diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,9584 @@
+{
+  "best_global_step": 1362,
+  "best_metric": 0.5826267600059509,
+  "best_model_checkpoint": "outputs/checkpoint-1362",
+  "epoch": 2.0,
+  "eval_steps": 500,
+  "global_step": 1362,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0014695077149155032,
+      "grad_norm": 22.578336715698242,
+      "learning_rate": 0.0,
+      "loss": 1.0461,
+      "step": 1
+    },
+    {
+      "epoch": 0.0029390154298310064,
+      "grad_norm": 22.790294647216797,
+      "learning_rate": 1.4705882352941178e-07,
+      "loss": 1.0136,
+      "step": 2
+    },
+    {
+      "epoch": 0.00440852314474651,
+      "grad_norm": 28.41132926940918,
+      "learning_rate": 2.9411764705882356e-07,
+      "loss": 1.2222,
+      "step": 3
+    },
+    {
+      "epoch": 0.005878030859662013,
+      "grad_norm": 25.634981155395508,
+      "learning_rate": 4.411764705882353e-07,
+      "loss": 1.2297,
+      "step": 4
+    },
+    {
+      "epoch": 0.0073475385745775165,
+      "grad_norm": 18.39181900024414,
+      "learning_rate": 5.882352941176471e-07,
+      "loss": 0.9497,
+      "step": 5
+    },
+    {
+      "epoch": 0.00881704628949302,
+      "grad_norm": 28.88524055480957,
+      "learning_rate": 7.352941176470589e-07,
+      "loss": 1.2902,
+      "step": 6
+    },
+    {
+      "epoch": 0.010286554004408524,
+      "grad_norm": 23.266002655029297,
+      "learning_rate": 8.823529411764706e-07,
+      "loss": 1.1428,
+      "step": 7
+    },
+    {
+      "epoch": 0.011756061719324026,
+      "grad_norm": 11.630635261535645,
+      "learning_rate": 1.0294117647058825e-06,
+      "loss": 0.761,
+      "step": 8
+    },
+    {
+      "epoch": 0.01322556943423953,
+      "grad_norm": 11.13463306427002,
+      "learning_rate": 1.1764705882352942e-06,
+      "loss": 0.771,
+      "step": 9
+    },
+    {
+      "epoch": 0.014695077149155033,
+      "grad_norm": 12.781632423400879,
+      "learning_rate": 1.323529411764706e-06,
+      "loss": 0.7198,
+      "step": 10
+    },
+    {
+      "epoch": 0.016164584864070537,
+      "grad_norm": 15.341560363769531,
+      "learning_rate": 1.4705882352941177e-06,
+      "loss": 0.8302,
+      "step": 11
+    },
+    {
+      "epoch": 0.01763409257898604,
+      "grad_norm": 10.644145965576172,
+      "learning_rate": 1.6176470588235297e-06,
+      "loss": 0.7705,
+      "step": 12
+    },
+    {
+      "epoch": 0.019103600293901544,
+      "grad_norm": 11.552690505981445,
+      "learning_rate": 1.7647058823529412e-06,
+      "loss": 0.8254,
+      "step": 13
+    },
+    {
+      "epoch": 0.020573108008817047,
+      "grad_norm": 13.58668327331543,
+      "learning_rate": 1.911764705882353e-06,
+      "loss": 0.7605,
+      "step": 14
+    },
+    {
+      "epoch": 0.02204261572373255,
+      "grad_norm": 12.238264083862305,
+      "learning_rate": 2.058823529411765e-06,
+      "loss": 0.9417,
+      "step": 15
+    },
+    {
+      "epoch": 0.02351212343864805,
+      "grad_norm": 8.899881362915039,
+      "learning_rate": 2.2058823529411767e-06,
+      "loss": 0.6737,
+      "step": 16
+    },
+    {
+      "epoch": 0.024981631153563555,
+      "grad_norm": 12.173700332641602,
+      "learning_rate": 2.3529411764705885e-06,
+      "loss": 0.7554,
+      "step": 17
+    },
+    {
+      "epoch": 0.02645113886847906,
+      "grad_norm": 7.917135238647461,
+      "learning_rate": 2.5e-06,
+      "loss": 0.5946,
+      "step": 18
+    },
+    {
+      "epoch": 0.027920646583394562,
+      "grad_norm": 6.633326530456543,
+      "learning_rate": 2.647058823529412e-06,
+      "loss": 0.5663,
+      "step": 19
+    },
+    {
+      "epoch": 0.029390154298310066,
+      "grad_norm": 10.154461860656738,
+      "learning_rate": 2.7941176470588237e-06,
+      "loss": 0.676,
+      "step": 20
+    },
+    {
+      "epoch": 0.03085966201322557,
+      "grad_norm": 7.985710620880127,
+      "learning_rate": 2.9411764705882355e-06,
+      "loss": 0.6576,
+      "step": 21
+    },
+    {
+      "epoch": 0.03232916972814107,
+      "grad_norm": 7.101803302764893,
+      "learning_rate": 3.0882352941176472e-06,
+      "loss": 0.594,
+      "step": 22
+    },
+    {
+      "epoch": 0.03379867744305658,
+      "grad_norm": 6.337625026702881,
+      "learning_rate": 3.2352941176470594e-06,
+      "loss": 0.5942,
+      "step": 23
+    },
+    {
+      "epoch": 0.03526818515797208,
+      "grad_norm": 4.286959648132324,
+      "learning_rate": 3.3823529411764707e-06,
+      "loss": 0.5817,
+      "step": 24
+    },
+    {
+      "epoch": 0.036737692872887584,
+      "grad_norm": 3.2818689346313477,
+      "learning_rate": 3.5294117647058825e-06,
+      "loss": 0.5307,
+      "step": 25
+    },
+    {
+      "epoch": 0.03820720058780309,
+      "grad_norm": 5.579320907592773,
+      "learning_rate": 3.6764705882352942e-06,
+      "loss": 0.5756,
+      "step": 26
+    },
+    {
+      "epoch": 0.03967670830271859,
+      "grad_norm": 4.195005416870117,
+      "learning_rate": 3.823529411764706e-06,
+      "loss": 0.5842,
+      "step": 27
+    },
+    {
+      "epoch": 0.041146216017634095,
+      "grad_norm": 4.223459243774414,
+      "learning_rate": 3.970588235294117e-06,
+      "loss": 0.5616,
+      "step": 28
+    },
+    {
+      "epoch": 0.0426157237325496,
+      "grad_norm": 3.02809739112854,
+      "learning_rate": 4.11764705882353e-06,
+      "loss": 0.5423,
+      "step": 29
+    },
+    {
+      "epoch": 0.0440852314474651,
+      "grad_norm": 3.734792947769165,
+      "learning_rate": 4.264705882352942e-06,
+      "loss": 0.6108,
+      "step": 30
+    },
+    {
+      "epoch": 0.045554739162380606,
+      "grad_norm": 3.4407219886779785,
+      "learning_rate": 4.411764705882353e-06,
+      "loss": 0.5621,
+      "step": 31
+    },
+    {
+      "epoch": 0.0470242468772961,
+      "grad_norm": 3.1143059730529785,
+      "learning_rate": 4.558823529411764e-06,
+      "loss": 0.5892,
+      "step": 32
+    },
+    {
+      "epoch": 0.048493754592211606,
+      "grad_norm": 4.139797210693359,
+      "learning_rate": 4.705882352941177e-06,
+      "loss": 0.6139,
+      "step": 33
+    },
+    {
+      "epoch": 0.04996326230712711,
+      "grad_norm": 5.330122947692871,
+      "learning_rate": 4.852941176470589e-06,
+      "loss": 0.6804,
+      "step": 34
+    },
+    {
+      "epoch": 0.05143277002204261,
+      "grad_norm": 4.1978535652160645,
+      "learning_rate": 5e-06,
+      "loss": 0.7257,
+      "step": 35
+    },
+    {
+      "epoch": 0.05290227773695812,
+      "grad_norm": 3.80385160446167,
+      "learning_rate": 5.147058823529412e-06,
+      "loss": 0.4641,
+      "step": 36
+    },
+    {
+      "epoch": 0.05437178545187362,
+      "grad_norm": 4.585588455200195,
+      "learning_rate": 5.294117647058824e-06,
+      "loss": 0.5409,
+      "step": 37
+    },
+    {
+      "epoch": 0.055841293166789124,
+      "grad_norm": 4.218822479248047,
+      "learning_rate": 5.441176470588236e-06,
+      "loss": 0.5671,
+      "step": 38
+    },
+    {
+      "epoch": 0.05731080088170463,
+      "grad_norm": 3.764843225479126,
+      "learning_rate": 5.588235294117647e-06,
+      "loss": 0.6958,
+      "step": 39
+    },
+    {
+      "epoch": 0.05878030859662013,
+      "grad_norm": 3.5275750160217285,
+      "learning_rate": 5.735294117647059e-06,
+      "loss": 0.546,
+      "step": 40
+    },
+    {
+      "epoch": 0.060249816311535635,
+      "grad_norm": 3.597670078277588,
+      "learning_rate": 5.882352941176471e-06,
+      "loss": 0.5062,
+      "step": 41
+    },
+    {
+      "epoch": 0.06171932402645114,
+      "grad_norm": 3.4608442783355713,
+      "learning_rate": 6.029411764705883e-06,
+      "loss": 0.5852,
+      "step": 42
+    },
+    {
+      "epoch": 0.06318883174136664,
+      "grad_norm": 3.483283281326294,
+      "learning_rate": 6.1764705882352944e-06,
+      "loss": 0.4127,
+      "step": 43
+    },
+    {
+      "epoch": 0.06465833945628215,
+      "grad_norm": 3.497749090194702,
+      "learning_rate": 6.323529411764706e-06,
+      "loss": 0.4803,
+      "step": 44
+    },
+    {
+      "epoch": 0.06612784717119764,
+      "grad_norm": 3.289926767349243,
+      "learning_rate": 6.470588235294119e-06,
+      "loss": 0.5393,
+      "step": 45
+    },
+    {
+      "epoch": 0.06759735488611315,
+      "grad_norm": 3.5752060413360596,
+      "learning_rate": 6.61764705882353e-06,
+      "loss": 0.4876,
+      "step": 46
+    },
+    {
+      "epoch": 0.06906686260102865,
+      "grad_norm": 3.4211418628692627,
+      "learning_rate": 6.7647058823529414e-06,
+      "loss": 0.6216,
+      "step": 47
+    },
+    {
+      "epoch": 0.07053637031594416,
+      "grad_norm": 3.455683708190918,
+      "learning_rate": 6.911764705882354e-06,
+      "loss": 0.5437,
+      "step": 48
+    },
+    {
+      "epoch": 0.07200587803085966,
+      "grad_norm": 4.6219682693481445,
+      "learning_rate": 7.058823529411765e-06,
+      "loss": 0.5181,
+      "step": 49
+    },
+    {
+      "epoch": 0.07347538574577517,
+      "grad_norm": 4.440234661102295,
+      "learning_rate": 7.205882352941176e-06,
+      "loss": 0.6207,
+      "step": 50
+    },
+    {
+      "epoch": 0.07494489346069066,
+      "grad_norm": 2.749244213104248,
+      "learning_rate": 7.3529411764705884e-06,
+      "loss": 0.4703,
+      "step": 51
+    },
+    {
+      "epoch": 0.07641440117560618,
+      "grad_norm": 3.0789012908935547,
+      "learning_rate": 7.5e-06,
+      "loss": 0.4894,
+      "step": 52
+    },
+    {
+      "epoch": 0.07788390889052167,
+      "grad_norm": 2.9591236114501953,
+      "learning_rate": 7.647058823529413e-06,
+      "loss": 0.4633,
+      "step": 53
+    },
+    {
+      "epoch": 0.07935341660543718,
+      "grad_norm": 3.2681379318237305,
+      "learning_rate": 7.794117647058825e-06,
+      "loss": 0.5108,
+      "step": 54
+    },
+    {
+      "epoch": 0.08082292432035268,
+      "grad_norm": 3.008309841156006,
+      "learning_rate": 7.941176470588235e-06,
+      "loss": 0.5267,
+      "step": 55
+    },
+    {
+      "epoch": 0.08229243203526819,
+      "grad_norm": 3.5393495559692383,
+      "learning_rate": 8.088235294117648e-06,
+      "loss": 0.5221,
+      "step": 56
+    },
+    {
+      "epoch": 0.08376193975018369,
+      "grad_norm": 4.431081295013428,
+      "learning_rate": 8.23529411764706e-06,
+      "loss": 0.6512,
+      "step": 57
+    },
+    {
+      "epoch": 0.0852314474650992,
+      "grad_norm": 3.7774767875671387,
+      "learning_rate": 8.38235294117647e-06,
+      "loss": 0.5913,
+      "step": 58
+    },
+    {
+      "epoch": 0.0867009551800147,
+      "grad_norm": 3.5482702255249023,
+      "learning_rate": 8.529411764705883e-06,
+      "loss": 0.5278,
+      "step": 59
+    },
+    {
+      "epoch": 0.0881704628949302,
+      "grad_norm": 3.515857219696045,
+      "learning_rate": 8.676470588235295e-06,
+      "loss": 0.5809,
+      "step": 60
+    },
+    {
+      "epoch": 0.0896399706098457,
+      "grad_norm": 2.9448065757751465,
+      "learning_rate": 8.823529411764707e-06,
+      "loss": 0.5242,
+      "step": 61
+    },
+    {
+      "epoch": 0.09110947832476121,
+      "grad_norm": 4.011646747589111,
+      "learning_rate": 8.970588235294119e-06,
+      "loss": 0.6344,
+      "step": 62
+    },
+    {
+      "epoch": 0.09257898603967671,
+      "grad_norm": 2.8951897621154785,
+      "learning_rate": 9.117647058823529e-06,
+      "loss": 0.4924,
+      "step": 63
+    },
+    {
+      "epoch": 0.0940484937545922,
+      "grad_norm": 3.22359299659729,
+      "learning_rate": 9.264705882352942e-06,
+      "loss": 0.5625,
+      "step": 64
+    },
+    {
+      "epoch": 0.09551800146950772,
+      "grad_norm": 3.2425975799560547,
+      "learning_rate": 9.411764705882354e-06,
+      "loss": 0.5251,
+      "step": 65
+    },
+    {
+      "epoch": 0.09698750918442321,
+      "grad_norm": 3.1809842586517334,
+      "learning_rate": 9.558823529411764e-06,
+      "loss": 0.4617,
+      "step": 66
+    },
+    {
+      "epoch": 0.09845701689933872,
+      "grad_norm": 3.5795326232910156,
+      "learning_rate": 9.705882352941177e-06,
+      "loss": 0.446,
+      "step": 67
+    },
+    {
+      "epoch": 0.09992652461425422,
+      "grad_norm": 4.245700359344482,
+      "learning_rate": 9.852941176470589e-06,
+      "loss": 0.5272,
+      "step": 68
+    },
+    {
+      "epoch": 0.10139603232916973,
+      "grad_norm": 4.003259658813477,
+      "learning_rate": 1e-05,
+      "loss": 0.4678,
+      "step": 69
+    },
+    {
+      "epoch": 0.10286554004408523,
+      "grad_norm": 3.5002801418304443,
+      "learning_rate": 1.0147058823529413e-05,
+      "loss": 0.5254,
+      "step": 70
+    },
+    {
+      "epoch": 0.10433504775900074,
+      "grad_norm": 2.787693500518799,
+      "learning_rate": 1.0294117647058824e-05,
+      "loss": 0.4538,
+      "step": 71
+    },
+    {
+      "epoch": 0.10580455547391623,
+      "grad_norm": 3.346015691757202,
+      "learning_rate": 1.0441176470588236e-05,
+      "loss": 0.5388,
+      "step": 72
+    },
+    {
+      "epoch": 0.10727406318883174,
+      "grad_norm": 3.3355085849761963,
+      "learning_rate": 1.0588235294117648e-05,
+      "loss": 0.4319,
+      "step": 73
+    },
+    {
+      "epoch": 0.10874357090374724,
+      "grad_norm": 4.5379791259765625,
+      "learning_rate": 1.0735294117647058e-05,
+      "loss": 0.7287,
+      "step": 74
+    },
+    {
+      "epoch": 0.11021307861866275,
+      "grad_norm": 3.1718356609344482,
+      "learning_rate": 1.0882352941176471e-05,
+      "loss": 0.533,
+      "step": 75
+    },
+    {
+      "epoch": 0.11168258633357825,
+      "grad_norm": 3.2101500034332275,
+      "learning_rate": 1.1029411764705883e-05,
+      "loss": 0.5286,
+      "step": 76
+    },
+    {
+      "epoch": 0.11315209404849376,
+      "grad_norm": 3.1723055839538574,
+      "learning_rate": 1.1176470588235295e-05,
+      "loss": 0.5988,
+      "step": 77
+    },
+    {
+      "epoch": 0.11462160176340926,
+      "grad_norm": 3.2591841220855713,
+      "learning_rate": 1.1323529411764707e-05,
+      "loss": 0.4868,
+      "step": 78
+    },
+    {
+      "epoch": 0.11609110947832477,
+      "grad_norm": 3.221942663192749,
+      "learning_rate": 1.1470588235294118e-05,
+      "loss": 0.517,
+      "step": 79
+    },
+    {
+      "epoch": 0.11756061719324026,
+      "grad_norm": 3.9792253971099854,
+      "learning_rate": 1.161764705882353e-05,
+      "loss": 0.4954,
+      "step": 80
+    },
+    {
+      "epoch": 0.11903012490815577,
+      "grad_norm": 3.054436206817627,
+      "learning_rate": 1.1764705882352942e-05,
+      "loss": 0.4435,
+      "step": 81
+    },
+    {
+      "epoch": 0.12049963262307127,
+      "grad_norm": 3.3728082180023193,
+      "learning_rate": 1.1911764705882354e-05,
+      "loss": 0.5807,
+      "step": 82
+    },
+    {
+      "epoch": 0.12196914033798678,
+      "grad_norm": 3.846057653427124,
+      "learning_rate": 1.2058823529411765e-05,
+      "loss": 0.5765,
+      "step": 83
+    },
+    {
+      "epoch": 0.12343864805290228,
+      "grad_norm": 3.1933460235595703,
+      "learning_rate": 1.2205882352941177e-05,
+      "loss": 0.5214,
+      "step": 84
+    },
+    {
+      "epoch": 0.12490815576781777,
+      "grad_norm": 3.793858766555786,
+      "learning_rate": 1.2352941176470589e-05,
+      "loss": 0.5818,
+      "step": 85
+    },
+    {
+      "epoch": 0.12637766348273327,
+      "grad_norm": 4.533748149871826,
+      "learning_rate": 1.25e-05,
+      "loss": 0.6809,
+      "step": 86
+    },
+    {
+      "epoch": 0.1278471711976488,
+      "grad_norm": 2.8002965450286865,
+      "learning_rate": 1.2647058823529412e-05,
+      "loss": 0.3995,
+      "step": 87
+    },
+    {
+      "epoch": 0.1293166789125643,
+      "grad_norm": 3.9104654788970947,
+      "learning_rate": 1.2794117647058822e-05,
+      "loss": 0.521,
+      "step": 88
+    },
+    {
+      "epoch": 0.1307861866274798,
+      "grad_norm": 3.5884015560150146,
+      "learning_rate": 1.2941176470588238e-05,
+      "loss": 0.5017,
+      "step": 89
+    },
+    {
+      "epoch": 0.13225569434239529,
+      "grad_norm": 3.435830593109131,
+      "learning_rate": 1.3088235294117648e-05,
+      "loss": 0.5439,
+      "step": 90
+    },
+    {
+      "epoch": 0.1337252020573108,
+      "grad_norm": 3.4616539478302,
+      "learning_rate": 1.323529411764706e-05,
+      "loss": 0.4221,
+      "step": 91
+    },
+    {
+      "epoch": 0.1351947097722263,
+      "grad_norm": 3.1809732913970947,
+      "learning_rate": 1.3382352941176471e-05,
+      "loss": 0.5477,
+      "step": 92
+    },
+    {
+      "epoch": 0.1366642174871418,
+      "grad_norm": 4.029687881469727,
+      "learning_rate": 1.3529411764705883e-05,
+      "loss": 0.564,
+      "step": 93
+    },
+    {
+      "epoch": 0.1381337252020573,
+      "grad_norm": 3.2925710678100586,
+      "learning_rate": 1.3676470588235296e-05,
+      "loss": 0.478,
+      "step": 94
+    },
+    {
+      "epoch": 0.13960323291697282,
+      "grad_norm": 3.4250714778900146,
+      "learning_rate": 1.3823529411764708e-05,
+      "loss": 0.5198,
+      "step": 95
+    },
+    {
+      "epoch": 0.14107274063188832,
+      "grad_norm": 3.3394150733947754,
+      "learning_rate": 1.3970588235294118e-05,
+      "loss": 0.5684,
+      "step": 96
+    },
+    {
+      "epoch": 0.14254224834680382,
+      "grad_norm": 3.4030964374542236,
+      "learning_rate": 1.411764705882353e-05,
+      "loss": 0.5745,
+      "step": 97
+    },
+    {
+      "epoch": 0.14401175606171931,
+      "grad_norm": 3.406029462814331,
+      "learning_rate": 1.4264705882352942e-05,
+      "loss": 0.4411,
+      "step": 98
+    },
+    {
+      "epoch": 0.14548126377663484,
+      "grad_norm": 3.003526210784912,
+      "learning_rate": 1.4411764705882352e-05,
+      "loss": 0.5679,
+      "step": 99
+    },
+    {
+      "epoch": 0.14695077149155034,
+      "grad_norm": 3.566894769668579,
+      "learning_rate": 1.4558823529411767e-05,
+      "loss": 0.5987,
+      "step": 100
+    },
+    {
+      "epoch": 0.14842027920646583,
+      "grad_norm": 3.1969480514526367,
+      "learning_rate": 1.4705882352941177e-05,
+      "loss": 0.5644,
+      "step": 101
+    },
+    {
+      "epoch": 0.14988978692138133,
+      "grad_norm": 3.9642298221588135,
+      "learning_rate": 1.4852941176470589e-05,
+      "loss": 0.5198,
+      "step": 102
+    },
+    {
+      "epoch": 0.15135929463629685,
+      "grad_norm": 3.1511409282684326,
+      "learning_rate": 1.5e-05,
+      "loss": 0.5587,
+      "step": 103
+    },
+    {
+      "epoch": 0.15282880235121235,
+      "grad_norm": 3.996485471725464,
+      "learning_rate": 1.5147058823529412e-05,
+      "loss": 0.6093,
+      "step": 104
+    },
+    {
+      "epoch": 0.15429831006612785,
+      "grad_norm": 4.1770195960998535,
+      "learning_rate": 1.5294117647058826e-05,
+      "loss": 0.5916,
+      "step": 105
+    },
+    {
+      "epoch": 0.15576781778104334,
+      "grad_norm": 3.429640531539917,
+      "learning_rate": 1.5441176470588237e-05,
+      "loss": 0.4946,
+      "step": 106
+    },
+    {
+      "epoch": 0.15723732549595884,
+      "grad_norm": 4.722433567047119,
+      "learning_rate": 1.558823529411765e-05,
+      "loss": 0.5997,
+      "step": 107
+    },
+    {
+      "epoch": 0.15870683321087437,
+      "grad_norm": 3.5222153663635254,
+      "learning_rate": 1.573529411764706e-05,
+      "loss": 0.5659,
+      "step": 108
+    },
+    {
+      "epoch": 0.16017634092578986,
+      "grad_norm": 3.7971980571746826,
+      "learning_rate": 1.588235294117647e-05,
+      "loss": 0.5285,
+      "step": 109
+    },
+    {
+      "epoch": 0.16164584864070536,
+      "grad_norm": 3.2351953983306885,
+      "learning_rate": 1.6029411764705884e-05,
+      "loss": 0.3978,
+      "step": 110
+    },
+    {
+      "epoch": 0.16311535635562086,
+      "grad_norm": 3.426316499710083,
+      "learning_rate": 1.6176470588235296e-05,
+      "loss": 0.6016,
+      "step": 111
+    },
+    {
+      "epoch": 0.16458486407053638,
+      "grad_norm": 3.185073137283325,
+      "learning_rate": 1.6323529411764708e-05,
+      "loss": 0.5822,
+      "step": 112
+    },
+    {
+      "epoch": 0.16605437178545188,
+      "grad_norm": 4.3362812995910645,
+      "learning_rate": 1.647058823529412e-05,
+      "loss": 0.5252,
+      "step": 113
+    },
+    {
+      "epoch": 0.16752387950036737,
+      "grad_norm": 3.6521811485290527,
+      "learning_rate": 1.6617647058823528e-05,
+      "loss": 0.4782,
+      "step": 114
+    },
+    {
+      "epoch": 0.16899338721528287,
+      "grad_norm": 3.922279119491577,
+      "learning_rate": 1.676470588235294e-05,
+      "loss": 0.5595,
+      "step": 115
+    },
+    {
+      "epoch": 0.1704628949301984,
+      "grad_norm": 3.0742456912994385,
+      "learning_rate": 1.6911764705882355e-05,
+      "loss": 0.5258,
+      "step": 116
+    },
+    {
+      "epoch": 0.1719324026451139,
+      "grad_norm": 3.638026237487793,
+      "learning_rate": 1.7058823529411767e-05,
+      "loss": 0.539,
+      "step": 117
+    },
+    {
+      "epoch": 0.1734019103600294,
+      "grad_norm": 3.462275505065918,
+      "learning_rate": 1.720588235294118e-05,
+      "loss": 0.7588,
+      "step": 118
+    },
+    {
+      "epoch": 0.17487141807494488,
+      "grad_norm": 2.845709800720215,
+      "learning_rate": 1.735294117647059e-05,
+      "loss": 0.4423,
+      "step": 119
+    },
+    {
+      "epoch": 0.1763409257898604,
+      "grad_norm": 3.4925496578216553,
+      "learning_rate": 1.75e-05,
+      "loss": 0.514,
+      "step": 120
+    },
+    {
+      "epoch": 0.1778104335047759,
+      "grad_norm": 2.81315541267395,
+      "learning_rate": 1.7647058823529414e-05,
+      "loss": 0.4339,
+      "step": 121
+    },
+    {
+      "epoch": 0.1792799412196914,
+      "grad_norm": 3.095677137374878,
+      "learning_rate": 1.7794117647058825e-05,
+      "loss": 0.4815,
+      "step": 122
+    },
+    {
+      "epoch": 0.1807494489346069,
+      "grad_norm": 3.100302219390869,
+      "learning_rate": 1.7941176470588237e-05,
+      "loss": 0.4787,
+      "step": 123
+    },
+    {
+      "epoch": 0.18221895664952242,
+      "grad_norm": 3.541459321975708,
+      "learning_rate": 1.808823529411765e-05,
+      "loss": 0.5338,
+      "step": 124
+    },
+    {
+      "epoch": 0.18368846436443792,
+      "grad_norm": 3.2955095767974854,
+      "learning_rate": 1.8235294117647057e-05,
+      "loss": 0.4417,
+      "step": 125
+    },
+    {
+      "epoch": 0.18515797207935342,
+      "grad_norm": 3.6846189498901367,
+      "learning_rate": 1.8382352941176472e-05,
+      "loss": 0.7231,
+      "step": 126
+    },
+    {
+      "epoch": 0.1866274797942689,
+      "grad_norm": 3.9915876388549805,
+      "learning_rate": 1.8529411764705884e-05,
+      "loss": 0.6129,
+      "step": 127
+    },
+    {
+      "epoch": 0.1880969875091844,
+      "grad_norm": 3.5288774967193604,
+      "learning_rate": 1.8676470588235296e-05,
+      "loss": 0.4824,
+      "step": 128
+    },
+    {
+      "epoch": 0.18956649522409993,
+      "grad_norm": 3.445746660232544,
+      "learning_rate": 1.8823529411764708e-05,
+      "loss": 0.564,
+      "step": 129
+    },
+    {
+      "epoch": 0.19103600293901543,
+      "grad_norm": 3.193707227706909,
+      "learning_rate": 1.897058823529412e-05,
+      "loss": 0.3647,
+      "step": 130
+    },
+    {
+      "epoch": 0.19250551065393093,
+      "grad_norm": 2.579481363296509,
+      "learning_rate": 1.9117647058823528e-05,
+      "loss": 0.4823,
+      "step": 131
+    },
+    {
+      "epoch": 0.19397501836884642,
+      "grad_norm": 4.319785118103027,
+      "learning_rate": 1.9264705882352943e-05,
+      "loss": 0.5184,
+      "step": 132
+    },
+    {
+      "epoch": 0.19544452608376195,
+      "grad_norm": 3.671367883682251,
+      "learning_rate": 1.9411764705882355e-05,
+      "loss": 0.5647,
+      "step": 133
+    },
+    {
+      "epoch": 0.19691403379867745,
+      "grad_norm": 3.5213143825531006,
+      "learning_rate": 1.9558823529411766e-05,
+      "loss": 0.4418,
+      "step": 134
+    },
+    {
+      "epoch": 0.19838354151359294,
+      "grad_norm": 5.03970193862915,
+      "learning_rate": 1.9705882352941178e-05,
+      "loss": 0.5535,
+      "step": 135
+    },
+    {
+      "epoch": 0.19985304922850844,
+      "grad_norm": 3.3518738746643066,
+      "learning_rate": 1.9852941176470586e-05,
+      "loss": 0.479,
+      "step": 136
+    },
+    {
+      "epoch": 0.20132255694342396,
+      "grad_norm": 3.7726588249206543,
+      "learning_rate": 2e-05,
+      "loss": 0.6307,
+      "step": 137
+    },
+    {
+      "epoch": 0.20279206465833946,
+      "grad_norm": 2.8609042167663574,
+      "learning_rate": 2.0147058823529413e-05,
+      "loss": 0.4273,
+      "step": 138
+    },
+    {
+      "epoch": 0.20426157237325496,
+      "grad_norm": 3.688164710998535,
+      "learning_rate": 2.0294117647058825e-05,
+      "loss": 0.6667,
+      "step": 139
+    },
+    {
+      "epoch": 0.20573108008817045,
+      "grad_norm": 3.444841146469116,
+      "learning_rate": 2.0441176470588237e-05,
+      "loss": 0.528,
+      "step": 140
+    },
+    {
+      "epoch": 0.20720058780308598,
+      "grad_norm": 2.591963291168213,
+      "learning_rate": 2.058823529411765e-05,
+      "loss": 0.4456,
+      "step": 141
+    },
+    {
+      "epoch": 0.20867009551800147,
+      "grad_norm": 3.3562300205230713,
+      "learning_rate": 2.073529411764706e-05,
+      "loss": 0.4975,
+      "step": 142
+    },
+    {
+      "epoch": 0.21013960323291697,
+      "grad_norm": 2.7175211906433105,
+      "learning_rate": 2.0882352941176472e-05,
+      "loss": 0.4557,
+      "step": 143
+    },
+    {
+      "epoch": 0.21160911094783247,
+      "grad_norm": 3.073939323425293,
+      "learning_rate": 2.1029411764705884e-05,
+      "loss": 0.4363,
+      "step": 144
+    },
+    {
+      "epoch": 0.213078618662748,
+      "grad_norm": 3.202542543411255,
+      "learning_rate": 2.1176470588235296e-05,
+      "loss": 0.5844,
+      "step": 145
+    },
+    {
+      "epoch": 0.2145481263776635,
+      "grad_norm": 3.2442450523376465,
+      "learning_rate": 2.1323529411764707e-05,
+      "loss": 0.4973,
+      "step": 146
+    },
+    {
+      "epoch": 0.216017634092579,
+      "grad_norm": 3.624577045440674,
+      "learning_rate": 2.1470588235294116e-05,
+      "loss": 0.5957,
+      "step": 147
+    },
+    {
+      "epoch": 0.21748714180749448,
+      "grad_norm": 3.1504297256469727,
+      "learning_rate": 2.161764705882353e-05,
+      "loss": 0.5249,
+      "step": 148
+    },
+    {
+      "epoch": 0.21895664952240998,
+      "grad_norm": 3.2034592628479004,
+      "learning_rate": 2.1764705882352943e-05,
+      "loss": 0.6105,
+      "step": 149
+    },
+    {
+      "epoch": 0.2204261572373255,
+      "grad_norm": 3.0447440147399902,
+      "learning_rate": 2.1911764705882354e-05,
+      "loss": 0.5701,
+      "step": 150
+    },
+    {
+      "epoch": 0.221895664952241,
+      "grad_norm": 3.6605937480926514,
+      "learning_rate": 2.2058823529411766e-05,
+      "loss": 0.5491,
+      "step": 151
+    },
+    {
+      "epoch": 0.2233651726671565,
+      "grad_norm": 2.822965383529663,
+      "learning_rate": 2.2205882352941178e-05,
+      "loss": 0.4951,
+      "step": 152
+    },
+    {
+      "epoch": 0.224834680382072,
+      "grad_norm": 3.3248889446258545,
+      "learning_rate": 2.235294117647059e-05,
+      "loss": 0.6472,
+      "step": 153
+    },
+    {
+      "epoch": 0.22630418809698752,
+      "grad_norm": 3.4236090183258057,
+      "learning_rate": 2.25e-05,
+      "loss": 0.5102,
+      "step": 154
+    },
+    {
+      "epoch": 0.22777369581190302,
+      "grad_norm": 2.8016133308410645,
+      "learning_rate": 2.2647058823529413e-05,
+      "loss": 0.4996,
+      "step": 155
+    },
+    {
+      "epoch": 0.2292432035268185,
+      "grad_norm": 2.961437225341797,
+      "learning_rate": 2.2794117647058825e-05,
+      "loss": 0.5497,
+      "step": 156
+    },
+    {
+      "epoch": 0.230712711241734,
+      "grad_norm": 2.4956557750701904,
+      "learning_rate": 2.2941176470588237e-05,
+      "loss": 0.484,
+      "step": 157
+    },
+    {
+      "epoch": 0.23218221895664953,
+      "grad_norm": 3.490821361541748,
+      "learning_rate": 2.308823529411765e-05,
+      "loss": 0.4476,
+      "step": 158
+    },
+    {
+      "epoch": 0.23365172667156503,
+      "grad_norm": 3.864422082901001,
+      "learning_rate": 2.323529411764706e-05,
+      "loss": 0.6115,
+      "step": 159
+    },
+    {
+      "epoch": 0.23512123438648053,
+      "grad_norm": 3.3472821712493896,
+      "learning_rate": 2.3382352941176472e-05,
+      "loss": 0.4655,
+      "step": 160
+    },
+    {
+      "epoch": 0.23659074210139602,
+      "grad_norm": 3.712374210357666,
+      "learning_rate": 2.3529411764705884e-05,
+      "loss": 0.5824,
+      "step": 161
+    },
+    {
+      "epoch": 0.23806024981631155,
+      "grad_norm": 3.10835862159729,
+      "learning_rate": 2.3676470588235295e-05,
+      "loss": 0.5884,
+      "step": 162
+    },
+    {
+      "epoch": 0.23952975753122704,
+      "grad_norm": 3.157806396484375,
+      "learning_rate": 2.3823529411764707e-05,
+      "loss": 0.6148,
+      "step": 163
+    },
+    {
+      "epoch": 0.24099926524614254,
+      "grad_norm": 3.5799741744995117,
+      "learning_rate": 2.397058823529412e-05,
+      "loss": 0.5431,
+      "step": 164
+    },
+    {
+      "epoch": 0.24246877296105804,
+      "grad_norm": 3.1053357124328613,
+      "learning_rate": 2.411764705882353e-05,
+      "loss": 0.5671,
+      "step": 165
+    },
+    {
+      "epoch": 0.24393828067597356,
+      "grad_norm": 2.7615926265716553,
+      "learning_rate": 2.4264705882352942e-05,
+      "loss": 0.4742,
+      "step": 166
+    },
+    {
+      "epoch": 0.24540778839088906,
+      "grad_norm": 3.182527780532837,
+      "learning_rate": 2.4411764705882354e-05,
+      "loss": 0.5403,
+      "step": 167
+    },
+    {
+      "epoch": 0.24687729610580456,
+      "grad_norm": 3.5012803077697754,
+      "learning_rate": 2.4558823529411766e-05,
+      "loss": 0.4603,
+      "step": 168
+    },
+    {
+      "epoch": 0.24834680382072005,
+      "grad_norm": 2.719846725463867,
+      "learning_rate": 2.4705882352941178e-05,
+      "loss": 0.5458,
+      "step": 169
+    },
+    {
+      "epoch": 0.24981631153563555,
+      "grad_norm": 3.413074016571045,
+      "learning_rate": 2.485294117647059e-05,
+      "loss": 0.6409,
+      "step": 170
+    },
+    {
+      "epoch": 0.25128581925055105,
+      "grad_norm": 3.378170967102051,
+      "learning_rate": 2.5e-05,
+      "loss": 0.5118,
+      "step": 171
+    },
+    {
+      "epoch": 0.25275532696546654,
+      "grad_norm": 2.9952166080474854,
+      "learning_rate": 2.5147058823529413e-05,
+      "loss": 0.5447,
+      "step": 172
+    },
+    {
+      "epoch": 0.2542248346803821,
+      "grad_norm": 2.8200743198394775,
+      "learning_rate": 2.5294117647058825e-05,
+      "loss": 0.5631,
+      "step": 173
+    },
+    {
+      "epoch": 0.2556943423952976,
+      "grad_norm": 2.450199604034424,
+      "learning_rate": 2.5441176470588236e-05,
+      "loss": 0.4231,
+      "step": 174
+    },
+    {
+      "epoch": 0.2571638501102131,
+      "grad_norm": 2.3949763774871826,
+      "learning_rate": 2.5588235294117645e-05,
+      "loss": 0.4515,
+      "step": 175
+    },
+    {
+      "epoch": 0.2586333578251286,
+      "grad_norm": 3.7633867263793945,
+      "learning_rate": 2.5735294117647057e-05,
+      "loss": 0.6116,
+      "step": 176
+    },
+    {
+      "epoch": 0.2601028655400441,
+      "grad_norm": 3.0055153369903564,
+      "learning_rate": 2.5882352941176475e-05,
+      "loss": 0.5734,
+      "step": 177
+    },
+    {
+      "epoch": 0.2615723732549596,
+      "grad_norm": 3.38706111907959,
+      "learning_rate": 2.6029411764705887e-05,
+      "loss": 0.6166,
+      "step": 178
+    },
+    {
+      "epoch": 0.2630418809698751,
+      "grad_norm": 2.9373891353607178,
+      "learning_rate": 2.6176470588235295e-05,
+      "loss": 0.5583,
+      "step": 179
+    },
+    {
+      "epoch": 0.26451138868479057,
+      "grad_norm": 3.001413345336914,
+      "learning_rate": 2.6323529411764707e-05,
+      "loss": 0.4753,
+      "step": 180
+    },
+    {
+      "epoch": 0.2659808963997061,
+      "grad_norm": 3.2107958793640137,
+      "learning_rate": 2.647058823529412e-05,
+      "loss": 0.489,
+      "step": 181
+    },
+    {
+      "epoch": 0.2674504041146216,
+      "grad_norm": 2.873209238052368,
+      "learning_rate": 2.661764705882353e-05,
+      "loss": 0.5361,
+      "step": 182
+    },
+    {
+      "epoch": 0.2689199118295371,
+      "grad_norm": 2.6571831703186035,
+      "learning_rate": 2.6764705882352942e-05,
+      "loss": 0.4929,
+      "step": 183
+    },
+    {
+      "epoch": 0.2703894195444526,
+      "grad_norm": 2.9449052810668945,
+      "learning_rate": 2.6911764705882354e-05,
+      "loss": 0.5905,
+      "step": 184
+    },
+    {
+      "epoch": 0.2718589272593681,
+      "grad_norm": 3.7342214584350586,
+      "learning_rate": 2.7058823529411766e-05,
+      "loss": 0.5882,
+      "step": 185
+    },
+    {
+      "epoch": 0.2733284349742836,
+      "grad_norm": 3.3349862098693848,
+      "learning_rate": 2.7205882352941174e-05,
+      "loss": 0.5838,
+      "step": 186
+    },
+    {
+      "epoch": 0.2747979426891991,
+      "grad_norm": 2.5733118057250977,
+      "learning_rate": 2.7352941176470593e-05,
+      "loss": 0.4708,
+      "step": 187
+    },
+    {
+      "epoch": 0.2762674504041146,
+      "grad_norm": 2.8958659172058105,
+      "learning_rate": 2.7500000000000004e-05,
+      "loss": 0.5153,
+      "step": 188
+    },
+    {
+      "epoch": 0.2777369581190301,
+      "grad_norm": 3.176270008087158,
+      "learning_rate": 2.7647058823529416e-05,
+      "loss": 0.5056,
+      "step": 189
+    },
+    {
+      "epoch": 0.27920646583394565,
+      "grad_norm": 3.607558250427246,
+      "learning_rate": 2.7794117647058824e-05,
+      "loss": 0.5886,
+      "step": 190
+    },
+    {
+      "epoch": 0.28067597354886115,
+      "grad_norm": 2.9143762588500977,
+      "learning_rate": 2.7941176470588236e-05,
+      "loss": 0.5612,
+      "step": 191
+    },
+    {
+      "epoch": 0.28214548126377664,
+      "grad_norm": 3.327091932296753,
+      "learning_rate": 2.8088235294117648e-05,
+      "loss": 0.4453,
+      "step": 192
+    },
+    {
+      "epoch": 0.28361498897869214,
+      "grad_norm": 2.931485414505005,
+      "learning_rate": 2.823529411764706e-05,
+      "loss": 0.4678,
+      "step": 193
+    },
+    {
+      "epoch": 0.28508449669360764,
+      "grad_norm": 2.5880470275878906,
+      "learning_rate": 2.838235294117647e-05,
+      "loss": 0.5334,
+      "step": 194
+    },
+    {
+      "epoch": 0.28655400440852313,
+      "grad_norm": 3.2717576026916504,
+      "learning_rate": 2.8529411764705883e-05,
+      "loss": 0.6545,
+      "step": 195
+    },
+    {
+      "epoch": 0.28802351212343863,
+      "grad_norm": 3.946833848953247,
+      "learning_rate": 2.8676470588235295e-05,
+      "loss": 0.6508,
+      "step": 196
+    },
+    {
+      "epoch": 0.2894930198383541,
+      "grad_norm": 3.1211342811584473,
+      "learning_rate": 2.8823529411764703e-05,
+      "loss": 0.4095,
+      "step": 197
+    },
+    {
+      "epoch": 0.2909625275532697,
+      "grad_norm": 3.0994009971618652,
+      "learning_rate": 2.8970588235294122e-05,
+      "loss": 0.5237,
+      "step": 198
+    },
+    {
+      "epoch": 0.2924320352681852,
+      "grad_norm": 2.904717206954956,
+      "learning_rate": 2.9117647058823534e-05,
+      "loss": 0.4408,
+      "step": 199
+    },
+    {
+      "epoch": 0.29390154298310067,
+      "grad_norm": 4.112053871154785,
+      "learning_rate": 2.9264705882352945e-05,
+      "loss": 0.6551,
+      "step": 200
+    },
+    {
+      "epoch": 0.29537105069801617,
+      "grad_norm": 3.9534692764282227,
+      "learning_rate": 2.9411764705882354e-05,
+      "loss": 0.6079,
+      "step": 201
+    },
+    {
+      "epoch": 0.29684055841293167,
+      "grad_norm": 2.8646299839019775,
+      "learning_rate": 2.9558823529411766e-05,
+      "loss": 0.4568,
+      "step": 202
+    },
+    {
+      "epoch": 0.29831006612784716,
+      "grad_norm": 3.326456308364868,
+      "learning_rate": 2.9705882352941177e-05,
+      "loss": 0.5381,
+      "step": 203
+    },
+    {
+      "epoch": 0.29977957384276266,
+      "grad_norm": 3.2934560775756836,
+      "learning_rate": 2.985294117647059e-05,
+      "loss": 0.4478,
+      "step": 204
+    },
+    {
+      "epoch": 0.30124908155767816,
+      "grad_norm": 3.0764334201812744,
+      "learning_rate": 3e-05,
+      "loss": 0.5584,
+      "step": 205
+    },
+    {
+      "epoch": 0.3027185892725937,
+      "grad_norm": 2.4040350914001465,
+      "learning_rate": 3.0147058823529413e-05,
+      "loss": 0.4228,
+      "step": 206
+    },
+    {
+      "epoch": 0.3041880969875092,
+      "grad_norm": 3.1341681480407715,
+      "learning_rate": 3.0294117647058824e-05,
+      "loss": 0.5952,
+      "step": 207
+    },
+    {
+      "epoch": 0.3056576047024247,
+      "grad_norm": 2.556406259536743,
+      "learning_rate": 3.0441176470588233e-05,
+      "loss": 0.5482,
+      "step": 208
+    },
+    {
+      "epoch": 0.3071271124173402,
+      "grad_norm": 2.7843801975250244,
+      "learning_rate": 3.058823529411765e-05,
+      "loss": 0.5153,
+      "step": 209
+    },
+    {
+      "epoch": 0.3085966201322557,
+      "grad_norm": 3.5703351497650146,
+      "learning_rate": 3.073529411764706e-05,
+      "loss": 0.5735,
+      "step": 210
+    },
+    {
+      "epoch": 0.3100661278471712,
+      "grad_norm": 3.8750274181365967,
+      "learning_rate": 3.0882352941176475e-05,
+      "loss": 0.5595,
+      "step": 211
+    },
+    {
+      "epoch": 0.3115356355620867,
+      "grad_norm": 3.2917017936706543,
+      "learning_rate": 3.1029411764705886e-05,
+      "loss": 0.4831,
+      "step": 212
+    },
+    {
+      "epoch": 0.3130051432770022,
+      "grad_norm": 3.376647710800171,
+      "learning_rate": 3.11764705882353e-05,
+      "loss": 0.6932,
+      "step": 213
+    },
+    {
+      "epoch": 0.3144746509919177,
+      "grad_norm": 3.5780632495880127,
+      "learning_rate": 3.132352941176471e-05,
+      "loss": 0.539,
+      "step": 214
+    },
+    {
+      "epoch": 0.31594415870683323,
+      "grad_norm": 3.5975916385650635,
+      "learning_rate": 3.147058823529412e-05,
+      "loss": 0.5253,
+      "step": 215
+    },
+    {
+      "epoch": 0.31741366642174873,
+      "grad_norm": 3.5709407329559326,
+      "learning_rate": 3.161764705882353e-05,
+      "loss": 0.6725,
+      "step": 216
+    },
+    {
+      "epoch": 0.3188831741366642,
+      "grad_norm": 3.1600329875946045,
+      "learning_rate": 3.176470588235294e-05,
+      "loss": 0.6515,
+      "step": 217
+    },
+    {
+      "epoch": 0.3203526818515797,
+      "grad_norm": 3.6528427600860596,
+      "learning_rate": 3.191176470588235e-05,
+      "loss": 0.6085,
+      "step": 218
+    },
+    {
+      "epoch": 0.3218221895664952,
+      "grad_norm": 3.4629173278808594,
+      "learning_rate": 3.205882352941177e-05,
+      "loss": 0.6383,
+      "step": 219
+    },
+    {
+      "epoch": 0.3232916972814107,
+      "grad_norm": 3.09975004196167,
+      "learning_rate": 3.220588235294118e-05,
+      "loss": 0.6611,
+      "step": 220
+    },
+    {
+      "epoch": 0.3247612049963262,
+      "grad_norm": 3.4542508125305176,
+      "learning_rate": 3.235294117647059e-05,
+      "loss": 0.654,
+      "step": 221
+    },
+    {
+      "epoch": 0.3262307127112417,
+      "grad_norm": 3.4997575283050537,
+      "learning_rate": 3.2500000000000004e-05,
+      "loss": 0.7129,
+      "step": 222
+    },
+    {
+      "epoch": 0.32770022042615726,
+      "grad_norm": 2.7205755710601807,
+      "learning_rate": 3.2647058823529416e-05,
+      "loss": 0.4552,
+      "step": 223
+    },
+    {
+      "epoch": 0.32916972814107276,
+      "grad_norm": 3.345202684402466,
+      "learning_rate": 3.279411764705883e-05,
+      "loss": 0.6208,
+      "step": 224
+    },
+    {
+      "epoch": 0.33063923585598826,
+      "grad_norm": 2.565737009048462,
+      "learning_rate": 3.294117647058824e-05,
+      "loss": 0.4836,
+      "step": 225
+    },
+    {
+      "epoch": 0.33210874357090375,
+      "grad_norm": 3.0795998573303223,
+      "learning_rate": 3.308823529411765e-05,
+      "loss": 0.6143,
+      "step": 226
+    },
+    {
+      "epoch": 0.33357825128581925,
+      "grad_norm": 3.0101351737976074,
+      "learning_rate": 3.3235294117647056e-05,
+      "loss": 0.5109,
+      "step": 227
+    },
+    {
+      "epoch": 0.33504775900073475,
+      "grad_norm": 3.604487419128418,
+      "learning_rate": 3.338235294117647e-05,
+      "loss": 0.6052,
+      "step": 228
+    },
+    {
+      "epoch": 0.33651726671565024,
+      "grad_norm": 3.319622278213501,
+      "learning_rate": 3.352941176470588e-05,
+      "loss": 0.5324,
+      "step": 229
+    },
+    {
+      "epoch": 0.33798677443056574,
+      "grad_norm": 2.7687244415283203,
+      "learning_rate": 3.36764705882353e-05,
+      "loss": 0.4414,
+      "step": 230
+    },
+    {
+      "epoch": 0.33945628214548124,
+      "grad_norm": 2.7827277183532715,
+      "learning_rate": 3.382352941176471e-05,
+      "loss": 0.4949,
+      "step": 231
+    },
+    {
+      "epoch": 0.3409257898603968,
+      "grad_norm": 3.6244659423828125,
+      "learning_rate": 3.397058823529412e-05,
+      "loss": 0.5701,
+      "step": 232
+    },
+    {
+      "epoch": 0.3423952975753123,
+      "grad_norm": 2.823563575744629,
+      "learning_rate": 3.411764705882353e-05,
+      "loss": 0.6129,
+      "step": 233
+    },
+    {
+      "epoch": 0.3438648052902278,
+      "grad_norm": 2.6408374309539795,
+      "learning_rate": 3.4264705882352945e-05,
+      "loss": 0.5112,
+      "step": 234
+    },
+    {
+      "epoch": 0.3453343130051433,
+      "grad_norm": 3.164153814315796,
+      "learning_rate": 3.441176470588236e-05,
+      "loss": 0.6774,
+      "step": 235
+    },
+    {
+      "epoch": 0.3468038207200588,
+      "grad_norm": 3.2631454467773438,
+      "learning_rate": 3.455882352941177e-05,
+      "loss": 0.4671,
+      "step": 236
+    },
+    {
+      "epoch": 0.34827332843497427,
+      "grad_norm": 2.908346176147461,
+      "learning_rate": 3.470588235294118e-05,
+      "loss": 0.6594,
+      "step": 237
+    },
+    {
+      "epoch": 0.34974283614988977,
+      "grad_norm": 3.2295093536376953,
+      "learning_rate": 3.4852941176470585e-05,
+      "loss": 0.6931,
+      "step": 238
+    },
+    {
+      "epoch": 0.35121234386480527,
+      "grad_norm": 3.515854835510254,
+      "learning_rate": 3.5e-05,
+      "loss": 0.5779,
+      "step": 239
+    },
+    {
+      "epoch": 0.3526818515797208,
+      "grad_norm": 3.1064531803131104,
+      "learning_rate": 3.514705882352941e-05,
+      "loss": 0.5536,
+      "step": 240
+    },
+    {
+      "epoch": 0.3541513592946363,
+      "grad_norm": 2.755269765853882,
+      "learning_rate": 3.529411764705883e-05,
+      "loss": 0.5057,
+      "step": 241
+    },
+    {
+      "epoch": 0.3556208670095518,
+      "grad_norm": 3.360236644744873,
+      "learning_rate": 3.544117647058824e-05,
+      "loss": 0.5552,
+      "step": 242
+    },
+    {
+      "epoch": 0.3570903747244673,
+      "grad_norm": 3.122957229614258,
+      "learning_rate": 3.558823529411765e-05,
+      "loss": 0.6216,
+      "step": 243
+    },
+    {
+      "epoch": 0.3585598824393828,
+      "grad_norm": 3.41726016998291,
+      "learning_rate": 3.573529411764706e-05,
+      "loss": 0.7104,
+      "step": 244
+    },
+    {
+      "epoch": 0.3600293901542983,
+      "grad_norm": 2.9083242416381836,
+      "learning_rate": 3.5882352941176474e-05,
+      "loss": 0.5609,
+      "step": 245
+    },
+    {
+      "epoch": 0.3614988978692138,
+      "grad_norm": 2.8559906482696533,
+      "learning_rate": 3.6029411764705886e-05,
+      "loss": 0.5991,
+      "step": 246
+    },
+    {
+      "epoch": 0.3629684055841293,
+      "grad_norm": 2.4915928840637207,
+      "learning_rate": 3.61764705882353e-05,
+      "loss": 0.4951,
+      "step": 247
+    },
+    {
+      "epoch": 0.36443791329904485,
+      "grad_norm": 3.297513484954834,
+      "learning_rate": 3.632352941176471e-05,
+      "loss": 0.4639,
+      "step": 248
+    },
+    {
+      "epoch": 0.36590742101396034,
+      "grad_norm": 2.8595054149627686,
+      "learning_rate": 3.6470588235294114e-05,
+      "loss": 0.5286,
+      "step": 249
+    },
+    {
+      "epoch": 0.36737692872887584,
+      "grad_norm": 3.114201307296753,
+      "learning_rate": 3.6617647058823526e-05,
+      "loss": 0.556,
+      "step": 250
+    },
+    {
+      "epoch": 0.36884643644379134,
+      "grad_norm": 3.1130385398864746,
+      "learning_rate": 3.6764705882352945e-05,
+      "loss": 0.4841,
+      "step": 251
+    },
+    {
+      "epoch": 0.37031594415870683,
+      "grad_norm": 2.736334800720215,
+      "learning_rate": 3.6911764705882356e-05,
+      "loss": 0.5214,
+      "step": 252
+    },
+    {
+      "epoch": 0.37178545187362233,
+      "grad_norm": 3.4331860542297363,
+      "learning_rate": 3.705882352941177e-05,
+      "loss": 0.6258,
+      "step": 253
+    },
+    {
+      "epoch": 0.3732549595885378,
+      "grad_norm": 2.8038713932037354,
+      "learning_rate": 3.720588235294118e-05,
+      "loss": 0.5936,
+      "step": 254
+    },
+    {
+      "epoch": 0.3747244673034533,
+      "grad_norm": 3.2930872440338135,
+      "learning_rate": 3.735294117647059e-05,
+      "loss": 0.5234,
+      "step": 255
+    },
+    {
+      "epoch": 0.3761939750183688,
+      "grad_norm": 3.6463680267333984,
+      "learning_rate": 3.7500000000000003e-05,
+      "loss": 0.6615,
+      "step": 256
+    },
+    {
+      "epoch": 0.3776634827332844,
+      "grad_norm": 2.9523704051971436,
+      "learning_rate": 3.7647058823529415e-05,
+      "loss": 0.565,
+      "step": 257
+    },
+    {
+      "epoch": 0.37913299044819987,
+      "grad_norm": 3.490185499191284,
+      "learning_rate": 3.779411764705883e-05,
+      "loss": 0.6828,
+      "step": 258
+    },
+    {
+      "epoch": 0.38060249816311537,
+      "grad_norm": 2.8911149501800537,
+      "learning_rate": 3.794117647058824e-05,
+      "loss": 0.5508,
+      "step": 259
+    },
+    {
+      "epoch": 0.38207200587803086,
+      "grad_norm": 3.775752544403076,
+      "learning_rate": 3.8088235294117644e-05,
+      "loss": 0.8089,
+      "step": 260
+    },
+    {
+      "epoch": 0.38354151359294636,
+      "grad_norm": 2.331620454788208,
+      "learning_rate": 3.8235294117647055e-05,
+      "loss": 0.422,
+      "step": 261
+    },
+    {
+      "epoch": 0.38501102130786186,
+      "grad_norm": 2.896380662918091,
+      "learning_rate": 3.8382352941176474e-05,
+      "loss": 0.6068,
+      "step": 262
+    },
+    {
+      "epoch": 0.38648052902277735,
+      "grad_norm": 2.7637434005737305,
+      "learning_rate": 3.8529411764705886e-05,
+      "loss": 0.5533,
+      "step": 263
+    },
+    {
+      "epoch": 0.38795003673769285,
+      "grad_norm": 2.8808846473693848,
+      "learning_rate": 3.86764705882353e-05,
+      "loss": 0.5322,
+      "step": 264
+    },
+    {
+      "epoch": 0.3894195444526084,
+      "grad_norm": 3.3949851989746094,
+      "learning_rate": 3.882352941176471e-05,
+      "loss": 0.5741,
+      "step": 265
+    },
+    {
+      "epoch": 0.3908890521675239,
+      "grad_norm": 6.345247268676758,
+      "learning_rate": 3.897058823529412e-05,
+      "loss": 0.6859,
+      "step": 266
+    },
+    {
+      "epoch": 0.3923585598824394,
+      "grad_norm": 2.397280216217041,
+      "learning_rate": 3.911764705882353e-05,
+      "loss": 0.6376,
+      "step": 267
+    },
+    {
+      "epoch": 0.3938280675973549,
+      "grad_norm": 2.9388270378112793,
+      "learning_rate": 3.9264705882352945e-05,
+      "loss": 0.4369,
+      "step": 268
+    },
+    {
+      "epoch": 0.3952975753122704,
+      "grad_norm": 3.113617181777954,
+      "learning_rate": 3.9411764705882356e-05,
+      "loss": 0.7229,
+      "step": 269
+    },
+    {
+      "epoch": 0.3967670830271859,
+      "grad_norm": 2.8402743339538574,
+      "learning_rate": 3.955882352941177e-05,
+      "loss": 0.5108,
+      "step": 270
+    },
+    {
+      "epoch": 0.3982365907421014,
+      "grad_norm": 2.757765531539917,
+      "learning_rate": 3.970588235294117e-05,
+      "loss": 0.4669,
+      "step": 271
+    },
+    {
+      "epoch": 0.3997060984570169,
+      "grad_norm": 3.26387882232666,
+      "learning_rate": 3.985294117647059e-05,
+      "loss": 0.6565,
+      "step": 272
+    },
+    {
+      "epoch": 0.4011756061719324,
+      "grad_norm": 3.0189812183380127,
+      "learning_rate": 4e-05,
+      "loss": 0.6389,
+      "step": 273
+    },
+    {
+      "epoch": 0.4026451138868479,
+      "grad_norm": 3.361266613006592,
+      "learning_rate": 4.0147058823529415e-05,
+      "loss": 0.4381,
+      "step": 274
+    },
+    {
+      "epoch": 0.4041146216017634,
+      "grad_norm": 3.5814902782440186,
+      "learning_rate": 4.029411764705883e-05,
+      "loss": 0.6312,
+      "step": 275
+    },
+    {
+      "epoch": 0.4055841293166789,
+      "grad_norm": 2.5322272777557373,
+      "learning_rate": 4.044117647058824e-05,
+      "loss": 0.4938,
+      "step": 276
+    },
+    {
+      "epoch": 0.4070536370315944,
+      "grad_norm": 2.8250925540924072,
+      "learning_rate": 4.058823529411765e-05,
+      "loss": 0.4793,
+      "step": 277
+    },
+    {
+      "epoch": 0.4085231447465099,
+      "grad_norm": 2.933051824569702,
+      "learning_rate": 4.073529411764706e-05,
+      "loss": 0.6603,
+      "step": 278
+    },
+    {
+      "epoch": 0.4099926524614254,
+      "grad_norm": 2.563755989074707,
+      "learning_rate": 4.0882352941176474e-05,
+      "loss": 0.5593,
+      "step": 279
+    },
+    {
+      "epoch": 0.4114621601763409,
+      "grad_norm": 4.022047519683838,
+      "learning_rate": 4.1029411764705886e-05,
+      "loss": 0.6655,
+      "step": 280
+    },
+    {
+      "epoch": 0.4129316678912564,
+      "grad_norm": 2.368408679962158,
+      "learning_rate": 4.11764705882353e-05,
+      "loss": 0.5346,
+      "step": 281
+    },
+    {
+      "epoch": 0.41440117560617196,
+      "grad_norm": 2.8808212280273438,
+      "learning_rate": 4.13235294117647e-05,
+      "loss": 0.6577,
+      "step": 282
+    },
+    {
+      "epoch": 0.41587068332108745,
+      "grad_norm": 3.438518762588501,
+      "learning_rate": 4.147058823529412e-05,
+      "loss": 0.5975,
+      "step": 283
+    },
+    {
+      "epoch": 0.41734019103600295,
+      "grad_norm": 3.1126911640167236,
+      "learning_rate": 4.161764705882353e-05,
+      "loss": 0.5632,
+      "step": 284
+    },
+    {
+      "epoch": 0.41880969875091845,
+      "grad_norm": 2.6921920776367188,
+      "learning_rate": 4.1764705882352944e-05,
+      "loss": 0.605,
+      "step": 285
+    },
+    {
+      "epoch": 0.42027920646583394,
+      "grad_norm": 2.8262081146240234,
+      "learning_rate": 4.1911764705882356e-05,
+      "loss": 0.691,
+      "step": 286
+    },
+    {
+      "epoch": 0.42174871418074944,
+      "grad_norm": 3.654421806335449,
+      "learning_rate": 4.205882352941177e-05,
+      "loss": 0.7957,
+      "step": 287
+    },
+    {
+      "epoch": 0.42321822189566494,
+      "grad_norm": 2.444572687149048,
+      "learning_rate": 4.220588235294118e-05,
+      "loss": 0.5601,
+      "step": 288
+    },
+    {
+      "epoch": 0.42468772961058043,
+      "grad_norm": 2.6849148273468018,
+      "learning_rate": 4.235294117647059e-05,
+      "loss": 0.5602,
+      "step": 289
+    },
+    {
+      "epoch": 0.426157237325496,
+      "grad_norm": 3.1374027729034424,
+      "learning_rate": 4.25e-05,
+      "loss": 0.5926,
+      "step": 290
+    },
+    {
+      "epoch": 0.4276267450404115,
+      "grad_norm": 2.706740617752075,
+      "learning_rate": 4.2647058823529415e-05,
+      "loss": 0.6482,
+      "step": 291
+    },
+    {
+      "epoch": 0.429096252755327,
+      "grad_norm": 3.215224027633667,
+      "learning_rate": 4.2794117647058827e-05,
+      "loss": 0.615,
+      "step": 292
+    },
+    {
+      "epoch": 0.4305657604702425,
+      "grad_norm": 2.5421245098114014,
+      "learning_rate": 4.294117647058823e-05,
+      "loss": 0.5802,
+      "step": 293
+    },
+    {
+      "epoch": 0.432035268185158,
+      "grad_norm": 2.6008574962615967,
+      "learning_rate": 4.308823529411765e-05,
+      "loss": 0.5417,
+      "step": 294
+    },
+    {
+      "epoch": 0.43350477590007347,
+      "grad_norm": 2.7563812732696533,
+      "learning_rate": 4.323529411764706e-05,
+      "loss": 0.5405,
+      "step": 295
+    },
+    {
+      "epoch": 0.43497428361498897,
+      "grad_norm": 3.032604694366455,
+      "learning_rate": 4.3382352941176474e-05,
+      "loss": 0.4956,
+      "step": 296
+    },
+    {
+      "epoch": 0.43644379132990446,
+      "grad_norm": 3.1097381114959717,
+      "learning_rate": 4.3529411764705885e-05,
+      "loss": 0.6136,
+      "step": 297
+    },
+    {
+      "epoch": 0.43791329904481996,
+      "grad_norm": 2.5641071796417236,
+      "learning_rate": 4.36764705882353e-05,
+      "loss": 0.4774,
+      "step": 298
+    },
+    {
+      "epoch": 0.4393828067597355,
+      "grad_norm": 2.95414400100708,
+      "learning_rate": 4.382352941176471e-05,
+      "loss": 0.5811,
+      "step": 299
+    },
+    {
+      "epoch": 0.440852314474651,
+      "grad_norm": 2.7911477088928223,
+      "learning_rate": 4.397058823529412e-05,
+      "loss": 0.5519,
+      "step": 300
+    },
+    {
+      "epoch": 0.4423218221895665,
+      "grad_norm": 3.4301278591156006,
+      "learning_rate": 4.411764705882353e-05,
+      "loss": 0.734,
+      "step": 301
+    },
+    {
+      "epoch": 0.443791329904482,
+      "grad_norm": 3.060013771057129,
+      "learning_rate": 4.4264705882352944e-05,
+      "loss": 0.4575,
+      "step": 302
+    },
+    {
+      "epoch": 0.4452608376193975,
+      "grad_norm": 2.912816286087036,
+      "learning_rate": 4.4411764705882356e-05,
+      "loss": 0.7212,
+      "step": 303
+    },
+    {
+      "epoch": 0.446730345334313,
+      "grad_norm": 2.5897419452667236,
+      "learning_rate": 4.455882352941177e-05,
+      "loss": 0.6676,
+      "step": 304
+    },
+    {
+      "epoch": 0.4481998530492285,
+      "grad_norm": 2.570291519165039,
+      "learning_rate": 4.470588235294118e-05,
+      "loss": 0.4783,
+      "step": 305
+    },
+    {
+      "epoch": 0.449669360764144,
+      "grad_norm": 2.1856844425201416,
+      "learning_rate": 4.485294117647059e-05,
+      "loss": 0.392,
+      "step": 306
+    },
+    {
+      "epoch": 0.45113886847905954,
+      "grad_norm": 2.918060541152954,
+      "learning_rate": 4.5e-05,
+      "loss": 0.6983,
+      "step": 307
+    },
+    {
+      "epoch": 0.45260837619397504,
+      "grad_norm": 2.4142963886260986,
+      "learning_rate": 4.5147058823529415e-05,
+      "loss": 0.5644,
+      "step": 308
+    },
+    {
+      "epoch": 0.45407788390889053,
+      "grad_norm": 2.714462995529175,
+      "learning_rate": 4.5294117647058826e-05,
+      "loss": 0.6435,
+      "step": 309
+    },
+    {
+      "epoch": 0.45554739162380603,
+      "grad_norm": 2.725534677505493,
+      "learning_rate": 4.544117647058824e-05,
+      "loss": 0.4935,
+      "step": 310
+    },
+    {
+      "epoch": 0.4570168993387215,
+      "grad_norm": 2.608978509902954,
+      "learning_rate": 4.558823529411765e-05,
+      "loss": 0.6736,
+      "step": 311
+    },
+    {
+      "epoch": 0.458486407053637,
+      "grad_norm": 3.3339529037475586,
+      "learning_rate": 4.573529411764706e-05,
+      "loss": 0.8257,
+      "step": 312
+    },
+    {
+      "epoch": 0.4599559147685525,
+      "grad_norm": 3.411391019821167,
+      "learning_rate": 4.588235294117647e-05,
+      "loss": 0.688,
+      "step": 313
+    },
+    {
+      "epoch": 0.461425422483468,
+      "grad_norm": 2.3950774669647217,
+      "learning_rate": 4.6029411764705885e-05,
+      "loss": 0.5289,
+      "step": 314
+    },
+    {
+      "epoch": 0.4628949301983835,
+      "grad_norm": 2.6945137977600098,
+      "learning_rate": 4.61764705882353e-05,
+      "loss": 0.6017,
+      "step": 315
+    },
+    {
+      "epoch": 0.46436443791329907,
+      "grad_norm": 2.063464879989624,
+      "learning_rate": 4.632352941176471e-05,
+      "loss": 0.3948,
+      "step": 316
+    },
+    {
+      "epoch": 0.46583394562821456,
+      "grad_norm": 2.8425235748291016,
+      "learning_rate": 4.647058823529412e-05,
+      "loss": 0.6894,
+      "step": 317
+    },
+    {
+      "epoch": 0.46730345334313006,
+      "grad_norm": 2.727858781814575,
+      "learning_rate": 4.661764705882353e-05,
+      "loss": 0.5763,
+      "step": 318
+    },
+    {
+      "epoch": 0.46877296105804556,
+      "grad_norm": 2.5340576171875,
+      "learning_rate": 4.6764705882352944e-05,
+      "loss": 0.5808,
+      "step": 319
+    },
+    {
+      "epoch": 0.47024246877296105,
+      "grad_norm": 2.1478378772735596,
+      "learning_rate": 4.6911764705882356e-05,
+      "loss": 0.5717,
+      "step": 320
+    },
+    {
+      "epoch": 0.47171197648787655,
+      "grad_norm": 2.880711317062378,
+      "learning_rate": 4.705882352941177e-05,
+      "loss": 0.5937,
+      "step": 321
+    },
+    {
+      "epoch": 0.47318148420279205,
+      "grad_norm": 2.9537229537963867,
+      "learning_rate": 4.720588235294118e-05,
+      "loss": 0.6531,
+      "step": 322
+    },
+    {
+      "epoch": 0.47465099191770754,
+      "grad_norm": 2.751915216445923,
+      "learning_rate": 4.735294117647059e-05,
+      "loss": 0.6433,
+      "step": 323
+    },
+    {
+      "epoch": 0.4761204996326231,
+      "grad_norm": 2.478581428527832,
+      "learning_rate": 4.75e-05,
+      "loss": 0.5794,
+      "step": 324
+    },
+    {
+      "epoch": 0.4775900073475386,
+      "grad_norm": 2.1114606857299805,
+      "learning_rate": 4.7647058823529414e-05,
+      "loss": 0.3648,
+      "step": 325
+    },
+    {
+      "epoch": 0.4790595150624541,
+      "grad_norm": 2.9210312366485596,
+      "learning_rate": 4.7794117647058826e-05,
+      "loss": 0.5878,
+      "step": 326
+    },
+    {
+      "epoch": 0.4805290227773696,
+      "grad_norm": 2.3768012523651123,
+      "learning_rate": 4.794117647058824e-05,
+      "loss": 0.5162,
+      "step": 327
+    },
+    {
+      "epoch": 0.4819985304922851,
+      "grad_norm": 2.504443407058716,
+      "learning_rate": 4.808823529411765e-05,
+      "loss": 0.6139,
+      "step": 328
+    },
+    {
+      "epoch": 0.4834680382072006,
+      "grad_norm": 2.818042755126953,
+      "learning_rate": 4.823529411764706e-05,
+      "loss": 0.6788,
+      "step": 329
+    },
+    {
+      "epoch": 0.4849375459221161,
+      "grad_norm": 3.370159864425659,
+      "learning_rate": 4.838235294117647e-05,
+      "loss": 0.7414,
+      "step": 330
+    },
+    {
+      "epoch": 0.4864070536370316,
+      "grad_norm": 2.7343268394470215,
+      "learning_rate": 4.8529411764705885e-05,
+      "loss": 0.6546,
+      "step": 331
+    },
+    {
+      "epoch": 0.4878765613519471,
+      "grad_norm": 2.7900588512420654,
+      "learning_rate": 4.86764705882353e-05,
+      "loss": 0.5745,
+      "step": 332
+    },
+    {
+      "epoch": 0.4893460690668626,
+      "grad_norm": 2.438570022583008,
+      "learning_rate": 4.882352941176471e-05,
+      "loss": 0.6329,
+      "step": 333
+    },
+    {
+      "epoch": 0.4908155767817781,
+      "grad_norm": 2.42809796333313,
+      "learning_rate": 4.897058823529412e-05,
+      "loss": 0.6054,
+      "step": 334
+    },
+    {
+      "epoch": 0.4922850844966936,
+      "grad_norm": 2.933128833770752,
+      "learning_rate": 4.911764705882353e-05,
+      "loss": 0.6667,
+      "step": 335
+    },
+    {
+      "epoch": 0.4937545922116091,
+      "grad_norm": 2.8789055347442627,
+      "learning_rate": 4.9264705882352944e-05,
+      "loss": 0.6352,
+      "step": 336
+    },
+    {
+      "epoch": 0.4952240999265246,
+      "grad_norm": 2.602020502090454,
+      "learning_rate": 4.9411764705882355e-05,
+      "loss": 0.5764,
+      "step": 337
+    },
+    {
+      "epoch": 0.4966936076414401,
+      "grad_norm": 3.5491220951080322,
+      "learning_rate": 4.955882352941177e-05,
+      "loss": 0.6715,
+      "step": 338
+    },
+    {
+      "epoch": 0.4981631153563556,
+      "grad_norm": 2.516988515853882,
+      "learning_rate": 4.970588235294118e-05,
+      "loss": 0.6705,
+      "step": 339
+    },
+    {
+      "epoch": 0.4996326230712711,
+      "grad_norm": 3.1182806491851807,
+      "learning_rate": 4.985294117647059e-05,
+      "loss": 0.6958,
+      "step": 340
+    },
+    {
+      "epoch": 0.5011021307861866,
+      "grad_norm": 2.866764783859253,
+      "learning_rate": 5e-05,
+      "loss": 0.6331,
+      "step": 341
+    },
+    {
+      "epoch": 0.5025716385011021,
+      "grad_norm": 2.47072696685791,
+      "learning_rate": 4.999998682450723e-05,
+      "loss": 0.6247,
+      "step": 342
+    },
+    {
+      "epoch": 0.5040411462160176,
+      "grad_norm": 2.377047300338745,
+      "learning_rate": 4.999994729804278e-05,
+      "loss": 0.5233,
+      "step": 343
+    },
+    {
+      "epoch": 0.5055106539309331,
+      "grad_norm": 2.81021785736084,
+      "learning_rate": 4.999988142064832e-05,
+      "loss": 0.5731,
+      "step": 344
+    },
+    {
+      "epoch": 0.5069801616458487,
+      "grad_norm": 2.468865156173706,
+      "learning_rate": 4.99997891923933e-05,
+      "loss": 0.6667,
+      "step": 345
+    },
+    {
+      "epoch": 0.5084496693607642,
+      "grad_norm": 2.6718835830688477,
+      "learning_rate": 4.9999670613374924e-05,
+      "loss": 0.7398,
+      "step": 346
+    },
+    {
+      "epoch": 0.5099191770756797,
+      "grad_norm": 2.7223904132843018,
+      "learning_rate": 4.9999525683718174e-05,
+      "loss": 0.6194,
+      "step": 347
+    },
+    {
+      "epoch": 0.5113886847905952,
+      "grad_norm": 2.5309524536132812,
+      "learning_rate": 4.9999354403575816e-05,
+      "loss": 0.6541,
+      "step": 348
+    },
+    {
+      "epoch": 0.5128581925055107,
+      "grad_norm": 3.3481900691986084,
+      "learning_rate": 4.999915677312839e-05,
+      "loss": 0.6602,
+      "step": 349
+    },
+    {
+      "epoch": 0.5143277002204262,
+      "grad_norm": 3.8400092124938965,
+      "learning_rate": 4.99989327925842e-05,
+      "loss": 0.918,
+      "step": 350
+    },
+    {
+      "epoch": 0.5157972079353417,
+      "grad_norm": 3.0437402725219727,
+      "learning_rate": 4.999868246217933e-05,
+      "loss": 0.7748,
+      "step": 351
+    },
+    {
+      "epoch": 0.5172667156502572,
+      "grad_norm": 2.908165454864502,
+      "learning_rate": 4.9998405782177646e-05,
+      "loss": 0.6077,
+      "step": 352
+    },
+    {
+      "epoch": 0.5187362233651727,
+      "grad_norm": 2.642967939376831,
+      "learning_rate": 4.9998102752870765e-05,
+      "loss": 0.7106,
+      "step": 353
+    },
+    {
+      "epoch": 0.5202057310800882,
+      "grad_norm": 2.8907039165496826,
+      "learning_rate": 4.9997773374578105e-05,
+      "loss": 0.6963,
+      "step": 354
+    },
+    {
+      "epoch": 0.5216752387950037,
+      "grad_norm": 2.6937923431396484,
+      "learning_rate": 4.999741764764684e-05,
+      "loss": 0.593,
+      "step": 355
+    },
+    {
+      "epoch": 0.5231447465099192,
+      "grad_norm": 3.0114598274230957,
+      "learning_rate": 4.999703557245192e-05,
+      "loss": 0.7375,
+      "step": 356
+    },
+    {
+      "epoch": 0.5246142542248347,
+      "grad_norm": 2.906250238418579,
+      "learning_rate": 4.9996627149396075e-05,
+      "loss": 0.8464,
+      "step": 357
+    },
+    {
+      "epoch": 0.5260837619397501,
+      "grad_norm": 2.270758628845215,
+      "learning_rate": 4.9996192378909786e-05,
+      "loss": 0.5146,
+      "step": 358
+    },
+    {
+      "epoch": 0.5275532696546656,
+      "grad_norm": 2.3181324005126953,
+      "learning_rate": 4.999573126145132e-05,
+      "loss": 0.5463,
+      "step": 359
+    },
+    {
+      "epoch": 0.5290227773695811,
+      "grad_norm": 2.1571543216705322,
+      "learning_rate": 4.999524379750672e-05,
+      "loss": 0.4927,
+      "step": 360
+    },
+    {
+      "epoch": 0.5304922850844966,
+      "grad_norm": 2.7695159912109375,
+      "learning_rate": 4.999472998758978e-05,
+      "loss": 0.5966,
+      "step": 361
+    },
+    {
+      "epoch": 0.5319617927994122,
+      "grad_norm": 3.023146867752075,
+      "learning_rate": 4.99941898322421e-05,
+      "loss": 0.662,
+      "step": 362
+    },
+    {
+      "epoch": 0.5334313005143277,
+      "grad_norm": 2.5564310550689697,
+      "learning_rate": 4.9993623332033004e-05,
+      "loss": 0.7129,
+      "step": 363
+    },
+    {
+      "epoch": 0.5349008082292432,
+      "grad_norm": 3.416302442550659,
+      "learning_rate": 4.99930304875596e-05,
+      "loss": 0.8221,
+      "step": 364
+    },
+    {
+      "epoch": 0.5363703159441587,
+      "grad_norm": 2.5907373428344727,
+      "learning_rate": 4.99924112994468e-05,
+      "loss": 0.693,
+      "step": 365
+    },
+    {
+      "epoch": 0.5378398236590742,
+      "grad_norm": 3.254452705383301,
+      "learning_rate": 4.9991765768347214e-05,
+      "loss": 0.727,
+      "step": 366
+    },
+    {
+      "epoch": 0.5393093313739897,
+      "grad_norm": 2.4297525882720947,
+      "learning_rate": 4.999109389494129e-05,
+      "loss": 0.5491,
+      "step": 367
+    },
+    {
+      "epoch": 0.5407788390889052,
+      "grad_norm": 2.8267903327941895,
+      "learning_rate": 4.999039567993719e-05,
+      "loss": 0.5792,
+      "step": 368
+    },
+    {
+      "epoch": 0.5422483468038207,
+      "grad_norm": 2.3427703380584717,
+      "learning_rate": 4.9989671124070866e-05,
+      "loss": 0.6866,
+      "step": 369
+    },
+    {
+      "epoch": 0.5437178545187362,
+      "grad_norm": 2.5801312923431396,
+      "learning_rate": 4.998892022810603e-05,
+      "loss": 0.6821,
+      "step": 370
+    },
+    {
+      "epoch": 0.5451873622336517,
+      "grad_norm": 2.6683170795440674,
+      "learning_rate": 4.998814299283415e-05,
+      "loss": 0.6552,
+      "step": 371
+    },
+    {
+      "epoch": 0.5466568699485672,
+      "grad_norm": 2.7003276348114014,
+      "learning_rate": 4.998733941907447e-05,
+      "loss": 0.6912,
+      "step": 372
+    },
+    {
+      "epoch": 0.5481263776634827,
+      "grad_norm": 2.683290481567383,
+      "learning_rate": 4.9986509507673986e-05,
+      "loss": 0.5951,
+      "step": 373
+    },
+    {
+      "epoch": 0.5495958853783982,
+      "grad_norm": 2.4549474716186523,
+      "learning_rate": 4.998565325950745e-05,
+      "loss": 0.623,
+      "step": 374
+    },
+    {
+      "epoch": 0.5510653930933137,
+      "grad_norm": 3.1299986839294434,
+      "learning_rate": 4.99847706754774e-05,
+      "loss": 0.604,
+      "step": 375
+    },
+    {
+      "epoch": 0.5525349008082292,
+      "grad_norm": 2.6778576374053955,
+      "learning_rate": 4.9983861756514095e-05,
+      "loss": 0.6648,
+      "step": 376
+    },
+    {
+      "epoch": 0.5540044085231447,
+      "grad_norm": 3.003086805343628,
+      "learning_rate": 4.998292650357558e-05,
+      "loss": 0.6857,
+      "step": 377
+    },
+    {
+      "epoch": 0.5554739162380602,
+      "grad_norm": 2.682598829269409,
+      "learning_rate": 4.9981964917647636e-05,
+      "loss": 0.4993,
+      "step": 378
+    },
+    {
+      "epoch": 0.5569434239529758,
+      "grad_norm": 2.7633721828460693,
+      "learning_rate": 4.9980976999743835e-05,
+      "loss": 0.6114,
+      "step": 379
+    },
+    {
+      "epoch": 0.5584129316678913,
+      "grad_norm": 2.5254220962524414,
+      "learning_rate": 4.997996275090547e-05,
+      "loss": 0.6483,
+      "step": 380
+    },
+    {
+      "epoch": 0.5598824393828068,
+      "grad_norm": 2.355877637863159,
+      "learning_rate": 4.99789221722016e-05,
+      "loss": 0.591,
+      "step": 381
+    },
+    {
+      "epoch": 0.5613519470977223,
+      "grad_norm": 2.745267629623413,
+      "learning_rate": 4.997785526472903e-05,
+      "loss": 0.6491,
+      "step": 382
+    },
+    {
+      "epoch": 0.5628214548126378,
+      "grad_norm": 2.442295789718628,
+      "learning_rate": 4.9976762029612335e-05,
+      "loss": 0.5434,
+      "step": 383
+    },
+    {
+      "epoch": 0.5642909625275533,
+      "grad_norm": 2.0727195739746094,
+      "learning_rate": 4.9975642468003814e-05,
+      "loss": 0.4434,
+      "step": 384
+    },
+    {
+      "epoch": 0.5657604702424688,
+      "grad_norm": 2.6510517597198486,
+      "learning_rate": 4.997449658108354e-05,
+      "loss": 0.5683,
+      "step": 385
+    },
+    {
+      "epoch": 0.5672299779573843,
+      "grad_norm": 2.9652810096740723,
+      "learning_rate": 4.997332437005931e-05,
+      "loss": 0.8551,
+      "step": 386
+    },
+    {
+      "epoch": 0.5686994856722998,
+      "grad_norm": 2.6075005531311035,
+      "learning_rate": 4.99721258361667e-05,
+      "loss": 0.5236,
+      "step": 387
+    },
+    {
+      "epoch": 0.5701689933872153,
+      "grad_norm": 3.1933131217956543,
+      "learning_rate": 4.9970900980669e-05,
+      "loss": 0.8335,
+      "step": 388
+    },
+    {
+      "epoch": 0.5716385011021308,
+      "grad_norm": 3.4058713912963867,
+      "learning_rate": 4.9969649804857257e-05,
+      "loss": 0.8279,
+      "step": 389
+    },
+    {
+      "epoch": 0.5731080088170463,
+      "grad_norm": 2.2080886363983154,
+      "learning_rate": 4.996837231005025e-05,
+      "loss": 0.5386,
+      "step": 390
+    },
+    {
+      "epoch": 0.5745775165319618,
+      "grad_norm": 2.260094404220581,
+      "learning_rate": 4.996706849759453e-05,
+      "loss": 0.6659,
+      "step": 391
+    },
+    {
+      "epoch": 0.5760470242468773,
+      "grad_norm": 2.776089906692505,
+      "learning_rate": 4.996573836886435e-05,
+      "loss": 0.6339,
+      "step": 392
+    },
+    {
+      "epoch": 0.5775165319617928,
+      "grad_norm": 2.6933529376983643,
+      "learning_rate": 4.9964381925261726e-05,
+      "loss": 0.5848,
+      "step": 393
+    },
+    {
+      "epoch": 0.5789860396767083,
+      "grad_norm": 2.3724892139434814,
+      "learning_rate": 4.996299916821639e-05,
+      "loss": 0.6801,
+      "step": 394
+    },
+    {
+      "epoch": 0.5804555473916239,
+      "grad_norm": 2.7813003063201904,
+      "learning_rate": 4.996159009918585e-05,
+      "loss": 0.6224,
+      "step": 395
+    },
+    {
+      "epoch": 0.5819250551065394,
+      "grad_norm": 2.2573020458221436,
+      "learning_rate": 4.9960154719655296e-05,
+      "loss": 0.5707,
+      "step": 396
+    },
+    {
+      "epoch": 0.5833945628214549,
+      "grad_norm": 2.6481475830078125,
+      "learning_rate": 4.9958693031137684e-05,
+      "loss": 0.6053,
+      "step": 397
+    },
+    {
+      "epoch": 0.5848640705363704,
+      "grad_norm": 2.738497018814087,
+      "learning_rate": 4.995720503517369e-05,
+      "loss": 0.6683,
+      "step": 398
+    },
+    {
+      "epoch": 0.5863335782512858,
+      "grad_norm": 2.8765463829040527,
+      "learning_rate": 4.9955690733331716e-05,
+      "loss": 0.6826,
+      "step": 399
+    },
+    {
+      "epoch": 0.5878030859662013,
+      "grad_norm": 3.3743786811828613,
+      "learning_rate": 4.9954150127207896e-05,
+      "loss": 0.6374,
+      "step": 400
+    },
+    {
+      "epoch": 0.5892725936811168,
+      "grad_norm": 2.5285439491271973,
+      "learning_rate": 4.995258321842611e-05,
+      "loss": 0.7114,
+      "step": 401
+    },
+    {
+      "epoch": 0.5907421013960323,
+      "grad_norm": 2.819162607192993,
+      "learning_rate": 4.995099000863791e-05,
+      "loss": 0.8048,
+      "step": 402
+    },
+    {
+      "epoch": 0.5922116091109478,
+      "grad_norm": 2.3715450763702393,
+      "learning_rate": 4.994937049952262e-05,
+      "loss": 0.5249,
+      "step": 403
+    },
+    {
+      "epoch": 0.5936811168258633,
+      "grad_norm": 2.583075761795044,
+      "learning_rate": 4.994772469278726e-05,
+      "loss": 0.6024,
+      "step": 404
+    },
+    {
+      "epoch": 0.5951506245407788,
+      "grad_norm": 2.8252193927764893,
+      "learning_rate": 4.994605259016658e-05,
+      "loss": 0.7375,
+      "step": 405
+    },
+    {
+      "epoch": 0.5966201322556943,
+      "grad_norm": 2.380434989929199,
+      "learning_rate": 4.9944354193423046e-05,
+      "loss": 0.6004,
+      "step": 406
+    },
+    {
+      "epoch": 0.5980896399706098,
+      "grad_norm": 2.782414674758911,
+      "learning_rate": 4.9942629504346827e-05,
+      "loss": 0.5657,
+      "step": 407
+    },
+    {
+      "epoch": 0.5995591476855253,
+      "grad_norm": 2.390254259109497,
+      "learning_rate": 4.994087852475581e-05,
+      "loss": 0.6545,
+      "step": 408
+    },
+    {
+      "epoch": 0.6010286554004408,
+      "grad_norm": 2.0151374340057373,
+      "learning_rate": 4.993910125649561e-05,
+      "loss": 0.4254,
+      "step": 409
+    },
+    {
+      "epoch": 0.6024981631153563,
+      "grad_norm": 2.577204704284668,
+      "learning_rate": 4.993729770143952e-05,
+      "loss": 0.7533,
+      "step": 410
+    },
+    {
+      "epoch": 0.6039676708302718,
+      "grad_norm": 2.825247049331665,
+      "learning_rate": 4.993546786148858e-05,
+      "loss": 0.7267,
+      "step": 411
+    },
+    {
+      "epoch": 0.6054371785451874,
+      "grad_norm": 2.2481977939605713,
+      "learning_rate": 4.993361173857149e-05,
+      "loss": 0.4952,
+      "step": 412
+    },
+    {
+      "epoch": 0.6069066862601029,
+      "grad_norm": 2.5996899604797363,
+      "learning_rate": 4.993172933464471e-05,
+      "loss": 0.6686,
+      "step": 413
+    },
+    {
+      "epoch": 0.6083761939750184,
+      "grad_norm": 2.340848684310913,
+      "learning_rate": 4.992982065169233e-05,
+      "loss": 0.5931,
+      "step": 414
+    },
+    {
+      "epoch": 0.6098457016899339,
+      "grad_norm": 2.3505592346191406,
+      "learning_rate": 4.9927885691726196e-05,
+      "loss": 0.56,
+      "step": 415
+    },
+    {
+      "epoch": 0.6113152094048494,
+      "grad_norm": 3.110799551010132,
+      "learning_rate": 4.992592445678582e-05,
+      "loss": 0.5413,
+      "step": 416
+    },
+    {
+      "epoch": 0.6127847171197649,
+      "grad_norm": 2.730454206466675,
+      "learning_rate": 4.992393694893844e-05,
+      "loss": 0.5749,
+      "step": 417
+    },
+    {
+      "epoch": 0.6142542248346804,
+      "grad_norm": 2.2132699489593506,
+      "learning_rate": 4.992192317027896e-05,
+      "loss": 0.4575,
+      "step": 418
+    },
+    {
+      "epoch": 0.6157237325495959,
+      "grad_norm": 2.815307378768921,
+      "learning_rate": 4.9919883122929976e-05,
+      "loss": 0.7793,
+      "step": 419
+    },
+    {
+      "epoch": 0.6171932402645114,
+      "grad_norm": 2.772037982940674,
+      "learning_rate": 4.991781680904179e-05,
+      "loss": 0.6576,
+      "step": 420
+    },
+    {
+      "epoch": 0.6186627479794269,
+      "grad_norm": 2.4888830184936523,
+      "learning_rate": 4.991572423079236e-05,
+      "loss": 0.5748,
+      "step": 421
+    },
+    {
+      "epoch": 0.6201322556943424,
+      "grad_norm": 2.401238203048706,
+      "learning_rate": 4.9913605390387365e-05,
+      "loss": 0.6055,
+      "step": 422
+    },
+    {
+      "epoch": 0.6216017634092579,
+      "grad_norm": 2.544219970703125,
+      "learning_rate": 4.991146029006013e-05,
+      "loss": 0.5938,
+      "step": 423
+    },
+    {
+      "epoch": 0.6230712711241734,
+      "grad_norm": 2.199223279953003,
+      "learning_rate": 4.9909288932071695e-05,
+      "loss": 0.4522,
+      "step": 424
+    },
+    {
+      "epoch": 0.6245407788390889,
+      "grad_norm": 2.139239549636841,
+      "learning_rate": 4.990709131871074e-05,
+      "loss": 0.5363,
+      "step": 425
+    },
+    {
+      "epoch": 0.6260102865540044,
+      "grad_norm": 2.8282506465911865,
+      "learning_rate": 4.990486745229364e-05,
+      "loss": 0.7737,
+      "step": 426
+    },
+    {
+      "epoch": 0.6274797942689199,
+      "grad_norm": 3.6818435192108154,
+      "learning_rate": 4.9902617335164444e-05,
+      "loss": 0.6536,
+      "step": 427
+    },
+    {
+      "epoch": 0.6289493019838354,
+      "grad_norm": 2.704010248184204,
+      "learning_rate": 4.9900340969694855e-05,
+      "loss": 0.7121,
+      "step": 428
+    },
+    {
+      "epoch": 0.630418809698751,
+      "grad_norm": 2.3553385734558105,
+      "learning_rate": 4.989803835828426e-05,
+      "loss": 0.466,
+      "step": 429
+    },
+    {
+      "epoch": 0.6318883174136665,
+      "grad_norm": 2.4212934970855713,
+      "learning_rate": 4.989570950335969e-05,
+      "loss": 0.6004,
+      "step": 430
+    },
+    {
+      "epoch": 0.633357825128582,
+      "grad_norm": 2.4454026222229004,
+      "learning_rate": 4.989335440737586e-05,
+      "loss": 0.5314,
+      "step": 431
+    },
+    {
+      "epoch": 0.6348273328434975,
+      "grad_norm": 2.5171499252319336,
+      "learning_rate": 4.989097307281514e-05,
+      "loss": 0.5342,
+      "step": 432
+    },
+    {
+      "epoch": 0.636296840558413,
+      "grad_norm": 2.800564765930176,
+      "learning_rate": 4.9888565502187545e-05,
+      "loss": 0.7134,
+      "step": 433
+    },
+    {
+      "epoch": 0.6377663482733285,
+      "grad_norm": 2.5622341632843018,
+      "learning_rate": 4.988613169803074e-05,
+      "loss": 0.6495,
+      "step": 434
+    },
+    {
+      "epoch": 0.639235855988244,
+      "grad_norm": 2.0342607498168945,
+      "learning_rate": 4.988367166291006e-05,
+      "loss": 0.6043,
+      "step": 435
+    },
+    {
+      "epoch": 0.6407053637031594,
+      "grad_norm": 3.0105364322662354,
+      "learning_rate": 4.988118539941848e-05,
+      "loss": 0.7853,
+      "step": 436
+    },
+    {
+      "epoch": 0.6421748714180749,
+      "grad_norm": 2.254004716873169,
+      "learning_rate": 4.9878672910176615e-05,
+      "loss": 0.5621,
+      "step": 437
+    },
+    {
+      "epoch": 0.6436443791329904,
+      "grad_norm": 2.531363010406494,
+      "learning_rate": 4.9876134197832725e-05,
+      "loss": 0.6911,
+      "step": 438
+    },
+    {
+      "epoch": 0.6451138868479059,
+      "grad_norm": 2.1498162746429443,
+      "learning_rate": 4.987356926506272e-05,
+      "loss": 0.5288,
+      "step": 439
+    },
+    {
+      "epoch": 0.6465833945628214,
+      "grad_norm": 3.122633457183838,
+      "learning_rate": 4.987097811457014e-05,
+      "loss": 0.6046,
+      "step": 440
+    },
+    {
+      "epoch": 0.6480529022777369,
+      "grad_norm": 2.7230827808380127,
+      "learning_rate": 4.986836074908616e-05,
+      "loss": 0.8692,
+      "step": 441
+    },
+    {
+      "epoch": 0.6495224099926524,
+      "grad_norm": 2.3932485580444336,
+      "learning_rate": 4.986571717136958e-05,
+      "loss": 0.5097,
+      "step": 442
+    },
+    {
+      "epoch": 0.6509919177075679,
+      "grad_norm": 2.253246545791626,
+      "learning_rate": 4.9863047384206835e-05,
+      "loss": 0.5718,
+      "step": 443
+    },
+    {
+      "epoch": 0.6524614254224834,
+      "grad_norm": 2.589494228363037,
+      "learning_rate": 4.9860351390411994e-05,
+      "loss": 0.654,
+      "step": 444
+    },
+    {
+      "epoch": 0.6539309331373989,
+      "grad_norm": 2.5916192531585693,
+      "learning_rate": 4.985762919282674e-05,
+      "loss": 0.62,
+      "step": 445
+    },
+    {
+      "epoch": 0.6554004408523145,
+      "grad_norm": 2.543355703353882,
+      "learning_rate": 4.985488079432037e-05,
+      "loss": 0.6703,
+      "step": 446
+    },
+    {
+      "epoch": 0.65686994856723,
+      "grad_norm": 2.750793218612671,
+      "learning_rate": 4.98521061977898e-05,
+      "loss": 0.6038,
+      "step": 447
+    },
+    {
+      "epoch": 0.6583394562821455,
+      "grad_norm": 2.897343873977661,
+      "learning_rate": 4.984930540615958e-05,
+      "loss": 0.803,
+      "step": 448
+    },
+    {
+      "epoch": 0.659808963997061,
+      "grad_norm": 2.5928101539611816,
+      "learning_rate": 4.984647842238185e-05,
+      "loss": 0.5722,
+      "step": 449
+    },
+    {
+      "epoch": 0.6612784717119765,
+      "grad_norm": 3.081439733505249,
+      "learning_rate": 4.984362524943634e-05,
+      "loss": 0.6498,
+      "step": 450
+    },
+    {
+      "epoch": 0.662747979426892,
+      "grad_norm": 2.3990509510040283,
+      "learning_rate": 4.984074589033044e-05,
+      "loss": 0.5255,
+      "step": 451
+    },
+    {
+      "epoch": 0.6642174871418075,
+      "grad_norm": 3.276865243911743,
+      "learning_rate": 4.9837840348099095e-05,
+      "loss": 0.7157,
+      "step": 452
+    },
+    {
+      "epoch": 0.665686994856723,
+      "grad_norm": 2.175750970840454,
+      "learning_rate": 4.983490862580486e-05,
+      "loss": 0.5651,
+      "step": 453
+    },
+    {
+      "epoch": 0.6671565025716385,
+      "grad_norm": 2.028820753097534,
+      "learning_rate": 4.983195072653788e-05,
+      "loss": 0.6101,
+      "step": 454
+    },
+    {
+      "epoch": 0.668626010286554,
+      "grad_norm": 2.1656150817871094,
+      "learning_rate": 4.982896665341591e-05,
+      "loss": 0.6312,
+      "step": 455
+    },
+    {
+      "epoch": 0.6700955180014695,
+      "grad_norm": 2.318148136138916,
+      "learning_rate": 4.9825956409584266e-05,
+      "loss": 0.5994,
+      "step": 456
+    },
+    {
+      "epoch": 0.671565025716385,
+      "grad_norm": 3.0571541786193848,
+      "learning_rate": 4.982291999821587e-05,
+      "loss": 0.6476,
+      "step": 457
+    },
+    {
+      "epoch": 0.6730345334313005,
+      "grad_norm": 2.604518175125122,
+      "learning_rate": 4.981985742251123e-05,
+      "loss": 0.4552,
+      "step": 458
+    },
+    {
+      "epoch": 0.674504041146216,
+      "grad_norm": 2.1126625537872314,
+      "learning_rate": 4.98167686856984e-05,
+      "loss": 0.594,
+      "step": 459
+    },
+    {
+      "epoch": 0.6759735488611315,
+      "grad_norm": 2.7541537284851074,
+      "learning_rate": 4.9813653791033057e-05,
+      "loss": 0.697,
+      "step": 460
+    },
+    {
+      "epoch": 0.677443056576047,
+      "grad_norm": 4.093698024749756,
+      "learning_rate": 4.98105127417984e-05,
+      "loss": 0.8774,
+      "step": 461
+    },
+    {
+      "epoch": 0.6789125642909625,
+      "grad_norm": 2.2070083618164062,
+      "learning_rate": 4.9807345541305224e-05,
+      "loss": 0.6611,
+      "step": 462
+    },
+    {
+      "epoch": 0.6803820720058781,
+      "grad_norm": 2.1070473194122314,
+      "learning_rate": 4.980415219289189e-05,
+      "loss": 0.635,
+      "step": 463
+    },
+    {
+      "epoch": 0.6818515797207936,
+      "grad_norm": 2.496419668197632,
+      "learning_rate": 4.980093269992431e-05,
+      "loss": 0.6512,
+      "step": 464
+    },
+    {
+      "epoch": 0.6833210874357091,
+      "grad_norm": 2.0512053966522217,
+      "learning_rate": 4.979768706579595e-05,
+      "loss": 0.6132,
+      "step": 465
+    },
+    {
+      "epoch": 0.6847905951506246,
+      "grad_norm": 2.995798349380493,
+      "learning_rate": 4.9794415293927846e-05,
+      "loss": 0.802,
+      "step": 466
+    },
+    {
+      "epoch": 0.6862601028655401,
+      "grad_norm": 2.3697338104248047,
+      "learning_rate": 4.979111738776857e-05,
+      "loss": 0.689,
+      "step": 467
+    },
+    {
+      "epoch": 0.6877296105804556,
+      "grad_norm": 2.598092555999756,
+      "learning_rate": 4.978779335079425e-05,
+      "loss": 0.7214,
+      "step": 468
+    },
+    {
+      "epoch": 0.6891991182953711,
+      "grad_norm": 2.862518072128296,
+      "learning_rate": 4.9784443186508545e-05,
+      "loss": 0.7442,
+      "step": 469
+    },
+    {
+      "epoch": 0.6906686260102866,
+      "grad_norm": 2.325723886489868,
+      "learning_rate": 4.978106689844266e-05,
+      "loss": 0.6064,
+      "step": 470
+    },
+    {
+      "epoch": 0.692138133725202,
+      "grad_norm": 2.716212511062622,
+      "learning_rate": 4.977766449015534e-05,
+      "loss": 0.7211,
+      "step": 471
+    },
+    {
+      "epoch": 0.6936076414401176,
+      "grad_norm": 2.5306997299194336,
+      "learning_rate": 4.9774235965232855e-05,
+      "loss": 0.5997,
+      "step": 472
+    },
+    {
+      "epoch": 0.695077149155033,
+      "grad_norm": 2.6134307384490967,
+      "learning_rate": 4.977078132728901e-05,
+      "loss": 0.6601,
+      "step": 473
+    },
+    {
+      "epoch": 0.6965466568699485,
+      "grad_norm": 2.220085382461548,
+      "learning_rate": 4.9767300579965126e-05,
+      "loss": 0.5497,
+      "step": 474
+    },
+    {
+      "epoch": 0.698016164584864,
+      "grad_norm": 2.1364073753356934,
+      "learning_rate": 4.976379372693004e-05,
+      "loss": 0.5988,
+      "step": 475
+    },
+    {
+      "epoch": 0.6994856722997795,
+      "grad_norm": 2.08339786529541,
+      "learning_rate": 4.976026077188013e-05,
+      "loss": 0.5257,
+      "step": 476
+    },
+    {
+      "epoch": 0.700955180014695,
+      "grad_norm": 2.223860025405884,
+      "learning_rate": 4.975670171853926e-05,
+      "loss": 0.5586,
+      "step": 477
+    },
+    {
+      "epoch": 0.7024246877296105,
+      "grad_norm": 2.2933597564697266,
+      "learning_rate": 4.9753116570658816e-05,
+      "loss": 0.6017,
+      "step": 478
+    },
+    {
+      "epoch": 0.7038941954445261,
+      "grad_norm": 2.100961685180664,
+      "learning_rate": 4.974950533201767e-05,
+      "loss": 0.5555,
+      "step": 479
+    },
+    {
+      "epoch": 0.7053637031594416,
+      "grad_norm": 2.059403419494629,
+      "learning_rate": 4.974586800642224e-05,
+      "loss": 0.5,
+      "step": 480
+    },
+    {
+      "epoch": 0.7068332108743571,
+      "grad_norm": 3.0029308795928955,
+      "learning_rate": 4.974220459770639e-05,
+      "loss": 0.6993,
+      "step": 481
+    },
+    {
+      "epoch": 0.7083027185892726,
+      "grad_norm": 2.48919939994812,
+      "learning_rate": 4.973851510973149e-05,
+      "loss": 0.6255,
+      "step": 482
+    },
+    {
+      "epoch": 0.7097722263041881,
+      "grad_norm": 2.4225497245788574,
+      "learning_rate": 4.973479954638642e-05,
+      "loss": 0.5851,
+      "step": 483
+    },
+    {
+      "epoch": 0.7112417340191036,
+      "grad_norm": 2.0806729793548584,
+      "learning_rate": 4.973105791158753e-05,
+      "loss": 0.4984,
+      "step": 484
+    },
+    {
+      "epoch": 0.7127112417340191,
+      "grad_norm": 2.5935800075531006,
+      "learning_rate": 4.972729020927865e-05,
+      "loss": 0.6968,
+      "step": 485
+    },
+    {
+      "epoch": 0.7141807494489346,
+      "grad_norm": 2.607990264892578,
+      "learning_rate": 4.972349644343108e-05,
+      "loss": 0.6313,
+      "step": 486
+    },
+    {
+      "epoch": 0.7156502571638501,
+      "grad_norm": 2.6710383892059326,
+      "learning_rate": 4.9719676618043606e-05,
+      "loss": 0.7649,
+      "step": 487
+    },
+    {
+      "epoch": 0.7171197648787656,
+      "grad_norm": 2.6919174194335938,
+      "learning_rate": 4.971583073714247e-05,
+      "loss": 0.8269,
+      "step": 488
+    },
+    {
+      "epoch": 0.7185892725936811,
+      "grad_norm": 2.8584353923797607,
+      "learning_rate": 4.9711958804781385e-05,
+      "loss": 0.5916,
+      "step": 489
+    },
+    {
+      "epoch": 0.7200587803085966,
+      "grad_norm": 2.439382314682007,
+      "learning_rate": 4.9708060825041515e-05,
+      "loss": 0.5237,
+      "step": 490
+    },
+    {
+      "epoch": 0.7215282880235121,
+      "grad_norm": 2.1240954399108887,
+      "learning_rate": 4.9704136802031485e-05,
+      "loss": 0.5921,
+      "step": 491
+    },
+    {
+      "epoch": 0.7229977957384276,
+      "grad_norm": 2.438596487045288,
+      "learning_rate": 4.9700186739887376e-05,
+      "loss": 0.4835,
+      "step": 492
+    },
+    {
+      "epoch": 0.7244673034533431,
+      "grad_norm": 3.331761360168457,
+      "learning_rate": 4.969621064277271e-05,
+      "loss": 0.6622,
+      "step": 493
+    },
+    {
+      "epoch": 0.7259368111682586,
+      "grad_norm": 2.1253693103790283,
+      "learning_rate": 4.9692208514878444e-05,
+      "loss": 0.5441,
+      "step": 494
+    },
+    {
+      "epoch": 0.7274063188831741,
+      "grad_norm": 2.170849561691284,
+      "learning_rate": 4.968818036042298e-05,
+      "loss": 0.5926,
+      "step": 495
+    },
+    {
+      "epoch": 0.7288758265980897,
+      "grad_norm": 2.6995272636413574,
+      "learning_rate": 4.968412618365216e-05,
+      "loss": 0.5765,
+      "step": 496
+    },
+    {
+      "epoch": 0.7303453343130052,
+      "grad_norm": 2.1400372982025146,
+      "learning_rate": 4.968004598883923e-05,
+      "loss": 0.6524,
+      "step": 497
+    },
+    {
+      "epoch": 0.7318148420279207,
+      "grad_norm": 2.528055191040039,
+      "learning_rate": 4.9675939780284884e-05,
+      "loss": 0.7347,
+      "step": 498
+    },
+    {
+      "epoch": 0.7332843497428362,
+      "grad_norm": 2.847480297088623,
+      "learning_rate": 4.967180756231723e-05,
+      "loss": 0.8763,
+      "step": 499
+    },
+    {
+      "epoch": 0.7347538574577517,
+      "grad_norm": 2.8384506702423096,
+      "learning_rate": 4.9667649339291786e-05,
+      "loss": 0.7365,
+      "step": 500
+    },
+    {
+      "epoch": 0.7362233651726672,
+      "grad_norm": 2.2153868675231934,
+      "learning_rate": 4.966346511559149e-05,
+      "loss": 0.6064,
+      "step": 501
+    },
+    {
+      "epoch": 0.7376928728875827,
+      "grad_norm": 2.2680118083953857,
+      "learning_rate": 4.965925489562667e-05,
+      "loss": 0.4272,
+      "step": 502
+    },
+    {
+      "epoch": 0.7391623806024982,
+      "grad_norm": 2.492192268371582,
+      "learning_rate": 4.965501868383506e-05,
+      "loss": 0.5955,
+      "step": 503
+    },
+    {
+      "epoch": 0.7406318883174137,
+      "grad_norm": 3.8302063941955566,
+      "learning_rate": 4.9650756484681816e-05,
+      "loss": 0.556,
+      "step": 504
+    },
+    {
+      "epoch": 0.7421013960323292,
+      "grad_norm": 2.8113222122192383,
+      "learning_rate": 4.964646830265944e-05,
+      "loss": 0.8223,
+      "step": 505
+    },
+    {
+      "epoch": 0.7435709037472447,
+      "grad_norm": 2.135054588317871,
+      "learning_rate": 4.964215414228786e-05,
+      "loss": 0.6246,
+      "step": 506
+    },
+    {
+      "epoch": 0.7450404114621602,
+      "grad_norm": 2.5363314151763916,
+      "learning_rate": 4.9637814008114345e-05,
+      "loss": 0.6389,
+      "step": 507
+    },
+    {
+      "epoch": 0.7465099191770757,
+      "grad_norm": 2.5335984230041504,
+      "learning_rate": 4.96334479047136e-05,
+      "loss": 0.5798,
+      "step": 508
+    },
+    {
+      "epoch": 0.7479794268919912,
+      "grad_norm": 2.232638120651245,
+      "learning_rate": 4.9629055836687665e-05,
+      "loss": 0.522,
+      "step": 509
+    },
+    {
+      "epoch": 0.7494489346069066,
+      "grad_norm": 2.3462014198303223,
+      "learning_rate": 4.962463780866594e-05,
+      "loss": 0.6007,
+      "step": 510
+    },
+    {
+      "epoch": 0.7509184423218221,
+      "grad_norm": 2.783234119415283,
+      "learning_rate": 4.962019382530521e-05,
+      "loss": 0.7309,
+      "step": 511
+    },
+    {
+      "epoch": 0.7523879500367376,
+      "grad_norm": 2.3357062339782715,
+      "learning_rate": 4.961572389128959e-05,
+      "loss": 0.5258,
+      "step": 512
+    },
+    {
+      "epoch": 0.7538574577516532,
+      "grad_norm": 2.442301034927368,
+      "learning_rate": 4.961122801133059e-05,
+      "loss": 0.7335,
+      "step": 513
+    },
+    {
+      "epoch": 0.7553269654665687,
+      "grad_norm": 2.1370739936828613,
+      "learning_rate": 4.960670619016704e-05,
+      "loss": 0.5849,
+      "step": 514
+    },
+    {
+      "epoch": 0.7567964731814842,
+      "grad_norm": 2.2204878330230713,
+      "learning_rate": 4.960215843256511e-05,
+      "loss": 0.55,
+      "step": 515
+    },
+    {
+      "epoch": 0.7582659808963997,
+      "grad_norm": 2.583329916000366,
+      "learning_rate": 4.959758474331833e-05,
+      "loss": 0.7699,
+      "step": 516
+    },
+    {
+      "epoch": 0.7597354886113152,
+      "grad_norm": 2.358966588973999,
+      "learning_rate": 4.959298512724753e-05,
+      "loss": 0.7367,
+      "step": 517
+    },
+    {
+      "epoch": 0.7612049963262307,
+      "grad_norm": 2.416844606399536,
+      "learning_rate": 4.9588359589200884e-05,
+      "loss": 0.7033,
+      "step": 518
+    },
+    {
+      "epoch": 0.7626745040411462,
+      "grad_norm": 3.295530319213867,
+      "learning_rate": 4.958370813405392e-05,
+      "loss": 0.7474,
+      "step": 519
+    },
+    {
+      "epoch": 0.7641440117560617,
+      "grad_norm": 2.5781569480895996,
+      "learning_rate": 4.957903076670942e-05,
+      "loss": 0.6652,
+      "step": 520
+    },
+    {
+      "epoch": 0.7656135194709772,
+      "grad_norm": 2.1570255756378174,
+      "learning_rate": 4.957432749209755e-05,
+      "loss": 0.6353,
+      "step": 521
+    },
+    {
+      "epoch": 0.7670830271858927,
+      "grad_norm": 2.01401948928833,
+      "learning_rate": 4.956959831517572e-05,
+      "loss": 0.4917,
+      "step": 522
+    },
+    {
+      "epoch": 0.7685525349008082,
+      "grad_norm": 2.164358139038086,
+      "learning_rate": 4.956484324092867e-05,
+      "loss": 0.6899,
+      "step": 523
+    },
+    {
+      "epoch": 0.7700220426157237,
+      "grad_norm": 3.4753246307373047,
+      "learning_rate": 4.956006227436845e-05,
+      "loss": 0.6401,
+      "step": 524
+    },
+    {
+      "epoch": 0.7714915503306392,
+      "grad_norm": 2.4580509662628174,
+      "learning_rate": 4.955525542053438e-05,
+      "loss": 0.61,
+      "step": 525
+    },
+    {
+      "epoch": 0.7729610580455547,
+      "grad_norm": 2.6388328075408936,
+      "learning_rate": 4.955042268449307e-05,
+      "loss": 0.7727,
+      "step": 526
+    },
+    {
+      "epoch": 0.7744305657604702,
+      "grad_norm": 2.131551504135132,
+      "learning_rate": 4.9545564071338426e-05,
+      "loss": 0.6592,
+      "step": 527
+    },
+    {
+      "epoch": 0.7759000734753857,
+      "grad_norm": 1.9104920625686646,
+      "learning_rate": 4.9540679586191605e-05,
+      "loss": 0.4329,
+      "step": 528
+    },
+    {
+      "epoch": 0.7773695811903012,
+      "grad_norm": 2.6334924697875977,
+      "learning_rate": 4.953576923420105e-05,
+      "loss": 0.5623,
+      "step": 529
+    },
+    {
+      "epoch": 0.7788390889052168,
+      "grad_norm": 2.2900774478912354,
+      "learning_rate": 4.953083302054247e-05,
+      "loss": 0.6244,
+      "step": 530
+    },
+    {
+      "epoch": 0.7803085966201323,
+      "grad_norm": 3.094778537750244,
+      "learning_rate": 4.952587095041882e-05,
+      "loss": 0.8521,
+      "step": 531
+    },
+    {
+      "epoch": 0.7817781043350478,
+      "grad_norm": 2.6516799926757812,
+      "learning_rate": 4.952088302906032e-05,
+      "loss": 0.6623,
+      "step": 532
+    },
+    {
+      "epoch": 0.7832476120499633,
+      "grad_norm": 2.6826205253601074,
+      "learning_rate": 4.9515869261724446e-05,
+      "loss": 0.721,
+      "step": 533
+    },
+    {
+      "epoch": 0.7847171197648788,
+      "grad_norm": 2.3135852813720703,
+      "learning_rate": 4.95108296536959e-05,
+      "loss": 0.4557,
+      "step": 534
+    },
+    {
+      "epoch": 0.7861866274797943,
+      "grad_norm": 2.823601007461548,
+      "learning_rate": 4.950576421028662e-05,
+      "loss": 0.7058,
+      "step": 535
+    },
+    {
+      "epoch": 0.7876561351947098,
+      "grad_norm": 2.627410888671875,
+      "learning_rate": 4.95006729368358e-05,
+      "loss": 0.4995,
+      "step": 536
+    },
+    {
+      "epoch": 0.7891256429096253,
+      "grad_norm": 2.2301442623138428,
+      "learning_rate": 4.949555583870983e-05,
+      "loss": 0.5666,
+      "step": 537
+    },
+    {
+      "epoch": 0.7905951506245408,
+      "grad_norm": 2.74621844291687,
+      "learning_rate": 4.949041292130233e-05,
+      "loss": 0.7564,
+      "step": 538
+    },
+    {
+      "epoch": 0.7920646583394563,
+      "grad_norm": 2.9513447284698486,
+      "learning_rate": 4.948524419003415e-05,
+      "loss": 0.7231,
+      "step": 539
+    },
+    {
+      "epoch": 0.7935341660543718,
+      "grad_norm": 3.148510217666626,
+      "learning_rate": 4.948004965035333e-05,
+      "loss": 0.6325,
+      "step": 540
+    },
+    {
+      "epoch": 0.7950036737692873,
+      "grad_norm": 3.302402973175049,
+      "learning_rate": 4.9474829307735115e-05,
+      "loss": 0.8296,
+      "step": 541
+    },
+    {
+      "epoch": 0.7964731814842028,
+      "grad_norm": 2.212381362915039,
+      "learning_rate": 4.946958316768196e-05,
+      "loss": 0.5355,
+      "step": 542
+    },
+    {
+      "epoch": 0.7979426891991183,
+      "grad_norm": 2.0545992851257324,
+      "learning_rate": 4.9464311235723506e-05,
+      "loss": 0.4572,
+      "step": 543
+    },
+    {
+      "epoch": 0.7994121969140338,
+      "grad_norm": 2.227982759475708,
+      "learning_rate": 4.9459013517416565e-05,
+      "loss": 0.62,
+      "step": 544
+    },
+    {
+      "epoch": 0.8008817046289493,
+      "grad_norm": 2.044691801071167,
+      "learning_rate": 4.9453690018345144e-05,
+      "loss": 0.5263,
+      "step": 545
+    },
+    {
+      "epoch": 0.8023512123438648,
+      "grad_norm": 3.104004144668579,
+      "learning_rate": 4.9448340744120424e-05,
+      "loss": 0.8096,
+      "step": 546
+    },
+    {
+      "epoch": 0.8038207200587804,
+      "grad_norm": 2.1079485416412354,
+      "learning_rate": 4.944296570038076e-05,
+      "loss": 0.6216,
+      "step": 547
+    },
+    {
+      "epoch": 0.8052902277736959,
+      "grad_norm": 2.836124897003174,
+      "learning_rate": 4.943756489279164e-05,
+      "loss": 0.7396,
+      "step": 548
+    },
+    {
+      "epoch": 0.8067597354886114,
+      "grad_norm": 2.9169132709503174,
+      "learning_rate": 4.943213832704575e-05,
+      "loss": 0.6457,
+      "step": 549
+    },
+    {
+      "epoch": 0.8082292432035268,
+      "grad_norm": 3.0923452377319336,
+      "learning_rate": 4.9426686008862886e-05,
+      "loss": 0.7446,
+      "step": 550
+    },
+    {
+      "epoch": 0.8096987509184423,
+      "grad_norm": 1.96324622631073,
+      "learning_rate": 4.942120794399002e-05,
+      "loss": 0.6166,
+      "step": 551
+    },
+    {
+      "epoch": 0.8111682586333578,
+      "grad_norm": 2.473320245742798,
+      "learning_rate": 4.941570413820124e-05,
+      "loss": 0.5753,
+      "step": 552
+    },
+    {
+      "epoch": 0.8126377663482733,
+      "grad_norm": 2.902329683303833,
+      "learning_rate": 4.941017459729778e-05,
+      "loss": 0.6765,
+      "step": 553
+    },
+    {
+      "epoch": 0.8141072740631888,
+      "grad_norm": 1.9809579849243164,
+      "learning_rate": 4.940461932710798e-05,
+      "loss": 0.3917,
+      "step": 554
+    },
+    {
+      "epoch": 0.8155767817781043,
+      "grad_norm": 2.079895257949829,
+      "learning_rate": 4.939903833348733e-05,
+      "loss": 0.635,
+      "step": 555
+    },
+    {
+      "epoch": 0.8170462894930198,
+      "grad_norm": 2.7767322063446045,
+      "learning_rate": 4.939343162231841e-05,
+      "loss": 0.6231,
+      "step": 556
+    },
+    {
+      "epoch": 0.8185157972079353,
+      "grad_norm": 2.6445229053497314,
+      "learning_rate": 4.938779919951092e-05,
+      "loss": 0.6512,
+      "step": 557
+    },
+    {
+      "epoch": 0.8199853049228508,
+      "grad_norm": 2.1407949924468994,
+      "learning_rate": 4.938214107100165e-05,
+      "loss": 0.6185,
+      "step": 558
+    },
+    {
+      "epoch": 0.8214548126377663,
+      "grad_norm": 2.4886536598205566,
+      "learning_rate": 4.937645724275449e-05,
+      "loss": 0.7984,
+      "step": 559
+    },
+    {
+      "epoch": 0.8229243203526818,
+      "grad_norm": 2.6786420345306396,
+      "learning_rate": 4.9370747720760426e-05,
+      "loss": 0.6552,
+      "step": 560
+    },
+    {
+      "epoch": 0.8243938280675973,
+      "grad_norm": 1.9730219841003418,
+      "learning_rate": 4.9365012511037514e-05,
+      "loss": 0.6236,
+      "step": 561
+    },
+    {
+      "epoch": 0.8258633357825128,
+      "grad_norm": 2.32598614692688,
+      "learning_rate": 4.9359251619630886e-05,
+      "loss": 0.8109,
+      "step": 562
+    },
+    {
+      "epoch": 0.8273328434974284,
+      "grad_norm": 2.27353572845459,
+      "learning_rate": 4.935346505261276e-05,
+      "loss": 0.5999,
+      "step": 563
+    },
+    {
+      "epoch": 0.8288023512123439,
+      "grad_norm": 2.688149929046631,
+      "learning_rate": 4.934765281608239e-05,
+      "loss": 0.7602,
+      "step": 564
+    },
+    {
+      "epoch": 0.8302718589272594,
+      "grad_norm": 2.15217924118042,
+      "learning_rate": 4.9341814916166124e-05,
+      "loss": 0.4012,
+      "step": 565
+    },
+    {
+      "epoch": 0.8317413666421749,
+      "grad_norm": 2.782214641571045,
+      "learning_rate": 4.933595135901732e-05,
+      "loss": 0.5736,
+      "step": 566
+    },
+    {
+      "epoch": 0.8332108743570904,
+      "grad_norm": 2.277165174484253,
+      "learning_rate": 4.933006215081641e-05,
+      "loss": 0.6248,
+      "step": 567
+    },
+    {
+      "epoch": 0.8346803820720059,
+      "grad_norm": 2.5467677116394043,
+      "learning_rate": 4.932414729777085e-05,
+      "loss": 0.5196,
+      "step": 568
+    },
+    {
+      "epoch": 0.8361498897869214,
+      "grad_norm": 2.49916672706604,
+      "learning_rate": 4.931820680611512e-05,
+      "loss": 0.5921,
+      "step": 569
+    },
+    {
+      "epoch": 0.8376193975018369,
+      "grad_norm": 2.6123595237731934,
+      "learning_rate": 4.931224068211074e-05,
+      "loss": 0.6147,
+      "step": 570
+    },
+    {
+      "epoch": 0.8390889052167524,
+      "grad_norm": 2.139331340789795,
+      "learning_rate": 4.930624893204624e-05,
+      "loss": 0.5555,
+      "step": 571
+    },
+    {
+      "epoch": 0.8405584129316679,
+      "grad_norm": 2.3458757400512695,
+      "learning_rate": 4.930023156223715e-05,
+      "loss": 0.6356,
+      "step": 572
+    },
+    {
+      "epoch": 0.8420279206465834,
+      "grad_norm": 2.045133113861084,
+      "learning_rate": 4.929418857902604e-05,
+      "loss": 0.5469,
+      "step": 573
+    },
+    {
+      "epoch": 0.8434974283614989,
+      "grad_norm": 2.3724586963653564,
+      "learning_rate": 4.928811998878242e-05,
+      "loss": 0.584,
+      "step": 574
+    },
+    {
+      "epoch": 0.8449669360764144,
+      "grad_norm": 2.31768536567688,
+      "learning_rate": 4.928202579790285e-05,
+      "loss": 0.5466,
+      "step": 575
+    },
+    {
+      "epoch": 0.8464364437913299,
+      "grad_norm": 2.7359490394592285,
+      "learning_rate": 4.9275906012810835e-05,
+      "loss": 0.6034,
+      "step": 576
+    },
+    {
+      "epoch": 0.8479059515062454,
+      "grad_norm": 2.7302405834198,
+      "learning_rate": 4.926976063995686e-05,
+      "loss": 0.7608,
+      "step": 577
+    },
+    {
+      "epoch": 0.8493754592211609,
+      "grad_norm": 2.052567720413208,
+      "learning_rate": 4.926358968581842e-05,
+      "loss": 0.5011,
+      "step": 578
+    },
+    {
+      "epoch": 0.8508449669360764,
+      "grad_norm": 2.2667741775512695,
+      "learning_rate": 4.925739315689991e-05,
+      "loss": 0.5769,
+      "step": 579
+    },
+    {
+      "epoch": 0.852314474650992,
+      "grad_norm": 2.2745754718780518,
+      "learning_rate": 4.9251171059732744e-05,
+      "loss": 0.5368,
+      "step": 580
+    },
+    {
+      "epoch": 0.8537839823659075,
+      "grad_norm": 2.3800880908966064,
+      "learning_rate": 4.9244923400875245e-05,
+      "loss": 0.669,
+      "step": 581
+    },
+    {
+      "epoch": 0.855253490080823,
+      "grad_norm": 3.142043352127075,
+      "learning_rate": 4.9238650186912683e-05,
+      "loss": 0.9251,
+      "step": 582
+    },
+    {
+      "epoch": 0.8567229977957385,
+      "grad_norm": 1.944888949394226,
+      "learning_rate": 4.923235142445728e-05,
+      "loss": 0.5754,
+      "step": 583
+    },
+    {
+      "epoch": 0.858192505510654,
+      "grad_norm": 2.640141487121582,
+      "learning_rate": 4.9226027120148195e-05,
+      "loss": 0.6702,
+      "step": 584
+    },
+    {
+      "epoch": 0.8596620132255695,
+      "grad_norm": 1.6972280740737915,
+      "learning_rate": 4.9219677280651475e-05,
+      "loss": 0.438,
+      "step": 585
+    },
+    {
+      "epoch": 0.861131520940485,
+      "grad_norm": 2.24550199508667,
+      "learning_rate": 4.9213301912660106e-05,
+      "loss": 0.5744,
+      "step": 586
+    },
+    {
+      "epoch": 0.8626010286554004,
+      "grad_norm": 1.7848341464996338,
+      "learning_rate": 4.920690102289397e-05,
+      "loss": 0.5117,
+      "step": 587
+    },
+    {
+      "epoch": 0.864070536370316,
+      "grad_norm": 2.259721279144287,
+      "learning_rate": 4.920047461809988e-05,
+      "loss": 0.5655,
+      "step": 588
+    },
+    {
+      "epoch": 0.8655400440852314,
+      "grad_norm": 2.361989736557007,
+      "learning_rate": 4.919402270505149e-05,
+      "loss": 0.6164,
+      "step": 589
+    },
+    {
+      "epoch": 0.8670095518001469,
+      "grad_norm": 2.7982707023620605,
+      "learning_rate": 4.91875452905494e-05,
+      "loss": 0.7654,
+      "step": 590
+    },
+    {
+      "epoch": 0.8684790595150624,
+      "grad_norm": 3.4847824573516846,
+      "learning_rate": 4.918104238142104e-05,
+      "loss": 0.7426,
+      "step": 591
+    },
+    {
+      "epoch": 0.8699485672299779,
+      "grad_norm": 2.416091203689575,
+      "learning_rate": 4.9174513984520744e-05,
+      "loss": 0.5251,
+      "step": 592
+    },
+    {
+      "epoch": 0.8714180749448934,
+      "grad_norm": 2.3640871047973633,
+      "learning_rate": 4.916796010672969e-05,
+      "loss": 0.6661,
+      "step": 593
+    },
+    {
+      "epoch": 0.8728875826598089,
+      "grad_norm": 2.4868736267089844,
+      "learning_rate": 4.9161380754955924e-05,
+      "loss": 0.8162,
+      "step": 594
+    },
+    {
+      "epoch": 0.8743570903747244,
+      "grad_norm": 2.2972755432128906,
+      "learning_rate": 4.915477593613436e-05,
+      "loss": 0.5511,
+      "step": 595
+    },
+    {
+      "epoch": 0.8758265980896399,
+      "grad_norm": 2.199916362762451,
+      "learning_rate": 4.914814565722671e-05,
+      "loss": 0.4608,
+      "step": 596
+    },
+    {
+      "epoch": 0.8772961058045555,
+      "grad_norm": 1.8572275638580322,
+      "learning_rate": 4.914148992522157e-05,
+      "loss": 0.436,
+      "step": 597
+    },
+    {
+      "epoch": 0.878765613519471,
+      "grad_norm": 2.1603405475616455,
+      "learning_rate": 4.913480874713433e-05,
+      "loss": 0.6311,
+      "step": 598
+    },
+    {
+      "epoch": 0.8802351212343865,
+      "grad_norm": 2.023216485977173,
+      "learning_rate": 4.9128102130007225e-05,
+      "loss": 0.544,
+      "step": 599
+    },
+    {
+      "epoch": 0.881704628949302,
+      "grad_norm": 2.329108238220215,
+      "learning_rate": 4.91213700809093e-05,
+      "loss": 0.5528,
+      "step": 600
+    },
+    {
+      "epoch": 0.8831741366642175,
+      "grad_norm": 2.149566650390625,
+      "learning_rate": 4.911461260693638e-05,
+      "loss": 0.468,
+      "step": 601
+    },
+    {
+      "epoch": 0.884643644379133,
+      "grad_norm": 2.9256246089935303,
+      "learning_rate": 4.9107829715211124e-05,
+      "loss": 0.7522,
+      "step": 602
+    },
+    {
+      "epoch": 0.8861131520940485,
+      "grad_norm": 2.787365674972534,
+      "learning_rate": 4.9101021412882965e-05,
+      "loss": 0.7629,
+      "step": 603
+    },
+    {
+      "epoch": 0.887582659808964,
+      "grad_norm": 1.7118382453918457,
+      "learning_rate": 4.9094187707128115e-05,
+      "loss": 0.5101,
+      "step": 604
+    },
+    {
+      "epoch": 0.8890521675238795,
+      "grad_norm": 4.151994705200195,
+      "learning_rate": 4.908732860514958e-05,
+      "loss": 0.9405,
+      "step": 605
+    },
+    {
+      "epoch": 0.890521675238795,
+      "grad_norm": 3.0542750358581543,
+      "learning_rate": 4.908044411417711e-05,
+      "loss": 0.733,
+      "step": 606
+    },
+    {
+      "epoch": 0.8919911829537105,
+      "grad_norm": 2.025644302368164,
+      "learning_rate": 4.907353424146726e-05,
+      "loss": 0.5462,
+      "step": 607
+    },
+    {
+      "epoch": 0.893460690668626,
+      "grad_norm": 2.461653232574463,
+      "learning_rate": 4.906659899430326e-05,
+      "loss": 0.5904,
+      "step": 608
+    },
+    {
+      "epoch": 0.8949301983835415,
+      "grad_norm": 2.174624443054199,
+      "learning_rate": 4.905963837999518e-05,
+      "loss": 0.5938,
+      "step": 609
+    },
+    {
+      "epoch": 0.896399706098457,
+      "grad_norm": 2.574413299560547,
+      "learning_rate": 4.905265240587975e-05,
+      "loss": 0.6716,
+      "step": 610
+    },
+    {
+      "epoch": 0.8978692138133725,
+      "grad_norm": 2.692120313644409,
+      "learning_rate": 4.9045641079320484e-05,
+      "loss": 0.7348,
+      "step": 611
+    },
+    {
+      "epoch": 0.899338721528288,
+      "grad_norm": 2.104729175567627,
+      "learning_rate": 4.903860440770758e-05,
+      "loss": 0.513,
+      "step": 612
+    },
+    {
+      "epoch": 0.9008082292432035,
+      "grad_norm": 2.376796245574951,
+      "learning_rate": 4.9031542398457974e-05,
+      "loss": 0.6363,
+      "step": 613
+    },
+    {
+      "epoch": 0.9022777369581191,
+      "grad_norm": 2.1267740726470947,
+      "learning_rate": 4.902445505901531e-05,
+      "loss": 0.5722,
+      "step": 614
+    },
+    {
+      "epoch": 0.9037472446730346,
+      "grad_norm": 2.4500656127929688,
+      "learning_rate": 4.901734239684991e-05,
+      "loss": 0.6533,
+      "step": 615
+    },
+    {
+      "epoch": 0.9052167523879501,
+      "grad_norm": 2.1529133319854736,
+      "learning_rate": 4.9010204419458804e-05,
+      "loss": 0.6065,
+      "step": 616
+    },
+    {
+      "epoch": 0.9066862601028656,
+      "grad_norm": 2.8337793350219727,
+      "learning_rate": 4.900304113436571e-05,
+      "loss": 0.5843,
+      "step": 617
+    },
+    {
+      "epoch": 0.9081557678177811,
+      "grad_norm": 2.0218799114227295,
+      "learning_rate": 4.8995852549121e-05,
+      "loss": 0.5309,
+      "step": 618
+    },
+    {
+      "epoch": 0.9096252755326966,
+      "grad_norm": 1.9721390008926392,
+      "learning_rate": 4.898863867130174e-05,
+      "loss": 0.525,
+      "step": 619
+    },
+    {
+      "epoch": 0.9110947832476121,
+      "grad_norm": 2.184145927429199,
+      "learning_rate": 4.8981399508511624e-05,
+      "loss": 0.5714,
+      "step": 620
+    },
+    {
+      "epoch": 0.9125642909625276,
+      "grad_norm": 2.6820855140686035,
+      "learning_rate": 4.8974135068381036e-05,
+      "loss": 0.7192,
+      "step": 621
+    },
+    {
+      "epoch": 0.914033798677443,
+      "grad_norm": 2.075873851776123,
+      "learning_rate": 4.896684535856696e-05,
+      "loss": 0.613,
+      "step": 622
+    },
+    {
+      "epoch": 0.9155033063923586,
+      "grad_norm": 2.1703760623931885,
+      "learning_rate": 4.895953038675306e-05,
+      "loss": 0.5115,
+      "step": 623
+    },
+    {
+      "epoch": 0.916972814107274,
+      "grad_norm": 1.9222508668899536,
+      "learning_rate": 4.895219016064959e-05,
+      "loss": 0.4879,
+      "step": 624
+    },
+    {
+      "epoch": 0.9184423218221895,
+      "grad_norm": 2.156532049179077,
+      "learning_rate": 4.894482468799344e-05,
+      "loss": 0.5716,
+      "step": 625
+    },
+    {
+      "epoch": 0.919911829537105,
+      "grad_norm": 2.6406233310699463,
+      "learning_rate": 4.893743397654811e-05,
+      "loss": 0.5198,
+      "step": 626
+    },
+    {
+      "epoch": 0.9213813372520205,
+      "grad_norm": 2.0648574829101562,
+      "learning_rate": 4.893001803410371e-05,
+      "loss": 0.5294,
+      "step": 627
+    },
+    {
+      "epoch": 0.922850844966936,
+      "grad_norm": 2.779963731765747,
+      "learning_rate": 4.8922576868476914e-05,
+      "loss": 0.6801,
+      "step": 628
+    },
+    {
+      "epoch": 0.9243203526818515,
+      "grad_norm": 2.54862904548645,
+      "learning_rate": 4.891511048751102e-05,
+      "loss": 0.6496,
+      "step": 629
+    },
+    {
+      "epoch": 0.925789860396767,
+      "grad_norm": 2.2401037216186523,
+      "learning_rate": 4.890761889907589e-05,
+      "loss": 0.6116,
+      "step": 630
+    },
+    {
+      "epoch": 0.9272593681116826,
+      "grad_norm": 2.6670756340026855,
+      "learning_rate": 4.890010211106795e-05,
+      "loss": 0.6842,
+      "step": 631
+    },
+    {
+      "epoch": 0.9287288758265981,
+      "grad_norm": 2.214515447616577,
+      "learning_rate": 4.889256013141018e-05,
+      "loss": 0.5209,
+      "step": 632
+    },
+    {
+      "epoch": 0.9301983835415136,
+      "grad_norm": 2.364154100418091,
+      "learning_rate": 4.888499296805214e-05,
+      "loss": 0.615,
+      "step": 633
+    },
+    {
+      "epoch": 0.9316678912564291,
+      "grad_norm": 2.3102328777313232,
+      "learning_rate": 4.8877400628969906e-05,
+      "loss": 0.6594,
+      "step": 634
+    },
+    {
+      "epoch": 0.9331373989713446,
+      "grad_norm": 2.227790117263794,
+      "learning_rate": 4.886978312216611e-05,
+      "loss": 0.6514,
+      "step": 635
+    },
+    {
+      "epoch": 0.9346069066862601,
+      "grad_norm": 2.3620498180389404,
+      "learning_rate": 4.88621404556699e-05,
+      "loss": 0.5684,
+      "step": 636
+    },
+    {
+      "epoch": 0.9360764144011756,
+      "grad_norm": 2.1664891242980957,
+      "learning_rate": 4.885447263753696e-05,
+      "loss": 0.4358,
+      "step": 637
+    },
+    {
+      "epoch": 0.9375459221160911,
+      "grad_norm": 2.7512993812561035,
+      "learning_rate": 4.884677967584945e-05,
+      "loss": 0.6542,
+      "step": 638
+    },
+    {
+      "epoch": 0.9390154298310066,
+      "grad_norm": 2.8794596195220947,
+      "learning_rate": 4.883906157871608e-05,
+      "loss": 0.7703,
+      "step": 639
+    },
+    {
+      "epoch": 0.9404849375459221,
+      "grad_norm": 2.872744560241699,
+      "learning_rate": 4.8831318354272015e-05,
+      "loss": 0.8021,
+      "step": 640
+    },
+    {
+      "epoch": 0.9419544452608376,
+      "grad_norm": 2.1877517700195312,
+      "learning_rate": 4.882355001067892e-05,
+      "loss": 0.4885,
+      "step": 641
+    },
+    {
+      "epoch": 0.9434239529757531,
+      "grad_norm": 2.613196611404419,
+      "learning_rate": 4.881575655612493e-05,
+      "loss": 0.7238,
+      "step": 642
+    },
+    {
+      "epoch": 0.9448934606906686,
+      "grad_norm": 3.4381418228149414,
+      "learning_rate": 4.880793799882466e-05,
+      "loss": 0.7552,
+      "step": 643
+    },
+    {
+      "epoch": 0.9463629684055841,
+      "grad_norm": 2.633586883544922,
+      "learning_rate": 4.8800094347019184e-05,
+      "loss": 0.7187,
+      "step": 644
+    },
+    {
+      "epoch": 0.9478324761204996,
+      "grad_norm": 2.2368478775024414,
+      "learning_rate": 4.8792225608976e-05,
+      "loss": 0.5735,
+      "step": 645
+    },
+    {
+      "epoch": 0.9493019838354151,
+      "grad_norm": 2.1886346340179443,
+      "learning_rate": 4.878433179298909e-05,
+      "loss": 0.6552,
+      "step": 646
+    },
+    {
+      "epoch": 0.9507714915503307,
+      "grad_norm": 2.1210744380950928,
+      "learning_rate": 4.877641290737884e-05,
+      "loss": 0.6248,
+      "step": 647
+    },
+    {
+      "epoch": 0.9522409992652462,
+      "grad_norm": 2.004408359527588,
+      "learning_rate": 4.876846896049207e-05,
+      "loss": 0.5047,
+      "step": 648
+    },
+    {
+      "epoch": 0.9537105069801617,
+      "grad_norm": 3.0902316570281982,
+      "learning_rate": 4.8760499960701996e-05,
+      "loss": 0.7987,
+      "step": 649
+    },
+    {
+      "epoch": 0.9551800146950772,
+      "grad_norm": 2.3769571781158447,
+      "learning_rate": 4.875250591640828e-05,
+      "loss": 0.6185,
+      "step": 650
+    },
+    {
+      "epoch": 0.9566495224099927,
+      "grad_norm": 1.8998838663101196,
+      "learning_rate": 4.874448683603695e-05,
+      "loss": 0.4395,
+      "step": 651
+    },
+    {
+      "epoch": 0.9581190301249082,
+      "grad_norm": 2.5027642250061035,
+      "learning_rate": 4.8736442728040435e-05,
+      "loss": 0.5241,
+      "step": 652
+    },
+    {
+      "epoch": 0.9595885378398237,
+      "grad_norm": 2.833956003189087,
+      "learning_rate": 4.872837360089754e-05,
+      "loss": 0.7559,
+      "step": 653
+    },
+    {
+      "epoch": 0.9610580455547392,
+      "grad_norm": 2.516974687576294,
+      "learning_rate": 4.872027946311344e-05,
+      "loss": 0.6711,
+      "step": 654
+    },
+    {
+      "epoch": 0.9625275532696547,
+      "grad_norm": 2.4347918033599854,
+      "learning_rate": 4.8712160323219677e-05,
+      "loss": 0.5306,
+      "step": 655
+    },
+    {
+      "epoch": 0.9639970609845702,
+      "grad_norm": 1.8865374326705933,
+      "learning_rate": 4.870401618977415e-05,
+      "loss": 0.4762,
+      "step": 656
+    },
+    {
+      "epoch": 0.9654665686994857,
+      "grad_norm": 2.1358063220977783,
+      "learning_rate": 4.869584707136109e-05,
+      "loss": 0.4672,
+      "step": 657
+    },
+    {
+      "epoch": 0.9669360764144012,
+      "grad_norm": 2.0028703212738037,
+      "learning_rate": 4.868765297659108e-05,
+      "loss": 0.5768,
+      "step": 658
+    },
+    {
+      "epoch": 0.9684055841293167,
+      "grad_norm": 2.3724420070648193,
+      "learning_rate": 4.8679433914101006e-05,
+      "loss": 0.4921,
+      "step": 659
+    },
+    {
+      "epoch": 0.9698750918442322,
+      "grad_norm": 1.8501986265182495,
+      "learning_rate": 4.867118989255409e-05,
+      "loss": 0.4748,
+      "step": 660
+    },
+    {
+      "epoch": 0.9713445995591476,
+      "grad_norm": 1.954079508781433,
+      "learning_rate": 4.8662920920639866e-05,
+      "loss": 0.4378,
+      "step": 661
+    },
+    {
+      "epoch": 0.9728141072740631,
+      "grad_norm": 2.556483507156372,
+      "learning_rate": 4.8654627007074135e-05,
+      "loss": 0.8358,
+      "step": 662
+    },
+    {
+      "epoch": 0.9742836149889786,
+      "grad_norm": 2.5413832664489746,
+      "learning_rate": 4.864630816059903e-05,
+      "loss": 0.7369,
+      "step": 663
+    },
+    {
+      "epoch": 0.9757531227038942,
+      "grad_norm": 2.136522054672241,
+      "learning_rate": 4.8637964389982926e-05,
+      "loss": 0.5887,
+      "step": 664
+    },
+    {
+      "epoch": 0.9772226304188097,
+      "grad_norm": 2.746372699737549,
+      "learning_rate": 4.862959570402049e-05,
+      "loss": 0.6063,
+      "step": 665
+    },
+    {
+      "epoch": 0.9786921381337252,
+      "grad_norm": 2.1235287189483643,
+      "learning_rate": 4.862120211153266e-05,
+      "loss": 0.6117,
+      "step": 666
+    },
+    {
+      "epoch": 0.9801616458486407,
+      "grad_norm": 2.0644116401672363,
+      "learning_rate": 4.8612783621366595e-05,
+      "loss": 0.5506,
+      "step": 667
+    },
+    {
+      "epoch": 0.9816311535635562,
+      "grad_norm": 2.639031171798706,
+      "learning_rate": 4.8604340242395724e-05,
+      "loss": 0.7263,
+      "step": 668
+    },
+    {
+      "epoch": 0.9831006612784717,
+      "grad_norm": 2.18987774848938,
+      "learning_rate": 4.8595871983519704e-05,
+      "loss": 0.7063,
+      "step": 669
+    },
+    {
+      "epoch": 0.9845701689933872,
+      "grad_norm": 2.563066005706787,
+      "learning_rate": 4.858737885366441e-05,
+      "loss": 0.674,
+      "step": 670
+    },
+    {
+      "epoch": 0.9860396767083027,
+      "grad_norm": 2.5474982261657715,
+      "learning_rate": 4.857886086178194e-05,
+      "loss": 0.6576,
+      "step": 671
+    },
+    {
+      "epoch": 0.9875091844232182,
+      "grad_norm": 2.199758291244507,
+      "learning_rate": 4.857031801685058e-05,
+      "loss": 0.4906,
+      "step": 672
+    },
+    {
+      "epoch": 0.9889786921381337,
+      "grad_norm": 2.1076645851135254,
+      "learning_rate": 4.856175032787485e-05,
+      "loss": 0.4187,
+      "step": 673
+    },
+    {
+      "epoch": 0.9904481998530492,
+      "grad_norm": 2.205265998840332,
+      "learning_rate": 4.8553157803885404e-05,
+      "loss": 0.5849,
+      "step": 674
+    },
+    {
+      "epoch": 0.9919177075679647,
+      "grad_norm": 2.188340663909912,
+      "learning_rate": 4.8544540453939124e-05,
+      "loss": 0.5217,
+      "step": 675
+    },
+    {
+      "epoch": 0.9933872152828802,
+      "grad_norm": 1.7291617393493652,
+      "learning_rate": 4.853589828711903e-05,
+      "loss": 0.47,
+      "step": 676
+    },
+    {
+      "epoch": 0.9948567229977957,
+      "grad_norm": 2.1585612297058105,
+      "learning_rate": 4.85272313125343e-05,
+      "loss": 0.5893,
+      "step": 677
+    },
+    {
+      "epoch": 0.9963262307127112,
+      "grad_norm": 2.17795991897583,
+      "learning_rate": 4.851853953932026e-05,
+      "loss": 0.5602,
+      "step": 678
+    },
+    {
+      "epoch": 0.9977957384276267,
+      "grad_norm": 2.1016292572021484,
+      "learning_rate": 4.8509822976638395e-05,
+      "loss": 0.6033,
+      "step": 679
+    },
+    {
+      "epoch": 0.9992652461425422,
+      "grad_norm": 2.5465779304504395,
+      "learning_rate": 4.850108163367631e-05,
+      "loss": 0.7276,
+      "step": 680
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.688074827194214,
+      "learning_rate": 4.849231551964771e-05,
+      "loss": 0.594,
+      "step": 681
+    },
+    {
+      "epoch": 1.0,
+      "eval_loss": 0.6192097067832947,
+      "eval_runtime": 30.9747,
+      "eval_samples_per_second": 4.649,
+      "eval_steps_per_second": 4.649,
+      "step": 681
+    },
+    {
+      "epoch": 1.0014695077149156,
+      "grad_norm": 2.3628222942352295,
+      "learning_rate": 4.8483524643792436e-05,
+      "loss": 0.471,
+      "step": 682
+    },
+    {
+      "epoch": 1.002939015429831,
+      "grad_norm": 1.538787603378296,
+      "learning_rate": 4.8474709015376416e-05,
+      "loss": 0.3746,
+      "step": 683
+    },
+    {
+      "epoch": 1.0044085231447466,
+      "grad_norm": 1.7308114767074585,
+      "learning_rate": 4.846586864369167e-05,
+      "loss": 0.3796,
+      "step": 684
+    },
+    {
+      "epoch": 1.005878030859662,
+      "grad_norm": 1.6080310344696045,
+      "learning_rate": 4.845700353805629e-05,
+      "loss": 0.3661,
+      "step": 685
+    },
+    {
+      "epoch": 1.0073475385745776,
+      "grad_norm": 2.0169012546539307,
+      "learning_rate": 4.844811370781446e-05,
+      "loss": 0.4163,
+      "step": 686
+    },
+    {
+      "epoch": 1.008817046289493,
+      "grad_norm": 1.9985781908035278,
+      "learning_rate": 4.843919916233639e-05,
+      "loss": 0.4462,
+      "step": 687
+    },
+    {
+      "epoch": 1.0102865540044086,
+      "grad_norm": 2.1932809352874756,
+      "learning_rate": 4.8430259911018396e-05,
+      "loss": 0.3695,
+      "step": 688
+    },
+    {
+      "epoch": 1.011756061719324,
+      "grad_norm": 2.6596124172210693,
+      "learning_rate": 4.842129596328277e-05,
+      "loss": 0.55,
+      "step": 689
+    },
+    {
+      "epoch": 1.0132255694342396,
+      "grad_norm": 2.9033000469207764,
+      "learning_rate": 4.841230732857788e-05,
+      "loss": 0.386,
+      "step": 690
+    },
+    {
+      "epoch": 1.014695077149155,
+      "grad_norm": 3.7226250171661377,
+      "learning_rate": 4.8403294016378095e-05,
+      "loss": 0.4591,
+      "step": 691
+    },
+    {
+      "epoch": 1.0161645848640706,
+      "grad_norm": 2.84456467628479,
+      "learning_rate": 4.8394256036183816e-05,
+      "loss": 0.3732,
+      "step": 692
+    },
+    {
+      "epoch": 1.017634092578986,
+      "grad_norm": 2.2714834213256836,
+      "learning_rate": 4.8385193397521425e-05,
+      "loss": 0.385,
+      "step": 693
+    },
+    {
+      "epoch": 1.0191036002939016,
+      "grad_norm": 2.8564984798431396,
+      "learning_rate": 4.8376106109943286e-05,
+      "loss": 0.4787,
+      "step": 694
+    },
+    {
+      "epoch": 1.020573108008817,
+      "grad_norm": 2.3014914989471436,
+      "learning_rate": 4.836699418302777e-05,
+      "loss": 0.4418,
+      "step": 695
+    },
+    {
+      "epoch": 1.0220426157237326,
+      "grad_norm": 2.216087579727173,
+      "learning_rate": 4.835785762637921e-05,
+      "loss": 0.3056,
+      "step": 696
+    },
+    {
+      "epoch": 1.023512123438648,
+      "grad_norm": 2.395174264907837,
+      "learning_rate": 4.8348696449627886e-05,
+      "loss": 0.3789,
+      "step": 697
+    },
+    {
+      "epoch": 1.0249816311535636,
+      "grad_norm": 2.1030936241149902,
+      "learning_rate": 4.8339510662430046e-05,
+      "loss": 0.2633,
+      "step": 698
+    },
+    {
+      "epoch": 1.0264511388684792,
+      "grad_norm": 2.26007080078125,
+      "learning_rate": 4.8330300274467874e-05,
+      "loss": 0.2648,
+      "step": 699
+    },
+    {
+      "epoch": 1.0279206465833945,
+      "grad_norm": 2.8508899211883545,
+      "learning_rate": 4.832106529544948e-05,
+      "loss": 0.607,
+      "step": 700
+    },
+    {
+      "epoch": 1.0293901542983102,
+      "grad_norm": 2.3766891956329346,
+      "learning_rate": 4.8311805735108894e-05,
+      "loss": 0.4333,
+      "step": 701
+    },
+    {
+      "epoch": 1.0308596620132255,
+      "grad_norm": 2.003467082977295,
+      "learning_rate": 4.830252160320606e-05,
+      "loss": 0.3499,
+      "step": 702
+    },
+    {
+      "epoch": 1.0323291697281411,
+      "grad_norm": 2.1121034622192383,
+      "learning_rate": 4.829321290952683e-05,
+      "loss": 0.4199,
+      "step": 703
+    },
+    {
+      "epoch": 1.0337986774430565,
+      "grad_norm": 2.2768359184265137,
+      "learning_rate": 4.8283879663882906e-05,
+      "loss": 0.3619,
+      "step": 704
+    },
+    {
+      "epoch": 1.0352681851579721,
+      "grad_norm": 2.2349584102630615,
+      "learning_rate": 4.8274521876111925e-05,
+      "loss": 0.3972,
+      "step": 705
+    },
+    {
+      "epoch": 1.0367376928728875,
+      "grad_norm": 2.003754138946533,
+      "learning_rate": 4.8265139556077346e-05,
+      "loss": 0.3727,
+      "step": 706
+    },
+    {
+      "epoch": 1.0382072005878031,
+      "grad_norm": 2.281745433807373,
+      "learning_rate": 4.825573271366851e-05,
+      "loss": 0.4102,
+      "step": 707
+    },
+    {
+      "epoch": 1.0396767083027185,
+      "grad_norm": 2.683112144470215,
+      "learning_rate": 4.8246301358800604e-05,
+      "loss": 0.4972,
+      "step": 708
+    },
+    {
+      "epoch": 1.0411462160176341,
+      "grad_norm": 2.282343626022339,
+      "learning_rate": 4.823684550141464e-05,
+      "loss": 0.4038,
+      "step": 709
+    },
+    {
+      "epoch": 1.0426157237325495,
+      "grad_norm": 2.050415277481079,
+      "learning_rate": 4.822736515147748e-05,
+      "loss": 0.3468,
+      "step": 710
+    },
+    {
+      "epoch": 1.0440852314474651,
+      "grad_norm": 1.8552284240722656,
+      "learning_rate": 4.821786031898176e-05,
+      "loss": 0.4349,
+      "step": 711
+    },
+    {
+      "epoch": 1.0455547391623805,
+      "grad_norm": 2.133265733718872,
+      "learning_rate": 4.820833101394597e-05,
+      "loss": 0.3335,
+      "step": 712
+    },
+    {
+      "epoch": 1.0470242468772961,
+      "grad_norm": 1.6545878648757935,
+      "learning_rate": 4.8198777246414373e-05,
+      "loss": 0.2708,
+      "step": 713
+    },
+    {
+      "epoch": 1.0484937545922115,
+      "grad_norm": 1.7532316446304321,
+      "learning_rate": 4.8189199026457e-05,
+      "loss": 0.3285,
+      "step": 714
+    },
+    {
+      "epoch": 1.049963262307127,
+      "grad_norm": 2.213719367980957,
+      "learning_rate": 4.817959636416969e-05,
+      "loss": 0.4325,
+      "step": 715
+    },
+    {
+      "epoch": 1.0514327700220427,
+      "grad_norm": 2.514274835586548,
+      "learning_rate": 4.8169969269674016e-05,
+      "loss": 0.4303,
+      "step": 716
+    },
+    {
+      "epoch": 1.052902277736958,
+      "grad_norm": 2.3625645637512207,
+      "learning_rate": 4.816031775311732e-05,
+      "loss": 0.4934,
+      "step": 717
+    },
+    {
+      "epoch": 1.0543717854518737,
+      "grad_norm": 2.4587574005126953,
+      "learning_rate": 4.8150641824672685e-05,
+      "loss": 0.3869,
+      "step": 718
+    },
+    {
+      "epoch": 1.055841293166789,
+      "grad_norm": 2.7009196281433105,
+      "learning_rate": 4.814094149453891e-05,
+      "loss": 0.4905,
+      "step": 719
+    },
+    {
+      "epoch": 1.0573108008817047,
+      "grad_norm": 1.9971612691879272,
+      "learning_rate": 4.813121677294054e-05,
+      "loss": 0.3295,
+      "step": 720
+    },
+    {
+      "epoch": 1.05878030859662,
+      "grad_norm": 2.463809013366699,
+      "learning_rate": 4.81214676701278e-05,
+      "loss": 0.4822,
+      "step": 721
+    },
+    {
+      "epoch": 1.0602498163115357,
+      "grad_norm": 2.7179911136627197,
+      "learning_rate": 4.811169419637663e-05,
+      "loss": 0.434,
+      "step": 722
+    },
+    {
+      "epoch": 1.061719324026451,
+      "grad_norm": 1.7837159633636475,
+      "learning_rate": 4.8101896361988675e-05,
+      "loss": 0.2667,
+      "step": 723
+    },
+    {
+      "epoch": 1.0631888317413667,
+      "grad_norm": 2.085952043533325,
+      "learning_rate": 4.8092074177291215e-05,
+      "loss": 0.4337,
+      "step": 724
+    },
+    {
+      "epoch": 1.064658339456282,
+      "grad_norm": 2.304379940032959,
+      "learning_rate": 4.808222765263724e-05,
+      "loss": 0.3583,
+      "step": 725
+    },
+    {
+      "epoch": 1.0661278471711977,
+      "grad_norm": 2.5706167221069336,
+      "learning_rate": 4.807235679840536e-05,
+      "loss": 0.4504,
+      "step": 726
+    },
+    {
+      "epoch": 1.067597354886113,
+      "grad_norm": 2.233501434326172,
+      "learning_rate": 4.806246162499985e-05,
+      "loss": 0.4371,
+      "step": 727
+    },
+    {
+      "epoch": 1.0690668626010287,
+      "grad_norm": 3.602660655975342,
+      "learning_rate": 4.805254214285061e-05,
+      "loss": 0.5581,
+      "step": 728
+    },
+    {
+      "epoch": 1.070536370315944,
+      "grad_norm": 1.9689836502075195,
+      "learning_rate": 4.804259836241318e-05,
+      "loss": 0.4281,
+      "step": 729
+    },
+    {
+      "epoch": 1.0720058780308597,
+      "grad_norm": 2.070441246032715,
+      "learning_rate": 4.8032630294168675e-05,
+      "loss": 0.3118,
+      "step": 730
+    },
+    {
+      "epoch": 1.073475385745775,
+      "grad_norm": 1.7218447923660278,
+      "learning_rate": 4.802263794862385e-05,
+      "loss": 0.3799,
+      "step": 731
+    },
+    {
+      "epoch": 1.0749448934606907,
+      "grad_norm": 2.1031031608581543,
+      "learning_rate": 4.8012621336311016e-05,
+      "loss": 0.2832,
+      "step": 732
+    },
+    {
+      "epoch": 1.0764144011756063,
+      "grad_norm": 1.6383007764816284,
+      "learning_rate": 4.8002580467788086e-05,
+      "loss": 0.296,
+      "step": 733
+    },
+    {
+      "epoch": 1.0778839088905217,
+      "grad_norm": 1.9484933614730835,
+      "learning_rate": 4.7992515353638525e-05,
+      "loss": 0.3157,
+      "step": 734
+    },
+    {
+      "epoch": 1.0793534166054373,
+      "grad_norm": 2.494256019592285,
+      "learning_rate": 4.798242600447137e-05,
+      "loss": 0.4569,
+      "step": 735
+    },
+    {
+      "epoch": 1.0808229243203527,
+      "grad_norm": 1.3053300380706787,
+      "learning_rate": 4.797231243092118e-05,
+      "loss": 0.2223,
+      "step": 736
+    },
+    {
+      "epoch": 1.0822924320352683,
+      "grad_norm": 2.5402936935424805,
+      "learning_rate": 4.796217464364808e-05,
+      "loss": 0.4689,
+      "step": 737
+    },
+    {
+      "epoch": 1.0837619397501836,
+      "grad_norm": 2.2366368770599365,
+      "learning_rate": 4.795201265333767e-05,
+      "loss": 0.3203,
+      "step": 738
+    },
+    {
+      "epoch": 1.0852314474650993,
+      "grad_norm": 2.116938352584839,
+      "learning_rate": 4.7941826470701115e-05,
+      "loss": 0.3649,
+      "step": 739
+    },
+    {
+      "epoch": 1.0867009551800146,
+      "grad_norm": 2.5721476078033447,
+      "learning_rate": 4.793161610647504e-05,
+      "loss": 0.5025,
+      "step": 740
+    },
+    {
+      "epoch": 1.0881704628949302,
+      "grad_norm": 2.3269119262695312,
+      "learning_rate": 4.792138157142158e-05,
+      "loss": 0.3952,
+      "step": 741
+    },
+    {
+      "epoch": 1.0896399706098456,
+      "grad_norm": 2.008136749267578,
+      "learning_rate": 4.7911122876328317e-05,
+      "loss": 0.3048,
+      "step": 742
+    },
+    {
+      "epoch": 1.0911094783247612,
+      "grad_norm": 2.795898199081421,
+      "learning_rate": 4.790084003200834e-05,
+      "loss": 0.4398,
+      "step": 743
+    },
+    {
+      "epoch": 1.0925789860396766,
+      "grad_norm": 3.4373717308044434,
+      "learning_rate": 4.7890533049300166e-05,
+      "loss": 0.5439,
+      "step": 744
+    },
+    {
+      "epoch": 1.0940484937545922,
+      "grad_norm": 2.261160135269165,
+      "learning_rate": 4.788020193906776e-05,
+      "loss": 0.4454,
+      "step": 745
+    },
+    {
+      "epoch": 1.0955180014695076,
+      "grad_norm": 1.9496560096740723,
+      "learning_rate": 4.786984671220053e-05,
+      "loss": 0.384,
+      "step": 746
+    },
+    {
+      "epoch": 1.0969875091844232,
+      "grad_norm": 2.191901922225952,
+      "learning_rate": 4.785946737961327e-05,
+      "loss": 0.3248,
+      "step": 747
+    },
+    {
+      "epoch": 1.0984570168993386,
+      "grad_norm": 2.0085935592651367,
+      "learning_rate": 4.784906395224623e-05,
+      "loss": 0.3802,
+      "step": 748
+    },
+    {
+      "epoch": 1.0999265246142542,
+      "grad_norm": 1.74215829372406,
+      "learning_rate": 4.783863644106502e-05,
+      "loss": 0.3574,
+      "step": 749
+    },
+    {
+      "epoch": 1.1013960323291698,
+      "grad_norm": 1.9457191228866577,
+      "learning_rate": 4.782818485706065e-05,
+      "loss": 0.3618,
+      "step": 750
+    },
+    {
+      "epoch": 1.1028655400440852,
+      "grad_norm": 2.26507306098938,
+      "learning_rate": 4.7817709211249515e-05,
+      "loss": 0.3816,
+      "step": 751
+    },
+    {
+      "epoch": 1.1043350477590008,
+      "grad_norm": 1.966572642326355,
+      "learning_rate": 4.7807209514673336e-05,
+      "loss": 0.3867,
+      "step": 752
+    },
+    {
+      "epoch": 1.1058045554739162,
+      "grad_norm": 3.1015684604644775,
+      "learning_rate": 4.779668577839921e-05,
+      "loss": 0.4897,
+      "step": 753
+    },
+    {
+      "epoch": 1.1072740631888318,
+      "grad_norm": 2.573324203491211,
+      "learning_rate": 4.7786138013519585e-05,
+      "loss": 0.4608,
+      "step": 754
+    },
+    {
+      "epoch": 1.1087435709037472,
+      "grad_norm": 2.2066636085510254,
+      "learning_rate": 4.777556623115221e-05,
+      "loss": 0.4472,
+      "step": 755
+    },
+    {
+      "epoch": 1.1102130786186628,
+      "grad_norm": 1.8826298713684082,
+      "learning_rate": 4.7764970442440164e-05,
+      "loss": 0.295,
+      "step": 756
+    },
+    {
+      "epoch": 1.1116825863335782,
+      "grad_norm": 3.54438853263855,
+      "learning_rate": 4.775435065855183e-05,
+      "loss": 0.4784,
+      "step": 757
+    },
+    {
+      "epoch": 1.1131520940484938,
+      "grad_norm": 2.284369945526123,
+      "learning_rate": 4.7743706890680865e-05,
+      "loss": 0.3844,
+      "step": 758
+    },
+    {
+      "epoch": 1.1146216017634092,
+      "grad_norm": 2.363003969192505,
+      "learning_rate": 4.7733039150046234e-05,
+      "loss": 0.4169,
+      "step": 759
+    },
+    {
+      "epoch": 1.1160911094783248,
+      "grad_norm": 1.8094215393066406,
+      "learning_rate": 4.7722347447892155e-05,
+      "loss": 0.315,
+      "step": 760
+    },
+    {
+      "epoch": 1.1175606171932402,
+      "grad_norm": 2.5151467323303223,
+      "learning_rate": 4.7711631795488096e-05,
+      "loss": 0.3489,
+      "step": 761
+    },
+    {
+      "epoch": 1.1190301249081558,
+      "grad_norm": 1.907175064086914,
+      "learning_rate": 4.7700892204128775e-05,
+      "loss": 0.3152,
+      "step": 762
+    },
+    {
+      "epoch": 1.1204996326230712,
+      "grad_norm": 2.0628597736358643,
+      "learning_rate": 4.769012868513416e-05,
+      "loss": 0.3712,
+      "step": 763
+    },
+    {
+      "epoch": 1.1219691403379868,
+      "grad_norm": 1.771032691001892,
+      "learning_rate": 4.767934124984941e-05,
+      "loss": 0.3461,
+      "step": 764
+    },
+    {
+      "epoch": 1.1234386480529022,
+      "grad_norm": 2.039865493774414,
+      "learning_rate": 4.766852990964492e-05,
+      "loss": 0.3558,
+      "step": 765
+    },
+    {
+      "epoch": 1.1249081557678178,
+      "grad_norm": 1.9117597341537476,
+      "learning_rate": 4.765769467591625e-05,
+      "loss": 0.3033,
+      "step": 766
+    },
+    {
+      "epoch": 1.1263776634827334,
+      "grad_norm": 1.814257025718689,
+      "learning_rate": 4.764683556008418e-05,
+      "loss": 0.2829,
+      "step": 767
+    },
+    {
+      "epoch": 1.1278471711976488,
+      "grad_norm": 2.197000741958618,
+      "learning_rate": 4.7635952573594637e-05,
+      "loss": 0.3621,
+      "step": 768
+    },
+    {
+      "epoch": 1.1293166789125644,
+      "grad_norm": 1.8606325387954712,
+      "learning_rate": 4.7625045727918724e-05,
+      "loss": 0.2652,
+      "step": 769
+    },
+    {
+      "epoch": 1.1307861866274798,
+      "grad_norm": 2.251941442489624,
+      "learning_rate": 4.7614115034552686e-05,
+      "loss": 0.3392,
+      "step": 770
+    },
+    {
+      "epoch": 1.1322556943423954,
+      "grad_norm": 2.1944212913513184,
+      "learning_rate": 4.76031605050179e-05,
+      "loss": 0.3834,
+      "step": 771
+    },
+    {
+      "epoch": 1.1337252020573108,
+      "grad_norm": 1.9881553649902344,
+      "learning_rate": 4.7592182150860876e-05,
+      "loss": 0.3702,
+      "step": 772
+    },
+    {
+      "epoch": 1.1351947097722264,
+      "grad_norm": 3.4269587993621826,
+      "learning_rate": 4.758117998365322e-05,
+      "loss": 0.5975,
+      "step": 773
+    },
+    {
+      "epoch": 1.1366642174871417,
+      "grad_norm": 1.7125985622406006,
+      "learning_rate": 4.757015401499167e-05,
+      "loss": 0.2979,
+      "step": 774
+    },
+    {
+      "epoch": 1.1381337252020574,
+      "grad_norm": 1.8743551969528198,
+      "learning_rate": 4.7559104256498023e-05,
+      "loss": 0.3676,
+      "step": 775
+    },
+    {
+      "epoch": 1.1396032329169727,
+      "grad_norm": 1.8048675060272217,
+      "learning_rate": 4.754803071981916e-05,
+      "loss": 0.335,
+      "step": 776
+    },
+    {
+      "epoch": 1.1410727406318883,
+      "grad_norm": 2.1437811851501465,
+      "learning_rate": 4.753693341662702e-05,
+      "loss": 0.3599,
+      "step": 777
+    },
+    {
+      "epoch": 1.1425422483468037,
+      "grad_norm": 2.03254771232605,
+      "learning_rate": 4.7525812358618596e-05,
+      "loss": 0.3443,
+      "step": 778
+    },
+    {
+      "epoch": 1.1440117560617193,
+      "grad_norm": 2.6306824684143066,
+      "learning_rate": 4.751466755751594e-05,
+      "loss": 0.3835,
+      "step": 779
+    },
+    {
+      "epoch": 1.145481263776635,
+      "grad_norm": 2.6445915699005127,
+      "learning_rate": 4.750349902506609e-05,
+      "loss": 0.4525,
+      "step": 780
+    },
+    {
+      "epoch": 1.1469507714915503,
+      "grad_norm": 1.7030630111694336,
+      "learning_rate": 4.749230677304114e-05,
+      "loss": 0.3131,
+      "step": 781
+    },
+    {
+      "epoch": 1.1484202792064657,
+      "grad_norm": 2.2668240070343018,
+      "learning_rate": 4.748109081323814e-05,
+      "loss": 0.4031,
+      "step": 782
+    },
+    {
+      "epoch": 1.1498897869213813,
+      "grad_norm": 1.869971752166748,
+      "learning_rate": 4.7469851157479177e-05,
+      "loss": 0.3444,
+      "step": 783
+    },
+    {
+      "epoch": 1.151359294636297,
+      "grad_norm": 3.237847089767456,
+      "learning_rate": 4.7458587817611276e-05,
+      "loss": 0.6194,
+      "step": 784
+    },
+    {
+      "epoch": 1.1528288023512123,
+      "grad_norm": 2.437415361404419,
+      "learning_rate": 4.7447300805506455e-05,
+      "loss": 0.3726,
+      "step": 785
+    },
+    {
+      "epoch": 1.154298310066128,
+      "grad_norm": 3.0447607040405273,
+      "learning_rate": 4.7435990133061656e-05,
+      "loss": 0.4758,
+      "step": 786
+    },
+    {
+      "epoch": 1.1557678177810433,
+      "grad_norm": 1.934402346611023,
+      "learning_rate": 4.742465581219878e-05,
+      "loss": 0.3368,
+      "step": 787
+    },
+    {
+      "epoch": 1.157237325495959,
+      "grad_norm": 1.9989057779312134,
+      "learning_rate": 4.7413297854864655e-05,
+      "loss": 0.3692,
+      "step": 788
+    },
+    {
+      "epoch": 1.1587068332108743,
+      "grad_norm": 3.5975356101989746,
+      "learning_rate": 4.7401916273031e-05,
+      "loss": 0.4549,
+      "step": 789
+    },
+    {
+      "epoch": 1.16017634092579,
+      "grad_norm": 1.9463549852371216,
+      "learning_rate": 4.739051107869446e-05,
+      "loss": 0.3776,
+      "step": 790
+    },
+    {
+      "epoch": 1.1616458486407053,
+      "grad_norm": 3.553086280822754,
+      "learning_rate": 4.7379082283876566e-05,
+      "loss": 0.7636,
+      "step": 791
+    },
+    {
+      "epoch": 1.163115356355621,
+      "grad_norm": 1.7739683389663696,
+      "learning_rate": 4.73676299006237e-05,
+      "loss": 0.3197,
+      "step": 792
+    },
+    {
+      "epoch": 1.1645848640705363,
+      "grad_norm": 2.253840446472168,
+      "learning_rate": 4.7356153941007143e-05,
+      "loss": 0.4219,
+      "step": 793
+    },
+    {
+      "epoch": 1.166054371785452,
+      "grad_norm": 2.4121429920196533,
+      "learning_rate": 4.7344654417123e-05,
+      "loss": 0.4198,
+      "step": 794
+    },
+    {
+      "epoch": 1.1675238795003673,
+      "grad_norm": 1.746691346168518,
+      "learning_rate": 4.733313134109223e-05,
+      "loss": 0.3551,
+      "step": 795
+    },
+    {
+      "epoch": 1.168993387215283,
+      "grad_norm": 2.172924041748047,
+      "learning_rate": 4.7321584725060597e-05,
+      "loss": 0.4276,
+      "step": 796
+    },
+    {
+      "epoch": 1.1704628949301985,
+      "grad_norm": 2.308032512664795,
+      "learning_rate": 4.7310014581198695e-05,
+      "loss": 0.4204,
+      "step": 797
+    },
+    {
+      "epoch": 1.171932402645114,
+      "grad_norm": 2.0524096488952637,
+      "learning_rate": 4.7298420921701926e-05,
+      "loss": 0.2878,
+      "step": 798
+    },
+    {
+      "epoch": 1.1734019103600293,
+      "grad_norm": 1.982011318206787,
+      "learning_rate": 4.728680375879044e-05,
+      "loss": 0.4045,
+      "step": 799
+    },
+    {
+      "epoch": 1.1748714180749449,
+      "grad_norm": 1.6454319953918457,
+      "learning_rate": 4.72751631047092e-05,
+      "loss": 0.3604,
+      "step": 800
+    },
+    {
+      "epoch": 1.1763409257898605,
+      "grad_norm": 1.9815795421600342,
+      "learning_rate": 4.726349897172791e-05,
+      "loss": 0.4038,
+      "step": 801
+    },
+    {
+      "epoch": 1.1778104335047759,
+      "grad_norm": 2.0404534339904785,
+      "learning_rate": 4.725181137214103e-05,
+      "loss": 0.4065,
+      "step": 802
+    },
+    {
+      "epoch": 1.1792799412196915,
+      "grad_norm": 2.8057069778442383,
+      "learning_rate": 4.7240100318267746e-05,
+      "loss": 0.4932,
+      "step": 803
+    },
+    {
+      "epoch": 1.1807494489346069,
+      "grad_norm": 2.239226818084717,
+      "learning_rate": 4.722836582245197e-05,
+      "loss": 0.5191,
+      "step": 804
+    },
+    {
+      "epoch": 1.1822189566495225,
+      "grad_norm": 1.8206257820129395,
+      "learning_rate": 4.721660789706232e-05,
+      "loss": 0.3657,
+      "step": 805
+    },
+    {
+      "epoch": 1.1836884643644379,
+      "grad_norm": 1.8755921125411987,
+      "learning_rate": 4.720482655449212e-05,
+      "loss": 0.3797,
+      "step": 806
+    },
+    {
+      "epoch": 1.1851579720793535,
+      "grad_norm": 2.219597101211548,
+      "learning_rate": 4.719302180715938e-05,
+      "loss": 0.4876,
+      "step": 807
+    },
+    {
+      "epoch": 1.1866274797942689,
+      "grad_norm": 2.8050894737243652,
+      "learning_rate": 4.718119366750674e-05,
+      "loss": 0.5485,
+      "step": 808
+    },
+    {
+      "epoch": 1.1880969875091845,
+      "grad_norm": 2.2481648921966553,
+      "learning_rate": 4.716934214800155e-05,
+      "loss": 0.4205,
+      "step": 809
+    },
+    {
+      "epoch": 1.1895664952240999,
+      "grad_norm": 2.0312628746032715,
+      "learning_rate": 4.7157467261135765e-05,
+      "loss": 0.348,
+      "step": 810
+    },
+    {
+      "epoch": 1.1910360029390155,
+      "grad_norm": 2.0684547424316406,
+      "learning_rate": 4.7145569019425996e-05,
+      "loss": 0.3363,
+      "step": 811
+    },
+    {
+      "epoch": 1.1925055106539308,
+      "grad_norm": 2.1075944900512695,
+      "learning_rate": 4.713364743541345e-05,
+      "loss": 0.4618,
+      "step": 812
+    },
+    {
+      "epoch": 1.1939750183688465,
+      "grad_norm": 2.303858757019043,
+      "learning_rate": 4.7121702521663955e-05,
+      "loss": 0.4936,
+      "step": 813
+    },
+    {
+      "epoch": 1.195444526083762,
+      "grad_norm": 2.0791971683502197,
+      "learning_rate": 4.7109734290767905e-05,
+      "loss": 0.3503,
+      "step": 814
+    },
+    {
+      "epoch": 1.1969140337986774,
+      "grad_norm": 1.8620028495788574,
+      "learning_rate": 4.70977427553403e-05,
+      "loss": 0.3307,
+      "step": 815
+    },
+    {
+      "epoch": 1.1983835415135928,
+      "grad_norm": 2.7779459953308105,
+      "learning_rate": 4.70857279280207e-05,
+      "loss": 0.4942,
+      "step": 816
+    },
+    {
+      "epoch": 1.1998530492285084,
+      "grad_norm": 1.9387404918670654,
+      "learning_rate": 4.707368982147318e-05,
+      "loss": 0.3974,
+      "step": 817
+    },
+    {
+      "epoch": 1.201322556943424,
+      "grad_norm": 2.336399793624878,
+      "learning_rate": 4.70616284483864e-05,
+      "loss": 0.4428,
+      "step": 818
+    },
+    {
+      "epoch": 1.2027920646583394,
+      "grad_norm": 1.8475910425186157,
+      "learning_rate": 4.7049543821473506e-05,
+      "loss": 0.3464,
+      "step": 819
+    },
+    {
+      "epoch": 1.204261572373255,
+      "grad_norm": 1.9967962503433228,
+      "learning_rate": 4.703743595347219e-05,
+      "loss": 0.3196,
+      "step": 820
+    },
+    {
+      "epoch": 1.2057310800881704,
+      "grad_norm": 2.3950350284576416,
+      "learning_rate": 4.702530485714461e-05,
+      "loss": 0.4294,
+      "step": 821
+    },
+    {
+      "epoch": 1.207200587803086,
+      "grad_norm": 2.478525400161743,
+      "learning_rate": 4.701315054527743e-05,
+      "loss": 0.386,
+      "step": 822
+    },
+    {
+      "epoch": 1.2086700955180014,
+      "grad_norm": 2.5162229537963867,
+      "learning_rate": 4.700097303068177e-05,
+      "loss": 0.4049,
+      "step": 823
+    },
+    {
+      "epoch": 1.210139603232917,
+      "grad_norm": 2.346830368041992,
+      "learning_rate": 4.69887723261932e-05,
+      "loss": 0.497,
+      "step": 824
+    },
+    {
+      "epoch": 1.2116091109478324,
+      "grad_norm": 1.982297658920288,
+      "learning_rate": 4.6976548444671756e-05,
+      "loss": 0.2998,
+      "step": 825
+    },
+    {
+      "epoch": 1.213078618662748,
+      "grad_norm": 2.1650495529174805,
+      "learning_rate": 4.696430139900188e-05,
+      "loss": 0.4464,
+      "step": 826
+    },
+    {
+      "epoch": 1.2145481263776634,
+      "grad_norm": 1.5292145013809204,
+      "learning_rate": 4.695203120209245e-05,
+      "loss": 0.356,
+      "step": 827
+    },
+    {
+      "epoch": 1.216017634092579,
+      "grad_norm": 2.100552558898926,
+      "learning_rate": 4.6939737866876735e-05,
+      "loss": 0.3128,
+      "step": 828
+    },
+    {
+      "epoch": 1.2174871418074944,
+      "grad_norm": 1.9075181484222412,
+      "learning_rate": 4.69274214063124e-05,
+      "loss": 0.2672,
+      "step": 829
+    },
+    {
+      "epoch": 1.21895664952241,
+      "grad_norm": 2.0911717414855957,
+      "learning_rate": 4.691508183338147e-05,
+      "loss": 0.3383,
+      "step": 830
+    },
+    {
+      "epoch": 1.2204261572373256,
+      "grad_norm": 2.2416839599609375,
+      "learning_rate": 4.6902719161090345e-05,
+      "loss": 0.3254,
+      "step": 831
+    },
+    {
+      "epoch": 1.221895664952241,
+      "grad_norm": 2.680610179901123,
+      "learning_rate": 4.689033340246977e-05,
+      "loss": 0.5349,
+      "step": 832
+    },
+    {
+      "epoch": 1.2233651726671564,
+      "grad_norm": 2.8520190715789795,
+      "learning_rate": 4.6877924570574817e-05,
+      "loss": 0.4473,
+      "step": 833
+    },
+    {
+      "epoch": 1.224834680382072,
+      "grad_norm": 2.128718852996826,
+      "learning_rate": 4.6865492678484895e-05,
+      "loss": 0.3812,
+      "step": 834
+    },
+    {
+      "epoch": 1.2263041880969876,
+      "grad_norm": 2.2420125007629395,
+      "learning_rate": 4.685303773930371e-05,
+      "loss": 0.2951,
+      "step": 835
+    },
+    {
+      "epoch": 1.227773695811903,
+      "grad_norm": 1.9147225618362427,
+      "learning_rate": 4.684055976615924e-05,
+      "loss": 0.3715,
+      "step": 836
+    },
+    {
+      "epoch": 1.2292432035268186,
+      "grad_norm": 3.0332610607147217,
+      "learning_rate": 4.682805877220378e-05,
+      "loss": 0.4689,
+      "step": 837
+    },
+    {
+      "epoch": 1.230712711241734,
+      "grad_norm": 1.9656281471252441,
+      "learning_rate": 4.681553477061386e-05,
+      "loss": 0.3188,
+      "step": 838
+    },
+    {
+      "epoch": 1.2321822189566496,
+      "grad_norm": 2.277771472930908,
+      "learning_rate": 4.6802987774590275e-05,
+      "loss": 0.6169,
+      "step": 839
+    },
+    {
+      "epoch": 1.233651726671565,
+      "grad_norm": 2.310939073562622,
+      "learning_rate": 4.6790417797358055e-05,
+      "loss": 0.3917,
+      "step": 840
+    },
+    {
+      "epoch": 1.2351212343864806,
+      "grad_norm": 2.0788984298706055,
+      "learning_rate": 4.677782485216644e-05,
+      "loss": 0.4154,
+      "step": 841
+    },
+    {
+      "epoch": 1.236590742101396,
+      "grad_norm": 1.9377974271774292,
+      "learning_rate": 4.6765208952288906e-05,
+      "loss": 0.3674,
+      "step": 842
+    },
+    {
+      "epoch": 1.2380602498163116,
+      "grad_norm": 2.1602776050567627,
+      "learning_rate": 4.675257011102311e-05,
+      "loss": 0.4403,
+      "step": 843
+    },
+    {
+      "epoch": 1.239529757531227,
+      "grad_norm": 2.3176870346069336,
+      "learning_rate": 4.673990834169087e-05,
+      "loss": 0.439,
+      "step": 844
+    },
+    {
+      "epoch": 1.2409992652461426,
+      "grad_norm": 1.746071457862854,
+      "learning_rate": 4.672722365763821e-05,
+      "loss": 0.2822,
+      "step": 845
+    },
+    {
+      "epoch": 1.242468772961058,
+      "grad_norm": 1.963962435722351,
+      "learning_rate": 4.671451607223527e-05,
+      "loss": 0.4306,
+      "step": 846
+    },
+    {
+      "epoch": 1.2439382806759736,
+      "grad_norm": 1.6478948593139648,
+      "learning_rate": 4.670178559887637e-05,
+      "loss": 0.339,
+      "step": 847
+    },
+    {
+      "epoch": 1.2454077883908892,
+      "grad_norm": 1.8290305137634277,
+      "learning_rate": 4.66890322509799e-05,
+      "loss": 0.3283,
+      "step": 848
+    },
+    {
+      "epoch": 1.2468772961058046,
+      "grad_norm": 2.300769805908203,
+      "learning_rate": 4.667625604198842e-05,
+      "loss": 0.4195,
+      "step": 849
+    },
+    {
+      "epoch": 1.24834680382072,
+      "grad_norm": 2.0550520420074463,
+      "learning_rate": 4.6663456985368536e-05,
+      "loss": 0.4262,
+      "step": 850
+    },
+    {
+      "epoch": 1.2498163115356355,
+      "grad_norm": 1.9349777698516846,
+      "learning_rate": 4.665063509461097e-05,
+      "loss": 0.3464,
+      "step": 851
+    },
+    {
+      "epoch": 1.2512858192505512,
+      "grad_norm": 2.3614187240600586,
+      "learning_rate": 4.66377903832305e-05,
+      "loss": 0.4552,
+      "step": 852
+    },
+    {
+      "epoch": 1.2527553269654665,
+      "grad_norm": 2.105259895324707,
+      "learning_rate": 4.662492286476595e-05,
+      "loss": 0.3956,
+      "step": 853
+    },
+    {
+      "epoch": 1.2542248346803822,
+      "grad_norm": 2.745619058609009,
+      "learning_rate": 4.66120325527802e-05,
+      "loss": 0.5343,
+      "step": 854
+    },
+    {
+      "epoch": 1.2556943423952975,
+      "grad_norm": 1.8009859323501587,
+      "learning_rate": 4.659911946086014e-05,
+      "loss": 0.3343,
+      "step": 855
+    },
+    {
+      "epoch": 1.2571638501102131,
+      "grad_norm": 2.2524428367614746,
+      "learning_rate": 4.658618360261669e-05,
+      "loss": 0.3889,
+      "step": 856
+    },
+    {
+      "epoch": 1.2586333578251285,
+      "grad_norm": 2.2497031688690186,
+      "learning_rate": 4.6573224991684744e-05,
+      "loss": 0.4552,
+      "step": 857
+    },
+    {
+      "epoch": 1.2601028655400441,
+      "grad_norm": 2.406367063522339,
+      "learning_rate": 4.656024364172319e-05,
+      "loss": 0.3909,
+      "step": 858
+    },
+    {
+      "epoch": 1.2615723732549595,
+      "grad_norm": 2.156977891921997,
+      "learning_rate": 4.654723956641489e-05,
+      "loss": 0.4246,
+      "step": 859
+    },
+    {
+      "epoch": 1.2630418809698751,
+      "grad_norm": 1.9126993417739868,
+      "learning_rate": 4.653421277946664e-05,
+      "loss": 0.298,
+      "step": 860
+    },
+    {
+      "epoch": 1.2645113886847905,
+      "grad_norm": 1.9334651231765747,
+      "learning_rate": 4.6521163294609196e-05,
+      "loss": 0.3033,
+      "step": 861
+    },
+    {
+      "epoch": 1.2659808963997061,
+      "grad_norm": 2.1871142387390137,
+      "learning_rate": 4.6508091125597224e-05,
+      "loss": 0.3802,
+      "step": 862
+    },
+    {
+      "epoch": 1.2674504041146215,
+      "grad_norm": 2.4042274951934814,
+      "learning_rate": 4.649499628620931e-05,
+      "loss": 0.3972,
+      "step": 863
+    },
+    {
+      "epoch": 1.2689199118295371,
+      "grad_norm": 3.6092514991760254,
+      "learning_rate": 4.648187879024793e-05,
+      "loss": 0.4275,
+      "step": 864
+    },
+    {
+      "epoch": 1.2703894195444527,
+      "grad_norm": 1.8425012826919556,
+      "learning_rate": 4.6468738651539446e-05,
+      "loss": 0.355,
+      "step": 865
+    },
+    {
+      "epoch": 1.271858927259368,
+      "grad_norm": 2.6961333751678467,
+      "learning_rate": 4.645557588393407e-05,
+      "loss": 0.4864,
+      "step": 866
+    },
+    {
+      "epoch": 1.2733284349742835,
+      "grad_norm": 1.813185691833496,
+      "learning_rate": 4.6442390501305886e-05,
+      "loss": 0.3245,
+      "step": 867
+    },
+    {
+      "epoch": 1.274797942689199,
+      "grad_norm": 1.9211902618408203,
+      "learning_rate": 4.642918251755281e-05,
+      "loss": 0.3567,
+      "step": 868
+    },
+    {
+      "epoch": 1.2762674504041147,
+      "grad_norm": 2.289109706878662,
+      "learning_rate": 4.641595194659657e-05,
+      "loss": 0.3711,
+      "step": 869
+    },
+    {
+      "epoch": 1.27773695811903,
+      "grad_norm": 2.191418170928955,
+      "learning_rate": 4.640269880238271e-05,
+      "loss": 0.4196,
+      "step": 870
+    },
+    {
+      "epoch": 1.2792064658339457,
+      "grad_norm": 2.1594879627227783,
+      "learning_rate": 4.638942309888058e-05,
+      "loss": 0.3824,
+      "step": 871
+    },
+    {
+      "epoch": 1.280675973548861,
+      "grad_norm": 2.3400375843048096,
+      "learning_rate": 4.637612485008328e-05,
+      "loss": 0.2897,
+      "step": 872
+    },
+    {
+      "epoch": 1.2821454812637767,
+      "grad_norm": 2.3774490356445312,
+      "learning_rate": 4.6362804070007694e-05,
+      "loss": 0.3294,
+      "step": 873
+    },
+    {
+      "epoch": 1.283614988978692,
+      "grad_norm": 2.4574666023254395,
+      "learning_rate": 4.634946077269445e-05,
+      "loss": 0.3684,
+      "step": 874
+    },
+    {
+      "epoch": 1.2850844966936077,
+      "grad_norm": 2.0533554553985596,
+      "learning_rate": 4.63360949722079e-05,
+      "loss": 0.3748,
+      "step": 875
+    },
+    {
+      "epoch": 1.286554004408523,
+      "grad_norm": 1.8845067024230957,
+      "learning_rate": 4.6322706682636144e-05,
+      "loss": 0.3658,
+      "step": 876
+    },
+    {
+      "epoch": 1.2880235121234387,
+      "grad_norm": 2.023440361022949,
+      "learning_rate": 4.630929591809095e-05,
+      "loss": 0.5008,
+      "step": 877
+    },
+    {
+      "epoch": 1.289493019838354,
+      "grad_norm": 2.3095834255218506,
+      "learning_rate": 4.6295862692707794e-05,
+      "loss": 0.3932,
+      "step": 878
+    },
+    {
+      "epoch": 1.2909625275532697,
+      "grad_norm": 3.2307188510894775,
+      "learning_rate": 4.628240702064583e-05,
+      "loss": 0.4189,
+      "step": 879
+    },
+    {
+      "epoch": 1.292432035268185,
+      "grad_norm": 2.0710816383361816,
+      "learning_rate": 4.626892891608786e-05,
+      "loss": 0.3563,
+      "step": 880
+    },
+    {
+      "epoch": 1.2939015429831007,
+      "grad_norm": 1.947065830230713,
+      "learning_rate": 4.625542839324036e-05,
+      "loss": 0.2936,
+      "step": 881
+    },
+    {
+      "epoch": 1.2953710506980163,
+      "grad_norm": 2.159735918045044,
+      "learning_rate": 4.624190546633339e-05,
+      "loss": 0.359,
+      "step": 882
+    },
+    {
+      "epoch": 1.2968405584129317,
+      "grad_norm": 2.540349006652832,
+      "learning_rate": 4.622836014962064e-05,
+      "loss": 0.3715,
+      "step": 883
+    },
+    {
+      "epoch": 1.298310066127847,
+      "grad_norm": 1.9199589490890503,
+      "learning_rate": 4.621479245737944e-05,
+      "loss": 0.3873,
+      "step": 884
+    },
+    {
+      "epoch": 1.2997795738427627,
+      "grad_norm": 4.1916117668151855,
+      "learning_rate": 4.620120240391065e-05,
+      "loss": 0.4038,
+      "step": 885
+    },
+    {
+      "epoch": 1.3012490815576783,
+      "grad_norm": 2.578864812850952,
+      "learning_rate": 4.6187590003538725e-05,
+      "loss": 0.433,
+      "step": 886
+    },
+    {
+      "epoch": 1.3027185892725937,
+      "grad_norm": 2.4232017993927,
+      "learning_rate": 4.617395527061168e-05,
+      "loss": 0.4335,
+      "step": 887
+    },
+    {
+      "epoch": 1.3041880969875093,
+      "grad_norm": 1.9666262865066528,
+      "learning_rate": 4.6160298219501053e-05,
+      "loss": 0.4192,
+      "step": 888
+    },
+    {
+      "epoch": 1.3056576047024246,
+      "grad_norm": 1.940856695175171,
+      "learning_rate": 4.614661886460191e-05,
+      "loss": 0.2791,
+      "step": 889
+    },
+    {
+      "epoch": 1.3071271124173403,
+      "grad_norm": 2.348318576812744,
+      "learning_rate": 4.6132917220332846e-05,
+      "loss": 0.3846,
+      "step": 890
+    },
+    {
+      "epoch": 1.3085966201322556,
+      "grad_norm": 2.381927967071533,
+      "learning_rate": 4.611919330113592e-05,
+      "loss": 0.3669,
+      "step": 891
+    },
+    {
+      "epoch": 1.3100661278471712,
+      "grad_norm": 2.271601676940918,
+      "learning_rate": 4.6105447121476684e-05,
+      "loss": 0.3899,
+      "step": 892
+    },
+    {
+      "epoch": 1.3115356355620866,
+      "grad_norm": 2.3410778045654297,
+      "learning_rate": 4.6091678695844164e-05,
+      "loss": 0.4928,
+      "step": 893
+    },
+    {
+      "epoch": 1.3130051432770022,
+      "grad_norm": 1.928640604019165,
+      "learning_rate": 4.607788803875081e-05,
+      "loss": 0.3435,
+      "step": 894
+    },
+    {
+      "epoch": 1.3144746509919176,
+      "grad_norm": 1.8532216548919678,
+      "learning_rate": 4.6064075164732536e-05,
+      "loss": 0.31,
+      "step": 895
+    },
+    {
+      "epoch": 1.3159441587068332,
+      "grad_norm": 3.1321182250976562,
+      "learning_rate": 4.605024008834863e-05,
+      "loss": 0.5072,
+      "step": 896
+    },
+    {
+      "epoch": 1.3174136664217486,
+      "grad_norm": 2.3073890209198,
+      "learning_rate": 4.603638282418183e-05,
+      "loss": 0.3358,
+      "step": 897
+    },
+    {
+      "epoch": 1.3188831741366642,
+      "grad_norm": 2.5726428031921387,
+      "learning_rate": 4.6022503386838236e-05,
+      "loss": 0.5135,
+      "step": 898
+    },
+    {
+      "epoch": 1.3203526818515798,
+      "grad_norm": 3.0411410331726074,
+      "learning_rate": 4.600860179094732e-05,
+      "loss": 0.4825,
+      "step": 899
+    },
+    {
+      "epoch": 1.3218221895664952,
+      "grad_norm": 2.687187910079956,
+      "learning_rate": 4.599467805116191e-05,
+      "loss": 0.5792,
+      "step": 900
+    },
+    {
+      "epoch": 1.3232916972814106,
+      "grad_norm": 2.221806764602661,
+      "learning_rate": 4.598073218215817e-05,
+      "loss": 0.3317,
+      "step": 901
+    },
+    {
+      "epoch": 1.3247612049963262,
+      "grad_norm": 3.0394415855407715,
+      "learning_rate": 4.5966764198635606e-05,
+      "loss": 0.5182,
+      "step": 902
+    },
+    {
+      "epoch": 1.3262307127112418,
+      "grad_norm": 2.4513041973114014,
+      "learning_rate": 4.595277411531702e-05,
+      "loss": 0.5657,
+      "step": 903
+    },
+    {
+      "epoch": 1.3277002204261572,
+      "grad_norm": 1.8227571249008179,
+      "learning_rate": 4.593876194694851e-05,
+      "loss": 0.3485,
+      "step": 904
+    },
+    {
+      "epoch": 1.3291697281410728,
+      "grad_norm": 1.8964450359344482,
+      "learning_rate": 4.5924727708299445e-05,
+      "loss": 0.3705,
+      "step": 905
+    },
+    {
+      "epoch": 1.3306392358559882,
+      "grad_norm": 2.145548105239868,
+      "learning_rate": 4.591067141416249e-05,
+      "loss": 0.4215,
+      "step": 906
+    },
+    {
+      "epoch": 1.3321087435709038,
+      "grad_norm": 2.3101871013641357,
+      "learning_rate": 4.5896593079353514e-05,
+      "loss": 0.5327,
+      "step": 907
+    },
+    {
+      "epoch": 1.3335782512858192,
+      "grad_norm": 2.4980099201202393,
+      "learning_rate": 4.588249271871164e-05,
+      "loss": 0.5556,
+      "step": 908
+    },
+    {
+      "epoch": 1.3350477590007348,
+      "grad_norm": 1.993589997291565,
+      "learning_rate": 4.586837034709921e-05,
+      "loss": 0.3913,
+      "step": 909
+    },
+    {
+      "epoch": 1.3365172667156502,
+      "grad_norm": 2.0869038105010986,
+      "learning_rate": 4.585422597940175e-05,
+      "loss": 0.3261,
+      "step": 910
+    },
+    {
+      "epoch": 1.3379867744305658,
+      "grad_norm": 1.6393184661865234,
+      "learning_rate": 4.584005963052799e-05,
+      "loss": 0.3721,
+      "step": 911
+    },
+    {
+      "epoch": 1.3394562821454812,
+      "grad_norm": 1.7490593194961548,
+      "learning_rate": 4.582587131540982e-05,
+      "loss": 0.3587,
+      "step": 912
+    },
+    {
+      "epoch": 1.3409257898603968,
+      "grad_norm": 1.9467417001724243,
+      "learning_rate": 4.581166104900227e-05,
+      "loss": 0.4231,
+      "step": 913
+    },
+    {
+      "epoch": 1.3423952975753122,
+      "grad_norm": 2.518550395965576,
+      "learning_rate": 4.5797428846283546e-05,
+      "loss": 0.4732,
+      "step": 914
+    },
+    {
+      "epoch": 1.3438648052902278,
+      "grad_norm": 1.9926445484161377,
+      "learning_rate": 4.5783174722254926e-05,
+      "loss": 0.3655,
+      "step": 915
+    },
+    {
+      "epoch": 1.3453343130051434,
+      "grad_norm": 1.7111753225326538,
+      "learning_rate": 4.5768898691940845e-05,
+      "loss": 0.3305,
+      "step": 916
+    },
+    {
+      "epoch": 1.3468038207200588,
+      "grad_norm": 2.241863489151001,
+      "learning_rate": 4.575460077038877e-05,
+      "loss": 0.3248,
+      "step": 917
+    },
+    {
+      "epoch": 1.3482733284349742,
+      "grad_norm": 2.070793867111206,
+      "learning_rate": 4.574028097266929e-05,
+      "loss": 0.3726,
+      "step": 918
+    },
+    {
+      "epoch": 1.3497428361498898,
+      "grad_norm": 2.2661964893341064,
+      "learning_rate": 4.572593931387604e-05,
+      "loss": 0.4132,
+      "step": 919
+    },
+    {
+      "epoch": 1.3512123438648054,
+      "grad_norm": 2.121584415435791,
+      "learning_rate": 4.57115758091257e-05,
+      "loss": 0.4092,
+      "step": 920
+    },
+    {
+      "epoch": 1.3526818515797208,
+      "grad_norm": 2.526932954788208,
+      "learning_rate": 4.5697190473557946e-05,
+      "loss": 0.4619,
+      "step": 921
+    },
+    {
+      "epoch": 1.3541513592946364,
+      "grad_norm": 1.7800441980361938,
+      "learning_rate": 4.568278332233551e-05,
+      "loss": 0.3659,
+      "step": 922
+    },
+    {
+      "epoch": 1.3556208670095518,
+      "grad_norm": 2.759404420852661,
+      "learning_rate": 4.5668354370644086e-05,
+      "loss": 0.4792,
+      "step": 923
+    },
+    {
+      "epoch": 1.3570903747244674,
+      "grad_norm": 2.3154211044311523,
+      "learning_rate": 4.5653903633692366e-05,
+      "loss": 0.5306,
+      "step": 924
+    },
+    {
+      "epoch": 1.3585598824393827,
+      "grad_norm": 1.6706632375717163,
+      "learning_rate": 4.5639431126711996e-05,
+      "loss": 0.3179,
+      "step": 925
+    },
+    {
+      "epoch": 1.3600293901542984,
+      "grad_norm": 2.0957283973693848,
+      "learning_rate": 4.5624936864957556e-05,
+      "loss": 0.3635,
+      "step": 926
+    },
+    {
+      "epoch": 1.3614988978692137,
+      "grad_norm": 1.9344760179519653,
+      "learning_rate": 4.561042086370659e-05,
+      "loss": 0.4141,
+      "step": 927
+    },
+    {
+      "epoch": 1.3629684055841293,
+      "grad_norm": 2.0966579914093018,
+      "learning_rate": 4.559588313825952e-05,
+      "loss": 0.4708,
+      "step": 928
+    },
+    {
+      "epoch": 1.364437913299045,
+      "grad_norm": 1.9113317728042603,
+      "learning_rate": 4.558132370393968e-05,
+      "loss": 0.3108,
+      "step": 929
+    },
+    {
+      "epoch": 1.3659074210139603,
+      "grad_norm": 1.994572639465332,
+      "learning_rate": 4.5566742576093305e-05,
+      "loss": 0.412,
+      "step": 930
+    },
+    {
+      "epoch": 1.3673769287288757,
+      "grad_norm": 1.913077712059021,
+      "learning_rate": 4.555213977008946e-05,
+      "loss": 0.328,
+      "step": 931
+    },
+    {
+      "epoch": 1.3688464364437913,
+      "grad_norm": 2.3127646446228027,
+      "learning_rate": 4.553751530132009e-05,
+      "loss": 0.4803,
+      "step": 932
+    },
+    {
+      "epoch": 1.370315944158707,
+      "grad_norm": 1.5695472955703735,
+      "learning_rate": 4.552286918519996e-05,
+      "loss": 0.3776,
+      "step": 933
+    },
+    {
+      "epoch": 1.3717854518736223,
+      "grad_norm": 2.153719425201416,
+      "learning_rate": 4.550820143716664e-05,
+      "loss": 0.378,
+      "step": 934
+    },
+    {
+      "epoch": 1.3732549595885377,
+      "grad_norm": 1.870944857597351,
+      "learning_rate": 4.5493512072680536e-05,
+      "loss": 0.3433,
+      "step": 935
+    },
+    {
+      "epoch": 1.3747244673034533,
+      "grad_norm": 2.656707763671875,
+      "learning_rate": 4.54788011072248e-05,
+      "loss": 0.5418,
+      "step": 936
+    },
+    {
+      "epoch": 1.376193975018369,
+      "grad_norm": 1.7983543872833252,
+      "learning_rate": 4.546406855630537e-05,
+      "loss": 0.3586,
+      "step": 937
+    },
+    {
+      "epoch": 1.3776634827332843,
+      "grad_norm": 1.9111722707748413,
+      "learning_rate": 4.5449314435450955e-05,
+      "loss": 0.2981,
+      "step": 938
+    },
+    {
+      "epoch": 1.3791329904482,
+      "grad_norm": 1.833272933959961,
+      "learning_rate": 4.543453876021296e-05,
+      "loss": 0.3126,
+      "step": 939
+    },
+    {
+      "epoch": 1.3806024981631153,
+      "grad_norm": 1.9480398893356323,
+      "learning_rate": 4.541974154616555e-05,
+      "loss": 0.3941,
+      "step": 940
+    },
+    {
+      "epoch": 1.382072005878031,
+      "grad_norm": 1.8195396661758423,
+      "learning_rate": 4.540492280890555e-05,
+      "loss": 0.2753,
+      "step": 941
+    },
+    {
+      "epoch": 1.3835415135929463,
+      "grad_norm": 1.871058702468872,
+      "learning_rate": 4.5390082564052496e-05,
+      "loss": 0.3781,
+      "step": 942
+    },
+    {
+      "epoch": 1.385011021307862,
+      "grad_norm": 2.0355100631713867,
+      "learning_rate": 4.5375220827248596e-05,
+      "loss": 0.3576,
+      "step": 943
+    },
+    {
+      "epoch": 1.3864805290227773,
+      "grad_norm": 1.6167826652526855,
+      "learning_rate": 4.536033761415871e-05,
+      "loss": 0.3504,
+      "step": 944
+    },
+    {
+      "epoch": 1.387950036737693,
+      "grad_norm": 1.9695193767547607,
+      "learning_rate": 4.534543294047033e-05,
+      "loss": 0.3979,
+      "step": 945
+    },
+    {
+      "epoch": 1.3894195444526085,
+      "grad_norm": 2.6076338291168213,
+      "learning_rate": 4.533050682189357e-05,
+      "loss": 0.3761,
+      "step": 946
+    },
+    {
+      "epoch": 1.390889052167524,
+      "grad_norm": 1.9334447383880615,
+      "learning_rate": 4.5315559274161144e-05,
+      "loss": 0.3591,
+      "step": 947
+    },
+    {
+      "epoch": 1.3923585598824393,
+      "grad_norm": 2.5892813205718994,
+      "learning_rate": 4.5300590313028355e-05,
+      "loss": 0.4656,
+      "step": 948
+    },
+    {
+      "epoch": 1.393828067597355,
+      "grad_norm": 2.281627655029297,
+      "learning_rate": 4.528559995427308e-05,
+      "loss": 0.4175,
+      "step": 949
+    },
+    {
+      "epoch": 1.3952975753122705,
+      "grad_norm": 2.482177257537842,
+      "learning_rate": 4.5270588213695754e-05,
+      "loss": 0.5575,
+      "step": 950
+    },
+    {
+      "epoch": 1.3967670830271859,
+      "grad_norm": 2.0755059719085693,
+      "learning_rate": 4.525555510711934e-05,
+      "loss": 0.348,
+      "step": 951
+    },
+    {
+      "epoch": 1.3982365907421013,
+      "grad_norm": 2.6475231647491455,
+      "learning_rate": 4.524050065038932e-05,
+      "loss": 0.4768,
+      "step": 952
+    },
+    {
+      "epoch": 1.3997060984570169,
+      "grad_norm": 2.458592176437378,
+      "learning_rate": 4.522542485937369e-05,
+      "loss": 0.6306,
+      "step": 953
+    },
+    {
+      "epoch": 1.4011756061719325,
+      "grad_norm": 1.8067344427108765,
+      "learning_rate": 4.521032774996292e-05,
+      "loss": 0.3011,
+      "step": 954
+    },
+    {
+      "epoch": 1.4026451138868479,
+      "grad_norm": 1.9009654521942139,
+      "learning_rate": 4.5195209338069974e-05,
+      "loss": 0.4376,
+      "step": 955
+    },
+    {
+      "epoch": 1.4041146216017635,
+      "grad_norm": 2.49210262298584,
+      "learning_rate": 4.518006963963024e-05,
+      "loss": 0.3288,
+      "step": 956
+    },
+    {
+      "epoch": 1.4055841293166789,
+      "grad_norm": 2.1522045135498047,
+      "learning_rate": 4.516490867060156e-05,
+      "loss": 0.3231,
+      "step": 957
+    },
+    {
+      "epoch": 1.4070536370315945,
+      "grad_norm": 2.199289560317993,
+      "learning_rate": 4.51497264469642e-05,
+      "loss": 0.4247,
+      "step": 958
+    },
+    {
+      "epoch": 1.4085231447465099,
+      "grad_norm": 1.9931894540786743,
+      "learning_rate": 4.5134522984720816e-05,
+      "loss": 0.3478,
+      "step": 959
+    },
+    {
+      "epoch": 1.4099926524614255,
+      "grad_norm": 1.938806414604187,
+      "learning_rate": 4.511929829989646e-05,
+      "loss": 0.4182,
+      "step": 960
+    },
+    {
+      "epoch": 1.4114621601763409,
+      "grad_norm": 2.57068133354187,
+      "learning_rate": 4.510405240853854e-05,
+      "loss": 0.4351,
+      "step": 961
+    },
+    {
+      "epoch": 1.4129316678912565,
+      "grad_norm": 1.675110936164856,
+      "learning_rate": 4.5088785326716844e-05,
+      "loss": 0.3641,
+      "step": 962
+    },
+    {
+      "epoch": 1.414401175606172,
+      "grad_norm": 1.7026541233062744,
+      "learning_rate": 4.507349707052346e-05,
+      "loss": 0.2882,
+      "step": 963
+    },
+    {
+      "epoch": 1.4158706833210875,
+      "grad_norm": 1.7799955606460571,
+      "learning_rate": 4.505818765607283e-05,
+      "loss": 0.3217,
+      "step": 964
+    },
+    {
+      "epoch": 1.4173401910360028,
+      "grad_norm": 1.6473877429962158,
+      "learning_rate": 4.504285709950167e-05,
+      "loss": 0.3324,
+      "step": 965
+    },
+    {
+      "epoch": 1.4188096987509184,
+      "grad_norm": 1.6083662509918213,
+      "learning_rate": 4.502750541696899e-05,
+      "loss": 0.3082,
+      "step": 966
+    },
+    {
+      "epoch": 1.420279206465834,
+      "grad_norm": 1.809044361114502,
+      "learning_rate": 4.5012132624656065e-05,
+      "loss": 0.3753,
+      "step": 967
+    },
+    {
+      "epoch": 1.4217487141807494,
+      "grad_norm": 2.5478386878967285,
+      "learning_rate": 4.499673873876644e-05,
+      "loss": 0.4225,
+      "step": 968
+    },
+    {
+      "epoch": 1.4232182218956648,
+      "grad_norm": 1.5637617111206055,
+      "learning_rate": 4.498132377552587e-05,
+      "loss": 0.3196,
+      "step": 969
+    },
+    {
+      "epoch": 1.4246877296105804,
+      "grad_norm": 2.050851821899414,
+      "learning_rate": 4.496588775118232e-05,
+      "loss": 0.3449,
+      "step": 970
+    },
+    {
+      "epoch": 1.426157237325496,
+      "grad_norm": 1.80287504196167,
+      "learning_rate": 4.4950430682006e-05,
+      "loss": 0.3201,
+      "step": 971
+    },
+    {
+      "epoch": 1.4276267450404114,
+      "grad_norm": 2.2093443870544434,
+      "learning_rate": 4.493495258428923e-05,
+      "loss": 0.5092,
+      "step": 972
+    },
+    {
+      "epoch": 1.429096252755327,
+      "grad_norm": 2.3426804542541504,
+      "learning_rate": 4.491945347434656e-05,
+      "loss": 0.5347,
+      "step": 973
+    },
+    {
+      "epoch": 1.4305657604702424,
+      "grad_norm": 1.721139907836914,
+      "learning_rate": 4.490393336851466e-05,
+      "loss": 0.3234,
+      "step": 974
+    },
+    {
+      "epoch": 1.432035268185158,
+      "grad_norm": 2.013716459274292,
+      "learning_rate": 4.4888392283152333e-05,
+      "loss": 0.3683,
+      "step": 975
+    },
+    {
+      "epoch": 1.4335047759000734,
+      "grad_norm": 1.8883419036865234,
+      "learning_rate": 4.48728302346405e-05,
+      "loss": 0.387,
+      "step": 976
+    },
+    {
+      "epoch": 1.434974283614989,
+      "grad_norm": 2.614715099334717,
+      "learning_rate": 4.485724723938215e-05,
+      "loss": 0.505,
+      "step": 977
+    },
+    {
+      "epoch": 1.4364437913299044,
+      "grad_norm": 2.5659191608428955,
+      "learning_rate": 4.484164331380241e-05,
+      "loss": 0.4898,
+      "step": 978
+    },
+    {
+      "epoch": 1.43791329904482,
+      "grad_norm": 1.8044931888580322,
+      "learning_rate": 4.482601847434841e-05,
+      "loss": 0.3407,
+      "step": 979
+    },
+    {
+      "epoch": 1.4393828067597356,
+      "grad_norm": 1.6509901285171509,
+      "learning_rate": 4.4810372737489345e-05,
+      "loss": 0.3064,
+      "step": 980
+    },
+    {
+      "epoch": 1.440852314474651,
+      "grad_norm": 2.4285407066345215,
+      "learning_rate": 4.4794706119716455e-05,
+      "loss": 0.5257,
+      "step": 981
+    },
+    {
+      "epoch": 1.4423218221895664,
+      "grad_norm": 1.9130295515060425,
+      "learning_rate": 4.477901863754296e-05,
+      "loss": 0.3485,
+      "step": 982
+    },
+    {
+      "epoch": 1.443791329904482,
+      "grad_norm": 2.3371636867523193,
+      "learning_rate": 4.476331030750408e-05,
+      "loss": 0.4456,
+      "step": 983
+    },
+    {
+      "epoch": 1.4452608376193976,
+      "grad_norm": 1.8473620414733887,
+      "learning_rate": 4.474758114615704e-05,
+      "loss": 0.3679,
+      "step": 984
+    },
+    {
+      "epoch": 1.446730345334313,
+      "grad_norm": 1.9587016105651855,
+      "learning_rate": 4.4731831170080966e-05,
+      "loss": 0.334,
+      "step": 985
+    },
+    {
+      "epoch": 1.4481998530492284,
+      "grad_norm": 1.7284753322601318,
+      "learning_rate": 4.4716060395876956e-05,
+      "loss": 0.3362,
+      "step": 986
+    },
+    {
+      "epoch": 1.449669360764144,
+      "grad_norm": 2.397571563720703,
+      "learning_rate": 4.4700268840168045e-05,
+      "loss": 0.373,
+      "step": 987
+    },
+    {
+      "epoch": 1.4511388684790596,
+      "grad_norm": 1.8131027221679688,
+      "learning_rate": 4.468445651959916e-05,
+      "loss": 0.3501,
+      "step": 988
+    },
+    {
+      "epoch": 1.452608376193975,
+      "grad_norm": 1.8367456197738647,
+      "learning_rate": 4.4668623450837085e-05,
+      "loss": 0.381,
+      "step": 989
+    },
+    {
+      "epoch": 1.4540778839088906,
+      "grad_norm": 2.5923097133636475,
+      "learning_rate": 4.4652769650570515e-05,
+      "loss": 0.6026,
+      "step": 990
+    },
+    {
+      "epoch": 1.455547391623806,
+      "grad_norm": 2.8078975677490234,
+      "learning_rate": 4.463689513550997e-05,
+      "loss": 0.3752,
+      "step": 991
+    },
+    {
+      "epoch": 1.4570168993387216,
+      "grad_norm": 2.237123727798462,
+      "learning_rate": 4.4620999922387826e-05,
+      "loss": 0.3774,
+      "step": 992
+    },
+    {
+      "epoch": 1.458486407053637,
+      "grad_norm": 2.1396095752716064,
+      "learning_rate": 4.460508402795827e-05,
+      "loss": 0.3894,
+      "step": 993
+    },
+    {
+      "epoch": 1.4599559147685526,
+      "grad_norm": 2.783582925796509,
+      "learning_rate": 4.458914746899726e-05,
+      "loss": 0.3599,
+      "step": 994
+    },
+    {
+      "epoch": 1.461425422483468,
+      "grad_norm": 1.8704665899276733,
+      "learning_rate": 4.457319026230257e-05,
+      "loss": 0.3209,
+      "step": 995
+    },
+    {
+      "epoch": 1.4628949301983836,
+      "grad_norm": 2.3405473232269287,
+      "learning_rate": 4.455721242469373e-05,
+      "loss": 0.4338,
+      "step": 996
+    },
+    {
+      "epoch": 1.4643644379132992,
+      "grad_norm": 1.7970101833343506,
+      "learning_rate": 4.4541213973012004e-05,
+      "loss": 0.3527,
+      "step": 997
+    },
+    {
+      "epoch": 1.4658339456282146,
+      "grad_norm": 2.3542556762695312,
+      "learning_rate": 4.452519492412039e-05,
+      "loss": 0.4169,
+      "step": 998
+    },
+    {
+      "epoch": 1.46730345334313,
+      "grad_norm": 1.6511187553405762,
+      "learning_rate": 4.4509155294903595e-05,
+      "loss": 0.3325,
+      "step": 999
+    },
+    {
+      "epoch": 1.4687729610580456,
+      "grad_norm": 2.0340991020202637,
+      "learning_rate": 4.449309510226803e-05,
+      "loss": 0.4163,
+      "step": 1000
+    },
+    {
+      "epoch": 1.4702424687729612,
+      "grad_norm": 1.5841100215911865,
+      "learning_rate": 4.447701436314176e-05,
+      "loss": 0.29,
+      "step": 1001
+    },
+    {
+      "epoch": 1.4717119764878765,
+      "grad_norm": 2.4417927265167236,
+      "learning_rate": 4.446091309447453e-05,
+      "loss": 0.4366,
+      "step": 1002
+    },
+    {
+      "epoch": 1.473181484202792,
+      "grad_norm": 1.768143892288208,
+      "learning_rate": 4.44447913132377e-05,
+      "loss": 0.324,
+      "step": 1003
+    },
+    {
+      "epoch": 1.4746509919177075,
+      "grad_norm": 2.3296210765838623,
+      "learning_rate": 4.442864903642428e-05,
+      "loss": 0.4574,
+      "step": 1004
+    },
+    {
+      "epoch": 1.4761204996326232,
+      "grad_norm": 1.6528271436691284,
+      "learning_rate": 4.4412486281048836e-05,
+      "loss": 0.3129,
+      "step": 1005
+    },
+    {
+      "epoch": 1.4775900073475385,
+      "grad_norm": 1.8390206098556519,
+      "learning_rate": 4.439630306414758e-05,
+      "loss": 0.2933,
+      "step": 1006
+    },
+    {
+      "epoch": 1.4790595150624541,
+      "grad_norm": 2.4844603538513184,
+      "learning_rate": 4.4380099402778244e-05,
+      "loss": 0.4202,
+      "step": 1007
+    },
+    {
+      "epoch": 1.4805290227773695,
+      "grad_norm": 3.0346226692199707,
+      "learning_rate": 4.4363875314020134e-05,
+      "loss": 0.5391,
+      "step": 1008
+    },
+    {
+      "epoch": 1.4819985304922851,
+      "grad_norm": 2.3181450366973877,
+      "learning_rate": 4.4347630814974076e-05,
+      "loss": 0.4339,
+      "step": 1009
+    },
+    {
+      "epoch": 1.4834680382072005,
+      "grad_norm": 2.108584403991699,
+      "learning_rate": 4.4331365922762414e-05,
+      "loss": 0.3883,
+      "step": 1010
+    },
+    {
+      "epoch": 1.4849375459221161,
+      "grad_norm": 2.2387866973876953,
+      "learning_rate": 4.431508065452897e-05,
+      "loss": 0.4758,
+      "step": 1011
+    },
+    {
+      "epoch": 1.4864070536370315,
+      "grad_norm": 2.6580286026000977,
+      "learning_rate": 4.4298775027439085e-05,
+      "loss": 0.4769,
+      "step": 1012
+    },
+    {
+      "epoch": 1.4878765613519471,
+      "grad_norm": 2.0424692630767822,
+      "learning_rate": 4.428244905867952e-05,
+      "loss": 0.3528,
+      "step": 1013
+    },
+    {
+      "epoch": 1.4893460690668627,
+      "grad_norm": 2.4300434589385986,
+      "learning_rate": 4.426610276545848e-05,
+      "loss": 0.4731,
+      "step": 1014
+    },
+    {
+      "epoch": 1.4908155767817781,
+      "grad_norm": 2.8303451538085938,
+      "learning_rate": 4.4249736165005626e-05,
+      "loss": 0.577,
+      "step": 1015
+    },
+    {
+      "epoch": 1.4922850844966935,
+      "grad_norm": 2.0848405361175537,
+      "learning_rate": 4.423334927457198e-05,
+      "loss": 0.4429,
+      "step": 1016
+    },
+    {
+      "epoch": 1.4937545922116091,
+      "grad_norm": 1.8962504863739014,
+      "learning_rate": 4.421694211142997e-05,
+      "loss": 0.3414,
+      "step": 1017
+    },
+    {
+      "epoch": 1.4952240999265247,
+      "grad_norm": 1.480811595916748,
+      "learning_rate": 4.420051469287341e-05,
+      "loss": 0.3152,
+      "step": 1018
+    },
+    {
+      "epoch": 1.49669360764144,
+      "grad_norm": 1.9934083223342896,
+      "learning_rate": 4.418406703621743e-05,
+      "loss": 0.4217,
+      "step": 1019
+    },
+    {
+      "epoch": 1.4981631153563555,
+      "grad_norm": 2.196218490600586,
+      "learning_rate": 4.416759915879851e-05,
+      "loss": 0.4017,
+      "step": 1020
+    },
+    {
+      "epoch": 1.499632623071271,
+      "grad_norm": 2.1255695819854736,
+      "learning_rate": 4.415111107797445e-05,
+      "loss": 0.459,
+      "step": 1021
+    },
+    {
+      "epoch": 1.5011021307861867,
+      "grad_norm": 2.1786110401153564,
+      "learning_rate": 4.4134602811124335e-05,
+      "loss": 0.3518,
+      "step": 1022
+    },
+    {
+      "epoch": 1.502571638501102,
+      "grad_norm": 1.638182520866394,
+      "learning_rate": 4.411807437564853e-05,
+      "loss": 0.3374,
+      "step": 1023
+    },
+    {
+      "epoch": 1.5040411462160175,
+      "grad_norm": 2.3170716762542725,
+      "learning_rate": 4.4101525788968655e-05,
+      "loss": 0.4431,
+      "step": 1024
+    },
+    {
+      "epoch": 1.505510653930933,
+      "grad_norm": 3.0402963161468506,
+      "learning_rate": 4.408495706852758e-05,
+      "loss": 0.5101,
+      "step": 1025
+    },
+    {
+      "epoch": 1.5069801616458487,
+      "grad_norm": 2.4391536712646484,
+      "learning_rate": 4.406836823178937e-05,
+      "loss": 0.4693,
+      "step": 1026
+    },
+    {
+      "epoch": 1.5084496693607643,
+      "grad_norm": 2.1560895442962646,
+      "learning_rate": 4.4051759296239336e-05,
+      "loss": 0.3989,
+      "step": 1027
+    },
+    {
+      "epoch": 1.5099191770756797,
+      "grad_norm": 1.6178730726242065,
+      "learning_rate": 4.4035130279383934e-05,
+      "loss": 0.2816,
+      "step": 1028
+    },
+    {
+      "epoch": 1.511388684790595,
+      "grad_norm": 1.7810813188552856,
+      "learning_rate": 4.401848119875081e-05,
+      "loss": 0.3193,
+      "step": 1029
+    },
+    {
+      "epoch": 1.5128581925055107,
+      "grad_norm": 1.773915410041809,
+      "learning_rate": 4.400181207188876e-05,
+      "loss": 0.3484,
+      "step": 1030
+    },
+    {
+      "epoch": 1.5143277002204263,
+      "grad_norm": 1.9315768480300903,
+      "learning_rate": 4.398512291636768e-05,
+      "loss": 0.3103,
+      "step": 1031
+    },
+    {
+      "epoch": 1.5157972079353417,
+      "grad_norm": 2.5469584465026855,
+      "learning_rate": 4.3968413749778616e-05,
+      "loss": 0.4905,
+      "step": 1032
+    },
+    {
+      "epoch": 1.517266715650257,
+      "grad_norm": 2.421135187149048,
+      "learning_rate": 4.3951684589733683e-05,
+      "loss": 0.3615,
+      "step": 1033
+    },
+    {
+      "epoch": 1.5187362233651727,
+      "grad_norm": 2.0811352729797363,
+      "learning_rate": 4.393493545386607e-05,
+      "loss": 0.4307,
+      "step": 1034
+    },
+    {
+      "epoch": 1.5202057310800883,
+      "grad_norm": 1.7437090873718262,
+      "learning_rate": 4.3918166359830034e-05,
+      "loss": 0.3554,
+      "step": 1035
+    },
+    {
+      "epoch": 1.5216752387950037,
+      "grad_norm": 3.370530843734741,
+      "learning_rate": 4.3901377325300865e-05,
+      "loss": 0.5472,
+      "step": 1036
+    },
+    {
+      "epoch": 1.523144746509919,
+      "grad_norm": 2.1887168884277344,
+      "learning_rate": 4.3884568367974845e-05,
+      "loss": 0.4753,
+      "step": 1037
+    },
+    {
+      "epoch": 1.5246142542248347,
+      "grad_norm": 2.938441753387451,
+      "learning_rate": 4.386773950556931e-05,
+      "loss": 0.3596,
+      "step": 1038
+    },
+    {
+      "epoch": 1.5260837619397503,
+      "grad_norm": 1.9315327405929565,
+      "learning_rate": 4.385089075582252e-05,
+      "loss": 0.4129,
+      "step": 1039
+    },
+    {
+      "epoch": 1.5275532696546656,
+      "grad_norm": 1.9231863021850586,
+      "learning_rate": 4.3834022136493716e-05,
+      "loss": 0.4717,
+      "step": 1040
+    },
+    {
+      "epoch": 1.529022777369581,
+      "grad_norm": 1.7257550954818726,
+      "learning_rate": 4.381713366536311e-05,
+      "loss": 0.3103,
+      "step": 1041
+    },
+    {
+      "epoch": 1.5304922850844966,
+      "grad_norm": 2.638929843902588,
+      "learning_rate": 4.380022536023182e-05,
+      "loss": 0.4881,
+      "step": 1042
+    },
+    {
+      "epoch": 1.5319617927994122,
+      "grad_norm": 2.0922393798828125,
+      "learning_rate": 4.378329723892184e-05,
+      "loss": 0.4635,
+      "step": 1043
+    },
+    {
+      "epoch": 1.5334313005143279,
+      "grad_norm": 1.6405478715896606,
+      "learning_rate": 4.376634931927609e-05,
+      "loss": 0.2653,
+      "step": 1044
+    },
+    {
+      "epoch": 1.5349008082292432,
+      "grad_norm": 1.5531071424484253,
+      "learning_rate": 4.374938161915835e-05,
+      "loss": 0.2806,
+      "step": 1045
+    },
+    {
+      "epoch": 1.5363703159441586,
+      "grad_norm": 1.8074288368225098,
+      "learning_rate": 4.373239415645324e-05,
+      "loss": 0.352,
+      "step": 1046
+    },
+    {
+      "epoch": 1.5378398236590742,
+      "grad_norm": 2.4195077419281006,
+      "learning_rate": 4.3715386949066204e-05,
+      "loss": 0.3829,
+      "step": 1047
+    },
+    {
+      "epoch": 1.5393093313739898,
+      "grad_norm": 2.142836570739746,
+      "learning_rate": 4.3698360014923533e-05,
+      "loss": 0.4052,
+      "step": 1048
+    },
+    {
+      "epoch": 1.5407788390889052,
+      "grad_norm": 2.1885037422180176,
+      "learning_rate": 4.3681313371972276e-05,
+      "loss": 0.4877,
+      "step": 1049
+    },
+    {
+      "epoch": 1.5422483468038206,
+      "grad_norm": 2.3549177646636963,
+      "learning_rate": 4.3664247038180264e-05,
+      "loss": 0.3998,
+      "step": 1050
+    },
+    {
+      "epoch": 1.5437178545187362,
+      "grad_norm": 2.37020206451416,
+      "learning_rate": 4.3647161031536085e-05,
+      "loss": 0.3283,
+      "step": 1051
+    },
+    {
+      "epoch": 1.5451873622336518,
+      "grad_norm": 2.675374984741211,
+      "learning_rate": 4.363005537004907e-05,
+      "loss": 0.5809,
+      "step": 1052
+    },
+    {
+      "epoch": 1.5466568699485672,
+      "grad_norm": 1.9372869729995728,
+      "learning_rate": 4.361293007174925e-05,
+      "loss": 0.3467,
+      "step": 1053
+    },
+    {
+      "epoch": 1.5481263776634826,
+      "grad_norm": 2.6747262477874756,
+      "learning_rate": 4.359578515468737e-05,
+      "loss": 0.4417,
+      "step": 1054
+    },
+    {
+      "epoch": 1.5495958853783982,
+      "grad_norm": 2.2514960765838623,
+      "learning_rate": 4.357862063693486e-05,
+      "loss": 0.4392,
+      "step": 1055
+    },
+    {
+      "epoch": 1.5510653930933138,
+      "grad_norm": 2.179076671600342,
+      "learning_rate": 4.3561436536583776e-05,
+      "loss": 0.3952,
+      "step": 1056
+    },
+    {
+      "epoch": 1.5525349008082292,
+      "grad_norm": 1.4125500917434692,
+      "learning_rate": 4.354423287174686e-05,
+      "loss": 0.2298,
+      "step": 1057
+    },
+    {
+      "epoch": 1.5540044085231446,
+      "grad_norm": 1.6750937700271606,
+      "learning_rate": 4.3527009660557434e-05,
+      "loss": 0.322,
+      "step": 1058
+    },
+    {
+      "epoch": 1.5554739162380602,
+      "grad_norm": 2.4319887161254883,
+      "learning_rate": 4.350976692116945e-05,
+      "loss": 0.5423,
+      "step": 1059
+    },
+    {
+      "epoch": 1.5569434239529758,
+      "grad_norm": 1.950619101524353,
+      "learning_rate": 4.349250467175744e-05,
+      "loss": 0.3405,
+      "step": 1060
+    },
+    {
+      "epoch": 1.5584129316678914,
+      "grad_norm": 2.1191647052764893,
+      "learning_rate": 4.347522293051648e-05,
+      "loss": 0.4261,
+      "step": 1061
+    },
+    {
+      "epoch": 1.5598824393828068,
+      "grad_norm": 1.7134963274002075,
+      "learning_rate": 4.345792171566222e-05,
+      "loss": 0.3649,
+      "step": 1062
+    },
+    {
+      "epoch": 1.5613519470977222,
+      "grad_norm": 2.498394727706909,
+      "learning_rate": 4.3440601045430824e-05,
+      "loss": 0.501,
+      "step": 1063
+    },
+    {
+      "epoch": 1.5628214548126378,
+      "grad_norm": 2.60149884223938,
+      "learning_rate": 4.342326093807895e-05,
+      "loss": 0.5734,
+      "step": 1064
+    },
+    {
+      "epoch": 1.5642909625275534,
+      "grad_norm": 1.6078531742095947,
+      "learning_rate": 4.3405901411883764e-05,
+      "loss": 0.2928,
+      "step": 1065
+    },
+    {
+      "epoch": 1.5657604702424688,
+      "grad_norm": 4.084954261779785,
+      "learning_rate": 4.338852248514289e-05,
+      "loss": 0.5999,
+      "step": 1066
+    },
+    {
+      "epoch": 1.5672299779573842,
+      "grad_norm": 1.7487298250198364,
+      "learning_rate": 4.337112417617439e-05,
+      "loss": 0.3384,
+      "step": 1067
+    },
+    {
+      "epoch": 1.5686994856722998,
+      "grad_norm": 2.3559834957122803,
+      "learning_rate": 4.335370650331679e-05,
+      "loss": 0.4421,
+      "step": 1068
+    },
+    {
+      "epoch": 1.5701689933872154,
+      "grad_norm": 1.8020366430282593,
+      "learning_rate": 4.333626948492898e-05,
+      "loss": 0.3816,
+      "step": 1069
+    },
+    {
+      "epoch": 1.5716385011021308,
+      "grad_norm": 2.049919843673706,
+      "learning_rate": 4.331881313939029e-05,
+      "loss": 0.3824,
+      "step": 1070
+    },
+    {
+      "epoch": 1.5731080088170462,
+      "grad_norm": 1.8263096809387207,
+      "learning_rate": 4.330133748510037e-05,
+      "loss": 0.321,
+      "step": 1071
+    },
+    {
+      "epoch": 1.5745775165319618,
+      "grad_norm": 2.2093987464904785,
+      "learning_rate": 4.3283842540479264e-05,
+      "loss": 0.4709,
+      "step": 1072
+    },
+    {
+      "epoch": 1.5760470242468774,
+      "grad_norm": 2.030076503753662,
+      "learning_rate": 4.3266328323967333e-05,
+      "loss": 0.4114,
+      "step": 1073
+    },
+    {
+      "epoch": 1.5775165319617928,
+      "grad_norm": 1.963865041732788,
+      "learning_rate": 4.324879485402524e-05,
+      "loss": 0.3445,
+      "step": 1074
+    },
+    {
+      "epoch": 1.5789860396767081,
+      "grad_norm": 1.9706882238388062,
+      "learning_rate": 4.323124214913397e-05,
+      "loss": 0.3189,
+      "step": 1075
+    },
+    {
+      "epoch": 1.5804555473916237,
+      "grad_norm": 1.8552851676940918,
+      "learning_rate": 4.321367022779476e-05,
+      "loss": 0.3493,
+      "step": 1076
+    },
+    {
+      "epoch": 1.5819250551065394,
+      "grad_norm": 2.499727249145508,
+      "learning_rate": 4.3196079108529106e-05,
+      "loss": 0.472,
+      "step": 1077
+    },
+    {
+      "epoch": 1.583394562821455,
+      "grad_norm": 1.721739411354065,
+      "learning_rate": 4.3178468809878734e-05,
+      "loss": 0.3161,
+      "step": 1078
+    },
+    {
+      "epoch": 1.5848640705363704,
+      "grad_norm": 2.2469680309295654,
+      "learning_rate": 4.3160839350405606e-05,
+      "loss": 0.3631,
+      "step": 1079
+    },
+    {
+      "epoch": 1.5863335782512857,
+      "grad_norm": 2.3527121543884277,
+      "learning_rate": 4.3143190748691856e-05,
+      "loss": 0.4135,
+      "step": 1080
+    },
+    {
+      "epoch": 1.5878030859662013,
+      "grad_norm": 1.9278106689453125,
+      "learning_rate": 4.312552302333982e-05,
+      "loss": 0.2692,
+      "step": 1081
+    },
+    {
+      "epoch": 1.589272593681117,
+      "grad_norm": 2.0070223808288574,
+      "learning_rate": 4.310783619297196e-05,
+      "loss": 0.4337,
+      "step": 1082
+    },
+    {
+      "epoch": 1.5907421013960323,
+      "grad_norm": 1.7902767658233643,
+      "learning_rate": 4.309013027623091e-05,
+      "loss": 0.3657,
+      "step": 1083
+    },
+    {
+      "epoch": 1.5922116091109477,
+      "grad_norm": 2.2472829818725586,
+      "learning_rate": 4.307240529177939e-05,
+      "loss": 0.4494,
+      "step": 1084
+    },
+    {
+      "epoch": 1.5936811168258633,
+      "grad_norm": 2.5411903858184814,
+      "learning_rate": 4.305466125830023e-05,
+      "loss": 0.3289,
+      "step": 1085
+    },
+    {
+      "epoch": 1.595150624540779,
+      "grad_norm": 2.3176913261413574,
+      "learning_rate": 4.303689819449636e-05,
+      "loss": 0.3911,
+      "step": 1086
+    },
+    {
+      "epoch": 1.5966201322556943,
+      "grad_norm": 2.154841899871826,
+      "learning_rate": 4.301911611909074e-05,
+      "loss": 0.4358,
+      "step": 1087
+    },
+    {
+      "epoch": 1.5980896399706097,
+      "grad_norm": 1.8012100458145142,
+      "learning_rate": 4.300131505082637e-05,
+      "loss": 0.3721,
+      "step": 1088
+    },
+    {
+      "epoch": 1.5995591476855253,
+      "grad_norm": 1.4884535074234009,
+      "learning_rate": 4.2983495008466276e-05,
+      "loss": 0.267,
+      "step": 1089
+    },
+    {
+      "epoch": 1.601028655400441,
+      "grad_norm": 2.1323859691619873,
+      "learning_rate": 4.296565601079351e-05,
+      "loss": 0.4128,
+      "step": 1090
+    },
+    {
+      "epoch": 1.6024981631153563,
+      "grad_norm": 2.1527538299560547,
+      "learning_rate": 4.294779807661105e-05,
+      "loss": 0.422,
+      "step": 1091
+    },
+    {
+      "epoch": 1.6039676708302717,
+      "grad_norm": 1.842810034751892,
+      "learning_rate": 4.292992122474188e-05,
+      "loss": 0.2854,
+      "step": 1092
+    },
+    {
+      "epoch": 1.6054371785451873,
+      "grad_norm": 1.740471363067627,
+      "learning_rate": 4.29120254740289e-05,
+      "loss": 0.3322,
+      "step": 1093
+    },
+    {
+      "epoch": 1.606906686260103,
+      "grad_norm": 1.844868779182434,
+      "learning_rate": 4.289411084333494e-05,
+      "loss": 0.3112,
+      "step": 1094
+    },
+    {
+      "epoch": 1.6083761939750185,
+      "grad_norm": 2.241093397140503,
+      "learning_rate": 4.287617735154272e-05,
+      "loss": 0.3685,
+      "step": 1095
+    },
+    {
+      "epoch": 1.609845701689934,
+      "grad_norm": 1.8865793943405151,
+      "learning_rate": 4.285822501755486e-05,
+      "loss": 0.3833,
+      "step": 1096
+    },
+    {
+      "epoch": 1.6113152094048493,
+      "grad_norm": 2.118736982345581,
+      "learning_rate": 4.284025386029381e-05,
+      "loss": 0.3788,
+      "step": 1097
+    },
+    {
+      "epoch": 1.612784717119765,
+      "grad_norm": 1.8158824443817139,
+      "learning_rate": 4.282226389870189e-05,
+      "loss": 0.3307,
+      "step": 1098
+    },
+    {
+      "epoch": 1.6142542248346805,
+      "grad_norm": 2.1578421592712402,
+      "learning_rate": 4.280425515174123e-05,
+      "loss": 0.3601,
+      "step": 1099
+    },
+    {
+      "epoch": 1.615723732549596,
+      "grad_norm": 1.9328081607818604,
+      "learning_rate": 4.278622763839376e-05,
+      "loss": 0.3316,
+      "step": 1100
+    },
+    {
+      "epoch": 1.6171932402645113,
+      "grad_norm": 1.6253280639648438,
+      "learning_rate": 4.276818137766118e-05,
+      "loss": 0.2713,
+      "step": 1101
+    },
+    {
+      "epoch": 1.6186627479794269,
+      "grad_norm": 2.368690252304077,
+      "learning_rate": 4.275011638856497e-05,
+      "loss": 0.4208,
+      "step": 1102
+    },
+    {
+      "epoch": 1.6201322556943425,
+      "grad_norm": 2.1482365131378174,
+      "learning_rate": 4.273203269014634e-05,
+      "loss": 0.477,
+      "step": 1103
+    },
+    {
+      "epoch": 1.6216017634092579,
+      "grad_norm": 1.9967759847640991,
+      "learning_rate": 4.271393030146621e-05,
+      "loss": 0.3707,
+      "step": 1104
+    },
+    {
+      "epoch": 1.6230712711241733,
+      "grad_norm": 2.9055097103118896,
+      "learning_rate": 4.269580924160523e-05,
+      "loss": 0.4022,
+      "step": 1105
+    },
+    {
+      "epoch": 1.6245407788390889,
+      "grad_norm": 1.7754637002944946,
+      "learning_rate": 4.267766952966369e-05,
+      "loss": 0.3405,
+      "step": 1106
+    },
+    {
+      "epoch": 1.6260102865540045,
+      "grad_norm": 1.8170151710510254,
+      "learning_rate": 4.265951118476158e-05,
+      "loss": 0.345,
+      "step": 1107
+    },
+    {
+      "epoch": 1.6274797942689199,
+      "grad_norm": 2.035055160522461,
+      "learning_rate": 4.26413342260385e-05,
+      "loss": 0.3531,
+      "step": 1108
+    },
+    {
+      "epoch": 1.6289493019838353,
+      "grad_norm": 1.9751542806625366,
+      "learning_rate": 4.2623138672653684e-05,
+      "loss": 0.4448,
+      "step": 1109
+    },
+    {
+      "epoch": 1.6304188096987509,
+      "grad_norm": 2.1838934421539307,
+      "learning_rate": 4.2604924543785965e-05,
+      "loss": 0.4536,
+      "step": 1110
+    },
+    {
+      "epoch": 1.6318883174136665,
+      "grad_norm": 2.1768908500671387,
+      "learning_rate": 4.2586691858633747e-05,
+      "loss": 0.424,
+      "step": 1111
+    },
+    {
+      "epoch": 1.633357825128582,
+      "grad_norm": 1.7382866144180298,
+      "learning_rate": 4.2568440636415007e-05,
+      "loss": 0.3456,
+      "step": 1112
+    },
+    {
+      "epoch": 1.6348273328434975,
+      "grad_norm": 1.695993185043335,
+      "learning_rate": 4.2550170896367245e-05,
+      "loss": 0.3152,
+      "step": 1113
+    },
+    {
+      "epoch": 1.6362968405584128,
+      "grad_norm": 1.7283481359481812,
+      "learning_rate": 4.253188265774749e-05,
+      "loss": 0.3876,
+      "step": 1114
+    },
+    {
+      "epoch": 1.6377663482733285,
+      "grad_norm": 1.4989573955535889,
+      "learning_rate": 4.2513575939832275e-05,
+      "loss": 0.3087,
+      "step": 1115
+    },
+    {
+      "epoch": 1.639235855988244,
+      "grad_norm": 2.858825206756592,
+      "learning_rate": 4.249525076191759e-05,
+      "loss": 0.4843,
+      "step": 1116
+    },
+    {
+      "epoch": 1.6407053637031594,
+      "grad_norm": 1.8042891025543213,
+      "learning_rate": 4.24769071433189e-05,
+      "loss": 0.3262,
+      "step": 1117
+    },
+    {
+      "epoch": 1.6421748714180748,
+      "grad_norm": 1.9758857488632202,
+      "learning_rate": 4.2458545103371105e-05,
+      "loss": 0.3661,
+      "step": 1118
+    },
+    {
+      "epoch": 1.6436443791329904,
+      "grad_norm": 1.842680811882019,
+      "learning_rate": 4.2440164661428515e-05,
+      "loss": 0.3983,
+      "step": 1119
+    },
+    {
+      "epoch": 1.645113886847906,
+      "grad_norm": 1.7211142778396606,
+      "learning_rate": 4.2421765836864844e-05,
+      "loss": 0.2519,
+      "step": 1120
+    },
+    {
+      "epoch": 1.6465833945628214,
+      "grad_norm": 2.680864095687866,
+      "learning_rate": 4.2403348649073174e-05,
+      "loss": 0.5681,
+      "step": 1121
+    },
+    {
+      "epoch": 1.6480529022777368,
+      "grad_norm": 2.027303457260132,
+      "learning_rate": 4.238491311746594e-05,
+      "loss": 0.3507,
+      "step": 1122
+    },
+    {
+      "epoch": 1.6495224099926524,
+      "grad_norm": 1.9026082754135132,
+      "learning_rate": 4.2366459261474933e-05,
+      "loss": 0.408,
+      "step": 1123
+    },
+    {
+      "epoch": 1.650991917707568,
+      "grad_norm": 1.9093105792999268,
+      "learning_rate": 4.234798710055125e-05,
+      "loss": 0.4044,
+      "step": 1124
+    },
+    {
+      "epoch": 1.6524614254224834,
+      "grad_norm": 2.508981704711914,
+      "learning_rate": 4.232949665416525e-05,
+      "loss": 0.4933,
+      "step": 1125
+    },
+    {
+      "epoch": 1.6539309331373988,
+      "grad_norm": 2.8525471687316895,
+      "learning_rate": 4.2310987941806616e-05,
+      "loss": 0.5198,
+      "step": 1126
+    },
+    {
+      "epoch": 1.6554004408523144,
+      "grad_norm": 1.7180473804473877,
+      "learning_rate": 4.2292460982984255e-05,
+      "loss": 0.3506,
+      "step": 1127
+    },
+    {
+      "epoch": 1.65686994856723,
+      "grad_norm": 1.6653529405593872,
+      "learning_rate": 4.22739157972263e-05,
+      "loss": 0.3042,
+      "step": 1128
+    },
+    {
+      "epoch": 1.6583394562821456,
+      "grad_norm": 2.2206649780273438,
+      "learning_rate": 4.2255352404080134e-05,
+      "loss": 0.358,
+      "step": 1129
+    },
+    {
+      "epoch": 1.659808963997061,
+      "grad_norm": 2.0602469444274902,
+      "learning_rate": 4.223677082311228e-05,
+      "loss": 0.4192,
+      "step": 1130
+    },
+    {
+      "epoch": 1.6612784717119764,
+      "grad_norm": 1.8140531778335571,
+      "learning_rate": 4.221817107390847e-05,
+      "loss": 0.2925,
+      "step": 1131
+    },
+    {
+      "epoch": 1.662747979426892,
+      "grad_norm": 2.064891815185547,
+      "learning_rate": 4.2199553176073565e-05,
+      "loss": 0.4016,
+      "step": 1132
+    },
+    {
+      "epoch": 1.6642174871418076,
+      "grad_norm": 2.393601894378662,
+      "learning_rate": 4.218091714923157e-05,
+      "loss": 0.4643,
+      "step": 1133
+    },
+    {
+      "epoch": 1.665686994856723,
+      "grad_norm": 1.5757851600646973,
+      "learning_rate": 4.216226301302559e-05,
+      "loss": 0.269,
+      "step": 1134
+    },
+    {
+      "epoch": 1.6671565025716384,
+      "grad_norm": 2.0919063091278076,
+      "learning_rate": 4.2143590787117814e-05,
+      "loss": 0.4503,
+      "step": 1135
+    },
+    {
+      "epoch": 1.668626010286554,
+      "grad_norm": 2.2089176177978516,
+      "learning_rate": 4.2124900491189514e-05,
+      "loss": 0.4531,
+      "step": 1136
+    },
+    {
+      "epoch": 1.6700955180014696,
+      "grad_norm": 2.535834789276123,
+      "learning_rate": 4.210619214494099e-05,
+      "loss": 0.4219,
+      "step": 1137
+    },
+    {
+      "epoch": 1.671565025716385,
+      "grad_norm": 1.4843156337738037,
+      "learning_rate": 4.208746576809159e-05,
+      "loss": 0.2882,
+      "step": 1138
+    },
+    {
+      "epoch": 1.6730345334313004,
+      "grad_norm": 2.475684404373169,
+      "learning_rate": 4.206872138037964e-05,
+      "loss": 0.4238,
+      "step": 1139
+    },
+    {
+      "epoch": 1.674504041146216,
+      "grad_norm": 2.036083459854126,
+      "learning_rate": 4.2049959001562464e-05,
+      "loss": 0.3526,
+      "step": 1140
+    },
+    {
+      "epoch": 1.6759735488611316,
+      "grad_norm": 1.8914906978607178,
+      "learning_rate": 4.203117865141635e-05,
+      "loss": 0.3908,
+      "step": 1141
+    },
+    {
+      "epoch": 1.677443056576047,
+      "grad_norm": 1.924364447593689,
+      "learning_rate": 4.201238034973654e-05,
+      "loss": 0.4249,
+      "step": 1142
+    },
+    {
+      "epoch": 1.6789125642909624,
+      "grad_norm": 1.4568886756896973,
+      "learning_rate": 4.199356411633717e-05,
+      "loss": 0.2765,
+      "step": 1143
+    },
+    {
+      "epoch": 1.680382072005878,
+      "grad_norm": 1.8813097476959229,
+      "learning_rate": 4.19747299710513e-05,
+      "loss": 0.3561,
+      "step": 1144
+    },
+    {
+      "epoch": 1.6818515797207936,
+      "grad_norm": 2.0658230781555176,
+      "learning_rate": 4.1955877933730855e-05,
+      "loss": 0.4128,
+      "step": 1145
+    },
+    {
+      "epoch": 1.6833210874357092,
+      "grad_norm": 2.402867078781128,
+      "learning_rate": 4.193700802424663e-05,
+      "loss": 0.4368,
+      "step": 1146
+    },
+    {
+      "epoch": 1.6847905951506246,
+      "grad_norm": 2.06353497505188,
+      "learning_rate": 4.191812026248825e-05,
+      "loss": 0.4086,
+      "step": 1147
+    },
+    {
+      "epoch": 1.68626010286554,
+      "grad_norm": 1.5699948072433472,
+      "learning_rate": 4.189921466836416e-05,
+      "loss": 0.2687,
+      "step": 1148
+    },
+    {
+      "epoch": 1.6877296105804556,
+      "grad_norm": 2.1702752113342285,
+      "learning_rate": 4.188029126180161e-05,
+      "loss": 0.3689,
+      "step": 1149
+    },
+    {
+      "epoch": 1.6891991182953712,
+      "grad_norm": 1.977005958557129,
+      "learning_rate": 4.186135006274661e-05,
+      "loss": 0.4708,
+      "step": 1150
+    },
+    {
+      "epoch": 1.6906686260102866,
+      "grad_norm": 2.121612310409546,
+      "learning_rate": 4.184239109116393e-05,
+      "loss": 0.3858,
+      "step": 1151
+    },
+    {
+      "epoch": 1.692138133725202,
+      "grad_norm": 1.6749706268310547,
+      "learning_rate": 4.182341436703708e-05,
+      "loss": 0.3735,
+      "step": 1152
+    },
+    {
+      "epoch": 1.6936076414401176,
+      "grad_norm": 1.9160727262496948,
+      "learning_rate": 4.1804419910368265e-05,
+      "loss": 0.3745,
+      "step": 1153
+    },
+    {
+      "epoch": 1.6950771491550332,
+      "grad_norm": 1.9433330297470093,
+      "learning_rate": 4.17854077411784e-05,
+      "loss": 0.4537,
+      "step": 1154
+    },
+    {
+      "epoch": 1.6965466568699485,
+      "grad_norm": 2.3486084938049316,
+      "learning_rate": 4.1766377879507056e-05,
+      "loss": 0.3884,
+      "step": 1155
+    },
+    {
+      "epoch": 1.698016164584864,
+      "grad_norm": 1.4544788599014282,
+      "learning_rate": 4.1747330345412455e-05,
+      "loss": 0.2848,
+      "step": 1156
+    },
+    {
+      "epoch": 1.6994856722997795,
+      "grad_norm": 1.782996654510498,
+      "learning_rate": 4.172826515897146e-05,
+      "loss": 0.366,
+      "step": 1157
+    },
+    {
+      "epoch": 1.7009551800146951,
+      "grad_norm": 1.9279265403747559,
+      "learning_rate": 4.1709182340279516e-05,
+      "loss": 0.3759,
+      "step": 1158
+    },
+    {
+      "epoch": 1.7024246877296105,
+      "grad_norm": 2.2062206268310547,
+      "learning_rate": 4.1690081909450674e-05,
+      "loss": 0.5587,
+      "step": 1159
+    },
+    {
+      "epoch": 1.7038941954445261,
+      "grad_norm": 1.9330358505249023,
+      "learning_rate": 4.1670963886617535e-05,
+      "loss": 0.3143,
+      "step": 1160
+    },
+    {
+      "epoch": 1.7053637031594415,
+      "grad_norm": 2.0875959396362305,
+      "learning_rate": 4.1651828291931264e-05,
+      "loss": 0.3901,
+      "step": 1161
+    },
+    {
+      "epoch": 1.7068332108743571,
+      "grad_norm": 1.4467285871505737,
+      "learning_rate": 4.1632675145561504e-05,
+      "loss": 0.2656,
+      "step": 1162
+    },
+    {
+      "epoch": 1.7083027185892727,
+      "grad_norm": 1.5371090173721313,
+      "learning_rate": 4.161350446769645e-05,
+      "loss": 0.3253,
+      "step": 1163
+    },
+    {
+      "epoch": 1.7097722263041881,
+      "grad_norm": 2.5458199977874756,
+      "learning_rate": 4.159431627854275e-05,
+      "loss": 0.5397,
+      "step": 1164
+    },
+    {
+      "epoch": 1.7112417340191035,
+      "grad_norm": 3.022674083709717,
+      "learning_rate": 4.157511059832551e-05,
+      "loss": 0.5779,
+      "step": 1165
+    },
+    {
+      "epoch": 1.7127112417340191,
+      "grad_norm": 2.2920355796813965,
+      "learning_rate": 4.155588744728826e-05,
+      "loss": 0.5333,
+      "step": 1166
+    },
+    {
+      "epoch": 1.7141807494489347,
+      "grad_norm": 3.3242571353912354,
+      "learning_rate": 4.153664684569297e-05,
+      "loss": 0.534,
+      "step": 1167
+    },
+    {
+      "epoch": 1.7156502571638501,
+      "grad_norm": 1.80581533908844,
+      "learning_rate": 4.151738881382e-05,
+      "loss": 0.2885,
+      "step": 1168
+    },
+    {
+      "epoch": 1.7171197648787655,
+      "grad_norm": 1.7348268032073975,
+      "learning_rate": 4.149811337196807e-05,
+      "loss": 0.2902,
+      "step": 1169
+    },
+    {
+      "epoch": 1.718589272593681,
+      "grad_norm": 2.031843662261963,
+      "learning_rate": 4.1478820540454246e-05,
+      "loss": 0.3879,
+      "step": 1170
+    },
+    {
+      "epoch": 1.7200587803085967,
+      "grad_norm": 2.259049892425537,
+      "learning_rate": 4.145951033961395e-05,
+      "loss": 0.3906,
+      "step": 1171
+    },
+    {
+      "epoch": 1.721528288023512,
+      "grad_norm": 2.7384800910949707,
+      "learning_rate": 4.144018278980088e-05,
+      "loss": 0.6318,
+      "step": 1172
+    },
+    {
+      "epoch": 1.7229977957384275,
+      "grad_norm": 2.0385689735412598,
+      "learning_rate": 4.142083791138703e-05,
+      "loss": 0.3626,
+      "step": 1173
+    },
+    {
+      "epoch": 1.724467303453343,
+      "grad_norm": 2.582695245742798,
+      "learning_rate": 4.140147572476268e-05,
+      "loss": 0.4428,
+      "step": 1174
+    },
+    {
+      "epoch": 1.7259368111682587,
+      "grad_norm": 1.631521463394165,
+      "learning_rate": 4.1382096250336346e-05,
+      "loss": 0.371,
+      "step": 1175
+    },
+    {
+      "epoch": 1.727406318883174,
+      "grad_norm": 1.9476027488708496,
+      "learning_rate": 4.136269950853473e-05,
+      "loss": 0.3495,
+      "step": 1176
+    },
+    {
+      "epoch": 1.7288758265980897,
+      "grad_norm": 2.1872546672821045,
+      "learning_rate": 4.134328551980279e-05,
+      "loss": 0.5318,
+      "step": 1177
+    },
+    {
+      "epoch": 1.730345334313005,
+      "grad_norm": 2.1452524662017822,
+      "learning_rate": 4.132385430460361e-05,
+      "loss": 0.3599,
+      "step": 1178
+    },
+    {
+      "epoch": 1.7318148420279207,
+      "grad_norm": 2.413681745529175,
+      "learning_rate": 4.130440588341847e-05,
+      "loss": 0.3778,
+      "step": 1179
+    },
+    {
+      "epoch": 1.7332843497428363,
+      "grad_norm": 1.960511565208435,
+      "learning_rate": 4.128494027674678e-05,
+      "loss": 0.3784,
+      "step": 1180
+    },
+    {
+      "epoch": 1.7347538574577517,
+      "grad_norm": 2.0622100830078125,
+      "learning_rate": 4.126545750510605e-05,
+      "loss": 0.4684,
+      "step": 1181
+    },
+    {
+      "epoch": 1.736223365172667,
+      "grad_norm": 2.4012298583984375,
+      "learning_rate": 4.124595758903189e-05,
+      "loss": 0.5726,
+      "step": 1182
+    },
+    {
+      "epoch": 1.7376928728875827,
+      "grad_norm": 2.0459372997283936,
+      "learning_rate": 4.122644054907798e-05,
+      "loss": 0.4455,
+      "step": 1183
+    },
+    {
+      "epoch": 1.7391623806024983,
+      "grad_norm": 2.459186553955078,
+      "learning_rate": 4.1206906405816036e-05,
+      "loss": 0.4605,
+      "step": 1184
+    },
+    {
+      "epoch": 1.7406318883174137,
+      "grad_norm": 1.8493303060531616,
+      "learning_rate": 4.1187355179835844e-05,
+      "loss": 0.3392,
+      "step": 1185
+    },
+    {
+      "epoch": 1.742101396032329,
+      "grad_norm": 1.5771507024765015,
+      "learning_rate": 4.1167786891745144e-05,
+      "loss": 0.3145,
+      "step": 1186
+    },
+    {
+      "epoch": 1.7435709037472447,
+      "grad_norm": 1.961456537246704,
+      "learning_rate": 4.1148201562169685e-05,
+      "loss": 0.3375,
+      "step": 1187
+    },
+    {
+      "epoch": 1.7450404114621603,
+      "grad_norm": 1.811500072479248,
+      "learning_rate": 4.1128599211753184e-05,
+      "loss": 0.351,
+      "step": 1188
+    },
+    {
+      "epoch": 1.7465099191770757,
+      "grad_norm": 2.158924102783203,
+      "learning_rate": 4.110897986115729e-05,
+      "loss": 0.4753,
+      "step": 1189
+    },
+    {
+      "epoch": 1.747979426891991,
+      "grad_norm": 1.5972789525985718,
+      "learning_rate": 4.108934353106156e-05,
+      "loss": 0.3453,
+      "step": 1190
+    },
+    {
+      "epoch": 1.7494489346069066,
+      "grad_norm": 2.1524298191070557,
+      "learning_rate": 4.1069690242163484e-05,
+      "loss": 0.4603,
+      "step": 1191
+    },
+    {
+      "epoch": 1.7509184423218223,
+      "grad_norm": 1.6004942655563354,
+      "learning_rate": 4.105002001517838e-05,
+      "loss": 0.2946,
+      "step": 1192
+    },
+    {
+      "epoch": 1.7523879500367376,
+      "grad_norm": 2.209228754043579,
+      "learning_rate": 4.103033287083946e-05,
+      "loss": 0.3327,
+      "step": 1193
+    },
+    {
+      "epoch": 1.7538574577516532,
+      "grad_norm": 2.0604615211486816,
+      "learning_rate": 4.101062882989774e-05,
+      "loss": 0.4422,
+      "step": 1194
+    },
+    {
+      "epoch": 1.7553269654665686,
+      "grad_norm": 2.261294364929199,
+      "learning_rate": 4.0990907913122056e-05,
+      "loss": 0.4412,
+      "step": 1195
+    },
+    {
+      "epoch": 1.7567964731814842,
+      "grad_norm": 1.7700508832931519,
+      "learning_rate": 4.097117014129903e-05,
+      "loss": 0.4055,
+      "step": 1196
+    },
+    {
+      "epoch": 1.7582659808963998,
+      "grad_norm": 1.5785746574401855,
+      "learning_rate": 4.095141553523306e-05,
+      "loss": 0.3161,
+      "step": 1197
+    },
+    {
+      "epoch": 1.7597354886113152,
+      "grad_norm": 1.6752835512161255,
+      "learning_rate": 4.093164411574628e-05,
+      "loss": 0.3669,
+      "step": 1198
+    },
+    {
+      "epoch": 1.7612049963262306,
+      "grad_norm": 1.90375554561615,
+      "learning_rate": 4.0911855903678534e-05,
+      "loss": 0.3296,
+      "step": 1199
+    },
+    {
+      "epoch": 1.7626745040411462,
+      "grad_norm": 1.7571159601211548,
+      "learning_rate": 4.0892050919887394e-05,
+      "loss": 0.2528,
+      "step": 1200
+    },
+    {
+      "epoch": 1.7641440117560618,
+      "grad_norm": 2.4218101501464844,
+      "learning_rate": 4.0872229185248075e-05,
+      "loss": 0.3307,
+      "step": 1201
+    },
+    {
+      "epoch": 1.7656135194709772,
+      "grad_norm": 1.7807369232177734,
+      "learning_rate": 4.085239072065348e-05,
+      "loss": 0.3797,
+      "step": 1202
+    },
+    {
+      "epoch": 1.7670830271858926,
+      "grad_norm": 2.155738353729248,
+      "learning_rate": 4.083253554701412e-05,
+      "loss": 0.361,
+      "step": 1203
+    },
+    {
+      "epoch": 1.7685525349008082,
+      "grad_norm": 2.1835241317749023,
+      "learning_rate": 4.081266368525815e-05,
+      "loss": 0.3446,
+      "step": 1204
+    },
+    {
+      "epoch": 1.7700220426157238,
+      "grad_norm": 2.2820212841033936,
+      "learning_rate": 4.079277515633127e-05,
+      "loss": 0.4502,
+      "step": 1205
+    },
+    {
+      "epoch": 1.7714915503306392,
+      "grad_norm": 2.1415176391601562,
+      "learning_rate": 4.0772869981196806e-05,
+      "loss": 0.4256,
+      "step": 1206
+    },
+    {
+      "epoch": 1.7729610580455546,
+      "grad_norm": 1.8384228944778442,
+      "learning_rate": 4.075294818083557e-05,
+      "loss": 0.3573,
+      "step": 1207
+    },
+    {
+      "epoch": 1.7744305657604702,
+      "grad_norm": 2.28544282913208,
+      "learning_rate": 4.073300977624594e-05,
+      "loss": 0.3772,
+      "step": 1208
+    },
+    {
+      "epoch": 1.7759000734753858,
+      "grad_norm": 1.811877965927124,
+      "learning_rate": 4.071305478844377e-05,
+      "loss": 0.3606,
+      "step": 1209
+    },
+    {
+      "epoch": 1.7773695811903012,
+      "grad_norm": 2.3214685916900635,
+      "learning_rate": 4.069308323846241e-05,
+      "loss": 0.378,
+      "step": 1210
+    },
+    {
+      "epoch": 1.7788390889052168,
+      "grad_norm": 2.096066951751709,
+      "learning_rate": 4.067309514735267e-05,
+      "loss": 0.3782,
+      "step": 1211
+    },
+    {
+      "epoch": 1.7803085966201322,
+      "grad_norm": 2.678314447402954,
+      "learning_rate": 4.065309053618276e-05,
+      "loss": 0.4899,
+      "step": 1212
+    },
+    {
+      "epoch": 1.7817781043350478,
+      "grad_norm": 2.141754388809204,
+      "learning_rate": 4.063306942603835e-05,
+      "loss": 0.3541,
+      "step": 1213
+    },
+    {
+      "epoch": 1.7832476120499634,
+      "grad_norm": 2.173258066177368,
+      "learning_rate": 4.0613031838022486e-05,
+      "loss": 0.3608,
+      "step": 1214
+    },
+    {
+      "epoch": 1.7847171197648788,
+      "grad_norm": 2.4709153175354004,
+      "learning_rate": 4.059297779325555e-05,
+      "loss": 0.4862,
+      "step": 1215
+    },
+    {
+      "epoch": 1.7861866274797942,
+      "grad_norm": 1.8630024194717407,
+      "learning_rate": 4.0572907312875314e-05,
+      "loss": 0.2964,
+      "step": 1216
+    },
+    {
+      "epoch": 1.7876561351947098,
+      "grad_norm": 2.19879150390625,
+      "learning_rate": 4.055282041803685e-05,
+      "loss": 0.3767,
+      "step": 1217
+    },
+    {
+      "epoch": 1.7891256429096254,
+      "grad_norm": 1.854219913482666,
+      "learning_rate": 4.053271712991254e-05,
+      "loss": 0.379,
+      "step": 1218
+    },
+    {
+      "epoch": 1.7905951506245408,
+      "grad_norm": 1.8051823377609253,
+      "learning_rate": 4.0512597469692035e-05,
+      "loss": 0.3468,
+      "step": 1219
+    },
+    {
+      "epoch": 1.7920646583394562,
+      "grad_norm": 2.1510658264160156,
+      "learning_rate": 4.049246145858227e-05,
+      "loss": 0.4421,
+      "step": 1220
+    },
+    {
+      "epoch": 1.7935341660543718,
+      "grad_norm": 1.7533938884735107,
+      "learning_rate": 4.047230911780737e-05,
+      "loss": 0.3409,
+      "step": 1221
+    },
+    {
+      "epoch": 1.7950036737692874,
+      "grad_norm": 1.7940617799758911,
+      "learning_rate": 4.04521404686087e-05,
+      "loss": 0.3209,
+      "step": 1222
+    },
+    {
+      "epoch": 1.7964731814842028,
+      "grad_norm": 1.8579769134521484,
+      "learning_rate": 4.0431955532244827e-05,
+      "loss": 0.3443,
+      "step": 1223
+    },
+    {
+      "epoch": 1.7979426891991181,
+      "grad_norm": 1.817435383796692,
+      "learning_rate": 4.041175432999146e-05,
+      "loss": 0.2474,
+      "step": 1224
+    },
+    {
+      "epoch": 1.7994121969140338,
+      "grad_norm": 2.2651281356811523,
+      "learning_rate": 4.039153688314145e-05,
+      "loss": 0.4446,
+      "step": 1225
+    },
+    {
+      "epoch": 1.8008817046289494,
+      "grad_norm": 1.63055419921875,
+      "learning_rate": 4.037130321300482e-05,
+      "loss": 0.297,
+      "step": 1226
+    },
+    {
+      "epoch": 1.8023512123438648,
+      "grad_norm": 1.7453168630599976,
+      "learning_rate": 4.0351053340908615e-05,
+      "loss": 0.2949,
+      "step": 1227
+    },
+    {
+      "epoch": 1.8038207200587804,
+      "grad_norm": 2.2204678058624268,
+      "learning_rate": 4.0330787288197026e-05,
+      "loss": 0.4797,
+      "step": 1228
+    },
+    {
+      "epoch": 1.8052902277736957,
+      "grad_norm": 2.0886471271514893,
+      "learning_rate": 4.031050507623125e-05,
+      "loss": 0.4281,
+      "step": 1229
+    },
+    {
+      "epoch": 1.8067597354886114,
+      "grad_norm": 1.7772088050842285,
+      "learning_rate": 4.0290206726389556e-05,
+      "loss": 0.3814,
+      "step": 1230
+    },
+    {
+      "epoch": 1.808229243203527,
+      "grad_norm": 1.7894595861434937,
+      "learning_rate": 4.02698922600672e-05,
+      "loss": 0.392,
+      "step": 1231
+    },
+    {
+      "epoch": 1.8096987509184423,
+      "grad_norm": 2.2798566818237305,
+      "learning_rate": 4.0249561698676416e-05,
+      "loss": 0.4131,
+      "step": 1232
+    },
+    {
+      "epoch": 1.8111682586333577,
+      "grad_norm": 2.1407337188720703,
+      "learning_rate": 4.022921506364644e-05,
+      "loss": 0.4209,
+      "step": 1233
+    },
+    {
+      "epoch": 1.8126377663482733,
+      "grad_norm": 1.8522671461105347,
+      "learning_rate": 4.020885237642342e-05,
+      "loss": 0.3764,
+      "step": 1234
+    },
+    {
+      "epoch": 1.814107274063189,
+      "grad_norm": 1.8190054893493652,
+      "learning_rate": 4.0188473658470426e-05,
+      "loss": 0.3032,
+      "step": 1235
+    },
+    {
+      "epoch": 1.8155767817781043,
+      "grad_norm": 2.3600475788116455,
+      "learning_rate": 4.0168078931267425e-05,
+      "loss": 0.5616,
+      "step": 1236
+    },
+    {
+      "epoch": 1.8170462894930197,
+      "grad_norm": 1.51101553440094,
+      "learning_rate": 4.014766821631128e-05,
+      "loss": 0.2724,
+      "step": 1237
+    },
+    {
+      "epoch": 1.8185157972079353,
+      "grad_norm": 3.2337160110473633,
+      "learning_rate": 4.0127241535115676e-05,
+      "loss": 0.5413,
+      "step": 1238
+    },
+    {
+      "epoch": 1.819985304922851,
+      "grad_norm": 2.5443623065948486,
+      "learning_rate": 4.0106798909211144e-05,
+      "loss": 0.4103,
+      "step": 1239
+    },
+    {
+      "epoch": 1.8214548126377663,
+      "grad_norm": 1.919850468635559,
+      "learning_rate": 4.008634036014501e-05,
+      "loss": 0.5127,
+      "step": 1240
+    },
+    {
+      "epoch": 1.8229243203526817,
+      "grad_norm": 2.3663110733032227,
+      "learning_rate": 4.0065865909481417e-05,
+      "loss": 0.4036,
+      "step": 1241
+    },
+    {
+      "epoch": 1.8243938280675973,
+      "grad_norm": 1.786495327949524,
+      "learning_rate": 4.0045375578801214e-05,
+      "loss": 0.3716,
+      "step": 1242
+    },
+    {
+      "epoch": 1.825863335782513,
+      "grad_norm": 1.9357807636260986,
+      "learning_rate": 4.002486938970202e-05,
+      "loss": 0.3744,
+      "step": 1243
+    },
+    {
+      "epoch": 1.8273328434974285,
+      "grad_norm": 2.187225341796875,
+      "learning_rate": 4.000434736379819e-05,
+      "loss": 0.4375,
+      "step": 1244
+    },
+    {
+      "epoch": 1.828802351212344,
+      "grad_norm": 1.7615622282028198,
+      "learning_rate": 3.998380952272073e-05,
+      "loss": 0.3301,
+      "step": 1245
+    },
+    {
+      "epoch": 1.8302718589272593,
+      "grad_norm": 1.3644670248031616,
+      "learning_rate": 3.9963255888117326e-05,
+      "loss": 0.2679,
+      "step": 1246
+    },
+    {
+      "epoch": 1.831741366642175,
+      "grad_norm": 2.007152557373047,
+      "learning_rate": 3.9942686481652344e-05,
+      "loss": 0.4503,
+      "step": 1247
+    },
+    {
+      "epoch": 1.8332108743570905,
+      "grad_norm": 1.4597958326339722,
+      "learning_rate": 3.9922101325006736e-05,
+      "loss": 0.2412,
+      "step": 1248
+    },
+    {
+      "epoch": 1.834680382072006,
+      "grad_norm": 1.8932536840438843,
+      "learning_rate": 3.990150043987806e-05,
+      "loss": 0.3197,
+      "step": 1249
+    },
+    {
+      "epoch": 1.8361498897869213,
+      "grad_norm": 2.2570950984954834,
+      "learning_rate": 3.988088384798047e-05,
+      "loss": 0.4996,
+      "step": 1250
+    },
+    {
+      "epoch": 1.837619397501837,
+      "grad_norm": 1.6372095346450806,
+      "learning_rate": 3.986025157104467e-05,
+      "loss": 0.3378,
+      "step": 1251
+    },
+    {
+      "epoch": 1.8390889052167525,
+      "grad_norm": 1.887732982635498,
+      "learning_rate": 3.983960363081789e-05,
+      "loss": 0.3594,
+      "step": 1252
+    },
+    {
+      "epoch": 1.8405584129316679,
+      "grad_norm": 2.1195032596588135,
+      "learning_rate": 3.981894004906388e-05,
+      "loss": 0.34,
+      "step": 1253
+    },
+    {
+      "epoch": 1.8420279206465833,
+      "grad_norm": 2.2256240844726562,
+      "learning_rate": 3.979826084756285e-05,
+      "loss": 0.4189,
+      "step": 1254
+    },
+    {
+      "epoch": 1.8434974283614989,
+      "grad_norm": 2.555799961090088,
+      "learning_rate": 3.9777566048111516e-05,
+      "loss": 0.4987,
+      "step": 1255
+    },
+    {
+      "epoch": 1.8449669360764145,
+      "grad_norm": 2.3646445274353027,
+      "learning_rate": 3.9756855672522993e-05,
+      "loss": 0.3647,
+      "step": 1256
+    },
+    {
+      "epoch": 1.8464364437913299,
+      "grad_norm": 1.5152289867401123,
+      "learning_rate": 3.9736129742626846e-05,
+      "loss": 0.242,
+      "step": 1257
+    },
+    {
+      "epoch": 1.8479059515062453,
+      "grad_norm": 1.6087247133255005,
+      "learning_rate": 3.9715388280269016e-05,
+      "loss": 0.2982,
+      "step": 1258
+    },
+    {
+      "epoch": 1.8493754592211609,
+      "grad_norm": 2.2901556491851807,
+      "learning_rate": 3.969463130731183e-05,
+      "loss": 0.4348,
+      "step": 1259
+    },
+    {
+      "epoch": 1.8508449669360765,
+      "grad_norm": 2.135261058807373,
+      "learning_rate": 3.967385884563395e-05,
+      "loss": 0.3982,
+      "step": 1260
+    },
+    {
+      "epoch": 1.852314474650992,
+      "grad_norm": 2.0445687770843506,
+      "learning_rate": 3.965307091713037e-05,
+      "loss": 0.3749,
+      "step": 1261
+    },
+    {
+      "epoch": 1.8537839823659075,
+      "grad_norm": 1.9856984615325928,
+      "learning_rate": 3.9632267543712387e-05,
+      "loss": 0.3509,
+      "step": 1262
+    },
+    {
+      "epoch": 1.8552534900808229,
+      "grad_norm": 1.4575191736221313,
+      "learning_rate": 3.961144874730758e-05,
+      "loss": 0.2431,
+      "step": 1263
+    },
+    {
+      "epoch": 1.8567229977957385,
+      "grad_norm": 1.7664800882339478,
+      "learning_rate": 3.9590614549859764e-05,
+      "loss": 0.2916,
+      "step": 1264
+    },
+    {
+      "epoch": 1.858192505510654,
+      "grad_norm": 1.9742193222045898,
+      "learning_rate": 3.9569764973329026e-05,
+      "loss": 0.3162,
+      "step": 1265
+    },
+    {
+      "epoch": 1.8596620132255695,
+      "grad_norm": 2.2139391899108887,
+      "learning_rate": 3.9548900039691635e-05,
+      "loss": 0.523,
+      "step": 1266
+    },
+    {
+      "epoch": 1.8611315209404848,
+      "grad_norm": 2.7148094177246094,
+      "learning_rate": 3.952801977094005e-05,
+      "loss": 0.5337,
+      "step": 1267
+    },
+    {
+      "epoch": 1.8626010286554004,
+      "grad_norm": 2.2409133911132812,
+      "learning_rate": 3.950712418908289e-05,
+      "loss": 0.4315,
+      "step": 1268
+    },
+    {
+      "epoch": 1.864070536370316,
+      "grad_norm": 2.1831066608428955,
+      "learning_rate": 3.948621331614495e-05,
+      "loss": 0.3519,
+      "step": 1269
+    },
+    {
+      "epoch": 1.8655400440852314,
+      "grad_norm": 2.3470141887664795,
+      "learning_rate": 3.9465287174167085e-05,
+      "loss": 0.3559,
+      "step": 1270
+    },
+    {
+      "epoch": 1.8670095518001468,
+      "grad_norm": 2.164759874343872,
+      "learning_rate": 3.9444345785206285e-05,
+      "loss": 0.3346,
+      "step": 1271
+    },
+    {
+      "epoch": 1.8684790595150624,
+      "grad_norm": 1.7910867929458618,
+      "learning_rate": 3.94233891713356e-05,
+      "loss": 0.3206,
+      "step": 1272
+    },
+    {
+      "epoch": 1.869948567229978,
+      "grad_norm": 2.1369168758392334,
+      "learning_rate": 3.940241735464412e-05,
+      "loss": 0.3781,
+      "step": 1273
+    },
+    {
+      "epoch": 1.8714180749448934,
+      "grad_norm": 2.3138229846954346,
+      "learning_rate": 3.938143035723697e-05,
+      "loss": 0.5165,
+      "step": 1274
+    },
+    {
+      "epoch": 1.8728875826598088,
+      "grad_norm": 2.3118338584899902,
+      "learning_rate": 3.936042820123529e-05,
+      "loss": 0.494,
+      "step": 1275
+    },
+    {
+      "epoch": 1.8743570903747244,
+      "grad_norm": 1.8929877281188965,
+      "learning_rate": 3.933941090877615e-05,
+      "loss": 0.2745,
+      "step": 1276
+    },
+    {
+      "epoch": 1.87582659808964,
+      "grad_norm": 1.9124139547348022,
+      "learning_rate": 3.931837850201263e-05,
+      "loss": 0.3496,
+      "step": 1277
+    },
+    {
+      "epoch": 1.8772961058045556,
+      "grad_norm": 2.2020812034606934,
+      "learning_rate": 3.9297331003113705e-05,
+      "loss": 0.4284,
+      "step": 1278
+    },
+    {
+      "epoch": 1.878765613519471,
+      "grad_norm": 2.100343704223633,
+      "learning_rate": 3.927626843426427e-05,
+      "loss": 0.385,
+      "step": 1279
+    },
+    {
+      "epoch": 1.8802351212343864,
+      "grad_norm": 1.9978796243667603,
+      "learning_rate": 3.92551908176651e-05,
+      "loss": 0.3771,
+      "step": 1280
+    },
+    {
+      "epoch": 1.881704628949302,
+      "grad_norm": 2.136000633239746,
+      "learning_rate": 3.923409817553284e-05,
+      "loss": 0.3915,
+      "step": 1281
+    },
+    {
+      "epoch": 1.8831741366642176,
+      "grad_norm": 2.046025514602661,
+      "learning_rate": 3.9212990530099956e-05,
+      "loss": 0.407,
+      "step": 1282
+    },
+    {
+      "epoch": 1.884643644379133,
+      "grad_norm": 1.8668262958526611,
+      "learning_rate": 3.919186790361475e-05,
+      "loss": 0.4574,
+      "step": 1283
+    },
+    {
+      "epoch": 1.8861131520940484,
+      "grad_norm": 2.0905561447143555,
+      "learning_rate": 3.9170730318341304e-05,
+      "loss": 0.3743,
+      "step": 1284
+    },
+    {
+      "epoch": 1.887582659808964,
+      "grad_norm": 2.2915642261505127,
+      "learning_rate": 3.914957779655945e-05,
+      "loss": 0.4263,
+      "step": 1285
+    },
+    {
+      "epoch": 1.8890521675238796,
+      "grad_norm": 1.7058526277542114,
+      "learning_rate": 3.91284103605648e-05,
+      "loss": 0.3441,
+      "step": 1286
+    },
+    {
+      "epoch": 1.890521675238795,
+      "grad_norm": 2.4302480220794678,
+      "learning_rate": 3.910722803266865e-05,
+      "loss": 0.5921,
+      "step": 1287
+    },
+    {
+      "epoch": 1.8919911829537104,
+      "grad_norm": 2.3298866748809814,
+      "learning_rate": 3.908603083519802e-05,
+      "loss": 0.3793,
+      "step": 1288
+    },
+    {
+      "epoch": 1.893460690668626,
+      "grad_norm": Infinity,
+      "learning_rate": 3.906481879049558e-05,
+      "loss": 0.3615,
+      "step": 1289
+    },
+    {
+      "epoch": 1.8949301983835416,
+      "grad_norm": 2.2267322540283203,
+      "learning_rate": 3.906481879049558e-05,
+      "loss": 0.5245,
+      "step": 1290
+    },
+    {
+      "epoch": 1.896399706098457,
+      "grad_norm": 2.3007447719573975,
+      "learning_rate": 3.904359192091967e-05,
+      "loss": 0.3558,
+      "step": 1291
+    },
+    {
+      "epoch": 1.8978692138133724,
+      "grad_norm": 2.4194750785827637,
+      "learning_rate": 3.902235024884425e-05,
+      "loss": 0.3725,
+      "step": 1292
+    },
+    {
+      "epoch": 1.899338721528288,
+      "grad_norm": 2.0349271297454834,
+      "learning_rate": 3.900109379665887e-05,
+      "loss": 0.3785,
+      "step": 1293
+    },
+    {
+      "epoch": 1.9008082292432036,
+      "grad_norm": 1.733258605003357,
+      "learning_rate": 3.897982258676867e-05,
+      "loss": 0.3955,
+      "step": 1294
+    },
+    {
+      "epoch": 1.9022777369581192,
+      "grad_norm": 1.9881511926651,
+      "learning_rate": 3.895853664159436e-05,
+      "loss": 0.3387,
+      "step": 1295
+    },
+    {
+      "epoch": 1.9037472446730346,
+      "grad_norm": 1.9925830364227295,
+      "learning_rate": 3.893723598357214e-05,
+      "loss": 0.4346,
+      "step": 1296
+    },
+    {
+      "epoch": 1.90521675238795,
+      "grad_norm": 1.5361812114715576,
+      "learning_rate": 3.891592063515376e-05,
+      "loss": 0.3259,
+      "step": 1297
+    },
+    {
+      "epoch": 1.9066862601028656,
+      "grad_norm": 1.8852967023849487,
+      "learning_rate": 3.8894590618806436e-05,
+      "loss": 0.4228,
+      "step": 1298
+    },
+    {
+      "epoch": 1.9081557678177812,
+      "grad_norm": 2.0385208129882812,
+      "learning_rate": 3.8873245957012846e-05,
+      "loss": 0.4588,
+      "step": 1299
+    },
+    {
+      "epoch": 1.9096252755326966,
+      "grad_norm": 2.2459235191345215,
+      "learning_rate": 3.88518866722711e-05,
+      "loss": 0.4182,
+      "step": 1300
+    },
+    {
+      "epoch": 1.911094783247612,
+      "grad_norm": 2.0538039207458496,
+      "learning_rate": 3.883051278709473e-05,
+      "loss": 0.3665,
+      "step": 1301
+    },
+    {
+      "epoch": 1.9125642909625276,
+      "grad_norm": 2.7344133853912354,
+      "learning_rate": 3.880912432401265e-05,
+      "loss": 0.4823,
+      "step": 1302
+    },
+    {
+      "epoch": 1.9140337986774432,
+      "grad_norm": 2.383896827697754,
+      "learning_rate": 3.878772130556915e-05,
+      "loss": 0.3915,
+      "step": 1303
+    },
+    {
+      "epoch": 1.9155033063923586,
+      "grad_norm": 2.1083223819732666,
+      "learning_rate": 3.876630375432384e-05,
+      "loss": 0.3837,
+      "step": 1304
+    },
+    {
+      "epoch": 1.916972814107274,
+      "grad_norm": 1.769742727279663,
+      "learning_rate": 3.874487169285168e-05,
+      "loss": 0.37,
+      "step": 1305
+    },
+    {
+      "epoch": 1.9184423218221895,
+      "grad_norm": 1.6477030515670776,
+      "learning_rate": 3.872342514374291e-05,
+      "loss": 0.354,
+      "step": 1306
+    },
+    {
+      "epoch": 1.9199118295371052,
+      "grad_norm": 2.375729560852051,
+      "learning_rate": 3.8701964129603025e-05,
+      "loss": 0.503,
+      "step": 1307
+    },
+    {
+      "epoch": 1.9213813372520205,
+      "grad_norm": 2.493929147720337,
+      "learning_rate": 3.8680488673052784e-05,
+      "loss": 0.4054,
+      "step": 1308
+    },
+    {
+      "epoch": 1.922850844966936,
+      "grad_norm": 1.9855644702911377,
+      "learning_rate": 3.865899879672816e-05,
+      "loss": 0.3285,
+      "step": 1309
+    },
+    {
+      "epoch": 1.9243203526818515,
+      "grad_norm": 1.6614248752593994,
+      "learning_rate": 3.863749452328035e-05,
+      "loss": 0.3092,
+      "step": 1310
+    },
+    {
+      "epoch": 1.9257898603967671,
+      "grad_norm": 2.6410045623779297,
+      "learning_rate": 3.861597587537568e-05,
+      "loss": 0.5674,
+      "step": 1311
+    },
+    {
+      "epoch": 1.9272593681116827,
+      "grad_norm": 2.8698296546936035,
+      "learning_rate": 3.859444287569567e-05,
+      "loss": 0.4176,
+      "step": 1312
+    },
+    {
+      "epoch": 1.9287288758265981,
+      "grad_norm": 1.8873811960220337,
+      "learning_rate": 3.8572895546936944e-05,
+      "loss": 0.3602,
+      "step": 1313
+    },
+    {
+      "epoch": 1.9301983835415135,
+      "grad_norm": 1.8994073867797852,
+      "learning_rate": 3.855133391181124e-05,
+      "loss": 0.4213,
+      "step": 1314
+    },
+    {
+      "epoch": 1.9316678912564291,
+      "grad_norm": 1.6846975088119507,
+      "learning_rate": 3.852975799304536e-05,
+      "loss": 0.3403,
+      "step": 1315
+    },
+    {
+      "epoch": 1.9331373989713447,
+      "grad_norm": 1.9634159803390503,
+      "learning_rate": 3.850816781338119e-05,
+      "loss": 0.3362,
+      "step": 1316
+    },
+    {
+      "epoch": 1.9346069066862601,
+      "grad_norm": 1.6626337766647339,
+      "learning_rate": 3.8486563395575625e-05,
+      "loss": 0.2846,
+      "step": 1317
+    },
+    {
+      "epoch": 1.9360764144011755,
+      "grad_norm": 1.66458261013031,
+      "learning_rate": 3.8464944762400565e-05,
+      "loss": 0.3063,
+      "step": 1318
+    },
+    {
+      "epoch": 1.9375459221160911,
+      "grad_norm": 1.3462938070297241,
+      "learning_rate": 3.844331193664291e-05,
+      "loss": 0.2753,
+      "step": 1319
+    },
+    {
+      "epoch": 1.9390154298310067,
+      "grad_norm": 1.5900206565856934,
+      "learning_rate": 3.842166494110451e-05,
+      "loss": 0.287,
+      "step": 1320
+    },
+    {
+      "epoch": 1.940484937545922,
+      "grad_norm": 1.6456480026245117,
+      "learning_rate": 3.840000379860214e-05,
+      "loss": 0.3431,
+      "step": 1321
+    },
+    {
+      "epoch": 1.9419544452608375,
+      "grad_norm": 1.6924060583114624,
+      "learning_rate": 3.837832853196751e-05,
+      "loss": 0.3163,
+      "step": 1322
+    },
+    {
+      "epoch": 1.943423952975753,
+      "grad_norm": 1.6371219158172607,
+      "learning_rate": 3.835663916404721e-05,
+      "loss": 0.2795,
+      "step": 1323
+    },
+    {
+      "epoch": 1.9448934606906687,
+      "grad_norm": 2.0110068321228027,
+      "learning_rate": 3.833493571770268e-05,
+      "loss": 0.3043,
+      "step": 1324
+    },
+    {
+      "epoch": 1.946362968405584,
+      "grad_norm": 2.4278488159179688,
+      "learning_rate": 3.831321821581021e-05,
+      "loss": 0.4693,
+      "step": 1325
+    },
+    {
+      "epoch": 1.9478324761204995,
+      "grad_norm": 1.6662366390228271,
+      "learning_rate": 3.82914866812609e-05,
+      "loss": 0.3702,
+      "step": 1326
+    },
+    {
+      "epoch": 1.949301983835415,
+      "grad_norm": 1.8097858428955078,
+      "learning_rate": 3.826974113696065e-05,
+      "loss": 0.3711,
+      "step": 1327
+    },
+    {
+      "epoch": 1.9507714915503307,
+      "grad_norm": 1.7795965671539307,
+      "learning_rate": 3.824798160583012e-05,
+      "loss": 0.3499,
+      "step": 1328
+    },
+    {
+      "epoch": 1.9522409992652463,
+      "grad_norm": 1.9042242765426636,
+      "learning_rate": 3.8226208110804715e-05,
+      "loss": 0.3937,
+      "step": 1329
+    },
+    {
+      "epoch": 1.9537105069801617,
+      "grad_norm": 2.387477159500122,
+      "learning_rate": 3.820442067483455e-05,
+      "loss": 0.3811,
+      "step": 1330
+    },
+    {
+      "epoch": 1.955180014695077,
+      "grad_norm": 2.42753529548645,
+      "learning_rate": 3.818261932088445e-05,
+      "loss": 0.5244,
+      "step": 1331
+    },
+    {
+      "epoch": 1.9566495224099927,
+      "grad_norm": 2.490438461303711,
+      "learning_rate": 3.81608040719339e-05,
+      "loss": 0.4055,
+      "step": 1332
+    },
+    {
+      "epoch": 1.9581190301249083,
+      "grad_norm": 1.426076054573059,
+      "learning_rate": 3.813897495097702e-05,
+      "loss": 0.2711,
+      "step": 1333
+    },
+    {
+      "epoch": 1.9595885378398237,
+      "grad_norm": 2.255260467529297,
+      "learning_rate": 3.811713198102258e-05,
+      "loss": 0.3992,
+      "step": 1334
+    },
+    {
+      "epoch": 1.961058045554739,
+      "grad_norm": 1.7828693389892578,
+      "learning_rate": 3.8095275185093923e-05,
+      "loss": 0.3478,
+      "step": 1335
+    },
+    {
+      "epoch": 1.9625275532696547,
+      "grad_norm": 2.4091989994049072,
+      "learning_rate": 3.807340458622898e-05,
+      "loss": 0.4888,
+      "step": 1336
+    },
+    {
+      "epoch": 1.9639970609845703,
+      "grad_norm": 1.830181360244751,
+      "learning_rate": 3.805152020748021e-05,
+      "loss": 0.4193,
+      "step": 1337
+    },
+    {
+      "epoch": 1.9654665686994857,
+      "grad_norm": 2.222064256668091,
+      "learning_rate": 3.802962207191463e-05,
+      "loss": 0.4584,
+      "step": 1338
+    },
+    {
+      "epoch": 1.966936076414401,
+      "grad_norm": 2.126588821411133,
+      "learning_rate": 3.800771020261373e-05,
+      "loss": 0.374,
+      "step": 1339
+    },
+    {
+      "epoch": 1.9684055841293167,
+      "grad_norm": 1.7062770128250122,
+      "learning_rate": 3.798578462267347e-05,
+      "loss": 0.3289,
+      "step": 1340
+    },
+    {
+      "epoch": 1.9698750918442323,
+      "grad_norm": 1.9276719093322754,
+      "learning_rate": 3.7963845355204304e-05,
+      "loss": 0.3197,
+      "step": 1341
+    },
+    {
+      "epoch": 1.9713445995591476,
+      "grad_norm": 2.0036489963531494,
+      "learning_rate": 3.794189242333106e-05,
+      "loss": 0.4164,
+      "step": 1342
+    },
+    {
+      "epoch": 1.972814107274063,
+      "grad_norm": 2.4382541179656982,
+      "learning_rate": 3.791992585019301e-05,
+      "loss": 0.4221,
+      "step": 1343
+    },
+    {
+      "epoch": 1.9742836149889786,
+      "grad_norm": 1.8204247951507568,
+      "learning_rate": 3.789794565894378e-05,
+      "loss": 0.3495,
+      "step": 1344
+    },
+    {
+      "epoch": 1.9757531227038942,
+      "grad_norm": 2.125206470489502,
+      "learning_rate": 3.787595187275136e-05,
+      "loss": 0.3531,
+      "step": 1345
+    },
+    {
+      "epoch": 1.9772226304188099,
+      "grad_norm": 2.210862874984741,
+      "learning_rate": 3.7853944514798064e-05,
+      "loss": 0.5755,
+      "step": 1346
+    },
+    {
+      "epoch": 1.9786921381337252,
+      "grad_norm": 2.612302541732788,
+      "learning_rate": 3.783192360828052e-05,
+      "loss": 0.5063,
+      "step": 1347
+    },
+    {
+      "epoch": 1.9801616458486406,
+      "grad_norm": 3.0008304119110107,
+      "learning_rate": 3.780988917640963e-05,
+      "loss": 0.5169,
+      "step": 1348
+    },
+    {
+      "epoch": 1.9816311535635562,
+      "grad_norm": 2.036790132522583,
+      "learning_rate": 3.7787841242410546e-05,
+      "loss": 0.3697,
+      "step": 1349
+    },
+    {
+      "epoch": 1.9831006612784718,
+      "grad_norm": 2.5431113243103027,
+      "learning_rate": 3.7765779829522675e-05,
+      "loss": 0.554,
+      "step": 1350
+    },
+    {
+      "epoch": 1.9845701689933872,
+      "grad_norm": 2.9467356204986572,
+      "learning_rate": 3.774370496099961e-05,
+      "loss": 0.5331,
+      "step": 1351
+    },
+    {
+      "epoch": 1.9860396767083026,
+      "grad_norm": 1.8796031475067139,
+      "learning_rate": 3.7721616660109125e-05,
+      "loss": 0.4045,
+      "step": 1352
+    },
+    {
+      "epoch": 1.9875091844232182,
+      "grad_norm": 1.6509068012237549,
+      "learning_rate": 3.7699514950133166e-05,
+      "loss": 0.3365,
+      "step": 1353
+    },
+    {
+      "epoch": 1.9889786921381338,
+      "grad_norm": 1.8875857591629028,
+      "learning_rate": 3.767739985436781e-05,
+      "loss": 0.3887,
+      "step": 1354
+    },
+    {
+      "epoch": 1.9904481998530492,
+      "grad_norm": 2.4470999240875244,
+      "learning_rate": 3.765527139612323e-05,
+      "loss": 0.5218,
+      "step": 1355
+    },
+    {
+      "epoch": 1.9919177075679646,
+      "grad_norm": 2.116738796234131,
+      "learning_rate": 3.7633129598723704e-05,
+      "loss": 0.3825,
+      "step": 1356
+    },
+    {
+      "epoch": 1.9933872152828802,
+      "grad_norm": 1.7693043947219849,
+      "learning_rate": 3.7610974485507554e-05,
+      "loss": 0.3976,
+      "step": 1357
+    },
+    {
+      "epoch": 1.9948567229977958,
+      "grad_norm": 1.5983883142471313,
+      "learning_rate": 3.758880607982714e-05,
+      "loss": 0.3048,
+      "step": 1358
+    },
+    {
+      "epoch": 1.9963262307127112,
+      "grad_norm": 2.0453860759735107,
+      "learning_rate": 3.7566624405048844e-05,
+      "loss": 0.4425,
+      "step": 1359
+    },
+    {
+      "epoch": 1.9977957384276266,
+      "grad_norm": 1.7836405038833618,
+      "learning_rate": 3.754442948455302e-05,
+      "loss": 0.3618,
+      "step": 1360
+    },
+    {
+      "epoch": 1.9992652461425422,
+      "grad_norm": 1.9100253582000732,
+      "learning_rate": 3.752222134173399e-05,
+      "loss": 0.3335,
+      "step": 1361
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 2.937403678894043,
+      "learning_rate": 3.7500000000000003e-05,
+      "loss": 0.3931,
+      "step": 1362
+    },
+    {
+      "epoch": 2.0,
+      "eval_loss": 0.5826267600059509,
+      "eval_runtime": 21.086,
+      "eval_samples_per_second": 6.829,
+      "eval_steps_per_second": 6.829,
+      "step": 1362
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 3400,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 5,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 2.220819772607693e+16,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}