diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,24969 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.9991577765300392,
+  "eval_steps": 500,
+  "global_step": 3561,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0008422234699606962,
+      "grad_norm": 2.7139885425567627,
+      "learning_rate": 2.8011204481792718e-08,
+      "loss": 0.7099,
+      "step": 1
+    },
+    {
+      "epoch": 0.0016844469399213925,
+      "grad_norm": 2.5785422325134277,
+      "learning_rate": 5.6022408963585437e-08,
+      "loss": 0.7135,
+      "step": 2
+    },
+    {
+      "epoch": 0.0025266704098820887,
+      "grad_norm": 2.7374205589294434,
+      "learning_rate": 8.403361344537815e-08,
+      "loss": 0.7394,
+      "step": 3
+    },
+    {
+      "epoch": 0.003368893879842785,
+      "grad_norm": 2.7776577472686768,
+      "learning_rate": 1.1204481792717087e-07,
+      "loss": 0.7385,
+      "step": 4
+    },
+    {
+      "epoch": 0.004211117349803481,
+      "grad_norm": 2.700944662094116,
+      "learning_rate": 1.400560224089636e-07,
+      "loss": 0.753,
+      "step": 5
+    },
+    {
+      "epoch": 0.0050533408197641775,
+      "grad_norm": 2.9223380088806152,
+      "learning_rate": 1.680672268907563e-07,
+      "loss": 0.7637,
+      "step": 6
+    },
+    {
+      "epoch": 0.005895564289724873,
+      "grad_norm": 2.558424234390259,
+      "learning_rate": 1.9607843137254904e-07,
+      "loss": 0.7837,
+      "step": 7
+    },
+    {
+      "epoch": 0.00673778775968557,
+      "grad_norm": 2.620910882949829,
+      "learning_rate": 2.2408963585434175e-07,
+      "loss": 0.7524,
+      "step": 8
+    },
+    {
+      "epoch": 0.007580011229646266,
+      "grad_norm": 2.769366979598999,
+      "learning_rate": 2.5210084033613445e-07,
+      "loss": 0.7467,
+      "step": 9
+    },
+    {
+      "epoch": 0.008422234699606962,
+      "grad_norm": 2.6973936557769775,
+      "learning_rate": 2.801120448179272e-07,
+      "loss": 0.7416,
+      "step": 10
+    },
+    {
+      "epoch": 0.009264458169567658,
+      "grad_norm": 2.6612558364868164,
+      "learning_rate": 3.081232492997199e-07,
+      "loss": 0.7446,
+      "step": 11
+    },
+    {
+      "epoch": 0.010106681639528355,
+      "grad_norm": 2.729600667953491,
+      "learning_rate": 3.361344537815126e-07,
+      "loss": 0.7453,
+      "step": 12
+    },
+    {
+      "epoch": 0.010948905109489052,
+      "grad_norm": 2.687103271484375,
+      "learning_rate": 3.641456582633054e-07,
+      "loss": 0.7578,
+      "step": 13
+    },
+    {
+      "epoch": 0.011791128579449747,
+      "grad_norm": 2.611377239227295,
+      "learning_rate": 3.921568627450981e-07,
+      "loss": 0.7663,
+      "step": 14
+    },
+    {
+      "epoch": 0.012633352049410443,
+      "grad_norm": 2.788295030593872,
+      "learning_rate": 4.201680672268908e-07,
+      "loss": 0.7648,
+      "step": 15
+    },
+    {
+      "epoch": 0.01347557551937114,
+      "grad_norm": 2.528625726699829,
+      "learning_rate": 4.481792717086835e-07,
+      "loss": 0.6989,
+      "step": 16
+    },
+    {
+      "epoch": 0.014317798989331837,
+      "grad_norm": 2.6373348236083984,
+      "learning_rate": 4.7619047619047623e-07,
+      "loss": 0.7643,
+      "step": 17
+    },
+    {
+      "epoch": 0.015160022459292532,
+      "grad_norm": 2.454690456390381,
+      "learning_rate": 5.042016806722689e-07,
+      "loss": 0.7315,
+      "step": 18
+    },
+    {
+      "epoch": 0.016002245929253228,
+      "grad_norm": 2.1451756954193115,
+      "learning_rate": 5.322128851540616e-07,
+      "loss": 0.7307,
+      "step": 19
+    },
+    {
+      "epoch": 0.016844469399213923,
+      "grad_norm": 2.230375289916992,
+      "learning_rate": 5.602240896358544e-07,
+      "loss": 0.7181,
+      "step": 20
+    },
+    {
+      "epoch": 0.01768669286917462,
+      "grad_norm": 2.100219488143921,
+      "learning_rate": 5.882352941176471e-07,
+      "loss": 0.7203,
+      "step": 21
+    },
+    {
+      "epoch": 0.018528916339135316,
+      "grad_norm": 1.9923067092895508,
+      "learning_rate": 6.162464985994398e-07,
+      "loss": 0.7079,
+      "step": 22
+    },
+    {
+      "epoch": 0.019371139809096015,
+      "grad_norm": 1.9194575548171997,
+      "learning_rate": 6.442577030812325e-07,
+      "loss": 0.6887,
+      "step": 23
+    },
+    {
+      "epoch": 0.02021336327905671,
+      "grad_norm": 1.9598820209503174,
+      "learning_rate": 6.722689075630252e-07,
+      "loss": 0.7282,
+      "step": 24
+    },
+    {
+      "epoch": 0.021055586749017405,
+      "grad_norm": 1.6673228740692139,
+      "learning_rate": 7.002801120448179e-07,
+      "loss": 0.7221,
+      "step": 25
+    },
+    {
+      "epoch": 0.021897810218978103,
+      "grad_norm": 1.5177627801895142,
+      "learning_rate": 7.282913165266108e-07,
+      "loss": 0.7328,
+      "step": 26
+    },
+    {
+      "epoch": 0.022740033688938798,
+      "grad_norm": 1.3995906114578247,
+      "learning_rate": 7.563025210084034e-07,
+      "loss": 0.6929,
+      "step": 27
+    },
+    {
+      "epoch": 0.023582257158899493,
+      "grad_norm": 1.4189306497573853,
+      "learning_rate": 7.843137254901962e-07,
+      "loss": 0.6901,
+      "step": 28
+    },
+    {
+      "epoch": 0.02442448062886019,
+      "grad_norm": 1.362115502357483,
+      "learning_rate": 8.123249299719889e-07,
+      "loss": 0.6601,
+      "step": 29
+    },
+    {
+      "epoch": 0.025266704098820886,
+      "grad_norm": 1.3577219247817993,
+      "learning_rate": 8.403361344537816e-07,
+      "loss": 0.7119,
+      "step": 30
+    },
+    {
+      "epoch": 0.026108927568781585,
+      "grad_norm": 1.3313069343566895,
+      "learning_rate": 8.683473389355742e-07,
+      "loss": 0.6776,
+      "step": 31
+    },
+    {
+      "epoch": 0.02695115103874228,
+      "grad_norm": 1.2647556066513062,
+      "learning_rate": 8.96358543417367e-07,
+      "loss": 0.6799,
+      "step": 32
+    },
+    {
+      "epoch": 0.027793374508702975,
+      "grad_norm": 1.2313024997711182,
+      "learning_rate": 9.243697478991598e-07,
+      "loss": 0.6693,
+      "step": 33
+    },
+    {
+      "epoch": 0.028635597978663673,
+      "grad_norm": 1.1124472618103027,
+      "learning_rate": 9.523809523809525e-07,
+      "loss": 0.6515,
+      "step": 34
+    },
+    {
+      "epoch": 0.029477821448624368,
+      "grad_norm": 1.0712497234344482,
+      "learning_rate": 9.80392156862745e-07,
+      "loss": 0.6746,
+      "step": 35
+    },
+    {
+      "epoch": 0.030320044918585063,
+      "grad_norm": 1.0045796632766724,
+      "learning_rate": 1.0084033613445378e-06,
+      "loss": 0.6326,
+      "step": 36
+    },
+    {
+      "epoch": 0.03116226838854576,
+      "grad_norm": 0.9366388320922852,
+      "learning_rate": 1.0364145658263308e-06,
+      "loss": 0.6184,
+      "step": 37
+    },
+    {
+      "epoch": 0.032004491858506456,
+      "grad_norm": 0.9612988233566284,
+      "learning_rate": 1.0644257703081233e-06,
+      "loss": 0.6551,
+      "step": 38
+    },
+    {
+      "epoch": 0.032846715328467155,
+      "grad_norm": 0.9342404007911682,
+      "learning_rate": 1.092436974789916e-06,
+      "loss": 0.6553,
+      "step": 39
+    },
+    {
+      "epoch": 0.033688938798427846,
+      "grad_norm": 0.8526981472969055,
+      "learning_rate": 1.1204481792717088e-06,
+      "loss": 0.6212,
+      "step": 40
+    },
+    {
+      "epoch": 0.034531162268388545,
+      "grad_norm": 0.8903021216392517,
+      "learning_rate": 1.1484593837535015e-06,
+      "loss": 0.6381,
+      "step": 41
+    },
+    {
+      "epoch": 0.03537338573834924,
+      "grad_norm": 0.8123112320899963,
+      "learning_rate": 1.1764705882352942e-06,
+      "loss": 0.6237,
+      "step": 42
+    },
+    {
+      "epoch": 0.03621560920830994,
+      "grad_norm": 0.8697153925895691,
+      "learning_rate": 1.204481792717087e-06,
+      "loss": 0.637,
+      "step": 43
+    },
+    {
+      "epoch": 0.03705783267827063,
+      "grad_norm": 0.7646113038063049,
+      "learning_rate": 1.2324929971988797e-06,
+      "loss": 0.6099,
+      "step": 44
+    },
+    {
+      "epoch": 0.03790005614823133,
+      "grad_norm": 0.7301273941993713,
+      "learning_rate": 1.2605042016806724e-06,
+      "loss": 0.6307,
+      "step": 45
+    },
+    {
+      "epoch": 0.03874227961819203,
+      "grad_norm": 0.6388236880302429,
+      "learning_rate": 1.288515406162465e-06,
+      "loss": 0.6004,
+      "step": 46
+    },
+    {
+      "epoch": 0.03958450308815272,
+      "grad_norm": 0.6535987854003906,
+      "learning_rate": 1.316526610644258e-06,
+      "loss": 0.6248,
+      "step": 47
+    },
+    {
+      "epoch": 0.04042672655811342,
+      "grad_norm": 0.5928495526313782,
+      "learning_rate": 1.3445378151260504e-06,
+      "loss": 0.5803,
+      "step": 48
+    },
+    {
+      "epoch": 0.04126895002807412,
+      "grad_norm": 0.6195793151855469,
+      "learning_rate": 1.3725490196078434e-06,
+      "loss": 0.5997,
+      "step": 49
+    },
+    {
+      "epoch": 0.04211117349803481,
+      "grad_norm": 0.5354277491569519,
+      "learning_rate": 1.4005602240896359e-06,
+      "loss": 0.5537,
+      "step": 50
+    },
+    {
+      "epoch": 0.04295339696799551,
+      "grad_norm": 0.5489490032196045,
+      "learning_rate": 1.4285714285714286e-06,
+      "loss": 0.5818,
+      "step": 51
+    },
+    {
+      "epoch": 0.043795620437956206,
+      "grad_norm": 0.5248365998268127,
+      "learning_rate": 1.4565826330532216e-06,
+      "loss": 0.5803,
+      "step": 52
+    },
+    {
+      "epoch": 0.0446378439079169,
+      "grad_norm": 0.5383778214454651,
+      "learning_rate": 1.484593837535014e-06,
+      "loss": 0.6108,
+      "step": 53
+    },
+    {
+      "epoch": 0.045480067377877596,
+      "grad_norm": 0.5217686295509338,
+      "learning_rate": 1.5126050420168068e-06,
+      "loss": 0.5962,
+      "step": 54
+    },
+    {
+      "epoch": 0.046322290847838295,
+      "grad_norm": 0.5311843752861023,
+      "learning_rate": 1.5406162464985996e-06,
+      "loss": 0.5983,
+      "step": 55
+    },
+    {
+      "epoch": 0.047164514317798986,
+      "grad_norm": 0.5228126645088196,
+      "learning_rate": 1.5686274509803923e-06,
+      "loss": 0.5789,
+      "step": 56
+    },
+    {
+      "epoch": 0.048006737787759685,
+      "grad_norm": 0.4893929958343506,
+      "learning_rate": 1.5966386554621848e-06,
+      "loss": 0.5695,
+      "step": 57
+    },
+    {
+      "epoch": 0.04884896125772038,
+      "grad_norm": 0.46164101362228394,
+      "learning_rate": 1.6246498599439778e-06,
+      "loss": 0.5611,
+      "step": 58
+    },
+    {
+      "epoch": 0.04969118472768108,
+      "grad_norm": 0.42507484555244446,
+      "learning_rate": 1.6526610644257705e-06,
+      "loss": 0.5262,
+      "step": 59
+    },
+    {
+      "epoch": 0.05053340819764177,
+      "grad_norm": 0.41358301043510437,
+      "learning_rate": 1.6806722689075632e-06,
+      "loss": 0.5948,
+      "step": 60
+    },
+    {
+      "epoch": 0.05137563166760247,
+      "grad_norm": 0.3956189751625061,
+      "learning_rate": 1.708683473389356e-06,
+      "loss": 0.5569,
+      "step": 61
+    },
+    {
+      "epoch": 0.05221785513756317,
+      "grad_norm": 0.41548147797584534,
+      "learning_rate": 1.7366946778711485e-06,
+      "loss": 0.5444,
+      "step": 62
+    },
+    {
+      "epoch": 0.05306007860752386,
+      "grad_norm": 0.4024750888347626,
+      "learning_rate": 1.7647058823529414e-06,
+      "loss": 0.5549,
+      "step": 63
+    },
+    {
+      "epoch": 0.05390230207748456,
+      "grad_norm": 0.36441898345947266,
+      "learning_rate": 1.792717086834734e-06,
+      "loss": 0.5219,
+      "step": 64
+    },
+    {
+      "epoch": 0.05474452554744526,
+      "grad_norm": 0.37034735083580017,
+      "learning_rate": 1.8207282913165267e-06,
+      "loss": 0.5264,
+      "step": 65
+    },
+    {
+      "epoch": 0.05558674901740595,
+      "grad_norm": 0.34774619340896606,
+      "learning_rate": 1.8487394957983196e-06,
+      "loss": 0.5511,
+      "step": 66
+    },
+    {
+      "epoch": 0.05642897248736665,
+      "grad_norm": 0.3381839096546173,
+      "learning_rate": 1.8767507002801122e-06,
+      "loss": 0.5242,
+      "step": 67
+    },
+    {
+      "epoch": 0.057271195957327346,
+      "grad_norm": 0.34560632705688477,
+      "learning_rate": 1.904761904761905e-06,
+      "loss": 0.5143,
+      "step": 68
+    },
+    {
+      "epoch": 0.05811341942728804,
+      "grad_norm": 0.35959887504577637,
+      "learning_rate": 1.932773109243698e-06,
+      "loss": 0.5333,
+      "step": 69
+    },
+    {
+      "epoch": 0.058955642897248736,
+      "grad_norm": 0.3450910449028015,
+      "learning_rate": 1.96078431372549e-06,
+      "loss": 0.5181,
+      "step": 70
+    },
+    {
+      "epoch": 0.059797866367209435,
+      "grad_norm": 0.3333073854446411,
+      "learning_rate": 1.988795518207283e-06,
+      "loss": 0.5358,
+      "step": 71
+    },
+    {
+      "epoch": 0.060640089837170126,
+      "grad_norm": 0.34593814611434937,
+      "learning_rate": 2.0168067226890756e-06,
+      "loss": 0.5351,
+      "step": 72
+    },
+    {
+      "epoch": 0.061482313307130824,
+      "grad_norm": 0.30022236704826355,
+      "learning_rate": 2.0448179271708684e-06,
+      "loss": 0.5296,
+      "step": 73
+    },
+    {
+      "epoch": 0.06232453677709152,
+      "grad_norm": 0.31735390424728394,
+      "learning_rate": 2.0728291316526615e-06,
+      "loss": 0.5198,
+      "step": 74
+    },
+    {
+      "epoch": 0.06316676024705221,
+      "grad_norm": 0.40318942070007324,
+      "learning_rate": 2.100840336134454e-06,
+      "loss": 0.5037,
+      "step": 75
+    },
+    {
+      "epoch": 0.06400898371701291,
+      "grad_norm": 0.2909386456012726,
+      "learning_rate": 2.1288515406162466e-06,
+      "loss": 0.4936,
+      "step": 76
+    },
+    {
+      "epoch": 0.06485120718697361,
+      "grad_norm": 0.30838459730148315,
+      "learning_rate": 2.1568627450980393e-06,
+      "loss": 0.5073,
+      "step": 77
+    },
+    {
+      "epoch": 0.06569343065693431,
+      "grad_norm": 0.2972123324871063,
+      "learning_rate": 2.184873949579832e-06,
+      "loss": 0.5445,
+      "step": 78
+    },
+    {
+      "epoch": 0.06653565412689501,
+      "grad_norm": 0.29268571734428406,
+      "learning_rate": 2.2128851540616248e-06,
+      "loss": 0.5006,
+      "step": 79
+    },
+    {
+      "epoch": 0.06737787759685569,
+      "grad_norm": 0.2976105213165283,
+      "learning_rate": 2.2408963585434175e-06,
+      "loss": 0.5266,
+      "step": 80
+    },
+    {
+      "epoch": 0.06822010106681639,
+      "grad_norm": 0.2998995780944824,
+      "learning_rate": 2.2689075630252102e-06,
+      "loss": 0.5093,
+      "step": 81
+    },
+    {
+      "epoch": 0.06906232453677709,
+      "grad_norm": 0.3065139055252075,
+      "learning_rate": 2.296918767507003e-06,
+      "loss": 0.5078,
+      "step": 82
+    },
+    {
+      "epoch": 0.06990454800673779,
+      "grad_norm": 0.2579464614391327,
+      "learning_rate": 2.3249299719887957e-06,
+      "loss": 0.484,
+      "step": 83
+    },
+    {
+      "epoch": 0.07074677147669849,
+      "grad_norm": 0.28767305612564087,
+      "learning_rate": 2.3529411764705885e-06,
+      "loss": 0.5066,
+      "step": 84
+    },
+    {
+      "epoch": 0.07158899494665918,
+      "grad_norm": 0.27491238713264465,
+      "learning_rate": 2.380952380952381e-06,
+      "loss": 0.4865,
+      "step": 85
+    },
+    {
+      "epoch": 0.07243121841661988,
+      "grad_norm": 0.2776690721511841,
+      "learning_rate": 2.408963585434174e-06,
+      "loss": 0.5279,
+      "step": 86
+    },
+    {
+      "epoch": 0.07327344188658057,
+      "grad_norm": 0.2866682708263397,
+      "learning_rate": 2.4369747899159667e-06,
+      "loss": 0.506,
+      "step": 87
+    },
+    {
+      "epoch": 0.07411566535654127,
+      "grad_norm": 0.27375006675720215,
+      "learning_rate": 2.4649859943977594e-06,
+      "loss": 0.5203,
+      "step": 88
+    },
+    {
+      "epoch": 0.07495788882650196,
+      "grad_norm": 0.2791830897331238,
+      "learning_rate": 2.492997198879552e-06,
+      "loss": 0.5098,
+      "step": 89
+    },
+    {
+      "epoch": 0.07580011229646266,
+      "grad_norm": 0.2743302881717682,
+      "learning_rate": 2.521008403361345e-06,
+      "loss": 0.5236,
+      "step": 90
+    },
+    {
+      "epoch": 0.07664233576642336,
+      "grad_norm": 0.2687312066555023,
+      "learning_rate": 2.549019607843137e-06,
+      "loss": 0.4928,
+      "step": 91
+    },
+    {
+      "epoch": 0.07748455923638406,
+      "grad_norm": 0.25979599356651306,
+      "learning_rate": 2.57703081232493e-06,
+      "loss": 0.4872,
+      "step": 92
+    },
+    {
+      "epoch": 0.07832678270634474,
+      "grad_norm": 0.2747167646884918,
+      "learning_rate": 2.605042016806723e-06,
+      "loss": 0.4934,
+      "step": 93
+    },
+    {
+      "epoch": 0.07916900617630544,
+      "grad_norm": 0.25741496682167053,
+      "learning_rate": 2.633053221288516e-06,
+      "loss": 0.5163,
+      "step": 94
+    },
+    {
+      "epoch": 0.08001122964626614,
+      "grad_norm": 0.2904193103313446,
+      "learning_rate": 2.6610644257703085e-06,
+      "loss": 0.494,
+      "step": 95
+    },
+    {
+      "epoch": 0.08085345311622684,
+      "grad_norm": 0.25851333141326904,
+      "learning_rate": 2.689075630252101e-06,
+      "loss": 0.4831,
+      "step": 96
+    },
+    {
+      "epoch": 0.08169567658618754,
+      "grad_norm": 0.27450624108314514,
+      "learning_rate": 2.7170868347338936e-06,
+      "loss": 0.5113,
+      "step": 97
+    },
+    {
+      "epoch": 0.08253790005614824,
+      "grad_norm": 0.2518622875213623,
+      "learning_rate": 2.7450980392156867e-06,
+      "loss": 0.4762,
+      "step": 98
+    },
+    {
+      "epoch": 0.08338012352610892,
+      "grad_norm": 0.25959309935569763,
+      "learning_rate": 2.7731092436974795e-06,
+      "loss": 0.4694,
+      "step": 99
+    },
+    {
+      "epoch": 0.08422234699606962,
+      "grad_norm": 0.28160223364830017,
+      "learning_rate": 2.8011204481792718e-06,
+      "loss": 0.5058,
+      "step": 100
+    },
+    {
+      "epoch": 0.08506457046603032,
+      "grad_norm": 0.27110520005226135,
+      "learning_rate": 2.8291316526610645e-06,
+      "loss": 0.4883,
+      "step": 101
+    },
+    {
+      "epoch": 0.08590679393599102,
+      "grad_norm": 0.2723713219165802,
+      "learning_rate": 2.8571428571428573e-06,
+      "loss": 0.5112,
+      "step": 102
+    },
+    {
+      "epoch": 0.08674901740595171,
+      "grad_norm": 0.2606522738933563,
+      "learning_rate": 2.88515406162465e-06,
+      "loss": 0.4779,
+      "step": 103
+    },
+    {
+      "epoch": 0.08759124087591241,
+      "grad_norm": 0.2593904435634613,
+      "learning_rate": 2.913165266106443e-06,
+      "loss": 0.5065,
+      "step": 104
+    },
+    {
+      "epoch": 0.08843346434587311,
+      "grad_norm": 0.2900582551956177,
+      "learning_rate": 2.9411764705882355e-06,
+      "loss": 0.49,
+      "step": 105
+    },
+    {
+      "epoch": 0.0892756878158338,
+      "grad_norm": 0.2618088126182556,
+      "learning_rate": 2.969187675070028e-06,
+      "loss": 0.4668,
+      "step": 106
+    },
+    {
+      "epoch": 0.0901179112857945,
+      "grad_norm": 0.28109776973724365,
+      "learning_rate": 2.997198879551821e-06,
+      "loss": 0.4702,
+      "step": 107
+    },
+    {
+      "epoch": 0.09096013475575519,
+      "grad_norm": 0.23819291591644287,
+      "learning_rate": 3.0252100840336137e-06,
+      "loss": 0.4843,
+      "step": 108
+    },
+    {
+      "epoch": 0.09180235822571589,
+      "grad_norm": 0.2603215277194977,
+      "learning_rate": 3.053221288515407e-06,
+      "loss": 0.4834,
+      "step": 109
+    },
+    {
+      "epoch": 0.09264458169567659,
+      "grad_norm": 0.2843914031982422,
+      "learning_rate": 3.081232492997199e-06,
+      "loss": 0.4787,
+      "step": 110
+    },
+    {
+      "epoch": 0.09348680516563729,
+      "grad_norm": 0.27807438373565674,
+      "learning_rate": 3.109243697478992e-06,
+      "loss": 0.4966,
+      "step": 111
+    },
+    {
+      "epoch": 0.09432902863559797,
+      "grad_norm": 0.25857606530189514,
+      "learning_rate": 3.1372549019607846e-06,
+      "loss": 0.4654,
+      "step": 112
+    },
+    {
+      "epoch": 0.09517125210555867,
+      "grad_norm": 0.3008019030094147,
+      "learning_rate": 3.1652661064425773e-06,
+      "loss": 0.4811,
+      "step": 113
+    },
+    {
+      "epoch": 0.09601347557551937,
+      "grad_norm": 0.2548486590385437,
+      "learning_rate": 3.1932773109243696e-06,
+      "loss": 0.486,
+      "step": 114
+    },
+    {
+      "epoch": 0.09685569904548007,
+      "grad_norm": 0.2549389600753784,
+      "learning_rate": 3.221288515406163e-06,
+      "loss": 0.4674,
+      "step": 115
+    },
+    {
+      "epoch": 0.09769792251544077,
+      "grad_norm": 0.2595871090888977,
+      "learning_rate": 3.2492997198879555e-06,
+      "loss": 0.4681,
+      "step": 116
+    },
+    {
+      "epoch": 0.09854014598540146,
+      "grad_norm": 0.24076548218727112,
+      "learning_rate": 3.2773109243697483e-06,
+      "loss": 0.4794,
+      "step": 117
+    },
+    {
+      "epoch": 0.09938236945536216,
+      "grad_norm": 0.2617177367210388,
+      "learning_rate": 3.305322128851541e-06,
+      "loss": 0.4741,
+      "step": 118
+    },
+    {
+      "epoch": 0.10022459292532285,
+      "grad_norm": 0.2643377184867859,
+      "learning_rate": 3.3333333333333333e-06,
+      "loss": 0.4778,
+      "step": 119
+    },
+    {
+      "epoch": 0.10106681639528355,
+      "grad_norm": 0.25778520107269287,
+      "learning_rate": 3.3613445378151265e-06,
+      "loss": 0.4667,
+      "step": 120
+    },
+    {
+      "epoch": 0.10190903986524424,
+      "grad_norm": 0.2485887110233307,
+      "learning_rate": 3.3893557422969192e-06,
+      "loss": 0.4942,
+      "step": 121
+    },
+    {
+      "epoch": 0.10275126333520494,
+      "grad_norm": 0.2509766221046448,
+      "learning_rate": 3.417366946778712e-06,
+      "loss": 0.4623,
+      "step": 122
+    },
+    {
+      "epoch": 0.10359348680516564,
+      "grad_norm": 0.2518707513809204,
+      "learning_rate": 3.4453781512605043e-06,
+      "loss": 0.4549,
+      "step": 123
+    },
+    {
+      "epoch": 0.10443571027512634,
+      "grad_norm": 0.2753753960132599,
+      "learning_rate": 3.473389355742297e-06,
+      "loss": 0.4455,
+      "step": 124
+    },
+    {
+      "epoch": 0.10527793374508702,
+      "grad_norm": 0.25094544887542725,
+      "learning_rate": 3.5014005602240897e-06,
+      "loss": 0.4582,
+      "step": 125
+    },
+    {
+      "epoch": 0.10612015721504772,
+      "grad_norm": 0.2611958980560303,
+      "learning_rate": 3.529411764705883e-06,
+      "loss": 0.4981,
+      "step": 126
+    },
+    {
+      "epoch": 0.10696238068500842,
+      "grad_norm": 0.2824934422969818,
+      "learning_rate": 3.5574229691876756e-06,
+      "loss": 0.4671,
+      "step": 127
+    },
+    {
+      "epoch": 0.10780460415496912,
+      "grad_norm": 0.235825777053833,
+      "learning_rate": 3.585434173669468e-06,
+      "loss": 0.4477,
+      "step": 128
+    },
+    {
+      "epoch": 0.10864682762492982,
+      "grad_norm": 0.25024521350860596,
+      "learning_rate": 3.6134453781512607e-06,
+      "loss": 0.4579,
+      "step": 129
+    },
+    {
+      "epoch": 0.10948905109489052,
+      "grad_norm": 0.26830458641052246,
+      "learning_rate": 3.6414565826330534e-06,
+      "loss": 0.4848,
+      "step": 130
+    },
+    {
+      "epoch": 0.1103312745648512,
+      "grad_norm": 0.27017107605934143,
+      "learning_rate": 3.669467787114846e-06,
+      "loss": 0.4639,
+      "step": 131
+    },
+    {
+      "epoch": 0.1111734980348119,
+      "grad_norm": 0.2695944607257843,
+      "learning_rate": 3.6974789915966393e-06,
+      "loss": 0.4513,
+      "step": 132
+    },
+    {
+      "epoch": 0.1120157215047726,
+      "grad_norm": 0.29847121238708496,
+      "learning_rate": 3.7254901960784316e-06,
+      "loss": 0.4732,
+      "step": 133
+    },
+    {
+      "epoch": 0.1128579449747333,
+      "grad_norm": 0.26092973351478577,
+      "learning_rate": 3.7535014005602243e-06,
+      "loss": 0.4714,
+      "step": 134
+    },
+    {
+      "epoch": 0.113700168444694,
+      "grad_norm": 0.3229716718196869,
+      "learning_rate": 3.781512605042017e-06,
+      "loss": 0.4414,
+      "step": 135
+    },
+    {
+      "epoch": 0.11454239191465469,
+      "grad_norm": 0.2888365685939789,
+      "learning_rate": 3.80952380952381e-06,
+      "loss": 0.4798,
+      "step": 136
+    },
+    {
+      "epoch": 0.11538461538461539,
+      "grad_norm": 0.2670290470123291,
+      "learning_rate": 3.8375350140056026e-06,
+      "loss": 0.4413,
+      "step": 137
+    },
+    {
+      "epoch": 0.11622683885457608,
+      "grad_norm": 0.2552930414676666,
+      "learning_rate": 3.865546218487396e-06,
+      "loss": 0.4621,
+      "step": 138
+    },
+    {
+      "epoch": 0.11706906232453677,
+      "grad_norm": 0.255060613155365,
+      "learning_rate": 3.893557422969188e-06,
+      "loss": 0.4274,
+      "step": 139
+    },
+    {
+      "epoch": 0.11791128579449747,
+      "grad_norm": 0.2504236698150635,
+      "learning_rate": 3.92156862745098e-06,
+      "loss": 0.4631,
+      "step": 140
+    },
+    {
+      "epoch": 0.11875350926445817,
+      "grad_norm": 0.25774672627449036,
+      "learning_rate": 3.9495798319327735e-06,
+      "loss": 0.463,
+      "step": 141
+    },
+    {
+      "epoch": 0.11959573273441887,
+      "grad_norm": 0.2837006151676178,
+      "learning_rate": 3.977591036414566e-06,
+      "loss": 0.45,
+      "step": 142
+    },
+    {
+      "epoch": 0.12043795620437957,
+      "grad_norm": 0.24713779985904694,
+      "learning_rate": 4.005602240896359e-06,
+      "loss": 0.4438,
+      "step": 143
+    },
+    {
+      "epoch": 0.12128017967434025,
+      "grad_norm": 0.2868185043334961,
+      "learning_rate": 4.033613445378151e-06,
+      "loss": 0.4702,
+      "step": 144
+    },
+    {
+      "epoch": 0.12212240314430095,
+      "grad_norm": 0.29100534319877625,
+      "learning_rate": 4.0616246498599444e-06,
+      "loss": 0.4584,
+      "step": 145
+    },
+    {
+      "epoch": 0.12296462661426165,
+      "grad_norm": 0.24189263582229614,
+      "learning_rate": 4.089635854341737e-06,
+      "loss": 0.4599,
+      "step": 146
+    },
+    {
+      "epoch": 0.12380685008422235,
+      "grad_norm": 0.26137450337409973,
+      "learning_rate": 4.11764705882353e-06,
+      "loss": 0.4549,
+      "step": 147
+    },
+    {
+      "epoch": 0.12464907355418305,
+      "grad_norm": 0.25882503390312195,
+      "learning_rate": 4.145658263305323e-06,
+      "loss": 0.4521,
+      "step": 148
+    },
+    {
+      "epoch": 0.12549129702414374,
+      "grad_norm": 0.26184239983558655,
+      "learning_rate": 4.173669467787115e-06,
+      "loss": 0.4576,
+      "step": 149
+    },
+    {
+      "epoch": 0.12633352049410443,
+      "grad_norm": 0.2604025602340698,
+      "learning_rate": 4.201680672268908e-06,
+      "loss": 0.4806,
+      "step": 150
+    },
+    {
+      "epoch": 0.12717574396406514,
+      "grad_norm": 0.2570963203907013,
+      "learning_rate": 4.229691876750701e-06,
+      "loss": 0.4283,
+      "step": 151
+    },
+    {
+      "epoch": 0.12801796743402583,
+      "grad_norm": 0.2603912353515625,
+      "learning_rate": 4.257703081232493e-06,
+      "loss": 0.4516,
+      "step": 152
+    },
+    {
+      "epoch": 0.12886019090398654,
+      "grad_norm": 0.2573627829551697,
+      "learning_rate": 4.2857142857142855e-06,
+      "loss": 0.4642,
+      "step": 153
+    },
+    {
+      "epoch": 0.12970241437394722,
+      "grad_norm": 0.2628520429134369,
+      "learning_rate": 4.313725490196079e-06,
+      "loss": 0.4614,
+      "step": 154
+    },
+    {
+      "epoch": 0.1305446378439079,
+      "grad_norm": 0.26437580585479736,
+      "learning_rate": 4.341736694677872e-06,
+      "loss": 0.4758,
+      "step": 155
+    },
+    {
+      "epoch": 0.13138686131386862,
+      "grad_norm": 0.24924257397651672,
+      "learning_rate": 4.369747899159664e-06,
+      "loss": 0.4466,
+      "step": 156
+    },
+    {
+      "epoch": 0.1322290847838293,
+      "grad_norm": 0.265608549118042,
+      "learning_rate": 4.397759103641457e-06,
+      "loss": 0.4732,
+      "step": 157
+    },
+    {
+      "epoch": 0.13307130825379002,
+      "grad_norm": 0.24660134315490723,
+      "learning_rate": 4.4257703081232496e-06,
+      "loss": 0.4453,
+      "step": 158
+    },
+    {
+      "epoch": 0.1339135317237507,
+      "grad_norm": 0.2537635266780853,
+      "learning_rate": 4.453781512605043e-06,
+      "loss": 0.4711,
+      "step": 159
+    },
+    {
+      "epoch": 0.13475575519371139,
+      "grad_norm": 0.2631363570690155,
+      "learning_rate": 4.481792717086835e-06,
+      "loss": 0.4556,
+      "step": 160
+    },
+    {
+      "epoch": 0.1355979786636721,
+      "grad_norm": 0.2548125088214874,
+      "learning_rate": 4.509803921568628e-06,
+      "loss": 0.4691,
+      "step": 161
+    },
+    {
+      "epoch": 0.13644020213363278,
+      "grad_norm": 0.2381230890750885,
+      "learning_rate": 4.5378151260504205e-06,
+      "loss": 0.4396,
+      "step": 162
+    },
+    {
+      "epoch": 0.1372824256035935,
+      "grad_norm": 0.2523537576198578,
+      "learning_rate": 4.565826330532213e-06,
+      "loss": 0.4418,
+      "step": 163
+    },
+    {
+      "epoch": 0.13812464907355418,
+      "grad_norm": 0.26021164655685425,
+      "learning_rate": 4.593837535014006e-06,
+      "loss": 0.4818,
+      "step": 164
+    },
+    {
+      "epoch": 0.1389668725435149,
+      "grad_norm": 0.2858008146286011,
+      "learning_rate": 4.621848739495799e-06,
+      "loss": 0.4398,
+      "step": 165
+    },
+    {
+      "epoch": 0.13980909601347558,
+      "grad_norm": 0.2752104699611664,
+      "learning_rate": 4.6498599439775914e-06,
+      "loss": 0.4766,
+      "step": 166
+    },
+    {
+      "epoch": 0.14065131948343626,
+      "grad_norm": 0.2509008049964905,
+      "learning_rate": 4.677871148459384e-06,
+      "loss": 0.4499,
+      "step": 167
+    },
+    {
+      "epoch": 0.14149354295339697,
+      "grad_norm": 0.2699715197086334,
+      "learning_rate": 4.705882352941177e-06,
+      "loss": 0.4696,
+      "step": 168
+    },
+    {
+      "epoch": 0.14233576642335766,
+      "grad_norm": 0.29950863122940063,
+      "learning_rate": 4.733893557422969e-06,
+      "loss": 0.4532,
+      "step": 169
+    },
+    {
+      "epoch": 0.14317798989331837,
+      "grad_norm": 0.2432369738817215,
+      "learning_rate": 4.761904761904762e-06,
+      "loss": 0.448,
+      "step": 170
+    },
+    {
+      "epoch": 0.14402021336327905,
+      "grad_norm": 0.25826480984687805,
+      "learning_rate": 4.7899159663865555e-06,
+      "loss": 0.4349,
+      "step": 171
+    },
+    {
+      "epoch": 0.14486243683323977,
+      "grad_norm": 0.2995217740535736,
+      "learning_rate": 4.817927170868348e-06,
+      "loss": 0.4619,
+      "step": 172
+    },
+    {
+      "epoch": 0.14570466030320045,
+      "grad_norm": 0.2739027440547943,
+      "learning_rate": 4.84593837535014e-06,
+      "loss": 0.4486,
+      "step": 173
+    },
+    {
+      "epoch": 0.14654688377316114,
+      "grad_norm": 0.22998765110969543,
+      "learning_rate": 4.873949579831933e-06,
+      "loss": 0.4239,
+      "step": 174
+    },
+    {
+      "epoch": 0.14738910724312185,
+      "grad_norm": 0.3160231113433838,
+      "learning_rate": 4.901960784313726e-06,
+      "loss": 0.4522,
+      "step": 175
+    },
+    {
+      "epoch": 0.14823133071308253,
+      "grad_norm": 0.2635315954685211,
+      "learning_rate": 4.929971988795519e-06,
+      "loss": 0.4372,
+      "step": 176
+    },
+    {
+      "epoch": 0.14907355418304324,
+      "grad_norm": 0.24901020526885986,
+      "learning_rate": 4.957983193277311e-06,
+      "loss": 0.4459,
+      "step": 177
+    },
+    {
+      "epoch": 0.14991577765300393,
+      "grad_norm": 0.27069857716560364,
+      "learning_rate": 4.985994397759104e-06,
+      "loss": 0.4833,
+      "step": 178
+    },
+    {
+      "epoch": 0.1507580011229646,
+      "grad_norm": 0.2763567268848419,
+      "learning_rate": 5.0140056022408966e-06,
+      "loss": 0.4369,
+      "step": 179
+    },
+    {
+      "epoch": 0.15160022459292533,
+      "grad_norm": 0.26479318737983704,
+      "learning_rate": 5.04201680672269e-06,
+      "loss": 0.4509,
+      "step": 180
+    },
+    {
+      "epoch": 0.152442448062886,
+      "grad_norm": 0.22581776976585388,
+      "learning_rate": 5.070028011204482e-06,
+      "loss": 0.4173,
+      "step": 181
+    },
+    {
+      "epoch": 0.15328467153284672,
+      "grad_norm": 0.2581729292869568,
+      "learning_rate": 5.098039215686274e-06,
+      "loss": 0.4512,
+      "step": 182
+    },
+    {
+      "epoch": 0.1541268950028074,
+      "grad_norm": 0.2569541931152344,
+      "learning_rate": 5.1260504201680675e-06,
+      "loss": 0.4441,
+      "step": 183
+    },
+    {
+      "epoch": 0.15496911847276812,
+      "grad_norm": 0.27027472853660583,
+      "learning_rate": 5.15406162464986e-06,
+      "loss": 0.4612,
+      "step": 184
+    },
+    {
+      "epoch": 0.1558113419427288,
+      "grad_norm": 0.2799546420574188,
+      "learning_rate": 5.182072829131654e-06,
+      "loss": 0.4713,
+      "step": 185
+    },
+    {
+      "epoch": 0.1566535654126895,
+      "grad_norm": 0.27860912680625916,
+      "learning_rate": 5.210084033613446e-06,
+      "loss": 0.4368,
+      "step": 186
+    },
+    {
+      "epoch": 0.1574957888826502,
+      "grad_norm": 0.2584918439388275,
+      "learning_rate": 5.2380952380952384e-06,
+      "loss": 0.4169,
+      "step": 187
+    },
+    {
+      "epoch": 0.15833801235261089,
+      "grad_norm": 0.2801576554775238,
+      "learning_rate": 5.266106442577032e-06,
+      "loss": 0.438,
+      "step": 188
+    },
+    {
+      "epoch": 0.1591802358225716,
+      "grad_norm": 0.3123912811279297,
+      "learning_rate": 5.294117647058824e-06,
+      "loss": 0.4763,
+      "step": 189
+    },
+    {
+      "epoch": 0.16002245929253228,
+      "grad_norm": 0.2466737926006317,
+      "learning_rate": 5.322128851540617e-06,
+      "loss": 0.4384,
+      "step": 190
+    },
+    {
+      "epoch": 0.160864682762493,
+      "grad_norm": 0.24402092397212982,
+      "learning_rate": 5.350140056022409e-06,
+      "loss": 0.4226,
+      "step": 191
+    },
+    {
+      "epoch": 0.16170690623245368,
+      "grad_norm": 0.3021142780780792,
+      "learning_rate": 5.378151260504202e-06,
+      "loss": 0.4303,
+      "step": 192
+    },
+    {
+      "epoch": 0.16254912970241436,
+      "grad_norm": 0.2734925150871277,
+      "learning_rate": 5.406162464985995e-06,
+      "loss": 0.4308,
+      "step": 193
+    },
+    {
+      "epoch": 0.16339135317237508,
+      "grad_norm": 0.24589981138706207,
+      "learning_rate": 5.434173669467787e-06,
+      "loss": 0.4317,
+      "step": 194
+    },
+    {
+      "epoch": 0.16423357664233576,
+      "grad_norm": 0.2625247538089752,
+      "learning_rate": 5.4621848739495795e-06,
+      "loss": 0.4357,
+      "step": 195
+    },
+    {
+      "epoch": 0.16507580011229647,
+      "grad_norm": 0.25053343176841736,
+      "learning_rate": 5.4901960784313735e-06,
+      "loss": 0.4175,
+      "step": 196
+    },
+    {
+      "epoch": 0.16591802358225716,
+      "grad_norm": 0.23492728173732758,
+      "learning_rate": 5.518207282913166e-06,
+      "loss": 0.4506,
+      "step": 197
+    },
+    {
+      "epoch": 0.16676024705221784,
+      "grad_norm": 0.2485855370759964,
+      "learning_rate": 5.546218487394959e-06,
+      "loss": 0.4517,
+      "step": 198
+    },
+    {
+      "epoch": 0.16760247052217855,
+      "grad_norm": 0.22515664994716644,
+      "learning_rate": 5.574229691876751e-06,
+      "loss": 0.437,
+      "step": 199
+    },
+    {
+      "epoch": 0.16844469399213924,
+      "grad_norm": 0.2672145962715149,
+      "learning_rate": 5.6022408963585436e-06,
+      "loss": 0.4512,
+      "step": 200
+    },
+    {
+      "epoch": 0.16928691746209995,
+      "grad_norm": 0.28508293628692627,
+      "learning_rate": 5.630252100840337e-06,
+      "loss": 0.4401,
+      "step": 201
+    },
+    {
+      "epoch": 0.17012914093206064,
+      "grad_norm": 0.2558955252170563,
+      "learning_rate": 5.658263305322129e-06,
+      "loss": 0.4215,
+      "step": 202
+    },
+    {
+      "epoch": 0.17097136440202135,
+      "grad_norm": 0.24676844477653503,
+      "learning_rate": 5.686274509803922e-06,
+      "loss": 0.4283,
+      "step": 203
+    },
+    {
+      "epoch": 0.17181358787198203,
+      "grad_norm": 0.2635219991207123,
+      "learning_rate": 5.7142857142857145e-06,
+      "loss": 0.4641,
+      "step": 204
+    },
+    {
+      "epoch": 0.17265581134194272,
+      "grad_norm": 0.23264090716838837,
+      "learning_rate": 5.742296918767507e-06,
+      "loss": 0.4257,
+      "step": 205
+    },
+    {
+      "epoch": 0.17349803481190343,
+      "grad_norm": 0.2503959536552429,
+      "learning_rate": 5.7703081232493e-06,
+      "loss": 0.4256,
+      "step": 206
+    },
+    {
+      "epoch": 0.1743402582818641,
+      "grad_norm": 0.25033625960350037,
+      "learning_rate": 5.798319327731093e-06,
+      "loss": 0.4291,
+      "step": 207
+    },
+    {
+      "epoch": 0.17518248175182483,
+      "grad_norm": 0.27226942777633667,
+      "learning_rate": 5.826330532212886e-06,
+      "loss": 0.4503,
+      "step": 208
+    },
+    {
+      "epoch": 0.1760247052217855,
+      "grad_norm": 0.25907737016677856,
+      "learning_rate": 5.854341736694679e-06,
+      "loss": 0.4339,
+      "step": 209
+    },
+    {
+      "epoch": 0.17686692869174622,
+      "grad_norm": 0.25802192091941833,
+      "learning_rate": 5.882352941176471e-06,
+      "loss": 0.4413,
+      "step": 210
+    },
+    {
+      "epoch": 0.1777091521617069,
+      "grad_norm": 0.26198717951774597,
+      "learning_rate": 5.910364145658264e-06,
+      "loss": 0.4412,
+      "step": 211
+    },
+    {
+      "epoch": 0.1785513756316676,
+      "grad_norm": 0.2522122859954834,
+      "learning_rate": 5.938375350140056e-06,
+      "loss": 0.4265,
+      "step": 212
+    },
+    {
+      "epoch": 0.1793935991016283,
+      "grad_norm": 0.2897355258464813,
+      "learning_rate": 5.9663865546218495e-06,
+      "loss": 0.4426,
+      "step": 213
+    },
+    {
+      "epoch": 0.180235822571589,
+      "grad_norm": 0.2690928876399994,
+      "learning_rate": 5.994397759103642e-06,
+      "loss": 0.4573,
+      "step": 214
+    },
+    {
+      "epoch": 0.1810780460415497,
+      "grad_norm": 0.2453453242778778,
+      "learning_rate": 6.022408963585434e-06,
+      "loss": 0.4462,
+      "step": 215
+    },
+    {
+      "epoch": 0.18192026951151039,
+      "grad_norm": 0.23455454409122467,
+      "learning_rate": 6.050420168067227e-06,
+      "loss": 0.4422,
+      "step": 216
+    },
+    {
+      "epoch": 0.1827624929814711,
+      "grad_norm": 0.25252288579940796,
+      "learning_rate": 6.07843137254902e-06,
+      "loss": 0.4347,
+      "step": 217
+    },
+    {
+      "epoch": 0.18360471645143178,
+      "grad_norm": 0.2373313456773758,
+      "learning_rate": 6.106442577030814e-06,
+      "loss": 0.4172,
+      "step": 218
+    },
+    {
+      "epoch": 0.18444693992139247,
+      "grad_norm": 0.22695258259773254,
+      "learning_rate": 6.134453781512606e-06,
+      "loss": 0.4394,
+      "step": 219
+    },
+    {
+      "epoch": 0.18528916339135318,
+      "grad_norm": 0.24199916422367096,
+      "learning_rate": 6.162464985994398e-06,
+      "loss": 0.4357,
+      "step": 220
+    },
+    {
+      "epoch": 0.18613138686131386,
+      "grad_norm": 0.27336186170578003,
+      "learning_rate": 6.1904761904761914e-06,
+      "loss": 0.4304,
+      "step": 221
+    },
+    {
+      "epoch": 0.18697361033127458,
+      "grad_norm": 0.23748163878917694,
+      "learning_rate": 6.218487394957984e-06,
+      "loss": 0.4407,
+      "step": 222
+    },
+    {
+      "epoch": 0.18781583380123526,
+      "grad_norm": 0.2980271279811859,
+      "learning_rate": 6.246498599439776e-06,
+      "loss": 0.4307,
+      "step": 223
+    },
+    {
+      "epoch": 0.18865805727119594,
+      "grad_norm": 0.25661277770996094,
+      "learning_rate": 6.274509803921569e-06,
+      "loss": 0.4273,
+      "step": 224
+    },
+    {
+      "epoch": 0.18950028074115666,
+      "grad_norm": 0.27232545614242554,
+      "learning_rate": 6.3025210084033615e-06,
+      "loss": 0.4142,
+      "step": 225
+    },
+    {
+      "epoch": 0.19034250421111734,
+      "grad_norm": 0.2807347774505615,
+      "learning_rate": 6.330532212885155e-06,
+      "loss": 0.4444,
+      "step": 226
+    },
+    {
+      "epoch": 0.19118472768107805,
+      "grad_norm": 0.25483405590057373,
+      "learning_rate": 6.358543417366947e-06,
+      "loss": 0.4433,
+      "step": 227
+    },
+    {
+      "epoch": 0.19202695115103874,
+      "grad_norm": 0.263897180557251,
+      "learning_rate": 6.386554621848739e-06,
+      "loss": 0.4229,
+      "step": 228
+    },
+    {
+      "epoch": 0.19286917462099945,
+      "grad_norm": 0.27256882190704346,
+      "learning_rate": 6.414565826330533e-06,
+      "loss": 0.4273,
+      "step": 229
+    },
+    {
+      "epoch": 0.19371139809096014,
+      "grad_norm": 0.30847159028053284,
+      "learning_rate": 6.442577030812326e-06,
+      "loss": 0.4364,
+      "step": 230
+    },
+    {
+      "epoch": 0.19455362156092082,
+      "grad_norm": 0.24873819947242737,
+      "learning_rate": 6.470588235294119e-06,
+      "loss": 0.4267,
+      "step": 231
+    },
+    {
+      "epoch": 0.19539584503088153,
+      "grad_norm": 0.30360686779022217,
+      "learning_rate": 6.498599439775911e-06,
+      "loss": 0.4096,
+      "step": 232
+    },
+    {
+      "epoch": 0.19623806850084222,
+      "grad_norm": 0.29067298769950867,
+      "learning_rate": 6.526610644257703e-06,
+      "loss": 0.4656,
+      "step": 233
+    },
+    {
+      "epoch": 0.19708029197080293,
+      "grad_norm": 0.280111700296402,
+      "learning_rate": 6.5546218487394966e-06,
+      "loss": 0.4368,
+      "step": 234
+    },
+    {
+      "epoch": 0.1979225154407636,
+      "grad_norm": 0.26713109016418457,
+      "learning_rate": 6.582633053221289e-06,
+      "loss": 0.4215,
+      "step": 235
+    },
+    {
+      "epoch": 0.19876473891072433,
+      "grad_norm": 0.2671687602996826,
+      "learning_rate": 6.610644257703082e-06,
+      "loss": 0.4087,
+      "step": 236
+    },
+    {
+      "epoch": 0.199606962380685,
+      "grad_norm": 0.2896401584148407,
+      "learning_rate": 6.638655462184874e-06,
+      "loss": 0.4387,
+      "step": 237
+    },
+    {
+      "epoch": 0.2004491858506457,
+      "grad_norm": 0.24639546871185303,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 0.4213,
+      "step": 238
+    },
+    {
+      "epoch": 0.2012914093206064,
+      "grad_norm": 0.29163262248039246,
+      "learning_rate": 6.69467787114846e-06,
+      "loss": 0.4198,
+      "step": 239
+    },
+    {
+      "epoch": 0.2021336327905671,
+      "grad_norm": 0.26881664991378784,
+      "learning_rate": 6.722689075630253e-06,
+      "loss": 0.4308,
+      "step": 240
+    },
+    {
+      "epoch": 0.2029758562605278,
+      "grad_norm": 0.26591241359710693,
+      "learning_rate": 6.750700280112046e-06,
+      "loss": 0.4266,
+      "step": 241
+    },
+    {
+      "epoch": 0.2038180797304885,
+      "grad_norm": 0.29882800579071045,
+      "learning_rate": 6.7787114845938384e-06,
+      "loss": 0.4246,
+      "step": 242
+    },
+    {
+      "epoch": 0.20466030320044917,
+      "grad_norm": 0.28500762581825256,
+      "learning_rate": 6.806722689075631e-06,
+      "loss": 0.4616,
+      "step": 243
+    },
+    {
+      "epoch": 0.20550252667040989,
+      "grad_norm": 0.3008909523487091,
+      "learning_rate": 6.834733893557424e-06,
+      "loss": 0.4245,
+      "step": 244
+    },
+    {
+      "epoch": 0.20634475014037057,
+      "grad_norm": 0.2836599349975586,
+      "learning_rate": 6.862745098039216e-06,
+      "loss": 0.418,
+      "step": 245
+    },
+    {
+      "epoch": 0.20718697361033128,
+      "grad_norm": 0.27612629532814026,
+      "learning_rate": 6.8907563025210085e-06,
+      "loss": 0.4347,
+      "step": 246
+    },
+    {
+      "epoch": 0.20802919708029197,
+      "grad_norm": 0.283955842256546,
+      "learning_rate": 6.918767507002802e-06,
+      "loss": 0.4455,
+      "step": 247
+    },
+    {
+      "epoch": 0.20887142055025268,
+      "grad_norm": 0.2872370183467865,
+      "learning_rate": 6.946778711484594e-06,
+      "loss": 0.4213,
+      "step": 248
+    },
+    {
+      "epoch": 0.20971364402021336,
+      "grad_norm": 0.2736302316188812,
+      "learning_rate": 6.974789915966387e-06,
+      "loss": 0.4213,
+      "step": 249
+    },
+    {
+      "epoch": 0.21055586749017405,
+      "grad_norm": 0.2974297106266022,
+      "learning_rate": 7.0028011204481795e-06,
+      "loss": 0.4053,
+      "step": 250
+    },
+    {
+      "epoch": 0.21139809096013476,
+      "grad_norm": 0.33730605244636536,
+      "learning_rate": 7.030812324929972e-06,
+      "loss": 0.4246,
+      "step": 251
+    },
+    {
+      "epoch": 0.21224031443009544,
+      "grad_norm": 0.2622481882572174,
+      "learning_rate": 7.058823529411766e-06,
+      "loss": 0.4202,
+      "step": 252
+    },
+    {
+      "epoch": 0.21308253790005616,
+      "grad_norm": 0.28689515590667725,
+      "learning_rate": 7.086834733893558e-06,
+      "loss": 0.4252,
+      "step": 253
+    },
+    {
+      "epoch": 0.21392476137001684,
+      "grad_norm": 0.2937809228897095,
+      "learning_rate": 7.114845938375351e-06,
+      "loss": 0.4495,
+      "step": 254
+    },
+    {
+      "epoch": 0.21476698483997755,
+      "grad_norm": 0.2611549198627472,
+      "learning_rate": 7.1428571428571436e-06,
+      "loss": 0.4294,
+      "step": 255
+    },
+    {
+      "epoch": 0.21560920830993824,
+      "grad_norm": 0.24436011910438538,
+      "learning_rate": 7.170868347338936e-06,
+      "loss": 0.4318,
+      "step": 256
+    },
+    {
+      "epoch": 0.21645143177989892,
+      "grad_norm": 0.2528916895389557,
+      "learning_rate": 7.198879551820729e-06,
+      "loss": 0.4309,
+      "step": 257
+    },
+    {
+      "epoch": 0.21729365524985964,
+      "grad_norm": 0.25828465819358826,
+      "learning_rate": 7.226890756302521e-06,
+      "loss": 0.4122,
+      "step": 258
+    },
+    {
+      "epoch": 0.21813587871982032,
+      "grad_norm": 0.24670042097568512,
+      "learning_rate": 7.2549019607843145e-06,
+      "loss": 0.4197,
+      "step": 259
+    },
+    {
+      "epoch": 0.21897810218978103,
+      "grad_norm": 0.29221484065055847,
+      "learning_rate": 7.282913165266107e-06,
+      "loss": 0.4158,
+      "step": 260
+    },
+    {
+      "epoch": 0.21982032565974172,
+      "grad_norm": 0.2650546133518219,
+      "learning_rate": 7.310924369747899e-06,
+      "loss": 0.4098,
+      "step": 261
+    },
+    {
+      "epoch": 0.2206625491297024,
+      "grad_norm": 0.27381178736686707,
+      "learning_rate": 7.338935574229692e-06,
+      "loss": 0.4165,
+      "step": 262
+    },
+    {
+      "epoch": 0.2215047725996631,
+      "grad_norm": 0.2621234357357025,
+      "learning_rate": 7.3669467787114854e-06,
+      "loss": 0.4566,
+      "step": 263
+    },
+    {
+      "epoch": 0.2223469960696238,
+      "grad_norm": 0.27825459837913513,
+      "learning_rate": 7.394957983193279e-06,
+      "loss": 0.4297,
+      "step": 264
+    },
+    {
+      "epoch": 0.2231892195395845,
+      "grad_norm": 0.27097970247268677,
+      "learning_rate": 7.422969187675071e-06,
+      "loss": 0.4318,
+      "step": 265
+    },
+    {
+      "epoch": 0.2240314430095452,
+      "grad_norm": 0.29523909091949463,
+      "learning_rate": 7.450980392156863e-06,
+      "loss": 0.4327,
+      "step": 266
+    },
+    {
+      "epoch": 0.2248736664795059,
+      "grad_norm": 0.3030112683773041,
+      "learning_rate": 7.478991596638656e-06,
+      "loss": 0.4317,
+      "step": 267
+    },
+    {
+      "epoch": 0.2257158899494666,
+      "grad_norm": 0.25813236832618713,
+      "learning_rate": 7.507002801120449e-06,
+      "loss": 0.4137,
+      "step": 268
+    },
+    {
+      "epoch": 0.22655811341942728,
+      "grad_norm": 0.2950754761695862,
+      "learning_rate": 7.535014005602241e-06,
+      "loss": 0.4026,
+      "step": 269
+    },
+    {
+      "epoch": 0.227400336889388,
+      "grad_norm": 0.2944609522819519,
+      "learning_rate": 7.563025210084034e-06,
+      "loss": 0.4171,
+      "step": 270
+    },
+    {
+      "epoch": 0.22824256035934867,
+      "grad_norm": 0.25540435314178467,
+      "learning_rate": 7.5910364145658265e-06,
+      "loss": 0.4385,
+      "step": 271
+    },
+    {
+      "epoch": 0.22908478382930939,
+      "grad_norm": 0.28693222999572754,
+      "learning_rate": 7.61904761904762e-06,
+      "loss": 0.4317,
+      "step": 272
+    },
+    {
+      "epoch": 0.22992700729927007,
+      "grad_norm": 0.32201504707336426,
+      "learning_rate": 7.647058823529411e-06,
+      "loss": 0.4383,
+      "step": 273
+    },
+    {
+      "epoch": 0.23076923076923078,
+      "grad_norm": 0.2618683874607086,
+      "learning_rate": 7.675070028011205e-06,
+      "loss": 0.3924,
+      "step": 274
+    },
+    {
+      "epoch": 0.23161145423919147,
+      "grad_norm": 0.2778761684894562,
+      "learning_rate": 7.703081232492997e-06,
+      "loss": 0.4213,
+      "step": 275
+    },
+    {
+      "epoch": 0.23245367770915215,
+      "grad_norm": 0.28987208008766174,
+      "learning_rate": 7.731092436974791e-06,
+      "loss": 0.4325,
+      "step": 276
+    },
+    {
+      "epoch": 0.23329590117911286,
+      "grad_norm": 0.28098398447036743,
+      "learning_rate": 7.759103641456584e-06,
+      "loss": 0.4246,
+      "step": 277
+    },
+    {
+      "epoch": 0.23413812464907355,
+      "grad_norm": 0.29261812567710876,
+      "learning_rate": 7.787114845938376e-06,
+      "loss": 0.4511,
+      "step": 278
+    },
+    {
+      "epoch": 0.23498034811903426,
+      "grad_norm": 0.2923807203769684,
+      "learning_rate": 7.815126050420168e-06,
+      "loss": 0.4305,
+      "step": 279
+    },
+    {
+      "epoch": 0.23582257158899494,
+      "grad_norm": 0.2776165008544922,
+      "learning_rate": 7.84313725490196e-06,
+      "loss": 0.4071,
+      "step": 280
+    },
+    {
+      "epoch": 0.23666479505895563,
+      "grad_norm": 0.3058721423149109,
+      "learning_rate": 7.871148459383755e-06,
+      "loss": 0.4349,
+      "step": 281
+    },
+    {
+      "epoch": 0.23750701852891634,
+      "grad_norm": 0.24573291838169098,
+      "learning_rate": 7.899159663865547e-06,
+      "loss": 0.4177,
+      "step": 282
+    },
+    {
+      "epoch": 0.23834924199887703,
+      "grad_norm": 0.29806849360466003,
+      "learning_rate": 7.92717086834734e-06,
+      "loss": 0.433,
+      "step": 283
+    },
+    {
+      "epoch": 0.23919146546883774,
+      "grad_norm": 0.2590548098087311,
+      "learning_rate": 7.955182072829132e-06,
+      "loss": 0.4082,
+      "step": 284
+    },
+    {
+      "epoch": 0.24003368893879842,
+      "grad_norm": 0.31977927684783936,
+      "learning_rate": 7.983193277310926e-06,
+      "loss": 0.4477,
+      "step": 285
+    },
+    {
+      "epoch": 0.24087591240875914,
+      "grad_norm": 0.2705984115600586,
+      "learning_rate": 8.011204481792718e-06,
+      "loss": 0.4006,
+      "step": 286
+    },
+    {
+      "epoch": 0.24171813587871982,
+      "grad_norm": 0.2714398205280304,
+      "learning_rate": 8.03921568627451e-06,
+      "loss": 0.4317,
+      "step": 287
+    },
+    {
+      "epoch": 0.2425603593486805,
+      "grad_norm": 0.27493196725845337,
+      "learning_rate": 8.067226890756303e-06,
+      "loss": 0.429,
+      "step": 288
+    },
+    {
+      "epoch": 0.24340258281864122,
+      "grad_norm": 0.28182482719421387,
+      "learning_rate": 8.095238095238097e-06,
+      "loss": 0.3974,
+      "step": 289
+    },
+    {
+      "epoch": 0.2442448062886019,
+      "grad_norm": 0.26035940647125244,
+      "learning_rate": 8.123249299719889e-06,
+      "loss": 0.4025,
+      "step": 290
+    },
+    {
+      "epoch": 0.2450870297585626,
+      "grad_norm": 0.2671283781528473,
+      "learning_rate": 8.151260504201681e-06,
+      "loss": 0.4375,
+      "step": 291
+    },
+    {
+      "epoch": 0.2459292532285233,
+      "grad_norm": 0.24599888920783997,
+      "learning_rate": 8.179271708683473e-06,
+      "loss": 0.3932,
+      "step": 292
+    },
+    {
+      "epoch": 0.246771476698484,
+      "grad_norm": 0.26801490783691406,
+      "learning_rate": 8.207282913165266e-06,
+      "loss": 0.4219,
+      "step": 293
+    },
+    {
+      "epoch": 0.2476137001684447,
+      "grad_norm": 0.2903979420661926,
+      "learning_rate": 8.23529411764706e-06,
+      "loss": 0.4306,
+      "step": 294
+    },
+    {
+      "epoch": 0.24845592363840538,
+      "grad_norm": 0.26690545678138733,
+      "learning_rate": 8.263305322128852e-06,
+      "loss": 0.4009,
+      "step": 295
+    },
+    {
+      "epoch": 0.2492981471083661,
+      "grad_norm": 0.2960893213748932,
+      "learning_rate": 8.291316526610646e-06,
+      "loss": 0.4198,
+      "step": 296
+    },
+    {
+      "epoch": 0.2501403705783268,
+      "grad_norm": 0.24158838391304016,
+      "learning_rate": 8.319327731092438e-06,
+      "loss": 0.4145,
+      "step": 297
+    },
+    {
+      "epoch": 0.2509825940482875,
+      "grad_norm": 0.26518386602401733,
+      "learning_rate": 8.34733893557423e-06,
+      "loss": 0.4194,
+      "step": 298
+    },
+    {
+      "epoch": 0.2518248175182482,
+      "grad_norm": 0.2885102927684784,
+      "learning_rate": 8.375350140056023e-06,
+      "loss": 0.428,
+      "step": 299
+    },
+    {
+      "epoch": 0.25266704098820886,
+      "grad_norm": 0.24481067061424255,
+      "learning_rate": 8.403361344537815e-06,
+      "loss": 0.4162,
+      "step": 300
+    },
+    {
+      "epoch": 0.25350926445816957,
+      "grad_norm": 0.2774904668331146,
+      "learning_rate": 8.43137254901961e-06,
+      "loss": 0.4359,
+      "step": 301
+    },
+    {
+      "epoch": 0.2543514879281303,
+      "grad_norm": 0.25240957736968994,
+      "learning_rate": 8.459383753501402e-06,
+      "loss": 0.3918,
+      "step": 302
+    },
+    {
+      "epoch": 0.25519371139809094,
+      "grad_norm": 0.259308397769928,
+      "learning_rate": 8.487394957983194e-06,
+      "loss": 0.4297,
+      "step": 303
+    },
+    {
+      "epoch": 0.25603593486805165,
+      "grad_norm": 0.25079861283302307,
+      "learning_rate": 8.515406162464986e-06,
+      "loss": 0.4217,
+      "step": 304
+    },
+    {
+      "epoch": 0.25687815833801236,
+      "grad_norm": 0.2630213499069214,
+      "learning_rate": 8.543417366946779e-06,
+      "loss": 0.4308,
+      "step": 305
+    },
+    {
+      "epoch": 0.2577203818079731,
+      "grad_norm": 0.24344508349895477,
+      "learning_rate": 8.571428571428571e-06,
+      "loss": 0.3946,
+      "step": 306
+    },
+    {
+      "epoch": 0.25856260527793373,
+      "grad_norm": 0.28903988003730774,
+      "learning_rate": 8.599439775910365e-06,
+      "loss": 0.4279,
+      "step": 307
+    },
+    {
+      "epoch": 0.25940482874789444,
+      "grad_norm": 0.2642119526863098,
+      "learning_rate": 8.627450980392157e-06,
+      "loss": 0.4097,
+      "step": 308
+    },
+    {
+      "epoch": 0.26024705221785516,
+      "grad_norm": 0.2755482494831085,
+      "learning_rate": 8.655462184873951e-06,
+      "loss": 0.4129,
+      "step": 309
+    },
+    {
+      "epoch": 0.2610892756878158,
+      "grad_norm": 0.29126107692718506,
+      "learning_rate": 8.683473389355744e-06,
+      "loss": 0.4117,
+      "step": 310
+    },
+    {
+      "epoch": 0.2619314991577765,
+      "grad_norm": 0.25518178939819336,
+      "learning_rate": 8.711484593837536e-06,
+      "loss": 0.4167,
+      "step": 311
+    },
+    {
+      "epoch": 0.26277372262773724,
+      "grad_norm": 0.27598628401756287,
+      "learning_rate": 8.739495798319328e-06,
+      "loss": 0.4351,
+      "step": 312
+    },
+    {
+      "epoch": 0.2636159460976979,
+      "grad_norm": 0.2577887773513794,
+      "learning_rate": 8.76750700280112e-06,
+      "loss": 0.4299,
+      "step": 313
+    },
+    {
+      "epoch": 0.2644581695676586,
+      "grad_norm": 0.2999492883682251,
+      "learning_rate": 8.795518207282914e-06,
+      "loss": 0.4389,
+      "step": 314
+    },
+    {
+      "epoch": 0.2653003930376193,
+      "grad_norm": 0.27518129348754883,
+      "learning_rate": 8.823529411764707e-06,
+      "loss": 0.4036,
+      "step": 315
+    },
+    {
+      "epoch": 0.26614261650758003,
+      "grad_norm": 0.28360292315483093,
+      "learning_rate": 8.851540616246499e-06,
+      "loss": 0.4355,
+      "step": 316
+    },
+    {
+      "epoch": 0.2669848399775407,
+      "grad_norm": 0.27708709239959717,
+      "learning_rate": 8.879551820728291e-06,
+      "loss": 0.4011,
+      "step": 317
+    },
+    {
+      "epoch": 0.2678270634475014,
+      "grad_norm": 0.27756384015083313,
+      "learning_rate": 8.907563025210085e-06,
+      "loss": 0.4207,
+      "step": 318
+    },
+    {
+      "epoch": 0.2686692869174621,
+      "grad_norm": 0.28298452496528625,
+      "learning_rate": 8.935574229691878e-06,
+      "loss": 0.4381,
+      "step": 319
+    },
+    {
+      "epoch": 0.26951151038742277,
+      "grad_norm": 0.2583518326282501,
+      "learning_rate": 8.96358543417367e-06,
+      "loss": 0.4043,
+      "step": 320
+    },
+    {
+      "epoch": 0.2703537338573835,
+      "grad_norm": 0.27131637930870056,
+      "learning_rate": 8.991596638655462e-06,
+      "loss": 0.4223,
+      "step": 321
+    },
+    {
+      "epoch": 0.2711959573273442,
+      "grad_norm": 0.2701607644557953,
+      "learning_rate": 9.019607843137256e-06,
+      "loss": 0.4179,
+      "step": 322
+    },
+    {
+      "epoch": 0.2720381807973049,
+      "grad_norm": 0.2683499753475189,
+      "learning_rate": 9.047619047619049e-06,
+      "loss": 0.3957,
+      "step": 323
+    },
+    {
+      "epoch": 0.27288040426726556,
+      "grad_norm": 0.25922417640686035,
+      "learning_rate": 9.075630252100841e-06,
+      "loss": 0.4058,
+      "step": 324
+    },
+    {
+      "epoch": 0.2737226277372263,
+      "grad_norm": 0.2564760744571686,
+      "learning_rate": 9.103641456582633e-06,
+      "loss": 0.4108,
+      "step": 325
+    },
+    {
+      "epoch": 0.274564851207187,
+      "grad_norm": 0.26044026017189026,
+      "learning_rate": 9.131652661064426e-06,
+      "loss": 0.4112,
+      "step": 326
+    },
+    {
+      "epoch": 0.27540707467714765,
+      "grad_norm": 0.272298127412796,
+      "learning_rate": 9.15966386554622e-06,
+      "loss": 0.4191,
+      "step": 327
+    },
+    {
+      "epoch": 0.27624929814710836,
+      "grad_norm": 0.2655772268772125,
+      "learning_rate": 9.187675070028012e-06,
+      "loss": 0.4091,
+      "step": 328
+    },
+    {
+      "epoch": 0.27709152161706907,
+      "grad_norm": 0.26740503311157227,
+      "learning_rate": 9.215686274509804e-06,
+      "loss": 0.3991,
+      "step": 329
+    },
+    {
+      "epoch": 0.2779337450870298,
+      "grad_norm": 0.28810393810272217,
+      "learning_rate": 9.243697478991598e-06,
+      "loss": 0.4245,
+      "step": 330
+    },
+    {
+      "epoch": 0.27877596855699044,
+      "grad_norm": 0.29594841599464417,
+      "learning_rate": 9.27170868347339e-06,
+      "loss": 0.4275,
+      "step": 331
+    },
+    {
+      "epoch": 0.27961819202695115,
+      "grad_norm": 0.30016207695007324,
+      "learning_rate": 9.299719887955183e-06,
+      "loss": 0.3934,
+      "step": 332
+    },
+    {
+      "epoch": 0.28046041549691186,
+      "grad_norm": 0.25892147421836853,
+      "learning_rate": 9.327731092436975e-06,
+      "loss": 0.4425,
+      "step": 333
+    },
+    {
+      "epoch": 0.2813026389668725,
+      "grad_norm": 0.305174320936203,
+      "learning_rate": 9.355742296918767e-06,
+      "loss": 0.429,
+      "step": 334
+    },
+    {
+      "epoch": 0.28214486243683323,
+      "grad_norm": 0.29193490743637085,
+      "learning_rate": 9.383753501400561e-06,
+      "loss": 0.4226,
+      "step": 335
+    },
+    {
+      "epoch": 0.28298708590679394,
+      "grad_norm": 0.3096065819263458,
+      "learning_rate": 9.411764705882354e-06,
+      "loss": 0.4501,
+      "step": 336
+    },
+    {
+      "epoch": 0.28382930937675466,
+      "grad_norm": 0.28678637742996216,
+      "learning_rate": 9.439775910364146e-06,
+      "loss": 0.424,
+      "step": 337
+    },
+    {
+      "epoch": 0.2846715328467153,
+      "grad_norm": 0.2797674536705017,
+      "learning_rate": 9.467787114845938e-06,
+      "loss": 0.3909,
+      "step": 338
+    },
+    {
+      "epoch": 0.285513756316676,
+      "grad_norm": 0.30397042632102966,
+      "learning_rate": 9.49579831932773e-06,
+      "loss": 0.4042,
+      "step": 339
+    },
+    {
+      "epoch": 0.28635597978663674,
+      "grad_norm": 0.28562748432159424,
+      "learning_rate": 9.523809523809525e-06,
+      "loss": 0.4316,
+      "step": 340
+    },
+    {
+      "epoch": 0.2871982032565974,
+      "grad_norm": 0.27482637763023376,
+      "learning_rate": 9.551820728291317e-06,
+      "loss": 0.415,
+      "step": 341
+    },
+    {
+      "epoch": 0.2880404267265581,
+      "grad_norm": 0.31606414914131165,
+      "learning_rate": 9.579831932773111e-06,
+      "loss": 0.4069,
+      "step": 342
+    },
+    {
+      "epoch": 0.2888826501965188,
+      "grad_norm": 0.348613440990448,
+      "learning_rate": 9.607843137254903e-06,
+      "loss": 0.4326,
+      "step": 343
+    },
+    {
+      "epoch": 0.28972487366647953,
+      "grad_norm": 0.2946453094482422,
+      "learning_rate": 9.635854341736696e-06,
+      "loss": 0.4271,
+      "step": 344
+    },
+    {
+      "epoch": 0.2905670971364402,
+      "grad_norm": 0.2972312867641449,
+      "learning_rate": 9.663865546218488e-06,
+      "loss": 0.4226,
+      "step": 345
+    },
+    {
+      "epoch": 0.2914093206064009,
+      "grad_norm": 0.3148708641529083,
+      "learning_rate": 9.69187675070028e-06,
+      "loss": 0.3889,
+      "step": 346
+    },
+    {
+      "epoch": 0.2922515440763616,
+      "grad_norm": 0.3089003562927246,
+      "learning_rate": 9.719887955182074e-06,
+      "loss": 0.3934,
+      "step": 347
+    },
+    {
+      "epoch": 0.29309376754632227,
+      "grad_norm": 0.29248514771461487,
+      "learning_rate": 9.747899159663867e-06,
+      "loss": 0.4003,
+      "step": 348
+    },
+    {
+      "epoch": 0.293935991016283,
+      "grad_norm": 0.3220504820346832,
+      "learning_rate": 9.775910364145659e-06,
+      "loss": 0.4282,
+      "step": 349
+    },
+    {
+      "epoch": 0.2947782144862437,
+      "grad_norm": 0.3207250237464905,
+      "learning_rate": 9.803921568627451e-06,
+      "loss": 0.4099,
+      "step": 350
+    },
+    {
+      "epoch": 0.2956204379562044,
+      "grad_norm": 0.31499800086021423,
+      "learning_rate": 9.831932773109244e-06,
+      "loss": 0.3907,
+      "step": 351
+    },
+    {
+      "epoch": 0.29646266142616506,
+      "grad_norm": 0.2804590165615082,
+      "learning_rate": 9.859943977591038e-06,
+      "loss": 0.3929,
+      "step": 352
+    },
+    {
+      "epoch": 0.2973048848961258,
+      "grad_norm": 0.3583865165710449,
+      "learning_rate": 9.88795518207283e-06,
+      "loss": 0.4331,
+      "step": 353
+    },
+    {
+      "epoch": 0.2981471083660865,
+      "grad_norm": 0.2689453065395355,
+      "learning_rate": 9.915966386554622e-06,
+      "loss": 0.4261,
+      "step": 354
+    },
+    {
+      "epoch": 0.29898933183604715,
+      "grad_norm": 0.24902856349945068,
+      "learning_rate": 9.943977591036416e-06,
+      "loss": 0.3995,
+      "step": 355
+    },
+    {
+      "epoch": 0.29983155530600786,
+      "grad_norm": 0.28560537099838257,
+      "learning_rate": 9.971988795518209e-06,
+      "loss": 0.413,
+      "step": 356
+    },
+    {
+      "epoch": 0.30067377877596857,
+      "grad_norm": 0.2845311164855957,
+      "learning_rate": 1e-05,
+      "loss": 0.4229,
+      "step": 357
+    },
+    {
+      "epoch": 0.3015160022459292,
+      "grad_norm": 0.2546791732311249,
+      "learning_rate": 9.99999759644146e-06,
+      "loss": 0.4159,
+      "step": 358
+    },
+    {
+      "epoch": 0.30235822571588994,
+      "grad_norm": 0.2738746702671051,
+      "learning_rate": 9.999990385768144e-06,
+      "loss": 0.4301,
+      "step": 359
+    },
+    {
+      "epoch": 0.30320044918585065,
+      "grad_norm": 0.30176639556884766,
+      "learning_rate": 9.999978367986988e-06,
+      "loss": 0.3877,
+      "step": 360
+    },
+    {
+      "epoch": 0.30404267265581136,
+      "grad_norm": 0.26767653226852417,
+      "learning_rate": 9.999961543109546e-06,
+      "loss": 0.4062,
+      "step": 361
+    },
+    {
+      "epoch": 0.304884896125772,
+      "grad_norm": 0.30675870180130005,
+      "learning_rate": 9.999939911151992e-06,
+      "loss": 0.4241,
+      "step": 362
+    },
+    {
+      "epoch": 0.30572711959573273,
+      "grad_norm": 0.2657410204410553,
+      "learning_rate": 9.999913472135126e-06,
+      "loss": 0.4107,
+      "step": 363
+    },
+    {
+      "epoch": 0.30656934306569344,
+      "grad_norm": 0.28486159443855286,
+      "learning_rate": 9.999882226084366e-06,
+      "loss": 0.4193,
+      "step": 364
+    },
+    {
+      "epoch": 0.3074115665356541,
+      "grad_norm": 0.271361768245697,
+      "learning_rate": 9.999846173029752e-06,
+      "loss": 0.4184,
+      "step": 365
+    },
+    {
+      "epoch": 0.3082537900056148,
+      "grad_norm": 0.26552683115005493,
+      "learning_rate": 9.999805313005946e-06,
+      "loss": 0.4155,
+      "step": 366
+    },
+    {
+      "epoch": 0.3090960134755755,
+      "grad_norm": 0.2777312994003296,
+      "learning_rate": 9.999759646052234e-06,
+      "loss": 0.4078,
+      "step": 367
+    },
+    {
+      "epoch": 0.30993823694553624,
+      "grad_norm": 0.26210469007492065,
+      "learning_rate": 9.99970917221252e-06,
+      "loss": 0.4068,
+      "step": 368
+    },
+    {
+      "epoch": 0.3107804604154969,
+      "grad_norm": 0.2592101991176605,
+      "learning_rate": 9.99965389153533e-06,
+      "loss": 0.4084,
+      "step": 369
+    },
+    {
+      "epoch": 0.3116226838854576,
+      "grad_norm": 0.26666998863220215,
+      "learning_rate": 9.999593804073812e-06,
+      "loss": 0.4118,
+      "step": 370
+    },
+    {
+      "epoch": 0.3124649073554183,
+      "grad_norm": 0.28305909037590027,
+      "learning_rate": 9.999528909885738e-06,
+      "loss": 0.3965,
+      "step": 371
+    },
+    {
+      "epoch": 0.313307130825379,
+      "grad_norm": 0.25685542821884155,
+      "learning_rate": 9.999459209033495e-06,
+      "loss": 0.4255,
+      "step": 372
+    },
+    {
+      "epoch": 0.3141493542953397,
+      "grad_norm": 0.2907734811306,
+      "learning_rate": 9.999384701584098e-06,
+      "loss": 0.4074,
+      "step": 373
+    },
+    {
+      "epoch": 0.3149915777653004,
+      "grad_norm": 0.27968016266822815,
+      "learning_rate": 9.99930538760918e-06,
+      "loss": 0.4105,
+      "step": 374
+    },
+    {
+      "epoch": 0.3158338012352611,
+      "grad_norm": 0.30917319655418396,
+      "learning_rate": 9.999221267184993e-06,
+      "loss": 0.3744,
+      "step": 375
+    },
+    {
+      "epoch": 0.31667602470522177,
+      "grad_norm": 0.3107389807701111,
+      "learning_rate": 9.999132340392416e-06,
+      "loss": 0.4255,
+      "step": 376
+    },
+    {
+      "epoch": 0.3175182481751825,
+      "grad_norm": 0.3284875154495239,
+      "learning_rate": 9.999038607316942e-06,
+      "loss": 0.4103,
+      "step": 377
+    },
+    {
+      "epoch": 0.3183604716451432,
+      "grad_norm": 0.3265838921070099,
+      "learning_rate": 9.998940068048688e-06,
+      "loss": 0.4163,
+      "step": 378
+    },
+    {
+      "epoch": 0.31920269511510385,
+      "grad_norm": 0.34593111276626587,
+      "learning_rate": 9.998836722682397e-06,
+      "loss": 0.4144,
+      "step": 379
+    },
+    {
+      "epoch": 0.32004491858506456,
+      "grad_norm": 0.309358686208725,
+      "learning_rate": 9.998728571317422e-06,
+      "loss": 0.4178,
+      "step": 380
+    },
+    {
+      "epoch": 0.3208871420550253,
+      "grad_norm": 0.2778068482875824,
+      "learning_rate": 9.998615614057743e-06,
+      "loss": 0.4281,
+      "step": 381
+    },
+    {
+      "epoch": 0.321729365524986,
+      "grad_norm": 0.40132778882980347,
+      "learning_rate": 9.998497851011963e-06,
+      "loss": 0.4237,
+      "step": 382
+    },
+    {
+      "epoch": 0.32257158899494665,
+      "grad_norm": 0.2529246509075165,
+      "learning_rate": 9.998375282293298e-06,
+      "loss": 0.415,
+      "step": 383
+    },
+    {
+      "epoch": 0.32341381246490736,
+      "grad_norm": 0.3505932092666626,
+      "learning_rate": 9.998247908019594e-06,
+      "loss": 0.4394,
+      "step": 384
+    },
+    {
+      "epoch": 0.32425603593486807,
+      "grad_norm": 0.3261929154396057,
+      "learning_rate": 9.998115728313305e-06,
+      "loss": 0.3895,
+      "step": 385
+    },
+    {
+      "epoch": 0.3250982594048287,
+      "grad_norm": 0.37192705273628235,
+      "learning_rate": 9.997978743301516e-06,
+      "loss": 0.4179,
+      "step": 386
+    },
+    {
+      "epoch": 0.32594048287478944,
+      "grad_norm": 0.25225892663002014,
+      "learning_rate": 9.997836953115927e-06,
+      "loss": 0.417,
+      "step": 387
+    },
+    {
+      "epoch": 0.32678270634475015,
+      "grad_norm": 0.34606847167015076,
+      "learning_rate": 9.997690357892857e-06,
+      "loss": 0.4351,
+      "step": 388
+    },
+    {
+      "epoch": 0.32762492981471086,
+      "grad_norm": 0.308843731880188,
+      "learning_rate": 9.997538957773248e-06,
+      "loss": 0.4046,
+      "step": 389
+    },
+    {
+      "epoch": 0.3284671532846715,
+      "grad_norm": 0.2955940365791321,
+      "learning_rate": 9.997382752902658e-06,
+      "loss": 0.3789,
+      "step": 390
+    },
+    {
+      "epoch": 0.32930937675463223,
+      "grad_norm": 0.303615003824234,
+      "learning_rate": 9.997221743431267e-06,
+      "loss": 0.4044,
+      "step": 391
+    },
+    {
+      "epoch": 0.33015160022459294,
+      "grad_norm": 0.30333343148231506,
+      "learning_rate": 9.997055929513873e-06,
+      "loss": 0.4393,
+      "step": 392
+    },
+    {
+      "epoch": 0.3309938236945536,
+      "grad_norm": 0.29766735434532166,
+      "learning_rate": 9.996885311309892e-06,
+      "loss": 0.422,
+      "step": 393
+    },
+    {
+      "epoch": 0.3318360471645143,
+      "grad_norm": 0.28912481665611267,
+      "learning_rate": 9.996709888983362e-06,
+      "loss": 0.411,
+      "step": 394
+    },
+    {
+      "epoch": 0.332678270634475,
+      "grad_norm": 0.30706316232681274,
+      "learning_rate": 9.99652966270294e-06,
+      "loss": 0.4136,
+      "step": 395
+    },
+    {
+      "epoch": 0.3335204941044357,
+      "grad_norm": 0.31515517830848694,
+      "learning_rate": 9.996344632641895e-06,
+      "loss": 0.4105,
+      "step": 396
+    },
+    {
+      "epoch": 0.3343627175743964,
+      "grad_norm": 0.2824975252151489,
+      "learning_rate": 9.996154798978122e-06,
+      "loss": 0.409,
+      "step": 397
+    },
+    {
+      "epoch": 0.3352049410443571,
+      "grad_norm": 0.2654586434364319,
+      "learning_rate": 9.995960161894132e-06,
+      "loss": 0.4352,
+      "step": 398
+    },
+    {
+      "epoch": 0.3360471645143178,
+      "grad_norm": 0.3069636821746826,
+      "learning_rate": 9.995760721577053e-06,
+      "loss": 0.4292,
+      "step": 399
+    },
+    {
+      "epoch": 0.3368893879842785,
+      "grad_norm": 0.2730976641178131,
+      "learning_rate": 9.99555647821863e-06,
+      "loss": 0.4101,
+      "step": 400
+    },
+    {
+      "epoch": 0.3377316114542392,
+      "grad_norm": 0.26647499203681946,
+      "learning_rate": 9.99534743201523e-06,
+      "loss": 0.3737,
+      "step": 401
+    },
+    {
+      "epoch": 0.3385738349241999,
+      "grad_norm": 0.2886996269226074,
+      "learning_rate": 9.995133583167833e-06,
+      "loss": 0.4339,
+      "step": 402
+    },
+    {
+      "epoch": 0.33941605839416056,
+      "grad_norm": 0.2916286587715149,
+      "learning_rate": 9.99491493188204e-06,
+      "loss": 0.4101,
+      "step": 403
+    },
+    {
+      "epoch": 0.34025828186412127,
+      "grad_norm": 0.32273969054222107,
+      "learning_rate": 9.994691478368067e-06,
+      "loss": 0.3825,
+      "step": 404
+    },
+    {
+      "epoch": 0.341100505334082,
+      "grad_norm": 0.26445654034614563,
+      "learning_rate": 9.994463222840748e-06,
+      "loss": 0.4172,
+      "step": 405
+    },
+    {
+      "epoch": 0.3419427288040427,
+      "grad_norm": 0.31402212381362915,
+      "learning_rate": 9.994230165519529e-06,
+      "loss": 0.4256,
+      "step": 406
+    },
+    {
+      "epoch": 0.34278495227400335,
+      "grad_norm": 0.2932952344417572,
+      "learning_rate": 9.993992306628481e-06,
+      "loss": 0.4206,
+      "step": 407
+    },
+    {
+      "epoch": 0.34362717574396406,
+      "grad_norm": 0.2703147232532501,
+      "learning_rate": 9.993749646396286e-06,
+      "loss": 0.4054,
+      "step": 408
+    },
+    {
+      "epoch": 0.3444693992139248,
+      "grad_norm": 0.31178292632102966,
+      "learning_rate": 9.993502185056244e-06,
+      "loss": 0.4237,
+      "step": 409
+    },
+    {
+      "epoch": 0.34531162268388543,
+      "grad_norm": 0.2774251103401184,
+      "learning_rate": 9.993249922846269e-06,
+      "loss": 0.4118,
+      "step": 410
+    },
+    {
+      "epoch": 0.34615384615384615,
+      "grad_norm": 0.2867864668369293,
+      "learning_rate": 9.992992860008893e-06,
+      "loss": 0.4218,
+      "step": 411
+    },
+    {
+      "epoch": 0.34699606962380686,
+      "grad_norm": 0.2802028954029083,
+      "learning_rate": 9.99273099679126e-06,
+      "loss": 0.4167,
+      "step": 412
+    },
+    {
+      "epoch": 0.34783829309376757,
+      "grad_norm": 0.24620549380779266,
+      "learning_rate": 9.992464333445134e-06,
+      "loss": 0.4018,
+      "step": 413
+    },
+    {
+      "epoch": 0.3486805165637282,
+      "grad_norm": 0.2835536301136017,
+      "learning_rate": 9.99219287022689e-06,
+      "loss": 0.4086,
+      "step": 414
+    },
+    {
+      "epoch": 0.34952274003368894,
+      "grad_norm": 0.29051610827445984,
+      "learning_rate": 9.99191660739752e-06,
+      "loss": 0.3848,
+      "step": 415
+    },
+    {
+      "epoch": 0.35036496350364965,
+      "grad_norm": 0.2779700756072998,
+      "learning_rate": 9.991635545222628e-06,
+      "loss": 0.4122,
+      "step": 416
+    },
+    {
+      "epoch": 0.3512071869736103,
+      "grad_norm": 0.25075340270996094,
+      "learning_rate": 9.991349683972435e-06,
+      "loss": 0.3967,
+      "step": 417
+    },
+    {
+      "epoch": 0.352049410443571,
+      "grad_norm": 0.26911744475364685,
+      "learning_rate": 9.991059023921773e-06,
+      "loss": 0.4128,
+      "step": 418
+    },
+    {
+      "epoch": 0.35289163391353173,
+      "grad_norm": 0.2757958471775055,
+      "learning_rate": 9.990763565350092e-06,
+      "loss": 0.4236,
+      "step": 419
+    },
+    {
+      "epoch": 0.35373385738349244,
+      "grad_norm": 0.27440062165260315,
+      "learning_rate": 9.990463308541452e-06,
+      "loss": 0.4243,
+      "step": 420
+    },
+    {
+      "epoch": 0.3545760808534531,
+      "grad_norm": 0.28325390815734863,
+      "learning_rate": 9.990158253784525e-06,
+      "loss": 0.4127,
+      "step": 421
+    },
+    {
+      "epoch": 0.3554183043234138,
+      "grad_norm": 0.25825008749961853,
+      "learning_rate": 9.989848401372602e-06,
+      "loss": 0.3925,
+      "step": 422
+    },
+    {
+      "epoch": 0.3562605277933745,
+      "grad_norm": 0.2866584062576294,
+      "learning_rate": 9.989533751603578e-06,
+      "loss": 0.4258,
+      "step": 423
+    },
+    {
+      "epoch": 0.3571027512633352,
+      "grad_norm": 0.31460416316986084,
+      "learning_rate": 9.989214304779965e-06,
+      "loss": 0.4357,
+      "step": 424
+    },
+    {
+      "epoch": 0.3579449747332959,
+      "grad_norm": 0.26211369037628174,
+      "learning_rate": 9.988890061208889e-06,
+      "loss": 0.398,
+      "step": 425
+    },
+    {
+      "epoch": 0.3587871982032566,
+      "grad_norm": 0.2859407365322113,
+      "learning_rate": 9.988561021202083e-06,
+      "loss": 0.4132,
+      "step": 426
+    },
+    {
+      "epoch": 0.3596294216732173,
+      "grad_norm": 0.28992462158203125,
+      "learning_rate": 9.988227185075897e-06,
+      "loss": 0.4053,
+      "step": 427
+    },
+    {
+      "epoch": 0.360471645143178,
+      "grad_norm": 0.3033839166164398,
+      "learning_rate": 9.987888553151285e-06,
+      "loss": 0.4287,
+      "step": 428
+    },
+    {
+      "epoch": 0.3613138686131387,
+      "grad_norm": 0.28602495789527893,
+      "learning_rate": 9.987545125753818e-06,
+      "loss": 0.4176,
+      "step": 429
+    },
+    {
+      "epoch": 0.3621560920830994,
+      "grad_norm": 0.313308984041214,
+      "learning_rate": 9.987196903213677e-06,
+      "loss": 0.404,
+      "step": 430
+    },
+    {
+      "epoch": 0.36299831555306006,
+      "grad_norm": 0.26802730560302734,
+      "learning_rate": 9.986843885865649e-06,
+      "loss": 0.4092,
+      "step": 431
+    },
+    {
+      "epoch": 0.36384053902302077,
+      "grad_norm": 0.2948782742023468,
+      "learning_rate": 9.986486074049131e-06,
+      "loss": 0.3985,
+      "step": 432
+    },
+    {
+      "epoch": 0.3646827624929815,
+      "grad_norm": 0.3537862300872803,
+      "learning_rate": 9.986123468108134e-06,
+      "loss": 0.413,
+      "step": 433
+    },
+    {
+      "epoch": 0.3655249859629422,
+      "grad_norm": 0.2874160706996918,
+      "learning_rate": 9.985756068391276e-06,
+      "loss": 0.4193,
+      "step": 434
+    },
+    {
+      "epoch": 0.36636720943290285,
+      "grad_norm": 0.45029184222221375,
+      "learning_rate": 9.985383875251783e-06,
+      "loss": 0.4117,
+      "step": 435
+    },
+    {
+      "epoch": 0.36720943290286356,
+      "grad_norm": 0.2880908250808716,
+      "learning_rate": 9.985006889047492e-06,
+      "loss": 0.4198,
+      "step": 436
+    },
+    {
+      "epoch": 0.3680516563728243,
+      "grad_norm": 0.3384968638420105,
+      "learning_rate": 9.984625110140844e-06,
+      "loss": 0.4295,
+      "step": 437
+    },
+    {
+      "epoch": 0.36889387984278493,
+      "grad_norm": 0.27980169653892517,
+      "learning_rate": 9.98423853889889e-06,
+      "loss": 0.4093,
+      "step": 438
+    },
+    {
+      "epoch": 0.36973610331274565,
+      "grad_norm": 0.3056717813014984,
+      "learning_rate": 9.983847175693291e-06,
+      "loss": 0.3861,
+      "step": 439
+    },
+    {
+      "epoch": 0.37057832678270636,
+      "grad_norm": 0.29720255732536316,
+      "learning_rate": 9.983451020900312e-06,
+      "loss": 0.4159,
+      "step": 440
+    },
+    {
+      "epoch": 0.371420550252667,
+      "grad_norm": 0.3048990070819855,
+      "learning_rate": 9.983050074900824e-06,
+      "loss": 0.4171,
+      "step": 441
+    },
+    {
+      "epoch": 0.3722627737226277,
+      "grad_norm": 0.338155061006546,
+      "learning_rate": 9.982644338080308e-06,
+      "loss": 0.4201,
+      "step": 442
+    },
+    {
+      "epoch": 0.37310499719258844,
+      "grad_norm": 0.3175585865974426,
+      "learning_rate": 9.982233810828846e-06,
+      "loss": 0.4204,
+      "step": 443
+    },
+    {
+      "epoch": 0.37394722066254915,
+      "grad_norm": 0.31735533475875854,
+      "learning_rate": 9.98181849354113e-06,
+      "loss": 0.4142,
+      "step": 444
+    },
+    {
+      "epoch": 0.3747894441325098,
+      "grad_norm": 0.33125582337379456,
+      "learning_rate": 9.98139838661646e-06,
+      "loss": 0.4166,
+      "step": 445
+    },
+    {
+      "epoch": 0.3756316676024705,
+      "grad_norm": 0.2836975157260895,
+      "learning_rate": 9.980973490458728e-06,
+      "loss": 0.4033,
+      "step": 446
+    },
+    {
+      "epoch": 0.37647389107243123,
+      "grad_norm": 0.303907573223114,
+      "learning_rate": 9.980543805476447e-06,
+      "loss": 0.4017,
+      "step": 447
+    },
+    {
+      "epoch": 0.3773161145423919,
+      "grad_norm": 0.3196185231208801,
+      "learning_rate": 9.980109332082722e-06,
+      "loss": 0.41,
+      "step": 448
+    },
+    {
+      "epoch": 0.3781583380123526,
+      "grad_norm": 0.3260578513145447,
+      "learning_rate": 9.979670070695265e-06,
+      "loss": 0.4247,
+      "step": 449
+    },
+    {
+      "epoch": 0.3790005614823133,
+      "grad_norm": 0.32000672817230225,
+      "learning_rate": 9.979226021736396e-06,
+      "loss": 0.4291,
+      "step": 450
+    },
+    {
+      "epoch": 0.379842784952274,
+      "grad_norm": 0.3371357321739197,
+      "learning_rate": 9.978777185633032e-06,
+      "loss": 0.3954,
+      "step": 451
+    },
+    {
+      "epoch": 0.3806850084222347,
+      "grad_norm": 0.2443530112504959,
+      "learning_rate": 9.978323562816693e-06,
+      "loss": 0.3982,
+      "step": 452
+    },
+    {
+      "epoch": 0.3815272318921954,
+      "grad_norm": 0.31013694405555725,
+      "learning_rate": 9.977865153723508e-06,
+      "loss": 0.4137,
+      "step": 453
+    },
+    {
+      "epoch": 0.3823694553621561,
+      "grad_norm": 0.31999990344047546,
+      "learning_rate": 9.977401958794194e-06,
+      "loss": 0.4103,
+      "step": 454
+    },
+    {
+      "epoch": 0.38321167883211676,
+      "grad_norm": 0.2706303298473358,
+      "learning_rate": 9.976933978474085e-06,
+      "loss": 0.4014,
+      "step": 455
+    },
+    {
+      "epoch": 0.3840539023020775,
+      "grad_norm": 0.27447131276130676,
+      "learning_rate": 9.976461213213104e-06,
+      "loss": 0.3868,
+      "step": 456
+    },
+    {
+      "epoch": 0.3848961257720382,
+      "grad_norm": 0.31807786226272583,
+      "learning_rate": 9.97598366346578e-06,
+      "loss": 0.3846,
+      "step": 457
+    },
+    {
+      "epoch": 0.3857383492419989,
+      "grad_norm": 0.271933376789093,
+      "learning_rate": 9.975501329691241e-06,
+      "loss": 0.4182,
+      "step": 458
+    },
+    {
+      "epoch": 0.38658057271195956,
+      "grad_norm": 0.2885846197605133,
+      "learning_rate": 9.975014212353212e-06,
+      "loss": 0.4128,
+      "step": 459
+    },
+    {
+      "epoch": 0.38742279618192027,
+      "grad_norm": 0.28178519010543823,
+      "learning_rate": 9.974522311920021e-06,
+      "loss": 0.4237,
+      "step": 460
+    },
+    {
+      "epoch": 0.388265019651881,
+      "grad_norm": 0.29402226209640503,
+      "learning_rate": 9.974025628864592e-06,
+      "loss": 0.3933,
+      "step": 461
+    },
+    {
+      "epoch": 0.38910724312184164,
+      "grad_norm": 0.26406511664390564,
+      "learning_rate": 9.973524163664447e-06,
+      "loss": 0.3735,
+      "step": 462
+    },
+    {
+      "epoch": 0.38994946659180235,
+      "grad_norm": 0.3236297369003296,
+      "learning_rate": 9.973017916801708e-06,
+      "loss": 0.4083,
+      "step": 463
+    },
+    {
+      "epoch": 0.39079169006176306,
+      "grad_norm": 0.2857525646686554,
+      "learning_rate": 9.972506888763092e-06,
+      "loss": 0.4027,
+      "step": 464
+    },
+    {
+      "epoch": 0.3916339135317238,
+      "grad_norm": 0.25883105397224426,
+      "learning_rate": 9.971991080039912e-06,
+      "loss": 0.4105,
+      "step": 465
+    },
+    {
+      "epoch": 0.39247613700168443,
+      "grad_norm": 0.2922370731830597,
+      "learning_rate": 9.971470491128077e-06,
+      "loss": 0.3932,
+      "step": 466
+    },
+    {
+      "epoch": 0.39331836047164515,
+      "grad_norm": 0.28805452585220337,
+      "learning_rate": 9.9709451225281e-06,
+      "loss": 0.4136,
+      "step": 467
+    },
+    {
+      "epoch": 0.39416058394160586,
+      "grad_norm": 0.26007330417633057,
+      "learning_rate": 9.970414974745077e-06,
+      "loss": 0.4072,
+      "step": 468
+    },
+    {
+      "epoch": 0.3950028074115665,
+      "grad_norm": 0.24989785254001617,
+      "learning_rate": 9.969880048288704e-06,
+      "loss": 0.4081,
+      "step": 469
+    },
+    {
+      "epoch": 0.3958450308815272,
+      "grad_norm": 0.26404762268066406,
+      "learning_rate": 9.969340343673277e-06,
+      "loss": 0.3937,
+      "step": 470
+    },
+    {
+      "epoch": 0.39668725435148794,
+      "grad_norm": 0.24274301528930664,
+      "learning_rate": 9.968795861417676e-06,
+      "loss": 0.417,
+      "step": 471
+    },
+    {
+      "epoch": 0.39752947782144865,
+      "grad_norm": 0.25516611337661743,
+      "learning_rate": 9.96824660204538e-06,
+      "loss": 0.4068,
+      "step": 472
+    },
+    {
+      "epoch": 0.3983717012914093,
+      "grad_norm": 0.24564239382743835,
+      "learning_rate": 9.96769256608446e-06,
+      "loss": 0.4197,
+      "step": 473
+    },
+    {
+      "epoch": 0.39921392476137,
+      "grad_norm": 0.2595030665397644,
+      "learning_rate": 9.967133754067581e-06,
+      "loss": 0.3948,
+      "step": 474
+    },
+    {
+      "epoch": 0.40005614823133073,
+      "grad_norm": 0.25603294372558594,
+      "learning_rate": 9.966570166531997e-06,
+      "loss": 0.4141,
+      "step": 475
+    },
+    {
+      "epoch": 0.4008983717012914,
+      "grad_norm": 0.2584262192249298,
+      "learning_rate": 9.966001804019552e-06,
+      "loss": 0.4257,
+      "step": 476
+    },
+    {
+      "epoch": 0.4017405951712521,
+      "grad_norm": 0.2369280606508255,
+      "learning_rate": 9.965428667076687e-06,
+      "loss": 0.4094,
+      "step": 477
+    },
+    {
+      "epoch": 0.4025828186412128,
+      "grad_norm": 0.2712903320789337,
+      "learning_rate": 9.964850756254426e-06,
+      "loss": 0.3783,
+      "step": 478
+    },
+    {
+      "epoch": 0.40342504211117347,
+      "grad_norm": 0.24917183816432953,
+      "learning_rate": 9.964268072108385e-06,
+      "loss": 0.3973,
+      "step": 479
+    },
+    {
+      "epoch": 0.4042672655811342,
+      "grad_norm": 0.2436331957578659,
+      "learning_rate": 9.963680615198774e-06,
+      "loss": 0.4184,
+      "step": 480
+    },
+    {
+      "epoch": 0.4051094890510949,
+      "grad_norm": 0.275935560464859,
+      "learning_rate": 9.963088386090386e-06,
+      "loss": 0.3807,
+      "step": 481
+    },
+    {
+      "epoch": 0.4059517125210556,
+      "grad_norm": 0.2412864714860916,
+      "learning_rate": 9.962491385352601e-06,
+      "loss": 0.4139,
+      "step": 482
+    },
+    {
+      "epoch": 0.40679393599101626,
+      "grad_norm": 0.26072415709495544,
+      "learning_rate": 9.961889613559396e-06,
+      "loss": 0.4119,
+      "step": 483
+    },
+    {
+      "epoch": 0.407636159460977,
+      "grad_norm": 0.27970102429389954,
+      "learning_rate": 9.961283071289323e-06,
+      "loss": 0.4231,
+      "step": 484
+    },
+    {
+      "epoch": 0.4084783829309377,
+      "grad_norm": 0.2597492039203644,
+      "learning_rate": 9.960671759125529e-06,
+      "loss": 0.418,
+      "step": 485
+    },
+    {
+      "epoch": 0.40932060640089835,
+      "grad_norm": 0.2641339898109436,
+      "learning_rate": 9.960055677655743e-06,
+      "loss": 0.3961,
+      "step": 486
+    },
+    {
+      "epoch": 0.41016282987085906,
+      "grad_norm": 0.2758226692676544,
+      "learning_rate": 9.959434827472278e-06,
+      "loss": 0.4075,
+      "step": 487
+    },
+    {
+      "epoch": 0.41100505334081977,
+      "grad_norm": 0.256102591753006,
+      "learning_rate": 9.958809209172038e-06,
+      "loss": 0.4084,
+      "step": 488
+    },
+    {
+      "epoch": 0.4118472768107805,
+      "grad_norm": 0.27487775683403015,
+      "learning_rate": 9.958178823356503e-06,
+      "loss": 0.3878,
+      "step": 489
+    },
+    {
+      "epoch": 0.41268950028074114,
+      "grad_norm": 0.24837318062782288,
+      "learning_rate": 9.957543670631743e-06,
+      "loss": 0.3978,
+      "step": 490
+    },
+    {
+      "epoch": 0.41353172375070185,
+      "grad_norm": 0.2772219777107239,
+      "learning_rate": 9.956903751608409e-06,
+      "loss": 0.4179,
+      "step": 491
+    },
+    {
+      "epoch": 0.41437394722066256,
+      "grad_norm": 0.30891284346580505,
+      "learning_rate": 9.956259066901733e-06,
+      "loss": 0.3918,
+      "step": 492
+    },
+    {
+      "epoch": 0.4152161706906232,
+      "grad_norm": 0.2519027888774872,
+      "learning_rate": 9.95560961713153e-06,
+      "loss": 0.4007,
+      "step": 493
+    },
+    {
+      "epoch": 0.41605839416058393,
+      "grad_norm": 0.31477248668670654,
+      "learning_rate": 9.954955402922195e-06,
+      "loss": 0.3806,
+      "step": 494
+    },
+    {
+      "epoch": 0.41690061763054465,
+      "grad_norm": 0.30688488483428955,
+      "learning_rate": 9.954296424902709e-06,
+      "loss": 0.402,
+      "step": 495
+    },
+    {
+      "epoch": 0.41774284110050536,
+      "grad_norm": 0.2900298833847046,
+      "learning_rate": 9.953632683706624e-06,
+      "loss": 0.4308,
+      "step": 496
+    },
+    {
+      "epoch": 0.418585064570466,
+      "grad_norm": 0.33165663480758667,
+      "learning_rate": 9.95296417997208e-06,
+      "loss": 0.4131,
+      "step": 497
+    },
+    {
+      "epoch": 0.4194272880404267,
+      "grad_norm": 0.254812628030777,
+      "learning_rate": 9.95229091434179e-06,
+      "loss": 0.3977,
+      "step": 498
+    },
+    {
+      "epoch": 0.42026951151038744,
+      "grad_norm": 0.3099961578845978,
+      "learning_rate": 9.95161288746305e-06,
+      "loss": 0.4235,
+      "step": 499
+    },
+    {
+      "epoch": 0.4211117349803481,
+      "grad_norm": 0.25541451573371887,
+      "learning_rate": 9.950930099987728e-06,
+      "loss": 0.4061,
+      "step": 500
+    },
+    {
+      "epoch": 0.4219539584503088,
+      "grad_norm": 0.2893903851509094,
+      "learning_rate": 9.950242552572272e-06,
+      "loss": 0.4148,
+      "step": 501
+    },
+    {
+      "epoch": 0.4227961819202695,
+      "grad_norm": 0.26284852623939514,
+      "learning_rate": 9.949550245877708e-06,
+      "loss": 0.3981,
+      "step": 502
+    },
+    {
+      "epoch": 0.42363840539023023,
+      "grad_norm": 0.27225029468536377,
+      "learning_rate": 9.948853180569635e-06,
+      "loss": 0.4128,
+      "step": 503
+    },
+    {
+      "epoch": 0.4244806288601909,
+      "grad_norm": 0.28634268045425415,
+      "learning_rate": 9.948151357318228e-06,
+      "loss": 0.4004,
+      "step": 504
+    },
+    {
+      "epoch": 0.4253228523301516,
+      "grad_norm": 0.28133031725883484,
+      "learning_rate": 9.947444776798235e-06,
+      "loss": 0.4123,
+      "step": 505
+    },
+    {
+      "epoch": 0.4261650758001123,
+      "grad_norm": 0.27984684705734253,
+      "learning_rate": 9.946733439688982e-06,
+      "loss": 0.3934,
+      "step": 506
+    },
+    {
+      "epoch": 0.42700729927007297,
+      "grad_norm": 0.2432078868150711,
+      "learning_rate": 9.946017346674362e-06,
+      "loss": 0.423,
+      "step": 507
+    },
+    {
+      "epoch": 0.4278495227400337,
+      "grad_norm": 0.28992959856987,
+      "learning_rate": 9.945296498442845e-06,
+      "loss": 0.3977,
+      "step": 508
+    },
+    {
+      "epoch": 0.4286917462099944,
+      "grad_norm": 0.2634131610393524,
+      "learning_rate": 9.944570895687471e-06,
+      "loss": 0.395,
+      "step": 509
+    },
+    {
+      "epoch": 0.4295339696799551,
+      "grad_norm": 0.255519300699234,
+      "learning_rate": 9.943840539105853e-06,
+      "loss": 0.3913,
+      "step": 510
+    },
+    {
+      "epoch": 0.43037619314991576,
+      "grad_norm": 0.2576368451118469,
+      "learning_rate": 9.943105429400171e-06,
+      "loss": 0.3904,
+      "step": 511
+    },
+    {
+      "epoch": 0.4312184166198765,
+      "grad_norm": 0.2746698558330536,
+      "learning_rate": 9.942365567277178e-06,
+      "loss": 0.3899,
+      "step": 512
+    },
+    {
+      "epoch": 0.4320606400898372,
+      "grad_norm": 0.2807859480381012,
+      "learning_rate": 9.941620953448195e-06,
+      "loss": 0.4123,
+      "step": 513
+    },
+    {
+      "epoch": 0.43290286355979785,
+      "grad_norm": 0.2983086407184601,
+      "learning_rate": 9.940871588629108e-06,
+      "loss": 0.4049,
+      "step": 514
+    },
+    {
+      "epoch": 0.43374508702975856,
+      "grad_norm": 0.2526918053627014,
+      "learning_rate": 9.940117473540377e-06,
+      "loss": 0.4206,
+      "step": 515
+    },
+    {
+      "epoch": 0.43458731049971927,
+      "grad_norm": 0.25227421522140503,
+      "learning_rate": 9.939358608907026e-06,
+      "loss": 0.4008,
+      "step": 516
+    },
+    {
+      "epoch": 0.43542953396968,
+      "grad_norm": 0.255201131105423,
+      "learning_rate": 9.938594995458644e-06,
+      "loss": 0.407,
+      "step": 517
+    },
+    {
+      "epoch": 0.43627175743964064,
+      "grad_norm": 0.2799823582172394,
+      "learning_rate": 9.937826633929388e-06,
+      "loss": 0.4022,
+      "step": 518
+    },
+    {
+      "epoch": 0.43711398090960135,
+      "grad_norm": 0.24459460377693176,
+      "learning_rate": 9.937053525057977e-06,
+      "loss": 0.4058,
+      "step": 519
+    },
+    {
+      "epoch": 0.43795620437956206,
+      "grad_norm": 0.26577839255332947,
+      "learning_rate": 9.936275669587697e-06,
+      "loss": 0.3902,
+      "step": 520
+    },
+    {
+      "epoch": 0.4387984278495227,
+      "grad_norm": 0.28700780868530273,
+      "learning_rate": 9.935493068266396e-06,
+      "loss": 0.4186,
+      "step": 521
+    },
+    {
+      "epoch": 0.43964065131948343,
+      "grad_norm": 0.2666192948818207,
+      "learning_rate": 9.934705721846487e-06,
+      "loss": 0.416,
+      "step": 522
+    },
+    {
+      "epoch": 0.44048287478944415,
+      "grad_norm": 0.2625783681869507,
+      "learning_rate": 9.933913631084942e-06,
+      "loss": 0.394,
+      "step": 523
+    },
+    {
+      "epoch": 0.4413250982594048,
+      "grad_norm": 0.24555350840091705,
+      "learning_rate": 9.933116796743294e-06,
+      "loss": 0.393,
+      "step": 524
+    },
+    {
+      "epoch": 0.4421673217293655,
+      "grad_norm": 0.2609098553657532,
+      "learning_rate": 9.932315219587641e-06,
+      "loss": 0.3874,
+      "step": 525
+    },
+    {
+      "epoch": 0.4430095451993262,
+      "grad_norm": 0.2597399055957794,
+      "learning_rate": 9.931508900388635e-06,
+      "loss": 0.3821,
+      "step": 526
+    },
+    {
+      "epoch": 0.44385176866928694,
+      "grad_norm": 0.27358579635620117,
+      "learning_rate": 9.930697839921496e-06,
+      "loss": 0.3947,
+      "step": 527
+    },
+    {
+      "epoch": 0.4446939921392476,
+      "grad_norm": 0.25108766555786133,
+      "learning_rate": 9.92988203896599e-06,
+      "loss": 0.3992,
+      "step": 528
+    },
+    {
+      "epoch": 0.4455362156092083,
+      "grad_norm": 0.2552182078361511,
+      "learning_rate": 9.929061498306448e-06,
+      "loss": 0.4085,
+      "step": 529
+    },
+    {
+      "epoch": 0.446378439079169,
+      "grad_norm": 0.24164170026779175,
+      "learning_rate": 9.92823621873176e-06,
+      "loss": 0.3816,
+      "step": 530
+    },
+    {
+      "epoch": 0.4472206625491297,
+      "grad_norm": 0.2571217119693756,
+      "learning_rate": 9.927406201035368e-06,
+      "loss": 0.3978,
+      "step": 531
+    },
+    {
+      "epoch": 0.4480628860190904,
+      "grad_norm": 0.2748754322528839,
+      "learning_rate": 9.926571446015271e-06,
+      "loss": 0.3985,
+      "step": 532
+    },
+    {
+      "epoch": 0.4489051094890511,
+      "grad_norm": 0.232895165681839,
+      "learning_rate": 9.92573195447402e-06,
+      "loss": 0.3957,
+      "step": 533
+    },
+    {
+      "epoch": 0.4497473329590118,
+      "grad_norm": 0.2499205768108368,
+      "learning_rate": 9.924887727218724e-06,
+      "loss": 0.4001,
+      "step": 534
+    },
+    {
+      "epoch": 0.45058955642897247,
+      "grad_norm": 0.25635528564453125,
+      "learning_rate": 9.924038765061042e-06,
+      "loss": 0.3903,
+      "step": 535
+    },
+    {
+      "epoch": 0.4514317798989332,
+      "grad_norm": 0.2474844753742218,
+      "learning_rate": 9.923185068817184e-06,
+      "loss": 0.4026,
+      "step": 536
+    },
+    {
+      "epoch": 0.4522740033688939,
+      "grad_norm": 0.24200493097305298,
+      "learning_rate": 9.922326639307918e-06,
+      "loss": 0.4116,
+      "step": 537
+    },
+    {
+      "epoch": 0.45311622683885455,
+      "grad_norm": 0.2580752968788147,
+      "learning_rate": 9.921463477358555e-06,
+      "loss": 0.4286,
+      "step": 538
+    },
+    {
+      "epoch": 0.45395845030881526,
+      "grad_norm": 0.24984806776046753,
+      "learning_rate": 9.920595583798959e-06,
+      "loss": 0.4263,
+      "step": 539
+    },
+    {
+      "epoch": 0.454800673778776,
+      "grad_norm": 0.2657654881477356,
+      "learning_rate": 9.919722959463545e-06,
+      "loss": 0.4097,
+      "step": 540
+    },
+    {
+      "epoch": 0.4556428972487367,
+      "grad_norm": 0.29026904702186584,
+      "learning_rate": 9.918845605191274e-06,
+      "loss": 0.4042,
+      "step": 541
+    },
+    {
+      "epoch": 0.45648512071869735,
+      "grad_norm": 0.2547171711921692,
+      "learning_rate": 9.917963521825653e-06,
+      "loss": 0.3989,
+      "step": 542
+    },
+    {
+      "epoch": 0.45732734418865806,
+      "grad_norm": 0.2613818347454071,
+      "learning_rate": 9.917076710214739e-06,
+      "loss": 0.3997,
+      "step": 543
+    },
+    {
+      "epoch": 0.45816956765861877,
+      "grad_norm": 0.2573147118091583,
+      "learning_rate": 9.916185171211135e-06,
+      "loss": 0.4151,
+      "step": 544
+    },
+    {
+      "epoch": 0.4590117911285794,
+      "grad_norm": 0.24024684727191925,
+      "learning_rate": 9.915288905671986e-06,
+      "loss": 0.3656,
+      "step": 545
+    },
+    {
+      "epoch": 0.45985401459854014,
+      "grad_norm": 0.2507457435131073,
+      "learning_rate": 9.914387914458983e-06,
+      "loss": 0.402,
+      "step": 546
+    },
+    {
+      "epoch": 0.46069623806850085,
+      "grad_norm": 0.25308138132095337,
+      "learning_rate": 9.913482198438357e-06,
+      "loss": 0.4187,
+      "step": 547
+    },
+    {
+      "epoch": 0.46153846153846156,
+      "grad_norm": 0.2678643465042114,
+      "learning_rate": 9.912571758480892e-06,
+      "loss": 0.4229,
+      "step": 548
+    },
+    {
+      "epoch": 0.4623806850084222,
+      "grad_norm": 0.252312570810318,
+      "learning_rate": 9.911656595461899e-06,
+      "loss": 0.4014,
+      "step": 549
+    },
+    {
+      "epoch": 0.46322290847838293,
+      "grad_norm": 0.24333544075489044,
+      "learning_rate": 9.910736710261238e-06,
+      "loss": 0.3984,
+      "step": 550
+    },
+    {
+      "epoch": 0.46406513194834365,
+      "grad_norm": 0.26091766357421875,
+      "learning_rate": 9.909812103763312e-06,
+      "loss": 0.3842,
+      "step": 551
+    },
+    {
+      "epoch": 0.4649073554183043,
+      "grad_norm": 0.25327688455581665,
+      "learning_rate": 9.908882776857057e-06,
+      "loss": 0.4054,
+      "step": 552
+    },
+    {
+      "epoch": 0.465749578888265,
+      "grad_norm": 0.30325254797935486,
+      "learning_rate": 9.90794873043595e-06,
+      "loss": 0.3966,
+      "step": 553
+    },
+    {
+      "epoch": 0.4665918023582257,
+      "grad_norm": 0.29707014560699463,
+      "learning_rate": 9.907009965398005e-06,
+      "loss": 0.3941,
+      "step": 554
+    },
+    {
+      "epoch": 0.46743402582818644,
+      "grad_norm": 0.27899348735809326,
+      "learning_rate": 9.906066482645774e-06,
+      "loss": 0.4064,
+      "step": 555
+    },
+    {
+      "epoch": 0.4682762492981471,
+      "grad_norm": 0.2798828184604645,
+      "learning_rate": 9.905118283086341e-06,
+      "loss": 0.397,
+      "step": 556
+    },
+    {
+      "epoch": 0.4691184727681078,
+      "grad_norm": 0.2678147554397583,
+      "learning_rate": 9.904165367631329e-06,
+      "loss": 0.3808,
+      "step": 557
+    },
+    {
+      "epoch": 0.4699606962380685,
+      "grad_norm": 0.25181877613067627,
+      "learning_rate": 9.903207737196892e-06,
+      "loss": 0.3951,
+      "step": 558
+    },
+    {
+      "epoch": 0.4708029197080292,
+      "grad_norm": 0.3140571117401123,
+      "learning_rate": 9.902245392703719e-06,
+      "loss": 0.3898,
+      "step": 559
+    },
+    {
+      "epoch": 0.4716451431779899,
+      "grad_norm": 0.24063082039356232,
+      "learning_rate": 9.901278335077031e-06,
+      "loss": 0.3981,
+      "step": 560
+    },
+    {
+      "epoch": 0.4724873666479506,
+      "grad_norm": 0.23210649192333221,
+      "learning_rate": 9.900306565246579e-06,
+      "loss": 0.3929,
+      "step": 561
+    },
+    {
+      "epoch": 0.47332959011791126,
+      "grad_norm": 0.25003889203071594,
+      "learning_rate": 9.899330084146646e-06,
+      "loss": 0.4097,
+      "step": 562
+    },
+    {
+      "epoch": 0.47417181358787197,
+      "grad_norm": 0.2555879056453705,
+      "learning_rate": 9.898348892716042e-06,
+      "loss": 0.3941,
+      "step": 563
+    },
+    {
+      "epoch": 0.4750140370578327,
+      "grad_norm": 0.2842501103878021,
+      "learning_rate": 9.89736299189811e-06,
+      "loss": 0.4004,
+      "step": 564
+    },
+    {
+      "epoch": 0.4758562605277934,
+      "grad_norm": 0.24519623816013336,
+      "learning_rate": 9.896372382640718e-06,
+      "loss": 0.3935,
+      "step": 565
+    },
+    {
+      "epoch": 0.47669848399775405,
+      "grad_norm": 0.2366226464509964,
+      "learning_rate": 9.895377065896259e-06,
+      "loss": 0.4067,
+      "step": 566
+    },
+    {
+      "epoch": 0.47754070746771476,
+      "grad_norm": 0.2646145820617676,
+      "learning_rate": 9.894377042621654e-06,
+      "loss": 0.4011,
+      "step": 567
+    },
+    {
+      "epoch": 0.4783829309376755,
+      "grad_norm": 0.2705758512020111,
+      "learning_rate": 9.89337231377835e-06,
+      "loss": 0.4131,
+      "step": 568
+    },
+    {
+      "epoch": 0.47922515440763613,
+      "grad_norm": 0.243195042014122,
+      "learning_rate": 9.892362880332316e-06,
+      "loss": 0.382,
+      "step": 569
+    },
+    {
+      "epoch": 0.48006737787759685,
+      "grad_norm": 0.25831979513168335,
+      "learning_rate": 9.891348743254046e-06,
+      "loss": 0.3843,
+      "step": 570
+    },
+    {
+      "epoch": 0.48090960134755756,
+      "grad_norm": 0.2449856847524643,
+      "learning_rate": 9.890329903518554e-06,
+      "loss": 0.4043,
+      "step": 571
+    },
+    {
+      "epoch": 0.48175182481751827,
+      "grad_norm": 0.25311365723609924,
+      "learning_rate": 9.889306362105377e-06,
+      "loss": 0.4008,
+      "step": 572
+    },
+    {
+      "epoch": 0.4825940482874789,
+      "grad_norm": 0.22990000247955322,
+      "learning_rate": 9.888278119998573e-06,
+      "loss": 0.4042,
+      "step": 573
+    },
+    {
+      "epoch": 0.48343627175743964,
+      "grad_norm": 0.23013539612293243,
+      "learning_rate": 9.887245178186715e-06,
+      "loss": 0.4142,
+      "step": 574
+    },
+    {
+      "epoch": 0.48427849522740035,
+      "grad_norm": 0.2604409158229828,
+      "learning_rate": 9.886207537662899e-06,
+      "loss": 0.4039,
+      "step": 575
+    },
+    {
+      "epoch": 0.485120718697361,
+      "grad_norm": 0.2513580918312073,
+      "learning_rate": 9.885165199424738e-06,
+      "loss": 0.385,
+      "step": 576
+    },
+    {
+      "epoch": 0.4859629421673217,
+      "grad_norm": 0.2692544460296631,
+      "learning_rate": 9.884118164474359e-06,
+      "loss": 0.4241,
+      "step": 577
+    },
+    {
+      "epoch": 0.48680516563728243,
+      "grad_norm": 0.2501692473888397,
+      "learning_rate": 9.883066433818404e-06,
+      "loss": 0.4002,
+      "step": 578
+    },
+    {
+      "epoch": 0.48764738910724315,
+      "grad_norm": 0.27484849095344543,
+      "learning_rate": 9.882010008468038e-06,
+      "loss": 0.378,
+      "step": 579
+    },
+    {
+      "epoch": 0.4884896125772038,
+      "grad_norm": 0.25672003626823425,
+      "learning_rate": 9.880948889438923e-06,
+      "loss": 0.4199,
+      "step": 580
+    },
+    {
+      "epoch": 0.4893318360471645,
+      "grad_norm": 0.26332417130470276,
+      "learning_rate": 9.879883077751255e-06,
+      "loss": 0.4101,
+      "step": 581
+    },
+    {
+      "epoch": 0.4901740595171252,
+      "grad_norm": 0.23965123295783997,
+      "learning_rate": 9.878812574429722e-06,
+      "loss": 0.4109,
+      "step": 582
+    },
+    {
+      "epoch": 0.4910162829870859,
+      "grad_norm": 0.272664338350296,
+      "learning_rate": 9.877737380503534e-06,
+      "loss": 0.3989,
+      "step": 583
+    },
+    {
+      "epoch": 0.4918585064570466,
+      "grad_norm": 0.23640109598636627,
+      "learning_rate": 9.876657497006408e-06,
+      "loss": 0.3999,
+      "step": 584
+    },
+    {
+      "epoch": 0.4927007299270073,
+      "grad_norm": 0.26969826221466064,
+      "learning_rate": 9.875572924976568e-06,
+      "loss": 0.3823,
+      "step": 585
+    },
+    {
+      "epoch": 0.493542953396968,
+      "grad_norm": 0.2869883179664612,
+      "learning_rate": 9.874483665456746e-06,
+      "loss": 0.4051,
+      "step": 586
+    },
+    {
+      "epoch": 0.4943851768669287,
+      "grad_norm": 0.2534290850162506,
+      "learning_rate": 9.873389719494186e-06,
+      "loss": 0.4012,
+      "step": 587
+    },
+    {
+      "epoch": 0.4952274003368894,
+      "grad_norm": 0.28421682119369507,
+      "learning_rate": 9.87229108814063e-06,
+      "loss": 0.3961,
+      "step": 588
+    },
+    {
+      "epoch": 0.4960696238068501,
+      "grad_norm": 0.25548046827316284,
+      "learning_rate": 9.871187772452327e-06,
+      "loss": 0.4219,
+      "step": 589
+    },
+    {
+      "epoch": 0.49691184727681076,
+      "grad_norm": 0.23864027857780457,
+      "learning_rate": 9.870079773490033e-06,
+      "loss": 0.4051,
+      "step": 590
+    },
+    {
+      "epoch": 0.49775407074677147,
+      "grad_norm": 0.27701038122177124,
+      "learning_rate": 9.868967092319003e-06,
+      "loss": 0.3948,
+      "step": 591
+    },
+    {
+      "epoch": 0.4985962942167322,
+      "grad_norm": 0.25791436433792114,
+      "learning_rate": 9.867849730008994e-06,
+      "loss": 0.4001,
+      "step": 592
+    },
+    {
+      "epoch": 0.4994385176866929,
+      "grad_norm": 0.22449485957622528,
+      "learning_rate": 9.866727687634266e-06,
+      "loss": 0.4034,
+      "step": 593
+    },
+    {
+      "epoch": 0.5002807411566536,
+      "grad_norm": 0.2631644308567047,
+      "learning_rate": 9.865600966273576e-06,
+      "loss": 0.3675,
+      "step": 594
+    },
+    {
+      "epoch": 0.5011229646266142,
+      "grad_norm": 0.28211626410484314,
+      "learning_rate": 9.86446956701018e-06,
+      "loss": 0.4139,
+      "step": 595
+    },
+    {
+      "epoch": 0.501965188096575,
+      "grad_norm": 0.22687473893165588,
+      "learning_rate": 9.86333349093183e-06,
+      "loss": 0.4066,
+      "step": 596
+    },
+    {
+      "epoch": 0.5028074115665356,
+      "grad_norm": 0.27827274799346924,
+      "learning_rate": 9.86219273913078e-06,
+      "loss": 0.3923,
+      "step": 597
+    },
+    {
+      "epoch": 0.5036496350364964,
+      "grad_norm": 0.24931105971336365,
+      "learning_rate": 9.861047312703772e-06,
+      "loss": 0.3816,
+      "step": 598
+    },
+    {
+      "epoch": 0.5044918585064571,
+      "grad_norm": 0.29707902669906616,
+      "learning_rate": 9.859897212752049e-06,
+      "loss": 0.4201,
+      "step": 599
+    },
+    {
+      "epoch": 0.5053340819764177,
+      "grad_norm": 0.26794612407684326,
+      "learning_rate": 9.858742440381343e-06,
+      "loss": 0.4042,
+      "step": 600
+    },
+    {
+      "epoch": 0.5061763054463785,
+      "grad_norm": 0.24728211760520935,
+      "learning_rate": 9.857582996701878e-06,
+      "loss": 0.4209,
+      "step": 601
+    },
+    {
+      "epoch": 0.5070185289163391,
+      "grad_norm": 0.2674620747566223,
+      "learning_rate": 9.856418882828368e-06,
+      "loss": 0.4173,
+      "step": 602
+    },
+    {
+      "epoch": 0.5078607523862998,
+      "grad_norm": 0.2867263853549957,
+      "learning_rate": 9.855250099880026e-06,
+      "loss": 0.3963,
+      "step": 603
+    },
+    {
+      "epoch": 0.5087029758562606,
+      "grad_norm": 0.262423574924469,
+      "learning_rate": 9.854076648980543e-06,
+      "loss": 0.4141,
+      "step": 604
+    },
+    {
+      "epoch": 0.5095451993262212,
+      "grad_norm": 0.2728235125541687,
+      "learning_rate": 9.852898531258102e-06,
+      "loss": 0.4097,
+      "step": 605
+    },
+    {
+      "epoch": 0.5103874227961819,
+      "grad_norm": 0.2570253908634186,
+      "learning_rate": 9.851715747845372e-06,
+      "loss": 0.4089,
+      "step": 606
+    },
+    {
+      "epoch": 0.5112296462661426,
+      "grad_norm": 0.2520364224910736,
+      "learning_rate": 9.850528299879513e-06,
+      "loss": 0.3943,
+      "step": 607
+    },
+    {
+      "epoch": 0.5120718697361033,
+      "grad_norm": 0.24955777823925018,
+      "learning_rate": 9.84933618850216e-06,
+      "loss": 0.393,
+      "step": 608
+    },
+    {
+      "epoch": 0.512914093206064,
+      "grad_norm": 0.25666603446006775,
+      "learning_rate": 9.848139414859441e-06,
+      "loss": 0.4066,
+      "step": 609
+    },
+    {
+      "epoch": 0.5137563166760247,
+      "grad_norm": 0.2686915099620819,
+      "learning_rate": 9.84693798010196e-06,
+      "loss": 0.4023,
+      "step": 610
+    },
+    {
+      "epoch": 0.5145985401459854,
+      "grad_norm": 0.24033565819263458,
+      "learning_rate": 9.845731885384806e-06,
+      "loss": 0.3966,
+      "step": 611
+    },
+    {
+      "epoch": 0.5154407636159462,
+      "grad_norm": 0.24050825834274292,
+      "learning_rate": 9.844521131867546e-06,
+      "loss": 0.386,
+      "step": 612
+    },
+    {
+      "epoch": 0.5162829870859068,
+      "grad_norm": 0.2602809965610504,
+      "learning_rate": 9.843305720714227e-06,
+      "loss": 0.3914,
+      "step": 613
+    },
+    {
+      "epoch": 0.5171252105558675,
+      "grad_norm": 0.2609127461910248,
+      "learning_rate": 9.842085653093372e-06,
+      "loss": 0.4117,
+      "step": 614
+    },
+    {
+      "epoch": 0.5179674340258282,
+      "grad_norm": 0.2795320749282837,
+      "learning_rate": 9.840860930177984e-06,
+      "loss": 0.4067,
+      "step": 615
+    },
+    {
+      "epoch": 0.5188096574957889,
+      "grad_norm": 0.25793221592903137,
+      "learning_rate": 9.83963155314554e-06,
+      "loss": 0.3827,
+      "step": 616
+    },
+    {
+      "epoch": 0.5196518809657495,
+      "grad_norm": 0.2973584830760956,
+      "learning_rate": 9.838397523177993e-06,
+      "loss": 0.4203,
+      "step": 617
+    },
+    {
+      "epoch": 0.5204941044357103,
+      "grad_norm": 0.2485421746969223,
+      "learning_rate": 9.837158841461767e-06,
+      "loss": 0.4058,
+      "step": 618
+    },
+    {
+      "epoch": 0.521336327905671,
+      "grad_norm": 0.28625357151031494,
+      "learning_rate": 9.835915509187759e-06,
+      "loss": 0.4052,
+      "step": 619
+    },
+    {
+      "epoch": 0.5221785513756316,
+      "grad_norm": 0.28024089336395264,
+      "learning_rate": 9.834667527551341e-06,
+      "loss": 0.3896,
+      "step": 620
+    },
+    {
+      "epoch": 0.5230207748455924,
+      "grad_norm": 0.2571643590927124,
+      "learning_rate": 9.833414897752346e-06,
+      "loss": 0.3892,
+      "step": 621
+    },
+    {
+      "epoch": 0.523862998315553,
+      "grad_norm": 0.2675620913505554,
+      "learning_rate": 9.832157620995088e-06,
+      "loss": 0.4076,
+      "step": 622
+    },
+    {
+      "epoch": 0.5247052217855137,
+      "grad_norm": 0.26316148042678833,
+      "learning_rate": 9.830895698488341e-06,
+      "loss": 0.3937,
+      "step": 623
+    },
+    {
+      "epoch": 0.5255474452554745,
+      "grad_norm": 0.3073652684688568,
+      "learning_rate": 9.829629131445342e-06,
+      "loss": 0.4122,
+      "step": 624
+    },
+    {
+      "epoch": 0.5263896687254351,
+      "grad_norm": 0.24735543131828308,
+      "learning_rate": 9.828357921083803e-06,
+      "loss": 0.3816,
+      "step": 625
+    },
+    {
+      "epoch": 0.5272318921953958,
+      "grad_norm": 0.3236400783061981,
+      "learning_rate": 9.827082068625893e-06,
+      "loss": 0.3938,
+      "step": 626
+    },
+    {
+      "epoch": 0.5280741156653566,
+      "grad_norm": 0.339495986700058,
+      "learning_rate": 9.825801575298248e-06,
+      "loss": 0.4254,
+      "step": 627
+    },
+    {
+      "epoch": 0.5289163391353172,
+      "grad_norm": 0.2840985655784607,
+      "learning_rate": 9.824516442331963e-06,
+      "loss": 0.423,
+      "step": 628
+    },
+    {
+      "epoch": 0.529758562605278,
+      "grad_norm": 0.3441496789455414,
+      "learning_rate": 9.823226670962598e-06,
+      "loss": 0.3984,
+      "step": 629
+    },
+    {
+      "epoch": 0.5306007860752386,
+      "grad_norm": 0.26104456186294556,
+      "learning_rate": 9.821932262430164e-06,
+      "loss": 0.4058,
+      "step": 630
+    },
+    {
+      "epoch": 0.5314430095451993,
+      "grad_norm": 0.2665572762489319,
+      "learning_rate": 9.82063321797914e-06,
+      "loss": 0.3975,
+      "step": 631
+    },
+    {
+      "epoch": 0.5322852330151601,
+      "grad_norm": 0.25185590982437134,
+      "learning_rate": 9.819329538858458e-06,
+      "loss": 0.4018,
+      "step": 632
+    },
+    {
+      "epoch": 0.5331274564851207,
+      "grad_norm": 0.25625938177108765,
+      "learning_rate": 9.818021226321502e-06,
+      "loss": 0.3823,
+      "step": 633
+    },
+    {
+      "epoch": 0.5339696799550814,
+      "grad_norm": 0.32790085673332214,
+      "learning_rate": 9.816708281626116e-06,
+      "loss": 0.3869,
+      "step": 634
+    },
+    {
+      "epoch": 0.5348119034250421,
+      "grad_norm": 0.25382882356643677,
+      "learning_rate": 9.815390706034598e-06,
+      "loss": 0.415,
+      "step": 635
+    },
+    {
+      "epoch": 0.5356541268950028,
+      "grad_norm": 0.2376747578382492,
+      "learning_rate": 9.814068500813692e-06,
+      "loss": 0.4161,
+      "step": 636
+    },
+    {
+      "epoch": 0.5364963503649635,
+      "grad_norm": 0.25517722964286804,
+      "learning_rate": 9.812741667234599e-06,
+      "loss": 0.3881,
+      "step": 637
+    },
+    {
+      "epoch": 0.5373385738349242,
+      "grad_norm": 0.24864782392978668,
+      "learning_rate": 9.811410206572972e-06,
+      "loss": 0.3703,
+      "step": 638
+    },
+    {
+      "epoch": 0.5381807973048849,
+      "grad_norm": 0.2584892809391022,
+      "learning_rate": 9.8100741201089e-06,
+      "loss": 0.4057,
+      "step": 639
+    },
+    {
+      "epoch": 0.5390230207748455,
+      "grad_norm": 0.2750840187072754,
+      "learning_rate": 9.808733409126934e-06,
+      "loss": 0.4113,
+      "step": 640
+    },
+    {
+      "epoch": 0.5398652442448063,
+      "grad_norm": 0.24435681104660034,
+      "learning_rate": 9.807388074916064e-06,
+      "loss": 0.3987,
+      "step": 641
+    },
+    {
+      "epoch": 0.540707467714767,
+      "grad_norm": 0.2840283513069153,
+      "learning_rate": 9.806038118769724e-06,
+      "loss": 0.3964,
+      "step": 642
+    },
+    {
+      "epoch": 0.5415496911847277,
+      "grad_norm": 0.2739749848842621,
+      "learning_rate": 9.804683541985796e-06,
+      "loss": 0.4262,
+      "step": 643
+    },
+    {
+      "epoch": 0.5423919146546884,
+      "grad_norm": 0.25117605924606323,
+      "learning_rate": 9.803324345866599e-06,
+      "loss": 0.3956,
+      "step": 644
+    },
+    {
+      "epoch": 0.543234138124649,
+      "grad_norm": 0.26271775364875793,
+      "learning_rate": 9.801960531718898e-06,
+      "loss": 0.4025,
+      "step": 645
+    },
+    {
+      "epoch": 0.5440763615946098,
+      "grad_norm": 0.25807446241378784,
+      "learning_rate": 9.800592100853894e-06,
+      "loss": 0.4004,
+      "step": 646
+    },
+    {
+      "epoch": 0.5449185850645705,
+      "grad_norm": 0.25606074929237366,
+      "learning_rate": 9.79921905458723e-06,
+      "loss": 0.3868,
+      "step": 647
+    },
+    {
+      "epoch": 0.5457608085345311,
+      "grad_norm": 0.27761006355285645,
+      "learning_rate": 9.797841394238987e-06,
+      "loss": 0.4067,
+      "step": 648
+    },
+    {
+      "epoch": 0.5466030320044919,
+      "grad_norm": 0.2372005730867386,
+      "learning_rate": 9.796459121133675e-06,
+      "loss": 0.4075,
+      "step": 649
+    },
+    {
+      "epoch": 0.5474452554744526,
+      "grad_norm": 0.2324095219373703,
+      "learning_rate": 9.795072236600247e-06,
+      "loss": 0.3867,
+      "step": 650
+    },
+    {
+      "epoch": 0.5482874789444132,
+      "grad_norm": 0.2631465792655945,
+      "learning_rate": 9.793680741972084e-06,
+      "loss": 0.402,
+      "step": 651
+    },
+    {
+      "epoch": 0.549129702414374,
+      "grad_norm": 0.2458457201719284,
+      "learning_rate": 9.792284638587005e-06,
+      "loss": 0.3924,
+      "step": 652
+    },
+    {
+      "epoch": 0.5499719258843346,
+      "grad_norm": 0.2652607858181,
+      "learning_rate": 9.790883927787254e-06,
+      "loss": 0.3838,
+      "step": 653
+    },
+    {
+      "epoch": 0.5508141493542953,
+      "grad_norm": 0.2583690881729126,
+      "learning_rate": 9.789478610919508e-06,
+      "loss": 0.3694,
+      "step": 654
+    },
+    {
+      "epoch": 0.5516563728242561,
+      "grad_norm": 0.3279140591621399,
+      "learning_rate": 9.78806868933487e-06,
+      "loss": 0.398,
+      "step": 655
+    },
+    {
+      "epoch": 0.5524985962942167,
+      "grad_norm": 0.22404353320598602,
+      "learning_rate": 9.786654164388873e-06,
+      "loss": 0.3939,
+      "step": 656
+    },
+    {
+      "epoch": 0.5533408197641775,
+      "grad_norm": 0.2810705602169037,
+      "learning_rate": 9.785235037441473e-06,
+      "loss": 0.3991,
+      "step": 657
+    },
+    {
+      "epoch": 0.5541830432341381,
+      "grad_norm": 0.26564010977745056,
+      "learning_rate": 9.783811309857057e-06,
+      "loss": 0.4042,
+      "step": 658
+    },
+    {
+      "epoch": 0.5550252667040988,
+      "grad_norm": 0.2543601393699646,
+      "learning_rate": 9.782382983004424e-06,
+      "loss": 0.3967,
+      "step": 659
+    },
+    {
+      "epoch": 0.5558674901740596,
+      "grad_norm": 0.27026471495628357,
+      "learning_rate": 9.780950058256802e-06,
+      "loss": 0.4063,
+      "step": 660
+    },
+    {
+      "epoch": 0.5567097136440202,
+      "grad_norm": 0.2524589002132416,
+      "learning_rate": 9.779512536991839e-06,
+      "loss": 0.4245,
+      "step": 661
+    },
+    {
+      "epoch": 0.5575519371139809,
+      "grad_norm": 0.2756974697113037,
+      "learning_rate": 9.778070420591603e-06,
+      "loss": 0.4059,
+      "step": 662
+    },
+    {
+      "epoch": 0.5583941605839416,
+      "grad_norm": 0.23817111551761627,
+      "learning_rate": 9.77662371044258e-06,
+      "loss": 0.3775,
+      "step": 663
+    },
+    {
+      "epoch": 0.5592363840539023,
+      "grad_norm": 0.27896976470947266,
+      "learning_rate": 9.775172407935664e-06,
+      "loss": 0.4178,
+      "step": 664
+    },
+    {
+      "epoch": 0.560078607523863,
+      "grad_norm": 0.282566636800766,
+      "learning_rate": 9.773716514466179e-06,
+      "loss": 0.4463,
+      "step": 665
+    },
+    {
+      "epoch": 0.5609208309938237,
+      "grad_norm": 0.2578970491886139,
+      "learning_rate": 9.77225603143385e-06,
+      "loss": 0.3849,
+      "step": 666
+    },
+    {
+      "epoch": 0.5617630544637844,
+      "grad_norm": 0.2897636592388153,
+      "learning_rate": 9.770790960242821e-06,
+      "loss": 0.4002,
+      "step": 667
+    },
+    {
+      "epoch": 0.562605277933745,
+      "grad_norm": 0.23036324977874756,
+      "learning_rate": 9.769321302301648e-06,
+      "loss": 0.3838,
+      "step": 668
+    },
+    {
+      "epoch": 0.5634475014037058,
+      "grad_norm": 0.2721245586872101,
+      "learning_rate": 9.767847059023292e-06,
+      "loss": 0.3993,
+      "step": 669
+    },
+    {
+      "epoch": 0.5642897248736665,
+      "grad_norm": 0.266300767660141,
+      "learning_rate": 9.766368231825126e-06,
+      "loss": 0.4184,
+      "step": 670
+    },
+    {
+      "epoch": 0.5651319483436271,
+      "grad_norm": 0.3153010606765747,
+      "learning_rate": 9.764884822128928e-06,
+      "loss": 0.3748,
+      "step": 671
+    },
+    {
+      "epoch": 0.5659741718135879,
+      "grad_norm": 0.2557030916213989,
+      "learning_rate": 9.763396831360884e-06,
+      "loss": 0.3771,
+      "step": 672
+    },
+    {
+      "epoch": 0.5668163952835485,
+      "grad_norm": 0.2733110189437866,
+      "learning_rate": 9.761904260951583e-06,
+      "loss": 0.3881,
+      "step": 673
+    },
+    {
+      "epoch": 0.5676586187535093,
+      "grad_norm": 0.2519077658653259,
+      "learning_rate": 9.760407112336016e-06,
+      "loss": 0.3807,
+      "step": 674
+    },
+    {
+      "epoch": 0.56850084222347,
+      "grad_norm": 0.2971459627151489,
+      "learning_rate": 9.75890538695358e-06,
+      "loss": 0.4062,
+      "step": 675
+    },
+    {
+      "epoch": 0.5693430656934306,
+      "grad_norm": 0.27987900376319885,
+      "learning_rate": 9.757399086248062e-06,
+      "loss": 0.3901,
+      "step": 676
+    },
+    {
+      "epoch": 0.5701852891633914,
+      "grad_norm": 0.24134734272956848,
+      "learning_rate": 9.755888211667663e-06,
+      "loss": 0.3777,
+      "step": 677
+    },
+    {
+      "epoch": 0.571027512633352,
+      "grad_norm": 0.2651788592338562,
+      "learning_rate": 9.75437276466497e-06,
+      "loss": 0.4159,
+      "step": 678
+    },
+    {
+      "epoch": 0.5718697361033127,
+      "grad_norm": 0.22584442794322968,
+      "learning_rate": 9.752852746696968e-06,
+      "loss": 0.3922,
+      "step": 679
+    },
+    {
+      "epoch": 0.5727119595732735,
+      "grad_norm": 0.2702990770339966,
+      "learning_rate": 9.751328159225037e-06,
+      "loss": 0.3937,
+      "step": 680
+    },
+    {
+      "epoch": 0.5735541830432341,
+      "grad_norm": 0.27012670040130615,
+      "learning_rate": 9.749799003714954e-06,
+      "loss": 0.4118,
+      "step": 681
+    },
+    {
+      "epoch": 0.5743964065131948,
+      "grad_norm": 0.24326875805854797,
+      "learning_rate": 9.748265281636885e-06,
+      "loss": 0.4085,
+      "step": 682
+    },
+    {
+      "epoch": 0.5752386299831556,
+      "grad_norm": 0.257748544216156,
+      "learning_rate": 9.746726994465383e-06,
+      "loss": 0.3849,
+      "step": 683
+    },
+    {
+      "epoch": 0.5760808534531162,
+      "grad_norm": 0.2539883255958557,
+      "learning_rate": 9.745184143679398e-06,
+      "loss": 0.3977,
+      "step": 684
+    },
+    {
+      "epoch": 0.5769230769230769,
+      "grad_norm": 0.21800824999809265,
+      "learning_rate": 9.743636730762259e-06,
+      "loss": 0.4062,
+      "step": 685
+    },
+    {
+      "epoch": 0.5777653003930376,
+      "grad_norm": 0.2953537106513977,
+      "learning_rate": 9.742084757201684e-06,
+      "loss": 0.4001,
+      "step": 686
+    },
+    {
+      "epoch": 0.5786075238629983,
+      "grad_norm": 0.2531120777130127,
+      "learning_rate": 9.74052822448978e-06,
+      "loss": 0.3975,
+      "step": 687
+    },
+    {
+      "epoch": 0.5794497473329591,
+      "grad_norm": 0.25123146176338196,
+      "learning_rate": 9.738967134123035e-06,
+      "loss": 0.3841,
+      "step": 688
+    },
+    {
+      "epoch": 0.5802919708029197,
+      "grad_norm": 0.3006608188152313,
+      "learning_rate": 9.737401487602314e-06,
+      "loss": 0.3995,
+      "step": 689
+    },
+    {
+      "epoch": 0.5811341942728804,
+      "grad_norm": 0.2671349346637726,
+      "learning_rate": 9.735831286432869e-06,
+      "loss": 0.4051,
+      "step": 690
+    },
+    {
+      "epoch": 0.5819764177428411,
+      "grad_norm": 0.2805735170841217,
+      "learning_rate": 9.734256532124326e-06,
+      "loss": 0.3813,
+      "step": 691
+    },
+    {
+      "epoch": 0.5828186412128018,
+      "grad_norm": 0.25094088912010193,
+      "learning_rate": 9.732677226190692e-06,
+      "loss": 0.3693,
+      "step": 692
+    },
+    {
+      "epoch": 0.5836608646827625,
+      "grad_norm": 0.2721113860607147,
+      "learning_rate": 9.731093370150349e-06,
+      "loss": 0.4002,
+      "step": 693
+    },
+    {
+      "epoch": 0.5845030881527232,
+      "grad_norm": 0.24327607452869415,
+      "learning_rate": 9.729504965526053e-06,
+      "loss": 0.3991,
+      "step": 694
+    },
+    {
+      "epoch": 0.5853453116226839,
+      "grad_norm": 0.25750425457954407,
+      "learning_rate": 9.727912013844933e-06,
+      "loss": 0.4134,
+      "step": 695
+    },
+    {
+      "epoch": 0.5861875350926445,
+      "grad_norm": 0.25728413462638855,
+      "learning_rate": 9.72631451663849e-06,
+      "loss": 0.3895,
+      "step": 696
+    },
+    {
+      "epoch": 0.5870297585626053,
+      "grad_norm": 0.3020685613155365,
+      "learning_rate": 9.724712475442597e-06,
+      "loss": 0.3877,
+      "step": 697
+    },
+    {
+      "epoch": 0.587871982032566,
+      "grad_norm": 0.23866966366767883,
+      "learning_rate": 9.72310589179749e-06,
+      "loss": 0.4152,
+      "step": 698
+    },
+    {
+      "epoch": 0.5887142055025266,
+      "grad_norm": 0.28489524126052856,
+      "learning_rate": 9.721494767247779e-06,
+      "loss": 0.39,
+      "step": 699
+    },
+    {
+      "epoch": 0.5895564289724874,
+      "grad_norm": 0.27395954728126526,
+      "learning_rate": 9.719879103342438e-06,
+      "loss": 0.3756,
+      "step": 700
+    },
+    {
+      "epoch": 0.590398652442448,
+      "grad_norm": 0.28621479868888855,
+      "learning_rate": 9.718258901634802e-06,
+      "loss": 0.3948,
+      "step": 701
+    },
+    {
+      "epoch": 0.5912408759124088,
+      "grad_norm": 0.23523105680942535,
+      "learning_rate": 9.71663416368257e-06,
+      "loss": 0.385,
+      "step": 702
+    },
+    {
+      "epoch": 0.5920830993823695,
+      "grad_norm": 0.2857535481452942,
+      "learning_rate": 9.715004891047805e-06,
+      "loss": 0.3767,
+      "step": 703
+    },
+    {
+      "epoch": 0.5929253228523301,
+      "grad_norm": 0.26084885001182556,
+      "learning_rate": 9.71337108529693e-06,
+      "loss": 0.3769,
+      "step": 704
+    },
+    {
+      "epoch": 0.5937675463222909,
+      "grad_norm": 0.276625394821167,
+      "learning_rate": 9.71173274800072e-06,
+      "loss": 0.3924,
+      "step": 705
+    },
+    {
+      "epoch": 0.5946097697922516,
+      "grad_norm": 0.239909365773201,
+      "learning_rate": 9.71008988073431e-06,
+      "loss": 0.3989,
+      "step": 706
+    },
+    {
+      "epoch": 0.5954519932622122,
+      "grad_norm": 0.2926711142063141,
+      "learning_rate": 9.708442485077197e-06,
+      "loss": 0.3969,
+      "step": 707
+    },
+    {
+      "epoch": 0.596294216732173,
+      "grad_norm": 0.2879006266593933,
+      "learning_rate": 9.70679056261322e-06,
+      "loss": 0.3816,
+      "step": 708
+    },
+    {
+      "epoch": 0.5971364402021336,
+      "grad_norm": 0.24086812138557434,
+      "learning_rate": 9.70513411493058e-06,
+      "loss": 0.3813,
+      "step": 709
+    },
+    {
+      "epoch": 0.5979786636720943,
+      "grad_norm": 0.28989389538764954,
+      "learning_rate": 9.70347314362182e-06,
+      "loss": 0.3992,
+      "step": 710
+    },
+    {
+      "epoch": 0.5988208871420551,
+      "grad_norm": 0.25863921642303467,
+      "learning_rate": 9.70180765028384e-06,
+      "loss": 0.4085,
+      "step": 711
+    },
+    {
+      "epoch": 0.5996631106120157,
+      "grad_norm": 0.2894207239151001,
+      "learning_rate": 9.700137636517884e-06,
+      "loss": 0.4016,
+      "step": 712
+    },
+    {
+      "epoch": 0.6005053340819764,
+      "grad_norm": 0.25992056727409363,
+      "learning_rate": 9.698463103929542e-06,
+      "loss": 0.3675,
+      "step": 713
+    },
+    {
+      "epoch": 0.6013475575519371,
+      "grad_norm": 0.28006356954574585,
+      "learning_rate": 9.696784054128749e-06,
+      "loss": 0.393,
+      "step": 714
+    },
+    {
+      "epoch": 0.6021897810218978,
+      "grad_norm": 0.2698265612125397,
+      "learning_rate": 9.695100488729784e-06,
+      "loss": 0.4124,
+      "step": 715
+    },
+    {
+      "epoch": 0.6030320044918585,
+      "grad_norm": 0.24214014410972595,
+      "learning_rate": 9.693412409351264e-06,
+      "loss": 0.4018,
+      "step": 716
+    },
+    {
+      "epoch": 0.6038742279618192,
+      "grad_norm": 0.2883848547935486,
+      "learning_rate": 9.691719817616148e-06,
+      "loss": 0.4001,
+      "step": 717
+    },
+    {
+      "epoch": 0.6047164514317799,
+      "grad_norm": 0.27353766560554504,
+      "learning_rate": 9.690022715151734e-06,
+      "loss": 0.408,
+      "step": 718
+    },
+    {
+      "epoch": 0.6055586749017406,
+      "grad_norm": 0.253030389547348,
+      "learning_rate": 9.688321103589659e-06,
+      "loss": 0.3887,
+      "step": 719
+    },
+    {
+      "epoch": 0.6064008983717013,
+      "grad_norm": 0.255454957485199,
+      "learning_rate": 9.686614984565888e-06,
+      "loss": 0.3715,
+      "step": 720
+    },
+    {
+      "epoch": 0.607243121841662,
+      "grad_norm": 0.25838613510131836,
+      "learning_rate": 9.684904359720724e-06,
+      "loss": 0.4027,
+      "step": 721
+    },
+    {
+      "epoch": 0.6080853453116227,
+      "grad_norm": 0.23131829500198364,
+      "learning_rate": 9.683189230698804e-06,
+      "loss": 0.4158,
+      "step": 722
+    },
+    {
+      "epoch": 0.6089275687815834,
+      "grad_norm": 0.26560577750205994,
+      "learning_rate": 9.681469599149093e-06,
+      "loss": 0.3927,
+      "step": 723
+    },
+    {
+      "epoch": 0.609769792251544,
+      "grad_norm": 0.2801228165626526,
+      "learning_rate": 9.679745466724884e-06,
+      "loss": 0.3835,
+      "step": 724
+    },
+    {
+      "epoch": 0.6106120157215048,
+      "grad_norm": 0.2449226677417755,
+      "learning_rate": 9.678016835083798e-06,
+      "loss": 0.4011,
+      "step": 725
+    },
+    {
+      "epoch": 0.6114542391914655,
+      "grad_norm": 0.2523360252380371,
+      "learning_rate": 9.676283705887783e-06,
+      "loss": 0.3691,
+      "step": 726
+    },
+    {
+      "epoch": 0.6122964626614261,
+      "grad_norm": 0.2678155303001404,
+      "learning_rate": 9.674546080803109e-06,
+      "loss": 0.3874,
+      "step": 727
+    },
+    {
+      "epoch": 0.6131386861313869,
+      "grad_norm": 0.2483142614364624,
+      "learning_rate": 9.67280396150037e-06,
+      "loss": 0.4032,
+      "step": 728
+    },
+    {
+      "epoch": 0.6139809096013475,
+      "grad_norm": 0.29895925521850586,
+      "learning_rate": 9.671057349654481e-06,
+      "loss": 0.3938,
+      "step": 729
+    },
+    {
+      "epoch": 0.6148231330713082,
+      "grad_norm": 0.222129225730896,
+      "learning_rate": 9.669306246944674e-06,
+      "loss": 0.3928,
+      "step": 730
+    },
+    {
+      "epoch": 0.615665356541269,
+      "grad_norm": 0.29450759291648865,
+      "learning_rate": 9.6675506550545e-06,
+      "loss": 0.4152,
+      "step": 731
+    },
+    {
+      "epoch": 0.6165075800112296,
+      "grad_norm": 0.2805439829826355,
+      "learning_rate": 9.66579057567183e-06,
+      "loss": 0.4006,
+      "step": 732
+    },
+    {
+      "epoch": 0.6173498034811904,
+      "grad_norm": 0.234604150056839,
+      "learning_rate": 9.66402601048884e-06,
+      "loss": 0.3831,
+      "step": 733
+    },
+    {
+      "epoch": 0.618192026951151,
+      "grad_norm": 0.31237849593162537,
+      "learning_rate": 9.662256961202028e-06,
+      "loss": 0.3922,
+      "step": 734
+    },
+    {
+      "epoch": 0.6190342504211117,
+      "grad_norm": 0.2697100043296814,
+      "learning_rate": 9.660483429512198e-06,
+      "loss": 0.392,
+      "step": 735
+    },
+    {
+      "epoch": 0.6198764738910725,
+      "grad_norm": 0.31797346472740173,
+      "learning_rate": 9.658705417124466e-06,
+      "loss": 0.3838,
+      "step": 736
+    },
+    {
+      "epoch": 0.6207186973610331,
+      "grad_norm": 0.275126188993454,
+      "learning_rate": 9.656922925748254e-06,
+      "loss": 0.3969,
+      "step": 737
+    },
+    {
+      "epoch": 0.6215609208309938,
+      "grad_norm": 0.2596983015537262,
+      "learning_rate": 9.65513595709729e-06,
+      "loss": 0.3882,
+      "step": 738
+    },
+    {
+      "epoch": 0.6224031443009546,
+      "grad_norm": 0.2853798568248749,
+      "learning_rate": 9.653344512889608e-06,
+      "loss": 0.4083,
+      "step": 739
+    },
+    {
+      "epoch": 0.6232453677709152,
+      "grad_norm": 0.23947489261627197,
+      "learning_rate": 9.651548594847546e-06,
+      "loss": 0.3982,
+      "step": 740
+    },
+    {
+      "epoch": 0.6240875912408759,
+      "grad_norm": 0.26423704624176025,
+      "learning_rate": 9.649748204697741e-06,
+      "loss": 0.3913,
+      "step": 741
+    },
+    {
+      "epoch": 0.6249298147108366,
+      "grad_norm": 0.2359628826379776,
+      "learning_rate": 9.647943344171129e-06,
+      "loss": 0.3866,
+      "step": 742
+    },
+    {
+      "epoch": 0.6257720381807973,
+      "grad_norm": 0.2493540644645691,
+      "learning_rate": 9.646134015002946e-06,
+      "loss": 0.3946,
+      "step": 743
+    },
+    {
+      "epoch": 0.626614261650758,
+      "grad_norm": 0.2681938111782074,
+      "learning_rate": 9.644320218932723e-06,
+      "loss": 0.4084,
+      "step": 744
+    },
+    {
+      "epoch": 0.6274564851207187,
+      "grad_norm": 0.2515985667705536,
+      "learning_rate": 9.642501957704287e-06,
+      "loss": 0.3729,
+      "step": 745
+    },
+    {
+      "epoch": 0.6282987085906794,
+      "grad_norm": 0.25220364332199097,
+      "learning_rate": 9.640679233065755e-06,
+      "loss": 0.3944,
+      "step": 746
+    },
+    {
+      "epoch": 0.62914093206064,
+      "grad_norm": 0.2675838768482208,
+      "learning_rate": 9.63885204676954e-06,
+      "loss": 0.3733,
+      "step": 747
+    },
+    {
+      "epoch": 0.6299831555306008,
+      "grad_norm": 0.2608799636363983,
+      "learning_rate": 9.637020400572339e-06,
+      "loss": 0.4032,
+      "step": 748
+    },
+    {
+      "epoch": 0.6308253790005615,
+      "grad_norm": 0.23328998684883118,
+      "learning_rate": 9.63518429623514e-06,
+      "loss": 0.4149,
+      "step": 749
+    },
+    {
+      "epoch": 0.6316676024705222,
+      "grad_norm": 0.23251040279865265,
+      "learning_rate": 9.63334373552322e-06,
+      "loss": 0.3944,
+      "step": 750
+    },
+    {
+      "epoch": 0.6325098259404829,
+      "grad_norm": 0.2581608295440674,
+      "learning_rate": 9.631498720206132e-06,
+      "loss": 0.4023,
+      "step": 751
+    },
+    {
+      "epoch": 0.6333520494104435,
+      "grad_norm": 0.2426811009645462,
+      "learning_rate": 9.62964925205772e-06,
+      "loss": 0.3795,
+      "step": 752
+    },
+    {
+      "epoch": 0.6341942728804043,
+      "grad_norm": 0.25329363346099854,
+      "learning_rate": 9.627795332856107e-06,
+      "loss": 0.3948,
+      "step": 753
+    },
+    {
+      "epoch": 0.635036496350365,
+      "grad_norm": 0.25250154733657837,
+      "learning_rate": 9.625936964383691e-06,
+      "loss": 0.3767,
+      "step": 754
+    },
+    {
+      "epoch": 0.6358787198203256,
+      "grad_norm": 0.2705284357070923,
+      "learning_rate": 9.624074148427154e-06,
+      "loss": 0.3975,
+      "step": 755
+    },
+    {
+      "epoch": 0.6367209432902864,
+      "grad_norm": 0.24624527990818024,
+      "learning_rate": 9.622206886777448e-06,
+      "loss": 0.3792,
+      "step": 756
+    },
+    {
+      "epoch": 0.637563166760247,
+      "grad_norm": 0.2937239706516266,
+      "learning_rate": 9.620335181229805e-06,
+      "loss": 0.397,
+      "step": 757
+    },
+    {
+      "epoch": 0.6384053902302077,
+      "grad_norm": 0.24023059010505676,
+      "learning_rate": 9.618459033583725e-06,
+      "loss": 0.4001,
+      "step": 758
+    },
+    {
+      "epoch": 0.6392476137001685,
+      "grad_norm": 0.278189092874527,
+      "learning_rate": 9.616578445642982e-06,
+      "loss": 0.3803,
+      "step": 759
+    },
+    {
+      "epoch": 0.6400898371701291,
+      "grad_norm": 0.31741783022880554,
+      "learning_rate": 9.614693419215613e-06,
+      "loss": 0.3926,
+      "step": 760
+    },
+    {
+      "epoch": 0.6409320606400898,
+      "grad_norm": 0.26135605573654175,
+      "learning_rate": 9.612803956113932e-06,
+      "loss": 0.3572,
+      "step": 761
+    },
+    {
+      "epoch": 0.6417742841100506,
+      "grad_norm": 0.2977483868598938,
+      "learning_rate": 9.61091005815451e-06,
+      "loss": 0.3933,
+      "step": 762
+    },
+    {
+      "epoch": 0.6426165075800112,
+      "grad_norm": 0.29880109429359436,
+      "learning_rate": 9.609011727158184e-06,
+      "loss": 0.4027,
+      "step": 763
+    },
+    {
+      "epoch": 0.643458731049972,
+      "grad_norm": 0.23161263763904572,
+      "learning_rate": 9.607108964950056e-06,
+      "loss": 0.3636,
+      "step": 764
+    },
+    {
+      "epoch": 0.6443009545199326,
+      "grad_norm": 0.24532750248908997,
+      "learning_rate": 9.605201773359485e-06,
+      "loss": 0.3887,
+      "step": 765
+    },
+    {
+      "epoch": 0.6451431779898933,
+      "grad_norm": 0.2845829725265503,
+      "learning_rate": 9.603290154220091e-06,
+      "loss": 0.3963,
+      "step": 766
+    },
+    {
+      "epoch": 0.6459854014598541,
+      "grad_norm": 0.24565917253494263,
+      "learning_rate": 9.601374109369746e-06,
+      "loss": 0.3946,
+      "step": 767
+    },
+    {
+      "epoch": 0.6468276249298147,
+      "grad_norm": 0.24358156323432922,
+      "learning_rate": 9.599453640650585e-06,
+      "loss": 0.3983,
+      "step": 768
+    },
+    {
+      "epoch": 0.6476698483997754,
+      "grad_norm": 0.2880096137523651,
+      "learning_rate": 9.59752874990899e-06,
+      "loss": 0.3927,
+      "step": 769
+    },
+    {
+      "epoch": 0.6485120718697361,
+      "grad_norm": 0.25155845284461975,
+      "learning_rate": 9.595599438995593e-06,
+      "loss": 0.4066,
+      "step": 770
+    },
+    {
+      "epoch": 0.6493542953396968,
+      "grad_norm": 0.27330100536346436,
+      "learning_rate": 9.59366570976528e-06,
+      "loss": 0.4022,
+      "step": 771
+    },
+    {
+      "epoch": 0.6501965188096575,
+      "grad_norm": 0.2611733078956604,
+      "learning_rate": 9.591727564077189e-06,
+      "loss": 0.3915,
+      "step": 772
+    },
+    {
+      "epoch": 0.6510387422796182,
+      "grad_norm": 0.2380668818950653,
+      "learning_rate": 9.589785003794692e-06,
+      "loss": 0.4014,
+      "step": 773
+    },
+    {
+      "epoch": 0.6518809657495789,
+      "grad_norm": 0.2736964821815491,
+      "learning_rate": 9.587838030785413e-06,
+      "loss": 0.394,
+      "step": 774
+    },
+    {
+      "epoch": 0.6527231892195395,
+      "grad_norm": 0.23338758945465088,
+      "learning_rate": 9.585886646921221e-06,
+      "loss": 0.3826,
+      "step": 775
+    },
+    {
+      "epoch": 0.6535654126895003,
+      "grad_norm": 0.2619832754135132,
+      "learning_rate": 9.583930854078219e-06,
+      "loss": 0.4194,
+      "step": 776
+    },
+    {
+      "epoch": 0.654407636159461,
+      "grad_norm": 0.2791779339313507,
+      "learning_rate": 9.581970654136752e-06,
+      "loss": 0.3891,
+      "step": 777
+    },
+    {
+      "epoch": 0.6552498596294217,
+      "grad_norm": 0.2457047700881958,
+      "learning_rate": 9.580006048981403e-06,
+      "loss": 0.3794,
+      "step": 778
+    },
+    {
+      "epoch": 0.6560920830993824,
+      "grad_norm": 0.24340137839317322,
+      "learning_rate": 9.578037040500992e-06,
+      "loss": 0.396,
+      "step": 779
+    },
+    {
+      "epoch": 0.656934306569343,
+      "grad_norm": 0.3096691071987152,
+      "learning_rate": 9.576063630588563e-06,
+      "loss": 0.3789,
+      "step": 780
+    },
+    {
+      "epoch": 0.6577765300393038,
+      "grad_norm": 0.23291133344173431,
+      "learning_rate": 9.574085821141406e-06,
+      "loss": 0.393,
+      "step": 781
+    },
+    {
+      "epoch": 0.6586187535092645,
+      "grad_norm": 0.2542714476585388,
+      "learning_rate": 9.572103614061029e-06,
+      "loss": 0.385,
+      "step": 782
+    },
+    {
+      "epoch": 0.6594609769792251,
+      "grad_norm": 0.26308146119117737,
+      "learning_rate": 9.570117011253173e-06,
+      "loss": 0.4036,
+      "step": 783
+    },
+    {
+      "epoch": 0.6603032004491859,
+      "grad_norm": 0.2650177776813507,
+      "learning_rate": 9.568126014627805e-06,
+      "loss": 0.4113,
+      "step": 784
+    },
+    {
+      "epoch": 0.6611454239191465,
+      "grad_norm": 0.25345557928085327,
+      "learning_rate": 9.566130626099118e-06,
+      "loss": 0.3891,
+      "step": 785
+    },
+    {
+      "epoch": 0.6619876473891072,
+      "grad_norm": 0.2507096529006958,
+      "learning_rate": 9.56413084758552e-06,
+      "loss": 0.3891,
+      "step": 786
+    },
+    {
+      "epoch": 0.662829870859068,
+      "grad_norm": 0.24637813866138458,
+      "learning_rate": 9.562126681009649e-06,
+      "loss": 0.3737,
+      "step": 787
+    },
+    {
+      "epoch": 0.6636720943290286,
+      "grad_norm": 0.2659986615180969,
+      "learning_rate": 9.560118128298355e-06,
+      "loss": 0.4135,
+      "step": 788
+    },
+    {
+      "epoch": 0.6645143177989893,
+      "grad_norm": 0.2727932035923004,
+      "learning_rate": 9.55810519138271e-06,
+      "loss": 0.3974,
+      "step": 789
+    },
+    {
+      "epoch": 0.66535654126895,
+      "grad_norm": 0.26761728525161743,
+      "learning_rate": 9.556087872197997e-06,
+      "loss": 0.3718,
+      "step": 790
+    },
+    {
+      "epoch": 0.6661987647389107,
+      "grad_norm": 0.28363317251205444,
+      "learning_rate": 9.554066172683715e-06,
+      "loss": 0.4138,
+      "step": 791
+    },
+    {
+      "epoch": 0.6670409882088714,
+      "grad_norm": 0.2563317120075226,
+      "learning_rate": 9.552040094783575e-06,
+      "loss": 0.4085,
+      "step": 792
+    },
+    {
+      "epoch": 0.6678832116788321,
+      "grad_norm": 0.29577916860580444,
+      "learning_rate": 9.550009640445492e-06,
+      "loss": 0.3808,
+      "step": 793
+    },
+    {
+      "epoch": 0.6687254351487928,
+      "grad_norm": 0.30510273575782776,
+      "learning_rate": 9.547974811621594e-06,
+      "loss": 0.3998,
+      "step": 794
+    },
+    {
+      "epoch": 0.6695676586187536,
+      "grad_norm": 0.2771177887916565,
+      "learning_rate": 9.545935610268213e-06,
+      "loss": 0.3904,
+      "step": 795
+    },
+    {
+      "epoch": 0.6704098820887142,
+      "grad_norm": 0.23797883093357086,
+      "learning_rate": 9.543892038345885e-06,
+      "loss": 0.3888,
+      "step": 796
+    },
+    {
+      "epoch": 0.6712521055586749,
+      "grad_norm": 0.276486873626709,
+      "learning_rate": 9.541844097819347e-06,
+      "loss": 0.3909,
+      "step": 797
+    },
+    {
+      "epoch": 0.6720943290286356,
+      "grad_norm": 0.25211796164512634,
+      "learning_rate": 9.53979179065754e-06,
+      "loss": 0.4019,
+      "step": 798
+    },
+    {
+      "epoch": 0.6729365524985963,
+      "grad_norm": 0.2426571249961853,
+      "learning_rate": 9.537735118833595e-06,
+      "loss": 0.3907,
+      "step": 799
+    },
+    {
+      "epoch": 0.673778775968557,
+      "grad_norm": 0.23780083656311035,
+      "learning_rate": 9.53567408432485e-06,
+      "loss": 0.3757,
+      "step": 800
+    },
+    {
+      "epoch": 0.6746209994385177,
+      "grad_norm": 0.3084465265274048,
+      "learning_rate": 9.533608689112827e-06,
+      "loss": 0.3994,
+      "step": 801
+    },
+    {
+      "epoch": 0.6754632229084784,
+      "grad_norm": 0.2585592567920685,
+      "learning_rate": 9.531538935183252e-06,
+      "loss": 0.4141,
+      "step": 802
+    },
+    {
+      "epoch": 0.676305446378439,
+      "grad_norm": 0.23807695508003235,
+      "learning_rate": 9.529464824526027e-06,
+      "loss": 0.4032,
+      "step": 803
+    },
+    {
+      "epoch": 0.6771476698483998,
+      "grad_norm": 0.2669808864593506,
+      "learning_rate": 9.527386359135254e-06,
+      "loss": 0.4069,
+      "step": 804
+    },
+    {
+      "epoch": 0.6779898933183605,
+      "grad_norm": 0.24769054353237152,
+      "learning_rate": 9.525303541009218e-06,
+      "loss": 0.3805,
+      "step": 805
+    },
+    {
+      "epoch": 0.6788321167883211,
+      "grad_norm": 0.25517210364341736,
+      "learning_rate": 9.523216372150393e-06,
+      "loss": 0.4199,
+      "step": 806
+    },
+    {
+      "epoch": 0.6796743402582819,
+      "grad_norm": 0.22805273532867432,
+      "learning_rate": 9.521124854565425e-06,
+      "loss": 0.4066,
+      "step": 807
+    },
+    {
+      "epoch": 0.6805165637282425,
+      "grad_norm": 0.26646801829338074,
+      "learning_rate": 9.519028990265153e-06,
+      "loss": 0.4145,
+      "step": 808
+    },
+    {
+      "epoch": 0.6813587871982033,
+      "grad_norm": 0.22494232654571533,
+      "learning_rate": 9.516928781264588e-06,
+      "loss": 0.3941,
+      "step": 809
+    },
+    {
+      "epoch": 0.682201010668164,
+      "grad_norm": 0.2086746245622635,
+      "learning_rate": 9.514824229582922e-06,
+      "loss": 0.395,
+      "step": 810
+    },
+    {
+      "epoch": 0.6830432341381246,
+      "grad_norm": 0.22179856896400452,
+      "learning_rate": 9.512715337243517e-06,
+      "loss": 0.3974,
+      "step": 811
+    },
+    {
+      "epoch": 0.6838854576080854,
+      "grad_norm": 0.23663990199565887,
+      "learning_rate": 9.510602106273914e-06,
+      "loss": 0.3894,
+      "step": 812
+    },
+    {
+      "epoch": 0.684727681078046,
+      "grad_norm": 0.23997731506824493,
+      "learning_rate": 9.508484538705823e-06,
+      "loss": 0.391,
+      "step": 813
+    },
+    {
+      "epoch": 0.6855699045480067,
+      "grad_norm": 0.2230270504951477,
+      "learning_rate": 9.506362636575122e-06,
+      "loss": 0.4036,
+      "step": 814
+    },
+    {
+      "epoch": 0.6864121280179675,
+      "grad_norm": 0.23250775039196014,
+      "learning_rate": 9.504236401921856e-06,
+      "loss": 0.4077,
+      "step": 815
+    },
+    {
+      "epoch": 0.6872543514879281,
+      "grad_norm": 0.2370990365743637,
+      "learning_rate": 9.50210583679024e-06,
+      "loss": 0.4115,
+      "step": 816
+    },
+    {
+      "epoch": 0.6880965749578888,
+      "grad_norm": 0.25787898898124695,
+      "learning_rate": 9.499970943228646e-06,
+      "loss": 0.3852,
+      "step": 817
+    },
+    {
+      "epoch": 0.6889387984278496,
+      "grad_norm": 0.24739859998226166,
+      "learning_rate": 9.497831723289615e-06,
+      "loss": 0.3975,
+      "step": 818
+    },
+    {
+      "epoch": 0.6897810218978102,
+      "grad_norm": 0.22881582379341125,
+      "learning_rate": 9.495688179029838e-06,
+      "loss": 0.3766,
+      "step": 819
+    },
+    {
+      "epoch": 0.6906232453677709,
+      "grad_norm": 0.25048691034317017,
+      "learning_rate": 9.493540312510173e-06,
+      "loss": 0.4028,
+      "step": 820
+    },
+    {
+      "epoch": 0.6914654688377316,
+      "grad_norm": 0.22022280097007751,
+      "learning_rate": 9.491388125795623e-06,
+      "loss": 0.4066,
+      "step": 821
+    },
+    {
+      "epoch": 0.6923076923076923,
+      "grad_norm": 0.24745136499404907,
+      "learning_rate": 9.48923162095536e-06,
+      "loss": 0.4013,
+      "step": 822
+    },
+    {
+      "epoch": 0.6931499157776531,
+      "grad_norm": 0.24586893618106842,
+      "learning_rate": 9.487070800062689e-06,
+      "loss": 0.4051,
+      "step": 823
+    },
+    {
+      "epoch": 0.6939921392476137,
+      "grad_norm": 0.22258132696151733,
+      "learning_rate": 9.48490566519508e-06,
+      "loss": 0.4183,
+      "step": 824
+    },
+    {
+      "epoch": 0.6948343627175744,
+      "grad_norm": 0.22149792313575745,
+      "learning_rate": 9.482736218434144e-06,
+      "loss": 0.3884,
+      "step": 825
+    },
+    {
+      "epoch": 0.6956765861875351,
+      "grad_norm": 0.22729527950286865,
+      "learning_rate": 9.480562461865634e-06,
+      "loss": 0.3696,
+      "step": 826
+    },
+    {
+      "epoch": 0.6965188096574958,
+      "grad_norm": 0.22042086720466614,
+      "learning_rate": 9.478384397579452e-06,
+      "loss": 0.3929,
+      "step": 827
+    },
+    {
+      "epoch": 0.6973610331274565,
+      "grad_norm": 0.25043973326683044,
+      "learning_rate": 9.476202027669644e-06,
+      "loss": 0.3997,
+      "step": 828
+    },
+    {
+      "epoch": 0.6982032565974172,
+      "grad_norm": 0.2305058091878891,
+      "learning_rate": 9.474015354234385e-06,
+      "loss": 0.4011,
+      "step": 829
+    },
+    {
+      "epoch": 0.6990454800673779,
+      "grad_norm": 0.25610336661338806,
+      "learning_rate": 9.471824379375998e-06,
+      "loss": 0.4031,
+      "step": 830
+    },
+    {
+      "epoch": 0.6998877035373385,
+      "grad_norm": 0.21562042832374573,
+      "learning_rate": 9.469629105200937e-06,
+      "loss": 0.3763,
+      "step": 831
+    },
+    {
+      "epoch": 0.7007299270072993,
+      "grad_norm": 0.2788180410861969,
+      "learning_rate": 9.46742953381979e-06,
+      "loss": 0.3857,
+      "step": 832
+    },
+    {
+      "epoch": 0.70157215047726,
+      "grad_norm": 0.4841333031654358,
+      "learning_rate": 9.465225667347275e-06,
+      "loss": 0.3926,
+      "step": 833
+    },
+    {
+      "epoch": 0.7024143739472206,
+      "grad_norm": 0.26868411898612976,
+      "learning_rate": 9.463017507902245e-06,
+      "loss": 0.3917,
+      "step": 834
+    },
+    {
+      "epoch": 0.7032565974171814,
+      "grad_norm": 0.2521004378795624,
+      "learning_rate": 9.460805057607671e-06,
+      "loss": 0.4101,
+      "step": 835
+    },
+    {
+      "epoch": 0.704098820887142,
+      "grad_norm": 0.24040234088897705,
+      "learning_rate": 9.458588318590659e-06,
+      "loss": 0.3656,
+      "step": 836
+    },
+    {
+      "epoch": 0.7049410443571027,
+      "grad_norm": 0.2537100613117218,
+      "learning_rate": 9.45636729298243e-06,
+      "loss": 0.3813,
+      "step": 837
+    },
+    {
+      "epoch": 0.7057832678270635,
+      "grad_norm": 0.23415488004684448,
+      "learning_rate": 9.45414198291833e-06,
+      "loss": 0.3937,
+      "step": 838
+    },
+    {
+      "epoch": 0.7066254912970241,
+      "grad_norm": 0.25770246982574463,
+      "learning_rate": 9.451912390537828e-06,
+      "loss": 0.3935,
+      "step": 839
+    },
+    {
+      "epoch": 0.7074677147669849,
+      "grad_norm": 0.25172489881515503,
+      "learning_rate": 9.449678517984503e-06,
+      "loss": 0.3832,
+      "step": 840
+    },
+    {
+      "epoch": 0.7083099382369455,
+      "grad_norm": 0.23774784803390503,
+      "learning_rate": 9.447440367406053e-06,
+      "loss": 0.3819,
+      "step": 841
+    },
+    {
+      "epoch": 0.7091521617069062,
+      "grad_norm": 0.25593727827072144,
+      "learning_rate": 9.445197940954292e-06,
+      "loss": 0.3968,
+      "step": 842
+    },
+    {
+      "epoch": 0.709994385176867,
+      "grad_norm": 0.24409858882427216,
+      "learning_rate": 9.442951240785135e-06,
+      "loss": 0.4202,
+      "step": 843
+    },
+    {
+      "epoch": 0.7108366086468276,
+      "grad_norm": 0.2531510293483734,
+      "learning_rate": 9.440700269058617e-06,
+      "loss": 0.3811,
+      "step": 844
+    },
+    {
+      "epoch": 0.7116788321167883,
+      "grad_norm": 0.27462396025657654,
+      "learning_rate": 9.438445027938873e-06,
+      "loss": 0.4055,
+      "step": 845
+    },
+    {
+      "epoch": 0.712521055586749,
+      "grad_norm": 0.25302422046661377,
+      "learning_rate": 9.436185519594145e-06,
+      "loss": 0.3937,
+      "step": 846
+    },
+    {
+      "epoch": 0.7133632790567097,
+      "grad_norm": 0.24274687469005585,
+      "learning_rate": 9.433921746196777e-06,
+      "loss": 0.4038,
+      "step": 847
+    },
+    {
+      "epoch": 0.7142055025266704,
+      "grad_norm": 0.7182270884513855,
+      "learning_rate": 9.431653709923214e-06,
+      "loss": 0.4178,
+      "step": 848
+    },
+    {
+      "epoch": 0.7150477259966311,
+      "grad_norm": 0.26638856530189514,
+      "learning_rate": 9.429381412954e-06,
+      "loss": 0.3985,
+      "step": 849
+    },
+    {
+      "epoch": 0.7158899494665918,
+      "grad_norm": 0.2720615565776825,
+      "learning_rate": 9.427104857473773e-06,
+      "loss": 0.3844,
+      "step": 850
+    },
+    {
+      "epoch": 0.7167321729365524,
+      "grad_norm": 0.2869759798049927,
+      "learning_rate": 9.424824045671267e-06,
+      "loss": 0.4208,
+      "step": 851
+    },
+    {
+      "epoch": 0.7175743964065132,
+      "grad_norm": 0.242397740483284,
+      "learning_rate": 9.422538979739307e-06,
+      "loss": 0.4087,
+      "step": 852
+    },
+    {
+      "epoch": 0.7184166198764739,
+      "grad_norm": 0.25077950954437256,
+      "learning_rate": 9.420249661874812e-06,
+      "loss": 0.403,
+      "step": 853
+    },
+    {
+      "epoch": 0.7192588433464346,
+      "grad_norm": 0.28628140687942505,
+      "learning_rate": 9.417956094278784e-06,
+      "loss": 0.3731,
+      "step": 854
+    },
+    {
+      "epoch": 0.7201010668163953,
+      "grad_norm": 0.2310740351676941,
+      "learning_rate": 9.415658279156312e-06,
+      "loss": 0.3969,
+      "step": 855
+    },
+    {
+      "epoch": 0.720943290286356,
+      "grad_norm": 0.28179624676704407,
+      "learning_rate": 9.41335621871657e-06,
+      "loss": 0.4003,
+      "step": 856
+    },
+    {
+      "epoch": 0.7217855137563167,
+      "grad_norm": 0.26215457916259766,
+      "learning_rate": 9.41104991517281e-06,
+      "loss": 0.3803,
+      "step": 857
+    },
+    {
+      "epoch": 0.7226277372262774,
+      "grad_norm": 0.24661622941493988,
+      "learning_rate": 9.408739370742372e-06,
+      "loss": 0.3907,
+      "step": 858
+    },
+    {
+      "epoch": 0.723469960696238,
+      "grad_norm": 0.2931444048881531,
+      "learning_rate": 9.406424587646664e-06,
+      "loss": 0.4037,
+      "step": 859
+    },
+    {
+      "epoch": 0.7243121841661988,
+      "grad_norm": 0.21865636110305786,
+      "learning_rate": 9.404105568111173e-06,
+      "loss": 0.392,
+      "step": 860
+    },
+    {
+      "epoch": 0.7251544076361595,
+      "grad_norm": 0.2503955364227295,
+      "learning_rate": 9.401782314365458e-06,
+      "loss": 0.3787,
+      "step": 861
+    },
+    {
+      "epoch": 0.7259966311061201,
+      "grad_norm": 0.2230653613805771,
+      "learning_rate": 9.39945482864315e-06,
+      "loss": 0.4023,
+      "step": 862
+    },
+    {
+      "epoch": 0.7268388545760809,
+      "grad_norm": 0.2315695583820343,
+      "learning_rate": 9.39712311318195e-06,
+      "loss": 0.3925,
+      "step": 863
+    },
+    {
+      "epoch": 0.7276810780460415,
+      "grad_norm": 0.2381676435470581,
+      "learning_rate": 9.39478717022362e-06,
+      "loss": 0.3936,
+      "step": 864
+    },
+    {
+      "epoch": 0.7285233015160022,
+      "grad_norm": 0.24494384229183197,
+      "learning_rate": 9.392447002013996e-06,
+      "loss": 0.3849,
+      "step": 865
+    },
+    {
+      "epoch": 0.729365524985963,
+      "grad_norm": 0.2394935041666031,
+      "learning_rate": 9.390102610802965e-06,
+      "loss": 0.3882,
+      "step": 866
+    },
+    {
+      "epoch": 0.7302077484559236,
+      "grad_norm": 0.24465805292129517,
+      "learning_rate": 9.387753998844482e-06,
+      "loss": 0.3765,
+      "step": 867
+    },
+    {
+      "epoch": 0.7310499719258844,
+      "grad_norm": 0.23144975304603577,
+      "learning_rate": 9.385401168396558e-06,
+      "loss": 0.3837,
+      "step": 868
+    },
+    {
+      "epoch": 0.731892195395845,
+      "grad_norm": 0.2408526986837387,
+      "learning_rate": 9.383044121721257e-06,
+      "loss": 0.3716,
+      "step": 869
+    },
+    {
+      "epoch": 0.7327344188658057,
+      "grad_norm": 0.23346202075481415,
+      "learning_rate": 9.380682861084703e-06,
+      "loss": 0.3978,
+      "step": 870
+    },
+    {
+      "epoch": 0.7335766423357665,
+      "grad_norm": 0.2438187599182129,
+      "learning_rate": 9.378317388757062e-06,
+      "loss": 0.3938,
+      "step": 871
+    },
+    {
+      "epoch": 0.7344188658057271,
+      "grad_norm": 0.23521101474761963,
+      "learning_rate": 9.375947707012558e-06,
+      "loss": 0.3637,
+      "step": 872
+    },
+    {
+      "epoch": 0.7352610892756878,
+      "grad_norm": 0.2562772035598755,
+      "learning_rate": 9.37357381812946e-06,
+      "loss": 0.3888,
+      "step": 873
+    },
+    {
+      "epoch": 0.7361033127456486,
+      "grad_norm": 0.24033290147781372,
+      "learning_rate": 9.371195724390075e-06,
+      "loss": 0.3761,
+      "step": 874
+    },
+    {
+      "epoch": 0.7369455362156092,
+      "grad_norm": 0.25617146492004395,
+      "learning_rate": 9.368813428080763e-06,
+      "loss": 0.3926,
+      "step": 875
+    },
+    {
+      "epoch": 0.7377877596855699,
+      "grad_norm": 0.24614277482032776,
+      "learning_rate": 9.366426931491917e-06,
+      "loss": 0.3763,
+      "step": 876
+    },
+    {
+      "epoch": 0.7386299831555306,
+      "grad_norm": 0.24150444567203522,
+      "learning_rate": 9.364036236917972e-06,
+      "loss": 0.4064,
+      "step": 877
+    },
+    {
+      "epoch": 0.7394722066254913,
+      "grad_norm": 0.26152312755584717,
+      "learning_rate": 9.361641346657396e-06,
+      "loss": 0.3918,
+      "step": 878
+    },
+    {
+      "epoch": 0.740314430095452,
+      "grad_norm": 0.20616672933101654,
+      "learning_rate": 9.359242263012693e-06,
+      "loss": 0.4021,
+      "step": 879
+    },
+    {
+      "epoch": 0.7411566535654127,
+      "grad_norm": 0.23581033945083618,
+      "learning_rate": 9.356838988290401e-06,
+      "loss": 0.3902,
+      "step": 880
+    },
+    {
+      "epoch": 0.7419988770353734,
+      "grad_norm": 0.24657794833183289,
+      "learning_rate": 9.354431524801082e-06,
+      "loss": 0.4069,
+      "step": 881
+    },
+    {
+      "epoch": 0.742841100505334,
+      "grad_norm": 0.21929241716861725,
+      "learning_rate": 9.352019874859326e-06,
+      "loss": 0.3884,
+      "step": 882
+    },
+    {
+      "epoch": 0.7436833239752948,
+      "grad_norm": 0.24844641983509064,
+      "learning_rate": 9.349604040783754e-06,
+      "loss": 0.3814,
+      "step": 883
+    },
+    {
+      "epoch": 0.7445255474452555,
+      "grad_norm": 0.24313275516033173,
+      "learning_rate": 9.347184024897003e-06,
+      "loss": 0.4,
+      "step": 884
+    },
+    {
+      "epoch": 0.7453677709152162,
+      "grad_norm": 0.24789856374263763,
+      "learning_rate": 9.344759829525734e-06,
+      "loss": 0.4097,
+      "step": 885
+    },
+    {
+      "epoch": 0.7462099943851769,
+      "grad_norm": 0.24578550457954407,
+      "learning_rate": 9.342331457000621e-06,
+      "loss": 0.4079,
+      "step": 886
+    },
+    {
+      "epoch": 0.7470522178551375,
+      "grad_norm": 0.23694393038749695,
+      "learning_rate": 9.339898909656364e-06,
+      "loss": 0.4013,
+      "step": 887
+    },
+    {
+      "epoch": 0.7478944413250983,
+      "grad_norm": 0.2592124342918396,
+      "learning_rate": 9.33746218983167e-06,
+      "loss": 0.3991,
+      "step": 888
+    },
+    {
+      "epoch": 0.748736664795059,
+      "grad_norm": 0.23632554709911346,
+      "learning_rate": 9.335021299869256e-06,
+      "loss": 0.4267,
+      "step": 889
+    },
+    {
+      "epoch": 0.7495788882650196,
+      "grad_norm": 0.24402032792568207,
+      "learning_rate": 9.332576242115852e-06,
+      "loss": 0.3786,
+      "step": 890
+    },
+    {
+      "epoch": 0.7504211117349804,
+      "grad_norm": 0.22527766227722168,
+      "learning_rate": 9.330127018922195e-06,
+      "loss": 0.3789,
+      "step": 891
+    },
+    {
+      "epoch": 0.751263335204941,
+      "grad_norm": 0.2471362203359604,
+      "learning_rate": 9.327673632643021e-06,
+      "loss": 0.4083,
+      "step": 892
+    },
+    {
+      "epoch": 0.7521055586749017,
+      "grad_norm": 0.23946036398410797,
+      "learning_rate": 9.32521608563708e-06,
+      "loss": 0.3869,
+      "step": 893
+    },
+    {
+      "epoch": 0.7529477821448625,
+      "grad_norm": 0.275928795337677,
+      "learning_rate": 9.32275438026711e-06,
+      "loss": 0.3786,
+      "step": 894
+    },
+    {
+      "epoch": 0.7537900056148231,
+      "grad_norm": 0.21995076537132263,
+      "learning_rate": 9.320288518899853e-06,
+      "loss": 0.3825,
+      "step": 895
+    },
+    {
+      "epoch": 0.7546322290847838,
+      "grad_norm": 0.24087098240852356,
+      "learning_rate": 9.317818503906046e-06,
+      "loss": 0.3836,
+      "step": 896
+    },
+    {
+      "epoch": 0.7554744525547445,
+      "grad_norm": 0.23902052640914917,
+      "learning_rate": 9.315344337660422e-06,
+      "loss": 0.3879,
+      "step": 897
+    },
+    {
+      "epoch": 0.7563166760247052,
+      "grad_norm": 0.23242321610450745,
+      "learning_rate": 9.312866022541697e-06,
+      "loss": 0.3837,
+      "step": 898
+    },
+    {
+      "epoch": 0.757158899494666,
+      "grad_norm": 0.23153363168239594,
+      "learning_rate": 9.310383560932587e-06,
+      "loss": 0.3842,
+      "step": 899
+    },
+    {
+      "epoch": 0.7580011229646266,
+      "grad_norm": 0.24144352972507477,
+      "learning_rate": 9.307896955219787e-06,
+      "loss": 0.3744,
+      "step": 900
+    },
+    {
+      "epoch": 0.7588433464345873,
+      "grad_norm": 0.3434450626373291,
+      "learning_rate": 9.305406207793974e-06,
+      "loss": 0.4151,
+      "step": 901
+    },
+    {
+      "epoch": 0.759685569904548,
+      "grad_norm": 0.23582684993743896,
+      "learning_rate": 9.302911321049818e-06,
+      "loss": 0.3812,
+      "step": 902
+    },
+    {
+      "epoch": 0.7605277933745087,
+      "grad_norm": 0.24953168630599976,
+      "learning_rate": 9.300412297385954e-06,
+      "loss": 0.39,
+      "step": 903
+    },
+    {
+      "epoch": 0.7613700168444694,
+      "grad_norm": 0.30013614892959595,
+      "learning_rate": 9.297909139205005e-06,
+      "loss": 0.4118,
+      "step": 904
+    },
+    {
+      "epoch": 0.7622122403144301,
+      "grad_norm": 0.24813398718833923,
+      "learning_rate": 9.295401848913569e-06,
+      "loss": 0.3836,
+      "step": 905
+    },
+    {
+      "epoch": 0.7630544637843908,
+      "grad_norm": 0.226630300283432,
+      "learning_rate": 9.29289042892221e-06,
+      "loss": 0.3875,
+      "step": 906
+    },
+    {
+      "epoch": 0.7638966872543514,
+      "grad_norm": 0.23602192103862762,
+      "learning_rate": 9.290374881645465e-06,
+      "loss": 0.4,
+      "step": 907
+    },
+    {
+      "epoch": 0.7647389107243122,
+      "grad_norm": 0.24765224754810333,
+      "learning_rate": 9.287855209501844e-06,
+      "loss": 0.4083,
+      "step": 908
+    },
+    {
+      "epoch": 0.7655811341942729,
+      "grad_norm": 0.27086493372917175,
+      "learning_rate": 9.285331414913816e-06,
+      "loss": 0.4144,
+      "step": 909
+    },
+    {
+      "epoch": 0.7664233576642335,
+      "grad_norm": 0.22973863780498505,
+      "learning_rate": 9.282803500307818e-06,
+      "loss": 0.3661,
+      "step": 910
+    },
+    {
+      "epoch": 0.7672655811341943,
+      "grad_norm": 0.30480989813804626,
+      "learning_rate": 9.280271468114243e-06,
+      "loss": 0.3913,
+      "step": 911
+    },
+    {
+      "epoch": 0.768107804604155,
+      "grad_norm": 0.26645076274871826,
+      "learning_rate": 9.277735320767449e-06,
+      "loss": 0.4083,
+      "step": 912
+    },
+    {
+      "epoch": 0.7689500280741156,
+      "grad_norm": 0.27367421984672546,
+      "learning_rate": 9.275195060705749e-06,
+      "loss": 0.3795,
+      "step": 913
+    },
+    {
+      "epoch": 0.7697922515440764,
+      "grad_norm": 0.3100326359272003,
+      "learning_rate": 9.272650690371403e-06,
+      "loss": 0.3931,
+      "step": 914
+    },
+    {
+      "epoch": 0.770634475014037,
+      "grad_norm": 0.2507917582988739,
+      "learning_rate": 9.270102212210632e-06,
+      "loss": 0.3892,
+      "step": 915
+    },
+    {
+      "epoch": 0.7714766984839978,
+      "grad_norm": 0.245370551943779,
+      "learning_rate": 9.267549628673603e-06,
+      "loss": 0.3953,
+      "step": 916
+    },
+    {
+      "epoch": 0.7723189219539585,
+      "grad_norm": 0.2668488621711731,
+      "learning_rate": 9.264992942214427e-06,
+      "loss": 0.4083,
+      "step": 917
+    },
+    {
+      "epoch": 0.7731611454239191,
+      "grad_norm": 0.257581502199173,
+      "learning_rate": 9.262432155291167e-06,
+      "loss": 0.3755,
+      "step": 918
+    },
+    {
+      "epoch": 0.7740033688938799,
+      "grad_norm": 0.24320945143699646,
+      "learning_rate": 9.25986727036582e-06,
+      "loss": 0.3707,
+      "step": 919
+    },
+    {
+      "epoch": 0.7748455923638405,
+      "grad_norm": 0.2300320714712143,
+      "learning_rate": 9.257298289904324e-06,
+      "loss": 0.376,
+      "step": 920
+    },
+    {
+      "epoch": 0.7756878158338012,
+      "grad_norm": 0.2596640884876251,
+      "learning_rate": 9.254725216376562e-06,
+      "loss": 0.3885,
+      "step": 921
+    },
+    {
+      "epoch": 0.776530039303762,
+      "grad_norm": 0.2934633791446686,
+      "learning_rate": 9.252148052256343e-06,
+      "loss": 0.3939,
+      "step": 922
+    },
+    {
+      "epoch": 0.7773722627737226,
+      "grad_norm": 0.23935237526893616,
+      "learning_rate": 9.249566800021417e-06,
+      "loss": 0.4103,
+      "step": 923
+    },
+    {
+      "epoch": 0.7782144862436833,
+      "grad_norm": 0.25070446729660034,
+      "learning_rate": 9.246981462153456e-06,
+      "loss": 0.4276,
+      "step": 924
+    },
+    {
+      "epoch": 0.779056709713644,
+      "grad_norm": 0.23785822093486786,
+      "learning_rate": 9.244392041138068e-06,
+      "loss": 0.3645,
+      "step": 925
+    },
+    {
+      "epoch": 0.7798989331836047,
+      "grad_norm": 0.32643628120422363,
+      "learning_rate": 9.24179853946478e-06,
+      "loss": 0.4096,
+      "step": 926
+    },
+    {
+      "epoch": 0.7807411566535654,
+      "grad_norm": 0.23632368445396423,
+      "learning_rate": 9.239200959627048e-06,
+      "loss": 0.3805,
+      "step": 927
+    },
+    {
+      "epoch": 0.7815833801235261,
+      "grad_norm": 0.23155541718006134,
+      "learning_rate": 9.236599304122246e-06,
+      "loss": 0.3825,
+      "step": 928
+    },
+    {
+      "epoch": 0.7824256035934868,
+      "grad_norm": 0.25782865285873413,
+      "learning_rate": 9.233993575451663e-06,
+      "loss": 0.3925,
+      "step": 929
+    },
+    {
+      "epoch": 0.7832678270634476,
+      "grad_norm": 0.25167566537857056,
+      "learning_rate": 9.231383776120512e-06,
+      "loss": 0.3869,
+      "step": 930
+    },
+    {
+      "epoch": 0.7841100505334082,
+      "grad_norm": 0.3077695369720459,
+      "learning_rate": 9.228769908637912e-06,
+      "loss": 0.4098,
+      "step": 931
+    },
+    {
+      "epoch": 0.7849522740033689,
+      "grad_norm": 0.231857568025589,
+      "learning_rate": 9.226151975516897e-06,
+      "loss": 0.4046,
+      "step": 932
+    },
+    {
+      "epoch": 0.7857944974733296,
+      "grad_norm": 0.25358080863952637,
+      "learning_rate": 9.223529979274411e-06,
+      "loss": 0.3868,
+      "step": 933
+    },
+    {
+      "epoch": 0.7866367209432903,
+      "grad_norm": 0.23811326920986176,
+      "learning_rate": 9.220903922431302e-06,
+      "loss": 0.3926,
+      "step": 934
+    },
+    {
+      "epoch": 0.787478944413251,
+      "grad_norm": 0.2356197088956833,
+      "learning_rate": 9.218273807512318e-06,
+      "loss": 0.4042,
+      "step": 935
+    },
+    {
+      "epoch": 0.7883211678832117,
+      "grad_norm": 0.25609177350997925,
+      "learning_rate": 9.215639637046121e-06,
+      "loss": 0.3894,
+      "step": 936
+    },
+    {
+      "epoch": 0.7891633913531724,
+      "grad_norm": 0.2308328002691269,
+      "learning_rate": 9.213001413565259e-06,
+      "loss": 0.4005,
+      "step": 937
+    },
+    {
+      "epoch": 0.790005614823133,
+      "grad_norm": 0.2398657649755478,
+      "learning_rate": 9.210359139606183e-06,
+      "loss": 0.3909,
+      "step": 938
+    },
+    {
+      "epoch": 0.7908478382930938,
+      "grad_norm": 0.2556980550289154,
+      "learning_rate": 9.207712817709237e-06,
+      "loss": 0.3864,
+      "step": 939
+    },
+    {
+      "epoch": 0.7916900617630545,
+      "grad_norm": 0.2326350063085556,
+      "learning_rate": 9.205062450418655e-06,
+      "loss": 0.3678,
+      "step": 940
+    },
+    {
+      "epoch": 0.7925322852330151,
+      "grad_norm": 0.2267482876777649,
+      "learning_rate": 9.202408040282567e-06,
+      "loss": 0.3648,
+      "step": 941
+    },
+    {
+      "epoch": 0.7933745087029759,
+      "grad_norm": 0.24878454208374023,
+      "learning_rate": 9.19974958985298e-06,
+      "loss": 0.3939,
+      "step": 942
+    },
+    {
+      "epoch": 0.7942167321729365,
+      "grad_norm": 0.24260170757770538,
+      "learning_rate": 9.197087101685794e-06,
+      "loss": 0.3923,
+      "step": 943
+    },
+    {
+      "epoch": 0.7950589556428973,
+      "grad_norm": 0.24691571295261383,
+      "learning_rate": 9.194420578340785e-06,
+      "loss": 0.3934,
+      "step": 944
+    },
+    {
+      "epoch": 0.795901179112858,
+      "grad_norm": 0.24263276159763336,
+      "learning_rate": 9.191750022381613e-06,
+      "loss": 0.3892,
+      "step": 945
+    },
+    {
+      "epoch": 0.7967434025828186,
+      "grad_norm": 0.26982370018959045,
+      "learning_rate": 9.189075436375813e-06,
+      "loss": 0.4148,
+      "step": 946
+    },
+    {
+      "epoch": 0.7975856260527794,
+      "grad_norm": 0.2315046787261963,
+      "learning_rate": 9.186396822894792e-06,
+      "loss": 0.3889,
+      "step": 947
+    },
+    {
+      "epoch": 0.79842784952274,
+      "grad_norm": 0.25008097290992737,
+      "learning_rate": 9.183714184513832e-06,
+      "loss": 0.3772,
+      "step": 948
+    },
+    {
+      "epoch": 0.7992700729927007,
+      "grad_norm": 0.24061961472034454,
+      "learning_rate": 9.181027523812088e-06,
+      "loss": 0.3838,
+      "step": 949
+    },
+    {
+      "epoch": 0.8001122964626615,
+      "grad_norm": 0.25396090745925903,
+      "learning_rate": 9.178336843372576e-06,
+      "loss": 0.3902,
+      "step": 950
+    },
+    {
+      "epoch": 0.8009545199326221,
+      "grad_norm": 0.24839384853839874,
+      "learning_rate": 9.175642145782179e-06,
+      "loss": 0.3847,
+      "step": 951
+    },
+    {
+      "epoch": 0.8017967434025828,
+      "grad_norm": 0.273550420999527,
+      "learning_rate": 9.172943433631642e-06,
+      "loss": 0.3958,
+      "step": 952
+    },
+    {
+      "epoch": 0.8026389668725435,
+      "grad_norm": 0.2734324634075165,
+      "learning_rate": 9.170240709515573e-06,
+      "loss": 0.3939,
+      "step": 953
+    },
+    {
+      "epoch": 0.8034811903425042,
+      "grad_norm": 0.2695125341415405,
+      "learning_rate": 9.16753397603243e-06,
+      "loss": 0.387,
+      "step": 954
+    },
+    {
+      "epoch": 0.8043234138124649,
+      "grad_norm": 0.2518448233604431,
+      "learning_rate": 9.164823235784535e-06,
+      "loss": 0.391,
+      "step": 955
+    },
+    {
+      "epoch": 0.8051656372824256,
+      "grad_norm": 0.2905693054199219,
+      "learning_rate": 9.162108491378051e-06,
+      "loss": 0.4143,
+      "step": 956
+    },
+    {
+      "epoch": 0.8060078607523863,
+      "grad_norm": 0.2615605890750885,
+      "learning_rate": 9.159389745423003e-06,
+      "loss": 0.3993,
+      "step": 957
+    },
+    {
+      "epoch": 0.8068500842223469,
+      "grad_norm": 0.2930067777633667,
+      "learning_rate": 9.156667000533251e-06,
+      "loss": 0.3892,
+      "step": 958
+    },
+    {
+      "epoch": 0.8076923076923077,
+      "grad_norm": 0.2685520648956299,
+      "learning_rate": 9.153940259326511e-06,
+      "loss": 0.4065,
+      "step": 959
+    },
+    {
+      "epoch": 0.8085345311622684,
+      "grad_norm": 0.2457975298166275,
+      "learning_rate": 9.151209524424333e-06,
+      "loss": 0.3706,
+      "step": 960
+    },
+    {
+      "epoch": 0.8093767546322291,
+      "grad_norm": 0.27907201647758484,
+      "learning_rate": 9.14847479845211e-06,
+      "loss": 0.4101,
+      "step": 961
+    },
+    {
+      "epoch": 0.8102189781021898,
+      "grad_norm": 0.2515687942504883,
+      "learning_rate": 9.145736084039073e-06,
+      "loss": 0.3965,
+      "step": 962
+    },
+    {
+      "epoch": 0.8110612015721504,
+      "grad_norm": 0.24229024350643158,
+      "learning_rate": 9.142993383818284e-06,
+      "loss": 0.3875,
+      "step": 963
+    },
+    {
+      "epoch": 0.8119034250421112,
+      "grad_norm": 0.27047234773635864,
+      "learning_rate": 9.14024670042664e-06,
+      "loss": 0.4292,
+      "step": 964
+    },
+    {
+      "epoch": 0.8127456485120719,
+      "grad_norm": 0.26469686627388,
+      "learning_rate": 9.137496036504868e-06,
+      "loss": 0.372,
+      "step": 965
+    },
+    {
+      "epoch": 0.8135878719820325,
+      "grad_norm": 0.24354217946529388,
+      "learning_rate": 9.134741394697517e-06,
+      "loss": 0.3789,
+      "step": 966
+    },
+    {
+      "epoch": 0.8144300954519933,
+      "grad_norm": 0.23837910592556,
+      "learning_rate": 9.131982777652967e-06,
+      "loss": 0.3942,
+      "step": 967
+    },
+    {
+      "epoch": 0.815272318921954,
+      "grad_norm": 0.22872686386108398,
+      "learning_rate": 9.129220188023419e-06,
+      "loss": 0.3917,
+      "step": 968
+    },
+    {
+      "epoch": 0.8161145423919146,
+      "grad_norm": 0.22326874732971191,
+      "learning_rate": 9.126453628464889e-06,
+      "loss": 0.3837,
+      "step": 969
+    },
+    {
+      "epoch": 0.8169567658618754,
+      "grad_norm": 0.24706344306468964,
+      "learning_rate": 9.12368310163721e-06,
+      "loss": 0.4004,
+      "step": 970
+    },
+    {
+      "epoch": 0.817798989331836,
+      "grad_norm": 0.24931980669498444,
+      "learning_rate": 9.120908610204036e-06,
+      "loss": 0.4093,
+      "step": 971
+    },
+    {
+      "epoch": 0.8186412128017967,
+      "grad_norm": 0.23322880268096924,
+      "learning_rate": 9.118130156832823e-06,
+      "loss": 0.3866,
+      "step": 972
+    },
+    {
+      "epoch": 0.8194834362717575,
+      "grad_norm": 0.22499699890613556,
+      "learning_rate": 9.115347744194844e-06,
+      "loss": 0.3999,
+      "step": 973
+    },
+    {
+      "epoch": 0.8203256597417181,
+      "grad_norm": 0.22839175164699554,
+      "learning_rate": 9.112561374965177e-06,
+      "loss": 0.4033,
+      "step": 974
+    },
+    {
+      "epoch": 0.8211678832116789,
+      "grad_norm": 0.22353333234786987,
+      "learning_rate": 9.109771051822702e-06,
+      "loss": 0.3648,
+      "step": 975
+    },
+    {
+      "epoch": 0.8220101066816395,
+      "grad_norm": 0.3328852653503418,
+      "learning_rate": 9.106976777450099e-06,
+      "loss": 0.4012,
+      "step": 976
+    },
+    {
+      "epoch": 0.8228523301516002,
+      "grad_norm": 0.26589012145996094,
+      "learning_rate": 9.10417855453385e-06,
+      "loss": 0.3886,
+      "step": 977
+    },
+    {
+      "epoch": 0.823694553621561,
+      "grad_norm": 0.2546745538711548,
+      "learning_rate": 9.10137638576423e-06,
+      "loss": 0.3997,
+      "step": 978
+    },
+    {
+      "epoch": 0.8245367770915216,
+      "grad_norm": 0.23469386994838715,
+      "learning_rate": 9.098570273835314e-06,
+      "loss": 0.3859,
+      "step": 979
+    },
+    {
+      "epoch": 0.8253790005614823,
+      "grad_norm": 0.23561683297157288,
+      "learning_rate": 9.09576022144496e-06,
+      "loss": 0.3831,
+      "step": 980
+    },
+    {
+      "epoch": 0.826221224031443,
+      "grad_norm": 0.28168466687202454,
+      "learning_rate": 9.09294623129482e-06,
+      "loss": 0.3641,
+      "step": 981
+    },
+    {
+      "epoch": 0.8270634475014037,
+      "grad_norm": 0.24064616858959198,
+      "learning_rate": 9.090128306090329e-06,
+      "loss": 0.4049,
+      "step": 982
+    },
+    {
+      "epoch": 0.8279056709713644,
+      "grad_norm": 0.2215161919593811,
+      "learning_rate": 9.087306448540707e-06,
+      "loss": 0.3992,
+      "step": 983
+    },
+    {
+      "epoch": 0.8287478944413251,
+      "grad_norm": 0.2595074772834778,
+      "learning_rate": 9.084480661358954e-06,
+      "loss": 0.3823,
+      "step": 984
+    },
+    {
+      "epoch": 0.8295901179112858,
+      "grad_norm": 0.23889648914337158,
+      "learning_rate": 9.081650947261847e-06,
+      "loss": 0.3656,
+      "step": 985
+    },
+    {
+      "epoch": 0.8304323413812464,
+      "grad_norm": 0.24093107879161835,
+      "learning_rate": 9.07881730896994e-06,
+      "loss": 0.3919,
+      "step": 986
+    },
+    {
+      "epoch": 0.8312745648512072,
+      "grad_norm": 0.2698620557785034,
+      "learning_rate": 9.07597974920756e-06,
+      "loss": 0.3986,
+      "step": 987
+    },
+    {
+      "epoch": 0.8321167883211679,
+      "grad_norm": 0.2287716567516327,
+      "learning_rate": 9.073138270702804e-06,
+      "loss": 0.3815,
+      "step": 988
+    },
+    {
+      "epoch": 0.8329590117911286,
+      "grad_norm": 0.237266406416893,
+      "learning_rate": 9.070292876187532e-06,
+      "loss": 0.3726,
+      "step": 989
+    },
+    {
+      "epoch": 0.8338012352610893,
+      "grad_norm": 0.2581973075866699,
+      "learning_rate": 9.067443568397378e-06,
+      "loss": 0.4076,
+      "step": 990
+    },
+    {
+      "epoch": 0.83464345873105,
+      "grad_norm": 0.24701803922653198,
+      "learning_rate": 9.06459035007173e-06,
+      "loss": 0.4063,
+      "step": 991
+    },
+    {
+      "epoch": 0.8354856822010107,
+      "grad_norm": 0.24334250390529633,
+      "learning_rate": 9.061733223953738e-06,
+      "loss": 0.3882,
+      "step": 992
+    },
+    {
+      "epoch": 0.8363279056709714,
+      "grad_norm": 0.2147740125656128,
+      "learning_rate": 9.058872192790314e-06,
+      "loss": 0.3798,
+      "step": 993
+    },
+    {
+      "epoch": 0.837170129140932,
+      "grad_norm": 0.24799306690692902,
+      "learning_rate": 9.056007259332115e-06,
+      "loss": 0.415,
+      "step": 994
+    },
+    {
+      "epoch": 0.8380123526108928,
+      "grad_norm": 0.23995551466941833,
+      "learning_rate": 9.053138426333562e-06,
+      "loss": 0.4021,
+      "step": 995
+    },
+    {
+      "epoch": 0.8388545760808535,
+      "grad_norm": 0.24283404648303986,
+      "learning_rate": 9.05026569655281e-06,
+      "loss": 0.3929,
+      "step": 996
+    },
+    {
+      "epoch": 0.8396967995508141,
+      "grad_norm": 0.2712905704975128,
+      "learning_rate": 9.047389072751777e-06,
+      "loss": 0.3717,
+      "step": 997
+    },
+    {
+      "epoch": 0.8405390230207749,
+      "grad_norm": 0.27794942259788513,
+      "learning_rate": 9.044508557696111e-06,
+      "loss": 0.3774,
+      "step": 998
+    },
+    {
+      "epoch": 0.8413812464907355,
+      "grad_norm": 0.26650184392929077,
+      "learning_rate": 9.041624154155208e-06,
+      "loss": 0.3967,
+      "step": 999
+    },
+    {
+      "epoch": 0.8422234699606962,
+      "grad_norm": 0.30274587869644165,
+      "learning_rate": 9.038735864902201e-06,
+      "loss": 0.3873,
+      "step": 1000
+    },
+    {
+      "epoch": 0.843065693430657,
+      "grad_norm": 0.2814517617225647,
+      "learning_rate": 9.035843692713961e-06,
+      "loss": 0.391,
+      "step": 1001
+    },
+    {
+      "epoch": 0.8439079169006176,
+      "grad_norm": 0.23480753600597382,
+      "learning_rate": 9.032947640371086e-06,
+      "loss": 0.3931,
+      "step": 1002
+    },
+    {
+      "epoch": 0.8447501403705783,
+      "grad_norm": 0.2912794351577759,
+      "learning_rate": 9.030047710657912e-06,
+      "loss": 0.3905,
+      "step": 1003
+    },
+    {
+      "epoch": 0.845592363840539,
+      "grad_norm": 0.26708126068115234,
+      "learning_rate": 9.027143906362499e-06,
+      "loss": 0.3829,
+      "step": 1004
+    },
+    {
+      "epoch": 0.8464345873104997,
+      "grad_norm": 0.2397577464580536,
+      "learning_rate": 9.02423623027663e-06,
+      "loss": 0.3829,
+      "step": 1005
+    },
+    {
+      "epoch": 0.8472768107804605,
+      "grad_norm": 0.2438420057296753,
+      "learning_rate": 9.021324685195814e-06,
+      "loss": 0.3888,
+      "step": 1006
+    },
+    {
+      "epoch": 0.8481190342504211,
+      "grad_norm": 0.24794094264507294,
+      "learning_rate": 9.018409273919279e-06,
+      "loss": 0.378,
+      "step": 1007
+    },
+    {
+      "epoch": 0.8489612577203818,
+      "grad_norm": 0.2339514195919037,
+      "learning_rate": 9.01548999924997e-06,
+      "loss": 0.4014,
+      "step": 1008
+    },
+    {
+      "epoch": 0.8498034811903425,
+      "grad_norm": 0.2431575357913971,
+      "learning_rate": 9.012566863994548e-06,
+      "loss": 0.3975,
+      "step": 1009
+    },
+    {
+      "epoch": 0.8506457046603032,
+      "grad_norm": 0.23573164641857147,
+      "learning_rate": 9.00963987096338e-06,
+      "loss": 0.3963,
+      "step": 1010
+    },
+    {
+      "epoch": 0.8514879281302639,
+      "grad_norm": 0.245757058262825,
+      "learning_rate": 9.006709022970547e-06,
+      "loss": 0.4054,
+      "step": 1011
+    },
+    {
+      "epoch": 0.8523301516002246,
+      "grad_norm": 0.24755309522151947,
+      "learning_rate": 9.003774322833835e-06,
+      "loss": 0.3954,
+      "step": 1012
+    },
+    {
+      "epoch": 0.8531723750701853,
+      "grad_norm": 0.25644731521606445,
+      "learning_rate": 9.000835773374733e-06,
+      "loss": 0.3977,
+      "step": 1013
+    },
+    {
+      "epoch": 0.8540145985401459,
+      "grad_norm": 0.253381609916687,
+      "learning_rate": 8.997893377418432e-06,
+      "loss": 0.4008,
+      "step": 1014
+    },
+    {
+      "epoch": 0.8548568220101067,
+      "grad_norm": 0.2430805116891861,
+      "learning_rate": 8.99494713779382e-06,
+      "loss": 0.3724,
+      "step": 1015
+    },
+    {
+      "epoch": 0.8556990454800674,
+      "grad_norm": 0.24902412295341492,
+      "learning_rate": 8.991997057333481e-06,
+      "loss": 0.3809,
+      "step": 1016
+    },
+    {
+      "epoch": 0.856541268950028,
+      "grad_norm": 0.2504844665527344,
+      "learning_rate": 8.98904313887369e-06,
+      "loss": 0.397,
+      "step": 1017
+    },
+    {
+      "epoch": 0.8573834924199888,
+      "grad_norm": 0.28356367349624634,
+      "learning_rate": 8.986085385254417e-06,
+      "loss": 0.4272,
+      "step": 1018
+    },
+    {
+      "epoch": 0.8582257158899494,
+      "grad_norm": 0.2447170913219452,
+      "learning_rate": 8.983123799319312e-06,
+      "loss": 0.3974,
+      "step": 1019
+    },
+    {
+      "epoch": 0.8590679393599102,
+      "grad_norm": 0.21878619492053986,
+      "learning_rate": 8.980158383915714e-06,
+      "loss": 0.3867,
+      "step": 1020
+    },
+    {
+      "epoch": 0.8599101628298709,
+      "grad_norm": 0.254640132188797,
+      "learning_rate": 8.977189141894645e-06,
+      "loss": 0.4221,
+      "step": 1021
+    },
+    {
+      "epoch": 0.8607523862998315,
+      "grad_norm": 0.2748672664165497,
+      "learning_rate": 8.9742160761108e-06,
+      "loss": 0.3782,
+      "step": 1022
+    },
+    {
+      "epoch": 0.8615946097697923,
+      "grad_norm": 0.24861536920070648,
+      "learning_rate": 8.971239189422555e-06,
+      "loss": 0.374,
+      "step": 1023
+    },
+    {
+      "epoch": 0.862436833239753,
+      "grad_norm": 0.28852665424346924,
+      "learning_rate": 8.968258484691961e-06,
+      "loss": 0.4046,
+      "step": 1024
+    },
+    {
+      "epoch": 0.8632790567097136,
+      "grad_norm": 0.24399378895759583,
+      "learning_rate": 8.965273964784735e-06,
+      "loss": 0.3799,
+      "step": 1025
+    },
+    {
+      "epoch": 0.8641212801796744,
+      "grad_norm": 0.24471896886825562,
+      "learning_rate": 8.962285632570266e-06,
+      "loss": 0.3918,
+      "step": 1026
+    },
+    {
+      "epoch": 0.864963503649635,
+      "grad_norm": 0.2678011953830719,
+      "learning_rate": 8.959293490921606e-06,
+      "loss": 0.3867,
+      "step": 1027
+    },
+    {
+      "epoch": 0.8658057271195957,
+      "grad_norm": 0.2465362846851349,
+      "learning_rate": 8.956297542715469e-06,
+      "loss": 0.3904,
+      "step": 1028
+    },
+    {
+      "epoch": 0.8666479505895565,
+      "grad_norm": 0.268699586391449,
+      "learning_rate": 8.953297790832231e-06,
+      "loss": 0.3987,
+      "step": 1029
+    },
+    {
+      "epoch": 0.8674901740595171,
+      "grad_norm": 0.23215501010417938,
+      "learning_rate": 8.950294238155924e-06,
+      "loss": 0.3802,
+      "step": 1030
+    },
+    {
+      "epoch": 0.8683323975294778,
+      "grad_norm": 0.24207451939582825,
+      "learning_rate": 8.947286887574234e-06,
+      "loss": 0.402,
+      "step": 1031
+    },
+    {
+      "epoch": 0.8691746209994385,
+      "grad_norm": 0.2715492248535156,
+      "learning_rate": 8.944275741978495e-06,
+      "loss": 0.3945,
+      "step": 1032
+    },
+    {
+      "epoch": 0.8700168444693992,
+      "grad_norm": 0.2414940595626831,
+      "learning_rate": 8.941260804263697e-06,
+      "loss": 0.4273,
+      "step": 1033
+    },
+    {
+      "epoch": 0.87085906793936,
+      "grad_norm": 0.2749345600605011,
+      "learning_rate": 8.938242077328469e-06,
+      "loss": 0.4063,
+      "step": 1034
+    },
+    {
+      "epoch": 0.8717012914093206,
+      "grad_norm": 0.24765220284461975,
+      "learning_rate": 8.935219564075087e-06,
+      "loss": 0.3833,
+      "step": 1035
+    },
+    {
+      "epoch": 0.8725435148792813,
+      "grad_norm": 0.24345092475414276,
+      "learning_rate": 8.932193267409465e-06,
+      "loss": 0.3933,
+      "step": 1036
+    },
+    {
+      "epoch": 0.873385738349242,
+      "grad_norm": 0.2580495774745941,
+      "learning_rate": 8.929163190241157e-06,
+      "loss": 0.3869,
+      "step": 1037
+    },
+    {
+      "epoch": 0.8742279618192027,
+      "grad_norm": 0.2621128559112549,
+      "learning_rate": 8.92612933548335e-06,
+      "loss": 0.3825,
+      "step": 1038
+    },
+    {
+      "epoch": 0.8750701852891634,
+      "grad_norm": 0.24727575480937958,
+      "learning_rate": 8.923091706052863e-06,
+      "loss": 0.3659,
+      "step": 1039
+    },
+    {
+      "epoch": 0.8759124087591241,
+      "grad_norm": 0.25439199805259705,
+      "learning_rate": 8.920050304870142e-06,
+      "loss": 0.3817,
+      "step": 1040
+    },
+    {
+      "epoch": 0.8767546322290848,
+      "grad_norm": 0.2469315528869629,
+      "learning_rate": 8.917005134859263e-06,
+      "loss": 0.3881,
+      "step": 1041
+    },
+    {
+      "epoch": 0.8775968556990454,
+      "grad_norm": 0.21084031462669373,
+      "learning_rate": 8.913956198947923e-06,
+      "loss": 0.3784,
+      "step": 1042
+    },
+    {
+      "epoch": 0.8784390791690062,
+      "grad_norm": 0.24740955233573914,
+      "learning_rate": 8.910903500067443e-06,
+      "loss": 0.3902,
+      "step": 1043
+    },
+    {
+      "epoch": 0.8792813026389669,
+      "grad_norm": 0.25529807806015015,
+      "learning_rate": 8.907847041152757e-06,
+      "loss": 0.391,
+      "step": 1044
+    },
+    {
+      "epoch": 0.8801235261089275,
+      "grad_norm": 0.24274209141731262,
+      "learning_rate": 8.904786825142416e-06,
+      "loss": 0.3742,
+      "step": 1045
+    },
+    {
+      "epoch": 0.8809657495788883,
+      "grad_norm": 0.2258424311876297,
+      "learning_rate": 8.901722854978582e-06,
+      "loss": 0.3998,
+      "step": 1046
+    },
+    {
+      "epoch": 0.881807973048849,
+      "grad_norm": 0.24744433164596558,
+      "learning_rate": 8.89865513360703e-06,
+      "loss": 0.3859,
+      "step": 1047
+    },
+    {
+      "epoch": 0.8826501965188096,
+      "grad_norm": 0.2778688669204712,
+      "learning_rate": 8.89558366397714e-06,
+      "loss": 0.3787,
+      "step": 1048
+    },
+    {
+      "epoch": 0.8834924199887704,
+      "grad_norm": 0.22339139878749847,
+      "learning_rate": 8.892508449041893e-06,
+      "loss": 0.3919,
+      "step": 1049
+    },
+    {
+      "epoch": 0.884334643458731,
+      "grad_norm": 0.2479926347732544,
+      "learning_rate": 8.889429491757872e-06,
+      "loss": 0.3939,
+      "step": 1050
+    },
+    {
+      "epoch": 0.8851768669286918,
+      "grad_norm": 0.22446775436401367,
+      "learning_rate": 8.88634679508526e-06,
+      "loss": 0.3878,
+      "step": 1051
+    },
+    {
+      "epoch": 0.8860190903986525,
+      "grad_norm": 0.2394167184829712,
+      "learning_rate": 8.883260361987833e-06,
+      "loss": 0.3815,
+      "step": 1052
+    },
+    {
+      "epoch": 0.8868613138686131,
+      "grad_norm": 0.22700078785419464,
+      "learning_rate": 8.88017019543296e-06,
+      "loss": 0.3829,
+      "step": 1053
+    },
+    {
+      "epoch": 0.8877035373385739,
+      "grad_norm": 0.2256840020418167,
+      "learning_rate": 8.8770762983916e-06,
+      "loss": 0.4209,
+      "step": 1054
+    },
+    {
+      "epoch": 0.8885457608085345,
+      "grad_norm": 0.25572511553764343,
+      "learning_rate": 8.8739786738383e-06,
+      "loss": 0.3911,
+      "step": 1055
+    },
+    {
+      "epoch": 0.8893879842784952,
+      "grad_norm": 0.2469644397497177,
+      "learning_rate": 8.870877324751186e-06,
+      "loss": 0.4057,
+      "step": 1056
+    },
+    {
+      "epoch": 0.890230207748456,
+      "grad_norm": 0.23722214996814728,
+      "learning_rate": 8.867772254111966e-06,
+      "loss": 0.3695,
+      "step": 1057
+    },
+    {
+      "epoch": 0.8910724312184166,
+      "grad_norm": 0.22236628830432892,
+      "learning_rate": 8.864663464905933e-06,
+      "loss": 0.3813,
+      "step": 1058
+    },
+    {
+      "epoch": 0.8919146546883773,
+      "grad_norm": 0.21829207241535187,
+      "learning_rate": 8.861550960121946e-06,
+      "loss": 0.3859,
+      "step": 1059
+    },
+    {
+      "epoch": 0.892756878158338,
+      "grad_norm": 0.23551766574382782,
+      "learning_rate": 8.85843474275244e-06,
+      "loss": 0.4064,
+      "step": 1060
+    },
+    {
+      "epoch": 0.8935991016282987,
+      "grad_norm": 0.23779690265655518,
+      "learning_rate": 8.85531481579342e-06,
+      "loss": 0.3987,
+      "step": 1061
+    },
+    {
+      "epoch": 0.8944413250982594,
+      "grad_norm": 0.21609297394752502,
+      "learning_rate": 8.852191182244456e-06,
+      "loss": 0.3892,
+      "step": 1062
+    },
+    {
+      "epoch": 0.8952835485682201,
+      "grad_norm": 0.21627114713191986,
+      "learning_rate": 8.849063845108685e-06,
+      "loss": 0.3788,
+      "step": 1063
+    },
+    {
+      "epoch": 0.8961257720381808,
+      "grad_norm": 0.2338523417711258,
+      "learning_rate": 8.8459328073928e-06,
+      "loss": 0.3692,
+      "step": 1064
+    },
+    {
+      "epoch": 0.8969679955081415,
+      "grad_norm": 0.2511400282382965,
+      "learning_rate": 8.842798072107055e-06,
+      "loss": 0.3882,
+      "step": 1065
+    },
+    {
+      "epoch": 0.8978102189781022,
+      "grad_norm": 0.2453722059726715,
+      "learning_rate": 8.839659642265259e-06,
+      "loss": 0.3827,
+      "step": 1066
+    },
+    {
+      "epoch": 0.8986524424480629,
+      "grad_norm": 0.23561310768127441,
+      "learning_rate": 8.836517520884768e-06,
+      "loss": 0.3657,
+      "step": 1067
+    },
+    {
+      "epoch": 0.8994946659180236,
+      "grad_norm": 0.24510298669338226,
+      "learning_rate": 8.833371710986493e-06,
+      "loss": 0.3761,
+      "step": 1068
+    },
+    {
+      "epoch": 0.9003368893879843,
+      "grad_norm": 0.24968662858009338,
+      "learning_rate": 8.83022221559489e-06,
+      "loss": 0.3942,
+      "step": 1069
+    },
+    {
+      "epoch": 0.9011791128579449,
+      "grad_norm": 0.23057861626148224,
+      "learning_rate": 8.827069037737958e-06,
+      "loss": 0.3946,
+      "step": 1070
+    },
+    {
+      "epoch": 0.9020213363279057,
+      "grad_norm": 0.23505951464176178,
+      "learning_rate": 8.823912180447237e-06,
+      "loss": 0.3683,
+      "step": 1071
+    },
+    {
+      "epoch": 0.9028635597978664,
+      "grad_norm": 0.2223382592201233,
+      "learning_rate": 8.820751646757798e-06,
+      "loss": 0.3973,
+      "step": 1072
+    },
+    {
+      "epoch": 0.903705783267827,
+      "grad_norm": 0.21368998289108276,
+      "learning_rate": 8.81758743970826e-06,
+      "loss": 0.3947,
+      "step": 1073
+    },
+    {
+      "epoch": 0.9045480067377878,
+      "grad_norm": 0.2535130977630615,
+      "learning_rate": 8.81441956234076e-06,
+      "loss": 0.4189,
+      "step": 1074
+    },
+    {
+      "epoch": 0.9053902302077484,
+      "grad_norm": 0.24544082581996918,
+      "learning_rate": 8.81124801770097e-06,
+      "loss": 0.4065,
+      "step": 1075
+    },
+    {
+      "epoch": 0.9062324536777091,
+      "grad_norm": 0.22568069398403168,
+      "learning_rate": 8.80807280883809e-06,
+      "loss": 0.3902,
+      "step": 1076
+    },
+    {
+      "epoch": 0.9070746771476699,
+      "grad_norm": 0.24304959177970886,
+      "learning_rate": 8.804893938804839e-06,
+      "loss": 0.3855,
+      "step": 1077
+    },
+    {
+      "epoch": 0.9079169006176305,
+      "grad_norm": 0.24036061763763428,
+      "learning_rate": 8.801711410657456e-06,
+      "loss": 0.4042,
+      "step": 1078
+    },
+    {
+      "epoch": 0.9087591240875912,
+      "grad_norm": 0.252481609582901,
+      "learning_rate": 8.7985252274557e-06,
+      "loss": 0.3835,
+      "step": 1079
+    },
+    {
+      "epoch": 0.909601347557552,
+      "grad_norm": 0.2750056982040405,
+      "learning_rate": 8.795335392262841e-06,
+      "loss": 0.3889,
+      "step": 1080
+    },
+    {
+      "epoch": 0.9104435710275126,
+      "grad_norm": 0.3957831561565399,
+      "learning_rate": 8.79214190814566e-06,
+      "loss": 0.3763,
+      "step": 1081
+    },
+    {
+      "epoch": 0.9112857944974734,
+      "grad_norm": 0.24878466129302979,
+      "learning_rate": 8.78894477817445e-06,
+      "loss": 0.3649,
+      "step": 1082
+    },
+    {
+      "epoch": 0.912128017967434,
+      "grad_norm": 0.2372635006904602,
+      "learning_rate": 8.785744005423003e-06,
+      "loss": 0.3851,
+      "step": 1083
+    },
+    {
+      "epoch": 0.9129702414373947,
+      "grad_norm": 0.24955178797245026,
+      "learning_rate": 8.78253959296862e-06,
+      "loss": 0.3817,
+      "step": 1084
+    },
+    {
+      "epoch": 0.9138124649073555,
+      "grad_norm": 0.23643504083156586,
+      "learning_rate": 8.779331543892097e-06,
+      "loss": 0.3902,
+      "step": 1085
+    },
+    {
+      "epoch": 0.9146546883773161,
+      "grad_norm": 0.24997515976428986,
+      "learning_rate": 8.77611986127773e-06,
+      "loss": 0.3826,
+      "step": 1086
+    },
+    {
+      "epoch": 0.9154969118472768,
+      "grad_norm": 0.25293558835983276,
+      "learning_rate": 8.772904548213301e-06,
+      "loss": 0.3749,
+      "step": 1087
+    },
+    {
+      "epoch": 0.9163391353172375,
+      "grad_norm": 0.2371770143508911,
+      "learning_rate": 8.769685607790091e-06,
+      "loss": 0.3956,
+      "step": 1088
+    },
+    {
+      "epoch": 0.9171813587871982,
+      "grad_norm": 0.26978716254234314,
+      "learning_rate": 8.766463043102864e-06,
+      "loss": 0.3848,
+      "step": 1089
+    },
+    {
+      "epoch": 0.9180235822571589,
+      "grad_norm": 0.23945683240890503,
+      "learning_rate": 8.76323685724987e-06,
+      "loss": 0.3754,
+      "step": 1090
+    },
+    {
+      "epoch": 0.9188658057271196,
+      "grad_norm": 0.24600526690483093,
+      "learning_rate": 8.760007053332837e-06,
+      "loss": 0.3933,
+      "step": 1091
+    },
+    {
+      "epoch": 0.9197080291970803,
+      "grad_norm": 0.2496621310710907,
+      "learning_rate": 8.756773634456975e-06,
+      "loss": 0.4009,
+      "step": 1092
+    },
+    {
+      "epoch": 0.9205502526670409,
+      "grad_norm": 0.24451924860477448,
+      "learning_rate": 8.75353660373097e-06,
+      "loss": 0.4051,
+      "step": 1093
+    },
+    {
+      "epoch": 0.9213924761370017,
+      "grad_norm": 0.21420931816101074,
+      "learning_rate": 8.750295964266979e-06,
+      "loss": 0.3956,
+      "step": 1094
+    },
+    {
+      "epoch": 0.9222346996069624,
+      "grad_norm": 0.2390458881855011,
+      "learning_rate": 8.747051719180626e-06,
+      "loss": 0.4074,
+      "step": 1095
+    },
+    {
+      "epoch": 0.9230769230769231,
+      "grad_norm": 0.23600642383098602,
+      "learning_rate": 8.743803871591008e-06,
+      "loss": 0.3944,
+      "step": 1096
+    },
+    {
+      "epoch": 0.9239191465468838,
+      "grad_norm": 0.22579824924468994,
+      "learning_rate": 8.740552424620679e-06,
+      "loss": 0.3882,
+      "step": 1097
+    },
+    {
+      "epoch": 0.9247613700168444,
+      "grad_norm": 0.23936140537261963,
+      "learning_rate": 8.737297381395657e-06,
+      "loss": 0.3863,
+      "step": 1098
+    },
+    {
+      "epoch": 0.9256035934868052,
+      "grad_norm": 0.23146341741085052,
+      "learning_rate": 8.734038745045419e-06,
+      "loss": 0.3862,
+      "step": 1099
+    },
+    {
+      "epoch": 0.9264458169567659,
+      "grad_norm": 0.23204730451107025,
+      "learning_rate": 8.730776518702891e-06,
+      "loss": 0.3848,
+      "step": 1100
+    },
+    {
+      "epoch": 0.9272880404267265,
+      "grad_norm": 0.22360917925834656,
+      "learning_rate": 8.727510705504453e-06,
+      "loss": 0.3761,
+      "step": 1101
+    },
+    {
+      "epoch": 0.9281302638966873,
+      "grad_norm": 0.24749885499477386,
+      "learning_rate": 8.72424130858994e-06,
+      "loss": 0.4139,
+      "step": 1102
+    },
+    {
+      "epoch": 0.928972487366648,
+      "grad_norm": 0.21169915795326233,
+      "learning_rate": 8.72096833110262e-06,
+      "loss": 0.3653,
+      "step": 1103
+    },
+    {
+      "epoch": 0.9298147108366086,
+      "grad_norm": 0.24596849083900452,
+      "learning_rate": 8.717691776189214e-06,
+      "loss": 0.3915,
+      "step": 1104
+    },
+    {
+      "epoch": 0.9306569343065694,
+      "grad_norm": 0.2101602554321289,
+      "learning_rate": 8.714411646999878e-06,
+      "loss": 0.3699,
+      "step": 1105
+    },
+    {
+      "epoch": 0.93149915777653,
+      "grad_norm": 0.22556227445602417,
+      "learning_rate": 8.711127946688207e-06,
+      "loss": 0.4065,
+      "step": 1106
+    },
+    {
+      "epoch": 0.9323413812464907,
+      "grad_norm": 0.22517412900924683,
+      "learning_rate": 8.707840678411223e-06,
+      "loss": 0.3841,
+      "step": 1107
+    },
+    {
+      "epoch": 0.9331836047164515,
+      "grad_norm": 0.23653604090213776,
+      "learning_rate": 8.704549845329386e-06,
+      "loss": 0.3882,
+      "step": 1108
+    },
+    {
+      "epoch": 0.9340258281864121,
+      "grad_norm": 0.21519114077091217,
+      "learning_rate": 8.701255450606579e-06,
+      "loss": 0.3755,
+      "step": 1109
+    },
+    {
+      "epoch": 0.9348680516563729,
+      "grad_norm": 0.25170084834098816,
+      "learning_rate": 8.69795749741011e-06,
+      "loss": 0.408,
+      "step": 1110
+    },
+    {
+      "epoch": 0.9357102751263335,
+      "grad_norm": 0.23984849452972412,
+      "learning_rate": 8.694655988910707e-06,
+      "loss": 0.4137,
+      "step": 1111
+    },
+    {
+      "epoch": 0.9365524985962942,
+      "grad_norm": 0.22293125092983246,
+      "learning_rate": 8.69135092828252e-06,
+      "loss": 0.3885,
+      "step": 1112
+    },
+    {
+      "epoch": 0.937394722066255,
+      "grad_norm": 0.23699672520160675,
+      "learning_rate": 8.688042318703111e-06,
+      "loss": 0.3867,
+      "step": 1113
+    },
+    {
+      "epoch": 0.9382369455362156,
+      "grad_norm": 0.252868115901947,
+      "learning_rate": 8.684730163353457e-06,
+      "loss": 0.3769,
+      "step": 1114
+    },
+    {
+      "epoch": 0.9390791690061763,
+      "grad_norm": 0.24049875140190125,
+      "learning_rate": 8.681414465417936e-06,
+      "loss": 0.3782,
+      "step": 1115
+    },
+    {
+      "epoch": 0.939921392476137,
+      "grad_norm": 0.22941753268241882,
+      "learning_rate": 8.678095228084343e-06,
+      "loss": 0.4092,
+      "step": 1116
+    },
+    {
+      "epoch": 0.9407636159460977,
+      "grad_norm": 0.20538175106048584,
+      "learning_rate": 8.674772454543869e-06,
+      "loss": 0.4033,
+      "step": 1117
+    },
+    {
+      "epoch": 0.9416058394160584,
+      "grad_norm": 0.23638539016246796,
+      "learning_rate": 8.671446147991103e-06,
+      "loss": 0.3618,
+      "step": 1118
+    },
+    {
+      "epoch": 0.9424480628860191,
+      "grad_norm": 0.22866426408290863,
+      "learning_rate": 8.66811631162404e-06,
+      "loss": 0.3864,
+      "step": 1119
+    },
+    {
+      "epoch": 0.9432902863559798,
+      "grad_norm": 0.2309676706790924,
+      "learning_rate": 8.664782948644058e-06,
+      "loss": 0.4028,
+      "step": 1120
+    },
+    {
+      "epoch": 0.9441325098259404,
+      "grad_norm": 0.23818331956863403,
+      "learning_rate": 8.661446062255931e-06,
+      "loss": 0.3819,
+      "step": 1121
+    },
+    {
+      "epoch": 0.9449747332959012,
+      "grad_norm": 0.221882164478302,
+      "learning_rate": 8.65810565566782e-06,
+      "loss": 0.3762,
+      "step": 1122
+    },
+    {
+      "epoch": 0.9458169567658619,
+      "grad_norm": 0.25735414028167725,
+      "learning_rate": 8.654761732091271e-06,
+      "loss": 0.4059,
+      "step": 1123
+    },
+    {
+      "epoch": 0.9466591802358225,
+      "grad_norm": 0.23950034379959106,
+      "learning_rate": 8.65141429474121e-06,
+      "loss": 0.3819,
+      "step": 1124
+    },
+    {
+      "epoch": 0.9475014037057833,
+      "grad_norm": 0.22316765785217285,
+      "learning_rate": 8.648063346835943e-06,
+      "loss": 0.3974,
+      "step": 1125
+    },
+    {
+      "epoch": 0.9483436271757439,
+      "grad_norm": 0.23847147822380066,
+      "learning_rate": 8.644708891597147e-06,
+      "loss": 0.3841,
+      "step": 1126
+    },
+    {
+      "epoch": 0.9491858506457047,
+      "grad_norm": 0.21706604957580566,
+      "learning_rate": 8.641350932249876e-06,
+      "loss": 0.3754,
+      "step": 1127
+    },
+    {
+      "epoch": 0.9500280741156654,
+      "grad_norm": 0.22496441006660461,
+      "learning_rate": 8.637989472022548e-06,
+      "loss": 0.3879,
+      "step": 1128
+    },
+    {
+      "epoch": 0.950870297585626,
+      "grad_norm": 0.22241145372390747,
+      "learning_rate": 8.634624514146954e-06,
+      "loss": 0.3895,
+      "step": 1129
+    },
+    {
+      "epoch": 0.9517125210555868,
+      "grad_norm": 0.23969368636608124,
+      "learning_rate": 8.631256061858238e-06,
+      "loss": 0.3706,
+      "step": 1130
+    },
+    {
+      "epoch": 0.9525547445255474,
+      "grad_norm": 0.20851540565490723,
+      "learning_rate": 8.627884118394913e-06,
+      "loss": 0.4077,
+      "step": 1131
+    },
+    {
+      "epoch": 0.9533969679955081,
+      "grad_norm": 0.21636967360973358,
+      "learning_rate": 8.624508686998846e-06,
+      "loss": 0.3916,
+      "step": 1132
+    },
+    {
+      "epoch": 0.9542391914654689,
+      "grad_norm": 0.251487672328949,
+      "learning_rate": 8.621129770915248e-06,
+      "loss": 0.3586,
+      "step": 1133
+    },
+    {
+      "epoch": 0.9550814149354295,
+      "grad_norm": 0.33848926424980164,
+      "learning_rate": 8.617747373392697e-06,
+      "loss": 0.376,
+      "step": 1134
+    },
+    {
+      "epoch": 0.9559236384053902,
+      "grad_norm": 0.23164796829223633,
+      "learning_rate": 8.614361497683102e-06,
+      "loss": 0.3761,
+      "step": 1135
+    },
+    {
+      "epoch": 0.956765861875351,
+      "grad_norm": 0.2592570185661316,
+      "learning_rate": 8.61097214704173e-06,
+      "loss": 0.3959,
+      "step": 1136
+    },
+    {
+      "epoch": 0.9576080853453116,
+      "grad_norm": 0.22420254349708557,
+      "learning_rate": 8.607579324727175e-06,
+      "loss": 0.398,
+      "step": 1137
+    },
+    {
+      "epoch": 0.9584503088152723,
+      "grad_norm": 0.24111324548721313,
+      "learning_rate": 8.60418303400138e-06,
+      "loss": 0.3896,
+      "step": 1138
+    },
+    {
+      "epoch": 0.959292532285233,
+      "grad_norm": 0.2716215252876282,
+      "learning_rate": 8.600783278129617e-06,
+      "loss": 0.4066,
+      "step": 1139
+    },
+    {
+      "epoch": 0.9601347557551937,
+      "grad_norm": 0.25397905707359314,
+      "learning_rate": 8.597380060380493e-06,
+      "loss": 0.4086,
+      "step": 1140
+    },
+    {
+      "epoch": 0.9609769792251545,
+      "grad_norm": 0.23300889134407043,
+      "learning_rate": 8.59397338402594e-06,
+      "loss": 0.3715,
+      "step": 1141
+    },
+    {
+      "epoch": 0.9618192026951151,
+      "grad_norm": 0.2561149299144745,
+      "learning_rate": 8.590563252341216e-06,
+      "loss": 0.3704,
+      "step": 1142
+    },
+    {
+      "epoch": 0.9626614261650758,
+      "grad_norm": 0.24080070853233337,
+      "learning_rate": 8.5871496686049e-06,
+      "loss": 0.3583,
+      "step": 1143
+    },
+    {
+      "epoch": 0.9635036496350365,
+      "grad_norm": 0.28162842988967896,
+      "learning_rate": 8.583732636098895e-06,
+      "loss": 0.3998,
+      "step": 1144
+    },
+    {
+      "epoch": 0.9643458731049972,
+      "grad_norm": 0.24916066229343414,
+      "learning_rate": 8.580312158108413e-06,
+      "loss": 0.3781,
+      "step": 1145
+    },
+    {
+      "epoch": 0.9651880965749579,
+      "grad_norm": 0.24553634226322174,
+      "learning_rate": 8.576888237921983e-06,
+      "loss": 0.3674,
+      "step": 1146
+    },
+    {
+      "epoch": 0.9660303200449186,
+      "grad_norm": 0.24306859076023102,
+      "learning_rate": 8.57346087883144e-06,
+      "loss": 0.3797,
+      "step": 1147
+    },
+    {
+      "epoch": 0.9668725435148793,
+      "grad_norm": 0.24062088131904602,
+      "learning_rate": 8.570030084131933e-06,
+      "loss": 0.3868,
+      "step": 1148
+    },
+    {
+      "epoch": 0.9677147669848399,
+      "grad_norm": 0.2654595375061035,
+      "learning_rate": 8.566595857121902e-06,
+      "loss": 0.3875,
+      "step": 1149
+    },
+    {
+      "epoch": 0.9685569904548007,
+      "grad_norm": 0.24021129310131073,
+      "learning_rate": 8.563158201103096e-06,
+      "loss": 0.385,
+      "step": 1150
+    },
+    {
+      "epoch": 0.9693992139247614,
+      "grad_norm": 0.2687273919582367,
+      "learning_rate": 8.559717119380558e-06,
+      "loss": 0.3873,
+      "step": 1151
+    },
+    {
+      "epoch": 0.970241437394722,
+      "grad_norm": 0.28822076320648193,
+      "learning_rate": 8.556272615262623e-06,
+      "loss": 0.372,
+      "step": 1152
+    },
+    {
+      "epoch": 0.9710836608646828,
+      "grad_norm": 0.23251932859420776,
+      "learning_rate": 8.55282469206092e-06,
+      "loss": 0.371,
+      "step": 1153
+    },
+    {
+      "epoch": 0.9719258843346434,
+      "grad_norm": 0.231855109333992,
+      "learning_rate": 8.549373353090362e-06,
+      "loss": 0.374,
+      "step": 1154
+    },
+    {
+      "epoch": 0.9727681078046042,
+      "grad_norm": 0.2325243204832077,
+      "learning_rate": 8.545918601669147e-06,
+      "loss": 0.3623,
+      "step": 1155
+    },
+    {
+      "epoch": 0.9736103312745649,
+      "grad_norm": 0.24317185580730438,
+      "learning_rate": 8.542460441118756e-06,
+      "loss": 0.3985,
+      "step": 1156
+    },
+    {
+      "epoch": 0.9744525547445255,
+      "grad_norm": 0.2409314215183258,
+      "learning_rate": 8.538998874763942e-06,
+      "loss": 0.4072,
+      "step": 1157
+    },
+    {
+      "epoch": 0.9752947782144863,
+      "grad_norm": 0.22322307527065277,
+      "learning_rate": 8.535533905932739e-06,
+      "loss": 0.3968,
+      "step": 1158
+    },
+    {
+      "epoch": 0.976137001684447,
+      "grad_norm": 0.22121012210845947,
+      "learning_rate": 8.532065537956446e-06,
+      "loss": 0.3817,
+      "step": 1159
+    },
+    {
+      "epoch": 0.9769792251544076,
+      "grad_norm": 0.23832759261131287,
+      "learning_rate": 8.528593774169637e-06,
+      "loss": 0.4015,
+      "step": 1160
+    },
+    {
+      "epoch": 0.9778214486243684,
+      "grad_norm": 0.2314673662185669,
+      "learning_rate": 8.525118617910144e-06,
+      "loss": 0.3869,
+      "step": 1161
+    },
+    {
+      "epoch": 0.978663672094329,
+      "grad_norm": 0.23528188467025757,
+      "learning_rate": 8.521640072519066e-06,
+      "loss": 0.3786,
+      "step": 1162
+    },
+    {
+      "epoch": 0.9795058955642897,
+      "grad_norm": 0.2287149727344513,
+      "learning_rate": 8.518158141340755e-06,
+      "loss": 0.4026,
+      "step": 1163
+    },
+    {
+      "epoch": 0.9803481190342505,
+      "grad_norm": 0.2339918166399002,
+      "learning_rate": 8.514672827722824e-06,
+      "loss": 0.3845,
+      "step": 1164
+    },
+    {
+      "epoch": 0.9811903425042111,
+      "grad_norm": 0.24464187026023865,
+      "learning_rate": 8.511184135016134e-06,
+      "loss": 0.3893,
+      "step": 1165
+    },
+    {
+      "epoch": 0.9820325659741718,
+      "grad_norm": 0.2102963775396347,
+      "learning_rate": 8.507692066574795e-06,
+      "loss": 0.3766,
+      "step": 1166
+    },
+    {
+      "epoch": 0.9828747894441325,
+      "grad_norm": 0.20385922491550446,
+      "learning_rate": 8.504196625756166e-06,
+      "loss": 0.3832,
+      "step": 1167
+    },
+    {
+      "epoch": 0.9837170129140932,
+      "grad_norm": 0.2298925369977951,
+      "learning_rate": 8.500697815920843e-06,
+      "loss": 0.3788,
+      "step": 1168
+    },
+    {
+      "epoch": 0.9845592363840538,
+      "grad_norm": 0.22536714375019073,
+      "learning_rate": 8.497195640432664e-06,
+      "loss": 0.3747,
+      "step": 1169
+    },
+    {
+      "epoch": 0.9854014598540146,
+      "grad_norm": 0.240451842546463,
+      "learning_rate": 8.493690102658703e-06,
+      "loss": 0.4021,
+      "step": 1170
+    },
+    {
+      "epoch": 0.9862436833239753,
+      "grad_norm": 0.24420315027236938,
+      "learning_rate": 8.490181205969268e-06,
+      "loss": 0.4098,
+      "step": 1171
+    },
+    {
+      "epoch": 0.987085906793936,
+      "grad_norm": 0.23833104968070984,
+      "learning_rate": 8.486668953737891e-06,
+      "loss": 0.392,
+      "step": 1172
+    },
+    {
+      "epoch": 0.9879281302638967,
+      "grad_norm": 0.24562418460845947,
+      "learning_rate": 8.483153349341336e-06,
+      "loss": 0.4039,
+      "step": 1173
+    },
+    {
+      "epoch": 0.9887703537338574,
+      "grad_norm": 0.20323196053504944,
+      "learning_rate": 8.479634396159587e-06,
+      "loss": 0.3997,
+      "step": 1174
+    },
+    {
+      "epoch": 0.9896125772038181,
+      "grad_norm": 0.24542491137981415,
+      "learning_rate": 8.476112097575845e-06,
+      "loss": 0.3791,
+      "step": 1175
+    },
+    {
+      "epoch": 0.9904548006737788,
+      "grad_norm": 0.22892843186855316,
+      "learning_rate": 8.472586456976534e-06,
+      "loss": 0.3702,
+      "step": 1176
+    },
+    {
+      "epoch": 0.9912970241437394,
+      "grad_norm": 0.21064841747283936,
+      "learning_rate": 8.46905747775129e-06,
+      "loss": 0.3862,
+      "step": 1177
+    },
+    {
+      "epoch": 0.9921392476137002,
+      "grad_norm": 0.24221305549144745,
+      "learning_rate": 8.465525163292948e-06,
+      "loss": 0.3888,
+      "step": 1178
+    },
+    {
+      "epoch": 0.9929814710836609,
+      "grad_norm": 0.2127092331647873,
+      "learning_rate": 8.461989516997565e-06,
+      "loss": 0.4031,
+      "step": 1179
+    },
+    {
+      "epoch": 0.9938236945536215,
+      "grad_norm": 0.24036602675914764,
+      "learning_rate": 8.458450542264391e-06,
+      "loss": 0.3967,
+      "step": 1180
+    },
+    {
+      "epoch": 0.9946659180235823,
+      "grad_norm": 0.2219998985528946,
+      "learning_rate": 8.45490824249588e-06,
+      "loss": 0.3774,
+      "step": 1181
+    },
+    {
+      "epoch": 0.9955081414935429,
+      "grad_norm": 0.2132783681154251,
+      "learning_rate": 8.45136262109768e-06,
+      "loss": 0.3636,
+      "step": 1182
+    },
+    {
+      "epoch": 0.9963503649635036,
+      "grad_norm": 0.23521870374679565,
+      "learning_rate": 8.447813681478638e-06,
+      "loss": 0.3985,
+      "step": 1183
+    },
+    {
+      "epoch": 0.9971925884334644,
+      "grad_norm": 0.2551150918006897,
+      "learning_rate": 8.444261427050786e-06,
+      "loss": 0.4037,
+      "step": 1184
+    },
+    {
+      "epoch": 0.998034811903425,
+      "grad_norm": 0.21768134832382202,
+      "learning_rate": 8.440705861229344e-06,
+      "loss": 0.3734,
+      "step": 1185
+    },
+    {
+      "epoch": 0.9988770353733858,
+      "grad_norm": 0.22760789096355438,
+      "learning_rate": 8.437146987432717e-06,
+      "loss": 0.3958,
+      "step": 1186
+    },
+    {
+      "epoch": 0.9997192588433464,
+      "grad_norm": 0.22547492384910583,
+      "learning_rate": 8.43358480908249e-06,
+      "loss": 0.3673,
+      "step": 1187
+    },
+    {
+      "epoch": 1.000561482313307,
+      "grad_norm": 0.41303884983062744,
+      "learning_rate": 8.430019329603423e-06,
+      "loss": 0.5973,
+      "step": 1188
+    },
+    {
+      "epoch": 1.0014037057832679,
+      "grad_norm": 0.26283347606658936,
+      "learning_rate": 8.426450552423451e-06,
+      "loss": 0.3644,
+      "step": 1189
+    },
+    {
+      "epoch": 1.0022459292532284,
+      "grad_norm": 0.25336721539497375,
+      "learning_rate": 8.422878480973681e-06,
+      "loss": 0.3421,
+      "step": 1190
+    },
+    {
+      "epoch": 1.0030881527231892,
+      "grad_norm": 0.263960599899292,
+      "learning_rate": 8.41930311868839e-06,
+      "loss": 0.3677,
+      "step": 1191
+    },
+    {
+      "epoch": 1.00393037619315,
+      "grad_norm": 0.24778808653354645,
+      "learning_rate": 8.41572446900501e-06,
+      "loss": 0.3819,
+      "step": 1192
+    },
+    {
+      "epoch": 1.0047725996631107,
+      "grad_norm": 0.2429429590702057,
+      "learning_rate": 8.412142535364139e-06,
+      "loss": 0.3124,
+      "step": 1193
+    },
+    {
+      "epoch": 1.0056148231330713,
+      "grad_norm": 0.22689130902290344,
+      "learning_rate": 8.408557321209534e-06,
+      "loss": 0.365,
+      "step": 1194
+    },
+    {
+      "epoch": 1.006457046603032,
+      "grad_norm": 0.26087427139282227,
+      "learning_rate": 8.404968829988102e-06,
+      "loss": 0.3763,
+      "step": 1195
+    },
+    {
+      "epoch": 1.0072992700729928,
+      "grad_norm": 0.2285303771495819,
+      "learning_rate": 8.401377065149904e-06,
+      "loss": 0.3504,
+      "step": 1196
+    },
+    {
+      "epoch": 1.0081414935429533,
+      "grad_norm": 0.2542676031589508,
+      "learning_rate": 8.397782030148147e-06,
+      "loss": 0.3699,
+      "step": 1197
+    },
+    {
+      "epoch": 1.0089837170129141,
+      "grad_norm": 0.22694192826747894,
+      "learning_rate": 8.39418372843918e-06,
+      "loss": 0.3319,
+      "step": 1198
+    },
+    {
+      "epoch": 1.0098259404828749,
+      "grad_norm": 0.23784025013446808,
+      "learning_rate": 8.390582163482497e-06,
+      "loss": 0.3467,
+      "step": 1199
+    },
+    {
+      "epoch": 1.0106681639528354,
+      "grad_norm": 0.24005892872810364,
+      "learning_rate": 8.386977338740724e-06,
+      "loss": 0.351,
+      "step": 1200
+    },
+    {
+      "epoch": 1.0115103874227962,
+      "grad_norm": 0.2603375315666199,
+      "learning_rate": 8.383369257679625e-06,
+      "loss": 0.3975,
+      "step": 1201
+    },
+    {
+      "epoch": 1.012352610892757,
+      "grad_norm": 0.22591696679592133,
+      "learning_rate": 8.379757923768094e-06,
+      "loss": 0.3371,
+      "step": 1202
+    },
+    {
+      "epoch": 1.0131948343627175,
+      "grad_norm": 0.23937468230724335,
+      "learning_rate": 8.376143340478153e-06,
+      "loss": 0.384,
+      "step": 1203
+    },
+    {
+      "epoch": 1.0140370578326783,
+      "grad_norm": 0.25090497732162476,
+      "learning_rate": 8.372525511284945e-06,
+      "loss": 0.3625,
+      "step": 1204
+    },
+    {
+      "epoch": 1.014879281302639,
+      "grad_norm": 0.2606372535228729,
+      "learning_rate": 8.368904439666739e-06,
+      "loss": 0.3607,
+      "step": 1205
+    },
+    {
+      "epoch": 1.0157215047725996,
+      "grad_norm": 0.20606310665607452,
+      "learning_rate": 8.365280129104912e-06,
+      "loss": 0.3482,
+      "step": 1206
+    },
+    {
+      "epoch": 1.0165637282425604,
+      "grad_norm": 0.2573789060115814,
+      "learning_rate": 8.361652583083968e-06,
+      "loss": 0.3303,
+      "step": 1207
+    },
+    {
+      "epoch": 1.0174059517125211,
+      "grad_norm": 0.23522531986236572,
+      "learning_rate": 8.358021805091509e-06,
+      "loss": 0.3569,
+      "step": 1208
+    },
+    {
+      "epoch": 1.0182481751824817,
+      "grad_norm": 0.22316160798072815,
+      "learning_rate": 8.354387798618254e-06,
+      "loss": 0.3443,
+      "step": 1209
+    },
+    {
+      "epoch": 1.0190903986524424,
+      "grad_norm": 0.2838685214519501,
+      "learning_rate": 8.35075056715802e-06,
+      "loss": 0.3429,
+      "step": 1210
+    },
+    {
+      "epoch": 1.0199326221224032,
+      "grad_norm": 0.22736883163452148,
+      "learning_rate": 8.347110114207727e-06,
+      "loss": 0.3963,
+      "step": 1211
+    },
+    {
+      "epoch": 1.0207748455923638,
+      "grad_norm": 0.2243269979953766,
+      "learning_rate": 8.34346644326739e-06,
+      "loss": 0.3273,
+      "step": 1212
+    },
+    {
+      "epoch": 1.0216170690623245,
+      "grad_norm": 0.23876528441905975,
+      "learning_rate": 8.339819557840124e-06,
+      "loss": 0.392,
+      "step": 1213
+    },
+    {
+      "epoch": 1.0224592925322853,
+      "grad_norm": 0.2140982449054718,
+      "learning_rate": 8.336169461432125e-06,
+      "loss": 0.3563,
+      "step": 1214
+    },
+    {
+      "epoch": 1.0233015160022458,
+      "grad_norm": 0.24384154379367828,
+      "learning_rate": 8.332516157552684e-06,
+      "loss": 0.3981,
+      "step": 1215
+    },
+    {
+      "epoch": 1.0241437394722066,
+      "grad_norm": 0.1963784545660019,
+      "learning_rate": 8.328859649714171e-06,
+      "loss": 0.3368,
+      "step": 1216
+    },
+    {
+      "epoch": 1.0249859629421674,
+      "grad_norm": 0.23279255628585815,
+      "learning_rate": 8.32519994143204e-06,
+      "loss": 0.3863,
+      "step": 1217
+    },
+    {
+      "epoch": 1.025828186412128,
+      "grad_norm": 0.19129885733127594,
+      "learning_rate": 8.321537036224822e-06,
+      "loss": 0.3284,
+      "step": 1218
+    },
+    {
+      "epoch": 1.0266704098820887,
+      "grad_norm": 0.22419750690460205,
+      "learning_rate": 8.317870937614115e-06,
+      "loss": 0.3691,
+      "step": 1219
+    },
+    {
+      "epoch": 1.0275126333520495,
+      "grad_norm": 0.22040456533432007,
+      "learning_rate": 8.314201649124595e-06,
+      "loss": 0.3606,
+      "step": 1220
+    },
+    {
+      "epoch": 1.02835485682201,
+      "grad_norm": 0.2307283580303192,
+      "learning_rate": 8.310529174284004e-06,
+      "loss": 0.3419,
+      "step": 1221
+    },
+    {
+      "epoch": 1.0291970802919708,
+      "grad_norm": 0.23188059031963348,
+      "learning_rate": 8.30685351662314e-06,
+      "loss": 0.3544,
+      "step": 1222
+    },
+    {
+      "epoch": 1.0300393037619315,
+      "grad_norm": 0.2442709058523178,
+      "learning_rate": 8.30317467967587e-06,
+      "loss": 0.3719,
+      "step": 1223
+    },
+    {
+      "epoch": 1.0308815272318923,
+      "grad_norm": 0.26356396079063416,
+      "learning_rate": 8.299492666979114e-06,
+      "loss": 0.4023,
+      "step": 1224
+    },
+    {
+      "epoch": 1.0317237507018528,
+      "grad_norm": 0.22132344543933868,
+      "learning_rate": 8.295807482072842e-06,
+      "loss": 0.3593,
+      "step": 1225
+    },
+    {
+      "epoch": 1.0325659741718136,
+      "grad_norm": 0.23799309134483337,
+      "learning_rate": 8.292119128500082e-06,
+      "loss": 0.371,
+      "step": 1226
+    },
+    {
+      "epoch": 1.0334081976417744,
+      "grad_norm": 0.23775412142276764,
+      "learning_rate": 8.288427609806899e-06,
+      "loss": 0.3607,
+      "step": 1227
+    },
+    {
+      "epoch": 1.034250421111735,
+      "grad_norm": 0.2484200894832611,
+      "learning_rate": 8.28473292954241e-06,
+      "loss": 0.3769,
+      "step": 1228
+    },
+    {
+      "epoch": 1.0350926445816957,
+      "grad_norm": 0.2280905544757843,
+      "learning_rate": 8.281035091258762e-06,
+      "loss": 0.3716,
+      "step": 1229
+    },
+    {
+      "epoch": 1.0359348680516565,
+      "grad_norm": 0.24847450852394104,
+      "learning_rate": 8.277334098511147e-06,
+      "loss": 0.3766,
+      "step": 1230
+    },
+    {
+      "epoch": 1.036777091521617,
+      "grad_norm": 0.2286071479320526,
+      "learning_rate": 8.273629954857784e-06,
+      "loss": 0.3481,
+      "step": 1231
+    },
+    {
+      "epoch": 1.0376193149915778,
+      "grad_norm": 0.22194388508796692,
+      "learning_rate": 8.269922663859926e-06,
+      "loss": 0.3461,
+      "step": 1232
+    },
+    {
+      "epoch": 1.0384615384615385,
+      "grad_norm": 0.2565535604953766,
+      "learning_rate": 8.266212229081846e-06,
+      "loss": 0.3983,
+      "step": 1233
+    },
+    {
+      "epoch": 1.039303761931499,
+      "grad_norm": 0.23098818957805634,
+      "learning_rate": 8.262498654090846e-06,
+      "loss": 0.3492,
+      "step": 1234
+    },
+    {
+      "epoch": 1.0401459854014599,
+      "grad_norm": 0.25537002086639404,
+      "learning_rate": 8.258781942457244e-06,
+      "loss": 0.3692,
+      "step": 1235
+    },
+    {
+      "epoch": 1.0409882088714206,
+      "grad_norm": 0.2224876582622528,
+      "learning_rate": 8.255062097754371e-06,
+      "loss": 0.339,
+      "step": 1236
+    },
+    {
+      "epoch": 1.0418304323413812,
+      "grad_norm": 0.2605500817298889,
+      "learning_rate": 8.251339123558573e-06,
+      "loss": 0.3695,
+      "step": 1237
+    },
+    {
+      "epoch": 1.042672655811342,
+      "grad_norm": 0.22719444334506989,
+      "learning_rate": 8.247613023449209e-06,
+      "loss": 0.344,
+      "step": 1238
+    },
+    {
+      "epoch": 1.0435148792813027,
+      "grad_norm": 0.2446812242269516,
+      "learning_rate": 8.243883801008632e-06,
+      "loss": 0.407,
+      "step": 1239
+    },
+    {
+      "epoch": 1.0443571027512633,
+      "grad_norm": 0.21748603880405426,
+      "learning_rate": 8.240151459822207e-06,
+      "loss": 0.3408,
+      "step": 1240
+    },
+    {
+      "epoch": 1.045199326221224,
+      "grad_norm": 0.2330400049686432,
+      "learning_rate": 8.236416003478295e-06,
+      "loss": 0.3796,
+      "step": 1241
+    },
+    {
+      "epoch": 1.0460415496911848,
+      "grad_norm": 0.2480519860982895,
+      "learning_rate": 8.232677435568252e-06,
+      "loss": 0.3703,
+      "step": 1242
+    },
+    {
+      "epoch": 1.0468837731611453,
+      "grad_norm": 0.22414913773536682,
+      "learning_rate": 8.228935759686424e-06,
+      "loss": 0.3212,
+      "step": 1243
+    },
+    {
+      "epoch": 1.047725996631106,
+      "grad_norm": 0.24647359549999237,
+      "learning_rate": 8.225190979430145e-06,
+      "loss": 0.3837,
+      "step": 1244
+    },
+    {
+      "epoch": 1.0485682201010669,
+      "grad_norm": 0.24247120320796967,
+      "learning_rate": 8.221443098399733e-06,
+      "loss": 0.3617,
+      "step": 1245
+    },
+    {
+      "epoch": 1.0494104435710274,
+      "grad_norm": 0.24783526360988617,
+      "learning_rate": 8.217692120198492e-06,
+      "loss": 0.3767,
+      "step": 1246
+    },
+    {
+      "epoch": 1.0502526670409882,
+      "grad_norm": 0.23246386647224426,
+      "learning_rate": 8.213938048432697e-06,
+      "loss": 0.3665,
+      "step": 1247
+    },
+    {
+      "epoch": 1.051094890510949,
+      "grad_norm": 0.2402683049440384,
+      "learning_rate": 8.210180886711603e-06,
+      "loss": 0.379,
+      "step": 1248
+    },
+    {
+      "epoch": 1.0519371139809095,
+      "grad_norm": 0.2358441948890686,
+      "learning_rate": 8.206420638647433e-06,
+      "loss": 0.363,
+      "step": 1249
+    },
+    {
+      "epoch": 1.0527793374508703,
+      "grad_norm": 0.21942484378814697,
+      "learning_rate": 8.202657307855376e-06,
+      "loss": 0.3326,
+      "step": 1250
+    },
+    {
+      "epoch": 1.053621560920831,
+      "grad_norm": 0.22994016110897064,
+      "learning_rate": 8.198890897953586e-06,
+      "loss": 0.3677,
+      "step": 1251
+    },
+    {
+      "epoch": 1.0544637843907916,
+      "grad_norm": 0.23945726454257965,
+      "learning_rate": 8.19512141256318e-06,
+      "loss": 0.3956,
+      "step": 1252
+    },
+    {
+      "epoch": 1.0553060078607523,
+      "grad_norm": 0.20319385826587677,
+      "learning_rate": 8.191348855308229e-06,
+      "loss": 0.3523,
+      "step": 1253
+    },
+    {
+      "epoch": 1.0561482313307131,
+      "grad_norm": 0.22652249038219452,
+      "learning_rate": 8.187573229815757e-06,
+      "loss": 0.3419,
+      "step": 1254
+    },
+    {
+      "epoch": 1.0569904548006739,
+      "grad_norm": 0.2570735216140747,
+      "learning_rate": 8.18379453971574e-06,
+      "loss": 0.4001,
+      "step": 1255
+    },
+    {
+      "epoch": 1.0578326782706344,
+      "grad_norm": 0.22578343749046326,
+      "learning_rate": 8.180012788641097e-06,
+      "loss": 0.3478,
+      "step": 1256
+    },
+    {
+      "epoch": 1.0586749017405952,
+      "grad_norm": 0.2511613667011261,
+      "learning_rate": 8.176227980227693e-06,
+      "loss": 0.3678,
+      "step": 1257
+    },
+    {
+      "epoch": 1.059517125210556,
+      "grad_norm": 0.25504982471466064,
+      "learning_rate": 8.172440118114332e-06,
+      "loss": 0.3973,
+      "step": 1258
+    },
+    {
+      "epoch": 1.0603593486805165,
+      "grad_norm": 0.20504139363765717,
+      "learning_rate": 8.168649205942753e-06,
+      "loss": 0.3428,
+      "step": 1259
+    },
+    {
+      "epoch": 1.0612015721504773,
+      "grad_norm": 0.22897997498512268,
+      "learning_rate": 8.164855247357628e-06,
+      "loss": 0.3671,
+      "step": 1260
+    },
+    {
+      "epoch": 1.062043795620438,
+      "grad_norm": 0.2271280735731125,
+      "learning_rate": 8.161058246006558e-06,
+      "loss": 0.359,
+      "step": 1261
+    },
+    {
+      "epoch": 1.0628860190903986,
+      "grad_norm": 0.23936693370342255,
+      "learning_rate": 8.157258205540069e-06,
+      "loss": 0.3493,
+      "step": 1262
+    },
+    {
+      "epoch": 1.0637282425603594,
+      "grad_norm": 0.2141476571559906,
+      "learning_rate": 8.153455129611605e-06,
+      "loss": 0.3652,
+      "step": 1263
+    },
+    {
+      "epoch": 1.0645704660303201,
+      "grad_norm": 0.23102352023124695,
+      "learning_rate": 8.14964902187754e-06,
+      "loss": 0.3678,
+      "step": 1264
+    },
+    {
+      "epoch": 1.0654126895002807,
+      "grad_norm": 0.237512469291687,
+      "learning_rate": 8.145839885997146e-06,
+      "loss": 0.3335,
+      "step": 1265
+    },
+    {
+      "epoch": 1.0662549129702414,
+      "grad_norm": 0.2297290712594986,
+      "learning_rate": 8.142027725632622e-06,
+      "loss": 0.3592,
+      "step": 1266
+    },
+    {
+      "epoch": 1.0670971364402022,
+      "grad_norm": 0.25016435980796814,
+      "learning_rate": 8.138212544449067e-06,
+      "loss": 0.3785,
+      "step": 1267
+    },
+    {
+      "epoch": 1.0679393599101628,
+      "grad_norm": 0.22096030414104462,
+      "learning_rate": 8.134394346114486e-06,
+      "loss": 0.3609,
+      "step": 1268
+    },
+    {
+      "epoch": 1.0687815833801235,
+      "grad_norm": 0.26206687092781067,
+      "learning_rate": 8.130573134299782e-06,
+      "loss": 0.3661,
+      "step": 1269
+    },
+    {
+      "epoch": 1.0696238068500843,
+      "grad_norm": 0.24707172811031342,
+      "learning_rate": 8.126748912678757e-06,
+      "loss": 0.3965,
+      "step": 1270
+    },
+    {
+      "epoch": 1.0704660303200448,
+      "grad_norm": 0.21675467491149902,
+      "learning_rate": 8.122921684928111e-06,
+      "loss": 0.3157,
+      "step": 1271
+    },
+    {
+      "epoch": 1.0713082537900056,
+      "grad_norm": 0.22279198467731476,
+      "learning_rate": 8.119091454727427e-06,
+      "loss": 0.3631,
+      "step": 1272
+    },
+    {
+      "epoch": 1.0721504772599664,
+      "grad_norm": 0.2054135948419571,
+      "learning_rate": 8.11525822575918e-06,
+      "loss": 0.3735,
+      "step": 1273
+    },
+    {
+      "epoch": 1.072992700729927,
+      "grad_norm": 0.226912721991539,
+      "learning_rate": 8.111422001708725e-06,
+      "loss": 0.3479,
+      "step": 1274
+    },
+    {
+      "epoch": 1.0738349241998877,
+      "grad_norm": 0.23292496800422668,
+      "learning_rate": 8.107582786264299e-06,
+      "loss": 0.3569,
+      "step": 1275
+    },
+    {
+      "epoch": 1.0746771476698485,
+      "grad_norm": 0.2658878564834595,
+      "learning_rate": 8.10374058311701e-06,
+      "loss": 0.3859,
+      "step": 1276
+    },
+    {
+      "epoch": 1.075519371139809,
+      "grad_norm": 0.25955137610435486,
+      "learning_rate": 8.099895395960847e-06,
+      "loss": 0.3635,
+      "step": 1277
+    },
+    {
+      "epoch": 1.0763615946097698,
+      "grad_norm": 0.23224274814128876,
+      "learning_rate": 8.09604722849266e-06,
+      "loss": 0.36,
+      "step": 1278
+    },
+    {
+      "epoch": 1.0772038180797305,
+      "grad_norm": 0.26262158155441284,
+      "learning_rate": 8.092196084412167e-06,
+      "loss": 0.3731,
+      "step": 1279
+    },
+    {
+      "epoch": 1.078046041549691,
+      "grad_norm": 0.23350927233695984,
+      "learning_rate": 8.08834196742195e-06,
+      "loss": 0.3493,
+      "step": 1280
+    },
+    {
+      "epoch": 1.0788882650196518,
+      "grad_norm": 0.22659815847873688,
+      "learning_rate": 8.084484881227449e-06,
+      "loss": 0.3472,
+      "step": 1281
+    },
+    {
+      "epoch": 1.0797304884896126,
+      "grad_norm": 0.27644115686416626,
+      "learning_rate": 8.080624829536949e-06,
+      "loss": 0.3706,
+      "step": 1282
+    },
+    {
+      "epoch": 1.0805727119595732,
+      "grad_norm": 0.20871034264564514,
+      "learning_rate": 8.076761816061603e-06,
+      "loss": 0.3301,
+      "step": 1283
+    },
+    {
+      "epoch": 1.081414935429534,
+      "grad_norm": 0.2264663428068161,
+      "learning_rate": 8.072895844515398e-06,
+      "loss": 0.3711,
+      "step": 1284
+    },
+    {
+      "epoch": 1.0822571588994947,
+      "grad_norm": 0.22878475487232208,
+      "learning_rate": 8.069026918615173e-06,
+      "loss": 0.3576,
+      "step": 1285
+    },
+    {
+      "epoch": 1.0830993823694555,
+      "grad_norm": 0.2420300841331482,
+      "learning_rate": 8.065155042080599e-06,
+      "loss": 0.3634,
+      "step": 1286
+    },
+    {
+      "epoch": 1.083941605839416,
+      "grad_norm": 0.2258838415145874,
+      "learning_rate": 8.061280218634192e-06,
+      "loss": 0.3494,
+      "step": 1287
+    },
+    {
+      "epoch": 1.0847838293093768,
+      "grad_norm": 0.2405247539281845,
+      "learning_rate": 8.057402452001298e-06,
+      "loss": 0.3793,
+      "step": 1288
+    },
+    {
+      "epoch": 1.0856260527793375,
+      "grad_norm": 0.2510263919830322,
+      "learning_rate": 8.05352174591009e-06,
+      "loss": 0.3694,
+      "step": 1289
+    },
+    {
+      "epoch": 1.086468276249298,
+      "grad_norm": 0.23418664932250977,
+      "learning_rate": 8.049638104091575e-06,
+      "loss": 0.3633,
+      "step": 1290
+    },
+    {
+      "epoch": 1.0873104997192589,
+      "grad_norm": 0.27216869592666626,
+      "learning_rate": 8.04575153027957e-06,
+      "loss": 0.3878,
+      "step": 1291
+    },
+    {
+      "epoch": 1.0881527231892196,
+      "grad_norm": 0.22842253744602203,
+      "learning_rate": 8.041862028210725e-06,
+      "loss": 0.3384,
+      "step": 1292
+    },
+    {
+      "epoch": 1.0889949466591802,
+      "grad_norm": 0.24223503470420837,
+      "learning_rate": 8.037969601624495e-06,
+      "loss": 0.3674,
+      "step": 1293
+    },
+    {
+      "epoch": 1.089837170129141,
+      "grad_norm": 0.24100640416145325,
+      "learning_rate": 8.034074254263152e-06,
+      "loss": 0.3353,
+      "step": 1294
+    },
+    {
+      "epoch": 1.0906793935991017,
+      "grad_norm": 0.2577701210975647,
+      "learning_rate": 8.030175989871769e-06,
+      "loss": 0.3725,
+      "step": 1295
+    },
+    {
+      "epoch": 1.0915216170690623,
+      "grad_norm": 0.21651145815849304,
+      "learning_rate": 8.026274812198235e-06,
+      "loss": 0.3519,
+      "step": 1296
+    },
+    {
+      "epoch": 1.092363840539023,
+      "grad_norm": 0.22987744212150574,
+      "learning_rate": 8.022370724993229e-06,
+      "loss": 0.3541,
+      "step": 1297
+    },
+    {
+      "epoch": 1.0932060640089838,
+      "grad_norm": 0.221238911151886,
+      "learning_rate": 8.018463732010235e-06,
+      "loss": 0.4107,
+      "step": 1298
+    },
+    {
+      "epoch": 1.0940482874789443,
+      "grad_norm": 0.20016083121299744,
+      "learning_rate": 8.014553837005527e-06,
+      "loss": 0.305,
+      "step": 1299
+    },
+    {
+      "epoch": 1.094890510948905,
+      "grad_norm": 0.21442735195159912,
+      "learning_rate": 8.010641043738167e-06,
+      "loss": 0.371,
+      "step": 1300
+    },
+    {
+      "epoch": 1.0957327344188659,
+      "grad_norm": 0.24419724941253662,
+      "learning_rate": 8.006725355970008e-06,
+      "loss": 0.3874,
+      "step": 1301
+    },
+    {
+      "epoch": 1.0965749578888264,
+      "grad_norm": 0.2159401923418045,
+      "learning_rate": 8.002806777465685e-06,
+      "loss": 0.3563,
+      "step": 1302
+    },
+    {
+      "epoch": 1.0974171813587872,
+      "grad_norm": 0.20667701959609985,
+      "learning_rate": 7.99888531199261e-06,
+      "loss": 0.3299,
+      "step": 1303
+    },
+    {
+      "epoch": 1.098259404828748,
+      "grad_norm": 0.23032718896865845,
+      "learning_rate": 7.99496096332097e-06,
+      "loss": 0.3662,
+      "step": 1304
+    },
+    {
+      "epoch": 1.0991016282987085,
+      "grad_norm": 0.21561412513256073,
+      "learning_rate": 7.99103373522373e-06,
+      "loss": 0.3741,
+      "step": 1305
+    },
+    {
+      "epoch": 1.0999438517686693,
+      "grad_norm": 0.20311473309993744,
+      "learning_rate": 7.987103631476615e-06,
+      "loss": 0.3382,
+      "step": 1306
+    },
+    {
+      "epoch": 1.10078607523863,
+      "grad_norm": 0.2105664610862732,
+      "learning_rate": 7.98317065585812e-06,
+      "loss": 0.3733,
+      "step": 1307
+    },
+    {
+      "epoch": 1.1016282987085906,
+      "grad_norm": 0.20192131400108337,
+      "learning_rate": 7.9792348121495e-06,
+      "loss": 0.345,
+      "step": 1308
+    },
+    {
+      "epoch": 1.1024705221785513,
+      "grad_norm": 0.25903552770614624,
+      "learning_rate": 7.975296104134768e-06,
+      "loss": 0.4128,
+      "step": 1309
+    },
+    {
+      "epoch": 1.1033127456485121,
+      "grad_norm": 0.20803354680538177,
+      "learning_rate": 7.97135453560069e-06,
+      "loss": 0.3646,
+      "step": 1310
+    },
+    {
+      "epoch": 1.1041549691184729,
+      "grad_norm": 0.2241167426109314,
+      "learning_rate": 7.967410110336782e-06,
+      "loss": 0.3339,
+      "step": 1311
+    },
+    {
+      "epoch": 1.1049971925884334,
+      "grad_norm": 0.2666442096233368,
+      "learning_rate": 7.963462832135307e-06,
+      "loss": 0.4023,
+      "step": 1312
+    },
+    {
+      "epoch": 1.1058394160583942,
+      "grad_norm": 0.20622016489505768,
+      "learning_rate": 7.959512704791269e-06,
+      "loss": 0.3379,
+      "step": 1313
+    },
+    {
+      "epoch": 1.1066816395283547,
+      "grad_norm": 0.26777777075767517,
+      "learning_rate": 7.955559732102414e-06,
+      "loss": 0.405,
+      "step": 1314
+    },
+    {
+      "epoch": 1.1075238629983155,
+      "grad_norm": 0.2196749597787857,
+      "learning_rate": 7.951603917869223e-06,
+      "loss": 0.3563,
+      "step": 1315
+    },
+    {
+      "epoch": 1.1083660864682763,
+      "grad_norm": 0.22223137319087982,
+      "learning_rate": 7.94764526589491e-06,
+      "loss": 0.3455,
+      "step": 1316
+    },
+    {
+      "epoch": 1.109208309938237,
+      "grad_norm": 0.2447323054075241,
+      "learning_rate": 7.943683779985412e-06,
+      "loss": 0.3364,
+      "step": 1317
+    },
+    {
+      "epoch": 1.1100505334081976,
+      "grad_norm": 0.22005987167358398,
+      "learning_rate": 7.939719463949398e-06,
+      "loss": 0.341,
+      "step": 1318
+    },
+    {
+      "epoch": 1.1108927568781584,
+      "grad_norm": 0.2396015226840973,
+      "learning_rate": 7.93575232159825e-06,
+      "loss": 0.3537,
+      "step": 1319
+    },
+    {
+      "epoch": 1.1117349803481191,
+      "grad_norm": 0.2083560973405838,
+      "learning_rate": 7.931782356746076e-06,
+      "loss": 0.3618,
+      "step": 1320
+    },
+    {
+      "epoch": 1.1125772038180797,
+      "grad_norm": 0.24214932322502136,
+      "learning_rate": 7.927809573209691e-06,
+      "loss": 0.3566,
+      "step": 1321
+    },
+    {
+      "epoch": 1.1134194272880404,
+      "grad_norm": 0.22593237459659576,
+      "learning_rate": 7.923833974808622e-06,
+      "loss": 0.3349,
+      "step": 1322
+    },
+    {
+      "epoch": 1.1142616507580012,
+      "grad_norm": 0.22273516654968262,
+      "learning_rate": 7.919855565365102e-06,
+      "loss": 0.39,
+      "step": 1323
+    },
+    {
+      "epoch": 1.1151038742279618,
+      "grad_norm": 0.2301865667104721,
+      "learning_rate": 7.91587434870407e-06,
+      "loss": 0.3674,
+      "step": 1324
+    },
+    {
+      "epoch": 1.1159460976979225,
+      "grad_norm": 0.251726359128952,
+      "learning_rate": 7.911890328653156e-06,
+      "loss": 0.3483,
+      "step": 1325
+    },
+    {
+      "epoch": 1.1167883211678833,
+      "grad_norm": 0.21106411516666412,
+      "learning_rate": 7.907903509042696e-06,
+      "loss": 0.3813,
+      "step": 1326
+    },
+    {
+      "epoch": 1.1176305446378438,
+      "grad_norm": 0.20069539546966553,
+      "learning_rate": 7.903913893705706e-06,
+      "loss": 0.3464,
+      "step": 1327
+    },
+    {
+      "epoch": 1.1184727681078046,
+      "grad_norm": 0.2140786051750183,
+      "learning_rate": 7.899921486477899e-06,
+      "loss": 0.3796,
+      "step": 1328
+    },
+    {
+      "epoch": 1.1193149915777654,
+      "grad_norm": 0.21972276270389557,
+      "learning_rate": 7.895926291197667e-06,
+      "loss": 0.3763,
+      "step": 1329
+    },
+    {
+      "epoch": 1.120157215047726,
+      "grad_norm": 0.19542910158634186,
+      "learning_rate": 7.891928311706088e-06,
+      "loss": 0.3485,
+      "step": 1330
+    },
+    {
+      "epoch": 1.1209994385176867,
+      "grad_norm": 0.2048928141593933,
+      "learning_rate": 7.887927551846908e-06,
+      "loss": 0.3583,
+      "step": 1331
+    },
+    {
+      "epoch": 1.1218416619876475,
+      "grad_norm": 0.203351229429245,
+      "learning_rate": 7.883924015466554e-06,
+      "loss": 0.3453,
+      "step": 1332
+    },
+    {
+      "epoch": 1.122683885457608,
+      "grad_norm": 0.23957213759422302,
+      "learning_rate": 7.87991770641412e-06,
+      "loss": 0.3803,
+      "step": 1333
+    },
+    {
+      "epoch": 1.1235261089275688,
+      "grad_norm": 0.2266377955675125,
+      "learning_rate": 7.875908628541363e-06,
+      "loss": 0.3803,
+      "step": 1334
+    },
+    {
+      "epoch": 1.1243683323975295,
+      "grad_norm": 0.21484608948230743,
+      "learning_rate": 7.871896785702707e-06,
+      "loss": 0.3659,
+      "step": 1335
+    },
+    {
+      "epoch": 1.12521055586749,
+      "grad_norm": 0.21197688579559326,
+      "learning_rate": 7.86788218175523e-06,
+      "loss": 0.3423,
+      "step": 1336
+    },
+    {
+      "epoch": 1.1260527793374508,
+      "grad_norm": 0.2307911217212677,
+      "learning_rate": 7.863864820558669e-06,
+      "loss": 0.3362,
+      "step": 1337
+    },
+    {
+      "epoch": 1.1268950028074116,
+      "grad_norm": 0.22809703648090363,
+      "learning_rate": 7.859844705975405e-06,
+      "loss": 0.4024,
+      "step": 1338
+    },
+    {
+      "epoch": 1.1277372262773722,
+      "grad_norm": 0.2245447188615799,
+      "learning_rate": 7.855821841870472e-06,
+      "loss": 0.355,
+      "step": 1339
+    },
+    {
+      "epoch": 1.128579449747333,
+      "grad_norm": 0.21683157980442047,
+      "learning_rate": 7.851796232111546e-06,
+      "loss": 0.3697,
+      "step": 1340
+    },
+    {
+      "epoch": 1.1294216732172937,
+      "grad_norm": 0.22042052447795868,
+      "learning_rate": 7.847767880568944e-06,
+      "loss": 0.3597,
+      "step": 1341
+    },
+    {
+      "epoch": 1.1302638966872545,
+      "grad_norm": 0.2326141595840454,
+      "learning_rate": 7.843736791115614e-06,
+      "loss": 0.3603,
+      "step": 1342
+    },
+    {
+      "epoch": 1.131106120157215,
+      "grad_norm": 0.2275954782962799,
+      "learning_rate": 7.839702967627145e-06,
+      "loss": 0.3732,
+      "step": 1343
+    },
+    {
+      "epoch": 1.1319483436271758,
+      "grad_norm": 0.21167173981666565,
+      "learning_rate": 7.835666413981744e-06,
+      "loss": 0.3263,
+      "step": 1344
+    },
+    {
+      "epoch": 1.1327905670971363,
+      "grad_norm": 0.22949613630771637,
+      "learning_rate": 7.831627134060249e-06,
+      "loss": 0.3584,
+      "step": 1345
+    },
+    {
+      "epoch": 1.133632790567097,
+      "grad_norm": 0.24151138961315155,
+      "learning_rate": 7.827585131746122e-06,
+      "loss": 0.3904,
+      "step": 1346
+    },
+    {
+      "epoch": 1.1344750140370579,
+      "grad_norm": 0.23588694632053375,
+      "learning_rate": 7.823540410925434e-06,
+      "loss": 0.3839,
+      "step": 1347
+    },
+    {
+      "epoch": 1.1353172375070186,
+      "grad_norm": 0.2144988477230072,
+      "learning_rate": 7.81949297548688e-06,
+      "loss": 0.3653,
+      "step": 1348
+    },
+    {
+      "epoch": 1.1361594609769792,
+      "grad_norm": 0.23595893383026123,
+      "learning_rate": 7.815442829321754e-06,
+      "loss": 0.3538,
+      "step": 1349
+    },
+    {
+      "epoch": 1.13700168444694,
+      "grad_norm": 0.26966625452041626,
+      "learning_rate": 7.811389976323963e-06,
+      "loss": 0.4052,
+      "step": 1350
+    },
+    {
+      "epoch": 1.1378439079169007,
+      "grad_norm": 0.21856364607810974,
+      "learning_rate": 7.807334420390014e-06,
+      "loss": 0.3603,
+      "step": 1351
+    },
+    {
+      "epoch": 1.1386861313868613,
+      "grad_norm": 0.22447790205478668,
+      "learning_rate": 7.803276165419015e-06,
+      "loss": 0.3687,
+      "step": 1352
+    },
+    {
+      "epoch": 1.139528354856822,
+      "grad_norm": 0.2570057809352875,
+      "learning_rate": 7.799215215312667e-06,
+      "loss": 0.387,
+      "step": 1353
+    },
+    {
+      "epoch": 1.1403705783267828,
+      "grad_norm": 0.22275447845458984,
+      "learning_rate": 7.795151573975262e-06,
+      "loss": 0.345,
+      "step": 1354
+    },
+    {
+      "epoch": 1.1412128017967433,
+      "grad_norm": 0.23103633522987366,
+      "learning_rate": 7.79108524531368e-06,
+      "loss": 0.3427,
+      "step": 1355
+    },
+    {
+      "epoch": 1.142055025266704,
+      "grad_norm": 0.24295048415660858,
+      "learning_rate": 7.787016233237387e-06,
+      "loss": 0.3451,
+      "step": 1356
+    },
+    {
+      "epoch": 1.1428972487366649,
+      "grad_norm": 0.2352520227432251,
+      "learning_rate": 7.782944541658423e-06,
+      "loss": 0.3529,
+      "step": 1357
+    },
+    {
+      "epoch": 1.1437394722066254,
+      "grad_norm": 0.24101249873638153,
+      "learning_rate": 7.778870174491408e-06,
+      "loss": 0.362,
+      "step": 1358
+    },
+    {
+      "epoch": 1.1445816956765862,
+      "grad_norm": 0.22046388685703278,
+      "learning_rate": 7.774793135653537e-06,
+      "loss": 0.353,
+      "step": 1359
+    },
+    {
+      "epoch": 1.145423919146547,
+      "grad_norm": 0.220865398645401,
+      "learning_rate": 7.770713429064567e-06,
+      "loss": 0.3512,
+      "step": 1360
+    },
+    {
+      "epoch": 1.1462661426165075,
+      "grad_norm": 0.24877040088176727,
+      "learning_rate": 7.766631058646826e-06,
+      "loss": 0.3671,
+      "step": 1361
+    },
+    {
+      "epoch": 1.1471083660864683,
+      "grad_norm": 0.2663626968860626,
+      "learning_rate": 7.7625460283252e-06,
+      "loss": 0.3741,
+      "step": 1362
+    },
+    {
+      "epoch": 1.147950589556429,
+      "grad_norm": 0.21680448949337006,
+      "learning_rate": 7.75845834202713e-06,
+      "loss": 0.3703,
+      "step": 1363
+    },
+    {
+      "epoch": 1.1487928130263896,
+      "grad_norm": 0.1968405395746231,
+      "learning_rate": 7.754368003682617e-06,
+      "loss": 0.316,
+      "step": 1364
+    },
+    {
+      "epoch": 1.1496350364963503,
+      "grad_norm": 0.23480704426765442,
+      "learning_rate": 7.750275017224208e-06,
+      "loss": 0.3912,
+      "step": 1365
+    },
+    {
+      "epoch": 1.1504772599663111,
+      "grad_norm": 0.23217934370040894,
+      "learning_rate": 7.746179386586994e-06,
+      "loss": 0.3454,
+      "step": 1366
+    },
+    {
+      "epoch": 1.1513194834362717,
+      "grad_norm": 0.2343100905418396,
+      "learning_rate": 7.74208111570861e-06,
+      "loss": 0.3808,
+      "step": 1367
+    },
+    {
+      "epoch": 1.1521617069062324,
+      "grad_norm": 0.19447638094425201,
+      "learning_rate": 7.737980208529232e-06,
+      "loss": 0.3704,
+      "step": 1368
+    },
+    {
+      "epoch": 1.1530039303761932,
+      "grad_norm": 0.22011230885982513,
+      "learning_rate": 7.733876668991565e-06,
+      "loss": 0.3608,
+      "step": 1369
+    },
+    {
+      "epoch": 1.1538461538461537,
+      "grad_norm": 0.2208113670349121,
+      "learning_rate": 7.72977050104085e-06,
+      "loss": 0.3579,
+      "step": 1370
+    },
+    {
+      "epoch": 1.1546883773161145,
+      "grad_norm": 0.2177201509475708,
+      "learning_rate": 7.725661708624855e-06,
+      "loss": 0.3611,
+      "step": 1371
+    },
+    {
+      "epoch": 1.1555306007860753,
+      "grad_norm": 0.21167387068271637,
+      "learning_rate": 7.721550295693865e-06,
+      "loss": 0.3815,
+      "step": 1372
+    },
+    {
+      "epoch": 1.156372824256036,
+      "grad_norm": 0.22142820060253143,
+      "learning_rate": 7.71743626620069e-06,
+      "loss": 0.3751,
+      "step": 1373
+    },
+    {
+      "epoch": 1.1572150477259966,
+      "grad_norm": 0.25120094418525696,
+      "learning_rate": 7.713319624100657e-06,
+      "loss": 0.3277,
+      "step": 1374
+    },
+    {
+      "epoch": 1.1580572711959574,
+      "grad_norm": 0.23623235523700714,
+      "learning_rate": 7.7092003733516e-06,
+      "loss": 0.3389,
+      "step": 1375
+    },
+    {
+      "epoch": 1.158899494665918,
+      "grad_norm": 0.2505205571651459,
+      "learning_rate": 7.705078517913862e-06,
+      "loss": 0.3903,
+      "step": 1376
+    },
+    {
+      "epoch": 1.1597417181358787,
+      "grad_norm": 0.2453787922859192,
+      "learning_rate": 7.700954061750295e-06,
+      "loss": 0.3693,
+      "step": 1377
+    },
+    {
+      "epoch": 1.1605839416058394,
+      "grad_norm": 0.23586124181747437,
+      "learning_rate": 7.696827008826242e-06,
+      "loss": 0.3852,
+      "step": 1378
+    },
+    {
+      "epoch": 1.1614261650758002,
+      "grad_norm": 0.25180327892303467,
+      "learning_rate": 7.692697363109553e-06,
+      "loss": 0.3902,
+      "step": 1379
+    },
+    {
+      "epoch": 1.1622683885457608,
+      "grad_norm": 0.22355681657791138,
+      "learning_rate": 7.688565128570564e-06,
+      "loss": 0.3558,
+      "step": 1380
+    },
+    {
+      "epoch": 1.1631106120157215,
+      "grad_norm": 0.2151123583316803,
+      "learning_rate": 7.684430309182106e-06,
+      "loss": 0.3371,
+      "step": 1381
+    },
+    {
+      "epoch": 1.1639528354856823,
+      "grad_norm": 0.22661295533180237,
+      "learning_rate": 7.680292908919485e-06,
+      "loss": 0.3904,
+      "step": 1382
+    },
+    {
+      "epoch": 1.1647950589556428,
+      "grad_norm": 0.2297997623682022,
+      "learning_rate": 7.676152931760496e-06,
+      "loss": 0.3482,
+      "step": 1383
+    },
+    {
+      "epoch": 1.1656372824256036,
+      "grad_norm": 0.238944873213768,
+      "learning_rate": 7.672010381685416e-06,
+      "loss": 0.3758,
+      "step": 1384
+    },
+    {
+      "epoch": 1.1664795058955644,
+      "grad_norm": 0.2109581083059311,
+      "learning_rate": 7.667865262676981e-06,
+      "loss": 0.3397,
+      "step": 1385
+    },
+    {
+      "epoch": 1.167321729365525,
+      "grad_norm": 0.23245356976985931,
+      "learning_rate": 7.663717578720412e-06,
+      "loss": 0.377,
+      "step": 1386
+    },
+    {
+      "epoch": 1.1681639528354857,
+      "grad_norm": 0.2091844230890274,
+      "learning_rate": 7.659567333803386e-06,
+      "loss": 0.3256,
+      "step": 1387
+    },
+    {
+      "epoch": 1.1690061763054465,
+      "grad_norm": 0.2057802677154541,
+      "learning_rate": 7.655414531916048e-06,
+      "loss": 0.3674,
+      "step": 1388
+    },
+    {
+      "epoch": 1.169848399775407,
+      "grad_norm": 0.2157420814037323,
+      "learning_rate": 7.651259177050996e-06,
+      "loss": 0.3274,
+      "step": 1389
+    },
+    {
+      "epoch": 1.1706906232453678,
+      "grad_norm": 0.21391423046588898,
+      "learning_rate": 7.647101273203289e-06,
+      "loss": 0.3504,
+      "step": 1390
+    },
+    {
+      "epoch": 1.1715328467153285,
+      "grad_norm": 0.2303875833749771,
+      "learning_rate": 7.642940824370429e-06,
+      "loss": 0.3423,
+      "step": 1391
+    },
+    {
+      "epoch": 1.172375070185289,
+      "grad_norm": 0.26220864057540894,
+      "learning_rate": 7.638777834552372e-06,
+      "loss": 0.3806,
+      "step": 1392
+    },
+    {
+      "epoch": 1.1732172936552498,
+      "grad_norm": 0.20384414494037628,
+      "learning_rate": 7.634612307751513e-06,
+      "loss": 0.3344,
+      "step": 1393
+    },
+    {
+      "epoch": 1.1740595171252106,
+      "grad_norm": 0.22354431450366974,
+      "learning_rate": 7.630444247972688e-06,
+      "loss": 0.3843,
+      "step": 1394
+    },
+    {
+      "epoch": 1.1749017405951712,
+      "grad_norm": 0.21157680451869965,
+      "learning_rate": 7.626273659223166e-06,
+      "loss": 0.3636,
+      "step": 1395
+    },
+    {
+      "epoch": 1.175743964065132,
+      "grad_norm": 0.2282872051000595,
+      "learning_rate": 7.622100545512648e-06,
+      "loss": 0.3841,
+      "step": 1396
+    },
+    {
+      "epoch": 1.1765861875350927,
+      "grad_norm": 0.22645720839500427,
+      "learning_rate": 7.617924910853266e-06,
+      "loss": 0.3808,
+      "step": 1397
+    },
+    {
+      "epoch": 1.1774284110050532,
+      "grad_norm": 0.21912354230880737,
+      "learning_rate": 7.61374675925957e-06,
+      "loss": 0.3486,
+      "step": 1398
+    },
+    {
+      "epoch": 1.178270634475014,
+      "grad_norm": 0.2137390673160553,
+      "learning_rate": 7.609566094748535e-06,
+      "loss": 0.36,
+      "step": 1399
+    },
+    {
+      "epoch": 1.1791128579449748,
+      "grad_norm": 0.23046202957630157,
+      "learning_rate": 7.605382921339548e-06,
+      "loss": 0.3637,
+      "step": 1400
+    },
+    {
+      "epoch": 1.1799550814149353,
+      "grad_norm": 0.24850760400295258,
+      "learning_rate": 7.601197243054411e-06,
+      "loss": 0.3757,
+      "step": 1401
+    },
+    {
+      "epoch": 1.180797304884896,
+      "grad_norm": 0.20614586770534515,
+      "learning_rate": 7.597009063917333e-06,
+      "loss": 0.3346,
+      "step": 1402
+    },
+    {
+      "epoch": 1.1816395283548569,
+      "grad_norm": 0.2218198925256729,
+      "learning_rate": 7.5928183879549274e-06,
+      "loss": 0.3858,
+      "step": 1403
+    },
+    {
+      "epoch": 1.1824817518248176,
+      "grad_norm": 0.21687179803848267,
+      "learning_rate": 7.588625219196208e-06,
+      "loss": 0.3879,
+      "step": 1404
+    },
+    {
+      "epoch": 1.1833239752947782,
+      "grad_norm": 0.2420112043619156,
+      "learning_rate": 7.584429561672586e-06,
+      "loss": 0.3586,
+      "step": 1405
+    },
+    {
+      "epoch": 1.184166198764739,
+      "grad_norm": 0.25380006432533264,
+      "learning_rate": 7.580231419417863e-06,
+      "loss": 0.3845,
+      "step": 1406
+    },
+    {
+      "epoch": 1.1850084222346995,
+      "grad_norm": 0.21595925092697144,
+      "learning_rate": 7.576030796468233e-06,
+      "loss": 0.3644,
+      "step": 1407
+    },
+    {
+      "epoch": 1.1858506457046603,
+      "grad_norm": 0.2313537299633026,
+      "learning_rate": 7.571827696862274e-06,
+      "loss": 0.3331,
+      "step": 1408
+    },
+    {
+      "epoch": 1.186692869174621,
+      "grad_norm": 0.2036680281162262,
+      "learning_rate": 7.567622124640942e-06,
+      "loss": 0.3618,
+      "step": 1409
+    },
+    {
+      "epoch": 1.1875350926445818,
+      "grad_norm": 0.2853444814682007,
+      "learning_rate": 7.563414083847573e-06,
+      "loss": 0.4122,
+      "step": 1410
+    },
+    {
+      "epoch": 1.1883773161145423,
+      "grad_norm": 0.23605458438396454,
+      "learning_rate": 7.55920357852788e-06,
+      "loss": 0.334,
+      "step": 1411
+    },
+    {
+      "epoch": 1.189219539584503,
+      "grad_norm": 0.20309481024742126,
+      "learning_rate": 7.554990612729936e-06,
+      "loss": 0.3682,
+      "step": 1412
+    },
+    {
+      "epoch": 1.1900617630544639,
+      "grad_norm": 0.21416065096855164,
+      "learning_rate": 7.5507751905041885e-06,
+      "loss": 0.3548,
+      "step": 1413
+    },
+    {
+      "epoch": 1.1909039865244244,
+      "grad_norm": 0.2469504326581955,
+      "learning_rate": 7.5465573159034396e-06,
+      "loss": 0.3669,
+      "step": 1414
+    },
+    {
+      "epoch": 1.1917462099943852,
+      "grad_norm": 0.23954510688781738,
+      "learning_rate": 7.542336992982857e-06,
+      "loss": 0.356,
+      "step": 1415
+    },
+    {
+      "epoch": 1.192588433464346,
+      "grad_norm": 0.2099873125553131,
+      "learning_rate": 7.538114225799955e-06,
+      "loss": 0.376,
+      "step": 1416
+    },
+    {
+      "epoch": 1.1934306569343065,
+      "grad_norm": 0.21912400424480438,
+      "learning_rate": 7.533889018414602e-06,
+      "loss": 0.387,
+      "step": 1417
+    },
+    {
+      "epoch": 1.1942728804042673,
+      "grad_norm": 0.21182800829410553,
+      "learning_rate": 7.529661374889011e-06,
+      "loss": 0.3244,
+      "step": 1418
+    },
+    {
+      "epoch": 1.195115103874228,
+      "grad_norm": 0.2408171445131302,
+      "learning_rate": 7.525431299287737e-06,
+      "loss": 0.367,
+      "step": 1419
+    },
+    {
+      "epoch": 1.1959573273441886,
+      "grad_norm": 0.2102893888950348,
+      "learning_rate": 7.5211987956776755e-06,
+      "loss": 0.3645,
+      "step": 1420
+    },
+    {
+      "epoch": 1.1967995508141493,
+      "grad_norm": 0.25745439529418945,
+      "learning_rate": 7.516963868128054e-06,
+      "loss": 0.3827,
+      "step": 1421
+    },
+    {
+      "epoch": 1.1976417742841101,
+      "grad_norm": 0.22646570205688477,
+      "learning_rate": 7.512726520710429e-06,
+      "loss": 0.3589,
+      "step": 1422
+    },
+    {
+      "epoch": 1.1984839977540707,
+      "grad_norm": 0.2319384515285492,
+      "learning_rate": 7.508486757498687e-06,
+      "loss": 0.3483,
+      "step": 1423
+    },
+    {
+      "epoch": 1.1993262212240314,
+      "grad_norm": 0.2112833559513092,
+      "learning_rate": 7.5042445825690344e-06,
+      "loss": 0.3366,
+      "step": 1424
+    },
+    {
+      "epoch": 1.2001684446939922,
+      "grad_norm": 0.2383369654417038,
+      "learning_rate": 7.500000000000001e-06,
+      "loss": 0.3742,
+      "step": 1425
+    },
+    {
+      "epoch": 1.2010106681639527,
+      "grad_norm": 0.22696097195148468,
+      "learning_rate": 7.4957530138724245e-06,
+      "loss": 0.3443,
+      "step": 1426
+    },
+    {
+      "epoch": 1.2018528916339135,
+      "grad_norm": 0.21109351515769958,
+      "learning_rate": 7.491503628269458e-06,
+      "loss": 0.3698,
+      "step": 1427
+    },
+    {
+      "epoch": 1.2026951151038743,
+      "grad_norm": 0.24133235216140747,
+      "learning_rate": 7.4872518472765594e-06,
+      "loss": 0.4057,
+      "step": 1428
+    },
+    {
+      "epoch": 1.203537338573835,
+      "grad_norm": 0.21957817673683167,
+      "learning_rate": 7.4829976749814935e-06,
+      "loss": 0.3022,
+      "step": 1429
+    },
+    {
+      "epoch": 1.2043795620437956,
+      "grad_norm": 0.23803481459617615,
+      "learning_rate": 7.4787411154743175e-06,
+      "loss": 0.3377,
+      "step": 1430
+    },
+    {
+      "epoch": 1.2052217855137564,
+      "grad_norm": 0.23232616484165192,
+      "learning_rate": 7.474482172847391e-06,
+      "loss": 0.3841,
+      "step": 1431
+    },
+    {
+      "epoch": 1.206064008983717,
+      "grad_norm": 0.23307104408740997,
+      "learning_rate": 7.470220851195356e-06,
+      "loss": 0.3665,
+      "step": 1432
+    },
+    {
+      "epoch": 1.2069062324536777,
+      "grad_norm": 0.2201298177242279,
+      "learning_rate": 7.46595715461515e-06,
+      "loss": 0.3665,
+      "step": 1433
+    },
+    {
+      "epoch": 1.2077484559236384,
+      "grad_norm": 0.19988930225372314,
+      "learning_rate": 7.461691087205993e-06,
+      "loss": 0.346,
+      "step": 1434
+    },
+    {
+      "epoch": 1.2085906793935992,
+      "grad_norm": 0.2632721960544586,
+      "learning_rate": 7.457422653069379e-06,
+      "loss": 0.3814,
+      "step": 1435
+    },
+    {
+      "epoch": 1.2094329028635598,
+      "grad_norm": 0.2116016000509262,
+      "learning_rate": 7.45315185630908e-06,
+      "loss": 0.355,
+      "step": 1436
+    },
+    {
+      "epoch": 1.2102751263335205,
+      "grad_norm": 0.20903754234313965,
+      "learning_rate": 7.4488787010311425e-06,
+      "loss": 0.3569,
+      "step": 1437
+    },
+    {
+      "epoch": 1.211117349803481,
+      "grad_norm": 0.23268146812915802,
+      "learning_rate": 7.444603191343878e-06,
+      "loss": 0.3403,
+      "step": 1438
+    },
+    {
+      "epoch": 1.2119595732734418,
+      "grad_norm": 0.23514798283576965,
+      "learning_rate": 7.440325331357858e-06,
+      "loss": 0.3461,
+      "step": 1439
+    },
+    {
+      "epoch": 1.2128017967434026,
+      "grad_norm": 0.22160202264785767,
+      "learning_rate": 7.436045125185923e-06,
+      "loss": 0.375,
+      "step": 1440
+    },
+    {
+      "epoch": 1.2136440202133634,
+      "grad_norm": 0.20510455965995789,
+      "learning_rate": 7.431762576943157e-06,
+      "loss": 0.3538,
+      "step": 1441
+    },
+    {
+      "epoch": 1.214486243683324,
+      "grad_norm": 0.2111036628484726,
+      "learning_rate": 7.427477690746906e-06,
+      "loss": 0.3451,
+      "step": 1442
+    },
+    {
+      "epoch": 1.2153284671532847,
+      "grad_norm": 0.23642785847187042,
+      "learning_rate": 7.423190470716761e-06,
+      "loss": 0.3427,
+      "step": 1443
+    },
+    {
+      "epoch": 1.2161706906232455,
+      "grad_norm": 0.24378204345703125,
+      "learning_rate": 7.418900920974552e-06,
+      "loss": 0.3692,
+      "step": 1444
+    },
+    {
+      "epoch": 1.217012914093206,
+      "grad_norm": 0.23461374640464783,
+      "learning_rate": 7.414609045644356e-06,
+      "loss": 0.3889,
+      "step": 1445
+    },
+    {
+      "epoch": 1.2178551375631668,
+      "grad_norm": 0.22334659099578857,
+      "learning_rate": 7.4103148488524824e-06,
+      "loss": 0.349,
+      "step": 1446
+    },
+    {
+      "epoch": 1.2186973610331275,
+      "grad_norm": 0.2602890729904175,
+      "learning_rate": 7.40601833472747e-06,
+      "loss": 0.3786,
+      "step": 1447
+    },
+    {
+      "epoch": 1.219539584503088,
+      "grad_norm": 0.20589013397693634,
+      "learning_rate": 7.401719507400088e-06,
+      "loss": 0.3318,
+      "step": 1448
+    },
+    {
+      "epoch": 1.2203818079730488,
+      "grad_norm": 0.24295538663864136,
+      "learning_rate": 7.3974183710033334e-06,
+      "loss": 0.3628,
+      "step": 1449
+    },
+    {
+      "epoch": 1.2212240314430096,
+      "grad_norm": 0.2678720951080322,
+      "learning_rate": 7.393114929672414e-06,
+      "loss": 0.3951,
+      "step": 1450
+    },
+    {
+      "epoch": 1.2220662549129702,
+      "grad_norm": 0.22849512100219727,
+      "learning_rate": 7.388809187544764e-06,
+      "loss": 0.3606,
+      "step": 1451
+    },
+    {
+      "epoch": 1.222908478382931,
+      "grad_norm": 0.22386722266674042,
+      "learning_rate": 7.384501148760024e-06,
+      "loss": 0.3588,
+      "step": 1452
+    },
+    {
+      "epoch": 1.2237507018528917,
+      "grad_norm": 0.1996832638978958,
+      "learning_rate": 7.38019081746004e-06,
+      "loss": 0.3541,
+      "step": 1453
+    },
+    {
+      "epoch": 1.2245929253228522,
+      "grad_norm": 0.2293432205915451,
+      "learning_rate": 7.3758781977888684e-06,
+      "loss": 0.3545,
+      "step": 1454
+    },
+    {
+      "epoch": 1.225435148792813,
+      "grad_norm": 0.21567665040493011,
+      "learning_rate": 7.371563293892761e-06,
+      "loss": 0.3423,
+      "step": 1455
+    },
+    {
+      "epoch": 1.2262773722627738,
+      "grad_norm": 0.22745153307914734,
+      "learning_rate": 7.367246109920171e-06,
+      "loss": 0.3332,
+      "step": 1456
+    },
+    {
+      "epoch": 1.2271195957327343,
+      "grad_norm": 0.20745089650154114,
+      "learning_rate": 7.362926650021736e-06,
+      "loss": 0.3856,
+      "step": 1457
+    },
+    {
+      "epoch": 1.227961819202695,
+      "grad_norm": 0.24365340173244476,
+      "learning_rate": 7.3586049183502875e-06,
+      "loss": 0.391,
+      "step": 1458
+    },
+    {
+      "epoch": 1.2288040426726559,
+      "grad_norm": 0.21581348776817322,
+      "learning_rate": 7.354280919060839e-06,
+      "loss": 0.36,
+      "step": 1459
+    },
+    {
+      "epoch": 1.2296462661426166,
+      "grad_norm": 0.20025712251663208,
+      "learning_rate": 7.349954656310585e-06,
+      "loss": 0.3253,
+      "step": 1460
+    },
+    {
+      "epoch": 1.2304884896125772,
+      "grad_norm": 0.25201642513275146,
+      "learning_rate": 7.345626134258897e-06,
+      "loss": 0.3997,
+      "step": 1461
+    },
+    {
+      "epoch": 1.231330713082538,
+      "grad_norm": 0.23212586343288422,
+      "learning_rate": 7.341295357067315e-06,
+      "loss": 0.4041,
+      "step": 1462
+    },
+    {
+      "epoch": 1.2321729365524985,
+      "grad_norm": 0.20653077960014343,
+      "learning_rate": 7.336962328899553e-06,
+      "loss": 0.3153,
+      "step": 1463
+    },
+    {
+      "epoch": 1.2330151600224593,
+      "grad_norm": 0.24971120059490204,
+      "learning_rate": 7.3326270539214826e-06,
+      "loss": 0.3813,
+      "step": 1464
+    },
+    {
+      "epoch": 1.23385738349242,
+      "grad_norm": 0.2304530292749405,
+      "learning_rate": 7.3282895363011405e-06,
+      "loss": 0.3431,
+      "step": 1465
+    },
+    {
+      "epoch": 1.2346996069623808,
+      "grad_norm": 0.2110995203256607,
+      "learning_rate": 7.323949780208717e-06,
+      "loss": 0.3634,
+      "step": 1466
+    },
+    {
+      "epoch": 1.2355418304323413,
+      "grad_norm": 0.2557746171951294,
+      "learning_rate": 7.319607789816555e-06,
+      "loss": 0.3821,
+      "step": 1467
+    },
+    {
+      "epoch": 1.236384053902302,
+      "grad_norm": 0.23002184927463531,
+      "learning_rate": 7.315263569299147e-06,
+      "loss": 0.3528,
+      "step": 1468
+    },
+    {
+      "epoch": 1.2372262773722627,
+      "grad_norm": 0.21381570398807526,
+      "learning_rate": 7.310917122833127e-06,
+      "loss": 0.4101,
+      "step": 1469
+    },
+    {
+      "epoch": 1.2380685008422234,
+      "grad_norm": 0.21428106725215912,
+      "learning_rate": 7.306568454597269e-06,
+      "loss": 0.3479,
+      "step": 1470
+    },
+    {
+      "epoch": 1.2389107243121842,
+      "grad_norm": 0.25851118564605713,
+      "learning_rate": 7.302217568772488e-06,
+      "loss": 0.3517,
+      "step": 1471
+    },
+    {
+      "epoch": 1.239752947782145,
+      "grad_norm": 0.25823184847831726,
+      "learning_rate": 7.297864469541826e-06,
+      "loss": 0.3613,
+      "step": 1472
+    },
+    {
+      "epoch": 1.2405951712521055,
+      "grad_norm": 0.20786811411380768,
+      "learning_rate": 7.293509161090453e-06,
+      "loss": 0.3739,
+      "step": 1473
+    },
+    {
+      "epoch": 1.2414373947220663,
+      "grad_norm": 0.2107485681772232,
+      "learning_rate": 7.289151647605668e-06,
+      "loss": 0.3242,
+      "step": 1474
+    },
+    {
+      "epoch": 1.242279618192027,
+      "grad_norm": 0.248404860496521,
+      "learning_rate": 7.284791933276883e-06,
+      "loss": 0.3697,
+      "step": 1475
+    },
+    {
+      "epoch": 1.2431218416619876,
+      "grad_norm": 0.21293829381465912,
+      "learning_rate": 7.28043002229563e-06,
+      "loss": 0.3594,
+      "step": 1476
+    },
+    {
+      "epoch": 1.2439640651319483,
+      "grad_norm": 0.22747156023979187,
+      "learning_rate": 7.276065918855554e-06,
+      "loss": 0.3614,
+      "step": 1477
+    },
+    {
+      "epoch": 1.2448062886019091,
+      "grad_norm": 0.20419429242610931,
+      "learning_rate": 7.271699627152406e-06,
+      "loss": 0.3457,
+      "step": 1478
+    },
+    {
+      "epoch": 1.2456485120718697,
+      "grad_norm": 0.23443926870822906,
+      "learning_rate": 7.2673311513840395e-06,
+      "loss": 0.3576,
+      "step": 1479
+    },
+    {
+      "epoch": 1.2464907355418304,
+      "grad_norm": 0.25050753355026245,
+      "learning_rate": 7.26296049575041e-06,
+      "loss": 0.3875,
+      "step": 1480
+    },
+    {
+      "epoch": 1.2473329590117912,
+      "grad_norm": 0.23791559040546417,
+      "learning_rate": 7.2585876644535705e-06,
+      "loss": 0.3904,
+      "step": 1481
+    },
+    {
+      "epoch": 1.2481751824817517,
+      "grad_norm": 0.19877417385578156,
+      "learning_rate": 7.2542126616976596e-06,
+      "loss": 0.3321,
+      "step": 1482
+    },
+    {
+      "epoch": 1.2490174059517125,
+      "grad_norm": 0.21887525916099548,
+      "learning_rate": 7.24983549168891e-06,
+      "loss": 0.3517,
+      "step": 1483
+    },
+    {
+      "epoch": 1.2498596294216733,
+      "grad_norm": 0.2395276427268982,
+      "learning_rate": 7.2454561586356355e-06,
+      "loss": 0.3421,
+      "step": 1484
+    },
+    {
+      "epoch": 1.250701852891634,
+      "grad_norm": 0.21983036398887634,
+      "learning_rate": 7.241074666748228e-06,
+      "loss": 0.3496,
+      "step": 1485
+    },
+    {
+      "epoch": 1.2515440763615946,
+      "grad_norm": 0.20137713849544525,
+      "learning_rate": 7.236691020239157e-06,
+      "loss": 0.3631,
+      "step": 1486
+    },
+    {
+      "epoch": 1.2523862998315554,
+      "grad_norm": 0.22294463217258453,
+      "learning_rate": 7.232305223322963e-06,
+      "loss": 0.3653,
+      "step": 1487
+    },
+    {
+      "epoch": 1.253228523301516,
+      "grad_norm": 0.19250662624835968,
+      "learning_rate": 7.227917280216254e-06,
+      "loss": 0.33,
+      "step": 1488
+    },
+    {
+      "epoch": 1.2540707467714767,
+      "grad_norm": 0.2341872900724411,
+      "learning_rate": 7.2235271951377005e-06,
+      "loss": 0.3987,
+      "step": 1489
+    },
+    {
+      "epoch": 1.2549129702414374,
+      "grad_norm": 0.2020316869020462,
+      "learning_rate": 7.219134972308035e-06,
+      "loss": 0.3569,
+      "step": 1490
+    },
+    {
+      "epoch": 1.2557551937113982,
+      "grad_norm": 0.20235149562358856,
+      "learning_rate": 7.214740615950041e-06,
+      "loss": 0.3583,
+      "step": 1491
+    },
+    {
+      "epoch": 1.2565974171813588,
+      "grad_norm": 0.21482989192008972,
+      "learning_rate": 7.210344130288558e-06,
+      "loss": 0.3396,
+      "step": 1492
+    },
+    {
+      "epoch": 1.2574396406513195,
+      "grad_norm": 0.20860536396503448,
+      "learning_rate": 7.205945519550467e-06,
+      "loss": 0.3759,
+      "step": 1493
+    },
+    {
+      "epoch": 1.25828186412128,
+      "grad_norm": 0.20563741028308868,
+      "learning_rate": 7.201544787964698e-06,
+      "loss": 0.3644,
+      "step": 1494
+    },
+    {
+      "epoch": 1.2591240875912408,
+      "grad_norm": 0.21251443028450012,
+      "learning_rate": 7.197141939762217e-06,
+      "loss": 0.3238,
+      "step": 1495
+    },
+    {
+      "epoch": 1.2599663110612016,
+      "grad_norm": 0.24338217079639435,
+      "learning_rate": 7.192736979176025e-06,
+      "loss": 0.3891,
+      "step": 1496
+    },
+    {
+      "epoch": 1.2608085345311624,
+      "grad_norm": 0.22656427323818207,
+      "learning_rate": 7.188329910441154e-06,
+      "loss": 0.364,
+      "step": 1497
+    },
+    {
+      "epoch": 1.261650758001123,
+      "grad_norm": 0.19089765846729279,
+      "learning_rate": 7.183920737794663e-06,
+      "loss": 0.3476,
+      "step": 1498
+    },
+    {
+      "epoch": 1.2624929814710837,
+      "grad_norm": 0.2502041459083557,
+      "learning_rate": 7.179509465475636e-06,
+      "loss": 0.3563,
+      "step": 1499
+    },
+    {
+      "epoch": 1.2633352049410442,
+      "grad_norm": 0.22060425579547882,
+      "learning_rate": 7.175096097725169e-06,
+      "loss": 0.3802,
+      "step": 1500
+    },
+    {
+      "epoch": 1.264177428411005,
+      "grad_norm": 0.25501200556755066,
+      "learning_rate": 7.170680638786383e-06,
+      "loss": 0.3532,
+      "step": 1501
+    },
+    {
+      "epoch": 1.2650196518809658,
+      "grad_norm": 0.21981732547283173,
+      "learning_rate": 7.166263092904399e-06,
+      "loss": 0.3528,
+      "step": 1502
+    },
+    {
+      "epoch": 1.2658618753509265,
+      "grad_norm": 0.23550871014595032,
+      "learning_rate": 7.161843464326349e-06,
+      "loss": 0.3695,
+      "step": 1503
+    },
+    {
+      "epoch": 1.266704098820887,
+      "grad_norm": 0.24225080013275146,
+      "learning_rate": 7.157421757301371e-06,
+      "loss": 0.3578,
+      "step": 1504
+    },
+    {
+      "epoch": 1.2675463222908478,
+      "grad_norm": 0.20522600412368774,
+      "learning_rate": 7.1529979760805946e-06,
+      "loss": 0.3542,
+      "step": 1505
+    },
+    {
+      "epoch": 1.2683885457608086,
+      "grad_norm": 0.23005728423595428,
+      "learning_rate": 7.148572124917148e-06,
+      "loss": 0.3521,
+      "step": 1506
+    },
+    {
+      "epoch": 1.2692307692307692,
+      "grad_norm": 0.22665822505950928,
+      "learning_rate": 7.144144208066148e-06,
+      "loss": 0.3414,
+      "step": 1507
+    },
+    {
+      "epoch": 1.27007299270073,
+      "grad_norm": 0.21564829349517822,
+      "learning_rate": 7.1397142297846975e-06,
+      "loss": 0.3615,
+      "step": 1508
+    },
+    {
+      "epoch": 1.2709152161706907,
+      "grad_norm": 0.23631811141967773,
+      "learning_rate": 7.135282194331881e-06,
+      "loss": 0.3869,
+      "step": 1509
+    },
+    {
+      "epoch": 1.2717574396406512,
+      "grad_norm": 0.2472311556339264,
+      "learning_rate": 7.130848105968762e-06,
+      "loss": 0.3572,
+      "step": 1510
+    },
+    {
+      "epoch": 1.272599663110612,
+      "grad_norm": 0.20352278649806976,
+      "learning_rate": 7.126411968958374e-06,
+      "loss": 0.3525,
+      "step": 1511
+    },
+    {
+      "epoch": 1.2734418865805728,
+      "grad_norm": 0.24103017151355743,
+      "learning_rate": 7.121973787565727e-06,
+      "loss": 0.3856,
+      "step": 1512
+    },
+    {
+      "epoch": 1.2742841100505333,
+      "grad_norm": 0.2155137062072754,
+      "learning_rate": 7.1175335660577906e-06,
+      "loss": 0.3688,
+      "step": 1513
+    },
+    {
+      "epoch": 1.275126333520494,
+      "grad_norm": 0.22255980968475342,
+      "learning_rate": 7.113091308703498e-06,
+      "loss": 0.3707,
+      "step": 1514
+    },
+    {
+      "epoch": 1.2759685569904549,
+      "grad_norm": 0.22150452435016632,
+      "learning_rate": 7.1086470197737405e-06,
+      "loss": 0.3538,
+      "step": 1515
+    },
+    {
+      "epoch": 1.2768107804604156,
+      "grad_norm": 0.261150985956192,
+      "learning_rate": 7.104200703541358e-06,
+      "loss": 0.3684,
+      "step": 1516
+    },
+    {
+      "epoch": 1.2776530039303762,
+      "grad_norm": 0.2053704559803009,
+      "learning_rate": 7.099752364281147e-06,
+      "loss": 0.3627,
+      "step": 1517
+    },
+    {
+      "epoch": 1.278495227400337,
+      "grad_norm": 0.19474367797374725,
+      "learning_rate": 7.095302006269842e-06,
+      "loss": 0.3467,
+      "step": 1518
+    },
+    {
+      "epoch": 1.2793374508702975,
+      "grad_norm": 0.22994056344032288,
+      "learning_rate": 7.090849633786125e-06,
+      "loss": 0.3585,
+      "step": 1519
+    },
+    {
+      "epoch": 1.2801796743402583,
+      "grad_norm": 0.22903107106685638,
+      "learning_rate": 7.0863952511106075e-06,
+      "loss": 0.3829,
+      "step": 1520
+    },
+    {
+      "epoch": 1.281021897810219,
+      "grad_norm": 0.1935136318206787,
+      "learning_rate": 7.0819388625258385e-06,
+      "loss": 0.308,
+      "step": 1521
+    },
+    {
+      "epoch": 1.2818641212801798,
+      "grad_norm": 0.21748337149620056,
+      "learning_rate": 7.077480472316296e-06,
+      "loss": 0.4124,
+      "step": 1522
+    },
+    {
+      "epoch": 1.2827063447501403,
+      "grad_norm": 0.20140865445137024,
+      "learning_rate": 7.0730200847683795e-06,
+      "loss": 0.3328,
+      "step": 1523
+    },
+    {
+      "epoch": 1.283548568220101,
+      "grad_norm": 0.2253345400094986,
+      "learning_rate": 7.06855770417041e-06,
+      "loss": 0.3933,
+      "step": 1524
+    },
+    {
+      "epoch": 1.2843907916900617,
+      "grad_norm": 0.223196879029274,
+      "learning_rate": 7.0640933348126235e-06,
+      "loss": 0.358,
+      "step": 1525
+    },
+    {
+      "epoch": 1.2852330151600224,
+      "grad_norm": 0.20166324079036713,
+      "learning_rate": 7.059626980987172e-06,
+      "loss": 0.3446,
+      "step": 1526
+    },
+    {
+      "epoch": 1.2860752386299832,
+      "grad_norm": 0.230132058262825,
+      "learning_rate": 7.05515864698811e-06,
+      "loss": 0.3775,
+      "step": 1527
+    },
+    {
+      "epoch": 1.286917462099944,
+      "grad_norm": 0.22569893300533295,
+      "learning_rate": 7.0506883371114e-06,
+      "loss": 0.3603,
+      "step": 1528
+    },
+    {
+      "epoch": 1.2877596855699045,
+      "grad_norm": 0.2010277509689331,
+      "learning_rate": 7.046216055654902e-06,
+      "loss": 0.3565,
+      "step": 1529
+    },
+    {
+      "epoch": 1.2886019090398653,
+      "grad_norm": 0.22424009442329407,
+      "learning_rate": 7.041741806918372e-06,
+      "loss": 0.3822,
+      "step": 1530
+    },
+    {
+      "epoch": 1.2894441325098258,
+      "grad_norm": 0.20551267266273499,
+      "learning_rate": 7.0372655952034575e-06,
+      "loss": 0.3243,
+      "step": 1531
+    },
+    {
+      "epoch": 1.2902863559797866,
+      "grad_norm": 0.27740082144737244,
+      "learning_rate": 7.032787424813694e-06,
+      "loss": 0.3836,
+      "step": 1532
+    },
+    {
+      "epoch": 1.2911285794497473,
+      "grad_norm": 0.21728472411632538,
+      "learning_rate": 7.028307300054499e-06,
+      "loss": 0.3624,
+      "step": 1533
+    },
+    {
+      "epoch": 1.2919708029197081,
+      "grad_norm": 0.21663859486579895,
+      "learning_rate": 7.023825225233169e-06,
+      "loss": 0.3399,
+      "step": 1534
+    },
+    {
+      "epoch": 1.2928130263896687,
+      "grad_norm": 0.2152295559644699,
+      "learning_rate": 7.019341204658876e-06,
+      "loss": 0.3849,
+      "step": 1535
+    },
+    {
+      "epoch": 1.2936552498596294,
+      "grad_norm": 0.23276902735233307,
+      "learning_rate": 7.014855242642662e-06,
+      "loss": 0.3789,
+      "step": 1536
+    },
+    {
+      "epoch": 1.2944974733295902,
+      "grad_norm": 0.21636129915714264,
+      "learning_rate": 7.0103673434974375e-06,
+      "loss": 0.3439,
+      "step": 1537
+    },
+    {
+      "epoch": 1.2953396967995507,
+      "grad_norm": 0.23212984204292297,
+      "learning_rate": 7.0058775115379705e-06,
+      "loss": 0.3799,
+      "step": 1538
+    },
+    {
+      "epoch": 1.2961819202695115,
+      "grad_norm": 0.20423316955566406,
+      "learning_rate": 7.0013857510808934e-06,
+      "loss": 0.3313,
+      "step": 1539
+    },
+    {
+      "epoch": 1.2970241437394723,
+      "grad_norm": 0.20709416270256042,
+      "learning_rate": 6.99689206644469e-06,
+      "loss": 0.3255,
+      "step": 1540
+    },
+    {
+      "epoch": 1.2978663672094328,
+      "grad_norm": 0.21995380520820618,
+      "learning_rate": 6.992396461949693e-06,
+      "loss": 0.3736,
+      "step": 1541
+    },
+    {
+      "epoch": 1.2987085906793936,
+      "grad_norm": 0.21833443641662598,
+      "learning_rate": 6.987898941918082e-06,
+      "loss": 0.3641,
+      "step": 1542
+    },
+    {
+      "epoch": 1.2995508141493544,
+      "grad_norm": 0.2526569068431854,
+      "learning_rate": 6.9833995106738774e-06,
+      "loss": 0.3496,
+      "step": 1543
+    },
+    {
+      "epoch": 1.300393037619315,
+      "grad_norm": 0.23018354177474976,
+      "learning_rate": 6.978898172542939e-06,
+      "loss": 0.3682,
+      "step": 1544
+    },
+    {
+      "epoch": 1.3012352610892757,
+      "grad_norm": 0.2187051624059677,
+      "learning_rate": 6.974394931852957e-06,
+      "loss": 0.3324,
+      "step": 1545
+    },
+    {
+      "epoch": 1.3020774845592364,
+      "grad_norm": 0.23275905847549438,
+      "learning_rate": 6.969889792933454e-06,
+      "loss": 0.3668,
+      "step": 1546
+    },
+    {
+      "epoch": 1.3029197080291972,
+      "grad_norm": 0.2110632210969925,
+      "learning_rate": 6.965382760115775e-06,
+      "loss": 0.332,
+      "step": 1547
+    },
+    {
+      "epoch": 1.3037619314991578,
+      "grad_norm": 0.231571763753891,
+      "learning_rate": 6.960873837733089e-06,
+      "loss": 0.3932,
+      "step": 1548
+    },
+    {
+      "epoch": 1.3046041549691185,
+      "grad_norm": 0.21139703691005707,
+      "learning_rate": 6.956363030120377e-06,
+      "loss": 0.3755,
+      "step": 1549
+    },
+    {
+      "epoch": 1.305446378439079,
+      "grad_norm": 0.20921073853969574,
+      "learning_rate": 6.951850341614436e-06,
+      "loss": 0.3632,
+      "step": 1550
+    },
+    {
+      "epoch": 1.3062886019090398,
+      "grad_norm": 0.23165875673294067,
+      "learning_rate": 6.94733577655387e-06,
+      "loss": 0.362,
+      "step": 1551
+    },
+    {
+      "epoch": 1.3071308253790006,
+      "grad_norm": 0.23026436567306519,
+      "learning_rate": 6.942819339279089e-06,
+      "loss": 0.3451,
+      "step": 1552
+    },
+    {
+      "epoch": 1.3079730488489614,
+      "grad_norm": 0.21386614441871643,
+      "learning_rate": 6.9383010341323e-06,
+      "loss": 0.3528,
+      "step": 1553
+    },
+    {
+      "epoch": 1.308815272318922,
+      "grad_norm": 0.2159416675567627,
+      "learning_rate": 6.933780865457508e-06,
+      "loss": 0.3315,
+      "step": 1554
+    },
+    {
+      "epoch": 1.3096574957888827,
+      "grad_norm": 0.21278788149356842,
+      "learning_rate": 6.9292588376005095e-06,
+      "loss": 0.3949,
+      "step": 1555
+    },
+    {
+      "epoch": 1.3104997192588432,
+      "grad_norm": 0.20024999976158142,
+      "learning_rate": 6.924734954908887e-06,
+      "loss": 0.3288,
+      "step": 1556
+    },
+    {
+      "epoch": 1.311341942728804,
+      "grad_norm": 0.24197152256965637,
+      "learning_rate": 6.920209221732007e-06,
+      "loss": 0.4007,
+      "step": 1557
+    },
+    {
+      "epoch": 1.3121841661987648,
+      "grad_norm": 0.19979821145534515,
+      "learning_rate": 6.9156816424210175e-06,
+      "loss": 0.3112,
+      "step": 1558
+    },
+    {
+      "epoch": 1.3130263896687255,
+      "grad_norm": 0.2120949625968933,
+      "learning_rate": 6.911152221328837e-06,
+      "loss": 0.3552,
+      "step": 1559
+    },
+    {
+      "epoch": 1.313868613138686,
+      "grad_norm": 0.23370873928070068,
+      "learning_rate": 6.90662096281016e-06,
+      "loss": 0.375,
+      "step": 1560
+    },
+    {
+      "epoch": 1.3147108366086468,
+      "grad_norm": 0.21100081503391266,
+      "learning_rate": 6.902087871221439e-06,
+      "loss": 0.3498,
+      "step": 1561
+    },
+    {
+      "epoch": 1.3155530600786074,
+      "grad_norm": 0.21074166893959045,
+      "learning_rate": 6.897552950920898e-06,
+      "loss": 0.3611,
+      "step": 1562
+    },
+    {
+      "epoch": 1.3163952835485682,
+      "grad_norm": 0.22161081433296204,
+      "learning_rate": 6.893016206268518e-06,
+      "loss": 0.3671,
+      "step": 1563
+    },
+    {
+      "epoch": 1.317237507018529,
+      "grad_norm": 0.22640085220336914,
+      "learning_rate": 6.888477641626027e-06,
+      "loss": 0.329,
+      "step": 1564
+    },
+    {
+      "epoch": 1.3180797304884897,
+      "grad_norm": 0.2186802476644516,
+      "learning_rate": 6.88393726135691e-06,
+      "loss": 0.3733,
+      "step": 1565
+    },
+    {
+      "epoch": 1.3189219539584502,
+      "grad_norm": 0.2328466773033142,
+      "learning_rate": 6.879395069826394e-06,
+      "loss": 0.3746,
+      "step": 1566
+    },
+    {
+      "epoch": 1.319764177428411,
+      "grad_norm": 0.2159368246793747,
+      "learning_rate": 6.874851071401448e-06,
+      "loss": 0.3492,
+      "step": 1567
+    },
+    {
+      "epoch": 1.3206064008983718,
+      "grad_norm": 0.21284975111484528,
+      "learning_rate": 6.870305270450779e-06,
+      "loss": 0.3656,
+      "step": 1568
+    },
+    {
+      "epoch": 1.3214486243683323,
+      "grad_norm": 0.21185383200645447,
+      "learning_rate": 6.865757671344827e-06,
+      "loss": 0.3669,
+      "step": 1569
+    },
+    {
+      "epoch": 1.322290847838293,
+      "grad_norm": 0.2309747338294983,
+      "learning_rate": 6.861208278455759e-06,
+      "loss": 0.413,
+      "step": 1570
+    },
+    {
+      "epoch": 1.3231330713082539,
+      "grad_norm": 0.2102832794189453,
+      "learning_rate": 6.856657096157469e-06,
+      "loss": 0.3168,
+      "step": 1571
+    },
+    {
+      "epoch": 1.3239752947782144,
+      "grad_norm": 0.22655142843723297,
+      "learning_rate": 6.85210412882557e-06,
+      "loss": 0.3767,
+      "step": 1572
+    },
+    {
+      "epoch": 1.3248175182481752,
+      "grad_norm": 0.2309560477733612,
+      "learning_rate": 6.8475493808373895e-06,
+      "loss": 0.3696,
+      "step": 1573
+    },
+    {
+      "epoch": 1.325659741718136,
+      "grad_norm": 0.22664643824100494,
+      "learning_rate": 6.8429928565719724e-06,
+      "loss": 0.3709,
+      "step": 1574
+    },
+    {
+      "epoch": 1.3265019651880965,
+      "grad_norm": 0.2255297303199768,
+      "learning_rate": 6.838434560410064e-06,
+      "loss": 0.3626,
+      "step": 1575
+    },
+    {
+      "epoch": 1.3273441886580573,
+      "grad_norm": 0.22523364424705505,
+      "learning_rate": 6.833874496734122e-06,
+      "loss": 0.3761,
+      "step": 1576
+    },
+    {
+      "epoch": 1.328186412128018,
+      "grad_norm": 0.21672458946704865,
+      "learning_rate": 6.829312669928293e-06,
+      "loss": 0.3518,
+      "step": 1577
+    },
+    {
+      "epoch": 1.3290286355979788,
+      "grad_norm": 0.20885531604290009,
+      "learning_rate": 6.824749084378428e-06,
+      "loss": 0.3456,
+      "step": 1578
+    },
+    {
+      "epoch": 1.3298708590679393,
+      "grad_norm": 0.2158428132534027,
+      "learning_rate": 6.820183744472062e-06,
+      "loss": 0.3732,
+      "step": 1579
+    },
+    {
+      "epoch": 1.3307130825379,
+      "grad_norm": 0.22627563774585724,
+      "learning_rate": 6.81561665459842e-06,
+      "loss": 0.376,
+      "step": 1580
+    },
+    {
+      "epoch": 1.3315553060078607,
+      "grad_norm": 0.21666988730430603,
+      "learning_rate": 6.811047819148413e-06,
+      "loss": 0.3684,
+      "step": 1581
+    },
+    {
+      "epoch": 1.3323975294778214,
+      "grad_norm": 0.20151835680007935,
+      "learning_rate": 6.806477242514623e-06,
+      "loss": 0.3128,
+      "step": 1582
+    },
+    {
+      "epoch": 1.3332397529477822,
+      "grad_norm": 0.22538265585899353,
+      "learning_rate": 6.801904929091311e-06,
+      "loss": 0.3822,
+      "step": 1583
+    },
+    {
+      "epoch": 1.334081976417743,
+      "grad_norm": 0.2307758331298828,
+      "learning_rate": 6.7973308832744035e-06,
+      "loss": 0.398,
+      "step": 1584
+    },
+    {
+      "epoch": 1.3349241998877035,
+      "grad_norm": 0.21107590198516846,
+      "learning_rate": 6.792755109461498e-06,
+      "loss": 0.3483,
+      "step": 1585
+    },
+    {
+      "epoch": 1.3357664233576643,
+      "grad_norm": 0.2177739441394806,
+      "learning_rate": 6.78817761205185e-06,
+      "loss": 0.3518,
+      "step": 1586
+    },
+    {
+      "epoch": 1.3366086468276248,
+      "grad_norm": 0.20612265169620514,
+      "learning_rate": 6.783598395446371e-06,
+      "loss": 0.3495,
+      "step": 1587
+    },
+    {
+      "epoch": 1.3374508702975856,
+      "grad_norm": 0.20119178295135498,
+      "learning_rate": 6.779017464047629e-06,
+      "loss": 0.3432,
+      "step": 1588
+    },
+    {
+      "epoch": 1.3382930937675463,
+      "grad_norm": 0.22750814259052277,
+      "learning_rate": 6.7744348222598386e-06,
+      "loss": 0.3889,
+      "step": 1589
+    },
+    {
+      "epoch": 1.3391353172375071,
+      "grad_norm": 0.22598476707935333,
+      "learning_rate": 6.769850474488859e-06,
+      "loss": 0.3693,
+      "step": 1590
+    },
+    {
+      "epoch": 1.3399775407074677,
+      "grad_norm": 0.1982087939977646,
+      "learning_rate": 6.7652644251421875e-06,
+      "loss": 0.3141,
+      "step": 1591
+    },
+    {
+      "epoch": 1.3408197641774284,
+      "grad_norm": 0.22299589216709137,
+      "learning_rate": 6.7606766786289624e-06,
+      "loss": 0.3586,
+      "step": 1592
+    },
+    {
+      "epoch": 1.341661987647389,
+      "grad_norm": 0.20625364780426025,
+      "learning_rate": 6.756087239359948e-06,
+      "loss": 0.356,
+      "step": 1593
+    },
+    {
+      "epoch": 1.3425042111173497,
+      "grad_norm": 0.21917255222797394,
+      "learning_rate": 6.75149611174754e-06,
+      "loss": 0.3581,
+      "step": 1594
+    },
+    {
+      "epoch": 1.3433464345873105,
+      "grad_norm": 0.22012881934642792,
+      "learning_rate": 6.746903300205756e-06,
+      "loss": 0.3746,
+      "step": 1595
+    },
+    {
+      "epoch": 1.3441886580572713,
+      "grad_norm": 0.20556309819221497,
+      "learning_rate": 6.742308809150232e-06,
+      "loss": 0.3787,
+      "step": 1596
+    },
+    {
+      "epoch": 1.3450308815272318,
+      "grad_norm": 0.19921176135540009,
+      "learning_rate": 6.737712642998219e-06,
+      "loss": 0.3461,
+      "step": 1597
+    },
+    {
+      "epoch": 1.3458731049971926,
+      "grad_norm": 0.2157866656780243,
+      "learning_rate": 6.7331148061685796e-06,
+      "loss": 0.3814,
+      "step": 1598
+    },
+    {
+      "epoch": 1.3467153284671534,
+      "grad_norm": 0.213243767619133,
+      "learning_rate": 6.728515303081782e-06,
+      "loss": 0.3405,
+      "step": 1599
+    },
+    {
+      "epoch": 1.347557551937114,
+      "grad_norm": 0.23965325951576233,
+      "learning_rate": 6.723914138159895e-06,
+      "loss": 0.3728,
+      "step": 1600
+    },
+    {
+      "epoch": 1.3483997754070747,
+      "grad_norm": 0.22270913422107697,
+      "learning_rate": 6.719311315826589e-06,
+      "loss": 0.3781,
+      "step": 1601
+    },
+    {
+      "epoch": 1.3492419988770354,
+      "grad_norm": 0.20295816659927368,
+      "learning_rate": 6.714706840507122e-06,
+      "loss": 0.3256,
+      "step": 1602
+    },
+    {
+      "epoch": 1.350084222346996,
+      "grad_norm": 0.24144160747528076,
+      "learning_rate": 6.710100716628345e-06,
+      "loss": 0.3677,
+      "step": 1603
+    },
+    {
+      "epoch": 1.3509264458169568,
+      "grad_norm": 0.24592921137809753,
+      "learning_rate": 6.705492948618694e-06,
+      "loss": 0.3813,
+      "step": 1604
+    },
+    {
+      "epoch": 1.3517686692869175,
+      "grad_norm": 0.20310568809509277,
+      "learning_rate": 6.700883540908185e-06,
+      "loss": 0.3532,
+      "step": 1605
+    },
+    {
+      "epoch": 1.352610892756878,
+      "grad_norm": 0.30135607719421387,
+      "learning_rate": 6.696272497928411e-06,
+      "loss": 0.4034,
+      "step": 1606
+    },
+    {
+      "epoch": 1.3534531162268388,
+      "grad_norm": 0.20306457579135895,
+      "learning_rate": 6.691659824112535e-06,
+      "loss": 0.3345,
+      "step": 1607
+    },
+    {
+      "epoch": 1.3542953396967996,
+      "grad_norm": 0.1948816031217575,
+      "learning_rate": 6.687045523895292e-06,
+      "loss": 0.3221,
+      "step": 1608
+    },
+    {
+      "epoch": 1.3551375631667604,
+      "grad_norm": 0.23035724461078644,
+      "learning_rate": 6.682429601712976e-06,
+      "loss": 0.3724,
+      "step": 1609
+    },
+    {
+      "epoch": 1.355979786636721,
+      "grad_norm": 0.2039736956357956,
+      "learning_rate": 6.6778120620034455e-06,
+      "loss": 0.3477,
+      "step": 1610
+    },
+    {
+      "epoch": 1.3568220101066817,
+      "grad_norm": 0.22455595433712006,
+      "learning_rate": 6.673192909206109e-06,
+      "loss": 0.3985,
+      "step": 1611
+    },
+    {
+      "epoch": 1.3576642335766422,
+      "grad_norm": 0.20514552295207977,
+      "learning_rate": 6.668572147761929e-06,
+      "loss": 0.3361,
+      "step": 1612
+    },
+    {
+      "epoch": 1.358506457046603,
+      "grad_norm": 0.20776930451393127,
+      "learning_rate": 6.663949782113413e-06,
+      "loss": 0.3481,
+      "step": 1613
+    },
+    {
+      "epoch": 1.3593486805165638,
+      "grad_norm": 0.27368056774139404,
+      "learning_rate": 6.6593258167046115e-06,
+      "loss": 0.3668,
+      "step": 1614
+    },
+    {
+      "epoch": 1.3601909039865245,
+      "grad_norm": 0.19447094202041626,
+      "learning_rate": 6.654700255981115e-06,
+      "loss": 0.3312,
+      "step": 1615
+    },
+    {
+      "epoch": 1.361033127456485,
+      "grad_norm": 0.20037733018398285,
+      "learning_rate": 6.6500731043900425e-06,
+      "loss": 0.354,
+      "step": 1616
+    },
+    {
+      "epoch": 1.3618753509264458,
+      "grad_norm": 0.2207939624786377,
+      "learning_rate": 6.64544436638005e-06,
+      "loss": 0.3597,
+      "step": 1617
+    },
+    {
+      "epoch": 1.3627175743964064,
+      "grad_norm": 0.22954882681369781,
+      "learning_rate": 6.640814046401312e-06,
+      "loss": 0.4111,
+      "step": 1618
+    },
+    {
+      "epoch": 1.3635597978663672,
+      "grad_norm": 0.2107580602169037,
+      "learning_rate": 6.6361821489055275e-06,
+      "loss": 0.3634,
+      "step": 1619
+    },
+    {
+      "epoch": 1.364402021336328,
+      "grad_norm": 0.25928613543510437,
+      "learning_rate": 6.63154867834591e-06,
+      "loss": 0.3793,
+      "step": 1620
+    },
+    {
+      "epoch": 1.3652442448062887,
+      "grad_norm": 0.19150257110595703,
+      "learning_rate": 6.626913639177189e-06,
+      "loss": 0.3234,
+      "step": 1621
+    },
+    {
+      "epoch": 1.3660864682762492,
+      "grad_norm": 0.23382145166397095,
+      "learning_rate": 6.622277035855596e-06,
+      "loss": 0.3803,
+      "step": 1622
+    },
+    {
+      "epoch": 1.36692869174621,
+      "grad_norm": 0.21147483587265015,
+      "learning_rate": 6.617638872838874e-06,
+      "loss": 0.3307,
+      "step": 1623
+    },
+    {
+      "epoch": 1.3677709152161706,
+      "grad_norm": 0.19452020525932312,
+      "learning_rate": 6.61299915458626e-06,
+      "loss": 0.3397,
+      "step": 1624
+    },
+    {
+      "epoch": 1.3686131386861313,
+      "grad_norm": 0.21936218440532684,
+      "learning_rate": 6.608357885558485e-06,
+      "loss": 0.3613,
+      "step": 1625
+    },
+    {
+      "epoch": 1.369455362156092,
+      "grad_norm": 0.21089312434196472,
+      "learning_rate": 6.603715070217779e-06,
+      "loss": 0.367,
+      "step": 1626
+    },
+    {
+      "epoch": 1.3702975856260529,
+      "grad_norm": 0.22670963406562805,
+      "learning_rate": 6.599070713027849e-06,
+      "loss": 0.3447,
+      "step": 1627
+    },
+    {
+      "epoch": 1.3711398090960134,
+      "grad_norm": 0.21916204690933228,
+      "learning_rate": 6.594424818453891e-06,
+      "loss": 0.3479,
+      "step": 1628
+    },
+    {
+      "epoch": 1.3719820325659742,
+      "grad_norm": 0.23116189241409302,
+      "learning_rate": 6.589777390962575e-06,
+      "loss": 0.366,
+      "step": 1629
+    },
+    {
+      "epoch": 1.372824256035935,
+      "grad_norm": 0.21582533419132233,
+      "learning_rate": 6.58512843502205e-06,
+      "loss": 0.354,
+      "step": 1630
+    },
+    {
+      "epoch": 1.3736664795058955,
+      "grad_norm": 0.2179097682237625,
+      "learning_rate": 6.580477955101927e-06,
+      "loss": 0.3345,
+      "step": 1631
+    },
+    {
+      "epoch": 1.3745087029758563,
+      "grad_norm": 0.20926107466220856,
+      "learning_rate": 6.5758259556732896e-06,
+      "loss": 0.358,
+      "step": 1632
+    },
+    {
+      "epoch": 1.375350926445817,
+      "grad_norm": 0.22556179761886597,
+      "learning_rate": 6.571172441208678e-06,
+      "loss": 0.3683,
+      "step": 1633
+    },
+    {
+      "epoch": 1.3761931499157778,
+      "grad_norm": 0.21263204514980316,
+      "learning_rate": 6.566517416182088e-06,
+      "loss": 0.3358,
+      "step": 1634
+    },
+    {
+      "epoch": 1.3770353733857383,
+      "grad_norm": 0.34583309292793274,
+      "learning_rate": 6.561860885068972e-06,
+      "loss": 0.3307,
+      "step": 1635
+    },
+    {
+      "epoch": 1.377877596855699,
+      "grad_norm": 0.2124481499195099,
+      "learning_rate": 6.5572028523462275e-06,
+      "loss": 0.376,
+      "step": 1636
+    },
+    {
+      "epoch": 1.3787198203256597,
+      "grad_norm": 0.2069571316242218,
+      "learning_rate": 6.552543322492195e-06,
+      "loss": 0.3406,
+      "step": 1637
+    },
+    {
+      "epoch": 1.3795620437956204,
+      "grad_norm": 0.22916629910469055,
+      "learning_rate": 6.547882299986658e-06,
+      "loss": 0.3512,
+      "step": 1638
+    },
+    {
+      "epoch": 1.3804042672655812,
+      "grad_norm": 0.22375811636447906,
+      "learning_rate": 6.54321978931083e-06,
+      "loss": 0.3636,
+      "step": 1639
+    },
+    {
+      "epoch": 1.381246490735542,
+      "grad_norm": 0.21804262697696686,
+      "learning_rate": 6.53855579494736e-06,
+      "loss": 0.3548,
+      "step": 1640
+    },
+    {
+      "epoch": 1.3820887142055025,
+      "grad_norm": 0.2238789200782776,
+      "learning_rate": 6.53389032138032e-06,
+      "loss": 0.3794,
+      "step": 1641
+    },
+    {
+      "epoch": 1.3829309376754633,
+      "grad_norm": 0.21942567825317383,
+      "learning_rate": 6.5292233730952074e-06,
+      "loss": 0.3548,
+      "step": 1642
+    },
+    {
+      "epoch": 1.3837731611454238,
+      "grad_norm": 0.2897321581840515,
+      "learning_rate": 6.5245549545789335e-06,
+      "loss": 0.37,
+      "step": 1643
+    },
+    {
+      "epoch": 1.3846153846153846,
+      "grad_norm": 0.2234940081834793,
+      "learning_rate": 6.519885070319827e-06,
+      "loss": 0.3306,
+      "step": 1644
+    },
+    {
+      "epoch": 1.3854576080853453,
+      "grad_norm": 0.22789642214775085,
+      "learning_rate": 6.515213724807621e-06,
+      "loss": 0.3668,
+      "step": 1645
+    },
+    {
+      "epoch": 1.3862998315553061,
+      "grad_norm": 0.23737990856170654,
+      "learning_rate": 6.51054092253346e-06,
+      "loss": 0.3863,
+      "step": 1646
+    },
+    {
+      "epoch": 1.3871420550252667,
+      "grad_norm": 0.22608612477779388,
+      "learning_rate": 6.505866667989884e-06,
+      "loss": 0.3305,
+      "step": 1647
+    },
+    {
+      "epoch": 1.3879842784952274,
+      "grad_norm": 0.2412174493074417,
+      "learning_rate": 6.5011909656708305e-06,
+      "loss": 0.3769,
+      "step": 1648
+    },
+    {
+      "epoch": 1.388826501965188,
+      "grad_norm": 0.19889724254608154,
+      "learning_rate": 6.49651382007163e-06,
+      "loss": 0.3208,
+      "step": 1649
+    },
+    {
+      "epoch": 1.3896687254351487,
+      "grad_norm": 0.22150558233261108,
+      "learning_rate": 6.491835235688999e-06,
+      "loss": 0.3595,
+      "step": 1650
+    },
+    {
+      "epoch": 1.3905109489051095,
+      "grad_norm": 0.21915128827095032,
+      "learning_rate": 6.487155217021039e-06,
+      "loss": 0.3518,
+      "step": 1651
+    },
+    {
+      "epoch": 1.3913531723750703,
+      "grad_norm": 0.2163122445344925,
+      "learning_rate": 6.482473768567228e-06,
+      "loss": 0.3458,
+      "step": 1652
+    },
+    {
+      "epoch": 1.3921953958450308,
+      "grad_norm": 0.21529380977153778,
+      "learning_rate": 6.477790894828422e-06,
+      "loss": 0.3574,
+      "step": 1653
+    },
+    {
+      "epoch": 1.3930376193149916,
+      "grad_norm": 0.23291248083114624,
+      "learning_rate": 6.473106600306842e-06,
+      "loss": 0.3687,
+      "step": 1654
+    },
+    {
+      "epoch": 1.3938798427849521,
+      "grad_norm": 0.2185925394296646,
+      "learning_rate": 6.468420889506084e-06,
+      "loss": 0.3416,
+      "step": 1655
+    },
+    {
+      "epoch": 1.394722066254913,
+      "grad_norm": 0.22031629085540771,
+      "learning_rate": 6.463733766931096e-06,
+      "loss": 0.3562,
+      "step": 1656
+    },
+    {
+      "epoch": 1.3955642897248737,
+      "grad_norm": 0.24342696368694305,
+      "learning_rate": 6.459045237088189e-06,
+      "loss": 0.3705,
+      "step": 1657
+    },
+    {
+      "epoch": 1.3964065131948344,
+      "grad_norm": 0.23241513967514038,
+      "learning_rate": 6.454355304485024e-06,
+      "loss": 0.3376,
+      "step": 1658
+    },
+    {
+      "epoch": 1.397248736664795,
+      "grad_norm": 0.2287467122077942,
+      "learning_rate": 6.449663973630613e-06,
+      "loss": 0.3353,
+      "step": 1659
+    },
+    {
+      "epoch": 1.3980909601347558,
+      "grad_norm": 0.21655339002609253,
+      "learning_rate": 6.444971249035312e-06,
+      "loss": 0.3637,
+      "step": 1660
+    },
+    {
+      "epoch": 1.3989331836047165,
+      "grad_norm": 0.21061834692955017,
+      "learning_rate": 6.440277135210815e-06,
+      "loss": 0.3745,
+      "step": 1661
+    },
+    {
+      "epoch": 1.399775407074677,
+      "grad_norm": 0.2187851071357727,
+      "learning_rate": 6.435581636670154e-06,
+      "loss": 0.3574,
+      "step": 1662
+    },
+    {
+      "epoch": 1.4006176305446378,
+      "grad_norm": 0.21468669176101685,
+      "learning_rate": 6.43088475792769e-06,
+      "loss": 0.3555,
+      "step": 1663
+    },
+    {
+      "epoch": 1.4014598540145986,
+      "grad_norm": 0.21525272727012634,
+      "learning_rate": 6.426186503499114e-06,
+      "loss": 0.3474,
+      "step": 1664
+    },
+    {
+      "epoch": 1.4023020774845594,
+      "grad_norm": 0.2308477908372879,
+      "learning_rate": 6.421486877901436e-06,
+      "loss": 0.3469,
+      "step": 1665
+    },
+    {
+      "epoch": 1.40314430095452,
+      "grad_norm": 0.23242992162704468,
+      "learning_rate": 6.4167858856529875e-06,
+      "loss": 0.3822,
+      "step": 1666
+    },
+    {
+      "epoch": 1.4039865244244807,
+      "grad_norm": 0.20637556910514832,
+      "learning_rate": 6.412083531273411e-06,
+      "loss": 0.3638,
+      "step": 1667
+    },
+    {
+      "epoch": 1.4048287478944412,
+      "grad_norm": 0.2414320558309555,
+      "learning_rate": 6.407379819283661e-06,
+      "loss": 0.3518,
+      "step": 1668
+    },
+    {
+      "epoch": 1.405670971364402,
+      "grad_norm": 0.23745858669281006,
+      "learning_rate": 6.402674754205998e-06,
+      "loss": 0.3575,
+      "step": 1669
+    },
+    {
+      "epoch": 1.4065131948343628,
+      "grad_norm": 0.23913931846618652,
+      "learning_rate": 6.397968340563978e-06,
+      "loss": 0.371,
+      "step": 1670
+    },
+    {
+      "epoch": 1.4073554183043235,
+      "grad_norm": 0.22450561821460724,
+      "learning_rate": 6.393260582882462e-06,
+      "loss": 0.3219,
+      "step": 1671
+    },
+    {
+      "epoch": 1.408197641774284,
+      "grad_norm": 0.23129063844680786,
+      "learning_rate": 6.3885514856875945e-06,
+      "loss": 0.3987,
+      "step": 1672
+    },
+    {
+      "epoch": 1.4090398652442448,
+      "grad_norm": 0.24473051726818085,
+      "learning_rate": 6.383841053506813e-06,
+      "loss": 0.382,
+      "step": 1673
+    },
+    {
+      "epoch": 1.4098820887142054,
+      "grad_norm": 0.21250121295452118,
+      "learning_rate": 6.379129290868837e-06,
+      "loss": 0.3419,
+      "step": 1674
+    },
+    {
+      "epoch": 1.4107243121841662,
+      "grad_norm": 0.223237544298172,
+      "learning_rate": 6.3744162023036685e-06,
+      "loss": 0.3889,
+      "step": 1675
+    },
+    {
+      "epoch": 1.411566535654127,
+      "grad_norm": 0.22709648311138153,
+      "learning_rate": 6.369701792342576e-06,
+      "loss": 0.3552,
+      "step": 1676
+    },
+    {
+      "epoch": 1.4124087591240877,
+      "grad_norm": 0.20558980107307434,
+      "learning_rate": 6.364986065518106e-06,
+      "loss": 0.3356,
+      "step": 1677
+    },
+    {
+      "epoch": 1.4132509825940482,
+      "grad_norm": 0.23428533971309662,
+      "learning_rate": 6.360269026364071e-06,
+      "loss": 0.3917,
+      "step": 1678
+    },
+    {
+      "epoch": 1.414093206064009,
+      "grad_norm": 0.2525196671485901,
+      "learning_rate": 6.35555067941554e-06,
+      "loss": 0.3648,
+      "step": 1679
+    },
+    {
+      "epoch": 1.4149354295339696,
+      "grad_norm": 0.2390659749507904,
+      "learning_rate": 6.350831029208844e-06,
+      "loss": 0.3376,
+      "step": 1680
+    },
+    {
+      "epoch": 1.4157776530039303,
+      "grad_norm": 0.21221652626991272,
+      "learning_rate": 6.3461100802815625e-06,
+      "loss": 0.3609,
+      "step": 1681
+    },
+    {
+      "epoch": 1.416619876473891,
+      "grad_norm": 0.22451327741146088,
+      "learning_rate": 6.34138783717253e-06,
+      "loss": 0.3531,
+      "step": 1682
+    },
+    {
+      "epoch": 1.4174620999438519,
+      "grad_norm": 0.2102045863866806,
+      "learning_rate": 6.336664304421818e-06,
+      "loss": 0.3108,
+      "step": 1683
+    },
+    {
+      "epoch": 1.4183043234138124,
+      "grad_norm": 0.21954309940338135,
+      "learning_rate": 6.331939486570745e-06,
+      "loss": 0.3837,
+      "step": 1684
+    },
+    {
+      "epoch": 1.4191465468837732,
+      "grad_norm": 0.20022211968898773,
+      "learning_rate": 6.3272133881618596e-06,
+      "loss": 0.3487,
+      "step": 1685
+    },
+    {
+      "epoch": 1.4199887703537337,
+      "grad_norm": 0.22107923030853271,
+      "learning_rate": 6.322486013738942e-06,
+      "loss": 0.3499,
+      "step": 1686
+    },
+    {
+      "epoch": 1.4208309938236945,
+      "grad_norm": 0.22098489105701447,
+      "learning_rate": 6.317757367847005e-06,
+      "loss": 0.3242,
+      "step": 1687
+    },
+    {
+      "epoch": 1.4216732172936553,
+      "grad_norm": 0.2224896252155304,
+      "learning_rate": 6.313027455032274e-06,
+      "loss": 0.4122,
+      "step": 1688
+    },
+    {
+      "epoch": 1.422515440763616,
+      "grad_norm": 0.20495165884494781,
+      "learning_rate": 6.308296279842204e-06,
+      "loss": 0.3367,
+      "step": 1689
+    },
+    {
+      "epoch": 1.4233576642335766,
+      "grad_norm": 0.23125077784061432,
+      "learning_rate": 6.303563846825453e-06,
+      "loss": 0.3776,
+      "step": 1690
+    },
+    {
+      "epoch": 1.4241998877035373,
+      "grad_norm": 0.2269773632287979,
+      "learning_rate": 6.298830160531895e-06,
+      "loss": 0.3647,
+      "step": 1691
+    },
+    {
+      "epoch": 1.425042111173498,
+      "grad_norm": 0.20680685341358185,
+      "learning_rate": 6.294095225512604e-06,
+      "loss": 0.3503,
+      "step": 1692
+    },
+    {
+      "epoch": 1.4258843346434587,
+      "grad_norm": 0.22571782767772675,
+      "learning_rate": 6.289359046319862e-06,
+      "loss": 0.4192,
+      "step": 1693
+    },
+    {
+      "epoch": 1.4267265581134194,
+      "grad_norm": 0.21243885159492493,
+      "learning_rate": 6.2846216275071395e-06,
+      "loss": 0.3199,
+      "step": 1694
+    },
+    {
+      "epoch": 1.4275687815833802,
+      "grad_norm": 0.2743223309516907,
+      "learning_rate": 6.279882973629101e-06,
+      "loss": 0.3622,
+      "step": 1695
+    },
+    {
+      "epoch": 1.428411005053341,
+      "grad_norm": 0.19054637849330902,
+      "learning_rate": 6.275143089241603e-06,
+      "loss": 0.3307,
+      "step": 1696
+    },
+    {
+      "epoch": 1.4292532285233015,
+      "grad_norm": 0.27557373046875,
+      "learning_rate": 6.270401978901678e-06,
+      "loss": 0.4077,
+      "step": 1697
+    },
+    {
+      "epoch": 1.4300954519932623,
+      "grad_norm": 0.24843667447566986,
+      "learning_rate": 6.265659647167542e-06,
+      "loss": 0.369,
+      "step": 1698
+    },
+    {
+      "epoch": 1.4309376754632228,
+      "grad_norm": 0.23135574162006378,
+      "learning_rate": 6.260916098598584e-06,
+      "loss": 0.3172,
+      "step": 1699
+    },
+    {
+      "epoch": 1.4317798989331836,
+      "grad_norm": 0.19605474174022675,
+      "learning_rate": 6.256171337755362e-06,
+      "loss": 0.342,
+      "step": 1700
+    },
+    {
+      "epoch": 1.4326221224031443,
+      "grad_norm": 0.2249126434326172,
+      "learning_rate": 6.2514253691996e-06,
+      "loss": 0.3809,
+      "step": 1701
+    },
+    {
+      "epoch": 1.4334643458731051,
+      "grad_norm": 0.22616976499557495,
+      "learning_rate": 6.246678197494185e-06,
+      "loss": 0.3946,
+      "step": 1702
+    },
+    {
+      "epoch": 1.4343065693430657,
+      "grad_norm": 0.1968989074230194,
+      "learning_rate": 6.241929827203156e-06,
+      "loss": 0.3583,
+      "step": 1703
+    },
+    {
+      "epoch": 1.4351487928130264,
+      "grad_norm": 0.21847349405288696,
+      "learning_rate": 6.237180262891709e-06,
+      "loss": 0.3921,
+      "step": 1704
+    },
+    {
+      "epoch": 1.435991016282987,
+      "grad_norm": 0.19567416608333588,
+      "learning_rate": 6.2324295091261885e-06,
+      "loss": 0.3554,
+      "step": 1705
+    },
+    {
+      "epoch": 1.4368332397529477,
+      "grad_norm": 0.21993115544319153,
+      "learning_rate": 6.227677570474077e-06,
+      "loss": 0.3655,
+      "step": 1706
+    },
+    {
+      "epoch": 1.4376754632229085,
+      "grad_norm": 0.19446784257888794,
+      "learning_rate": 6.222924451504001e-06,
+      "loss": 0.3544,
+      "step": 1707
+    },
+    {
+      "epoch": 1.4385176866928693,
+      "grad_norm": 0.2094394713640213,
+      "learning_rate": 6.21817015678572e-06,
+      "loss": 0.337,
+      "step": 1708
+    },
+    {
+      "epoch": 1.4393599101628298,
+      "grad_norm": 0.22094875574111938,
+      "learning_rate": 6.213414690890125e-06,
+      "loss": 0.3941,
+      "step": 1709
+    },
+    {
+      "epoch": 1.4402021336327906,
+      "grad_norm": 0.1867690086364746,
+      "learning_rate": 6.208658058389232e-06,
+      "loss": 0.3237,
+      "step": 1710
+    },
+    {
+      "epoch": 1.4410443571027511,
+      "grad_norm": 0.2281683087348938,
+      "learning_rate": 6.203900263856177e-06,
+      "loss": 0.3747,
+      "step": 1711
+    },
+    {
+      "epoch": 1.441886580572712,
+      "grad_norm": 0.18243515491485596,
+      "learning_rate": 6.19914131186522e-06,
+      "loss": 0.3331,
+      "step": 1712
+    },
+    {
+      "epoch": 1.4427288040426727,
+      "grad_norm": 0.18891458213329315,
+      "learning_rate": 6.194381206991723e-06,
+      "loss": 0.3709,
+      "step": 1713
+    },
+    {
+      "epoch": 1.4435710275126334,
+      "grad_norm": 0.21232585608959198,
+      "learning_rate": 6.189619953812167e-06,
+      "loss": 0.3641,
+      "step": 1714
+    },
+    {
+      "epoch": 1.444413250982594,
+      "grad_norm": 0.21884340047836304,
+      "learning_rate": 6.184857556904129e-06,
+      "loss": 0.354,
+      "step": 1715
+    },
+    {
+      "epoch": 1.4452554744525548,
+      "grad_norm": 0.284342497587204,
+      "learning_rate": 6.180094020846291e-06,
+      "loss": 0.3789,
+      "step": 1716
+    },
+    {
+      "epoch": 1.4460976979225155,
+      "grad_norm": 0.1979561448097229,
+      "learning_rate": 6.175329350218426e-06,
+      "loss": 0.3294,
+      "step": 1717
+    },
+    {
+      "epoch": 1.446939921392476,
+      "grad_norm": 0.26564377546310425,
+      "learning_rate": 6.170563549601402e-06,
+      "loss": 0.3644,
+      "step": 1718
+    },
+    {
+      "epoch": 1.4477821448624368,
+      "grad_norm": 0.21076220273971558,
+      "learning_rate": 6.165796623577171e-06,
+      "loss": 0.3588,
+      "step": 1719
+    },
+    {
+      "epoch": 1.4486243683323976,
+      "grad_norm": 0.22534281015396118,
+      "learning_rate": 6.161028576728767e-06,
+      "loss": 0.3614,
+      "step": 1720
+    },
+    {
+      "epoch": 1.4494665918023582,
+      "grad_norm": 0.22138455510139465,
+      "learning_rate": 6.156259413640302e-06,
+      "loss": 0.3963,
+      "step": 1721
+    },
+    {
+      "epoch": 1.450308815272319,
+      "grad_norm": 0.20443442463874817,
+      "learning_rate": 6.15148913889696e-06,
+      "loss": 0.351,
+      "step": 1722
+    },
+    {
+      "epoch": 1.4511510387422797,
+      "grad_norm": 0.21905840933322906,
+      "learning_rate": 6.146717757084995e-06,
+      "loss": 0.4063,
+      "step": 1723
+    },
+    {
+      "epoch": 1.4519932622122402,
+      "grad_norm": 0.20653998851776123,
+      "learning_rate": 6.141945272791727e-06,
+      "loss": 0.3584,
+      "step": 1724
+    },
+    {
+      "epoch": 1.452835485682201,
+      "grad_norm": 0.1866150051355362,
+      "learning_rate": 6.1371716906055336e-06,
+      "loss": 0.3513,
+      "step": 1725
+    },
+    {
+      "epoch": 1.4536777091521618,
+      "grad_norm": 0.1922345757484436,
+      "learning_rate": 6.132397015115846e-06,
+      "loss": 0.3332,
+      "step": 1726
+    },
+    {
+      "epoch": 1.4545199326221225,
+      "grad_norm": 0.2280290275812149,
+      "learning_rate": 6.127621250913152e-06,
+      "loss": 0.389,
+      "step": 1727
+    },
+    {
+      "epoch": 1.455362156092083,
+      "grad_norm": 0.2247847467660904,
+      "learning_rate": 6.122844402588982e-06,
+      "loss": 0.3664,
+      "step": 1728
+    },
+    {
+      "epoch": 1.4562043795620438,
+      "grad_norm": 0.21930207312107086,
+      "learning_rate": 6.11806647473591e-06,
+      "loss": 0.3643,
+      "step": 1729
+    },
+    {
+      "epoch": 1.4570466030320044,
+      "grad_norm": 0.21347032487392426,
+      "learning_rate": 6.113287471947547e-06,
+      "loss": 0.3611,
+      "step": 1730
+    },
+    {
+      "epoch": 1.4578888265019652,
+      "grad_norm": 0.20544639229774475,
+      "learning_rate": 6.10850739881854e-06,
+      "loss": 0.3591,
+      "step": 1731
+    },
+    {
+      "epoch": 1.458731049971926,
+      "grad_norm": 0.20113444328308105,
+      "learning_rate": 6.103726259944562e-06,
+      "loss": 0.3545,
+      "step": 1732
+    },
+    {
+      "epoch": 1.4595732734418867,
+      "grad_norm": 0.19256292283535004,
+      "learning_rate": 6.098944059922311e-06,
+      "loss": 0.3378,
+      "step": 1733
+    },
+    {
+      "epoch": 1.4604154969118472,
+      "grad_norm": 0.22689050436019897,
+      "learning_rate": 6.094160803349508e-06,
+      "loss": 0.3587,
+      "step": 1734
+    },
+    {
+      "epoch": 1.461257720381808,
+      "grad_norm": 0.23645469546318054,
+      "learning_rate": 6.089376494824886e-06,
+      "loss": 0.4021,
+      "step": 1735
+    },
+    {
+      "epoch": 1.4620999438517686,
+      "grad_norm": 0.19880850613117218,
+      "learning_rate": 6.084591138948192e-06,
+      "loss": 0.3314,
+      "step": 1736
+    },
+    {
+      "epoch": 1.4629421673217293,
+      "grad_norm": 0.21392516791820526,
+      "learning_rate": 6.079804740320181e-06,
+      "loss": 0.3278,
+      "step": 1737
+    },
+    {
+      "epoch": 1.46378439079169,
+      "grad_norm": 0.20528754591941833,
+      "learning_rate": 6.075017303542605e-06,
+      "loss": 0.4016,
+      "step": 1738
+    },
+    {
+      "epoch": 1.4646266142616509,
+      "grad_norm": 0.19299480319023132,
+      "learning_rate": 6.070228833218221e-06,
+      "loss": 0.3255,
+      "step": 1739
+    },
+    {
+      "epoch": 1.4654688377316114,
+      "grad_norm": 0.21759140491485596,
+      "learning_rate": 6.065439333950776e-06,
+      "loss": 0.3951,
+      "step": 1740
+    },
+    {
+      "epoch": 1.4663110612015722,
+      "grad_norm": 0.20918811857700348,
+      "learning_rate": 6.060648810345006e-06,
+      "loss": 0.3355,
+      "step": 1741
+    },
+    {
+      "epoch": 1.4671532846715327,
+      "grad_norm": 0.22659289836883545,
+      "learning_rate": 6.055857267006631e-06,
+      "loss": 0.3633,
+      "step": 1742
+    },
+    {
+      "epoch": 1.4679955081414935,
+      "grad_norm": 0.2189176082611084,
+      "learning_rate": 6.051064708542357e-06,
+      "loss": 0.3777,
+      "step": 1743
+    },
+    {
+      "epoch": 1.4688377316114543,
+      "grad_norm": 0.20815801620483398,
+      "learning_rate": 6.046271139559859e-06,
+      "loss": 0.3448,
+      "step": 1744
+    },
+    {
+      "epoch": 1.469679955081415,
+      "grad_norm": 0.2091105580329895,
+      "learning_rate": 6.041476564667785e-06,
+      "loss": 0.3552,
+      "step": 1745
+    },
+    {
+      "epoch": 1.4705221785513756,
+      "grad_norm": 0.203644797205925,
+      "learning_rate": 6.036680988475756e-06,
+      "loss": 0.3248,
+      "step": 1746
+    },
+    {
+      "epoch": 1.4713644020213363,
+      "grad_norm": 0.20944565534591675,
+      "learning_rate": 6.031884415594347e-06,
+      "loss": 0.3752,
+      "step": 1747
+    },
+    {
+      "epoch": 1.472206625491297,
+      "grad_norm": 0.215575709939003,
+      "learning_rate": 6.0270868506351e-06,
+      "loss": 0.333,
+      "step": 1748
+    },
+    {
+      "epoch": 1.4730488489612577,
+      "grad_norm": 0.233964204788208,
+      "learning_rate": 6.022288298210502e-06,
+      "loss": 0.4178,
+      "step": 1749
+    },
+    {
+      "epoch": 1.4738910724312184,
+      "grad_norm": 0.22134220600128174,
+      "learning_rate": 6.017488762933996e-06,
+      "loss": 0.3831,
+      "step": 1750
+    },
+    {
+      "epoch": 1.4747332959011792,
+      "grad_norm": 0.2089032381772995,
+      "learning_rate": 6.012688249419966e-06,
+      "loss": 0.3641,
+      "step": 1751
+    },
+    {
+      "epoch": 1.4755755193711397,
+      "grad_norm": 0.2150684893131256,
+      "learning_rate": 6.00788676228374e-06,
+      "loss": 0.3739,
+      "step": 1752
+    },
+    {
+      "epoch": 1.4764177428411005,
+      "grad_norm": 0.20269526541233063,
+      "learning_rate": 6.003084306141579e-06,
+      "loss": 0.3398,
+      "step": 1753
+    },
+    {
+      "epoch": 1.4772599663110613,
+      "grad_norm": 0.20560164749622345,
+      "learning_rate": 5.998280885610677e-06,
+      "loss": 0.338,
+      "step": 1754
+    },
+    {
+      "epoch": 1.4781021897810218,
+      "grad_norm": 0.22493983805179596,
+      "learning_rate": 5.993476505309154e-06,
+      "loss": 0.4067,
+      "step": 1755
+    },
+    {
+      "epoch": 1.4789444132509826,
+      "grad_norm": 0.20303654670715332,
+      "learning_rate": 5.988671169856056e-06,
+      "loss": 0.3289,
+      "step": 1756
+    },
+    {
+      "epoch": 1.4797866367209433,
+      "grad_norm": 0.21149663627147675,
+      "learning_rate": 5.983864883871344e-06,
+      "loss": 0.3467,
+      "step": 1757
+    },
+    {
+      "epoch": 1.4806288601909041,
+      "grad_norm": 0.21716906130313873,
+      "learning_rate": 5.979057651975893e-06,
+      "loss": 0.3735,
+      "step": 1758
+    },
+    {
+      "epoch": 1.4814710836608647,
+      "grad_norm": 0.23328666388988495,
+      "learning_rate": 5.974249478791489e-06,
+      "loss": 0.3605,
+      "step": 1759
+    },
+    {
+      "epoch": 1.4823133071308254,
+      "grad_norm": 0.22171860933303833,
+      "learning_rate": 5.969440368940823e-06,
+      "loss": 0.3493,
+      "step": 1760
+    },
+    {
+      "epoch": 1.483155530600786,
+      "grad_norm": 0.1985427439212799,
+      "learning_rate": 5.964630327047485e-06,
+      "loss": 0.3656,
+      "step": 1761
+    },
+    {
+      "epoch": 1.4839977540707467,
+      "grad_norm": 0.22209234535694122,
+      "learning_rate": 5.9598193577359606e-06,
+      "loss": 0.3545,
+      "step": 1762
+    },
+    {
+      "epoch": 1.4848399775407075,
+      "grad_norm": 0.20604878664016724,
+      "learning_rate": 5.955007465631632e-06,
+      "loss": 0.348,
+      "step": 1763
+    },
+    {
+      "epoch": 1.4856822010106683,
+      "grad_norm": 0.2349865734577179,
+      "learning_rate": 5.9501946553607615e-06,
+      "loss": 0.3721,
+      "step": 1764
+    },
+    {
+      "epoch": 1.4865244244806288,
+      "grad_norm": 0.22401748597621918,
+      "learning_rate": 5.945380931550497e-06,
+      "loss": 0.3978,
+      "step": 1765
+    },
+    {
+      "epoch": 1.4873666479505896,
+      "grad_norm": 0.20538973808288574,
+      "learning_rate": 5.940566298828871e-06,
+      "loss": 0.3581,
+      "step": 1766
+    },
+    {
+      "epoch": 1.4882088714205501,
+      "grad_norm": 0.22036215662956238,
+      "learning_rate": 5.935750761824777e-06,
+      "loss": 0.3681,
+      "step": 1767
+    },
+    {
+      "epoch": 1.489051094890511,
+      "grad_norm": 0.22985632717609406,
+      "learning_rate": 5.93093432516799e-06,
+      "loss": 0.349,
+      "step": 1768
+    },
+    {
+      "epoch": 1.4898933183604717,
+      "grad_norm": 0.21334858238697052,
+      "learning_rate": 5.926116993489143e-06,
+      "loss": 0.3491,
+      "step": 1769
+    },
+    {
+      "epoch": 1.4907355418304324,
+      "grad_norm": 0.2376968264579773,
+      "learning_rate": 5.921298771419731e-06,
+      "loss": 0.3549,
+      "step": 1770
+    },
+    {
+      "epoch": 1.491577765300393,
+      "grad_norm": 0.2066209763288498,
+      "learning_rate": 5.916479663592107e-06,
+      "loss": 0.3337,
+      "step": 1771
+    },
+    {
+      "epoch": 1.4924199887703538,
+      "grad_norm": 0.21085307002067566,
+      "learning_rate": 5.911659674639473e-06,
+      "loss": 0.3653,
+      "step": 1772
+    },
+    {
+      "epoch": 1.4932622122403143,
+      "grad_norm": 0.22900329530239105,
+      "learning_rate": 5.906838809195879e-06,
+      "loss": 0.382,
+      "step": 1773
+    },
+    {
+      "epoch": 1.494104435710275,
+      "grad_norm": 0.2079591453075409,
+      "learning_rate": 5.90201707189622e-06,
+      "loss": 0.3536,
+      "step": 1774
+    },
+    {
+      "epoch": 1.4949466591802358,
+      "grad_norm": 0.21597476303577423,
+      "learning_rate": 5.897194467376226e-06,
+      "loss": 0.349,
+      "step": 1775
+    },
+    {
+      "epoch": 1.4957888826501966,
+      "grad_norm": 0.2699699401855469,
+      "learning_rate": 5.8923710002724595e-06,
+      "loss": 0.3775,
+      "step": 1776
+    },
+    {
+      "epoch": 1.4966311061201572,
+      "grad_norm": 0.2011500895023346,
+      "learning_rate": 5.887546675222319e-06,
+      "loss": 0.357,
+      "step": 1777
+    },
+    {
+      "epoch": 1.497473329590118,
+      "grad_norm": 0.22972598671913147,
+      "learning_rate": 5.8827214968640215e-06,
+      "loss": 0.3893,
+      "step": 1778
+    },
+    {
+      "epoch": 1.4983155530600787,
+      "grad_norm": 0.22463274002075195,
+      "learning_rate": 5.877895469836604e-06,
+      "loss": 0.3571,
+      "step": 1779
+    },
+    {
+      "epoch": 1.4991577765300392,
+      "grad_norm": 0.2096172422170639,
+      "learning_rate": 5.873068598779926e-06,
+      "loss": 0.3571,
+      "step": 1780
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.21031080186367035,
+      "learning_rate": 5.8682408883346535e-06,
+      "loss": 0.3461,
+      "step": 1781
+    },
+    {
+      "epoch": 1.5008422234699608,
+      "grad_norm": 0.23288646340370178,
+      "learning_rate": 5.863412343142258e-06,
+      "loss": 0.356,
+      "step": 1782
+    },
+    {
+      "epoch": 1.5016844469399215,
+      "grad_norm": 0.23997929692268372,
+      "learning_rate": 5.858582967845018e-06,
+      "loss": 0.3703,
+      "step": 1783
+    },
+    {
+      "epoch": 1.502526670409882,
+      "grad_norm": 0.25117379426956177,
+      "learning_rate": 5.853752767086007e-06,
+      "loss": 0.371,
+      "step": 1784
+    },
+    {
+      "epoch": 1.5033688938798426,
+      "grad_norm": 0.20585308969020844,
+      "learning_rate": 5.848921745509094e-06,
+      "loss": 0.3594,
+      "step": 1785
+    },
+    {
+      "epoch": 1.5042111173498034,
+      "grad_norm": 0.217906191945076,
+      "learning_rate": 5.844089907758935e-06,
+      "loss": 0.3606,
+      "step": 1786
+    },
+    {
+      "epoch": 1.5050533408197642,
+      "grad_norm": 0.2501089870929718,
+      "learning_rate": 5.839257258480974e-06,
+      "loss": 0.3863,
+      "step": 1787
+    },
+    {
+      "epoch": 1.505895564289725,
+      "grad_norm": 0.25611549615859985,
+      "learning_rate": 5.8344238023214305e-06,
+      "loss": 0.3325,
+      "step": 1788
+    },
+    {
+      "epoch": 1.5067377877596857,
+      "grad_norm": 0.26314201951026917,
+      "learning_rate": 5.829589543927305e-06,
+      "loss": 0.3971,
+      "step": 1789
+    },
+    {
+      "epoch": 1.5075800112296462,
+      "grad_norm": 0.19663043320178986,
+      "learning_rate": 5.824754487946366e-06,
+      "loss": 0.3366,
+      "step": 1790
+    },
+    {
+      "epoch": 1.508422234699607,
+      "grad_norm": 0.22584089636802673,
+      "learning_rate": 5.819918639027149e-06,
+      "loss": 0.3916,
+      "step": 1791
+    },
+    {
+      "epoch": 1.5092644581695676,
+      "grad_norm": 0.21082425117492676,
+      "learning_rate": 5.815082001818951e-06,
+      "loss": 0.3225,
+      "step": 1792
+    },
+    {
+      "epoch": 1.5101066816395283,
+      "grad_norm": 0.23728309571743011,
+      "learning_rate": 5.8102445809718325e-06,
+      "loss": 0.3816,
+      "step": 1793
+    },
+    {
+      "epoch": 1.510948905109489,
+      "grad_norm": 0.2071065604686737,
+      "learning_rate": 5.805406381136598e-06,
+      "loss": 0.3668,
+      "step": 1794
+    },
+    {
+      "epoch": 1.5117911285794499,
+      "grad_norm": 0.21563392877578735,
+      "learning_rate": 5.80056740696481e-06,
+      "loss": 0.3587,
+      "step": 1795
+    },
+    {
+      "epoch": 1.5126333520494104,
+      "grad_norm": 0.22941777110099792,
+      "learning_rate": 5.79572766310877e-06,
+      "loss": 0.2965,
+      "step": 1796
+    },
+    {
+      "epoch": 1.5134755755193712,
+      "grad_norm": 0.23965167999267578,
+      "learning_rate": 5.790887154221521e-06,
+      "loss": 0.3838,
+      "step": 1797
+    },
+    {
+      "epoch": 1.5143177989893317,
+      "grad_norm": 0.23982979357242584,
+      "learning_rate": 5.7860458849568425e-06,
+      "loss": 0.3777,
+      "step": 1798
+    },
+    {
+      "epoch": 1.5151600224592925,
+      "grad_norm": 0.20600104331970215,
+      "learning_rate": 5.781203859969242e-06,
+      "loss": 0.3523,
+      "step": 1799
+    },
+    {
+      "epoch": 1.5160022459292533,
+      "grad_norm": 0.24267232418060303,
+      "learning_rate": 5.776361083913959e-06,
+      "loss": 0.3497,
+      "step": 1800
+    },
+    {
+      "epoch": 1.516844469399214,
+      "grad_norm": 0.23190760612487793,
+      "learning_rate": 5.771517561446949e-06,
+      "loss": 0.3207,
+      "step": 1801
+    },
+    {
+      "epoch": 1.5176866928691746,
+      "grad_norm": 0.21379292011260986,
+      "learning_rate": 5.766673297224889e-06,
+      "loss": 0.3448,
+      "step": 1802
+    },
+    {
+      "epoch": 1.5185289163391353,
+      "grad_norm": 0.2329062819480896,
+      "learning_rate": 5.7618282959051685e-06,
+      "loss": 0.3896,
+      "step": 1803
+    },
+    {
+      "epoch": 1.5193711398090959,
+      "grad_norm": 0.21625331044197083,
+      "learning_rate": 5.756982562145884e-06,
+      "loss": 0.3653,
+      "step": 1804
+    },
+    {
+      "epoch": 1.5202133632790567,
+      "grad_norm": 0.20603640377521515,
+      "learning_rate": 5.75213610060584e-06,
+      "loss": 0.3589,
+      "step": 1805
+    },
+    {
+      "epoch": 1.5210555867490174,
+      "grad_norm": 0.20811544358730316,
+      "learning_rate": 5.747288915944533e-06,
+      "loss": 0.3272,
+      "step": 1806
+    },
+    {
+      "epoch": 1.5218978102189782,
+      "grad_norm": 0.21714171767234802,
+      "learning_rate": 5.742441012822166e-06,
+      "loss": 0.38,
+      "step": 1807
+    },
+    {
+      "epoch": 1.522740033688939,
+      "grad_norm": 0.20149259269237518,
+      "learning_rate": 5.737592395899623e-06,
+      "loss": 0.3756,
+      "step": 1808
+    },
+    {
+      "epoch": 1.5235822571588995,
+      "grad_norm": 0.20166918635368347,
+      "learning_rate": 5.7327430698384775e-06,
+      "loss": 0.3489,
+      "step": 1809
+    },
+    {
+      "epoch": 1.52442448062886,
+      "grad_norm": 0.20611999928951263,
+      "learning_rate": 5.727893039300987e-06,
+      "loss": 0.3676,
+      "step": 1810
+    },
+    {
+      "epoch": 1.5252667040988208,
+      "grad_norm": 0.21965749561786652,
+      "learning_rate": 5.7230423089500845e-06,
+      "loss": 0.361,
+      "step": 1811
+    },
+    {
+      "epoch": 1.5261089275687816,
+      "grad_norm": 0.217516228556633,
+      "learning_rate": 5.718190883449373e-06,
+      "loss": 0.3783,
+      "step": 1812
+    },
+    {
+      "epoch": 1.5269511510387423,
+      "grad_norm": 0.19376209378242493,
+      "learning_rate": 5.713338767463129e-06,
+      "loss": 0.3281,
+      "step": 1813
+    },
+    {
+      "epoch": 1.5277933745087031,
+      "grad_norm": 0.22235795855522156,
+      "learning_rate": 5.708485965656291e-06,
+      "loss": 0.3736,
+      "step": 1814
+    },
+    {
+      "epoch": 1.5286355979786637,
+      "grad_norm": 0.2107383906841278,
+      "learning_rate": 5.703632482694453e-06,
+      "loss": 0.3482,
+      "step": 1815
+    },
+    {
+      "epoch": 1.5294778214486242,
+      "grad_norm": 0.24591943621635437,
+      "learning_rate": 5.698778323243871e-06,
+      "loss": 0.3923,
+      "step": 1816
+    },
+    {
+      "epoch": 1.530320044918585,
+      "grad_norm": 0.2085351198911667,
+      "learning_rate": 5.693923491971445e-06,
+      "loss": 0.3583,
+      "step": 1817
+    },
+    {
+      "epoch": 1.5311622683885457,
+      "grad_norm": 0.21462799608707428,
+      "learning_rate": 5.689067993544726e-06,
+      "loss": 0.3653,
+      "step": 1818
+    },
+    {
+      "epoch": 1.5320044918585065,
+      "grad_norm": 0.21209967136383057,
+      "learning_rate": 5.6842118326318996e-06,
+      "loss": 0.385,
+      "step": 1819
+    },
+    {
+      "epoch": 1.5328467153284673,
+      "grad_norm": 0.24777087569236755,
+      "learning_rate": 5.679355013901797e-06,
+      "loss": 0.3857,
+      "step": 1820
+    },
+    {
+      "epoch": 1.5336889387984278,
+      "grad_norm": 0.2122153341770172,
+      "learning_rate": 5.674497542023875e-06,
+      "loss": 0.3491,
+      "step": 1821
+    },
+    {
+      "epoch": 1.5345311622683886,
+      "grad_norm": 0.2072700560092926,
+      "learning_rate": 5.669639421668221e-06,
+      "loss": 0.3753,
+      "step": 1822
+    },
+    {
+      "epoch": 1.5353733857383491,
+      "grad_norm": 0.20608697831630707,
+      "learning_rate": 5.664780657505547e-06,
+      "loss": 0.3453,
+      "step": 1823
+    },
+    {
+      "epoch": 1.53621560920831,
+      "grad_norm": 0.23337386548519135,
+      "learning_rate": 5.659921254207183e-06,
+      "loss": 0.3751,
+      "step": 1824
+    },
+    {
+      "epoch": 1.5370578326782707,
+      "grad_norm": 0.18582665920257568,
+      "learning_rate": 5.65506121644507e-06,
+      "loss": 0.3205,
+      "step": 1825
+    },
+    {
+      "epoch": 1.5379000561482314,
+      "grad_norm": 0.2243811935186386,
+      "learning_rate": 5.650200548891764e-06,
+      "loss": 0.3701,
+      "step": 1826
+    },
+    {
+      "epoch": 1.538742279618192,
+      "grad_norm": 0.1962759792804718,
+      "learning_rate": 5.645339256220427e-06,
+      "loss": 0.3491,
+      "step": 1827
+    },
+    {
+      "epoch": 1.5395845030881528,
+      "grad_norm": 0.20144011080265045,
+      "learning_rate": 5.640477343104815e-06,
+      "loss": 0.3763,
+      "step": 1828
+    },
+    {
+      "epoch": 1.5404267265581133,
+      "grad_norm": 0.19987617433071136,
+      "learning_rate": 5.635614814219289e-06,
+      "loss": 0.3559,
+      "step": 1829
+    },
+    {
+      "epoch": 1.541268950028074,
+      "grad_norm": 0.21206045150756836,
+      "learning_rate": 5.630751674238796e-06,
+      "loss": 0.3374,
+      "step": 1830
+    },
+    {
+      "epoch": 1.5421111734980348,
+      "grad_norm": 0.22243531048297882,
+      "learning_rate": 5.625887927838872e-06,
+      "loss": 0.3783,
+      "step": 1831
+    },
+    {
+      "epoch": 1.5429533969679956,
+      "grad_norm": 0.20843222737312317,
+      "learning_rate": 5.6210235796956395e-06,
+      "loss": 0.3295,
+      "step": 1832
+    },
+    {
+      "epoch": 1.5437956204379562,
+      "grad_norm": 0.2444627583026886,
+      "learning_rate": 5.616158634485793e-06,
+      "loss": 0.3971,
+      "step": 1833
+    },
+    {
+      "epoch": 1.544637843907917,
+      "grad_norm": 0.23421430587768555,
+      "learning_rate": 5.61129309688661e-06,
+      "loss": 0.349,
+      "step": 1834
+    },
+    {
+      "epoch": 1.5454800673778775,
+      "grad_norm": 0.2181040346622467,
+      "learning_rate": 5.606426971575926e-06,
+      "loss": 0.387,
+      "step": 1835
+    },
+    {
+      "epoch": 1.5463222908478382,
+      "grad_norm": 0.18999342620372772,
+      "learning_rate": 5.601560263232153e-06,
+      "loss": 0.355,
+      "step": 1836
+    },
+    {
+      "epoch": 1.547164514317799,
+      "grad_norm": 0.21678003668785095,
+      "learning_rate": 5.596692976534256e-06,
+      "loss": 0.3697,
+      "step": 1837
+    },
+    {
+      "epoch": 1.5480067377877598,
+      "grad_norm": 0.23101647198200226,
+      "learning_rate": 5.591825116161758e-06,
+      "loss": 0.3819,
+      "step": 1838
+    },
+    {
+      "epoch": 1.5488489612577205,
+      "grad_norm": 0.2279675304889679,
+      "learning_rate": 5.5869566867947344e-06,
+      "loss": 0.3661,
+      "step": 1839
+    },
+    {
+      "epoch": 1.549691184727681,
+      "grad_norm": 0.20913884043693542,
+      "learning_rate": 5.582087693113808e-06,
+      "loss": 0.3679,
+      "step": 1840
+    },
+    {
+      "epoch": 1.5505334081976416,
+      "grad_norm": 0.21088729798793793,
+      "learning_rate": 5.577218139800143e-06,
+      "loss": 0.3428,
+      "step": 1841
+    },
+    {
+      "epoch": 1.5513756316676024,
+      "grad_norm": 0.2020246833562851,
+      "learning_rate": 5.572348031535442e-06,
+      "loss": 0.3491,
+      "step": 1842
+    },
+    {
+      "epoch": 1.5522178551375632,
+      "grad_norm": 0.21002264320850372,
+      "learning_rate": 5.567477373001942e-06,
+      "loss": 0.3529,
+      "step": 1843
+    },
+    {
+      "epoch": 1.553060078607524,
+      "grad_norm": 0.2384791225194931,
+      "learning_rate": 5.562606168882404e-06,
+      "loss": 0.3819,
+      "step": 1844
+    },
+    {
+      "epoch": 1.5539023020774847,
+      "grad_norm": 0.2081071436405182,
+      "learning_rate": 5.557734423860122e-06,
+      "loss": 0.3635,
+      "step": 1845
+    },
+    {
+      "epoch": 1.5547445255474452,
+      "grad_norm": 0.1915816217660904,
+      "learning_rate": 5.552862142618906e-06,
+      "loss": 0.3083,
+      "step": 1846
+    },
+    {
+      "epoch": 1.5555867490174058,
+      "grad_norm": 0.20684143900871277,
+      "learning_rate": 5.547989329843079e-06,
+      "loss": 0.3731,
+      "step": 1847
+    },
+    {
+      "epoch": 1.5564289724873666,
+      "grad_norm": 0.23124712705612183,
+      "learning_rate": 5.543115990217478e-06,
+      "loss": 0.3839,
+      "step": 1848
+    },
+    {
+      "epoch": 1.5572711959573273,
+      "grad_norm": 0.25232207775115967,
+      "learning_rate": 5.538242128427444e-06,
+      "loss": 0.3614,
+      "step": 1849
+    },
+    {
+      "epoch": 1.558113419427288,
+      "grad_norm": 0.21806760132312775,
+      "learning_rate": 5.533367749158829e-06,
+      "loss": 0.3809,
+      "step": 1850
+    },
+    {
+      "epoch": 1.5589556428972489,
+      "grad_norm": 0.21748806536197662,
+      "learning_rate": 5.528492857097966e-06,
+      "loss": 0.3599,
+      "step": 1851
+    },
+    {
+      "epoch": 1.5597978663672094,
+      "grad_norm": 0.229938343167305,
+      "learning_rate": 5.523617456931696e-06,
+      "loss": 0.3265,
+      "step": 1852
+    },
+    {
+      "epoch": 1.5606400898371702,
+      "grad_norm": 0.21047593653202057,
+      "learning_rate": 5.518741553347341e-06,
+      "loss": 0.3902,
+      "step": 1853
+    },
+    {
+      "epoch": 1.5614823133071307,
+      "grad_norm": 0.2067413330078125,
+      "learning_rate": 5.513865151032709e-06,
+      "loss": 0.3387,
+      "step": 1854
+    },
+    {
+      "epoch": 1.5623245367770915,
+      "grad_norm": 0.20057551562786102,
+      "learning_rate": 5.508988254676087e-06,
+      "loss": 0.3508,
+      "step": 1855
+    },
+    {
+      "epoch": 1.5631667602470523,
+      "grad_norm": 0.20652174949645996,
+      "learning_rate": 5.504110868966239e-06,
+      "loss": 0.3392,
+      "step": 1856
+    },
+    {
+      "epoch": 1.564008983717013,
+      "grad_norm": 0.200935959815979,
+      "learning_rate": 5.499232998592399e-06,
+      "loss": 0.3447,
+      "step": 1857
+    },
+    {
+      "epoch": 1.5648512071869736,
+      "grad_norm": 0.20910383760929108,
+      "learning_rate": 5.49435464824426e-06,
+      "loss": 0.3653,
+      "step": 1858
+    },
+    {
+      "epoch": 1.5656934306569343,
+      "grad_norm": 0.1886066496372223,
+      "learning_rate": 5.489475822611988e-06,
+      "loss": 0.3388,
+      "step": 1859
+    },
+    {
+      "epoch": 1.5665356541268949,
+      "grad_norm": 0.21101483702659607,
+      "learning_rate": 5.484596526386198e-06,
+      "loss": 0.3786,
+      "step": 1860
+    },
+    {
+      "epoch": 1.5673778775968557,
+      "grad_norm": 0.2070988267660141,
+      "learning_rate": 5.479716764257961e-06,
+      "loss": 0.3624,
+      "step": 1861
+    },
+    {
+      "epoch": 1.5682201010668164,
+      "grad_norm": 0.20134082436561584,
+      "learning_rate": 5.474836540918791e-06,
+      "loss": 0.3558,
+      "step": 1862
+    },
+    {
+      "epoch": 1.5690623245367772,
+      "grad_norm": 0.20253299176692963,
+      "learning_rate": 5.469955861060653e-06,
+      "loss": 0.3575,
+      "step": 1863
+    },
+    {
+      "epoch": 1.5699045480067377,
+      "grad_norm": 0.19125215709209442,
+      "learning_rate": 5.465074729375944e-06,
+      "loss": 0.3635,
+      "step": 1864
+    },
+    {
+      "epoch": 1.5707467714766985,
+      "grad_norm": 0.21241691708564758,
+      "learning_rate": 5.4601931505575e-06,
+      "loss": 0.3698,
+      "step": 1865
+    },
+    {
+      "epoch": 1.571588994946659,
+      "grad_norm": 0.1975466012954712,
+      "learning_rate": 5.455311129298586e-06,
+      "loss": 0.3218,
+      "step": 1866
+    },
+    {
+      "epoch": 1.5724312184166198,
+      "grad_norm": 0.21139660477638245,
+      "learning_rate": 5.450428670292889e-06,
+      "loss": 0.3605,
+      "step": 1867
+    },
+    {
+      "epoch": 1.5732734418865806,
+      "grad_norm": 0.22668103873729706,
+      "learning_rate": 5.445545778234522e-06,
+      "loss": 0.3883,
+      "step": 1868
+    },
+    {
+      "epoch": 1.5741156653565413,
+      "grad_norm": 0.19831670820713043,
+      "learning_rate": 5.44066245781801e-06,
+      "loss": 0.3758,
+      "step": 1869
+    },
+    {
+      "epoch": 1.5749578888265021,
+      "grad_norm": 0.2144181728363037,
+      "learning_rate": 5.435778713738292e-06,
+      "loss": 0.3292,
+      "step": 1870
+    },
+    {
+      "epoch": 1.5758001122964627,
+      "grad_norm": 0.20668131113052368,
+      "learning_rate": 5.430894550690714e-06,
+      "loss": 0.3601,
+      "step": 1871
+    },
+    {
+      "epoch": 1.5766423357664232,
+      "grad_norm": 0.1921137273311615,
+      "learning_rate": 5.426009973371026e-06,
+      "loss": 0.3536,
+      "step": 1872
+    },
+    {
+      "epoch": 1.577484559236384,
+      "grad_norm": 0.1983177214860916,
+      "learning_rate": 5.421124986475371e-06,
+      "loss": 0.3605,
+      "step": 1873
+    },
+    {
+      "epoch": 1.5783267827063447,
+      "grad_norm": 0.211525559425354,
+      "learning_rate": 5.416239594700294e-06,
+      "loss": 0.3763,
+      "step": 1874
+    },
+    {
+      "epoch": 1.5791690061763055,
+      "grad_norm": 0.2196986824274063,
+      "learning_rate": 5.4113538027427245e-06,
+      "loss": 0.387,
+      "step": 1875
+    },
+    {
+      "epoch": 1.5800112296462663,
+      "grad_norm": 0.18906791508197784,
+      "learning_rate": 5.4064676152999765e-06,
+      "loss": 0.3233,
+      "step": 1876
+    },
+    {
+      "epoch": 1.5808534531162268,
+      "grad_norm": 0.1828877031803131,
+      "learning_rate": 5.4015810370697445e-06,
+      "loss": 0.3364,
+      "step": 1877
+    },
+    {
+      "epoch": 1.5816956765861874,
+      "grad_norm": 0.23177126049995422,
+      "learning_rate": 5.396694072750099e-06,
+      "loss": 0.3752,
+      "step": 1878
+    },
+    {
+      "epoch": 1.5825379000561481,
+      "grad_norm": 0.22009316086769104,
+      "learning_rate": 5.391806727039484e-06,
+      "loss": 0.3812,
+      "step": 1879
+    },
+    {
+      "epoch": 1.583380123526109,
+      "grad_norm": 0.19118615984916687,
+      "learning_rate": 5.386919004636706e-06,
+      "loss": 0.3442,
+      "step": 1880
+    },
+    {
+      "epoch": 1.5842223469960697,
+      "grad_norm": 0.20579873025417328,
+      "learning_rate": 5.382030910240936e-06,
+      "loss": 0.3754,
+      "step": 1881
+    },
+    {
+      "epoch": 1.5850645704660304,
+      "grad_norm": 0.18276536464691162,
+      "learning_rate": 5.3771424485517034e-06,
+      "loss": 0.3288,
+      "step": 1882
+    },
+    {
+      "epoch": 1.585906793935991,
+      "grad_norm": 0.23342770338058472,
+      "learning_rate": 5.3722536242688895e-06,
+      "loss": 0.342,
+      "step": 1883
+    },
+    {
+      "epoch": 1.5867490174059518,
+      "grad_norm": 0.20647521317005157,
+      "learning_rate": 5.367364442092724e-06,
+      "loss": 0.3435,
+      "step": 1884
+    },
+    {
+      "epoch": 1.5875912408759123,
+      "grad_norm": 0.2187911421060562,
+      "learning_rate": 5.362474906723781e-06,
+      "loss": 0.3579,
+      "step": 1885
+    },
+    {
+      "epoch": 1.588433464345873,
+      "grad_norm": 0.1964806765317917,
+      "learning_rate": 5.357585022862977e-06,
+      "loss": 0.38,
+      "step": 1886
+    },
+    {
+      "epoch": 1.5892756878158338,
+      "grad_norm": 0.21945969760417938,
+      "learning_rate": 5.352694795211555e-06,
+      "loss": 0.3737,
+      "step": 1887
+    },
+    {
+      "epoch": 1.5901179112857946,
+      "grad_norm": 0.22322909533977509,
+      "learning_rate": 5.347804228471101e-06,
+      "loss": 0.3511,
+      "step": 1888
+    },
+    {
+      "epoch": 1.5909601347557552,
+      "grad_norm": 0.1914677619934082,
+      "learning_rate": 5.342913327343515e-06,
+      "loss": 0.3429,
+      "step": 1889
+    },
+    {
+      "epoch": 1.591802358225716,
+      "grad_norm": 1.419649362564087,
+      "learning_rate": 5.338022096531028e-06,
+      "loss": 0.3701,
+      "step": 1890
+    },
+    {
+      "epoch": 1.5926445816956765,
+      "grad_norm": 0.20444141328334808,
+      "learning_rate": 5.33313054073618e-06,
+      "loss": 0.3279,
+      "step": 1891
+    },
+    {
+      "epoch": 1.5934868051656372,
+      "grad_norm": 0.23807506263256073,
+      "learning_rate": 5.32823866466183e-06,
+      "loss": 0.4132,
+      "step": 1892
+    },
+    {
+      "epoch": 1.594329028635598,
+      "grad_norm": 0.2279578149318695,
+      "learning_rate": 5.3233464730111426e-06,
+      "loss": 0.3564,
+      "step": 1893
+    },
+    {
+      "epoch": 1.5951712521055588,
+      "grad_norm": 0.19392019510269165,
+      "learning_rate": 5.318453970487582e-06,
+      "loss": 0.3329,
+      "step": 1894
+    },
+    {
+      "epoch": 1.5960134755755195,
+      "grad_norm": 0.2214561104774475,
+      "learning_rate": 5.31356116179492e-06,
+      "loss": 0.3618,
+      "step": 1895
+    },
+    {
+      "epoch": 1.59685569904548,
+      "grad_norm": 0.22227728366851807,
+      "learning_rate": 5.308668051637213e-06,
+      "loss": 0.3457,
+      "step": 1896
+    },
+    {
+      "epoch": 1.5976979225154406,
+      "grad_norm": 0.22292588651180267,
+      "learning_rate": 5.303774644718813e-06,
+      "loss": 0.3571,
+      "step": 1897
+    },
+    {
+      "epoch": 1.5985401459854014,
+      "grad_norm": 0.1955156922340393,
+      "learning_rate": 5.298880945744356e-06,
+      "loss": 0.3253,
+      "step": 1898
+    },
+    {
+      "epoch": 1.5993823694553622,
+      "grad_norm": 0.21948570013046265,
+      "learning_rate": 5.29398695941876e-06,
+      "loss": 0.3733,
+      "step": 1899
+    },
+    {
+      "epoch": 1.600224592925323,
+      "grad_norm": 0.22376330196857452,
+      "learning_rate": 5.289092690447215e-06,
+      "loss": 0.3589,
+      "step": 1900
+    },
+    {
+      "epoch": 1.6010668163952837,
+      "grad_norm": 0.20499028265476227,
+      "learning_rate": 5.284198143535188e-06,
+      "loss": 0.3581,
+      "step": 1901
+    },
+    {
+      "epoch": 1.6019090398652442,
+      "grad_norm": 0.20056955516338348,
+      "learning_rate": 5.279303323388413e-06,
+      "loss": 0.3406,
+      "step": 1902
+    },
+    {
+      "epoch": 1.6027512633352048,
+      "grad_norm": 0.2071678787469864,
+      "learning_rate": 5.274408234712881e-06,
+      "loss": 0.3567,
+      "step": 1903
+    },
+    {
+      "epoch": 1.6035934868051656,
+      "grad_norm": 0.20119628310203552,
+      "learning_rate": 5.2695128822148466e-06,
+      "loss": 0.337,
+      "step": 1904
+    },
+    {
+      "epoch": 1.6044357102751263,
+      "grad_norm": 0.2180289775133133,
+      "learning_rate": 5.2646172706008154e-06,
+      "loss": 0.4007,
+      "step": 1905
+    },
+    {
+      "epoch": 1.605277933745087,
+      "grad_norm": 0.1883217990398407,
+      "learning_rate": 5.259721404577546e-06,
+      "loss": 0.3784,
+      "step": 1906
+    },
+    {
+      "epoch": 1.6061201572150479,
+      "grad_norm": 0.21434904634952545,
+      "learning_rate": 5.254825288852033e-06,
+      "loss": 0.3512,
+      "step": 1907
+    },
+    {
+      "epoch": 1.6069623806850084,
+      "grad_norm": 0.22084085643291473,
+      "learning_rate": 5.249928928131523e-06,
+      "loss": 0.3534,
+      "step": 1908
+    },
+    {
+      "epoch": 1.607804604154969,
+      "grad_norm": 0.21422971785068512,
+      "learning_rate": 5.245032327123488e-06,
+      "loss": 0.3967,
+      "step": 1909
+    },
+    {
+      "epoch": 1.6086468276249297,
+      "grad_norm": 0.21373017132282257,
+      "learning_rate": 5.240135490535635e-06,
+      "loss": 0.3509,
+      "step": 1910
+    },
+    {
+      "epoch": 1.6094890510948905,
+      "grad_norm": 0.19849702715873718,
+      "learning_rate": 5.235238423075899e-06,
+      "loss": 0.3481,
+      "step": 1911
+    },
+    {
+      "epoch": 1.6103312745648513,
+      "grad_norm": 0.20527733862400055,
+      "learning_rate": 5.230341129452434e-06,
+      "loss": 0.3659,
+      "step": 1912
+    },
+    {
+      "epoch": 1.611173498034812,
+      "grad_norm": 0.21651127934455872,
+      "learning_rate": 5.225443614373614e-06,
+      "loss": 0.3777,
+      "step": 1913
+    },
+    {
+      "epoch": 1.6120157215047726,
+      "grad_norm": 0.22285105288028717,
+      "learning_rate": 5.220545882548024e-06,
+      "loss": 0.3802,
+      "step": 1914
+    },
+    {
+      "epoch": 1.6128579449747333,
+      "grad_norm": 0.22284671664237976,
+      "learning_rate": 5.215647938684458e-06,
+      "loss": 0.3693,
+      "step": 1915
+    },
+    {
+      "epoch": 1.6137001684446939,
+      "grad_norm": 0.18833483755588531,
+      "learning_rate": 5.210749787491913e-06,
+      "loss": 0.3456,
+      "step": 1916
+    },
+    {
+      "epoch": 1.6145423919146547,
+      "grad_norm": 0.19601304829120636,
+      "learning_rate": 5.20585143367959e-06,
+      "loss": 0.3472,
+      "step": 1917
+    },
+    {
+      "epoch": 1.6153846153846154,
+      "grad_norm": 0.21516947448253632,
+      "learning_rate": 5.200952881956875e-06,
+      "loss": 0.3466,
+      "step": 1918
+    },
+    {
+      "epoch": 1.6162268388545762,
+      "grad_norm": 0.1969214677810669,
+      "learning_rate": 5.196054137033354e-06,
+      "loss": 0.3382,
+      "step": 1919
+    },
+    {
+      "epoch": 1.6170690623245367,
+      "grad_norm": 0.19827716052532196,
+      "learning_rate": 5.191155203618796e-06,
+      "loss": 0.367,
+      "step": 1920
+    },
+    {
+      "epoch": 1.6179112857944975,
+      "grad_norm": 0.19620153307914734,
+      "learning_rate": 5.186256086423148e-06,
+      "loss": 0.3505,
+      "step": 1921
+    },
+    {
+      "epoch": 1.618753509264458,
+      "grad_norm": 0.22013582289218903,
+      "learning_rate": 5.181356790156539e-06,
+      "loss": 0.3871,
+      "step": 1922
+    },
+    {
+      "epoch": 1.6195957327344188,
+      "grad_norm": 0.19816164672374725,
+      "learning_rate": 5.176457319529264e-06,
+      "loss": 0.3493,
+      "step": 1923
+    },
+    {
+      "epoch": 1.6204379562043796,
+      "grad_norm": 0.1959572434425354,
+      "learning_rate": 5.171557679251788e-06,
+      "loss": 0.3681,
+      "step": 1924
+    },
+    {
+      "epoch": 1.6212801796743403,
+      "grad_norm": 0.21222509443759918,
+      "learning_rate": 5.166657874034745e-06,
+      "loss": 0.371,
+      "step": 1925
+    },
+    {
+      "epoch": 1.6221224031443011,
+      "grad_norm": 0.19689340889453888,
+      "learning_rate": 5.161757908588917e-06,
+      "loss": 0.3541,
+      "step": 1926
+    },
+    {
+      "epoch": 1.6229646266142617,
+      "grad_norm": 0.19198477268218994,
+      "learning_rate": 5.156857787625249e-06,
+      "loss": 0.3305,
+      "step": 1927
+    },
+    {
+      "epoch": 1.6238068500842222,
+      "grad_norm": 0.24948351085186005,
+      "learning_rate": 5.15195751585483e-06,
+      "loss": 0.3904,
+      "step": 1928
+    },
+    {
+      "epoch": 1.624649073554183,
+      "grad_norm": 0.20237040519714355,
+      "learning_rate": 5.147057097988898e-06,
+      "loss": 0.3091,
+      "step": 1929
+    },
+    {
+      "epoch": 1.6254912970241437,
+      "grad_norm": 0.20857195556163788,
+      "learning_rate": 5.142156538738827e-06,
+      "loss": 0.3853,
+      "step": 1930
+    },
+    {
+      "epoch": 1.6263335204941045,
+      "grad_norm": 0.2776411175727844,
+      "learning_rate": 5.137255842816132e-06,
+      "loss": 0.3281,
+      "step": 1931
+    },
+    {
+      "epoch": 1.6271757439640653,
+      "grad_norm": 0.191249817609787,
+      "learning_rate": 5.132355014932455e-06,
+      "loss": 0.3557,
+      "step": 1932
+    },
+    {
+      "epoch": 1.6280179674340258,
+      "grad_norm": 0.19988055527210236,
+      "learning_rate": 5.127454059799567e-06,
+      "loss": 0.3846,
+      "step": 1933
+    },
+    {
+      "epoch": 1.6288601909039864,
+      "grad_norm": 0.22994700074195862,
+      "learning_rate": 5.122552982129362e-06,
+      "loss": 0.3628,
+      "step": 1934
+    },
+    {
+      "epoch": 1.6297024143739471,
+      "grad_norm": 0.20393429696559906,
+      "learning_rate": 5.1176517866338495e-06,
+      "loss": 0.329,
+      "step": 1935
+    },
+    {
+      "epoch": 1.630544637843908,
+      "grad_norm": 0.19987967610359192,
+      "learning_rate": 5.112750478025156e-06,
+      "loss": 0.348,
+      "step": 1936
+    },
+    {
+      "epoch": 1.6313868613138687,
+      "grad_norm": 0.1870167851448059,
+      "learning_rate": 5.1078490610155105e-06,
+      "loss": 0.3124,
+      "step": 1937
+    },
+    {
+      "epoch": 1.6322290847838294,
+      "grad_norm": 0.20360225439071655,
+      "learning_rate": 5.102947540317254e-06,
+      "loss": 0.3941,
+      "step": 1938
+    },
+    {
+      "epoch": 1.63307130825379,
+      "grad_norm": 0.1947534829378128,
+      "learning_rate": 5.09804592064282e-06,
+      "loss": 0.3386,
+      "step": 1939
+    },
+    {
+      "epoch": 1.6339135317237508,
+      "grad_norm": 0.21810080111026764,
+      "learning_rate": 5.093144206704743e-06,
+      "loss": 0.3688,
+      "step": 1940
+    },
+    {
+      "epoch": 1.6347557551937113,
+      "grad_norm": 0.21921783685684204,
+      "learning_rate": 5.088242403215644e-06,
+      "loss": 0.3516,
+      "step": 1941
+    },
+    {
+      "epoch": 1.635597978663672,
+      "grad_norm": 0.19858203828334808,
+      "learning_rate": 5.083340514888232e-06,
+      "loss": 0.3498,
+      "step": 1942
+    },
+    {
+      "epoch": 1.6364402021336328,
+      "grad_norm": 0.2206740379333496,
+      "learning_rate": 5.078438546435298e-06,
+      "loss": 0.4053,
+      "step": 1943
+    },
+    {
+      "epoch": 1.6372824256035936,
+      "grad_norm": 0.193548783659935,
+      "learning_rate": 5.073536502569708e-06,
+      "loss": 0.3109,
+      "step": 1944
+    },
+    {
+      "epoch": 1.6381246490735542,
+      "grad_norm": 0.22453932464122772,
+      "learning_rate": 5.0686343880044044e-06,
+      "loss": 0.3971,
+      "step": 1945
+    },
+    {
+      "epoch": 1.638966872543515,
+      "grad_norm": 0.22631575167179108,
+      "learning_rate": 5.063732207452391e-06,
+      "loss": 0.3603,
+      "step": 1946
+    },
+    {
+      "epoch": 1.6398090960134755,
+      "grad_norm": 0.21678605675697327,
+      "learning_rate": 5.058829965626742e-06,
+      "loss": 0.3326,
+      "step": 1947
+    },
+    {
+      "epoch": 1.6406513194834362,
+      "grad_norm": 0.21846479177474976,
+      "learning_rate": 5.053927667240585e-06,
+      "loss": 0.3264,
+      "step": 1948
+    },
+    {
+      "epoch": 1.641493542953397,
+      "grad_norm": 0.21924428641796112,
+      "learning_rate": 5.049025317007108e-06,
+      "loss": 0.3638,
+      "step": 1949
+    },
+    {
+      "epoch": 1.6423357664233578,
+      "grad_norm": 0.21155060827732086,
+      "learning_rate": 5.0441229196395416e-06,
+      "loss": 0.374,
+      "step": 1950
+    },
+    {
+      "epoch": 1.6431779898933183,
+      "grad_norm": 0.21062292158603668,
+      "learning_rate": 5.039220479851167e-06,
+      "loss": 0.3419,
+      "step": 1951
+    },
+    {
+      "epoch": 1.644020213363279,
+      "grad_norm": 0.2159111499786377,
+      "learning_rate": 5.034318002355305e-06,
+      "loss": 0.3643,
+      "step": 1952
+    },
+    {
+      "epoch": 1.6448624368332396,
+      "grad_norm": 0.21283845603466034,
+      "learning_rate": 5.029415491865311e-06,
+      "loss": 0.3666,
+      "step": 1953
+    },
+    {
+      "epoch": 1.6457046603032004,
+      "grad_norm": 0.19281776249408722,
+      "learning_rate": 5.024512953094577e-06,
+      "loss": 0.3391,
+      "step": 1954
+    },
+    {
+      "epoch": 1.6465468837731612,
+      "grad_norm": 0.21010302007198334,
+      "learning_rate": 5.019610390756513e-06,
+      "loss": 0.3065,
+      "step": 1955
+    },
+    {
+      "epoch": 1.647389107243122,
+      "grad_norm": 0.21600353717803955,
+      "learning_rate": 5.014707809564562e-06,
+      "loss": 0.3885,
+      "step": 1956
+    },
+    {
+      "epoch": 1.6482313307130827,
+      "grad_norm": 0.19259744882583618,
+      "learning_rate": 5.009805214232177e-06,
+      "loss": 0.3272,
+      "step": 1957
+    },
+    {
+      "epoch": 1.6490735541830432,
+      "grad_norm": 0.2523140609264374,
+      "learning_rate": 5.004902609472831e-06,
+      "loss": 0.3671,
+      "step": 1958
+    },
+    {
+      "epoch": 1.6499157776530038,
+      "grad_norm": 0.1950521618127823,
+      "learning_rate": 5e-06,
+      "loss": 0.3745,
+      "step": 1959
+    },
+    {
+      "epoch": 1.6507580011229646,
+      "grad_norm": 0.21287277340888977,
+      "learning_rate": 4.995097390527171e-06,
+      "loss": 0.3727,
+      "step": 1960
+    },
+    {
+      "epoch": 1.6516002245929253,
+      "grad_norm": 0.2051495909690857,
+      "learning_rate": 4.990194785767824e-06,
+      "loss": 0.3648,
+      "step": 1961
+    },
+    {
+      "epoch": 1.652442448062886,
+      "grad_norm": 0.20224712789058685,
+      "learning_rate": 4.98529219043544e-06,
+      "loss": 0.3372,
+      "step": 1962
+    },
+    {
+      "epoch": 1.6532846715328469,
+      "grad_norm": 0.21337299048900604,
+      "learning_rate": 4.980389609243488e-06,
+      "loss": 0.3896,
+      "step": 1963
+    },
+    {
+      "epoch": 1.6541268950028074,
+      "grad_norm": 0.20423704385757446,
+      "learning_rate": 4.975487046905426e-06,
+      "loss": 0.3244,
+      "step": 1964
+    },
+    {
+      "epoch": 1.654969118472768,
+      "grad_norm": 0.22124212980270386,
+      "learning_rate": 4.97058450813469e-06,
+      "loss": 0.3616,
+      "step": 1965
+    },
+    {
+      "epoch": 1.6558113419427287,
+      "grad_norm": 0.20289303362369537,
+      "learning_rate": 4.9656819976446975e-06,
+      "loss": 0.3701,
+      "step": 1966
+    },
+    {
+      "epoch": 1.6566535654126895,
+      "grad_norm": 0.2149975448846817,
+      "learning_rate": 4.960779520148835e-06,
+      "loss": 0.3912,
+      "step": 1967
+    },
+    {
+      "epoch": 1.6574957888826503,
+      "grad_norm": 0.18951010704040527,
+      "learning_rate": 4.955877080360462e-06,
+      "loss": 0.3175,
+      "step": 1968
+    },
+    {
+      "epoch": 1.658338012352611,
+      "grad_norm": 0.22500726580619812,
+      "learning_rate": 4.950974682992894e-06,
+      "loss": 0.4012,
+      "step": 1969
+    },
+    {
+      "epoch": 1.6591802358225716,
+      "grad_norm": 0.19183504581451416,
+      "learning_rate": 4.9460723327594175e-06,
+      "loss": 0.3442,
+      "step": 1970
+    },
+    {
+      "epoch": 1.6600224592925323,
+      "grad_norm": 0.18563278019428253,
+      "learning_rate": 4.94117003437326e-06,
+      "loss": 0.345,
+      "step": 1971
+    },
+    {
+      "epoch": 1.6608646827624929,
+      "grad_norm": 0.2061086893081665,
+      "learning_rate": 4.9362677925476124e-06,
+      "loss": 0.3471,
+      "step": 1972
+    },
+    {
+      "epoch": 1.6617069062324537,
+      "grad_norm": 0.21592488884925842,
+      "learning_rate": 4.931365611995598e-06,
+      "loss": 0.374,
+      "step": 1973
+    },
+    {
+      "epoch": 1.6625491297024144,
+      "grad_norm": 0.19126929342746735,
+      "learning_rate": 4.926463497430293e-06,
+      "loss": 0.3394,
+      "step": 1974
+    },
+    {
+      "epoch": 1.6633913531723752,
+      "grad_norm": 0.2073042094707489,
+      "learning_rate": 4.921561453564704e-06,
+      "loss": 0.3551,
+      "step": 1975
+    },
+    {
+      "epoch": 1.6642335766423357,
+      "grad_norm": 0.20550045371055603,
+      "learning_rate": 4.9166594851117696e-06,
+      "loss": 0.349,
+      "step": 1976
+    },
+    {
+      "epoch": 1.6650758001122965,
+      "grad_norm": 0.23410096764564514,
+      "learning_rate": 4.911757596784358e-06,
+      "loss": 0.375,
+      "step": 1977
+    },
+    {
+      "epoch": 1.665918023582257,
+      "grad_norm": 0.19714966416358948,
+      "learning_rate": 4.906855793295259e-06,
+      "loss": 0.3579,
+      "step": 1978
+    },
+    {
+      "epoch": 1.6667602470522178,
+      "grad_norm": 0.21757762134075165,
+      "learning_rate": 4.901954079357182e-06,
+      "loss": 0.2941,
+      "step": 1979
+    },
+    {
+      "epoch": 1.6676024705221786,
+      "grad_norm": 0.2046147584915161,
+      "learning_rate": 4.897052459682749e-06,
+      "loss": 0.3857,
+      "step": 1980
+    },
+    {
+      "epoch": 1.6684446939921393,
+      "grad_norm": 0.21447475254535675,
+      "learning_rate": 4.892150938984491e-06,
+      "loss": 0.3776,
+      "step": 1981
+    },
+    {
+      "epoch": 1.6692869174621,
+      "grad_norm": 0.2030257284641266,
+      "learning_rate": 4.887249521974848e-06,
+      "loss": 0.3783,
+      "step": 1982
+    },
+    {
+      "epoch": 1.6701291409320607,
+      "grad_norm": 0.1929410845041275,
+      "learning_rate": 4.882348213366152e-06,
+      "loss": 0.3402,
+      "step": 1983
+    },
+    {
+      "epoch": 1.6709713644020212,
+      "grad_norm": 0.2044433206319809,
+      "learning_rate": 4.8774470178706405e-06,
+      "loss": 0.3749,
+      "step": 1984
+    },
+    {
+      "epoch": 1.671813587871982,
+      "grad_norm": 0.18149399757385254,
+      "learning_rate": 4.872545940200435e-06,
+      "loss": 0.3396,
+      "step": 1985
+    },
+    {
+      "epoch": 1.6726558113419427,
+      "grad_norm": 0.22244606912136078,
+      "learning_rate": 4.867644985067548e-06,
+      "loss": 0.3704,
+      "step": 1986
+    },
+    {
+      "epoch": 1.6734980348119035,
+      "grad_norm": 0.2242031842470169,
+      "learning_rate": 4.862744157183869e-06,
+      "loss": 0.379,
+      "step": 1987
+    },
+    {
+      "epoch": 1.6743402582818643,
+      "grad_norm": 0.22645780444145203,
+      "learning_rate": 4.857843461261176e-06,
+      "loss": 0.3882,
+      "step": 1988
+    },
+    {
+      "epoch": 1.6751824817518248,
+      "grad_norm": 0.18309590220451355,
+      "learning_rate": 4.8529429020111035e-06,
+      "loss": 0.3464,
+      "step": 1989
+    },
+    {
+      "epoch": 1.6760247052217854,
+      "grad_norm": 0.19142934679985046,
+      "learning_rate": 4.8480424841451725e-06,
+      "loss": 0.3794,
+      "step": 1990
+    },
+    {
+      "epoch": 1.6768669286917461,
+      "grad_norm": 0.2454836666584015,
+      "learning_rate": 4.8431422123747524e-06,
+      "loss": 0.3674,
+      "step": 1991
+    },
+    {
+      "epoch": 1.677709152161707,
+      "grad_norm": 0.20532433688640594,
+      "learning_rate": 4.838242091411085e-06,
+      "loss": 0.3232,
+      "step": 1992
+    },
+    {
+      "epoch": 1.6785513756316677,
+      "grad_norm": 0.21484729647636414,
+      "learning_rate": 4.833342125965257e-06,
+      "loss": 0.3612,
+      "step": 1993
+    },
+    {
+      "epoch": 1.6793935991016284,
+      "grad_norm": 0.20386093854904175,
+      "learning_rate": 4.828442320748213e-06,
+      "loss": 0.3322,
+      "step": 1994
+    },
+    {
+      "epoch": 1.680235822571589,
+      "grad_norm": 0.20775291323661804,
+      "learning_rate": 4.823542680470738e-06,
+      "loss": 0.329,
+      "step": 1995
+    },
+    {
+      "epoch": 1.6810780460415495,
+      "grad_norm": 0.2114599198102951,
+      "learning_rate": 4.818643209843463e-06,
+      "loss": 0.3909,
+      "step": 1996
+    },
+    {
+      "epoch": 1.6819202695115103,
+      "grad_norm": 0.19673332571983337,
+      "learning_rate": 4.813743913576852e-06,
+      "loss": 0.3436,
+      "step": 1997
+    },
+    {
+      "epoch": 1.682762492981471,
+      "grad_norm": 0.2197994440793991,
+      "learning_rate": 4.808844796381205e-06,
+      "loss": 0.3784,
+      "step": 1998
+    },
+    {
+      "epoch": 1.6836047164514318,
+      "grad_norm": 0.18247802555561066,
+      "learning_rate": 4.803945862966646e-06,
+      "loss": 0.3212,
+      "step": 1999
+    },
+    {
+      "epoch": 1.6844469399213926,
+      "grad_norm": 0.20470093190670013,
+      "learning_rate": 4.799047118043126e-06,
+      "loss": 0.3754,
+      "step": 2000
+    },
+    {
+      "epoch": 1.6852891633913532,
+      "grad_norm": 0.19941620528697968,
+      "learning_rate": 4.794148566320412e-06,
+      "loss": 0.3699,
+      "step": 2001
+    },
+    {
+      "epoch": 1.686131386861314,
+      "grad_norm": 0.18958649039268494,
+      "learning_rate": 4.789250212508088e-06,
+      "loss": 0.3426,
+      "step": 2002
+    },
+    {
+      "epoch": 1.6869736103312745,
+      "grad_norm": 0.2122664898633957,
+      "learning_rate": 4.7843520613155434e-06,
+      "loss": 0.3775,
+      "step": 2003
+    },
+    {
+      "epoch": 1.6878158338012352,
+      "grad_norm": 0.198108971118927,
+      "learning_rate": 4.779454117451978e-06,
+      "loss": 0.3637,
+      "step": 2004
+    },
+    {
+      "epoch": 1.688658057271196,
+      "grad_norm": 0.1983402669429779,
+      "learning_rate": 4.774556385626386e-06,
+      "loss": 0.3666,
+      "step": 2005
+    },
+    {
+      "epoch": 1.6895002807411568,
+      "grad_norm": 0.1770869344472885,
+      "learning_rate": 4.769658870547567e-06,
+      "loss": 0.3214,
+      "step": 2006
+    },
+    {
+      "epoch": 1.6903425042111173,
+      "grad_norm": 0.19760024547576904,
+      "learning_rate": 4.7647615769241e-06,
+      "loss": 0.3505,
+      "step": 2007
+    },
+    {
+      "epoch": 1.691184727681078,
+      "grad_norm": 0.1891362965106964,
+      "learning_rate": 4.759864509464366e-06,
+      "loss": 0.344,
+      "step": 2008
+    },
+    {
+      "epoch": 1.6920269511510386,
+      "grad_norm": 0.19179244339466095,
+      "learning_rate": 4.754967672876513e-06,
+      "loss": 0.3583,
+      "step": 2009
+    },
+    {
+      "epoch": 1.6928691746209994,
+      "grad_norm": 0.21558877825737,
+      "learning_rate": 4.750071071868478e-06,
+      "loss": 0.3713,
+      "step": 2010
+    },
+    {
+      "epoch": 1.6937113980909602,
+      "grad_norm": 0.20262889564037323,
+      "learning_rate": 4.745174711147967e-06,
+      "loss": 0.357,
+      "step": 2011
+    },
+    {
+      "epoch": 1.694553621560921,
+      "grad_norm": 0.21164673566818237,
+      "learning_rate": 4.7402785954224565e-06,
+      "loss": 0.37,
+      "step": 2012
+    },
+    {
+      "epoch": 1.6953958450308815,
+      "grad_norm": 0.20908448100090027,
+      "learning_rate": 4.7353827293991845e-06,
+      "loss": 0.368,
+      "step": 2013
+    },
+    {
+      "epoch": 1.6962380685008422,
+      "grad_norm": 0.2034013271331787,
+      "learning_rate": 4.730487117785155e-06,
+      "loss": 0.3431,
+      "step": 2014
+    },
+    {
+      "epoch": 1.6970802919708028,
+      "grad_norm": 0.22839775681495667,
+      "learning_rate": 4.725591765287119e-06,
+      "loss": 0.3753,
+      "step": 2015
+    },
+    {
+      "epoch": 1.6979225154407636,
+      "grad_norm": 0.22359232604503632,
+      "learning_rate": 4.720696676611589e-06,
+      "loss": 0.3214,
+      "step": 2016
+    },
+    {
+      "epoch": 1.6987647389107243,
+      "grad_norm": 0.22633424401283264,
+      "learning_rate": 4.715801856464812e-06,
+      "loss": 0.3904,
+      "step": 2017
+    },
+    {
+      "epoch": 1.699606962380685,
+      "grad_norm": 0.20711787045001984,
+      "learning_rate": 4.710907309552787e-06,
+      "loss": 0.3368,
+      "step": 2018
+    },
+    {
+      "epoch": 1.7004491858506459,
+      "grad_norm": 0.23692451417446136,
+      "learning_rate": 4.706013040581242e-06,
+      "loss": 0.375,
+      "step": 2019
+    },
+    {
+      "epoch": 1.7012914093206064,
+      "grad_norm": 0.2271483689546585,
+      "learning_rate": 4.701119054255646e-06,
+      "loss": 0.3536,
+      "step": 2020
+    },
+    {
+      "epoch": 1.702133632790567,
+      "grad_norm": 0.21336886286735535,
+      "learning_rate": 4.6962253552811885e-06,
+      "loss": 0.3488,
+      "step": 2021
+    },
+    {
+      "epoch": 1.7029758562605277,
+      "grad_norm": 0.1952803134918213,
+      "learning_rate": 4.691331948362789e-06,
+      "loss": 0.3107,
+      "step": 2022
+    },
+    {
+      "epoch": 1.7038180797304885,
+      "grad_norm": 0.2554273307323456,
+      "learning_rate": 4.6864388382050804e-06,
+      "loss": 0.3854,
+      "step": 2023
+    },
+    {
+      "epoch": 1.7046603032004493,
+      "grad_norm": 0.21529453992843628,
+      "learning_rate": 4.6815460295124185e-06,
+      "loss": 0.3762,
+      "step": 2024
+    },
+    {
+      "epoch": 1.70550252667041,
+      "grad_norm": 0.199816033244133,
+      "learning_rate": 4.676653526988858e-06,
+      "loss": 0.3406,
+      "step": 2025
+    },
+    {
+      "epoch": 1.7063447501403706,
+      "grad_norm": 0.2005997598171234,
+      "learning_rate": 4.671761335338171e-06,
+      "loss": 0.3618,
+      "step": 2026
+    },
+    {
+      "epoch": 1.7071869736103311,
+      "grad_norm": 0.22079318761825562,
+      "learning_rate": 4.666869459263821e-06,
+      "loss": 0.3761,
+      "step": 2027
+    },
+    {
+      "epoch": 1.7080291970802919,
+      "grad_norm": 0.21227099001407623,
+      "learning_rate": 4.661977903468974e-06,
+      "loss": 0.3523,
+      "step": 2028
+    },
+    {
+      "epoch": 1.7088714205502527,
+      "grad_norm": 0.19983810186386108,
+      "learning_rate": 4.657086672656486e-06,
+      "loss": 0.3221,
+      "step": 2029
+    },
+    {
+      "epoch": 1.7097136440202134,
+      "grad_norm": 0.2068592607975006,
+      "learning_rate": 4.652195771528901e-06,
+      "loss": 0.3619,
+      "step": 2030
+    },
+    {
+      "epoch": 1.7105558674901742,
+      "grad_norm": 0.21467489004135132,
+      "learning_rate": 4.647305204788445e-06,
+      "loss": 0.3632,
+      "step": 2031
+    },
+    {
+      "epoch": 1.7113980909601347,
+      "grad_norm": 0.20065774023532867,
+      "learning_rate": 4.642414977137026e-06,
+      "loss": 0.3784,
+      "step": 2032
+    },
+    {
+      "epoch": 1.7122403144300955,
+      "grad_norm": 0.19884279370307922,
+      "learning_rate": 4.63752509327622e-06,
+      "loss": 0.3419,
+      "step": 2033
+    },
+    {
+      "epoch": 1.713082537900056,
+      "grad_norm": 0.2171352505683899,
+      "learning_rate": 4.632635557907277e-06,
+      "loss": 0.352,
+      "step": 2034
+    },
+    {
+      "epoch": 1.7139247613700168,
+      "grad_norm": 0.22259283065795898,
+      "learning_rate": 4.627746375731112e-06,
+      "loss": 0.4014,
+      "step": 2035
+    },
+    {
+      "epoch": 1.7147669848399776,
+      "grad_norm": 0.22084489464759827,
+      "learning_rate": 4.622857551448297e-06,
+      "loss": 0.3455,
+      "step": 2036
+    },
+    {
+      "epoch": 1.7156092083099383,
+      "grad_norm": 0.2158217579126358,
+      "learning_rate": 4.617969089759066e-06,
+      "loss": 0.3451,
+      "step": 2037
+    },
+    {
+      "epoch": 1.716451431779899,
+      "grad_norm": 0.22856098413467407,
+      "learning_rate": 4.613080995363296e-06,
+      "loss": 0.3886,
+      "step": 2038
+    },
+    {
+      "epoch": 1.7172936552498597,
+      "grad_norm": 0.2910906970500946,
+      "learning_rate": 4.608193272960519e-06,
+      "loss": 0.3484,
+      "step": 2039
+    },
+    {
+      "epoch": 1.7181358787198202,
+      "grad_norm": 0.21710596978664398,
+      "learning_rate": 4.603305927249902e-06,
+      "loss": 0.3511,
+      "step": 2040
+    },
+    {
+      "epoch": 1.718978102189781,
+      "grad_norm": 0.21317730844020844,
+      "learning_rate": 4.598418962930258e-06,
+      "loss": 0.3707,
+      "step": 2041
+    },
+    {
+      "epoch": 1.7198203256597417,
+      "grad_norm": 0.20781542360782623,
+      "learning_rate": 4.593532384700026e-06,
+      "loss": 0.4028,
+      "step": 2042
+    },
+    {
+      "epoch": 1.7206625491297025,
+      "grad_norm": 0.2237594872713089,
+      "learning_rate": 4.588646197257278e-06,
+      "loss": 0.3914,
+      "step": 2043
+    },
+    {
+      "epoch": 1.721504772599663,
+      "grad_norm": 0.18635544180870056,
+      "learning_rate": 4.583760405299707e-06,
+      "loss": 0.3365,
+      "step": 2044
+    },
+    {
+      "epoch": 1.7223469960696238,
+      "grad_norm": 0.20491984486579895,
+      "learning_rate": 4.57887501352463e-06,
+      "loss": 0.3303,
+      "step": 2045
+    },
+    {
+      "epoch": 1.7231892195395844,
+      "grad_norm": 0.1990174651145935,
+      "learning_rate": 4.573990026628976e-06,
+      "loss": 0.3646,
+      "step": 2046
+    },
+    {
+      "epoch": 1.7240314430095451,
+      "grad_norm": 0.1963694989681244,
+      "learning_rate": 4.569105449309289e-06,
+      "loss": 0.3729,
+      "step": 2047
+    },
+    {
+      "epoch": 1.724873666479506,
+      "grad_norm": 0.18767237663269043,
+      "learning_rate": 4.564221286261709e-06,
+      "loss": 0.3316,
+      "step": 2048
+    },
+    {
+      "epoch": 1.7257158899494667,
+      "grad_norm": 0.23530213534832,
+      "learning_rate": 4.559337542181993e-06,
+      "loss": 0.3956,
+      "step": 2049
+    },
+    {
+      "epoch": 1.7265581134194274,
+      "grad_norm": 0.18765990436077118,
+      "learning_rate": 4.554454221765479e-06,
+      "loss": 0.3267,
+      "step": 2050
+    },
+    {
+      "epoch": 1.727400336889388,
+      "grad_norm": 0.22365470230579376,
+      "learning_rate": 4.549571329707113e-06,
+      "loss": 0.3459,
+      "step": 2051
+    },
+    {
+      "epoch": 1.7282425603593485,
+      "grad_norm": 0.21409985423088074,
+      "learning_rate": 4.544688870701416e-06,
+      "loss": 0.3421,
+      "step": 2052
+    },
+    {
+      "epoch": 1.7290847838293093,
+      "grad_norm": 0.21623964607715607,
+      "learning_rate": 4.539806849442501e-06,
+      "loss": 0.3462,
+      "step": 2053
+    },
+    {
+      "epoch": 1.72992700729927,
+      "grad_norm": 0.22499880194664001,
+      "learning_rate": 4.534925270624057e-06,
+      "loss": 0.3901,
+      "step": 2054
+    },
+    {
+      "epoch": 1.7307692307692308,
+      "grad_norm": 0.19443631172180176,
+      "learning_rate": 4.53004413893935e-06,
+      "loss": 0.3437,
+      "step": 2055
+    },
+    {
+      "epoch": 1.7316114542391916,
+      "grad_norm": 0.21552076935768127,
+      "learning_rate": 4.52516345908121e-06,
+      "loss": 0.3848,
+      "step": 2056
+    },
+    {
+      "epoch": 1.7324536777091522,
+      "grad_norm": 0.2129429429769516,
+      "learning_rate": 4.520283235742042e-06,
+      "loss": 0.3924,
+      "step": 2057
+    },
+    {
+      "epoch": 1.7332959011791127,
+      "grad_norm": 0.1930941939353943,
+      "learning_rate": 4.5154034736138035e-06,
+      "loss": 0.3357,
+      "step": 2058
+    },
+    {
+      "epoch": 1.7341381246490735,
+      "grad_norm": 0.20328223705291748,
+      "learning_rate": 4.510524177388014e-06,
+      "loss": 0.3711,
+      "step": 2059
+    },
+    {
+      "epoch": 1.7349803481190342,
+      "grad_norm": 0.1964006870985031,
+      "learning_rate": 4.505645351755741e-06,
+      "loss": 0.3599,
+      "step": 2060
+    },
+    {
+      "epoch": 1.735822571588995,
+      "grad_norm": 0.19247199594974518,
+      "learning_rate": 4.500767001407604e-06,
+      "loss": 0.3902,
+      "step": 2061
+    },
+    {
+      "epoch": 1.7366647950589558,
+      "grad_norm": 0.19466297328472137,
+      "learning_rate": 4.495889131033762e-06,
+      "loss": 0.351,
+      "step": 2062
+    },
+    {
+      "epoch": 1.7375070185289163,
+      "grad_norm": 0.20538896322250366,
+      "learning_rate": 4.491011745323914e-06,
+      "loss": 0.3439,
+      "step": 2063
+    },
+    {
+      "epoch": 1.738349241998877,
+      "grad_norm": 0.21315395832061768,
+      "learning_rate": 4.486134848967292e-06,
+      "loss": 0.3412,
+      "step": 2064
+    },
+    {
+      "epoch": 1.7391914654688376,
+      "grad_norm": 0.20313483476638794,
+      "learning_rate": 4.481258446652662e-06,
+      "loss": 0.3713,
+      "step": 2065
+    },
+    {
+      "epoch": 1.7400336889387984,
+      "grad_norm": 0.19733814895153046,
+      "learning_rate": 4.4763825430683055e-06,
+      "loss": 0.3602,
+      "step": 2066
+    },
+    {
+      "epoch": 1.7408759124087592,
+      "grad_norm": 0.19734057784080505,
+      "learning_rate": 4.471507142902036e-06,
+      "loss": 0.3461,
+      "step": 2067
+    },
+    {
+      "epoch": 1.74171813587872,
+      "grad_norm": 0.20122112333774567,
+      "learning_rate": 4.466632250841173e-06,
+      "loss": 0.3566,
+      "step": 2068
+    },
+    {
+      "epoch": 1.7425603593486805,
+      "grad_norm": 0.1987045258283615,
+      "learning_rate": 4.4617578715725565e-06,
+      "loss": 0.3677,
+      "step": 2069
+    },
+    {
+      "epoch": 1.7434025828186412,
+      "grad_norm": 0.2019500732421875,
+      "learning_rate": 4.4568840097825225e-06,
+      "loss": 0.3665,
+      "step": 2070
+    },
+    {
+      "epoch": 1.7442448062886018,
+      "grad_norm": 0.1923629343509674,
+      "learning_rate": 4.452010670156922e-06,
+      "loss": 0.3234,
+      "step": 2071
+    },
+    {
+      "epoch": 1.7450870297585626,
+      "grad_norm": 0.22685682773590088,
+      "learning_rate": 4.447137857381095e-06,
+      "loss": 0.3724,
+      "step": 2072
+    },
+    {
+      "epoch": 1.7459292532285233,
+      "grad_norm": 0.19463816285133362,
+      "learning_rate": 4.4422655761398785e-06,
+      "loss": 0.3531,
+      "step": 2073
+    },
+    {
+      "epoch": 1.746771476698484,
+      "grad_norm": 0.22609847784042358,
+      "learning_rate": 4.437393831117596e-06,
+      "loss": 0.4006,
+      "step": 2074
+    },
+    {
+      "epoch": 1.7476137001684446,
+      "grad_norm": 0.2065419703722,
+      "learning_rate": 4.432522626998061e-06,
+      "loss": 0.3708,
+      "step": 2075
+    },
+    {
+      "epoch": 1.7484559236384054,
+      "grad_norm": 0.18311607837677002,
+      "learning_rate": 4.427651968464559e-06,
+      "loss": 0.3563,
+      "step": 2076
+    },
+    {
+      "epoch": 1.749298147108366,
+      "grad_norm": 0.18132922053337097,
+      "learning_rate": 4.4227818601998575e-06,
+      "loss": 0.3192,
+      "step": 2077
+    },
+    {
+      "epoch": 1.7501403705783267,
+      "grad_norm": 0.2064925730228424,
+      "learning_rate": 4.417912306886192e-06,
+      "loss": 0.3578,
+      "step": 2078
+    },
+    {
+      "epoch": 1.7509825940482875,
+      "grad_norm": 0.21537601947784424,
+      "learning_rate": 4.413043313205266e-06,
+      "loss": 0.3696,
+      "step": 2079
+    },
+    {
+      "epoch": 1.7518248175182483,
+      "grad_norm": 0.1957070529460907,
+      "learning_rate": 4.408174883838243e-06,
+      "loss": 0.3298,
+      "step": 2080
+    },
+    {
+      "epoch": 1.752667040988209,
+      "grad_norm": 0.2095799446105957,
+      "learning_rate": 4.403307023465746e-06,
+      "loss": 0.3488,
+      "step": 2081
+    },
+    {
+      "epoch": 1.7535092644581696,
+      "grad_norm": 0.2147350013256073,
+      "learning_rate": 4.3984397367678475e-06,
+      "loss": 0.3798,
+      "step": 2082
+    },
+    {
+      "epoch": 1.7543514879281301,
+      "grad_norm": 0.2071705162525177,
+      "learning_rate": 4.393573028424075e-06,
+      "loss": 0.3623,
+      "step": 2083
+    },
+    {
+      "epoch": 1.7551937113980909,
+      "grad_norm": 0.2035631388425827,
+      "learning_rate": 4.388706903113391e-06,
+      "loss": 0.3553,
+      "step": 2084
+    },
+    {
+      "epoch": 1.7560359348680517,
+      "grad_norm": 0.20198489725589752,
+      "learning_rate": 4.383841365514208e-06,
+      "loss": 0.3733,
+      "step": 2085
+    },
+    {
+      "epoch": 1.7568781583380124,
+      "grad_norm": 0.2736494541168213,
+      "learning_rate": 4.378976420304361e-06,
+      "loss": 0.3946,
+      "step": 2086
+    },
+    {
+      "epoch": 1.7577203818079732,
+      "grad_norm": 0.21576163172721863,
+      "learning_rate": 4.374112072161129e-06,
+      "loss": 0.357,
+      "step": 2087
+    },
+    {
+      "epoch": 1.7585626052779337,
+      "grad_norm": 0.2382166087627411,
+      "learning_rate": 4.369248325761205e-06,
+      "loss": 0.3318,
+      "step": 2088
+    },
+    {
+      "epoch": 1.7594048287478943,
+      "grad_norm": 0.21679919958114624,
+      "learning_rate": 4.364385185780712e-06,
+      "loss": 0.3622,
+      "step": 2089
+    },
+    {
+      "epoch": 1.760247052217855,
+      "grad_norm": 0.1898692101240158,
+      "learning_rate": 4.359522656895185e-06,
+      "loss": 0.2974,
+      "step": 2090
+    },
+    {
+      "epoch": 1.7610892756878158,
+      "grad_norm": 0.2273748368024826,
+      "learning_rate": 4.354660743779575e-06,
+      "loss": 0.3993,
+      "step": 2091
+    },
+    {
+      "epoch": 1.7619314991577766,
+      "grad_norm": 0.23613716661930084,
+      "learning_rate": 4.349799451108236e-06,
+      "loss": 0.3663,
+      "step": 2092
+    },
+    {
+      "epoch": 1.7627737226277373,
+      "grad_norm": 0.20020337402820587,
+      "learning_rate": 4.3449387835549305e-06,
+      "loss": 0.3606,
+      "step": 2093
+    },
+    {
+      "epoch": 1.763615946097698,
+      "grad_norm": 0.18953366577625275,
+      "learning_rate": 4.340078745792818e-06,
+      "loss": 0.3435,
+      "step": 2094
+    },
+    {
+      "epoch": 1.7644581695676587,
+      "grad_norm": 0.20416444540023804,
+      "learning_rate": 4.3352193424944535e-06,
+      "loss": 0.3636,
+      "step": 2095
+    },
+    {
+      "epoch": 1.7653003930376192,
+      "grad_norm": 0.19791902601718903,
+      "learning_rate": 4.3303605783317794e-06,
+      "loss": 0.3573,
+      "step": 2096
+    },
+    {
+      "epoch": 1.76614261650758,
+      "grad_norm": 0.22840075194835663,
+      "learning_rate": 4.325502457976126e-06,
+      "loss": 0.3499,
+      "step": 2097
+    },
+    {
+      "epoch": 1.7669848399775407,
+      "grad_norm": 0.20278143882751465,
+      "learning_rate": 4.320644986098204e-06,
+      "loss": 0.396,
+      "step": 2098
+    },
+    {
+      "epoch": 1.7678270634475015,
+      "grad_norm": 2.3698928356170654,
+      "learning_rate": 4.315788167368102e-06,
+      "loss": 0.3408,
+      "step": 2099
+    },
+    {
+      "epoch": 1.768669286917462,
+      "grad_norm": 0.21950553357601166,
+      "learning_rate": 4.310932006455276e-06,
+      "loss": 0.3197,
+      "step": 2100
+    },
+    {
+      "epoch": 1.7695115103874228,
+      "grad_norm": 0.19593040645122528,
+      "learning_rate": 4.306076508028557e-06,
+      "loss": 0.344,
+      "step": 2101
+    },
+    {
+      "epoch": 1.7703537338573834,
+      "grad_norm": 0.2054298222064972,
+      "learning_rate": 4.301221676756129e-06,
+      "loss": 0.3511,
+      "step": 2102
+    },
+    {
+      "epoch": 1.7711959573273441,
+      "grad_norm": 0.20474767684936523,
+      "learning_rate": 4.296367517305548e-06,
+      "loss": 0.3553,
+      "step": 2103
+    },
+    {
+      "epoch": 1.772038180797305,
+      "grad_norm": 0.1904682219028473,
+      "learning_rate": 4.29151403434371e-06,
+      "loss": 0.322,
+      "step": 2104
+    },
+    {
+      "epoch": 1.7728804042672657,
+      "grad_norm": 0.21345622837543488,
+      "learning_rate": 4.286661232536873e-06,
+      "loss": 0.3496,
+      "step": 2105
+    },
+    {
+      "epoch": 1.7737226277372264,
+      "grad_norm": 0.2012915164232254,
+      "learning_rate": 4.281809116550629e-06,
+      "loss": 0.3608,
+      "step": 2106
+    },
+    {
+      "epoch": 1.774564851207187,
+      "grad_norm": 0.18038402497768402,
+      "learning_rate": 4.276957691049917e-06,
+      "loss": 0.3155,
+      "step": 2107
+    },
+    {
+      "epoch": 1.7754070746771475,
+      "grad_norm": 0.20159615576267242,
+      "learning_rate": 4.272106960699015e-06,
+      "loss": 0.392,
+      "step": 2108
+    },
+    {
+      "epoch": 1.7762492981471083,
+      "grad_norm": 0.19055233895778656,
+      "learning_rate": 4.267256930161523e-06,
+      "loss": 0.3248,
+      "step": 2109
+    },
+    {
+      "epoch": 1.777091521617069,
+      "grad_norm": 0.20288807153701782,
+      "learning_rate": 4.2624076041003794e-06,
+      "loss": 0.3645,
+      "step": 2110
+    },
+    {
+      "epoch": 1.7779337450870298,
+      "grad_norm": 0.19328925013542175,
+      "learning_rate": 4.257558987177835e-06,
+      "loss": 0.3432,
+      "step": 2111
+    },
+    {
+      "epoch": 1.7787759685569906,
+      "grad_norm": 0.20823122560977936,
+      "learning_rate": 4.252711084055468e-06,
+      "loss": 0.3514,
+      "step": 2112
+    },
+    {
+      "epoch": 1.7796181920269512,
+      "grad_norm": 0.1911734789609909,
+      "learning_rate": 4.247863899394162e-06,
+      "loss": 0.3436,
+      "step": 2113
+    },
+    {
+      "epoch": 1.7804604154969117,
+      "grad_norm": 0.2023228406906128,
+      "learning_rate": 4.243017437854117e-06,
+      "loss": 0.38,
+      "step": 2114
+    },
+    {
+      "epoch": 1.7813026389668725,
+      "grad_norm": 0.19723080098628998,
+      "learning_rate": 4.238171704094833e-06,
+      "loss": 0.3404,
+      "step": 2115
+    },
+    {
+      "epoch": 1.7821448624368332,
+      "grad_norm": 0.19566403329372406,
+      "learning_rate": 4.2333267027751125e-06,
+      "loss": 0.339,
+      "step": 2116
+    },
+    {
+      "epoch": 1.782987085906794,
+      "grad_norm": 0.23212406039237976,
+      "learning_rate": 4.228482438553052e-06,
+      "loss": 0.3674,
+      "step": 2117
+    },
+    {
+      "epoch": 1.7838293093767548,
+      "grad_norm": 0.1956307590007782,
+      "learning_rate": 4.223638916086044e-06,
+      "loss": 0.3553,
+      "step": 2118
+    },
+    {
+      "epoch": 1.7846715328467153,
+      "grad_norm": 0.18204696476459503,
+      "learning_rate": 4.218796140030759e-06,
+      "loss": 0.3242,
+      "step": 2119
+    },
+    {
+      "epoch": 1.7855137563166759,
+      "grad_norm": 0.1948997676372528,
+      "learning_rate": 4.21395411504316e-06,
+      "loss": 0.3601,
+      "step": 2120
+    },
+    {
+      "epoch": 1.7863559797866366,
+      "grad_norm": 0.1899755448102951,
+      "learning_rate": 4.209112845778481e-06,
+      "loss": 0.3525,
+      "step": 2121
+    },
+    {
+      "epoch": 1.7871982032565974,
+      "grad_norm": 0.212238147854805,
+      "learning_rate": 4.204272336891232e-06,
+      "loss": 0.353,
+      "step": 2122
+    },
+    {
+      "epoch": 1.7880404267265582,
+      "grad_norm": 0.1929008811712265,
+      "learning_rate": 4.199432593035192e-06,
+      "loss": 0.3801,
+      "step": 2123
+    },
+    {
+      "epoch": 1.788882650196519,
+      "grad_norm": 0.1923249065876007,
+      "learning_rate": 4.194593618863404e-06,
+      "loss": 0.3367,
+      "step": 2124
+    },
+    {
+      "epoch": 1.7897248736664795,
+      "grad_norm": 0.21146802604198456,
+      "learning_rate": 4.189755419028169e-06,
+      "loss": 0.3914,
+      "step": 2125
+    },
+    {
+      "epoch": 1.7905670971364402,
+      "grad_norm": 0.21049576997756958,
+      "learning_rate": 4.1849179981810506e-06,
+      "loss": 0.3775,
+      "step": 2126
+    },
+    {
+      "epoch": 1.7914093206064008,
+      "grad_norm": 0.18470361828804016,
+      "learning_rate": 4.180081360972852e-06,
+      "loss": 0.3293,
+      "step": 2127
+    },
+    {
+      "epoch": 1.7922515440763616,
+      "grad_norm": 0.19915157556533813,
+      "learning_rate": 4.175245512053637e-06,
+      "loss": 0.3579,
+      "step": 2128
+    },
+    {
+      "epoch": 1.7930937675463223,
+      "grad_norm": 0.19717763364315033,
+      "learning_rate": 4.1704104560726955e-06,
+      "loss": 0.3564,
+      "step": 2129
+    },
+    {
+      "epoch": 1.793935991016283,
+      "grad_norm": 0.21021628379821777,
+      "learning_rate": 4.165576197678571e-06,
+      "loss": 0.3781,
+      "step": 2130
+    },
+    {
+      "epoch": 1.7947782144862436,
+      "grad_norm": 0.1901855617761612,
+      "learning_rate": 4.160742741519028e-06,
+      "loss": 0.3428,
+      "step": 2131
+    },
+    {
+      "epoch": 1.7956204379562044,
+      "grad_norm": 0.21265830099582672,
+      "learning_rate": 4.1559100922410665e-06,
+      "loss": 0.3477,
+      "step": 2132
+    },
+    {
+      "epoch": 1.796462661426165,
+      "grad_norm": 0.19887612760066986,
+      "learning_rate": 4.151078254490908e-06,
+      "loss": 0.3552,
+      "step": 2133
+    },
+    {
+      "epoch": 1.7973048848961257,
+      "grad_norm": 0.22088709473609924,
+      "learning_rate": 4.146247232913996e-06,
+      "loss": 0.3513,
+      "step": 2134
+    },
+    {
+      "epoch": 1.7981471083660865,
+      "grad_norm": 0.1965794712305069,
+      "learning_rate": 4.141417032154984e-06,
+      "loss": 0.3753,
+      "step": 2135
+    },
+    {
+      "epoch": 1.7989893318360473,
+      "grad_norm": 0.357808381319046,
+      "learning_rate": 4.136587656857744e-06,
+      "loss": 0.3688,
+      "step": 2136
+    },
+    {
+      "epoch": 1.799831555306008,
+      "grad_norm": 0.20281732082366943,
+      "learning_rate": 4.131759111665349e-06,
+      "loss": 0.3474,
+      "step": 2137
+    },
+    {
+      "epoch": 1.8006737787759686,
+      "grad_norm": 0.20766642689704895,
+      "learning_rate": 4.126931401220075e-06,
+      "loss": 0.3685,
+      "step": 2138
+    },
+    {
+      "epoch": 1.8015160022459291,
+      "grad_norm": 0.2085840106010437,
+      "learning_rate": 4.122104530163397e-06,
+      "loss": 0.3612,
+      "step": 2139
+    },
+    {
+      "epoch": 1.8023582257158899,
+      "grad_norm": 0.2228211760520935,
+      "learning_rate": 4.117278503135981e-06,
+      "loss": 0.3581,
+      "step": 2140
+    },
+    {
+      "epoch": 1.8032004491858507,
+      "grad_norm": 0.22840428352355957,
+      "learning_rate": 4.112453324777683e-06,
+      "loss": 0.3242,
+      "step": 2141
+    },
+    {
+      "epoch": 1.8040426726558114,
+      "grad_norm": 0.22499187290668488,
+      "learning_rate": 4.107628999727542e-06,
+      "loss": 0.4048,
+      "step": 2142
+    },
+    {
+      "epoch": 1.8048848961257722,
+      "grad_norm": 0.1917266845703125,
+      "learning_rate": 4.102805532623775e-06,
+      "loss": 0.3339,
+      "step": 2143
+    },
+    {
+      "epoch": 1.8057271195957327,
+      "grad_norm": 0.2015596628189087,
+      "learning_rate": 4.097982928103782e-06,
+      "loss": 0.362,
+      "step": 2144
+    },
+    {
+      "epoch": 1.8065693430656933,
+      "grad_norm": 0.24050593376159668,
+      "learning_rate": 4.09316119080412e-06,
+      "loss": 0.3884,
+      "step": 2145
+    },
+    {
+      "epoch": 1.807411566535654,
+      "grad_norm": 0.18397676944732666,
+      "learning_rate": 4.088340325360529e-06,
+      "loss": 0.3047,
+      "step": 2146
+    },
+    {
+      "epoch": 1.8082537900056148,
+      "grad_norm": 0.20920848846435547,
+      "learning_rate": 4.083520336407894e-06,
+      "loss": 0.365,
+      "step": 2147
+    },
+    {
+      "epoch": 1.8090960134755756,
+      "grad_norm": 0.1963934749364853,
+      "learning_rate": 4.0787012285802695e-06,
+      "loss": 0.3462,
+      "step": 2148
+    },
+    {
+      "epoch": 1.8099382369455363,
+      "grad_norm": 0.19430825114250183,
+      "learning_rate": 4.073883006510858e-06,
+      "loss": 0.3678,
+      "step": 2149
+    },
+    {
+      "epoch": 1.810780460415497,
+      "grad_norm": 0.197347491979599,
+      "learning_rate": 4.069065674832011e-06,
+      "loss": 0.3406,
+      "step": 2150
+    },
+    {
+      "epoch": 1.8116226838854577,
+      "grad_norm": 0.22697080671787262,
+      "learning_rate": 4.064249238175223e-06,
+      "loss": 0.39,
+      "step": 2151
+    },
+    {
+      "epoch": 1.8124649073554182,
+      "grad_norm": 0.19796526432037354,
+      "learning_rate": 4.059433701171131e-06,
+      "loss": 0.3796,
+      "step": 2152
+    },
+    {
+      "epoch": 1.813307130825379,
+      "grad_norm": 0.19998793303966522,
+      "learning_rate": 4.054619068449502e-06,
+      "loss": 0.3628,
+      "step": 2153
+    },
+    {
+      "epoch": 1.8141493542953397,
+      "grad_norm": 0.21857374906539917,
+      "learning_rate": 4.04980534463924e-06,
+      "loss": 0.3922,
+      "step": 2154
+    },
+    {
+      "epoch": 1.8149915777653005,
+      "grad_norm": 0.27835121750831604,
+      "learning_rate": 4.044992534368369e-06,
+      "loss": 0.343,
+      "step": 2155
+    },
+    {
+      "epoch": 1.815833801235261,
+      "grad_norm": 0.22096095979213715,
+      "learning_rate": 4.04018064226404e-06,
+      "loss": 0.3341,
+      "step": 2156
+    },
+    {
+      "epoch": 1.8166760247052218,
+      "grad_norm": 0.21028409898281097,
+      "learning_rate": 4.035369672952516e-06,
+      "loss": 0.346,
+      "step": 2157
+    },
+    {
+      "epoch": 1.8175182481751824,
+      "grad_norm": 0.19172050058841705,
+      "learning_rate": 4.030559631059179e-06,
+      "loss": 0.3601,
+      "step": 2158
+    },
+    {
+      "epoch": 1.8183604716451431,
+      "grad_norm": 0.20794759690761566,
+      "learning_rate": 4.025750521208512e-06,
+      "loss": 0.3691,
+      "step": 2159
+    },
+    {
+      "epoch": 1.819202695115104,
+      "grad_norm": 0.22366368770599365,
+      "learning_rate": 4.020942348024108e-06,
+      "loss": 0.396,
+      "step": 2160
+    },
+    {
+      "epoch": 1.8200449185850647,
+      "grad_norm": 0.20643411576747894,
+      "learning_rate": 4.016135116128656e-06,
+      "loss": 0.3429,
+      "step": 2161
+    },
+    {
+      "epoch": 1.8208871420550252,
+      "grad_norm": 0.2257109135389328,
+      "learning_rate": 4.011328830143945e-06,
+      "loss": 0.326,
+      "step": 2162
+    },
+    {
+      "epoch": 1.821729365524986,
+      "grad_norm": 0.18669435381889343,
+      "learning_rate": 4.0065234946908456e-06,
+      "loss": 0.3354,
+      "step": 2163
+    },
+    {
+      "epoch": 1.8225715889949465,
+      "grad_norm": 0.20572814345359802,
+      "learning_rate": 4.001719114389325e-06,
+      "loss": 0.3673,
+      "step": 2164
+    },
+    {
+      "epoch": 1.8234138124649073,
+      "grad_norm": 0.2180599868297577,
+      "learning_rate": 3.996915693858422e-06,
+      "loss": 0.3897,
+      "step": 2165
+    },
+    {
+      "epoch": 1.824256035934868,
+      "grad_norm": 0.19118410348892212,
+      "learning_rate": 3.992113237716261e-06,
+      "loss": 0.3432,
+      "step": 2166
+    },
+    {
+      "epoch": 1.8250982594048288,
+      "grad_norm": 0.20306925475597382,
+      "learning_rate": 3.987311750580035e-06,
+      "loss": 0.3452,
+      "step": 2167
+    },
+    {
+      "epoch": 1.8259404828747896,
+      "grad_norm": 0.19168779253959656,
+      "learning_rate": 3.9825112370660055e-06,
+      "loss": 0.3373,
+      "step": 2168
+    },
+    {
+      "epoch": 1.8267827063447502,
+      "grad_norm": 0.19850198924541473,
+      "learning_rate": 3.977711701789499e-06,
+      "loss": 0.3646,
+      "step": 2169
+    },
+    {
+      "epoch": 1.8276249298147107,
+      "grad_norm": 0.20444419980049133,
+      "learning_rate": 3.972913149364902e-06,
+      "loss": 0.3841,
+      "step": 2170
+    },
+    {
+      "epoch": 1.8284671532846715,
+      "grad_norm": 0.1842080056667328,
+      "learning_rate": 3.9681155844056525e-06,
+      "loss": 0.3444,
+      "step": 2171
+    },
+    {
+      "epoch": 1.8293093767546322,
+      "grad_norm": 0.18652406334877014,
+      "learning_rate": 3.963319011524246e-06,
+      "loss": 0.3487,
+      "step": 2172
+    },
+    {
+      "epoch": 1.830151600224593,
+      "grad_norm": 0.1885792762041092,
+      "learning_rate": 3.9585234353322155e-06,
+      "loss": 0.3647,
+      "step": 2173
+    },
+    {
+      "epoch": 1.8309938236945538,
+      "grad_norm": 0.2128867506980896,
+      "learning_rate": 3.953728860440144e-06,
+      "loss": 0.3386,
+      "step": 2174
+    },
+    {
+      "epoch": 1.8318360471645143,
+      "grad_norm": 0.1972573846578598,
+      "learning_rate": 3.948935291457645e-06,
+      "loss": 0.3125,
+      "step": 2175
+    },
+    {
+      "epoch": 1.8326782706344749,
+      "grad_norm": 0.21564050018787384,
+      "learning_rate": 3.94414273299337e-06,
+      "loss": 0.3508,
+      "step": 2176
+    },
+    {
+      "epoch": 1.8335204941044356,
+      "grad_norm": 0.2020157277584076,
+      "learning_rate": 3.939351189654996e-06,
+      "loss": 0.3616,
+      "step": 2177
+    },
+    {
+      "epoch": 1.8343627175743964,
+      "grad_norm": 0.21299459040164948,
+      "learning_rate": 3.934560666049226e-06,
+      "loss": 0.3513,
+      "step": 2178
+    },
+    {
+      "epoch": 1.8352049410443572,
+      "grad_norm": 0.18165704607963562,
+      "learning_rate": 3.929771166781781e-06,
+      "loss": 0.3205,
+      "step": 2179
+    },
+    {
+      "epoch": 1.836047164514318,
+      "grad_norm": 0.21832971274852753,
+      "learning_rate": 3.9249826964573965e-06,
+      "loss": 0.3825,
+      "step": 2180
+    },
+    {
+      "epoch": 1.8368893879842785,
+      "grad_norm": 0.19859738647937775,
+      "learning_rate": 3.920195259679822e-06,
+      "loss": 0.3642,
+      "step": 2181
+    },
+    {
+      "epoch": 1.8377316114542392,
+      "grad_norm": 0.1959538608789444,
+      "learning_rate": 3.915408861051809e-06,
+      "loss": 0.3482,
+      "step": 2182
+    },
+    {
+      "epoch": 1.8385738349241998,
+      "grad_norm": 0.20815512537956238,
+      "learning_rate": 3.910623505175116e-06,
+      "loss": 0.3406,
+      "step": 2183
+    },
+    {
+      "epoch": 1.8394160583941606,
+      "grad_norm": 0.19060750305652618,
+      "learning_rate": 3.905839196650494e-06,
+      "loss": 0.3643,
+      "step": 2184
+    },
+    {
+      "epoch": 1.8402582818641213,
+      "grad_norm": 0.20968878269195557,
+      "learning_rate": 3.901055940077691e-06,
+      "loss": 0.3546,
+      "step": 2185
+    },
+    {
+      "epoch": 1.841100505334082,
+      "grad_norm": 0.23598092794418335,
+      "learning_rate": 3.8962737400554395e-06,
+      "loss": 0.3687,
+      "step": 2186
+    },
+    {
+      "epoch": 1.8419427288040426,
+      "grad_norm": 0.1882329136133194,
+      "learning_rate": 3.891492601181462e-06,
+      "loss": 0.3631,
+      "step": 2187
+    },
+    {
+      "epoch": 1.8427849522740034,
+      "grad_norm": 0.18433696031570435,
+      "learning_rate": 3.8867125280524535e-06,
+      "loss": 0.3527,
+      "step": 2188
+    },
+    {
+      "epoch": 1.843627175743964,
+      "grad_norm": 0.20385466516017914,
+      "learning_rate": 3.881933525264092e-06,
+      "loss": 0.336,
+      "step": 2189
+    },
+    {
+      "epoch": 1.8444693992139247,
+      "grad_norm": 0.22509856522083282,
+      "learning_rate": 3.877155597411019e-06,
+      "loss": 0.3987,
+      "step": 2190
+    },
+    {
+      "epoch": 1.8453116226838855,
+      "grad_norm": 0.19194535911083221,
+      "learning_rate": 3.87237874908685e-06,
+      "loss": 0.3319,
+      "step": 2191
+    },
+    {
+      "epoch": 1.8461538461538463,
+      "grad_norm": 0.23621989786624908,
+      "learning_rate": 3.867602984884155e-06,
+      "loss": 0.3902,
+      "step": 2192
+    },
+    {
+      "epoch": 1.8469960696238068,
+      "grad_norm": 0.19871260225772858,
+      "learning_rate": 3.862828309394469e-06,
+      "loss": 0.3398,
+      "step": 2193
+    },
+    {
+      "epoch": 1.8478382930937676,
+      "grad_norm": 0.19352886080741882,
+      "learning_rate": 3.8580547272082746e-06,
+      "loss": 0.3259,
+      "step": 2194
+    },
+    {
+      "epoch": 1.8486805165637281,
+      "grad_norm": 0.19926981627941132,
+      "learning_rate": 3.853282242915007e-06,
+      "loss": 0.3544,
+      "step": 2195
+    },
+    {
+      "epoch": 1.8495227400336889,
+      "grad_norm": 0.22587287425994873,
+      "learning_rate": 3.8485108611030415e-06,
+      "loss": 0.3695,
+      "step": 2196
+    },
+    {
+      "epoch": 1.8503649635036497,
+      "grad_norm": 0.20771846175193787,
+      "learning_rate": 3.843740586359701e-06,
+      "loss": 0.3504,
+      "step": 2197
+    },
+    {
+      "epoch": 1.8512071869736104,
+      "grad_norm": 0.1965019553899765,
+      "learning_rate": 3.8389714232712346e-06,
+      "loss": 0.3503,
+      "step": 2198
+    },
+    {
+      "epoch": 1.8520494104435712,
+      "grad_norm": 0.21217331290245056,
+      "learning_rate": 3.834203376422831e-06,
+      "loss": 0.3286,
+      "step": 2199
+    },
+    {
+      "epoch": 1.8528916339135317,
+      "grad_norm": 0.19798773527145386,
+      "learning_rate": 3.829436450398599e-06,
+      "loss": 0.3543,
+      "step": 2200
+    },
+    {
+      "epoch": 1.8537338573834923,
+      "grad_norm": 0.2020939290523529,
+      "learning_rate": 3.824670649781576e-06,
+      "loss": 0.376,
+      "step": 2201
+    },
+    {
+      "epoch": 1.854576080853453,
+      "grad_norm": 0.19974105060100555,
+      "learning_rate": 3.8199059791537105e-06,
+      "loss": 0.3364,
+      "step": 2202
+    },
+    {
+      "epoch": 1.8554183043234138,
+      "grad_norm": 0.1852501779794693,
+      "learning_rate": 3.815142443095873e-06,
+      "loss": 0.365,
+      "step": 2203
+    },
+    {
+      "epoch": 1.8562605277933746,
+      "grad_norm": 0.20142561197280884,
+      "learning_rate": 3.8103800461878344e-06,
+      "loss": 0.3541,
+      "step": 2204
+    },
+    {
+      "epoch": 1.8571027512633353,
+      "grad_norm": 0.21110711991786957,
+      "learning_rate": 3.805618793008279e-06,
+      "loss": 0.3517,
+      "step": 2205
+    },
+    {
+      "epoch": 1.857944974733296,
+      "grad_norm": 0.2029833346605301,
+      "learning_rate": 3.8008586881347815e-06,
+      "loss": 0.3593,
+      "step": 2206
+    },
+    {
+      "epoch": 1.8587871982032564,
+      "grad_norm": 0.1974041759967804,
+      "learning_rate": 3.7960997361438235e-06,
+      "loss": 0.324,
+      "step": 2207
+    },
+    {
+      "epoch": 1.8596294216732172,
+      "grad_norm": 0.19707661867141724,
+      "learning_rate": 3.7913419416107692e-06,
+      "loss": 0.358,
+      "step": 2208
+    },
+    {
+      "epoch": 1.860471645143178,
+      "grad_norm": 0.18754833936691284,
+      "learning_rate": 3.786585309109877e-06,
+      "loss": 0.3383,
+      "step": 2209
+    },
+    {
+      "epoch": 1.8613138686131387,
+      "grad_norm": 0.20792751014232635,
+      "learning_rate": 3.7818298432142814e-06,
+      "loss": 0.3767,
+      "step": 2210
+    },
+    {
+      "epoch": 1.8621560920830995,
+      "grad_norm": 0.2125100940465927,
+      "learning_rate": 3.777075548496001e-06,
+      "loss": 0.3741,
+      "step": 2211
+    },
+    {
+      "epoch": 1.86299831555306,
+      "grad_norm": 0.20231522619724274,
+      "learning_rate": 3.7723224295259247e-06,
+      "loss": 0.3425,
+      "step": 2212
+    },
+    {
+      "epoch": 1.8638405390230208,
+      "grad_norm": 0.19012172520160675,
+      "learning_rate": 3.7675704908738136e-06,
+      "loss": 0.3723,
+      "step": 2213
+    },
+    {
+      "epoch": 1.8646827624929814,
+      "grad_norm": 0.18218067288398743,
+      "learning_rate": 3.7628197371082916e-06,
+      "loss": 0.3243,
+      "step": 2214
+    },
+    {
+      "epoch": 1.8655249859629421,
+      "grad_norm": 0.1938510239124298,
+      "learning_rate": 3.758070172796846e-06,
+      "loss": 0.3692,
+      "step": 2215
+    },
+    {
+      "epoch": 1.866367209432903,
+      "grad_norm": 0.20477046072483063,
+      "learning_rate": 3.753321802505817e-06,
+      "loss": 0.4046,
+      "step": 2216
+    },
+    {
+      "epoch": 1.8672094329028637,
+      "grad_norm": 0.19240835309028625,
+      "learning_rate": 3.7485746308004013e-06,
+      "loss": 0.3404,
+      "step": 2217
+    },
+    {
+      "epoch": 1.8680516563728242,
+      "grad_norm": 0.17858603596687317,
+      "learning_rate": 3.743828662244639e-06,
+      "loss": 0.3911,
+      "step": 2218
+    },
+    {
+      "epoch": 1.868893879842785,
+      "grad_norm": 0.18475256860256195,
+      "learning_rate": 3.739083901401418e-06,
+      "loss": 0.3246,
+      "step": 2219
+    },
+    {
+      "epoch": 1.8697361033127455,
+      "grad_norm": 0.19433172047138214,
+      "learning_rate": 3.7343403528324574e-06,
+      "loss": 0.3648,
+      "step": 2220
+    },
+    {
+      "epoch": 1.8705783267827063,
+      "grad_norm": 0.20487923920154572,
+      "learning_rate": 3.7295980210983233e-06,
+      "loss": 0.3835,
+      "step": 2221
+    },
+    {
+      "epoch": 1.871420550252667,
+      "grad_norm": 0.19962626695632935,
+      "learning_rate": 3.7248569107583976e-06,
+      "loss": 0.3572,
+      "step": 2222
+    },
+    {
+      "epoch": 1.8722627737226278,
+      "grad_norm": 0.18625806272029877,
+      "learning_rate": 3.7201170263709004e-06,
+      "loss": 0.3539,
+      "step": 2223
+    },
+    {
+      "epoch": 1.8731049971925884,
+      "grad_norm": 0.1880733072757721,
+      "learning_rate": 3.7153783724928617e-06,
+      "loss": 0.338,
+      "step": 2224
+    },
+    {
+      "epoch": 1.8739472206625492,
+      "grad_norm": 0.22529815137386322,
+      "learning_rate": 3.71064095368014e-06,
+      "loss": 0.3857,
+      "step": 2225
+    },
+    {
+      "epoch": 1.8747894441325097,
+      "grad_norm": 0.19889788329601288,
+      "learning_rate": 3.705904774487396e-06,
+      "loss": 0.3197,
+      "step": 2226
+    },
+    {
+      "epoch": 1.8756316676024705,
+      "grad_norm": 0.22110477089881897,
+      "learning_rate": 3.7011698394681075e-06,
+      "loss": 0.3915,
+      "step": 2227
+    },
+    {
+      "epoch": 1.8764738910724312,
+      "grad_norm": 0.18708573281764984,
+      "learning_rate": 3.696436153174548e-06,
+      "loss": 0.3457,
+      "step": 2228
+    },
+    {
+      "epoch": 1.877316114542392,
+      "grad_norm": 0.1997976154088974,
+      "learning_rate": 3.6917037201577977e-06,
+      "loss": 0.3416,
+      "step": 2229
+    },
+    {
+      "epoch": 1.8781583380123528,
+      "grad_norm": 0.19073273241519928,
+      "learning_rate": 3.6869725449677254e-06,
+      "loss": 0.3595,
+      "step": 2230
+    },
+    {
+      "epoch": 1.8790005614823133,
+      "grad_norm": 0.204877570271492,
+      "learning_rate": 3.6822426321529967e-06,
+      "loss": 0.3675,
+      "step": 2231
+    },
+    {
+      "epoch": 1.8798427849522739,
+      "grad_norm": 0.19971223175525665,
+      "learning_rate": 3.6775139862610577e-06,
+      "loss": 0.3554,
+      "step": 2232
+    },
+    {
+      "epoch": 1.8806850084222346,
+      "grad_norm": 0.18837471306324005,
+      "learning_rate": 3.672786611838142e-06,
+      "loss": 0.3191,
+      "step": 2233
+    },
+    {
+      "epoch": 1.8815272318921954,
+      "grad_norm": 0.18985477089881897,
+      "learning_rate": 3.668060513429256e-06,
+      "loss": 0.3756,
+      "step": 2234
+    },
+    {
+      "epoch": 1.8823694553621562,
+      "grad_norm": 0.32632893323898315,
+      "learning_rate": 3.6633356955781827e-06,
+      "loss": 0.3784,
+      "step": 2235
+    },
+    {
+      "epoch": 1.883211678832117,
+      "grad_norm": 0.22074228525161743,
+      "learning_rate": 3.658612162827472e-06,
+      "loss": 0.3536,
+      "step": 2236
+    },
+    {
+      "epoch": 1.8840539023020775,
+      "grad_norm": 0.20336230099201202,
+      "learning_rate": 3.653889919718439e-06,
+      "loss": 0.3252,
+      "step": 2237
+    },
+    {
+      "epoch": 1.884896125772038,
+      "grad_norm": 0.20199203491210938,
+      "learning_rate": 3.649168970791157e-06,
+      "loss": 0.3427,
+      "step": 2238
+    },
+    {
+      "epoch": 1.8857383492419988,
+      "grad_norm": 0.18671944737434387,
+      "learning_rate": 3.644449320584462e-06,
+      "loss": 0.3506,
+      "step": 2239
+    },
+    {
+      "epoch": 1.8865805727119596,
+      "grad_norm": 0.21687693893909454,
+      "learning_rate": 3.639730973635929e-06,
+      "loss": 0.3996,
+      "step": 2240
+    },
+    {
+      "epoch": 1.8874227961819203,
+      "grad_norm": 0.19979308545589447,
+      "learning_rate": 3.635013934481895e-06,
+      "loss": 0.3454,
+      "step": 2241
+    },
+    {
+      "epoch": 1.888265019651881,
+      "grad_norm": 0.18458348512649536,
+      "learning_rate": 3.6302982076574244e-06,
+      "loss": 0.3344,
+      "step": 2242
+    },
+    {
+      "epoch": 1.8891072431218416,
+      "grad_norm": 0.21379539370536804,
+      "learning_rate": 3.6255837976963336e-06,
+      "loss": 0.3483,
+      "step": 2243
+    },
+    {
+      "epoch": 1.8899494665918024,
+      "grad_norm": 0.1934133619070053,
+      "learning_rate": 3.620870709131163e-06,
+      "loss": 0.36,
+      "step": 2244
+    },
+    {
+      "epoch": 1.890791690061763,
+      "grad_norm": 0.20163948833942413,
+      "learning_rate": 3.616158946493188e-06,
+      "loss": 0.3736,
+      "step": 2245
+    },
+    {
+      "epoch": 1.8916339135317237,
+      "grad_norm": 0.19730351865291595,
+      "learning_rate": 3.6114485143124068e-06,
+      "loss": 0.3244,
+      "step": 2246
+    },
+    {
+      "epoch": 1.8924761370016845,
+      "grad_norm": 0.21036989986896515,
+      "learning_rate": 3.6067394171175397e-06,
+      "loss": 0.3571,
+      "step": 2247
+    },
+    {
+      "epoch": 1.8933183604716453,
+      "grad_norm": 0.19844365119934082,
+      "learning_rate": 3.602031659436022e-06,
+      "loss": 0.3596,
+      "step": 2248
+    },
+    {
+      "epoch": 1.8941605839416058,
+      "grad_norm": 0.20970189571380615,
+      "learning_rate": 3.5973252457940034e-06,
+      "loss": 0.3491,
+      "step": 2249
+    },
+    {
+      "epoch": 1.8950028074115666,
+      "grad_norm": 0.20957820117473602,
+      "learning_rate": 3.5926201807163384e-06,
+      "loss": 0.3727,
+      "step": 2250
+    },
+    {
+      "epoch": 1.8958450308815271,
+      "grad_norm": 0.20133230090141296,
+      "learning_rate": 3.58791646872659e-06,
+      "loss": 0.3719,
+      "step": 2251
+    },
+    {
+      "epoch": 1.8966872543514879,
+      "grad_norm": 0.2148762345314026,
+      "learning_rate": 3.5832141143470146e-06,
+      "loss": 0.351,
+      "step": 2252
+    },
+    {
+      "epoch": 1.8975294778214487,
+      "grad_norm": 0.19707302749156952,
+      "learning_rate": 3.578513122098566e-06,
+      "loss": 0.323,
+      "step": 2253
+    },
+    {
+      "epoch": 1.8983717012914094,
+      "grad_norm": 0.22635704278945923,
+      "learning_rate": 3.5738134965008885e-06,
+      "loss": 0.3896,
+      "step": 2254
+    },
+    {
+      "epoch": 1.89921392476137,
+      "grad_norm": 0.4272863268852234,
+      "learning_rate": 3.5691152420723115e-06,
+      "loss": 0.3598,
+      "step": 2255
+    },
+    {
+      "epoch": 1.9000561482313307,
+      "grad_norm": 0.1947227418422699,
+      "learning_rate": 3.564418363329848e-06,
+      "loss": 0.3521,
+      "step": 2256
+    },
+    {
+      "epoch": 1.9008983717012913,
+      "grad_norm": 0.2008480727672577,
+      "learning_rate": 3.559722864789187e-06,
+      "loss": 0.3246,
+      "step": 2257
+    },
+    {
+      "epoch": 1.901740595171252,
+      "grad_norm": 0.22643692791461945,
+      "learning_rate": 3.5550287509646902e-06,
+      "loss": 0.3798,
+      "step": 2258
+    },
+    {
+      "epoch": 1.9025828186412128,
+      "grad_norm": 0.18489694595336914,
+      "learning_rate": 3.5503360263693887e-06,
+      "loss": 0.3446,
+      "step": 2259
+    },
+    {
+      "epoch": 1.9034250421111736,
+      "grad_norm": 0.19398696720600128,
+      "learning_rate": 3.5456446955149783e-06,
+      "loss": 0.3824,
+      "step": 2260
+    },
+    {
+      "epoch": 1.9042672655811343,
+      "grad_norm": 0.2685796916484833,
+      "learning_rate": 3.5409547629118124e-06,
+      "loss": 0.3266,
+      "step": 2261
+    },
+    {
+      "epoch": 1.905109489051095,
+      "grad_norm": 0.18848860263824463,
+      "learning_rate": 3.5362662330689067e-06,
+      "loss": 0.3595,
+      "step": 2262
+    },
+    {
+      "epoch": 1.9059517125210554,
+      "grad_norm": 0.20824608206748962,
+      "learning_rate": 3.531579110493917e-06,
+      "loss": 0.3746,
+      "step": 2263
+    },
+    {
+      "epoch": 1.9067939359910162,
+      "grad_norm": 0.20500995218753815,
+      "learning_rate": 3.5268933996931596e-06,
+      "loss": 0.3158,
+      "step": 2264
+    },
+    {
+      "epoch": 1.907636159460977,
+      "grad_norm": 0.18048281967639923,
+      "learning_rate": 3.5222091051715803e-06,
+      "loss": 0.3517,
+      "step": 2265
+    },
+    {
+      "epoch": 1.9084783829309377,
+      "grad_norm": 0.21019864082336426,
+      "learning_rate": 3.517526231432775e-06,
+      "loss": 0.3784,
+      "step": 2266
+    },
+    {
+      "epoch": 1.9093206064008985,
+      "grad_norm": 0.1963644176721573,
+      "learning_rate": 3.512844782978963e-06,
+      "loss": 0.3428,
+      "step": 2267
+    },
+    {
+      "epoch": 1.910162829870859,
+      "grad_norm": 0.2139882594347,
+      "learning_rate": 3.5081647643110028e-06,
+      "loss": 0.3766,
+      "step": 2268
+    },
+    {
+      "epoch": 1.9110050533408196,
+      "grad_norm": 0.17365844547748566,
+      "learning_rate": 3.5034861799283713e-06,
+      "loss": 0.3341,
+      "step": 2269
+    },
+    {
+      "epoch": 1.9118472768107804,
+      "grad_norm": 0.18427762389183044,
+      "learning_rate": 3.498809034329171e-06,
+      "loss": 0.3435,
+      "step": 2270
+    },
+    {
+      "epoch": 1.9126895002807411,
+      "grad_norm": 0.1972460150718689,
+      "learning_rate": 3.4941333320101173e-06,
+      "loss": 0.3402,
+      "step": 2271
+    },
+    {
+      "epoch": 1.913531723750702,
+      "grad_norm": 0.22721323370933533,
+      "learning_rate": 3.4894590774665414e-06,
+      "loss": 0.3715,
+      "step": 2272
+    },
+    {
+      "epoch": 1.9143739472206627,
+      "grad_norm": 0.204066202044487,
+      "learning_rate": 3.48478627519238e-06,
+      "loss": 0.3372,
+      "step": 2273
+    },
+    {
+      "epoch": 1.9152161706906232,
+      "grad_norm": 0.21919427812099457,
+      "learning_rate": 3.480114929680176e-06,
+      "loss": 0.3593,
+      "step": 2274
+    },
+    {
+      "epoch": 1.916058394160584,
+      "grad_norm": 0.24482612311840057,
+      "learning_rate": 3.4754450454210686e-06,
+      "loss": 0.3878,
+      "step": 2275
+    },
+    {
+      "epoch": 1.9169006176305445,
+      "grad_norm": 0.21800611913204193,
+      "learning_rate": 3.470776626904795e-06,
+      "loss": 0.3533,
+      "step": 2276
+    },
+    {
+      "epoch": 1.9177428411005053,
+      "grad_norm": 0.21003371477127075,
+      "learning_rate": 3.466109678619681e-06,
+      "loss": 0.3648,
+      "step": 2277
+    },
+    {
+      "epoch": 1.918585064570466,
+      "grad_norm": 0.20837385952472687,
+      "learning_rate": 3.4614442050526424e-06,
+      "loss": 0.3303,
+      "step": 2278
+    },
+    {
+      "epoch": 1.9194272880404268,
+      "grad_norm": 0.25349435210227966,
+      "learning_rate": 3.4567802106891724e-06,
+      "loss": 0.3891,
+      "step": 2279
+    },
+    {
+      "epoch": 1.9202695115103874,
+      "grad_norm": 0.23070620000362396,
+      "learning_rate": 3.4521177000133456e-06,
+      "loss": 0.3497,
+      "step": 2280
+    },
+    {
+      "epoch": 1.9211117349803482,
+      "grad_norm": 0.21101213991641998,
+      "learning_rate": 3.4474566775078055e-06,
+      "loss": 0.3595,
+      "step": 2281
+    },
+    {
+      "epoch": 1.9219539584503087,
+      "grad_norm": 0.18892095983028412,
+      "learning_rate": 3.442797147653776e-06,
+      "loss": 0.3136,
+      "step": 2282
+    },
+    {
+      "epoch": 1.9227961819202695,
+      "grad_norm": 0.22585587203502655,
+      "learning_rate": 3.4381391149310294e-06,
+      "loss": 0.3791,
+      "step": 2283
+    },
+    {
+      "epoch": 1.9236384053902302,
+      "grad_norm": 0.21867422759532928,
+      "learning_rate": 3.4334825838179143e-06,
+      "loss": 0.3722,
+      "step": 2284
+    },
+    {
+      "epoch": 1.924480628860191,
+      "grad_norm": 0.20635460317134857,
+      "learning_rate": 3.4288275587913235e-06,
+      "loss": 0.3633,
+      "step": 2285
+    },
+    {
+      "epoch": 1.9253228523301515,
+      "grad_norm": 0.18385303020477295,
+      "learning_rate": 3.4241740443267112e-06,
+      "loss": 0.3453,
+      "step": 2286
+    },
+    {
+      "epoch": 1.9261650758001123,
+      "grad_norm": 0.2156953364610672,
+      "learning_rate": 3.419522044898073e-06,
+      "loss": 0.3443,
+      "step": 2287
+    },
+    {
+      "epoch": 1.9270072992700729,
+      "grad_norm": 0.20479954779148102,
+      "learning_rate": 3.414871564977951e-06,
+      "loss": 0.3783,
+      "step": 2288
+    },
+    {
+      "epoch": 1.9278495227400336,
+      "grad_norm": 0.1958267241716385,
+      "learning_rate": 3.4102226090374246e-06,
+      "loss": 0.3814,
+      "step": 2289
+    },
+    {
+      "epoch": 1.9286917462099944,
+      "grad_norm": 0.20148569345474243,
+      "learning_rate": 3.4055751815461102e-06,
+      "loss": 0.341,
+      "step": 2290
+    },
+    {
+      "epoch": 1.9295339696799552,
+      "grad_norm": 0.20417562127113342,
+      "learning_rate": 3.4009292869721516e-06,
+      "loss": 0.3586,
+      "step": 2291
+    },
+    {
+      "epoch": 1.930376193149916,
+      "grad_norm": 0.22307829558849335,
+      "learning_rate": 3.3962849297822225e-06,
+      "loss": 0.3653,
+      "step": 2292
+    },
+    {
+      "epoch": 1.9312184166198765,
+      "grad_norm": 0.2065296322107315,
+      "learning_rate": 3.3916421144415146e-06,
+      "loss": 0.3578,
+      "step": 2293
+    },
+    {
+      "epoch": 1.932060640089837,
+      "grad_norm": 0.17637860774993896,
+      "learning_rate": 3.387000845413742e-06,
+      "loss": 0.3092,
+      "step": 2294
+    },
+    {
+      "epoch": 1.9329028635597978,
+      "grad_norm": 0.2161666601896286,
+      "learning_rate": 3.3823611271611266e-06,
+      "loss": 0.3947,
+      "step": 2295
+    },
+    {
+      "epoch": 1.9337450870297586,
+      "grad_norm": 0.19415435194969177,
+      "learning_rate": 3.377722964144405e-06,
+      "loss": 0.3314,
+      "step": 2296
+    },
+    {
+      "epoch": 1.9345873104997193,
+      "grad_norm": 0.26362645626068115,
+      "learning_rate": 3.3730863608228125e-06,
+      "loss": 0.3795,
+      "step": 2297
+    },
+    {
+      "epoch": 1.93542953396968,
+      "grad_norm": 0.192129984498024,
+      "learning_rate": 3.368451321654091e-06,
+      "loss": 0.3675,
+      "step": 2298
+    },
+    {
+      "epoch": 1.9362717574396406,
+      "grad_norm": 0.19030117988586426,
+      "learning_rate": 3.363817851094473e-06,
+      "loss": 0.3225,
+      "step": 2299
+    },
+    {
+      "epoch": 1.9371139809096012,
+      "grad_norm": 0.2569832503795624,
+      "learning_rate": 3.3591859535986894e-06,
+      "loss": 0.3636,
+      "step": 2300
+    },
+    {
+      "epoch": 1.937956204379562,
+      "grad_norm": 0.19425804913043976,
+      "learning_rate": 3.35455563361995e-06,
+      "loss": 0.3706,
+      "step": 2301
+    },
+    {
+      "epoch": 1.9387984278495227,
+      "grad_norm": 0.18746623396873474,
+      "learning_rate": 3.3499268956099583e-06,
+      "loss": 0.3475,
+      "step": 2302
+    },
+    {
+      "epoch": 1.9396406513194835,
+      "grad_norm": 0.20306524634361267,
+      "learning_rate": 3.345299744018886e-06,
+      "loss": 0.3673,
+      "step": 2303
+    },
+    {
+      "epoch": 1.9404828747894443,
+      "grad_norm": 0.20163676142692566,
+      "learning_rate": 3.3406741832953893e-06,
+      "loss": 0.36,
+      "step": 2304
+    },
+    {
+      "epoch": 1.9413250982594048,
+      "grad_norm": 0.23225265741348267,
+      "learning_rate": 3.336050217886588e-06,
+      "loss": 0.3633,
+      "step": 2305
+    },
+    {
+      "epoch": 1.9421673217293656,
+      "grad_norm": 0.19788064062595367,
+      "learning_rate": 3.331427852238073e-06,
+      "loss": 0.3652,
+      "step": 2306
+    },
+    {
+      "epoch": 1.9430095451993261,
+      "grad_norm": 0.19441500306129456,
+      "learning_rate": 3.3268070907938915e-06,
+      "loss": 0.3412,
+      "step": 2307
+    },
+    {
+      "epoch": 1.9438517686692869,
+      "grad_norm": 0.18719305098056793,
+      "learning_rate": 3.3221879379965553e-06,
+      "loss": 0.3476,
+      "step": 2308
+    },
+    {
+      "epoch": 1.9446939921392477,
+      "grad_norm": 0.1911689043045044,
+      "learning_rate": 3.3175703982870232e-06,
+      "loss": 0.3511,
+      "step": 2309
+    },
+    {
+      "epoch": 1.9455362156092084,
+      "grad_norm": 0.18693746626377106,
+      "learning_rate": 3.3129544761047093e-06,
+      "loss": 0.3135,
+      "step": 2310
+    },
+    {
+      "epoch": 1.946378439079169,
+      "grad_norm": 0.2037518471479416,
+      "learning_rate": 3.3083401758874655e-06,
+      "loss": 0.3657,
+      "step": 2311
+    },
+    {
+      "epoch": 1.9472206625491297,
+      "grad_norm": 0.21990913152694702,
+      "learning_rate": 3.303727502071591e-06,
+      "loss": 0.3737,
+      "step": 2312
+    },
+    {
+      "epoch": 1.9480628860190903,
+      "grad_norm": 0.17429234087467194,
+      "learning_rate": 3.2991164590918162e-06,
+      "loss": 0.331,
+      "step": 2313
+    },
+    {
+      "epoch": 1.948905109489051,
+      "grad_norm": 0.18694309890270233,
+      "learning_rate": 3.2945070513813082e-06,
+      "loss": 0.3524,
+      "step": 2314
+    },
+    {
+      "epoch": 1.9497473329590118,
+      "grad_norm": 0.2149062603712082,
+      "learning_rate": 3.289899283371657e-06,
+      "loss": 0.3825,
+      "step": 2315
+    },
+    {
+      "epoch": 1.9505895564289726,
+      "grad_norm": 0.19212117791175842,
+      "learning_rate": 3.2852931594928804e-06,
+      "loss": 0.3433,
+      "step": 2316
+    },
+    {
+      "epoch": 1.9514317798989333,
+      "grad_norm": 0.1927662193775177,
+      "learning_rate": 3.280688684173412e-06,
+      "loss": 0.3673,
+      "step": 2317
+    },
+    {
+      "epoch": 1.952274003368894,
+      "grad_norm": 0.20035819709300995,
+      "learning_rate": 3.276085861840106e-06,
+      "loss": 0.3546,
+      "step": 2318
+    },
+    {
+      "epoch": 1.9531162268388544,
+      "grad_norm": 0.20155903697013855,
+      "learning_rate": 3.271484696918218e-06,
+      "loss": 0.3196,
+      "step": 2319
+    },
+    {
+      "epoch": 1.9539584503088152,
+      "grad_norm": 0.2146608829498291,
+      "learning_rate": 3.2668851938314217e-06,
+      "loss": 0.3786,
+      "step": 2320
+    },
+    {
+      "epoch": 1.954800673778776,
+      "grad_norm": 0.20258279144763947,
+      "learning_rate": 3.262287357001781e-06,
+      "loss": 0.3656,
+      "step": 2321
+    },
+    {
+      "epoch": 1.9556428972487367,
+      "grad_norm": 0.20564958453178406,
+      "learning_rate": 3.2576911908497695e-06,
+      "loss": 0.3539,
+      "step": 2322
+    },
+    {
+      "epoch": 1.9564851207186975,
+      "grad_norm": 0.20844140648841858,
+      "learning_rate": 3.253096699794245e-06,
+      "loss": 0.3663,
+      "step": 2323
+    },
+    {
+      "epoch": 1.957327344188658,
+      "grad_norm": 0.20116738975048065,
+      "learning_rate": 3.248503888252461e-06,
+      "loss": 0.3497,
+      "step": 2324
+    },
+    {
+      "epoch": 1.9581695676586186,
+      "grad_norm": 0.18670654296875,
+      "learning_rate": 3.2439127606400546e-06,
+      "loss": 0.3396,
+      "step": 2325
+    },
+    {
+      "epoch": 1.9590117911285794,
+      "grad_norm": 0.2075943797826767,
+      "learning_rate": 3.239323321371039e-06,
+      "loss": 0.3482,
+      "step": 2326
+    },
+    {
+      "epoch": 1.9598540145985401,
+      "grad_norm": 0.21039381623268127,
+      "learning_rate": 3.2347355748578134e-06,
+      "loss": 0.4028,
+      "step": 2327
+    },
+    {
+      "epoch": 1.960696238068501,
+      "grad_norm": 0.1902557611465454,
+      "learning_rate": 3.2301495255111426e-06,
+      "loss": 0.3305,
+      "step": 2328
+    },
+    {
+      "epoch": 1.9615384615384617,
+      "grad_norm": 0.210167795419693,
+      "learning_rate": 3.225565177740163e-06,
+      "loss": 0.3658,
+      "step": 2329
+    },
+    {
+      "epoch": 1.9623806850084222,
+      "grad_norm": 0.18592344224452972,
+      "learning_rate": 3.2209825359523717e-06,
+      "loss": 0.3383,
+      "step": 2330
+    },
+    {
+      "epoch": 1.9632229084783828,
+      "grad_norm": 0.19514751434326172,
+      "learning_rate": 3.2164016045536306e-06,
+      "loss": 0.3599,
+      "step": 2331
+    },
+    {
+      "epoch": 1.9640651319483435,
+      "grad_norm": 0.18985408544540405,
+      "learning_rate": 3.2118223879481525e-06,
+      "loss": 0.3548,
+      "step": 2332
+    },
+    {
+      "epoch": 1.9649073554183043,
+      "grad_norm": 0.19115465879440308,
+      "learning_rate": 3.2072448905385046e-06,
+      "loss": 0.3573,
+      "step": 2333
+    },
+    {
+      "epoch": 1.965749578888265,
+      "grad_norm": 0.20761296153068542,
+      "learning_rate": 3.202669116725598e-06,
+      "loss": 0.3612,
+      "step": 2334
+    },
+    {
+      "epoch": 1.9665918023582258,
+      "grad_norm": 0.20256203413009644,
+      "learning_rate": 3.1980950709086923e-06,
+      "loss": 0.3644,
+      "step": 2335
+    },
+    {
+      "epoch": 1.9674340258281864,
+      "grad_norm": 0.20262305438518524,
+      "learning_rate": 3.193522757485378e-06,
+      "loss": 0.3643,
+      "step": 2336
+    },
+    {
+      "epoch": 1.9682762492981472,
+      "grad_norm": 0.18650726974010468,
+      "learning_rate": 3.1889521808515888e-06,
+      "loss": 0.3278,
+      "step": 2337
+    },
+    {
+      "epoch": 1.9691184727681077,
+      "grad_norm": 0.18252995610237122,
+      "learning_rate": 3.1843833454015804e-06,
+      "loss": 0.357,
+      "step": 2338
+    },
+    {
+      "epoch": 1.9699606962380685,
+      "grad_norm": 0.2056460976600647,
+      "learning_rate": 3.179816255527941e-06,
+      "loss": 0.3342,
+      "step": 2339
+    },
+    {
+      "epoch": 1.9708029197080292,
+      "grad_norm": 0.21376261115074158,
+      "learning_rate": 3.1752509156215738e-06,
+      "loss": 0.3602,
+      "step": 2340
+    },
+    {
+      "epoch": 1.97164514317799,
+      "grad_norm": 0.18783794343471527,
+      "learning_rate": 3.1706873300717094e-06,
+      "loss": 0.3479,
+      "step": 2341
+    },
+    {
+      "epoch": 1.9724873666479505,
+      "grad_norm": 0.18589752912521362,
+      "learning_rate": 3.16612550326588e-06,
+      "loss": 0.3471,
+      "step": 2342
+    },
+    {
+      "epoch": 1.9733295901179113,
+      "grad_norm": 0.2014445960521698,
+      "learning_rate": 3.1615654395899377e-06,
+      "loss": 0.3662,
+      "step": 2343
+    },
+    {
+      "epoch": 1.9741718135878719,
+      "grad_norm": 0.19668903946876526,
+      "learning_rate": 3.1570071434280292e-06,
+      "loss": 0.3346,
+      "step": 2344
+    },
+    {
+      "epoch": 1.9750140370578326,
+      "grad_norm": 0.20105549693107605,
+      "learning_rate": 3.152450619162612e-06,
+      "loss": 0.3417,
+      "step": 2345
+    },
+    {
+      "epoch": 1.9758562605277934,
+      "grad_norm": 0.1995558738708496,
+      "learning_rate": 3.1478958711744324e-06,
+      "loss": 0.3672,
+      "step": 2346
+    },
+    {
+      "epoch": 1.9766984839977542,
+      "grad_norm": 0.19072867929935455,
+      "learning_rate": 3.1433429038425334e-06,
+      "loss": 0.3369,
+      "step": 2347
+    },
+    {
+      "epoch": 1.977540707467715,
+      "grad_norm": 0.20282776653766632,
+      "learning_rate": 3.1387917215442427e-06,
+      "loss": 0.3741,
+      "step": 2348
+    },
+    {
+      "epoch": 1.9783829309376755,
+      "grad_norm": 0.19809283316135406,
+      "learning_rate": 3.1342423286551756e-06,
+      "loss": 0.3668,
+      "step": 2349
+    },
+    {
+      "epoch": 1.979225154407636,
+      "grad_norm": 0.21220876276493073,
+      "learning_rate": 3.1296947295492226e-06,
+      "loss": 0.3635,
+      "step": 2350
+    },
+    {
+      "epoch": 1.9800673778775968,
+      "grad_norm": 0.21924105286598206,
+      "learning_rate": 3.125148928598554e-06,
+      "loss": 0.3749,
+      "step": 2351
+    },
+    {
+      "epoch": 1.9809096013475576,
+      "grad_norm": 0.19084849953651428,
+      "learning_rate": 3.120604930173608e-06,
+      "loss": 0.3156,
+      "step": 2352
+    },
+    {
+      "epoch": 1.9817518248175183,
+      "grad_norm": 0.2195017784833908,
+      "learning_rate": 3.116062738643092e-06,
+      "loss": 0.3913,
+      "step": 2353
+    },
+    {
+      "epoch": 1.982594048287479,
+      "grad_norm": 0.18741287291049957,
+      "learning_rate": 3.1115223583739746e-06,
+      "loss": 0.3305,
+      "step": 2354
+    },
+    {
+      "epoch": 1.9834362717574396,
+      "grad_norm": 0.226642906665802,
+      "learning_rate": 3.1069837937314846e-06,
+      "loss": 0.3552,
+      "step": 2355
+    },
+    {
+      "epoch": 1.9842784952274002,
+      "grad_norm": 0.2061690092086792,
+      "learning_rate": 3.1024470490791027e-06,
+      "loss": 0.3734,
+      "step": 2356
+    },
+    {
+      "epoch": 1.985120718697361,
+      "grad_norm": 0.18772165477275848,
+      "learning_rate": 3.097912128778563e-06,
+      "loss": 0.3295,
+      "step": 2357
+    },
+    {
+      "epoch": 1.9859629421673217,
+      "grad_norm": 0.18409810960292816,
+      "learning_rate": 3.093379037189842e-06,
+      "loss": 0.3459,
+      "step": 2358
+    },
+    {
+      "epoch": 1.9868051656372825,
+      "grad_norm": 0.1886821687221527,
+      "learning_rate": 3.0888477786711646e-06,
+      "loss": 0.3526,
+      "step": 2359
+    },
+    {
+      "epoch": 1.9876473891072433,
+      "grad_norm": 0.2152273952960968,
+      "learning_rate": 3.0843183575789824e-06,
+      "loss": 0.3898,
+      "step": 2360
+    },
+    {
+      "epoch": 1.9884896125772038,
+      "grad_norm": 0.17739993333816528,
+      "learning_rate": 3.0797907782679944e-06,
+      "loss": 0.3301,
+      "step": 2361
+    },
+    {
+      "epoch": 1.9893318360471643,
+      "grad_norm": 0.2642650306224823,
+      "learning_rate": 3.075265045091114e-06,
+      "loss": 0.3241,
+      "step": 2362
+    },
+    {
+      "epoch": 1.9901740595171251,
+      "grad_norm": 0.20823827385902405,
+      "learning_rate": 3.070741162399492e-06,
+      "loss": 0.415,
+      "step": 2363
+    },
+    {
+      "epoch": 1.9910162829870859,
+      "grad_norm": 0.20513373613357544,
+      "learning_rate": 3.0662191345424925e-06,
+      "loss": 0.3595,
+      "step": 2364
+    },
+    {
+      "epoch": 1.9918585064570467,
+      "grad_norm": 0.20934171974658966,
+      "learning_rate": 3.061698965867701e-06,
+      "loss": 0.3682,
+      "step": 2365
+    },
+    {
+      "epoch": 1.9927007299270074,
+      "grad_norm": 0.1961492896080017,
+      "learning_rate": 3.057180660720912e-06,
+      "loss": 0.3486,
+      "step": 2366
+    },
+    {
+      "epoch": 1.993542953396968,
+      "grad_norm": 0.1857389658689499,
+      "learning_rate": 3.0526642234461313e-06,
+      "loss": 0.3278,
+      "step": 2367
+    },
+    {
+      "epoch": 1.9943851768669287,
+      "grad_norm": 0.21698589622974396,
+      "learning_rate": 3.048149658385565e-06,
+      "loss": 0.3702,
+      "step": 2368
+    },
+    {
+      "epoch": 1.9952274003368893,
+      "grad_norm": 0.18934296071529388,
+      "learning_rate": 3.043636969879625e-06,
+      "loss": 0.359,
+      "step": 2369
+    },
+    {
+      "epoch": 1.99606962380685,
+      "grad_norm": 0.1986536830663681,
+      "learning_rate": 3.039126162266912e-06,
+      "loss": 0.3645,
+      "step": 2370
+    },
+    {
+      "epoch": 1.9969118472768108,
+      "grad_norm": 0.18482689559459686,
+      "learning_rate": 3.0346172398842254e-06,
+      "loss": 0.3557,
+      "step": 2371
+    },
+    {
+      "epoch": 1.9977540707467716,
+      "grad_norm": 0.19014687836170197,
+      "learning_rate": 3.0301102070665466e-06,
+      "loss": 0.3637,
+      "step": 2372
+    },
+    {
+      "epoch": 1.9985962942167321,
+      "grad_norm": 0.18836571276187897,
+      "learning_rate": 3.0256050681470446e-06,
+      "loss": 0.3267,
+      "step": 2373
+    },
+    {
+      "epoch": 1.999438517686693,
+      "grad_norm": 0.21261648833751678,
+      "learning_rate": 3.0211018274570625e-06,
+      "loss": 0.3664,
+      "step": 2374
+    },
+    {
+      "epoch": 2.0002807411566534,
+      "grad_norm": 0.4335678815841675,
+      "learning_rate": 3.0166004893261247e-06,
+      "loss": 0.5585,
+      "step": 2375
+    },
+    {
+      "epoch": 2.001122964626614,
+      "grad_norm": 0.18163718283176422,
+      "learning_rate": 3.012101058081919e-06,
+      "loss": 0.2978,
+      "step": 2376
+    },
+    {
+      "epoch": 2.001965188096575,
+      "grad_norm": 0.18798938393592834,
+      "learning_rate": 3.007603538050309e-06,
+      "loss": 0.3536,
+      "step": 2377
+    },
+    {
+      "epoch": 2.0028074115665357,
+      "grad_norm": 0.19393697381019592,
+      "learning_rate": 3.0031079335553097e-06,
+      "loss": 0.2928,
+      "step": 2378
+    },
+    {
+      "epoch": 2.0036496350364965,
+      "grad_norm": 0.22345486283302307,
+      "learning_rate": 2.9986142489191074e-06,
+      "loss": 0.3708,
+      "step": 2379
+    },
+    {
+      "epoch": 2.004491858506457,
+      "grad_norm": 0.20450833439826965,
+      "learning_rate": 2.994122488462029e-06,
+      "loss": 0.3271,
+      "step": 2380
+    },
+    {
+      "epoch": 2.0053340819764176,
+      "grad_norm": 0.18012557923793793,
+      "learning_rate": 2.989632656502564e-06,
+      "loss": 0.3147,
+      "step": 2381
+    },
+    {
+      "epoch": 2.0061763054463784,
+      "grad_norm": 0.1874862015247345,
+      "learning_rate": 2.9851447573573383e-06,
+      "loss": 0.3424,
+      "step": 2382
+    },
+    {
+      "epoch": 2.007018528916339,
+      "grad_norm": 0.20006859302520752,
+      "learning_rate": 2.980658795341125e-06,
+      "loss": 0.3219,
+      "step": 2383
+    },
+    {
+      "epoch": 2.0078607523863,
+      "grad_norm": 0.20945049822330475,
+      "learning_rate": 2.9761747747668314e-06,
+      "loss": 0.3499,
+      "step": 2384
+    },
+    {
+      "epoch": 2.0087029758562607,
+      "grad_norm": 0.20440226793289185,
+      "learning_rate": 2.971692699945502e-06,
+      "loss": 0.3626,
+      "step": 2385
+    },
+    {
+      "epoch": 2.0095451993262214,
+      "grad_norm": 0.19525331258773804,
+      "learning_rate": 2.9672125751863067e-06,
+      "loss": 0.3229,
+      "step": 2386
+    },
+    {
+      "epoch": 2.0103874227961818,
+      "grad_norm": 0.1915249079465866,
+      "learning_rate": 2.9627344047965433e-06,
+      "loss": 0.3042,
+      "step": 2387
+    },
+    {
+      "epoch": 2.0112296462661425,
+      "grad_norm": 0.2285671979188919,
+      "learning_rate": 2.958258193081629e-06,
+      "loss": 0.3975,
+      "step": 2388
+    },
+    {
+      "epoch": 2.0120718697361033,
+      "grad_norm": 0.17777401208877563,
+      "learning_rate": 2.9537839443451e-06,
+      "loss": 0.294,
+      "step": 2389
+    },
+    {
+      "epoch": 2.012914093206064,
+      "grad_norm": 0.21443428099155426,
+      "learning_rate": 2.949311662888601e-06,
+      "loss": 0.322,
+      "step": 2390
+    },
+    {
+      "epoch": 2.013756316676025,
+      "grad_norm": 0.205448180437088,
+      "learning_rate": 2.9448413530118912e-06,
+      "loss": 0.3371,
+      "step": 2391
+    },
+    {
+      "epoch": 2.0145985401459856,
+      "grad_norm": 0.2116105556488037,
+      "learning_rate": 2.94037301901283e-06,
+      "loss": 0.3382,
+      "step": 2392
+    },
+    {
+      "epoch": 2.015440763615946,
+      "grad_norm": 0.18419043719768524,
+      "learning_rate": 2.935906665187378e-06,
+      "loss": 0.3019,
+      "step": 2393
+    },
+    {
+      "epoch": 2.0162829870859067,
+      "grad_norm": 0.21632736921310425,
+      "learning_rate": 2.9314422958295906e-06,
+      "loss": 0.336,
+      "step": 2394
+    },
+    {
+      "epoch": 2.0171252105558675,
+      "grad_norm": 0.20205894112586975,
+      "learning_rate": 2.9269799152316226e-06,
+      "loss": 0.3463,
+      "step": 2395
+    },
+    {
+      "epoch": 2.0179674340258282,
+      "grad_norm": 0.20102480053901672,
+      "learning_rate": 2.922519527683706e-06,
+      "loss": 0.3536,
+      "step": 2396
+    },
+    {
+      "epoch": 2.018809657495789,
+      "grad_norm": 0.18754532933235168,
+      "learning_rate": 2.9180611374741623e-06,
+      "loss": 0.308,
+      "step": 2397
+    },
+    {
+      "epoch": 2.0196518809657498,
+      "grad_norm": 0.17986422777175903,
+      "learning_rate": 2.913604748889395e-06,
+      "loss": 0.3044,
+      "step": 2398
+    },
+    {
+      "epoch": 2.02049410443571,
+      "grad_norm": 0.20123037695884705,
+      "learning_rate": 2.9091503662138764e-06,
+      "loss": 0.3692,
+      "step": 2399
+    },
+    {
+      "epoch": 2.021336327905671,
+      "grad_norm": 0.1836530864238739,
+      "learning_rate": 2.904697993730159e-06,
+      "loss": 0.3328,
+      "step": 2400
+    },
+    {
+      "epoch": 2.0221785513756316,
+      "grad_norm": 0.19825614988803864,
+      "learning_rate": 2.900247635718856e-06,
+      "loss": 0.3433,
+      "step": 2401
+    },
+    {
+      "epoch": 2.0230207748455924,
+      "grad_norm": 0.17435987293720245,
+      "learning_rate": 2.8957992964586445e-06,
+      "loss": 0.2942,
+      "step": 2402
+    },
+    {
+      "epoch": 2.023862998315553,
+      "grad_norm": 0.19725970923900604,
+      "learning_rate": 2.891352980226262e-06,
+      "loss": 0.342,
+      "step": 2403
+    },
+    {
+      "epoch": 2.024705221785514,
+      "grad_norm": 0.16922511160373688,
+      "learning_rate": 2.886908691296504e-06,
+      "loss": 0.288,
+      "step": 2404
+    },
+    {
+      "epoch": 2.0255474452554743,
+      "grad_norm": 0.20383970439434052,
+      "learning_rate": 2.8824664339422115e-06,
+      "loss": 0.322,
+      "step": 2405
+    },
+    {
+      "epoch": 2.026389668725435,
+      "grad_norm": 0.1849295198917389,
+      "learning_rate": 2.8780262124342755e-06,
+      "loss": 0.2902,
+      "step": 2406
+    },
+    {
+      "epoch": 2.027231892195396,
+      "grad_norm": 0.20193511247634888,
+      "learning_rate": 2.873588031041627e-06,
+      "loss": 0.3766,
+      "step": 2407
+    },
+    {
+      "epoch": 2.0280741156653566,
+      "grad_norm": 0.19835780560970306,
+      "learning_rate": 2.8691518940312413e-06,
+      "loss": 0.3577,
+      "step": 2408
+    },
+    {
+      "epoch": 2.0289163391353173,
+      "grad_norm": 0.1865420639514923,
+      "learning_rate": 2.8647178056681197e-06,
+      "loss": 0.3205,
+      "step": 2409
+    },
+    {
+      "epoch": 2.029758562605278,
+      "grad_norm": 0.19542618095874786,
+      "learning_rate": 2.8602857702153054e-06,
+      "loss": 0.3441,
+      "step": 2410
+    },
+    {
+      "epoch": 2.0306007860752384,
+      "grad_norm": 0.20656581223011017,
+      "learning_rate": 2.8558557919338537e-06,
+      "loss": 0.3295,
+      "step": 2411
+    },
+    {
+      "epoch": 2.031443009545199,
+      "grad_norm": 0.20013107359409332,
+      "learning_rate": 2.8514278750828537e-06,
+      "loss": 0.3104,
+      "step": 2412
+    },
+    {
+      "epoch": 2.03228523301516,
+      "grad_norm": 0.19680936634540558,
+      "learning_rate": 2.847002023919406e-06,
+      "loss": 0.3641,
+      "step": 2413
+    },
+    {
+      "epoch": 2.0331274564851207,
+      "grad_norm": 0.20290818810462952,
+      "learning_rate": 2.8425782426986304e-06,
+      "loss": 0.3606,
+      "step": 2414
+    },
+    {
+      "epoch": 2.0339696799550815,
+      "grad_norm": 0.2051219791173935,
+      "learning_rate": 2.838156535673652e-06,
+      "loss": 0.3302,
+      "step": 2415
+    },
+    {
+      "epoch": 2.0348119034250423,
+      "grad_norm": 0.1948593556880951,
+      "learning_rate": 2.833736907095604e-06,
+      "loss": 0.3358,
+      "step": 2416
+    },
+    {
+      "epoch": 2.035654126895003,
+      "grad_norm": 0.19389194250106812,
+      "learning_rate": 2.8293193612136183e-06,
+      "loss": 0.3352,
+      "step": 2417
+    },
+    {
+      "epoch": 2.0364963503649633,
+      "grad_norm": 0.19726473093032837,
+      "learning_rate": 2.8249039022748315e-06,
+      "loss": 0.3391,
+      "step": 2418
+    },
+    {
+      "epoch": 2.037338573834924,
+      "grad_norm": 0.19745996594429016,
+      "learning_rate": 2.8204905345243664e-06,
+      "loss": 0.3313,
+      "step": 2419
+    },
+    {
+      "epoch": 2.038180797304885,
+      "grad_norm": 0.18009109795093536,
+      "learning_rate": 2.816079262205339e-06,
+      "loss": 0.3175,
+      "step": 2420
+    },
+    {
+      "epoch": 2.0390230207748457,
+      "grad_norm": 0.18453653156757355,
+      "learning_rate": 2.8116700895588473e-06,
+      "loss": 0.3362,
+      "step": 2421
+    },
+    {
+      "epoch": 2.0398652442448064,
+      "grad_norm": 0.18108515441417694,
+      "learning_rate": 2.807263020823977e-06,
+      "loss": 0.3319,
+      "step": 2422
+    },
+    {
+      "epoch": 2.040707467714767,
+      "grad_norm": 0.18968205153942108,
+      "learning_rate": 2.8028580602377852e-06,
+      "loss": 0.3198,
+      "step": 2423
+    },
+    {
+      "epoch": 2.0415496911847275,
+      "grad_norm": 0.20090173184871674,
+      "learning_rate": 2.798455212035305e-06,
+      "loss": 0.3436,
+      "step": 2424
+    },
+    {
+      "epoch": 2.0423919146546883,
+      "grad_norm": 0.2093648612499237,
+      "learning_rate": 2.7940544804495345e-06,
+      "loss": 0.3239,
+      "step": 2425
+    },
+    {
+      "epoch": 2.043234138124649,
+      "grad_norm": 0.18286184966564178,
+      "learning_rate": 2.789655869711445e-06,
+      "loss": 0.298,
+      "step": 2426
+    },
+    {
+      "epoch": 2.04407636159461,
+      "grad_norm": 0.21552641689777374,
+      "learning_rate": 2.785259384049959e-06,
+      "loss": 0.3602,
+      "step": 2427
+    },
+    {
+      "epoch": 2.0449185850645706,
+      "grad_norm": 0.1874585598707199,
+      "learning_rate": 2.780865027691968e-06,
+      "loss": 0.3115,
+      "step": 2428
+    },
+    {
+      "epoch": 2.0457608085345313,
+      "grad_norm": 0.18681229650974274,
+      "learning_rate": 2.7764728048623003e-06,
+      "loss": 0.348,
+      "step": 2429
+    },
+    {
+      "epoch": 2.0466030320044917,
+      "grad_norm": 0.18479189276695251,
+      "learning_rate": 2.7720827197837475e-06,
+      "loss": 0.3516,
+      "step": 2430
+    },
+    {
+      "epoch": 2.0474452554744524,
+      "grad_norm": 0.20195545256137848,
+      "learning_rate": 2.7676947766770367e-06,
+      "loss": 0.3282,
+      "step": 2431
+    },
+    {
+      "epoch": 2.048287478944413,
+      "grad_norm": 0.19953538477420807,
+      "learning_rate": 2.7633089797608435e-06,
+      "loss": 0.3537,
+      "step": 2432
+    },
+    {
+      "epoch": 2.049129702414374,
+      "grad_norm": 0.17841529846191406,
+      "learning_rate": 2.7589253332517736e-06,
+      "loss": 0.314,
+      "step": 2433
+    },
+    {
+      "epoch": 2.0499719258843347,
+      "grad_norm": 0.2024816870689392,
+      "learning_rate": 2.7545438413643666e-06,
+      "loss": 0.3343,
+      "step": 2434
+    },
+    {
+      "epoch": 2.0508141493542955,
+      "grad_norm": 0.2072976976633072,
+      "learning_rate": 2.7501645083110893e-06,
+      "loss": 0.3357,
+      "step": 2435
+    },
+    {
+      "epoch": 2.051656372824256,
+      "grad_norm": 0.19924525916576385,
+      "learning_rate": 2.745787338302341e-06,
+      "loss": 0.3382,
+      "step": 2436
+    },
+    {
+      "epoch": 2.0524985962942166,
+      "grad_norm": 0.18899208307266235,
+      "learning_rate": 2.741412335546431e-06,
+      "loss": 0.3391,
+      "step": 2437
+    },
+    {
+      "epoch": 2.0533408197641774,
+      "grad_norm": 0.183904767036438,
+      "learning_rate": 2.7370395042495913e-06,
+      "loss": 0.3212,
+      "step": 2438
+    },
+    {
+      "epoch": 2.054183043234138,
+      "grad_norm": 0.17699691653251648,
+      "learning_rate": 2.7326688486159613e-06,
+      "loss": 0.3036,
+      "step": 2439
+    },
+    {
+      "epoch": 2.055025266704099,
+      "grad_norm": 0.19395457208156586,
+      "learning_rate": 2.7283003728475952e-06,
+      "loss": 0.3218,
+      "step": 2440
+    },
+    {
+      "epoch": 2.0558674901740597,
+      "grad_norm": 0.19687485694885254,
+      "learning_rate": 2.7239340811444476e-06,
+      "loss": 0.3498,
+      "step": 2441
+    },
+    {
+      "epoch": 2.05670971364402,
+      "grad_norm": 0.18807673454284668,
+      "learning_rate": 2.7195699777043723e-06,
+      "loss": 0.3553,
+      "step": 2442
+    },
+    {
+      "epoch": 2.0575519371139808,
+      "grad_norm": 0.1877565085887909,
+      "learning_rate": 2.7152080667231185e-06,
+      "loss": 0.3251,
+      "step": 2443
+    },
+    {
+      "epoch": 2.0583941605839415,
+      "grad_norm": 0.19361265003681183,
+      "learning_rate": 2.710848352394334e-06,
+      "loss": 0.3404,
+      "step": 2444
+    },
+    {
+      "epoch": 2.0592363840539023,
+      "grad_norm": 0.1828143149614334,
+      "learning_rate": 2.706490838909547e-06,
+      "loss": 0.2938,
+      "step": 2445
+    },
+    {
+      "epoch": 2.060078607523863,
+      "grad_norm": 0.20654404163360596,
+      "learning_rate": 2.7021355304581765e-06,
+      "loss": 0.3361,
+      "step": 2446
+    },
+    {
+      "epoch": 2.060920830993824,
+      "grad_norm": 0.2480696141719818,
+      "learning_rate": 2.6977824312275123e-06,
+      "loss": 0.3332,
+      "step": 2447
+    },
+    {
+      "epoch": 2.0617630544637846,
+      "grad_norm": 0.18790100514888763,
+      "learning_rate": 2.6934315454027323e-06,
+      "loss": 0.3492,
+      "step": 2448
+    },
+    {
+      "epoch": 2.062605277933745,
+      "grad_norm": 0.18553946912288666,
+      "learning_rate": 2.6890828771668742e-06,
+      "loss": 0.3542,
+      "step": 2449
+    },
+    {
+      "epoch": 2.0634475014037057,
+      "grad_norm": 0.1775999516248703,
+      "learning_rate": 2.684736430700854e-06,
+      "loss": 0.3335,
+      "step": 2450
+    },
+    {
+      "epoch": 2.0642897248736665,
+      "grad_norm": 0.1797809600830078,
+      "learning_rate": 2.680392210183446e-06,
+      "loss": 0.324,
+      "step": 2451
+    },
+    {
+      "epoch": 2.0651319483436272,
+      "grad_norm": 0.1839604526758194,
+      "learning_rate": 2.6760502197912842e-06,
+      "loss": 0.3312,
+      "step": 2452
+    },
+    {
+      "epoch": 2.065974171813588,
+      "grad_norm": 0.20131711661815643,
+      "learning_rate": 2.671710463698859e-06,
+      "loss": 0.3376,
+      "step": 2453
+    },
+    {
+      "epoch": 2.0668163952835488,
+      "grad_norm": 0.1908835917711258,
+      "learning_rate": 2.6673729460785174e-06,
+      "loss": 0.2927,
+      "step": 2454
+    },
+    {
+      "epoch": 2.067658618753509,
+      "grad_norm": 0.1810787320137024,
+      "learning_rate": 2.663037671100448e-06,
+      "loss": 0.3633,
+      "step": 2455
+    },
+    {
+      "epoch": 2.06850084222347,
+      "grad_norm": 0.17869506776332855,
+      "learning_rate": 2.6587046429326855e-06,
+      "loss": 0.3156,
+      "step": 2456
+    },
+    {
+      "epoch": 2.0693430656934306,
+      "grad_norm": 0.17782184481620789,
+      "learning_rate": 2.6543738657411033e-06,
+      "loss": 0.3061,
+      "step": 2457
+    },
+    {
+      "epoch": 2.0701852891633914,
+      "grad_norm": 0.19159172475337982,
+      "learning_rate": 2.6500453436894157e-06,
+      "loss": 0.3426,
+      "step": 2458
+    },
+    {
+      "epoch": 2.071027512633352,
+      "grad_norm": 0.1826389580965042,
+      "learning_rate": 2.6457190809391627e-06,
+      "loss": 0.3113,
+      "step": 2459
+    },
+    {
+      "epoch": 2.071869736103313,
+      "grad_norm": 0.17592735588550568,
+      "learning_rate": 2.6413950816497146e-06,
+      "loss": 0.344,
+      "step": 2460
+    },
+    {
+      "epoch": 2.0727119595732733,
+      "grad_norm": 0.18362398445606232,
+      "learning_rate": 2.6370733499782654e-06,
+      "loss": 0.3353,
+      "step": 2461
+    },
+    {
+      "epoch": 2.073554183043234,
+      "grad_norm": 0.18338549137115479,
+      "learning_rate": 2.6327538900798306e-06,
+      "loss": 0.2975,
+      "step": 2462
+    },
+    {
+      "epoch": 2.074396406513195,
+      "grad_norm": 0.18559111654758453,
+      "learning_rate": 2.628436706107238e-06,
+      "loss": 0.3489,
+      "step": 2463
+    },
+    {
+      "epoch": 2.0752386299831556,
+      "grad_norm": 0.20324550569057465,
+      "learning_rate": 2.6241218022111336e-06,
+      "loss": 0.3407,
+      "step": 2464
+    },
+    {
+      "epoch": 2.0760808534531163,
+      "grad_norm": 0.19184359908103943,
+      "learning_rate": 2.6198091825399606e-06,
+      "loss": 0.3328,
+      "step": 2465
+    },
+    {
+      "epoch": 2.076923076923077,
+      "grad_norm": 0.19564636051654816,
+      "learning_rate": 2.6154988512399784e-06,
+      "loss": 0.3096,
+      "step": 2466
+    },
+    {
+      "epoch": 2.0777653003930374,
+      "grad_norm": 0.18796582520008087,
+      "learning_rate": 2.6111908124552355e-06,
+      "loss": 0.3161,
+      "step": 2467
+    },
+    {
+      "epoch": 2.078607523862998,
+      "grad_norm": 0.18963384628295898,
+      "learning_rate": 2.6068850703275856e-06,
+      "loss": 0.3167,
+      "step": 2468
+    },
+    {
+      "epoch": 2.079449747332959,
+      "grad_norm": 0.20601922273635864,
+      "learning_rate": 2.6025816289966703e-06,
+      "loss": 0.3675,
+      "step": 2469
+    },
+    {
+      "epoch": 2.0802919708029197,
+      "grad_norm": 0.17115628719329834,
+      "learning_rate": 2.598280492599913e-06,
+      "loss": 0.3148,
+      "step": 2470
+    },
+    {
+      "epoch": 2.0811341942728805,
+      "grad_norm": 0.1847679615020752,
+      "learning_rate": 2.5939816652725324e-06,
+      "loss": 0.337,
+      "step": 2471
+    },
+    {
+      "epoch": 2.0819764177428413,
+      "grad_norm": 0.1880049705505371,
+      "learning_rate": 2.5896851511475184e-06,
+      "loss": 0.3443,
+      "step": 2472
+    },
+    {
+      "epoch": 2.0828186412128016,
+      "grad_norm": 0.1756199449300766,
+      "learning_rate": 2.5853909543556444e-06,
+      "loss": 0.2916,
+      "step": 2473
+    },
+    {
+      "epoch": 2.0836608646827623,
+      "grad_norm": 0.19857759773731232,
+      "learning_rate": 2.5810990790254486e-06,
+      "loss": 0.352,
+      "step": 2474
+    },
+    {
+      "epoch": 2.084503088152723,
+      "grad_norm": 0.17780746519565582,
+      "learning_rate": 2.5768095292832412e-06,
+      "loss": 0.3154,
+      "step": 2475
+    },
+    {
+      "epoch": 2.085345311622684,
+      "grad_norm": 0.1851738542318344,
+      "learning_rate": 2.5725223092530937e-06,
+      "loss": 0.33,
+      "step": 2476
+    },
+    {
+      "epoch": 2.0861875350926447,
+      "grad_norm": 0.18334592878818512,
+      "learning_rate": 2.568237423056844e-06,
+      "loss": 0.3127,
+      "step": 2477
+    },
+    {
+      "epoch": 2.0870297585626054,
+      "grad_norm": 0.19777873158454895,
+      "learning_rate": 2.5639548748140803e-06,
+      "loss": 0.3236,
+      "step": 2478
+    },
+    {
+      "epoch": 2.087871982032566,
+      "grad_norm": 0.19690145552158356,
+      "learning_rate": 2.5596746686421436e-06,
+      "loss": 0.3272,
+      "step": 2479
+    },
+    {
+      "epoch": 2.0887142055025265,
+      "grad_norm": 0.20011959969997406,
+      "learning_rate": 2.5553968086561244e-06,
+      "loss": 0.3663,
+      "step": 2480
+    },
+    {
+      "epoch": 2.0895564289724873,
+      "grad_norm": 0.18417225778102875,
+      "learning_rate": 2.5511212989688587e-06,
+      "loss": 0.2957,
+      "step": 2481
+    },
+    {
+      "epoch": 2.090398652442448,
+      "grad_norm": 0.18653437495231628,
+      "learning_rate": 2.546848143690922e-06,
+      "loss": 0.3384,
+      "step": 2482
+    },
+    {
+      "epoch": 2.091240875912409,
+      "grad_norm": 0.1805136352777481,
+      "learning_rate": 2.5425773469306247e-06,
+      "loss": 0.3047,
+      "step": 2483
+    },
+    {
+      "epoch": 2.0920830993823696,
+      "grad_norm": 0.19548816978931427,
+      "learning_rate": 2.5383089127940087e-06,
+      "loss": 0.3162,
+      "step": 2484
+    },
+    {
+      "epoch": 2.0929253228523303,
+      "grad_norm": 0.1792110949754715,
+      "learning_rate": 2.534042845384851e-06,
+      "loss": 0.2979,
+      "step": 2485
+    },
+    {
+      "epoch": 2.0937675463222907,
+      "grad_norm": 0.19322973489761353,
+      "learning_rate": 2.5297791488046445e-06,
+      "loss": 0.3261,
+      "step": 2486
+    },
+    {
+      "epoch": 2.0946097697922514,
+      "grad_norm": 0.19382603466510773,
+      "learning_rate": 2.525517827152614e-06,
+      "loss": 0.3536,
+      "step": 2487
+    },
+    {
+      "epoch": 2.095451993262212,
+      "grad_norm": 0.18523024022579193,
+      "learning_rate": 2.5212588845256837e-06,
+      "loss": 0.3117,
+      "step": 2488
+    },
+    {
+      "epoch": 2.096294216732173,
+      "grad_norm": 0.2027691900730133,
+      "learning_rate": 2.517002325018508e-06,
+      "loss": 0.32,
+      "step": 2489
+    },
+    {
+      "epoch": 2.0971364402021337,
+      "grad_norm": 0.19315862655639648,
+      "learning_rate": 2.5127481527234397e-06,
+      "loss": 0.3182,
+      "step": 2490
+    },
+    {
+      "epoch": 2.0979786636720945,
+      "grad_norm": 0.20692229270935059,
+      "learning_rate": 2.508496371730543e-06,
+      "loss": 0.4019,
+      "step": 2491
+    },
+    {
+      "epoch": 2.098820887142055,
+      "grad_norm": 0.16789032518863678,
+      "learning_rate": 2.5042469861275768e-06,
+      "loss": 0.2814,
+      "step": 2492
+    },
+    {
+      "epoch": 2.0996631106120156,
+      "grad_norm": 0.1847231537103653,
+      "learning_rate": 2.5000000000000015e-06,
+      "loss": 0.3051,
+      "step": 2493
+    },
+    {
+      "epoch": 2.1005053340819764,
+      "grad_norm": 0.1905345767736435,
+      "learning_rate": 2.4957554174309655e-06,
+      "loss": 0.3466,
+      "step": 2494
+    },
+    {
+      "epoch": 2.101347557551937,
+      "grad_norm": 0.20040138065814972,
+      "learning_rate": 2.491513242501315e-06,
+      "loss": 0.3337,
+      "step": 2495
+    },
+    {
+      "epoch": 2.102189781021898,
+      "grad_norm": 0.1926167607307434,
+      "learning_rate": 2.487273479289574e-06,
+      "loss": 0.31,
+      "step": 2496
+    },
+    {
+      "epoch": 2.1030320044918587,
+      "grad_norm": 0.18961243331432343,
+      "learning_rate": 2.4830361318719493e-06,
+      "loss": 0.343,
+      "step": 2497
+    },
+    {
+      "epoch": 2.103874227961819,
+      "grad_norm": 0.16875000298023224,
+      "learning_rate": 2.4788012043223253e-06,
+      "loss": 0.2956,
+      "step": 2498
+    },
+    {
+      "epoch": 2.1047164514317798,
+      "grad_norm": 0.23073846101760864,
+      "learning_rate": 2.4745687007122636e-06,
+      "loss": 0.3803,
+      "step": 2499
+    },
+    {
+      "epoch": 2.1055586749017405,
+      "grad_norm": 0.2010485678911209,
+      "learning_rate": 2.470338625110991e-06,
+      "loss": 0.3403,
+      "step": 2500
+    },
+    {
+      "epoch": 2.1064008983717013,
+      "grad_norm": 0.18841859698295593,
+      "learning_rate": 2.4661109815854005e-06,
+      "loss": 0.3306,
+      "step": 2501
+    },
+    {
+      "epoch": 2.107243121841662,
+      "grad_norm": 0.19118495285511017,
+      "learning_rate": 2.4618857742000463e-06,
+      "loss": 0.3179,
+      "step": 2502
+    },
+    {
+      "epoch": 2.108085345311623,
+      "grad_norm": 0.19626981019973755,
+      "learning_rate": 2.4576630070171447e-06,
+      "loss": 0.3228,
+      "step": 2503
+    },
+    {
+      "epoch": 2.108927568781583,
+      "grad_norm": 0.18817634880542755,
+      "learning_rate": 2.4534426840965604e-06,
+      "loss": 0.3178,
+      "step": 2504
+    },
+    {
+      "epoch": 2.109769792251544,
+      "grad_norm": 0.1974536031484604,
+      "learning_rate": 2.449224809495815e-06,
+      "loss": 0.3492,
+      "step": 2505
+    },
+    {
+      "epoch": 2.1106120157215047,
+      "grad_norm": 0.17316868901252747,
+      "learning_rate": 2.4450093872700648e-06,
+      "loss": 0.3129,
+      "step": 2506
+    },
+    {
+      "epoch": 2.1114542391914655,
+      "grad_norm": 0.1934424191713333,
+      "learning_rate": 2.440796421472122e-06,
+      "loss": 0.341,
+      "step": 2507
+    },
+    {
+      "epoch": 2.1122964626614262,
+      "grad_norm": 0.17776821553707123,
+      "learning_rate": 2.436585916152426e-06,
+      "loss": 0.336,
+      "step": 2508
+    },
+    {
+      "epoch": 2.113138686131387,
+      "grad_norm": 0.18335790932178497,
+      "learning_rate": 2.4323778753590582e-06,
+      "loss": 0.356,
+      "step": 2509
+    },
+    {
+      "epoch": 2.1139809096013478,
+      "grad_norm": 0.18221352994441986,
+      "learning_rate": 2.4281723031377275e-06,
+      "loss": 0.3176,
+      "step": 2510
+    },
+    {
+      "epoch": 2.114823133071308,
+      "grad_norm": 0.19676458835601807,
+      "learning_rate": 2.423969203531768e-06,
+      "loss": 0.3308,
+      "step": 2511
+    },
+    {
+      "epoch": 2.115665356541269,
+      "grad_norm": 0.18670789897441864,
+      "learning_rate": 2.419768580582137e-06,
+      "loss": 0.3109,
+      "step": 2512
+    },
+    {
+      "epoch": 2.1165075800112296,
+      "grad_norm": 0.19073185324668884,
+      "learning_rate": 2.4155704383274154e-06,
+      "loss": 0.343,
+      "step": 2513
+    },
+    {
+      "epoch": 2.1173498034811904,
+      "grad_norm": 0.20621013641357422,
+      "learning_rate": 2.411374780803793e-06,
+      "loss": 0.36,
+      "step": 2514
+    },
+    {
+      "epoch": 2.118192026951151,
+      "grad_norm": 0.19015243649482727,
+      "learning_rate": 2.4071816120450742e-06,
+      "loss": 0.3454,
+      "step": 2515
+    },
+    {
+      "epoch": 2.119034250421112,
+      "grad_norm": 0.17132285237312317,
+      "learning_rate": 2.402990936082667e-06,
+      "loss": 0.3016,
+      "step": 2516
+    },
+    {
+      "epoch": 2.1198764738910723,
+      "grad_norm": 0.18261852860450745,
+      "learning_rate": 2.3988027569455895e-06,
+      "loss": 0.355,
+      "step": 2517
+    },
+    {
+      "epoch": 2.120718697361033,
+      "grad_norm": 0.18323764204978943,
+      "learning_rate": 2.3946170786604526e-06,
+      "loss": 0.3355,
+      "step": 2518
+    },
+    {
+      "epoch": 2.121560920830994,
+      "grad_norm": 0.18560270965099335,
+      "learning_rate": 2.390433905251467e-06,
+      "loss": 0.3491,
+      "step": 2519
+    },
+    {
+      "epoch": 2.1224031443009546,
+      "grad_norm": 0.19938194751739502,
+      "learning_rate": 2.3862532407404306e-06,
+      "loss": 0.3359,
+      "step": 2520
+    },
+    {
+      "epoch": 2.1232453677709153,
+      "grad_norm": 0.17980343103408813,
+      "learning_rate": 2.3820750891467355e-06,
+      "loss": 0.3232,
+      "step": 2521
+    },
+    {
+      "epoch": 2.124087591240876,
+      "grad_norm": 0.19313830137252808,
+      "learning_rate": 2.377899454487351e-06,
+      "loss": 0.328,
+      "step": 2522
+    },
+    {
+      "epoch": 2.1249298147108364,
+      "grad_norm": 0.17931406199932098,
+      "learning_rate": 2.373726340776837e-06,
+      "loss": 0.3344,
+      "step": 2523
+    },
+    {
+      "epoch": 2.125772038180797,
+      "grad_norm": 0.20262593030929565,
+      "learning_rate": 2.369555752027313e-06,
+      "loss": 0.3525,
+      "step": 2524
+    },
+    {
+      "epoch": 2.126614261650758,
+      "grad_norm": 0.20063352584838867,
+      "learning_rate": 2.365387692248488e-06,
+      "loss": 0.3515,
+      "step": 2525
+    },
+    {
+      "epoch": 2.1274564851207187,
+      "grad_norm": 0.2100335657596588,
+      "learning_rate": 2.361222165447628e-06,
+      "loss": 0.3118,
+      "step": 2526
+    },
+    {
+      "epoch": 2.1282987085906795,
+      "grad_norm": 0.1892918199300766,
+      "learning_rate": 2.3570591756295717e-06,
+      "loss": 0.3397,
+      "step": 2527
+    },
+    {
+      "epoch": 2.1291409320606403,
+      "grad_norm": 0.1774381697177887,
+      "learning_rate": 2.3528987267967135e-06,
+      "loss": 0.2957,
+      "step": 2528
+    },
+    {
+      "epoch": 2.1299831555306006,
+      "grad_norm": 0.1967756599187851,
+      "learning_rate": 2.348740822949006e-06,
+      "loss": 0.3509,
+      "step": 2529
+    },
+    {
+      "epoch": 2.1308253790005613,
+      "grad_norm": 0.1812041848897934,
+      "learning_rate": 2.3445854680839534e-06,
+      "loss": 0.3119,
+      "step": 2530
+    },
+    {
+      "epoch": 2.131667602470522,
+      "grad_norm": 0.16971538960933685,
+      "learning_rate": 2.3404326661966148e-06,
+      "loss": 0.3074,
+      "step": 2531
+    },
+    {
+      "epoch": 2.132509825940483,
+      "grad_norm": 0.18924599885940552,
+      "learning_rate": 2.33628242127959e-06,
+      "loss": 0.3577,
+      "step": 2532
+    },
+    {
+      "epoch": 2.1333520494104437,
+      "grad_norm": 0.18778833746910095,
+      "learning_rate": 2.33213473732302e-06,
+      "loss": 0.3468,
+      "step": 2533
+    },
+    {
+      "epoch": 2.1341942728804044,
+      "grad_norm": 0.1919918656349182,
+      "learning_rate": 2.3279896183145857e-06,
+      "loss": 0.3034,
+      "step": 2534
+    },
+    {
+      "epoch": 2.1350364963503647,
+      "grad_norm": 0.180246502161026,
+      "learning_rate": 2.323847068239504e-06,
+      "loss": 0.3239,
+      "step": 2535
+    },
+    {
+      "epoch": 2.1358787198203255,
+      "grad_norm": 0.1909414827823639,
+      "learning_rate": 2.319707091080517e-06,
+      "loss": 0.3546,
+      "step": 2536
+    },
+    {
+      "epoch": 2.1367209432902863,
+      "grad_norm": 0.20118744671344757,
+      "learning_rate": 2.3155696908178974e-06,
+      "loss": 0.328,
+      "step": 2537
+    },
+    {
+      "epoch": 2.137563166760247,
+      "grad_norm": 0.18613122403621674,
+      "learning_rate": 2.3114348714294355e-06,
+      "loss": 0.3143,
+      "step": 2538
+    },
+    {
+      "epoch": 2.138405390230208,
+      "grad_norm": 0.19210347533226013,
+      "learning_rate": 2.3073026368904478e-06,
+      "loss": 0.3156,
+      "step": 2539
+    },
+    {
+      "epoch": 2.1392476137001686,
+      "grad_norm": 0.17713823914527893,
+      "learning_rate": 2.3031729911737576e-06,
+      "loss": 0.3152,
+      "step": 2540
+    },
+    {
+      "epoch": 2.1400898371701293,
+      "grad_norm": 0.19556507468223572,
+      "learning_rate": 2.2990459382497086e-06,
+      "loss": 0.3505,
+      "step": 2541
+    },
+    {
+      "epoch": 2.1409320606400897,
+      "grad_norm": 0.18178249895572662,
+      "learning_rate": 2.2949214820861403e-06,
+      "loss": 0.3466,
+      "step": 2542
+    },
+    {
+      "epoch": 2.1417742841100504,
+      "grad_norm": 0.17022491991519928,
+      "learning_rate": 2.290799626648402e-06,
+      "loss": 0.3005,
+      "step": 2543
+    },
+    {
+      "epoch": 2.142616507580011,
+      "grad_norm": 0.20459839701652527,
+      "learning_rate": 2.2866803758993446e-06,
+      "loss": 0.3601,
+      "step": 2544
+    },
+    {
+      "epoch": 2.143458731049972,
+      "grad_norm": 0.19675502181053162,
+      "learning_rate": 2.2825637337993094e-06,
+      "loss": 0.3271,
+      "step": 2545
+    },
+    {
+      "epoch": 2.1443009545199327,
+      "grad_norm": 0.21750758588314056,
+      "learning_rate": 2.2784497043061384e-06,
+      "loss": 0.3476,
+      "step": 2546
+    },
+    {
+      "epoch": 2.1451431779898935,
+      "grad_norm": 0.18072174489498138,
+      "learning_rate": 2.274338291375147e-06,
+      "loss": 0.3143,
+      "step": 2547
+    },
+    {
+      "epoch": 2.145985401459854,
+      "grad_norm": 0.1931690126657486,
+      "learning_rate": 2.2702294989591513e-06,
+      "loss": 0.362,
+      "step": 2548
+    },
+    {
+      "epoch": 2.1468276249298146,
+      "grad_norm": 0.18842744827270508,
+      "learning_rate": 2.266123331008436e-06,
+      "loss": 0.3127,
+      "step": 2549
+    },
+    {
+      "epoch": 2.1476698483997754,
+      "grad_norm": 0.22671958804130554,
+      "learning_rate": 2.262019791470772e-06,
+      "loss": 0.3548,
+      "step": 2550
+    },
+    {
+      "epoch": 2.148512071869736,
+      "grad_norm": 0.19913806021213531,
+      "learning_rate": 2.257918884291392e-06,
+      "loss": 0.3126,
+      "step": 2551
+    },
+    {
+      "epoch": 2.149354295339697,
+      "grad_norm": 0.17924943566322327,
+      "learning_rate": 2.253820613413009e-06,
+      "loss": 0.3431,
+      "step": 2552
+    },
+    {
+      "epoch": 2.1501965188096577,
+      "grad_norm": 0.19935238361358643,
+      "learning_rate": 2.2497249827757933e-06,
+      "loss": 0.3337,
+      "step": 2553
+    },
+    {
+      "epoch": 2.151038742279618,
+      "grad_norm": 0.18192946910858154,
+      "learning_rate": 2.245631996317384e-06,
+      "loss": 0.2829,
+      "step": 2554
+    },
+    {
+      "epoch": 2.1518809657495788,
+      "grad_norm": 0.20926830172538757,
+      "learning_rate": 2.2415416579728714e-06,
+      "loss": 0.37,
+      "step": 2555
+    },
+    {
+      "epoch": 2.1527231892195395,
+      "grad_norm": 0.18061847984790802,
+      "learning_rate": 2.2374539716748034e-06,
+      "loss": 0.3359,
+      "step": 2556
+    },
+    {
+      "epoch": 2.1535654126895003,
+      "grad_norm": 0.1798250526189804,
+      "learning_rate": 2.233368941353175e-06,
+      "loss": 0.3111,
+      "step": 2557
+    },
+    {
+      "epoch": 2.154407636159461,
+      "grad_norm": 0.1831042319536209,
+      "learning_rate": 2.2292865709354346e-06,
+      "loss": 0.337,
+      "step": 2558
+    },
+    {
+      "epoch": 2.155249859629422,
+      "grad_norm": 0.18504156172275543,
+      "learning_rate": 2.225206864346465e-06,
+      "loss": 0.331,
+      "step": 2559
+    },
+    {
+      "epoch": 2.156092083099382,
+      "grad_norm": 0.2094961702823639,
+      "learning_rate": 2.221129825508593e-06,
+      "loss": 0.3424,
+      "step": 2560
+    },
+    {
+      "epoch": 2.156934306569343,
+      "grad_norm": 0.16761691868305206,
+      "learning_rate": 2.2170554583415782e-06,
+      "loss": 0.2776,
+      "step": 2561
+    },
+    {
+      "epoch": 2.1577765300393037,
+      "grad_norm": 0.17078322172164917,
+      "learning_rate": 2.2129837667626147e-06,
+      "loss": 0.348,
+      "step": 2562
+    },
+    {
+      "epoch": 2.1586187535092645,
+      "grad_norm": 0.18788252770900726,
+      "learning_rate": 2.2089147546863187e-06,
+      "loss": 0.3638,
+      "step": 2563
+    },
+    {
+      "epoch": 2.1594609769792252,
+      "grad_norm": 0.1982835978269577,
+      "learning_rate": 2.20484842602474e-06,
+      "loss": 0.3234,
+      "step": 2564
+    },
+    {
+      "epoch": 2.160303200449186,
+      "grad_norm": 0.18814495205879211,
+      "learning_rate": 2.2007847846873342e-06,
+      "loss": 0.3536,
+      "step": 2565
+    },
+    {
+      "epoch": 2.1611454239191463,
+      "grad_norm": 0.16890032589435577,
+      "learning_rate": 2.196723834580987e-06,
+      "loss": 0.305,
+      "step": 2566
+    },
+    {
+      "epoch": 2.161987647389107,
+      "grad_norm": 0.1856263428926468,
+      "learning_rate": 2.1926655796099873e-06,
+      "loss": 0.3459,
+      "step": 2567
+    },
+    {
+      "epoch": 2.162829870859068,
+      "grad_norm": 0.19124473631381989,
+      "learning_rate": 2.188610023676041e-06,
+      "loss": 0.3144,
+      "step": 2568
+    },
+    {
+      "epoch": 2.1636720943290286,
+      "grad_norm": 0.21970148384571075,
+      "learning_rate": 2.1845571706782486e-06,
+      "loss": 0.3411,
+      "step": 2569
+    },
+    {
+      "epoch": 2.1645143177989894,
+      "grad_norm": 0.20076794922351837,
+      "learning_rate": 2.1805070245131234e-06,
+      "loss": 0.3265,
+      "step": 2570
+    },
+    {
+      "epoch": 2.16535654126895,
+      "grad_norm": 0.18565763533115387,
+      "learning_rate": 2.176459589074566e-06,
+      "loss": 0.2936,
+      "step": 2571
+    },
+    {
+      "epoch": 2.166198764738911,
+      "grad_norm": 0.20779716968536377,
+      "learning_rate": 2.17241486825388e-06,
+      "loss": 0.3521,
+      "step": 2572
+    },
+    {
+      "epoch": 2.1670409882088713,
+      "grad_norm": 0.18811194598674774,
+      "learning_rate": 2.1683728659397517e-06,
+      "loss": 0.2978,
+      "step": 2573
+    },
+    {
+      "epoch": 2.167883211678832,
+      "grad_norm": 0.19322511553764343,
+      "learning_rate": 2.164333586018259e-06,
+      "loss": 0.3225,
+      "step": 2574
+    },
+    {
+      "epoch": 2.168725435148793,
+      "grad_norm": 0.19798986613750458,
+      "learning_rate": 2.160297032372857e-06,
+      "loss": 0.3435,
+      "step": 2575
+    },
+    {
+      "epoch": 2.1695676586187536,
+      "grad_norm": 0.19660674035549164,
+      "learning_rate": 2.156263208884386e-06,
+      "loss": 0.3775,
+      "step": 2576
+    },
+    {
+      "epoch": 2.1704098820887143,
+      "grad_norm": 0.18910837173461914,
+      "learning_rate": 2.1522321194310577e-06,
+      "loss": 0.3182,
+      "step": 2577
+    },
+    {
+      "epoch": 2.171252105558675,
+      "grad_norm": 0.18349318206310272,
+      "learning_rate": 2.148203767888455e-06,
+      "loss": 0.353,
+      "step": 2578
+    },
+    {
+      "epoch": 2.1720943290286354,
+      "grad_norm": 0.1835123598575592,
+      "learning_rate": 2.1441781581295286e-06,
+      "loss": 0.3201,
+      "step": 2579
+    },
+    {
+      "epoch": 2.172936552498596,
+      "grad_norm": 0.18625721335411072,
+      "learning_rate": 2.1401552940245962e-06,
+      "loss": 0.3305,
+      "step": 2580
+    },
+    {
+      "epoch": 2.173778775968557,
+      "grad_norm": 0.18136513233184814,
+      "learning_rate": 2.1361351794413334e-06,
+      "loss": 0.3301,
+      "step": 2581
+    },
+    {
+      "epoch": 2.1746209994385177,
+      "grad_norm": 0.18543168902397156,
+      "learning_rate": 2.132117818244771e-06,
+      "loss": 0.325,
+      "step": 2582
+    },
+    {
+      "epoch": 2.1754632229084785,
+      "grad_norm": 0.20783573389053345,
+      "learning_rate": 2.1281032142972933e-06,
+      "loss": 0.3577,
+      "step": 2583
+    },
+    {
+      "epoch": 2.1763054463784393,
+      "grad_norm": 0.18684588372707367,
+      "learning_rate": 2.124091371458638e-06,
+      "loss": 0.3422,
+      "step": 2584
+    },
+    {
+      "epoch": 2.1771476698483996,
+      "grad_norm": 0.21282611787319183,
+      "learning_rate": 2.1200822935858807e-06,
+      "loss": 0.3752,
+      "step": 2585
+    },
+    {
+      "epoch": 2.1779898933183603,
+      "grad_norm": 0.17147836089134216,
+      "learning_rate": 2.1160759845334483e-06,
+      "loss": 0.3291,
+      "step": 2586
+    },
+    {
+      "epoch": 2.178832116788321,
+      "grad_norm": 0.17594896256923676,
+      "learning_rate": 2.1120724481530937e-06,
+      "loss": 0.3099,
+      "step": 2587
+    },
+    {
+      "epoch": 2.179674340258282,
+      "grad_norm": 0.1904868483543396,
+      "learning_rate": 2.1080716882939145e-06,
+      "loss": 0.3487,
+      "step": 2588
+    },
+    {
+      "epoch": 2.1805165637282427,
+      "grad_norm": 0.19627432525157928,
+      "learning_rate": 2.1040737088023323e-06,
+      "loss": 0.3376,
+      "step": 2589
+    },
+    {
+      "epoch": 2.1813587871982034,
+      "grad_norm": 0.18196117877960205,
+      "learning_rate": 2.100078513522102e-06,
+      "loss": 0.2858,
+      "step": 2590
+    },
+    {
+      "epoch": 2.182201010668164,
+      "grad_norm": 0.1847255825996399,
+      "learning_rate": 2.0960861062942956e-06,
+      "loss": 0.3203,
+      "step": 2591
+    },
+    {
+      "epoch": 2.1830432341381245,
+      "grad_norm": 0.1873438060283661,
+      "learning_rate": 2.0920964909573065e-06,
+      "loss": 0.3488,
+      "step": 2592
+    },
+    {
+      "epoch": 2.1838854576080853,
+      "grad_norm": 0.18644773960113525,
+      "learning_rate": 2.0881096713468435e-06,
+      "loss": 0.3249,
+      "step": 2593
+    },
+    {
+      "epoch": 2.184727681078046,
+      "grad_norm": 0.1977187842130661,
+      "learning_rate": 2.0841256512959314e-06,
+      "loss": 0.3654,
+      "step": 2594
+    },
+    {
+      "epoch": 2.185569904548007,
+      "grad_norm": 0.18865470588207245,
+      "learning_rate": 2.080144434634898e-06,
+      "loss": 0.3108,
+      "step": 2595
+    },
+    {
+      "epoch": 2.1864121280179676,
+      "grad_norm": 0.19848677515983582,
+      "learning_rate": 2.0761660251913795e-06,
+      "loss": 0.3296,
+      "step": 2596
+    },
+    {
+      "epoch": 2.187254351487928,
+      "grad_norm": 0.3032154142856598,
+      "learning_rate": 2.0721904267903097e-06,
+      "loss": 0.3087,
+      "step": 2597
+    },
+    {
+      "epoch": 2.1880965749578887,
+      "grad_norm": 0.21433380246162415,
+      "learning_rate": 2.068217643253925e-06,
+      "loss": 0.3281,
+      "step": 2598
+    },
+    {
+      "epoch": 2.1889387984278494,
+      "grad_norm": 0.20142246782779694,
+      "learning_rate": 2.0642476784017507e-06,
+      "loss": 0.3495,
+      "step": 2599
+    },
+    {
+      "epoch": 2.18978102189781,
+      "grad_norm": 0.18407846987247467,
+      "learning_rate": 2.0602805360506044e-06,
+      "loss": 0.3514,
+      "step": 2600
+    },
+    {
+      "epoch": 2.190623245367771,
+      "grad_norm": 0.18453073501586914,
+      "learning_rate": 2.056316220014588e-06,
+      "loss": 0.3087,
+      "step": 2601
+    },
+    {
+      "epoch": 2.1914654688377317,
+      "grad_norm": 0.18886882066726685,
+      "learning_rate": 2.0523547341050913e-06,
+      "loss": 0.3196,
+      "step": 2602
+    },
+    {
+      "epoch": 2.1923076923076925,
+      "grad_norm": 0.18989959359169006,
+      "learning_rate": 2.0483960821307757e-06,
+      "loss": 0.3228,
+      "step": 2603
+    },
+    {
+      "epoch": 2.193149915777653,
+      "grad_norm": 0.20251639187335968,
+      "learning_rate": 2.0444402678975876e-06,
+      "loss": 0.3731,
+      "step": 2604
+    },
+    {
+      "epoch": 2.1939921392476136,
+      "grad_norm": 0.19764868915081024,
+      "learning_rate": 2.040487295208732e-06,
+      "loss": 0.344,
+      "step": 2605
+    },
+    {
+      "epoch": 2.1948343627175744,
+      "grad_norm": 0.19254057109355927,
+      "learning_rate": 2.036537167864695e-06,
+      "loss": 0.3603,
+      "step": 2606
+    },
+    {
+      "epoch": 2.195676586187535,
+      "grad_norm": 0.18026889860630035,
+      "learning_rate": 2.0325898896632178e-06,
+      "loss": 0.3185,
+      "step": 2607
+    },
+    {
+      "epoch": 2.196518809657496,
+      "grad_norm": 0.1734847128391266,
+      "learning_rate": 2.0286454643993097e-06,
+      "loss": 0.3448,
+      "step": 2608
+    },
+    {
+      "epoch": 2.1973610331274567,
+      "grad_norm": 0.18341229856014252,
+      "learning_rate": 2.024703895865232e-06,
+      "loss": 0.3319,
+      "step": 2609
+    },
+    {
+      "epoch": 2.198203256597417,
+      "grad_norm": 0.19103066623210907,
+      "learning_rate": 2.0207651878505e-06,
+      "loss": 0.3184,
+      "step": 2610
+    },
+    {
+      "epoch": 2.1990454800673778,
+      "grad_norm": 0.18200139701366425,
+      "learning_rate": 2.0168293441418798e-06,
+      "loss": 0.3245,
+      "step": 2611
+    },
+    {
+      "epoch": 2.1998877035373385,
+      "grad_norm": 0.21664761006832123,
+      "learning_rate": 2.012896368523386e-06,
+      "loss": 0.3056,
+      "step": 2612
+    },
+    {
+      "epoch": 2.2007299270072993,
+      "grad_norm": 0.20730926096439362,
+      "learning_rate": 2.0089662647762716e-06,
+      "loss": 0.3537,
+      "step": 2613
+    },
+    {
+      "epoch": 2.20157215047726,
+      "grad_norm": 0.18876250088214874,
+      "learning_rate": 2.0050390366790307e-06,
+      "loss": 0.3144,
+      "step": 2614
+    },
+    {
+      "epoch": 2.202414373947221,
+      "grad_norm": 0.191916361451149,
+      "learning_rate": 2.001114688007393e-06,
+      "loss": 0.3202,
+      "step": 2615
+    },
+    {
+      "epoch": 2.203256597417181,
+      "grad_norm": 0.18677039444446564,
+      "learning_rate": 1.997193222534316e-06,
+      "loss": 0.339,
+      "step": 2616
+    },
+    {
+      "epoch": 2.204098820887142,
+      "grad_norm": 0.18321190774440765,
+      "learning_rate": 1.9932746440299926e-06,
+      "loss": 0.3164,
+      "step": 2617
+    },
+    {
+      "epoch": 2.2049410443571027,
+      "grad_norm": 0.18801504373550415,
+      "learning_rate": 1.989358956261835e-06,
+      "loss": 0.3168,
+      "step": 2618
+    },
+    {
+      "epoch": 2.2057832678270635,
+      "grad_norm": 0.2072179764509201,
+      "learning_rate": 1.9854461629944764e-06,
+      "loss": 0.3332,
+      "step": 2619
+    },
+    {
+      "epoch": 2.2066254912970242,
+      "grad_norm": 0.1983804851770401,
+      "learning_rate": 1.981536267989766e-06,
+      "loss": 0.342,
+      "step": 2620
+    },
+    {
+      "epoch": 2.207467714766985,
+      "grad_norm": 0.20064833760261536,
+      "learning_rate": 1.977629275006772e-06,
+      "loss": 0.3352,
+      "step": 2621
+    },
+    {
+      "epoch": 2.2083099382369458,
+      "grad_norm": 0.19819088280200958,
+      "learning_rate": 1.9737251878017678e-06,
+      "loss": 0.312,
+      "step": 2622
+    },
+    {
+      "epoch": 2.209152161706906,
+      "grad_norm": 0.18463195860385895,
+      "learning_rate": 1.969824010128233e-06,
+      "loss": 0.3287,
+      "step": 2623
+    },
+    {
+      "epoch": 2.209994385176867,
+      "grad_norm": 0.18770988285541534,
+      "learning_rate": 1.9659257457368503e-06,
+      "loss": 0.3288,
+      "step": 2624
+    },
+    {
+      "epoch": 2.2108366086468276,
+      "grad_norm": 0.20165736973285675,
+      "learning_rate": 1.962030398375506e-06,
+      "loss": 0.352,
+      "step": 2625
+    },
+    {
+      "epoch": 2.2116788321167884,
+      "grad_norm": 0.21232005953788757,
+      "learning_rate": 1.9581379717892748e-06,
+      "loss": 0.3418,
+      "step": 2626
+    },
+    {
+      "epoch": 2.212521055586749,
+      "grad_norm": 0.17897962033748627,
+      "learning_rate": 1.954248469720431e-06,
+      "loss": 0.3076,
+      "step": 2627
+    },
+    {
+      "epoch": 2.2133632790567095,
+      "grad_norm": 0.19365070760250092,
+      "learning_rate": 1.950361895908427e-06,
+      "loss": 0.333,
+      "step": 2628
+    },
+    {
+      "epoch": 2.2142055025266703,
+      "grad_norm": 0.1850166767835617,
+      "learning_rate": 1.946478254089911e-06,
+      "loss": 0.3242,
+      "step": 2629
+    },
+    {
+      "epoch": 2.215047725996631,
+      "grad_norm": 0.18275009095668793,
+      "learning_rate": 1.942597547998703e-06,
+      "loss": 0.2998,
+      "step": 2630
+    },
+    {
+      "epoch": 2.215889949466592,
+      "grad_norm": 0.21025508642196655,
+      "learning_rate": 1.9387197813658092e-06,
+      "loss": 0.3675,
+      "step": 2631
+    },
+    {
+      "epoch": 2.2167321729365526,
+      "grad_norm": 0.20004096627235413,
+      "learning_rate": 1.934844957919403e-06,
+      "loss": 0.3285,
+      "step": 2632
+    },
+    {
+      "epoch": 2.2175743964065133,
+      "grad_norm": 0.18686963617801666,
+      "learning_rate": 1.9309730813848302e-06,
+      "loss": 0.3258,
+      "step": 2633
+    },
+    {
+      "epoch": 2.218416619876474,
+      "grad_norm": 0.17905724048614502,
+      "learning_rate": 1.927104155484602e-06,
+      "loss": 0.3276,
+      "step": 2634
+    },
+    {
+      "epoch": 2.2192588433464344,
+      "grad_norm": 0.18085958063602448,
+      "learning_rate": 1.923238183938398e-06,
+      "loss": 0.3042,
+      "step": 2635
+    },
+    {
+      "epoch": 2.220101066816395,
+      "grad_norm": 0.21374529600143433,
+      "learning_rate": 1.919375170463052e-06,
+      "loss": 0.3559,
+      "step": 2636
+    },
+    {
+      "epoch": 2.220943290286356,
+      "grad_norm": 0.18827268481254578,
+      "learning_rate": 1.915515118772555e-06,
+      "loss": 0.3277,
+      "step": 2637
+    },
+    {
+      "epoch": 2.2217855137563167,
+      "grad_norm": 0.18231862783432007,
+      "learning_rate": 1.9116580325780505e-06,
+      "loss": 0.3211,
+      "step": 2638
+    },
+    {
+      "epoch": 2.2226277372262775,
+      "grad_norm": 0.20695286989212036,
+      "learning_rate": 1.9078039155878338e-06,
+      "loss": 0.3148,
+      "step": 2639
+    },
+    {
+      "epoch": 2.2234699606962383,
+      "grad_norm": 0.17526499927043915,
+      "learning_rate": 1.9039527715073424e-06,
+      "loss": 0.2952,
+      "step": 2640
+    },
+    {
+      "epoch": 2.2243121841661986,
+      "grad_norm": 0.17523185908794403,
+      "learning_rate": 1.9001046040391558e-06,
+      "loss": 0.3309,
+      "step": 2641
+    },
+    {
+      "epoch": 2.2251544076361593,
+      "grad_norm": 0.1895107626914978,
+      "learning_rate": 1.8962594168829907e-06,
+      "loss": 0.3431,
+      "step": 2642
+    },
+    {
+      "epoch": 2.22599663110612,
+      "grad_norm": 0.19150540232658386,
+      "learning_rate": 1.8924172137357038e-06,
+      "loss": 0.3294,
+      "step": 2643
+    },
+    {
+      "epoch": 2.226838854576081,
+      "grad_norm": 0.277058482170105,
+      "learning_rate": 1.8885779982912756e-06,
+      "loss": 0.3089,
+      "step": 2644
+    },
+    {
+      "epoch": 2.2276810780460417,
+      "grad_norm": 0.18266412615776062,
+      "learning_rate": 1.884741774240823e-06,
+      "loss": 0.3234,
+      "step": 2645
+    },
+    {
+      "epoch": 2.2285233015160024,
+      "grad_norm": 0.17445224523544312,
+      "learning_rate": 1.8809085452725744e-06,
+      "loss": 0.3406,
+      "step": 2646
+    },
+    {
+      "epoch": 2.2293655249859627,
+      "grad_norm": 0.18090897798538208,
+      "learning_rate": 1.8770783150718913e-06,
+      "loss": 0.3168,
+      "step": 2647
+    },
+    {
+      "epoch": 2.2302077484559235,
+      "grad_norm": 0.1880083978176117,
+      "learning_rate": 1.8732510873212428e-06,
+      "loss": 0.3188,
+      "step": 2648
+    },
+    {
+      "epoch": 2.2310499719258843,
+      "grad_norm": 0.19110964238643646,
+      "learning_rate": 1.8694268657002197e-06,
+      "loss": 0.2974,
+      "step": 2649
+    },
+    {
+      "epoch": 2.231892195395845,
+      "grad_norm": 0.18959249556064606,
+      "learning_rate": 1.865605653885516e-06,
+      "loss": 0.3262,
+      "step": 2650
+    },
+    {
+      "epoch": 2.232734418865806,
+      "grad_norm": 0.22355952858924866,
+      "learning_rate": 1.8617874555509342e-06,
+      "loss": 0.3323,
+      "step": 2651
+    },
+    {
+      "epoch": 2.2335766423357666,
+      "grad_norm": 0.21287395060062408,
+      "learning_rate": 1.8579722743673773e-06,
+      "loss": 0.3184,
+      "step": 2652
+    },
+    {
+      "epoch": 2.2344188658057273,
+      "grad_norm": 0.17951267957687378,
+      "learning_rate": 1.8541601140028542e-06,
+      "loss": 0.3534,
+      "step": 2653
+    },
+    {
+      "epoch": 2.2352610892756877,
+      "grad_norm": 0.1638944149017334,
+      "learning_rate": 1.8503509781224627e-06,
+      "loss": 0.308,
+      "step": 2654
+    },
+    {
+      "epoch": 2.2361033127456484,
+      "grad_norm": 0.1907835453748703,
+      "learning_rate": 1.8465448703883959e-06,
+      "loss": 0.3423,
+      "step": 2655
+    },
+    {
+      "epoch": 2.236945536215609,
+      "grad_norm": 0.21052592992782593,
+      "learning_rate": 1.8427417944599325e-06,
+      "loss": 0.3469,
+      "step": 2656
+    },
+    {
+      "epoch": 2.23778775968557,
+      "grad_norm": 0.16829697787761688,
+      "learning_rate": 1.8389417539934428e-06,
+      "loss": 0.2978,
+      "step": 2657
+    },
+    {
+      "epoch": 2.2386299831555307,
+      "grad_norm": 0.19448906183242798,
+      "learning_rate": 1.8351447526423728e-06,
+      "loss": 0.3445,
+      "step": 2658
+    },
+    {
+      "epoch": 2.239472206625491,
+      "grad_norm": 0.1964830756187439,
+      "learning_rate": 1.8313507940572477e-06,
+      "loss": 0.3412,
+      "step": 2659
+    },
+    {
+      "epoch": 2.240314430095452,
+      "grad_norm": 0.18357017636299133,
+      "learning_rate": 1.8275598818856682e-06,
+      "loss": 0.3451,
+      "step": 2660
+    },
+    {
+      "epoch": 2.2411566535654126,
+      "grad_norm": 0.1971561759710312,
+      "learning_rate": 1.8237720197723075e-06,
+      "loss": 0.3274,
+      "step": 2661
+    },
+    {
+      "epoch": 2.2419988770353734,
+      "grad_norm": 0.19698187708854675,
+      "learning_rate": 1.819987211358903e-06,
+      "loss": 0.3322,
+      "step": 2662
+    },
+    {
+      "epoch": 2.242841100505334,
+      "grad_norm": 0.18066270649433136,
+      "learning_rate": 1.8162054602842621e-06,
+      "loss": 0.3321,
+      "step": 2663
+    },
+    {
+      "epoch": 2.243683323975295,
+      "grad_norm": 0.19735048711299896,
+      "learning_rate": 1.812426770184243e-06,
+      "loss": 0.2899,
+      "step": 2664
+    },
+    {
+      "epoch": 2.2445255474452557,
+      "grad_norm": 0.16894842684268951,
+      "learning_rate": 1.8086511446917715e-06,
+      "loss": 0.3217,
+      "step": 2665
+    },
+    {
+      "epoch": 2.245367770915216,
+      "grad_norm": 0.18202324211597443,
+      "learning_rate": 1.8048785874368191e-06,
+      "loss": 0.3128,
+      "step": 2666
+    },
+    {
+      "epoch": 2.2462099943851768,
+      "grad_norm": 0.20821116864681244,
+      "learning_rate": 1.8011091020464138e-06,
+      "loss": 0.3589,
+      "step": 2667
+    },
+    {
+      "epoch": 2.2470522178551375,
+      "grad_norm": 0.17096319794654846,
+      "learning_rate": 1.7973426921446258e-06,
+      "loss": 0.3224,
+      "step": 2668
+    },
+    {
+      "epoch": 2.2478944413250983,
+      "grad_norm": 0.1881159394979477,
+      "learning_rate": 1.7935793613525693e-06,
+      "loss": 0.336,
+      "step": 2669
+    },
+    {
+      "epoch": 2.248736664795059,
+      "grad_norm": 0.17406538128852844,
+      "learning_rate": 1.789819113288397e-06,
+      "loss": 0.3212,
+      "step": 2670
+    },
+    {
+      "epoch": 2.24957888826502,
+      "grad_norm": 0.1771821230649948,
+      "learning_rate": 1.7860619515673034e-06,
+      "loss": 0.3411,
+      "step": 2671
+    },
+    {
+      "epoch": 2.25042111173498,
+      "grad_norm": 0.1870764195919037,
+      "learning_rate": 1.7823078798015098e-06,
+      "loss": 0.3485,
+      "step": 2672
+    },
+    {
+      "epoch": 2.251263335204941,
+      "grad_norm": 0.19147291779518127,
+      "learning_rate": 1.7785569016002686e-06,
+      "loss": 0.2787,
+      "step": 2673
+    },
+    {
+      "epoch": 2.2521055586749017,
+      "grad_norm": 0.19009776413440704,
+      "learning_rate": 1.7748090205698565e-06,
+      "loss": 0.3568,
+      "step": 2674
+    },
+    {
+      "epoch": 2.2529477821448625,
+      "grad_norm": 0.18951989710330963,
+      "learning_rate": 1.7710642403135768e-06,
+      "loss": 0.3371,
+      "step": 2675
+    },
+    {
+      "epoch": 2.2537900056148232,
+      "grad_norm": 0.18312712013721466,
+      "learning_rate": 1.7673225644317487e-06,
+      "loss": 0.3404,
+      "step": 2676
+    },
+    {
+      "epoch": 2.254632229084784,
+      "grad_norm": 0.20596158504486084,
+      "learning_rate": 1.7635839965217055e-06,
+      "loss": 0.3215,
+      "step": 2677
+    },
+    {
+      "epoch": 2.2554744525547443,
+      "grad_norm": 0.20948167145252228,
+      "learning_rate": 1.7598485401777932e-06,
+      "loss": 0.3267,
+      "step": 2678
+    },
+    {
+      "epoch": 2.256316676024705,
+      "grad_norm": 0.20294427871704102,
+      "learning_rate": 1.75611619899137e-06,
+      "loss": 0.3713,
+      "step": 2679
+    },
+    {
+      "epoch": 2.257158899494666,
+      "grad_norm": 0.1707998812198639,
+      "learning_rate": 1.7523869765507928e-06,
+      "loss": 0.3182,
+      "step": 2680
+    },
+    {
+      "epoch": 2.2580011229646266,
+      "grad_norm": 0.17670342326164246,
+      "learning_rate": 1.748660876441428e-06,
+      "loss": 0.3587,
+      "step": 2681
+    },
+    {
+      "epoch": 2.2588433464345874,
+      "grad_norm": 0.18761256337165833,
+      "learning_rate": 1.7449379022456297e-06,
+      "loss": 0.3356,
+      "step": 2682
+    },
+    {
+      "epoch": 2.259685569904548,
+      "grad_norm": 0.19832022488117218,
+      "learning_rate": 1.7412180575427572e-06,
+      "loss": 0.3478,
+      "step": 2683
+    },
+    {
+      "epoch": 2.260527793374509,
+      "grad_norm": 0.19921515882015228,
+      "learning_rate": 1.7375013459091529e-06,
+      "loss": 0.3493,
+      "step": 2684
+    },
+    {
+      "epoch": 2.2613700168444693,
+      "grad_norm": 0.18138949573040009,
+      "learning_rate": 1.7337877709181527e-06,
+      "loss": 0.3108,
+      "step": 2685
+    },
+    {
+      "epoch": 2.26221224031443,
+      "grad_norm": 0.19637282192707062,
+      "learning_rate": 1.7300773361400746e-06,
+      "loss": 0.3033,
+      "step": 2686
+    },
+    {
+      "epoch": 2.263054463784391,
+      "grad_norm": 0.20320606231689453,
+      "learning_rate": 1.7263700451422166e-06,
+      "loss": 0.3553,
+      "step": 2687
+    },
+    {
+      "epoch": 2.2638966872543516,
+      "grad_norm": 0.3034162223339081,
+      "learning_rate": 1.7226659014888548e-06,
+      "loss": 0.3312,
+      "step": 2688
+    },
+    {
+      "epoch": 2.2647389107243123,
+      "grad_norm": 0.18244308233261108,
+      "learning_rate": 1.7189649087412385e-06,
+      "loss": 0.3552,
+      "step": 2689
+    },
+    {
+      "epoch": 2.2655811341942727,
+      "grad_norm": 0.1824118047952652,
+      "learning_rate": 1.7152670704575919e-06,
+      "loss": 0.3138,
+      "step": 2690
+    },
+    {
+      "epoch": 2.2664233576642334,
+      "grad_norm": 0.1936587542295456,
+      "learning_rate": 1.711572390193102e-06,
+      "loss": 0.3388,
+      "step": 2691
+    },
+    {
+      "epoch": 2.267265581134194,
+      "grad_norm": 0.2071102261543274,
+      "learning_rate": 1.7078808714999207e-06,
+      "loss": 0.3675,
+      "step": 2692
+    },
+    {
+      "epoch": 2.268107804604155,
+      "grad_norm": 0.19295620918273926,
+      "learning_rate": 1.7041925179271584e-06,
+      "loss": 0.3218,
+      "step": 2693
+    },
+    {
+      "epoch": 2.2689500280741157,
+      "grad_norm": 0.18506266176700592,
+      "learning_rate": 1.7005073330208881e-06,
+      "loss": 0.3429,
+      "step": 2694
+    },
+    {
+      "epoch": 2.2697922515440765,
+      "grad_norm": 0.17563672363758087,
+      "learning_rate": 1.696825320324132e-06,
+      "loss": 0.339,
+      "step": 2695
+    },
+    {
+      "epoch": 2.2706344750140373,
+      "grad_norm": 0.17654356360435486,
+      "learning_rate": 1.6931464833768624e-06,
+      "loss": 0.2947,
+      "step": 2696
+    },
+    {
+      "epoch": 2.2714766984839976,
+      "grad_norm": 0.19551007449626923,
+      "learning_rate": 1.689470825715998e-06,
+      "loss": 0.3435,
+      "step": 2697
+    },
+    {
+      "epoch": 2.2723189219539583,
+      "grad_norm": 0.1912021040916443,
+      "learning_rate": 1.6857983508754056e-06,
+      "loss": 0.3226,
+      "step": 2698
+    },
+    {
+      "epoch": 2.273161145423919,
+      "grad_norm": 0.17643921077251434,
+      "learning_rate": 1.6821290623858865e-06,
+      "loss": 0.3279,
+      "step": 2699
+    },
+    {
+      "epoch": 2.27400336889388,
+      "grad_norm": 0.20375093817710876,
+      "learning_rate": 1.6784629637751814e-06,
+      "loss": 0.3578,
+      "step": 2700
+    },
+    {
+      "epoch": 2.2748455923638407,
+      "grad_norm": 0.19104832410812378,
+      "learning_rate": 1.6748000585679602e-06,
+      "loss": 0.328,
+      "step": 2701
+    },
+    {
+      "epoch": 2.2756878158338014,
+      "grad_norm": 0.19647930562496185,
+      "learning_rate": 1.6711403502858302e-06,
+      "loss": 0.3139,
+      "step": 2702
+    },
+    {
+      "epoch": 2.2765300393037617,
+      "grad_norm": 0.1835135668516159,
+      "learning_rate": 1.6674838424473172e-06,
+      "loss": 0.3185,
+      "step": 2703
+    },
+    {
+      "epoch": 2.2773722627737225,
+      "grad_norm": 0.1910325586795807,
+      "learning_rate": 1.6638305385678783e-06,
+      "loss": 0.3367,
+      "step": 2704
+    },
+    {
+      "epoch": 2.2782144862436833,
+      "grad_norm": 0.17093531787395477,
+      "learning_rate": 1.6601804421598787e-06,
+      "loss": 0.2972,
+      "step": 2705
+    },
+    {
+      "epoch": 2.279056709713644,
+      "grad_norm": 0.19414487481117249,
+      "learning_rate": 1.6565335567326112e-06,
+      "loss": 0.372,
+      "step": 2706
+    },
+    {
+      "epoch": 2.279898933183605,
+      "grad_norm": 0.1806740164756775,
+      "learning_rate": 1.6528898857922747e-06,
+      "loss": 0.2987,
+      "step": 2707
+    },
+    {
+      "epoch": 2.2807411566535656,
+      "grad_norm": 0.19277624785900116,
+      "learning_rate": 1.6492494328419816e-06,
+      "loss": 0.355,
+      "step": 2708
+    },
+    {
+      "epoch": 2.281583380123526,
+      "grad_norm": 0.1879538744688034,
+      "learning_rate": 1.6456122013817477e-06,
+      "loss": 0.3711,
+      "step": 2709
+    },
+    {
+      "epoch": 2.2824256035934867,
+      "grad_norm": 0.26613113284111023,
+      "learning_rate": 1.6419781949084928e-06,
+      "loss": 0.3437,
+      "step": 2710
+    },
+    {
+      "epoch": 2.2832678270634474,
+      "grad_norm": 0.18551747500896454,
+      "learning_rate": 1.6383474169160334e-06,
+      "loss": 0.3551,
+      "step": 2711
+    },
+    {
+      "epoch": 2.284110050533408,
+      "grad_norm": 0.18406280875205994,
+      "learning_rate": 1.6347198708950884e-06,
+      "loss": 0.327,
+      "step": 2712
+    },
+    {
+      "epoch": 2.284952274003369,
+      "grad_norm": 0.1676313877105713,
+      "learning_rate": 1.631095560333264e-06,
+      "loss": 0.3267,
+      "step": 2713
+    },
+    {
+      "epoch": 2.2857944974733297,
+      "grad_norm": 0.18933825194835663,
+      "learning_rate": 1.6274744887150562e-06,
+      "loss": 0.3433,
+      "step": 2714
+    },
+    {
+      "epoch": 2.2866367209432905,
+      "grad_norm": 0.19097904860973358,
+      "learning_rate": 1.6238566595218475e-06,
+      "loss": 0.3314,
+      "step": 2715
+    },
+    {
+      "epoch": 2.287478944413251,
+      "grad_norm": 0.1931888610124588,
+      "learning_rate": 1.6202420762319065e-06,
+      "loss": 0.3265,
+      "step": 2716
+    },
+    {
+      "epoch": 2.2883211678832116,
+      "grad_norm": 0.18155337870121002,
+      "learning_rate": 1.6166307423203765e-06,
+      "loss": 0.326,
+      "step": 2717
+    },
+    {
+      "epoch": 2.2891633913531724,
+      "grad_norm": 0.196844682097435,
+      "learning_rate": 1.6130226612592787e-06,
+      "loss": 0.3373,
+      "step": 2718
+    },
+    {
+      "epoch": 2.290005614823133,
+      "grad_norm": 0.18781322240829468,
+      "learning_rate": 1.6094178365175044e-06,
+      "loss": 0.3457,
+      "step": 2719
+    },
+    {
+      "epoch": 2.290847838293094,
+      "grad_norm": 0.17167329788208008,
+      "learning_rate": 1.6058162715608205e-06,
+      "loss": 0.3206,
+      "step": 2720
+    },
+    {
+      "epoch": 2.2916900617630542,
+      "grad_norm": 0.16969625651836395,
+      "learning_rate": 1.6022179698518525e-06,
+      "loss": 0.3458,
+      "step": 2721
+    },
+    {
+      "epoch": 2.292532285233015,
+      "grad_norm": 0.1915416568517685,
+      "learning_rate": 1.598622934850097e-06,
+      "loss": 0.3676,
+      "step": 2722
+    },
+    {
+      "epoch": 2.2933745087029758,
+      "grad_norm": 0.1874096542596817,
+      "learning_rate": 1.595031170011898e-06,
+      "loss": 0.3355,
+      "step": 2723
+    },
+    {
+      "epoch": 2.2942167321729365,
+      "grad_norm": 0.1758255809545517,
+      "learning_rate": 1.591442678790467e-06,
+      "loss": 0.3155,
+      "step": 2724
+    },
+    {
+      "epoch": 2.2950589556428973,
+      "grad_norm": 0.18766756355762482,
+      "learning_rate": 1.5878574646358608e-06,
+      "loss": 0.3806,
+      "step": 2725
+    },
+    {
+      "epoch": 2.295901179112858,
+      "grad_norm": 0.46558520197868347,
+      "learning_rate": 1.584275530994991e-06,
+      "loss": 0.317,
+      "step": 2726
+    },
+    {
+      "epoch": 2.296743402582819,
+      "grad_norm": 0.17401887476444244,
+      "learning_rate": 1.580696881311611e-06,
+      "loss": 0.3636,
+      "step": 2727
+    },
+    {
+      "epoch": 2.297585626052779,
+      "grad_norm": 0.1759478896856308,
+      "learning_rate": 1.5771215190263183e-06,
+      "loss": 0.301,
+      "step": 2728
+    },
+    {
+      "epoch": 2.29842784952274,
+      "grad_norm": 0.1894080638885498,
+      "learning_rate": 1.573549447576549e-06,
+      "loss": 0.3249,
+      "step": 2729
+    },
+    {
+      "epoch": 2.2992700729927007,
+      "grad_norm": 0.1710214465856552,
+      "learning_rate": 1.5699806703965787e-06,
+      "loss": 0.3388,
+      "step": 2730
+    },
+    {
+      "epoch": 2.3001122964626615,
+      "grad_norm": 0.18094199895858765,
+      "learning_rate": 1.5664151909175124e-06,
+      "loss": 0.3442,
+      "step": 2731
+    },
+    {
+      "epoch": 2.3009545199326222,
+      "grad_norm": 0.19348318874835968,
+      "learning_rate": 1.5628530125672848e-06,
+      "loss": 0.3502,
+      "step": 2732
+    },
+    {
+      "epoch": 2.301796743402583,
+      "grad_norm": 0.1840953230857849,
+      "learning_rate": 1.5592941387706562e-06,
+      "loss": 0.3152,
+      "step": 2733
+    },
+    {
+      "epoch": 2.3026389668725433,
+      "grad_norm": 0.17521828413009644,
+      "learning_rate": 1.555738572949214e-06,
+      "loss": 0.3359,
+      "step": 2734
+    },
+    {
+      "epoch": 2.303481190342504,
+      "grad_norm": 0.2008764147758484,
+      "learning_rate": 1.5521863185213626e-06,
+      "loss": 0.4031,
+      "step": 2735
+    },
+    {
+      "epoch": 2.304323413812465,
+      "grad_norm": 0.16677844524383545,
+      "learning_rate": 1.5486373789023206e-06,
+      "loss": 0.2943,
+      "step": 2736
+    },
+    {
+      "epoch": 2.3051656372824256,
+      "grad_norm": 0.19051140546798706,
+      "learning_rate": 1.5450917575041209e-06,
+      "loss": 0.3354,
+      "step": 2737
+    },
+    {
+      "epoch": 2.3060078607523864,
+      "grad_norm": 0.16995146870613098,
+      "learning_rate": 1.54154945773561e-06,
+      "loss": 0.3129,
+      "step": 2738
+    },
+    {
+      "epoch": 2.306850084222347,
+      "grad_norm": 0.17699390649795532,
+      "learning_rate": 1.538010483002435e-06,
+      "loss": 0.3423,
+      "step": 2739
+    },
+    {
+      "epoch": 2.3076923076923075,
+      "grad_norm": 0.16745799779891968,
+      "learning_rate": 1.5344748367070534e-06,
+      "loss": 0.3244,
+      "step": 2740
+    },
+    {
+      "epoch": 2.3085345311622683,
+      "grad_norm": 0.193995401263237,
+      "learning_rate": 1.5309425222487119e-06,
+      "loss": 0.3781,
+      "step": 2741
+    },
+    {
+      "epoch": 2.309376754632229,
+      "grad_norm": 0.19723963737487793,
+      "learning_rate": 1.5274135430234654e-06,
+      "loss": 0.3255,
+      "step": 2742
+    },
+    {
+      "epoch": 2.31021897810219,
+      "grad_norm": 0.19136905670166016,
+      "learning_rate": 1.5238879024241544e-06,
+      "loss": 0.3166,
+      "step": 2743
+    },
+    {
+      "epoch": 2.3110612015721506,
+      "grad_norm": 0.1685858815908432,
+      "learning_rate": 1.5203656038404146e-06,
+      "loss": 0.3125,
+      "step": 2744
+    },
+    {
+      "epoch": 2.3119034250421113,
+      "grad_norm": 0.17219972610473633,
+      "learning_rate": 1.5168466506586654e-06,
+      "loss": 0.3216,
+      "step": 2745
+    },
+    {
+      "epoch": 2.312745648512072,
+      "grad_norm": 0.19580884277820587,
+      "learning_rate": 1.5133310462621103e-06,
+      "loss": 0.3772,
+      "step": 2746
+    },
+    {
+      "epoch": 2.3135878719820324,
+      "grad_norm": 0.1875855177640915,
+      "learning_rate": 1.509818794030733e-06,
+      "loss": 0.3521,
+      "step": 2747
+    },
+    {
+      "epoch": 2.314430095451993,
+      "grad_norm": 0.17141541838645935,
+      "learning_rate": 1.506309897341297e-06,
+      "loss": 0.2875,
+      "step": 2748
+    },
+    {
+      "epoch": 2.315272318921954,
+      "grad_norm": 0.19664089381694794,
+      "learning_rate": 1.502804359567337e-06,
+      "loss": 0.3451,
+      "step": 2749
+    },
+    {
+      "epoch": 2.3161145423919147,
+      "grad_norm": 0.18293555080890656,
+      "learning_rate": 1.499302184079159e-06,
+      "loss": 0.3167,
+      "step": 2750
+    },
+    {
+      "epoch": 2.3169567658618755,
+      "grad_norm": 0.18555696308612823,
+      "learning_rate": 1.4958033742438348e-06,
+      "loss": 0.3243,
+      "step": 2751
+    },
+    {
+      "epoch": 2.317798989331836,
+      "grad_norm": 0.16527746617794037,
+      "learning_rate": 1.492307933425205e-06,
+      "loss": 0.299,
+      "step": 2752
+    },
+    {
+      "epoch": 2.3186412128017966,
+      "grad_norm": 0.18836185336112976,
+      "learning_rate": 1.4888158649838675e-06,
+      "loss": 0.3353,
+      "step": 2753
+    },
+    {
+      "epoch": 2.3194834362717573,
+      "grad_norm": 0.18642380833625793,
+      "learning_rate": 1.4853271722771772e-06,
+      "loss": 0.3335,
+      "step": 2754
+    },
+    {
+      "epoch": 2.320325659741718,
+      "grad_norm": 0.1658937633037567,
+      "learning_rate": 1.4818418586592448e-06,
+      "loss": 0.3213,
+      "step": 2755
+    },
+    {
+      "epoch": 2.321167883211679,
+      "grad_norm": 0.1841190606355667,
+      "learning_rate": 1.478359927480935e-06,
+      "loss": 0.3417,
+      "step": 2756
+    },
+    {
+      "epoch": 2.3220101066816397,
+      "grad_norm": 0.17181068658828735,
+      "learning_rate": 1.4748813820898554e-06,
+      "loss": 0.3131,
+      "step": 2757
+    },
+    {
+      "epoch": 2.3228523301516004,
+      "grad_norm": 0.17926974594593048,
+      "learning_rate": 1.4714062258303653e-06,
+      "loss": 0.3535,
+      "step": 2758
+    },
+    {
+      "epoch": 2.3236945536215607,
+      "grad_norm": 0.18430732190608978,
+      "learning_rate": 1.4679344620435543e-06,
+      "loss": 0.3513,
+      "step": 2759
+    },
+    {
+      "epoch": 2.3245367770915215,
+      "grad_norm": 0.18636822700500488,
+      "learning_rate": 1.4644660940672628e-06,
+      "loss": 0.3289,
+      "step": 2760
+    },
+    {
+      "epoch": 2.3253790005614823,
+      "grad_norm": 0.22061777114868164,
+      "learning_rate": 1.4610011252360594e-06,
+      "loss": 0.3499,
+      "step": 2761
+    },
+    {
+      "epoch": 2.326221224031443,
+      "grad_norm": 0.16940012574195862,
+      "learning_rate": 1.4575395588812452e-06,
+      "loss": 0.3058,
+      "step": 2762
+    },
+    {
+      "epoch": 2.327063447501404,
+      "grad_norm": 0.17286397516727448,
+      "learning_rate": 1.454081398330855e-06,
+      "loss": 0.3498,
+      "step": 2763
+    },
+    {
+      "epoch": 2.3279056709713646,
+      "grad_norm": 0.17675434052944183,
+      "learning_rate": 1.450626646909639e-06,
+      "loss": 0.3084,
+      "step": 2764
+    },
+    {
+      "epoch": 2.328747894441325,
+      "grad_norm": 0.18519286811351776,
+      "learning_rate": 1.4471753079390815e-06,
+      "loss": 0.3194,
+      "step": 2765
+    },
+    {
+      "epoch": 2.3295901179112857,
+      "grad_norm": 0.18685199320316315,
+      "learning_rate": 1.4437273847373778e-06,
+      "loss": 0.3492,
+      "step": 2766
+    },
+    {
+      "epoch": 2.3304323413812464,
+      "grad_norm": 0.1702100932598114,
+      "learning_rate": 1.4402828806194436e-06,
+      "loss": 0.3163,
+      "step": 2767
+    },
+    {
+      "epoch": 2.331274564851207,
+      "grad_norm": 0.15651482343673706,
+      "learning_rate": 1.4368417988969058e-06,
+      "loss": 0.2848,
+      "step": 2768
+    },
+    {
+      "epoch": 2.332116788321168,
+      "grad_norm": 0.20350609719753265,
+      "learning_rate": 1.4334041428781003e-06,
+      "loss": 0.3811,
+      "step": 2769
+    },
+    {
+      "epoch": 2.3329590117911287,
+      "grad_norm": 0.18207013607025146,
+      "learning_rate": 1.429969915868068e-06,
+      "loss": 0.3147,
+      "step": 2770
+    },
+    {
+      "epoch": 2.333801235261089,
+      "grad_norm": 0.17690841853618622,
+      "learning_rate": 1.4265391211685597e-06,
+      "loss": 0.3298,
+      "step": 2771
+    },
+    {
+      "epoch": 2.33464345873105,
+      "grad_norm": 0.18768583238124847,
+      "learning_rate": 1.4231117620780188e-06,
+      "loss": 0.318,
+      "step": 2772
+    },
+    {
+      "epoch": 2.3354856822010106,
+      "grad_norm": 0.18087123334407806,
+      "learning_rate": 1.4196878418915894e-06,
+      "loss": 0.3227,
+      "step": 2773
+    },
+    {
+      "epoch": 2.3363279056709714,
+      "grad_norm": 0.19178766012191772,
+      "learning_rate": 1.4162673639011065e-06,
+      "loss": 0.3507,
+      "step": 2774
+    },
+    {
+      "epoch": 2.337170129140932,
+      "grad_norm": 0.2115650326013565,
+      "learning_rate": 1.4128503313951008e-06,
+      "loss": 0.3659,
+      "step": 2775
+    },
+    {
+      "epoch": 2.338012352610893,
+      "grad_norm": 0.1869487762451172,
+      "learning_rate": 1.4094367476587867e-06,
+      "loss": 0.3332,
+      "step": 2776
+    },
+    {
+      "epoch": 2.3388545760808537,
+      "grad_norm": 0.19197297096252441,
+      "learning_rate": 1.4060266159740627e-06,
+      "loss": 0.3117,
+      "step": 2777
+    },
+    {
+      "epoch": 2.339696799550814,
+      "grad_norm": 0.16871175169944763,
+      "learning_rate": 1.4026199396195078e-06,
+      "loss": 0.3073,
+      "step": 2778
+    },
+    {
+      "epoch": 2.3405390230207748,
+      "grad_norm": 0.1799178570508957,
+      "learning_rate": 1.399216721870384e-06,
+      "loss": 0.364,
+      "step": 2779
+    },
+    {
+      "epoch": 2.3413812464907355,
+      "grad_norm": 0.17059600353240967,
+      "learning_rate": 1.3958169659986204e-06,
+      "loss": 0.3244,
+      "step": 2780
+    },
+    {
+      "epoch": 2.3422234699606963,
+      "grad_norm": 0.184001624584198,
+      "learning_rate": 1.3924206752728282e-06,
+      "loss": 0.3267,
+      "step": 2781
+    },
+    {
+      "epoch": 2.343065693430657,
+      "grad_norm": 0.1990918219089508,
+      "learning_rate": 1.389027852958273e-06,
+      "loss": 0.3968,
+      "step": 2782
+    },
+    {
+      "epoch": 2.3439079169006174,
+      "grad_norm": 0.17149695754051208,
+      "learning_rate": 1.385638502316899e-06,
+      "loss": 0.3008,
+      "step": 2783
+    },
+    {
+      "epoch": 2.344750140370578,
+      "grad_norm": 0.1957639753818512,
+      "learning_rate": 1.3822526266073044e-06,
+      "loss": 0.3502,
+      "step": 2784
+    },
+    {
+      "epoch": 2.345592363840539,
+      "grad_norm": 0.1924923062324524,
+      "learning_rate": 1.3788702290847517e-06,
+      "loss": 0.2741,
+      "step": 2785
+    },
+    {
+      "epoch": 2.3464345873104997,
+      "grad_norm": 0.20024049282073975,
+      "learning_rate": 1.3754913130011566e-06,
+      "loss": 0.3951,
+      "step": 2786
+    },
+    {
+      "epoch": 2.3472768107804605,
+      "grad_norm": 0.1801765114068985,
+      "learning_rate": 1.3721158816050872e-06,
+      "loss": 0.2943,
+      "step": 2787
+    },
+    {
+      "epoch": 2.3481190342504212,
+      "grad_norm": 0.20196901261806488,
+      "learning_rate": 1.3687439381417616e-06,
+      "loss": 0.3332,
+      "step": 2788
+    },
+    {
+      "epoch": 2.348961257720382,
+      "grad_norm": 0.18237796425819397,
+      "learning_rate": 1.3653754858530477e-06,
+      "loss": 0.3503,
+      "step": 2789
+    },
+    {
+      "epoch": 2.3498034811903423,
+      "grad_norm": 0.18183031678199768,
+      "learning_rate": 1.3620105279774532e-06,
+      "loss": 0.327,
+      "step": 2790
+    },
+    {
+      "epoch": 2.350645704660303,
+      "grad_norm": 0.1969723403453827,
+      "learning_rate": 1.3586490677501269e-06,
+      "loss": 0.3508,
+      "step": 2791
+    },
+    {
+      "epoch": 2.351487928130264,
+      "grad_norm": 0.18273185193538666,
+      "learning_rate": 1.3552911084028536e-06,
+      "loss": 0.3099,
+      "step": 2792
+    },
+    {
+      "epoch": 2.3523301516002246,
+      "grad_norm": 0.18478451669216156,
+      "learning_rate": 1.3519366531640589e-06,
+      "loss": 0.3326,
+      "step": 2793
+    },
+    {
+      "epoch": 2.3531723750701854,
+      "grad_norm": 0.19646549224853516,
+      "learning_rate": 1.3485857052587908e-06,
+      "loss": 0.3583,
+      "step": 2794
+    },
+    {
+      "epoch": 2.354014598540146,
+      "grad_norm": 0.17265410721302032,
+      "learning_rate": 1.3452382679087307e-06,
+      "loss": 0.3372,
+      "step": 2795
+    },
+    {
+      "epoch": 2.3548568220101065,
+      "grad_norm": 0.19773848354816437,
+      "learning_rate": 1.3418943443321807e-06,
+      "loss": 0.3436,
+      "step": 2796
+    },
+    {
+      "epoch": 2.3556990454800673,
+      "grad_norm": 0.17929218709468842,
+      "learning_rate": 1.3385539377440709e-06,
+      "loss": 0.2815,
+      "step": 2797
+    },
+    {
+      "epoch": 2.356541268950028,
+      "grad_norm": 0.19509731233119965,
+      "learning_rate": 1.3352170513559432e-06,
+      "loss": 0.3589,
+      "step": 2798
+    },
+    {
+      "epoch": 2.357383492419989,
+      "grad_norm": 0.189386248588562,
+      "learning_rate": 1.3318836883759634e-06,
+      "loss": 0.3374,
+      "step": 2799
+    },
+    {
+      "epoch": 2.3582257158899496,
+      "grad_norm": 0.18874704837799072,
+      "learning_rate": 1.3285538520088976e-06,
+      "loss": 0.3145,
+      "step": 2800
+    },
+    {
+      "epoch": 2.3590679393599103,
+      "grad_norm": 0.19876275956630707,
+      "learning_rate": 1.3252275454561337e-06,
+      "loss": 0.3612,
+      "step": 2801
+    },
+    {
+      "epoch": 2.3599101628298707,
+      "grad_norm": 0.17174991965293884,
+      "learning_rate": 1.3219047719156575e-06,
+      "loss": 0.3259,
+      "step": 2802
+    },
+    {
+      "epoch": 2.3607523862998314,
+      "grad_norm": 0.16203835606575012,
+      "learning_rate": 1.318585534582064e-06,
+      "loss": 0.309,
+      "step": 2803
+    },
+    {
+      "epoch": 2.361594609769792,
+      "grad_norm": 0.1940368413925171,
+      "learning_rate": 1.3152698366465449e-06,
+      "loss": 0.3591,
+      "step": 2804
+    },
+    {
+      "epoch": 2.362436833239753,
+      "grad_norm": 0.1763688176870346,
+      "learning_rate": 1.3119576812968893e-06,
+      "loss": 0.3434,
+      "step": 2805
+    },
+    {
+      "epoch": 2.3632790567097137,
+      "grad_norm": 0.1753963679075241,
+      "learning_rate": 1.30864907171748e-06,
+      "loss": 0.3003,
+      "step": 2806
+    },
+    {
+      "epoch": 2.3641212801796745,
+      "grad_norm": 0.1836460530757904,
+      "learning_rate": 1.305344011089294e-06,
+      "loss": 0.3296,
+      "step": 2807
+    },
+    {
+      "epoch": 2.3649635036496353,
+      "grad_norm": 0.18466545641422272,
+      "learning_rate": 1.3020425025898926e-06,
+      "loss": 0.3264,
+      "step": 2808
+    },
+    {
+      "epoch": 2.3658057271195956,
+      "grad_norm": 0.18666084110736847,
+      "learning_rate": 1.2987445493934236e-06,
+      "loss": 0.3428,
+      "step": 2809
+    },
+    {
+      "epoch": 2.3666479505895563,
+      "grad_norm": 0.17960001528263092,
+      "learning_rate": 1.295450154670615e-06,
+      "loss": 0.3471,
+      "step": 2810
+    },
+    {
+      "epoch": 2.367490174059517,
+      "grad_norm": 0.18435019254684448,
+      "learning_rate": 1.292159321588778e-06,
+      "loss": 0.3479,
+      "step": 2811
+    },
+    {
+      "epoch": 2.368332397529478,
+      "grad_norm": 0.16246698796749115,
+      "learning_rate": 1.288872053311795e-06,
+      "loss": 0.3057,
+      "step": 2812
+    },
+    {
+      "epoch": 2.3691746209994387,
+      "grad_norm": 0.1754060536623001,
+      "learning_rate": 1.2855883530001228e-06,
+      "loss": 0.3296,
+      "step": 2813
+    },
+    {
+      "epoch": 2.370016844469399,
+      "grad_norm": 0.3555665612220764,
+      "learning_rate": 1.282308223810786e-06,
+      "loss": 0.3046,
+      "step": 2814
+    },
+    {
+      "epoch": 2.3708590679393597,
+      "grad_norm": 0.17933470010757446,
+      "learning_rate": 1.2790316688973809e-06,
+      "loss": 0.3344,
+      "step": 2815
+    },
+    {
+      "epoch": 2.3717012914093205,
+      "grad_norm": 0.18554142117500305,
+      "learning_rate": 1.2757586914100612e-06,
+      "loss": 0.3695,
+      "step": 2816
+    },
+    {
+      "epoch": 2.3725435148792813,
+      "grad_norm": 0.17822647094726562,
+      "learning_rate": 1.272489294495548e-06,
+      "loss": 0.3418,
+      "step": 2817
+    },
+    {
+      "epoch": 2.373385738349242,
+      "grad_norm": 0.18284878134727478,
+      "learning_rate": 1.2692234812971106e-06,
+      "loss": 0.3372,
+      "step": 2818
+    },
+    {
+      "epoch": 2.374227961819203,
+      "grad_norm": 0.18310308456420898,
+      "learning_rate": 1.265961254954583e-06,
+      "loss": 0.3264,
+      "step": 2819
+    },
+    {
+      "epoch": 2.3750701852891636,
+      "grad_norm": 0.17785710096359253,
+      "learning_rate": 1.2627026186043423e-06,
+      "loss": 0.3195,
+      "step": 2820
+    },
+    {
+      "epoch": 2.375912408759124,
+      "grad_norm": 0.17308805882930756,
+      "learning_rate": 1.2594475753793211e-06,
+      "loss": 0.3216,
+      "step": 2821
+    },
+    {
+      "epoch": 2.3767546322290847,
+      "grad_norm": 0.19008275866508484,
+      "learning_rate": 1.256196128408993e-06,
+      "loss": 0.3308,
+      "step": 2822
+    },
+    {
+      "epoch": 2.3775968556990454,
+      "grad_norm": 0.18448509275913239,
+      "learning_rate": 1.252948280819375e-06,
+      "loss": 0.3341,
+      "step": 2823
+    },
+    {
+      "epoch": 2.378439079169006,
+      "grad_norm": 0.201051726937294,
+      "learning_rate": 1.249704035733022e-06,
+      "loss": 0.3335,
+      "step": 2824
+    },
+    {
+      "epoch": 2.379281302638967,
+      "grad_norm": 0.19201770424842834,
+      "learning_rate": 1.2464633962690304e-06,
+      "loss": 0.3422,
+      "step": 2825
+    },
+    {
+      "epoch": 2.3801235261089277,
+      "grad_norm": 0.1856694221496582,
+      "learning_rate": 1.243226365543026e-06,
+      "loss": 0.3521,
+      "step": 2826
+    },
+    {
+      "epoch": 2.3809657495788885,
+      "grad_norm": 0.17718330025672913,
+      "learning_rate": 1.239992946667165e-06,
+      "loss": 0.3353,
+      "step": 2827
+    },
+    {
+      "epoch": 2.381807973048849,
+      "grad_norm": 0.18037939071655273,
+      "learning_rate": 1.2367631427501308e-06,
+      "loss": 0.3386,
+      "step": 2828
+    },
+    {
+      "epoch": 2.3826501965188096,
+      "grad_norm": 0.18260610103607178,
+      "learning_rate": 1.2335369568971362e-06,
+      "loss": 0.3399,
+      "step": 2829
+    },
+    {
+      "epoch": 2.3834924199887704,
+      "grad_norm": 0.18438413739204407,
+      "learning_rate": 1.2303143922099092e-06,
+      "loss": 0.3626,
+      "step": 2830
+    },
+    {
+      "epoch": 2.384334643458731,
+      "grad_norm": 0.18501873314380646,
+      "learning_rate": 1.2270954517867e-06,
+      "loss": 0.3314,
+      "step": 2831
+    },
+    {
+      "epoch": 2.385176866928692,
+      "grad_norm": 0.1844661831855774,
+      "learning_rate": 1.2238801387222716e-06,
+      "loss": 0.3208,
+      "step": 2832
+    },
+    {
+      "epoch": 2.3860190903986522,
+      "grad_norm": 0.17237085103988647,
+      "learning_rate": 1.2206684561079035e-06,
+      "loss": 0.299,
+      "step": 2833
+    },
+    {
+      "epoch": 2.386861313868613,
+      "grad_norm": 0.1830216497182846,
+      "learning_rate": 1.2174604070313811e-06,
+      "loss": 0.3126,
+      "step": 2834
+    },
+    {
+      "epoch": 2.3877035373385738,
+      "grad_norm": 0.17193448543548584,
+      "learning_rate": 1.2142559945769995e-06,
+      "loss": 0.3232,
+      "step": 2835
+    },
+    {
+      "epoch": 2.3885457608085345,
+      "grad_norm": 0.18704761564731598,
+      "learning_rate": 1.211055221825554e-06,
+      "loss": 0.3653,
+      "step": 2836
+    },
+    {
+      "epoch": 2.3893879842784953,
+      "grad_norm": 0.18288898468017578,
+      "learning_rate": 1.207858091854342e-06,
+      "loss": 0.3709,
+      "step": 2837
+    },
+    {
+      "epoch": 2.390230207748456,
+      "grad_norm": 0.17680326104164124,
+      "learning_rate": 1.2046646077371615e-06,
+      "loss": 0.2964,
+      "step": 2838
+    },
+    {
+      "epoch": 2.391072431218417,
+      "grad_norm": 0.18553590774536133,
+      "learning_rate": 1.2014747725443004e-06,
+      "loss": 0.3341,
+      "step": 2839
+    },
+    {
+      "epoch": 2.391914654688377,
+      "grad_norm": 0.18301990628242493,
+      "learning_rate": 1.1982885893425455e-06,
+      "loss": 0.3643,
+      "step": 2840
+    },
+    {
+      "epoch": 2.392756878158338,
+      "grad_norm": 0.1753537952899933,
+      "learning_rate": 1.1951060611951615e-06,
+      "loss": 0.3195,
+      "step": 2841
+    },
+    {
+      "epoch": 2.3935991016282987,
+      "grad_norm": 0.184483602643013,
+      "learning_rate": 1.1919271911619106e-06,
+      "loss": 0.3384,
+      "step": 2842
+    },
+    {
+      "epoch": 2.3944413250982595,
+      "grad_norm": 0.19095274806022644,
+      "learning_rate": 1.1887519822990296e-06,
+      "loss": 0.3437,
+      "step": 2843
+    },
+    {
+      "epoch": 2.3952835485682202,
+      "grad_norm": 0.17648278176784515,
+      "learning_rate": 1.185580437659241e-06,
+      "loss": 0.3209,
+      "step": 2844
+    },
+    {
+      "epoch": 2.3961257720381806,
+      "grad_norm": 0.18717674911022186,
+      "learning_rate": 1.1824125602917414e-06,
+      "loss": 0.3459,
+      "step": 2845
+    },
+    {
+      "epoch": 2.3969679955081413,
+      "grad_norm": 0.18629314005374908,
+      "learning_rate": 1.1792483532422021e-06,
+      "loss": 0.3246,
+      "step": 2846
+    },
+    {
+      "epoch": 2.397810218978102,
+      "grad_norm": 0.19330480694770813,
+      "learning_rate": 1.1760878195527642e-06,
+      "loss": 0.3543,
+      "step": 2847
+    },
+    {
+      "epoch": 2.398652442448063,
+      "grad_norm": 0.17509174346923828,
+      "learning_rate": 1.1729309622620422e-06,
+      "loss": 0.3044,
+      "step": 2848
+    },
+    {
+      "epoch": 2.3994946659180236,
+      "grad_norm": 0.18171800673007965,
+      "learning_rate": 1.1697777844051105e-06,
+      "loss": 0.3116,
+      "step": 2849
+    },
+    {
+      "epoch": 2.4003368893879844,
+      "grad_norm": 0.19177602231502533,
+      "learning_rate": 1.1666282890135083e-06,
+      "loss": 0.3456,
+      "step": 2850
+    },
+    {
+      "epoch": 2.401179112857945,
+      "grad_norm": 0.18695443868637085,
+      "learning_rate": 1.1634824791152334e-06,
+      "loss": 0.3575,
+      "step": 2851
+    },
+    {
+      "epoch": 2.4020213363279055,
+      "grad_norm": 0.1723720133304596,
+      "learning_rate": 1.1603403577347434e-06,
+      "loss": 0.2962,
+      "step": 2852
+    },
+    {
+      "epoch": 2.4028635597978663,
+      "grad_norm": 0.18195010721683502,
+      "learning_rate": 1.1572019278929457e-06,
+      "loss": 0.3296,
+      "step": 2853
+    },
+    {
+      "epoch": 2.403705783267827,
+      "grad_norm": 0.1848120242357254,
+      "learning_rate": 1.1540671926072012e-06,
+      "loss": 0.3466,
+      "step": 2854
+    },
+    {
+      "epoch": 2.404548006737788,
+      "grad_norm": 0.17766617238521576,
+      "learning_rate": 1.1509361548913151e-06,
+      "loss": 0.3489,
+      "step": 2855
+    },
+    {
+      "epoch": 2.4053902302077486,
+      "grad_norm": 0.17193879187107086,
+      "learning_rate": 1.147808817755544e-06,
+      "loss": 0.3216,
+      "step": 2856
+    },
+    {
+      "epoch": 2.4062324536777093,
+      "grad_norm": 0.18416596949100494,
+      "learning_rate": 1.1446851842065804e-06,
+      "loss": 0.3361,
+      "step": 2857
+    },
+    {
+      "epoch": 2.40707467714767,
+      "grad_norm": 0.18768543004989624,
+      "learning_rate": 1.1415652572475628e-06,
+      "loss": 0.3183,
+      "step": 2858
+    },
+    {
+      "epoch": 2.4079169006176304,
+      "grad_norm": 0.1932467818260193,
+      "learning_rate": 1.1384490398780563e-06,
+      "loss": 0.3496,
+      "step": 2859
+    },
+    {
+      "epoch": 2.408759124087591,
+      "grad_norm": 0.18695873022079468,
+      "learning_rate": 1.1353365350940688e-06,
+      "loss": 0.3069,
+      "step": 2860
+    },
+    {
+      "epoch": 2.409601347557552,
+      "grad_norm": 0.17973104119300842,
+      "learning_rate": 1.1322277458880337e-06,
+      "loss": 0.3243,
+      "step": 2861
+    },
+    {
+      "epoch": 2.4104435710275127,
+      "grad_norm": 0.18068546056747437,
+      "learning_rate": 1.129122675248816e-06,
+      "loss": 0.342,
+      "step": 2862
+    },
+    {
+      "epoch": 2.4112857944974735,
+      "grad_norm": 0.16292357444763184,
+      "learning_rate": 1.1260213261617015e-06,
+      "loss": 0.3167,
+      "step": 2863
+    },
+    {
+      "epoch": 2.412128017967434,
+      "grad_norm": 0.18908433616161346,
+      "learning_rate": 1.1229237016084005e-06,
+      "loss": 0.3719,
+      "step": 2864
+    },
+    {
+      "epoch": 2.4129702414373946,
+      "grad_norm": 0.1725766956806183,
+      "learning_rate": 1.1198298045670402e-06,
+      "loss": 0.298,
+      "step": 2865
+    },
+    {
+      "epoch": 2.4138124649073553,
+      "grad_norm": 0.19303034245967865,
+      "learning_rate": 1.116739638012168e-06,
+      "loss": 0.3337,
+      "step": 2866
+    },
+    {
+      "epoch": 2.414654688377316,
+      "grad_norm": 0.17117536067962646,
+      "learning_rate": 1.113653204914742e-06,
+      "loss": 0.3324,
+      "step": 2867
+    },
+    {
+      "epoch": 2.415496911847277,
+      "grad_norm": 0.17073644697666168,
+      "learning_rate": 1.1105705082421303e-06,
+      "loss": 0.3238,
+      "step": 2868
+    },
+    {
+      "epoch": 2.4163391353172377,
+      "grad_norm": 0.1811438351869583,
+      "learning_rate": 1.1074915509581086e-06,
+      "loss": 0.3241,
+      "step": 2869
+    },
+    {
+      "epoch": 2.4171813587871984,
+      "grad_norm": 0.18533632159233093,
+      "learning_rate": 1.104416336022861e-06,
+      "loss": 0.3437,
+      "step": 2870
+    },
+    {
+      "epoch": 2.4180235822571587,
+      "grad_norm": 0.17663425207138062,
+      "learning_rate": 1.1013448663929704e-06,
+      "loss": 0.3603,
+      "step": 2871
+    },
+    {
+      "epoch": 2.4188658057271195,
+      "grad_norm": 0.1756037026643753,
+      "learning_rate": 1.0982771450214197e-06,
+      "loss": 0.3411,
+      "step": 2872
+    },
+    {
+      "epoch": 2.4197080291970803,
+      "grad_norm": 0.18769052624702454,
+      "learning_rate": 1.0952131748575855e-06,
+      "loss": 0.3106,
+      "step": 2873
+    },
+    {
+      "epoch": 2.420550252667041,
+      "grad_norm": 0.18603095412254333,
+      "learning_rate": 1.0921529588472446e-06,
+      "loss": 0.334,
+      "step": 2874
+    },
+    {
+      "epoch": 2.421392476137002,
+      "grad_norm": 0.19260810315608978,
+      "learning_rate": 1.0890964999325566e-06,
+      "loss": 0.3317,
+      "step": 2875
+    },
+    {
+      "epoch": 2.422234699606962,
+      "grad_norm": 0.17378182709217072,
+      "learning_rate": 1.0860438010520773e-06,
+      "loss": 0.3022,
+      "step": 2876
+    },
+    {
+      "epoch": 2.423076923076923,
+      "grad_norm": 0.18065233528614044,
+      "learning_rate": 1.0829948651407374e-06,
+      "loss": 0.3406,
+      "step": 2877
+    },
+    {
+      "epoch": 2.4239191465468837,
+      "grad_norm": 0.16788865625858307,
+      "learning_rate": 1.0799496951298595e-06,
+      "loss": 0.3158,
+      "step": 2878
+    },
+    {
+      "epoch": 2.4247613700168444,
+      "grad_norm": 0.16910141706466675,
+      "learning_rate": 1.0769082939471382e-06,
+      "loss": 0.3344,
+      "step": 2879
+    },
+    {
+      "epoch": 2.425603593486805,
+      "grad_norm": 0.1839236170053482,
+      "learning_rate": 1.0738706645166508e-06,
+      "loss": 0.3453,
+      "step": 2880
+    },
+    {
+      "epoch": 2.426445816956766,
+      "grad_norm": 0.16786138713359833,
+      "learning_rate": 1.0708368097588435e-06,
+      "loss": 0.3309,
+      "step": 2881
+    },
+    {
+      "epoch": 2.4272880404267267,
+      "grad_norm": 0.17052896320819855,
+      "learning_rate": 1.0678067325905362e-06,
+      "loss": 0.3456,
+      "step": 2882
+    },
+    {
+      "epoch": 2.428130263896687,
+      "grad_norm": 0.17799879610538483,
+      "learning_rate": 1.0647804359249143e-06,
+      "loss": 0.319,
+      "step": 2883
+    },
+    {
+      "epoch": 2.428972487366648,
+      "grad_norm": 0.1701173037290573,
+      "learning_rate": 1.0617579226715324e-06,
+      "loss": 0.3113,
+      "step": 2884
+    },
+    {
+      "epoch": 2.4298147108366086,
+      "grad_norm": 0.18232646584510803,
+      "learning_rate": 1.0587391957363053e-06,
+      "loss": 0.3778,
+      "step": 2885
+    },
+    {
+      "epoch": 2.4306569343065694,
+      "grad_norm": 0.15923668444156647,
+      "learning_rate": 1.0557242580215066e-06,
+      "loss": 0.3241,
+      "step": 2886
+    },
+    {
+      "epoch": 2.43149915777653,
+      "grad_norm": 0.17628398537635803,
+      "learning_rate": 1.0527131124257677e-06,
+      "loss": 0.3057,
+      "step": 2887
+    },
+    {
+      "epoch": 2.432341381246491,
+      "grad_norm": 0.18149569630622864,
+      "learning_rate": 1.0497057618440765e-06,
+      "loss": 0.3219,
+      "step": 2888
+    },
+    {
+      "epoch": 2.4331836047164517,
+      "grad_norm": 0.18324145674705505,
+      "learning_rate": 1.0467022091677692e-06,
+      "loss": 0.3294,
+      "step": 2889
+    },
+    {
+      "epoch": 2.434025828186412,
+      "grad_norm": 0.1819029152393341,
+      "learning_rate": 1.0437024572845317e-06,
+      "loss": 0.3265,
+      "step": 2890
+    },
+    {
+      "epoch": 2.4348680516563728,
+      "grad_norm": 0.18516305088996887,
+      "learning_rate": 1.040706509078394e-06,
+      "loss": 0.342,
+      "step": 2891
+    },
+    {
+      "epoch": 2.4357102751263335,
+      "grad_norm": 0.1963118612766266,
+      "learning_rate": 1.037714367429734e-06,
+      "loss": 0.3462,
+      "step": 2892
+    },
+    {
+      "epoch": 2.4365524985962943,
+      "grad_norm": 0.17552661895751953,
+      "learning_rate": 1.0347260352152644e-06,
+      "loss": 0.2975,
+      "step": 2893
+    },
+    {
+      "epoch": 2.437394722066255,
+      "grad_norm": 0.18327851593494415,
+      "learning_rate": 1.0317415153080406e-06,
+      "loss": 0.3481,
+      "step": 2894
+    },
+    {
+      "epoch": 2.4382369455362154,
+      "grad_norm": 0.17430095374584198,
+      "learning_rate": 1.0287608105774456e-06,
+      "loss": 0.3583,
+      "step": 2895
+    },
+    {
+      "epoch": 2.439079169006176,
+      "grad_norm": 0.1779869645833969,
+      "learning_rate": 1.025783923889202e-06,
+      "loss": 0.3114,
+      "step": 2896
+    },
+    {
+      "epoch": 2.439921392476137,
+      "grad_norm": 0.1845777928829193,
+      "learning_rate": 1.0228108581053565e-06,
+      "loss": 0.354,
+      "step": 2897
+    },
+    {
+      "epoch": 2.4407636159460977,
+      "grad_norm": 0.1726304590702057,
+      "learning_rate": 1.019841616084286e-06,
+      "loss": 0.323,
+      "step": 2898
+    },
+    {
+      "epoch": 2.4416058394160585,
+      "grad_norm": 0.17444516718387604,
+      "learning_rate": 1.0168762006806886e-06,
+      "loss": 0.3091,
+      "step": 2899
+    },
+    {
+      "epoch": 2.4424480628860192,
+      "grad_norm": 0.1818244457244873,
+      "learning_rate": 1.0139146147455842e-06,
+      "loss": 0.3464,
+      "step": 2900
+    },
+    {
+      "epoch": 2.44329028635598,
+      "grad_norm": 0.17454013228416443,
+      "learning_rate": 1.0109568611263094e-06,
+      "loss": 0.3061,
+      "step": 2901
+    },
+    {
+      "epoch": 2.4441325098259403,
+      "grad_norm": 0.17296692728996277,
+      "learning_rate": 1.0080029426665194e-06,
+      "loss": 0.3122,
+      "step": 2902
+    },
+    {
+      "epoch": 2.444974733295901,
+      "grad_norm": 0.17483364045619965,
+      "learning_rate": 1.0050528622061805e-06,
+      "loss": 0.3076,
+      "step": 2903
+    },
+    {
+      "epoch": 2.445816956765862,
+      "grad_norm": 0.1877381056547165,
+      "learning_rate": 1.002106622581569e-06,
+      "loss": 0.3669,
+      "step": 2904
+    },
+    {
+      "epoch": 2.4466591802358226,
+      "grad_norm": 0.1751059740781784,
+      "learning_rate": 9.991642266252672e-07,
+      "loss": 0.3014,
+      "step": 2905
+    },
+    {
+      "epoch": 2.4475014037057834,
+      "grad_norm": 0.19321346282958984,
+      "learning_rate": 9.96225677166166e-07,
+      "loss": 0.3562,
+      "step": 2906
+    },
+    {
+      "epoch": 2.4483436271757437,
+      "grad_norm": 0.1792708784341812,
+      "learning_rate": 9.932909770294542e-07,
+      "loss": 0.3156,
+      "step": 2907
+    },
+    {
+      "epoch": 2.4491858506457045,
+      "grad_norm": 0.1754908710718155,
+      "learning_rate": 9.903601290366217e-07,
+      "loss": 0.3201,
+      "step": 2908
+    },
+    {
+      "epoch": 2.4500280741156653,
+      "grad_norm": 0.18347208201885223,
+      "learning_rate": 9.87433136005454e-07,
+      "loss": 0.3294,
+      "step": 2909
+    },
+    {
+      "epoch": 2.450870297585626,
+      "grad_norm": 0.19159957766532898,
+      "learning_rate": 9.845100007500292e-07,
+      "loss": 0.3835,
+      "step": 2910
+    },
+    {
+      "epoch": 2.451712521055587,
+      "grad_norm": 0.16591417789459229,
+      "learning_rate": 9.81590726080721e-07,
+      "loss": 0.2709,
+      "step": 2911
+    },
+    {
+      "epoch": 2.4525547445255476,
+      "grad_norm": 0.19382108747959137,
+      "learning_rate": 9.786753148041871e-07,
+      "loss": 0.3401,
+      "step": 2912
+    },
+    {
+      "epoch": 2.4533969679955083,
+      "grad_norm": 0.1861683875322342,
+      "learning_rate": 9.757637697233723e-07,
+      "loss": 0.3328,
+      "step": 2913
+    },
+    {
+      "epoch": 2.4542391914654687,
+      "grad_norm": 0.1860036998987198,
+      "learning_rate": 9.728560936375032e-07,
+      "loss": 0.3403,
+      "step": 2914
+    },
+    {
+      "epoch": 2.4550814149354294,
+      "grad_norm": 0.19895556569099426,
+      "learning_rate": 9.699522893420894e-07,
+      "loss": 0.3314,
+      "step": 2915
+    },
+    {
+      "epoch": 2.45592363840539,
+      "grad_norm": 0.18061065673828125,
+      "learning_rate": 9.670523596289138e-07,
+      "loss": 0.3118,
+      "step": 2916
+    },
+    {
+      "epoch": 2.456765861875351,
+      "grad_norm": 0.17831942439079285,
+      "learning_rate": 9.641563072860416e-07,
+      "loss": 0.3282,
+      "step": 2917
+    },
+    {
+      "epoch": 2.4576080853453117,
+      "grad_norm": 0.1895962655544281,
+      "learning_rate": 9.61264135097799e-07,
+      "loss": 0.3429,
+      "step": 2918
+    },
+    {
+      "epoch": 2.4584503088152725,
+      "grad_norm": 0.18186987936496735,
+      "learning_rate": 9.58375845844793e-07,
+      "loss": 0.322,
+      "step": 2919
+    },
+    {
+      "epoch": 2.4592925322852333,
+      "grad_norm": 0.1867487132549286,
+      "learning_rate": 9.55491442303889e-07,
+      "loss": 0.3304,
+      "step": 2920
+    },
+    {
+      "epoch": 2.4601347557551936,
+      "grad_norm": 0.1775023490190506,
+      "learning_rate": 9.526109272482237e-07,
+      "loss": 0.3296,
+      "step": 2921
+    },
+    {
+      "epoch": 2.4609769792251543,
+      "grad_norm": 0.18398189544677734,
+      "learning_rate": 9.497343034471896e-07,
+      "loss": 0.3703,
+      "step": 2922
+    },
+    {
+      "epoch": 2.461819202695115,
+      "grad_norm": 0.16970793902873993,
+      "learning_rate": 9.468615736664405e-07,
+      "loss": 0.3224,
+      "step": 2923
+    },
+    {
+      "epoch": 2.462661426165076,
+      "grad_norm": 0.18678072094917297,
+      "learning_rate": 9.439927406678845e-07,
+      "loss": 0.3703,
+      "step": 2924
+    },
+    {
+      "epoch": 2.4635036496350367,
+      "grad_norm": 0.17588737607002258,
+      "learning_rate": 9.41127807209688e-07,
+      "loss": 0.35,
+      "step": 2925
+    },
+    {
+      "epoch": 2.464345873104997,
+      "grad_norm": 0.15737545490264893,
+      "learning_rate": 9.382667760462633e-07,
+      "loss": 0.2765,
+      "step": 2926
+    },
+    {
+      "epoch": 2.4651880965749577,
+      "grad_norm": 0.19520333409309387,
+      "learning_rate": 9.354096499282728e-07,
+      "loss": 0.3557,
+      "step": 2927
+    },
+    {
+      "epoch": 2.4660303200449185,
+      "grad_norm": 0.17047427594661713,
+      "learning_rate": 9.325564316026236e-07,
+      "loss": 0.2859,
+      "step": 2928
+    },
+    {
+      "epoch": 2.4668725435148793,
+      "grad_norm": 0.19578218460083008,
+      "learning_rate": 9.297071238124683e-07,
+      "loss": 0.3369,
+      "step": 2929
+    },
+    {
+      "epoch": 2.46771476698484,
+      "grad_norm": 0.19017136096954346,
+      "learning_rate": 9.268617292971982e-07,
+      "loss": 0.3502,
+      "step": 2930
+    },
+    {
+      "epoch": 2.468556990454801,
+      "grad_norm": 0.17374084889888763,
+      "learning_rate": 9.240202507924412e-07,
+      "loss": 0.3248,
+      "step": 2931
+    },
+    {
+      "epoch": 2.4693992139247616,
+      "grad_norm": 0.18185202777385712,
+      "learning_rate": 9.211826910300598e-07,
+      "loss": 0.3337,
+      "step": 2932
+    },
+    {
+      "epoch": 2.470241437394722,
+      "grad_norm": 0.17019344866275787,
+      "learning_rate": 9.183490527381539e-07,
+      "loss": 0.3557,
+      "step": 2933
+    },
+    {
+      "epoch": 2.4710836608646827,
+      "grad_norm": 0.17223379015922546,
+      "learning_rate": 9.155193386410466e-07,
+      "loss": 0.3242,
+      "step": 2934
+    },
+    {
+      "epoch": 2.4719258843346434,
+      "grad_norm": 0.18503350019454956,
+      "learning_rate": 9.126935514592949e-07,
+      "loss": 0.3253,
+      "step": 2935
+    },
+    {
+      "epoch": 2.472768107804604,
+      "grad_norm": 0.1900034248828888,
+      "learning_rate": 9.098716939096719e-07,
+      "loss": 0.3339,
+      "step": 2936
+    },
+    {
+      "epoch": 2.473610331274565,
+      "grad_norm": 0.20346327126026154,
+      "learning_rate": 9.070537687051817e-07,
+      "loss": 0.376,
+      "step": 2937
+    },
+    {
+      "epoch": 2.4744525547445253,
+      "grad_norm": 0.18149623274803162,
+      "learning_rate": 9.042397785550405e-07,
+      "loss": 0.333,
+      "step": 2938
+    },
+    {
+      "epoch": 2.475294778214486,
+      "grad_norm": 0.18477939069271088,
+      "learning_rate": 9.014297261646876e-07,
+      "loss": 0.3282,
+      "step": 2939
+    },
+    {
+      "epoch": 2.476137001684447,
+      "grad_norm": 0.18960389494895935,
+      "learning_rate": 8.986236142357707e-07,
+      "loss": 0.31,
+      "step": 2940
+    },
+    {
+      "epoch": 2.4769792251544076,
+      "grad_norm": 0.1833685338497162,
+      "learning_rate": 8.958214454661529e-07,
+      "loss": 0.356,
+      "step": 2941
+    },
+    {
+      "epoch": 2.4778214486243684,
+      "grad_norm": 0.17379865050315857,
+      "learning_rate": 8.930232225499025e-07,
+      "loss": 0.3437,
+      "step": 2942
+    },
+    {
+      "epoch": 2.478663672094329,
+      "grad_norm": 0.1800137162208557,
+      "learning_rate": 8.902289481772996e-07,
+      "loss": 0.3334,
+      "step": 2943
+    },
+    {
+      "epoch": 2.47950589556429,
+      "grad_norm": 0.17582203447818756,
+      "learning_rate": 8.874386250348232e-07,
+      "loss": 0.3356,
+      "step": 2944
+    },
+    {
+      "epoch": 2.4803481190342502,
+      "grad_norm": 0.17575836181640625,
+      "learning_rate": 8.846522558051563e-07,
+      "loss": 0.3388,
+      "step": 2945
+    },
+    {
+      "epoch": 2.481190342504211,
+      "grad_norm": 0.2217385172843933,
+      "learning_rate": 8.818698431671774e-07,
+      "loss": 0.3411,
+      "step": 2946
+    },
+    {
+      "epoch": 2.4820325659741718,
+      "grad_norm": 0.20173339545726776,
+      "learning_rate": 8.790913897959663e-07,
+      "loss": 0.3125,
+      "step": 2947
+    },
+    {
+      "epoch": 2.4828747894441325,
+      "grad_norm": 0.1942809671163559,
+      "learning_rate": 8.763168983627912e-07,
+      "loss": 0.3455,
+      "step": 2948
+    },
+    {
+      "epoch": 2.4837170129140933,
+      "grad_norm": 0.1882442831993103,
+      "learning_rate": 8.735463715351139e-07,
+      "loss": 0.2995,
+      "step": 2949
+    },
+    {
+      "epoch": 2.484559236384054,
+      "grad_norm": 0.1855713129043579,
+      "learning_rate": 8.70779811976582e-07,
+      "loss": 0.3468,
+      "step": 2950
+    },
+    {
+      "epoch": 2.485401459854015,
+      "grad_norm": 0.41547971963882446,
+      "learning_rate": 8.680172223470329e-07,
+      "loss": 0.3343,
+      "step": 2951
+    },
+    {
+      "epoch": 2.486243683323975,
+      "grad_norm": 0.18178273737430573,
+      "learning_rate": 8.652586053024836e-07,
+      "loss": 0.3185,
+      "step": 2952
+    },
+    {
+      "epoch": 2.487085906793936,
+      "grad_norm": 0.20105557143688202,
+      "learning_rate": 8.625039634951354e-07,
+      "loss": 0.3473,
+      "step": 2953
+    },
+    {
+      "epoch": 2.4879281302638967,
+      "grad_norm": 0.1699233204126358,
+      "learning_rate": 8.597532995733615e-07,
+      "loss": 0.2968,
+      "step": 2954
+    },
+    {
+      "epoch": 2.4887703537338575,
+      "grad_norm": 0.1811775267124176,
+      "learning_rate": 8.570066161817176e-07,
+      "loss": 0.3389,
+      "step": 2955
+    },
+    {
+      "epoch": 2.4896125772038182,
+      "grad_norm": 0.18310898542404175,
+      "learning_rate": 8.542639159609278e-07,
+      "loss": 0.333,
+      "step": 2956
+    },
+    {
+      "epoch": 2.4904548006737786,
+      "grad_norm": 0.173427015542984,
+      "learning_rate": 8.515252015478915e-07,
+      "loss": 0.3211,
+      "step": 2957
+    },
+    {
+      "epoch": 2.4912970241437393,
+      "grad_norm": 0.19226166605949402,
+      "learning_rate": 8.487904755756676e-07,
+      "loss": 0.3345,
+      "step": 2958
+    },
+    {
+      "epoch": 2.4921392476137,
+      "grad_norm": 0.17834272980690002,
+      "learning_rate": 8.460597406734905e-07,
+      "loss": 0.3325,
+      "step": 2959
+    },
+    {
+      "epoch": 2.492981471083661,
+      "grad_norm": 0.17401964962482452,
+      "learning_rate": 8.433329994667488e-07,
+      "loss": 0.3249,
+      "step": 2960
+    },
+    {
+      "epoch": 2.4938236945536216,
+      "grad_norm": 0.19734671711921692,
+      "learning_rate": 8.406102545769989e-07,
+      "loss": 0.3586,
+      "step": 2961
+    },
+    {
+      "epoch": 2.4946659180235824,
+      "grad_norm": 0.1604025810956955,
+      "learning_rate": 8.378915086219497e-07,
+      "loss": 0.2906,
+      "step": 2962
+    },
+    {
+      "epoch": 2.495508141493543,
+      "grad_norm": 0.17842164635658264,
+      "learning_rate": 8.351767642154673e-07,
+      "loss": 0.336,
+      "step": 2963
+    },
+    {
+      "epoch": 2.4963503649635035,
+      "grad_norm": 0.38500332832336426,
+      "learning_rate": 8.324660239675697e-07,
+      "loss": 0.3368,
+      "step": 2964
+    },
+    {
+      "epoch": 2.4971925884334643,
+      "grad_norm": 0.17705510556697845,
+      "learning_rate": 8.297592904844282e-07,
+      "loss": 0.3123,
+      "step": 2965
+    },
+    {
+      "epoch": 2.498034811903425,
+      "grad_norm": 0.18042810261249542,
+      "learning_rate": 8.270565663683583e-07,
+      "loss": 0.3285,
+      "step": 2966
+    },
+    {
+      "epoch": 2.498877035373386,
+      "grad_norm": 0.18032227456569672,
+      "learning_rate": 8.243578542178227e-07,
+      "loss": 0.3032,
+      "step": 2967
+    },
+    {
+      "epoch": 2.4997192588433466,
+      "grad_norm": 0.19498011469841003,
+      "learning_rate": 8.216631566274252e-07,
+      "loss": 0.3297,
+      "step": 2968
+    },
+    {
+      "epoch": 2.500561482313307,
+      "grad_norm": 0.17469584941864014,
+      "learning_rate": 8.189724761879131e-07,
+      "loss": 0.3529,
+      "step": 2969
+    },
+    {
+      "epoch": 2.501403705783268,
+      "grad_norm": 0.1745612919330597,
+      "learning_rate": 8.16285815486168e-07,
+      "loss": 0.3364,
+      "step": 2970
+    },
+    {
+      "epoch": 2.5022459292532284,
+      "grad_norm": 0.18163642287254333,
+      "learning_rate": 8.13603177105211e-07,
+      "loss": 0.3452,
+      "step": 2971
+    },
+    {
+      "epoch": 2.503088152723189,
+      "grad_norm": 0.1809900403022766,
+      "learning_rate": 8.109245636241892e-07,
+      "loss": 0.3404,
+      "step": 2972
+    },
+    {
+      "epoch": 2.50393037619315,
+      "grad_norm": 0.16917380690574646,
+      "learning_rate": 8.082499776183883e-07,
+      "loss": 0.2734,
+      "step": 2973
+    },
+    {
+      "epoch": 2.5047725996631107,
+      "grad_norm": 0.19013775885105133,
+      "learning_rate": 8.05579421659215e-07,
+      "loss": 0.3715,
+      "step": 2974
+    },
+    {
+      "epoch": 2.5056148231330715,
+      "grad_norm": 0.18309292197227478,
+      "learning_rate": 8.029128983142076e-07,
+      "loss": 0.3175,
+      "step": 2975
+    },
+    {
+      "epoch": 2.506457046603032,
+      "grad_norm": 0.17404715716838837,
+      "learning_rate": 8.002504101470204e-07,
+      "loss": 0.2973,
+      "step": 2976
+    },
+    {
+      "epoch": 2.5072992700729926,
+      "grad_norm": 0.17378441989421844,
+      "learning_rate": 7.975919597174342e-07,
+      "loss": 0.3214,
+      "step": 2977
+    },
+    {
+      "epoch": 2.5081414935429533,
+      "grad_norm": 0.17642168700695038,
+      "learning_rate": 7.949375495813439e-07,
+      "loss": 0.3262,
+      "step": 2978
+    },
+    {
+      "epoch": 2.508983717012914,
+      "grad_norm": 0.1785106658935547,
+      "learning_rate": 7.922871822907641e-07,
+      "loss": 0.3236,
+      "step": 2979
+    },
+    {
+      "epoch": 2.509825940482875,
+      "grad_norm": 0.1916131228208542,
+      "learning_rate": 7.896408603938194e-07,
+      "loss": 0.2964,
+      "step": 2980
+    },
+    {
+      "epoch": 2.5106681639528357,
+      "grad_norm": 0.2003284990787506,
+      "learning_rate": 7.869985864347424e-07,
+      "loss": 0.3688,
+      "step": 2981
+    },
+    {
+      "epoch": 2.5115103874227964,
+      "grad_norm": 0.1730331927537918,
+      "learning_rate": 7.843603629538804e-07,
+      "loss": 0.3459,
+      "step": 2982
+    },
+    {
+      "epoch": 2.5123526108927567,
+      "grad_norm": 0.1806933879852295,
+      "learning_rate": 7.817261924876812e-07,
+      "loss": 0.2955,
+      "step": 2983
+    },
+    {
+      "epoch": 2.5131948343627175,
+      "grad_norm": 0.18728403747081757,
+      "learning_rate": 7.790960775687001e-07,
+      "loss": 0.3404,
+      "step": 2984
+    },
+    {
+      "epoch": 2.5140370578326783,
+      "grad_norm": 0.19027812778949738,
+      "learning_rate": 7.764700207255904e-07,
+      "loss": 0.3419,
+      "step": 2985
+    },
+    {
+      "epoch": 2.514879281302639,
+      "grad_norm": 0.17174813151359558,
+      "learning_rate": 7.738480244831042e-07,
+      "loss": 0.3349,
+      "step": 2986
+    },
+    {
+      "epoch": 2.5157215047726,
+      "grad_norm": 0.2719394266605377,
+      "learning_rate": 7.71230091362089e-07,
+      "loss": 0.332,
+      "step": 2987
+    },
+    {
+      "epoch": 2.51656372824256,
+      "grad_norm": 0.1689806878566742,
+      "learning_rate": 7.686162238794898e-07,
+      "loss": 0.314,
+      "step": 2988
+    },
+    {
+      "epoch": 2.517405951712521,
+      "grad_norm": 0.18157345056533813,
+      "learning_rate": 7.660064245483384e-07,
+      "loss": 0.3358,
+      "step": 2989
+    },
+    {
+      "epoch": 2.5182481751824817,
+      "grad_norm": 0.16696105897426605,
+      "learning_rate": 7.634006958777568e-07,
+      "loss": 0.3185,
+      "step": 2990
+    },
+    {
+      "epoch": 2.5190903986524424,
+      "grad_norm": 0.17663368582725525,
+      "learning_rate": 7.607990403729526e-07,
+      "loss": 0.3406,
+      "step": 2991
+    },
+    {
+      "epoch": 2.519932622122403,
+      "grad_norm": 0.18124790489673615,
+      "learning_rate": 7.582014605352206e-07,
+      "loss": 0.3177,
+      "step": 2992
+    },
+    {
+      "epoch": 2.520774845592364,
+      "grad_norm": 0.18290624022483826,
+      "learning_rate": 7.556079588619341e-07,
+      "loss": 0.3303,
+      "step": 2993
+    },
+    {
+      "epoch": 2.5216170690623247,
+      "grad_norm": 0.16427665948867798,
+      "learning_rate": 7.530185378465459e-07,
+      "loss": 0.3084,
+      "step": 2994
+    },
+    {
+      "epoch": 2.522459292532285,
+      "grad_norm": 0.19871890544891357,
+      "learning_rate": 7.504331999785852e-07,
+      "loss": 0.3531,
+      "step": 2995
+    },
+    {
+      "epoch": 2.523301516002246,
+      "grad_norm": 0.17493316531181335,
+      "learning_rate": 7.47851947743658e-07,
+      "loss": 0.3122,
+      "step": 2996
+    },
+    {
+      "epoch": 2.5241437394722066,
+      "grad_norm": 0.1823892593383789,
+      "learning_rate": 7.452747836234392e-07,
+      "loss": 0.333,
+      "step": 2997
+    },
+    {
+      "epoch": 2.5249859629421674,
+      "grad_norm": 0.1649978756904602,
+      "learning_rate": 7.427017100956779e-07,
+      "loss": 0.2861,
+      "step": 2998
+    },
+    {
+      "epoch": 2.525828186412128,
+      "grad_norm": 0.17946523427963257,
+      "learning_rate": 7.401327296341826e-07,
+      "loss": 0.3261,
+      "step": 2999
+    },
+    {
+      "epoch": 2.5266704098820885,
+      "grad_norm": 0.18829745054244995,
+      "learning_rate": 7.375678447088347e-07,
+      "loss": 0.3552,
+      "step": 3000
+    },
+    {
+      "epoch": 2.5275126333520497,
+      "grad_norm": 0.18125388026237488,
+      "learning_rate": 7.350070577855716e-07,
+      "loss": 0.34,
+      "step": 3001
+    },
+    {
+      "epoch": 2.52835485682201,
+      "grad_norm": 0.18347607553005219,
+      "learning_rate": 7.324503713263975e-07,
+      "loss": 0.3344,
+      "step": 3002
+    },
+    {
+      "epoch": 2.5291970802919708,
+      "grad_norm": 0.16962610185146332,
+      "learning_rate": 7.298977877893688e-07,
+      "loss": 0.3393,
+      "step": 3003
+    },
+    {
+      "epoch": 2.5300393037619315,
+      "grad_norm": 0.17049525678157806,
+      "learning_rate": 7.273493096285989e-07,
+      "loss": 0.3219,
+      "step": 3004
+    },
+    {
+      "epoch": 2.5308815272318923,
+      "grad_norm": 0.18462485074996948,
+      "learning_rate": 7.24804939294253e-07,
+      "loss": 0.3143,
+      "step": 3005
+    },
+    {
+      "epoch": 2.531723750701853,
+      "grad_norm": 0.16474902629852295,
+      "learning_rate": 7.222646792325516e-07,
+      "loss": 0.3203,
+      "step": 3006
+    },
+    {
+      "epoch": 2.5325659741718134,
+      "grad_norm": 0.17993924021720886,
+      "learning_rate": 7.197285318857584e-07,
+      "loss": 0.3542,
+      "step": 3007
+    },
+    {
+      "epoch": 2.533408197641774,
+      "grad_norm": 0.20460404455661774,
+      "learning_rate": 7.171964996921848e-07,
+      "loss": 0.3391,
+      "step": 3008
+    },
+    {
+      "epoch": 2.534250421111735,
+      "grad_norm": 0.1803237944841385,
+      "learning_rate": 7.146685850861851e-07,
+      "loss": 0.3286,
+      "step": 3009
+    },
+    {
+      "epoch": 2.5350926445816957,
+      "grad_norm": 0.19286656379699707,
+      "learning_rate": 7.121447904981571e-07,
+      "loss": 0.3516,
+      "step": 3010
+    },
+    {
+      "epoch": 2.5359348680516565,
+      "grad_norm": 0.1730838119983673,
+      "learning_rate": 7.096251183545355e-07,
+      "loss": 0.3244,
+      "step": 3011
+    },
+    {
+      "epoch": 2.5367770915216172,
+      "grad_norm": 0.18125079572200775,
+      "learning_rate": 7.071095710777925e-07,
+      "loss": 0.3491,
+      "step": 3012
+    },
+    {
+      "epoch": 2.537619314991578,
+      "grad_norm": 0.18218089640140533,
+      "learning_rate": 7.045981510864319e-07,
+      "loss": 0.3676,
+      "step": 3013
+    },
+    {
+      "epoch": 2.5384615384615383,
+      "grad_norm": 0.17780695855617523,
+      "learning_rate": 7.02090860794995e-07,
+      "loss": 0.3324,
+      "step": 3014
+    },
+    {
+      "epoch": 2.539303761931499,
+      "grad_norm": 0.1743364930152893,
+      "learning_rate": 6.995877026140468e-07,
+      "loss": 0.3368,
+      "step": 3015
+    },
+    {
+      "epoch": 2.54014598540146,
+      "grad_norm": 0.1723746955394745,
+      "learning_rate": 6.970886789501851e-07,
+      "loss": 0.3074,
+      "step": 3016
+    },
+    {
+      "epoch": 2.5409882088714206,
+      "grad_norm": 0.19902728497982025,
+      "learning_rate": 6.945937922060259e-07,
+      "loss": 0.3647,
+      "step": 3017
+    },
+    {
+      "epoch": 2.5418304323413814,
+      "grad_norm": 0.19043061137199402,
+      "learning_rate": 6.921030447802146e-07,
+      "loss": 0.3365,
+      "step": 3018
+    },
+    {
+      "epoch": 2.5426726558113417,
+      "grad_norm": 0.1770019680261612,
+      "learning_rate": 6.896164390674125e-07,
+      "loss": 0.3303,
+      "step": 3019
+    },
+    {
+      "epoch": 2.5435148792813025,
+      "grad_norm": 0.17343956232070923,
+      "learning_rate": 6.871339774583025e-07,
+      "loss": 0.3198,
+      "step": 3020
+    },
+    {
+      "epoch": 2.5443571027512633,
+      "grad_norm": 0.19765575230121613,
+      "learning_rate": 6.846556623395795e-07,
+      "loss": 0.3465,
+      "step": 3021
+    },
+    {
+      "epoch": 2.545199326221224,
+      "grad_norm": 0.1678420454263687,
+      "learning_rate": 6.821814960939549e-07,
+      "loss": 0.3295,
+      "step": 3022
+    },
+    {
+      "epoch": 2.546041549691185,
+      "grad_norm": 0.16188831627368927,
+      "learning_rate": 6.797114811001482e-07,
+      "loss": 0.305,
+      "step": 3023
+    },
+    {
+      "epoch": 2.5468837731611456,
+      "grad_norm": 0.1842898577451706,
+      "learning_rate": 6.772456197328919e-07,
+      "loss": 0.372,
+      "step": 3024
+    },
+    {
+      "epoch": 2.5477259966311063,
+      "grad_norm": 0.18573446571826935,
+      "learning_rate": 6.74783914362922e-07,
+      "loss": 0.333,
+      "step": 3025
+    },
+    {
+      "epoch": 2.5485682201010667,
+      "grad_norm": 0.16520562767982483,
+      "learning_rate": 6.723263673569796e-07,
+      "loss": 0.2889,
+      "step": 3026
+    },
+    {
+      "epoch": 2.5494104435710274,
+      "grad_norm": 0.19662410020828247,
+      "learning_rate": 6.698729810778065e-07,
+      "loss": 0.3318,
+      "step": 3027
+    },
+    {
+      "epoch": 2.550252667040988,
+      "grad_norm": 0.18073995411396027,
+      "learning_rate": 6.674237578841486e-07,
+      "loss": 0.3625,
+      "step": 3028
+    },
+    {
+      "epoch": 2.551094890510949,
+      "grad_norm": 0.17097055912017822,
+      "learning_rate": 6.649787001307451e-07,
+      "loss": 0.2885,
+      "step": 3029
+    },
+    {
+      "epoch": 2.5519371139809097,
+      "grad_norm": 0.18093878030776978,
+      "learning_rate": 6.625378101683317e-07,
+      "loss": 0.338,
+      "step": 3030
+    },
+    {
+      "epoch": 2.55277933745087,
+      "grad_norm": 0.17633871734142303,
+      "learning_rate": 6.601010903436355e-07,
+      "loss": 0.3234,
+      "step": 3031
+    },
+    {
+      "epoch": 2.5536215609208313,
+      "grad_norm": 0.17590993642807007,
+      "learning_rate": 6.57668542999379e-07,
+      "loss": 0.3093,
+      "step": 3032
+    },
+    {
+      "epoch": 2.5544637843907916,
+      "grad_norm": 0.1865670382976532,
+      "learning_rate": 6.552401704742678e-07,
+      "loss": 0.3262,
+      "step": 3033
+    },
+    {
+      "epoch": 2.5553060078607523,
+      "grad_norm": 0.18153560161590576,
+      "learning_rate": 6.528159751029988e-07,
+      "loss": 0.3753,
+      "step": 3034
+    },
+    {
+      "epoch": 2.556148231330713,
+      "grad_norm": 0.1866806298494339,
+      "learning_rate": 6.503959592162468e-07,
+      "loss": 0.3234,
+      "step": 3035
+    },
+    {
+      "epoch": 2.556990454800674,
+      "grad_norm": 0.18803523480892181,
+      "learning_rate": 6.479801251406748e-07,
+      "loss": 0.3966,
+      "step": 3036
+    },
+    {
+      "epoch": 2.5578326782706347,
+      "grad_norm": 0.1652144193649292,
+      "learning_rate": 6.455684751989194e-07,
+      "loss": 0.33,
+      "step": 3037
+    },
+    {
+      "epoch": 2.558674901740595,
+      "grad_norm": 0.18305109441280365,
+      "learning_rate": 6.431610117095999e-07,
+      "loss": 0.3395,
+      "step": 3038
+    },
+    {
+      "epoch": 2.5595171252105557,
+      "grad_norm": 0.21174503862857819,
+      "learning_rate": 6.40757736987307e-07,
+      "loss": 0.327,
+      "step": 3039
+    },
+    {
+      "epoch": 2.5603593486805165,
+      "grad_norm": 0.17883852124214172,
+      "learning_rate": 6.383586533426051e-07,
+      "loss": 0.3276,
+      "step": 3040
+    },
+    {
+      "epoch": 2.5612015721504773,
+      "grad_norm": 0.17874382436275482,
+      "learning_rate": 6.359637630820292e-07,
+      "loss": 0.3482,
+      "step": 3041
+    },
+    {
+      "epoch": 2.562043795620438,
+      "grad_norm": 0.17511144280433655,
+      "learning_rate": 6.335730685080838e-07,
+      "loss": 0.3241,
+      "step": 3042
+    },
+    {
+      "epoch": 2.562886019090399,
+      "grad_norm": 0.18171054124832153,
+      "learning_rate": 6.311865719192384e-07,
+      "loss": 0.3565,
+      "step": 3043
+    },
+    {
+      "epoch": 2.5637282425603596,
+      "grad_norm": 0.16010089218616486,
+      "learning_rate": 6.28804275609926e-07,
+      "loss": 0.3,
+      "step": 3044
+    },
+    {
+      "epoch": 2.56457046603032,
+      "grad_norm": 0.17939692735671997,
+      "learning_rate": 6.26426181870542e-07,
+      "loss": 0.3338,
+      "step": 3045
+    },
+    {
+      "epoch": 2.5654126895002807,
+      "grad_norm": 0.18418268859386444,
+      "learning_rate": 6.24052292987442e-07,
+      "loss": 0.3296,
+      "step": 3046
+    },
+    {
+      "epoch": 2.5662549129702414,
+      "grad_norm": 0.1782510131597519,
+      "learning_rate": 6.216826112429391e-07,
+      "loss": 0.3392,
+      "step": 3047
+    },
+    {
+      "epoch": 2.567097136440202,
+      "grad_norm": 0.17586378753185272,
+      "learning_rate": 6.193171389152996e-07,
+      "loss": 0.3384,
+      "step": 3048
+    },
+    {
+      "epoch": 2.567939359910163,
+      "grad_norm": 0.19140559434890747,
+      "learning_rate": 6.169558782787438e-07,
+      "loss": 0.3345,
+      "step": 3049
+    },
+    {
+      "epoch": 2.5687815833801233,
+      "grad_norm": 0.18451711535453796,
+      "learning_rate": 6.145988316034441e-07,
+      "loss": 0.3388,
+      "step": 3050
+    },
+    {
+      "epoch": 2.569623806850084,
+      "grad_norm": 0.16443663835525513,
+      "learning_rate": 6.122460011555187e-07,
+      "loss": 0.3415,
+      "step": 3051
+    },
+    {
+      "epoch": 2.570466030320045,
+      "grad_norm": 0.17171230912208557,
+      "learning_rate": 6.098973891970373e-07,
+      "loss": 0.3185,
+      "step": 3052
+    },
+    {
+      "epoch": 2.5713082537900056,
+      "grad_norm": 0.17013221979141235,
+      "learning_rate": 6.075529979860068e-07,
+      "loss": 0.3115,
+      "step": 3053
+    },
+    {
+      "epoch": 2.5721504772599664,
+      "grad_norm": 0.18915019929409027,
+      "learning_rate": 6.052128297763804e-07,
+      "loss": 0.3331,
+      "step": 3054
+    },
+    {
+      "epoch": 2.572992700729927,
+      "grad_norm": 0.17950762808322906,
+      "learning_rate": 6.028768868180523e-07,
+      "loss": 0.3327,
+      "step": 3055
+    },
+    {
+      "epoch": 2.573834924199888,
+      "grad_norm": 0.17014242708683014,
+      "learning_rate": 6.005451713568505e-07,
+      "loss": 0.3283,
+      "step": 3056
+    },
+    {
+      "epoch": 2.5746771476698482,
+      "grad_norm": 0.1866944283246994,
+      "learning_rate": 5.982176856345445e-07,
+      "loss": 0.3299,
+      "step": 3057
+    },
+    {
+      "epoch": 2.575519371139809,
+      "grad_norm": 0.18184182047843933,
+      "learning_rate": 5.958944318888287e-07,
+      "loss": 0.3335,
+      "step": 3058
+    },
+    {
+      "epoch": 2.5763615946097698,
+      "grad_norm": 0.19136478006839752,
+      "learning_rate": 5.935754123533378e-07,
+      "loss": 0.3218,
+      "step": 3059
+    },
+    {
+      "epoch": 2.5772038180797305,
+      "grad_norm": 0.16770581901073456,
+      "learning_rate": 5.912606292576284e-07,
+      "loss": 0.3255,
+      "step": 3060
+    },
+    {
+      "epoch": 2.5780460415496913,
+      "grad_norm": 0.18040478229522705,
+      "learning_rate": 5.889500848271901e-07,
+      "loss": 0.3051,
+      "step": 3061
+    },
+    {
+      "epoch": 2.5788882650196516,
+      "grad_norm": 0.17083778977394104,
+      "learning_rate": 5.866437812834325e-07,
+      "loss": 0.3578,
+      "step": 3062
+    },
+    {
+      "epoch": 2.579730488489613,
+      "grad_norm": 0.18398484587669373,
+      "learning_rate": 5.843417208436908e-07,
+      "loss": 0.3524,
+      "step": 3063
+    },
+    {
+      "epoch": 2.580572711959573,
+      "grad_norm": 0.17296041548252106,
+      "learning_rate": 5.82043905721218e-07,
+      "loss": 0.345,
+      "step": 3064
+    },
+    {
+      "epoch": 2.581414935429534,
+      "grad_norm": 0.18131209909915924,
+      "learning_rate": 5.797503381251896e-07,
+      "loss": 0.334,
+      "step": 3065
+    },
+    {
+      "epoch": 2.5822571588994947,
+      "grad_norm": 0.17383840680122375,
+      "learning_rate": 5.774610202606939e-07,
+      "loss": 0.3408,
+      "step": 3066
+    },
+    {
+      "epoch": 2.5830993823694555,
+      "grad_norm": 0.17166505753993988,
+      "learning_rate": 5.751759543287355e-07,
+      "loss": 0.3521,
+      "step": 3067
+    },
+    {
+      "epoch": 2.5839416058394162,
+      "grad_norm": 0.17635107040405273,
+      "learning_rate": 5.728951425262292e-07,
+      "loss": 0.2984,
+      "step": 3068
+    },
+    {
+      "epoch": 2.5847838293093766,
+      "grad_norm": 0.16817434132099152,
+      "learning_rate": 5.706185870460018e-07,
+      "loss": 0.3215,
+      "step": 3069
+    },
+    {
+      "epoch": 2.5856260527793373,
+      "grad_norm": 0.1779753714799881,
+      "learning_rate": 5.683462900767873e-07,
+      "loss": 0.3313,
+      "step": 3070
+    },
+    {
+      "epoch": 2.586468276249298,
+      "grad_norm": 0.17176592350006104,
+      "learning_rate": 5.660782538032245e-07,
+      "loss": 0.3281,
+      "step": 3071
+    },
+    {
+      "epoch": 2.587310499719259,
+      "grad_norm": 0.18446749448776245,
+      "learning_rate": 5.63814480405856e-07,
+      "loss": 0.3245,
+      "step": 3072
+    },
+    {
+      "epoch": 2.5881527231892196,
+      "grad_norm": 0.17266134917736053,
+      "learning_rate": 5.61554972061128e-07,
+      "loss": 0.3175,
+      "step": 3073
+    },
+    {
+      "epoch": 2.5889949466591804,
+      "grad_norm": 0.17002305388450623,
+      "learning_rate": 5.592997309413834e-07,
+      "loss": 0.3193,
+      "step": 3074
+    },
+    {
+      "epoch": 2.589837170129141,
+      "grad_norm": 0.19935524463653564,
+      "learning_rate": 5.570487592148666e-07,
+      "loss": 0.3508,
+      "step": 3075
+    },
+    {
+      "epoch": 2.5906793935991015,
+      "grad_norm": 0.15772496163845062,
+      "learning_rate": 5.548020590457098e-07,
+      "loss": 0.3333,
+      "step": 3076
+    },
+    {
+      "epoch": 2.5915216170690623,
+      "grad_norm": 0.1764228343963623,
+      "learning_rate": 5.525596325939469e-07,
+      "loss": 0.3406,
+      "step": 3077
+    },
+    {
+      "epoch": 2.592363840539023,
+      "grad_norm": 0.18433359265327454,
+      "learning_rate": 5.503214820154979e-07,
+      "loss": 0.3054,
+      "step": 3078
+    },
+    {
+      "epoch": 2.593206064008984,
+      "grad_norm": 0.17767508327960968,
+      "learning_rate": 5.480876094621734e-07,
+      "loss": 0.319,
+      "step": 3079
+    },
+    {
+      "epoch": 2.5940482874789446,
+      "grad_norm": 0.17714276909828186,
+      "learning_rate": 5.458580170816713e-07,
+      "loss": 0.3555,
+      "step": 3080
+    },
+    {
+      "epoch": 2.594890510948905,
+      "grad_norm": 0.16562534868717194,
+      "learning_rate": 5.436327070175729e-07,
+      "loss": 0.2983,
+      "step": 3081
+    },
+    {
+      "epoch": 2.5957327344188657,
+      "grad_norm": 0.2066238671541214,
+      "learning_rate": 5.414116814093434e-07,
+      "loss": 0.3616,
+      "step": 3082
+    },
+    {
+      "epoch": 2.5965749578888264,
+      "grad_norm": 0.18519745767116547,
+      "learning_rate": 5.391949423923298e-07,
+      "loss": 0.3018,
+      "step": 3083
+    },
+    {
+      "epoch": 2.597417181358787,
+      "grad_norm": 0.18363292515277863,
+      "learning_rate": 5.369824920977567e-07,
+      "loss": 0.3116,
+      "step": 3084
+    },
+    {
+      "epoch": 2.598259404828748,
+      "grad_norm": 0.18081538379192352,
+      "learning_rate": 5.347743326527255e-07,
+      "loss": 0.3443,
+      "step": 3085
+    },
+    {
+      "epoch": 2.5991016282987087,
+      "grad_norm": 0.185759037733078,
+      "learning_rate": 5.325704661802106e-07,
+      "loss": 0.353,
+      "step": 3086
+    },
+    {
+      "epoch": 2.5999438517686695,
+      "grad_norm": 0.1803555190563202,
+      "learning_rate": 5.303708947990638e-07,
+      "loss": 0.3493,
+      "step": 3087
+    },
+    {
+      "epoch": 2.60078607523863,
+      "grad_norm": 0.1740846186876297,
+      "learning_rate": 5.281756206240035e-07,
+      "loss": 0.2993,
+      "step": 3088
+    },
+    {
+      "epoch": 2.6016282987085906,
+      "grad_norm": 0.17752055823802948,
+      "learning_rate": 5.25984645765617e-07,
+      "loss": 0.347,
+      "step": 3089
+    },
+    {
+      "epoch": 2.6024705221785513,
+      "grad_norm": 0.17252019047737122,
+      "learning_rate": 5.237979723303582e-07,
+      "loss": 0.3308,
+      "step": 3090
+    },
+    {
+      "epoch": 2.603312745648512,
+      "grad_norm": 0.1650722175836563,
+      "learning_rate": 5.216156024205482e-07,
+      "loss": 0.3362,
+      "step": 3091
+    },
+    {
+      "epoch": 2.604154969118473,
+      "grad_norm": 0.16832077503204346,
+      "learning_rate": 5.194375381343664e-07,
+      "loss": 0.3179,
+      "step": 3092
+    },
+    {
+      "epoch": 2.604997192588433,
+      "grad_norm": 0.1857997328042984,
+      "learning_rate": 5.172637815658583e-07,
+      "loss": 0.3459,
+      "step": 3093
+    },
+    {
+      "epoch": 2.6058394160583944,
+      "grad_norm": 0.17449375987052917,
+      "learning_rate": 5.150943348049198e-07,
+      "loss": 0.3144,
+      "step": 3094
+    },
+    {
+      "epoch": 2.6066816395283547,
+      "grad_norm": 0.19014085829257965,
+      "learning_rate": 5.129291999373109e-07,
+      "loss": 0.3837,
+      "step": 3095
+    },
+    {
+      "epoch": 2.6075238629983155,
+      "grad_norm": 0.1807759404182434,
+      "learning_rate": 5.107683790446411e-07,
+      "loss": 0.3193,
+      "step": 3096
+    },
+    {
+      "epoch": 2.6083660864682763,
+      "grad_norm": 0.17734181880950928,
+      "learning_rate": 5.086118742043761e-07,
+      "loss": 0.3575,
+      "step": 3097
+    },
+    {
+      "epoch": 2.609208309938237,
+      "grad_norm": 0.16949152946472168,
+      "learning_rate": 5.064596874898292e-07,
+      "loss": 0.293,
+      "step": 3098
+    },
+    {
+      "epoch": 2.610050533408198,
+      "grad_norm": 0.17707063257694244,
+      "learning_rate": 5.04311820970163e-07,
+      "loss": 0.3613,
+      "step": 3099
+    },
+    {
+      "epoch": 2.610892756878158,
+      "grad_norm": 0.18580542504787445,
+      "learning_rate": 5.021682767103858e-07,
+      "loss": 0.3379,
+      "step": 3100
+    },
+    {
+      "epoch": 2.611734980348119,
+      "grad_norm": 0.18288764357566833,
+      "learning_rate": 5.000290567713533e-07,
+      "loss": 0.3342,
+      "step": 3101
+    },
+    {
+      "epoch": 2.6125772038180797,
+      "grad_norm": 0.19285543262958527,
+      "learning_rate": 4.978941632097612e-07,
+      "loss": 0.3157,
+      "step": 3102
+    },
+    {
+      "epoch": 2.6134194272880404,
+      "grad_norm": 0.17595289647579193,
+      "learning_rate": 4.957635980781445e-07,
+      "loss": 0.3196,
+      "step": 3103
+    },
+    {
+      "epoch": 2.614261650758001,
+      "grad_norm": 0.18875975906848907,
+      "learning_rate": 4.936373634248792e-07,
+      "loss": 0.3137,
+      "step": 3104
+    },
+    {
+      "epoch": 2.615103874227962,
+      "grad_norm": 0.17160099744796753,
+      "learning_rate": 4.915154612941781e-07,
+      "loss": 0.312,
+      "step": 3105
+    },
+    {
+      "epoch": 2.6159460976979227,
+      "grad_norm": 0.1704033613204956,
+      "learning_rate": 4.893978937260868e-07,
+      "loss": 0.3431,
+      "step": 3106
+    },
+    {
+      "epoch": 2.616788321167883,
+      "grad_norm": 0.18089723587036133,
+      "learning_rate": 4.872846627564842e-07,
+      "loss": 0.3543,
+      "step": 3107
+    },
+    {
+      "epoch": 2.617630544637844,
+      "grad_norm": 0.1861312836408615,
+      "learning_rate": 4.851757704170796e-07,
+      "loss": 0.3494,
+      "step": 3108
+    },
+    {
+      "epoch": 2.6184727681078046,
+      "grad_norm": 0.19477856159210205,
+      "learning_rate": 4.830712187354125e-07,
+      "loss": 0.3508,
+      "step": 3109
+    },
+    {
+      "epoch": 2.6193149915777654,
+      "grad_norm": 0.1828373372554779,
+      "learning_rate": 4.809710097348469e-07,
+      "loss": 0.2927,
+      "step": 3110
+    },
+    {
+      "epoch": 2.620157215047726,
+      "grad_norm": 0.19689728319644928,
+      "learning_rate": 4.788751454345763e-07,
+      "loss": 0.3149,
+      "step": 3111
+    },
+    {
+      "epoch": 2.6209994385176865,
+      "grad_norm": 0.19670797884464264,
+      "learning_rate": 4.767836278496085e-07,
+      "loss": 0.3386,
+      "step": 3112
+    },
+    {
+      "epoch": 2.6218416619876472,
+      "grad_norm": 0.16965490579605103,
+      "learning_rate": 4.7469645899078153e-07,
+      "loss": 0.3201,
+      "step": 3113
+    },
+    {
+      "epoch": 2.622683885457608,
+      "grad_norm": 0.1843360811471939,
+      "learning_rate": 4.726136408647464e-07,
+      "loss": 0.332,
+      "step": 3114
+    },
+    {
+      "epoch": 2.6235261089275688,
+      "grad_norm": 0.17019793391227722,
+      "learning_rate": 4.7053517547397454e-07,
+      "loss": 0.3233,
+      "step": 3115
+    },
+    {
+      "epoch": 2.6243683323975295,
+      "grad_norm": 0.16619959473609924,
+      "learning_rate": 4.6846106481675035e-07,
+      "loss": 0.3347,
+      "step": 3116
+    },
+    {
+      "epoch": 2.6252105558674903,
+      "grad_norm": 0.1637767255306244,
+      "learning_rate": 4.663913108871726e-07,
+      "loss": 0.3206,
+      "step": 3117
+    },
+    {
+      "epoch": 2.626052779337451,
+      "grad_norm": 0.17986620962619781,
+      "learning_rate": 4.643259156751506e-07,
+      "loss": 0.3438,
+      "step": 3118
+    },
+    {
+      "epoch": 2.6268950028074114,
+      "grad_norm": 0.18076659739017487,
+      "learning_rate": 4.622648811664049e-07,
+      "loss": 0.3557,
+      "step": 3119
+    },
+    {
+      "epoch": 2.627737226277372,
+      "grad_norm": 0.16573551297187805,
+      "learning_rate": 4.60208209342462e-07,
+      "loss": 0.3173,
+      "step": 3120
+    },
+    {
+      "epoch": 2.628579449747333,
+      "grad_norm": 0.19350749254226685,
+      "learning_rate": 4.581559021806542e-07,
+      "loss": 0.3453,
+      "step": 3121
+    },
+    {
+      "epoch": 2.6294216732172937,
+      "grad_norm": 0.18189962208271027,
+      "learning_rate": 4.561079616541164e-07,
+      "loss": 0.3085,
+      "step": 3122
+    },
+    {
+      "epoch": 2.6302638966872545,
+      "grad_norm": 0.18365921080112457,
+      "learning_rate": 4.540643897317887e-07,
+      "loss": 0.3258,
+      "step": 3123
+    },
+    {
+      "epoch": 2.631106120157215,
+      "grad_norm": 0.16131576895713806,
+      "learning_rate": 4.520251883784077e-07,
+      "loss": 0.3052,
+      "step": 3124
+    },
+    {
+      "epoch": 2.631948343627176,
+      "grad_norm": 0.1853943169116974,
+      "learning_rate": 4.4999035955450964e-07,
+      "loss": 0.3786,
+      "step": 3125
+    },
+    {
+      "epoch": 2.6327905670971363,
+      "grad_norm": 0.1730596274137497,
+      "learning_rate": 4.4795990521642684e-07,
+      "loss": 0.322,
+      "step": 3126
+    },
+    {
+      "epoch": 2.633632790567097,
+      "grad_norm": 0.18700529634952545,
+      "learning_rate": 4.459338273162844e-07,
+      "loss": 0.3538,
+      "step": 3127
+    },
+    {
+      "epoch": 2.634475014037058,
+      "grad_norm": 0.17436066269874573,
+      "learning_rate": 4.439121278020031e-07,
+      "loss": 0.3491,
+      "step": 3128
+    },
+    {
+      "epoch": 2.6353172375070186,
+      "grad_norm": 0.1638750582933426,
+      "learning_rate": 4.4189480861729137e-07,
+      "loss": 0.3038,
+      "step": 3129
+    },
+    {
+      "epoch": 2.6361594609769794,
+      "grad_norm": 0.16320183873176575,
+      "learning_rate": 4.3988187170164673e-07,
+      "loss": 0.3137,
+      "step": 3130
+    },
+    {
+      "epoch": 2.6370016844469397,
+      "grad_norm": 0.16656161844730377,
+      "learning_rate": 4.378733189903528e-07,
+      "loss": 0.3385,
+      "step": 3131
+    },
+    {
+      "epoch": 2.6378439079169005,
+      "grad_norm": 0.1839463710784912,
+      "learning_rate": 4.35869152414482e-07,
+      "loss": 0.3445,
+      "step": 3132
+    },
+    {
+      "epoch": 2.6386861313868613,
+      "grad_norm": 0.18246237933635712,
+      "learning_rate": 4.3386937390088366e-07,
+      "loss": 0.3685,
+      "step": 3133
+    },
+    {
+      "epoch": 2.639528354856822,
+      "grad_norm": 0.1700492799282074,
+      "learning_rate": 4.3187398537219593e-07,
+      "loss": 0.3194,
+      "step": 3134
+    },
+    {
+      "epoch": 2.640370578326783,
+      "grad_norm": 0.17533659934997559,
+      "learning_rate": 4.2988298874682754e-07,
+      "loss": 0.3096,
+      "step": 3135
+    },
+    {
+      "epoch": 2.6412128017967436,
+      "grad_norm": 0.17206905782222748,
+      "learning_rate": 4.278963859389723e-07,
+      "loss": 0.3508,
+      "step": 3136
+    },
+    {
+      "epoch": 2.6420550252667043,
+      "grad_norm": 0.20786888897418976,
+      "learning_rate": 4.259141788585947e-07,
+      "loss": 0.3397,
+      "step": 3137
+    },
+    {
+      "epoch": 2.6428972487366647,
+      "grad_norm": 0.1743757575750351,
+      "learning_rate": 4.239363694114368e-07,
+      "loss": 0.3344,
+      "step": 3138
+    },
+    {
+      "epoch": 2.6437394722066254,
+      "grad_norm": 0.16983045637607574,
+      "learning_rate": 4.2196295949901044e-07,
+      "loss": 0.3174,
+      "step": 3139
+    },
+    {
+      "epoch": 2.644581695676586,
+      "grad_norm": 0.1640872210264206,
+      "learning_rate": 4.1999395101859796e-07,
+      "loss": 0.3219,
+      "step": 3140
+    },
+    {
+      "epoch": 2.645423919146547,
+      "grad_norm": 0.17452681064605713,
+      "learning_rate": 4.1802934586324897e-07,
+      "loss": 0.3523,
+      "step": 3141
+    },
+    {
+      "epoch": 2.6462661426165077,
+      "grad_norm": 0.18185783922672272,
+      "learning_rate": 4.160691459217825e-07,
+      "loss": 0.3166,
+      "step": 3142
+    },
+    {
+      "epoch": 2.647108366086468,
+      "grad_norm": 0.1847081035375595,
+      "learning_rate": 4.1411335307878056e-07,
+      "loss": 0.3256,
+      "step": 3143
+    },
+    {
+      "epoch": 2.647950589556429,
+      "grad_norm": 0.17259185016155243,
+      "learning_rate": 4.1216196921458786e-07,
+      "loss": 0.333,
+      "step": 3144
+    },
+    {
+      "epoch": 2.6487928130263896,
+      "grad_norm": 0.18488125503063202,
+      "learning_rate": 4.102149962053098e-07,
+      "loss": 0.3293,
+      "step": 3145
+    },
+    {
+      "epoch": 2.6496350364963503,
+      "grad_norm": 0.16416329145431519,
+      "learning_rate": 4.0827243592281294e-07,
+      "loss": 0.3229,
+      "step": 3146
+    },
+    {
+      "epoch": 2.650477259966311,
+      "grad_norm": 0.17134909331798553,
+      "learning_rate": 4.0633429023472004e-07,
+      "loss": 0.3599,
+      "step": 3147
+    },
+    {
+      "epoch": 2.651319483436272,
+      "grad_norm": 0.16451704502105713,
+      "learning_rate": 4.044005610044094e-07,
+      "loss": 0.3238,
+      "step": 3148
+    },
+    {
+      "epoch": 2.6521617069062327,
+      "grad_norm": 0.17452752590179443,
+      "learning_rate": 4.0247125009101275e-07,
+      "loss": 0.3491,
+      "step": 3149
+    },
+    {
+      "epoch": 2.653003930376193,
+      "grad_norm": 0.19579990208148956,
+      "learning_rate": 4.0054635934941633e-07,
+      "loss": 0.3426,
+      "step": 3150
+    },
+    {
+      "epoch": 2.6538461538461537,
+      "grad_norm": 0.16347888112068176,
+      "learning_rate": 3.986258906302543e-07,
+      "loss": 0.3198,
+      "step": 3151
+    },
+    {
+      "epoch": 2.6546883773161145,
+      "grad_norm": 0.17911840975284576,
+      "learning_rate": 3.967098457799118e-07,
+      "loss": 0.3578,
+      "step": 3152
+    },
+    {
+      "epoch": 2.6555306007860753,
+      "grad_norm": 0.16491587460041046,
+      "learning_rate": 3.947982266405159e-07,
+      "loss": 0.3344,
+      "step": 3153
+    },
+    {
+      "epoch": 2.656372824256036,
+      "grad_norm": 0.17907950282096863,
+      "learning_rate": 3.928910350499454e-07,
+      "loss": 0.345,
+      "step": 3154
+    },
+    {
+      "epoch": 2.6572150477259964,
+      "grad_norm": 0.16243170201778412,
+      "learning_rate": 3.9098827284181683e-07,
+      "loss": 0.3067,
+      "step": 3155
+    },
+    {
+      "epoch": 2.6580572711959576,
+      "grad_norm": 0.19042733311653137,
+      "learning_rate": 3.890899418454913e-07,
+      "loss": 0.333,
+      "step": 3156
+    },
+    {
+      "epoch": 2.658899494665918,
+      "grad_norm": 0.18333835899829865,
+      "learning_rate": 3.871960438860689e-07,
+      "loss": 0.3522,
+      "step": 3157
+    },
+    {
+      "epoch": 2.6597417181358787,
+      "grad_norm": 0.17561234533786774,
+      "learning_rate": 3.8530658078438754e-07,
+      "loss": 0.3267,
+      "step": 3158
+    },
+    {
+      "epoch": 2.6605839416058394,
+      "grad_norm": 0.17840707302093506,
+      "learning_rate": 3.834215543570191e-07,
+      "loss": 0.33,
+      "step": 3159
+    },
+    {
+      "epoch": 2.6614261650758,
+      "grad_norm": 0.17847512662410736,
+      "learning_rate": 3.81540966416275e-07,
+      "loss": 0.3327,
+      "step": 3160
+    },
+    {
+      "epoch": 2.662268388545761,
+      "grad_norm": 0.16171397268772125,
+      "learning_rate": 3.796648187701957e-07,
+      "loss": 0.319,
+      "step": 3161
+    },
+    {
+      "epoch": 2.6631106120157213,
+      "grad_norm": 0.1658945530653,
+      "learning_rate": 3.777931132225526e-07,
+      "loss": 0.3392,
+      "step": 3162
+    },
+    {
+      "epoch": 2.663952835485682,
+      "grad_norm": 0.20843006670475006,
+      "learning_rate": 3.75925851572847e-07,
+      "loss": 0.3263,
+      "step": 3163
+    },
+    {
+      "epoch": 2.664795058955643,
+      "grad_norm": 0.17359398305416107,
+      "learning_rate": 3.7406303561630996e-07,
+      "loss": 0.3257,
+      "step": 3164
+    },
+    {
+      "epoch": 2.6656372824256036,
+      "grad_norm": 0.17524690926074982,
+      "learning_rate": 3.72204667143895e-07,
+      "loss": 0.3084,
+      "step": 3165
+    },
+    {
+      "epoch": 2.6664795058955644,
+      "grad_norm": 0.1768205612897873,
+      "learning_rate": 3.703507479422813e-07,
+      "loss": 0.3398,
+      "step": 3166
+    },
+    {
+      "epoch": 2.667321729365525,
+      "grad_norm": 0.18176667392253876,
+      "learning_rate": 3.6850127979386917e-07,
+      "loss": 0.3399,
+      "step": 3167
+    },
+    {
+      "epoch": 2.668163952835486,
+      "grad_norm": 0.18099865317344666,
+      "learning_rate": 3.666562644767824e-07,
+      "loss": 0.3444,
+      "step": 3168
+    },
+    {
+      "epoch": 2.6690061763054462,
+      "grad_norm": 0.1720365285873413,
+      "learning_rate": 3.648157037648598e-07,
+      "loss": 0.3564,
+      "step": 3169
+    },
+    {
+      "epoch": 2.669848399775407,
+      "grad_norm": 0.18111947178840637,
+      "learning_rate": 3.6297959942766303e-07,
+      "loss": 0.3414,
+      "step": 3170
+    },
+    {
+      "epoch": 2.6706906232453678,
+      "grad_norm": 0.17587114870548248,
+      "learning_rate": 3.611479532304618e-07,
+      "loss": 0.2977,
+      "step": 3171
+    },
+    {
+      "epoch": 2.6715328467153285,
+      "grad_norm": 0.16626040637493134,
+      "learning_rate": 3.593207669342463e-07,
+      "loss": 0.3231,
+      "step": 3172
+    },
+    {
+      "epoch": 2.6723750701852893,
+      "grad_norm": 0.6724367737770081,
+      "learning_rate": 3.574980422957147e-07,
+      "loss": 0.3574,
+      "step": 3173
+    },
+    {
+      "epoch": 2.6732172936552496,
+      "grad_norm": 0.17822889983654022,
+      "learning_rate": 3.556797810672785e-07,
+      "loss": 0.3476,
+      "step": 3174
+    },
+    {
+      "epoch": 2.6740595171252104,
+      "grad_norm": 0.17662286758422852,
+      "learning_rate": 3.538659849970555e-07,
+      "loss": 0.3207,
+      "step": 3175
+    },
+    {
+      "epoch": 2.674901740595171,
+      "grad_norm": 0.17599961161613464,
+      "learning_rate": 3.5205665582887296e-07,
+      "loss": 0.3628,
+      "step": 3176
+    },
+    {
+      "epoch": 2.675743964065132,
+      "grad_norm": 0.17218033969402313,
+      "learning_rate": 3.5025179530225995e-07,
+      "loss": 0.3065,
+      "step": 3177
+    },
+    {
+      "epoch": 2.6765861875350927,
+      "grad_norm": 0.17550210654735565,
+      "learning_rate": 3.484514051524546e-07,
+      "loss": 0.3106,
+      "step": 3178
+    },
+    {
+      "epoch": 2.6774284110050535,
+      "grad_norm": 0.1761777400970459,
+      "learning_rate": 3.466554871103922e-07,
+      "loss": 0.3175,
+      "step": 3179
+    },
+    {
+      "epoch": 2.6782706344750142,
+      "grad_norm": 0.17350716888904572,
+      "learning_rate": 3.4486404290271115e-07,
+      "loss": 0.3517,
+      "step": 3180
+    },
+    {
+      "epoch": 2.6791128579449746,
+      "grad_norm": 0.16766268014907837,
+      "learning_rate": 3.43077074251747e-07,
+      "loss": 0.3212,
+      "step": 3181
+    },
+    {
+      "epoch": 2.6799550814149353,
+      "grad_norm": 0.19454942643642426,
+      "learning_rate": 3.4129458287553487e-07,
+      "loss": 0.3421,
+      "step": 3182
+    },
+    {
+      "epoch": 2.680797304884896,
+      "grad_norm": 0.1776033639907837,
+      "learning_rate": 3.395165704878023e-07,
+      "loss": 0.3429,
+      "step": 3183
+    },
+    {
+      "epoch": 2.681639528354857,
+      "grad_norm": 0.1730322688817978,
+      "learning_rate": 3.3774303879797297e-07,
+      "loss": 0.3139,
+      "step": 3184
+    },
+    {
+      "epoch": 2.6824817518248176,
+      "grad_norm": 0.16911497712135315,
+      "learning_rate": 3.359739895111602e-07,
+      "loss": 0.3132,
+      "step": 3185
+    },
+    {
+      "epoch": 2.683323975294778,
+      "grad_norm": 0.17976872622966766,
+      "learning_rate": 3.3420942432817127e-07,
+      "loss": 0.3274,
+      "step": 3186
+    },
+    {
+      "epoch": 2.684166198764739,
+      "grad_norm": 0.19177532196044922,
+      "learning_rate": 3.324493449454991e-07,
+      "loss": 0.3444,
+      "step": 3187
+    },
+    {
+      "epoch": 2.6850084222346995,
+      "grad_norm": 0.17109844088554382,
+      "learning_rate": 3.3069375305532725e-07,
+      "loss": 0.3241,
+      "step": 3188
+    },
+    {
+      "epoch": 2.6858506457046603,
+      "grad_norm": 0.18712511658668518,
+      "learning_rate": 3.289426503455201e-07,
+      "loss": 0.3379,
+      "step": 3189
+    },
+    {
+      "epoch": 2.686692869174621,
+      "grad_norm": 0.1836530715227127,
+      "learning_rate": 3.271960384996309e-07,
+      "loss": 0.3118,
+      "step": 3190
+    },
+    {
+      "epoch": 2.687535092644582,
+      "grad_norm": 0.18208006024360657,
+      "learning_rate": 3.2545391919689193e-07,
+      "loss": 0.3328,
+      "step": 3191
+    },
+    {
+      "epoch": 2.6883773161145426,
+      "grad_norm": 0.18471787869930267,
+      "learning_rate": 3.237162941122185e-07,
+      "loss": 0.336,
+      "step": 3192
+    },
+    {
+      "epoch": 2.689219539584503,
+      "grad_norm": 0.16804207861423492,
+      "learning_rate": 3.2198316491620305e-07,
+      "loss": 0.3044,
+      "step": 3193
+    },
+    {
+      "epoch": 2.6900617630544637,
+      "grad_norm": 0.18219810724258423,
+      "learning_rate": 3.202545332751178e-07,
+      "loss": 0.3453,
+      "step": 3194
+    },
+    {
+      "epoch": 2.6909039865244244,
+      "grad_norm": 0.17537109553813934,
+      "learning_rate": 3.185304008509077e-07,
+      "loss": 0.3089,
+      "step": 3195
+    },
+    {
+      "epoch": 2.691746209994385,
+      "grad_norm": 0.1631346195936203,
+      "learning_rate": 3.1681076930119626e-07,
+      "loss": 0.3375,
+      "step": 3196
+    },
+    {
+      "epoch": 2.692588433464346,
+      "grad_norm": 0.18209043145179749,
+      "learning_rate": 3.150956402792765e-07,
+      "loss": 0.3504,
+      "step": 3197
+    },
+    {
+      "epoch": 2.6934306569343067,
+      "grad_norm": 0.1706717163324356,
+      "learning_rate": 3.133850154341139e-07,
+      "loss": 0.3147,
+      "step": 3198
+    },
+    {
+      "epoch": 2.6942728804042675,
+      "grad_norm": 0.16930033266544342,
+      "learning_rate": 3.116788964103429e-07,
+      "loss": 0.3296,
+      "step": 3199
+    },
+    {
+      "epoch": 2.695115103874228,
+      "grad_norm": 0.17827804386615753,
+      "learning_rate": 3.099772848482657e-07,
+      "loss": 0.3239,
+      "step": 3200
+    },
+    {
+      "epoch": 2.6959573273441886,
+      "grad_norm": 0.1688344031572342,
+      "learning_rate": 3.082801823838527e-07,
+      "loss": 0.3129,
+      "step": 3201
+    },
+    {
+      "epoch": 2.6967995508141493,
+      "grad_norm": 0.18355657160282135,
+      "learning_rate": 3.0658759064873755e-07,
+      "loss": 0.3664,
+      "step": 3202
+    },
+    {
+      "epoch": 2.69764177428411,
+      "grad_norm": 0.17502044141292572,
+      "learning_rate": 3.0489951127021744e-07,
+      "loss": 0.3255,
+      "step": 3203
+    },
+    {
+      "epoch": 2.698483997754071,
+      "grad_norm": 0.17082899808883667,
+      "learning_rate": 3.0321594587125083e-07,
+      "loss": 0.2905,
+      "step": 3204
+    },
+    {
+      "epoch": 2.699326221224031,
+      "grad_norm": 0.15576323866844177,
+      "learning_rate": 3.015368960704584e-07,
+      "loss": 0.329,
+      "step": 3205
+    },
+    {
+      "epoch": 2.700168444693992,
+      "grad_norm": 0.17618028819561005,
+      "learning_rate": 2.9986236348211684e-07,
+      "loss": 0.351,
+      "step": 3206
+    },
+    {
+      "epoch": 2.7010106681639527,
+      "grad_norm": 0.15897849202156067,
+      "learning_rate": 2.9819234971616154e-07,
+      "loss": 0.3138,
+      "step": 3207
+    },
+    {
+      "epoch": 2.7018528916339135,
+      "grad_norm": 0.17301517724990845,
+      "learning_rate": 2.9652685637818147e-07,
+      "loss": 0.3238,
+      "step": 3208
+    },
+    {
+      "epoch": 2.7026951151038743,
+      "grad_norm": 0.18447841703891754,
+      "learning_rate": 2.9486588506942303e-07,
+      "loss": 0.3684,
+      "step": 3209
+    },
+    {
+      "epoch": 2.703537338573835,
+      "grad_norm": 0.19456276297569275,
+      "learning_rate": 2.932094373867811e-07,
+      "loss": 0.3428,
+      "step": 3210
+    },
+    {
+      "epoch": 2.704379562043796,
+      "grad_norm": 0.1816757172346115,
+      "learning_rate": 2.915575149228056e-07,
+      "loss": 0.3259,
+      "step": 3211
+    },
+    {
+      "epoch": 2.705221785513756,
+      "grad_norm": 0.20598746836185455,
+      "learning_rate": 2.8991011926569003e-07,
+      "loss": 0.3233,
+      "step": 3212
+    },
+    {
+      "epoch": 2.706064008983717,
+      "grad_norm": 0.16217656433582306,
+      "learning_rate": 2.882672519992824e-07,
+      "loss": 0.3118,
+      "step": 3213
+    },
+    {
+      "epoch": 2.7069062324536777,
+      "grad_norm": 0.17557232081890106,
+      "learning_rate": 2.8662891470307154e-07,
+      "loss": 0.3467,
+      "step": 3214
+    },
+    {
+      "epoch": 2.7077484559236384,
+      "grad_norm": 0.17025545239448547,
+      "learning_rate": 2.8499510895219464e-07,
+      "loss": 0.308,
+      "step": 3215
+    },
+    {
+      "epoch": 2.708590679393599,
+      "grad_norm": 0.16908609867095947,
+      "learning_rate": 2.833658363174302e-07,
+      "loss": 0.3198,
+      "step": 3216
+    },
+    {
+      "epoch": 2.7094329028635595,
+      "grad_norm": 0.16965502500534058,
+      "learning_rate": 2.817410983651997e-07,
+      "loss": 0.3376,
+      "step": 3217
+    },
+    {
+      "epoch": 2.7102751263335207,
+      "grad_norm": 0.17812147736549377,
+      "learning_rate": 2.80120896657563e-07,
+      "loss": 0.3437,
+      "step": 3218
+    },
+    {
+      "epoch": 2.711117349803481,
+      "grad_norm": 0.17128512263298035,
+      "learning_rate": 2.785052327522214e-07,
+      "loss": 0.3224,
+      "step": 3219
+    },
+    {
+      "epoch": 2.711959573273442,
+      "grad_norm": 0.1748804748058319,
+      "learning_rate": 2.768941082025112e-07,
+      "loss": 0.34,
+      "step": 3220
+    },
+    {
+      "epoch": 2.7128017967434026,
+      "grad_norm": 0.17742173373699188,
+      "learning_rate": 2.7528752455740606e-07,
+      "loss": 0.3349,
+      "step": 3221
+    },
+    {
+      "epoch": 2.7136440202133634,
+      "grad_norm": 0.1696932166814804,
+      "learning_rate": 2.73685483361511e-07,
+      "loss": 0.3446,
+      "step": 3222
+    },
+    {
+      "epoch": 2.714486243683324,
+      "grad_norm": 0.1748046725988388,
+      "learning_rate": 2.720879861550685e-07,
+      "loss": 0.3052,
+      "step": 3223
+    },
+    {
+      "epoch": 2.7153284671532845,
+      "grad_norm": 0.1804862767457962,
+      "learning_rate": 2.7049503447394874e-07,
+      "loss": 0.3297,
+      "step": 3224
+    },
+    {
+      "epoch": 2.7161706906232452,
+      "grad_norm": 0.18347153067588806,
+      "learning_rate": 2.6890662984965234e-07,
+      "loss": 0.3496,
+      "step": 3225
+    },
+    {
+      "epoch": 2.717012914093206,
+      "grad_norm": 0.1759885549545288,
+      "learning_rate": 2.6732277380930873e-07,
+      "loss": 0.3362,
+      "step": 3226
+    },
+    {
+      "epoch": 2.7178551375631668,
+      "grad_norm": 0.16716201603412628,
+      "learning_rate": 2.657434678756754e-07,
+      "loss": 0.2997,
+      "step": 3227
+    },
+    {
+      "epoch": 2.7186973610331275,
+      "grad_norm": 0.18232782185077667,
+      "learning_rate": 2.6416871356713224e-07,
+      "loss": 0.352,
+      "step": 3228
+    },
+    {
+      "epoch": 2.7195395845030883,
+      "grad_norm": 0.185092493891716,
+      "learning_rate": 2.625985123976876e-07,
+      "loss": 0.3423,
+      "step": 3229
+    },
+    {
+      "epoch": 2.720381807973049,
+      "grad_norm": 0.4390159249305725,
+      "learning_rate": 2.6103286587696674e-07,
+      "loss": 0.3557,
+      "step": 3230
+    },
+    {
+      "epoch": 2.7212240314430094,
+      "grad_norm": 0.16824904084205627,
+      "learning_rate": 2.594717755102205e-07,
+      "loss": 0.3386,
+      "step": 3231
+    },
+    {
+      "epoch": 2.72206625491297,
+      "grad_norm": 0.17320628464221954,
+      "learning_rate": 2.5791524279831613e-07,
+      "loss": 0.3302,
+      "step": 3232
+    },
+    {
+      "epoch": 2.722908478382931,
+      "grad_norm": 0.18359090387821198,
+      "learning_rate": 2.5636326923774325e-07,
+      "loss": 0.3163,
+      "step": 3233
+    },
+    {
+      "epoch": 2.7237507018528917,
+      "grad_norm": 0.17598529160022736,
+      "learning_rate": 2.548158563206038e-07,
+      "loss": 0.2999,
+      "step": 3234
+    },
+    {
+      "epoch": 2.7245929253228525,
+      "grad_norm": 0.17214341461658478,
+      "learning_rate": 2.532730055346172e-07,
+      "loss": 0.3218,
+      "step": 3235
+    },
+    {
+      "epoch": 2.725435148792813,
+      "grad_norm": 0.17373663187026978,
+      "learning_rate": 2.517347183631158e-07,
+      "loss": 0.3266,
+      "step": 3236
+    },
+    {
+      "epoch": 2.7262773722627736,
+      "grad_norm": 0.17507827281951904,
+      "learning_rate": 2.5020099628504603e-07,
+      "loss": 0.3468,
+      "step": 3237
+    },
+    {
+      "epoch": 2.7271195957327343,
+      "grad_norm": 0.1673845499753952,
+      "learning_rate": 2.4867184077496333e-07,
+      "loss": 0.3034,
+      "step": 3238
+    },
+    {
+      "epoch": 2.727961819202695,
+      "grad_norm": 0.16584512591362,
+      "learning_rate": 2.471472533030339e-07,
+      "loss": 0.3437,
+      "step": 3239
+    },
+    {
+      "epoch": 2.728804042672656,
+      "grad_norm": 0.16631992161273956,
+      "learning_rate": 2.4562723533503084e-07,
+      "loss": 0.2933,
+      "step": 3240
+    },
+    {
+      "epoch": 2.7296462661426166,
+      "grad_norm": 0.1780492663383484,
+      "learning_rate": 2.441117883323374e-07,
+      "loss": 0.345,
+      "step": 3241
+    },
+    {
+      "epoch": 2.7304884896125774,
+      "grad_norm": 0.17436213791370392,
+      "learning_rate": 2.426009137519375e-07,
+      "loss": 0.3417,
+      "step": 3242
+    },
+    {
+      "epoch": 2.7313307130825377,
+      "grad_norm": 0.16060450673103333,
+      "learning_rate": 2.4109461304642254e-07,
+      "loss": 0.3304,
+      "step": 3243
+    },
+    {
+      "epoch": 2.7321729365524985,
+      "grad_norm": 0.16995938122272491,
+      "learning_rate": 2.395928876639847e-07,
+      "loss": 0.3158,
+      "step": 3244
+    },
+    {
+      "epoch": 2.7330151600224593,
+      "grad_norm": 0.1702856421470642,
+      "learning_rate": 2.3809573904841844e-07,
+      "loss": 0.3076,
+      "step": 3245
+    },
+    {
+      "epoch": 2.73385738349242,
+      "grad_norm": 0.1806149184703827,
+      "learning_rate": 2.3660316863911682e-07,
+      "loss": 0.3237,
+      "step": 3246
+    },
+    {
+      "epoch": 2.734699606962381,
+      "grad_norm": 0.18080653250217438,
+      "learning_rate": 2.3511517787107363e-07,
+      "loss": 0.3547,
+      "step": 3247
+    },
+    {
+      "epoch": 2.735541830432341,
+      "grad_norm": 0.1690075546503067,
+      "learning_rate": 2.336317681748751e-07,
+      "loss": 0.3174,
+      "step": 3248
+    },
+    {
+      "epoch": 2.7363840539023023,
+      "grad_norm": 0.16987837851047516,
+      "learning_rate": 2.3215294097670927e-07,
+      "loss": 0.3013,
+      "step": 3249
+    },
+    {
+      "epoch": 2.7372262773722627,
+      "grad_norm": 0.16689546406269073,
+      "learning_rate": 2.3067869769835215e-07,
+      "loss": 0.3377,
+      "step": 3250
+    },
+    {
+      "epoch": 2.7380685008422234,
+      "grad_norm": 0.18353791534900665,
+      "learning_rate": 2.292090397571789e-07,
+      "loss": 0.3565,
+      "step": 3251
+    },
+    {
+      "epoch": 2.738910724312184,
+      "grad_norm": 0.17699800431728363,
+      "learning_rate": 2.277439685661509e-07,
+      "loss": 0.2995,
+      "step": 3252
+    },
+    {
+      "epoch": 2.739752947782145,
+      "grad_norm": 0.1799086481332779,
+      "learning_rate": 2.262834855338225e-07,
+      "loss": 0.3096,
+      "step": 3253
+    },
+    {
+      "epoch": 2.7405951712521057,
+      "grad_norm": 0.18038147687911987,
+      "learning_rate": 2.2482759206433613e-07,
+      "loss": 0.34,
+      "step": 3254
+    },
+    {
+      "epoch": 2.741437394722066,
+      "grad_norm": 0.18209418654441833,
+      "learning_rate": 2.2337628955742263e-07,
+      "loss": 0.3362,
+      "step": 3255
+    },
+    {
+      "epoch": 2.742279618192027,
+      "grad_norm": 0.175188347697258,
+      "learning_rate": 2.21929579408397e-07,
+      "loss": 0.3516,
+      "step": 3256
+    },
+    {
+      "epoch": 2.7431218416619876,
+      "grad_norm": 0.1681680530309677,
+      "learning_rate": 2.204874630081616e-07,
+      "loss": 0.3223,
+      "step": 3257
+    },
+    {
+      "epoch": 2.7439640651319483,
+      "grad_norm": 0.1675475686788559,
+      "learning_rate": 2.1904994174319903e-07,
+      "loss": 0.3522,
+      "step": 3258
+    },
+    {
+      "epoch": 2.744806288601909,
+      "grad_norm": 0.17860925197601318,
+      "learning_rate": 2.1761701699557824e-07,
+      "loss": 0.3132,
+      "step": 3259
+    },
+    {
+      "epoch": 2.74564851207187,
+      "grad_norm": 0.1875511258840561,
+      "learning_rate": 2.1618869014294498e-07,
+      "loss": 0.3268,
+      "step": 3260
+    },
+    {
+      "epoch": 2.7464907355418307,
+      "grad_norm": 0.1863333284854889,
+      "learning_rate": 2.1476496255852685e-07,
+      "loss": 0.3233,
+      "step": 3261
+    },
+    {
+      "epoch": 2.747332959011791,
+      "grad_norm": 0.1790715754032135,
+      "learning_rate": 2.1334583561112786e-07,
+      "loss": 0.3351,
+      "step": 3262
+    },
+    {
+      "epoch": 2.7481751824817517,
+      "grad_norm": 0.16683176159858704,
+      "learning_rate": 2.1193131066513107e-07,
+      "loss": 0.3015,
+      "step": 3263
+    },
+    {
+      "epoch": 2.7490174059517125,
+      "grad_norm": 0.16594216227531433,
+      "learning_rate": 2.1052138908049303e-07,
+      "loss": 0.3334,
+      "step": 3264
+    },
+    {
+      "epoch": 2.7498596294216733,
+      "grad_norm": 0.15618757903575897,
+      "learning_rate": 2.091160722127472e-07,
+      "loss": 0.303,
+      "step": 3265
+    },
+    {
+      "epoch": 2.750701852891634,
+      "grad_norm": 0.16715048253536224,
+      "learning_rate": 2.0771536141299565e-07,
+      "loss": 0.3438,
+      "step": 3266
+    },
+    {
+      "epoch": 2.7515440763615944,
+      "grad_norm": 0.16841080784797668,
+      "learning_rate": 2.0631925802791608e-07,
+      "loss": 0.3173,
+      "step": 3267
+    },
+    {
+      "epoch": 2.7523862998315556,
+      "grad_norm": 0.1783439964056015,
+      "learning_rate": 2.0492776339975374e-07,
+      "loss": 0.3484,
+      "step": 3268
+    },
+    {
+      "epoch": 2.753228523301516,
+      "grad_norm": 0.18109945952892303,
+      "learning_rate": 2.0354087886632623e-07,
+      "loss": 0.3232,
+      "step": 3269
+    },
+    {
+      "epoch": 2.7540707467714767,
+      "grad_norm": 0.1643245667219162,
+      "learning_rate": 2.0215860576101532e-07,
+      "loss": 0.323,
+      "step": 3270
+    },
+    {
+      "epoch": 2.7549129702414374,
+      "grad_norm": 0.1697859913110733,
+      "learning_rate": 2.0078094541277016e-07,
+      "loss": 0.323,
+      "step": 3271
+    },
+    {
+      "epoch": 2.755755193711398,
+      "grad_norm": 0.17192378640174866,
+      "learning_rate": 1.9940789914610682e-07,
+      "loss": 0.3397,
+      "step": 3272
+    },
+    {
+      "epoch": 2.756597417181359,
+      "grad_norm": 0.16300977766513824,
+      "learning_rate": 1.9803946828110376e-07,
+      "loss": 0.3215,
+      "step": 3273
+    },
+    {
+      "epoch": 2.7574396406513193,
+      "grad_norm": 0.17897097766399384,
+      "learning_rate": 1.966756541334025e-07,
+      "loss": 0.3411,
+      "step": 3274
+    },
+    {
+      "epoch": 2.75828186412128,
+      "grad_norm": 0.17692428827285767,
+      "learning_rate": 1.953164580142064e-07,
+      "loss": 0.3413,
+      "step": 3275
+    },
+    {
+      "epoch": 2.759124087591241,
+      "grad_norm": 0.17106205224990845,
+      "learning_rate": 1.9396188123027736e-07,
+      "loss": 0.3171,
+      "step": 3276
+    },
+    {
+      "epoch": 2.7599663110612016,
+      "grad_norm": 0.16745461523532867,
+      "learning_rate": 1.9261192508393755e-07,
+      "loss": 0.2923,
+      "step": 3277
+    },
+    {
+      "epoch": 2.7608085345311624,
+      "grad_norm": 0.19608838856220245,
+      "learning_rate": 1.912665908730671e-07,
+      "loss": 0.344,
+      "step": 3278
+    },
+    {
+      "epoch": 2.7616507580011227,
+      "grad_norm": 0.18816998600959778,
+      "learning_rate": 1.8992587989110133e-07,
+      "loss": 0.3349,
+      "step": 3279
+    },
+    {
+      "epoch": 2.762492981471084,
+      "grad_norm": 0.16697731614112854,
+      "learning_rate": 1.8858979342703088e-07,
+      "loss": 0.3318,
+      "step": 3280
+    },
+    {
+      "epoch": 2.7633352049410442,
+      "grad_norm": 0.16499722003936768,
+      "learning_rate": 1.8725833276540095e-07,
+      "loss": 0.32,
+      "step": 3281
+    },
+    {
+      "epoch": 2.764177428411005,
+      "grad_norm": 0.17865091562271118,
+      "learning_rate": 1.8593149918630927e-07,
+      "loss": 0.363,
+      "step": 3282
+    },
+    {
+      "epoch": 2.7650196518809658,
+      "grad_norm": 0.18328674137592316,
+      "learning_rate": 1.8460929396540428e-07,
+      "loss": 0.3149,
+      "step": 3283
+    },
+    {
+      "epoch": 2.7658618753509265,
+      "grad_norm": 0.1787954419851303,
+      "learning_rate": 1.8329171837388527e-07,
+      "loss": 0.3257,
+      "step": 3284
+    },
+    {
+      "epoch": 2.7667040988208873,
+      "grad_norm": 0.1814635545015335,
+      "learning_rate": 1.8197877367849948e-07,
+      "loss": 0.3336,
+      "step": 3285
+    },
+    {
+      "epoch": 2.7675463222908476,
+      "grad_norm": 0.1731037199497223,
+      "learning_rate": 1.8067046114154386e-07,
+      "loss": 0.3036,
+      "step": 3286
+    },
+    {
+      "epoch": 2.7683885457608084,
+      "grad_norm": 0.17906337976455688,
+      "learning_rate": 1.7936678202085945e-07,
+      "loss": 0.3525,
+      "step": 3287
+    },
+    {
+      "epoch": 2.769230769230769,
+      "grad_norm": 0.17718879878520966,
+      "learning_rate": 1.7806773756983641e-07,
+      "loss": 0.3007,
+      "step": 3288
+    },
+    {
+      "epoch": 2.77007299270073,
+      "grad_norm": 0.19208289682865143,
+      "learning_rate": 1.7677332903740296e-07,
+      "loss": 0.3397,
+      "step": 3289
+    },
+    {
+      "epoch": 2.7709152161706907,
+      "grad_norm": 0.16983190178871155,
+      "learning_rate": 1.7548355766803638e-07,
+      "loss": 0.327,
+      "step": 3290
+    },
+    {
+      "epoch": 2.7717574396406515,
+      "grad_norm": 0.1772148162126541,
+      "learning_rate": 1.7419842470175196e-07,
+      "loss": 0.3454,
+      "step": 3291
+    },
+    {
+      "epoch": 2.7725996631106122,
+      "grad_norm": 0.16711288690567017,
+      "learning_rate": 1.7291793137410695e-07,
+      "loss": 0.3494,
+      "step": 3292
+    },
+    {
+      "epoch": 2.7734418865805726,
+      "grad_norm": 1.8478301763534546,
+      "learning_rate": 1.7164207891619823e-07,
+      "loss": 0.3289,
+      "step": 3293
+    },
+    {
+      "epoch": 2.7742841100505333,
+      "grad_norm": 0.17004255950450897,
+      "learning_rate": 1.7037086855465902e-07,
+      "loss": 0.335,
+      "step": 3294
+    },
+    {
+      "epoch": 2.775126333520494,
+      "grad_norm": 0.1759692281484604,
+      "learning_rate": 1.6910430151166058e-07,
+      "loss": 0.3325,
+      "step": 3295
+    },
+    {
+      "epoch": 2.775968556990455,
+      "grad_norm": 1.8058511018753052,
+      "learning_rate": 1.6784237900491163e-07,
+      "loss": 0.3419,
+      "step": 3296
+    },
+    {
+      "epoch": 2.7768107804604156,
+      "grad_norm": 0.1898031383752823,
+      "learning_rate": 1.6658510224765333e-07,
+      "loss": 0.3411,
+      "step": 3297
+    },
+    {
+      "epoch": 2.777653003930376,
+      "grad_norm": 0.1677130162715912,
+      "learning_rate": 1.6533247244866102e-07,
+      "loss": 0.2952,
+      "step": 3298
+    },
+    {
+      "epoch": 2.778495227400337,
+      "grad_norm": 0.17073382437229156,
+      "learning_rate": 1.6408449081224131e-07,
+      "loss": 0.3464,
+      "step": 3299
+    },
+    {
+      "epoch": 2.7793374508702975,
+      "grad_norm": 0.1733926236629486,
+      "learning_rate": 1.6284115853823445e-07,
+      "loss": 0.3215,
+      "step": 3300
+    },
+    {
+      "epoch": 2.7801796743402583,
+      "grad_norm": 0.16632869839668274,
+      "learning_rate": 1.6160247682200813e-07,
+      "loss": 0.3229,
+      "step": 3301
+    },
+    {
+      "epoch": 2.781021897810219,
+      "grad_norm": 0.18543268740177155,
+      "learning_rate": 1.6036844685446084e-07,
+      "loss": 0.3336,
+      "step": 3302
+    },
+    {
+      "epoch": 2.78186412128018,
+      "grad_norm": 0.15792086720466614,
+      "learning_rate": 1.5913906982201744e-07,
+      "loss": 0.2924,
+      "step": 3303
+    },
+    {
+      "epoch": 2.7827063447501406,
+      "grad_norm": 0.18348781764507294,
+      "learning_rate": 1.5791434690662966e-07,
+      "loss": 0.3631,
+      "step": 3304
+    },
+    {
+      "epoch": 2.783548568220101,
+      "grad_norm": 0.1935282051563263,
+      "learning_rate": 1.566942792857745e-07,
+      "loss": 0.3336,
+      "step": 3305
+    },
+    {
+      "epoch": 2.7843907916900617,
+      "grad_norm": 0.1989465206861496,
+      "learning_rate": 1.554788681324554e-07,
+      "loss": 0.347,
+      "step": 3306
+    },
+    {
+      "epoch": 2.7852330151600224,
+      "grad_norm": 0.17033520340919495,
+      "learning_rate": 1.5426811461519419e-07,
+      "loss": 0.3008,
+      "step": 3307
+    },
+    {
+      "epoch": 2.786075238629983,
+      "grad_norm": 0.16638170182704926,
+      "learning_rate": 1.530620198980398e-07,
+      "loss": 0.3154,
+      "step": 3308
+    },
+    {
+      "epoch": 2.786917462099944,
+      "grad_norm": 0.18002824485301971,
+      "learning_rate": 1.5186058514055912e-07,
+      "loss": 0.3428,
+      "step": 3309
+    },
+    {
+      "epoch": 2.7877596855699043,
+      "grad_norm": 0.17526303231716156,
+      "learning_rate": 1.506638114978398e-07,
+      "loss": 0.296,
+      "step": 3310
+    },
+    {
+      "epoch": 2.7886019090398655,
+      "grad_norm": 0.18935514986515045,
+      "learning_rate": 1.4947170012048872e-07,
+      "loss": 0.3352,
+      "step": 3311
+    },
+    {
+      "epoch": 2.789444132509826,
+      "grad_norm": 0.17067907750606537,
+      "learning_rate": 1.482842521546285e-07,
+      "loss": 0.3025,
+      "step": 3312
+    },
+    {
+      "epoch": 2.7902863559797866,
+      "grad_norm": 0.18163152039051056,
+      "learning_rate": 1.471014687418998e-07,
+      "loss": 0.3591,
+      "step": 3313
+    },
+    {
+      "epoch": 2.7911285794497473,
+      "grad_norm": 0.19213518500328064,
+      "learning_rate": 1.4592335101945855e-07,
+      "loss": 0.3448,
+      "step": 3314
+    },
+    {
+      "epoch": 2.791970802919708,
+      "grad_norm": 0.17021815478801727,
+      "learning_rate": 1.447499001199748e-07,
+      "loss": 0.3027,
+      "step": 3315
+    },
+    {
+      "epoch": 2.792813026389669,
+      "grad_norm": 0.18658633530139923,
+      "learning_rate": 1.435811171716317e-07,
+      "loss": 0.3625,
+      "step": 3316
+    },
+    {
+      "epoch": 2.793655249859629,
+      "grad_norm": 0.17915920913219452,
+      "learning_rate": 1.4241700329812368e-07,
+      "loss": 0.3426,
+      "step": 3317
+    },
+    {
+      "epoch": 2.79449747332959,
+      "grad_norm": 0.16904447972774506,
+      "learning_rate": 1.4125755961865827e-07,
+      "loss": 0.32,
+      "step": 3318
+    },
+    {
+      "epoch": 2.7953396967995507,
+      "grad_norm": 0.17508530616760254,
+      "learning_rate": 1.4010278724795157e-07,
+      "loss": 0.303,
+      "step": 3319
+    },
+    {
+      "epoch": 2.7961819202695115,
+      "grad_norm": 0.2018783986568451,
+      "learning_rate": 1.3895268729622824e-07,
+      "loss": 0.3717,
+      "step": 3320
+    },
+    {
+      "epoch": 2.7970241437394723,
+      "grad_norm": 0.17020922899246216,
+      "learning_rate": 1.3780726086922103e-07,
+      "loss": 0.3199,
+      "step": 3321
+    },
+    {
+      "epoch": 2.797866367209433,
+      "grad_norm": 0.17910431325435638,
+      "learning_rate": 1.366665090681707e-07,
+      "loss": 0.3188,
+      "step": 3322
+    },
+    {
+      "epoch": 2.798708590679394,
+      "grad_norm": 0.19550931453704834,
+      "learning_rate": 1.355304329898216e-07,
+      "loss": 0.3448,
+      "step": 3323
+    },
+    {
+      "epoch": 2.799550814149354,
+      "grad_norm": 0.17510132491588593,
+      "learning_rate": 1.3439903372642615e-07,
+      "loss": 0.3019,
+      "step": 3324
+    },
+    {
+      "epoch": 2.800393037619315,
+      "grad_norm": 0.16419118642807007,
+      "learning_rate": 1.332723123657348e-07,
+      "loss": 0.3421,
+      "step": 3325
+    },
+    {
+      "epoch": 2.8012352610892757,
+      "grad_norm": 0.16772206127643585,
+      "learning_rate": 1.3215026999100655e-07,
+      "loss": 0.3579,
+      "step": 3326
+    },
+    {
+      "epoch": 2.8020774845592364,
+      "grad_norm": 0.153725266456604,
+      "learning_rate": 1.3103290768099796e-07,
+      "loss": 0.3065,
+      "step": 3327
+    },
+    {
+      "epoch": 2.802919708029197,
+      "grad_norm": 0.19064374268054962,
+      "learning_rate": 1.299202265099675e-07,
+      "loss": 0.339,
+      "step": 3328
+    },
+    {
+      "epoch": 2.8037619314991575,
+      "grad_norm": 0.18251188099384308,
+      "learning_rate": 1.288122275476733e-07,
+      "loss": 0.3348,
+      "step": 3329
+    },
+    {
+      "epoch": 2.8046041549691187,
+      "grad_norm": 0.17828452587127686,
+      "learning_rate": 1.2770891185937106e-07,
+      "loss": 0.3311,
+      "step": 3330
+    },
+    {
+      "epoch": 2.805446378439079,
+      "grad_norm": 0.16462412476539612,
+      "learning_rate": 1.2661028050581446e-07,
+      "loss": 0.3522,
+      "step": 3331
+    },
+    {
+      "epoch": 2.80628860190904,
+      "grad_norm": 0.2014075666666031,
+      "learning_rate": 1.2551633454325362e-07,
+      "loss": 0.332,
+      "step": 3332
+    },
+    {
+      "epoch": 2.8071308253790006,
+      "grad_norm": 0.169560968875885,
+      "learning_rate": 1.244270750234333e-07,
+      "loss": 0.341,
+      "step": 3333
+    },
+    {
+      "epoch": 2.8079730488489614,
+      "grad_norm": 0.16761159896850586,
+      "learning_rate": 1.2334250299359362e-07,
+      "loss": 0.3033,
+      "step": 3334
+    },
+    {
+      "epoch": 2.808815272318922,
+      "grad_norm": 0.1883648782968521,
+      "learning_rate": 1.2226261949646656e-07,
+      "loss": 0.355,
+      "step": 3335
+    },
+    {
+      "epoch": 2.8096574957888825,
+      "grad_norm": 0.16675567626953125,
+      "learning_rate": 1.2118742557027885e-07,
+      "loss": 0.3185,
+      "step": 3336
+    },
+    {
+      "epoch": 2.8104997192588432,
+      "grad_norm": 0.17546634376049042,
+      "learning_rate": 1.201169222487464e-07,
+      "loss": 0.328,
+      "step": 3337
+    },
+    {
+      "epoch": 2.811341942728804,
+      "grad_norm": 0.16184347867965698,
+      "learning_rate": 1.1905111056107644e-07,
+      "loss": 0.3366,
+      "step": 3338
+    },
+    {
+      "epoch": 2.8121841661987648,
+      "grad_norm": 0.1763359010219574,
+      "learning_rate": 1.1798999153196433e-07,
+      "loss": 0.3356,
+      "step": 3339
+    },
+    {
+      "epoch": 2.8130263896687255,
+      "grad_norm": 0.1889248639345169,
+      "learning_rate": 1.1693356618159568e-07,
+      "loss": 0.3127,
+      "step": 3340
+    },
+    {
+      "epoch": 2.813868613138686,
+      "grad_norm": 0.18146924674510956,
+      "learning_rate": 1.1588183552564247e-07,
+      "loss": 0.3352,
+      "step": 3341
+    },
+    {
+      "epoch": 2.814710836608647,
+      "grad_norm": 0.17269383370876312,
+      "learning_rate": 1.1483480057526364e-07,
+      "loss": 0.3399,
+      "step": 3342
+    },
+    {
+      "epoch": 2.8155530600786074,
+      "grad_norm": 0.16894157230854034,
+      "learning_rate": 1.1379246233710172e-07,
+      "loss": 0.3349,
+      "step": 3343
+    },
+    {
+      "epoch": 2.816395283548568,
+      "grad_norm": 0.1881352812051773,
+      "learning_rate": 1.1275482181328568e-07,
+      "loss": 0.3627,
+      "step": 3344
+    },
+    {
+      "epoch": 2.817237507018529,
+      "grad_norm": 0.19253571331501007,
+      "learning_rate": 1.1172188000142803e-07,
+      "loss": 0.3266,
+      "step": 3345
+    },
+    {
+      "epoch": 2.8180797304884897,
+      "grad_norm": 0.17829467356204987,
+      "learning_rate": 1.1069363789462273e-07,
+      "loss": 0.3348,
+      "step": 3346
+    },
+    {
+      "epoch": 2.8189219539584505,
+      "grad_norm": 0.16123495995998383,
+      "learning_rate": 1.0967009648144621e-07,
+      "loss": 0.2821,
+      "step": 3347
+    },
+    {
+      "epoch": 2.819764177428411,
+      "grad_norm": 0.18511976301670074,
+      "learning_rate": 1.0865125674595467e-07,
+      "loss": 0.3455,
+      "step": 3348
+    },
+    {
+      "epoch": 2.8206064008983716,
+      "grad_norm": 0.18319936096668243,
+      "learning_rate": 1.0763711966768453e-07,
+      "loss": 0.3471,
+      "step": 3349
+    },
+    {
+      "epoch": 2.8214486243683323,
+      "grad_norm": 0.16767950356006622,
+      "learning_rate": 1.0662768622165087e-07,
+      "loss": 0.3123,
+      "step": 3350
+    },
+    {
+      "epoch": 2.822290847838293,
+      "grad_norm": 0.1737443506717682,
+      "learning_rate": 1.0562295737834738e-07,
+      "loss": 0.3511,
+      "step": 3351
+    },
+    {
+      "epoch": 2.823133071308254,
+      "grad_norm": 0.184846431016922,
+      "learning_rate": 1.0462293410374303e-07,
+      "loss": 0.3405,
+      "step": 3352
+    },
+    {
+      "epoch": 2.8239752947782146,
+      "grad_norm": 0.1793210804462433,
+      "learning_rate": 1.0362761735928372e-07,
+      "loss": 0.3326,
+      "step": 3353
+    },
+    {
+      "epoch": 2.8248175182481754,
+      "grad_norm": 0.17455455660820007,
+      "learning_rate": 1.026370081018907e-07,
+      "loss": 0.2945,
+      "step": 3354
+    },
+    {
+      "epoch": 2.8256597417181357,
+      "grad_norm": 0.17478932440280914,
+      "learning_rate": 1.0165110728395878e-07,
+      "loss": 0.3273,
+      "step": 3355
+    },
+    {
+      "epoch": 2.8265019651880965,
+      "grad_norm": 0.16787944734096527,
+      "learning_rate": 1.0066991585335583e-07,
+      "loss": 0.3345,
+      "step": 3356
+    },
+    {
+      "epoch": 2.8273441886580573,
+      "grad_norm": 0.16388723254203796,
+      "learning_rate": 9.969343475342285e-08,
+      "loss": 0.3339,
+      "step": 3357
+    },
+    {
+      "epoch": 2.828186412128018,
+      "grad_norm": 0.1678238958120346,
+      "learning_rate": 9.872166492297052e-08,
+      "loss": 0.3409,
+      "step": 3358
+    },
+    {
+      "epoch": 2.829028635597979,
+      "grad_norm": 0.16698604822158813,
+      "learning_rate": 9.775460729628262e-08,
+      "loss": 0.3203,
+      "step": 3359
+    },
+    {
+      "epoch": 2.829870859067939,
+      "grad_norm": 0.1685776263475418,
+      "learning_rate": 9.679226280310982e-08,
+      "loss": 0.317,
+      "step": 3360
+    },
+    {
+      "epoch": 2.8307130825379003,
+      "grad_norm": 0.17047767341136932,
+      "learning_rate": 9.583463236867318e-08,
+      "loss": 0.3277,
+      "step": 3361
+    },
+    {
+      "epoch": 2.8315553060078607,
+      "grad_norm": 0.18030725419521332,
+      "learning_rate": 9.48817169136601e-08,
+      "loss": 0.3193,
+      "step": 3362
+    },
+    {
+      "epoch": 2.8323975294778214,
+      "grad_norm": 0.18441860377788544,
+      "learning_rate": 9.393351735422773e-08,
+      "loss": 0.3523,
+      "step": 3363
+    },
+    {
+      "epoch": 2.833239752947782,
+      "grad_norm": 0.15680761635303497,
+      "learning_rate": 9.299003460199519e-08,
+      "loss": 0.3235,
+      "step": 3364
+    },
+    {
+      "epoch": 2.834081976417743,
+      "grad_norm": 0.17292484641075134,
+      "learning_rate": 9.205126956405075e-08,
+      "loss": 0.3305,
+      "step": 3365
+    },
+    {
+      "epoch": 2.8349241998877037,
+      "grad_norm": 0.1728994995355606,
+      "learning_rate": 9.111722314294358e-08,
+      "loss": 0.3431,
+      "step": 3366
+    },
+    {
+      "epoch": 2.835766423357664,
+      "grad_norm": 0.18224498629570007,
+      "learning_rate": 9.018789623668866e-08,
+      "loss": 0.3662,
+      "step": 3367
+    },
+    {
+      "epoch": 2.836608646827625,
+      "grad_norm": 0.19175554811954498,
+      "learning_rate": 8.926328973876242e-08,
+      "loss": 0.3096,
+      "step": 3368
+    },
+    {
+      "epoch": 2.8374508702975856,
+      "grad_norm": 0.1898619830608368,
+      "learning_rate": 8.834340453810375e-08,
+      "loss": 0.3613,
+      "step": 3369
+    },
+    {
+      "epoch": 2.8382930937675463,
+      "grad_norm": 0.17442317306995392,
+      "learning_rate": 8.742824151911022e-08,
+      "loss": 0.3078,
+      "step": 3370
+    },
+    {
+      "epoch": 2.839135317237507,
+      "grad_norm": 0.1779947429895401,
+      "learning_rate": 8.651780156164302e-08,
+      "loss": 0.3168,
+      "step": 3371
+    },
+    {
+      "epoch": 2.8399775407074674,
+      "grad_norm": 0.17783810198307037,
+      "learning_rate": 8.561208554101863e-08,
+      "loss": 0.3499,
+      "step": 3372
+    },
+    {
+      "epoch": 2.8408197641774287,
+      "grad_norm": 0.16697177290916443,
+      "learning_rate": 8.471109432801494e-08,
+      "loss": 0.3265,
+      "step": 3373
+    },
+    {
+      "epoch": 2.841661987647389,
+      "grad_norm": 0.160588338971138,
+      "learning_rate": 8.381482878886571e-08,
+      "loss": 0.3002,
+      "step": 3374
+    },
+    {
+      "epoch": 2.8425042111173497,
+      "grad_norm": 0.1759585589170456,
+      "learning_rate": 8.29232897852611e-08,
+      "loss": 0.397,
+      "step": 3375
+    },
+    {
+      "epoch": 2.8433464345873105,
+      "grad_norm": 0.16448669135570526,
+      "learning_rate": 8.203647817434823e-08,
+      "loss": 0.3006,
+      "step": 3376
+    },
+    {
+      "epoch": 2.8441886580572713,
+      "grad_norm": 0.17329882085323334,
+      "learning_rate": 8.11543948087279e-08,
+      "loss": 0.293,
+      "step": 3377
+    },
+    {
+      "epoch": 2.845030881527232,
+      "grad_norm": 0.18138037621974945,
+      "learning_rate": 8.027704053645613e-08,
+      "loss": 0.3608,
+      "step": 3378
+    },
+    {
+      "epoch": 2.8458731049971924,
+      "grad_norm": 0.1575583964586258,
+      "learning_rate": 7.94044162010421e-08,
+      "loss": 0.3175,
+      "step": 3379
+    },
+    {
+      "epoch": 2.846715328467153,
+      "grad_norm": 0.17390009760856628,
+      "learning_rate": 7.85365226414464e-08,
+      "loss": 0.3437,
+      "step": 3380
+    },
+    {
+      "epoch": 2.847557551937114,
+      "grad_norm": 0.17777961492538452,
+      "learning_rate": 7.76733606920832e-08,
+      "loss": 0.3743,
+      "step": 3381
+    },
+    {
+      "epoch": 2.8483997754070747,
+      "grad_norm": 0.1718461662530899,
+      "learning_rate": 7.681493118281646e-08,
+      "loss": 0.3202,
+      "step": 3382
+    },
+    {
+      "epoch": 2.8492419988770354,
+      "grad_norm": 0.15905161201953888,
+      "learning_rate": 7.59612349389599e-08,
+      "loss": 0.2956,
+      "step": 3383
+    },
+    {
+      "epoch": 2.850084222346996,
+      "grad_norm": 0.18780933320522308,
+      "learning_rate": 7.511227278127697e-08,
+      "loss": 0.3396,
+      "step": 3384
+    },
+    {
+      "epoch": 2.850926445816957,
+      "grad_norm": 0.17712104320526123,
+      "learning_rate": 7.426804552598088e-08,
+      "loss": 0.3616,
+      "step": 3385
+    },
+    {
+      "epoch": 2.8517686692869173,
+      "grad_norm": 0.1937163770198822,
+      "learning_rate": 7.342855398472958e-08,
+      "loss": 0.3452,
+      "step": 3386
+    },
+    {
+      "epoch": 2.852610892756878,
+      "grad_norm": 0.16116411983966827,
+      "learning_rate": 7.259379896463248e-08,
+      "loss": 0.3204,
+      "step": 3387
+    },
+    {
+      "epoch": 2.853453116226839,
+      "grad_norm": 0.16988983750343323,
+      "learning_rate": 7.176378126824035e-08,
+      "loss": 0.3418,
+      "step": 3388
+    },
+    {
+      "epoch": 2.8542953396967996,
+      "grad_norm": 0.16600120067596436,
+      "learning_rate": 7.093850169355266e-08,
+      "loss": 0.3093,
+      "step": 3389
+    },
+    {
+      "epoch": 2.8551375631667604,
+      "grad_norm": 0.16801238059997559,
+      "learning_rate": 7.011796103401192e-08,
+      "loss": 0.335,
+      "step": 3390
+    },
+    {
+      "epoch": 2.8559797866367207,
+      "grad_norm": 0.18370668590068817,
+      "learning_rate": 6.930216007850598e-08,
+      "loss": 0.3361,
+      "step": 3391
+    },
+    {
+      "epoch": 2.856822010106682,
+      "grad_norm": 0.18052010238170624,
+      "learning_rate": 6.849109961136468e-08,
+      "loss": 0.3241,
+      "step": 3392
+    },
+    {
+      "epoch": 2.8576642335766422,
+      "grad_norm": 0.16566221415996552,
+      "learning_rate": 6.768478041236037e-08,
+      "loss": 0.3119,
+      "step": 3393
+    },
+    {
+      "epoch": 2.858506457046603,
+      "grad_norm": 0.15954545140266418,
+      "learning_rate": 6.688320325670628e-08,
+      "loss": 0.3244,
+      "step": 3394
+    },
+    {
+      "epoch": 2.8593486805165638,
+      "grad_norm": 0.16784389317035675,
+      "learning_rate": 6.608636891505982e-08,
+      "loss": 0.3279,
+      "step": 3395
+    },
+    {
+      "epoch": 2.8601909039865245,
+      "grad_norm": 0.18025429546833038,
+      "learning_rate": 6.529427815351374e-08,
+      "loss": 0.3239,
+      "step": 3396
+    },
+    {
+      "epoch": 2.8610331274564853,
+      "grad_norm": 0.17966340482234955,
+      "learning_rate": 6.450693173360445e-08,
+      "loss": 0.3264,
+      "step": 3397
+    },
+    {
+      "epoch": 2.8618753509264456,
+      "grad_norm": 0.16547337174415588,
+      "learning_rate": 6.372433041230364e-08,
+      "loss": 0.3351,
+      "step": 3398
+    },
+    {
+      "epoch": 2.8627175743964064,
+      "grad_norm": 0.16610006988048553,
+      "learning_rate": 6.294647494202444e-08,
+      "loss": 0.3254,
+      "step": 3399
+    },
+    {
+      "epoch": 2.863559797866367,
+      "grad_norm": 0.1672658920288086,
+      "learning_rate": 6.217336607061364e-08,
+      "loss": 0.3208,
+      "step": 3400
+    },
+    {
+      "epoch": 2.864402021336328,
+      "grad_norm": 0.18397913873195648,
+      "learning_rate": 6.140500454135668e-08,
+      "loss": 0.3249,
+      "step": 3401
+    },
+    {
+      "epoch": 2.8652442448062887,
+      "grad_norm": 0.19070270657539368,
+      "learning_rate": 6.064139109297485e-08,
+      "loss": 0.3895,
+      "step": 3402
+    },
+    {
+      "epoch": 2.866086468276249,
+      "grad_norm": 0.17317475378513336,
+      "learning_rate": 5.988252645962367e-08,
+      "loss": 0.3131,
+      "step": 3403
+    },
+    {
+      "epoch": 2.8669286917462102,
+      "grad_norm": 0.16814860701560974,
+      "learning_rate": 5.912841137089287e-08,
+      "loss": 0.3339,
+      "step": 3404
+    },
+    {
+      "epoch": 2.8677709152161706,
+      "grad_norm": 0.17419366538524628,
+      "learning_rate": 5.8379046551807486e-08,
+      "loss": 0.3121,
+      "step": 3405
+    },
+    {
+      "epoch": 2.8686131386861313,
+      "grad_norm": 0.16927020251750946,
+      "learning_rate": 5.7634432722822875e-08,
+      "loss": 0.3349,
+      "step": 3406
+    },
+    {
+      "epoch": 2.869455362156092,
+      "grad_norm": 0.16771690547466278,
+      "learning_rate": 5.6894570599829726e-08,
+      "loss": 0.3062,
+      "step": 3407
+    },
+    {
+      "epoch": 2.870297585626053,
+      "grad_norm": 0.18303582072257996,
+      "learning_rate": 5.615946089414737e-08,
+      "loss": 0.3431,
+      "step": 3408
+    },
+    {
+      "epoch": 2.8711398090960136,
+      "grad_norm": 0.1693166345357895,
+      "learning_rate": 5.542910431252935e-08,
+      "loss": 0.3123,
+      "step": 3409
+    },
+    {
+      "epoch": 2.871982032565974,
+      "grad_norm": 0.17601299285888672,
+      "learning_rate": 5.470350155715565e-08,
+      "loss": 0.3492,
+      "step": 3410
+    },
+    {
+      "epoch": 2.8728242560359347,
+      "grad_norm": 0.17169730365276337,
+      "learning_rate": 5.398265332563935e-08,
+      "loss": 0.3386,
+      "step": 3411
+    },
+    {
+      "epoch": 2.8736664795058955,
+      "grad_norm": 0.17456136643886566,
+      "learning_rate": 5.32665603110194e-08,
+      "loss": 0.3393,
+      "step": 3412
+    },
+    {
+      "epoch": 2.8745087029758563,
+      "grad_norm": 0.16657783091068268,
+      "learning_rate": 5.255522320176565e-08,
+      "loss": 0.3061,
+      "step": 3413
+    },
+    {
+      "epoch": 2.875350926445817,
+      "grad_norm": 0.1727527678012848,
+      "learning_rate": 5.1848642681773254e-08,
+      "loss": 0.3089,
+      "step": 3414
+    },
+    {
+      "epoch": 2.876193149915778,
+      "grad_norm": 0.17237474024295807,
+      "learning_rate": 5.114681943036603e-08,
+      "loss": 0.3215,
+      "step": 3415
+    },
+    {
+      "epoch": 2.8770353733857386,
+      "grad_norm": 0.17007453739643097,
+      "learning_rate": 5.0449754122292585e-08,
+      "loss": 0.368,
+      "step": 3416
+    },
+    {
+      "epoch": 2.877877596855699,
+      "grad_norm": 0.1603289097547531,
+      "learning_rate": 4.975744742772848e-08,
+      "loss": 0.3295,
+      "step": 3417
+    },
+    {
+      "epoch": 2.8787198203256597,
+      "grad_norm": 0.18231481313705444,
+      "learning_rate": 4.906990001227296e-08,
+      "loss": 0.3121,
+      "step": 3418
+    },
+    {
+      "epoch": 2.8795620437956204,
+      "grad_norm": 0.1824287474155426,
+      "learning_rate": 4.838711253695061e-08,
+      "loss": 0.3582,
+      "step": 3419
+    },
+    {
+      "epoch": 2.880404267265581,
+      "grad_norm": 0.17323684692382812,
+      "learning_rate": 4.770908565820964e-08,
+      "loss": 0.3265,
+      "step": 3420
+    },
+    {
+      "epoch": 2.881246490735542,
+      "grad_norm": 0.17707312107086182,
+      "learning_rate": 4.7035820027920284e-08,
+      "loss": 0.2921,
+      "step": 3421
+    },
+    {
+      "epoch": 2.8820887142055023,
+      "grad_norm": 0.1917116492986679,
+      "learning_rate": 4.636731629337587e-08,
+      "loss": 0.3445,
+      "step": 3422
+    },
+    {
+      "epoch": 2.8829309376754635,
+      "grad_norm": 0.1672520935535431,
+      "learning_rate": 4.5703575097292286e-08,
+      "loss": 0.327,
+      "step": 3423
+    },
+    {
+      "epoch": 2.883773161145424,
+      "grad_norm": 0.168136328458786,
+      "learning_rate": 4.5044597077805175e-08,
+      "loss": 0.3227,
+      "step": 3424
+    },
+    {
+      "epoch": 2.8846153846153846,
+      "grad_norm": 0.1622181236743927,
+      "learning_rate": 4.439038286847164e-08,
+      "loss": 0.322,
+      "step": 3425
+    },
+    {
+      "epoch": 2.8854576080853453,
+      "grad_norm": 0.16611912846565247,
+      "learning_rate": 4.37409330982691e-08,
+      "loss": 0.3496,
+      "step": 3426
+    },
+    {
+      "epoch": 2.886299831555306,
+      "grad_norm": 0.17439675331115723,
+      "learning_rate": 4.309624839159254e-08,
+      "loss": 0.3544,
+      "step": 3427
+    },
+    {
+      "epoch": 2.887142055025267,
+      "grad_norm": 0.17138072848320007,
+      "learning_rate": 4.245632936825783e-08,
+      "loss": 0.3419,
+      "step": 3428
+    },
+    {
+      "epoch": 2.887984278495227,
+      "grad_norm": 0.17312675714492798,
+      "learning_rate": 4.182117664349783e-08,
+      "loss": 0.2944,
+      "step": 3429
+    },
+    {
+      "epoch": 2.888826501965188,
+      "grad_norm": 0.18604117631912231,
+      "learning_rate": 4.119079082796351e-08,
+      "loss": 0.3626,
+      "step": 3430
+    },
+    {
+      "epoch": 2.8896687254351487,
+      "grad_norm": 0.16157913208007812,
+      "learning_rate": 4.056517252772229e-08,
+      "loss": 0.3313,
+      "step": 3431
+    },
+    {
+      "epoch": 2.8905109489051095,
+      "grad_norm": 0.1615222543478012,
+      "learning_rate": 3.99443223442586e-08,
+      "loss": 0.3249,
+      "step": 3432
+    },
+    {
+      "epoch": 2.8913531723750703,
+      "grad_norm": 0.19957011938095093,
+      "learning_rate": 3.9328240874471624e-08,
+      "loss": 0.3522,
+      "step": 3433
+    },
+    {
+      "epoch": 2.892195395845031,
+      "grad_norm": 0.17030540108680725,
+      "learning_rate": 3.871692871067756e-08,
+      "loss": 0.3024,
+      "step": 3434
+    },
+    {
+      "epoch": 2.893037619314992,
+      "grad_norm": 0.17600853741168976,
+      "learning_rate": 3.8110386440605164e-08,
+      "loss": 0.3452,
+      "step": 3435
+    },
+    {
+      "epoch": 2.893879842784952,
+      "grad_norm": 0.1735491305589676,
+      "learning_rate": 3.750861464739908e-08,
+      "loss": 0.3244,
+      "step": 3436
+    },
+    {
+      "epoch": 2.894722066254913,
+      "grad_norm": 0.17435042560100555,
+      "learning_rate": 3.6911613909616505e-08,
+      "loss": 0.3274,
+      "step": 3437
+    },
+    {
+      "epoch": 2.8955642897248737,
+      "grad_norm": 0.28229641914367676,
+      "learning_rate": 3.631938480122777e-08,
+      "loss": 0.3657,
+      "step": 3438
+    },
+    {
+      "epoch": 2.8964065131948344,
+      "grad_norm": 0.1704999953508377,
+      "learning_rate": 3.573192789161628e-08,
+      "loss": 0.297,
+      "step": 3439
+    },
+    {
+      "epoch": 2.897248736664795,
+      "grad_norm": 0.17650829255580902,
+      "learning_rate": 3.514924374557638e-08,
+      "loss": 0.3409,
+      "step": 3440
+    },
+    {
+      "epoch": 2.8980909601347555,
+      "grad_norm": 0.18400131165981293,
+      "learning_rate": 3.457133292331494e-08,
+      "loss": 0.3751,
+      "step": 3441
+    },
+    {
+      "epoch": 2.8989331836047163,
+      "grad_norm": 0.17719951272010803,
+      "learning_rate": 3.3998195980448065e-08,
+      "loss": 0.3225,
+      "step": 3442
+    },
+    {
+      "epoch": 2.899775407074677,
+      "grad_norm": 0.177615225315094,
+      "learning_rate": 3.342983346800388e-08,
+      "loss": 0.3262,
+      "step": 3443
+    },
+    {
+      "epoch": 2.900617630544638,
+      "grad_norm": 0.16130474209785461,
+      "learning_rate": 3.2866245932418606e-08,
+      "loss": 0.2945,
+      "step": 3444
+    },
+    {
+      "epoch": 2.9014598540145986,
+      "grad_norm": 0.16819952428340912,
+      "learning_rate": 3.230743391553881e-08,
+      "loss": 0.3536,
+      "step": 3445
+    },
+    {
+      "epoch": 2.9023020774845594,
+      "grad_norm": 0.16620105504989624,
+      "learning_rate": 3.175339795462029e-08,
+      "loss": 0.3125,
+      "step": 3446
+    },
+    {
+      "epoch": 2.90314430095452,
+      "grad_norm": 0.1779838651418686,
+      "learning_rate": 3.120413858232474e-08,
+      "loss": 0.3397,
+      "step": 3447
+    },
+    {
+      "epoch": 2.9039865244244805,
+      "grad_norm": 0.17595644295215607,
+      "learning_rate": 3.0659656326724186e-08,
+      "loss": 0.3416,
+      "step": 3448
+    },
+    {
+      "epoch": 2.9048287478944412,
+      "grad_norm": 0.17794086039066315,
+      "learning_rate": 3.011995171129545e-08,
+      "loss": 0.3403,
+      "step": 3449
+    },
+    {
+      "epoch": 2.905670971364402,
+      "grad_norm": 0.16569305956363678,
+      "learning_rate": 2.9585025254924572e-08,
+      "loss": 0.3333,
+      "step": 3450
+    },
+    {
+      "epoch": 2.9065131948343628,
+      "grad_norm": 0.17164397239685059,
+      "learning_rate": 2.9054877471901277e-08,
+      "loss": 0.3338,
+      "step": 3451
+    },
+    {
+      "epoch": 2.9073554183043235,
+      "grad_norm": 0.1742820292711258,
+      "learning_rate": 2.852950887192285e-08,
+      "loss": 0.2963,
+      "step": 3452
+    },
+    {
+      "epoch": 2.908197641774284,
+      "grad_norm": 0.1825890839099884,
+      "learning_rate": 2.8008919960090253e-08,
+      "loss": 0.355,
+      "step": 3453
+    },
+    {
+      "epoch": 2.909039865244245,
+      "grad_norm": 0.15666228532791138,
+      "learning_rate": 2.7493111236909787e-08,
+      "loss": 0.2954,
+      "step": 3454
+    },
+    {
+      "epoch": 2.9098820887142054,
+      "grad_norm": 0.17284969985485077,
+      "learning_rate": 2.6982083198293096e-08,
+      "loss": 0.3231,
+      "step": 3455
+    },
+    {
+      "epoch": 2.910724312184166,
+      "grad_norm": 0.16500769555568695,
+      "learning_rate": 2.6475836335553838e-08,
+      "loss": 0.3283,
+      "step": 3456
+    },
+    {
+      "epoch": 2.911566535654127,
+      "grad_norm": 0.17349562048912048,
+      "learning_rate": 2.5974371135408792e-08,
+      "loss": 0.3398,
+      "step": 3457
+    },
+    {
+      "epoch": 2.9124087591240877,
+      "grad_norm": 0.18464182317256927,
+      "learning_rate": 2.5477688079979522e-08,
+      "loss": 0.3366,
+      "step": 3458
+    },
+    {
+      "epoch": 2.9132509825940485,
+      "grad_norm": 0.16741546988487244,
+      "learning_rate": 2.4985787646788497e-08,
+      "loss": 0.3306,
+      "step": 3459
+    },
+    {
+      "epoch": 2.914093206064009,
+      "grad_norm": 0.16009648144245148,
+      "learning_rate": 2.4498670308760742e-08,
+      "loss": 0.3007,
+      "step": 3460
+    },
+    {
+      "epoch": 2.9149354295339696,
+      "grad_norm": 0.17391559481620789,
+      "learning_rate": 2.401633653422053e-08,
+      "loss": 0.3207,
+      "step": 3461
+    },
+    {
+      "epoch": 2.9157776530039303,
+      "grad_norm": 0.1760876476764679,
+      "learning_rate": 2.3538786786896918e-08,
+      "loss": 0.35,
+      "step": 3462
+    },
+    {
+      "epoch": 2.916619876473891,
+      "grad_norm": 0.16919660568237305,
+      "learning_rate": 2.306602152591597e-08,
+      "loss": 0.3329,
+      "step": 3463
+    },
+    {
+      "epoch": 2.917462099943852,
+      "grad_norm": 0.17137733101844788,
+      "learning_rate": 2.2598041205806333e-08,
+      "loss": 0.3125,
+      "step": 3464
+    },
+    {
+      "epoch": 2.9183043234138126,
+      "grad_norm": 0.1716642528772354,
+      "learning_rate": 2.2134846276494205e-08,
+      "loss": 0.3861,
+      "step": 3465
+    },
+    {
+      "epoch": 2.9191465468837734,
+      "grad_norm": 0.1804467886686325,
+      "learning_rate": 2.1676437183306697e-08,
+      "loss": 0.3229,
+      "step": 3466
+    },
+    {
+      "epoch": 2.9199887703537337,
+      "grad_norm": 0.17677058279514313,
+      "learning_rate": 2.1222814366969048e-08,
+      "loss": 0.3313,
+      "step": 3467
+    },
+    {
+      "epoch": 2.9208309938236945,
+      "grad_norm": 0.170046865940094,
+      "learning_rate": 2.0773978263605164e-08,
+      "loss": 0.3434,
+      "step": 3468
+    },
+    {
+      "epoch": 2.9216732172936553,
+      "grad_norm": 0.18440905213356018,
+      "learning_rate": 2.032992930473543e-08,
+      "loss": 0.3705,
+      "step": 3469
+    },
+    {
+      "epoch": 2.922515440763616,
+      "grad_norm": 0.16616974771022797,
+      "learning_rate": 1.9890667917280006e-08,
+      "loss": 0.3014,
+      "step": 3470
+    },
+    {
+      "epoch": 2.923357664233577,
+      "grad_norm": 0.1813036948442459,
+      "learning_rate": 1.9456194523554404e-08,
+      "loss": 0.3493,
+      "step": 3471
+    },
+    {
+      "epoch": 2.924199887703537,
+      "grad_norm": 0.17219789326190948,
+      "learning_rate": 1.9026509541272276e-08,
+      "loss": 0.3366,
+      "step": 3472
+    },
+    {
+      "epoch": 2.925042111173498,
+      "grad_norm": 0.17453472316265106,
+      "learning_rate": 1.860161338354205e-08,
+      "loss": 0.3411,
+      "step": 3473
+    },
+    {
+      "epoch": 2.9258843346434587,
+      "grad_norm": 0.17616593837738037,
+      "learning_rate": 1.8181506458869735e-08,
+      "loss": 0.3178,
+      "step": 3474
+    },
+    {
+      "epoch": 2.9267265581134194,
+      "grad_norm": 0.1713724583387375,
+      "learning_rate": 1.7766189171154468e-08,
+      "loss": 0.305,
+      "step": 3475
+    },
+    {
+      "epoch": 2.92756878158338,
+      "grad_norm": 0.16618001461029053,
+      "learning_rate": 1.7355661919693513e-08,
+      "loss": 0.3304,
+      "step": 3476
+    },
+    {
+      "epoch": 2.928411005053341,
+      "grad_norm": 0.16214674711227417,
+      "learning_rate": 1.69499250991767e-08,
+      "loss": 0.3337,
+      "step": 3477
+    },
+    {
+      "epoch": 2.9292532285233017,
+      "grad_norm": 0.18076547980308533,
+      "learning_rate": 1.654897909968922e-08,
+      "loss": 0.3099,
+      "step": 3478
+    },
+    {
+      "epoch": 2.930095451993262,
+      "grad_norm": 0.17332997918128967,
+      "learning_rate": 1.6152824306709392e-08,
+      "loss": 0.3344,
+      "step": 3479
+    },
+    {
+      "epoch": 2.930937675463223,
+      "grad_norm": 0.18455633521080017,
+      "learning_rate": 1.576146110111032e-08,
+      "loss": 0.3437,
+      "step": 3480
+    },
+    {
+      "epoch": 2.9317798989331836,
+      "grad_norm": 0.1677646040916443,
+      "learning_rate": 1.5374889859157137e-08,
+      "loss": 0.3167,
+      "step": 3481
+    },
+    {
+      "epoch": 2.9326221224031443,
+      "grad_norm": 0.17141175270080566,
+      "learning_rate": 1.4993110952509215e-08,
+      "loss": 0.3195,
+      "step": 3482
+    },
+    {
+      "epoch": 2.933464345873105,
+      "grad_norm": 0.17640945315361023,
+      "learning_rate": 1.4616124748217387e-08,
+      "loss": 0.3656,
+      "step": 3483
+    },
+    {
+      "epoch": 2.9343065693430654,
+      "grad_norm": 0.17052777111530304,
+      "learning_rate": 1.424393160872506e-08,
+      "loss": 0.3364,
+      "step": 3484
+    },
+    {
+      "epoch": 2.9351487928130267,
+      "grad_norm": 0.19747313857078552,
+      "learning_rate": 1.3876531891867106e-08,
+      "loss": 0.2958,
+      "step": 3485
+    },
+    {
+      "epoch": 2.935991016282987,
+      "grad_norm": 0.1779537945985794,
+      "learning_rate": 1.351392595087042e-08,
+      "loss": 0.2915,
+      "step": 3486
+    },
+    {
+      "epoch": 2.9368332397529477,
+      "grad_norm": 0.18276581168174744,
+      "learning_rate": 1.3156114134352805e-08,
+      "loss": 0.3353,
+      "step": 3487
+    },
+    {
+      "epoch": 2.9376754632229085,
+      "grad_norm": 0.17492914199829102,
+      "learning_rate": 1.2803096786323521e-08,
+      "loss": 0.3233,
+      "step": 3488
+    },
+    {
+      "epoch": 2.9385176866928693,
+      "grad_norm": 0.17329710721969604,
+      "learning_rate": 1.2454874246181081e-08,
+      "loss": 0.3359,
+      "step": 3489
+    },
+    {
+      "epoch": 2.93935991016283,
+      "grad_norm": 0.17337341606616974,
+      "learning_rate": 1.2111446848714347e-08,
+      "loss": 0.3201,
+      "step": 3490
+    },
+    {
+      "epoch": 2.9402021336327904,
+      "grad_norm": 0.18885523080825806,
+      "learning_rate": 1.1772814924103649e-08,
+      "loss": 0.3499,
+      "step": 3491
+    },
+    {
+      "epoch": 2.941044357102751,
+      "grad_norm": 0.16768944263458252,
+      "learning_rate": 1.1438978797916888e-08,
+      "loss": 0.2999,
+      "step": 3492
+    },
+    {
+      "epoch": 2.941886580572712,
+      "grad_norm": 0.16767726838588715,
+      "learning_rate": 1.1109938791112328e-08,
+      "loss": 0.3321,
+      "step": 3493
+    },
+    {
+      "epoch": 2.9427288040426727,
+      "grad_norm": 0.1708662509918213,
+      "learning_rate": 1.0785695220035809e-08,
+      "loss": 0.3277,
+      "step": 3494
+    },
+    {
+      "epoch": 2.9435710275126334,
+      "grad_norm": 0.1800270676612854,
+      "learning_rate": 1.0466248396424072e-08,
+      "loss": 0.3193,
+      "step": 3495
+    },
+    {
+      "epoch": 2.944413250982594,
+      "grad_norm": 0.17361044883728027,
+      "learning_rate": 1.0151598627399784e-08,
+      "loss": 0.3189,
+      "step": 3496
+    },
+    {
+      "epoch": 2.945255474452555,
+      "grad_norm": 0.1791125237941742,
+      "learning_rate": 9.841746215474845e-09,
+      "loss": 0.3096,
+      "step": 3497
+    },
+    {
+      "epoch": 2.9460976979225153,
+      "grad_norm": 0.1758781522512436,
+      "learning_rate": 9.536691458548741e-09,
+      "loss": 0.3276,
+      "step": 3498
+    },
+    {
+      "epoch": 2.946939921392476,
+      "grad_norm": 0.16645842790603638,
+      "learning_rate": 9.236434649908532e-09,
+      "loss": 0.3395,
+      "step": 3499
+    },
+    {
+      "epoch": 2.947782144862437,
+      "grad_norm": 0.17300006747245789,
+      "learning_rate": 8.940976078227193e-09,
+      "loss": 0.3613,
+      "step": 3500
+    },
+    {
+      "epoch": 2.9486243683323976,
+      "grad_norm": 0.16995273530483246,
+      "learning_rate": 8.650316027566386e-09,
+      "loss": 0.3028,
+      "step": 3501
+    },
+    {
+      "epoch": 2.9494665918023584,
+      "grad_norm": 0.18239375948905945,
+      "learning_rate": 8.364454777373132e-09,
+      "loss": 0.3571,
+      "step": 3502
+    },
+    {
+      "epoch": 2.9503088152723187,
+      "grad_norm": 0.17194171249866486,
+      "learning_rate": 8.083392602481477e-09,
+      "loss": 0.3263,
+      "step": 3503
+    },
+    {
+      "epoch": 2.9511510387422795,
+      "grad_norm": 0.1683514416217804,
+      "learning_rate": 7.807129773110822e-09,
+      "loss": 0.3286,
+      "step": 3504
+    },
+    {
+      "epoch": 2.9519932622122402,
+      "grad_norm": 0.17565521597862244,
+      "learning_rate": 7.535666554866483e-09,
+      "loss": 0.3318,
+      "step": 3505
+    },
+    {
+      "epoch": 2.952835485682201,
+      "grad_norm": 0.1656530350446701,
+      "learning_rate": 7.269003208740244e-09,
+      "loss": 0.3311,
+      "step": 3506
+    },
+    {
+      "epoch": 2.9536777091521618,
+      "grad_norm": 0.1599675863981247,
+      "learning_rate": 7.007139991108136e-09,
+      "loss": 0.3183,
+      "step": 3507
+    },
+    {
+      "epoch": 2.9545199326221225,
+      "grad_norm": 0.17979347705841064,
+      "learning_rate": 6.750077153731549e-09,
+      "loss": 0.3415,
+      "step": 3508
+    },
+    {
+      "epoch": 2.9553621560920833,
+      "grad_norm": 0.1761392205953598,
+      "learning_rate": 6.497814943756675e-09,
+      "loss": 0.3236,
+      "step": 3509
+    },
+    {
+      "epoch": 2.9562043795620436,
+      "grad_norm": 0.17616896331310272,
+      "learning_rate": 6.25035360371451e-09,
+      "loss": 0.3507,
+      "step": 3510
+    },
+    {
+      "epoch": 2.9570466030320044,
+      "grad_norm": 0.1791347861289978,
+      "learning_rate": 6.00769337151974e-09,
+      "loss": 0.3461,
+      "step": 3511
+    },
+    {
+      "epoch": 2.957888826501965,
+      "grad_norm": 0.17160698771476746,
+      "learning_rate": 5.769834480472414e-09,
+      "loss": 0.3022,
+      "step": 3512
+    },
+    {
+      "epoch": 2.958731049971926,
+      "grad_norm": 0.17749731242656708,
+      "learning_rate": 5.536777159254603e-09,
+      "loss": 0.3267,
+      "step": 3513
+    },
+    {
+      "epoch": 2.9595732734418867,
+      "grad_norm": 0.1847856491804123,
+      "learning_rate": 5.308521631934294e-09,
+      "loss": 0.3781,
+      "step": 3514
+    },
+    {
+      "epoch": 2.960415496911847,
+      "grad_norm": 0.16491389274597168,
+      "learning_rate": 5.08506811796039e-09,
+      "loss": 0.3129,
+      "step": 3515
+    },
+    {
+      "epoch": 2.9612577203818082,
+      "grad_norm": 0.17291562259197235,
+      "learning_rate": 4.866416832167153e-09,
+      "loss": 0.3353,
+      "step": 3516
+    },
+    {
+      "epoch": 2.9620999438517686,
+      "grad_norm": 0.17772185802459717,
+      "learning_rate": 4.652567984770873e-09,
+      "loss": 0.3455,
+      "step": 3517
+    },
+    {
+      "epoch": 2.9629421673217293,
+      "grad_norm": 0.18350841104984283,
+      "learning_rate": 4.443521781370974e-09,
+      "loss": 0.3165,
+      "step": 3518
+    },
+    {
+      "epoch": 2.96378439079169,
+      "grad_norm": 0.18145957589149475,
+      "learning_rate": 4.239278422948911e-09,
+      "loss": 0.359,
+      "step": 3519
+    },
+    {
+      "epoch": 2.964626614261651,
+      "grad_norm": 0.17118234932422638,
+      "learning_rate": 4.0398381058692755e-09,
+      "loss": 0.3054,
+      "step": 3520
+    },
+    {
+      "epoch": 2.9654688377316116,
+      "grad_norm": 0.16678743064403534,
+      "learning_rate": 3.845201021879241e-09,
+      "loss": 0.3201,
+      "step": 3521
+    },
+    {
+      "epoch": 2.966311061201572,
+      "grad_norm": 0.1834692806005478,
+      "learning_rate": 3.655367358106343e-09,
+      "loss": 0.3251,
+      "step": 3522
+    },
+    {
+      "epoch": 2.9671532846715327,
+      "grad_norm": 0.17456305027008057,
+      "learning_rate": 3.470337297062365e-09,
+      "loss": 0.3449,
+      "step": 3523
+    },
+    {
+      "epoch": 2.9679955081414935,
+      "grad_norm": 0.18448472023010254,
+      "learning_rate": 3.290111016638342e-09,
+      "loss": 0.3214,
+      "step": 3524
+    },
+    {
+      "epoch": 2.9688377316114543,
+      "grad_norm": 0.15986642241477966,
+      "learning_rate": 3.1146886901090024e-09,
+      "loss": 0.3225,
+      "step": 3525
+    },
+    {
+      "epoch": 2.969679955081415,
+      "grad_norm": 0.15964114665985107,
+      "learning_rate": 2.9440704861288804e-09,
+      "loss": 0.3156,
+      "step": 3526
+    },
+    {
+      "epoch": 2.970522178551376,
+      "grad_norm": 0.17358997464179993,
+      "learning_rate": 2.7782565687339836e-09,
+      "loss": 0.3403,
+      "step": 3527
+    },
+    {
+      "epoch": 2.9713644020213366,
+      "grad_norm": 0.17132069170475006,
+      "learning_rate": 2.617247097342901e-09,
+      "loss": 0.3299,
+      "step": 3528
+    },
+    {
+      "epoch": 2.972206625491297,
+      "grad_norm": 0.16908591985702515,
+      "learning_rate": 2.461042226752919e-09,
+      "loss": 0.3103,
+      "step": 3529
+    },
+    {
+      "epoch": 2.9730488489612577,
+      "grad_norm": 0.16565345227718353,
+      "learning_rate": 2.3096421071433508e-09,
+      "loss": 0.35,
+      "step": 3530
+    },
+    {
+      "epoch": 2.9738910724312184,
+      "grad_norm": 0.18640905618667603,
+      "learning_rate": 2.1630468840738716e-09,
+      "loss": 0.322,
+      "step": 3531
+    },
+    {
+      "epoch": 2.974733295901179,
+      "grad_norm": 0.1704546958208084,
+      "learning_rate": 2.0212566984845194e-09,
+      "loss": 0.3314,
+      "step": 3532
+    },
+    {
+      "epoch": 2.97557551937114,
+      "grad_norm": 0.16451433300971985,
+      "learning_rate": 1.8842716866956935e-09,
+      "loss": 0.3142,
+      "step": 3533
+    },
+    {
+      "epoch": 2.9764177428411003,
+      "grad_norm": 0.18481574952602386,
+      "learning_rate": 1.7520919804075997e-09,
+      "loss": 0.359,
+      "step": 3534
+    },
+    {
+      "epoch": 2.977259966311061,
+      "grad_norm": 0.16199319064617157,
+      "learning_rate": 1.624717706701917e-09,
+      "loss": 0.2873,
+      "step": 3535
+    },
+    {
+      "epoch": 2.978102189781022,
+      "grad_norm": 0.1770927757024765,
+      "learning_rate": 1.5021489880384653e-09,
+      "loss": 0.348,
+      "step": 3536
+    },
+    {
+      "epoch": 2.9789444132509826,
+      "grad_norm": 0.15922176837921143,
+      "learning_rate": 1.3843859422574269e-09,
+      "loss": 0.3204,
+      "step": 3537
+    },
+    {
+      "epoch": 2.9797866367209433,
+      "grad_norm": 0.17661967873573303,
+      "learning_rate": 1.2714286825793453e-09,
+      "loss": 0.36,
+      "step": 3538
+    },
+    {
+      "epoch": 2.980628860190904,
+      "grad_norm": 0.17170146107673645,
+      "learning_rate": 1.163277317604572e-09,
+      "loss": 0.3373,
+      "step": 3539
+    },
+    {
+      "epoch": 2.981471083660865,
+      "grad_norm": 0.1677255630493164,
+      "learning_rate": 1.0599319513115992e-09,
+      "loss": 0.3,
+      "step": 3540
+    },
+    {
+      "epoch": 2.982313307130825,
+      "grad_norm": 0.1660800576210022,
+      "learning_rate": 9.613926830587262e-10,
+      "loss": 0.2874,
+      "step": 3541
+    },
+    {
+      "epoch": 2.983155530600786,
+      "grad_norm": 0.19125357270240784,
+      "learning_rate": 8.676596075851696e-10,
+      "loss": 0.3728,
+      "step": 3542
+    },
+    {
+      "epoch": 2.9839977540707467,
+      "grad_norm": 0.1575414538383484,
+      "learning_rate": 7.787328150071771e-10,
+      "loss": 0.3193,
+      "step": 3543
+    },
+    {
+      "epoch": 2.9848399775407075,
+      "grad_norm": 0.17667099833488464,
+      "learning_rate": 6.946123908208036e-10,
+      "loss": 0.3592,
+      "step": 3544
+    },
+    {
+      "epoch": 2.9856822010106683,
+      "grad_norm": 0.16778381168842316,
+      "learning_rate": 6.152984159024655e-10,
+      "loss": 0.3025,
+      "step": 3545
+    },
+    {
+      "epoch": 2.9865244244806286,
+      "grad_norm": 0.17686714231967926,
+      "learning_rate": 5.40790966505611e-10,
+      "loss": 0.3393,
+      "step": 3546
+    },
+    {
+      "epoch": 2.98736664795059,
+      "grad_norm": 0.1615176796913147,
+      "learning_rate": 4.710901142634949e-10,
+      "loss": 0.2975,
+      "step": 3547
+    },
+    {
+      "epoch": 2.98820887142055,
+      "grad_norm": 0.17146608233451843,
+      "learning_rate": 4.061959261886239e-10,
+      "loss": 0.3541,
+      "step": 3548
+    },
+    {
+      "epoch": 2.989051094890511,
+      "grad_norm": 0.18955619633197784,
+      "learning_rate": 3.4610846467109106e-10,
+      "loss": 0.3537,
+      "step": 3549
+    },
+    {
+      "epoch": 2.9898933183604717,
+      "grad_norm": 0.17441588640213013,
+      "learning_rate": 2.9082778748135146e-10,
+      "loss": 0.3191,
+      "step": 3550
+    },
+    {
+      "epoch": 2.9907355418304324,
+      "grad_norm": 0.1792432814836502,
+      "learning_rate": 2.403539477668915e-10,
+      "loss": 0.308,
+      "step": 3551
+    },
+    {
+      "epoch": 2.991577765300393,
+      "grad_norm": 0.186254620552063,
+      "learning_rate": 1.9468699405444936e-10,
+      "loss": 0.3485,
+      "step": 3552
+    },
+    {
+      "epoch": 2.9924199887703535,
+      "grad_norm": 0.17337492108345032,
+      "learning_rate": 1.538269702494599e-10,
+      "loss": 0.3396,
+      "step": 3553
+    },
+    {
+      "epoch": 2.9932622122403143,
+      "grad_norm": 0.15600022673606873,
+      "learning_rate": 1.1777391563549956e-10,
+      "loss": 0.2946,
+      "step": 3554
+    },
+    {
+      "epoch": 2.994104435710275,
+      "grad_norm": 0.18066754937171936,
+      "learning_rate": 8.652786487484133e-11,
+      "loss": 0.3236,
+      "step": 3555
+    },
+    {
+      "epoch": 2.994946659180236,
+      "grad_norm": 0.18084211647510529,
+      "learning_rate": 6.008884800845494e-11,
+      "loss": 0.3694,
+      "step": 3556
+    },
+    {
+      "epoch": 2.9957888826501966,
+      "grad_norm": 0.16585589945316315,
+      "learning_rate": 3.8456890455451646e-11,
+      "loss": 0.3137,
+      "step": 3557
+    },
+    {
+      "epoch": 2.9966311061201574,
+      "grad_norm": 0.159687802195549,
+      "learning_rate": 2.1632013013084265e-11,
+      "loss": 0.3002,
+      "step": 3558
+    },
+    {
+      "epoch": 2.997473329590118,
+      "grad_norm": 0.16455858945846558,
+      "learning_rate": 9.614231857302258e-12,
+      "loss": 0.3421,
+      "step": 3559
+    },
+    {
+      "epoch": 2.9983155530600785,
+      "grad_norm": 0.15924279391765594,
+      "learning_rate": 2.403558542196649e-12,
+      "loss": 0.3125,
+      "step": 3560
+    },
+    {
+      "epoch": 2.9991577765300392,
+      "grad_norm": 0.1690434068441391,
+      "learning_rate": 0.0,
+      "loss": 0.335,
+      "step": 3561
+    },
+    {
+      "epoch": 2.9991577765300392,
+      "step": 3561,
+      "total_flos": 1.5509698741010432e+16,
+      "train_loss": 0.37121493492157265,
+      "train_runtime": 260447.3904,
+      "train_samples_per_second": 1.313,
+      "train_steps_per_second": 0.014
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 3561,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 100,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.5509698741010432e+16,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}