+      "epoch": 0.27854034147974555,
+      "grad_norm": 0.12235147505998611,
+      "learning_rate": 0.0001,
+      "loss": 1.6104,
+      "step": 832
+    },
+    {
+      "epoch": 0.2788751255440241,
+      "grad_norm": 0.11765027791261673,
+      "learning_rate": 0.0001,
+      "loss": 1.5489,
+      "step": 833
+    },
+    {
+      "epoch": 0.27920990960830266,
+      "grad_norm": 0.12349414080381393,
+      "learning_rate": 0.0001,
+      "loss": 1.6089,
+      "step": 834
+    },
+    {
+      "epoch": 0.27954469367258117,
+      "grad_norm": 0.12419009208679199,
+      "learning_rate": 0.0001,
+      "loss": 1.6418,
+      "step": 835
+    },
+    {
+      "epoch": 0.2798794777368597,
+      "grad_norm": 0.12406160682439804,
+      "learning_rate": 0.0001,
+      "loss": 1.5774,
+      "step": 836
+    },
+    {
+      "epoch": 0.2802142618011383,
+      "grad_norm": 0.11722970008850098,
+      "learning_rate": 0.0001,
+      "loss": 1.5634,
+      "step": 837
+    },
+    {
+      "epoch": 0.2805490458654168,
+      "grad_norm": 0.1188267171382904,
+      "learning_rate": 0.0001,
+      "loss": 1.5005,
+      "step": 838
+    },
+    {
+      "epoch": 0.28088382992969535,
+      "grad_norm": 0.11977598071098328,
+      "learning_rate": 0.0001,
+      "loss": 1.5556,
+      "step": 839
+    },
+    {
+      "epoch": 0.2812186139939739,
+      "grad_norm": 0.12196852266788483,
+      "learning_rate": 0.0001,
+      "loss": 1.604,
+      "step": 840
+    },
+    {
+      "epoch": 0.2815533980582524,
+      "grad_norm": 0.12035735696554184,
+      "learning_rate": 0.0001,
+      "loss": 1.5304,
+      "step": 841
+    },
+    {
+      "epoch": 0.281888182122531,
+      "grad_norm": 0.12375766038894653,
+      "learning_rate": 0.0001,
+      "loss": 1.5929,
+      "step": 842
+    },
+    {
+      "epoch": 0.28222296618680953,
+      "grad_norm": 0.1304924190044403,
+      "learning_rate": 0.0001,
+      "loss": 1.6148,
+      "step": 843
+    },
+    {
+      "epoch": 0.28255775025108804,
+      "grad_norm": 0.12864375114440918,
+      "learning_rate": 0.0001,
+      "loss": 1.4907,
+      "step": 844
+    },
+    {
+      "epoch": 0.2828925343153666,
+      "grad_norm": 0.12013059109449387,
+      "learning_rate": 0.0001,
+      "loss": 1.5051,
+      "step": 845
+    },
+    {
+      "epoch": 0.28322731837964515,
+      "grad_norm": 0.1277569979429245,
+      "learning_rate": 0.0001,
+      "loss": 1.5942,
+      "step": 846
+    },
+    {
+      "epoch": 0.28356210244392366,
+      "grad_norm": 0.13474377989768982,
+      "learning_rate": 0.0001,
+      "loss": 1.6098,
+      "step": 847
+    },
+    {
+      "epoch": 0.2838968865082022,
+      "grad_norm": 0.12635944783687592,
+      "learning_rate": 0.0001,
+      "loss": 1.6217,
+      "step": 848
+    },
+    {
+      "epoch": 0.2842316705724808,
+      "grad_norm": 0.12218885123729706,
+      "learning_rate": 0.0001,
+      "loss": 1.578,
+      "step": 849
+    },
+    {
+      "epoch": 0.2845664546367593,
+      "grad_norm": 0.12037128210067749,
+      "learning_rate": 0.0001,
+      "loss": 1.5502,
+      "step": 850
+    },
+    {
+      "epoch": 0.28490123870103784,
+      "grad_norm": 0.12386499345302582,
+      "learning_rate": 0.0001,
+      "loss": 1.6922,
+      "step": 851
+    },
+    {
+      "epoch": 0.28523602276531634,
+      "grad_norm": 0.1298052966594696,
+      "learning_rate": 0.0001,
+      "loss": 1.6589,
+      "step": 852
+    },
+    {
+      "epoch": 0.2855708068295949,
+      "grad_norm": 0.12143804877996445,
+      "learning_rate": 0.0001,
+      "loss": 1.5856,
+      "step": 853
+    },
+    {
+      "epoch": 0.28590559089387346,
+      "grad_norm": 0.11675681918859482,
+      "learning_rate": 0.0001,
+      "loss": 1.4877,
+      "step": 854
+    },
+    {
+      "epoch": 0.28624037495815197,
+      "grad_norm": 0.11870943009853363,
+      "learning_rate": 0.0001,
+      "loss": 1.5699,
+      "step": 855
+    },
+    {
+      "epoch": 0.2865751590224305,
+      "grad_norm": 0.12752340734004974,
+      "learning_rate": 0.0001,
+      "loss": 1.5648,
+      "step": 856
+    },
+    {
+      "epoch": 0.2869099430867091,
+      "grad_norm": 0.1254730522632599,
+      "learning_rate": 0.0001,
+      "loss": 1.5331,
+      "step": 857
+    },
+    {
+      "epoch": 0.2872447271509876,
+      "grad_norm": 0.12351144105195999,
+      "learning_rate": 0.0001,
+      "loss": 1.5984,
+      "step": 858
+    },
+    {
+      "epoch": 0.28757951121526615,
+      "grad_norm": 0.12823925912380219,
+      "learning_rate": 0.0001,
+      "loss": 1.4704,
+      "step": 859
+    },
+    {
+      "epoch": 0.2879142952795447,
+      "grad_norm": 0.12884090840816498,
+      "learning_rate": 0.0001,
+      "loss": 1.5302,
+      "step": 860
+    },
+    {
+      "epoch": 0.2882490793438232,
+      "grad_norm": 0.12310319393873215,
+      "learning_rate": 0.0001,
+      "loss": 1.5554,
+      "step": 861
+    },
+    {
+      "epoch": 0.28858386340810177,
+      "grad_norm": 0.12592901289463043,
+      "learning_rate": 0.0001,
+      "loss": 1.573,
+      "step": 862
+    },
+    {
+      "epoch": 0.28891864747238033,
+      "grad_norm": 0.12326246500015259,
+      "learning_rate": 0.0001,
+      "loss": 1.5408,
+      "step": 863
+    },
+    {
+      "epoch": 0.28925343153665883,
+      "grad_norm": 0.12688298523426056,
+      "learning_rate": 0.0001,
+      "loss": 1.609,
+      "step": 864
+    },
+    {
+      "epoch": 0.2895882156009374,
+      "grad_norm": 0.13284268975257874,
+      "learning_rate": 0.0001,
+      "loss": 1.4774,
+      "step": 865
+    },
+    {
+      "epoch": 0.28992299966521595,
+      "grad_norm": 0.12346718460321426,
+      "learning_rate": 0.0001,
+      "loss": 1.5675,
+      "step": 866
+    },
+    {
+      "epoch": 0.29025778372949446,
+      "grad_norm": 0.12501643598079681,
+      "learning_rate": 0.0001,
+      "loss": 1.54,
+      "step": 867
+    },
+    {
+      "epoch": 0.290592567793773,
+      "grad_norm": 0.14129911363124847,
+      "learning_rate": 0.0001,
+      "loss": 1.5707,
+      "step": 868
+    },
+    {
+      "epoch": 0.2909273518580516,
+      "grad_norm": 0.11998032033443451,
+      "learning_rate": 0.0001,
+      "loss": 1.6025,
+      "step": 869
+    },
+    {
+      "epoch": 0.2912621359223301,
+      "grad_norm": 0.14502458274364471,
+      "learning_rate": 0.0001,
+      "loss": 1.6066,
+      "step": 870
+    },
+    {
+      "epoch": 0.29159691998660864,
+      "grad_norm": 0.13429078459739685,
+      "learning_rate": 0.0001,
+      "loss": 1.5773,
+      "step": 871
+    },
+    {
+      "epoch": 0.2919317040508872,
+      "grad_norm": 0.12702088057994843,
+      "learning_rate": 0.0001,
+      "loss": 1.5331,
+      "step": 872
+    },
+    {
+      "epoch": 0.2922664881151657,
+      "grad_norm": 0.1450689435005188,
+      "learning_rate": 0.0001,
+      "loss": 1.6426,
+      "step": 873
+    },
+    {
+      "epoch": 0.29260127217944426,
+      "grad_norm": 0.12571430206298828,
+      "learning_rate": 0.0001,
+      "loss": 1.5702,
+      "step": 874
+    },
+    {
+      "epoch": 0.2929360562437228,
+      "grad_norm": 0.15491126477718353,
+      "learning_rate": 0.0001,
+      "loss": 1.6229,
+      "step": 875
+    },
+    {
+      "epoch": 0.2932708403080013,
+      "grad_norm": 0.1497523933649063,
+      "learning_rate": 0.0001,
+      "loss": 1.6073,
+      "step": 876
+    },
+    {
+      "epoch": 0.2936056243722799,
+      "grad_norm": 0.12279631197452545,
+      "learning_rate": 0.0001,
+      "loss": 1.5836,
+      "step": 877
+    },
+    {
+      "epoch": 0.29394040843655844,
+      "grad_norm": 0.16039283573627472,
+      "learning_rate": 0.0001,
+      "loss": 1.6125,
+      "step": 878
+    },
+    {
+      "epoch": 0.29427519250083695,
+      "grad_norm": 0.1275695562362671,
+      "learning_rate": 0.0001,
+      "loss": 1.5279,
+      "step": 879
+    },
+    {
+      "epoch": 0.2946099765651155,
+      "grad_norm": 0.12885813415050507,
+      "learning_rate": 0.0001,
+      "loss": 1.5662,
+      "step": 880
+    },
+    {
+      "epoch": 0.29494476062939406,
+      "grad_norm": 0.1439967006444931,
+      "learning_rate": 0.0001,
+      "loss": 1.6408,
+      "step": 881
+    },
+    {
+      "epoch": 0.29527954469367257,
+      "grad_norm": 0.12064056098461151,
+      "learning_rate": 0.0001,
+      "loss": 1.5292,
+      "step": 882
+    },
+    {
+      "epoch": 0.29561432875795113,
+      "grad_norm": 0.12883847951889038,
+      "learning_rate": 0.0001,
+      "loss": 1.6024,
+      "step": 883
+    },
+    {
+      "epoch": 0.2959491128222297,
+      "grad_norm": 0.12654966115951538,
+      "learning_rate": 0.0001,
+      "loss": 1.5838,
+      "step": 884
+    },
+    {
+      "epoch": 0.2962838968865082,
+      "grad_norm": 0.13914820551872253,
+      "learning_rate": 0.0001,
+      "loss": 1.5345,
+      "step": 885
+    },
+    {
+      "epoch": 0.29661868095078675,
+      "grad_norm": 0.12559537589550018,
+      "learning_rate": 0.0001,
+      "loss": 1.515,
+      "step": 886
+    },
+    {
+      "epoch": 0.2969534650150653,
+      "grad_norm": 0.1451893299818039,
+      "learning_rate": 0.0001,
+      "loss": 1.5924,
+      "step": 887
+    },
+    {
+      "epoch": 0.2972882490793438,
+      "grad_norm": 0.13416925072669983,
+      "learning_rate": 0.0001,
+      "loss": 1.6371,
+      "step": 888
+    },
+    {
+      "epoch": 0.2976230331436224,
+      "grad_norm": 0.12274248152971268,
+      "learning_rate": 0.0001,
+      "loss": 1.6539,
+      "step": 889
+    },
+    {
+      "epoch": 0.2979578172079009,
+      "grad_norm": 0.143101766705513,
+      "learning_rate": 0.0001,
+      "loss": 1.5748,
+      "step": 890
+    },
+    {
+      "epoch": 0.29829260127217944,
+      "grad_norm": 0.12564097344875336,
+      "learning_rate": 0.0001,
+      "loss": 1.5875,
+      "step": 891
+    },
+    {
+      "epoch": 0.298627385336458,
+      "grad_norm": 0.12403486669063568,
+      "learning_rate": 0.0001,
+      "loss": 1.5765,
+      "step": 892
+    },
+    {
+      "epoch": 0.2989621694007365,
+      "grad_norm": 0.13099223375320435,
+      "learning_rate": 0.0001,
+      "loss": 1.5656,
+      "step": 893
+    },
+    {
+      "epoch": 0.29929695346501506,
+      "grad_norm": 0.12135787308216095,
+      "learning_rate": 0.0001,
+      "loss": 1.4958,
+      "step": 894
+    },
+    {
+      "epoch": 0.2996317375292936,
+      "grad_norm": 0.12442804127931595,
+      "learning_rate": 0.0001,
+      "loss": 1.6222,
+      "step": 895
+    },
+    {
+      "epoch": 0.2999665215935721,
+      "grad_norm": 0.12768028676509857,
+      "learning_rate": 0.0001,
+      "loss": 1.6719,
+      "step": 896
+    },
+    {
+      "epoch": 0.3003013056578507,
+      "grad_norm": 0.1240835040807724,
+      "learning_rate": 0.0001,
+      "loss": 1.5114,
+      "step": 897
+    },
+    {
+      "epoch": 0.30063608972212924,
+      "grad_norm": 0.12057949602603912,
+      "learning_rate": 0.0001,
+      "loss": 1.5864,
+      "step": 898
+    },
+    {
+      "epoch": 0.30097087378640774,
+      "grad_norm": 0.1332257241010666,
+      "learning_rate": 0.0001,
+      "loss": 1.652,
+      "step": 899
+    },
+    {
+      "epoch": 0.3013056578506863,
+      "grad_norm": 0.12191877514123917,
+      "learning_rate": 0.0001,
+      "loss": 1.6016,
+      "step": 900
+    },
+    {
+      "epoch": 0.30164044191496486,
+      "grad_norm": 0.13481038808822632,
+      "learning_rate": 0.0001,
+      "loss": 1.5724,
+      "step": 901
+    },
+    {
+      "epoch": 0.30197522597924337,
+      "grad_norm": 0.12434981763362885,
+      "learning_rate": 0.0001,
+      "loss": 1.5873,
+      "step": 902
+    },
+    {
+      "epoch": 0.3023100100435219,
+      "grad_norm": 0.12398968636989594,
+      "learning_rate": 0.0001,
+      "loss": 1.5917,
+      "step": 903
+    },
+    {
+      "epoch": 0.3026447941078005,
+      "grad_norm": 0.13455741107463837,
+      "learning_rate": 0.0001,
+      "loss": 1.6293,
+      "step": 904
+    },
+    {
+      "epoch": 0.302979578172079,
+      "grad_norm": 0.12864330410957336,
+      "learning_rate": 0.0001,
+      "loss": 1.6671,
+      "step": 905
+    },
+    {
+      "epoch": 0.30331436223635755,
+      "grad_norm": 0.1306915581226349,
+      "learning_rate": 0.0001,
+      "loss": 1.5669,
+      "step": 906
+    },
+    {
+      "epoch": 0.3036491463006361,
+      "grad_norm": 0.12770214676856995,
+      "learning_rate": 0.0001,
+      "loss": 1.515,
+      "step": 907
+    },
+    {
+      "epoch": 0.3039839303649146,
+      "grad_norm": 0.12244972586631775,
+      "learning_rate": 0.0001,
+      "loss": 1.7102,
+      "step": 908
+    },
+    {
+      "epoch": 0.30431871442919317,
+      "grad_norm": 0.12544330954551697,
+      "learning_rate": 0.0001,
+      "loss": 1.5809,
+      "step": 909
+    },
+    {
+      "epoch": 0.30465349849347173,
+      "grad_norm": 0.12653569877147675,
+      "learning_rate": 0.0001,
+      "loss": 1.5504,
+      "step": 910
+    },
+    {
+      "epoch": 0.30498828255775023,
+      "grad_norm": 0.1295597404241562,
+      "learning_rate": 0.0001,
+      "loss": 1.6077,
+      "step": 911
+    },
+    {
+      "epoch": 0.3053230666220288,
+      "grad_norm": 0.13423195481300354,
+      "learning_rate": 0.0001,
+      "loss": 1.6433,
+      "step": 912
+    },
+    {
+      "epoch": 0.30565785068630735,
+      "grad_norm": 0.12957747280597687,
+      "learning_rate": 0.0001,
+      "loss": 1.72,
+      "step": 913
+    },
+    {
+      "epoch": 0.30599263475058586,
+      "grad_norm": 0.1274273693561554,
+      "learning_rate": 0.0001,
+      "loss": 1.5916,
+      "step": 914
+    },
+    {
+      "epoch": 0.3063274188148644,
+      "grad_norm": 0.12693728506565094,
+      "learning_rate": 0.0001,
+      "loss": 1.5582,
+      "step": 915
+    },
+    {
+      "epoch": 0.306662202879143,
+      "grad_norm": 0.12224942445755005,
+      "learning_rate": 0.0001,
+      "loss": 1.6431,
+      "step": 916
+    },
+    {
+      "epoch": 0.3069969869434215,
+      "grad_norm": 0.12495341151952744,
+      "learning_rate": 0.0001,
+      "loss": 1.6554,
+      "step": 917
+    },
+    {
+      "epoch": 0.30733177100770004,
+      "grad_norm": 0.12348316609859467,
+      "learning_rate": 0.0001,
+      "loss": 1.5617,
+      "step": 918
+    },
+    {
+      "epoch": 0.3076665550719786,
+      "grad_norm": 0.12086449563503265,
+      "learning_rate": 0.0001,
+      "loss": 1.5866,
+      "step": 919
+    },
+    {
+      "epoch": 0.3080013391362571,
+      "grad_norm": 0.12970371544361115,
+      "learning_rate": 0.0001,
+      "loss": 1.6444,
+      "step": 920
+    },
+    {
+      "epoch": 0.30833612320053566,
+      "grad_norm": 0.115717314183712,
+      "learning_rate": 0.0001,
+      "loss": 1.4493,
+      "step": 921
+    },
+    {
+      "epoch": 0.3086709072648142,
+      "grad_norm": 0.1250089704990387,
+      "learning_rate": 0.0001,
+      "loss": 1.5889,
+      "step": 922
+    },
+    {
+      "epoch": 0.3090056913290927,
+      "grad_norm": 0.11084622144699097,
+      "learning_rate": 0.0001,
+      "loss": 1.3815,
+      "step": 923
+    },
+    {
+      "epoch": 0.3093404753933713,
+      "grad_norm": 0.12127161026000977,
+      "learning_rate": 0.0001,
+      "loss": 1.5558,
+      "step": 924
+    },
+    {
+      "epoch": 0.30967525945764984,
+      "grad_norm": 0.12244665622711182,
+      "learning_rate": 0.0001,
+      "loss": 1.6409,
+      "step": 925
+    },
+    {
+      "epoch": 0.31001004352192835,
+      "grad_norm": 0.12553781270980835,
+      "learning_rate": 0.0001,
+      "loss": 1.6205,
+      "step": 926
+    },
+    {
+      "epoch": 0.3103448275862069,
+      "grad_norm": 0.12222031503915787,
+      "learning_rate": 0.0001,
+      "loss": 1.6323,
+      "step": 927
+    },
+    {
+      "epoch": 0.3106796116504854,
+      "grad_norm": 0.1246923953294754,
+      "learning_rate": 0.0001,
+      "loss": 1.719,
+      "step": 928
+    },
+    {
+      "epoch": 0.31101439571476397,
+      "grad_norm": 0.13237862288951874,
+      "learning_rate": 0.0001,
+      "loss": 1.6517,
+      "step": 929
+    },
+    {
+      "epoch": 0.31134917977904253,
+      "grad_norm": 0.11562683433294296,
+      "learning_rate": 0.0001,
+      "loss": 1.5043,
+      "step": 930
+    },
+    {
+      "epoch": 0.31168396384332103,
+      "grad_norm": 0.12860921025276184,
+      "learning_rate": 0.0001,
+      "loss": 1.5939,
+      "step": 931
+    },
+    {
+      "epoch": 0.3120187479075996,
+      "grad_norm": 0.11789809912443161,
+      "learning_rate": 0.0001,
+      "loss": 1.4763,
+      "step": 932
+    },
+    {
+      "epoch": 0.31235353197187815,
+      "grad_norm": 0.12612248957157135,
+      "learning_rate": 0.0001,
+      "loss": 1.6355,
+      "step": 933
+    },
+    {
+      "epoch": 0.31268831603615665,
+      "grad_norm": 0.14561748504638672,
+      "learning_rate": 0.0001,
+      "loss": 1.6897,
+      "step": 934
+    },
+    {
+      "epoch": 0.3130231001004352,
+      "grad_norm": 0.1276092380285263,
+      "learning_rate": 0.0001,
+      "loss": 1.6438,
+      "step": 935
+    },
+    {
+      "epoch": 0.3133578841647138,
+      "grad_norm": 0.13539274036884308,
+      "learning_rate": 0.0001,
+      "loss": 1.5562,
+      "step": 936
+    },
+    {
+      "epoch": 0.3136926682289923,
+      "grad_norm": 0.12490363419055939,
+      "learning_rate": 0.0001,
+      "loss": 1.5592,
+      "step": 937
+    },
+    {
+      "epoch": 0.31402745229327084,
+      "grad_norm": 0.12392627447843552,
+      "learning_rate": 0.0001,
+      "loss": 1.6344,
+      "step": 938
+    },
+    {
+      "epoch": 0.3143622363575494,
+      "grad_norm": 0.13469712436199188,
+      "learning_rate": 0.0001,
+      "loss": 1.7123,
+      "step": 939
+    },
+    {
+      "epoch": 0.3146970204218279,
+      "grad_norm": 0.13380196690559387,
+      "learning_rate": 0.0001,
+      "loss": 1.6485,
+      "step": 940
+    },
+    {
+      "epoch": 0.31503180448610646,
+      "grad_norm": 0.12370868027210236,
+      "learning_rate": 0.0001,
+      "loss": 1.5663,
+      "step": 941
+    },
+    {
+      "epoch": 0.315366588550385,
+      "grad_norm": 0.1381116360425949,
+      "learning_rate": 0.0001,
+      "loss": 1.5682,
+      "step": 942
+    },
+    {
+      "epoch": 0.3157013726146635,
+      "grad_norm": 0.15112708508968353,
+      "learning_rate": 0.0001,
+      "loss": 1.6236,
+      "step": 943
+    },
+    {
+      "epoch": 0.3160361566789421,
+      "grad_norm": 0.13402314484119415,
+      "learning_rate": 0.0001,
+      "loss": 1.67,
+      "step": 944
+    },
+    {
+      "epoch": 0.31637094074322064,
+      "grad_norm": 0.13505329191684723,
+      "learning_rate": 0.0001,
+      "loss": 1.5149,
+      "step": 945
+    },
+    {
+      "epoch": 0.31670572480749914,
+      "grad_norm": 0.1328267902135849,
+      "learning_rate": 0.0001,
+      "loss": 1.5129,
+      "step": 946
+    },
+    {
+      "epoch": 0.3170405088717777,
+      "grad_norm": 0.12792791426181793,
+      "learning_rate": 0.0001,
+      "loss": 1.5868,
+      "step": 947
+    },
+    {
+      "epoch": 0.31737529293605626,
+      "grad_norm": 0.11726494878530502,
+      "learning_rate": 0.0001,
+      "loss": 1.5581,
+      "step": 948
+    },
+    {
+      "epoch": 0.31771007700033477,
+      "grad_norm": 0.12302982062101364,
+      "learning_rate": 0.0001,
+      "loss": 1.5296,
+      "step": 949
+    },
+    {
+      "epoch": 0.3180448610646133,
+      "grad_norm": 0.1206970065832138,
+      "learning_rate": 0.0001,
+      "loss": 1.5066,
+      "step": 950
+    },
+    {
+      "epoch": 0.3183796451288919,
+      "grad_norm": 0.1165679469704628,
+      "learning_rate": 0.0001,
+      "loss": 1.5486,
+      "step": 951
+    },
+    {
+      "epoch": 0.3187144291931704,
+      "grad_norm": 0.12752187252044678,
+      "learning_rate": 0.0001,
+      "loss": 1.6441,
+      "step": 952
+    },
+    {
+      "epoch": 0.31904921325744895,
+      "grad_norm": 0.12091311067342758,
+      "learning_rate": 0.0001,
+      "loss": 1.5482,
+      "step": 953
+    },
+    {
+      "epoch": 0.3193839973217275,
+      "grad_norm": 0.12838125228881836,
+      "learning_rate": 0.0001,
+      "loss": 1.6027,
+      "step": 954
+    },
+    {
+      "epoch": 0.319718781386006,
+      "grad_norm": 0.11839887499809265,
+      "learning_rate": 0.0001,
+      "loss": 1.5533,
+      "step": 955
+    },
+    {
+      "epoch": 0.32005356545028457,
+      "grad_norm": 0.1277683675289154,
+      "learning_rate": 0.0001,
+      "loss": 1.5461,
+      "step": 956
+    },
+    {
+      "epoch": 0.32038834951456313,
+      "grad_norm": 0.12134066224098206,
+      "learning_rate": 0.0001,
+      "loss": 1.5649,
+      "step": 957
+    },
+    {
+      "epoch": 0.32072313357884163,
+      "grad_norm": 0.12735500931739807,
+      "learning_rate": 0.0001,
+      "loss": 1.608,
+      "step": 958
+    },
+    {
+      "epoch": 0.3210579176431202,
+      "grad_norm": 0.133828267455101,
+      "learning_rate": 0.0001,
+      "loss": 1.5675,
+      "step": 959
+    },
+    {
+      "epoch": 0.32139270170739875,
+      "grad_norm": 0.12437241524457932,
+      "learning_rate": 0.0001,
+      "loss": 1.6325,
+      "step": 960
+    },
+    {
+      "epoch": 0.32172748577167726,
+      "grad_norm": 0.12489302456378937,
+      "learning_rate": 0.0001,
+      "loss": 1.6441,
+      "step": 961
+    },
+    {
+      "epoch": 0.3220622698359558,
+      "grad_norm": 0.12957216799259186,
+      "learning_rate": 0.0001,
+      "loss": 1.5328,
+      "step": 962
+    },
+    {
+      "epoch": 0.3223970539002344,
+      "grad_norm": 0.1317603886127472,
+      "learning_rate": 0.0001,
+      "loss": 1.6061,
+      "step": 963
+    },
+    {
+      "epoch": 0.3227318379645129,
+      "grad_norm": 0.12075690180063248,
+      "learning_rate": 0.0001,
+      "loss": 1.5508,
+      "step": 964
+    },
+    {
+      "epoch": 0.32306662202879144,
+      "grad_norm": 0.11924642324447632,
+      "learning_rate": 0.0001,
+      "loss": 1.4772,
+      "step": 965
+    },
+    {
+      "epoch": 0.32340140609306994,
+      "grad_norm": 0.12515272200107574,
+      "learning_rate": 0.0001,
+      "loss": 1.5748,
+      "step": 966
+    },
+    {
+      "epoch": 0.3237361901573485,
+      "grad_norm": 0.11952123045921326,
+      "learning_rate": 0.0001,
+      "loss": 1.5852,
+      "step": 967
+    },
+    {
+      "epoch": 0.32407097422162706,
+      "grad_norm": 0.125240296125412,
+      "learning_rate": 0.0001,
+      "loss": 1.5388,
+      "step": 968
+    },
+    {
+      "epoch": 0.32440575828590557,
+      "grad_norm": 0.12284346669912338,
+      "learning_rate": 0.0001,
+      "loss": 1.6059,
+      "step": 969
+    },
+    {
+      "epoch": 0.3247405423501841,
+      "grad_norm": 0.11825854331254959,
+      "learning_rate": 0.0001,
+      "loss": 1.52,
+      "step": 970
+    },
+    {
+      "epoch": 0.3250753264144627,
+      "grad_norm": 0.1247822567820549,
+      "learning_rate": 0.0001,
+      "loss": 1.6265,
+      "step": 971
+    },
+    {
+      "epoch": 0.3254101104787412,
+      "grad_norm": 0.12490460276603699,
+      "learning_rate": 0.0001,
+      "loss": 1.6047,
+      "step": 972
+    },
+    {
+      "epoch": 0.32574489454301975,
+      "grad_norm": 0.11784359812736511,
+      "learning_rate": 0.0001,
+      "loss": 1.451,
+      "step": 973
+    },
+    {
+      "epoch": 0.3260796786072983,
+      "grad_norm": 0.12558013200759888,
+      "learning_rate": 0.0001,
+      "loss": 1.6244,
+      "step": 974
+    },
+    {
+      "epoch": 0.3264144626715768,
+      "grad_norm": 0.12492769211530685,
+      "learning_rate": 0.0001,
+      "loss": 1.6821,
+      "step": 975
+    },
+    {
+      "epoch": 0.32674924673585537,
+      "grad_norm": 0.11894410103559494,
+      "learning_rate": 0.0001,
+      "loss": 1.5476,
+      "step": 976
+    },
+    {
+      "epoch": 0.32708403080013393,
+      "grad_norm": 0.12406729906797409,
+      "learning_rate": 0.0001,
+      "loss": 1.5954,
+      "step": 977
+    },
+    {
+      "epoch": 0.32741881486441243,
+      "grad_norm": 0.12805567681789398,
+      "learning_rate": 0.0001,
+      "loss": 1.5216,
+      "step": 978
+    },
+    {
+      "epoch": 0.327753598928691,
+      "grad_norm": 0.12648111581802368,
+      "learning_rate": 0.0001,
+      "loss": 1.6923,
+      "step": 979
+    },
+    {
+      "epoch": 0.32808838299296955,
+      "grad_norm": 0.12503187358379364,
+      "learning_rate": 0.0001,
+      "loss": 1.6204,
+      "step": 980
+    },
+    {
+      "epoch": 0.32842316705724806,
+      "grad_norm": 0.12180895358324051,
+      "learning_rate": 0.0001,
+      "loss": 1.5764,
+      "step": 981
+    },
+    {
+      "epoch": 0.3287579511215266,
+      "grad_norm": 0.12118836492300034,
+      "learning_rate": 0.0001,
+      "loss": 1.4937,
+      "step": 982
+    },
+    {
+      "epoch": 0.3290927351858052,
+      "grad_norm": 0.12758868932724,
+      "learning_rate": 0.0001,
+      "loss": 1.6198,
+      "step": 983
+    },
+    {
+      "epoch": 0.3294275192500837,
+      "grad_norm": 0.1190565824508667,
+      "learning_rate": 0.0001,
+      "loss": 1.587,
+      "step": 984
+    },
+    {
+      "epoch": 0.32976230331436224,
+      "grad_norm": 0.12521426379680634,
+      "learning_rate": 0.0001,
+      "loss": 1.5403,
+      "step": 985
+    },
+    {
+      "epoch": 0.3300970873786408,
+      "grad_norm": 0.1259697824716568,
+      "learning_rate": 0.0001,
+      "loss": 1.5356,
+      "step": 986
+    },
+    {
+      "epoch": 0.3304318714429193,
+      "grad_norm": 0.12639686465263367,
+      "learning_rate": 0.0001,
+      "loss": 1.5941,
+      "step": 987
+    },
+    {
+      "epoch": 0.33076665550719786,
+      "grad_norm": 0.12533701956272125,
+      "learning_rate": 0.0001,
+      "loss": 1.6826,
+      "step": 988
+    },
+    {
+      "epoch": 0.3311014395714764,
+      "grad_norm": 0.1349916309118271,
+      "learning_rate": 0.0001,
+      "loss": 1.6818,
+      "step": 989
+    },
+    {
+      "epoch": 0.3314362236357549,
+      "grad_norm": 0.12522515654563904,
+      "learning_rate": 0.0001,
+      "loss": 1.531,
+      "step": 990
+    },
+    {
+      "epoch": 0.3317710077000335,
+      "grad_norm": 0.12278946489095688,
+      "learning_rate": 0.0001,
+      "loss": 1.5098,
+      "step": 991
+    },
+    {
+      "epoch": 0.33210579176431204,
+      "grad_norm": 0.1286853700876236,
+      "learning_rate": 0.0001,
+      "loss": 1.5117,
+      "step": 992
+    },
+    {
+      "epoch": 0.33244057582859055,
+      "grad_norm": 0.1212511882185936,
+      "learning_rate": 0.0001,
+      "loss": 1.4762,
+      "step": 993
+    },
+    {
+      "epoch": 0.3327753598928691,
+      "grad_norm": 0.1347900927066803,
+      "learning_rate": 0.0001,
+      "loss": 1.6793,
+      "step": 994
+    },
+    {
+      "epoch": 0.33311014395714766,
+      "grad_norm": 0.11994650959968567,
+      "learning_rate": 0.0001,
+      "loss": 1.6026,
+      "step": 995
+    },
+    {
+      "epoch": 0.33344492802142617,
+      "grad_norm": 0.13167862594127655,
+      "learning_rate": 0.0001,
+      "loss": 1.6341,
+      "step": 996
+    },
+    {
+      "epoch": 0.3337797120857047,
+      "grad_norm": 0.13315805792808533,
+      "learning_rate": 0.0001,
+      "loss": 1.5414,
+      "step": 997
+    },
+    {
+      "epoch": 0.3341144961499833,
+      "grad_norm": 0.12088074535131454,
+      "learning_rate": 0.0001,
+      "loss": 1.5769,
+      "step": 998
+    },
+    {
+      "epoch": 0.3344492802142618,
+      "grad_norm": 0.13783089816570282,
+      "learning_rate": 0.0001,
+      "loss": 1.5365,
+      "step": 999
+    },
+    {
+      "epoch": 0.33478406427854035,
+      "grad_norm": 0.13187260925769806,
+      "learning_rate": 0.0001,
+      "loss": 1.5929,
+      "step": 1000
+    },
+    {
+      "epoch": 0.3351188483428189,
+      "grad_norm": 0.13189886510372162,
+      "learning_rate": 0.0001,
+      "loss": 1.5591,
+      "step": 1001
+    },
+    {
+      "epoch": 0.3354536324070974,
+      "grad_norm": 0.1421831101179123,
+      "learning_rate": 0.0001,
+      "loss": 1.5674,
+      "step": 1002
+    },
+    {
+      "epoch": 0.335788416471376,
+      "grad_norm": 0.1282414346933365,
+      "learning_rate": 0.0001,
+      "loss": 1.5696,
+      "step": 1003
+    },
+    {
+      "epoch": 0.3361232005356545,
+      "grad_norm": 0.13641226291656494,
+      "learning_rate": 0.0001,
+      "loss": 1.5336,
+      "step": 1004
+    },
+    {
+      "epoch": 0.33645798459993304,
+      "grad_norm": 0.14396816492080688,
+      "learning_rate": 0.0001,
+      "loss": 1.5648,
+      "step": 1005
+    },
+    {
+      "epoch": 0.3367927686642116,
+      "grad_norm": 0.12792754173278809,
+      "learning_rate": 0.0001,
+      "loss": 1.631,
+      "step": 1006
+    },
+    {
+      "epoch": 0.3371275527284901,
+      "grad_norm": 0.1327052116394043,
+      "learning_rate": 0.0001,
+      "loss": 1.5746,
+      "step": 1007
+    },
+    {
+      "epoch": 0.33746233679276866,
+      "grad_norm": 0.14353278279304504,
+      "learning_rate": 0.0001,
+      "loss": 1.5345,
+      "step": 1008
+    },
+    {
+      "epoch": 0.3377971208570472,
+      "grad_norm": 0.137548565864563,
+      "learning_rate": 0.0001,
+      "loss": 1.6771,
+      "step": 1009
+    },
+    {
+      "epoch": 0.3381319049213257,
+      "grad_norm": 0.13727347552776337,
+      "learning_rate": 0.0001,
+      "loss": 1.6451,
+      "step": 1010
+    },
+    {
+      "epoch": 0.3384666889856043,
+      "grad_norm": 0.13395574688911438,
+      "learning_rate": 0.0001,
+      "loss": 1.5378,
+      "step": 1011
+    },
+    {
+      "epoch": 0.33880147304988284,
+      "grad_norm": 0.12692630290985107,
+      "learning_rate": 0.0001,
+      "loss": 1.5555,
+      "step": 1012
+    },
+    {
+      "epoch": 0.33913625711416134,
+      "grad_norm": 0.12900549173355103,
+      "learning_rate": 0.0001,
+      "loss": 1.5451,
+      "step": 1013
+    },
+    {
+      "epoch": 0.3394710411784399,
+      "grad_norm": 0.11654023826122284,
+      "learning_rate": 0.0001,
+      "loss": 1.5063,
+      "step": 1014
+    },
+    {
+      "epoch": 0.33980582524271846,
+      "grad_norm": 0.13518574833869934,
+      "learning_rate": 0.0001,
+      "loss": 1.5578,
+      "step": 1015
+    },
+    {
+      "epoch": 0.34014060930699697,
+      "grad_norm": 0.126609668135643,
+      "learning_rate": 0.0001,
+      "loss": 1.4299,
+      "step": 1016
+    },
+    {
+      "epoch": 0.3404753933712755,
+      "grad_norm": 0.12412185966968536,
+      "learning_rate": 0.0001,
+      "loss": 1.5083,
+      "step": 1017
+    },
+    {
+      "epoch": 0.3408101774355541,
+      "grad_norm": 0.12521536648273468,
+      "learning_rate": 0.0001,
+      "loss": 1.5264,
+      "step": 1018
+    },
+    {
+      "epoch": 0.3411449614998326,
+      "grad_norm": 0.12396744638681412,
+      "learning_rate": 0.0001,
+      "loss": 1.5984,
+      "step": 1019
+    },
+    {
+      "epoch": 0.34147974556411115,
+      "grad_norm": 0.12353380024433136,
+      "learning_rate": 0.0001,
+      "loss": 1.5615,
+      "step": 1020
+    },
+    {
+      "epoch": 0.3418145296283897,
+      "grad_norm": 0.1337115615606308,
+      "learning_rate": 0.0001,
+      "loss": 1.5777,
+      "step": 1021
+    },
+    {
+      "epoch": 0.3421493136926682,
+      "grad_norm": 0.13354641199111938,
+      "learning_rate": 0.0001,
+      "loss": 1.5417,
+      "step": 1022
+    },
+    {
+      "epoch": 0.34248409775694677,
+      "grad_norm": 0.12444625794887543,
+      "learning_rate": 0.0001,
+      "loss": 1.579,
+      "step": 1023
+    },
+    {
+      "epoch": 0.34281888182122533,
+      "grad_norm": 0.12876839935779572,
+      "learning_rate": 0.0001,
+      "loss": 1.4921,
+      "step": 1024
+    },
+    {
+      "epoch": 0.34315366588550383,
+      "grad_norm": 0.13097478449344635,
+      "learning_rate": 0.0001,
+      "loss": 1.5756,
+      "step": 1025
+    },
+    {
+      "epoch": 0.3434884499497824,
+      "grad_norm": 0.1257512867450714,
+      "learning_rate": 0.0001,
+      "loss": 1.5273,
+      "step": 1026
+    },
+    {
+      "epoch": 0.34382323401406095,
+      "grad_norm": 0.13378176093101501,
+      "learning_rate": 0.0001,
+      "loss": 1.5484,
+      "step": 1027
+    },
+    {
+      "epoch": 0.34415801807833946,
+      "grad_norm": 0.1325940638780594,
+      "learning_rate": 0.0001,
+      "loss": 1.6229,
+      "step": 1028
+    },
+    {
+      "epoch": 0.344492802142618,
+      "grad_norm": 0.11962547153234482,
+      "learning_rate": 0.0001,
+      "loss": 1.4859,
+      "step": 1029
+    },
+    {
+      "epoch": 0.3448275862068966,
+      "grad_norm": 0.12927503883838654,
+      "learning_rate": 0.0001,
+      "loss": 1.6788,
+      "step": 1030
+    },
+    {
+      "epoch": 0.3451623702711751,
+      "grad_norm": 0.13427825272083282,
+      "learning_rate": 0.0001,
+      "loss": 1.5514,
+      "step": 1031
+    },
+    {
+      "epoch": 0.34549715433545364,
+      "grad_norm": 0.13139428198337555,
+      "learning_rate": 0.0001,
+      "loss": 1.6164,
+      "step": 1032
+    },
+    {
+      "epoch": 0.3458319383997322,
+      "grad_norm": 0.12266752868890762,
+      "learning_rate": 0.0001,
+      "loss": 1.5226,
+      "step": 1033
+    },
+    {
+      "epoch": 0.3461667224640107,
+      "grad_norm": 0.14490726590156555,
+      "learning_rate": 0.0001,
+      "loss": 1.5562,
+      "step": 1034
+    },
+    {
+      "epoch": 0.34650150652828926,
+      "grad_norm": 0.11922143399715424,
+      "learning_rate": 0.0001,
+      "loss": 1.465,
+      "step": 1035
+    },
+    {
+      "epoch": 0.3468362905925678,
+      "grad_norm": 0.12442134320735931,
+      "learning_rate": 0.0001,
+      "loss": 1.5653,
+      "step": 1036
+    },
+    {
+      "epoch": 0.3471710746568463,
+      "grad_norm": 0.1383199840784073,
+      "learning_rate": 0.0001,
+      "loss": 1.5509,
+      "step": 1037
+    },
+    {
+      "epoch": 0.3475058587211249,
+      "grad_norm": 0.12311188876628876,
+      "learning_rate": 0.0001,
+      "loss": 1.5429,
+      "step": 1038
+    },
+    {
+      "epoch": 0.34784064278540344,
+      "grad_norm": 0.12368562817573547,
+      "learning_rate": 0.0001,
+      "loss": 1.6099,
+      "step": 1039
+    },
+    {
+      "epoch": 0.34817542684968195,
+      "grad_norm": 0.13235348463058472,
+      "learning_rate": 0.0001,
+      "loss": 1.586,
+      "step": 1040
+    },
+    {
+      "epoch": 0.3485102109139605,
+      "grad_norm": 0.12543101608753204,
+      "learning_rate": 0.0001,
+      "loss": 1.5094,
+      "step": 1041
+    },
+    {
+      "epoch": 0.348844994978239,
+      "grad_norm": 0.12461157888174057,
+      "learning_rate": 0.0001,
+      "loss": 1.6067,
+      "step": 1042
+    },
+    {
+      "epoch": 0.34917977904251757,
+      "grad_norm": 0.12375465035438538,
+      "learning_rate": 0.0001,
+      "loss": 1.5953,
+      "step": 1043
+    },
+    {
+      "epoch": 0.34951456310679613,
+      "grad_norm": 0.13041523098945618,
+      "learning_rate": 0.0001,
+      "loss": 1.6088,
+      "step": 1044
+    },
+    {
+      "epoch": 0.34984934717107463,
+      "grad_norm": 0.12022354453802109,
+      "learning_rate": 0.0001,
+      "loss": 1.4805,
+      "step": 1045
+    },
+    {
+      "epoch": 0.3501841312353532,
+      "grad_norm": 0.1251700222492218,
+      "learning_rate": 0.0001,
+      "loss": 1.5457,
+      "step": 1046
+    },
+    {
+      "epoch": 0.35051891529963175,
+      "grad_norm": 0.12562930583953857,
+      "learning_rate": 0.0001,
+      "loss": 1.501,
+      "step": 1047
+    },
+    {
+      "epoch": 0.35085369936391025,
+      "grad_norm": 0.13178695738315582,
+      "learning_rate": 0.0001,
+      "loss": 1.6332,
+      "step": 1048
+    },
+    {
+      "epoch": 0.3511884834281888,
+      "grad_norm": 0.12346772104501724,
+      "learning_rate": 0.0001,
+      "loss": 1.5875,
+      "step": 1049
+    },
+    {
+      "epoch": 0.3515232674924674,
+      "grad_norm": 0.12000037729740143,
+      "learning_rate": 0.0001,
+      "loss": 1.5166,
+      "step": 1050
+    },
+    {
+      "epoch": 0.3518580515567459,
+      "grad_norm": 0.13240620493888855,
+      "learning_rate": 0.0001,
+      "loss": 1.5801,
+      "step": 1051
+    },
+    {
+      "epoch": 0.35219283562102444,
+      "grad_norm": 0.12688381969928741,
+      "learning_rate": 0.0001,
+      "loss": 1.5581,
+      "step": 1052
+    },
+    {
+      "epoch": 0.352527619685303,
+      "grad_norm": 0.12421749532222748,
+      "learning_rate": 0.0001,
+      "loss": 1.5626,
+      "step": 1053
+    },
+    {
+      "epoch": 0.3528624037495815,
+      "grad_norm": 0.12876258790493011,
+      "learning_rate": 0.0001,
+      "loss": 1.4921,
+      "step": 1054
+    },
+    {
+      "epoch": 0.35319718781386006,
+      "grad_norm": 0.13299116492271423,
+      "learning_rate": 0.0001,
+      "loss": 1.5828,
+      "step": 1055
+    },
+    {
+      "epoch": 0.3535319718781386,
+      "grad_norm": 0.12605415284633636,
+      "learning_rate": 0.0001,
+      "loss": 1.5963,
+      "step": 1056
+    },
+    {
+      "epoch": 0.3538667559424171,
+      "grad_norm": 0.13100145757198334,
+      "learning_rate": 0.0001,
+      "loss": 1.6035,
+      "step": 1057
+    },
+    {
+      "epoch": 0.3542015400066957,
+      "grad_norm": 0.12380324304103851,
+      "learning_rate": 0.0001,
+      "loss": 1.5784,
+      "step": 1058
+    },
+    {
+      "epoch": 0.35453632407097424,
+      "grad_norm": 0.1288285106420517,
+      "learning_rate": 0.0001,
+      "loss": 1.5454,
+      "step": 1059
+    },
+    {
+      "epoch": 0.35487110813525274,
+      "grad_norm": 0.12464431673288345,
+      "learning_rate": 0.0001,
+      "loss": 1.5622,
+      "step": 1060
+    },
+    {
+      "epoch": 0.3552058921995313,
+      "grad_norm": 0.12694504857063293,
+      "learning_rate": 0.0001,
+      "loss": 1.5361,
+      "step": 1061
+    },
+    {
+      "epoch": 0.35554067626380986,
+      "grad_norm": 0.12736117839813232,
+      "learning_rate": 0.0001,
+      "loss": 1.5931,
+      "step": 1062
+    },
+    {
+      "epoch": 0.35587546032808837,
+      "grad_norm": 0.12816745042800903,
+      "learning_rate": 0.0001,
+      "loss": 1.584,
+      "step": 1063
+    },
+    {
+      "epoch": 0.3562102443923669,
+      "grad_norm": 0.12096529453992844,
+      "learning_rate": 0.0001,
+      "loss": 1.4851,
+      "step": 1064
+    },
+    {
+      "epoch": 0.3565450284566455,
+      "grad_norm": 0.12956807017326355,
+      "learning_rate": 0.0001,
+      "loss": 1.5296,
+      "step": 1065
+    },
+    {
+      "epoch": 0.356879812520924,
+      "grad_norm": 0.12413816154003143,
+      "learning_rate": 0.0001,
+      "loss": 1.5634,
+      "step": 1066
+    },
+    {
+      "epoch": 0.35721459658520255,
+      "grad_norm": 0.13675865530967712,
+      "learning_rate": 0.0001,
+      "loss": 1.498,
+      "step": 1067
+    },
+    {
+      "epoch": 0.3575493806494811,
+      "grad_norm": 0.12694036960601807,
+      "learning_rate": 0.0001,
+      "loss": 1.6586,
+      "step": 1068
+    },
+    {
+      "epoch": 0.3578841647137596,
+      "grad_norm": 0.13280896842479706,
+      "learning_rate": 0.0001,
+      "loss": 1.4662,
+      "step": 1069
+    },
+    {
+      "epoch": 0.35821894877803817,
+      "grad_norm": 0.13775292038917542,
+      "learning_rate": 0.0001,
+      "loss": 1.5833,
+      "step": 1070
+    },
+    {
+      "epoch": 0.35855373284231673,
+      "grad_norm": 0.12691499292850494,
+      "learning_rate": 0.0001,
+      "loss": 1.6034,
+      "step": 1071
+    },
+    {
+      "epoch": 0.35888851690659523,
+      "grad_norm": 0.13247890770435333,
+      "learning_rate": 0.0001,
+      "loss": 1.5617,
+      "step": 1072
+    },
+    {
+      "epoch": 0.3592233009708738,
+      "grad_norm": 0.1524164378643036,
+      "learning_rate": 0.0001,
+      "loss": 1.7153,
+      "step": 1073
+    },
+    {
+      "epoch": 0.35955808503515235,
+      "grad_norm": 0.12795189023017883,
+      "learning_rate": 0.0001,
+      "loss": 1.5657,
+      "step": 1074
+    },
+    {
+      "epoch": 0.35989286909943086,
+      "grad_norm": 0.12827672064304352,
+      "learning_rate": 0.0001,
+      "loss": 1.4345,
+      "step": 1075
+    },
+    {
+      "epoch": 0.3602276531637094,
+      "grad_norm": 0.13488048315048218,
+      "learning_rate": 0.0001,
+      "loss": 1.5137,
+      "step": 1076
+    },
+    {
+      "epoch": 0.360562437227988,
+      "grad_norm": 0.11891927570104599,
+      "learning_rate": 0.0001,
+      "loss": 1.508,
+      "step": 1077
+    },
+    {
+      "epoch": 0.3608972212922665,
+      "grad_norm": 0.1263907551765442,
+      "learning_rate": 0.0001,
+      "loss": 1.5969,
+      "step": 1078
+    },
+    {
+      "epoch": 0.36123200535654504,
+      "grad_norm": 0.12749949097633362,
+      "learning_rate": 0.0001,
+      "loss": 1.5646,
+      "step": 1079
+    },
+    {
+      "epoch": 0.36156678942082354,
+      "grad_norm": 0.12221404910087585,
+      "learning_rate": 0.0001,
+      "loss": 1.5279,
+      "step": 1080
+    },
+    {
+      "epoch": 0.3619015734851021,
+      "grad_norm": 0.12473400682210922,
+      "learning_rate": 0.0001,
+      "loss": 1.507,
+      "step": 1081
+    },
+    {
+      "epoch": 0.36223635754938066,
+      "grad_norm": 0.13297304511070251,
+      "learning_rate": 0.0001,
+      "loss": 1.5636,
+      "step": 1082
+    },
+    {
+      "epoch": 0.36257114161365916,
+      "grad_norm": 0.1260288655757904,
+      "learning_rate": 0.0001,
+      "loss": 1.5429,
+      "step": 1083
+    },
+    {
+      "epoch": 0.3629059256779377,
+      "grad_norm": 0.12271251529455185,
+      "learning_rate": 0.0001,
+      "loss": 1.6139,
+      "step": 1084
+    },
+    {
+      "epoch": 0.3632407097422163,
+      "grad_norm": 0.13517338037490845,
+      "learning_rate": 0.0001,
+      "loss": 1.59,
+      "step": 1085
+    },
+    {
+      "epoch": 0.3635754938064948,
+      "grad_norm": 0.12335921078920364,
+      "learning_rate": 0.0001,
+      "loss": 1.5477,
+      "step": 1086
+    },
+    {
+      "epoch": 0.36391027787077335,
+      "grad_norm": 0.12416140735149384,
+      "learning_rate": 0.0001,
+      "loss": 1.5792,
+      "step": 1087
+    },
+    {
+      "epoch": 0.3642450619350519,
+      "grad_norm": 0.1330622136592865,
+      "learning_rate": 0.0001,
+      "loss": 1.6416,
+      "step": 1088
+    },
+    {
+      "epoch": 0.3645798459993304,
+      "grad_norm": 0.11882945895195007,
+      "learning_rate": 0.0001,
+      "loss": 1.5633,
+      "step": 1089
+    },
+    {
+      "epoch": 0.36491463006360897,
+      "grad_norm": 0.12056804448366165,
+      "learning_rate": 0.0001,
+      "loss": 1.5639,
+      "step": 1090
+    },
+    {
+      "epoch": 0.36524941412788753,
+      "grad_norm": 0.12773139774799347,
+      "learning_rate": 0.0001,
+      "loss": 1.5221,
+      "step": 1091
+    },
+    {
+      "epoch": 0.36558419819216603,
+      "grad_norm": 0.12159121781587601,
+      "learning_rate": 0.0001,
+      "loss": 1.5255,
+      "step": 1092
+    },
+    {
+      "epoch": 0.3659189822564446,
+      "grad_norm": 0.12454614788293839,
+      "learning_rate": 0.0001,
+      "loss": 1.5685,
+      "step": 1093
+    },
+    {
+      "epoch": 0.36625376632072315,
+      "grad_norm": 0.1252131462097168,
+      "learning_rate": 0.0001,
+      "loss": 1.5721,
+      "step": 1094
+    },
+    {
+      "epoch": 0.36658855038500165,
+      "grad_norm": 0.12228623777627945,
+      "learning_rate": 0.0001,
+      "loss": 1.5488,
+      "step": 1095
+    },
+    {
+      "epoch": 0.3669233344492802,
+      "grad_norm": 0.1220550686120987,
+      "learning_rate": 0.0001,
+      "loss": 1.524,
+      "step": 1096
+    },
+    {
+      "epoch": 0.3672581185135588,
+      "grad_norm": 0.12096890807151794,
+      "learning_rate": 0.0001,
+      "loss": 1.4846,
+      "step": 1097
+    },
+    {
+      "epoch": 0.3675929025778373,
+      "grad_norm": 0.12377587705850601,
+      "learning_rate": 0.0001,
+      "loss": 1.6305,
+      "step": 1098
+    },
+    {
+      "epoch": 0.36792768664211584,
+      "grad_norm": 0.12515562772750854,
+      "learning_rate": 0.0001,
+      "loss": 1.6078,
+      "step": 1099
+    },
+    {
+      "epoch": 0.3682624707063944,
+      "grad_norm": 0.12402921915054321,
+      "learning_rate": 0.0001,
+      "loss": 1.532,
+      "step": 1100
+    },
+    {
+      "epoch": 0.3685972547706729,
+      "grad_norm": 0.12373632192611694,
+      "learning_rate": 0.0001,
+      "loss": 1.512,
+      "step": 1101
+    },
+    {
+      "epoch": 0.36893203883495146,
+      "grad_norm": 0.12751725316047668,
+      "learning_rate": 0.0001,
+      "loss": 1.5799,
+      "step": 1102
+    },
+    {
+      "epoch": 0.36926682289923,
+      "grad_norm": 0.12221360951662064,
+      "learning_rate": 0.0001,
+      "loss": 1.4454,
+      "step": 1103
+    },
+    {
+      "epoch": 0.3696016069635085,
+      "grad_norm": 0.12299706041812897,
+      "learning_rate": 0.0001,
+      "loss": 1.5994,
+      "step": 1104
+    },
+    {
+      "epoch": 0.3699363910277871,
+      "grad_norm": 0.1294013112783432,
+      "learning_rate": 0.0001,
+      "loss": 1.6196,
+      "step": 1105
+    },
+    {
+      "epoch": 0.37027117509206564,
+      "grad_norm": 0.1240616887807846,
+      "learning_rate": 0.0001,
+      "loss": 1.5548,
+      "step": 1106
+    },
+    {
+      "epoch": 0.37060595915634414,
+      "grad_norm": 0.12403808534145355,
+      "learning_rate": 0.0001,
+      "loss": 1.6311,
+      "step": 1107
+    },
+    {
+      "epoch": 0.3709407432206227,
+      "grad_norm": 0.11872854828834534,
+      "learning_rate": 0.0001,
+      "loss": 1.4156,
+      "step": 1108
+    },
+    {
+      "epoch": 0.37127552728490126,
+      "grad_norm": 0.12752331793308258,
+      "learning_rate": 0.0001,
+      "loss": 1.6212,
+      "step": 1109
+    },
+    {
+      "epoch": 0.37161031134917977,
+      "grad_norm": 0.12329373508691788,
+      "learning_rate": 0.0001,
+      "loss": 1.5868,
+      "step": 1110
+    },
+    {
+      "epoch": 0.3719450954134583,
+      "grad_norm": 0.12340104579925537,
+      "learning_rate": 0.0001,
+      "loss": 1.5292,
+      "step": 1111
+    },
+    {
+      "epoch": 0.3722798794777369,
+      "grad_norm": 0.11669819802045822,
+      "learning_rate": 0.0001,
+      "loss": 1.5188,
+      "step": 1112
+    },
+    {
+      "epoch": 0.3726146635420154,
+      "grad_norm": 0.11677731573581696,
+      "learning_rate": 0.0001,
+      "loss": 1.5151,
+      "step": 1113
+    },
+    {
+      "epoch": 0.37294944760629395,
+      "grad_norm": 0.12206505239009857,
+      "learning_rate": 0.0001,
+      "loss": 1.6733,
+      "step": 1114
+    },
+    {
+      "epoch": 0.3732842316705725,
+      "grad_norm": 0.12234992533922195,
+      "learning_rate": 0.0001,
+      "loss": 1.5242,
+      "step": 1115
+    },
+    {
+      "epoch": 0.373619015734851,
+      "grad_norm": 0.12357670813798904,
+      "learning_rate": 0.0001,
+      "loss": 1.5432,
+      "step": 1116
+    },
+    {
+      "epoch": 0.37395379979912957,
+      "grad_norm": 0.12345674633979797,
+      "learning_rate": 0.0001,
+      "loss": 1.6483,
+      "step": 1117
+    },
+    {
+      "epoch": 0.3742885838634081,
+      "grad_norm": 0.1179901510477066,
+      "learning_rate": 0.0001,
+      "loss": 1.5899,
+      "step": 1118
+    },
+    {
+      "epoch": 0.37462336792768663,
+      "grad_norm": 0.12135247141122818,
+      "learning_rate": 0.0001,
+      "loss": 1.554,
+      "step": 1119
+    },
+    {
+      "epoch": 0.3749581519919652,
+      "grad_norm": 0.12836892902851105,
+      "learning_rate": 0.0001,
+      "loss": 1.6242,
+      "step": 1120
+    },
+    {
+      "epoch": 0.3752929360562437,
+      "grad_norm": 0.12851716578006744,
+      "learning_rate": 0.0001,
+      "loss": 1.6372,
+      "step": 1121
+    },
+    {
+      "epoch": 0.37562772012052226,
+      "grad_norm": 0.12096036225557327,
+      "learning_rate": 0.0001,
+      "loss": 1.5042,
+      "step": 1122
+    },
+    {
+      "epoch": 0.3759625041848008,
+      "grad_norm": 0.121758371591568,
+      "learning_rate": 0.0001,
+      "loss": 1.5561,
+      "step": 1123
+    },
+    {
+      "epoch": 0.3762972882490793,
+      "grad_norm": 0.12547370791435242,
+      "learning_rate": 0.0001,
+      "loss": 1.571,
+      "step": 1124
+    },
+    {
+      "epoch": 0.3766320723133579,
+      "grad_norm": 0.12488459795713425,
+      "learning_rate": 0.0001,
+      "loss": 1.6101,
+      "step": 1125
+    },
+    {
+      "epoch": 0.37696685637763644,
+      "grad_norm": 0.12440396845340729,
+      "learning_rate": 0.0001,
+      "loss": 1.4978,
+      "step": 1126
+    },
+    {
+      "epoch": 0.37730164044191494,
+      "grad_norm": 0.1293293535709381,
+      "learning_rate": 0.0001,
+      "loss": 1.6226,
+      "step": 1127
+    },
+    {
+      "epoch": 0.3776364245061935,
+      "grad_norm": 0.1270667314529419,
+      "learning_rate": 0.0001,
+      "loss": 1.5403,
+      "step": 1128
+    },
+    {
+      "epoch": 0.37797120857047206,
+      "grad_norm": 0.13023768365383148,
+      "learning_rate": 0.0001,
+      "loss": 1.6641,
+      "step": 1129
+    },
+    {
+      "epoch": 0.37830599263475057,
+      "grad_norm": 0.12713496387004852,
+      "learning_rate": 0.0001,
+      "loss": 1.5685,
+      "step": 1130
+    },
+    {
+      "epoch": 0.3786407766990291,
+      "grad_norm": 0.126458540558815,
+      "learning_rate": 0.0001,
+      "loss": 1.5624,
+      "step": 1131
+    },
+    {
+      "epoch": 0.3789755607633077,
+      "grad_norm": 0.12100820988416672,
+      "learning_rate": 0.0001,
+      "loss": 1.5158,
+      "step": 1132
+    },
+    {
+      "epoch": 0.3793103448275862,
+      "grad_norm": 0.13373976945877075,
+      "learning_rate": 0.0001,
+      "loss": 1.5151,
+      "step": 1133
+    },
+    {
+      "epoch": 0.37964512889186475,
+      "grad_norm": 0.12730540335178375,
+      "learning_rate": 0.0001,
+      "loss": 1.5701,
+      "step": 1134
+    },
+    {
+      "epoch": 0.3799799129561433,
+      "grad_norm": 0.13641048967838287,
+      "learning_rate": 0.0001,
+      "loss": 1.5144,
+      "step": 1135
+    },
+    {
+      "epoch": 0.3803146970204218,
+      "grad_norm": 0.13271461427211761,
+      "learning_rate": 0.0001,
+      "loss": 1.5884,
+      "step": 1136
+    },
+    {
+      "epoch": 0.38064948108470037,
+      "grad_norm": 0.12385160475969315,
+      "learning_rate": 0.0001,
+      "loss": 1.5374,
+      "step": 1137
+    },
+    {
+      "epoch": 0.38098426514897893,
+      "grad_norm": 0.12949350476264954,
+      "learning_rate": 0.0001,
+      "loss": 1.546,
+      "step": 1138
+    },
+    {
+      "epoch": 0.38131904921325743,
+      "grad_norm": 0.135132297873497,
+      "learning_rate": 0.0001,
+      "loss": 1.5913,
+      "step": 1139
+    },
+    {
+      "epoch": 0.381653833277536,
+      "grad_norm": 0.11533955484628677,
+      "learning_rate": 0.0001,
+      "loss": 1.3968,
+      "step": 1140
+    },
+    {
+      "epoch": 0.38198861734181455,
+      "grad_norm": 0.13532719016075134,
+      "learning_rate": 0.0001,
+      "loss": 1.5534,
+      "step": 1141
+    },
+    {
+      "epoch": 0.38232340140609306,
+      "grad_norm": 0.14101184904575348,
+      "learning_rate": 0.0001,
+      "loss": 1.557,
+      "step": 1142
+    },
+    {
+      "epoch": 0.3826581854703716,
+      "grad_norm": 0.12038899213075638,
+      "learning_rate": 0.0001,
+      "loss": 1.4831,
+      "step": 1143
+    },
+    {
+      "epoch": 0.3829929695346502,
+      "grad_norm": 0.13053514063358307,
+      "learning_rate": 0.0001,
+      "loss": 1.5882,
+      "step": 1144
+    },
+    {
+      "epoch": 0.3833277535989287,
+      "grad_norm": 0.12372793257236481,
+      "learning_rate": 0.0001,
+      "loss": 1.6047,
+      "step": 1145
+    },
+    {
+      "epoch": 0.38366253766320724,
+      "grad_norm": 0.12823140621185303,
+      "learning_rate": 0.0001,
+      "loss": 1.6126,
+      "step": 1146
+    },
+    {
+      "epoch": 0.3839973217274858,
+      "grad_norm": 0.12058600783348083,
+      "learning_rate": 0.0001,
+      "loss": 1.4713,
+      "step": 1147
+    },
+    {
+      "epoch": 0.3843321057917643,
+      "grad_norm": 0.12674620747566223,
+      "learning_rate": 0.0001,
+      "loss": 1.6126,
+      "step": 1148
+    },
+    {
+      "epoch": 0.38466688985604286,
+      "grad_norm": 0.1214526891708374,
+      "learning_rate": 0.0001,
+      "loss": 1.6317,
+      "step": 1149
+    },
+    {
+      "epoch": 0.3850016739203214,
+      "grad_norm": 0.12831653654575348,
+      "learning_rate": 0.0001,
+      "loss": 1.5479,
+      "step": 1150
+    },
+    {
+      "epoch": 0.3853364579845999,
+      "grad_norm": 0.12079459428787231,
+      "learning_rate": 0.0001,
+      "loss": 1.5544,
+      "step": 1151
+    },
+    {
+      "epoch": 0.3856712420488785,
+      "grad_norm": 0.12021779268980026,
+      "learning_rate": 0.0001,
+      "loss": 1.5536,
+      "step": 1152
+    },
+    {
+      "epoch": 0.38600602611315704,
+      "grad_norm": 0.13052217662334442,
+      "learning_rate": 0.0001,
+      "loss": 1.5482,
+      "step": 1153
+    },
+    {
+      "epoch": 0.38634081017743555,
+      "grad_norm": 0.12613235414028168,
+      "learning_rate": 0.0001,
+      "loss": 1.6056,
+      "step": 1154
+    },
+    {
+      "epoch": 0.3866755942417141,
+      "grad_norm": 0.12751324474811554,
+      "learning_rate": 0.0001,
+      "loss": 1.5513,
+      "step": 1155
+    },
+    {
+      "epoch": 0.3870103783059926,
+      "grad_norm": 0.11987000703811646,
+      "learning_rate": 0.0001,
+      "loss": 1.4836,
+      "step": 1156
+    },
+    {
+      "epoch": 0.38734516237027117,
+      "grad_norm": 0.13999362289905548,
+      "learning_rate": 0.0001,
+      "loss": 1.6763,
+      "step": 1157
+    },
+    {
+      "epoch": 0.3876799464345497,
+      "grad_norm": 0.128611221909523,
+      "learning_rate": 0.0001,
+      "loss": 1.6281,
+      "step": 1158
+    },
+    {
+      "epoch": 0.38801473049882823,
+      "grad_norm": 0.1292606145143509,
+      "learning_rate": 0.0001,
+      "loss": 1.6846,
+      "step": 1159
+    },
+    {
+      "epoch": 0.3883495145631068,
+      "grad_norm": 0.13090923428535461,
+      "learning_rate": 0.0001,
+      "loss": 1.628,
+      "step": 1160
+    },
+    {
+      "epoch": 0.38868429862738535,
+      "grad_norm": 0.12356492131948471,
+      "learning_rate": 0.0001,
+      "loss": 1.5158,
+      "step": 1161
+    },
+    {
+      "epoch": 0.38901908269166385,
+      "grad_norm": 0.12005447596311569,
+      "learning_rate": 0.0001,
+      "loss": 1.62,
+      "step": 1162
+    },
+    {
+      "epoch": 0.3893538667559424,
+      "grad_norm": 0.12113460153341293,
+      "learning_rate": 0.0001,
+      "loss": 1.4954,
+      "step": 1163
+    },
+    {
+      "epoch": 0.38968865082022097,
+      "grad_norm": 0.11953802406787872,
+      "learning_rate": 0.0001,
+      "loss": 1.4891,
+      "step": 1164
+    },
+    {
+      "epoch": 0.3900234348844995,
+      "grad_norm": 0.1292644739151001,
+      "learning_rate": 0.0001,
+      "loss": 1.555,
+      "step": 1165
+    },
+    {
+      "epoch": 0.39035821894877804,
+      "grad_norm": 0.12345704436302185,
+      "learning_rate": 0.0001,
+      "loss": 1.4939,
+      "step": 1166
+    },
+    {
+      "epoch": 0.3906930030130566,
+      "grad_norm": 0.12334253638982773,
+      "learning_rate": 0.0001,
+      "loss": 1.6058,
+      "step": 1167
+    },
+    {
+      "epoch": 0.3910277870773351,
+      "grad_norm": 0.13044217228889465,
+      "learning_rate": 0.0001,
+      "loss": 1.5349,
+      "step": 1168
+    },
+    {
+      "epoch": 0.39136257114161366,
+      "grad_norm": 0.12309286743402481,
+      "learning_rate": 0.0001,
+      "loss": 1.5007,
+      "step": 1169
+    },
+    {
+      "epoch": 0.3916973552058922,
+      "grad_norm": 0.12565681338310242,
+      "learning_rate": 0.0001,
+      "loss": 1.5172,
+      "step": 1170
+    },
+    {
+      "epoch": 0.3920321392701707,
+      "grad_norm": 0.13335129618644714,
+      "learning_rate": 0.0001,
+      "loss": 1.5666,
+      "step": 1171
+    },
+    {
+      "epoch": 0.3923669233344493,
+      "grad_norm": 0.12664766609668732,
+      "learning_rate": 0.0001,
+      "loss": 1.5471,
+      "step": 1172
+    },
+    {
+      "epoch": 0.39270170739872784,
+      "grad_norm": 0.12703973054885864,
+      "learning_rate": 0.0001,
+      "loss": 1.545,
+      "step": 1173
+    },
+    {
+      "epoch": 0.39303649146300634,
+      "grad_norm": 0.12242884933948517,
+      "learning_rate": 0.0001,
+      "loss": 1.4768,
+      "step": 1174
+    },
+    {
+      "epoch": 0.3933712755272849,
+      "grad_norm": 0.13055263459682465,
+      "learning_rate": 0.0001,
+      "loss": 1.4782,
+      "step": 1175
+    },
+    {
+      "epoch": 0.39370605959156346,
+      "grad_norm": 0.13161849975585938,
+      "learning_rate": 0.0001,
+      "loss": 1.621,
+      "step": 1176
+    },
+    {
+      "epoch": 0.39404084365584197,
+      "grad_norm": 0.1257203370332718,
+      "learning_rate": 0.0001,
+      "loss": 1.5655,
+      "step": 1177
+    },
+    {
+      "epoch": 0.3943756277201205,
+      "grad_norm": 0.14164592325687408,
+      "learning_rate": 0.0001,
+      "loss": 1.4884,
+      "step": 1178
+    },
+    {
+      "epoch": 0.3947104117843991,
+      "grad_norm": 0.12696050107479095,
+      "learning_rate": 0.0001,
+      "loss": 1.5829,
+      "step": 1179
+    },
+    {
+      "epoch": 0.3950451958486776,
+      "grad_norm": 0.12652398645877838,
+      "learning_rate": 0.0001,
+      "loss": 1.6345,
+      "step": 1180
+    },
+    {
+      "epoch": 0.39537997991295615,
+      "grad_norm": 0.12333660572767258,
+      "learning_rate": 0.0001,
+      "loss": 1.5375,
+      "step": 1181
+    },
+    {
+      "epoch": 0.3957147639772347,
+      "grad_norm": 0.13108794391155243,
+      "learning_rate": 0.0001,
+      "loss": 1.6441,
+      "step": 1182
+    },
+    {
+      "epoch": 0.3960495480415132,
+      "grad_norm": 0.13195887207984924,
+      "learning_rate": 0.0001,
+      "loss": 1.5939,
+      "step": 1183
+    },
+    {
+      "epoch": 0.39638433210579177,
+      "grad_norm": 0.12931646406650543,
+      "learning_rate": 0.0001,
+      "loss": 1.5317,
+      "step": 1184
+    },
+    {
+      "epoch": 0.39671911617007033,
+      "grad_norm": 0.12439566105604172,
+      "learning_rate": 0.0001,
+      "loss": 1.5391,
+      "step": 1185
+    },
+    {
+      "epoch": 0.39705390023434883,
+      "grad_norm": 0.12557551264762878,
+      "learning_rate": 0.0001,
+      "loss": 1.5723,
+      "step": 1186
+    },
+    {
+      "epoch": 0.3973886842986274,
+      "grad_norm": 0.13013330101966858,
+      "learning_rate": 0.0001,
+      "loss": 1.4812,
+      "step": 1187
+    },
+    {
+      "epoch": 0.39772346836290595,
+      "grad_norm": 0.12955336272716522,
+      "learning_rate": 0.0001,
+      "loss": 1.5799,
+      "step": 1188
+    },
+    {
+      "epoch": 0.39805825242718446,
+      "grad_norm": 0.1347295343875885,
+      "learning_rate": 0.0001,
+      "loss": 1.6634,
+      "step": 1189
+    },
+    {
+      "epoch": 0.398393036491463,
+      "grad_norm": 0.13187319040298462,
+      "learning_rate": 0.0001,
+      "loss": 1.5146,
+      "step": 1190
+    },
+    {
+      "epoch": 0.3987278205557416,
+      "grad_norm": 0.13010048866271973,
+      "learning_rate": 0.0001,
+      "loss": 1.5003,
+      "step": 1191
+    },
+    {
+      "epoch": 0.3990626046200201,
+      "grad_norm": 0.12330204248428345,
+      "learning_rate": 0.0001,
+      "loss": 1.5765,
+      "step": 1192
+    },
+    {
+      "epoch": 0.39939738868429864,
+      "grad_norm": 0.1346241533756256,
+      "learning_rate": 0.0001,
+      "loss": 1.5979,
+      "step": 1193
+    },
+    {
+      "epoch": 0.39973217274857714,
+      "grad_norm": 0.13725797832012177,
+      "learning_rate": 0.0001,
+      "loss": 1.5813,
+      "step": 1194
+    },
+    {
+      "epoch": 0.4000669568128557,
+      "grad_norm": 0.12039465457201004,
+      "learning_rate": 0.0001,
+      "loss": 1.4363,
+      "step": 1195
+    },
+    {
+      "epoch": 0.40040174087713426,
+      "grad_norm": 0.1276928186416626,
+      "learning_rate": 0.0001,
+      "loss": 1.6575,
+      "step": 1196
+    },
+    {
+      "epoch": 0.40073652494141276,
+      "grad_norm": 0.12903235852718353,
+      "learning_rate": 0.0001,
+      "loss": 1.6059,
+      "step": 1197
+    },
+    {
+      "epoch": 0.4010713090056913,
+      "grad_norm": 0.12678353488445282,
+      "learning_rate": 0.0001,
+      "loss": 1.5624,
+      "step": 1198
+    },
+    {
+      "epoch": 0.4014060930699699,
+      "grad_norm": 0.12884308397769928,
+      "learning_rate": 0.0001,
+      "loss": 1.5995,
+      "step": 1199
+    },
+    {
+      "epoch": 0.4017408771342484,
+      "grad_norm": 0.11986846476793289,
+      "learning_rate": 0.0001,
+      "loss": 1.4767,
+      "step": 1200
+    },
+    {
+      "epoch": 0.40207566119852695,
+      "grad_norm": 0.12227410078048706,
+      "learning_rate": 0.0001,
+      "loss": 1.5056,
+      "step": 1201
+    },
+    {
+      "epoch": 0.4024104452628055,
+      "grad_norm": 0.12593914568424225,
+      "learning_rate": 0.0001,
+      "loss": 1.5836,
+      "step": 1202
+    },
+    {
+      "epoch": 0.402745229327084,
+      "grad_norm": 0.12477041780948639,
+      "learning_rate": 0.0001,
+      "loss": 1.5745,
+      "step": 1203
+    },
+    {
+      "epoch": 0.40308001339136257,
+      "grad_norm": 0.1216067373752594,
+      "learning_rate": 0.0001,
+      "loss": 1.5824,
+      "step": 1204
+    },
+    {
+      "epoch": 0.4034147974556411,
+      "grad_norm": 0.13550971448421478,
+      "learning_rate": 0.0001,
+      "loss": 1.6635,
+      "step": 1205
+    },
+    {
+      "epoch": 0.40374958151991963,
+      "grad_norm": 0.12963739037513733,
+      "learning_rate": 0.0001,
+      "loss": 1.6586,
+      "step": 1206
+    },
+    {
+      "epoch": 0.4040843655841982,
+      "grad_norm": 0.11887506395578384,
+      "learning_rate": 0.0001,
+      "loss": 1.4933,
+      "step": 1207
+    },
+    {
+      "epoch": 0.40441914964847675,
+      "grad_norm": 0.13262464106082916,
+      "learning_rate": 0.0001,
+      "loss": 1.5759,
+      "step": 1208
+    },
+    {
+      "epoch": 0.40475393371275525,
+      "grad_norm": 0.13952501118183136,
+      "learning_rate": 0.0001,
+      "loss": 1.6918,
+      "step": 1209
+    },
+    {
+      "epoch": 0.4050887177770338,
+      "grad_norm": 0.13401460647583008,
+      "learning_rate": 0.0001,
+      "loss": 1.5102,
+      "step": 1210
+    },
+    {
+      "epoch": 0.4054235018413124,
+      "grad_norm": 0.14476630091667175,
+      "learning_rate": 0.0001,
+      "loss": 1.6817,
+      "step": 1211
+    },
+    {
+      "epoch": 0.4057582859055909,
+      "grad_norm": 0.1285640001296997,
+      "learning_rate": 0.0001,
+      "loss": 1.653,
+      "step": 1212
+    },
+    {
+      "epoch": 0.40609306996986944,
+      "grad_norm": 0.13845203816890717,
+      "learning_rate": 0.0001,
+      "loss": 1.5996,
+      "step": 1213
+    },
+    {
+      "epoch": 0.406427854034148,
+      "grad_norm": 0.13416174054145813,
+      "learning_rate": 0.0001,
+      "loss": 1.6222,
+      "step": 1214
+    },
+    {
+      "epoch": 0.4067626380984265,
+      "grad_norm": 0.1267634481191635,
+      "learning_rate": 0.0001,
+      "loss": 1.5257,
+      "step": 1215
+    },
+    {
+      "epoch": 0.40709742216270506,
+      "grad_norm": 0.13453447818756104,
+      "learning_rate": 0.0001,
+      "loss": 1.5745,
+      "step": 1216
+    },
+    {
+      "epoch": 0.4074322062269836,
+      "grad_norm": 0.12069771438837051,
+      "learning_rate": 0.0001,
+      "loss": 1.5516,
+      "step": 1217
+    },
+    {
+      "epoch": 0.4077669902912621,
+      "grad_norm": 0.12483450770378113,
+      "learning_rate": 0.0001,
+      "loss": 1.5899,
+      "step": 1218
+    },
+    {
+      "epoch": 0.4081017743555407,
+      "grad_norm": 0.14123085141181946,
+      "learning_rate": 0.0001,
+      "loss": 1.6334,
+      "step": 1219
+    },
+    {
+      "epoch": 0.40843655841981924,
+      "grad_norm": 0.12844936549663544,
+      "learning_rate": 0.0001,
+      "loss": 1.4936,
+      "step": 1220
+    },
+    {
+      "epoch": 0.40877134248409774,
+      "grad_norm": 0.13094481825828552,
+      "learning_rate": 0.0001,
+      "loss": 1.6554,
+      "step": 1221
+    },
+    {
+      "epoch": 0.4091061265483763,
+      "grad_norm": 0.12563113868236542,
+      "learning_rate": 0.0001,
+      "loss": 1.4708,
+      "step": 1222
+    },
+    {
+      "epoch": 0.40944091061265486,
+      "grad_norm": 0.12495769560337067,
+      "learning_rate": 0.0001,
+      "loss": 1.5012,
+      "step": 1223
+    },
+    {
+      "epoch": 0.40977569467693337,
+      "grad_norm": 0.12314360588788986,
+      "learning_rate": 0.0001,
+      "loss": 1.5769,
+      "step": 1224
+    },
+    {
+      "epoch": 0.4101104787412119,
+      "grad_norm": 0.1389753818511963,
+      "learning_rate": 0.0001,
+      "loss": 1.5978,
+      "step": 1225
+    },
+    {
+      "epoch": 0.4104452628054905,
+      "grad_norm": 0.12703324854373932,
+      "learning_rate": 0.0001,
+      "loss": 1.5349,
+      "step": 1226
+    },
+    {
+      "epoch": 0.410780046869769,
+      "grad_norm": 0.11995337903499603,
+      "learning_rate": 0.0001,
+      "loss": 1.5307,
+      "step": 1227
+    },
+    {
+      "epoch": 0.41111483093404755,
+      "grad_norm": 0.1330454796552658,
+      "learning_rate": 0.0001,
+      "loss": 1.6277,
+      "step": 1228
+    },
+    {
+      "epoch": 0.41144961499832605,
+      "grad_norm": 0.12632183730602264,
+      "learning_rate": 0.0001,
+      "loss": 1.507,
+      "step": 1229
+    },
+    {
+      "epoch": 0.4117843990626046,
+      "grad_norm": 0.13255640864372253,
+      "learning_rate": 0.0001,
+      "loss": 1.5797,
+      "step": 1230
+    },
+    {
+      "epoch": 0.41211918312688317,
+      "grad_norm": 0.13822025060653687,
+      "learning_rate": 0.0001,
+      "loss": 1.5945,
+      "step": 1231
+    },
+    {
+      "epoch": 0.4124539671911617,
+      "grad_norm": 0.1303391307592392,
+      "learning_rate": 0.0001,
+      "loss": 1.5928,
+      "step": 1232
+    },
+    {
+      "epoch": 0.41278875125544023,
+      "grad_norm": 0.12309371680021286,
+      "learning_rate": 0.0001,
+      "loss": 1.4794,
+      "step": 1233
+    },
+    {
+      "epoch": 0.4131235353197188,
+      "grad_norm": 0.12375032901763916,
+      "learning_rate": 0.0001,
+      "loss": 1.5133,
+      "step": 1234
+    },
+    {
+      "epoch": 0.4134583193839973,
+      "grad_norm": 0.13613499701023102,
+      "learning_rate": 0.0001,
+      "loss": 1.621,
+      "step": 1235
+    },
+    {
+      "epoch": 0.41379310344827586,
+      "grad_norm": 0.13198764622211456,
+      "learning_rate": 0.0001,
+      "loss": 1.5762,
+      "step": 1236
+    },
+    {
+      "epoch": 0.4141278875125544,
+      "grad_norm": 0.1294814646244049,
+      "learning_rate": 0.0001,
+      "loss": 1.5836,
+      "step": 1237
+    },
+    {
+      "epoch": 0.4144626715768329,
+      "grad_norm": 0.12597797811031342,
+      "learning_rate": 0.0001,
+      "loss": 1.5988,
+      "step": 1238
+    },
+    {
+      "epoch": 0.4147974556411115,
+      "grad_norm": 0.12371232360601425,
+      "learning_rate": 0.0001,
+      "loss": 1.5432,
+      "step": 1239
+    },
+    {
+      "epoch": 0.41513223970539004,
+      "grad_norm": 0.12919354438781738,
+      "learning_rate": 0.0001,
+      "loss": 1.5507,
+      "step": 1240
+    },
+    {
+      "epoch": 0.41546702376966854,
+      "grad_norm": 0.12919418513774872,
+      "learning_rate": 0.0001,
+      "loss": 1.7431,
+      "step": 1241
+    },
+    {
+      "epoch": 0.4158018078339471,
+      "grad_norm": 0.12314452975988388,
+      "learning_rate": 0.0001,
+      "loss": 1.5407,
+      "step": 1242
+    },
+    {
+      "epoch": 0.41613659189822566,
+      "grad_norm": 0.1360636204481125,
+      "learning_rate": 0.0001,
+      "loss": 1.5872,
+      "step": 1243
+    },
+    {
+      "epoch": 0.41647137596250416,
+      "grad_norm": 0.12739785015583038,
+      "learning_rate": 0.0001,
+      "loss": 1.4998,
+      "step": 1244
+    },
+    {
+      "epoch": 0.4168061600267827,
+      "grad_norm": 0.12558461725711823,
+      "learning_rate": 0.0001,
+      "loss": 1.6422,
+      "step": 1245
+    },
+    {
+      "epoch": 0.4171409440910613,
+      "grad_norm": 0.130743145942688,
+      "learning_rate": 0.0001,
+      "loss": 1.6537,
+      "step": 1246
+    },
+    {
+      "epoch": 0.4174757281553398,
+      "grad_norm": 0.12714166939258575,
+      "learning_rate": 0.0001,
+      "loss": 1.4309,
+      "step": 1247
+    },
+    {
+      "epoch": 0.41781051221961835,
+      "grad_norm": 0.12849892675876617,
+      "learning_rate": 0.0001,
+      "loss": 1.514,
+      "step": 1248
+    },
+    {
+      "epoch": 0.4181452962838969,
+      "grad_norm": 0.1366477757692337,
+      "learning_rate": 0.0001,
+      "loss": 1.6397,
+      "step": 1249
+    },
+    {
+      "epoch": 0.4184800803481754,
+      "grad_norm": 0.1324029415845871,
+      "learning_rate": 0.0001,
+      "loss": 1.5647,
+      "step": 1250
+    },
+    {
+      "epoch": 0.41881486441245397,
+      "grad_norm": 0.1272830069065094,
+      "learning_rate": 0.0001,
+      "loss": 1.633,
+      "step": 1251
+    },
+    {
+      "epoch": 0.41914964847673253,
+      "grad_norm": 0.12891270220279694,
+      "learning_rate": 0.0001,
+      "loss": 1.5571,
+      "step": 1252
+    },
+    {
+      "epoch": 0.41948443254101103,
+      "grad_norm": 0.1334099918603897,
+      "learning_rate": 0.0001,
+      "loss": 1.4905,
+      "step": 1253
+    },
+    {
+      "epoch": 0.4198192166052896,
+      "grad_norm": 0.12439723312854767,
+      "learning_rate": 0.0001,
+      "loss": 1.5859,
+      "step": 1254
+    },
+    {
+      "epoch": 0.42015400066956815,
+      "grad_norm": 0.13870543241500854,
+      "learning_rate": 0.0001,
+      "loss": 1.6226,
+      "step": 1255
+    },
+    {
+      "epoch": 0.42048878473384665,
+      "grad_norm": 0.13232079148292542,
+      "learning_rate": 0.0001,
+      "loss": 1.6566,
+      "step": 1256
+    },
+    {
+      "epoch": 0.4208235687981252,
+      "grad_norm": 0.12575885653495789,
+      "learning_rate": 0.0001,
+      "loss": 1.5629,
+      "step": 1257
+    },
+    {
+      "epoch": 0.4211583528624038,
+      "grad_norm": 0.12995895743370056,
+      "learning_rate": 0.0001,
+      "loss": 1.5703,
+      "step": 1258
+    },
+    {
+      "epoch": 0.4214931369266823,
+      "grad_norm": 0.12801054120063782,
+      "learning_rate": 0.0001,
+      "loss": 1.6326,
+      "step": 1259
+    },
+    {
+      "epoch": 0.42182792099096084,
+      "grad_norm": 0.12584693729877472,
+      "learning_rate": 0.0001,
+      "loss": 1.6329,
+      "step": 1260
+    },
+    {
+      "epoch": 0.4221627050552394,
+      "grad_norm": 0.13142889738082886,
+      "learning_rate": 0.0001,
+      "loss": 1.7081,
+      "step": 1261
+    },
+    {
+      "epoch": 0.4224974891195179,
+      "grad_norm": 0.12793239951133728,
+      "learning_rate": 0.0001,
+      "loss": 1.6032,
+      "step": 1262
+    },
+    {
+      "epoch": 0.42283227318379646,
+      "grad_norm": 0.12368165701627731,
+      "learning_rate": 0.0001,
+      "loss": 1.5468,
+      "step": 1263
+    },
+    {
+      "epoch": 0.423167057248075,
+      "grad_norm": 0.13081911206245422,
+      "learning_rate": 0.0001,
+      "loss": 1.6175,
+      "step": 1264
+    },
+    {
+      "epoch": 0.4235018413123535,
+      "grad_norm": 0.12801037728786469,
+      "learning_rate": 0.0001,
+      "loss": 1.537,
+      "step": 1265
+    },
+    {
+      "epoch": 0.4238366253766321,
+      "grad_norm": 0.1274782121181488,
+      "learning_rate": 0.0001,
+      "loss": 1.5277,
+      "step": 1266
+    },
+    {
+      "epoch": 0.4241714094409106,
+      "grad_norm": 0.1194332018494606,
+      "learning_rate": 0.0001,
+      "loss": 1.496,
+      "step": 1267
+    },
+    {
+      "epoch": 0.42450619350518914,
+      "grad_norm": 0.13174927234649658,
+      "learning_rate": 0.0001,
+      "loss": 1.5975,
+      "step": 1268
+    },
+    {
+      "epoch": 0.4248409775694677,
+      "grad_norm": 0.1254556030035019,
+      "learning_rate": 0.0001,
+      "loss": 1.6119,
+      "step": 1269
+    },
+    {
+      "epoch": 0.4251757616337462,
+      "grad_norm": 0.13203318417072296,
+      "learning_rate": 0.0001,
+      "loss": 1.5564,
+      "step": 1270
+    },
+    {
+      "epoch": 0.42551054569802477,
+      "grad_norm": 0.12941622734069824,
+      "learning_rate": 0.0001,
+      "loss": 1.6285,
+      "step": 1271
+    },
+    {
+      "epoch": 0.4258453297623033,
+      "grad_norm": 0.12527894973754883,
+      "learning_rate": 0.0001,
+      "loss": 1.5703,
+      "step": 1272
+    },
+    {
+      "epoch": 0.42618011382658183,
+      "grad_norm": 0.12617714703083038,
+      "learning_rate": 0.0001,
+      "loss": 1.6523,
+      "step": 1273
+    },
+    {
+      "epoch": 0.4265148978908604,
+      "grad_norm": 0.12326870858669281,
+      "learning_rate": 0.0001,
+      "loss": 1.5533,
+      "step": 1274
+    },
+    {
+      "epoch": 0.42684968195513895,
+      "grad_norm": 0.1295124888420105,
+      "learning_rate": 0.0001,
+      "loss": 1.5587,
+      "step": 1275
+    },
+    {
+      "epoch": 0.42718446601941745,
+      "grad_norm": 0.12248773872852325,
+      "learning_rate": 0.0001,
+      "loss": 1.5762,
+      "step": 1276
+    },
+    {
+      "epoch": 0.427519250083696,
+      "grad_norm": 0.12932232022285461,
+      "learning_rate": 0.0001,
+      "loss": 1.6162,
+      "step": 1277
+    },
+    {
+      "epoch": 0.42785403414797457,
+      "grad_norm": 0.1178537905216217,
+      "learning_rate": 0.0001,
+      "loss": 1.472,
+      "step": 1278
+    },
+    {
+      "epoch": 0.4281888182122531,
+      "grad_norm": 0.1269647628068924,
+      "learning_rate": 0.0001,
+      "loss": 1.5551,
+      "step": 1279
+    },
+    {
+      "epoch": 0.42852360227653163,
+      "grad_norm": 0.120000459253788,
+      "learning_rate": 0.0001,
+      "loss": 1.509,
+      "step": 1280
+    },
+    {
+      "epoch": 0.4288583863408102,
+      "grad_norm": 0.12708665430545807,
+      "learning_rate": 0.0001,
+      "loss": 1.5293,
+      "step": 1281
+    },
+    {
+      "epoch": 0.4291931704050887,
+      "grad_norm": 0.13209426403045654,
+      "learning_rate": 0.0001,
+      "loss": 1.6311,
+      "step": 1282
+    },
+    {
+      "epoch": 0.42952795446936726,
+      "grad_norm": 0.1305491328239441,
+      "learning_rate": 0.0001,
+      "loss": 1.5505,
+      "step": 1283
+    },
+    {
+      "epoch": 0.4298627385336458,
+      "grad_norm": 0.1237809956073761,
+      "learning_rate": 0.0001,
+      "loss": 1.5457,
+      "step": 1284
+    },
+    {
+      "epoch": 0.4301975225979243,
+      "grad_norm": 0.13375982642173767,
+      "learning_rate": 0.0001,
+      "loss": 1.5321,
+      "step": 1285
+    },
+    {
+      "epoch": 0.4305323066622029,
+      "grad_norm": 0.13597902655601501,
+      "learning_rate": 0.0001,
+      "loss": 1.6229,
+      "step": 1286
+    },
+    {
+      "epoch": 0.43086709072648144,
+      "grad_norm": 0.12488207966089249,
+      "learning_rate": 0.0001,
+      "loss": 1.5231,
+      "step": 1287
+    },
+    {
+      "epoch": 0.43120187479075994,
+      "grad_norm": 0.12950995564460754,
+      "learning_rate": 0.0001,
+      "loss": 1.7162,
+      "step": 1288
+    },
+    {
+      "epoch": 0.4315366588550385,
+      "grad_norm": 0.12734153866767883,
+      "learning_rate": 0.0001,
+      "loss": 1.5735,
+      "step": 1289
+    },
+    {
+      "epoch": 0.43187144291931706,
+      "grad_norm": 0.13684290647506714,
+      "learning_rate": 0.0001,
+      "loss": 1.5866,
+      "step": 1290
+    },
+    {
+      "epoch": 0.43220622698359557,
+      "grad_norm": 0.12665408849716187,
+      "learning_rate": 0.0001,
+      "loss": 1.5236,
+      "step": 1291
+    },
+    {
+      "epoch": 0.4325410110478741,
+      "grad_norm": 0.12092933058738708,
+      "learning_rate": 0.0001,
+      "loss": 1.4859,
+      "step": 1292
+    },
+    {
+      "epoch": 0.4328757951121527,
+      "grad_norm": 0.14012545347213745,
+      "learning_rate": 0.0001,
+      "loss": 1.6158,
+      "step": 1293
+    },
+    {
+      "epoch": 0.4332105791764312,
+      "grad_norm": 0.12820059061050415,
+      "learning_rate": 0.0001,
+      "loss": 1.5108,
+      "step": 1294
+    },
+    {
+      "epoch": 0.43354536324070975,
+      "grad_norm": 0.13247036933898926,
+      "learning_rate": 0.0001,
+      "loss": 1.6031,
+      "step": 1295
+    },
+    {
+      "epoch": 0.4338801473049883,
+      "grad_norm": 0.12412893772125244,
+      "learning_rate": 0.0001,
+      "loss": 1.5829,
+      "step": 1296
+    },
+    {
+      "epoch": 0.4342149313692668,
+      "grad_norm": 0.12657597661018372,
+      "learning_rate": 0.0001,
+      "loss": 1.5139,
+      "step": 1297
+    },
+    {
+      "epoch": 0.43454971543354537,
+      "grad_norm": 0.13494263589382172,
+      "learning_rate": 0.0001,
+      "loss": 1.6264,
+      "step": 1298
+    },
+    {
+      "epoch": 0.43488449949782393,
+      "grad_norm": 0.12553179264068604,
+      "learning_rate": 0.0001,
+      "loss": 1.5587,
+      "step": 1299
+    },
+    {
+      "epoch": 0.43521928356210243,
+      "grad_norm": 0.12029055505990982,
+      "learning_rate": 0.0001,
+      "loss": 1.5177,
+      "step": 1300
+    },
+    {
+      "epoch": 0.435554067626381,
+      "grad_norm": 0.12742608785629272,
+      "learning_rate": 0.0001,
+      "loss": 1.6345,
+      "step": 1301
+    },
+    {
+      "epoch": 0.43588885169065955,
+      "grad_norm": 0.12749677896499634,
+      "learning_rate": 0.0001,
+      "loss": 1.5183,
+      "step": 1302
+    },
+    {
+      "epoch": 0.43622363575493805,
+      "grad_norm": 0.13716910779476166,
+      "learning_rate": 0.0001,
+      "loss": 1.6064,
+      "step": 1303
+    },
+    {
+      "epoch": 0.4365584198192166,
+      "grad_norm": 0.11626800149679184,
+      "learning_rate": 0.0001,
+      "loss": 1.461,
+      "step": 1304
+    },
+    {
+      "epoch": 0.4368932038834951,
+      "grad_norm": 0.12892816960811615,
+      "learning_rate": 0.0001,
+      "loss": 1.5856,
+      "step": 1305
+    },
+    {
+      "epoch": 0.4372279879477737,
+      "grad_norm": 0.12171407043933868,
+      "learning_rate": 0.0001,
+      "loss": 1.5669,
+      "step": 1306
+    },
+    {
+      "epoch": 0.43756277201205224,
+      "grad_norm": 0.12705732882022858,
+      "learning_rate": 0.0001,
+      "loss": 1.5392,
+      "step": 1307
+    },
+    {
+      "epoch": 0.43789755607633074,
+      "grad_norm": 0.12489151209592819,
+      "learning_rate": 0.0001,
+      "loss": 1.5621,
+      "step": 1308
+    },
+    {
+      "epoch": 0.4382323401406093,
+      "grad_norm": 0.1306968778371811,
+      "learning_rate": 0.0001,
+      "loss": 1.5601,
+      "step": 1309
+    },
+    {
+      "epoch": 0.43856712420488786,
+      "grad_norm": 0.12457779794931412,
+      "learning_rate": 0.0001,
+      "loss": 1.5292,
+      "step": 1310
+    },
+    {
+      "epoch": 0.43890190826916636,
+      "grad_norm": 0.1351223587989807,
+      "learning_rate": 0.0001,
+      "loss": 1.6364,
+      "step": 1311
+    },
+    {
+      "epoch": 0.4392366923334449,
+      "grad_norm": 0.16403745114803314,
+      "learning_rate": 0.0001,
+      "loss": 1.6135,
+      "step": 1312
+    },
+    {
+      "epoch": 0.4395714763977235,
+      "grad_norm": 0.1373598426580429,
+      "learning_rate": 0.0001,
+      "loss": 1.6102,
+      "step": 1313
+    },
+    {
+      "epoch": 0.439906260462002,
+      "grad_norm": 0.12474294006824493,
+      "learning_rate": 0.0001,
+      "loss": 1.4732,
+      "step": 1314
+    },
+    {
+      "epoch": 0.44024104452628054,
+      "grad_norm": 0.13775482773780823,
+      "learning_rate": 0.0001,
+      "loss": 1.4623,
+      "step": 1315
+    },
+    {
+      "epoch": 0.4405758285905591,
+      "grad_norm": 0.12874817848205566,
+      "learning_rate": 0.0001,
+      "loss": 1.5885,
+      "step": 1316
+    },
+    {
+      "epoch": 0.4409106126548376,
+      "grad_norm": 0.13382995128631592,
+      "learning_rate": 0.0001,
+      "loss": 1.4458,
+      "step": 1317
+    },
+    {
+      "epoch": 0.44124539671911617,
+      "grad_norm": 0.1267126202583313,
+      "learning_rate": 0.0001,
+      "loss": 1.5709,
+      "step": 1318
+    },
+    {
+      "epoch": 0.4415801807833947,
+      "grad_norm": 0.12839357554912567,
+      "learning_rate": 0.0001,
+      "loss": 1.5377,
+      "step": 1319
+    },
+    {
+      "epoch": 0.44191496484767323,
+      "grad_norm": 0.13176332414150238,
+      "learning_rate": 0.0001,
+      "loss": 1.4342,
+      "step": 1320
+    },
+    {
+      "epoch": 0.4422497489119518,
+      "grad_norm": 0.13202795386314392,
+      "learning_rate": 0.0001,
+      "loss": 1.5997,
+      "step": 1321
+    },
+    {
+      "epoch": 0.44258453297623035,
+      "grad_norm": 0.12316932529211044,
+      "learning_rate": 0.0001,
+      "loss": 1.4323,
+      "step": 1322
+    },
+    {
+      "epoch": 0.44291931704050885,
+      "grad_norm": 0.1301979273557663,
+      "learning_rate": 0.0001,
+      "loss": 1.5882,
+      "step": 1323
+    },
+    {
+      "epoch": 0.4432541011047874,
+      "grad_norm": 0.1263076364994049,
+      "learning_rate": 0.0001,
+      "loss": 1.4469,
+      "step": 1324
+    },
+    {
+      "epoch": 0.44358888516906597,
+      "grad_norm": 0.12310474365949631,
+      "learning_rate": 0.0001,
+      "loss": 1.4898,
+      "step": 1325
+    },
+    {
+      "epoch": 0.4439236692333445,
+      "grad_norm": 0.12039102613925934,
+      "learning_rate": 0.0001,
+      "loss": 1.5324,
+      "step": 1326
+    },
+    {
+      "epoch": 0.44425845329762303,
+      "grad_norm": 0.12545818090438843,
+      "learning_rate": 0.0001,
+      "loss": 1.6171,
+      "step": 1327
+    },
+    {
+      "epoch": 0.4445932373619016,
+      "grad_norm": 0.1259836107492447,
+      "learning_rate": 0.0001,
+      "loss": 1.5059,
+      "step": 1328
+    },
+    {
+      "epoch": 0.4449280214261801,
+      "grad_norm": 0.12518031895160675,
+      "learning_rate": 0.0001,
+      "loss": 1.5958,
+      "step": 1329
+    },
+    {
+      "epoch": 0.44526280549045866,
+      "grad_norm": 0.12583878636360168,
+      "learning_rate": 0.0001,
+      "loss": 1.4837,
+      "step": 1330
+    },
+    {
+      "epoch": 0.4455975895547372,
+      "grad_norm": 0.12569929659366608,
+      "learning_rate": 0.0001,
+      "loss": 1.536,
+      "step": 1331
+    },
+    {
+      "epoch": 0.4459323736190157,
+      "grad_norm": 0.1288549304008484,
+      "learning_rate": 0.0001,
+      "loss": 1.5525,
+      "step": 1332
+    },
+    {
+      "epoch": 0.4462671576832943,
+      "grad_norm": 0.13198384642601013,
+      "learning_rate": 0.0001,
+      "loss": 1.542,
+      "step": 1333
+    },
+    {
+      "epoch": 0.44660194174757284,
+      "grad_norm": 0.1238170713186264,
+      "learning_rate": 0.0001,
+      "loss": 1.4021,
+      "step": 1334
+    },
+    {
+      "epoch": 0.44693672581185134,
+      "grad_norm": 0.13295157253742218,
+      "learning_rate": 0.0001,
+      "loss": 1.5553,
+      "step": 1335
+    },
+    {
+      "epoch": 0.4472715098761299,
+      "grad_norm": 0.13403776288032532,
+      "learning_rate": 0.0001,
+      "loss": 1.4761,
+      "step": 1336
+    },
+    {
+      "epoch": 0.44760629394040846,
+      "grad_norm": 0.13343052566051483,
+      "learning_rate": 0.0001,
+      "loss": 1.573,
+      "step": 1337
+    },
+    {
+      "epoch": 0.44794107800468697,
+      "grad_norm": 0.125327467918396,
+      "learning_rate": 0.0001,
+      "loss": 1.5682,
+      "step": 1338
+    },
+    {
+      "epoch": 0.4482758620689655,
+      "grad_norm": 0.12958160042762756,
+      "learning_rate": 0.0001,
+      "loss": 1.5294,
+      "step": 1339
+    },
+    {
+      "epoch": 0.4486106461332441,
+      "grad_norm": 0.1384599506855011,
+      "learning_rate": 0.0001,
+      "loss": 1.5791,
+      "step": 1340
+    },
+    {
+      "epoch": 0.4489454301975226,
+      "grad_norm": 0.1257963478565216,
+      "learning_rate": 0.0001,
+      "loss": 1.5732,
+      "step": 1341
+    },
+    {
+      "epoch": 0.44928021426180115,
+      "grad_norm": 0.12630927562713623,
+      "learning_rate": 0.0001,
+      "loss": 1.5558,
+      "step": 1342
+    },
+    {
+      "epoch": 0.44961499832607965,
+      "grad_norm": 0.1268066167831421,
+      "learning_rate": 0.0001,
+      "loss": 1.5958,
+      "step": 1343
+    },
+    {
+      "epoch": 0.4499497823903582,
+      "grad_norm": 0.12455032020807266,
+      "learning_rate": 0.0001,
+      "loss": 1.5607,
+      "step": 1344
+    },
+    {
+      "epoch": 0.45028456645463677,
+      "grad_norm": 0.12265735119581223,
+      "learning_rate": 0.0001,
+      "loss": 1.5197,
+      "step": 1345
+    },
+    {
+      "epoch": 0.4506193505189153,
+      "grad_norm": 0.1307050883769989,
+      "learning_rate": 0.0001,
+      "loss": 1.6407,
+      "step": 1346
+    },
+    {
+      "epoch": 0.45095413458319383,
+      "grad_norm": 0.13128429651260376,
+      "learning_rate": 0.0001,
+      "loss": 1.5559,
+      "step": 1347
+    },
+    {
+      "epoch": 0.4512889186474724,
+      "grad_norm": 0.13010568916797638,
+      "learning_rate": 0.0001,
+      "loss": 1.5332,
+      "step": 1348
+    },
+    {
+      "epoch": 0.4516237027117509,
+      "grad_norm": 0.12650929391384125,
+      "learning_rate": 0.0001,
+      "loss": 1.6047,
+      "step": 1349
+    },
+    {
+      "epoch": 0.45195848677602946,
+      "grad_norm": 0.12306904792785645,
+      "learning_rate": 0.0001,
+      "loss": 1.5499,
+      "step": 1350
+    },
+    {
+      "epoch": 0.452293270840308,
+      "grad_norm": 0.13351021707057953,
+      "learning_rate": 0.0001,
+      "loss": 1.4737,
+      "step": 1351
+    },
+    {
+      "epoch": 0.4526280549045865,
+      "grad_norm": 0.12178155779838562,
+      "learning_rate": 0.0001,
+      "loss": 1.4775,
+      "step": 1352
+    },
+    {
+      "epoch": 0.4529628389688651,
+      "grad_norm": 0.13516512513160706,
+      "learning_rate": 0.0001,
+      "loss": 1.6391,
+      "step": 1353
+    },
+    {
+      "epoch": 0.45329762303314364,
+      "grad_norm": 0.12909267842769623,
+      "learning_rate": 0.0001,
+      "loss": 1.4684,
+      "step": 1354
+    },
+    {
+      "epoch": 0.45363240709742214,
+      "grad_norm": 0.12209142744541168,
+      "learning_rate": 0.0001,
+      "loss": 1.5198,
+      "step": 1355
+    },
+    {
+      "epoch": 0.4539671911617007,
+      "grad_norm": 0.1269826740026474,
+      "learning_rate": 0.0001,
+      "loss": 1.5294,
+      "step": 1356
+    },
+    {
+      "epoch": 0.45430197522597926,
+      "grad_norm": 0.13762542605400085,
+      "learning_rate": 0.0001,
+      "loss": 1.5567,
+      "step": 1357
+    },
+    {
+      "epoch": 0.45463675929025776,
+      "grad_norm": 0.1306358128786087,
+      "learning_rate": 0.0001,
+      "loss": 1.5829,
+      "step": 1358
+    },
+    {
+      "epoch": 0.4549715433545363,
+      "grad_norm": 0.1383924037218094,
+      "learning_rate": 0.0001,
+      "loss": 1.6382,
+      "step": 1359
+    },
+    {
+      "epoch": 0.4553063274188149,
+      "grad_norm": 0.13577204942703247,
+      "learning_rate": 0.0001,
+      "loss": 1.6067,
+      "step": 1360
+    },
+    {
+      "epoch": 0.4556411114830934,
+      "grad_norm": 0.12534180283546448,
+      "learning_rate": 0.0001,
+      "loss": 1.574,
+      "step": 1361
+    },
+    {
+      "epoch": 0.45597589554737195,
+      "grad_norm": 0.12367561459541321,
+      "learning_rate": 0.0001,
+      "loss": 1.5089,
+      "step": 1362
+    },
+    {
+      "epoch": 0.4563106796116505,
+      "grad_norm": 0.14012429118156433,
+      "learning_rate": 0.0001,
+      "loss": 1.6044,
+      "step": 1363
+    },
+    {
+      "epoch": 0.456645463675929,
+      "grad_norm": 0.13164697587490082,
+      "learning_rate": 0.0001,
+      "loss": 1.6058,
+      "step": 1364
+    },
+    {
+      "epoch": 0.45698024774020757,
+      "grad_norm": 0.14275015890598297,
+      "learning_rate": 0.0001,
+      "loss": 1.6945,
+      "step": 1365
+    },
+    {
+      "epoch": 0.4573150318044861,
+      "grad_norm": 0.1312190294265747,
+      "learning_rate": 0.0001,
+      "loss": 1.5595,
+      "step": 1366
+    },
+    {
+      "epoch": 0.45764981586876463,
+      "grad_norm": 0.1276426464319229,
+      "learning_rate": 0.0001,
+      "loss": 1.5639,
+      "step": 1367
+    },
+    {
+      "epoch": 0.4579845999330432,
+      "grad_norm": 0.12928691506385803,
+      "learning_rate": 0.0001,
+      "loss": 1.6555,
+      "step": 1368
+    },
+    {
+      "epoch": 0.45831938399732175,
+      "grad_norm": 0.12562155723571777,
+      "learning_rate": 0.0001,
+      "loss": 1.5017,
+      "step": 1369
+    },
+    {
+      "epoch": 0.45865416806160025,
+      "grad_norm": 0.12555162608623505,
+      "learning_rate": 0.0001,
+      "loss": 1.5133,
+      "step": 1370
+    },
+    {
+      "epoch": 0.4589889521258788,
+      "grad_norm": 0.13354945182800293,
+      "learning_rate": 0.0001,
+      "loss": 1.5802,
+      "step": 1371
+    },
+    {
+      "epoch": 0.4593237361901574,
+      "grad_norm": 0.13059929013252258,
+      "learning_rate": 0.0001,
+      "loss": 1.5152,
+      "step": 1372
+    },
+    {
+      "epoch": 0.4596585202544359,
+      "grad_norm": 0.1313420981168747,
+      "learning_rate": 0.0001,
+      "loss": 1.5411,
+      "step": 1373
+    },
+    {
+      "epoch": 0.45999330431871444,
+      "grad_norm": 0.13619214296340942,
+      "learning_rate": 0.0001,
+      "loss": 1.5348,
+      "step": 1374
+    },
+    {
+      "epoch": 0.460328088382993,
+      "grad_norm": 0.12227842211723328,
+      "learning_rate": 0.0001,
+      "loss": 1.5258,
+      "step": 1375
+    },
+    {
+      "epoch": 0.4606628724472715,
+      "grad_norm": 0.12962037324905396,
+      "learning_rate": 0.0001,
+      "loss": 1.6469,
+      "step": 1376
+    },
+    {
+      "epoch": 0.46099765651155006,
+      "grad_norm": 0.128581240773201,
+      "learning_rate": 0.0001,
+      "loss": 1.6151,
+      "step": 1377
+    },
+    {
+      "epoch": 0.4613324405758286,
+      "grad_norm": 0.12887564301490784,
+      "learning_rate": 0.0001,
+      "loss": 1.5741,
+      "step": 1378
+    },
+    {
+      "epoch": 0.4616672246401071,
+      "grad_norm": 0.12684863805770874,
+      "learning_rate": 0.0001,
+      "loss": 1.6168,
+      "step": 1379
+    },
+    {
+      "epoch": 0.4620020087043857,
+      "grad_norm": 0.11986137181520462,
+      "learning_rate": 0.0001,
+      "loss": 1.5278,
+      "step": 1380
+    },
+    {
+      "epoch": 0.4623367927686642,
+      "grad_norm": 0.12904709577560425,
+      "learning_rate": 0.0001,
+      "loss": 1.5247,
+      "step": 1381
+    },
+    {
+      "epoch": 0.46267157683294274,
+      "grad_norm": 0.12737007439136505,
+      "learning_rate": 0.0001,
+      "loss": 1.6354,
+      "step": 1382
+    },
+    {
+      "epoch": 0.4630063608972213,
+      "grad_norm": 0.13845406472682953,
+      "learning_rate": 0.0001,
+      "loss": 1.5696,
+      "step": 1383
+    },
+    {
+      "epoch": 0.4633411449614998,
+      "grad_norm": 0.1215730682015419,
+      "learning_rate": 0.0001,
+      "loss": 1.5277,
+      "step": 1384
+    },
+    {
+      "epoch": 0.46367592902577837,
+      "grad_norm": 0.12643855810165405,
+      "learning_rate": 0.0001,
+      "loss": 1.5691,
+      "step": 1385
+    },
+    {
+      "epoch": 0.4640107130900569,
+      "grad_norm": 0.12575271725654602,
+      "learning_rate": 0.0001,
+      "loss": 1.5075,
+      "step": 1386
+    },
+    {
+      "epoch": 0.46434549715433543,
+      "grad_norm": 0.13134850561618805,
+      "learning_rate": 0.0001,
+      "loss": 1.6195,
+      "step": 1387
+    },
+    {
+      "epoch": 0.464680281218614,
+      "grad_norm": 0.12751908600330353,
+      "learning_rate": 0.0001,
+      "loss": 1.5396,
+      "step": 1388
+    },
+    {
+      "epoch": 0.46501506528289255,
+      "grad_norm": 0.1260857880115509,
+      "learning_rate": 0.0001,
+      "loss": 1.581,
+      "step": 1389
+    },
+    {
+      "epoch": 0.46534984934717105,
+      "grad_norm": 0.13056620955467224,
+      "learning_rate": 0.0001,
+      "loss": 1.5604,
+      "step": 1390
+    },
+    {
+      "epoch": 0.4656846334114496,
+      "grad_norm": 0.12854252755641937,
+      "learning_rate": 0.0001,
+      "loss": 1.5729,
+      "step": 1391
+    },
+    {
+      "epoch": 0.46601941747572817,
+      "grad_norm": 0.12587207555770874,
+      "learning_rate": 0.0001,
+      "loss": 1.5685,
+      "step": 1392
+    },
+    {
+      "epoch": 0.4663542015400067,
+      "grad_norm": 0.13984687626361847,
+      "learning_rate": 0.0001,
+      "loss": 1.5327,
+      "step": 1393
+    },
+    {
+      "epoch": 0.46668898560428523,
+      "grad_norm": 0.1340693235397339,
+      "learning_rate": 0.0001,
+      "loss": 1.5047,
+      "step": 1394
+    },
+    {
+      "epoch": 0.4670237696685638,
+      "grad_norm": 0.12426851689815521,
+      "learning_rate": 0.0001,
+      "loss": 1.5614,
+      "step": 1395
+    },
+    {
+      "epoch": 0.4673585537328423,
+      "grad_norm": 0.14335423707962036,
+      "learning_rate": 0.0001,
+      "loss": 1.5968,
+      "step": 1396
+    },
+    {
+      "epoch": 0.46769333779712086,
+      "grad_norm": 0.1285167783498764,
+      "learning_rate": 0.0001,
+      "loss": 1.4816,
+      "step": 1397
+    },
+    {
+      "epoch": 0.4680281218613994,
+      "grad_norm": 0.12221338599920273,
+      "learning_rate": 0.0001,
+      "loss": 1.5412,
+      "step": 1398
+    },
+    {
+      "epoch": 0.4683629059256779,
+      "grad_norm": 0.13749419152736664,
+      "learning_rate": 0.0001,
+      "loss": 1.6426,
+      "step": 1399
+    },
+    {
+      "epoch": 0.4686976899899565,
+      "grad_norm": 0.1292765736579895,
+      "learning_rate": 0.0001,
+      "loss": 1.4826,
+      "step": 1400
+    },
+    {
+      "epoch": 0.46903247405423504,
+      "grad_norm": 0.12175814807415009,
+      "learning_rate": 0.0001,
+      "loss": 1.4674,
+      "step": 1401
+    },
+    {
+      "epoch": 0.46936725811851354,
+      "grad_norm": 0.13381820917129517,
+      "learning_rate": 0.0001,
+      "loss": 1.515,
+      "step": 1402
+    },
+    {
+      "epoch": 0.4697020421827921,
+      "grad_norm": 0.13659454882144928,
+      "learning_rate": 0.0001,
+      "loss": 1.5513,
+      "step": 1403
+    },
+    {
+      "epoch": 0.47003682624707066,
+      "grad_norm": 0.12511052191257477,
+      "learning_rate": 0.0001,
+      "loss": 1.5457,
+      "step": 1404
+    },
+    {
+      "epoch": 0.47037161031134916,
+      "grad_norm": 0.13325883448123932,
+      "learning_rate": 0.0001,
+      "loss": 1.5893,
+      "step": 1405
+    },
+    {
+      "epoch": 0.4707063943756277,
+      "grad_norm": 0.12582562863826752,
+      "learning_rate": 0.0001,
+      "loss": 1.5285,
+      "step": 1406
+    },
+    {
+      "epoch": 0.4710411784399063,
+      "grad_norm": 0.13141517341136932,
+      "learning_rate": 0.0001,
+      "loss": 1.5865,
+      "step": 1407
+    },
+    {
+      "epoch": 0.4713759625041848,
+      "grad_norm": 0.13099296391010284,
+      "learning_rate": 0.0001,
+      "loss": 1.5322,
+      "step": 1408
+    },
+    {
+      "epoch": 0.47171074656846335,
+      "grad_norm": 0.146238312125206,
+      "learning_rate": 0.0001,
+      "loss": 1.6397,
+      "step": 1409
+    },
+    {
+      "epoch": 0.4720455306327419,
+      "grad_norm": 0.12129180878400803,
+      "learning_rate": 0.0001,
+      "loss": 1.5033,
+      "step": 1410
+    },
+    {
+      "epoch": 0.4723803146970204,
+      "grad_norm": 0.125573992729187,
+      "learning_rate": 0.0001,
+      "loss": 1.571,
+      "step": 1411
+    },
+    {
+      "epoch": 0.47271509876129897,
+      "grad_norm": 0.14334800839424133,
+      "learning_rate": 0.0001,
+      "loss": 1.5323,
+      "step": 1412
+    },
+    {
+      "epoch": 0.47304988282557753,
+      "grad_norm": 0.1354663372039795,
+      "learning_rate": 0.0001,
+      "loss": 1.5733,
+      "step": 1413
+    },
+    {
+      "epoch": 0.47338466688985603,
+      "grad_norm": 0.13040928542613983,
+      "learning_rate": 0.0001,
+      "loss": 1.4702,
+      "step": 1414
+    },
+    {
+      "epoch": 0.4737194509541346,
+      "grad_norm": 0.12931925058364868,
+      "learning_rate": 0.0001,
+      "loss": 1.6017,
+      "step": 1415
+    },
+    {
+      "epoch": 0.47405423501841315,
+      "grad_norm": 0.13492871820926666,
+      "learning_rate": 0.0001,
+      "loss": 1.5827,
+      "step": 1416
+    },
+    {
+      "epoch": 0.47438901908269165,
+      "grad_norm": 0.12549789249897003,
+      "learning_rate": 0.0001,
+      "loss": 1.5856,
+      "step": 1417
+    },
+    {
+      "epoch": 0.4747238031469702,
+      "grad_norm": 0.13328687846660614,
+      "learning_rate": 0.0001,
+      "loss": 1.6163,
+      "step": 1418
+    },
+    {
+      "epoch": 0.4750585872112487,
+      "grad_norm": 0.13430629670619965,
+      "learning_rate": 0.0001,
+      "loss": 1.5663,
+      "step": 1419
+    },
+    {
+      "epoch": 0.4753933712755273,
+      "grad_norm": 0.12909024953842163,
+      "learning_rate": 0.0001,
+      "loss": 1.6085,
+      "step": 1420
+    },
+    {
+      "epoch": 0.47572815533980584,
+      "grad_norm": 0.13095097243785858,
+      "learning_rate": 0.0001,
+      "loss": 1.585,
+      "step": 1421
+    },
+    {
+      "epoch": 0.47606293940408434,
+      "grad_norm": 0.1313266009092331,
+      "learning_rate": 0.0001,
+      "loss": 1.5279,
+      "step": 1422
+    },
+    {
+      "epoch": 0.4763977234683629,
+      "grad_norm": 0.12739764153957367,
+      "learning_rate": 0.0001,
+      "loss": 1.6473,
+      "step": 1423
+    },
+    {
+      "epoch": 0.47673250753264146,
+      "grad_norm": 0.12780874967575073,
+      "learning_rate": 0.0001,
+      "loss": 1.5566,
+      "step": 1424
+    },
+    {
+      "epoch": 0.47706729159691996,
+      "grad_norm": 0.12299945950508118,
+      "learning_rate": 0.0001,
+      "loss": 1.5632,
+      "step": 1425
+    },
+    {
+      "epoch": 0.4774020756611985,
+      "grad_norm": 0.12845619022846222,
+      "learning_rate": 0.0001,
+      "loss": 1.5799,
+      "step": 1426
+    },
+    {
+      "epoch": 0.4777368597254771,
+      "grad_norm": 0.12429885566234589,
+      "learning_rate": 0.0001,
+      "loss": 1.565,
+      "step": 1427
+    },
+    {
+      "epoch": 0.4780716437897556,
+      "grad_norm": 0.12623021006584167,
+      "learning_rate": 0.0001,
+      "loss": 1.5579,
+      "step": 1428
+    },
+    {
+      "epoch": 0.47840642785403414,
+      "grad_norm": 0.121118925511837,
+      "learning_rate": 0.0001,
+      "loss": 1.5044,
+      "step": 1429
+    },
+    {
+      "epoch": 0.4787412119183127,
+      "grad_norm": 0.13029584288597107,
+      "learning_rate": 0.0001,
+      "loss": 1.5945,
+      "step": 1430
+    },
+    {
+      "epoch": 0.4790759959825912,
+      "grad_norm": 0.1309075504541397,
+      "learning_rate": 0.0001,
+      "loss": 1.5638,
+      "step": 1431
+    },
+    {
+      "epoch": 0.47941078004686977,
+      "grad_norm": 0.12302339822053909,
+      "learning_rate": 0.0001,
+      "loss": 1.553,
+      "step": 1432
+    },
+    {
+      "epoch": 0.4797455641111483,
+      "grad_norm": 0.13640674948692322,
+      "learning_rate": 0.0001,
+      "loss": 1.6299,
+      "step": 1433
+    },
+    {
+      "epoch": 0.48008034817542683,
+      "grad_norm": 0.12669233977794647,
+      "learning_rate": 0.0001,
+      "loss": 1.5603,
+      "step": 1434
+    },
+    {
+      "epoch": 0.4804151322397054,
+      "grad_norm": 0.14192534983158112,
+      "learning_rate": 0.0001,
+      "loss": 1.5648,
+      "step": 1435
+    },
+    {
+      "epoch": 0.48074991630398395,
+      "grad_norm": 0.12855654954910278,
+      "learning_rate": 0.0001,
+      "loss": 1.5782,
+      "step": 1436
+    },
+    {
+      "epoch": 0.48108470036826245,
+      "grad_norm": 0.13193868100643158,
+      "learning_rate": 0.0001,
+      "loss": 1.4815,
+      "step": 1437
+    },
+    {
+      "epoch": 0.481419484432541,
+      "grad_norm": 0.1313331574201584,
+      "learning_rate": 0.0001,
+      "loss": 1.597,
+      "step": 1438
+    },
+    {
+      "epoch": 0.48175426849681957,
+      "grad_norm": 0.14010664820671082,
+      "learning_rate": 0.0001,
+      "loss": 1.5911,
+      "step": 1439
+    },
+    {
+      "epoch": 0.4820890525610981,
+      "grad_norm": 0.12899306416511536,
+      "learning_rate": 0.0001,
+      "loss": 1.5346,
+      "step": 1440
+    },
+    {
+      "epoch": 0.48242383662537663,
+      "grad_norm": 0.14157001674175262,
+      "learning_rate": 0.0001,
+      "loss": 1.4947,
+      "step": 1441
+    },
+    {
+      "epoch": 0.4827586206896552,
+      "grad_norm": 0.12598420679569244,
+      "learning_rate": 0.0001,
+      "loss": 1.5713,
+      "step": 1442
+    },
+    {
+      "epoch": 0.4830934047539337,
+      "grad_norm": 0.12368304282426834,
+      "learning_rate": 0.0001,
+      "loss": 1.4691,
+      "step": 1443
+    },
+    {
+      "epoch": 0.48342818881821226,
+      "grad_norm": 0.15252211689949036,
+      "learning_rate": 0.0001,
+      "loss": 1.5298,
+      "step": 1444
+    },
+    {
+      "epoch": 0.4837629728824908,
+      "grad_norm": 0.12461958080530167,
+      "learning_rate": 0.0001,
+      "loss": 1.5377,
+      "step": 1445
+    },
+    {
+      "epoch": 0.4840977569467693,
+      "grad_norm": 0.13883721828460693,
+      "learning_rate": 0.0001,
+      "loss": 1.5754,
+      "step": 1446
+    },
+    {
+      "epoch": 0.4844325410110479,
+      "grad_norm": 0.14833161234855652,
+      "learning_rate": 0.0001,
+      "loss": 1.514,
+      "step": 1447
+    },
+    {
+      "epoch": 0.48476732507532644,
+      "grad_norm": 0.12511619925498962,
+      "learning_rate": 0.0001,
+      "loss": 1.5765,
+      "step": 1448
+    },
+    {
+      "epoch": 0.48510210913960494,
+      "grad_norm": 0.1352238804101944,
+      "learning_rate": 0.0001,
+      "loss": 1.5231,
+      "step": 1449
+    },
+    {
+      "epoch": 0.4854368932038835,
+      "grad_norm": 0.14310289919376373,
+      "learning_rate": 0.0001,
+      "loss": 1.5516,
+      "step": 1450
+    },
+    {
+      "epoch": 0.48577167726816206,
+      "grad_norm": 0.1293793022632599,
+      "learning_rate": 0.0001,
+      "loss": 1.6124,
+      "step": 1451
+    },
+    {
+      "epoch": 0.48610646133244056,
+      "grad_norm": 0.1351606398820877,
+      "learning_rate": 0.0001,
+      "loss": 1.5535,
+      "step": 1452
+    },
+    {
+      "epoch": 0.4864412453967191,
+      "grad_norm": 0.1305823028087616,
+      "learning_rate": 0.0001,
+      "loss": 1.505,
+      "step": 1453
+    },
+    {
+      "epoch": 0.4867760294609977,
+      "grad_norm": 0.12973332405090332,
+      "learning_rate": 0.0001,
+      "loss": 1.6027,
+      "step": 1454
+    },
+    {
+      "epoch": 0.4871108135252762,
+      "grad_norm": 0.1279638260602951,
+      "learning_rate": 0.0001,
+      "loss": 1.5664,
+      "step": 1455
+    },
+    {
+      "epoch": 0.48744559758955475,
+      "grad_norm": 0.1322777271270752,
+      "learning_rate": 0.0001,
+      "loss": 1.605,
+      "step": 1456
+    },
+    {
+      "epoch": 0.48778038165383325,
+      "grad_norm": 0.14680039882659912,
+      "learning_rate": 0.0001,
+      "loss": 1.5243,
+      "step": 1457
+    },
+    {
+      "epoch": 0.4881151657181118,
+      "grad_norm": 0.12435714155435562,
+      "learning_rate": 0.0001,
+      "loss": 1.4835,
+      "step": 1458
+    },
+    {
+      "epoch": 0.48844994978239037,
+      "grad_norm": 0.13253144919872284,
+      "learning_rate": 0.0001,
+      "loss": 1.5797,
+      "step": 1459
+    },
+    {
+      "epoch": 0.4887847338466689,
+      "grad_norm": 0.14123192429542542,
+      "learning_rate": 0.0001,
+      "loss": 1.5795,
+      "step": 1460
+    },
+    {
+      "epoch": 0.48911951791094743,
+      "grad_norm": 0.1254579871892929,
+      "learning_rate": 0.0001,
+      "loss": 1.4829,
+      "step": 1461
+    },
+    {
+      "epoch": 0.489454301975226,
+      "grad_norm": 0.1407458633184433,
+      "learning_rate": 0.0001,
+      "loss": 1.5746,
+      "step": 1462
+    },
+    {
+      "epoch": 0.4897890860395045,
+      "grad_norm": 0.13967539370059967,
+      "learning_rate": 0.0001,
+      "loss": 1.611,
+      "step": 1463
+    },
+    {
+      "epoch": 0.49012387010378305,
+      "grad_norm": 0.13044650852680206,
+      "learning_rate": 0.0001,
+      "loss": 1.5614,
+      "step": 1464
+    },
+    {
+      "epoch": 0.4904586541680616,
+      "grad_norm": 0.13819964230060577,
+      "learning_rate": 0.0001,
+      "loss": 1.5579,
+      "step": 1465
+    },
+    {
+      "epoch": 0.4907934382323401,
+      "grad_norm": 0.12795104086399078,
+      "learning_rate": 0.0001,
+      "loss": 1.5373,
+      "step": 1466
+    },
+    {
+      "epoch": 0.4911282222966187,
+      "grad_norm": 0.13034126162528992,
+      "learning_rate": 0.0001,
+      "loss": 1.5077,
+      "step": 1467
+    },
+    {
+      "epoch": 0.49146300636089724,
+      "grad_norm": 0.1358436644077301,
+      "learning_rate": 0.0001,
+      "loss": 1.6376,
+      "step": 1468
+    },
+    {
+      "epoch": 0.49179779042517574,
+      "grad_norm": 0.12750184535980225,
+      "learning_rate": 0.0001,
+      "loss": 1.5638,
+      "step": 1469
+    },
+    {
+      "epoch": 0.4921325744894543,
+      "grad_norm": 0.13034793734550476,
+      "learning_rate": 0.0001,
+      "loss": 1.5053,
+      "step": 1470
+    },
+    {
+      "epoch": 0.49246735855373286,
+      "grad_norm": 0.1303941309452057,
+      "learning_rate": 0.0001,
+      "loss": 1.5342,
+      "step": 1471
+    },
+    {
+      "epoch": 0.49280214261801136,
+      "grad_norm": 0.12955164909362793,
+      "learning_rate": 0.0001,
+      "loss": 1.5396,
+      "step": 1472
+    },
+    {
+      "epoch": 0.4931369266822899,
+      "grad_norm": 0.12884975969791412,
+      "learning_rate": 0.0001,
+      "loss": 1.5389,
+      "step": 1473
+    },
+    {
+      "epoch": 0.4934717107465685,
+      "grad_norm": 0.1278049647808075,
+      "learning_rate": 0.0001,
+      "loss": 1.5937,
+      "step": 1474
+    },
+    {
+      "epoch": 0.493806494810847,
+      "grad_norm": 0.12420760840177536,
+      "learning_rate": 0.0001,
+      "loss": 1.4753,
+      "step": 1475
+    },
+    {
+      "epoch": 0.49414127887512554,
+      "grad_norm": 0.12760096788406372,
+      "learning_rate": 0.0001,
+      "loss": 1.647,
+      "step": 1476
+    },
+    {
+      "epoch": 0.4944760629394041,
+      "grad_norm": 0.1320486068725586,
+      "learning_rate": 0.0001,
+      "loss": 1.5758,
+      "step": 1477
+    },
+    {
+      "epoch": 0.4948108470036826,
+      "grad_norm": 0.13898344337940216,
+      "learning_rate": 0.0001,
+      "loss": 1.6265,
+      "step": 1478
+    },
+    {
+      "epoch": 0.49514563106796117,
+      "grad_norm": 0.12908297777175903,
+      "learning_rate": 0.0001,
+      "loss": 1.6294,
+      "step": 1479
+    },
+    {
+      "epoch": 0.4954804151322397,
+      "grad_norm": 0.13149291276931763,
+      "learning_rate": 0.0001,
+      "loss": 1.5297,
+      "step": 1480
+    },
+    {
+      "epoch": 0.49581519919651823,
+      "grad_norm": 0.13526497781276703,
+      "learning_rate": 0.0001,
+      "loss": 1.5374,
+      "step": 1481
+    },
+    {
+      "epoch": 0.4961499832607968,
+      "grad_norm": 0.12223420292139053,
+      "learning_rate": 0.0001,
+      "loss": 1.5424,
+      "step": 1482
+    },
+    {
+      "epoch": 0.49648476732507535,
+      "grad_norm": 0.1266697198152542,
+      "learning_rate": 0.0001,
+      "loss": 1.5847,
+      "step": 1483
+    },
+    {
+      "epoch": 0.49681955138935385,
+      "grad_norm": 0.14440171420574188,
+      "learning_rate": 0.0001,
+      "loss": 1.5362,
+      "step": 1484
+    },
+    {
+      "epoch": 0.4971543354536324,
+      "grad_norm": 0.12831640243530273,
+      "learning_rate": 0.0001,
+      "loss": 1.5803,
+      "step": 1485
+    },
+    {
+      "epoch": 0.49748911951791097,
+      "grad_norm": 0.13665077090263367,
+      "learning_rate": 0.0001,
+      "loss": 1.5741,
+      "step": 1486
+    },
+    {
+      "epoch": 0.4978239035821895,
+      "grad_norm": 0.13725218176841736,
+      "learning_rate": 0.0001,
+      "loss": 1.6207,
+      "step": 1487
+    },
+    {
+      "epoch": 0.49815868764646803,
+      "grad_norm": 0.1271527111530304,
+      "learning_rate": 0.0001,
+      "loss": 1.6129,
+      "step": 1488
+    },
+    {
+      "epoch": 0.4984934717107466,
+      "grad_norm": 0.15319159626960754,
+      "learning_rate": 0.0001,
+      "loss": 1.6247,
+      "step": 1489
+    },
+    {
+      "epoch": 0.4988282557750251,
+      "grad_norm": 0.12440894544124603,
+      "learning_rate": 0.0001,
+      "loss": 1.4354,
+      "step": 1490
+    },
+    {
+      "epoch": 0.49916303983930366,
+      "grad_norm": 0.1261643022298813,
+      "learning_rate": 0.0001,
+      "loss": 1.609,
+      "step": 1491
+    },
+    {
+      "epoch": 0.4994978239035822,
+      "grad_norm": 0.14216668903827667,
+      "learning_rate": 0.0001,
+      "loss": 1.5599,
+      "step": 1492
+    },
+    {
+      "epoch": 0.4998326079678607,
+      "grad_norm": 0.13173174858093262,
+      "learning_rate": 0.0001,
+      "loss": 1.5056,
+      "step": 1493
+    },
+    {
+      "epoch": 0.5001673920321392,
+      "grad_norm": 0.12335377931594849,
+      "learning_rate": 0.0001,
+      "loss": 1.5544,
+      "step": 1494
+    },
+    {
+      "epoch": 0.5005021760964178,
+      "grad_norm": 0.13367588818073273,
+      "learning_rate": 0.0001,
+      "loss": 1.4908,
+      "step": 1495
+    },
+    {
+      "epoch": 0.5008369601606963,
+      "grad_norm": 0.13830317556858063,
+      "learning_rate": 0.0001,
+      "loss": 1.6147,
+      "step": 1496
+    },
+    {
+      "epoch": 0.5011717442249749,
+      "grad_norm": 0.13441935181617737,
+      "learning_rate": 0.0001,
+      "loss": 1.6855,
+      "step": 1497
+    },
+    {
+      "epoch": 0.5015065282892535,
+      "grad_norm": 0.14937585592269897,
+      "learning_rate": 0.0001,
+      "loss": 1.6021,
+      "step": 1498
+    },
+    {
+      "epoch": 0.501841312353532,
+      "grad_norm": 0.1289912909269333,
+      "learning_rate": 0.0001,
+      "loss": 1.5516,
+      "step": 1499
+    },
+    {
+      "epoch": 0.5021760964178105,
+      "grad_norm": 0.12371324002742767,
+      "learning_rate": 0.0001,
+      "loss": 1.5842,
+      "step": 1500
+    },
+    {
+      "epoch": 0.502510880482089,
+      "grad_norm": 0.12764602899551392,
+      "learning_rate": 0.0001,
+      "loss": 1.5836,
+      "step": 1501
+    },
+    {
+      "epoch": 0.5028456645463676,
+      "grad_norm": 0.12929953634738922,
+      "learning_rate": 0.0001,
+      "loss": 1.5656,
+      "step": 1502
+    },
+    {
+      "epoch": 0.5031804486106461,
+      "grad_norm": 0.1252906322479248,
+      "learning_rate": 0.0001,
+      "loss": 1.4856,
+      "step": 1503
+    },
+    {
+      "epoch": 0.5035152326749247,
+      "grad_norm": 0.13477809727191925,
+      "learning_rate": 0.0001,
+      "loss": 1.6185,
+      "step": 1504
+    },
+    {
+      "epoch": 0.5038500167392033,
+      "grad_norm": 0.12459214776754379,
+      "learning_rate": 0.0001,
+      "loss": 1.5323,
+      "step": 1505
+    },
+    {
+      "epoch": 0.5041848008034817,
+      "grad_norm": 0.12989842891693115,
+      "learning_rate": 0.0001,
+      "loss": 1.5325,
+      "step": 1506
+    },
+    {
+      "epoch": 0.5045195848677603,
+      "grad_norm": 0.12878334522247314,
+      "learning_rate": 0.0001,
+      "loss": 1.6504,
+      "step": 1507
+    },
+    {
+      "epoch": 0.5048543689320388,
+      "grad_norm": 0.14765828847885132,
+      "learning_rate": 0.0001,
+      "loss": 1.5978,
+      "step": 1508
+    },
+    {
+      "epoch": 0.5051891529963174,
+      "grad_norm": 0.1294100284576416,
+      "learning_rate": 0.0001,
+      "loss": 1.6909,
+      "step": 1509
+    },
+    {
+      "epoch": 0.505523937060596,
+      "grad_norm": 0.1304991990327835,
+      "learning_rate": 0.0001,
+      "loss": 1.6513,
+      "step": 1510
+    },
+    {
+      "epoch": 0.5058587211248745,
+      "grad_norm": 0.1318545788526535,
+      "learning_rate": 0.0001,
+      "loss": 1.5489,
+      "step": 1511
+    },
+    {
+      "epoch": 0.506193505189153,
+      "grad_norm": 0.13185527920722961,
+      "learning_rate": 0.0001,
+      "loss": 1.6317,
+      "step": 1512
+    },
+    {
+      "epoch": 0.5065282892534315,
+      "grad_norm": 0.13133597373962402,
+      "learning_rate": 0.0001,
+      "loss": 1.5853,
+      "step": 1513
+    },
+    {
+      "epoch": 0.5068630733177101,
+      "grad_norm": 0.14132916927337646,
+      "learning_rate": 0.0001,
+      "loss": 1.6844,
+      "step": 1514
+    },
+    {
+      "epoch": 0.5071978573819886,
+      "grad_norm": 0.12680397927761078,
+      "learning_rate": 0.0001,
+      "loss": 1.6048,
+      "step": 1515
+    },
+    {
+      "epoch": 0.5075326414462672,
+      "grad_norm": 0.125723198056221,
+      "learning_rate": 0.0001,
+      "loss": 1.5296,
+      "step": 1516
+    },
+    {
+      "epoch": 0.5078674255105456,
+      "grad_norm": 0.135573148727417,
+      "learning_rate": 0.0001,
+      "loss": 1.6619,
+      "step": 1517
+    },
+    {
+      "epoch": 0.5082022095748242,
+      "grad_norm": 0.12755006551742554,
+      "learning_rate": 0.0001,
+      "loss": 1.5376,
+      "step": 1518
+    },
+    {
+      "epoch": 0.5085369936391028,
+      "grad_norm": 0.1527450680732727,
+      "learning_rate": 0.0001,
+      "loss": 1.4984,
+      "step": 1519
+    },
+    {
+      "epoch": 0.5088717777033813,
+      "grad_norm": 0.12978217005729675,
+      "learning_rate": 0.0001,
+      "loss": 1.514,
+      "step": 1520
+    },
+    {
+      "epoch": 0.5092065617676599,
+      "grad_norm": 0.13393737375736237,
+      "learning_rate": 0.0001,
+      "loss": 1.5267,
+      "step": 1521
+    },
+    {
+      "epoch": 0.5095413458319384,
+      "grad_norm": 0.13406458497047424,
+      "learning_rate": 0.0001,
+      "loss": 1.4858,
+      "step": 1522
+    },
+    {
+      "epoch": 0.5098761298962169,
+      "grad_norm": 0.13214215636253357,
+      "learning_rate": 0.0001,
+      "loss": 1.5391,
+      "step": 1523
+    },
+    {
+      "epoch": 0.5102109139604954,
+      "grad_norm": 0.13335101306438446,
+      "learning_rate": 0.0001,
+      "loss": 1.5791,
+      "step": 1524
+    },
+    {
+      "epoch": 0.510545698024774,
+      "grad_norm": 0.12885718047618866,
+      "learning_rate": 0.0001,
+      "loss": 1.532,
+      "step": 1525
+    },
+    {
+      "epoch": 0.5108804820890526,
+      "grad_norm": 0.12838226556777954,
+      "learning_rate": 0.0001,
+      "loss": 1.5186,
+      "step": 1526
+    },
+    {
+      "epoch": 0.5112152661533311,
+      "grad_norm": 0.13160903751850128,
+      "learning_rate": 0.0001,
+      "loss": 1.5792,
+      "step": 1527
+    },
+    {
+      "epoch": 0.5115500502176097,
+      "grad_norm": 0.1264614462852478,
+      "learning_rate": 0.0001,
+      "loss": 1.6005,
+      "step": 1528
+    },
+    {
+      "epoch": 0.5118848342818881,
+      "grad_norm": 0.13425403833389282,
+      "learning_rate": 0.0001,
+      "loss": 1.5413,
+      "step": 1529
+    },
+    {
+      "epoch": 0.5122196183461667,
+      "grad_norm": 0.12175809592008591,
+      "learning_rate": 0.0001,
+      "loss": 1.5128,
+      "step": 1530
+    },
+    {
+      "epoch": 0.5125544024104453,
+      "grad_norm": 0.1299484223127365,
+      "learning_rate": 0.0001,
+      "loss": 1.4981,
+      "step": 1531
+    },
+    {
+      "epoch": 0.5128891864747238,
+      "grad_norm": 0.12358542531728745,
+      "learning_rate": 0.0001,
+      "loss": 1.4794,
+      "step": 1532
+    },
+    {
+      "epoch": 0.5132239705390024,
+      "grad_norm": 0.12457676231861115,
+      "learning_rate": 0.0001,
+      "loss": 1.462,
+      "step": 1533
+    },
+    {
+      "epoch": 0.5135587546032809,
+      "grad_norm": 0.12775678932666779,
+      "learning_rate": 0.0001,
+      "loss": 1.4993,
+      "step": 1534
+    },
+    {
+      "epoch": 0.5138935386675594,
+      "grad_norm": 0.12386265397071838,
+      "learning_rate": 0.0001,
+      "loss": 1.504,
+      "step": 1535
+    },
+    {
+      "epoch": 0.5142283227318379,
+      "grad_norm": 0.13995805382728577,
+      "learning_rate": 0.0001,
+      "loss": 1.5912,
+      "step": 1536
+    },
+    {
+      "epoch": 0.5145631067961165,
+      "grad_norm": 0.1274706870317459,
+      "learning_rate": 0.0001,
+      "loss": 1.6514,
+      "step": 1537
+    },
+    {
+      "epoch": 0.5148978908603951,
+      "grad_norm": 0.12781144678592682,
+      "learning_rate": 0.0001,
+      "loss": 1.5379,
+      "step": 1538
+    },
+    {
+      "epoch": 0.5152326749246736,
+      "grad_norm": 0.12408823519945145,
+      "learning_rate": 0.0001,
+      "loss": 1.4709,
+      "step": 1539
+    },
+    {
+      "epoch": 0.5155674589889522,
+      "grad_norm": 0.12711866199970245,
+      "learning_rate": 0.0001,
+      "loss": 1.5529,
+      "step": 1540
+    },
+    {
+      "epoch": 0.5159022430532306,
+      "grad_norm": 0.12433881312608719,
+      "learning_rate": 0.0001,
+      "loss": 1.4641,
+      "step": 1541
+    },
+    {
+      "epoch": 0.5162370271175092,
+      "grad_norm": 0.13031256198883057,
+      "learning_rate": 0.0001,
+      "loss": 1.6042,
+      "step": 1542
+    },
+    {
+      "epoch": 0.5165718111817877,
+      "grad_norm": 0.1294173002243042,
+      "learning_rate": 0.0001,
+      "loss": 1.5269,
+      "step": 1543
+    },
+    {
+      "epoch": 0.5169065952460663,
+      "grad_norm": 0.1273273229598999,
+      "learning_rate": 0.0001,
+      "loss": 1.5984,
+      "step": 1544
+    },
+    {
+      "epoch": 0.5172413793103449,
+      "grad_norm": 0.13191919028759003,
+      "learning_rate": 0.0001,
+      "loss": 1.5684,
+      "step": 1545
+    },
+    {
+      "epoch": 0.5175761633746234,
+      "grad_norm": 0.13768093287944794,
+      "learning_rate": 0.0001,
+      "loss": 1.555,
+      "step": 1546
+    },
+    {
+      "epoch": 0.5179109474389019,
+      "grad_norm": 0.12926150858402252,
+      "learning_rate": 0.0001,
+      "loss": 1.4731,
+      "step": 1547
+    },
+    {
+      "epoch": 0.5182457315031804,
+      "grad_norm": 0.12586715817451477,
+      "learning_rate": 0.0001,
+      "loss": 1.4794,
+      "step": 1548
+    },
+    {
+      "epoch": 0.518580515567459,
+      "grad_norm": 0.12548579275608063,
+      "learning_rate": 0.0001,
+      "loss": 1.5266,
+      "step": 1549
+    },
+    {
+      "epoch": 0.5189152996317375,
+      "grad_norm": 0.12171539664268494,
+      "learning_rate": 0.0001,
+      "loss": 1.4205,
+      "step": 1550
+    },
+    {
+      "epoch": 0.5192500836960161,
+      "grad_norm": 0.13130709528923035,
+      "learning_rate": 0.0001,
+      "loss": 1.5927,
+      "step": 1551
+    },
+    {
+      "epoch": 0.5195848677602946,
+      "grad_norm": 0.1342555582523346,
+      "learning_rate": 0.0001,
+      "loss": 1.5756,
+      "step": 1552
+    },
+    {
+      "epoch": 0.5199196518245731,
+      "grad_norm": 0.12991021573543549,
+      "learning_rate": 0.0001,
+      "loss": 1.646,
+      "step": 1553
+    },
+    {
+      "epoch": 0.5202544358888517,
+      "grad_norm": 0.13074184954166412,
+      "learning_rate": 0.0001,
+      "loss": 1.4619,
+      "step": 1554
+    },
+    {
+      "epoch": 0.5205892199531302,
+      "grad_norm": 0.12969058752059937,
+      "learning_rate": 0.0001,
+      "loss": 1.5048,
+      "step": 1555
+    },
+    {
+      "epoch": 0.5209240040174088,
+      "grad_norm": 0.12283259630203247,
+      "learning_rate": 0.0001,
+      "loss": 1.4968,
+      "step": 1556
+    },
+    {
+      "epoch": 0.5212587880816874,
+      "grad_norm": 0.14244720339775085,
+      "learning_rate": 0.0001,
+      "loss": 1.5984,
+      "step": 1557
+    },
+    {
+      "epoch": 0.5215935721459658,
+      "grad_norm": 0.12856322526931763,
+      "learning_rate": 0.0001,
+      "loss": 1.5382,
+      "step": 1558
+    },
+    {
+      "epoch": 0.5219283562102444,
+      "grad_norm": 0.1262657344341278,
+      "learning_rate": 0.0001,
+      "loss": 1.5191,
+      "step": 1559
+    },
+    {
+      "epoch": 0.5222631402745229,
+      "grad_norm": 0.1350589543581009,
+      "learning_rate": 0.0001,
+      "loss": 1.5812,
+      "step": 1560
+    },
+    {
+      "epoch": 0.5225979243388015,
+      "grad_norm": 0.13602742552757263,
+      "learning_rate": 0.0001,
+      "loss": 1.6252,
+      "step": 1561
+    },
+    {
+      "epoch": 0.52293270840308,
+      "grad_norm": 0.1273350566625595,
+      "learning_rate": 0.0001,
+      "loss": 1.5607,
+      "step": 1562
+    },
+    {
+      "epoch": 0.5232674924673586,
+      "grad_norm": 0.1261235773563385,
+      "learning_rate": 0.0001,
+      "loss": 1.4537,
+      "step": 1563
+    },
+    {
+      "epoch": 0.523602276531637,
+      "grad_norm": 0.123395174741745,
+      "learning_rate": 0.0001,
+      "loss": 1.4839,
+      "step": 1564
+    },
+    {
+      "epoch": 0.5239370605959156,
+      "grad_norm": 0.12707623839378357,
+      "learning_rate": 0.0001,
+      "loss": 1.5671,
+      "step": 1565
+    },
+    {
+      "epoch": 0.5242718446601942,
+      "grad_norm": 0.119587741792202,
+      "learning_rate": 0.0001,
+      "loss": 1.4637,
+      "step": 1566
+    },
+    {
+      "epoch": 0.5246066287244727,
+      "grad_norm": 0.12568604946136475,
+      "learning_rate": 0.0001,
+      "loss": 1.5196,
+      "step": 1567
+    },
+    {
+      "epoch": 0.5249414127887513,
+      "grad_norm": 0.13292740285396576,
+      "learning_rate": 0.0001,
+      "loss": 1.5909,
+      "step": 1568
+    },
+    {
+      "epoch": 0.5252761968530298,
+      "grad_norm": 0.13198155164718628,
+      "learning_rate": 0.0001,
+      "loss": 1.6039,
+      "step": 1569
+    },
+    {
+      "epoch": 0.5256109809173083,
+      "grad_norm": 0.12587766349315643,
+      "learning_rate": 0.0001,
+      "loss": 1.5418,
+      "step": 1570
+    },
+    {
+      "epoch": 0.5259457649815868,
+      "grad_norm": 0.12726300954818726,
+      "learning_rate": 0.0001,
+      "loss": 1.5366,
+      "step": 1571
+    },
+    {
+      "epoch": 0.5262805490458654,
+      "grad_norm": 0.12479355186223984,
+      "learning_rate": 0.0001,
+      "loss": 1.5486,
+      "step": 1572
+    },
+    {
+      "epoch": 0.526615333110144,
+      "grad_norm": 0.1242307722568512,
+      "learning_rate": 0.0001,
+      "loss": 1.4547,
+      "step": 1573
+    },
+    {
+      "epoch": 0.5269501171744225,
+      "grad_norm": 0.12753188610076904,
+      "learning_rate": 0.0001,
+      "loss": 1.6649,
+      "step": 1574
+    },
+    {
+      "epoch": 0.5272849012387011,
+      "grad_norm": 0.12815521657466888,
+      "learning_rate": 0.0001,
+      "loss": 1.4489,
+      "step": 1575
+    },
+    {
+      "epoch": 0.5276196853029795,
+      "grad_norm": 0.1192578375339508,
+      "learning_rate": 0.0001,
+      "loss": 1.4078,
+      "step": 1576
+    },
+    {
+      "epoch": 0.5279544693672581,
+      "grad_norm": 0.12596169114112854,
+      "learning_rate": 0.0001,
+      "loss": 1.5369,
+      "step": 1577
+    },
+    {
+      "epoch": 0.5282892534315367,
+      "grad_norm": 0.13193419575691223,
+      "learning_rate": 0.0001,
+      "loss": 1.5601,
+      "step": 1578
+    },
+    {
+      "epoch": 0.5286240374958152,
+      "grad_norm": 0.1277266889810562,
+      "learning_rate": 0.0001,
+      "loss": 1.5336,
+      "step": 1579
+    },
+    {
+      "epoch": 0.5289588215600938,
+      "grad_norm": 0.12819704413414001,
+      "learning_rate": 0.0001,
+      "loss": 1.4713,
+      "step": 1580
+    },
+    {
+      "epoch": 0.5292936056243723,
+      "grad_norm": 0.1399090439081192,
+      "learning_rate": 0.0001,
+      "loss": 1.5978,
+      "step": 1581
+    },
+    {
+      "epoch": 0.5296283896886508,
+      "grad_norm": 0.1373160183429718,
+      "learning_rate": 0.0001,
+      "loss": 1.6614,
+      "step": 1582
+    },
+    {
+      "epoch": 0.5299631737529293,
+      "grad_norm": 0.1253012716770172,
+      "learning_rate": 0.0001,
+      "loss": 1.5317,
+      "step": 1583
+    },
+    {
+      "epoch": 0.5302979578172079,
+      "grad_norm": 0.124544158577919,
+      "learning_rate": 0.0001,
+      "loss": 1.4947,
+      "step": 1584
+    },
+    {
+      "epoch": 0.5306327418814865,
+      "grad_norm": 0.13060353696346283,
+      "learning_rate": 0.0001,
+      "loss": 1.5342,
+      "step": 1585
+    },
+    {
+      "epoch": 0.530967525945765,
+      "grad_norm": 0.12680500745773315,
+      "learning_rate": 0.0001,
+      "loss": 1.4597,
+      "step": 1586
+    },
+    {
+      "epoch": 0.5313023100100436,
+      "grad_norm": 0.13112664222717285,
+      "learning_rate": 0.0001,
+      "loss": 1.5978,
+      "step": 1587
+    },
+    {
+      "epoch": 0.531637094074322,
+      "grad_norm": 0.13016077876091003,
+      "learning_rate": 0.0001,
+      "loss": 1.5575,
+      "step": 1588
+    },
+    {
+      "epoch": 0.5319718781386006,
+      "grad_norm": 0.1273767054080963,
+      "learning_rate": 0.0001,
+      "loss": 1.607,
+      "step": 1589
+    },
+    {
+      "epoch": 0.5323066622028791,
+      "grad_norm": 0.1310475915670395,
+      "learning_rate": 0.0001,
+      "loss": 1.5066,
+      "step": 1590
+    },
+    {
+      "epoch": 0.5326414462671577,
+      "grad_norm": 0.12938565015792847,
+      "learning_rate": 0.0001,
+      "loss": 1.4933,
+      "step": 1591
+    },
+    {
+      "epoch": 0.5329762303314363,
+      "grad_norm": 0.12316200882196426,
+      "learning_rate": 0.0001,
+      "loss": 1.4752,
+      "step": 1592
+    },
+    {
+      "epoch": 0.5333110143957147,
+      "grad_norm": 0.13205035030841827,
+      "learning_rate": 0.0001,
+      "loss": 1.5061,
+      "step": 1593
+    },
+    {
+      "epoch": 0.5336457984599933,
+      "grad_norm": 0.12517520785331726,
+      "learning_rate": 0.0001,
+      "loss": 1.5237,
+      "step": 1594
+    },
+    {
+      "epoch": 0.5339805825242718,
+      "grad_norm": 0.1309306025505066,
+      "learning_rate": 0.0001,
+      "loss": 1.5975,
+      "step": 1595
+    },
+    {
+      "epoch": 0.5343153665885504,
+      "grad_norm": 0.13565212488174438,
+      "learning_rate": 0.0001,
+      "loss": 1.6888,
+      "step": 1596
+    },
+    {
+      "epoch": 0.534650150652829,
+      "grad_norm": 0.13044795393943787,
+      "learning_rate": 0.0001,
+      "loss": 1.547,
+      "step": 1597
+    },
+    {
+      "epoch": 0.5349849347171075,
+      "grad_norm": 0.12757791578769684,
+      "learning_rate": 0.0001,
+      "loss": 1.5788,
+      "step": 1598
+    },
+    {
+      "epoch": 0.535319718781386,
+      "grad_norm": 0.12625539302825928,
+      "learning_rate": 0.0001,
+      "loss": 1.6271,
+      "step": 1599
+    },
+    {
+      "epoch": 0.5356545028456645,
+      "grad_norm": 0.12980274856090546,
+      "learning_rate": 0.0001,
+      "loss": 1.4808,
+      "step": 1600
+    },
+    {
+      "epoch": 0.5359892869099431,
+      "grad_norm": 0.1339329481124878,
+      "learning_rate": 0.0001,
+      "loss": 1.5838,
+      "step": 1601
+    },
+    {
+      "epoch": 0.5363240709742216,
+      "grad_norm": 0.13570533692836761,
+      "learning_rate": 0.0001,
+      "loss": 1.5526,
+      "step": 1602
+    },
+    {
+      "epoch": 0.5366588550385002,
+      "grad_norm": 0.13043223321437836,
+      "learning_rate": 0.0001,
+      "loss": 1.5046,
+      "step": 1603
+    },
+    {
+      "epoch": 0.5369936391027788,
+      "grad_norm": 0.1268492341041565,
+      "learning_rate": 0.0001,
+      "loss": 1.4846,
+      "step": 1604
+    },
+    {
+      "epoch": 0.5373284231670572,
+      "grad_norm": 0.12844318151474,
+      "learning_rate": 0.0001,
+      "loss": 1.622,
+      "step": 1605
+    },
+    {
+      "epoch": 0.5376632072313358,
+      "grad_norm": 0.12543794512748718,
+      "learning_rate": 0.0001,
+      "loss": 1.4895,
+      "step": 1606
+    },
+    {
+      "epoch": 0.5379979912956143,
+      "grad_norm": 0.13247263431549072,
+      "learning_rate": 0.0001,
+      "loss": 1.5431,
+      "step": 1607
+    },
+    {
+      "epoch": 0.5383327753598929,
+      "grad_norm": 0.12495877593755722,
+      "learning_rate": 0.0001,
+      "loss": 1.5534,
+      "step": 1608
+    },
+    {
+      "epoch": 0.5386675594241714,
+      "grad_norm": 0.12770773470401764,
+      "learning_rate": 0.0001,
+      "loss": 1.5296,
+      "step": 1609
+    },
+    {
+      "epoch": 0.53900234348845,
+      "grad_norm": 0.1249793991446495,
+      "learning_rate": 0.0001,
+      "loss": 1.549,
+      "step": 1610
+    },
+    {
+      "epoch": 0.5393371275527284,
+      "grad_norm": 0.13602420687675476,
+      "learning_rate": 0.0001,
+      "loss": 1.6911,
+      "step": 1611
+    },
+    {
+      "epoch": 0.539671911617007,
+      "grad_norm": 0.1260257512331009,
+      "learning_rate": 0.0001,
+      "loss": 1.6155,
+      "step": 1612
+    },
+    {
+      "epoch": 0.5400066956812856,
+      "grad_norm": 0.13716067373752594,
+      "learning_rate": 0.0001,
+      "loss": 1.5017,
+      "step": 1613
+    },
+    {
+      "epoch": 0.5403414797455641,
+      "grad_norm": 0.12322457879781723,
+      "learning_rate": 0.0001,
+      "loss": 1.4567,
+      "step": 1614
+    },
+    {
+      "epoch": 0.5406762638098427,
+      "grad_norm": 0.1295168548822403,
+      "learning_rate": 0.0001,
+      "loss": 1.5388,
+      "step": 1615
+    },
+    {
+      "epoch": 0.5410110478741212,
+      "grad_norm": 0.13598200678825378,
+      "learning_rate": 0.0001,
+      "loss": 1.6189,
+      "step": 1616
+    },
+    {
+      "epoch": 0.5413458319383997,
+      "grad_norm": 0.12514351308345795,
+      "learning_rate": 0.0001,
+      "loss": 1.5957,
+      "step": 1617
+    },
+    {
+      "epoch": 0.5416806160026783,
+      "grad_norm": 0.13243642449378967,
+      "learning_rate": 0.0001,
+      "loss": 1.5211,
+      "step": 1618
+    },
+    {
+      "epoch": 0.5420154000669568,
+      "grad_norm": 0.14331547915935516,
+      "learning_rate": 0.0001,
+      "loss": 1.628,
+      "step": 1619
+    },
+    {
+      "epoch": 0.5423501841312354,
+      "grad_norm": 0.13204847276210785,
+      "learning_rate": 0.0001,
+      "loss": 1.6131,
+      "step": 1620
+    },
+    {
+      "epoch": 0.5426849681955139,
+      "grad_norm": 0.13828937709331512,
+      "learning_rate": 0.0001,
+      "loss": 1.6206,
+      "step": 1621
+    },
+    {
+      "epoch": 0.5430197522597925,
+      "grad_norm": 0.13166444003582,
+      "learning_rate": 0.0001,
+      "loss": 1.556,
+      "step": 1622
+    },
+    {
+      "epoch": 0.5433545363240709,
+      "grad_norm": 0.131551131606102,
+      "learning_rate": 0.0001,
+      "loss": 1.5884,
+      "step": 1623
+    },
+    {
+      "epoch": 0.5436893203883495,
+      "grad_norm": 0.1386868953704834,
+      "learning_rate": 0.0001,
+      "loss": 1.626,
+      "step": 1624
+    },
+    {
+      "epoch": 0.544024104452628,
+      "grad_norm": 0.12754793465137482,
+      "learning_rate": 0.0001,
+      "loss": 1.5419,
+      "step": 1625
+    },
+    {
+      "epoch": 0.5443588885169066,
+      "grad_norm": 0.13059911131858826,
+      "learning_rate": 0.0001,
+      "loss": 1.5886,
+      "step": 1626
+    },
+    {
+      "epoch": 0.5446936725811852,
+      "grad_norm": 0.13056625425815582,
+      "learning_rate": 0.0001,
+      "loss": 1.5093,
+      "step": 1627
+    },
+    {
+      "epoch": 0.5450284566454636,
+      "grad_norm": 0.12965354323387146,
+      "learning_rate": 0.0001,
+      "loss": 1.5766,
+      "step": 1628
+    },
+    {
+      "epoch": 0.5453632407097422,
+      "grad_norm": 0.12052886188030243,
+      "learning_rate": 0.0001,
+      "loss": 1.5315,
+      "step": 1629
+    },
+    {
+      "epoch": 0.5456980247740207,
+      "grad_norm": 0.12897798418998718,
+      "learning_rate": 0.0001,
+      "loss": 1.6129,
+      "step": 1630
+    },
+    {
+      "epoch": 0.5460328088382993,
+      "grad_norm": 0.12880270183086395,
+      "learning_rate": 0.0001,
+      "loss": 1.6111,
+      "step": 1631
+    },
+    {
+      "epoch": 0.5463675929025779,
+      "grad_norm": 0.13251414895057678,
+      "learning_rate": 0.0001,
+      "loss": 1.5786,
+      "step": 1632
+    },
+    {
+      "epoch": 0.5467023769668564,
+      "grad_norm": 0.13067522644996643,
+      "learning_rate": 0.0001,
+      "loss": 1.5724,
+      "step": 1633
+    },
+    {
+      "epoch": 0.5470371610311349,
+      "grad_norm": 0.127615824341774,
+      "learning_rate": 0.0001,
+      "loss": 1.4672,
+      "step": 1634
+    },
+    {
+      "epoch": 0.5473719450954134,
+      "grad_norm": 0.12785358726978302,
+      "learning_rate": 0.0001,
+      "loss": 1.4379,
+      "step": 1635
+    },
+    {
+      "epoch": 0.547706729159692,
+      "grad_norm": 0.1336808055639267,
+      "learning_rate": 0.0001,
+      "loss": 1.5894,
+      "step": 1636
+    },
+    {
+      "epoch": 0.5480415132239705,
+      "grad_norm": 0.12709666788578033,
+      "learning_rate": 0.0001,
+      "loss": 1.5646,
+      "step": 1637
+    },
+    {
+      "epoch": 0.5483762972882491,
+      "grad_norm": 0.1278083175420761,
+      "learning_rate": 0.0001,
+      "loss": 1.5481,
+      "step": 1638
+    },
+    {
+      "epoch": 0.5487110813525277,
+      "grad_norm": 0.1273607462644577,
+      "learning_rate": 0.0001,
+      "loss": 1.6099,
+      "step": 1639
+    },
+    {
+      "epoch": 0.5490458654168061,
+      "grad_norm": 0.13073420524597168,
+      "learning_rate": 0.0001,
+      "loss": 1.6554,
+      "step": 1640
+    },
+    {
+      "epoch": 0.5493806494810847,
+      "grad_norm": 0.12339271605014801,
+      "learning_rate": 0.0001,
+      "loss": 1.4866,
+      "step": 1641
+    },
+    {
+      "epoch": 0.5497154335453632,
+      "grad_norm": 0.12296874821186066,
+      "learning_rate": 0.0001,
+      "loss": 1.4542,
+      "step": 1642
+    },
+    {
+      "epoch": 0.5500502176096418,
+      "grad_norm": 0.12228816747665405,
+      "learning_rate": 0.0001,
+      "loss": 1.5008,
+      "step": 1643
+    },
+    {
+      "epoch": 0.5503850016739203,
+      "grad_norm": 0.12167999148368835,
+      "learning_rate": 0.0001,
+      "loss": 1.4793,
+      "step": 1644
+    },
+    {
+      "epoch": 0.5507197857381989,
+      "grad_norm": 0.1323646754026413,
+      "learning_rate": 0.0001,
+      "loss": 1.6053,
+      "step": 1645
+    },
+    {
+      "epoch": 0.5510545698024774,
+      "grad_norm": 0.13682882487773895,
+      "learning_rate": 0.0001,
+      "loss": 1.5962,
+      "step": 1646
+    },
+    {
+      "epoch": 0.5513893538667559,
+      "grad_norm": 0.13337336480617523,
+      "learning_rate": 0.0001,
+      "loss": 1.6422,
+      "step": 1647
+    },
+    {
+      "epoch": 0.5517241379310345,
+      "grad_norm": 0.12662284076213837,
+      "learning_rate": 0.0001,
+      "loss": 1.4729,
+      "step": 1648
+    },
+    {
+      "epoch": 0.552058921995313,
+      "grad_norm": 0.13070893287658691,
+      "learning_rate": 0.0001,
+      "loss": 1.5548,
+      "step": 1649
+    },
+    {
+      "epoch": 0.5523937060595916,
+      "grad_norm": 0.1237405389547348,
+      "learning_rate": 0.0001,
+      "loss": 1.5731,
+      "step": 1650
+    },
+    {
+      "epoch": 0.5527284901238702,
+      "grad_norm": 0.12684407830238342,
+      "learning_rate": 0.0001,
+      "loss": 1.5927,
+      "step": 1651
+    },
+    {
+      "epoch": 0.5530632741881486,
+      "grad_norm": 0.13257922232151031,
+      "learning_rate": 0.0001,
+      "loss": 1.6194,
+      "step": 1652
+    },
+    {
+      "epoch": 0.5533980582524272,
+      "grad_norm": 0.12506547570228577,
+      "learning_rate": 0.0001,
+      "loss": 1.4954,
+      "step": 1653
+    },
+    {
+      "epoch": 0.5537328423167057,
+      "grad_norm": 0.13652825355529785,
+      "learning_rate": 0.0001,
+      "loss": 1.5936,
+      "step": 1654
+    },
+    {
+      "epoch": 0.5540676263809843,
+      "grad_norm": 0.1281632035970688,
+      "learning_rate": 0.0001,
+      "loss": 1.5239,
+      "step": 1655
+    },
+    {
+      "epoch": 0.5544024104452628,
+      "grad_norm": 0.1302935630083084,
+      "learning_rate": 0.0001,
+      "loss": 1.5731,
+      "step": 1656
+    },
+    {
+      "epoch": 0.5547371945095414,
+      "grad_norm": 0.13843512535095215,
+      "learning_rate": 0.0001,
+      "loss": 1.6028,
+      "step": 1657
+    },
+    {
+      "epoch": 0.5550719785738198,
+      "grad_norm": 0.13132615387439728,
+      "learning_rate": 0.0001,
+      "loss": 1.5167,
+      "step": 1658
+    },
+    {
+      "epoch": 0.5554067626380984,
+      "grad_norm": 0.1269274204969406,
+      "learning_rate": 0.0001,
+      "loss": 1.3276,
+      "step": 1659
+    },
+    {
+      "epoch": 0.555741546702377,
+      "grad_norm": 0.14026238024234772,
+      "learning_rate": 0.0001,
+      "loss": 1.5699,
+      "step": 1660
+    },
+    {
+      "epoch": 0.5560763307666555,
+      "grad_norm": 0.13259948790073395,
+      "learning_rate": 0.0001,
+      "loss": 1.5627,
+      "step": 1661
+    },
+    {
+      "epoch": 0.5564111148309341,
+      "grad_norm": 0.1282505840063095,
+      "learning_rate": 0.0001,
+      "loss": 1.601,
+      "step": 1662
+    },
+    {
+      "epoch": 0.5567458988952126,
+      "grad_norm": 0.14385761320590973,
+      "learning_rate": 0.0001,
+      "loss": 1.5731,
+      "step": 1663
+    },
+    {
+      "epoch": 0.5570806829594911,
+      "grad_norm": 0.12249067425727844,
+      "learning_rate": 0.0001,
+      "loss": 1.5416,
+      "step": 1664
+    },
+    {
+      "epoch": 0.5574154670237697,
+      "grad_norm": 0.13182908296585083,
+      "learning_rate": 0.0001,
+      "loss": 1.5313,
+      "step": 1665
+    },
+    {
+      "epoch": 0.5577502510880482,
+      "grad_norm": 0.14085689187049866,
+      "learning_rate": 0.0001,
+      "loss": 1.5736,
+      "step": 1666
+    },
+    {
+      "epoch": 0.5580850351523268,
+      "grad_norm": 0.14808295667171478,
+      "learning_rate": 0.0001,
+      "loss": 1.6265,
+      "step": 1667
+    },
+    {
+      "epoch": 0.5584198192166053,
+      "grad_norm": 0.13931553065776825,
+      "learning_rate": 0.0001,
+      "loss": 1.5729,
+      "step": 1668
+    },
+    {
+      "epoch": 0.5587546032808838,
+      "grad_norm": 0.14633771777153015,
+      "learning_rate": 0.0001,
+      "loss": 1.5433,
+      "step": 1669
+    },
+    {
+      "epoch": 0.5590893873451623,
+      "grad_norm": 0.1228380873799324,
+      "learning_rate": 0.0001,
+      "loss": 1.544,
+      "step": 1670
+    },
+    {
+      "epoch": 0.5594241714094409,
+      "grad_norm": 0.12809088826179504,
+      "learning_rate": 0.0001,
+      "loss": 1.5724,
+      "step": 1671
+    },
+    {
+      "epoch": 0.5597589554737195,
+      "grad_norm": 0.13453969359397888,
+      "learning_rate": 0.0001,
+      "loss": 1.5062,
+      "step": 1672
+    },
+    {
+      "epoch": 0.560093739537998,
+      "grad_norm": 0.13969993591308594,
+      "learning_rate": 0.0001,
+      "loss": 1.6302,
+      "step": 1673
+    },
+    {
+      "epoch": 0.5604285236022766,
+      "grad_norm": 0.13022400438785553,
+      "learning_rate": 0.0001,
+      "loss": 1.6323,
+      "step": 1674
+    },
+    {
+      "epoch": 0.560763307666555,
+      "grad_norm": 0.13372890651226044,
+      "learning_rate": 0.0001,
+      "loss": 1.6017,
+      "step": 1675
+    },
+    {
+      "epoch": 0.5610980917308336,
+      "grad_norm": 0.1426994502544403,
+      "learning_rate": 0.0001,
+      "loss": 1.5737,
+      "step": 1676
+    },
+    {
+      "epoch": 0.5614328757951121,
+      "grad_norm": 0.1358005702495575,
+      "learning_rate": 0.0001,
+      "loss": 1.5812,
+      "step": 1677
+    },
+    {
+      "epoch": 0.5617676598593907,
+      "grad_norm": 0.1320638507604599,
+      "learning_rate": 0.0001,
+      "loss": 1.5414,
+      "step": 1678
+    },
+    {
+      "epoch": 0.5621024439236693,
+      "grad_norm": 0.13449324667453766,
+      "learning_rate": 0.0001,
+      "loss": 1.4752,
+      "step": 1679
+    },
+    {
+      "epoch": 0.5624372279879478,
+      "grad_norm": 0.13063769042491913,
+      "learning_rate": 0.0001,
+      "loss": 1.5002,
+      "step": 1680
+    },
+    {
+      "epoch": 0.5627720120522263,
+      "grad_norm": 0.12591435015201569,
+      "learning_rate": 0.0001,
+      "loss": 1.5331,
+      "step": 1681
+    },
+    {
+      "epoch": 0.5631067961165048,
+      "grad_norm": 0.144126296043396,
+      "learning_rate": 0.0001,
+      "loss": 1.6207,
+      "step": 1682
+    },
+    {
+      "epoch": 0.5634415801807834,
+      "grad_norm": 0.13355223834514618,
+      "learning_rate": 0.0001,
+      "loss": 1.546,
+      "step": 1683
+    },
+    {
+      "epoch": 0.563776364245062,
+      "grad_norm": 0.12519478797912598,
+      "learning_rate": 0.0001,
+      "loss": 1.5836,
+      "step": 1684
+    },
+    {
+      "epoch": 0.5641111483093405,
+      "grad_norm": 0.1350811868906021,
+      "learning_rate": 0.0001,
+      "loss": 1.577,
+      "step": 1685
+    },
+    {
+      "epoch": 0.5644459323736191,
+      "grad_norm": 0.14059753715991974,
+      "learning_rate": 0.0001,
+      "loss": 1.5457,
+      "step": 1686
+    },
+    {
+      "epoch": 0.5647807164378975,
+      "grad_norm": 0.13620074093341827,
+      "learning_rate": 0.0001,
+      "loss": 1.5318,
+      "step": 1687
+    },
+    {
+      "epoch": 0.5651155005021761,
+      "grad_norm": 0.13117417693138123,
+      "learning_rate": 0.0001,
+      "loss": 1.5413,
+      "step": 1688
+    },
+    {
+      "epoch": 0.5654502845664546,
+      "grad_norm": 0.14555278420448303,
+      "learning_rate": 0.0001,
+      "loss": 1.5775,
+      "step": 1689
+    },
+    {
+      "epoch": 0.5657850686307332,
+      "grad_norm": 0.12660092115402222,
+      "learning_rate": 0.0001,
+      "loss": 1.5034,
+      "step": 1690
+    },
+    {
+      "epoch": 0.5661198526950117,
+      "grad_norm": 0.12967108190059662,
+      "learning_rate": 0.0001,
+      "loss": 1.5755,
+      "step": 1691
+    },
+    {
+      "epoch": 0.5664546367592903,
+      "grad_norm": 0.13999544084072113,
+      "learning_rate": 0.0001,
+      "loss": 1.4471,
+      "step": 1692
+    },
+    {
+      "epoch": 0.5667894208235688,
+      "grad_norm": 0.13235735893249512,
+      "learning_rate": 0.0001,
+      "loss": 1.4967,
+      "step": 1693
+    },
+    {
+      "epoch": 0.5671242048878473,
+      "grad_norm": 0.1373562067747116,
+      "learning_rate": 0.0001,
+      "loss": 1.6267,
+      "step": 1694
+    },
+    {
+      "epoch": 0.5674589889521259,
+      "grad_norm": 0.1320851445198059,
+      "learning_rate": 0.0001,
+      "loss": 1.5259,
+      "step": 1695
+    },
+    {
+      "epoch": 0.5677937730164044,
+      "grad_norm": 0.13309001922607422,
+      "learning_rate": 0.0001,
+      "loss": 1.5604,
+      "step": 1696
+    },
+    {
+      "epoch": 0.568128557080683,
+      "grad_norm": 0.12666000425815582,
+      "learning_rate": 0.0001,
+      "loss": 1.5352,
+      "step": 1697
+    },
+    {
+      "epoch": 0.5684633411449616,
+      "grad_norm": 0.12397143244743347,
+      "learning_rate": 0.0001,
+      "loss": 1.5474,
+      "step": 1698
+    },
+    {
+      "epoch": 0.56879812520924,
+      "grad_norm": 0.1286936104297638,
+      "learning_rate": 0.0001,
+      "loss": 1.5125,
+      "step": 1699
+    },
+    {
+      "epoch": 0.5691329092735186,
+      "grad_norm": 0.12525172531604767,
+      "learning_rate": 0.0001,
+      "loss": 1.4172,
+      "step": 1700
+    },
+    {
+      "epoch": 0.5694676933377971,
+      "grad_norm": 0.13234922289848328,
+      "learning_rate": 0.0001,
+      "loss": 1.5374,
+      "step": 1701
+    },
+    {
+      "epoch": 0.5698024774020757,
+      "grad_norm": 0.13341423869132996,
+      "learning_rate": 0.0001,
+      "loss": 1.5615,
+      "step": 1702
+    },
+    {
+      "epoch": 0.5701372614663542,
+      "grad_norm": 0.12672466039657593,
+      "learning_rate": 0.0001,
+      "loss": 1.4147,
+      "step": 1703
+    },
+    {
+      "epoch": 0.5704720455306327,
+      "grad_norm": 0.13073183596134186,
+      "learning_rate": 0.0001,
+      "loss": 1.5237,
+      "step": 1704
+    },
+    {
+      "epoch": 0.5708068295949112,
+      "grad_norm": 0.13044412434101105,
+      "learning_rate": 0.0001,
+      "loss": 1.6044,
+      "step": 1705
+    },
+    {
+      "epoch": 0.5711416136591898,
+      "grad_norm": 0.13865146040916443,
+      "learning_rate": 0.0001,
+      "loss": 1.5648,
+      "step": 1706
+    },
+    {
+      "epoch": 0.5714763977234684,
+      "grad_norm": 0.13418787717819214,
+      "learning_rate": 0.0001,
+      "loss": 1.5948,
+      "step": 1707
+    },
+    {
+      "epoch": 0.5718111817877469,
+      "grad_norm": 0.1279216855764389,
+      "learning_rate": 0.0001,
+      "loss": 1.5465,
+      "step": 1708
+    },
+    {
+      "epoch": 0.5721459658520255,
+      "grad_norm": 0.13305789232254028,
+      "learning_rate": 0.0001,
+      "loss": 1.5768,
+      "step": 1709
+    },
+    {
+      "epoch": 0.5724807499163039,
+      "grad_norm": 0.12358289957046509,
+      "learning_rate": 0.0001,
+      "loss": 1.4377,
+      "step": 1710
+    },
+    {
+      "epoch": 0.5728155339805825,
+      "grad_norm": 0.128280371427536,
+      "learning_rate": 0.0001,
+      "loss": 1.5684,
+      "step": 1711
+    },
+    {
+      "epoch": 0.573150318044861,
+      "grad_norm": 0.1336420327425003,
+      "learning_rate": 0.0001,
+      "loss": 1.5438,
+      "step": 1712
+    },
+    {
+      "epoch": 0.5734851021091396,
+      "grad_norm": 0.13142135739326477,
+      "learning_rate": 0.0001,
+      "loss": 1.5821,
+      "step": 1713
+    },
+    {
+      "epoch": 0.5738198861734182,
+      "grad_norm": 0.1367759257555008,
+      "learning_rate": 0.0001,
+      "loss": 1.5294,
+      "step": 1714
+    },
+    {
+      "epoch": 0.5741546702376967,
+      "grad_norm": 0.1364768147468567,
+      "learning_rate": 0.0001,
+      "loss": 1.4889,
+      "step": 1715
+    },
+    {
+      "epoch": 0.5744894543019752,
+      "grad_norm": 0.12675487995147705,
+      "learning_rate": 0.0001,
+      "loss": 1.5789,
+      "step": 1716
+    },
+    {
+      "epoch": 0.5748242383662537,
+      "grad_norm": 0.13054460287094116,
+      "learning_rate": 0.0001,
+      "loss": 1.5653,
+      "step": 1717
+    },
+    {
+      "epoch": 0.5751590224305323,
+      "grad_norm": 0.14481523633003235,
+      "learning_rate": 0.0001,
+      "loss": 1.6135,
+      "step": 1718
+    },
+    {
+      "epoch": 0.5754938064948109,
+      "grad_norm": 0.1317768394947052,
+      "learning_rate": 0.0001,
+      "loss": 1.5015,
+      "step": 1719
+    },
+    {
+      "epoch": 0.5758285905590894,
+      "grad_norm": 0.13205017149448395,
+      "learning_rate": 0.0001,
+      "loss": 1.5667,
+      "step": 1720
+    },
+    {
+      "epoch": 0.576163374623368,
+      "grad_norm": 0.13702328503131866,
+      "learning_rate": 0.0001,
+      "loss": 1.5487,
+      "step": 1721
+    },
+    {
+      "epoch": 0.5764981586876464,
+      "grad_norm": 0.13435296714305878,
+      "learning_rate": 0.0001,
+      "loss": 1.6059,
+      "step": 1722
+    },
+    {
+      "epoch": 0.576832942751925,
+      "grad_norm": 0.13013921678066254,
+      "learning_rate": 0.0001,
+      "loss": 1.5948,
+      "step": 1723
+    },
+    {
+      "epoch": 0.5771677268162035,
+      "grad_norm": 0.12254009395837784,
+      "learning_rate": 0.0001,
+      "loss": 1.485,
+      "step": 1724
+    },
+    {
+      "epoch": 0.5775025108804821,
+      "grad_norm": 0.13023540377616882,
+      "learning_rate": 0.0001,
+      "loss": 1.6237,
+      "step": 1725
+    },
+    {
+      "epoch": 0.5778372949447607,
+      "grad_norm": 0.1339290589094162,
+      "learning_rate": 0.0001,
+      "loss": 1.5983,
+      "step": 1726
+    },
+    {
+      "epoch": 0.5781720790090392,
+      "grad_norm": 0.13126787543296814,
+      "learning_rate": 0.0001,
+      "loss": 1.5947,
+      "step": 1727
+    },
+    {
+      "epoch": 0.5785068630733177,
+      "grad_norm": 0.12525591254234314,
+      "learning_rate": 0.0001,
+      "loss": 1.4519,
+      "step": 1728
+    },
+    {
+      "epoch": 0.5788416471375962,
+      "grad_norm": 0.12789173424243927,
+      "learning_rate": 0.0001,
+      "loss": 1.5293,
+      "step": 1729
+    },
+    {
+      "epoch": 0.5791764312018748,
+      "grad_norm": 0.12775948643684387,
+      "learning_rate": 0.0001,
+      "loss": 1.5971,
+      "step": 1730
+    },
+    {
+      "epoch": 0.5795112152661533,
+      "grad_norm": 0.13437266647815704,
+      "learning_rate": 0.0001,
+      "loss": 1.595,
+      "step": 1731
+    },
+    {
+      "epoch": 0.5798459993304319,
+      "grad_norm": 0.13249057531356812,
+      "learning_rate": 0.0001,
+      "loss": 1.5524,
+      "step": 1732
+    },
+    {
+      "epoch": 0.5801807833947105,
+      "grad_norm": 0.12838158011436462,
+      "learning_rate": 0.0001,
+      "loss": 1.4641,
+      "step": 1733
+    },
+    {
+      "epoch": 0.5805155674589889,
+      "grad_norm": 0.1311095654964447,
+      "learning_rate": 0.0001,
+      "loss": 1.5964,
+      "step": 1734
+    },
+    {
+      "epoch": 0.5808503515232675,
+      "grad_norm": 0.12928825616836548,
+      "learning_rate": 0.0001,
+      "loss": 1.5153,
+      "step": 1735
+    },
+    {
+      "epoch": 0.581185135587546,
+      "grad_norm": 0.1317373663187027,
+      "learning_rate": 0.0001,
+      "loss": 1.5805,
+      "step": 1736
+    },
+    {
+      "epoch": 0.5815199196518246,
+      "grad_norm": 0.1291595846414566,
+      "learning_rate": 0.0001,
+      "loss": 1.4974,
+      "step": 1737
+    },
+    {
+      "epoch": 0.5818547037161031,
+      "grad_norm": 0.12890678644180298,
+      "learning_rate": 0.0001,
+      "loss": 1.5778,
+      "step": 1738
+    },
+    {
+      "epoch": 0.5821894877803817,
+      "grad_norm": 0.13605663180351257,
+      "learning_rate": 0.0001,
+      "loss": 1.5206,
+      "step": 1739
+    },
+    {
+      "epoch": 0.5825242718446602,
+      "grad_norm": 0.12535326182842255,
+      "learning_rate": 0.0001,
+      "loss": 1.4989,
+      "step": 1740
+    },
+    {
+      "epoch": 0.5828590559089387,
+      "grad_norm": 0.13682806491851807,
+      "learning_rate": 0.0001,
+      "loss": 1.5558,
+      "step": 1741
+    },
+    {
+      "epoch": 0.5831938399732173,
+      "grad_norm": 0.12900637090206146,
+      "learning_rate": 0.0001,
+      "loss": 1.5687,
+      "step": 1742
+    },
+    {
+      "epoch": 0.5835286240374958,
+      "grad_norm": 0.1287071406841278,
+      "learning_rate": 0.0001,
+      "loss": 1.5349,
+      "step": 1743
+    },
+    {
+      "epoch": 0.5838634081017744,
+      "grad_norm": 0.12810088694095612,
+      "learning_rate": 0.0001,
+      "loss": 1.5363,
+      "step": 1744
+    },
+    {
+      "epoch": 0.5841981921660528,
+      "grad_norm": 0.13105565309524536,
+      "learning_rate": 0.0001,
+      "loss": 1.5633,
+      "step": 1745
+    },
+    {
+      "epoch": 0.5845329762303314,
+      "grad_norm": 0.13414978981018066,
+      "learning_rate": 0.0001,
+      "loss": 1.5965,
+      "step": 1746
+    },
+    {
+      "epoch": 0.58486776029461,
+      "grad_norm": 0.12767766416072845,
+      "learning_rate": 0.0001,
+      "loss": 1.517,
+      "step": 1747
+    },
+    {
+      "epoch": 0.5852025443588885,
+      "grad_norm": 0.12798413634300232,
+      "learning_rate": 0.0001,
+      "loss": 1.4184,
+      "step": 1748
+    },
+    {
+      "epoch": 0.5855373284231671,
+      "grad_norm": 0.13183465600013733,
+      "learning_rate": 0.0001,
+      "loss": 1.4812,
+      "step": 1749
+    },
+    {
+      "epoch": 0.5858721124874456,
+      "grad_norm": 0.12950639426708221,
+      "learning_rate": 0.0001,
+      "loss": 1.4371,
+      "step": 1750
+    },
+    {
+      "epoch": 0.5862068965517241,
+      "grad_norm": 0.1397038698196411,
+      "learning_rate": 0.0001,
+      "loss": 1.5023,
+      "step": 1751
+    },
+    {
+      "epoch": 0.5865416806160026,
+      "grad_norm": 0.1396951824426651,
+      "learning_rate": 0.0001,
+      "loss": 1.5174,
+      "step": 1752
+    },
+    {
+      "epoch": 0.5868764646802812,
+      "grad_norm": 0.13188160955905914,
+      "learning_rate": 0.0001,
+      "loss": 1.511,
+      "step": 1753
+    },
+    {
+      "epoch": 0.5872112487445598,
+      "grad_norm": 0.13433519005775452,
+      "learning_rate": 0.0001,
+      "loss": 1.5214,
+      "step": 1754
+    },
+    {
+      "epoch": 0.5875460328088383,
+      "grad_norm": 0.13022519648075104,
+      "learning_rate": 0.0001,
+      "loss": 1.5629,
+      "step": 1755
+    },
+    {
+      "epoch": 0.5878808168731169,
+      "grad_norm": 0.12651024758815765,
+      "learning_rate": 0.0001,
+      "loss": 1.4469,
+      "step": 1756
+    },
+    {
+      "epoch": 0.5882156009373953,
+      "grad_norm": 0.13489894568920135,
+      "learning_rate": 0.0001,
+      "loss": 1.5363,
+      "step": 1757
+    },
+    {
+      "epoch": 0.5885503850016739,
+      "grad_norm": 0.13707391917705536,
+      "learning_rate": 0.0001,
+      "loss": 1.6495,
+      "step": 1758
+    },
+    {
+      "epoch": 0.5888851690659525,
+      "grad_norm": 0.12528660893440247,
+      "learning_rate": 0.0001,
+      "loss": 1.5296,
+      "step": 1759
+    },
+    {
+      "epoch": 0.589219953130231,
+      "grad_norm": 0.14160814881324768,
+      "learning_rate": 0.0001,
+      "loss": 1.5977,
+      "step": 1760
+    },
+    {
+      "epoch": 0.5895547371945096,
+      "grad_norm": 0.12557724118232727,
+      "learning_rate": 0.0001,
+      "loss": 1.4915,
+      "step": 1761
+    },
+    {
+      "epoch": 0.5898895212587881,
+      "grad_norm": 0.12706881761550903,
+      "learning_rate": 0.0001,
+      "loss": 1.5775,
+      "step": 1762
+    },
+    {
+      "epoch": 0.5902243053230666,
+      "grad_norm": 0.13343869149684906,
+      "learning_rate": 0.0001,
+      "loss": 1.6033,
+      "step": 1763
+    },
+    {
+      "epoch": 0.5905590893873451,
+      "grad_norm": 0.1284165382385254,
+      "learning_rate": 0.0001,
+      "loss": 1.5255,
+      "step": 1764
+    },
+    {
+      "epoch": 0.5908938734516237,
+      "grad_norm": 0.12860101461410522,
+      "learning_rate": 0.0001,
+      "loss": 1.4694,
+      "step": 1765
+    },
+    {
+      "epoch": 0.5912286575159023,
+      "grad_norm": 0.12808945775032043,
+      "learning_rate": 0.0001,
+      "loss": 1.6068,
+      "step": 1766
+    },
+    {
+      "epoch": 0.5915634415801808,
+      "grad_norm": 0.13219839334487915,
+      "learning_rate": 0.0001,
+      "loss": 1.5519,
+      "step": 1767
+    },
+    {
+      "epoch": 0.5918982256444594,
+      "grad_norm": 0.12471086531877518,
+      "learning_rate": 0.0001,
+      "loss": 1.4465,
+      "step": 1768
+    },
+    {
+      "epoch": 0.5922330097087378,
+      "grad_norm": 0.13721035420894623,
+      "learning_rate": 0.0001,
+      "loss": 1.5656,
+      "step": 1769
+    },
+    {
+      "epoch": 0.5925677937730164,
+      "grad_norm": 0.1299833208322525,
+      "learning_rate": 0.0001,
+      "loss": 1.4767,
+      "step": 1770
+    },
+    {
+      "epoch": 0.5929025778372949,
+      "grad_norm": 0.13570041954517365,
+      "learning_rate": 0.0001,
+      "loss": 1.5929,
+      "step": 1771
+    },
+    {
+      "epoch": 0.5932373619015735,
+      "grad_norm": 0.12360662966966629,
+      "learning_rate": 0.0001,
+      "loss": 1.4179,
+      "step": 1772
+    },
+    {
+      "epoch": 0.5935721459658521,
+      "grad_norm": 0.138414204120636,
+      "learning_rate": 0.0001,
+      "loss": 1.6123,
+      "step": 1773
+    },
+    {
+      "epoch": 0.5939069300301306,
+      "grad_norm": 0.1347961127758026,
+      "learning_rate": 0.0001,
+      "loss": 1.6135,
+      "step": 1774
+    },
+    {
+      "epoch": 0.5942417140944091,
+      "grad_norm": 0.1333123743534088,
+      "learning_rate": 0.0001,
+      "loss": 1.3935,
+      "step": 1775
+    },
+    {
+      "epoch": 0.5945764981586876,
+      "grad_norm": 0.13112439215183258,
+      "learning_rate": 0.0001,
+      "loss": 1.5531,
+      "step": 1776
+    },
+    {
+      "epoch": 0.5949112822229662,
+      "grad_norm": 0.1356613039970398,
+      "learning_rate": 0.0001,
+      "loss": 1.5338,
+      "step": 1777
+    },
+    {
+      "epoch": 0.5952460662872447,
+      "grad_norm": 0.13762056827545166,
+      "learning_rate": 0.0001,
+      "loss": 1.5684,
+      "step": 1778
+    },
+    {
+      "epoch": 0.5955808503515233,
+      "grad_norm": 0.13242678344249725,
+      "learning_rate": 0.0001,
+      "loss": 1.5946,
+      "step": 1779
+    },
+    {
+      "epoch": 0.5959156344158018,
+      "grad_norm": 0.1304038166999817,
+      "learning_rate": 0.0001,
+      "loss": 1.5634,
+      "step": 1780
+    },
+    {
+      "epoch": 0.5962504184800803,
+      "grad_norm": 0.13004854321479797,
+      "learning_rate": 0.0001,
+      "loss": 1.5612,
+      "step": 1781
+    },
+    {
+      "epoch": 0.5965852025443589,
+      "grad_norm": 0.13909399509429932,
+      "learning_rate": 0.0001,
+      "loss": 1.5613,
+      "step": 1782
+    },
+    {
+      "epoch": 0.5969199866086374,
+      "grad_norm": 0.13109537959098816,
+      "learning_rate": 0.0001,
+      "loss": 1.5769,
+      "step": 1783
+    },
+    {
+      "epoch": 0.597254770672916,
+      "grad_norm": 0.13889670372009277,
+      "learning_rate": 0.0001,
+      "loss": 1.5788,
+      "step": 1784
+    },
+    {
+      "epoch": 0.5975895547371946,
+      "grad_norm": 0.12981747090816498,
+      "learning_rate": 0.0001,
+      "loss": 1.5294,
+      "step": 1785
+    },
+    {
+      "epoch": 0.597924338801473,
+      "grad_norm": 0.12865106761455536,
+      "learning_rate": 0.0001,
+      "loss": 1.5907,
+      "step": 1786
+    },
+    {
+      "epoch": 0.5982591228657516,
+      "grad_norm": 0.13081815838813782,
+      "learning_rate": 0.0001,
+      "loss": 1.6513,
+      "step": 1787
+    },
+    {
+      "epoch": 0.5985939069300301,
+      "grad_norm": 0.1357847899198532,
+      "learning_rate": 0.0001,
+      "loss": 1.6925,
+      "step": 1788
+    },
+    {
+      "epoch": 0.5989286909943087,
+      "grad_norm": 0.1296125054359436,
+      "learning_rate": 0.0001,
+      "loss": 1.5362,
+      "step": 1789
+    },
+    {
+      "epoch": 0.5992634750585872,
+      "grad_norm": 0.13272371888160706,
+      "learning_rate": 0.0001,
+      "loss": 1.669,
+      "step": 1790
+    },
+    {
+      "epoch": 0.5995982591228658,
+      "grad_norm": 0.1340399980545044,
+      "learning_rate": 0.0001,
+      "loss": 1.5674,
+      "step": 1791
+    },
+    {
+      "epoch": 0.5999330431871442,
+      "grad_norm": 0.12497217208147049,
+      "learning_rate": 0.0001,
+      "loss": 1.4629,
+      "step": 1792
+    },
+    {
+      "epoch": 0.6002678272514228,
+      "grad_norm": 0.14285002648830414,
+      "learning_rate": 0.0001,
+      "loss": 1.5278,
+      "step": 1793
+    },
+    {
+      "epoch": 0.6006026113157014,
+      "grad_norm": 0.1328384429216385,
+      "learning_rate": 0.0001,
+      "loss": 1.5532,
+      "step": 1794
+    },
+    {
+      "epoch": 0.6009373953799799,
+      "grad_norm": 0.13168397545814514,
+      "learning_rate": 0.0001,
+      "loss": 1.6406,
+      "step": 1795
+    },
+    {
+      "epoch": 0.6012721794442585,
+      "grad_norm": 0.12567539513111115,
+      "learning_rate": 0.0001,
+      "loss": 1.5389,
+      "step": 1796
+    },
+    {
+      "epoch": 0.601606963508537,
+      "grad_norm": 0.13105528056621552,
+      "learning_rate": 0.0001,
+      "loss": 1.5754,
+      "step": 1797
+    },
+    {
+      "epoch": 0.6019417475728155,
+      "grad_norm": 0.1292327493429184,
+      "learning_rate": 0.0001,
+      "loss": 1.4713,
+      "step": 1798
+    },
+    {
+      "epoch": 0.602276531637094,
+      "grad_norm": 0.12788547575473785,
+      "learning_rate": 0.0001,
+      "loss": 1.5787,
+      "step": 1799
+    },
+    {
+      "epoch": 0.6026113157013726,
+      "grad_norm": 0.1307074874639511,
+      "learning_rate": 0.0001,
+      "loss": 1.6191,
+      "step": 1800
+    },
+    {
+      "epoch": 0.6029460997656512,
+      "grad_norm": 0.136485293507576,
+      "learning_rate": 0.0001,
+      "loss": 1.6063,
+      "step": 1801
+    },
+    {
+      "epoch": 0.6032808838299297,
+      "grad_norm": 0.12938566505908966,
+      "learning_rate": 0.0001,
+      "loss": 1.5466,
+      "step": 1802
+    },
+    {
+      "epoch": 0.6036156678942083,
+      "grad_norm": 0.12429405003786087,
+      "learning_rate": 0.0001,
+      "loss": 1.4672,
+      "step": 1803
+    },
+    {
+      "epoch": 0.6039504519584867,
+      "grad_norm": 0.12657684087753296,
+      "learning_rate": 0.0001,
+      "loss": 1.5159,
+      "step": 1804
+    },
+    {
+      "epoch": 0.6042852360227653,
+      "grad_norm": 0.13287223875522614,
+      "learning_rate": 0.0001,
+      "loss": 1.5838,
+      "step": 1805
+    },
+    {
+      "epoch": 0.6046200200870439,
+      "grad_norm": 0.13268281519412994,
+      "learning_rate": 0.0001,
+      "loss": 1.5282,
+      "step": 1806
+    },
+    {
+      "epoch": 0.6049548041513224,
+      "grad_norm": 0.1264685094356537,
+      "learning_rate": 0.0001,
+      "loss": 1.5795,
+      "step": 1807
+    },
+    {
+      "epoch": 0.605289588215601,
+      "grad_norm": 0.1276138424873352,
+      "learning_rate": 0.0001,
+      "loss": 1.4648,
+      "step": 1808
+    },
+    {
+      "epoch": 0.6056243722798795,
+      "grad_norm": 0.13063056766986847,
+      "learning_rate": 0.0001,
+      "loss": 1.5692,
+      "step": 1809
+    },
+    {
+      "epoch": 0.605959156344158,
+      "grad_norm": 0.12172877043485641,
+      "learning_rate": 0.0001,
+      "loss": 1.4785,
+      "step": 1810
+    },
+    {
+      "epoch": 0.6062939404084365,
+      "grad_norm": 0.13516037166118622,
+      "learning_rate": 0.0001,
+      "loss": 1.5316,
+      "step": 1811
+    },
+    {
+      "epoch": 0.6066287244727151,
+      "grad_norm": 0.12978719174861908,
+      "learning_rate": 0.0001,
+      "loss": 1.5103,
+      "step": 1812
+    },
+    {
+      "epoch": 0.6069635085369937,
+      "grad_norm": 0.1354977786540985,
+      "learning_rate": 0.0001,
+      "loss": 1.5368,
+      "step": 1813
+    },
+    {
+      "epoch": 0.6072982926012722,
+      "grad_norm": 0.12445911020040512,
+      "learning_rate": 0.0001,
+      "loss": 1.4966,
+      "step": 1814
+    },
+    {
+      "epoch": 0.6076330766655507,
+      "grad_norm": 0.13546685874462128,
+      "learning_rate": 0.0001,
+      "loss": 1.62,
+      "step": 1815
+    },
+    {
+      "epoch": 0.6079678607298292,
+      "grad_norm": 0.12861642241477966,
+      "learning_rate": 0.0001,
+      "loss": 1.5895,
+      "step": 1816
+    },
+    {
+      "epoch": 0.6083026447941078,
+      "grad_norm": 0.13455091416835785,
+      "learning_rate": 0.0001,
+      "loss": 1.5217,
+      "step": 1817
+    },
+    {
+      "epoch": 0.6086374288583863,
+      "grad_norm": 0.13514240086078644,
+      "learning_rate": 0.0001,
+      "loss": 1.5947,
+      "step": 1818
+    },
+    {
+      "epoch": 0.6089722129226649,
+      "grad_norm": 0.12753477692604065,
+      "learning_rate": 0.0001,
+      "loss": 1.492,
+      "step": 1819
+    },
+    {
+      "epoch": 0.6093069969869435,
+      "grad_norm": 0.1335463970899582,
+      "learning_rate": 0.0001,
+      "loss": 1.5806,
+      "step": 1820
+    },
+    {
+      "epoch": 0.6096417810512219,
+      "grad_norm": 0.14587751030921936,
+      "learning_rate": 0.0001,
+      "loss": 1.5679,
+      "step": 1821
+    },
+    {
+      "epoch": 0.6099765651155005,
+      "grad_norm": 0.13787920773029327,
+      "learning_rate": 0.0001,
+      "loss": 1.4759,
+      "step": 1822
+    },
+    {
+      "epoch": 0.610311349179779,
+      "grad_norm": 0.135360449552536,
+      "learning_rate": 0.0001,
+      "loss": 1.4968,
+      "step": 1823
+    },
+    {
+      "epoch": 0.6106461332440576,
+      "grad_norm": 0.13543657958507538,
+      "learning_rate": 0.0001,
+      "loss": 1.5321,
+      "step": 1824
+    },
+    {
+      "epoch": 0.6109809173083361,
+      "grad_norm": 0.127221018075943,
+      "learning_rate": 0.0001,
+      "loss": 1.5239,
+      "step": 1825
+    },
+    {
+      "epoch": 0.6113157013726147,
+      "grad_norm": 0.1439230740070343,
+      "learning_rate": 0.0001,
+      "loss": 1.6458,
+      "step": 1826
+    },
+    {
+      "epoch": 0.6116504854368932,
+      "grad_norm": 0.13141925632953644,
+      "learning_rate": 0.0001,
+      "loss": 1.504,
+      "step": 1827
+    },
+    {
+      "epoch": 0.6119852695011717,
+      "grad_norm": 0.12811610102653503,
+      "learning_rate": 0.0001,
+      "loss": 1.6137,
+      "step": 1828
+    },
+    {
+      "epoch": 0.6123200535654503,
+      "grad_norm": 0.13353578746318817,
+      "learning_rate": 0.0001,
+      "loss": 1.5209,
+      "step": 1829
+    },
+    {
+      "epoch": 0.6126548376297288,
+      "grad_norm": 0.13006985187530518,
+      "learning_rate": 0.0001,
+      "loss": 1.4776,
+      "step": 1830
+    },
+    {
+      "epoch": 0.6129896216940074,
+      "grad_norm": 0.1350172609090805,
+      "learning_rate": 0.0001,
+      "loss": 1.5994,
+      "step": 1831
+    },
+    {
+      "epoch": 0.613324405758286,
+      "grad_norm": 0.13640815019607544,
+      "learning_rate": 0.0001,
+      "loss": 1.6383,
+      "step": 1832
+    },
+    {
+      "epoch": 0.6136591898225644,
+      "grad_norm": 0.14161550998687744,
+      "learning_rate": 0.0001,
+      "loss": 1.5486,
+      "step": 1833
+    },
+    {
+      "epoch": 0.613993973886843,
+      "grad_norm": 0.12927186489105225,
+      "learning_rate": 0.0001,
+      "loss": 1.5166,
+      "step": 1834
+    },
+    {
+      "epoch": 0.6143287579511215,
+      "grad_norm": 0.1287536919116974,
+      "learning_rate": 0.0001,
+      "loss": 1.496,
+      "step": 1835
+    },
+    {
+      "epoch": 0.6146635420154001,
+      "grad_norm": 0.13734175264835358,
+      "learning_rate": 0.0001,
+      "loss": 1.5638,
+      "step": 1836
+    },
+    {
+      "epoch": 0.6149983260796786,
+      "grad_norm": 0.13784490525722504,
+      "learning_rate": 0.0001,
+      "loss": 1.593,
+      "step": 1837
+    },
+    {
+      "epoch": 0.6153331101439572,
+      "grad_norm": 0.1259312480688095,
+      "learning_rate": 0.0001,
+      "loss": 1.5208,
+      "step": 1838
+    },
+    {
+      "epoch": 0.6156678942082356,
+      "grad_norm": 0.15089771151542664,
+      "learning_rate": 0.0001,
+      "loss": 1.5251,
+      "step": 1839
+    },
+    {
+      "epoch": 0.6160026782725142,
+      "grad_norm": 0.14801523089408875,
+      "learning_rate": 0.0001,
+      "loss": 1.5706,
+      "step": 1840
+    },
+    {
+      "epoch": 0.6163374623367928,
+      "grad_norm": 0.1345253735780716,
+      "learning_rate": 0.0001,
+      "loss": 1.5695,
+      "step": 1841
+    },
+    {
+      "epoch": 0.6166722464010713,
+      "grad_norm": 0.15094773471355438,
+      "learning_rate": 0.0001,
+      "loss": 1.5744,
+      "step": 1842
+    },
+    {
+      "epoch": 0.6170070304653499,
+      "grad_norm": 0.13193759322166443,
+      "learning_rate": 0.0001,
+      "loss": 1.5345,
+      "step": 1843
+    },
+    {
+      "epoch": 0.6173418145296284,
+      "grad_norm": 0.12728765606880188,
+      "learning_rate": 0.0001,
+      "loss": 1.5026,
+      "step": 1844
+    },
+    {
+      "epoch": 0.6176765985939069,
+      "grad_norm": 0.14725570380687714,
+      "learning_rate": 0.0001,
+      "loss": 1.581,
+      "step": 1845
+    },
+    {
+      "epoch": 0.6180113826581854,
+      "grad_norm": 0.13824598491191864,
+      "learning_rate": 0.0001,
+      "loss": 1.5359,
+      "step": 1846
+    },
+    {
+      "epoch": 0.618346166722464,
+      "grad_norm": 0.12178414314985275,
+      "learning_rate": 0.0001,
+      "loss": 1.4936,
+      "step": 1847
+    },
+    {
+      "epoch": 0.6186809507867426,
+      "grad_norm": 0.156047984957695,
+      "learning_rate": 0.0001,
+      "loss": 1.5737,
+      "step": 1848
+    },
+    {
+      "epoch": 0.6190157348510211,
+      "grad_norm": 0.15707126259803772,
+      "learning_rate": 0.0001,
+      "loss": 1.6287,
+      "step": 1849
+    },
+    {
+      "epoch": 0.6193505189152997,
+      "grad_norm": 0.1378837376832962,
+      "learning_rate": 0.0001,
+      "loss": 1.616,
+      "step": 1850
+    },
+    {
+      "epoch": 0.6196853029795781,
+      "grad_norm": 0.1423729658126831,
+      "learning_rate": 0.0001,
+      "loss": 1.5409,
+      "step": 1851
+    },
+    {
+      "epoch": 0.6200200870438567,
+      "grad_norm": 0.16630493104457855,
+      "learning_rate": 0.0001,
+      "loss": 1.6264,
+      "step": 1852
+    },
+    {
+      "epoch": 0.6203548711081353,
+      "grad_norm": 0.13753686845302582,
+      "learning_rate": 0.0001,
+      "loss": 1.6104,
+      "step": 1853
+    },
+    {
+      "epoch": 0.6206896551724138,
+      "grad_norm": 0.13337332010269165,
+      "learning_rate": 0.0001,
+      "loss": 1.5104,
+      "step": 1854
+    },
+    {
+      "epoch": 0.6210244392366924,
+      "grad_norm": 0.14229977130889893,
+      "learning_rate": 0.0001,
+      "loss": 1.4228,
+      "step": 1855
+    },
+    {
+      "epoch": 0.6213592233009708,
+      "grad_norm": 0.1403966248035431,
+      "learning_rate": 0.0001,
+      "loss": 1.5623,
+      "step": 1856
+    },
+    {
+      "epoch": 0.6216940073652494,
+      "grad_norm": 0.12786665558815002,
+      "learning_rate": 0.0001,
+      "loss": 1.5058,
+      "step": 1857
+    },
+    {
+      "epoch": 0.6220287914295279,
+      "grad_norm": 0.14748771488666534,
+      "learning_rate": 0.0001,
+      "loss": 1.5004,
+      "step": 1858
+    },
+    {
+      "epoch": 0.6223635754938065,
+      "grad_norm": 0.14041772484779358,
+      "learning_rate": 0.0001,
+      "loss": 1.6154,
+      "step": 1859
+    },
+    {
+      "epoch": 0.6226983595580851,
+      "grad_norm": 0.1256851702928543,
+      "learning_rate": 0.0001,
+      "loss": 1.4634,
+      "step": 1860
+    },
+    {
+      "epoch": 0.6230331436223636,
+      "grad_norm": 0.12676502764225006,
+      "learning_rate": 0.0001,
+      "loss": 1.5163,
+      "step": 1861
+    },
+    {
+      "epoch": 0.6233679276866421,
+      "grad_norm": 0.14927968382835388,
+      "learning_rate": 0.0001,
+      "loss": 1.5686,
+      "step": 1862
+    },
+    {
+      "epoch": 0.6237027117509206,
+      "grad_norm": 0.1308298408985138,
+      "learning_rate": 0.0001,
+      "loss": 1.5032,
+      "step": 1863
+    },
+    {
+      "epoch": 0.6240374958151992,
+      "grad_norm": 0.13208165764808655,
+      "learning_rate": 0.0001,
+      "loss": 1.5519,
+      "step": 1864
+    },
+    {
+      "epoch": 0.6243722798794777,
+      "grad_norm": 0.13822416961193085,
+      "learning_rate": 0.0001,
+      "loss": 1.5664,
+      "step": 1865
+    },
+    {
+      "epoch": 0.6247070639437563,
+      "grad_norm": 0.13646993041038513,
+      "learning_rate": 0.0001,
+      "loss": 1.5361,
+      "step": 1866
+    },
+    {
+      "epoch": 0.6250418480080349,
+      "grad_norm": 0.1273556500673294,
+      "learning_rate": 0.0001,
+      "loss": 1.546,
+      "step": 1867
+    },
+    {
+      "epoch": 0.6253766320723133,
+      "grad_norm": 0.13555049896240234,
+      "learning_rate": 0.0001,
+      "loss": 1.5288,
+      "step": 1868
+    },
+    {
+      "epoch": 0.6257114161365919,
+      "grad_norm": 0.13126762211322784,
+      "learning_rate": 0.0001,
+      "loss": 1.4659,
+      "step": 1869
+    },
+    {
+      "epoch": 0.6260462002008704,
+      "grad_norm": 0.1348927766084671,
+      "learning_rate": 0.0001,
+      "loss": 1.5812,
+      "step": 1870
+    },
+    {
+      "epoch": 0.626380984265149,
+      "grad_norm": 0.1363980621099472,
+      "learning_rate": 0.0001,
+      "loss": 1.6506,
+      "step": 1871
+    },
+    {
+      "epoch": 0.6267157683294275,
+      "grad_norm": 0.13422980904579163,
+      "learning_rate": 0.0001,
+      "loss": 1.5298,
+      "step": 1872
+    },
+    {
+      "epoch": 0.6270505523937061,
+      "grad_norm": 0.12745925784111023,
+      "learning_rate": 0.0001,
+      "loss": 1.4898,
+      "step": 1873
+    },
+    {
+      "epoch": 0.6273853364579846,
+      "grad_norm": 0.1292264759540558,
+      "learning_rate": 0.0001,
+      "loss": 1.548,
+      "step": 1874
+    },
+    {
+      "epoch": 0.6277201205222631,
+      "grad_norm": 0.1412927806377411,
+      "learning_rate": 0.0001,
+      "loss": 1.5228,
+      "step": 1875
+    },
+    {
+      "epoch": 0.6280549045865417,
+      "grad_norm": 0.1328163594007492,
+      "learning_rate": 0.0001,
+      "loss": 1.5521,
+      "step": 1876
+    },
+    {
+      "epoch": 0.6283896886508202,
+      "grad_norm": 0.1258804351091385,
+      "learning_rate": 0.0001,
+      "loss": 1.4781,
+      "step": 1877
+    },
+    {
+      "epoch": 0.6287244727150988,
+      "grad_norm": 0.128944993019104,
+      "learning_rate": 0.0001,
+      "loss": 1.5123,
+      "step": 1878
+    },
+    {
+      "epoch": 0.6290592567793774,
+      "grad_norm": 0.1244087815284729,
+      "learning_rate": 0.0001,
+      "loss": 1.4386,
+      "step": 1879
+    },
+    {
+      "epoch": 0.6293940408436558,
+      "grad_norm": 0.12890097498893738,
+      "learning_rate": 0.0001,
+      "loss": 1.5266,
+      "step": 1880
+    },
+    {
+      "epoch": 0.6297288249079344,
+      "grad_norm": 0.1312391459941864,
+      "learning_rate": 0.0001,
+      "loss": 1.5395,
+      "step": 1881
+    },
+    {
+      "epoch": 0.6300636089722129,
+      "grad_norm": 0.13363149762153625,
+      "learning_rate": 0.0001,
+      "loss": 1.5721,
+      "step": 1882
+    },
+    {
+      "epoch": 0.6303983930364915,
+      "grad_norm": 0.13130998611450195,
+      "learning_rate": 0.0001,
+      "loss": 1.5542,
+      "step": 1883
+    },
+    {
+      "epoch": 0.63073317710077,
+      "grad_norm": 0.13050179183483124,
+      "learning_rate": 0.0001,
+      "loss": 1.5422,
+      "step": 1884
+    },
+    {
+      "epoch": 0.6310679611650486,
+      "grad_norm": 0.13548725843429565,
+      "learning_rate": 0.0001,
+      "loss": 1.5597,
+      "step": 1885
+    },
+    {
+      "epoch": 0.631402745229327,
+      "grad_norm": 0.13810521364212036,
+      "learning_rate": 0.0001,
+      "loss": 1.6428,
+      "step": 1886
+    },
+    {
+      "epoch": 0.6317375292936056,
+      "grad_norm": 0.12898769974708557,
+      "learning_rate": 0.0001,
+      "loss": 1.5091,
+      "step": 1887
+    },
+    {
+      "epoch": 0.6320723133578842,
+      "grad_norm": 0.13874949514865875,
+      "learning_rate": 0.0001,
+      "loss": 1.473,
+      "step": 1888
+    },
+    {
+      "epoch": 0.6324070974221627,
+      "grad_norm": 0.1275644749403,
+      "learning_rate": 0.0001,
+      "loss": 1.5844,
+      "step": 1889
+    },
+    {
+      "epoch": 0.6327418814864413,
+      "grad_norm": 0.13245896995067596,
+      "learning_rate": 0.0001,
+      "loss": 1.602,
+      "step": 1890
+    },
+    {
+      "epoch": 0.6330766655507197,
+      "grad_norm": 0.13937050104141235,
+      "learning_rate": 0.0001,
+      "loss": 1.6106,
+      "step": 1891
+    },
+    {
+      "epoch": 0.6334114496149983,
+      "grad_norm": 0.13569729030132294,
+      "learning_rate": 0.0001,
+      "loss": 1.523,
+      "step": 1892
+    },
+    {
+      "epoch": 0.6337462336792768,
+      "grad_norm": 0.1360468864440918,
+      "learning_rate": 0.0001,
+      "loss": 1.5032,
+      "step": 1893
+    },
+    {
+      "epoch": 0.6340810177435554,
+      "grad_norm": 0.12757538259029388,
+      "learning_rate": 0.0001,
+      "loss": 1.487,
+      "step": 1894
+    },
+    {
+      "epoch": 0.634415801807834,
+      "grad_norm": 0.13325755298137665,
+      "learning_rate": 0.0001,
+      "loss": 1.5386,
+      "step": 1895
+    },
+    {
+      "epoch": 0.6347505858721125,
+      "grad_norm": 0.1348341703414917,
+      "learning_rate": 0.0001,
+      "loss": 1.6195,
+      "step": 1896
+    },
+    {
+      "epoch": 0.635085369936391,
+      "grad_norm": 0.14284925162792206,
+      "learning_rate": 0.0001,
+      "loss": 1.636,
+      "step": 1897
+    },
+    {
+      "epoch": 0.6354201540006695,
+      "grad_norm": 0.12641146779060364,
+      "learning_rate": 0.0001,
+      "loss": 1.5172,
+      "step": 1898
+    },
+    {
+      "epoch": 0.6357549380649481,
+      "grad_norm": 0.1327671855688095,
+      "learning_rate": 0.0001,
+      "loss": 1.6519,
+      "step": 1899
+    },
+    {
+      "epoch": 0.6360897221292267,
+      "grad_norm": 0.13408274948596954,
+      "learning_rate": 0.0001,
+      "loss": 1.4722,
+      "step": 1900
+    },
+    {
+      "epoch": 0.6364245061935052,
+      "grad_norm": 0.13136939704418182,
+      "learning_rate": 0.0001,
+      "loss": 1.56,
+      "step": 1901
+    },
+    {
+      "epoch": 0.6367592902577838,
+      "grad_norm": 0.13018733263015747,
+      "learning_rate": 0.0001,
+      "loss": 1.5499,
+      "step": 1902
+    },
+    {
+      "epoch": 0.6370940743220622,
+      "grad_norm": 0.137217178940773,
+      "learning_rate": 0.0001,
+      "loss": 1.6224,
+      "step": 1903
+    },
+    {
+      "epoch": 0.6374288583863408,
+      "grad_norm": 0.12886135280132294,
+      "learning_rate": 0.0001,
+      "loss": 1.5993,
+      "step": 1904
+    },
+    {
+      "epoch": 0.6377636424506193,
+      "grad_norm": 0.12878277897834778,
+      "learning_rate": 0.0001,
+      "loss": 1.4407,
+      "step": 1905
+    },
+    {
+      "epoch": 0.6380984265148979,
+      "grad_norm": 0.12817195057868958,
+      "learning_rate": 0.0001,
+      "loss": 1.5113,
+      "step": 1906
+    },
+    {
+      "epoch": 0.6384332105791765,
+      "grad_norm": 0.12779603898525238,
+      "learning_rate": 0.0001,
+      "loss": 1.573,
+      "step": 1907
+    },
+    {
+      "epoch": 0.638767994643455,
+      "grad_norm": 0.13575701415538788,
+      "learning_rate": 0.0001,
+      "loss": 1.5689,
+      "step": 1908
+    },
+    {
+      "epoch": 0.6391027787077335,
+      "grad_norm": 0.1292586326599121,
+      "learning_rate": 0.0001,
+      "loss": 1.5853,
+      "step": 1909
+    },
+    {
+      "epoch": 0.639437562772012,
+      "grad_norm": 0.13209429383277893,
+      "learning_rate": 0.0001,
+      "loss": 1.5374,
+      "step": 1910
+    },
+    {
+      "epoch": 0.6397723468362906,
+      "grad_norm": 0.13795161247253418,
+      "learning_rate": 0.0001,
+      "loss": 1.5752,
+      "step": 1911
+    },
+    {
+      "epoch": 0.6401071309005691,
+      "grad_norm": 0.13106195628643036,
+      "learning_rate": 0.0001,
+      "loss": 1.5074,
+      "step": 1912
+    },
+    {
+      "epoch": 0.6404419149648477,
+      "grad_norm": 0.1364029496908188,
+      "learning_rate": 0.0001,
+      "loss": 1.4415,
+      "step": 1913
+    },
+    {
+      "epoch": 0.6407766990291263,
+      "grad_norm": 0.13437704741954803,
+      "learning_rate": 0.0001,
+      "loss": 1.5179,
+      "step": 1914
+    },
+    {
+      "epoch": 0.6411114830934047,
+      "grad_norm": 0.12899838387966156,
+      "learning_rate": 0.0001,
+      "loss": 1.4437,
+      "step": 1915
+    },
+    {
+      "epoch": 0.6414462671576833,
+      "grad_norm": 0.1336640864610672,
+      "learning_rate": 0.0001,
+      "loss": 1.4988,
+      "step": 1916
+    },
+    {
+      "epoch": 0.6417810512219618,
+      "grad_norm": 0.13116469979286194,
+      "learning_rate": 0.0001,
+      "loss": 1.5944,
+      "step": 1917
+    },
+    {
+      "epoch": 0.6421158352862404,
+      "grad_norm": 0.1323315054178238,
+      "learning_rate": 0.0001,
+      "loss": 1.6378,
+      "step": 1918
+    },
+    {
+      "epoch": 0.642450619350519,
+      "grad_norm": 0.13012604415416718,
+      "learning_rate": 0.0001,
+      "loss": 1.591,
+      "step": 1919
+    },
+    {
+      "epoch": 0.6427854034147975,
+      "grad_norm": 0.13358043134212494,
+      "learning_rate": 0.0001,
+      "loss": 1.4948,
+      "step": 1920
+    },
+    {
+      "epoch": 0.643120187479076,
+      "grad_norm": 0.13027198612689972,
+      "learning_rate": 0.0001,
+      "loss": 1.5749,
+      "step": 1921
+    },
+    {
+      "epoch": 0.6434549715433545,
+      "grad_norm": 0.11880921572446823,
+      "learning_rate": 0.0001,
+      "loss": 1.434,
+      "step": 1922
+    },
+    {
+      "epoch": 0.6437897556076331,
+      "grad_norm": 0.1275249421596527,
+      "learning_rate": 0.0001,
+      "loss": 1.5074,
+      "step": 1923
+    },
+    {
+      "epoch": 0.6441245396719116,
+      "grad_norm": 0.13402846455574036,
+      "learning_rate": 0.0001,
+      "loss": 1.6019,
+      "step": 1924
+    },
+    {
+      "epoch": 0.6444593237361902,
+      "grad_norm": 0.1263839304447174,
+      "learning_rate": 0.0001,
+      "loss": 1.494,
+      "step": 1925
+    },
+    {
+      "epoch": 0.6447941078004688,
+      "grad_norm": 0.12889358401298523,
+      "learning_rate": 0.0001,
+      "loss": 1.4811,
+      "step": 1926
+    },
+    {
+      "epoch": 0.6451288918647472,
+      "grad_norm": 0.13030682504177094,
+      "learning_rate": 0.0001,
+      "loss": 1.5573,
+      "step": 1927
+    },
+    {
+      "epoch": 0.6454636759290258,
+      "grad_norm": 0.12815749645233154,
+      "learning_rate": 0.0001,
+      "loss": 1.5839,
+      "step": 1928
+    },
+    {
+      "epoch": 0.6457984599933043,
+      "grad_norm": 0.13763943314552307,
+      "learning_rate": 0.0001,
+      "loss": 1.4967,
+      "step": 1929
+    },
+    {
+      "epoch": 0.6461332440575829,
+      "grad_norm": 0.12890425324440002,
+      "learning_rate": 0.0001,
+      "loss": 1.4861,
+      "step": 1930
+    },
+    {
+      "epoch": 0.6464680281218614,
+      "grad_norm": 0.13768140971660614,
+      "learning_rate": 0.0001,
+      "loss": 1.5095,
+      "step": 1931
+    },
+    {
+      "epoch": 0.6468028121861399,
+      "grad_norm": 0.1268666833639145,
+      "learning_rate": 0.0001,
+      "loss": 1.5237,
+      "step": 1932
+    },
+    {
+      "epoch": 0.6471375962504184,
+      "grad_norm": 0.13325713574886322,
+      "learning_rate": 0.0001,
+      "loss": 1.593,
+      "step": 1933
+    },
+    {
+      "epoch": 0.647472380314697,
+      "grad_norm": 0.13848131895065308,
+      "learning_rate": 0.0001,
+      "loss": 1.4935,
+      "step": 1934
+    },
+    {
+      "epoch": 0.6478071643789756,
+      "grad_norm": 0.1393735706806183,
+      "learning_rate": 0.0001,
+      "loss": 1.6234,
+      "step": 1935
+    },
+    {
+      "epoch": 0.6481419484432541,
+      "grad_norm": 0.1441955864429474,
+      "learning_rate": 0.0001,
+      "loss": 1.6218,
+      "step": 1936
+    },
+    {
+      "epoch": 0.6484767325075327,
+      "grad_norm": 0.13111312687397003,
+      "learning_rate": 0.0001,
+      "loss": 1.5639,
+      "step": 1937
+    },
+    {
+      "epoch": 0.6488115165718111,
+      "grad_norm": 0.12940305471420288,
+      "learning_rate": 0.0001,
+      "loss": 1.5864,
+      "step": 1938
+    },
+    {
+      "epoch": 0.6491463006360897,
+      "grad_norm": 0.13657227158546448,
+      "learning_rate": 0.0001,
+      "loss": 1.5125,
+      "step": 1939
+    },
+    {
+      "epoch": 0.6494810847003682,
+      "grad_norm": 0.12390992790460587,
+      "learning_rate": 0.0001,
+      "loss": 1.4631,
+      "step": 1940
+    },
+    {
+      "epoch": 0.6498158687646468,
+      "grad_norm": 0.1316480040550232,
+      "learning_rate": 0.0001,
+      "loss": 1.5343,
+      "step": 1941
+    },
+    {
+      "epoch": 0.6501506528289254,
+      "grad_norm": 0.13427673280239105,
+      "learning_rate": 0.0001,
+      "loss": 1.5456,
+      "step": 1942
+    },
+    {
+      "epoch": 0.6504854368932039,
+      "grad_norm": 0.1284562200307846,
+      "learning_rate": 0.0001,
+      "loss": 1.5017,
+      "step": 1943
+    },
+    {
+      "epoch": 0.6508202209574824,
+      "grad_norm": 0.13431181013584137,
+      "learning_rate": 0.0001,
+      "loss": 1.45,
+      "step": 1944
+    },
+    {
+      "epoch": 0.6511550050217609,
+      "grad_norm": 0.13080428540706635,
+      "learning_rate": 0.0001,
+      "loss": 1.5035,
+      "step": 1945
+    },
+    {
+      "epoch": 0.6514897890860395,
+      "grad_norm": 0.13691136240959167,
+      "learning_rate": 0.0001,
+      "loss": 1.5145,
+      "step": 1946
+    },
+    {
+      "epoch": 0.651824573150318,
+      "grad_norm": 0.12990237772464752,
+      "learning_rate": 0.0001,
+      "loss": 1.5393,
+      "step": 1947
+    },
+    {
+      "epoch": 0.6521593572145966,
+      "grad_norm": 0.12529443204402924,
+      "learning_rate": 0.0001,
+      "loss": 1.468,
+      "step": 1948
+    },
+    {
+      "epoch": 0.6524941412788752,
+      "grad_norm": 0.13029485940933228,
+      "learning_rate": 0.0001,
+      "loss": 1.5229,
+      "step": 1949
+    },
+    {
+      "epoch": 0.6528289253431536,
+      "grad_norm": 0.13873140513896942,
+      "learning_rate": 0.0001,
+      "loss": 1.5667,
+      "step": 1950
+    },
+    {
+      "epoch": 0.6531637094074322,
+      "grad_norm": 0.13176368176937103,
+      "learning_rate": 0.0001,
+      "loss": 1.4231,
+      "step": 1951
+    },
+    {
+      "epoch": 0.6534984934717107,
+      "grad_norm": 0.13046538829803467,
+      "learning_rate": 0.0001,
+      "loss": 1.5151,
+      "step": 1952
+    },
+    {
+      "epoch": 0.6538332775359893,
+      "grad_norm": 0.1290617287158966,
+      "learning_rate": 0.0001,
+      "loss": 1.6184,
+      "step": 1953
+    },
+    {
+      "epoch": 0.6541680616002679,
+      "grad_norm": 0.13826888799667358,
+      "learning_rate": 0.0001,
+      "loss": 1.5597,
+      "step": 1954
+    },
+    {
+      "epoch": 0.6545028456645464,
+      "grad_norm": 0.1341448426246643,
+      "learning_rate": 0.0001,
+      "loss": 1.5763,
+      "step": 1955
+    },
+    {
+      "epoch": 0.6548376297288249,
+      "grad_norm": 0.1293526589870453,
+      "learning_rate": 0.0001,
+      "loss": 1.4475,
+      "step": 1956
+    },
+    {
+      "epoch": 0.6551724137931034,
+      "grad_norm": 0.12727828323841095,
+      "learning_rate": 0.0001,
+      "loss": 1.5195,
+      "step": 1957
+    },
+    {
+      "epoch": 0.655507197857382,
+      "grad_norm": 0.13981108367443085,
+      "learning_rate": 0.0001,
+      "loss": 1.6515,
+      "step": 1958
+    },
+    {
+      "epoch": 0.6558419819216605,
+      "grad_norm": 0.1339573711156845,
+      "learning_rate": 0.0001,
+      "loss": 1.4506,
+      "step": 1959
+    },
+    {
+      "epoch": 0.6561767659859391,
+      "grad_norm": 0.13203227519989014,
+      "learning_rate": 0.0001,
+      "loss": 1.5553,
+      "step": 1960
+    },
+    {
+      "epoch": 0.6565115500502177,
+      "grad_norm": 0.1276148110628128,
+      "learning_rate": 0.0001,
+      "loss": 1.5442,
+      "step": 1961
+    },
+    {
+      "epoch": 0.6568463341144961,
+      "grad_norm": 0.13206414878368378,
+      "learning_rate": 0.0001,
+      "loss": 1.4193,
+      "step": 1962
+    },
+    {
+      "epoch": 0.6571811181787747,
+      "grad_norm": 0.14616969227790833,
+      "learning_rate": 0.0001,
+      "loss": 1.6147,
+      "step": 1963
+    },
+    {
+      "epoch": 0.6575159022430532,
+      "grad_norm": 0.13604846596717834,
+      "learning_rate": 0.0001,
+      "loss": 1.5652,
+      "step": 1964
+    },
+    {
+      "epoch": 0.6578506863073318,
+      "grad_norm": 0.13196608424186707,
+      "learning_rate": 0.0001,
+      "loss": 1.565,
+      "step": 1965
+    },
+    {
+      "epoch": 0.6581854703716103,
+      "grad_norm": 0.14214178919792175,
+      "learning_rate": 0.0001,
+      "loss": 1.5692,
+      "step": 1966
+    },
+    {
+      "epoch": 0.6585202544358888,
+      "grad_norm": 0.1290048062801361,
+      "learning_rate": 0.0001,
+      "loss": 1.5004,
+      "step": 1967
+    },
+    {
+      "epoch": 0.6588550385001674,
+      "grad_norm": 0.13306178152561188,
+      "learning_rate": 0.0001,
+      "loss": 1.5913,
+      "step": 1968
+    },
+    {
+      "epoch": 0.6591898225644459,
+      "grad_norm": 0.1337195485830307,
+      "learning_rate": 0.0001,
+      "loss": 1.5888,
+      "step": 1969
+    },
+    {
+      "epoch": 0.6595246066287245,
+      "grad_norm": 0.1345224380493164,
+      "learning_rate": 0.0001,
+      "loss": 1.5513,
+      "step": 1970
+    },
+    {
+      "epoch": 0.659859390693003,
+      "grad_norm": 0.12885946035385132,
+      "learning_rate": 0.0001,
+      "loss": 1.4686,
+      "step": 1971
+    },
+    {
+      "epoch": 0.6601941747572816,
+      "grad_norm": 0.1352531760931015,
+      "learning_rate": 0.0001,
+      "loss": 1.5958,
+      "step": 1972
+    },
+    {
+      "epoch": 0.66052895882156,
+      "grad_norm": 0.12501929700374603,
+      "learning_rate": 0.0001,
+      "loss": 1.4162,
+      "step": 1973
+    },
+    {
+      "epoch": 0.6608637428858386,
+      "grad_norm": 0.1291869580745697,
+      "learning_rate": 0.0001,
+      "loss": 1.4463,
+      "step": 1974
+    },
+    {
+      "epoch": 0.6611985269501172,
+      "grad_norm": 0.14670369029045105,
+      "learning_rate": 0.0001,
+      "loss": 1.4661,
+      "step": 1975
+    },
+    {
+      "epoch": 0.6615333110143957,
+      "grad_norm": 0.13643884658813477,
+      "learning_rate": 0.0001,
+      "loss": 1.5677,
+      "step": 1976
+    },
+    {
+      "epoch": 0.6618680950786743,
+      "grad_norm": 0.13746634125709534,
+      "learning_rate": 0.0001,
+      "loss": 1.4903,
+      "step": 1977
+    },
+    {
+      "epoch": 0.6622028791429528,
+      "grad_norm": 0.14677157998085022,
+      "learning_rate": 0.0001,
+      "loss": 1.5492,
+      "step": 1978
+    },
+    {
+      "epoch": 0.6625376632072313,
+      "grad_norm": 0.1345069259405136,
+      "learning_rate": 0.0001,
+      "loss": 1.6059,
+      "step": 1979
+    },
+    {
+      "epoch": 0.6628724472715098,
+      "grad_norm": 0.13783417642116547,
+      "learning_rate": 0.0001,
+      "loss": 1.5546,
+      "step": 1980
+    },
+    {
+      "epoch": 0.6632072313357884,
+      "grad_norm": 0.13266097009181976,
+      "learning_rate": 0.0001,
+      "loss": 1.4469,
+      "step": 1981
+    },
+    {
+      "epoch": 0.663542015400067,
+      "grad_norm": 0.13931085169315338,
+      "learning_rate": 0.0001,
+      "loss": 1.5797,
+      "step": 1982
+    },
+    {
+      "epoch": 0.6638767994643455,
+      "grad_norm": 0.13039837777614594,
+      "learning_rate": 0.0001,
+      "loss": 1.4508,
+      "step": 1983
+    },
+    {
+      "epoch": 0.6642115835286241,
+      "grad_norm": 0.13921616971492767,
+      "learning_rate": 0.0001,
+      "loss": 1.6177,
+      "step": 1984
+    },
+    {
+      "epoch": 0.6645463675929025,
+      "grad_norm": 0.1381753534078598,
+      "learning_rate": 0.0001,
+      "loss": 1.6578,
+      "step": 1985
+    },
+    {
+      "epoch": 0.6648811516571811,
+      "grad_norm": 0.1361846625804901,
+      "learning_rate": 0.0001,
+      "loss": 1.5422,
+      "step": 1986
+    },
+    {
+      "epoch": 0.6652159357214597,
+      "grad_norm": 0.14170324802398682,
+      "learning_rate": 0.0001,
+      "loss": 1.6339,
+      "step": 1987
+    },
+    {
+      "epoch": 0.6655507197857382,
+      "grad_norm": 0.13164804875850677,
+      "learning_rate": 0.0001,
+      "loss": 1.5623,
+      "step": 1988
+    },
+    {
+      "epoch": 0.6658855038500168,
+      "grad_norm": 0.13766439259052277,
+      "learning_rate": 0.0001,
+      "loss": 1.5661,
+      "step": 1989
+    },
+    {
+      "epoch": 0.6662202879142953,
+      "grad_norm": 0.1340639889240265,
+      "learning_rate": 0.0001,
+      "loss": 1.6035,
+      "step": 1990
+    },
+    {
+      "epoch": 0.6665550719785738,
+      "grad_norm": 0.132024347782135,
+      "learning_rate": 0.0001,
+      "loss": 1.6319,
+      "step": 1991
+    },
+    {
+      "epoch": 0.6668898560428523,
+      "grad_norm": 0.13272161781787872,
+      "learning_rate": 0.0001,
+      "loss": 1.4522,
+      "step": 1992
+    },
+    {
+      "epoch": 0.6672246401071309,
+      "grad_norm": 0.14372223615646362,
+      "learning_rate": 0.0001,
+      "loss": 1.581,
+      "step": 1993
+    },
+    {
+      "epoch": 0.6675594241714095,
+      "grad_norm": 0.13869139552116394,
+      "learning_rate": 0.0001,
+      "loss": 1.6178,
+      "step": 1994
+    },
+    {
+      "epoch": 0.667894208235688,
+      "grad_norm": 0.12776124477386475,
+      "learning_rate": 0.0001,
+      "loss": 1.51,
+      "step": 1995
+    },
+    {
+      "epoch": 0.6682289922999666,
+      "grad_norm": 0.13583005964756012,
+      "learning_rate": 0.0001,
+      "loss": 1.5771,
+      "step": 1996
+    },
+    {
+      "epoch": 0.668563776364245,
+      "grad_norm": 0.13394635915756226,
+      "learning_rate": 0.0001,
+      "loss": 1.5605,
+      "step": 1997
+    },
+    {
+      "epoch": 0.6688985604285236,
+      "grad_norm": 0.13842739164829254,
+      "learning_rate": 0.0001,
+      "loss": 1.5541,
+      "step": 1998
+    },
+    {
+      "epoch": 0.6692333444928021,
+      "grad_norm": 0.13265378773212433,
+      "learning_rate": 0.0001,
+      "loss": 1.5772,
+      "step": 1999
+    },
+    {
+      "epoch": 0.6695681285570807,
+      "grad_norm": 0.13662943243980408,
+      "learning_rate": 0.0001,
+      "loss": 1.591,
+      "step": 2000
+    },
+    {
+      "epoch": 0.6699029126213593,
+      "grad_norm": 0.12512929737567902,
+      "learning_rate": 0.0001,
+      "loss": 1.4162,
+      "step": 2001
+    },
+    {
+      "epoch": 0.6702376966856378,
+      "grad_norm": 0.1327543556690216,
+      "learning_rate": 0.0001,
+      "loss": 1.4978,
+      "step": 2002
+    },
+    {
+      "epoch": 0.6705724807499163,
+      "grad_norm": 0.13269194960594177,
+      "learning_rate": 0.0001,
+      "loss": 1.5998,
+      "step": 2003
+    },
+    {
+      "epoch": 0.6709072648141948,
+      "grad_norm": 0.14017336070537567,
+      "learning_rate": 0.0001,
+      "loss": 1.5785,
+      "step": 2004
+    },
+    {
+      "epoch": 0.6712420488784734,
+      "grad_norm": 0.1304367482662201,
+      "learning_rate": 0.0001,
+      "loss": 1.4781,
+      "step": 2005
+    },
+    {
+      "epoch": 0.671576832942752,
+      "grad_norm": 0.13442495465278625,
+      "learning_rate": 0.0001,
+      "loss": 1.5358,
+      "step": 2006
+    },
+    {
+      "epoch": 0.6719116170070305,
+      "grad_norm": 0.13490137457847595,
+      "learning_rate": 0.0001,
+      "loss": 1.6273,
+      "step": 2007
+    },
+    {
+      "epoch": 0.672246401071309,
+      "grad_norm": 0.1324394941329956,
+      "learning_rate": 0.0001,
+      "loss": 1.5884,
+      "step": 2008
+    },
+    {
+      "epoch": 0.6725811851355875,
+      "grad_norm": 0.12797103822231293,
+      "learning_rate": 0.0001,
+      "loss": 1.551,
+      "step": 2009
+    },
+    {
+      "epoch": 0.6729159691998661,
+      "grad_norm": 0.13374999165534973,
+      "learning_rate": 0.0001,
+      "loss": 1.5571,
+      "step": 2010
+    },
+    {
+      "epoch": 0.6732507532641446,
+      "grad_norm": 0.13020572066307068,
+      "learning_rate": 0.0001,
+      "loss": 1.4756,
+      "step": 2011
+    },
+    {
+      "epoch": 0.6735855373284232,
+      "grad_norm": 0.12501733005046844,
+      "learning_rate": 0.0001,
+      "loss": 1.5073,
+      "step": 2012
+    },
+    {
+      "epoch": 0.6739203213927017,
+      "grad_norm": 0.12433689087629318,
+      "learning_rate": 0.0001,
+      "loss": 1.4574,
+      "step": 2013
+    },
+    {
+      "epoch": 0.6742551054569802,
+      "grad_norm": 0.14026397466659546,
+      "learning_rate": 0.0001,
+      "loss": 1.5513,
+      "step": 2014
+    },
+    {
+      "epoch": 0.6745898895212588,
+      "grad_norm": 0.1340554803609848,
+      "learning_rate": 0.0001,
+      "loss": 1.5686,
+      "step": 2015
+    },
+    {
+      "epoch": 0.6749246735855373,
+      "grad_norm": 0.12796646356582642,
+      "learning_rate": 0.0001,
+      "loss": 1.4842,
+      "step": 2016
+    },
+    {
+      "epoch": 0.6752594576498159,
+      "grad_norm": 0.1362949162721634,
+      "learning_rate": 0.0001,
+      "loss": 1.5763,
+      "step": 2017
+    },
+    {
+      "epoch": 0.6755942417140944,
+      "grad_norm": 0.1347300410270691,
+      "learning_rate": 0.0001,
+      "loss": 1.5975,
+      "step": 2018
+    },
+    {
+      "epoch": 0.675929025778373,
+      "grad_norm": 0.13647662103176117,
+      "learning_rate": 0.0001,
+      "loss": 1.5395,
+      "step": 2019
+    },
+    {
+      "epoch": 0.6762638098426514,
+      "grad_norm": 0.13441947102546692,
+      "learning_rate": 0.0001,
+      "loss": 1.5726,
+      "step": 2020
+    },
+    {
+      "epoch": 0.67659859390693,
+      "grad_norm": 0.13435856997966766,
+      "learning_rate": 0.0001,
+      "loss": 1.6806,
+      "step": 2021
+    },
+    {
+      "epoch": 0.6769333779712086,
+      "grad_norm": 0.1239754781126976,
+      "learning_rate": 0.0001,
+      "loss": 1.4045,
+      "step": 2022
+    },
+    {
+      "epoch": 0.6772681620354871,
+      "grad_norm": 0.13493669033050537,
+      "learning_rate": 0.0001,
+      "loss": 1.5606,
+      "step": 2023
+    },
+    {
+      "epoch": 0.6776029460997657,
+      "grad_norm": 0.12938407063484192,
+      "learning_rate": 0.0001,
+      "loss": 1.5201,
+      "step": 2024
+    },
+    {
+      "epoch": 0.6779377301640442,
+      "grad_norm": 0.12213901430368423,
+      "learning_rate": 0.0001,
+      "loss": 1.4436,
+      "step": 2025
+    },
+    {
+      "epoch": 0.6782725142283227,
+      "grad_norm": 0.14107517898082733,
+      "learning_rate": 0.0001,
+      "loss": 1.5584,
+      "step": 2026
+    },
+    {
+      "epoch": 0.6786072982926012,
+      "grad_norm": 0.13082027435302734,
+      "learning_rate": 0.0001,
+      "loss": 1.5278,
+      "step": 2027
+    },
+    {
+      "epoch": 0.6789420823568798,
+      "grad_norm": 0.14623381197452545,
+      "learning_rate": 0.0001,
+      "loss": 1.668,
+      "step": 2028
+    },
+    {
+      "epoch": 0.6792768664211584,
+      "grad_norm": 0.12862159311771393,
+      "learning_rate": 0.0001,
+      "loss": 1.5534,
+      "step": 2029
+    },
+    {
+      "epoch": 0.6796116504854369,
+      "grad_norm": 0.13177117705345154,
+      "learning_rate": 0.0001,
+      "loss": 1.5564,
+      "step": 2030
+    },
+    {
+      "epoch": 0.6799464345497155,
+      "grad_norm": 0.12835298478603363,
+      "learning_rate": 0.0001,
+      "loss": 1.479,
+      "step": 2031
+    },
+    {
+      "epoch": 0.6802812186139939,
+      "grad_norm": 0.14096349477767944,
+      "learning_rate": 0.0001,
+      "loss": 1.6175,
+      "step": 2032
+    },
+    {
+      "epoch": 0.6806160026782725,
+      "grad_norm": 0.12646090984344482,
+      "learning_rate": 0.0001,
+      "loss": 1.4861,
+      "step": 2033
+    },
+    {
+      "epoch": 0.680950786742551,
+      "grad_norm": 0.137931689620018,
+      "learning_rate": 0.0001,
+      "loss": 1.5051,
+      "step": 2034
+    },
+    {
+      "epoch": 0.6812855708068296,
+      "grad_norm": 0.13240592181682587,
+      "learning_rate": 0.0001,
+      "loss": 1.5868,
+      "step": 2035
+    },
+    {
+      "epoch": 0.6816203548711082,
+      "grad_norm": 0.1362670511007309,
+      "learning_rate": 0.0001,
+      "loss": 1.5899,
+      "step": 2036
+    },
+    {
+      "epoch": 0.6819551389353867,
+      "grad_norm": 0.13148629665374756,
+      "learning_rate": 0.0001,
+      "loss": 1.521,
+      "step": 2037
+    },
+    {
+      "epoch": 0.6822899229996652,
+      "grad_norm": 0.13285885751247406,
+      "learning_rate": 0.0001,
+      "loss": 1.5122,
+      "step": 2038
+    },
+    {
+      "epoch": 0.6826247070639437,
+      "grad_norm": 0.1264655739068985,
+      "learning_rate": 0.0001,
+      "loss": 1.4886,
+      "step": 2039
+    },
+    {
+      "epoch": 0.6829594911282223,
+      "grad_norm": 0.12677529454231262,
+      "learning_rate": 0.0001,
+      "loss": 1.5068,
+      "step": 2040
+    },
+    {
+      "epoch": 0.6832942751925009,
+      "grad_norm": 0.13277101516723633,
+      "learning_rate": 0.0001,
+      "loss": 1.6065,
+      "step": 2041
+    },
+    {
+      "epoch": 0.6836290592567794,
+      "grad_norm": 0.13291488587856293,
+      "learning_rate": 0.0001,
+      "loss": 1.5755,
+      "step": 2042
+    },
+    {
+      "epoch": 0.6839638433210579,
+      "grad_norm": 0.13058260083198547,
+      "learning_rate": 0.0001,
+      "loss": 1.5286,
+      "step": 2043
+    },
+    {
+      "epoch": 0.6842986273853364,
+      "grad_norm": 0.13059435784816742,
+      "learning_rate": 0.0001,
+      "loss": 1.5803,
+      "step": 2044
+    },
+    {
+      "epoch": 0.684633411449615,
+      "grad_norm": 0.12917304039001465,
+      "learning_rate": 0.0001,
+      "loss": 1.576,
+      "step": 2045
+    },
+    {
+      "epoch": 0.6849681955138935,
+      "grad_norm": 0.12822791934013367,
+      "learning_rate": 0.0001,
+      "loss": 1.5201,
+      "step": 2046
+    },
+    {
+      "epoch": 0.6853029795781721,
+      "grad_norm": 0.14006927609443665,
+      "learning_rate": 0.0001,
+      "loss": 1.5445,
+      "step": 2047
+    },
+    {
+      "epoch": 0.6856377636424507,
+      "grad_norm": 0.13502942025661469,
+      "learning_rate": 0.0001,
+      "loss": 1.543,
+      "step": 2048
+    },
+    {
+      "epoch": 0.6859725477067291,
+      "grad_norm": 0.1351221352815628,
+      "learning_rate": 0.0001,
+      "loss": 1.5594,
+      "step": 2049
+    },
+    {
+      "epoch": 0.6863073317710077,
+      "grad_norm": 0.13474461436271667,
+      "learning_rate": 0.0001,
+      "loss": 1.5984,
+      "step": 2050
+    },
+    {
+      "epoch": 0.6866421158352862,
+      "grad_norm": 0.1317591369152069,
+      "learning_rate": 0.0001,
+      "loss": 1.5681,
+      "step": 2051
+    },
+    {
+      "epoch": 0.6869768998995648,
+      "grad_norm": 0.1300475299358368,
+      "learning_rate": 0.0001,
+      "loss": 1.5426,
+      "step": 2052
+    },
+    {
+      "epoch": 0.6873116839638433,
+      "grad_norm": 0.1308741718530655,
+      "learning_rate": 0.0001,
+      "loss": 1.5649,
+      "step": 2053
+    },
+    {
+      "epoch": 0.6876464680281219,
+      "grad_norm": 0.1339602768421173,
+      "learning_rate": 0.0001,
+      "loss": 1.5422,
+      "step": 2054
+    },
+    {
+      "epoch": 0.6879812520924004,
+      "grad_norm": 0.12556122243404388,
+      "learning_rate": 0.0001,
+      "loss": 1.3939,
+      "step": 2055
+    },
+    {
+      "epoch": 0.6883160361566789,
+      "grad_norm": 0.1331097036600113,
+      "learning_rate": 0.0001,
+      "loss": 1.5725,
+      "step": 2056
+    },
+    {
+      "epoch": 0.6886508202209575,
+      "grad_norm": 0.12769033014774323,
+      "learning_rate": 0.0001,
+      "loss": 1.5133,
+      "step": 2057
+    },
+    {
+      "epoch": 0.688985604285236,
+      "grad_norm": 0.13246020674705505,
+      "learning_rate": 0.0001,
+      "loss": 1.5533,
+      "step": 2058
+    },
+    {
+      "epoch": 0.6893203883495146,
+      "grad_norm": 0.13371361792087555,
+      "learning_rate": 0.0001,
+      "loss": 1.6253,
+      "step": 2059
+    },
+    {
+      "epoch": 0.6896551724137931,
+      "grad_norm": 0.1314792037010193,
+      "learning_rate": 0.0001,
+      "loss": 1.4943,
+      "step": 2060
+    },
+    {
+      "epoch": 0.6899899564780716,
+      "grad_norm": 0.13194666802883148,
+      "learning_rate": 0.0001,
+      "loss": 1.5983,
+      "step": 2061
+    },
+    {
+      "epoch": 0.6903247405423502,
+      "grad_norm": 0.13631388545036316,
+      "learning_rate": 0.0001,
+      "loss": 1.4932,
+      "step": 2062
+    },
+    {
+      "epoch": 0.6906595246066287,
+      "grad_norm": 0.1319463849067688,
+      "learning_rate": 0.0001,
+      "loss": 1.5848,
+      "step": 2063
+    },
+    {
+      "epoch": 0.6909943086709073,
+      "grad_norm": 0.14124637842178345,
+      "learning_rate": 0.0001,
+      "loss": 1.6066,
+      "step": 2064
+    },
+    {
+      "epoch": 0.6913290927351858,
+      "grad_norm": 0.12954577803611755,
+      "learning_rate": 0.0001,
+      "loss": 1.4153,
+      "step": 2065
+    },
+    {
+      "epoch": 0.6916638767994644,
+      "grad_norm": 0.1325748711824417,
+      "learning_rate": 0.0001,
+      "loss": 1.5766,
+      "step": 2066
+    },
+    {
+      "epoch": 0.6919986608637428,
+      "grad_norm": 0.13064290583133698,
+      "learning_rate": 0.0001,
+      "loss": 1.4995,
+      "step": 2067
+    },
+    {
+      "epoch": 0.6923334449280214,
+      "grad_norm": 0.1248745545744896,
+      "learning_rate": 0.0001,
+      "loss": 1.5077,
+      "step": 2068
+    },
+    {
+      "epoch": 0.6926682289923,
+      "grad_norm": 0.1278417706489563,
+      "learning_rate": 0.0001,
+      "loss": 1.5449,
+      "step": 2069
+    },
+    {
+      "epoch": 0.6930030130565785,
+      "grad_norm": 0.13311515748500824,
+      "learning_rate": 0.0001,
+      "loss": 1.5251,
+      "step": 2070
+    },
+    {
+      "epoch": 0.6933377971208571,
+      "grad_norm": 0.13218218088150024,
+      "learning_rate": 0.0001,
+      "loss": 1.5359,
+      "step": 2071
+    },
+    {
+      "epoch": 0.6936725811851356,
+      "grad_norm": 0.13042452931404114,
+      "learning_rate": 0.0001,
+      "loss": 1.5534,
+      "step": 2072
+    },
+    {
+      "epoch": 0.6940073652494141,
+      "grad_norm": 0.1393493264913559,
+      "learning_rate": 0.0001,
+      "loss": 1.594,
+      "step": 2073
+    },
+    {
+      "epoch": 0.6943421493136926,
+      "grad_norm": 0.1298573911190033,
+      "learning_rate": 0.0001,
+      "loss": 1.518,
+      "step": 2074
+    },
+    {
+      "epoch": 0.6946769333779712,
+      "grad_norm": 0.13325051963329315,
+      "learning_rate": 0.0001,
+      "loss": 1.5068,
+      "step": 2075
+    },
+    {
+      "epoch": 0.6950117174422498,
+      "grad_norm": 0.1269649714231491,
+      "learning_rate": 0.0001,
+      "loss": 1.4805,
+      "step": 2076
+    },
+    {
+      "epoch": 0.6953465015065283,
+      "grad_norm": 0.12699490785598755,
+      "learning_rate": 0.0001,
+      "loss": 1.4228,
+      "step": 2077
+    },
+    {
+      "epoch": 0.6956812855708069,
+      "grad_norm": 0.1379399597644806,
+      "learning_rate": 0.0001,
+      "loss": 1.4138,
+      "step": 2078
+    },
+    {
+      "epoch": 0.6960160696350853,
+      "grad_norm": 0.13343951106071472,
+      "learning_rate": 0.0001,
+      "loss": 1.5947,
+      "step": 2079
+    },
+    {
+      "epoch": 0.6963508536993639,
+      "grad_norm": 0.13461847603321075,
+      "learning_rate": 0.0001,
+      "loss": 1.5333,
+      "step": 2080
+    },
+    {
+      "epoch": 0.6966856377636425,
+      "grad_norm": 0.1299065202474594,
+      "learning_rate": 0.0001,
+      "loss": 1.5415,
+      "step": 2081
+    },
+    {
+      "epoch": 0.697020421827921,
+      "grad_norm": 0.1272873431444168,
+      "learning_rate": 0.0001,
+      "loss": 1.4443,
+      "step": 2082
+    },
+    {
+      "epoch": 0.6973552058921996,
+      "grad_norm": 0.136282280087471,
+      "learning_rate": 0.0001,
+      "loss": 1.4996,
+      "step": 2083
+    },
+    {
+      "epoch": 0.697689989956478,
+      "grad_norm": 0.12842769920825958,
+      "learning_rate": 0.0001,
+      "loss": 1.5574,
+      "step": 2084
+    },
+    {
+      "epoch": 0.6980247740207566,
+      "grad_norm": 0.12897315621376038,
+      "learning_rate": 0.0001,
+      "loss": 1.6162,
+      "step": 2085
+    },
+    {
+      "epoch": 0.6983595580850351,
+      "grad_norm": 0.13097885251045227,
+      "learning_rate": 0.0001,
+      "loss": 1.4949,
+      "step": 2086
+    },
+    {
+      "epoch": 0.6986943421493137,
+      "grad_norm": 0.13251438736915588,
+      "learning_rate": 0.0001,
+      "loss": 1.5041,
+      "step": 2087
+    },
+    {
+      "epoch": 0.6990291262135923,
+      "grad_norm": 0.1319066435098648,
+      "learning_rate": 0.0001,
+      "loss": 1.5499,
+      "step": 2088
+    },
+    {
+      "epoch": 0.6993639102778708,
+      "grad_norm": 0.13142657279968262,
+      "learning_rate": 0.0001,
+      "loss": 1.452,
+      "step": 2089
+    },
+    {
+      "epoch": 0.6996986943421493,
+      "grad_norm": 0.13348999619483948,
+      "learning_rate": 0.0001,
+      "loss": 1.4905,
+      "step": 2090
+    },
+    {
+      "epoch": 0.7000334784064278,
+      "grad_norm": 0.13037413358688354,
+      "learning_rate": 0.0001,
+      "loss": 1.4949,
+      "step": 2091
+    },
+    {
+      "epoch": 0.7003682624707064,
+      "grad_norm": 0.15308037400245667,
+      "learning_rate": 0.0001,
+      "loss": 1.6023,
+      "step": 2092
+    },
+    {
+      "epoch": 0.7007030465349849,
+      "grad_norm": 0.128286212682724,
+      "learning_rate": 0.0001,
+      "loss": 1.5298,
+      "step": 2093
+    },
+    {
+      "epoch": 0.7010378305992635,
+      "grad_norm": 0.13967067003250122,
+      "learning_rate": 0.0001,
+      "loss": 1.5577,
+      "step": 2094
+    },
+    {
+      "epoch": 0.7013726146635421,
+      "grad_norm": 0.13320837914943695,
+      "learning_rate": 0.0001,
+      "loss": 1.5923,
+      "step": 2095
+    },
+    {
+      "epoch": 0.7017073987278205,
+      "grad_norm": 0.12857401371002197,
+      "learning_rate": 0.0001,
+      "loss": 1.4623,
+      "step": 2096
+    },
+    {
+      "epoch": 0.7020421827920991,
+      "grad_norm": 0.12525291740894318,
+      "learning_rate": 0.0001,
+      "loss": 1.5126,
+      "step": 2097
+    },
+    {
+      "epoch": 0.7023769668563776,
+      "grad_norm": 0.1316770762205124,
+      "learning_rate": 0.0001,
+      "loss": 1.5433,
+      "step": 2098
+    },
+    {
+      "epoch": 0.7027117509206562,
+      "grad_norm": 0.1343490481376648,
+      "learning_rate": 0.0001,
+      "loss": 1.5085,
+      "step": 2099
+    },
+    {
+      "epoch": 0.7030465349849347,
+      "grad_norm": 0.12864871323108673,
+      "learning_rate": 0.0001,
+      "loss": 1.46,
+      "step": 2100
+    },
+    {
+      "epoch": 0.7033813190492133,
+      "grad_norm": 0.13915804028511047,
+      "learning_rate": 0.0001,
+      "loss": 1.6961,
+      "step": 2101
+    },
+    {
+      "epoch": 0.7037161031134918,
+      "grad_norm": 0.12709419429302216,
+      "learning_rate": 0.0001,
+      "loss": 1.4931,
+      "step": 2102
+    },
+    {
+      "epoch": 0.7040508871777703,
+      "grad_norm": 0.1383008360862732,
+      "learning_rate": 0.0001,
+      "loss": 1.5925,
+      "step": 2103
+    },
+    {
+      "epoch": 0.7043856712420489,
+      "grad_norm": 0.1338641494512558,
+      "learning_rate": 0.0001,
+      "loss": 1.4715,
+      "step": 2104
+    },
+    {
+      "epoch": 0.7047204553063274,
+      "grad_norm": 0.12291635572910309,
+      "learning_rate": 0.0001,
+      "loss": 1.3746,
+      "step": 2105
+    },
+    {
+      "epoch": 0.705055239370606,
+      "grad_norm": 0.13391555845737457,
+      "learning_rate": 0.0001,
+      "loss": 1.627,
+      "step": 2106
+    },
+    {
+      "epoch": 0.7053900234348845,
+      "grad_norm": 0.13259120285511017,
+      "learning_rate": 0.0001,
+      "loss": 1.6069,
+      "step": 2107
+    },
+    {
+      "epoch": 0.705724807499163,
+      "grad_norm": 0.13009488582611084,
+      "learning_rate": 0.0001,
+      "loss": 1.534,
+      "step": 2108
+    },
+    {
+      "epoch": 0.7060595915634416,
+      "grad_norm": 0.12612484395503998,
+      "learning_rate": 0.0001,
+      "loss": 1.4612,
+      "step": 2109
+    },
+    {
+      "epoch": 0.7063943756277201,
+      "grad_norm": 0.12470883876085281,
+      "learning_rate": 0.0001,
+      "loss": 1.4388,
+      "step": 2110
+    },
+    {
+      "epoch": 0.7067291596919987,
+      "grad_norm": 0.13072682917118073,
+      "learning_rate": 0.0001,
+      "loss": 1.5083,
+      "step": 2111
+    },
+    {
+      "epoch": 0.7070639437562772,
+      "grad_norm": 0.13037820160388947,
+      "learning_rate": 0.0001,
+      "loss": 1.4514,
+      "step": 2112
+    },
+    {
+      "epoch": 0.7073987278205558,
+      "grad_norm": 0.1304703801870346,
+      "learning_rate": 0.0001,
+      "loss": 1.4644,
+      "step": 2113
+    },
+    {
+      "epoch": 0.7077335118848342,
+      "grad_norm": 0.1345730423927307,
+      "learning_rate": 0.0001,
+      "loss": 1.4849,
+      "step": 2114
+    },
+    {
+      "epoch": 0.7080682959491128,
+      "grad_norm": 0.14024527370929718,
+      "learning_rate": 0.0001,
+      "loss": 1.5851,
+      "step": 2115
+    },
+    {
+      "epoch": 0.7084030800133914,
+      "grad_norm": 0.13666972517967224,
+      "learning_rate": 0.0001,
+      "loss": 1.4858,
+      "step": 2116
+    },
+    {
+      "epoch": 0.7087378640776699,
+      "grad_norm": 0.13574914634227753,
+      "learning_rate": 0.0001,
+      "loss": 1.5258,
+      "step": 2117
+    },
+    {
+      "epoch": 0.7090726481419485,
+      "grad_norm": 0.1362755447626114,
+      "learning_rate": 0.0001,
+      "loss": 1.5592,
+      "step": 2118
+    },
+    {
+      "epoch": 0.7094074322062269,
+      "grad_norm": 0.12771886587142944,
+      "learning_rate": 0.0001,
+      "loss": 1.459,
+      "step": 2119
+    },
+    {
+      "epoch": 0.7097422162705055,
+      "grad_norm": 0.13762152194976807,
+      "learning_rate": 0.0001,
+      "loss": 1.5934,
+      "step": 2120
+    },
+    {
+      "epoch": 0.710077000334784,
+      "grad_norm": 0.13554149866104126,
+      "learning_rate": 0.0001,
+      "loss": 1.5728,
+      "step": 2121
+    },
+    {
+      "epoch": 0.7104117843990626,
+      "grad_norm": 0.1313951313495636,
+      "learning_rate": 0.0001,
+      "loss": 1.517,
+      "step": 2122
+    },
+    {
+      "epoch": 0.7107465684633412,
+      "grad_norm": 0.12920212745666504,
+      "learning_rate": 0.0001,
+      "loss": 1.4647,
+      "step": 2123
+    },
+    {
+      "epoch": 0.7110813525276197,
+      "grad_norm": 0.13671697676181793,
+      "learning_rate": 0.0001,
+      "loss": 1.4987,
+      "step": 2124
+    },
+    {
+      "epoch": 0.7114161365918982,
+      "grad_norm": 0.12860006093978882,
+      "learning_rate": 0.0001,
+      "loss": 1.5304,
+      "step": 2125
+    },
+    {
+      "epoch": 0.7117509206561767,
+      "grad_norm": 0.12372934073209763,
+      "learning_rate": 0.0001,
+      "loss": 1.4964,
+      "step": 2126
+    },
+    {
+      "epoch": 0.7120857047204553,
+      "grad_norm": 0.13640989363193512,
+      "learning_rate": 0.0001,
+      "loss": 1.521,
+      "step": 2127
+    },
+    {
+      "epoch": 0.7124204887847339,
+      "grad_norm": 0.13121746480464935,
+      "learning_rate": 0.0001,
+      "loss": 1.5118,
+      "step": 2128
+    },
+    {
+      "epoch": 0.7127552728490124,
+      "grad_norm": 0.1307837963104248,
+      "learning_rate": 0.0001,
+      "loss": 1.5688,
+      "step": 2129
+    },
+    {
+      "epoch": 0.713090056913291,
+      "grad_norm": 0.13141870498657227,
+      "learning_rate": 0.0001,
+      "loss": 1.5435,
+      "step": 2130
+    },
+    {
+      "epoch": 0.7134248409775694,
+      "grad_norm": 0.13490049540996552,
+      "learning_rate": 0.0001,
+      "loss": 1.5421,
+      "step": 2131
+    },
+    {
+      "epoch": 0.713759625041848,
+      "grad_norm": 0.13801416754722595,
+      "learning_rate": 0.0001,
+      "loss": 1.6097,
+      "step": 2132
+    },
+    {
+      "epoch": 0.7140944091061265,
+      "grad_norm": 0.13066011667251587,
+      "learning_rate": 0.0001,
+      "loss": 1.4629,
+      "step": 2133
+    },
+    {
+      "epoch": 0.7144291931704051,
+      "grad_norm": 0.13355465233325958,
+      "learning_rate": 0.0001,
+      "loss": 1.6363,
+      "step": 2134
+    },
+    {
+      "epoch": 0.7147639772346837,
+      "grad_norm": 0.12968328595161438,
+      "learning_rate": 0.0001,
+      "loss": 1.4454,
+      "step": 2135
+    },
+    {
+      "epoch": 0.7150987612989622,
+      "grad_norm": 0.14093713462352753,
+      "learning_rate": 0.0001,
+      "loss": 1.6115,
+      "step": 2136
+    },
+    {
+      "epoch": 0.7154335453632407,
+      "grad_norm": 0.13097916543483734,
+      "learning_rate": 0.0001,
+      "loss": 1.5531,
+      "step": 2137
+    },
+    {
+      "epoch": 0.7157683294275192,
+      "grad_norm": 0.1295294314622879,
+      "learning_rate": 0.0001,
+      "loss": 1.5923,
+      "step": 2138
+    },
+    {
+      "epoch": 0.7161031134917978,
+      "grad_norm": 0.13776849210262299,
+      "learning_rate": 0.0001,
+      "loss": 1.5992,
+      "step": 2139
+    },
+    {
+      "epoch": 0.7164378975560763,
+      "grad_norm": 0.13502860069274902,
+      "learning_rate": 0.0001,
+      "loss": 1.4677,
+      "step": 2140
+    },
+    {
+      "epoch": 0.7167726816203549,
+      "grad_norm": 0.13480490446090698,
+      "learning_rate": 0.0001,
+      "loss": 1.6244,
+      "step": 2141
+    },
+    {
+      "epoch": 0.7171074656846335,
+      "grad_norm": 0.13483154773712158,
+      "learning_rate": 0.0001,
+      "loss": 1.616,
+      "step": 2142
+    },
+    {
+      "epoch": 0.7174422497489119,
+      "grad_norm": 0.14340271055698395,
+      "learning_rate": 0.0001,
+      "loss": 1.6287,
+      "step": 2143
+    },
+    {
+      "epoch": 0.7177770338131905,
+      "grad_norm": 0.13620589673519135,
+      "learning_rate": 0.0001,
+      "loss": 1.5193,
+      "step": 2144
+    },
+    {
+      "epoch": 0.718111817877469,
+      "grad_norm": 0.13150522112846375,
+      "learning_rate": 0.0001,
+      "loss": 1.5038,
+      "step": 2145
+    },
+    {
+      "epoch": 0.7184466019417476,
+      "grad_norm": 0.13259613513946533,
+      "learning_rate": 0.0001,
+      "loss": 1.5666,
+      "step": 2146
+    },
+    {
+      "epoch": 0.7187813860060261,
+      "grad_norm": 0.1307973563671112,
+      "learning_rate": 0.0001,
+      "loss": 1.5762,
+      "step": 2147
+    },
+    {
+      "epoch": 0.7191161700703047,
+      "grad_norm": 0.13372613489627838,
+      "learning_rate": 0.0001,
+      "loss": 1.5352,
+      "step": 2148
+    },
+    {
+      "epoch": 0.7194509541345832,
+      "grad_norm": 0.13534867763519287,
+      "learning_rate": 0.0001,
+      "loss": 1.4652,
+      "step": 2149
+    },
+    {
+      "epoch": 0.7197857381988617,
+      "grad_norm": 0.1332571804523468,
+      "learning_rate": 0.0001,
+      "loss": 1.5532,
+      "step": 2150
+    },
+    {
+      "epoch": 0.7201205222631403,
+      "grad_norm": 0.13172098994255066,
+      "learning_rate": 0.0001,
+      "loss": 1.4728,
+      "step": 2151
+    },
+    {
+      "epoch": 0.7204553063274188,
+      "grad_norm": 0.12765897810459137,
+      "learning_rate": 0.0001,
+      "loss": 1.4597,
+      "step": 2152
+    },
+    {
+      "epoch": 0.7207900903916974,
+      "grad_norm": 0.13026951253414154,
+      "learning_rate": 0.0001,
+      "loss": 1.4877,
+      "step": 2153
+    },
+    {
+      "epoch": 0.721124874455976,
+      "grad_norm": 0.1389724761247635,
+      "learning_rate": 0.0001,
+      "loss": 1.5332,
+      "step": 2154
+    },
+    {
+      "epoch": 0.7214596585202544,
+      "grad_norm": 0.13382194936275482,
+      "learning_rate": 0.0001,
+      "loss": 1.5179,
+      "step": 2155
+    },
+    {
+      "epoch": 0.721794442584533,
+      "grad_norm": 0.12780801951885223,
+      "learning_rate": 0.0001,
+      "loss": 1.4393,
+      "step": 2156
+    },
+    {
+      "epoch": 0.7221292266488115,
+      "grad_norm": 0.1323569118976593,
+      "learning_rate": 0.0001,
+      "loss": 1.5528,
+      "step": 2157
+    },
+    {
+      "epoch": 0.7224640107130901,
+      "grad_norm": 0.1358579397201538,
+      "learning_rate": 0.0001,
+      "loss": 1.4996,
+      "step": 2158
+    },
+    {
+      "epoch": 0.7227987947773686,
+      "grad_norm": 0.13905704021453857,
+      "learning_rate": 0.0001,
+      "loss": 1.5979,
+      "step": 2159
+    },
+    {
+      "epoch": 0.7231335788416471,
+      "grad_norm": 0.1356305480003357,
+      "learning_rate": 0.0001,
+      "loss": 1.5851,
+      "step": 2160
+    },
+    {
+      "epoch": 0.7234683629059256,
+      "grad_norm": 0.13545480370521545,
+      "learning_rate": 0.0001,
+      "loss": 1.5622,
+      "step": 2161
+    },
+    {
+      "epoch": 0.7238031469702042,
+      "grad_norm": 0.13289092481136322,
+      "learning_rate": 0.0001,
+      "loss": 1.5253,
+      "step": 2162
+    },
+    {
+      "epoch": 0.7241379310344828,
+      "grad_norm": 0.130274698138237,
+      "learning_rate": 0.0001,
+      "loss": 1.4498,
+      "step": 2163
+    },
+    {
+      "epoch": 0.7244727150987613,
+      "grad_norm": 0.13009384274482727,
+      "learning_rate": 0.0001,
+      "loss": 1.5593,
+      "step": 2164
+    },
+    {
+      "epoch": 0.7248074991630399,
+      "grad_norm": 0.13778330385684967,
+      "learning_rate": 0.0001,
+      "loss": 1.4054,
+      "step": 2165
+    },
+    {
+      "epoch": 0.7251422832273183,
+      "grad_norm": 0.14639288187026978,
+      "learning_rate": 0.0001,
+      "loss": 1.5563,
+      "step": 2166
+    },
+    {
+      "epoch": 0.7254770672915969,
+      "grad_norm": 0.14019513130187988,
+      "learning_rate": 0.0001,
+      "loss": 1.6143,
+      "step": 2167
+    },
+    {
+      "epoch": 0.7258118513558754,
+      "grad_norm": 0.15255634486675262,
+      "learning_rate": 0.0001,
+      "loss": 1.4999,
+      "step": 2168
+    },
+    {
+      "epoch": 0.726146635420154,
+      "grad_norm": 0.133973628282547,
+      "learning_rate": 0.0001,
+      "loss": 1.5648,
+      "step": 2169
+    },
+    {
+      "epoch": 0.7264814194844326,
+      "grad_norm": 0.14227105677127838,
+      "learning_rate": 0.0001,
+      "loss": 1.5372,
+      "step": 2170
+    },
+    {
+      "epoch": 0.7268162035487111,
+      "grad_norm": 0.13694263994693756,
+      "learning_rate": 0.0001,
+      "loss": 1.5454,
+      "step": 2171
+    },
+    {
+      "epoch": 0.7271509876129896,
+      "grad_norm": 0.1395786851644516,
+      "learning_rate": 0.0001,
+      "loss": 1.6018,
+      "step": 2172
+    },
+    {
+      "epoch": 0.7274857716772681,
+      "grad_norm": 0.13695751130580902,
+      "learning_rate": 0.0001,
+      "loss": 1.5542,
+      "step": 2173
+    },
+    {
+      "epoch": 0.7278205557415467,
+      "grad_norm": 0.14114227890968323,
+      "learning_rate": 0.0001,
+      "loss": 1.4742,
+      "step": 2174
+    },
+    {
+      "epoch": 0.7281553398058253,
+      "grad_norm": 0.14633609354496002,
+      "learning_rate": 0.0001,
+      "loss": 1.5335,
+      "step": 2175
+    },
+    {
+      "epoch": 0.7284901238701038,
+      "grad_norm": 0.12929964065551758,
+      "learning_rate": 0.0001,
+      "loss": 1.4759,
+      "step": 2176
+    },
+    {
+      "epoch": 0.7288249079343824,
+      "grad_norm": 0.14383701980113983,
+      "learning_rate": 0.0001,
+      "loss": 1.5744,
+      "step": 2177
+    },
+    {
+      "epoch": 0.7291596919986608,
+      "grad_norm": 0.14609093964099884,
+      "learning_rate": 0.0001,
+      "loss": 1.4927,
+      "step": 2178
+    },
+    {
+      "epoch": 0.7294944760629394,
+      "grad_norm": 0.13813704252243042,
+      "learning_rate": 0.0001,
+      "loss": 1.535,
+      "step": 2179
+    },
+    {
+      "epoch": 0.7298292601272179,
+      "grad_norm": 0.13343721628189087,
+      "learning_rate": 0.0001,
+      "loss": 1.5239,
+      "step": 2180
+    },
+    {
+      "epoch": 0.7301640441914965,
+      "grad_norm": 0.13793961703777313,
+      "learning_rate": 0.0001,
+      "loss": 1.4959,
+      "step": 2181
+    },
+    {
+      "epoch": 0.7304988282557751,
+      "grad_norm": 0.14635740220546722,
+      "learning_rate": 0.0001,
+      "loss": 1.5759,
+      "step": 2182
+    },
+    {
+      "epoch": 0.7308336123200536,
+      "grad_norm": 0.13331273198127747,
+      "learning_rate": 0.0001,
+      "loss": 1.5169,
+      "step": 2183
+    },
+    {
+      "epoch": 0.7311683963843321,
+      "grad_norm": 0.13492250442504883,
+      "learning_rate": 0.0001,
+      "loss": 1.4711,
+      "step": 2184
+    },
+    {
+      "epoch": 0.7315031804486106,
+      "grad_norm": 0.14489556849002838,
+      "learning_rate": 0.0001,
+      "loss": 1.584,
+      "step": 2185
+    },
+    {
+      "epoch": 0.7318379645128892,
+      "grad_norm": 0.13701508939266205,
+      "learning_rate": 0.0001,
+      "loss": 1.5844,
+      "step": 2186
+    },
+    {
+      "epoch": 0.7321727485771677,
+      "grad_norm": 0.1370009034872055,
+      "learning_rate": 0.0001,
+      "loss": 1.5287,
+      "step": 2187
+    },
+    {
+      "epoch": 0.7325075326414463,
+      "grad_norm": 0.14577260613441467,
+      "learning_rate": 0.0001,
+      "loss": 1.4752,
+      "step": 2188
+    },
+    {
+      "epoch": 0.7328423167057249,
+      "grad_norm": 0.1377391368150711,
+      "learning_rate": 0.0001,
+      "loss": 1.5484,
+      "step": 2189
+    },
+    {
+      "epoch": 0.7331771007700033,
+      "grad_norm": 0.1396346390247345,
+      "learning_rate": 0.0001,
+      "loss": 1.5405,
+      "step": 2190
+    },
+    {
+      "epoch": 0.7335118848342819,
+      "grad_norm": 0.1492149382829666,
+      "learning_rate": 0.0001,
+      "loss": 1.5028,
+      "step": 2191
+    },
+    {
+      "epoch": 0.7338466688985604,
+      "grad_norm": 0.13928255438804626,
+      "learning_rate": 0.0001,
+      "loss": 1.6229,
+      "step": 2192
+    },
+    {
+      "epoch": 0.734181452962839,
+      "grad_norm": 0.13838155567646027,
+      "learning_rate": 0.0001,
+      "loss": 1.5661,
+      "step": 2193
+    },
+    {
+      "epoch": 0.7345162370271175,
+      "grad_norm": 0.1435183733701706,
+      "learning_rate": 0.0001,
+      "loss": 1.6133,
+      "step": 2194
+    },
+    {
+      "epoch": 0.734851021091396,
+      "grad_norm": 0.13500259816646576,
+      "learning_rate": 0.0001,
+      "loss": 1.5728,
+      "step": 2195
+    },
+    {
+      "epoch": 0.7351858051556746,
+      "grad_norm": 0.13238045573234558,
+      "learning_rate": 0.0001,
+      "loss": 1.5435,
+      "step": 2196
+    },
+    {
+      "epoch": 0.7355205892199531,
+      "grad_norm": 0.13493601977825165,
+      "learning_rate": 0.0001,
+      "loss": 1.5117,
+      "step": 2197
+    },
+    {
+      "epoch": 0.7358553732842317,
+      "grad_norm": 0.1433602273464203,
+      "learning_rate": 0.0001,
+      "loss": 1.5921,
+      "step": 2198
+    },
+    {
+      "epoch": 0.7361901573485102,
+      "grad_norm": 0.13165898621082306,
+      "learning_rate": 0.0001,
+      "loss": 1.5648,
+      "step": 2199
+    },
+    {
+      "epoch": 0.7365249414127888,
+      "grad_norm": 0.1355050653219223,
+      "learning_rate": 0.0001,
+      "loss": 1.5998,
+      "step": 2200
+    },
+    {
+      "epoch": 0.7368597254770672,
+      "grad_norm": 0.1296299695968628,
+      "learning_rate": 0.0001,
+      "loss": 1.3903,
+      "step": 2201
+    },
+    {
+      "epoch": 0.7371945095413458,
+      "grad_norm": 0.13563255965709686,
+      "learning_rate": 0.0001,
+      "loss": 1.5462,
+      "step": 2202
+    },
+    {
+      "epoch": 0.7375292936056244,
+      "grad_norm": 0.13449116051197052,
+      "learning_rate": 0.0001,
+      "loss": 1.5344,
+      "step": 2203
+    },
+    {
+      "epoch": 0.7378640776699029,
+      "grad_norm": 0.12928107380867004,
+      "learning_rate": 0.0001,
+      "loss": 1.5212,
+      "step": 2204
+    },
+    {
+      "epoch": 0.7381988617341815,
+      "grad_norm": 0.13199785351753235,
+      "learning_rate": 0.0001,
+      "loss": 1.5408,
+      "step": 2205
+    },
+    {
+      "epoch": 0.73853364579846,
+      "grad_norm": 0.13608896732330322,
+      "learning_rate": 0.0001,
+      "loss": 1.6036,
+      "step": 2206
+    },
+    {
+      "epoch": 0.7388684298627385,
+      "grad_norm": 0.1248575821518898,
+      "learning_rate": 0.0001,
+      "loss": 1.4513,
+      "step": 2207
+    },
+    {
+      "epoch": 0.739203213927017,
+      "grad_norm": 0.1319798231124878,
+      "learning_rate": 0.0001,
+      "loss": 1.5231,
+      "step": 2208
+    },
+    {
+      "epoch": 0.7395379979912956,
+      "grad_norm": 0.1297694742679596,
+      "learning_rate": 0.0001,
+      "loss": 1.492,
+      "step": 2209
+    },
+    {
+      "epoch": 0.7398727820555742,
+      "grad_norm": 0.13263830542564392,
+      "learning_rate": 0.0001,
+      "loss": 1.5746,
+      "step": 2210
+    },
+    {
+      "epoch": 0.7402075661198527,
+      "grad_norm": 0.1352548599243164,
+      "learning_rate": 0.0001,
+      "loss": 1.567,
+      "step": 2211
+    },
+    {
+      "epoch": 0.7405423501841313,
+      "grad_norm": 0.13107185065746307,
+      "learning_rate": 0.0001,
+      "loss": 1.5053,
+      "step": 2212
+    },
+    {
+      "epoch": 0.7408771342484097,
+      "grad_norm": 0.13326485455036163,
+      "learning_rate": 0.0001,
+      "loss": 1.5838,
+      "step": 2213
+    },
+    {
+      "epoch": 0.7412119183126883,
+      "grad_norm": 0.14211507141590118,
+      "learning_rate": 0.0001,
+      "loss": 1.5694,
+      "step": 2214
+    },
+    {
+      "epoch": 0.7415467023769668,
+      "grad_norm": 0.13121196627616882,
+      "learning_rate": 0.0001,
+      "loss": 1.4977,
+      "step": 2215
+    },
+    {
+      "epoch": 0.7418814864412454,
+      "grad_norm": 0.13140466809272766,
+      "learning_rate": 0.0001,
+      "loss": 1.568,
+      "step": 2216
+    },
+    {
+      "epoch": 0.742216270505524,
+      "grad_norm": 0.1365407258272171,
+      "learning_rate": 0.0001,
+      "loss": 1.6667,
+      "step": 2217
+    },
+    {
+      "epoch": 0.7425510545698025,
+      "grad_norm": 0.13460293412208557,
+      "learning_rate": 0.0001,
+      "loss": 1.5813,
+      "step": 2218
+    },
+    {
+      "epoch": 0.742885838634081,
+      "grad_norm": 0.13729612529277802,
+      "learning_rate": 0.0001,
+      "loss": 1.5491,
+      "step": 2219
+    },
+    {
+      "epoch": 0.7432206226983595,
+      "grad_norm": 0.13383755087852478,
+      "learning_rate": 0.0001,
+      "loss": 1.5678,
+      "step": 2220
+    },
+    {
+      "epoch": 0.7435554067626381,
+      "grad_norm": 0.13744328916072845,
+      "learning_rate": 0.0001,
+      "loss": 1.5336,
+      "step": 2221
+    },
+    {
+      "epoch": 0.7438901908269167,
+      "grad_norm": 0.12934266030788422,
+      "learning_rate": 0.0001,
+      "loss": 1.5429,
+      "step": 2222
+    },
+    {
+      "epoch": 0.7442249748911952,
+      "grad_norm": 0.1308993250131607,
+      "learning_rate": 0.0001,
+      "loss": 1.5449,
+      "step": 2223
+    },
+    {
+      "epoch": 0.7445597589554738,
+      "grad_norm": 0.1382169872522354,
+      "learning_rate": 0.0001,
+      "loss": 1.6019,
+      "step": 2224
+    },
+    {
+      "epoch": 0.7448945430197522,
+      "grad_norm": 0.13184891641139984,
+      "learning_rate": 0.0001,
+      "loss": 1.5357,
+      "step": 2225
+    },
+    {
+      "epoch": 0.7452293270840308,
+      "grad_norm": 0.1404266655445099,
+      "learning_rate": 0.0001,
+      "loss": 1.5935,
+      "step": 2226
+    },
+    {
+      "epoch": 0.7455641111483093,
+      "grad_norm": 0.13625003397464752,
+      "learning_rate": 0.0001,
+      "loss": 1.5588,
+      "step": 2227
+    },
+    {
+      "epoch": 0.7458988952125879,
+      "grad_norm": 0.1287645548582077,
+      "learning_rate": 0.0001,
+      "loss": 1.435,
+      "step": 2228
+    },
+    {
+      "epoch": 0.7462336792768665,
+      "grad_norm": 0.13726918399333954,
+      "learning_rate": 0.0001,
+      "loss": 1.5453,
+      "step": 2229
+    },
+    {
+      "epoch": 0.746568463341145,
+      "grad_norm": 0.13299064338207245,
+      "learning_rate": 0.0001,
+      "loss": 1.4996,
+      "step": 2230
+    },
+    {
+      "epoch": 0.7469032474054235,
+      "grad_norm": 0.13553793728351593,
+      "learning_rate": 0.0001,
+      "loss": 1.5395,
+      "step": 2231
+    },
+    {
+      "epoch": 0.747238031469702,
+      "grad_norm": 0.13683359324932098,
+      "learning_rate": 0.0001,
+      "loss": 1.661,
+      "step": 2232
+    },
+    {
+      "epoch": 0.7475728155339806,
+      "grad_norm": 0.13002213835716248,
+      "learning_rate": 0.0001,
+      "loss": 1.5245,
+      "step": 2233
+    },
+    {
+      "epoch": 0.7479075995982591,
+      "grad_norm": 0.13479109108448029,
+      "learning_rate": 0.0001,
+      "loss": 1.5724,
+      "step": 2234
+    },
+    {
+      "epoch": 0.7482423836625377,
+      "grad_norm": 0.13677366077899933,
+      "learning_rate": 0.0001,
+      "loss": 1.6276,
+      "step": 2235
+    },
+    {
+      "epoch": 0.7485771677268162,
+      "grad_norm": 0.14970214664936066,
+      "learning_rate": 0.0001,
+      "loss": 1.6145,
+      "step": 2236
+    },
+    {
+      "epoch": 0.7489119517910947,
+      "grad_norm": 0.1285363882780075,
+      "learning_rate": 0.0001,
+      "loss": 1.4591,
+      "step": 2237
+    },
+    {
+      "epoch": 0.7492467358553733,
+      "grad_norm": 0.14044371247291565,
+      "learning_rate": 0.0001,
+      "loss": 1.511,
+      "step": 2238
+    },
+    {
+      "epoch": 0.7495815199196518,
+      "grad_norm": 0.13310682773590088,
+      "learning_rate": 0.0001,
+      "loss": 1.5777,
+      "step": 2239
+    },
+    {
+      "epoch": 0.7499163039839304,
+      "grad_norm": 0.14290130138397217,
+      "learning_rate": 0.0001,
+      "loss": 1.5075,
+      "step": 2240
+    },
+    {
+      "epoch": 0.750251088048209,
+      "grad_norm": 0.1509731411933899,
+      "learning_rate": 0.0001,
+      "loss": 1.6198,
+      "step": 2241
+    },
+    {
+      "epoch": 0.7505858721124874,
+      "grad_norm": 0.13322798907756805,
+      "learning_rate": 0.0001,
+      "loss": 1.5722,
+      "step": 2242
+    },
+    {
+      "epoch": 0.750920656176766,
+      "grad_norm": 0.1355818659067154,
+      "learning_rate": 0.0001,
+      "loss": 1.4922,
+      "step": 2243
+    },
+    {
+      "epoch": 0.7512554402410445,
+      "grad_norm": 0.14394080638885498,
+      "learning_rate": 0.0001,
+      "loss": 1.5976,
+      "step": 2244
+    },
+    {
+      "epoch": 0.7515902243053231,
+      "grad_norm": 0.135832279920578,
+      "learning_rate": 0.0001,
+      "loss": 1.5138,
+      "step": 2245
+    },
+    {
+      "epoch": 0.7519250083696016,
+      "grad_norm": 0.13906393945217133,
+      "learning_rate": 0.0001,
+      "loss": 1.5351,
+      "step": 2246
+    },
+    {
+      "epoch": 0.7522597924338802,
+      "grad_norm": 0.13090325891971588,
+      "learning_rate": 0.0001,
+      "loss": 1.4505,
+      "step": 2247
+    },
+    {
+      "epoch": 0.7525945764981586,
+      "grad_norm": 0.13537496328353882,
+      "learning_rate": 0.0001,
+      "loss": 1.4955,
+      "step": 2248
+    },
+    {
+      "epoch": 0.7529293605624372,
+      "grad_norm": 0.1373416930437088,
+      "learning_rate": 0.0001,
+      "loss": 1.541,
+      "step": 2249
+    },
+    {
+      "epoch": 0.7532641446267158,
+      "grad_norm": 0.1294248253107071,
+      "learning_rate": 0.0001,
+      "loss": 1.4943,
+      "step": 2250
+    },
+    {
+      "epoch": 0.7535989286909943,
+      "grad_norm": 0.12977437674999237,
+      "learning_rate": 0.0001,
+      "loss": 1.5315,
+      "step": 2251
+    },
+    {
+      "epoch": 0.7539337127552729,
+      "grad_norm": 0.13353915512561798,
+      "learning_rate": 0.0001,
+      "loss": 1.4855,
+      "step": 2252
+    },
+    {
+      "epoch": 0.7542684968195514,
+      "grad_norm": 0.1338808536529541,
+      "learning_rate": 0.0001,
+      "loss": 1.5483,
+      "step": 2253
+    },
+    {
+      "epoch": 0.7546032808838299,
+      "grad_norm": 0.13082879781723022,
+      "learning_rate": 0.0001,
+      "loss": 1.5276,
+      "step": 2254
+    },
+    {
+      "epoch": 0.7549380649481084,
+      "grad_norm": 0.12903323769569397,
+      "learning_rate": 0.0001,
+      "loss": 1.5506,
+      "step": 2255
+    },
+    {
+      "epoch": 0.755272849012387,
+      "grad_norm": 0.1312693953514099,
+      "learning_rate": 0.0001,
+      "loss": 1.4347,
+      "step": 2256
+    },
+    {
+      "epoch": 0.7556076330766656,
+      "grad_norm": 0.13503922522068024,
+      "learning_rate": 0.0001,
+      "loss": 1.5089,
+      "step": 2257
+    },
+    {
+      "epoch": 0.7559424171409441,
+      "grad_norm": 0.13478560745716095,
+      "learning_rate": 0.0001,
+      "loss": 1.4717,
+      "step": 2258
+    },
+    {
+      "epoch": 0.7562772012052227,
+      "grad_norm": 0.14111362397670746,
+      "learning_rate": 0.0001,
+      "loss": 1.4982,
+      "step": 2259
+    },
+    {
+      "epoch": 0.7566119852695011,
+      "grad_norm": 0.13715283572673798,
+      "learning_rate": 0.0001,
+      "loss": 1.5166,
+      "step": 2260
+    },
+    {
+      "epoch": 0.7569467693337797,
+      "grad_norm": 0.14457426965236664,
+      "learning_rate": 0.0001,
+      "loss": 1.6322,
+      "step": 2261
+    },
+    {
+      "epoch": 0.7572815533980582,
+      "grad_norm": 0.13212622702121735,
+      "learning_rate": 0.0001,
+      "loss": 1.4653,
+      "step": 2262
+    },
+    {
+      "epoch": 0.7576163374623368,
+      "grad_norm": 0.136484295129776,
+      "learning_rate": 0.0001,
+      "loss": 1.4416,
+      "step": 2263
+    },
+    {
+      "epoch": 0.7579511215266154,
+      "grad_norm": 0.13701216876506805,
+      "learning_rate": 0.0001,
+      "loss": 1.5158,
+      "step": 2264
+    },
+    {
+      "epoch": 0.7582859055908939,
+      "grad_norm": 0.13045822083950043,
+      "learning_rate": 0.0001,
+      "loss": 1.4805,
+      "step": 2265
+    },
+    {
+      "epoch": 0.7586206896551724,
+      "grad_norm": 0.13484729826450348,
+      "learning_rate": 0.0001,
+      "loss": 1.4919,
+      "step": 2266
+    },
+    {
+      "epoch": 0.7589554737194509,
+      "grad_norm": 0.1352708488702774,
+      "learning_rate": 0.0001,
+      "loss": 1.5632,
+      "step": 2267
+    },
+    {
+      "epoch": 0.7592902577837295,
+      "grad_norm": 0.13968177139759064,
+      "learning_rate": 0.0001,
+      "loss": 1.5983,
+      "step": 2268
+    },
+    {
+      "epoch": 0.759625041848008,
+      "grad_norm": 0.13527031242847443,
+      "learning_rate": 0.0001,
+      "loss": 1.5361,
+      "step": 2269
+    },
+    {
+      "epoch": 0.7599598259122866,
+      "grad_norm": 0.13342413306236267,
+      "learning_rate": 0.0001,
+      "loss": 1.5487,
+      "step": 2270
+    },
+    {
+      "epoch": 0.7602946099765651,
+      "grad_norm": 0.13037632405757904,
+      "learning_rate": 0.0001,
+      "loss": 1.4433,
+      "step": 2271
+    },
+    {
+      "epoch": 0.7606293940408436,
+      "grad_norm": 0.12888109683990479,
+      "learning_rate": 0.0001,
+      "loss": 1.5565,
+      "step": 2272
+    },
+    {
+      "epoch": 0.7609641781051222,
+      "grad_norm": 0.13160650432109833,
+      "learning_rate": 0.0001,
+      "loss": 1.6344,
+      "step": 2273
+    },
+    {
+      "epoch": 0.7612989621694007,
+      "grad_norm": 0.13456179201602936,
+      "learning_rate": 0.0001,
+      "loss": 1.5983,
+      "step": 2274
+    },
+    {
+      "epoch": 0.7616337462336793,
+      "grad_norm": 0.12624886631965637,
+      "learning_rate": 0.0001,
+      "loss": 1.4877,
+      "step": 2275
+    },
+    {
+      "epoch": 0.7619685302979579,
+      "grad_norm": 0.13493984937667847,
+      "learning_rate": 0.0001,
+      "loss": 1.6083,
+      "step": 2276
+    },
+    {
+      "epoch": 0.7623033143622363,
+      "grad_norm": 0.13616621494293213,
+      "learning_rate": 0.0001,
+      "loss": 1.59,
+      "step": 2277
+    },
+    {
+      "epoch": 0.7626380984265149,
+      "grad_norm": 0.1309913843870163,
+      "learning_rate": 0.0001,
+      "loss": 1.5356,
+      "step": 2278
+    },
+    {
+      "epoch": 0.7629728824907934,
+      "grad_norm": 0.1269841343164444,
+      "learning_rate": 0.0001,
+      "loss": 1.442,
+      "step": 2279
+    },
+    {
+      "epoch": 0.763307666555072,
+      "grad_norm": 0.13083530962467194,
+      "learning_rate": 0.0001,
+      "loss": 1.4919,
+      "step": 2280
+    },
+    {
+      "epoch": 0.7636424506193505,
+      "grad_norm": 0.13288795948028564,
+      "learning_rate": 0.0001,
+      "loss": 1.5919,
+      "step": 2281
+    },
+    {
+      "epoch": 0.7639772346836291,
+      "grad_norm": 0.1334894597530365,
+      "learning_rate": 0.0001,
+      "loss": 1.5203,
+      "step": 2282
+    },
+    {
+      "epoch": 0.7643120187479076,
+      "grad_norm": 0.1322222203016281,
+      "learning_rate": 0.0001,
+      "loss": 1.4987,
+      "step": 2283
+    },
+    {
+      "epoch": 0.7646468028121861,
+      "grad_norm": 0.13740068674087524,
+      "learning_rate": 0.0001,
+      "loss": 1.5966,
+      "step": 2284
+    },
+    {
+      "epoch": 0.7649815868764647,
+      "grad_norm": 0.13021446764469147,
+      "learning_rate": 0.0001,
+      "loss": 1.5163,
+      "step": 2285
+    },
+    {
+      "epoch": 0.7653163709407432,
+      "grad_norm": 0.13992641866207123,
+      "learning_rate": 0.0001,
+      "loss": 1.5116,
+      "step": 2286
+    },
+    {
+      "epoch": 0.7656511550050218,
+      "grad_norm": 0.13332848250865936,
+      "learning_rate": 0.0001,
+      "loss": 1.5066,
+      "step": 2287
+    },
+    {
+      "epoch": 0.7659859390693003,
+      "grad_norm": 0.12683235108852386,
+      "learning_rate": 0.0001,
+      "loss": 1.4933,
+      "step": 2288
+    },
+    {
+      "epoch": 0.7663207231335788,
+      "grad_norm": 0.13610418140888214,
+      "learning_rate": 0.0001,
+      "loss": 1.5115,
+      "step": 2289
+    },
+    {
+      "epoch": 0.7666555071978574,
+      "grad_norm": 0.13530276715755463,
+      "learning_rate": 0.0001,
+      "loss": 1.5899,
+      "step": 2290
+    },
+    {
+      "epoch": 0.7669902912621359,
+      "grad_norm": 0.13067664206027985,
+      "learning_rate": 0.0001,
+      "loss": 1.4806,
+      "step": 2291
+    },
+    {
+      "epoch": 0.7673250753264145,
+      "grad_norm": 0.12956401705741882,
+      "learning_rate": 0.0001,
+      "loss": 1.4432,
+      "step": 2292
+    },
+    {
+      "epoch": 0.767659859390693,
+      "grad_norm": 0.1368110626935959,
+      "learning_rate": 0.0001,
+      "loss": 1.5858,
+      "step": 2293
+    },
+    {
+      "epoch": 0.7679946434549716,
+      "grad_norm": 0.13342629373073578,
+      "learning_rate": 0.0001,
+      "loss": 1.4773,
+      "step": 2294
+    },
+    {
+      "epoch": 0.76832942751925,
+      "grad_norm": 0.13525448739528656,
+      "learning_rate": 0.0001,
+      "loss": 1.5574,
+      "step": 2295
+    },
+    {
+      "epoch": 0.7686642115835286,
+      "grad_norm": 0.14219002425670624,
+      "learning_rate": 0.0001,
+      "loss": 1.6207,
+      "step": 2296
+    },
+    {
+      "epoch": 0.7689989956478072,
+      "grad_norm": 0.13410523533821106,
+      "learning_rate": 0.0001,
+      "loss": 1.5414,
+      "step": 2297
+    },
+    {
+      "epoch": 0.7693337797120857,
+      "grad_norm": 0.1366255283355713,
+      "learning_rate": 0.0001,
+      "loss": 1.5588,
+      "step": 2298
+    },
+    {
+      "epoch": 0.7696685637763643,
+      "grad_norm": 0.14335733652114868,
+      "learning_rate": 0.0001,
+      "loss": 1.4797,
+      "step": 2299
+    },
+    {
+      "epoch": 0.7700033478406428,
+      "grad_norm": 0.13368913531303406,
+      "learning_rate": 0.0001,
+      "loss": 1.5068,
+      "step": 2300
+    },
+    {
+      "epoch": 0.7703381319049213,
+      "grad_norm": 0.14045390486717224,
+      "learning_rate": 0.0001,
+      "loss": 1.5532,
+      "step": 2301
+    },
+    {
+      "epoch": 0.7706729159691998,
+      "grad_norm": 0.13820236921310425,
+      "learning_rate": 0.0001,
+      "loss": 1.4334,
+      "step": 2302
+    },
+    {
+      "epoch": 0.7710077000334784,
+      "grad_norm": 0.13486477732658386,
+      "learning_rate": 0.0001,
+      "loss": 1.6277,
+      "step": 2303
+    },
+    {
+      "epoch": 0.771342484097757,
+      "grad_norm": 0.1374381184577942,
+      "learning_rate": 0.0001,
+      "loss": 1.4995,
+      "step": 2304
+    },
+    {
+      "epoch": 0.7716772681620355,
+      "grad_norm": 0.14841946959495544,
+      "learning_rate": 0.0001,
+      "loss": 1.6044,
+      "step": 2305
+    },
+    {
+      "epoch": 0.7720120522263141,
+      "grad_norm": 0.13106206059455872,
+      "learning_rate": 0.0001,
+      "loss": 1.5009,
+      "step": 2306
+    },
+    {
+      "epoch": 0.7723468362905925,
+      "grad_norm": 0.13768276572227478,
+      "learning_rate": 0.0001,
+      "loss": 1.5289,
+      "step": 2307
+    },
+    {
+      "epoch": 0.7726816203548711,
+      "grad_norm": 0.14987289905548096,
+      "learning_rate": 0.0001,
+      "loss": 1.5654,
+      "step": 2308
+    },
+    {
+      "epoch": 0.7730164044191496,
+      "grad_norm": 0.13422365486621857,
+      "learning_rate": 0.0001,
+      "loss": 1.5781,
+      "step": 2309
+    },
+    {
+      "epoch": 0.7733511884834282,
+      "grad_norm": 0.14007548987865448,
+      "learning_rate": 0.0001,
+      "loss": 1.468,
+      "step": 2310
+    },
+    {
+      "epoch": 0.7736859725477068,
+      "grad_norm": 0.140237495303154,
+      "learning_rate": 0.0001,
+      "loss": 1.4408,
+      "step": 2311
+    },
+    {
+      "epoch": 0.7740207566119852,
+      "grad_norm": 0.1331593543291092,
+      "learning_rate": 0.0001,
+      "loss": 1.5213,
+      "step": 2312
+    },
+    {
+      "epoch": 0.7743555406762638,
+      "grad_norm": 0.13670580089092255,
+      "learning_rate": 0.0001,
+      "loss": 1.5034,
+      "step": 2313
+    },
+    {
+      "epoch": 0.7746903247405423,
+      "grad_norm": 0.13198411464691162,
+      "learning_rate": 0.0001,
+      "loss": 1.4633,
+      "step": 2314
+    },
+    {
+      "epoch": 0.7750251088048209,
+      "grad_norm": 0.14384810626506805,
+      "learning_rate": 0.0001,
+      "loss": 1.6254,
+      "step": 2315
+    },
+    {
+      "epoch": 0.7753598928690995,
+      "grad_norm": 0.12808088958263397,
+      "learning_rate": 0.0001,
+      "loss": 1.4751,
+      "step": 2316
+    },
+    {
+      "epoch": 0.775694676933378,
+      "grad_norm": 0.14130346477031708,
+      "learning_rate": 0.0001,
+      "loss": 1.5306,
+      "step": 2317
+    },
+    {
+      "epoch": 0.7760294609976565,
+      "grad_norm": 0.13153797388076782,
+      "learning_rate": 0.0001,
+      "loss": 1.5046,
+      "step": 2318
+    },
+    {
+      "epoch": 0.776364245061935,
+      "grad_norm": 0.13447383046150208,
+      "learning_rate": 0.0001,
+      "loss": 1.5288,
+      "step": 2319
+    },
+    {
+      "epoch": 0.7766990291262136,
+      "grad_norm": 0.13588428497314453,
+      "learning_rate": 0.0001,
+      "loss": 1.5792,
+      "step": 2320
+    },
+    {
+      "epoch": 0.7770338131904921,
+      "grad_norm": 0.1414654701948166,
+      "learning_rate": 0.0001,
+      "loss": 1.6252,
+      "step": 2321
+    },
+    {
+      "epoch": 0.7773685972547707,
+      "grad_norm": 0.14798319339752197,
+      "learning_rate": 0.0001,
+      "loss": 1.5182,
+      "step": 2322
+    },
+    {
+      "epoch": 0.7777033813190493,
+      "grad_norm": 0.13594651222229004,
+      "learning_rate": 0.0001,
+      "loss": 1.59,
+      "step": 2323
+    },
+    {
+      "epoch": 0.7780381653833277,
+      "grad_norm": 0.13689537346363068,
+      "learning_rate": 0.0001,
+      "loss": 1.5312,
+      "step": 2324
+    },
+    {
+      "epoch": 0.7783729494476063,
+      "grad_norm": 0.13842853903770447,
+      "learning_rate": 0.0001,
+      "loss": 1.5453,
+      "step": 2325
+    },
+    {
+      "epoch": 0.7787077335118848,
+      "grad_norm": 0.14006944000720978,
+      "learning_rate": 0.0001,
+      "loss": 1.5789,
+      "step": 2326
+    },
+    {
+      "epoch": 0.7790425175761634,
+      "grad_norm": 0.1328335702419281,
+      "learning_rate": 0.0001,
+      "loss": 1.5183,
+      "step": 2327
+    },
+    {
+      "epoch": 0.7793773016404419,
+      "grad_norm": 0.1366383582353592,
+      "learning_rate": 0.0001,
+      "loss": 1.5861,
+      "step": 2328
+    },
+    {
+      "epoch": 0.7797120857047205,
+      "grad_norm": 0.1384078413248062,
+      "learning_rate": 0.0001,
+      "loss": 1.4768,
+      "step": 2329
+    },
+    {
+      "epoch": 0.780046869768999,
+      "grad_norm": 0.13138563930988312,
+      "learning_rate": 0.0001,
+      "loss": 1.5415,
+      "step": 2330
+    },
+    {
+      "epoch": 0.7803816538332775,
+      "grad_norm": 0.13533802330493927,
+      "learning_rate": 0.0001,
+      "loss": 1.5351,
+      "step": 2331
+    },
+    {
+      "epoch": 0.7807164378975561,
+      "grad_norm": 0.12634359300136566,
+      "learning_rate": 0.0001,
+      "loss": 1.4854,
+      "step": 2332
+    },
+    {
+      "epoch": 0.7810512219618346,
+      "grad_norm": 0.14045196771621704,
+      "learning_rate": 0.0001,
+      "loss": 1.5979,
+      "step": 2333
+    },
+    {
+      "epoch": 0.7813860060261132,
+      "grad_norm": 0.12970393896102905,
+      "learning_rate": 0.0001,
+      "loss": 1.4883,
+      "step": 2334
+    },
+    {
+      "epoch": 0.7817207900903917,
+      "grad_norm": 0.13416926562786102,
+      "learning_rate": 0.0001,
+      "loss": 1.538,
+      "step": 2335
+    },
+    {
+      "epoch": 0.7820555741546702,
+      "grad_norm": 0.12993508577346802,
+      "learning_rate": 0.0001,
+      "loss": 1.3861,
+      "step": 2336
+    },
+    {
+      "epoch": 0.7823903582189488,
+      "grad_norm": 0.1441780924797058,
+      "learning_rate": 0.0001,
+      "loss": 1.5082,
+      "step": 2337
+    },
+    {
+      "epoch": 0.7827251422832273,
+      "grad_norm": 0.1340634673833847,
+      "learning_rate": 0.0001,
+      "loss": 1.5308,
+      "step": 2338
+    },
+    {
+      "epoch": 0.7830599263475059,
+      "grad_norm": 0.1375696063041687,
+      "learning_rate": 0.0001,
+      "loss": 1.4726,
+      "step": 2339
+    },
+    {
+      "epoch": 0.7833947104117844,
+      "grad_norm": 0.13143296539783478,
+      "learning_rate": 0.0001,
+      "loss": 1.5403,
+      "step": 2340
+    },
+    {
+      "epoch": 0.783729494476063,
+      "grad_norm": 0.14144007861614227,
+      "learning_rate": 0.0001,
+      "loss": 1.5596,
+      "step": 2341
+    },
+    {
+      "epoch": 0.7840642785403414,
+      "grad_norm": 0.1288491189479828,
+      "learning_rate": 0.0001,
+      "loss": 1.4793,
+      "step": 2342
+    },
+    {
+      "epoch": 0.78439906260462,
+      "grad_norm": 0.13762634992599487,
+      "learning_rate": 0.0001,
+      "loss": 1.5224,
+      "step": 2343
+    },
+    {
+      "epoch": 0.7847338466688986,
+      "grad_norm": 0.1369268000125885,
+      "learning_rate": 0.0001,
+      "loss": 1.5678,
+      "step": 2344
+    },
+    {
+      "epoch": 0.7850686307331771,
+      "grad_norm": 0.1348867565393448,
+      "learning_rate": 0.0001,
+      "loss": 1.5764,
+      "step": 2345
+    },
+    {
+      "epoch": 0.7854034147974557,
+      "grad_norm": 0.13499613106250763,
+      "learning_rate": 0.0001,
+      "loss": 1.5317,
+      "step": 2346
+    },
+    {
+      "epoch": 0.7857381988617341,
+      "grad_norm": 0.136494979262352,
+      "learning_rate": 0.0001,
+      "loss": 1.6178,
+      "step": 2347
+    },
+    {
+      "epoch": 0.7860729829260127,
+      "grad_norm": 0.13742174208164215,
+      "learning_rate": 0.0001,
+      "loss": 1.5524,
+      "step": 2348
+    },
+    {
+      "epoch": 0.7864077669902912,
+      "grad_norm": 0.1315702348947525,
+      "learning_rate": 0.0001,
+      "loss": 1.5199,
+      "step": 2349
+    },
+    {
+      "epoch": 0.7867425510545698,
+      "grad_norm": 0.1344085931777954,
+      "learning_rate": 0.0001,
+      "loss": 1.5222,
+      "step": 2350
+    },
+    {
+      "epoch": 0.7870773351188484,
+      "grad_norm": 0.1331881582736969,
+      "learning_rate": 0.0001,
+      "loss": 1.4746,
+      "step": 2351
+    },
+    {
+      "epoch": 0.7874121191831269,
+      "grad_norm": 0.13880756497383118,
+      "learning_rate": 0.0001,
+      "loss": 1.5027,
+      "step": 2352
+    },
+    {
+      "epoch": 0.7877469032474054,
+      "grad_norm": 0.1315576285123825,
+      "learning_rate": 0.0001,
+      "loss": 1.5833,
+      "step": 2353
+    },
+    {
+      "epoch": 0.7880816873116839,
+      "grad_norm": 0.1278029829263687,
+      "learning_rate": 0.0001,
+      "loss": 1.4475,
+      "step": 2354
+    },
+    {
+      "epoch": 0.7884164713759625,
+      "grad_norm": 0.14114075899124146,
+      "learning_rate": 0.0001,
+      "loss": 1.4451,
+      "step": 2355
+    },
+    {
+      "epoch": 0.788751255440241,
+      "grad_norm": 0.1352827101945877,
+      "learning_rate": 0.0001,
+      "loss": 1.4816,
+      "step": 2356
+    },
+    {
+      "epoch": 0.7890860395045196,
+      "grad_norm": 0.1316574364900589,
+      "learning_rate": 0.0001,
+      "loss": 1.4572,
+      "step": 2357
+    },
+    {
+      "epoch": 0.7894208235687982,
+      "grad_norm": 0.13792237639427185,
+      "learning_rate": 0.0001,
+      "loss": 1.6108,
+      "step": 2358
+    },
+    {
+      "epoch": 0.7897556076330766,
+      "grad_norm": 0.1365162879228592,
+      "learning_rate": 0.0001,
+      "loss": 1.5303,
+      "step": 2359
+    },
+    {
+      "epoch": 0.7900903916973552,
+      "grad_norm": 0.13918493688106537,
+      "learning_rate": 0.0001,
+      "loss": 1.6387,
+      "step": 2360
+    },
+    {
+      "epoch": 0.7904251757616337,
+      "grad_norm": 0.1277536302804947,
+      "learning_rate": 0.0001,
+      "loss": 1.5365,
+      "step": 2361
+    },
+    {
+      "epoch": 0.7907599598259123,
+      "grad_norm": 0.13407327234745026,
+      "learning_rate": 0.0001,
+      "loss": 1.4571,
+      "step": 2362
+    },
+    {
+      "epoch": 0.7910947438901909,
+      "grad_norm": 0.1346539407968521,
+      "learning_rate": 0.0001,
+      "loss": 1.4506,
+      "step": 2363
+    },
+    {
+      "epoch": 0.7914295279544694,
+      "grad_norm": 0.13160093128681183,
+      "learning_rate": 0.0001,
+      "loss": 1.4457,
+      "step": 2364
+    },
+    {
+      "epoch": 0.7917643120187479,
+      "grad_norm": 0.13025003671646118,
+      "learning_rate": 0.0001,
+      "loss": 1.56,
+      "step": 2365
+    },
+    {
+      "epoch": 0.7920990960830264,
+      "grad_norm": 0.14476409554481506,
+      "learning_rate": 0.0001,
+      "loss": 1.5876,
+      "step": 2366
+    },
+    {
+      "epoch": 0.792433880147305,
+      "grad_norm": 0.13053929805755615,
+      "learning_rate": 0.0001,
+      "loss": 1.4338,
+      "step": 2367
+    },
+    {
+      "epoch": 0.7927686642115835,
+      "grad_norm": 0.13872520625591278,
+      "learning_rate": 0.0001,
+      "loss": 1.6427,
+      "step": 2368
+    },
+    {
+      "epoch": 0.7931034482758621,
+      "grad_norm": 0.14061668515205383,
+      "learning_rate": 0.0001,
+      "loss": 1.4886,
+      "step": 2369
+    },
+    {
+      "epoch": 0.7934382323401407,
+      "grad_norm": 0.130232036113739,
+      "learning_rate": 0.0001,
+      "loss": 1.4023,
+      "step": 2370
+    },
+    {
+      "epoch": 0.7937730164044191,
+      "grad_norm": 0.23358748853206635,
+      "learning_rate": 0.0001,
+      "loss": 1.457,
+      "step": 2371
+    },
+    {
+      "epoch": 0.7941078004686977,
+      "grad_norm": 0.13233914971351624,
+      "learning_rate": 0.0001,
+      "loss": 1.4307,
+      "step": 2372
+    },
+    {
+      "epoch": 0.7944425845329762,
+      "grad_norm": 0.13504283130168915,
+      "learning_rate": 0.0001,
+      "loss": 1.5976,
+      "step": 2373
+    },
+    {
+      "epoch": 0.7947773685972548,
+      "grad_norm": 0.13976161181926727,
+      "learning_rate": 0.0001,
+      "loss": 1.6455,
+      "step": 2374
+    },
+    {
+      "epoch": 0.7951121526615333,
+      "grad_norm": 0.1336098313331604,
+      "learning_rate": 0.0001,
+      "loss": 1.4469,
+      "step": 2375
+    },
+    {
+      "epoch": 0.7954469367258119,
+      "grad_norm": 0.13648861646652222,
+      "learning_rate": 0.0001,
+      "loss": 1.4964,
+      "step": 2376
+    },
+    {
+      "epoch": 0.7957817207900904,
+      "grad_norm": 0.13627798855304718,
+      "learning_rate": 0.0001,
+      "loss": 1.5834,
+      "step": 2377
+    },
+    {
+      "epoch": 0.7961165048543689,
+      "grad_norm": 0.14114542305469513,
+      "learning_rate": 0.0001,
+      "loss": 1.5566,
+      "step": 2378
+    },
+    {
+      "epoch": 0.7964512889186475,
+      "grad_norm": 0.13499446213245392,
+      "learning_rate": 0.0001,
+      "loss": 1.5174,
+      "step": 2379
+    },
+    {
+      "epoch": 0.796786072982926,
+      "grad_norm": 0.14620280265808105,
+      "learning_rate": 0.0001,
+      "loss": 1.6778,
+      "step": 2380
+    },
+    {
+      "epoch": 0.7971208570472046,
+      "grad_norm": 0.13239939510822296,
+      "learning_rate": 0.0001,
+      "loss": 1.5274,
+      "step": 2381
+    },
+    {
+      "epoch": 0.7974556411114831,
+      "grad_norm": 0.13517913222312927,
+      "learning_rate": 0.0001,
+      "loss": 1.5291,
+      "step": 2382
+    },
+    {
+      "epoch": 0.7977904251757616,
+      "grad_norm": 0.1352391242980957,
+      "learning_rate": 0.0001,
+      "loss": 1.5285,
+      "step": 2383
+    },
+    {
+      "epoch": 0.7981252092400402,
+      "grad_norm": 0.14000670611858368,
+      "learning_rate": 0.0001,
+      "loss": 1.6194,
+      "step": 2384
+    },
+    {
+      "epoch": 0.7984599933043187,
+      "grad_norm": 0.1349296271800995,
+      "learning_rate": 0.0001,
+      "loss": 1.5001,
+      "step": 2385
+    },
+    {
+      "epoch": 0.7987947773685973,
+      "grad_norm": 0.1352308988571167,
+      "learning_rate": 0.0001,
+      "loss": 1.6213,
+      "step": 2386
+    },
+    {
+      "epoch": 0.7991295614328758,
+      "grad_norm": 0.1368694305419922,
+      "learning_rate": 0.0001,
+      "loss": 1.5861,
+      "step": 2387
+    },
+    {
+      "epoch": 0.7994643454971543,
+      "grad_norm": 0.1355554759502411,
+      "learning_rate": 0.0001,
+      "loss": 1.5377,
+      "step": 2388
+    },
+    {
+      "epoch": 0.7997991295614328,
+      "grad_norm": 0.13328254222869873,
+      "learning_rate": 0.0001,
+      "loss": 1.5517,
+      "step": 2389
+    },
+    {
+      "epoch": 0.8001339136257114,
+      "grad_norm": 0.13724930584430695,
+      "learning_rate": 0.0001,
+      "loss": 1.5987,
+      "step": 2390
+    },
+    {
+      "epoch": 0.80046869768999,
+      "grad_norm": 0.13542616367340088,
+      "learning_rate": 0.0001,
+      "loss": 1.6654,
+      "step": 2391
+    },
+    {
+      "epoch": 0.8008034817542685,
+      "grad_norm": 0.1366943120956421,
+      "learning_rate": 0.0001,
+      "loss": 1.6196,
+      "step": 2392
+    },
+    {
+      "epoch": 0.8011382658185471,
+      "grad_norm": 0.13868063688278198,
+      "learning_rate": 0.0001,
+      "loss": 1.587,
+      "step": 2393
+    },
+    {
+      "epoch": 0.8014730498828255,
+      "grad_norm": 0.1393207311630249,
+      "learning_rate": 0.0001,
+      "loss": 1.5559,
+      "step": 2394
+    },
+    {
+      "epoch": 0.8018078339471041,
+      "grad_norm": 0.13909262418746948,
+      "learning_rate": 0.0001,
+      "loss": 1.5007,
+      "step": 2395
+    },
+    {
+      "epoch": 0.8021426180113826,
+      "grad_norm": 0.12949267029762268,
+      "learning_rate": 0.0001,
+      "loss": 1.5108,
+      "step": 2396
+    },
+    {
+      "epoch": 0.8024774020756612,
+      "grad_norm": 0.12755730748176575,
+      "learning_rate": 0.0001,
+      "loss": 1.5008,
+      "step": 2397
+    },
+    {
+      "epoch": 0.8028121861399398,
+      "grad_norm": 0.12899887561798096,
+      "learning_rate": 0.0001,
+      "loss": 1.3877,
+      "step": 2398
+    },
+    {
+      "epoch": 0.8031469702042183,
+      "grad_norm": 0.1423116773366928,
+      "learning_rate": 0.0001,
+      "loss": 1.4996,
+      "step": 2399
+    },
+    {
+      "epoch": 0.8034817542684968,
+      "grad_norm": 0.13548225164413452,
+      "learning_rate": 0.0001,
+      "loss": 1.5214,
+      "step": 2400
+    },
+    {
+      "epoch": 0.8038165383327753,
+      "grad_norm": 0.13150808215141296,
+      "learning_rate": 0.0001,
+      "loss": 1.4772,
+      "step": 2401
+    },
+    {
+      "epoch": 0.8041513223970539,
+      "grad_norm": 0.13790038228034973,
+      "learning_rate": 0.0001,
+      "loss": 1.5704,
+      "step": 2402
+    },
+    {
+      "epoch": 0.8044861064613325,
+      "grad_norm": 0.13106264173984528,
+      "learning_rate": 0.0001,
+      "loss": 1.5073,
+      "step": 2403
+    },
+    {
+      "epoch": 0.804820890525611,
+      "grad_norm": 0.13568797707557678,
+      "learning_rate": 0.0001,
+      "loss": 1.6371,
+      "step": 2404
+    },
+    {
+      "epoch": 0.8051556745898896,
+      "grad_norm": 0.13882842659950256,
+      "learning_rate": 0.0001,
+      "loss": 1.5571,
+      "step": 2405
+    },
+    {
+      "epoch": 0.805490458654168,
+      "grad_norm": 0.1312180459499359,
+      "learning_rate": 0.0001,
+      "loss": 1.5625,
+      "step": 2406
+    },
+    {
+      "epoch": 0.8058252427184466,
+      "grad_norm": 0.12823453545570374,
+      "learning_rate": 0.0001,
+      "loss": 1.5046,
+      "step": 2407
+    },
+    {
+      "epoch": 0.8061600267827251,
+      "grad_norm": 0.13207179307937622,
+      "learning_rate": 0.0001,
+      "loss": 1.5031,
+      "step": 2408
+    },
+    {
+      "epoch": 0.8064948108470037,
+      "grad_norm": 0.1277305632829666,
+      "learning_rate": 0.0001,
+      "loss": 1.4867,
+      "step": 2409
+    },
+    {
+      "epoch": 0.8068295949112823,
+      "grad_norm": 0.13227322697639465,
+      "learning_rate": 0.0001,
+      "loss": 1.5019,
+      "step": 2410
+    },
+    {
+      "epoch": 0.8071643789755608,
+      "grad_norm": 0.1336304098367691,
+      "learning_rate": 0.0001,
+      "loss": 1.4424,
+      "step": 2411
+    },
+    {
+      "epoch": 0.8074991630398393,
+      "grad_norm": 0.13859078288078308,
+      "learning_rate": 0.0001,
+      "loss": 1.5301,
+      "step": 2412
+    },
+    {
+      "epoch": 0.8078339471041178,
+      "grad_norm": 0.1342136412858963,
+      "learning_rate": 0.0001,
+      "loss": 1.485,
+      "step": 2413
+    },
+    {
+      "epoch": 0.8081687311683964,
+      "grad_norm": 0.14003999531269073,
+      "learning_rate": 0.0001,
+      "loss": 1.5313,
+      "step": 2414
+    },
+    {
+      "epoch": 0.8085035152326749,
+      "grad_norm": 0.13216662406921387,
+      "learning_rate": 0.0001,
+      "loss": 1.52,
+      "step": 2415
+    },
+    {
+      "epoch": 0.8088382992969535,
+      "grad_norm": 0.1373407393693924,
+      "learning_rate": 0.0001,
+      "loss": 1.5157,
+      "step": 2416
+    },
+    {
+      "epoch": 0.8091730833612321,
+      "grad_norm": 0.13850343227386475,
+      "learning_rate": 0.0001,
+      "loss": 1.4971,
+      "step": 2417
+    },
+    {
+      "epoch": 0.8095078674255105,
+      "grad_norm": 0.1334608793258667,
+      "learning_rate": 0.0001,
+      "loss": 1.5237,
+      "step": 2418
+    },
+    {
+      "epoch": 0.8098426514897891,
+      "grad_norm": 0.13133668899536133,
+      "learning_rate": 0.0001,
+      "loss": 1.5053,
+      "step": 2419
+    },
+    {
+      "epoch": 0.8101774355540676,
+      "grad_norm": 0.13715368509292603,
+      "learning_rate": 0.0001,
+      "loss": 1.6357,
+      "step": 2420
+    },
+    {
+      "epoch": 0.8105122196183462,
+      "grad_norm": 0.14129430055618286,
+      "learning_rate": 0.0001,
+      "loss": 1.5736,
+      "step": 2421
+    },
+    {
+      "epoch": 0.8108470036826247,
+      "grad_norm": 0.133287250995636,
+      "learning_rate": 0.0001,
+      "loss": 1.4701,
+      "step": 2422
+    },
+    {
+      "epoch": 0.8111817877469032,
+      "grad_norm": 0.137081116437912,
+      "learning_rate": 0.0001,
+      "loss": 1.4562,
+      "step": 2423
+    },
+    {
+      "epoch": 0.8115165718111818,
+      "grad_norm": 0.13136571645736694,
+      "learning_rate": 0.0001,
+      "loss": 1.5014,
+      "step": 2424
+    },
+    {
+      "epoch": 0.8118513558754603,
+      "grad_norm": 0.13660964369773865,
+      "learning_rate": 0.0001,
+      "loss": 1.5533,
+      "step": 2425
+    },
+    {
+      "epoch": 0.8121861399397389,
+      "grad_norm": 0.145840123295784,
+      "learning_rate": 0.0001,
+      "loss": 1.6406,
+      "step": 2426
+    },
+    {
+      "epoch": 0.8125209240040174,
+      "grad_norm": 0.13612517714500427,
+      "learning_rate": 0.0001,
+      "loss": 1.4968,
+      "step": 2427
+    },
+    {
+      "epoch": 0.812855708068296,
+      "grad_norm": 0.14182846248149872,
+      "learning_rate": 0.0001,
+      "loss": 1.5507,
+      "step": 2428
+    },
+    {
+      "epoch": 0.8131904921325744,
+      "grad_norm": 0.13697752356529236,
+      "learning_rate": 0.0001,
+      "loss": 1.5241,
+      "step": 2429
+    },
+    {
+      "epoch": 0.813525276196853,
+      "grad_norm": 0.14000248908996582,
+      "learning_rate": 0.0001,
+      "loss": 1.6002,
+      "step": 2430
+    },
+    {
+      "epoch": 0.8138600602611316,
+      "grad_norm": 0.13774293661117554,
+      "learning_rate": 0.0001,
+      "loss": 1.5198,
+      "step": 2431
+    },
+    {
+      "epoch": 0.8141948443254101,
+      "grad_norm": 0.13524143397808075,
+      "learning_rate": 0.0001,
+      "loss": 1.5326,
+      "step": 2432
+    },
+    {
+      "epoch": 0.8145296283896887,
+      "grad_norm": 0.13584178686141968,
+      "learning_rate": 0.0001,
+      "loss": 1.5313,
+      "step": 2433
+    },
+    {
+      "epoch": 0.8148644124539672,
+      "grad_norm": 0.13589173555374146,
+      "learning_rate": 0.0001,
+      "loss": 1.5097,
+      "step": 2434
+    },
+    {
+      "epoch": 0.8151991965182457,
+      "grad_norm": 0.1420723795890808,
+      "learning_rate": 0.0001,
+      "loss": 1.593,
+      "step": 2435
+    },
+    {
+      "epoch": 0.8155339805825242,
+      "grad_norm": 0.13078542053699493,
+      "learning_rate": 0.0001,
+      "loss": 1.4239,
+      "step": 2436
+    },
+    {
+      "epoch": 0.8158687646468028,
+      "grad_norm": 0.14007273316383362,
+      "learning_rate": 0.0001,
+      "loss": 1.5912,
+      "step": 2437
+    },
+    {
+      "epoch": 0.8162035487110814,
+      "grad_norm": 0.13472947478294373,
+      "learning_rate": 0.0001,
+      "loss": 1.5146,
+      "step": 2438
+    },
+    {
+      "epoch": 0.8165383327753599,
+      "grad_norm": 0.13456539809703827,
+      "learning_rate": 0.0001,
+      "loss": 1.5277,
+      "step": 2439
+    },
+    {
+      "epoch": 0.8168731168396385,
+      "grad_norm": 0.13376279175281525,
+      "learning_rate": 0.0001,
+      "loss": 1.4554,
+      "step": 2440
+    },
+    {
+      "epoch": 0.8172079009039169,
+      "grad_norm": 0.13720721006393433,
+      "learning_rate": 0.0001,
+      "loss": 1.5463,
+      "step": 2441
+    },
+    {
+      "epoch": 0.8175426849681955,
+      "grad_norm": 0.1363624483346939,
+      "learning_rate": 0.0001,
+      "loss": 1.537,
+      "step": 2442
+    },
+    {
+      "epoch": 0.817877469032474,
+      "grad_norm": 0.13379956781864166,
+      "learning_rate": 0.0001,
+      "loss": 1.5831,
+      "step": 2443
+    },
+    {
+      "epoch": 0.8182122530967526,
+      "grad_norm": 0.13432839512825012,
+      "learning_rate": 0.0001,
+      "loss": 1.5511,
+      "step": 2444
+    },
+    {
+      "epoch": 0.8185470371610312,
+      "grad_norm": 0.1365717500448227,
+      "learning_rate": 0.0001,
+      "loss": 1.4519,
+      "step": 2445
+    },
+    {
+      "epoch": 0.8188818212253097,
+      "grad_norm": 0.13430190086364746,
+      "learning_rate": 0.0001,
+      "loss": 1.4878,
+      "step": 2446
+    },
+    {
+      "epoch": 0.8192166052895882,
+      "grad_norm": 0.13606110215187073,
+      "learning_rate": 0.0001,
+      "loss": 1.5585,
+      "step": 2447
+    },
+    {
+      "epoch": 0.8195513893538667,
+      "grad_norm": 0.13404667377471924,
+      "learning_rate": 0.0001,
+      "loss": 1.5156,
+      "step": 2448
+    },
+    {
+      "epoch": 0.8198861734181453,
+      "grad_norm": 0.14223212003707886,
+      "learning_rate": 0.0001,
+      "loss": 1.5904,
+      "step": 2449
+    },
+    {
+      "epoch": 0.8202209574824239,
+      "grad_norm": 0.13209384679794312,
+      "learning_rate": 0.0001,
+      "loss": 1.551,
+      "step": 2450
+    },
+    {
+      "epoch": 0.8205557415467024,
+      "grad_norm": 0.13522854447364807,
+      "learning_rate": 0.0001,
+      "loss": 1.5325,
+      "step": 2451
+    },
+    {
+      "epoch": 0.820890525610981,
+      "grad_norm": 0.13555531203746796,
+      "learning_rate": 0.0001,
+      "loss": 1.5327,
+      "step": 2452
+    },
+    {
+      "epoch": 0.8212253096752594,
+      "grad_norm": 0.13121196627616882,
+      "learning_rate": 0.0001,
+      "loss": 1.5208,
+      "step": 2453
+    },
+    {
+      "epoch": 0.821560093739538,
+      "grad_norm": 0.13988123834133148,
+      "learning_rate": 0.0001,
+      "loss": 1.6188,
+      "step": 2454
+    },
+    {
+      "epoch": 0.8218948778038165,
+      "grad_norm": 0.1347675770521164,
+      "learning_rate": 0.0001,
+      "loss": 1.5212,
+      "step": 2455
+    },
+    {
+      "epoch": 0.8222296618680951,
+      "grad_norm": 0.13975632190704346,
+      "learning_rate": 0.0001,
+      "loss": 1.6152,
+      "step": 2456
+    },
+    {
+      "epoch": 0.8225644459323737,
+      "grad_norm": 0.1271917223930359,
+      "learning_rate": 0.0001,
+      "loss": 1.4209,
+      "step": 2457
+    },
+    {
+      "epoch": 0.8228992299966521,
+      "grad_norm": 0.13226144015789032,
+      "learning_rate": 0.0001,
+      "loss": 1.5397,
+      "step": 2458
+    },
+    {
+      "epoch": 0.8232340140609307,
+      "grad_norm": 0.1391698569059372,
+      "learning_rate": 0.0001,
+      "loss": 1.5394,
+      "step": 2459
+    },
+    {
+      "epoch": 0.8235687981252092,
+      "grad_norm": 0.13757720589637756,
+      "learning_rate": 0.0001,
+      "loss": 1.5465,
+      "step": 2460
+    },
+    {
+      "epoch": 0.8239035821894878,
+      "grad_norm": 0.13116374611854553,
+      "learning_rate": 0.0001,
+      "loss": 1.5072,
+      "step": 2461
+    },
+    {
+      "epoch": 0.8242383662537663,
+      "grad_norm": 0.13408921658992767,
+      "learning_rate": 0.0001,
+      "loss": 1.5398,
+      "step": 2462
+    },
+    {
+      "epoch": 0.8245731503180449,
+      "grad_norm": 0.13682673871517181,
+      "learning_rate": 0.0001,
+      "loss": 1.574,
+      "step": 2463
+    },
+    {
+      "epoch": 0.8249079343823233,
+      "grad_norm": 0.12918630242347717,
+      "learning_rate": 0.0001,
+      "loss": 1.4619,
+      "step": 2464
+    },
+    {
+      "epoch": 0.8252427184466019,
+      "grad_norm": 0.14337001740932465,
+      "learning_rate": 0.0001,
+      "loss": 1.5494,
+      "step": 2465
+    },
+    {
+      "epoch": 0.8255775025108805,
+      "grad_norm": 0.13083745539188385,
+      "learning_rate": 0.0001,
+      "loss": 1.4594,
+      "step": 2466
+    },
+    {
+      "epoch": 0.825912286575159,
+      "grad_norm": 0.13452093303203583,
+      "learning_rate": 0.0001,
+      "loss": 1.5114,
+      "step": 2467
+    },
+    {
+      "epoch": 0.8262470706394376,
+      "grad_norm": 0.1375538408756256,
+      "learning_rate": 0.0001,
+      "loss": 1.5472,
+      "step": 2468
+    },
+    {
+      "epoch": 0.8265818547037161,
+      "grad_norm": 0.13618512451648712,
+      "learning_rate": 0.0001,
+      "loss": 1.5067,
+      "step": 2469
+    },
+    {
+      "epoch": 0.8269166387679946,
+      "grad_norm": 0.13334475457668304,
+      "learning_rate": 0.0001,
+      "loss": 1.5626,
+      "step": 2470
+    },
+    {
+      "epoch": 0.8272514228322732,
+      "grad_norm": 0.12935003638267517,
+      "learning_rate": 0.0001,
+      "loss": 1.4524,
+      "step": 2471
+    },
+    {
+      "epoch": 0.8275862068965517,
+      "grad_norm": 0.1333768367767334,
+      "learning_rate": 0.0001,
+      "loss": 1.4809,
+      "step": 2472
+    },
+    {
+      "epoch": 0.8279209909608303,
+      "grad_norm": 0.139461949467659,
+      "learning_rate": 0.0001,
+      "loss": 1.5265,
+      "step": 2473
+    },
+    {
+      "epoch": 0.8282557750251088,
+      "grad_norm": 0.14345921576023102,
+      "learning_rate": 0.0001,
+      "loss": 1.5911,
+      "step": 2474
+    },
+    {
+      "epoch": 0.8285905590893874,
+      "grad_norm": 0.12835142016410828,
+      "learning_rate": 0.0001,
+      "loss": 1.4934,
+      "step": 2475
+    },
+    {
+      "epoch": 0.8289253431536658,
+      "grad_norm": 0.13207587599754333,
+      "learning_rate": 0.0001,
+      "loss": 1.5013,
+      "step": 2476
+    },
+    {
+      "epoch": 0.8292601272179444,
+      "grad_norm": 0.14216424524784088,
+      "learning_rate": 0.0001,
+      "loss": 1.5783,
+      "step": 2477
+    },
+    {
+      "epoch": 0.829594911282223,
+      "grad_norm": 0.1372382938861847,
+      "learning_rate": 0.0001,
+      "loss": 1.5487,
+      "step": 2478
+    },
+    {
+      "epoch": 0.8299296953465015,
+      "grad_norm": 0.14100505411624908,
+      "learning_rate": 0.0001,
+      "loss": 1.5893,
+      "step": 2479
+    },
+    {
+      "epoch": 0.8302644794107801,
+      "grad_norm": 0.13831539452075958,
+      "learning_rate": 0.0001,
+      "loss": 1.5308,
+      "step": 2480
+    },
+    {
+      "epoch": 0.8305992634750586,
+      "grad_norm": 0.13254091143608093,
+      "learning_rate": 0.0001,
+      "loss": 1.509,
+      "step": 2481
+    },
+    {
+      "epoch": 0.8309340475393371,
+      "grad_norm": 0.13434451818466187,
+      "learning_rate": 0.0001,
+      "loss": 1.4544,
+      "step": 2482
+    },
+    {
+      "epoch": 0.8312688316036156,
+      "grad_norm": 0.13452693819999695,
+      "learning_rate": 0.0001,
+      "loss": 1.4875,
+      "step": 2483
+    },
+    {
+      "epoch": 0.8316036156678942,
+      "grad_norm": 0.13497060537338257,
+      "learning_rate": 0.0001,
+      "loss": 1.4973,
+      "step": 2484
+    },
+    {
+      "epoch": 0.8319383997321728,
+      "grad_norm": 0.13919363915920258,
+      "learning_rate": 0.0001,
+      "loss": 1.4425,
+      "step": 2485
+    },
+    {
+      "epoch": 0.8322731837964513,
+      "grad_norm": 0.14376235008239746,
+      "learning_rate": 0.0001,
+      "loss": 1.5438,
+      "step": 2486
+    },
+    {
+      "epoch": 0.8326079678607299,
+      "grad_norm": 0.13027647137641907,
+      "learning_rate": 0.0001,
+      "loss": 1.4899,
+      "step": 2487
+    },
+    {
+      "epoch": 0.8329427519250083,
+      "grad_norm": 0.1342213749885559,
+      "learning_rate": 0.0001,
+      "loss": 1.4716,
+      "step": 2488
+    },
+    {
+      "epoch": 0.8332775359892869,
+      "grad_norm": 0.1298682540655136,
+      "learning_rate": 0.0001,
+      "loss": 1.4359,
+      "step": 2489
+    },
+    {
+      "epoch": 0.8336123200535654,
+      "grad_norm": 0.13764667510986328,
+      "learning_rate": 0.0001,
+      "loss": 1.6205,
+      "step": 2490
+    },
+    {
+      "epoch": 0.833947104117844,
+      "grad_norm": 0.13023105263710022,
+      "learning_rate": 0.0001,
+      "loss": 1.4276,
+      "step": 2491
+    },
+    {
+      "epoch": 0.8342818881821226,
+      "grad_norm": 0.1355689913034439,
+      "learning_rate": 0.0001,
+      "loss": 1.4635,
+      "step": 2492
+    },
+    {
+      "epoch": 0.8346166722464011,
+      "grad_norm": 0.13397172093391418,
+      "learning_rate": 0.0001,
+      "loss": 1.5855,
+      "step": 2493
+    },
+    {
+      "epoch": 0.8349514563106796,
+      "grad_norm": 0.13192683458328247,
+      "learning_rate": 0.0001,
+      "loss": 1.5209,
+      "step": 2494
+    },
+    {
+      "epoch": 0.8352862403749581,
+      "grad_norm": 0.13405252993106842,
+      "learning_rate": 0.0001,
+      "loss": 1.5144,
+      "step": 2495
+    },
+    {
+      "epoch": 0.8356210244392367,
+      "grad_norm": 0.13375818729400635,
+      "learning_rate": 0.0001,
+      "loss": 1.4467,
+      "step": 2496
+    },
+    {
+      "epoch": 0.8359558085035153,
+      "grad_norm": 0.12543916702270508,
+      "learning_rate": 0.0001,
+      "loss": 1.3992,
+      "step": 2497
+    },
+    {
+      "epoch": 0.8362905925677938,
+      "grad_norm": 0.13587196171283722,
+      "learning_rate": 0.0001,
+      "loss": 1.487,
+      "step": 2498
+    },
+    {
+      "epoch": 0.8366253766320723,
+      "grad_norm": 0.13462427258491516,
+      "learning_rate": 0.0001,
+      "loss": 1.5455,
+      "step": 2499
+    },
+    {
+      "epoch": 0.8369601606963508,
+      "grad_norm": 0.13338516652584076,
+      "learning_rate": 0.0001,
+      "loss": 1.5612,
+      "step": 2500
+    },
+    {
+      "epoch": 0.0003348961821835231,
+      "grad_norm": 0.1373075246810913,
+      "learning_rate": 0.0001,
+      "loss": 1.5044,
+      "step": 2501
+    },
+    {
+      "epoch": 0.0006697923643670462,
+      "grad_norm": 0.13018828630447388,
+      "learning_rate": 0.0001,
+      "loss": 1.4832,
+      "step": 2502
+    },
+    {
+      "epoch": 0.0010046885465505692,
+      "grad_norm": 0.1399035006761551,
+      "learning_rate": 0.0001,
+      "loss": 1.5154,
+      "step": 2503
+    },
+    {
+      "epoch": 0.0013395847287340924,
+      "grad_norm": 0.13555970788002014,
+      "learning_rate": 0.0001,
+      "loss": 1.5546,
+      "step": 2504
+    },
+    {
+      "epoch": 0.0016744809109176155,
+      "grad_norm": 0.13555824756622314,
+      "learning_rate": 0.0001,
+      "loss": 1.4698,
+      "step": 2505
+    },
+    {
+      "epoch": 0.0020093770931011385,
+      "grad_norm": 0.13882486522197723,
+      "learning_rate": 0.0001,
+      "loss": 1.527,
+      "step": 2506
+    },
+    {
+      "epoch": 0.002344273275284662,
+      "grad_norm": 0.1495855301618576,
+      "learning_rate": 0.0001,
+      "loss": 1.5397,
+      "step": 2507
+    },
+    {
+      "epoch": 0.0026791694574681848,
+      "grad_norm": 0.13565705716609955,
+      "learning_rate": 0.0001,
+      "loss": 1.4942,
+      "step": 2508
+    },
+    {
+      "epoch": 0.003014065639651708,
+      "grad_norm": 0.13897860050201416,
+      "learning_rate": 0.0001,
+      "loss": 1.5108,
+      "step": 2509
+    },
+    {
+      "epoch": 0.003348961821835231,
+      "grad_norm": 0.13958251476287842,
+      "learning_rate": 0.0001,
+      "loss": 1.4347,
+      "step": 2510
+    },
+    {
+      "epoch": 0.003683858004018754,
+      "grad_norm": 0.14172150194644928,
+      "learning_rate": 0.0001,
+      "loss": 1.6219,
+      "step": 2511
+    },
+    {
+      "epoch": 0.004018754186202277,
+      "grad_norm": 0.14103008806705475,
+      "learning_rate": 0.0001,
+      "loss": 1.4782,
+      "step": 2512
+    },
+    {
+      "epoch": 0.004353650368385801,
+      "grad_norm": 0.1375630795955658,
+      "learning_rate": 0.0001,
+      "loss": 1.4742,
+      "step": 2513
+    },
+    {
+      "epoch": 0.004688546550569324,
+      "grad_norm": 0.13548851013183594,
+      "learning_rate": 0.0001,
+      "loss": 1.4329,
+      "step": 2514
+    },
+    {
+      "epoch": 0.005023442732752847,
+      "grad_norm": 0.1332726776599884,
+      "learning_rate": 0.0001,
+      "loss": 1.4935,
+      "step": 2515
+    },
+    {
+      "epoch": 0.0053583389149363695,
+      "grad_norm": 0.1388276070356369,
+      "learning_rate": 0.0001,
+      "loss": 1.5153,
+      "step": 2516
+    },
+    {
+      "epoch": 0.0056932350971198925,
+      "grad_norm": 0.14141787588596344,
+      "learning_rate": 0.0001,
+      "loss": 1.5171,
+      "step": 2517
+    },
+    {
+      "epoch": 0.006028131279303416,
+      "grad_norm": 0.13511618971824646,
+      "learning_rate": 0.0001,
+      "loss": 1.4681,
+      "step": 2518
+    },
+    {
+      "epoch": 0.006363027461486939,
+      "grad_norm": 0.13339103758335114,
+      "learning_rate": 0.0001,
+      "loss": 1.5153,
+      "step": 2519
+    },
+    {
+      "epoch": 0.006697923643670462,
+      "grad_norm": 0.13787244260311127,
+      "learning_rate": 0.0001,
+      "loss": 1.4729,
+      "step": 2520
+    },
+    {
+      "epoch": 0.007032819825853985,
+      "grad_norm": 0.13926245272159576,
+      "learning_rate": 0.0001,
+      "loss": 1.5248,
+      "step": 2521
+    },
+    {
+      "epoch": 0.007367716008037508,
+      "grad_norm": 0.1414000689983368,
+      "learning_rate": 0.0001,
+      "loss": 1.5346,
+      "step": 2522
+    },
+    {
+      "epoch": 0.007702612190221032,
+      "grad_norm": 0.13984519243240356,
+      "learning_rate": 0.0001,
+      "loss": 1.47,
+      "step": 2523
+    },
+    {
+      "epoch": 0.008037508372404554,
+      "grad_norm": 0.13971388339996338,
+      "learning_rate": 0.0001,
+      "loss": 1.519,
+      "step": 2524
+    },
+    {
+      "epoch": 0.008372404554588079,
+      "grad_norm": 0.13793839514255524,
+      "learning_rate": 0.0001,
+      "loss": 1.4681,
+      "step": 2525
+    },
+    {
+      "epoch": 0.008707300736771601,
+      "grad_norm": 0.136174738407135,
+      "learning_rate": 0.0001,
+      "loss": 1.3919,
+      "step": 2526
+    },
+    {
+      "epoch": 0.009042196918955124,
+      "grad_norm": 0.13941626250743866,
+      "learning_rate": 0.0001,
+      "loss": 1.5496,
+      "step": 2527
+    },
+    {
+      "epoch": 0.009377093101138647,
+      "grad_norm": 0.13217207789421082,
+      "learning_rate": 0.0001,
+      "loss": 1.4442,
+      "step": 2528
+    },
+    {
+      "epoch": 0.00971198928332217,
+      "grad_norm": 0.14001114666461945,
+      "learning_rate": 0.0001,
+      "loss": 1.5094,
+      "step": 2529
+    },
+    {
+      "epoch": 0.010046885465505693,
+      "grad_norm": 0.14493252336978912,
+      "learning_rate": 0.0001,
+      "loss": 1.5396,
+      "step": 2530
+    },
+    {
+      "epoch": 0.010381781647689216,
+      "grad_norm": 0.1410461664199829,
+      "learning_rate": 0.0001,
+      "loss": 1.4883,
+      "step": 2531
+    },
+    {
+      "epoch": 0.010716677829872739,
+      "grad_norm": 0.13383805751800537,
+      "learning_rate": 0.0001,
+      "loss": 1.4484,
+      "step": 2532
+    },
+    {
+      "epoch": 0.011051574012056262,
+      "grad_norm": 0.1375914365053177,
+      "learning_rate": 0.0001,
+      "loss": 1.4361,
+      "step": 2533
+    },
+    {
+      "epoch": 0.011386470194239785,
+      "grad_norm": 0.13821835815906525,
+      "learning_rate": 0.0001,
+      "loss": 1.4563,
+      "step": 2534
+    },
+    {
+      "epoch": 0.01172136637642331,
+      "grad_norm": 0.14385046064853668,
+      "learning_rate": 0.0001,
+      "loss": 1.5055,
+      "step": 2535
+    },
+    {
+      "epoch": 0.012056262558606833,
+      "grad_norm": 0.13575609028339386,
+      "learning_rate": 0.0001,
+      "loss": 1.4273,
+      "step": 2536
+    },
+    {
+      "epoch": 0.012391158740790355,
+      "grad_norm": 0.14160922169685364,
+      "learning_rate": 0.0001,
+      "loss": 1.4898,
+      "step": 2537
+    },
+    {
+      "epoch": 0.012726054922973878,
+      "grad_norm": 0.1364385187625885,
+      "learning_rate": 0.0001,
+      "loss": 1.3803,
+      "step": 2538
+    },
+    {
+      "epoch": 0.013060951105157401,
+      "grad_norm": 0.14626376330852509,
+      "learning_rate": 0.0001,
+      "loss": 1.5782,
+      "step": 2539
+    },
+    {
+      "epoch": 0.013395847287340924,
+      "grad_norm": 0.140034481883049,
+      "learning_rate": 0.0001,
+      "loss": 1.4472,
+      "step": 2540
+    },
+    {
+      "epoch": 0.013730743469524447,
+      "grad_norm": 0.13497884571552277,
+      "learning_rate": 0.0001,
+      "loss": 1.4742,
+      "step": 2541
+    },
+    {
+      "epoch": 0.01406563965170797,
+      "grad_norm": 0.13669952750205994,
+      "learning_rate": 0.0001,
+      "loss": 1.4631,
+      "step": 2542
+    },
+    {
+      "epoch": 0.014400535833891493,
+      "grad_norm": 0.14313064515590668,
+      "learning_rate": 0.0001,
+      "loss": 1.5265,
+      "step": 2543
+    },
+    {
+      "epoch": 0.014735432016075016,
+      "grad_norm": 0.14192776381969452,
+      "learning_rate": 0.0001,
+      "loss": 1.5202,
+      "step": 2544
+    },
+    {
+      "epoch": 0.01507032819825854,
+      "grad_norm": 0.13427315652370453,
+      "learning_rate": 0.0001,
+      "loss": 1.4417,
+      "step": 2545
+    },
+    {
+      "epoch": 0.015405224380442064,
+      "grad_norm": 0.13986799120903015,
+      "learning_rate": 0.0001,
+      "loss": 1.5255,
+      "step": 2546
+    },
+    {
+      "epoch": 0.015740120562625585,
+      "grad_norm": 0.13709712028503418,
+      "learning_rate": 0.0001,
+      "loss": 1.4431,
+      "step": 2547
+    },
+    {
+      "epoch": 0.016075016744809108,
+      "grad_norm": 0.13961805403232574,
+      "learning_rate": 0.0001,
+      "loss": 1.486,
+      "step": 2548
+    },
+    {
+      "epoch": 0.01640991292699263,
+      "grad_norm": 0.1424587219953537,
+      "learning_rate": 0.0001,
+      "loss": 1.5153,
+      "step": 2549
+    },
+    {
+      "epoch": 0.016744809109176157,
+      "grad_norm": 0.141858771443367,
+      "learning_rate": 0.0001,
+      "loss": 1.4655,
+      "step": 2550
+    },
+    {
+      "epoch": 0.01707970529135968,
+      "grad_norm": 0.1496685892343521,
+      "learning_rate": 0.0001,
+      "loss": 1.4828,
+      "step": 2551
+    },
+    {
+      "epoch": 0.017414601473543203,
+      "grad_norm": 0.14911946654319763,
+      "learning_rate": 0.0001,
+      "loss": 1.5248,
+      "step": 2552
+    },
+    {
+      "epoch": 0.017749497655726726,
+      "grad_norm": 0.1409626305103302,
+      "learning_rate": 0.0001,
+      "loss": 1.4795,
+      "step": 2553
+    },
+    {
+      "epoch": 0.01808439383791025,
+      "grad_norm": 0.1455659717321396,
+      "learning_rate": 0.0001,
+      "loss": 1.5117,
+      "step": 2554
+    },
+    {
+      "epoch": 0.018419290020093772,
+      "grad_norm": 0.1373487412929535,
+      "learning_rate": 0.0001,
+      "loss": 1.4457,
+      "step": 2555
+    },
+    {
+      "epoch": 0.018754186202277295,
+      "grad_norm": 0.1419982612133026,
+      "learning_rate": 0.0001,
+      "loss": 1.4763,
+      "step": 2556
+    },
+    {
+      "epoch": 0.019089082384460818,
+      "grad_norm": 0.13673114776611328,
+      "learning_rate": 0.0001,
+      "loss": 1.4535,
+      "step": 2557
+    },
+    {
+      "epoch": 0.01942397856664434,
+      "grad_norm": 0.1524597704410553,
+      "learning_rate": 0.0001,
+      "loss": 1.5723,
+      "step": 2558
+    },
+    {
+      "epoch": 0.019758874748827863,
+      "grad_norm": 0.14004305005073547,
+      "learning_rate": 0.0001,
+      "loss": 1.5207,
+      "step": 2559
+    },
+    {
+      "epoch": 0.020093770931011386,
+      "grad_norm": 0.13520407676696777,
+      "learning_rate": 0.0001,
+      "loss": 1.4977,
+      "step": 2560
+    },
+    {
+      "epoch": 0.02042866711319491,
+      "grad_norm": 0.14743876457214355,
+      "learning_rate": 0.0001,
+      "loss": 1.4901,
+      "step": 2561
+    },
+    {
+      "epoch": 0.020763563295378432,
+      "grad_norm": 0.1442539095878601,
+      "learning_rate": 0.0001,
+      "loss": 1.5069,
+      "step": 2562
+    },
+    {
+      "epoch": 0.021098459477561955,
+      "grad_norm": 0.14434701204299927,
+      "learning_rate": 0.0001,
+      "loss": 1.5274,
+      "step": 2563
+    },
+    {
+      "epoch": 0.021433355659745478,
+      "grad_norm": 0.1463836133480072,
+      "learning_rate": 0.0001,
+      "loss": 1.5482,
+      "step": 2564
+    },
+    {
+      "epoch": 0.021768251841929,
+      "grad_norm": 0.1377059519290924,
+      "learning_rate": 0.0001,
+      "loss": 1.4227,
+      "step": 2565
+    },
+    {
+      "epoch": 0.022103148024112524,
+      "grad_norm": 0.13638120889663696,
+      "learning_rate": 0.0001,
+      "loss": 1.4764,
+      "step": 2566
+    },
+    {
+      "epoch": 0.022438044206296047,
+      "grad_norm": 0.15557897090911865,
+      "learning_rate": 0.0001,
+      "loss": 1.5515,
+      "step": 2567
+    },
+    {
+      "epoch": 0.02277294038847957,
+      "grad_norm": 0.14417023956775665,
+      "learning_rate": 0.0001,
+      "loss": 1.636,
+      "step": 2568
+    },
+    {
+      "epoch": 0.023107836570663093,
+      "grad_norm": 0.1372545212507248,
+      "learning_rate": 0.0001,
+      "loss": 1.4169,
+      "step": 2569
+    },
+    {
+      "epoch": 0.02344273275284662,
+      "grad_norm": 0.1555853635072708,
+      "learning_rate": 0.0001,
+      "loss": 1.4845,
+      "step": 2570
+    },
+    {
+      "epoch": 0.023777628935030142,
+      "grad_norm": 0.15075907111167908,
+      "learning_rate": 0.0001,
+      "loss": 1.4158,
+      "step": 2571
+    },
+    {
+      "epoch": 0.024112525117213665,
+      "grad_norm": 0.132953479886055,
+      "learning_rate": 0.0001,
+      "loss": 1.3884,
+      "step": 2572
+    },
+    {
+      "epoch": 0.024447421299397188,
+      "grad_norm": 0.1466650664806366,
+      "learning_rate": 0.0001,
+      "loss": 1.4772,
+      "step": 2573
+    },
+    {
+      "epoch": 0.02478231748158071,
+      "grad_norm": 0.1482735276222229,
+      "learning_rate": 0.0001,
+      "loss": 1.5177,
+      "step": 2574
+    },
+    {
+      "epoch": 0.025117213663764234,
+      "grad_norm": 0.13384978473186493,
+      "learning_rate": 0.0001,
+      "loss": 1.4631,
+      "step": 2575
+    },
+    {
+      "epoch": 0.025452109845947757,
+      "grad_norm": 0.14189793169498444,
+      "learning_rate": 0.0001,
+      "loss": 1.4961,
+      "step": 2576
+    },
+    {
+      "epoch": 0.02578700602813128,
+      "grad_norm": 0.1458515226840973,
+      "learning_rate": 0.0001,
+      "loss": 1.4043,
+      "step": 2577
+    },
+    {
+      "epoch": 0.026121902210314803,
+      "grad_norm": 0.1455136090517044,
+      "learning_rate": 0.0001,
+      "loss": 1.4856,
+      "step": 2578
+    },
+    {
+      "epoch": 0.026456798392498326,
+      "grad_norm": 0.1397375911474228,
+      "learning_rate": 0.0001,
+      "loss": 1.4641,
+      "step": 2579
+    },
+    {
+      "epoch": 0.02679169457468185,
+      "grad_norm": 0.15315301716327667,
+      "learning_rate": 0.0001,
+      "loss": 1.5293,
+      "step": 2580
+    },
+    {
+      "epoch": 0.02712659075686537,
+      "grad_norm": 0.14230938255786896,
+      "learning_rate": 0.0001,
+      "loss": 1.5069,
+      "step": 2581
+    },
+    {
+      "epoch": 0.027461486939048894,
+      "grad_norm": 0.1534007042646408,
+      "learning_rate": 0.0001,
+      "loss": 1.4823,
+      "step": 2582
+    },
+    {
+      "epoch": 0.027796383121232417,
+      "grad_norm": 0.1446549892425537,
+      "learning_rate": 0.0001,
+      "loss": 1.5113,
+      "step": 2583
+    },
+    {
+      "epoch": 0.02813127930341594,
+      "grad_norm": 0.14052511751651764,
+      "learning_rate": 0.0001,
+      "loss": 1.4786,
+      "step": 2584
+    },
+    {
+      "epoch": 0.028466175485599463,
+      "grad_norm": 0.1368207037448883,
+      "learning_rate": 0.0001,
+      "loss": 1.4235,
+      "step": 2585
+    },
+    {
+      "epoch": 0.028801071667782986,
+      "grad_norm": 0.15163308382034302,
+      "learning_rate": 0.0001,
+      "loss": 1.5705,
+      "step": 2586
+    },
+    {
+      "epoch": 0.02913596784996651,
+      "grad_norm": 0.13964402675628662,
+      "learning_rate": 0.0001,
+      "loss": 1.4089,
+      "step": 2587
+    },
+    {
+      "epoch": 0.029470864032150032,
+      "grad_norm": 0.13717153668403625,
+      "learning_rate": 0.0001,
+      "loss": 1.4685,
+      "step": 2588
+    },
+    {
+      "epoch": 0.029805760214333555,
+      "grad_norm": 0.15048347413539886,
+      "learning_rate": 0.0001,
+      "loss": 1.5594,
+      "step": 2589
+    },
+    {
+      "epoch": 0.03014065639651708,
+      "grad_norm": 0.14722289144992828,
+      "learning_rate": 0.0001,
+      "loss": 1.4536,
+      "step": 2590
+    },
+    {
+      "epoch": 0.030475552578700604,
+      "grad_norm": 0.14477623999118805,
+      "learning_rate": 0.0001,
+      "loss": 1.4998,
+      "step": 2591
+    },
+    {
+      "epoch": 0.030810448760884127,
+      "grad_norm": 0.14907695353031158,
+      "learning_rate": 0.0001,
+      "loss": 1.5043,
+      "step": 2592
+    },
+    {
+      "epoch": 0.03114534494306765,
+      "grad_norm": 0.14083077013492584,
+      "learning_rate": 0.0001,
+      "loss": 1.4757,
+      "step": 2593
+    },
+    {
+      "epoch": 0.03148024112525117,
+      "grad_norm": 0.13702519237995148,
+      "learning_rate": 0.0001,
+      "loss": 1.4473,
+      "step": 2594
+    },
+    {
+      "epoch": 0.031815137307434696,
+      "grad_norm": 0.14483878016471863,
+      "learning_rate": 0.0001,
+      "loss": 1.5523,
+      "step": 2595
+    },
+    {
+      "epoch": 0.032150033489618215,
+      "grad_norm": 0.14438624680042267,
+      "learning_rate": 0.0001,
+      "loss": 1.5179,
+      "step": 2596
+    },
+    {
+      "epoch": 0.03248492967180174,
+      "grad_norm": 0.13237686455249786,
+      "learning_rate": 0.0001,
+      "loss": 1.4097,
+      "step": 2597
+    },
+    {
+      "epoch": 0.03281982585398526,
+      "grad_norm": 0.1421421617269516,
+      "learning_rate": 0.0001,
+      "loss": 1.4931,
+      "step": 2598
+    },
+    {
+      "epoch": 0.03315472203616879,
+      "grad_norm": 0.1436019092798233,
+      "learning_rate": 0.0001,
+      "loss": 1.493,
+      "step": 2599
+    },
+    {
+      "epoch": 0.033489618218352314,
+      "grad_norm": 0.14162643253803253,
+      "learning_rate": 0.0001,
+      "loss": 1.4743,
+      "step": 2600
+    },
+    {
+      "epoch": 0.033824514400535834,
+      "grad_norm": 0.13866618275642395,
+      "learning_rate": 0.0001,
+      "loss": 1.4849,
+      "step": 2601
+    },
+    {
+      "epoch": 0.03415941058271936,
+      "grad_norm": 0.1426554024219513,
+      "learning_rate": 0.0001,
+      "loss": 1.5222,
+      "step": 2602
+    },
+    {
+      "epoch": 0.03449430676490288,
+      "grad_norm": 0.14943377673625946,
+      "learning_rate": 0.0001,
+      "loss": 1.6258,
+      "step": 2603
+    },
+    {
+      "epoch": 0.034829202947086406,
+      "grad_norm": 0.14554472267627716,
+      "learning_rate": 0.0001,
+      "loss": 1.5683,
+      "step": 2604
+    },
+    {
+      "epoch": 0.035164099129269925,
+      "grad_norm": 0.15278968214988708,
+      "learning_rate": 0.0001,
+      "loss": 1.5867,
+      "step": 2605
+    },
+    {
+      "epoch": 0.03549899531145345,
+      "grad_norm": 0.1403978019952774,
+      "learning_rate": 0.0001,
+      "loss": 1.4167,
+      "step": 2606
+    },
+    {
+      "epoch": 0.03583389149363697,
+      "grad_norm": 0.14401455223560333,
+      "learning_rate": 0.0001,
+      "loss": 1.5629,
+      "step": 2607
+    },
+    {
+      "epoch": 0.0361687876758205,
+      "grad_norm": 0.14237728714942932,
+      "learning_rate": 0.0001,
+      "loss": 1.4703,
+      "step": 2608
+    },
+    {
+      "epoch": 0.03650368385800402,
+      "grad_norm": 0.15331515669822693,
+      "learning_rate": 0.0001,
+      "loss": 1.4727,
+      "step": 2609
+    },
+    {
+      "epoch": 0.036838580040187544,
+      "grad_norm": 0.13513866066932678,
+      "learning_rate": 0.0001,
+      "loss": 1.4841,
+      "step": 2610
+    },
+    {
+      "epoch": 0.03717347622237106,
+      "grad_norm": 0.1455763429403305,
+      "learning_rate": 0.0001,
+      "loss": 1.5234,
+      "step": 2611
+    },
+    {
+      "epoch": 0.03750837240455459,
+      "grad_norm": 0.15021777153015137,
+      "learning_rate": 0.0001,
+      "loss": 1.4737,
+      "step": 2612
+    },
+    {
+      "epoch": 0.03784326858673811,
+      "grad_norm": 0.1495695263147354,
+      "learning_rate": 0.0001,
+      "loss": 1.4734,
+      "step": 2613
+    },
+    {
+      "epoch": 0.038178164768921635,
+      "grad_norm": 0.14363637566566467,
+      "learning_rate": 0.0001,
+      "loss": 1.5417,
+      "step": 2614
+    },
+    {
+      "epoch": 0.038513060951105155,
+      "grad_norm": 0.14860506355762482,
+      "learning_rate": 0.0001,
+      "loss": 1.4967,
+      "step": 2615
+    },
+    {
+      "epoch": 0.03884795713328868,
+      "grad_norm": 0.138936385512352,
+      "learning_rate": 0.0001,
+      "loss": 1.5451,
+      "step": 2616
+    },
+    {
+      "epoch": 0.0391828533154722,
+      "grad_norm": 0.13959284126758575,
+      "learning_rate": 0.0001,
+      "loss": 1.4186,
+      "step": 2617
+    },
+    {
+      "epoch": 0.03951774949765573,
+      "grad_norm": 0.13905645906925201,
+      "learning_rate": 0.0001,
+      "loss": 1.4981,
+      "step": 2618
+    },
+    {
+      "epoch": 0.039852645679839246,
+      "grad_norm": 0.13835424184799194,
+      "learning_rate": 0.0001,
+      "loss": 1.4646,
+      "step": 2619
+    },
+    {
+      "epoch": 0.04018754186202277,
+      "grad_norm": 0.1373290866613388,
+      "learning_rate": 0.0001,
+      "loss": 1.461,
+      "step": 2620
+    },
+    {
+      "epoch": 0.0405224380442063,
+      "grad_norm": 0.1450372189283371,
+      "learning_rate": 0.0001,
+      "loss": 1.4827,
+      "step": 2621
+    },
+    {
+      "epoch": 0.04085733422638982,
+      "grad_norm": 0.1416875571012497,
+      "learning_rate": 0.0001,
+      "loss": 1.4285,
+      "step": 2622
+    },
+    {
+      "epoch": 0.041192230408573345,
+      "grad_norm": 0.14558644592761993,
+      "learning_rate": 0.0001,
+      "loss": 1.4835,
+      "step": 2623
+    },
+    {
+      "epoch": 0.041527126590756865,
+      "grad_norm": 0.14103256165981293,
+      "learning_rate": 0.0001,
+      "loss": 1.4838,
+      "step": 2624
+    },
+    {
+      "epoch": 0.04186202277294039,
+      "grad_norm": 0.13511677086353302,
+      "learning_rate": 0.0001,
+      "loss": 1.4154,
+      "step": 2625
+    },
+    {
+      "epoch": 0.04219691895512391,
+      "grad_norm": 0.1407540738582611,
+      "learning_rate": 0.0001,
+      "loss": 1.3846,
+      "step": 2626
+    },
+    {
+      "epoch": 0.04253181513730744,
+      "grad_norm": 0.13562853634357452,
+      "learning_rate": 0.0001,
+      "loss": 1.3934,
+      "step": 2627
+    },
+    {
+      "epoch": 0.042866711319490956,
+      "grad_norm": 0.14359675347805023,
+      "learning_rate": 0.0001,
+      "loss": 1.5502,
+      "step": 2628
+    },
+    {
+      "epoch": 0.04320160750167448,
+      "grad_norm": 0.14287690818309784,
+      "learning_rate": 0.0001,
+      "loss": 1.5875,
+      "step": 2629
+    },
+    {
+      "epoch": 0.043536503683858,
+      "grad_norm": 0.14544172585010529,
+      "learning_rate": 0.0001,
+      "loss": 1.5727,
+      "step": 2630
+    },
+    {
+      "epoch": 0.04387139986604153,
+      "grad_norm": 0.13922414183616638,
+      "learning_rate": 0.0001,
+      "loss": 1.3776,
+      "step": 2631
+    },
+    {
+      "epoch": 0.04420629604822505,
+      "grad_norm": 0.13899439573287964,
+      "learning_rate": 0.0001,
+      "loss": 1.4755,
+      "step": 2632
+    },
+    {
+      "epoch": 0.044541192230408574,
+      "grad_norm": 0.14173531532287598,
+      "learning_rate": 0.0001,
+      "loss": 1.4999,
+      "step": 2633
+    },
+    {
+      "epoch": 0.044876088412592094,
+      "grad_norm": 0.13438910245895386,
+      "learning_rate": 0.0001,
+      "loss": 1.4053,
+      "step": 2634
+    },
+    {
+      "epoch": 0.04521098459477562,
+      "grad_norm": 0.14383186399936676,
+      "learning_rate": 0.0001,
+      "loss": 1.515,
+      "step": 2635
+    },
+    {
+      "epoch": 0.04554588077695914,
+      "grad_norm": 0.1358998566865921,
+      "learning_rate": 0.0001,
+      "loss": 1.4713,
+      "step": 2636
+    },
+    {
+      "epoch": 0.045880776959142666,
+      "grad_norm": 0.13926327228546143,
+      "learning_rate": 0.0001,
+      "loss": 1.443,
+      "step": 2637
+    },
+    {
+      "epoch": 0.046215673141326186,
+      "grad_norm": 0.14442448318004608,
+      "learning_rate": 0.0001,
+      "loss": 1.5082,
+      "step": 2638
+    },
+    {
+      "epoch": 0.04655056932350971,
+      "grad_norm": 0.13538411259651184,
+      "learning_rate": 0.0001,
+      "loss": 1.4567,
+      "step": 2639
+    },
+    {
+      "epoch": 0.04688546550569324,
+      "grad_norm": 0.14301930367946625,
+      "learning_rate": 0.0001,
+      "loss": 1.4305,
+      "step": 2640
+    },
+    {
+      "epoch": 0.04722036168787676,
+      "grad_norm": 0.13956156373023987,
+      "learning_rate": 0.0001,
+      "loss": 1.4996,
+      "step": 2641
+    },
+    {
+      "epoch": 0.047555257870060284,
+      "grad_norm": 0.13752375543117523,
+      "learning_rate": 0.0001,
+      "loss": 1.4959,
+      "step": 2642
+    },
+    {
+      "epoch": 0.047890154052243804,
+      "grad_norm": 0.14346663653850555,
+      "learning_rate": 0.0001,
+      "loss": 1.5298,
+      "step": 2643
+    },
+    {
+      "epoch": 0.04822505023442733,
+      "grad_norm": 0.14569582045078278,
+      "learning_rate": 0.0001,
+      "loss": 1.4791,
+      "step": 2644
+    },
+    {
+      "epoch": 0.04855994641661085,
+      "grad_norm": 0.1401805430650711,
+      "learning_rate": 0.0001,
+      "loss": 1.446,
+      "step": 2645
+    },
+    {
+      "epoch": 0.048894842598794376,
+      "grad_norm": 0.14419414103031158,
+      "learning_rate": 0.0001,
+      "loss": 1.4684,
+      "step": 2646
+    },
+    {
+      "epoch": 0.049229738780977896,
+      "grad_norm": 0.15199334919452667,
+      "learning_rate": 0.0001,
+      "loss": 1.4544,
+      "step": 2647
+    },
+    {
+      "epoch": 0.04956463496316142,
+      "grad_norm": 0.13346783816814423,
+      "learning_rate": 0.0001,
+      "loss": 1.4106,
+      "step": 2648
+    },
+    {
+      "epoch": 0.04989953114534494,
+      "grad_norm": 0.14965283870697021,
+      "learning_rate": 0.0001,
+      "loss": 1.4939,
+      "step": 2649
+    },
+    {
+      "epoch": 0.05023442732752847,
+      "grad_norm": 0.15457184612751007,
+      "learning_rate": 0.0001,
+      "loss": 1.5083,
+      "step": 2650
+    },
+    {
+      "epoch": 0.05056932350971199,
+      "grad_norm": 0.13663454353809357,
+      "learning_rate": 0.0001,
+      "loss": 1.5019,
+      "step": 2651
+    },
+    {
+      "epoch": 0.050904219691895514,
+      "grad_norm": 0.14018338918685913,
+      "learning_rate": 0.0001,
+      "loss": 1.4765,
+      "step": 2652
+    },
+    {
+      "epoch": 0.05123911587407903,
+      "grad_norm": 0.14040859043598175,
+      "learning_rate": 0.0001,
+      "loss": 1.4675,
+      "step": 2653
+    },
+    {
+      "epoch": 0.05157401205626256,
+      "grad_norm": 0.14722508192062378,
+      "learning_rate": 0.0001,
+      "loss": 1.4704,
+      "step": 2654
+    },
+    {
+      "epoch": 0.05190890823844608,
+      "grad_norm": 0.1608511507511139,
+      "learning_rate": 0.0001,
+      "loss": 1.5832,
+      "step": 2655
+    },
+    {
+      "epoch": 0.052243804420629605,
+      "grad_norm": 0.14078736305236816,
+      "learning_rate": 0.0001,
+      "loss": 1.4143,
+      "step": 2656
+    },
+    {
+      "epoch": 0.052578700602813125,
+      "grad_norm": 0.14444111287593842,
+      "learning_rate": 0.0001,
+      "loss": 1.5433,
+      "step": 2657
+    },
+    {
+      "epoch": 0.05291359678499665,
+      "grad_norm": 0.1437847912311554,
+      "learning_rate": 0.0001,
+      "loss": 1.5246,
+      "step": 2658
+    },
+    {
+      "epoch": 0.05324849296718017,
+      "grad_norm": 0.13621410727500916,
+      "learning_rate": 0.0001,
+      "loss": 1.4587,
+      "step": 2659
+    },
+    {
+      "epoch": 0.0535833891493637,
+      "grad_norm": 0.1466796100139618,
+      "learning_rate": 0.0001,
+      "loss": 1.4839,
+      "step": 2660
+    },
+    {
+      "epoch": 0.053918285331547224,
+      "grad_norm": 0.14355647563934326,
+      "learning_rate": 0.0001,
+      "loss": 1.5311,
+      "step": 2661
+    },
+    {
+      "epoch": 0.05425318151373074,
+      "grad_norm": 0.14155465364456177,
+      "learning_rate": 0.0001,
+      "loss": 1.4321,
+      "step": 2662
+    },
+    {
+      "epoch": 0.05458807769591427,
+      "grad_norm": 0.13682101666927338,
+      "learning_rate": 0.0001,
+      "loss": 1.436,
+      "step": 2663
+    },
+    {
+      "epoch": 0.05492297387809779,
+      "grad_norm": 0.14461514353752136,
+      "learning_rate": 0.0001,
+      "loss": 1.491,
+      "step": 2664
+    },
+    {
+      "epoch": 0.055257870060281315,
+      "grad_norm": 0.14589102566242218,
+      "learning_rate": 0.0001,
+      "loss": 1.452,
+      "step": 2665
+    },
+    {
+      "epoch": 0.055592766242464835,
+      "grad_norm": 0.13621005415916443,
+      "learning_rate": 0.0001,
+      "loss": 1.4085,
+      "step": 2666
+    },
+    {
+      "epoch": 0.05592766242464836,
+      "grad_norm": 0.13911636173725128,
+      "learning_rate": 0.0001,
+      "loss": 1.5222,
+      "step": 2667
+    },
+    {
+      "epoch": 0.05626255860683188,
+      "grad_norm": 0.14434079825878143,
+      "learning_rate": 0.0001,
+      "loss": 1.5304,
+      "step": 2668
+    },
+    {
+      "epoch": 0.05659745478901541,
+      "grad_norm": 0.1485336571931839,
+      "learning_rate": 0.0001,
+      "loss": 1.5097,
+      "step": 2669
+    },
+    {
+      "epoch": 0.056932350971198926,
+      "grad_norm": 0.14089222252368927,
+      "learning_rate": 0.0001,
+      "loss": 1.4742,
+      "step": 2670
+    },
+    {
+      "epoch": 0.05726724715338245,
+      "grad_norm": 0.1406714767217636,
+      "learning_rate": 0.0001,
+      "loss": 1.4688,
+      "step": 2671
+    },
+    {
+      "epoch": 0.05760214333556597,
+      "grad_norm": 0.13756093382835388,
+      "learning_rate": 0.0001,
+      "loss": 1.4576,
+      "step": 2672
+    },
+    {
+      "epoch": 0.0579370395177495,
+      "grad_norm": 0.13884751498699188,
+      "learning_rate": 0.0001,
+      "loss": 1.5325,
+      "step": 2673
+    },
+    {
+      "epoch": 0.05827193569993302,
+      "grad_norm": 0.14231450855731964,
+      "learning_rate": 0.0001,
+      "loss": 1.4978,
+      "step": 2674
+    },
+    {
+      "epoch": 0.058606831882116545,
+      "grad_norm": 0.14200443029403687,
+      "learning_rate": 0.0001,
+      "loss": 1.5245,
+      "step": 2675
+    },
+    {
+      "epoch": 0.058941728064300064,
+      "grad_norm": 0.1436898410320282,
+      "learning_rate": 0.0001,
+      "loss": 1.4993,
+      "step": 2676
+    },
+    {
+      "epoch": 0.05927662424648359,
+      "grad_norm": 0.1421404331922531,
+      "learning_rate": 0.0001,
+      "loss": 1.4759,
+      "step": 2677
+    },
+    {
+      "epoch": 0.05961152042866711,
+      "grad_norm": 0.14410701394081116,
+      "learning_rate": 0.0001,
+      "loss": 1.4643,
+      "step": 2678
+    },
+    {
+      "epoch": 0.059946416610850636,
+      "grad_norm": 0.14235490560531616,
+      "learning_rate": 0.0001,
+      "loss": 1.5351,
+      "step": 2679
+    },
+    {
+      "epoch": 0.06028131279303416,
+      "grad_norm": 0.1442299336194992,
+      "learning_rate": 0.0001,
+      "loss": 1.5266,
+      "step": 2680
+    },
+    {
+      "epoch": 0.06061620897521768,
+      "grad_norm": 0.1418464481830597,
+      "learning_rate": 0.0001,
+      "loss": 1.5767,
+      "step": 2681
+    },
+    {
+      "epoch": 0.06095110515740121,
+      "grad_norm": 0.14591242372989655,
+      "learning_rate": 0.0001,
+      "loss": 1.4796,
+      "step": 2682
+    },
+    {
+      "epoch": 0.06128600133958473,
+      "grad_norm": 0.14850719273090363,
+      "learning_rate": 0.0001,
+      "loss": 1.4836,
+      "step": 2683
+    },
+    {
+      "epoch": 0.061620897521768254,
+      "grad_norm": 0.14169856905937195,
+      "learning_rate": 0.0001,
+      "loss": 1.4769,
+      "step": 2684
+    },
+    {
+      "epoch": 0.061955793703951774,
+      "grad_norm": 0.14849288761615753,
+      "learning_rate": 0.0001,
+      "loss": 1.5431,
+      "step": 2685
+    },
+    {
+      "epoch": 0.0622906898861353,
+      "grad_norm": 0.14057648181915283,
+      "learning_rate": 0.0001,
+      "loss": 1.4802,
+      "step": 2686
+    },
+    {
+      "epoch": 0.06262558606831882,
+      "grad_norm": 0.15280254185199738,
+      "learning_rate": 0.0001,
+      "loss": 1.4881,
+      "step": 2687
+    },
+    {
+      "epoch": 0.06296048225050234,
+      "grad_norm": 0.13690420985221863,
+      "learning_rate": 0.0001,
+      "loss": 1.374,
+      "step": 2688
+    },
+    {
+      "epoch": 0.06329537843268587,
+      "grad_norm": 0.14229778945446014,
+      "learning_rate": 0.0001,
+      "loss": 1.4981,
+      "step": 2689
+    },
+    {
+      "epoch": 0.06363027461486939,
+      "grad_norm": 0.14442794024944305,
+      "learning_rate": 0.0001,
+      "loss": 1.4799,
+      "step": 2690
+    },
+    {
+      "epoch": 0.06396517079705291,
+      "grad_norm": 0.15053285658359528,
+      "learning_rate": 0.0001,
+      "loss": 1.566,
+      "step": 2691
+    },
+    {
+      "epoch": 0.06430006697923643,
+      "grad_norm": 0.1395815759897232,
+      "learning_rate": 0.0001,
+      "loss": 1.4605,
+      "step": 2692
+    },
+    {
+      "epoch": 0.06463496316141996,
+      "grad_norm": 0.15144221484661102,
+      "learning_rate": 0.0001,
+      "loss": 1.4642,
+      "step": 2693
+    },
+    {
+      "epoch": 0.06496985934360348,
+      "grad_norm": 0.14683520793914795,
+      "learning_rate": 0.0001,
+      "loss": 1.4647,
+      "step": 2694
+    },
+    {
+      "epoch": 0.065304755525787,
+      "grad_norm": 0.1422366201877594,
+      "learning_rate": 0.0001,
+      "loss": 1.5096,
+      "step": 2695
+    },
+    {
+      "epoch": 0.06563965170797052,
+      "grad_norm": 0.14606067538261414,
+      "learning_rate": 0.0001,
+      "loss": 1.5396,
+      "step": 2696
+    },
+    {
+      "epoch": 0.06597454789015406,
+      "grad_norm": 0.13931868970394135,
+      "learning_rate": 0.0001,
+      "loss": 1.4693,
+      "step": 2697
+    },
+    {
+      "epoch": 0.06630944407233758,
+      "grad_norm": 0.14257977902889252,
+      "learning_rate": 0.0001,
+      "loss": 1.5156,
+      "step": 2698
+    },
+    {
+      "epoch": 0.0666443402545211,
+      "grad_norm": 0.14814600348472595,
+      "learning_rate": 0.0001,
+      "loss": 1.5187,
+      "step": 2699
+    },
+    {
+      "epoch": 0.06697923643670463,
+      "grad_norm": 0.14432038366794586,
+      "learning_rate": 0.0001,
+      "loss": 1.4105,
+      "step": 2700
+    },
+    {
+      "epoch": 0.06731413261888815,
+      "grad_norm": 0.14682446420192719,
+      "learning_rate": 0.0001,
+      "loss": 1.5516,
+      "step": 2701
+    },
+    {
+      "epoch": 0.06764902880107167,
+      "grad_norm": 0.14887043833732605,
+      "learning_rate": 0.0001,
+      "loss": 1.5903,
+      "step": 2702
+    },
+    {
+      "epoch": 0.06798392498325519,
+      "grad_norm": 0.14733371138572693,
+      "learning_rate": 0.0001,
+      "loss": 1.5192,
+      "step": 2703
+    },
+    {
+      "epoch": 0.06831882116543872,
+      "grad_norm": 0.14175830781459808,
+      "learning_rate": 0.0001,
+      "loss": 1.468,
+      "step": 2704
+    },
+    {
+      "epoch": 0.06865371734762224,
+      "grad_norm": 0.13812625408172607,
+      "learning_rate": 0.0001,
+      "loss": 1.4148,
+      "step": 2705
+    },
+    {
+      "epoch": 0.06898861352980576,
+      "grad_norm": 0.14035607874393463,
+      "learning_rate": 0.0001,
+      "loss": 1.4531,
+      "step": 2706
+    },
+    {
+      "epoch": 0.06932350971198928,
+      "grad_norm": 0.14815551042556763,
+      "learning_rate": 0.0001,
+      "loss": 1.5359,
+      "step": 2707
+    },
+    {
+      "epoch": 0.06965840589417281,
+      "grad_norm": 0.14825250208377838,
+      "learning_rate": 0.0001,
+      "loss": 1.4737,
+      "step": 2708
+    },
+    {
+      "epoch": 0.06999330207635633,
+      "grad_norm": 0.14412347972393036,
+      "learning_rate": 0.0001,
+      "loss": 1.4155,
+      "step": 2709
+    },
+    {
+      "epoch": 0.07032819825853985,
+      "grad_norm": 0.14373992383480072,
+      "learning_rate": 0.0001,
+      "loss": 1.5142,
+      "step": 2710
+    },
+    {
+      "epoch": 0.07066309444072337,
+      "grad_norm": 0.14845426380634308,
+      "learning_rate": 0.0001,
+      "loss": 1.6058,
+      "step": 2711
+    },
+    {
+      "epoch": 0.0709979906229069,
+      "grad_norm": 0.15275675058364868,
+      "learning_rate": 0.0001,
+      "loss": 1.5081,
+      "step": 2712
+    },
+    {
+      "epoch": 0.07133288680509042,
+      "grad_norm": 0.1481783539056778,
+      "learning_rate": 0.0001,
+      "loss": 1.5469,
+      "step": 2713
+    },
+    {
+      "epoch": 0.07166778298727394,
+      "grad_norm": 0.14404213428497314,
+      "learning_rate": 0.0001,
+      "loss": 1.6133,
+      "step": 2714
+    },
+    {
+      "epoch": 0.07200267916945746,
+      "grad_norm": 0.138307586312294,
+      "learning_rate": 0.0001,
+      "loss": 1.5124,
+      "step": 2715
+    },
+    {
+      "epoch": 0.072337575351641,
+      "grad_norm": 0.14422239363193512,
+      "learning_rate": 0.0001,
+      "loss": 1.4536,
+      "step": 2716
+    },
+    {
+      "epoch": 0.07267247153382451,
+      "grad_norm": 0.14184726774692535,
+      "learning_rate": 0.0001,
+      "loss": 1.4547,
+      "step": 2717
+    },
+    {
+      "epoch": 0.07300736771600803,
+      "grad_norm": 0.1343645304441452,
+      "learning_rate": 0.0001,
+      "loss": 1.4741,
+      "step": 2718
+    },
+    {
+      "epoch": 0.07334226389819157,
+      "grad_norm": 0.14284183084964752,
+      "learning_rate": 0.0001,
+      "loss": 1.5052,
+      "step": 2719
+    },
+    {
+      "epoch": 0.07367716008037509,
+      "grad_norm": 0.14086835086345673,
+      "learning_rate": 0.0001,
+      "loss": 1.5035,
+      "step": 2720
+    },
+    {
+      "epoch": 0.0740120562625586,
+      "grad_norm": 0.146831676363945,
+      "learning_rate": 0.0001,
+      "loss": 1.4494,
+      "step": 2721
+    },
+    {
+      "epoch": 0.07434695244474213,
+      "grad_norm": 0.14096811413764954,
+      "learning_rate": 0.0001,
+      "loss": 1.3873,
+      "step": 2722
+    },
+    {
+      "epoch": 0.07468184862692566,
+      "grad_norm": 0.1556435525417328,
+      "learning_rate": 0.0001,
+      "loss": 1.5451,
+      "step": 2723
+    },
+    {
+      "epoch": 0.07501674480910918,
+      "grad_norm": 0.1460019052028656,
+      "learning_rate": 0.0001,
+      "loss": 1.573,
+      "step": 2724
+    },
+    {
+      "epoch": 0.0753516409912927,
+      "grad_norm": 0.14590851962566376,
+      "learning_rate": 0.0001,
+      "loss": 1.4472,
+      "step": 2725
+    },
+    {
+      "epoch": 0.07568653717347622,
+      "grad_norm": 0.14320936799049377,
+      "learning_rate": 0.0001,
+      "loss": 1.5049,
+      "step": 2726
+    },
+    {
+      "epoch": 0.07602143335565975,
+      "grad_norm": 0.15445268154144287,
+      "learning_rate": 0.0001,
+      "loss": 1.5577,
+      "step": 2727
+    },
+    {
+      "epoch": 0.07635632953784327,
+      "grad_norm": 0.1448332965373993,
+      "learning_rate": 0.0001,
+      "loss": 1.5076,
+      "step": 2728
+    },
+    {
+      "epoch": 0.07669122572002679,
+      "grad_norm": 0.1451125591993332,
+      "learning_rate": 0.0001,
+      "loss": 1.5353,
+      "step": 2729
+    },
+    {
+      "epoch": 0.07702612190221031,
+      "grad_norm": 0.1508006453514099,
+      "learning_rate": 0.0001,
+      "loss": 1.653,
+      "step": 2730
+    },
+    {
+      "epoch": 0.07736101808439384,
+      "grad_norm": 0.13818366825580597,
+      "learning_rate": 0.0001,
+      "loss": 1.3885,
+      "step": 2731
+    },
+    {
+      "epoch": 0.07769591426657736,
+      "grad_norm": 0.13172827661037445,
+      "learning_rate": 0.0001,
+      "loss": 1.3705,
+      "step": 2732
+    },
+    {
+      "epoch": 0.07803081044876088,
+      "grad_norm": 0.1504860669374466,
+      "learning_rate": 0.0001,
+      "loss": 1.5108,
+      "step": 2733
+    },
+    {
+      "epoch": 0.0783657066309444,
+      "grad_norm": 0.1418769508600235,
+      "learning_rate": 0.0001,
+      "loss": 1.4078,
+      "step": 2734
+    },
+    {
+      "epoch": 0.07870060281312793,
+      "grad_norm": 0.14419272541999817,
+      "learning_rate": 0.0001,
+      "loss": 1.4953,
+      "step": 2735
+    },
+    {
+      "epoch": 0.07903549899531145,
+      "grad_norm": 0.14663195610046387,
+      "learning_rate": 0.0001,
+      "loss": 1.4626,
+      "step": 2736
+    },
+    {
+      "epoch": 0.07937039517749497,
+      "grad_norm": 0.13987226784229279,
+      "learning_rate": 0.0001,
+      "loss": 1.5421,
+      "step": 2737
+    },
+    {
+      "epoch": 0.07970529135967849,
+      "grad_norm": 0.1439124494791031,
+      "learning_rate": 0.0001,
+      "loss": 1.5178,
+      "step": 2738
+    },
+    {
+      "epoch": 0.08004018754186203,
+      "grad_norm": 0.1464778035879135,
+      "learning_rate": 0.0001,
+      "loss": 1.4798,
+      "step": 2739
+    },
+    {
+      "epoch": 0.08037508372404555,
+      "grad_norm": 0.13677044212818146,
+      "learning_rate": 0.0001,
+      "loss": 1.449,
+      "step": 2740
+    },
+    {
+      "epoch": 0.08070997990622907,
+      "grad_norm": 0.1404646635055542,
+      "learning_rate": 0.0001,
+      "loss": 1.5006,
+      "step": 2741
+    },
+    {
+      "epoch": 0.0810448760884126,
+      "grad_norm": 0.14357681572437286,
+      "learning_rate": 0.0001,
+      "loss": 1.5219,
+      "step": 2742
+    },
+    {
+      "epoch": 0.08137977227059612,
+      "grad_norm": 0.14373867213726044,
+      "learning_rate": 0.0001,
+      "loss": 1.4428,
+      "step": 2743
+    },
+    {
+      "epoch": 0.08171466845277964,
+      "grad_norm": 0.1410626769065857,
+      "learning_rate": 0.0001,
+      "loss": 1.4882,
+      "step": 2744
+    },
+    {
+      "epoch": 0.08204956463496316,
+      "grad_norm": 0.13724640011787415,
+      "learning_rate": 0.0001,
+      "loss": 1.4109,
+      "step": 2745
+    },
+    {
+      "epoch": 0.08238446081714669,
+      "grad_norm": 0.1481046825647354,
+      "learning_rate": 0.0001,
+      "loss": 1.614,
+      "step": 2746
+    },
+    {
+      "epoch": 0.08271935699933021,
+      "grad_norm": 0.1424095779657364,
+      "learning_rate": 0.0001,
+      "loss": 1.5868,
+      "step": 2747
+    },
+    {
+      "epoch": 0.08305425318151373,
+      "grad_norm": 0.14285503327846527,
+      "learning_rate": 0.0001,
+      "loss": 1.3974,
+      "step": 2748
+    },
+    {
+      "epoch": 0.08338914936369725,
+      "grad_norm": 0.1423054337501526,
+      "learning_rate": 0.0001,
+      "loss": 1.5738,
+      "step": 2749
+    },
+    {
+      "epoch": 0.08372404554588078,
+      "grad_norm": 0.14483210444450378,
+      "learning_rate": 0.0001,
+      "loss": 1.4587,
+      "step": 2750
+    },
+    {
+      "epoch": 0.0840589417280643,
+      "grad_norm": 0.14319851994514465,
+      "learning_rate": 0.0001,
+      "loss": 1.4642,
+      "step": 2751
+    },
+    {
+      "epoch": 0.08439383791024782,
+      "grad_norm": 0.13955651223659515,
+      "learning_rate": 0.0001,
+      "loss": 1.3657,
+      "step": 2752
+    },
+    {
+      "epoch": 0.08472873409243134,
+      "grad_norm": 0.1478879302740097,
+      "learning_rate": 0.0001,
+      "loss": 1.6474,
+      "step": 2753
+    },
+    {
+      "epoch": 0.08506363027461487,
+      "grad_norm": 0.14244432747364044,
+      "learning_rate": 0.0001,
+      "loss": 1.5459,
+      "step": 2754
+    },
+    {
+      "epoch": 0.0853985264567984,
+      "grad_norm": 0.14115175604820251,
+      "learning_rate": 0.0001,
+      "loss": 1.4614,
+      "step": 2755
+    },
+    {
+      "epoch": 0.08573342263898191,
+      "grad_norm": 0.13785597681999207,
+      "learning_rate": 0.0001,
+      "loss": 1.4257,
+      "step": 2756
+    },
+    {
+      "epoch": 0.08606831882116543,
+      "grad_norm": 0.14487232267856598,
+      "learning_rate": 0.0001,
+      "loss": 1.4983,
+      "step": 2757
+    },
+    {
+      "epoch": 0.08640321500334897,
+      "grad_norm": 0.14556938409805298,
+      "learning_rate": 0.0001,
+      "loss": 1.4263,
+      "step": 2758
+    },
+    {
+      "epoch": 0.08673811118553248,
+      "grad_norm": 0.1452976018190384,
+      "learning_rate": 0.0001,
+      "loss": 1.5328,
+      "step": 2759
+    },
+    {
+      "epoch": 0.087073007367716,
+      "grad_norm": 0.14344626665115356,
+      "learning_rate": 0.0001,
+      "loss": 1.4995,
+      "step": 2760
+    },
+    {
+      "epoch": 0.08740790354989954,
+      "grad_norm": 0.14486606419086456,
+      "learning_rate": 0.0001,
+      "loss": 1.5118,
+      "step": 2761
+    },
+    {
+      "epoch": 0.08774279973208306,
+      "grad_norm": 0.1462741494178772,
+      "learning_rate": 0.0001,
+      "loss": 1.4929,
+      "step": 2762
+    },
+    {
+      "epoch": 0.08807769591426658,
+      "grad_norm": 0.14603829383850098,
+      "learning_rate": 0.0001,
+      "loss": 1.5696,
+      "step": 2763
+    },
+    {
+      "epoch": 0.0884125920964501,
+      "grad_norm": 0.14310294389724731,
+      "learning_rate": 0.0001,
+      "loss": 1.4989,
+      "step": 2764
+    },
+    {
+      "epoch": 0.08874748827863363,
+      "grad_norm": 0.14336185157299042,
+      "learning_rate": 0.0001,
+      "loss": 1.4546,
+      "step": 2765
+    },
+    {
+      "epoch": 0.08908238446081715,
+      "grad_norm": 0.15111830830574036,
+      "learning_rate": 0.0001,
+      "loss": 1.5161,
+      "step": 2766
+    },
+    {
+      "epoch": 0.08941728064300067,
+      "grad_norm": 0.1415042281150818,
+      "learning_rate": 0.0001,
+      "loss": 1.4944,
+      "step": 2767
+    },
+    {
+      "epoch": 0.08975217682518419,
+      "grad_norm": 0.14492426812648773,
+      "learning_rate": 0.0001,
+      "loss": 1.4775,
+      "step": 2768
+    },
+    {
+      "epoch": 0.09008707300736772,
+      "grad_norm": 0.14594174921512604,
+      "learning_rate": 0.0001,
+      "loss": 1.4341,
+      "step": 2769
+    },
+    {
+      "epoch": 0.09042196918955124,
+      "grad_norm": 0.1443568468093872,
+      "learning_rate": 0.0001,
+      "loss": 1.4261,
+      "step": 2770
+    },
+    {
+      "epoch": 0.09075686537173476,
+      "grad_norm": 0.14747354388237,
+      "learning_rate": 0.0001,
+      "loss": 1.4934,
+      "step": 2771
+    },
+    {
+      "epoch": 0.09109176155391828,
+      "grad_norm": 0.14336217939853668,
+      "learning_rate": 0.0001,
+      "loss": 1.5243,
+      "step": 2772
+    },
+    {
+      "epoch": 0.09142665773610181,
+      "grad_norm": 0.1489473432302475,
+      "learning_rate": 0.0001,
+      "loss": 1.4997,
+      "step": 2773
+    },
+    {
+      "epoch": 0.09176155391828533,
+      "grad_norm": 0.140666201710701,
+      "learning_rate": 0.0001,
+      "loss": 1.4384,
+      "step": 2774
+    },
+    {
+      "epoch": 0.09209645010046885,
+      "grad_norm": 0.14660325646400452,
+      "learning_rate": 0.0001,
+      "loss": 1.4997,
+      "step": 2775
+    },
+    {
+      "epoch": 0.09243134628265237,
+      "grad_norm": 0.14790892601013184,
+      "learning_rate": 0.0001,
+      "loss": 1.4342,
+      "step": 2776
+    },
+    {
+      "epoch": 0.0927662424648359,
+      "grad_norm": 0.14560778439044952,
+      "learning_rate": 0.0001,
+      "loss": 1.5703,
+      "step": 2777
+    },
+    {
+      "epoch": 0.09310113864701942,
+      "grad_norm": 0.14003904163837433,
+      "learning_rate": 0.0001,
+      "loss": 1.5091,
+      "step": 2778
+    },
+    {
+      "epoch": 0.09343603482920294,
+      "grad_norm": 0.14353299140930176,
+      "learning_rate": 0.0001,
+      "loss": 1.5347,
+      "step": 2779
+    },
+    {
+      "epoch": 0.09377093101138648,
+      "grad_norm": 0.14216402173042297,
+      "learning_rate": 0.0001,
+      "loss": 1.4911,
+      "step": 2780
+    },
+    {
+      "epoch": 0.09410582719357,
+      "grad_norm": 0.14118923246860504,
+      "learning_rate": 0.0001,
+      "loss": 1.454,
+      "step": 2781
+    },
+    {
+      "epoch": 0.09444072337575352,
+      "grad_norm": 0.1451854258775711,
+      "learning_rate": 0.0001,
+      "loss": 1.5258,
+      "step": 2782
+    },
+    {
+      "epoch": 0.09477561955793704,
+      "grad_norm": 0.1455250382423401,
+      "learning_rate": 0.0001,
+      "loss": 1.508,
+      "step": 2783
+    },
+    {
+      "epoch": 0.09511051574012057,
+      "grad_norm": 0.1398870199918747,
+      "learning_rate": 0.0001,
+      "loss": 1.4157,
+      "step": 2784
+    },
+    {
+      "epoch": 0.09544541192230409,
+      "grad_norm": 0.1541321873664856,
+      "learning_rate": 0.0001,
+      "loss": 1.5076,
+      "step": 2785
+    },
+    {
+      "epoch": 0.09578030810448761,
+      "grad_norm": 0.14037346839904785,
+      "learning_rate": 0.0001,
+      "loss": 1.4294,
+      "step": 2786
+    },
+    {
+      "epoch": 0.09611520428667113,
+      "grad_norm": 0.15060150623321533,
+      "learning_rate": 0.0001,
+      "loss": 1.4742,
+      "step": 2787
+    },
+    {
+      "epoch": 0.09645010046885466,
+      "grad_norm": 0.1445634961128235,
+      "learning_rate": 0.0001,
+      "loss": 1.4149,
+      "step": 2788
+    },
+    {
+      "epoch": 0.09678499665103818,
+      "grad_norm": 0.1478995382785797,
+      "learning_rate": 0.0001,
+      "loss": 1.5271,
+      "step": 2789
+    },
+    {
+      "epoch": 0.0971198928332217,
+      "grad_norm": 0.1673794984817505,
+      "learning_rate": 0.0001,
+      "loss": 1.5274,
+      "step": 2790
+    },
+    {
+      "epoch": 0.09745478901540522,
+      "grad_norm": 0.14559200406074524,
+      "learning_rate": 0.0001,
+      "loss": 1.4834,
+      "step": 2791
+    },
+    {
+      "epoch": 0.09778968519758875,
+      "grad_norm": 0.14184166491031647,
+      "learning_rate": 0.0001,
+      "loss": 1.5399,
+      "step": 2792
+    },
+    {
+      "epoch": 0.09812458137977227,
+      "grad_norm": 0.14942292869091034,
+      "learning_rate": 0.0001,
+      "loss": 1.3988,
+      "step": 2793
+    },
+    {
+      "epoch": 0.09845947756195579,
+      "grad_norm": 0.14354535937309265,
+      "learning_rate": 0.0001,
+      "loss": 1.4625,
+      "step": 2794
+    },
+    {
+      "epoch": 0.09879437374413931,
+      "grad_norm": 0.14293980598449707,
+      "learning_rate": 0.0001,
+      "loss": 1.5965,
+      "step": 2795
+    },
+    {
+      "epoch": 0.09912926992632284,
+      "grad_norm": 0.1488644778728485,
+      "learning_rate": 0.0001,
+      "loss": 1.4126,
+      "step": 2796
+    },
+    {
+      "epoch": 0.09946416610850636,
+      "grad_norm": 0.13816925883293152,
+      "learning_rate": 0.0001,
+      "loss": 1.4674,
+      "step": 2797
+    },
+    {
+      "epoch": 0.09979906229068988,
+      "grad_norm": 0.1448117345571518,
+      "learning_rate": 0.0001,
+      "loss": 1.5199,
+      "step": 2798
+    },
+    {
+      "epoch": 0.10013395847287342,
+      "grad_norm": 0.14358732104301453,
+      "learning_rate": 0.0001,
+      "loss": 1.3845,
+      "step": 2799
+    },
+    {
+      "epoch": 0.10046885465505694,
+      "grad_norm": 0.1465434730052948,
+      "learning_rate": 0.0001,
+      "loss": 1.4358,
+      "step": 2800
+    },
+    {
+      "epoch": 0.10080375083724046,
+      "grad_norm": 0.14081235229969025,
+      "learning_rate": 0.0001,
+      "loss": 1.4383,
+      "step": 2801
+    },
+    {
+      "epoch": 0.10113864701942397,
+      "grad_norm": 0.14658094942569733,
+      "learning_rate": 0.0001,
+      "loss": 1.4229,
+      "step": 2802
+    },
+    {
+      "epoch": 0.10147354320160751,
+      "grad_norm": 0.14403977990150452,
+      "learning_rate": 0.0001,
+      "loss": 1.5351,
+      "step": 2803
+    },
+    {
+      "epoch": 0.10180843938379103,
+      "grad_norm": 0.1446755975484848,
+      "learning_rate": 0.0001,
+      "loss": 1.4485,
+      "step": 2804
+    },
+    {
+      "epoch": 0.10214333556597455,
+      "grad_norm": 0.14856703579425812,
+      "learning_rate": 0.0001,
+      "loss": 1.5629,
+      "step": 2805
+    },
+    {
+      "epoch": 0.10247823174815807,
+      "grad_norm": 0.14590945839881897,
+      "learning_rate": 0.0001,
+      "loss": 1.4851,
+      "step": 2806
+    },
+    {
+      "epoch": 0.1028131279303416,
+      "grad_norm": 0.14451883733272552,
+      "learning_rate": 0.0001,
+      "loss": 1.46,
+      "step": 2807
+    },
+    {
+      "epoch": 0.10314802411252512,
+      "grad_norm": 0.15517374873161316,
+      "learning_rate": 0.0001,
+      "loss": 1.5061,
+      "step": 2808
+    },
+    {
+      "epoch": 0.10348292029470864,
+      "grad_norm": 0.14735953509807587,
+      "learning_rate": 0.0001,
+      "loss": 1.514,
+      "step": 2809
+    },
+    {
+      "epoch": 0.10381781647689216,
+      "grad_norm": 0.1464541256427765,
+      "learning_rate": 0.0001,
+      "loss": 1.5269,
+      "step": 2810
+    },
+    {
+      "epoch": 0.10415271265907569,
+      "grad_norm": 0.15126463770866394,
+      "learning_rate": 0.0001,
+      "loss": 1.411,
+      "step": 2811
+    },
+    {
+      "epoch": 0.10448760884125921,
+      "grad_norm": 0.1394396275281906,
+      "learning_rate": 0.0001,
+      "loss": 1.5222,
+      "step": 2812
+    },
+    {
+      "epoch": 0.10482250502344273,
+      "grad_norm": 0.14099687337875366,
+      "learning_rate": 0.0001,
+      "loss": 1.4384,
+      "step": 2813
+    },
+    {
+      "epoch": 0.10515740120562625,
+      "grad_norm": 0.15170758962631226,
+      "learning_rate": 0.0001,
+      "loss": 1.5148,
+      "step": 2814
+    },
+    {
+      "epoch": 0.10549229738780978,
+      "grad_norm": 0.14324727654457092,
+      "learning_rate": 0.0001,
+      "loss": 1.4224,
+      "step": 2815
+    },
+    {
+      "epoch": 0.1058271935699933,
+      "grad_norm": 0.1460292786359787,
+      "learning_rate": 0.0001,
+      "loss": 1.558,
+      "step": 2816
+    },
+    {
+      "epoch": 0.10616208975217682,
+      "grad_norm": 0.1621580272912979,
+      "learning_rate": 0.0001,
+      "loss": 1.5001,
+      "step": 2817
+    },
+    {
+      "epoch": 0.10649698593436034,
+      "grad_norm": 0.14594002068042755,
+      "learning_rate": 0.0001,
+      "loss": 1.5567,
+      "step": 2818
+    },
+    {
+      "epoch": 0.10683188211654387,
+      "grad_norm": 0.1518278867006302,
+      "learning_rate": 0.0001,
+      "loss": 1.5166,
+      "step": 2819
+    },
+    {
+      "epoch": 0.1071667782987274,
+      "grad_norm": 0.1512691229581833,
+      "learning_rate": 0.0001,
+      "loss": 1.3993,
+      "step": 2820
+    },
+    {
+      "epoch": 0.10750167448091091,
+      "grad_norm": 0.14213557541370392,
+      "learning_rate": 0.0001,
+      "loss": 1.3772,
+      "step": 2821
+    },
+    {
+      "epoch": 0.10783657066309445,
+      "grad_norm": 0.14900755882263184,
+      "learning_rate": 0.0001,
+      "loss": 1.5243,
+      "step": 2822
+    },
+    {
+      "epoch": 0.10817146684527797,
+      "grad_norm": 0.15196503698825836,
+      "learning_rate": 0.0001,
+      "loss": 1.5012,
+      "step": 2823
+    },
+    {
+      "epoch": 0.10850636302746149,
+      "grad_norm": 0.14796164631843567,
+      "learning_rate": 0.0001,
+      "loss": 1.4849,
+      "step": 2824
+    },
+    {
+      "epoch": 0.108841259209645,
+      "grad_norm": 0.15282899141311646,
+      "learning_rate": 0.0001,
+      "loss": 1.5008,
+      "step": 2825
+    },
+    {
+      "epoch": 0.10917615539182854,
+      "grad_norm": 0.14410486817359924,
+      "learning_rate": 0.0001,
+      "loss": 1.4907,
+      "step": 2826
+    },
+    {
+      "epoch": 0.10951105157401206,
+      "grad_norm": 0.14892512559890747,
+      "learning_rate": 0.0001,
+      "loss": 1.4086,
+      "step": 2827
+    },
+    {
+      "epoch": 0.10984594775619558,
+      "grad_norm": 0.14852510392665863,
+      "learning_rate": 0.0001,
+      "loss": 1.5008,
+      "step": 2828
+    },
+    {
+      "epoch": 0.1101808439383791,
+      "grad_norm": 0.14441753923892975,
+      "learning_rate": 0.0001,
+      "loss": 1.4408,
+      "step": 2829
+    },
+    {
+      "epoch": 0.11051574012056263,
+      "grad_norm": 0.14280498027801514,
+      "learning_rate": 0.0001,
+      "loss": 1.5208,
+      "step": 2830
+    },
+    {
+      "epoch": 0.11085063630274615,
+      "grad_norm": 0.15437792241573334,
+      "learning_rate": 0.0001,
+      "loss": 1.5327,
+      "step": 2831
+    },
+    {
+      "epoch": 0.11118553248492967,
+      "grad_norm": 0.1393211930990219,
+      "learning_rate": 0.0001,
+      "loss": 1.492,
+      "step": 2832
+    },
+    {
+      "epoch": 0.11152042866711319,
+      "grad_norm": 0.15891778469085693,
+      "learning_rate": 0.0001,
+      "loss": 1.5739,
+      "step": 2833
+    },
+    {
+      "epoch": 0.11185532484929672,
+      "grad_norm": 0.15602029860019684,
+      "learning_rate": 0.0001,
+      "loss": 1.4942,
+      "step": 2834
+    },
+    {
+      "epoch": 0.11219022103148024,
+      "grad_norm": 0.15723031759262085,
+      "learning_rate": 0.0001,
+      "loss": 1.5449,
+      "step": 2835
+    },
+    {
+      "epoch": 0.11252511721366376,
+      "grad_norm": 0.1397925466299057,
+      "learning_rate": 0.0001,
+      "loss": 1.3486,
+      "step": 2836
+    },
+    {
+      "epoch": 0.11286001339584728,
+      "grad_norm": 0.16877742111682892,
+      "learning_rate": 0.0001,
+      "loss": 1.551,
+      "step": 2837
+    },
+    {
+      "epoch": 0.11319490957803081,
+      "grad_norm": 0.1419568657875061,
+      "learning_rate": 0.0001,
+      "loss": 1.4831,
+      "step": 2838
+    },
+    {
+      "epoch": 0.11352980576021433,
+      "grad_norm": 0.14331306517124176,
+      "learning_rate": 0.0001,
+      "loss": 1.3837,
+      "step": 2839
+    },
+    {
+      "epoch": 0.11386470194239785,
+      "grad_norm": 0.1620616614818573,
+      "learning_rate": 0.0001,
+      "loss": 1.4805,
+      "step": 2840
+    },
+    {
+      "epoch": 0.11419959812458139,
+      "grad_norm": 0.14861570298671722,
+      "learning_rate": 0.0001,
+      "loss": 1.5874,
+      "step": 2841
+    },
+    {
+      "epoch": 0.1145344943067649,
+      "grad_norm": 0.15495343506336212,
+      "learning_rate": 0.0001,
+      "loss": 1.5645,
+      "step": 2842
+    },
+    {
+      "epoch": 0.11486939048894843,
+      "grad_norm": 0.15632815659046173,
+      "learning_rate": 0.0001,
+      "loss": 1.4557,
+      "step": 2843
+    },
+    {
+      "epoch": 0.11520428667113194,
+      "grad_norm": 0.14908194541931152,
+      "learning_rate": 0.0001,
+      "loss": 1.5085,
+      "step": 2844
+    },
+    {
+      "epoch": 0.11553918285331548,
+      "grad_norm": 0.14396606385707855,
+      "learning_rate": 0.0001,
+      "loss": 1.4502,
+      "step": 2845
+    },
+    {
+      "epoch": 0.115874079035499,
+      "grad_norm": 0.16483518481254578,
+      "learning_rate": 0.0001,
+      "loss": 1.5383,
+      "step": 2846
+    },
+    {
+      "epoch": 0.11620897521768252,
+      "grad_norm": 0.14482854306697845,
+      "learning_rate": 0.0001,
+      "loss": 1.4928,
+      "step": 2847
+    },
+    {
+      "epoch": 0.11654387139986604,
+      "grad_norm": 0.15183790028095245,
+      "learning_rate": 0.0001,
+      "loss": 1.551,
+      "step": 2848
+    },
+    {
+      "epoch": 0.11687876758204957,
+      "grad_norm": 0.1584623008966446,
+      "learning_rate": 0.0001,
+      "loss": 1.5034,
+      "step": 2849
+    },
+    {
+      "epoch": 0.11721366376423309,
+      "grad_norm": 0.15252350270748138,
+      "learning_rate": 0.0001,
+      "loss": 1.6486,
+      "step": 2850
+    },
+    {
+      "epoch": 0.11754855994641661,
+      "grad_norm": 0.15280024707317352,
+      "learning_rate": 0.0001,
+      "loss": 1.5209,
+      "step": 2851
+    },
+    {
+      "epoch": 0.11788345612860013,
+      "grad_norm": 0.13550280034542084,
+      "learning_rate": 0.0001,
+      "loss": 1.372,
+      "step": 2852
+    },
+    {
+      "epoch": 0.11821835231078366,
+      "grad_norm": 0.14424198865890503,
+      "learning_rate": 0.0001,
+      "loss": 1.4848,
+      "step": 2853
+    },
+    {
+      "epoch": 0.11855324849296718,
+      "grad_norm": 0.14939643442630768,
+      "learning_rate": 0.0001,
+      "loss": 1.4816,
+      "step": 2854
+    },
+    {
+      "epoch": 0.1188881446751507,
+      "grad_norm": 0.15128619968891144,
+      "learning_rate": 0.0001,
+      "loss": 1.4966,
+      "step": 2855
+    },
+    {
+      "epoch": 0.11922304085733422,
+      "grad_norm": 0.1442454755306244,
+      "learning_rate": 0.0001,
+      "loss": 1.5416,
+      "step": 2856
+    },
+    {
+      "epoch": 0.11955793703951775,
+      "grad_norm": 0.14224649965763092,
+      "learning_rate": 0.0001,
+      "loss": 1.545,
+      "step": 2857
+    },
+    {
+      "epoch": 0.11989283322170127,
+      "grad_norm": 0.14138370752334595,
+      "learning_rate": 0.0001,
+      "loss": 1.5045,
+      "step": 2858
+    },
+    {
+      "epoch": 0.12022772940388479,
+      "grad_norm": 0.15062503516674042,
+      "learning_rate": 0.0001,
+      "loss": 1.4984,
+      "step": 2859
+    },
+    {
+      "epoch": 0.12056262558606833,
+      "grad_norm": 0.15108346939086914,
+      "learning_rate": 0.0001,
+      "loss": 1.5688,
+      "step": 2860
+    },
+    {
+      "epoch": 0.12089752176825184,
+      "grad_norm": 0.1456029713153839,
+      "learning_rate": 0.0001,
+      "loss": 1.4947,
+      "step": 2861
+    },
+    {
+      "epoch": 0.12123241795043536,
+      "grad_norm": 0.14550332725048065,
+      "learning_rate": 0.0001,
+      "loss": 1.3615,
+      "step": 2862
+    },
+    {
+      "epoch": 0.12156731413261888,
+      "grad_norm": 0.14609716832637787,
+      "learning_rate": 0.0001,
+      "loss": 1.4911,
+      "step": 2863
+    },
+    {
+      "epoch": 0.12190221031480242,
+      "grad_norm": 0.14730757474899292,
+      "learning_rate": 0.0001,
+      "loss": 1.4903,
+      "step": 2864
+    },
+    {
+      "epoch": 0.12223710649698594,
+      "grad_norm": 0.1497606337070465,
+      "learning_rate": 0.0001,
+      "loss": 1.4883,
+      "step": 2865
+    },
+    {
+      "epoch": 0.12257200267916946,
+      "grad_norm": 0.14975416660308838,
+      "learning_rate": 0.0001,
+      "loss": 1.5021,
+      "step": 2866
+    },
+    {
+      "epoch": 0.12290689886135298,
+      "grad_norm": 0.14952072501182556,
+      "learning_rate": 0.0001,
+      "loss": 1.5539,
+      "step": 2867
+    },
+    {
+      "epoch": 0.12324179504353651,
+      "grad_norm": 0.15092499554157257,
+      "learning_rate": 0.0001,
+      "loss": 1.459,
+      "step": 2868
+    },
+    {
+      "epoch": 0.12357669122572003,
+      "grad_norm": 0.14270053803920746,
+      "learning_rate": 0.0001,
+      "loss": 1.4523,
+      "step": 2869
+    },
+    {
+      "epoch": 0.12391158740790355,
+      "grad_norm": 0.14697770774364471,
+      "learning_rate": 0.0001,
+      "loss": 1.5208,
+      "step": 2870
+    },
+    {
+      "epoch": 0.12424648359008707,
+      "grad_norm": 0.1468277871608734,
+      "learning_rate": 0.0001,
+      "loss": 1.5413,
+      "step": 2871
+    },
+    {
+      "epoch": 0.1245813797722706,
+      "grad_norm": 0.1417912393808365,
+      "learning_rate": 0.0001,
+      "loss": 1.4422,
+      "step": 2872
+    },
+    {
+      "epoch": 0.12491627595445412,
+      "grad_norm": 0.1466628909111023,
+      "learning_rate": 0.0001,
+      "loss": 1.5536,
+      "step": 2873
+    },
+    {
+      "epoch": 0.12525117213663764,
+      "grad_norm": 0.15047575533390045,
+      "learning_rate": 0.0001,
+      "loss": 1.5134,
+      "step": 2874
+    },
+    {
+      "epoch": 0.12558606831882116,
+      "grad_norm": 0.14367663860321045,
+      "learning_rate": 0.0001,
+      "loss": 1.5137,
+      "step": 2875
+    },
+    {
+      "epoch": 0.12592096450100468,
+      "grad_norm": 0.14631038904190063,
+      "learning_rate": 0.0001,
+      "loss": 1.5526,
+      "step": 2876
+    },
+    {
+      "epoch": 0.1262558606831882,
+      "grad_norm": 0.1465481072664261,
+      "learning_rate": 0.0001,
+      "loss": 1.5348,
+      "step": 2877
+    },
+    {
+      "epoch": 0.12659075686537175,
+      "grad_norm": 0.141470804810524,
+      "learning_rate": 0.0001,
+      "loss": 1.4179,
+      "step": 2878
+    },
+    {
+      "epoch": 0.12692565304755526,
+      "grad_norm": 0.14299000799655914,
+      "learning_rate": 0.0001,
+      "loss": 1.447,
+      "step": 2879
+    },
+    {
+      "epoch": 0.12726054922973878,
+      "grad_norm": 0.13756218552589417,
+      "learning_rate": 0.0001,
+      "loss": 1.3867,
+      "step": 2880
+    },
+    {
+      "epoch": 0.1275954454119223,
+      "grad_norm": 0.141182541847229,
+      "learning_rate": 0.0001,
+      "loss": 1.4574,
+      "step": 2881
+    },
+    {
+      "epoch": 0.12793034159410582,
+      "grad_norm": 0.1464800387620926,
+      "learning_rate": 0.0001,
+      "loss": 1.4442,
+      "step": 2882
+    },
+    {
+      "epoch": 0.12826523777628934,
+      "grad_norm": 0.14528337121009827,
+      "learning_rate": 0.0001,
+      "loss": 1.4991,
+      "step": 2883
+    },
+    {
+      "epoch": 0.12860013395847286,
+      "grad_norm": 0.14793065190315247,
+      "learning_rate": 0.0001,
+      "loss": 1.5269,
+      "step": 2884
+    },
+    {
+      "epoch": 0.1289350301406564,
+      "grad_norm": 0.1435883641242981,
+      "learning_rate": 0.0001,
+      "loss": 1.4521,
+      "step": 2885
+    },
+    {
+      "epoch": 0.12926992632283993,
+      "grad_norm": 0.14673016965389252,
+      "learning_rate": 0.0001,
+      "loss": 1.5129,
+      "step": 2886
+    },
+    {
+      "epoch": 0.12960482250502345,
+      "grad_norm": 0.13962426781654358,
+      "learning_rate": 0.0001,
+      "loss": 1.467,
+      "step": 2887
+    },
+    {
+      "epoch": 0.12993971868720697,
+      "grad_norm": 0.1506863832473755,
+      "learning_rate": 0.0001,
+      "loss": 1.5761,
+      "step": 2888
+    },
+    {
+      "epoch": 0.1302746148693905,
+      "grad_norm": 0.14275865256786346,
+      "learning_rate": 0.0001,
+      "loss": 1.4139,
+      "step": 2889
+    },
+    {
+      "epoch": 0.130609511051574,
+      "grad_norm": 0.14221537113189697,
+      "learning_rate": 0.0001,
+      "loss": 1.4269,
+      "step": 2890
+    },
+    {
+      "epoch": 0.13094440723375753,
+      "grad_norm": 0.1372966170310974,
+      "learning_rate": 0.0001,
+      "loss": 1.4217,
+      "step": 2891
+    },
+    {
+      "epoch": 0.13127930341594105,
+      "grad_norm": 0.1523875892162323,
+      "learning_rate": 0.0001,
+      "loss": 1.528,
+      "step": 2892
+    },
+    {
+      "epoch": 0.1316141995981246,
+      "grad_norm": 0.13655850291252136,
+      "learning_rate": 0.0001,
+      "loss": 1.4749,
+      "step": 2893
+    },
+    {
+      "epoch": 0.1319490957803081,
+      "grad_norm": 0.14285436272621155,
+      "learning_rate": 0.0001,
+      "loss": 1.416,
+      "step": 2894
+    },
+    {
+      "epoch": 0.13228399196249163,
+      "grad_norm": 0.1540876179933548,
+      "learning_rate": 0.0001,
+      "loss": 1.604,
+      "step": 2895
+    },
+    {
+      "epoch": 0.13261888814467515,
+      "grad_norm": 0.14926448464393616,
+      "learning_rate": 0.0001,
+      "loss": 1.5095,
+      "step": 2896
+    },
+    {
+      "epoch": 0.13295378432685867,
+      "grad_norm": 0.15166886150836945,
+      "learning_rate": 0.0001,
+      "loss": 1.5766,
+      "step": 2897
+    },
+    {
+      "epoch": 0.1332886805090422,
+      "grad_norm": 0.15036708116531372,
+      "learning_rate": 0.0001,
+      "loss": 1.4603,
+      "step": 2898
+    },
+    {
+      "epoch": 0.1336235766912257,
+      "grad_norm": 0.14928917586803436,
+      "learning_rate": 0.0001,
+      "loss": 1.5102,
+      "step": 2899
+    },
+    {
+      "epoch": 0.13395847287340926,
+      "grad_norm": 0.14783331751823425,
+      "learning_rate": 0.0001,
+      "loss": 1.4389,
+      "step": 2900
+    },
+    {
+      "epoch": 0.13429336905559278,
+      "grad_norm": 0.1541386842727661,
+      "learning_rate": 0.0001,
+      "loss": 1.643,
+      "step": 2901
+    },
+    {
+      "epoch": 0.1346282652377763,
+      "grad_norm": 0.15611632168293,
+      "learning_rate": 0.0001,
+      "loss": 1.5295,
+      "step": 2902
+    },
+    {
+      "epoch": 0.13496316141995982,
+      "grad_norm": 0.1471196413040161,
+      "learning_rate": 0.0001,
+      "loss": 1.4305,
+      "step": 2903
+    },
+    {
+      "epoch": 0.13529805760214333,
+      "grad_norm": 0.15436099469661713,
+      "learning_rate": 0.0001,
+      "loss": 1.5258,
+      "step": 2904
+    },
+    {
+      "epoch": 0.13563295378432685,
+      "grad_norm": 0.14856241643428802,
+      "learning_rate": 0.0001,
+      "loss": 1.432,
+      "step": 2905
+    },
+    {
+      "epoch": 0.13596784996651037,
+      "grad_norm": 0.1435495764017105,
+      "learning_rate": 0.0001,
+      "loss": 1.4063,
+      "step": 2906
+    },
+    {
+      "epoch": 0.1363027461486939,
+      "grad_norm": 0.14317458868026733,
+      "learning_rate": 0.0001,
+      "loss": 1.4238,
+      "step": 2907
+    },
+    {
+      "epoch": 0.13663764233087744,
+      "grad_norm": 0.14765316247940063,
+      "learning_rate": 0.0001,
+      "loss": 1.5453,
+      "step": 2908
+    },
+    {
+      "epoch": 0.13697253851306096,
+      "grad_norm": 0.15258818864822388,
+      "learning_rate": 0.0001,
+      "loss": 1.559,
+      "step": 2909
+    },
+    {
+      "epoch": 0.13730743469524448,
+      "grad_norm": 0.14648139476776123,
+      "learning_rate": 0.0001,
+      "loss": 1.4419,
+      "step": 2910
+    },
+    {
+      "epoch": 0.137642330877428,
+      "grad_norm": 0.14375437796115875,
+      "learning_rate": 0.0001,
+      "loss": 1.4301,
+      "step": 2911
+    },
+    {
+      "epoch": 0.13797722705961152,
+      "grad_norm": 0.1457594335079193,
+      "learning_rate": 0.0001,
+      "loss": 1.4593,
+      "step": 2912
+    },
+    {
+      "epoch": 0.13831212324179504,
+      "grad_norm": 0.14920419454574585,
+      "learning_rate": 0.0001,
+      "loss": 1.5514,
+      "step": 2913
+    },
+    {
+      "epoch": 0.13864701942397856,
+      "grad_norm": 0.1460302621126175,
+      "learning_rate": 0.0001,
+      "loss": 1.47,
+      "step": 2914
+    },
+    {
+      "epoch": 0.13898191560616208,
+      "grad_norm": 0.1491982638835907,
+      "learning_rate": 0.0001,
+      "loss": 1.4894,
+      "step": 2915
+    },
+    {
+      "epoch": 0.13931681178834562,
+      "grad_norm": 0.14638325572013855,
+      "learning_rate": 0.0001,
+      "loss": 1.4824,
+      "step": 2916
+    },
+    {
+      "epoch": 0.13965170797052914,
+      "grad_norm": 0.14499424397945404,
+      "learning_rate": 0.0001,
+      "loss": 1.4292,
+      "step": 2917
+    },
+    {
+      "epoch": 0.13998660415271266,
+      "grad_norm": 0.14576716721057892,
+      "learning_rate": 0.0001,
+      "loss": 1.5192,
+      "step": 2918
+    },
+    {
+      "epoch": 0.14032150033489618,
+      "grad_norm": 0.1488197296857834,
+      "learning_rate": 0.0001,
+      "loss": 1.442,
+      "step": 2919
+    },
+    {
+      "epoch": 0.1406563965170797,
+      "grad_norm": 0.14574909210205078,
+      "learning_rate": 0.0001,
+      "loss": 1.4224,
+      "step": 2920
+    },
+    {
+      "epoch": 0.14099129269926322,
+      "grad_norm": 0.15139609575271606,
+      "learning_rate": 0.0001,
+      "loss": 1.5193,
+      "step": 2921
+    },
+    {
+      "epoch": 0.14132618888144674,
+      "grad_norm": 0.15167635679244995,
+      "learning_rate": 0.0001,
+      "loss": 1.5037,
+      "step": 2922
+    },
+    {
+      "epoch": 0.1416610850636303,
+      "grad_norm": 0.14516854286193848,
+      "learning_rate": 0.0001,
+      "loss": 1.4547,
+      "step": 2923
+    },
+    {
+      "epoch": 0.1419959812458138,
+      "grad_norm": 0.14109300076961517,
+      "learning_rate": 0.0001,
+      "loss": 1.4066,
+      "step": 2924
+    },
+    {
+      "epoch": 0.14233087742799733,
+      "grad_norm": 0.1450345367193222,
+      "learning_rate": 0.0001,
+      "loss": 1.512,
+      "step": 2925
+    },
+    {
+      "epoch": 0.14266577361018085,
+      "grad_norm": 0.14865948259830475,
+      "learning_rate": 0.0001,
+      "loss": 1.4758,
+      "step": 2926
+    },
+    {
+      "epoch": 0.14300066979236437,
+      "grad_norm": 0.1497637927532196,
+      "learning_rate": 0.0001,
+      "loss": 1.4643,
+      "step": 2927
+    },
+    {
+      "epoch": 0.14333556597454788,
+      "grad_norm": 0.1432199478149414,
+      "learning_rate": 0.0001,
+      "loss": 1.467,
+      "step": 2928
+    },
+    {
+      "epoch": 0.1436704621567314,
+      "grad_norm": 0.15253181755542755,
+      "learning_rate": 0.0001,
+      "loss": 1.5088,
+      "step": 2929
+    },
+    {
+      "epoch": 0.14400535833891492,
+      "grad_norm": 0.1489313691854477,
+      "learning_rate": 0.0001,
+      "loss": 1.475,
+      "step": 2930
+    },
+    {
+      "epoch": 0.14434025452109847,
+      "grad_norm": 0.15359234809875488,
+      "learning_rate": 0.0001,
+      "loss": 1.5527,
+      "step": 2931
+    },
+    {
+      "epoch": 0.144675150703282,
+      "grad_norm": 0.14580143988132477,
+      "learning_rate": 0.0001,
+      "loss": 1.4946,
+      "step": 2932
+    },
+    {
+      "epoch": 0.1450100468854655,
+      "grad_norm": 0.14856304228305817,
+      "learning_rate": 0.0001,
+      "loss": 1.4547,
+      "step": 2933
+    },
+    {
+      "epoch": 0.14534494306764903,
+      "grad_norm": 0.14780816435813904,
+      "learning_rate": 0.0001,
+      "loss": 1.4839,
+      "step": 2934
+    },
+    {
+      "epoch": 0.14567983924983255,
+      "grad_norm": 0.15428078174591064,
+      "learning_rate": 0.0001,
+      "loss": 1.5118,
+      "step": 2935
+    },
+    {
+      "epoch": 0.14601473543201607,
+      "grad_norm": 0.13885167241096497,
+      "learning_rate": 0.0001,
+      "loss": 1.4458,
+      "step": 2936
+    },
+    {
+      "epoch": 0.1463496316141996,
+      "grad_norm": 0.14830608665943146,
+      "learning_rate": 0.0001,
+      "loss": 1.441,
+      "step": 2937
+    },
+    {
+      "epoch": 0.14668452779638314,
+      "grad_norm": 0.14985600113868713,
+      "learning_rate": 0.0001,
+      "loss": 1.4694,
+      "step": 2938
+    },
+    {
+      "epoch": 0.14701942397856665,
+      "grad_norm": 0.14177678525447845,
+      "learning_rate": 0.0001,
+      "loss": 1.5314,
+      "step": 2939
+    },
+    {
+      "epoch": 0.14735432016075017,
+      "grad_norm": 0.1454205960035324,
+      "learning_rate": 0.0001,
+      "loss": 1.4811,
+      "step": 2940
+    },
+    {
+      "epoch": 0.1476892163429337,
+      "grad_norm": 0.14319349825382233,
+      "learning_rate": 0.0001,
+      "loss": 1.4508,
+      "step": 2941
+    },
+    {
+      "epoch": 0.1480241125251172,
+      "grad_norm": 0.14416146278381348,
+      "learning_rate": 0.0001,
+      "loss": 1.4633,
+      "step": 2942
+    },
+    {
+      "epoch": 0.14835900870730073,
+      "grad_norm": 0.14271238446235657,
+      "learning_rate": 0.0001,
+      "loss": 1.4124,
+      "step": 2943
+    },
+    {
+      "epoch": 0.14869390488948425,
+      "grad_norm": 0.149638369679451,
+      "learning_rate": 0.0001,
+      "loss": 1.5879,
+      "step": 2944
+    },
+    {
+      "epoch": 0.14902880107166777,
+      "grad_norm": 0.15497125685214996,
+      "learning_rate": 0.0001,
+      "loss": 1.5543,
+      "step": 2945
+    },
+    {
+      "epoch": 0.14936369725385132,
+      "grad_norm": 0.14638566970825195,
+      "learning_rate": 0.0001,
+      "loss": 1.52,
+      "step": 2946
+    },
+    {
+      "epoch": 0.14969859343603484,
+      "grad_norm": 0.13843445479869843,
+      "learning_rate": 0.0001,
+      "loss": 1.4464,
+      "step": 2947
+    },
+    {
+      "epoch": 0.15003348961821836,
+      "grad_norm": 0.1452428102493286,
+      "learning_rate": 0.0001,
+      "loss": 1.5037,
+      "step": 2948
+    },
+    {
+      "epoch": 0.15036838580040188,
+      "grad_norm": 0.1429130882024765,
+      "learning_rate": 0.0001,
+      "loss": 1.4431,
+      "step": 2949
+    },
+    {
+      "epoch": 0.1507032819825854,
+      "grad_norm": 0.14541609585285187,
+      "learning_rate": 0.0001,
+      "loss": 1.502,
+      "step": 2950
+    },
+    {
+      "epoch": 0.15103817816476892,
+      "grad_norm": 0.1436072587966919,
+      "learning_rate": 0.0001,
+      "loss": 1.5754,
+      "step": 2951
+    },
+    {
+      "epoch": 0.15137307434695244,
+      "grad_norm": 0.1478150635957718,
+      "learning_rate": 0.0001,
+      "loss": 1.4515,
+      "step": 2952
+    },
+    {
+      "epoch": 0.15170797052913595,
+      "grad_norm": 0.14313754439353943,
+      "learning_rate": 0.0001,
+      "loss": 1.4284,
+      "step": 2953
+    },
+    {
+      "epoch": 0.1520428667113195,
+      "grad_norm": 0.14626973867416382,
+      "learning_rate": 0.0001,
+      "loss": 1.4427,
+      "step": 2954
+    },
+    {
+      "epoch": 0.15237776289350302,
+      "grad_norm": 0.14305603504180908,
+      "learning_rate": 0.0001,
+      "loss": 1.4067,
+      "step": 2955
+    },
+    {
+      "epoch": 0.15271265907568654,
+      "grad_norm": 0.15285871922969818,
+      "learning_rate": 0.0001,
+      "loss": 1.4773,
+      "step": 2956
+    },
+    {
+      "epoch": 0.15304755525787006,
+      "grad_norm": 0.14073546230793,
+      "learning_rate": 0.0001,
+      "loss": 1.394,
+      "step": 2957
+    },
+    {
+      "epoch": 0.15338245144005358,
+      "grad_norm": 0.15410828590393066,
+      "learning_rate": 0.0001,
+      "loss": 1.5485,
+      "step": 2958
+    },
+    {
+      "epoch": 0.1537173476222371,
+      "grad_norm": 0.162978857755661,
+      "learning_rate": 0.0001,
+      "loss": 1.601,
+      "step": 2959
+    },
+    {
+      "epoch": 0.15405224380442062,
+      "grad_norm": 0.14464987814426422,
+      "learning_rate": 0.0001,
+      "loss": 1.5635,
+      "step": 2960
+    },
+    {
+      "epoch": 0.15438713998660417,
+      "grad_norm": 0.1483234167098999,
+      "learning_rate": 0.0001,
+      "loss": 1.481,
+      "step": 2961
+    },
+    {
+      "epoch": 0.15472203616878769,
+      "grad_norm": 0.151525616645813,
+      "learning_rate": 0.0001,
+      "loss": 1.4573,
+      "step": 2962
+    },
+    {
+      "epoch": 0.1550569323509712,
+      "grad_norm": 0.14589788019657135,
+      "learning_rate": 0.0001,
+      "loss": 1.5513,
+      "step": 2963
+    },
+    {
+      "epoch": 0.15539182853315472,
+      "grad_norm": 0.15947841107845306,
+      "learning_rate": 0.0001,
+      "loss": 1.5102,
+      "step": 2964
+    },
+    {
+      "epoch": 0.15572672471533824,
+      "grad_norm": 0.14565446972846985,
+      "learning_rate": 0.0001,
+      "loss": 1.4366,
+      "step": 2965
+    },
+    {
+      "epoch": 0.15606162089752176,
+      "grad_norm": 0.1432366818189621,
+      "learning_rate": 0.0001,
+      "loss": 1.5625,
+      "step": 2966
+    },
+    {
+      "epoch": 0.15639651707970528,
+      "grad_norm": 0.15333931148052216,
+      "learning_rate": 0.0001,
+      "loss": 1.458,
+      "step": 2967
+    },
+    {
+      "epoch": 0.1567314132618888,
+      "grad_norm": 0.15091878175735474,
+      "learning_rate": 0.0001,
+      "loss": 1.4619,
+      "step": 2968
+    },
+    {
+      "epoch": 0.15706630944407235,
+      "grad_norm": 0.1437712013721466,
+      "learning_rate": 0.0001,
+      "loss": 1.5207,
+      "step": 2969
+    },
+    {
+      "epoch": 0.15740120562625587,
+      "grad_norm": 0.16013203561306,
+      "learning_rate": 0.0001,
+      "loss": 1.5625,
+      "step": 2970
+    },
+    {
+      "epoch": 0.1577361018084394,
+      "grad_norm": 0.1535797417163849,
+      "learning_rate": 0.0001,
+      "loss": 1.5372,
+      "step": 2971
+    },
+    {
+      "epoch": 0.1580709979906229,
+      "grad_norm": 0.14695370197296143,
+      "learning_rate": 0.0001,
+      "loss": 1.492,
+      "step": 2972
+    },
+    {
+      "epoch": 0.15840589417280643,
+      "grad_norm": 0.15556497871875763,
+      "learning_rate": 0.0001,
+      "loss": 1.5317,
+      "step": 2973
+    },
+    {
+      "epoch": 0.15874079035498995,
+      "grad_norm": 0.15451493859291077,
+      "learning_rate": 0.0001,
+      "loss": 1.5248,
+      "step": 2974
+    },
+    {
+      "epoch": 0.15907568653717347,
+      "grad_norm": 0.14186540246009827,
+      "learning_rate": 0.0001,
+      "loss": 1.4655,
+      "step": 2975
+    },
+    {
+      "epoch": 0.15941058271935699,
+      "grad_norm": 0.15657803416252136,
+      "learning_rate": 0.0001,
+      "loss": 1.4686,
+      "step": 2976
+    },
+    {
+      "epoch": 0.15974547890154053,
+      "grad_norm": 0.15531529486179352,
+      "learning_rate": 0.0001,
+      "loss": 1.512,
+      "step": 2977
+    },
+    {
+      "epoch": 0.16008037508372405,
+      "grad_norm": 0.14492064714431763,
+      "learning_rate": 0.0001,
+      "loss": 1.4434,
+      "step": 2978
+    },
+    {
+      "epoch": 0.16041527126590757,
+      "grad_norm": 0.1516708880662918,
+      "learning_rate": 0.0001,
+      "loss": 1.4656,
+      "step": 2979
+    },
+    {
+      "epoch": 0.1607501674480911,
+      "grad_norm": 0.14958037436008453,
+      "learning_rate": 0.0001,
+      "loss": 1.4626,
+      "step": 2980
+    },
+    {
+      "epoch": 0.1610850636302746,
+      "grad_norm": 0.1415179967880249,
+      "learning_rate": 0.0001,
+      "loss": 1.4767,
+      "step": 2981
+    },
+    {
+      "epoch": 0.16141995981245813,
+      "grad_norm": 0.1595795750617981,
+      "learning_rate": 0.0001,
+      "loss": 1.5476,
+      "step": 2982
+    },
+    {
+      "epoch": 0.16175485599464165,
+      "grad_norm": 0.1559695601463318,
+      "learning_rate": 0.0001,
+      "loss": 1.5846,
+      "step": 2983
+    },
+    {
+      "epoch": 0.1620897521768252,
+      "grad_norm": 0.14177894592285156,
+      "learning_rate": 0.0001,
+      "loss": 1.3714,
+      "step": 2984
+    },
+    {
+      "epoch": 0.16242464835900872,
+      "grad_norm": 0.15120206773281097,
+      "learning_rate": 0.0001,
+      "loss": 1.4896,
+      "step": 2985
+    },
+    {
+      "epoch": 0.16275954454119224,
+      "grad_norm": 0.1491135209798813,
+      "learning_rate": 0.0001,
+      "loss": 1.4239,
+      "step": 2986
+    },
+    {
+      "epoch": 0.16309444072337576,
+      "grad_norm": 0.15394911170005798,
+      "learning_rate": 0.0001,
+      "loss": 1.5397,
+      "step": 2987
+    },
+    {
+      "epoch": 0.16342933690555927,
+      "grad_norm": 0.14349506795406342,
+      "learning_rate": 0.0001,
+      "loss": 1.4447,
+      "step": 2988
+    },
+    {
+      "epoch": 0.1637642330877428,
+      "grad_norm": 0.15076670050621033,
+      "learning_rate": 0.0001,
+      "loss": 1.4206,
+      "step": 2989
+    },
+    {
+      "epoch": 0.1640991292699263,
+      "grad_norm": 0.14125433564186096,
+      "learning_rate": 0.0001,
+      "loss": 1.426,
+      "step": 2990
+    },
+    {
+      "epoch": 0.16443402545210983,
+      "grad_norm": 0.15920014679431915,
+      "learning_rate": 0.0001,
+      "loss": 1.5571,
+      "step": 2991
+    },
+    {
+      "epoch": 0.16476892163429338,
+      "grad_norm": 0.1530846655368805,
+      "learning_rate": 0.0001,
+      "loss": 1.4773,
+      "step": 2992
+    },
+    {
+      "epoch": 0.1651038178164769,
+      "grad_norm": 0.14886631071567535,
+      "learning_rate": 0.0001,
+      "loss": 1.5835,
+      "step": 2993
+    },
+    {
+      "epoch": 0.16543871399866042,
+      "grad_norm": 0.1467156857252121,
+      "learning_rate": 0.0001,
+      "loss": 1.5853,
+      "step": 2994
+    },
+    {
+      "epoch": 0.16577361018084394,
+      "grad_norm": 0.14700840413570404,
+      "learning_rate": 0.0001,
+      "loss": 1.5046,
+      "step": 2995
+    },
+    {
+      "epoch": 0.16610850636302746,
+      "grad_norm": 0.1469178944826126,
+      "learning_rate": 0.0001,
+      "loss": 1.5363,
+      "step": 2996
+    },
+    {
+      "epoch": 0.16644340254521098,
+      "grad_norm": 0.15248429775238037,
+      "learning_rate": 0.0001,
+      "loss": 1.4008,
+      "step": 2997
+    },
+    {
+      "epoch": 0.1667782987273945,
+      "grad_norm": 0.15299245715141296,
+      "learning_rate": 0.0001,
+      "loss": 1.5311,
+      "step": 2998
+    },
+    {
+      "epoch": 0.16711319490957804,
+      "grad_norm": 0.1501973271369934,
+      "learning_rate": 0.0001,
+      "loss": 1.481,
+      "step": 2999
+    },
+    {
+      "epoch": 0.16744809109176156,
+      "grad_norm": 0.14169913530349731,
+      "learning_rate": 0.0001,
+      "loss": 1.3979,
+      "step": 3000
+    },
+    {
+      "epoch": 0.16778298727394508,
+      "grad_norm": 0.14365142583847046,
+      "learning_rate": 0.0001,
+      "loss": 1.492,
+      "step": 3001
+    },
+    {
+      "epoch": 0.1681178834561286,
+      "grad_norm": 0.1459331214427948,
+      "learning_rate": 0.0001,
+      "loss": 1.4924,
+      "step": 3002
+    },
+    {
+      "epoch": 0.16845277963831212,
+      "grad_norm": 0.1534721553325653,
+      "learning_rate": 0.0001,
+      "loss": 1.5352,
+      "step": 3003
+    },
+    {
+      "epoch": 0.16878767582049564,
+      "grad_norm": 0.14717139303684235,
+      "learning_rate": 0.0001,
+      "loss": 1.5159,
+      "step": 3004
+    },
+    {
+      "epoch": 0.16912257200267916,
+      "grad_norm": 0.14475423097610474,
+      "learning_rate": 0.0001,
+      "loss": 1.4526,
+      "step": 3005
+    },
+    {
+      "epoch": 0.16945746818486268,
+      "grad_norm": 0.14285942912101746,
+      "learning_rate": 0.0001,
+      "loss": 1.4722,
+      "step": 3006
+    },
+    {
+      "epoch": 0.16979236436704623,
+      "grad_norm": 0.14912879467010498,
+      "learning_rate": 0.0001,
+      "loss": 1.4364,
+      "step": 3007
+    },
+    {
+      "epoch": 0.17012726054922975,
+      "grad_norm": 0.15310518443584442,
+      "learning_rate": 0.0001,
+      "loss": 1.541,
+      "step": 3008
+    },
+    {
+      "epoch": 0.17046215673141327,
+      "grad_norm": 0.14872018992900848,
+      "learning_rate": 0.0001,
+      "loss": 1.5321,
+      "step": 3009
+    },
+    {
+      "epoch": 0.1707970529135968,
+      "grad_norm": 0.1429416537284851,
+      "learning_rate": 0.0001,
+      "loss": 1.4555,
+      "step": 3010
+    },
+    {
+      "epoch": 0.1711319490957803,
+      "grad_norm": 0.14461266994476318,
+      "learning_rate": 0.0001,
+      "loss": 1.4324,
+      "step": 3011
+    },
+    {
+      "epoch": 0.17146684527796383,
+      "grad_norm": 0.15209822356700897,
+      "learning_rate": 0.0001,
+      "loss": 1.5429,
+      "step": 3012
+    },
+    {
+      "epoch": 0.17180174146014734,
+      "grad_norm": 0.148372620344162,
+      "learning_rate": 0.0001,
+      "loss": 1.4373,
+      "step": 3013
+    },
+    {
+      "epoch": 0.17213663764233086,
+      "grad_norm": 0.1498633325099945,
+      "learning_rate": 0.0001,
+      "loss": 1.5929,
+      "step": 3014
+    },
+    {
+      "epoch": 0.1724715338245144,
+      "grad_norm": 0.15261657536029816,
+      "learning_rate": 0.0001,
+      "loss": 1.5334,
+      "step": 3015
+    },
+    {
+      "epoch": 0.17280643000669793,
+      "grad_norm": 0.14222516119480133,
+      "learning_rate": 0.0001,
+      "loss": 1.4574,
+      "step": 3016
+    },
+    {
+      "epoch": 0.17314132618888145,
+      "grad_norm": 0.14974059164524078,
+      "learning_rate": 0.0001,
+      "loss": 1.4512,
+      "step": 3017
+    },
+    {
+      "epoch": 0.17347622237106497,
+      "grad_norm": 0.14610859751701355,
+      "learning_rate": 0.0001,
+      "loss": 1.4675,
+      "step": 3018
+    },
+    {
+      "epoch": 0.1738111185532485,
+      "grad_norm": 0.14841997623443604,
+      "learning_rate": 0.0001,
+      "loss": 1.399,
+      "step": 3019
+    },
+    {
+      "epoch": 0.174146014735432,
+      "grad_norm": 0.14367492496967316,
+      "learning_rate": 0.0001,
+      "loss": 1.4205,
+      "step": 3020
+    },
+    {
+      "epoch": 0.17448091091761553,
+      "grad_norm": 0.14827755093574524,
+      "learning_rate": 0.0001,
+      "loss": 1.4087,
+      "step": 3021
+    },
+    {
+      "epoch": 0.17481580709979908,
+      "grad_norm": 0.1429951936006546,
+      "learning_rate": 0.0001,
+      "loss": 1.4383,
+      "step": 3022
+    },
+    {
+      "epoch": 0.1751507032819826,
+      "grad_norm": 0.14675314724445343,
+      "learning_rate": 0.0001,
+      "loss": 1.4825,
+      "step": 3023
+    },
+    {
+      "epoch": 0.17548559946416611,
+      "grad_norm": 0.1578107476234436,
+      "learning_rate": 0.0001,
+      "loss": 1.5891,
+      "step": 3024
+    },
+    {
+      "epoch": 0.17582049564634963,
+      "grad_norm": 0.15326331555843353,
+      "learning_rate": 0.0001,
+      "loss": 1.5382,
+      "step": 3025
+    },
+    {
+      "epoch": 0.17615539182853315,
+      "grad_norm": 0.15587475895881653,
+      "learning_rate": 0.0001,
+      "loss": 1.4613,
+      "step": 3026
+    },
+    {
+      "epoch": 0.17649028801071667,
+      "grad_norm": 0.14984256029129028,
+      "learning_rate": 0.0001,
+      "loss": 1.4881,
+      "step": 3027
+    },
+    {
+      "epoch": 0.1768251841929002,
+      "grad_norm": 0.15645815432071686,
+      "learning_rate": 0.0001,
+      "loss": 1.5756,
+      "step": 3028
+    },
+    {
+      "epoch": 0.1771600803750837,
+      "grad_norm": 0.14840419590473175,
+      "learning_rate": 0.0001,
+      "loss": 1.5134,
+      "step": 3029
+    },
+    {
+      "epoch": 0.17749497655726726,
+      "grad_norm": 0.1488848477602005,
+      "learning_rate": 0.0001,
+      "loss": 1.3932,
+      "step": 3030
+    },
+    {
+      "epoch": 0.17782987273945078,
+      "grad_norm": 0.14206485450267792,
+      "learning_rate": 0.0001,
+      "loss": 1.4531,
+      "step": 3031
+    },
+    {
+      "epoch": 0.1781647689216343,
+      "grad_norm": 0.14872042834758759,
+      "learning_rate": 0.0001,
+      "loss": 1.4467,
+      "step": 3032
+    },
+    {
+      "epoch": 0.17849966510381782,
+      "grad_norm": 0.1473948210477829,
+      "learning_rate": 0.0001,
+      "loss": 1.5119,
+      "step": 3033
+    },
+    {
+      "epoch": 0.17883456128600134,
+      "grad_norm": 0.14164240658283234,
+      "learning_rate": 0.0001,
+      "loss": 1.4372,
+      "step": 3034
+    },
+    {
+      "epoch": 0.17916945746818486,
+      "grad_norm": 0.15156961977481842,
+      "learning_rate": 0.0001,
+      "loss": 1.5728,
+      "step": 3035
+    },
+    {
+      "epoch": 0.17950435365036838,
+      "grad_norm": 0.14617635309696198,
+      "learning_rate": 0.0001,
+      "loss": 1.4933,
+      "step": 3036
+    },
+    {
+      "epoch": 0.1798392498325519,
+      "grad_norm": 0.14830727875232697,
+      "learning_rate": 0.0001,
+      "loss": 1.5332,
+      "step": 3037
+    },
+    {
+      "epoch": 0.18017414601473544,
+      "grad_norm": 0.1488112509250641,
+      "learning_rate": 0.0001,
+      "loss": 1.4318,
+      "step": 3038
+    },
+    {
+      "epoch": 0.18050904219691896,
+      "grad_norm": 0.14796082675457,
+      "learning_rate": 0.0001,
+      "loss": 1.6183,
+      "step": 3039
+    },
+    {
+      "epoch": 0.18084393837910248,
+      "grad_norm": 0.15212929248809814,
+      "learning_rate": 0.0001,
+      "loss": 1.554,
+      "step": 3040
+    },
+    {
+      "epoch": 0.181178834561286,
+      "grad_norm": 0.1485069841146469,
+      "learning_rate": 0.0001,
+      "loss": 1.4418,
+      "step": 3041
+    },
+    {
+      "epoch": 0.18151373074346952,
+      "grad_norm": 0.1428901106119156,
+      "learning_rate": 0.0001,
+      "loss": 1.4409,
+      "step": 3042
+    },
+    {
+      "epoch": 0.18184862692565304,
+      "grad_norm": 0.1394181102514267,
+      "learning_rate": 0.0001,
+      "loss": 1.4253,
+      "step": 3043
+    },
+    {
+      "epoch": 0.18218352310783656,
+      "grad_norm": 0.15848679840564728,
+      "learning_rate": 0.0001,
+      "loss": 1.4266,
+      "step": 3044
+    },
+    {
+      "epoch": 0.1825184192900201,
+      "grad_norm": 0.15975135564804077,
+      "learning_rate": 0.0001,
+      "loss": 1.5339,
+      "step": 3045
+    },
+    {
+      "epoch": 0.18285331547220363,
+      "grad_norm": 0.1499025970697403,
+      "learning_rate": 0.0001,
+      "loss": 1.4731,
+      "step": 3046
+    },
+    {
+      "epoch": 0.18318821165438715,
+      "grad_norm": 0.1490570306777954,
+      "learning_rate": 0.0001,
+      "loss": 1.5787,
+      "step": 3047
+    },
+    {
+      "epoch": 0.18352310783657066,
+      "grad_norm": 0.16007208824157715,
+      "learning_rate": 0.0001,
+      "loss": 1.5333,
+      "step": 3048
+    },
+    {
+      "epoch": 0.18385800401875418,
+      "grad_norm": 0.14733049273490906,
+      "learning_rate": 0.0001,
+      "loss": 1.4289,
+      "step": 3049
+    },
+    {
+      "epoch": 0.1841929002009377,
+      "grad_norm": 0.14483436942100525,
+      "learning_rate": 0.0001,
+      "loss": 1.4398,
+      "step": 3050
+    },
+    {
+      "epoch": 0.18452779638312122,
+      "grad_norm": 0.1574944406747818,
+      "learning_rate": 0.0001,
+      "loss": 1.5116,
+      "step": 3051
+    },
+    {
+      "epoch": 0.18486269256530474,
+      "grad_norm": 0.14573010802268982,
+      "learning_rate": 0.0001,
+      "loss": 1.4275,
+      "step": 3052
+    },
+    {
+      "epoch": 0.1851975887474883,
+      "grad_norm": 0.14518269896507263,
+      "learning_rate": 0.0001,
+      "loss": 1.4883,
+      "step": 3053
+    },
+    {
+      "epoch": 0.1855324849296718,
+      "grad_norm": 0.14543338119983673,
+      "learning_rate": 0.0001,
+      "loss": 1.442,
+      "step": 3054
+    },
+    {
+      "epoch": 0.18586738111185533,
+      "grad_norm": 0.157647043466568,
+      "learning_rate": 0.0001,
+      "loss": 1.3932,
+      "step": 3055
+    },
+    {
+      "epoch": 0.18620227729403885,
+      "grad_norm": 0.15597252547740936,
+      "learning_rate": 0.0001,
+      "loss": 1.5283,
+      "step": 3056
+    },
+    {
+      "epoch": 0.18653717347622237,
+      "grad_norm": 0.14394451677799225,
+      "learning_rate": 0.0001,
+      "loss": 1.4146,
+      "step": 3057
+    },
+    {
+      "epoch": 0.1868720696584059,
+      "grad_norm": 0.14798322319984436,
+      "learning_rate": 0.0001,
+      "loss": 1.4653,
+      "step": 3058
+    },
+    {
+      "epoch": 0.1872069658405894,
+      "grad_norm": 0.1463840901851654,
+      "learning_rate": 0.0001,
+      "loss": 1.5469,
+      "step": 3059
+    },
+    {
+      "epoch": 0.18754186202277295,
+      "grad_norm": 0.14234453439712524,
+      "learning_rate": 0.0001,
+      "loss": 1.4405,
+      "step": 3060
+    },
+    {
+      "epoch": 0.18787675820495647,
+      "grad_norm": 0.1502765715122223,
+      "learning_rate": 0.0001,
+      "loss": 1.6293,
+      "step": 3061
+    },
+    {
+      "epoch": 0.18821165438714,
+      "grad_norm": 0.14024700224399567,
+      "learning_rate": 0.0001,
+      "loss": 1.4284,
+      "step": 3062
+    },
+    {
+      "epoch": 0.1885465505693235,
+      "grad_norm": 0.1412835270166397,
+      "learning_rate": 0.0001,
+      "loss": 1.4675,
+      "step": 3063
+    },
+    {
+      "epoch": 0.18888144675150703,
+      "grad_norm": 0.1488562822341919,
+      "learning_rate": 0.0001,
+      "loss": 1.5071,
+      "step": 3064
+    },
+    {
+      "epoch": 0.18921634293369055,
+      "grad_norm": 0.14808917045593262,
+      "learning_rate": 0.0001,
+      "loss": 1.4587,
+      "step": 3065
+    },
+    {
+      "epoch": 0.18955123911587407,
+      "grad_norm": 0.14762601256370544,
+      "learning_rate": 0.0001,
+      "loss": 1.5168,
+      "step": 3066
+    },
+    {
+      "epoch": 0.1898861352980576,
+      "grad_norm": 0.14939825236797333,
+      "learning_rate": 0.0001,
+      "loss": 1.5294,
+      "step": 3067
+    },
+    {
+      "epoch": 0.19022103148024114,
+      "grad_norm": 0.1474798023700714,
+      "learning_rate": 0.0001,
+      "loss": 1.4373,
+      "step": 3068
+    },
+    {
+      "epoch": 0.19055592766242466,
+      "grad_norm": 0.1497553139925003,
+      "learning_rate": 0.0001,
+      "loss": 1.4871,
+      "step": 3069
+    },
+    {
+      "epoch": 0.19089082384460818,
+      "grad_norm": 0.1531815230846405,
+      "learning_rate": 0.0001,
+      "loss": 1.4986,
+      "step": 3070
+    },
+    {
+      "epoch": 0.1912257200267917,
+      "grad_norm": 0.1441792994737625,
+      "learning_rate": 0.0001,
+      "loss": 1.4386,
+      "step": 3071
+    },
+    {
+      "epoch": 0.19156061620897522,
+      "grad_norm": 0.14361928403377533,
+      "learning_rate": 0.0001,
+      "loss": 1.3961,
+      "step": 3072
+    },
+    {
+      "epoch": 0.19189551239115873,
+      "grad_norm": 0.15087366104125977,
+      "learning_rate": 0.0001,
+      "loss": 1.5987,
+      "step": 3073
+    },
+    {
+      "epoch": 0.19223040857334225,
+      "grad_norm": 0.15212655067443848,
+      "learning_rate": 0.0001,
+      "loss": 1.5697,
+      "step": 3074
+    },
+    {
+      "epoch": 0.19256530475552577,
+      "grad_norm": 0.15329968929290771,
+      "learning_rate": 0.0001,
+      "loss": 1.5129,
+      "step": 3075
+    },
+    {
+      "epoch": 0.19290020093770932,
+      "grad_norm": 0.14272186160087585,
+      "learning_rate": 0.0001,
+      "loss": 1.431,
+      "step": 3076
+    },
+    {
+      "epoch": 0.19323509711989284,
+      "grad_norm": 0.14385369420051575,
+      "learning_rate": 0.0001,
+      "loss": 1.5065,
+      "step": 3077
+    },
+    {
+      "epoch": 0.19356999330207636,
+      "grad_norm": 0.14664167165756226,
+      "learning_rate": 0.0001,
+      "loss": 1.472,
+      "step": 3078
+    },
+    {
+      "epoch": 0.19390488948425988,
+      "grad_norm": 0.15277238190174103,
+      "learning_rate": 0.0001,
+      "loss": 1.4597,
+      "step": 3079
+    },
+    {
+      "epoch": 0.1942397856664434,
+      "grad_norm": 0.15171590447425842,
+      "learning_rate": 0.0001,
+      "loss": 1.5161,
+      "step": 3080
+    },
+    {
+      "epoch": 0.19457468184862692,
+      "grad_norm": 0.14549346268177032,
+      "learning_rate": 0.0001,
+      "loss": 1.4613,
+      "step": 3081
+    },
+    {
+      "epoch": 0.19490957803081044,
+      "grad_norm": 0.14628693461418152,
+      "learning_rate": 0.0001,
+      "loss": 1.4514,
+      "step": 3082
+    },
+    {
+      "epoch": 0.19524447421299398,
+      "grad_norm": 0.14419186115264893,
+      "learning_rate": 0.0001,
+      "loss": 1.4797,
+      "step": 3083
+    },
+    {
+      "epoch": 0.1955793703951775,
+      "grad_norm": 0.14945529401302338,
+      "learning_rate": 0.0001,
+      "loss": 1.452,
+      "step": 3084
+    },
+    {
+      "epoch": 0.19591426657736102,
+      "grad_norm": 0.1542719602584839,
+      "learning_rate": 0.0001,
+      "loss": 1.5125,
+      "step": 3085
+    },
+    {
+      "epoch": 0.19624916275954454,
+      "grad_norm": 0.15316370129585266,
+      "learning_rate": 0.0001,
+      "loss": 1.597,
+      "step": 3086
+    },
+    {
+      "epoch": 0.19658405894172806,
+      "grad_norm": 0.15390264987945557,
+      "learning_rate": 0.0001,
+      "loss": 1.4559,
+      "step": 3087
+    },
+    {
+      "epoch": 0.19691895512391158,
+      "grad_norm": 0.15577276051044464,
+      "learning_rate": 0.0001,
+      "loss": 1.5045,
+      "step": 3088
+    },
+    {
+      "epoch": 0.1972538513060951,
+      "grad_norm": 0.14395619928836823,
+      "learning_rate": 0.0001,
+      "loss": 1.485,
+      "step": 3089
+    },
+    {
+      "epoch": 0.19758874748827862,
+      "grad_norm": 0.14563584327697754,
+      "learning_rate": 0.0001,
+      "loss": 1.4736,
+      "step": 3090
+    },
+    {
+      "epoch": 0.19792364367046217,
+      "grad_norm": 0.15247657895088196,
+      "learning_rate": 0.0001,
+      "loss": 1.4362,
+      "step": 3091
+    },
+    {
+      "epoch": 0.1982585398526457,
+      "grad_norm": 0.1521885246038437,
+      "learning_rate": 0.0001,
+      "loss": 1.4417,
+      "step": 3092
+    },
+    {
+      "epoch": 0.1985934360348292,
+      "grad_norm": 0.14599932730197906,
+      "learning_rate": 0.0001,
+      "loss": 1.4764,
+      "step": 3093
+    },
+    {
+      "epoch": 0.19892833221701273,
+      "grad_norm": 0.14040043950080872,
+      "learning_rate": 0.0001,
+      "loss": 1.3132,
+      "step": 3094
+    },
+    {
+      "epoch": 0.19926322839919625,
+      "grad_norm": 0.15043559670448303,
+      "learning_rate": 0.0001,
+      "loss": 1.4917,
+      "step": 3095
+    },
+    {
+      "epoch": 0.19959812458137977,
+      "grad_norm": 0.15356433391571045,
+      "learning_rate": 0.0001,
+      "loss": 1.6173,
+      "step": 3096
+    },
+    {
+      "epoch": 0.19993302076356329,
+      "grad_norm": 0.15018680691719055,
+      "learning_rate": 0.0001,
+      "loss": 1.5324,
+      "step": 3097
+    },
+    {
+      "epoch": 0.20026791694574683,
+      "grad_norm": 0.15176072716712952,
+      "learning_rate": 0.0001,
+      "loss": 1.4686,
+      "step": 3098
+    },
+    {
+      "epoch": 0.20060281312793035,
+      "grad_norm": 0.15239951014518738,
+      "learning_rate": 0.0001,
+      "loss": 1.5726,
+      "step": 3099
+    },
+    {
+      "epoch": 0.20093770931011387,
+      "grad_norm": 0.1486234813928604,
+      "learning_rate": 0.0001,
+      "loss": 1.4154,
+      "step": 3100
+    },
+    {
+      "epoch": 0.2012726054922974,
+      "grad_norm": 0.1507226973772049,
+      "learning_rate": 0.0001,
+      "loss": 1.5301,
+      "step": 3101
+    },
+    {
+      "epoch": 0.2016075016744809,
+      "grad_norm": 0.14854057133197784,
+      "learning_rate": 0.0001,
+      "loss": 1.3898,
+      "step": 3102
+    },
+    {
+      "epoch": 0.20194239785666443,
+      "grad_norm": 0.15393656492233276,
+      "learning_rate": 0.0001,
+      "loss": 1.4901,
+      "step": 3103
+    },
+    {
+      "epoch": 0.20227729403884795,
+      "grad_norm": 0.15347276628017426,
+      "learning_rate": 0.0001,
+      "loss": 1.5183,
+      "step": 3104
+    },
+    {
+      "epoch": 0.20261219022103147,
+      "grad_norm": 0.16099943220615387,
+      "learning_rate": 0.0001,
+      "loss": 1.524,
+      "step": 3105
+    },
+    {
+      "epoch": 0.20294708640321502,
+      "grad_norm": 0.14733991026878357,
+      "learning_rate": 0.0001,
+      "loss": 1.4457,
+      "step": 3106
+    },
+    {
+      "epoch": 0.20328198258539854,
+      "grad_norm": 0.147960364818573,
+      "learning_rate": 0.0001,
+      "loss": 1.4295,
+      "step": 3107
+    },
+    {
+      "epoch": 0.20361687876758205,
+      "grad_norm": 0.15491634607315063,
+      "learning_rate": 0.0001,
+      "loss": 1.4399,
+      "step": 3108
+    },
+    {
+      "epoch": 0.20395177494976557,
+      "grad_norm": 0.14620526134967804,
+      "learning_rate": 0.0001,
+      "loss": 1.4536,
+      "step": 3109
+    },
+    {
+      "epoch": 0.2042866711319491,
+      "grad_norm": 0.14189660549163818,
+      "learning_rate": 0.0001,
+      "loss": 1.3921,
+      "step": 3110
+    },
+    {
+      "epoch": 0.2046215673141326,
+      "grad_norm": 0.14943283796310425,
+      "learning_rate": 0.0001,
+      "loss": 1.342,
+      "step": 3111
+    },
+    {
+      "epoch": 0.20495646349631613,
+      "grad_norm": 0.1576606184244156,
+      "learning_rate": 0.0001,
+      "loss": 1.4832,
+      "step": 3112
+    },
+    {
+      "epoch": 0.20529135967849965,
+      "grad_norm": 0.1504100263118744,
+      "learning_rate": 0.0001,
+      "loss": 1.4987,
+      "step": 3113
+    },
+    {
+      "epoch": 0.2056262558606832,
+      "grad_norm": 0.15109190344810486,
+      "learning_rate": 0.0001,
+      "loss": 1.4591,
+      "step": 3114
+    },
+    {
+      "epoch": 0.20596115204286672,
+      "grad_norm": 0.15667004883289337,
+      "learning_rate": 0.0001,
+      "loss": 1.3823,
+      "step": 3115
+    },
+    {
+      "epoch": 0.20629604822505024,
+      "grad_norm": 0.14729659259319305,
+      "learning_rate": 0.0001,
+      "loss": 1.4394,
+      "step": 3116
+    },
+    {
+      "epoch": 0.20663094440723376,
+      "grad_norm": 0.1476004719734192,
+      "learning_rate": 0.0001,
+      "loss": 1.443,
+      "step": 3117
+    },
+    {
+      "epoch": 0.20696584058941728,
+      "grad_norm": 0.17168119549751282,
+      "learning_rate": 0.0001,
+      "loss": 1.5484,
+      "step": 3118
+    },
+    {
+      "epoch": 0.2073007367716008,
+      "grad_norm": 0.14790430665016174,
+      "learning_rate": 0.0001,
+      "loss": 1.469,
+      "step": 3119
+    },
+    {
+      "epoch": 0.20763563295378432,
+      "grad_norm": 0.16001714766025543,
+      "learning_rate": 0.0001,
+      "loss": 1.4979,
+      "step": 3120
+    },
+    {
+      "epoch": 0.20797052913596786,
+      "grad_norm": 0.15084713697433472,
+      "learning_rate": 0.0001,
+      "loss": 1.3852,
+      "step": 3121
+    },
+    {
+      "epoch": 0.20830542531815138,
+      "grad_norm": 0.14759619534015656,
+      "learning_rate": 0.0001,
+      "loss": 1.5474,
+      "step": 3122
+    },
+    {
+      "epoch": 0.2086403215003349,
+      "grad_norm": 0.15545868873596191,
+      "learning_rate": 0.0001,
+      "loss": 1.5908,
+      "step": 3123
+    },
+    {
+      "epoch": 0.20897521768251842,
+      "grad_norm": 0.15997780859470367,
+      "learning_rate": 0.0001,
+      "loss": 1.4449,
+      "step": 3124
+    },
+    {
+      "epoch": 0.20931011386470194,
+      "grad_norm": 0.14409230649471283,
+      "learning_rate": 0.0001,
+      "loss": 1.5113,
+      "step": 3125
+    },
+    {
+      "epoch": 0.20964501004688546,
+      "grad_norm": 0.14863921701908112,
+      "learning_rate": 0.0001,
+      "loss": 1.5332,
+      "step": 3126
+    },
+    {
+      "epoch": 0.20997990622906898,
+      "grad_norm": 0.14341822266578674,
+      "learning_rate": 0.0001,
+      "loss": 1.4174,
+      "step": 3127
+    },
+    {
+      "epoch": 0.2103148024112525,
+      "grad_norm": 0.14988672733306885,
+      "learning_rate": 0.0001,
+      "loss": 1.4759,
+      "step": 3128
+    },
+    {
+      "epoch": 0.21064969859343605,
+      "grad_norm": 0.14667005836963654,
+      "learning_rate": 0.0001,
+      "loss": 1.4981,
+      "step": 3129
+    },
+    {
+      "epoch": 0.21098459477561957,
+      "grad_norm": 0.14991022646427155,
+      "learning_rate": 0.0001,
+      "loss": 1.5047,
+      "step": 3130
+    },
+    {
+      "epoch": 0.21131949095780309,
+      "grad_norm": 0.14525125920772552,
+      "learning_rate": 0.0001,
+      "loss": 1.3783,
+      "step": 3131
+    },
+    {
+      "epoch": 0.2116543871399866,
+      "grad_norm": 0.15049485862255096,
+      "learning_rate": 0.0001,
+      "loss": 1.4224,
+      "step": 3132
+    },
+    {
+      "epoch": 0.21198928332217012,
+      "grad_norm": 0.15043260157108307,
+      "learning_rate": 0.0001,
+      "loss": 1.4528,
+      "step": 3133
+    },
+    {
+      "epoch": 0.21232417950435364,
+      "grad_norm": 0.15994007885456085,
+      "learning_rate": 0.0001,
+      "loss": 1.5085,
+      "step": 3134
+    },
+    {
+      "epoch": 0.21265907568653716,
+      "grad_norm": 0.14406290650367737,
+      "learning_rate": 0.0001,
+      "loss": 1.464,
+      "step": 3135
+    },
+    {
+      "epoch": 0.21299397186872068,
+      "grad_norm": 0.15055160224437714,
+      "learning_rate": 0.0001,
+      "loss": 1.4788,
+      "step": 3136
+    },
+    {
+      "epoch": 0.21332886805090423,
+      "grad_norm": 0.150338813662529,
+      "learning_rate": 0.0001,
+      "loss": 1.5285,
+      "step": 3137
+    },
+    {
+      "epoch": 0.21366376423308775,
+      "grad_norm": 0.1500924527645111,
+      "learning_rate": 0.0001,
+      "loss": 1.4794,
+      "step": 3138
+    },
+    {
+      "epoch": 0.21399866041527127,
+      "grad_norm": 0.14780132472515106,
+      "learning_rate": 0.0001,
+      "loss": 1.5283,
+      "step": 3139
+    },
+    {
+      "epoch": 0.2143335565974548,
+      "grad_norm": 0.14827805757522583,
+      "learning_rate": 0.0001,
+      "loss": 1.4136,
+      "step": 3140
+    },
+    {
+      "epoch": 0.2146684527796383,
+      "grad_norm": 0.14717240631580353,
+      "learning_rate": 0.0001,
+      "loss": 1.4812,
+      "step": 3141
+    },
+    {
+      "epoch": 0.21500334896182183,
+      "grad_norm": 0.15452304482460022,
+      "learning_rate": 0.0001,
+      "loss": 1.492,
+      "step": 3142
+    },
+    {
+      "epoch": 0.21533824514400535,
+      "grad_norm": 0.17032486200332642,
+      "learning_rate": 0.0001,
+      "loss": 1.5352,
+      "step": 3143
+    },
+    {
+      "epoch": 0.2156731413261889,
+      "grad_norm": 0.15467940270900726,
+      "learning_rate": 0.0001,
+      "loss": 1.5039,
+      "step": 3144
+    },
+    {
+      "epoch": 0.2160080375083724,
+      "grad_norm": 0.15712405741214752,
+      "learning_rate": 0.0001,
+      "loss": 1.507,
+      "step": 3145
+    },
+    {
+      "epoch": 0.21634293369055593,
+      "grad_norm": 0.158954918384552,
+      "learning_rate": 0.0001,
+      "loss": 1.5357,
+      "step": 3146
+    },
+    {
+      "epoch": 0.21667782987273945,
+      "grad_norm": 0.14702315628528595,
+      "learning_rate": 0.0001,
+      "loss": 1.5295,
+      "step": 3147
+    },
+    {
+      "epoch": 0.21701272605492297,
+      "grad_norm": 0.14634455740451813,
+      "learning_rate": 0.0001,
+      "loss": 1.4175,
+      "step": 3148
+    },
+    {
+      "epoch": 0.2173476222371065,
+      "grad_norm": 0.15392786264419556,
+      "learning_rate": 0.0001,
+      "loss": 1.5544,
+      "step": 3149
+    },
+    {
+      "epoch": 0.21768251841929,
+      "grad_norm": 0.16009333729743958,
+      "learning_rate": 0.0001,
+      "loss": 1.63,
+      "step": 3150
+    },
+    {
+      "epoch": 0.21801741460147353,
+      "grad_norm": 0.14513467252254486,
+      "learning_rate": 0.0001,
+      "loss": 1.4659,
+      "step": 3151
+    },
+    {
+      "epoch": 0.21835231078365708,
+      "grad_norm": 0.14836783707141876,
+      "learning_rate": 0.0001,
+      "loss": 1.4878,
+      "step": 3152
+    },
+    {
+      "epoch": 0.2186872069658406,
+      "grad_norm": 0.15242692828178406,
+      "learning_rate": 0.0001,
+      "loss": 1.4305,
+      "step": 3153
+    },
+    {
+      "epoch": 0.21902210314802412,
+      "grad_norm": 0.14825765788555145,
+      "learning_rate": 0.0001,
+      "loss": 1.4887,
+      "step": 3154
+    },
+    {
+      "epoch": 0.21935699933020764,
+      "grad_norm": 0.15620997548103333,
+      "learning_rate": 0.0001,
+      "loss": 1.5047,
+      "step": 3155
+    },
+    {
+      "epoch": 0.21969189551239116,
+      "grad_norm": 0.15094633400440216,
+      "learning_rate": 0.0001,
+      "loss": 1.4292,
+      "step": 3156
+    },
+    {
+      "epoch": 0.22002679169457467,
+      "grad_norm": 0.1550116240978241,
+      "learning_rate": 0.0001,
+      "loss": 1.4542,
+      "step": 3157
+    },
+    {
+      "epoch": 0.2203616878767582,
+      "grad_norm": 0.15815100073814392,
+      "learning_rate": 0.0001,
+      "loss": 1.5237,
+      "step": 3158
+    },
+    {
+      "epoch": 0.22069658405894174,
+      "grad_norm": 0.1517287939786911,
+      "learning_rate": 0.0001,
+      "loss": 1.4601,
+      "step": 3159
+    },
+    {
+      "epoch": 0.22103148024112526,
+      "grad_norm": 0.15398970246315002,
+      "learning_rate": 0.0001,
+      "loss": 1.4871,
+      "step": 3160
+    },
+    {
+      "epoch": 0.22136637642330878,
+      "grad_norm": 0.1565081775188446,
+      "learning_rate": 0.0001,
+      "loss": 1.4783,
+      "step": 3161
+    },
+    {
+      "epoch": 0.2217012726054923,
+      "grad_norm": 0.14842985570430756,
+      "learning_rate": 0.0001,
+      "loss": 1.3362,
+      "step": 3162
+    },
+    {
+      "epoch": 0.22203616878767582,
+      "grad_norm": 0.15767332911491394,
+      "learning_rate": 0.0001,
+      "loss": 1.4659,
+      "step": 3163
+    },
+    {
+      "epoch": 0.22237106496985934,
+      "grad_norm": 0.15569786727428436,
+      "learning_rate": 0.0001,
+      "loss": 1.5224,
+      "step": 3164
+    },
+    {
+      "epoch": 0.22270596115204286,
+      "grad_norm": 0.15435588359832764,
+      "learning_rate": 0.0001,
+      "loss": 1.5205,
+      "step": 3165
+    },
+    {
+      "epoch": 0.22304085733422638,
+      "grad_norm": 0.15310215950012207,
+      "learning_rate": 0.0001,
+      "loss": 1.4772,
+      "step": 3166
+    },
+    {
+      "epoch": 0.22337575351640993,
+      "grad_norm": 0.16117234528064728,
+      "learning_rate": 0.0001,
+      "loss": 1.4884,
+      "step": 3167
+    },
+    {
+      "epoch": 0.22371064969859344,
+      "grad_norm": 0.15367351472377777,
+      "learning_rate": 0.0001,
+      "loss": 1.5431,
+      "step": 3168
+    },
+    {
+      "epoch": 0.22404554588077696,
+      "grad_norm": 0.15488213300704956,
+      "learning_rate": 0.0001,
+      "loss": 1.4126,
+      "step": 3169
+    },
+    {
+      "epoch": 0.22438044206296048,
+      "grad_norm": 0.15359318256378174,
+      "learning_rate": 0.0001,
+      "loss": 1.5055,
+      "step": 3170
+    },
+    {
+      "epoch": 0.224715338245144,
+      "grad_norm": 0.14719818532466888,
+      "learning_rate": 0.0001,
+      "loss": 1.4483,
+      "step": 3171
+    },
+    {
+      "epoch": 0.22505023442732752,
+      "grad_norm": 0.15634144842624664,
+      "learning_rate": 0.0001,
+      "loss": 1.57,
+      "step": 3172
+    },
+    {
+      "epoch": 0.22538513060951104,
+      "grad_norm": 0.1440373957157135,
+      "learning_rate": 0.0001,
+      "loss": 1.4468,
+      "step": 3173
+    },
+    {
+      "epoch": 0.22572002679169456,
+      "grad_norm": 0.14679275453090668,
+      "learning_rate": 0.0001,
+      "loss": 1.4967,
+      "step": 3174
+    },
+    {
+      "epoch": 0.2260549229738781,
+      "grad_norm": 0.1491180658340454,
+      "learning_rate": 0.0001,
+      "loss": 1.5348,
+      "step": 3175
+    },
+    {
+      "epoch": 0.22638981915606163,
+      "grad_norm": 0.16220654547214508,
+      "learning_rate": 0.0001,
+      "loss": 1.51,
+      "step": 3176
+    },
+    {
+      "epoch": 0.22672471533824515,
+      "grad_norm": 0.14397957921028137,
+      "learning_rate": 0.0001,
+      "loss": 1.3946,
+      "step": 3177
+    },
+    {
+      "epoch": 0.22705961152042867,
+      "grad_norm": 0.1529865860939026,
+      "learning_rate": 0.0001,
+      "loss": 1.5672,
+      "step": 3178
+    },
+    {
+      "epoch": 0.2273945077026122,
+      "grad_norm": 0.14720121026039124,
+      "learning_rate": 0.0001,
+      "loss": 1.4336,
+      "step": 3179
+    },
+    {
+      "epoch": 0.2277294038847957,
+      "grad_norm": 0.14547964930534363,
+      "learning_rate": 0.0001,
+      "loss": 1.5099,
+      "step": 3180
+    },
+    {
+      "epoch": 0.22806430006697923,
+      "grad_norm": 0.15165194869041443,
+      "learning_rate": 0.0001,
+      "loss": 1.4584,
+      "step": 3181
+    },
+    {
+      "epoch": 0.22839919624916277,
+      "grad_norm": 0.14752137660980225,
+      "learning_rate": 0.0001,
+      "loss": 1.4914,
+      "step": 3182
+    },
+    {
+      "epoch": 0.2287340924313463,
+      "grad_norm": 0.142874076962471,
+      "learning_rate": 0.0001,
+      "loss": 1.4233,
+      "step": 3183
+    },
+    {
+      "epoch": 0.2290689886135298,
+      "grad_norm": 0.1451757699251175,
+      "learning_rate": 0.0001,
+      "loss": 1.5243,
+      "step": 3184
+    },
+    {
+      "epoch": 0.22940388479571333,
+      "grad_norm": 0.14648723602294922,
+      "learning_rate": 0.0001,
+      "loss": 1.4939,
+      "step": 3185
+    },
+    {
+      "epoch": 0.22973878097789685,
+      "grad_norm": 0.15457214415073395,
+      "learning_rate": 0.0001,
+      "loss": 1.5448,
+      "step": 3186
+    },
+    {
+      "epoch": 0.23007367716008037,
+      "grad_norm": 0.14498350024223328,
+      "learning_rate": 0.0001,
+      "loss": 1.4727,
+      "step": 3187
+    },
+    {
+      "epoch": 0.2304085733422639,
+      "grad_norm": 0.13952688872814178,
+      "learning_rate": 0.0001,
+      "loss": 1.3022,
+      "step": 3188
+    },
+    {
+      "epoch": 0.2307434695244474,
+      "grad_norm": 0.15119028091430664,
+      "learning_rate": 0.0001,
+      "loss": 1.5231,
+      "step": 3189
+    },
+    {
+      "epoch": 0.23107836570663096,
+      "grad_norm": 0.15895050764083862,
+      "learning_rate": 0.0001,
+      "loss": 1.5578,
+      "step": 3190
+    },
+    {
+      "epoch": 0.23141326188881448,
+      "grad_norm": 0.15064969658851624,
+      "learning_rate": 0.0001,
+      "loss": 1.4316,
+      "step": 3191
+    },
+    {
+      "epoch": 0.231748158070998,
+      "grad_norm": 0.14713798463344574,
+      "learning_rate": 0.0001,
+      "loss": 1.4909,
+      "step": 3192
+    },
+    {
+      "epoch": 0.23208305425318151,
+      "grad_norm": 0.17914965748786926,
+      "learning_rate": 0.0001,
+      "loss": 1.47,
+      "step": 3193
+    },
+    {
+      "epoch": 0.23241795043536503,
+      "grad_norm": 0.1502309888601303,
+      "learning_rate": 0.0001,
+      "loss": 1.4022,
+      "step": 3194
+    },
+    {
+      "epoch": 0.23275284661754855,
+      "grad_norm": 0.14691807329654694,
+      "learning_rate": 0.0001,
+      "loss": 1.4684,
+      "step": 3195
+    },
+    {
+      "epoch": 0.23308774279973207,
+      "grad_norm": 0.1478799283504486,
+      "learning_rate": 0.0001,
+      "loss": 1.4575,
+      "step": 3196
+    },
+    {
+      "epoch": 0.23342263898191562,
+      "grad_norm": 0.14784233272075653,
+      "learning_rate": 0.0001,
+      "loss": 1.3907,
+      "step": 3197
+    },
+    {
+      "epoch": 0.23375753516409914,
+      "grad_norm": 0.15338735282421112,
+      "learning_rate": 0.0001,
+      "loss": 1.5106,
+      "step": 3198
+    },
+    {
+      "epoch": 0.23409243134628266,
+      "grad_norm": 0.15106314420700073,
+      "learning_rate": 0.0001,
+      "loss": 1.5257,
+      "step": 3199
+    },
+    {
+      "epoch": 0.23442732752846618,
+      "grad_norm": 0.1492011845111847,
+      "learning_rate": 0.0001,
+      "loss": 1.4579,
+      "step": 3200
+    },
+    {
+      "epoch": 0.2347622237106497,
+      "grad_norm": 0.14995889365673065,
+      "learning_rate": 0.0001,
+      "loss": 1.4645,
+      "step": 3201
+    },
+    {
+      "epoch": 0.23509711989283322,
+      "grad_norm": 0.15579792857170105,
+      "learning_rate": 0.0001,
+      "loss": 1.5242,
+      "step": 3202
+    },
+    {
+      "epoch": 0.23543201607501674,
+      "grad_norm": 0.1529867798089981,
+      "learning_rate": 0.0001,
+      "loss": 1.3896,
+      "step": 3203
+    },
+    {
+      "epoch": 0.23576691225720026,
+      "grad_norm": 0.1536225974559784,
+      "learning_rate": 0.0001,
+      "loss": 1.4913,
+      "step": 3204
+    },
+    {
+      "epoch": 0.2361018084393838,
+      "grad_norm": 0.14973454177379608,
+      "learning_rate": 0.0001,
+      "loss": 1.4563,
+      "step": 3205
+    },
+    {
+      "epoch": 0.23643670462156732,
+      "grad_norm": 0.16134217381477356,
+      "learning_rate": 0.0001,
+      "loss": 1.5128,
+      "step": 3206
+    },
+    {
+      "epoch": 0.23677160080375084,
+      "grad_norm": 0.1515876054763794,
+      "learning_rate": 0.0001,
+      "loss": 1.5311,
+      "step": 3207
+    },
+    {
+      "epoch": 0.23710649698593436,
+      "grad_norm": 0.14970046281814575,
+      "learning_rate": 0.0001,
+      "loss": 1.5453,
+      "step": 3208
+    },
+    {
+      "epoch": 0.23744139316811788,
+      "grad_norm": 0.16384047269821167,
+      "learning_rate": 0.0001,
+      "loss": 1.4953,
+      "step": 3209
+    },
+    {
+      "epoch": 0.2377762893503014,
+      "grad_norm": 0.1614055335521698,
+      "learning_rate": 0.0001,
+      "loss": 1.3486,
+      "step": 3210
+    },
+    {
+      "epoch": 0.23811118553248492,
+      "grad_norm": 0.15903745591640472,
+      "learning_rate": 0.0001,
+      "loss": 1.5484,
+      "step": 3211
+    },
+    {
+      "epoch": 0.23844608171466844,
+      "grad_norm": 0.15176984667778015,
+      "learning_rate": 0.0001,
+      "loss": 1.4096,
+      "step": 3212
+    },
+    {
+      "epoch": 0.238780977896852,
+      "grad_norm": 0.1553134322166443,
+      "learning_rate": 0.0001,
+      "loss": 1.5071,
+      "step": 3213
+    },
+    {
+      "epoch": 0.2391158740790355,
+      "grad_norm": 0.15207169950008392,
+      "learning_rate": 0.0001,
+      "loss": 1.4846,
+      "step": 3214
+    },
+    {
+      "epoch": 0.23945077026121903,
+      "grad_norm": 0.14777179062366486,
+      "learning_rate": 0.0001,
+      "loss": 1.5013,
+      "step": 3215
+    },
+    {
+      "epoch": 0.23978566644340255,
+      "grad_norm": 0.15911854803562164,
+      "learning_rate": 0.0001,
+      "loss": 1.5292,
+      "step": 3216
+    },
+    {
+      "epoch": 0.24012056262558606,
+      "grad_norm": 0.16325682401657104,
+      "learning_rate": 0.0001,
+      "loss": 1.5035,
+      "step": 3217
+    },
+    {
+      "epoch": 0.24045545880776958,
+      "grad_norm": 0.1513979583978653,
+      "learning_rate": 0.0001,
+      "loss": 1.4521,
+      "step": 3218
+    },
+    {
+      "epoch": 0.2407903549899531,
+      "grad_norm": 0.16634339094161987,
+      "learning_rate": 0.0001,
+      "loss": 1.5492,
+      "step": 3219
+    },
+    {
+      "epoch": 0.24112525117213665,
+      "grad_norm": 0.15186385810375214,
+      "learning_rate": 0.0001,
+      "loss": 1.4537,
+      "step": 3220
+    },
+    {
+      "epoch": 0.24146014735432017,
+      "grad_norm": 0.15369641780853271,
+      "learning_rate": 0.0001,
+      "loss": 1.4837,
+      "step": 3221
+    },
+    {
+      "epoch": 0.2417950435365037,
+      "grad_norm": 0.15841279923915863,
+      "learning_rate": 0.0001,
+      "loss": 1.4912,
+      "step": 3222
+    },
+    {
+      "epoch": 0.2421299397186872,
+      "grad_norm": 0.15733489394187927,
+      "learning_rate": 0.0001,
+      "loss": 1.4796,
+      "step": 3223
+    },
+    {
+      "epoch": 0.24246483590087073,
+      "grad_norm": 0.14990831911563873,
+      "learning_rate": 0.0001,
+      "loss": 1.5174,
+      "step": 3224
+    },
+    {
+      "epoch": 0.24279973208305425,
+      "grad_norm": 0.1500232070684433,
+      "learning_rate": 0.0001,
+      "loss": 1.479,
+      "step": 3225
+    },
+    {
+      "epoch": 0.24313462826523777,
+      "grad_norm": 0.14930838346481323,
+      "learning_rate": 0.0001,
+      "loss": 1.4311,
+      "step": 3226
+    },
+    {
+      "epoch": 0.2434695244474213,
+      "grad_norm": 0.14843516051769257,
+      "learning_rate": 0.0001,
+      "loss": 1.4714,
+      "step": 3227
+    },
+    {
+      "epoch": 0.24380442062960483,
+      "grad_norm": 0.14314772188663483,
+      "learning_rate": 0.0001,
+      "loss": 1.4189,
+      "step": 3228
+    },
+    {
+      "epoch": 0.24413931681178835,
+      "grad_norm": 0.15701232850551605,
+      "learning_rate": 0.0001,
+      "loss": 1.5089,
+      "step": 3229
+    },
+    {
+      "epoch": 0.24447421299397187,
+      "grad_norm": 0.14976859092712402,
+      "learning_rate": 0.0001,
+      "loss": 1.4099,
+      "step": 3230
+    },
+    {
+      "epoch": 0.2448091091761554,
+      "grad_norm": 0.14823557436466217,
+      "learning_rate": 0.0001,
+      "loss": 1.5241,
+      "step": 3231
+    },
+    {
+      "epoch": 0.2451440053583389,
+      "grad_norm": 0.1463557332754135,
+      "learning_rate": 0.0001,
+      "loss": 1.4538,
+      "step": 3232
+    },
+    {
+      "epoch": 0.24547890154052243,
+      "grad_norm": 0.15305958688259125,
+      "learning_rate": 0.0001,
+      "loss": 1.5595,
+      "step": 3233
+    },
+    {
+      "epoch": 0.24581379772270595,
+      "grad_norm": 0.15024501085281372,
+      "learning_rate": 0.0001,
+      "loss": 1.4731,
+      "step": 3234
+    },
+    {
+      "epoch": 0.24614869390488947,
+      "grad_norm": 0.15059971809387207,
+      "learning_rate": 0.0001,
+      "loss": 1.5031,
+      "step": 3235
+    },
+    {
+      "epoch": 0.24648359008707302,
+      "grad_norm": 0.15700224041938782,
+      "learning_rate": 0.0001,
+      "loss": 1.518,
+      "step": 3236
+    },
+    {
+      "epoch": 0.24681848626925654,
+      "grad_norm": 0.1432359218597412,
+      "learning_rate": 0.0001,
+      "loss": 1.472,
+      "step": 3237
+    },
+    {
+      "epoch": 0.24715338245144006,
+      "grad_norm": 0.14348803460597992,
+      "learning_rate": 0.0001,
+      "loss": 1.4363,
+      "step": 3238
+    },
+    {
+      "epoch": 0.24748827863362358,
+      "grad_norm": 0.15466760098934174,
+      "learning_rate": 0.0001,
+      "loss": 1.4862,
+      "step": 3239
+    },
+    {
+      "epoch": 0.2478231748158071,
+      "grad_norm": 0.1510024070739746,
+      "learning_rate": 0.0001,
+      "loss": 1.5311,
+      "step": 3240
+    },
+    {
+      "epoch": 0.24815807099799062,
+      "grad_norm": 0.15169547498226166,
+      "learning_rate": 0.0001,
+      "loss": 1.5277,
+      "step": 3241
+    },
+    {
+      "epoch": 0.24849296718017413,
+      "grad_norm": 0.1525082141160965,
+      "learning_rate": 0.0001,
+      "loss": 1.4847,
+      "step": 3242
+    },
+    {
+      "epoch": 0.24882786336235768,
+      "grad_norm": 0.1485402137041092,
+      "learning_rate": 0.0001,
+      "loss": 1.5632,
+      "step": 3243
+    },
+    {
+      "epoch": 0.2491627595445412,
+      "grad_norm": 0.1473839432001114,
+      "learning_rate": 0.0001,
+      "loss": 1.4764,
+      "step": 3244
+    },
+    {
+      "epoch": 0.24949765572672472,
+      "grad_norm": 0.1522548496723175,
+      "learning_rate": 0.0001,
+      "loss": 1.6454,
+      "step": 3245
+    },
+    {
+      "epoch": 0.24983255190890824,
+      "grad_norm": 0.1536744087934494,
+      "learning_rate": 0.0001,
+      "loss": 1.479,
+      "step": 3246
+    },
+    {
+      "epoch": 0.25016744809109176,
+      "grad_norm": 0.14527705311775208,
+      "learning_rate": 0.0001,
+      "loss": 1.5183,
+      "step": 3247
+    },
+    {
+      "epoch": 0.2505023442732753,
+      "grad_norm": 0.1515124887228012,
+      "learning_rate": 0.0001,
+      "loss": 1.477,
+      "step": 3248
+    },
+    {
+      "epoch": 0.2508372404554588,
+      "grad_norm": 0.1502838134765625,
+      "learning_rate": 0.0001,
+      "loss": 1.4847,
+      "step": 3249
+    },
+    {
+      "epoch": 0.2511721366376423,
+      "grad_norm": 0.14173923432826996,
+      "learning_rate": 0.0001,
+      "loss": 1.4254,
+      "step": 3250
+    },
+    {
+      "epoch": 0.25150703281982584,
+      "grad_norm": 0.14899542927742004,
+      "learning_rate": 0.0001,
+      "loss": 1.4805,
+      "step": 3251
+    },
+    {
+      "epoch": 0.25184192900200936,
+      "grad_norm": 0.1492302268743515,
+      "learning_rate": 0.0001,
+      "loss": 1.5133,
+      "step": 3252
+    },
+    {
+      "epoch": 0.2521768251841929,
+      "grad_norm": 0.15040142834186554,
+      "learning_rate": 0.0001,
+      "loss": 1.4799,
+      "step": 3253
+    },
+    {
+      "epoch": 0.2525117213663764,
+      "grad_norm": 0.14641860127449036,
+      "learning_rate": 0.0001,
+      "loss": 1.4095,
+      "step": 3254
+    },
+    {
+      "epoch": 0.25284661754855997,
+      "grad_norm": 0.15745055675506592,
+      "learning_rate": 0.0001,
+      "loss": 1.5862,
+      "step": 3255
+    },
+    {
+      "epoch": 0.2531815137307435,
+      "grad_norm": 0.1484472006559372,
+      "learning_rate": 0.0001,
+      "loss": 1.5026,
+      "step": 3256
+    },
+    {
+      "epoch": 0.253516409912927,
+      "grad_norm": 0.14549176394939423,
+      "learning_rate": 0.0001,
+      "loss": 1.4881,
+      "step": 3257
+    },
+    {
+      "epoch": 0.25385130609511053,
+      "grad_norm": 0.15266196429729462,
+      "learning_rate": 0.0001,
+      "loss": 1.595,
+      "step": 3258
+    },
+    {
+      "epoch": 0.25418620227729405,
+      "grad_norm": 0.149305522441864,
+      "learning_rate": 0.0001,
+      "loss": 1.4982,
+      "step": 3259
+    },
+    {
+      "epoch": 0.25452109845947757,
+      "grad_norm": 0.14763863384723663,
+      "learning_rate": 0.0001,
+      "loss": 1.4299,
+      "step": 3260
+    },
+    {
+      "epoch": 0.2548559946416611,
+      "grad_norm": 0.15491554141044617,
+      "learning_rate": 0.0001,
+      "loss": 1.4403,
+      "step": 3261
+    },
+    {
+      "epoch": 0.2551908908238446,
+      "grad_norm": 0.14249491691589355,
+      "learning_rate": 0.0001,
+      "loss": 1.469,
+      "step": 3262
+    },
+    {
+      "epoch": 0.2555257870060281,
+      "grad_norm": 0.14917422831058502,
+      "learning_rate": 0.0001,
+      "loss": 1.401,
+      "step": 3263
+    },
+    {
+      "epoch": 0.25586068318821165,
+      "grad_norm": 0.14510908722877502,
+      "learning_rate": 0.0001,
+      "loss": 1.4544,
+      "step": 3264
+    },
+    {
+      "epoch": 0.25619557937039517,
+      "grad_norm": 0.15895402431488037,
+      "learning_rate": 0.0001,
+      "loss": 1.5992,
+      "step": 3265
+    },
+    {
+      "epoch": 0.2565304755525787,
+      "grad_norm": 0.15244704484939575,
+      "learning_rate": 0.0001,
+      "loss": 1.5201,
+      "step": 3266
+    },
+    {
+      "epoch": 0.2568653717347622,
+      "grad_norm": 0.15109996497631073,
+      "learning_rate": 0.0001,
+      "loss": 1.3982,
+      "step": 3267
+    },
+    {
+      "epoch": 0.2572002679169457,
+      "grad_norm": 0.15099941194057465,
+      "learning_rate": 0.0001,
+      "loss": 1.4949,
+      "step": 3268
+    },
+    {
+      "epoch": 0.25753516409912924,
+      "grad_norm": 0.15113183856010437,
+      "learning_rate": 0.0001,
+      "loss": 1.4185,
+      "step": 3269
+    },
+    {
+      "epoch": 0.2578700602813128,
+      "grad_norm": 0.15029966831207275,
+      "learning_rate": 0.0001,
+      "loss": 1.4401,
+      "step": 3270
+    },
+    {
+      "epoch": 0.25820495646349634,
+      "grad_norm": 0.15110653638839722,
+      "learning_rate": 0.0001,
+      "loss": 1.4606,
+      "step": 3271
+    },
+    {
+      "epoch": 0.25853985264567986,
+      "grad_norm": 0.14699125289916992,
+      "learning_rate": 0.0001,
+      "loss": 1.3752,
+      "step": 3272
+    },
+    {
+      "epoch": 0.2588747488278634,
+      "grad_norm": 0.1514248549938202,
+      "learning_rate": 0.0001,
+      "loss": 1.4327,
+      "step": 3273
+    },
+    {
+      "epoch": 0.2592096450100469,
+      "grad_norm": 0.14915835857391357,
+      "learning_rate": 0.0001,
+      "loss": 1.4541,
+      "step": 3274
+    },
+    {
+      "epoch": 0.2595445411922304,
+      "grad_norm": 0.1568090319633484,
+      "learning_rate": 0.0001,
+      "loss": 1.5556,
+      "step": 3275
+    },
+    {
+      "epoch": 0.25987943737441394,
+      "grad_norm": 0.15016591548919678,
+      "learning_rate": 0.0001,
+      "loss": 1.4523,
+      "step": 3276
+    },
+    {
+      "epoch": 0.26021433355659745,
+      "grad_norm": 0.14889879524707794,
+      "learning_rate": 0.0001,
+      "loss": 1.5291,
+      "step": 3277
+    },
+    {
+      "epoch": 0.260549229738781,
+      "grad_norm": 0.15212146937847137,
+      "learning_rate": 0.0001,
+      "loss": 1.4512,
+      "step": 3278
+    },
+    {
+      "epoch": 0.2608841259209645,
+      "grad_norm": 0.1520780622959137,
+      "learning_rate": 0.0001,
+      "loss": 1.4806,
+      "step": 3279
+    },
+    {
+      "epoch": 0.261219022103148,
+      "grad_norm": 0.14372298121452332,
+      "learning_rate": 0.0001,
+      "loss": 1.4424,
+      "step": 3280
+    },
+    {
+      "epoch": 0.26155391828533153,
+      "grad_norm": 0.14275674521923065,
+      "learning_rate": 0.0001,
+      "loss": 1.399,
+      "step": 3281
+    },
+    {
+      "epoch": 0.26188881446751505,
+      "grad_norm": 0.16535142064094543,
+      "learning_rate": 0.0001,
+      "loss": 1.6046,
+      "step": 3282
+    },
+    {
+      "epoch": 0.26222371064969857,
+      "grad_norm": 0.15277263522148132,
+      "learning_rate": 0.0001,
+      "loss": 1.4929,
+      "step": 3283
+    },
+    {
+      "epoch": 0.2625586068318821,
+      "grad_norm": 0.15697571635246277,
+      "learning_rate": 0.0001,
+      "loss": 1.5432,
+      "step": 3284
+    },
+    {
+      "epoch": 0.26289350301406567,
+      "grad_norm": 0.1490306705236435,
+      "learning_rate": 0.0001,
+      "loss": 1.5217,
+      "step": 3285
+    },
+    {
+      "epoch": 0.2632283991962492,
+      "grad_norm": 0.144150048494339,
+      "learning_rate": 0.0001,
+      "loss": 1.3888,
+      "step": 3286
+    },
+    {
+      "epoch": 0.2635632953784327,
+      "grad_norm": 0.14768853783607483,
+      "learning_rate": 0.0001,
+      "loss": 1.434,
+      "step": 3287
+    },
+    {
+      "epoch": 0.2638981915606162,
+      "grad_norm": 0.15402550995349884,
+      "learning_rate": 0.0001,
+      "loss": 1.4649,
+      "step": 3288
+    },
+    {
+      "epoch": 0.26423308774279974,
+      "grad_norm": 0.14135251939296722,
+      "learning_rate": 0.0001,
+      "loss": 1.3352,
+      "step": 3289
+    },
+    {
+      "epoch": 0.26456798392498326,
+      "grad_norm": 0.14637348055839539,
+      "learning_rate": 0.0001,
+      "loss": 1.4109,
+      "step": 3290
+    },
+    {
+      "epoch": 0.2649028801071668,
+      "grad_norm": 0.16085872054100037,
+      "learning_rate": 0.0001,
+      "loss": 1.5278,
+      "step": 3291
+    },
+    {
+      "epoch": 0.2652377762893503,
+      "grad_norm": 0.1754620373249054,
+      "learning_rate": 0.0001,
+      "loss": 1.6455,
+      "step": 3292
+    },
+    {
+      "epoch": 0.2655726724715338,
+      "grad_norm": 0.15829911828041077,
+      "learning_rate": 0.0001,
+      "loss": 1.5336,
+      "step": 3293
+    },
+    {
+      "epoch": 0.26590756865371734,
+      "grad_norm": 0.15245473384857178,
+      "learning_rate": 0.0001,
+      "loss": 1.3699,
+      "step": 3294
+    },
+    {
+      "epoch": 0.26624246483590086,
+      "grad_norm": 0.16027551889419556,
+      "learning_rate": 0.0001,
+      "loss": 1.5331,
+      "step": 3295
+    },
+    {
+      "epoch": 0.2665773610180844,
+      "grad_norm": 0.1470411866903305,
+      "learning_rate": 0.0001,
+      "loss": 1.447,
+      "step": 3296
+    },
+    {
+      "epoch": 0.2669122572002679,
+      "grad_norm": 0.15567657351493835,
+      "learning_rate": 0.0001,
+      "loss": 1.4364,
+      "step": 3297
+    },
+    {
+      "epoch": 0.2672471533824514,
+      "grad_norm": 0.15527381002902985,
+      "learning_rate": 0.0001,
+      "loss": 1.5166,
+      "step": 3298
+    },
+    {
+      "epoch": 0.26758204956463494,
+      "grad_norm": 0.14668625593185425,
+      "learning_rate": 0.0001,
+      "loss": 1.5355,
+      "step": 3299
+    },
+    {
+      "epoch": 0.2679169457468185,
+      "grad_norm": 0.15117013454437256,
+      "learning_rate": 0.0001,
+      "loss": 1.4292,
+      "step": 3300
+    },
+    {
+      "epoch": 0.26825184192900203,
+      "grad_norm": 0.15294520556926727,
+      "learning_rate": 0.0001,
+      "loss": 1.4845,
+      "step": 3301
+    },
+    {
+      "epoch": 0.26858673811118555,
+      "grad_norm": 0.15113379061222076,
+      "learning_rate": 0.0001,
+      "loss": 1.4522,
+      "step": 3302
+    },
+    {
+      "epoch": 0.26892163429336907,
+      "grad_norm": 0.14972607791423798,
+      "learning_rate": 0.0001,
+      "loss": 1.5287,
+      "step": 3303
+    },
+    {
+      "epoch": 0.2692565304755526,
+      "grad_norm": 0.15561367571353912,
+      "learning_rate": 0.0001,
+      "loss": 1.5022,
+      "step": 3304
+    },
+    {
+      "epoch": 0.2695914266577361,
+      "grad_norm": 0.15228398144245148,
+      "learning_rate": 0.0001,
+      "loss": 1.4579,
+      "step": 3305
+    },
+    {
+      "epoch": 0.26992632283991963,
+      "grad_norm": 0.15206193923950195,
+      "learning_rate": 0.0001,
+      "loss": 1.4691,
+      "step": 3306
+    },
+    {
+      "epoch": 0.27026121902210315,
+      "grad_norm": 0.15127262473106384,
+      "learning_rate": 0.0001,
+      "loss": 1.4826,
+      "step": 3307
+    },
+    {
+      "epoch": 0.27059611520428667,
+      "grad_norm": 0.14557790756225586,
+      "learning_rate": 0.0001,
+      "loss": 1.5215,
+      "step": 3308
+    },
+    {
+      "epoch": 0.2709310113864702,
+      "grad_norm": 0.15763740241527557,
+      "learning_rate": 0.0001,
+      "loss": 1.5212,
+      "step": 3309
+    },
+    {
+      "epoch": 0.2712659075686537,
+      "grad_norm": 0.1479531228542328,
+      "learning_rate": 0.0001,
+      "loss": 1.4819,
+      "step": 3310
+    },
+    {
+      "epoch": 0.2716008037508372,
+      "grad_norm": 0.14689850807189941,
+      "learning_rate": 0.0001,
+      "loss": 1.4905,
+      "step": 3311
+    },
+    {
+      "epoch": 0.27193569993302075,
+      "grad_norm": 0.15508733689785004,
+      "learning_rate": 0.0001,
+      "loss": 1.5279,
+      "step": 3312
+    },
+    {
+      "epoch": 0.27227059611520427,
+      "grad_norm": 0.1534416526556015,
+      "learning_rate": 0.0001,
+      "loss": 1.4621,
+      "step": 3313
+    },
+    {
+      "epoch": 0.2726054922973878,
+      "grad_norm": 0.14797250926494598,
+      "learning_rate": 0.0001,
+      "loss": 1.429,
+      "step": 3314
+    },
+    {
+      "epoch": 0.2729403884795713,
+      "grad_norm": 0.15446336567401886,
+      "learning_rate": 0.0001,
+      "loss": 1.4702,
+      "step": 3315
+    },
+    {
+      "epoch": 0.2732752846617549,
+      "grad_norm": 0.1508505940437317,
+      "learning_rate": 0.0001,
+      "loss": 1.5452,
+      "step": 3316
+    },
+    {
+      "epoch": 0.2736101808439384,
+      "grad_norm": 0.15143181383609772,
+      "learning_rate": 0.0001,
+      "loss": 1.4926,
+      "step": 3317
+    },
+    {
+      "epoch": 0.2739450770261219,
+      "grad_norm": 0.15116530656814575,
+      "learning_rate": 0.0001,
+      "loss": 1.4111,
+      "step": 3318
+    },
+    {
+      "epoch": 0.27427997320830544,
+      "grad_norm": 0.1512419730424881,
+      "learning_rate": 0.0001,
+      "loss": 1.4943,
+      "step": 3319
+    },
+    {
+      "epoch": 0.27461486939048896,
+      "grad_norm": 0.1462402641773224,
+      "learning_rate": 0.0001,
+      "loss": 1.5575,
+      "step": 3320
+    },
+    {
+      "epoch": 0.2749497655726725,
+      "grad_norm": 0.1448834389448166,
+      "learning_rate": 0.0001,
+      "loss": 1.423,
+      "step": 3321
+    },
+    {
+      "epoch": 0.275284661754856,
+      "grad_norm": 0.1444164365530014,
+      "learning_rate": 0.0001,
+      "loss": 1.4009,
+      "step": 3322
+    },
+    {
+      "epoch": 0.2756195579370395,
+      "grad_norm": 0.14190183579921722,
+      "learning_rate": 0.0001,
+      "loss": 1.4412,
+      "step": 3323
+    },
+    {
+      "epoch": 0.27595445411922304,
+      "grad_norm": 0.144227534532547,
+      "learning_rate": 0.0001,
+      "loss": 1.4297,
+      "step": 3324
+    },
+    {
+      "epoch": 0.27628935030140656,
+      "grad_norm": 0.146895632147789,
+      "learning_rate": 0.0001,
+      "loss": 1.4689,
+      "step": 3325
+    },
+    {
+      "epoch": 0.2766242464835901,
+      "grad_norm": 0.15591838955879211,
+      "learning_rate": 0.0001,
+      "loss": 1.4893,
+      "step": 3326
+    },
+    {
+      "epoch": 0.2769591426657736,
+      "grad_norm": 0.15250559151172638,
+      "learning_rate": 0.0001,
+      "loss": 1.519,
+      "step": 3327
+    },
+    {
+      "epoch": 0.2772940388479571,
+      "grad_norm": 0.14862549304962158,
+      "learning_rate": 0.0001,
+      "loss": 1.5158,
+      "step": 3328
+    },
+    {
+      "epoch": 0.27762893503014063,
+      "grad_norm": 0.1499813199043274,
+      "learning_rate": 0.0001,
+      "loss": 1.4556,
+      "step": 3329
+    },
+    {
+      "epoch": 0.27796383121232415,
+      "grad_norm": 0.15060973167419434,
+      "learning_rate": 0.0001,
+      "loss": 1.4599,
+      "step": 3330
+    },
+    {
+      "epoch": 0.27829872739450773,
+      "grad_norm": 0.14788423478603363,
+      "learning_rate": 0.0001,
+      "loss": 1.4697,
+      "step": 3331
+    },
+    {
+      "epoch": 0.27863362357669125,
+      "grad_norm": 0.14984434843063354,
+      "learning_rate": 0.0001,
+      "loss": 1.4535,
+      "step": 3332
+    },
+    {
+      "epoch": 0.27896851975887477,
+      "grad_norm": 0.15621770918369293,
+      "learning_rate": 0.0001,
+      "loss": 1.6125,
+      "step": 3333
+    },
+    {
+      "epoch": 0.2793034159410583,
+      "grad_norm": 0.15792222321033478,
+      "learning_rate": 0.0001,
+      "loss": 1.5877,
+      "step": 3334
+    },
+    {
+      "epoch": 0.2796383121232418,
+      "grad_norm": 0.15121862292289734,
+      "learning_rate": 0.0001,
+      "loss": 1.4266,
+      "step": 3335
+    },
+    {
+      "epoch": 0.2799732083054253,
+      "grad_norm": 0.15701161324977875,
+      "learning_rate": 0.0001,
+      "loss": 1.4681,
+      "step": 3336
+    },
+    {
+      "epoch": 0.28030810448760884,
+      "grad_norm": 0.14958228170871735,
+      "learning_rate": 0.0001,
+      "loss": 1.4815,
+      "step": 3337
+    },
+    {
+      "epoch": 0.28064300066979236,
+      "grad_norm": 0.14307574927806854,
+      "learning_rate": 0.0001,
+      "loss": 1.4184,
+      "step": 3338
+    },
+    {
+      "epoch": 0.2809778968519759,
+      "grad_norm": 0.1685991883277893,
+      "learning_rate": 0.0001,
+      "loss": 1.5861,
+      "step": 3339
+    },
+    {
+      "epoch": 0.2813127930341594,
+      "grad_norm": 0.144222691655159,
+      "learning_rate": 0.0001,
+      "loss": 1.4452,
+      "step": 3340
+    },
+    {
+      "epoch": 0.2816476892163429,
+      "grad_norm": 0.14686745405197144,
+      "learning_rate": 0.0001,
+      "loss": 1.4747,
+      "step": 3341
+    },
+    {
+      "epoch": 0.28198258539852644,
+      "grad_norm": 0.15471439063549042,
+      "learning_rate": 0.0001,
+      "loss": 1.4458,
+      "step": 3342
+    },
+    {
+      "epoch": 0.28231748158070996,
+      "grad_norm": 0.1510482132434845,
+      "learning_rate": 0.0001,
+      "loss": 1.4489,
+      "step": 3343
+    },
+    {
+      "epoch": 0.2826523777628935,
+      "grad_norm": 0.14976145327091217,
+      "learning_rate": 0.0001,
+      "loss": 1.3924,
+      "step": 3344
+    },
+    {
+      "epoch": 0.282987273945077,
+      "grad_norm": 0.15701331198215485,
+      "learning_rate": 0.0001,
+      "loss": 1.5704,
+      "step": 3345
+    },
+    {
+      "epoch": 0.2833221701272606,
+      "grad_norm": 0.14849096536636353,
+      "learning_rate": 0.0001,
+      "loss": 1.4541,
+      "step": 3346
+    },
+    {
+      "epoch": 0.2836570663094441,
+      "grad_norm": 0.1467583030462265,
+      "learning_rate": 0.0001,
+      "loss": 1.4102,
+      "step": 3347
+    },
+    {
+      "epoch": 0.2839919624916276,
+      "grad_norm": 0.1470215618610382,
+      "learning_rate": 0.0001,
+      "loss": 1.5082,
+      "step": 3348
+    },
+    {
+      "epoch": 0.28432685867381113,
+      "grad_norm": 0.14759418368339539,
+      "learning_rate": 0.0001,
+      "loss": 1.5136,
+      "step": 3349
+    },
+    {
+      "epoch": 0.28466175485599465,
+      "grad_norm": 0.15090933442115784,
+      "learning_rate": 0.0001,
+      "loss": 1.4585,
+      "step": 3350
+    },
+    {
+      "epoch": 0.2849966510381782,
+      "grad_norm": 0.15021245181560516,
+      "learning_rate": 0.0001,
+      "loss": 1.5347,
+      "step": 3351
+    },
+    {
+      "epoch": 0.2853315472203617,
+      "grad_norm": 0.14365991950035095,
+      "learning_rate": 0.0001,
+      "loss": 1.4526,
+      "step": 3352
+    },
+    {
+      "epoch": 0.2856664434025452,
+      "grad_norm": 0.14594535529613495,
+      "learning_rate": 0.0001,
+      "loss": 1.4779,
+      "step": 3353
+    },
+    {
+      "epoch": 0.28600133958472873,
+      "grad_norm": 0.15110114216804504,
+      "learning_rate": 0.0001,
+      "loss": 1.5095,
+      "step": 3354
+    },
+    {
+      "epoch": 0.28633623576691225,
+      "grad_norm": 0.1482629030942917,
+      "learning_rate": 0.0001,
+      "loss": 1.5056,
+      "step": 3355
+    },
+    {
+      "epoch": 0.28667113194909577,
+      "grad_norm": 0.1559736132621765,
+      "learning_rate": 0.0001,
+      "loss": 1.4846,
+      "step": 3356
+    },
+    {
+      "epoch": 0.2870060281312793,
+      "grad_norm": 0.14480571448802948,
+      "learning_rate": 0.0001,
+      "loss": 1.423,
+      "step": 3357
+    },
+    {
+      "epoch": 0.2873409243134628,
+      "grad_norm": 0.14984071254730225,
+      "learning_rate": 0.0001,
+      "loss": 1.4767,
+      "step": 3358
+    },
+    {
+      "epoch": 0.28767582049564633,
+      "grad_norm": 0.15247821807861328,
+      "learning_rate": 0.0001,
+      "loss": 1.5178,
+      "step": 3359
+    },
+    {
+      "epoch": 0.28801071667782985,
+      "grad_norm": 0.1461106687784195,
+      "learning_rate": 0.0001,
+      "loss": 1.4228,
+      "step": 3360
+    },
+    {
+      "epoch": 0.2883456128600134,
+      "grad_norm": 0.1496150642633438,
+      "learning_rate": 0.0001,
+      "loss": 1.4189,
+      "step": 3361
+    },
+    {
+      "epoch": 0.28868050904219694,
+      "grad_norm": 0.16044890880584717,
+      "learning_rate": 0.0001,
+      "loss": 1.5286,
+      "step": 3362
+    },
+    {
+      "epoch": 0.28901540522438046,
+      "grad_norm": 0.14375263452529907,
+      "learning_rate": 0.0001,
+      "loss": 1.4536,
+      "step": 3363
+    },
+    {
+      "epoch": 0.289350301406564,
+      "grad_norm": 0.14571574330329895,
+      "learning_rate": 0.0001,
+      "loss": 1.5264,
+      "step": 3364
+    },
+    {
+      "epoch": 0.2896851975887475,
+      "grad_norm": 0.14890584349632263,
+      "learning_rate": 0.0001,
+      "loss": 1.4577,
+      "step": 3365
+    },
+    {
+      "epoch": 0.290020093770931,
+      "grad_norm": 0.1530643105506897,
+      "learning_rate": 0.0001,
+      "loss": 1.5208,
+      "step": 3366
+    },
+    {
+      "epoch": 0.29035498995311454,
+      "grad_norm": 0.15293872356414795,
+      "learning_rate": 0.0001,
+      "loss": 1.4503,
+      "step": 3367
+    },
+    {
+      "epoch": 0.29068988613529806,
+      "grad_norm": 0.15317441523075104,
+      "learning_rate": 0.0001,
+      "loss": 1.5716,
+      "step": 3368
+    },
+    {
+      "epoch": 0.2910247823174816,
+      "grad_norm": 0.1484214812517166,
+      "learning_rate": 0.0001,
+      "loss": 1.4718,
+      "step": 3369
+    },
+    {
+      "epoch": 0.2913596784996651,
+      "grad_norm": 0.15597650408744812,
+      "learning_rate": 0.0001,
+      "loss": 1.4813,
+      "step": 3370
+    },
+    {
+      "epoch": 0.2916945746818486,
+      "grad_norm": 0.14535312354564667,
+      "learning_rate": 0.0001,
+      "loss": 1.5349,
+      "step": 3371
+    },
+    {
+      "epoch": 0.29202947086403214,
+      "grad_norm": 0.15586082637310028,
+      "learning_rate": 0.0001,
+      "loss": 1.5989,
+      "step": 3372
+    },
+    {
+      "epoch": 0.29236436704621566,
+      "grad_norm": 0.1462761014699936,
+      "learning_rate": 0.0001,
+      "loss": 1.3777,
+      "step": 3373
+    },
+    {
+      "epoch": 0.2926992632283992,
+      "grad_norm": 0.15477129817008972,
+      "learning_rate": 0.0001,
+      "loss": 1.5025,
+      "step": 3374
+    },
+    {
+      "epoch": 0.2930341594105827,
+      "grad_norm": 0.15040694177150726,
+      "learning_rate": 0.0001,
+      "loss": 1.4232,
+      "step": 3375
+    },
+    {
+      "epoch": 0.29336905559276627,
+      "grad_norm": 0.14610585570335388,
+      "learning_rate": 0.0001,
+      "loss": 1.4791,
+      "step": 3376
+    },
+    {
+      "epoch": 0.2937039517749498,
+      "grad_norm": 0.1511322408914566,
+      "learning_rate": 0.0001,
+      "loss": 1.5075,
+      "step": 3377
+    },
+    {
+      "epoch": 0.2940388479571333,
+      "grad_norm": 0.15154308080673218,
+      "learning_rate": 0.0001,
+      "loss": 1.5467,
+      "step": 3378
+    },
+    {
+      "epoch": 0.29437374413931683,
+      "grad_norm": 0.1455857902765274,
+      "learning_rate": 0.0001,
+      "loss": 1.5215,
+      "step": 3379
+    },
+    {
+      "epoch": 0.29470864032150035,
+      "grad_norm": 0.15827228128910065,
+      "learning_rate": 0.0001,
+      "loss": 1.5196,
+      "step": 3380
+    },
+    {
+      "epoch": 0.29504353650368387,
+      "grad_norm": 0.15383900701999664,
+      "learning_rate": 0.0001,
+      "loss": 1.561,
+      "step": 3381
+    },
+    {
+      "epoch": 0.2953784326858674,
+      "grad_norm": 0.14216770231723785,
+      "learning_rate": 0.0001,
+      "loss": 1.3841,
+      "step": 3382
+    },
+    {
+      "epoch": 0.2957133288680509,
+      "grad_norm": 0.15554247796535492,
+      "learning_rate": 0.0001,
+      "loss": 1.3912,
+      "step": 3383
+    },
+    {
+      "epoch": 0.2960482250502344,
+      "grad_norm": 0.15362423658370972,
+      "learning_rate": 0.0001,
+      "loss": 1.5188,
+      "step": 3384
+    },
+    {
+      "epoch": 0.29638312123241795,
+      "grad_norm": 0.15400175750255585,
+      "learning_rate": 0.0001,
+      "loss": 1.4673,
+      "step": 3385
+    },
+    {
+      "epoch": 0.29671801741460146,
+      "grad_norm": 0.15115594863891602,
+      "learning_rate": 0.0001,
+      "loss": 1.4775,
+      "step": 3386
+    },
+    {
+      "epoch": 0.297052913596785,
+      "grad_norm": 0.14465996623039246,
+      "learning_rate": 0.0001,
+      "loss": 1.4302,
+      "step": 3387
+    },
+    {
+      "epoch": 0.2973878097789685,
+      "grad_norm": 0.1533147394657135,
+      "learning_rate": 0.0001,
+      "loss": 1.5944,
+      "step": 3388
+    },
+    {
+      "epoch": 0.297722705961152,
+      "grad_norm": 0.1591259390115738,
+      "learning_rate": 0.0001,
+      "loss": 1.5124,
+      "step": 3389
+    },
+    {
+      "epoch": 0.29805760214333554,
+      "grad_norm": 0.15579280257225037,
+      "learning_rate": 0.0001,
+      "loss": 1.5061,
+      "step": 3390
+    },
+    {
+      "epoch": 0.29839249832551906,
+      "grad_norm": 0.15362292528152466,
+      "learning_rate": 0.0001,
+      "loss": 1.5147,
+      "step": 3391
+    },
+    {
+      "epoch": 0.29872739450770264,
+      "grad_norm": 0.16088083386421204,
+      "learning_rate": 0.0001,
+      "loss": 1.4932,
+      "step": 3392
+    },
+    {
+      "epoch": 0.29906229068988616,
+      "grad_norm": 0.15710797905921936,
+      "learning_rate": 0.0001,
+      "loss": 1.4535,
+      "step": 3393
+    },
+    {
+      "epoch": 0.2993971868720697,
+      "grad_norm": 0.14893610775470734,
+      "learning_rate": 0.0001,
+      "loss": 1.4464,
+      "step": 3394
+    },
+    {
+      "epoch": 0.2997320830542532,
+      "grad_norm": 0.1565815806388855,
+      "learning_rate": 0.0001,
+      "loss": 1.4875,
+      "step": 3395
+    },
+    {
+      "epoch": 0.3000669792364367,
+      "grad_norm": 0.14962182939052582,
+      "learning_rate": 0.0001,
+      "loss": 1.4647,
+      "step": 3396
+    },
+    {
+      "epoch": 0.30040187541862023,
+      "grad_norm": 0.14253853261470795,
+      "learning_rate": 0.0001,
+      "loss": 1.4166,
+      "step": 3397
+    },
+    {
+      "epoch": 0.30073677160080375,
+      "grad_norm": 0.13929632306098938,
+      "learning_rate": 0.0001,
+      "loss": 1.3655,
+      "step": 3398
+    },
+    {
+      "epoch": 0.3010716677829873,
+      "grad_norm": 0.1565397083759308,
+      "learning_rate": 0.0001,
+      "loss": 1.5822,
+      "step": 3399
+    },
+    {
+      "epoch": 0.3014065639651708,
+      "grad_norm": 0.15752102434635162,
+      "learning_rate": 0.0001,
+      "loss": 1.5227,
+      "step": 3400
+    },
+    {
+      "epoch": 0.3017414601473543,
+      "grad_norm": 0.15279462933540344,
+      "learning_rate": 0.0001,
+      "loss": 1.4461,
+      "step": 3401
+    },
+    {
+      "epoch": 0.30207635632953783,
+      "grad_norm": 0.14432330429553986,
+      "learning_rate": 0.0001,
+      "loss": 1.5224,
+      "step": 3402
+    },
+    {
+      "epoch": 0.30241125251172135,
+      "grad_norm": 0.1433674395084381,
+      "learning_rate": 0.0001,
+      "loss": 1.364,
+      "step": 3403
+    },
+    {
+      "epoch": 0.30274614869390487,
+      "grad_norm": 0.14757895469665527,
+      "learning_rate": 0.0001,
+      "loss": 1.5272,
+      "step": 3404
+    },
+    {
+      "epoch": 0.3030810448760884,
+      "grad_norm": 0.14946860074996948,
+      "learning_rate": 0.0001,
+      "loss": 1.4806,
+      "step": 3405
+    },
+    {
+      "epoch": 0.3034159410582719,
+      "grad_norm": 0.146403968334198,
+      "learning_rate": 0.0001,
+      "loss": 1.4675,
+      "step": 3406
+    },
+    {
+      "epoch": 0.3037508372404555,
+      "grad_norm": 0.15197888016700745,
+      "learning_rate": 0.0001,
+      "loss": 1.3588,
+      "step": 3407
+    },
+    {
+      "epoch": 0.304085733422639,
+      "grad_norm": 0.14806897938251495,
+      "learning_rate": 0.0001,
+      "loss": 1.4443,
+      "step": 3408
+    },
+    {
+      "epoch": 0.3044206296048225,
+      "grad_norm": 0.14273184537887573,
+      "learning_rate": 0.0001,
+      "loss": 1.4078,
+      "step": 3409
+    },
+    {
+      "epoch": 0.30475552578700604,
+      "grad_norm": 0.1511409729719162,
+      "learning_rate": 0.0001,
+      "loss": 1.4823,
+      "step": 3410
+    },
+    {
+      "epoch": 0.30509042196918956,
+      "grad_norm": 0.15107282996177673,
+      "learning_rate": 0.0001,
+      "loss": 1.4685,
+      "step": 3411
+    },
+    {
+      "epoch": 0.3054253181513731,
+      "grad_norm": 0.14973436295986176,
+      "learning_rate": 0.0001,
+      "loss": 1.5428,
+      "step": 3412
+    },
+    {
+      "epoch": 0.3057602143335566,
+      "grad_norm": 0.15601438283920288,
+      "learning_rate": 0.0001,
+      "loss": 1.5471,
+      "step": 3413
+    },
+    {
+      "epoch": 0.3060951105157401,
+      "grad_norm": 0.15377770364284515,
+      "learning_rate": 0.0001,
+      "loss": 1.4574,
+      "step": 3414
+    },
+    {
+      "epoch": 0.30643000669792364,
+      "grad_norm": 0.14803405106067657,
+      "learning_rate": 0.0001,
+      "loss": 1.4975,
+      "step": 3415
+    },
+    {
+      "epoch": 0.30676490288010716,
+      "grad_norm": 0.15249978005886078,
+      "learning_rate": 0.0001,
+      "loss": 1.3705,
+      "step": 3416
+    },
+    {
+      "epoch": 0.3070997990622907,
+      "grad_norm": 0.15126726031303406,
+      "learning_rate": 0.0001,
+      "loss": 1.4297,
+      "step": 3417
+    },
+    {
+      "epoch": 0.3074346952444742,
+      "grad_norm": 0.14738833904266357,
+      "learning_rate": 0.0001,
+      "loss": 1.3943,
+      "step": 3418
+    },
+    {
+      "epoch": 0.3077695914266577,
+      "grad_norm": 0.1609349548816681,
+      "learning_rate": 0.0001,
+      "loss": 1.491,
+      "step": 3419
+    },
+    {
+      "epoch": 0.30810448760884124,
+      "grad_norm": 0.14774426817893982,
+      "learning_rate": 0.0001,
+      "loss": 1.4311,
+      "step": 3420
+    },
+    {
+      "epoch": 0.30843938379102476,
+      "grad_norm": 0.1546633392572403,
+      "learning_rate": 0.0001,
+      "loss": 1.4578,
+      "step": 3421
+    },
+    {
+      "epoch": 0.30877427997320833,
+      "grad_norm": 0.15193051099777222,
+      "learning_rate": 0.0001,
+      "loss": 1.4649,
+      "step": 3422
+    },
+    {
+      "epoch": 0.30910917615539185,
+      "grad_norm": 0.15336847305297852,
+      "learning_rate": 0.0001,
+      "loss": 1.6061,
+      "step": 3423
+    },
+    {
+      "epoch": 0.30944407233757537,
+      "grad_norm": 0.14489242434501648,
+      "learning_rate": 0.0001,
+      "loss": 1.468,
+      "step": 3424
+    },
+    {
+      "epoch": 0.3097789685197589,
+      "grad_norm": 0.15843623876571655,
+      "learning_rate": 0.0001,
+      "loss": 1.4897,
+      "step": 3425
+    },
+    {
+      "epoch": 0.3101138647019424,
+      "grad_norm": 0.15312214195728302,
+      "learning_rate": 0.0001,
+      "loss": 1.4869,
+      "step": 3426
+    },
+    {
+      "epoch": 0.31044876088412593,
+      "grad_norm": 0.15313860774040222,
+      "learning_rate": 0.0001,
+      "loss": 1.4169,
+      "step": 3427
+    },
+    {
+      "epoch": 0.31078365706630945,
+      "grad_norm": 0.15683285892009735,
+      "learning_rate": 0.0001,
+      "loss": 1.4916,
+      "step": 3428
+    },
+    {
+      "epoch": 0.31111855324849297,
+      "grad_norm": 0.15352609753608704,
+      "learning_rate": 0.0001,
+      "loss": 1.4138,
+      "step": 3429
+    },
+    {
+      "epoch": 0.3114534494306765,
+      "grad_norm": 0.16370314359664917,
+      "learning_rate": 0.0001,
+      "loss": 1.5103,
+      "step": 3430
+    },
+    {
+      "epoch": 0.31178834561286,
+      "grad_norm": 0.15114983916282654,
+      "learning_rate": 0.0001,
+      "loss": 1.4491,
+      "step": 3431
+    },
+    {
+      "epoch": 0.3121232417950435,
+      "grad_norm": 0.15531513094902039,
+      "learning_rate": 0.0001,
+      "loss": 1.5423,
+      "step": 3432
+    },
+    {
+      "epoch": 0.31245813797722705,
+      "grad_norm": 0.14949560165405273,
+      "learning_rate": 0.0001,
+      "loss": 1.4947,
+      "step": 3433
+    },
+    {
+      "epoch": 0.31279303415941057,
+      "grad_norm": 0.14659979939460754,
+      "learning_rate": 0.0001,
+      "loss": 1.4453,
+      "step": 3434
+    },
+    {
+      "epoch": 0.3131279303415941,
+      "grad_norm": 0.147125706076622,
+      "learning_rate": 0.0001,
+      "loss": 1.474,
+      "step": 3435
+    },
+    {
+      "epoch": 0.3134628265237776,
+      "grad_norm": 0.1630190759897232,
+      "learning_rate": 0.0001,
+      "loss": 1.5404,
+      "step": 3436
+    },
+    {
+      "epoch": 0.3137977227059612,
+      "grad_norm": 0.1427680253982544,
+      "learning_rate": 0.0001,
+      "loss": 1.3876,
+      "step": 3437
+    },
+    {
+      "epoch": 0.3141326188881447,
+      "grad_norm": 0.15487343072891235,
+      "learning_rate": 0.0001,
+      "loss": 1.4852,
+      "step": 3438
+    },
+    {
+      "epoch": 0.3144675150703282,
+      "grad_norm": 0.15448251366615295,
+      "learning_rate": 0.0001,
+      "loss": 1.4936,
+      "step": 3439
+    },
+    {
+      "epoch": 0.31480241125251174,
+      "grad_norm": 0.14923791587352753,
+      "learning_rate": 0.0001,
+      "loss": 1.5198,
+      "step": 3440
+    },
+    {
+      "epoch": 0.31513730743469526,
+      "grad_norm": 0.15218065679073334,
+      "learning_rate": 0.0001,
+      "loss": 1.4689,
+      "step": 3441
+    },
+    {
+      "epoch": 0.3154722036168788,
+      "grad_norm": 0.15223555266857147,
+      "learning_rate": 0.0001,
+      "loss": 1.5374,
+      "step": 3442
+    },
+    {
+      "epoch": 0.3158070997990623,
+      "grad_norm": 0.14683569967746735,
+      "learning_rate": 0.0001,
+      "loss": 1.5073,
+      "step": 3443
+    },
+    {
+      "epoch": 0.3161419959812458,
+      "grad_norm": 0.152296245098114,
+      "learning_rate": 0.0001,
+      "loss": 1.3803,
+      "step": 3444
+    },
+    {
+      "epoch": 0.31647689216342934,
+      "grad_norm": 0.149747833609581,
+      "learning_rate": 0.0001,
+      "loss": 1.4896,
+      "step": 3445
+    },
+    {
+      "epoch": 0.31681178834561285,
+      "grad_norm": 0.153005912899971,
+      "learning_rate": 0.0001,
+      "loss": 1.5106,
+      "step": 3446
+    },
+    {
+      "epoch": 0.3171466845277964,
+      "grad_norm": 0.15335361659526825,
+      "learning_rate": 0.0001,
+      "loss": 1.4979,
+      "step": 3447
+    },
+    {
+      "epoch": 0.3174815807099799,
+      "grad_norm": 0.15364506840705872,
+      "learning_rate": 0.0001,
+      "loss": 1.5059,
+      "step": 3448
+    },
+    {
+      "epoch": 0.3178164768921634,
+      "grad_norm": 0.15296977758407593,
+      "learning_rate": 0.0001,
+      "loss": 1.5079,
+      "step": 3449
+    },
+    {
+      "epoch": 0.31815137307434693,
+      "grad_norm": 0.1484413892030716,
+      "learning_rate": 0.0001,
+      "loss": 1.3828,
+      "step": 3450
+    },
+    {
+      "epoch": 0.31848626925653045,
+      "grad_norm": 0.16037903726100922,
+      "learning_rate": 0.0001,
+      "loss": 1.4062,
+      "step": 3451
+    },
+    {
+      "epoch": 0.31882116543871397,
+      "grad_norm": 0.15443116426467896,
+      "learning_rate": 0.0001,
+      "loss": 1.4373,
+      "step": 3452
+    },
+    {
+      "epoch": 0.31915606162089755,
+      "grad_norm": 0.1552223265171051,
+      "learning_rate": 0.0001,
+      "loss": 1.5142,
+      "step": 3453
+    },
+    {
+      "epoch": 0.31949095780308107,
+      "grad_norm": 0.14638473093509674,
+      "learning_rate": 0.0001,
+      "loss": 1.439,
+      "step": 3454
+    },
+    {
+      "epoch": 0.3198258539852646,
+      "grad_norm": 0.14468824863433838,
+      "learning_rate": 0.0001,
+      "loss": 1.401,
+      "step": 3455
+    },
+    {
+      "epoch": 0.3201607501674481,
+      "grad_norm": 0.15405170619487762,
+      "learning_rate": 0.0001,
+      "loss": 1.4824,
+      "step": 3456
+    },
+    {
+      "epoch": 0.3204956463496316,
+      "grad_norm": 0.15160967409610748,
+      "learning_rate": 0.0001,
+      "loss": 1.4575,
+      "step": 3457
+    },
+    {
+      "epoch": 0.32083054253181514,
+      "grad_norm": 0.1501069813966751,
+      "learning_rate": 0.0001,
+      "loss": 1.4433,
+      "step": 3458
+    },
+    {
+      "epoch": 0.32116543871399866,
+      "grad_norm": 0.15485960245132446,
+      "learning_rate": 0.0001,
+      "loss": 1.4768,
+      "step": 3459
+    },
+    {
+      "epoch": 0.3215003348961822,
+      "grad_norm": 0.16001489758491516,
+      "learning_rate": 0.0001,
+      "loss": 1.456,
+      "step": 3460
+    },
+    {
+      "epoch": 0.3218352310783657,
+      "grad_norm": 0.15336276590824127,
+      "learning_rate": 0.0001,
+      "loss": 1.5219,
+      "step": 3461
+    },
+    {
+      "epoch": 0.3221701272605492,
+      "grad_norm": 0.15447048842906952,
+      "learning_rate": 0.0001,
+      "loss": 1.4807,
+      "step": 3462
+    },
+    {
+      "epoch": 0.32250502344273274,
+      "grad_norm": 0.15134736895561218,
+      "learning_rate": 0.0001,
+      "loss": 1.4368,
+      "step": 3463
+    },
+    {
+      "epoch": 0.32283991962491626,
+      "grad_norm": 0.1590096652507782,
+      "learning_rate": 0.0001,
+      "loss": 1.4353,
+      "step": 3464
+    },
+    {
+      "epoch": 0.3231748158070998,
+      "grad_norm": 0.15310245752334595,
+      "learning_rate": 0.0001,
+      "loss": 1.5494,
+      "step": 3465
+    },
+    {
+      "epoch": 0.3235097119892833,
+      "grad_norm": 0.14869531989097595,
+      "learning_rate": 0.0001,
+      "loss": 1.4673,
+      "step": 3466
+    },
+    {
+      "epoch": 0.3238446081714668,
+      "grad_norm": 0.15101325511932373,
+      "learning_rate": 0.0001,
+      "loss": 1.4633,
+      "step": 3467
+    },
+    {
+      "epoch": 0.3241795043536504,
+      "grad_norm": 0.1566777378320694,
+      "learning_rate": 0.0001,
+      "loss": 1.5121,
+      "step": 3468
+    },
+    {
+      "epoch": 0.3245144005358339,
+      "grad_norm": 0.1482858955860138,
+      "learning_rate": 0.0001,
+      "loss": 1.3693,
+      "step": 3469
+    },
+    {
+      "epoch": 0.32484929671801743,
+      "grad_norm": 0.15713654458522797,
+      "learning_rate": 0.0001,
+      "loss": 1.5532,
+      "step": 3470
+    },
+    {
+      "epoch": 0.32518419290020095,
+      "grad_norm": 0.15125976502895355,
+      "learning_rate": 0.0001,
+      "loss": 1.5104,
+      "step": 3471
+    },
+    {
+      "epoch": 0.32551908908238447,
+      "grad_norm": 0.1499818116426468,
+      "learning_rate": 0.0001,
+      "loss": 1.378,
+      "step": 3472
+    },
+    {
+      "epoch": 0.325853985264568,
+      "grad_norm": 0.16104479134082794,
+      "learning_rate": 0.0001,
+      "loss": 1.6239,
+      "step": 3473
+    },
+    {
+      "epoch": 0.3261888814467515,
+      "grad_norm": 0.15305879712104797,
+      "learning_rate": 0.0001,
+      "loss": 1.4882,
+      "step": 3474
+    },
+    {
+      "epoch": 0.32652377762893503,
+      "grad_norm": 0.14504723250865936,
+      "learning_rate": 0.0001,
+      "loss": 1.4882,
+      "step": 3475
+    },
+    {
+      "epoch": 0.32685867381111855,
+      "grad_norm": 0.15625393390655518,
+      "learning_rate": 0.0001,
+      "loss": 1.546,
+      "step": 3476
+    },
+    {
+      "epoch": 0.32719356999330207,
+      "grad_norm": 0.1602199822664261,
+      "learning_rate": 0.0001,
+      "loss": 1.5666,
+      "step": 3477
+    },
+    {
+      "epoch": 0.3275284661754856,
+      "grad_norm": 0.14354944229125977,
+      "learning_rate": 0.0001,
+      "loss": 1.4286,
+      "step": 3478
+    },
+    {
+      "epoch": 0.3278633623576691,
+      "grad_norm": 0.15154190361499786,
+      "learning_rate": 0.0001,
+      "loss": 1.4966,
+      "step": 3479
+    },
+    {
+      "epoch": 0.3281982585398526,
+      "grad_norm": 0.15154610574245453,
+      "learning_rate": 0.0001,
+      "loss": 1.5694,
+      "step": 3480
+    },
+    {
+      "epoch": 0.32853315472203615,
+      "grad_norm": 0.15371978282928467,
+      "learning_rate": 0.0001,
+      "loss": 1.4273,
+      "step": 3481
+    },
+    {
+      "epoch": 0.32886805090421967,
+      "grad_norm": 0.15090040862560272,
+      "learning_rate": 0.0001,
+      "loss": 1.4088,
+      "step": 3482
+    },
+    {
+      "epoch": 0.32920294708640324,
+      "grad_norm": 0.15387658774852753,
+      "learning_rate": 0.0001,
+      "loss": 1.5402,
+      "step": 3483
+    },
+    {
+      "epoch": 0.32953784326858676,
+      "grad_norm": 0.15785494446754456,
+      "learning_rate": 0.0001,
+      "loss": 1.5964,
+      "step": 3484
+    },
+    {
+      "epoch": 0.3298727394507703,
+      "grad_norm": 0.15188738703727722,
+      "learning_rate": 0.0001,
+      "loss": 1.527,
+      "step": 3485
+    },
+    {
+      "epoch": 0.3302076356329538,
+      "grad_norm": 0.14585669338703156,
+      "learning_rate": 0.0001,
+      "loss": 1.3841,
+      "step": 3486
+    },
+    {
+      "epoch": 0.3305425318151373,
+      "grad_norm": 0.15196645259857178,
+      "learning_rate": 0.0001,
+      "loss": 1.3851,
+      "step": 3487
+    },
+    {
+      "epoch": 0.33087742799732084,
+      "grad_norm": 0.1498725563287735,
+      "learning_rate": 0.0001,
+      "loss": 1.5218,
+      "step": 3488
+    },
+    {
+      "epoch": 0.33121232417950436,
+      "grad_norm": 0.15197834372520447,
+      "learning_rate": 0.0001,
+      "loss": 1.4579,
+      "step": 3489
+    },
+    {
+      "epoch": 0.3315472203616879,
+      "grad_norm": 0.15742060542106628,
+      "learning_rate": 0.0001,
+      "loss": 1.5561,
+      "step": 3490
+    },
+    {
+      "epoch": 0.3318821165438714,
+      "grad_norm": 0.15078939497470856,
+      "learning_rate": 0.0001,
+      "loss": 1.4181,
+      "step": 3491
+    },
+    {
+      "epoch": 0.3322170127260549,
+      "grad_norm": 0.1436023712158203,
+      "learning_rate": 0.0001,
+      "loss": 1.4586,
+      "step": 3492
+    },
+    {
+      "epoch": 0.33255190890823844,
+      "grad_norm": 0.14628535509109497,
+      "learning_rate": 0.0001,
+      "loss": 1.4207,
+      "step": 3493
+    },
+    {
+      "epoch": 0.33288680509042196,
+      "grad_norm": 0.16081663966178894,
+      "learning_rate": 0.0001,
+      "loss": 1.5292,
+      "step": 3494
+    },
+    {
+      "epoch": 0.3332217012726055,
+      "grad_norm": 0.1588294953107834,
+      "learning_rate": 0.0001,
+      "loss": 1.5263,
+      "step": 3495
+    },
+    {
+      "epoch": 0.333556597454789,
+      "grad_norm": 0.14988917112350464,
+      "learning_rate": 0.0001,
+      "loss": 1.4401,
+      "step": 3496
+    },
+    {
+      "epoch": 0.3338914936369725,
+      "grad_norm": 0.17191949486732483,
+      "learning_rate": 0.0001,
+      "loss": 1.6032,
+      "step": 3497
+    },
+    {
+      "epoch": 0.3342263898191561,
+      "grad_norm": 0.14784646034240723,
+      "learning_rate": 0.0001,
+      "loss": 1.4044,
+      "step": 3498
+    },
+    {
+      "epoch": 0.3345612860013396,
+      "grad_norm": 0.1579027622938156,
+      "learning_rate": 0.0001,
+      "loss": 1.4487,
+      "step": 3499
+    },
+    {
+      "epoch": 0.33489618218352313,
+      "grad_norm": 0.15323099493980408,
+      "learning_rate": 0.0001,
+      "loss": 1.4331,
+      "step": 3500
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 5972,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 2,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.986928858418381e+19,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}