diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,14073 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 400,
+  "global_step": 2000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0005,
+      "grad_norm": 1.9212751388549805,
+      "learning_rate": 2e-05,
+      "loss": 2.215,
+      "step": 1
+    },
+    {
+      "epoch": 0.001,
+      "grad_norm": 2.2275137901306152,
+      "learning_rate": 4e-05,
+      "loss": 2.0641,
+      "step": 2
+    },
+    {
+      "epoch": 0.0015,
+      "grad_norm": 2.32753586769104,
+      "learning_rate": 6e-05,
+      "loss": 2.2095,
+      "step": 3
+    },
+    {
+      "epoch": 0.002,
+      "grad_norm": 1.946838140487671,
+      "learning_rate": 8e-05,
+      "loss": 2.1757,
+      "step": 4
+    },
+    {
+      "epoch": 0.0025,
+      "grad_norm": 1.39604651927948,
+      "learning_rate": 0.0001,
+      "loss": 2.0656,
+      "step": 5
+    },
+    {
+      "epoch": 0.003,
+      "grad_norm": 1.571901798248291,
+      "learning_rate": 0.00012,
+      "loss": 2.0274,
+      "step": 6
+    },
+    {
+      "epoch": 0.0035,
+      "grad_norm": 1.2693225145339966,
+      "learning_rate": 0.00014,
+      "loss": 1.865,
+      "step": 7
+    },
+    {
+      "epoch": 0.004,
+      "grad_norm": 1.1292457580566406,
+      "learning_rate": 0.00016,
+      "loss": 1.5891,
+      "step": 8
+    },
+    {
+      "epoch": 0.0045,
+      "grad_norm": 1.3623621463775635,
+      "learning_rate": 0.00018,
+      "loss": 1.5916,
+      "step": 9
+    },
+    {
+      "epoch": 0.005,
+      "grad_norm": 1.3371833562850952,
+      "learning_rate": 0.0002,
+      "loss": 1.4759,
+      "step": 10
+    },
+    {
+      "epoch": 0.0055,
+      "grad_norm": 1.2279239892959595,
+      "learning_rate": 0.0001998994974874372,
+      "loss": 1.3904,
+      "step": 11
+    },
+    {
+      "epoch": 0.006,
+      "grad_norm": 1.3857628107070923,
+      "learning_rate": 0.00019979899497487438,
+      "loss": 1.151,
+      "step": 12
+    },
+    {
+      "epoch": 0.0065,
+      "grad_norm": 1.2530171871185303,
+      "learning_rate": 0.00019969849246231157,
+      "loss": 0.9745,
+      "step": 13
+    },
+    {
+      "epoch": 0.007,
+      "grad_norm": 1.349227786064148,
+      "learning_rate": 0.00019959798994974876,
+      "loss": 1.0997,
+      "step": 14
+    },
+    {
+      "epoch": 0.0075,
+      "grad_norm": 1.2657450437545776,
+      "learning_rate": 0.00019949748743718594,
+      "loss": 1.2381,
+      "step": 15
+    },
+    {
+      "epoch": 0.008,
+      "grad_norm": 1.084928274154663,
+      "learning_rate": 0.00019939698492462313,
+      "loss": 0.786,
+      "step": 16
+    },
+    {
+      "epoch": 0.0085,
+      "grad_norm": 1.0906790494918823,
+      "learning_rate": 0.00019929648241206031,
+      "loss": 0.8476,
+      "step": 17
+    },
+    {
+      "epoch": 0.009,
+      "grad_norm": 1.0389244556427002,
+      "learning_rate": 0.0001991959798994975,
+      "loss": 0.9624,
+      "step": 18
+    },
+    {
+      "epoch": 0.0095,
+      "grad_norm": 0.9856970310211182,
+      "learning_rate": 0.0001990954773869347,
+      "loss": 0.802,
+      "step": 19
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.033990740776062,
+      "learning_rate": 0.00019899497487437187,
+      "loss": 0.9143,
+      "step": 20
+    },
+    {
+      "epoch": 0.0105,
+      "grad_norm": 1.1213363409042358,
+      "learning_rate": 0.00019889447236180906,
+      "loss": 0.9041,
+      "step": 21
+    },
+    {
+      "epoch": 0.011,
+      "grad_norm": 0.9037308096885681,
+      "learning_rate": 0.00019879396984924622,
+      "loss": 0.786,
+      "step": 22
+    },
+    {
+      "epoch": 0.0115,
+      "grad_norm": 0.9636969566345215,
+      "learning_rate": 0.00019869346733668343,
+      "loss": 0.8478,
+      "step": 23
+    },
+    {
+      "epoch": 0.012,
+      "grad_norm": 0.9868170022964478,
+      "learning_rate": 0.00019859296482412062,
+      "loss": 0.8824,
+      "step": 24
+    },
+    {
+      "epoch": 0.0125,
+      "grad_norm": 0.9793375730514526,
+      "learning_rate": 0.0001984924623115578,
+      "loss": 0.8472,
+      "step": 25
+    },
+    {
+      "epoch": 0.013,
+      "grad_norm": 0.9688933491706848,
+      "learning_rate": 0.000198391959798995,
+      "loss": 0.7469,
+      "step": 26
+    },
+    {
+      "epoch": 0.0135,
+      "grad_norm": 0.8956223130226135,
+      "learning_rate": 0.00019829145728643218,
+      "loss": 0.7871,
+      "step": 27
+    },
+    {
+      "epoch": 0.014,
+      "grad_norm": 1.047118067741394,
+      "learning_rate": 0.00019819095477386937,
+      "loss": 0.7728,
+      "step": 28
+    },
+    {
+      "epoch": 0.0145,
+      "grad_norm": 1.0644763708114624,
+      "learning_rate": 0.00019809045226130653,
+      "loss": 0.8429,
+      "step": 29
+    },
+    {
+      "epoch": 0.015,
+      "grad_norm": 1.1773219108581543,
+      "learning_rate": 0.0001979899497487437,
+      "loss": 0.9289,
+      "step": 30
+    },
+    {
+      "epoch": 0.0155,
+      "grad_norm": 0.96980881690979,
+      "learning_rate": 0.00019788944723618093,
+      "loss": 0.7551,
+      "step": 31
+    },
+    {
+      "epoch": 0.016,
+      "grad_norm": 0.9512197375297546,
+      "learning_rate": 0.0001977889447236181,
+      "loss": 0.8559,
+      "step": 32
+    },
+    {
+      "epoch": 0.0165,
+      "grad_norm": 0.8439690470695496,
+      "learning_rate": 0.0001976884422110553,
+      "loss": 0.6484,
+      "step": 33
+    },
+    {
+      "epoch": 0.017,
+      "grad_norm": 0.8353943824768066,
+      "learning_rate": 0.00019758793969849249,
+      "loss": 0.6201,
+      "step": 34
+    },
+    {
+      "epoch": 0.0175,
+      "grad_norm": 0.920733630657196,
+      "learning_rate": 0.00019748743718592964,
+      "loss": 0.6189,
+      "step": 35
+    },
+    {
+      "epoch": 0.018,
+      "grad_norm": 0.8841115236282349,
+      "learning_rate": 0.00019738693467336683,
+      "loss": 0.825,
+      "step": 36
+    },
+    {
+      "epoch": 0.0185,
+      "grad_norm": 0.8891990780830383,
+      "learning_rate": 0.00019728643216080402,
+      "loss": 0.6734,
+      "step": 37
+    },
+    {
+      "epoch": 0.019,
+      "grad_norm": 0.9868722558021545,
+      "learning_rate": 0.0001971859296482412,
+      "loss": 0.7867,
+      "step": 38
+    },
+    {
+      "epoch": 0.0195,
+      "grad_norm": 1.098929524421692,
+      "learning_rate": 0.0001970854271356784,
+      "loss": 0.738,
+      "step": 39
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9769437909126282,
+      "learning_rate": 0.0001969849246231156,
+      "loss": 0.6644,
+      "step": 40
+    },
+    {
+      "epoch": 0.0205,
+      "grad_norm": 0.9597660899162292,
+      "learning_rate": 0.0001968844221105528,
+      "loss": 0.7086,
+      "step": 41
+    },
+    {
+      "epoch": 0.021,
+      "grad_norm": 0.9977144002914429,
+      "learning_rate": 0.00019678391959798995,
+      "loss": 0.7227,
+      "step": 42
+    },
+    {
+      "epoch": 0.0215,
+      "grad_norm": 1.0582157373428345,
+      "learning_rate": 0.00019668341708542714,
+      "loss": 0.8746,
+      "step": 43
+    },
+    {
+      "epoch": 0.022,
+      "grad_norm": 1.2232478857040405,
+      "learning_rate": 0.00019658291457286432,
+      "loss": 0.7717,
+      "step": 44
+    },
+    {
+      "epoch": 0.0225,
+      "grad_norm": 1.1404426097869873,
+      "learning_rate": 0.0001964824120603015,
+      "loss": 0.6966,
+      "step": 45
+    },
+    {
+      "epoch": 0.023,
+      "grad_norm": 0.8900367617607117,
+      "learning_rate": 0.0001963819095477387,
+      "loss": 0.5641,
+      "step": 46
+    },
+    {
+      "epoch": 0.0235,
+      "grad_norm": 0.9628854393959045,
+      "learning_rate": 0.00019628140703517588,
+      "loss": 0.6375,
+      "step": 47
+    },
+    {
+      "epoch": 0.024,
+      "grad_norm": 1.04586923122406,
+      "learning_rate": 0.0001961809045226131,
+      "loss": 0.559,
+      "step": 48
+    },
+    {
+      "epoch": 0.0245,
+      "grad_norm": 1.414943814277649,
+      "learning_rate": 0.00019608040201005026,
+      "loss": 0.6797,
+      "step": 49
+    },
+    {
+      "epoch": 0.025,
+      "grad_norm": 1.2836331129074097,
+      "learning_rate": 0.00019597989949748744,
+      "loss": 0.5336,
+      "step": 50
+    },
+    {
+      "epoch": 0.0255,
+      "grad_norm": 0.9944679737091064,
+      "learning_rate": 0.00019587939698492463,
+      "loss": 1.0081,
+      "step": 51
+    },
+    {
+      "epoch": 0.026,
+      "grad_norm": 0.9618650078773499,
+      "learning_rate": 0.00019577889447236181,
+      "loss": 0.9708,
+      "step": 52
+    },
+    {
+      "epoch": 0.0265,
+      "grad_norm": 0.9560412764549255,
+      "learning_rate": 0.000195678391959799,
+      "loss": 1.0262,
+      "step": 53
+    },
+    {
+      "epoch": 0.027,
+      "grad_norm": 0.8408827781677246,
+      "learning_rate": 0.0001955778894472362,
+      "loss": 0.7476,
+      "step": 54
+    },
+    {
+      "epoch": 0.0275,
+      "grad_norm": 0.852470338344574,
+      "learning_rate": 0.00019547738693467337,
+      "loss": 0.7852,
+      "step": 55
+    },
+    {
+      "epoch": 0.028,
+      "grad_norm": 0.9377810955047607,
+      "learning_rate": 0.00019537688442211056,
+      "loss": 0.862,
+      "step": 56
+    },
+    {
+      "epoch": 0.0285,
+      "grad_norm": 0.8051367402076721,
+      "learning_rate": 0.00019527638190954775,
+      "loss": 0.7589,
+      "step": 57
+    },
+    {
+      "epoch": 0.029,
+      "grad_norm": 0.9737776517868042,
+      "learning_rate": 0.00019517587939698493,
+      "loss": 0.8373,
+      "step": 58
+    },
+    {
+      "epoch": 0.0295,
+      "grad_norm": 0.9199332594871521,
+      "learning_rate": 0.00019507537688442212,
+      "loss": 0.6826,
+      "step": 59
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0166407823562622,
+      "learning_rate": 0.0001949748743718593,
+      "loss": 0.6999,
+      "step": 60
+    },
+    {
+      "epoch": 0.0305,
+      "grad_norm": 1.0075502395629883,
+      "learning_rate": 0.0001948743718592965,
+      "loss": 0.842,
+      "step": 61
+    },
+    {
+      "epoch": 0.031,
+      "grad_norm": 0.8241415619850159,
+      "learning_rate": 0.00019477386934673368,
+      "loss": 0.6479,
+      "step": 62
+    },
+    {
+      "epoch": 0.0315,
+      "grad_norm": 0.7866672873497009,
+      "learning_rate": 0.00019467336683417087,
+      "loss": 0.6068,
+      "step": 63
+    },
+    {
+      "epoch": 0.032,
+      "grad_norm": 0.7412344813346863,
+      "learning_rate": 0.00019457286432160805,
+      "loss": 0.5115,
+      "step": 64
+    },
+    {
+      "epoch": 0.0325,
+      "grad_norm": 0.9863675236701965,
+      "learning_rate": 0.00019447236180904524,
+      "loss": 0.7101,
+      "step": 65
+    },
+    {
+      "epoch": 0.033,
+      "grad_norm": 0.8766986727714539,
+      "learning_rate": 0.00019437185929648243,
+      "loss": 0.5552,
+      "step": 66
+    },
+    {
+      "epoch": 0.0335,
+      "grad_norm": 0.8189011812210083,
+      "learning_rate": 0.0001942713567839196,
+      "loss": 0.6418,
+      "step": 67
+    },
+    {
+      "epoch": 0.034,
+      "grad_norm": 0.9847667813301086,
+      "learning_rate": 0.0001941708542713568,
+      "loss": 0.6014,
+      "step": 68
+    },
+    {
+      "epoch": 0.0345,
+      "grad_norm": 1.0095819234848022,
+      "learning_rate": 0.00019407035175879398,
+      "loss": 0.8432,
+      "step": 69
+    },
+    {
+      "epoch": 0.035,
+      "grad_norm": 0.8844761848449707,
+      "learning_rate": 0.00019396984924623117,
+      "loss": 0.5534,
+      "step": 70
+    },
+    {
+      "epoch": 0.0355,
+      "grad_norm": 0.9700236916542053,
+      "learning_rate": 0.00019386934673366836,
+      "loss": 0.6588,
+      "step": 71
+    },
+    {
+      "epoch": 0.036,
+      "grad_norm": 0.9961576461791992,
+      "learning_rate": 0.00019376884422110552,
+      "loss": 0.6668,
+      "step": 72
+    },
+    {
+      "epoch": 0.0365,
+      "grad_norm": 1.026727318763733,
+      "learning_rate": 0.00019366834170854273,
+      "loss": 0.6486,
+      "step": 73
+    },
+    {
+      "epoch": 0.037,
+      "grad_norm": 0.9477052688598633,
+      "learning_rate": 0.00019356783919597992,
+      "loss": 0.5957,
+      "step": 74
+    },
+    {
+      "epoch": 0.0375,
+      "grad_norm": 0.8860141634941101,
+      "learning_rate": 0.0001934673366834171,
+      "loss": 0.6697,
+      "step": 75
+    },
+    {
+      "epoch": 0.038,
+      "grad_norm": 0.7626288533210754,
+      "learning_rate": 0.0001933668341708543,
+      "loss": 0.4807,
+      "step": 76
+    },
+    {
+      "epoch": 0.0385,
+      "grad_norm": 0.8229109048843384,
+      "learning_rate": 0.00019326633165829148,
+      "loss": 0.5105,
+      "step": 77
+    },
+    {
+      "epoch": 0.039,
+      "grad_norm": 0.8756014108657837,
+      "learning_rate": 0.00019316582914572864,
+      "loss": 0.5917,
+      "step": 78
+    },
+    {
+      "epoch": 0.0395,
+      "grad_norm": 0.8144027590751648,
+      "learning_rate": 0.00019306532663316582,
+      "loss": 0.5343,
+      "step": 79
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8877262473106384,
+      "learning_rate": 0.000192964824120603,
+      "loss": 0.5681,
+      "step": 80
+    },
+    {
+      "epoch": 0.0405,
+      "grad_norm": 0.8442334532737732,
+      "learning_rate": 0.0001928643216080402,
+      "loss": 0.616,
+      "step": 81
+    },
+    {
+      "epoch": 0.041,
+      "grad_norm": 1.0273793935775757,
+      "learning_rate": 0.0001927638190954774,
+      "loss": 0.7154,
+      "step": 82
+    },
+    {
+      "epoch": 0.0415,
+      "grad_norm": 1.0146235227584839,
+      "learning_rate": 0.0001926633165829146,
+      "loss": 0.7314,
+      "step": 83
+    },
+    {
+      "epoch": 0.042,
+      "grad_norm": 1.0425784587860107,
+      "learning_rate": 0.00019256281407035178,
+      "loss": 0.5275,
+      "step": 84
+    },
+    {
+      "epoch": 0.0425,
+      "grad_norm": 0.8293735384941101,
+      "learning_rate": 0.00019246231155778894,
+      "loss": 0.5404,
+      "step": 85
+    },
+    {
+      "epoch": 0.043,
+      "grad_norm": 0.8221484422683716,
+      "learning_rate": 0.00019236180904522613,
+      "loss": 0.4605,
+      "step": 86
+    },
+    {
+      "epoch": 0.0435,
+      "grad_norm": 0.8663842082023621,
+      "learning_rate": 0.00019226130653266331,
+      "loss": 0.5512,
+      "step": 87
+    },
+    {
+      "epoch": 0.044,
+      "grad_norm": 1.364332914352417,
+      "learning_rate": 0.0001921608040201005,
+      "loss": 0.7857,
+      "step": 88
+    },
+    {
+      "epoch": 0.0445,
+      "grad_norm": 0.8523952960968018,
+      "learning_rate": 0.0001920603015075377,
+      "loss": 0.5245,
+      "step": 89
+    },
+    {
+      "epoch": 0.045,
+      "grad_norm": 0.9850642085075378,
+      "learning_rate": 0.0001919597989949749,
+      "loss": 0.6477,
+      "step": 90
+    },
+    {
+      "epoch": 0.0455,
+      "grad_norm": 0.8323433995246887,
+      "learning_rate": 0.0001918592964824121,
+      "loss": 0.5972,
+      "step": 91
+    },
+    {
+      "epoch": 0.046,
+      "grad_norm": 0.9467252492904663,
+      "learning_rate": 0.00019175879396984925,
+      "loss": 0.5334,
+      "step": 92
+    },
+    {
+      "epoch": 0.0465,
+      "grad_norm": 0.8340337872505188,
+      "learning_rate": 0.00019165829145728643,
+      "loss": 0.5002,
+      "step": 93
+    },
+    {
+      "epoch": 0.047,
+      "grad_norm": 0.8522113561630249,
+      "learning_rate": 0.00019155778894472362,
+      "loss": 0.4849,
+      "step": 94
+    },
+    {
+      "epoch": 0.0475,
+      "grad_norm": 0.9448826909065247,
+      "learning_rate": 0.0001914572864321608,
+      "loss": 0.5973,
+      "step": 95
+    },
+    {
+      "epoch": 0.048,
+      "grad_norm": 0.8943369388580322,
+      "learning_rate": 0.000191356783919598,
+      "loss": 0.5314,
+      "step": 96
+    },
+    {
+      "epoch": 0.0485,
+      "grad_norm": 1.0275241136550903,
+      "learning_rate": 0.00019125628140703518,
+      "loss": 0.6002,
+      "step": 97
+    },
+    {
+      "epoch": 0.049,
+      "grad_norm": 1.067549467086792,
+      "learning_rate": 0.0001911557788944724,
+      "loss": 0.7125,
+      "step": 98
+    },
+    {
+      "epoch": 0.0495,
+      "grad_norm": 0.7812333106994629,
+      "learning_rate": 0.00019105527638190955,
+      "loss": 0.4519,
+      "step": 99
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1348077058792114,
+      "learning_rate": 0.00019095477386934674,
+      "loss": 0.4832,
+      "step": 100
+    },
+    {
+      "epoch": 0.0505,
+      "grad_norm": 0.6455503106117249,
+      "learning_rate": 0.00019085427135678393,
+      "loss": 0.7981,
+      "step": 101
+    },
+    {
+      "epoch": 0.051,
+      "grad_norm": 0.65630704164505,
+      "learning_rate": 0.0001907537688442211,
+      "loss": 0.6972,
+      "step": 102
+    },
+    {
+      "epoch": 0.0515,
+      "grad_norm": 0.7021996974945068,
+      "learning_rate": 0.0001906532663316583,
+      "loss": 0.6986,
+      "step": 103
+    },
+    {
+      "epoch": 0.052,
+      "grad_norm": 0.8909450769424438,
+      "learning_rate": 0.00019055276381909548,
+      "loss": 0.7501,
+      "step": 104
+    },
+    {
+      "epoch": 0.0525,
+      "grad_norm": 0.8215047717094421,
+      "learning_rate": 0.00019045226130653267,
+      "loss": 0.7754,
+      "step": 105
+    },
+    {
+      "epoch": 0.053,
+      "grad_norm": 0.8323522806167603,
+      "learning_rate": 0.00019035175879396986,
+      "loss": 0.6716,
+      "step": 106
+    },
+    {
+      "epoch": 0.0535,
+      "grad_norm": 0.7359536290168762,
+      "learning_rate": 0.00019025125628140704,
+      "loss": 0.6585,
+      "step": 107
+    },
+    {
+      "epoch": 0.054,
+      "grad_norm": 0.801367461681366,
+      "learning_rate": 0.00019015075376884423,
+      "loss": 0.7739,
+      "step": 108
+    },
+    {
+      "epoch": 0.0545,
+      "grad_norm": 0.791890561580658,
+      "learning_rate": 0.00019005025125628142,
+      "loss": 0.7107,
+      "step": 109
+    },
+    {
+      "epoch": 0.055,
+      "grad_norm": 0.7103503942489624,
+      "learning_rate": 0.0001899497487437186,
+      "loss": 0.536,
+      "step": 110
+    },
+    {
+      "epoch": 0.0555,
+      "grad_norm": 0.645163893699646,
+      "learning_rate": 0.0001898492462311558,
+      "loss": 0.5613,
+      "step": 111
+    },
+    {
+      "epoch": 0.056,
+      "grad_norm": 0.8729243278503418,
+      "learning_rate": 0.00018974874371859298,
+      "loss": 0.6209,
+      "step": 112
+    },
+    {
+      "epoch": 0.0565,
+      "grad_norm": 0.834465503692627,
+      "learning_rate": 0.00018964824120603016,
+      "loss": 0.5785,
+      "step": 113
+    },
+    {
+      "epoch": 0.057,
+      "grad_norm": 0.7811848521232605,
+      "learning_rate": 0.00018954773869346732,
+      "loss": 0.5672,
+      "step": 114
+    },
+    {
+      "epoch": 0.0575,
+      "grad_norm": 0.8115241527557373,
+      "learning_rate": 0.00018944723618090454,
+      "loss": 0.5989,
+      "step": 115
+    },
+    {
+      "epoch": 0.058,
+      "grad_norm": 0.7859218716621399,
+      "learning_rate": 0.00018934673366834172,
+      "loss": 0.5023,
+      "step": 116
+    },
+    {
+      "epoch": 0.0585,
+      "grad_norm": 0.9395287036895752,
+      "learning_rate": 0.0001892462311557789,
+      "loss": 0.6109,
+      "step": 117
+    },
+    {
+      "epoch": 0.059,
+      "grad_norm": 0.9204758405685425,
+      "learning_rate": 0.0001891457286432161,
+      "loss": 0.6337,
+      "step": 118
+    },
+    {
+      "epoch": 0.0595,
+      "grad_norm": 0.9071129560470581,
+      "learning_rate": 0.00018904522613065328,
+      "loss": 0.5274,
+      "step": 119
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0141112804412842,
+      "learning_rate": 0.00018894472361809047,
+      "loss": 0.7084,
+      "step": 120
+    },
+    {
+      "epoch": 0.0605,
+      "grad_norm": 0.8796365261077881,
+      "learning_rate": 0.00018884422110552763,
+      "loss": 0.5223,
+      "step": 121
+    },
+    {
+      "epoch": 0.061,
+      "grad_norm": 0.805503785610199,
+      "learning_rate": 0.00018874371859296481,
+      "loss": 0.5468,
+      "step": 122
+    },
+    {
+      "epoch": 0.0615,
+      "grad_norm": 0.8110489249229431,
+      "learning_rate": 0.00018864321608040203,
+      "loss": 0.5013,
+      "step": 123
+    },
+    {
+      "epoch": 0.062,
+      "grad_norm": 1.105877161026001,
+      "learning_rate": 0.00018854271356783921,
+      "loss": 0.7481,
+      "step": 124
+    },
+    {
+      "epoch": 0.0625,
+      "grad_norm": 0.8206800818443298,
+      "learning_rate": 0.0001884422110552764,
+      "loss": 0.5693,
+      "step": 125
+    },
+    {
+      "epoch": 0.063,
+      "grad_norm": 0.8398736119270325,
+      "learning_rate": 0.0001883417085427136,
+      "loss": 0.5275,
+      "step": 126
+    },
+    {
+      "epoch": 0.0635,
+      "grad_norm": 1.0161406993865967,
+      "learning_rate": 0.00018824120603015077,
+      "loss": 0.5812,
+      "step": 127
+    },
+    {
+      "epoch": 0.064,
+      "grad_norm": 0.8265745043754578,
+      "learning_rate": 0.00018814070351758793,
+      "loss": 0.7212,
+      "step": 128
+    },
+    {
+      "epoch": 0.0645,
+      "grad_norm": 0.7982003092765808,
+      "learning_rate": 0.00018804020100502512,
+      "loss": 0.4451,
+      "step": 129
+    },
+    {
+      "epoch": 0.065,
+      "grad_norm": 0.8722413182258606,
+      "learning_rate": 0.0001879396984924623,
+      "loss": 0.5745,
+      "step": 130
+    },
+    {
+      "epoch": 0.0655,
+      "grad_norm": 0.8562734127044678,
+      "learning_rate": 0.0001878391959798995,
+      "loss": 0.5377,
+      "step": 131
+    },
+    {
+      "epoch": 0.066,
+      "grad_norm": 0.7483986020088196,
+      "learning_rate": 0.0001877386934673367,
+      "loss": 0.4236,
+      "step": 132
+    },
+    {
+      "epoch": 0.0665,
+      "grad_norm": 0.7342506647109985,
+      "learning_rate": 0.0001876381909547739,
+      "loss": 0.4599,
+      "step": 133
+    },
+    {
+      "epoch": 0.067,
+      "grad_norm": 1.0230991840362549,
+      "learning_rate": 0.00018753768844221108,
+      "loss": 0.6086,
+      "step": 134
+    },
+    {
+      "epoch": 0.0675,
+      "grad_norm": 0.7798928618431091,
+      "learning_rate": 0.00018743718592964824,
+      "loss": 0.5426,
+      "step": 135
+    },
+    {
+      "epoch": 0.068,
+      "grad_norm": 0.8807337880134583,
+      "learning_rate": 0.00018733668341708543,
+      "loss": 0.5807,
+      "step": 136
+    },
+    {
+      "epoch": 0.0685,
+      "grad_norm": 0.7064052820205688,
+      "learning_rate": 0.0001872361809045226,
+      "loss": 0.4591,
+      "step": 137
+    },
+    {
+      "epoch": 0.069,
+      "grad_norm": 0.8630251884460449,
+      "learning_rate": 0.0001871356783919598,
+      "loss": 0.6175,
+      "step": 138
+    },
+    {
+      "epoch": 0.0695,
+      "grad_norm": 0.8436501622200012,
+      "learning_rate": 0.00018703517587939698,
+      "loss": 0.5085,
+      "step": 139
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8576134443283081,
+      "learning_rate": 0.0001869346733668342,
+      "loss": 0.4768,
+      "step": 140
+    },
+    {
+      "epoch": 0.0705,
+      "grad_norm": 0.7777182459831238,
+      "learning_rate": 0.00018683417085427138,
+      "loss": 0.562,
+      "step": 141
+    },
+    {
+      "epoch": 0.071,
+      "grad_norm": 0.7800930738449097,
+      "learning_rate": 0.00018673366834170854,
+      "loss": 0.4415,
+      "step": 142
+    },
+    {
+      "epoch": 0.0715,
+      "grad_norm": 0.879585325717926,
+      "learning_rate": 0.00018663316582914573,
+      "loss": 0.5074,
+      "step": 143
+    },
+    {
+      "epoch": 0.072,
+      "grad_norm": 0.7874137163162231,
+      "learning_rate": 0.00018653266331658292,
+      "loss": 0.4745,
+      "step": 144
+    },
+    {
+      "epoch": 0.0725,
+      "grad_norm": 0.8690528869628906,
+      "learning_rate": 0.0001864321608040201,
+      "loss": 0.4785,
+      "step": 145
+    },
+    {
+      "epoch": 0.073,
+      "grad_norm": 0.9143624901771545,
+      "learning_rate": 0.0001863316582914573,
+      "loss": 0.5306,
+      "step": 146
+    },
+    {
+      "epoch": 0.0735,
+      "grad_norm": 0.8383007049560547,
+      "learning_rate": 0.00018623115577889448,
+      "loss": 0.4454,
+      "step": 147
+    },
+    {
+      "epoch": 0.074,
+      "grad_norm": 0.9677513837814331,
+      "learning_rate": 0.0001861306532663317,
+      "loss": 0.4453,
+      "step": 148
+    },
+    {
+      "epoch": 0.0745,
+      "grad_norm": 0.700311541557312,
+      "learning_rate": 0.00018603015075376885,
+      "loss": 0.327,
+      "step": 149
+    },
+    {
+      "epoch": 0.075,
+      "grad_norm": 0.9320207238197327,
+      "learning_rate": 0.00018592964824120604,
+      "loss": 0.4189,
+      "step": 150
+    },
+    {
+      "epoch": 0.0755,
+      "grad_norm": 0.7870379686355591,
+      "learning_rate": 0.00018582914572864322,
+      "loss": 0.8049,
+      "step": 151
+    },
+    {
+      "epoch": 0.076,
+      "grad_norm": 0.7801129817962646,
+      "learning_rate": 0.0001857286432160804,
+      "loss": 0.7435,
+      "step": 152
+    },
+    {
+      "epoch": 0.0765,
+      "grad_norm": 0.9339824914932251,
+      "learning_rate": 0.0001856281407035176,
+      "loss": 0.6658,
+      "step": 153
+    },
+    {
+      "epoch": 0.077,
+      "grad_norm": 0.7770904302597046,
+      "learning_rate": 0.00018552763819095478,
+      "loss": 0.7067,
+      "step": 154
+    },
+    {
+      "epoch": 0.0775,
+      "grad_norm": 0.7401093244552612,
+      "learning_rate": 0.00018542713567839197,
+      "loss": 0.8342,
+      "step": 155
+    },
+    {
+      "epoch": 0.078,
+      "grad_norm": 0.7223836779594421,
+      "learning_rate": 0.00018532663316582915,
+      "loss": 0.7242,
+      "step": 156
+    },
+    {
+      "epoch": 0.0785,
+      "grad_norm": 0.8037763237953186,
+      "learning_rate": 0.00018522613065326634,
+      "loss": 0.76,
+      "step": 157
+    },
+    {
+      "epoch": 0.079,
+      "grad_norm": 0.7356771230697632,
+      "learning_rate": 0.00018512562814070353,
+      "loss": 0.6327,
+      "step": 158
+    },
+    {
+      "epoch": 0.0795,
+      "grad_norm": 0.7682873010635376,
+      "learning_rate": 0.00018502512562814071,
+      "loss": 0.7354,
+      "step": 159
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8977128267288208,
+      "learning_rate": 0.0001849246231155779,
+      "loss": 0.803,
+      "step": 160
+    },
+    {
+      "epoch": 0.0805,
+      "grad_norm": 0.8244638442993164,
+      "learning_rate": 0.0001848241206030151,
+      "loss": 0.6348,
+      "step": 161
+    },
+    {
+      "epoch": 0.081,
+      "grad_norm": 0.7262585759162903,
+      "learning_rate": 0.00018472361809045227,
+      "loss": 0.6588,
+      "step": 162
+    },
+    {
+      "epoch": 0.0815,
+      "grad_norm": 0.6660133600234985,
+      "learning_rate": 0.00018462311557788946,
+      "loss": 0.4494,
+      "step": 163
+    },
+    {
+      "epoch": 0.082,
+      "grad_norm": 0.6688201427459717,
+      "learning_rate": 0.00018452261306532662,
+      "loss": 0.5766,
+      "step": 164
+    },
+    {
+      "epoch": 0.0825,
+      "grad_norm": 0.7148305177688599,
+      "learning_rate": 0.00018442211055276383,
+      "loss": 0.715,
+      "step": 165
+    },
+    {
+      "epoch": 0.083,
+      "grad_norm": 0.6823235154151917,
+      "learning_rate": 0.00018432160804020102,
+      "loss": 0.6551,
+      "step": 166
+    },
+    {
+      "epoch": 0.0835,
+      "grad_norm": 0.7435379028320312,
+      "learning_rate": 0.0001842211055276382,
+      "loss": 0.5591,
+      "step": 167
+    },
+    {
+      "epoch": 0.084,
+      "grad_norm": 0.7870805859565735,
+      "learning_rate": 0.0001841206030150754,
+      "loss": 0.7267,
+      "step": 168
+    },
+    {
+      "epoch": 0.0845,
+      "grad_norm": 0.826301634311676,
+      "learning_rate": 0.00018402010050251258,
+      "loss": 0.626,
+      "step": 169
+    },
+    {
+      "epoch": 0.085,
+      "grad_norm": 0.8056604862213135,
+      "learning_rate": 0.00018391959798994977,
+      "loss": 0.4988,
+      "step": 170
+    },
+    {
+      "epoch": 0.0855,
+      "grad_norm": 0.766531765460968,
+      "learning_rate": 0.00018381909547738692,
+      "loss": 0.6145,
+      "step": 171
+    },
+    {
+      "epoch": 0.086,
+      "grad_norm": 0.7824142575263977,
+      "learning_rate": 0.0001837185929648241,
+      "loss": 0.5293,
+      "step": 172
+    },
+    {
+      "epoch": 0.0865,
+      "grad_norm": 0.8327628970146179,
+      "learning_rate": 0.0001836180904522613,
+      "loss": 0.6871,
+      "step": 173
+    },
+    {
+      "epoch": 0.087,
+      "grad_norm": 0.7102224826812744,
+      "learning_rate": 0.0001835175879396985,
+      "loss": 0.5561,
+      "step": 174
+    },
+    {
+      "epoch": 0.0875,
+      "grad_norm": 0.8501052260398865,
+      "learning_rate": 0.0001834170854271357,
+      "loss": 0.6198,
+      "step": 175
+    },
+    {
+      "epoch": 0.088,
+      "grad_norm": 0.818057656288147,
+      "learning_rate": 0.00018331658291457288,
+      "loss": 0.5636,
+      "step": 176
+    },
+    {
+      "epoch": 0.0885,
+      "grad_norm": 0.818524956703186,
+      "learning_rate": 0.00018321608040201007,
+      "loss": 0.5692,
+      "step": 177
+    },
+    {
+      "epoch": 0.089,
+      "grad_norm": 0.7823440432548523,
+      "learning_rate": 0.00018311557788944723,
+      "loss": 0.5973,
+      "step": 178
+    },
+    {
+      "epoch": 0.0895,
+      "grad_norm": 0.814060628414154,
+      "learning_rate": 0.00018301507537688442,
+      "loss": 0.5802,
+      "step": 179
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7701743841171265,
+      "learning_rate": 0.0001829145728643216,
+      "loss": 0.5659,
+      "step": 180
+    },
+    {
+      "epoch": 0.0905,
+      "grad_norm": 0.7541936039924622,
+      "learning_rate": 0.0001828140703517588,
+      "loss": 0.553,
+      "step": 181
+    },
+    {
+      "epoch": 0.091,
+      "grad_norm": 0.8096820712089539,
+      "learning_rate": 0.000182713567839196,
+      "loss": 0.6466,
+      "step": 182
+    },
+    {
+      "epoch": 0.0915,
+      "grad_norm": 0.7970617413520813,
+      "learning_rate": 0.0001826130653266332,
+      "loss": 0.5595,
+      "step": 183
+    },
+    {
+      "epoch": 0.092,
+      "grad_norm": 0.7757616639137268,
+      "learning_rate": 0.00018251256281407038,
+      "loss": 0.6264,
+      "step": 184
+    },
+    {
+      "epoch": 0.0925,
+      "grad_norm": 0.8462669253349304,
+      "learning_rate": 0.00018241206030150754,
+      "loss": 0.561,
+      "step": 185
+    },
+    {
+      "epoch": 0.093,
+      "grad_norm": 0.7140375971794128,
+      "learning_rate": 0.00018231155778894472,
+      "loss": 0.5594,
+      "step": 186
+    },
+    {
+      "epoch": 0.0935,
+      "grad_norm": 0.7507144212722778,
+      "learning_rate": 0.0001822110552763819,
+      "loss": 0.5286,
+      "step": 187
+    },
+    {
+      "epoch": 0.094,
+      "grad_norm": 0.7307922840118408,
+      "learning_rate": 0.0001821105527638191,
+      "loss": 0.4066,
+      "step": 188
+    },
+    {
+      "epoch": 0.0945,
+      "grad_norm": 0.7994447946548462,
+      "learning_rate": 0.00018201005025125628,
+      "loss": 0.3905,
+      "step": 189
+    },
+    {
+      "epoch": 0.095,
+      "grad_norm": 0.8033040761947632,
+      "learning_rate": 0.0001819095477386935,
+      "loss": 0.5124,
+      "step": 190
+    },
+    {
+      "epoch": 0.0955,
+      "grad_norm": 0.7406694293022156,
+      "learning_rate": 0.00018180904522613068,
+      "loss": 0.4055,
+      "step": 191
+    },
+    {
+      "epoch": 0.096,
+      "grad_norm": 0.7335091829299927,
+      "learning_rate": 0.00018170854271356784,
+      "loss": 0.4383,
+      "step": 192
+    },
+    {
+      "epoch": 0.0965,
+      "grad_norm": 0.7533824443817139,
+      "learning_rate": 0.00018160804020100503,
+      "loss": 0.3858,
+      "step": 193
+    },
+    {
+      "epoch": 0.097,
+      "grad_norm": 0.7952653765678406,
+      "learning_rate": 0.00018150753768844221,
+      "loss": 0.4816,
+      "step": 194
+    },
+    {
+      "epoch": 0.0975,
+      "grad_norm": 0.7465129494667053,
+      "learning_rate": 0.0001814070351758794,
+      "loss": 0.3906,
+      "step": 195
+    },
+    {
+      "epoch": 0.098,
+      "grad_norm": 0.8752571940422058,
+      "learning_rate": 0.0001813065326633166,
+      "loss": 0.4701,
+      "step": 196
+    },
+    {
+      "epoch": 0.0985,
+      "grad_norm": 0.8893344402313232,
+      "learning_rate": 0.00018120603015075377,
+      "loss": 0.4432,
+      "step": 197
+    },
+    {
+      "epoch": 0.099,
+      "grad_norm": 0.9299802184104919,
+      "learning_rate": 0.00018110552763819096,
+      "loss": 0.458,
+      "step": 198
+    },
+    {
+      "epoch": 0.0995,
+      "grad_norm": 0.6995590329170227,
+      "learning_rate": 0.00018100502512562815,
+      "loss": 0.2739,
+      "step": 199
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8476576805114746,
+      "learning_rate": 0.00018090452261306533,
+      "loss": 0.3622,
+      "step": 200
+    },
+    {
+      "epoch": 0.1005,
+      "grad_norm": 0.6347658038139343,
+      "learning_rate": 0.00018080402010050252,
+      "loss": 0.6535,
+      "step": 201
+    },
+    {
+      "epoch": 0.101,
+      "grad_norm": 0.6387312412261963,
+      "learning_rate": 0.0001807035175879397,
+      "loss": 0.5971,
+      "step": 202
+    },
+    {
+      "epoch": 0.1015,
+      "grad_norm": 0.6946901082992554,
+      "learning_rate": 0.0001806030150753769,
+      "loss": 0.6498,
+      "step": 203
+    },
+    {
+      "epoch": 0.102,
+      "grad_norm": 0.9761562347412109,
+      "learning_rate": 0.00018050251256281408,
+      "loss": 0.854,
+      "step": 204
+    },
+    {
+      "epoch": 0.1025,
+      "grad_norm": 0.82000333070755,
+      "learning_rate": 0.00018040201005025127,
+      "loss": 0.5644,
+      "step": 205
+    },
+    {
+      "epoch": 0.103,
+      "grad_norm": 0.806016743183136,
+      "learning_rate": 0.00018030150753768845,
+      "loss": 0.6871,
+      "step": 206
+    },
+    {
+      "epoch": 0.1035,
+      "grad_norm": 0.8057332038879395,
+      "learning_rate": 0.00018020100502512564,
+      "loss": 0.6893,
+      "step": 207
+    },
+    {
+      "epoch": 0.104,
+      "grad_norm": 0.6865416169166565,
+      "learning_rate": 0.00018010050251256282,
+      "loss": 0.531,
+      "step": 208
+    },
+    {
+      "epoch": 0.1045,
+      "grad_norm": 0.7269001603126526,
+      "learning_rate": 0.00018,
+      "loss": 0.6496,
+      "step": 209
+    },
+    {
+      "epoch": 0.105,
+      "grad_norm": 0.7222223281860352,
+      "learning_rate": 0.0001798994974874372,
+      "loss": 0.552,
+      "step": 210
+    },
+    {
+      "epoch": 0.1055,
+      "grad_norm": 0.6749189496040344,
+      "learning_rate": 0.00017979899497487438,
+      "loss": 0.6567,
+      "step": 211
+    },
+    {
+      "epoch": 0.106,
+      "grad_norm": 0.7353073954582214,
+      "learning_rate": 0.00017969849246231157,
+      "loss": 0.6922,
+      "step": 212
+    },
+    {
+      "epoch": 0.1065,
+      "grad_norm": 0.8278074264526367,
+      "learning_rate": 0.00017959798994974876,
+      "loss": 0.7094,
+      "step": 213
+    },
+    {
+      "epoch": 0.107,
+      "grad_norm": 0.6379684209823608,
+      "learning_rate": 0.00017949748743718592,
+      "loss": 0.4995,
+      "step": 214
+    },
+    {
+      "epoch": 0.1075,
+      "grad_norm": 0.7294252514839172,
+      "learning_rate": 0.00017939698492462313,
+      "loss": 0.5262,
+      "step": 215
+    },
+    {
+      "epoch": 0.108,
+      "grad_norm": 0.7357264161109924,
+      "learning_rate": 0.00017929648241206032,
+      "loss": 0.5618,
+      "step": 216
+    },
+    {
+      "epoch": 0.1085,
+      "grad_norm": 0.8066862225532532,
+      "learning_rate": 0.0001791959798994975,
+      "loss": 0.5278,
+      "step": 217
+    },
+    {
+      "epoch": 0.109,
+      "grad_norm": 0.7261362075805664,
+      "learning_rate": 0.0001790954773869347,
+      "loss": 0.5847,
+      "step": 218
+    },
+    {
+      "epoch": 0.1095,
+      "grad_norm": 0.7264408469200134,
+      "learning_rate": 0.00017899497487437188,
+      "loss": 0.4799,
+      "step": 219
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6803249716758728,
+      "learning_rate": 0.00017889447236180906,
+      "loss": 0.4219,
+      "step": 220
+    },
+    {
+      "epoch": 0.1105,
+      "grad_norm": 0.7424831390380859,
+      "learning_rate": 0.00017879396984924622,
+      "loss": 0.5594,
+      "step": 221
+    },
+    {
+      "epoch": 0.111,
+      "grad_norm": 0.7046506404876709,
+      "learning_rate": 0.0001786934673366834,
+      "loss": 0.5511,
+      "step": 222
+    },
+    {
+      "epoch": 0.1115,
+      "grad_norm": 0.7620507478713989,
+      "learning_rate": 0.0001785929648241206,
+      "loss": 0.6128,
+      "step": 223
+    },
+    {
+      "epoch": 0.112,
+      "grad_norm": 0.7486132383346558,
+      "learning_rate": 0.0001784924623115578,
+      "loss": 0.4746,
+      "step": 224
+    },
+    {
+      "epoch": 0.1125,
+      "grad_norm": 0.721565306186676,
+      "learning_rate": 0.000178391959798995,
+      "loss": 0.5537,
+      "step": 225
+    },
+    {
+      "epoch": 0.113,
+      "grad_norm": 0.9783015847206116,
+      "learning_rate": 0.00017829145728643218,
+      "loss": 0.6283,
+      "step": 226
+    },
+    {
+      "epoch": 0.1135,
+      "grad_norm": 0.7669968605041504,
+      "learning_rate": 0.00017819095477386937,
+      "loss": 0.5103,
+      "step": 227
+    },
+    {
+      "epoch": 0.114,
+      "grad_norm": 0.743038535118103,
+      "learning_rate": 0.00017809045226130653,
+      "loss": 0.5431,
+      "step": 228
+    },
+    {
+      "epoch": 0.1145,
+      "grad_norm": 0.8427644371986389,
+      "learning_rate": 0.00017798994974874371,
+      "loss": 0.5874,
+      "step": 229
+    },
+    {
+      "epoch": 0.115,
+      "grad_norm": 0.8389422297477722,
+      "learning_rate": 0.0001778894472361809,
+      "loss": 0.4673,
+      "step": 230
+    },
+    {
+      "epoch": 0.1155,
+      "grad_norm": 0.8205832839012146,
+      "learning_rate": 0.0001777889447236181,
+      "loss": 0.465,
+      "step": 231
+    },
+    {
+      "epoch": 0.116,
+      "grad_norm": 0.7751673460006714,
+      "learning_rate": 0.0001776884422110553,
+      "loss": 0.4555,
+      "step": 232
+    },
+    {
+      "epoch": 0.1165,
+      "grad_norm": 0.7582501769065857,
+      "learning_rate": 0.0001775879396984925,
+      "loss": 0.439,
+      "step": 233
+    },
+    {
+      "epoch": 0.117,
+      "grad_norm": 0.7710978984832764,
+      "learning_rate": 0.00017748743718592967,
+      "loss": 0.4656,
+      "step": 234
+    },
+    {
+      "epoch": 0.1175,
+      "grad_norm": 0.7527711987495422,
+      "learning_rate": 0.00017738693467336683,
+      "loss": 0.4489,
+      "step": 235
+    },
+    {
+      "epoch": 0.118,
+      "grad_norm": 0.8192912936210632,
+      "learning_rate": 0.00017728643216080402,
+      "loss": 0.5772,
+      "step": 236
+    },
+    {
+      "epoch": 0.1185,
+      "grad_norm": 0.7981165647506714,
+      "learning_rate": 0.0001771859296482412,
+      "loss": 0.4579,
+      "step": 237
+    },
+    {
+      "epoch": 0.119,
+      "grad_norm": 0.739599347114563,
+      "learning_rate": 0.0001770854271356784,
+      "loss": 0.3964,
+      "step": 238
+    },
+    {
+      "epoch": 0.1195,
+      "grad_norm": 0.6874092221260071,
+      "learning_rate": 0.00017698492462311558,
+      "loss": 0.3753,
+      "step": 239
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7658870220184326,
+      "learning_rate": 0.0001768844221105528,
+      "loss": 0.4417,
+      "step": 240
+    },
+    {
+      "epoch": 0.1205,
+      "grad_norm": 0.6945328116416931,
+      "learning_rate": 0.00017678391959798998,
+      "loss": 0.4079,
+      "step": 241
+    },
+    {
+      "epoch": 0.121,
+      "grad_norm": 0.8424277901649475,
+      "learning_rate": 0.00017668341708542714,
+      "loss": 0.5034,
+      "step": 242
+    },
+    {
+      "epoch": 0.1215,
+      "grad_norm": 0.7296550869941711,
+      "learning_rate": 0.00017658291457286432,
+      "loss": 0.4075,
+      "step": 243
+    },
+    {
+      "epoch": 0.122,
+      "grad_norm": 0.6844024062156677,
+      "learning_rate": 0.0001764824120603015,
+      "loss": 0.3183,
+      "step": 244
+    },
+    {
+      "epoch": 0.1225,
+      "grad_norm": 0.7940804958343506,
+      "learning_rate": 0.0001763819095477387,
+      "loss": 0.4526,
+      "step": 245
+    },
+    {
+      "epoch": 0.123,
+      "grad_norm": 0.831255316734314,
+      "learning_rate": 0.00017628140703517588,
+      "loss": 0.4748,
+      "step": 246
+    },
+    {
+      "epoch": 0.1235,
+      "grad_norm": 0.9887934923171997,
+      "learning_rate": 0.00017618090452261307,
+      "loss": 0.5011,
+      "step": 247
+    },
+    {
+      "epoch": 0.124,
+      "grad_norm": 0.789718747138977,
+      "learning_rate": 0.00017608040201005026,
+      "loss": 0.3772,
+      "step": 248
+    },
+    {
+      "epoch": 0.1245,
+      "grad_norm": 0.9952366352081299,
+      "learning_rate": 0.00017597989949748744,
+      "loss": 0.5481,
+      "step": 249
+    },
+    {
+      "epoch": 0.125,
+      "grad_norm": 0.8396393060684204,
+      "learning_rate": 0.00017587939698492463,
+      "loss": 0.4619,
+      "step": 250
+    },
+    {
+      "epoch": 0.1255,
+      "grad_norm": 0.7401638627052307,
+      "learning_rate": 0.00017577889447236182,
+      "loss": 0.5709,
+      "step": 251
+    },
+    {
+      "epoch": 0.126,
+      "grad_norm": 0.7383289337158203,
+      "learning_rate": 0.000175678391959799,
+      "loss": 0.6041,
+      "step": 252
+    },
+    {
+      "epoch": 0.1265,
+      "grad_norm": 0.6855509877204895,
+      "learning_rate": 0.0001755778894472362,
+      "loss": 0.5268,
+      "step": 253
+    },
+    {
+      "epoch": 0.127,
+      "grad_norm": 0.6161994338035583,
+      "learning_rate": 0.00017547738693467338,
+      "loss": 0.4399,
+      "step": 254
+    },
+    {
+      "epoch": 0.1275,
+      "grad_norm": 0.6935853958129883,
+      "learning_rate": 0.00017537688442211056,
+      "loss": 0.5343,
+      "step": 255
+    },
+    {
+      "epoch": 0.128,
+      "grad_norm": 0.8727939128875732,
+      "learning_rate": 0.00017527638190954775,
+      "loss": 0.8002,
+      "step": 256
+    },
+    {
+      "epoch": 0.1285,
+      "grad_norm": 0.7109793424606323,
+      "learning_rate": 0.00017517587939698494,
+      "loss": 0.6442,
+      "step": 257
+    },
+    {
+      "epoch": 0.129,
+      "grad_norm": 0.7154646515846252,
+      "learning_rate": 0.00017507537688442212,
+      "loss": 0.5754,
+      "step": 258
+    },
+    {
+      "epoch": 0.1295,
+      "grad_norm": 0.7552391290664673,
+      "learning_rate": 0.0001749748743718593,
+      "loss": 0.614,
+      "step": 259
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.648891031742096,
+      "learning_rate": 0.0001748743718592965,
+      "loss": 0.4961,
+      "step": 260
+    },
+    {
+      "epoch": 0.1305,
+      "grad_norm": 0.7435721755027771,
+      "learning_rate": 0.00017477386934673368,
+      "loss": 0.7027,
+      "step": 261
+    },
+    {
+      "epoch": 0.131,
+      "grad_norm": 0.6650488972663879,
+      "learning_rate": 0.00017467336683417087,
+      "loss": 0.5429,
+      "step": 262
+    },
+    {
+      "epoch": 0.1315,
+      "grad_norm": 0.6643757224082947,
+      "learning_rate": 0.00017457286432160805,
+      "loss": 0.5206,
+      "step": 263
+    },
+    {
+      "epoch": 0.132,
+      "grad_norm": 0.5941948890686035,
+      "learning_rate": 0.00017447236180904521,
+      "loss": 0.4143,
+      "step": 264
+    },
+    {
+      "epoch": 0.1325,
+      "grad_norm": 0.6611546874046326,
+      "learning_rate": 0.0001743718592964824,
+      "loss": 0.4542,
+      "step": 265
+    },
+    {
+      "epoch": 0.133,
+      "grad_norm": 0.769702672958374,
+      "learning_rate": 0.00017427135678391961,
+      "loss": 0.6865,
+      "step": 266
+    },
+    {
+      "epoch": 0.1335,
+      "grad_norm": 0.7664195895195007,
+      "learning_rate": 0.0001741708542713568,
+      "loss": 0.5243,
+      "step": 267
+    },
+    {
+      "epoch": 0.134,
+      "grad_norm": 0.6314630508422852,
+      "learning_rate": 0.000174070351758794,
+      "loss": 0.4554,
+      "step": 268
+    },
+    {
+      "epoch": 0.1345,
+      "grad_norm": 0.7502909302711487,
+      "learning_rate": 0.00017396984924623117,
+      "loss": 0.5743,
+      "step": 269
+    },
+    {
+      "epoch": 0.135,
+      "grad_norm": 0.714921236038208,
+      "learning_rate": 0.00017386934673366836,
+      "loss": 0.5093,
+      "step": 270
+    },
+    {
+      "epoch": 0.1355,
+      "grad_norm": 0.6410713195800781,
+      "learning_rate": 0.00017376884422110552,
+      "loss": 0.4267,
+      "step": 271
+    },
+    {
+      "epoch": 0.136,
+      "grad_norm": 0.77266526222229,
+      "learning_rate": 0.0001736683417085427,
+      "loss": 0.4684,
+      "step": 272
+    },
+    {
+      "epoch": 0.1365,
+      "grad_norm": 0.7584907412528992,
+      "learning_rate": 0.0001735678391959799,
+      "loss": 0.5401,
+      "step": 273
+    },
+    {
+      "epoch": 0.137,
+      "grad_norm": 0.8193687200546265,
+      "learning_rate": 0.0001734673366834171,
+      "loss": 0.5039,
+      "step": 274
+    },
+    {
+      "epoch": 0.1375,
+      "grad_norm": 0.7617428302764893,
+      "learning_rate": 0.0001733668341708543,
+      "loss": 0.6295,
+      "step": 275
+    },
+    {
+      "epoch": 0.138,
+      "grad_norm": 0.963472306728363,
+      "learning_rate": 0.00017326633165829148,
+      "loss": 0.6334,
+      "step": 276
+    },
+    {
+      "epoch": 0.1385,
+      "grad_norm": 0.748769223690033,
+      "learning_rate": 0.00017316582914572867,
+      "loss": 0.5854,
+      "step": 277
+    },
+    {
+      "epoch": 0.139,
+      "grad_norm": 0.7013671398162842,
+      "learning_rate": 0.00017306532663316582,
+      "loss": 0.509,
+      "step": 278
+    },
+    {
+      "epoch": 0.1395,
+      "grad_norm": 1.1528773307800293,
+      "learning_rate": 0.000172964824120603,
+      "loss": 0.5977,
+      "step": 279
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7551888823509216,
+      "learning_rate": 0.0001728643216080402,
+      "loss": 0.4998,
+      "step": 280
+    },
+    {
+      "epoch": 0.1405,
+      "grad_norm": 0.6252229809761047,
+      "learning_rate": 0.00017276381909547738,
+      "loss": 0.4239,
+      "step": 281
+    },
+    {
+      "epoch": 0.141,
+      "grad_norm": 0.6933767199516296,
+      "learning_rate": 0.0001726633165829146,
+      "loss": 0.4463,
+      "step": 282
+    },
+    {
+      "epoch": 0.1415,
+      "grad_norm": 0.7303416132926941,
+      "learning_rate": 0.00017256281407035178,
+      "loss": 0.5525,
+      "step": 283
+    },
+    {
+      "epoch": 0.142,
+      "grad_norm": 0.6901214718818665,
+      "learning_rate": 0.00017246231155778897,
+      "loss": 0.5067,
+      "step": 284
+    },
+    {
+      "epoch": 0.1425,
+      "grad_norm": 0.7375679016113281,
+      "learning_rate": 0.00017236180904522613,
+      "loss": 0.5384,
+      "step": 285
+    },
+    {
+      "epoch": 0.143,
+      "grad_norm": 0.7633487582206726,
+      "learning_rate": 0.00017226130653266332,
+      "loss": 0.4812,
+      "step": 286
+    },
+    {
+      "epoch": 0.1435,
+      "grad_norm": 0.7331466674804688,
+      "learning_rate": 0.0001721608040201005,
+      "loss": 0.4281,
+      "step": 287
+    },
+    {
+      "epoch": 0.144,
+      "grad_norm": 0.6968885660171509,
+      "learning_rate": 0.0001720603015075377,
+      "loss": 0.4906,
+      "step": 288
+    },
+    {
+      "epoch": 0.1445,
+      "grad_norm": 0.7250369787216187,
+      "learning_rate": 0.00017195979899497488,
+      "loss": 0.4734,
+      "step": 289
+    },
+    {
+      "epoch": 0.145,
+      "grad_norm": 0.7470961809158325,
+      "learning_rate": 0.00017185929648241206,
+      "loss": 0.5223,
+      "step": 290
+    },
+    {
+      "epoch": 0.1455,
+      "grad_norm": 0.8535013794898987,
+      "learning_rate": 0.00017175879396984928,
+      "loss": 0.5311,
+      "step": 291
+    },
+    {
+      "epoch": 0.146,
+      "grad_norm": 0.7980099320411682,
+      "learning_rate": 0.00017165829145728644,
+      "loss": 0.4995,
+      "step": 292
+    },
+    {
+      "epoch": 0.1465,
+      "grad_norm": 0.8489773273468018,
+      "learning_rate": 0.00017155778894472362,
+      "loss": 0.4885,
+      "step": 293
+    },
+    {
+      "epoch": 0.147,
+      "grad_norm": 0.8632635474205017,
+      "learning_rate": 0.0001714572864321608,
+      "loss": 0.5617,
+      "step": 294
+    },
+    {
+      "epoch": 0.1475,
+      "grad_norm": 0.6674500107765198,
+      "learning_rate": 0.000171356783919598,
+      "loss": 0.3971,
+      "step": 295
+    },
+    {
+      "epoch": 0.148,
+      "grad_norm": 0.7467655539512634,
+      "learning_rate": 0.00017125628140703518,
+      "loss": 0.3979,
+      "step": 296
+    },
+    {
+      "epoch": 0.1485,
+      "grad_norm": 0.7403197884559631,
+      "learning_rate": 0.00017115577889447237,
+      "loss": 0.4306,
+      "step": 297
+    },
+    {
+      "epoch": 0.149,
+      "grad_norm": 0.8875041007995605,
+      "learning_rate": 0.00017105527638190955,
+      "loss": 0.4453,
+      "step": 298
+    },
+    {
+      "epoch": 0.1495,
+      "grad_norm": 0.6805903911590576,
+      "learning_rate": 0.00017095477386934674,
+      "loss": 0.3169,
+      "step": 299
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8135784268379211,
+      "learning_rate": 0.00017085427135678393,
+      "loss": 0.3775,
+      "step": 300
+    },
+    {
+      "epoch": 0.1505,
+      "grad_norm": 0.6680217981338501,
+      "learning_rate": 0.00017075376884422111,
+      "loss": 0.6488,
+      "step": 301
+    },
+    {
+      "epoch": 0.151,
+      "grad_norm": 0.7143925428390503,
+      "learning_rate": 0.0001706532663316583,
+      "loss": 0.6792,
+      "step": 302
+    },
+    {
+      "epoch": 0.1515,
+      "grad_norm": 0.7646191716194153,
+      "learning_rate": 0.0001705527638190955,
+      "loss": 0.5628,
+      "step": 303
+    },
+    {
+      "epoch": 0.152,
+      "grad_norm": 0.707794725894928,
+      "learning_rate": 0.00017045226130653267,
+      "loss": 0.5163,
+      "step": 304
+    },
+    {
+      "epoch": 0.1525,
+      "grad_norm": 0.7562041282653809,
+      "learning_rate": 0.00017035175879396986,
+      "loss": 0.6485,
+      "step": 305
+    },
+    {
+      "epoch": 0.153,
+      "grad_norm": 0.6837920546531677,
+      "learning_rate": 0.00017025125628140705,
+      "loss": 0.5763,
+      "step": 306
+    },
+    {
+      "epoch": 0.1535,
+      "grad_norm": 0.7620954513549805,
+      "learning_rate": 0.00017015075376884423,
+      "loss": 0.7071,
+      "step": 307
+    },
+    {
+      "epoch": 0.154,
+      "grad_norm": 0.7150668501853943,
+      "learning_rate": 0.00017005025125628142,
+      "loss": 0.6262,
+      "step": 308
+    },
+    {
+      "epoch": 0.1545,
+      "grad_norm": 0.7056312561035156,
+      "learning_rate": 0.0001699497487437186,
+      "loss": 0.6862,
+      "step": 309
+    },
+    {
+      "epoch": 0.155,
+      "grad_norm": 0.7380791306495667,
+      "learning_rate": 0.0001698492462311558,
+      "loss": 0.5437,
+      "step": 310
+    },
+    {
+      "epoch": 0.1555,
+      "grad_norm": 0.7088383436203003,
+      "learning_rate": 0.00016974874371859298,
+      "loss": 0.5417,
+      "step": 311
+    },
+    {
+      "epoch": 0.156,
+      "grad_norm": 0.6764571070671082,
+      "learning_rate": 0.00016964824120603016,
+      "loss": 0.4749,
+      "step": 312
+    },
+    {
+      "epoch": 0.1565,
+      "grad_norm": 0.6379378437995911,
+      "learning_rate": 0.00016954773869346735,
+      "loss": 0.496,
+      "step": 313
+    },
+    {
+      "epoch": 0.157,
+      "grad_norm": 0.7442959547042847,
+      "learning_rate": 0.0001694472361809045,
+      "loss": 0.6728,
+      "step": 314
+    },
+    {
+      "epoch": 0.1575,
+      "grad_norm": 0.6742901802062988,
+      "learning_rate": 0.0001693467336683417,
+      "loss": 0.5436,
+      "step": 315
+    },
+    {
+      "epoch": 0.158,
+      "grad_norm": 0.7418420314788818,
+      "learning_rate": 0.0001692462311557789,
+      "loss": 0.507,
+      "step": 316
+    },
+    {
+      "epoch": 0.1585,
+      "grad_norm": 0.7401244044303894,
+      "learning_rate": 0.0001691457286432161,
+      "loss": 0.5694,
+      "step": 317
+    },
+    {
+      "epoch": 0.159,
+      "grad_norm": 0.6102404594421387,
+      "learning_rate": 0.00016904522613065328,
+      "loss": 0.4973,
+      "step": 318
+    },
+    {
+      "epoch": 0.1595,
+      "grad_norm": 0.7396304607391357,
+      "learning_rate": 0.00016894472361809047,
+      "loss": 0.4643,
+      "step": 319
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7031675577163696,
+      "learning_rate": 0.00016884422110552766,
+      "loss": 0.5225,
+      "step": 320
+    },
+    {
+      "epoch": 0.1605,
+      "grad_norm": 0.7107573747634888,
+      "learning_rate": 0.00016874371859296482,
+      "loss": 0.5175,
+      "step": 321
+    },
+    {
+      "epoch": 0.161,
+      "grad_norm": 0.7950466871261597,
+      "learning_rate": 0.000168643216080402,
+      "loss": 0.5918,
+      "step": 322
+    },
+    {
+      "epoch": 0.1615,
+      "grad_norm": 0.8171111941337585,
+      "learning_rate": 0.0001685427135678392,
+      "loss": 0.6103,
+      "step": 323
+    },
+    {
+      "epoch": 0.162,
+      "grad_norm": 0.7460869550704956,
+      "learning_rate": 0.0001684422110552764,
+      "loss": 0.5362,
+      "step": 324
+    },
+    {
+      "epoch": 0.1625,
+      "grad_norm": 0.8052082061767578,
+      "learning_rate": 0.0001683417085427136,
+      "loss": 0.5237,
+      "step": 325
+    },
+    {
+      "epoch": 0.163,
+      "grad_norm": 0.7894774079322815,
+      "learning_rate": 0.00016824120603015078,
+      "loss": 0.4607,
+      "step": 326
+    },
+    {
+      "epoch": 0.1635,
+      "grad_norm": 0.8383395075798035,
+      "learning_rate": 0.00016814070351758796,
+      "loss": 0.5701,
+      "step": 327
+    },
+    {
+      "epoch": 0.164,
+      "grad_norm": 0.7205836772918701,
+      "learning_rate": 0.00016804020100502512,
+      "loss": 0.4555,
+      "step": 328
+    },
+    {
+      "epoch": 0.1645,
+      "grad_norm": 0.6374972462654114,
+      "learning_rate": 0.0001679396984924623,
+      "loss": 0.3601,
+      "step": 329
+    },
+    {
+      "epoch": 0.165,
+      "grad_norm": 0.6702671647071838,
+      "learning_rate": 0.0001678391959798995,
+      "loss": 0.4028,
+      "step": 330
+    },
+    {
+      "epoch": 0.1655,
+      "grad_norm": 0.734506368637085,
+      "learning_rate": 0.00016773869346733668,
+      "loss": 0.4537,
+      "step": 331
+    },
+    {
+      "epoch": 0.166,
+      "grad_norm": 0.7993707656860352,
+      "learning_rate": 0.0001676381909547739,
+      "loss": 0.4872,
+      "step": 332
+    },
+    {
+      "epoch": 0.1665,
+      "grad_norm": 0.8910720348358154,
+      "learning_rate": 0.00016753768844221108,
+      "loss": 0.5551,
+      "step": 333
+    },
+    {
+      "epoch": 0.167,
+      "grad_norm": 0.827415943145752,
+      "learning_rate": 0.00016743718592964827,
+      "loss": 0.4656,
+      "step": 334
+    },
+    {
+      "epoch": 0.1675,
+      "grad_norm": 0.7187349796295166,
+      "learning_rate": 0.00016733668341708543,
+      "loss": 0.4162,
+      "step": 335
+    },
+    {
+      "epoch": 0.168,
+      "grad_norm": 0.7324169874191284,
+      "learning_rate": 0.0001672361809045226,
+      "loss": 0.4599,
+      "step": 336
+    },
+    {
+      "epoch": 0.1685,
+      "grad_norm": 0.8108154535293579,
+      "learning_rate": 0.0001671356783919598,
+      "loss": 0.5099,
+      "step": 337
+    },
+    {
+      "epoch": 0.169,
+      "grad_norm": 0.7957372665405273,
+      "learning_rate": 0.00016703517587939699,
+      "loss": 0.4547,
+      "step": 338
+    },
+    {
+      "epoch": 0.1695,
+      "grad_norm": 0.6938037872314453,
+      "learning_rate": 0.00016693467336683417,
+      "loss": 0.4329,
+      "step": 339
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7371119260787964,
+      "learning_rate": 0.00016683417085427136,
+      "loss": 0.4157,
+      "step": 340
+    },
+    {
+      "epoch": 0.1705,
+      "grad_norm": 0.7351891398429871,
+      "learning_rate": 0.00016673366834170857,
+      "loss": 0.3646,
+      "step": 341
+    },
+    {
+      "epoch": 0.171,
+      "grad_norm": 0.7066468000411987,
+      "learning_rate": 0.00016663316582914573,
+      "loss": 0.3744,
+      "step": 342
+    },
+    {
+      "epoch": 0.1715,
+      "grad_norm": 0.6542836427688599,
+      "learning_rate": 0.00016653266331658292,
+      "loss": 0.3551,
+      "step": 343
+    },
+    {
+      "epoch": 0.172,
+      "grad_norm": 0.7765346169471741,
+      "learning_rate": 0.0001664321608040201,
+      "loss": 0.4117,
+      "step": 344
+    },
+    {
+      "epoch": 0.1725,
+      "grad_norm": 0.8277013897895813,
+      "learning_rate": 0.0001663316582914573,
+      "loss": 0.5079,
+      "step": 345
+    },
+    {
+      "epoch": 0.173,
+      "grad_norm": 0.7319324612617493,
+      "learning_rate": 0.00016623115577889448,
+      "loss": 0.4316,
+      "step": 346
+    },
+    {
+      "epoch": 0.1735,
+      "grad_norm": 0.9099276661872864,
+      "learning_rate": 0.00016613065326633166,
+      "loss": 0.5007,
+      "step": 347
+    },
+    {
+      "epoch": 0.174,
+      "grad_norm": 0.7808762192726135,
+      "learning_rate": 0.00016603015075376885,
+      "loss": 0.4041,
+      "step": 348
+    },
+    {
+      "epoch": 0.1745,
+      "grad_norm": 0.93597811460495,
+      "learning_rate": 0.00016592964824120604,
+      "loss": 0.5765,
+      "step": 349
+    },
+    {
+      "epoch": 0.175,
+      "grad_norm": 0.9638497233390808,
+      "learning_rate": 0.00016582914572864322,
+      "loss": 0.4794,
+      "step": 350
+    },
+    {
+      "epoch": 0.1755,
+      "grad_norm": 0.6557117104530334,
+      "learning_rate": 0.0001657286432160804,
+      "loss": 0.6915,
+      "step": 351
+    },
+    {
+      "epoch": 0.176,
+      "grad_norm": 0.599416971206665,
+      "learning_rate": 0.0001656281407035176,
+      "loss": 0.4561,
+      "step": 352
+    },
+    {
+      "epoch": 0.1765,
+      "grad_norm": 0.6415974497795105,
+      "learning_rate": 0.00016552763819095478,
+      "loss": 0.5266,
+      "step": 353
+    },
+    {
+      "epoch": 0.177,
+      "grad_norm": 0.6229732632637024,
+      "learning_rate": 0.00016542713567839197,
+      "loss": 0.5588,
+      "step": 354
+    },
+    {
+      "epoch": 0.1775,
+      "grad_norm": 0.7231833934783936,
+      "learning_rate": 0.00016532663316582916,
+      "loss": 0.4866,
+      "step": 355
+    },
+    {
+      "epoch": 0.178,
+      "grad_norm": 0.8384533524513245,
+      "learning_rate": 0.00016522613065326634,
+      "loss": 0.7369,
+      "step": 356
+    },
+    {
+      "epoch": 0.1785,
+      "grad_norm": 0.6029393076896667,
+      "learning_rate": 0.00016512562814070353,
+      "loss": 0.481,
+      "step": 357
+    },
+    {
+      "epoch": 0.179,
+      "grad_norm": 0.6602717638015747,
+      "learning_rate": 0.00016502512562814072,
+      "loss": 0.5202,
+      "step": 358
+    },
+    {
+      "epoch": 0.1795,
+      "grad_norm": 0.8618465065956116,
+      "learning_rate": 0.0001649246231155779,
+      "loss": 0.7511,
+      "step": 359
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6275924444198608,
+      "learning_rate": 0.0001648241206030151,
+      "loss": 0.4586,
+      "step": 360
+    },
+    {
+      "epoch": 0.1805,
+      "grad_norm": 0.7058954834938049,
+      "learning_rate": 0.00016472361809045228,
+      "loss": 0.515,
+      "step": 361
+    },
+    {
+      "epoch": 0.181,
+      "grad_norm": 0.661612331867218,
+      "learning_rate": 0.00016462311557788946,
+      "loss": 0.5178,
+      "step": 362
+    },
+    {
+      "epoch": 0.1815,
+      "grad_norm": 0.6896937489509583,
+      "learning_rate": 0.00016452261306532665,
+      "loss": 0.5794,
+      "step": 363
+    },
+    {
+      "epoch": 0.182,
+      "grad_norm": 0.8359775543212891,
+      "learning_rate": 0.0001644221105527638,
+      "loss": 0.5896,
+      "step": 364
+    },
+    {
+      "epoch": 0.1825,
+      "grad_norm": 0.8894909024238586,
+      "learning_rate": 0.000164321608040201,
+      "loss": 0.6271,
+      "step": 365
+    },
+    {
+      "epoch": 0.183,
+      "grad_norm": 0.9869162440299988,
+      "learning_rate": 0.0001642211055276382,
+      "loss": 0.6255,
+      "step": 366
+    },
+    {
+      "epoch": 0.1835,
+      "grad_norm": 0.8036964535713196,
+      "learning_rate": 0.0001641206030150754,
+      "loss": 0.5796,
+      "step": 367
+    },
+    {
+      "epoch": 0.184,
+      "grad_norm": 0.6997955441474915,
+      "learning_rate": 0.00016402010050251258,
+      "loss": 0.518,
+      "step": 368
+    },
+    {
+      "epoch": 0.1845,
+      "grad_norm": 0.7880759835243225,
+      "learning_rate": 0.00016391959798994977,
+      "loss": 0.5399,
+      "step": 369
+    },
+    {
+      "epoch": 0.185,
+      "grad_norm": 0.7499764561653137,
+      "learning_rate": 0.00016381909547738695,
+      "loss": 0.6129,
+      "step": 370
+    },
+    {
+      "epoch": 0.1855,
+      "grad_norm": 0.7060808539390564,
+      "learning_rate": 0.0001637185929648241,
+      "loss": 0.5648,
+      "step": 371
+    },
+    {
+      "epoch": 0.186,
+      "grad_norm": 0.6893482208251953,
+      "learning_rate": 0.0001636180904522613,
+      "loss": 0.4841,
+      "step": 372
+    },
+    {
+      "epoch": 0.1865,
+      "grad_norm": 0.7040382027626038,
+      "learning_rate": 0.00016351758793969849,
+      "loss": 0.4929,
+      "step": 373
+    },
+    {
+      "epoch": 0.187,
+      "grad_norm": 0.6883841156959534,
+      "learning_rate": 0.0001634170854271357,
+      "loss": 0.5268,
+      "step": 374
+    },
+    {
+      "epoch": 0.1875,
+      "grad_norm": 0.6659446358680725,
+      "learning_rate": 0.00016331658291457289,
+      "loss": 0.4911,
+      "step": 375
+    },
+    {
+      "epoch": 0.188,
+      "grad_norm": 0.6387864351272583,
+      "learning_rate": 0.00016321608040201007,
+      "loss": 0.4292,
+      "step": 376
+    },
+    {
+      "epoch": 0.1885,
+      "grad_norm": 0.7901704907417297,
+      "learning_rate": 0.00016311557788944726,
+      "loss": 0.5813,
+      "step": 377
+    },
+    {
+      "epoch": 0.189,
+      "grad_norm": 0.7493123412132263,
+      "learning_rate": 0.00016301507537688442,
+      "loss": 0.5928,
+      "step": 378
+    },
+    {
+      "epoch": 0.1895,
+      "grad_norm": 0.6282365322113037,
+      "learning_rate": 0.0001629145728643216,
+      "loss": 0.445,
+      "step": 379
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7925242185592651,
+      "learning_rate": 0.0001628140703517588,
+      "loss": 0.5562,
+      "step": 380
+    },
+    {
+      "epoch": 0.1905,
+      "grad_norm": 0.6898102760314941,
+      "learning_rate": 0.00016271356783919598,
+      "loss": 0.4087,
+      "step": 381
+    },
+    {
+      "epoch": 0.191,
+      "grad_norm": 0.7045435905456543,
+      "learning_rate": 0.00016261306532663316,
+      "loss": 0.4987,
+      "step": 382
+    },
+    {
+      "epoch": 0.1915,
+      "grad_norm": 0.8399732112884521,
+      "learning_rate": 0.00016251256281407038,
+      "loss": 0.5215,
+      "step": 383
+    },
+    {
+      "epoch": 0.192,
+      "grad_norm": 0.6442425847053528,
+      "learning_rate": 0.00016241206030150756,
+      "loss": 0.4272,
+      "step": 384
+    },
+    {
+      "epoch": 0.1925,
+      "grad_norm": 0.9033990502357483,
+      "learning_rate": 0.00016231155778894472,
+      "loss": 0.6684,
+      "step": 385
+    },
+    {
+      "epoch": 0.193,
+      "grad_norm": 0.8048330545425415,
+      "learning_rate": 0.0001622110552763819,
+      "loss": 0.5655,
+      "step": 386
+    },
+    {
+      "epoch": 0.1935,
+      "grad_norm": 0.7665651440620422,
+      "learning_rate": 0.0001621105527638191,
+      "loss": 0.5133,
+      "step": 387
+    },
+    {
+      "epoch": 0.194,
+      "grad_norm": 0.6844928860664368,
+      "learning_rate": 0.00016201005025125628,
+      "loss": 0.3911,
+      "step": 388
+    },
+    {
+      "epoch": 0.1945,
+      "grad_norm": 0.7046768665313721,
+      "learning_rate": 0.00016190954773869347,
+      "loss": 0.4391,
+      "step": 389
+    },
+    {
+      "epoch": 0.195,
+      "grad_norm": 0.6305338740348816,
+      "learning_rate": 0.00016180904522613066,
+      "loss": 0.4001,
+      "step": 390
+    },
+    {
+      "epoch": 0.1955,
+      "grad_norm": 0.8463757038116455,
+      "learning_rate": 0.00016170854271356787,
+      "loss": 0.5693,
+      "step": 391
+    },
+    {
+      "epoch": 0.196,
+      "grad_norm": 0.6951143145561218,
+      "learning_rate": 0.00016160804020100503,
+      "loss": 0.402,
+      "step": 392
+    },
+    {
+      "epoch": 0.1965,
+      "grad_norm": 0.7149537205696106,
+      "learning_rate": 0.00016150753768844222,
+      "loss": 0.3903,
+      "step": 393
+    },
+    {
+      "epoch": 0.197,
+      "grad_norm": 0.579406201839447,
+      "learning_rate": 0.0001614070351758794,
+      "loss": 0.2981,
+      "step": 394
+    },
+    {
+      "epoch": 0.1975,
+      "grad_norm": 0.7180090546607971,
+      "learning_rate": 0.0001613065326633166,
+      "loss": 0.4418,
+      "step": 395
+    },
+    {
+      "epoch": 0.198,
+      "grad_norm": 0.7786920070648193,
+      "learning_rate": 0.00016120603015075378,
+      "loss": 0.4532,
+      "step": 396
+    },
+    {
+      "epoch": 0.1985,
+      "grad_norm": 0.7447192072868347,
+      "learning_rate": 0.00016110552763819096,
+      "loss": 0.3753,
+      "step": 397
+    },
+    {
+      "epoch": 0.199,
+      "grad_norm": 0.7644127011299133,
+      "learning_rate": 0.00016100502512562815,
+      "loss": 0.4325,
+      "step": 398
+    },
+    {
+      "epoch": 0.1995,
+      "grad_norm": 0.8857213854789734,
+      "learning_rate": 0.00016090452261306533,
+      "loss": 0.4291,
+      "step": 399
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8126779794692993,
+      "learning_rate": 0.00016080402010050252,
+      "loss": 0.3671,
+      "step": 400
+    },
+    {
+      "epoch": 0.2,
+      "eval_loss": 0.5126131772994995,
+      "eval_runtime": 109.0134,
+      "eval_samples_per_second": 4.587,
+      "eval_steps_per_second": 4.587,
+      "step": 400
+    },
+    {
+      "epoch": 0.2005,
+      "grad_norm": 0.7218247056007385,
+      "learning_rate": 0.0001607035175879397,
+      "loss": 0.7444,
+      "step": 401
+    },
+    {
+      "epoch": 0.201,
+      "grad_norm": 0.6251023411750793,
+      "learning_rate": 0.0001606030150753769,
+      "loss": 0.6868,
+      "step": 402
+    },
+    {
+      "epoch": 0.2015,
+      "grad_norm": 0.7314577698707581,
+      "learning_rate": 0.00016050251256281408,
+      "loss": 0.7422,
+      "step": 403
+    },
+    {
+      "epoch": 0.202,
+      "grad_norm": 0.6857109665870667,
+      "learning_rate": 0.00016040201005025127,
+      "loss": 0.6468,
+      "step": 404
+    },
+    {
+      "epoch": 0.2025,
+      "grad_norm": 0.7708049416542053,
+      "learning_rate": 0.00016030150753768845,
+      "loss": 0.6255,
+      "step": 405
+    },
+    {
+      "epoch": 0.203,
+      "grad_norm": 0.8734421133995056,
+      "learning_rate": 0.00016020100502512564,
+      "loss": 0.7415,
+      "step": 406
+    },
+    {
+      "epoch": 0.2035,
+      "grad_norm": 0.6194499135017395,
+      "learning_rate": 0.0001601005025125628,
+      "loss": 0.4667,
+      "step": 407
+    },
+    {
+      "epoch": 0.204,
+      "grad_norm": 0.6940017938613892,
+      "learning_rate": 0.00016,
+      "loss": 0.5843,
+      "step": 408
+    },
+    {
+      "epoch": 0.2045,
+      "grad_norm": 0.738488495349884,
+      "learning_rate": 0.0001598994974874372,
+      "loss": 0.6505,
+      "step": 409
+    },
+    {
+      "epoch": 0.205,
+      "grad_norm": 0.6275967955589294,
+      "learning_rate": 0.00015979899497487439,
+      "loss": 0.4914,
+      "step": 410
+    },
+    {
+      "epoch": 0.2055,
+      "grad_norm": 0.564113199710846,
+      "learning_rate": 0.00015969849246231157,
+      "loss": 0.4367,
+      "step": 411
+    },
+    {
+      "epoch": 0.206,
+      "grad_norm": 0.6475598216056824,
+      "learning_rate": 0.00015959798994974876,
+      "loss": 0.4509,
+      "step": 412
+    },
+    {
+      "epoch": 0.2065,
+      "grad_norm": 0.7129649519920349,
+      "learning_rate": 0.00015949748743718595,
+      "loss": 0.5064,
+      "step": 413
+    },
+    {
+      "epoch": 0.207,
+      "grad_norm": 0.7318679690361023,
+      "learning_rate": 0.0001593969849246231,
+      "loss": 0.6094,
+      "step": 414
+    },
+    {
+      "epoch": 0.2075,
+      "grad_norm": 0.6968581676483154,
+      "learning_rate": 0.0001592964824120603,
+      "loss": 0.5121,
+      "step": 415
+    },
+    {
+      "epoch": 0.208,
+      "grad_norm": 0.7797547578811646,
+      "learning_rate": 0.0001591959798994975,
+      "loss": 0.5492,
+      "step": 416
+    },
+    {
+      "epoch": 0.2085,
+      "grad_norm": 0.733516275882721,
+      "learning_rate": 0.0001590954773869347,
+      "loss": 0.5657,
+      "step": 417
+    },
+    {
+      "epoch": 0.209,
+      "grad_norm": 0.7635998129844666,
+      "learning_rate": 0.00015899497487437188,
+      "loss": 0.6398,
+      "step": 418
+    },
+    {
+      "epoch": 0.2095,
+      "grad_norm": 0.7976422905921936,
+      "learning_rate": 0.00015889447236180906,
+      "loss": 0.6026,
+      "step": 419
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.650238037109375,
+      "learning_rate": 0.00015879396984924625,
+      "loss": 0.4502,
+      "step": 420
+    },
+    {
+      "epoch": 0.2105,
+      "grad_norm": 0.685032844543457,
+      "learning_rate": 0.0001586934673366834,
+      "loss": 0.5073,
+      "step": 421
+    },
+    {
+      "epoch": 0.211,
+      "grad_norm": 0.7014485597610474,
+      "learning_rate": 0.0001585929648241206,
+      "loss": 0.4889,
+      "step": 422
+    },
+    {
+      "epoch": 0.2115,
+      "grad_norm": 0.6189039349555969,
+      "learning_rate": 0.00015849246231155778,
+      "loss": 0.4129,
+      "step": 423
+    },
+    {
+      "epoch": 0.212,
+      "grad_norm": 0.952081024646759,
+      "learning_rate": 0.000158391959798995,
+      "loss": 0.9464,
+      "step": 424
+    },
+    {
+      "epoch": 0.2125,
+      "grad_norm": 0.5379677414894104,
+      "learning_rate": 0.00015829145728643218,
+      "loss": 0.3873,
+      "step": 425
+    },
+    {
+      "epoch": 0.213,
+      "grad_norm": 0.7126659154891968,
+      "learning_rate": 0.00015819095477386937,
+      "loss": 0.507,
+      "step": 426
+    },
+    {
+      "epoch": 0.2135,
+      "grad_norm": 0.6439905762672424,
+      "learning_rate": 0.00015809045226130656,
+      "loss": 0.4714,
+      "step": 427
+    },
+    {
+      "epoch": 0.214,
+      "grad_norm": 0.8148710131645203,
+      "learning_rate": 0.00015798994974874372,
+      "loss": 0.5019,
+      "step": 428
+    },
+    {
+      "epoch": 0.2145,
+      "grad_norm": 0.6386300921440125,
+      "learning_rate": 0.0001578894472361809,
+      "loss": 0.4777,
+      "step": 429
+    },
+    {
+      "epoch": 0.215,
+      "grad_norm": 0.709976077079773,
+      "learning_rate": 0.0001577889447236181,
+      "loss": 0.4958,
+      "step": 430
+    },
+    {
+      "epoch": 0.2155,
+      "grad_norm": 0.6431065201759338,
+      "learning_rate": 0.00015768844221105528,
+      "loss": 0.4194,
+      "step": 431
+    },
+    {
+      "epoch": 0.216,
+      "grad_norm": 0.6088384985923767,
+      "learning_rate": 0.00015758793969849246,
+      "loss": 0.3718,
+      "step": 432
+    },
+    {
+      "epoch": 0.2165,
+      "grad_norm": 0.715979278087616,
+      "learning_rate": 0.00015748743718592968,
+      "loss": 0.4786,
+      "step": 433
+    },
+    {
+      "epoch": 0.217,
+      "grad_norm": 0.5891948342323303,
+      "learning_rate": 0.00015738693467336686,
+      "loss": 0.3824,
+      "step": 434
+    },
+    {
+      "epoch": 0.2175,
+      "grad_norm": 0.7159863114356995,
+      "learning_rate": 0.00015728643216080402,
+      "loss": 0.5132,
+      "step": 435
+    },
+    {
+      "epoch": 0.218,
+      "grad_norm": 0.7554921507835388,
+      "learning_rate": 0.0001571859296482412,
+      "loss": 0.5579,
+      "step": 436
+    },
+    {
+      "epoch": 0.2185,
+      "grad_norm": 0.8158654570579529,
+      "learning_rate": 0.0001570854271356784,
+      "loss": 0.5902,
+      "step": 437
+    },
+    {
+      "epoch": 0.219,
+      "grad_norm": 0.6420375108718872,
+      "learning_rate": 0.00015698492462311558,
+      "loss": 0.3833,
+      "step": 438
+    },
+    {
+      "epoch": 0.2195,
+      "grad_norm": 0.6936222910881042,
+      "learning_rate": 0.00015688442211055277,
+      "loss": 0.4193,
+      "step": 439
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6749586462974548,
+      "learning_rate": 0.00015678391959798995,
+      "loss": 0.4043,
+      "step": 440
+    },
+    {
+      "epoch": 0.2205,
+      "grad_norm": 0.7509293556213379,
+      "learning_rate": 0.00015668341708542714,
+      "loss": 0.5394,
+      "step": 441
+    },
+    {
+      "epoch": 0.221,
+      "grad_norm": 0.9507200121879578,
+      "learning_rate": 0.00015658291457286433,
+      "loss": 0.4352,
+      "step": 442
+    },
+    {
+      "epoch": 0.2215,
+      "grad_norm": 0.5460655689239502,
+      "learning_rate": 0.0001564824120603015,
+      "loss": 0.2889,
+      "step": 443
+    },
+    {
+      "epoch": 0.222,
+      "grad_norm": 0.7076512575149536,
+      "learning_rate": 0.0001563819095477387,
+      "loss": 0.4019,
+      "step": 444
+    },
+    {
+      "epoch": 0.2225,
+      "grad_norm": 0.7339255213737488,
+      "learning_rate": 0.00015628140703517589,
+      "loss": 0.4112,
+      "step": 445
+    },
+    {
+      "epoch": 0.223,
+      "grad_norm": 0.6644828915596008,
+      "learning_rate": 0.00015618090452261307,
+      "loss": 0.3527,
+      "step": 446
+    },
+    {
+      "epoch": 0.2235,
+      "grad_norm": 0.6336199045181274,
+      "learning_rate": 0.00015608040201005026,
+      "loss": 0.3868,
+      "step": 447
+    },
+    {
+      "epoch": 0.224,
+      "grad_norm": 0.6502320170402527,
+      "learning_rate": 0.00015597989949748745,
+      "loss": 0.3598,
+      "step": 448
+    },
+    {
+      "epoch": 0.2245,
+      "grad_norm": 0.7250773310661316,
+      "learning_rate": 0.00015587939698492463,
+      "loss": 0.3531,
+      "step": 449
+    },
+    {
+      "epoch": 0.225,
+      "grad_norm": 0.7061754465103149,
+      "learning_rate": 0.00015577889447236182,
+      "loss": 0.3828,
+      "step": 450
+    },
+    {
+      "epoch": 0.2255,
+      "grad_norm": 0.5939537286758423,
+      "learning_rate": 0.000155678391959799,
+      "loss": 0.5969,
+      "step": 451
+    },
+    {
+      "epoch": 0.226,
+      "grad_norm": 0.6014315485954285,
+      "learning_rate": 0.0001555778894472362,
+      "loss": 0.556,
+      "step": 452
+    },
+    {
+      "epoch": 0.2265,
+      "grad_norm": 0.6123878359794617,
+      "learning_rate": 0.00015547738693467338,
+      "loss": 0.488,
+      "step": 453
+    },
+    {
+      "epoch": 0.227,
+      "grad_norm": 0.740699291229248,
+      "learning_rate": 0.00015537688442211056,
+      "loss": 0.7585,
+      "step": 454
+    },
+    {
+      "epoch": 0.2275,
+      "grad_norm": 0.6283687353134155,
+      "learning_rate": 0.00015527638190954775,
+      "loss": 0.6057,
+      "step": 455
+    },
+    {
+      "epoch": 0.228,
+      "grad_norm": 0.5816089510917664,
+      "learning_rate": 0.00015517587939698494,
+      "loss": 0.4493,
+      "step": 456
+    },
+    {
+      "epoch": 0.2285,
+      "grad_norm": 0.7151514291763306,
+      "learning_rate": 0.0001550753768844221,
+      "loss": 0.6391,
+      "step": 457
+    },
+    {
+      "epoch": 0.229,
+      "grad_norm": 0.7102727890014648,
+      "learning_rate": 0.0001549748743718593,
+      "loss": 0.6949,
+      "step": 458
+    },
+    {
+      "epoch": 0.2295,
+      "grad_norm": 0.7023615837097168,
+      "learning_rate": 0.0001548743718592965,
+      "loss": 0.6183,
+      "step": 459
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6406300067901611,
+      "learning_rate": 0.00015477386934673368,
+      "loss": 0.4607,
+      "step": 460
+    },
+    {
+      "epoch": 0.2305,
+      "grad_norm": 0.5949312448501587,
+      "learning_rate": 0.00015467336683417087,
+      "loss": 0.4277,
+      "step": 461
+    },
+    {
+      "epoch": 0.231,
+      "grad_norm": 0.6687455773353577,
+      "learning_rate": 0.00015457286432160806,
+      "loss": 0.4933,
+      "step": 462
+    },
+    {
+      "epoch": 0.2315,
+      "grad_norm": 0.6369645595550537,
+      "learning_rate": 0.00015447236180904524,
+      "loss": 0.5078,
+      "step": 463
+    },
+    {
+      "epoch": 0.232,
+      "grad_norm": 0.7452518343925476,
+      "learning_rate": 0.0001543718592964824,
+      "loss": 0.6395,
+      "step": 464
+    },
+    {
+      "epoch": 0.2325,
+      "grad_norm": 0.5993404984474182,
+      "learning_rate": 0.0001542713567839196,
+      "loss": 0.5417,
+      "step": 465
+    },
+    {
+      "epoch": 0.233,
+      "grad_norm": 0.7856594920158386,
+      "learning_rate": 0.0001541708542713568,
+      "loss": 0.7321,
+      "step": 466
+    },
+    {
+      "epoch": 0.2335,
+      "grad_norm": 0.6237269043922424,
+      "learning_rate": 0.000154070351758794,
+      "loss": 0.4503,
+      "step": 467
+    },
+    {
+      "epoch": 0.234,
+      "grad_norm": 0.664532482624054,
+      "learning_rate": 0.00015396984924623117,
+      "loss": 0.4881,
+      "step": 468
+    },
+    {
+      "epoch": 0.2345,
+      "grad_norm": 0.5276694893836975,
+      "learning_rate": 0.00015386934673366836,
+      "loss": 0.3298,
+      "step": 469
+    },
+    {
+      "epoch": 0.235,
+      "grad_norm": 0.7690772414207458,
+      "learning_rate": 0.00015376884422110555,
+      "loss": 0.6312,
+      "step": 470
+    },
+    {
+      "epoch": 0.2355,
+      "grad_norm": 0.6965339779853821,
+      "learning_rate": 0.0001536683417085427,
+      "loss": 0.4933,
+      "step": 471
+    },
+    {
+      "epoch": 0.236,
+      "grad_norm": 0.745960533618927,
+      "learning_rate": 0.0001535678391959799,
+      "loss": 0.536,
+      "step": 472
+    },
+    {
+      "epoch": 0.2365,
+      "grad_norm": 0.9550654888153076,
+      "learning_rate": 0.00015346733668341708,
+      "loss": 0.6102,
+      "step": 473
+    },
+    {
+      "epoch": 0.237,
+      "grad_norm": 0.7985636591911316,
+      "learning_rate": 0.00015336683417085427,
+      "loss": 0.4807,
+      "step": 474
+    },
+    {
+      "epoch": 0.2375,
+      "grad_norm": 0.6231959462165833,
+      "learning_rate": 0.00015326633165829148,
+      "loss": 0.3927,
+      "step": 475
+    },
+    {
+      "epoch": 0.238,
+      "grad_norm": 0.8460021018981934,
+      "learning_rate": 0.00015316582914572867,
+      "loss": 0.5317,
+      "step": 476
+    },
+    {
+      "epoch": 0.2385,
+      "grad_norm": 0.7687966823577881,
+      "learning_rate": 0.00015306532663316583,
+      "loss": 0.4479,
+      "step": 477
+    },
+    {
+      "epoch": 0.239,
+      "grad_norm": 0.7508051991462708,
+      "learning_rate": 0.000152964824120603,
+      "loss": 0.5205,
+      "step": 478
+    },
+    {
+      "epoch": 0.2395,
+      "grad_norm": 0.7614192366600037,
+      "learning_rate": 0.0001528643216080402,
+      "loss": 0.4413,
+      "step": 479
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6554893255233765,
+      "learning_rate": 0.00015276381909547739,
+      "loss": 0.427,
+      "step": 480
+    },
+    {
+      "epoch": 0.2405,
+      "grad_norm": 0.6568048000335693,
+      "learning_rate": 0.00015266331658291457,
+      "loss": 0.454,
+      "step": 481
+    },
+    {
+      "epoch": 0.241,
+      "grad_norm": 0.815615713596344,
+      "learning_rate": 0.00015256281407035176,
+      "loss": 0.6328,
+      "step": 482
+    },
+    {
+      "epoch": 0.2415,
+      "grad_norm": 0.8455066084861755,
+      "learning_rate": 0.00015246231155778897,
+      "loss": 0.434,
+      "step": 483
+    },
+    {
+      "epoch": 0.242,
+      "grad_norm": 0.7144169211387634,
+      "learning_rate": 0.00015236180904522613,
+      "loss": 0.5158,
+      "step": 484
+    },
+    {
+      "epoch": 0.2425,
+      "grad_norm": 0.7430387735366821,
+      "learning_rate": 0.00015226130653266332,
+      "loss": 0.4375,
+      "step": 485
+    },
+    {
+      "epoch": 0.243,
+      "grad_norm": 0.6197020411491394,
+      "learning_rate": 0.0001521608040201005,
+      "loss": 0.3327,
+      "step": 486
+    },
+    {
+      "epoch": 0.2435,
+      "grad_norm": 0.8677177429199219,
+      "learning_rate": 0.0001520603015075377,
+      "loss": 0.4224,
+      "step": 487
+    },
+    {
+      "epoch": 0.244,
+      "grad_norm": 0.7367002367973328,
+      "learning_rate": 0.00015195979899497488,
+      "loss": 0.4811,
+      "step": 488
+    },
+    {
+      "epoch": 0.2445,
+      "grad_norm": 0.7841548323631287,
+      "learning_rate": 0.00015185929648241206,
+      "loss": 0.4686,
+      "step": 489
+    },
+    {
+      "epoch": 0.245,
+      "grad_norm": 0.8038194179534912,
+      "learning_rate": 0.00015175879396984925,
+      "loss": 0.571,
+      "step": 490
+    },
+    {
+      "epoch": 0.2455,
+      "grad_norm": 0.5407405495643616,
+      "learning_rate": 0.00015165829145728644,
+      "loss": 0.2954,
+      "step": 491
+    },
+    {
+      "epoch": 0.246,
+      "grad_norm": 0.6732304096221924,
+      "learning_rate": 0.00015155778894472362,
+      "loss": 0.385,
+      "step": 492
+    },
+    {
+      "epoch": 0.2465,
+      "grad_norm": 0.7976018786430359,
+      "learning_rate": 0.0001514572864321608,
+      "loss": 0.489,
+      "step": 493
+    },
+    {
+      "epoch": 0.247,
+      "grad_norm": 0.6538392901420593,
+      "learning_rate": 0.000151356783919598,
+      "loss": 0.3371,
+      "step": 494
+    },
+    {
+      "epoch": 0.2475,
+      "grad_norm": 0.6547286510467529,
+      "learning_rate": 0.00015125628140703518,
+      "loss": 0.3885,
+      "step": 495
+    },
+    {
+      "epoch": 0.248,
+      "grad_norm": 1.0107066631317139,
+      "learning_rate": 0.00015115577889447237,
+      "loss": 0.4436,
+      "step": 496
+    },
+    {
+      "epoch": 0.2485,
+      "grad_norm": 0.678554892539978,
+      "learning_rate": 0.00015105527638190956,
+      "loss": 0.3546,
+      "step": 497
+    },
+    {
+      "epoch": 0.249,
+      "grad_norm": 0.5999384522438049,
+      "learning_rate": 0.00015095477386934674,
+      "loss": 0.29,
+      "step": 498
+    },
+    {
+      "epoch": 0.2495,
+      "grad_norm": 0.8595788478851318,
+      "learning_rate": 0.00015085427135678393,
+      "loss": 0.4862,
+      "step": 499
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6345055103302002,
+      "learning_rate": 0.00015075376884422112,
+      "loss": 0.3281,
+      "step": 500
+    },
+    {
+      "epoch": 0.2505,
+      "grad_norm": 0.6969207525253296,
+      "learning_rate": 0.0001506532663316583,
+      "loss": 0.8152,
+      "step": 501
+    },
+    {
+      "epoch": 0.251,
+      "grad_norm": 0.6502382755279541,
+      "learning_rate": 0.0001505527638190955,
+      "loss": 0.4541,
+      "step": 502
+    },
+    {
+      "epoch": 0.2515,
+      "grad_norm": 0.5796345472335815,
+      "learning_rate": 0.00015045226130653267,
+      "loss": 0.487,
+      "step": 503
+    },
+    {
+      "epoch": 0.252,
+      "grad_norm": 0.6514685750007629,
+      "learning_rate": 0.00015035175879396986,
+      "loss": 0.502,
+      "step": 504
+    },
+    {
+      "epoch": 0.2525,
+      "grad_norm": 0.7342237234115601,
+      "learning_rate": 0.00015025125628140705,
+      "loss": 0.6062,
+      "step": 505
+    },
+    {
+      "epoch": 0.253,
+      "grad_norm": 0.6466059684753418,
+      "learning_rate": 0.00015015075376884423,
+      "loss": 0.5351,
+      "step": 506
+    },
+    {
+      "epoch": 0.2535,
+      "grad_norm": 0.7463874220848083,
+      "learning_rate": 0.0001500502512562814,
+      "loss": 0.6853,
+      "step": 507
+    },
+    {
+      "epoch": 0.254,
+      "grad_norm": 0.7068241238594055,
+      "learning_rate": 0.0001499497487437186,
+      "loss": 0.5303,
+      "step": 508
+    },
+    {
+      "epoch": 0.2545,
+      "grad_norm": 0.8023015260696411,
+      "learning_rate": 0.0001498492462311558,
+      "loss": 0.5227,
+      "step": 509
+    },
+    {
+      "epoch": 0.255,
+      "grad_norm": 0.6587471961975098,
+      "learning_rate": 0.00014974874371859298,
+      "loss": 0.5734,
+      "step": 510
+    },
+    {
+      "epoch": 0.2555,
+      "grad_norm": 0.7357711791992188,
+      "learning_rate": 0.00014964824120603017,
+      "loss": 0.5802,
+      "step": 511
+    },
+    {
+      "epoch": 0.256,
+      "grad_norm": 0.6654890775680542,
+      "learning_rate": 0.00014954773869346735,
+      "loss": 0.4874,
+      "step": 512
+    },
+    {
+      "epoch": 0.2565,
+      "grad_norm": 0.6203697919845581,
+      "learning_rate": 0.00014944723618090454,
+      "loss": 0.4442,
+      "step": 513
+    },
+    {
+      "epoch": 0.257,
+      "grad_norm": 0.8192847967147827,
+      "learning_rate": 0.0001493467336683417,
+      "loss": 0.6098,
+      "step": 514
+    },
+    {
+      "epoch": 0.2575,
+      "grad_norm": 0.6767122149467468,
+      "learning_rate": 0.00014924623115577889,
+      "loss": 0.5199,
+      "step": 515
+    },
+    {
+      "epoch": 0.258,
+      "grad_norm": 0.7172583341598511,
+      "learning_rate": 0.0001491457286432161,
+      "loss": 0.551,
+      "step": 516
+    },
+    {
+      "epoch": 0.2585,
+      "grad_norm": 0.6525209546089172,
+      "learning_rate": 0.00014904522613065329,
+      "loss": 0.5102,
+      "step": 517
+    },
+    {
+      "epoch": 0.259,
+      "grad_norm": 0.6344095468521118,
+      "learning_rate": 0.00014894472361809047,
+      "loss": 0.4957,
+      "step": 518
+    },
+    {
+      "epoch": 0.2595,
+      "grad_norm": 0.6586654186248779,
+      "learning_rate": 0.00014884422110552766,
+      "loss": 0.4918,
+      "step": 519
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6742819547653198,
+      "learning_rate": 0.00014874371859296482,
+      "loss": 0.4332,
+      "step": 520
+    },
+    {
+      "epoch": 0.2605,
+      "grad_norm": 0.6519845128059387,
+      "learning_rate": 0.000148643216080402,
+      "loss": 0.5,
+      "step": 521
+    },
+    {
+      "epoch": 0.261,
+      "grad_norm": 0.7158078551292419,
+      "learning_rate": 0.0001485427135678392,
+      "loss": 0.4842,
+      "step": 522
+    },
+    {
+      "epoch": 0.2615,
+      "grad_norm": 0.6570281386375427,
+      "learning_rate": 0.00014844221105527638,
+      "loss": 0.4664,
+      "step": 523
+    },
+    {
+      "epoch": 0.262,
+      "grad_norm": 0.668930172920227,
+      "learning_rate": 0.00014834170854271356,
+      "loss": 0.4051,
+      "step": 524
+    },
+    {
+      "epoch": 0.2625,
+      "grad_norm": 0.6977079510688782,
+      "learning_rate": 0.00014824120603015078,
+      "loss": 0.5306,
+      "step": 525
+    },
+    {
+      "epoch": 0.263,
+      "grad_norm": 0.6891977190971375,
+      "learning_rate": 0.00014814070351758796,
+      "loss": 0.4944,
+      "step": 526
+    },
+    {
+      "epoch": 0.2635,
+      "grad_norm": 0.6917031407356262,
+      "learning_rate": 0.00014804020100502512,
+      "loss": 0.4367,
+      "step": 527
+    },
+    {
+      "epoch": 0.264,
+      "grad_norm": 0.7421239614486694,
+      "learning_rate": 0.0001479396984924623,
+      "loss": 0.5103,
+      "step": 528
+    },
+    {
+      "epoch": 0.2645,
+      "grad_norm": 0.8465878963470459,
+      "learning_rate": 0.0001478391959798995,
+      "loss": 0.5728,
+      "step": 529
+    },
+    {
+      "epoch": 0.265,
+      "grad_norm": 0.8142993450164795,
+      "learning_rate": 0.00014773869346733668,
+      "loss": 0.6326,
+      "step": 530
+    },
+    {
+      "epoch": 0.2655,
+      "grad_norm": 0.6576914191246033,
+      "learning_rate": 0.00014763819095477387,
+      "loss": 0.4805,
+      "step": 531
+    },
+    {
+      "epoch": 0.266,
+      "grad_norm": 0.6926950812339783,
+      "learning_rate": 0.00014753768844221106,
+      "loss": 0.4588,
+      "step": 532
+    },
+    {
+      "epoch": 0.2665,
+      "grad_norm": 0.6741633415222168,
+      "learning_rate": 0.00014743718592964827,
+      "loss": 0.3791,
+      "step": 533
+    },
+    {
+      "epoch": 0.267,
+      "grad_norm": 0.7339944243431091,
+      "learning_rate": 0.00014733668341708543,
+      "loss": 0.4235,
+      "step": 534
+    },
+    {
+      "epoch": 0.2675,
+      "grad_norm": 0.8096172213554382,
+      "learning_rate": 0.00014723618090452262,
+      "loss": 0.4893,
+      "step": 535
+    },
+    {
+      "epoch": 0.268,
+      "grad_norm": 0.7587369680404663,
+      "learning_rate": 0.0001471356783919598,
+      "loss": 0.5256,
+      "step": 536
+    },
+    {
+      "epoch": 0.2685,
+      "grad_norm": 0.5496168732643127,
+      "learning_rate": 0.000147035175879397,
+      "loss": 0.3748,
+      "step": 537
+    },
+    {
+      "epoch": 0.269,
+      "grad_norm": 0.673461377620697,
+      "learning_rate": 0.00014693467336683417,
+      "loss": 0.3851,
+      "step": 538
+    },
+    {
+      "epoch": 0.2695,
+      "grad_norm": 0.7310826778411865,
+      "learning_rate": 0.00014683417085427136,
+      "loss": 0.554,
+      "step": 539
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6193651556968689,
+      "learning_rate": 0.00014673366834170855,
+      "loss": 0.4016,
+      "step": 540
+    },
+    {
+      "epoch": 0.2705,
+      "grad_norm": 0.6466208696365356,
+      "learning_rate": 0.00014663316582914573,
+      "loss": 0.4439,
+      "step": 541
+    },
+    {
+      "epoch": 0.271,
+      "grad_norm": 0.5416639447212219,
+      "learning_rate": 0.00014653266331658292,
+      "loss": 0.3109,
+      "step": 542
+    },
+    {
+      "epoch": 0.2715,
+      "grad_norm": 0.6309217810630798,
+      "learning_rate": 0.0001464321608040201,
+      "loss": 0.3493,
+      "step": 543
+    },
+    {
+      "epoch": 0.272,
+      "grad_norm": 0.6398806571960449,
+      "learning_rate": 0.0001463316582914573,
+      "loss": 0.3968,
+      "step": 544
+    },
+    {
+      "epoch": 0.2725,
+      "grad_norm": 0.7098992466926575,
+      "learning_rate": 0.00014623115577889448,
+      "loss": 0.4022,
+      "step": 545
+    },
+    {
+      "epoch": 0.273,
+      "grad_norm": 0.620333731174469,
+      "learning_rate": 0.00014613065326633167,
+      "loss": 0.3054,
+      "step": 546
+    },
+    {
+      "epoch": 0.2735,
+      "grad_norm": 0.6589905619621277,
+      "learning_rate": 0.00014603015075376885,
+      "loss": 0.3363,
+      "step": 547
+    },
+    {
+      "epoch": 0.274,
+      "grad_norm": 0.7924469709396362,
+      "learning_rate": 0.00014592964824120604,
+      "loss": 0.4354,
+      "step": 548
+    },
+    {
+      "epoch": 0.2745,
+      "grad_norm": 0.7733848690986633,
+      "learning_rate": 0.00014582914572864323,
+      "loss": 0.4251,
+      "step": 549
+    },
+    {
+      "epoch": 0.275,
+      "grad_norm": 0.6761373281478882,
+      "learning_rate": 0.0001457286432160804,
+      "loss": 0.3261,
+      "step": 550
+    },
+    {
+      "epoch": 0.2755,
+      "grad_norm": 0.7072325944900513,
+      "learning_rate": 0.0001456281407035176,
+      "loss": 0.6651,
+      "step": 551
+    },
+    {
+      "epoch": 0.276,
+      "grad_norm": 0.6927074790000916,
+      "learning_rate": 0.00014552763819095479,
+      "loss": 0.5225,
+      "step": 552
+    },
+    {
+      "epoch": 0.2765,
+      "grad_norm": 0.7367109060287476,
+      "learning_rate": 0.00014542713567839197,
+      "loss": 0.6584,
+      "step": 553
+    },
+    {
+      "epoch": 0.277,
+      "grad_norm": 0.6004040241241455,
+      "learning_rate": 0.00014532663316582916,
+      "loss": 0.4373,
+      "step": 554
+    },
+    {
+      "epoch": 0.2775,
+      "grad_norm": 0.6807554960250854,
+      "learning_rate": 0.00014522613065326634,
+      "loss": 0.5658,
+      "step": 555
+    },
+    {
+      "epoch": 0.278,
+      "grad_norm": 0.6944225430488586,
+      "learning_rate": 0.00014512562814070353,
+      "loss": 0.5686,
+      "step": 556
+    },
+    {
+      "epoch": 0.2785,
+      "grad_norm": 0.685875654220581,
+      "learning_rate": 0.0001450251256281407,
+      "loss": 0.5658,
+      "step": 557
+    },
+    {
+      "epoch": 0.279,
+      "grad_norm": 0.7746342420578003,
+      "learning_rate": 0.0001449246231155779,
+      "loss": 0.6185,
+      "step": 558
+    },
+    {
+      "epoch": 0.2795,
+      "grad_norm": 0.6917704939842224,
+      "learning_rate": 0.0001448241206030151,
+      "loss": 0.5502,
+      "step": 559
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5866712331771851,
+      "learning_rate": 0.00014472361809045228,
+      "loss": 0.405,
+      "step": 560
+    },
+    {
+      "epoch": 0.2805,
+      "grad_norm": 0.630342960357666,
+      "learning_rate": 0.00014462311557788946,
+      "loss": 0.456,
+      "step": 561
+    },
+    {
+      "epoch": 0.281,
+      "grad_norm": 0.7325962781906128,
+      "learning_rate": 0.00014452261306532665,
+      "loss": 0.5823,
+      "step": 562
+    },
+    {
+      "epoch": 0.2815,
+      "grad_norm": 0.6631718277931213,
+      "learning_rate": 0.0001444221105527638,
+      "loss": 0.5188,
+      "step": 563
+    },
+    {
+      "epoch": 0.282,
+      "grad_norm": 0.6104342341423035,
+      "learning_rate": 0.000144321608040201,
+      "loss": 0.4024,
+      "step": 564
+    },
+    {
+      "epoch": 0.2825,
+      "grad_norm": 0.7061834335327148,
+      "learning_rate": 0.00014422110552763818,
+      "loss": 0.5988,
+      "step": 565
+    },
+    {
+      "epoch": 0.283,
+      "grad_norm": 0.6309420466423035,
+      "learning_rate": 0.00014412060301507537,
+      "loss": 0.4114,
+      "step": 566
+    },
+    {
+      "epoch": 0.2835,
+      "grad_norm": 0.6901776790618896,
+      "learning_rate": 0.00014402010050251258,
+      "loss": 0.5477,
+      "step": 567
+    },
+    {
+      "epoch": 0.284,
+      "grad_norm": 0.7117554545402527,
+      "learning_rate": 0.00014391959798994977,
+      "loss": 0.4933,
+      "step": 568
+    },
+    {
+      "epoch": 0.2845,
+      "grad_norm": 0.6800329089164734,
+      "learning_rate": 0.00014381909547738696,
+      "loss": 0.4624,
+      "step": 569
+    },
+    {
+      "epoch": 0.285,
+      "grad_norm": 0.7547156810760498,
+      "learning_rate": 0.00014371859296482411,
+      "loss": 0.635,
+      "step": 570
+    },
+    {
+      "epoch": 0.2855,
+      "grad_norm": 0.718701183795929,
+      "learning_rate": 0.0001436180904522613,
+      "loss": 0.4968,
+      "step": 571
+    },
+    {
+      "epoch": 0.286,
+      "grad_norm": 0.6961089968681335,
+      "learning_rate": 0.0001435175879396985,
+      "loss": 0.5344,
+      "step": 572
+    },
+    {
+      "epoch": 0.2865,
+      "grad_norm": 0.5895490646362305,
+      "learning_rate": 0.00014341708542713567,
+      "loss": 0.3577,
+      "step": 573
+    },
+    {
+      "epoch": 0.287,
+      "grad_norm": 0.6628783941268921,
+      "learning_rate": 0.00014331658291457286,
+      "loss": 0.442,
+      "step": 574
+    },
+    {
+      "epoch": 0.2875,
+      "grad_norm": 0.6606783866882324,
+      "learning_rate": 0.00014321608040201007,
+      "loss": 0.4569,
+      "step": 575
+    },
+    {
+      "epoch": 0.288,
+      "grad_norm": 0.7671996355056763,
+      "learning_rate": 0.00014311557788944726,
+      "loss": 0.5128,
+      "step": 576
+    },
+    {
+      "epoch": 0.2885,
+      "grad_norm": 0.720435380935669,
+      "learning_rate": 0.00014301507537688442,
+      "loss": 0.3899,
+      "step": 577
+    },
+    {
+      "epoch": 0.289,
+      "grad_norm": 0.783862829208374,
+      "learning_rate": 0.0001429145728643216,
+      "loss": 0.5848,
+      "step": 578
+    },
+    {
+      "epoch": 0.2895,
+      "grad_norm": 0.5570368766784668,
+      "learning_rate": 0.0001428140703517588,
+      "loss": 0.3292,
+      "step": 579
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6207924485206604,
+      "learning_rate": 0.00014271356783919598,
+      "loss": 0.3673,
+      "step": 580
+    },
+    {
+      "epoch": 0.2905,
+      "grad_norm": 0.6351432800292969,
+      "learning_rate": 0.00014261306532663317,
+      "loss": 0.3525,
+      "step": 581
+    },
+    {
+      "epoch": 0.291,
+      "grad_norm": 0.7852205634117126,
+      "learning_rate": 0.00014251256281407035,
+      "loss": 0.5761,
+      "step": 582
+    },
+    {
+      "epoch": 0.2915,
+      "grad_norm": 0.7996278405189514,
+      "learning_rate": 0.00014241206030150757,
+      "loss": 0.4439,
+      "step": 583
+    },
+    {
+      "epoch": 0.292,
+      "grad_norm": 0.6674121022224426,
+      "learning_rate": 0.00014231155778894473,
+      "loss": 0.4504,
+      "step": 584
+    },
+    {
+      "epoch": 0.2925,
+      "grad_norm": 0.6028682589530945,
+      "learning_rate": 0.0001422110552763819,
+      "loss": 0.3496,
+      "step": 585
+    },
+    {
+      "epoch": 0.293,
+      "grad_norm": 0.6133868098258972,
+      "learning_rate": 0.0001421105527638191,
+      "loss": 0.3939,
+      "step": 586
+    },
+    {
+      "epoch": 0.2935,
+      "grad_norm": 0.8932775855064392,
+      "learning_rate": 0.00014201005025125629,
+      "loss": 0.4891,
+      "step": 587
+    },
+    {
+      "epoch": 0.294,
+      "grad_norm": 0.7422942519187927,
+      "learning_rate": 0.00014190954773869347,
+      "loss": 0.4427,
+      "step": 588
+    },
+    {
+      "epoch": 0.2945,
+      "grad_norm": 0.70126873254776,
+      "learning_rate": 0.00014180904522613066,
+      "loss": 0.4527,
+      "step": 589
+    },
+    {
+      "epoch": 0.295,
+      "grad_norm": 0.7380238771438599,
+      "learning_rate": 0.00014170854271356784,
+      "loss": 0.4972,
+      "step": 590
+    },
+    {
+      "epoch": 0.2955,
+      "grad_norm": 0.7120311856269836,
+      "learning_rate": 0.00014160804020100503,
+      "loss": 0.3918,
+      "step": 591
+    },
+    {
+      "epoch": 0.296,
+      "grad_norm": 0.5820732116699219,
+      "learning_rate": 0.00014150753768844222,
+      "loss": 0.3135,
+      "step": 592
+    },
+    {
+      "epoch": 0.2965,
+      "grad_norm": 0.6390547156333923,
+      "learning_rate": 0.0001414070351758794,
+      "loss": 0.3355,
+      "step": 593
+    },
+    {
+      "epoch": 0.297,
+      "grad_norm": 0.6904809474945068,
+      "learning_rate": 0.0001413065326633166,
+      "loss": 0.4079,
+      "step": 594
+    },
+    {
+      "epoch": 0.2975,
+      "grad_norm": 0.8338470458984375,
+      "learning_rate": 0.00014120603015075378,
+      "loss": 0.4803,
+      "step": 595
+    },
+    {
+      "epoch": 0.298,
+      "grad_norm": 0.7458425760269165,
+      "learning_rate": 0.00014110552763819096,
+      "loss": 0.38,
+      "step": 596
+    },
+    {
+      "epoch": 0.2985,
+      "grad_norm": 0.7649307250976562,
+      "learning_rate": 0.00014100502512562815,
+      "loss": 0.3665,
+      "step": 597
+    },
+    {
+      "epoch": 0.299,
+      "grad_norm": 0.6732955574989319,
+      "learning_rate": 0.00014090452261306534,
+      "loss": 0.3766,
+      "step": 598
+    },
+    {
+      "epoch": 0.2995,
+      "grad_norm": 0.6341168880462646,
+      "learning_rate": 0.00014080402010050252,
+      "loss": 0.3749,
+      "step": 599
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6906387209892273,
+      "learning_rate": 0.0001407035175879397,
+      "loss": 0.3339,
+      "step": 600
+    },
+    {
+      "epoch": 0.3005,
+      "grad_norm": 0.6381351947784424,
+      "learning_rate": 0.0001406030150753769,
+      "loss": 0.6461,
+      "step": 601
+    },
+    {
+      "epoch": 0.301,
+      "grad_norm": 0.5286738872528076,
+      "learning_rate": 0.00014050251256281408,
+      "loss": 0.4357,
+      "step": 602
+    },
+    {
+      "epoch": 0.3015,
+      "grad_norm": 0.6979527473449707,
+      "learning_rate": 0.00014040201005025127,
+      "loss": 0.5872,
+      "step": 603
+    },
+    {
+      "epoch": 0.302,
+      "grad_norm": 0.6415486335754395,
+      "learning_rate": 0.00014030150753768846,
+      "loss": 0.4808,
+      "step": 604
+    },
+    {
+      "epoch": 0.3025,
+      "grad_norm": 0.7876259684562683,
+      "learning_rate": 0.00014020100502512564,
+      "loss": 0.5757,
+      "step": 605
+    },
+    {
+      "epoch": 0.303,
+      "grad_norm": 0.6682291626930237,
+      "learning_rate": 0.0001401005025125628,
+      "loss": 0.4953,
+      "step": 606
+    },
+    {
+      "epoch": 0.3035,
+      "grad_norm": 0.6934815049171448,
+      "learning_rate": 0.00014,
+      "loss": 0.5974,
+      "step": 607
+    },
+    {
+      "epoch": 0.304,
+      "grad_norm": 0.7640621662139893,
+      "learning_rate": 0.0001398994974874372,
+      "loss": 0.5136,
+      "step": 608
+    },
+    {
+      "epoch": 0.3045,
+      "grad_norm": 0.7184918522834778,
+      "learning_rate": 0.0001397989949748744,
+      "loss": 0.587,
+      "step": 609
+    },
+    {
+      "epoch": 0.305,
+      "grad_norm": 0.7544426321983337,
+      "learning_rate": 0.00013969849246231157,
+      "loss": 0.5078,
+      "step": 610
+    },
+    {
+      "epoch": 0.3055,
+      "grad_norm": 0.6475300192832947,
+      "learning_rate": 0.00013959798994974876,
+      "loss": 0.5699,
+      "step": 611
+    },
+    {
+      "epoch": 0.306,
+      "grad_norm": 0.6399070620536804,
+      "learning_rate": 0.00013949748743718595,
+      "loss": 0.4793,
+      "step": 612
+    },
+    {
+      "epoch": 0.3065,
+      "grad_norm": 0.6058080792427063,
+      "learning_rate": 0.0001393969849246231,
+      "loss": 0.3834,
+      "step": 613
+    },
+    {
+      "epoch": 0.307,
+      "grad_norm": 0.7285460829734802,
+      "learning_rate": 0.0001392964824120603,
+      "loss": 0.5604,
+      "step": 614
+    },
+    {
+      "epoch": 0.3075,
+      "grad_norm": 0.6031981706619263,
+      "learning_rate": 0.00013919597989949748,
+      "loss": 0.4131,
+      "step": 615
+    },
+    {
+      "epoch": 0.308,
+      "grad_norm": 0.6865850687026978,
+      "learning_rate": 0.00013909547738693467,
+      "loss": 0.4833,
+      "step": 616
+    },
+    {
+      "epoch": 0.3085,
+      "grad_norm": 0.5910267233848572,
+      "learning_rate": 0.00013899497487437188,
+      "loss": 0.449,
+      "step": 617
+    },
+    {
+      "epoch": 0.309,
+      "grad_norm": 0.7655642032623291,
+      "learning_rate": 0.00013889447236180907,
+      "loss": 0.5358,
+      "step": 618
+    },
+    {
+      "epoch": 0.3095,
+      "grad_norm": 0.7173675298690796,
+      "learning_rate": 0.00013879396984924625,
+      "loss": 0.5345,
+      "step": 619
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6728115677833557,
+      "learning_rate": 0.0001386934673366834,
+      "loss": 0.4891,
+      "step": 620
+    },
+    {
+      "epoch": 0.3105,
+      "grad_norm": 0.7066881656646729,
+      "learning_rate": 0.0001385929648241206,
+      "loss": 0.4769,
+      "step": 621
+    },
+    {
+      "epoch": 0.311,
+      "grad_norm": 0.7541621327400208,
+      "learning_rate": 0.00013849246231155778,
+      "loss": 0.5441,
+      "step": 622
+    },
+    {
+      "epoch": 0.3115,
+      "grad_norm": 0.7071688771247864,
+      "learning_rate": 0.00013839195979899497,
+      "loss": 0.5814,
+      "step": 623
+    },
+    {
+      "epoch": 0.312,
+      "grad_norm": 0.7858960032463074,
+      "learning_rate": 0.00013829145728643216,
+      "loss": 0.5657,
+      "step": 624
+    },
+    {
+      "epoch": 0.3125,
+      "grad_norm": 0.6961463093757629,
+      "learning_rate": 0.00013819095477386937,
+      "loss": 0.4398,
+      "step": 625
+    },
+    {
+      "epoch": 0.313,
+      "grad_norm": 0.7086874842643738,
+      "learning_rate": 0.00013809045226130656,
+      "loss": 0.503,
+      "step": 626
+    },
+    {
+      "epoch": 0.3135,
+      "grad_norm": 0.689068615436554,
+      "learning_rate": 0.00013798994974874372,
+      "loss": 0.45,
+      "step": 627
+    },
+    {
+      "epoch": 0.314,
+      "grad_norm": 0.6770050525665283,
+      "learning_rate": 0.0001378894472361809,
+      "loss": 0.4117,
+      "step": 628
+    },
+    {
+      "epoch": 0.3145,
+      "grad_norm": 0.597040057182312,
+      "learning_rate": 0.0001377889447236181,
+      "loss": 0.4329,
+      "step": 629
+    },
+    {
+      "epoch": 0.315,
+      "grad_norm": 0.7402741312980652,
+      "learning_rate": 0.00013768844221105528,
+      "loss": 0.5511,
+      "step": 630
+    },
+    {
+      "epoch": 0.3155,
+      "grad_norm": 0.6317721009254456,
+      "learning_rate": 0.00013758793969849246,
+      "loss": 0.4625,
+      "step": 631
+    },
+    {
+      "epoch": 0.316,
+      "grad_norm": 0.740571141242981,
+      "learning_rate": 0.00013748743718592965,
+      "loss": 0.563,
+      "step": 632
+    },
+    {
+      "epoch": 0.3165,
+      "grad_norm": 0.6318831443786621,
+      "learning_rate": 0.00013738693467336686,
+      "loss": 0.4378,
+      "step": 633
+    },
+    {
+      "epoch": 0.317,
+      "grad_norm": 0.7050504088401794,
+      "learning_rate": 0.00013728643216080402,
+      "loss": 0.4075,
+      "step": 634
+    },
+    {
+      "epoch": 0.3175,
+      "grad_norm": 0.7124577164649963,
+      "learning_rate": 0.0001371859296482412,
+      "loss": 0.4653,
+      "step": 635
+    },
+    {
+      "epoch": 0.318,
+      "grad_norm": 0.7688040733337402,
+      "learning_rate": 0.0001370854271356784,
+      "loss": 0.507,
+      "step": 636
+    },
+    {
+      "epoch": 0.3185,
+      "grad_norm": 0.695570170879364,
+      "learning_rate": 0.00013698492462311558,
+      "loss": 0.3921,
+      "step": 637
+    },
+    {
+      "epoch": 0.319,
+      "grad_norm": 0.8030912280082703,
+      "learning_rate": 0.00013688442211055277,
+      "loss": 0.5324,
+      "step": 638
+    },
+    {
+      "epoch": 0.3195,
+      "grad_norm": 0.7891384363174438,
+      "learning_rate": 0.00013678391959798996,
+      "loss": 0.4917,
+      "step": 639
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6823398470878601,
+      "learning_rate": 0.00013668341708542714,
+      "loss": 0.465,
+      "step": 640
+    },
+    {
+      "epoch": 0.3205,
+      "grad_norm": 0.7522284388542175,
+      "learning_rate": 0.00013658291457286433,
+      "loss": 0.558,
+      "step": 641
+    },
+    {
+      "epoch": 0.321,
+      "grad_norm": 0.6949052810668945,
+      "learning_rate": 0.00013648241206030151,
+      "loss": 0.4192,
+      "step": 642
+    },
+    {
+      "epoch": 0.3215,
+      "grad_norm": 0.7988911867141724,
+      "learning_rate": 0.0001363819095477387,
+      "loss": 0.4899,
+      "step": 643
+    },
+    {
+      "epoch": 0.322,
+      "grad_norm": 0.6017119288444519,
+      "learning_rate": 0.0001362814070351759,
+      "loss": 0.3431,
+      "step": 644
+    },
+    {
+      "epoch": 0.3225,
+      "grad_norm": 0.6157026886940002,
+      "learning_rate": 0.00013618090452261307,
+      "loss": 0.3228,
+      "step": 645
+    },
+    {
+      "epoch": 0.323,
+      "grad_norm": 0.6884551644325256,
+      "learning_rate": 0.00013608040201005026,
+      "loss": 0.3733,
+      "step": 646
+    },
+    {
+      "epoch": 0.3235,
+      "grad_norm": 0.7864928245544434,
+      "learning_rate": 0.00013597989949748745,
+      "loss": 0.4991,
+      "step": 647
+    },
+    {
+      "epoch": 0.324,
+      "grad_norm": 0.6312656402587891,
+      "learning_rate": 0.00013587939698492463,
+      "loss": 0.3908,
+      "step": 648
+    },
+    {
+      "epoch": 0.3245,
+      "grad_norm": 0.6704145669937134,
+      "learning_rate": 0.0001357788944723618,
+      "loss": 0.3807,
+      "step": 649
+    },
+    {
+      "epoch": 0.325,
+      "grad_norm": 0.5958676338195801,
+      "learning_rate": 0.000135678391959799,
+      "loss": 0.2863,
+      "step": 650
+    },
+    {
+      "epoch": 0.3255,
+      "grad_norm": 0.661532461643219,
+      "learning_rate": 0.0001355778894472362,
+      "loss": 0.5436,
+      "step": 651
+    },
+    {
+      "epoch": 0.326,
+      "grad_norm": 0.4808809757232666,
+      "learning_rate": 0.00013547738693467338,
+      "loss": 0.424,
+      "step": 652
+    },
+    {
+      "epoch": 0.3265,
+      "grad_norm": 0.8177589178085327,
+      "learning_rate": 0.00013537688442211057,
+      "loss": 0.8762,
+      "step": 653
+    },
+    {
+      "epoch": 0.327,
+      "grad_norm": 0.5700751543045044,
+      "learning_rate": 0.00013527638190954775,
+      "loss": 0.4943,
+      "step": 654
+    },
+    {
+      "epoch": 0.3275,
+      "grad_norm": 0.6347500085830688,
+      "learning_rate": 0.00013517587939698494,
+      "loss": 0.5179,
+      "step": 655
+    },
+    {
+      "epoch": 0.328,
+      "grad_norm": 0.5706446766853333,
+      "learning_rate": 0.0001350753768844221,
+      "loss": 0.4358,
+      "step": 656
+    },
+    {
+      "epoch": 0.3285,
+      "grad_norm": 0.6627894043922424,
+      "learning_rate": 0.00013497487437185928,
+      "loss": 0.4683,
+      "step": 657
+    },
+    {
+      "epoch": 0.329,
+      "grad_norm": 0.6122452020645142,
+      "learning_rate": 0.00013487437185929647,
+      "loss": 0.6253,
+      "step": 658
+    },
+    {
+      "epoch": 0.3295,
+      "grad_norm": 0.6348839998245239,
+      "learning_rate": 0.00013477386934673368,
+      "loss": 0.455,
+      "step": 659
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6455675959587097,
+      "learning_rate": 0.00013467336683417087,
+      "loss": 0.5343,
+      "step": 660
+    },
+    {
+      "epoch": 0.3305,
+      "grad_norm": 0.6887580752372742,
+      "learning_rate": 0.00013457286432160806,
+      "loss": 0.5367,
+      "step": 661
+    },
+    {
+      "epoch": 0.331,
+      "grad_norm": 0.6344459652900696,
+      "learning_rate": 0.00013447236180904524,
+      "loss": 0.4823,
+      "step": 662
+    },
+    {
+      "epoch": 0.3315,
+      "grad_norm": 0.6811181902885437,
+      "learning_rate": 0.0001343718592964824,
+      "loss": 0.4992,
+      "step": 663
+    },
+    {
+      "epoch": 0.332,
+      "grad_norm": 0.6942533254623413,
+      "learning_rate": 0.0001342713567839196,
+      "loss": 0.5113,
+      "step": 664
+    },
+    {
+      "epoch": 0.3325,
+      "grad_norm": 0.6795490980148315,
+      "learning_rate": 0.00013417085427135678,
+      "loss": 0.5383,
+      "step": 665
+    },
+    {
+      "epoch": 0.333,
+      "grad_norm": 0.7495938539505005,
+      "learning_rate": 0.00013407035175879396,
+      "loss": 0.6548,
+      "step": 666
+    },
+    {
+      "epoch": 0.3335,
+      "grad_norm": 0.6919872164726257,
+      "learning_rate": 0.00013396984924623118,
+      "loss": 0.4895,
+      "step": 667
+    },
+    {
+      "epoch": 0.334,
+      "grad_norm": 0.6273533701896667,
+      "learning_rate": 0.00013386934673366836,
+      "loss": 0.5156,
+      "step": 668
+    },
+    {
+      "epoch": 0.3345,
+      "grad_norm": 0.7559247612953186,
+      "learning_rate": 0.00013376884422110555,
+      "loss": 0.6158,
+      "step": 669
+    },
+    {
+      "epoch": 0.335,
+      "grad_norm": 0.5783503651618958,
+      "learning_rate": 0.0001336683417085427,
+      "loss": 0.4315,
+      "step": 670
+    },
+    {
+      "epoch": 0.3355,
+      "grad_norm": 0.6685211658477783,
+      "learning_rate": 0.0001335678391959799,
+      "loss": 0.436,
+      "step": 671
+    },
+    {
+      "epoch": 0.336,
+      "grad_norm": 0.7402259707450867,
+      "learning_rate": 0.00013346733668341708,
+      "loss": 0.5645,
+      "step": 672
+    },
+    {
+      "epoch": 0.3365,
+      "grad_norm": 0.6705771088600159,
+      "learning_rate": 0.00013336683417085427,
+      "loss": 0.4795,
+      "step": 673
+    },
+    {
+      "epoch": 0.337,
+      "grad_norm": 0.6640343070030212,
+      "learning_rate": 0.00013326633165829146,
+      "loss": 0.4768,
+      "step": 674
+    },
+    {
+      "epoch": 0.3375,
+      "grad_norm": 0.6481262445449829,
+      "learning_rate": 0.00013316582914572867,
+      "loss": 0.5249,
+      "step": 675
+    },
+    {
+      "epoch": 0.338,
+      "grad_norm": 0.705291211605072,
+      "learning_rate": 0.00013306532663316586,
+      "loss": 0.542,
+      "step": 676
+    },
+    {
+      "epoch": 0.3385,
+      "grad_norm": 0.6769446134567261,
+      "learning_rate": 0.00013296482412060301,
+      "loss": 0.4471,
+      "step": 677
+    },
+    {
+      "epoch": 0.339,
+      "grad_norm": 0.6781315803527832,
+      "learning_rate": 0.0001328643216080402,
+      "loss": 0.4816,
+      "step": 678
+    },
+    {
+      "epoch": 0.3395,
+      "grad_norm": 0.626255989074707,
+      "learning_rate": 0.0001327638190954774,
+      "loss": 0.2999,
+      "step": 679
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6601517200469971,
+      "learning_rate": 0.00013266331658291457,
+      "loss": 0.5085,
+      "step": 680
+    },
+    {
+      "epoch": 0.3405,
+      "grad_norm": 0.7327180504798889,
+      "learning_rate": 0.00013256281407035176,
+      "loss": 0.5062,
+      "step": 681
+    },
+    {
+      "epoch": 0.341,
+      "grad_norm": 0.6068310737609863,
+      "learning_rate": 0.00013246231155778895,
+      "loss": 0.3631,
+      "step": 682
+    },
+    {
+      "epoch": 0.3415,
+      "grad_norm": 0.6374022364616394,
+      "learning_rate": 0.00013236180904522613,
+      "loss": 0.4209,
+      "step": 683
+    },
+    {
+      "epoch": 0.342,
+      "grad_norm": 0.7234434485435486,
+      "learning_rate": 0.00013226130653266332,
+      "loss": 0.5698,
+      "step": 684
+    },
+    {
+      "epoch": 0.3425,
+      "grad_norm": 0.6477595567703247,
+      "learning_rate": 0.0001321608040201005,
+      "loss": 0.4489,
+      "step": 685
+    },
+    {
+      "epoch": 0.343,
+      "grad_norm": 0.6408647894859314,
+      "learning_rate": 0.0001320603015075377,
+      "loss": 0.4106,
+      "step": 686
+    },
+    {
+      "epoch": 0.3435,
+      "grad_norm": 0.6780766248703003,
+      "learning_rate": 0.00013195979899497488,
+      "loss": 0.4253,
+      "step": 687
+    },
+    {
+      "epoch": 0.344,
+      "grad_norm": 0.6178062558174133,
+      "learning_rate": 0.00013185929648241207,
+      "loss": 0.344,
+      "step": 688
+    },
+    {
+      "epoch": 0.3445,
+      "grad_norm": 0.6815159916877747,
+      "learning_rate": 0.00013175879396984925,
+      "loss": 0.4383,
+      "step": 689
+    },
+    {
+      "epoch": 0.345,
+      "grad_norm": 0.6829226613044739,
+      "learning_rate": 0.00013165829145728644,
+      "loss": 0.4394,
+      "step": 690
+    },
+    {
+      "epoch": 0.3455,
+      "grad_norm": 0.6620256900787354,
+      "learning_rate": 0.00013155778894472363,
+      "loss": 0.4372,
+      "step": 691
+    },
+    {
+      "epoch": 0.346,
+      "grad_norm": 0.7038753032684326,
+      "learning_rate": 0.0001314572864321608,
+      "loss": 0.4021,
+      "step": 692
+    },
+    {
+      "epoch": 0.3465,
+      "grad_norm": 0.7720176577568054,
+      "learning_rate": 0.000131356783919598,
+      "loss": 0.398,
+      "step": 693
+    },
+    {
+      "epoch": 0.347,
+      "grad_norm": 0.6240848898887634,
+      "learning_rate": 0.00013125628140703518,
+      "loss": 0.3443,
+      "step": 694
+    },
+    {
+      "epoch": 0.3475,
+      "grad_norm": 0.6680148839950562,
+      "learning_rate": 0.00013115577889447237,
+      "loss": 0.3312,
+      "step": 695
+    },
+    {
+      "epoch": 0.348,
+      "grad_norm": 0.8039118051528931,
+      "learning_rate": 0.00013105527638190956,
+      "loss": 0.4164,
+      "step": 696
+    },
+    {
+      "epoch": 0.3485,
+      "grad_norm": 0.7914130687713623,
+      "learning_rate": 0.00013095477386934674,
+      "loss": 0.2903,
+      "step": 697
+    },
+    {
+      "epoch": 0.349,
+      "grad_norm": 0.7591024041175842,
+      "learning_rate": 0.00013085427135678393,
+      "loss": 0.451,
+      "step": 698
+    },
+    {
+      "epoch": 0.3495,
+      "grad_norm": 0.7589693069458008,
+      "learning_rate": 0.0001307537688442211,
+      "loss": 0.3544,
+      "step": 699
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6071057915687561,
+      "learning_rate": 0.0001306532663316583,
+      "loss": 0.2999,
+      "step": 700
+    },
+    {
+      "epoch": 0.3505,
+      "grad_norm": 0.6010034680366516,
+      "learning_rate": 0.0001305527638190955,
+      "loss": 0.569,
+      "step": 701
+    },
+    {
+      "epoch": 0.351,
+      "grad_norm": 0.5435221195220947,
+      "learning_rate": 0.00013045226130653268,
+      "loss": 0.4087,
+      "step": 702
+    },
+    {
+      "epoch": 0.3515,
+      "grad_norm": 0.6977417469024658,
+      "learning_rate": 0.00013035175879396986,
+      "loss": 0.6146,
+      "step": 703
+    },
+    {
+      "epoch": 0.352,
+      "grad_norm": 0.7874389886856079,
+      "learning_rate": 0.00013025125628140705,
+      "loss": 0.7083,
+      "step": 704
+    },
+    {
+      "epoch": 0.3525,
+      "grad_norm": 0.7938291430473328,
+      "learning_rate": 0.00013015075376884424,
+      "loss": 0.645,
+      "step": 705
+    },
+    {
+      "epoch": 0.353,
+      "grad_norm": 0.6631731986999512,
+      "learning_rate": 0.0001300502512562814,
+      "loss": 0.5422,
+      "step": 706
+    },
+    {
+      "epoch": 0.3535,
+      "grad_norm": 0.6268270611763,
+      "learning_rate": 0.00012994974874371858,
+      "loss": 0.4887,
+      "step": 707
+    },
+    {
+      "epoch": 0.354,
+      "grad_norm": 0.6497275829315186,
+      "learning_rate": 0.00012984924623115577,
+      "loss": 0.4815,
+      "step": 708
+    },
+    {
+      "epoch": 0.3545,
+      "grad_norm": 0.6512343287467957,
+      "learning_rate": 0.00012974874371859298,
+      "loss": 0.5415,
+      "step": 709
+    },
+    {
+      "epoch": 0.355,
+      "grad_norm": 0.7381061911582947,
+      "learning_rate": 0.00012964824120603017,
+      "loss": 0.5219,
+      "step": 710
+    },
+    {
+      "epoch": 0.3555,
+      "grad_norm": 0.6348230242729187,
+      "learning_rate": 0.00012954773869346735,
+      "loss": 0.4564,
+      "step": 711
+    },
+    {
+      "epoch": 0.356,
+      "grad_norm": 0.7903686165809631,
+      "learning_rate": 0.00012944723618090454,
+      "loss": 0.4909,
+      "step": 712
+    },
+    {
+      "epoch": 0.3565,
+      "grad_norm": 0.6956429481506348,
+      "learning_rate": 0.0001293467336683417,
+      "loss": 0.5682,
+      "step": 713
+    },
+    {
+      "epoch": 0.357,
+      "grad_norm": 0.7090478539466858,
+      "learning_rate": 0.0001292462311557789,
+      "loss": 0.6102,
+      "step": 714
+    },
+    {
+      "epoch": 0.3575,
+      "grad_norm": 0.669255793094635,
+      "learning_rate": 0.00012914572864321607,
+      "loss": 0.5908,
+      "step": 715
+    },
+    {
+      "epoch": 0.358,
+      "grad_norm": 0.6955499649047852,
+      "learning_rate": 0.00012904522613065326,
+      "loss": 0.5039,
+      "step": 716
+    },
+    {
+      "epoch": 0.3585,
+      "grad_norm": 0.6489084959030151,
+      "learning_rate": 0.00012894472361809047,
+      "loss": 0.4189,
+      "step": 717
+    },
+    {
+      "epoch": 0.359,
+      "grad_norm": 0.6561848521232605,
+      "learning_rate": 0.00012884422110552766,
+      "loss": 0.5097,
+      "step": 718
+    },
+    {
+      "epoch": 0.3595,
+      "grad_norm": 0.5511810779571533,
+      "learning_rate": 0.00012874371859296485,
+      "loss": 0.3753,
+      "step": 719
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6526991724967957,
+      "learning_rate": 0.000128643216080402,
+      "loss": 0.4468,
+      "step": 720
+    },
+    {
+      "epoch": 0.3605,
+      "grad_norm": 0.7432270050048828,
+      "learning_rate": 0.0001285427135678392,
+      "loss": 0.5779,
+      "step": 721
+    },
+    {
+      "epoch": 0.361,
+      "grad_norm": 0.6847168803215027,
+      "learning_rate": 0.00012844221105527638,
+      "loss": 0.5052,
+      "step": 722
+    },
+    {
+      "epoch": 0.3615,
+      "grad_norm": 0.5510095953941345,
+      "learning_rate": 0.00012834170854271357,
+      "loss": 0.4615,
+      "step": 723
+    },
+    {
+      "epoch": 0.362,
+      "grad_norm": 0.663713812828064,
+      "learning_rate": 0.00012824120603015075,
+      "loss": 0.3982,
+      "step": 724
+    },
+    {
+      "epoch": 0.3625,
+      "grad_norm": 0.5830754637718201,
+      "learning_rate": 0.00012814070351758797,
+      "loss": 0.3999,
+      "step": 725
+    },
+    {
+      "epoch": 0.363,
+      "grad_norm": 0.6276527643203735,
+      "learning_rate": 0.00012804020100502515,
+      "loss": 0.4345,
+      "step": 726
+    },
+    {
+      "epoch": 0.3635,
+      "grad_norm": 0.6939334273338318,
+      "learning_rate": 0.0001279396984924623,
+      "loss": 0.4639,
+      "step": 727
+    },
+    {
+      "epoch": 0.364,
+      "grad_norm": 0.8046755194664001,
+      "learning_rate": 0.0001278391959798995,
+      "loss": 0.6184,
+      "step": 728
+    },
+    {
+      "epoch": 0.3645,
+      "grad_norm": 0.6341655850410461,
+      "learning_rate": 0.00012773869346733668,
+      "loss": 0.4596,
+      "step": 729
+    },
+    {
+      "epoch": 0.365,
+      "grad_norm": 0.6571204662322998,
+      "learning_rate": 0.00012763819095477387,
+      "loss": 0.5166,
+      "step": 730
+    },
+    {
+      "epoch": 0.3655,
+      "grad_norm": 0.6901340484619141,
+      "learning_rate": 0.00012753768844221106,
+      "loss": 0.4398,
+      "step": 731
+    },
+    {
+      "epoch": 0.366,
+      "grad_norm": 0.6806580424308777,
+      "learning_rate": 0.00012743718592964824,
+      "loss": 0.3942,
+      "step": 732
+    },
+    {
+      "epoch": 0.3665,
+      "grad_norm": 0.5810626149177551,
+      "learning_rate": 0.00012733668341708543,
+      "loss": 0.3348,
+      "step": 733
+    },
+    {
+      "epoch": 0.367,
+      "grad_norm": 0.6607562899589539,
+      "learning_rate": 0.00012723618090452262,
+      "loss": 0.357,
+      "step": 734
+    },
+    {
+      "epoch": 0.3675,
+      "grad_norm": 0.5651497840881348,
+      "learning_rate": 0.0001271356783919598,
+      "loss": 0.3278,
+      "step": 735
+    },
+    {
+      "epoch": 0.368,
+      "grad_norm": 0.9776192307472229,
+      "learning_rate": 0.000127035175879397,
+      "loss": 0.4794,
+      "step": 736
+    },
+    {
+      "epoch": 0.3685,
+      "grad_norm": 0.7543694376945496,
+      "learning_rate": 0.00012693467336683418,
+      "loss": 0.521,
+      "step": 737
+    },
+    {
+      "epoch": 0.369,
+      "grad_norm": 0.6802242994308472,
+      "learning_rate": 0.00012683417085427136,
+      "loss": 0.4235,
+      "step": 738
+    },
+    {
+      "epoch": 0.3695,
+      "grad_norm": 0.6219953894615173,
+      "learning_rate": 0.00012673366834170855,
+      "loss": 0.3349,
+      "step": 739
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6835973262786865,
+      "learning_rate": 0.00012663316582914574,
+      "loss": 0.4398,
+      "step": 740
+    },
+    {
+      "epoch": 0.3705,
+      "grad_norm": 0.6314231753349304,
+      "learning_rate": 0.00012653266331658292,
+      "loss": 0.3476,
+      "step": 741
+    },
+    {
+      "epoch": 0.371,
+      "grad_norm": 0.7550387382507324,
+      "learning_rate": 0.0001264321608040201,
+      "loss": 0.4731,
+      "step": 742
+    },
+    {
+      "epoch": 0.3715,
+      "grad_norm": 0.7602382302284241,
+      "learning_rate": 0.0001263316582914573,
+      "loss": 0.4541,
+      "step": 743
+    },
+    {
+      "epoch": 0.372,
+      "grad_norm": 0.6269517540931702,
+      "learning_rate": 0.00012623115577889448,
+      "loss": 0.3915,
+      "step": 744
+    },
+    {
+      "epoch": 0.3725,
+      "grad_norm": 0.7010166049003601,
+      "learning_rate": 0.00012613065326633167,
+      "loss": 0.4221,
+      "step": 745
+    },
+    {
+      "epoch": 0.373,
+      "grad_norm": 0.684403657913208,
+      "learning_rate": 0.00012603015075376885,
+      "loss": 0.3427,
+      "step": 746
+    },
+    {
+      "epoch": 0.3735,
+      "grad_norm": 0.6829254627227783,
+      "learning_rate": 0.00012592964824120604,
+      "loss": 0.3196,
+      "step": 747
+    },
+    {
+      "epoch": 0.374,
+      "grad_norm": 0.6489712595939636,
+      "learning_rate": 0.00012582914572864323,
+      "loss": 0.3978,
+      "step": 748
+    },
+    {
+      "epoch": 0.3745,
+      "grad_norm": 0.6540708541870117,
+      "learning_rate": 0.0001257286432160804,
+      "loss": 0.3613,
+      "step": 749
+    },
+    {
+      "epoch": 0.375,
+      "grad_norm": 0.6953591704368591,
+      "learning_rate": 0.0001256281407035176,
+      "loss": 0.3339,
+      "step": 750
+    },
+    {
+      "epoch": 0.3755,
+      "grad_norm": 0.5229864120483398,
+      "learning_rate": 0.0001255276381909548,
+      "loss": 0.4174,
+      "step": 751
+    },
+    {
+      "epoch": 0.376,
+      "grad_norm": 0.5876749157905579,
+      "learning_rate": 0.00012542713567839197,
+      "loss": 0.428,
+      "step": 752
+    },
+    {
+      "epoch": 0.3765,
+      "grad_norm": 0.7164889574050903,
+      "learning_rate": 0.00012532663316582916,
+      "loss": 0.6847,
+      "step": 753
+    },
+    {
+      "epoch": 0.377,
+      "grad_norm": 0.6268723011016846,
+      "learning_rate": 0.00012522613065326635,
+      "loss": 0.569,
+      "step": 754
+    },
+    {
+      "epoch": 0.3775,
+      "grad_norm": 0.6305063962936401,
+      "learning_rate": 0.00012512562814070353,
+      "loss": 0.5488,
+      "step": 755
+    },
+    {
+      "epoch": 0.378,
+      "grad_norm": 0.6433343887329102,
+      "learning_rate": 0.0001250251256281407,
+      "loss": 0.4665,
+      "step": 756
+    },
+    {
+      "epoch": 0.3785,
+      "grad_norm": 0.6525387167930603,
+      "learning_rate": 0.00012492462311557788,
+      "loss": 0.5072,
+      "step": 757
+    },
+    {
+      "epoch": 0.379,
+      "grad_norm": 0.6392046809196472,
+      "learning_rate": 0.00012482412060301507,
+      "loss": 0.4828,
+      "step": 758
+    },
+    {
+      "epoch": 0.3795,
+      "grad_norm": 0.6635848879814148,
+      "learning_rate": 0.00012472361809045228,
+      "loss": 0.6234,
+      "step": 759
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6538762450218201,
+      "learning_rate": 0.00012462311557788947,
+      "loss": 0.4496,
+      "step": 760
+    },
+    {
+      "epoch": 0.3805,
+      "grad_norm": 0.6655104160308838,
+      "learning_rate": 0.00012452261306532665,
+      "loss": 0.4716,
+      "step": 761
+    },
+    {
+      "epoch": 0.381,
+      "grad_norm": 0.6546260118484497,
+      "learning_rate": 0.00012442211055276384,
+      "loss": 0.4537,
+      "step": 762
+    },
+    {
+      "epoch": 0.3815,
+      "grad_norm": 0.6713255643844604,
+      "learning_rate": 0.000124321608040201,
+      "loss": 0.5567,
+      "step": 763
+    },
+    {
+      "epoch": 0.382,
+      "grad_norm": 0.618012011051178,
+      "learning_rate": 0.00012422110552763818,
+      "loss": 0.4621,
+      "step": 764
+    },
+    {
+      "epoch": 0.3825,
+      "grad_norm": 0.7027286887168884,
+      "learning_rate": 0.00012412060301507537,
+      "loss": 0.5078,
+      "step": 765
+    },
+    {
+      "epoch": 0.383,
+      "grad_norm": 0.5655239224433899,
+      "learning_rate": 0.00012402010050251256,
+      "loss": 0.3493,
+      "step": 766
+    },
+    {
+      "epoch": 0.3835,
+      "grad_norm": 0.7134120464324951,
+      "learning_rate": 0.00012391959798994977,
+      "loss": 0.5821,
+      "step": 767
+    },
+    {
+      "epoch": 0.384,
+      "grad_norm": 0.6549416184425354,
+      "learning_rate": 0.00012381909547738696,
+      "loss": 0.5422,
+      "step": 768
+    },
+    {
+      "epoch": 0.3845,
+      "grad_norm": 0.7009751200675964,
+      "learning_rate": 0.00012371859296482414,
+      "loss": 0.5362,
+      "step": 769
+    },
+    {
+      "epoch": 0.385,
+      "grad_norm": 0.590291440486908,
+      "learning_rate": 0.0001236180904522613,
+      "loss": 0.3847,
+      "step": 770
+    },
+    {
+      "epoch": 0.3855,
+      "grad_norm": 0.6242939829826355,
+      "learning_rate": 0.0001235175879396985,
+      "loss": 0.4765,
+      "step": 771
+    },
+    {
+      "epoch": 0.386,
+      "grad_norm": 0.6846091747283936,
+      "learning_rate": 0.00012341708542713568,
+      "loss": 0.4358,
+      "step": 772
+    },
+    {
+      "epoch": 0.3865,
+      "grad_norm": 0.6702343225479126,
+      "learning_rate": 0.00012331658291457286,
+      "loss": 0.4448,
+      "step": 773
+    },
+    {
+      "epoch": 0.387,
+      "grad_norm": 0.6505341529846191,
+      "learning_rate": 0.00012321608040201005,
+      "loss": 0.5132,
+      "step": 774
+    },
+    {
+      "epoch": 0.3875,
+      "grad_norm": 0.8075600862503052,
+      "learning_rate": 0.00012311557788944724,
+      "loss": 0.535,
+      "step": 775
+    },
+    {
+      "epoch": 0.388,
+      "grad_norm": 0.5987162590026855,
+      "learning_rate": 0.00012301507537688445,
+      "loss": 0.3797,
+      "step": 776
+    },
+    {
+      "epoch": 0.3885,
+      "grad_norm": 0.6283299326896667,
+      "learning_rate": 0.0001229145728643216,
+      "loss": 0.4211,
+      "step": 777
+    },
+    {
+      "epoch": 0.389,
+      "grad_norm": 0.5995007157325745,
+      "learning_rate": 0.0001228140703517588,
+      "loss": 0.3634,
+      "step": 778
+    },
+    {
+      "epoch": 0.3895,
+      "grad_norm": 0.6714475154876709,
+      "learning_rate": 0.00012271356783919598,
+      "loss": 0.4098,
+      "step": 779
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7879155874252319,
+      "learning_rate": 0.00012261306532663317,
+      "loss": 0.4826,
+      "step": 780
+    },
+    {
+      "epoch": 0.3905,
+      "grad_norm": 0.6633744835853577,
+      "learning_rate": 0.00012251256281407035,
+      "loss": 0.4538,
+      "step": 781
+    },
+    {
+      "epoch": 0.391,
+      "grad_norm": 0.7721315622329712,
+      "learning_rate": 0.00012241206030150754,
+      "loss": 0.6062,
+      "step": 782
+    },
+    {
+      "epoch": 0.3915,
+      "grad_norm": 0.6604461073875427,
+      "learning_rate": 0.00012231155778894473,
+      "loss": 0.4672,
+      "step": 783
+    },
+    {
+      "epoch": 0.392,
+      "grad_norm": 0.6523184180259705,
+      "learning_rate": 0.00012221105527638191,
+      "loss": 0.4377,
+      "step": 784
+    },
+    {
+      "epoch": 0.3925,
+      "grad_norm": 0.689082145690918,
+      "learning_rate": 0.0001221105527638191,
+      "loss": 0.4995,
+      "step": 785
+    },
+    {
+      "epoch": 0.393,
+      "grad_norm": 0.7039391994476318,
+      "learning_rate": 0.00012201005025125629,
+      "loss": 0.4629,
+      "step": 786
+    },
+    {
+      "epoch": 0.3935,
+      "grad_norm": 0.6609583497047424,
+      "learning_rate": 0.00012190954773869347,
+      "loss": 0.4276,
+      "step": 787
+    },
+    {
+      "epoch": 0.394,
+      "grad_norm": 0.8710364699363708,
+      "learning_rate": 0.00012180904522613066,
+      "loss": 0.4326,
+      "step": 788
+    },
+    {
+      "epoch": 0.3945,
+      "grad_norm": 0.6564118266105652,
+      "learning_rate": 0.00012170854271356785,
+      "loss": 0.4562,
+      "step": 789
+    },
+    {
+      "epoch": 0.395,
+      "grad_norm": 0.8167173862457275,
+      "learning_rate": 0.00012160804020100502,
+      "loss": 0.5081,
+      "step": 790
+    },
+    {
+      "epoch": 0.3955,
+      "grad_norm": 0.5859940648078918,
+      "learning_rate": 0.0001215075376884422,
+      "loss": 0.3782,
+      "step": 791
+    },
+    {
+      "epoch": 0.396,
+      "grad_norm": 0.6078713536262512,
+      "learning_rate": 0.00012140703517587942,
+      "loss": 0.3503,
+      "step": 792
+    },
+    {
+      "epoch": 0.3965,
+      "grad_norm": 0.7475302219390869,
+      "learning_rate": 0.00012130653266331659,
+      "loss": 0.4884,
+      "step": 793
+    },
+    {
+      "epoch": 0.397,
+      "grad_norm": 0.6197308897972107,
+      "learning_rate": 0.00012120603015075378,
+      "loss": 0.3515,
+      "step": 794
+    },
+    {
+      "epoch": 0.3975,
+      "grad_norm": 0.5162922739982605,
+      "learning_rate": 0.00012110552763819097,
+      "loss": 0.2955,
+      "step": 795
+    },
+    {
+      "epoch": 0.398,
+      "grad_norm": 0.6324794888496399,
+      "learning_rate": 0.00012100502512562815,
+      "loss": 0.3578,
+      "step": 796
+    },
+    {
+      "epoch": 0.3985,
+      "grad_norm": 0.6863030791282654,
+      "learning_rate": 0.00012090452261306532,
+      "loss": 0.4104,
+      "step": 797
+    },
+    {
+      "epoch": 0.399,
+      "grad_norm": 0.564802348613739,
+      "learning_rate": 0.00012080402010050251,
+      "loss": 0.321,
+      "step": 798
+    },
+    {
+      "epoch": 0.3995,
+      "grad_norm": 0.6706806421279907,
+      "learning_rate": 0.0001207035175879397,
+      "loss": 0.315,
+      "step": 799
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6451688408851624,
+      "learning_rate": 0.00012060301507537688,
+      "loss": 0.3046,
+      "step": 800
+    },
+    {
+      "epoch": 0.4,
+      "eval_loss": 0.46057480573654175,
+      "eval_runtime": 109.3011,
+      "eval_samples_per_second": 4.575,
+      "eval_steps_per_second": 4.575,
+      "step": 800
+    },
+    {
+      "epoch": 0.4005,
+      "grad_norm": 0.695269763469696,
+      "learning_rate": 0.00012050251256281408,
+      "loss": 0.6059,
+      "step": 801
+    },
+    {
+      "epoch": 0.401,
+      "grad_norm": 0.5415158867835999,
+      "learning_rate": 0.00012040201005025127,
+      "loss": 0.4146,
+      "step": 802
+    },
+    {
+      "epoch": 0.4015,
+      "grad_norm": 0.6972713470458984,
+      "learning_rate": 0.00012030150753768846,
+      "loss": 0.6355,
+      "step": 803
+    },
+    {
+      "epoch": 0.402,
+      "grad_norm": 0.5803679823875427,
+      "learning_rate": 0.00012020100502512563,
+      "loss": 0.441,
+      "step": 804
+    },
+    {
+      "epoch": 0.4025,
+      "grad_norm": 0.5423812866210938,
+      "learning_rate": 0.00012010050251256282,
+      "loss": 0.4207,
+      "step": 805
+    },
+    {
+      "epoch": 0.403,
+      "grad_norm": 0.7284221649169922,
+      "learning_rate": 0.00012,
+      "loss": 0.5337,
+      "step": 806
+    },
+    {
+      "epoch": 0.4035,
+      "grad_norm": 0.6033008694648743,
+      "learning_rate": 0.00011989949748743719,
+      "loss": 0.4185,
+      "step": 807
+    },
+    {
+      "epoch": 0.404,
+      "grad_norm": 0.6643309593200684,
+      "learning_rate": 0.00011979899497487436,
+      "loss": 0.5535,
+      "step": 808
+    },
+    {
+      "epoch": 0.4045,
+      "grad_norm": 0.7048445343971252,
+      "learning_rate": 0.00011969849246231158,
+      "loss": 0.5375,
+      "step": 809
+    },
+    {
+      "epoch": 0.405,
+      "grad_norm": 0.6209299564361572,
+      "learning_rate": 0.00011959798994974876,
+      "loss": 0.4217,
+      "step": 810
+    },
+    {
+      "epoch": 0.4055,
+      "grad_norm": 0.6066304445266724,
+      "learning_rate": 0.00011949748743718594,
+      "loss": 0.4436,
+      "step": 811
+    },
+    {
+      "epoch": 0.406,
+      "grad_norm": 0.7337306141853333,
+      "learning_rate": 0.00011939698492462312,
+      "loss": 0.573,
+      "step": 812
+    },
+    {
+      "epoch": 0.4065,
+      "grad_norm": 0.5860584378242493,
+      "learning_rate": 0.00011929648241206031,
+      "loss": 0.3965,
+      "step": 813
+    },
+    {
+      "epoch": 0.407,
+      "grad_norm": 0.6481097936630249,
+      "learning_rate": 0.0001191959798994975,
+      "loss": 0.4245,
+      "step": 814
+    },
+    {
+      "epoch": 0.4075,
+      "grad_norm": 0.6164637207984924,
+      "learning_rate": 0.00011909547738693467,
+      "loss": 0.3482,
+      "step": 815
+    },
+    {
+      "epoch": 0.408,
+      "grad_norm": 0.73078852891922,
+      "learning_rate": 0.00011899497487437185,
+      "loss": 0.435,
+      "step": 816
+    },
+    {
+      "epoch": 0.4085,
+      "grad_norm": 0.6557646989822388,
+      "learning_rate": 0.00011889447236180907,
+      "loss": 0.4816,
+      "step": 817
+    },
+    {
+      "epoch": 0.409,
+      "grad_norm": 0.6213710308074951,
+      "learning_rate": 0.00011879396984924624,
+      "loss": 0.4027,
+      "step": 818
+    },
+    {
+      "epoch": 0.4095,
+      "grad_norm": 0.6748867630958557,
+      "learning_rate": 0.00011869346733668343,
+      "loss": 0.4573,
+      "step": 819
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6431273818016052,
+      "learning_rate": 0.00011859296482412061,
+      "loss": 0.4309,
+      "step": 820
+    },
+    {
+      "epoch": 0.4105,
+      "grad_norm": 0.6067095398902893,
+      "learning_rate": 0.0001184924623115578,
+      "loss": 0.3773,
+      "step": 821
+    },
+    {
+      "epoch": 0.411,
+      "grad_norm": 0.8941280245780945,
+      "learning_rate": 0.00011839195979899497,
+      "loss": 0.5296,
+      "step": 822
+    },
+    {
+      "epoch": 0.4115,
+      "grad_norm": 0.7774582505226135,
+      "learning_rate": 0.00011829145728643216,
+      "loss": 0.494,
+      "step": 823
+    },
+    {
+      "epoch": 0.412,
+      "grad_norm": 0.6341260075569153,
+      "learning_rate": 0.00011819095477386935,
+      "loss": 0.4489,
+      "step": 824
+    },
+    {
+      "epoch": 0.4125,
+      "grad_norm": 0.579179584980011,
+      "learning_rate": 0.00011809045226130653,
+      "loss": 0.3788,
+      "step": 825
+    },
+    {
+      "epoch": 0.413,
+      "grad_norm": 0.6744787693023682,
+      "learning_rate": 0.00011798994974874373,
+      "loss": 0.4423,
+      "step": 826
+    },
+    {
+      "epoch": 0.4135,
+      "grad_norm": 0.6709211468696594,
+      "learning_rate": 0.00011788944723618092,
+      "loss": 0.4655,
+      "step": 827
+    },
+    {
+      "epoch": 0.414,
+      "grad_norm": 0.6809451580047607,
+      "learning_rate": 0.0001177889447236181,
+      "loss": 0.4243,
+      "step": 828
+    },
+    {
+      "epoch": 0.4145,
+      "grad_norm": 0.6023095846176147,
+      "learning_rate": 0.00011768844221105528,
+      "loss": 0.3846,
+      "step": 829
+    },
+    {
+      "epoch": 0.415,
+      "grad_norm": 0.726060152053833,
+      "learning_rate": 0.00011758793969849247,
+      "loss": 0.4482,
+      "step": 830
+    },
+    {
+      "epoch": 0.4155,
+      "grad_norm": 0.6115005016326904,
+      "learning_rate": 0.00011748743718592965,
+      "loss": 0.28,
+      "step": 831
+    },
+    {
+      "epoch": 0.416,
+      "grad_norm": 0.6593244075775146,
+      "learning_rate": 0.00011738693467336684,
+      "loss": 0.4203,
+      "step": 832
+    },
+    {
+      "epoch": 0.4165,
+      "grad_norm": 0.6594223380088806,
+      "learning_rate": 0.00011728643216080401,
+      "loss": 0.407,
+      "step": 833
+    },
+    {
+      "epoch": 0.417,
+      "grad_norm": 0.6577668190002441,
+      "learning_rate": 0.00011718592964824122,
+      "loss": 0.4019,
+      "step": 834
+    },
+    {
+      "epoch": 0.4175,
+      "grad_norm": 0.8204243779182434,
+      "learning_rate": 0.00011708542713567841,
+      "loss": 0.4649,
+      "step": 835
+    },
+    {
+      "epoch": 0.418,
+      "grad_norm": 0.7047533392906189,
+      "learning_rate": 0.00011698492462311558,
+      "loss": 0.3948,
+      "step": 836
+    },
+    {
+      "epoch": 0.4185,
+      "grad_norm": 0.7383876442909241,
+      "learning_rate": 0.00011688442211055277,
+      "loss": 0.5201,
+      "step": 837
+    },
+    {
+      "epoch": 0.419,
+      "grad_norm": 0.7413540482521057,
+      "learning_rate": 0.00011678391959798996,
+      "loss": 0.3794,
+      "step": 838
+    },
+    {
+      "epoch": 0.4195,
+      "grad_norm": 0.7076113224029541,
+      "learning_rate": 0.00011668341708542714,
+      "loss": 0.534,
+      "step": 839
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6135266423225403,
+      "learning_rate": 0.00011658291457286432,
+      "loss": 0.3603,
+      "step": 840
+    },
+    {
+      "epoch": 0.4205,
+      "grad_norm": 0.7112100124359131,
+      "learning_rate": 0.0001164824120603015,
+      "loss": 0.4675,
+      "step": 841
+    },
+    {
+      "epoch": 0.421,
+      "grad_norm": 0.6987900733947754,
+      "learning_rate": 0.00011638190954773872,
+      "loss": 0.3901,
+      "step": 842
+    },
+    {
+      "epoch": 0.4215,
+      "grad_norm": 0.6804396510124207,
+      "learning_rate": 0.00011628140703517589,
+      "loss": 0.4463,
+      "step": 843
+    },
+    {
+      "epoch": 0.422,
+      "grad_norm": 0.7364391088485718,
+      "learning_rate": 0.00011618090452261308,
+      "loss": 0.3215,
+      "step": 844
+    },
+    {
+      "epoch": 0.4225,
+      "grad_norm": 0.7415553331375122,
+      "learning_rate": 0.00011608040201005026,
+      "loss": 0.446,
+      "step": 845
+    },
+    {
+      "epoch": 0.423,
+      "grad_norm": 0.8232871890068054,
+      "learning_rate": 0.00011597989949748745,
+      "loss": 0.4136,
+      "step": 846
+    },
+    {
+      "epoch": 0.4235,
+      "grad_norm": 0.636400043964386,
+      "learning_rate": 0.00011587939698492462,
+      "loss": 0.3481,
+      "step": 847
+    },
+    {
+      "epoch": 0.424,
+      "grad_norm": 0.5906121134757996,
+      "learning_rate": 0.00011577889447236181,
+      "loss": 0.2645,
+      "step": 848
+    },
+    {
+      "epoch": 0.4245,
+      "grad_norm": 0.7287012338638306,
+      "learning_rate": 0.000115678391959799,
+      "loss": 0.4016,
+      "step": 849
+    },
+    {
+      "epoch": 0.425,
+      "grad_norm": 0.5770605206489563,
+      "learning_rate": 0.00011557788944723618,
+      "loss": 0.3078,
+      "step": 850
+    },
+    {
+      "epoch": 0.4255,
+      "grad_norm": 0.5685473084449768,
+      "learning_rate": 0.00011547738693467338,
+      "loss": 0.5386,
+      "step": 851
+    },
+    {
+      "epoch": 0.426,
+      "grad_norm": 0.5275063514709473,
+      "learning_rate": 0.00011537688442211057,
+      "loss": 0.4218,
+      "step": 852
+    },
+    {
+      "epoch": 0.4265,
+      "grad_norm": 0.6230756044387817,
+      "learning_rate": 0.00011527638190954775,
+      "loss": 0.5573,
+      "step": 853
+    },
+    {
+      "epoch": 0.427,
+      "grad_norm": 0.6235846281051636,
+      "learning_rate": 0.00011517587939698493,
+      "loss": 0.4519,
+      "step": 854
+    },
+    {
+      "epoch": 0.4275,
+      "grad_norm": 0.6024863123893738,
+      "learning_rate": 0.00011507537688442211,
+      "loss": 0.5701,
+      "step": 855
+    },
+    {
+      "epoch": 0.428,
+      "grad_norm": 0.699385404586792,
+      "learning_rate": 0.0001149748743718593,
+      "loss": 0.5108,
+      "step": 856
+    },
+    {
+      "epoch": 0.4285,
+      "grad_norm": 0.5543192625045776,
+      "learning_rate": 0.00011487437185929649,
+      "loss": 0.4929,
+      "step": 857
+    },
+    {
+      "epoch": 0.429,
+      "grad_norm": 0.6173766851425171,
+      "learning_rate": 0.00011477386934673366,
+      "loss": 0.4686,
+      "step": 858
+    },
+    {
+      "epoch": 0.4295,
+      "grad_norm": 0.6076511144638062,
+      "learning_rate": 0.00011467336683417087,
+      "loss": 0.4329,
+      "step": 859
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6499843597412109,
+      "learning_rate": 0.00011457286432160806,
+      "loss": 0.5409,
+      "step": 860
+    },
+    {
+      "epoch": 0.4305,
+      "grad_norm": 0.628150224685669,
+      "learning_rate": 0.00011447236180904523,
+      "loss": 0.4076,
+      "step": 861
+    },
+    {
+      "epoch": 0.431,
+      "grad_norm": 0.801193118095398,
+      "learning_rate": 0.00011437185929648242,
+      "loss": 0.6134,
+      "step": 862
+    },
+    {
+      "epoch": 0.4315,
+      "grad_norm": 0.8064663410186768,
+      "learning_rate": 0.0001142713567839196,
+      "loss": 0.6258,
+      "step": 863
+    },
+    {
+      "epoch": 0.432,
+      "grad_norm": 0.533795952796936,
+      "learning_rate": 0.00011417085427135679,
+      "loss": 0.3837,
+      "step": 864
+    },
+    {
+      "epoch": 0.4325,
+      "grad_norm": 0.6335496306419373,
+      "learning_rate": 0.00011407035175879396,
+      "loss": 0.4206,
+      "step": 865
+    },
+    {
+      "epoch": 0.433,
+      "grad_norm": 0.6104476451873779,
+      "learning_rate": 0.00011396984924623115,
+      "loss": 0.3927,
+      "step": 866
+    },
+    {
+      "epoch": 0.4335,
+      "grad_norm": 0.7701111435890198,
+      "learning_rate": 0.00011386934673366834,
+      "loss": 0.611,
+      "step": 867
+    },
+    {
+      "epoch": 0.434,
+      "grad_norm": 0.6561410427093506,
+      "learning_rate": 0.00011376884422110554,
+      "loss": 0.4196,
+      "step": 868
+    },
+    {
+      "epoch": 0.4345,
+      "grad_norm": 0.6631766557693481,
+      "learning_rate": 0.00011366834170854272,
+      "loss": 0.4261,
+      "step": 869
+    },
+    {
+      "epoch": 0.435,
+      "grad_norm": 0.6072076559066772,
+      "learning_rate": 0.00011356783919597991,
+      "loss": 0.4374,
+      "step": 870
+    },
+    {
+      "epoch": 0.4355,
+      "grad_norm": 0.6271715760231018,
+      "learning_rate": 0.0001134673366834171,
+      "loss": 0.4496,
+      "step": 871
+    },
+    {
+      "epoch": 0.436,
+      "grad_norm": 0.7398489117622375,
+      "learning_rate": 0.00011336683417085427,
+      "loss": 0.4324,
+      "step": 872
+    },
+    {
+      "epoch": 0.4365,
+      "grad_norm": 0.7068373560905457,
+      "learning_rate": 0.00011326633165829146,
+      "loss": 0.4233,
+      "step": 873
+    },
+    {
+      "epoch": 0.437,
+      "grad_norm": 0.7282095551490784,
+      "learning_rate": 0.00011316582914572864,
+      "loss": 0.4489,
+      "step": 874
+    },
+    {
+      "epoch": 0.4375,
+      "grad_norm": 0.6563122868537903,
+      "learning_rate": 0.00011306532663316583,
+      "loss": 0.4254,
+      "step": 875
+    },
+    {
+      "epoch": 0.438,
+      "grad_norm": 0.6277741193771362,
+      "learning_rate": 0.00011296482412060303,
+      "loss": 0.4467,
+      "step": 876
+    },
+    {
+      "epoch": 0.4385,
+      "grad_norm": 0.6733102798461914,
+      "learning_rate": 0.00011286432160804022,
+      "loss": 0.4265,
+      "step": 877
+    },
+    {
+      "epoch": 0.439,
+      "grad_norm": 0.6597092151641846,
+      "learning_rate": 0.0001127638190954774,
+      "loss": 0.4302,
+      "step": 878
+    },
+    {
+      "epoch": 0.4395,
+      "grad_norm": 0.7238799333572388,
+      "learning_rate": 0.00011266331658291458,
+      "loss": 0.4951,
+      "step": 879
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5781562328338623,
+      "learning_rate": 0.00011256281407035176,
+      "loss": 0.3626,
+      "step": 880
+    },
+    {
+      "epoch": 0.4405,
+      "grad_norm": 0.7262015342712402,
+      "learning_rate": 0.00011246231155778895,
+      "loss": 0.4424,
+      "step": 881
+    },
+    {
+      "epoch": 0.441,
+      "grad_norm": 0.6868186593055725,
+      "learning_rate": 0.00011236180904522614,
+      "loss": 0.393,
+      "step": 882
+    },
+    {
+      "epoch": 0.4415,
+      "grad_norm": 0.5810798406600952,
+      "learning_rate": 0.00011226130653266331,
+      "loss": 0.3387,
+      "step": 883
+    },
+    {
+      "epoch": 0.442,
+      "grad_norm": 0.6988510489463806,
+      "learning_rate": 0.00011216080402010052,
+      "loss": 0.4306,
+      "step": 884
+    },
+    {
+      "epoch": 0.4425,
+      "grad_norm": 0.6653384566307068,
+      "learning_rate": 0.00011206030150753771,
+      "loss": 0.4097,
+      "step": 885
+    },
+    {
+      "epoch": 0.443,
+      "grad_norm": 0.7394219040870667,
+      "learning_rate": 0.00011195979899497488,
+      "loss": 0.5136,
+      "step": 886
+    },
+    {
+      "epoch": 0.4435,
+      "grad_norm": 0.8427491188049316,
+      "learning_rate": 0.00011185929648241207,
+      "loss": 0.4564,
+      "step": 887
+    },
+    {
+      "epoch": 0.444,
+      "grad_norm": 0.888885498046875,
+      "learning_rate": 0.00011175879396984925,
+      "loss": 0.5678,
+      "step": 888
+    },
+    {
+      "epoch": 0.4445,
+      "grad_norm": 0.6608572602272034,
+      "learning_rate": 0.00011165829145728644,
+      "loss": 0.3872,
+      "step": 889
+    },
+    {
+      "epoch": 0.445,
+      "grad_norm": 0.6415966153144836,
+      "learning_rate": 0.00011155778894472361,
+      "loss": 0.4032,
+      "step": 890
+    },
+    {
+      "epoch": 0.4455,
+      "grad_norm": 0.6585486531257629,
+      "learning_rate": 0.0001114572864321608,
+      "loss": 0.4175,
+      "step": 891
+    },
+    {
+      "epoch": 0.446,
+      "grad_norm": 0.6461412906646729,
+      "learning_rate": 0.00011135678391959799,
+      "loss": 0.3505,
+      "step": 892
+    },
+    {
+      "epoch": 0.4465,
+      "grad_norm": 0.6923301219940186,
+      "learning_rate": 0.00011125628140703519,
+      "loss": 0.4843,
+      "step": 893
+    },
+    {
+      "epoch": 0.447,
+      "grad_norm": 0.6094756722450256,
+      "learning_rate": 0.00011115577889447237,
+      "loss": 0.3533,
+      "step": 894
+    },
+    {
+      "epoch": 0.4475,
+      "grad_norm": 0.6534079909324646,
+      "learning_rate": 0.00011105527638190956,
+      "loss": 0.3639,
+      "step": 895
+    },
+    {
+      "epoch": 0.448,
+      "grad_norm": 0.7944132089614868,
+      "learning_rate": 0.00011095477386934675,
+      "loss": 0.5237,
+      "step": 896
+    },
+    {
+      "epoch": 0.4485,
+      "grad_norm": 0.813023030757904,
+      "learning_rate": 0.00011085427135678392,
+      "loss": 0.4301,
+      "step": 897
+    },
+    {
+      "epoch": 0.449,
+      "grad_norm": 0.5883468985557556,
+      "learning_rate": 0.0001107537688442211,
+      "loss": 0.3299,
+      "step": 898
+    },
+    {
+      "epoch": 0.4495,
+      "grad_norm": 0.5350788831710815,
+      "learning_rate": 0.00011065326633165829,
+      "loss": 0.2701,
+      "step": 899
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.843183159828186,
+      "learning_rate": 0.00011055276381909548,
+      "loss": 0.4112,
+      "step": 900
+    },
+    {
+      "epoch": 0.4505,
+      "grad_norm": 0.5139120221138,
+      "learning_rate": 0.00011045226130653268,
+      "loss": 0.46,
+      "step": 901
+    },
+    {
+      "epoch": 0.451,
+      "grad_norm": 0.6451822519302368,
+      "learning_rate": 0.00011035175879396986,
+      "loss": 0.6365,
+      "step": 902
+    },
+    {
+      "epoch": 0.4515,
+      "grad_norm": 0.566187858581543,
+      "learning_rate": 0.00011025125628140705,
+      "loss": 0.5234,
+      "step": 903
+    },
+    {
+      "epoch": 0.452,
+      "grad_norm": 0.5185102820396423,
+      "learning_rate": 0.00011015075376884422,
+      "loss": 0.3566,
+      "step": 904
+    },
+    {
+      "epoch": 0.4525,
+      "grad_norm": 0.5954529047012329,
+      "learning_rate": 0.00011005025125628141,
+      "loss": 0.5403,
+      "step": 905
+    },
+    {
+      "epoch": 0.453,
+      "grad_norm": 0.5809349417686462,
+      "learning_rate": 0.0001099497487437186,
+      "loss": 0.3885,
+      "step": 906
+    },
+    {
+      "epoch": 0.4535,
+      "grad_norm": 0.5929345488548279,
+      "learning_rate": 0.00010984924623115578,
+      "loss": 0.4501,
+      "step": 907
+    },
+    {
+      "epoch": 0.454,
+      "grad_norm": 0.6421545147895813,
+      "learning_rate": 0.00010974874371859296,
+      "loss": 0.5641,
+      "step": 908
+    },
+    {
+      "epoch": 0.4545,
+      "grad_norm": 0.5908579230308533,
+      "learning_rate": 0.00010964824120603017,
+      "loss": 0.4894,
+      "step": 909
+    },
+    {
+      "epoch": 0.455,
+      "grad_norm": 0.6923574805259705,
+      "learning_rate": 0.00010954773869346736,
+      "loss": 0.5802,
+      "step": 910
+    },
+    {
+      "epoch": 0.4555,
+      "grad_norm": 0.604924201965332,
+      "learning_rate": 0.00010944723618090453,
+      "loss": 0.4278,
+      "step": 911
+    },
+    {
+      "epoch": 0.456,
+      "grad_norm": 0.7144548892974854,
+      "learning_rate": 0.00010934673366834172,
+      "loss": 0.5649,
+      "step": 912
+    },
+    {
+      "epoch": 0.4565,
+      "grad_norm": 0.6487929224967957,
+      "learning_rate": 0.0001092462311557789,
+      "loss": 0.5238,
+      "step": 913
+    },
+    {
+      "epoch": 0.457,
+      "grad_norm": 0.6377905607223511,
+      "learning_rate": 0.00010914572864321609,
+      "loss": 0.4382,
+      "step": 914
+    },
+    {
+      "epoch": 0.4575,
+      "grad_norm": 0.619849443435669,
+      "learning_rate": 0.00010904522613065326,
+      "loss": 0.4302,
+      "step": 915
+    },
+    {
+      "epoch": 0.458,
+      "grad_norm": 0.6150761842727661,
+      "learning_rate": 0.00010894472361809045,
+      "loss": 0.3966,
+      "step": 916
+    },
+    {
+      "epoch": 0.4585,
+      "grad_norm": 0.5515437722206116,
+      "learning_rate": 0.00010884422110552763,
+      "loss": 0.3679,
+      "step": 917
+    },
+    {
+      "epoch": 0.459,
+      "grad_norm": 0.6204591393470764,
+      "learning_rate": 0.00010874371859296483,
+      "loss": 0.4727,
+      "step": 918
+    },
+    {
+      "epoch": 0.4595,
+      "grad_norm": 0.6051822304725647,
+      "learning_rate": 0.00010864321608040202,
+      "loss": 0.4197,
+      "step": 919
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7983821630477905,
+      "learning_rate": 0.00010854271356783921,
+      "loss": 0.5484,
+      "step": 920
+    },
+    {
+      "epoch": 0.4605,
+      "grad_norm": 0.6586692333221436,
+      "learning_rate": 0.0001084422110552764,
+      "loss": 0.4537,
+      "step": 921
+    },
+    {
+      "epoch": 0.461,
+      "grad_norm": 0.71335369348526,
+      "learning_rate": 0.00010834170854271357,
+      "loss": 0.3546,
+      "step": 922
+    },
+    {
+      "epoch": 0.4615,
+      "grad_norm": 0.5574732422828674,
+      "learning_rate": 0.00010824120603015075,
+      "loss": 0.3377,
+      "step": 923
+    },
+    {
+      "epoch": 0.462,
+      "grad_norm": 0.6498692631721497,
+      "learning_rate": 0.00010814070351758794,
+      "loss": 0.4477,
+      "step": 924
+    },
+    {
+      "epoch": 0.4625,
+      "grad_norm": 0.6718057990074158,
+      "learning_rate": 0.00010804020100502513,
+      "loss": 0.4368,
+      "step": 925
+    },
+    {
+      "epoch": 0.463,
+      "grad_norm": 0.7341487407684326,
+      "learning_rate": 0.00010793969849246233,
+      "loss": 0.4737,
+      "step": 926
+    },
+    {
+      "epoch": 0.4635,
+      "grad_norm": 0.6624234318733215,
+      "learning_rate": 0.00010783919597989951,
+      "loss": 0.4086,
+      "step": 927
+    },
+    {
+      "epoch": 0.464,
+      "grad_norm": 0.6753014922142029,
+      "learning_rate": 0.0001077386934673367,
+      "loss": 0.4881,
+      "step": 928
+    },
+    {
+      "epoch": 0.4645,
+      "grad_norm": 0.7022872567176819,
+      "learning_rate": 0.00010763819095477387,
+      "loss": 0.4249,
+      "step": 929
+    },
+    {
+      "epoch": 0.465,
+      "grad_norm": 0.6022166609764099,
+      "learning_rate": 0.00010753768844221106,
+      "loss": 0.3666,
+      "step": 930
+    },
+    {
+      "epoch": 0.4655,
+      "grad_norm": 0.7409123182296753,
+      "learning_rate": 0.00010743718592964825,
+      "loss": 0.503,
+      "step": 931
+    },
+    {
+      "epoch": 0.466,
+      "grad_norm": 0.9237306714057922,
+      "learning_rate": 0.00010733668341708543,
+      "loss": 0.4396,
+      "step": 932
+    },
+    {
+      "epoch": 0.4665,
+      "grad_norm": 0.7058731913566589,
+      "learning_rate": 0.0001072361809045226,
+      "loss": 0.373,
+      "step": 933
+    },
+    {
+      "epoch": 0.467,
+      "grad_norm": 0.6318204402923584,
+      "learning_rate": 0.00010713567839195982,
+      "loss": 0.4561,
+      "step": 934
+    },
+    {
+      "epoch": 0.4675,
+      "grad_norm": 0.6592294573783875,
+      "learning_rate": 0.00010703517587939699,
+      "loss": 0.378,
+      "step": 935
+    },
+    {
+      "epoch": 0.468,
+      "grad_norm": 0.7681623697280884,
+      "learning_rate": 0.00010693467336683418,
+      "loss": 0.6095,
+      "step": 936
+    },
+    {
+      "epoch": 0.4685,
+      "grad_norm": 0.7231519818305969,
+      "learning_rate": 0.00010683417085427136,
+      "loss": 0.4299,
+      "step": 937
+    },
+    {
+      "epoch": 0.469,
+      "grad_norm": 0.6852970123291016,
+      "learning_rate": 0.00010673366834170855,
+      "loss": 0.3859,
+      "step": 938
+    },
+    {
+      "epoch": 0.4695,
+      "grad_norm": 0.8438256978988647,
+      "learning_rate": 0.00010663316582914574,
+      "loss": 0.5052,
+      "step": 939
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6184353828430176,
+      "learning_rate": 0.00010653266331658291,
+      "loss": 0.3172,
+      "step": 940
+    },
+    {
+      "epoch": 0.4705,
+      "grad_norm": 0.6000133156776428,
+      "learning_rate": 0.0001064321608040201,
+      "loss": 0.3075,
+      "step": 941
+    },
+    {
+      "epoch": 0.471,
+      "grad_norm": 0.7016804814338684,
+      "learning_rate": 0.00010633165829145728,
+      "loss": 0.453,
+      "step": 942
+    },
+    {
+      "epoch": 0.4715,
+      "grad_norm": 0.5929070115089417,
+      "learning_rate": 0.00010623115577889448,
+      "loss": 0.3368,
+      "step": 943
+    },
+    {
+      "epoch": 0.472,
+      "grad_norm": 0.6257913112640381,
+      "learning_rate": 0.00010613065326633167,
+      "loss": 0.3401,
+      "step": 944
+    },
+    {
+      "epoch": 0.4725,
+      "grad_norm": 0.7872753143310547,
+      "learning_rate": 0.00010603015075376886,
+      "loss": 0.5095,
+      "step": 945
+    },
+    {
+      "epoch": 0.473,
+      "grad_norm": 0.6129568815231323,
+      "learning_rate": 0.00010592964824120604,
+      "loss": 0.3022,
+      "step": 946
+    },
+    {
+      "epoch": 0.4735,
+      "grad_norm": 0.5946791768074036,
+      "learning_rate": 0.00010582914572864322,
+      "loss": 0.311,
+      "step": 947
+    },
+    {
+      "epoch": 0.474,
+      "grad_norm": 0.7823770046234131,
+      "learning_rate": 0.0001057286432160804,
+      "loss": 0.443,
+      "step": 948
+    },
+    {
+      "epoch": 0.4745,
+      "grad_norm": 0.6603112816810608,
+      "learning_rate": 0.00010562814070351759,
+      "loss": 0.3118,
+      "step": 949
+    },
+    {
+      "epoch": 0.475,
+      "grad_norm": 0.9254936575889587,
+      "learning_rate": 0.00010552763819095478,
+      "loss": 0.3847,
+      "step": 950
+    },
+    {
+      "epoch": 0.4755,
+      "grad_norm": 0.4870583713054657,
+      "learning_rate": 0.00010542713567839198,
+      "loss": 0.3878,
+      "step": 951
+    },
+    {
+      "epoch": 0.476,
+      "grad_norm": 0.6767250299453735,
+      "learning_rate": 0.00010532663316582916,
+      "loss": 0.5613,
+      "step": 952
+    },
+    {
+      "epoch": 0.4765,
+      "grad_norm": 0.5829386711120605,
+      "learning_rate": 0.00010522613065326635,
+      "loss": 0.34,
+      "step": 953
+    },
+    {
+      "epoch": 0.477,
+      "grad_norm": 0.5983999967575073,
+      "learning_rate": 0.00010512562814070352,
+      "loss": 0.4066,
+      "step": 954
+    },
+    {
+      "epoch": 0.4775,
+      "grad_norm": 0.6167608499526978,
+      "learning_rate": 0.00010502512562814071,
+      "loss": 0.4689,
+      "step": 955
+    },
+    {
+      "epoch": 0.478,
+      "grad_norm": 0.6133597493171692,
+      "learning_rate": 0.0001049246231155779,
+      "loss": 0.441,
+      "step": 956
+    },
+    {
+      "epoch": 0.4785,
+      "grad_norm": 0.6292927861213684,
+      "learning_rate": 0.00010482412060301508,
+      "loss": 0.5114,
+      "step": 957
+    },
+    {
+      "epoch": 0.479,
+      "grad_norm": 0.5658861398696899,
+      "learning_rate": 0.00010472361809045225,
+      "loss": 0.3908,
+      "step": 958
+    },
+    {
+      "epoch": 0.4795,
+      "grad_norm": 0.7186408638954163,
+      "learning_rate": 0.00010462311557788944,
+      "loss": 0.7191,
+      "step": 959
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6248448491096497,
+      "learning_rate": 0.00010452261306532664,
+      "loss": 0.5158,
+      "step": 960
+    },
+    {
+      "epoch": 0.4805,
+      "grad_norm": 0.6519450545310974,
+      "learning_rate": 0.00010442211055276383,
+      "loss": 0.437,
+      "step": 961
+    },
+    {
+      "epoch": 0.481,
+      "grad_norm": 0.6064294576644897,
+      "learning_rate": 0.00010432160804020101,
+      "loss": 0.3649,
+      "step": 962
+    },
+    {
+      "epoch": 0.4815,
+      "grad_norm": 0.7025746703147888,
+      "learning_rate": 0.0001042211055276382,
+      "loss": 0.549,
+      "step": 963
+    },
+    {
+      "epoch": 0.482,
+      "grad_norm": 0.7231892943382263,
+      "learning_rate": 0.00010412060301507539,
+      "loss": 0.5307,
+      "step": 964
+    },
+    {
+      "epoch": 0.4825,
+      "grad_norm": 0.5593419671058655,
+      "learning_rate": 0.00010402010050251256,
+      "loss": 0.3937,
+      "step": 965
+    },
+    {
+      "epoch": 0.483,
+      "grad_norm": 0.6470860838890076,
+      "learning_rate": 0.00010391959798994975,
+      "loss": 0.4257,
+      "step": 966
+    },
+    {
+      "epoch": 0.4835,
+      "grad_norm": 0.698374330997467,
+      "learning_rate": 0.00010381909547738693,
+      "loss": 0.4742,
+      "step": 967
+    },
+    {
+      "epoch": 0.484,
+      "grad_norm": 0.6035527586936951,
+      "learning_rate": 0.00010371859296482413,
+      "loss": 0.4315,
+      "step": 968
+    },
+    {
+      "epoch": 0.4845,
+      "grad_norm": 0.7081201076507568,
+      "learning_rate": 0.00010361809045226132,
+      "loss": 0.5021,
+      "step": 969
+    },
+    {
+      "epoch": 0.485,
+      "grad_norm": 0.7783226370811462,
+      "learning_rate": 0.0001035175879396985,
+      "loss": 0.5462,
+      "step": 970
+    },
+    {
+      "epoch": 0.4855,
+      "grad_norm": 0.8135401606559753,
+      "learning_rate": 0.00010341708542713569,
+      "loss": 0.6459,
+      "step": 971
+    },
+    {
+      "epoch": 0.486,
+      "grad_norm": 0.6614494919776917,
+      "learning_rate": 0.00010331658291457286,
+      "loss": 0.4222,
+      "step": 972
+    },
+    {
+      "epoch": 0.4865,
+      "grad_norm": 0.6248466968536377,
+      "learning_rate": 0.00010321608040201005,
+      "loss": 0.4398,
+      "step": 973
+    },
+    {
+      "epoch": 0.487,
+      "grad_norm": 0.7052192091941833,
+      "learning_rate": 0.00010311557788944724,
+      "loss": 0.5272,
+      "step": 974
+    },
+    {
+      "epoch": 0.4875,
+      "grad_norm": 0.5158700942993164,
+      "learning_rate": 0.00010301507537688442,
+      "loss": 0.3126,
+      "step": 975
+    },
+    {
+      "epoch": 0.488,
+      "grad_norm": 0.6931352019309998,
+      "learning_rate": 0.00010291457286432162,
+      "loss": 0.4884,
+      "step": 976
+    },
+    {
+      "epoch": 0.4885,
+      "grad_norm": 0.7146469354629517,
+      "learning_rate": 0.00010281407035175881,
+      "loss": 0.5665,
+      "step": 977
+    },
+    {
+      "epoch": 0.489,
+      "grad_norm": 0.7529250383377075,
+      "learning_rate": 0.00010271356783919598,
+      "loss": 0.5632,
+      "step": 978
+    },
+    {
+      "epoch": 0.4895,
+      "grad_norm": 0.7609817385673523,
+      "learning_rate": 0.00010261306532663317,
+      "loss": 0.464,
+      "step": 979
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7257074117660522,
+      "learning_rate": 0.00010251256281407036,
+      "loss": 0.5138,
+      "step": 980
+    },
+    {
+      "epoch": 0.4905,
+      "grad_norm": 0.7832579016685486,
+      "learning_rate": 0.00010241206030150754,
+      "loss": 0.5021,
+      "step": 981
+    },
+    {
+      "epoch": 0.491,
+      "grad_norm": 0.6656126379966736,
+      "learning_rate": 0.00010231155778894473,
+      "loss": 0.409,
+      "step": 982
+    },
+    {
+      "epoch": 0.4915,
+      "grad_norm": 0.652890682220459,
+      "learning_rate": 0.0001022110552763819,
+      "loss": 0.4339,
+      "step": 983
+    },
+    {
+      "epoch": 0.492,
+      "grad_norm": 0.8393736481666565,
+      "learning_rate": 0.00010211055276381909,
+      "loss": 0.4787,
+      "step": 984
+    },
+    {
+      "epoch": 0.4925,
+      "grad_norm": 0.7713173031806946,
+      "learning_rate": 0.00010201005025125629,
+      "loss": 0.4713,
+      "step": 985
+    },
+    {
+      "epoch": 0.493,
+      "grad_norm": 0.5690344572067261,
+      "learning_rate": 0.00010190954773869348,
+      "loss": 0.3239,
+      "step": 986
+    },
+    {
+      "epoch": 0.4935,
+      "grad_norm": 0.7821685075759888,
+      "learning_rate": 0.00010180904522613066,
+      "loss": 0.533,
+      "step": 987
+    },
+    {
+      "epoch": 0.494,
+      "grad_norm": 0.6914284229278564,
+      "learning_rate": 0.00010170854271356785,
+      "loss": 0.3809,
+      "step": 988
+    },
+    {
+      "epoch": 0.4945,
+      "grad_norm": 0.5622161626815796,
+      "learning_rate": 0.00010160804020100503,
+      "loss": 0.3488,
+      "step": 989
+    },
+    {
+      "epoch": 0.495,
+      "grad_norm": 0.7163942456245422,
+      "learning_rate": 0.00010150753768844221,
+      "loss": 0.4119,
+      "step": 990
+    },
+    {
+      "epoch": 0.4955,
+      "grad_norm": 0.6110238432884216,
+      "learning_rate": 0.0001014070351758794,
+      "loss": 0.3785,
+      "step": 991
+    },
+    {
+      "epoch": 0.496,
+      "grad_norm": 0.6549639105796814,
+      "learning_rate": 0.00010130653266331658,
+      "loss": 0.3723,
+      "step": 992
+    },
+    {
+      "epoch": 0.4965,
+      "grad_norm": 0.7609460353851318,
+      "learning_rate": 0.00010120603015075378,
+      "loss": 0.4579,
+      "step": 993
+    },
+    {
+      "epoch": 0.497,
+      "grad_norm": 0.9316160678863525,
+      "learning_rate": 0.00010110552763819097,
+      "loss": 0.4469,
+      "step": 994
+    },
+    {
+      "epoch": 0.4975,
+      "grad_norm": 0.818123996257782,
+      "learning_rate": 0.00010100502512562815,
+      "loss": 0.5094,
+      "step": 995
+    },
+    {
+      "epoch": 0.498,
+      "grad_norm": 0.625618577003479,
+      "learning_rate": 0.00010090452261306533,
+      "loss": 0.3278,
+      "step": 996
+    },
+    {
+      "epoch": 0.4985,
+      "grad_norm": 0.7071962952613831,
+      "learning_rate": 0.00010080402010050251,
+      "loss": 0.3882,
+      "step": 997
+    },
+    {
+      "epoch": 0.499,
+      "grad_norm": 0.8827985525131226,
+      "learning_rate": 0.0001007035175879397,
+      "loss": 0.4107,
+      "step": 998
+    },
+    {
+      "epoch": 0.4995,
+      "grad_norm": 0.6694146990776062,
+      "learning_rate": 0.00010060301507537689,
+      "loss": 0.3112,
+      "step": 999
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6154847741127014,
+      "learning_rate": 0.00010050251256281407,
+      "loss": 0.3106,
+      "step": 1000
+    },
+    {
+      "epoch": 0.5005,
+      "grad_norm": 0.5844072103500366,
+      "learning_rate": 0.00010040201005025127,
+      "loss": 0.6231,
+      "step": 1001
+    },
+    {
+      "epoch": 0.501,
+      "grad_norm": 0.5295509696006775,
+      "learning_rate": 0.00010030150753768846,
+      "loss": 0.4053,
+      "step": 1002
+    },
+    {
+      "epoch": 0.5015,
+      "grad_norm": 0.6414412260055542,
+      "learning_rate": 0.00010020100502512563,
+      "loss": 0.632,
+      "step": 1003
+    },
+    {
+      "epoch": 0.502,
+      "grad_norm": 0.6924289464950562,
+      "learning_rate": 0.00010010050251256282,
+      "loss": 0.6125,
+      "step": 1004
+    },
+    {
+      "epoch": 0.5025,
+      "grad_norm": 0.6129923462867737,
+      "learning_rate": 0.0001,
+      "loss": 0.4712,
+      "step": 1005
+    },
+    {
+      "epoch": 0.503,
+      "grad_norm": 0.6037617921829224,
+      "learning_rate": 9.989949748743719e-05,
+      "loss": 0.5032,
+      "step": 1006
+    },
+    {
+      "epoch": 0.5035,
+      "grad_norm": 0.7331117987632751,
+      "learning_rate": 9.979899497487438e-05,
+      "loss": 0.6122,
+      "step": 1007
+    },
+    {
+      "epoch": 0.504,
+      "grad_norm": 0.6575391292572021,
+      "learning_rate": 9.969849246231156e-05,
+      "loss": 0.5571,
+      "step": 1008
+    },
+    {
+      "epoch": 0.5045,
+      "grad_norm": 0.6595961451530457,
+      "learning_rate": 9.959798994974875e-05,
+      "loss": 0.5673,
+      "step": 1009
+    },
+    {
+      "epoch": 0.505,
+      "grad_norm": 0.6597159504890442,
+      "learning_rate": 9.949748743718594e-05,
+      "loss": 0.519,
+      "step": 1010
+    },
+    {
+      "epoch": 0.5055,
+      "grad_norm": 0.6271268725395203,
+      "learning_rate": 9.939698492462311e-05,
+      "loss": 0.5459,
+      "step": 1011
+    },
+    {
+      "epoch": 0.506,
+      "grad_norm": 0.6806215643882751,
+      "learning_rate": 9.929648241206031e-05,
+      "loss": 0.4549,
+      "step": 1012
+    },
+    {
+      "epoch": 0.5065,
+      "grad_norm": 0.5191720128059387,
+      "learning_rate": 9.91959798994975e-05,
+      "loss": 0.3891,
+      "step": 1013
+    },
+    {
+      "epoch": 0.507,
+      "grad_norm": 0.7007437944412231,
+      "learning_rate": 9.909547738693468e-05,
+      "loss": 0.6303,
+      "step": 1014
+    },
+    {
+      "epoch": 0.5075,
+      "grad_norm": 0.5263321995735168,
+      "learning_rate": 9.899497487437186e-05,
+      "loss": 0.3472,
+      "step": 1015
+    },
+    {
+      "epoch": 0.508,
+      "grad_norm": 0.6780453324317932,
+      "learning_rate": 9.889447236180906e-05,
+      "loss": 0.5451,
+      "step": 1016
+    },
+    {
+      "epoch": 0.5085,
+      "grad_norm": 0.5239865183830261,
+      "learning_rate": 9.879396984924624e-05,
+      "loss": 0.336,
+      "step": 1017
+    },
+    {
+      "epoch": 0.509,
+      "grad_norm": 0.6514967679977417,
+      "learning_rate": 9.869346733668342e-05,
+      "loss": 0.4567,
+      "step": 1018
+    },
+    {
+      "epoch": 0.5095,
+      "grad_norm": 0.6086786389350891,
+      "learning_rate": 9.85929648241206e-05,
+      "loss": 0.4689,
+      "step": 1019
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6586118340492249,
+      "learning_rate": 9.84924623115578e-05,
+      "loss": 0.4911,
+      "step": 1020
+    },
+    {
+      "epoch": 0.5105,
+      "grad_norm": 0.6027961373329163,
+      "learning_rate": 9.839195979899497e-05,
+      "loss": 0.4397,
+      "step": 1021
+    },
+    {
+      "epoch": 0.511,
+      "grad_norm": 0.6764188408851624,
+      "learning_rate": 9.829145728643216e-05,
+      "loss": 0.4311,
+      "step": 1022
+    },
+    {
+      "epoch": 0.5115,
+      "grad_norm": 0.6616021990776062,
+      "learning_rate": 9.819095477386935e-05,
+      "loss": 0.4784,
+      "step": 1023
+    },
+    {
+      "epoch": 0.512,
+      "grad_norm": 0.7057104110717773,
+      "learning_rate": 9.809045226130655e-05,
+      "loss": 0.5114,
+      "step": 1024
+    },
+    {
+      "epoch": 0.5125,
+      "grad_norm": 0.7854090929031372,
+      "learning_rate": 9.798994974874372e-05,
+      "loss": 0.6063,
+      "step": 1025
+    },
+    {
+      "epoch": 0.513,
+      "grad_norm": 0.5893591046333313,
+      "learning_rate": 9.788944723618091e-05,
+      "loss": 0.4168,
+      "step": 1026
+    },
+    {
+      "epoch": 0.5135,
+      "grad_norm": 0.6921731233596802,
+      "learning_rate": 9.77889447236181e-05,
+      "loss": 0.4274,
+      "step": 1027
+    },
+    {
+      "epoch": 0.514,
+      "grad_norm": 0.5702146291732788,
+      "learning_rate": 9.768844221105528e-05,
+      "loss": 0.3499,
+      "step": 1028
+    },
+    {
+      "epoch": 0.5145,
+      "grad_norm": 0.662333071231842,
+      "learning_rate": 9.758793969849247e-05,
+      "loss": 0.4208,
+      "step": 1029
+    },
+    {
+      "epoch": 0.515,
+      "grad_norm": 0.6480969190597534,
+      "learning_rate": 9.748743718592965e-05,
+      "loss": 0.441,
+      "step": 1030
+    },
+    {
+      "epoch": 0.5155,
+      "grad_norm": 0.7060957551002502,
+      "learning_rate": 9.738693467336684e-05,
+      "loss": 0.4636,
+      "step": 1031
+    },
+    {
+      "epoch": 0.516,
+      "grad_norm": 0.6555445194244385,
+      "learning_rate": 9.728643216080403e-05,
+      "loss": 0.4068,
+      "step": 1032
+    },
+    {
+      "epoch": 0.5165,
+      "grad_norm": 0.642917275428772,
+      "learning_rate": 9.718592964824121e-05,
+      "loss": 0.4064,
+      "step": 1033
+    },
+    {
+      "epoch": 0.517,
+      "grad_norm": 0.598086953163147,
+      "learning_rate": 9.70854271356784e-05,
+      "loss": 0.3381,
+      "step": 1034
+    },
+    {
+      "epoch": 0.5175,
+      "grad_norm": 0.675029993057251,
+      "learning_rate": 9.698492462311559e-05,
+      "loss": 0.3879,
+      "step": 1035
+    },
+    {
+      "epoch": 0.518,
+      "grad_norm": 0.750040590763092,
+      "learning_rate": 9.688442211055276e-05,
+      "loss": 0.5107,
+      "step": 1036
+    },
+    {
+      "epoch": 0.5185,
+      "grad_norm": 0.6424611806869507,
+      "learning_rate": 9.678391959798996e-05,
+      "loss": 0.3954,
+      "step": 1037
+    },
+    {
+      "epoch": 0.519,
+      "grad_norm": 0.8020350337028503,
+      "learning_rate": 9.668341708542715e-05,
+      "loss": 0.4697,
+      "step": 1038
+    },
+    {
+      "epoch": 0.5195,
+      "grad_norm": 0.6393791437149048,
+      "learning_rate": 9.658291457286432e-05,
+      "loss": 0.3678,
+      "step": 1039
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7228180766105652,
+      "learning_rate": 9.64824120603015e-05,
+      "loss": 0.4354,
+      "step": 1040
+    },
+    {
+      "epoch": 0.5205,
+      "grad_norm": 0.7023571133613586,
+      "learning_rate": 9.63819095477387e-05,
+      "loss": 0.3797,
+      "step": 1041
+    },
+    {
+      "epoch": 0.521,
+      "grad_norm": 0.571578860282898,
+      "learning_rate": 9.628140703517589e-05,
+      "loss": 0.3097,
+      "step": 1042
+    },
+    {
+      "epoch": 0.5215,
+      "grad_norm": 0.661182165145874,
+      "learning_rate": 9.618090452261306e-05,
+      "loss": 0.3303,
+      "step": 1043
+    },
+    {
+      "epoch": 0.522,
+      "grad_norm": 0.6941015720367432,
+      "learning_rate": 9.608040201005025e-05,
+      "loss": 0.3458,
+      "step": 1044
+    },
+    {
+      "epoch": 0.5225,
+      "grad_norm": 0.7076554894447327,
+      "learning_rate": 9.597989949748745e-05,
+      "loss": 0.3389,
+      "step": 1045
+    },
+    {
+      "epoch": 0.523,
+      "grad_norm": 0.6422895193099976,
+      "learning_rate": 9.587939698492462e-05,
+      "loss": 0.3601,
+      "step": 1046
+    },
+    {
+      "epoch": 0.5235,
+      "grad_norm": 0.682384192943573,
+      "learning_rate": 9.577889447236181e-05,
+      "loss": 0.3901,
+      "step": 1047
+    },
+    {
+      "epoch": 0.524,
+      "grad_norm": 0.7263312935829163,
+      "learning_rate": 9.5678391959799e-05,
+      "loss": 0.3551,
+      "step": 1048
+    },
+    {
+      "epoch": 0.5245,
+      "grad_norm": 0.7216951251029968,
+      "learning_rate": 9.55778894472362e-05,
+      "loss": 0.3802,
+      "step": 1049
+    },
+    {
+      "epoch": 0.525,
+      "grad_norm": 0.7573158740997314,
+      "learning_rate": 9.547738693467337e-05,
+      "loss": 0.3542,
+      "step": 1050
+    },
+    {
+      "epoch": 0.5255,
+      "grad_norm": 0.4903378486633301,
+      "learning_rate": 9.537688442211056e-05,
+      "loss": 0.3604,
+      "step": 1051
+    },
+    {
+      "epoch": 0.526,
+      "grad_norm": 0.5113053917884827,
+      "learning_rate": 9.527638190954774e-05,
+      "loss": 0.3601,
+      "step": 1052
+    },
+    {
+      "epoch": 0.5265,
+      "grad_norm": 0.5002259016036987,
+      "learning_rate": 9.517587939698493e-05,
+      "loss": 0.3407,
+      "step": 1053
+    },
+    {
+      "epoch": 0.527,
+      "grad_norm": 0.6394708752632141,
+      "learning_rate": 9.507537688442212e-05,
+      "loss": 0.4852,
+      "step": 1054
+    },
+    {
+      "epoch": 0.5275,
+      "grad_norm": 0.6457741856575012,
+      "learning_rate": 9.49748743718593e-05,
+      "loss": 0.5529,
+      "step": 1055
+    },
+    {
+      "epoch": 0.528,
+      "grad_norm": 0.6499139666557312,
+      "learning_rate": 9.487437185929649e-05,
+      "loss": 0.4806,
+      "step": 1056
+    },
+    {
+      "epoch": 0.5285,
+      "grad_norm": 0.6220322847366333,
+      "learning_rate": 9.477386934673366e-05,
+      "loss": 0.4212,
+      "step": 1057
+    },
+    {
+      "epoch": 0.529,
+      "grad_norm": 0.6322537660598755,
+      "learning_rate": 9.467336683417086e-05,
+      "loss": 0.4739,
+      "step": 1058
+    },
+    {
+      "epoch": 0.5295,
+      "grad_norm": 0.5648052096366882,
+      "learning_rate": 9.457286432160805e-05,
+      "loss": 0.3844,
+      "step": 1059
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.663206934928894,
+      "learning_rate": 9.447236180904523e-05,
+      "loss": 0.493,
+      "step": 1060
+    },
+    {
+      "epoch": 0.5305,
+      "grad_norm": 0.6135300993919373,
+      "learning_rate": 9.437185929648241e-05,
+      "loss": 0.4022,
+      "step": 1061
+    },
+    {
+      "epoch": 0.531,
+      "grad_norm": 0.6415677666664124,
+      "learning_rate": 9.427135678391961e-05,
+      "loss": 0.482,
+      "step": 1062
+    },
+    {
+      "epoch": 0.5315,
+      "grad_norm": 0.7749935388565063,
+      "learning_rate": 9.41708542713568e-05,
+      "loss": 0.4762,
+      "step": 1063
+    },
+    {
+      "epoch": 0.532,
+      "grad_norm": 0.7408579587936401,
+      "learning_rate": 9.407035175879397e-05,
+      "loss": 0.6317,
+      "step": 1064
+    },
+    {
+      "epoch": 0.5325,
+      "grad_norm": 0.6152773499488831,
+      "learning_rate": 9.396984924623115e-05,
+      "loss": 0.422,
+      "step": 1065
+    },
+    {
+      "epoch": 0.533,
+      "grad_norm": 0.7029319405555725,
+      "learning_rate": 9.386934673366835e-05,
+      "loss": 0.5893,
+      "step": 1066
+    },
+    {
+      "epoch": 0.5335,
+      "grad_norm": 0.6096486449241638,
+      "learning_rate": 9.376884422110554e-05,
+      "loss": 0.4512,
+      "step": 1067
+    },
+    {
+      "epoch": 0.534,
+      "grad_norm": 0.6801251769065857,
+      "learning_rate": 9.366834170854271e-05,
+      "loss": 0.4588,
+      "step": 1068
+    },
+    {
+      "epoch": 0.5345,
+      "grad_norm": 0.5919814705848694,
+      "learning_rate": 9.35678391959799e-05,
+      "loss": 0.3553,
+      "step": 1069
+    },
+    {
+      "epoch": 0.535,
+      "grad_norm": 0.6783584952354431,
+      "learning_rate": 9.34673366834171e-05,
+      "loss": 0.4355,
+      "step": 1070
+    },
+    {
+      "epoch": 0.5355,
+      "grad_norm": 0.7030637860298157,
+      "learning_rate": 9.336683417085427e-05,
+      "loss": 0.4849,
+      "step": 1071
+    },
+    {
+      "epoch": 0.536,
+      "grad_norm": 0.5400413274765015,
+      "learning_rate": 9.326633165829146e-05,
+      "loss": 0.2966,
+      "step": 1072
+    },
+    {
+      "epoch": 0.5365,
+      "grad_norm": 0.6417726874351501,
+      "learning_rate": 9.316582914572864e-05,
+      "loss": 0.3807,
+      "step": 1073
+    },
+    {
+      "epoch": 0.537,
+      "grad_norm": 0.6106700897216797,
+      "learning_rate": 9.306532663316585e-05,
+      "loss": 0.4131,
+      "step": 1074
+    },
+    {
+      "epoch": 0.5375,
+      "grad_norm": 0.7506864070892334,
+      "learning_rate": 9.296482412060302e-05,
+      "loss": 0.4795,
+      "step": 1075
+    },
+    {
+      "epoch": 0.538,
+      "grad_norm": 0.7540200352668762,
+      "learning_rate": 9.28643216080402e-05,
+      "loss": 0.4763,
+      "step": 1076
+    },
+    {
+      "epoch": 0.5385,
+      "grad_norm": 0.6447717547416687,
+      "learning_rate": 9.276381909547739e-05,
+      "loss": 0.3871,
+      "step": 1077
+    },
+    {
+      "epoch": 0.539,
+      "grad_norm": 0.6654728055000305,
+      "learning_rate": 9.266331658291458e-05,
+      "loss": 0.416,
+      "step": 1078
+    },
+    {
+      "epoch": 0.5395,
+      "grad_norm": 0.6780759692192078,
+      "learning_rate": 9.256281407035176e-05,
+      "loss": 0.4099,
+      "step": 1079
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5935434103012085,
+      "learning_rate": 9.246231155778895e-05,
+      "loss": 0.3922,
+      "step": 1080
+    },
+    {
+      "epoch": 0.5405,
+      "grad_norm": 0.6018031239509583,
+      "learning_rate": 9.236180904522614e-05,
+      "loss": 0.2972,
+      "step": 1081
+    },
+    {
+      "epoch": 0.541,
+      "grad_norm": 0.7132726311683655,
+      "learning_rate": 9.226130653266331e-05,
+      "loss": 0.4413,
+      "step": 1082
+    },
+    {
+      "epoch": 0.5415,
+      "grad_norm": 0.6959756016731262,
+      "learning_rate": 9.216080402010051e-05,
+      "loss": 0.3463,
+      "step": 1083
+    },
+    {
+      "epoch": 0.542,
+      "grad_norm": 0.6992121934890747,
+      "learning_rate": 9.20603015075377e-05,
+      "loss": 0.4327,
+      "step": 1084
+    },
+    {
+      "epoch": 0.5425,
+      "grad_norm": 0.6635663509368896,
+      "learning_rate": 9.195979899497488e-05,
+      "loss": 0.3912,
+      "step": 1085
+    },
+    {
+      "epoch": 0.543,
+      "grad_norm": 0.7486253976821899,
+      "learning_rate": 9.185929648241206e-05,
+      "loss": 0.4889,
+      "step": 1086
+    },
+    {
+      "epoch": 0.5435,
+      "grad_norm": 1.0061055421829224,
+      "learning_rate": 9.175879396984926e-05,
+      "loss": 0.5145,
+      "step": 1087
+    },
+    {
+      "epoch": 0.544,
+      "grad_norm": 0.7225772738456726,
+      "learning_rate": 9.165829145728644e-05,
+      "loss": 0.3539,
+      "step": 1088
+    },
+    {
+      "epoch": 0.5445,
+      "grad_norm": 0.726796567440033,
+      "learning_rate": 9.155778894472362e-05,
+      "loss": 0.4262,
+      "step": 1089
+    },
+    {
+      "epoch": 0.545,
+      "grad_norm": 0.6302950978279114,
+      "learning_rate": 9.14572864321608e-05,
+      "loss": 0.3255,
+      "step": 1090
+    },
+    {
+      "epoch": 0.5455,
+      "grad_norm": 0.6314042806625366,
+      "learning_rate": 9.1356783919598e-05,
+      "loss": 0.3376,
+      "step": 1091
+    },
+    {
+      "epoch": 0.546,
+      "grad_norm": 0.5843379497528076,
+      "learning_rate": 9.125628140703519e-05,
+      "loss": 0.3048,
+      "step": 1092
+    },
+    {
+      "epoch": 0.5465,
+      "grad_norm": 0.7276121377944946,
+      "learning_rate": 9.115577889447236e-05,
+      "loss": 0.4047,
+      "step": 1093
+    },
+    {
+      "epoch": 0.547,
+      "grad_norm": 0.7290487885475159,
+      "learning_rate": 9.105527638190955e-05,
+      "loss": 0.4046,
+      "step": 1094
+    },
+    {
+      "epoch": 0.5475,
+      "grad_norm": 0.6247122287750244,
+      "learning_rate": 9.095477386934675e-05,
+      "loss": 0.3543,
+      "step": 1095
+    },
+    {
+      "epoch": 0.548,
+      "grad_norm": 0.6410446166992188,
+      "learning_rate": 9.085427135678392e-05,
+      "loss": 0.3401,
+      "step": 1096
+    },
+    {
+      "epoch": 0.5485,
+      "grad_norm": 0.60970538854599,
+      "learning_rate": 9.075376884422111e-05,
+      "loss": 0.3,
+      "step": 1097
+    },
+    {
+      "epoch": 0.549,
+      "grad_norm": 0.7480593919754028,
+      "learning_rate": 9.06532663316583e-05,
+      "loss": 0.3891,
+      "step": 1098
+    },
+    {
+      "epoch": 0.5495,
+      "grad_norm": 0.6796474456787109,
+      "learning_rate": 9.055276381909548e-05,
+      "loss": 0.3583,
+      "step": 1099
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7244755625724792,
+      "learning_rate": 9.045226130653267e-05,
+      "loss": 0.3465,
+      "step": 1100
+    },
+    {
+      "epoch": 0.5505,
+      "grad_norm": 0.6946145296096802,
+      "learning_rate": 9.035175879396985e-05,
+      "loss": 0.6943,
+      "step": 1101
+    },
+    {
+      "epoch": 0.551,
+      "grad_norm": 0.6643890142440796,
+      "learning_rate": 9.025125628140704e-05,
+      "loss": 0.6727,
+      "step": 1102
+    },
+    {
+      "epoch": 0.5515,
+      "grad_norm": 0.5823101997375488,
+      "learning_rate": 9.015075376884423e-05,
+      "loss": 0.5153,
+      "step": 1103
+    },
+    {
+      "epoch": 0.552,
+      "grad_norm": 0.47915422916412354,
+      "learning_rate": 9.005025125628141e-05,
+      "loss": 0.3779,
+      "step": 1104
+    },
+    {
+      "epoch": 0.5525,
+      "grad_norm": 0.6620208024978638,
+      "learning_rate": 8.99497487437186e-05,
+      "loss": 0.6675,
+      "step": 1105
+    },
+    {
+      "epoch": 0.553,
+      "grad_norm": 0.5241115093231201,
+      "learning_rate": 8.984924623115579e-05,
+      "loss": 0.3835,
+      "step": 1106
+    },
+    {
+      "epoch": 0.5535,
+      "grad_norm": 0.6850830316543579,
+      "learning_rate": 8.974874371859296e-05,
+      "loss": 0.6798,
+      "step": 1107
+    },
+    {
+      "epoch": 0.554,
+      "grad_norm": 0.5499701499938965,
+      "learning_rate": 8.964824120603016e-05,
+      "loss": 0.4342,
+      "step": 1108
+    },
+    {
+      "epoch": 0.5545,
+      "grad_norm": 0.6417487859725952,
+      "learning_rate": 8.954773869346734e-05,
+      "loss": 0.5344,
+      "step": 1109
+    },
+    {
+      "epoch": 0.555,
+      "grad_norm": 0.4796616733074188,
+      "learning_rate": 8.944723618090453e-05,
+      "loss": 0.3332,
+      "step": 1110
+    },
+    {
+      "epoch": 0.5555,
+      "grad_norm": 0.7013306617736816,
+      "learning_rate": 8.93467336683417e-05,
+      "loss": 0.6137,
+      "step": 1111
+    },
+    {
+      "epoch": 0.556,
+      "grad_norm": 0.5321751832962036,
+      "learning_rate": 8.92462311557789e-05,
+      "loss": 0.3642,
+      "step": 1112
+    },
+    {
+      "epoch": 0.5565,
+      "grad_norm": 0.6158413290977478,
+      "learning_rate": 8.914572864321609e-05,
+      "loss": 0.4137,
+      "step": 1113
+    },
+    {
+      "epoch": 0.557,
+      "grad_norm": 0.644307017326355,
+      "learning_rate": 8.904522613065326e-05,
+      "loss": 0.4125,
+      "step": 1114
+    },
+    {
+      "epoch": 0.5575,
+      "grad_norm": 0.8024719953536987,
+      "learning_rate": 8.894472361809045e-05,
+      "loss": 0.5208,
+      "step": 1115
+    },
+    {
+      "epoch": 0.558,
+      "grad_norm": 0.6219522953033447,
+      "learning_rate": 8.884422110552765e-05,
+      "loss": 0.4893,
+      "step": 1116
+    },
+    {
+      "epoch": 0.5585,
+      "grad_norm": 0.6608858704566956,
+      "learning_rate": 8.874371859296484e-05,
+      "loss": 0.4242,
+      "step": 1117
+    },
+    {
+      "epoch": 0.559,
+      "grad_norm": 0.6402149796485901,
+      "learning_rate": 8.864321608040201e-05,
+      "loss": 0.3416,
+      "step": 1118
+    },
+    {
+      "epoch": 0.5595,
+      "grad_norm": 0.6571792364120483,
+      "learning_rate": 8.85427135678392e-05,
+      "loss": 0.3799,
+      "step": 1119
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6188195943832397,
+      "learning_rate": 8.84422110552764e-05,
+      "loss": 0.4357,
+      "step": 1120
+    },
+    {
+      "epoch": 0.5605,
+      "grad_norm": 0.645401656627655,
+      "learning_rate": 8.834170854271357e-05,
+      "loss": 0.5226,
+      "step": 1121
+    },
+    {
+      "epoch": 0.561,
+      "grad_norm": 0.624073326587677,
+      "learning_rate": 8.824120603015076e-05,
+      "loss": 0.4199,
+      "step": 1122
+    },
+    {
+      "epoch": 0.5615,
+      "grad_norm": 0.6683102250099182,
+      "learning_rate": 8.814070351758794e-05,
+      "loss": 0.4546,
+      "step": 1123
+    },
+    {
+      "epoch": 0.562,
+      "grad_norm": 0.6345419883728027,
+      "learning_rate": 8.804020100502513e-05,
+      "loss": 0.3589,
+      "step": 1124
+    },
+    {
+      "epoch": 0.5625,
+      "grad_norm": 0.6211184859275818,
+      "learning_rate": 8.793969849246232e-05,
+      "loss": 0.4435,
+      "step": 1125
+    },
+    {
+      "epoch": 0.563,
+      "grad_norm": 0.6419554948806763,
+      "learning_rate": 8.78391959798995e-05,
+      "loss": 0.4246,
+      "step": 1126
+    },
+    {
+      "epoch": 0.5635,
+      "grad_norm": 0.7413788437843323,
+      "learning_rate": 8.773869346733669e-05,
+      "loss": 0.4205,
+      "step": 1127
+    },
+    {
+      "epoch": 0.564,
+      "grad_norm": 0.7201132774353027,
+      "learning_rate": 8.763819095477387e-05,
+      "loss": 0.4918,
+      "step": 1128
+    },
+    {
+      "epoch": 0.5645,
+      "grad_norm": 0.5728611946105957,
+      "learning_rate": 8.753768844221106e-05,
+      "loss": 0.3073,
+      "step": 1129
+    },
+    {
+      "epoch": 0.565,
+      "grad_norm": 0.5985531806945801,
+      "learning_rate": 8.743718592964825e-05,
+      "loss": 0.3352,
+      "step": 1130
+    },
+    {
+      "epoch": 0.5655,
+      "grad_norm": 0.6621812582015991,
+      "learning_rate": 8.733668341708543e-05,
+      "loss": 0.4379,
+      "step": 1131
+    },
+    {
+      "epoch": 0.566,
+      "grad_norm": 0.6426060199737549,
+      "learning_rate": 8.723618090452261e-05,
+      "loss": 0.3993,
+      "step": 1132
+    },
+    {
+      "epoch": 0.5665,
+      "grad_norm": 0.6904336810112,
+      "learning_rate": 8.713567839195981e-05,
+      "loss": 0.4256,
+      "step": 1133
+    },
+    {
+      "epoch": 0.567,
+      "grad_norm": 0.6327126622200012,
+      "learning_rate": 8.7035175879397e-05,
+      "loss": 0.3443,
+      "step": 1134
+    },
+    {
+      "epoch": 0.5675,
+      "grad_norm": 0.6015544533729553,
+      "learning_rate": 8.693467336683418e-05,
+      "loss": 0.3251,
+      "step": 1135
+    },
+    {
+      "epoch": 0.568,
+      "grad_norm": 0.569023847579956,
+      "learning_rate": 8.683417085427135e-05,
+      "loss": 0.3085,
+      "step": 1136
+    },
+    {
+      "epoch": 0.5685,
+      "grad_norm": 0.6571947336196899,
+      "learning_rate": 8.673366834170855e-05,
+      "loss": 0.4107,
+      "step": 1137
+    },
+    {
+      "epoch": 0.569,
+      "grad_norm": 0.5744337439537048,
+      "learning_rate": 8.663316582914574e-05,
+      "loss": 0.3578,
+      "step": 1138
+    },
+    {
+      "epoch": 0.5695,
+      "grad_norm": 0.7693044543266296,
+      "learning_rate": 8.653266331658291e-05,
+      "loss": 0.3769,
+      "step": 1139
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9207112789154053,
+      "learning_rate": 8.64321608040201e-05,
+      "loss": 0.4633,
+      "step": 1140
+    },
+    {
+      "epoch": 0.5705,
+      "grad_norm": 0.5837090611457825,
+      "learning_rate": 8.63316582914573e-05,
+      "loss": 0.2766,
+      "step": 1141
+    },
+    {
+      "epoch": 0.571,
+      "grad_norm": 0.6629645824432373,
+      "learning_rate": 8.623115577889449e-05,
+      "loss": 0.4416,
+      "step": 1142
+    },
+    {
+      "epoch": 0.5715,
+      "grad_norm": 0.7339362502098083,
+      "learning_rate": 8.613065326633166e-05,
+      "loss": 0.4691,
+      "step": 1143
+    },
+    {
+      "epoch": 0.572,
+      "grad_norm": 0.6327295899391174,
+      "learning_rate": 8.603015075376884e-05,
+      "loss": 0.3366,
+      "step": 1144
+    },
+    {
+      "epoch": 0.5725,
+      "grad_norm": 0.6540222764015198,
+      "learning_rate": 8.592964824120603e-05,
+      "loss": 0.3413,
+      "step": 1145
+    },
+    {
+      "epoch": 0.573,
+      "grad_norm": 0.7128530740737915,
+      "learning_rate": 8.582914572864322e-05,
+      "loss": 0.4074,
+      "step": 1146
+    },
+    {
+      "epoch": 0.5735,
+      "grad_norm": 0.661177396774292,
+      "learning_rate": 8.57286432160804e-05,
+      "loss": 0.3458,
+      "step": 1147
+    },
+    {
+      "epoch": 0.574,
+      "grad_norm": 0.7002956867218018,
+      "learning_rate": 8.562814070351759e-05,
+      "loss": 0.3593,
+      "step": 1148
+    },
+    {
+      "epoch": 0.5745,
+      "grad_norm": 0.6258227229118347,
+      "learning_rate": 8.552763819095478e-05,
+      "loss": 0.2799,
+      "step": 1149
+    },
+    {
+      "epoch": 0.575,
+      "grad_norm": 0.7016198039054871,
+      "learning_rate": 8.542713567839196e-05,
+      "loss": 0.3297,
+      "step": 1150
+    },
+    {
+      "epoch": 0.5755,
+      "grad_norm": 0.592373788356781,
+      "learning_rate": 8.532663316582915e-05,
+      "loss": 0.498,
+      "step": 1151
+    },
+    {
+      "epoch": 0.576,
+      "grad_norm": 0.6065918803215027,
+      "learning_rate": 8.522613065326634e-05,
+      "loss": 0.5713,
+      "step": 1152
+    },
+    {
+      "epoch": 0.5765,
+      "grad_norm": 0.6737627387046814,
+      "learning_rate": 8.512562814070352e-05,
+      "loss": 0.5781,
+      "step": 1153
+    },
+    {
+      "epoch": 0.577,
+      "grad_norm": 0.6311325430870056,
+      "learning_rate": 8.502512562814071e-05,
+      "loss": 0.486,
+      "step": 1154
+    },
+    {
+      "epoch": 0.5775,
+      "grad_norm": 0.5257154107093811,
+      "learning_rate": 8.49246231155779e-05,
+      "loss": 0.4074,
+      "step": 1155
+    },
+    {
+      "epoch": 0.578,
+      "grad_norm": 0.5722939968109131,
+      "learning_rate": 8.482412060301508e-05,
+      "loss": 0.4531,
+      "step": 1156
+    },
+    {
+      "epoch": 0.5785,
+      "grad_norm": 0.5869833827018738,
+      "learning_rate": 8.472361809045226e-05,
+      "loss": 0.4346,
+      "step": 1157
+    },
+    {
+      "epoch": 0.579,
+      "grad_norm": 0.6782426238059998,
+      "learning_rate": 8.462311557788946e-05,
+      "loss": 0.5387,
+      "step": 1158
+    },
+    {
+      "epoch": 0.5795,
+      "grad_norm": 0.649428129196167,
+      "learning_rate": 8.452261306532664e-05,
+      "loss": 0.5437,
+      "step": 1159
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5602666139602661,
+      "learning_rate": 8.442211055276383e-05,
+      "loss": 0.4291,
+      "step": 1160
+    },
+    {
+      "epoch": 0.5805,
+      "grad_norm": 0.5818144083023071,
+      "learning_rate": 8.4321608040201e-05,
+      "loss": 0.364,
+      "step": 1161
+    },
+    {
+      "epoch": 0.581,
+      "grad_norm": 0.636055588722229,
+      "learning_rate": 8.42211055276382e-05,
+      "loss": 0.4542,
+      "step": 1162
+    },
+    {
+      "epoch": 0.5815,
+      "grad_norm": 0.7070976495742798,
+      "learning_rate": 8.412060301507539e-05,
+      "loss": 0.6046,
+      "step": 1163
+    },
+    {
+      "epoch": 0.582,
+      "grad_norm": 0.621055543422699,
+      "learning_rate": 8.402010050251256e-05,
+      "loss": 0.3558,
+      "step": 1164
+    },
+    {
+      "epoch": 0.5825,
+      "grad_norm": 0.5671415328979492,
+      "learning_rate": 8.391959798994975e-05,
+      "loss": 0.3778,
+      "step": 1165
+    },
+    {
+      "epoch": 0.583,
+      "grad_norm": 0.7483795285224915,
+      "learning_rate": 8.381909547738695e-05,
+      "loss": 0.5184,
+      "step": 1166
+    },
+    {
+      "epoch": 0.5835,
+      "grad_norm": 0.6384664177894592,
+      "learning_rate": 8.371859296482413e-05,
+      "loss": 0.528,
+      "step": 1167
+    },
+    {
+      "epoch": 0.584,
+      "grad_norm": 0.6245007514953613,
+      "learning_rate": 8.36180904522613e-05,
+      "loss": 0.4056,
+      "step": 1168
+    },
+    {
+      "epoch": 0.5845,
+      "grad_norm": 0.6813990473747253,
+      "learning_rate": 8.351758793969849e-05,
+      "loss": 0.5091,
+      "step": 1169
+    },
+    {
+      "epoch": 0.585,
+      "grad_norm": 0.5709593296051025,
+      "learning_rate": 8.341708542713568e-05,
+      "loss": 0.4151,
+      "step": 1170
+    },
+    {
+      "epoch": 0.5855,
+      "grad_norm": 0.8480190634727478,
+      "learning_rate": 8.331658291457287e-05,
+      "loss": 0.4725,
+      "step": 1171
+    },
+    {
+      "epoch": 0.586,
+      "grad_norm": 0.6905885338783264,
+      "learning_rate": 8.321608040201005e-05,
+      "loss": 0.4419,
+      "step": 1172
+    },
+    {
+      "epoch": 0.5865,
+      "grad_norm": 0.6872347593307495,
+      "learning_rate": 8.311557788944724e-05,
+      "loss": 0.5255,
+      "step": 1173
+    },
+    {
+      "epoch": 0.587,
+      "grad_norm": 0.636440634727478,
+      "learning_rate": 8.301507537688443e-05,
+      "loss": 0.3374,
+      "step": 1174
+    },
+    {
+      "epoch": 0.5875,
+      "grad_norm": 0.6185802221298218,
+      "learning_rate": 8.291457286432161e-05,
+      "loss": 0.4424,
+      "step": 1175
+    },
+    {
+      "epoch": 0.588,
+      "grad_norm": 0.6180161833763123,
+      "learning_rate": 8.28140703517588e-05,
+      "loss": 0.364,
+      "step": 1176
+    },
+    {
+      "epoch": 0.5885,
+      "grad_norm": 0.6018222570419312,
+      "learning_rate": 8.271356783919599e-05,
+      "loss": 0.3526,
+      "step": 1177
+    },
+    {
+      "epoch": 0.589,
+      "grad_norm": 0.6561729311943054,
+      "learning_rate": 8.261306532663317e-05,
+      "loss": 0.3839,
+      "step": 1178
+    },
+    {
+      "epoch": 0.5895,
+      "grad_norm": 0.646457314491272,
+      "learning_rate": 8.251256281407036e-05,
+      "loss": 0.4416,
+      "step": 1179
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6187090873718262,
+      "learning_rate": 8.241206030150754e-05,
+      "loss": 0.3837,
+      "step": 1180
+    },
+    {
+      "epoch": 0.5905,
+      "grad_norm": 0.7008528709411621,
+      "learning_rate": 8.231155778894473e-05,
+      "loss": 0.4095,
+      "step": 1181
+    },
+    {
+      "epoch": 0.591,
+      "grad_norm": 0.7083922028541565,
+      "learning_rate": 8.22110552763819e-05,
+      "loss": 0.4747,
+      "step": 1182
+    },
+    {
+      "epoch": 0.5915,
+      "grad_norm": 0.543174147605896,
+      "learning_rate": 8.21105527638191e-05,
+      "loss": 0.3108,
+      "step": 1183
+    },
+    {
+      "epoch": 0.592,
+      "grad_norm": 0.5825045108795166,
+      "learning_rate": 8.201005025125629e-05,
+      "loss": 0.3735,
+      "step": 1184
+    },
+    {
+      "epoch": 0.5925,
+      "grad_norm": 0.7645877599716187,
+      "learning_rate": 8.190954773869348e-05,
+      "loss": 0.4322,
+      "step": 1185
+    },
+    {
+      "epoch": 0.593,
+      "grad_norm": 0.656465470790863,
+      "learning_rate": 8.180904522613065e-05,
+      "loss": 0.3755,
+      "step": 1186
+    },
+    {
+      "epoch": 0.5935,
+      "grad_norm": 0.6986780166625977,
+      "learning_rate": 8.170854271356785e-05,
+      "loss": 0.4258,
+      "step": 1187
+    },
+    {
+      "epoch": 0.594,
+      "grad_norm": 0.5855197906494141,
+      "learning_rate": 8.160804020100504e-05,
+      "loss": 0.3124,
+      "step": 1188
+    },
+    {
+      "epoch": 0.5945,
+      "grad_norm": 0.7552366852760315,
+      "learning_rate": 8.150753768844221e-05,
+      "loss": 0.4813,
+      "step": 1189
+    },
+    {
+      "epoch": 0.595,
+      "grad_norm": 0.6567428708076477,
+      "learning_rate": 8.14070351758794e-05,
+      "loss": 0.4362,
+      "step": 1190
+    },
+    {
+      "epoch": 0.5955,
+      "grad_norm": 0.7465741634368896,
+      "learning_rate": 8.130653266331658e-05,
+      "loss": 0.493,
+      "step": 1191
+    },
+    {
+      "epoch": 0.596,
+      "grad_norm": 0.7557632923126221,
+      "learning_rate": 8.120603015075378e-05,
+      "loss": 0.4529,
+      "step": 1192
+    },
+    {
+      "epoch": 0.5965,
+      "grad_norm": 0.6540181636810303,
+      "learning_rate": 8.110552763819096e-05,
+      "loss": 0.351,
+      "step": 1193
+    },
+    {
+      "epoch": 0.597,
+      "grad_norm": 0.732714056968689,
+      "learning_rate": 8.100502512562814e-05,
+      "loss": 0.4848,
+      "step": 1194
+    },
+    {
+      "epoch": 0.5975,
+      "grad_norm": 0.5961761474609375,
+      "learning_rate": 8.090452261306533e-05,
+      "loss": 0.3702,
+      "step": 1195
+    },
+    {
+      "epoch": 0.598,
+      "grad_norm": 0.5561414361000061,
+      "learning_rate": 8.080402010050251e-05,
+      "loss": 0.3228,
+      "step": 1196
+    },
+    {
+      "epoch": 0.5985,
+      "grad_norm": 0.7552709579467773,
+      "learning_rate": 8.07035175879397e-05,
+      "loss": 0.3896,
+      "step": 1197
+    },
+    {
+      "epoch": 0.599,
+      "grad_norm": 0.5895271301269531,
+      "learning_rate": 8.060301507537689e-05,
+      "loss": 0.3164,
+      "step": 1198
+    },
+    {
+      "epoch": 0.5995,
+      "grad_norm": 0.5493632555007935,
+      "learning_rate": 8.050251256281407e-05,
+      "loss": 0.2763,
+      "step": 1199
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6455180644989014,
+      "learning_rate": 8.040201005025126e-05,
+      "loss": 0.2804,
+      "step": 1200
+    },
+    {
+      "epoch": 0.6,
+      "eval_loss": 0.4328170716762543,
+      "eval_runtime": 109.0774,
+      "eval_samples_per_second": 4.584,
+      "eval_steps_per_second": 4.584,
+      "step": 1200
+    },
+    {
+      "epoch": 0.6005,
+      "grad_norm": 0.5865734219551086,
+      "learning_rate": 8.030150753768845e-05,
+      "loss": 0.5191,
+      "step": 1201
+    },
+    {
+      "epoch": 0.601,
+      "grad_norm": 0.6249243021011353,
+      "learning_rate": 8.020100502512563e-05,
+      "loss": 0.5507,
+      "step": 1202
+    },
+    {
+      "epoch": 0.6015,
+      "grad_norm": 0.6485846638679504,
+      "learning_rate": 8.010050251256282e-05,
+      "loss": 0.6103,
+      "step": 1203
+    },
+    {
+      "epoch": 0.602,
+      "grad_norm": 0.7078679800033569,
+      "learning_rate": 8e-05,
+      "loss": 0.544,
+      "step": 1204
+    },
+    {
+      "epoch": 0.6025,
+      "grad_norm": 0.8002246618270874,
+      "learning_rate": 7.989949748743719e-05,
+      "loss": 0.7316,
+      "step": 1205
+    },
+    {
+      "epoch": 0.603,
+      "grad_norm": 0.658182680606842,
+      "learning_rate": 7.979899497487438e-05,
+      "loss": 0.517,
+      "step": 1206
+    },
+    {
+      "epoch": 0.6035,
+      "grad_norm": 0.6149115562438965,
+      "learning_rate": 7.969849246231155e-05,
+      "loss": 0.474,
+      "step": 1207
+    },
+    {
+      "epoch": 0.604,
+      "grad_norm": 0.5622962713241577,
+      "learning_rate": 7.959798994974875e-05,
+      "loss": 0.423,
+      "step": 1208
+    },
+    {
+      "epoch": 0.6045,
+      "grad_norm": 0.7385901808738708,
+      "learning_rate": 7.949748743718594e-05,
+      "loss": 0.5962,
+      "step": 1209
+    },
+    {
+      "epoch": 0.605,
+      "grad_norm": 0.5881628394126892,
+      "learning_rate": 7.939698492462313e-05,
+      "loss": 0.542,
+      "step": 1210
+    },
+    {
+      "epoch": 0.6055,
+      "grad_norm": 0.636054277420044,
+      "learning_rate": 7.92964824120603e-05,
+      "loss": 0.4508,
+      "step": 1211
+    },
+    {
+      "epoch": 0.606,
+      "grad_norm": 0.5950518846511841,
+      "learning_rate": 7.91959798994975e-05,
+      "loss": 0.5311,
+      "step": 1212
+    },
+    {
+      "epoch": 0.6065,
+      "grad_norm": 0.6514744758605957,
+      "learning_rate": 7.909547738693468e-05,
+      "loss": 0.5175,
+      "step": 1213
+    },
+    {
+      "epoch": 0.607,
+      "grad_norm": 0.5444443821907043,
+      "learning_rate": 7.899497487437186e-05,
+      "loss": 0.2971,
+      "step": 1214
+    },
+    {
+      "epoch": 0.6075,
+      "grad_norm": 0.47983551025390625,
+      "learning_rate": 7.889447236180904e-05,
+      "loss": 0.3345,
+      "step": 1215
+    },
+    {
+      "epoch": 0.608,
+      "grad_norm": 0.5972909331321716,
+      "learning_rate": 7.879396984924623e-05,
+      "loss": 0.4744,
+      "step": 1216
+    },
+    {
+      "epoch": 0.6085,
+      "grad_norm": 0.5156551003456116,
+      "learning_rate": 7.869346733668343e-05,
+      "loss": 0.3455,
+      "step": 1217
+    },
+    {
+      "epoch": 0.609,
+      "grad_norm": 0.7201193571090698,
+      "learning_rate": 7.85929648241206e-05,
+      "loss": 0.5067,
+      "step": 1218
+    },
+    {
+      "epoch": 0.6095,
+      "grad_norm": 1.5245405435562134,
+      "learning_rate": 7.849246231155779e-05,
+      "loss": 0.453,
+      "step": 1219
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5629554986953735,
+      "learning_rate": 7.839195979899498e-05,
+      "loss": 0.406,
+      "step": 1220
+    },
+    {
+      "epoch": 0.6105,
+      "grad_norm": 0.5909186601638794,
+      "learning_rate": 7.829145728643216e-05,
+      "loss": 0.4452,
+      "step": 1221
+    },
+    {
+      "epoch": 0.611,
+      "grad_norm": 0.6694748997688293,
+      "learning_rate": 7.819095477386935e-05,
+      "loss": 0.3793,
+      "step": 1222
+    },
+    {
+      "epoch": 0.6115,
+      "grad_norm": 0.8270274996757507,
+      "learning_rate": 7.809045226130654e-05,
+      "loss": 0.6646,
+      "step": 1223
+    },
+    {
+      "epoch": 0.612,
+      "grad_norm": 0.5169193148612976,
+      "learning_rate": 7.798994974874372e-05,
+      "loss": 0.319,
+      "step": 1224
+    },
+    {
+      "epoch": 0.6125,
+      "grad_norm": 0.629927396774292,
+      "learning_rate": 7.788944723618091e-05,
+      "loss": 0.4553,
+      "step": 1225
+    },
+    {
+      "epoch": 0.613,
+      "grad_norm": 0.5944894552230835,
+      "learning_rate": 7.77889447236181e-05,
+      "loss": 0.439,
+      "step": 1226
+    },
+    {
+      "epoch": 0.6135,
+      "grad_norm": 0.7143748998641968,
+      "learning_rate": 7.768844221105528e-05,
+      "loss": 0.4937,
+      "step": 1227
+    },
+    {
+      "epoch": 0.614,
+      "grad_norm": 0.687521755695343,
+      "learning_rate": 7.758793969849247e-05,
+      "loss": 0.4304,
+      "step": 1228
+    },
+    {
+      "epoch": 0.6145,
+      "grad_norm": 0.6217719912528992,
+      "learning_rate": 7.748743718592966e-05,
+      "loss": 0.3289,
+      "step": 1229
+    },
+    {
+      "epoch": 0.615,
+      "grad_norm": 0.7772793173789978,
+      "learning_rate": 7.738693467336684e-05,
+      "loss": 0.4461,
+      "step": 1230
+    },
+    {
+      "epoch": 0.6155,
+      "grad_norm": 0.5530400276184082,
+      "learning_rate": 7.728643216080403e-05,
+      "loss": 0.3078,
+      "step": 1231
+    },
+    {
+      "epoch": 0.616,
+      "grad_norm": 0.591573178768158,
+      "learning_rate": 7.71859296482412e-05,
+      "loss": 0.3848,
+      "step": 1232
+    },
+    {
+      "epoch": 0.6165,
+      "grad_norm": 0.5898609757423401,
+      "learning_rate": 7.70854271356784e-05,
+      "loss": 0.3805,
+      "step": 1233
+    },
+    {
+      "epoch": 0.617,
+      "grad_norm": 0.6441884636878967,
+      "learning_rate": 7.698492462311559e-05,
+      "loss": 0.3713,
+      "step": 1234
+    },
+    {
+      "epoch": 0.6175,
+      "grad_norm": 0.7090359330177307,
+      "learning_rate": 7.688442211055277e-05,
+      "loss": 0.4603,
+      "step": 1235
+    },
+    {
+      "epoch": 0.618,
+      "grad_norm": 0.679050087928772,
+      "learning_rate": 7.678391959798995e-05,
+      "loss": 0.4351,
+      "step": 1236
+    },
+    {
+      "epoch": 0.6185,
+      "grad_norm": 0.5958268642425537,
+      "learning_rate": 7.668341708542713e-05,
+      "loss": 0.3772,
+      "step": 1237
+    },
+    {
+      "epoch": 0.619,
+      "grad_norm": 0.5951745510101318,
+      "learning_rate": 7.658291457286433e-05,
+      "loss": 0.317,
+      "step": 1238
+    },
+    {
+      "epoch": 0.6195,
+      "grad_norm": 0.6101677417755127,
+      "learning_rate": 7.64824120603015e-05,
+      "loss": 0.379,
+      "step": 1239
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.659229040145874,
+      "learning_rate": 7.638190954773869e-05,
+      "loss": 0.4124,
+      "step": 1240
+    },
+    {
+      "epoch": 0.6205,
+      "grad_norm": 0.734038233757019,
+      "learning_rate": 7.628140703517588e-05,
+      "loss": 0.4467,
+      "step": 1241
+    },
+    {
+      "epoch": 0.621,
+      "grad_norm": 0.6723963022232056,
+      "learning_rate": 7.618090452261307e-05,
+      "loss": 0.3766,
+      "step": 1242
+    },
+    {
+      "epoch": 0.6215,
+      "grad_norm": 0.689439594745636,
+      "learning_rate": 7.608040201005025e-05,
+      "loss": 0.35,
+      "step": 1243
+    },
+    {
+      "epoch": 0.622,
+      "grad_norm": 0.7784453630447388,
+      "learning_rate": 7.597989949748744e-05,
+      "loss": 0.5154,
+      "step": 1244
+    },
+    {
+      "epoch": 0.6225,
+      "grad_norm": 0.5916337966918945,
+      "learning_rate": 7.587939698492463e-05,
+      "loss": 0.3443,
+      "step": 1245
+    },
+    {
+      "epoch": 0.623,
+      "grad_norm": 0.5388657450675964,
+      "learning_rate": 7.577889447236181e-05,
+      "loss": 0.2675,
+      "step": 1246
+    },
+    {
+      "epoch": 0.6235,
+      "grad_norm": 0.5964484214782715,
+      "learning_rate": 7.5678391959799e-05,
+      "loss": 0.2826,
+      "step": 1247
+    },
+    {
+      "epoch": 0.624,
+      "grad_norm": 0.5806629657745361,
+      "learning_rate": 7.557788944723618e-05,
+      "loss": 0.2511,
+      "step": 1248
+    },
+    {
+      "epoch": 0.6245,
+      "grad_norm": 0.7091418504714966,
+      "learning_rate": 7.547738693467337e-05,
+      "loss": 0.3281,
+      "step": 1249
+    },
+    {
+      "epoch": 0.625,
+      "grad_norm": 0.6654065847396851,
+      "learning_rate": 7.537688442211056e-05,
+      "loss": 0.305,
+      "step": 1250
+    },
+    {
+      "epoch": 0.6255,
+      "grad_norm": 0.5841928720474243,
+      "learning_rate": 7.527638190954774e-05,
+      "loss": 0.6432,
+      "step": 1251
+    },
+    {
+      "epoch": 0.626,
+      "grad_norm": 0.5990925431251526,
+      "learning_rate": 7.517587939698493e-05,
+      "loss": 0.5952,
+      "step": 1252
+    },
+    {
+      "epoch": 0.6265,
+      "grad_norm": 0.6245496869087219,
+      "learning_rate": 7.507537688442212e-05,
+      "loss": 0.6091,
+      "step": 1253
+    },
+    {
+      "epoch": 0.627,
+      "grad_norm": 0.5987966060638428,
+      "learning_rate": 7.49748743718593e-05,
+      "loss": 0.4897,
+      "step": 1254
+    },
+    {
+      "epoch": 0.6275,
+      "grad_norm": 0.6903722882270813,
+      "learning_rate": 7.487437185929649e-05,
+      "loss": 0.5878,
+      "step": 1255
+    },
+    {
+      "epoch": 0.628,
+      "grad_norm": 0.6584540605545044,
+      "learning_rate": 7.477386934673368e-05,
+      "loss": 0.4908,
+      "step": 1256
+    },
+    {
+      "epoch": 0.6285,
+      "grad_norm": 0.5866811871528625,
+      "learning_rate": 7.467336683417085e-05,
+      "loss": 0.4501,
+      "step": 1257
+    },
+    {
+      "epoch": 0.629,
+      "grad_norm": 0.6148390769958496,
+      "learning_rate": 7.457286432160805e-05,
+      "loss": 0.5051,
+      "step": 1258
+    },
+    {
+      "epoch": 0.6295,
+      "grad_norm": 0.6225212812423706,
+      "learning_rate": 7.447236180904524e-05,
+      "loss": 0.4878,
+      "step": 1259
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7189927101135254,
+      "learning_rate": 7.437185929648241e-05,
+      "loss": 0.5058,
+      "step": 1260
+    },
+    {
+      "epoch": 0.6305,
+      "grad_norm": 0.6307305097579956,
+      "learning_rate": 7.42713567839196e-05,
+      "loss": 0.4234,
+      "step": 1261
+    },
+    {
+      "epoch": 0.631,
+      "grad_norm": 0.5336734652519226,
+      "learning_rate": 7.417085427135678e-05,
+      "loss": 0.2822,
+      "step": 1262
+    },
+    {
+      "epoch": 0.6315,
+      "grad_norm": 0.6724408864974976,
+      "learning_rate": 7.407035175879398e-05,
+      "loss": 0.5804,
+      "step": 1263
+    },
+    {
+      "epoch": 0.632,
+      "grad_norm": 0.6323987245559692,
+      "learning_rate": 7.396984924623115e-05,
+      "loss": 0.4274,
+      "step": 1264
+    },
+    {
+      "epoch": 0.6325,
+      "grad_norm": 0.65557861328125,
+      "learning_rate": 7.386934673366834e-05,
+      "loss": 0.4601,
+      "step": 1265
+    },
+    {
+      "epoch": 0.633,
+      "grad_norm": 0.5712767243385315,
+      "learning_rate": 7.376884422110553e-05,
+      "loss": 0.3198,
+      "step": 1266
+    },
+    {
+      "epoch": 0.6335,
+      "grad_norm": 0.5992964506149292,
+      "learning_rate": 7.366834170854271e-05,
+      "loss": 0.3294,
+      "step": 1267
+    },
+    {
+      "epoch": 0.634,
+      "grad_norm": 0.6699780821800232,
+      "learning_rate": 7.35678391959799e-05,
+      "loss": 0.523,
+      "step": 1268
+    },
+    {
+      "epoch": 0.6345,
+      "grad_norm": 0.6404006481170654,
+      "learning_rate": 7.346733668341709e-05,
+      "loss": 0.4108,
+      "step": 1269
+    },
+    {
+      "epoch": 0.635,
+      "grad_norm": 0.6395924687385559,
+      "learning_rate": 7.336683417085427e-05,
+      "loss": 0.3843,
+      "step": 1270
+    },
+    {
+      "epoch": 0.6355,
+      "grad_norm": 0.5462537407875061,
+      "learning_rate": 7.326633165829146e-05,
+      "loss": 0.3098,
+      "step": 1271
+    },
+    {
+      "epoch": 0.636,
+      "grad_norm": 0.7069851756095886,
+      "learning_rate": 7.316582914572865e-05,
+      "loss": 0.4733,
+      "step": 1272
+    },
+    {
+      "epoch": 0.6365,
+      "grad_norm": 0.6006671786308289,
+      "learning_rate": 7.306532663316583e-05,
+      "loss": 0.387,
+      "step": 1273
+    },
+    {
+      "epoch": 0.637,
+      "grad_norm": 0.6217454671859741,
+      "learning_rate": 7.296482412060302e-05,
+      "loss": 0.4223,
+      "step": 1274
+    },
+    {
+      "epoch": 0.6375,
+      "grad_norm": 0.8078276515007019,
+      "learning_rate": 7.28643216080402e-05,
+      "loss": 0.5341,
+      "step": 1275
+    },
+    {
+      "epoch": 0.638,
+      "grad_norm": 0.6509358286857605,
+      "learning_rate": 7.276381909547739e-05,
+      "loss": 0.4029,
+      "step": 1276
+    },
+    {
+      "epoch": 0.6385,
+      "grad_norm": 0.7253496646881104,
+      "learning_rate": 7.266331658291458e-05,
+      "loss": 0.4976,
+      "step": 1277
+    },
+    {
+      "epoch": 0.639,
+      "grad_norm": 0.7190517783164978,
+      "learning_rate": 7.256281407035177e-05,
+      "loss": 0.4793,
+      "step": 1278
+    },
+    {
+      "epoch": 0.6395,
+      "grad_norm": 0.689811110496521,
+      "learning_rate": 7.246231155778895e-05,
+      "loss": 0.414,
+      "step": 1279
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6280933022499084,
+      "learning_rate": 7.236180904522614e-05,
+      "loss": 0.4134,
+      "step": 1280
+    },
+    {
+      "epoch": 0.6405,
+      "grad_norm": 0.7857996225357056,
+      "learning_rate": 7.226130653266333e-05,
+      "loss": 0.5303,
+      "step": 1281
+    },
+    {
+      "epoch": 0.641,
+      "grad_norm": 0.7987685203552246,
+      "learning_rate": 7.21608040201005e-05,
+      "loss": 0.5393,
+      "step": 1282
+    },
+    {
+      "epoch": 0.6415,
+      "grad_norm": 0.62496417760849,
+      "learning_rate": 7.206030150753768e-05,
+      "loss": 0.3449,
+      "step": 1283
+    },
+    {
+      "epoch": 0.642,
+      "grad_norm": 0.6730352640151978,
+      "learning_rate": 7.195979899497488e-05,
+      "loss": 0.4286,
+      "step": 1284
+    },
+    {
+      "epoch": 0.6425,
+      "grad_norm": 0.653498649597168,
+      "learning_rate": 7.185929648241206e-05,
+      "loss": 0.4226,
+      "step": 1285
+    },
+    {
+      "epoch": 0.643,
+      "grad_norm": 0.6921342015266418,
+      "learning_rate": 7.175879396984924e-05,
+      "loss": 0.51,
+      "step": 1286
+    },
+    {
+      "epoch": 0.6435,
+      "grad_norm": 0.798220694065094,
+      "learning_rate": 7.165829145728643e-05,
+      "loss": 0.459,
+      "step": 1287
+    },
+    {
+      "epoch": 0.644,
+      "grad_norm": 0.735780656337738,
+      "learning_rate": 7.155778894472363e-05,
+      "loss": 0.4373,
+      "step": 1288
+    },
+    {
+      "epoch": 0.6445,
+      "grad_norm": 0.7888179421424866,
+      "learning_rate": 7.14572864321608e-05,
+      "loss": 0.4069,
+      "step": 1289
+    },
+    {
+      "epoch": 0.645,
+      "grad_norm": 0.6111482381820679,
+      "learning_rate": 7.135678391959799e-05,
+      "loss": 0.396,
+      "step": 1290
+    },
+    {
+      "epoch": 0.6455,
+      "grad_norm": 0.6213611960411072,
+      "learning_rate": 7.125628140703518e-05,
+      "loss": 0.4239,
+      "step": 1291
+    },
+    {
+      "epoch": 0.646,
+      "grad_norm": 0.7928662300109863,
+      "learning_rate": 7.115577889447236e-05,
+      "loss": 0.4214,
+      "step": 1292
+    },
+    {
+      "epoch": 0.6465,
+      "grad_norm": 0.9043471813201904,
+      "learning_rate": 7.105527638190955e-05,
+      "loss": 0.4061,
+      "step": 1293
+    },
+    {
+      "epoch": 0.647,
+      "grad_norm": 0.8142223954200745,
+      "learning_rate": 7.095477386934674e-05,
+      "loss": 0.4773,
+      "step": 1294
+    },
+    {
+      "epoch": 0.6475,
+      "grad_norm": 0.8139143586158752,
+      "learning_rate": 7.085427135678392e-05,
+      "loss": 0.4772,
+      "step": 1295
+    },
+    {
+      "epoch": 0.648,
+      "grad_norm": 0.6583029627799988,
+      "learning_rate": 7.075376884422111e-05,
+      "loss": 0.4054,
+      "step": 1296
+    },
+    {
+      "epoch": 0.6485,
+      "grad_norm": 0.8101404905319214,
+      "learning_rate": 7.06532663316583e-05,
+      "loss": 0.4239,
+      "step": 1297
+    },
+    {
+      "epoch": 0.649,
+      "grad_norm": 0.6830359101295471,
+      "learning_rate": 7.055276381909548e-05,
+      "loss": 0.2966,
+      "step": 1298
+    },
+    {
+      "epoch": 0.6495,
+      "grad_norm": 0.573408305644989,
+      "learning_rate": 7.045226130653267e-05,
+      "loss": 0.3043,
+      "step": 1299
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5185285210609436,
+      "learning_rate": 7.035175879396985e-05,
+      "loss": 0.2372,
+      "step": 1300
+    },
+    {
+      "epoch": 0.6505,
+      "grad_norm": 0.544872522354126,
+      "learning_rate": 7.025125628140704e-05,
+      "loss": 0.4139,
+      "step": 1301
+    },
+    {
+      "epoch": 0.651,
+      "grad_norm": 0.4939993917942047,
+      "learning_rate": 7.015075376884423e-05,
+      "loss": 0.3128,
+      "step": 1302
+    },
+    {
+      "epoch": 0.6515,
+      "grad_norm": 0.6160438656806946,
+      "learning_rate": 7.00502512562814e-05,
+      "loss": 0.5182,
+      "step": 1303
+    },
+    {
+      "epoch": 0.652,
+      "grad_norm": 0.7090399265289307,
+      "learning_rate": 6.99497487437186e-05,
+      "loss": 0.5506,
+      "step": 1304
+    },
+    {
+      "epoch": 0.6525,
+      "grad_norm": 0.6979255080223083,
+      "learning_rate": 6.984924623115579e-05,
+      "loss": 0.5918,
+      "step": 1305
+    },
+    {
+      "epoch": 0.653,
+      "grad_norm": 0.5997445583343506,
+      "learning_rate": 6.974874371859297e-05,
+      "loss": 0.3252,
+      "step": 1306
+    },
+    {
+      "epoch": 0.6535,
+      "grad_norm": 0.700102686882019,
+      "learning_rate": 6.964824120603015e-05,
+      "loss": 0.5162,
+      "step": 1307
+    },
+    {
+      "epoch": 0.654,
+      "grad_norm": 0.5877028107643127,
+      "learning_rate": 6.954773869346733e-05,
+      "loss": 0.422,
+      "step": 1308
+    },
+    {
+      "epoch": 0.6545,
+      "grad_norm": 0.6030882000923157,
+      "learning_rate": 6.944723618090453e-05,
+      "loss": 0.3911,
+      "step": 1309
+    },
+    {
+      "epoch": 0.655,
+      "grad_norm": 0.6222591400146484,
+      "learning_rate": 6.93467336683417e-05,
+      "loss": 0.3729,
+      "step": 1310
+    },
+    {
+      "epoch": 0.6555,
+      "grad_norm": 0.8279755711555481,
+      "learning_rate": 6.924623115577889e-05,
+      "loss": 0.6466,
+      "step": 1311
+    },
+    {
+      "epoch": 0.656,
+      "grad_norm": 0.6300482153892517,
+      "learning_rate": 6.914572864321608e-05,
+      "loss": 0.4615,
+      "step": 1312
+    },
+    {
+      "epoch": 0.6565,
+      "grad_norm": 0.5657908916473389,
+      "learning_rate": 6.904522613065328e-05,
+      "loss": 0.3288,
+      "step": 1313
+    },
+    {
+      "epoch": 0.657,
+      "grad_norm": 0.6319009065628052,
+      "learning_rate": 6.894472361809045e-05,
+      "loss": 0.4014,
+      "step": 1314
+    },
+    {
+      "epoch": 0.6575,
+      "grad_norm": 0.6748180985450745,
+      "learning_rate": 6.884422110552764e-05,
+      "loss": 0.4489,
+      "step": 1315
+    },
+    {
+      "epoch": 0.658,
+      "grad_norm": 0.7351542115211487,
+      "learning_rate": 6.874371859296482e-05,
+      "loss": 0.413,
+      "step": 1316
+    },
+    {
+      "epoch": 0.6585,
+      "grad_norm": 0.6244130730628967,
+      "learning_rate": 6.864321608040201e-05,
+      "loss": 0.3728,
+      "step": 1317
+    },
+    {
+      "epoch": 0.659,
+      "grad_norm": 0.5785526037216187,
+      "learning_rate": 6.85427135678392e-05,
+      "loss": 0.3855,
+      "step": 1318
+    },
+    {
+      "epoch": 0.6595,
+      "grad_norm": 0.7869665622711182,
+      "learning_rate": 6.844221105527638e-05,
+      "loss": 0.5584,
+      "step": 1319
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6740867495536804,
+      "learning_rate": 6.834170854271357e-05,
+      "loss": 0.4518,
+      "step": 1320
+    },
+    {
+      "epoch": 0.6605,
+      "grad_norm": 0.5851686000823975,
+      "learning_rate": 6.824120603015076e-05,
+      "loss": 0.3499,
+      "step": 1321
+    },
+    {
+      "epoch": 0.661,
+      "grad_norm": 0.6631923317909241,
+      "learning_rate": 6.814070351758794e-05,
+      "loss": 0.3647,
+      "step": 1322
+    },
+    {
+      "epoch": 0.6615,
+      "grad_norm": 0.776313066482544,
+      "learning_rate": 6.804020100502513e-05,
+      "loss": 0.5563,
+      "step": 1323
+    },
+    {
+      "epoch": 0.662,
+      "grad_norm": 0.7072078585624695,
+      "learning_rate": 6.793969849246232e-05,
+      "loss": 0.4466,
+      "step": 1324
+    },
+    {
+      "epoch": 0.6625,
+      "grad_norm": 0.6975109577178955,
+      "learning_rate": 6.78391959798995e-05,
+      "loss": 0.4338,
+      "step": 1325
+    },
+    {
+      "epoch": 0.663,
+      "grad_norm": 0.624419093132019,
+      "learning_rate": 6.773869346733669e-05,
+      "loss": 0.3736,
+      "step": 1326
+    },
+    {
+      "epoch": 0.6635,
+      "grad_norm": 0.6096883416175842,
+      "learning_rate": 6.763819095477388e-05,
+      "loss": 0.3949,
+      "step": 1327
+    },
+    {
+      "epoch": 0.664,
+      "grad_norm": 0.564062237739563,
+      "learning_rate": 6.753768844221105e-05,
+      "loss": 0.2985,
+      "step": 1328
+    },
+    {
+      "epoch": 0.6645,
+      "grad_norm": 0.657784640789032,
+      "learning_rate": 6.743718592964824e-05,
+      "loss": 0.4453,
+      "step": 1329
+    },
+    {
+      "epoch": 0.665,
+      "grad_norm": 0.7055421471595764,
+      "learning_rate": 6.733668341708544e-05,
+      "loss": 0.4926,
+      "step": 1330
+    },
+    {
+      "epoch": 0.6655,
+      "grad_norm": 0.7406333684921265,
+      "learning_rate": 6.723618090452262e-05,
+      "loss": 0.4783,
+      "step": 1331
+    },
+    {
+      "epoch": 0.666,
+      "grad_norm": 0.5958945751190186,
+      "learning_rate": 6.71356783919598e-05,
+      "loss": 0.3955,
+      "step": 1332
+    },
+    {
+      "epoch": 0.6665,
+      "grad_norm": 0.6093408465385437,
+      "learning_rate": 6.703517587939698e-05,
+      "loss": 0.384,
+      "step": 1333
+    },
+    {
+      "epoch": 0.667,
+      "grad_norm": 0.8250641226768494,
+      "learning_rate": 6.693467336683418e-05,
+      "loss": 0.6,
+      "step": 1334
+    },
+    {
+      "epoch": 0.6675,
+      "grad_norm": 0.7149648666381836,
+      "learning_rate": 6.683417085427135e-05,
+      "loss": 0.4491,
+      "step": 1335
+    },
+    {
+      "epoch": 0.668,
+      "grad_norm": 0.7480332255363464,
+      "learning_rate": 6.673366834170854e-05,
+      "loss": 0.4812,
+      "step": 1336
+    },
+    {
+      "epoch": 0.6685,
+      "grad_norm": 0.6271966099739075,
+      "learning_rate": 6.663316582914573e-05,
+      "loss": 0.4455,
+      "step": 1337
+    },
+    {
+      "epoch": 0.669,
+      "grad_norm": 0.6300346851348877,
+      "learning_rate": 6.653266331658293e-05,
+      "loss": 0.3629,
+      "step": 1338
+    },
+    {
+      "epoch": 0.6695,
+      "grad_norm": 0.57159423828125,
+      "learning_rate": 6.64321608040201e-05,
+      "loss": 0.3039,
+      "step": 1339
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7275968790054321,
+      "learning_rate": 6.633165829145729e-05,
+      "loss": 0.363,
+      "step": 1340
+    },
+    {
+      "epoch": 0.6705,
+      "grad_norm": 0.6649594902992249,
+      "learning_rate": 6.623115577889447e-05,
+      "loss": 0.4179,
+      "step": 1341
+    },
+    {
+      "epoch": 0.671,
+      "grad_norm": 0.6205945014953613,
+      "learning_rate": 6.613065326633166e-05,
+      "loss": 0.4119,
+      "step": 1342
+    },
+    {
+      "epoch": 0.6715,
+      "grad_norm": 0.7332936525344849,
+      "learning_rate": 6.603015075376885e-05,
+      "loss": 0.3343,
+      "step": 1343
+    },
+    {
+      "epoch": 0.672,
+      "grad_norm": 0.6883128881454468,
+      "learning_rate": 6.592964824120603e-05,
+      "loss": 0.4038,
+      "step": 1344
+    },
+    {
+      "epoch": 0.6725,
+      "grad_norm": 0.5815170407295227,
+      "learning_rate": 6.582914572864322e-05,
+      "loss": 0.2877,
+      "step": 1345
+    },
+    {
+      "epoch": 0.673,
+      "grad_norm": 0.6479829549789429,
+      "learning_rate": 6.57286432160804e-05,
+      "loss": 0.341,
+      "step": 1346
+    },
+    {
+      "epoch": 0.6735,
+      "grad_norm": 0.6782565116882324,
+      "learning_rate": 6.562814070351759e-05,
+      "loss": 0.3147,
+      "step": 1347
+    },
+    {
+      "epoch": 0.674,
+      "grad_norm": 0.6235268115997314,
+      "learning_rate": 6.552763819095478e-05,
+      "loss": 0.2829,
+      "step": 1348
+    },
+    {
+      "epoch": 0.6745,
+      "grad_norm": 0.6052720546722412,
+      "learning_rate": 6.542713567839197e-05,
+      "loss": 0.2743,
+      "step": 1349
+    },
+    {
+      "epoch": 0.675,
+      "grad_norm": 0.7172004580497742,
+      "learning_rate": 6.532663316582915e-05,
+      "loss": 0.3076,
+      "step": 1350
+    },
+    {
+      "epoch": 0.6755,
+      "grad_norm": 0.5431641936302185,
+      "learning_rate": 6.522613065326634e-05,
+      "loss": 0.473,
+      "step": 1351
+    },
+    {
+      "epoch": 0.676,
+      "grad_norm": 0.5774933099746704,
+      "learning_rate": 6.512562814070352e-05,
+      "loss": 0.4656,
+      "step": 1352
+    },
+    {
+      "epoch": 0.6765,
+      "grad_norm": 0.6277914643287659,
+      "learning_rate": 6.50251256281407e-05,
+      "loss": 0.5575,
+      "step": 1353
+    },
+    {
+      "epoch": 0.677,
+      "grad_norm": 0.6048242449760437,
+      "learning_rate": 6.492462311557788e-05,
+      "loss": 0.4797,
+      "step": 1354
+    },
+    {
+      "epoch": 0.6775,
+      "grad_norm": 0.6998909115791321,
+      "learning_rate": 6.482412060301508e-05,
+      "loss": 0.6131,
+      "step": 1355
+    },
+    {
+      "epoch": 0.678,
+      "grad_norm": 0.7205404043197632,
+      "learning_rate": 6.472361809045227e-05,
+      "loss": 0.6343,
+      "step": 1356
+    },
+    {
+      "epoch": 0.6785,
+      "grad_norm": 0.5799533128738403,
+      "learning_rate": 6.462311557788944e-05,
+      "loss": 0.4388,
+      "step": 1357
+    },
+    {
+      "epoch": 0.679,
+      "grad_norm": 0.6117823719978333,
+      "learning_rate": 6.452261306532663e-05,
+      "loss": 0.5037,
+      "step": 1358
+    },
+    {
+      "epoch": 0.6795,
+      "grad_norm": 0.49690335988998413,
+      "learning_rate": 6.442211055276383e-05,
+      "loss": 0.3548,
+      "step": 1359
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5126916766166687,
+      "learning_rate": 6.4321608040201e-05,
+      "loss": 0.38,
+      "step": 1360
+    },
+    {
+      "epoch": 0.6805,
+      "grad_norm": 0.6659131050109863,
+      "learning_rate": 6.422110552763819e-05,
+      "loss": 0.4413,
+      "step": 1361
+    },
+    {
+      "epoch": 0.681,
+      "grad_norm": 0.5937123894691467,
+      "learning_rate": 6.412060301507538e-05,
+      "loss": 0.4497,
+      "step": 1362
+    },
+    {
+      "epoch": 0.6815,
+      "grad_norm": 0.6383756995201111,
+      "learning_rate": 6.402010050251258e-05,
+      "loss": 0.5112,
+      "step": 1363
+    },
+    {
+      "epoch": 0.682,
+      "grad_norm": 0.6325298547744751,
+      "learning_rate": 6.391959798994975e-05,
+      "loss": 0.4359,
+      "step": 1364
+    },
+    {
+      "epoch": 0.6825,
+      "grad_norm": 0.8342583179473877,
+      "learning_rate": 6.381909547738694e-05,
+      "loss": 0.5505,
+      "step": 1365
+    },
+    {
+      "epoch": 0.683,
+      "grad_norm": 0.8640339970588684,
+      "learning_rate": 6.371859296482412e-05,
+      "loss": 0.5571,
+      "step": 1366
+    },
+    {
+      "epoch": 0.6835,
+      "grad_norm": 0.6175604462623596,
+      "learning_rate": 6.361809045226131e-05,
+      "loss": 0.3627,
+      "step": 1367
+    },
+    {
+      "epoch": 0.684,
+      "grad_norm": 0.6928995251655579,
+      "learning_rate": 6.35175879396985e-05,
+      "loss": 0.4831,
+      "step": 1368
+    },
+    {
+      "epoch": 0.6845,
+      "grad_norm": 0.8149310946464539,
+      "learning_rate": 6.341708542713568e-05,
+      "loss": 0.5165,
+      "step": 1369
+    },
+    {
+      "epoch": 0.685,
+      "grad_norm": 0.6305818557739258,
+      "learning_rate": 6.331658291457287e-05,
+      "loss": 0.4253,
+      "step": 1370
+    },
+    {
+      "epoch": 0.6855,
+      "grad_norm": 0.6397230625152588,
+      "learning_rate": 6.321608040201005e-05,
+      "loss": 0.4466,
+      "step": 1371
+    },
+    {
+      "epoch": 0.686,
+      "grad_norm": 0.6917291283607483,
+      "learning_rate": 6.311557788944724e-05,
+      "loss": 0.3722,
+      "step": 1372
+    },
+    {
+      "epoch": 0.6865,
+      "grad_norm": 0.7245936393737793,
+      "learning_rate": 6.301507537688443e-05,
+      "loss": 0.4009,
+      "step": 1373
+    },
+    {
+      "epoch": 0.687,
+      "grad_norm": 0.8297839760780334,
+      "learning_rate": 6.291457286432161e-05,
+      "loss": 0.6313,
+      "step": 1374
+    },
+    {
+      "epoch": 0.6875,
+      "grad_norm": 0.6506665349006653,
+      "learning_rate": 6.28140703517588e-05,
+      "loss": 0.4067,
+      "step": 1375
+    },
+    {
+      "epoch": 0.688,
+      "grad_norm": 0.6748162508010864,
+      "learning_rate": 6.271356783919599e-05,
+      "loss": 0.4562,
+      "step": 1376
+    },
+    {
+      "epoch": 0.6885,
+      "grad_norm": 0.6416603922843933,
+      "learning_rate": 6.261306532663317e-05,
+      "loss": 0.4194,
+      "step": 1377
+    },
+    {
+      "epoch": 0.689,
+      "grad_norm": 0.7227828502655029,
+      "learning_rate": 6.251256281407035e-05,
+      "loss": 0.5512,
+      "step": 1378
+    },
+    {
+      "epoch": 0.6895,
+      "grad_norm": 0.6089235544204712,
+      "learning_rate": 6.241206030150753e-05,
+      "loss": 0.3512,
+      "step": 1379
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7788892388343811,
+      "learning_rate": 6.231155778894473e-05,
+      "loss": 0.5234,
+      "step": 1380
+    },
+    {
+      "epoch": 0.6905,
+      "grad_norm": 0.6106981635093689,
+      "learning_rate": 6.221105527638192e-05,
+      "loss": 0.3677,
+      "step": 1381
+    },
+    {
+      "epoch": 0.691,
+      "grad_norm": 0.6866602897644043,
+      "learning_rate": 6.211055276381909e-05,
+      "loss": 0.4085,
+      "step": 1382
+    },
+    {
+      "epoch": 0.6915,
+      "grad_norm": 0.6046152114868164,
+      "learning_rate": 6.201005025125628e-05,
+      "loss": 0.3521,
+      "step": 1383
+    },
+    {
+      "epoch": 0.692,
+      "grad_norm": 0.5847232937812805,
+      "learning_rate": 6.190954773869348e-05,
+      "loss": 0.3632,
+      "step": 1384
+    },
+    {
+      "epoch": 0.6925,
+      "grad_norm": 0.5511264801025391,
+      "learning_rate": 6.180904522613065e-05,
+      "loss": 0.3364,
+      "step": 1385
+    },
+    {
+      "epoch": 0.693,
+      "grad_norm": 0.6893430948257446,
+      "learning_rate": 6.170854271356784e-05,
+      "loss": 0.4139,
+      "step": 1386
+    },
+    {
+      "epoch": 0.6935,
+      "grad_norm": 0.7392801642417908,
+      "learning_rate": 6.160804020100502e-05,
+      "loss": 0.4353,
+      "step": 1387
+    },
+    {
+      "epoch": 0.694,
+      "grad_norm": 0.7249146103858948,
+      "learning_rate": 6.150753768844222e-05,
+      "loss": 0.3777,
+      "step": 1388
+    },
+    {
+      "epoch": 0.6945,
+      "grad_norm": 0.632563591003418,
+      "learning_rate": 6.14070351758794e-05,
+      "loss": 0.3498,
+      "step": 1389
+    },
+    {
+      "epoch": 0.695,
+      "grad_norm": 0.7016903162002563,
+      "learning_rate": 6.130653266331658e-05,
+      "loss": 0.4229,
+      "step": 1390
+    },
+    {
+      "epoch": 0.6955,
+      "grad_norm": 0.6217426657676697,
+      "learning_rate": 6.120603015075377e-05,
+      "loss": 0.3733,
+      "step": 1391
+    },
+    {
+      "epoch": 0.696,
+      "grad_norm": 0.66749507188797,
+      "learning_rate": 6.110552763819096e-05,
+      "loss": 0.4123,
+      "step": 1392
+    },
+    {
+      "epoch": 0.6965,
+      "grad_norm": 0.6326702833175659,
+      "learning_rate": 6.1005025125628143e-05,
+      "loss": 0.3232,
+      "step": 1393
+    },
+    {
+      "epoch": 0.697,
+      "grad_norm": 0.7051452994346619,
+      "learning_rate": 6.090452261306533e-05,
+      "loss": 0.3888,
+      "step": 1394
+    },
+    {
+      "epoch": 0.6975,
+      "grad_norm": 0.6341792941093445,
+      "learning_rate": 6.080402010050251e-05,
+      "loss": 0.368,
+      "step": 1395
+    },
+    {
+      "epoch": 0.698,
+      "grad_norm": 0.5280935764312744,
+      "learning_rate": 6.070351758793971e-05,
+      "loss": 0.2593,
+      "step": 1396
+    },
+    {
+      "epoch": 0.6985,
+      "grad_norm": 0.7069610357284546,
+      "learning_rate": 6.060301507537689e-05,
+      "loss": 0.3837,
+      "step": 1397
+    },
+    {
+      "epoch": 0.699,
+      "grad_norm": 0.697541356086731,
+      "learning_rate": 6.0502512562814076e-05,
+      "loss": 0.34,
+      "step": 1398
+    },
+    {
+      "epoch": 0.6995,
+      "grad_norm": 0.7502248883247375,
+      "learning_rate": 6.0402010050251256e-05,
+      "loss": 0.3738,
+      "step": 1399
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.661486804485321,
+      "learning_rate": 6.030150753768844e-05,
+      "loss": 0.2664,
+      "step": 1400
+    },
+    {
+      "epoch": 0.7005,
+      "grad_norm": 0.5114116072654724,
+      "learning_rate": 6.0201005025125635e-05,
+      "loss": 0.4844,
+      "step": 1401
+    },
+    {
+      "epoch": 0.701,
+      "grad_norm": 0.515103280544281,
+      "learning_rate": 6.0100502512562815e-05,
+      "loss": 0.4186,
+      "step": 1402
+    },
+    {
+      "epoch": 0.7015,
+      "grad_norm": 0.538327693939209,
+      "learning_rate": 6e-05,
+      "loss": 0.3682,
+      "step": 1403
+    },
+    {
+      "epoch": 0.702,
+      "grad_norm": 0.7211799025535583,
+      "learning_rate": 5.989949748743718e-05,
+      "loss": 0.5736,
+      "step": 1404
+    },
+    {
+      "epoch": 0.7025,
+      "grad_norm": 0.6347200274467468,
+      "learning_rate": 5.979899497487438e-05,
+      "loss": 0.5333,
+      "step": 1405
+    },
+    {
+      "epoch": 0.703,
+      "grad_norm": 0.6995561718940735,
+      "learning_rate": 5.969849246231156e-05,
+      "loss": 0.4808,
+      "step": 1406
+    },
+    {
+      "epoch": 0.7035,
+      "grad_norm": 0.6232349276542664,
+      "learning_rate": 5.959798994974875e-05,
+      "loss": 0.4996,
+      "step": 1407
+    },
+    {
+      "epoch": 0.704,
+      "grad_norm": 0.6309375166893005,
+      "learning_rate": 5.949748743718593e-05,
+      "loss": 0.566,
+      "step": 1408
+    },
+    {
+      "epoch": 0.7045,
+      "grad_norm": 0.5563838481903076,
+      "learning_rate": 5.939698492462312e-05,
+      "loss": 0.3862,
+      "step": 1409
+    },
+    {
+      "epoch": 0.705,
+      "grad_norm": 0.6890807151794434,
+      "learning_rate": 5.929648241206031e-05,
+      "loss": 0.4698,
+      "step": 1410
+    },
+    {
+      "epoch": 0.7055,
+      "grad_norm": 0.5582006573677063,
+      "learning_rate": 5.9195979899497487e-05,
+      "loss": 0.3818,
+      "step": 1411
+    },
+    {
+      "epoch": 0.706,
+      "grad_norm": 0.6586242318153381,
+      "learning_rate": 5.909547738693467e-05,
+      "loss": 0.4409,
+      "step": 1412
+    },
+    {
+      "epoch": 0.7065,
+      "grad_norm": 0.6055617928504944,
+      "learning_rate": 5.8994974874371866e-05,
+      "loss": 0.458,
+      "step": 1413
+    },
+    {
+      "epoch": 0.707,
+      "grad_norm": 0.6218027472496033,
+      "learning_rate": 5.889447236180905e-05,
+      "loss": 0.4944,
+      "step": 1414
+    },
+    {
+      "epoch": 0.7075,
+      "grad_norm": 0.6920521259307861,
+      "learning_rate": 5.879396984924623e-05,
+      "loss": 0.5255,
+      "step": 1415
+    },
+    {
+      "epoch": 0.708,
+      "grad_norm": 0.7051585912704468,
+      "learning_rate": 5.869346733668342e-05,
+      "loss": 0.5235,
+      "step": 1416
+    },
+    {
+      "epoch": 0.7085,
+      "grad_norm": 0.6307564973831177,
+      "learning_rate": 5.859296482412061e-05,
+      "loss": 0.4355,
+      "step": 1417
+    },
+    {
+      "epoch": 0.709,
+      "grad_norm": 0.7005728483200073,
+      "learning_rate": 5.849246231155779e-05,
+      "loss": 0.3887,
+      "step": 1418
+    },
+    {
+      "epoch": 0.7095,
+      "grad_norm": 0.6718925833702087,
+      "learning_rate": 5.839195979899498e-05,
+      "loss": 0.4081,
+      "step": 1419
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7462560534477234,
+      "learning_rate": 5.829145728643216e-05,
+      "loss": 0.4955,
+      "step": 1420
+    },
+    {
+      "epoch": 0.7105,
+      "grad_norm": 0.6004988551139832,
+      "learning_rate": 5.819095477386936e-05,
+      "loss": 0.4488,
+      "step": 1421
+    },
+    {
+      "epoch": 0.711,
+      "grad_norm": 0.6675456166267395,
+      "learning_rate": 5.809045226130654e-05,
+      "loss": 0.4702,
+      "step": 1422
+    },
+    {
+      "epoch": 0.7115,
+      "grad_norm": 0.6721092462539673,
+      "learning_rate": 5.7989949748743724e-05,
+      "loss": 0.4753,
+      "step": 1423
+    },
+    {
+      "epoch": 0.712,
+      "grad_norm": 0.7630552649497986,
+      "learning_rate": 5.7889447236180904e-05,
+      "loss": 0.5787,
+      "step": 1424
+    },
+    {
+      "epoch": 0.7125,
+      "grad_norm": 0.6053659319877625,
+      "learning_rate": 5.778894472361809e-05,
+      "loss": 0.3587,
+      "step": 1425
+    },
+    {
+      "epoch": 0.713,
+      "grad_norm": 0.6678106784820557,
+      "learning_rate": 5.7688442211055284e-05,
+      "loss": 0.4391,
+      "step": 1426
+    },
+    {
+      "epoch": 0.7135,
+      "grad_norm": 0.5987004041671753,
+      "learning_rate": 5.7587939698492464e-05,
+      "loss": 0.362,
+      "step": 1427
+    },
+    {
+      "epoch": 0.714,
+      "grad_norm": 0.8072169423103333,
+      "learning_rate": 5.748743718592965e-05,
+      "loss": 0.4969,
+      "step": 1428
+    },
+    {
+      "epoch": 0.7145,
+      "grad_norm": 0.5814105868339539,
+      "learning_rate": 5.738693467336683e-05,
+      "loss": 0.3358,
+      "step": 1429
+    },
+    {
+      "epoch": 0.715,
+      "grad_norm": 0.7491559982299805,
+      "learning_rate": 5.728643216080403e-05,
+      "loss": 0.4568,
+      "step": 1430
+    },
+    {
+      "epoch": 0.7155,
+      "grad_norm": 0.6187725067138672,
+      "learning_rate": 5.718592964824121e-05,
+      "loss": 0.3608,
+      "step": 1431
+    },
+    {
+      "epoch": 0.716,
+      "grad_norm": 0.7313234210014343,
+      "learning_rate": 5.7085427135678396e-05,
+      "loss": 0.524,
+      "step": 1432
+    },
+    {
+      "epoch": 0.7165,
+      "grad_norm": 0.5682476162910461,
+      "learning_rate": 5.6984924623115576e-05,
+      "loss": 0.3746,
+      "step": 1433
+    },
+    {
+      "epoch": 0.717,
+      "grad_norm": 0.6439304947853088,
+      "learning_rate": 5.688442211055277e-05,
+      "loss": 0.4235,
+      "step": 1434
+    },
+    {
+      "epoch": 0.7175,
+      "grad_norm": 0.5471600890159607,
+      "learning_rate": 5.6783919597989955e-05,
+      "loss": 0.2955,
+      "step": 1435
+    },
+    {
+      "epoch": 0.718,
+      "grad_norm": 0.7312886118888855,
+      "learning_rate": 5.6683417085427135e-05,
+      "loss": 0.4672,
+      "step": 1436
+    },
+    {
+      "epoch": 0.7185,
+      "grad_norm": 0.626193642616272,
+      "learning_rate": 5.658291457286432e-05,
+      "loss": 0.4818,
+      "step": 1437
+    },
+    {
+      "epoch": 0.719,
+      "grad_norm": 0.6074478030204773,
+      "learning_rate": 5.6482412060301515e-05,
+      "loss": 0.4019,
+      "step": 1438
+    },
+    {
+      "epoch": 0.7195,
+      "grad_norm": 0.7731311917304993,
+      "learning_rate": 5.63819095477387e-05,
+      "loss": 0.463,
+      "step": 1439
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6433038115501404,
+      "learning_rate": 5.628140703517588e-05,
+      "loss": 0.3389,
+      "step": 1440
+    },
+    {
+      "epoch": 0.7205,
+      "grad_norm": 0.6128441691398621,
+      "learning_rate": 5.618090452261307e-05,
+      "loss": 0.3186,
+      "step": 1441
+    },
+    {
+      "epoch": 0.721,
+      "grad_norm": 0.6955687999725342,
+      "learning_rate": 5.608040201005026e-05,
+      "loss": 0.424,
+      "step": 1442
+    },
+    {
+      "epoch": 0.7215,
+      "grad_norm": 0.7239277958869934,
+      "learning_rate": 5.597989949748744e-05,
+      "loss": 0.3893,
+      "step": 1443
+    },
+    {
+      "epoch": 0.722,
+      "grad_norm": 0.6157107949256897,
+      "learning_rate": 5.587939698492463e-05,
+      "loss": 0.3179,
+      "step": 1444
+    },
+    {
+      "epoch": 0.7225,
+      "grad_norm": 0.6023396849632263,
+      "learning_rate": 5.577889447236181e-05,
+      "loss": 0.2951,
+      "step": 1445
+    },
+    {
+      "epoch": 0.723,
+      "grad_norm": 0.7131041884422302,
+      "learning_rate": 5.567839195979899e-05,
+      "loss": 0.3762,
+      "step": 1446
+    },
+    {
+      "epoch": 0.7235,
+      "grad_norm": 0.648732602596283,
+      "learning_rate": 5.5577889447236186e-05,
+      "loss": 0.2701,
+      "step": 1447
+    },
+    {
+      "epoch": 0.724,
+      "grad_norm": 0.599702000617981,
+      "learning_rate": 5.547738693467337e-05,
+      "loss": 0.298,
+      "step": 1448
+    },
+    {
+      "epoch": 0.7245,
+      "grad_norm": 0.620100200176239,
+      "learning_rate": 5.537688442211055e-05,
+      "loss": 0.3175,
+      "step": 1449
+    },
+    {
+      "epoch": 0.725,
+      "grad_norm": 0.6005038022994995,
+      "learning_rate": 5.527638190954774e-05,
+      "loss": 0.2934,
+      "step": 1450
+    },
+    {
+      "epoch": 0.7255,
+      "grad_norm": 0.5175260901451111,
+      "learning_rate": 5.517587939698493e-05,
+      "loss": 0.4184,
+      "step": 1451
+    },
+    {
+      "epoch": 0.726,
+      "grad_norm": 0.5498479604721069,
+      "learning_rate": 5.507537688442211e-05,
+      "loss": 0.4432,
+      "step": 1452
+    },
+    {
+      "epoch": 0.7265,
+      "grad_norm": 0.670079231262207,
+      "learning_rate": 5.49748743718593e-05,
+      "loss": 0.6993,
+      "step": 1453
+    },
+    {
+      "epoch": 0.727,
+      "grad_norm": 0.5345937013626099,
+      "learning_rate": 5.487437185929648e-05,
+      "loss": 0.4785,
+      "step": 1454
+    },
+    {
+      "epoch": 0.7275,
+      "grad_norm": 0.550835132598877,
+      "learning_rate": 5.477386934673368e-05,
+      "loss": 0.4725,
+      "step": 1455
+    },
+    {
+      "epoch": 0.728,
+      "grad_norm": 0.5656582117080688,
+      "learning_rate": 5.467336683417086e-05,
+      "loss": 0.404,
+      "step": 1456
+    },
+    {
+      "epoch": 0.7285,
+      "grad_norm": 0.6900332570075989,
+      "learning_rate": 5.4572864321608045e-05,
+      "loss": 0.537,
+      "step": 1457
+    },
+    {
+      "epoch": 0.729,
+      "grad_norm": 0.5815069079399109,
+      "learning_rate": 5.4472361809045224e-05,
+      "loss": 0.468,
+      "step": 1458
+    },
+    {
+      "epoch": 0.7295,
+      "grad_norm": 0.61700439453125,
+      "learning_rate": 5.437185929648242e-05,
+      "loss": 0.4669,
+      "step": 1459
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6132184863090515,
+      "learning_rate": 5.4271356783919604e-05,
+      "loss": 0.4476,
+      "step": 1460
+    },
+    {
+      "epoch": 0.7305,
+      "grad_norm": 0.5600353479385376,
+      "learning_rate": 5.4170854271356784e-05,
+      "loss": 0.3899,
+      "step": 1461
+    },
+    {
+      "epoch": 0.731,
+      "grad_norm": 0.5819986462593079,
+      "learning_rate": 5.407035175879397e-05,
+      "loss": 0.3557,
+      "step": 1462
+    },
+    {
+      "epoch": 0.7315,
+      "grad_norm": 0.5940659642219543,
+      "learning_rate": 5.3969849246231163e-05,
+      "loss": 0.3601,
+      "step": 1463
+    },
+    {
+      "epoch": 0.732,
+      "grad_norm": 0.563206136226654,
+      "learning_rate": 5.386934673366835e-05,
+      "loss": 0.3525,
+      "step": 1464
+    },
+    {
+      "epoch": 0.7325,
+      "grad_norm": 0.7950224876403809,
+      "learning_rate": 5.376884422110553e-05,
+      "loss": 0.6904,
+      "step": 1465
+    },
+    {
+      "epoch": 0.733,
+      "grad_norm": 0.6856942176818848,
+      "learning_rate": 5.3668341708542716e-05,
+      "loss": 0.4796,
+      "step": 1466
+    },
+    {
+      "epoch": 0.7335,
+      "grad_norm": 0.7660188674926758,
+      "learning_rate": 5.356783919597991e-05,
+      "loss": 0.4309,
+      "step": 1467
+    },
+    {
+      "epoch": 0.734,
+      "grad_norm": 0.6034350991249084,
+      "learning_rate": 5.346733668341709e-05,
+      "loss": 0.3998,
+      "step": 1468
+    },
+    {
+      "epoch": 0.7345,
+      "grad_norm": 0.623004138469696,
+      "learning_rate": 5.3366834170854276e-05,
+      "loss": 0.3942,
+      "step": 1469
+    },
+    {
+      "epoch": 0.735,
+      "grad_norm": 0.641361653804779,
+      "learning_rate": 5.3266331658291455e-05,
+      "loss": 0.3651,
+      "step": 1470
+    },
+    {
+      "epoch": 0.7355,
+      "grad_norm": 0.7048323750495911,
+      "learning_rate": 5.316582914572864e-05,
+      "loss": 0.4759,
+      "step": 1471
+    },
+    {
+      "epoch": 0.736,
+      "grad_norm": 0.8306154608726501,
+      "learning_rate": 5.3065326633165835e-05,
+      "loss": 0.6084,
+      "step": 1472
+    },
+    {
+      "epoch": 0.7365,
+      "grad_norm": 0.6112460494041443,
+      "learning_rate": 5.296482412060302e-05,
+      "loss": 0.3178,
+      "step": 1473
+    },
+    {
+      "epoch": 0.737,
+      "grad_norm": 0.8105069994926453,
+      "learning_rate": 5.28643216080402e-05,
+      "loss": 0.48,
+      "step": 1474
+    },
+    {
+      "epoch": 0.7375,
+      "grad_norm": 0.5714189410209656,
+      "learning_rate": 5.276381909547739e-05,
+      "loss": 0.3541,
+      "step": 1475
+    },
+    {
+      "epoch": 0.738,
+      "grad_norm": 0.6870591044425964,
+      "learning_rate": 5.266331658291458e-05,
+      "loss": 0.4274,
+      "step": 1476
+    },
+    {
+      "epoch": 0.7385,
+      "grad_norm": 0.8062255382537842,
+      "learning_rate": 5.256281407035176e-05,
+      "loss": 0.4652,
+      "step": 1477
+    },
+    {
+      "epoch": 0.739,
+      "grad_norm": 0.8148736953735352,
+      "learning_rate": 5.246231155778895e-05,
+      "loss": 0.5113,
+      "step": 1478
+    },
+    {
+      "epoch": 0.7395,
+      "grad_norm": 0.815363883972168,
+      "learning_rate": 5.236180904522613e-05,
+      "loss": 0.4746,
+      "step": 1479
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6996128559112549,
+      "learning_rate": 5.226130653266332e-05,
+      "loss": 0.4496,
+      "step": 1480
+    },
+    {
+      "epoch": 0.7405,
+      "grad_norm": 0.5806979537010193,
+      "learning_rate": 5.2160804020100507e-05,
+      "loss": 0.3588,
+      "step": 1481
+    },
+    {
+      "epoch": 0.741,
+      "grad_norm": 0.7564975619316101,
+      "learning_rate": 5.206030150753769e-05,
+      "loss": 0.5173,
+      "step": 1482
+    },
+    {
+      "epoch": 0.7415,
+      "grad_norm": 0.6974437236785889,
+      "learning_rate": 5.195979899497487e-05,
+      "loss": 0.3436,
+      "step": 1483
+    },
+    {
+      "epoch": 0.742,
+      "grad_norm": 0.690948486328125,
+      "learning_rate": 5.1859296482412066e-05,
+      "loss": 0.3862,
+      "step": 1484
+    },
+    {
+      "epoch": 0.7425,
+      "grad_norm": 0.7029353380203247,
+      "learning_rate": 5.175879396984925e-05,
+      "loss": 0.4503,
+      "step": 1485
+    },
+    {
+      "epoch": 0.743,
+      "grad_norm": 0.69179767370224,
+      "learning_rate": 5.165829145728643e-05,
+      "loss": 0.3972,
+      "step": 1486
+    },
+    {
+      "epoch": 0.7435,
+      "grad_norm": 0.6400958299636841,
+      "learning_rate": 5.155778894472362e-05,
+      "loss": 0.3819,
+      "step": 1487
+    },
+    {
+      "epoch": 0.744,
+      "grad_norm": 0.6433467864990234,
+      "learning_rate": 5.145728643216081e-05,
+      "loss": 0.4206,
+      "step": 1488
+    },
+    {
+      "epoch": 0.7445,
+      "grad_norm": 0.5991151928901672,
+      "learning_rate": 5.135678391959799e-05,
+      "loss": 0.3537,
+      "step": 1489
+    },
+    {
+      "epoch": 0.745,
+      "grad_norm": 0.6432679295539856,
+      "learning_rate": 5.125628140703518e-05,
+      "loss": 0.3306,
+      "step": 1490
+    },
+    {
+      "epoch": 0.7455,
+      "grad_norm": 0.6517552733421326,
+      "learning_rate": 5.1155778894472365e-05,
+      "loss": 0.3081,
+      "step": 1491
+    },
+    {
+      "epoch": 0.746,
+      "grad_norm": 0.506137490272522,
+      "learning_rate": 5.1055276381909544e-05,
+      "loss": 0.2409,
+      "step": 1492
+    },
+    {
+      "epoch": 0.7465,
+      "grad_norm": 0.5608853697776794,
+      "learning_rate": 5.095477386934674e-05,
+      "loss": 0.2784,
+      "step": 1493
+    },
+    {
+      "epoch": 0.747,
+      "grad_norm": 0.5695006251335144,
+      "learning_rate": 5.0854271356783924e-05,
+      "loss": 0.2163,
+      "step": 1494
+    },
+    {
+      "epoch": 0.7475,
+      "grad_norm": 0.6422480940818787,
+      "learning_rate": 5.0753768844221104e-05,
+      "loss": 0.3484,
+      "step": 1495
+    },
+    {
+      "epoch": 0.748,
+      "grad_norm": 0.6399369835853577,
+      "learning_rate": 5.065326633165829e-05,
+      "loss": 0.3092,
+      "step": 1496
+    },
+    {
+      "epoch": 0.7485,
+      "grad_norm": 0.5133620500564575,
+      "learning_rate": 5.0552763819095484e-05,
+      "loss": 0.2231,
+      "step": 1497
+    },
+    {
+      "epoch": 0.749,
+      "grad_norm": 0.6734706163406372,
+      "learning_rate": 5.045226130653266e-05,
+      "loss": 0.3048,
+      "step": 1498
+    },
+    {
+      "epoch": 0.7495,
+      "grad_norm": 0.8790683150291443,
+      "learning_rate": 5.035175879396985e-05,
+      "loss": 0.4484,
+      "step": 1499
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6768514513969421,
+      "learning_rate": 5.0251256281407036e-05,
+      "loss": 0.3395,
+      "step": 1500
+    },
+    {
+      "epoch": 0.7505,
+      "grad_norm": 0.5504356622695923,
+      "learning_rate": 5.015075376884423e-05,
+      "loss": 0.522,
+      "step": 1501
+    },
+    {
+      "epoch": 0.751,
+      "grad_norm": 0.6592190861701965,
+      "learning_rate": 5.005025125628141e-05,
+      "loss": 0.578,
+      "step": 1502
+    },
+    {
+      "epoch": 0.7515,
+      "grad_norm": 0.6240267157554626,
+      "learning_rate": 4.9949748743718596e-05,
+      "loss": 0.5037,
+      "step": 1503
+    },
+    {
+      "epoch": 0.752,
+      "grad_norm": 0.6037539839744568,
+      "learning_rate": 4.984924623115578e-05,
+      "loss": 0.4466,
+      "step": 1504
+    },
+    {
+      "epoch": 0.7525,
+      "grad_norm": 0.7022961974143982,
+      "learning_rate": 4.974874371859297e-05,
+      "loss": 0.5211,
+      "step": 1505
+    },
+    {
+      "epoch": 0.753,
+      "grad_norm": 0.6277244091033936,
+      "learning_rate": 4.9648241206030155e-05,
+      "loss": 0.507,
+      "step": 1506
+    },
+    {
+      "epoch": 0.7535,
+      "grad_norm": 0.782058596611023,
+      "learning_rate": 4.954773869346734e-05,
+      "loss": 0.6366,
+      "step": 1507
+    },
+    {
+      "epoch": 0.754,
+      "grad_norm": 0.6159875988960266,
+      "learning_rate": 4.944723618090453e-05,
+      "loss": 0.3698,
+      "step": 1508
+    },
+    {
+      "epoch": 0.7545,
+      "grad_norm": 0.5792503952980042,
+      "learning_rate": 4.934673366834171e-05,
+      "loss": 0.4215,
+      "step": 1509
+    },
+    {
+      "epoch": 0.755,
+      "grad_norm": 0.69441819190979,
+      "learning_rate": 4.92462311557789e-05,
+      "loss": 0.4794,
+      "step": 1510
+    },
+    {
+      "epoch": 0.7555,
+      "grad_norm": 0.6756144165992737,
+      "learning_rate": 4.914572864321608e-05,
+      "loss": 0.4104,
+      "step": 1511
+    },
+    {
+      "epoch": 0.756,
+      "grad_norm": 0.6601399779319763,
+      "learning_rate": 4.9045226130653274e-05,
+      "loss": 0.4223,
+      "step": 1512
+    },
+    {
+      "epoch": 0.7565,
+      "grad_norm": 0.5763709545135498,
+      "learning_rate": 4.8944723618090454e-05,
+      "loss": 0.3465,
+      "step": 1513
+    },
+    {
+      "epoch": 0.757,
+      "grad_norm": 0.5687897205352783,
+      "learning_rate": 4.884422110552764e-05,
+      "loss": 0.3901,
+      "step": 1514
+    },
+    {
+      "epoch": 0.7575,
+      "grad_norm": 0.7059202790260315,
+      "learning_rate": 4.874371859296483e-05,
+      "loss": 0.4109,
+      "step": 1515
+    },
+    {
+      "epoch": 0.758,
+      "grad_norm": 0.6452916264533997,
+      "learning_rate": 4.864321608040201e-05,
+      "loss": 0.4459,
+      "step": 1516
+    },
+    {
+      "epoch": 0.7585,
+      "grad_norm": 0.7344638705253601,
+      "learning_rate": 4.85427135678392e-05,
+      "loss": 0.4713,
+      "step": 1517
+    },
+    {
+      "epoch": 0.759,
+      "grad_norm": 0.7311956882476807,
+      "learning_rate": 4.844221105527638e-05,
+      "loss": 0.4966,
+      "step": 1518
+    },
+    {
+      "epoch": 0.7595,
+      "grad_norm": 0.724567174911499,
+      "learning_rate": 4.834170854271357e-05,
+      "loss": 0.4757,
+      "step": 1519
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7957019209861755,
+      "learning_rate": 4.824120603015075e-05,
+      "loss": 0.6176,
+      "step": 1520
+    },
+    {
+      "epoch": 0.7605,
+      "grad_norm": 0.6060484051704407,
+      "learning_rate": 4.8140703517587946e-05,
+      "loss": 0.3572,
+      "step": 1521
+    },
+    {
+      "epoch": 0.761,
+      "grad_norm": 0.695808470249176,
+      "learning_rate": 4.8040201005025125e-05,
+      "loss": 0.3737,
+      "step": 1522
+    },
+    {
+      "epoch": 0.7615,
+      "grad_norm": 0.6713405251502991,
+      "learning_rate": 4.793969849246231e-05,
+      "loss": 0.4212,
+      "step": 1523
+    },
+    {
+      "epoch": 0.762,
+      "grad_norm": 0.6103047132492065,
+      "learning_rate": 4.78391959798995e-05,
+      "loss": 0.3422,
+      "step": 1524
+    },
+    {
+      "epoch": 0.7625,
+      "grad_norm": 0.720363199710846,
+      "learning_rate": 4.7738693467336685e-05,
+      "loss": 0.436,
+      "step": 1525
+    },
+    {
+      "epoch": 0.763,
+      "grad_norm": 0.6375768780708313,
+      "learning_rate": 4.763819095477387e-05,
+      "loss": 0.3806,
+      "step": 1526
+    },
+    {
+      "epoch": 0.7635,
+      "grad_norm": 0.592395007610321,
+      "learning_rate": 4.753768844221106e-05,
+      "loss": 0.3644,
+      "step": 1527
+    },
+    {
+      "epoch": 0.764,
+      "grad_norm": 0.5654150247573853,
+      "learning_rate": 4.7437185929648244e-05,
+      "loss": 0.3369,
+      "step": 1528
+    },
+    {
+      "epoch": 0.7645,
+      "grad_norm": 0.5492650270462036,
+      "learning_rate": 4.733668341708543e-05,
+      "loss": 0.3046,
+      "step": 1529
+    },
+    {
+      "epoch": 0.765,
+      "grad_norm": 0.6452381014823914,
+      "learning_rate": 4.723618090452262e-05,
+      "loss": 0.4324,
+      "step": 1530
+    },
+    {
+      "epoch": 0.7655,
+      "grad_norm": 0.6236452460289001,
+      "learning_rate": 4.7135678391959804e-05,
+      "loss": 0.3978,
+      "step": 1531
+    },
+    {
+      "epoch": 0.766,
+      "grad_norm": 0.6496348977088928,
+      "learning_rate": 4.703517587939698e-05,
+      "loss": 0.4425,
+      "step": 1532
+    },
+    {
+      "epoch": 0.7665,
+      "grad_norm": 0.8068647980690002,
+      "learning_rate": 4.6934673366834177e-05,
+      "loss": 0.4255,
+      "step": 1533
+    },
+    {
+      "epoch": 0.767,
+      "grad_norm": 0.7566070556640625,
+      "learning_rate": 4.6834170854271356e-05,
+      "loss": 0.4497,
+      "step": 1534
+    },
+    {
+      "epoch": 0.7675,
+      "grad_norm": 0.6734246015548706,
+      "learning_rate": 4.673366834170855e-05,
+      "loss": 0.431,
+      "step": 1535
+    },
+    {
+      "epoch": 0.768,
+      "grad_norm": 0.5811039805412292,
+      "learning_rate": 4.663316582914573e-05,
+      "loss": 0.3123,
+      "step": 1536
+    },
+    {
+      "epoch": 0.7685,
+      "grad_norm": 0.7292871475219727,
+      "learning_rate": 4.653266331658292e-05,
+      "loss": 0.3435,
+      "step": 1537
+    },
+    {
+      "epoch": 0.769,
+      "grad_norm": 0.6840037107467651,
+      "learning_rate": 4.64321608040201e-05,
+      "loss": 0.4229,
+      "step": 1538
+    },
+    {
+      "epoch": 0.7695,
+      "grad_norm": 0.5893954634666443,
+      "learning_rate": 4.633165829145729e-05,
+      "loss": 0.3388,
+      "step": 1539
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6605488657951355,
+      "learning_rate": 4.6231155778894475e-05,
+      "loss": 0.3851,
+      "step": 1540
+    },
+    {
+      "epoch": 0.7705,
+      "grad_norm": 0.62589031457901,
+      "learning_rate": 4.6130653266331655e-05,
+      "loss": 0.3658,
+      "step": 1541
+    },
+    {
+      "epoch": 0.771,
+      "grad_norm": 0.7203136682510376,
+      "learning_rate": 4.603015075376885e-05,
+      "loss": 0.4152,
+      "step": 1542
+    },
+    {
+      "epoch": 0.7715,
+      "grad_norm": 0.6860759258270264,
+      "learning_rate": 4.592964824120603e-05,
+      "loss": 0.3796,
+      "step": 1543
+    },
+    {
+      "epoch": 0.772,
+      "grad_norm": 0.624264657497406,
+      "learning_rate": 4.582914572864322e-05,
+      "loss": 0.3524,
+      "step": 1544
+    },
+    {
+      "epoch": 0.7725,
+      "grad_norm": 0.6847519874572754,
+      "learning_rate": 4.57286432160804e-05,
+      "loss": 0.3784,
+      "step": 1545
+    },
+    {
+      "epoch": 0.773,
+      "grad_norm": 0.5414067506790161,
+      "learning_rate": 4.5628140703517594e-05,
+      "loss": 0.2336,
+      "step": 1546
+    },
+    {
+      "epoch": 0.7735,
+      "grad_norm": 0.7270304560661316,
+      "learning_rate": 4.5527638190954774e-05,
+      "loss": 0.3126,
+      "step": 1547
+    },
+    {
+      "epoch": 0.774,
+      "grad_norm": 0.6397119760513306,
+      "learning_rate": 4.542713567839196e-05,
+      "loss": 0.2994,
+      "step": 1548
+    },
+    {
+      "epoch": 0.7745,
+      "grad_norm": 0.782112717628479,
+      "learning_rate": 4.532663316582915e-05,
+      "loss": 0.4381,
+      "step": 1549
+    },
+    {
+      "epoch": 0.775,
+      "grad_norm": 0.5597090721130371,
+      "learning_rate": 4.522613065326633e-05,
+      "loss": 0.2844,
+      "step": 1550
+    },
+    {
+      "epoch": 0.7755,
+      "grad_norm": 0.5405400395393372,
+      "learning_rate": 4.512562814070352e-05,
+      "loss": 0.5066,
+      "step": 1551
+    },
+    {
+      "epoch": 0.776,
+      "grad_norm": 0.5049933791160583,
+      "learning_rate": 4.5025125628140706e-05,
+      "loss": 0.4278,
+      "step": 1552
+    },
+    {
+      "epoch": 0.7765,
+      "grad_norm": 0.5136622786521912,
+      "learning_rate": 4.492462311557789e-05,
+      "loss": 0.4384,
+      "step": 1553
+    },
+    {
+      "epoch": 0.777,
+      "grad_norm": 0.45853346586227417,
+      "learning_rate": 4.482412060301508e-05,
+      "loss": 0.3439,
+      "step": 1554
+    },
+    {
+      "epoch": 0.7775,
+      "grad_norm": 0.5482397675514221,
+      "learning_rate": 4.4723618090452266e-05,
+      "loss": 0.4884,
+      "step": 1555
+    },
+    {
+      "epoch": 0.778,
+      "grad_norm": 0.5246921181678772,
+      "learning_rate": 4.462311557788945e-05,
+      "loss": 0.4469,
+      "step": 1556
+    },
+    {
+      "epoch": 0.7785,
+      "grad_norm": 0.5995600819587708,
+      "learning_rate": 4.452261306532663e-05,
+      "loss": 0.5074,
+      "step": 1557
+    },
+    {
+      "epoch": 0.779,
+      "grad_norm": 0.5938295722007751,
+      "learning_rate": 4.4422110552763825e-05,
+      "loss": 0.4978,
+      "step": 1558
+    },
+    {
+      "epoch": 0.7795,
+      "grad_norm": 0.6469851136207581,
+      "learning_rate": 4.4321608040201005e-05,
+      "loss": 0.4825,
+      "step": 1559
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7019020318984985,
+      "learning_rate": 4.42211055276382e-05,
+      "loss": 0.5591,
+      "step": 1560
+    },
+    {
+      "epoch": 0.7805,
+      "grad_norm": 0.5239007472991943,
+      "learning_rate": 4.412060301507538e-05,
+      "loss": 0.3799,
+      "step": 1561
+    },
+    {
+      "epoch": 0.781,
+      "grad_norm": 0.6300528049468994,
+      "learning_rate": 4.4020100502512564e-05,
+      "loss": 0.4266,
+      "step": 1562
+    },
+    {
+      "epoch": 0.7815,
+      "grad_norm": 0.6061190366744995,
+      "learning_rate": 4.391959798994975e-05,
+      "loss": 0.4175,
+      "step": 1563
+    },
+    {
+      "epoch": 0.782,
+      "grad_norm": 0.5768359899520874,
+      "learning_rate": 4.381909547738694e-05,
+      "loss": 0.3765,
+      "step": 1564
+    },
+    {
+      "epoch": 0.7825,
+      "grad_norm": 0.6297916769981384,
+      "learning_rate": 4.3718592964824124e-05,
+      "loss": 0.3845,
+      "step": 1565
+    },
+    {
+      "epoch": 0.783,
+      "grad_norm": 0.648495614528656,
+      "learning_rate": 4.3618090452261303e-05,
+      "loss": 0.3762,
+      "step": 1566
+    },
+    {
+      "epoch": 0.7835,
+      "grad_norm": 0.6243414282798767,
+      "learning_rate": 4.35175879396985e-05,
+      "loss": 0.3552,
+      "step": 1567
+    },
+    {
+      "epoch": 0.784,
+      "grad_norm": 0.6298936009407043,
+      "learning_rate": 4.3417085427135676e-05,
+      "loss": 0.4894,
+      "step": 1568
+    },
+    {
+      "epoch": 0.7845,
+      "grad_norm": 0.5340626239776611,
+      "learning_rate": 4.331658291457287e-05,
+      "loss": 0.3039,
+      "step": 1569
+    },
+    {
+      "epoch": 0.785,
+      "grad_norm": 0.5981042385101318,
+      "learning_rate": 4.321608040201005e-05,
+      "loss": 0.4766,
+      "step": 1570
+    },
+    {
+      "epoch": 0.7855,
+      "grad_norm": 0.663563072681427,
+      "learning_rate": 4.311557788944724e-05,
+      "loss": 0.4351,
+      "step": 1571
+    },
+    {
+      "epoch": 0.786,
+      "grad_norm": 0.6706258654594421,
+      "learning_rate": 4.301507537688442e-05,
+      "loss": 0.4223,
+      "step": 1572
+    },
+    {
+      "epoch": 0.7865,
+      "grad_norm": 0.8861648440361023,
+      "learning_rate": 4.291457286432161e-05,
+      "loss": 0.4441,
+      "step": 1573
+    },
+    {
+      "epoch": 0.787,
+      "grad_norm": 0.6051937937736511,
+      "learning_rate": 4.2814070351758795e-05,
+      "loss": 0.3071,
+      "step": 1574
+    },
+    {
+      "epoch": 0.7875,
+      "grad_norm": 0.7022461295127869,
+      "learning_rate": 4.271356783919598e-05,
+      "loss": 0.4318,
+      "step": 1575
+    },
+    {
+      "epoch": 0.788,
+      "grad_norm": 0.6715277433395386,
+      "learning_rate": 4.261306532663317e-05,
+      "loss": 0.3442,
+      "step": 1576
+    },
+    {
+      "epoch": 0.7885,
+      "grad_norm": 0.7998569011688232,
+      "learning_rate": 4.2512562814070355e-05,
+      "loss": 0.3586,
+      "step": 1577
+    },
+    {
+      "epoch": 0.789,
+      "grad_norm": 0.6366814374923706,
+      "learning_rate": 4.241206030150754e-05,
+      "loss": 0.3634,
+      "step": 1578
+    },
+    {
+      "epoch": 0.7895,
+      "grad_norm": 0.6236107349395752,
+      "learning_rate": 4.231155778894473e-05,
+      "loss": 0.3973,
+      "step": 1579
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6412162184715271,
+      "learning_rate": 4.2211055276381914e-05,
+      "loss": 0.3558,
+      "step": 1580
+    },
+    {
+      "epoch": 0.7905,
+      "grad_norm": 0.6055232882499695,
+      "learning_rate": 4.21105527638191e-05,
+      "loss": 0.3044,
+      "step": 1581
+    },
+    {
+      "epoch": 0.791,
+      "grad_norm": 0.6669800281524658,
+      "learning_rate": 4.201005025125628e-05,
+      "loss": 0.4398,
+      "step": 1582
+    },
+    {
+      "epoch": 0.7915,
+      "grad_norm": 0.7510372996330261,
+      "learning_rate": 4.1909547738693474e-05,
+      "loss": 0.4653,
+      "step": 1583
+    },
+    {
+      "epoch": 0.792,
+      "grad_norm": 0.730017364025116,
+      "learning_rate": 4.180904522613065e-05,
+      "loss": 0.3591,
+      "step": 1584
+    },
+    {
+      "epoch": 0.7925,
+      "grad_norm": 0.6909388899803162,
+      "learning_rate": 4.170854271356784e-05,
+      "loss": 0.4409,
+      "step": 1585
+    },
+    {
+      "epoch": 0.793,
+      "grad_norm": 0.7873001098632812,
+      "learning_rate": 4.1608040201005026e-05,
+      "loss": 0.4716,
+      "step": 1586
+    },
+    {
+      "epoch": 0.7935,
+      "grad_norm": 0.7261164784431458,
+      "learning_rate": 4.150753768844221e-05,
+      "loss": 0.4739,
+      "step": 1587
+    },
+    {
+      "epoch": 0.794,
+      "grad_norm": 0.6229093074798584,
+      "learning_rate": 4.14070351758794e-05,
+      "loss": 0.3132,
+      "step": 1588
+    },
+    {
+      "epoch": 0.7945,
+      "grad_norm": 0.6704588532447815,
+      "learning_rate": 4.1306532663316586e-05,
+      "loss": 0.4346,
+      "step": 1589
+    },
+    {
+      "epoch": 0.795,
+      "grad_norm": 0.6334629058837891,
+      "learning_rate": 4.120603015075377e-05,
+      "loss": 0.3577,
+      "step": 1590
+    },
+    {
+      "epoch": 0.7955,
+      "grad_norm": 0.7896953821182251,
+      "learning_rate": 4.110552763819095e-05,
+      "loss": 0.3784,
+      "step": 1591
+    },
+    {
+      "epoch": 0.796,
+      "grad_norm": 0.611592173576355,
+      "learning_rate": 4.1005025125628145e-05,
+      "loss": 0.2669,
+      "step": 1592
+    },
+    {
+      "epoch": 0.7965,
+      "grad_norm": 0.6428189277648926,
+      "learning_rate": 4.0904522613065325e-05,
+      "loss": 0.3268,
+      "step": 1593
+    },
+    {
+      "epoch": 0.797,
+      "grad_norm": 0.7141292095184326,
+      "learning_rate": 4.080402010050252e-05,
+      "loss": 0.4236,
+      "step": 1594
+    },
+    {
+      "epoch": 0.7975,
+      "grad_norm": 0.5845600962638855,
+      "learning_rate": 4.07035175879397e-05,
+      "loss": 0.3443,
+      "step": 1595
+    },
+    {
+      "epoch": 0.798,
+      "grad_norm": 0.8169146180152893,
+      "learning_rate": 4.060301507537689e-05,
+      "loss": 0.3373,
+      "step": 1596
+    },
+    {
+      "epoch": 0.7985,
+      "grad_norm": 0.7912828326225281,
+      "learning_rate": 4.050251256281407e-05,
+      "loss": 0.4673,
+      "step": 1597
+    },
+    {
+      "epoch": 0.799,
+      "grad_norm": 0.5616620182991028,
+      "learning_rate": 4.040201005025126e-05,
+      "loss": 0.265,
+      "step": 1598
+    },
+    {
+      "epoch": 0.7995,
+      "grad_norm": 0.7994551062583923,
+      "learning_rate": 4.0301507537688444e-05,
+      "loss": 0.3427,
+      "step": 1599
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5124440789222717,
+      "learning_rate": 4.020100502512563e-05,
+      "loss": 0.2373,
+      "step": 1600
+    },
+    {
+      "epoch": 0.8,
+      "eval_loss": 0.4139021635055542,
+      "eval_runtime": 109.0333,
+      "eval_samples_per_second": 4.586,
+      "eval_steps_per_second": 4.586,
+      "step": 1600
+    },
+    {
+      "epoch": 0.8005,
+      "grad_norm": 0.5576016902923584,
+      "learning_rate": 4.010050251256282e-05,
+      "loss": 0.488,
+      "step": 1601
+    },
+    {
+      "epoch": 0.801,
+      "grad_norm": 0.6127573251724243,
+      "learning_rate": 4e-05,
+      "loss": 0.6135,
+      "step": 1602
+    },
+    {
+      "epoch": 0.8015,
+      "grad_norm": 0.5914636254310608,
+      "learning_rate": 3.989949748743719e-05,
+      "loss": 0.5322,
+      "step": 1603
+    },
+    {
+      "epoch": 0.802,
+      "grad_norm": 0.6623550057411194,
+      "learning_rate": 3.9798994974874376e-05,
+      "loss": 0.5313,
+      "step": 1604
+    },
+    {
+      "epoch": 0.8025,
+      "grad_norm": 0.6122729182243347,
+      "learning_rate": 3.969849246231156e-05,
+      "loss": 0.4221,
+      "step": 1605
+    },
+    {
+      "epoch": 0.803,
+      "grad_norm": 0.6187363266944885,
+      "learning_rate": 3.959798994974875e-05,
+      "loss": 0.5149,
+      "step": 1606
+    },
+    {
+      "epoch": 0.8035,
+      "grad_norm": 0.6618751287460327,
+      "learning_rate": 3.949748743718593e-05,
+      "loss": 0.5521,
+      "step": 1607
+    },
+    {
+      "epoch": 0.804,
+      "grad_norm": 0.6277629733085632,
+      "learning_rate": 3.9396984924623115e-05,
+      "loss": 0.471,
+      "step": 1608
+    },
+    {
+      "epoch": 0.8045,
+      "grad_norm": 0.6420491337776184,
+      "learning_rate": 3.92964824120603e-05,
+      "loss": 0.4397,
+      "step": 1609
+    },
+    {
+      "epoch": 0.805,
+      "grad_norm": 0.6185916662216187,
+      "learning_rate": 3.919597989949749e-05,
+      "loss": 0.4181,
+      "step": 1610
+    },
+    {
+      "epoch": 0.8055,
+      "grad_norm": 0.566835343837738,
+      "learning_rate": 3.9095477386934675e-05,
+      "loss": 0.3734,
+      "step": 1611
+    },
+    {
+      "epoch": 0.806,
+      "grad_norm": 0.6276429891586304,
+      "learning_rate": 3.899497487437186e-05,
+      "loss": 0.4512,
+      "step": 1612
+    },
+    {
+      "epoch": 0.8065,
+      "grad_norm": 0.7259454727172852,
+      "learning_rate": 3.889447236180905e-05,
+      "loss": 0.4727,
+      "step": 1613
+    },
+    {
+      "epoch": 0.807,
+      "grad_norm": 0.6564999222755432,
+      "learning_rate": 3.8793969849246234e-05,
+      "loss": 0.509,
+      "step": 1614
+    },
+    {
+      "epoch": 0.8075,
+      "grad_norm": 0.7257003784179688,
+      "learning_rate": 3.869346733668342e-05,
+      "loss": 0.519,
+      "step": 1615
+    },
+    {
+      "epoch": 0.808,
+      "grad_norm": 0.6809136867523193,
+      "learning_rate": 3.85929648241206e-05,
+      "loss": 0.4134,
+      "step": 1616
+    },
+    {
+      "epoch": 0.8085,
+      "grad_norm": 0.5957052707672119,
+      "learning_rate": 3.8492462311557794e-05,
+      "loss": 0.3074,
+      "step": 1617
+    },
+    {
+      "epoch": 0.809,
+      "grad_norm": 0.7124346494674683,
+      "learning_rate": 3.8391959798994973e-05,
+      "loss": 0.5571,
+      "step": 1618
+    },
+    {
+      "epoch": 0.8095,
+      "grad_norm": 0.7949675917625427,
+      "learning_rate": 3.829145728643217e-05,
+      "loss": 0.5415,
+      "step": 1619
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6418821811676025,
+      "learning_rate": 3.8190954773869346e-05,
+      "loss": 0.4471,
+      "step": 1620
+    },
+    {
+      "epoch": 0.8105,
+      "grad_norm": 0.6463232636451721,
+      "learning_rate": 3.809045226130653e-05,
+      "loss": 0.3637,
+      "step": 1621
+    },
+    {
+      "epoch": 0.811,
+      "grad_norm": 0.8067373633384705,
+      "learning_rate": 3.798994974874372e-05,
+      "loss": 0.4909,
+      "step": 1622
+    },
+    {
+      "epoch": 0.8115,
+      "grad_norm": 0.7840099334716797,
+      "learning_rate": 3.7889447236180906e-05,
+      "loss": 0.5712,
+      "step": 1623
+    },
+    {
+      "epoch": 0.812,
+      "grad_norm": 0.6682786345481873,
+      "learning_rate": 3.778894472361809e-05,
+      "loss": 0.3734,
+      "step": 1624
+    },
+    {
+      "epoch": 0.8125,
+      "grad_norm": 0.6121772527694702,
+      "learning_rate": 3.768844221105528e-05,
+      "loss": 0.3653,
+      "step": 1625
+    },
+    {
+      "epoch": 0.813,
+      "grad_norm": 0.5917052626609802,
+      "learning_rate": 3.7587939698492465e-05,
+      "loss": 0.3681,
+      "step": 1626
+    },
+    {
+      "epoch": 0.8135,
+      "grad_norm": 0.5569581985473633,
+      "learning_rate": 3.748743718592965e-05,
+      "loss": 0.337,
+      "step": 1627
+    },
+    {
+      "epoch": 0.814,
+      "grad_norm": 0.6722996234893799,
+      "learning_rate": 3.738693467336684e-05,
+      "loss": 0.3872,
+      "step": 1628
+    },
+    {
+      "epoch": 0.8145,
+      "grad_norm": 0.6076365113258362,
+      "learning_rate": 3.7286432160804025e-05,
+      "loss": 0.3806,
+      "step": 1629
+    },
+    {
+      "epoch": 0.815,
+      "grad_norm": 0.7922750115394592,
+      "learning_rate": 3.7185929648241204e-05,
+      "loss": 0.5109,
+      "step": 1630
+    },
+    {
+      "epoch": 0.8155,
+      "grad_norm": 0.5152855515480042,
+      "learning_rate": 3.708542713567839e-05,
+      "loss": 0.2795,
+      "step": 1631
+    },
+    {
+      "epoch": 0.816,
+      "grad_norm": 0.5825745463371277,
+      "learning_rate": 3.698492462311558e-05,
+      "loss": 0.3782,
+      "step": 1632
+    },
+    {
+      "epoch": 0.8165,
+      "grad_norm": 0.7476134300231934,
+      "learning_rate": 3.6884422110552764e-05,
+      "loss": 0.5121,
+      "step": 1633
+    },
+    {
+      "epoch": 0.817,
+      "grad_norm": 0.6665414571762085,
+      "learning_rate": 3.678391959798995e-05,
+      "loss": 0.3393,
+      "step": 1634
+    },
+    {
+      "epoch": 0.8175,
+      "grad_norm": 0.6580054759979248,
+      "learning_rate": 3.668341708542714e-05,
+      "loss": 0.443,
+      "step": 1635
+    },
+    {
+      "epoch": 0.818,
+      "grad_norm": 0.6498279571533203,
+      "learning_rate": 3.658291457286432e-05,
+      "loss": 0.3953,
+      "step": 1636
+    },
+    {
+      "epoch": 0.8185,
+      "grad_norm": 0.6642691493034363,
+      "learning_rate": 3.648241206030151e-05,
+      "loss": 0.4102,
+      "step": 1637
+    },
+    {
+      "epoch": 0.819,
+      "grad_norm": 0.783823549747467,
+      "learning_rate": 3.6381909547738696e-05,
+      "loss": 0.498,
+      "step": 1638
+    },
+    {
+      "epoch": 0.8195,
+      "grad_norm": 0.7009001970291138,
+      "learning_rate": 3.628140703517588e-05,
+      "loss": 0.4327,
+      "step": 1639
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6497199535369873,
+      "learning_rate": 3.618090452261307e-05,
+      "loss": 0.4206,
+      "step": 1640
+    },
+    {
+      "epoch": 0.8205,
+      "grad_norm": 0.6588165760040283,
+      "learning_rate": 3.608040201005025e-05,
+      "loss": 0.343,
+      "step": 1641
+    },
+    {
+      "epoch": 0.821,
+      "grad_norm": 0.9206802845001221,
+      "learning_rate": 3.597989949748744e-05,
+      "loss": 0.5308,
+      "step": 1642
+    },
+    {
+      "epoch": 0.8215,
+      "grad_norm": 0.7294663786888123,
+      "learning_rate": 3.587939698492462e-05,
+      "loss": 0.3933,
+      "step": 1643
+    },
+    {
+      "epoch": 0.822,
+      "grad_norm": 0.7004202604293823,
+      "learning_rate": 3.5778894472361815e-05,
+      "loss": 0.3952,
+      "step": 1644
+    },
+    {
+      "epoch": 0.8225,
+      "grad_norm": 0.6495623588562012,
+      "learning_rate": 3.5678391959798995e-05,
+      "loss": 0.3693,
+      "step": 1645
+    },
+    {
+      "epoch": 0.823,
+      "grad_norm": 0.5871478319168091,
+      "learning_rate": 3.557788944723618e-05,
+      "loss": 0.3012,
+      "step": 1646
+    },
+    {
+      "epoch": 0.8235,
+      "grad_norm": 0.6222943663597107,
+      "learning_rate": 3.547738693467337e-05,
+      "loss": 0.2897,
+      "step": 1647
+    },
+    {
+      "epoch": 0.824,
+      "grad_norm": 0.6424744129180908,
+      "learning_rate": 3.5376884422110554e-05,
+      "loss": 0.3685,
+      "step": 1648
+    },
+    {
+      "epoch": 0.8245,
+      "grad_norm": 0.896592915058136,
+      "learning_rate": 3.527638190954774e-05,
+      "loss": 0.4839,
+      "step": 1649
+    },
+    {
+      "epoch": 0.825,
+      "grad_norm": 0.6496111154556274,
+      "learning_rate": 3.517587939698493e-05,
+      "loss": 0.3061,
+      "step": 1650
+    },
+    {
+      "epoch": 0.8255,
+      "grad_norm": 0.629523754119873,
+      "learning_rate": 3.5075376884422114e-05,
+      "loss": 0.4046,
+      "step": 1651
+    },
+    {
+      "epoch": 0.826,
+      "grad_norm": 0.6341447830200195,
+      "learning_rate": 3.49748743718593e-05,
+      "loss": 0.5433,
+      "step": 1652
+    },
+    {
+      "epoch": 0.8265,
+      "grad_norm": 0.6374323964118958,
+      "learning_rate": 3.487437185929649e-05,
+      "loss": 0.5251,
+      "step": 1653
+    },
+    {
+      "epoch": 0.827,
+      "grad_norm": 0.5337071418762207,
+      "learning_rate": 3.4773869346733667e-05,
+      "loss": 0.4093,
+      "step": 1654
+    },
+    {
+      "epoch": 0.8275,
+      "grad_norm": 0.48636487126350403,
+      "learning_rate": 3.467336683417085e-05,
+      "loss": 0.3134,
+      "step": 1655
+    },
+    {
+      "epoch": 0.828,
+      "grad_norm": 0.5469774603843689,
+      "learning_rate": 3.457286432160804e-05,
+      "loss": 0.4012,
+      "step": 1656
+    },
+    {
+      "epoch": 0.8285,
+      "grad_norm": 0.7101122736930847,
+      "learning_rate": 3.4472361809045226e-05,
+      "loss": 0.5545,
+      "step": 1657
+    },
+    {
+      "epoch": 0.829,
+      "grad_norm": 0.6524446606636047,
+      "learning_rate": 3.437185929648241e-05,
+      "loss": 0.54,
+      "step": 1658
+    },
+    {
+      "epoch": 0.8295,
+      "grad_norm": 0.6362601518630981,
+      "learning_rate": 3.42713567839196e-05,
+      "loss": 0.4927,
+      "step": 1659
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5991029143333435,
+      "learning_rate": 3.4170854271356785e-05,
+      "loss": 0.4229,
+      "step": 1660
+    },
+    {
+      "epoch": 0.8305,
+      "grad_norm": 0.6555331349372864,
+      "learning_rate": 3.407035175879397e-05,
+      "loss": 0.4118,
+      "step": 1661
+    },
+    {
+      "epoch": 0.831,
+      "grad_norm": 0.6281815767288208,
+      "learning_rate": 3.396984924623116e-05,
+      "loss": 0.4318,
+      "step": 1662
+    },
+    {
+      "epoch": 0.8315,
+      "grad_norm": 0.5101354718208313,
+      "learning_rate": 3.3869346733668345e-05,
+      "loss": 0.3252,
+      "step": 1663
+    },
+    {
+      "epoch": 0.832,
+      "grad_norm": 0.7071707248687744,
+      "learning_rate": 3.3768844221105525e-05,
+      "loss": 0.5282,
+      "step": 1664
+    },
+    {
+      "epoch": 0.8325,
+      "grad_norm": 0.6886499524116516,
+      "learning_rate": 3.366834170854272e-05,
+      "loss": 0.4304,
+      "step": 1665
+    },
+    {
+      "epoch": 0.833,
+      "grad_norm": 0.7029287219047546,
+      "learning_rate": 3.35678391959799e-05,
+      "loss": 0.4935,
+      "step": 1666
+    },
+    {
+      "epoch": 0.8335,
+      "grad_norm": 0.6921351552009583,
+      "learning_rate": 3.346733668341709e-05,
+      "loss": 0.4387,
+      "step": 1667
+    },
+    {
+      "epoch": 0.834,
+      "grad_norm": 0.6940300464630127,
+      "learning_rate": 3.336683417085427e-05,
+      "loss": 0.3919,
+      "step": 1668
+    },
+    {
+      "epoch": 0.8345,
+      "grad_norm": 0.6877486705780029,
+      "learning_rate": 3.3266331658291464e-05,
+      "loss": 0.419,
+      "step": 1669
+    },
+    {
+      "epoch": 0.835,
+      "grad_norm": 0.576531708240509,
+      "learning_rate": 3.3165829145728643e-05,
+      "loss": 0.3301,
+      "step": 1670
+    },
+    {
+      "epoch": 0.8355,
+      "grad_norm": 0.5293341875076294,
+      "learning_rate": 3.306532663316583e-05,
+      "loss": 0.2883,
+      "step": 1671
+    },
+    {
+      "epoch": 0.836,
+      "grad_norm": 0.6576932668685913,
+      "learning_rate": 3.2964824120603016e-05,
+      "loss": 0.3858,
+      "step": 1672
+    },
+    {
+      "epoch": 0.8365,
+      "grad_norm": 0.6273109912872314,
+      "learning_rate": 3.28643216080402e-05,
+      "loss": 0.3038,
+      "step": 1673
+    },
+    {
+      "epoch": 0.837,
+      "grad_norm": 0.6952325105667114,
+      "learning_rate": 3.276381909547739e-05,
+      "loss": 0.4064,
+      "step": 1674
+    },
+    {
+      "epoch": 0.8375,
+      "grad_norm": 0.6681288480758667,
+      "learning_rate": 3.2663316582914576e-05,
+      "loss": 0.4149,
+      "step": 1675
+    },
+    {
+      "epoch": 0.838,
+      "grad_norm": 0.6957809925079346,
+      "learning_rate": 3.256281407035176e-05,
+      "loss": 0.4711,
+      "step": 1676
+    },
+    {
+      "epoch": 0.8385,
+      "grad_norm": 0.513947069644928,
+      "learning_rate": 3.246231155778894e-05,
+      "loss": 0.3031,
+      "step": 1677
+    },
+    {
+      "epoch": 0.839,
+      "grad_norm": 0.7426998019218445,
+      "learning_rate": 3.2361809045226135e-05,
+      "loss": 0.4274,
+      "step": 1678
+    },
+    {
+      "epoch": 0.8395,
+      "grad_norm": 0.7126373052597046,
+      "learning_rate": 3.2261306532663315e-05,
+      "loss": 0.4352,
+      "step": 1679
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5618422627449036,
+      "learning_rate": 3.21608040201005e-05,
+      "loss": 0.2993,
+      "step": 1680
+    },
+    {
+      "epoch": 0.8405,
+      "grad_norm": 0.6278876066207886,
+      "learning_rate": 3.206030150753769e-05,
+      "loss": 0.3242,
+      "step": 1681
+    },
+    {
+      "epoch": 0.841,
+      "grad_norm": 0.6913187503814697,
+      "learning_rate": 3.1959798994974875e-05,
+      "loss": 0.4115,
+      "step": 1682
+    },
+    {
+      "epoch": 0.8415,
+      "grad_norm": 0.6365401744842529,
+      "learning_rate": 3.185929648241206e-05,
+      "loss": 0.3421,
+      "step": 1683
+    },
+    {
+      "epoch": 0.842,
+      "grad_norm": 0.6104812026023865,
+      "learning_rate": 3.175879396984925e-05,
+      "loss": 0.4239,
+      "step": 1684
+    },
+    {
+      "epoch": 0.8425,
+      "grad_norm": 0.6641998291015625,
+      "learning_rate": 3.1658291457286434e-05,
+      "loss": 0.3463,
+      "step": 1685
+    },
+    {
+      "epoch": 0.843,
+      "grad_norm": 0.6832804083824158,
+      "learning_rate": 3.155778894472362e-05,
+      "loss": 0.405,
+      "step": 1686
+    },
+    {
+      "epoch": 0.8435,
+      "grad_norm": 0.7006424069404602,
+      "learning_rate": 3.145728643216081e-05,
+      "loss": 0.3416,
+      "step": 1687
+    },
+    {
+      "epoch": 0.844,
+      "grad_norm": 0.5932193994522095,
+      "learning_rate": 3.1356783919597993e-05,
+      "loss": 0.3124,
+      "step": 1688
+    },
+    {
+      "epoch": 0.8445,
+      "grad_norm": 0.6610186696052551,
+      "learning_rate": 3.125628140703517e-05,
+      "loss": 0.3661,
+      "step": 1689
+    },
+    {
+      "epoch": 0.845,
+      "grad_norm": 0.8019976019859314,
+      "learning_rate": 3.1155778894472366e-05,
+      "loss": 0.4785,
+      "step": 1690
+    },
+    {
+      "epoch": 0.8455,
+      "grad_norm": 0.6109830141067505,
+      "learning_rate": 3.1055276381909546e-05,
+      "loss": 0.301,
+      "step": 1691
+    },
+    {
+      "epoch": 0.846,
+      "grad_norm": 0.7036943435668945,
+      "learning_rate": 3.095477386934674e-05,
+      "loss": 0.392,
+      "step": 1692
+    },
+    {
+      "epoch": 0.8465,
+      "grad_norm": 0.6208803057670593,
+      "learning_rate": 3.085427135678392e-05,
+      "loss": 0.3157,
+      "step": 1693
+    },
+    {
+      "epoch": 0.847,
+      "grad_norm": 0.6118980050086975,
+      "learning_rate": 3.075376884422111e-05,
+      "loss": 0.2804,
+      "step": 1694
+    },
+    {
+      "epoch": 0.8475,
+      "grad_norm": 0.6996687650680542,
+      "learning_rate": 3.065326633165829e-05,
+      "loss": 0.3252,
+      "step": 1695
+    },
+    {
+      "epoch": 0.848,
+      "grad_norm": 0.684698760509491,
+      "learning_rate": 3.055276381909548e-05,
+      "loss": 0.3531,
+      "step": 1696
+    },
+    {
+      "epoch": 0.8485,
+      "grad_norm": 0.5951422452926636,
+      "learning_rate": 3.0452261306532665e-05,
+      "loss": 0.2787,
+      "step": 1697
+    },
+    {
+      "epoch": 0.849,
+      "grad_norm": 0.5187300443649292,
+      "learning_rate": 3.0351758793969855e-05,
+      "loss": 0.2421,
+      "step": 1698
+    },
+    {
+      "epoch": 0.8495,
+      "grad_norm": 0.6378542184829712,
+      "learning_rate": 3.0251256281407038e-05,
+      "loss": 0.3036,
+      "step": 1699
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7344485521316528,
+      "learning_rate": 3.015075376884422e-05,
+      "loss": 0.3503,
+      "step": 1700
+    },
+    {
+      "epoch": 0.8505,
+      "grad_norm": 0.4944750666618347,
+      "learning_rate": 3.0050251256281408e-05,
+      "loss": 0.3573,
+      "step": 1701
+    },
+    {
+      "epoch": 0.851,
+      "grad_norm": 0.6047641634941101,
+      "learning_rate": 2.994974874371859e-05,
+      "loss": 0.4533,
+      "step": 1702
+    },
+    {
+      "epoch": 0.8515,
+      "grad_norm": 0.5493037104606628,
+      "learning_rate": 2.984924623115578e-05,
+      "loss": 0.4003,
+      "step": 1703
+    },
+    {
+      "epoch": 0.852,
+      "grad_norm": 0.6913791298866272,
+      "learning_rate": 2.9748743718592964e-05,
+      "loss": 0.6033,
+      "step": 1704
+    },
+    {
+      "epoch": 0.8525,
+      "grad_norm": 0.8119217753410339,
+      "learning_rate": 2.9648241206030153e-05,
+      "loss": 0.5477,
+      "step": 1705
+    },
+    {
+      "epoch": 0.853,
+      "grad_norm": 0.5093177556991577,
+      "learning_rate": 2.9547738693467337e-05,
+      "loss": 0.3588,
+      "step": 1706
+    },
+    {
+      "epoch": 0.8535,
+      "grad_norm": 0.5311038494110107,
+      "learning_rate": 2.9447236180904526e-05,
+      "loss": 0.3509,
+      "step": 1707
+    },
+    {
+      "epoch": 0.854,
+      "grad_norm": 0.5833839178085327,
+      "learning_rate": 2.934673366834171e-05,
+      "loss": 0.3684,
+      "step": 1708
+    },
+    {
+      "epoch": 0.8545,
+      "grad_norm": 0.5847336649894714,
+      "learning_rate": 2.9246231155778896e-05,
+      "loss": 0.4002,
+      "step": 1709
+    },
+    {
+      "epoch": 0.855,
+      "grad_norm": 0.6113662719726562,
+      "learning_rate": 2.914572864321608e-05,
+      "loss": 0.3701,
+      "step": 1710
+    },
+    {
+      "epoch": 0.8555,
+      "grad_norm": 0.6850919723510742,
+      "learning_rate": 2.904522613065327e-05,
+      "loss": 0.5344,
+      "step": 1711
+    },
+    {
+      "epoch": 0.856,
+      "grad_norm": 0.6709843277931213,
+      "learning_rate": 2.8944723618090452e-05,
+      "loss": 0.5288,
+      "step": 1712
+    },
+    {
+      "epoch": 0.8565,
+      "grad_norm": 0.6538962721824646,
+      "learning_rate": 2.8844221105527642e-05,
+      "loss": 0.5231,
+      "step": 1713
+    },
+    {
+      "epoch": 0.857,
+      "grad_norm": 0.6653441190719604,
+      "learning_rate": 2.8743718592964825e-05,
+      "loss": 0.4538,
+      "step": 1714
+    },
+    {
+      "epoch": 0.8575,
+      "grad_norm": 0.6139349341392517,
+      "learning_rate": 2.8643216080402015e-05,
+      "loss": 0.4759,
+      "step": 1715
+    },
+    {
+      "epoch": 0.858,
+      "grad_norm": 0.7201113700866699,
+      "learning_rate": 2.8542713567839198e-05,
+      "loss": 0.4387,
+      "step": 1716
+    },
+    {
+      "epoch": 0.8585,
+      "grad_norm": 0.6542975902557373,
+      "learning_rate": 2.8442211055276384e-05,
+      "loss": 0.3516,
+      "step": 1717
+    },
+    {
+      "epoch": 0.859,
+      "grad_norm": 0.6745914816856384,
+      "learning_rate": 2.8341708542713568e-05,
+      "loss": 0.4265,
+      "step": 1718
+    },
+    {
+      "epoch": 0.8595,
+      "grad_norm": 0.5845760107040405,
+      "learning_rate": 2.8241206030150757e-05,
+      "loss": 0.3527,
+      "step": 1719
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6596361398696899,
+      "learning_rate": 2.814070351758794e-05,
+      "loss": 0.4588,
+      "step": 1720
+    },
+    {
+      "epoch": 0.8605,
+      "grad_norm": 0.6279307007789612,
+      "learning_rate": 2.804020100502513e-05,
+      "loss": 0.4831,
+      "step": 1721
+    },
+    {
+      "epoch": 0.861,
+      "grad_norm": 0.6030738353729248,
+      "learning_rate": 2.7939698492462314e-05,
+      "loss": 0.3424,
+      "step": 1722
+    },
+    {
+      "epoch": 0.8615,
+      "grad_norm": 0.6914360523223877,
+      "learning_rate": 2.7839195979899497e-05,
+      "loss": 0.4534,
+      "step": 1723
+    },
+    {
+      "epoch": 0.862,
+      "grad_norm": 0.5811136960983276,
+      "learning_rate": 2.7738693467336686e-05,
+      "loss": 0.3369,
+      "step": 1724
+    },
+    {
+      "epoch": 0.8625,
+      "grad_norm": 0.6894232034683228,
+      "learning_rate": 2.763819095477387e-05,
+      "loss": 0.3871,
+      "step": 1725
+    },
+    {
+      "epoch": 0.863,
+      "grad_norm": 0.695335328578949,
+      "learning_rate": 2.7537688442211056e-05,
+      "loss": 0.5093,
+      "step": 1726
+    },
+    {
+      "epoch": 0.8635,
+      "grad_norm": 0.7209401726722717,
+      "learning_rate": 2.743718592964824e-05,
+      "loss": 0.4357,
+      "step": 1727
+    },
+    {
+      "epoch": 0.864,
+      "grad_norm": 0.6519538760185242,
+      "learning_rate": 2.733668341708543e-05,
+      "loss": 0.3773,
+      "step": 1728
+    },
+    {
+      "epoch": 0.8645,
+      "grad_norm": 0.6212464570999146,
+      "learning_rate": 2.7236180904522612e-05,
+      "loss": 0.3247,
+      "step": 1729
+    },
+    {
+      "epoch": 0.865,
+      "grad_norm": 0.6528953909873962,
+      "learning_rate": 2.7135678391959802e-05,
+      "loss": 0.3732,
+      "step": 1730
+    },
+    {
+      "epoch": 0.8655,
+      "grad_norm": 0.7183499932289124,
+      "learning_rate": 2.7035175879396985e-05,
+      "loss": 0.4657,
+      "step": 1731
+    },
+    {
+      "epoch": 0.866,
+      "grad_norm": 0.7579643130302429,
+      "learning_rate": 2.6934673366834175e-05,
+      "loss": 0.4051,
+      "step": 1732
+    },
+    {
+      "epoch": 0.8665,
+      "grad_norm": 0.6558851003646851,
+      "learning_rate": 2.6834170854271358e-05,
+      "loss": 0.3312,
+      "step": 1733
+    },
+    {
+      "epoch": 0.867,
+      "grad_norm": 0.6520231366157532,
+      "learning_rate": 2.6733668341708545e-05,
+      "loss": 0.3518,
+      "step": 1734
+    },
+    {
+      "epoch": 0.8675,
+      "grad_norm": 0.7598215341567993,
+      "learning_rate": 2.6633165829145728e-05,
+      "loss": 0.3937,
+      "step": 1735
+    },
+    {
+      "epoch": 0.868,
+      "grad_norm": 0.7400127649307251,
+      "learning_rate": 2.6532663316582917e-05,
+      "loss": 0.4207,
+      "step": 1736
+    },
+    {
+      "epoch": 0.8685,
+      "grad_norm": 0.7712739109992981,
+      "learning_rate": 2.64321608040201e-05,
+      "loss": 0.5268,
+      "step": 1737
+    },
+    {
+      "epoch": 0.869,
+      "grad_norm": 0.6129913330078125,
+      "learning_rate": 2.633165829145729e-05,
+      "loss": 0.3295,
+      "step": 1738
+    },
+    {
+      "epoch": 0.8695,
+      "grad_norm": 0.7502526044845581,
+      "learning_rate": 2.6231155778894474e-05,
+      "loss": 0.4469,
+      "step": 1739
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6569104790687561,
+      "learning_rate": 2.613065326633166e-05,
+      "loss": 0.3767,
+      "step": 1740
+    },
+    {
+      "epoch": 0.8705,
+      "grad_norm": 0.7063089609146118,
+      "learning_rate": 2.6030150753768847e-05,
+      "loss": 0.327,
+      "step": 1741
+    },
+    {
+      "epoch": 0.871,
+      "grad_norm": 1.0344231128692627,
+      "learning_rate": 2.5929648241206033e-05,
+      "loss": 0.4914,
+      "step": 1742
+    },
+    {
+      "epoch": 0.8715,
+      "grad_norm": 0.6093212962150574,
+      "learning_rate": 2.5829145728643216e-05,
+      "loss": 0.289,
+      "step": 1743
+    },
+    {
+      "epoch": 0.872,
+      "grad_norm": 0.6271454095840454,
+      "learning_rate": 2.5728643216080406e-05,
+      "loss": 0.3237,
+      "step": 1744
+    },
+    {
+      "epoch": 0.8725,
+      "grad_norm": 0.5010676980018616,
+      "learning_rate": 2.562814070351759e-05,
+      "loss": 0.2448,
+      "step": 1745
+    },
+    {
+      "epoch": 0.873,
+      "grad_norm": 0.6918530464172363,
+      "learning_rate": 2.5527638190954772e-05,
+      "loss": 0.3375,
+      "step": 1746
+    },
+    {
+      "epoch": 0.8735,
+      "grad_norm": 0.7002788782119751,
+      "learning_rate": 2.5427135678391962e-05,
+      "loss": 0.3918,
+      "step": 1747
+    },
+    {
+      "epoch": 0.874,
+      "grad_norm": 0.6033680438995361,
+      "learning_rate": 2.5326633165829145e-05,
+      "loss": 0.2687,
+      "step": 1748
+    },
+    {
+      "epoch": 0.8745,
+      "grad_norm": 0.7553046345710754,
+      "learning_rate": 2.522613065326633e-05,
+      "loss": 0.2674,
+      "step": 1749
+    },
+    {
+      "epoch": 0.875,
+      "grad_norm": 0.6541697382926941,
+      "learning_rate": 2.5125628140703518e-05,
+      "loss": 0.2882,
+      "step": 1750
+    },
+    {
+      "epoch": 0.8755,
+      "grad_norm": 0.5916823744773865,
+      "learning_rate": 2.5025125628140705e-05,
+      "loss": 0.5418,
+      "step": 1751
+    },
+    {
+      "epoch": 0.876,
+      "grad_norm": 0.49342840909957886,
+      "learning_rate": 2.492462311557789e-05,
+      "loss": 0.3741,
+      "step": 1752
+    },
+    {
+      "epoch": 0.8765,
+      "grad_norm": 0.5878573060035706,
+      "learning_rate": 2.4824120603015078e-05,
+      "loss": 0.5341,
+      "step": 1753
+    },
+    {
+      "epoch": 0.877,
+      "grad_norm": 0.6414496302604675,
+      "learning_rate": 2.4723618090452264e-05,
+      "loss": 0.6448,
+      "step": 1754
+    },
+    {
+      "epoch": 0.8775,
+      "grad_norm": 0.6701349020004272,
+      "learning_rate": 2.462311557788945e-05,
+      "loss": 0.664,
+      "step": 1755
+    },
+    {
+      "epoch": 0.878,
+      "grad_norm": 0.6071045398712158,
+      "learning_rate": 2.4522613065326637e-05,
+      "loss": 0.4645,
+      "step": 1756
+    },
+    {
+      "epoch": 0.8785,
+      "grad_norm": 0.654080331325531,
+      "learning_rate": 2.442211055276382e-05,
+      "loss": 0.4636,
+      "step": 1757
+    },
+    {
+      "epoch": 0.879,
+      "grad_norm": 0.602473258972168,
+      "learning_rate": 2.4321608040201007e-05,
+      "loss": 0.4677,
+      "step": 1758
+    },
+    {
+      "epoch": 0.8795,
+      "grad_norm": 0.5240753293037415,
+      "learning_rate": 2.422110552763819e-05,
+      "loss": 0.3698,
+      "step": 1759
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6632701754570007,
+      "learning_rate": 2.4120603015075376e-05,
+      "loss": 0.503,
+      "step": 1760
+    },
+    {
+      "epoch": 0.8805,
+      "grad_norm": 0.5482516884803772,
+      "learning_rate": 2.4020100502512563e-05,
+      "loss": 0.37,
+      "step": 1761
+    },
+    {
+      "epoch": 0.881,
+      "grad_norm": 0.7788616418838501,
+      "learning_rate": 2.391959798994975e-05,
+      "loss": 0.6022,
+      "step": 1762
+    },
+    {
+      "epoch": 0.8815,
+      "grad_norm": 0.7823307514190674,
+      "learning_rate": 2.3819095477386936e-05,
+      "loss": 0.5395,
+      "step": 1763
+    },
+    {
+      "epoch": 0.882,
+      "grad_norm": 0.7380117774009705,
+      "learning_rate": 2.3718592964824122e-05,
+      "loss": 0.4886,
+      "step": 1764
+    },
+    {
+      "epoch": 0.8825,
+      "grad_norm": 0.5757150053977966,
+      "learning_rate": 2.361809045226131e-05,
+      "loss": 0.3305,
+      "step": 1765
+    },
+    {
+      "epoch": 0.883,
+      "grad_norm": 0.6612235903739929,
+      "learning_rate": 2.351758793969849e-05,
+      "loss": 0.4434,
+      "step": 1766
+    },
+    {
+      "epoch": 0.8835,
+      "grad_norm": 0.7172591090202332,
+      "learning_rate": 2.3417085427135678e-05,
+      "loss": 0.498,
+      "step": 1767
+    },
+    {
+      "epoch": 0.884,
+      "grad_norm": 0.6225490570068359,
+      "learning_rate": 2.3316582914572865e-05,
+      "loss": 0.4555,
+      "step": 1768
+    },
+    {
+      "epoch": 0.8845,
+      "grad_norm": 0.685167670249939,
+      "learning_rate": 2.321608040201005e-05,
+      "loss": 0.4831,
+      "step": 1769
+    },
+    {
+      "epoch": 0.885,
+      "grad_norm": 0.5485687255859375,
+      "learning_rate": 2.3115577889447238e-05,
+      "loss": 0.2988,
+      "step": 1770
+    },
+    {
+      "epoch": 0.8855,
+      "grad_norm": 0.8124721646308899,
+      "learning_rate": 2.3015075376884424e-05,
+      "loss": 0.6016,
+      "step": 1771
+    },
+    {
+      "epoch": 0.886,
+      "grad_norm": 0.6508186459541321,
+      "learning_rate": 2.291457286432161e-05,
+      "loss": 0.4901,
+      "step": 1772
+    },
+    {
+      "epoch": 0.8865,
+      "grad_norm": 0.7299104928970337,
+      "learning_rate": 2.2814070351758797e-05,
+      "loss": 0.4045,
+      "step": 1773
+    },
+    {
+      "epoch": 0.887,
+      "grad_norm": 0.7113926410675049,
+      "learning_rate": 2.271356783919598e-05,
+      "loss": 0.4389,
+      "step": 1774
+    },
+    {
+      "epoch": 0.8875,
+      "grad_norm": 0.6425351500511169,
+      "learning_rate": 2.2613065326633167e-05,
+      "loss": 0.3902,
+      "step": 1775
+    },
+    {
+      "epoch": 0.888,
+      "grad_norm": 0.5860398411750793,
+      "learning_rate": 2.2512562814070353e-05,
+      "loss": 0.29,
+      "step": 1776
+    },
+    {
+      "epoch": 0.8885,
+      "grad_norm": 0.6536673903465271,
+      "learning_rate": 2.241206030150754e-05,
+      "loss": 0.4359,
+      "step": 1777
+    },
+    {
+      "epoch": 0.889,
+      "grad_norm": 0.8302834033966064,
+      "learning_rate": 2.2311557788944726e-05,
+      "loss": 0.491,
+      "step": 1778
+    },
+    {
+      "epoch": 0.8895,
+      "grad_norm": 0.6673223376274109,
+      "learning_rate": 2.2211055276381913e-05,
+      "loss": 0.3016,
+      "step": 1779
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6284530758857727,
+      "learning_rate": 2.21105527638191e-05,
+      "loss": 0.337,
+      "step": 1780
+    },
+    {
+      "epoch": 0.8905,
+      "grad_norm": 0.632018506526947,
+      "learning_rate": 2.2010050251256282e-05,
+      "loss": 0.3538,
+      "step": 1781
+    },
+    {
+      "epoch": 0.891,
+      "grad_norm": 0.5739997625350952,
+      "learning_rate": 2.190954773869347e-05,
+      "loss": 0.3817,
+      "step": 1782
+    },
+    {
+      "epoch": 0.8915,
+      "grad_norm": 0.6176656484603882,
+      "learning_rate": 2.1809045226130652e-05,
+      "loss": 0.3691,
+      "step": 1783
+    },
+    {
+      "epoch": 0.892,
+      "grad_norm": 0.7113850712776184,
+      "learning_rate": 2.1708542713567838e-05,
+      "loss": 0.4276,
+      "step": 1784
+    },
+    {
+      "epoch": 0.8925,
+      "grad_norm": 0.6931306719779968,
+      "learning_rate": 2.1608040201005025e-05,
+      "loss": 0.3706,
+      "step": 1785
+    },
+    {
+      "epoch": 0.893,
+      "grad_norm": 0.7537098526954651,
+      "learning_rate": 2.150753768844221e-05,
+      "loss": 0.4704,
+      "step": 1786
+    },
+    {
+      "epoch": 0.8935,
+      "grad_norm": 0.6854217052459717,
+      "learning_rate": 2.1407035175879398e-05,
+      "loss": 0.4135,
+      "step": 1787
+    },
+    {
+      "epoch": 0.894,
+      "grad_norm": 0.6507512331008911,
+      "learning_rate": 2.1306532663316584e-05,
+      "loss": 0.3677,
+      "step": 1788
+    },
+    {
+      "epoch": 0.8945,
+      "grad_norm": 0.682126522064209,
+      "learning_rate": 2.120603015075377e-05,
+      "loss": 0.3649,
+      "step": 1789
+    },
+    {
+      "epoch": 0.895,
+      "grad_norm": 0.6120560765266418,
+      "learning_rate": 2.1105527638190957e-05,
+      "loss": 0.2983,
+      "step": 1790
+    },
+    {
+      "epoch": 0.8955,
+      "grad_norm": 0.653452455997467,
+      "learning_rate": 2.100502512562814e-05,
+      "loss": 0.3952,
+      "step": 1791
+    },
+    {
+      "epoch": 0.896,
+      "grad_norm": 0.7148404121398926,
+      "learning_rate": 2.0904522613065327e-05,
+      "loss": 0.4731,
+      "step": 1792
+    },
+    {
+      "epoch": 0.8965,
+      "grad_norm": 0.6948829293251038,
+      "learning_rate": 2.0804020100502513e-05,
+      "loss": 0.3641,
+      "step": 1793
+    },
+    {
+      "epoch": 0.897,
+      "grad_norm": 0.7577362656593323,
+      "learning_rate": 2.07035175879397e-05,
+      "loss": 0.3981,
+      "step": 1794
+    },
+    {
+      "epoch": 0.8975,
+      "grad_norm": 0.7123954892158508,
+      "learning_rate": 2.0603015075376886e-05,
+      "loss": 0.4041,
+      "step": 1795
+    },
+    {
+      "epoch": 0.898,
+      "grad_norm": 0.8327853083610535,
+      "learning_rate": 2.0502512562814073e-05,
+      "loss": 0.3742,
+      "step": 1796
+    },
+    {
+      "epoch": 0.8985,
+      "grad_norm": 0.6596772074699402,
+      "learning_rate": 2.040201005025126e-05,
+      "loss": 0.3543,
+      "step": 1797
+    },
+    {
+      "epoch": 0.899,
+      "grad_norm": 0.677960216999054,
+      "learning_rate": 2.0301507537688446e-05,
+      "loss": 0.3578,
+      "step": 1798
+    },
+    {
+      "epoch": 0.8995,
+      "grad_norm": 0.7781803607940674,
+      "learning_rate": 2.020100502512563e-05,
+      "loss": 0.335,
+      "step": 1799
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6539471745491028,
+      "learning_rate": 2.0100502512562815e-05,
+      "loss": 0.3558,
+      "step": 1800
+    },
+    {
+      "epoch": 0.9005,
+      "grad_norm": 0.5422185063362122,
+      "learning_rate": 2e-05,
+      "loss": 0.5247,
+      "step": 1801
+    },
+    {
+      "epoch": 0.901,
+      "grad_norm": 0.6491250395774841,
+      "learning_rate": 1.9899497487437188e-05,
+      "loss": 0.5673,
+      "step": 1802
+    },
+    {
+      "epoch": 0.9015,
+      "grad_norm": 0.5200637578964233,
+      "learning_rate": 1.9798994974874375e-05,
+      "loss": 0.4511,
+      "step": 1803
+    },
+    {
+      "epoch": 0.902,
+      "grad_norm": 0.6333795189857483,
+      "learning_rate": 1.9698492462311558e-05,
+      "loss": 0.5226,
+      "step": 1804
+    },
+    {
+      "epoch": 0.9025,
+      "grad_norm": 0.660571277141571,
+      "learning_rate": 1.9597989949748744e-05,
+      "loss": 0.6071,
+      "step": 1805
+    },
+    {
+      "epoch": 0.903,
+      "grad_norm": 0.612684965133667,
+      "learning_rate": 1.949748743718593e-05,
+      "loss": 0.5114,
+      "step": 1806
+    },
+    {
+      "epoch": 0.9035,
+      "grad_norm": 0.5527110695838928,
+      "learning_rate": 1.9396984924623117e-05,
+      "loss": 0.4395,
+      "step": 1807
+    },
+    {
+      "epoch": 0.904,
+      "grad_norm": 0.606997549533844,
+      "learning_rate": 1.92964824120603e-05,
+      "loss": 0.4663,
+      "step": 1808
+    },
+    {
+      "epoch": 0.9045,
+      "grad_norm": 0.6626200079917908,
+      "learning_rate": 1.9195979899497487e-05,
+      "loss": 0.516,
+      "step": 1809
+    },
+    {
+      "epoch": 0.905,
+      "grad_norm": 0.5738974809646606,
+      "learning_rate": 1.9095477386934673e-05,
+      "loss": 0.3971,
+      "step": 1810
+    },
+    {
+      "epoch": 0.9055,
+      "grad_norm": 0.6293787956237793,
+      "learning_rate": 1.899497487437186e-05,
+      "loss": 0.4386,
+      "step": 1811
+    },
+    {
+      "epoch": 0.906,
+      "grad_norm": 0.6659508943557739,
+      "learning_rate": 1.8894472361809046e-05,
+      "loss": 0.4526,
+      "step": 1812
+    },
+    {
+      "epoch": 0.9065,
+      "grad_norm": 0.671139657497406,
+      "learning_rate": 1.8793969849246233e-05,
+      "loss": 0.511,
+      "step": 1813
+    },
+    {
+      "epoch": 0.907,
+      "grad_norm": 0.7288784980773926,
+      "learning_rate": 1.869346733668342e-05,
+      "loss": 0.4266,
+      "step": 1814
+    },
+    {
+      "epoch": 0.9075,
+      "grad_norm": 0.6416025161743164,
+      "learning_rate": 1.8592964824120602e-05,
+      "loss": 0.4007,
+      "step": 1815
+    },
+    {
+      "epoch": 0.908,
+      "grad_norm": 0.6718318462371826,
+      "learning_rate": 1.849246231155779e-05,
+      "loss": 0.4011,
+      "step": 1816
+    },
+    {
+      "epoch": 0.9085,
+      "grad_norm": 0.6760593056678772,
+      "learning_rate": 1.8391959798994975e-05,
+      "loss": 0.4514,
+      "step": 1817
+    },
+    {
+      "epoch": 0.909,
+      "grad_norm": 0.7032837271690369,
+      "learning_rate": 1.829145728643216e-05,
+      "loss": 0.4521,
+      "step": 1818
+    },
+    {
+      "epoch": 0.9095,
+      "grad_norm": 0.7102841734886169,
+      "learning_rate": 1.8190954773869348e-05,
+      "loss": 0.5345,
+      "step": 1819
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7079935669898987,
+      "learning_rate": 1.8090452261306535e-05,
+      "loss": 0.3763,
+      "step": 1820
+    },
+    {
+      "epoch": 0.9105,
+      "grad_norm": 0.5373201966285706,
+      "learning_rate": 1.798994974874372e-05,
+      "loss": 0.2976,
+      "step": 1821
+    },
+    {
+      "epoch": 0.911,
+      "grad_norm": 0.6126549243927002,
+      "learning_rate": 1.7889447236180908e-05,
+      "loss": 0.3429,
+      "step": 1822
+    },
+    {
+      "epoch": 0.9115,
+      "grad_norm": 0.5091574788093567,
+      "learning_rate": 1.778894472361809e-05,
+      "loss": 0.3014,
+      "step": 1823
+    },
+    {
+      "epoch": 0.912,
+      "grad_norm": 1.1894861459732056,
+      "learning_rate": 1.7688442211055277e-05,
+      "loss": 0.5561,
+      "step": 1824
+    },
+    {
+      "epoch": 0.9125,
+      "grad_norm": 0.703696608543396,
+      "learning_rate": 1.7587939698492464e-05,
+      "loss": 0.385,
+      "step": 1825
+    },
+    {
+      "epoch": 0.913,
+      "grad_norm": 0.688389778137207,
+      "learning_rate": 1.748743718592965e-05,
+      "loss": 0.3531,
+      "step": 1826
+    },
+    {
+      "epoch": 0.9135,
+      "grad_norm": 0.6290487051010132,
+      "learning_rate": 1.7386934673366833e-05,
+      "loss": 0.4108,
+      "step": 1827
+    },
+    {
+      "epoch": 0.914,
+      "grad_norm": 0.7750195860862732,
+      "learning_rate": 1.728643216080402e-05,
+      "loss": 0.3725,
+      "step": 1828
+    },
+    {
+      "epoch": 0.9145,
+      "grad_norm": 0.7465023398399353,
+      "learning_rate": 1.7185929648241206e-05,
+      "loss": 0.4541,
+      "step": 1829
+    },
+    {
+      "epoch": 0.915,
+      "grad_norm": 0.6750847101211548,
+      "learning_rate": 1.7085427135678393e-05,
+      "loss": 0.3853,
+      "step": 1830
+    },
+    {
+      "epoch": 0.9155,
+      "grad_norm": 0.6402559876441956,
+      "learning_rate": 1.698492462311558e-05,
+      "loss": 0.4236,
+      "step": 1831
+    },
+    {
+      "epoch": 0.916,
+      "grad_norm": 0.6501702666282654,
+      "learning_rate": 1.6884422110552762e-05,
+      "loss": 0.366,
+      "step": 1832
+    },
+    {
+      "epoch": 0.9165,
+      "grad_norm": 0.5888473987579346,
+      "learning_rate": 1.678391959798995e-05,
+      "loss": 0.3588,
+      "step": 1833
+    },
+    {
+      "epoch": 0.917,
+      "grad_norm": 0.5912230014801025,
+      "learning_rate": 1.6683417085427135e-05,
+      "loss": 0.3234,
+      "step": 1834
+    },
+    {
+      "epoch": 0.9175,
+      "grad_norm": 0.6947012543678284,
+      "learning_rate": 1.6582914572864322e-05,
+      "loss": 0.2833,
+      "step": 1835
+    },
+    {
+      "epoch": 0.918,
+      "grad_norm": 0.7913910746574402,
+      "learning_rate": 1.6482412060301508e-05,
+      "loss": 0.4497,
+      "step": 1836
+    },
+    {
+      "epoch": 0.9185,
+      "grad_norm": 0.8567656874656677,
+      "learning_rate": 1.6381909547738695e-05,
+      "loss": 0.4549,
+      "step": 1837
+    },
+    {
+      "epoch": 0.919,
+      "grad_norm": 0.5096683502197266,
+      "learning_rate": 1.628140703517588e-05,
+      "loss": 0.2712,
+      "step": 1838
+    },
+    {
+      "epoch": 0.9195,
+      "grad_norm": 0.6716606020927429,
+      "learning_rate": 1.6180904522613068e-05,
+      "loss": 0.428,
+      "step": 1839
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.579538881778717,
+      "learning_rate": 1.608040201005025e-05,
+      "loss": 0.2754,
+      "step": 1840
+    },
+    {
+      "epoch": 0.9205,
+      "grad_norm": 0.583088755607605,
+      "learning_rate": 1.5979899497487437e-05,
+      "loss": 0.2607,
+      "step": 1841
+    },
+    {
+      "epoch": 0.921,
+      "grad_norm": 0.6406164765357971,
+      "learning_rate": 1.5879396984924624e-05,
+      "loss": 0.3536,
+      "step": 1842
+    },
+    {
+      "epoch": 0.9215,
+      "grad_norm": 0.6029505133628845,
+      "learning_rate": 1.577889447236181e-05,
+      "loss": 0.3427,
+      "step": 1843
+    },
+    {
+      "epoch": 0.922,
+      "grad_norm": 0.6062425374984741,
+      "learning_rate": 1.5678391959798997e-05,
+      "loss": 0.3866,
+      "step": 1844
+    },
+    {
+      "epoch": 0.9225,
+      "grad_norm": 0.7379024028778076,
+      "learning_rate": 1.5577889447236183e-05,
+      "loss": 0.3361,
+      "step": 1845
+    },
+    {
+      "epoch": 0.923,
+      "grad_norm": 0.5891954302787781,
+      "learning_rate": 1.547738693467337e-05,
+      "loss": 0.2654,
+      "step": 1846
+    },
+    {
+      "epoch": 0.9235,
+      "grad_norm": 0.5837146043777466,
+      "learning_rate": 1.5376884422110556e-05,
+      "loss": 0.2554,
+      "step": 1847
+    },
+    {
+      "epoch": 0.924,
+      "grad_norm": 0.6203508973121643,
+      "learning_rate": 1.527638190954774e-05,
+      "loss": 0.2755,
+      "step": 1848
+    },
+    {
+      "epoch": 0.9245,
+      "grad_norm": 0.6655300855636597,
+      "learning_rate": 1.5175879396984927e-05,
+      "loss": 0.2975,
+      "step": 1849
+    },
+    {
+      "epoch": 0.925,
+      "grad_norm": 0.6980761885643005,
+      "learning_rate": 1.507537688442211e-05,
+      "loss": 0.3236,
+      "step": 1850
+    },
+    {
+      "epoch": 0.9255,
+      "grad_norm": 0.6375998854637146,
+      "learning_rate": 1.4974874371859295e-05,
+      "loss": 0.6517,
+      "step": 1851
+    },
+    {
+      "epoch": 0.926,
+      "grad_norm": 0.4773874282836914,
+      "learning_rate": 1.4874371859296482e-05,
+      "loss": 0.3724,
+      "step": 1852
+    },
+    {
+      "epoch": 0.9265,
+      "grad_norm": 0.5012878775596619,
+      "learning_rate": 1.4773869346733668e-05,
+      "loss": 0.2976,
+      "step": 1853
+    },
+    {
+      "epoch": 0.927,
+      "grad_norm": 0.5998973846435547,
+      "learning_rate": 1.4673366834170855e-05,
+      "loss": 0.3732,
+      "step": 1854
+    },
+    {
+      "epoch": 0.9275,
+      "grad_norm": 0.6029618978500366,
+      "learning_rate": 1.457286432160804e-05,
+      "loss": 0.5431,
+      "step": 1855
+    },
+    {
+      "epoch": 0.928,
+      "grad_norm": 0.63199383020401,
+      "learning_rate": 1.4472361809045226e-05,
+      "loss": 0.4427,
+      "step": 1856
+    },
+    {
+      "epoch": 0.9285,
+      "grad_norm": 0.6722904443740845,
+      "learning_rate": 1.4371859296482413e-05,
+      "loss": 0.4317,
+      "step": 1857
+    },
+    {
+      "epoch": 0.929,
+      "grad_norm": 0.616225004196167,
+      "learning_rate": 1.4271356783919599e-05,
+      "loss": 0.3822,
+      "step": 1858
+    },
+    {
+      "epoch": 0.9295,
+      "grad_norm": 0.8107300996780396,
+      "learning_rate": 1.4170854271356784e-05,
+      "loss": 0.6871,
+      "step": 1859
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6912587881088257,
+      "learning_rate": 1.407035175879397e-05,
+      "loss": 0.4726,
+      "step": 1860
+    },
+    {
+      "epoch": 0.9305,
+      "grad_norm": 0.769458532333374,
+      "learning_rate": 1.3969849246231157e-05,
+      "loss": 0.6493,
+      "step": 1861
+    },
+    {
+      "epoch": 0.931,
+      "grad_norm": 0.5853328108787537,
+      "learning_rate": 1.3869346733668343e-05,
+      "loss": 0.4456,
+      "step": 1862
+    },
+    {
+      "epoch": 0.9315,
+      "grad_norm": 0.5630549788475037,
+      "learning_rate": 1.3768844221105528e-05,
+      "loss": 0.3283,
+      "step": 1863
+    },
+    {
+      "epoch": 0.932,
+      "grad_norm": 0.6847948431968689,
+      "learning_rate": 1.3668341708542715e-05,
+      "loss": 0.4354,
+      "step": 1864
+    },
+    {
+      "epoch": 0.9325,
+      "grad_norm": 0.7098371982574463,
+      "learning_rate": 1.3567839195979901e-05,
+      "loss": 0.47,
+      "step": 1865
+    },
+    {
+      "epoch": 0.933,
+      "grad_norm": 0.7356335520744324,
+      "learning_rate": 1.3467336683417087e-05,
+      "loss": 0.5036,
+      "step": 1866
+    },
+    {
+      "epoch": 0.9335,
+      "grad_norm": 0.556207001209259,
+      "learning_rate": 1.3366834170854272e-05,
+      "loss": 0.3381,
+      "step": 1867
+    },
+    {
+      "epoch": 0.934,
+      "grad_norm": 0.6592797636985779,
+      "learning_rate": 1.3266331658291459e-05,
+      "loss": 0.4394,
+      "step": 1868
+    },
+    {
+      "epoch": 0.9345,
+      "grad_norm": 0.7090339660644531,
+      "learning_rate": 1.3165829145728645e-05,
+      "loss": 0.4159,
+      "step": 1869
+    },
+    {
+      "epoch": 0.935,
+      "grad_norm": 0.6776400804519653,
+      "learning_rate": 1.306532663316583e-05,
+      "loss": 0.4927,
+      "step": 1870
+    },
+    {
+      "epoch": 0.9355,
+      "grad_norm": 0.6419298648834229,
+      "learning_rate": 1.2964824120603017e-05,
+      "loss": 0.4192,
+      "step": 1871
+    },
+    {
+      "epoch": 0.936,
+      "grad_norm": 0.6618615388870239,
+      "learning_rate": 1.2864321608040203e-05,
+      "loss": 0.3946,
+      "step": 1872
+    },
+    {
+      "epoch": 0.9365,
+      "grad_norm": 0.8378037214279175,
+      "learning_rate": 1.2763819095477386e-05,
+      "loss": 0.6463,
+      "step": 1873
+    },
+    {
+      "epoch": 0.937,
+      "grad_norm": 0.7007853388786316,
+      "learning_rate": 1.2663316582914573e-05,
+      "loss": 0.4827,
+      "step": 1874
+    },
+    {
+      "epoch": 0.9375,
+      "grad_norm": 0.6386650204658508,
+      "learning_rate": 1.2562814070351759e-05,
+      "loss": 0.3523,
+      "step": 1875
+    },
+    {
+      "epoch": 0.938,
+      "grad_norm": 0.7903352379798889,
+      "learning_rate": 1.2462311557788946e-05,
+      "loss": 0.5302,
+      "step": 1876
+    },
+    {
+      "epoch": 0.9385,
+      "grad_norm": 0.6161884665489197,
+      "learning_rate": 1.2361809045226132e-05,
+      "loss": 0.3644,
+      "step": 1877
+    },
+    {
+      "epoch": 0.939,
+      "grad_norm": 0.6710797548294067,
+      "learning_rate": 1.2261306532663318e-05,
+      "loss": 0.4377,
+      "step": 1878
+    },
+    {
+      "epoch": 0.9395,
+      "grad_norm": 0.754216730594635,
+      "learning_rate": 1.2160804020100503e-05,
+      "loss": 0.4504,
+      "step": 1879
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.68315589427948,
+      "learning_rate": 1.2060301507537688e-05,
+      "loss": 0.4021,
+      "step": 1880
+    },
+    {
+      "epoch": 0.9405,
+      "grad_norm": 0.6273193359375,
+      "learning_rate": 1.1959798994974875e-05,
+      "loss": 0.3841,
+      "step": 1881
+    },
+    {
+      "epoch": 0.941,
+      "grad_norm": 0.7164735198020935,
+      "learning_rate": 1.1859296482412061e-05,
+      "loss": 0.3729,
+      "step": 1882
+    },
+    {
+      "epoch": 0.9415,
+      "grad_norm": 0.9396920800209045,
+      "learning_rate": 1.1758793969849246e-05,
+      "loss": 0.3661,
+      "step": 1883
+    },
+    {
+      "epoch": 0.942,
+      "grad_norm": 0.7597629427909851,
+      "learning_rate": 1.1658291457286432e-05,
+      "loss": 0.4472,
+      "step": 1884
+    },
+    {
+      "epoch": 0.9425,
+      "grad_norm": 0.6072019338607788,
+      "learning_rate": 1.1557788944723619e-05,
+      "loss": 0.3421,
+      "step": 1885
+    },
+    {
+      "epoch": 0.943,
+      "grad_norm": 0.5514841079711914,
+      "learning_rate": 1.1457286432160805e-05,
+      "loss": 0.27,
+      "step": 1886
+    },
+    {
+      "epoch": 0.9435,
+      "grad_norm": 0.8010035753250122,
+      "learning_rate": 1.135678391959799e-05,
+      "loss": 0.4521,
+      "step": 1887
+    },
+    {
+      "epoch": 0.944,
+      "grad_norm": 0.7580119967460632,
+      "learning_rate": 1.1256281407035177e-05,
+      "loss": 0.5071,
+      "step": 1888
+    },
+    {
+      "epoch": 0.9445,
+      "grad_norm": 0.6949161887168884,
+      "learning_rate": 1.1155778894472363e-05,
+      "loss": 0.4048,
+      "step": 1889
+    },
+    {
+      "epoch": 0.945,
+      "grad_norm": 0.5998812317848206,
+      "learning_rate": 1.105527638190955e-05,
+      "loss": 0.3079,
+      "step": 1890
+    },
+    {
+      "epoch": 0.9455,
+      "grad_norm": 0.6919785737991333,
+      "learning_rate": 1.0954773869346734e-05,
+      "loss": 0.3748,
+      "step": 1891
+    },
+    {
+      "epoch": 0.946,
+      "grad_norm": 0.6881744861602783,
+      "learning_rate": 1.0854271356783919e-05,
+      "loss": 0.3891,
+      "step": 1892
+    },
+    {
+      "epoch": 0.9465,
+      "grad_norm": 0.6427577137947083,
+      "learning_rate": 1.0753768844221106e-05,
+      "loss": 0.3481,
+      "step": 1893
+    },
+    {
+      "epoch": 0.947,
+      "grad_norm": 0.6709974408149719,
+      "learning_rate": 1.0653266331658292e-05,
+      "loss": 0.3747,
+      "step": 1894
+    },
+    {
+      "epoch": 0.9475,
+      "grad_norm": 0.6137408018112183,
+      "learning_rate": 1.0552763819095479e-05,
+      "loss": 0.2918,
+      "step": 1895
+    },
+    {
+      "epoch": 0.948,
+      "grad_norm": 0.6262741088867188,
+      "learning_rate": 1.0452261306532663e-05,
+      "loss": 0.3704,
+      "step": 1896
+    },
+    {
+      "epoch": 0.9485,
+      "grad_norm": 0.6251107454299927,
+      "learning_rate": 1.035175879396985e-05,
+      "loss": 0.2891,
+      "step": 1897
+    },
+    {
+      "epoch": 0.949,
+      "grad_norm": 0.6505588889122009,
+      "learning_rate": 1.0251256281407036e-05,
+      "loss": 0.3033,
+      "step": 1898
+    },
+    {
+      "epoch": 0.9495,
+      "grad_norm": 0.7799379825592041,
+      "learning_rate": 1.0150753768844223e-05,
+      "loss": 0.3041,
+      "step": 1899
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.4991627633571625,
+      "learning_rate": 1.0050251256281408e-05,
+      "loss": 0.2449,
+      "step": 1900
+    },
+    {
+      "epoch": 0.9505,
+      "grad_norm": 0.5111928582191467,
+      "learning_rate": 9.949748743718594e-06,
+      "loss": 0.3913,
+      "step": 1901
+    },
+    {
+      "epoch": 0.951,
+      "grad_norm": 0.5456981062889099,
+      "learning_rate": 9.849246231155779e-06,
+      "loss": 0.4758,
+      "step": 1902
+    },
+    {
+      "epoch": 0.9515,
+      "grad_norm": 0.5930306315422058,
+      "learning_rate": 9.748743718592965e-06,
+      "loss": 0.4378,
+      "step": 1903
+    },
+    {
+      "epoch": 0.952,
+      "grad_norm": 0.5851122140884399,
+      "learning_rate": 9.64824120603015e-06,
+      "loss": 0.4086,
+      "step": 1904
+    },
+    {
+      "epoch": 0.9525,
+      "grad_norm": 0.6449226140975952,
+      "learning_rate": 9.547738693467337e-06,
+      "loss": 0.4598,
+      "step": 1905
+    },
+    {
+      "epoch": 0.953,
+      "grad_norm": 0.5486339330673218,
+      "learning_rate": 9.447236180904523e-06,
+      "loss": 0.4781,
+      "step": 1906
+    },
+    {
+      "epoch": 0.9535,
+      "grad_norm": 0.5541635751724243,
+      "learning_rate": 9.34673366834171e-06,
+      "loss": 0.4218,
+      "step": 1907
+    },
+    {
+      "epoch": 0.954,
+      "grad_norm": 0.7386255860328674,
+      "learning_rate": 9.246231155778894e-06,
+      "loss": 0.5639,
+      "step": 1908
+    },
+    {
+      "epoch": 0.9545,
+      "grad_norm": 0.5018959045410156,
+      "learning_rate": 9.14572864321608e-06,
+      "loss": 0.3113,
+      "step": 1909
+    },
+    {
+      "epoch": 0.955,
+      "grad_norm": 0.7117149233818054,
+      "learning_rate": 9.045226130653267e-06,
+      "loss": 0.5029,
+      "step": 1910
+    },
+    {
+      "epoch": 0.9555,
+      "grad_norm": 0.5913116335868835,
+      "learning_rate": 8.944723618090454e-06,
+      "loss": 0.3828,
+      "step": 1911
+    },
+    {
+      "epoch": 0.956,
+      "grad_norm": 0.5680027008056641,
+      "learning_rate": 8.844221105527639e-06,
+      "loss": 0.4075,
+      "step": 1912
+    },
+    {
+      "epoch": 0.9565,
+      "grad_norm": 0.638623833656311,
+      "learning_rate": 8.743718592964825e-06,
+      "loss": 0.3964,
+      "step": 1913
+    },
+    {
+      "epoch": 0.957,
+      "grad_norm": 0.6504754424095154,
+      "learning_rate": 8.64321608040201e-06,
+      "loss": 0.4906,
+      "step": 1914
+    },
+    {
+      "epoch": 0.9575,
+      "grad_norm": 0.67131507396698,
+      "learning_rate": 8.542713567839196e-06,
+      "loss": 0.4042,
+      "step": 1915
+    },
+    {
+      "epoch": 0.958,
+      "grad_norm": 0.6521419286727905,
+      "learning_rate": 8.442211055276381e-06,
+      "loss": 0.3655,
+      "step": 1916
+    },
+    {
+      "epoch": 0.9585,
+      "grad_norm": 0.5560711622238159,
+      "learning_rate": 8.341708542713568e-06,
+      "loss": 0.371,
+      "step": 1917
+    },
+    {
+      "epoch": 0.959,
+      "grad_norm": 0.6484144330024719,
+      "learning_rate": 8.241206030150754e-06,
+      "loss": 0.4587,
+      "step": 1918
+    },
+    {
+      "epoch": 0.9595,
+      "grad_norm": 0.683219850063324,
+      "learning_rate": 8.14070351758794e-06,
+      "loss": 0.4604,
+      "step": 1919
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5811185240745544,
+      "learning_rate": 8.040201005025125e-06,
+      "loss": 0.3766,
+      "step": 1920
+    },
+    {
+      "epoch": 0.9605,
+      "grad_norm": 0.5768767595291138,
+      "learning_rate": 7.939698492462312e-06,
+      "loss": 0.3182,
+      "step": 1921
+    },
+    {
+      "epoch": 0.961,
+      "grad_norm": 0.7092142105102539,
+      "learning_rate": 7.839195979899498e-06,
+      "loss": 0.451,
+      "step": 1922
+    },
+    {
+      "epoch": 0.9615,
+      "grad_norm": 0.5659924745559692,
+      "learning_rate": 7.738693467336685e-06,
+      "loss": 0.3128,
+      "step": 1923
+    },
+    {
+      "epoch": 0.962,
+      "grad_norm": 0.5894472599029541,
+      "learning_rate": 7.63819095477387e-06,
+      "loss": 0.3441,
+      "step": 1924
+    },
+    {
+      "epoch": 0.9625,
+      "grad_norm": 0.6528652906417847,
+      "learning_rate": 7.537688442211055e-06,
+      "loss": 0.4354,
+      "step": 1925
+    },
+    {
+      "epoch": 0.963,
+      "grad_norm": 0.7244471311569214,
+      "learning_rate": 7.437185929648241e-06,
+      "loss": 0.5047,
+      "step": 1926
+    },
+    {
+      "epoch": 0.9635,
+      "grad_norm": 0.8269768953323364,
+      "learning_rate": 7.336683417085427e-06,
+      "loss": 0.4364,
+      "step": 1927
+    },
+    {
+      "epoch": 0.964,
+      "grad_norm": 0.7433727979660034,
+      "learning_rate": 7.236180904522613e-06,
+      "loss": 0.466,
+      "step": 1928
+    },
+    {
+      "epoch": 0.9645,
+      "grad_norm": 0.6624360680580139,
+      "learning_rate": 7.1356783919597995e-06,
+      "loss": 0.3748,
+      "step": 1929
+    },
+    {
+      "epoch": 0.965,
+      "grad_norm": 0.6394107937812805,
+      "learning_rate": 7.035175879396985e-06,
+      "loss": 0.3761,
+      "step": 1930
+    },
+    {
+      "epoch": 0.9655,
+      "grad_norm": 0.6123507022857666,
+      "learning_rate": 6.934673366834172e-06,
+      "loss": 0.317,
+      "step": 1931
+    },
+    {
+      "epoch": 0.966,
+      "grad_norm": 0.7256487607955933,
+      "learning_rate": 6.834170854271357e-06,
+      "loss": 0.51,
+      "step": 1932
+    },
+    {
+      "epoch": 0.9665,
+      "grad_norm": 0.624494731426239,
+      "learning_rate": 6.733668341708544e-06,
+      "loss": 0.3712,
+      "step": 1933
+    },
+    {
+      "epoch": 0.967,
+      "grad_norm": 0.5863220691680908,
+      "learning_rate": 6.633165829145729e-06,
+      "loss": 0.3121,
+      "step": 1934
+    },
+    {
+      "epoch": 0.9675,
+      "grad_norm": 0.7814472913742065,
+      "learning_rate": 6.532663316582915e-06,
+      "loss": 0.42,
+      "step": 1935
+    },
+    {
+      "epoch": 0.968,
+      "grad_norm": 0.7488368153572083,
+      "learning_rate": 6.4321608040201015e-06,
+      "loss": 0.4493,
+      "step": 1936
+    },
+    {
+      "epoch": 0.9685,
+      "grad_norm": 0.5412418842315674,
+      "learning_rate": 6.331658291457286e-06,
+      "loss": 0.2907,
+      "step": 1937
+    },
+    {
+      "epoch": 0.969,
+      "grad_norm": 0.6010940670967102,
+      "learning_rate": 6.231155778894473e-06,
+      "loss": 0.3504,
+      "step": 1938
+    },
+    {
+      "epoch": 0.9695,
+      "grad_norm": 0.7745607495307922,
+      "learning_rate": 6.130653266331659e-06,
+      "loss": 0.3468,
+      "step": 1939
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6169768571853638,
+      "learning_rate": 6.030150753768844e-06,
+      "loss": 0.3231,
+      "step": 1940
+    },
+    {
+      "epoch": 0.9705,
+      "grad_norm": 0.6768513321876526,
+      "learning_rate": 5.9296482412060305e-06,
+      "loss": 0.365,
+      "step": 1941
+    },
+    {
+      "epoch": 0.971,
+      "grad_norm": 0.7564877271652222,
+      "learning_rate": 5.829145728643216e-06,
+      "loss": 0.4096,
+      "step": 1942
+    },
+    {
+      "epoch": 0.9715,
+      "grad_norm": 0.5392837524414062,
+      "learning_rate": 5.728643216080403e-06,
+      "loss": 0.2562,
+      "step": 1943
+    },
+    {
+      "epoch": 0.972,
+      "grad_norm": 0.6356944441795349,
+      "learning_rate": 5.628140703517588e-06,
+      "loss": 0.3145,
+      "step": 1944
+    },
+    {
+      "epoch": 0.9725,
+      "grad_norm": 0.6749258041381836,
+      "learning_rate": 5.527638190954775e-06,
+      "loss": 0.3627,
+      "step": 1945
+    },
+    {
+      "epoch": 0.973,
+      "grad_norm": 0.6194499135017395,
+      "learning_rate": 5.4271356783919595e-06,
+      "loss": 0.2775,
+      "step": 1946
+    },
+    {
+      "epoch": 0.9735,
+      "grad_norm": 0.8884201049804688,
+      "learning_rate": 5.326633165829146e-06,
+      "loss": 0.4044,
+      "step": 1947
+    },
+    {
+      "epoch": 0.974,
+      "grad_norm": 0.5936954617500305,
+      "learning_rate": 5.226130653266332e-06,
+      "loss": 0.284,
+      "step": 1948
+    },
+    {
+      "epoch": 0.9745,
+      "grad_norm": 0.7279488444328308,
+      "learning_rate": 5.125628140703518e-06,
+      "loss": 0.3186,
+      "step": 1949
+    },
+    {
+      "epoch": 0.975,
+      "grad_norm": 0.6850216388702393,
+      "learning_rate": 5.025125628140704e-06,
+      "loss": 0.3216,
+      "step": 1950
+    },
+    {
+      "epoch": 0.9755,
+      "grad_norm": 0.6496731638908386,
+      "learning_rate": 4.9246231155778894e-06,
+      "loss": 0.5885,
+      "step": 1951
+    },
+    {
+      "epoch": 0.976,
+      "grad_norm": 0.589713454246521,
+      "learning_rate": 4.824120603015075e-06,
+      "loss": 0.3863,
+      "step": 1952
+    },
+    {
+      "epoch": 0.9765,
+      "grad_norm": 0.554733395576477,
+      "learning_rate": 4.7236180904522615e-06,
+      "loss": 0.4463,
+      "step": 1953
+    },
+    {
+      "epoch": 0.977,
+      "grad_norm": 0.5910104513168335,
+      "learning_rate": 4.623115577889447e-06,
+      "loss": 0.4756,
+      "step": 1954
+    },
+    {
+      "epoch": 0.9775,
+      "grad_norm": 0.645351767539978,
+      "learning_rate": 4.522613065326634e-06,
+      "loss": 0.4705,
+      "step": 1955
+    },
+    {
+      "epoch": 0.978,
+      "grad_norm": 0.5496158003807068,
+      "learning_rate": 4.422110552763819e-06,
+      "loss": 0.3909,
+      "step": 1956
+    },
+    {
+      "epoch": 0.9785,
+      "grad_norm": 0.5388986468315125,
+      "learning_rate": 4.321608040201005e-06,
+      "loss": 0.4269,
+      "step": 1957
+    },
+    {
+      "epoch": 0.979,
+      "grad_norm": 0.6822681427001953,
+      "learning_rate": 4.2211055276381906e-06,
+      "loss": 0.7039,
+      "step": 1958
+    },
+    {
+      "epoch": 0.9795,
+      "grad_norm": 0.6257898807525635,
+      "learning_rate": 4.120603015075377e-06,
+      "loss": 0.4763,
+      "step": 1959
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6473307013511658,
+      "learning_rate": 4.020100502512563e-06,
+      "loss": 0.4179,
+      "step": 1960
+    },
+    {
+      "epoch": 0.9805,
+      "grad_norm": 0.683695912361145,
+      "learning_rate": 3.919597989949749e-06,
+      "loss": 0.5051,
+      "step": 1961
+    },
+    {
+      "epoch": 0.981,
+      "grad_norm": 0.6073151230812073,
+      "learning_rate": 3.819095477386935e-06,
+      "loss": 0.395,
+      "step": 1962
+    },
+    {
+      "epoch": 0.9815,
+      "grad_norm": 0.604476809501648,
+      "learning_rate": 3.7185929648241204e-06,
+      "loss": 0.3757,
+      "step": 1963
+    },
+    {
+      "epoch": 0.982,
+      "grad_norm": 0.6297847032546997,
+      "learning_rate": 3.6180904522613065e-06,
+      "loss": 0.4872,
+      "step": 1964
+    },
+    {
+      "epoch": 0.9825,
+      "grad_norm": 0.6326467990875244,
+      "learning_rate": 3.5175879396984926e-06,
+      "loss": 0.3618,
+      "step": 1965
+    },
+    {
+      "epoch": 0.983,
+      "grad_norm": 0.6912254095077515,
+      "learning_rate": 3.4170854271356786e-06,
+      "loss": 0.4817,
+      "step": 1966
+    },
+    {
+      "epoch": 0.9835,
+      "grad_norm": 0.6747575998306274,
+      "learning_rate": 3.3165829145728647e-06,
+      "loss": 0.4248,
+      "step": 1967
+    },
+    {
+      "epoch": 0.984,
+      "grad_norm": 0.6581198573112488,
+      "learning_rate": 3.2160804020100507e-06,
+      "loss": 0.4989,
+      "step": 1968
+    },
+    {
+      "epoch": 0.9845,
+      "grad_norm": 0.6973407864570618,
+      "learning_rate": 3.1155778894472364e-06,
+      "loss": 0.417,
+      "step": 1969
+    },
+    {
+      "epoch": 0.985,
+      "grad_norm": 0.5891755819320679,
+      "learning_rate": 3.015075376884422e-06,
+      "loss": 0.3256,
+      "step": 1970
+    },
+    {
+      "epoch": 0.9855,
+      "grad_norm": 0.6015627980232239,
+      "learning_rate": 2.914572864321608e-06,
+      "loss": 0.4124,
+      "step": 1971
+    },
+    {
+      "epoch": 0.986,
+      "grad_norm": 0.6696973443031311,
+      "learning_rate": 2.814070351758794e-06,
+      "loss": 0.3954,
+      "step": 1972
+    },
+    {
+      "epoch": 0.9865,
+      "grad_norm": 0.8548389077186584,
+      "learning_rate": 2.7135678391959798e-06,
+      "loss": 0.5053,
+      "step": 1973
+    },
+    {
+      "epoch": 0.987,
+      "grad_norm": 0.9650516510009766,
+      "learning_rate": 2.613065326633166e-06,
+      "loss": 0.5065,
+      "step": 1974
+    },
+    {
+      "epoch": 0.9875,
+      "grad_norm": 0.7417435646057129,
+      "learning_rate": 2.512562814070352e-06,
+      "loss": 0.5381,
+      "step": 1975
+    },
+    {
+      "epoch": 0.988,
+      "grad_norm": 0.6715131402015686,
+      "learning_rate": 2.4120603015075375e-06,
+      "loss": 0.3484,
+      "step": 1976
+    },
+    {
+      "epoch": 0.9885,
+      "grad_norm": 0.7092119455337524,
+      "learning_rate": 2.3115577889447236e-06,
+      "loss": 0.4511,
+      "step": 1977
+    },
+    {
+      "epoch": 0.989,
+      "grad_norm": 0.6347234845161438,
+      "learning_rate": 2.2110552763819096e-06,
+      "loss": 0.4059,
+      "step": 1978
+    },
+    {
+      "epoch": 0.9895,
+      "grad_norm": 0.703213632106781,
+      "learning_rate": 2.1105527638190953e-06,
+      "loss": 0.4079,
+      "step": 1979
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6775699853897095,
+      "learning_rate": 2.0100502512562813e-06,
+      "loss": 0.4594,
+      "step": 1980
+    },
+    {
+      "epoch": 0.9905,
+      "grad_norm": 0.5209073424339294,
+      "learning_rate": 1.9095477386934674e-06,
+      "loss": 0.2439,
+      "step": 1981
+    },
+    {
+      "epoch": 0.991,
+      "grad_norm": 0.6620062589645386,
+      "learning_rate": 1.8090452261306533e-06,
+      "loss": 0.4323,
+      "step": 1982
+    },
+    {
+      "epoch": 0.9915,
+      "grad_norm": 0.6307405233383179,
+      "learning_rate": 1.7085427135678393e-06,
+      "loss": 0.3855,
+      "step": 1983
+    },
+    {
+      "epoch": 0.992,
+      "grad_norm": 0.6738427877426147,
+      "learning_rate": 1.6080402010050254e-06,
+      "loss": 0.3884,
+      "step": 1984
+    },
+    {
+      "epoch": 0.9925,
+      "grad_norm": 0.7500790357589722,
+      "learning_rate": 1.507537688442211e-06,
+      "loss": 0.4589,
+      "step": 1985
+    },
+    {
+      "epoch": 0.993,
+      "grad_norm": 0.6440097093582153,
+      "learning_rate": 1.407035175879397e-06,
+      "loss": 0.3745,
+      "step": 1986
+    },
+    {
+      "epoch": 0.9935,
+      "grad_norm": 0.6714029908180237,
+      "learning_rate": 1.306532663316583e-06,
+      "loss": 0.4175,
+      "step": 1987
+    },
+    {
+      "epoch": 0.994,
+      "grad_norm": 0.6461104154586792,
+      "learning_rate": 1.2060301507537688e-06,
+      "loss": 0.3573,
+      "step": 1988
+    },
+    {
+      "epoch": 0.9945,
+      "grad_norm": 0.6324477195739746,
+      "learning_rate": 1.1055276381909548e-06,
+      "loss": 0.3312,
+      "step": 1989
+    },
+    {
+      "epoch": 0.995,
+      "grad_norm": 0.811026930809021,
+      "learning_rate": 1.0050251256281407e-06,
+      "loss": 0.4623,
+      "step": 1990
+    },
+    {
+      "epoch": 0.9955,
+      "grad_norm": 0.5602400898933411,
+      "learning_rate": 9.045226130653266e-07,
+      "loss": 0.2947,
+      "step": 1991
+    },
+    {
+      "epoch": 0.996,
+      "grad_norm": 0.5518646240234375,
+      "learning_rate": 8.040201005025127e-07,
+      "loss": 0.3501,
+      "step": 1992
+    },
+    {
+      "epoch": 0.9965,
+      "grad_norm": 0.5984534025192261,
+      "learning_rate": 7.035175879396985e-07,
+      "loss": 0.3294,
+      "step": 1993
+    },
+    {
+      "epoch": 0.997,
+      "grad_norm": 0.6465786695480347,
+      "learning_rate": 6.030150753768844e-07,
+      "loss": 0.3384,
+      "step": 1994
+    },
+    {
+      "epoch": 0.9975,
+      "grad_norm": 0.688086986541748,
+      "learning_rate": 5.025125628140703e-07,
+      "loss": 0.3244,
+      "step": 1995
+    },
+    {
+      "epoch": 0.998,
+      "grad_norm": 0.8641793131828308,
+      "learning_rate": 4.0201005025125634e-07,
+      "loss": 0.4195,
+      "step": 1996
+    },
+    {
+      "epoch": 0.9985,
+      "grad_norm": 0.7611424922943115,
+      "learning_rate": 3.015075376884422e-07,
+      "loss": 0.3944,
+      "step": 1997
+    },
+    {
+      "epoch": 0.999,
+      "grad_norm": 0.7514485716819763,
+      "learning_rate": 2.0100502512562817e-07,
+      "loss": 0.4138,
+      "step": 1998
+    },
+    {
+      "epoch": 0.9995,
+      "grad_norm": 0.6244973540306091,
+      "learning_rate": 1.0050251256281409e-07,
+      "loss": 0.2933,
+      "step": 1999
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.4967952370643616,
+      "learning_rate": 0.0,
+      "loss": 0.2244,
+      "step": 2000
+    },
+    {
+      "epoch": 1.0,
+      "eval_loss": 0.4038906991481781,
+      "eval_runtime": 108.5268,
+      "eval_samples_per_second": 4.607,
+      "eval_steps_per_second": 4.607,
+      "step": 2000
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 2000,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.604399166532608e+16,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}