lillian039's picture
Model save
3bdd32d verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 3.0,
"eval_steps": 500,
"global_step": 456,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.006578947368421052,
"grad_norm": 0.9301734884942886,
"learning_rate": 4.347826086956522e-06,
"loss": 1.3345,
"step": 1
},
{
"epoch": 0.013157894736842105,
"grad_norm": 0.8660026064620485,
"learning_rate": 8.695652173913044e-06,
"loss": 1.2699,
"step": 2
},
{
"epoch": 0.019736842105263157,
"grad_norm": 0.9006761056531956,
"learning_rate": 1.3043478260869566e-05,
"loss": 1.3134,
"step": 3
},
{
"epoch": 0.02631578947368421,
"grad_norm": 0.8334670285502003,
"learning_rate": 1.739130434782609e-05,
"loss": 1.2548,
"step": 4
},
{
"epoch": 0.03289473684210526,
"grad_norm": 0.8574803030028485,
"learning_rate": 2.173913043478261e-05,
"loss": 1.2843,
"step": 5
},
{
"epoch": 0.039473684210526314,
"grad_norm": 0.8390817055257793,
"learning_rate": 2.608695652173913e-05,
"loss": 1.2982,
"step": 6
},
{
"epoch": 0.046052631578947366,
"grad_norm": 0.7141521506417161,
"learning_rate": 3.0434782608695656e-05,
"loss": 1.1906,
"step": 7
},
{
"epoch": 0.05263157894736842,
"grad_norm": 0.6047790773556203,
"learning_rate": 3.478260869565218e-05,
"loss": 1.1499,
"step": 8
},
{
"epoch": 0.05921052631578947,
"grad_norm": 0.5265109630433017,
"learning_rate": 3.91304347826087e-05,
"loss": 1.0976,
"step": 9
},
{
"epoch": 0.06578947368421052,
"grad_norm": 0.5245802695709278,
"learning_rate": 4.347826086956522e-05,
"loss": 1.0108,
"step": 10
},
{
"epoch": 0.07236842105263158,
"grad_norm": 0.5599855466933923,
"learning_rate": 4.782608695652174e-05,
"loss": 0.9675,
"step": 11
},
{
"epoch": 0.07894736842105263,
"grad_norm": 0.5570743219368649,
"learning_rate": 5.217391304347826e-05,
"loss": 0.8945,
"step": 12
},
{
"epoch": 0.08552631578947369,
"grad_norm": 0.5874369878213753,
"learning_rate": 5.652173913043478e-05,
"loss": 0.8503,
"step": 13
},
{
"epoch": 0.09210526315789473,
"grad_norm": 0.522753964863263,
"learning_rate": 6.086956521739131e-05,
"loss": 0.7773,
"step": 14
},
{
"epoch": 0.09868421052631579,
"grad_norm": 0.5330305040727032,
"learning_rate": 6.521739130434783e-05,
"loss": 0.7115,
"step": 15
},
{
"epoch": 0.10526315789473684,
"grad_norm": 0.46085140769365757,
"learning_rate": 6.956521739130436e-05,
"loss": 0.6463,
"step": 16
},
{
"epoch": 0.1118421052631579,
"grad_norm": 0.40848688620706747,
"learning_rate": 7.391304347826086e-05,
"loss": 0.6151,
"step": 17
},
{
"epoch": 0.11842105263157894,
"grad_norm": 0.2704450670383811,
"learning_rate": 7.82608695652174e-05,
"loss": 0.5724,
"step": 18
},
{
"epoch": 0.125,
"grad_norm": 0.26085072447544416,
"learning_rate": 8.260869565217392e-05,
"loss": 0.5528,
"step": 19
},
{
"epoch": 0.13157894736842105,
"grad_norm": 0.2339583161738833,
"learning_rate": 8.695652173913044e-05,
"loss": 0.5369,
"step": 20
},
{
"epoch": 0.13815789473684212,
"grad_norm": 0.1949857904626772,
"learning_rate": 9.130434782608696e-05,
"loss": 0.512,
"step": 21
},
{
"epoch": 0.14473684210526316,
"grad_norm": 0.2435718076959428,
"learning_rate": 9.565217391304348e-05,
"loss": 0.5202,
"step": 22
},
{
"epoch": 0.1513157894736842,
"grad_norm": 0.22205719081769748,
"learning_rate": 0.0001,
"loss": 0.5221,
"step": 23
},
{
"epoch": 0.15789473684210525,
"grad_norm": 0.21129557114647404,
"learning_rate": 0.00010434782608695653,
"loss": 0.5128,
"step": 24
},
{
"epoch": 0.16447368421052633,
"grad_norm": 0.18916988013906638,
"learning_rate": 0.00010869565217391305,
"loss": 0.4869,
"step": 25
},
{
"epoch": 0.17105263157894737,
"grad_norm": 0.14949200071673818,
"learning_rate": 0.00011304347826086956,
"loss": 0.4821,
"step": 26
},
{
"epoch": 0.17763157894736842,
"grad_norm": 0.1581846054226008,
"learning_rate": 0.0001173913043478261,
"loss": 0.4661,
"step": 27
},
{
"epoch": 0.18421052631578946,
"grad_norm": 0.134913498610646,
"learning_rate": 0.00012173913043478263,
"loss": 0.4758,
"step": 28
},
{
"epoch": 0.19078947368421054,
"grad_norm": 0.12244728979819958,
"learning_rate": 0.00012608695652173915,
"loss": 0.4504,
"step": 29
},
{
"epoch": 0.19736842105263158,
"grad_norm": 0.12177957867071984,
"learning_rate": 0.00013043478260869567,
"loss": 0.4788,
"step": 30
},
{
"epoch": 0.20394736842105263,
"grad_norm": 0.109229502136198,
"learning_rate": 0.0001347826086956522,
"loss": 0.4434,
"step": 31
},
{
"epoch": 0.21052631578947367,
"grad_norm": 0.11628917220798876,
"learning_rate": 0.0001391304347826087,
"loss": 0.4536,
"step": 32
},
{
"epoch": 0.21710526315789475,
"grad_norm": 0.11733881064063928,
"learning_rate": 0.0001434782608695652,
"loss": 0.4742,
"step": 33
},
{
"epoch": 0.2236842105263158,
"grad_norm": 0.11627552292165463,
"learning_rate": 0.00014782608695652173,
"loss": 0.4494,
"step": 34
},
{
"epoch": 0.23026315789473684,
"grad_norm": 0.11476166833257723,
"learning_rate": 0.00015217391304347827,
"loss": 0.4319,
"step": 35
},
{
"epoch": 0.23684210526315788,
"grad_norm": 0.10237139066203954,
"learning_rate": 0.0001565217391304348,
"loss": 0.4263,
"step": 36
},
{
"epoch": 0.24342105263157895,
"grad_norm": 0.10918908589611533,
"learning_rate": 0.00016086956521739132,
"loss": 0.4254,
"step": 37
},
{
"epoch": 0.25,
"grad_norm": 0.10679450577166048,
"learning_rate": 0.00016521739130434784,
"loss": 0.4363,
"step": 38
},
{
"epoch": 0.2565789473684211,
"grad_norm": 0.10517151889540283,
"learning_rate": 0.00016956521739130436,
"loss": 0.4353,
"step": 39
},
{
"epoch": 0.2631578947368421,
"grad_norm": 0.10453003257689569,
"learning_rate": 0.00017391304347826088,
"loss": 0.4313,
"step": 40
},
{
"epoch": 0.26973684210526316,
"grad_norm": 0.09705593623180399,
"learning_rate": 0.0001782608695652174,
"loss": 0.4146,
"step": 41
},
{
"epoch": 0.27631578947368424,
"grad_norm": 0.09840834125049662,
"learning_rate": 0.00018260869565217392,
"loss": 0.414,
"step": 42
},
{
"epoch": 0.28289473684210525,
"grad_norm": 0.09733214025720571,
"learning_rate": 0.00018695652173913045,
"loss": 0.4163,
"step": 43
},
{
"epoch": 0.2894736842105263,
"grad_norm": 0.09403117831890086,
"learning_rate": 0.00019130434782608697,
"loss": 0.4095,
"step": 44
},
{
"epoch": 0.29605263157894735,
"grad_norm": 0.0998536829474165,
"learning_rate": 0.0001956521739130435,
"loss": 0.4075,
"step": 45
},
{
"epoch": 0.3026315789473684,
"grad_norm": 0.09752527766251139,
"learning_rate": 0.0002,
"loss": 0.4194,
"step": 46
},
{
"epoch": 0.3092105263157895,
"grad_norm": 0.09784770986893948,
"learning_rate": 0.0001999970643795027,
"loss": 0.4109,
"step": 47
},
{
"epoch": 0.3157894736842105,
"grad_norm": 0.09501975538592294,
"learning_rate": 0.00019998825769036819,
"loss": 0.4116,
"step": 48
},
{
"epoch": 0.3223684210526316,
"grad_norm": 0.09520308666764123,
"learning_rate": 0.0001999735804496583,
"loss": 0.4016,
"step": 49
},
{
"epoch": 0.32894736842105265,
"grad_norm": 0.09458518327177581,
"learning_rate": 0.0001999530335191093,
"loss": 0.4038,
"step": 50
},
{
"epoch": 0.3355263157894737,
"grad_norm": 0.08881816935147156,
"learning_rate": 0.000199926618105081,
"loss": 0.3881,
"step": 51
},
{
"epoch": 0.34210526315789475,
"grad_norm": 0.09507811168988711,
"learning_rate": 0.000199894335758486,
"loss": 0.4101,
"step": 52
},
{
"epoch": 0.34868421052631576,
"grad_norm": 0.0892408897509568,
"learning_rate": 0.00019985618837469864,
"loss": 0.3976,
"step": 53
},
{
"epoch": 0.35526315789473684,
"grad_norm": 0.08932074970893863,
"learning_rate": 0.0001998121781934438,
"loss": 0.3869,
"step": 54
},
{
"epoch": 0.3618421052631579,
"grad_norm": 0.09109394290790748,
"learning_rate": 0.00019976230779866525,
"loss": 0.3898,
"step": 55
},
{
"epoch": 0.3684210526315789,
"grad_norm": 0.09116027892556224,
"learning_rate": 0.00019970658011837404,
"loss": 0.398,
"step": 56
},
{
"epoch": 0.375,
"grad_norm": 0.09891633296064052,
"learning_rate": 0.00019964499842447663,
"loss": 0.4046,
"step": 57
},
{
"epoch": 0.3815789473684211,
"grad_norm": 0.0939928712835388,
"learning_rate": 0.00019957756633258265,
"loss": 0.3864,
"step": 58
},
{
"epoch": 0.3881578947368421,
"grad_norm": 0.09374915175716102,
"learning_rate": 0.00019950428780179274,
"loss": 0.3951,
"step": 59
},
{
"epoch": 0.39473684210526316,
"grad_norm": 0.09474512693386154,
"learning_rate": 0.000199425167134466,
"loss": 0.3911,
"step": 60
},
{
"epoch": 0.40131578947368424,
"grad_norm": 0.09534030624628262,
"learning_rate": 0.0001993402089759675,
"loss": 0.3917,
"step": 61
},
{
"epoch": 0.40789473684210525,
"grad_norm": 0.09308439623268501,
"learning_rate": 0.0001992494183143955,
"loss": 0.3799,
"step": 62
},
{
"epoch": 0.4144736842105263,
"grad_norm": 0.0963241819886369,
"learning_rate": 0.00019915280048028852,
"loss": 0.3958,
"step": 63
},
{
"epoch": 0.42105263157894735,
"grad_norm": 0.09034072116748955,
"learning_rate": 0.00019905036114631247,
"loss": 0.3735,
"step": 64
},
{
"epoch": 0.4276315789473684,
"grad_norm": 0.09570337959260364,
"learning_rate": 0.00019894210632692745,
"loss": 0.3854,
"step": 65
},
{
"epoch": 0.4342105263157895,
"grad_norm": 0.091943482076957,
"learning_rate": 0.00019882804237803488,
"loss": 0.3839,
"step": 66
},
{
"epoch": 0.4407894736842105,
"grad_norm": 0.09392178800966167,
"learning_rate": 0.000198708175996604,
"loss": 0.3922,
"step": 67
},
{
"epoch": 0.4473684210526316,
"grad_norm": 0.09032901472780272,
"learning_rate": 0.00019858251422027903,
"loss": 0.3827,
"step": 68
},
{
"epoch": 0.45394736842105265,
"grad_norm": 0.08955279616077026,
"learning_rate": 0.00019845106442696562,
"loss": 0.3782,
"step": 69
},
{
"epoch": 0.4605263157894737,
"grad_norm": 0.08888768341850095,
"learning_rate": 0.00019831383433439797,
"loss": 0.3803,
"step": 70
},
{
"epoch": 0.46710526315789475,
"grad_norm": 0.10135869270529443,
"learning_rate": 0.00019817083199968552,
"loss": 0.3876,
"step": 71
},
{
"epoch": 0.47368421052631576,
"grad_norm": 0.09494931509695588,
"learning_rate": 0.00019802206581883994,
"loss": 0.3878,
"step": 72
},
{
"epoch": 0.48026315789473684,
"grad_norm": 0.09731569285786043,
"learning_rate": 0.00019786754452628225,
"loss": 0.3731,
"step": 73
},
{
"epoch": 0.4868421052631579,
"grad_norm": 0.09286168859706302,
"learning_rate": 0.00019770727719432994,
"loss": 0.3852,
"step": 74
},
{
"epoch": 0.4934210526315789,
"grad_norm": 0.09331628340360382,
"learning_rate": 0.00019754127323266428,
"loss": 0.3679,
"step": 75
},
{
"epoch": 0.5,
"grad_norm": 0.0985838743425182,
"learning_rate": 0.00019736954238777792,
"loss": 0.3751,
"step": 76
},
{
"epoch": 0.506578947368421,
"grad_norm": 0.08784090808893116,
"learning_rate": 0.00019719209474240262,
"loss": 0.3857,
"step": 77
},
{
"epoch": 0.5131578947368421,
"grad_norm": 0.09595397552494361,
"learning_rate": 0.00019700894071491732,
"loss": 0.3772,
"step": 78
},
{
"epoch": 0.5197368421052632,
"grad_norm": 0.09751786827076832,
"learning_rate": 0.00019682009105873635,
"loss": 0.3783,
"step": 79
},
{
"epoch": 0.5263157894736842,
"grad_norm": 0.08800370843142873,
"learning_rate": 0.00019662555686167808,
"loss": 0.3662,
"step": 80
},
{
"epoch": 0.5328947368421053,
"grad_norm": 0.0897695028264937,
"learning_rate": 0.0001964253495453141,
"loss": 0.3524,
"step": 81
},
{
"epoch": 0.5394736842105263,
"grad_norm": 0.09955413713238825,
"learning_rate": 0.00019621948086429844,
"loss": 0.3893,
"step": 82
},
{
"epoch": 0.5460526315789473,
"grad_norm": 0.09349570369317932,
"learning_rate": 0.00019600796290567747,
"loss": 0.3677,
"step": 83
},
{
"epoch": 0.5526315789473685,
"grad_norm": 0.08892170997103152,
"learning_rate": 0.00019579080808818032,
"loss": 0.3695,
"step": 84
},
{
"epoch": 0.5592105263157895,
"grad_norm": 0.09487589068069692,
"learning_rate": 0.00019556802916148962,
"loss": 0.3758,
"step": 85
},
{
"epoch": 0.5657894736842105,
"grad_norm": 0.08802853628490082,
"learning_rate": 0.00019533963920549306,
"loss": 0.3611,
"step": 86
},
{
"epoch": 0.5723684210526315,
"grad_norm": 0.0952039827447668,
"learning_rate": 0.00019510565162951537,
"loss": 0.3748,
"step": 87
},
{
"epoch": 0.5789473684210527,
"grad_norm": 0.08848254195231814,
"learning_rate": 0.000194866080171531,
"loss": 0.3599,
"step": 88
},
{
"epoch": 0.5855263157894737,
"grad_norm": 0.09613545225239202,
"learning_rate": 0.00019462093889735765,
"loss": 0.3733,
"step": 89
},
{
"epoch": 0.5921052631578947,
"grad_norm": 0.08924377105304611,
"learning_rate": 0.00019437024219983028,
"loss": 0.3788,
"step": 90
},
{
"epoch": 0.5986842105263158,
"grad_norm": 0.0970473798319214,
"learning_rate": 0.00019411400479795617,
"loss": 0.3768,
"step": 91
},
{
"epoch": 0.6052631578947368,
"grad_norm": 0.09487853197595246,
"learning_rate": 0.0001938522417360507,
"loss": 0.3707,
"step": 92
},
{
"epoch": 0.6118421052631579,
"grad_norm": 0.08878860279782896,
"learning_rate": 0.00019358496838285406,
"loss": 0.3709,
"step": 93
},
{
"epoch": 0.618421052631579,
"grad_norm": 0.0894046820861241,
"learning_rate": 0.00019331220043062894,
"loss": 0.3732,
"step": 94
},
{
"epoch": 0.625,
"grad_norm": 0.0873587537636765,
"learning_rate": 0.00019303395389423918,
"loss": 0.378,
"step": 95
},
{
"epoch": 0.631578947368421,
"grad_norm": 0.09689373631878609,
"learning_rate": 0.0001927502451102095,
"loss": 0.3696,
"step": 96
},
{
"epoch": 0.6381578947368421,
"grad_norm": 0.09705847084532389,
"learning_rate": 0.00019246109073576628,
"loss": 0.3604,
"step": 97
},
{
"epoch": 0.6447368421052632,
"grad_norm": 0.09181701291383483,
"learning_rate": 0.00019216650774785972,
"loss": 0.3711,
"step": 98
},
{
"epoch": 0.6513157894736842,
"grad_norm": 0.09684384392230812,
"learning_rate": 0.00019186651344216702,
"loss": 0.3736,
"step": 99
},
{
"epoch": 0.6578947368421053,
"grad_norm": 0.10298393644504046,
"learning_rate": 0.00019156112543207673,
"loss": 0.3802,
"step": 100
},
{
"epoch": 0.6644736842105263,
"grad_norm": 0.09420976346161887,
"learning_rate": 0.000191250361647655,
"loss": 0.3672,
"step": 101
},
{
"epoch": 0.6710526315789473,
"grad_norm": 0.08630124052676108,
"learning_rate": 0.00019093424033459248,
"loss": 0.3709,
"step": 102
},
{
"epoch": 0.6776315789473685,
"grad_norm": 0.09501717321245914,
"learning_rate": 0.0001906127800531333,
"loss": 0.366,
"step": 103
},
{
"epoch": 0.6842105263157895,
"grad_norm": 0.08821786734234693,
"learning_rate": 0.00019028599967698531,
"loss": 0.3689,
"step": 104
},
{
"epoch": 0.6907894736842105,
"grad_norm": 0.08658177205492697,
"learning_rate": 0.0001899539183922119,
"loss": 0.3541,
"step": 105
},
{
"epoch": 0.6973684210526315,
"grad_norm": 0.09325609783945248,
"learning_rate": 0.00018961655569610557,
"loss": 0.3721,
"step": 106
},
{
"epoch": 0.7039473684210527,
"grad_norm": 0.08841280838509634,
"learning_rate": 0.00018927393139604325,
"loss": 0.3519,
"step": 107
},
{
"epoch": 0.7105263157894737,
"grad_norm": 0.09690514694100973,
"learning_rate": 0.00018892606560832337,
"loss": 0.3774,
"step": 108
},
{
"epoch": 0.7171052631578947,
"grad_norm": 0.09722105218508262,
"learning_rate": 0.00018857297875698455,
"loss": 0.3558,
"step": 109
},
{
"epoch": 0.7236842105263158,
"grad_norm": 0.08839435869483182,
"learning_rate": 0.00018821469157260685,
"loss": 0.3565,
"step": 110
},
{
"epoch": 0.7302631578947368,
"grad_norm": 0.08814358581045939,
"learning_rate": 0.00018785122509109426,
"loss": 0.362,
"step": 111
},
{
"epoch": 0.7368421052631579,
"grad_norm": 0.08994997062082188,
"learning_rate": 0.00018748260065243986,
"loss": 0.3455,
"step": 112
},
{
"epoch": 0.743421052631579,
"grad_norm": 0.08543302223033665,
"learning_rate": 0.0001871088398994728,
"loss": 0.362,
"step": 113
},
{
"epoch": 0.75,
"grad_norm": 0.08894111865879421,
"learning_rate": 0.00018672996477658767,
"loss": 0.3643,
"step": 114
},
{
"epoch": 0.756578947368421,
"grad_norm": 0.08853456191177575,
"learning_rate": 0.00018634599752845592,
"loss": 0.3658,
"step": 115
},
{
"epoch": 0.7631578947368421,
"grad_norm": 0.09412265410116702,
"learning_rate": 0.00018595696069872013,
"loss": 0.3732,
"step": 116
},
{
"epoch": 0.7697368421052632,
"grad_norm": 0.08843510819092883,
"learning_rate": 0.00018556287712867005,
"loss": 0.3498,
"step": 117
},
{
"epoch": 0.7763157894736842,
"grad_norm": 0.09005908717348306,
"learning_rate": 0.00018516376995590187,
"loss": 0.343,
"step": 118
},
{
"epoch": 0.7828947368421053,
"grad_norm": 0.09003456135164496,
"learning_rate": 0.00018475966261295948,
"loss": 0.3708,
"step": 119
},
{
"epoch": 0.7894736842105263,
"grad_norm": 0.09116449509404734,
"learning_rate": 0.00018435057882595882,
"loss": 0.3548,
"step": 120
},
{
"epoch": 0.7960526315789473,
"grad_norm": 0.09200874674641456,
"learning_rate": 0.000183936542613195,
"loss": 0.3728,
"step": 121
},
{
"epoch": 0.8026315789473685,
"grad_norm": 0.08854192816932756,
"learning_rate": 0.0001835175782837318,
"loss": 0.3554,
"step": 122
},
{
"epoch": 0.8092105263157895,
"grad_norm": 0.0902705983711389,
"learning_rate": 0.00018309371043597473,
"loss": 0.3708,
"step": 123
},
{
"epoch": 0.8157894736842105,
"grad_norm": 0.08478855503527433,
"learning_rate": 0.0001826649639562266,
"loss": 0.3455,
"step": 124
},
{
"epoch": 0.8223684210526315,
"grad_norm": 0.08988565845179716,
"learning_rate": 0.0001822313640172265,
"loss": 0.3552,
"step": 125
},
{
"epoch": 0.8289473684210527,
"grad_norm": 0.08874881990127954,
"learning_rate": 0.00018179293607667178,
"loss": 0.3587,
"step": 126
},
{
"epoch": 0.8355263157894737,
"grad_norm": 0.09180650586751762,
"learning_rate": 0.00018134970587572343,
"loss": 0.3577,
"step": 127
},
{
"epoch": 0.8421052631578947,
"grad_norm": 0.09068943599513157,
"learning_rate": 0.00018090169943749476,
"loss": 0.3616,
"step": 128
},
{
"epoch": 0.8486842105263158,
"grad_norm": 0.09063538270613061,
"learning_rate": 0.00018044894306552337,
"loss": 0.3732,
"step": 129
},
{
"epoch": 0.8552631578947368,
"grad_norm": 0.0901373223301407,
"learning_rate": 0.00017999146334222695,
"loss": 0.3345,
"step": 130
},
{
"epoch": 0.8618421052631579,
"grad_norm": 0.0885794928608872,
"learning_rate": 0.00017952928712734268,
"loss": 0.3498,
"step": 131
},
{
"epoch": 0.868421052631579,
"grad_norm": 0.08860093481405781,
"learning_rate": 0.00017906244155634982,
"loss": 0.3658,
"step": 132
},
{
"epoch": 0.875,
"grad_norm": 0.08977436227076488,
"learning_rate": 0.00017859095403887696,
"loss": 0.3579,
"step": 133
},
{
"epoch": 0.881578947368421,
"grad_norm": 0.08940660194253588,
"learning_rate": 0.00017811485225709256,
"loss": 0.3537,
"step": 134
},
{
"epoch": 0.8881578947368421,
"grad_norm": 0.08080512906819583,
"learning_rate": 0.00017763416416407952,
"loss": 0.3379,
"step": 135
},
{
"epoch": 0.8947368421052632,
"grad_norm": 0.08837755359593251,
"learning_rate": 0.0001771489179821943,
"loss": 0.3524,
"step": 136
},
{
"epoch": 0.9013157894736842,
"grad_norm": 0.08523392262655982,
"learning_rate": 0.00017665914220140963,
"loss": 0.3386,
"step": 137
},
{
"epoch": 0.9078947368421053,
"grad_norm": 0.09103630477831169,
"learning_rate": 0.00017616486557764187,
"loss": 0.3659,
"step": 138
},
{
"epoch": 0.9144736842105263,
"grad_norm": 0.09441590695909564,
"learning_rate": 0.00017566611713106287,
"loss": 0.3507,
"step": 139
},
{
"epoch": 0.9210526315789473,
"grad_norm": 0.09451841778053577,
"learning_rate": 0.00017516292614439585,
"loss": 0.357,
"step": 140
},
{
"epoch": 0.9276315789473685,
"grad_norm": 0.08792822153458119,
"learning_rate": 0.00017465532216119625,
"loss": 0.3369,
"step": 141
},
{
"epoch": 0.9342105263157895,
"grad_norm": 0.09847169501416486,
"learning_rate": 0.00017414333498411733,
"loss": 0.3537,
"step": 142
},
{
"epoch": 0.9407894736842105,
"grad_norm": 0.09535814851654961,
"learning_rate": 0.00017362699467316005,
"loss": 0.3688,
"step": 143
},
{
"epoch": 0.9473684210526315,
"grad_norm": 0.0894204821997548,
"learning_rate": 0.0001731063315439084,
"loss": 0.3562,
"step": 144
},
{
"epoch": 0.9539473684210527,
"grad_norm": 0.08747335966415704,
"learning_rate": 0.0001725813761657495,
"loss": 0.3431,
"step": 145
},
{
"epoch": 0.9605263157894737,
"grad_norm": 0.09306693938804542,
"learning_rate": 0.0001720521593600787,
"loss": 0.3463,
"step": 146
},
{
"epoch": 0.9671052631578947,
"grad_norm": 0.09389776996796227,
"learning_rate": 0.00017151871219849008,
"loss": 0.3585,
"step": 147
},
{
"epoch": 0.9736842105263158,
"grad_norm": 0.08560502715505004,
"learning_rate": 0.00017098106600095205,
"loss": 0.3465,
"step": 148
},
{
"epoch": 0.9802631578947368,
"grad_norm": 0.08501659921152092,
"learning_rate": 0.00017043925233396854,
"loss": 0.3481,
"step": 149
},
{
"epoch": 0.9868421052631579,
"grad_norm": 0.09086546243483316,
"learning_rate": 0.00016989330300872576,
"loss": 0.355,
"step": 150
},
{
"epoch": 0.993421052631579,
"grad_norm": 0.08269896391914834,
"learning_rate": 0.00016934325007922417,
"loss": 0.347,
"step": 151
},
{
"epoch": 1.0,
"grad_norm": 0.08946228022278917,
"learning_rate": 0.000168789125840397,
"loss": 0.3519,
"step": 152
},
{
"epoch": 1.0,
"eval_loss": 0.3528538942337036,
"eval_runtime": 165.9892,
"eval_samples_per_second": 30.755,
"eval_steps_per_second": 0.964,
"step": 152
},
{
"epoch": 1.006578947368421,
"grad_norm": 0.08542638599767739,
"learning_rate": 0.00016823096282621365,
"loss": 0.3457,
"step": 153
},
{
"epoch": 1.013157894736842,
"grad_norm": 0.08540863800720608,
"learning_rate": 0.0001676687938077698,
"loss": 0.3323,
"step": 154
},
{
"epoch": 1.019736842105263,
"grad_norm": 0.08863302632538074,
"learning_rate": 0.0001671026517913634,
"loss": 0.3447,
"step": 155
},
{
"epoch": 1.0263157894736843,
"grad_norm": 0.08334205967004667,
"learning_rate": 0.00016653257001655652,
"loss": 0.3337,
"step": 156
},
{
"epoch": 1.0328947368421053,
"grad_norm": 0.08918626067677893,
"learning_rate": 0.0001659585819542241,
"loss": 0.343,
"step": 157
},
{
"epoch": 1.0394736842105263,
"grad_norm": 0.09996420454196991,
"learning_rate": 0.00016538072130458853,
"loss": 0.3529,
"step": 158
},
{
"epoch": 1.0460526315789473,
"grad_norm": 0.08514205522788368,
"learning_rate": 0.00016479902199524116,
"loss": 0.3338,
"step": 159
},
{
"epoch": 1.0526315789473684,
"grad_norm": 0.08426879308880401,
"learning_rate": 0.00016421351817915024,
"loss": 0.3347,
"step": 160
},
{
"epoch": 1.0592105263157894,
"grad_norm": 0.09335534193359232,
"learning_rate": 0.00016362424423265598,
"loss": 0.3396,
"step": 161
},
{
"epoch": 1.0657894736842106,
"grad_norm": 0.09585726718482877,
"learning_rate": 0.00016303123475345182,
"loss": 0.3471,
"step": 162
},
{
"epoch": 1.0723684210526316,
"grad_norm": 0.08521089701402258,
"learning_rate": 0.00016243452455855342,
"loss": 0.333,
"step": 163
},
{
"epoch": 1.0789473684210527,
"grad_norm": 0.08924725173949923,
"learning_rate": 0.00016183414868225433,
"loss": 0.3311,
"step": 164
},
{
"epoch": 1.0855263157894737,
"grad_norm": 0.08988680540431337,
"learning_rate": 0.00016123014237406913,
"loss": 0.3298,
"step": 165
},
{
"epoch": 1.0921052631578947,
"grad_norm": 0.09878832595880599,
"learning_rate": 0.0001606225410966638,
"loss": 0.3322,
"step": 166
},
{
"epoch": 1.0986842105263157,
"grad_norm": 0.08804802839862416,
"learning_rate": 0.0001600113805237737,
"loss": 0.3265,
"step": 167
},
{
"epoch": 1.1052631578947367,
"grad_norm": 0.08862630237882663,
"learning_rate": 0.0001593966965381088,
"loss": 0.3273,
"step": 168
},
{
"epoch": 1.111842105263158,
"grad_norm": 0.09796877182879883,
"learning_rate": 0.00015877852522924732,
"loss": 0.35,
"step": 169
},
{
"epoch": 1.118421052631579,
"grad_norm": 0.08598713651781048,
"learning_rate": 0.00015815690289151658,
"loss": 0.3297,
"step": 170
},
{
"epoch": 1.125,
"grad_norm": 0.08765944123128291,
"learning_rate": 0.00015753186602186209,
"loss": 0.3214,
"step": 171
},
{
"epoch": 1.131578947368421,
"grad_norm": 0.09209895613253609,
"learning_rate": 0.00015690345131770474,
"loss": 0.3315,
"step": 172
},
{
"epoch": 1.138157894736842,
"grad_norm": 0.09108557919395985,
"learning_rate": 0.0001562716956747863,
"loss": 0.3367,
"step": 173
},
{
"epoch": 1.1447368421052633,
"grad_norm": 0.08895539817943555,
"learning_rate": 0.00015563663618500302,
"loss": 0.3341,
"step": 174
},
{
"epoch": 1.1513157894736843,
"grad_norm": 0.08973951614406138,
"learning_rate": 0.00015499831013422803,
"loss": 0.3339,
"step": 175
},
{
"epoch": 1.1578947368421053,
"grad_norm": 0.08620142874315022,
"learning_rate": 0.00015435675500012212,
"loss": 0.335,
"step": 176
},
{
"epoch": 1.1644736842105263,
"grad_norm": 0.09126492238951715,
"learning_rate": 0.00015371200844993332,
"loss": 0.3312,
"step": 177
},
{
"epoch": 1.1710526315789473,
"grad_norm": 0.09587526018453588,
"learning_rate": 0.00015306410833828535,
"loss": 0.3294,
"step": 178
},
{
"epoch": 1.1776315789473684,
"grad_norm": 0.09087674252888882,
"learning_rate": 0.00015241309270495524,
"loss": 0.3372,
"step": 179
},
{
"epoch": 1.1842105263157894,
"grad_norm": 0.0883641678692255,
"learning_rate": 0.00015175899977263964,
"loss": 0.3386,
"step": 180
},
{
"epoch": 1.1907894736842106,
"grad_norm": 0.09379689103796256,
"learning_rate": 0.00015110186794471103,
"loss": 0.3377,
"step": 181
},
{
"epoch": 1.1973684210526316,
"grad_norm": 0.09279921976435307,
"learning_rate": 0.00015044173580296266,
"loss": 0.3331,
"step": 182
},
{
"epoch": 1.2039473684210527,
"grad_norm": 0.09523467660834971,
"learning_rate": 0.0001497786421053434,
"loss": 0.3468,
"step": 183
},
{
"epoch": 1.2105263157894737,
"grad_norm": 0.09692763912842987,
"learning_rate": 0.00014911262578368232,
"loss": 0.3254,
"step": 184
},
{
"epoch": 1.2171052631578947,
"grad_norm": 0.0905572094984609,
"learning_rate": 0.0001484437259414027,
"loss": 0.3428,
"step": 185
},
{
"epoch": 1.2236842105263157,
"grad_norm": 0.0895303347607193,
"learning_rate": 0.0001477719818512263,
"loss": 0.3343,
"step": 186
},
{
"epoch": 1.2302631578947367,
"grad_norm": 0.09084314989641666,
"learning_rate": 0.00014709743295286752,
"loss": 0.3428,
"step": 187
},
{
"epoch": 1.236842105263158,
"grad_norm": 0.09188016999870963,
"learning_rate": 0.0001464201188507178,
"loss": 0.3392,
"step": 188
},
{
"epoch": 1.243421052631579,
"grad_norm": 0.09034095867956521,
"learning_rate": 0.00014574007931152037,
"loss": 0.3305,
"step": 189
},
{
"epoch": 1.25,
"grad_norm": 0.09422416865836616,
"learning_rate": 0.00014505735426203543,
"loss": 0.3519,
"step": 190
},
{
"epoch": 1.256578947368421,
"grad_norm": 0.09146229714998408,
"learning_rate": 0.00014437198378669598,
"loss": 0.3246,
"step": 191
},
{
"epoch": 1.263157894736842,
"grad_norm": 0.09150826987109315,
"learning_rate": 0.00014368400812525434,
"loss": 0.3385,
"step": 192
},
{
"epoch": 1.2697368421052633,
"grad_norm": 0.09170822182104935,
"learning_rate": 0.00014299346767041954,
"loss": 0.3202,
"step": 193
},
{
"epoch": 1.2763157894736843,
"grad_norm": 0.08611790992292927,
"learning_rate": 0.0001423004029654859,
"loss": 0.3309,
"step": 194
},
{
"epoch": 1.2828947368421053,
"grad_norm": 0.09296272729378167,
"learning_rate": 0.00014160485470195245,
"loss": 0.3334,
"step": 195
},
{
"epoch": 1.2894736842105263,
"grad_norm": 0.09167809377202646,
"learning_rate": 0.00014090686371713402,
"loss": 0.3311,
"step": 196
},
{
"epoch": 1.2960526315789473,
"grad_norm": 0.08975846040243303,
"learning_rate": 0.0001402064709917634,
"loss": 0.3435,
"step": 197
},
{
"epoch": 1.3026315789473684,
"grad_norm": 0.09083148769815752,
"learning_rate": 0.00013950371764758542,
"loss": 0.3252,
"step": 198
},
{
"epoch": 1.3092105263157894,
"grad_norm": 0.09070017702344292,
"learning_rate": 0.0001387986449449425,
"loss": 0.3417,
"step": 199
},
{
"epoch": 1.3157894736842106,
"grad_norm": 0.09406507310996726,
"learning_rate": 0.00013809129428035227,
"loss": 0.3474,
"step": 200
},
{
"epoch": 1.3223684210526316,
"grad_norm": 0.08965350812782857,
"learning_rate": 0.00013738170718407687,
"loss": 0.3309,
"step": 201
},
{
"epoch": 1.3289473684210527,
"grad_norm": 0.08959778822119953,
"learning_rate": 0.00013666992531768482,
"loss": 0.3402,
"step": 202
},
{
"epoch": 1.3355263157894737,
"grad_norm": 0.092084592265277,
"learning_rate": 0.0001359559904716048,
"loss": 0.3375,
"step": 203
},
{
"epoch": 1.3421052631578947,
"grad_norm": 0.09457199959869429,
"learning_rate": 0.0001352399445626722,
"loss": 0.3389,
"step": 204
},
{
"epoch": 1.3486842105263157,
"grad_norm": 0.09197177008454786,
"learning_rate": 0.00013452182963166794,
"loss": 0.3374,
"step": 205
},
{
"epoch": 1.3552631578947367,
"grad_norm": 0.09399945833367072,
"learning_rate": 0.00013380168784085027,
"loss": 0.3393,
"step": 206
},
{
"epoch": 1.361842105263158,
"grad_norm": 0.08785100494819592,
"learning_rate": 0.00013307956147147923,
"loss": 0.339,
"step": 207
},
{
"epoch": 1.368421052631579,
"grad_norm": 0.09128615601816815,
"learning_rate": 0.00013235549292133425,
"loss": 0.3319,
"step": 208
},
{
"epoch": 1.375,
"grad_norm": 0.08674366087344106,
"learning_rate": 0.00013162952470222486,
"loss": 0.3262,
"step": 209
},
{
"epoch": 1.381578947368421,
"grad_norm": 0.09650968662374713,
"learning_rate": 0.00013090169943749476,
"loss": 0.3218,
"step": 210
},
{
"epoch": 1.388157894736842,
"grad_norm": 0.09647571976953839,
"learning_rate": 0.00013017205985951926,
"loss": 0.3297,
"step": 211
},
{
"epoch": 1.3947368421052633,
"grad_norm": 0.08840745611440821,
"learning_rate": 0.00012944064880719632,
"loss": 0.3183,
"step": 212
},
{
"epoch": 1.4013157894736843,
"grad_norm": 0.09429827798710846,
"learning_rate": 0.0001287075092234316,
"loss": 0.3441,
"step": 213
},
{
"epoch": 1.4078947368421053,
"grad_norm": 0.09400151382755063,
"learning_rate": 0.00012797268415261682,
"loss": 0.3374,
"step": 214
},
{
"epoch": 1.4144736842105263,
"grad_norm": 0.094388991307675,
"learning_rate": 0.00012723621673810275,
"loss": 0.3425,
"step": 215
},
{
"epoch": 1.4210526315789473,
"grad_norm": 0.09040719138127835,
"learning_rate": 0.0001264981502196662,
"loss": 0.3466,
"step": 216
},
{
"epoch": 1.4276315789473684,
"grad_norm": 0.08410156088705376,
"learning_rate": 0.0001257585279309711,
"loss": 0.3293,
"step": 217
},
{
"epoch": 1.4342105263157894,
"grad_norm": 0.09012762447734174,
"learning_rate": 0.00012501739329702453,
"loss": 0.3241,
"step": 218
},
{
"epoch": 1.4407894736842106,
"grad_norm": 0.08981619956473173,
"learning_rate": 0.00012427478983162695,
"loss": 0.3309,
"step": 219
},
{
"epoch": 1.4473684210526316,
"grad_norm": 0.09690483937945461,
"learning_rate": 0.0001235307611348174,
"loss": 0.3353,
"step": 220
},
{
"epoch": 1.4539473684210527,
"grad_norm": 0.09199281173602707,
"learning_rate": 0.00012278535089031378,
"loss": 0.3478,
"step": 221
},
{
"epoch": 1.4605263157894737,
"grad_norm": 0.09383616607484997,
"learning_rate": 0.0001220386028629479,
"loss": 0.3226,
"step": 222
},
{
"epoch": 1.4671052631578947,
"grad_norm": 0.0888565150769418,
"learning_rate": 0.00012129056089609609,
"loss": 0.3368,
"step": 223
},
{
"epoch": 1.4736842105263157,
"grad_norm": 0.08846739233573914,
"learning_rate": 0.00012054126890910498,
"loss": 0.3363,
"step": 224
},
{
"epoch": 1.4802631578947367,
"grad_norm": 0.08828966342515125,
"learning_rate": 0.00011979077089471287,
"loss": 0.3426,
"step": 225
},
{
"epoch": 1.486842105263158,
"grad_norm": 0.09066446045227519,
"learning_rate": 0.00011903911091646684,
"loss": 0.3305,
"step": 226
},
{
"epoch": 1.493421052631579,
"grad_norm": 0.0885766678361643,
"learning_rate": 0.00011828633310613568,
"loss": 0.3359,
"step": 227
},
{
"epoch": 1.5,
"grad_norm": 0.08447486357210816,
"learning_rate": 0.00011753248166111878,
"loss": 0.342,
"step": 228
},
{
"epoch": 1.506578947368421,
"grad_norm": 0.08995420125881844,
"learning_rate": 0.00011677760084185123,
"loss": 0.3239,
"step": 229
},
{
"epoch": 1.513157894736842,
"grad_norm": 0.08976025618793185,
"learning_rate": 0.0001160217349692051,
"loss": 0.3234,
"step": 230
},
{
"epoch": 1.5197368421052633,
"grad_norm": 0.08403343805686422,
"learning_rate": 0.00011526492842188745,
"loss": 0.3241,
"step": 231
},
{
"epoch": 1.526315789473684,
"grad_norm": 0.08731505199027503,
"learning_rate": 0.00011450722563383449,
"loss": 0.323,
"step": 232
},
{
"epoch": 1.5328947368421053,
"grad_norm": 0.08919530304419715,
"learning_rate": 0.00011374867109160294,
"loss": 0.335,
"step": 233
},
{
"epoch": 1.5394736842105263,
"grad_norm": 0.08543608378136278,
"learning_rate": 0.00011298930933175804,
"loss": 0.3257,
"step": 234
},
{
"epoch": 1.5460526315789473,
"grad_norm": 0.0843872841143867,
"learning_rate": 0.00011222918493825876,
"loss": 0.322,
"step": 235
},
{
"epoch": 1.5526315789473686,
"grad_norm": 0.09015699624357591,
"learning_rate": 0.00011146834253984006,
"loss": 0.3393,
"step": 236
},
{
"epoch": 1.5592105263157894,
"grad_norm": 0.08527720509613376,
"learning_rate": 0.00011070682680739275,
"loss": 0.3228,
"step": 237
},
{
"epoch": 1.5657894736842106,
"grad_norm": 0.08882512008213382,
"learning_rate": 0.00010994468245134071,
"loss": 0.3307,
"step": 238
},
{
"epoch": 1.5723684210526314,
"grad_norm": 0.0873699179145412,
"learning_rate": 0.00010918195421901583,
"loss": 0.3315,
"step": 239
},
{
"epoch": 1.5789473684210527,
"grad_norm": 0.0856032512902339,
"learning_rate": 0.00010841868689203071,
"loss": 0.3249,
"step": 240
},
{
"epoch": 1.5855263157894737,
"grad_norm": 0.08508957717963754,
"learning_rate": 0.0001076549252836496,
"loss": 0.3198,
"step": 241
},
{
"epoch": 1.5921052631578947,
"grad_norm": 0.08646553066877591,
"learning_rate": 0.00010689071423615709,
"loss": 0.3201,
"step": 242
},
{
"epoch": 1.598684210526316,
"grad_norm": 0.08651961917444705,
"learning_rate": 0.00010612609861822556,
"loss": 0.3273,
"step": 243
},
{
"epoch": 1.6052631578947367,
"grad_norm": 0.08675562470697942,
"learning_rate": 0.00010536112332228057,
"loss": 0.3344,
"step": 244
},
{
"epoch": 1.611842105263158,
"grad_norm": 0.09458426431125393,
"learning_rate": 0.00010459583326186533,
"loss": 0.3411,
"step": 245
},
{
"epoch": 1.618421052631579,
"grad_norm": 0.09120208666424241,
"learning_rate": 0.00010383027336900355,
"loss": 0.3359,
"step": 246
},
{
"epoch": 1.625,
"grad_norm": 0.09024986688209398,
"learning_rate": 0.00010306448859156155,
"loss": 0.3244,
"step": 247
},
{
"epoch": 1.631578947368421,
"grad_norm": 0.08766784765589664,
"learning_rate": 0.00010229852389060905,
"loss": 0.3165,
"step": 248
},
{
"epoch": 1.638157894736842,
"grad_norm": 0.08799575986715236,
"learning_rate": 0.00010153242423777965,
"loss": 0.3274,
"step": 249
},
{
"epoch": 1.6447368421052633,
"grad_norm": 0.0920628008990826,
"learning_rate": 0.00010076623461263018,
"loss": 0.3228,
"step": 250
},
{
"epoch": 1.651315789473684,
"grad_norm": 0.09132220875461658,
"learning_rate": 0.0001,
"loss": 0.3291,
"step": 251
},
{
"epoch": 1.6578947368421053,
"grad_norm": 0.0861257482862008,
"learning_rate": 9.923376538736985e-05,
"loss": 0.3257,
"step": 252
},
{
"epoch": 1.6644736842105263,
"grad_norm": 0.08970245386690526,
"learning_rate": 9.846757576222038e-05,
"loss": 0.333,
"step": 253
},
{
"epoch": 1.6710526315789473,
"grad_norm": 0.08745399881851679,
"learning_rate": 9.770147610939097e-05,
"loss": 0.3243,
"step": 254
},
{
"epoch": 1.6776315789473686,
"grad_norm": 0.0942879278062166,
"learning_rate": 9.693551140843847e-05,
"loss": 0.3305,
"step": 255
},
{
"epoch": 1.6842105263157894,
"grad_norm": 0.08491033600054997,
"learning_rate": 9.616972663099647e-05,
"loss": 0.3219,
"step": 256
},
{
"epoch": 1.6907894736842106,
"grad_norm": 0.0824170398183598,
"learning_rate": 9.54041667381347e-05,
"loss": 0.3067,
"step": 257
},
{
"epoch": 1.6973684210526314,
"grad_norm": 0.08689158883776966,
"learning_rate": 9.463887667771946e-05,
"loss": 0.3243,
"step": 258
},
{
"epoch": 1.7039473684210527,
"grad_norm": 0.09175347122606316,
"learning_rate": 9.387390138177448e-05,
"loss": 0.3315,
"step": 259
},
{
"epoch": 1.7105263157894737,
"grad_norm": 0.08682953977254314,
"learning_rate": 9.310928576384293e-05,
"loss": 0.3183,
"step": 260
},
{
"epoch": 1.7171052631578947,
"grad_norm": 0.0844793823883941,
"learning_rate": 9.234507471635043e-05,
"loss": 0.3262,
"step": 261
},
{
"epoch": 1.723684210526316,
"grad_norm": 0.08904925607635152,
"learning_rate": 9.15813131079693e-05,
"loss": 0.3318,
"step": 262
},
{
"epoch": 1.7302631578947367,
"grad_norm": 0.08897767966029022,
"learning_rate": 9.08180457809842e-05,
"loss": 0.3338,
"step": 263
},
{
"epoch": 1.736842105263158,
"grad_norm": 0.08418525864024147,
"learning_rate": 9.005531754865929e-05,
"loss": 0.3224,
"step": 264
},
{
"epoch": 1.743421052631579,
"grad_norm": 0.08369708951027545,
"learning_rate": 8.929317319260726e-05,
"loss": 0.3344,
"step": 265
},
{
"epoch": 1.75,
"grad_norm": 0.08344688234687514,
"learning_rate": 8.853165746015997e-05,
"loss": 0.3186,
"step": 266
},
{
"epoch": 1.756578947368421,
"grad_norm": 0.08882311328649488,
"learning_rate": 8.777081506174127e-05,
"loss": 0.3189,
"step": 267
},
{
"epoch": 1.763157894736842,
"grad_norm": 0.08911077215219455,
"learning_rate": 8.701069066824194e-05,
"loss": 0.347,
"step": 268
},
{
"epoch": 1.7697368421052633,
"grad_norm": 0.08845395362382234,
"learning_rate": 8.625132890839706e-05,
"loss": 0.3401,
"step": 269
},
{
"epoch": 1.776315789473684,
"grad_norm": 0.0890814727471271,
"learning_rate": 8.549277436616551e-05,
"loss": 0.3357,
"step": 270
},
{
"epoch": 1.7828947368421053,
"grad_norm": 0.08688996127676986,
"learning_rate": 8.473507157811254e-05,
"loss": 0.3255,
"step": 271
},
{
"epoch": 1.7894736842105263,
"grad_norm": 0.08839344810308368,
"learning_rate": 8.397826503079489e-05,
"loss": 0.3485,
"step": 272
},
{
"epoch": 1.7960526315789473,
"grad_norm": 0.08616082603732687,
"learning_rate": 8.32223991581488e-05,
"loss": 0.3333,
"step": 273
},
{
"epoch": 1.8026315789473686,
"grad_norm": 0.0873042236962117,
"learning_rate": 8.246751833888122e-05,
"loss": 0.3293,
"step": 274
},
{
"epoch": 1.8092105263157894,
"grad_norm": 0.08637643463516168,
"learning_rate": 8.171366689386432e-05,
"loss": 0.3366,
"step": 275
},
{
"epoch": 1.8157894736842106,
"grad_norm": 0.08702343026083556,
"learning_rate": 8.096088908353315e-05,
"loss": 0.317,
"step": 276
},
{
"epoch": 1.8223684210526314,
"grad_norm": 0.08671977794554171,
"learning_rate": 8.020922910528717e-05,
"loss": 0.3243,
"step": 277
},
{
"epoch": 1.8289473684210527,
"grad_norm": 0.08800970811383262,
"learning_rate": 7.945873109089503e-05,
"loss": 0.328,
"step": 278
},
{
"epoch": 1.8355263157894737,
"grad_norm": 0.08715460079863534,
"learning_rate": 7.870943910390392e-05,
"loss": 0.3393,
"step": 279
},
{
"epoch": 1.8421052631578947,
"grad_norm": 0.08522036841567814,
"learning_rate": 7.796139713705213e-05,
"loss": 0.3264,
"step": 280
},
{
"epoch": 1.848684210526316,
"grad_norm": 0.08867443005533418,
"learning_rate": 7.721464910968627e-05,
"loss": 0.3245,
"step": 281
},
{
"epoch": 1.8552631578947367,
"grad_norm": 0.085291540055969,
"learning_rate": 7.646923886518264e-05,
"loss": 0.3193,
"step": 282
},
{
"epoch": 1.861842105263158,
"grad_norm": 0.0865604586807815,
"learning_rate": 7.57252101683731e-05,
"loss": 0.3265,
"step": 283
},
{
"epoch": 1.868421052631579,
"grad_norm": 0.08796722757240084,
"learning_rate": 7.49826067029755e-05,
"loss": 0.3315,
"step": 284
},
{
"epoch": 1.875,
"grad_norm": 0.08552421405911567,
"learning_rate": 7.424147206902891e-05,
"loss": 0.3193,
"step": 285
},
{
"epoch": 1.881578947368421,
"grad_norm": 0.08359126397078406,
"learning_rate": 7.350184978033386e-05,
"loss": 0.3283,
"step": 286
},
{
"epoch": 1.888157894736842,
"grad_norm": 0.09295820408600056,
"learning_rate": 7.276378326189728e-05,
"loss": 0.34,
"step": 287
},
{
"epoch": 1.8947368421052633,
"grad_norm": 0.08607086022906524,
"learning_rate": 7.202731584738322e-05,
"loss": 0.3224,
"step": 288
},
{
"epoch": 1.901315789473684,
"grad_norm": 0.08511420128792765,
"learning_rate": 7.129249077656843e-05,
"loss": 0.3297,
"step": 289
},
{
"epoch": 1.9078947368421053,
"grad_norm": 0.08811788288471267,
"learning_rate": 7.055935119280369e-05,
"loss": 0.327,
"step": 290
},
{
"epoch": 1.9144736842105263,
"grad_norm": 0.08477145974314237,
"learning_rate": 6.982794014048077e-05,
"loss": 0.3161,
"step": 291
},
{
"epoch": 1.9210526315789473,
"grad_norm": 0.08372118058279919,
"learning_rate": 6.909830056250527e-05,
"loss": 0.3306,
"step": 292
},
{
"epoch": 1.9276315789473686,
"grad_norm": 0.08686106996534168,
"learning_rate": 6.837047529777516e-05,
"loss": 0.32,
"step": 293
},
{
"epoch": 1.9342105263157894,
"grad_norm": 0.08515118561718467,
"learning_rate": 6.764450707866576e-05,
"loss": 0.3252,
"step": 294
},
{
"epoch": 1.9407894736842106,
"grad_norm": 0.09265542100862634,
"learning_rate": 6.692043852852079e-05,
"loss": 0.3325,
"step": 295
},
{
"epoch": 1.9473684210526314,
"grad_norm": 0.09595543410539613,
"learning_rate": 6.619831215914974e-05,
"loss": 0.3333,
"step": 296
},
{
"epoch": 1.9539473684210527,
"grad_norm": 0.08927193916247748,
"learning_rate": 6.547817036833208e-05,
"loss": 0.34,
"step": 297
},
{
"epoch": 1.9605263157894737,
"grad_norm": 0.08244175050863302,
"learning_rate": 6.476005543732782e-05,
"loss": 0.3206,
"step": 298
},
{
"epoch": 1.9671052631578947,
"grad_norm": 0.08541613997019516,
"learning_rate": 6.404400952839522e-05,
"loss": 0.3095,
"step": 299
},
{
"epoch": 1.973684210526316,
"grad_norm": 0.08585952116970799,
"learning_rate": 6.333007468231521e-05,
"loss": 0.3143,
"step": 300
},
{
"epoch": 1.9802631578947367,
"grad_norm": 0.09128909476408122,
"learning_rate": 6.261829281592314e-05,
"loss": 0.3345,
"step": 301
},
{
"epoch": 1.986842105263158,
"grad_norm": 0.09002004932775928,
"learning_rate": 6.190870571964774e-05,
"loss": 0.3257,
"step": 302
},
{
"epoch": 1.993421052631579,
"grad_norm": 0.08541319081987805,
"learning_rate": 6.120135505505748e-05,
"loss": 0.3269,
"step": 303
},
{
"epoch": 2.0,
"grad_norm": 0.08848658071924796,
"learning_rate": 6.049628235241459e-05,
"loss": 0.3233,
"step": 304
},
{
"epoch": 2.0,
"eval_loss": 0.33567601442337036,
"eval_runtime": 160.3905,
"eval_samples_per_second": 31.829,
"eval_steps_per_second": 0.998,
"step": 304
},
{
"epoch": 2.0065789473684212,
"grad_norm": 0.08452043290010046,
"learning_rate": 5.9793529008236625e-05,
"loss": 0.308,
"step": 305
},
{
"epoch": 2.013157894736842,
"grad_norm": 0.084981364890352,
"learning_rate": 5.909313628286601e-05,
"loss": 0.3138,
"step": 306
},
{
"epoch": 2.0197368421052633,
"grad_norm": 0.08373368193427122,
"learning_rate": 5.839514529804757e-05,
"loss": 0.3051,
"step": 307
},
{
"epoch": 2.026315789473684,
"grad_norm": 0.08658031473872292,
"learning_rate": 5.769959703451413e-05,
"loss": 0.3149,
"step": 308
},
{
"epoch": 2.0328947368421053,
"grad_norm": 0.0872119478961657,
"learning_rate": 5.7006532329580464e-05,
"loss": 0.2952,
"step": 309
},
{
"epoch": 2.039473684210526,
"grad_norm": 0.08985312096854285,
"learning_rate": 5.63159918747457e-05,
"loss": 0.31,
"step": 310
},
{
"epoch": 2.0460526315789473,
"grad_norm": 0.08981442223402118,
"learning_rate": 5.5628016213304025e-05,
"loss": 0.2987,
"step": 311
},
{
"epoch": 2.0526315789473686,
"grad_norm": 0.08888884200687737,
"learning_rate": 5.49426457379646e-05,
"loss": 0.314,
"step": 312
},
{
"epoch": 2.0592105263157894,
"grad_norm": 0.09263884628501892,
"learning_rate": 5.425992068847965e-05,
"loss": 0.3105,
"step": 313
},
{
"epoch": 2.0657894736842106,
"grad_norm": 0.09695553478937406,
"learning_rate": 5.3579881149282204e-05,
"loss": 0.2961,
"step": 314
},
{
"epoch": 2.0723684210526314,
"grad_norm": 0.09301145489202128,
"learning_rate": 5.29025670471325e-05,
"loss": 0.3121,
"step": 315
},
{
"epoch": 2.0789473684210527,
"grad_norm": 0.09056322468061577,
"learning_rate": 5.222801814877369e-05,
"loss": 0.3084,
"step": 316
},
{
"epoch": 2.085526315789474,
"grad_norm": 0.09036800264131283,
"learning_rate": 5.155627405859731e-05,
"loss": 0.3068,
"step": 317
},
{
"epoch": 2.0921052631578947,
"grad_norm": 0.09181257128694434,
"learning_rate": 5.088737421631767e-05,
"loss": 0.3113,
"step": 318
},
{
"epoch": 2.098684210526316,
"grad_norm": 0.09316264301444047,
"learning_rate": 5.0221357894656605e-05,
"loss": 0.3067,
"step": 319
},
{
"epoch": 2.1052631578947367,
"grad_norm": 0.09246714604570534,
"learning_rate": 4.955826419703735e-05,
"loss": 0.3035,
"step": 320
},
{
"epoch": 2.111842105263158,
"grad_norm": 0.0933519192978327,
"learning_rate": 4.889813205528895e-05,
"loss": 0.3079,
"step": 321
},
{
"epoch": 2.1184210526315788,
"grad_norm": 0.08840630002939555,
"learning_rate": 4.8241000227360354e-05,
"loss": 0.3082,
"step": 322
},
{
"epoch": 2.125,
"grad_norm": 0.090468509851056,
"learning_rate": 4.758690729504478e-05,
"loss": 0.3013,
"step": 323
},
{
"epoch": 2.1315789473684212,
"grad_norm": 0.09269514982319728,
"learning_rate": 4.693589166171466e-05,
"loss": 0.3045,
"step": 324
},
{
"epoch": 2.138157894736842,
"grad_norm": 0.0944000183696063,
"learning_rate": 4.628799155006669e-05,
"loss": 0.3104,
"step": 325
},
{
"epoch": 2.1447368421052633,
"grad_norm": 0.09277066986876077,
"learning_rate": 4.56432449998779e-05,
"loss": 0.3061,
"step": 326
},
{
"epoch": 2.151315789473684,
"grad_norm": 0.09140609684482606,
"learning_rate": 4.500168986577197e-05,
"loss": 0.2987,
"step": 327
},
{
"epoch": 2.1578947368421053,
"grad_norm": 0.09135597016385583,
"learning_rate": 4.4363363814997004e-05,
"loss": 0.3138,
"step": 328
},
{
"epoch": 2.1644736842105265,
"grad_norm": 0.09228058446081117,
"learning_rate": 4.3728304325213764e-05,
"loss": 0.3093,
"step": 329
},
{
"epoch": 2.1710526315789473,
"grad_norm": 0.09444538193500554,
"learning_rate": 4.30965486822953e-05,
"loss": 0.3121,
"step": 330
},
{
"epoch": 2.1776315789473686,
"grad_norm": 0.09278379473856056,
"learning_rate": 4.2468133978137945e-05,
"loss": 0.3085,
"step": 331
},
{
"epoch": 2.1842105263157894,
"grad_norm": 0.09002068234283259,
"learning_rate": 4.1843097108483465e-05,
"loss": 0.3115,
"step": 332
},
{
"epoch": 2.1907894736842106,
"grad_norm": 0.09182734300425907,
"learning_rate": 4.12214747707527e-05,
"loss": 0.3041,
"step": 333
},
{
"epoch": 2.1973684210526314,
"grad_norm": 0.09480621705211878,
"learning_rate": 4.0603303461891253e-05,
"loss": 0.3119,
"step": 334
},
{
"epoch": 2.2039473684210527,
"grad_norm": 0.09462062961478308,
"learning_rate": 3.9988619476226355e-05,
"loss": 0.3097,
"step": 335
},
{
"epoch": 2.2105263157894735,
"grad_norm": 0.09047742114008103,
"learning_rate": 3.937745890333623e-05,
"loss": 0.2977,
"step": 336
},
{
"epoch": 2.2171052631578947,
"grad_norm": 0.09210517994431433,
"learning_rate": 3.8769857625930906e-05,
"loss": 0.3045,
"step": 337
},
{
"epoch": 2.223684210526316,
"grad_norm": 0.09051025798694169,
"learning_rate": 3.816585131774571e-05,
"loss": 0.3042,
"step": 338
},
{
"epoch": 2.2302631578947367,
"grad_norm": 0.0944091820926416,
"learning_rate": 3.756547544144664e-05,
"loss": 0.3146,
"step": 339
},
{
"epoch": 2.236842105263158,
"grad_norm": 0.09225272107547029,
"learning_rate": 3.69687652465482e-05,
"loss": 0.2979,
"step": 340
},
{
"epoch": 2.2434210526315788,
"grad_norm": 0.09466054564763007,
"learning_rate": 3.637575576734404e-05,
"loss": 0.3268,
"step": 341
},
{
"epoch": 2.25,
"grad_norm": 0.09415533738249948,
"learning_rate": 3.578648182084975e-05,
"loss": 0.3138,
"step": 342
},
{
"epoch": 2.2565789473684212,
"grad_norm": 0.09287732881109488,
"learning_rate": 3.5200978004758886e-05,
"loss": 0.3034,
"step": 343
},
{
"epoch": 2.263157894736842,
"grad_norm": 0.09052642673165286,
"learning_rate": 3.46192786954115e-05,
"loss": 0.3136,
"step": 344
},
{
"epoch": 2.2697368421052633,
"grad_norm": 0.0941549755498813,
"learning_rate": 3.4041418045775895e-05,
"loss": 0.3102,
"step": 345
},
{
"epoch": 2.276315789473684,
"grad_norm": 0.09081461979766672,
"learning_rate": 3.346742998344348e-05,
"loss": 0.3156,
"step": 346
},
{
"epoch": 2.2828947368421053,
"grad_norm": 0.09188038212201086,
"learning_rate": 3.2897348208636614e-05,
"loss": 0.2981,
"step": 347
},
{
"epoch": 2.2894736842105265,
"grad_norm": 0.09139958010871077,
"learning_rate": 3.2331206192230214e-05,
"loss": 0.3016,
"step": 348
},
{
"epoch": 2.2960526315789473,
"grad_norm": 0.09340863611831572,
"learning_rate": 3.1769037173786375e-05,
"loss": 0.2942,
"step": 349
},
{
"epoch": 2.3026315789473686,
"grad_norm": 0.09731529194622152,
"learning_rate": 3.121087415960304e-05,
"loss": 0.32,
"step": 350
},
{
"epoch": 2.3092105263157894,
"grad_norm": 0.09400113350744722,
"learning_rate": 3.065674992077584e-05,
"loss": 0.3121,
"step": 351
},
{
"epoch": 2.3157894736842106,
"grad_norm": 0.09370429535454715,
"learning_rate": 3.0106696991274276e-05,
"loss": 0.2998,
"step": 352
},
{
"epoch": 2.3223684210526314,
"grad_norm": 0.0931264318395488,
"learning_rate": 2.9560747666031475e-05,
"loss": 0.3045,
"step": 353
},
{
"epoch": 2.3289473684210527,
"grad_norm": 0.09457813125113766,
"learning_rate": 2.901893399904797e-05,
"loss": 0.3102,
"step": 354
},
{
"epoch": 2.3355263157894735,
"grad_norm": 0.09787356394017462,
"learning_rate": 2.848128780150995e-05,
"loss": 0.3124,
"step": 355
},
{
"epoch": 2.3421052631578947,
"grad_norm": 0.09294514140726695,
"learning_rate": 2.794784063992131e-05,
"loss": 0.3029,
"step": 356
},
{
"epoch": 2.348684210526316,
"grad_norm": 0.09472220447378009,
"learning_rate": 2.74186238342505e-05,
"loss": 0.3026,
"step": 357
},
{
"epoch": 2.3552631578947367,
"grad_norm": 0.09382282625627145,
"learning_rate": 2.6893668456091625e-05,
"loss": 0.3067,
"step": 358
},
{
"epoch": 2.361842105263158,
"grad_norm": 0.09371704952441416,
"learning_rate": 2.6373005326839972e-05,
"loss": 0.3065,
"step": 359
},
{
"epoch": 2.3684210526315788,
"grad_norm": 0.098774530422876,
"learning_rate": 2.5856665015882685e-05,
"loss": 0.3098,
"step": 360
},
{
"epoch": 2.375,
"grad_norm": 0.09126307975009218,
"learning_rate": 2.5344677838803733e-05,
"loss": 0.3044,
"step": 361
},
{
"epoch": 2.3815789473684212,
"grad_norm": 0.09144641630016763,
"learning_rate": 2.4837073855604188e-05,
"loss": 0.2908,
"step": 362
},
{
"epoch": 2.388157894736842,
"grad_norm": 0.09468863277888301,
"learning_rate": 2.433388286893714e-05,
"loss": 0.3101,
"step": 363
},
{
"epoch": 2.3947368421052633,
"grad_norm": 0.09354400587822048,
"learning_rate": 2.383513442235812e-05,
"loss": 0.3136,
"step": 364
},
{
"epoch": 2.401315789473684,
"grad_norm": 0.09390078122691951,
"learning_rate": 2.334085779859041e-05,
"loss": 0.3079,
"step": 365
},
{
"epoch": 2.4078947368421053,
"grad_norm": 0.09248111646485846,
"learning_rate": 2.2851082017805703e-05,
"loss": 0.305,
"step": 366
},
{
"epoch": 2.4144736842105265,
"grad_norm": 0.09244167723770774,
"learning_rate": 2.236583583592049e-05,
"loss": 0.302,
"step": 367
},
{
"epoch": 2.4210526315789473,
"grad_norm": 0.09129005795575806,
"learning_rate": 2.1885147742907465e-05,
"loss": 0.3027,
"step": 368
},
{
"epoch": 2.4276315789473686,
"grad_norm": 0.09432198973734413,
"learning_rate": 2.1409045961123065e-05,
"loss": 0.3217,
"step": 369
},
{
"epoch": 2.4342105263157894,
"grad_norm": 0.09309602245093605,
"learning_rate": 2.0937558443650206e-05,
"loss": 0.3063,
"step": 370
},
{
"epoch": 2.4407894736842106,
"grad_norm": 0.09208894600051501,
"learning_rate": 2.047071287265735e-05,
"loss": 0.3052,
"step": 371
},
{
"epoch": 2.4473684210526314,
"grad_norm": 0.09266552082076217,
"learning_rate": 2.0008536657773046e-05,
"loss": 0.3089,
"step": 372
},
{
"epoch": 2.4539473684210527,
"grad_norm": 0.09140806866621644,
"learning_rate": 1.9551056934476654e-05,
"loss": 0.3013,
"step": 373
},
{
"epoch": 2.4605263157894735,
"grad_norm": 0.09231696012858444,
"learning_rate": 1.9098300562505266e-05,
"loss": 0.3086,
"step": 374
},
{
"epoch": 2.4671052631578947,
"grad_norm": 0.08976907732352662,
"learning_rate": 1.8650294124276556e-05,
"loss": 0.2908,
"step": 375
},
{
"epoch": 2.473684210526316,
"grad_norm": 0.09089868074014618,
"learning_rate": 1.8207063923328237e-05,
"loss": 0.2934,
"step": 376
},
{
"epoch": 2.4802631578947367,
"grad_norm": 0.09124223709758555,
"learning_rate": 1.7768635982773528e-05,
"loss": 0.3025,
"step": 377
},
{
"epoch": 2.486842105263158,
"grad_norm": 0.09141228045415595,
"learning_rate": 1.73350360437734e-05,
"loss": 0.2984,
"step": 378
},
{
"epoch": 2.4934210526315788,
"grad_norm": 0.09338433503873096,
"learning_rate": 1.6906289564025278e-05,
"loss": 0.3015,
"step": 379
},
{
"epoch": 2.5,
"grad_norm": 0.09381766067303669,
"learning_rate": 1.6482421716268215e-05,
"loss": 0.3034,
"step": 380
},
{
"epoch": 2.5065789473684212,
"grad_norm": 0.09841665883274005,
"learning_rate": 1.6063457386805004e-05,
"loss": 0.3097,
"step": 381
},
{
"epoch": 2.513157894736842,
"grad_norm": 0.09562609257672962,
"learning_rate": 1.564942117404119e-05,
"loss": 0.3194,
"step": 382
},
{
"epoch": 2.5197368421052633,
"grad_norm": 0.09236226141715091,
"learning_rate": 1.524033738704056e-05,
"loss": 0.3116,
"step": 383
},
{
"epoch": 2.526315789473684,
"grad_norm": 0.09295111594898633,
"learning_rate": 1.4836230044098166e-05,
"loss": 0.2926,
"step": 384
},
{
"epoch": 2.5328947368421053,
"grad_norm": 0.09050178834061187,
"learning_rate": 1.4437122871329955e-05,
"loss": 0.3048,
"step": 385
},
{
"epoch": 2.5394736842105265,
"grad_norm": 0.09354446958969953,
"learning_rate": 1.4043039301279903e-05,
"loss": 0.3094,
"step": 386
},
{
"epoch": 2.5460526315789473,
"grad_norm": 0.094528346545483,
"learning_rate": 1.3654002471544091e-05,
"loss": 0.2976,
"step": 387
},
{
"epoch": 2.5526315789473686,
"grad_norm": 0.09437075440256357,
"learning_rate": 1.3270035223412369e-05,
"loss": 0.303,
"step": 388
},
{
"epoch": 2.5592105263157894,
"grad_norm": 0.09299672016467998,
"learning_rate": 1.2891160100527222e-05,
"loss": 0.2991,
"step": 389
},
{
"epoch": 2.5657894736842106,
"grad_norm": 0.09295508003371206,
"learning_rate": 1.2517399347560167e-05,
"loss": 0.3111,
"step": 390
},
{
"epoch": 2.5723684210526314,
"grad_norm": 0.09416195404382027,
"learning_rate": 1.214877490890578e-05,
"loss": 0.3175,
"step": 391
},
{
"epoch": 2.5789473684210527,
"grad_norm": 0.09276583875255047,
"learning_rate": 1.1785308427393183e-05,
"loss": 0.299,
"step": 392
},
{
"epoch": 2.5855263157894735,
"grad_norm": 0.09118856826069724,
"learning_rate": 1.1427021243015467e-05,
"loss": 0.3029,
"step": 393
},
{
"epoch": 2.5921052631578947,
"grad_norm": 0.09153124872331597,
"learning_rate": 1.1073934391676667e-05,
"loss": 0.3053,
"step": 394
},
{
"epoch": 2.598684210526316,
"grad_norm": 0.0931741744031297,
"learning_rate": 1.0726068603956741e-05,
"loss": 0.2962,
"step": 395
},
{
"epoch": 2.6052631578947367,
"grad_norm": 0.0939187765998417,
"learning_rate": 1.0383444303894452e-05,
"loss": 0.3008,
"step": 396
},
{
"epoch": 2.611842105263158,
"grad_norm": 0.09383471721688587,
"learning_rate": 1.0046081607788127e-05,
"loss": 0.312,
"step": 397
},
{
"epoch": 2.6184210526315788,
"grad_norm": 0.09081224477516317,
"learning_rate": 9.714000323014705e-06,
"loss": 0.2947,
"step": 398
},
{
"epoch": 2.625,
"grad_norm": 0.09356165649233147,
"learning_rate": 9.3872199468667e-06,
"loss": 0.3016,
"step": 399
},
{
"epoch": 2.6315789473684212,
"grad_norm": 0.09192904823893439,
"learning_rate": 9.065759665407515e-06,
"loss": 0.2971,
"step": 400
},
{
"epoch": 2.638157894736842,
"grad_norm": 0.09313138523215327,
"learning_rate": 8.749638352345002e-06,
"loss": 0.3042,
"step": 401
},
{
"epoch": 2.6447368421052633,
"grad_norm": 0.0921977880339759,
"learning_rate": 8.43887456792326e-06,
"loss": 0.3016,
"step": 402
},
{
"epoch": 2.651315789473684,
"grad_norm": 0.09055704810960578,
"learning_rate": 8.133486557833015e-06,
"loss": 0.3003,
"step": 403
},
{
"epoch": 2.6578947368421053,
"grad_norm": 0.08952929321885197,
"learning_rate": 7.833492252140284e-06,
"loss": 0.2912,
"step": 404
},
{
"epoch": 2.6644736842105265,
"grad_norm": 0.0917144195389881,
"learning_rate": 7.538909264233751e-06,
"loss": 0.2908,
"step": 405
},
{
"epoch": 2.6710526315789473,
"grad_norm": 0.09112348602865261,
"learning_rate": 7.249754889790539e-06,
"loss": 0.2765,
"step": 406
},
{
"epoch": 2.6776315789473686,
"grad_norm": 0.09188227591595896,
"learning_rate": 6.966046105760826e-06,
"loss": 0.2998,
"step": 407
},
{
"epoch": 2.6842105263157894,
"grad_norm": 0.09193344799591113,
"learning_rate": 6.6877995693710785e-06,
"loss": 0.3073,
"step": 408
},
{
"epoch": 2.6907894736842106,
"grad_norm": 0.09125510240984186,
"learning_rate": 6.41503161714595e-06,
"loss": 0.2856,
"step": 409
},
{
"epoch": 2.6973684210526314,
"grad_norm": 0.09093722035003914,
"learning_rate": 6.147758263949321e-06,
"loss": 0.3029,
"step": 410
},
{
"epoch": 2.7039473684210527,
"grad_norm": 0.09720543266247605,
"learning_rate": 5.885995202043848e-06,
"loss": 0.3069,
"step": 411
},
{
"epoch": 2.7105263157894735,
"grad_norm": 0.09427472370082618,
"learning_rate": 5.629757800169733e-06,
"loss": 0.3084,
"step": 412
},
{
"epoch": 2.7171052631578947,
"grad_norm": 0.089990452112005,
"learning_rate": 5.379061102642358e-06,
"loss": 0.2999,
"step": 413
},
{
"epoch": 2.723684210526316,
"grad_norm": 0.09500774461582016,
"learning_rate": 5.1339198284689916e-06,
"loss": 0.3027,
"step": 414
},
{
"epoch": 2.7302631578947367,
"grad_norm": 0.09140486608400628,
"learning_rate": 4.8943483704846475e-06,
"loss": 0.2943,
"step": 415
},
{
"epoch": 2.736842105263158,
"grad_norm": 0.09254297353923395,
"learning_rate": 4.660360794506946e-06,
"loss": 0.2929,
"step": 416
},
{
"epoch": 2.7434210526315788,
"grad_norm": 0.09259358584474216,
"learning_rate": 4.431970838510391e-06,
"loss": 0.3057,
"step": 417
},
{
"epoch": 2.75,
"grad_norm": 0.0933006035630196,
"learning_rate": 4.209191911819688e-06,
"loss": 0.3066,
"step": 418
},
{
"epoch": 2.7565789473684212,
"grad_norm": 0.09175027584175248,
"learning_rate": 3.992037094322532e-06,
"loss": 0.2974,
"step": 419
},
{
"epoch": 2.763157894736842,
"grad_norm": 0.09612639615242942,
"learning_rate": 3.7805191357015657e-06,
"loss": 0.3152,
"step": 420
},
{
"epoch": 2.7697368421052633,
"grad_norm": 0.09267277826656926,
"learning_rate": 3.5746504546859018e-06,
"loss": 0.3122,
"step": 421
},
{
"epoch": 2.776315789473684,
"grad_norm": 0.09173196582903508,
"learning_rate": 3.3744431383219367e-06,
"loss": 0.3068,
"step": 422
},
{
"epoch": 2.7828947368421053,
"grad_norm": 0.09141455521971137,
"learning_rate": 3.1799089412636762e-06,
"loss": 0.3,
"step": 423
},
{
"epoch": 2.7894736842105265,
"grad_norm": 0.09007921765376342,
"learning_rate": 2.991059285082698e-06,
"loss": 0.3009,
"step": 424
},
{
"epoch": 2.7960526315789473,
"grad_norm": 0.0953184723452783,
"learning_rate": 2.8079052575973765e-06,
"loss": 0.3043,
"step": 425
},
{
"epoch": 2.8026315789473686,
"grad_norm": 0.09157757129707106,
"learning_rate": 2.6304576122221035e-06,
"loss": 0.3,
"step": 426
},
{
"epoch": 2.8092105263157894,
"grad_norm": 0.09169344011676386,
"learning_rate": 2.4587267673357394e-06,
"loss": 0.2917,
"step": 427
},
{
"epoch": 2.8157894736842106,
"grad_norm": 0.09140035394059028,
"learning_rate": 2.2927228056700777e-06,
"loss": 0.3086,
"step": 428
},
{
"epoch": 2.8223684210526314,
"grad_norm": 0.08884580894240417,
"learning_rate": 2.1324554737177648e-06,
"loss": 0.2974,
"step": 429
},
{
"epoch": 2.8289473684210527,
"grad_norm": 0.09227882207821884,
"learning_rate": 1.9779341811600794e-06,
"loss": 0.2993,
"step": 430
},
{
"epoch": 2.8355263157894735,
"grad_norm": 0.09230636915969623,
"learning_rate": 1.8291680003145073e-06,
"loss": 0.3014,
"step": 431
},
{
"epoch": 2.8421052631578947,
"grad_norm": 0.09126820223778219,
"learning_rate": 1.6861656656020464e-06,
"loss": 0.301,
"step": 432
},
{
"epoch": 2.848684210526316,
"grad_norm": 0.09346572537156347,
"learning_rate": 1.5489355730343934e-06,
"loss": 0.3047,
"step": 433
},
{
"epoch": 2.8552631578947367,
"grad_norm": 0.09159445394674477,
"learning_rate": 1.417485779720995e-06,
"loss": 0.3111,
"step": 434
},
{
"epoch": 2.861842105263158,
"grad_norm": 0.09263259410862734,
"learning_rate": 1.2918240033960027e-06,
"loss": 0.3064,
"step": 435
},
{
"epoch": 2.8684210526315788,
"grad_norm": 0.09195985713654573,
"learning_rate": 1.1719576219651585e-06,
"loss": 0.3085,
"step": 436
},
{
"epoch": 2.875,
"grad_norm": 0.09287155646470045,
"learning_rate": 1.057893673072563e-06,
"loss": 0.3123,
"step": 437
},
{
"epoch": 2.8815789473684212,
"grad_norm": 0.09261252125029774,
"learning_rate": 9.496388536875623e-07,
"loss": 0.3014,
"step": 438
},
{
"epoch": 2.888157894736842,
"grad_norm": 0.09510242096519511,
"learning_rate": 8.471995197114835e-07,
"loss": 0.3117,
"step": 439
},
{
"epoch": 2.8947368421052633,
"grad_norm": 0.09143211071538708,
"learning_rate": 7.505816856045012e-07,
"loss": 0.3021,
"step": 440
},
{
"epoch": 2.901315789473684,
"grad_norm": 0.0920917445117736,
"learning_rate": 6.597910240324967e-07,
"loss": 0.3049,
"step": 441
},
{
"epoch": 2.9078947368421053,
"grad_norm": 0.09323526996440369,
"learning_rate": 5.74832865534014e-07,
"loss": 0.3044,
"step": 442
},
{
"epoch": 2.9144736842105265,
"grad_norm": 0.08995868186598065,
"learning_rate": 4.957121982072765e-07,
"loss": 0.3074,
"step": 443
},
{
"epoch": 2.9210526315789473,
"grad_norm": 0.09113776795368703,
"learning_rate": 4.224336674173546e-07,
"loss": 0.3109,
"step": 444
},
{
"epoch": 2.9276315789473686,
"grad_norm": 0.09224937503409146,
"learning_rate": 3.5500157552337264e-07,
"loss": 0.3017,
"step": 445
},
{
"epoch": 2.9342105263157894,
"grad_norm": 0.09468784220174042,
"learning_rate": 2.934198816259559e-07,
"loss": 0.2992,
"step": 446
},
{
"epoch": 2.9407894736842106,
"grad_norm": 0.09123611642303985,
"learning_rate": 2.3769220133477156e-07,
"loss": 0.3068,
"step": 447
},
{
"epoch": 2.9473684210526314,
"grad_norm": 0.09217795191626804,
"learning_rate": 1.8782180655622095e-07,
"loss": 0.3032,
"step": 448
},
{
"epoch": 2.9539473684210527,
"grad_norm": 0.09439928587738844,
"learning_rate": 1.4381162530135993e-07,
"loss": 0.3149,
"step": 449
},
{
"epoch": 2.9605263157894735,
"grad_norm": 0.09155507167334377,
"learning_rate": 1.0566424151401411e-07,
"loss": 0.3064,
"step": 450
},
{
"epoch": 2.9671052631578947,
"grad_norm": 0.09353543798817081,
"learning_rate": 7.338189491900016e-08,
"loss": 0.3079,
"step": 451
},
{
"epoch": 2.973684210526316,
"grad_norm": 0.09071512231087696,
"learning_rate": 4.696648089068667e-08,
"loss": 0.3019,
"step": 452
},
{
"epoch": 2.9802631578947367,
"grad_norm": 0.09100478976490751,
"learning_rate": 2.641955034170529e-08,
"loss": 0.3038,
"step": 453
},
{
"epoch": 2.986842105263158,
"grad_norm": 0.09180608030721753,
"learning_rate": 1.1742309631845861e-08,
"loss": 0.305,
"step": 454
},
{
"epoch": 2.9934210526315788,
"grad_norm": 0.09166087142524931,
"learning_rate": 2.935620497301894e-09,
"loss": 0.2995,
"step": 455
},
{
"epoch": 3.0,
"grad_norm": 0.08984730795754493,
"learning_rate": 0.0,
"loss": 0.2989,
"step": 456
},
{
"epoch": 3.0,
"eval_loss": 0.3327427804470062,
"eval_runtime": 158.4336,
"eval_samples_per_second": 32.222,
"eval_steps_per_second": 1.01,
"step": 456
},
{
"epoch": 3.0,
"step": 456,
"total_flos": 1.6172409563591475e+17,
"train_loss": 0.366262159849468,
"train_runtime": 5417.5508,
"train_samples_per_second": 10.741,
"train_steps_per_second": 0.084
}
],
"logging_steps": 1,
"max_steps": 456,
"num_input_tokens_seen": 0,
"num_train_epochs": 3,
"save_steps": 500,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 1.6172409563591475e+17,
"train_batch_size": 8,
"trial_name": null,
"trial_params": null
}