{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 0,
  "global_step": 118,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.00847457627118644,
      "grad_norm": 35.74056625366211,
      "learning_rate": 1.008e-05,
      "loss": 2.2767,
      "step": 1
    },
    {
      "epoch": 0.01694915254237288,
      "grad_norm": 13.115900039672852,
      "learning_rate": 1.0078213880062201e-05,
      "loss": 1.6586,
      "step": 2
    },
    {
      "epoch": 0.025423728813559324,
      "grad_norm": 8.92249870300293,
      "learning_rate": 1.007285678621088e-05,
      "loss": 1.1126,
      "step": 3
    },
    {
      "epoch": 0.03389830508474576,
      "grad_norm": 12.179636001586914,
      "learning_rate": 1.0063932515434981e-05,
      "loss": 0.9935,
      "step": 4
    },
    {
      "epoch": 0.0423728813559322,
      "grad_norm": 4.910240650177002,
      "learning_rate": 1.005144739305909e-05,
      "loss": 0.8416,
      "step": 5
    },
    {
      "epoch": 0.05084745762711865,
      "grad_norm": 2.3368802070617676,
      "learning_rate": 1.0035410268260196e-05,
      "loss": 0.8057,
      "step": 6
    },
    {
      "epoch": 0.059322033898305086,
      "grad_norm": 3.5738086700439453,
      "learning_rate": 1.0015832507795577e-05,
      "loss": 0.7107,
      "step": 7
    },
    {
      "epoch": 0.06779661016949153,
      "grad_norm": 2.041591167449951,
      "learning_rate": 9.992727987946306e-06,
      "loss": 0.7948,
      "step": 8
    },
    {
      "epoch": 0.07627118644067797,
      "grad_norm": 1.9192763566970825,
      "learning_rate": 9.966113084682048e-06,
      "loss": 0.7323,
      "step": 9
    },
    {
      "epoch": 0.0847457627118644,
      "grad_norm": 2.181330919265747,
      "learning_rate": 9.936006662054137e-06,
      "loss": 0.7378,
      "step": 10
    },
    {
      "epoch": 0.09322033898305085,
      "grad_norm": 1.7582520246505737,
      "learning_rate": 9.902430058825162e-06,
      "loss": 0.7146,
      "step": 11
    },
    {
      "epoch": 0.1016949152542373,
      "grad_norm": 1.3104244470596313,
      "learning_rate": 9.865407073344506e-06,
      "loss": 0.6574,
      "step": 12
    },
    {
      "epoch": 0.11016949152542373,
      "grad_norm": 1.1520931720733643,
      "learning_rate": 9.824963946680641e-06,
      "loss": 0.674,
      "step": 13
    },
    {
      "epoch": 0.11864406779661017,
      "grad_norm": 1.3915627002716064,
      "learning_rate": 9.781129344022008e-06,
      "loss": 0.675,
      "step": 14
    },
    {
      "epoch": 0.1271186440677966,
      "grad_norm": 1.3260822296142578,
      "learning_rate": 9.733934334359786e-06,
      "loss": 0.6156,
      "step": 15
    },
    {
      "epoch": 0.13559322033898305,
      "grad_norm": 1.1522308588027954,
      "learning_rate": 9.683412368466881e-06,
      "loss": 0.6107,
      "step": 16
    },
    {
      "epoch": 0.1440677966101695,
      "grad_norm": 1.0242621898651123,
      "learning_rate": 9.629599255188763e-06,
      "loss": 0.544,
      "step": 17
    },
    {
      "epoch": 0.15254237288135594,
      "grad_norm": 1.5411088466644287,
      "learning_rate": 9.572533136062945e-06,
      "loss": 0.6265,
      "step": 18
    },
    {
      "epoch": 0.16101694915254236,
      "grad_norm": 1.2036213874816895,
      "learning_rate": 9.51225445828512e-06,
      "loss": 0.6261,
      "step": 19
    },
    {
      "epoch": 0.1694915254237288,
      "grad_norm": 1.004852533340454,
      "learning_rate": 9.448805946041084e-06,
      "loss": 0.5895,
      "step": 20
    },
    {
      "epoch": 0.17796610169491525,
      "grad_norm": 1.0301568508148193,
      "learning_rate": 9.382232570224783e-06,
      "loss": 0.6057,
      "step": 21
    },
    {
      "epoch": 0.1864406779661017,
      "grad_norm": 1.0345704555511475,
      "learning_rate": 9.312581516563942e-06,
      "loss": 0.5539,
      "step": 22
    },
    {
      "epoch": 0.19491525423728814,
      "grad_norm": 1.0441306829452515,
      "learning_rate": 9.239902152175876e-06,
      "loss": 0.5699,
      "step": 23
    },
    {
      "epoch": 0.2033898305084746,
      "grad_norm": 1.1769165992736816,
      "learning_rate": 9.164245990577172e-06,
      "loss": 0.573,
      "step": 24
    },
    {
      "epoch": 0.211864406779661,
      "grad_norm": 1.0624712705612183,
      "learning_rate": 9.085666655172057e-06,
      "loss": 0.634,
      "step": 25
    },
    {
      "epoch": 0.22033898305084745,
      "grad_norm": 1.0083361864089966,
      "learning_rate": 9.004219841245318e-06,
      "loss": 0.5617,
      "step": 26
    },
    {
      "epoch": 0.2288135593220339,
      "grad_norm": 1.0477335453033447,
      "learning_rate": 8.919963276486727e-06,
      "loss": 0.5579,
      "step": 27
    },
    {
      "epoch": 0.23728813559322035,
      "grad_norm": 1.229489803314209,
      "learning_rate": 8.832956680074937e-06,
      "loss": 0.6022,
      "step": 28
    },
    {
      "epoch": 0.2457627118644068,
      "grad_norm": 1.093316912651062,
      "learning_rate": 8.74326172034986e-06,
      "loss": 0.6176,
      "step": 29
    },
    {
      "epoch": 0.2542372881355932,
      "grad_norm": 1.174489974975586,
      "learning_rate": 8.650941971103509e-06,
      "loss": 0.5965,
      "step": 30
    },
    {
      "epoch": 0.2627118644067797,
      "grad_norm": 0.9779760837554932,
      "learning_rate": 8.556062866520323e-06,
      "loss": 0.5618,
      "step": 31
    },
    {
      "epoch": 0.2711864406779661,
      "grad_norm": 1.1349478960037231,
      "learning_rate": 8.458691654798865e-06,
      "loss": 0.6187,
      "step": 32
    },
    {
      "epoch": 0.2796610169491525,
      "grad_norm": 1.027665138244629,
      "learning_rate": 8.358897350487796e-06,
      "loss": 0.5923,
      "step": 33
    },
    {
      "epoch": 0.288135593220339,
      "grad_norm": 1.2860366106033325,
      "learning_rate": 8.256750685569892e-06,
      "loss": 0.6993,
      "step": 34
    },
    {
      "epoch": 0.2966101694915254,
      "grad_norm": 1.019265055656433,
      "learning_rate": 8.152324059328807e-06,
      "loss": 0.561,
      "step": 35
    },
    {
      "epoch": 0.3050847457627119,
      "grad_norm": 1.0647754669189453,
      "learning_rate": 8.045691487034047e-06,
      "loss": 0.5601,
      "step": 36
    },
    {
      "epoch": 0.3135593220338983,
      "grad_norm": 0.9456817507743835,
      "learning_rate": 7.936928547480611e-06,
      "loss": 0.5355,
      "step": 37
    },
    {
      "epoch": 0.3220338983050847,
      "grad_norm": 0.9935976266860962,
      "learning_rate": 7.826112329420414e-06,
      "loss": 0.5562,
      "step": 38
    },
    {
      "epoch": 0.3305084745762712,
      "grad_norm": 1.0699551105499268,
      "learning_rate": 7.713321376923502e-06,
      "loss": 0.5859,
      "step": 39
    },
    {
      "epoch": 0.3389830508474576,
      "grad_norm": 0.9607884287834167,
      "learning_rate": 7.598635633707753e-06,
      "loss": 0.5634,
      "step": 40
    },
    {
      "epoch": 0.3474576271186441,
      "grad_norm": 1.007079839706421,
      "learning_rate": 7.482136386476557e-06,
      "loss": 0.5743,
      "step": 41
    },
    {
      "epoch": 0.3559322033898305,
      "grad_norm": 1.0889322757720947,
      "learning_rate": 7.3639062073046025e-06,
      "loss": 0.6155,
      "step": 42
    },
    {
      "epoch": 0.3644067796610169,
      "grad_norm": 1.0192927122116089,
      "learning_rate": 7.244028895112632e-06,
      "loss": 0.5434,
      "step": 43
    },
    {
      "epoch": 0.3728813559322034,
      "grad_norm": 1.0717029571533203,
      "learning_rate": 7.122589416272626e-06,
      "loss": 0.5939,
      "step": 44
    },
    {
      "epoch": 0.3813559322033898,
      "grad_norm": 0.958198606967926,
      "learning_rate": 6.999673844385537e-06,
      "loss": 0.6275,
      "step": 45
    },
    {
      "epoch": 0.3898305084745763,
      "grad_norm": 1.0603076219558716,
      "learning_rate": 6.875369299274224e-06,
      "loss": 0.5774,
      "step": 46
    },
    {
      "epoch": 0.3983050847457627,
      "grad_norm": 1.0000678300857544,
      "learning_rate": 6.7497638852348716e-06,
      "loss": 0.569,
      "step": 47
    },
    {
      "epoch": 0.4067796610169492,
      "grad_norm": 1.054443359375,
      "learning_rate": 6.622946628590623e-06,
      "loss": 0.6047,
      "step": 48
    },
    {
      "epoch": 0.4152542372881356,
      "grad_norm": 1.0552459955215454,
      "learning_rate": 6.49500741459169e-06,
      "loss": 0.5411,
      "step": 49
    },
    {
      "epoch": 0.423728813559322,
      "grad_norm": 1.0494464635849,
      "learning_rate": 6.366036923706707e-06,
      "loss": 0.5714,
      "step": 50
    },
    {
      "epoch": 0.4322033898305085,
      "grad_norm": 0.915934681892395,
      "learning_rate": 6.23612656735042e-06,
      "loss": 0.5167,
      "step": 51
    },
    {
      "epoch": 0.4406779661016949,
      "grad_norm": 1.5039361715316772,
      "learning_rate": 6.105368423093307e-06,
      "loss": 0.6265,
      "step": 52
    },
    {
      "epoch": 0.4491525423728814,
      "grad_norm": 1.0584355592727661,
      "learning_rate": 5.973855169399057e-06,
      "loss": 0.5589,
      "step": 53
    },
    {
      "epoch": 0.4576271186440678,
      "grad_norm": 1.064541220664978,
      "learning_rate": 5.841680019936123e-06,
      "loss": 0.5228,
      "step": 54
    },
    {
      "epoch": 0.4661016949152542,
      "grad_norm": 1.0165396928787231,
      "learning_rate": 5.708936657509959e-06,
      "loss": 0.6116,
      "step": 55
    },
    {
      "epoch": 0.4745762711864407,
      "grad_norm": 1.0315970182418823,
      "learning_rate": 5.5757191676627215e-06,
      "loss": 0.5637,
      "step": 56
    },
    {
      "epoch": 0.4830508474576271,
      "grad_norm": 0.9197196960449219,
      "learning_rate": 5.442121971987539e-06,
      "loss": 0.5194,
      "step": 57
    },
    {
      "epoch": 0.4915254237288136,
      "grad_norm": 0.9129324555397034,
      "learning_rate": 5.3082397612045805e-06,
      "loss": 0.5541,
      "step": 58
    },
    {
      "epoch": 0.5,
      "grad_norm": 1.0417673587799072,
      "learning_rate": 5.174167428046385e-06,
      "loss": 0.5888,
      "step": 59
    },
    {
      "epoch": 0.5084745762711864,
      "grad_norm": 0.9219163656234741,
      "learning_rate": 5.04e-06,
      "loss": 0.5358,
      "step": 60
    },
    {
      "epoch": 0.5169491525423728,
      "grad_norm": 0.998306393623352,
      "learning_rate": 4.905832571953615e-06,
      "loss": 0.5352,
      "step": 61
    },
    {
      "epoch": 0.5254237288135594,
      "grad_norm": 1.092921495437622,
      "learning_rate": 4.771760238795421e-06,
      "loss": 0.5527,
      "step": 62
    },
    {
      "epoch": 0.5338983050847458,
      "grad_norm": 1.0643359422683716,
      "learning_rate": 4.637878028012461e-06,
      "loss": 0.6151,
      "step": 63
    },
    {
      "epoch": 0.5423728813559322,
      "grad_norm": 1.1080788373947144,
      "learning_rate": 4.504280832337279e-06,
      "loss": 0.5372,
      "step": 64
    },
    {
      "epoch": 0.5508474576271186,
      "grad_norm": 0.9985234141349792,
      "learning_rate": 4.371063342490041e-06,
      "loss": 0.5466,
      "step": 65
    },
    {
      "epoch": 0.559322033898305,
      "grad_norm": 1.0811630487442017,
      "learning_rate": 4.238319980063878e-06,
      "loss": 0.5848,
      "step": 66
    },
    {
      "epoch": 0.5677966101694916,
      "grad_norm": 1.0621752738952637,
      "learning_rate": 4.106144830600945e-06,
      "loss": 0.5546,
      "step": 67
    },
    {
      "epoch": 0.576271186440678,
      "grad_norm": 1.1586964130401611,
      "learning_rate": 3.974631576906693e-06,
      "loss": 0.5268,
      "step": 68
    },
    {
      "epoch": 0.5847457627118644,
      "grad_norm": 0.9288192987442017,
      "learning_rate": 3.843873432649581e-06,
      "loss": 0.564,
      "step": 69
    },
    {
      "epoch": 0.5932203389830508,
      "grad_norm": 0.9237483739852905,
      "learning_rate": 3.713963076293294e-06,
      "loss": 0.5849,
      "step": 70
    },
    {
      "epoch": 0.6016949152542372,
      "grad_norm": 1.2002196311950684,
      "learning_rate": 3.5849925854083115e-06,
      "loss": 0.5766,
      "step": 71
    },
    {
      "epoch": 0.6101694915254238,
      "grad_norm": 0.9985136389732361,
      "learning_rate": 3.4570533714093793e-06,
      "loss": 0.5443,
      "step": 72
    },
    {
      "epoch": 0.6186440677966102,
      "grad_norm": 1.052380084991455,
      "learning_rate": 3.3302361147651272e-06,
      "loss": 0.5679,
      "step": 73
    },
    {
      "epoch": 0.6271186440677966,
      "grad_norm": 1.0863173007965088,
      "learning_rate": 3.204630700725776e-06,
      "loss": 0.5995,
      "step": 74
    },
    {
      "epoch": 0.635593220338983,
      "grad_norm": 0.9472033381462097,
      "learning_rate": 3.080326155614464e-06,
      "loss": 0.5524,
      "step": 75
    },
    {
      "epoch": 0.6440677966101694,
      "grad_norm": 0.9481269717216492,
      "learning_rate": 2.9574105837273754e-06,
      "loss": 0.5578,
      "step": 76
    },
    {
      "epoch": 0.652542372881356,
      "grad_norm": 0.8579362630844116,
      "learning_rate": 2.835971104887369e-06,
      "loss": 0.5076,
      "step": 77
    },
    {
      "epoch": 0.6610169491525424,
      "grad_norm": 0.8677413463592529,
      "learning_rate": 2.7160937926953967e-06,
      "loss": 0.513,
      "step": 78
    },
    {
      "epoch": 0.6694915254237288,
      "grad_norm": 1.145965814590454,
      "learning_rate": 2.5978636135234437e-06,
      "loss": 0.5288,
      "step": 79
    },
    {
      "epoch": 0.6779661016949152,
      "grad_norm": 1.1086536645889282,
      "learning_rate": 2.481364366292248e-06,
      "loss": 0.6084,
      "step": 80
    },
    {
      "epoch": 0.6864406779661016,
      "grad_norm": 1.011196494102478,
      "learning_rate": 2.3666786230764983e-06,
      "loss": 0.6131,
      "step": 81
    },
    {
      "epoch": 0.6949152542372882,
      "grad_norm": 0.903338611125946,
      "learning_rate": 2.253887670579587e-06,
      "loss": 0.5771,
      "step": 82
    },
    {
      "epoch": 0.7033898305084746,
      "grad_norm": 1.0274643898010254,
      "learning_rate": 2.1430714525193897e-06,
      "loss": 0.5836,
      "step": 83
    },
    {
      "epoch": 0.711864406779661,
      "grad_norm": 1.0280394554138184,
      "learning_rate": 2.0343085129659523e-06,
      "loss": 0.5476,
      "step": 84
    },
    {
      "epoch": 0.7203389830508474,
      "grad_norm": 0.8955240845680237,
      "learning_rate": 1.9276759406711933e-06,
      "loss": 0.5065,
      "step": 85
    },
    {
      "epoch": 0.7288135593220338,
      "grad_norm": 0.9722875952720642,
      "learning_rate": 1.823249314430108e-06,
      "loss": 0.5386,
      "step": 86
    },
    {
      "epoch": 0.7372881355932204,
      "grad_norm": 1.010117769241333,
      "learning_rate": 1.7211026495122058e-06,
      "loss": 0.5898,
      "step": 87
    },
    {
      "epoch": 0.7457627118644068,
      "grad_norm": 0.9292120933532715,
      "learning_rate": 1.6213083452011349e-06,
      "loss": 0.5245,
      "step": 88
    },
    {
      "epoch": 0.7542372881355932,
      "grad_norm": 0.8837317228317261,
      "learning_rate": 1.5239371334796776e-06,
      "loss": 0.5642,
      "step": 89
    },
    {
      "epoch": 0.7627118644067796,
      "grad_norm": 1.1663438081741333,
      "learning_rate": 1.4290580288964914e-06,
      "loss": 0.593,
      "step": 90
    },
    {
      "epoch": 0.7711864406779662,
      "grad_norm": 1.0487372875213623,
      "learning_rate": 1.336738279650142e-06,
      "loss": 0.5621,
      "step": 91
    },
    {
      "epoch": 0.7796610169491526,
      "grad_norm": 1.05958092212677,
      "learning_rate": 1.247043319925062e-06,
      "loss": 0.5409,
      "step": 92
    },
    {
      "epoch": 0.788135593220339,
      "grad_norm": 0.839719831943512,
      "learning_rate": 1.1600367235132717e-06,
      "loss": 0.5105,
      "step": 93
    },
    {
      "epoch": 0.7966101694915254,
      "grad_norm": 1.0145214796066284,
      "learning_rate": 1.075780158754682e-06,
      "loss": 0.5873,
      "step": 94
    },
    {
      "epoch": 0.8050847457627118,
      "grad_norm": 1.016517996788025,
      "learning_rate": 9.943333448279442e-07,
      "loss": 0.564,
      "step": 95
    },
    {
      "epoch": 0.8135593220338984,
      "grad_norm": 0.8958804607391357,
      "learning_rate": 9.157540094228282e-07,
      "loss": 0.58,
      "step": 96
    },
    {
      "epoch": 0.8220338983050848,
      "grad_norm": 0.916645884513855,
      "learning_rate": 8.400978478241241e-07,
      "loss": 0.4649,
      "step": 97
    },
    {
      "epoch": 0.8305084745762712,
      "grad_norm": 0.8990329504013062,
      "learning_rate": 7.674184834360587e-07,
      "loss": 0.5424,
      "step": 98
    },
    {
      "epoch": 0.8389830508474576,
      "grad_norm": 0.9331949949264526,
      "learning_rate": 6.977674297752169e-07,
      "loss": 0.5427,
      "step": 99
    },
    {
      "epoch": 0.847457627118644,
      "grad_norm": 0.9024314284324646,
      "learning_rate": 6.311940539589152e-07,
      "loss": 0.588,
      "step": 100
    },
    {
      "epoch": 0.8559322033898306,
      "grad_norm": 0.9858136177062988,
      "learning_rate": 5.677455417148799e-07,
      "loss": 0.6236,
      "step": 101
    },
    {
      "epoch": 0.864406779661017,
      "grad_norm": 0.924796462059021,
      "learning_rate": 5.074668639370556e-07,
      "loss": 0.5011,
      "step": 102
    },
    {
      "epoch": 0.8728813559322034,
      "grad_norm": 0.9715655446052551,
      "learning_rate": 4.504007448112373e-07,
      "loss": 0.568,
      "step": 103
    },
    {
      "epoch": 0.8813559322033898,
      "grad_norm": 0.9612396955490112,
      "learning_rate": 3.9658763153311863e-07,
      "loss": 0.5957,
      "step": 104
    },
    {
      "epoch": 0.8898305084745762,
      "grad_norm": 0.8345690369606018,
      "learning_rate": 3.460656656402143e-07,
      "loss": 0.5547,
      "step": 105
    },
    {
      "epoch": 0.8983050847457628,
      "grad_norm": 0.9040005207061768,
      "learning_rate": 2.9887065597799247e-07,
      "loss": 0.5367,
      "step": 106
    },
    {
      "epoch": 0.9067796610169492,
      "grad_norm": 0.7813464403152466,
      "learning_rate": 2.5503605331935833e-07,
      "loss": 0.5397,
      "step": 107
    },
    {
      "epoch": 0.9152542372881356,
      "grad_norm": 1.0138887166976929,
      "learning_rate": 2.145929266554925e-07,
      "loss": 0.5478,
      "step": 108
    },
    {
      "epoch": 0.923728813559322,
      "grad_norm": 1.2826368808746338,
      "learning_rate": 1.7756994117483972e-07,
      "loss": 0.5731,
      "step": 109
    },
    {
      "epoch": 0.9322033898305084,
      "grad_norm": 1.0911537408828735,
      "learning_rate": 1.4399333794586233e-07,
      "loss": 0.585,
      "step": 110
    },
    {
      "epoch": 0.940677966101695,
      "grad_norm": 0.9331554174423218,
      "learning_rate": 1.1388691531795196e-07,
      "loss": 0.5553,
      "step": 111
    },
    {
      "epoch": 0.9491525423728814,
      "grad_norm": 1.0142954587936401,
      "learning_rate": 8.727201205369431e-08,
      "loss": 0.5371,
      "step": 112
    },
    {
      "epoch": 0.9576271186440678,
      "grad_norm": 0.9292749762535095,
      "learning_rate": 6.416749220442344e-08,
      "loss": 0.5288,
      "step": 113
    },
    {
      "epoch": 0.9661016949152542,
      "grad_norm": 0.8662624955177307,
      "learning_rate": 4.4589731739804295e-08,
      "loss": 0.5724,
      "step": 114
    },
    {
      "epoch": 0.9745762711864406,
      "grad_norm": 0.8610790371894836,
      "learning_rate": 2.855260694090905e-08,
      "loss": 0.4941,
      "step": 115
    },
    {
      "epoch": 0.9830508474576272,
      "grad_norm": 0.8874425292015076,
      "learning_rate": 1.6067484565019372e-08,
      "loss": 0.5518,
      "step": 116
    },
    {
      "epoch": 0.9915254237288136,
      "grad_norm": 0.9031229615211487,
      "learning_rate": 7.1432137891198e-09,
      "loss": 0.5562,
      "step": 117
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.8897117972373962,
      "learning_rate": 1.7861199377987623e-09,
      "loss": 0.562,
      "step": 118
    }
  ],
  "logging_steps": 1.0,
  "max_steps": 118,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "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": 1.3048368484214374e+17,
  "train_batch_size": 3,
  "trial_name": null,
  "trial_params": null
}