diff --git "a/checkpoint-1092/trainer_state.json" "b/checkpoint-1092/trainer_state.json"
deleted file mode 100644--- "a/checkpoint-1092/trainer_state.json"
+++ /dev/null
@@ -1,7781 +0,0 @@
-{
-  "best_metric": null,
-  "best_model_checkpoint": null,
-  "epoch": 13.0,
-  "eval_steps": 500,
-  "global_step": 1092,
-  "is_hyper_param_search": false,
-  "is_local_process_zero": true,
-  "is_world_process_zero": true,
-  "log_history": [
-    {
-      "epoch": 0.011904761904761904,
-      "grad_norm": 17.036167336296355,
-      "learning_rate": 3.921568627450981e-07,
-      "loss": 9.2899,
-      "step": 1
-    },
-    {
-      "epoch": 0.023809523809523808,
-      "grad_norm": 16.51040791982049,
-      "learning_rate": 7.843137254901962e-07,
-      "loss": 9.1432,
-      "step": 2
-    },
-    {
-      "epoch": 0.03571428571428571,
-      "grad_norm": 17.409029658709134,
-      "learning_rate": 1.1764705882352942e-06,
-      "loss": 9.5443,
-      "step": 3
-    },
-    {
-      "epoch": 0.047619047619047616,
-      "grad_norm": 13.165316535403809,
-      "learning_rate": 1.5686274509803923e-06,
-      "loss": 7.9008,
-      "step": 4
-    },
-    {
-      "epoch": 0.05952380952380952,
-      "grad_norm": 13.598211666391473,
-      "learning_rate": 1.96078431372549e-06,
-      "loss": 7.9033,
-      "step": 5
-    },
-    {
-      "epoch": 0.07142857142857142,
-      "grad_norm": 6.219902614589629,
-      "learning_rate": 2.3529411764705885e-06,
-      "loss": 4.9995,
-      "step": 6
-    },
-    {
-      "epoch": 0.08333333333333333,
-      "grad_norm": 9.734044836676956,
-      "learning_rate": 2.7450980392156867e-06,
-      "loss": 6.5143,
-      "step": 7
-    },
-    {
-      "epoch": 0.09523809523809523,
-      "grad_norm": 16.731742931555345,
-      "learning_rate": 3.1372549019607846e-06,
-      "loss": 9.0697,
-      "step": 8
-    },
-    {
-      "epoch": 0.10714285714285714,
-      "grad_norm": 13.231944799550764,
-      "learning_rate": 3.529411764705883e-06,
-      "loss": 7.8908,
-      "step": 9
-    },
-    {
-      "epoch": 0.11904761904761904,
-      "grad_norm": 9.617661005903635,
-      "learning_rate": 3.92156862745098e-06,
-      "loss": 6.4021,
-      "step": 10
-    },
-    {
-      "epoch": 0.13095238095238096,
-      "grad_norm": 13.333162661233034,
-      "learning_rate": 4.313725490196079e-06,
-      "loss": 7.8622,
-      "step": 11
-    },
-    {
-      "epoch": 0.14285714285714285,
-      "grad_norm": 10.173955662246277,
-      "learning_rate": 4.705882352941177e-06,
-      "loss": 6.4799,
-      "step": 12
-    },
-    {
-      "epoch": 0.15476190476190477,
-      "grad_norm": 13.53421872626686,
-      "learning_rate": 5.098039215686274e-06,
-      "loss": 7.615,
-      "step": 13
-    },
-    {
-      "epoch": 0.16666666666666666,
-      "grad_norm": 13.528091158193236,
-      "learning_rate": 5.4901960784313735e-06,
-      "loss": 7.8859,
-      "step": 14
-    },
-    {
-      "epoch": 0.17857142857142858,
-      "grad_norm": 13.46530715470059,
-      "learning_rate": 5.882352941176471e-06,
-      "loss": 7.7087,
-      "step": 15
-    },
-    {
-      "epoch": 0.19047619047619047,
-      "grad_norm": 13.186854066383047,
-      "learning_rate": 6.274509803921569e-06,
-      "loss": 7.7375,
-      "step": 16
-    },
-    {
-      "epoch": 0.20238095238095238,
-      "grad_norm": 13.709261078338196,
-      "learning_rate": 6.666666666666667e-06,
-      "loss": 7.7001,
-      "step": 17
-    },
-    {
-      "epoch": 0.21428571428571427,
-      "grad_norm": 13.694779123750742,
-      "learning_rate": 7.058823529411766e-06,
-      "loss": 7.5827,
-      "step": 18
-    },
-    {
-      "epoch": 0.2261904761904762,
-      "grad_norm": 16.830863696016568,
-      "learning_rate": 7.450980392156863e-06,
-      "loss": 8.9601,
-      "step": 19
-    },
-    {
-      "epoch": 0.23809523809523808,
-      "grad_norm": 17.22287672142912,
-      "learning_rate": 7.84313725490196e-06,
-      "loss": 8.9006,
-      "step": 20
-    },
-    {
-      "epoch": 0.25,
-      "grad_norm": 17.92814388484799,
-      "learning_rate": 8.23529411764706e-06,
-      "loss": 9.1349,
-      "step": 21
-    },
-    {
-      "epoch": 0.2619047619047619,
-      "grad_norm": 17.193339613649197,
-      "learning_rate": 8.627450980392157e-06,
-      "loss": 8.7754,
-      "step": 22
-    },
-    {
-      "epoch": 0.27380952380952384,
-      "grad_norm": 17.286206959133764,
-      "learning_rate": 9.019607843137256e-06,
-      "loss": 8.7204,
-      "step": 23
-    },
-    {
-      "epoch": 0.2857142857142857,
-      "grad_norm": 10.070105901874978,
-      "learning_rate": 9.411764705882354e-06,
-      "loss": 6.1837,
-      "step": 24
-    },
-    {
-      "epoch": 0.2976190476190476,
-      "grad_norm": 13.810184661186407,
-      "learning_rate": 9.803921568627451e-06,
-      "loss": 7.4576,
-      "step": 25
-    },
-    {
-      "epoch": 0.30952380952380953,
-      "grad_norm": 6.7625117928823775,
-      "learning_rate": 1.0196078431372549e-05,
-      "loss": 4.806,
-      "step": 26
-    },
-    {
-      "epoch": 0.32142857142857145,
-      "grad_norm": 13.956198625918494,
-      "learning_rate": 1.0588235294117648e-05,
-      "loss": 7.0487,
-      "step": 27
-    },
-    {
-      "epoch": 0.3333333333333333,
-      "grad_norm": 10.260558995935384,
-      "learning_rate": 1.0980392156862747e-05,
-      "loss": 5.6256,
-      "step": 28
-    },
-    {
-      "epoch": 0.34523809523809523,
-      "grad_norm": 17.35991907597269,
-      "learning_rate": 1.1372549019607844e-05,
-      "loss": 8.0574,
-      "step": 29
-    },
-    {
-      "epoch": 0.35714285714285715,
-      "grad_norm": 9.721422858798402,
-      "learning_rate": 1.1764705882352942e-05,
-      "loss": 5.5548,
-      "step": 30
-    },
-    {
-      "epoch": 0.36904761904761907,
-      "grad_norm": 13.840485577854023,
-      "learning_rate": 1.215686274509804e-05,
-      "loss": 6.7663,
-      "step": 31
-    },
-    {
-      "epoch": 0.38095238095238093,
-      "grad_norm": 10.118512030653148,
-      "learning_rate": 1.2549019607843138e-05,
-      "loss": 5.543,
-      "step": 32
-    },
-    {
-      "epoch": 0.39285714285714285,
-      "grad_norm": 14.32625990673996,
-      "learning_rate": 1.2941176470588238e-05,
-      "loss": 6.2332,
-      "step": 33
-    },
-    {
-      "epoch": 0.40476190476190477,
-      "grad_norm": 10.011793459979055,
-      "learning_rate": 1.3333333333333333e-05,
-      "loss": 5.3651,
-      "step": 34
-    },
-    {
-      "epoch": 0.4166666666666667,
-      "grad_norm": 10.123714137969845,
-      "learning_rate": 1.3725490196078432e-05,
-      "loss": 5.3173,
-      "step": 35
-    },
-    {
-      "epoch": 0.42857142857142855,
-      "grad_norm": 10.251943500911867,
-      "learning_rate": 1.4117647058823532e-05,
-      "loss": 4.9469,
-      "step": 36
-    },
-    {
-      "epoch": 0.44047619047619047,
-      "grad_norm": 15.220443465155252,
-      "learning_rate": 1.4509803921568629e-05,
-      "loss": 5.5487,
-      "step": 37
-    },
-    {
-      "epoch": 0.4523809523809524,
-      "grad_norm": 10.713358439711659,
-      "learning_rate": 1.4901960784313726e-05,
-      "loss": 4.8903,
-      "step": 38
-    },
-    {
-      "epoch": 0.4642857142857143,
-      "grad_norm": 10.755404929616764,
-      "learning_rate": 1.5294117647058822e-05,
-      "loss": 4.8086,
-      "step": 39
-    },
-    {
-      "epoch": 0.47619047619047616,
-      "grad_norm": 11.89621232727623,
-      "learning_rate": 1.568627450980392e-05,
-      "loss": 4.5261,
-      "step": 40
-    },
-    {
-      "epoch": 0.4880952380952381,
-      "grad_norm": 7.524808327717391,
-      "learning_rate": 1.607843137254902e-05,
-      "loss": 3.7146,
-      "step": 41
-    },
-    {
-      "epoch": 0.5,
-      "grad_norm": 12.838155901645372,
-      "learning_rate": 1.647058823529412e-05,
-      "loss": 4.2837,
-      "step": 42
-    },
-    {
-      "epoch": 0.5119047619047619,
-      "grad_norm": 12.287630066184375,
-      "learning_rate": 1.686274509803922e-05,
-      "loss": 4.2075,
-      "step": 43
-    },
-    {
-      "epoch": 0.5238095238095238,
-      "grad_norm": 25.407055834196797,
-      "learning_rate": 1.7254901960784314e-05,
-      "loss": 4.4444,
-      "step": 44
-    },
-    {
-      "epoch": 0.5357142857142857,
-      "grad_norm": 14.241934754354682,
-      "learning_rate": 1.7647058823529414e-05,
-      "loss": 3.9014,
-      "step": 45
-    },
-    {
-      "epoch": 0.5476190476190477,
-      "grad_norm": 14.566784082629397,
-      "learning_rate": 1.8039215686274513e-05,
-      "loss": 3.5172,
-      "step": 46
-    },
-    {
-      "epoch": 0.5595238095238095,
-      "grad_norm": 13.542772667104277,
-      "learning_rate": 1.843137254901961e-05,
-      "loss": 3.296,
-      "step": 47
-    },
-    {
-      "epoch": 0.5714285714285714,
-      "grad_norm": 16.124803348222727,
-      "learning_rate": 1.8823529411764708e-05,
-      "loss": 3.0504,
-      "step": 48
-    },
-    {
-      "epoch": 0.5833333333333334,
-      "grad_norm": 12.95227008504844,
-      "learning_rate": 1.9215686274509807e-05,
-      "loss": 2.88,
-      "step": 49
-    },
-    {
-      "epoch": 0.5952380952380952,
-      "grad_norm": 3.595023968242296,
-      "learning_rate": 1.9607843137254903e-05,
-      "loss": 2.6277,
-      "step": 50
-    },
-    {
-      "epoch": 0.6071428571428571,
-      "grad_norm": 14.324134170867602,
-      "learning_rate": 2e-05,
-      "loss": 2.5481,
-      "step": 51
-    },
-    {
-      "epoch": 0.6190476190476191,
-      "grad_norm": 14.936425671150648,
-      "learning_rate": 1.9999981403661347e-05,
-      "loss": 2.4101,
-      "step": 52
-    },
-    {
-      "epoch": 0.6309523809523809,
-      "grad_norm": 11.892934392488657,
-      "learning_rate": 1.9999925614714537e-05,
-      "loss": 2.2475,
-      "step": 53
-    },
-    {
-      "epoch": 0.6428571428571429,
-      "grad_norm": 6.548368217885392,
-      "learning_rate": 1.9999832633367076e-05,
-      "loss": 2.5809,
-      "step": 54
-    },
-    {
-      "epoch": 0.6547619047619048,
-      "grad_norm": 11.621015010841388,
-      "learning_rate": 1.999970245996478e-05,
-      "loss": 2.0091,
-      "step": 55
-    },
-    {
-      "epoch": 0.6666666666666666,
-      "grad_norm": 8.696284471720817,
-      "learning_rate": 1.99995350949918e-05,
-      "loss": 2.1676,
-      "step": 56
-    },
-    {
-      "epoch": 0.6785714285714286,
-      "grad_norm": 12.002026381893657,
-      "learning_rate": 1.9999330539070615e-05,
-      "loss": 1.665,
-      "step": 57
-    },
-    {
-      "epoch": 0.6904761904761905,
-      "grad_norm": 11.065824262620708,
-      "learning_rate": 1.9999088792962016e-05,
-      "loss": 1.5071,
-      "step": 58
-    },
-    {
-      "epoch": 0.7023809523809523,
-      "grad_norm": 10.994414191496874,
-      "learning_rate": 1.999880985756513e-05,
-      "loss": 1.2489,
-      "step": 59
-    },
-    {
-      "epoch": 0.7142857142857143,
-      "grad_norm": 5.605348233753531,
-      "learning_rate": 1.9998493733917385e-05,
-      "loss": 2.1398,
-      "step": 60
-    },
-    {
-      "epoch": 0.7261904761904762,
-      "grad_norm": 11.91282152884282,
-      "learning_rate": 1.9998140423194534e-05,
-      "loss": 0.7581,
-      "step": 61
-    },
-    {
-      "epoch": 0.7380952380952381,
-      "grad_norm": 5.840433347765395,
-      "learning_rate": 1.9997749926710634e-05,
-      "loss": 1.4768,
-      "step": 62
-    },
-    {
-      "epoch": 0.75,
-      "grad_norm": 6.105791712493414,
-      "learning_rate": 1.999732224591804e-05,
-      "loss": 1.0261,
-      "step": 63
-    },
-    {
-      "epoch": 0.7619047619047619,
-      "grad_norm": 6.785960731870506,
-      "learning_rate": 1.999685738240742e-05,
-      "loss": 0.906,
-      "step": 64
-    },
-    {
-      "epoch": 0.7738095238095238,
-      "grad_norm": 6.2306364026893375,
-      "learning_rate": 1.999635533790772e-05,
-      "loss": 0.3629,
-      "step": 65
-    },
-    {
-      "epoch": 0.7857142857142857,
-      "grad_norm": 4.129747678855472,
-      "learning_rate": 1.999581611428618e-05,
-      "loss": 1.4474,
-      "step": 66
-    },
-    {
-      "epoch": 0.7976190476190477,
-      "grad_norm": 3.619952492724209,
-      "learning_rate": 1.9995239713548318e-05,
-      "loss": 1.207,
-      "step": 67
-    },
-    {
-      "epoch": 0.8095238095238095,
-      "grad_norm": 4.0590676768999305,
-      "learning_rate": 1.9994626137837917e-05,
-      "loss": 1.3818,
-      "step": 68
-    },
-    {
-      "epoch": 0.8214285714285714,
-      "grad_norm": 2.7517047479750576,
-      "learning_rate": 1.999397538943704e-05,
-      "loss": 1.2052,
-      "step": 69
-    },
-    {
-      "epoch": 0.8333333333333334,
-      "grad_norm": 6.041576469016595,
-      "learning_rate": 1.9993287470765986e-05,
-      "loss": 0.8238,
-      "step": 70
-    },
-    {
-      "epoch": 0.8452380952380952,
-      "grad_norm": 7.266622078598528,
-      "learning_rate": 1.999256238438331e-05,
-      "loss": 1.5338,
-      "step": 71
-    },
-    {
-      "epoch": 0.8571428571428571,
-      "grad_norm": 5.912372435902364,
-      "learning_rate": 1.9991800132985803e-05,
-      "loss": 2.3784,
-      "step": 72
-    },
-    {
-      "epoch": 0.8690476190476191,
-      "grad_norm": 4.956631346595567,
-      "learning_rate": 1.9991000719408483e-05,
-      "loss": 1.9622,
-      "step": 73
-    },
-    {
-      "epoch": 0.8809523809523809,
-      "grad_norm": 6.416670196307194,
-      "learning_rate": 1.9990164146624583e-05,
-      "loss": 0.8154,
-      "step": 74
-    },
-    {
-      "epoch": 0.8928571428571429,
-      "grad_norm": 12.54481609589616,
-      "learning_rate": 1.998929041774554e-05,
-      "loss": 1.4481,
-      "step": 75
-    },
-    {
-      "epoch": 0.9047619047619048,
-      "grad_norm": 3.1042285039390634,
-      "learning_rate": 1.9988379536020988e-05,
-      "loss": 1.2713,
-      "step": 76
-    },
-    {
-      "epoch": 0.9166666666666666,
-      "grad_norm": 4.825919181696002,
-      "learning_rate": 1.9987431504838737e-05,
-      "loss": 2.2873,
-      "step": 77
-    },
-    {
-      "epoch": 0.9285714285714286,
-      "grad_norm": 5.304665169610741,
-      "learning_rate": 1.998644632772477e-05,
-      "loss": 0.6765,
-      "step": 78
-    },
-    {
-      "epoch": 0.9404761904761905,
-      "grad_norm": 3.0999380965325196,
-      "learning_rate": 1.9985424008343226e-05,
-      "loss": 1.1929,
-      "step": 79
-    },
-    {
-      "epoch": 0.9523809523809523,
-      "grad_norm": 2.692769030054589,
-      "learning_rate": 1.998436455049638e-05,
-      "loss": 0.6175,
-      "step": 80
-    },
-    {
-      "epoch": 0.9642857142857143,
-      "grad_norm": 3.104212672349469,
-      "learning_rate": 1.9983267958124647e-05,
-      "loss": 1.2298,
-      "step": 81
-    },
-    {
-      "epoch": 0.9761904761904762,
-      "grad_norm": 3.3640203554586683,
-      "learning_rate": 1.998213423530654e-05,
-      "loss": 1.1316,
-      "step": 82
-    },
-    {
-      "epoch": 0.9880952380952381,
-      "grad_norm": 4.1473861422585525,
-      "learning_rate": 1.9980963386258682e-05,
-      "loss": 0.6541,
-      "step": 83
-    },
-    {
-      "epoch": 1.0,
-      "grad_norm": 4.372535318480156,
-      "learning_rate": 1.997975541533577e-05,
-      "loss": 1.3033,
-      "step": 84
-    },
-    {
-      "epoch": 1.0,
-      "eval_loss": 0.9026409983634949,
-      "eval_runtime": 57.3906,
-      "eval_samples_per_second": 1.045,
-      "eval_steps_per_second": 1.045,
-      "step": 84
-    },
-    {
-      "epoch": 1.0119047619047619,
-      "grad_norm": 2.777482345866408,
-      "learning_rate": 1.9978510327030577e-05,
-      "loss": 1.1553,
-      "step": 85
-    },
-    {
-      "epoch": 1.0238095238095237,
-      "grad_norm": 4.0464154340106475,
-      "learning_rate": 1.9977228125973917e-05,
-      "loss": 1.2478,
-      "step": 86
-    },
-    {
-      "epoch": 1.0357142857142858,
-      "grad_norm": 3.0692784835075306,
-      "learning_rate": 1.997590881693464e-05,
-      "loss": 1.1854,
-      "step": 87
-    },
-    {
-      "epoch": 1.0476190476190477,
-      "grad_norm": 2.3415895627089442,
-      "learning_rate": 1.9974552404819606e-05,
-      "loss": 1.5326,
-      "step": 88
-    },
-    {
-      "epoch": 1.0595238095238095,
-      "grad_norm": 2.336319119026421,
-      "learning_rate": 1.9973158894673677e-05,
-      "loss": 1.5136,
-      "step": 89
-    },
-    {
-      "epoch": 1.0714285714285714,
-      "grad_norm": 3.167429137483161,
-      "learning_rate": 1.9971728291679692e-05,
-      "loss": 0.2382,
-      "step": 90
-    },
-    {
-      "epoch": 1.0833333333333333,
-      "grad_norm": 2.8915816180219527,
-      "learning_rate": 1.9970260601158444e-05,
-      "loss": 1.0799,
-      "step": 91
-    },
-    {
-      "epoch": 1.0952380952380953,
-      "grad_norm": 3.2474222004118514,
-      "learning_rate": 1.996875582856867e-05,
-      "loss": 0.6601,
-      "step": 92
-    },
-    {
-      "epoch": 1.1071428571428572,
-      "grad_norm": 3.0071143059194436,
-      "learning_rate": 1.996721397950702e-05,
-      "loss": 1.0593,
-      "step": 93
-    },
-    {
-      "epoch": 1.119047619047619,
-      "grad_norm": 2.3487483579678625,
-      "learning_rate": 1.996563505970804e-05,
-      "loss": 0.5617,
-      "step": 94
-    },
-    {
-      "epoch": 1.130952380952381,
-      "grad_norm": 1.608206071362919,
-      "learning_rate": 1.9964019075044164e-05,
-      "loss": 1.1265,
-      "step": 95
-    },
-    {
-      "epoch": 1.1428571428571428,
-      "grad_norm": 2.1739566515425603,
-      "learning_rate": 1.9962366031525663e-05,
-      "loss": 0.638,
-      "step": 96
-    },
-    {
-      "epoch": 1.1547619047619047,
-      "grad_norm": 4.254923691373443,
-      "learning_rate": 1.9960675935300653e-05,
-      "loss": 1.5095,
-      "step": 97
-    },
-    {
-      "epoch": 1.1666666666666667,
-      "grad_norm": 2.1819566840909195,
-      "learning_rate": 1.9958948792655056e-05,
-      "loss": 0.5706,
-      "step": 98
-    },
-    {
-      "epoch": 1.1785714285714286,
-      "grad_norm": 2.0311354931594963,
-      "learning_rate": 1.995718461001257e-05,
-      "loss": 0.634,
-      "step": 99
-    },
-    {
-      "epoch": 1.1904761904761905,
-      "grad_norm": 1.5715938441379123,
-      "learning_rate": 1.9955383393934677e-05,
-      "loss": 0.9344,
-      "step": 100
-    },
-    {
-      "epoch": 1.2023809523809523,
-      "grad_norm": 2.230216937882692,
-      "learning_rate": 1.9953545151120565e-05,
-      "loss": 0.5713,
-      "step": 101
-    },
-    {
-      "epoch": 1.2142857142857142,
-      "grad_norm": 1.9246970673958967,
-      "learning_rate": 1.9951669888407162e-05,
-      "loss": 0.5854,
-      "step": 102
-    },
-    {
-      "epoch": 1.2261904761904763,
-      "grad_norm": 5.010856503469265,
-      "learning_rate": 1.9949757612769068e-05,
-      "loss": 0.6193,
-      "step": 103
-    },
-    {
-      "epoch": 1.2380952380952381,
-      "grad_norm": 1.6031600884223667,
-      "learning_rate": 1.994780833131855e-05,
-      "loss": 1.0085,
-      "step": 104
-    },
-    {
-      "epoch": 1.25,
-      "grad_norm": 2.749900960307275,
-      "learning_rate": 1.994582205130551e-05,
-      "loss": 1.4096,
-      "step": 105
-    },
-    {
-      "epoch": 1.2619047619047619,
-      "grad_norm": 1.0626905444801218,
-      "learning_rate": 1.9943798780117448e-05,
-      "loss": 0.5127,
-      "step": 106
-    },
-    {
-      "epoch": 1.2738095238095237,
-      "grad_norm": 1.574752679982518,
-      "learning_rate": 1.9941738525279456e-05,
-      "loss": 0.9511,
-      "step": 107
-    },
-    {
-      "epoch": 1.2857142857142856,
-      "grad_norm": 3.0210041542837596,
-      "learning_rate": 1.9939641294454172e-05,
-      "loss": 0.5167,
-      "step": 108
-    },
-    {
-      "epoch": 1.2976190476190477,
-      "grad_norm": 6.451462231840578,
-      "learning_rate": 1.993750709544176e-05,
-      "loss": 1.4284,
-      "step": 109
-    },
-    {
-      "epoch": 1.3095238095238095,
-      "grad_norm": 9.197366751047214,
-      "learning_rate": 1.9935335936179876e-05,
-      "loss": 0.2286,
-      "step": 110
-    },
-    {
-      "epoch": 1.3214285714285714,
-      "grad_norm": 0.9774070171128464,
-      "learning_rate": 1.9933127824743646e-05,
-      "loss": 0.4533,
-      "step": 111
-    },
-    {
-      "epoch": 1.3333333333333333,
-      "grad_norm": 1.5276991777107454,
-      "learning_rate": 1.9930882769345623e-05,
-      "loss": 0.529,
-      "step": 112
-    },
-    {
-      "epoch": 1.3452380952380953,
-      "grad_norm": 1.5612981824321013,
-      "learning_rate": 1.9928600778335774e-05,
-      "loss": 0.9313,
-      "step": 113
-    },
-    {
-      "epoch": 1.3571428571428572,
-      "grad_norm": 2.3815979703196852,
-      "learning_rate": 1.9926281860201427e-05,
-      "loss": 0.0845,
-      "step": 114
-    },
-    {
-      "epoch": 1.369047619047619,
-      "grad_norm": 4.8026812662489355,
-      "learning_rate": 1.992392602356727e-05,
-      "loss": 0.9438,
-      "step": 115
-    },
-    {
-      "epoch": 1.380952380952381,
-      "grad_norm": 5.356973288599696,
-      "learning_rate": 1.9921533277195282e-05,
-      "loss": 0.5935,
-      "step": 116
-    },
-    {
-      "epoch": 1.3928571428571428,
-      "grad_norm": 2.316890722187289,
-      "learning_rate": 1.9919103629984727e-05,
-      "loss": 1.2909,
-      "step": 117
-    },
-    {
-      "epoch": 1.4047619047619047,
-      "grad_norm": 3.1719727319199853,
-      "learning_rate": 1.991663709097212e-05,
-      "loss": 0.5475,
-      "step": 118
-    },
-    {
-      "epoch": 1.4166666666666667,
-      "grad_norm": 1.6801319700044128,
-      "learning_rate": 1.9914133669331174e-05,
-      "loss": 0.9584,
-      "step": 119
-    },
-    {
-      "epoch": 1.4285714285714286,
-      "grad_norm": 2.038248435291157,
-      "learning_rate": 1.991159337437279e-05,
-      "loss": 0.4672,
-      "step": 120
-    },
-    {
-      "epoch": 1.4404761904761905,
-      "grad_norm": 2.281787866316906,
-      "learning_rate": 1.9909016215544998e-05,
-      "loss": 0.5058,
-      "step": 121
-    },
-    {
-      "epoch": 1.4523809523809523,
-      "grad_norm": 6.05492934604481,
-      "learning_rate": 1.9906402202432945e-05,
-      "loss": 0.6174,
-      "step": 122
-    },
-    {
-      "epoch": 1.4642857142857144,
-      "grad_norm": 4.542635490766034,
-      "learning_rate": 1.990375134475885e-05,
-      "loss": 0.9407,
-      "step": 123
-    },
-    {
-      "epoch": 1.4761904761904763,
-      "grad_norm": 4.770874980454038,
-      "learning_rate": 1.9901063652381954e-05,
-      "loss": 0.1024,
-      "step": 124
-    },
-    {
-      "epoch": 1.4880952380952381,
-      "grad_norm": 1.8092172752263767,
-      "learning_rate": 1.989833913529851e-05,
-      "loss": 0.137,
-      "step": 125
-    },
-    {
-      "epoch": 1.5,
-      "grad_norm": 4.513986632423435,
-      "learning_rate": 1.9895577803641726e-05,
-      "loss": 0.5451,
-      "step": 126
-    },
-    {
-      "epoch": 1.5119047619047619,
-      "grad_norm": 1.4614473604306615,
-      "learning_rate": 1.9892779667681733e-05,
-      "loss": 0.8946,
-      "step": 127
-    },
-    {
-      "epoch": 1.5238095238095237,
-      "grad_norm": 4.049432365383836,
-      "learning_rate": 1.9889944737825546e-05,
-      "loss": 0.4902,
-      "step": 128
-    },
-    {
-      "epoch": 1.5357142857142856,
-      "grad_norm": 3.6084671463423876,
-      "learning_rate": 1.988707302461703e-05,
-      "loss": 0.0791,
-      "step": 129
-    },
-    {
-      "epoch": 1.5476190476190477,
-      "grad_norm": 1.8953178018348253,
-      "learning_rate": 1.9884164538736858e-05,
-      "loss": 0.1778,
-      "step": 130
-    },
-    {
-      "epoch": 1.5595238095238095,
-      "grad_norm": 1.692060013017211,
-      "learning_rate": 1.9881219291002462e-05,
-      "loss": 0.9511,
-      "step": 131
-    },
-    {
-      "epoch": 1.5714285714285714,
-      "grad_norm": 2.745211390170471,
-      "learning_rate": 1.9878237292368014e-05,
-      "loss": 0.9219,
-      "step": 132
-    },
-    {
-      "epoch": 1.5833333333333335,
-      "grad_norm": 6.211027429665557,
-      "learning_rate": 1.9875218553924357e-05,
-      "loss": 0.5817,
-      "step": 133
-    },
-    {
-      "epoch": 1.5952380952380953,
-      "grad_norm": 12.205614792958544,
-      "learning_rate": 1.9872163086898992e-05,
-      "loss": 0.2321,
-      "step": 134
-    },
-    {
-      "epoch": 1.6071428571428572,
-      "grad_norm": 4.263263812234368,
-      "learning_rate": 1.986907090265602e-05,
-      "loss": 0.0934,
-      "step": 135
-    },
-    {
-      "epoch": 1.619047619047619,
-      "grad_norm": 4.372915900877127,
-      "learning_rate": 1.98659420126961e-05,
-      "loss": 0.0622,
-      "step": 136
-    },
-    {
-      "epoch": 1.630952380952381,
-      "grad_norm": 1.6281660636980724,
-      "learning_rate": 1.9862776428656412e-05,
-      "loss": 0.9627,
-      "step": 137
-    },
-    {
-      "epoch": 1.6428571428571428,
-      "grad_norm": 1.466484459747661,
-      "learning_rate": 1.985957416231061e-05,
-      "loss": 0.1446,
-      "step": 138
-    },
-    {
-      "epoch": 1.6547619047619047,
-      "grad_norm": 1.3764747389198715,
-      "learning_rate": 1.9856335225568778e-05,
-      "loss": 0.8076,
-      "step": 139
-    },
-    {
-      "epoch": 1.6666666666666665,
-      "grad_norm": 11.454638511224184,
-      "learning_rate": 1.9853059630477396e-05,
-      "loss": 0.7499,
-      "step": 140
-    },
-    {
-      "epoch": 1.6785714285714286,
-      "grad_norm": 1.3551997471007746,
-      "learning_rate": 1.9849747389219272e-05,
-      "loss": 0.8396,
-      "step": 141
-    },
-    {
-      "epoch": 1.6904761904761905,
-      "grad_norm": 5.589546478044774,
-      "learning_rate": 1.984639851411352e-05,
-      "loss": 1.0755,
-      "step": 142
-    },
-    {
-      "epoch": 1.7023809523809523,
-      "grad_norm": 6.308349485763978,
-      "learning_rate": 1.9843013017615505e-05,
-      "loss": 0.599,
-      "step": 143
-    },
-    {
-      "epoch": 1.7142857142857144,
-      "grad_norm": 3.113963021942783,
-      "learning_rate": 1.9839590912316794e-05,
-      "loss": 0.839,
-      "step": 144
-    },
-    {
-      "epoch": 1.7261904761904763,
-      "grad_norm": 8.55365310087414,
-      "learning_rate": 1.9836132210945108e-05,
-      "loss": 0.193,
-      "step": 145
-    },
-    {
-      "epoch": 1.7380952380952381,
-      "grad_norm": 2.797343688675298,
-      "learning_rate": 1.9832636926364294e-05,
-      "loss": 1.2497,
-      "step": 146
-    },
-    {
-      "epoch": 1.75,
-      "grad_norm": 2.7347494170989566,
-      "learning_rate": 1.9829105071574243e-05,
-      "loss": 0.5603,
-      "step": 147
-    },
-    {
-      "epoch": 1.7619047619047619,
-      "grad_norm": 3.8951141138135257,
-      "learning_rate": 1.9825536659710866e-05,
-      "loss": 0.8488,
-      "step": 148
-    },
-    {
-      "epoch": 1.7738095238095237,
-      "grad_norm": 3.5133582001834225,
-      "learning_rate": 1.982193170404605e-05,
-      "loss": 0.8789,
-      "step": 149
-    },
-    {
-      "epoch": 1.7857142857142856,
-      "grad_norm": 4.91423782997247,
-      "learning_rate": 1.9818290217987587e-05,
-      "loss": 0.5037,
-      "step": 150
-    },
-    {
-      "epoch": 1.7976190476190477,
-      "grad_norm": 2.700610142348175,
-      "learning_rate": 1.981461221507914e-05,
-      "loss": 0.4866,
-      "step": 151
-    },
-    {
-      "epoch": 1.8095238095238095,
-      "grad_norm": 2.03214158550786,
-      "learning_rate": 1.9810897709000183e-05,
-      "loss": 0.872,
-      "step": 152
-    },
-    {
-      "epoch": 1.8214285714285714,
-      "grad_norm": 3.272072664954593,
-      "learning_rate": 1.9807146713565957e-05,
-      "loss": 1.2727,
-      "step": 153
-    },
-    {
-      "epoch": 1.8333333333333335,
-      "grad_norm": 1.581429785806821,
-      "learning_rate": 1.9803359242727427e-05,
-      "loss": 0.7566,
-      "step": 154
-    },
-    {
-      "epoch": 1.8452380952380953,
-      "grad_norm": 9.22961810487271,
-      "learning_rate": 1.9799535310571205e-05,
-      "loss": 0.6788,
-      "step": 155
-    },
-    {
-      "epoch": 1.8571428571428572,
-      "grad_norm": 0.7637949995107111,
-      "learning_rate": 1.9795674931319515e-05,
-      "loss": 0.418,
-      "step": 156
-    },
-    {
-      "epoch": 1.869047619047619,
-      "grad_norm": 1.2819972814953018,
-      "learning_rate": 1.979177811933015e-05,
-      "loss": 0.8087,
-      "step": 157
-    },
-    {
-      "epoch": 1.880952380952381,
-      "grad_norm": 1.6876409305315814,
-      "learning_rate": 1.978784488909639e-05,
-      "loss": 0.848,
-      "step": 158
-    },
-    {
-      "epoch": 1.8928571428571428,
-      "grad_norm": 3.4822248028190512,
-      "learning_rate": 1.9783875255246972e-05,
-      "loss": 0.5048,
-      "step": 159
-    },
-    {
-      "epoch": 1.9047619047619047,
-      "grad_norm": 1.9167539626251773,
-      "learning_rate": 1.9779869232546033e-05,
-      "loss": 0.8171,
-      "step": 160
-    },
-    {
-      "epoch": 1.9166666666666665,
-      "grad_norm": 2.2347320112562006,
-      "learning_rate": 1.977582683589304e-05,
-      "loss": 1.2693,
-      "step": 161
-    },
-    {
-      "epoch": 1.9285714285714286,
-      "grad_norm": 1.2561177879376604,
-      "learning_rate": 1.9771748080322746e-05,
-      "loss": 0.3953,
-      "step": 162
-    },
-    {
-      "epoch": 1.9404761904761905,
-      "grad_norm": 2.0408657215196495,
-      "learning_rate": 1.9767632981005138e-05,
-      "loss": 1.1772,
-      "step": 163
-    },
-    {
-      "epoch": 1.9523809523809523,
-      "grad_norm": 2.233965174646808,
-      "learning_rate": 1.976348155324537e-05,
-      "loss": 0.8335,
-      "step": 164
-    },
-    {
-      "epoch": 1.9642857142857144,
-      "grad_norm": 2.4594307933578605,
-      "learning_rate": 1.9759293812483717e-05,
-      "loss": 0.0393,
-      "step": 165
-    },
-    {
-      "epoch": 1.9761904761904763,
-      "grad_norm": 2.4993255857117513,
-      "learning_rate": 1.9755069774295502e-05,
-      "loss": 0.8245,
-      "step": 166
-    },
-    {
-      "epoch": 1.9880952380952381,
-      "grad_norm": 1.5646336637997305,
-      "learning_rate": 1.975080945439106e-05,
-      "loss": 0.4099,
-      "step": 167
-    },
-    {
-      "epoch": 2.0,
-      "grad_norm": 2.1272339637343025,
-      "learning_rate": 1.9746512868615656e-05,
-      "loss": 1.1694,
-      "step": 168
-    },
-    {
-      "epoch": 2.0,
-      "eval_loss": 0.6088699102401733,
-      "eval_runtime": 39.1086,
-      "eval_samples_per_second": 1.534,
-      "eval_steps_per_second": 1.534,
-      "step": 168
-    },
-    {
-      "epoch": 2.011904761904762,
-      "grad_norm": 1.0416876526645193,
-      "learning_rate": 1.974218003294945e-05,
-      "loss": 0.3786,
-      "step": 169
-    },
-    {
-      "epoch": 2.0238095238095237,
-      "grad_norm": 2.1337629148828365,
-      "learning_rate": 1.973781096350741e-05,
-      "loss": 0.4309,
-      "step": 170
-    },
-    {
-      "epoch": 2.0357142857142856,
-      "grad_norm": 2.2107303096576625,
-      "learning_rate": 1.9733405676539283e-05,
-      "loss": 1.0759,
-      "step": 171
-    },
-    {
-      "epoch": 2.0476190476190474,
-      "grad_norm": 2.0554217724529966,
-      "learning_rate": 1.97289641884295e-05,
-      "loss": 1.0771,
-      "step": 172
-    },
-    {
-      "epoch": 2.0595238095238093,
-      "grad_norm": 2.1586243563623366,
-      "learning_rate": 1.9724486515697157e-05,
-      "loss": 1.1039,
-      "step": 173
-    },
-    {
-      "epoch": 2.0714285714285716,
-      "grad_norm": 4.584438173048313,
-      "learning_rate": 1.9719972674995905e-05,
-      "loss": 0.8035,
-      "step": 174
-    },
-    {
-      "epoch": 2.0833333333333335,
-      "grad_norm": 1.2828049201231282,
-      "learning_rate": 1.971542268311394e-05,
-      "loss": 0.739,
-      "step": 175
-    },
-    {
-      "epoch": 2.0952380952380953,
-      "grad_norm": 2.5469482800959047,
-      "learning_rate": 1.9710836556973887e-05,
-      "loss": 0.4056,
-      "step": 176
-    },
-    {
-      "epoch": 2.107142857142857,
-      "grad_norm": 3.1810387560100524,
-      "learning_rate": 1.9706214313632786e-05,
-      "loss": 0.8045,
-      "step": 177
-    },
-    {
-      "epoch": 2.119047619047619,
-      "grad_norm": 3.2420213991910956,
-      "learning_rate": 1.970155597028199e-05,
-      "loss": 0.3999,
-      "step": 178
-    },
-    {
-      "epoch": 2.130952380952381,
-      "grad_norm": 1.9699411778652773,
-      "learning_rate": 1.969686154424713e-05,
-      "loss": 0.9999,
-      "step": 179
-    },
-    {
-      "epoch": 2.142857142857143,
-      "grad_norm": 2.627806590925809,
-      "learning_rate": 1.9692131052988035e-05,
-      "loss": 0.1001,
-      "step": 180
-    },
-    {
-      "epoch": 2.1547619047619047,
-      "grad_norm": 3.7819808176512772,
-      "learning_rate": 1.9687364514098664e-05,
-      "loss": 0.0601,
-      "step": 181
-    },
-    {
-      "epoch": 2.1666666666666665,
-      "grad_norm": 2.9442257019853844,
-      "learning_rate": 1.9682561945307052e-05,
-      "loss": 0.7165,
-      "step": 182
-    },
-    {
-      "epoch": 2.1785714285714284,
-      "grad_norm": 2.7812375288605446,
-      "learning_rate": 1.9677723364475237e-05,
-      "loss": 0.7077,
-      "step": 183
-    },
-    {
-      "epoch": 2.1904761904761907,
-      "grad_norm": 2.662139054171453,
-      "learning_rate": 1.9672848789599204e-05,
-      "loss": 0.0975,
-      "step": 184
-    },
-    {
-      "epoch": 2.2023809523809526,
-      "grad_norm": 3.13638542402792,
-      "learning_rate": 1.9667938238808796e-05,
-      "loss": 0.1782,
-      "step": 185
-    },
-    {
-      "epoch": 2.2142857142857144,
-      "grad_norm": 1.533138893606418,
-      "learning_rate": 1.9662991730367664e-05,
-      "loss": 0.7253,
-      "step": 186
-    },
-    {
-      "epoch": 2.2261904761904763,
-      "grad_norm": 1.6326899334327516,
-      "learning_rate": 1.9658009282673202e-05,
-      "loss": 0.4389,
-      "step": 187
-    },
-    {
-      "epoch": 2.238095238095238,
-      "grad_norm": 4.129746342978926,
-      "learning_rate": 1.9652990914256466e-05,
-      "loss": 0.0696,
-      "step": 188
-    },
-    {
-      "epoch": 2.25,
-      "grad_norm": 4.866194278395534,
-      "learning_rate": 1.964793664378211e-05,
-      "loss": 0.6483,
-      "step": 189
-    },
-    {
-      "epoch": 2.261904761904762,
-      "grad_norm": 4.040961944862707,
-      "learning_rate": 1.964284649004832e-05,
-      "loss": 0.3702,
-      "step": 190
-    },
-    {
-      "epoch": 2.2738095238095237,
-      "grad_norm": 5.955269361913289,
-      "learning_rate": 1.9637720471986736e-05,
-      "loss": 0.729,
-      "step": 191
-    },
-    {
-      "epoch": 2.2857142857142856,
-      "grad_norm": 2.416827364213375,
-      "learning_rate": 1.9632558608662403e-05,
-      "loss": 1.0398,
-      "step": 192
-    },
-    {
-      "epoch": 2.2976190476190474,
-      "grad_norm": 4.386696083224722,
-      "learning_rate": 1.962736091927366e-05,
-      "loss": 0.0767,
-      "step": 193
-    },
-    {
-      "epoch": 2.3095238095238093,
-      "grad_norm": 3.5619821484086467,
-      "learning_rate": 1.9622127423152114e-05,
-      "loss": 0.4072,
-      "step": 194
-    },
-    {
-      "epoch": 2.3214285714285716,
-      "grad_norm": 3.265850499691294,
-      "learning_rate": 1.9616858139762534e-05,
-      "loss": 0.6642,
-      "step": 195
-    },
-    {
-      "epoch": 2.3333333333333335,
-      "grad_norm": 2.5415289262555687,
-      "learning_rate": 1.96115530887028e-05,
-      "loss": 0.8033,
-      "step": 196
-    },
-    {
-      "epoch": 2.3452380952380953,
-      "grad_norm": 2.5712529524190577,
-      "learning_rate": 1.9606212289703813e-05,
-      "loss": 0.0824,
-      "step": 197
-    },
-    {
-      "epoch": 2.357142857142857,
-      "grad_norm": 5.474104888697016,
-      "learning_rate": 1.9600835762629434e-05,
-      "loss": 0.4142,
-      "step": 198
-    },
-    {
-      "epoch": 2.369047619047619,
-      "grad_norm": 2.879836797156971,
-      "learning_rate": 1.9595423527476407e-05,
-      "loss": 0.086,
-      "step": 199
-    },
-    {
-      "epoch": 2.380952380952381,
-      "grad_norm": 1.62480689368467,
-      "learning_rate": 1.9589975604374287e-05,
-      "loss": 0.6521,
-      "step": 200
-    },
-    {
-      "epoch": 2.392857142857143,
-      "grad_norm": 4.477628590267687,
-      "learning_rate": 1.9584492013585358e-05,
-      "loss": 0.4574,
-      "step": 201
-    },
-    {
-      "epoch": 2.4047619047619047,
-      "grad_norm": 3.7371630710469907,
-      "learning_rate": 1.9578972775504556e-05,
-      "loss": 0.7116,
-      "step": 202
-    },
-    {
-      "epoch": 2.4166666666666665,
-      "grad_norm": 4.388820750003364,
-      "learning_rate": 1.9573417910659414e-05,
-      "loss": 0.8658,
-      "step": 203
-    },
-    {
-      "epoch": 2.4285714285714284,
-      "grad_norm": 5.264780771949968,
-      "learning_rate": 1.9567827439709954e-05,
-      "loss": 0.5218,
-      "step": 204
-    },
-    {
-      "epoch": 2.4404761904761907,
-      "grad_norm": 2.8263063565249014,
-      "learning_rate": 1.9562201383448637e-05,
-      "loss": 0.589,
-      "step": 205
-    },
-    {
-      "epoch": 2.4523809523809526,
-      "grad_norm": 1.9326495868158482,
-      "learning_rate": 1.9556539762800276e-05,
-      "loss": 0.6548,
-      "step": 206
-    },
-    {
-      "epoch": 2.4642857142857144,
-      "grad_norm": 1.4295352535114343,
-      "learning_rate": 1.9550842598821954e-05,
-      "loss": 0.2921,
-      "step": 207
-    },
-    {
-      "epoch": 2.4761904761904763,
-      "grad_norm": 4.131679625334346,
-      "learning_rate": 1.954510991270294e-05,
-      "loss": 0.0552,
-      "step": 208
-    },
-    {
-      "epoch": 2.488095238095238,
-      "grad_norm": 1.2944968084779456,
-      "learning_rate": 1.953934172576464e-05,
-      "loss": 0.5584,
-      "step": 209
-    },
-    {
-      "epoch": 2.5,
-      "grad_norm": 2.3368325920978656,
-      "learning_rate": 1.9533538059460475e-05,
-      "loss": 0.8149,
-      "step": 210
-    },
-    {
-      "epoch": 2.511904761904762,
-      "grad_norm": 2.3435541020549957,
-      "learning_rate": 1.952769893537584e-05,
-      "loss": 0.7674,
-      "step": 211
-    },
-    {
-      "epoch": 2.5238095238095237,
-      "grad_norm": 1.4945515120342354,
-      "learning_rate": 1.9521824375228005e-05,
-      "loss": 0.6053,
-      "step": 212
-    },
-    {
-      "epoch": 2.5357142857142856,
-      "grad_norm": 1.6994434814237505,
-      "learning_rate": 1.9515914400866022e-05,
-      "loss": 0.6523,
-      "step": 213
-    },
-    {
-      "epoch": 2.5476190476190474,
-      "grad_norm": 1.7583638651152655,
-      "learning_rate": 1.9509969034270675e-05,
-      "loss": 0.324,
-      "step": 214
-    },
-    {
-      "epoch": 2.5595238095238093,
-      "grad_norm": 2.303351195831185,
-      "learning_rate": 1.950398829755437e-05,
-      "loss": 0.7802,
-      "step": 215
-    },
-    {
-      "epoch": 2.571428571428571,
-      "grad_norm": 1.2085607857296046,
-      "learning_rate": 1.9497972212961073e-05,
-      "loss": 0.5606,
-      "step": 216
-    },
-    {
-      "epoch": 2.5833333333333335,
-      "grad_norm": 4.953999376334107,
-      "learning_rate": 1.9491920802866207e-05,
-      "loss": 0.6656,
-      "step": 217
-    },
-    {
-      "epoch": 2.5952380952380953,
-      "grad_norm": 10.462260026499376,
-      "learning_rate": 1.9485834089776587e-05,
-      "loss": 0.1941,
-      "step": 218
-    },
-    {
-      "epoch": 2.607142857142857,
-      "grad_norm": 2.098150876215288,
-      "learning_rate": 1.9479712096330335e-05,
-      "loss": 0.371,
-      "step": 219
-    },
-    {
-      "epoch": 2.619047619047619,
-      "grad_norm": 6.9790355144844805,
-      "learning_rate": 1.947355484529678e-05,
-      "loss": 0.3854,
-      "step": 220
-    },
-    {
-      "epoch": 2.630952380952381,
-      "grad_norm": 4.988557486539209,
-      "learning_rate": 1.9467362359576388e-05,
-      "loss": 0.333,
-      "step": 221
-    },
-    {
-      "epoch": 2.642857142857143,
-      "grad_norm": 7.063238285522866,
-      "learning_rate": 1.9461134662200667e-05,
-      "loss": 0.3903,
-      "step": 222
-    },
-    {
-      "epoch": 2.6547619047619047,
-      "grad_norm": 1.5164015264404487,
-      "learning_rate": 1.9454871776332096e-05,
-      "loss": 0.5085,
-      "step": 223
-    },
-    {
-      "epoch": 2.6666666666666665,
-      "grad_norm": 2.522359534562382,
-      "learning_rate": 1.944857372526402e-05,
-      "loss": 0.738,
-      "step": 224
-    },
-    {
-      "epoch": 2.678571428571429,
-      "grad_norm": 1.555259463726646,
-      "learning_rate": 1.9442240532420585e-05,
-      "loss": 0.4806,
-      "step": 225
-    },
-    {
-      "epoch": 2.6904761904761907,
-      "grad_norm": 7.951689819472121,
-      "learning_rate": 1.9435872221356623e-05,
-      "loss": 0.4552,
-      "step": 226
-    },
-    {
-      "epoch": 2.7023809523809526,
-      "grad_norm": 1.5044684923460068,
-      "learning_rate": 1.9429468815757587e-05,
-      "loss": 0.4904,
-      "step": 227
-    },
-    {
-      "epoch": 2.7142857142857144,
-      "grad_norm": 1.4782862605961744,
-      "learning_rate": 1.9423030339439464e-05,
-      "loss": 0.459,
-      "step": 228
-    },
-    {
-      "epoch": 2.7261904761904763,
-      "grad_norm": 7.30930822807625,
-      "learning_rate": 1.9416556816348663e-05,
-      "loss": 0.4524,
-      "step": 229
-    },
-    {
-      "epoch": 2.738095238095238,
-      "grad_norm": 0.8563665535357093,
-      "learning_rate": 1.9410048270561958e-05,
-      "loss": 0.2686,
-      "step": 230
-    },
-    {
-      "epoch": 2.75,
-      "grad_norm": 6.36705833497127,
-      "learning_rate": 1.9403504726286367e-05,
-      "loss": 0.4109,
-      "step": 231
-    },
-    {
-      "epoch": 2.761904761904762,
-      "grad_norm": 1.6877844007234946,
-      "learning_rate": 1.9396926207859085e-05,
-      "loss": 0.0287,
-      "step": 232
-    },
-    {
-      "epoch": 2.7738095238095237,
-      "grad_norm": 1.5933904005469899,
-      "learning_rate": 1.9390312739747384e-05,
-      "loss": 0.25,
-      "step": 233
-    },
-    {
-      "epoch": 2.7857142857142856,
-      "grad_norm": 3.8409331200672043,
-      "learning_rate": 1.938366434654852e-05,
-      "loss": 0.1199,
-      "step": 234
-    },
-    {
-      "epoch": 2.7976190476190474,
-      "grad_norm": 1.9144885212012308,
-      "learning_rate": 1.9376981052989655e-05,
-      "loss": 0.4406,
-      "step": 235
-    },
-    {
-      "epoch": 2.8095238095238093,
-      "grad_norm": 3.138591834492908,
-      "learning_rate": 1.9370262883927737e-05,
-      "loss": 0.3056,
-      "step": 236
-    },
-    {
-      "epoch": 2.821428571428571,
-      "grad_norm": 1.5020413167589903,
-      "learning_rate": 1.9363509864349438e-05,
-      "loss": 0.3429,
-      "step": 237
-    },
-    {
-      "epoch": 2.8333333333333335,
-      "grad_norm": 3.3495647793942003,
-      "learning_rate": 1.935672201937105e-05,
-      "loss": 0.0534,
-      "step": 238
-    },
-    {
-      "epoch": 2.8452380952380953,
-      "grad_norm": 2.362065083308391,
-      "learning_rate": 1.9349899374238384e-05,
-      "loss": 0.5528,
-      "step": 239
-    },
-    {
-      "epoch": 2.857142857142857,
-      "grad_norm": 2.0160995396477723,
-      "learning_rate": 1.934304195432668e-05,
-      "loss": 0.6616,
-      "step": 240
-    },
-    {
-      "epoch": 2.869047619047619,
-      "grad_norm": 3.3847007434339496,
-      "learning_rate": 1.9336149785140524e-05,
-      "loss": 0.0527,
-      "step": 241
-    },
-    {
-      "epoch": 2.880952380952381,
-      "grad_norm": 2.934474785594088,
-      "learning_rate": 1.9329222892313738e-05,
-      "loss": 0.3513,
-      "step": 242
-    },
-    {
-      "epoch": 2.892857142857143,
-      "grad_norm": 2.1712614824222967,
-      "learning_rate": 1.9322261301609286e-05,
-      "loss": 0.0815,
-      "step": 243
-    },
-    {
-      "epoch": 2.9047619047619047,
-      "grad_norm": 1.421255557551219,
-      "learning_rate": 1.9315265038919194e-05,
-      "loss": 0.2604,
-      "step": 244
-    },
-    {
-      "epoch": 2.9166666666666665,
-      "grad_norm": 2.208193199321934,
-      "learning_rate": 1.930823413026443e-05,
-      "loss": 0.4335,
-      "step": 245
-    },
-    {
-      "epoch": 2.928571428571429,
-      "grad_norm": 1.5103556530084337,
-      "learning_rate": 1.9301168601794832e-05,
-      "loss": 0.2859,
-      "step": 246
-    },
-    {
-      "epoch": 2.9404761904761907,
-      "grad_norm": 2.499861790171546,
-      "learning_rate": 1.9294068479788987e-05,
-      "loss": 0.6764,
-      "step": 247
-    },
-    {
-      "epoch": 2.9523809523809526,
-      "grad_norm": 3.708903007657656,
-      "learning_rate": 1.928693379065415e-05,
-      "loss": 0.5288,
-      "step": 248
-    },
-    {
-      "epoch": 2.9642857142857144,
-      "grad_norm": 22.915584132176974,
-      "learning_rate": 1.9279764560926142e-05,
-      "loss": 0.9066,
-      "step": 249
-    },
-    {
-      "epoch": 2.9761904761904763,
-      "grad_norm": 4.272981133884936,
-      "learning_rate": 1.927256081726925e-05,
-      "loss": 0.0868,
-      "step": 250
-    },
-    {
-      "epoch": 2.988095238095238,
-      "grad_norm": 1.4918399220179503,
-      "learning_rate": 1.926532258647612e-05,
-      "loss": 0.4034,
-      "step": 251
-    },
-    {
-      "epoch": 3.0,
-      "grad_norm": 4.341407962457437,
-      "learning_rate": 1.9258049895467672e-05,
-      "loss": 0.5374,
-      "step": 252
-    },
-    {
-      "epoch": 3.0,
-      "eval_loss": 0.39895328879356384,
-      "eval_runtime": 38.7955,
-      "eval_samples_per_second": 1.547,
-      "eval_steps_per_second": 1.547,
-      "step": 252
-    },
-    {
-      "epoch": 3.011904761904762,
-      "grad_norm": 2.999227280578599,
-      "learning_rate": 1.9250742771292993e-05,
-      "loss": 0.2335,
-      "step": 253
-    },
-    {
-      "epoch": 3.0238095238095237,
-      "grad_norm": 2.977688909060128,
-      "learning_rate": 1.9243401241129232e-05,
-      "loss": 0.3035,
-      "step": 254
-    },
-    {
-      "epoch": 3.0357142857142856,
-      "grad_norm": 4.856998071383737,
-      "learning_rate": 1.9236025332281506e-05,
-      "loss": 0.4267,
-      "step": 255
-    },
-    {
-      "epoch": 3.0476190476190474,
-      "grad_norm": 3.6416277575589784,
-      "learning_rate": 1.9228615072182798e-05,
-      "loss": 0.0806,
-      "step": 256
-    },
-    {
-      "epoch": 3.0595238095238093,
-      "grad_norm": 3.8536704247985667,
-      "learning_rate": 1.9221170488393844e-05,
-      "loss": 0.5701,
-      "step": 257
-    },
-    {
-      "epoch": 3.0714285714285716,
-      "grad_norm": 5.006059228225849,
-      "learning_rate": 1.9213691608603046e-05,
-      "loss": 0.2624,
-      "step": 258
-    },
-    {
-      "epoch": 3.0833333333333335,
-      "grad_norm": 3.477751917820131,
-      "learning_rate": 1.920617846062636e-05,
-      "loss": 0.4209,
-      "step": 259
-    },
-    {
-      "epoch": 3.0952380952380953,
-      "grad_norm": 7.008269361085479,
-      "learning_rate": 1.91986310724072e-05,
-      "loss": 0.5247,
-      "step": 260
-    },
-    {
-      "epoch": 3.107142857142857,
-      "grad_norm": 2.4584658954882044,
-      "learning_rate": 1.9191049472016313e-05,
-      "loss": 0.0346,
-      "step": 261
-    },
-    {
-      "epoch": 3.119047619047619,
-      "grad_norm": 5.012502345173128,
-      "learning_rate": 1.918343368765171e-05,
-      "loss": 0.0991,
-      "step": 262
-    },
-    {
-      "epoch": 3.130952380952381,
-      "grad_norm": 4.016227848571736,
-      "learning_rate": 1.917578374763853e-05,
-      "loss": 0.4093,
-      "step": 263
-    },
-    {
-      "epoch": 3.142857142857143,
-      "grad_norm": 3.284851594075178,
-      "learning_rate": 1.9168099680428943e-05,
-      "loss": 0.4856,
-      "step": 264
-    },
-    {
-      "epoch": 3.1547619047619047,
-      "grad_norm": 1.439490683728371,
-      "learning_rate": 1.916038151460206e-05,
-      "loss": 0.3762,
-      "step": 265
-    },
-    {
-      "epoch": 3.1666666666666665,
-      "grad_norm": 2.5758200075686895,
-      "learning_rate": 1.91526292788638e-05,
-      "loss": 0.3824,
-      "step": 266
-    },
-    {
-      "epoch": 3.1785714285714284,
-      "grad_norm": 1.5227377147219847,
-      "learning_rate": 1.9144843002046805e-05,
-      "loss": 0.3724,
-      "step": 267
-    },
-    {
-      "epoch": 3.1904761904761907,
-      "grad_norm": 2.607295269246451,
-      "learning_rate": 1.9137022713110323e-05,
-      "loss": 0.4212,
-      "step": 268
-    },
-    {
-      "epoch": 3.2023809523809526,
-      "grad_norm": 4.777066718838521,
-      "learning_rate": 1.9129168441140104e-05,
-      "loss": 0.2489,
-      "step": 269
-    },
-    {
-      "epoch": 3.2142857142857144,
-      "grad_norm": 0.8359102150893967,
-      "learning_rate": 1.912128021534829e-05,
-      "loss": 0.1657,
-      "step": 270
-    },
-    {
-      "epoch": 3.2261904761904763,
-      "grad_norm": 2.594995245986025,
-      "learning_rate": 1.9113358065073297e-05,
-      "loss": 0.5461,
-      "step": 271
-    },
-    {
-      "epoch": 3.238095238095238,
-      "grad_norm": 0.9681291220123174,
-      "learning_rate": 1.9105402019779728e-05,
-      "loss": 0.1927,
-      "step": 272
-    },
-    {
-      "epoch": 3.25,
-      "grad_norm": 1.0698917247005957,
-      "learning_rate": 1.9097412109058247e-05,
-      "loss": 0.0175,
-      "step": 273
-    },
-    {
-      "epoch": 3.261904761904762,
-      "grad_norm": 2.5131905938872694,
-      "learning_rate": 1.9089388362625468e-05,
-      "loss": 0.5642,
-      "step": 274
-    },
-    {
-      "epoch": 3.2738095238095237,
-      "grad_norm": 1.0075136405580187,
-      "learning_rate": 1.9081330810323852e-05,
-      "loss": 0.1631,
-      "step": 275
-    },
-    {
-      "epoch": 3.2857142857142856,
-      "grad_norm": 6.448857884660445,
-      "learning_rate": 1.9073239482121598e-05,
-      "loss": 0.4545,
-      "step": 276
-    },
-    {
-      "epoch": 3.2976190476190474,
-      "grad_norm": 10.404043140853894,
-      "learning_rate": 1.906511440811252e-05,
-      "loss": 0.4511,
-      "step": 277
-    },
-    {
-      "epoch": 3.3095238095238093,
-      "grad_norm": 3.246806863815276,
-      "learning_rate": 1.9056955618515934e-05,
-      "loss": 0.2476,
-      "step": 278
-    },
-    {
-      "epoch": 3.3214285714285716,
-      "grad_norm": 2.289193427727155,
-      "learning_rate": 1.904876314367658e-05,
-      "loss": 0.1657,
-      "step": 279
-    },
-    {
-      "epoch": 3.3333333333333335,
-      "grad_norm": 2.980090661433879,
-      "learning_rate": 1.904053701406445e-05,
-      "loss": 0.1611,
-      "step": 280
-    },
-    {
-      "epoch": 3.3452380952380953,
-      "grad_norm": 7.457260763434641,
-      "learning_rate": 1.9032277260274733e-05,
-      "loss": 0.138,
-      "step": 281
-    },
-    {
-      "epoch": 3.357142857142857,
-      "grad_norm": 2.607895595891462,
-      "learning_rate": 1.9023983913027655e-05,
-      "loss": 0.3192,
-      "step": 282
-    },
-    {
-      "epoch": 3.369047619047619,
-      "grad_norm": 7.092330079559744,
-      "learning_rate": 1.9015657003168405e-05,
-      "loss": 0.1357,
-      "step": 283
-    },
-    {
-      "epoch": 3.380952380952381,
-      "grad_norm": 1.24814877201098,
-      "learning_rate": 1.9007296561666987e-05,
-      "loss": 0.2687,
-      "step": 284
-    },
-    {
-      "epoch": 3.392857142857143,
-      "grad_norm": 1.8637801294405039,
-      "learning_rate": 1.8998902619618117e-05,
-      "loss": 0.2858,
-      "step": 285
-    },
-    {
-      "epoch": 3.4047619047619047,
-      "grad_norm": 4.888040446870592,
-      "learning_rate": 1.8990475208241114e-05,
-      "loss": 0.0924,
-      "step": 286
-    },
-    {
-      "epoch": 3.4166666666666665,
-      "grad_norm": 4.001525244593556,
-      "learning_rate": 1.898201435887978e-05,
-      "loss": 0.186,
-      "step": 287
-    },
-    {
-      "epoch": 3.4285714285714284,
-      "grad_norm": 1.5303656577521532,
-      "learning_rate": 1.8973520103002278e-05,
-      "loss": 0.2531,
-      "step": 288
-    },
-    {
-      "epoch": 3.4404761904761907,
-      "grad_norm": 2.6407383882318136,
-      "learning_rate": 1.896499247220102e-05,
-      "loss": 0.0644,
-      "step": 289
-    },
-    {
-      "epoch": 3.4523809523809526,
-      "grad_norm": 5.013301155952652,
-      "learning_rate": 1.8956431498192547e-05,
-      "loss": 0.0803,
-      "step": 290
-    },
-    {
-      "epoch": 3.4642857142857144,
-      "grad_norm": 2.0289616983127865,
-      "learning_rate": 1.8947837212817414e-05,
-      "loss": 0.0326,
-      "step": 291
-    },
-    {
-      "epoch": 3.4761904761904763,
-      "grad_norm": 2.6152472211362783,
-      "learning_rate": 1.893920964804007e-05,
-      "loss": 0.4119,
-      "step": 292
-    },
-    {
-      "epoch": 3.488095238095238,
-      "grad_norm": 4.4777909223010495,
-      "learning_rate": 1.8930548835948738e-05,
-      "loss": 0.3507,
-      "step": 293
-    },
-    {
-      "epoch": 3.5,
-      "grad_norm": 0.6366794435472163,
-      "learning_rate": 1.8921854808755295e-05,
-      "loss": 0.0115,
-      "step": 294
-    },
-    {
-      "epoch": 3.511904761904762,
-      "grad_norm": 2.4420373988670883,
-      "learning_rate": 1.8913127598795157e-05,
-      "loss": 0.4161,
-      "step": 295
-    },
-    {
-      "epoch": 3.5238095238095237,
-      "grad_norm": 2.4446578383404365,
-      "learning_rate": 1.8904367238527156e-05,
-      "loss": 0.1434,
-      "step": 296
-    },
-    {
-      "epoch": 3.5357142857142856,
-      "grad_norm": 1.9529776383968345,
-      "learning_rate": 1.8895573760533413e-05,
-      "loss": 0.2163,
-      "step": 297
-    },
-    {
-      "epoch": 3.5476190476190474,
-      "grad_norm": 10.168211887481995,
-      "learning_rate": 1.8886747197519232e-05,
-      "loss": 0.3446,
-      "step": 298
-    },
-    {
-      "epoch": 3.5595238095238093,
-      "grad_norm": 0.8708790799659621,
-      "learning_rate": 1.887788758231296e-05,
-      "loss": 0.172,
-      "step": 299
-    },
-    {
-      "epoch": 3.571428571428571,
-      "grad_norm": 1.6300564668273962,
-      "learning_rate": 1.8868994947865882e-05,
-      "loss": 0.3293,
-      "step": 300
-    },
-    {
-      "epoch": 3.5833333333333335,
-      "grad_norm": 1.5955068953197555,
-      "learning_rate": 1.8860069327252085e-05,
-      "loss": 0.1722,
-      "step": 301
-    },
-    {
-      "epoch": 3.5952380952380953,
-      "grad_norm": 3.5420643404756045,
-      "learning_rate": 1.885111075366834e-05,
-      "loss": 0.3466,
-      "step": 302
-    },
-    {
-      "epoch": 3.607142857142857,
-      "grad_norm": 0.8255964441201371,
-      "learning_rate": 1.8842119260433984e-05,
-      "loss": 0.1426,
-      "step": 303
-    },
-    {
-      "epoch": 3.619047619047619,
-      "grad_norm": 2.2416092982597835,
-      "learning_rate": 1.8833094880990777e-05,
-      "loss": 0.1176,
-      "step": 304
-    },
-    {
-      "epoch": 3.630952380952381,
-      "grad_norm": 3.17281147686109,
-      "learning_rate": 1.882403764890282e-05,
-      "loss": 0.2363,
-      "step": 305
-    },
-    {
-      "epoch": 3.642857142857143,
-      "grad_norm": 3.50801962281108,
-      "learning_rate": 1.8814947597856366e-05,
-      "loss": 0.1298,
-      "step": 306
-    },
-    {
-      "epoch": 3.6547619047619047,
-      "grad_norm": 2.015451984498423,
-      "learning_rate": 1.8805824761659766e-05,
-      "loss": 0.0503,
-      "step": 307
-    },
-    {
-      "epoch": 3.6666666666666665,
-      "grad_norm": 1.8138556454595522,
-      "learning_rate": 1.8796669174243274e-05,
-      "loss": 0.1708,
-      "step": 308
-    },
-    {
-      "epoch": 3.678571428571429,
-      "grad_norm": 1.350180232476615,
-      "learning_rate": 1.878748086965898e-05,
-      "loss": 0.2102,
-      "step": 309
-    },
-    {
-      "epoch": 3.6904761904761907,
-      "grad_norm": 5.44990543805186,
-      "learning_rate": 1.877825988208065e-05,
-      "loss": 0.1776,
-      "step": 310
-    },
-    {
-      "epoch": 3.7023809523809526,
-      "grad_norm": 5.429618864013119,
-      "learning_rate": 1.8769006245803597e-05,
-      "loss": 0.2871,
-      "step": 311
-    },
-    {
-      "epoch": 3.7142857142857144,
-      "grad_norm": 9.615841204468909,
-      "learning_rate": 1.875971999524458e-05,
-      "loss": 0.4327,
-      "step": 312
-    },
-    {
-      "epoch": 3.7261904761904763,
-      "grad_norm": 13.832201073491575,
-      "learning_rate": 1.875040116494165e-05,
-      "loss": 0.3311,
-      "step": 313
-    },
-    {
-      "epoch": 3.738095238095238,
-      "grad_norm": 3.1133674287556383,
-      "learning_rate": 1.874104978955403e-05,
-      "loss": 0.1555,
-      "step": 314
-    },
-    {
-      "epoch": 3.75,
-      "grad_norm": 2.347399033576179,
-      "learning_rate": 1.8731665903861987e-05,
-      "loss": 0.1674,
-      "step": 315
-    },
-    {
-      "epoch": 3.761904761904762,
-      "grad_norm": 9.849958200867494,
-      "learning_rate": 1.8722249542766704e-05,
-      "loss": 0.1878,
-      "step": 316
-    },
-    {
-      "epoch": 3.7738095238095237,
-      "grad_norm": 2.435572970484653,
-      "learning_rate": 1.8712800741290155e-05,
-      "loss": 0.1737,
-      "step": 317
-    },
-    {
-      "epoch": 3.7857142857142856,
-      "grad_norm": 1.3696724790752792,
-      "learning_rate": 1.8703319534574954e-05,
-      "loss": 0.2522,
-      "step": 318
-    },
-    {
-      "epoch": 3.7976190476190474,
-      "grad_norm": 2.3593401003654937,
-      "learning_rate": 1.869380595788426e-05,
-      "loss": 0.1571,
-      "step": 319
-    },
-    {
-      "epoch": 3.8095238095238093,
-      "grad_norm": 1.0006750257594192,
-      "learning_rate": 1.8684260046601595e-05,
-      "loss": 0.131,
-      "step": 320
-    },
-    {
-      "epoch": 3.821428571428571,
-      "grad_norm": 8.568117470038834,
-      "learning_rate": 1.867468183623077e-05,
-      "loss": 0.2712,
-      "step": 321
-    },
-    {
-      "epoch": 3.8333333333333335,
-      "grad_norm": 4.128637677242795,
-      "learning_rate": 1.8665071362395714e-05,
-      "loss": 0.2944,
-      "step": 322
-    },
-    {
-      "epoch": 3.8452380952380953,
-      "grad_norm": 8.30896311650967,
-      "learning_rate": 1.8655428660840344e-05,
-      "loss": 0.2637,
-      "step": 323
-    },
-    {
-      "epoch": 3.857142857142857,
-      "grad_norm": 2.837983814094986,
-      "learning_rate": 1.8645753767428457e-05,
-      "loss": 0.26,
-      "step": 324
-    },
-    {
-      "epoch": 3.869047619047619,
-      "grad_norm": 1.4937224948435774,
-      "learning_rate": 1.863604671814357e-05,
-      "loss": 0.1746,
-      "step": 325
-    },
-    {
-      "epoch": 3.880952380952381,
-      "grad_norm": 2.841367725391816,
-      "learning_rate": 1.8626307549088794e-05,
-      "loss": 0.2785,
-      "step": 326
-    },
-    {
-      "epoch": 3.892857142857143,
-      "grad_norm": 5.964953926450381,
-      "learning_rate": 1.861653629648671e-05,
-      "loss": 0.2995,
-      "step": 327
-    },
-    {
-      "epoch": 3.9047619047619047,
-      "grad_norm": 2.3586129170487817,
-      "learning_rate": 1.8606732996679224e-05,
-      "loss": 0.1655,
-      "step": 328
-    },
-    {
-      "epoch": 3.9166666666666665,
-      "grad_norm": 4.983039463607254,
-      "learning_rate": 1.8596897686127428e-05,
-      "loss": 0.1406,
-      "step": 329
-    },
-    {
-      "epoch": 3.928571428571429,
-      "grad_norm": 2.6513317896940753,
-      "learning_rate": 1.858703040141148e-05,
-      "loss": 0.2406,
-      "step": 330
-    },
-    {
-      "epoch": 3.9404761904761907,
-      "grad_norm": 2.4589121027162415,
-      "learning_rate": 1.8577131179230447e-05,
-      "loss": 0.263,
-      "step": 331
-    },
-    {
-      "epoch": 3.9523809523809526,
-      "grad_norm": 5.9418158438751805,
-      "learning_rate": 1.8567200056402197e-05,
-      "loss": 0.1849,
-      "step": 332
-    },
-    {
-      "epoch": 3.9642857142857144,
-      "grad_norm": 2.717129397334188,
-      "learning_rate": 1.8557237069863224e-05,
-      "loss": 0.1202,
-      "step": 333
-    },
-    {
-      "epoch": 3.9761904761904763,
-      "grad_norm": 1.5063133605816246,
-      "learning_rate": 1.8547242256668546e-05,
-      "loss": 0.2361,
-      "step": 334
-    },
-    {
-      "epoch": 3.988095238095238,
-      "grad_norm": 1.8546470741755403,
-      "learning_rate": 1.8537215653991554e-05,
-      "loss": 0.1449,
-      "step": 335
-    },
-    {
-      "epoch": 4.0,
-      "grad_norm": 1.0304658609012838,
-      "learning_rate": 1.852715729912386e-05,
-      "loss": 0.1486,
-      "step": 336
-    },
-    {
-      "epoch": 4.0,
-      "eval_loss": 0.21311835944652557,
-      "eval_runtime": 38.6432,
-      "eval_samples_per_second": 1.553,
-      "eval_steps_per_second": 1.553,
-      "step": 336
-    },
-    {
-      "epoch": 4.011904761904762,
-      "grad_norm": 2.733522684873606,
-      "learning_rate": 1.8517067229475184e-05,
-      "loss": 0.0409,
-      "step": 337
-    },
-    {
-      "epoch": 4.023809523809524,
-      "grad_norm": 1.8388646319823272,
-      "learning_rate": 1.8506945482573197e-05,
-      "loss": 0.1691,
-      "step": 338
-    },
-    {
-      "epoch": 4.035714285714286,
-      "grad_norm": 1.8958492787854766,
-      "learning_rate": 1.849679209606338e-05,
-      "loss": 0.108,
-      "step": 339
-    },
-    {
-      "epoch": 4.0476190476190474,
-      "grad_norm": 7.540918072075598,
-      "learning_rate": 1.8486607107708902e-05,
-      "loss": 0.1848,
-      "step": 340
-    },
-    {
-      "epoch": 4.059523809523809,
-      "grad_norm": 5.8327310752377155,
-      "learning_rate": 1.8476390555390458e-05,
-      "loss": 0.1757,
-      "step": 341
-    },
-    {
-      "epoch": 4.071428571428571,
-      "grad_norm": 2.37604795133662,
-      "learning_rate": 1.846614247710614e-05,
-      "loss": 0.0432,
-      "step": 342
-    },
-    {
-      "epoch": 4.083333333333333,
-      "grad_norm": 3.956049610570956,
-      "learning_rate": 1.84558629109713e-05,
-      "loss": 0.1275,
-      "step": 343
-    },
-    {
-      "epoch": 4.095238095238095,
-      "grad_norm": 2.2566528797985366,
-      "learning_rate": 1.8445551895218393e-05,
-      "loss": 0.3215,
-      "step": 344
-    },
-    {
-      "epoch": 4.107142857142857,
-      "grad_norm": 5.222667806863405,
-      "learning_rate": 1.843520946819685e-05,
-      "loss": 0.0958,
-      "step": 345
-    },
-    {
-      "epoch": 4.119047619047619,
-      "grad_norm": 1.4292479650404268,
-      "learning_rate": 1.842483566837292e-05,
-      "loss": 0.1721,
-      "step": 346
-    },
-    {
-      "epoch": 4.130952380952381,
-      "grad_norm": 2.0768979787883897,
-      "learning_rate": 1.8414430534329552e-05,
-      "loss": 0.1847,
-      "step": 347
-    },
-    {
-      "epoch": 4.142857142857143,
-      "grad_norm": 2.323248373240265,
-      "learning_rate": 1.8403994104766214e-05,
-      "loss": 0.2216,
-      "step": 348
-    },
-    {
-      "epoch": 4.154761904761905,
-      "grad_norm": 3.092855168358233,
-      "learning_rate": 1.8393526418498786e-05,
-      "loss": 0.3547,
-      "step": 349
-    },
-    {
-      "epoch": 4.166666666666667,
-      "grad_norm": 2.820694205381939,
-      "learning_rate": 1.8383027514459403e-05,
-      "loss": 0.0892,
-      "step": 350
-    },
-    {
-      "epoch": 4.178571428571429,
-      "grad_norm": 5.19609590141203,
-      "learning_rate": 1.8372497431696287e-05,
-      "loss": 0.2335,
-      "step": 351
-    },
-    {
-      "epoch": 4.190476190476191,
-      "grad_norm": 1.8357548516888786,
-      "learning_rate": 1.8361936209373646e-05,
-      "loss": 0.1042,
-      "step": 352
-    },
-    {
-      "epoch": 4.2023809523809526,
-      "grad_norm": 3.143774052079354,
-      "learning_rate": 1.835134388677149e-05,
-      "loss": 0.1141,
-      "step": 353
-    },
-    {
-      "epoch": 4.214285714285714,
-      "grad_norm": 1.3502370439171336,
-      "learning_rate": 1.83407205032855e-05,
-      "loss": 0.1851,
-      "step": 354
-    },
-    {
-      "epoch": 4.226190476190476,
-      "grad_norm": 2.207079287438361,
-      "learning_rate": 1.833006609842688e-05,
-      "loss": 0.1518,
-      "step": 355
-    },
-    {
-      "epoch": 4.238095238095238,
-      "grad_norm": 2.366363737001346,
-      "learning_rate": 1.8319380711822225e-05,
-      "loss": 0.2351,
-      "step": 356
-    },
-    {
-      "epoch": 4.25,
-      "grad_norm": 2.3720004969109003,
-      "learning_rate": 1.8308664383213343e-05,
-      "loss": 0.0997,
-      "step": 357
-    },
-    {
-      "epoch": 4.261904761904762,
-      "grad_norm": 1.6380161132514732,
-      "learning_rate": 1.8297917152457128e-05,
-      "loss": 0.1759,
-      "step": 358
-    },
-    {
-      "epoch": 4.273809523809524,
-      "grad_norm": 3.0271158281302695,
-      "learning_rate": 1.8287139059525413e-05,
-      "loss": 0.1826,
-      "step": 359
-    },
-    {
-      "epoch": 4.285714285714286,
-      "grad_norm": 24.711766628918838,
-      "learning_rate": 1.8276330144504802e-05,
-      "loss": 0.3111,
-      "step": 360
-    },
-    {
-      "epoch": 4.2976190476190474,
-      "grad_norm": 3.8497648128182815,
-      "learning_rate": 1.8265490447596552e-05,
-      "loss": 0.1551,
-      "step": 361
-    },
-    {
-      "epoch": 4.309523809523809,
-      "grad_norm": 1.2126662151093983,
-      "learning_rate": 1.8254620009116396e-05,
-      "loss": 0.1191,
-      "step": 362
-    },
-    {
-      "epoch": 4.321428571428571,
-      "grad_norm": 10.322330853048463,
-      "learning_rate": 1.824371886949441e-05,
-      "loss": 0.1496,
-      "step": 363
-    },
-    {
-      "epoch": 4.333333333333333,
-      "grad_norm": 3.86269481191417,
-      "learning_rate": 1.823278706927484e-05,
-      "loss": 0.0617,
-      "step": 364
-    },
-    {
-      "epoch": 4.345238095238095,
-      "grad_norm": 2.409129687324055,
-      "learning_rate": 1.8221824649115986e-05,
-      "loss": 0.1082,
-      "step": 365
-    },
-    {
-      "epoch": 4.357142857142857,
-      "grad_norm": 6.72802746402091,
-      "learning_rate": 1.8210831649790016e-05,
-      "loss": 0.2257,
-      "step": 366
-    },
-    {
-      "epoch": 4.369047619047619,
-      "grad_norm": 3.286057633238495,
-      "learning_rate": 1.819980811218285e-05,
-      "loss": 0.13,
-      "step": 367
-    },
-    {
-      "epoch": 4.380952380952381,
-      "grad_norm": 2.0442328951186446,
-      "learning_rate": 1.8188754077293964e-05,
-      "loss": 0.0307,
-      "step": 368
-    },
-    {
-      "epoch": 4.392857142857143,
-      "grad_norm": 0.7743849368361936,
-      "learning_rate": 1.8177669586236276e-05,
-      "loss": 0.068,
-      "step": 369
-    },
-    {
-      "epoch": 4.404761904761905,
-      "grad_norm": 1.8828834323895858,
-      "learning_rate": 1.8166554680235982e-05,
-      "loss": 0.079,
-      "step": 370
-    },
-    {
-      "epoch": 4.416666666666667,
-      "grad_norm": 2.035350787727451,
-      "learning_rate": 1.8155409400632388e-05,
-      "loss": 0.1643,
-      "step": 371
-    },
-    {
-      "epoch": 4.428571428571429,
-      "grad_norm": 11.240616104668772,
-      "learning_rate": 1.814423378887777e-05,
-      "loss": 0.254,
-      "step": 372
-    },
-    {
-      "epoch": 4.440476190476191,
-      "grad_norm": 3.3550314941524144,
-      "learning_rate": 1.8133027886537227e-05,
-      "loss": 0.1288,
-      "step": 373
-    },
-    {
-      "epoch": 4.4523809523809526,
-      "grad_norm": 0.8521875859194227,
-      "learning_rate": 1.8121791735288504e-05,
-      "loss": 0.0143,
-      "step": 374
-    },
-    {
-      "epoch": 4.464285714285714,
-      "grad_norm": 0.7346037032427043,
-      "learning_rate": 1.8110525376921863e-05,
-      "loss": 0.0699,
-      "step": 375
-    },
-    {
-      "epoch": 4.476190476190476,
-      "grad_norm": 1.6709740168329212,
-      "learning_rate": 1.80992288533399e-05,
-      "loss": 0.125,
-      "step": 376
-    },
-    {
-      "epoch": 4.488095238095238,
-      "grad_norm": 1.9138466367058902,
-      "learning_rate": 1.8087902206557413e-05,
-      "loss": 0.2358,
-      "step": 377
-    },
-    {
-      "epoch": 4.5,
-      "grad_norm": 0.8814173547123243,
-      "learning_rate": 1.8076545478701235e-05,
-      "loss": 0.0804,
-      "step": 378
-    },
-    {
-      "epoch": 4.511904761904762,
-      "grad_norm": 1.6206474959698678,
-      "learning_rate": 1.8065158712010074e-05,
-      "loss": 0.0219,
-      "step": 379
-    },
-    {
-      "epoch": 4.523809523809524,
-      "grad_norm": 2.3891226185750383,
-      "learning_rate": 1.8053741948834373e-05,
-      "loss": 0.0307,
-      "step": 380
-    },
-    {
-      "epoch": 4.535714285714286,
-      "grad_norm": 7.206521173807237,
-      "learning_rate": 1.8042295231636116e-05,
-      "loss": 0.2891,
-      "step": 381
-    },
-    {
-      "epoch": 4.5476190476190474,
-      "grad_norm": 2.064018339358741,
-      "learning_rate": 1.803081860298872e-05,
-      "loss": 0.193,
-      "step": 382
-    },
-    {
-      "epoch": 4.559523809523809,
-      "grad_norm": 1.0485599754303636,
-      "learning_rate": 1.801931210557684e-05,
-      "loss": 0.0177,
-      "step": 383
-    },
-    {
-      "epoch": 4.571428571428571,
-      "grad_norm": 1.4754707293160358,
-      "learning_rate": 1.8007775782196216e-05,
-      "loss": 0.1202,
-      "step": 384
-    },
-    {
-      "epoch": 4.583333333333333,
-      "grad_norm": 1.9792453212923111,
-      "learning_rate": 1.7996209675753524e-05,
-      "loss": 0.091,
-      "step": 385
-    },
-    {
-      "epoch": 4.595238095238095,
-      "grad_norm": 10.95594851386592,
-      "learning_rate": 1.798461382926621e-05,
-      "loss": 0.2459,
-      "step": 386
-    },
-    {
-      "epoch": 4.607142857142857,
-      "grad_norm": 2.0796924560180905,
-      "learning_rate": 1.7972988285862337e-05,
-      "loss": 0.193,
-      "step": 387
-    },
-    {
-      "epoch": 4.619047619047619,
-      "grad_norm": 1.843392372727349,
-      "learning_rate": 1.7961333088780404e-05,
-      "loss": 0.1797,
-      "step": 388
-    },
-    {
-      "epoch": 4.630952380952381,
-      "grad_norm": 1.5749872984474034,
-      "learning_rate": 1.794964828136922e-05,
-      "loss": 0.135,
-      "step": 389
-    },
-    {
-      "epoch": 4.642857142857143,
-      "grad_norm": 6.267524676104374,
-      "learning_rate": 1.7937933907087702e-05,
-      "loss": 0.1105,
-      "step": 390
-    },
-    {
-      "epoch": 4.654761904761905,
-      "grad_norm": 10.307838193086383,
-      "learning_rate": 1.792619000950475e-05,
-      "loss": 0.2006,
-      "step": 391
-    },
-    {
-      "epoch": 4.666666666666667,
-      "grad_norm": 1.6810655378072146,
-      "learning_rate": 1.7914416632299066e-05,
-      "loss": 0.0731,
-      "step": 392
-    },
-    {
-      "epoch": 4.678571428571429,
-      "grad_norm": 1.60907336690843,
-      "learning_rate": 1.7902613819258983e-05,
-      "loss": 0.1373,
-      "step": 393
-    },
-    {
-      "epoch": 4.690476190476191,
-      "grad_norm": 1.6877915961502108,
-      "learning_rate": 1.7890781614282333e-05,
-      "loss": 0.0464,
-      "step": 394
-    },
-    {
-      "epoch": 4.7023809523809526,
-      "grad_norm": 1.540430941284871,
-      "learning_rate": 1.787892006137625e-05,
-      "loss": 0.1439,
-      "step": 395
-    },
-    {
-      "epoch": 4.714285714285714,
-      "grad_norm": 1.0699166796243846,
-      "learning_rate": 1.786702920465702e-05,
-      "loss": 0.1025,
-      "step": 396
-    },
-    {
-      "epoch": 4.726190476190476,
-      "grad_norm": 3.376295432593385,
-      "learning_rate": 1.7855109088349927e-05,
-      "loss": 0.1607,
-      "step": 397
-    },
-    {
-      "epoch": 4.738095238095238,
-      "grad_norm": 1.4488284754381477,
-      "learning_rate": 1.7843159756789078e-05,
-      "loss": 0.137,
-      "step": 398
-    },
-    {
-      "epoch": 4.75,
-      "grad_norm": 2.3771736790466584,
-      "learning_rate": 1.783118125441723e-05,
-      "loss": 0.0942,
-      "step": 399
-    },
-    {
-      "epoch": 4.761904761904762,
-      "grad_norm": 6.776410312402479,
-      "learning_rate": 1.7819173625785644e-05,
-      "loss": 0.1516,
-      "step": 400
-    },
-    {
-      "epoch": 4.773809523809524,
-      "grad_norm": 1.4340400006613512,
-      "learning_rate": 1.7807136915553904e-05,
-      "loss": 0.108,
-      "step": 401
-    },
-    {
-      "epoch": 4.785714285714286,
-      "grad_norm": 1.160536686716552,
-      "learning_rate": 1.779507116848976e-05,
-      "loss": 0.1759,
-      "step": 402
-    },
-    {
-      "epoch": 4.7976190476190474,
-      "grad_norm": 4.003407777046596,
-      "learning_rate": 1.7782976429468957e-05,
-      "loss": 0.1845,
-      "step": 403
-    },
-    {
-      "epoch": 4.809523809523809,
-      "grad_norm": 1.4582337088270905,
-      "learning_rate": 1.7770852743475067e-05,
-      "loss": 0.1125,
-      "step": 404
-    },
-    {
-      "epoch": 4.821428571428571,
-      "grad_norm": 2.5480560830370718,
-      "learning_rate": 1.775870015559932e-05,
-      "loss": 0.1306,
-      "step": 405
-    },
-    {
-      "epoch": 4.833333333333333,
-      "grad_norm": 7.841245613706642,
-      "learning_rate": 1.7746518711040445e-05,
-      "loss": 0.1568,
-      "step": 406
-    },
-    {
-      "epoch": 4.845238095238095,
-      "grad_norm": 2.036257140830808,
-      "learning_rate": 1.7734308455104496e-05,
-      "loss": 0.1595,
-      "step": 407
-    },
-    {
-      "epoch": 4.857142857142857,
-      "grad_norm": 1.4002028813999992,
-      "learning_rate": 1.7722069433204687e-05,
-      "loss": 0.119,
-      "step": 408
-    },
-    {
-      "epoch": 4.869047619047619,
-      "grad_norm": 2.032627252571408,
-      "learning_rate": 1.7709801690861214e-05,
-      "loss": 0.1404,
-      "step": 409
-    },
-    {
-      "epoch": 4.880952380952381,
-      "grad_norm": 5.533353485198094,
-      "learning_rate": 1.7697505273701097e-05,
-      "loss": 0.1607,
-      "step": 410
-    },
-    {
-      "epoch": 4.892857142857143,
-      "grad_norm": 3.6514331337266337,
-      "learning_rate": 1.7685180227458003e-05,
-      "loss": 0.1803,
-      "step": 411
-    },
-    {
-      "epoch": 4.904761904761905,
-      "grad_norm": 4.680921109039639,
-      "learning_rate": 1.767282659797208e-05,
-      "loss": 0.1033,
-      "step": 412
-    },
-    {
-      "epoch": 4.916666666666667,
-      "grad_norm": 2.621923262817786,
-      "learning_rate": 1.766044443118978e-05,
-      "loss": 0.1073,
-      "step": 413
-    },
-    {
-      "epoch": 4.928571428571429,
-      "grad_norm": 2.0892833469600305,
-      "learning_rate": 1.7648033773163703e-05,
-      "loss": 0.1565,
-      "step": 414
-    },
-    {
-      "epoch": 4.940476190476191,
-      "grad_norm": 1.4629703816596813,
-      "learning_rate": 1.76355946700524e-05,
-      "loss": 0.0821,
-      "step": 415
-    },
-    {
-      "epoch": 4.9523809523809526,
-      "grad_norm": 2.7567759069015794,
-      "learning_rate": 1.7623127168120232e-05,
-      "loss": 0.1481,
-      "step": 416
-    },
-    {
-      "epoch": 4.964285714285714,
-      "grad_norm": 2.8093220612131207,
-      "learning_rate": 1.7610631313737174e-05,
-      "loss": 0.0601,
-      "step": 417
-    },
-    {
-      "epoch": 4.976190476190476,
-      "grad_norm": 1.510038616834676,
-      "learning_rate": 1.7598107153378655e-05,
-      "loss": 0.1432,
-      "step": 418
-    },
-    {
-      "epoch": 4.988095238095238,
-      "grad_norm": 3.6744872821306007,
-      "learning_rate": 1.7585554733625384e-05,
-      "loss": 0.0908,
-      "step": 419
-    },
-    {
-      "epoch": 5.0,
-      "grad_norm": 1.6035869270834606,
-      "learning_rate": 1.7572974101163166e-05,
-      "loss": 0.1099,
-      "step": 420
-    },
-    {
-      "epoch": 5.0,
-      "eval_loss": 0.16849547624588013,
-      "eval_runtime": 39.0956,
-      "eval_samples_per_second": 1.535,
-      "eval_steps_per_second": 1.535,
-      "step": 420
-    },
-    {
-      "epoch": 5.011904761904762,
-      "grad_norm": 2.0984543226784163,
-      "learning_rate": 1.7560365302782738e-05,
-      "loss": 0.1384,
-      "step": 421
-    },
-    {
-      "epoch": 5.023809523809524,
-      "grad_norm": 1.811823585704519,
-      "learning_rate": 1.7547728385379606e-05,
-      "loss": 0.0963,
-      "step": 422
-    },
-    {
-      "epoch": 5.035714285714286,
-      "grad_norm": 1.0199746089219917,
-      "learning_rate": 1.753506339595384e-05,
-      "loss": 0.0982,
-      "step": 423
-    },
-    {
-      "epoch": 5.0476190476190474,
-      "grad_norm": 1.450450087066948,
-      "learning_rate": 1.7522370381609937e-05,
-      "loss": 0.1163,
-      "step": 424
-    },
-    {
-      "epoch": 5.059523809523809,
-      "grad_norm": 2.188317483580504,
-      "learning_rate": 1.7509649389556605e-05,
-      "loss": 0.0727,
-      "step": 425
-    },
-    {
-      "epoch": 5.071428571428571,
-      "grad_norm": 3.81543761344269,
-      "learning_rate": 1.7496900467106627e-05,
-      "loss": 0.1037,
-      "step": 426
-    },
-    {
-      "epoch": 5.083333333333333,
-      "grad_norm": 3.1538637361704414,
-      "learning_rate": 1.7484123661676657e-05,
-      "loss": 0.2181,
-      "step": 427
-    },
-    {
-      "epoch": 5.095238095238095,
-      "grad_norm": 2.2005038350795028,
-      "learning_rate": 1.747131902078705e-05,
-      "loss": 0.1523,
-      "step": 428
-    },
-    {
-      "epoch": 5.107142857142857,
-      "grad_norm": 2.829273286758714,
-      "learning_rate": 1.7458486592061702e-05,
-      "loss": 0.0905,
-      "step": 429
-    },
-    {
-      "epoch": 5.119047619047619,
-      "grad_norm": 2.768932754858575,
-      "learning_rate": 1.7445626423227844e-05,
-      "loss": 0.1626,
-      "step": 430
-    },
-    {
-      "epoch": 5.130952380952381,
-      "grad_norm": 6.3593130187982245,
-      "learning_rate": 1.743273856211589e-05,
-      "loss": 0.1318,
-      "step": 431
-    },
-    {
-      "epoch": 5.142857142857143,
-      "grad_norm": 2.9334873540908104,
-      "learning_rate": 1.7419823056659245e-05,
-      "loss": 0.1621,
-      "step": 432
-    },
-    {
-      "epoch": 5.154761904761905,
-      "grad_norm": 1.5188109000144108,
-      "learning_rate": 1.7406879954894133e-05,
-      "loss": 0.1091,
-      "step": 433
-    },
-    {
-      "epoch": 5.166666666666667,
-      "grad_norm": 6.8222392966891565,
-      "learning_rate": 1.7393909304959414e-05,
-      "loss": 0.1203,
-      "step": 434
-    },
-    {
-      "epoch": 5.178571428571429,
-      "grad_norm": 2.4224667278776963,
-      "learning_rate": 1.738091115509641e-05,
-      "loss": 0.0776,
-      "step": 435
-    },
-    {
-      "epoch": 5.190476190476191,
-      "grad_norm": 2.5737345276650587,
-      "learning_rate": 1.736788555364872e-05,
-      "loss": 0.1427,
-      "step": 436
-    },
-    {
-      "epoch": 5.2023809523809526,
-      "grad_norm": 1.3719252993722968,
-      "learning_rate": 1.7354832549062036e-05,
-      "loss": 0.0647,
-      "step": 437
-    },
-    {
-      "epoch": 5.214285714285714,
-      "grad_norm": 2.3933922727041805,
-      "learning_rate": 1.7341752189883983e-05,
-      "loss": 0.0354,
-      "step": 438
-    },
-    {
-      "epoch": 5.226190476190476,
-      "grad_norm": 1.5221788352824002,
-      "learning_rate": 1.732864452476392e-05,
-      "loss": 0.0865,
-      "step": 439
-    },
-    {
-      "epoch": 5.238095238095238,
-      "grad_norm": 2.063804997908119,
-      "learning_rate": 1.731550960245276e-05,
-      "loss": 0.1055,
-      "step": 440
-    },
-    {
-      "epoch": 5.25,
-      "grad_norm": 1.5372127294983347,
-      "learning_rate": 1.7302347471802797e-05,
-      "loss": 0.0604,
-      "step": 441
-    },
-    {
-      "epoch": 5.261904761904762,
-      "grad_norm": 2.02376227571018,
-      "learning_rate": 1.728915818176752e-05,
-      "loss": 0.0868,
-      "step": 442
-    },
-    {
-      "epoch": 5.273809523809524,
-      "grad_norm": 1.2042690309607718,
-      "learning_rate": 1.7275941781401427e-05,
-      "loss": 0.0806,
-      "step": 443
-    },
-    {
-      "epoch": 5.285714285714286,
-      "grad_norm": 5.282799497759904,
-      "learning_rate": 1.726269831985985e-05,
-      "loss": 0.1004,
-      "step": 444
-    },
-    {
-      "epoch": 5.2976190476190474,
-      "grad_norm": 1.353248586923984,
-      "learning_rate": 1.724942784639877e-05,
-      "loss": 0.0166,
-      "step": 445
-    },
-    {
-      "epoch": 5.309523809523809,
-      "grad_norm": 1.6233751950823214,
-      "learning_rate": 1.7236130410374627e-05,
-      "loss": 0.1031,
-      "step": 446
-    },
-    {
-      "epoch": 5.321428571428571,
-      "grad_norm": 5.300462784341094,
-      "learning_rate": 1.7222806061244148e-05,
-      "loss": 0.0468,
-      "step": 447
-    },
-    {
-      "epoch": 5.333333333333333,
-      "grad_norm": 3.0190761834590174,
-      "learning_rate": 1.7209454848564155e-05,
-      "loss": 0.024,
-      "step": 448
-    },
-    {
-      "epoch": 5.345238095238095,
-      "grad_norm": 0.7706594623111424,
-      "learning_rate": 1.7196076821991384e-05,
-      "loss": 0.06,
-      "step": 449
-    },
-    {
-      "epoch": 5.357142857142857,
-      "grad_norm": 5.781992041615708,
-      "learning_rate": 1.7182672031282296e-05,
-      "loss": 0.1124,
-      "step": 450
-    },
-    {
-      "epoch": 5.369047619047619,
-      "grad_norm": 4.9262566902719325,
-      "learning_rate": 1.7169240526292898e-05,
-      "loss": 0.0687,
-      "step": 451
-    },
-    {
-      "epoch": 5.380952380952381,
-      "grad_norm": 2.44908284406808,
-      "learning_rate": 1.715578235697855e-05,
-      "loss": 0.1093,
-      "step": 452
-    },
-    {
-      "epoch": 5.392857142857143,
-      "grad_norm": 2.1858738956817505,
-      "learning_rate": 1.714229757339379e-05,
-      "loss": 0.0744,
-      "step": 453
-    },
-    {
-      "epoch": 5.404761904761905,
-      "grad_norm": 8.729299168938855,
-      "learning_rate": 1.7128786225692135e-05,
-      "loss": 0.1642,
-      "step": 454
-    },
-    {
-      "epoch": 5.416666666666667,
-      "grad_norm": 2.537620951119198,
-      "learning_rate": 1.7115248364125908e-05,
-      "loss": 0.1596,
-      "step": 455
-    },
-    {
-      "epoch": 5.428571428571429,
-      "grad_norm": 0.9225751240168636,
-      "learning_rate": 1.710168403904604e-05,
-      "loss": 0.0669,
-      "step": 456
-    },
-    {
-      "epoch": 5.440476190476191,
-      "grad_norm": 1.5986892707523244,
-      "learning_rate": 1.7088093300901882e-05,
-      "loss": 0.1076,
-      "step": 457
-    },
-    {
-      "epoch": 5.4523809523809526,
-      "grad_norm": 7.794554323075924,
-      "learning_rate": 1.7074476200241037e-05,
-      "loss": 0.0529,
-      "step": 458
-    },
-    {
-      "epoch": 5.464285714285714,
-      "grad_norm": 1.0068313155262074,
-      "learning_rate": 1.706083278770914e-05,
-      "loss": 0.0633,
-      "step": 459
-    },
-    {
-      "epoch": 5.476190476190476,
-      "grad_norm": 5.203004572929099,
-      "learning_rate": 1.7047163114049704e-05,
-      "loss": 0.1619,
-      "step": 460
-    },
-    {
-      "epoch": 5.488095238095238,
-      "grad_norm": 4.89789141073733,
-      "learning_rate": 1.7033467230103895e-05,
-      "loss": 0.0813,
-      "step": 461
-    },
-    {
-      "epoch": 5.5,
-      "grad_norm": 1.7611120130870677,
-      "learning_rate": 1.7019745186810378e-05,
-      "loss": 0.0284,
-      "step": 462
-    },
-    {
-      "epoch": 5.511904761904762,
-      "grad_norm": 2.8716423718777406,
-      "learning_rate": 1.700599703520511e-05,
-      "loss": 0.107,
-      "step": 463
-    },
-    {
-      "epoch": 5.523809523809524,
-      "grad_norm": 1.1827701240898811,
-      "learning_rate": 1.6992222826421136e-05,
-      "loss": 0.0614,
-      "step": 464
-    },
-    {
-      "epoch": 5.535714285714286,
-      "grad_norm": 2.711674669785039,
-      "learning_rate": 1.697842261168843e-05,
-      "loss": 0.1187,
-      "step": 465
-    },
-    {
-      "epoch": 5.5476190476190474,
-      "grad_norm": 8.047773822350543,
-      "learning_rate": 1.6964596442333696e-05,
-      "loss": 0.1916,
-      "step": 466
-    },
-    {
-      "epoch": 5.559523809523809,
-      "grad_norm": 16.2071619554794,
-      "learning_rate": 1.695074436978015e-05,
-      "loss": 0.1374,
-      "step": 467
-    },
-    {
-      "epoch": 5.571428571428571,
-      "grad_norm": 3.833330687856342,
-      "learning_rate": 1.6936866445547354e-05,
-      "loss": 0.0884,
-      "step": 468
-    },
-    {
-      "epoch": 5.583333333333333,
-      "grad_norm": 2.0568261929174056,
-      "learning_rate": 1.6922962721251038e-05,
-      "loss": 0.1201,
-      "step": 469
-    },
-    {
-      "epoch": 5.595238095238095,
-      "grad_norm": 3.0077364023768043,
-      "learning_rate": 1.6909033248602863e-05,
-      "loss": 0.0437,
-      "step": 470
-    },
-    {
-      "epoch": 5.607142857142857,
-      "grad_norm": 3.218198456214115,
-      "learning_rate": 1.689507807941027e-05,
-      "loss": 0.2151,
-      "step": 471
-    },
-    {
-      "epoch": 5.619047619047619,
-      "grad_norm": 1.3027730243067113,
-      "learning_rate": 1.6881097265576273e-05,
-      "loss": 0.1128,
-      "step": 472
-    },
-    {
-      "epoch": 5.630952380952381,
-      "grad_norm": 1.7258540427814248,
-      "learning_rate": 1.6867090859099256e-05,
-      "loss": 0.0805,
-      "step": 473
-    },
-    {
-      "epoch": 5.642857142857143,
-      "grad_norm": 0.8805724825057599,
-      "learning_rate": 1.68530589120728e-05,
-      "loss": 0.1066,
-      "step": 474
-    },
-    {
-      "epoch": 5.654761904761905,
-      "grad_norm": 1.9113079355582814,
-      "learning_rate": 1.6839001476685472e-05,
-      "loss": 0.1192,
-      "step": 475
-    },
-    {
-      "epoch": 5.666666666666667,
-      "grad_norm": 11.825279511941737,
-      "learning_rate": 1.682491860522063e-05,
-      "loss": 0.3232,
-      "step": 476
-    },
-    {
-      "epoch": 5.678571428571429,
-      "grad_norm": 1.9412063405946438,
-      "learning_rate": 1.681081035005626e-05,
-      "loss": 0.0642,
-      "step": 477
-    },
-    {
-      "epoch": 5.690476190476191,
-      "grad_norm": 1.9416750977216886,
-      "learning_rate": 1.6796676763664724e-05,
-      "loss": 0.0661,
-      "step": 478
-    },
-    {
-      "epoch": 5.7023809523809526,
-      "grad_norm": 6.736681970474298,
-      "learning_rate": 1.678251789861262e-05,
-      "loss": 0.1142,
-      "step": 479
-    },
-    {
-      "epoch": 5.714285714285714,
-      "grad_norm": 8.988950191247852,
-      "learning_rate": 1.676833380756056e-05,
-      "loss": 0.1257,
-      "step": 480
-    },
-    {
-      "epoch": 5.726190476190476,
-      "grad_norm": 2.6627643546016664,
-      "learning_rate": 1.6754124543262973e-05,
-      "loss": 0.0828,
-      "step": 481
-    },
-    {
-      "epoch": 5.738095238095238,
-      "grad_norm": 1.3707408293314916,
-      "learning_rate": 1.6739890158567917e-05,
-      "loss": 0.0432,
-      "step": 482
-    },
-    {
-      "epoch": 5.75,
-      "grad_norm": 1.203361483911428,
-      "learning_rate": 1.6725630706416882e-05,
-      "loss": 0.1027,
-      "step": 483
-    },
-    {
-      "epoch": 5.761904761904762,
-      "grad_norm": 1.1581274009447258,
-      "learning_rate": 1.671134623984459e-05,
-      "loss": 0.0586,
-      "step": 484
-    },
-    {
-      "epoch": 5.773809523809524,
-      "grad_norm": 2.7846087083215942,
-      "learning_rate": 1.6697036811978785e-05,
-      "loss": 0.0946,
-      "step": 485
-    },
-    {
-      "epoch": 5.785714285714286,
-      "grad_norm": 1.0055933651335667,
-      "learning_rate": 1.668270247604008e-05,
-      "loss": 0.0447,
-      "step": 486
-    },
-    {
-      "epoch": 5.7976190476190474,
-      "grad_norm": 8.469500559608552,
-      "learning_rate": 1.666834328534169e-05,
-      "loss": 0.1275,
-      "step": 487
-    },
-    {
-      "epoch": 5.809523809523809,
-      "grad_norm": 9.65420791197495,
-      "learning_rate": 1.6653959293289296e-05,
-      "loss": 0.1675,
-      "step": 488
-    },
-    {
-      "epoch": 5.821428571428571,
-      "grad_norm": 7.4143605449900605,
-      "learning_rate": 1.663955055338082e-05,
-      "loss": 0.175,
-      "step": 489
-    },
-    {
-      "epoch": 5.833333333333333,
-      "grad_norm": 3.2154530706581355,
-      "learning_rate": 1.6625117119206217e-05,
-      "loss": 0.0419,
-      "step": 490
-    },
-    {
-      "epoch": 5.845238095238095,
-      "grad_norm": 1.1263482849597752,
-      "learning_rate": 1.6610659044447296e-05,
-      "loss": 0.0443,
-      "step": 491
-    },
-    {
-      "epoch": 5.857142857142857,
-      "grad_norm": 3.0300211714461702,
-      "learning_rate": 1.6596176382877506e-05,
-      "loss": 0.1838,
-      "step": 492
-    },
-    {
-      "epoch": 5.869047619047619,
-      "grad_norm": 0.66866959793733,
-      "learning_rate": 1.658166918836175e-05,
-      "loss": 0.0541,
-      "step": 493
-    },
-    {
-      "epoch": 5.880952380952381,
-      "grad_norm": 4.3346266946381045,
-      "learning_rate": 1.6567137514856157e-05,
-      "loss": 0.1322,
-      "step": 494
-    },
-    {
-      "epoch": 5.892857142857143,
-      "grad_norm": 2.134678683799703,
-      "learning_rate": 1.655258141640792e-05,
-      "loss": 0.0875,
-      "step": 495
-    },
-    {
-      "epoch": 5.904761904761905,
-      "grad_norm": 6.126638160009494,
-      "learning_rate": 1.653800094715506e-05,
-      "loss": 0.0584,
-      "step": 496
-    },
-    {
-      "epoch": 5.916666666666667,
-      "grad_norm": 0.7344309410677271,
-      "learning_rate": 1.652339616132625e-05,
-      "loss": 0.0467,
-      "step": 497
-    },
-    {
-      "epoch": 5.928571428571429,
-      "grad_norm": 1.793606831473593,
-      "learning_rate": 1.65087671132406e-05,
-      "loss": 0.1195,
-      "step": 498
-    },
-    {
-      "epoch": 5.940476190476191,
-      "grad_norm": 4.880270027679669,
-      "learning_rate": 1.6494113857307454e-05,
-      "loss": 0.1688,
-      "step": 499
-    },
-    {
-      "epoch": 5.9523809523809526,
-      "grad_norm": 1.700000856351666,
-      "learning_rate": 1.6479436448026197e-05,
-      "loss": 0.1244,
-      "step": 500
-    },
-    {
-      "epoch": 5.964285714285714,
-      "grad_norm": 4.08041629571301,
-      "learning_rate": 1.646473493998604e-05,
-      "loss": 0.1081,
-      "step": 501
-    },
-    {
-      "epoch": 5.976190476190476,
-      "grad_norm": 1.4491524448140567,
-      "learning_rate": 1.645000938786582e-05,
-      "loss": 0.1233,
-      "step": 502
-    },
-    {
-      "epoch": 5.988095238095238,
-      "grad_norm": 0.9253565946532154,
-      "learning_rate": 1.6435259846433823e-05,
-      "loss": 0.0225,
-      "step": 503
-    },
-    {
-      "epoch": 6.0,
-      "grad_norm": 2.772680062939085,
-      "learning_rate": 1.6420486370547537e-05,
-      "loss": 0.1832,
-      "step": 504
-    },
-    {
-      "epoch": 6.0,
-      "eval_loss": 0.14325179159641266,
-      "eval_runtime": 38.6443,
-      "eval_samples_per_second": 1.553,
-      "eval_steps_per_second": 1.553,
-      "step": 504
-    },
-    {
-      "epoch": 6.011904761904762,
-      "grad_norm": 1.5019024766855187,
-      "learning_rate": 1.640568901515347e-05,
-      "loss": 0.106,
-      "step": 505
-    },
-    {
-      "epoch": 6.023809523809524,
-      "grad_norm": 1.195423816720282,
-      "learning_rate": 1.6390867835286954e-05,
-      "loss": 0.0226,
-      "step": 506
-    },
-    {
-      "epoch": 6.035714285714286,
-      "grad_norm": 2.0934313411664895,
-      "learning_rate": 1.6376022886071924e-05,
-      "loss": 0.1394,
-      "step": 507
-    },
-    {
-      "epoch": 6.0476190476190474,
-      "grad_norm": 0.9134568702424617,
-      "learning_rate": 1.6361154222720713e-05,
-      "loss": 0.0738,
-      "step": 508
-    },
-    {
-      "epoch": 6.059523809523809,
-      "grad_norm": 1.827946060377346,
-      "learning_rate": 1.634626190053387e-05,
-      "loss": 0.1004,
-      "step": 509
-    },
-    {
-      "epoch": 6.071428571428571,
-      "grad_norm": 4.291381510622088,
-      "learning_rate": 1.6331345974899923e-05,
-      "loss": 0.1089,
-      "step": 510
-    },
-    {
-      "epoch": 6.083333333333333,
-      "grad_norm": 1.793625854996647,
-      "learning_rate": 1.6316406501295198e-05,
-      "loss": 0.1218,
-      "step": 511
-    },
-    {
-      "epoch": 6.095238095238095,
-      "grad_norm": 2.706232219127682,
-      "learning_rate": 1.6301443535283593e-05,
-      "loss": 0.0865,
-      "step": 512
-    },
-    {
-      "epoch": 6.107142857142857,
-      "grad_norm": 3.43227842671449,
-      "learning_rate": 1.6286457132516383e-05,
-      "loss": 0.069,
-      "step": 513
-    },
-    {
-      "epoch": 6.119047619047619,
-      "grad_norm": 2.0565239329994323,
-      "learning_rate": 1.6271447348732023e-05,
-      "loss": 0.1169,
-      "step": 514
-    },
-    {
-      "epoch": 6.130952380952381,
-      "grad_norm": 4.730627764396558,
-      "learning_rate": 1.6256414239755902e-05,
-      "loss": 0.0821,
-      "step": 515
-    },
-    {
-      "epoch": 6.142857142857143,
-      "grad_norm": 3.504993607204566,
-      "learning_rate": 1.6241357861500184e-05,
-      "loss": 0.0732,
-      "step": 516
-    },
-    {
-      "epoch": 6.154761904761905,
-      "grad_norm": 1.1924649747006983,
-      "learning_rate": 1.6226278269963577e-05,
-      "loss": 0.097,
-      "step": 517
-    },
-    {
-      "epoch": 6.166666666666667,
-      "grad_norm": 1.1713451235970245,
-      "learning_rate": 1.621117552123111e-05,
-      "loss": 0.0721,
-      "step": 518
-    },
-    {
-      "epoch": 6.178571428571429,
-      "grad_norm": 1.278243637158708,
-      "learning_rate": 1.6196049671473954e-05,
-      "loss": 0.0847,
-      "step": 519
-    },
-    {
-      "epoch": 6.190476190476191,
-      "grad_norm": 1.7037872290115583,
-      "learning_rate": 1.618090077694919e-05,
-      "loss": 0.0646,
-      "step": 520
-    },
-    {
-      "epoch": 6.2023809523809526,
-      "grad_norm": 1.2759918767604022,
-      "learning_rate": 1.6165728893999616e-05,
-      "loss": 0.1003,
-      "step": 521
-    },
-    {
-      "epoch": 6.214285714285714,
-      "grad_norm": 1.5153242877488946,
-      "learning_rate": 1.6150534079053528e-05,
-      "loss": 0.0729,
-      "step": 522
-    },
-    {
-      "epoch": 6.226190476190476,
-      "grad_norm": 2.9422960437536996,
-      "learning_rate": 1.6135316388624508e-05,
-      "loss": 0.1581,
-      "step": 523
-    },
-    {
-      "epoch": 6.238095238095238,
-      "grad_norm": 0.9976260526302768,
-      "learning_rate": 1.612007587931122e-05,
-      "loss": 0.0425,
-      "step": 524
-    },
-    {
-      "epoch": 6.25,
-      "grad_norm": 1.917033625841123,
-      "learning_rate": 1.6104812607797204e-05,
-      "loss": 0.0811,
-      "step": 525
-    },
-    {
-      "epoch": 6.261904761904762,
-      "grad_norm": 1.6978119095102044,
-      "learning_rate": 1.6089526630850644e-05,
-      "loss": 0.1054,
-      "step": 526
-    },
-    {
-      "epoch": 6.273809523809524,
-      "grad_norm": 2.459854386034363,
-      "learning_rate": 1.607421800532419e-05,
-      "loss": 0.079,
-      "step": 527
-    },
-    {
-      "epoch": 6.285714285714286,
-      "grad_norm": 1.7906992011251934,
-      "learning_rate": 1.6058886788154713e-05,
-      "loss": 0.0679,
-      "step": 528
-    },
-    {
-      "epoch": 6.2976190476190474,
-      "grad_norm": 1.051587482130984,
-      "learning_rate": 1.6043533036363114e-05,
-      "loss": 0.0465,
-      "step": 529
-    },
-    {
-      "epoch": 6.309523809523809,
-      "grad_norm": 1.8489956604650926,
-      "learning_rate": 1.6028156807054113e-05,
-      "loss": 0.0602,
-      "step": 530
-    },
-    {
-      "epoch": 6.321428571428571,
-      "grad_norm": 1.365616441791178,
-      "learning_rate": 1.601275815741602e-05,
-      "loss": 0.0995,
-      "step": 531
-    },
-    {
-      "epoch": 6.333333333333333,
-      "grad_norm": 1.4026420836250593,
-      "learning_rate": 1.5997337144720534e-05,
-      "loss": 0.0887,
-      "step": 532
-    },
-    {
-      "epoch": 6.345238095238095,
-      "grad_norm": 1.7760000515082845,
-      "learning_rate": 1.598189382632253e-05,
-      "loss": 0.0714,
-      "step": 533
-    },
-    {
-      "epoch": 6.357142857142857,
-      "grad_norm": 3.151403903780976,
-      "learning_rate": 1.5966428259659844e-05,
-      "loss": 0.0232,
-      "step": 534
-    },
-    {
-      "epoch": 6.369047619047619,
-      "grad_norm": 2.0134779392823083,
-      "learning_rate": 1.5950940502253064e-05,
-      "loss": 0.12,
-      "step": 535
-    },
-    {
-      "epoch": 6.380952380952381,
-      "grad_norm": 1.3996798101959818,
-      "learning_rate": 1.59354306117053e-05,
-      "loss": 0.0676,
-      "step": 536
-    },
-    {
-      "epoch": 6.392857142857143,
-      "grad_norm": 0.983271468293818,
-      "learning_rate": 1.591989864570199e-05,
-      "loss": 0.0469,
-      "step": 537
-    },
-    {
-      "epoch": 6.404761904761905,
-      "grad_norm": 1.5159705113486506,
-      "learning_rate": 1.5904344662010672e-05,
-      "loss": 0.0971,
-      "step": 538
-    },
-    {
-      "epoch": 6.416666666666667,
-      "grad_norm": 1.8392738688203814,
-      "learning_rate": 1.5888768718480778e-05,
-      "loss": 0.0924,
-      "step": 539
-    },
-    {
-      "epoch": 6.428571428571429,
-      "grad_norm": 2.4551157718317986,
-      "learning_rate": 1.5873170873043412e-05,
-      "loss": 0.0658,
-      "step": 540
-    },
-    {
-      "epoch": 6.440476190476191,
-      "grad_norm": 1.5886407912063516,
-      "learning_rate": 1.5857551183711138e-05,
-      "loss": 0.0761,
-      "step": 541
-    },
-    {
-      "epoch": 6.4523809523809526,
-      "grad_norm": 1.4026787489563222,
-      "learning_rate": 1.584190970857776e-05,
-      "loss": 0.0649,
-      "step": 542
-    },
-    {
-      "epoch": 6.464285714285714,
-      "grad_norm": 1.1906948905495682,
-      "learning_rate": 1.582624650581811e-05,
-      "loss": 0.0768,
-      "step": 543
-    },
-    {
-      "epoch": 6.476190476190476,
-      "grad_norm": 1.7857819999749693,
-      "learning_rate": 1.5810561633687842e-05,
-      "loss": 0.0269,
-      "step": 544
-    },
-    {
-      "epoch": 6.488095238095238,
-      "grad_norm": 0.8665778712262205,
-      "learning_rate": 1.5794855150523183e-05,
-      "loss": 0.0427,
-      "step": 545
-    },
-    {
-      "epoch": 6.5,
-      "grad_norm": 0.9975530177048364,
-      "learning_rate": 1.5779127114740757e-05,
-      "loss": 0.033,
-      "step": 546
-    },
-    {
-      "epoch": 6.511904761904762,
-      "grad_norm": 1.3394274426999186,
-      "learning_rate": 1.5763377584837337e-05,
-      "loss": 0.0427,
-      "step": 547
-    },
-    {
-      "epoch": 6.523809523809524,
-      "grad_norm": 1.4827725196076444,
-      "learning_rate": 1.5747606619389642e-05,
-      "loss": 0.019,
-      "step": 548
-    },
-    {
-      "epoch": 6.535714285714286,
-      "grad_norm": 1.2995548812490796,
-      "learning_rate": 1.5731814277054112e-05,
-      "loss": 0.09,
-      "step": 549
-    },
-    {
-      "epoch": 6.5476190476190474,
-      "grad_norm": 0.8657770498669943,
-      "learning_rate": 1.57160006165667e-05,
-      "loss": 0.0163,
-      "step": 550
-    },
-    {
-      "epoch": 6.559523809523809,
-      "grad_norm": 0.7283838543975072,
-      "learning_rate": 1.570016569674264e-05,
-      "loss": 0.0568,
-      "step": 551
-    },
-    {
-      "epoch": 6.571428571428571,
-      "grad_norm": 1.470782870334043,
-      "learning_rate": 1.5684309576476248e-05,
-      "loss": 0.0925,
-      "step": 552
-    },
-    {
-      "epoch": 6.583333333333333,
-      "grad_norm": 1.9078912192128594,
-      "learning_rate": 1.5668432314740663e-05,
-      "loss": 0.0939,
-      "step": 553
-    },
-    {
-      "epoch": 6.595238095238095,
-      "grad_norm": 1.4741198916158764,
-      "learning_rate": 1.5652533970587688e-05,
-      "loss": 0.0621,
-      "step": 554
-    },
-    {
-      "epoch": 6.607142857142857,
-      "grad_norm": 0.8725669529266904,
-      "learning_rate": 1.5636614603147513e-05,
-      "loss": 0.0544,
-      "step": 555
-    },
-    {
-      "epoch": 6.619047619047619,
-      "grad_norm": 0.7432776800153558,
-      "learning_rate": 1.562067427162853e-05,
-      "loss": 0.0564,
-      "step": 556
-    },
-    {
-      "epoch": 6.630952380952381,
-      "grad_norm": 1.4317069015929436,
-      "learning_rate": 1.56047130353171e-05,
-      "loss": 0.0167,
-      "step": 557
-    },
-    {
-      "epoch": 6.642857142857143,
-      "grad_norm": 0.8325527210494423,
-      "learning_rate": 1.5588730953577336e-05,
-      "loss": 0.0402,
-      "step": 558
-    },
-    {
-      "epoch": 6.654761904761905,
-      "grad_norm": 3.2233010428483793,
-      "learning_rate": 1.5572728085850873e-05,
-      "loss": 0.1871,
-      "step": 559
-    },
-    {
-      "epoch": 6.666666666666667,
-      "grad_norm": 1.1007119123597107,
-      "learning_rate": 1.5556704491656666e-05,
-      "loss": 0.038,
-      "step": 560
-    },
-    {
-      "epoch": 6.678571428571429,
-      "grad_norm": 0.3049436586601034,
-      "learning_rate": 1.554066023059075e-05,
-      "loss": 0.0043,
-      "step": 561
-    },
-    {
-      "epoch": 6.690476190476191,
-      "grad_norm": 0.6116647914325464,
-      "learning_rate": 1.5524595362326028e-05,
-      "loss": 0.0448,
-      "step": 562
-    },
-    {
-      "epoch": 6.7023809523809526,
-      "grad_norm": 2.566952968870286,
-      "learning_rate": 1.5508509946612045e-05,
-      "loss": 0.1348,
-      "step": 563
-    },
-    {
-      "epoch": 6.714285714285714,
-      "grad_norm": 1.626464079459653,
-      "learning_rate": 1.549240404327477e-05,
-      "loss": 0.0739,
-      "step": 564
-    },
-    {
-      "epoch": 6.726190476190476,
-      "grad_norm": 1.301521334500284,
-      "learning_rate": 1.5476277712216364e-05,
-      "loss": 0.0583,
-      "step": 565
-    },
-    {
-      "epoch": 6.738095238095238,
-      "grad_norm": 0.8296339701668457,
-      "learning_rate": 1.546013101341498e-05,
-      "loss": 0.0125,
-      "step": 566
-    },
-    {
-      "epoch": 6.75,
-      "grad_norm": 1.9162341614965561,
-      "learning_rate": 1.544396400692451e-05,
-      "loss": 0.0764,
-      "step": 567
-    },
-    {
-      "epoch": 6.761904761904762,
-      "grad_norm": 0.8294531297211747,
-      "learning_rate": 1.5427776752874372e-05,
-      "loss": 0.0527,
-      "step": 568
-    },
-    {
-      "epoch": 6.773809523809524,
-      "grad_norm": 1.637333814661965,
-      "learning_rate": 1.5411569311469307e-05,
-      "loss": 0.0701,
-      "step": 569
-    },
-    {
-      "epoch": 6.785714285714286,
-      "grad_norm": 0.7654549021641782,
-      "learning_rate": 1.5395341742989126e-05,
-      "loss": 0.0392,
-      "step": 570
-    },
-    {
-      "epoch": 6.7976190476190474,
-      "grad_norm": 4.113607635248581,
-      "learning_rate": 1.53790941077885e-05,
-      "loss": 0.1606,
-      "step": 571
-    },
-    {
-      "epoch": 6.809523809523809,
-      "grad_norm": 2.894466788003413,
-      "learning_rate": 1.5362826466296736e-05,
-      "loss": 0.1243,
-      "step": 572
-    },
-    {
-      "epoch": 6.821428571428571,
-      "grad_norm": 1.681756853880995,
-      "learning_rate": 1.534653887901754e-05,
-      "loss": 0.0906,
-      "step": 573
-    },
-    {
-      "epoch": 6.833333333333333,
-      "grad_norm": 7.226542543292303,
-      "learning_rate": 1.533023140652882e-05,
-      "loss": 0.0968,
-      "step": 574
-    },
-    {
-      "epoch": 6.845238095238095,
-      "grad_norm": 2.7169022930643427,
-      "learning_rate": 1.531390410948243e-05,
-      "loss": 0.0198,
-      "step": 575
-    },
-    {
-      "epoch": 6.857142857142857,
-      "grad_norm": 5.199337298379754,
-      "learning_rate": 1.5297557048603963e-05,
-      "loss": 0.0724,
-      "step": 576
-    },
-    {
-      "epoch": 6.869047619047619,
-      "grad_norm": 1.4573244826016143,
-      "learning_rate": 1.52811902846925e-05,
-      "loss": 0.0829,
-      "step": 577
-    },
-    {
-      "epoch": 6.880952380952381,
-      "grad_norm": 1.8919827827452864,
-      "learning_rate": 1.526480387862043e-05,
-      "loss": 0.086,
-      "step": 578
-    },
-    {
-      "epoch": 6.892857142857143,
-      "grad_norm": 2.8140927699375267,
-      "learning_rate": 1.5248397891333183e-05,
-      "loss": 0.1588,
-      "step": 579
-    },
-    {
-      "epoch": 6.904761904761905,
-      "grad_norm": 2.0967385770460782,
-      "learning_rate": 1.5231972383849019e-05,
-      "loss": 0.061,
-      "step": 580
-    },
-    {
-      "epoch": 6.916666666666667,
-      "grad_norm": 2.8660947565892383,
-      "learning_rate": 1.5215527417258793e-05,
-      "loss": 0.1155,
-      "step": 581
-    },
-    {
-      "epoch": 6.928571428571429,
-      "grad_norm": 1.2394198268814334,
-      "learning_rate": 1.5199063052725746e-05,
-      "loss": 0.047,
-      "step": 582
-    },
-    {
-      "epoch": 6.940476190476191,
-      "grad_norm": 1.7366594536677353,
-      "learning_rate": 1.518257935148525e-05,
-      "loss": 0.1103,
-      "step": 583
-    },
-    {
-      "epoch": 6.9523809523809526,
-      "grad_norm": 1.4905106038915912,
-      "learning_rate": 1.5166076374844605e-05,
-      "loss": 0.1084,
-      "step": 584
-    },
-    {
-      "epoch": 6.964285714285714,
-      "grad_norm": 1.7466305557853612,
-      "learning_rate": 1.5149554184182802e-05,
-      "loss": 0.0601,
-      "step": 585
-    },
-    {
-      "epoch": 6.976190476190476,
-      "grad_norm": 2.2407234722112332,
-      "learning_rate": 1.5133012840950292e-05,
-      "loss": 0.1101,
-      "step": 586
-    },
-    {
-      "epoch": 6.988095238095238,
-      "grad_norm": 0.9285094911344285,
-      "learning_rate": 1.5116452406668758e-05,
-      "loss": 0.0419,
-      "step": 587
-    },
-    {
-      "epoch": 7.0,
-      "grad_norm": 1.836337641554133,
-      "learning_rate": 1.5099872942930886e-05,
-      "loss": 0.0258,
-      "step": 588
-    },
-    {
-      "epoch": 7.0,
-      "eval_loss": 0.14865067601203918,
-      "eval_runtime": 43.1081,
-      "eval_samples_per_second": 1.392,
-      "eval_steps_per_second": 1.392,
-      "step": 588
-    },
-    {
-      "epoch": 7.011904761904762,
-      "grad_norm": 1.2568935563245456,
-      "learning_rate": 1.5083274511400144e-05,
-      "loss": 0.0558,
-      "step": 589
-    },
-    {
-      "epoch": 7.023809523809524,
-      "grad_norm": 6.801089494891712,
-      "learning_rate": 1.5066657173810542e-05,
-      "loss": 0.1197,
-      "step": 590
-    },
-    {
-      "epoch": 7.035714285714286,
-      "grad_norm": 5.569312791725525,
-      "learning_rate": 1.5050020991966405e-05,
-      "loss": 0.0738,
-      "step": 591
-    },
-    {
-      "epoch": 7.0476190476190474,
-      "grad_norm": 2.759979048006171,
-      "learning_rate": 1.5033366027742156e-05,
-      "loss": 0.0367,
-      "step": 592
-    },
-    {
-      "epoch": 7.059523809523809,
-      "grad_norm": 1.6545276645155973,
-      "learning_rate": 1.5016692343082053e-05,
-      "loss": 0.0814,
-      "step": 593
-    },
-    {
-      "epoch": 7.071428571428571,
-      "grad_norm": 2.1169578211888727,
-      "learning_rate": 1.5000000000000002e-05,
-      "loss": 0.0903,
-      "step": 594
-    },
-    {
-      "epoch": 7.083333333333333,
-      "grad_norm": 1.829536723099167,
-      "learning_rate": 1.4983289060579294e-05,
-      "loss": 0.0679,
-      "step": 595
-    },
-    {
-      "epoch": 7.095238095238095,
-      "grad_norm": 2.063531967019674,
-      "learning_rate": 1.4966559586972387e-05,
-      "loss": 0.0285,
-      "step": 596
-    },
-    {
-      "epoch": 7.107142857142857,
-      "grad_norm": 0.9923944509007061,
-      "learning_rate": 1.4949811641400668e-05,
-      "loss": 0.0364,
-      "step": 597
-    },
-    {
-      "epoch": 7.119047619047619,
-      "grad_norm": 0.9674661161779017,
-      "learning_rate": 1.4933045286154242e-05,
-      "loss": 0.0599,
-      "step": 598
-    },
-    {
-      "epoch": 7.130952380952381,
-      "grad_norm": 2.761259352030199,
-      "learning_rate": 1.4916260583591659e-05,
-      "loss": 0.0284,
-      "step": 599
-    },
-    {
-      "epoch": 7.142857142857143,
-      "grad_norm": 3.4877717256597296,
-      "learning_rate": 1.4899457596139728e-05,
-      "loss": 0.1466,
-      "step": 600
-    },
-    {
-      "epoch": 7.154761904761905,
-      "grad_norm": 1.8885548051230485,
-      "learning_rate": 1.488263638629326e-05,
-      "loss": 0.0611,
-      "step": 601
-    },
-    {
-      "epoch": 7.166666666666667,
-      "grad_norm": 3.5085551469483733,
-      "learning_rate": 1.4865797016614839e-05,
-      "loss": 0.0335,
-      "step": 602
-    },
-    {
-      "epoch": 7.178571428571429,
-      "grad_norm": 2.1317182396352514,
-      "learning_rate": 1.4848939549734582e-05,
-      "loss": 0.1127,
-      "step": 603
-    },
-    {
-      "epoch": 7.190476190476191,
-      "grad_norm": 2.4596429083047906,
-      "learning_rate": 1.4832064048349928e-05,
-      "loss": 0.0533,
-      "step": 604
-    },
-    {
-      "epoch": 7.2023809523809526,
-      "grad_norm": 4.555208354333248,
-      "learning_rate": 1.4815170575225383e-05,
-      "loss": 0.0496,
-      "step": 605
-    },
-    {
-      "epoch": 7.214285714285714,
-      "grad_norm": 2.557372985143864,
-      "learning_rate": 1.4798259193192298e-05,
-      "loss": 0.1063,
-      "step": 606
-    },
-    {
-      "epoch": 7.226190476190476,
-      "grad_norm": 1.949653943995653,
-      "learning_rate": 1.4781329965148625e-05,
-      "loss": 0.0531,
-      "step": 607
-    },
-    {
-      "epoch": 7.238095238095238,
-      "grad_norm": 1.935893865742182,
-      "learning_rate": 1.47643829540587e-05,
-      "loss": 0.0963,
-      "step": 608
-    },
-    {
-      "epoch": 7.25,
-      "grad_norm": 1.7142332761042414,
-      "learning_rate": 1.4747418222952997e-05,
-      "loss": 0.0815,
-      "step": 609
-    },
-    {
-      "epoch": 7.261904761904762,
-      "grad_norm": 1.8054607129649187,
-      "learning_rate": 1.4730435834927885e-05,
-      "loss": 0.0824,
-      "step": 610
-    },
-    {
-      "epoch": 7.273809523809524,
-      "grad_norm": 1.3720448261221745,
-      "learning_rate": 1.471343585314542e-05,
-      "loss": 0.0772,
-      "step": 611
-    },
-    {
-      "epoch": 7.285714285714286,
-      "grad_norm": 1.565400646280693,
-      "learning_rate": 1.4696418340833081e-05,
-      "loss": 0.0704,
-      "step": 612
-    },
-    {
-      "epoch": 7.2976190476190474,
-      "grad_norm": 1.2730729423352545,
-      "learning_rate": 1.4679383361283554e-05,
-      "loss": 0.0519,
-      "step": 613
-    },
-    {
-      "epoch": 7.309523809523809,
-      "grad_norm": 2.1497377325557685,
-      "learning_rate": 1.4662330977854488e-05,
-      "loss": 0.0929,
-      "step": 614
-    },
-    {
-      "epoch": 7.321428571428571,
-      "grad_norm": 1.2135536719810307,
-      "learning_rate": 1.4645261253968262e-05,
-      "loss": 0.0136,
-      "step": 615
-    },
-    {
-      "epoch": 7.333333333333333,
-      "grad_norm": 1.9076749429523272,
-      "learning_rate": 1.4628174253111751e-05,
-      "loss": 0.0589,
-      "step": 616
-    },
-    {
-      "epoch": 7.345238095238095,
-      "grad_norm": 1.865336445497231,
-      "learning_rate": 1.4611070038836084e-05,
-      "loss": 0.1012,
-      "step": 617
-    },
-    {
-      "epoch": 7.357142857142857,
-      "grad_norm": 1.4531772649657178,
-      "learning_rate": 1.4593948674756418e-05,
-      "loss": 0.0185,
-      "step": 618
-    },
-    {
-      "epoch": 7.369047619047619,
-      "grad_norm": 1.3967163423107203,
-      "learning_rate": 1.4576810224551683e-05,
-      "loss": 0.0679,
-      "step": 619
-    },
-    {
-      "epoch": 7.380952380952381,
-      "grad_norm": 1.2127139438665828,
-      "learning_rate": 1.4559654751964364e-05,
-      "loss": 0.0585,
-      "step": 620
-    },
-    {
-      "epoch": 7.392857142857143,
-      "grad_norm": 0.8916239980755017,
-      "learning_rate": 1.4542482320800264e-05,
-      "loss": 0.0366,
-      "step": 621
-    },
-    {
-      "epoch": 7.404761904761905,
-      "grad_norm": 0.9741419411575285,
-      "learning_rate": 1.4525292994928247e-05,
-      "loss": 0.0559,
-      "step": 622
-    },
-    {
-      "epoch": 7.416666666666667,
-      "grad_norm": 1.0000739028449486,
-      "learning_rate": 1.4508086838280018e-05,
-      "loss": 0.0398,
-      "step": 623
-    },
-    {
-      "epoch": 7.428571428571429,
-      "grad_norm": 1.8740728890384972,
-      "learning_rate": 1.449086391484988e-05,
-      "loss": 0.0778,
-      "step": 624
-    },
-    {
-      "epoch": 7.440476190476191,
-      "grad_norm": 1.7375565170529184,
-      "learning_rate": 1.4473624288694499e-05,
-      "loss": 0.0827,
-      "step": 625
-    },
-    {
-      "epoch": 7.4523809523809526,
-      "grad_norm": 2.0591771711864473,
-      "learning_rate": 1.4456368023932657e-05,
-      "loss": 0.0797,
-      "step": 626
-    },
-    {
-      "epoch": 7.464285714285714,
-      "grad_norm": 3.089641917565682,
-      "learning_rate": 1.4439095184745025e-05,
-      "loss": 0.14,
-      "step": 627
-    },
-    {
-      "epoch": 7.476190476190476,
-      "grad_norm": 2.1378985179166894,
-      "learning_rate": 1.4421805835373917e-05,
-      "loss": 0.0224,
-      "step": 628
-    },
-    {
-      "epoch": 7.488095238095238,
-      "grad_norm": 2.8192981571711293,
-      "learning_rate": 1.4404500040123049e-05,
-      "loss": 0.1095,
-      "step": 629
-    },
-    {
-      "epoch": 7.5,
-      "grad_norm": 1.9613857654437483,
-      "learning_rate": 1.4387177863357307e-05,
-      "loss": 0.0938,
-      "step": 630
-    },
-    {
-      "epoch": 7.511904761904762,
-      "grad_norm": 1.3072569940934167,
-      "learning_rate": 1.4369839369502507e-05,
-      "loss": 0.0561,
-      "step": 631
-    },
-    {
-      "epoch": 7.523809523809524,
-      "grad_norm": 4.300152949971368,
-      "learning_rate": 1.4352484623045149e-05,
-      "loss": 0.0399,
-      "step": 632
-    },
-    {
-      "epoch": 7.535714285714286,
-      "grad_norm": 1.4294772622134482,
-      "learning_rate": 1.4335113688532183e-05,
-      "loss": 0.0449,
-      "step": 633
-    },
-    {
-      "epoch": 7.5476190476190474,
-      "grad_norm": 2.483155533373781,
-      "learning_rate": 1.431772663057076e-05,
-      "loss": 0.1037,
-      "step": 634
-    },
-    {
-      "epoch": 7.559523809523809,
-      "grad_norm": 1.2006274544896178,
-      "learning_rate": 1.4300323513828008e-05,
-      "loss": 0.04,
-      "step": 635
-    },
-    {
-      "epoch": 7.571428571428571,
-      "grad_norm": 1.3950943162774636,
-      "learning_rate": 1.4282904403030773e-05,
-      "loss": 0.0362,
-      "step": 636
-    },
-    {
-      "epoch": 7.583333333333333,
-      "grad_norm": 1.8558101279081418,
-      "learning_rate": 1.4265469362965398e-05,
-      "loss": 0.0518,
-      "step": 637
-    },
-    {
-      "epoch": 7.595238095238095,
-      "grad_norm": 1.6872639221568726,
-      "learning_rate": 1.4248018458477464e-05,
-      "loss": 0.064,
-      "step": 638
-    },
-    {
-      "epoch": 7.607142857142857,
-      "grad_norm": 1.6544975886322297,
-      "learning_rate": 1.4230551754471553e-05,
-      "loss": 0.0561,
-      "step": 639
-    },
-    {
-      "epoch": 7.619047619047619,
-      "grad_norm": 2.6900047123006035,
-      "learning_rate": 1.4213069315911014e-05,
-      "loss": 0.1032,
-      "step": 640
-    },
-    {
-      "epoch": 7.630952380952381,
-      "grad_norm": 1.4255009879275804,
-      "learning_rate": 1.419557120781772e-05,
-      "loss": 0.0729,
-      "step": 641
-    },
-    {
-      "epoch": 7.642857142857143,
-      "grad_norm": 1.2946732289623608,
-      "learning_rate": 1.4178057495271816e-05,
-      "loss": 0.0332,
-      "step": 642
-    },
-    {
-      "epoch": 7.654761904761905,
-      "grad_norm": 1.397832155407573,
-      "learning_rate": 1.4160528243411493e-05,
-      "loss": 0.057,
-      "step": 643
-    },
-    {
-      "epoch": 7.666666666666667,
-      "grad_norm": 2.258767217854666,
-      "learning_rate": 1.4142983517432724e-05,
-      "loss": 0.0877,
-      "step": 644
-    },
-    {
-      "epoch": 7.678571428571429,
-      "grad_norm": 1.734061257783898,
-      "learning_rate": 1.412542338258905e-05,
-      "loss": 0.0629,
-      "step": 645
-    },
-    {
-      "epoch": 7.690476190476191,
-      "grad_norm": 1.6703985042648415,
-      "learning_rate": 1.4107847904191309e-05,
-      "loss": 0.0752,
-      "step": 646
-    },
-    {
-      "epoch": 7.7023809523809526,
-      "grad_norm": 2.6765274588483208,
-      "learning_rate": 1.4090257147607414e-05,
-      "loss": 0.0718,
-      "step": 647
-    },
-    {
-      "epoch": 7.714285714285714,
-      "grad_norm": 2.2666127426059997,
-      "learning_rate": 1.4072651178262096e-05,
-      "loss": 0.0689,
-      "step": 648
-    },
-    {
-      "epoch": 7.726190476190476,
-      "grad_norm": 3.028603990936858,
-      "learning_rate": 1.405503006163667e-05,
-      "loss": 0.0728,
-      "step": 649
-    },
-    {
-      "epoch": 7.738095238095238,
-      "grad_norm": 1.6601742539557618,
-      "learning_rate": 1.4037393863268785e-05,
-      "loss": 0.0755,
-      "step": 650
-    },
-    {
-      "epoch": 7.75,
-      "grad_norm": 3.6530760121468,
-      "learning_rate": 1.4019742648752184e-05,
-      "loss": 0.06,
-      "step": 651
-    },
-    {
-      "epoch": 7.761904761904762,
-      "grad_norm": 1.909724163901433,
-      "learning_rate": 1.400207648373646e-05,
-      "loss": 0.0795,
-      "step": 652
-    },
-    {
-      "epoch": 7.773809523809524,
-      "grad_norm": 1.5217184466934186,
-      "learning_rate": 1.3984395433926816e-05,
-      "loss": 0.0647,
-      "step": 653
-    },
-    {
-      "epoch": 7.785714285714286,
-      "grad_norm": 2.349579191021632,
-      "learning_rate": 1.3966699565083803e-05,
-      "loss": 0.02,
-      "step": 654
-    },
-    {
-      "epoch": 7.7976190476190474,
-      "grad_norm": 1.5403983446470608,
-      "learning_rate": 1.3948988943023097e-05,
-      "loss": 0.0735,
-      "step": 655
-    },
-    {
-      "epoch": 7.809523809523809,
-      "grad_norm": 1.7826487054141489,
-      "learning_rate": 1.3931263633615241e-05,
-      "loss": 0.0704,
-      "step": 656
-    },
-    {
-      "epoch": 7.821428571428571,
-      "grad_norm": 1.6670060973825707,
-      "learning_rate": 1.3913523702785413e-05,
-      "loss": 0.0646,
-      "step": 657
-    },
-    {
-      "epoch": 7.833333333333333,
-      "grad_norm": 1.1293411386334704,
-      "learning_rate": 1.3895769216513158e-05,
-      "loss": 0.0122,
-      "step": 658
-    },
-    {
-      "epoch": 7.845238095238095,
-      "grad_norm": 0.8329512938247684,
-      "learning_rate": 1.3878000240832168e-05,
-      "loss": 0.0433,
-      "step": 659
-    },
-    {
-      "epoch": 7.857142857142857,
-      "grad_norm": 2.686692661037454,
-      "learning_rate": 1.3860216841830019e-05,
-      "loss": 0.0984,
-      "step": 660
-    },
-    {
-      "epoch": 7.869047619047619,
-      "grad_norm": 1.2301440232404512,
-      "learning_rate": 1.3842419085647933e-05,
-      "loss": 0.0417,
-      "step": 661
-    },
-    {
-      "epoch": 7.880952380952381,
-      "grad_norm": 2.0183227698474333,
-      "learning_rate": 1.3824607038480533e-05,
-      "loss": 0.0723,
-      "step": 662
-    },
-    {
-      "epoch": 7.892857142857143,
-      "grad_norm": 2.6389098823393358,
-      "learning_rate": 1.380678076657559e-05,
-      "loss": 0.09,
-      "step": 663
-    },
-    {
-      "epoch": 7.904761904761905,
-      "grad_norm": 3.4549373370939978,
-      "learning_rate": 1.3788940336233781e-05,
-      "loss": 0.1292,
-      "step": 664
-    },
-    {
-      "epoch": 7.916666666666667,
-      "grad_norm": 2.2667693069829227,
-      "learning_rate": 1.3771085813808442e-05,
-      "loss": 0.0828,
-      "step": 665
-    },
-    {
-      "epoch": 7.928571428571429,
-      "grad_norm": 0.47972584449219374,
-      "learning_rate": 1.3753217265705324e-05,
-      "loss": 0.0053,
-      "step": 666
-    },
-    {
-      "epoch": 7.940476190476191,
-      "grad_norm": 1.75002178495459,
-      "learning_rate": 1.3735334758382338e-05,
-      "loss": 0.04,
-      "step": 667
-    },
-    {
-      "epoch": 7.9523809523809526,
-      "grad_norm": 1.814029715731369,
-      "learning_rate": 1.371743835834932e-05,
-      "loss": 0.0647,
-      "step": 668
-    },
-    {
-      "epoch": 7.964285714285714,
-      "grad_norm": 1.7511940468473046,
-      "learning_rate": 1.3699528132167777e-05,
-      "loss": 0.0169,
-      "step": 669
-    },
-    {
-      "epoch": 7.976190476190476,
-      "grad_norm": 2.1541073840936753,
-      "learning_rate": 1.3681604146450625e-05,
-      "loss": 0.0931,
-      "step": 670
-    },
-    {
-      "epoch": 7.988095238095238,
-      "grad_norm": 2.3197341635579716,
-      "learning_rate": 1.3663666467861972e-05,
-      "loss": 0.0743,
-      "step": 671
-    },
-    {
-      "epoch": 8.0,
-      "grad_norm": 2.5918283864996052,
-      "learning_rate": 1.3645715163116846e-05,
-      "loss": 0.0388,
-      "step": 672
-    },
-    {
-      "epoch": 8.0,
-      "eval_loss": 0.11669167876243591,
-      "eval_runtime": 40.8048,
-      "eval_samples_per_second": 1.47,
-      "eval_steps_per_second": 1.47,
-      "step": 672
-    },
-    {
-      "epoch": 8.011904761904763,
-      "grad_norm": 1.8156831505652815,
-      "learning_rate": 1.362775029898096e-05,
-      "loss": 0.0193,
-      "step": 673
-    },
-    {
-      "epoch": 8.023809523809524,
-      "grad_norm": 1.014332919297201,
-      "learning_rate": 1.3609771942270444e-05,
-      "loss": 0.0618,
-      "step": 674
-    },
-    {
-      "epoch": 8.035714285714286,
-      "grad_norm": 2.740643902996938,
-      "learning_rate": 1.3591780159851629e-05,
-      "loss": 0.084,
-      "step": 675
-    },
-    {
-      "epoch": 8.047619047619047,
-      "grad_norm": 1.2215784460103545,
-      "learning_rate": 1.3573775018640765e-05,
-      "loss": 0.0265,
-      "step": 676
-    },
-    {
-      "epoch": 8.05952380952381,
-      "grad_norm": 1.2814221356115454,
-      "learning_rate": 1.3555756585603795e-05,
-      "loss": 0.0265,
-      "step": 677
-    },
-    {
-      "epoch": 8.071428571428571,
-      "grad_norm": 2.498831848618261,
-      "learning_rate": 1.3537724927756095e-05,
-      "loss": 0.0489,
-      "step": 678
-    },
-    {
-      "epoch": 8.083333333333334,
-      "grad_norm": 2.0916980358289528,
-      "learning_rate": 1.351968011216223e-05,
-      "loss": 0.0194,
-      "step": 679
-    },
-    {
-      "epoch": 8.095238095238095,
-      "grad_norm": 1.8436837623551254,
-      "learning_rate": 1.3501622205935698e-05,
-      "loss": 0.0624,
-      "step": 680
-    },
-    {
-      "epoch": 8.107142857142858,
-      "grad_norm": 2.998109666966864,
-      "learning_rate": 1.348355127623869e-05,
-      "loss": 0.1059,
-      "step": 681
-    },
-    {
-      "epoch": 8.119047619047619,
-      "grad_norm": 0.9572390376576435,
-      "learning_rate": 1.3465467390281826e-05,
-      "loss": 0.0254,
-      "step": 682
-    },
-    {
-      "epoch": 8.130952380952381,
-      "grad_norm": 1.1446640882773418,
-      "learning_rate": 1.3447370615323924e-05,
-      "loss": 0.0321,
-      "step": 683
-    },
-    {
-      "epoch": 8.142857142857142,
-      "grad_norm": 0.9314532514923329,
-      "learning_rate": 1.3429261018671735e-05,
-      "loss": 0.0335,
-      "step": 684
-    },
-    {
-      "epoch": 8.154761904761905,
-      "grad_norm": 0.9099154090085874,
-      "learning_rate": 1.3411138667679697e-05,
-      "loss": 0.0275,
-      "step": 685
-    },
-    {
-      "epoch": 8.166666666666666,
-      "grad_norm": 2.814442692268865,
-      "learning_rate": 1.3393003629749684e-05,
-      "loss": 0.0574,
-      "step": 686
-    },
-    {
-      "epoch": 8.178571428571429,
-      "grad_norm": 1.2925967965525935,
-      "learning_rate": 1.3374855972330758e-05,
-      "loss": 0.0318,
-      "step": 687
-    },
-    {
-      "epoch": 8.19047619047619,
-      "grad_norm": 4.6036673705862,
-      "learning_rate": 1.3356695762918915e-05,
-      "loss": 0.1407,
-      "step": 688
-    },
-    {
-      "epoch": 8.202380952380953,
-      "grad_norm": 2.3563878485370484,
-      "learning_rate": 1.3338523069056838e-05,
-      "loss": 0.0429,
-      "step": 689
-    },
-    {
-      "epoch": 8.214285714285714,
-      "grad_norm": 3.740777621839342,
-      "learning_rate": 1.332033795833364e-05,
-      "loss": 0.0477,
-      "step": 690
-    },
-    {
-      "epoch": 8.226190476190476,
-      "grad_norm": 4.188291251339514,
-      "learning_rate": 1.3302140498384617e-05,
-      "loss": 0.0449,
-      "step": 691
-    },
-    {
-      "epoch": 8.238095238095237,
-      "grad_norm": 1.884707086232278,
-      "learning_rate": 1.328393075689099e-05,
-      "loss": 0.0482,
-      "step": 692
-    },
-    {
-      "epoch": 8.25,
-      "grad_norm": 1.3968846684263698,
-      "learning_rate": 1.326570880157967e-05,
-      "loss": 0.0301,
-      "step": 693
-    },
-    {
-      "epoch": 8.261904761904763,
-      "grad_norm": 1.905739875764223,
-      "learning_rate": 1.3247474700222981e-05,
-      "loss": 0.0698,
-      "step": 694
-    },
-    {
-      "epoch": 8.273809523809524,
-      "grad_norm": 0.8840377855273591,
-      "learning_rate": 1.3229228520638435e-05,
-      "loss": 0.0482,
-      "step": 695
-    },
-    {
-      "epoch": 8.285714285714286,
-      "grad_norm": 1.681936237192879,
-      "learning_rate": 1.3210970330688453e-05,
-      "loss": 0.0501,
-      "step": 696
-    },
-    {
-      "epoch": 8.297619047619047,
-      "grad_norm": 3.012072033315285,
-      "learning_rate": 1.319270019828013e-05,
-      "loss": 0.0741,
-      "step": 697
-    },
-    {
-      "epoch": 8.30952380952381,
-      "grad_norm": 1.374140431461223,
-      "learning_rate": 1.3174418191364988e-05,
-      "loss": 0.038,
-      "step": 698
-    },
-    {
-      "epoch": 8.321428571428571,
-      "grad_norm": 1.810238997618556,
-      "learning_rate": 1.31561243779387e-05,
-      "loss": 0.0669,
-      "step": 699
-    },
-    {
-      "epoch": 8.333333333333334,
-      "grad_norm": 1.8326667848989575,
-      "learning_rate": 1.3137818826040856e-05,
-      "loss": 0.0763,
-      "step": 700
-    },
-    {
-      "epoch": 8.345238095238095,
-      "grad_norm": 1.5303005114433177,
-      "learning_rate": 1.3119501603754705e-05,
-      "loss": 0.0595,
-      "step": 701
-    },
-    {
-      "epoch": 8.357142857142858,
-      "grad_norm": 1.0204967306864992,
-      "learning_rate": 1.3101172779206902e-05,
-      "loss": 0.0251,
-      "step": 702
-    },
-    {
-      "epoch": 8.369047619047619,
-      "grad_norm": 2.6953720121220024,
-      "learning_rate": 1.308283242056725e-05,
-      "loss": 0.0852,
-      "step": 703
-    },
-    {
-      "epoch": 8.380952380952381,
-      "grad_norm": 1.091732342741422,
-      "learning_rate": 1.3064480596048454e-05,
-      "loss": 0.0637,
-      "step": 704
-    },
-    {
-      "epoch": 8.392857142857142,
-      "grad_norm": 1.2212925528454934,
-      "learning_rate": 1.3046117373905866e-05,
-      "loss": 0.0295,
-      "step": 705
-    },
-    {
-      "epoch": 8.404761904761905,
-      "grad_norm": 1.311921484987392,
-      "learning_rate": 1.3027742822437222e-05,
-      "loss": 0.0126,
-      "step": 706
-    },
-    {
-      "epoch": 8.416666666666666,
-      "grad_norm": 1.845203590104293,
-      "learning_rate": 1.3009357009982397e-05,
-      "loss": 0.0568,
-      "step": 707
-    },
-    {
-      "epoch": 8.428571428571429,
-      "grad_norm": 2.1638646922168934,
-      "learning_rate": 1.2990960004923154e-05,
-      "loss": 0.0477,
-      "step": 708
-    },
-    {
-      "epoch": 8.44047619047619,
-      "grad_norm": 3.170909990089901,
-      "learning_rate": 1.2972551875682882e-05,
-      "loss": 0.0851,
-      "step": 709
-    },
-    {
-      "epoch": 8.452380952380953,
-      "grad_norm": 2.0061950163071662,
-      "learning_rate": 1.2954132690726335e-05,
-      "loss": 0.0409,
-      "step": 710
-    },
-    {
-      "epoch": 8.464285714285714,
-      "grad_norm": 1.0898938904654067,
-      "learning_rate": 1.2935702518559399e-05,
-      "loss": 0.0104,
-      "step": 711
-    },
-    {
-      "epoch": 8.476190476190476,
-      "grad_norm": 3.8531317949706008,
-      "learning_rate": 1.2917261427728815e-05,
-      "loss": 0.079,
-      "step": 712
-    },
-    {
-      "epoch": 8.488095238095237,
-      "grad_norm": 3.2258612383912584,
-      "learning_rate": 1.289880948682194e-05,
-      "loss": 0.0624,
-      "step": 713
-    },
-    {
-      "epoch": 8.5,
-      "grad_norm": 2.408210971584371,
-      "learning_rate": 1.288034676446648e-05,
-      "loss": 0.0746,
-      "step": 714
-    },
-    {
-      "epoch": 8.511904761904763,
-      "grad_norm": 1.8900504075144005,
-      "learning_rate": 1.2861873329330248e-05,
-      "loss": 0.0598,
-      "step": 715
-    },
-    {
-      "epoch": 8.523809523809524,
-      "grad_norm": 5.2541248321128196,
-      "learning_rate": 1.2843389250120885e-05,
-      "loss": 0.0451,
-      "step": 716
-    },
-    {
-      "epoch": 8.535714285714286,
-      "grad_norm": 11.687616844777143,
-      "learning_rate": 1.2824894595585637e-05,
-      "loss": 0.0734,
-      "step": 717
-    },
-    {
-      "epoch": 8.547619047619047,
-      "grad_norm": 2.7273662847529643,
-      "learning_rate": 1.2806389434511078e-05,
-      "loss": 0.0887,
-      "step": 718
-    },
-    {
-      "epoch": 8.55952380952381,
-      "grad_norm": 0.8084461433790741,
-      "learning_rate": 1.2787873835722851e-05,
-      "loss": 0.0175,
-      "step": 719
-    },
-    {
-      "epoch": 8.571428571428571,
-      "grad_norm": 2.562516566277651,
-      "learning_rate": 1.276934786808543e-05,
-      "loss": 0.037,
-      "step": 720
-    },
-    {
-      "epoch": 8.583333333333334,
-      "grad_norm": 1.964626228993161,
-      "learning_rate": 1.2750811600501843e-05,
-      "loss": 0.0143,
-      "step": 721
-    },
-    {
-      "epoch": 8.595238095238095,
-      "grad_norm": 2.347260151226618,
-      "learning_rate": 1.2732265101913435e-05,
-      "loss": 0.0992,
-      "step": 722
-    },
-    {
-      "epoch": 8.607142857142858,
-      "grad_norm": 2.332067934310295,
-      "learning_rate": 1.2713708441299602e-05,
-      "loss": 0.0903,
-      "step": 723
-    },
-    {
-      "epoch": 8.619047619047619,
-      "grad_norm": 2.433504640915981,
-      "learning_rate": 1.2695141687677527e-05,
-      "loss": 0.0906,
-      "step": 724
-    },
-    {
-      "epoch": 8.630952380952381,
-      "grad_norm": 0.7365574346673278,
-      "learning_rate": 1.2676564910101948e-05,
-      "loss": 0.0066,
-      "step": 725
-    },
-    {
-      "epoch": 8.642857142857142,
-      "grad_norm": 2.1926068397741028,
-      "learning_rate": 1.2657978177664862e-05,
-      "loss": 0.0692,
-      "step": 726
-    },
-    {
-      "epoch": 8.654761904761905,
-      "grad_norm": 3.677952898983117,
-      "learning_rate": 1.263938155949531e-05,
-      "loss": 0.111,
-      "step": 727
-    },
-    {
-      "epoch": 8.666666666666666,
-      "grad_norm": 1.9850598171802365,
-      "learning_rate": 1.2620775124759093e-05,
-      "loss": 0.0676,
-      "step": 728
-    },
-    {
-      "epoch": 8.678571428571429,
-      "grad_norm": 4.3432896002801495,
-      "learning_rate": 1.2602158942658523e-05,
-      "loss": 0.1596,
-      "step": 729
-    },
-    {
-      "epoch": 8.69047619047619,
-      "grad_norm": 1.6243741730128927,
-      "learning_rate": 1.2583533082432168e-05,
-      "loss": 0.0642,
-      "step": 730
-    },
-    {
-      "epoch": 8.702380952380953,
-      "grad_norm": 1.27022170434261,
-      "learning_rate": 1.2564897613354586e-05,
-      "loss": 0.0247,
-      "step": 731
-    },
-    {
-      "epoch": 8.714285714285714,
-      "grad_norm": 1.323666707186927,
-      "learning_rate": 1.2546252604736073e-05,
-      "loss": 0.0427,
-      "step": 732
-    },
-    {
-      "epoch": 8.726190476190476,
-      "grad_norm": 3.3123883449785327,
-      "learning_rate": 1.2527598125922413e-05,
-      "loss": 0.0919,
-      "step": 733
-    },
-    {
-      "epoch": 8.738095238095237,
-      "grad_norm": 0.867843230559778,
-      "learning_rate": 1.2508934246294604e-05,
-      "loss": 0.0511,
-      "step": 734
-    },
-    {
-      "epoch": 8.75,
-      "grad_norm": 1.2662403709808365,
-      "learning_rate": 1.2490261035268614e-05,
-      "loss": 0.0355,
-      "step": 735
-    },
-    {
-      "epoch": 8.761904761904763,
-      "grad_norm": 1.3128545329767234,
-      "learning_rate": 1.2471578562295115e-05,
-      "loss": 0.0424,
-      "step": 736
-    },
-    {
-      "epoch": 8.773809523809524,
-      "grad_norm": 5.195758504027162,
-      "learning_rate": 1.245288689685922e-05,
-      "loss": 0.0486,
-      "step": 737
-    },
-    {
-      "epoch": 8.785714285714286,
-      "grad_norm": 2.397782390906139,
-      "learning_rate": 1.243418610848024e-05,
-      "loss": 0.0501,
-      "step": 738
-    },
-    {
-      "epoch": 8.797619047619047,
-      "grad_norm": 3.803902756894812,
-      "learning_rate": 1.2415476266711415e-05,
-      "loss": 0.0724,
-      "step": 739
-    },
-    {
-      "epoch": 8.80952380952381,
-      "grad_norm": 3.9391624067695212,
-      "learning_rate": 1.2396757441139655e-05,
-      "loss": 0.0669,
-      "step": 740
-    },
-    {
-      "epoch": 8.821428571428571,
-      "grad_norm": 2.82436087893358,
-      "learning_rate": 1.2378029701385288e-05,
-      "loss": 0.0941,
-      "step": 741
-    },
-    {
-      "epoch": 8.833333333333334,
-      "grad_norm": 1.3255227276592592,
-      "learning_rate": 1.2359293117101783e-05,
-      "loss": 0.0312,
-      "step": 742
-    },
-    {
-      "epoch": 8.845238095238095,
-      "grad_norm": 15.30188169590839,
-      "learning_rate": 1.234054775797552e-05,
-      "loss": 0.2154,
-      "step": 743
-    },
-    {
-      "epoch": 8.857142857142858,
-      "grad_norm": 1.5966516669338604,
-      "learning_rate": 1.2321793693725509e-05,
-      "loss": 0.0496,
-      "step": 744
-    },
-    {
-      "epoch": 8.869047619047619,
-      "grad_norm": 1.4094052298445738,
-      "learning_rate": 1.2303030994103133e-05,
-      "loss": 0.0403,
-      "step": 745
-    },
-    {
-      "epoch": 8.880952380952381,
-      "grad_norm": 1.3961941919580525,
-      "learning_rate": 1.2284259728891897e-05,
-      "loss": 0.0432,
-      "step": 746
-    },
-    {
-      "epoch": 8.892857142857142,
-      "grad_norm": 1.7789035378853075,
-      "learning_rate": 1.2265479967907158e-05,
-      "loss": 0.0558,
-      "step": 747
-    },
-    {
-      "epoch": 8.904761904761905,
-      "grad_norm": 1.3087355978594275,
-      "learning_rate": 1.2246691780995881e-05,
-      "loss": 0.0513,
-      "step": 748
-    },
-    {
-      "epoch": 8.916666666666666,
-      "grad_norm": 2.4584365408967748,
-      "learning_rate": 1.2227895238036358e-05,
-      "loss": 0.0899,
-      "step": 749
-    },
-    {
-      "epoch": 8.928571428571429,
-      "grad_norm": 2.4158077298850396,
-      "learning_rate": 1.2209090408937972e-05,
-      "loss": 0.0968,
-      "step": 750
-    },
-    {
-      "epoch": 8.94047619047619,
-      "grad_norm": 1.4637737987592996,
-      "learning_rate": 1.2190277363640908e-05,
-      "loss": 0.0245,
-      "step": 751
-    },
-    {
-      "epoch": 8.952380952380953,
-      "grad_norm": 3.702641437817495,
-      "learning_rate": 1.2171456172115922e-05,
-      "loss": 0.1082,
-      "step": 752
-    },
-    {
-      "epoch": 8.964285714285714,
-      "grad_norm": 2.349077228223263,
-      "learning_rate": 1.2152626904364066e-05,
-      "loss": 0.0929,
-      "step": 753
-    },
-    {
-      "epoch": 8.976190476190476,
-      "grad_norm": 1.1008670892256474,
-      "learning_rate": 1.2133789630416426e-05,
-      "loss": 0.0355,
-      "step": 754
-    },
-    {
-      "epoch": 8.988095238095237,
-      "grad_norm": 2.1587571145198017,
-      "learning_rate": 1.211494442033387e-05,
-      "loss": 0.0204,
-      "step": 755
-    },
-    {
-      "epoch": 9.0,
-      "grad_norm": 0.8256549231338911,
-      "learning_rate": 1.2096091344206778e-05,
-      "loss": 0.0252,
-      "step": 756
-    },
-    {
-      "epoch": 9.0,
-      "eval_loss": 0.09582231193780899,
-      "eval_runtime": 40.4622,
-      "eval_samples_per_second": 1.483,
-      "eval_steps_per_second": 1.483,
-      "step": 756
-    },
-    {
-      "epoch": 9.011904761904763,
-      "grad_norm": 1.282205989400026,
-      "learning_rate": 1.2077230472154787e-05,
-      "loss": 0.0295,
-      "step": 757
-    },
-    {
-      "epoch": 9.023809523809524,
-      "grad_norm": 2.1270661891136293,
-      "learning_rate": 1.2058361874326527e-05,
-      "loss": 0.0757,
-      "step": 758
-    },
-    {
-      "epoch": 9.035714285714286,
-      "grad_norm": 1.4062511264145792,
-      "learning_rate": 1.203948562089937e-05,
-      "loss": 0.044,
-      "step": 759
-    },
-    {
-      "epoch": 9.047619047619047,
-      "grad_norm": 2.1904184923247643,
-      "learning_rate": 1.2020601782079155e-05,
-      "loss": 0.0824,
-      "step": 760
-    },
-    {
-      "epoch": 9.05952380952381,
-      "grad_norm": 3.286148753711866,
-      "learning_rate": 1.2001710428099935e-05,
-      "loss": 0.0836,
-      "step": 761
-    },
-    {
-      "epoch": 9.071428571428571,
-      "grad_norm": 1.5939900895983135,
-      "learning_rate": 1.198281162922371e-05,
-      "loss": 0.0509,
-      "step": 762
-    },
-    {
-      "epoch": 9.083333333333334,
-      "grad_norm": 0.8241534897298811,
-      "learning_rate": 1.1963905455740176e-05,
-      "loss": 0.0228,
-      "step": 763
-    },
-    {
-      "epoch": 9.095238095238095,
-      "grad_norm": 1.3287066485870127,
-      "learning_rate": 1.1944991977966452e-05,
-      "loss": 0.033,
-      "step": 764
-    },
-    {
-      "epoch": 9.107142857142858,
-      "grad_norm": 1.4021161713675208,
-      "learning_rate": 1.1926071266246826e-05,
-      "loss": 0.0351,
-      "step": 765
-    },
-    {
-      "epoch": 9.119047619047619,
-      "grad_norm": 6.391793428333144,
-      "learning_rate": 1.1907143390952494e-05,
-      "loss": 0.0991,
-      "step": 766
-    },
-    {
-      "epoch": 9.130952380952381,
-      "grad_norm": 2.386900925645386,
-      "learning_rate": 1.1888208422481287e-05,
-      "loss": 0.0205,
-      "step": 767
-    },
-    {
-      "epoch": 9.142857142857142,
-      "grad_norm": 7.279317351490466,
-      "learning_rate": 1.1869266431257422e-05,
-      "loss": 0.0735,
-      "step": 768
-    },
-    {
-      "epoch": 9.154761904761905,
-      "grad_norm": 24.006127332322464,
-      "learning_rate": 1.185031748773124e-05,
-      "loss": 0.219,
-      "step": 769
-    },
-    {
-      "epoch": 9.166666666666666,
-      "grad_norm": 2.0743647255989,
-      "learning_rate": 1.1831361662378933e-05,
-      "loss": 0.0585,
-      "step": 770
-    },
-    {
-      "epoch": 9.178571428571429,
-      "grad_norm": 1.4495226552360447,
-      "learning_rate": 1.1812399025702291e-05,
-      "loss": 0.0279,
-      "step": 771
-    },
-    {
-      "epoch": 9.19047619047619,
-      "grad_norm": 2.482091671301266,
-      "learning_rate": 1.1793429648228437e-05,
-      "loss": 0.0756,
-      "step": 772
-    },
-    {
-      "epoch": 9.202380952380953,
-      "grad_norm": 2.2578166171442784,
-      "learning_rate": 1.177445360050956e-05,
-      "loss": 0.0687,
-      "step": 773
-    },
-    {
-      "epoch": 9.214285714285714,
-      "grad_norm": 1.3642263825505174,
-      "learning_rate": 1.1755470953122668e-05,
-      "loss": 0.033,
-      "step": 774
-    },
-    {
-      "epoch": 9.226190476190476,
-      "grad_norm": 0.43557975402669774,
-      "learning_rate": 1.1736481776669307e-05,
-      "loss": 0.0051,
-      "step": 775
-    },
-    {
-      "epoch": 9.238095238095237,
-      "grad_norm": 1.7819915407174811,
-      "learning_rate": 1.1717486141775305e-05,
-      "loss": 0.0145,
-      "step": 776
-    },
-    {
-      "epoch": 9.25,
-      "grad_norm": 2.524511395140908,
-      "learning_rate": 1.1698484119090518e-05,
-      "loss": 0.0752,
-      "step": 777
-    },
-    {
-      "epoch": 9.261904761904763,
-      "grad_norm": 1.3378846697745967,
-      "learning_rate": 1.1679475779288555e-05,
-      "loss": 0.054,
-      "step": 778
-    },
-    {
-      "epoch": 9.273809523809524,
-      "grad_norm": 2.864204351537188,
-      "learning_rate": 1.1660461193066521e-05,
-      "loss": 0.1019,
-      "step": 779
-    },
-    {
-      "epoch": 9.285714285714286,
-      "grad_norm": 3.044999321706026,
-      "learning_rate": 1.164144043114475e-05,
-      "loss": 0.1038,
-      "step": 780
-    },
-    {
-      "epoch": 9.297619047619047,
-      "grad_norm": 1.4627132525729252,
-      "learning_rate": 1.1622413564266554e-05,
-      "loss": 0.0573,
-      "step": 781
-    },
-    {
-      "epoch": 9.30952380952381,
-      "grad_norm": 2.677810759821976,
-      "learning_rate": 1.1603380663197941e-05,
-      "loss": 0.0715,
-      "step": 782
-    },
-    {
-      "epoch": 9.321428571428571,
-      "grad_norm": 2.063886050190795,
-      "learning_rate": 1.1584341798727365e-05,
-      "loss": 0.0799,
-      "step": 783
-    },
-    {
-      "epoch": 9.333333333333334,
-      "grad_norm": 0.669569446308257,
-      "learning_rate": 1.156529704166546e-05,
-      "loss": 0.0062,
-      "step": 784
-    },
-    {
-      "epoch": 9.345238095238095,
-      "grad_norm": 1.9073286324932273,
-      "learning_rate": 1.154624646284478e-05,
-      "loss": 0.0671,
-      "step": 785
-    },
-    {
-      "epoch": 9.357142857142858,
-      "grad_norm": 3.517001704800636,
-      "learning_rate": 1.1527190133119527e-05,
-      "loss": 0.0793,
-      "step": 786
-    },
-    {
-      "epoch": 9.369047619047619,
-      "grad_norm": 4.373994962095242,
-      "learning_rate": 1.150812812336529e-05,
-      "loss": 0.1071,
-      "step": 787
-    },
-    {
-      "epoch": 9.380952380952381,
-      "grad_norm": 4.416943179885893,
-      "learning_rate": 1.1489060504478788e-05,
-      "loss": 0.1162,
-      "step": 788
-    },
-    {
-      "epoch": 9.392857142857142,
-      "grad_norm": 1.7509893632815494,
-      "learning_rate": 1.1469987347377602e-05,
-      "loss": 0.0305,
-      "step": 789
-    },
-    {
-      "epoch": 9.404761904761905,
-      "grad_norm": 1.97331353114604,
-      "learning_rate": 1.145090872299991e-05,
-      "loss": 0.0588,
-      "step": 790
-    },
-    {
-      "epoch": 9.416666666666666,
-      "grad_norm": 2.3558012359385088,
-      "learning_rate": 1.1431824702304222e-05,
-      "loss": 0.0583,
-      "step": 791
-    },
-    {
-      "epoch": 9.428571428571429,
-      "grad_norm": 2.688493267114919,
-      "learning_rate": 1.1412735356269124e-05,
-      "loss": 0.07,
-      "step": 792
-    },
-    {
-      "epoch": 9.44047619047619,
-      "grad_norm": 1.9416496607434544,
-      "learning_rate": 1.1393640755893003e-05,
-      "loss": 0.0439,
-      "step": 793
-    },
-    {
-      "epoch": 9.452380952380953,
-      "grad_norm": 1.7273468969890204,
-      "learning_rate": 1.1374540972193787e-05,
-      "loss": 0.0523,
-      "step": 794
-    },
-    {
-      "epoch": 9.464285714285714,
-      "grad_norm": 2.9273956897516826,
-      "learning_rate": 1.1355436076208687e-05,
-      "loss": 0.0664,
-      "step": 795
-    },
-    {
-      "epoch": 9.476190476190476,
-      "grad_norm": 6.234566222242338,
-      "learning_rate": 1.1336326138993927e-05,
-      "loss": 0.0867,
-      "step": 796
-    },
-    {
-      "epoch": 9.488095238095237,
-      "grad_norm": 2.322428354346157,
-      "learning_rate": 1.1317211231624482e-05,
-      "loss": 0.0694,
-      "step": 797
-    },
-    {
-      "epoch": 9.5,
-      "grad_norm": 3.2692359052782787,
-      "learning_rate": 1.1298091425193807e-05,
-      "loss": 0.1264,
-      "step": 798
-    },
-    {
-      "epoch": 9.511904761904763,
-      "grad_norm": 1.6606364957031776,
-      "learning_rate": 1.1278966790813582e-05,
-      "loss": 0.044,
-      "step": 799
-    },
-    {
-      "epoch": 9.523809523809524,
-      "grad_norm": 1.1379465046983437,
-      "learning_rate": 1.125983739961344e-05,
-      "loss": 0.01,
-      "step": 800
-    },
-    {
-      "epoch": 9.535714285714286,
-      "grad_norm": 1.9162169523281654,
-      "learning_rate": 1.124070332274071e-05,
-      "loss": 0.0437,
-      "step": 801
-    },
-    {
-      "epoch": 9.547619047619047,
-      "grad_norm": 3.0304053559495294,
-      "learning_rate": 1.1221564631360154e-05,
-      "loss": 0.0766,
-      "step": 802
-    },
-    {
-      "epoch": 9.55952380952381,
-      "grad_norm": 5.340769713504394,
-      "learning_rate": 1.1202421396653678e-05,
-      "loss": 0.0384,
-      "step": 803
-    },
-    {
-      "epoch": 9.571428571428571,
-      "grad_norm": 1.453031390336659,
-      "learning_rate": 1.11832736898201e-05,
-      "loss": 0.0126,
-      "step": 804
-    },
-    {
-      "epoch": 9.583333333333334,
-      "grad_norm": 1.112891160662993,
-      "learning_rate": 1.1164121582074874e-05,
-      "loss": 0.0108,
-      "step": 805
-    },
-    {
-      "epoch": 9.595238095238095,
-      "grad_norm": 1.2126762063748622,
-      "learning_rate": 1.1144965144649809e-05,
-      "loss": 0.0354,
-      "step": 806
-    },
-    {
-      "epoch": 9.607142857142858,
-      "grad_norm": 1.2210629125933703,
-      "learning_rate": 1.1125804448792832e-05,
-      "loss": 0.0119,
-      "step": 807
-    },
-    {
-      "epoch": 9.619047619047619,
-      "grad_norm": 2.7345990636822264,
-      "learning_rate": 1.1106639565767691e-05,
-      "loss": 0.0862,
-      "step": 808
-    },
-    {
-      "epoch": 9.630952380952381,
-      "grad_norm": 1.4528274005758528,
-      "learning_rate": 1.1087470566853727e-05,
-      "loss": 0.0473,
-      "step": 809
-    },
-    {
-      "epoch": 9.642857142857142,
-      "grad_norm": 3.23106792107614,
-      "learning_rate": 1.1068297523345573e-05,
-      "loss": 0.1117,
-      "step": 810
-    },
-    {
-      "epoch": 9.654761904761905,
-      "grad_norm": 1.3500416514630418,
-      "learning_rate": 1.1049120506552913e-05,
-      "loss": 0.0196,
-      "step": 811
-    },
-    {
-      "epoch": 9.666666666666666,
-      "grad_norm": 2.5216815103287296,
-      "learning_rate": 1.1029939587800207e-05,
-      "loss": 0.0714,
-      "step": 812
-    },
-    {
-      "epoch": 9.678571428571429,
-      "grad_norm": 1.331833717446806,
-      "learning_rate": 1.1010754838426427e-05,
-      "loss": 0.0367,
-      "step": 813
-    },
-    {
-      "epoch": 9.69047619047619,
-      "grad_norm": 0.8528992509172784,
-      "learning_rate": 1.099156632978479e-05,
-      "loss": 0.0078,
-      "step": 814
-    },
-    {
-      "epoch": 9.702380952380953,
-      "grad_norm": 1.9692634466833878,
-      "learning_rate": 1.0972374133242503e-05,
-      "loss": 0.0357,
-      "step": 815
-    },
-    {
-      "epoch": 9.714285714285714,
-      "grad_norm": 2.40950210370164,
-      "learning_rate": 1.0953178320180475e-05,
-      "loss": 0.0901,
-      "step": 816
-    },
-    {
-      "epoch": 9.726190476190476,
-      "grad_norm": 0.35898819817840166,
-      "learning_rate": 1.0933978961993084e-05,
-      "loss": 0.0044,
-      "step": 817
-    },
-    {
-      "epoch": 9.738095238095237,
-      "grad_norm": 0.706113936829129,
-      "learning_rate": 1.0914776130087874e-05,
-      "loss": 0.0066,
-      "step": 818
-    },
-    {
-      "epoch": 9.75,
-      "grad_norm": 2.075397250889756,
-      "learning_rate": 1.0895569895885321e-05,
-      "loss": 0.0343,
-      "step": 819
-    },
-    {
-      "epoch": 9.761904761904763,
-      "grad_norm": 1.1878627675894107,
-      "learning_rate": 1.0876360330818554e-05,
-      "loss": 0.0112,
-      "step": 820
-    },
-    {
-      "epoch": 9.773809523809524,
-      "grad_norm": 2.0754258512016817,
-      "learning_rate": 1.0857147506333087e-05,
-      "loss": 0.0287,
-      "step": 821
-    },
-    {
-      "epoch": 9.785714285714286,
-      "grad_norm": 2.270181907997249,
-      "learning_rate": 1.0837931493886561e-05,
-      "loss": 0.053,
-      "step": 822
-    },
-    {
-      "epoch": 9.797619047619047,
-      "grad_norm": 2.996581841029044,
-      "learning_rate": 1.081871236494847e-05,
-      "loss": 0.1128,
-      "step": 823
-    },
-    {
-      "epoch": 9.80952380952381,
-      "grad_norm": 1.4375053769264925,
-      "learning_rate": 1.0799490190999893e-05,
-      "loss": 0.0428,
-      "step": 824
-    },
-    {
-      "epoch": 9.821428571428571,
-      "grad_norm": 1.9654964465679015,
-      "learning_rate": 1.0780265043533252e-05,
-      "loss": 0.0179,
-      "step": 825
-    },
-    {
-      "epoch": 9.833333333333334,
-      "grad_norm": 2.2834355055646607,
-      "learning_rate": 1.0761036994052008e-05,
-      "loss": 0.0764,
-      "step": 826
-    },
-    {
-      "epoch": 9.845238095238095,
-      "grad_norm": 2.677653120102437,
-      "learning_rate": 1.0741806114070434e-05,
-      "loss": 0.0541,
-      "step": 827
-    },
-    {
-      "epoch": 9.857142857142858,
-      "grad_norm": 2.3216873278592494,
-      "learning_rate": 1.0722572475113316e-05,
-      "loss": 0.0445,
-      "step": 828
-    },
-    {
-      "epoch": 9.869047619047619,
-      "grad_norm": 2.134658538525783,
-      "learning_rate": 1.0703336148715705e-05,
-      "loss": 0.0504,
-      "step": 829
-    },
-    {
-      "epoch": 9.880952380952381,
-      "grad_norm": 3.4003928781027457,
-      "learning_rate": 1.0684097206422654e-05,
-      "loss": 0.0889,
-      "step": 830
-    },
-    {
-      "epoch": 9.892857142857142,
-      "grad_norm": 2.8288540439979277,
-      "learning_rate": 1.0664855719788936e-05,
-      "loss": 0.0778,
-      "step": 831
-    },
-    {
-      "epoch": 9.904761904761905,
-      "grad_norm": 1.6914967044629623,
-      "learning_rate": 1.0645611760378795e-05,
-      "loss": 0.0546,
-      "step": 832
-    },
-    {
-      "epoch": 9.916666666666666,
-      "grad_norm": 3.2241458761178876,
-      "learning_rate": 1.0626365399765668e-05,
-      "loss": 0.0458,
-      "step": 833
-    },
-    {
-      "epoch": 9.928571428571429,
-      "grad_norm": 6.421780549816035,
-      "learning_rate": 1.060711670953192e-05,
-      "loss": 0.0422,
-      "step": 834
-    },
-    {
-      "epoch": 9.94047619047619,
-      "grad_norm": 2.2721390346454284,
-      "learning_rate": 1.0587865761268583e-05,
-      "loss": 0.0546,
-      "step": 835
-    },
-    {
-      "epoch": 9.952380952380953,
-      "grad_norm": 1.9385170559910914,
-      "learning_rate": 1.0568612626575093e-05,
-      "loss": 0.0524,
-      "step": 836
-    },
-    {
-      "epoch": 9.964285714285714,
-      "grad_norm": 2.182053081295405,
-      "learning_rate": 1.0549357377059007e-05,
-      "loss": 0.021,
-      "step": 837
-    },
-    {
-      "epoch": 9.976190476190476,
-      "grad_norm": 1.550297109310986,
-      "learning_rate": 1.053010008433576e-05,
-      "loss": 0.0316,
-      "step": 838
-    },
-    {
-      "epoch": 9.988095238095237,
-      "grad_norm": 1.3496799629334293,
-      "learning_rate": 1.051084082002837e-05,
-      "loss": 0.0369,
-      "step": 839
-    },
-    {
-      "epoch": 10.0,
-      "grad_norm": 2.806285360073082,
-      "learning_rate": 1.0491579655767203e-05,
-      "loss": 0.0744,
-      "step": 840
-    },
-    {
-      "epoch": 10.0,
-      "eval_loss": 0.17259296774864197,
-      "eval_runtime": 39.9462,
-      "eval_samples_per_second": 1.502,
-      "eval_steps_per_second": 1.502,
-      "step": 840
-    },
-    {
-      "epoch": 10.011904761904763,
-      "grad_norm": 2.1627843341088786,
-      "learning_rate": 1.0472316663189683e-05,
-      "loss": 0.0561,
-      "step": 841
-    },
-    {
-      "epoch": 10.023809523809524,
-      "grad_norm": 2.8309484064277246,
-      "learning_rate": 1.0453051913940042e-05,
-      "loss": 0.0822,
-      "step": 842
-    },
-    {
-      "epoch": 10.035714285714286,
-      "grad_norm": 4.436488618508568,
-      "learning_rate": 1.0433785479669038e-05,
-      "loss": 0.0324,
-      "step": 843
-    },
-    {
-      "epoch": 10.047619047619047,
-      "grad_norm": 2.812367342584338,
-      "learning_rate": 1.0414517432033695e-05,
-      "loss": 0.0688,
-      "step": 844
-    },
-    {
-      "epoch": 10.05952380952381,
-      "grad_norm": 1.9245163132317105,
-      "learning_rate": 1.039524784269704e-05,
-      "loss": 0.0476,
-      "step": 845
-    },
-    {
-      "epoch": 10.071428571428571,
-      "grad_norm": 1.7189647202551532,
-      "learning_rate": 1.0375976783327841e-05,
-      "loss": 0.0301,
-      "step": 846
-    },
-    {
-      "epoch": 10.083333333333334,
-      "grad_norm": 0.9817231239896711,
-      "learning_rate": 1.0356704325600323e-05,
-      "loss": 0.0077,
-      "step": 847
-    },
-    {
-      "epoch": 10.095238095238095,
-      "grad_norm": 1.3545517237783122,
-      "learning_rate": 1.0337430541193918e-05,
-      "loss": 0.0528,
-      "step": 848
-    },
-    {
-      "epoch": 10.107142857142858,
-      "grad_norm": 1.9013617849189701,
-      "learning_rate": 1.0318155501792988e-05,
-      "loss": 0.0469,
-      "step": 849
-    },
-    {
-      "epoch": 10.119047619047619,
-      "grad_norm": 1.5367684534473693,
-      "learning_rate": 1.0298879279086568e-05,
-      "loss": 0.0186,
-      "step": 850
-    },
-    {
-      "epoch": 10.130952380952381,
-      "grad_norm": 1.4082305565292088,
-      "learning_rate": 1.027960194476809e-05,
-      "loss": 0.0414,
-      "step": 851
-    },
-    {
-      "epoch": 10.142857142857142,
-      "grad_norm": 1.9714729317443556,
-      "learning_rate": 1.026032357053512e-05,
-      "loss": 0.0327,
-      "step": 852
-    },
-    {
-      "epoch": 10.154761904761905,
-      "grad_norm": 1.2486033765999593,
-      "learning_rate": 1.0241044228089096e-05,
-      "loss": 0.0153,
-      "step": 853
-    },
-    {
-      "epoch": 10.166666666666666,
-      "grad_norm": 1.8558173945665013,
-      "learning_rate": 1.0221763989135052e-05,
-      "loss": 0.0317,
-      "step": 854
-    },
-    {
-      "epoch": 10.178571428571429,
-      "grad_norm": 1.9976430929242892,
-      "learning_rate": 1.0202482925381359e-05,
-      "loss": 0.0674,
-      "step": 855
-    },
-    {
-      "epoch": 10.19047619047619,
-      "grad_norm": 2.2642858611465,
-      "learning_rate": 1.0183201108539453e-05,
-      "loss": 0.0148,
-      "step": 856
-    },
-    {
-      "epoch": 10.202380952380953,
-      "grad_norm": 3.740932380776293,
-      "learning_rate": 1.016391861032358e-05,
-      "loss": 0.0669,
-      "step": 857
-    },
-    {
-      "epoch": 10.214285714285714,
-      "grad_norm": 1.8188813093185374,
-      "learning_rate": 1.0144635502450509e-05,
-      "loss": 0.0379,
-      "step": 858
-    },
-    {
-      "epoch": 10.226190476190476,
-      "grad_norm": 2.7437019175338424,
-      "learning_rate": 1.0125351856639279e-05,
-      "loss": 0.0326,
-      "step": 859
-    },
-    {
-      "epoch": 10.238095238095237,
-      "grad_norm": 3.933447240473136,
-      "learning_rate": 1.0106067744610932e-05,
-      "loss": 0.0809,
-      "step": 860
-    },
-    {
-      "epoch": 10.25,
-      "grad_norm": 2.6212363917089987,
-      "learning_rate": 1.0086783238088244e-05,
-      "loss": 0.0181,
-      "step": 861
-    },
-    {
-      "epoch": 10.261904761904763,
-      "grad_norm": 1.4956127769577703,
-      "learning_rate": 1.0067498408795462e-05,
-      "loss": 0.0393,
-      "step": 862
-    },
-    {
-      "epoch": 10.273809523809524,
-      "grad_norm": 4.756355647768309,
-      "learning_rate": 1.0048213328458027e-05,
-      "loss": 0.0484,
-      "step": 863
-    },
-    {
-      "epoch": 10.285714285714286,
-      "grad_norm": 5.264351585446348,
-      "learning_rate": 1.0028928068802314e-05,
-      "loss": 0.0264,
-      "step": 864
-    },
-    {
-      "epoch": 10.297619047619047,
-      "grad_norm": 1.4844810359483214,
-      "learning_rate": 1.0009642701555369e-05,
-      "loss": 0.0303,
-      "step": 865
-    },
-    {
-      "epoch": 10.30952380952381,
-      "grad_norm": 2.603905301584864,
-      "learning_rate": 9.990357298444631e-06,
-      "loss": 0.0162,
-      "step": 866
-    },
-    {
-      "epoch": 10.321428571428571,
-      "grad_norm": 2.8184520996344444,
-      "learning_rate": 9.971071931197686e-06,
-      "loss": 0.0552,
-      "step": 867
-    },
-    {
-      "epoch": 10.333333333333334,
-      "grad_norm": 1.2787347559583708,
-      "learning_rate": 9.951786671541975e-06,
-      "loss": 0.0091,
-      "step": 868
-    },
-    {
-      "epoch": 10.345238095238095,
-      "grad_norm": 3.460050048787224,
-      "learning_rate": 9.932501591204538e-06,
-      "loss": 0.0852,
-      "step": 869
-    },
-    {
-      "epoch": 10.357142857142858,
-      "grad_norm": 3.4097967028653273,
-      "learning_rate": 9.913216761911754e-06,
-      "loss": 0.067,
-      "step": 870
-    },
-    {
-      "epoch": 10.369047619047619,
-      "grad_norm": 2.4386231982584436,
-      "learning_rate": 9.89393225538907e-06,
-      "loss": 0.0556,
-      "step": 871
-    },
-    {
-      "epoch": 10.380952380952381,
-      "grad_norm": 2.933272866461037,
-      "learning_rate": 9.874648143360723e-06,
-      "loss": 0.0496,
-      "step": 872
-    },
-    {
-      "epoch": 10.392857142857142,
-      "grad_norm": 1.8929442632741174,
-      "learning_rate": 9.855364497549495e-06,
-      "loss": 0.0413,
-      "step": 873
-    },
-    {
-      "epoch": 10.404761904761905,
-      "grad_norm": 2.3616431199563435,
-      "learning_rate": 9.836081389676422e-06,
-      "loss": 0.0538,
-      "step": 874
-    },
-    {
-      "epoch": 10.416666666666666,
-      "grad_norm": 2.4127824710359524,
-      "learning_rate": 9.816798891460545e-06,
-      "loss": 0.0191,
-      "step": 875
-    },
-    {
-      "epoch": 10.428571428571429,
-      "grad_norm": 3.907132416538552,
-      "learning_rate": 9.797517074618643e-06,
-      "loss": 0.0892,
-      "step": 876
-    },
-    {
-      "epoch": 10.44047619047619,
-      "grad_norm": 4.181403959687322,
-      "learning_rate": 9.778236010864948e-06,
-      "loss": 0.0902,
-      "step": 877
-    },
-    {
-      "epoch": 10.452380952380953,
-      "grad_norm": 2.3611084962335953,
-      "learning_rate": 9.758955771910906e-06,
-      "loss": 0.0162,
-      "step": 878
-    },
-    {
-      "epoch": 10.464285714285714,
-      "grad_norm": 1.9021196469452089,
-      "learning_rate": 9.739676429464881e-06,
-      "loss": 0.0535,
-      "step": 879
-    },
-    {
-      "epoch": 10.476190476190476,
-      "grad_norm": 3.988458784098384,
-      "learning_rate": 9.72039805523191e-06,
-      "loss": 0.0521,
-      "step": 880
-    },
-    {
-      "epoch": 10.488095238095237,
-      "grad_norm": 1.6135380627489257,
-      "learning_rate": 9.701120720913434e-06,
-      "loss": 0.0289,
-      "step": 881
-    },
-    {
-      "epoch": 10.5,
-      "grad_norm": 2.7761673436825713,
-      "learning_rate": 9.681844498207012e-06,
-      "loss": 0.0658,
-      "step": 882
-    },
-    {
-      "epoch": 10.511904761904763,
-      "grad_norm": 2.3672118880045385,
-      "learning_rate": 9.662569458806085e-06,
-      "loss": 0.0162,
-      "step": 883
-    },
-    {
-      "epoch": 10.523809523809524,
-      "grad_norm": 2.29142928632914,
-      "learning_rate": 9.64329567439968e-06,
-      "loss": 0.0429,
-      "step": 884
-    },
-    {
-      "epoch": 10.535714285714286,
-      "grad_norm": 2.6254810846588823,
-      "learning_rate": 9.624023216672162e-06,
-      "loss": 0.038,
-      "step": 885
-    },
-    {
-      "epoch": 10.547619047619047,
-      "grad_norm": 3.7113842881972587,
-      "learning_rate": 9.604752157302962e-06,
-      "loss": 0.0229,
-      "step": 886
-    },
-    {
-      "epoch": 10.55952380952381,
-      "grad_norm": 3.3798411722544457,
-      "learning_rate": 9.585482567966309e-06,
-      "loss": 0.0569,
-      "step": 887
-    },
-    {
-      "epoch": 10.571428571428571,
-      "grad_norm": 1.3225876420518712,
-      "learning_rate": 9.566214520330966e-06,
-      "loss": 0.0251,
-      "step": 888
-    },
-    {
-      "epoch": 10.583333333333334,
-      "grad_norm": 1.4216589719224715,
-      "learning_rate": 9.54694808605996e-06,
-      "loss": 0.0173,
-      "step": 889
-    },
-    {
-      "epoch": 10.595238095238095,
-      "grad_norm": 1.3910055558361236,
-      "learning_rate": 9.527683336810319e-06,
-      "loss": 0.0186,
-      "step": 890
-    },
-    {
-      "epoch": 10.607142857142858,
-      "grad_norm": 1.7766662091152174,
-      "learning_rate": 9.5084203442328e-06,
-      "loss": 0.0203,
-      "step": 891
-    },
-    {
-      "epoch": 10.619047619047619,
-      "grad_norm": 0.5740612737142285,
-      "learning_rate": 9.489159179971633e-06,
-      "loss": 0.0051,
-      "step": 892
-    },
-    {
-      "epoch": 10.630952380952381,
-      "grad_norm": 2.3148547584071126,
-      "learning_rate": 9.469899915664245e-06,
-      "loss": 0.0315,
-      "step": 893
-    },
-    {
-      "epoch": 10.642857142857142,
-      "grad_norm": 4.294237887718487,
-      "learning_rate": 9.450642622940995e-06,
-      "loss": 0.0989,
-      "step": 894
-    },
-    {
-      "epoch": 10.654761904761905,
-      "grad_norm": 1.9110906979245355,
-      "learning_rate": 9.43138737342491e-06,
-      "loss": 0.0375,
-      "step": 895
-    },
-    {
-      "epoch": 10.666666666666666,
-      "grad_norm": 3.080110364647247,
-      "learning_rate": 9.412134238731419e-06,
-      "loss": 0.0217,
-      "step": 896
-    },
-    {
-      "epoch": 10.678571428571429,
-      "grad_norm": 1.363341861610208,
-      "learning_rate": 9.392883290468084e-06,
-      "loss": 0.0435,
-      "step": 897
-    },
-    {
-      "epoch": 10.69047619047619,
-      "grad_norm": 4.827423440092501,
-      "learning_rate": 9.373634600234336e-06,
-      "loss": 0.068,
-      "step": 898
-    },
-    {
-      "epoch": 10.702380952380953,
-      "grad_norm": 4.225961183200878,
-      "learning_rate": 9.354388239621208e-06,
-      "loss": 0.0479,
-      "step": 899
-    },
-    {
-      "epoch": 10.714285714285714,
-      "grad_norm": 1.7229712679249933,
-      "learning_rate": 9.335144280211066e-06,
-      "loss": 0.0305,
-      "step": 900
-    },
-    {
-      "epoch": 10.726190476190476,
-      "grad_norm": 1.0787497487434379,
-      "learning_rate": 9.315902793577349e-06,
-      "loss": 0.0145,
-      "step": 901
-    },
-    {
-      "epoch": 10.738095238095237,
-      "grad_norm": 2.7023940269650053,
-      "learning_rate": 9.296663851284298e-06,
-      "loss": 0.0739,
-      "step": 902
-    },
-    {
-      "epoch": 10.75,
-      "grad_norm": 1.9107792283247966,
-      "learning_rate": 9.277427524886689e-06,
-      "loss": 0.0466,
-      "step": 903
-    },
-    {
-      "epoch": 10.761904761904763,
-      "grad_norm": 2.345839473140023,
-      "learning_rate": 9.258193885929569e-06,
-      "loss": 0.0598,
-      "step": 904
-    },
-    {
-      "epoch": 10.773809523809524,
-      "grad_norm": 2.29363713803529,
-      "learning_rate": 9.238963005947994e-06,
-      "loss": 0.0303,
-      "step": 905
-    },
-    {
-      "epoch": 10.785714285714286,
-      "grad_norm": 10.905447082859407,
-      "learning_rate": 9.219734956466753e-06,
-      "loss": 0.1111,
-      "step": 906
-    },
-    {
-      "epoch": 10.797619047619047,
-      "grad_norm": 3.4292885991616764,
-      "learning_rate": 9.200509809000108e-06,
-      "loss": 0.0189,
-      "step": 907
-    },
-    {
-      "epoch": 10.80952380952381,
-      "grad_norm": 3.960545973039541,
-      "learning_rate": 9.181287635051535e-06,
-      "loss": 0.0196,
-      "step": 908
-    },
-    {
-      "epoch": 10.821428571428571,
-      "grad_norm": 4.77796269861986,
-      "learning_rate": 9.16206850611344e-06,
-      "loss": 0.0641,
-      "step": 909
-    },
-    {
-      "epoch": 10.833333333333334,
-      "grad_norm": 2.248205151577021,
-      "learning_rate": 9.142852493666914e-06,
-      "loss": 0.0463,
-      "step": 910
-    },
-    {
-      "epoch": 10.845238095238095,
-      "grad_norm": 0.5220374804491678,
-      "learning_rate": 9.123639669181448e-06,
-      "loss": 0.0037,
-      "step": 911
-    },
-    {
-      "epoch": 10.857142857142858,
-      "grad_norm": 1.4827333676054641,
-      "learning_rate": 9.10443010411468e-06,
-      "loss": 0.0363,
-      "step": 912
-    },
-    {
-      "epoch": 10.869047619047619,
-      "grad_norm": 2.3046166185723367,
-      "learning_rate": 9.085223869912129e-06,
-      "loss": 0.0264,
-      "step": 913
-    },
-    {
-      "epoch": 10.880952380952381,
-      "grad_norm": 1.374770737645624,
-      "learning_rate": 9.066021038006919e-06,
-      "loss": 0.0242,
-      "step": 914
-    },
-    {
-      "epoch": 10.892857142857142,
-      "grad_norm": 0.5505032227795658,
-      "learning_rate": 9.046821679819528e-06,
-      "loss": 0.0107,
-      "step": 915
-    },
-    {
-      "epoch": 10.904761904761905,
-      "grad_norm": 0.6878170520689223,
-      "learning_rate": 9.027625866757502e-06,
-      "loss": 0.0057,
-      "step": 916
-    },
-    {
-      "epoch": 10.916666666666666,
-      "grad_norm": 0.7765123697963863,
-      "learning_rate": 9.008433670215211e-06,
-      "loss": 0.0101,
-      "step": 917
-    },
-    {
-      "epoch": 10.928571428571429,
-      "grad_norm": 2.1303477070805057,
-      "learning_rate": 8.989245161573576e-06,
-      "loss": 0.0443,
-      "step": 918
-    },
-    {
-      "epoch": 10.94047619047619,
-      "grad_norm": 1.3982329826936775,
-      "learning_rate": 8.970060412199796e-06,
-      "loss": 0.0207,
-      "step": 919
-    },
-    {
-      "epoch": 10.952380952380953,
-      "grad_norm": 0.1649606219604496,
-      "learning_rate": 8.95087949344709e-06,
-      "loss": 0.002,
-      "step": 920
-    },
-    {
-      "epoch": 10.964285714285714,
-      "grad_norm": 3.2199107038343215,
-      "learning_rate": 8.931702476654432e-06,
-      "loss": 0.0505,
-      "step": 921
-    },
-    {
-      "epoch": 10.976190476190476,
-      "grad_norm": 2.16680588658055,
-      "learning_rate": 8.912529433146277e-06,
-      "loss": 0.0499,
-      "step": 922
-    },
-    {
-      "epoch": 10.988095238095237,
-      "grad_norm": 3.2534119784727134,
-      "learning_rate": 8.893360434232312e-06,
-      "loss": 0.0592,
-      "step": 923
-    },
-    {
-      "epoch": 11.0,
-      "grad_norm": 1.1963933272226175,
-      "learning_rate": 8.874195551207173e-06,
-      "loss": 0.0413,
-      "step": 924
-    },
-    {
-      "epoch": 11.0,
-      "eval_loss": 0.11936289817094803,
-      "eval_runtime": 41.3432,
-      "eval_samples_per_second": 1.451,
-      "eval_steps_per_second": 1.451,
-      "step": 924
-    },
-    {
-      "epoch": 11.011904761904763,
-      "grad_norm": 4.071414945159717,
-      "learning_rate": 8.855034855350195e-06,
-      "loss": 0.0826,
-      "step": 925
-    },
-    {
-      "epoch": 11.023809523809524,
-      "grad_norm": 1.8132042735392948,
-      "learning_rate": 8.835878417925132e-06,
-      "loss": 0.0227,
-      "step": 926
-    },
-    {
-      "epoch": 11.035714285714286,
-      "grad_norm": 2.175362642713856,
-      "learning_rate": 8.816726310179904e-06,
-      "loss": 0.0147,
-      "step": 927
-    },
-    {
-      "epoch": 11.047619047619047,
-      "grad_norm": 3.3405606733438193,
-      "learning_rate": 8.797578603346329e-06,
-      "loss": 0.0446,
-      "step": 928
-    },
-    {
-      "epoch": 11.05952380952381,
-      "grad_norm": 1.467071763402714,
-      "learning_rate": 8.778435368639851e-06,
-      "loss": 0.0434,
-      "step": 929
-    },
-    {
-      "epoch": 11.071428571428571,
-      "grad_norm": 2.378574565354438,
-      "learning_rate": 8.759296677259291e-06,
-      "loss": 0.0179,
-      "step": 930
-    },
-    {
-      "epoch": 11.083333333333334,
-      "grad_norm": 2.286995252249508,
-      "learning_rate": 8.740162600386564e-06,
-      "loss": 0.0364,
-      "step": 931
-    },
-    {
-      "epoch": 11.095238095238095,
-      "grad_norm": 0.3706268897484962,
-      "learning_rate": 8.721033209186425e-06,
-      "loss": 0.0076,
-      "step": 932
-    },
-    {
-      "epoch": 11.107142857142858,
-      "grad_norm": 1.8385403898102013,
-      "learning_rate": 8.701908574806198e-06,
-      "loss": 0.0372,
-      "step": 933
-    },
-    {
-      "epoch": 11.119047619047619,
-      "grad_norm": 9.977918756470613,
-      "learning_rate": 8.682788768375521e-06,
-      "loss": 0.1343,
-      "step": 934
-    },
-    {
-      "epoch": 11.130952380952381,
-      "grad_norm": 11.53576686757209,
-      "learning_rate": 8.663673861006075e-06,
-      "loss": 0.0868,
-      "step": 935
-    },
-    {
-      "epoch": 11.142857142857142,
-      "grad_norm": 0.6141826906637148,
-      "learning_rate": 8.644563923791318e-06,
-      "loss": 0.0108,
-      "step": 936
-    },
-    {
-      "epoch": 11.154761904761905,
-      "grad_norm": 2.432596609300301,
-      "learning_rate": 8.625459027806215e-06,
-      "loss": 0.0339,
-      "step": 937
-    },
-    {
-      "epoch": 11.166666666666666,
-      "grad_norm": 3.8542871920099198,
-      "learning_rate": 8.606359244106998e-06,
-      "loss": 0.0668,
-      "step": 938
-    },
-    {
-      "epoch": 11.178571428571429,
-      "grad_norm": 2.671768123413697,
-      "learning_rate": 8.587264643730877e-06,
-      "loss": 0.0619,
-      "step": 939
-    },
-    {
-      "epoch": 11.19047619047619,
-      "grad_norm": 3.196564343224687,
-      "learning_rate": 8.568175297695777e-06,
-      "loss": 0.0576,
-      "step": 940
-    },
-    {
-      "epoch": 11.202380952380953,
-      "grad_norm": 1.768532734461991,
-      "learning_rate": 8.549091277000092e-06,
-      "loss": 0.0233,
-      "step": 941
-    },
-    {
-      "epoch": 11.214285714285714,
-      "grad_norm": 0.7924847176972385,
-      "learning_rate": 8.530012652622398e-06,
-      "loss": 0.0047,
-      "step": 942
-    },
-    {
-      "epoch": 11.226190476190476,
-      "grad_norm": 1.7637958145761288,
-      "learning_rate": 8.510939495521213e-06,
-      "loss": 0.0507,
-      "step": 943
-    },
-    {
-      "epoch": 11.238095238095237,
-      "grad_norm": 1.582931148820781,
-      "learning_rate": 8.491871876634712e-06,
-      "loss": 0.0212,
-      "step": 944
-    },
-    {
-      "epoch": 11.25,
-      "grad_norm": 0.5914931459652956,
-      "learning_rate": 8.472809866880475e-06,
-      "loss": 0.0038,
-      "step": 945
-    },
-    {
-      "epoch": 11.261904761904763,
-      "grad_norm": 1.1896930113797612,
-      "learning_rate": 8.453753537155222e-06,
-      "loss": 0.0066,
-      "step": 946
-    },
-    {
-      "epoch": 11.273809523809524,
-      "grad_norm": 1.9364330725484524,
-      "learning_rate": 8.43470295833454e-06,
-      "loss": 0.0404,
-      "step": 947
-    },
-    {
-      "epoch": 11.285714285714286,
-      "grad_norm": 2.4320944736811754,
-      "learning_rate": 8.415658201272636e-06,
-      "loss": 0.0477,
-      "step": 948
-    },
-    {
-      "epoch": 11.297619047619047,
-      "grad_norm": 0.7784757843303742,
-      "learning_rate": 8.39661933680206e-06,
-      "loss": 0.0048,
-      "step": 949
-    },
-    {
-      "epoch": 11.30952380952381,
-      "grad_norm": 2.42654239840075,
-      "learning_rate": 8.377586435733448e-06,
-      "loss": 0.0434,
-      "step": 950
-    },
-    {
-      "epoch": 11.321428571428571,
-      "grad_norm": 1.4246506756586026,
-      "learning_rate": 8.35855956885525e-06,
-      "loss": 0.008,
-      "step": 951
-    },
-    {
-      "epoch": 11.333333333333334,
-      "grad_norm": 2.1965500168826084,
-      "learning_rate": 8.33953880693348e-06,
-      "loss": 0.0703,
-      "step": 952
-    },
-    {
-      "epoch": 11.345238095238095,
-      "grad_norm": 2.932674016353764,
-      "learning_rate": 8.320524220711446e-06,
-      "loss": 0.0822,
-      "step": 953
-    },
-    {
-      "epoch": 11.357142857142858,
-      "grad_norm": 1.7763421701243831,
-      "learning_rate": 8.301515880909482e-06,
-      "loss": 0.0646,
-      "step": 954
-    },
-    {
-      "epoch": 11.369047619047619,
-      "grad_norm": 2.8036007920254216,
-      "learning_rate": 8.282513858224698e-06,
-      "loss": 0.0571,
-      "step": 955
-    },
-    {
-      "epoch": 11.380952380952381,
-      "grad_norm": 2.2812771220745045,
-      "learning_rate": 8.263518223330698e-06,
-      "loss": 0.0403,
-      "step": 956
-    },
-    {
-      "epoch": 11.392857142857142,
-      "grad_norm": 2.7481991959683905,
-      "learning_rate": 8.244529046877336e-06,
-      "loss": 0.0118,
-      "step": 957
-    },
-    {
-      "epoch": 11.404761904761905,
-      "grad_norm": 2.5728367074156058,
-      "learning_rate": 8.225546399490442e-06,
-      "loss": 0.0292,
-      "step": 958
-    },
-    {
-      "epoch": 11.416666666666666,
-      "grad_norm": 2.1394262109496527,
-      "learning_rate": 8.206570351771568e-06,
-      "loss": 0.058,
-      "step": 959
-    },
-    {
-      "epoch": 11.428571428571429,
-      "grad_norm": 3.5093300747987968,
-      "learning_rate": 8.187600974297714e-06,
-      "loss": 0.0594,
-      "step": 960
-    },
-    {
-      "epoch": 11.44047619047619,
-      "grad_norm": 2.1510794700494036,
-      "learning_rate": 8.16863833762107e-06,
-      "loss": 0.0526,
-      "step": 961
-    },
-    {
-      "epoch": 11.452380952380953,
-      "grad_norm": 0.6585429250012116,
-      "learning_rate": 8.149682512268762e-06,
-      "loss": 0.0116,
-      "step": 962
-    },
-    {
-      "epoch": 11.464285714285714,
-      "grad_norm": 6.3589705470591005,
-      "learning_rate": 8.13073356874258e-06,
-      "loss": 0.0423,
-      "step": 963
-    },
-    {
-      "epoch": 11.476190476190476,
-      "grad_norm": 2.254323022814341,
-      "learning_rate": 8.111791577518716e-06,
-      "loss": 0.0161,
-      "step": 964
-    },
-    {
-      "epoch": 11.488095238095237,
-      "grad_norm": 3.5461286235553846,
-      "learning_rate": 8.092856609047507e-06,
-      "loss": 0.0247,
-      "step": 965
-    },
-    {
-      "epoch": 11.5,
-      "grad_norm": 15.739331158348278,
-      "learning_rate": 8.073928733753176e-06,
-      "loss": 0.066,
-      "step": 966
-    },
-    {
-      "epoch": 11.511904761904763,
-      "grad_norm": 2.5580355333647584,
-      "learning_rate": 8.055008022033551e-06,
-      "loss": 0.0684,
-      "step": 967
-    },
-    {
-      "epoch": 11.523809523809524,
-      "grad_norm": 4.913451078682193,
-      "learning_rate": 8.036094544259827e-06,
-      "loss": 0.1084,
-      "step": 968
-    },
-    {
-      "epoch": 11.535714285714286,
-      "grad_norm": 3.2726257546182387,
-      "learning_rate": 8.017188370776291e-06,
-      "loss": 0.0523,
-      "step": 969
-    },
-    {
-      "epoch": 11.547619047619047,
-      "grad_norm": 2.6146492746923258,
-      "learning_rate": 7.998289571900067e-06,
-      "loss": 0.034,
-      "step": 970
-    },
-    {
-      "epoch": 11.55952380952381,
-      "grad_norm": 1.625512471741649,
-      "learning_rate": 7.979398217920848e-06,
-      "loss": 0.0222,
-      "step": 971
-    },
-    {
-      "epoch": 11.571428571428571,
-      "grad_norm": 0.8994325256152944,
-      "learning_rate": 7.960514379100632e-06,
-      "loss": 0.0073,
-      "step": 972
-    },
-    {
-      "epoch": 11.583333333333334,
-      "grad_norm": 0.7340148368002074,
-      "learning_rate": 7.941638125673475e-06,
-      "loss": 0.0084,
-      "step": 973
-    },
-    {
-      "epoch": 11.595238095238095,
-      "grad_norm": 1.9951166193182492,
-      "learning_rate": 7.922769527845218e-06,
-      "loss": 0.0249,
-      "step": 974
-    },
-    {
-      "epoch": 11.607142857142858,
-      "grad_norm": 3.943314618173986,
-      "learning_rate": 7.903908655793224e-06,
-      "loss": 0.0648,
-      "step": 975
-    },
-    {
-      "epoch": 11.619047619047619,
-      "grad_norm": 2.4252278861240155,
-      "learning_rate": 7.885055579666134e-06,
-      "loss": 0.0869,
-      "step": 976
-    },
-    {
-      "epoch": 11.630952380952381,
-      "grad_norm": 2.3616939840761737,
-      "learning_rate": 7.866210369583575e-06,
-      "loss": 0.0357,
-      "step": 977
-    },
-    {
-      "epoch": 11.642857142857142,
-      "grad_norm": 1.498225825498272,
-      "learning_rate": 7.847373095635937e-06,
-      "loss": 0.0311,
-      "step": 978
-    },
-    {
-      "epoch": 11.654761904761905,
-      "grad_norm": 5.646623285086258,
-      "learning_rate": 7.82854382788408e-06,
-      "loss": 0.0653,
-      "step": 979
-    },
-    {
-      "epoch": 11.666666666666666,
-      "grad_norm": 0.8452190367515048,
-      "learning_rate": 7.809722636359097e-06,
-      "loss": 0.0119,
-      "step": 980
-    },
-    {
-      "epoch": 11.678571428571429,
-      "grad_norm": 2.1528464430456014,
-      "learning_rate": 7.790909591062033e-06,
-      "loss": 0.0146,
-      "step": 981
-    },
-    {
-      "epoch": 11.69047619047619,
-      "grad_norm": 4.224232166285187,
-      "learning_rate": 7.772104761963645e-06,
-      "loss": 0.0638,
-      "step": 982
-    },
-    {
-      "epoch": 11.702380952380953,
-      "grad_norm": 2.8975023694327957,
-      "learning_rate": 7.753308219004124e-06,
-      "loss": 0.0534,
-      "step": 983
-    },
-    {
-      "epoch": 11.714285714285714,
-      "grad_norm": 1.4871431291781625,
-      "learning_rate": 7.734520032092845e-06,
-      "loss": 0.0147,
-      "step": 984
-    },
-    {
-      "epoch": 11.726190476190476,
-      "grad_norm": 2.817660054546793,
-      "learning_rate": 7.715740271108108e-06,
-      "loss": 0.0334,
-      "step": 985
-    },
-    {
-      "epoch": 11.738095238095237,
-      "grad_norm": 2.265439459023376,
-      "learning_rate": 7.69696900589687e-06,
-      "loss": 0.0351,
-      "step": 986
-    },
-    {
-      "epoch": 11.75,
-      "grad_norm": 2.820518716548115,
-      "learning_rate": 7.678206306274495e-06,
-      "loss": 0.0373,
-      "step": 987
-    },
-    {
-      "epoch": 11.761904761904763,
-      "grad_norm": 1.4555114365738053,
-      "learning_rate": 7.659452242024483e-06,
-      "loss": 0.0359,
-      "step": 988
-    },
-    {
-      "epoch": 11.773809523809524,
-      "grad_norm": 2.148007272965159,
-      "learning_rate": 7.64070688289822e-06,
-      "loss": 0.0427,
-      "step": 989
-    },
-    {
-      "epoch": 11.785714285714286,
-      "grad_norm": 2.6798202271627423,
-      "learning_rate": 7.621970298614717e-06,
-      "loss": 0.046,
-      "step": 990
-    },
-    {
-      "epoch": 11.797619047619047,
-      "grad_norm": 0.5172075857377311,
-      "learning_rate": 7.6032425588603465e-06,
-      "loss": 0.0085,
-      "step": 991
-    },
-    {
-      "epoch": 11.80952380952381,
-      "grad_norm": 2.8043769978606914,
-      "learning_rate": 7.584523733288589e-06,
-      "loss": 0.0491,
-      "step": 992
-    },
-    {
-      "epoch": 11.821428571428571,
-      "grad_norm": 1.37174029863172,
-      "learning_rate": 7.565813891519766e-06,
-      "loss": 0.0202,
-      "step": 993
-    },
-    {
-      "epoch": 11.833333333333334,
-      "grad_norm": 0.2854308071521844,
-      "learning_rate": 7.547113103140786e-06,
-      "loss": 0.0021,
-      "step": 994
-    },
-    {
-      "epoch": 11.845238095238095,
-      "grad_norm": 3.484971218422417,
-      "learning_rate": 7.528421437704891e-06,
-      "loss": 0.0494,
-      "step": 995
-    },
-    {
-      "epoch": 11.857142857142858,
-      "grad_norm": 4.439639213595532,
-      "learning_rate": 7.509738964731389e-06,
-      "loss": 0.0382,
-      "step": 996
-    },
-    {
-      "epoch": 11.869047619047619,
-      "grad_norm": 2.9936609268911947,
-      "learning_rate": 7.4910657537054e-06,
-      "loss": 0.076,
-      "step": 997
-    },
-    {
-      "epoch": 11.880952380952381,
-      "grad_norm": 1.3741364759132064,
-      "learning_rate": 7.472401874077593e-06,
-      "loss": 0.0159,
-      "step": 998
-    },
-    {
-      "epoch": 11.892857142857142,
-      "grad_norm": 1.7029666602982898,
-      "learning_rate": 7.453747395263932e-06,
-      "loss": 0.0197,
-      "step": 999
-    },
-    {
-      "epoch": 11.904761904761905,
-      "grad_norm": 3.210865127366338,
-      "learning_rate": 7.435102386645421e-06,
-      "loss": 0.0172,
-      "step": 1000
-    },
-    {
-      "epoch": 11.916666666666666,
-      "grad_norm": 2.230829971481931,
-      "learning_rate": 7.4164669175678376e-06,
-      "loss": 0.0169,
-      "step": 1001
-    },
-    {
-      "epoch": 11.928571428571429,
-      "grad_norm": 2.19343014063508,
-      "learning_rate": 7.3978410573414795e-06,
-      "loss": 0.0337,
-      "step": 1002
-    },
-    {
-      "epoch": 11.94047619047619,
-      "grad_norm": 1.8146120196303304,
-      "learning_rate": 7.3792248752409116e-06,
-      "loss": 0.0147,
-      "step": 1003
-    },
-    {
-      "epoch": 11.952380952380953,
-      "grad_norm": 2.044635364554011,
-      "learning_rate": 7.360618440504695e-06,
-      "loss": 0.0273,
-      "step": 1004
-    },
-    {
-      "epoch": 11.964285714285714,
-      "grad_norm": 1.5594003161322363,
-      "learning_rate": 7.342021822335144e-06,
-      "loss": 0.0408,
-      "step": 1005
-    },
-    {
-      "epoch": 11.976190476190476,
-      "grad_norm": 1.9622001194940828,
-      "learning_rate": 7.323435089898059e-06,
-      "loss": 0.0194,
-      "step": 1006
-    },
-    {
-      "epoch": 11.988095238095237,
-      "grad_norm": 1.1152284378601478,
-      "learning_rate": 7.3048583123224745e-06,
-      "loss": 0.0067,
-      "step": 1007
-    },
-    {
-      "epoch": 12.0,
-      "grad_norm": 1.4220342011365041,
-      "learning_rate": 7.2862915587004e-06,
-      "loss": 0.0099,
-      "step": 1008
-    },
-    {
-      "epoch": 12.0,
-      "eval_loss": 0.09689878672361374,
-      "eval_runtime": 40.9326,
-      "eval_samples_per_second": 1.466,
-      "eval_steps_per_second": 1.466,
-      "step": 1008
-    },
-    {
-      "epoch": 12.011904761904763,
-      "grad_norm": 1.0021159288146593,
-      "learning_rate": 7.267734898086565e-06,
-      "loss": 0.0068,
-      "step": 1009
-    },
-    {
-      "epoch": 12.023809523809524,
-      "grad_norm": 1.2619101524108916,
-      "learning_rate": 7.249188399498159e-06,
-      "loss": 0.0117,
-      "step": 1010
-    },
-    {
-      "epoch": 12.035714285714286,
-      "grad_norm": 0.734941316803662,
-      "learning_rate": 7.230652131914574e-06,
-      "loss": 0.0195,
-      "step": 1011
-    },
-    {
-      "epoch": 12.047619047619047,
-      "grad_norm": 2.079991861318786,
-      "learning_rate": 7.212126164277151e-06,
-      "loss": 0.0241,
-      "step": 1012
-    },
-    {
-      "epoch": 12.05952380952381,
-      "grad_norm": 3.0509428341350437,
-      "learning_rate": 7.1936105654889245e-06,
-      "loss": 0.0689,
-      "step": 1013
-    },
-    {
-      "epoch": 12.071428571428571,
-      "grad_norm": 2.619913977958957,
-      "learning_rate": 7.1751054044143616e-06,
-      "loss": 0.0477,
-      "step": 1014
-    },
-    {
-      "epoch": 12.083333333333334,
-      "grad_norm": 1.151849460495169,
-      "learning_rate": 7.156610749879116e-06,
-      "loss": 0.0156,
-      "step": 1015
-    },
-    {
-      "epoch": 12.095238095238095,
-      "grad_norm": 2.5505348766082467,
-      "learning_rate": 7.1381266706697564e-06,
-      "loss": 0.0425,
-      "step": 1016
-    },
-    {
-      "epoch": 12.107142857142858,
-      "grad_norm": 1.7373338551385542,
-      "learning_rate": 7.11965323553352e-06,
-      "loss": 0.0172,
-      "step": 1017
-    },
-    {
-      "epoch": 12.119047619047619,
-      "grad_norm": 2.940052483151144,
-      "learning_rate": 7.10119051317806e-06,
-      "loss": 0.0497,
-      "step": 1018
-    },
-    {
-      "epoch": 12.130952380952381,
-      "grad_norm": 3.3383244130827414,
-      "learning_rate": 7.082738572271185e-06,
-      "loss": 0.0464,
-      "step": 1019
-    },
-    {
-      "epoch": 12.142857142857142,
-      "grad_norm": 2.0013889730979533,
-      "learning_rate": 7.064297481440601e-06,
-      "loss": 0.0327,
-      "step": 1020
-    },
-    {
-      "epoch": 12.154761904761905,
-      "grad_norm": 1.8008055977885011,
-      "learning_rate": 7.045867309273664e-06,
-      "loss": 0.0357,
-      "step": 1021
-    },
-    {
-      "epoch": 12.166666666666666,
-      "grad_norm": 1.7122862048563616,
-      "learning_rate": 7.02744812431712e-06,
-      "loss": 0.0165,
-      "step": 1022
-    },
-    {
-      "epoch": 12.178571428571429,
-      "grad_norm": 0.5421740170140353,
-      "learning_rate": 7.009039995076845e-06,
-      "loss": 0.0102,
-      "step": 1023
-    },
-    {
-      "epoch": 12.19047619047619,
-      "grad_norm": 3.3659274571341875,
-      "learning_rate": 6.990642990017603e-06,
-      "loss": 0.0257,
-      "step": 1024
-    },
-    {
-      "epoch": 12.202380952380953,
-      "grad_norm": 2.6776534186312615,
-      "learning_rate": 6.97225717756278e-06,
-      "loss": 0.0464,
-      "step": 1025
-    },
-    {
-      "epoch": 12.214285714285714,
-      "grad_norm": 1.89868056337621,
-      "learning_rate": 6.953882626094136e-06,
-      "loss": 0.0363,
-      "step": 1026
-    },
-    {
-      "epoch": 12.226190476190476,
-      "grad_norm": 1.6557758156145048,
-      "learning_rate": 6.935519403951549e-06,
-      "loss": 0.0241,
-      "step": 1027
-    },
-    {
-      "epoch": 12.238095238095237,
-      "grad_norm": 2.6942588170276003,
-      "learning_rate": 6.917167579432753e-06,
-      "loss": 0.0116,
-      "step": 1028
-    },
-    {
-      "epoch": 12.25,
-      "grad_norm": 4.021190726312291,
-      "learning_rate": 6.898827220793103e-06,
-      "loss": 0.0134,
-      "step": 1029
-    },
-    {
-      "epoch": 12.261904761904763,
-      "grad_norm": 6.859494774187862,
-      "learning_rate": 6.880498396245298e-06,
-      "loss": 0.0254,
-      "step": 1030
-    },
-    {
-      "epoch": 12.273809523809524,
-      "grad_norm": 1.8169990791501511,
-      "learning_rate": 6.862181173959146e-06,
-      "loss": 0.0259,
-      "step": 1031
-    },
-    {
-      "epoch": 12.285714285714286,
-      "grad_norm": 2.420725120451796,
-      "learning_rate": 6.8438756220613045e-06,
-      "loss": 0.059,
-      "step": 1032
-    },
-    {
-      "epoch": 12.297619047619047,
-      "grad_norm": 3.0165246425408396,
-      "learning_rate": 6.825581808635016e-06,
-      "loss": 0.0133,
-      "step": 1033
-    },
-    {
-      "epoch": 12.30952380952381,
-      "grad_norm": 2.6905776992698254,
-      "learning_rate": 6.807299801719871e-06,
-      "loss": 0.0389,
-      "step": 1034
-    },
-    {
-      "epoch": 12.321428571428571,
-      "grad_norm": 1.7274248581300429,
-      "learning_rate": 6.789029669311551e-06,
-      "loss": 0.0077,
-      "step": 1035
-    },
-    {
-      "epoch": 12.333333333333334,
-      "grad_norm": 1.8001126965802912,
-      "learning_rate": 6.770771479361568e-06,
-      "loss": 0.0344,
-      "step": 1036
-    },
-    {
-      "epoch": 12.345238095238095,
-      "grad_norm": 2.7849018104793024,
-      "learning_rate": 6.752525299777021e-06,
-      "loss": 0.0181,
-      "step": 1037
-    },
-    {
-      "epoch": 12.357142857142858,
-      "grad_norm": 0.5733123316569594,
-      "learning_rate": 6.734291198420333e-06,
-      "loss": 0.0033,
-      "step": 1038
-    },
-    {
-      "epoch": 12.369047619047619,
-      "grad_norm": 3.4183200124225133,
-      "learning_rate": 6.716069243109011e-06,
-      "loss": 0.0348,
-      "step": 1039
-    },
-    {
-      "epoch": 12.380952380952381,
-      "grad_norm": 2.19111813065019,
-      "learning_rate": 6.697859501615387e-06,
-      "loss": 0.0262,
-      "step": 1040
-    },
-    {
-      "epoch": 12.392857142857142,
-      "grad_norm": 1.413820906012132,
-      "learning_rate": 6.679662041666361e-06,
-      "loss": 0.0157,
-      "step": 1041
-    },
-    {
-      "epoch": 12.404761904761905,
-      "grad_norm": 2.1775312629739765,
-      "learning_rate": 6.661476930943162e-06,
-      "loss": 0.0172,
-      "step": 1042
-    },
-    {
-      "epoch": 12.416666666666666,
-      "grad_norm": 3.0659515472526655,
-      "learning_rate": 6.643304237081088e-06,
-      "loss": 0.0335,
-      "step": 1043
-    },
-    {
-      "epoch": 12.428571428571429,
-      "grad_norm": 1.8628049550491268,
-      "learning_rate": 6.625144027669246e-06,
-      "loss": 0.0222,
-      "step": 1044
-    },
-    {
-      "epoch": 12.44047619047619,
-      "grad_norm": 0.7175217914778912,
-      "learning_rate": 6.606996370250319e-06,
-      "loss": 0.0092,
-      "step": 1045
-    },
-    {
-      "epoch": 12.452380952380953,
-      "grad_norm": 1.478121964073369,
-      "learning_rate": 6.588861332320307e-06,
-      "loss": 0.037,
-      "step": 1046
-    },
-    {
-      "epoch": 12.464285714285714,
-      "grad_norm": 2.3923451246112717,
-      "learning_rate": 6.570738981328266e-06,
-      "loss": 0.0746,
-      "step": 1047
-    },
-    {
-      "epoch": 12.476190476190476,
-      "grad_norm": 4.7820280255811225,
-      "learning_rate": 6.552629384676079e-06,
-      "loss": 0.0587,
-      "step": 1048
-    },
-    {
-      "epoch": 12.488095238095237,
-      "grad_norm": 5.042407963332626,
-      "learning_rate": 6.534532609718177e-06,
-      "loss": 0.0741,
-      "step": 1049
-    },
-    {
-      "epoch": 12.5,
-      "grad_norm": 1.992091322531725,
-      "learning_rate": 6.516448723761315e-06,
-      "loss": 0.0127,
-      "step": 1050
-    },
-    {
-      "epoch": 12.511904761904763,
-      "grad_norm": 2.0251462511451317,
-      "learning_rate": 6.498377794064303e-06,
-      "loss": 0.0354,
-      "step": 1051
-    },
-    {
-      "epoch": 12.523809523809524,
-      "grad_norm": 3.374309680561493,
-      "learning_rate": 6.480319887837772e-06,
-      "loss": 0.0326,
-      "step": 1052
-    },
-    {
-      "epoch": 12.535714285714286,
-      "grad_norm": 3.050395187529473,
-      "learning_rate": 6.4622750722439075e-06,
-      "loss": 0.022,
-      "step": 1053
-    },
-    {
-      "epoch": 12.547619047619047,
-      "grad_norm": 11.589276409101357,
-      "learning_rate": 6.4442434143962075e-06,
-      "loss": 0.054,
-      "step": 1054
-    },
-    {
-      "epoch": 12.55952380952381,
-      "grad_norm": 2.3449829130817137,
-      "learning_rate": 6.426224981359238e-06,
-      "loss": 0.0493,
-      "step": 1055
-    },
-    {
-      "epoch": 12.571428571428571,
-      "grad_norm": 2.44302873820673,
-      "learning_rate": 6.4082198401483755e-06,
-      "loss": 0.0238,
-      "step": 1056
-    },
-    {
-      "epoch": 12.583333333333334,
-      "grad_norm": 1.1669955142868396,
-      "learning_rate": 6.390228057729557e-06,
-      "loss": 0.0186,
-      "step": 1057
-    },
-    {
-      "epoch": 12.595238095238095,
-      "grad_norm": 0.9590754842243425,
-      "learning_rate": 6.372249701019044e-06,
-      "loss": 0.0099,
-      "step": 1058
-    },
-    {
-      "epoch": 12.607142857142858,
-      "grad_norm": 1.0699806800229257,
-      "learning_rate": 6.354284836883156e-06,
-      "loss": 0.0174,
-      "step": 1059
-    },
-    {
-      "epoch": 12.619047619047619,
-      "grad_norm": 2.8391167747003987,
-      "learning_rate": 6.336333532138032e-06,
-      "loss": 0.045,
-      "step": 1060
-    },
-    {
-      "epoch": 12.630952380952381,
-      "grad_norm": 0.6894773072049841,
-      "learning_rate": 6.318395853549379e-06,
-      "loss": 0.0041,
-      "step": 1061
-    },
-    {
-      "epoch": 12.642857142857142,
-      "grad_norm": 4.84209428246548,
-      "learning_rate": 6.30047186783223e-06,
-      "loss": 0.045,
-      "step": 1062
-    },
-    {
-      "epoch": 12.654761904761905,
-      "grad_norm": 0.7448463992353114,
-      "learning_rate": 6.282561641650682e-06,
-      "loss": 0.0094,
-      "step": 1063
-    },
-    {
-      "epoch": 12.666666666666666,
-      "grad_norm": 1.8024616854906685,
-      "learning_rate": 6.2646652416176665e-06,
-      "loss": 0.0093,
-      "step": 1064
-    },
-    {
-      "epoch": 12.678571428571429,
-      "grad_norm": 1.3429782506913301,
-      "learning_rate": 6.246782734294683e-06,
-      "loss": 0.0172,
-      "step": 1065
-    },
-    {
-      "epoch": 12.69047619047619,
-      "grad_norm": 2.963838717143966,
-      "learning_rate": 6.2289141861915635e-06,
-      "loss": 0.0208,
-      "step": 1066
-    },
-    {
-      "epoch": 12.702380952380953,
-      "grad_norm": 2.5846144874264043,
-      "learning_rate": 6.211059663766224e-06,
-      "loss": 0.0402,
-      "step": 1067
-    },
-    {
-      "epoch": 12.714285714285714,
-      "grad_norm": 2.560181092214124,
-      "learning_rate": 6.193219233424414e-06,
-      "loss": 0.0288,
-      "step": 1068
-    },
-    {
-      "epoch": 12.726190476190476,
-      "grad_norm": 2.0384912620129194,
-      "learning_rate": 6.1753929615194705e-06,
-      "loss": 0.0195,
-      "step": 1069
-    },
-    {
-      "epoch": 12.738095238095237,
-      "grad_norm": 1.4052060125033978,
-      "learning_rate": 6.15758091435207e-06,
-      "loss": 0.0074,
-      "step": 1070
-    },
-    {
-      "epoch": 12.75,
-      "grad_norm": 4.685196377107016,
-      "learning_rate": 6.139783158169985e-06,
-      "loss": 0.0337,
-      "step": 1071
-    },
-    {
-      "epoch": 12.761904761904763,
-      "grad_norm": 0.8218476974088521,
-      "learning_rate": 6.121999759167837e-06,
-      "loss": 0.0044,
-      "step": 1072
-    },
-    {
-      "epoch": 12.773809523809524,
-      "grad_norm": 4.0846199910982115,
-      "learning_rate": 6.104230783486846e-06,
-      "loss": 0.0523,
-      "step": 1073
-    },
-    {
-      "epoch": 12.785714285714286,
-      "grad_norm": 1.089610236862446,
-      "learning_rate": 6.086476297214594e-06,
-      "loss": 0.0127,
-      "step": 1074
-    },
-    {
-      "epoch": 12.797619047619047,
-      "grad_norm": 0.5229904428811587,
-      "learning_rate": 6.068736366384764e-06,
-      "loss": 0.0108,
-      "step": 1075
-    },
-    {
-      "epoch": 12.80952380952381,
-      "grad_norm": 0.5295945123203731,
-      "learning_rate": 6.05101105697691e-06,
-      "loss": 0.0069,
-      "step": 1076
-    },
-    {
-      "epoch": 12.821428571428571,
-      "grad_norm": 0.40635985353282955,
-      "learning_rate": 6.0333004349162025e-06,
-      "loss": 0.0065,
-      "step": 1077
-    },
-    {
-      "epoch": 12.833333333333334,
-      "grad_norm": 2.4730189484544107,
-      "learning_rate": 6.015604566073187e-06,
-      "loss": 0.013,
-      "step": 1078
-    },
-    {
-      "epoch": 12.845238095238095,
-      "grad_norm": 0.5170708412524836,
-      "learning_rate": 5.99792351626354e-06,
-      "loss": 0.0078,
-      "step": 1079
-    },
-    {
-      "epoch": 12.857142857142858,
-      "grad_norm": 0.7722760595256714,
-      "learning_rate": 5.980257351247818e-06,
-      "loss": 0.0122,
-      "step": 1080
-    },
-    {
-      "epoch": 12.869047619047619,
-      "grad_norm": 0.9562287992871482,
-      "learning_rate": 5.962606136731217e-06,
-      "loss": 0.015,
-      "step": 1081
-    },
-    {
-      "epoch": 12.880952380952381,
-      "grad_norm": 2.57487715422707,
-      "learning_rate": 5.944969938363332e-06,
-      "loss": 0.0558,
-      "step": 1082
-    },
-    {
-      "epoch": 12.892857142857142,
-      "grad_norm": 5.175353888734591,
-      "learning_rate": 5.927348821737906e-06,
-      "loss": 0.0775,
-      "step": 1083
-    },
-    {
-      "epoch": 12.904761904761905,
-      "grad_norm": 3.3863214768962635,
-      "learning_rate": 5.9097428523925874e-06,
-      "loss": 0.0587,
-      "step": 1084
-    },
-    {
-      "epoch": 12.916666666666666,
-      "grad_norm": 2.3710552627489503,
-      "learning_rate": 5.892152095808691e-06,
-      "loss": 0.0114,
-      "step": 1085
-    },
-    {
-      "epoch": 12.928571428571429,
-      "grad_norm": 1.7382649132331074,
-      "learning_rate": 5.87457661741095e-06,
-      "loss": 0.016,
-      "step": 1086
-    },
-    {
-      "epoch": 12.94047619047619,
-      "grad_norm": 0.7673832218317043,
-      "learning_rate": 5.857016482567275e-06,
-      "loss": 0.0144,
-      "step": 1087
-    },
-    {
-      "epoch": 12.952380952380953,
-      "grad_norm": 1.1565970722585457,
-      "learning_rate": 5.83947175658851e-06,
-      "loss": 0.016,
-      "step": 1088
-    },
-    {
-      "epoch": 12.964285714285714,
-      "grad_norm": 1.3519643426410404,
-      "learning_rate": 5.821942504728183e-06,
-      "loss": 0.0254,
-      "step": 1089
-    },
-    {
-      "epoch": 12.976190476190476,
-      "grad_norm": 2.91693108261119,
-      "learning_rate": 5.80442879218228e-06,
-      "loss": 0.0419,
-      "step": 1090
-    },
-    {
-      "epoch": 12.988095238095237,
-      "grad_norm": 1.747082227614345,
-      "learning_rate": 5.786930684088988e-06,
-      "loss": 0.0192,
-      "step": 1091
-    },
-    {
-      "epoch": 13.0,
-      "grad_norm": 2.0947996297444966,
-      "learning_rate": 5.769448245528451e-06,
-      "loss": 0.0347,
-      "step": 1092
-    },
-    {
-      "epoch": 13.0,
-      "eval_loss": 0.16690757870674133,
-      "eval_runtime": 41.0419,
-      "eval_samples_per_second": 1.462,
-      "eval_steps_per_second": 1.462,
-      "step": 1092
-    }
-  ],
-  "logging_steps": 1.0,
-  "max_steps": 1680,
-  "num_input_tokens_seen": 0,
-  "num_train_epochs": 20,
-  "save_steps": 500,
-  "stateful_callbacks": {
-    "TrainerControl": {
-      "args": {
-        "should_epoch_stop": false,
-        "should_evaluate": false,
-        "should_log": false,
-        "should_save": true,
-        "should_training_stop": false
-      },
-      "attributes": {}
-    }
-  },
-  "total_flos": 304017211392000.0,
-  "train_batch_size": 1,
-  "trial_name": null,
-  "trial_params": null
-}