{
  "best_metric": 0.6429920196533203,
  "best_model_checkpoint": "miner_id_24/checkpoint-200",
  "epoch": 0.03536380514543365,
  "eval_steps": 50,
  "global_step": 200,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.00017681902572716824,
      "grad_norm": 31.45827293395996,
      "learning_rate": 1e-05,
      "loss": 5.7904,
      "step": 1
    },
    {
      "epoch": 0.00017681902572716824,
      "eval_loss": 1.5057493448257446,
      "eval_runtime": 697.5779,
      "eval_samples_per_second": 13.654,
      "eval_steps_per_second": 3.415,
      "step": 1
    },
    {
      "epoch": 0.0003536380514543365,
      "grad_norm": 34.1806640625,
      "learning_rate": 2e-05,
      "loss": 6.1229,
      "step": 2
    },
    {
      "epoch": 0.0005304570771815047,
      "grad_norm": 22.680261611938477,
      "learning_rate": 3e-05,
      "loss": 5.8247,
      "step": 3
    },
    {
      "epoch": 0.000707276102908673,
      "grad_norm": 17.230327606201172,
      "learning_rate": 4e-05,
      "loss": 4.859,
      "step": 4
    },
    {
      "epoch": 0.0008840951286358412,
      "grad_norm": 15.331626892089844,
      "learning_rate": 5e-05,
      "loss": 4.4965,
      "step": 5
    },
    {
      "epoch": 0.0010609141543630094,
      "grad_norm": 12.390981674194336,
      "learning_rate": 6e-05,
      "loss": 3.9683,
      "step": 6
    },
    {
      "epoch": 0.0012377331800901778,
      "grad_norm": 10.394994735717773,
      "learning_rate": 7e-05,
      "loss": 3.7708,
      "step": 7
    },
    {
      "epoch": 0.001414552205817346,
      "grad_norm": 10.412393569946289,
      "learning_rate": 8e-05,
      "loss": 3.6026,
      "step": 8
    },
    {
      "epoch": 0.001591371231544514,
      "grad_norm": 10.695121765136719,
      "learning_rate": 9e-05,
      "loss": 3.4169,
      "step": 9
    },
    {
      "epoch": 0.0017681902572716825,
      "grad_norm": 9.828315734863281,
      "learning_rate": 0.0001,
      "loss": 3.6774,
      "step": 10
    },
    {
      "epoch": 0.0019450092829988506,
      "grad_norm": 14.590110778808594,
      "learning_rate": 9.999316524962345e-05,
      "loss": 3.6333,
      "step": 11
    },
    {
      "epoch": 0.002121828308726019,
      "grad_norm": 8.314573287963867,
      "learning_rate": 9.997266286704631e-05,
      "loss": 3.5488,
      "step": 12
    },
    {
      "epoch": 0.002298647334453187,
      "grad_norm": 8.966548919677734,
      "learning_rate": 9.993849845741524e-05,
      "loss": 3.4814,
      "step": 13
    },
    {
      "epoch": 0.0024754663601803555,
      "grad_norm": 7.582108497619629,
      "learning_rate": 9.989068136093873e-05,
      "loss": 3.3718,
      "step": 14
    },
    {
      "epoch": 0.0026522853859075237,
      "grad_norm": 6.994225025177002,
      "learning_rate": 9.98292246503335e-05,
      "loss": 3.2995,
      "step": 15
    },
    {
      "epoch": 0.002829104411634692,
      "grad_norm": 7.011326313018799,
      "learning_rate": 9.975414512725057e-05,
      "loss": 3.2847,
      "step": 16
    },
    {
      "epoch": 0.00300592343736186,
      "grad_norm": 6.379772663116455,
      "learning_rate": 9.966546331768191e-05,
      "loss": 3.1525,
      "step": 17
    },
    {
      "epoch": 0.003182742463089028,
      "grad_norm": 7.65142297744751,
      "learning_rate": 9.956320346634876e-05,
      "loss": 3.2565,
      "step": 18
    },
    {
      "epoch": 0.0033595614888161968,
      "grad_norm": 6.9947309494018555,
      "learning_rate": 9.944739353007344e-05,
      "loss": 3.109,
      "step": 19
    },
    {
      "epoch": 0.003536380514543365,
      "grad_norm": 6.524970054626465,
      "learning_rate": 9.931806517013612e-05,
      "loss": 3.1238,
      "step": 20
    },
    {
      "epoch": 0.003713199540270533,
      "grad_norm": 6.738452911376953,
      "learning_rate": 9.917525374361912e-05,
      "loss": 3.095,
      "step": 21
    },
    {
      "epoch": 0.0038900185659977013,
      "grad_norm": 6.129000186920166,
      "learning_rate": 9.901899829374047e-05,
      "loss": 3.178,
      "step": 22
    },
    {
      "epoch": 0.00406683759172487,
      "grad_norm": 5.906164646148682,
      "learning_rate": 9.884934153917997e-05,
      "loss": 3.1067,
      "step": 23
    },
    {
      "epoch": 0.004243656617452038,
      "grad_norm": 5.738907337188721,
      "learning_rate": 9.86663298624003e-05,
      "loss": 3.1284,
      "step": 24
    },
    {
      "epoch": 0.004420475643179206,
      "grad_norm": 6.172359943389893,
      "learning_rate": 9.847001329696653e-05,
      "loss": 3.0873,
      "step": 25
    },
    {
      "epoch": 0.004597294668906374,
      "grad_norm": 6.070489883422852,
      "learning_rate": 9.826044551386744e-05,
      "loss": 3.024,
      "step": 26
    },
    {
      "epoch": 0.0047741136946335425,
      "grad_norm": 5.165764331817627,
      "learning_rate": 9.803768380684242e-05,
      "loss": 2.6945,
      "step": 27
    },
    {
      "epoch": 0.004950932720360711,
      "grad_norm": 5.422625541687012,
      "learning_rate": 9.780178907671789e-05,
      "loss": 3.1162,
      "step": 28
    },
    {
      "epoch": 0.005127751746087879,
      "grad_norm": 5.724668979644775,
      "learning_rate": 9.755282581475769e-05,
      "loss": 2.8799,
      "step": 29
    },
    {
      "epoch": 0.005304570771815047,
      "grad_norm": 5.608127593994141,
      "learning_rate": 9.729086208503174e-05,
      "loss": 3.1417,
      "step": 30
    },
    {
      "epoch": 0.005481389797542215,
      "grad_norm": 5.916441440582275,
      "learning_rate": 9.701596950580806e-05,
      "loss": 2.7373,
      "step": 31
    },
    {
      "epoch": 0.005658208823269384,
      "grad_norm": 6.396985054016113,
      "learning_rate": 9.672822322997305e-05,
      "loss": 2.9008,
      "step": 32
    },
    {
      "epoch": 0.005835027848996552,
      "grad_norm": 5.534632205963135,
      "learning_rate": 9.642770192448536e-05,
      "loss": 2.9336,
      "step": 33
    },
    {
      "epoch": 0.00601184687472372,
      "grad_norm": 5.079275131225586,
      "learning_rate": 9.611448774886924e-05,
      "loss": 2.8344,
      "step": 34
    },
    {
      "epoch": 0.006188665900450889,
      "grad_norm": 5.050607681274414,
      "learning_rate": 9.578866633275288e-05,
      "loss": 3.0011,
      "step": 35
    },
    {
      "epoch": 0.006365484926178056,
      "grad_norm": 4.718279838562012,
      "learning_rate": 9.545032675245813e-05,
      "loss": 2.8885,
      "step": 36
    },
    {
      "epoch": 0.006542303951905225,
      "grad_norm": 5.671701431274414,
      "learning_rate": 9.509956150664796e-05,
      "loss": 3.0087,
      "step": 37
    },
    {
      "epoch": 0.0067191229776323936,
      "grad_norm": 5.0960187911987305,
      "learning_rate": 9.473646649103818e-05,
      "loss": 2.8506,
      "step": 38
    },
    {
      "epoch": 0.006895942003359561,
      "grad_norm": 5.12495231628418,
      "learning_rate": 9.43611409721806e-05,
      "loss": 2.9755,
      "step": 39
    },
    {
      "epoch": 0.00707276102908673,
      "grad_norm": 5.502079010009766,
      "learning_rate": 9.397368756032445e-05,
      "loss": 2.7671,
      "step": 40
    },
    {
      "epoch": 0.007249580054813898,
      "grad_norm": 5.2155022621154785,
      "learning_rate": 9.357421218136386e-05,
      "loss": 2.9465,
      "step": 41
    },
    {
      "epoch": 0.007426399080541066,
      "grad_norm": 5.489111423492432,
      "learning_rate": 9.316282404787871e-05,
      "loss": 2.8452,
      "step": 42
    },
    {
      "epoch": 0.007603218106268235,
      "grad_norm": 4.807673454284668,
      "learning_rate": 9.273963562927695e-05,
      "loss": 2.6389,
      "step": 43
    },
    {
      "epoch": 0.0077800371319954025,
      "grad_norm": 4.675958633422852,
      "learning_rate": 9.230476262104677e-05,
      "loss": 2.7592,
      "step": 44
    },
    {
      "epoch": 0.007956856157722571,
      "grad_norm": 4.997447967529297,
      "learning_rate": 9.185832391312644e-05,
      "loss": 2.7375,
      "step": 45
    },
    {
      "epoch": 0.00813367518344974,
      "grad_norm": 5.734372615814209,
      "learning_rate": 9.140044155740101e-05,
      "loss": 2.6703,
      "step": 46
    },
    {
      "epoch": 0.008310494209176908,
      "grad_norm": 5.159792423248291,
      "learning_rate": 9.093124073433463e-05,
      "loss": 2.7582,
      "step": 47
    },
    {
      "epoch": 0.008487313234904075,
      "grad_norm": 5.002233028411865,
      "learning_rate": 9.045084971874738e-05,
      "loss": 2.6302,
      "step": 48
    },
    {
      "epoch": 0.008664132260631244,
      "grad_norm": 7.093165397644043,
      "learning_rate": 8.995939984474624e-05,
      "loss": 2.654,
      "step": 49
    },
    {
      "epoch": 0.008840951286358412,
      "grad_norm": 8.698509216308594,
      "learning_rate": 8.945702546981969e-05,
      "loss": 2.8836,
      "step": 50
    },
    {
      "epoch": 0.008840951286358412,
      "eval_loss": 0.796090304851532,
      "eval_runtime": 702.1573,
      "eval_samples_per_second": 13.565,
      "eval_steps_per_second": 3.392,
      "step": 50
    },
    {
      "epoch": 0.009017770312085581,
      "grad_norm": 10.096370697021484,
      "learning_rate": 8.894386393810563e-05,
      "loss": 3.866,
      "step": 51
    },
    {
      "epoch": 0.009194589337812748,
      "grad_norm": 5.29091215133667,
      "learning_rate": 8.842005554284296e-05,
      "loss": 3.2826,
      "step": 52
    },
    {
      "epoch": 0.009371408363539916,
      "grad_norm": 8.30187702178955,
      "learning_rate": 8.788574348801675e-05,
      "loss": 3.2665,
      "step": 53
    },
    {
      "epoch": 0.009548227389267085,
      "grad_norm": 8.213489532470703,
      "learning_rate": 8.73410738492077e-05,
      "loss": 3.3482,
      "step": 54
    },
    {
      "epoch": 0.009725046414994254,
      "grad_norm": 4.872696876525879,
      "learning_rate": 8.678619553365659e-05,
      "loss": 3.0297,
      "step": 55
    },
    {
      "epoch": 0.009901865440721422,
      "grad_norm": 4.427055835723877,
      "learning_rate": 8.622126023955446e-05,
      "loss": 3.1758,
      "step": 56
    },
    {
      "epoch": 0.010078684466448589,
      "grad_norm": 4.391081809997559,
      "learning_rate": 8.564642241456986e-05,
      "loss": 3.0824,
      "step": 57
    },
    {
      "epoch": 0.010255503492175758,
      "grad_norm": 4.616461753845215,
      "learning_rate": 8.506183921362443e-05,
      "loss": 2.9979,
      "step": 58
    },
    {
      "epoch": 0.010432322517902926,
      "grad_norm": 5.5998854637146,
      "learning_rate": 8.44676704559283e-05,
      "loss": 3.1535,
      "step": 59
    },
    {
      "epoch": 0.010609141543630095,
      "grad_norm": 4.823612213134766,
      "learning_rate": 8.386407858128706e-05,
      "loss": 2.9943,
      "step": 60
    },
    {
      "epoch": 0.010785960569357263,
      "grad_norm": 4.2810773849487305,
      "learning_rate": 8.32512286056924e-05,
      "loss": 2.8258,
      "step": 61
    },
    {
      "epoch": 0.01096277959508443,
      "grad_norm": 4.464574813842773,
      "learning_rate": 8.262928807620843e-05,
      "loss": 3.0109,
      "step": 62
    },
    {
      "epoch": 0.011139598620811599,
      "grad_norm": 4.964107036590576,
      "learning_rate": 8.199842702516583e-05,
      "loss": 3.0339,
      "step": 63
    },
    {
      "epoch": 0.011316417646538767,
      "grad_norm": 4.688778400421143,
      "learning_rate": 8.135881792367686e-05,
      "loss": 2.8198,
      "step": 64
    },
    {
      "epoch": 0.011493236672265936,
      "grad_norm": 4.536004066467285,
      "learning_rate": 8.07106356344834e-05,
      "loss": 3.1435,
      "step": 65
    },
    {
      "epoch": 0.011670055697993105,
      "grad_norm": 4.5505475997924805,
      "learning_rate": 8.005405736415126e-05,
      "loss": 2.8332,
      "step": 66
    },
    {
      "epoch": 0.011846874723720272,
      "grad_norm": 3.98578143119812,
      "learning_rate": 7.938926261462366e-05,
      "loss": 2.7494,
      "step": 67
    },
    {
      "epoch": 0.01202369374944744,
      "grad_norm": 4.547332763671875,
      "learning_rate": 7.871643313414718e-05,
      "loss": 2.9165,
      "step": 68
    },
    {
      "epoch": 0.012200512775174609,
      "grad_norm": 3.9778659343719482,
      "learning_rate": 7.803575286758364e-05,
      "loss": 2.7262,
      "step": 69
    },
    {
      "epoch": 0.012377331800901777,
      "grad_norm": 4.422539710998535,
      "learning_rate": 7.734740790612136e-05,
      "loss": 2.8827,
      "step": 70
    },
    {
      "epoch": 0.012554150826628946,
      "grad_norm": 4.440034866333008,
      "learning_rate": 7.66515864363997e-05,
      "loss": 2.9343,
      "step": 71
    },
    {
      "epoch": 0.012730969852356113,
      "grad_norm": 4.438572883605957,
      "learning_rate": 7.594847868906076e-05,
      "loss": 2.8539,
      "step": 72
    },
    {
      "epoch": 0.012907788878083281,
      "grad_norm": 4.555628776550293,
      "learning_rate": 7.52382768867422e-05,
      "loss": 3.0262,
      "step": 73
    },
    {
      "epoch": 0.01308460790381045,
      "grad_norm": 4.12335729598999,
      "learning_rate": 7.452117519152542e-05,
      "loss": 2.85,
      "step": 74
    },
    {
      "epoch": 0.013261426929537619,
      "grad_norm": 4.22313928604126,
      "learning_rate": 7.379736965185368e-05,
      "loss": 2.7441,
      "step": 75
    },
    {
      "epoch": 0.013438245955264787,
      "grad_norm": 4.474804401397705,
      "learning_rate": 7.30670581489344e-05,
      "loss": 2.832,
      "step": 76
    },
    {
      "epoch": 0.013615064980991954,
      "grad_norm": 4.030309200286865,
      "learning_rate": 7.233044034264034e-05,
      "loss": 2.7783,
      "step": 77
    },
    {
      "epoch": 0.013791884006719123,
      "grad_norm": 4.222381591796875,
      "learning_rate": 7.158771761692464e-05,
      "loss": 2.9147,
      "step": 78
    },
    {
      "epoch": 0.013968703032446291,
      "grad_norm": 4.298746109008789,
      "learning_rate": 7.083909302476453e-05,
      "loss": 2.8797,
      "step": 79
    },
    {
      "epoch": 0.01414552205817346,
      "grad_norm": 4.495664119720459,
      "learning_rate": 7.008477123264848e-05,
      "loss": 3.0076,
      "step": 80
    },
    {
      "epoch": 0.014322341083900628,
      "grad_norm": 4.467094421386719,
      "learning_rate": 6.932495846462261e-05,
      "loss": 2.7632,
      "step": 81
    },
    {
      "epoch": 0.014499160109627795,
      "grad_norm": 4.033594608306885,
      "learning_rate": 6.855986244591104e-05,
      "loss": 2.8745,
      "step": 82
    },
    {
      "epoch": 0.014675979135354964,
      "grad_norm": 3.7200775146484375,
      "learning_rate": 6.778969234612584e-05,
      "loss": 2.7365,
      "step": 83
    },
    {
      "epoch": 0.014852798161082132,
      "grad_norm": 4.186917781829834,
      "learning_rate": 6.701465872208216e-05,
      "loss": 2.7117,
      "step": 84
    },
    {
      "epoch": 0.015029617186809301,
      "grad_norm": 4.097092151641846,
      "learning_rate": 6.623497346023418e-05,
      "loss": 2.9168,
      "step": 85
    },
    {
      "epoch": 0.01520643621253647,
      "grad_norm": 4.045411586761475,
      "learning_rate": 6.545084971874738e-05,
      "loss": 2.652,
      "step": 86
    },
    {
      "epoch": 0.015383255238263636,
      "grad_norm": 3.9682466983795166,
      "learning_rate": 6.466250186922325e-05,
      "loss": 2.7225,
      "step": 87
    },
    {
      "epoch": 0.015560074263990805,
      "grad_norm": 4.498869895935059,
      "learning_rate": 6.387014543809223e-05,
      "loss": 2.6963,
      "step": 88
    },
    {
      "epoch": 0.015736893289717974,
      "grad_norm": 4.063172817230225,
      "learning_rate": 6.307399704769099e-05,
      "loss": 2.4276,
      "step": 89
    },
    {
      "epoch": 0.015913712315445142,
      "grad_norm": 4.287618160247803,
      "learning_rate": 6.227427435703997e-05,
      "loss": 2.532,
      "step": 90
    },
    {
      "epoch": 0.01609053134117231,
      "grad_norm": 4.349016189575195,
      "learning_rate": 6.147119600233758e-05,
      "loss": 2.5787,
      "step": 91
    },
    {
      "epoch": 0.01626735036689948,
      "grad_norm": 4.420528888702393,
      "learning_rate": 6.066498153718735e-05,
      "loss": 2.7135,
      "step": 92
    },
    {
      "epoch": 0.016444169392626648,
      "grad_norm": 3.8942527770996094,
      "learning_rate": 5.985585137257401e-05,
      "loss": 2.4143,
      "step": 93
    },
    {
      "epoch": 0.016620988418353817,
      "grad_norm": 4.0477399826049805,
      "learning_rate": 5.90440267166055e-05,
      "loss": 2.4438,
      "step": 94
    },
    {
      "epoch": 0.01679780744408098,
      "grad_norm": 4.152792453765869,
      "learning_rate": 5.8229729514036705e-05,
      "loss": 2.6107,
      "step": 95
    },
    {
      "epoch": 0.01697462646980815,
      "grad_norm": 4.101984977722168,
      "learning_rate": 5.74131823855921e-05,
      "loss": 2.5281,
      "step": 96
    },
    {
      "epoch": 0.01715144549553532,
      "grad_norm": 4.297181129455566,
      "learning_rate": 5.6594608567103456e-05,
      "loss": 2.5312,
      "step": 97
    },
    {
      "epoch": 0.017328264521262488,
      "grad_norm": 4.133391857147217,
      "learning_rate": 5.577423184847932e-05,
      "loss": 2.4468,
      "step": 98
    },
    {
      "epoch": 0.017505083546989656,
      "grad_norm": 4.21942663192749,
      "learning_rate": 5.495227651252315e-05,
      "loss": 2.4245,
      "step": 99
    },
    {
      "epoch": 0.017681902572716825,
      "grad_norm": 4.527026176452637,
      "learning_rate": 5.4128967273616625e-05,
      "loss": 2.4347,
      "step": 100
    },
    {
      "epoch": 0.017681902572716825,
      "eval_loss": 0.7386493682861328,
      "eval_runtime": 702.0834,
      "eval_samples_per_second": 13.567,
      "eval_steps_per_second": 3.393,
      "step": 100
    },
    {
      "epoch": 0.017858721598443993,
      "grad_norm": 9.57063102722168,
      "learning_rate": 5.330452921628497e-05,
      "loss": 3.4226,
      "step": 101
    },
    {
      "epoch": 0.018035540624171162,
      "grad_norm": 7.75221061706543,
      "learning_rate": 5.247918773366112e-05,
      "loss": 3.4674,
      "step": 102
    },
    {
      "epoch": 0.01821235964989833,
      "grad_norm": 4.360025405883789,
      "learning_rate": 5.165316846586541e-05,
      "loss": 2.9915,
      "step": 103
    },
    {
      "epoch": 0.018389178675625496,
      "grad_norm": 4.865253448486328,
      "learning_rate": 5.0826697238317935e-05,
      "loss": 3.1289,
      "step": 104
    },
    {
      "epoch": 0.018565997701352664,
      "grad_norm": 5.143971920013428,
      "learning_rate": 5e-05,
      "loss": 3.178,
      "step": 105
    },
    {
      "epoch": 0.018742816727079833,
      "grad_norm": 4.194464683532715,
      "learning_rate": 4.917330276168208e-05,
      "loss": 3.0124,
      "step": 106
    },
    {
      "epoch": 0.018919635752807,
      "grad_norm": 3.829989194869995,
      "learning_rate": 4.834683153413459e-05,
      "loss": 3.0288,
      "step": 107
    },
    {
      "epoch": 0.01909645477853417,
      "grad_norm": 3.890665292739868,
      "learning_rate": 4.7520812266338885e-05,
      "loss": 2.8104,
      "step": 108
    },
    {
      "epoch": 0.01927327380426134,
      "grad_norm": 3.4673619270324707,
      "learning_rate": 4.669547078371504e-05,
      "loss": 2.7374,
      "step": 109
    },
    {
      "epoch": 0.019450092829988507,
      "grad_norm": 3.518317699432373,
      "learning_rate": 4.5871032726383386e-05,
      "loss": 2.8593,
      "step": 110
    },
    {
      "epoch": 0.019626911855715676,
      "grad_norm": 3.6939802169799805,
      "learning_rate": 4.504772348747687e-05,
      "loss": 2.6806,
      "step": 111
    },
    {
      "epoch": 0.019803730881442844,
      "grad_norm": 4.089910984039307,
      "learning_rate": 4.4225768151520694e-05,
      "loss": 3.0053,
      "step": 112
    },
    {
      "epoch": 0.019980549907170013,
      "grad_norm": 3.797542095184326,
      "learning_rate": 4.3405391432896555e-05,
      "loss": 2.8644,
      "step": 113
    },
    {
      "epoch": 0.020157368932897178,
      "grad_norm": 3.655669689178467,
      "learning_rate": 4.2586817614407895e-05,
      "loss": 2.4503,
      "step": 114
    },
    {
      "epoch": 0.020334187958624347,
      "grad_norm": 3.9156174659729004,
      "learning_rate": 4.17702704859633e-05,
      "loss": 2.7663,
      "step": 115
    },
    {
      "epoch": 0.020511006984351515,
      "grad_norm": 3.895479440689087,
      "learning_rate": 4.095597328339452e-05,
      "loss": 2.8286,
      "step": 116
    },
    {
      "epoch": 0.020687826010078684,
      "grad_norm": 3.7556660175323486,
      "learning_rate": 4.0144148627425993e-05,
      "loss": 2.7171,
      "step": 117
    },
    {
      "epoch": 0.020864645035805852,
      "grad_norm": 3.933443546295166,
      "learning_rate": 3.933501846281267e-05,
      "loss": 2.8017,
      "step": 118
    },
    {
      "epoch": 0.02104146406153302,
      "grad_norm": 3.722991943359375,
      "learning_rate": 3.852880399766243e-05,
      "loss": 2.7568,
      "step": 119
    },
    {
      "epoch": 0.02121828308726019,
      "grad_norm": 4.077044486999512,
      "learning_rate": 3.772572564296005e-05,
      "loss": 2.7161,
      "step": 120
    },
    {
      "epoch": 0.021395102112987358,
      "grad_norm": 3.83428955078125,
      "learning_rate": 3.6926002952309016e-05,
      "loss": 2.5725,
      "step": 121
    },
    {
      "epoch": 0.021571921138714527,
      "grad_norm": 3.630547285079956,
      "learning_rate": 3.612985456190778e-05,
      "loss": 2.6474,
      "step": 122
    },
    {
      "epoch": 0.021748740164441695,
      "grad_norm": 3.830258369445801,
      "learning_rate": 3.533749813077677e-05,
      "loss": 2.6005,
      "step": 123
    },
    {
      "epoch": 0.02192555919016886,
      "grad_norm": 3.7952754497528076,
      "learning_rate": 3.4549150281252636e-05,
      "loss": 2.6564,
      "step": 124
    },
    {
      "epoch": 0.02210237821589603,
      "grad_norm": 4.006258964538574,
      "learning_rate": 3.3765026539765834e-05,
      "loss": 2.6367,
      "step": 125
    },
    {
      "epoch": 0.022279197241623198,
      "grad_norm": 3.754654884338379,
      "learning_rate": 3.298534127791785e-05,
      "loss": 2.5622,
      "step": 126
    },
    {
      "epoch": 0.022456016267350366,
      "grad_norm": 3.727962017059326,
      "learning_rate": 3.221030765387417e-05,
      "loss": 2.6284,
      "step": 127
    },
    {
      "epoch": 0.022632835293077535,
      "grad_norm": 3.7966456413269043,
      "learning_rate": 3.144013755408895e-05,
      "loss": 2.735,
      "step": 128
    },
    {
      "epoch": 0.022809654318804704,
      "grad_norm": 3.8324391841888428,
      "learning_rate": 3.0675041535377405e-05,
      "loss": 2.5817,
      "step": 129
    },
    {
      "epoch": 0.022986473344531872,
      "grad_norm": 3.9248697757720947,
      "learning_rate": 2.991522876735154e-05,
      "loss": 2.8796,
      "step": 130
    },
    {
      "epoch": 0.02316329237025904,
      "grad_norm": 3.8856313228607178,
      "learning_rate": 2.916090697523549e-05,
      "loss": 2.5024,
      "step": 131
    },
    {
      "epoch": 0.02334011139598621,
      "grad_norm": 3.7371416091918945,
      "learning_rate": 2.8412282383075363e-05,
      "loss": 2.597,
      "step": 132
    },
    {
      "epoch": 0.023516930421713378,
      "grad_norm": 3.8303768634796143,
      "learning_rate": 2.766955965735968e-05,
      "loss": 2.4895,
      "step": 133
    },
    {
      "epoch": 0.023693749447440543,
      "grad_norm": 3.8751957416534424,
      "learning_rate": 2.693294185106562e-05,
      "loss": 2.5636,
      "step": 134
    },
    {
      "epoch": 0.02387056847316771,
      "grad_norm": 3.867126226425171,
      "learning_rate": 2.6202630348146324e-05,
      "loss": 2.4996,
      "step": 135
    },
    {
      "epoch": 0.02404738749889488,
      "grad_norm": 4.0624003410339355,
      "learning_rate": 2.547882480847461e-05,
      "loss": 2.7292,
      "step": 136
    },
    {
      "epoch": 0.02422420652462205,
      "grad_norm": 3.5355868339538574,
      "learning_rate": 2.476172311325783e-05,
      "loss": 2.4412,
      "step": 137
    },
    {
      "epoch": 0.024401025550349217,
      "grad_norm": 3.7237305641174316,
      "learning_rate": 2.405152131093926e-05,
      "loss": 2.6159,
      "step": 138
    },
    {
      "epoch": 0.024577844576076386,
      "grad_norm": 3.518690347671509,
      "learning_rate": 2.3348413563600325e-05,
      "loss": 2.4404,
      "step": 139
    },
    {
      "epoch": 0.024754663601803555,
      "grad_norm": 3.7103893756866455,
      "learning_rate": 2.2652592093878666e-05,
      "loss": 2.2584,
      "step": 140
    },
    {
      "epoch": 0.024931482627530723,
      "grad_norm": 3.6089131832122803,
      "learning_rate": 2.196424713241637e-05,
      "loss": 2.4054,
      "step": 141
    },
    {
      "epoch": 0.025108301653257892,
      "grad_norm": 3.8912951946258545,
      "learning_rate": 2.128356686585282e-05,
      "loss": 2.5531,
      "step": 142
    },
    {
      "epoch": 0.02528512067898506,
      "grad_norm": 3.9472362995147705,
      "learning_rate": 2.061073738537635e-05,
      "loss": 2.3604,
      "step": 143
    },
    {
      "epoch": 0.025461939704712225,
      "grad_norm": 3.8991811275482178,
      "learning_rate": 1.9945942635848748e-05,
      "loss": 2.4417,
      "step": 144
    },
    {
      "epoch": 0.025638758730439394,
      "grad_norm": 3.9949779510498047,
      "learning_rate": 1.928936436551661e-05,
      "loss": 2.3783,
      "step": 145
    },
    {
      "epoch": 0.025815577756166563,
      "grad_norm": 4.199954032897949,
      "learning_rate": 1.8641182076323148e-05,
      "loss": 2.5726,
      "step": 146
    },
    {
      "epoch": 0.02599239678189373,
      "grad_norm": 4.285723686218262,
      "learning_rate": 1.800157297483417e-05,
      "loss": 2.5949,
      "step": 147
    },
    {
      "epoch": 0.0261692158076209,
      "grad_norm": 3.848357915878296,
      "learning_rate": 1.7370711923791567e-05,
      "loss": 2.2872,
      "step": 148
    },
    {
      "epoch": 0.02634603483334807,
      "grad_norm": 3.925421953201294,
      "learning_rate": 1.6748771394307585e-05,
      "loss": 2.2459,
      "step": 149
    },
    {
      "epoch": 0.026522853859075237,
      "grad_norm": 4.618225574493408,
      "learning_rate": 1.6135921418712956e-05,
      "loss": 2.4605,
      "step": 150
    },
    {
      "epoch": 0.026522853859075237,
      "eval_loss": 0.6821020245552063,
      "eval_runtime": 702.2143,
      "eval_samples_per_second": 13.564,
      "eval_steps_per_second": 3.392,
      "step": 150
    },
    {
      "epoch": 0.026699672884802406,
      "grad_norm": 6.810517311096191,
      "learning_rate": 1.553232954407171e-05,
      "loss": 3.3892,
      "step": 151
    },
    {
      "epoch": 0.026876491910529574,
      "grad_norm": 6.7886810302734375,
      "learning_rate": 1.4938160786375572e-05,
      "loss": 3.2336,
      "step": 152
    },
    {
      "epoch": 0.027053310936256743,
      "grad_norm": 6.150904655456543,
      "learning_rate": 1.435357758543015e-05,
      "loss": 3.1348,
      "step": 153
    },
    {
      "epoch": 0.027230129961983908,
      "grad_norm": 5.942625522613525,
      "learning_rate": 1.3778739760445552e-05,
      "loss": 2.9942,
      "step": 154
    },
    {
      "epoch": 0.027406948987711077,
      "grad_norm": 4.729331970214844,
      "learning_rate": 1.3213804466343421e-05,
      "loss": 2.8072,
      "step": 155
    },
    {
      "epoch": 0.027583768013438245,
      "grad_norm": 3.9404985904693604,
      "learning_rate": 1.2658926150792322e-05,
      "loss": 2.7145,
      "step": 156
    },
    {
      "epoch": 0.027760587039165414,
      "grad_norm": 3.8660361766815186,
      "learning_rate": 1.2114256511983274e-05,
      "loss": 2.843,
      "step": 157
    },
    {
      "epoch": 0.027937406064892582,
      "grad_norm": 3.684872627258301,
      "learning_rate": 1.157994445715706e-05,
      "loss": 2.6958,
      "step": 158
    },
    {
      "epoch": 0.02811422509061975,
      "grad_norm": 4.2424468994140625,
      "learning_rate": 1.1056136061894384e-05,
      "loss": 2.7719,
      "step": 159
    },
    {
      "epoch": 0.02829104411634692,
      "grad_norm": 4.217720031738281,
      "learning_rate": 1.0542974530180327e-05,
      "loss": 2.9478,
      "step": 160
    },
    {
      "epoch": 0.028467863142074088,
      "grad_norm": 4.370022773742676,
      "learning_rate": 1.0040600155253765e-05,
      "loss": 2.8683,
      "step": 161
    },
    {
      "epoch": 0.028644682167801257,
      "grad_norm": 4.087332725524902,
      "learning_rate": 9.549150281252633e-06,
      "loss": 2.741,
      "step": 162
    },
    {
      "epoch": 0.028821501193528425,
      "grad_norm": 4.101226806640625,
      "learning_rate": 9.068759265665384e-06,
      "loss": 2.9603,
      "step": 163
    },
    {
      "epoch": 0.02899832021925559,
      "grad_norm": 3.877307891845703,
      "learning_rate": 8.599558442598998e-06,
      "loss": 2.7864,
      "step": 164
    },
    {
      "epoch": 0.02917513924498276,
      "grad_norm": 3.6642425060272217,
      "learning_rate": 8.141676086873572e-06,
      "loss": 2.6385,
      "step": 165
    },
    {
      "epoch": 0.029351958270709928,
      "grad_norm": 3.7589988708496094,
      "learning_rate": 7.695237378953223e-06,
      "loss": 2.6889,
      "step": 166
    },
    {
      "epoch": 0.029528777296437096,
      "grad_norm": 3.773329734802246,
      "learning_rate": 7.260364370723044e-06,
      "loss": 2.7377,
      "step": 167
    },
    {
      "epoch": 0.029705596322164265,
      "grad_norm": 3.6296513080596924,
      "learning_rate": 6.837175952121306e-06,
      "loss": 2.595,
      "step": 168
    },
    {
      "epoch": 0.029882415347891433,
      "grad_norm": 3.9712111949920654,
      "learning_rate": 6.425787818636131e-06,
      "loss": 2.7767,
      "step": 169
    },
    {
      "epoch": 0.030059234373618602,
      "grad_norm": 3.7664237022399902,
      "learning_rate": 6.026312439675552e-06,
      "loss": 2.6668,
      "step": 170
    },
    {
      "epoch": 0.03023605339934577,
      "grad_norm": 3.7158427238464355,
      "learning_rate": 5.6388590278194096e-06,
      "loss": 2.6469,
      "step": 171
    },
    {
      "epoch": 0.03041287242507294,
      "grad_norm": 3.6463067531585693,
      "learning_rate": 5.263533508961827e-06,
      "loss": 2.6483,
      "step": 172
    },
    {
      "epoch": 0.030589691450800108,
      "grad_norm": 3.6208555698394775,
      "learning_rate": 4.900438493352055e-06,
      "loss": 2.5328,
      "step": 173
    },
    {
      "epoch": 0.030766510476527273,
      "grad_norm": 3.5317137241363525,
      "learning_rate": 4.549673247541875e-06,
      "loss": 2.494,
      "step": 174
    },
    {
      "epoch": 0.03094332950225444,
      "grad_norm": 3.8325750827789307,
      "learning_rate": 4.2113336672471245e-06,
      "loss": 2.6988,
      "step": 175
    },
    {
      "epoch": 0.03112014852798161,
      "grad_norm": 3.9132578372955322,
      "learning_rate": 3.885512251130763e-06,
      "loss": 2.6593,
      "step": 176
    },
    {
      "epoch": 0.03129696755370878,
      "grad_norm": 3.598088026046753,
      "learning_rate": 3.5722980755146517e-06,
      "loss": 2.4802,
      "step": 177
    },
    {
      "epoch": 0.03147378657943595,
      "grad_norm": 3.5188252925872803,
      "learning_rate": 3.271776770026963e-06,
      "loss": 2.5409,
      "step": 178
    },
    {
      "epoch": 0.03165060560516311,
      "grad_norm": 3.72456431388855,
      "learning_rate": 2.9840304941919415e-06,
      "loss": 2.5696,
      "step": 179
    },
    {
      "epoch": 0.031827424630890284,
      "grad_norm": 3.7853140830993652,
      "learning_rate": 2.7091379149682685e-06,
      "loss": 2.7003,
      "step": 180
    },
    {
      "epoch": 0.03200424365661745,
      "grad_norm": 3.860959768295288,
      "learning_rate": 2.4471741852423237e-06,
      "loss": 2.5493,
      "step": 181
    },
    {
      "epoch": 0.03218106268234462,
      "grad_norm": 3.5097670555114746,
      "learning_rate": 2.1982109232821178e-06,
      "loss": 2.5367,
      "step": 182
    },
    {
      "epoch": 0.03235788170807179,
      "grad_norm": 3.6318600177764893,
      "learning_rate": 1.962316193157593e-06,
      "loss": 2.5658,
      "step": 183
    },
    {
      "epoch": 0.03253470073379896,
      "grad_norm": 3.9260454177856445,
      "learning_rate": 1.7395544861325718e-06,
      "loss": 2.6398,
      "step": 184
    },
    {
      "epoch": 0.032711519759526124,
      "grad_norm": 3.5062942504882812,
      "learning_rate": 1.5299867030334814e-06,
      "loss": 2.3665,
      "step": 185
    },
    {
      "epoch": 0.032888338785253296,
      "grad_norm": 3.8497798442840576,
      "learning_rate": 1.333670137599713e-06,
      "loss": 2.5584,
      "step": 186
    },
    {
      "epoch": 0.03306515781098046,
      "grad_norm": 3.772922992706299,
      "learning_rate": 1.1506584608200367e-06,
      "loss": 2.6179,
      "step": 187
    },
    {
      "epoch": 0.03324197683670763,
      "grad_norm": 3.7414562702178955,
      "learning_rate": 9.810017062595322e-07,
      "loss": 2.3508,
      "step": 188
    },
    {
      "epoch": 0.0334187958624348,
      "grad_norm": 3.7079050540924072,
      "learning_rate": 8.247462563808817e-07,
      "loss": 2.5306,
      "step": 189
    },
    {
      "epoch": 0.03359561488816196,
      "grad_norm": 3.4822869300842285,
      "learning_rate": 6.819348298638839e-07,
      "loss": 2.3852,
      "step": 190
    },
    {
      "epoch": 0.033772433913889136,
      "grad_norm": 3.7052745819091797,
      "learning_rate": 5.526064699265753e-07,
      "loss": 2.5681,
      "step": 191
    },
    {
      "epoch": 0.0339492529396163,
      "grad_norm": 3.7392704486846924,
      "learning_rate": 4.367965336512403e-07,
      "loss": 2.3412,
      "step": 192
    },
    {
      "epoch": 0.03412607196534347,
      "grad_norm": 3.666461944580078,
      "learning_rate": 3.3453668231809286e-07,
      "loss": 2.3487,
      "step": 193
    },
    {
      "epoch": 0.03430289099107064,
      "grad_norm": 3.8285269737243652,
      "learning_rate": 2.458548727494292e-07,
      "loss": 2.5071,
      "step": 194
    },
    {
      "epoch": 0.03447971001679781,
      "grad_norm": 3.8054890632629395,
      "learning_rate": 1.7077534966650766e-07,
      "loss": 2.3838,
      "step": 195
    },
    {
      "epoch": 0.034656529042524975,
      "grad_norm": 3.792473554611206,
      "learning_rate": 1.0931863906127327e-07,
      "loss": 2.3911,
      "step": 196
    },
    {
      "epoch": 0.03483334806825215,
      "grad_norm": 4.204066753387451,
      "learning_rate": 6.150154258476315e-08,
      "loss": 2.5493,
      "step": 197
    },
    {
      "epoch": 0.03501016709397931,
      "grad_norm": 4.200047492980957,
      "learning_rate": 2.7337132953697554e-08,
      "loss": 2.3861,
      "step": 198
    },
    {
      "epoch": 0.03518698611970648,
      "grad_norm": 4.204197883605957,
      "learning_rate": 6.834750376549792e-09,
      "loss": 2.4921,
      "step": 199
    },
    {
      "epoch": 0.03536380514543365,
      "grad_norm": 5.526068687438965,
      "learning_rate": 0.0,
      "loss": 2.665,
      "step": 200
    },
    {
      "epoch": 0.03536380514543365,
      "eval_loss": 0.6429920196533203,
      "eval_runtime": 702.134,
      "eval_samples_per_second": 13.566,
      "eval_steps_per_second": 3.393,
      "step": 200
    }
  ],
  "logging_steps": 1,
  "max_steps": 200,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 50,
  "stateful_callbacks": {
    "EarlyStoppingCallback": {
      "args": {
        "early_stopping_patience": 5,
        "early_stopping_threshold": 0.0
      },
      "attributes": {
        "early_stopping_patience_counter": 0
      }
    },
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 2.861999239200768e+17,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}