{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 5.0,
  "eval_steps": 83,
  "global_step": 415,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.012048192771084338,
      "grad_norm": 0.3426714241504669,
      "learning_rate": 0.0,
      "loss": 1.35498046875,
      "step": 1
    },
    {
      "epoch": 0.024096385542168676,
      "grad_norm": 0.267749160528183,
      "learning_rate": 4.7619047619047615e-06,
      "loss": 1.111328125,
      "step": 2
    },
    {
      "epoch": 0.03614457831325301,
      "grad_norm": 0.29530617594718933,
      "learning_rate": 9.523809523809523e-06,
      "loss": 1.376953125,
      "step": 3
    },
    {
      "epoch": 0.04819277108433735,
      "grad_norm": 0.32084405422210693,
      "learning_rate": 1.4285714285714285e-05,
      "loss": 1.408203125,
      "step": 4
    },
    {
      "epoch": 0.060240963855421686,
      "grad_norm": 0.3341103196144104,
      "learning_rate": 1.9047619047619046e-05,
      "loss": 1.232421875,
      "step": 5
    },
    {
      "epoch": 0.07228915662650602,
      "grad_norm": 0.2828045189380646,
      "learning_rate": 2.380952380952381e-05,
      "loss": 1.1787109375,
      "step": 6
    },
    {
      "epoch": 0.08433734939759036,
      "grad_norm": 0.1980012059211731,
      "learning_rate": 2.857142857142857e-05,
      "loss": 1.1982421875,
      "step": 7
    },
    {
      "epoch": 0.0963855421686747,
      "grad_norm": 0.11467871814966202,
      "learning_rate": 3.3333333333333335e-05,
      "loss": 1.1357421875,
      "step": 8
    },
    {
      "epoch": 0.10843373493975904,
      "grad_norm": 0.08303967863321304,
      "learning_rate": 3.809523809523809e-05,
      "loss": 1.0966796875,
      "step": 9
    },
    {
      "epoch": 0.12048192771084337,
      "grad_norm": 0.07549171894788742,
      "learning_rate": 4.2857142857142856e-05,
      "loss": 1.111328125,
      "step": 10
    },
    {
      "epoch": 0.13253012048192772,
      "grad_norm": 0.08018944412469864,
      "learning_rate": 4.761904761904762e-05,
      "loss": 1.03955078125,
      "step": 11
    },
    {
      "epoch": 0.14457831325301204,
      "grad_norm": 0.08259129524230957,
      "learning_rate": 5.2380952380952384e-05,
      "loss": 1.0576171875,
      "step": 12
    },
    {
      "epoch": 0.1566265060240964,
      "grad_norm": 0.08708677440881729,
      "learning_rate": 5.714285714285714e-05,
      "loss": 1.1103515625,
      "step": 13
    },
    {
      "epoch": 0.1686746987951807,
      "grad_norm": 0.10218311101198196,
      "learning_rate": 6.19047619047619e-05,
      "loss": 1.029296875,
      "step": 14
    },
    {
      "epoch": 0.18072289156626506,
      "grad_norm": 0.10239769518375397,
      "learning_rate": 6.666666666666667e-05,
      "loss": 1.07470703125,
      "step": 15
    },
    {
      "epoch": 0.1927710843373494,
      "grad_norm": 0.09094081819057465,
      "learning_rate": 7.142857142857143e-05,
      "loss": 1.0126953125,
      "step": 16
    },
    {
      "epoch": 0.20481927710843373,
      "grad_norm": 0.07679731398820877,
      "learning_rate": 7.619047619047618e-05,
      "loss": 0.99365234375,
      "step": 17
    },
    {
      "epoch": 0.21686746987951808,
      "grad_norm": 0.07608456909656525,
      "learning_rate": 8.095238095238096e-05,
      "loss": 1.09326171875,
      "step": 18
    },
    {
      "epoch": 0.2289156626506024,
      "grad_norm": 0.06455695629119873,
      "learning_rate": 8.571428571428571e-05,
      "loss": 1.0546875,
      "step": 19
    },
    {
      "epoch": 0.24096385542168675,
      "grad_norm": 0.06312122195959091,
      "learning_rate": 9.047619047619048e-05,
      "loss": 0.994140625,
      "step": 20
    },
    {
      "epoch": 0.25301204819277107,
      "grad_norm": 0.05908634513616562,
      "learning_rate": 9.523809523809524e-05,
      "loss": 1.06884765625,
      "step": 21
    },
    {
      "epoch": 0.26506024096385544,
      "grad_norm": 0.047742899507284164,
      "learning_rate": 0.0001,
      "loss": 0.828369140625,
      "step": 22
    },
    {
      "epoch": 0.27710843373493976,
      "grad_norm": 0.05446731299161911,
      "learning_rate": 9.999856950113066e-05,
      "loss": 0.99267578125,
      "step": 23
    },
    {
      "epoch": 0.2891566265060241,
      "grad_norm": 0.05094271898269653,
      "learning_rate": 9.999427809547046e-05,
      "loss": 0.92822265625,
      "step": 24
    },
    {
      "epoch": 0.30120481927710846,
      "grad_norm": 0.05215759575366974,
      "learning_rate": 9.998712605585727e-05,
      "loss": 0.863525390625,
      "step": 25
    },
    {
      "epoch": 0.3132530120481928,
      "grad_norm": 0.05302610993385315,
      "learning_rate": 9.997711383700148e-05,
      "loss": 0.9326171875,
      "step": 26
    },
    {
      "epoch": 0.3253012048192771,
      "grad_norm": 0.04989799112081528,
      "learning_rate": 9.996424207545721e-05,
      "loss": 1.03125,
      "step": 27
    },
    {
      "epoch": 0.3373493975903614,
      "grad_norm": 0.046004582196474075,
      "learning_rate": 9.994851158958184e-05,
      "loss": 0.916015625,
      "step": 28
    },
    {
      "epoch": 0.3493975903614458,
      "grad_norm": 0.03904594108462334,
      "learning_rate": 9.992992337948388e-05,
      "loss": 1.0302734375,
      "step": 29
    },
    {
      "epoch": 0.3614457831325301,
      "grad_norm": 0.035267237573862076,
      "learning_rate": 9.990847862695951e-05,
      "loss": 1.0078125,
      "step": 30
    },
    {
      "epoch": 0.37349397590361444,
      "grad_norm": 0.040470998734235764,
      "learning_rate": 9.988417869541736e-05,
      "loss": 0.8857421875,
      "step": 31
    },
    {
      "epoch": 0.3855421686746988,
      "grad_norm": 0.03800118714570999,
      "learning_rate": 9.985702512979186e-05,
      "loss": 0.962890625,
      "step": 32
    },
    {
      "epoch": 0.39759036144578314,
      "grad_norm": 0.041127946227788925,
      "learning_rate": 9.982701965644499e-05,
      "loss": 0.978515625,
      "step": 33
    },
    {
      "epoch": 0.40963855421686746,
      "grad_norm": 0.040262386202812195,
      "learning_rate": 9.979416418305655e-05,
      "loss": 0.9921875,
      "step": 34
    },
    {
      "epoch": 0.42168674698795183,
      "grad_norm": 0.03574175760149956,
      "learning_rate": 9.975846079850291e-05,
      "loss": 0.89111328125,
      "step": 35
    },
    {
      "epoch": 0.43373493975903615,
      "grad_norm": 0.03455124795436859,
      "learning_rate": 9.97199117727241e-05,
      "loss": 0.8896484375,
      "step": 36
    },
    {
      "epoch": 0.4457831325301205,
      "grad_norm": 0.03667663037776947,
      "learning_rate": 9.967851955657959e-05,
      "loss": 0.81298828125,
      "step": 37
    },
    {
      "epoch": 0.4578313253012048,
      "grad_norm": 0.03574740141630173,
      "learning_rate": 9.963428678169242e-05,
      "loss": 0.93896484375,
      "step": 38
    },
    {
      "epoch": 0.46987951807228917,
      "grad_norm": 0.03379588946700096,
      "learning_rate": 9.958721626028189e-05,
      "loss": 0.94580078125,
      "step": 39
    },
    {
      "epoch": 0.4819277108433735,
      "grad_norm": 0.03961994871497154,
      "learning_rate": 9.953731098498477e-05,
      "loss": 0.94677734375,
      "step": 40
    },
    {
      "epoch": 0.4939759036144578,
      "grad_norm": 0.03511177748441696,
      "learning_rate": 9.948457412866511e-05,
      "loss": 0.8525390625,
      "step": 41
    },
    {
      "epoch": 0.5060240963855421,
      "grad_norm": 0.033400677144527435,
      "learning_rate": 9.942900904421233e-05,
      "loss": 0.8955078125,
      "step": 42
    },
    {
      "epoch": 0.5180722891566265,
      "grad_norm": 0.19605278968811035,
      "learning_rate": 9.937061926432826e-05,
      "loss": 0.73614501953125,
      "step": 43
    },
    {
      "epoch": 0.5301204819277109,
      "grad_norm": 0.03378308564424515,
      "learning_rate": 9.930940850130241e-05,
      "loss": 0.86962890625,
      "step": 44
    },
    {
      "epoch": 0.5421686746987951,
      "grad_norm": 0.03614725172519684,
      "learning_rate": 9.9245380646776e-05,
      "loss": 0.8740234375,
      "step": 45
    },
    {
      "epoch": 0.5542168674698795,
      "grad_norm": 0.0337648019194603,
      "learning_rate": 9.91785397714945e-05,
      "loss": 0.94970703125,
      "step": 46
    },
    {
      "epoch": 0.5662650602409639,
      "grad_norm": 0.03873274847865105,
      "learning_rate": 9.910889012504889e-05,
      "loss": 0.984375,
      "step": 47
    },
    {
      "epoch": 0.5783132530120482,
      "grad_norm": 0.03375189006328583,
      "learning_rate": 9.90364361356054e-05,
      "loss": 0.91455078125,
      "step": 48
    },
    {
      "epoch": 0.5903614457831325,
      "grad_norm": 0.03698453679680824,
      "learning_rate": 9.896118240962404e-05,
      "loss": 0.82177734375,
      "step": 49
    },
    {
      "epoch": 0.6024096385542169,
      "grad_norm": 0.03549083322286606,
      "learning_rate": 9.888313373156568e-05,
      "loss": 0.8505859375,
      "step": 50
    },
    {
      "epoch": 0.6144578313253012,
      "grad_norm": 0.03702399879693985,
      "learning_rate": 9.880229506358792e-05,
      "loss": 0.85546875,
      "step": 51
    },
    {
      "epoch": 0.6265060240963856,
      "grad_norm": 0.03412725403904915,
      "learning_rate": 9.871867154522957e-05,
      "loss": 0.8720703125,
      "step": 52
    },
    {
      "epoch": 0.6385542168674698,
      "grad_norm": 0.03455042839050293,
      "learning_rate": 9.863226849308387e-05,
      "loss": 0.84326171875,
      "step": 53
    },
    {
      "epoch": 0.6506024096385542,
      "grad_norm": 0.034792136400938034,
      "learning_rate": 9.854309140046055e-05,
      "loss": 0.9248046875,
      "step": 54
    },
    {
      "epoch": 0.6626506024096386,
      "grad_norm": 0.038040246814489365,
      "learning_rate": 9.84511459370365e-05,
      "loss": 0.80810546875,
      "step": 55
    },
    {
      "epoch": 0.6746987951807228,
      "grad_norm": 0.033053476363420486,
      "learning_rate": 9.835643794849534e-05,
      "loss": 0.85302734375,
      "step": 56
    },
    {
      "epoch": 0.6867469879518072,
      "grad_norm": 0.0341220498085022,
      "learning_rate": 9.825897345615577e-05,
      "loss": 0.90673828125,
      "step": 57
    },
    {
      "epoch": 0.6987951807228916,
      "grad_norm": 0.03554167598485947,
      "learning_rate": 9.815875865658873e-05,
      "loss": 0.8984375,
      "step": 58
    },
    {
      "epoch": 0.7108433734939759,
      "grad_norm": 0.030804554000496864,
      "learning_rate": 9.805579992122339e-05,
      "loss": 0.9384765625,
      "step": 59
    },
    {
      "epoch": 0.7228915662650602,
      "grad_norm": 0.04017789289355278,
      "learning_rate": 9.795010379594226e-05,
      "loss": 0.8251953125,
      "step": 60
    },
    {
      "epoch": 0.7349397590361446,
      "grad_norm": 0.03208062797784805,
      "learning_rate": 9.784167700066472e-05,
      "loss": 0.8935546875,
      "step": 61
    },
    {
      "epoch": 0.7469879518072289,
      "grad_norm": 0.037682078778743744,
      "learning_rate": 9.773052642892003e-05,
      "loss": 0.83447265625,
      "step": 62
    },
    {
      "epoch": 0.7590361445783133,
      "grad_norm": 0.03160431608557701,
      "learning_rate": 9.761665914740899e-05,
      "loss": 0.701416015625,
      "step": 63
    },
    {
      "epoch": 0.7710843373493976,
      "grad_norm": 0.0371614545583725,
      "learning_rate": 9.750008239555457e-05,
      "loss": 0.8154296875,
      "step": 64
    },
    {
      "epoch": 0.7831325301204819,
      "grad_norm": 0.033163782209157944,
      "learning_rate": 9.738080358504176e-05,
      "loss": 0.9482421875,
      "step": 65
    },
    {
      "epoch": 0.7951807228915663,
      "grad_norm": 0.03720880672335625,
      "learning_rate": 9.725883029934624e-05,
      "loss": 0.7802734375,
      "step": 66
    },
    {
      "epoch": 0.8072289156626506,
      "grad_norm": 0.032983068376779556,
      "learning_rate": 9.713417029325235e-05,
      "loss": 0.87255859375,
      "step": 67
    },
    {
      "epoch": 0.8192771084337349,
      "grad_norm": 0.03193071112036705,
      "learning_rate": 9.700683149235999e-05,
      "loss": 0.62567138671875,
      "step": 68
    },
    {
      "epoch": 0.8313253012048193,
      "grad_norm": 0.03492351621389389,
      "learning_rate": 9.687682199258074e-05,
      "loss": 0.80859375,
      "step": 69
    },
    {
      "epoch": 0.8433734939759037,
      "grad_norm": 0.04678158462047577,
      "learning_rate": 9.674415005962318e-05,
      "loss": 0.78369140625,
      "step": 70
    },
    {
      "epoch": 0.8554216867469879,
      "grad_norm": 0.033582933247089386,
      "learning_rate": 9.660882412846729e-05,
      "loss": 0.8359375,
      "step": 71
    },
    {
      "epoch": 0.8674698795180723,
      "grad_norm": 0.03978513181209564,
      "learning_rate": 9.647085280282826e-05,
      "loss": 0.8193359375,
      "step": 72
    },
    {
      "epoch": 0.8795180722891566,
      "grad_norm": 0.03892577439546585,
      "learning_rate": 9.633024485460941e-05,
      "loss": 0.79833984375,
      "step": 73
    },
    {
      "epoch": 0.891566265060241,
      "grad_norm": 0.03606899827718735,
      "learning_rate": 9.618700922334459e-05,
      "loss": 0.837890625,
      "step": 74
    },
    {
      "epoch": 0.9036144578313253,
      "grad_norm": 0.04031959921121597,
      "learning_rate": 9.604115501562972e-05,
      "loss": 0.85400390625,
      "step": 75
    },
    {
      "epoch": 0.9156626506024096,
      "grad_norm": 0.03623352572321892,
      "learning_rate": 9.58926915045439e-05,
      "loss": 0.81982421875,
      "step": 76
    },
    {
      "epoch": 0.927710843373494,
      "grad_norm": 0.036078643053770065,
      "learning_rate": 9.574162812905977e-05,
      "loss": 0.91259765625,
      "step": 77
    },
    {
      "epoch": 0.9397590361445783,
      "grad_norm": 0.03492267429828644,
      "learning_rate": 9.558797449344345e-05,
      "loss": 0.8681640625,
      "step": 78
    },
    {
      "epoch": 0.9518072289156626,
      "grad_norm": 0.03617448732256889,
      "learning_rate": 9.543174036664392e-05,
      "loss": 0.85888671875,
      "step": 79
    },
    {
      "epoch": 0.963855421686747,
      "grad_norm": 0.03611322492361069,
      "learning_rate": 9.527293568167193e-05,
      "loss": 0.91162109375,
      "step": 80
    },
    {
      "epoch": 0.9759036144578314,
      "grad_norm": 0.08425167948007584,
      "learning_rate": 9.511157053496848e-05,
      "loss": 0.8359375,
      "step": 81
    },
    {
      "epoch": 0.9879518072289156,
      "grad_norm": 0.03861250355839729,
      "learning_rate": 9.494765518576287e-05,
      "loss": 0.86474609375,
      "step": 82
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.03831891342997551,
      "learning_rate": 9.478120005542054e-05,
      "loss": 0.7744140625,
      "step": 83
    },
    {
      "epoch": 1.0,
      "eval_loss": 0.810546875,
      "eval_runtime": 5.5664,
      "eval_samples_per_second": 1.078,
      "eval_steps_per_second": 0.18,
      "step": 83
    },
    {
      "epoch": 1.0120481927710843,
      "grad_norm": 0.03854953497648239,
      "learning_rate": 9.46122157267804e-05,
      "loss": 0.83154296875,
      "step": 84
    },
    {
      "epoch": 1.0240963855421688,
      "grad_norm": 0.0327351838350296,
      "learning_rate": 9.444071294348204e-05,
      "loss": 0.68896484375,
      "step": 85
    },
    {
      "epoch": 1.036144578313253,
      "grad_norm": 0.034474555402994156,
      "learning_rate": 9.426670260928272e-05,
      "loss": 0.888671875,
      "step": 86
    },
    {
      "epoch": 1.0481927710843373,
      "grad_norm": 0.03608318418264389,
      "learning_rate": 9.409019578736404e-05,
      "loss": 0.884765625,
      "step": 87
    },
    {
      "epoch": 1.0602409638554218,
      "grad_norm": 0.04214343801140785,
      "learning_rate": 9.39112036996286e-05,
      "loss": 0.75927734375,
      "step": 88
    },
    {
      "epoch": 1.072289156626506,
      "grad_norm": 0.0403887964785099,
      "learning_rate": 9.372973772598662e-05,
      "loss": 0.7587890625,
      "step": 89
    },
    {
      "epoch": 1.0843373493975903,
      "grad_norm": 0.04118107631802559,
      "learning_rate": 9.354580940363232e-05,
      "loss": 0.818359375,
      "step": 90
    },
    {
      "epoch": 1.0963855421686748,
      "grad_norm": 0.041133224964141846,
      "learning_rate": 9.335943042631045e-05,
      "loss": 0.7919921875,
      "step": 91
    },
    {
      "epoch": 1.108433734939759,
      "grad_norm": 0.039186157286167145,
      "learning_rate": 9.317061264357285e-05,
      "loss": 0.80029296875,
      "step": 92
    },
    {
      "epoch": 1.1204819277108433,
      "grad_norm": 0.041317328810691833,
      "learning_rate": 9.297936806002509e-05,
      "loss": 0.81884765625,
      "step": 93
    },
    {
      "epoch": 1.1325301204819278,
      "grad_norm": 0.044248566031455994,
      "learning_rate": 9.278570883456316e-05,
      "loss": 0.7509765625,
      "step": 94
    },
    {
      "epoch": 1.144578313253012,
      "grad_norm": 0.043425533920526505,
      "learning_rate": 9.258964727960053e-05,
      "loss": 0.78369140625,
      "step": 95
    },
    {
      "epoch": 1.1566265060240963,
      "grad_norm": 0.04103036969900131,
      "learning_rate": 9.239119586028535e-05,
      "loss": 0.83642578125,
      "step": 96
    },
    {
      "epoch": 1.1686746987951806,
      "grad_norm": 0.04539019241929054,
      "learning_rate": 9.219036719370786e-05,
      "loss": 0.76904296875,
      "step": 97
    },
    {
      "epoch": 1.180722891566265,
      "grad_norm": 0.04369170963764191,
      "learning_rate": 9.198717404809832e-05,
      "loss": 0.8232421875,
      "step": 98
    },
    {
      "epoch": 1.1927710843373494,
      "grad_norm": 0.04113304242491722,
      "learning_rate": 9.178162934201517e-05,
      "loss": 0.7744140625,
      "step": 99
    },
    {
      "epoch": 1.2048192771084336,
      "grad_norm": 0.04359579458832741,
      "learning_rate": 9.15737461435237e-05,
      "loss": 0.77099609375,
      "step": 100
    },
    {
      "epoch": 1.216867469879518,
      "grad_norm": 0.038704823702573776,
      "learning_rate": 9.136353766936529e-05,
      "loss": 0.87890625,
      "step": 101
    },
    {
      "epoch": 1.2289156626506024,
      "grad_norm": 0.04137009382247925,
      "learning_rate": 9.115101728411704e-05,
      "loss": 0.83984375,
      "step": 102
    },
    {
      "epoch": 1.2409638554216866,
      "grad_norm": 0.04400080069899559,
      "learning_rate": 9.093619849934208e-05,
      "loss": 0.77490234375,
      "step": 103
    },
    {
      "epoch": 1.2530120481927711,
      "grad_norm": 0.03810833767056465,
      "learning_rate": 9.071909497273059e-05,
      "loss": 0.87646484375,
      "step": 104
    },
    {
      "epoch": 1.2650602409638554,
      "grad_norm": 0.03435809537768364,
      "learning_rate": 9.049972050723142e-05,
      "loss": 0.6739501953125,
      "step": 105
    },
    {
      "epoch": 1.2771084337349397,
      "grad_norm": 0.03897792100906372,
      "learning_rate": 9.027808905017456e-05,
      "loss": 0.8046875,
      "step": 106
    },
    {
      "epoch": 1.2891566265060241,
      "grad_norm": 0.04031569883227348,
      "learning_rate": 9.005421469238441e-05,
      "loss": 0.74658203125,
      "step": 107
    },
    {
      "epoch": 1.3012048192771084,
      "grad_norm": 0.03262839838862419,
      "learning_rate": 8.98281116672839e-05,
      "loss": 0.7159423828125,
      "step": 108
    },
    {
      "epoch": 1.3132530120481927,
      "grad_norm": 0.040843553841114044,
      "learning_rate": 8.959979434998952e-05,
      "loss": 0.75830078125,
      "step": 109
    },
    {
      "epoch": 1.3253012048192772,
      "grad_norm": 0.036981672048568726,
      "learning_rate": 8.936927725639746e-05,
      "loss": 0.86181640625,
      "step": 110
    },
    {
      "epoch": 1.3373493975903614,
      "grad_norm": 0.03964255005121231,
      "learning_rate": 8.913657504226072e-05,
      "loss": 0.75732421875,
      "step": 111
    },
    {
      "epoch": 1.3493975903614457,
      "grad_norm": 0.036245595663785934,
      "learning_rate": 8.890170250225726e-05,
      "loss": 0.876953125,
      "step": 112
    },
    {
      "epoch": 1.3614457831325302,
      "grad_norm": 0.03682960197329521,
      "learning_rate": 8.86646745690494e-05,
      "loss": 0.85693359375,
      "step": 113
    },
    {
      "epoch": 1.3734939759036144,
      "grad_norm": 0.04356293007731438,
      "learning_rate": 8.842550631233451e-05,
      "loss": 0.73828125,
      "step": 114
    },
    {
      "epoch": 1.3855421686746987,
      "grad_norm": 0.03705068305134773,
      "learning_rate": 8.818421293788688e-05,
      "loss": 0.82373046875,
      "step": 115
    },
    {
      "epoch": 1.3975903614457832,
      "grad_norm": 0.037450194358825684,
      "learning_rate": 8.794080978659087e-05,
      "loss": 0.83837890625,
      "step": 116
    },
    {
      "epoch": 1.4096385542168675,
      "grad_norm": 0.03452831879258156,
      "learning_rate": 8.769531233346575e-05,
      "loss": 0.86181640625,
      "step": 117
    },
    {
      "epoch": 1.4216867469879517,
      "grad_norm": 0.04294472187757492,
      "learning_rate": 8.744773618668169e-05,
      "loss": 0.76123046875,
      "step": 118
    },
    {
      "epoch": 1.4337349397590362,
      "grad_norm": 0.03723621740937233,
      "learning_rate": 8.719809708656749e-05,
      "loss": 0.76953125,
      "step": 119
    },
    {
      "epoch": 1.4457831325301205,
      "grad_norm": 0.04655089229345322,
      "learning_rate": 8.694641090460985e-05,
      "loss": 0.68359375,
      "step": 120
    },
    {
      "epoch": 1.4578313253012047,
      "grad_norm": 0.03863748535513878,
      "learning_rate": 8.669269364244425e-05,
      "loss": 0.81298828125,
      "step": 121
    },
    {
      "epoch": 1.4698795180722892,
      "grad_norm": 0.04541182145476341,
      "learning_rate": 8.643696143083766e-05,
      "loss": 0.8173828125,
      "step": 122
    },
    {
      "epoch": 1.4819277108433735,
      "grad_norm": 0.039033275097608566,
      "learning_rate": 8.617923052866297e-05,
      "loss": 0.82861328125,
      "step": 123
    },
    {
      "epoch": 1.4939759036144578,
      "grad_norm": 0.05136038735508919,
      "learning_rate": 8.591951732186521e-05,
      "loss": 0.72314453125,
      "step": 124
    },
    {
      "epoch": 1.5060240963855422,
      "grad_norm": 0.03898656740784645,
      "learning_rate": 8.565783832241993e-05,
      "loss": 0.78369140625,
      "step": 125
    },
    {
      "epoch": 1.5180722891566265,
      "grad_norm": 0.03719359636306763,
      "learning_rate": 8.539421016728322e-05,
      "loss": 0.641021728515625,
      "step": 126
    },
    {
      "epoch": 1.5301204819277108,
      "grad_norm": 0.04546543210744858,
      "learning_rate": 8.512864961733409e-05,
      "loss": 0.75390625,
      "step": 127
    },
    {
      "epoch": 1.5421686746987953,
      "grad_norm": 0.0420440249145031,
      "learning_rate": 8.486117355630885e-05,
      "loss": 0.759765625,
      "step": 128
    },
    {
      "epoch": 1.5542168674698795,
      "grad_norm": 0.04084852710366249,
      "learning_rate": 8.459179898972763e-05,
      "loss": 0.85009765625,
      "step": 129
    },
    {
      "epoch": 1.5662650602409638,
      "grad_norm": 0.03632322698831558,
      "learning_rate": 8.43205430438132e-05,
      "loss": 0.883056640625,
      "step": 130
    },
    {
      "epoch": 1.5783132530120483,
      "grad_norm": 0.041238050907850266,
      "learning_rate": 8.40474229644022e-05,
      "loss": 0.81103515625,
      "step": 131
    },
    {
      "epoch": 1.5903614457831325,
      "grad_norm": 0.04097457602620125,
      "learning_rate": 8.377245611584862e-05,
      "loss": 0.7236328125,
      "step": 132
    },
    {
      "epoch": 1.6024096385542168,
      "grad_norm": 0.04292835295200348,
      "learning_rate": 8.349565997991982e-05,
      "loss": 0.74462890625,
      "step": 133
    },
    {
      "epoch": 1.6144578313253013,
      "grad_norm": 0.040858227759599686,
      "learning_rate": 8.321705215468514e-05,
      "loss": 0.75,
      "step": 134
    },
    {
      "epoch": 1.6265060240963856,
      "grad_norm": 0.048841387033462524,
      "learning_rate": 8.293665035339694e-05,
      "loss": 0.76904296875,
      "step": 135
    },
    {
      "epoch": 1.6385542168674698,
      "grad_norm": 0.044113028794527054,
      "learning_rate": 8.26544724033646e-05,
      "loss": 0.7353515625,
      "step": 136
    },
    {
      "epoch": 1.6506024096385543,
      "grad_norm": 0.04390498623251915,
      "learning_rate": 8.237053624482087e-05,
      "loss": 0.822265625,
      "step": 137
    },
    {
      "epoch": 1.6626506024096386,
      "grad_norm": 0.045735761523246765,
      "learning_rate": 8.208485992978153e-05,
      "loss": 0.70703125,
      "step": 138
    },
    {
      "epoch": 1.6746987951807228,
      "grad_norm": 0.042709387838840485,
      "learning_rate": 8.179746162089745e-05,
      "loss": 0.75927734375,
      "step": 139
    },
    {
      "epoch": 1.6867469879518073,
      "grad_norm": 0.04299396276473999,
      "learning_rate": 8.150835959030005e-05,
      "loss": 0.810546875,
      "step": 140
    },
    {
      "epoch": 1.6987951807228916,
      "grad_norm": 0.04076501727104187,
      "learning_rate": 8.121757221843943e-05,
      "loss": 0.8076171875,
      "step": 141
    },
    {
      "epoch": 1.7108433734939759,
      "grad_norm": 0.04018573462963104,
      "learning_rate": 8.092511799291588e-05,
      "loss": 0.853515625,
      "step": 142
    },
    {
      "epoch": 1.7228915662650603,
      "grad_norm": 0.04612232372164726,
      "learning_rate": 8.06310155073045e-05,
      "loss": 0.734375,
      "step": 143
    },
    {
      "epoch": 1.7349397590361446,
      "grad_norm": 0.04090199992060661,
      "learning_rate": 8.033528345997294e-05,
      "loss": 0.806640625,
      "step": 144
    },
    {
      "epoch": 1.7469879518072289,
      "grad_norm": 0.048365116119384766,
      "learning_rate": 8.003794065289276e-05,
      "loss": 0.74609375,
      "step": 145
    },
    {
      "epoch": 1.7590361445783134,
      "grad_norm": 0.04346657544374466,
      "learning_rate": 7.973900599044393e-05,
      "loss": 0.6346435546875,
      "step": 146
    },
    {
      "epoch": 1.7710843373493976,
      "grad_norm": 0.05970126762986183,
      "learning_rate": 7.943849847821294e-05,
      "loss": 0.72509765625,
      "step": 147
    },
    {
      "epoch": 1.783132530120482,
      "grad_norm": 0.04644714668393135,
      "learning_rate": 7.913643722178454e-05,
      "loss": 0.86572265625,
      "step": 148
    },
    {
      "epoch": 1.7951807228915664,
      "grad_norm": 0.07093939930200577,
      "learning_rate": 7.883284142552701e-05,
      "loss": 0.6953125,
      "step": 149
    },
    {
      "epoch": 1.8072289156626506,
      "grad_norm": 0.062447842210531235,
      "learning_rate": 7.852773039137115e-05,
      "loss": 0.78955078125,
      "step": 150
    },
    {
      "epoch": 1.819277108433735,
      "grad_norm": 0.049009568989276886,
      "learning_rate": 7.822112351758312e-05,
      "loss": 0.5615234375,
      "step": 151
    },
    {
      "epoch": 1.8313253012048194,
      "grad_norm": 0.05831490084528923,
      "learning_rate": 7.791304029753122e-05,
      "loss": 0.72900390625,
      "step": 152
    },
    {
      "epoch": 1.8433734939759037,
      "grad_norm": 0.05294172093272209,
      "learning_rate": 7.760350031844647e-05,
      "loss": 0.69970703125,
      "step": 153
    },
    {
      "epoch": 1.855421686746988,
      "grad_norm": 0.050907790660858154,
      "learning_rate": 7.729252326017735e-05,
      "loss": 0.75537109375,
      "step": 154
    },
    {
      "epoch": 1.8674698795180724,
      "grad_norm": 0.051166173070669174,
      "learning_rate": 7.69801288939385e-05,
      "loss": 0.73681640625,
      "step": 155
    },
    {
      "epoch": 1.8795180722891565,
      "grad_norm": 0.05042441189289093,
      "learning_rate": 7.666633708105381e-05,
      "loss": 0.71484375,
      "step": 156
    },
    {
      "epoch": 1.891566265060241,
      "grad_norm": 0.04705800488591194,
      "learning_rate": 7.635116777169372e-05,
      "loss": 0.75927734375,
      "step": 157
    },
    {
      "epoch": 1.9036144578313254,
      "grad_norm": 0.05736063793301582,
      "learning_rate": 7.603464100360662e-05,
      "loss": 0.77783203125,
      "step": 158
    },
    {
      "epoch": 1.9156626506024095,
      "grad_norm": 0.04634004458785057,
      "learning_rate": 7.57167769008452e-05,
      "loss": 0.74169921875,
      "step": 159
    },
    {
      "epoch": 1.927710843373494,
      "grad_norm": 0.04420914873480797,
      "learning_rate": 7.539759567248673e-05,
      "loss": 0.841796875,
      "step": 160
    },
    {
      "epoch": 1.9397590361445785,
      "grad_norm": 0.04374831169843674,
      "learning_rate": 7.507711761134841e-05,
      "loss": 0.796875,
      "step": 161
    },
    {
      "epoch": 1.9518072289156625,
      "grad_norm": 0.05857405066490173,
      "learning_rate": 7.475536309269705e-05,
      "loss": 0.7880859375,
      "step": 162
    },
    {
      "epoch": 1.963855421686747,
      "grad_norm": 0.043156642466783524,
      "learning_rate": 7.44323525729538e-05,
      "loss": 0.83837890625,
      "step": 163
    },
    {
      "epoch": 1.9759036144578315,
      "grad_norm": 0.04853413254022598,
      "learning_rate": 7.410810658839348e-05,
      "loss": 0.7578125,
      "step": 164
    },
    {
      "epoch": 1.9879518072289155,
      "grad_norm": 0.04652052000164986,
      "learning_rate": 7.378264575383893e-05,
      "loss": 0.79541015625,
      "step": 165
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.04825202003121376,
      "learning_rate": 7.34559907613504e-05,
      "loss": 0.6962890625,
      "step": 166
    },
    {
      "epoch": 2.0,
      "eval_loss": 0.76025390625,
      "eval_runtime": 5.5746,
      "eval_samples_per_second": 1.076,
      "eval_steps_per_second": 0.179,
      "step": 166
    },
    {
      "epoch": 2.0120481927710845,
      "grad_norm": 0.045713990926742554,
      "learning_rate": 7.312816237891001e-05,
      "loss": 0.75341796875,
      "step": 167
    },
    {
      "epoch": 2.0240963855421685,
      "grad_norm": 0.04203895479440689,
      "learning_rate": 7.279918144910134e-05,
      "loss": 0.6337890625,
      "step": 168
    },
    {
      "epoch": 2.036144578313253,
      "grad_norm": 0.04191376641392708,
      "learning_rate": 7.246906888778428e-05,
      "loss": 0.8203125,
      "step": 169
    },
    {
      "epoch": 2.0481927710843375,
      "grad_norm": 0.042297348380088806,
      "learning_rate": 7.213784568276534e-05,
      "loss": 0.81494140625,
      "step": 170
    },
    {
      "epoch": 2.0602409638554215,
      "grad_norm": 0.05187546834349632,
      "learning_rate": 7.18055328924632e-05,
      "loss": 0.68701171875,
      "step": 171
    },
    {
      "epoch": 2.072289156626506,
      "grad_norm": 0.04434656724333763,
      "learning_rate": 7.147215164456986e-05,
      "loss": 0.689453125,
      "step": 172
    },
    {
      "epoch": 2.0843373493975905,
      "grad_norm": 0.050876740366220474,
      "learning_rate": 7.11377231347075e-05,
      "loss": 0.74951171875,
      "step": 173
    },
    {
      "epoch": 2.0963855421686746,
      "grad_norm": 0.048498522490262985,
      "learning_rate": 7.080226862508077e-05,
      "loss": 0.720703125,
      "step": 174
    },
    {
      "epoch": 2.108433734939759,
      "grad_norm": 0.04706128314137459,
      "learning_rate": 7.046580944312509e-05,
      "loss": 0.7353515625,
      "step": 175
    },
    {
      "epoch": 2.1204819277108435,
      "grad_norm": 0.04583480581641197,
      "learning_rate": 7.012836698015066e-05,
      "loss": 0.7529296875,
      "step": 176
    },
    {
      "epoch": 2.1325301204819276,
      "grad_norm": 0.05084407702088356,
      "learning_rate": 6.978996268998244e-05,
      "loss": 0.68212890625,
      "step": 177
    },
    {
      "epoch": 2.144578313253012,
      "grad_norm": 0.045442067086696625,
      "learning_rate": 6.945061808759618e-05,
      "loss": 0.7177734375,
      "step": 178
    },
    {
      "epoch": 2.1566265060240966,
      "grad_norm": 0.04830284044146538,
      "learning_rate": 6.911035474775058e-05,
      "loss": 0.77294921875,
      "step": 179
    },
    {
      "epoch": 2.1686746987951806,
      "grad_norm": 0.050688792020082474,
      "learning_rate": 6.876919430361548e-05,
      "loss": 0.703125,
      "step": 180
    },
    {
      "epoch": 2.180722891566265,
      "grad_norm": 0.045072443783283234,
      "learning_rate": 6.842715844539671e-05,
      "loss": 0.7587890625,
      "step": 181
    },
    {
      "epoch": 2.1927710843373496,
      "grad_norm": 0.049108296632766724,
      "learning_rate": 6.808426891895683e-05,
      "loss": 0.7109375,
      "step": 182
    },
    {
      "epoch": 2.2048192771084336,
      "grad_norm": 0.047122225165367126,
      "learning_rate": 6.774054752443275e-05,
      "loss": 0.70654296875,
      "step": 183
    },
    {
      "epoch": 2.216867469879518,
      "grad_norm": 0.04391293227672577,
      "learning_rate": 6.73960161148496e-05,
      "loss": 0.8125,
      "step": 184
    },
    {
      "epoch": 2.2289156626506026,
      "grad_norm": 0.048237577080726624,
      "learning_rate": 6.705069659473149e-05,
      "loss": 0.7734375,
      "step": 185
    },
    {
      "epoch": 2.2409638554216866,
      "grad_norm": 0.050621919333934784,
      "learning_rate": 6.670461091870876e-05,
      "loss": 0.7109375,
      "step": 186
    },
    {
      "epoch": 2.253012048192771,
      "grad_norm": 0.044606756418943405,
      "learning_rate": 6.635778109012226e-05,
      "loss": 0.8154296875,
      "step": 187
    },
    {
      "epoch": 2.2650602409638556,
      "grad_norm": 0.03811057657003403,
      "learning_rate": 6.60102291596243e-05,
      "loss": 0.62548828125,
      "step": 188
    },
    {
      "epoch": 2.2771084337349397,
      "grad_norm": 0.045942217111587524,
      "learning_rate": 6.566197722377681e-05,
      "loss": 0.74365234375,
      "step": 189
    },
    {
      "epoch": 2.289156626506024,
      "grad_norm": 0.045799266546964645,
      "learning_rate": 6.53130474236465e-05,
      "loss": 0.685546875,
      "step": 190
    },
    {
      "epoch": 2.3012048192771086,
      "grad_norm": 0.038542598485946655,
      "learning_rate": 6.496346194339712e-05,
      "loss": 0.66650390625,
      "step": 191
    },
    {
      "epoch": 2.3132530120481927,
      "grad_norm": 0.046753592789173126,
      "learning_rate": 6.461324300887906e-05,
      "loss": 0.697265625,
      "step": 192
    },
    {
      "epoch": 2.325301204819277,
      "grad_norm": 0.043985772877931595,
      "learning_rate": 6.426241288621635e-05,
      "loss": 0.8017578125,
      "step": 193
    },
    {
      "epoch": 2.337349397590361,
      "grad_norm": 0.04346597567200661,
      "learning_rate": 6.391099388039094e-05,
      "loss": 0.69921875,
      "step": 194
    },
    {
      "epoch": 2.3493975903614457,
      "grad_norm": 0.041919730603694916,
      "learning_rate": 6.355900833382458e-05,
      "loss": 0.81884765625,
      "step": 195
    },
    {
      "epoch": 2.36144578313253,
      "grad_norm": 0.04315177723765373,
      "learning_rate": 6.320647862495848e-05,
      "loss": 0.8017578125,
      "step": 196
    },
    {
      "epoch": 2.3734939759036147,
      "grad_norm": 0.04781806841492653,
      "learning_rate": 6.285342716683042e-05,
      "loss": 0.67919921875,
      "step": 197
    },
    {
      "epoch": 2.3855421686746987,
      "grad_norm": 0.044159699231386185,
      "learning_rate": 6.249987640564978e-05,
      "loss": 0.7685546875,
      "step": 198
    },
    {
      "epoch": 2.397590361445783,
      "grad_norm": 0.04710109159350395,
      "learning_rate": 6.214584881937054e-05,
      "loss": 0.78369140625,
      "step": 199
    },
    {
      "epoch": 2.4096385542168672,
      "grad_norm": 0.04072146490216255,
      "learning_rate": 6.179136691626212e-05,
      "loss": 0.81005859375,
      "step": 200
    },
    {
      "epoch": 2.4216867469879517,
      "grad_norm": 0.05193505808711052,
      "learning_rate": 6.143645323347836e-05,
      "loss": 0.70556640625,
      "step": 201
    },
    {
      "epoch": 2.433734939759036,
      "grad_norm": 0.044399626553058624,
      "learning_rate": 6.108113033562468e-05,
      "loss": 0.71826171875,
      "step": 202
    },
    {
      "epoch": 2.4457831325301207,
      "grad_norm": 0.052263930439949036,
      "learning_rate": 6.072542081332348e-05,
      "loss": 0.629150390625,
      "step": 203
    },
    {
      "epoch": 2.4578313253012047,
      "grad_norm": 0.05204152315855026,
      "learning_rate": 6.0369347281777836e-05,
      "loss": 0.7587890625,
      "step": 204
    },
    {
      "epoch": 2.4698795180722892,
      "grad_norm": 0.04927836358547211,
      "learning_rate": 6.001293237933371e-05,
      "loss": 0.76025390625,
      "step": 205
    },
    {
      "epoch": 2.4819277108433733,
      "grad_norm": 0.04696602746844292,
      "learning_rate": 5.965619876604065e-05,
      "loss": 0.77490234375,
      "step": 206
    },
    {
      "epoch": 2.4939759036144578,
      "grad_norm": 0.049448322504758835,
      "learning_rate": 5.929916912221114e-05,
      "loss": 0.6650390625,
      "step": 207
    },
    {
      "epoch": 2.5060240963855422,
      "grad_norm": 0.047334328293800354,
      "learning_rate": 5.8941866146978586e-05,
      "loss": 0.73095703125,
      "step": 208
    },
    {
      "epoch": 2.5180722891566267,
      "grad_norm": 0.04117713123559952,
      "learning_rate": 5.8584312556854205e-05,
      "loss": 0.599090576171875,
      "step": 209
    },
    {
      "epoch": 2.5301204819277108,
      "grad_norm": 0.050235822796821594,
      "learning_rate": 5.82265310842827e-05,
      "loss": 0.7001953125,
      "step": 210
    },
    {
      "epoch": 2.5421686746987953,
      "grad_norm": 0.048958081752061844,
      "learning_rate": 5.786854447619706e-05,
      "loss": 0.7041015625,
      "step": 211
    },
    {
      "epoch": 2.5542168674698793,
      "grad_norm": 0.04437977448105812,
      "learning_rate": 5.751037549257232e-05,
      "loss": 0.79833984375,
      "step": 212
    },
    {
      "epoch": 2.566265060240964,
      "grad_norm": 0.045179564505815506,
      "learning_rate": 5.715204690497853e-05,
      "loss": 0.83447265625,
      "step": 213
    },
    {
      "epoch": 2.5783132530120483,
      "grad_norm": 0.047433678060770035,
      "learning_rate": 5.679358149513297e-05,
      "loss": 0.759765625,
      "step": 214
    },
    {
      "epoch": 2.5903614457831328,
      "grad_norm": 0.04841213300824165,
      "learning_rate": 5.6435002053451834e-05,
      "loss": 0.6708984375,
      "step": 215
    },
    {
      "epoch": 2.602409638554217,
      "grad_norm": 0.05304801091551781,
      "learning_rate": 5.6076331377601144e-05,
      "loss": 0.689453125,
      "step": 216
    },
    {
      "epoch": 2.6144578313253013,
      "grad_norm": 0.050705425441265106,
      "learning_rate": 5.5717592271047394e-05,
      "loss": 0.6962890625,
      "step": 217
    },
    {
      "epoch": 2.6265060240963853,
      "grad_norm": 0.0477730929851532,
      "learning_rate": 5.535880754160776e-05,
      "loss": 0.7177734375,
      "step": 218
    },
    {
      "epoch": 2.63855421686747,
      "grad_norm": 0.05534712225198746,
      "learning_rate": 5.500000000000001e-05,
      "loss": 0.6806640625,
      "step": 219
    },
    {
      "epoch": 2.6506024096385543,
      "grad_norm": 0.0478631854057312,
      "learning_rate": 5.464119245839227e-05,
      "loss": 0.7685546875,
      "step": 220
    },
    {
      "epoch": 2.662650602409639,
      "grad_norm": 0.051722150295972824,
      "learning_rate": 5.4282407728952624e-05,
      "loss": 0.6533203125,
      "step": 221
    },
    {
      "epoch": 2.674698795180723,
      "grad_norm": 0.04888847842812538,
      "learning_rate": 5.392366862239888e-05,
      "loss": 0.7080078125,
      "step": 222
    },
    {
      "epoch": 2.6867469879518073,
      "grad_norm": 0.04678454250097275,
      "learning_rate": 5.356499794654817e-05,
      "loss": 0.7587890625,
      "step": 223
    },
    {
      "epoch": 2.6987951807228914,
      "grad_norm": 0.046244245022535324,
      "learning_rate": 5.320641850486704e-05,
      "loss": 0.7587890625,
      "step": 224
    },
    {
      "epoch": 2.710843373493976,
      "grad_norm": 0.04210350662469864,
      "learning_rate": 5.284795309502149e-05,
      "loss": 0.80712890625,
      "step": 225
    },
    {
      "epoch": 2.7228915662650603,
      "grad_norm": 0.05145766958594322,
      "learning_rate": 5.248962450742769e-05,
      "loss": 0.68359375,
      "step": 226
    },
    {
      "epoch": 2.734939759036145,
      "grad_norm": 0.042611025273799896,
      "learning_rate": 5.213145552380296e-05,
      "loss": 0.759765625,
      "step": 227
    },
    {
      "epoch": 2.746987951807229,
      "grad_norm": 0.05434741452336311,
      "learning_rate": 5.177346891571732e-05,
      "loss": 0.697265625,
      "step": 228
    },
    {
      "epoch": 2.7590361445783134,
      "grad_norm": 0.04228973016142845,
      "learning_rate": 5.141568744314582e-05,
      "loss": 0.59759521484375,
      "step": 229
    },
    {
      "epoch": 2.7710843373493974,
      "grad_norm": 0.05489232763648033,
      "learning_rate": 5.105813385302142e-05,
      "loss": 0.67333984375,
      "step": 230
    },
    {
      "epoch": 2.783132530120482,
      "grad_norm": 0.044379498809576035,
      "learning_rate": 5.070083087778887e-05,
      "loss": 0.81787109375,
      "step": 231
    },
    {
      "epoch": 2.7951807228915664,
      "grad_norm": 0.054705750197172165,
      "learning_rate": 5.034380123395936e-05,
      "loss": 0.64501953125,
      "step": 232
    },
    {
      "epoch": 2.807228915662651,
      "grad_norm": 0.04854787513613701,
      "learning_rate": 4.998706762066632e-05,
      "loss": 0.74072265625,
      "step": 233
    },
    {
      "epoch": 2.819277108433735,
      "grad_norm": 0.04435088858008385,
      "learning_rate": 4.963065271822218e-05,
      "loss": 0.52410888671875,
      "step": 234
    },
    {
      "epoch": 2.8313253012048194,
      "grad_norm": 0.05073436722159386,
      "learning_rate": 4.927457918667653e-05,
      "loss": 0.681640625,
      "step": 235
    },
    {
      "epoch": 2.8433734939759034,
      "grad_norm": 0.053770314902067184,
      "learning_rate": 4.891886966437532e-05,
      "loss": 0.6513671875,
      "step": 236
    },
    {
      "epoch": 2.855421686746988,
      "grad_norm": 0.048878807574510574,
      "learning_rate": 4.856354676652165e-05,
      "loss": 0.70751953125,
      "step": 237
    },
    {
      "epoch": 2.8674698795180724,
      "grad_norm": 0.05484607815742493,
      "learning_rate": 4.820863308373789e-05,
      "loss": 0.6904296875,
      "step": 238
    },
    {
      "epoch": 2.8795180722891565,
      "grad_norm": 0.052399493753910065,
      "learning_rate": 4.785415118062947e-05,
      "loss": 0.66748046875,
      "step": 239
    },
    {
      "epoch": 2.891566265060241,
      "grad_norm": 0.05076400935649872,
      "learning_rate": 4.750012359435023e-05,
      "loss": 0.7109375,
      "step": 240
    },
    {
      "epoch": 2.9036144578313254,
      "grad_norm": 0.05134972184896469,
      "learning_rate": 4.714657283316961e-05,
      "loss": 0.729736328125,
      "step": 241
    },
    {
      "epoch": 2.9156626506024095,
      "grad_norm": 0.04737435653805733,
      "learning_rate": 4.6793521375041526e-05,
      "loss": 0.69482421875,
      "step": 242
    },
    {
      "epoch": 2.927710843373494,
      "grad_norm": 0.044630248099565506,
      "learning_rate": 4.6440991666175435e-05,
      "loss": 0.798828125,
      "step": 243
    },
    {
      "epoch": 2.9397590361445785,
      "grad_norm": 0.07215094566345215,
      "learning_rate": 4.608900611960908e-05,
      "loss": 0.75537109375,
      "step": 244
    },
    {
      "epoch": 2.9518072289156625,
      "grad_norm": 0.04662378504872322,
      "learning_rate": 4.5737587113783655e-05,
      "loss": 0.7451171875,
      "step": 245
    },
    {
      "epoch": 2.963855421686747,
      "grad_norm": 0.04592070356011391,
      "learning_rate": 4.5386756991120926e-05,
      "loss": 0.794921875,
      "step": 246
    },
    {
      "epoch": 2.9759036144578315,
      "grad_norm": 0.05287398770451546,
      "learning_rate": 4.5036538056602894e-05,
      "loss": 0.712890625,
      "step": 247
    },
    {
      "epoch": 2.9879518072289155,
      "grad_norm": 0.04912392050027847,
      "learning_rate": 4.46869525763535e-05,
      "loss": 0.75390625,
      "step": 248
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.05474797263741493,
      "learning_rate": 4.43380227762232e-05,
      "loss": 0.645751953125,
      "step": 249
    },
    {
      "epoch": 3.0,
      "eval_loss": 0.73974609375,
      "eval_runtime": 5.5892,
      "eval_samples_per_second": 1.073,
      "eval_steps_per_second": 0.179,
      "step": 249
    },
    {
      "epoch": 3.0120481927710845,
      "grad_norm": 0.04911520332098007,
      "learning_rate": 4.3989770840375724e-05,
      "loss": 0.703857421875,
      "step": 250
    },
    {
      "epoch": 3.0240963855421685,
      "grad_norm": 0.05203466862440109,
      "learning_rate": 4.364221890987774e-05,
      "loss": 0.5999755859375,
      "step": 251
    },
    {
      "epoch": 3.036144578313253,
      "grad_norm": 0.048401765525341034,
      "learning_rate": 4.329538908129124e-05,
      "loss": 0.779296875,
      "step": 252
    },
    {
      "epoch": 3.0481927710843375,
      "grad_norm": 0.04791688919067383,
      "learning_rate": 4.294930340526852e-05,
      "loss": 0.77294921875,
      "step": 253
    },
    {
      "epoch": 3.0602409638554215,
      "grad_norm": 0.06055872514843941,
      "learning_rate": 4.260398388515041e-05,
      "loss": 0.64404296875,
      "step": 254
    },
    {
      "epoch": 3.072289156626506,
      "grad_norm": 0.05343617498874664,
      "learning_rate": 4.2259452475567264e-05,
      "loss": 0.6484375,
      "step": 255
    },
    {
      "epoch": 3.0843373493975905,
      "grad_norm": 0.0541570745408535,
      "learning_rate": 4.1915731081043185e-05,
      "loss": 0.70556640625,
      "step": 256
    },
    {
      "epoch": 3.0963855421686746,
      "grad_norm": 0.06034017354249954,
      "learning_rate": 4.1572841554603315e-05,
      "loss": 0.67578125,
      "step": 257
    },
    {
      "epoch": 3.108433734939759,
      "grad_norm": 0.04963250085711479,
      "learning_rate": 4.1230805696384525e-05,
      "loss": 0.69580078125,
      "step": 258
    },
    {
      "epoch": 3.1204819277108435,
      "grad_norm": 0.06139222905039787,
      "learning_rate": 4.088964525224945e-05,
      "loss": 0.71337890625,
      "step": 259
    },
    {
      "epoch": 3.1325301204819276,
      "grad_norm": 0.0590483583509922,
      "learning_rate": 4.054938191240383e-05,
      "loss": 0.63916015625,
      "step": 260
    },
    {
      "epoch": 3.144578313253012,
      "grad_norm": 0.05276598408818245,
      "learning_rate": 4.021003731001758e-05,
      "loss": 0.6748046875,
      "step": 261
    },
    {
      "epoch": 3.1566265060240966,
      "grad_norm": 0.05339794605970383,
      "learning_rate": 3.987163301984935e-05,
      "loss": 0.73291015625,
      "step": 262
    },
    {
      "epoch": 3.1686746987951806,
      "grad_norm": 0.055302802473306656,
      "learning_rate": 3.953419055687492e-05,
      "loss": 0.6611328125,
      "step": 263
    },
    {
      "epoch": 3.180722891566265,
      "grad_norm": 0.055431053042411804,
      "learning_rate": 3.919773137491924e-05,
      "loss": 0.71923828125,
      "step": 264
    },
    {
      "epoch": 3.1927710843373496,
      "grad_norm": 0.058169297873973846,
      "learning_rate": 3.8862276865292514e-05,
      "loss": 0.6708984375,
      "step": 265
    },
    {
      "epoch": 3.2048192771084336,
      "grad_norm": 0.05146446079015732,
      "learning_rate": 3.852784835543014e-05,
      "loss": 0.666748046875,
      "step": 266
    },
    {
      "epoch": 3.216867469879518,
      "grad_norm": 0.050599105656147,
      "learning_rate": 3.8194467107536814e-05,
      "loss": 0.77490234375,
      "step": 267
    },
    {
      "epoch": 3.2289156626506026,
      "grad_norm": 0.05236315354704857,
      "learning_rate": 3.7862154317234654e-05,
      "loss": 0.73193359375,
      "step": 268
    },
    {
      "epoch": 3.2409638554216866,
      "grad_norm": 0.0552217997610569,
      "learning_rate": 3.753093111221572e-05,
      "loss": 0.67041015625,
      "step": 269
    },
    {
      "epoch": 3.253012048192771,
      "grad_norm": 0.04807985574007034,
      "learning_rate": 3.720081855089869e-05,
      "loss": 0.77685546875,
      "step": 270
    },
    {
      "epoch": 3.2650602409638556,
      "grad_norm": 0.04145972803235054,
      "learning_rate": 3.687183762109001e-05,
      "loss": 0.59503173828125,
      "step": 271
    },
    {
      "epoch": 3.2771084337349397,
      "grad_norm": 0.05418400093913078,
      "learning_rate": 3.6544009238649614e-05,
      "loss": 0.706298828125,
      "step": 272
    },
    {
      "epoch": 3.289156626506024,
      "grad_norm": 0.14572136104106903,
      "learning_rate": 3.621735424616108e-05,
      "loss": 0.646484375,
      "step": 273
    },
    {
      "epoch": 3.3012048192771086,
      "grad_norm": 0.04199405014514923,
      "learning_rate": 3.589189341160653e-05,
      "loss": 0.6346435546875,
      "step": 274
    },
    {
      "epoch": 3.3132530120481927,
      "grad_norm": 0.05933097377419472,
      "learning_rate": 3.556764742704619e-05,
      "loss": 0.6591796875,
      "step": 275
    },
    {
      "epoch": 3.325301204819277,
      "grad_norm": 0.04929123818874359,
      "learning_rate": 3.524463690730294e-05,
      "loss": 0.763671875,
      "step": 276
    },
    {
      "epoch": 3.337349397590361,
      "grad_norm": 0.05239742994308472,
      "learning_rate": 3.492288238865161e-05,
      "loss": 0.662109375,
      "step": 277
    },
    {
      "epoch": 3.3493975903614457,
      "grad_norm": 0.04863642156124115,
      "learning_rate": 3.460240432751328e-05,
      "loss": 0.78271484375,
      "step": 278
    },
    {
      "epoch": 3.36144578313253,
      "grad_norm": 0.04782761633396149,
      "learning_rate": 3.4283223099154815e-05,
      "loss": 0.76513671875,
      "step": 279
    },
    {
      "epoch": 3.3734939759036147,
      "grad_norm": 0.05435561388731003,
      "learning_rate": 3.3965358996393394e-05,
      "loss": 0.642578125,
      "step": 280
    },
    {
      "epoch": 3.3855421686746987,
      "grad_norm": 0.04814352095127106,
      "learning_rate": 3.3648832228306296e-05,
      "loss": 0.733154296875,
      "step": 281
    },
    {
      "epoch": 3.397590361445783,
      "grad_norm": 0.04872865229845047,
      "learning_rate": 3.333366291894619e-05,
      "loss": 0.751220703125,
      "step": 282
    },
    {
      "epoch": 3.4096385542168672,
      "grad_norm": 0.04587680473923683,
      "learning_rate": 3.301987110606151e-05,
      "loss": 0.776123046875,
      "step": 283
    },
    {
      "epoch": 3.4216867469879517,
      "grad_norm": 0.05650034174323082,
      "learning_rate": 3.2707476739822665e-05,
      "loss": 0.669921875,
      "step": 284
    },
    {
      "epoch": 3.433734939759036,
      "grad_norm": 0.046442095190286636,
      "learning_rate": 3.239649968155354e-05,
      "loss": 0.68603515625,
      "step": 285
    },
    {
      "epoch": 3.4457831325301207,
      "grad_norm": 0.05190088972449303,
      "learning_rate": 3.208695970246879e-05,
      "loss": 0.592529296875,
      "step": 286
    },
    {
      "epoch": 3.4578313253012047,
      "grad_norm": 0.049239631742239,
      "learning_rate": 3.177887648241691e-05,
      "loss": 0.7236328125,
      "step": 287
    },
    {
      "epoch": 3.4698795180722892,
      "grad_norm": 0.051178719848394394,
      "learning_rate": 3.147226960862887e-05,
      "loss": 0.725341796875,
      "step": 288
    },
    {
      "epoch": 3.4819277108433733,
      "grad_norm": 0.047779131680727005,
      "learning_rate": 3.1167158574473e-05,
      "loss": 0.73974609375,
      "step": 289
    },
    {
      "epoch": 3.4939759036144578,
      "grad_norm": 0.05449013411998749,
      "learning_rate": 3.086356277821545e-05,
      "loss": 0.62841796875,
      "step": 290
    },
    {
      "epoch": 3.5060240963855422,
      "grad_norm": 0.05054523050785065,
      "learning_rate": 3.056150152178707e-05,
      "loss": 0.696533203125,
      "step": 291
    },
    {
      "epoch": 3.5180722891566267,
      "grad_norm": 0.047927260398864746,
      "learning_rate": 3.0260994009556098e-05,
      "loss": 0.5734710693359375,
      "step": 292
    },
    {
      "epoch": 3.5301204819277108,
      "grad_norm": 0.053068626672029495,
      "learning_rate": 2.996205934710725e-05,
      "loss": 0.666015625,
      "step": 293
    },
    {
      "epoch": 3.5421686746987953,
      "grad_norm": 0.050974540412425995,
      "learning_rate": 2.9664716540027083e-05,
      "loss": 0.6708984375,
      "step": 294
    },
    {
      "epoch": 3.5542168674698793,
      "grad_norm": 0.0462355799973011,
      "learning_rate": 2.9368984492695524e-05,
      "loss": 0.76708984375,
      "step": 295
    },
    {
      "epoch": 3.566265060240964,
      "grad_norm": 0.04542889818549156,
      "learning_rate": 2.9074882007084135e-05,
      "loss": 0.80517578125,
      "step": 296
    },
    {
      "epoch": 3.5783132530120483,
      "grad_norm": 0.05186092108488083,
      "learning_rate": 2.8782427781560588e-05,
      "loss": 0.728271484375,
      "step": 297
    },
    {
      "epoch": 3.5903614457831328,
      "grad_norm": 0.05267176777124405,
      "learning_rate": 2.8491640409699977e-05,
      "loss": 0.63916015625,
      "step": 298
    },
    {
      "epoch": 3.602409638554217,
      "grad_norm": 0.051587752997875214,
      "learning_rate": 2.8202538379102562e-05,
      "loss": 0.65478515625,
      "step": 299
    },
    {
      "epoch": 3.6144578313253013,
      "grad_norm": 0.05620565637946129,
      "learning_rate": 2.7915140070218492e-05,
      "loss": 0.662353515625,
      "step": 300
    },
    {
      "epoch": 3.6265060240963853,
      "grad_norm": 0.05026882514357567,
      "learning_rate": 2.7629463755179152e-05,
      "loss": 0.684814453125,
      "step": 301
    },
    {
      "epoch": 3.63855421686747,
      "grad_norm": 0.05328973010182381,
      "learning_rate": 2.7345527596635424e-05,
      "loss": 0.6455078125,
      "step": 302
    },
    {
      "epoch": 3.6506024096385543,
      "grad_norm": 0.05063421651721001,
      "learning_rate": 2.7063349646603064e-05,
      "loss": 0.734375,
      "step": 303
    },
    {
      "epoch": 3.662650602409639,
      "grad_norm": 0.05420877784490585,
      "learning_rate": 2.6782947845314866e-05,
      "loss": 0.61962890625,
      "step": 304
    },
    {
      "epoch": 3.674698795180723,
      "grad_norm": 0.050907839089632034,
      "learning_rate": 2.6504340020080186e-05,
      "loss": 0.676513671875,
      "step": 305
    },
    {
      "epoch": 3.6867469879518073,
      "grad_norm": 0.048789966851472855,
      "learning_rate": 2.622754388415139e-05,
      "loss": 0.7265625,
      "step": 306
    },
    {
      "epoch": 3.6987951807228914,
      "grad_norm": 0.04995959252119064,
      "learning_rate": 2.5952577035597813e-05,
      "loss": 0.726806640625,
      "step": 307
    },
    {
      "epoch": 3.710843373493976,
      "grad_norm": 0.04701906815171242,
      "learning_rate": 2.567945695618682e-05,
      "loss": 0.77734375,
      "step": 308
    },
    {
      "epoch": 3.7228915662650603,
      "grad_norm": 0.05101192742586136,
      "learning_rate": 2.5408201010272393e-05,
      "loss": 0.65185546875,
      "step": 309
    },
    {
      "epoch": 3.734939759036145,
      "grad_norm": 0.04791393503546715,
      "learning_rate": 2.5138826443691167e-05,
      "loss": 0.73046875,
      "step": 310
    },
    {
      "epoch": 3.746987951807229,
      "grad_norm": 0.05365339666604996,
      "learning_rate": 2.487135038266592e-05,
      "loss": 0.66552734375,
      "step": 311
    },
    {
      "epoch": 3.7590361445783134,
      "grad_norm": 0.044052645564079285,
      "learning_rate": 2.4605789832716807e-05,
      "loss": 0.5758056640625,
      "step": 312
    },
    {
      "epoch": 3.7710843373493974,
      "grad_norm": 0.052461929619312286,
      "learning_rate": 2.434216167758008e-05,
      "loss": 0.64111328125,
      "step": 313
    },
    {
      "epoch": 3.783132530120482,
      "grad_norm": 0.047101572155952454,
      "learning_rate": 2.4080482678134796e-05,
      "loss": 0.788818359375,
      "step": 314
    },
    {
      "epoch": 3.7951807228915664,
      "grad_norm": 0.05791046842932701,
      "learning_rate": 2.3820769471337046e-05,
      "loss": 0.6142578125,
      "step": 315
    },
    {
      "epoch": 3.807228915662651,
      "grad_norm": 0.052243251353502274,
      "learning_rate": 2.356303856916235e-05,
      "loss": 0.708984375,
      "step": 316
    },
    {
      "epoch": 3.819277108433735,
      "grad_norm": 0.045897092670202255,
      "learning_rate": 2.3307306357555764e-05,
      "loss": 0.50128173828125,
      "step": 317
    },
    {
      "epoch": 3.8313253012048194,
      "grad_norm": 0.05350908637046814,
      "learning_rate": 2.3053589095390173e-05,
      "loss": 0.651611328125,
      "step": 318
    },
    {
      "epoch": 3.8433734939759034,
      "grad_norm": 0.05694294348359108,
      "learning_rate": 2.280190291343252e-05,
      "loss": 0.61962890625,
      "step": 319
    },
    {
      "epoch": 3.855421686746988,
      "grad_norm": 0.05226485803723335,
      "learning_rate": 2.2552263813318325e-05,
      "loss": 0.677734375,
      "step": 320
    },
    {
      "epoch": 3.8674698795180724,
      "grad_norm": 0.05336916819214821,
      "learning_rate": 2.2304687666534256e-05,
      "loss": 0.658935546875,
      "step": 321
    },
    {
      "epoch": 3.8795180722891565,
      "grad_norm": 0.057428572326898575,
      "learning_rate": 2.2059190213409138e-05,
      "loss": 0.63720703125,
      "step": 322
    },
    {
      "epoch": 3.891566265060241,
      "grad_norm": 0.050565581768751144,
      "learning_rate": 2.181578706211313e-05,
      "loss": 0.6826171875,
      "step": 323
    },
    {
      "epoch": 3.9036144578313254,
      "grad_norm": 0.051534149795770645,
      "learning_rate": 2.157449368766549e-05,
      "loss": 0.701416015625,
      "step": 324
    },
    {
      "epoch": 3.9156626506024095,
      "grad_norm": 0.05284786596894264,
      "learning_rate": 2.1335325430950605e-05,
      "loss": 0.66650390625,
      "step": 325
    },
    {
      "epoch": 3.927710843373494,
      "grad_norm": 0.04722355306148529,
      "learning_rate": 2.1098297497742753e-05,
      "loss": 0.77294921875,
      "step": 326
    },
    {
      "epoch": 3.9397590361445785,
      "grad_norm": 0.04605000093579292,
      "learning_rate": 2.0863424957739272e-05,
      "loss": 0.72802734375,
      "step": 327
    },
    {
      "epoch": 3.9518072289156625,
      "grad_norm": 0.04951898381114006,
      "learning_rate": 2.0630722743602536e-05,
      "loss": 0.71923828125,
      "step": 328
    },
    {
      "epoch": 3.963855421686747,
      "grad_norm": 0.04650385305285454,
      "learning_rate": 2.04002056500105e-05,
      "loss": 0.7685546875,
      "step": 329
    },
    {
      "epoch": 3.9759036144578315,
      "grad_norm": 0.052805136889219284,
      "learning_rate": 2.0171888332716117e-05,
      "loss": 0.685546875,
      "step": 330
    },
    {
      "epoch": 3.9879518072289155,
      "grad_norm": 0.04784088954329491,
      "learning_rate": 1.9945785307615597e-05,
      "loss": 0.727783203125,
      "step": 331
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.054690491408109665,
      "learning_rate": 1.972191094982544e-05,
      "loss": 0.6142578125,
      "step": 332
    },
    {
      "epoch": 4.0,
      "eval_loss": 0.729736328125,
      "eval_runtime": 5.5707,
      "eval_samples_per_second": 1.077,
      "eval_steps_per_second": 0.18,
      "step": 332
    },
    {
      "epoch": 4.0120481927710845,
      "grad_norm": 0.050628919154405594,
      "learning_rate": 1.9500279492768593e-05,
      "loss": 0.672607421875,
      "step": 333
    },
    {
      "epoch": 4.024096385542169,
      "grad_norm": 0.051577940583229065,
      "learning_rate": 1.9280905027269413e-05,
      "loss": 0.5792236328125,
      "step": 334
    },
    {
      "epoch": 4.036144578313253,
      "grad_norm": 0.05095500126481056,
      "learning_rate": 1.906380150065792e-05,
      "loss": 0.75341796875,
      "step": 335
    },
    {
      "epoch": 4.048192771084337,
      "grad_norm": 0.05450722202658653,
      "learning_rate": 1.8848982715882973e-05,
      "loss": 0.746826171875,
      "step": 336
    },
    {
      "epoch": 4.0602409638554215,
      "grad_norm": 0.053352318704128265,
      "learning_rate": 1.8636462330634713e-05,
      "loss": 0.61572265625,
      "step": 337
    },
    {
      "epoch": 4.072289156626506,
      "grad_norm": 0.05324487015604973,
      "learning_rate": 1.8426253856476316e-05,
      "loss": 0.6220703125,
      "step": 338
    },
    {
      "epoch": 4.0843373493975905,
      "grad_norm": 0.05741007998585701,
      "learning_rate": 1.8218370657984853e-05,
      "loss": 0.680908203125,
      "step": 339
    },
    {
      "epoch": 4.096385542168675,
      "grad_norm": 0.05961979180574417,
      "learning_rate": 1.8012825951901697e-05,
      "loss": 0.649658203125,
      "step": 340
    },
    {
      "epoch": 4.108433734939759,
      "grad_norm": 0.049268510192632675,
      "learning_rate": 1.7809632806292137e-05,
      "loss": 0.670654296875,
      "step": 341
    },
    {
      "epoch": 4.120481927710843,
      "grad_norm": 0.04999725520610809,
      "learning_rate": 1.7608804139714653e-05,
      "loss": 0.6865234375,
      "step": 342
    },
    {
      "epoch": 4.132530120481928,
      "grad_norm": 0.05462054908275604,
      "learning_rate": 1.7410352720399466e-05,
      "loss": 0.613037109375,
      "step": 343
    },
    {
      "epoch": 4.144578313253012,
      "grad_norm": 0.05475283041596413,
      "learning_rate": 1.7214291165436852e-05,
      "loss": 0.650146484375,
      "step": 344
    },
    {
      "epoch": 4.156626506024097,
      "grad_norm": 0.049491528421640396,
      "learning_rate": 1.7020631939974927e-05,
      "loss": 0.71044921875,
      "step": 345
    },
    {
      "epoch": 4.168674698795181,
      "grad_norm": 0.05138550326228142,
      "learning_rate": 1.682938735642715e-05,
      "loss": 0.634765625,
      "step": 346
    },
    {
      "epoch": 4.180722891566265,
      "grad_norm": 0.04960554838180542,
      "learning_rate": 1.6640569573689564e-05,
      "loss": 0.69384765625,
      "step": 347
    },
    {
      "epoch": 4.192771084337349,
      "grad_norm": 0.053525276482105255,
      "learning_rate": 1.645419059636769e-05,
      "loss": 0.64697265625,
      "step": 348
    },
    {
      "epoch": 4.204819277108434,
      "grad_norm": 0.05179784819483757,
      "learning_rate": 1.6270262274013388e-05,
      "loss": 0.642822265625,
      "step": 349
    },
    {
      "epoch": 4.216867469879518,
      "grad_norm": 0.048925552517175674,
      "learning_rate": 1.60887963003714e-05,
      "loss": 0.75048828125,
      "step": 350
    },
    {
      "epoch": 4.228915662650603,
      "grad_norm": 0.04987911134958267,
      "learning_rate": 1.590980421263598e-05,
      "loss": 0.709716796875,
      "step": 351
    },
    {
      "epoch": 4.240963855421687,
      "grad_norm": 0.05319344252347946,
      "learning_rate": 1.5733297390717287e-05,
      "loss": 0.6455078125,
      "step": 352
    },
    {
      "epoch": 4.253012048192771,
      "grad_norm": 0.04964512214064598,
      "learning_rate": 1.5559287056517955e-05,
      "loss": 0.754638671875,
      "step": 353
    },
    {
      "epoch": 4.265060240963855,
      "grad_norm": 0.0447506457567215,
      "learning_rate": 1.5387784273219606e-05,
      "loss": 0.57843017578125,
      "step": 354
    },
    {
      "epoch": 4.27710843373494,
      "grad_norm": 0.054329171776771545,
      "learning_rate": 1.5218799944579465e-05,
      "loss": 0.683837890625,
      "step": 355
    },
    {
      "epoch": 4.289156626506024,
      "grad_norm": 0.05268352851271629,
      "learning_rate": 1.5052344814237138e-05,
      "loss": 0.622802734375,
      "step": 356
    },
    {
      "epoch": 4.301204819277109,
      "grad_norm": 0.04670117422938347,
      "learning_rate": 1.4888429465031531e-05,
      "loss": 0.6168212890625,
      "step": 357
    },
    {
      "epoch": 4.313253012048193,
      "grad_norm": 0.05199913680553436,
      "learning_rate": 1.4727064318328077e-05,
      "loss": 0.6357421875,
      "step": 358
    },
    {
      "epoch": 4.325301204819277,
      "grad_norm": 0.05077880993485451,
      "learning_rate": 1.4568259633356091e-05,
      "loss": 0.7431640625,
      "step": 359
    },
    {
      "epoch": 4.337349397590361,
      "grad_norm": 0.0551830418407917,
      "learning_rate": 1.4412025506556558e-05,
      "loss": 0.640625,
      "step": 360
    },
    {
      "epoch": 4.349397590361446,
      "grad_norm": 0.04864334315061569,
      "learning_rate": 1.4258371870940245e-05,
      "loss": 0.7607421875,
      "step": 361
    },
    {
      "epoch": 4.36144578313253,
      "grad_norm": 0.0486895926296711,
      "learning_rate": 1.4107308495456101e-05,
      "loss": 0.74462890625,
      "step": 362
    },
    {
      "epoch": 4.373493975903615,
      "grad_norm": 0.05314972624182701,
      "learning_rate": 1.3958844984370281e-05,
      "loss": 0.62060546875,
      "step": 363
    },
    {
      "epoch": 4.385542168674699,
      "grad_norm": 0.04796022176742554,
      "learning_rate": 1.3812990776655413e-05,
      "loss": 0.713623046875,
      "step": 364
    },
    {
      "epoch": 4.397590361445783,
      "grad_norm": 0.05055151879787445,
      "learning_rate": 1.3669755145390601e-05,
      "loss": 0.72998046875,
      "step": 365
    },
    {
      "epoch": 4.409638554216867,
      "grad_norm": 0.049634531140327454,
      "learning_rate": 1.3529147197171766e-05,
      "loss": 0.756591796875,
      "step": 366
    },
    {
      "epoch": 4.421686746987952,
      "grad_norm": 0.06740517169237137,
      "learning_rate": 1.3391175871532716e-05,
      "loss": 0.648681640625,
      "step": 367
    },
    {
      "epoch": 4.433734939759036,
      "grad_norm": 0.04741332679986954,
      "learning_rate": 1.3255849940376827e-05,
      "loss": 0.66748046875,
      "step": 368
    },
    {
      "epoch": 4.445783132530121,
      "grad_norm": 0.05192714184522629,
      "learning_rate": 1.3123178007419256e-05,
      "loss": 0.57275390625,
      "step": 369
    },
    {
      "epoch": 4.457831325301205,
      "grad_norm": 0.05069812387228012,
      "learning_rate": 1.2993168507640022e-05,
      "loss": 0.70361328125,
      "step": 370
    },
    {
      "epoch": 4.469879518072289,
      "grad_norm": 0.050952933728694916,
      "learning_rate": 1.2865829706747661e-05,
      "loss": 0.704833984375,
      "step": 371
    },
    {
      "epoch": 4.481927710843373,
      "grad_norm": 0.05078313499689102,
      "learning_rate": 1.2741169700653776e-05,
      "loss": 0.72021484375,
      "step": 372
    },
    {
      "epoch": 4.493975903614458,
      "grad_norm": 0.053540535271167755,
      "learning_rate": 1.261919641495826e-05,
      "loss": 0.607666015625,
      "step": 373
    },
    {
      "epoch": 4.506024096385542,
      "grad_norm": 0.04874333739280701,
      "learning_rate": 1.2499917604445432e-05,
      "loss": 0.678955078125,
      "step": 374
    },
    {
      "epoch": 4.518072289156627,
      "grad_norm": 0.0443819984793663,
      "learning_rate": 1.2383340852591018e-05,
      "loss": 0.5587005615234375,
      "step": 375
    },
    {
      "epoch": 4.530120481927711,
      "grad_norm": 0.05096491798758507,
      "learning_rate": 1.2269473571079973e-05,
      "loss": 0.64697265625,
      "step": 376
    },
    {
      "epoch": 4.542168674698795,
      "grad_norm": 0.05113641917705536,
      "learning_rate": 1.2158322999335302e-05,
      "loss": 0.650390625,
      "step": 377
    },
    {
      "epoch": 4.554216867469879,
      "grad_norm": 0.04624852165579796,
      "learning_rate": 1.204989620405776e-05,
      "loss": 0.7490234375,
      "step": 378
    },
    {
      "epoch": 4.566265060240964,
      "grad_norm": 0.04788292571902275,
      "learning_rate": 1.1944200078776614e-05,
      "loss": 0.7890625,
      "step": 379
    },
    {
      "epoch": 4.578313253012048,
      "grad_norm": 0.051201291382312775,
      "learning_rate": 1.1841241343411292e-05,
      "loss": 0.709716796875,
      "step": 380
    },
    {
      "epoch": 4.590361445783133,
      "grad_norm": 0.05183528736233711,
      "learning_rate": 1.1741026543844239e-05,
      "loss": 0.621337890625,
      "step": 381
    },
    {
      "epoch": 4.602409638554217,
      "grad_norm": 0.052178479731082916,
      "learning_rate": 1.164356205150466e-05,
      "loss": 0.63427734375,
      "step": 382
    },
    {
      "epoch": 4.614457831325301,
      "grad_norm": 0.051191024482250214,
      "learning_rate": 1.1548854062963508e-05,
      "loss": 0.64306640625,
      "step": 383
    },
    {
      "epoch": 4.626506024096385,
      "grad_norm": 0.05055718496441841,
      "learning_rate": 1.1456908599539453e-05,
      "loss": 0.668701171875,
      "step": 384
    },
    {
      "epoch": 4.63855421686747,
      "grad_norm": 0.05491870269179344,
      "learning_rate": 1.1367731506916138e-05,
      "loss": 0.626953125,
      "step": 385
    },
    {
      "epoch": 4.650602409638554,
      "grad_norm": 0.05112716928124428,
      "learning_rate": 1.128132845477044e-05,
      "loss": 0.71630859375,
      "step": 386
    },
    {
      "epoch": 4.662650602409639,
      "grad_norm": 0.05581074208021164,
      "learning_rate": 1.1197704936412088e-05,
      "loss": 0.600341796875,
      "step": 387
    },
    {
      "epoch": 4.674698795180722,
      "grad_norm": 0.05165069177746773,
      "learning_rate": 1.1116866268434323e-05,
      "loss": 0.65966796875,
      "step": 388
    },
    {
      "epoch": 4.686746987951807,
      "grad_norm": 0.056944672018289566,
      "learning_rate": 1.1038817590375967e-05,
      "loss": 0.71142578125,
      "step": 389
    },
    {
      "epoch": 4.698795180722891,
      "grad_norm": 0.05819409340620041,
      "learning_rate": 1.0963563864394597e-05,
      "loss": 0.710205078125,
      "step": 390
    },
    {
      "epoch": 4.710843373493976,
      "grad_norm": 0.04591294750571251,
      "learning_rate": 1.089110987495111e-05,
      "loss": 0.762451171875,
      "step": 391
    },
    {
      "epoch": 4.72289156626506,
      "grad_norm": 0.052318621426820755,
      "learning_rate": 1.0821460228505495e-05,
      "loss": 0.63623046875,
      "step": 392
    },
    {
      "epoch": 4.734939759036145,
      "grad_norm": 0.04854559153318405,
      "learning_rate": 1.0754619353224007e-05,
      "loss": 0.71435546875,
      "step": 393
    },
    {
      "epoch": 4.746987951807229,
      "grad_norm": 0.050771284848451614,
      "learning_rate": 1.0690591498697591e-05,
      "loss": 0.649169921875,
      "step": 394
    },
    {
      "epoch": 4.759036144578313,
      "grad_norm": 0.042553313076496124,
      "learning_rate": 1.0629380735671746e-05,
      "loss": 0.56396484375,
      "step": 395
    },
    {
      "epoch": 4.771084337349397,
      "grad_norm": 0.05852508172392845,
      "learning_rate": 1.0570990955787675e-05,
      "loss": 0.62353515625,
      "step": 396
    },
    {
      "epoch": 4.783132530120482,
      "grad_norm": 0.04783790931105614,
      "learning_rate": 1.05154258713349e-05,
      "loss": 0.773193359375,
      "step": 397
    },
    {
      "epoch": 4.795180722891566,
      "grad_norm": 0.05407063663005829,
      "learning_rate": 1.0462689015015223e-05,
      "loss": 0.597900390625,
      "step": 398
    },
    {
      "epoch": 4.807228915662651,
      "grad_norm": 0.05101221054792404,
      "learning_rate": 1.0412783739718116e-05,
      "loss": 0.693359375,
      "step": 399
    },
    {
      "epoch": 4.8192771084337345,
      "grad_norm": 0.04693254828453064,
      "learning_rate": 1.0365713218307587e-05,
      "loss": 0.4884033203125,
      "step": 400
    },
    {
      "epoch": 4.831325301204819,
      "grad_norm": 0.05623774975538254,
      "learning_rate": 1.032148044342041e-05,
      "loss": 0.63720703125,
      "step": 401
    },
    {
      "epoch": 4.843373493975903,
      "grad_norm": 0.05694933980703354,
      "learning_rate": 1.0280088227275896e-05,
      "loss": 0.60400390625,
      "step": 402
    },
    {
      "epoch": 4.855421686746988,
      "grad_norm": 0.051737479865550995,
      "learning_rate": 1.0241539201497092e-05,
      "loss": 0.6630859375,
      "step": 403
    },
    {
      "epoch": 4.867469879518072,
      "grad_norm": 0.05680634081363678,
      "learning_rate": 1.020583581694345e-05,
      "loss": 0.64501953125,
      "step": 404
    },
    {
      "epoch": 4.879518072289157,
      "grad_norm": 0.0542740598320961,
      "learning_rate": 1.017298034355503e-05,
      "loss": 0.62109375,
      "step": 405
    },
    {
      "epoch": 4.891566265060241,
      "grad_norm": 0.052864715456962585,
      "learning_rate": 1.0142974870208151e-05,
      "loss": 0.6689453125,
      "step": 406
    },
    {
      "epoch": 4.903614457831325,
      "grad_norm": 0.05715819448232651,
      "learning_rate": 1.0115821304582643e-05,
      "loss": 0.686279296875,
      "step": 407
    },
    {
      "epoch": 4.9156626506024095,
      "grad_norm": 0.0525980219244957,
      "learning_rate": 1.0091521373040488e-05,
      "loss": 0.651123046875,
      "step": 408
    },
    {
      "epoch": 4.927710843373494,
      "grad_norm": 0.05436130613088608,
      "learning_rate": 1.0070076620516122e-05,
      "loss": 0.760986328125,
      "step": 409
    },
    {
      "epoch": 4.9397590361445785,
      "grad_norm": 0.04810437187552452,
      "learning_rate": 1.0051488410418169e-05,
      "loss": 0.715087890625,
      "step": 410
    },
    {
      "epoch": 4.951807228915663,
      "grad_norm": 0.05025848373770714,
      "learning_rate": 1.0035757924542791e-05,
      "loss": 0.705078125,
      "step": 411
    },
    {
      "epoch": 4.9638554216867465,
      "grad_norm": 0.04734313115477562,
      "learning_rate": 1.0022886162998531e-05,
      "loss": 0.754150390625,
      "step": 412
    },
    {
      "epoch": 4.975903614457831,
      "grad_norm": 0.054667290300130844,
      "learning_rate": 1.001287394414274e-05,
      "loss": 0.671875,
      "step": 413
    },
    {
      "epoch": 4.9879518072289155,
      "grad_norm": 0.04885823652148247,
      "learning_rate": 1.000572190452954e-05,
      "loss": 0.7158203125,
      "step": 414
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.058027319610118866,
      "learning_rate": 1.000143049886935e-05,
      "loss": 0.59765625,
      "step": 415
    },
    {
      "epoch": 5.0,
      "eval_loss": 0.727294921875,
      "eval_runtime": 5.5764,
      "eval_samples_per_second": 1.076,
      "eval_steps_per_second": 0.179,
      "step": 415
    }
  ],
  "logging_steps": 1.0,
  "max_steps": 415,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 0,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 3.6116289196455363e+19,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}