multisubject_medicine_mc / trainer_state.json
neginr's picture
End of training
009e3c1 verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 6.955414012738854,
"eval_steps": 500,
"global_step": 364,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.01910828025477707,
"grad_norm": 14.990140022095986,
"learning_rate": 5.405405405405406e-07,
"loss": 1.5145,
"step": 1
},
{
"epoch": 0.03821656050955414,
"grad_norm": 14.385565542515142,
"learning_rate": 1.0810810810810812e-06,
"loss": 1.4886,
"step": 2
},
{
"epoch": 0.05732484076433121,
"grad_norm": 14.479366215173409,
"learning_rate": 1.6216216216216219e-06,
"loss": 1.4943,
"step": 3
},
{
"epoch": 0.07643312101910828,
"grad_norm": 14.524746880859054,
"learning_rate": 2.1621621621621623e-06,
"loss": 1.5135,
"step": 4
},
{
"epoch": 0.09554140127388536,
"grad_norm": 12.585870030810783,
"learning_rate": 2.702702702702703e-06,
"loss": 1.4553,
"step": 5
},
{
"epoch": 0.11464968152866242,
"grad_norm": 10.656848802496679,
"learning_rate": 3.2432432432432437e-06,
"loss": 1.3436,
"step": 6
},
{
"epoch": 0.1337579617834395,
"grad_norm": 7.960921811444013,
"learning_rate": 3.7837837837837844e-06,
"loss": 1.2728,
"step": 7
},
{
"epoch": 0.15286624203821655,
"grad_norm": 6.812177612569246,
"learning_rate": 4.324324324324325e-06,
"loss": 1.2222,
"step": 8
},
{
"epoch": 0.17197452229299362,
"grad_norm": 4.986047582488592,
"learning_rate": 4.864864864864866e-06,
"loss": 1.1818,
"step": 9
},
{
"epoch": 0.1910828025477707,
"grad_norm": 6.412552449052916,
"learning_rate": 5.405405405405406e-06,
"loss": 1.1491,
"step": 10
},
{
"epoch": 0.21019108280254778,
"grad_norm": 5.494825671780005,
"learning_rate": 5.945945945945947e-06,
"loss": 1.0885,
"step": 11
},
{
"epoch": 0.22929936305732485,
"grad_norm": 4.430026646060167,
"learning_rate": 6.486486486486487e-06,
"loss": 1.0825,
"step": 12
},
{
"epoch": 0.2484076433121019,
"grad_norm": 4.42759103791487,
"learning_rate": 7.027027027027028e-06,
"loss": 1.0223,
"step": 13
},
{
"epoch": 0.267515923566879,
"grad_norm": 5.062626572152773,
"learning_rate": 7.567567567567569e-06,
"loss": 1.0311,
"step": 14
},
{
"epoch": 0.28662420382165604,
"grad_norm": 4.648519435630669,
"learning_rate": 8.108108108108109e-06,
"loss": 0.9974,
"step": 15
},
{
"epoch": 0.3057324840764331,
"grad_norm": 3.704566808626819,
"learning_rate": 8.64864864864865e-06,
"loss": 0.9995,
"step": 16
},
{
"epoch": 0.3248407643312102,
"grad_norm": 2.7515165681456812,
"learning_rate": 9.189189189189191e-06,
"loss": 0.944,
"step": 17
},
{
"epoch": 0.34394904458598724,
"grad_norm": 2.8097914129434836,
"learning_rate": 9.729729729729732e-06,
"loss": 0.9407,
"step": 18
},
{
"epoch": 0.3630573248407643,
"grad_norm": 3.2291111191358435,
"learning_rate": 1.027027027027027e-05,
"loss": 0.9214,
"step": 19
},
{
"epoch": 0.3821656050955414,
"grad_norm": 2.5625348519696294,
"learning_rate": 1.0810810810810812e-05,
"loss": 0.9183,
"step": 20
},
{
"epoch": 0.4012738853503185,
"grad_norm": 2.232749262686051,
"learning_rate": 1.1351351351351352e-05,
"loss": 0.9078,
"step": 21
},
{
"epoch": 0.42038216560509556,
"grad_norm": 1.9156573821747163,
"learning_rate": 1.1891891891891894e-05,
"loss": 0.878,
"step": 22
},
{
"epoch": 0.4394904458598726,
"grad_norm": 1.941941511919396,
"learning_rate": 1.2432432432432433e-05,
"loss": 0.8715,
"step": 23
},
{
"epoch": 0.4585987261146497,
"grad_norm": 2.0994478371302074,
"learning_rate": 1.2972972972972975e-05,
"loss": 0.8773,
"step": 24
},
{
"epoch": 0.47770700636942676,
"grad_norm": 1.6275791350191224,
"learning_rate": 1.3513513513513515e-05,
"loss": 0.8805,
"step": 25
},
{
"epoch": 0.4968152866242038,
"grad_norm": 1.515599235439942,
"learning_rate": 1.4054054054054055e-05,
"loss": 0.8635,
"step": 26
},
{
"epoch": 0.5159235668789809,
"grad_norm": 1.6443236891879496,
"learning_rate": 1.4594594594594596e-05,
"loss": 0.8514,
"step": 27
},
{
"epoch": 0.535031847133758,
"grad_norm": 2.1566568550834795,
"learning_rate": 1.5135135135135138e-05,
"loss": 0.8666,
"step": 28
},
{
"epoch": 0.554140127388535,
"grad_norm": 1.865335585637504,
"learning_rate": 1.5675675675675676e-05,
"loss": 0.8535,
"step": 29
},
{
"epoch": 0.5732484076433121,
"grad_norm": 1.513729648970103,
"learning_rate": 1.6216216216216218e-05,
"loss": 0.8754,
"step": 30
},
{
"epoch": 0.5923566878980892,
"grad_norm": 1.5420010766356778,
"learning_rate": 1.6756756756756757e-05,
"loss": 0.8624,
"step": 31
},
{
"epoch": 0.6114649681528662,
"grad_norm": 2.0504598465800044,
"learning_rate": 1.72972972972973e-05,
"loss": 0.8505,
"step": 32
},
{
"epoch": 0.6305732484076433,
"grad_norm": 1.6138987956162945,
"learning_rate": 1.783783783783784e-05,
"loss": 0.8609,
"step": 33
},
{
"epoch": 0.6496815286624203,
"grad_norm": 1.4520722307746579,
"learning_rate": 1.8378378378378383e-05,
"loss": 0.8408,
"step": 34
},
{
"epoch": 0.6687898089171974,
"grad_norm": 1.7223251738619296,
"learning_rate": 1.891891891891892e-05,
"loss": 0.8289,
"step": 35
},
{
"epoch": 0.6878980891719745,
"grad_norm": 1.5293784012050076,
"learning_rate": 1.9459459459459463e-05,
"loss": 0.8407,
"step": 36
},
{
"epoch": 0.7070063694267515,
"grad_norm": 1.640633324255724,
"learning_rate": 2e-05,
"loss": 0.8339,
"step": 37
},
{
"epoch": 0.7261146496815286,
"grad_norm": 1.4337126211032725,
"learning_rate": 1.9999538500851633e-05,
"loss": 0.8484,
"step": 38
},
{
"epoch": 0.7452229299363057,
"grad_norm": 1.3829040809921898,
"learning_rate": 1.9998154046002822e-05,
"loss": 0.8237,
"step": 39
},
{
"epoch": 0.7643312101910829,
"grad_norm": 1.3338599298109723,
"learning_rate": 1.9995846763238514e-05,
"loss": 0.7996,
"step": 40
},
{
"epoch": 0.7834394904458599,
"grad_norm": 1.6437827729744903,
"learning_rate": 1.9992616865520515e-05,
"loss": 0.8503,
"step": 41
},
{
"epoch": 0.802547770700637,
"grad_norm": 1.5035408873313147,
"learning_rate": 1.9988464650967834e-05,
"loss": 0.8288,
"step": 42
},
{
"epoch": 0.821656050955414,
"grad_norm": 1.9307767442896708,
"learning_rate": 1.9983390502829168e-05,
"loss": 0.836,
"step": 43
},
{
"epoch": 0.8407643312101911,
"grad_norm": 1.6121481786901808,
"learning_rate": 1.9977394889447526e-05,
"loss": 0.8268,
"step": 44
},
{
"epoch": 0.8598726114649682,
"grad_norm": 1.737736803264758,
"learning_rate": 1.9970478364217e-05,
"loss": 0.836,
"step": 45
},
{
"epoch": 0.8789808917197452,
"grad_norm": 1.7605640838470302,
"learning_rate": 1.9962641565531694e-05,
"loss": 0.8257,
"step": 46
},
{
"epoch": 0.8980891719745223,
"grad_norm": 1.6081290667717165,
"learning_rate": 1.9953885216726788e-05,
"loss": 0.8443,
"step": 47
},
{
"epoch": 0.9171974522292994,
"grad_norm": 1.7541275129594776,
"learning_rate": 1.994421012601179e-05,
"loss": 0.8366,
"step": 48
},
{
"epoch": 0.9363057324840764,
"grad_norm": 1.383033283548282,
"learning_rate": 1.9933617186395917e-05,
"loss": 0.8154,
"step": 49
},
{
"epoch": 0.9554140127388535,
"grad_norm": 1.6777566379598123,
"learning_rate": 1.99221073756057e-05,
"loss": 0.836,
"step": 50
},
{
"epoch": 0.9745222929936306,
"grad_norm": 1.3270894313186803,
"learning_rate": 1.990968175599471e-05,
"loss": 0.8229,
"step": 51
},
{
"epoch": 0.9936305732484076,
"grad_norm": 1.311413944058582,
"learning_rate": 1.9896341474445526e-05,
"loss": 0.7993,
"step": 52
},
{
"epoch": 1.0127388535031847,
"grad_norm": 1.5134518938324237,
"learning_rate": 1.9882087762263857e-05,
"loss": 0.7211,
"step": 53
},
{
"epoch": 1.0318471337579618,
"grad_norm": 1.4457472761668542,
"learning_rate": 1.9866921935064907e-05,
"loss": 0.65,
"step": 54
},
{
"epoch": 1.0509554140127388,
"grad_norm": 1.6984985951003597,
"learning_rate": 1.985084539265195e-05,
"loss": 0.6405,
"step": 55
},
{
"epoch": 1.070063694267516,
"grad_norm": 1.340217284611253,
"learning_rate": 1.983385961888711e-05,
"loss": 0.6557,
"step": 56
},
{
"epoch": 1.089171974522293,
"grad_norm": 1.4683215371294849,
"learning_rate": 1.9815966181554412e-05,
"loss": 0.6382,
"step": 57
},
{
"epoch": 1.10828025477707,
"grad_norm": 1.3073685584419914,
"learning_rate": 1.9797166732215078e-05,
"loss": 0.628,
"step": 58
},
{
"epoch": 1.127388535031847,
"grad_norm": 1.4163134544358502,
"learning_rate": 1.977746300605507e-05,
"loss": 0.5882,
"step": 59
},
{
"epoch": 1.1464968152866242,
"grad_norm": 1.4044561953506427,
"learning_rate": 1.975685682172497e-05,
"loss": 0.6272,
"step": 60
},
{
"epoch": 1.1656050955414012,
"grad_norm": 1.353708898359601,
"learning_rate": 1.973535008117207e-05,
"loss": 0.6121,
"step": 61
},
{
"epoch": 1.1847133757961783,
"grad_norm": 1.256242946923662,
"learning_rate": 1.9712944769464864e-05,
"loss": 0.6169,
"step": 62
},
{
"epoch": 1.2038216560509554,
"grad_norm": 1.2353934101359754,
"learning_rate": 1.9689642954609808e-05,
"loss": 0.6123,
"step": 63
},
{
"epoch": 1.2229299363057324,
"grad_norm": 1.242011368111671,
"learning_rate": 1.9665446787360444e-05,
"loss": 0.6738,
"step": 64
},
{
"epoch": 1.2420382165605095,
"grad_norm": 1.1948083066768818,
"learning_rate": 1.9640358501018885e-05,
"loss": 0.5984,
"step": 65
},
{
"epoch": 1.2611464968152866,
"grad_norm": 1.227900627562625,
"learning_rate": 1.9614380411229693e-05,
"loss": 0.6112,
"step": 66
},
{
"epoch": 1.2802547770700636,
"grad_norm": 1.2766612501501673,
"learning_rate": 1.9587514915766124e-05,
"loss": 0.6423,
"step": 67
},
{
"epoch": 1.2993630573248407,
"grad_norm": 1.3704435778321213,
"learning_rate": 1.9559764494308838e-05,
"loss": 0.5744,
"step": 68
},
{
"epoch": 1.3184713375796178,
"grad_norm": 1.4139525066526928,
"learning_rate": 1.9531131708217005e-05,
"loss": 0.5943,
"step": 69
},
{
"epoch": 1.3375796178343948,
"grad_norm": 1.259594103517775,
"learning_rate": 1.950161920029191e-05,
"loss": 0.63,
"step": 70
},
{
"epoch": 1.356687898089172,
"grad_norm": 1.5341795996533234,
"learning_rate": 1.9471229694533003e-05,
"loss": 0.6243,
"step": 71
},
{
"epoch": 1.3757961783439492,
"grad_norm": 1.2433858014815762,
"learning_rate": 1.943996599588649e-05,
"loss": 0.6225,
"step": 72
},
{
"epoch": 1.394904458598726,
"grad_norm": 1.2066678457467106,
"learning_rate": 1.940783098998643e-05,
"loss": 0.6018,
"step": 73
},
{
"epoch": 1.4140127388535033,
"grad_norm": 1.2748999280410294,
"learning_rate": 1.93748276428884e-05,
"loss": 0.5936,
"step": 74
},
{
"epoch": 1.4331210191082802,
"grad_norm": 1.2920229857639585,
"learning_rate": 1.9340959000795707e-05,
"loss": 0.6397,
"step": 75
},
{
"epoch": 1.4522292993630574,
"grad_norm": 1.3957421713319056,
"learning_rate": 1.9306228189778255e-05,
"loss": 0.6213,
"step": 76
},
{
"epoch": 1.4713375796178343,
"grad_norm": 1.4014972976675841,
"learning_rate": 1.927063841548398e-05,
"loss": 0.6195,
"step": 77
},
{
"epoch": 1.4904458598726116,
"grad_norm": 1.2887931174342688,
"learning_rate": 1.9234192962842996e-05,
"loss": 0.5891,
"step": 78
},
{
"epoch": 1.5095541401273884,
"grad_norm": 1.2715731353165856,
"learning_rate": 1.9196895195764363e-05,
"loss": 0.6315,
"step": 79
},
{
"epoch": 1.5286624203821657,
"grad_norm": 1.229751944932979,
"learning_rate": 1.9158748556825637e-05,
"loss": 0.6056,
"step": 80
},
{
"epoch": 1.5477707006369426,
"grad_norm": 1.442724258592494,
"learning_rate": 1.9119756566955092e-05,
"loss": 0.6364,
"step": 81
},
{
"epoch": 1.5668789808917198,
"grad_norm": 1.202348821649108,
"learning_rate": 1.907992282510675e-05,
"loss": 0.6271,
"step": 82
},
{
"epoch": 1.5859872611464967,
"grad_norm": 1.2836864625828608,
"learning_rate": 1.90392510079282e-05,
"loss": 0.6128,
"step": 83
},
{
"epoch": 1.605095541401274,
"grad_norm": 1.1800499309121613,
"learning_rate": 1.8997744869421248e-05,
"loss": 0.6083,
"step": 84
},
{
"epoch": 1.6242038216560508,
"grad_norm": 1.216225762413445,
"learning_rate": 1.8955408240595396e-05,
"loss": 0.6377,
"step": 85
},
{
"epoch": 1.643312101910828,
"grad_norm": 1.1638626931156888,
"learning_rate": 1.891224502911428e-05,
"loss": 0.6178,
"step": 86
},
{
"epoch": 1.662420382165605,
"grad_norm": 1.1974901367787416,
"learning_rate": 1.886825921893497e-05,
"loss": 0.5987,
"step": 87
},
{
"epoch": 1.6815286624203822,
"grad_norm": 1.169886620928099,
"learning_rate": 1.8823454869940243e-05,
"loss": 0.583,
"step": 88
},
{
"epoch": 1.700636942675159,
"grad_norm": 1.1742563632707246,
"learning_rate": 1.8777836117563894e-05,
"loss": 0.6144,
"step": 89
},
{
"epoch": 1.7197452229299364,
"grad_norm": 1.165573756913235,
"learning_rate": 1.873140717240899e-05,
"loss": 0.6324,
"step": 90
},
{
"epoch": 1.7388535031847132,
"grad_norm": 1.3576876293848648,
"learning_rate": 1.8684172319859258e-05,
"loss": 0.6505,
"step": 91
},
{
"epoch": 1.7579617834394905,
"grad_norm": 1.2097778554735674,
"learning_rate": 1.863613591968355e-05,
"loss": 0.6176,
"step": 92
},
{
"epoch": 1.7770700636942676,
"grad_norm": 1.3440536330118746,
"learning_rate": 1.858730240563342e-05,
"loss": 0.5928,
"step": 93
},
{
"epoch": 1.7961783439490446,
"grad_norm": 1.5421403847520048,
"learning_rate": 1.8537676285033886e-05,
"loss": 0.6094,
"step": 94
},
{
"epoch": 1.8152866242038217,
"grad_norm": 1.298905703769219,
"learning_rate": 1.848726213836744e-05,
"loss": 0.615,
"step": 95
},
{
"epoch": 1.8343949044585988,
"grad_norm": 1.5410691337183182,
"learning_rate": 1.8436064618851225e-05,
"loss": 0.5955,
"step": 96
},
{
"epoch": 1.8535031847133758,
"grad_norm": 1.4380886980637535,
"learning_rate": 1.838408845200758e-05,
"loss": 0.5845,
"step": 97
},
{
"epoch": 1.872611464968153,
"grad_norm": 1.2927175032806673,
"learning_rate": 1.8331338435227838e-05,
"loss": 0.5969,
"step": 98
},
{
"epoch": 1.89171974522293,
"grad_norm": 1.4661387293954078,
"learning_rate": 1.8277819437329577e-05,
"loss": 0.6284,
"step": 99
},
{
"epoch": 1.910828025477707,
"grad_norm": 1.4690243089475816,
"learning_rate": 1.8223536398107177e-05,
"loss": 0.6102,
"step": 100
},
{
"epoch": 1.929936305732484,
"grad_norm": 1.2407333702545924,
"learning_rate": 1.8168494327875918e-05,
"loss": 0.5917,
"step": 101
},
{
"epoch": 1.9490445859872612,
"grad_norm": 1.2765830269044587,
"learning_rate": 1.8112698307009506e-05,
"loss": 0.6346,
"step": 102
},
{
"epoch": 1.9681528662420382,
"grad_norm": 1.2005302634214932,
"learning_rate": 1.8056153485471167e-05,
"loss": 0.6232,
"step": 103
},
{
"epoch": 1.9872611464968153,
"grad_norm": 1.148963792464725,
"learning_rate": 1.799886508233829e-05,
"loss": 0.5957,
"step": 104
},
{
"epoch": 2.0063694267515926,
"grad_norm": 1.2628345318954612,
"learning_rate": 1.7940838385320732e-05,
"loss": 0.5155,
"step": 105
},
{
"epoch": 2.0254777070063694,
"grad_norm": 1.3070169831120713,
"learning_rate": 1.788207875027274e-05,
"loss": 0.4399,
"step": 106
},
{
"epoch": 2.0445859872611467,
"grad_norm": 1.3027394925308748,
"learning_rate": 1.7822591600698632e-05,
"loss": 0.3959,
"step": 107
},
{
"epoch": 2.0636942675159236,
"grad_norm": 2.0977150823190454,
"learning_rate": 1.776238242725217e-05,
"loss": 0.4197,
"step": 108
},
{
"epoch": 2.082802547770701,
"grad_norm": 1.5059789233190988,
"learning_rate": 1.7701456787229805e-05,
"loss": 0.3949,
"step": 109
},
{
"epoch": 2.1019108280254777,
"grad_norm": 3.3191966173358805,
"learning_rate": 1.7639820304057745e-05,
"loss": 0.3931,
"step": 110
},
{
"epoch": 2.121019108280255,
"grad_norm": 1.3624451723710866,
"learning_rate": 1.7577478666772886e-05,
"loss": 0.3843,
"step": 111
},
{
"epoch": 2.140127388535032,
"grad_norm": 1.1484719310505338,
"learning_rate": 1.751443762949772e-05,
"loss": 0.3745,
"step": 112
},
{
"epoch": 2.159235668789809,
"grad_norm": 1.2700892780904618,
"learning_rate": 1.7450703010909263e-05,
"loss": 0.413,
"step": 113
},
{
"epoch": 2.178343949044586,
"grad_norm": 1.0586189376543789,
"learning_rate": 1.738628069370195e-05,
"loss": 0.3464,
"step": 114
},
{
"epoch": 2.1974522292993632,
"grad_norm": 1.1783273985129143,
"learning_rate": 1.732117662404469e-05,
"loss": 0.3512,
"step": 115
},
{
"epoch": 2.21656050955414,
"grad_norm": 1.1532860416115633,
"learning_rate": 1.7255396811032014e-05,
"loss": 0.3493,
"step": 116
},
{
"epoch": 2.2356687898089174,
"grad_norm": 1.07196230298279,
"learning_rate": 1.718894732612947e-05,
"loss": 0.3543,
"step": 117
},
{
"epoch": 2.254777070063694,
"grad_norm": 1.09377483988167,
"learning_rate": 1.712183430261319e-05,
"loss": 0.3378,
"step": 118
},
{
"epoch": 2.2738853503184715,
"grad_norm": 1.1086380438083379,
"learning_rate": 1.7054063935003813e-05,
"loss": 0.3691,
"step": 119
},
{
"epoch": 2.2929936305732483,
"grad_norm": 1.1110135807811403,
"learning_rate": 1.698564247849473e-05,
"loss": 0.3703,
"step": 120
},
{
"epoch": 2.3121019108280256,
"grad_norm": 1.0959457554638712,
"learning_rate": 1.691657624837472e-05,
"loss": 0.3382,
"step": 121
},
{
"epoch": 2.3312101910828025,
"grad_norm": 1.0763216442984707,
"learning_rate": 1.684687161944506e-05,
"loss": 0.3655,
"step": 122
},
{
"epoch": 2.3503184713375798,
"grad_norm": 1.0644643193109964,
"learning_rate": 1.677653502543113e-05,
"loss": 0.3713,
"step": 123
},
{
"epoch": 2.3694267515923566,
"grad_norm": 1.0520091051033431,
"learning_rate": 1.6705572958388576e-05,
"loss": 0.3637,
"step": 124
},
{
"epoch": 2.388535031847134,
"grad_norm": 0.9715610282842169,
"learning_rate": 1.6633991968104095e-05,
"loss": 0.3342,
"step": 125
},
{
"epoch": 2.4076433121019107,
"grad_norm": 1.0495947222277981,
"learning_rate": 1.6561798661490904e-05,
"loss": 0.362,
"step": 126
},
{
"epoch": 2.426751592356688,
"grad_norm": 1.027768605633936,
"learning_rate": 1.6488999701978905e-05,
"loss": 0.3909,
"step": 127
},
{
"epoch": 2.445859872611465,
"grad_norm": 1.0430605883090662,
"learning_rate": 1.6415601808899658e-05,
"loss": 0.3707,
"step": 128
},
{
"epoch": 2.464968152866242,
"grad_norm": 0.9897344708027874,
"learning_rate": 1.63416117568662e-05,
"loss": 0.3716,
"step": 129
},
{
"epoch": 2.484076433121019,
"grad_norm": 1.0206268353107175,
"learning_rate": 1.6267036375147728e-05,
"loss": 0.3791,
"step": 130
},
{
"epoch": 2.5031847133757963,
"grad_norm": 0.9968464604996385,
"learning_rate": 1.619188254703927e-05,
"loss": 0.3959,
"step": 131
},
{
"epoch": 2.522292993630573,
"grad_norm": 1.0450892578829158,
"learning_rate": 1.6116157209226356e-05,
"loss": 0.3935,
"step": 132
},
{
"epoch": 2.5414012738853504,
"grad_norm": 1.0420327393892306,
"learning_rate": 1.6039867351144778e-05,
"loss": 0.3732,
"step": 133
},
{
"epoch": 2.5605095541401273,
"grad_norm": 1.0658140476870972,
"learning_rate": 1.5963020014335437e-05,
"loss": 0.3679,
"step": 134
},
{
"epoch": 2.5796178343949046,
"grad_norm": 1.015910488975753,
"learning_rate": 1.588562229179443e-05,
"loss": 0.4047,
"step": 135
},
{
"epoch": 2.5987261146496814,
"grad_norm": 1.0151425054438397,
"learning_rate": 1.5807681327318372e-05,
"loss": 0.3517,
"step": 136
},
{
"epoch": 2.6178343949044587,
"grad_norm": 1.1559403742194478,
"learning_rate": 1.5729204314845002e-05,
"loss": 0.3777,
"step": 137
},
{
"epoch": 2.6369426751592355,
"grad_norm": 1.1338157549104528,
"learning_rate": 1.56501984977892e-05,
"loss": 0.3729,
"step": 138
},
{
"epoch": 2.656050955414013,
"grad_norm": 1.064291963474977,
"learning_rate": 1.557067116837444e-05,
"loss": 0.3829,
"step": 139
},
{
"epoch": 2.6751592356687897,
"grad_norm": 1.0427156725933961,
"learning_rate": 1.5490629666959668e-05,
"loss": 0.39,
"step": 140
},
{
"epoch": 2.694267515923567,
"grad_norm": 1.0226624788733183,
"learning_rate": 1.541008138136183e-05,
"loss": 0.3889,
"step": 141
},
{
"epoch": 2.713375796178344,
"grad_norm": 1.0525328188947,
"learning_rate": 1.5329033746173975e-05,
"loss": 0.3612,
"step": 142
},
{
"epoch": 2.732484076433121,
"grad_norm": 1.078767271893382,
"learning_rate": 1.5247494242079024e-05,
"loss": 0.4071,
"step": 143
},
{
"epoch": 2.7515923566878984,
"grad_norm": 0.9854161293601749,
"learning_rate": 1.5165470395159314e-05,
"loss": 0.3517,
"step": 144
},
{
"epoch": 2.770700636942675,
"grad_norm": 1.0216191388051186,
"learning_rate": 1.5082969776201948e-05,
"loss": 0.3634,
"step": 145
},
{
"epoch": 2.789808917197452,
"grad_norm": 1.0260644601169646,
"learning_rate": 1.5000000000000002e-05,
"loss": 0.3458,
"step": 146
},
{
"epoch": 2.8089171974522293,
"grad_norm": 1.0679550777084548,
"learning_rate": 1.4916568724649688e-05,
"loss": 0.3719,
"step": 147
},
{
"epoch": 2.8280254777070066,
"grad_norm": 0.9833092792439724,
"learning_rate": 1.483268365084351e-05,
"loss": 0.3577,
"step": 148
},
{
"epoch": 2.8471337579617835,
"grad_norm": 1.0213618618762001,
"learning_rate": 1.4748352521159492e-05,
"loss": 0.3769,
"step": 149
},
{
"epoch": 2.8662420382165603,
"grad_norm": 0.9881143303734706,
"learning_rate": 1.466358311934654e-05,
"loss": 0.37,
"step": 150
},
{
"epoch": 2.8853503184713376,
"grad_norm": 1.0426528347787805,
"learning_rate": 1.4578383269606004e-05,
"loss": 0.3737,
"step": 151
},
{
"epoch": 2.904458598726115,
"grad_norm": 1.011316944159334,
"learning_rate": 1.4492760835869504e-05,
"loss": 0.3484,
"step": 152
},
{
"epoch": 2.9235668789808917,
"grad_norm": 1.0771107253387162,
"learning_rate": 1.4406723721073088e-05,
"loss": 0.3458,
"step": 153
},
{
"epoch": 2.9426751592356686,
"grad_norm": 1.0914736531721394,
"learning_rate": 1.4320279866427798e-05,
"loss": 0.3837,
"step": 154
},
{
"epoch": 2.961783439490446,
"grad_norm": 0.9879292075174337,
"learning_rate": 1.4233437250686695e-05,
"loss": 0.3379,
"step": 155
},
{
"epoch": 2.980891719745223,
"grad_norm": 1.0534558360040707,
"learning_rate": 1.4146203889408418e-05,
"loss": 0.3737,
"step": 156
},
{
"epoch": 3.0,
"grad_norm": 0.9546429746260057,
"learning_rate": 1.4058587834217356e-05,
"loss": 0.3077,
"step": 157
},
{
"epoch": 3.0191082802547773,
"grad_norm": 1.0794314661371742,
"learning_rate": 1.3970597172060482e-05,
"loss": 0.2445,
"step": 158
},
{
"epoch": 3.038216560509554,
"grad_norm": 0.8725593864518827,
"learning_rate": 1.3882240024460928e-05,
"loss": 0.251,
"step": 159
},
{
"epoch": 3.0573248407643314,
"grad_norm": 0.9528452182374068,
"learning_rate": 1.3793524546768358e-05,
"loss": 0.222,
"step": 160
},
{
"epoch": 3.0764331210191083,
"grad_norm": 1.243120588259197,
"learning_rate": 1.3704458927406261e-05,
"loss": 0.2438,
"step": 161
},
{
"epoch": 3.0955414012738856,
"grad_norm": 0.9927780427766021,
"learning_rate": 1.3615051387116131e-05,
"loss": 0.2071,
"step": 162
},
{
"epoch": 3.1146496815286624,
"grad_norm": 0.8760051519137424,
"learning_rate": 1.3525310178198707e-05,
"loss": 0.2072,
"step": 163
},
{
"epoch": 3.1337579617834397,
"grad_norm": 0.993441527748417,
"learning_rate": 1.3435243583752294e-05,
"loss": 0.1903,
"step": 164
},
{
"epoch": 3.1528662420382165,
"grad_norm": 1.0122823808035941,
"learning_rate": 1.3344859916908206e-05,
"loss": 0.223,
"step": 165
},
{
"epoch": 3.171974522292994,
"grad_norm": 0.8280127986581172,
"learning_rate": 1.325416752006351e-05,
"loss": 0.2167,
"step": 166
},
{
"epoch": 3.1910828025477707,
"grad_norm": 0.9207860411765015,
"learning_rate": 1.3163174764110985e-05,
"loss": 0.2234,
"step": 167
},
{
"epoch": 3.210191082802548,
"grad_norm": 1.0087540713450258,
"learning_rate": 1.3071890047666498e-05,
"loss": 0.2607,
"step": 168
},
{
"epoch": 3.229299363057325,
"grad_norm": 0.8754252182890954,
"learning_rate": 1.2980321796293838e-05,
"loss": 0.2321,
"step": 169
},
{
"epoch": 3.248407643312102,
"grad_norm": 0.8428251952961017,
"learning_rate": 1.288847846172701e-05,
"loss": 0.2115,
"step": 170
},
{
"epoch": 3.267515923566879,
"grad_norm": 0.854200909124553,
"learning_rate": 1.2796368521090143e-05,
"loss": 0.2184,
"step": 171
},
{
"epoch": 3.286624203821656,
"grad_norm": 0.869443567808356,
"learning_rate": 1.2704000476115079e-05,
"loss": 0.2382,
"step": 172
},
{
"epoch": 3.305732484076433,
"grad_norm": 0.8459308281223218,
"learning_rate": 1.2611382852356632e-05,
"loss": 0.2425,
"step": 173
},
{
"epoch": 3.3248407643312103,
"grad_norm": 0.8148693074167046,
"learning_rate": 1.2518524198405699e-05,
"loss": 0.1893,
"step": 174
},
{
"epoch": 3.343949044585987,
"grad_norm": 0.8713896320210411,
"learning_rate": 1.2425433085100224e-05,
"loss": 0.2613,
"step": 175
},
{
"epoch": 3.3630573248407645,
"grad_norm": 0.9166069241971261,
"learning_rate": 1.233211810473411e-05,
"loss": 0.2493,
"step": 176
},
{
"epoch": 3.3821656050955413,
"grad_norm": 0.8690652011432752,
"learning_rate": 1.2238587870264152e-05,
"loss": 0.2547,
"step": 177
},
{
"epoch": 3.4012738853503186,
"grad_norm": 0.862487386493507,
"learning_rate": 1.2144851014515055e-05,
"loss": 0.2292,
"step": 178
},
{
"epoch": 3.4203821656050954,
"grad_norm": 0.8815704028252178,
"learning_rate": 1.2050916189382646e-05,
"loss": 0.2064,
"step": 179
},
{
"epoch": 3.4394904458598727,
"grad_norm": 0.8675773397668213,
"learning_rate": 1.1956792065035281e-05,
"loss": 0.25,
"step": 180
},
{
"epoch": 3.4585987261146496,
"grad_norm": 0.8933077382821798,
"learning_rate": 1.1862487329113606e-05,
"loss": 0.2358,
"step": 181
},
{
"epoch": 3.477707006369427,
"grad_norm": 0.8903704341987667,
"learning_rate": 1.1768010685928686e-05,
"loss": 0.2264,
"step": 182
},
{
"epoch": 3.4968152866242037,
"grad_norm": 0.8383070066020237,
"learning_rate": 1.1673370855658592e-05,
"loss": 0.2043,
"step": 183
},
{
"epoch": 3.515923566878981,
"grad_norm": 0.8911649033404783,
"learning_rate": 1.1578576573543541e-05,
"loss": 0.2178,
"step": 184
},
{
"epoch": 3.535031847133758,
"grad_norm": 0.8585210661904665,
"learning_rate": 1.1483636589079627e-05,
"loss": 0.2156,
"step": 185
},
{
"epoch": 3.554140127388535,
"grad_norm": 0.8162221440030065,
"learning_rate": 1.1388559665211241e-05,
"loss": 0.2548,
"step": 186
},
{
"epoch": 3.573248407643312,
"grad_norm": 0.7963772067435857,
"learning_rate": 1.1293354577522264e-05,
"loss": 0.2381,
"step": 187
},
{
"epoch": 3.5923566878980893,
"grad_norm": 0.8677746097750828,
"learning_rate": 1.1198030113426074e-05,
"loss": 0.2054,
"step": 188
},
{
"epoch": 3.611464968152866,
"grad_norm": 0.895188923289957,
"learning_rate": 1.1102595071354471e-05,
"loss": 0.2171,
"step": 189
},
{
"epoch": 3.6305732484076434,
"grad_norm": 0.8791187036486455,
"learning_rate": 1.1007058259945584e-05,
"loss": 0.2477,
"step": 190
},
{
"epoch": 3.6496815286624202,
"grad_norm": 0.8155997120595588,
"learning_rate": 1.0911428497230834e-05,
"loss": 0.2045,
"step": 191
},
{
"epoch": 3.6687898089171975,
"grad_norm": 0.8770814295828485,
"learning_rate": 1.0815714609821027e-05,
"loss": 0.2383,
"step": 192
},
{
"epoch": 3.6878980891719744,
"grad_norm": 0.819184381321191,
"learning_rate": 1.0719925432091671e-05,
"loss": 0.2143,
"step": 193
},
{
"epoch": 3.7070063694267517,
"grad_norm": 0.8250575947986096,
"learning_rate": 1.0624069805367558e-05,
"loss": 0.1944,
"step": 194
},
{
"epoch": 3.7261146496815285,
"grad_norm": 0.8276542050458014,
"learning_rate": 1.0528156577106703e-05,
"loss": 0.2404,
"step": 195
},
{
"epoch": 3.745222929936306,
"grad_norm": 0.8410279981631184,
"learning_rate": 1.043219460008374e-05,
"loss": 0.2131,
"step": 196
},
{
"epoch": 3.7643312101910826,
"grad_norm": 0.8003839667268587,
"learning_rate": 1.0336192731572805e-05,
"loss": 0.2373,
"step": 197
},
{
"epoch": 3.78343949044586,
"grad_norm": 0.8003447028390943,
"learning_rate": 1.0240159832530007e-05,
"loss": 0.2228,
"step": 198
},
{
"epoch": 3.802547770700637,
"grad_norm": 0.8107666767644914,
"learning_rate": 1.0144104766775574e-05,
"loss": 0.2245,
"step": 199
},
{
"epoch": 3.821656050955414,
"grad_norm": 0.8283501745311356,
"learning_rate": 1.004803640017571e-05,
"loss": 0.1727,
"step": 200
},
{
"epoch": 3.840764331210191,
"grad_norm": 0.8430576593378581,
"learning_rate": 9.951963599824294e-06,
"loss": 0.241,
"step": 201
},
{
"epoch": 3.859872611464968,
"grad_norm": 0.842864753329655,
"learning_rate": 9.855895233224431e-06,
"loss": 0.2361,
"step": 202
},
{
"epoch": 3.8789808917197455,
"grad_norm": 0.7894891328436988,
"learning_rate": 9.759840167469995e-06,
"loss": 0.1798,
"step": 203
},
{
"epoch": 3.8980891719745223,
"grad_norm": 0.8043963124163499,
"learning_rate": 9.663807268427197e-06,
"loss": 0.2059,
"step": 204
},
{
"epoch": 3.917197452229299,
"grad_norm": 0.8321817031652909,
"learning_rate": 9.56780539991626e-06,
"loss": 0.212,
"step": 205
},
{
"epoch": 3.9363057324840764,
"grad_norm": 0.8074239055674105,
"learning_rate": 9.471843422893299e-06,
"loss": 0.2009,
"step": 206
},
{
"epoch": 3.9554140127388537,
"grad_norm": 0.8557990210783927,
"learning_rate": 9.375930194632447e-06,
"loss": 0.2145,
"step": 207
},
{
"epoch": 3.9745222929936306,
"grad_norm": 0.7812183187749983,
"learning_rate": 9.28007456790833e-06,
"loss": 0.1866,
"step": 208
},
{
"epoch": 3.9936305732484074,
"grad_norm": 0.7736504835220844,
"learning_rate": 9.184285390178978e-06,
"loss": 0.2221,
"step": 209
},
{
"epoch": 4.012738853503185,
"grad_norm": 0.6358381739054941,
"learning_rate": 9.08857150276917e-06,
"loss": 0.1033,
"step": 210
},
{
"epoch": 4.031847133757962,
"grad_norm": 0.6598743747199938,
"learning_rate": 8.992941740054418e-06,
"loss": 0.148,
"step": 211
},
{
"epoch": 4.050955414012739,
"grad_norm": 0.577662790844215,
"learning_rate": 8.897404928645529e-06,
"loss": 0.1086,
"step": 212
},
{
"epoch": 4.070063694267516,
"grad_norm": 0.667177834347151,
"learning_rate": 8.80196988657393e-06,
"loss": 0.1663,
"step": 213
},
{
"epoch": 4.089171974522293,
"grad_norm": 0.7876247378563334,
"learning_rate": 8.706645422477739e-06,
"loss": 0.1534,
"step": 214
},
{
"epoch": 4.10828025477707,
"grad_norm": 0.7587047405394588,
"learning_rate": 8.611440334788762e-06,
"loss": 0.1326,
"step": 215
},
{
"epoch": 4.127388535031847,
"grad_norm": 0.646869384984155,
"learning_rate": 8.516363410920376e-06,
"loss": 0.1116,
"step": 216
},
{
"epoch": 4.146496815286624,
"grad_norm": 0.6538670553340495,
"learning_rate": 8.42142342645646e-06,
"loss": 0.1243,
"step": 217
},
{
"epoch": 4.165605095541402,
"grad_norm": 0.6104543037349788,
"learning_rate": 8.326629144341408e-06,
"loss": 0.1251,
"step": 218
},
{
"epoch": 4.1847133757961785,
"grad_norm": 0.6638827104955027,
"learning_rate": 8.231989314071318e-06,
"loss": 0.1235,
"step": 219
},
{
"epoch": 4.203821656050955,
"grad_norm": 0.6157950485106373,
"learning_rate": 8.137512670886397e-06,
"loss": 0.1318,
"step": 220
},
{
"epoch": 4.222929936305732,
"grad_norm": 0.5807970143215085,
"learning_rate": 8.043207934964722e-06,
"loss": 0.1528,
"step": 221
},
{
"epoch": 4.24203821656051,
"grad_norm": 0.6166843267142346,
"learning_rate": 7.949083810617358e-06,
"loss": 0.1777,
"step": 222
},
{
"epoch": 4.261146496815287,
"grad_norm": 0.6029875516314757,
"learning_rate": 7.855148985484946e-06,
"loss": 0.1336,
"step": 223
},
{
"epoch": 4.280254777070064,
"grad_norm": 0.6416238599404323,
"learning_rate": 7.761412129735853e-06,
"loss": 0.1447,
"step": 224
},
{
"epoch": 4.2993630573248405,
"grad_norm": 0.5618037002140864,
"learning_rate": 7.667881895265895e-06,
"loss": 0.1153,
"step": 225
},
{
"epoch": 4.318471337579618,
"grad_norm": 0.5872125669535371,
"learning_rate": 7.574566914899779e-06,
"loss": 0.1218,
"step": 226
},
{
"epoch": 4.337579617834395,
"grad_norm": 0.6311165305084404,
"learning_rate": 7.481475801594302e-06,
"loss": 0.1191,
"step": 227
},
{
"epoch": 4.356687898089172,
"grad_norm": 0.5835850069158302,
"learning_rate": 7.388617147643371e-06,
"loss": 0.1175,
"step": 228
},
{
"epoch": 4.375796178343949,
"grad_norm": 0.5641329873077757,
"learning_rate": 7.295999523884921e-06,
"loss": 0.1268,
"step": 229
},
{
"epoch": 4.3949044585987265,
"grad_norm": 0.5710638661565366,
"learning_rate": 7.203631478909857e-06,
"loss": 0.1196,
"step": 230
},
{
"epoch": 4.414012738853503,
"grad_norm": 0.6304674375983539,
"learning_rate": 7.111521538272997e-06,
"loss": 0.162,
"step": 231
},
{
"epoch": 4.43312101910828,
"grad_norm": 0.5925648384154195,
"learning_rate": 7.019678203706164e-06,
"loss": 0.101,
"step": 232
},
{
"epoch": 4.452229299363057,
"grad_norm": 0.5759144673258558,
"learning_rate": 6.928109952333506e-06,
"loss": 0.1412,
"step": 233
},
{
"epoch": 4.471337579617835,
"grad_norm": 0.5908269020834356,
"learning_rate": 6.83682523588902e-06,
"loss": 0.1535,
"step": 234
},
{
"epoch": 4.490445859872612,
"grad_norm": 0.5973740508699396,
"learning_rate": 6.745832479936492e-06,
"loss": 0.1317,
"step": 235
},
{
"epoch": 4.509554140127388,
"grad_norm": 0.587279118111191,
"learning_rate": 6.655140083091794e-06,
"loss": 0.1214,
"step": 236
},
{
"epoch": 4.528662420382165,
"grad_norm": 0.6209961542756124,
"learning_rate": 6.564756416247712e-06,
"loss": 0.1518,
"step": 237
},
{
"epoch": 4.547770700636943,
"grad_norm": 0.579382073138381,
"learning_rate": 6.474689821801295e-06,
"loss": 0.1309,
"step": 238
},
{
"epoch": 4.56687898089172,
"grad_norm": 0.6065440171745169,
"learning_rate": 6.384948612883872e-06,
"loss": 0.1009,
"step": 239
},
{
"epoch": 4.585987261146497,
"grad_norm": 0.5585615132779307,
"learning_rate": 6.2955410725937405e-06,
"loss": 0.1217,
"step": 240
},
{
"epoch": 4.6050955414012735,
"grad_norm": 0.5924523782164844,
"learning_rate": 6.206475453231644e-06,
"loss": 0.1109,
"step": 241
},
{
"epoch": 4.624203821656051,
"grad_norm": 0.5995874401366601,
"learning_rate": 6.117759975539075e-06,
"loss": 0.1182,
"step": 242
},
{
"epoch": 4.643312101910828,
"grad_norm": 0.5576794993986748,
"learning_rate": 6.029402827939519e-06,
"loss": 0.1121,
"step": 243
},
{
"epoch": 4.662420382165605,
"grad_norm": 0.5770135050173938,
"learning_rate": 5.941412165782645e-06,
"loss": 0.1245,
"step": 244
},
{
"epoch": 4.681528662420382,
"grad_norm": 0.5929257692806121,
"learning_rate": 5.853796110591583e-06,
"loss": 0.1511,
"step": 245
},
{
"epoch": 4.7006369426751595,
"grad_norm": 0.5925990861687294,
"learning_rate": 5.766562749313309e-06,
"loss": 0.1279,
"step": 246
},
{
"epoch": 4.719745222929936,
"grad_norm": 0.5699434481371307,
"learning_rate": 5.6797201335722064e-06,
"loss": 0.1501,
"step": 247
},
{
"epoch": 4.738853503184713,
"grad_norm": 0.5785541560481104,
"learning_rate": 5.593276278926912e-06,
"loss": 0.1234,
"step": 248
},
{
"epoch": 4.757961783439491,
"grad_norm": 0.5944816781846355,
"learning_rate": 5.507239164130501e-06,
"loss": 0.1131,
"step": 249
},
{
"epoch": 4.777070063694268,
"grad_norm": 0.5951359908540822,
"learning_rate": 5.421616730394e-06,
"loss": 0.098,
"step": 250
},
{
"epoch": 4.796178343949045,
"grad_norm": 0.582629331373758,
"learning_rate": 5.336416880653461e-06,
"loss": 0.1258,
"step": 251
},
{
"epoch": 4.8152866242038215,
"grad_norm": 0.5834525291911234,
"learning_rate": 5.251647478840511e-06,
"loss": 0.1048,
"step": 252
},
{
"epoch": 4.834394904458598,
"grad_norm": 0.5991362378090112,
"learning_rate": 5.167316349156495e-06,
"loss": 0.1324,
"step": 253
},
{
"epoch": 4.853503184713376,
"grad_norm": 0.5588826595586873,
"learning_rate": 5.083431275350312e-06,
"loss": 0.0959,
"step": 254
},
{
"epoch": 4.872611464968153,
"grad_norm": 0.6020289968727252,
"learning_rate": 5.000000000000003e-06,
"loss": 0.1195,
"step": 255
},
{
"epoch": 4.89171974522293,
"grad_norm": 0.5775428372320027,
"learning_rate": 4.917030223798057e-06,
"loss": 0.1317,
"step": 256
},
{
"epoch": 4.9108280254777075,
"grad_norm": 0.5517699900430338,
"learning_rate": 4.834529604840686e-06,
"loss": 0.1062,
"step": 257
},
{
"epoch": 4.929936305732484,
"grad_norm": 0.5583818338051684,
"learning_rate": 4.7525057579209775e-06,
"loss": 0.1264,
"step": 258
},
{
"epoch": 4.949044585987261,
"grad_norm": 0.5703988666724562,
"learning_rate": 4.670966253826027e-06,
"loss": 0.0929,
"step": 259
},
{
"epoch": 4.968152866242038,
"grad_norm": 0.5587580672285489,
"learning_rate": 4.589918618638173e-06,
"loss": 0.1487,
"step": 260
},
{
"epoch": 4.987261146496815,
"grad_norm": 0.5754442184055244,
"learning_rate": 4.5093703330403385e-06,
"loss": 0.1104,
"step": 261
},
{
"epoch": 5.006369426751593,
"grad_norm": 0.47604933067213157,
"learning_rate": 4.429328831625565e-06,
"loss": 0.1053,
"step": 262
},
{
"epoch": 5.025477707006369,
"grad_norm": 0.44769491277310514,
"learning_rate": 4.349801502210801e-06,
"loss": 0.0756,
"step": 263
},
{
"epoch": 5.044585987261146,
"grad_norm": 0.43083222800739923,
"learning_rate": 4.270795685155001e-06,
"loss": 0.0857,
"step": 264
},
{
"epoch": 5.063694267515924,
"grad_norm": 0.4001233947597807,
"learning_rate": 4.192318672681631e-06,
"loss": 0.056,
"step": 265
},
{
"epoch": 5.082802547770701,
"grad_norm": 0.3941889030720719,
"learning_rate": 4.1143777082055715e-06,
"loss": 0.0757,
"step": 266
},
{
"epoch": 5.101910828025478,
"grad_norm": 0.4069434130632119,
"learning_rate": 4.036979985664566e-06,
"loss": 0.0881,
"step": 267
},
{
"epoch": 5.1210191082802545,
"grad_norm": 0.4402669609666307,
"learning_rate": 3.960132648855226e-06,
"loss": 0.0756,
"step": 268
},
{
"epoch": 5.140127388535032,
"grad_norm": 0.44049414089918293,
"learning_rate": 3.883842790773647e-06,
"loss": 0.0915,
"step": 269
},
{
"epoch": 5.159235668789809,
"grad_norm": 0.5617927406070873,
"learning_rate": 3.8081174529607346e-06,
"loss": 0.109,
"step": 270
},
{
"epoch": 5.178343949044586,
"grad_norm": 0.4723379246794441,
"learning_rate": 3.732963624852275e-06,
"loss": 0.0806,
"step": 271
},
{
"epoch": 5.197452229299363,
"grad_norm": 0.43295612565720193,
"learning_rate": 3.6583882431338047e-06,
"loss": 0.0627,
"step": 272
},
{
"epoch": 5.2165605095541405,
"grad_norm": 0.45553745226026043,
"learning_rate": 3.584398191100341e-06,
"loss": 0.1009,
"step": 273
},
{
"epoch": 5.235668789808917,
"grad_norm": 0.40108690481095655,
"learning_rate": 3.511000298021098e-06,
"loss": 0.0714,
"step": 274
},
{
"epoch": 5.254777070063694,
"grad_norm": 0.40423387388292076,
"learning_rate": 3.4382013385090985e-06,
"loss": 0.0777,
"step": 275
},
{
"epoch": 5.273885350318471,
"grad_norm": 0.36188868968156396,
"learning_rate": 3.3660080318959043e-06,
"loss": 0.0823,
"step": 276
},
{
"epoch": 5.292993630573249,
"grad_norm": 0.4276164374502514,
"learning_rate": 3.2944270416114256e-06,
"loss": 0.0817,
"step": 277
},
{
"epoch": 5.312101910828026,
"grad_norm": 0.4244618276711872,
"learning_rate": 3.223464974568874e-06,
"loss": 0.1013,
"step": 278
},
{
"epoch": 5.3312101910828025,
"grad_norm": 0.40711935753435863,
"learning_rate": 3.153128380554941e-06,
"loss": 0.085,
"step": 279
},
{
"epoch": 5.350318471337579,
"grad_norm": 0.4102731718512158,
"learning_rate": 3.0834237516252817e-06,
"loss": 0.126,
"step": 280
},
{
"epoch": 5.369426751592357,
"grad_norm": 0.3500596427441456,
"learning_rate": 3.0143575215052732e-06,
"loss": 0.0813,
"step": 281
},
{
"epoch": 5.388535031847134,
"grad_norm": 0.38548735206001494,
"learning_rate": 2.94593606499619e-06,
"loss": 0.1018,
"step": 282
},
{
"epoch": 5.407643312101911,
"grad_norm": 0.3755708662176069,
"learning_rate": 2.878165697386812e-06,
"loss": 0.0732,
"step": 283
},
{
"epoch": 5.426751592356688,
"grad_norm": 0.38027052274952294,
"learning_rate": 2.8110526738705345e-06,
"loss": 0.0651,
"step": 284
},
{
"epoch": 5.445859872611465,
"grad_norm": 0.40256402037943106,
"learning_rate": 2.7446031889679893e-06,
"loss": 0.0821,
"step": 285
},
{
"epoch": 5.464968152866242,
"grad_norm": 0.4191726432928864,
"learning_rate": 2.678823375955314e-06,
"loss": 0.0733,
"step": 286
},
{
"epoch": 5.484076433121019,
"grad_norm": 0.40632663104786704,
"learning_rate": 2.6137193062980506e-06,
"loss": 0.0764,
"step": 287
},
{
"epoch": 5.503184713375796,
"grad_norm": 0.3816589726850424,
"learning_rate": 2.5492969890907383e-06,
"loss": 0.0792,
"step": 288
},
{
"epoch": 5.522292993630574,
"grad_norm": 0.4064168090616113,
"learning_rate": 2.485562370502279e-06,
"loss": 0.0855,
"step": 289
},
{
"epoch": 5.54140127388535,
"grad_norm": 0.3955328560796272,
"learning_rate": 2.4225213332271203e-06,
"loss": 0.0756,
"step": 290
},
{
"epoch": 5.560509554140127,
"grad_norm": 0.38641025333156304,
"learning_rate": 2.3601796959422585e-06,
"loss": 0.0862,
"step": 291
},
{
"epoch": 5.579617834394904,
"grad_norm": 0.3698307234746336,
"learning_rate": 2.2985432127701945e-06,
"loss": 0.0476,
"step": 292
},
{
"epoch": 5.598726114649682,
"grad_norm": 0.39951138087208105,
"learning_rate": 2.2376175727478346e-06,
"loss": 0.0822,
"step": 293
},
{
"epoch": 5.617834394904459,
"grad_norm": 0.39799752448346915,
"learning_rate": 2.1774083993013715e-06,
"loss": 0.1094,
"step": 294
},
{
"epoch": 5.6369426751592355,
"grad_norm": 0.3997225350107758,
"learning_rate": 2.1179212497272582e-06,
"loss": 0.0845,
"step": 295
},
{
"epoch": 5.656050955414012,
"grad_norm": 0.3865557218719786,
"learning_rate": 2.0591616146792705e-06,
"loss": 0.0951,
"step": 296
},
{
"epoch": 5.67515923566879,
"grad_norm": 0.45083097110031933,
"learning_rate": 2.0011349176617133e-06,
"loss": 0.0891,
"step": 297
},
{
"epoch": 5.694267515923567,
"grad_norm": 0.401699893254878,
"learning_rate": 1.9438465145288377e-06,
"loss": 0.0694,
"step": 298
},
{
"epoch": 5.713375796178344,
"grad_norm": 0.39056593793324884,
"learning_rate": 1.8873016929904942e-06,
"loss": 0.107,
"step": 299
},
{
"epoch": 5.732484076433121,
"grad_norm": 0.40090489846349914,
"learning_rate": 1.8315056721240831e-06,
"loss": 0.0748,
"step": 300
},
{
"epoch": 5.751592356687898,
"grad_norm": 0.35412931082743904,
"learning_rate": 1.7764636018928249e-06,
"loss": 0.0733,
"step": 301
},
{
"epoch": 5.770700636942675,
"grad_norm": 0.40398464926101857,
"learning_rate": 1.722180562670428e-06,
"loss": 0.1067,
"step": 302
},
{
"epoch": 5.789808917197452,
"grad_norm": 0.39946500505114957,
"learning_rate": 1.6686615647721638e-06,
"loss": 0.1,
"step": 303
},
{
"epoch": 5.80891719745223,
"grad_norm": 0.39758527930540954,
"learning_rate": 1.6159115479924259e-06,
"loss": 0.0833,
"step": 304
},
{
"epoch": 5.828025477707007,
"grad_norm": 0.41708635213572515,
"learning_rate": 1.5639353811487744e-06,
"loss": 0.0515,
"step": 305
},
{
"epoch": 5.8471337579617835,
"grad_norm": 0.41188326189259594,
"learning_rate": 1.5127378616325606e-06,
"loss": 0.0773,
"step": 306
},
{
"epoch": 5.86624203821656,
"grad_norm": 0.3745565856400691,
"learning_rate": 1.462323714966114e-06,
"loss": 0.0602,
"step": 307
},
{
"epoch": 5.885350318471337,
"grad_norm": 0.40852051581658766,
"learning_rate": 1.4126975943665844e-06,
"loss": 0.0717,
"step": 308
},
{
"epoch": 5.904458598726115,
"grad_norm": 0.414260810777109,
"learning_rate": 1.3638640803164516e-06,
"loss": 0.1213,
"step": 309
},
{
"epoch": 5.923566878980892,
"grad_norm": 0.3884498532672599,
"learning_rate": 1.3158276801407432e-06,
"loss": 0.0908,
"step": 310
},
{
"epoch": 5.942675159235669,
"grad_norm": 0.384071233807421,
"learning_rate": 1.2685928275910142e-06,
"loss": 0.0812,
"step": 311
},
{
"epoch": 5.961783439490446,
"grad_norm": 0.34519516758720215,
"learning_rate": 1.222163882436107e-06,
"loss": 0.0597,
"step": 312
},
{
"epoch": 5.980891719745223,
"grad_norm": 0.3977373982189989,
"learning_rate": 1.1765451300597574e-06,
"loss": 0.0913,
"step": 313
},
{
"epoch": 6.0,
"grad_norm": 0.3878546571382882,
"learning_rate": 1.1317407810650372e-06,
"loss": 0.0884,
"step": 314
},
{
"epoch": 6.019108280254777,
"grad_norm": 0.33779153179080706,
"learning_rate": 1.0877549708857228e-06,
"loss": 0.0802,
"step": 315
},
{
"epoch": 6.038216560509555,
"grad_norm": 0.3179141578914388,
"learning_rate": 1.0445917594046073e-06,
"loss": 0.0645,
"step": 316
},
{
"epoch": 6.057324840764331,
"grad_norm": 0.3204502935987288,
"learning_rate": 1.0022551305787564e-06,
"loss": 0.0641,
"step": 317
},
{
"epoch": 6.076433121019108,
"grad_norm": 0.31828040793224005,
"learning_rate": 9.607489920717983e-07,
"loss": 0.0663,
"step": 318
},
{
"epoch": 6.095541401273885,
"grad_norm": 0.3009135223782124,
"learning_rate": 9.200771748932513e-07,
"loss": 0.0811,
"step": 319
},
{
"epoch": 6.114649681528663,
"grad_norm": 0.2578678128906716,
"learning_rate": 8.802434330449128e-07,
"loss": 0.0517,
"step": 320
},
{
"epoch": 6.13375796178344,
"grad_norm": 0.2790874578821567,
"learning_rate": 8.412514431743657e-07,
"loss": 0.0605,
"step": 321
},
{
"epoch": 6.1528662420382165,
"grad_norm": 0.3087854313268563,
"learning_rate": 8.031048042356393e-07,
"loss": 0.082,
"step": 322
},
{
"epoch": 6.171974522292993,
"grad_norm": 0.2950780829352482,
"learning_rate": 7.65807037157007e-07,
"loss": 0.0686,
"step": 323
},
{
"epoch": 6.191082802547771,
"grad_norm": 0.31169370667435575,
"learning_rate": 7.293615845160196e-07,
"loss": 0.0868,
"step": 324
},
{
"epoch": 6.210191082802548,
"grad_norm": 0.2857072437843116,
"learning_rate": 6.937718102217461e-07,
"loss": 0.0545,
"step": 325
},
{
"epoch": 6.229299363057325,
"grad_norm": 0.27631097197420473,
"learning_rate": 6.590409992042957e-07,
"loss": 0.0557,
"step": 326
},
{
"epoch": 6.248407643312102,
"grad_norm": 0.29321775110156156,
"learning_rate": 6.251723571116031e-07,
"loss": 0.0757,
"step": 327
},
{
"epoch": 6.267515923566879,
"grad_norm": 0.30475215808851175,
"learning_rate": 5.921690100135713e-07,
"loss": 0.0709,
"step": 328
},
{
"epoch": 6.286624203821656,
"grad_norm": 0.2966881009482061,
"learning_rate": 5.600340041135133e-07,
"loss": 0.0615,
"step": 329
},
{
"epoch": 6.305732484076433,
"grad_norm": 0.26393910697205086,
"learning_rate": 5.287703054670012e-07,
"loss": 0.0443,
"step": 330
},
{
"epoch": 6.32484076433121,
"grad_norm": 0.3597673902205209,
"learning_rate": 4.983807997080925e-07,
"loss": 0.0909,
"step": 331
},
{
"epoch": 6.343949044585988,
"grad_norm": 0.3013327090348725,
"learning_rate": 4.6886829178299676e-07,
"loss": 0.0667,
"step": 332
},
{
"epoch": 6.3630573248407645,
"grad_norm": 0.36243106458518387,
"learning_rate": 4.402355056911656e-07,
"loss": 0.0795,
"step": 333
},
{
"epoch": 6.382165605095541,
"grad_norm": 0.32028430641355277,
"learning_rate": 4.124850842338779e-07,
"loss": 0.0622,
"step": 334
},
{
"epoch": 6.401273885350318,
"grad_norm": 0.32768566090708884,
"learning_rate": 3.8561958877030957e-07,
"loss": 0.0587,
"step": 335
},
{
"epoch": 6.420382165605096,
"grad_norm": 0.303307793064129,
"learning_rate": 3.5964149898111587e-07,
"loss": 0.0921,
"step": 336
},
{
"epoch": 6.439490445859873,
"grad_norm": 0.2997336665485515,
"learning_rate": 3.345532126395579e-07,
"loss": 0.0621,
"step": 337
},
{
"epoch": 6.45859872611465,
"grad_norm": 0.3298346447543088,
"learning_rate": 3.1035704539019384e-07,
"loss": 0.0582,
"step": 338
},
{
"epoch": 6.477707006369426,
"grad_norm": 0.3298555361153554,
"learning_rate": 2.870552305351382e-07,
"loss": 0.0647,
"step": 339
},
{
"epoch": 6.496815286624204,
"grad_norm": 0.2996129728963715,
"learning_rate": 2.646499188279328e-07,
"loss": 0.0698,
"step": 340
},
{
"epoch": 6.515923566878981,
"grad_norm": 0.33491302387099714,
"learning_rate": 2.4314317827503375e-07,
"loss": 0.0894,
"step": 341
},
{
"epoch": 6.535031847133758,
"grad_norm": 0.302758634407778,
"learning_rate": 2.2253699394493066e-07,
"loss": 0.0741,
"step": 342
},
{
"epoch": 6.554140127388535,
"grad_norm": 0.3123601731674272,
"learning_rate": 2.028332677849254e-07,
"loss": 0.092,
"step": 343
},
{
"epoch": 6.573248407643312,
"grad_norm": 0.3009510561512071,
"learning_rate": 1.840338184455881e-07,
"loss": 0.0658,
"step": 344
},
{
"epoch": 6.592356687898089,
"grad_norm": 0.2863965586517129,
"learning_rate": 1.6614038111289034e-07,
"loss": 0.0767,
"step": 345
},
{
"epoch": 6.611464968152866,
"grad_norm": 0.30456439022535475,
"learning_rate": 1.49154607348051e-07,
"loss": 0.0679,
"step": 346
},
{
"epoch": 6.630573248407643,
"grad_norm": 0.2897959688201585,
"learning_rate": 1.330780649350938e-07,
"loss": 0.0817,
"step": 347
},
{
"epoch": 6.649681528662421,
"grad_norm": 0.3133503576071545,
"learning_rate": 1.1791223773614635e-07,
"loss": 0.0904,
"step": 348
},
{
"epoch": 6.6687898089171975,
"grad_norm": 0.30830973853117616,
"learning_rate": 1.0365852555447642e-07,
"loss": 0.0689,
"step": 349
},
{
"epoch": 6.687898089171974,
"grad_norm": 0.288367823780671,
"learning_rate": 9.031824400528854e-08,
"loss": 0.0587,
"step": 350
},
{
"epoch": 6.707006369426751,
"grad_norm": 0.3021284104708251,
"learning_rate": 7.789262439430012e-08,
"loss": 0.0702,
"step": 351
},
{
"epoch": 6.726114649681529,
"grad_norm": 0.2647136363664868,
"learning_rate": 6.638281360408339e-08,
"loss": 0.0474,
"step": 352
},
{
"epoch": 6.745222929936306,
"grad_norm": 0.2796508908753172,
"learning_rate": 5.578987398821345e-08,
"loss": 0.0721,
"step": 353
},
{
"epoch": 6.764331210191083,
"grad_norm": 0.2813296893013791,
"learning_rate": 4.6114783273213395e-08,
"loss": 0.0641,
"step": 354
},
{
"epoch": 6.7834394904458595,
"grad_norm": 0.29441527204389356,
"learning_rate": 3.735843446830867e-08,
"loss": 0.0818,
"step": 355
},
{
"epoch": 6.802547770700637,
"grad_norm": 0.2948771544793749,
"learning_rate": 2.9521635783001932e-08,
"loss": 0.0676,
"step": 356
},
{
"epoch": 6.821656050955414,
"grad_norm": 0.2887568298378835,
"learning_rate": 2.2605110552477162e-08,
"loss": 0.0926,
"step": 357
},
{
"epoch": 6.840764331210191,
"grad_norm": 0.31867546960884374,
"learning_rate": 1.6609497170834154e-08,
"loss": 0.0594,
"step": 358
},
{
"epoch": 6.859872611464969,
"grad_norm": 0.2774368466710001,
"learning_rate": 1.1535349032167908e-08,
"loss": 0.0604,
"step": 359
},
{
"epoch": 6.8789808917197455,
"grad_norm": 0.24190844368060577,
"learning_rate": 7.3831344794872415e-09,
"loss": 0.0539,
"step": 360
},
{
"epoch": 6.898089171974522,
"grad_norm": 0.2839464423499535,
"learning_rate": 4.153236761488266e-09,
"loss": 0.0409,
"step": 361
},
{
"epoch": 6.917197452229299,
"grad_norm": 0.2738860294303811,
"learning_rate": 1.8459539971804608e-09,
"loss": 0.0818,
"step": 362
},
{
"epoch": 6.936305732484076,
"grad_norm": 0.2921665833615179,
"learning_rate": 4.614991483686826e-10,
"loss": 0.0689,
"step": 363
},
{
"epoch": 6.955414012738854,
"grad_norm": 0.30226270013210504,
"learning_rate": 0.0,
"loss": 0.0803,
"step": 364
},
{
"epoch": 6.955414012738854,
"step": 364,
"total_flos": 9.018872245445427e+16,
"train_loss": 0.35375573943563543,
"train_runtime": 2255.9423,
"train_samples_per_second": 15.515,
"train_steps_per_second": 0.161
}
],
"logging_steps": 1,
"max_steps": 364,
"num_input_tokens_seen": 0,
"num_train_epochs": 7,
"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": 9.018872245445427e+16,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}