internvl15_2000_ / trainer_state.json
cjfcsjt's picture
Upload folder using huggingface_hub
06b08ad verified
Raw
History Blame Contribute Delete
348 kB
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 0.2547081204132639,
"eval_steps": 500,
"global_step": 2000,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.00012735406020663196,
"grad_norm": 7.271933690036313,
"learning_rate": 4.237288135593221e-08,
"loss": 1.8539,
"step": 1
},
{
"epoch": 0.0002547081204132639,
"grad_norm": 9.658086805732816,
"learning_rate": 8.474576271186442e-08,
"loss": 2.0669,
"step": 2
},
{
"epoch": 0.0003820621806198959,
"grad_norm": 11.484017585849884,
"learning_rate": 1.2711864406779662e-07,
"loss": 2.2215,
"step": 3
},
{
"epoch": 0.0005094162408265278,
"grad_norm": 5.540318356122022,
"learning_rate": 1.6949152542372883e-07,
"loss": 1.8601,
"step": 4
},
{
"epoch": 0.0006367703010331598,
"grad_norm": 8.72124267822763,
"learning_rate": 2.1186440677966102e-07,
"loss": 2.036,
"step": 5
},
{
"epoch": 0.0007641243612397918,
"grad_norm": 5.653927826902831,
"learning_rate": 2.5423728813559323e-07,
"loss": 1.6908,
"step": 6
},
{
"epoch": 0.0008914784214464238,
"grad_norm": 6.524121350290057,
"learning_rate": 2.966101694915255e-07,
"loss": 1.8968,
"step": 7
},
{
"epoch": 0.0010188324816530557,
"grad_norm": 6.459540108765332,
"learning_rate": 3.3898305084745766e-07,
"loss": 1.7855,
"step": 8
},
{
"epoch": 0.0011461865418596878,
"grad_norm": 9.04352851601738,
"learning_rate": 3.813559322033899e-07,
"loss": 2.0275,
"step": 9
},
{
"epoch": 0.0012735406020663196,
"grad_norm": 5.189147687544124,
"learning_rate": 4.2372881355932204e-07,
"loss": 1.7905,
"step": 10
},
{
"epoch": 0.0014008946622729515,
"grad_norm": 7.8253156604412375,
"learning_rate": 4.661016949152543e-07,
"loss": 1.9746,
"step": 11
},
{
"epoch": 0.0015282487224795836,
"grad_norm": 8.244971390044737,
"learning_rate": 5.084745762711865e-07,
"loss": 1.9675,
"step": 12
},
{
"epoch": 0.0016556027826862155,
"grad_norm": 9.407911793872701,
"learning_rate": 5.508474576271187e-07,
"loss": 1.8633,
"step": 13
},
{
"epoch": 0.0017829568428928476,
"grad_norm": 7.269078375368235,
"learning_rate": 5.93220338983051e-07,
"loss": 1.9822,
"step": 14
},
{
"epoch": 0.0019103109030994794,
"grad_norm": 6.836825277174009,
"learning_rate": 6.355932203389831e-07,
"loss": 1.8575,
"step": 15
},
{
"epoch": 0.0020376649633061113,
"grad_norm": 7.472620005302849,
"learning_rate": 6.779661016949153e-07,
"loss": 1.7522,
"step": 16
},
{
"epoch": 0.002165019023512743,
"grad_norm": 5.337881334605893,
"learning_rate": 7.203389830508476e-07,
"loss": 1.623,
"step": 17
},
{
"epoch": 0.0022923730837193755,
"grad_norm": 9.373863863448666,
"learning_rate": 7.627118644067798e-07,
"loss": 1.9805,
"step": 18
},
{
"epoch": 0.0024197271439260074,
"grad_norm": 7.718845629389268,
"learning_rate": 8.050847457627118e-07,
"loss": 1.807,
"step": 19
},
{
"epoch": 0.0025470812041326393,
"grad_norm": 7.130785818146366,
"learning_rate": 8.474576271186441e-07,
"loss": 1.824,
"step": 20
},
{
"epoch": 0.002674435264339271,
"grad_norm": 6.681287500260171,
"learning_rate": 8.898305084745763e-07,
"loss": 1.8496,
"step": 21
},
{
"epoch": 0.002801789324545903,
"grad_norm": 6.872143516134369,
"learning_rate": 9.322033898305086e-07,
"loss": 1.8806,
"step": 22
},
{
"epoch": 0.0029291433847525353,
"grad_norm": 4.705998046939443,
"learning_rate": 9.745762711864408e-07,
"loss": 1.634,
"step": 23
},
{
"epoch": 0.003056497444959167,
"grad_norm": 11.506522277892683,
"learning_rate": 1.016949152542373e-06,
"loss": 2.1845,
"step": 24
},
{
"epoch": 0.003183851505165799,
"grad_norm": 11.312083874209405,
"learning_rate": 1.059322033898305e-06,
"loss": 1.8857,
"step": 25
},
{
"epoch": 0.003311205565372431,
"grad_norm": 6.4627660818311226,
"learning_rate": 1.1016949152542374e-06,
"loss": 1.8913,
"step": 26
},
{
"epoch": 0.003438559625579063,
"grad_norm": 7.32229167356663,
"learning_rate": 1.1440677966101696e-06,
"loss": 1.9071,
"step": 27
},
{
"epoch": 0.003565913685785695,
"grad_norm": 5.958980854124657,
"learning_rate": 1.186440677966102e-06,
"loss": 1.7951,
"step": 28
},
{
"epoch": 0.003693267745992327,
"grad_norm": 7.968036694484309,
"learning_rate": 1.228813559322034e-06,
"loss": 1.9683,
"step": 29
},
{
"epoch": 0.003820621806198959,
"grad_norm": 6.087713606766322,
"learning_rate": 1.2711864406779662e-06,
"loss": 1.7463,
"step": 30
},
{
"epoch": 0.003947975866405591,
"grad_norm": 7.15380657743577,
"learning_rate": 1.3135593220338985e-06,
"loss": 1.8375,
"step": 31
},
{
"epoch": 0.004075329926612223,
"grad_norm": 6.99281487234341,
"learning_rate": 1.3559322033898307e-06,
"loss": 2.0248,
"step": 32
},
{
"epoch": 0.0042026839868188545,
"grad_norm": 6.702985570920233,
"learning_rate": 1.3983050847457628e-06,
"loss": 1.9517,
"step": 33
},
{
"epoch": 0.004330038047025486,
"grad_norm": 7.265232789313097,
"learning_rate": 1.4406779661016951e-06,
"loss": 1.9653,
"step": 34
},
{
"epoch": 0.004457392107232118,
"grad_norm": 11.115946654222988,
"learning_rate": 1.4830508474576273e-06,
"loss": 2.2556,
"step": 35
},
{
"epoch": 0.004584746167438751,
"grad_norm": 7.299990917879491,
"learning_rate": 1.5254237288135596e-06,
"loss": 1.8916,
"step": 36
},
{
"epoch": 0.004712100227645383,
"grad_norm": 8.757682697242801,
"learning_rate": 1.5677966101694915e-06,
"loss": 1.9587,
"step": 37
},
{
"epoch": 0.004839454287852015,
"grad_norm": 6.31538702786833,
"learning_rate": 1.6101694915254237e-06,
"loss": 1.6702,
"step": 38
},
{
"epoch": 0.004966808348058647,
"grad_norm": 5.678276211998329,
"learning_rate": 1.652542372881356e-06,
"loss": 1.6698,
"step": 39
},
{
"epoch": 0.0050941624082652785,
"grad_norm": 6.5502751882271255,
"learning_rate": 1.6949152542372882e-06,
"loss": 1.6604,
"step": 40
},
{
"epoch": 0.00522151646847191,
"grad_norm": 7.378033030484872,
"learning_rate": 1.7372881355932205e-06,
"loss": 1.8324,
"step": 41
},
{
"epoch": 0.005348870528678542,
"grad_norm": 15.420303713013,
"learning_rate": 1.7796610169491526e-06,
"loss": 2.0715,
"step": 42
},
{
"epoch": 0.005476224588885174,
"grad_norm": 5.720408048146975,
"learning_rate": 1.8220338983050848e-06,
"loss": 1.723,
"step": 43
},
{
"epoch": 0.005603578649091806,
"grad_norm": 6.303637032995317,
"learning_rate": 1.8644067796610171e-06,
"loss": 1.7073,
"step": 44
},
{
"epoch": 0.005730932709298438,
"grad_norm": 6.278397704866944,
"learning_rate": 1.9067796610169493e-06,
"loss": 1.6899,
"step": 45
},
{
"epoch": 0.005858286769505071,
"grad_norm": 9.400638363853234,
"learning_rate": 1.9491525423728816e-06,
"loss": 2.0536,
"step": 46
},
{
"epoch": 0.0059856408297117025,
"grad_norm": 11.772339058952785,
"learning_rate": 1.9915254237288137e-06,
"loss": 1.9004,
"step": 47
},
{
"epoch": 0.006112994889918334,
"grad_norm": 6.097740079901016,
"learning_rate": 2.033898305084746e-06,
"loss": 1.7205,
"step": 48
},
{
"epoch": 0.006240348950124966,
"grad_norm": 7.00755338498105,
"learning_rate": 2.076271186440678e-06,
"loss": 1.8041,
"step": 49
},
{
"epoch": 0.006367703010331598,
"grad_norm": 6.9732164515856,
"learning_rate": 2.11864406779661e-06,
"loss": 1.7358,
"step": 50
},
{
"epoch": 0.00649505707053823,
"grad_norm": 6.583437987191887,
"learning_rate": 2.1610169491525427e-06,
"loss": 1.7373,
"step": 51
},
{
"epoch": 0.006622411130744862,
"grad_norm": 8.700856431922421,
"learning_rate": 2.203389830508475e-06,
"loss": 1.8176,
"step": 52
},
{
"epoch": 0.006749765190951494,
"grad_norm": 6.11722055087655,
"learning_rate": 2.245762711864407e-06,
"loss": 1.6658,
"step": 53
},
{
"epoch": 0.006877119251158126,
"grad_norm": 6.397543745767114,
"learning_rate": 2.288135593220339e-06,
"loss": 1.761,
"step": 54
},
{
"epoch": 0.0070044733113647575,
"grad_norm": 9.611483508510796,
"learning_rate": 2.3305084745762712e-06,
"loss": 1.9658,
"step": 55
},
{
"epoch": 0.00713182737157139,
"grad_norm": 9.349587138643392,
"learning_rate": 2.372881355932204e-06,
"loss": 1.7335,
"step": 56
},
{
"epoch": 0.007259181431778022,
"grad_norm": 6.522174278890342,
"learning_rate": 2.415254237288136e-06,
"loss": 1.6166,
"step": 57
},
{
"epoch": 0.007386535491984654,
"grad_norm": 7.750941557755308,
"learning_rate": 2.457627118644068e-06,
"loss": 1.7544,
"step": 58
},
{
"epoch": 0.007513889552191286,
"grad_norm": 6.912297289671559,
"learning_rate": 2.5e-06,
"loss": 1.7011,
"step": 59
},
{
"epoch": 0.007641243612397918,
"grad_norm": 7.984840812457877,
"learning_rate": 2.5423728813559323e-06,
"loss": 1.6993,
"step": 60
},
{
"epoch": 0.00776859767260455,
"grad_norm": 7.121099642585329,
"learning_rate": 2.5847457627118645e-06,
"loss": 1.6092,
"step": 61
},
{
"epoch": 0.007895951732811182,
"grad_norm": 13.071363736303791,
"learning_rate": 2.627118644067797e-06,
"loss": 1.9803,
"step": 62
},
{
"epoch": 0.008023305793017814,
"grad_norm": 7.054327363108295,
"learning_rate": 2.669491525423729e-06,
"loss": 1.5912,
"step": 63
},
{
"epoch": 0.008150659853224445,
"grad_norm": 8.19110074835651,
"learning_rate": 2.7118644067796613e-06,
"loss": 1.6463,
"step": 64
},
{
"epoch": 0.008278013913431078,
"grad_norm": 7.584386759380691,
"learning_rate": 2.7542372881355934e-06,
"loss": 1.6676,
"step": 65
},
{
"epoch": 0.008405367973637709,
"grad_norm": 13.138200788018468,
"learning_rate": 2.7966101694915256e-06,
"loss": 1.8309,
"step": 66
},
{
"epoch": 0.008532722033844342,
"grad_norm": 9.75600502737869,
"learning_rate": 2.838983050847458e-06,
"loss": 1.8151,
"step": 67
},
{
"epoch": 0.008660076094050973,
"grad_norm": 9.03322022802435,
"learning_rate": 2.8813559322033903e-06,
"loss": 1.7743,
"step": 68
},
{
"epoch": 0.008787430154257606,
"grad_norm": 7.31942395137626,
"learning_rate": 2.9237288135593224e-06,
"loss": 1.4732,
"step": 69
},
{
"epoch": 0.008914784214464237,
"grad_norm": 7.29859110156687,
"learning_rate": 2.9661016949152545e-06,
"loss": 1.5354,
"step": 70
},
{
"epoch": 0.00904213827467087,
"grad_norm": 7.57838373430734,
"learning_rate": 3.0084745762711862e-06,
"loss": 1.6728,
"step": 71
},
{
"epoch": 0.009169492334877502,
"grad_norm": 8.627342016718561,
"learning_rate": 3.0508474576271192e-06,
"loss": 1.5038,
"step": 72
},
{
"epoch": 0.009296846395084133,
"grad_norm": 11.103801289476134,
"learning_rate": 3.0932203389830514e-06,
"loss": 1.6997,
"step": 73
},
{
"epoch": 0.009424200455290766,
"grad_norm": 7.695175185164474,
"learning_rate": 3.135593220338983e-06,
"loss": 1.6034,
"step": 74
},
{
"epoch": 0.009551554515497397,
"grad_norm": 7.476937502840767,
"learning_rate": 3.1779661016949152e-06,
"loss": 1.7081,
"step": 75
},
{
"epoch": 0.00967890857570403,
"grad_norm": 8.36338889468324,
"learning_rate": 3.2203389830508473e-06,
"loss": 1.7367,
"step": 76
},
{
"epoch": 0.00980626263591066,
"grad_norm": 7.377402480026078,
"learning_rate": 3.26271186440678e-06,
"loss": 1.5538,
"step": 77
},
{
"epoch": 0.009933616696117293,
"grad_norm": 7.682228824336897,
"learning_rate": 3.305084745762712e-06,
"loss": 1.6121,
"step": 78
},
{
"epoch": 0.010060970756323924,
"grad_norm": 9.349692811866577,
"learning_rate": 3.347457627118644e-06,
"loss": 1.6372,
"step": 79
},
{
"epoch": 0.010188324816530557,
"grad_norm": 7.284498284211861,
"learning_rate": 3.3898305084745763e-06,
"loss": 1.6672,
"step": 80
},
{
"epoch": 0.01031567887673719,
"grad_norm": 7.870893588490932,
"learning_rate": 3.4322033898305084e-06,
"loss": 1.581,
"step": 81
},
{
"epoch": 0.01044303293694382,
"grad_norm": 7.532864852561931,
"learning_rate": 3.474576271186441e-06,
"loss": 1.5485,
"step": 82
},
{
"epoch": 0.010570386997150454,
"grad_norm": 7.315126387572679,
"learning_rate": 3.516949152542373e-06,
"loss": 1.462,
"step": 83
},
{
"epoch": 0.010697741057357085,
"grad_norm": 9.982699879040739,
"learning_rate": 3.5593220338983053e-06,
"loss": 1.6831,
"step": 84
},
{
"epoch": 0.010825095117563717,
"grad_norm": 6.745164905565429,
"learning_rate": 3.6016949152542374e-06,
"loss": 1.4625,
"step": 85
},
{
"epoch": 0.010952449177770348,
"grad_norm": 7.396636518107798,
"learning_rate": 3.6440677966101695e-06,
"loss": 1.4898,
"step": 86
},
{
"epoch": 0.011079803237976981,
"grad_norm": 6.530327106595857,
"learning_rate": 3.686440677966102e-06,
"loss": 1.4851,
"step": 87
},
{
"epoch": 0.011207157298183612,
"grad_norm": 7.280031290511072,
"learning_rate": 3.7288135593220342e-06,
"loss": 1.5776,
"step": 88
},
{
"epoch": 0.011334511358390245,
"grad_norm": 7.258807653030983,
"learning_rate": 3.7711864406779664e-06,
"loss": 1.493,
"step": 89
},
{
"epoch": 0.011461865418596876,
"grad_norm": 6.780060975917414,
"learning_rate": 3.8135593220338985e-06,
"loss": 1.497,
"step": 90
},
{
"epoch": 0.011589219478803509,
"grad_norm": 5.299820792569274,
"learning_rate": 3.8559322033898315e-06,
"loss": 1.3391,
"step": 91
},
{
"epoch": 0.011716573539010141,
"grad_norm": 11.906512079833732,
"learning_rate": 3.898305084745763e-06,
"loss": 1.7044,
"step": 92
},
{
"epoch": 0.011843927599216772,
"grad_norm": 13.750145928735572,
"learning_rate": 3.940677966101695e-06,
"loss": 1.3373,
"step": 93
},
{
"epoch": 0.011971281659423405,
"grad_norm": 6.2094515960377015,
"learning_rate": 3.9830508474576275e-06,
"loss": 1.4218,
"step": 94
},
{
"epoch": 0.012098635719630036,
"grad_norm": 8.702043175020558,
"learning_rate": 4.025423728813559e-06,
"loss": 1.5958,
"step": 95
},
{
"epoch": 0.012225989779836669,
"grad_norm": 10.530891389712178,
"learning_rate": 4.067796610169492e-06,
"loss": 1.3869,
"step": 96
},
{
"epoch": 0.0123533438400433,
"grad_norm": 8.382484394062372,
"learning_rate": 4.110169491525424e-06,
"loss": 1.4319,
"step": 97
},
{
"epoch": 0.012480697900249933,
"grad_norm": 8.65764730585386,
"learning_rate": 4.152542372881356e-06,
"loss": 1.4986,
"step": 98
},
{
"epoch": 0.012608051960456564,
"grad_norm": 6.733364941110484,
"learning_rate": 4.1949152542372886e-06,
"loss": 1.3344,
"step": 99
},
{
"epoch": 0.012735406020663196,
"grad_norm": 7.635913980134383,
"learning_rate": 4.23728813559322e-06,
"loss": 1.4161,
"step": 100
},
{
"epoch": 0.012862760080869829,
"grad_norm": 10.236577619610213,
"learning_rate": 4.279661016949153e-06,
"loss": 1.5962,
"step": 101
},
{
"epoch": 0.01299011414107646,
"grad_norm": 5.526567621635844,
"learning_rate": 4.322033898305085e-06,
"loss": 1.2936,
"step": 102
},
{
"epoch": 0.013117468201283093,
"grad_norm": 7.496201634189791,
"learning_rate": 4.364406779661017e-06,
"loss": 1.4379,
"step": 103
},
{
"epoch": 0.013244822261489724,
"grad_norm": 6.365117809706692,
"learning_rate": 4.40677966101695e-06,
"loss": 1.3969,
"step": 104
},
{
"epoch": 0.013372176321696357,
"grad_norm": 4.970383669437514,
"learning_rate": 4.449152542372881e-06,
"loss": 1.4318,
"step": 105
},
{
"epoch": 0.013499530381902988,
"grad_norm": 6.540428319781601,
"learning_rate": 4.491525423728814e-06,
"loss": 1.4254,
"step": 106
},
{
"epoch": 0.01362688444210962,
"grad_norm": 5.759607332938477,
"learning_rate": 4.5338983050847465e-06,
"loss": 1.4332,
"step": 107
},
{
"epoch": 0.013754238502316251,
"grad_norm": 8.31020834173882,
"learning_rate": 4.576271186440678e-06,
"loss": 1.4719,
"step": 108
},
{
"epoch": 0.013881592562522884,
"grad_norm": 7.340003043857536,
"learning_rate": 4.618644067796611e-06,
"loss": 1.3971,
"step": 109
},
{
"epoch": 0.014008946622729515,
"grad_norm": 6.511784361547076,
"learning_rate": 4.6610169491525425e-06,
"loss": 1.4415,
"step": 110
},
{
"epoch": 0.014136300682936148,
"grad_norm": 6.731671633748867,
"learning_rate": 4.703389830508475e-06,
"loss": 1.4395,
"step": 111
},
{
"epoch": 0.01426365474314278,
"grad_norm": 6.905671660774626,
"learning_rate": 4.745762711864408e-06,
"loss": 1.365,
"step": 112
},
{
"epoch": 0.014391008803349412,
"grad_norm": 6.430425515815662,
"learning_rate": 4.788135593220339e-06,
"loss": 1.3832,
"step": 113
},
{
"epoch": 0.014518362863556044,
"grad_norm": 7.435499419092716,
"learning_rate": 4.830508474576272e-06,
"loss": 1.4609,
"step": 114
},
{
"epoch": 0.014645716923762675,
"grad_norm": 5.4419059003708625,
"learning_rate": 4.872881355932204e-06,
"loss": 1.3687,
"step": 115
},
{
"epoch": 0.014773070983969308,
"grad_norm": 5.480813981464255,
"learning_rate": 4.915254237288136e-06,
"loss": 1.3259,
"step": 116
},
{
"epoch": 0.014900425044175939,
"grad_norm": 19.487504134260305,
"learning_rate": 4.957627118644069e-06,
"loss": 1.2768,
"step": 117
},
{
"epoch": 0.015027779104382572,
"grad_norm": 6.2148980714352895,
"learning_rate": 5e-06,
"loss": 1.3555,
"step": 118
},
{
"epoch": 0.015155133164589203,
"grad_norm": 5.838187604837424,
"learning_rate": 5.042372881355932e-06,
"loss": 1.3501,
"step": 119
},
{
"epoch": 0.015282487224795836,
"grad_norm": 5.3091089533384,
"learning_rate": 5.084745762711865e-06,
"loss": 1.3454,
"step": 120
},
{
"epoch": 0.015409841285002468,
"grad_norm": 7.12967595373887,
"learning_rate": 5.127118644067796e-06,
"loss": 1.2718,
"step": 121
},
{
"epoch": 0.0155371953452091,
"grad_norm": 5.081241156275937,
"learning_rate": 5.169491525423729e-06,
"loss": 1.306,
"step": 122
},
{
"epoch": 0.015664549405415732,
"grad_norm": 5.1678133686722605,
"learning_rate": 5.211864406779662e-06,
"loss": 1.2662,
"step": 123
},
{
"epoch": 0.015791903465622363,
"grad_norm": 6.609829212931382,
"learning_rate": 5.254237288135594e-06,
"loss": 1.4107,
"step": 124
},
{
"epoch": 0.015919257525828994,
"grad_norm": 5.390686388633695,
"learning_rate": 5.296610169491526e-06,
"loss": 1.3657,
"step": 125
},
{
"epoch": 0.01604661158603563,
"grad_norm": 4.716304726524517,
"learning_rate": 5.338983050847458e-06,
"loss": 1.285,
"step": 126
},
{
"epoch": 0.01617396564624226,
"grad_norm": 7.442946547443793,
"learning_rate": 5.38135593220339e-06,
"loss": 1.5355,
"step": 127
},
{
"epoch": 0.01630131970644889,
"grad_norm": 6.08763747167752,
"learning_rate": 5.423728813559323e-06,
"loss": 1.225,
"step": 128
},
{
"epoch": 0.01642867376665552,
"grad_norm": 7.440177884180067,
"learning_rate": 5.466101694915254e-06,
"loss": 1.4228,
"step": 129
},
{
"epoch": 0.016556027826862156,
"grad_norm": 7.716718027865346,
"learning_rate": 5.508474576271187e-06,
"loss": 1.4903,
"step": 130
},
{
"epoch": 0.016683381887068787,
"grad_norm": 4.940675318037667,
"learning_rate": 5.550847457627119e-06,
"loss": 1.3298,
"step": 131
},
{
"epoch": 0.016810735947275418,
"grad_norm": 6.172436497895839,
"learning_rate": 5.593220338983051e-06,
"loss": 1.334,
"step": 132
},
{
"epoch": 0.016938090007482053,
"grad_norm": 4.577141665939358,
"learning_rate": 5.635593220338984e-06,
"loss": 1.3172,
"step": 133
},
{
"epoch": 0.017065444067688684,
"grad_norm": 4.928191476396523,
"learning_rate": 5.677966101694916e-06,
"loss": 1.2845,
"step": 134
},
{
"epoch": 0.017192798127895315,
"grad_norm": 5.175516465809592,
"learning_rate": 5.720338983050848e-06,
"loss": 1.3261,
"step": 135
},
{
"epoch": 0.017320152188101946,
"grad_norm": 5.868259215222263,
"learning_rate": 5.7627118644067805e-06,
"loss": 1.3385,
"step": 136
},
{
"epoch": 0.01744750624830858,
"grad_norm": 6.6858498507159405,
"learning_rate": 5.805084745762712e-06,
"loss": 1.3098,
"step": 137
},
{
"epoch": 0.01757486030851521,
"grad_norm": 5.492215282646657,
"learning_rate": 5.847457627118645e-06,
"loss": 1.2635,
"step": 138
},
{
"epoch": 0.017702214368721842,
"grad_norm": 10.420122310923675,
"learning_rate": 5.8898305084745765e-06,
"loss": 1.3717,
"step": 139
},
{
"epoch": 0.017829568428928473,
"grad_norm": 5.144524576447836,
"learning_rate": 5.932203389830509e-06,
"loss": 1.2734,
"step": 140
},
{
"epoch": 0.017956922489135108,
"grad_norm": 6.264504044638567,
"learning_rate": 5.974576271186441e-06,
"loss": 1.318,
"step": 141
},
{
"epoch": 0.01808427654934174,
"grad_norm": 5.740016397060673,
"learning_rate": 6.0169491525423725e-06,
"loss": 1.3783,
"step": 142
},
{
"epoch": 0.01821163060954837,
"grad_norm": 6.251052157530912,
"learning_rate": 6.059322033898306e-06,
"loss": 1.3595,
"step": 143
},
{
"epoch": 0.018338984669755004,
"grad_norm": 5.523655392073338,
"learning_rate": 6.1016949152542385e-06,
"loss": 1.1841,
"step": 144
},
{
"epoch": 0.018466338729961635,
"grad_norm": 5.937324601644495,
"learning_rate": 6.14406779661017e-06,
"loss": 1.328,
"step": 145
},
{
"epoch": 0.018593692790168266,
"grad_norm": 5.449422459229983,
"learning_rate": 6.186440677966103e-06,
"loss": 1.3431,
"step": 146
},
{
"epoch": 0.018721046850374897,
"grad_norm": 7.690879017131419,
"learning_rate": 6.2288135593220344e-06,
"loss": 1.3203,
"step": 147
},
{
"epoch": 0.01884840091058153,
"grad_norm": 5.27418216675118,
"learning_rate": 6.271186440677966e-06,
"loss": 1.2765,
"step": 148
},
{
"epoch": 0.018975754970788163,
"grad_norm": 6.17835986202785,
"learning_rate": 6.313559322033899e-06,
"loss": 1.2689,
"step": 149
},
{
"epoch": 0.019103109030994794,
"grad_norm": 5.636045335076978,
"learning_rate": 6.3559322033898304e-06,
"loss": 1.3782,
"step": 150
},
{
"epoch": 0.019230463091201428,
"grad_norm": 5.108084699461727,
"learning_rate": 6.398305084745763e-06,
"loss": 1.3202,
"step": 151
},
{
"epoch": 0.01935781715140806,
"grad_norm": 6.525588214420939,
"learning_rate": 6.440677966101695e-06,
"loss": 1.3457,
"step": 152
},
{
"epoch": 0.01948517121161469,
"grad_norm": 5.606250966219782,
"learning_rate": 6.483050847457628e-06,
"loss": 1.3161,
"step": 153
},
{
"epoch": 0.01961252527182132,
"grad_norm": 5.588206693263131,
"learning_rate": 6.52542372881356e-06,
"loss": 1.3518,
"step": 154
},
{
"epoch": 0.019739879332027956,
"grad_norm": 5.148154081610033,
"learning_rate": 6.567796610169492e-06,
"loss": 1.2552,
"step": 155
},
{
"epoch": 0.019867233392234587,
"grad_norm": 7.284755141437708,
"learning_rate": 6.610169491525424e-06,
"loss": 1.2592,
"step": 156
},
{
"epoch": 0.019994587452441218,
"grad_norm": 4.314617134869305,
"learning_rate": 6.652542372881357e-06,
"loss": 1.1922,
"step": 157
},
{
"epoch": 0.02012194151264785,
"grad_norm": 5.301650153890583,
"learning_rate": 6.694915254237288e-06,
"loss": 1.2398,
"step": 158
},
{
"epoch": 0.020249295572854483,
"grad_norm": 4.878119228831268,
"learning_rate": 6.737288135593221e-06,
"loss": 1.3827,
"step": 159
},
{
"epoch": 0.020376649633061114,
"grad_norm": 6.819622271416848,
"learning_rate": 6.779661016949153e-06,
"loss": 1.3324,
"step": 160
},
{
"epoch": 0.020504003693267745,
"grad_norm": 4.742557919660855,
"learning_rate": 6.822033898305085e-06,
"loss": 1.3472,
"step": 161
},
{
"epoch": 0.02063135775347438,
"grad_norm": 5.689883895788964,
"learning_rate": 6.864406779661017e-06,
"loss": 1.1704,
"step": 162
},
{
"epoch": 0.02075871181368101,
"grad_norm": 6.417111302943501,
"learning_rate": 6.90677966101695e-06,
"loss": 1.3367,
"step": 163
},
{
"epoch": 0.02088606587388764,
"grad_norm": 6.942220645458798,
"learning_rate": 6.949152542372882e-06,
"loss": 1.2618,
"step": 164
},
{
"epoch": 0.021013419934094273,
"grad_norm": 5.252453185448875,
"learning_rate": 6.9915254237288146e-06,
"loss": 1.2426,
"step": 165
},
{
"epoch": 0.021140773994300907,
"grad_norm": 5.9624264758553185,
"learning_rate": 7.033898305084746e-06,
"loss": 1.514,
"step": 166
},
{
"epoch": 0.021268128054507538,
"grad_norm": 7.006920527037309,
"learning_rate": 7.076271186440679e-06,
"loss": 1.2284,
"step": 167
},
{
"epoch": 0.02139548211471417,
"grad_norm": 6.09704000229216,
"learning_rate": 7.1186440677966106e-06,
"loss": 1.3126,
"step": 168
},
{
"epoch": 0.0215228361749208,
"grad_norm": 8.247528168259292,
"learning_rate": 7.161016949152543e-06,
"loss": 1.4338,
"step": 169
},
{
"epoch": 0.021650190235127435,
"grad_norm": 6.171994905876386,
"learning_rate": 7.203389830508475e-06,
"loss": 1.2461,
"step": 170
},
{
"epoch": 0.021777544295334066,
"grad_norm": 5.131775076203093,
"learning_rate": 7.2457627118644065e-06,
"loss": 1.2865,
"step": 171
},
{
"epoch": 0.021904898355540697,
"grad_norm": 6.94815098159131,
"learning_rate": 7.288135593220339e-06,
"loss": 1.1561,
"step": 172
},
{
"epoch": 0.02203225241574733,
"grad_norm": 6.040933218097003,
"learning_rate": 7.3305084745762725e-06,
"loss": 1.2328,
"step": 173
},
{
"epoch": 0.022159606475953962,
"grad_norm": 5.685342232662009,
"learning_rate": 7.372881355932204e-06,
"loss": 1.2557,
"step": 174
},
{
"epoch": 0.022286960536160593,
"grad_norm": 6.001107431572396,
"learning_rate": 7.415254237288137e-06,
"loss": 1.2557,
"step": 175
},
{
"epoch": 0.022414314596367224,
"grad_norm": 6.392164571865962,
"learning_rate": 7.4576271186440685e-06,
"loss": 1.2892,
"step": 176
},
{
"epoch": 0.02254166865657386,
"grad_norm": 6.654915116110552,
"learning_rate": 7.500000000000001e-06,
"loss": 1.2959,
"step": 177
},
{
"epoch": 0.02266902271678049,
"grad_norm": 7.458014397932309,
"learning_rate": 7.542372881355933e-06,
"loss": 1.3309,
"step": 178
},
{
"epoch": 0.02279637677698712,
"grad_norm": 7.500898180267626,
"learning_rate": 7.5847457627118645e-06,
"loss": 1.2727,
"step": 179
},
{
"epoch": 0.02292373083719375,
"grad_norm": 5.733225361598391,
"learning_rate": 7.627118644067797e-06,
"loss": 1.3032,
"step": 180
},
{
"epoch": 0.023051084897400386,
"grad_norm": 8.154695676999397,
"learning_rate": 7.66949152542373e-06,
"loss": 1.2128,
"step": 181
},
{
"epoch": 0.023178438957607017,
"grad_norm": 6.985545902969932,
"learning_rate": 7.711864406779663e-06,
"loss": 1.3754,
"step": 182
},
{
"epoch": 0.023305793017813648,
"grad_norm": 5.1080014310280655,
"learning_rate": 7.754237288135595e-06,
"loss": 1.2624,
"step": 183
},
{
"epoch": 0.023433147078020283,
"grad_norm": 5.456029954386173,
"learning_rate": 7.796610169491526e-06,
"loss": 1.3119,
"step": 184
},
{
"epoch": 0.023560501138226914,
"grad_norm": 5.870259096838373,
"learning_rate": 7.838983050847458e-06,
"loss": 1.3431,
"step": 185
},
{
"epoch": 0.023687855198433545,
"grad_norm": 4.844651464438866,
"learning_rate": 7.88135593220339e-06,
"loss": 1.1957,
"step": 186
},
{
"epoch": 0.023815209258640176,
"grad_norm": 4.881515257364843,
"learning_rate": 7.923728813559323e-06,
"loss": 1.2141,
"step": 187
},
{
"epoch": 0.02394256331884681,
"grad_norm": 5.3909046929818265,
"learning_rate": 7.966101694915255e-06,
"loss": 1.2117,
"step": 188
},
{
"epoch": 0.02406991737905344,
"grad_norm": 5.084112680862707,
"learning_rate": 8.008474576271187e-06,
"loss": 1.2686,
"step": 189
},
{
"epoch": 0.024197271439260072,
"grad_norm": 6.606274321627709,
"learning_rate": 8.050847457627118e-06,
"loss": 1.269,
"step": 190
},
{
"epoch": 0.024324625499466707,
"grad_norm": 4.587065315960808,
"learning_rate": 8.093220338983052e-06,
"loss": 1.1426,
"step": 191
},
{
"epoch": 0.024451979559673338,
"grad_norm": 5.427227269739341,
"learning_rate": 8.135593220338983e-06,
"loss": 1.2607,
"step": 192
},
{
"epoch": 0.02457933361987997,
"grad_norm": 7.6170325004261885,
"learning_rate": 8.177966101694917e-06,
"loss": 1.3308,
"step": 193
},
{
"epoch": 0.0247066876800866,
"grad_norm": 6.3427616345396345,
"learning_rate": 8.220338983050849e-06,
"loss": 1.213,
"step": 194
},
{
"epoch": 0.024834041740293234,
"grad_norm": 7.609799602341245,
"learning_rate": 8.26271186440678e-06,
"loss": 1.2924,
"step": 195
},
{
"epoch": 0.024961395800499865,
"grad_norm": 4.2632262922802315,
"learning_rate": 8.305084745762712e-06,
"loss": 1.145,
"step": 196
},
{
"epoch": 0.025088749860706496,
"grad_norm": 5.907637175889623,
"learning_rate": 8.347457627118645e-06,
"loss": 1.1883,
"step": 197
},
{
"epoch": 0.025216103920913127,
"grad_norm": 5.60309063911008,
"learning_rate": 8.389830508474577e-06,
"loss": 1.2593,
"step": 198
},
{
"epoch": 0.02534345798111976,
"grad_norm": 5.196493969572722,
"learning_rate": 8.432203389830509e-06,
"loss": 1.0949,
"step": 199
},
{
"epoch": 0.025470812041326393,
"grad_norm": 4.282351386970627,
"learning_rate": 8.47457627118644e-06,
"loss": 1.2098,
"step": 200
},
{
"epoch": 0.025598166101533024,
"grad_norm": 7.771190590780939,
"learning_rate": 8.516949152542372e-06,
"loss": 1.3432,
"step": 201
},
{
"epoch": 0.025725520161739658,
"grad_norm": 6.684773397321345,
"learning_rate": 8.559322033898306e-06,
"loss": 1.3336,
"step": 202
},
{
"epoch": 0.02585287422194629,
"grad_norm": 7.465165686739227,
"learning_rate": 8.601694915254239e-06,
"loss": 1.1817,
"step": 203
},
{
"epoch": 0.02598022828215292,
"grad_norm": 7.086114075804347,
"learning_rate": 8.64406779661017e-06,
"loss": 1.239,
"step": 204
},
{
"epoch": 0.02610758234235955,
"grad_norm": 5.267852583691561,
"learning_rate": 8.686440677966103e-06,
"loss": 1.2166,
"step": 205
},
{
"epoch": 0.026234936402566186,
"grad_norm": 6.204765199065724,
"learning_rate": 8.728813559322034e-06,
"loss": 1.1705,
"step": 206
},
{
"epoch": 0.026362290462772817,
"grad_norm": 5.8650519374099,
"learning_rate": 8.771186440677966e-06,
"loss": 1.1771,
"step": 207
},
{
"epoch": 0.026489644522979448,
"grad_norm": 4.473193540326503,
"learning_rate": 8.8135593220339e-06,
"loss": 1.2766,
"step": 208
},
{
"epoch": 0.02661699858318608,
"grad_norm": 6.466444701222969,
"learning_rate": 8.855932203389831e-06,
"loss": 1.1482,
"step": 209
},
{
"epoch": 0.026744352643392713,
"grad_norm": 5.215909996040192,
"learning_rate": 8.898305084745763e-06,
"loss": 1.3104,
"step": 210
},
{
"epoch": 0.026871706703599344,
"grad_norm": 4.8024398880969015,
"learning_rate": 8.940677966101694e-06,
"loss": 1.0803,
"step": 211
},
{
"epoch": 0.026999060763805975,
"grad_norm": 6.601632130272579,
"learning_rate": 8.983050847457628e-06,
"loss": 1.3942,
"step": 212
},
{
"epoch": 0.02712641482401261,
"grad_norm": 5.329161065013925,
"learning_rate": 9.02542372881356e-06,
"loss": 1.2741,
"step": 213
},
{
"epoch": 0.02725376888421924,
"grad_norm": 5.287158798138966,
"learning_rate": 9.067796610169493e-06,
"loss": 1.2516,
"step": 214
},
{
"epoch": 0.02738112294442587,
"grad_norm": 4.441068262343084,
"learning_rate": 9.110169491525425e-06,
"loss": 1.1225,
"step": 215
},
{
"epoch": 0.027508477004632503,
"grad_norm": 5.781878133106383,
"learning_rate": 9.152542372881356e-06,
"loss": 1.3163,
"step": 216
},
{
"epoch": 0.027635831064839137,
"grad_norm": 6.6835361034553715,
"learning_rate": 9.194915254237288e-06,
"loss": 1.2013,
"step": 217
},
{
"epoch": 0.027763185125045768,
"grad_norm": 6.766142298984512,
"learning_rate": 9.237288135593222e-06,
"loss": 1.2307,
"step": 218
},
{
"epoch": 0.0278905391852524,
"grad_norm": 5.126568996026534,
"learning_rate": 9.279661016949153e-06,
"loss": 1.204,
"step": 219
},
{
"epoch": 0.02801789324545903,
"grad_norm": 5.481113953245,
"learning_rate": 9.322033898305085e-06,
"loss": 1.2129,
"step": 220
},
{
"epoch": 0.028145247305665665,
"grad_norm": 9.019736795243459,
"learning_rate": 9.364406779661017e-06,
"loss": 1.3009,
"step": 221
},
{
"epoch": 0.028272601365872296,
"grad_norm": 6.11214577364218,
"learning_rate": 9.40677966101695e-06,
"loss": 1.1636,
"step": 222
},
{
"epoch": 0.028399955426078927,
"grad_norm": 6.208039688573386,
"learning_rate": 9.449152542372882e-06,
"loss": 1.2205,
"step": 223
},
{
"epoch": 0.02852730948628556,
"grad_norm": 6.883913954414822,
"learning_rate": 9.491525423728815e-06,
"loss": 1.2852,
"step": 224
},
{
"epoch": 0.028654663546492192,
"grad_norm": 7.770250504823821,
"learning_rate": 9.533898305084747e-06,
"loss": 1.2271,
"step": 225
},
{
"epoch": 0.028782017606698823,
"grad_norm": 6.430792119192894,
"learning_rate": 9.576271186440679e-06,
"loss": 1.2906,
"step": 226
},
{
"epoch": 0.028909371666905454,
"grad_norm": 7.010031460461512,
"learning_rate": 9.61864406779661e-06,
"loss": 1.2835,
"step": 227
},
{
"epoch": 0.02903672572711209,
"grad_norm": 7.645972753876864,
"learning_rate": 9.661016949152544e-06,
"loss": 1.2884,
"step": 228
},
{
"epoch": 0.02916407978731872,
"grad_norm": 6.254690622657227,
"learning_rate": 9.703389830508475e-06,
"loss": 1.2195,
"step": 229
},
{
"epoch": 0.02929143384752535,
"grad_norm": 5.283445725714637,
"learning_rate": 9.745762711864407e-06,
"loss": 1.2283,
"step": 230
},
{
"epoch": 0.02941878790773198,
"grad_norm": 5.300161776682745,
"learning_rate": 9.788135593220339e-06,
"loss": 1.2168,
"step": 231
},
{
"epoch": 0.029546141967938616,
"grad_norm": 4.599103442185643,
"learning_rate": 9.830508474576272e-06,
"loss": 1.1986,
"step": 232
},
{
"epoch": 0.029673496028145247,
"grad_norm": 4.231142055650315,
"learning_rate": 9.872881355932204e-06,
"loss": 1.1642,
"step": 233
},
{
"epoch": 0.029800850088351878,
"grad_norm": 5.711332252730025,
"learning_rate": 9.915254237288137e-06,
"loss": 1.1048,
"step": 234
},
{
"epoch": 0.029928204148558513,
"grad_norm": 4.869391712213411,
"learning_rate": 9.957627118644069e-06,
"loss": 1.0453,
"step": 235
},
{
"epoch": 0.030055558208765144,
"grad_norm": 5.5126782723662675,
"learning_rate": 1e-05,
"loss": 1.2603,
"step": 236
},
{
"epoch": 0.030182912268971775,
"grad_norm": 5.210563223006909,
"learning_rate": 1.0042372881355933e-05,
"loss": 1.2728,
"step": 237
},
{
"epoch": 0.030310266329178406,
"grad_norm": 4.462683768455516,
"learning_rate": 1.0084745762711864e-05,
"loss": 1.173,
"step": 238
},
{
"epoch": 0.03043762038938504,
"grad_norm": 6.872645859376026,
"learning_rate": 1.0127118644067798e-05,
"loss": 1.2844,
"step": 239
},
{
"epoch": 0.03056497444959167,
"grad_norm": 6.347134293444063,
"learning_rate": 1.016949152542373e-05,
"loss": 1.1263,
"step": 240
},
{
"epoch": 0.030692328509798302,
"grad_norm": 4.948369483974166,
"learning_rate": 1.0211864406779661e-05,
"loss": 1.1743,
"step": 241
},
{
"epoch": 0.030819682570004937,
"grad_norm": 5.091288527678461,
"learning_rate": 1.0254237288135593e-05,
"loss": 1.1484,
"step": 242
},
{
"epoch": 0.030947036630211568,
"grad_norm": 6.830186475942115,
"learning_rate": 1.0296610169491526e-05,
"loss": 1.2874,
"step": 243
},
{
"epoch": 0.0310743906904182,
"grad_norm": 4.416479046753212,
"learning_rate": 1.0338983050847458e-05,
"loss": 1.1908,
"step": 244
},
{
"epoch": 0.03120174475062483,
"grad_norm": 6.347450116845334,
"learning_rate": 1.038135593220339e-05,
"loss": 1.3404,
"step": 245
},
{
"epoch": 0.031329098810831464,
"grad_norm": 5.174297399324874,
"learning_rate": 1.0423728813559325e-05,
"loss": 1.3433,
"step": 246
},
{
"epoch": 0.03145645287103809,
"grad_norm": 6.830511797695472,
"learning_rate": 1.0466101694915256e-05,
"loss": 1.2365,
"step": 247
},
{
"epoch": 0.031583806931244726,
"grad_norm": 4.788501700444122,
"learning_rate": 1.0508474576271188e-05,
"loss": 1.186,
"step": 248
},
{
"epoch": 0.03171116099145136,
"grad_norm": 4.786448725323033,
"learning_rate": 1.055084745762712e-05,
"loss": 1.2575,
"step": 249
},
{
"epoch": 0.03183851505165799,
"grad_norm": 5.365572446233603,
"learning_rate": 1.0593220338983052e-05,
"loss": 1.1592,
"step": 250
},
{
"epoch": 0.03196586911186462,
"grad_norm": 5.338549308461248,
"learning_rate": 1.0635593220338985e-05,
"loss": 1.1911,
"step": 251
},
{
"epoch": 0.03209322317207126,
"grad_norm": 6.401895373542065,
"learning_rate": 1.0677966101694917e-05,
"loss": 1.2061,
"step": 252
},
{
"epoch": 0.032220577232277885,
"grad_norm": 7.118990857049467,
"learning_rate": 1.0720338983050848e-05,
"loss": 1.1411,
"step": 253
},
{
"epoch": 0.03234793129248452,
"grad_norm": 5.334239505238133,
"learning_rate": 1.076271186440678e-05,
"loss": 1.1481,
"step": 254
},
{
"epoch": 0.032475285352691154,
"grad_norm": 6.033395336247018,
"learning_rate": 1.0805084745762714e-05,
"loss": 1.304,
"step": 255
},
{
"epoch": 0.03260263941289778,
"grad_norm": 5.935809085397795,
"learning_rate": 1.0847457627118645e-05,
"loss": 1.2317,
"step": 256
},
{
"epoch": 0.032729993473104416,
"grad_norm": 6.56257120729597,
"learning_rate": 1.0889830508474577e-05,
"loss": 1.0834,
"step": 257
},
{
"epoch": 0.03285734753331104,
"grad_norm": 5.52426654901273,
"learning_rate": 1.0932203389830509e-05,
"loss": 1.2457,
"step": 258
},
{
"epoch": 0.03298470159351768,
"grad_norm": 5.3082240230657876,
"learning_rate": 1.0974576271186442e-05,
"loss": 1.109,
"step": 259
},
{
"epoch": 0.03311205565372431,
"grad_norm": 7.27848219758666,
"learning_rate": 1.1016949152542374e-05,
"loss": 1.1494,
"step": 260
},
{
"epoch": 0.03323940971393094,
"grad_norm": 6.186081232346015,
"learning_rate": 1.1059322033898305e-05,
"loss": 1.1951,
"step": 261
},
{
"epoch": 0.033366763774137574,
"grad_norm": 6.09704876157918,
"learning_rate": 1.1101694915254237e-05,
"loss": 1.2232,
"step": 262
},
{
"epoch": 0.03349411783434421,
"grad_norm": 5.6139682675869,
"learning_rate": 1.1144067796610169e-05,
"loss": 1.1192,
"step": 263
},
{
"epoch": 0.033621471894550836,
"grad_norm": 5.013522744422119,
"learning_rate": 1.1186440677966102e-05,
"loss": 1.1021,
"step": 264
},
{
"epoch": 0.03374882595475747,
"grad_norm": 5.189883867849798,
"learning_rate": 1.1228813559322034e-05,
"loss": 1.0947,
"step": 265
},
{
"epoch": 0.033876180014964105,
"grad_norm": 9.693514624172337,
"learning_rate": 1.1271186440677967e-05,
"loss": 1.2365,
"step": 266
},
{
"epoch": 0.03400353407517073,
"grad_norm": 4.465752509974851,
"learning_rate": 1.13135593220339e-05,
"loss": 1.2413,
"step": 267
},
{
"epoch": 0.03413088813537737,
"grad_norm": 5.235194811033555,
"learning_rate": 1.1355932203389833e-05,
"loss": 1.161,
"step": 268
},
{
"epoch": 0.034258242195583995,
"grad_norm": 6.032168619829575,
"learning_rate": 1.1398305084745764e-05,
"loss": 1.1826,
"step": 269
},
{
"epoch": 0.03438559625579063,
"grad_norm": 6.792682345529315,
"learning_rate": 1.1440677966101696e-05,
"loss": 1.1542,
"step": 270
},
{
"epoch": 0.034512950315997264,
"grad_norm": 6.934400261192159,
"learning_rate": 1.148305084745763e-05,
"loss": 1.2888,
"step": 271
},
{
"epoch": 0.03464030437620389,
"grad_norm": 5.887182388323694,
"learning_rate": 1.1525423728813561e-05,
"loss": 1.2312,
"step": 272
},
{
"epoch": 0.034767658436410526,
"grad_norm": 6.01624388050281,
"learning_rate": 1.1567796610169493e-05,
"loss": 1.2219,
"step": 273
},
{
"epoch": 0.03489501249661716,
"grad_norm": 7.370642457700537,
"learning_rate": 1.1610169491525424e-05,
"loss": 1.2457,
"step": 274
},
{
"epoch": 0.03502236655682379,
"grad_norm": 6.054900449884194,
"learning_rate": 1.1652542372881356e-05,
"loss": 1.1599,
"step": 275
},
{
"epoch": 0.03514972061703042,
"grad_norm": 7.087314997551028,
"learning_rate": 1.169491525423729e-05,
"loss": 1.2703,
"step": 276
},
{
"epoch": 0.03527707467723706,
"grad_norm": 6.126673567100073,
"learning_rate": 1.1737288135593221e-05,
"loss": 1.1172,
"step": 277
},
{
"epoch": 0.035404428737443684,
"grad_norm": 6.316984189121061,
"learning_rate": 1.1779661016949153e-05,
"loss": 1.2458,
"step": 278
},
{
"epoch": 0.03553178279765032,
"grad_norm": 7.160140799731012,
"learning_rate": 1.1822033898305085e-05,
"loss": 1.1956,
"step": 279
},
{
"epoch": 0.035659136857856946,
"grad_norm": 6.894648823802736,
"learning_rate": 1.1864406779661018e-05,
"loss": 1.1181,
"step": 280
},
{
"epoch": 0.03578649091806358,
"grad_norm": 6.664617446028264,
"learning_rate": 1.190677966101695e-05,
"loss": 1.1402,
"step": 281
},
{
"epoch": 0.035913844978270215,
"grad_norm": 6.941991778114701,
"learning_rate": 1.1949152542372882e-05,
"loss": 1.1371,
"step": 282
},
{
"epoch": 0.03604119903847684,
"grad_norm": 4.18847827017481,
"learning_rate": 1.1991525423728813e-05,
"loss": 1.1745,
"step": 283
},
{
"epoch": 0.03616855309868348,
"grad_norm": 5.438227637703072,
"learning_rate": 1.2033898305084745e-05,
"loss": 1.2066,
"step": 284
},
{
"epoch": 0.03629590715889011,
"grad_norm": 5.776477720439709,
"learning_rate": 1.2076271186440678e-05,
"loss": 1.0364,
"step": 285
},
{
"epoch": 0.03642326121909674,
"grad_norm": 4.473684720727284,
"learning_rate": 1.2118644067796612e-05,
"loss": 1.1551,
"step": 286
},
{
"epoch": 0.036550615279303374,
"grad_norm": 5.67647515168569,
"learning_rate": 1.2161016949152544e-05,
"loss": 1.2159,
"step": 287
},
{
"epoch": 0.03667796933951001,
"grad_norm": 5.7037797943882,
"learning_rate": 1.2203389830508477e-05,
"loss": 1.2122,
"step": 288
},
{
"epoch": 0.036805323399716636,
"grad_norm": 3.9661231771896124,
"learning_rate": 1.2245762711864409e-05,
"loss": 1.0775,
"step": 289
},
{
"epoch": 0.03693267745992327,
"grad_norm": 6.025963240112854,
"learning_rate": 1.228813559322034e-05,
"loss": 1.2435,
"step": 290
},
{
"epoch": 0.0370600315201299,
"grad_norm": 6.998972953655834,
"learning_rate": 1.2330508474576272e-05,
"loss": 1.2353,
"step": 291
},
{
"epoch": 0.03718738558033653,
"grad_norm": 5.012810128147621,
"learning_rate": 1.2372881355932205e-05,
"loss": 1.1848,
"step": 292
},
{
"epoch": 0.03731473964054317,
"grad_norm": 5.8944234970891864,
"learning_rate": 1.2415254237288137e-05,
"loss": 1.1496,
"step": 293
},
{
"epoch": 0.037442093700749794,
"grad_norm": 6.154631581068166,
"learning_rate": 1.2457627118644069e-05,
"loss": 1.2476,
"step": 294
},
{
"epoch": 0.03756944776095643,
"grad_norm": 8.244605872992144,
"learning_rate": 1.25e-05,
"loss": 1.2131,
"step": 295
},
{
"epoch": 0.03769680182116306,
"grad_norm": 5.418418288060157,
"learning_rate": 1.2542372881355932e-05,
"loss": 1.1227,
"step": 296
},
{
"epoch": 0.03782415588136969,
"grad_norm": 7.34352247819092,
"learning_rate": 1.2584745762711866e-05,
"loss": 1.0944,
"step": 297
},
{
"epoch": 0.037951509941576325,
"grad_norm": 5.990661665555994,
"learning_rate": 1.2627118644067797e-05,
"loss": 1.1886,
"step": 298
},
{
"epoch": 0.03807886400178296,
"grad_norm": 5.1388355607481895,
"learning_rate": 1.2669491525423729e-05,
"loss": 1.1942,
"step": 299
},
{
"epoch": 0.03820621806198959,
"grad_norm": 5.576874237453807,
"learning_rate": 1.2711864406779661e-05,
"loss": 1.2666,
"step": 300
},
{
"epoch": 0.03833357212219622,
"grad_norm": 6.33907171920546,
"learning_rate": 1.2754237288135594e-05,
"loss": 1.1965,
"step": 301
},
{
"epoch": 0.038460926182402856,
"grad_norm": 6.752063859556231,
"learning_rate": 1.2796610169491526e-05,
"loss": 1.1443,
"step": 302
},
{
"epoch": 0.038588280242609484,
"grad_norm": 5.783912499236838,
"learning_rate": 1.2838983050847458e-05,
"loss": 1.2141,
"step": 303
},
{
"epoch": 0.03871563430281612,
"grad_norm": 9.393457985080918,
"learning_rate": 1.288135593220339e-05,
"loss": 1.227,
"step": 304
},
{
"epoch": 0.038842988363022746,
"grad_norm": 3.728983414621008,
"learning_rate": 1.2923728813559324e-05,
"loss": 1.1179,
"step": 305
},
{
"epoch": 0.03897034242322938,
"grad_norm": 6.896312623976467,
"learning_rate": 1.2966101694915256e-05,
"loss": 1.1986,
"step": 306
},
{
"epoch": 0.039097696483436015,
"grad_norm": 5.28379985943732,
"learning_rate": 1.3008474576271188e-05,
"loss": 1.0703,
"step": 307
},
{
"epoch": 0.03922505054364264,
"grad_norm": 7.615355599028728,
"learning_rate": 1.305084745762712e-05,
"loss": 1.2172,
"step": 308
},
{
"epoch": 0.03935240460384928,
"grad_norm": 5.221403988975462,
"learning_rate": 1.3093220338983053e-05,
"loss": 1.0893,
"step": 309
},
{
"epoch": 0.03947975866405591,
"grad_norm": 4.33216729737341,
"learning_rate": 1.3135593220338985e-05,
"loss": 1.0594,
"step": 310
},
{
"epoch": 0.03960711272426254,
"grad_norm": 5.546533020653022,
"learning_rate": 1.3177966101694916e-05,
"loss": 1.0849,
"step": 311
},
{
"epoch": 0.03973446678446917,
"grad_norm": 6.177659040601369,
"learning_rate": 1.3220338983050848e-05,
"loss": 1.1686,
"step": 312
},
{
"epoch": 0.03986182084467581,
"grad_norm": 6.419279516083969,
"learning_rate": 1.3262711864406782e-05,
"loss": 1.1274,
"step": 313
},
{
"epoch": 0.039989174904882435,
"grad_norm": 5.585204517851212,
"learning_rate": 1.3305084745762713e-05,
"loss": 1.1679,
"step": 314
},
{
"epoch": 0.04011652896508907,
"grad_norm": 4.364550727716378,
"learning_rate": 1.3347457627118645e-05,
"loss": 1.1287,
"step": 315
},
{
"epoch": 0.0402438830252957,
"grad_norm": 7.1135740518403,
"learning_rate": 1.3389830508474577e-05,
"loss": 1.1679,
"step": 316
},
{
"epoch": 0.04037123708550233,
"grad_norm": 8.475439603763972,
"learning_rate": 1.343220338983051e-05,
"loss": 1.1221,
"step": 317
},
{
"epoch": 0.040498591145708966,
"grad_norm": 6.599847803383956,
"learning_rate": 1.3474576271186442e-05,
"loss": 1.2375,
"step": 318
},
{
"epoch": 0.040625945205915594,
"grad_norm": 5.760783160268134,
"learning_rate": 1.3516949152542374e-05,
"loss": 1.1498,
"step": 319
},
{
"epoch": 0.04075329926612223,
"grad_norm": 5.623168816789539,
"learning_rate": 1.3559322033898305e-05,
"loss": 1.0415,
"step": 320
},
{
"epoch": 0.04088065332632886,
"grad_norm": 5.145493633882771,
"learning_rate": 1.3601694915254237e-05,
"loss": 1.1116,
"step": 321
},
{
"epoch": 0.04100800738653549,
"grad_norm": 6.643310865385602,
"learning_rate": 1.364406779661017e-05,
"loss": 1.3068,
"step": 322
},
{
"epoch": 0.041135361446742125,
"grad_norm": 5.551218314953344,
"learning_rate": 1.3686440677966102e-05,
"loss": 1.2183,
"step": 323
},
{
"epoch": 0.04126271550694876,
"grad_norm": 6.227722749099259,
"learning_rate": 1.3728813559322034e-05,
"loss": 1.281,
"step": 324
},
{
"epoch": 0.04139006956715539,
"grad_norm": 7.083253598231905,
"learning_rate": 1.3771186440677969e-05,
"loss": 1.0694,
"step": 325
},
{
"epoch": 0.04151742362736202,
"grad_norm": 5.469741818543489,
"learning_rate": 1.38135593220339e-05,
"loss": 1.2052,
"step": 326
},
{
"epoch": 0.04164477768756865,
"grad_norm": 5.973156326146339,
"learning_rate": 1.3855932203389832e-05,
"loss": 1.2121,
"step": 327
},
{
"epoch": 0.04177213174777528,
"grad_norm": 5.59625972090016,
"learning_rate": 1.3898305084745764e-05,
"loss": 1.1219,
"step": 328
},
{
"epoch": 0.04189948580798192,
"grad_norm": 7.162602299730285,
"learning_rate": 1.3940677966101697e-05,
"loss": 1.1341,
"step": 329
},
{
"epoch": 0.042026839868188545,
"grad_norm": 5.91873415790762,
"learning_rate": 1.3983050847457629e-05,
"loss": 1.1412,
"step": 330
},
{
"epoch": 0.04215419392839518,
"grad_norm": 4.9801198560393685,
"learning_rate": 1.4025423728813561e-05,
"loss": 1.0811,
"step": 331
},
{
"epoch": 0.042281547988601814,
"grad_norm": 7.04494488794065,
"learning_rate": 1.4067796610169493e-05,
"loss": 1.1825,
"step": 332
},
{
"epoch": 0.04240890204880844,
"grad_norm": 4.976812191722208,
"learning_rate": 1.4110169491525424e-05,
"loss": 1.0741,
"step": 333
},
{
"epoch": 0.042536256109015076,
"grad_norm": 5.810975233685298,
"learning_rate": 1.4152542372881358e-05,
"loss": 1.222,
"step": 334
},
{
"epoch": 0.04266361016922171,
"grad_norm": 6.413104005442228,
"learning_rate": 1.419491525423729e-05,
"loss": 1.1863,
"step": 335
},
{
"epoch": 0.04279096422942834,
"grad_norm": 5.016552230357741,
"learning_rate": 1.4237288135593221e-05,
"loss": 1.1668,
"step": 336
},
{
"epoch": 0.04291831828963497,
"grad_norm": 3.332210844506039,
"learning_rate": 1.4279661016949153e-05,
"loss": 1.1309,
"step": 337
},
{
"epoch": 0.0430456723498416,
"grad_norm": 5.386934736650668,
"learning_rate": 1.4322033898305086e-05,
"loss": 1.1814,
"step": 338
},
{
"epoch": 0.043173026410048235,
"grad_norm": 6.338924282523835,
"learning_rate": 1.4364406779661018e-05,
"loss": 1.2045,
"step": 339
},
{
"epoch": 0.04330038047025487,
"grad_norm": 7.035777079943078,
"learning_rate": 1.440677966101695e-05,
"loss": 1.1769,
"step": 340
},
{
"epoch": 0.0434277345304615,
"grad_norm": 7.6157803688276555,
"learning_rate": 1.4449152542372881e-05,
"loss": 1.1015,
"step": 341
},
{
"epoch": 0.04355508859066813,
"grad_norm": 5.101772800568265,
"learning_rate": 1.4491525423728813e-05,
"loss": 1.153,
"step": 342
},
{
"epoch": 0.043682442650874766,
"grad_norm": 6.4044347656892215,
"learning_rate": 1.4533898305084746e-05,
"loss": 1.0963,
"step": 343
},
{
"epoch": 0.04380979671108139,
"grad_norm": 7.158946233171024,
"learning_rate": 1.4576271186440678e-05,
"loss": 1.1785,
"step": 344
},
{
"epoch": 0.04393715077128803,
"grad_norm": 5.614692399116299,
"learning_rate": 1.4618644067796612e-05,
"loss": 1.1204,
"step": 345
},
{
"epoch": 0.04406450483149466,
"grad_norm": 7.491907204007989,
"learning_rate": 1.4661016949152545e-05,
"loss": 1.0839,
"step": 346
},
{
"epoch": 0.04419185889170129,
"grad_norm": 6.626849797952939,
"learning_rate": 1.4703389830508477e-05,
"loss": 1.2041,
"step": 347
},
{
"epoch": 0.044319212951907924,
"grad_norm": 6.108738078212482,
"learning_rate": 1.4745762711864408e-05,
"loss": 1.1178,
"step": 348
},
{
"epoch": 0.04444656701211455,
"grad_norm": 5.6485756082916945,
"learning_rate": 1.478813559322034e-05,
"loss": 1.0452,
"step": 349
},
{
"epoch": 0.044573921072321186,
"grad_norm": 6.370962322851192,
"learning_rate": 1.4830508474576274e-05,
"loss": 1.17,
"step": 350
},
{
"epoch": 0.04470127513252782,
"grad_norm": 5.2103942862478325,
"learning_rate": 1.4872881355932205e-05,
"loss": 1.1943,
"step": 351
},
{
"epoch": 0.04482862919273445,
"grad_norm": 5.756424673207367,
"learning_rate": 1.4915254237288137e-05,
"loss": 1.0939,
"step": 352
},
{
"epoch": 0.04495598325294108,
"grad_norm": 5.8667551178477675,
"learning_rate": 1.4957627118644069e-05,
"loss": 1.1213,
"step": 353
},
{
"epoch": 0.04508333731314772,
"grad_norm": 5.172829346899861,
"learning_rate": 1.5000000000000002e-05,
"loss": 1.079,
"step": 354
},
{
"epoch": 0.045210691373354345,
"grad_norm": 4.573603040529301,
"learning_rate": 1.5042372881355934e-05,
"loss": 1.0208,
"step": 355
},
{
"epoch": 0.04533804543356098,
"grad_norm": 4.8197230839623035,
"learning_rate": 1.5084745762711865e-05,
"loss": 1.1836,
"step": 356
},
{
"epoch": 0.045465399493767614,
"grad_norm": 7.777097030397322,
"learning_rate": 1.5127118644067797e-05,
"loss": 1.1588,
"step": 357
},
{
"epoch": 0.04559275355397424,
"grad_norm": 5.278754553281395,
"learning_rate": 1.5169491525423729e-05,
"loss": 1.0581,
"step": 358
},
{
"epoch": 0.045720107614180876,
"grad_norm": 4.858938117795735,
"learning_rate": 1.5211864406779662e-05,
"loss": 1.0556,
"step": 359
},
{
"epoch": 0.0458474616743875,
"grad_norm": 5.661545103286113,
"learning_rate": 1.5254237288135594e-05,
"loss": 1.1903,
"step": 360
},
{
"epoch": 0.04597481573459414,
"grad_norm": 7.069885982905191,
"learning_rate": 1.5296610169491526e-05,
"loss": 1.1951,
"step": 361
},
{
"epoch": 0.04610216979480077,
"grad_norm": 4.610622017492172,
"learning_rate": 1.533898305084746e-05,
"loss": 1.0968,
"step": 362
},
{
"epoch": 0.0462295238550074,
"grad_norm": 5.0935084747906005,
"learning_rate": 1.538135593220339e-05,
"loss": 1.1064,
"step": 363
},
{
"epoch": 0.046356877915214034,
"grad_norm": 4.494028155508151,
"learning_rate": 1.5423728813559326e-05,
"loss": 1.0162,
"step": 364
},
{
"epoch": 0.04648423197542067,
"grad_norm": 5.347282268412717,
"learning_rate": 1.5466101694915256e-05,
"loss": 1.0498,
"step": 365
},
{
"epoch": 0.046611586035627296,
"grad_norm": 6.447015340870528,
"learning_rate": 1.550847457627119e-05,
"loss": 1.0885,
"step": 366
},
{
"epoch": 0.04673894009583393,
"grad_norm": 5.665803544376279,
"learning_rate": 1.555084745762712e-05,
"loss": 1.0579,
"step": 367
},
{
"epoch": 0.046866294156040565,
"grad_norm": 6.497470363477434,
"learning_rate": 1.5593220338983053e-05,
"loss": 1.147,
"step": 368
},
{
"epoch": 0.04699364821624719,
"grad_norm": 6.44022082264147,
"learning_rate": 1.5635593220338986e-05,
"loss": 1.1009,
"step": 369
},
{
"epoch": 0.04712100227645383,
"grad_norm": 5.137514048462012,
"learning_rate": 1.5677966101694916e-05,
"loss": 1.0439,
"step": 370
},
{
"epoch": 0.047248356336660455,
"grad_norm": 5.702533680817622,
"learning_rate": 1.572033898305085e-05,
"loss": 1.0592,
"step": 371
},
{
"epoch": 0.04737571039686709,
"grad_norm": 5.280215190201436,
"learning_rate": 1.576271186440678e-05,
"loss": 1.1372,
"step": 372
},
{
"epoch": 0.047503064457073724,
"grad_norm": 5.3070710210883725,
"learning_rate": 1.5805084745762713e-05,
"loss": 1.1866,
"step": 373
},
{
"epoch": 0.04763041851728035,
"grad_norm": 5.790683946551362,
"learning_rate": 1.5847457627118646e-05,
"loss": 1.122,
"step": 374
},
{
"epoch": 0.047757772577486986,
"grad_norm": 4.4342977225389655,
"learning_rate": 1.5889830508474576e-05,
"loss": 1.0376,
"step": 375
},
{
"epoch": 0.04788512663769362,
"grad_norm": 7.361533850601036,
"learning_rate": 1.593220338983051e-05,
"loss": 1.1915,
"step": 376
},
{
"epoch": 0.04801248069790025,
"grad_norm": 6.792856435960524,
"learning_rate": 1.5974576271186443e-05,
"loss": 1.13,
"step": 377
},
{
"epoch": 0.04813983475810688,
"grad_norm": 5.3009787897284015,
"learning_rate": 1.6016949152542373e-05,
"loss": 1.1738,
"step": 378
},
{
"epoch": 0.04826718881831352,
"grad_norm": 7.999690049842849,
"learning_rate": 1.6059322033898307e-05,
"loss": 1.1297,
"step": 379
},
{
"epoch": 0.048394542878520144,
"grad_norm": 5.551331527032272,
"learning_rate": 1.6101694915254237e-05,
"loss": 1.2117,
"step": 380
},
{
"epoch": 0.04852189693872678,
"grad_norm": 4.79399733321872,
"learning_rate": 1.614406779661017e-05,
"loss": 1.1036,
"step": 381
},
{
"epoch": 0.04864925099893341,
"grad_norm": 5.685210049924944,
"learning_rate": 1.6186440677966104e-05,
"loss": 1.2222,
"step": 382
},
{
"epoch": 0.04877660505914004,
"grad_norm": 8.528954948278576,
"learning_rate": 1.6228813559322034e-05,
"loss": 1.1457,
"step": 383
},
{
"epoch": 0.048903959119346675,
"grad_norm": 5.27660350563353,
"learning_rate": 1.6271186440677967e-05,
"loss": 0.9738,
"step": 384
},
{
"epoch": 0.0490313131795533,
"grad_norm": 8.813854789754753,
"learning_rate": 1.63135593220339e-05,
"loss": 1.1817,
"step": 385
},
{
"epoch": 0.04915866723975994,
"grad_norm": 4.481348378738923,
"learning_rate": 1.6355932203389834e-05,
"loss": 1.0746,
"step": 386
},
{
"epoch": 0.04928602129996657,
"grad_norm": 5.467818000940252,
"learning_rate": 1.6398305084745764e-05,
"loss": 1.087,
"step": 387
},
{
"epoch": 0.0494133753601732,
"grad_norm": 6.6069025415732945,
"learning_rate": 1.6440677966101697e-05,
"loss": 1.1628,
"step": 388
},
{
"epoch": 0.049540729420379834,
"grad_norm": 5.1796771893751075,
"learning_rate": 1.648305084745763e-05,
"loss": 1.1182,
"step": 389
},
{
"epoch": 0.04966808348058647,
"grad_norm": 4.147969192537622,
"learning_rate": 1.652542372881356e-05,
"loss": 1.1918,
"step": 390
},
{
"epoch": 0.049795437540793096,
"grad_norm": 7.153712193297355,
"learning_rate": 1.6567796610169494e-05,
"loss": 1.1153,
"step": 391
},
{
"epoch": 0.04992279160099973,
"grad_norm": 6.5428372771306,
"learning_rate": 1.6610169491525424e-05,
"loss": 1.2467,
"step": 392
},
{
"epoch": 0.050050145661206365,
"grad_norm": 6.720477498664704,
"learning_rate": 1.6652542372881357e-05,
"loss": 1.1426,
"step": 393
},
{
"epoch": 0.05017749972141299,
"grad_norm": 5.490675564806118,
"learning_rate": 1.669491525423729e-05,
"loss": 1.1069,
"step": 394
},
{
"epoch": 0.05030485378161963,
"grad_norm": 4.996180028819127,
"learning_rate": 1.673728813559322e-05,
"loss": 1.1453,
"step": 395
},
{
"epoch": 0.050432207841826254,
"grad_norm": 6.543552037899987,
"learning_rate": 1.6779661016949154e-05,
"loss": 1.1048,
"step": 396
},
{
"epoch": 0.05055956190203289,
"grad_norm": 6.797227170743476,
"learning_rate": 1.6822033898305084e-05,
"loss": 1.2534,
"step": 397
},
{
"epoch": 0.05068691596223952,
"grad_norm": 6.369776960986534,
"learning_rate": 1.6864406779661018e-05,
"loss": 1.175,
"step": 398
},
{
"epoch": 0.05081427002244615,
"grad_norm": 4.493815941253846,
"learning_rate": 1.690677966101695e-05,
"loss": 1.1982,
"step": 399
},
{
"epoch": 0.050941624082652785,
"grad_norm": 4.799971882419884,
"learning_rate": 1.694915254237288e-05,
"loss": 1.1094,
"step": 400
},
{
"epoch": 0.05106897814285942,
"grad_norm": 5.5972935199009175,
"learning_rate": 1.6991525423728815e-05,
"loss": 1.1708,
"step": 401
},
{
"epoch": 0.05119633220306605,
"grad_norm": 5.317337885054691,
"learning_rate": 1.7033898305084745e-05,
"loss": 1.0627,
"step": 402
},
{
"epoch": 0.05132368626327268,
"grad_norm": 5.190269190692385,
"learning_rate": 1.7076271186440678e-05,
"loss": 1.086,
"step": 403
},
{
"epoch": 0.051451040323479316,
"grad_norm": 7.00350346766054,
"learning_rate": 1.711864406779661e-05,
"loss": 1.1158,
"step": 404
},
{
"epoch": 0.051578394383685944,
"grad_norm": 5.506921747563612,
"learning_rate": 1.7161016949152545e-05,
"loss": 1.1417,
"step": 405
},
{
"epoch": 0.05170574844389258,
"grad_norm": 5.838009549432716,
"learning_rate": 1.7203389830508478e-05,
"loss": 1.2006,
"step": 406
},
{
"epoch": 0.051833102504099206,
"grad_norm": 6.087071911683631,
"learning_rate": 1.7245762711864408e-05,
"loss": 1.2103,
"step": 407
},
{
"epoch": 0.05196045656430584,
"grad_norm": 6.958343810418135,
"learning_rate": 1.728813559322034e-05,
"loss": 1.1675,
"step": 408
},
{
"epoch": 0.052087810624512475,
"grad_norm": 6.131831193883537,
"learning_rate": 1.733050847457627e-05,
"loss": 1.1751,
"step": 409
},
{
"epoch": 0.0522151646847191,
"grad_norm": 6.429994225862797,
"learning_rate": 1.7372881355932205e-05,
"loss": 1.0883,
"step": 410
},
{
"epoch": 0.05234251874492574,
"grad_norm": 5.681284096324905,
"learning_rate": 1.741525423728814e-05,
"loss": 1.072,
"step": 411
},
{
"epoch": 0.05246987280513237,
"grad_norm": 5.152898635657918,
"learning_rate": 1.745762711864407e-05,
"loss": 1.1444,
"step": 412
},
{
"epoch": 0.052597226865339,
"grad_norm": 8.122164657024648,
"learning_rate": 1.7500000000000002e-05,
"loss": 1.117,
"step": 413
},
{
"epoch": 0.05272458092554563,
"grad_norm": 7.388012285107703,
"learning_rate": 1.7542372881355932e-05,
"loss": 1.1604,
"step": 414
},
{
"epoch": 0.05285193498575227,
"grad_norm": 4.142600455576895,
"learning_rate": 1.7584745762711865e-05,
"loss": 1.0567,
"step": 415
},
{
"epoch": 0.052979289045958895,
"grad_norm": 5.341916132123986,
"learning_rate": 1.76271186440678e-05,
"loss": 1.1664,
"step": 416
},
{
"epoch": 0.05310664310616553,
"grad_norm": 4.098930284467975,
"learning_rate": 1.766949152542373e-05,
"loss": 1.0655,
"step": 417
},
{
"epoch": 0.05323399716637216,
"grad_norm": 5.283165939679673,
"learning_rate": 1.7711864406779662e-05,
"loss": 1.0878,
"step": 418
},
{
"epoch": 0.05336135122657879,
"grad_norm": 5.72671571034816,
"learning_rate": 1.7754237288135596e-05,
"loss": 1.1076,
"step": 419
},
{
"epoch": 0.053488705286785426,
"grad_norm": 4.181638800501704,
"learning_rate": 1.7796610169491526e-05,
"loss": 1.0477,
"step": 420
},
{
"epoch": 0.053616059346992054,
"grad_norm": 6.037518500068706,
"learning_rate": 1.783898305084746e-05,
"loss": 1.0744,
"step": 421
},
{
"epoch": 0.05374341340719869,
"grad_norm": 5.908386812743721,
"learning_rate": 1.788135593220339e-05,
"loss": 1.2553,
"step": 422
},
{
"epoch": 0.05387076746740532,
"grad_norm": 5.477098120428443,
"learning_rate": 1.7923728813559326e-05,
"loss": 1.1578,
"step": 423
},
{
"epoch": 0.05399812152761195,
"grad_norm": 5.269725074907827,
"learning_rate": 1.7966101694915256e-05,
"loss": 1.0921,
"step": 424
},
{
"epoch": 0.054125475587818585,
"grad_norm": 5.99473658806239,
"learning_rate": 1.800847457627119e-05,
"loss": 1.1249,
"step": 425
},
{
"epoch": 0.05425282964802522,
"grad_norm": 5.212141226810042,
"learning_rate": 1.805084745762712e-05,
"loss": 1.0942,
"step": 426
},
{
"epoch": 0.05438018370823185,
"grad_norm": 4.914321965340521,
"learning_rate": 1.8093220338983053e-05,
"loss": 1.0524,
"step": 427
},
{
"epoch": 0.05450753776843848,
"grad_norm": 4.934122789640228,
"learning_rate": 1.8135593220338986e-05,
"loss": 1.0692,
"step": 428
},
{
"epoch": 0.05463489182864511,
"grad_norm": 5.974731962823762,
"learning_rate": 1.8177966101694916e-05,
"loss": 1.0516,
"step": 429
},
{
"epoch": 0.05476224588885174,
"grad_norm": 5.18332734102514,
"learning_rate": 1.822033898305085e-05,
"loss": 1.1025,
"step": 430
},
{
"epoch": 0.05488959994905838,
"grad_norm": 4.61394209816082,
"learning_rate": 1.8262711864406783e-05,
"loss": 1.091,
"step": 431
},
{
"epoch": 0.055016954009265005,
"grad_norm": 5.894652591010023,
"learning_rate": 1.8305084745762713e-05,
"loss": 1.1953,
"step": 432
},
{
"epoch": 0.05514430806947164,
"grad_norm": 5.86790849951594,
"learning_rate": 1.8347457627118646e-05,
"loss": 1.0643,
"step": 433
},
{
"epoch": 0.055271662129678274,
"grad_norm": 5.628648380088236,
"learning_rate": 1.8389830508474576e-05,
"loss": 1.119,
"step": 434
},
{
"epoch": 0.0553990161898849,
"grad_norm": 4.991344207554816,
"learning_rate": 1.843220338983051e-05,
"loss": 1.0047,
"step": 435
},
{
"epoch": 0.055526370250091536,
"grad_norm": 6.211890650133721,
"learning_rate": 1.8474576271186443e-05,
"loss": 1.0869,
"step": 436
},
{
"epoch": 0.05565372431029817,
"grad_norm": 6.316762561456026,
"learning_rate": 1.8516949152542373e-05,
"loss": 1.0198,
"step": 437
},
{
"epoch": 0.0557810783705048,
"grad_norm": 4.2256016229559235,
"learning_rate": 1.8559322033898307e-05,
"loss": 1.0982,
"step": 438
},
{
"epoch": 0.05590843243071143,
"grad_norm": 7.262435245717664,
"learning_rate": 1.8601694915254237e-05,
"loss": 1.0517,
"step": 439
},
{
"epoch": 0.05603578649091806,
"grad_norm": 5.585281184126876,
"learning_rate": 1.864406779661017e-05,
"loss": 1.2128,
"step": 440
},
{
"epoch": 0.056163140551124695,
"grad_norm": 4.767034803637561,
"learning_rate": 1.8686440677966103e-05,
"loss": 1.1603,
"step": 441
},
{
"epoch": 0.05629049461133133,
"grad_norm": 6.468924552294705,
"learning_rate": 1.8728813559322033e-05,
"loss": 1.2014,
"step": 442
},
{
"epoch": 0.05641784867153796,
"grad_norm": 9.31974357153982,
"learning_rate": 1.877118644067797e-05,
"loss": 1.0906,
"step": 443
},
{
"epoch": 0.05654520273174459,
"grad_norm": 6.012844007647524,
"learning_rate": 1.88135593220339e-05,
"loss": 1.1387,
"step": 444
},
{
"epoch": 0.056672556791951226,
"grad_norm": 4.083626146262829,
"learning_rate": 1.8855932203389834e-05,
"loss": 1.1401,
"step": 445
},
{
"epoch": 0.05679991085215785,
"grad_norm": 3.9173329104567634,
"learning_rate": 1.8898305084745764e-05,
"loss": 1.0576,
"step": 446
},
{
"epoch": 0.05692726491236449,
"grad_norm": 5.714794354280716,
"learning_rate": 1.8940677966101697e-05,
"loss": 0.975,
"step": 447
},
{
"epoch": 0.05705461897257112,
"grad_norm": 7.531880079322617,
"learning_rate": 1.898305084745763e-05,
"loss": 1.1463,
"step": 448
},
{
"epoch": 0.05718197303277775,
"grad_norm": 4.742736684054098,
"learning_rate": 1.902542372881356e-05,
"loss": 0.9861,
"step": 449
},
{
"epoch": 0.057309327092984384,
"grad_norm": 4.8112771598542885,
"learning_rate": 1.9067796610169494e-05,
"loss": 1.1611,
"step": 450
},
{
"epoch": 0.05743668115319101,
"grad_norm": 6.124219883654539,
"learning_rate": 1.9110169491525424e-05,
"loss": 1.1772,
"step": 451
},
{
"epoch": 0.057564035213397646,
"grad_norm": 6.490103891067207,
"learning_rate": 1.9152542372881357e-05,
"loss": 1.1467,
"step": 452
},
{
"epoch": 0.05769138927360428,
"grad_norm": 4.884413799929866,
"learning_rate": 1.919491525423729e-05,
"loss": 1.0313,
"step": 453
},
{
"epoch": 0.05781874333381091,
"grad_norm": 17.31884120667972,
"learning_rate": 1.923728813559322e-05,
"loss": 1.0885,
"step": 454
},
{
"epoch": 0.05794609739401754,
"grad_norm": 6.044814909695786,
"learning_rate": 1.9279661016949154e-05,
"loss": 1.0058,
"step": 455
},
{
"epoch": 0.05807345145422418,
"grad_norm": 6.115362876923583,
"learning_rate": 1.9322033898305087e-05,
"loss": 1.1625,
"step": 456
},
{
"epoch": 0.058200805514430805,
"grad_norm": 7.167009685498636,
"learning_rate": 1.9364406779661017e-05,
"loss": 1.2125,
"step": 457
},
{
"epoch": 0.05832815957463744,
"grad_norm": 8.321018070875693,
"learning_rate": 1.940677966101695e-05,
"loss": 1.2028,
"step": 458
},
{
"epoch": 0.058455513634844074,
"grad_norm": 5.883357171000457,
"learning_rate": 1.944915254237288e-05,
"loss": 1.0797,
"step": 459
},
{
"epoch": 0.0585828676950507,
"grad_norm": 6.548593372992014,
"learning_rate": 1.9491525423728814e-05,
"loss": 1.2657,
"step": 460
},
{
"epoch": 0.058710221755257336,
"grad_norm": 6.503332017630883,
"learning_rate": 1.9533898305084748e-05,
"loss": 1.1852,
"step": 461
},
{
"epoch": 0.05883757581546396,
"grad_norm": 5.406696984660371,
"learning_rate": 1.9576271186440678e-05,
"loss": 1.0883,
"step": 462
},
{
"epoch": 0.0589649298756706,
"grad_norm": 6.705726669953014,
"learning_rate": 1.961864406779661e-05,
"loss": 1.0506,
"step": 463
},
{
"epoch": 0.05909228393587723,
"grad_norm": 4.629241235900978,
"learning_rate": 1.9661016949152545e-05,
"loss": 1.0315,
"step": 464
},
{
"epoch": 0.05921963799608386,
"grad_norm": 5.078040676884035,
"learning_rate": 1.9703389830508478e-05,
"loss": 1.1109,
"step": 465
},
{
"epoch": 0.059346992056290494,
"grad_norm": 7.199572698841122,
"learning_rate": 1.9745762711864408e-05,
"loss": 1.0686,
"step": 466
},
{
"epoch": 0.05947434611649713,
"grad_norm": 6.04655483295243,
"learning_rate": 1.978813559322034e-05,
"loss": 1.2287,
"step": 467
},
{
"epoch": 0.059601700176703756,
"grad_norm": 4.20413933182076,
"learning_rate": 1.9830508474576275e-05,
"loss": 1.1357,
"step": 468
},
{
"epoch": 0.05972905423691039,
"grad_norm": 5.884048636001412,
"learning_rate": 1.9872881355932205e-05,
"loss": 1.0665,
"step": 469
},
{
"epoch": 0.059856408297117025,
"grad_norm": 4.659764544387501,
"learning_rate": 1.9915254237288138e-05,
"loss": 1.0815,
"step": 470
},
{
"epoch": 0.05998376235732365,
"grad_norm": 5.519786788305731,
"learning_rate": 1.9957627118644068e-05,
"loss": 1.0704,
"step": 471
},
{
"epoch": 0.06011111641753029,
"grad_norm": 8.08682390597387,
"learning_rate": 2e-05,
"loss": 1.0464,
"step": 472
},
{
"epoch": 0.06023847047773692,
"grad_norm": 5.6181193336206166,
"learning_rate": 1.9999999787305685e-05,
"loss": 1.0311,
"step": 473
},
{
"epoch": 0.06036582453794355,
"grad_norm": 4.519360327229284,
"learning_rate": 1.9999999149222745e-05,
"loss": 1.0972,
"step": 474
},
{
"epoch": 0.060493178598150184,
"grad_norm": 6.846898856876654,
"learning_rate": 1.9999998085751203e-05,
"loss": 1.1147,
"step": 475
},
{
"epoch": 0.06062053265835681,
"grad_norm": 5.986633352740627,
"learning_rate": 1.9999996596891108e-05,
"loss": 1.1468,
"step": 476
},
{
"epoch": 0.060747886718563446,
"grad_norm": 4.984324679225318,
"learning_rate": 1.999999468264253e-05,
"loss": 1.1052,
"step": 477
},
{
"epoch": 0.06087524077877008,
"grad_norm": 7.075737893813151,
"learning_rate": 1.999999234300554e-05,
"loss": 1.0461,
"step": 478
},
{
"epoch": 0.06100259483897671,
"grad_norm": 5.73308430212171,
"learning_rate": 1.9999989577980245e-05,
"loss": 1.2228,
"step": 479
},
{
"epoch": 0.06112994889918334,
"grad_norm": 6.053575376693396,
"learning_rate": 1.9999986387566755e-05,
"loss": 1.2758,
"step": 480
},
{
"epoch": 0.06125730295938998,
"grad_norm": 6.078084832772162,
"learning_rate": 1.9999982771765212e-05,
"loss": 1.1119,
"step": 481
},
{
"epoch": 0.061384657019596604,
"grad_norm": 6.797968774241789,
"learning_rate": 1.9999978730575768e-05,
"loss": 1.0882,
"step": 482
},
{
"epoch": 0.06151201107980324,
"grad_norm": 6.273452093694109,
"learning_rate": 1.999997426399859e-05,
"loss": 1.1447,
"step": 483
},
{
"epoch": 0.06163936514000987,
"grad_norm": 7.10284083310267,
"learning_rate": 1.999996937203388e-05,
"loss": 0.9969,
"step": 484
},
{
"epoch": 0.0617667192002165,
"grad_norm": 5.530254727061982,
"learning_rate": 1.9999964054681835e-05,
"loss": 1.1789,
"step": 485
},
{
"epoch": 0.061894073260423135,
"grad_norm": 4.669864489858105,
"learning_rate": 1.9999958311942685e-05,
"loss": 1.1693,
"step": 486
},
{
"epoch": 0.06202142732062976,
"grad_norm": 5.757957922497859,
"learning_rate": 1.9999952143816677e-05,
"loss": 1.1089,
"step": 487
},
{
"epoch": 0.0621487813808364,
"grad_norm": 7.478914502779971,
"learning_rate": 1.999994555030407e-05,
"loss": 1.1695,
"step": 488
},
{
"epoch": 0.06227613544104303,
"grad_norm": 7.5854044450973825,
"learning_rate": 1.9999938531405142e-05,
"loss": 1.1188,
"step": 489
},
{
"epoch": 0.06240348950124966,
"grad_norm": 7.252243747122389,
"learning_rate": 1.9999931087120198e-05,
"loss": 1.1036,
"step": 490
},
{
"epoch": 0.06253084356145629,
"grad_norm": 5.098055652157508,
"learning_rate": 1.999992321744955e-05,
"loss": 1.1222,
"step": 491
},
{
"epoch": 0.06265819762166293,
"grad_norm": 5.617857743175649,
"learning_rate": 1.9999914922393536e-05,
"loss": 1.0715,
"step": 492
},
{
"epoch": 0.06278555168186956,
"grad_norm": 4.719171568373844,
"learning_rate": 1.9999906201952507e-05,
"loss": 1.0764,
"step": 493
},
{
"epoch": 0.06291290574207618,
"grad_norm": 6.104534602441233,
"learning_rate": 1.9999897056126832e-05,
"loss": 1.2059,
"step": 494
},
{
"epoch": 0.06304025980228282,
"grad_norm": 6.722891165636998,
"learning_rate": 1.9999887484916902e-05,
"loss": 1.1001,
"step": 495
},
{
"epoch": 0.06316761386248945,
"grad_norm": 5.75820105806048,
"learning_rate": 1.999987748832313e-05,
"loss": 1.0382,
"step": 496
},
{
"epoch": 0.06329496792269608,
"grad_norm": 5.701985786262386,
"learning_rate": 1.9999867066345927e-05,
"loss": 1.0952,
"step": 497
},
{
"epoch": 0.06342232198290272,
"grad_norm": 4.50835236964317,
"learning_rate": 1.9999856218985753e-05,
"loss": 1.0516,
"step": 498
},
{
"epoch": 0.06354967604310935,
"grad_norm": 6.243899305236716,
"learning_rate": 1.9999844946243055e-05,
"loss": 0.9709,
"step": 499
},
{
"epoch": 0.06367703010331598,
"grad_norm": 5.025162325082904,
"learning_rate": 1.999983324811832e-05,
"loss": 1.0088,
"step": 500
},
{
"epoch": 0.06380438416352262,
"grad_norm": 4.394774081832498,
"learning_rate": 1.9999821124612047e-05,
"loss": 1.0979,
"step": 501
},
{
"epoch": 0.06393173822372925,
"grad_norm": 5.236599122276942,
"learning_rate": 1.9999808575724747e-05,
"loss": 1.1553,
"step": 502
},
{
"epoch": 0.06405909228393587,
"grad_norm": 5.490313235937539,
"learning_rate": 1.9999795601456955e-05,
"loss": 1.0991,
"step": 503
},
{
"epoch": 0.06418644634414251,
"grad_norm": 7.443317350150323,
"learning_rate": 1.9999782201809227e-05,
"loss": 1.029,
"step": 504
},
{
"epoch": 0.06431380040434914,
"grad_norm": 6.019639298288074,
"learning_rate": 1.9999768376782126e-05,
"loss": 1.0034,
"step": 505
},
{
"epoch": 0.06444115446455577,
"grad_norm": 5.986425302535064,
"learning_rate": 1.9999754126376247e-05,
"loss": 1.2177,
"step": 506
},
{
"epoch": 0.06456850852476241,
"grad_norm": 6.693771701353711,
"learning_rate": 1.999973945059219e-05,
"loss": 1.1051,
"step": 507
},
{
"epoch": 0.06469586258496904,
"grad_norm": 5.945756812918701,
"learning_rate": 1.9999724349430588e-05,
"loss": 1.1274,
"step": 508
},
{
"epoch": 0.06482321664517567,
"grad_norm": 5.769953067264716,
"learning_rate": 1.9999708822892074e-05,
"loss": 1.1091,
"step": 509
},
{
"epoch": 0.06495057070538231,
"grad_norm": 5.927645566674118,
"learning_rate": 1.999969287097731e-05,
"loss": 1.229,
"step": 510
},
{
"epoch": 0.06507792476558893,
"grad_norm": 5.822752084151968,
"learning_rate": 1.9999676493686982e-05,
"loss": 1.068,
"step": 511
},
{
"epoch": 0.06520527882579556,
"grad_norm": 8.400925321930224,
"learning_rate": 1.9999659691021783e-05,
"loss": 1.1987,
"step": 512
},
{
"epoch": 0.0653326328860022,
"grad_norm": 3.869302190712713,
"learning_rate": 1.999964246298242e-05,
"loss": 1.0111,
"step": 513
},
{
"epoch": 0.06545998694620883,
"grad_norm": 8.700792421229472,
"learning_rate": 1.9999624809569635e-05,
"loss": 1.0577,
"step": 514
},
{
"epoch": 0.06558734100641546,
"grad_norm": 5.010001669634086,
"learning_rate": 1.9999606730784178e-05,
"loss": 1.1465,
"step": 515
},
{
"epoch": 0.06571469506662209,
"grad_norm": 8.145420177808317,
"learning_rate": 1.9999588226626814e-05,
"loss": 1.1327,
"step": 516
},
{
"epoch": 0.06584204912682873,
"grad_norm": 6.193503193792947,
"learning_rate": 1.9999569297098334e-05,
"loss": 1.144,
"step": 517
},
{
"epoch": 0.06596940318703536,
"grad_norm": 7.465247190743016,
"learning_rate": 1.999954994219954e-05,
"loss": 1.1173,
"step": 518
},
{
"epoch": 0.06609675724724198,
"grad_norm": 6.750553814809614,
"learning_rate": 1.999953016193126e-05,
"loss": 1.1759,
"step": 519
},
{
"epoch": 0.06622411130744862,
"grad_norm": 4.564339045028179,
"learning_rate": 1.9999509956294324e-05,
"loss": 1.0987,
"step": 520
},
{
"epoch": 0.06635146536765525,
"grad_norm": 5.428131802762721,
"learning_rate": 1.9999489325289607e-05,
"loss": 1.124,
"step": 521
},
{
"epoch": 0.06647881942786188,
"grad_norm": 7.130879168351056,
"learning_rate": 1.9999468268917978e-05,
"loss": 1.008,
"step": 522
},
{
"epoch": 0.06660617348806852,
"grad_norm": 6.866731977923751,
"learning_rate": 1.9999446787180338e-05,
"loss": 1.1246,
"step": 523
},
{
"epoch": 0.06673352754827515,
"grad_norm": 5.66268234801421,
"learning_rate": 1.9999424880077592e-05,
"loss": 1.1667,
"step": 524
},
{
"epoch": 0.06686088160848178,
"grad_norm": 6.135632375693495,
"learning_rate": 1.999940254761068e-05,
"loss": 1.1069,
"step": 525
},
{
"epoch": 0.06698823566868842,
"grad_norm": 6.9654472975314015,
"learning_rate": 1.9999379789780543e-05,
"loss": 1.2112,
"step": 526
},
{
"epoch": 0.06711558972889504,
"grad_norm": 4.0917355771810735,
"learning_rate": 1.999935660658816e-05,
"loss": 1.0041,
"step": 527
},
{
"epoch": 0.06724294378910167,
"grad_norm": 5.299854262165641,
"learning_rate": 1.9999332998034515e-05,
"loss": 1.1039,
"step": 528
},
{
"epoch": 0.06737029784930831,
"grad_norm": 5.03714288624225,
"learning_rate": 1.9999308964120604e-05,
"loss": 1.0356,
"step": 529
},
{
"epoch": 0.06749765190951494,
"grad_norm": 5.018361328580709,
"learning_rate": 1.999928450484746e-05,
"loss": 1.0958,
"step": 530
},
{
"epoch": 0.06762500596972157,
"grad_norm": 4.259050045925872,
"learning_rate": 1.9999259620216113e-05,
"loss": 1.0695,
"step": 531
},
{
"epoch": 0.06775236002992821,
"grad_norm": 4.36079760893954,
"learning_rate": 1.999923431022763e-05,
"loss": 1.1662,
"step": 532
},
{
"epoch": 0.06787971409013484,
"grad_norm": 5.849711758285983,
"learning_rate": 1.9999208574883086e-05,
"loss": 1.1017,
"step": 533
},
{
"epoch": 0.06800706815034147,
"grad_norm": 4.771071826465309,
"learning_rate": 1.999918241418357e-05,
"loss": 1.0827,
"step": 534
},
{
"epoch": 0.0681344222105481,
"grad_norm": 4.811774062854366,
"learning_rate": 1.99991558281302e-05,
"loss": 1.0242,
"step": 535
},
{
"epoch": 0.06826177627075473,
"grad_norm": 6.02176026421573,
"learning_rate": 1.999912881672411e-05,
"loss": 1.0808,
"step": 536
},
{
"epoch": 0.06838913033096136,
"grad_norm": 5.9725923802994405,
"learning_rate": 1.999910137996644e-05,
"loss": 1.0689,
"step": 537
},
{
"epoch": 0.06851648439116799,
"grad_norm": 4.913192016936453,
"learning_rate": 1.9999073517858365e-05,
"loss": 1.015,
"step": 538
},
{
"epoch": 0.06864383845137463,
"grad_norm": 5.977263286120293,
"learning_rate": 1.999904523040107e-05,
"loss": 1.1264,
"step": 539
},
{
"epoch": 0.06877119251158126,
"grad_norm": 6.274222774929446,
"learning_rate": 1.9999016517595752e-05,
"loss": 1.0752,
"step": 540
},
{
"epoch": 0.06889854657178789,
"grad_norm": 5.376708180660854,
"learning_rate": 1.999898737944364e-05,
"loss": 1.0579,
"step": 541
},
{
"epoch": 0.06902590063199453,
"grad_norm": 5.978380991215184,
"learning_rate": 1.9998957815945962e-05,
"loss": 0.9844,
"step": 542
},
{
"epoch": 0.06915325469220115,
"grad_norm": 5.110581375165403,
"learning_rate": 1.999892782710399e-05,
"loss": 1.1507,
"step": 543
},
{
"epoch": 0.06928060875240778,
"grad_norm": 5.5490173500932345,
"learning_rate": 1.999889741291899e-05,
"loss": 1.1175,
"step": 544
},
{
"epoch": 0.06940796281261442,
"grad_norm": 6.375743448626605,
"learning_rate": 1.9998866573392265e-05,
"loss": 1.1339,
"step": 545
},
{
"epoch": 0.06953531687282105,
"grad_norm": 5.936784560617764,
"learning_rate": 1.9998835308525115e-05,
"loss": 0.9726,
"step": 546
},
{
"epoch": 0.06966267093302768,
"grad_norm": 6.062258371473456,
"learning_rate": 1.9998803618318873e-05,
"loss": 1.0518,
"step": 547
},
{
"epoch": 0.06979002499323432,
"grad_norm": 5.482671141521196,
"learning_rate": 1.9998771502774895e-05,
"loss": 1.1076,
"step": 548
},
{
"epoch": 0.06991737905344095,
"grad_norm": 6.91431518782602,
"learning_rate": 1.9998738961894538e-05,
"loss": 1.2298,
"step": 549
},
{
"epoch": 0.07004473311364758,
"grad_norm": 8.527637205784295,
"learning_rate": 1.9998705995679195e-05,
"loss": 1.1865,
"step": 550
},
{
"epoch": 0.07017208717385422,
"grad_norm": 5.06632813192715,
"learning_rate": 1.999867260413026e-05,
"loss": 1.1329,
"step": 551
},
{
"epoch": 0.07029944123406084,
"grad_norm": 6.611392459528193,
"learning_rate": 1.9998638787249158e-05,
"loss": 1.023,
"step": 552
},
{
"epoch": 0.07042679529426747,
"grad_norm": 4.434471493854659,
"learning_rate": 1.9998604545037325e-05,
"loss": 1.0204,
"step": 553
},
{
"epoch": 0.07055414935447411,
"grad_norm": 5.885039821026563,
"learning_rate": 1.999856987749622e-05,
"loss": 1.1628,
"step": 554
},
{
"epoch": 0.07068150341468074,
"grad_norm": 7.757189707425133,
"learning_rate": 1.999853478462732e-05,
"loss": 1.0407,
"step": 555
},
{
"epoch": 0.07080885747488737,
"grad_norm": 5.097586842499232,
"learning_rate": 1.999849926643211e-05,
"loss": 1.0845,
"step": 556
},
{
"epoch": 0.07093621153509401,
"grad_norm": 6.9215922760756765,
"learning_rate": 1.999846332291211e-05,
"loss": 1.0299,
"step": 557
},
{
"epoch": 0.07106356559530064,
"grad_norm": 4.351081643511859,
"learning_rate": 1.9998426954068837e-05,
"loss": 1.035,
"step": 558
},
{
"epoch": 0.07119091965550726,
"grad_norm": 5.114464514362844,
"learning_rate": 1.9998390159903853e-05,
"loss": 1.0353,
"step": 559
},
{
"epoch": 0.07131827371571389,
"grad_norm": 7.082369241365838,
"learning_rate": 1.9998352940418713e-05,
"loss": 0.9955,
"step": 560
},
{
"epoch": 0.07144562777592053,
"grad_norm": 4.362420880931001,
"learning_rate": 1.9998315295615002e-05,
"loss": 1.1176,
"step": 561
},
{
"epoch": 0.07157298183612716,
"grad_norm": 6.0660644749385355,
"learning_rate": 1.999827722549432e-05,
"loss": 1.1078,
"step": 562
},
{
"epoch": 0.07170033589633379,
"grad_norm": 4.456082830731038,
"learning_rate": 1.99982387300583e-05,
"loss": 1.0612,
"step": 563
},
{
"epoch": 0.07182768995654043,
"grad_norm": 5.829329542602628,
"learning_rate": 1.9998199809308558e-05,
"loss": 1.0402,
"step": 564
},
{
"epoch": 0.07195504401674706,
"grad_norm": 5.921042673202974,
"learning_rate": 1.9998160463246762e-05,
"loss": 1.15,
"step": 565
},
{
"epoch": 0.07208239807695369,
"grad_norm": 5.450573868242138,
"learning_rate": 1.9998120691874586e-05,
"loss": 1.1177,
"step": 566
},
{
"epoch": 0.07220975213716033,
"grad_norm": 5.703991131790351,
"learning_rate": 1.9998080495193716e-05,
"loss": 1.1882,
"step": 567
},
{
"epoch": 0.07233710619736695,
"grad_norm": 5.68231568533917,
"learning_rate": 1.9998039873205866e-05,
"loss": 1.0749,
"step": 568
},
{
"epoch": 0.07246446025757358,
"grad_norm": 4.717088015083142,
"learning_rate": 1.999799882591277e-05,
"loss": 1.1654,
"step": 569
},
{
"epoch": 0.07259181431778022,
"grad_norm": 5.238015073517867,
"learning_rate": 1.999795735331616e-05,
"loss": 1.0809,
"step": 570
},
{
"epoch": 0.07271916837798685,
"grad_norm": 6.21417528126376,
"learning_rate": 1.9997915455417813e-05,
"loss": 1.1254,
"step": 571
},
{
"epoch": 0.07284652243819348,
"grad_norm": 4.179097870209438,
"learning_rate": 1.9997873132219502e-05,
"loss": 1.0508,
"step": 572
},
{
"epoch": 0.07297387649840012,
"grad_norm": 6.321086520222815,
"learning_rate": 1.9997830383723034e-05,
"loss": 1.1056,
"step": 573
},
{
"epoch": 0.07310123055860675,
"grad_norm": 5.323689961360213,
"learning_rate": 1.9997787209930222e-05,
"loss": 1.1261,
"step": 574
},
{
"epoch": 0.07322858461881337,
"grad_norm": 5.293979089457133,
"learning_rate": 1.999774361084291e-05,
"loss": 1.075,
"step": 575
},
{
"epoch": 0.07335593867902002,
"grad_norm": 6.901915419823858,
"learning_rate": 1.9997699586462947e-05,
"loss": 1.0958,
"step": 576
},
{
"epoch": 0.07348329273922664,
"grad_norm": 4.789539008525201,
"learning_rate": 1.9997655136792206e-05,
"loss": 1.0798,
"step": 577
},
{
"epoch": 0.07361064679943327,
"grad_norm": 6.839564328066072,
"learning_rate": 1.999761026183258e-05,
"loss": 1.0093,
"step": 578
},
{
"epoch": 0.07373800085963991,
"grad_norm": 6.244943022517674,
"learning_rate": 1.9997564961585976e-05,
"loss": 1.06,
"step": 579
},
{
"epoch": 0.07386535491984654,
"grad_norm": 3.9667931485780747,
"learning_rate": 1.999751923605432e-05,
"loss": 1.0282,
"step": 580
},
{
"epoch": 0.07399270898005317,
"grad_norm": 6.350330566452636,
"learning_rate": 1.999747308523956e-05,
"loss": 1.1528,
"step": 581
},
{
"epoch": 0.0741200630402598,
"grad_norm": 5.164516042394334,
"learning_rate": 1.999742650914366e-05,
"loss": 1.031,
"step": 582
},
{
"epoch": 0.07424741710046644,
"grad_norm": 4.923610417638436,
"learning_rate": 1.9997379507768596e-05,
"loss": 1.0159,
"step": 583
},
{
"epoch": 0.07437477116067306,
"grad_norm": 5.076996381492745,
"learning_rate": 1.9997332081116374e-05,
"loss": 1.1246,
"step": 584
},
{
"epoch": 0.07450212522087969,
"grad_norm": 3.4257266867804166,
"learning_rate": 1.9997284229189008e-05,
"loss": 1.1539,
"step": 585
},
{
"epoch": 0.07462947928108633,
"grad_norm": 5.9256759966138635,
"learning_rate": 1.9997235951988533e-05,
"loss": 1.0759,
"step": 586
},
{
"epoch": 0.07475683334129296,
"grad_norm": 5.6626644961145844,
"learning_rate": 1.9997187249517004e-05,
"loss": 1.1168,
"step": 587
},
{
"epoch": 0.07488418740149959,
"grad_norm": 5.7046274374791075,
"learning_rate": 1.999713812177649e-05,
"loss": 0.9921,
"step": 588
},
{
"epoch": 0.07501154146170623,
"grad_norm": 5.994923669426167,
"learning_rate": 1.9997088568769084e-05,
"loss": 1.2166,
"step": 589
},
{
"epoch": 0.07513889552191286,
"grad_norm": 5.193736544338015,
"learning_rate": 1.9997038590496892e-05,
"loss": 0.9917,
"step": 590
},
{
"epoch": 0.07526624958211948,
"grad_norm": 4.396294351179956,
"learning_rate": 1.9996988186962044e-05,
"loss": 1.0891,
"step": 591
},
{
"epoch": 0.07539360364232613,
"grad_norm": 5.49995734458335,
"learning_rate": 1.999693735816668e-05,
"loss": 0.9993,
"step": 592
},
{
"epoch": 0.07552095770253275,
"grad_norm": 5.253398884946806,
"learning_rate": 1.9996886104112963e-05,
"loss": 1.0713,
"step": 593
},
{
"epoch": 0.07564831176273938,
"grad_norm": 4.642680011765876,
"learning_rate": 1.9996834424803074e-05,
"loss": 1.0839,
"step": 594
},
{
"epoch": 0.07577566582294602,
"grad_norm": 6.37376182350069,
"learning_rate": 1.999678232023921e-05,
"loss": 1.1269,
"step": 595
},
{
"epoch": 0.07590301988315265,
"grad_norm": 5.763360134317636,
"learning_rate": 1.9996729790423588e-05,
"loss": 1.1208,
"step": 596
},
{
"epoch": 0.07603037394335928,
"grad_norm": 5.206828680769963,
"learning_rate": 1.9996676835358443e-05,
"loss": 1.0223,
"step": 597
},
{
"epoch": 0.07615772800356592,
"grad_norm": 4.44025776020388,
"learning_rate": 1.999662345504603e-05,
"loss": 1.07,
"step": 598
},
{
"epoch": 0.07628508206377255,
"grad_norm": 4.782135127297473,
"learning_rate": 1.9996569649488614e-05,
"loss": 1.0387,
"step": 599
},
{
"epoch": 0.07641243612397917,
"grad_norm": 5.107139915137725,
"learning_rate": 1.9996515418688493e-05,
"loss": 1.1048,
"step": 600
},
{
"epoch": 0.07653979018418582,
"grad_norm": 5.9873211732412495,
"learning_rate": 1.9996460762647962e-05,
"loss": 1.0631,
"step": 601
},
{
"epoch": 0.07666714424439244,
"grad_norm": 5.043658764095534,
"learning_rate": 1.9996405681369353e-05,
"loss": 1.0388,
"step": 602
},
{
"epoch": 0.07679449830459907,
"grad_norm": 4.477337036234279,
"learning_rate": 1.999635017485501e-05,
"loss": 1.0678,
"step": 603
},
{
"epoch": 0.07692185236480571,
"grad_norm": 3.4802276151035993,
"learning_rate": 1.999629424310729e-05,
"loss": 1.0386,
"step": 604
},
{
"epoch": 0.07704920642501234,
"grad_norm": 6.354251381564723,
"learning_rate": 1.999623788612858e-05,
"loss": 1.0509,
"step": 605
},
{
"epoch": 0.07717656048521897,
"grad_norm": 7.521001783969517,
"learning_rate": 1.9996181103921268e-05,
"loss": 1.0353,
"step": 606
},
{
"epoch": 0.0773039145454256,
"grad_norm": 5.629739248359252,
"learning_rate": 1.9996123896487775e-05,
"loss": 1.0046,
"step": 607
},
{
"epoch": 0.07743126860563224,
"grad_norm": 5.746840230765029,
"learning_rate": 1.9996066263830533e-05,
"loss": 1.0223,
"step": 608
},
{
"epoch": 0.07755862266583886,
"grad_norm": 6.298985584272212,
"learning_rate": 1.999600820595199e-05,
"loss": 1.0892,
"step": 609
},
{
"epoch": 0.07768597672604549,
"grad_norm": 6.803573780037953,
"learning_rate": 1.999594972285462e-05,
"loss": 1.0924,
"step": 610
},
{
"epoch": 0.07781333078625213,
"grad_norm": 6.37358376722561,
"learning_rate": 1.9995890814540915e-05,
"loss": 1.0744,
"step": 611
},
{
"epoch": 0.07794068484645876,
"grad_norm": 9.007917100839986,
"learning_rate": 1.9995831481013376e-05,
"loss": 1.2104,
"step": 612
},
{
"epoch": 0.07806803890666539,
"grad_norm": 8.287857126219594,
"learning_rate": 1.999577172227452e-05,
"loss": 1.0484,
"step": 613
},
{
"epoch": 0.07819539296687203,
"grad_norm": 4.04299092309741,
"learning_rate": 1.9995711538326903e-05,
"loss": 0.9676,
"step": 614
},
{
"epoch": 0.07832274702707866,
"grad_norm": 6.571260544897439,
"learning_rate": 1.9995650929173075e-05,
"loss": 1.0512,
"step": 615
},
{
"epoch": 0.07845010108728528,
"grad_norm": 6.1484744156888915,
"learning_rate": 1.999558989481562e-05,
"loss": 1.1788,
"step": 616
},
{
"epoch": 0.07857745514749193,
"grad_norm": 5.316568230621563,
"learning_rate": 1.999552843525713e-05,
"loss": 1.0587,
"step": 617
},
{
"epoch": 0.07870480920769855,
"grad_norm": 4.570047484971308,
"learning_rate": 1.999546655050022e-05,
"loss": 1.0535,
"step": 618
},
{
"epoch": 0.07883216326790518,
"grad_norm": 6.1608163386830785,
"learning_rate": 1.9995404240547527e-05,
"loss": 1.0144,
"step": 619
},
{
"epoch": 0.07895951732811182,
"grad_norm": 7.7568651371924595,
"learning_rate": 1.9995341505401697e-05,
"loss": 1.0017,
"step": 620
},
{
"epoch": 0.07908687138831845,
"grad_norm": 5.0343224284140184,
"learning_rate": 1.99952783450654e-05,
"loss": 0.9866,
"step": 621
},
{
"epoch": 0.07921422544852508,
"grad_norm": 4.733892334358776,
"learning_rate": 1.999521475954132e-05,
"loss": 1.0346,
"step": 622
},
{
"epoch": 0.07934157950873172,
"grad_norm": 8.245032462200156,
"learning_rate": 1.9995150748832167e-05,
"loss": 1.0559,
"step": 623
},
{
"epoch": 0.07946893356893835,
"grad_norm": 5.288275385428034,
"learning_rate": 1.9995086312940665e-05,
"loss": 1.1287,
"step": 624
},
{
"epoch": 0.07959628762914497,
"grad_norm": 4.692547128203939,
"learning_rate": 1.9995021451869548e-05,
"loss": 1.0788,
"step": 625
},
{
"epoch": 0.07972364168935162,
"grad_norm": 5.274171136757848,
"learning_rate": 1.999495616562158e-05,
"loss": 1.0668,
"step": 626
},
{
"epoch": 0.07985099574955824,
"grad_norm": 4.503948598565,
"learning_rate": 1.9994890454199537e-05,
"loss": 0.9856,
"step": 627
},
{
"epoch": 0.07997834980976487,
"grad_norm": 4.458866099436066,
"learning_rate": 1.999482431760621e-05,
"loss": 1.1141,
"step": 628
},
{
"epoch": 0.0801057038699715,
"grad_norm": 5.021075273227246,
"learning_rate": 1.9994757755844422e-05,
"loss": 1.0495,
"step": 629
},
{
"epoch": 0.08023305793017814,
"grad_norm": 6.947131370381861,
"learning_rate": 1.9994690768916997e-05,
"loss": 0.98,
"step": 630
},
{
"epoch": 0.08036041199038477,
"grad_norm": 7.41009293017886,
"learning_rate": 1.999462335682679e-05,
"loss": 1.1349,
"step": 631
},
{
"epoch": 0.0804877660505914,
"grad_norm": 5.172513156024174,
"learning_rate": 1.999455551957666e-05,
"loss": 1.0966,
"step": 632
},
{
"epoch": 0.08061512011079804,
"grad_norm": 5.951912183195801,
"learning_rate": 1.99944872571695e-05,
"loss": 1.0624,
"step": 633
},
{
"epoch": 0.08074247417100466,
"grad_norm": 7.68449951170285,
"learning_rate": 1.9994418569608212e-05,
"loss": 0.9459,
"step": 634
},
{
"epoch": 0.08086982823121129,
"grad_norm": 5.28572353524242,
"learning_rate": 1.999434945689572e-05,
"loss": 1.1071,
"step": 635
},
{
"epoch": 0.08099718229141793,
"grad_norm": 5.390054600709794,
"learning_rate": 1.9994279919034958e-05,
"loss": 1.1041,
"step": 636
},
{
"epoch": 0.08112453635162456,
"grad_norm": 6.091162812164593,
"learning_rate": 1.999420995602889e-05,
"loss": 1.0167,
"step": 637
},
{
"epoch": 0.08125189041183119,
"grad_norm": 9.582331419829577,
"learning_rate": 1.9994139567880492e-05,
"loss": 1.1483,
"step": 638
},
{
"epoch": 0.08137924447203783,
"grad_norm": 6.486061704103913,
"learning_rate": 1.999406875459275e-05,
"loss": 1.057,
"step": 639
},
{
"epoch": 0.08150659853224446,
"grad_norm": 5.762382774236402,
"learning_rate": 1.999399751616869e-05,
"loss": 1.1079,
"step": 640
},
{
"epoch": 0.08163395259245108,
"grad_norm": 5.748421825606348,
"learning_rate": 1.999392585261133e-05,
"loss": 1.0681,
"step": 641
},
{
"epoch": 0.08176130665265773,
"grad_norm": 6.038734654142748,
"learning_rate": 1.9993853763923724e-05,
"loss": 0.9983,
"step": 642
},
{
"epoch": 0.08188866071286435,
"grad_norm": 6.696020954623414,
"learning_rate": 1.9993781250108934e-05,
"loss": 0.9574,
"step": 643
},
{
"epoch": 0.08201601477307098,
"grad_norm": 4.880728461504636,
"learning_rate": 1.9993708311170055e-05,
"loss": 1.0897,
"step": 644
},
{
"epoch": 0.08214336883327762,
"grad_norm": 6.765789355953292,
"learning_rate": 1.999363494711018e-05,
"loss": 0.9652,
"step": 645
},
{
"epoch": 0.08227072289348425,
"grad_norm": 4.847223623737666,
"learning_rate": 1.999356115793243e-05,
"loss": 1.079,
"step": 646
},
{
"epoch": 0.08239807695369088,
"grad_norm": 5.234269143927291,
"learning_rate": 1.999348694363995e-05,
"loss": 1.0237,
"step": 647
},
{
"epoch": 0.08252543101389752,
"grad_norm": 4.89183541331242,
"learning_rate": 1.9993412304235898e-05,
"loss": 1.0104,
"step": 648
},
{
"epoch": 0.08265278507410415,
"grad_norm": 5.865712718277241,
"learning_rate": 1.999333723972344e-05,
"loss": 1.0922,
"step": 649
},
{
"epoch": 0.08278013913431077,
"grad_norm": 5.789907869913807,
"learning_rate": 1.999326175010578e-05,
"loss": 1.0889,
"step": 650
},
{
"epoch": 0.0829074931945174,
"grad_norm": 6.9597588942410065,
"learning_rate": 1.9993185835386118e-05,
"loss": 1.0146,
"step": 651
},
{
"epoch": 0.08303484725472404,
"grad_norm": 5.810287218463056,
"learning_rate": 1.999310949556769e-05,
"loss": 1.0555,
"step": 652
},
{
"epoch": 0.08316220131493067,
"grad_norm": 5.914241168247319,
"learning_rate": 1.999303273065374e-05,
"loss": 1.1138,
"step": 653
},
{
"epoch": 0.0832895553751373,
"grad_norm": 7.40780816406947,
"learning_rate": 1.9992955540647544e-05,
"loss": 1.0209,
"step": 654
},
{
"epoch": 0.08341690943534394,
"grad_norm": 5.139163287743998,
"learning_rate": 1.999287792555237e-05,
"loss": 1.1916,
"step": 655
},
{
"epoch": 0.08354426349555057,
"grad_norm": 7.338151071702125,
"learning_rate": 1.999279988537153e-05,
"loss": 1.0876,
"step": 656
},
{
"epoch": 0.0836716175557572,
"grad_norm": 8.346356045647864,
"learning_rate": 1.9992721420108338e-05,
"loss": 1.0237,
"step": 657
},
{
"epoch": 0.08379897161596384,
"grad_norm": 5.194369994982586,
"learning_rate": 1.9992642529766136e-05,
"loss": 1.019,
"step": 658
},
{
"epoch": 0.08392632567617046,
"grad_norm": 6.352857844734931,
"learning_rate": 1.999256321434828e-05,
"loss": 1.0938,
"step": 659
},
{
"epoch": 0.08405367973637709,
"grad_norm": 4.991216000871772,
"learning_rate": 1.9992483473858138e-05,
"loss": 1.1457,
"step": 660
},
{
"epoch": 0.08418103379658373,
"grad_norm": 5.119276246534278,
"learning_rate": 1.9992403308299112e-05,
"loss": 1.0455,
"step": 661
},
{
"epoch": 0.08430838785679036,
"grad_norm": 6.314325956131299,
"learning_rate": 1.9992322717674603e-05,
"loss": 1.1113,
"step": 662
},
{
"epoch": 0.08443574191699699,
"grad_norm": 7.77910023602844,
"learning_rate": 1.9992241701988042e-05,
"loss": 1.0262,
"step": 663
},
{
"epoch": 0.08456309597720363,
"grad_norm": 7.1596010849287195,
"learning_rate": 1.999216026124288e-05,
"loss": 1.0519,
"step": 664
},
{
"epoch": 0.08469045003741026,
"grad_norm": 4.897077210322237,
"learning_rate": 1.9992078395442574e-05,
"loss": 1.0238,
"step": 665
},
{
"epoch": 0.08481780409761688,
"grad_norm": 6.509385595436648,
"learning_rate": 1.9991996104590612e-05,
"loss": 1.0995,
"step": 666
},
{
"epoch": 0.08494515815782352,
"grad_norm": 6.02964231578448,
"learning_rate": 1.999191338869049e-05,
"loss": 1.0622,
"step": 667
},
{
"epoch": 0.08507251221803015,
"grad_norm": 6.841055561733911,
"learning_rate": 1.9991830247745732e-05,
"loss": 1.089,
"step": 668
},
{
"epoch": 0.08519986627823678,
"grad_norm": 6.576305581782257,
"learning_rate": 1.999174668175987e-05,
"loss": 1.0683,
"step": 669
},
{
"epoch": 0.08532722033844342,
"grad_norm": 5.960560877989974,
"learning_rate": 1.9991662690736462e-05,
"loss": 1.0535,
"step": 670
},
{
"epoch": 0.08545457439865005,
"grad_norm": 6.509239086210101,
"learning_rate": 1.999157827467908e-05,
"loss": 1.0301,
"step": 671
},
{
"epoch": 0.08558192845885668,
"grad_norm": 7.124308770095535,
"learning_rate": 1.9991493433591315e-05,
"loss": 1.1226,
"step": 672
},
{
"epoch": 0.0857092825190633,
"grad_norm": 5.026968704141392,
"learning_rate": 1.9991408167476772e-05,
"loss": 1.1574,
"step": 673
},
{
"epoch": 0.08583663657926995,
"grad_norm": 4.301359298900442,
"learning_rate": 1.9991322476339088e-05,
"loss": 1.0282,
"step": 674
},
{
"epoch": 0.08596399063947657,
"grad_norm": 11.74719529462412,
"learning_rate": 1.9991236360181897e-05,
"loss": 1.0841,
"step": 675
},
{
"epoch": 0.0860913446996832,
"grad_norm": 5.71392752342204,
"learning_rate": 1.999114981900887e-05,
"loss": 1.128,
"step": 676
},
{
"epoch": 0.08621869875988984,
"grad_norm": 5.9033570647377465,
"learning_rate": 1.9991062852823683e-05,
"loss": 1.0681,
"step": 677
},
{
"epoch": 0.08634605282009647,
"grad_norm": 5.322196996403861,
"learning_rate": 1.9990975461630043e-05,
"loss": 1.0203,
"step": 678
},
{
"epoch": 0.0864734068803031,
"grad_norm": 4.951072869448979,
"learning_rate": 1.999088764543166e-05,
"loss": 1.0799,
"step": 679
},
{
"epoch": 0.08660076094050974,
"grad_norm": 5.500998839799335,
"learning_rate": 1.999079940423227e-05,
"loss": 1.0914,
"step": 680
},
{
"epoch": 0.08672811500071637,
"grad_norm": 6.443149791036705,
"learning_rate": 1.9990710738035632e-05,
"loss": 1.212,
"step": 681
},
{
"epoch": 0.086855469060923,
"grad_norm": 5.5992443869211295,
"learning_rate": 1.999062164684551e-05,
"loss": 1.1295,
"step": 682
},
{
"epoch": 0.08698282312112963,
"grad_norm": 4.401937119017067,
"learning_rate": 1.9990532130665703e-05,
"loss": 1.0117,
"step": 683
},
{
"epoch": 0.08711017718133626,
"grad_norm": 6.193840092408911,
"learning_rate": 1.9990442189500016e-05,
"loss": 1.0194,
"step": 684
},
{
"epoch": 0.08723753124154289,
"grad_norm": 7.51254012466861,
"learning_rate": 1.9990351823352268e-05,
"loss": 0.9975,
"step": 685
},
{
"epoch": 0.08736488530174953,
"grad_norm": 5.877886124116896,
"learning_rate": 1.999026103222631e-05,
"loss": 1.1254,
"step": 686
},
{
"epoch": 0.08749223936195616,
"grad_norm": 5.472095272874463,
"learning_rate": 1.9990169816126005e-05,
"loss": 1.0221,
"step": 687
},
{
"epoch": 0.08761959342216279,
"grad_norm": 4.893392833864217,
"learning_rate": 1.999007817505523e-05,
"loss": 1.0455,
"step": 688
},
{
"epoch": 0.08774694748236943,
"grad_norm": 5.637380242608783,
"learning_rate": 1.9989986109017882e-05,
"loss": 1.02,
"step": 689
},
{
"epoch": 0.08787430154257606,
"grad_norm": 4.8513515752131635,
"learning_rate": 1.9989893618017882e-05,
"loss": 1.0035,
"step": 690
},
{
"epoch": 0.08800165560278268,
"grad_norm": 6.058231699570725,
"learning_rate": 1.9989800702059158e-05,
"loss": 1.0731,
"step": 691
},
{
"epoch": 0.08812900966298932,
"grad_norm": 5.045541876584629,
"learning_rate": 1.9989707361145672e-05,
"loss": 0.9995,
"step": 692
},
{
"epoch": 0.08825636372319595,
"grad_norm": 5.568628770940815,
"learning_rate": 1.9989613595281384e-05,
"loss": 1.0222,
"step": 693
},
{
"epoch": 0.08838371778340258,
"grad_norm": 5.02853680543003,
"learning_rate": 1.998951940447029e-05,
"loss": 1.0428,
"step": 694
},
{
"epoch": 0.08851107184360922,
"grad_norm": 4.792565214522304,
"learning_rate": 1.9989424788716397e-05,
"loss": 1.0065,
"step": 695
},
{
"epoch": 0.08863842590381585,
"grad_norm": 5.291371775342626,
"learning_rate": 1.9989329748023728e-05,
"loss": 1.0209,
"step": 696
},
{
"epoch": 0.08876577996402248,
"grad_norm": 6.034196045176283,
"learning_rate": 1.998923428239632e-05,
"loss": 1.1247,
"step": 697
},
{
"epoch": 0.0888931340242291,
"grad_norm": 5.512739512804424,
"learning_rate": 1.9989138391838243e-05,
"loss": 1.0025,
"step": 698
},
{
"epoch": 0.08902048808443574,
"grad_norm": 5.877097871265012,
"learning_rate": 1.9989042076353572e-05,
"loss": 1.0448,
"step": 699
},
{
"epoch": 0.08914784214464237,
"grad_norm": 4.461166883673963,
"learning_rate": 1.9988945335946407e-05,
"loss": 1.0821,
"step": 700
},
{
"epoch": 0.089275196204849,
"grad_norm": 6.024518144895393,
"learning_rate": 1.9988848170620857e-05,
"loss": 1.1252,
"step": 701
},
{
"epoch": 0.08940255026505564,
"grad_norm": 5.810905319826364,
"learning_rate": 1.998875058038106e-05,
"loss": 1.0291,
"step": 702
},
{
"epoch": 0.08952990432526227,
"grad_norm": 4.288768987566355,
"learning_rate": 1.9988652565231167e-05,
"loss": 0.951,
"step": 703
},
{
"epoch": 0.0896572583854689,
"grad_norm": 4.429403951945096,
"learning_rate": 1.9988554125175347e-05,
"loss": 1.1162,
"step": 704
},
{
"epoch": 0.08978461244567554,
"grad_norm": 5.102004214642084,
"learning_rate": 1.998845526021779e-05,
"loss": 1.02,
"step": 705
},
{
"epoch": 0.08991196650588217,
"grad_norm": 4.930098381995267,
"learning_rate": 1.9988355970362693e-05,
"loss": 1.0211,
"step": 706
},
{
"epoch": 0.09003932056608879,
"grad_norm": 5.773187737747096,
"learning_rate": 1.9988256255614292e-05,
"loss": 1.0348,
"step": 707
},
{
"epoch": 0.09016667462629543,
"grad_norm": 5.927086944668368,
"learning_rate": 1.9988156115976818e-05,
"loss": 1.1057,
"step": 708
},
{
"epoch": 0.09029402868650206,
"grad_norm": 5.025749279734705,
"learning_rate": 1.998805555145454e-05,
"loss": 0.9713,
"step": 709
},
{
"epoch": 0.09042138274670869,
"grad_norm": 5.882972340229742,
"learning_rate": 1.9987954562051724e-05,
"loss": 1.0112,
"step": 710
},
{
"epoch": 0.09054873680691533,
"grad_norm": 5.104455101530909,
"learning_rate": 1.9987853147772677e-05,
"loss": 1.1046,
"step": 711
},
{
"epoch": 0.09067609086712196,
"grad_norm": 4.674587113206371,
"learning_rate": 1.9987751308621714e-05,
"loss": 1.0097,
"step": 712
},
{
"epoch": 0.09080344492732859,
"grad_norm": 4.381587273734659,
"learning_rate": 1.9987649044603158e-05,
"loss": 1.0142,
"step": 713
},
{
"epoch": 0.09093079898753523,
"grad_norm": 9.46538823947152,
"learning_rate": 1.9987546355721363e-05,
"loss": 1.095,
"step": 714
},
{
"epoch": 0.09105815304774185,
"grad_norm": 4.511638425962468,
"learning_rate": 1.9987443241980696e-05,
"loss": 1.1166,
"step": 715
},
{
"epoch": 0.09118550710794848,
"grad_norm": 5.012728511329524,
"learning_rate": 1.9987339703385552e-05,
"loss": 1.1406,
"step": 716
},
{
"epoch": 0.09131286116815512,
"grad_norm": 6.335841769660585,
"learning_rate": 1.9987235739940325e-05,
"loss": 1.1545,
"step": 717
},
{
"epoch": 0.09144021522836175,
"grad_norm": 4.924962410565526,
"learning_rate": 1.9987131351649437e-05,
"loss": 1.1317,
"step": 718
},
{
"epoch": 0.09156756928856838,
"grad_norm": 6.1544385074745165,
"learning_rate": 1.998702653851734e-05,
"loss": 1.1361,
"step": 719
},
{
"epoch": 0.091694923348775,
"grad_norm": 6.154575178172681,
"learning_rate": 1.998692130054848e-05,
"loss": 0.9903,
"step": 720
},
{
"epoch": 0.09182227740898165,
"grad_norm": 6.154952943747507,
"learning_rate": 1.998681563774734e-05,
"loss": 1.0783,
"step": 721
},
{
"epoch": 0.09194963146918828,
"grad_norm": 4.838842453109084,
"learning_rate": 1.9986709550118417e-05,
"loss": 1.0136,
"step": 722
},
{
"epoch": 0.0920769855293949,
"grad_norm": 5.410325823718829,
"learning_rate": 1.9986603037666218e-05,
"loss": 1.0734,
"step": 723
},
{
"epoch": 0.09220433958960154,
"grad_norm": 7.933672605467293,
"learning_rate": 1.9986496100395276e-05,
"loss": 1.0677,
"step": 724
},
{
"epoch": 0.09233169364980817,
"grad_norm": 9.510319375372314,
"learning_rate": 1.9986388738310142e-05,
"loss": 1.0391,
"step": 725
},
{
"epoch": 0.0924590477100148,
"grad_norm": 5.218726609229018,
"learning_rate": 1.998628095141538e-05,
"loss": 1.0036,
"step": 726
},
{
"epoch": 0.09258640177022144,
"grad_norm": 6.188449979354496,
"learning_rate": 1.998617273971558e-05,
"loss": 1.0897,
"step": 727
},
{
"epoch": 0.09271375583042807,
"grad_norm": 5.214680072630314,
"learning_rate": 1.998606410321534e-05,
"loss": 1.0975,
"step": 728
},
{
"epoch": 0.0928411098906347,
"grad_norm": 5.1103404588248855,
"learning_rate": 1.998595504191928e-05,
"loss": 1.115,
"step": 729
},
{
"epoch": 0.09296846395084134,
"grad_norm": 7.251754548435966,
"learning_rate": 1.9985845555832047e-05,
"loss": 0.9956,
"step": 730
},
{
"epoch": 0.09309581801104796,
"grad_norm": 6.024552653946663,
"learning_rate": 1.9985735644958292e-05,
"loss": 1.0765,
"step": 731
},
{
"epoch": 0.09322317207125459,
"grad_norm": 4.944129259919682,
"learning_rate": 1.9985625309302692e-05,
"loss": 1.1096,
"step": 732
},
{
"epoch": 0.09335052613146123,
"grad_norm": 4.15991324847806,
"learning_rate": 1.9985514548869942e-05,
"loss": 0.9459,
"step": 733
},
{
"epoch": 0.09347788019166786,
"grad_norm": 3.7197710446687253,
"learning_rate": 1.998540336366475e-05,
"loss": 0.9994,
"step": 734
},
{
"epoch": 0.09360523425187449,
"grad_norm": 6.06962528536452,
"learning_rate": 1.9985291753691853e-05,
"loss": 0.9692,
"step": 735
},
{
"epoch": 0.09373258831208113,
"grad_norm": 4.293669347546974,
"learning_rate": 1.998517971895599e-05,
"loss": 1.163,
"step": 736
},
{
"epoch": 0.09385994237228776,
"grad_norm": 4.771429610435823,
"learning_rate": 1.9985067259461936e-05,
"loss": 1.0687,
"step": 737
},
{
"epoch": 0.09398729643249439,
"grad_norm": 5.6708848998084065,
"learning_rate": 1.9984954375214464e-05,
"loss": 0.9521,
"step": 738
},
{
"epoch": 0.09411465049270103,
"grad_norm": 7.222470864474539,
"learning_rate": 1.9984841066218387e-05,
"loss": 1.0992,
"step": 739
},
{
"epoch": 0.09424200455290765,
"grad_norm": 5.365241129523414,
"learning_rate": 1.9984727332478517e-05,
"loss": 1.1041,
"step": 740
},
{
"epoch": 0.09436935861311428,
"grad_norm": 6.039438648032581,
"learning_rate": 1.9984613173999694e-05,
"loss": 1.0716,
"step": 741
},
{
"epoch": 0.09449671267332091,
"grad_norm": 4.452459345808136,
"learning_rate": 1.9984498590786778e-05,
"loss": 1.0688,
"step": 742
},
{
"epoch": 0.09462406673352755,
"grad_norm": 4.496510318180523,
"learning_rate": 1.9984383582844636e-05,
"loss": 1.1061,
"step": 743
},
{
"epoch": 0.09475142079373418,
"grad_norm": 5.133958012277289,
"learning_rate": 1.998426815017817e-05,
"loss": 1.0172,
"step": 744
},
{
"epoch": 0.0948787748539408,
"grad_norm": 6.401996670637185,
"learning_rate": 1.998415229279228e-05,
"loss": 1.0715,
"step": 745
},
{
"epoch": 0.09500612891414745,
"grad_norm": 5.80380879011085,
"learning_rate": 1.99840360106919e-05,
"loss": 0.9658,
"step": 746
},
{
"epoch": 0.09513348297435407,
"grad_norm": 5.835257004125801,
"learning_rate": 1.998391930388198e-05,
"loss": 1.1158,
"step": 747
},
{
"epoch": 0.0952608370345607,
"grad_norm": 5.1207374896936955,
"learning_rate": 1.9983802172367477e-05,
"loss": 1.0881,
"step": 748
},
{
"epoch": 0.09538819109476734,
"grad_norm": 11.487577279944423,
"learning_rate": 1.9983684616153378e-05,
"loss": 1.1162,
"step": 749
},
{
"epoch": 0.09551554515497397,
"grad_norm": 5.542628422317417,
"learning_rate": 1.998356663524468e-05,
"loss": 1.1225,
"step": 750
},
{
"epoch": 0.0956428992151806,
"grad_norm": 6.967952118088262,
"learning_rate": 1.998344822964641e-05,
"loss": 1.1577,
"step": 751
},
{
"epoch": 0.09577025327538724,
"grad_norm": 5.47302950580714,
"learning_rate": 1.99833293993636e-05,
"loss": 1.0219,
"step": 752
},
{
"epoch": 0.09589760733559387,
"grad_norm": 5.362081572740628,
"learning_rate": 1.99832101444013e-05,
"loss": 1.0726,
"step": 753
},
{
"epoch": 0.0960249613958005,
"grad_norm": 6.812421045808244,
"learning_rate": 1.9983090464764587e-05,
"loss": 1.065,
"step": 754
},
{
"epoch": 0.09615231545600714,
"grad_norm": 8.852579514610417,
"learning_rate": 1.9982970360458557e-05,
"loss": 1.1051,
"step": 755
},
{
"epoch": 0.09627966951621376,
"grad_norm": 6.265131524266463,
"learning_rate": 1.998284983148831e-05,
"loss": 1.0406,
"step": 756
},
{
"epoch": 0.09640702357642039,
"grad_norm": 10.166042829874742,
"learning_rate": 1.9982728877858978e-05,
"loss": 1.0842,
"step": 757
},
{
"epoch": 0.09653437763662703,
"grad_norm": 5.6845487546028695,
"learning_rate": 1.998260749957571e-05,
"loss": 1.0445,
"step": 758
},
{
"epoch": 0.09666173169683366,
"grad_norm": 4.165910779417907,
"learning_rate": 1.9982485696643663e-05,
"loss": 0.9373,
"step": 759
},
{
"epoch": 0.09678908575704029,
"grad_norm": 4.396198685395114,
"learning_rate": 1.998236346906802e-05,
"loss": 1.0627,
"step": 760
},
{
"epoch": 0.09691643981724693,
"grad_norm": 6.08568614592465,
"learning_rate": 1.9982240816853983e-05,
"loss": 1.0471,
"step": 761
},
{
"epoch": 0.09704379387745356,
"grad_norm": 5.1671257225291,
"learning_rate": 1.9982117740006763e-05,
"loss": 1.1163,
"step": 762
},
{
"epoch": 0.09717114793766018,
"grad_norm": 5.916493408818419,
"learning_rate": 1.9981994238531603e-05,
"loss": 1.1089,
"step": 763
},
{
"epoch": 0.09729850199786683,
"grad_norm": 4.560569041032856,
"learning_rate": 1.9981870312433755e-05,
"loss": 0.9946,
"step": 764
},
{
"epoch": 0.09742585605807345,
"grad_norm": 4.936586319610627,
"learning_rate": 1.9981745961718486e-05,
"loss": 1.1241,
"step": 765
},
{
"epoch": 0.09755321011828008,
"grad_norm": 5.56261932587953,
"learning_rate": 1.998162118639109e-05,
"loss": 1.1137,
"step": 766
},
{
"epoch": 0.09768056417848671,
"grad_norm": 4.700296051765587,
"learning_rate": 1.9981495986456876e-05,
"loss": 0.9961,
"step": 767
},
{
"epoch": 0.09780791823869335,
"grad_norm": 5.784826543609449,
"learning_rate": 1.9981370361921165e-05,
"loss": 1.0256,
"step": 768
},
{
"epoch": 0.09793527229889998,
"grad_norm": 5.929088537148662,
"learning_rate": 1.9981244312789306e-05,
"loss": 0.9581,
"step": 769
},
{
"epoch": 0.0980626263591066,
"grad_norm": 5.179139535494771,
"learning_rate": 1.9981117839066657e-05,
"loss": 0.9951,
"step": 770
},
{
"epoch": 0.09818998041931325,
"grad_norm": 6.371153924167855,
"learning_rate": 1.99809909407586e-05,
"loss": 1.0445,
"step": 771
},
{
"epoch": 0.09831733447951987,
"grad_norm": 7.152752283867728,
"learning_rate": 1.998086361787053e-05,
"loss": 1.1133,
"step": 772
},
{
"epoch": 0.0984446885397265,
"grad_norm": 5.0226457841069605,
"learning_rate": 1.998073587040787e-05,
"loss": 1.086,
"step": 773
},
{
"epoch": 0.09857204259993314,
"grad_norm": 5.248569293580547,
"learning_rate": 1.998060769837605e-05,
"loss": 1.0538,
"step": 774
},
{
"epoch": 0.09869939666013977,
"grad_norm": 6.354143319842696,
"learning_rate": 1.9980479101780525e-05,
"loss": 1.0211,
"step": 775
},
{
"epoch": 0.0988267507203464,
"grad_norm": 4.431387976904693,
"learning_rate": 1.9980350080626755e-05,
"loss": 0.9861,
"step": 776
},
{
"epoch": 0.09895410478055304,
"grad_norm": 6.274642885540535,
"learning_rate": 1.998022063492024e-05,
"loss": 1.0126,
"step": 777
},
{
"epoch": 0.09908145884075967,
"grad_norm": 4.3373684558105,
"learning_rate": 1.9980090764666486e-05,
"loss": 1.0245,
"step": 778
},
{
"epoch": 0.0992088129009663,
"grad_norm": 5.5406400642685085,
"learning_rate": 1.997996046987101e-05,
"loss": 1.0031,
"step": 779
},
{
"epoch": 0.09933616696117294,
"grad_norm": 6.194165887352312,
"learning_rate": 1.997982975053936e-05,
"loss": 1.0883,
"step": 780
},
{
"epoch": 0.09946352102137956,
"grad_norm": 5.211934830651193,
"learning_rate": 1.9979698606677095e-05,
"loss": 1.1272,
"step": 781
},
{
"epoch": 0.09959087508158619,
"grad_norm": 7.170192413089251,
"learning_rate": 1.9979567038289796e-05,
"loss": 1.0528,
"step": 782
},
{
"epoch": 0.09971822914179283,
"grad_norm": 6.254459468621482,
"learning_rate": 1.997943504538306e-05,
"loss": 0.9662,
"step": 783
},
{
"epoch": 0.09984558320199946,
"grad_norm": 6.957908510804562,
"learning_rate": 1.9979302627962494e-05,
"loss": 1.0736,
"step": 784
},
{
"epoch": 0.09997293726220609,
"grad_norm": 6.320699069316108,
"learning_rate": 1.997916978603374e-05,
"loss": 1.0725,
"step": 785
},
{
"epoch": 0.10010029132241273,
"grad_norm": 5.789714163937426,
"learning_rate": 1.9979036519602446e-05,
"loss": 1.0178,
"step": 786
},
{
"epoch": 0.10022764538261936,
"grad_norm": 6.830007751817898,
"learning_rate": 1.997890282867428e-05,
"loss": 1.0137,
"step": 787
},
{
"epoch": 0.10035499944282598,
"grad_norm": 4.903635482520552,
"learning_rate": 1.9978768713254927e-05,
"loss": 1.0192,
"step": 788
},
{
"epoch": 0.10048235350303261,
"grad_norm": 18.90650180617438,
"learning_rate": 1.99786341733501e-05,
"loss": 0.9982,
"step": 789
},
{
"epoch": 0.10060970756323925,
"grad_norm": 4.963165407706841,
"learning_rate": 1.997849920896551e-05,
"loss": 0.9878,
"step": 790
},
{
"epoch": 0.10073706162344588,
"grad_norm": 4.880410542025724,
"learning_rate": 1.9978363820106912e-05,
"loss": 1.0394,
"step": 791
},
{
"epoch": 0.10086441568365251,
"grad_norm": 5.359966990525815,
"learning_rate": 1.9978228006780056e-05,
"loss": 1.0565,
"step": 792
},
{
"epoch": 0.10099176974385915,
"grad_norm": 5.27348234934632,
"learning_rate": 1.9978091768990723e-05,
"loss": 1.1053,
"step": 793
},
{
"epoch": 0.10111912380406578,
"grad_norm": 6.207040468644822,
"learning_rate": 1.9977955106744706e-05,
"loss": 1.047,
"step": 794
},
{
"epoch": 0.1012464778642724,
"grad_norm": 5.75935564622613,
"learning_rate": 1.9977818020047816e-05,
"loss": 0.9662,
"step": 795
},
{
"epoch": 0.10137383192447905,
"grad_norm": 3.5010736725984377,
"learning_rate": 1.9977680508905894e-05,
"loss": 1.0921,
"step": 796
},
{
"epoch": 0.10150118598468567,
"grad_norm": 7.0157745277425425,
"learning_rate": 1.9977542573324782e-05,
"loss": 0.9324,
"step": 797
},
{
"epoch": 0.1016285400448923,
"grad_norm": 4.361273239480551,
"learning_rate": 1.9977404213310347e-05,
"loss": 1.0017,
"step": 798
},
{
"epoch": 0.10175589410509894,
"grad_norm": 5.415330262261502,
"learning_rate": 1.997726542886848e-05,
"loss": 0.997,
"step": 799
},
{
"epoch": 0.10188324816530557,
"grad_norm": 6.053517402057396,
"learning_rate": 1.9977126220005082e-05,
"loss": 1.0096,
"step": 800
},
{
"epoch": 0.1020106022255122,
"grad_norm": 5.6546141224148165,
"learning_rate": 1.9976986586726072e-05,
"loss": 1.1677,
"step": 801
},
{
"epoch": 0.10213795628571884,
"grad_norm": 4.5596153860015125,
"learning_rate": 1.9976846529037396e-05,
"loss": 1.039,
"step": 802
},
{
"epoch": 0.10226531034592547,
"grad_norm": 6.434913485996966,
"learning_rate": 1.9976706046945003e-05,
"loss": 1.032,
"step": 803
},
{
"epoch": 0.1023926644061321,
"grad_norm": 6.885333254525804,
"learning_rate": 1.9976565140454877e-05,
"loss": 1.0179,
"step": 804
},
{
"epoch": 0.10252001846633874,
"grad_norm": 5.634192690488991,
"learning_rate": 1.9976423809573014e-05,
"loss": 1.007,
"step": 805
},
{
"epoch": 0.10264737252654536,
"grad_norm": 5.028451840415393,
"learning_rate": 1.9976282054305415e-05,
"loss": 1.0552,
"step": 806
},
{
"epoch": 0.10277472658675199,
"grad_norm": 5.18185209266238,
"learning_rate": 1.9976139874658117e-05,
"loss": 0.985,
"step": 807
},
{
"epoch": 0.10290208064695863,
"grad_norm": 5.351761194703276,
"learning_rate": 1.9975997270637172e-05,
"loss": 0.977,
"step": 808
},
{
"epoch": 0.10302943470716526,
"grad_norm": 5.0199059016310725,
"learning_rate": 1.997585424224864e-05,
"loss": 1.0321,
"step": 809
},
{
"epoch": 0.10315678876737189,
"grad_norm": 7.4969247235155425,
"learning_rate": 1.9975710789498603e-05,
"loss": 1.1377,
"step": 810
},
{
"epoch": 0.10328414282757852,
"grad_norm": 5.024824126093265,
"learning_rate": 1.9975566912393173e-05,
"loss": 1.0533,
"step": 811
},
{
"epoch": 0.10341149688778516,
"grad_norm": 9.205641086845837,
"learning_rate": 1.9975422610938463e-05,
"loss": 1.0876,
"step": 812
},
{
"epoch": 0.10353885094799178,
"grad_norm": 4.248778448141489,
"learning_rate": 1.9975277885140613e-05,
"loss": 0.9902,
"step": 813
},
{
"epoch": 0.10366620500819841,
"grad_norm": 5.970955805599458,
"learning_rate": 1.9975132735005778e-05,
"loss": 1.0198,
"step": 814
},
{
"epoch": 0.10379355906840505,
"grad_norm": 5.154292249370319,
"learning_rate": 1.9974987160540132e-05,
"loss": 1.0242,
"step": 815
},
{
"epoch": 0.10392091312861168,
"grad_norm": 5.2176426609278055,
"learning_rate": 1.9974841161749875e-05,
"loss": 1.1066,
"step": 816
},
{
"epoch": 0.10404826718881831,
"grad_norm": 5.4636206413600235,
"learning_rate": 1.997469473864121e-05,
"loss": 1.1187,
"step": 817
},
{
"epoch": 0.10417562124902495,
"grad_norm": 5.683022885312907,
"learning_rate": 1.997454789122037e-05,
"loss": 1.1843,
"step": 818
},
{
"epoch": 0.10430297530923158,
"grad_norm": 5.910681621146795,
"learning_rate": 1.9974400619493595e-05,
"loss": 1.0454,
"step": 819
},
{
"epoch": 0.1044303293694382,
"grad_norm": 5.36205187083293,
"learning_rate": 1.9974252923467162e-05,
"loss": 0.9426,
"step": 820
},
{
"epoch": 0.10455768342964485,
"grad_norm": 5.648169546775838,
"learning_rate": 1.9974104803147344e-05,
"loss": 1.0387,
"step": 821
},
{
"epoch": 0.10468503748985147,
"grad_norm": 4.938694869698999,
"learning_rate": 1.9973956258540438e-05,
"loss": 1.0432,
"step": 822
},
{
"epoch": 0.1048123915500581,
"grad_norm": 6.627397211487185,
"learning_rate": 1.9973807289652777e-05,
"loss": 1.1652,
"step": 823
},
{
"epoch": 0.10493974561026474,
"grad_norm": 6.178454019201148,
"learning_rate": 1.9973657896490687e-05,
"loss": 1.0397,
"step": 824
},
{
"epoch": 0.10506709967047137,
"grad_norm": 6.174888628368443,
"learning_rate": 1.9973508079060524e-05,
"loss": 0.978,
"step": 825
},
{
"epoch": 0.105194453730678,
"grad_norm": 6.708740158855683,
"learning_rate": 1.9973357837368664e-05,
"loss": 0.9996,
"step": 826
},
{
"epoch": 0.10532180779088464,
"grad_norm": 7.273286451170561,
"learning_rate": 1.99732071714215e-05,
"loss": 0.9919,
"step": 827
},
{
"epoch": 0.10544916185109127,
"grad_norm": 4.836429305944261,
"learning_rate": 1.997305608122544e-05,
"loss": 0.9303,
"step": 828
},
{
"epoch": 0.1055765159112979,
"grad_norm": 5.203196115551915,
"learning_rate": 1.9972904566786903e-05,
"loss": 0.9287,
"step": 829
},
{
"epoch": 0.10570386997150454,
"grad_norm": 5.355710926325952,
"learning_rate": 1.9972752628112344e-05,
"loss": 1.0132,
"step": 830
},
{
"epoch": 0.10583122403171116,
"grad_norm": 5.978964807733789,
"learning_rate": 1.9972600265208224e-05,
"loss": 1.0344,
"step": 831
},
{
"epoch": 0.10595857809191779,
"grad_norm": 4.315744257777001,
"learning_rate": 1.9972447478081026e-05,
"loss": 1.1191,
"step": 832
},
{
"epoch": 0.10608593215212442,
"grad_norm": 5.5114553971235605,
"learning_rate": 1.9972294266737243e-05,
"loss": 1.0505,
"step": 833
},
{
"epoch": 0.10621328621233106,
"grad_norm": 7.059071476639452,
"learning_rate": 1.9972140631183396e-05,
"loss": 1.2202,
"step": 834
},
{
"epoch": 0.10634064027253769,
"grad_norm": 4.830647719965635,
"learning_rate": 1.9971986571426024e-05,
"loss": 1.0377,
"step": 835
},
{
"epoch": 0.10646799433274431,
"grad_norm": 5.099515121448014,
"learning_rate": 1.9971832087471678e-05,
"loss": 1.0378,
"step": 836
},
{
"epoch": 0.10659534839295096,
"grad_norm": 5.874044137666145,
"learning_rate": 1.9971677179326925e-05,
"loss": 1.0322,
"step": 837
},
{
"epoch": 0.10672270245315758,
"grad_norm": 6.157219050476186,
"learning_rate": 1.9971521846998362e-05,
"loss": 1.0744,
"step": 838
},
{
"epoch": 0.10685005651336421,
"grad_norm": 4.74324840763616,
"learning_rate": 1.997136609049259e-05,
"loss": 0.9892,
"step": 839
},
{
"epoch": 0.10697741057357085,
"grad_norm": 3.737316042860656,
"learning_rate": 1.997120990981624e-05,
"loss": 1.082,
"step": 840
},
{
"epoch": 0.10710476463377748,
"grad_norm": 5.970295491277951,
"learning_rate": 1.9971053304975955e-05,
"loss": 1.0749,
"step": 841
},
{
"epoch": 0.10723211869398411,
"grad_norm": 8.05229262408578,
"learning_rate": 1.9970896275978392e-05,
"loss": 1.0926,
"step": 842
},
{
"epoch": 0.10735947275419075,
"grad_norm": 4.885471541605568,
"learning_rate": 1.9970738822830237e-05,
"loss": 0.9979,
"step": 843
},
{
"epoch": 0.10748682681439738,
"grad_norm": 6.75132907922118,
"learning_rate": 1.9970580945538187e-05,
"loss": 1.0148,
"step": 844
},
{
"epoch": 0.107614180874604,
"grad_norm": 5.0097069928342846,
"learning_rate": 1.997042264410895e-05,
"loss": 1.1069,
"step": 845
},
{
"epoch": 0.10774153493481065,
"grad_norm": 5.598578899449339,
"learning_rate": 1.9970263918549274e-05,
"loss": 1.0181,
"step": 846
},
{
"epoch": 0.10786888899501727,
"grad_norm": 5.37790388575956,
"learning_rate": 1.9970104768865895e-05,
"loss": 1.0247,
"step": 847
},
{
"epoch": 0.1079962430552239,
"grad_norm": 6.595977551571855,
"learning_rate": 1.9969945195065596e-05,
"loss": 1.0204,
"step": 848
},
{
"epoch": 0.10812359711543054,
"grad_norm": 7.964236907516185,
"learning_rate": 1.996978519715516e-05,
"loss": 0.9665,
"step": 849
},
{
"epoch": 0.10825095117563717,
"grad_norm": 6.392637201835179,
"learning_rate": 1.9969624775141393e-05,
"loss": 1.0868,
"step": 850
},
{
"epoch": 0.1083783052358438,
"grad_norm": 5.587715175504954,
"learning_rate": 1.9969463929031118e-05,
"loss": 0.9563,
"step": 851
},
{
"epoch": 0.10850565929605044,
"grad_norm": 4.658229330377023,
"learning_rate": 1.9969302658831182e-05,
"loss": 1.0354,
"step": 852
},
{
"epoch": 0.10863301335625707,
"grad_norm": 4.787778299474807,
"learning_rate": 1.996914096454844e-05,
"loss": 0.9508,
"step": 853
},
{
"epoch": 0.1087603674164637,
"grad_norm": 4.867284685024919,
"learning_rate": 1.996897884618977e-05,
"loss": 1.0619,
"step": 854
},
{
"epoch": 0.10888772147667033,
"grad_norm": 5.147589421211179,
"learning_rate": 1.9968816303762076e-05,
"loss": 0.9208,
"step": 855
},
{
"epoch": 0.10901507553687696,
"grad_norm": 6.599867455278636,
"learning_rate": 1.9968653337272262e-05,
"loss": 0.9198,
"step": 856
},
{
"epoch": 0.10914242959708359,
"grad_norm": 7.3878300162248465,
"learning_rate": 1.9968489946727265e-05,
"loss": 1.0832,
"step": 857
},
{
"epoch": 0.10926978365729022,
"grad_norm": 5.149081153750179,
"learning_rate": 1.996832613213404e-05,
"loss": 1.0392,
"step": 858
},
{
"epoch": 0.10939713771749686,
"grad_norm": 7.385644504779292,
"learning_rate": 1.9968161893499548e-05,
"loss": 1.0305,
"step": 859
},
{
"epoch": 0.10952449177770349,
"grad_norm": 6.126046850311139,
"learning_rate": 1.9967997230830784e-05,
"loss": 1.0887,
"step": 860
},
{
"epoch": 0.10965184583791011,
"grad_norm": 7.1861931981170715,
"learning_rate": 1.996783214413474e-05,
"loss": 1.109,
"step": 861
},
{
"epoch": 0.10977919989811676,
"grad_norm": 6.273300682466664,
"learning_rate": 1.9967666633418454e-05,
"loss": 1.0648,
"step": 862
},
{
"epoch": 0.10990655395832338,
"grad_norm": 6.694825340539664,
"learning_rate": 1.996750069868895e-05,
"loss": 1.1229,
"step": 863
},
{
"epoch": 0.11003390801853001,
"grad_norm": 4.287595849036246,
"learning_rate": 1.9967334339953303e-05,
"loss": 0.9974,
"step": 864
},
{
"epoch": 0.11016126207873665,
"grad_norm": 5.7722573112928215,
"learning_rate": 1.996716755721858e-05,
"loss": 1.0254,
"step": 865
},
{
"epoch": 0.11028861613894328,
"grad_norm": 4.54564301360154,
"learning_rate": 1.9967000350491873e-05,
"loss": 0.947,
"step": 866
},
{
"epoch": 0.11041597019914991,
"grad_norm": 4.937993278489573,
"learning_rate": 1.9966832719780305e-05,
"loss": 1.105,
"step": 867
},
{
"epoch": 0.11054332425935655,
"grad_norm": 5.060781116430827,
"learning_rate": 1.9966664665090997e-05,
"loss": 1.111,
"step": 868
},
{
"epoch": 0.11067067831956318,
"grad_norm": 6.785440723383181,
"learning_rate": 1.9966496186431106e-05,
"loss": 1.0737,
"step": 869
},
{
"epoch": 0.1107980323797698,
"grad_norm": 6.19654754989308,
"learning_rate": 1.9966327283807794e-05,
"loss": 1.0739,
"step": 870
},
{
"epoch": 0.11092538643997644,
"grad_norm": 6.420826475951312,
"learning_rate": 1.996615795722825e-05,
"loss": 1.0866,
"step": 871
},
{
"epoch": 0.11105274050018307,
"grad_norm": 5.052261359889531,
"learning_rate": 1.996598820669967e-05,
"loss": 1.0155,
"step": 872
},
{
"epoch": 0.1111800945603897,
"grad_norm": 4.816232002119726,
"learning_rate": 1.996581803222928e-05,
"loss": 1.0649,
"step": 873
},
{
"epoch": 0.11130744862059634,
"grad_norm": 4.5369893524939595,
"learning_rate": 1.9965647433824315e-05,
"loss": 0.9625,
"step": 874
},
{
"epoch": 0.11143480268080297,
"grad_norm": 5.307753652943214,
"learning_rate": 1.9965476411492043e-05,
"loss": 0.9882,
"step": 875
},
{
"epoch": 0.1115621567410096,
"grad_norm": 6.092648763675741,
"learning_rate": 1.9965304965239723e-05,
"loss": 0.8954,
"step": 876
},
{
"epoch": 0.11168951080121624,
"grad_norm": 6.984597836210311,
"learning_rate": 1.996513309507466e-05,
"loss": 1.0302,
"step": 877
},
{
"epoch": 0.11181686486142287,
"grad_norm": 5.913009731264718,
"learning_rate": 1.9964960801004164e-05,
"loss": 1.082,
"step": 878
},
{
"epoch": 0.11194421892162949,
"grad_norm": 4.500339071427008,
"learning_rate": 1.9964788083035554e-05,
"loss": 1.0465,
"step": 879
},
{
"epoch": 0.11207157298183612,
"grad_norm": 6.428641140272211,
"learning_rate": 1.9964614941176194e-05,
"loss": 1.054,
"step": 880
},
{
"epoch": 0.11219892704204276,
"grad_norm": 5.458770889720949,
"learning_rate": 1.9964441375433436e-05,
"loss": 0.9883,
"step": 881
},
{
"epoch": 0.11232628110224939,
"grad_norm": 5.406102525761324,
"learning_rate": 1.996426738581467e-05,
"loss": 0.9914,
"step": 882
},
{
"epoch": 0.11245363516245602,
"grad_norm": 5.6600517858286175,
"learning_rate": 1.996409297232729e-05,
"loss": 0.9404,
"step": 883
},
{
"epoch": 0.11258098922266266,
"grad_norm": 6.302272858935621,
"learning_rate": 1.996391813497872e-05,
"loss": 0.9588,
"step": 884
},
{
"epoch": 0.11270834328286929,
"grad_norm": 6.660119593007626,
"learning_rate": 1.99637428737764e-05,
"loss": 1.0316,
"step": 885
},
{
"epoch": 0.11283569734307591,
"grad_norm": 5.668058934930348,
"learning_rate": 1.9963567188727783e-05,
"loss": 1.0363,
"step": 886
},
{
"epoch": 0.11296305140328256,
"grad_norm": 4.359857566676804,
"learning_rate": 1.9963391079840344e-05,
"loss": 1.0509,
"step": 887
},
{
"epoch": 0.11309040546348918,
"grad_norm": 5.11794036685001,
"learning_rate": 1.9963214547121573e-05,
"loss": 1.0572,
"step": 888
},
{
"epoch": 0.11321775952369581,
"grad_norm": 8.069355260721297,
"learning_rate": 1.9963037590578977e-05,
"loss": 1.0512,
"step": 889
},
{
"epoch": 0.11334511358390245,
"grad_norm": 5.085180082630144,
"learning_rate": 1.9962860210220085e-05,
"loss": 0.9946,
"step": 890
},
{
"epoch": 0.11347246764410908,
"grad_norm": 6.3451816089753414,
"learning_rate": 1.9962682406052445e-05,
"loss": 1.0836,
"step": 891
},
{
"epoch": 0.1135998217043157,
"grad_norm": 4.30454150054419,
"learning_rate": 1.996250417808362e-05,
"loss": 1.0612,
"step": 892
},
{
"epoch": 0.11372717576452235,
"grad_norm": 6.121196889586544,
"learning_rate": 1.996232552632119e-05,
"loss": 1.1126,
"step": 893
},
{
"epoch": 0.11385452982472898,
"grad_norm": 5.352515632788303,
"learning_rate": 1.9962146450772756e-05,
"loss": 1.0238,
"step": 894
},
{
"epoch": 0.1139818838849356,
"grad_norm": 6.644230409121299,
"learning_rate": 1.9961966951445933e-05,
"loss": 1.008,
"step": 895
},
{
"epoch": 0.11410923794514224,
"grad_norm": 5.595505082119347,
"learning_rate": 1.9961787028348357e-05,
"loss": 1.085,
"step": 896
},
{
"epoch": 0.11423659200534887,
"grad_norm": 5.556965678651519,
"learning_rate": 1.9961606681487685e-05,
"loss": 1.0309,
"step": 897
},
{
"epoch": 0.1143639460655555,
"grad_norm": 4.581100382319801,
"learning_rate": 1.996142591087159e-05,
"loss": 1.0942,
"step": 898
},
{
"epoch": 0.11449130012576214,
"grad_norm": 5.835283153368007,
"learning_rate": 1.9961244716507757e-05,
"loss": 1.1027,
"step": 899
},
{
"epoch": 0.11461865418596877,
"grad_norm": 5.063142900500023,
"learning_rate": 1.9961063098403897e-05,
"loss": 0.9936,
"step": 900
},
{
"epoch": 0.1147460082461754,
"grad_norm": 4.51443898818577,
"learning_rate": 1.9960881056567732e-05,
"loss": 0.9346,
"step": 901
},
{
"epoch": 0.11487336230638202,
"grad_norm": 5.329502856059749,
"learning_rate": 1.9960698591007008e-05,
"loss": 1.0803,
"step": 902
},
{
"epoch": 0.11500071636658867,
"grad_norm": 4.504037740993079,
"learning_rate": 1.996051570172949e-05,
"loss": 0.9776,
"step": 903
},
{
"epoch": 0.11512807042679529,
"grad_norm": 6.117784882628449,
"learning_rate": 1.9960332388742952e-05,
"loss": 1.0179,
"step": 904
},
{
"epoch": 0.11525542448700192,
"grad_norm": 5.357360995256535,
"learning_rate": 1.99601486520552e-05,
"loss": 1.1039,
"step": 905
},
{
"epoch": 0.11538277854720856,
"grad_norm": 10.642215559466937,
"learning_rate": 1.995996449167404e-05,
"loss": 0.9746,
"step": 906
},
{
"epoch": 0.11551013260741519,
"grad_norm": 5.78019713016134,
"learning_rate": 1.9959779907607317e-05,
"loss": 0.9421,
"step": 907
},
{
"epoch": 0.11563748666762182,
"grad_norm": 6.034713778278319,
"learning_rate": 1.995959489986287e-05,
"loss": 1.1411,
"step": 908
},
{
"epoch": 0.11576484072782846,
"grad_norm": 5.582388650865901,
"learning_rate": 1.9959409468448582e-05,
"loss": 1.0087,
"step": 909
},
{
"epoch": 0.11589219478803509,
"grad_norm": 5.796074084454103,
"learning_rate": 1.995922361337233e-05,
"loss": 1.1047,
"step": 910
},
{
"epoch": 0.11601954884824171,
"grad_norm": 6.864811231180803,
"learning_rate": 1.9959037334642027e-05,
"loss": 1.0263,
"step": 911
},
{
"epoch": 0.11614690290844835,
"grad_norm": 5.806496093265156,
"learning_rate": 1.9958850632265595e-05,
"loss": 0.9675,
"step": 912
},
{
"epoch": 0.11627425696865498,
"grad_norm": 4.753198851564046,
"learning_rate": 1.995866350625098e-05,
"loss": 0.9769,
"step": 913
},
{
"epoch": 0.11640161102886161,
"grad_norm": 5.277047375560149,
"learning_rate": 1.9958475956606133e-05,
"loss": 1.0622,
"step": 914
},
{
"epoch": 0.11652896508906825,
"grad_norm": 4.805170397798593,
"learning_rate": 1.995828798333904e-05,
"loss": 1.0204,
"step": 915
},
{
"epoch": 0.11665631914927488,
"grad_norm": 3.8686471361527017,
"learning_rate": 1.9958099586457696e-05,
"loss": 1.0807,
"step": 916
},
{
"epoch": 0.1167836732094815,
"grad_norm": 5.415077015337016,
"learning_rate": 1.9957910765970114e-05,
"loss": 1.0983,
"step": 917
},
{
"epoch": 0.11691102726968815,
"grad_norm": 6.980435687788938,
"learning_rate": 1.9957721521884322e-05,
"loss": 1.1048,
"step": 918
},
{
"epoch": 0.11703838132989478,
"grad_norm": 4.371332647836583,
"learning_rate": 1.9957531854208378e-05,
"loss": 1.0857,
"step": 919
},
{
"epoch": 0.1171657353901014,
"grad_norm": 5.134828586707463,
"learning_rate": 1.9957341762950346e-05,
"loss": 1.0729,
"step": 920
},
{
"epoch": 0.11729308945030804,
"grad_norm": 5.138492594150226,
"learning_rate": 1.9957151248118314e-05,
"loss": 1.131,
"step": 921
},
{
"epoch": 0.11742044351051467,
"grad_norm": 4.016436424357209,
"learning_rate": 1.9956960309720385e-05,
"loss": 1.0131,
"step": 922
},
{
"epoch": 0.1175477975707213,
"grad_norm": 5.120844016543802,
"learning_rate": 1.995676894776468e-05,
"loss": 1.0525,
"step": 923
},
{
"epoch": 0.11767515163092793,
"grad_norm": 6.446405565457038,
"learning_rate": 1.995657716225934e-05,
"loss": 1.0622,
"step": 924
},
{
"epoch": 0.11780250569113457,
"grad_norm": 5.7319024599644814,
"learning_rate": 1.9956384953212526e-05,
"loss": 1.0326,
"step": 925
},
{
"epoch": 0.1179298597513412,
"grad_norm": 5.503297337568938,
"learning_rate": 1.995619232063241e-05,
"loss": 0.9426,
"step": 926
},
{
"epoch": 0.11805721381154782,
"grad_norm": 5.387966927033936,
"learning_rate": 1.9955999264527194e-05,
"loss": 1.0234,
"step": 927
},
{
"epoch": 0.11818456787175446,
"grad_norm": 4.988396723164246,
"learning_rate": 1.9955805784905083e-05,
"loss": 1.0124,
"step": 928
},
{
"epoch": 0.11831192193196109,
"grad_norm": 5.309249422003908,
"learning_rate": 1.9955611881774308e-05,
"loss": 1.1048,
"step": 929
},
{
"epoch": 0.11843927599216772,
"grad_norm": 5.357774679073277,
"learning_rate": 1.995541755514312e-05,
"loss": 0.9624,
"step": 930
},
{
"epoch": 0.11856663005237436,
"grad_norm": 6.42531343281874,
"learning_rate": 1.9955222805019786e-05,
"loss": 1.0409,
"step": 931
},
{
"epoch": 0.11869398411258099,
"grad_norm": 5.131309347239302,
"learning_rate": 1.9955027631412584e-05,
"loss": 1.1056,
"step": 932
},
{
"epoch": 0.11882133817278762,
"grad_norm": 4.4340370527963895,
"learning_rate": 1.9954832034329827e-05,
"loss": 0.9607,
"step": 933
},
{
"epoch": 0.11894869223299426,
"grad_norm": 5.347056475736151,
"learning_rate": 1.9954636013779826e-05,
"loss": 1.0851,
"step": 934
},
{
"epoch": 0.11907604629320089,
"grad_norm": 6.211339474511673,
"learning_rate": 1.995443956977093e-05,
"loss": 1.0095,
"step": 935
},
{
"epoch": 0.11920340035340751,
"grad_norm": 5.982173504781401,
"learning_rate": 1.995424270231148e-05,
"loss": 1.0042,
"step": 936
},
{
"epoch": 0.11933075441361415,
"grad_norm": 6.007968697173025,
"learning_rate": 1.9954045411409864e-05,
"loss": 1.0089,
"step": 937
},
{
"epoch": 0.11945810847382078,
"grad_norm": 4.726619196780303,
"learning_rate": 1.9953847697074472e-05,
"loss": 0.984,
"step": 938
},
{
"epoch": 0.11958546253402741,
"grad_norm": 4.7802597715905275,
"learning_rate": 1.995364955931371e-05,
"loss": 1.0665,
"step": 939
},
{
"epoch": 0.11971281659423405,
"grad_norm": 4.51709910688604,
"learning_rate": 1.995345099813601e-05,
"loss": 0.9949,
"step": 940
},
{
"epoch": 0.11984017065444068,
"grad_norm": 5.500703073126749,
"learning_rate": 1.9953252013549816e-05,
"loss": 1.0011,
"step": 941
},
{
"epoch": 0.1199675247146473,
"grad_norm": 6.24212088805979,
"learning_rate": 1.9953052605563596e-05,
"loss": 1.0222,
"step": 942
},
{
"epoch": 0.12009487877485395,
"grad_norm": 4.63458043245687,
"learning_rate": 1.995285277418583e-05,
"loss": 1.0849,
"step": 943
},
{
"epoch": 0.12022223283506057,
"grad_norm": 3.6977378837815587,
"learning_rate": 1.9952652519425016e-05,
"loss": 0.9489,
"step": 944
},
{
"epoch": 0.1203495868952672,
"grad_norm": 5.320802371961369,
"learning_rate": 1.995245184128968e-05,
"loss": 1.1189,
"step": 945
},
{
"epoch": 0.12047694095547384,
"grad_norm": 5.860292896853712,
"learning_rate": 1.9952250739788356e-05,
"loss": 0.9716,
"step": 946
},
{
"epoch": 0.12060429501568047,
"grad_norm": 5.153604094452106,
"learning_rate": 1.995204921492959e-05,
"loss": 0.9592,
"step": 947
},
{
"epoch": 0.1207316490758871,
"grad_norm": 7.173086026737213,
"learning_rate": 1.995184726672197e-05,
"loss": 1.0795,
"step": 948
},
{
"epoch": 0.12085900313609373,
"grad_norm": 6.421640656365196,
"learning_rate": 1.9951644895174076e-05,
"loss": 1.0193,
"step": 949
},
{
"epoch": 0.12098635719630037,
"grad_norm": 5.4544739024931195,
"learning_rate": 1.995144210029452e-05,
"loss": 0.9711,
"step": 950
},
{
"epoch": 0.121113711256507,
"grad_norm": 5.661363850647127,
"learning_rate": 1.9951238882091926e-05,
"loss": 1.0674,
"step": 951
},
{
"epoch": 0.12124106531671362,
"grad_norm": 7.360293450517743,
"learning_rate": 1.995103524057494e-05,
"loss": 1.0153,
"step": 952
},
{
"epoch": 0.12136841937692026,
"grad_norm": 4.63858305836024,
"learning_rate": 1.995083117575223e-05,
"loss": 1.1147,
"step": 953
},
{
"epoch": 0.12149577343712689,
"grad_norm": 5.686394552597674,
"learning_rate": 1.9950626687632466e-05,
"loss": 0.9923,
"step": 954
},
{
"epoch": 0.12162312749733352,
"grad_norm": 6.112128610386265,
"learning_rate": 1.9950421776224353e-05,
"loss": 0.9806,
"step": 955
},
{
"epoch": 0.12175048155754016,
"grad_norm": 6.282131759471412,
"learning_rate": 1.9950216441536613e-05,
"loss": 1.165,
"step": 956
},
{
"epoch": 0.12187783561774679,
"grad_norm": 5.361202895850754,
"learning_rate": 1.9950010683577968e-05,
"loss": 1.0751,
"step": 957
},
{
"epoch": 0.12200518967795342,
"grad_norm": 4.535062251730972,
"learning_rate": 1.9949804502357183e-05,
"loss": 1.0677,
"step": 958
},
{
"epoch": 0.12213254373816006,
"grad_norm": 6.974919756902898,
"learning_rate": 1.994959789788302e-05,
"loss": 1.0085,
"step": 959
},
{
"epoch": 0.12225989779836668,
"grad_norm": 5.319029621165861,
"learning_rate": 1.994939087016427e-05,
"loss": 1.004,
"step": 960
},
{
"epoch": 0.12238725185857331,
"grad_norm": 6.546630563474553,
"learning_rate": 1.9949183419209747e-05,
"loss": 0.9709,
"step": 961
},
{
"epoch": 0.12251460591877995,
"grad_norm": 5.408159569220149,
"learning_rate": 1.9948975545028263e-05,
"loss": 1.0071,
"step": 962
},
{
"epoch": 0.12264195997898658,
"grad_norm": 5.251176611431323,
"learning_rate": 1.994876724762867e-05,
"loss": 1.0617,
"step": 963
},
{
"epoch": 0.12276931403919321,
"grad_norm": 4.902891917186358,
"learning_rate": 1.9948558527019826e-05,
"loss": 1.0336,
"step": 964
},
{
"epoch": 0.12289666809939985,
"grad_norm": 6.705538939706773,
"learning_rate": 1.994834938321061e-05,
"loss": 1.0693,
"step": 965
},
{
"epoch": 0.12302402215960648,
"grad_norm": 6.087297201988893,
"learning_rate": 1.994813981620992e-05,
"loss": 1.0077,
"step": 966
},
{
"epoch": 0.1231513762198131,
"grad_norm": 4.56559021693041,
"learning_rate": 1.9947929826026668e-05,
"loss": 1.0278,
"step": 967
},
{
"epoch": 0.12327873028001975,
"grad_norm": 4.71483525462677,
"learning_rate": 1.9947719412669787e-05,
"loss": 1.1168,
"step": 968
},
{
"epoch": 0.12340608434022637,
"grad_norm": 6.379245316884144,
"learning_rate": 1.994750857614823e-05,
"loss": 1.0561,
"step": 969
},
{
"epoch": 0.123533438400433,
"grad_norm": 5.052821385327841,
"learning_rate": 1.9947297316470963e-05,
"loss": 1.1216,
"step": 970
},
{
"epoch": 0.12366079246063963,
"grad_norm": 4.041449839225834,
"learning_rate": 1.9947085633646977e-05,
"loss": 0.9505,
"step": 971
},
{
"epoch": 0.12378814652084627,
"grad_norm": 4.702999278064229,
"learning_rate": 1.994687352768527e-05,
"loss": 1.1083,
"step": 972
},
{
"epoch": 0.1239155005810529,
"grad_norm": 4.5330116793643835,
"learning_rate": 1.994666099859487e-05,
"loss": 1.0322,
"step": 973
},
{
"epoch": 0.12404285464125953,
"grad_norm": 5.515008735269388,
"learning_rate": 1.994644804638482e-05,
"loss": 1.005,
"step": 974
},
{
"epoch": 0.12417020870146617,
"grad_norm": 8.7381555226762,
"learning_rate": 1.9946234671064172e-05,
"loss": 0.9686,
"step": 975
},
{
"epoch": 0.1242975627616728,
"grad_norm": 5.550035376908558,
"learning_rate": 1.994602087264201e-05,
"loss": 1.108,
"step": 976
},
{
"epoch": 0.12442491682187942,
"grad_norm": 5.977919959197699,
"learning_rate": 1.994580665112742e-05,
"loss": 1.0092,
"step": 977
},
{
"epoch": 0.12455227088208606,
"grad_norm": 4.282609932418776,
"learning_rate": 1.994559200652952e-05,
"loss": 1.0182,
"step": 978
},
{
"epoch": 0.12467962494229269,
"grad_norm": 5.907460245135092,
"learning_rate": 1.9945376938857443e-05,
"loss": 0.9318,
"step": 979
},
{
"epoch": 0.12480697900249932,
"grad_norm": 4.831551629010998,
"learning_rate": 1.994516144812033e-05,
"loss": 1.0026,
"step": 980
},
{
"epoch": 0.12493433306270596,
"grad_norm": 4.311846255079303,
"learning_rate": 1.994494553432736e-05,
"loss": 0.9798,
"step": 981
},
{
"epoch": 0.12506168712291257,
"grad_norm": 5.6537046659323265,
"learning_rate": 1.9944729197487702e-05,
"loss": 1.0372,
"step": 982
},
{
"epoch": 0.12518904118311922,
"grad_norm": 8.135826527084301,
"learning_rate": 1.994451243761057e-05,
"loss": 1.1158,
"step": 983
},
{
"epoch": 0.12531639524332586,
"grad_norm": 7.315360789394048,
"learning_rate": 1.9944295254705187e-05,
"loss": 1.1006,
"step": 984
},
{
"epoch": 0.12544374930353247,
"grad_norm": 8.115377978949036,
"learning_rate": 1.994407764878078e-05,
"loss": 1.0766,
"step": 985
},
{
"epoch": 0.1255711033637391,
"grad_norm": 5.391810013262356,
"learning_rate": 1.9943859619846617e-05,
"loss": 1.0636,
"step": 986
},
{
"epoch": 0.12569845742394575,
"grad_norm": 6.15608052441614,
"learning_rate": 1.9943641167911965e-05,
"loss": 1.0649,
"step": 987
},
{
"epoch": 0.12582581148415237,
"grad_norm": 8.880531843600096,
"learning_rate": 1.994342229298612e-05,
"loss": 1.0144,
"step": 988
},
{
"epoch": 0.125953165544359,
"grad_norm": 6.462124439889047,
"learning_rate": 1.9943202995078394e-05,
"loss": 1.0238,
"step": 989
},
{
"epoch": 0.12608051960456565,
"grad_norm": 10.131981973721643,
"learning_rate": 1.994298327419811e-05,
"loss": 1.0622,
"step": 990
},
{
"epoch": 0.12620787366477226,
"grad_norm": 6.29611830033835,
"learning_rate": 1.9942763130354624e-05,
"loss": 1.1349,
"step": 991
},
{
"epoch": 0.1263352277249789,
"grad_norm": 5.458378138855743,
"learning_rate": 1.9942542563557294e-05,
"loss": 1.0549,
"step": 992
},
{
"epoch": 0.12646258178518555,
"grad_norm": 5.555032754246985,
"learning_rate": 1.9942321573815502e-05,
"loss": 0.9425,
"step": 993
},
{
"epoch": 0.12658993584539216,
"grad_norm": 4.502659753383141,
"learning_rate": 1.994210016113865e-05,
"loss": 1.1005,
"step": 994
},
{
"epoch": 0.1267172899055988,
"grad_norm": 5.353733663701035,
"learning_rate": 1.994187832553616e-05,
"loss": 1.0607,
"step": 995
},
{
"epoch": 0.12684464396580544,
"grad_norm": 3.80325171228991,
"learning_rate": 1.9941656067017466e-05,
"loss": 1.139,
"step": 996
},
{
"epoch": 0.12697199802601206,
"grad_norm": 4.255154345141497,
"learning_rate": 1.9941433385592022e-05,
"loss": 0.9297,
"step": 997
},
{
"epoch": 0.1270993520862187,
"grad_norm": 4.50577894650802,
"learning_rate": 1.99412102812693e-05,
"loss": 1.0254,
"step": 998
},
{
"epoch": 0.12722670614642534,
"grad_norm": 6.663816351330356,
"learning_rate": 1.9940986754058792e-05,
"loss": 1.1039,
"step": 999
},
{
"epoch": 0.12735406020663195,
"grad_norm": 6.1321546845655694,
"learning_rate": 1.9940762803970006e-05,
"loss": 1.0365,
"step": 1000
},
{
"epoch": 0.1274814142668386,
"grad_norm": 5.012189883106625,
"learning_rate": 1.9940538431012472e-05,
"loss": 1.0125,
"step": 1001
},
{
"epoch": 0.12760876832704524,
"grad_norm": 4.761917222778728,
"learning_rate": 1.9940313635195728e-05,
"loss": 1.0547,
"step": 1002
},
{
"epoch": 0.12773612238725185,
"grad_norm": 5.289529123387824,
"learning_rate": 1.9940088416529342e-05,
"loss": 1.0712,
"step": 1003
},
{
"epoch": 0.1278634764474585,
"grad_norm": 5.437705902059703,
"learning_rate": 1.9939862775022893e-05,
"loss": 1.0807,
"step": 1004
},
{
"epoch": 0.12799083050766513,
"grad_norm": 6.951708291416241,
"learning_rate": 1.993963671068598e-05,
"loss": 1.0059,
"step": 1005
},
{
"epoch": 0.12811818456787175,
"grad_norm": 4.9604876456962845,
"learning_rate": 1.9939410223528215e-05,
"loss": 1.0181,
"step": 1006
},
{
"epoch": 0.1282455386280784,
"grad_norm": 7.510870367376681,
"learning_rate": 1.993918331355924e-05,
"loss": 1.0302,
"step": 1007
},
{
"epoch": 0.12837289268828503,
"grad_norm": 5.293053432786597,
"learning_rate": 1.9938955980788703e-05,
"loss": 1.0705,
"step": 1008
},
{
"epoch": 0.12850024674849164,
"grad_norm": 6.103049982237845,
"learning_rate": 1.9938728225226273e-05,
"loss": 0.9717,
"step": 1009
},
{
"epoch": 0.12862760080869828,
"grad_norm": 4.882607026926755,
"learning_rate": 1.9938500046881643e-05,
"loss": 0.9793,
"step": 1010
},
{
"epoch": 0.12875495486890493,
"grad_norm": 5.646936831956459,
"learning_rate": 1.9938271445764515e-05,
"loss": 0.9924,
"step": 1011
},
{
"epoch": 0.12888230892911154,
"grad_norm": 6.42042187216709,
"learning_rate": 1.9938042421884617e-05,
"loss": 0.9512,
"step": 1012
},
{
"epoch": 0.12900966298931818,
"grad_norm": 4.131351782487081,
"learning_rate": 1.993781297525169e-05,
"loss": 1.0183,
"step": 1013
},
{
"epoch": 0.12913701704952482,
"grad_norm": 5.587315444117175,
"learning_rate": 1.9937583105875494e-05,
"loss": 0.9225,
"step": 1014
},
{
"epoch": 0.12926437110973144,
"grad_norm": 5.817599121088648,
"learning_rate": 1.9937352813765808e-05,
"loss": 0.9889,
"step": 1015
},
{
"epoch": 0.12939172516993808,
"grad_norm": 4.717808307383814,
"learning_rate": 1.9937122098932428e-05,
"loss": 0.9959,
"step": 1016
},
{
"epoch": 0.12951907923014472,
"grad_norm": 4.6219142208021475,
"learning_rate": 1.9936890961385168e-05,
"loss": 0.9909,
"step": 1017
},
{
"epoch": 0.12964643329035133,
"grad_norm": 5.360274258899001,
"learning_rate": 1.993665940113386e-05,
"loss": 1.0057,
"step": 1018
},
{
"epoch": 0.12977378735055797,
"grad_norm": 5.580861336737161,
"learning_rate": 1.9936427418188357e-05,
"loss": 0.9989,
"step": 1019
},
{
"epoch": 0.12990114141076461,
"grad_norm": 5.159156858475661,
"learning_rate": 1.9936195012558524e-05,
"loss": 0.9662,
"step": 1020
},
{
"epoch": 0.13002849547097123,
"grad_norm": 5.548067034656748,
"learning_rate": 1.993596218425425e-05,
"loss": 1.0041,
"step": 1021
},
{
"epoch": 0.13015584953117787,
"grad_norm": 4.9843518486800695,
"learning_rate": 1.9935728933285438e-05,
"loss": 1.0422,
"step": 1022
},
{
"epoch": 0.1302832035913845,
"grad_norm": 4.844871735661303,
"learning_rate": 1.9935495259662008e-05,
"loss": 1.0537,
"step": 1023
},
{
"epoch": 0.13041055765159112,
"grad_norm": 5.4011935363233645,
"learning_rate": 1.9935261163393904e-05,
"loss": 0.9057,
"step": 1024
},
{
"epoch": 0.13053791171179777,
"grad_norm": 5.113145096425448,
"learning_rate": 1.9935026644491082e-05,
"loss": 1.0,
"step": 1025
},
{
"epoch": 0.1306652657720044,
"grad_norm": 4.776971436010882,
"learning_rate": 1.9934791702963515e-05,
"loss": 1.0148,
"step": 1026
},
{
"epoch": 0.13079261983221102,
"grad_norm": 5.132128897887906,
"learning_rate": 1.9934556338821206e-05,
"loss": 0.9756,
"step": 1027
},
{
"epoch": 0.13091997389241766,
"grad_norm": 4.755116768575519,
"learning_rate": 1.9934320552074162e-05,
"loss": 1.0135,
"step": 1028
},
{
"epoch": 0.13104732795262428,
"grad_norm": 4.623487276696922,
"learning_rate": 1.9934084342732413e-05,
"loss": 0.9728,
"step": 1029
},
{
"epoch": 0.13117468201283092,
"grad_norm": 5.275131383046155,
"learning_rate": 1.9933847710806e-05,
"loss": 0.9805,
"step": 1030
},
{
"epoch": 0.13130203607303756,
"grad_norm": 5.539016357574028,
"learning_rate": 1.9933610656305006e-05,
"loss": 1.1283,
"step": 1031
},
{
"epoch": 0.13142939013324417,
"grad_norm": 5.368997460236626,
"learning_rate": 1.99333731792395e-05,
"loss": 1.0306,
"step": 1032
},
{
"epoch": 0.13155674419345081,
"grad_norm": 4.374498829065241,
"learning_rate": 1.9933135279619592e-05,
"loss": 0.9264,
"step": 1033
},
{
"epoch": 0.13168409825365746,
"grad_norm": 4.632795825947545,
"learning_rate": 1.9932896957455397e-05,
"loss": 0.9511,
"step": 1034
},
{
"epoch": 0.13181145231386407,
"grad_norm": 5.802893791461331,
"learning_rate": 1.9932658212757053e-05,
"loss": 1.0492,
"step": 1035
},
{
"epoch": 0.1319388063740707,
"grad_norm": 7.5402494612894975,
"learning_rate": 1.9932419045534724e-05,
"loss": 1.1128,
"step": 1036
},
{
"epoch": 0.13206616043427735,
"grad_norm": 5.214650079795567,
"learning_rate": 1.9932179455798574e-05,
"loss": 0.9812,
"step": 1037
},
{
"epoch": 0.13219351449448397,
"grad_norm": 5.761631024226228,
"learning_rate": 1.9931939443558803e-05,
"loss": 1.0368,
"step": 1038
},
{
"epoch": 0.1323208685546906,
"grad_norm": 5.505637660709738,
"learning_rate": 1.993169900882561e-05,
"loss": 1.0554,
"step": 1039
},
{
"epoch": 0.13244822261489725,
"grad_norm": 7.051467786001133,
"learning_rate": 1.9931458151609234e-05,
"loss": 1.0514,
"step": 1040
},
{
"epoch": 0.13257557667510386,
"grad_norm": 6.307248187578084,
"learning_rate": 1.9931216871919914e-05,
"loss": 1.0074,
"step": 1041
},
{
"epoch": 0.1327029307353105,
"grad_norm": 5.105711016930289,
"learning_rate": 1.9930975169767918e-05,
"loss": 1.0188,
"step": 1042
},
{
"epoch": 0.13283028479551715,
"grad_norm": 5.433468782261706,
"learning_rate": 1.9930733045163525e-05,
"loss": 0.9844,
"step": 1043
},
{
"epoch": 0.13295763885572376,
"grad_norm": 6.700120372544943,
"learning_rate": 1.9930490498117035e-05,
"loss": 1.0511,
"step": 1044
},
{
"epoch": 0.1330849929159304,
"grad_norm": 6.7738719831634775,
"learning_rate": 1.9930247528638768e-05,
"loss": 0.9928,
"step": 1045
},
{
"epoch": 0.13321234697613704,
"grad_norm": 6.388895187743395,
"learning_rate": 1.9930004136739058e-05,
"loss": 0.9695,
"step": 1046
},
{
"epoch": 0.13333970103634366,
"grad_norm": 5.2922802466753955,
"learning_rate": 1.9929760322428256e-05,
"loss": 0.9448,
"step": 1047
},
{
"epoch": 0.1334670550965503,
"grad_norm": 5.865390141642061,
"learning_rate": 1.9929516085716736e-05,
"loss": 1.1244,
"step": 1048
},
{
"epoch": 0.13359440915675694,
"grad_norm": 5.007676716360697,
"learning_rate": 1.992927142661489e-05,
"loss": 0.9693,
"step": 1049
},
{
"epoch": 0.13372176321696355,
"grad_norm": 5.057958001797684,
"learning_rate": 1.992902634513312e-05,
"loss": 0.949,
"step": 1050
},
{
"epoch": 0.1338491172771702,
"grad_norm": 4.937284440755442,
"learning_rate": 1.9928780841281858e-05,
"loss": 1.0291,
"step": 1051
},
{
"epoch": 0.13397647133737683,
"grad_norm": 5.169117541086143,
"learning_rate": 1.9928534915071543e-05,
"loss": 1.0791,
"step": 1052
},
{
"epoch": 0.13410382539758345,
"grad_norm": 4.1750885091753736,
"learning_rate": 1.9928288566512638e-05,
"loss": 1.0141,
"step": 1053
},
{
"epoch": 0.1342311794577901,
"grad_norm": 7.6227363603328575,
"learning_rate": 1.9928041795615616e-05,
"loss": 0.9137,
"step": 1054
},
{
"epoch": 0.13435853351799673,
"grad_norm": 5.799733570161933,
"learning_rate": 1.992779460239099e-05,
"loss": 1.039,
"step": 1055
},
{
"epoch": 0.13448588757820334,
"grad_norm": 5.177060573023828,
"learning_rate": 1.9927546986849258e-05,
"loss": 0.977,
"step": 1056
},
{
"epoch": 0.13461324163841,
"grad_norm": 7.3578280951204444,
"learning_rate": 1.9927298949000965e-05,
"loss": 0.9648,
"step": 1057
},
{
"epoch": 0.13474059569861663,
"grad_norm": 4.946775198994267,
"learning_rate": 1.9927050488856657e-05,
"loss": 1.1389,
"step": 1058
},
{
"epoch": 0.13486794975882324,
"grad_norm": 6.061502590914894,
"learning_rate": 1.9926801606426906e-05,
"loss": 1.0901,
"step": 1059
},
{
"epoch": 0.13499530381902988,
"grad_norm": 5.923565706250715,
"learning_rate": 1.992655230172229e-05,
"loss": 1.0891,
"step": 1060
},
{
"epoch": 0.13512265787923652,
"grad_norm": 4.96207252794465,
"learning_rate": 1.992630257475343e-05,
"loss": 0.9961,
"step": 1061
},
{
"epoch": 0.13525001193944314,
"grad_norm": 7.651995788732587,
"learning_rate": 1.9926052425530936e-05,
"loss": 0.9719,
"step": 1062
},
{
"epoch": 0.13537736599964978,
"grad_norm": 6.374845016689928,
"learning_rate": 1.9925801854065456e-05,
"loss": 0.9967,
"step": 1063
},
{
"epoch": 0.13550472005985642,
"grad_norm": 5.183903929938708,
"learning_rate": 1.9925550860367646e-05,
"loss": 0.986,
"step": 1064
},
{
"epoch": 0.13563207412006303,
"grad_norm": 5.011184200046132,
"learning_rate": 1.9925299444448183e-05,
"loss": 0.9907,
"step": 1065
},
{
"epoch": 0.13575942818026968,
"grad_norm": 6.122976786381155,
"learning_rate": 1.9925047606317766e-05,
"loss": 1.0059,
"step": 1066
},
{
"epoch": 0.13588678224047632,
"grad_norm": 5.423363560969625,
"learning_rate": 1.9924795345987103e-05,
"loss": 1.0698,
"step": 1067
},
{
"epoch": 0.13601413630068293,
"grad_norm": 5.226175473513691,
"learning_rate": 1.9924542663466925e-05,
"loss": 1.0515,
"step": 1068
},
{
"epoch": 0.13614149036088957,
"grad_norm": 6.246310861904402,
"learning_rate": 1.9924289558767982e-05,
"loss": 1.096,
"step": 1069
},
{
"epoch": 0.1362688444210962,
"grad_norm": 6.652299021796196,
"learning_rate": 1.9924036031901042e-05,
"loss": 1.0495,
"step": 1070
},
{
"epoch": 0.13639619848130283,
"grad_norm": 5.8620075466350805,
"learning_rate": 1.992378208287689e-05,
"loss": 0.9825,
"step": 1071
},
{
"epoch": 0.13652355254150947,
"grad_norm": 6.050447734312274,
"learning_rate": 1.992352771170633e-05,
"loss": 0.9908,
"step": 1072
},
{
"epoch": 0.13665090660171608,
"grad_norm": 5.240944454895212,
"learning_rate": 1.9923272918400175e-05,
"loss": 0.9909,
"step": 1073
},
{
"epoch": 0.13677826066192272,
"grad_norm": 5.091152909868271,
"learning_rate": 1.992301770296927e-05,
"loss": 1.0386,
"step": 1074
},
{
"epoch": 0.13690561472212937,
"grad_norm": 5.6087775377108064,
"learning_rate": 1.9922762065424474e-05,
"loss": 1.0048,
"step": 1075
},
{
"epoch": 0.13703296878233598,
"grad_norm": 6.818629036310338,
"learning_rate": 1.992250600577665e-05,
"loss": 0.9762,
"step": 1076
},
{
"epoch": 0.13716032284254262,
"grad_norm": 6.777493187743689,
"learning_rate": 1.9922249524036704e-05,
"loss": 1.1636,
"step": 1077
},
{
"epoch": 0.13728767690274926,
"grad_norm": 4.841139354431293,
"learning_rate": 1.9921992620215544e-05,
"loss": 1.0353,
"step": 1078
},
{
"epoch": 0.13741503096295588,
"grad_norm": 4.447936344624334,
"learning_rate": 1.992173529432409e-05,
"loss": 1.0011,
"step": 1079
},
{
"epoch": 0.13754238502316252,
"grad_norm": 7.258787157460566,
"learning_rate": 1.9921477546373296e-05,
"loss": 0.9964,
"step": 1080
},
{
"epoch": 0.13766973908336916,
"grad_norm": 6.816925011599404,
"learning_rate": 1.9921219376374123e-05,
"loss": 0.8766,
"step": 1081
},
{
"epoch": 0.13779709314357577,
"grad_norm": 5.154850070375168,
"learning_rate": 1.9920960784337552e-05,
"loss": 1.0945,
"step": 1082
},
{
"epoch": 0.1379244472037824,
"grad_norm": 3.4197004541756377,
"learning_rate": 1.992070177027459e-05,
"loss": 0.96,
"step": 1083
},
{
"epoch": 0.13805180126398905,
"grad_norm": 4.389033453340384,
"learning_rate": 1.9920442334196248e-05,
"loss": 1.0093,
"step": 1084
},
{
"epoch": 0.13817915532419567,
"grad_norm": 5.624768231173525,
"learning_rate": 1.9920182476113564e-05,
"loss": 0.9758,
"step": 1085
},
{
"epoch": 0.1383065093844023,
"grad_norm": 4.090395885605275,
"learning_rate": 1.9919922196037596e-05,
"loss": 0.984,
"step": 1086
},
{
"epoch": 0.13843386344460895,
"grad_norm": 4.890555384326312,
"learning_rate": 1.9919661493979408e-05,
"loss": 1.0094,
"step": 1087
},
{
"epoch": 0.13856121750481556,
"grad_norm": 9.665373310076777,
"learning_rate": 1.99194003699501e-05,
"loss": 1.0565,
"step": 1088
},
{
"epoch": 0.1386885715650222,
"grad_norm": 5.37681690823644,
"learning_rate": 1.991913882396077e-05,
"loss": 0.9501,
"step": 1089
},
{
"epoch": 0.13881592562522885,
"grad_norm": 4.849219402929612,
"learning_rate": 1.991887685602255e-05,
"loss": 0.9674,
"step": 1090
},
{
"epoch": 0.13894327968543546,
"grad_norm": 3.6688144346749265,
"learning_rate": 1.9918614466146586e-05,
"loss": 1.0883,
"step": 1091
},
{
"epoch": 0.1390706337456421,
"grad_norm": 6.9702190681083955,
"learning_rate": 1.9918351654344033e-05,
"loss": 1.0757,
"step": 1092
},
{
"epoch": 0.13919798780584874,
"grad_norm": 4.918950450259129,
"learning_rate": 1.991808842062607e-05,
"loss": 1.042,
"step": 1093
},
{
"epoch": 0.13932534186605536,
"grad_norm": 4.6766149658767615,
"learning_rate": 1.9917824765003905e-05,
"loss": 1.0009,
"step": 1094
},
{
"epoch": 0.139452695926262,
"grad_norm": 5.286134165579219,
"learning_rate": 1.9917560687488743e-05,
"loss": 0.9648,
"step": 1095
},
{
"epoch": 0.13958004998646864,
"grad_norm": 4.8150572484813035,
"learning_rate": 1.9917296188091823e-05,
"loss": 0.9133,
"step": 1096
},
{
"epoch": 0.13970740404667525,
"grad_norm": 5.7510746490669575,
"learning_rate": 1.9917031266824395e-05,
"loss": 0.8875,
"step": 1097
},
{
"epoch": 0.1398347581068819,
"grad_norm": 5.203916380671531,
"learning_rate": 1.991676592369773e-05,
"loss": 0.9772,
"step": 1098
},
{
"epoch": 0.13996211216708854,
"grad_norm": 4.728170729754482,
"learning_rate": 1.991650015872311e-05,
"loss": 1.087,
"step": 1099
},
{
"epoch": 0.14008946622729515,
"grad_norm": 4.143056479483173,
"learning_rate": 1.991623397191185e-05,
"loss": 1.0218,
"step": 1100
},
{
"epoch": 0.1402168202875018,
"grad_norm": 4.069189579388974,
"learning_rate": 1.9915967363275264e-05,
"loss": 0.9937,
"step": 1101
},
{
"epoch": 0.14034417434770843,
"grad_norm": 9.021029169953739,
"learning_rate": 1.9915700332824696e-05,
"loss": 0.9962,
"step": 1102
},
{
"epoch": 0.14047152840791505,
"grad_norm": 5.2339468254213175,
"learning_rate": 1.9915432880571508e-05,
"loss": 1.0367,
"step": 1103
},
{
"epoch": 0.1405988824681217,
"grad_norm": 5.262071083693544,
"learning_rate": 1.9915165006527076e-05,
"loss": 1.0258,
"step": 1104
},
{
"epoch": 0.14072623652832833,
"grad_norm": 5.158737784672319,
"learning_rate": 1.991489671070279e-05,
"loss": 0.9628,
"step": 1105
},
{
"epoch": 0.14085359058853494,
"grad_norm": 4.923590079715338,
"learning_rate": 1.9914627993110072e-05,
"loss": 0.9043,
"step": 1106
},
{
"epoch": 0.14098094464874159,
"grad_norm": 5.1952724828111085,
"learning_rate": 1.9914358853760346e-05,
"loss": 0.9633,
"step": 1107
},
{
"epoch": 0.14110829870894823,
"grad_norm": 4.745809212149704,
"learning_rate": 1.9914089292665065e-05,
"loss": 0.9792,
"step": 1108
},
{
"epoch": 0.14123565276915484,
"grad_norm": 5.416111413911312,
"learning_rate": 1.991381930983569e-05,
"loss": 1.0286,
"step": 1109
},
{
"epoch": 0.14136300682936148,
"grad_norm": 5.574909726083418,
"learning_rate": 1.9913548905283714e-05,
"loss": 1.083,
"step": 1110
},
{
"epoch": 0.14149036088956812,
"grad_norm": 3.9528528655095663,
"learning_rate": 1.9913278079020633e-05,
"loss": 0.9055,
"step": 1111
},
{
"epoch": 0.14161771494977474,
"grad_norm": 6.877030228906015,
"learning_rate": 1.9913006831057967e-05,
"loss": 1.0158,
"step": 1112
},
{
"epoch": 0.14174506900998138,
"grad_norm": 5.337862217782534,
"learning_rate": 1.9912735161407264e-05,
"loss": 1.0859,
"step": 1113
},
{
"epoch": 0.14187242307018802,
"grad_norm": 5.764308872267546,
"learning_rate": 1.991246307008007e-05,
"loss": 0.9251,
"step": 1114
},
{
"epoch": 0.14199977713039463,
"grad_norm": 5.64803125970419,
"learning_rate": 1.9912190557087964e-05,
"loss": 0.9771,
"step": 1115
},
{
"epoch": 0.14212713119060127,
"grad_norm": 4.898251762701724,
"learning_rate": 1.991191762244254e-05,
"loss": 1.0447,
"step": 1116
},
{
"epoch": 0.14225448525080792,
"grad_norm": 6.234584565811258,
"learning_rate": 1.9911644266155402e-05,
"loss": 1.0607,
"step": 1117
},
{
"epoch": 0.14238183931101453,
"grad_norm": 5.266679154554281,
"learning_rate": 1.9911370488238185e-05,
"loss": 0.9349,
"step": 1118
},
{
"epoch": 0.14250919337122117,
"grad_norm": 6.256153586880418,
"learning_rate": 1.9911096288702532e-05,
"loss": 1.2035,
"step": 1119
},
{
"epoch": 0.14263654743142778,
"grad_norm": 6.450600044668659,
"learning_rate": 1.9910821667560106e-05,
"loss": 1.0081,
"step": 1120
},
{
"epoch": 0.14276390149163443,
"grad_norm": 3.8362862919805774,
"learning_rate": 1.9910546624822596e-05,
"loss": 1.0396,
"step": 1121
},
{
"epoch": 0.14289125555184107,
"grad_norm": 5.81600157570588,
"learning_rate": 1.9910271160501694e-05,
"loss": 0.9653,
"step": 1122
},
{
"epoch": 0.14301860961204768,
"grad_norm": 4.879159571085699,
"learning_rate": 1.990999527460912e-05,
"loss": 0.9589,
"step": 1123
},
{
"epoch": 0.14314596367225432,
"grad_norm": 6.8926005396879795,
"learning_rate": 1.990971896715661e-05,
"loss": 1.0908,
"step": 1124
},
{
"epoch": 0.14327331773246096,
"grad_norm": 7.01316195540877,
"learning_rate": 1.990944223815592e-05,
"loss": 1.0536,
"step": 1125
},
{
"epoch": 0.14340067179266758,
"grad_norm": 4.769891348826958,
"learning_rate": 1.990916508761882e-05,
"loss": 1.0685,
"step": 1126
},
{
"epoch": 0.14352802585287422,
"grad_norm": 5.274988017136401,
"learning_rate": 1.9908887515557103e-05,
"loss": 0.9612,
"step": 1127
},
{
"epoch": 0.14365537991308086,
"grad_norm": 4.753994918159333,
"learning_rate": 1.990860952198257e-05,
"loss": 1.0291,
"step": 1128
},
{
"epoch": 0.14378273397328747,
"grad_norm": 4.599970958452185,
"learning_rate": 1.990833110690705e-05,
"loss": 1.0102,
"step": 1129
},
{
"epoch": 0.14391008803349412,
"grad_norm": 5.262607513561777,
"learning_rate": 1.990805227034239e-05,
"loss": 0.9505,
"step": 1130
},
{
"epoch": 0.14403744209370076,
"grad_norm": 4.750430438967478,
"learning_rate": 1.9907773012300442e-05,
"loss": 1.1432,
"step": 1131
},
{
"epoch": 0.14416479615390737,
"grad_norm": 5.5534318022376485,
"learning_rate": 1.99074933327931e-05,
"loss": 1.0166,
"step": 1132
},
{
"epoch": 0.144292150214114,
"grad_norm": 4.36148381243615,
"learning_rate": 1.9907213231832244e-05,
"loss": 0.9515,
"step": 1133
},
{
"epoch": 0.14441950427432065,
"grad_norm": 5.396342465843601,
"learning_rate": 1.99069327094298e-05,
"loss": 0.988,
"step": 1134
},
{
"epoch": 0.14454685833452727,
"grad_norm": 4.160370125444187,
"learning_rate": 1.99066517655977e-05,
"loss": 0.9402,
"step": 1135
},
{
"epoch": 0.1446742123947339,
"grad_norm": 4.221462705684877,
"learning_rate": 1.990637040034789e-05,
"loss": 0.923,
"step": 1136
},
{
"epoch": 0.14480156645494055,
"grad_norm": 6.253117203600368,
"learning_rate": 1.9906088613692348e-05,
"loss": 1.0339,
"step": 1137
},
{
"epoch": 0.14492892051514716,
"grad_norm": 6.017093469305539,
"learning_rate": 1.9905806405643053e-05,
"loss": 1.0136,
"step": 1138
},
{
"epoch": 0.1450562745753538,
"grad_norm": 4.399259175443589,
"learning_rate": 1.990552377621201e-05,
"loss": 1.0272,
"step": 1139
},
{
"epoch": 0.14518362863556045,
"grad_norm": 4.547342282958136,
"learning_rate": 1.990524072541125e-05,
"loss": 0.9937,
"step": 1140
},
{
"epoch": 0.14531098269576706,
"grad_norm": 4.855709152837835,
"learning_rate": 1.9904957253252804e-05,
"loss": 0.9244,
"step": 1141
},
{
"epoch": 0.1454383367559737,
"grad_norm": 5.5757047814984615,
"learning_rate": 1.9904673359748735e-05,
"loss": 0.9908,
"step": 1142
},
{
"epoch": 0.14556569081618034,
"grad_norm": 6.043856870540701,
"learning_rate": 1.9904389044911122e-05,
"loss": 1.0196,
"step": 1143
},
{
"epoch": 0.14569304487638696,
"grad_norm": 4.801416490086993,
"learning_rate": 1.9904104308752053e-05,
"loss": 1.0553,
"step": 1144
},
{
"epoch": 0.1458203989365936,
"grad_norm": 5.379496091195757,
"learning_rate": 1.9903819151283645e-05,
"loss": 1.0304,
"step": 1145
},
{
"epoch": 0.14594775299680024,
"grad_norm": 4.545430481586915,
"learning_rate": 1.9903533572518026e-05,
"loss": 0.9908,
"step": 1146
},
{
"epoch": 0.14607510705700685,
"grad_norm": 5.220752628775305,
"learning_rate": 1.9903247572467344e-05,
"loss": 1.1251,
"step": 1147
},
{
"epoch": 0.1462024611172135,
"grad_norm": 6.674829190815485,
"learning_rate": 1.990296115114377e-05,
"loss": 1.0095,
"step": 1148
},
{
"epoch": 0.14632981517742014,
"grad_norm": 5.507004179382678,
"learning_rate": 1.9902674308559483e-05,
"loss": 1.0024,
"step": 1149
},
{
"epoch": 0.14645716923762675,
"grad_norm": 4.71990972755875,
"learning_rate": 1.9902387044726686e-05,
"loss": 0.9865,
"step": 1150
},
{
"epoch": 0.1465845232978334,
"grad_norm": 4.478833113678234,
"learning_rate": 1.9902099359657597e-05,
"loss": 0.9298,
"step": 1151
},
{
"epoch": 0.14671187735804003,
"grad_norm": 6.364648530400351,
"learning_rate": 1.9901811253364458e-05,
"loss": 1.0776,
"step": 1152
},
{
"epoch": 0.14683923141824665,
"grad_norm": 6.283898853112022,
"learning_rate": 1.9901522725859523e-05,
"loss": 1.0698,
"step": 1153
},
{
"epoch": 0.1469665854784533,
"grad_norm": 4.844800872017544,
"learning_rate": 1.9901233777155062e-05,
"loss": 0.8839,
"step": 1154
},
{
"epoch": 0.14709393953865993,
"grad_norm": 6.435025526587445,
"learning_rate": 1.9900944407263373e-05,
"loss": 1.0899,
"step": 1155
},
{
"epoch": 0.14722129359886654,
"grad_norm": 6.109479898396119,
"learning_rate": 1.9900654616196765e-05,
"loss": 1.0966,
"step": 1156
},
{
"epoch": 0.14734864765907318,
"grad_norm": 4.981320392872287,
"learning_rate": 1.9900364403967555e-05,
"loss": 0.9802,
"step": 1157
},
{
"epoch": 0.14747600171927983,
"grad_norm": 6.337664612206646,
"learning_rate": 1.9900073770588104e-05,
"loss": 0.9783,
"step": 1158
},
{
"epoch": 0.14760335577948644,
"grad_norm": 7.134358467286747,
"learning_rate": 1.9899782716070764e-05,
"loss": 1.0323,
"step": 1159
},
{
"epoch": 0.14773070983969308,
"grad_norm": 6.583584293781033,
"learning_rate": 1.9899491240427917e-05,
"loss": 1.0655,
"step": 1160
},
{
"epoch": 0.14785806389989972,
"grad_norm": 7.120906523954685,
"learning_rate": 1.9899199343671968e-05,
"loss": 1.0935,
"step": 1161
},
{
"epoch": 0.14798541796010634,
"grad_norm": 5.433366980575255,
"learning_rate": 1.9898907025815327e-05,
"loss": 0.9979,
"step": 1162
},
{
"epoch": 0.14811277202031298,
"grad_norm": 6.2705095615215205,
"learning_rate": 1.9898614286870433e-05,
"loss": 1.1334,
"step": 1163
},
{
"epoch": 0.1482401260805196,
"grad_norm": 4.831201455882813,
"learning_rate": 1.9898321126849743e-05,
"loss": 1.0073,
"step": 1164
},
{
"epoch": 0.14836748014072623,
"grad_norm": 4.672358765322368,
"learning_rate": 1.9898027545765715e-05,
"loss": 1.0821,
"step": 1165
},
{
"epoch": 0.14849483420093287,
"grad_norm": 7.180955909703082,
"learning_rate": 1.989773354363085e-05,
"loss": 0.9288,
"step": 1166
},
{
"epoch": 0.1486221882611395,
"grad_norm": 4.403280569757165,
"learning_rate": 1.989743912045765e-05,
"loss": 0.961,
"step": 1167
},
{
"epoch": 0.14874954232134613,
"grad_norm": 8.908893607791116,
"learning_rate": 1.9897144276258637e-05,
"loss": 1.0988,
"step": 1168
},
{
"epoch": 0.14887689638155277,
"grad_norm": 4.88691858604188,
"learning_rate": 1.9896849011046356e-05,
"loss": 0.9468,
"step": 1169
},
{
"epoch": 0.14900425044175938,
"grad_norm": 4.439021534715612,
"learning_rate": 1.989655332483337e-05,
"loss": 1.09,
"step": 1170
},
{
"epoch": 0.14913160450196603,
"grad_norm": 4.584781117513443,
"learning_rate": 1.989625721763225e-05,
"loss": 1.1378,
"step": 1171
},
{
"epoch": 0.14925895856217267,
"grad_norm": 6.829460380092409,
"learning_rate": 1.9895960689455598e-05,
"loss": 1.0116,
"step": 1172
},
{
"epoch": 0.14938631262237928,
"grad_norm": 5.925553037691101,
"learning_rate": 1.9895663740316027e-05,
"loss": 0.9667,
"step": 1173
},
{
"epoch": 0.14951366668258592,
"grad_norm": 5.650634169645478,
"learning_rate": 1.9895366370226164e-05,
"loss": 1.0056,
"step": 1174
},
{
"epoch": 0.14964102074279256,
"grad_norm": 6.303749288830858,
"learning_rate": 1.9895068579198667e-05,
"loss": 0.9515,
"step": 1175
},
{
"epoch": 0.14976837480299918,
"grad_norm": 4.734757511813959,
"learning_rate": 1.9894770367246197e-05,
"loss": 1.0735,
"step": 1176
},
{
"epoch": 0.14989572886320582,
"grad_norm": 6.024659029227881,
"learning_rate": 1.9894471734381443e-05,
"loss": 1.086,
"step": 1177
},
{
"epoch": 0.15002308292341246,
"grad_norm": 7.097352579325189,
"learning_rate": 1.989417268061711e-05,
"loss": 0.951,
"step": 1178
},
{
"epoch": 0.15015043698361907,
"grad_norm": 5.378720969428306,
"learning_rate": 1.989387320596591e-05,
"loss": 1.0527,
"step": 1179
},
{
"epoch": 0.15027779104382571,
"grad_norm": 5.978861447650805,
"learning_rate": 1.9893573310440592e-05,
"loss": 1.0597,
"step": 1180
},
{
"epoch": 0.15040514510403236,
"grad_norm": 8.018983252708948,
"learning_rate": 1.989327299405391e-05,
"loss": 0.8077,
"step": 1181
},
{
"epoch": 0.15053249916423897,
"grad_norm": 5.923675182970437,
"learning_rate": 1.9892972256818642e-05,
"loss": 1.0515,
"step": 1182
},
{
"epoch": 0.1506598532244456,
"grad_norm": 5.0075340729200315,
"learning_rate": 1.989267109874758e-05,
"loss": 0.9554,
"step": 1183
},
{
"epoch": 0.15078720728465225,
"grad_norm": 4.6758452931338805,
"learning_rate": 1.989236951985353e-05,
"loss": 0.9426,
"step": 1184
},
{
"epoch": 0.15091456134485887,
"grad_norm": 4.204069010290642,
"learning_rate": 1.9892067520149325e-05,
"loss": 1.0049,
"step": 1185
},
{
"epoch": 0.1510419154050655,
"grad_norm": 5.048848336214381,
"learning_rate": 1.989176509964781e-05,
"loss": 1.0377,
"step": 1186
},
{
"epoch": 0.15116926946527215,
"grad_norm": 6.138944051196484,
"learning_rate": 1.9891462258361854e-05,
"loss": 0.985,
"step": 1187
},
{
"epoch": 0.15129662352547876,
"grad_norm": 5.009658449450558,
"learning_rate": 1.9891158996304332e-05,
"loss": 1.0061,
"step": 1188
},
{
"epoch": 0.1514239775856854,
"grad_norm": 5.76423574524903,
"learning_rate": 1.989085531348815e-05,
"loss": 1.1038,
"step": 1189
},
{
"epoch": 0.15155133164589205,
"grad_norm": 22.6411250699605,
"learning_rate": 1.9890551209926228e-05,
"loss": 1.0009,
"step": 1190
},
{
"epoch": 0.15167868570609866,
"grad_norm": 7.817797032377672,
"learning_rate": 1.9890246685631497e-05,
"loss": 1.0513,
"step": 1191
},
{
"epoch": 0.1518060397663053,
"grad_norm": 5.7520894110552065,
"learning_rate": 1.9889941740616915e-05,
"loss": 0.9986,
"step": 1192
},
{
"epoch": 0.15193339382651194,
"grad_norm": 18.699414183880535,
"learning_rate": 1.988963637489545e-05,
"loss": 1.0264,
"step": 1193
},
{
"epoch": 0.15206074788671856,
"grad_norm": 4.828371109600161,
"learning_rate": 1.9889330588480092e-05,
"loss": 1.1089,
"step": 1194
},
{
"epoch": 0.1521881019469252,
"grad_norm": 5.636301851353449,
"learning_rate": 1.9889024381383853e-05,
"loss": 1.1113,
"step": 1195
},
{
"epoch": 0.15231545600713184,
"grad_norm": 6.401306924892736,
"learning_rate": 1.9888717753619756e-05,
"loss": 1.0345,
"step": 1196
},
{
"epoch": 0.15244281006733845,
"grad_norm": 4.8383120621596785,
"learning_rate": 1.988841070520085e-05,
"loss": 1.0335,
"step": 1197
},
{
"epoch": 0.1525701641275451,
"grad_norm": 4.596736430265082,
"learning_rate": 1.9888103236140187e-05,
"loss": 0.9814,
"step": 1198
},
{
"epoch": 0.15269751818775174,
"grad_norm": 5.860190047479529,
"learning_rate": 1.988779534645085e-05,
"loss": 1.0338,
"step": 1199
},
{
"epoch": 0.15282487224795835,
"grad_norm": 8.045167730751048,
"learning_rate": 1.9887487036145942e-05,
"loss": 0.9122,
"step": 1200
},
{
"epoch": 0.152952226308165,
"grad_norm": 5.64218312561491,
"learning_rate": 1.988717830523857e-05,
"loss": 1.0578,
"step": 1201
},
{
"epoch": 0.15307958036837163,
"grad_norm": 5.637619781913279,
"learning_rate": 1.9886869153741873e-05,
"loss": 1.0015,
"step": 1202
},
{
"epoch": 0.15320693442857825,
"grad_norm": 5.581245639367872,
"learning_rate": 1.9886559581669e-05,
"loss": 1.0501,
"step": 1203
},
{
"epoch": 0.1533342884887849,
"grad_norm": 4.895855236680835,
"learning_rate": 1.9886249589033115e-05,
"loss": 1.0239,
"step": 1204
},
{
"epoch": 0.15346164254899153,
"grad_norm": 6.5685870323690345,
"learning_rate": 1.988593917584741e-05,
"loss": 1.0942,
"step": 1205
},
{
"epoch": 0.15358899660919814,
"grad_norm": 8.102356330141617,
"learning_rate": 1.9885628342125093e-05,
"loss": 1.093,
"step": 1206
},
{
"epoch": 0.15371635066940478,
"grad_norm": 4.9198943497393985,
"learning_rate": 1.9885317087879378e-05,
"loss": 1.049,
"step": 1207
},
{
"epoch": 0.15384370472961142,
"grad_norm": 4.840514055806769,
"learning_rate": 1.9885005413123515e-05,
"loss": 0.9917,
"step": 1208
},
{
"epoch": 0.15397105878981804,
"grad_norm": 5.329710962069987,
"learning_rate": 1.9884693317870754e-05,
"loss": 0.8854,
"step": 1209
},
{
"epoch": 0.15409841285002468,
"grad_norm": 6.40699250174998,
"learning_rate": 1.9884380802134374e-05,
"loss": 0.9925,
"step": 1210
},
{
"epoch": 0.1542257669102313,
"grad_norm": 4.920757832972157,
"learning_rate": 1.988406786592767e-05,
"loss": 0.959,
"step": 1211
},
{
"epoch": 0.15435312097043793,
"grad_norm": 5.794855058912914,
"learning_rate": 1.9883754509263952e-05,
"loss": 1.1196,
"step": 1212
},
{
"epoch": 0.15448047503064458,
"grad_norm": 5.76384991041534,
"learning_rate": 1.9883440732156553e-05,
"loss": 1.0452,
"step": 1213
},
{
"epoch": 0.1546078290908512,
"grad_norm": 5.213327114429976,
"learning_rate": 1.9883126534618818e-05,
"loss": 0.9344,
"step": 1214
},
{
"epoch": 0.15473518315105783,
"grad_norm": 6.551134712996828,
"learning_rate": 1.988281191666411e-05,
"loss": 1.0484,
"step": 1215
},
{
"epoch": 0.15486253721126447,
"grad_norm": 4.463543399675435,
"learning_rate": 1.988249687830582e-05,
"loss": 0.9069,
"step": 1216
},
{
"epoch": 0.1549898912714711,
"grad_norm": 4.702098037275322,
"learning_rate": 1.9882181419557342e-05,
"loss": 1.0329,
"step": 1217
},
{
"epoch": 0.15511724533167773,
"grad_norm": 5.438172923574985,
"learning_rate": 1.9881865540432104e-05,
"loss": 0.9534,
"step": 1218
},
{
"epoch": 0.15524459939188437,
"grad_norm": 5.295644431518071,
"learning_rate": 1.9881549240943533e-05,
"loss": 1.0289,
"step": 1219
},
{
"epoch": 0.15537195345209098,
"grad_norm": 4.500410485089305,
"learning_rate": 1.988123252110509e-05,
"loss": 1.0132,
"step": 1220
},
{
"epoch": 0.15549930751229762,
"grad_norm": 4.585819379334822,
"learning_rate": 1.9880915380930245e-05,
"loss": 1.0024,
"step": 1221
},
{
"epoch": 0.15562666157250427,
"grad_norm": 7.193516344506464,
"learning_rate": 1.9880597820432493e-05,
"loss": 0.9475,
"step": 1222
},
{
"epoch": 0.15575401563271088,
"grad_norm": 5.1443825760360244,
"learning_rate": 1.988027983962534e-05,
"loss": 1.0516,
"step": 1223
},
{
"epoch": 0.15588136969291752,
"grad_norm": 5.090082076634423,
"learning_rate": 1.9879961438522312e-05,
"loss": 0.966,
"step": 1224
},
{
"epoch": 0.15600872375312416,
"grad_norm": 4.790525527190118,
"learning_rate": 1.987964261713695e-05,
"loss": 0.8954,
"step": 1225
},
{
"epoch": 0.15613607781333078,
"grad_norm": 5.236927104927023,
"learning_rate": 1.9879323375482825e-05,
"loss": 1.0287,
"step": 1226
},
{
"epoch": 0.15626343187353742,
"grad_norm": 5.950369610245666,
"learning_rate": 1.987900371357351e-05,
"loss": 1.0874,
"step": 1227
},
{
"epoch": 0.15639078593374406,
"grad_norm": 5.524607144711156,
"learning_rate": 1.9878683631422605e-05,
"loss": 0.9406,
"step": 1228
},
{
"epoch": 0.15651813999395067,
"grad_norm": 5.079437180226869,
"learning_rate": 1.987836312904373e-05,
"loss": 1.039,
"step": 1229
},
{
"epoch": 0.1566454940541573,
"grad_norm": 6.124208048789126,
"learning_rate": 1.9878042206450515e-05,
"loss": 0.9518,
"step": 1230
},
{
"epoch": 0.15677284811436396,
"grad_norm": 4.603658564682635,
"learning_rate": 1.9877720863656605e-05,
"loss": 0.9797,
"step": 1231
},
{
"epoch": 0.15690020217457057,
"grad_norm": 5.297661376914641,
"learning_rate": 1.9877399100675684e-05,
"loss": 1.0236,
"step": 1232
},
{
"epoch": 0.1570275562347772,
"grad_norm": 5.3680721259125255,
"learning_rate": 1.987707691752143e-05,
"loss": 0.9727,
"step": 1233
},
{
"epoch": 0.15715491029498385,
"grad_norm": 5.842960079883927,
"learning_rate": 1.987675431420755e-05,
"loss": 0.9885,
"step": 1234
},
{
"epoch": 0.15728226435519047,
"grad_norm": 3.570953778995695,
"learning_rate": 1.9876431290747766e-05,
"loss": 0.9444,
"step": 1235
},
{
"epoch": 0.1574096184153971,
"grad_norm": 4.451910556028384,
"learning_rate": 1.987610784715582e-05,
"loss": 0.9464,
"step": 1236
},
{
"epoch": 0.15753697247560375,
"grad_norm": 5.669527939617677,
"learning_rate": 1.9875783983445473e-05,
"loss": 0.9582,
"step": 1237
},
{
"epoch": 0.15766432653581036,
"grad_norm": 5.212123844470004,
"learning_rate": 1.9875459699630503e-05,
"loss": 0.9536,
"step": 1238
},
{
"epoch": 0.157791680596017,
"grad_norm": 5.681033304276333,
"learning_rate": 1.9875134995724697e-05,
"loss": 0.9886,
"step": 1239
},
{
"epoch": 0.15791903465622364,
"grad_norm": 4.781488350306441,
"learning_rate": 1.9874809871741877e-05,
"loss": 1.0747,
"step": 1240
},
{
"epoch": 0.15804638871643026,
"grad_norm": 5.27408922477416,
"learning_rate": 1.9874484327695862e-05,
"loss": 1.0797,
"step": 1241
},
{
"epoch": 0.1581737427766369,
"grad_norm": 5.110825638930353,
"learning_rate": 1.9874158363600513e-05,
"loss": 0.9206,
"step": 1242
},
{
"epoch": 0.15830109683684354,
"grad_norm": 5.286355163870387,
"learning_rate": 1.9873831979469687e-05,
"loss": 0.9544,
"step": 1243
},
{
"epoch": 0.15842845089705015,
"grad_norm": 5.876306266847982,
"learning_rate": 1.9873505175317272e-05,
"loss": 0.9467,
"step": 1244
},
{
"epoch": 0.1585558049572568,
"grad_norm": 3.8290624599508156,
"learning_rate": 1.987317795115717e-05,
"loss": 0.99,
"step": 1245
},
{
"epoch": 0.15868315901746344,
"grad_norm": 6.334550171788543,
"learning_rate": 1.98728503070033e-05,
"loss": 1.0207,
"step": 1246
},
{
"epoch": 0.15881051307767005,
"grad_norm": 5.201014001722949,
"learning_rate": 1.9872522242869598e-05,
"loss": 1.0032,
"step": 1247
},
{
"epoch": 0.1589378671378767,
"grad_norm": 5.339655506580464,
"learning_rate": 1.987219375877002e-05,
"loss": 1.0571,
"step": 1248
},
{
"epoch": 0.15906522119808333,
"grad_norm": 5.917438223951082,
"learning_rate": 1.9871864854718545e-05,
"loss": 0.9298,
"step": 1249
},
{
"epoch": 0.15919257525828995,
"grad_norm": 5.84159509430854,
"learning_rate": 1.9871535530729154e-05,
"loss": 1.1235,
"step": 1250
},
{
"epoch": 0.1593199293184966,
"grad_norm": 6.647444797729966,
"learning_rate": 1.9871205786815865e-05,
"loss": 1.0961,
"step": 1251
},
{
"epoch": 0.15944728337870323,
"grad_norm": 4.506879845567531,
"learning_rate": 1.9870875622992697e-05,
"loss": 1.0051,
"step": 1252
},
{
"epoch": 0.15957463743890984,
"grad_norm": 4.978751044171506,
"learning_rate": 1.9870545039273704e-05,
"loss": 0.9519,
"step": 1253
},
{
"epoch": 0.15970199149911649,
"grad_norm": 4.884386465067718,
"learning_rate": 1.9870214035672945e-05,
"loss": 1.0405,
"step": 1254
},
{
"epoch": 0.1598293455593231,
"grad_norm": 6.01301752032264,
"learning_rate": 1.9869882612204496e-05,
"loss": 0.9768,
"step": 1255
},
{
"epoch": 0.15995669961952974,
"grad_norm": 5.66422076826293,
"learning_rate": 1.986955076888246e-05,
"loss": 0.8278,
"step": 1256
},
{
"epoch": 0.16008405367973638,
"grad_norm": 6.38059333314379,
"learning_rate": 1.986921850572095e-05,
"loss": 0.9132,
"step": 1257
},
{
"epoch": 0.160211407739943,
"grad_norm": 5.136288561774867,
"learning_rate": 1.9868885822734104e-05,
"loss": 1.0093,
"step": 1258
},
{
"epoch": 0.16033876180014964,
"grad_norm": 6.1340803451624435,
"learning_rate": 1.986855271993607e-05,
"loss": 0.9994,
"step": 1259
},
{
"epoch": 0.16046611586035628,
"grad_norm": 6.610357838234759,
"learning_rate": 1.9868219197341024e-05,
"loss": 0.8919,
"step": 1260
},
{
"epoch": 0.1605934699205629,
"grad_norm": 5.068832393028328,
"learning_rate": 1.9867885254963147e-05,
"loss": 0.958,
"step": 1261
},
{
"epoch": 0.16072082398076953,
"grad_norm": 5.098652919809279,
"learning_rate": 1.9867550892816646e-05,
"loss": 1.0166,
"step": 1262
},
{
"epoch": 0.16084817804097618,
"grad_norm": 5.54546086727413,
"learning_rate": 1.9867216110915745e-05,
"loss": 0.9558,
"step": 1263
},
{
"epoch": 0.1609755321011828,
"grad_norm": 3.6506732894506726,
"learning_rate": 1.986688090927469e-05,
"loss": 0.9749,
"step": 1264
},
{
"epoch": 0.16110288616138943,
"grad_norm": 6.630011336603153,
"learning_rate": 1.9866545287907732e-05,
"loss": 1.087,
"step": 1265
},
{
"epoch": 0.16123024022159607,
"grad_norm": 5.1520144550405025,
"learning_rate": 1.9866209246829152e-05,
"loss": 1.0564,
"step": 1266
},
{
"epoch": 0.16135759428180269,
"grad_norm": 4.567456623073952,
"learning_rate": 1.9865872786053245e-05,
"loss": 0.9943,
"step": 1267
},
{
"epoch": 0.16148494834200933,
"grad_norm": 5.429662910241082,
"learning_rate": 1.9865535905594326e-05,
"loss": 1.0046,
"step": 1268
},
{
"epoch": 0.16161230240221597,
"grad_norm": 4.745719989140955,
"learning_rate": 1.986519860546672e-05,
"loss": 0.9511,
"step": 1269
},
{
"epoch": 0.16173965646242258,
"grad_norm": 4.907924293524879,
"learning_rate": 1.986486088568478e-05,
"loss": 0.947,
"step": 1270
},
{
"epoch": 0.16186701052262922,
"grad_norm": 5.93928420715077,
"learning_rate": 1.9864522746262867e-05,
"loss": 1.0879,
"step": 1271
},
{
"epoch": 0.16199436458283586,
"grad_norm": 4.6692044079111,
"learning_rate": 1.986418418721537e-05,
"loss": 0.9495,
"step": 1272
},
{
"epoch": 0.16212171864304248,
"grad_norm": 6.435001221651496,
"learning_rate": 1.986384520855669e-05,
"loss": 0.9653,
"step": 1273
},
{
"epoch": 0.16224907270324912,
"grad_norm": 4.8527476756639745,
"learning_rate": 1.9863505810301246e-05,
"loss": 1.0149,
"step": 1274
},
{
"epoch": 0.16237642676345576,
"grad_norm": 5.268659764215143,
"learning_rate": 1.9863165992463477e-05,
"loss": 0.988,
"step": 1275
},
{
"epoch": 0.16250378082366237,
"grad_norm": 6.928202404369275,
"learning_rate": 1.986282575505783e-05,
"loss": 1.0151,
"step": 1276
},
{
"epoch": 0.16263113488386902,
"grad_norm": 5.708832025208047,
"learning_rate": 1.9862485098098796e-05,
"loss": 0.986,
"step": 1277
},
{
"epoch": 0.16275848894407566,
"grad_norm": 4.293735314223316,
"learning_rate": 1.986214402160085e-05,
"loss": 0.9516,
"step": 1278
},
{
"epoch": 0.16288584300428227,
"grad_norm": 5.128487865527736,
"learning_rate": 1.9861802525578508e-05,
"loss": 0.9776,
"step": 1279
},
{
"epoch": 0.1630131970644889,
"grad_norm": 5.926519580527757,
"learning_rate": 1.986146061004629e-05,
"loss": 0.9454,
"step": 1280
},
{
"epoch": 0.16314055112469555,
"grad_norm": 4.709791762168808,
"learning_rate": 1.9861118275018755e-05,
"loss": 1.055,
"step": 1281
},
{
"epoch": 0.16326790518490217,
"grad_norm": 3.9447872001005515,
"learning_rate": 1.9860775520510453e-05,
"loss": 1.0019,
"step": 1282
},
{
"epoch": 0.1633952592451088,
"grad_norm": 4.657867150816275,
"learning_rate": 1.9860432346535966e-05,
"loss": 0.9158,
"step": 1283
},
{
"epoch": 0.16352261330531545,
"grad_norm": 6.558586770395333,
"learning_rate": 1.9860088753109896e-05,
"loss": 1.0146,
"step": 1284
},
{
"epoch": 0.16364996736552206,
"grad_norm": 6.790804131467284,
"learning_rate": 1.985974474024686e-05,
"loss": 1.1179,
"step": 1285
},
{
"epoch": 0.1637773214257287,
"grad_norm": 4.380667149497051,
"learning_rate": 1.9859400307961486e-05,
"loss": 1.029,
"step": 1286
},
{
"epoch": 0.16390467548593535,
"grad_norm": 6.457342584561743,
"learning_rate": 1.985905545626843e-05,
"loss": 1.0326,
"step": 1287
},
{
"epoch": 0.16403202954614196,
"grad_norm": 4.491102535418511,
"learning_rate": 1.985871018518236e-05,
"loss": 1.0289,
"step": 1288
},
{
"epoch": 0.1641593836063486,
"grad_norm": 5.129222154149322,
"learning_rate": 1.9858364494717966e-05,
"loss": 1.0228,
"step": 1289
},
{
"epoch": 0.16428673766655524,
"grad_norm": 5.162285434060186,
"learning_rate": 1.9858018384889946e-05,
"loss": 0.9045,
"step": 1290
},
{
"epoch": 0.16441409172676186,
"grad_norm": 5.6943603054539915,
"learning_rate": 1.9857671855713038e-05,
"loss": 0.9674,
"step": 1291
},
{
"epoch": 0.1645414457869685,
"grad_norm": 5.6519327284473695,
"learning_rate": 1.9857324907201966e-05,
"loss": 0.9993,
"step": 1292
},
{
"epoch": 0.16466879984717514,
"grad_norm": 5.529623880960854,
"learning_rate": 1.98569775393715e-05,
"loss": 0.973,
"step": 1293
},
{
"epoch": 0.16479615390738175,
"grad_norm": 4.634977146985319,
"learning_rate": 1.9856629752236413e-05,
"loss": 1.0061,
"step": 1294
},
{
"epoch": 0.1649235079675884,
"grad_norm": 5.395211300534669,
"learning_rate": 1.9856281545811497e-05,
"loss": 0.9469,
"step": 1295
},
{
"epoch": 0.16505086202779504,
"grad_norm": 4.268896392501476,
"learning_rate": 1.9855932920111563e-05,
"loss": 0.9921,
"step": 1296
},
{
"epoch": 0.16517821608800165,
"grad_norm": 7.711568727329617,
"learning_rate": 1.9855583875151453e-05,
"loss": 0.9872,
"step": 1297
},
{
"epoch": 0.1653055701482083,
"grad_norm": 5.241036528942064,
"learning_rate": 1.9855234410946002e-05,
"loss": 0.9546,
"step": 1298
},
{
"epoch": 0.16543292420841493,
"grad_norm": 6.764661337078974,
"learning_rate": 1.985488452751008e-05,
"loss": 0.9521,
"step": 1299
},
{
"epoch": 0.16556027826862155,
"grad_norm": 6.087105126077417,
"learning_rate": 1.9854534224858574e-05,
"loss": 1.0039,
"step": 1300
},
{
"epoch": 0.1656876323288282,
"grad_norm": 5.684029327166587,
"learning_rate": 1.9854183503006383e-05,
"loss": 1.082,
"step": 1301
},
{
"epoch": 0.1658149863890348,
"grad_norm": 4.707341632660841,
"learning_rate": 1.9853832361968424e-05,
"loss": 0.9413,
"step": 1302
},
{
"epoch": 0.16594234044924144,
"grad_norm": 5.88976947430451,
"learning_rate": 1.9853480801759637e-05,
"loss": 1.0828,
"step": 1303
},
{
"epoch": 0.16606969450944808,
"grad_norm": 4.565661654367643,
"learning_rate": 1.9853128822394976e-05,
"loss": 1.0482,
"step": 1304
},
{
"epoch": 0.1661970485696547,
"grad_norm": 4.481935482188907,
"learning_rate": 1.9852776423889414e-05,
"loss": 1.1119,
"step": 1305
},
{
"epoch": 0.16632440262986134,
"grad_norm": 11.635677561258317,
"learning_rate": 1.9852423606257943e-05,
"loss": 0.9503,
"step": 1306
},
{
"epoch": 0.16645175669006798,
"grad_norm": 7.3166335722827265,
"learning_rate": 1.9852070369515566e-05,
"loss": 1.0055,
"step": 1307
},
{
"epoch": 0.1665791107502746,
"grad_norm": 5.1385735125753165,
"learning_rate": 1.9851716713677315e-05,
"loss": 0.9379,
"step": 1308
},
{
"epoch": 0.16670646481048124,
"grad_norm": 4.744954692274551,
"learning_rate": 1.9851362638758236e-05,
"loss": 1.0048,
"step": 1309
},
{
"epoch": 0.16683381887068788,
"grad_norm": 4.680360774292839,
"learning_rate": 1.9851008144773386e-05,
"loss": 0.9329,
"step": 1310
},
{
"epoch": 0.1669611729308945,
"grad_norm": 5.54574530474556,
"learning_rate": 1.9850653231737844e-05,
"loss": 1.0084,
"step": 1311
},
{
"epoch": 0.16708852699110113,
"grad_norm": 5.629096341017364,
"learning_rate": 1.985029789966671e-05,
"loss": 0.9647,
"step": 1312
},
{
"epoch": 0.16721588105130777,
"grad_norm": 4.247339201112461,
"learning_rate": 1.98499421485751e-05,
"loss": 1.018,
"step": 1313
},
{
"epoch": 0.1673432351115144,
"grad_norm": 4.8500695135109835,
"learning_rate": 1.984958597847815e-05,
"loss": 1.0345,
"step": 1314
},
{
"epoch": 0.16747058917172103,
"grad_norm": 5.242236255129088,
"learning_rate": 1.9849229389391e-05,
"loss": 1.0124,
"step": 1315
},
{
"epoch": 0.16759794323192767,
"grad_norm": 4.494156115860417,
"learning_rate": 1.984887238132883e-05,
"loss": 0.9282,
"step": 1316
},
{
"epoch": 0.16772529729213428,
"grad_norm": 5.527577705958565,
"learning_rate": 1.9848514954306827e-05,
"loss": 0.9766,
"step": 1317
},
{
"epoch": 0.16785265135234093,
"grad_norm": 7.153181059331445,
"learning_rate": 1.9848157108340186e-05,
"loss": 0.9793,
"step": 1318
},
{
"epoch": 0.16798000541254757,
"grad_norm": 4.274232333675409,
"learning_rate": 1.984779884344414e-05,
"loss": 0.8939,
"step": 1319
},
{
"epoch": 0.16810735947275418,
"grad_norm": 6.091392647998473,
"learning_rate": 1.9847440159633918e-05,
"loss": 1.0784,
"step": 1320
},
{
"epoch": 0.16823471353296082,
"grad_norm": 4.3504516717848,
"learning_rate": 1.9847081056924788e-05,
"loss": 1.0399,
"step": 1321
},
{
"epoch": 0.16836206759316746,
"grad_norm": 5.874738646842187,
"learning_rate": 1.984672153533202e-05,
"loss": 1.0156,
"step": 1322
},
{
"epoch": 0.16848942165337408,
"grad_norm": 6.273024606003035,
"learning_rate": 1.9846361594870914e-05,
"loss": 1.0532,
"step": 1323
},
{
"epoch": 0.16861677571358072,
"grad_norm": 9.43738116258996,
"learning_rate": 1.9846001235556775e-05,
"loss": 0.9758,
"step": 1324
},
{
"epoch": 0.16874412977378736,
"grad_norm": 3.951354463349465,
"learning_rate": 1.984564045740493e-05,
"loss": 1.0017,
"step": 1325
},
{
"epoch": 0.16887148383399397,
"grad_norm": 6.057988641405209,
"learning_rate": 1.984527926043074e-05,
"loss": 1.0013,
"step": 1326
},
{
"epoch": 0.16899883789420062,
"grad_norm": 5.121047938510769,
"learning_rate": 1.9844917644649553e-05,
"loss": 1.0206,
"step": 1327
},
{
"epoch": 0.16912619195440726,
"grad_norm": 6.34730347577412,
"learning_rate": 1.984455561007676e-05,
"loss": 1.0719,
"step": 1328
},
{
"epoch": 0.16925354601461387,
"grad_norm": 6.9857632411586525,
"learning_rate": 1.984419315672776e-05,
"loss": 1.1088,
"step": 1329
},
{
"epoch": 0.1693809000748205,
"grad_norm": 4.054283636482987,
"learning_rate": 1.9843830284617975e-05,
"loss": 0.9645,
"step": 1330
},
{
"epoch": 0.16950825413502715,
"grad_norm": 5.763261181254879,
"learning_rate": 1.9843466993762836e-05,
"loss": 1.0112,
"step": 1331
},
{
"epoch": 0.16963560819523377,
"grad_norm": 5.006710508504241,
"learning_rate": 1.98431032841778e-05,
"loss": 1.0243,
"step": 1332
},
{
"epoch": 0.1697629622554404,
"grad_norm": 5.33453967438594,
"learning_rate": 1.9842739155878337e-05,
"loss": 1.0415,
"step": 1333
},
{
"epoch": 0.16989031631564705,
"grad_norm": 4.239488505334722,
"learning_rate": 1.984237460887994e-05,
"loss": 1.0452,
"step": 1334
},
{
"epoch": 0.17001767037585366,
"grad_norm": 8.172692340070634,
"learning_rate": 1.9842009643198113e-05,
"loss": 0.9793,
"step": 1335
},
{
"epoch": 0.1701450244360603,
"grad_norm": 5.34305572461287,
"learning_rate": 1.984164425884838e-05,
"loss": 1.0048,
"step": 1336
},
{
"epoch": 0.17027237849626695,
"grad_norm": 5.792797865504691,
"learning_rate": 1.984127845584629e-05,
"loss": 0.9795,
"step": 1337
},
{
"epoch": 0.17039973255647356,
"grad_norm": 4.318874126588256,
"learning_rate": 1.9840912234207396e-05,
"loss": 0.9736,
"step": 1338
},
{
"epoch": 0.1705270866166802,
"grad_norm": 4.162413651770721,
"learning_rate": 1.9840545593947286e-05,
"loss": 1.0151,
"step": 1339
},
{
"epoch": 0.17065444067688684,
"grad_norm": 6.5925908086697325,
"learning_rate": 1.9840178535081548e-05,
"loss": 0.904,
"step": 1340
},
{
"epoch": 0.17078179473709346,
"grad_norm": 5.536949712479812,
"learning_rate": 1.98398110576258e-05,
"loss": 1.0522,
"step": 1341
},
{
"epoch": 0.1709091487973001,
"grad_norm": 7.280052774258995,
"learning_rate": 1.9839443161595668e-05,
"loss": 1.0073,
"step": 1342
},
{
"epoch": 0.17103650285750674,
"grad_norm": 5.838656403647462,
"learning_rate": 1.9839074847006815e-05,
"loss": 0.9766,
"step": 1343
},
{
"epoch": 0.17116385691771335,
"grad_norm": 6.120541876944528,
"learning_rate": 1.9838706113874896e-05,
"loss": 1.003,
"step": 1344
},
{
"epoch": 0.17129121097792,
"grad_norm": 6.272822405730764,
"learning_rate": 1.9838336962215606e-05,
"loss": 0.9179,
"step": 1345
},
{
"epoch": 0.1714185650381266,
"grad_norm": 4.571611477394392,
"learning_rate": 1.983796739204464e-05,
"loss": 1.1381,
"step": 1346
},
{
"epoch": 0.17154591909833325,
"grad_norm": 5.164721379109247,
"learning_rate": 1.9837597403377726e-05,
"loss": 1.0764,
"step": 1347
},
{
"epoch": 0.1716732731585399,
"grad_norm": 5.382269800120944,
"learning_rate": 1.98372269962306e-05,
"loss": 1.099,
"step": 1348
},
{
"epoch": 0.1718006272187465,
"grad_norm": 5.6010520628007985,
"learning_rate": 1.9836856170619018e-05,
"loss": 0.8125,
"step": 1349
},
{
"epoch": 0.17192798127895315,
"grad_norm": 6.5416130825043,
"learning_rate": 1.983648492655875e-05,
"loss": 0.9736,
"step": 1350
},
{
"epoch": 0.1720553353391598,
"grad_norm": 5.986232857622864,
"learning_rate": 1.9836113264065598e-05,
"loss": 1.1308,
"step": 1351
},
{
"epoch": 0.1721826893993664,
"grad_norm": 3.721914787944601,
"learning_rate": 1.983574118315537e-05,
"loss": 0.9768,
"step": 1352
},
{
"epoch": 0.17231004345957304,
"grad_norm": 5.081106208746815,
"learning_rate": 1.983536868384389e-05,
"loss": 1.0861,
"step": 1353
},
{
"epoch": 0.17243739751977968,
"grad_norm": 4.766787127771113,
"learning_rate": 1.9834995766147e-05,
"loss": 0.9764,
"step": 1354
},
{
"epoch": 0.1725647515799863,
"grad_norm": 5.844392628957778,
"learning_rate": 1.9834622430080574e-05,
"loss": 0.9675,
"step": 1355
},
{
"epoch": 0.17269210564019294,
"grad_norm": 4.266773111287161,
"learning_rate": 1.9834248675660484e-05,
"loss": 0.9577,
"step": 1356
},
{
"epoch": 0.17281945970039958,
"grad_norm": 5.339045366544977,
"learning_rate": 1.9833874502902636e-05,
"loss": 1.0691,
"step": 1357
},
{
"epoch": 0.1729468137606062,
"grad_norm": 5.321239149867997,
"learning_rate": 1.9833499911822944e-05,
"loss": 1.02,
"step": 1358
},
{
"epoch": 0.17307416782081284,
"grad_norm": 6.036001912839067,
"learning_rate": 1.983312490243734e-05,
"loss": 1.0136,
"step": 1359
},
{
"epoch": 0.17320152188101948,
"grad_norm": 5.842316966880855,
"learning_rate": 1.9832749474761782e-05,
"loss": 0.9487,
"step": 1360
},
{
"epoch": 0.1733288759412261,
"grad_norm": 5.268133726086745,
"learning_rate": 1.9832373628812235e-05,
"loss": 0.897,
"step": 1361
},
{
"epoch": 0.17345623000143273,
"grad_norm": 10.765103043746514,
"learning_rate": 1.9831997364604693e-05,
"loss": 1.057,
"step": 1362
},
{
"epoch": 0.17358358406163937,
"grad_norm": 6.986459988359291,
"learning_rate": 1.983162068215515e-05,
"loss": 0.984,
"step": 1363
},
{
"epoch": 0.173710938121846,
"grad_norm": 6.54607260801127,
"learning_rate": 1.9831243581479643e-05,
"loss": 1.1114,
"step": 1364
},
{
"epoch": 0.17383829218205263,
"grad_norm": 4.736502694914325,
"learning_rate": 1.983086606259421e-05,
"loss": 0.9659,
"step": 1365
},
{
"epoch": 0.17396564624225927,
"grad_norm": 5.3548235459378,
"learning_rate": 1.9830488125514907e-05,
"loss": 0.9164,
"step": 1366
},
{
"epoch": 0.17409300030246588,
"grad_norm": 5.012251530527317,
"learning_rate": 1.983010977025781e-05,
"loss": 1.0724,
"step": 1367
},
{
"epoch": 0.17422035436267252,
"grad_norm": 4.104810144678136,
"learning_rate": 1.982973099683902e-05,
"loss": 1.0126,
"step": 1368
},
{
"epoch": 0.17434770842287917,
"grad_norm": 4.606415189475125,
"learning_rate": 1.9829351805274643e-05,
"loss": 0.9828,
"step": 1369
},
{
"epoch": 0.17447506248308578,
"grad_norm": 4.9605551264425625,
"learning_rate": 1.9828972195580815e-05,
"loss": 1.0372,
"step": 1370
},
{
"epoch": 0.17460241654329242,
"grad_norm": 6.515749118272252,
"learning_rate": 1.9828592167773676e-05,
"loss": 0.9049,
"step": 1371
},
{
"epoch": 0.17472977060349906,
"grad_norm": 5.611944008796001,
"learning_rate": 1.9828211721869404e-05,
"loss": 1.0385,
"step": 1372
},
{
"epoch": 0.17485712466370568,
"grad_norm": 4.296467265811125,
"learning_rate": 1.9827830857884173e-05,
"loss": 1.0467,
"step": 1373
},
{
"epoch": 0.17498447872391232,
"grad_norm": 5.582847839807333,
"learning_rate": 1.9827449575834187e-05,
"loss": 0.9349,
"step": 1374
},
{
"epoch": 0.17511183278411896,
"grad_norm": 5.693181480389708,
"learning_rate": 1.9827067875735667e-05,
"loss": 0.8793,
"step": 1375
},
{
"epoch": 0.17523918684432557,
"grad_norm": 5.261651325431034,
"learning_rate": 1.982668575760485e-05,
"loss": 0.9581,
"step": 1376
},
{
"epoch": 0.17536654090453221,
"grad_norm": 4.977291802306023,
"learning_rate": 1.982630322145799e-05,
"loss": 0.9355,
"step": 1377
},
{
"epoch": 0.17549389496473886,
"grad_norm": 8.424306090442627,
"learning_rate": 1.9825920267311358e-05,
"loss": 1.0181,
"step": 1378
},
{
"epoch": 0.17562124902494547,
"grad_norm": 4.067377536948859,
"learning_rate": 1.9825536895181245e-05,
"loss": 0.9847,
"step": 1379
},
{
"epoch": 0.1757486030851521,
"grad_norm": 5.434342969610268,
"learning_rate": 1.982515310508396e-05,
"loss": 1.0621,
"step": 1380
},
{
"epoch": 0.17587595714535875,
"grad_norm": 3.951322725200578,
"learning_rate": 1.9824768897035833e-05,
"loss": 0.9415,
"step": 1381
},
{
"epoch": 0.17600331120556537,
"grad_norm": 4.20572669754507,
"learning_rate": 1.98243842710532e-05,
"loss": 0.9017,
"step": 1382
},
{
"epoch": 0.176130665265772,
"grad_norm": 5.642519645876438,
"learning_rate": 1.9823999227152426e-05,
"loss": 0.9919,
"step": 1383
},
{
"epoch": 0.17625801932597865,
"grad_norm": 5.8060276543263845,
"learning_rate": 1.9823613765349894e-05,
"loss": 0.9357,
"step": 1384
},
{
"epoch": 0.17638537338618526,
"grad_norm": 6.730220495639194,
"learning_rate": 1.9823227885661994e-05,
"loss": 1.0167,
"step": 1385
},
{
"epoch": 0.1765127274463919,
"grad_norm": 4.98295281189637,
"learning_rate": 1.982284158810515e-05,
"loss": 0.9644,
"step": 1386
},
{
"epoch": 0.17664008150659855,
"grad_norm": 5.611349539312602,
"learning_rate": 1.982245487269579e-05,
"loss": 0.9759,
"step": 1387
},
{
"epoch": 0.17676743556680516,
"grad_norm": 5.830986568146951,
"learning_rate": 1.982206773945036e-05,
"loss": 0.9505,
"step": 1388
},
{
"epoch": 0.1768947896270118,
"grad_norm": 4.7729098313991845,
"learning_rate": 1.9821680188385334e-05,
"loss": 1.0816,
"step": 1389
},
{
"epoch": 0.17702214368721844,
"grad_norm": 4.199991298848176,
"learning_rate": 1.982129221951719e-05,
"loss": 1.1747,
"step": 1390
},
{
"epoch": 0.17714949774742506,
"grad_norm": 5.562743963827968,
"learning_rate": 1.9820903832862445e-05,
"loss": 1.0474,
"step": 1391
},
{
"epoch": 0.1772768518076317,
"grad_norm": 6.610761492520081,
"learning_rate": 1.9820515028437612e-05,
"loss": 1.0511,
"step": 1392
},
{
"epoch": 0.1774042058678383,
"grad_norm": 4.700877720465981,
"learning_rate": 1.9820125806259233e-05,
"loss": 1.034,
"step": 1393
},
{
"epoch": 0.17753155992804495,
"grad_norm": 5.507829641761889,
"learning_rate": 1.981973616634386e-05,
"loss": 0.9607,
"step": 1394
},
{
"epoch": 0.1776589139882516,
"grad_norm": 4.55152367032817,
"learning_rate": 1.9819346108708074e-05,
"loss": 1.0074,
"step": 1395
},
{
"epoch": 0.1777862680484582,
"grad_norm": 5.155387667593372,
"learning_rate": 1.9818955633368464e-05,
"loss": 0.9597,
"step": 1396
},
{
"epoch": 0.17791362210866485,
"grad_norm": 5.73277489064217,
"learning_rate": 1.981856474034164e-05,
"loss": 1.0927,
"step": 1397
},
{
"epoch": 0.1780409761688715,
"grad_norm": 4.071601648882491,
"learning_rate": 1.9818173429644237e-05,
"loss": 0.9982,
"step": 1398
},
{
"epoch": 0.1781683302290781,
"grad_norm": 3.9397300581268513,
"learning_rate": 1.9817781701292892e-05,
"loss": 0.9403,
"step": 1399
},
{
"epoch": 0.17829568428928474,
"grad_norm": 4.89777375854755,
"learning_rate": 1.9817389555304274e-05,
"loss": 0.9848,
"step": 1400
},
{
"epoch": 0.1784230383494914,
"grad_norm": 5.384909433154051,
"learning_rate": 1.9816996991695057e-05,
"loss": 1.0004,
"step": 1401
},
{
"epoch": 0.178550392409698,
"grad_norm": 6.171274817534474,
"learning_rate": 1.9816604010481955e-05,
"loss": 0.9742,
"step": 1402
},
{
"epoch": 0.17867774646990464,
"grad_norm": 5.372669158161206,
"learning_rate": 1.981621061168167e-05,
"loss": 1.0258,
"step": 1403
},
{
"epoch": 0.17880510053011128,
"grad_norm": 5.42572394959394,
"learning_rate": 1.9815816795310945e-05,
"loss": 0.8953,
"step": 1404
},
{
"epoch": 0.1789324545903179,
"grad_norm": 4.888856025546814,
"learning_rate": 1.981542256138653e-05,
"loss": 0.9233,
"step": 1405
},
{
"epoch": 0.17905980865052454,
"grad_norm": 5.561367305342423,
"learning_rate": 1.9815027909925194e-05,
"loss": 0.9601,
"step": 1406
},
{
"epoch": 0.17918716271073118,
"grad_norm": 4.272674659833029,
"learning_rate": 1.9814632840943728e-05,
"loss": 1.0174,
"step": 1407
},
{
"epoch": 0.1793145167709378,
"grad_norm": 4.094915326073295,
"learning_rate": 1.9814237354458937e-05,
"loss": 0.9695,
"step": 1408
},
{
"epoch": 0.17944187083114443,
"grad_norm": 4.905545323784712,
"learning_rate": 1.981384145048764e-05,
"loss": 0.929,
"step": 1409
},
{
"epoch": 0.17956922489135108,
"grad_norm": 4.367058003097394,
"learning_rate": 1.9813445129046685e-05,
"loss": 0.9274,
"step": 1410
},
{
"epoch": 0.1796965789515577,
"grad_norm": 6.3573246680356394,
"learning_rate": 1.9813048390152926e-05,
"loss": 1.0155,
"step": 1411
},
{
"epoch": 0.17982393301176433,
"grad_norm": 6.87006575640397,
"learning_rate": 1.9812651233823245e-05,
"loss": 0.9062,
"step": 1412
},
{
"epoch": 0.17995128707197097,
"grad_norm": 4.697870202127253,
"learning_rate": 1.9812253660074532e-05,
"loss": 1.0905,
"step": 1413
},
{
"epoch": 0.18007864113217759,
"grad_norm": 5.959781639656172,
"learning_rate": 1.98118556689237e-05,
"loss": 1.119,
"step": 1414
},
{
"epoch": 0.18020599519238423,
"grad_norm": 4.111138375223385,
"learning_rate": 1.9811457260387683e-05,
"loss": 0.923,
"step": 1415
},
{
"epoch": 0.18033334925259087,
"grad_norm": 4.006177185634729,
"learning_rate": 1.9811058434483422e-05,
"loss": 1.0849,
"step": 1416
},
{
"epoch": 0.18046070331279748,
"grad_norm": 4.48631303225857,
"learning_rate": 1.981065919122789e-05,
"loss": 1.0728,
"step": 1417
},
{
"epoch": 0.18058805737300412,
"grad_norm": 6.525586168407294,
"learning_rate": 1.9810259530638064e-05,
"loss": 1.0408,
"step": 1418
},
{
"epoch": 0.18071541143321077,
"grad_norm": 6.020659799505925,
"learning_rate": 1.980985945273095e-05,
"loss": 1.0975,
"step": 1419
},
{
"epoch": 0.18084276549341738,
"grad_norm": 6.885051800346678,
"learning_rate": 1.9809458957523563e-05,
"loss": 0.9118,
"step": 1420
},
{
"epoch": 0.18097011955362402,
"grad_norm": 3.859705119358011,
"learning_rate": 1.9809058045032942e-05,
"loss": 1.0455,
"step": 1421
},
{
"epoch": 0.18109747361383066,
"grad_norm": 5.034443854459602,
"learning_rate": 1.980865671527614e-05,
"loss": 0.9053,
"step": 1422
},
{
"epoch": 0.18122482767403728,
"grad_norm": 4.970510494794013,
"learning_rate": 1.9808254968270236e-05,
"loss": 1.0684,
"step": 1423
},
{
"epoch": 0.18135218173424392,
"grad_norm": 6.993515553435381,
"learning_rate": 1.9807852804032306e-05,
"loss": 1.0819,
"step": 1424
},
{
"epoch": 0.18147953579445056,
"grad_norm": 4.3244256139826405,
"learning_rate": 1.980745022257947e-05,
"loss": 1.0832,
"step": 1425
},
{
"epoch": 0.18160688985465717,
"grad_norm": 5.41282386427537,
"learning_rate": 1.9807047223928847e-05,
"loss": 1.0109,
"step": 1426
},
{
"epoch": 0.1817342439148638,
"grad_norm": 5.5849581199995075,
"learning_rate": 1.980664380809758e-05,
"loss": 0.9701,
"step": 1427
},
{
"epoch": 0.18186159797507045,
"grad_norm": 7.5150214763849705,
"learning_rate": 1.980623997510283e-05,
"loss": 1.1214,
"step": 1428
},
{
"epoch": 0.18198895203527707,
"grad_norm": 5.378983019888339,
"learning_rate": 1.9805835724961783e-05,
"loss": 1.0311,
"step": 1429
},
{
"epoch": 0.1821163060954837,
"grad_norm": 5.328501578975879,
"learning_rate": 1.9805431057691627e-05,
"loss": 1.089,
"step": 1430
},
{
"epoch": 0.18224366015569035,
"grad_norm": 4.827672449525467,
"learning_rate": 1.9805025973309577e-05,
"loss": 0.9638,
"step": 1431
},
{
"epoch": 0.18237101421589696,
"grad_norm": 4.960158495677945,
"learning_rate": 1.980462047183287e-05,
"loss": 0.994,
"step": 1432
},
{
"epoch": 0.1824983682761036,
"grad_norm": 4.417898594090962,
"learning_rate": 1.9804214553278747e-05,
"loss": 0.9635,
"step": 1433
},
{
"epoch": 0.18262572233631025,
"grad_norm": 4.3511745752527755,
"learning_rate": 1.9803808217664483e-05,
"loss": 0.9231,
"step": 1434
},
{
"epoch": 0.18275307639651686,
"grad_norm": 7.574369841195433,
"learning_rate": 1.9803401465007363e-05,
"loss": 0.9365,
"step": 1435
},
{
"epoch": 0.1828804304567235,
"grad_norm": 4.817720034473403,
"learning_rate": 1.9802994295324685e-05,
"loss": 0.9657,
"step": 1436
},
{
"epoch": 0.18300778451693012,
"grad_norm": 3.9337316979442223,
"learning_rate": 1.980258670863377e-05,
"loss": 1.0757,
"step": 1437
},
{
"epoch": 0.18313513857713676,
"grad_norm": 4.912953650419165,
"learning_rate": 1.980217870495196e-05,
"loss": 0.8668,
"step": 1438
},
{
"epoch": 0.1832624926373434,
"grad_norm": 6.3601703568201255,
"learning_rate": 1.980177028429661e-05,
"loss": 1.0216,
"step": 1439
},
{
"epoch": 0.18338984669755,
"grad_norm": 4.361051722376627,
"learning_rate": 1.980136144668509e-05,
"loss": 1.0007,
"step": 1440
},
{
"epoch": 0.18351720075775665,
"grad_norm": 4.831064656441857,
"learning_rate": 1.98009521921348e-05,
"loss": 1.0569,
"step": 1441
},
{
"epoch": 0.1836445548179633,
"grad_norm": 4.673866077337553,
"learning_rate": 1.9800542520663136e-05,
"loss": 1.1306,
"step": 1442
},
{
"epoch": 0.1837719088781699,
"grad_norm": 5.025317562564007,
"learning_rate": 1.980013243228754e-05,
"loss": 1.0482,
"step": 1443
},
{
"epoch": 0.18389926293837655,
"grad_norm": 4.917699005025452,
"learning_rate": 1.979972192702544e-05,
"loss": 0.9926,
"step": 1444
},
{
"epoch": 0.1840266169985832,
"grad_norm": 6.695090006422005,
"learning_rate": 1.9799311004894314e-05,
"loss": 1.0492,
"step": 1445
},
{
"epoch": 0.1841539710587898,
"grad_norm": 5.783826429176605,
"learning_rate": 1.9798899665911636e-05,
"loss": 0.9291,
"step": 1446
},
{
"epoch": 0.18428132511899645,
"grad_norm": 8.583389528948576,
"learning_rate": 1.97984879100949e-05,
"loss": 1.0729,
"step": 1447
},
{
"epoch": 0.1844086791792031,
"grad_norm": 4.9845657551297196,
"learning_rate": 1.9798075737461627e-05,
"loss": 1.081,
"step": 1448
},
{
"epoch": 0.1845360332394097,
"grad_norm": 4.630384249468881,
"learning_rate": 1.9797663148029352e-05,
"loss": 0.9889,
"step": 1449
},
{
"epoch": 0.18466338729961634,
"grad_norm": 4.459600725990378,
"learning_rate": 1.9797250141815617e-05,
"loss": 1.0447,
"step": 1450
},
{
"epoch": 0.18479074135982299,
"grad_norm": 7.9994144225384165,
"learning_rate": 1.9796836718838e-05,
"loss": 1.0084,
"step": 1451
},
{
"epoch": 0.1849180954200296,
"grad_norm": 5.037288099049157,
"learning_rate": 1.9796422879114082e-05,
"loss": 0.9369,
"step": 1452
},
{
"epoch": 0.18504544948023624,
"grad_norm": 5.12163116638082,
"learning_rate": 1.9796008622661472e-05,
"loss": 1.0629,
"step": 1453
},
{
"epoch": 0.18517280354044288,
"grad_norm": 4.711496399073382,
"learning_rate": 1.9795593949497786e-05,
"loss": 1.0271,
"step": 1454
},
{
"epoch": 0.1853001576006495,
"grad_norm": 6.076948429563655,
"learning_rate": 1.979517885964067e-05,
"loss": 0.9624,
"step": 1455
},
{
"epoch": 0.18542751166085614,
"grad_norm": 4.85950570375367,
"learning_rate": 1.979476335310778e-05,
"loss": 1.0016,
"step": 1456
},
{
"epoch": 0.18555486572106278,
"grad_norm": 6.515755558306762,
"learning_rate": 1.9794347429916786e-05,
"loss": 0.9468,
"step": 1457
},
{
"epoch": 0.1856822197812694,
"grad_norm": 4.261883670402554,
"learning_rate": 1.9793931090085385e-05,
"loss": 0.9449,
"step": 1458
},
{
"epoch": 0.18580957384147603,
"grad_norm": 5.087873755370306,
"learning_rate": 1.9793514333631287e-05,
"loss": 1.0691,
"step": 1459
},
{
"epoch": 0.18593692790168267,
"grad_norm": 6.254327115367058,
"learning_rate": 1.9793097160572223e-05,
"loss": 1.0433,
"step": 1460
},
{
"epoch": 0.1860642819618893,
"grad_norm": 6.42593203660469,
"learning_rate": 1.9792679570925933e-05,
"loss": 1.0038,
"step": 1461
},
{
"epoch": 0.18619163602209593,
"grad_norm": 4.263729582608785,
"learning_rate": 1.9792261564710188e-05,
"loss": 1.041,
"step": 1462
},
{
"epoch": 0.18631899008230257,
"grad_norm": 4.486672480798122,
"learning_rate": 1.9791843141942763e-05,
"loss": 1.0097,
"step": 1463
},
{
"epoch": 0.18644634414250918,
"grad_norm": 8.109797883813664,
"learning_rate": 1.979142430264146e-05,
"loss": 0.9008,
"step": 1464
},
{
"epoch": 0.18657369820271583,
"grad_norm": 5.566391858534027,
"learning_rate": 1.97910050468241e-05,
"loss": 0.9353,
"step": 1465
},
{
"epoch": 0.18670105226292247,
"grad_norm": 5.093553410580432,
"learning_rate": 1.979058537450851e-05,
"loss": 0.9952,
"step": 1466
},
{
"epoch": 0.18682840632312908,
"grad_norm": 6.860038429797663,
"learning_rate": 1.979016528571255e-05,
"loss": 0.9609,
"step": 1467
},
{
"epoch": 0.18695576038333572,
"grad_norm": 6.188720582735326,
"learning_rate": 1.9789744780454082e-05,
"loss": 1.0192,
"step": 1468
},
{
"epoch": 0.18708311444354236,
"grad_norm": 5.187779614369765,
"learning_rate": 1.9789323858751e-05,
"loss": 1.1023,
"step": 1469
},
{
"epoch": 0.18721046850374898,
"grad_norm": 5.060137161989241,
"learning_rate": 1.978890252062121e-05,
"loss": 1.0415,
"step": 1470
},
{
"epoch": 0.18733782256395562,
"grad_norm": 4.245005646287378,
"learning_rate": 1.9788480766082626e-05,
"loss": 1.0162,
"step": 1471
},
{
"epoch": 0.18746517662416226,
"grad_norm": 5.212867024991071,
"learning_rate": 1.9788058595153202e-05,
"loss": 0.8992,
"step": 1472
},
{
"epoch": 0.18759253068436887,
"grad_norm": 3.5576216105774727,
"learning_rate": 1.978763600785089e-05,
"loss": 0.9837,
"step": 1473
},
{
"epoch": 0.18771988474457552,
"grad_norm": 4.995126447174946,
"learning_rate": 1.9787213004193665e-05,
"loss": 0.967,
"step": 1474
},
{
"epoch": 0.18784723880478216,
"grad_norm": 5.481986109719116,
"learning_rate": 1.9786789584199523e-05,
"loss": 0.9039,
"step": 1475
},
{
"epoch": 0.18797459286498877,
"grad_norm": 5.599221052763061,
"learning_rate": 1.9786365747886475e-05,
"loss": 1.0468,
"step": 1476
},
{
"epoch": 0.1881019469251954,
"grad_norm": 4.9890705819427374,
"learning_rate": 1.9785941495272553e-05,
"loss": 1.0024,
"step": 1477
},
{
"epoch": 0.18822930098540205,
"grad_norm": 6.845984246635087,
"learning_rate": 1.9785516826375805e-05,
"loss": 0.9939,
"step": 1478
},
{
"epoch": 0.18835665504560867,
"grad_norm": 4.7652065406059565,
"learning_rate": 1.978509174121429e-05,
"loss": 0.946,
"step": 1479
},
{
"epoch": 0.1884840091058153,
"grad_norm": 7.010725251690263,
"learning_rate": 1.978466623980609e-05,
"loss": 1.0261,
"step": 1480
},
{
"epoch": 0.18861136316602195,
"grad_norm": 5.342474712763604,
"learning_rate": 1.9784240322169316e-05,
"loss": 1.0171,
"step": 1481
},
{
"epoch": 0.18873871722622856,
"grad_norm": 6.2207481969521,
"learning_rate": 1.9783813988322076e-05,
"loss": 1.0655,
"step": 1482
},
{
"epoch": 0.1888660712864352,
"grad_norm": 6.051037206093783,
"learning_rate": 1.9783387238282513e-05,
"loss": 0.9079,
"step": 1483
},
{
"epoch": 0.18899342534664182,
"grad_norm": 5.941004953369172,
"learning_rate": 1.9782960072068772e-05,
"loss": 1.0933,
"step": 1484
},
{
"epoch": 0.18912077940684846,
"grad_norm": 3.9146807214716772,
"learning_rate": 1.978253248969903e-05,
"loss": 1.0632,
"step": 1485
},
{
"epoch": 0.1892481334670551,
"grad_norm": 5.4431079577859105,
"learning_rate": 1.9782104491191475e-05,
"loss": 0.9911,
"step": 1486
},
{
"epoch": 0.18937548752726172,
"grad_norm": 4.052988037995278,
"learning_rate": 1.9781676076564316e-05,
"loss": 1.0377,
"step": 1487
},
{
"epoch": 0.18950284158746836,
"grad_norm": 5.62963947124253,
"learning_rate": 1.978124724583577e-05,
"loss": 0.9062,
"step": 1488
},
{
"epoch": 0.189630195647675,
"grad_norm": 5.138572399027655,
"learning_rate": 1.978081799902409e-05,
"loss": 0.9558,
"step": 1489
},
{
"epoch": 0.1897575497078816,
"grad_norm": 4.889826407754536,
"learning_rate": 1.9780388336147524e-05,
"loss": 1.0003,
"step": 1490
},
{
"epoch": 0.18988490376808825,
"grad_norm": 7.62897875429576,
"learning_rate": 1.9779958257224355e-05,
"loss": 1.0335,
"step": 1491
},
{
"epoch": 0.1900122578282949,
"grad_norm": 6.335914922220099,
"learning_rate": 1.9779527762272877e-05,
"loss": 1.0124,
"step": 1492
},
{
"epoch": 0.1901396118885015,
"grad_norm": 4.470530335373219,
"learning_rate": 1.9779096851311406e-05,
"loss": 1.093,
"step": 1493
},
{
"epoch": 0.19026696594870815,
"grad_norm": 8.02603348111266,
"learning_rate": 1.977866552435827e-05,
"loss": 1.0535,
"step": 1494
},
{
"epoch": 0.1903943200089148,
"grad_norm": 4.870751901973264,
"learning_rate": 1.9778233781431814e-05,
"loss": 1.0418,
"step": 1495
},
{
"epoch": 0.1905216740691214,
"grad_norm": 4.832118681946968,
"learning_rate": 1.977780162255041e-05,
"loss": 1.0292,
"step": 1496
},
{
"epoch": 0.19064902812932805,
"grad_norm": 5.930854692540798,
"learning_rate": 1.9777369047732438e-05,
"loss": 0.9462,
"step": 1497
},
{
"epoch": 0.1907763821895347,
"grad_norm": 5.495050110245704,
"learning_rate": 1.9776936056996297e-05,
"loss": 0.8484,
"step": 1498
},
{
"epoch": 0.1909037362497413,
"grad_norm": 5.508620355758505,
"learning_rate": 1.977650265036041e-05,
"loss": 1.0204,
"step": 1499
},
{
"epoch": 0.19103109030994794,
"grad_norm": 5.195661444167357,
"learning_rate": 1.9776068827843214e-05,
"loss": 0.9216,
"step": 1500
},
{
"epoch": 0.19115844437015458,
"grad_norm": 7.5291732520655374,
"learning_rate": 1.9775634589463158e-05,
"loss": 0.9515,
"step": 1501
},
{
"epoch": 0.1912857984303612,
"grad_norm": 3.888663536687444,
"learning_rate": 1.977519993523872e-05,
"loss": 0.9452,
"step": 1502
},
{
"epoch": 0.19141315249056784,
"grad_norm": 5.3393483019284895,
"learning_rate": 1.9774764865188388e-05,
"loss": 0.9975,
"step": 1503
},
{
"epoch": 0.19154050655077448,
"grad_norm": 7.013216892528038,
"learning_rate": 1.977432937933067e-05,
"loss": 0.9489,
"step": 1504
},
{
"epoch": 0.1916678606109811,
"grad_norm": 4.091945570836384,
"learning_rate": 1.9773893477684086e-05,
"loss": 1.0263,
"step": 1505
},
{
"epoch": 0.19179521467118774,
"grad_norm": 6.133647497396183,
"learning_rate": 1.977345716026718e-05,
"loss": 0.9503,
"step": 1506
},
{
"epoch": 0.19192256873139438,
"grad_norm": 6.061068965688803,
"learning_rate": 1.9773020427098522e-05,
"loss": 0.996,
"step": 1507
},
{
"epoch": 0.192049922791601,
"grad_norm": 5.329895262413773,
"learning_rate": 1.9772583278196677e-05,
"loss": 0.988,
"step": 1508
},
{
"epoch": 0.19217727685180763,
"grad_norm": 6.2580468828024545,
"learning_rate": 1.977214571358025e-05,
"loss": 0.9768,
"step": 1509
},
{
"epoch": 0.19230463091201427,
"grad_norm": 5.9261243566210045,
"learning_rate": 1.9771707733267852e-05,
"loss": 0.9682,
"step": 1510
},
{
"epoch": 0.1924319849722209,
"grad_norm": 5.140494208715624,
"learning_rate": 1.977126933727811e-05,
"loss": 1.08,
"step": 1511
},
{
"epoch": 0.19255933903242753,
"grad_norm": 6.7300714252643825,
"learning_rate": 1.977083052562968e-05,
"loss": 1.11,
"step": 1512
},
{
"epoch": 0.19268669309263417,
"grad_norm": 5.585036667965532,
"learning_rate": 1.977039129834122e-05,
"loss": 1.1374,
"step": 1513
},
{
"epoch": 0.19281404715284078,
"grad_norm": 4.049696480009359,
"learning_rate": 1.9769951655431426e-05,
"loss": 0.9615,
"step": 1514
},
{
"epoch": 0.19294140121304743,
"grad_norm": 5.976439949723937,
"learning_rate": 1.976951159691899e-05,
"loss": 1.0128,
"step": 1515
},
{
"epoch": 0.19306875527325407,
"grad_norm": 4.616711174007855,
"learning_rate": 1.9769071122822634e-05,
"loss": 1.0134,
"step": 1516
},
{
"epoch": 0.19319610933346068,
"grad_norm": 6.802833628690229,
"learning_rate": 1.97686302331611e-05,
"loss": 0.8562,
"step": 1517
},
{
"epoch": 0.19332346339366732,
"grad_norm": 4.702031512207982,
"learning_rate": 1.9768188927953134e-05,
"loss": 1.0063,
"step": 1518
},
{
"epoch": 0.19345081745387396,
"grad_norm": 4.7894770830185385,
"learning_rate": 1.9767747207217516e-05,
"loss": 0.9469,
"step": 1519
},
{
"epoch": 0.19357817151408058,
"grad_norm": 5.886225263924752,
"learning_rate": 1.9767305070973033e-05,
"loss": 1.0466,
"step": 1520
},
{
"epoch": 0.19370552557428722,
"grad_norm": 4.56563617096565,
"learning_rate": 1.9766862519238493e-05,
"loss": 1.0864,
"step": 1521
},
{
"epoch": 0.19383287963449386,
"grad_norm": 5.274761480208912,
"learning_rate": 1.9766419552032723e-05,
"loss": 1.0344,
"step": 1522
},
{
"epoch": 0.19396023369470047,
"grad_norm": 6.4716598673917485,
"learning_rate": 1.9765976169374566e-05,
"loss": 1.0106,
"step": 1523
},
{
"epoch": 0.19408758775490711,
"grad_norm": 6.49377231098414,
"learning_rate": 1.9765532371282882e-05,
"loss": 1.0763,
"step": 1524
},
{
"epoch": 0.19421494181511376,
"grad_norm": 5.465689014999976,
"learning_rate": 1.976508815777655e-05,
"loss": 0.9861,
"step": 1525
},
{
"epoch": 0.19434229587532037,
"grad_norm": 5.770104464373687,
"learning_rate": 1.976464352887447e-05,
"loss": 1.0394,
"step": 1526
},
{
"epoch": 0.194469649935527,
"grad_norm": 7.618250626931756,
"learning_rate": 1.9764198484595553e-05,
"loss": 1.0921,
"step": 1527
},
{
"epoch": 0.19459700399573365,
"grad_norm": 4.240896460573713,
"learning_rate": 1.9763753024958724e-05,
"loss": 0.9663,
"step": 1528
},
{
"epoch": 0.19472435805594027,
"grad_norm": 4.724200807830366,
"learning_rate": 1.9763307149982945e-05,
"loss": 1.1107,
"step": 1529
},
{
"epoch": 0.1948517121161469,
"grad_norm": 6.047163924379785,
"learning_rate": 1.9762860859687173e-05,
"loss": 0.9791,
"step": 1530
},
{
"epoch": 0.19497906617635352,
"grad_norm": 5.820781427733166,
"learning_rate": 1.9762414154090398e-05,
"loss": 0.9669,
"step": 1531
},
{
"epoch": 0.19510642023656016,
"grad_norm": 6.023807657745205,
"learning_rate": 1.976196703321162e-05,
"loss": 0.953,
"step": 1532
},
{
"epoch": 0.1952337742967668,
"grad_norm": 6.153987255679557,
"learning_rate": 1.9761519497069863e-05,
"loss": 0.9781,
"step": 1533
},
{
"epoch": 0.19536112835697342,
"grad_norm": 6.374491054190222,
"learning_rate": 1.9761071545684158e-05,
"loss": 0.9647,
"step": 1534
},
{
"epoch": 0.19548848241718006,
"grad_norm": 4.914819316472806,
"learning_rate": 1.9760623179073568e-05,
"loss": 1.0278,
"step": 1535
},
{
"epoch": 0.1956158364773867,
"grad_norm": 5.046922051650349,
"learning_rate": 1.9760174397257158e-05,
"loss": 0.9767,
"step": 1536
},
{
"epoch": 0.19574319053759331,
"grad_norm": 6.035067442206626,
"learning_rate": 1.9759725200254027e-05,
"loss": 0.9667,
"step": 1537
},
{
"epoch": 0.19587054459779996,
"grad_norm": 7.227809225227867,
"learning_rate": 1.9759275588083273e-05,
"loss": 0.8844,
"step": 1538
},
{
"epoch": 0.1959978986580066,
"grad_norm": 5.3397633812304175,
"learning_rate": 1.975882556076403e-05,
"loss": 0.88,
"step": 1539
},
{
"epoch": 0.1961252527182132,
"grad_norm": 5.160049056182221,
"learning_rate": 1.975837511831544e-05,
"loss": 0.9729,
"step": 1540
},
{
"epoch": 0.19625260677841985,
"grad_norm": 5.5355201230080056,
"learning_rate": 1.9757924260756665e-05,
"loss": 0.983,
"step": 1541
},
{
"epoch": 0.1963799608386265,
"grad_norm": 5.863808218740776,
"learning_rate": 1.975747298810688e-05,
"loss": 0.9425,
"step": 1542
},
{
"epoch": 0.1965073148988331,
"grad_norm": 4.844791423453396,
"learning_rate": 1.9757021300385288e-05,
"loss": 0.9033,
"step": 1543
},
{
"epoch": 0.19663466895903975,
"grad_norm": 5.3108868113088405,
"learning_rate": 1.9756569197611098e-05,
"loss": 1.0034,
"step": 1544
},
{
"epoch": 0.1967620230192464,
"grad_norm": 5.669919856613489,
"learning_rate": 1.975611667980354e-05,
"loss": 1.0133,
"step": 1545
},
{
"epoch": 0.196889377079453,
"grad_norm": 5.563726225720538,
"learning_rate": 1.9755663746981873e-05,
"loss": 1.0517,
"step": 1546
},
{
"epoch": 0.19701673113965965,
"grad_norm": 4.65256759143498,
"learning_rate": 1.9755210399165358e-05,
"loss": 0.9638,
"step": 1547
},
{
"epoch": 0.1971440851998663,
"grad_norm": 5.692109304765883,
"learning_rate": 1.9754756636373277e-05,
"loss": 1.0341,
"step": 1548
},
{
"epoch": 0.1972714392600729,
"grad_norm": 5.938750125384786,
"learning_rate": 1.975430245862494e-05,
"loss": 1.0953,
"step": 1549
},
{
"epoch": 0.19739879332027954,
"grad_norm": 4.856150646003446,
"learning_rate": 1.9753847865939657e-05,
"loss": 0.9406,
"step": 1550
},
{
"epoch": 0.19752614738048618,
"grad_norm": 4.081784765659373,
"learning_rate": 1.9753392858336776e-05,
"loss": 1.0257,
"step": 1551
},
{
"epoch": 0.1976535014406928,
"grad_norm": 5.313787865208813,
"learning_rate": 1.975293743583565e-05,
"loss": 1.0489,
"step": 1552
},
{
"epoch": 0.19778085550089944,
"grad_norm": 5.256694158959123,
"learning_rate": 1.9752481598455648e-05,
"loss": 1.0289,
"step": 1553
},
{
"epoch": 0.19790820956110608,
"grad_norm": 3.9844003713959366,
"learning_rate": 1.9752025346216164e-05,
"loss": 1.0425,
"step": 1554
},
{
"epoch": 0.1980355636213127,
"grad_norm": 6.3238206329291895,
"learning_rate": 1.975156867913661e-05,
"loss": 0.9651,
"step": 1555
},
{
"epoch": 0.19816291768151933,
"grad_norm": 4.594948125801704,
"learning_rate": 1.97511115972364e-05,
"loss": 1.0036,
"step": 1556
},
{
"epoch": 0.19829027174172598,
"grad_norm": 5.107193694114138,
"learning_rate": 1.9750654100534992e-05,
"loss": 1.0419,
"step": 1557
},
{
"epoch": 0.1984176258019326,
"grad_norm": 5.134129835650002,
"learning_rate": 1.9750196189051837e-05,
"loss": 0.9756,
"step": 1558
},
{
"epoch": 0.19854497986213923,
"grad_norm": 4.923212553870282,
"learning_rate": 1.974973786280642e-05,
"loss": 0.9265,
"step": 1559
},
{
"epoch": 0.19867233392234587,
"grad_norm": 3.993883702005666,
"learning_rate": 1.9749279121818235e-05,
"loss": 0.9821,
"step": 1560
},
{
"epoch": 0.1987996879825525,
"grad_norm": 5.936873553507101,
"learning_rate": 1.97488199661068e-05,
"loss": 1.1268,
"step": 1561
},
{
"epoch": 0.19892704204275913,
"grad_norm": 5.895892714705748,
"learning_rate": 1.9748360395691644e-05,
"loss": 1.0248,
"step": 1562
},
{
"epoch": 0.19905439610296577,
"grad_norm": 6.590647469234013,
"learning_rate": 1.9747900410592314e-05,
"loss": 0.8957,
"step": 1563
},
{
"epoch": 0.19918175016317238,
"grad_norm": 4.850838673233757,
"learning_rate": 1.9747440010828384e-05,
"loss": 0.9782,
"step": 1564
},
{
"epoch": 0.19930910422337902,
"grad_norm": 4.491532730813909,
"learning_rate": 1.974697919641943e-05,
"loss": 1.0488,
"step": 1565
},
{
"epoch": 0.19943645828358567,
"grad_norm": 6.2761674501680424,
"learning_rate": 1.974651796738506e-05,
"loss": 0.9976,
"step": 1566
},
{
"epoch": 0.19956381234379228,
"grad_norm": 3.8229999723721324,
"learning_rate": 1.9746056323744897e-05,
"loss": 1.0592,
"step": 1567
},
{
"epoch": 0.19969116640399892,
"grad_norm": 6.720524043585452,
"learning_rate": 1.9745594265518574e-05,
"loss": 0.9593,
"step": 1568
},
{
"epoch": 0.19981852046420556,
"grad_norm": 4.296872336646902,
"learning_rate": 1.9745131792725748e-05,
"loss": 0.9543,
"step": 1569
},
{
"epoch": 0.19994587452441218,
"grad_norm": 3.8372903524284334,
"learning_rate": 1.974466890538609e-05,
"loss": 0.9131,
"step": 1570
},
{
"epoch": 0.20007322858461882,
"grad_norm": 5.553138656470616,
"learning_rate": 1.9744205603519293e-05,
"loss": 0.9076,
"step": 1571
},
{
"epoch": 0.20020058264482546,
"grad_norm": 6.149281231816345,
"learning_rate": 1.9743741887145067e-05,
"loss": 1.1001,
"step": 1572
},
{
"epoch": 0.20032793670503207,
"grad_norm": 4.25695904218789,
"learning_rate": 1.9743277756283133e-05,
"loss": 0.9958,
"step": 1573
},
{
"epoch": 0.20045529076523871,
"grad_norm": 5.06832475301277,
"learning_rate": 1.974281321095324e-05,
"loss": 0.9478,
"step": 1574
},
{
"epoch": 0.20058264482544533,
"grad_norm": 5.257481285421065,
"learning_rate": 1.9742348251175146e-05,
"loss": 0.9822,
"step": 1575
},
{
"epoch": 0.20070999888565197,
"grad_norm": 5.073481575950944,
"learning_rate": 1.974188287696863e-05,
"loss": 0.9306,
"step": 1576
},
{
"epoch": 0.2008373529458586,
"grad_norm": 5.268382632298335,
"learning_rate": 1.974141708835349e-05,
"loss": 1.0149,
"step": 1577
},
{
"epoch": 0.20096470700606522,
"grad_norm": 9.723198856770377,
"learning_rate": 1.9740950885349536e-05,
"loss": 1.0157,
"step": 1578
},
{
"epoch": 0.20109206106627187,
"grad_norm": 5.560161827806057,
"learning_rate": 1.9740484267976608e-05,
"loss": 0.9826,
"step": 1579
},
{
"epoch": 0.2012194151264785,
"grad_norm": 5.714445234524424,
"learning_rate": 1.974001723625455e-05,
"loss": 0.9877,
"step": 1580
},
{
"epoch": 0.20134676918668512,
"grad_norm": 4.86436044752818,
"learning_rate": 1.9739549790203224e-05,
"loss": 1.0302,
"step": 1581
},
{
"epoch": 0.20147412324689176,
"grad_norm": 5.9878916952751995,
"learning_rate": 1.973908192984252e-05,
"loss": 0.9946,
"step": 1582
},
{
"epoch": 0.2016014773070984,
"grad_norm": 6.879338230864136,
"learning_rate": 1.9738613655192345e-05,
"loss": 0.9916,
"step": 1583
},
{
"epoch": 0.20172883136730502,
"grad_norm": 4.571181099017479,
"learning_rate": 1.973814496627261e-05,
"loss": 1.1194,
"step": 1584
},
{
"epoch": 0.20185618542751166,
"grad_norm": 5.342571284702611,
"learning_rate": 1.9737675863103257e-05,
"loss": 0.9541,
"step": 1585
},
{
"epoch": 0.2019835394877183,
"grad_norm": 5.99493798663698,
"learning_rate": 1.9737206345704244e-05,
"loss": 1.0448,
"step": 1586
},
{
"epoch": 0.2021108935479249,
"grad_norm": 5.206679954039697,
"learning_rate": 1.9736736414095538e-05,
"loss": 0.9584,
"step": 1587
},
{
"epoch": 0.20223824760813155,
"grad_norm": 6.131071545787127,
"learning_rate": 1.973626606829713e-05,
"loss": 0.9996,
"step": 1588
},
{
"epoch": 0.2023656016683382,
"grad_norm": 5.296658401786897,
"learning_rate": 1.9735795308329037e-05,
"loss": 0.9626,
"step": 1589
},
{
"epoch": 0.2024929557285448,
"grad_norm": 4.480860805236891,
"learning_rate": 1.973532413421127e-05,
"loss": 0.8541,
"step": 1590
},
{
"epoch": 0.20262030978875145,
"grad_norm": 5.458814740488239,
"learning_rate": 1.973485254596388e-05,
"loss": 0.942,
"step": 1591
},
{
"epoch": 0.2027476638489581,
"grad_norm": 5.60844851522458,
"learning_rate": 1.9734380543606932e-05,
"loss": 1.0469,
"step": 1592
},
{
"epoch": 0.2028750179091647,
"grad_norm": 3.9807266586051577,
"learning_rate": 1.9733908127160495e-05,
"loss": 0.9929,
"step": 1593
},
{
"epoch": 0.20300237196937135,
"grad_norm": 5.214944879929989,
"learning_rate": 1.973343529664467e-05,
"loss": 0.8852,
"step": 1594
},
{
"epoch": 0.203129726029578,
"grad_norm": 5.241255425764933,
"learning_rate": 1.9732962052079575e-05,
"loss": 0.9598,
"step": 1595
},
{
"epoch": 0.2032570800897846,
"grad_norm": 5.1585897050212255,
"learning_rate": 1.9732488393485332e-05,
"loss": 1.0306,
"step": 1596
},
{
"epoch": 0.20338443414999124,
"grad_norm": 6.224674029413985,
"learning_rate": 1.9732014320882096e-05,
"loss": 0.9012,
"step": 1597
},
{
"epoch": 0.20351178821019789,
"grad_norm": 4.540211622372655,
"learning_rate": 1.9731539834290034e-05,
"loss": 1.0735,
"step": 1598
},
{
"epoch": 0.2036391422704045,
"grad_norm": 5.2255767211946855,
"learning_rate": 1.9731064933729324e-05,
"loss": 0.8904,
"step": 1599
},
{
"epoch": 0.20376649633061114,
"grad_norm": 5.265525749018251,
"learning_rate": 1.9730589619220177e-05,
"loss": 0.9952,
"step": 1600
},
{
"epoch": 0.20389385039081778,
"grad_norm": 4.648503766268739,
"learning_rate": 1.9730113890782804e-05,
"loss": 1.0357,
"step": 1601
},
{
"epoch": 0.2040212044510244,
"grad_norm": 5.008336936877452,
"learning_rate": 1.9729637748437448e-05,
"loss": 0.9865,
"step": 1602
},
{
"epoch": 0.20414855851123104,
"grad_norm": 4.832030263174414,
"learning_rate": 1.972916119220436e-05,
"loss": 1.0622,
"step": 1603
},
{
"epoch": 0.20427591257143768,
"grad_norm": 8.765669824278218,
"learning_rate": 1.9728684222103812e-05,
"loss": 0.9047,
"step": 1604
},
{
"epoch": 0.2044032666316443,
"grad_norm": 5.656930787651489,
"learning_rate": 1.9728206838156098e-05,
"loss": 0.9878,
"step": 1605
},
{
"epoch": 0.20453062069185093,
"grad_norm": 3.652824427343459,
"learning_rate": 1.9727729040381517e-05,
"loss": 0.906,
"step": 1606
},
{
"epoch": 0.20465797475205758,
"grad_norm": 4.8632365397041175,
"learning_rate": 1.97272508288004e-05,
"loss": 1.0788,
"step": 1607
},
{
"epoch": 0.2047853288122642,
"grad_norm": 6.17770195662139,
"learning_rate": 1.972677220343309e-05,
"loss": 1.0285,
"step": 1608
},
{
"epoch": 0.20491268287247083,
"grad_norm": 5.537076995055236,
"learning_rate": 1.972629316429995e-05,
"loss": 0.9512,
"step": 1609
},
{
"epoch": 0.20504003693267747,
"grad_norm": 4.737506296762595,
"learning_rate": 1.972581371142135e-05,
"loss": 0.9401,
"step": 1610
},
{
"epoch": 0.20516739099288409,
"grad_norm": 5.31851788920953,
"learning_rate": 1.9725333844817688e-05,
"loss": 1.0683,
"step": 1611
},
{
"epoch": 0.20529474505309073,
"grad_norm": 5.594261103461401,
"learning_rate": 1.972485356450938e-05,
"loss": 0.8846,
"step": 1612
},
{
"epoch": 0.20542209911329737,
"grad_norm": 5.099392439612177,
"learning_rate": 1.972437287051685e-05,
"loss": 1.0609,
"step": 1613
},
{
"epoch": 0.20554945317350398,
"grad_norm": 6.0986471208048725,
"learning_rate": 1.9723891762860557e-05,
"loss": 1.0413,
"step": 1614
},
{
"epoch": 0.20567680723371062,
"grad_norm": 5.832802375970654,
"learning_rate": 1.9723410241560958e-05,
"loss": 1.043,
"step": 1615
},
{
"epoch": 0.20580416129391726,
"grad_norm": 5.085532270422708,
"learning_rate": 1.972292830663854e-05,
"loss": 0.9361,
"step": 1616
},
{
"epoch": 0.20593151535412388,
"grad_norm": 7.483943281683596,
"learning_rate": 1.9722445958113803e-05,
"loss": 1.0771,
"step": 1617
},
{
"epoch": 0.20605886941433052,
"grad_norm": 7.7429546203268815,
"learning_rate": 1.9721963196007263e-05,
"loss": 0.9862,
"step": 1618
},
{
"epoch": 0.20618622347453716,
"grad_norm": 4.275373003118664,
"learning_rate": 1.972148002033946e-05,
"loss": 0.9736,
"step": 1619
},
{
"epoch": 0.20631357753474378,
"grad_norm": 4.442061108412303,
"learning_rate": 1.9720996431130946e-05,
"loss": 0.962,
"step": 1620
},
{
"epoch": 0.20644093159495042,
"grad_norm": 4.36612542464768,
"learning_rate": 1.9720512428402293e-05,
"loss": 0.9597,
"step": 1621
},
{
"epoch": 0.20656828565515703,
"grad_norm": 5.255502269780313,
"learning_rate": 1.9720028012174094e-05,
"loss": 0.9574,
"step": 1622
},
{
"epoch": 0.20669563971536367,
"grad_norm": 5.427818009431217,
"learning_rate": 1.9719543182466944e-05,
"loss": 1.0017,
"step": 1623
},
{
"epoch": 0.2068229937755703,
"grad_norm": 5.819726533003532,
"learning_rate": 1.9719057939301477e-05,
"loss": 0.9151,
"step": 1624
},
{
"epoch": 0.20695034783577693,
"grad_norm": 5.716644399764548,
"learning_rate": 1.9718572282698335e-05,
"loss": 1.0101,
"step": 1625
},
{
"epoch": 0.20707770189598357,
"grad_norm": 4.933284650167188,
"learning_rate": 1.971808621267817e-05,
"loss": 0.9385,
"step": 1626
},
{
"epoch": 0.2072050559561902,
"grad_norm": 4.628939626066382,
"learning_rate": 1.9717599729261666e-05,
"loss": 1.0506,
"step": 1627
},
{
"epoch": 0.20733241001639682,
"grad_norm": 5.3615858672677215,
"learning_rate": 1.971711283246951e-05,
"loss": 1.0117,
"step": 1628
},
{
"epoch": 0.20745976407660346,
"grad_norm": 6.214288232361231,
"learning_rate": 1.971662552232242e-05,
"loss": 0.9271,
"step": 1629
},
{
"epoch": 0.2075871181368101,
"grad_norm": 6.244529771655492,
"learning_rate": 1.971613779884113e-05,
"loss": 1.1197,
"step": 1630
},
{
"epoch": 0.20771447219701672,
"grad_norm": 6.029004879854435,
"learning_rate": 1.9715649662046378e-05,
"loss": 1.0605,
"step": 1631
},
{
"epoch": 0.20784182625722336,
"grad_norm": 5.489495215773398,
"learning_rate": 1.971516111195893e-05,
"loss": 0.9659,
"step": 1632
},
{
"epoch": 0.20796918031743,
"grad_norm": 5.247503867981655,
"learning_rate": 1.971467214859957e-05,
"loss": 0.8921,
"step": 1633
},
{
"epoch": 0.20809653437763662,
"grad_norm": 7.338572393551121,
"learning_rate": 1.97141827719891e-05,
"loss": 1.0235,
"step": 1634
},
{
"epoch": 0.20822388843784326,
"grad_norm": 6.658093694099257,
"learning_rate": 1.971369298214834e-05,
"loss": 1.0838,
"step": 1635
},
{
"epoch": 0.2083512424980499,
"grad_norm": 4.663091357163081,
"learning_rate": 1.9713202779098118e-05,
"loss": 1.0795,
"step": 1636
},
{
"epoch": 0.2084785965582565,
"grad_norm": 5.371071910466963,
"learning_rate": 1.9712712162859287e-05,
"loss": 1.1829,
"step": 1637
},
{
"epoch": 0.20860595061846315,
"grad_norm": 5.334058751998837,
"learning_rate": 1.9712221133452722e-05,
"loss": 1.0557,
"step": 1638
},
{
"epoch": 0.2087333046786698,
"grad_norm": 3.8505080445218827,
"learning_rate": 1.971172969089931e-05,
"loss": 0.9079,
"step": 1639
},
{
"epoch": 0.2088606587388764,
"grad_norm": 5.877852700967215,
"learning_rate": 1.9711237835219955e-05,
"loss": 0.9663,
"step": 1640
},
{
"epoch": 0.20898801279908305,
"grad_norm": 5.391358610350945,
"learning_rate": 1.9710745566435578e-05,
"loss": 0.993,
"step": 1641
},
{
"epoch": 0.2091153668592897,
"grad_norm": 5.2566674900155075,
"learning_rate": 1.9710252884567124e-05,
"loss": 1.0123,
"step": 1642
},
{
"epoch": 0.2092427209194963,
"grad_norm": 4.4334306997826625,
"learning_rate": 1.970975978963555e-05,
"loss": 0.928,
"step": 1643
},
{
"epoch": 0.20937007497970295,
"grad_norm": 4.981977214882108,
"learning_rate": 1.970926628166183e-05,
"loss": 1.0223,
"step": 1644
},
{
"epoch": 0.2094974290399096,
"grad_norm": 4.961837278454065,
"learning_rate": 1.9708772360666958e-05,
"loss": 0.9146,
"step": 1645
},
{
"epoch": 0.2096247831001162,
"grad_norm": 5.469682363322358,
"learning_rate": 1.970827802667194e-05,
"loss": 0.9955,
"step": 1646
},
{
"epoch": 0.20975213716032284,
"grad_norm": 7.40464849097527,
"learning_rate": 1.9707783279697816e-05,
"loss": 0.9915,
"step": 1647
},
{
"epoch": 0.20987949122052948,
"grad_norm": 4.157092245708658,
"learning_rate": 1.9707288119765625e-05,
"loss": 1.0057,
"step": 1648
},
{
"epoch": 0.2100068452807361,
"grad_norm": 5.203421186201981,
"learning_rate": 1.9706792546896425e-05,
"loss": 1.0078,
"step": 1649
},
{
"epoch": 0.21013419934094274,
"grad_norm": 4.969316942031503,
"learning_rate": 1.9706296561111308e-05,
"loss": 1.0448,
"step": 1650
},
{
"epoch": 0.21026155340114938,
"grad_norm": 4.850189456975334,
"learning_rate": 1.9705800162431365e-05,
"loss": 1.0313,
"step": 1651
},
{
"epoch": 0.210388907461356,
"grad_norm": 6.328925489493655,
"learning_rate": 1.9705303350877714e-05,
"loss": 1.047,
"step": 1652
},
{
"epoch": 0.21051626152156264,
"grad_norm": 5.55827205153297,
"learning_rate": 1.970480612647149e-05,
"loss": 1.0637,
"step": 1653
},
{
"epoch": 0.21064361558176928,
"grad_norm": 6.3722659961571955,
"learning_rate": 1.9704308489233846e-05,
"loss": 1.047,
"step": 1654
},
{
"epoch": 0.2107709696419759,
"grad_norm": 4.8103105593092375,
"learning_rate": 1.9703810439185946e-05,
"loss": 1.0598,
"step": 1655
},
{
"epoch": 0.21089832370218253,
"grad_norm": 8.339356140604846,
"learning_rate": 1.970331197634898e-05,
"loss": 0.9091,
"step": 1656
},
{
"epoch": 0.21102567776238917,
"grad_norm": 4.713283176254165,
"learning_rate": 1.970281310074415e-05,
"loss": 1.0329,
"step": 1657
},
{
"epoch": 0.2111530318225958,
"grad_norm": 5.932060239417839,
"learning_rate": 1.9702313812392683e-05,
"loss": 1.0818,
"step": 1658
},
{
"epoch": 0.21128038588280243,
"grad_norm": 5.183738355768398,
"learning_rate": 1.9701814111315813e-05,
"loss": 1.0107,
"step": 1659
},
{
"epoch": 0.21140773994300907,
"grad_norm": 5.967140180385663,
"learning_rate": 1.9701313997534798e-05,
"loss": 0.9086,
"step": 1660
},
{
"epoch": 0.21153509400321568,
"grad_norm": 4.887698336818495,
"learning_rate": 1.9700813471070907e-05,
"loss": 1.0905,
"step": 1661
},
{
"epoch": 0.21166244806342233,
"grad_norm": 6.2653132715471616,
"learning_rate": 1.9700312531945444e-05,
"loss": 0.963,
"step": 1662
},
{
"epoch": 0.21178980212362897,
"grad_norm": 6.325315553536174,
"learning_rate": 1.9699811180179704e-05,
"loss": 0.9581,
"step": 1663
},
{
"epoch": 0.21191715618383558,
"grad_norm": 4.819017233161328,
"learning_rate": 1.9699309415795027e-05,
"loss": 1.023,
"step": 1664
},
{
"epoch": 0.21204451024404222,
"grad_norm": 5.740437392739419,
"learning_rate": 1.9698807238812748e-05,
"loss": 0.9332,
"step": 1665
},
{
"epoch": 0.21217186430424884,
"grad_norm": 5.980074858946459,
"learning_rate": 1.9698304649254235e-05,
"loss": 1.0194,
"step": 1666
},
{
"epoch": 0.21229921836445548,
"grad_norm": 4.768733977244646,
"learning_rate": 1.9697801647140865e-05,
"loss": 1.0535,
"step": 1667
},
{
"epoch": 0.21242657242466212,
"grad_norm": 5.7997365299760535,
"learning_rate": 1.9697298232494037e-05,
"loss": 0.9816,
"step": 1668
},
{
"epoch": 0.21255392648486873,
"grad_norm": 5.853278729669982,
"learning_rate": 1.969679440533516e-05,
"loss": 0.9141,
"step": 1669
},
{
"epoch": 0.21268128054507537,
"grad_norm": 5.938993888444609,
"learning_rate": 1.9696290165685674e-05,
"loss": 0.9569,
"step": 1670
},
{
"epoch": 0.21280863460528202,
"grad_norm": 5.824831771058527,
"learning_rate": 1.9695785513567024e-05,
"loss": 1.076,
"step": 1671
},
{
"epoch": 0.21293598866548863,
"grad_norm": 9.641584104438895,
"learning_rate": 1.969528044900068e-05,
"loss": 1.0234,
"step": 1672
},
{
"epoch": 0.21306334272569527,
"grad_norm": 3.97168845182061,
"learning_rate": 1.969477497200812e-05,
"loss": 0.9538,
"step": 1673
},
{
"epoch": 0.2131906967859019,
"grad_norm": 8.086281884272529,
"learning_rate": 1.9694269082610856e-05,
"loss": 0.9032,
"step": 1674
},
{
"epoch": 0.21331805084610853,
"grad_norm": 6.182984833608822,
"learning_rate": 1.9693762780830404e-05,
"loss": 0.9367,
"step": 1675
},
{
"epoch": 0.21344540490631517,
"grad_norm": 6.781903072851365,
"learning_rate": 1.96932560666883e-05,
"loss": 1.0375,
"step": 1676
},
{
"epoch": 0.2135727589665218,
"grad_norm": 4.736966773952431,
"learning_rate": 1.96927489402061e-05,
"loss": 0.9725,
"step": 1677
},
{
"epoch": 0.21370011302672842,
"grad_norm": 5.598186757942846,
"learning_rate": 1.9692241401405376e-05,
"loss": 0.9808,
"step": 1678
},
{
"epoch": 0.21382746708693506,
"grad_norm": 5.052495985449508,
"learning_rate": 1.9691733450307723e-05,
"loss": 0.9624,
"step": 1679
},
{
"epoch": 0.2139548211471417,
"grad_norm": 6.730913657850143,
"learning_rate": 1.9691225086934743e-05,
"loss": 0.9979,
"step": 1680
},
{
"epoch": 0.21408217520734832,
"grad_norm": 6.456859330175002,
"learning_rate": 1.969071631130806e-05,
"loss": 1.0151,
"step": 1681
},
{
"epoch": 0.21420952926755496,
"grad_norm": 5.2489839206361335,
"learning_rate": 1.9690207123449322e-05,
"loss": 1.0065,
"step": 1682
},
{
"epoch": 0.2143368833277616,
"grad_norm": 5.993312287482855,
"learning_rate": 1.968969752338019e-05,
"loss": 1.0205,
"step": 1683
},
{
"epoch": 0.21446423738796822,
"grad_norm": 6.105879854601475,
"learning_rate": 1.968918751112233e-05,
"loss": 0.9294,
"step": 1684
},
{
"epoch": 0.21459159144817486,
"grad_norm": 4.22773700972384,
"learning_rate": 1.968867708669745e-05,
"loss": 0.9431,
"step": 1685
},
{
"epoch": 0.2147189455083815,
"grad_norm": 4.66537789305537,
"learning_rate": 1.968816625012726e-05,
"loss": 0.9277,
"step": 1686
},
{
"epoch": 0.2148462995685881,
"grad_norm": 5.7675083464920505,
"learning_rate": 1.968765500143349e-05,
"loss": 1.0131,
"step": 1687
},
{
"epoch": 0.21497365362879475,
"grad_norm": 4.252982047072925,
"learning_rate": 1.9687143340637885e-05,
"loss": 0.9675,
"step": 1688
},
{
"epoch": 0.2151010076890014,
"grad_norm": 4.236290755223603,
"learning_rate": 1.9686631267762216e-05,
"loss": 0.9622,
"step": 1689
},
{
"epoch": 0.215228361749208,
"grad_norm": 6.703800763121683,
"learning_rate": 1.968611878282826e-05,
"loss": 1.0032,
"step": 1690
},
{
"epoch": 0.21535571580941465,
"grad_norm": 3.940146147442758,
"learning_rate": 1.968560588585782e-05,
"loss": 0.9759,
"step": 1691
},
{
"epoch": 0.2154830698696213,
"grad_norm": 4.799271273927135,
"learning_rate": 1.9685092576872717e-05,
"loss": 1.0655,
"step": 1692
},
{
"epoch": 0.2156104239298279,
"grad_norm": 6.680157489861707,
"learning_rate": 1.9684578855894783e-05,
"loss": 0.9126,
"step": 1693
},
{
"epoch": 0.21573777799003455,
"grad_norm": 4.623934442616402,
"learning_rate": 1.968406472294587e-05,
"loss": 0.8562,
"step": 1694
},
{
"epoch": 0.2158651320502412,
"grad_norm": 6.912497891913811,
"learning_rate": 1.9683550178047852e-05,
"loss": 1.0111,
"step": 1695
},
{
"epoch": 0.2159924861104478,
"grad_norm": 4.770563687140798,
"learning_rate": 1.9683035221222617e-05,
"loss": 0.8775,
"step": 1696
},
{
"epoch": 0.21611984017065444,
"grad_norm": 6.298702154748113,
"learning_rate": 1.9682519852492066e-05,
"loss": 0.9678,
"step": 1697
},
{
"epoch": 0.21624719423086108,
"grad_norm": 4.4266732538319795,
"learning_rate": 1.9682004071878128e-05,
"loss": 0.9295,
"step": 1698
},
{
"epoch": 0.2163745482910677,
"grad_norm": 4.86071601836413,
"learning_rate": 1.968148787940274e-05,
"loss": 0.9146,
"step": 1699
},
{
"epoch": 0.21650190235127434,
"grad_norm": 4.516499048667697,
"learning_rate": 1.9680971275087862e-05,
"loss": 1.0534,
"step": 1700
},
{
"epoch": 0.21662925641148098,
"grad_norm": 3.7220862033214726,
"learning_rate": 1.968045425895547e-05,
"loss": 0.9642,
"step": 1701
},
{
"epoch": 0.2167566104716876,
"grad_norm": 6.026022429395702,
"learning_rate": 1.9679936831027558e-05,
"loss": 1.0535,
"step": 1702
},
{
"epoch": 0.21688396453189424,
"grad_norm": 6.490947875880115,
"learning_rate": 1.9679418991326134e-05,
"loss": 1.0937,
"step": 1703
},
{
"epoch": 0.21701131859210088,
"grad_norm": 5.384847093145054,
"learning_rate": 1.967890073987323e-05,
"loss": 0.9663,
"step": 1704
},
{
"epoch": 0.2171386726523075,
"grad_norm": 6.540822875724549,
"learning_rate": 1.9678382076690883e-05,
"loss": 0.9928,
"step": 1705
},
{
"epoch": 0.21726602671251413,
"grad_norm": 5.064704509223324,
"learning_rate": 1.9677863001801167e-05,
"loss": 1.0063,
"step": 1706
},
{
"epoch": 0.21739338077272077,
"grad_norm": 5.422335646351518,
"learning_rate": 1.967734351522616e-05,
"loss": 0.9905,
"step": 1707
},
{
"epoch": 0.2175207348329274,
"grad_norm": 4.819619597306839,
"learning_rate": 1.9676823616987958e-05,
"loss": 1.0503,
"step": 1708
},
{
"epoch": 0.21764808889313403,
"grad_norm": 4.844071309907478,
"learning_rate": 1.967630330710868e-05,
"loss": 1.0092,
"step": 1709
},
{
"epoch": 0.21777544295334067,
"grad_norm": 6.182490314826213,
"learning_rate": 1.967578258561045e-05,
"loss": 0.9271,
"step": 1710
},
{
"epoch": 0.21790279701354728,
"grad_norm": 4.720366952500596,
"learning_rate": 1.9675261452515434e-05,
"loss": 0.9913,
"step": 1711
},
{
"epoch": 0.21803015107375393,
"grad_norm": 5.772697432269532,
"learning_rate": 1.967473990784579e-05,
"loss": 0.8759,
"step": 1712
},
{
"epoch": 0.21815750513396054,
"grad_norm": 5.444777255796088,
"learning_rate": 1.967421795162371e-05,
"loss": 0.9836,
"step": 1713
},
{
"epoch": 0.21828485919416718,
"grad_norm": 5.684695044720064,
"learning_rate": 1.9673695583871392e-05,
"loss": 0.9398,
"step": 1714
},
{
"epoch": 0.21841221325437382,
"grad_norm": 5.116815311786866,
"learning_rate": 1.9673172804611057e-05,
"loss": 0.8838,
"step": 1715
},
{
"epoch": 0.21853956731458044,
"grad_norm": 5.017741199544419,
"learning_rate": 1.9672649613864946e-05,
"loss": 0.988,
"step": 1716
},
{
"epoch": 0.21866692137478708,
"grad_norm": 6.147091011630903,
"learning_rate": 1.967212601165531e-05,
"loss": 1.1042,
"step": 1717
},
{
"epoch": 0.21879427543499372,
"grad_norm": 4.417358464217229,
"learning_rate": 1.9671601998004436e-05,
"loss": 0.9115,
"step": 1718
},
{
"epoch": 0.21892162949520033,
"grad_norm": 6.183131669998234,
"learning_rate": 1.96710775729346e-05,
"loss": 1.053,
"step": 1719
},
{
"epoch": 0.21904898355540697,
"grad_norm": 5.625086126727791,
"learning_rate": 1.9670552736468117e-05,
"loss": 1.0128,
"step": 1720
},
{
"epoch": 0.21917633761561361,
"grad_norm": 4.117668804143339,
"learning_rate": 1.9670027488627314e-05,
"loss": 1.0047,
"step": 1721
},
{
"epoch": 0.21930369167582023,
"grad_norm": 5.02721456437311,
"learning_rate": 1.966950182943453e-05,
"loss": 0.9749,
"step": 1722
},
{
"epoch": 0.21943104573602687,
"grad_norm": 5.166146877763162,
"learning_rate": 1.966897575891213e-05,
"loss": 0.8736,
"step": 1723
},
{
"epoch": 0.2195583997962335,
"grad_norm": 5.609421073703104,
"learning_rate": 1.9668449277082492e-05,
"loss": 0.8433,
"step": 1724
},
{
"epoch": 0.21968575385644012,
"grad_norm": 5.286167721833049,
"learning_rate": 1.966792238396801e-05,
"loss": 0.8466,
"step": 1725
},
{
"epoch": 0.21981310791664677,
"grad_norm": 4.961751171277027,
"learning_rate": 1.9667395079591096e-05,
"loss": 0.9572,
"step": 1726
},
{
"epoch": 0.2199404619768534,
"grad_norm": 5.827625570298563,
"learning_rate": 1.9666867363974187e-05,
"loss": 1.0829,
"step": 1727
},
{
"epoch": 0.22006781603706002,
"grad_norm": 6.267381835602661,
"learning_rate": 1.9666339237139723e-05,
"loss": 1.0272,
"step": 1728
},
{
"epoch": 0.22019517009726666,
"grad_norm": 4.6393897470833165,
"learning_rate": 1.966581069911018e-05,
"loss": 0.941,
"step": 1729
},
{
"epoch": 0.2203225241574733,
"grad_norm": 4.630764230227775,
"learning_rate": 1.9665281749908034e-05,
"loss": 1.0309,
"step": 1730
},
{
"epoch": 0.22044987821767992,
"grad_norm": 5.536201975914104,
"learning_rate": 1.966475238955579e-05,
"loss": 1.0152,
"step": 1731
},
{
"epoch": 0.22057723227788656,
"grad_norm": 4.906899767570099,
"learning_rate": 1.9664222618075958e-05,
"loss": 1.0782,
"step": 1732
},
{
"epoch": 0.2207045863380932,
"grad_norm": 5.744214007840506,
"learning_rate": 1.9663692435491084e-05,
"loss": 0.9878,
"step": 1733
},
{
"epoch": 0.22083194039829981,
"grad_norm": 4.846756918555281,
"learning_rate": 1.966316184182372e-05,
"loss": 1.0303,
"step": 1734
},
{
"epoch": 0.22095929445850646,
"grad_norm": 3.883458903060236,
"learning_rate": 1.966263083709643e-05,
"loss": 1.0827,
"step": 1735
},
{
"epoch": 0.2210866485187131,
"grad_norm": 5.049937733181349,
"learning_rate": 1.966209942133181e-05,
"loss": 1.0007,
"step": 1736
},
{
"epoch": 0.2212140025789197,
"grad_norm": 3.77498605990836,
"learning_rate": 1.966156759455246e-05,
"loss": 0.9559,
"step": 1737
},
{
"epoch": 0.22134135663912635,
"grad_norm": 6.192356246123623,
"learning_rate": 1.9661035356781007e-05,
"loss": 0.9042,
"step": 1738
},
{
"epoch": 0.221468710699333,
"grad_norm": 7.090534035783035,
"learning_rate": 1.9660502708040094e-05,
"loss": 1.0616,
"step": 1739
},
{
"epoch": 0.2215960647595396,
"grad_norm": 5.8361028726987545,
"learning_rate": 1.965996964835237e-05,
"loss": 1.0419,
"step": 1740
},
{
"epoch": 0.22172341881974625,
"grad_norm": 5.605022984235532,
"learning_rate": 1.9659436177740518e-05,
"loss": 0.9657,
"step": 1741
},
{
"epoch": 0.2218507728799529,
"grad_norm": 5.7205697588158575,
"learning_rate": 1.965890229622723e-05,
"loss": 0.9856,
"step": 1742
},
{
"epoch": 0.2219781269401595,
"grad_norm": 5.185543196797189,
"learning_rate": 1.9658368003835223e-05,
"loss": 1.0353,
"step": 1743
},
{
"epoch": 0.22210548100036615,
"grad_norm": 9.360388597426082,
"learning_rate": 1.965783330058721e-05,
"loss": 1.1027,
"step": 1744
},
{
"epoch": 0.2222328350605728,
"grad_norm": 7.419286400959521,
"learning_rate": 1.9657298186505952e-05,
"loss": 1.0728,
"step": 1745
},
{
"epoch": 0.2223601891207794,
"grad_norm": 6.056021463434238,
"learning_rate": 1.96567626616142e-05,
"loss": 1.0296,
"step": 1746
},
{
"epoch": 0.22248754318098604,
"grad_norm": 4.617228809565627,
"learning_rate": 1.9656226725934745e-05,
"loss": 0.9012,
"step": 1747
},
{
"epoch": 0.22261489724119268,
"grad_norm": 4.107763634438576,
"learning_rate": 1.965569037949038e-05,
"loss": 0.9253,
"step": 1748
},
{
"epoch": 0.2227422513013993,
"grad_norm": 6.390962251377734,
"learning_rate": 1.9655153622303918e-05,
"loss": 1.0192,
"step": 1749
},
{
"epoch": 0.22286960536160594,
"grad_norm": 4.793158368103816,
"learning_rate": 1.9654616454398194e-05,
"loss": 0.9457,
"step": 1750
},
{
"epoch": 0.22299695942181258,
"grad_norm": 4.688639998095213,
"learning_rate": 1.9654078875796064e-05,
"loss": 0.9644,
"step": 1751
},
{
"epoch": 0.2231243134820192,
"grad_norm": 5.905266573942803,
"learning_rate": 1.9653540886520387e-05,
"loss": 0.9638,
"step": 1752
},
{
"epoch": 0.22325166754222583,
"grad_norm": 5.442243589670223,
"learning_rate": 1.9653002486594057e-05,
"loss": 1.0104,
"step": 1753
},
{
"epoch": 0.22337902160243248,
"grad_norm": 4.397132693549584,
"learning_rate": 1.965246367603997e-05,
"loss": 0.9889,
"step": 1754
},
{
"epoch": 0.2235063756626391,
"grad_norm": 5.403560785443919,
"learning_rate": 1.965192445488105e-05,
"loss": 0.9676,
"step": 1755
},
{
"epoch": 0.22363372972284573,
"grad_norm": 6.487656831407575,
"learning_rate": 1.9651384823140237e-05,
"loss": 1.1338,
"step": 1756
},
{
"epoch": 0.22376108378305234,
"grad_norm": 5.990041097203875,
"learning_rate": 1.9650844780840475e-05,
"loss": 1.0659,
"step": 1757
},
{
"epoch": 0.22388843784325899,
"grad_norm": 6.69981491060953,
"learning_rate": 1.9650304328004752e-05,
"loss": 0.9826,
"step": 1758
},
{
"epoch": 0.22401579190346563,
"grad_norm": 5.43961722710271,
"learning_rate": 1.9649763464656052e-05,
"loss": 0.9696,
"step": 1759
},
{
"epoch": 0.22414314596367224,
"grad_norm": 5.3903222925085155,
"learning_rate": 1.9649222190817382e-05,
"loss": 1.0604,
"step": 1760
},
{
"epoch": 0.22427050002387888,
"grad_norm": 6.023288987205302,
"learning_rate": 1.9648680506511763e-05,
"loss": 0.9395,
"step": 1761
},
{
"epoch": 0.22439785408408552,
"grad_norm": 6.729129031364457,
"learning_rate": 1.9648138411762245e-05,
"loss": 1.0276,
"step": 1762
},
{
"epoch": 0.22452520814429214,
"grad_norm": 5.303370631142946,
"learning_rate": 1.9647595906591884e-05,
"loss": 0.9666,
"step": 1763
},
{
"epoch": 0.22465256220449878,
"grad_norm": 4.003173046434199,
"learning_rate": 1.964705299102376e-05,
"loss": 0.9826,
"step": 1764
},
{
"epoch": 0.22477991626470542,
"grad_norm": 4.822631673720893,
"learning_rate": 1.9646509665080967e-05,
"loss": 0.9451,
"step": 1765
},
{
"epoch": 0.22490727032491203,
"grad_norm": 11.400257950926209,
"learning_rate": 1.9645965928786615e-05,
"loss": 0.9215,
"step": 1766
},
{
"epoch": 0.22503462438511868,
"grad_norm": 12.665581171471478,
"learning_rate": 1.9645421782163838e-05,
"loss": 0.8822,
"step": 1767
},
{
"epoch": 0.22516197844532532,
"grad_norm": 6.241716923360686,
"learning_rate": 1.964487722523578e-05,
"loss": 0.9472,
"step": 1768
},
{
"epoch": 0.22528933250553193,
"grad_norm": 4.949288889351066,
"learning_rate": 1.9644332258025604e-05,
"loss": 0.8934,
"step": 1769
},
{
"epoch": 0.22541668656573857,
"grad_norm": 4.900761541238467,
"learning_rate": 1.96437868805565e-05,
"loss": 1.0991,
"step": 1770
},
{
"epoch": 0.2255440406259452,
"grad_norm": 6.7477896568198945,
"learning_rate": 1.9643241092851664e-05,
"loss": 0.9397,
"step": 1771
},
{
"epoch": 0.22567139468615183,
"grad_norm": 4.4663847952956806,
"learning_rate": 1.964269489493431e-05,
"loss": 0.9228,
"step": 1772
},
{
"epoch": 0.22579874874635847,
"grad_norm": 4.780268150692939,
"learning_rate": 1.9642148286827674e-05,
"loss": 1.0703,
"step": 1773
},
{
"epoch": 0.2259261028065651,
"grad_norm": 3.7715650378964707,
"learning_rate": 1.964160126855501e-05,
"loss": 0.9436,
"step": 1774
},
{
"epoch": 0.22605345686677172,
"grad_norm": 4.992705937599941,
"learning_rate": 1.964105384013959e-05,
"loss": 1.0159,
"step": 1775
},
{
"epoch": 0.22618081092697837,
"grad_norm": 4.893049230375066,
"learning_rate": 1.9640506001604693e-05,
"loss": 0.9875,
"step": 1776
},
{
"epoch": 0.226308164987185,
"grad_norm": 5.730724700849259,
"learning_rate": 1.9639957752973632e-05,
"loss": 0.9165,
"step": 1777
},
{
"epoch": 0.22643551904739162,
"grad_norm": 5.960767591062262,
"learning_rate": 1.963940909426972e-05,
"loss": 1.0084,
"step": 1778
},
{
"epoch": 0.22656287310759826,
"grad_norm": 5.127943728705067,
"learning_rate": 1.9638860025516305e-05,
"loss": 0.9076,
"step": 1779
},
{
"epoch": 0.2266902271678049,
"grad_norm": 6.308130546075289,
"learning_rate": 1.963831054673674e-05,
"loss": 0.9977,
"step": 1780
},
{
"epoch": 0.22681758122801152,
"grad_norm": 7.182944803137629,
"learning_rate": 1.96377606579544e-05,
"loss": 1.0313,
"step": 1781
},
{
"epoch": 0.22694493528821816,
"grad_norm": 4.015812375932955,
"learning_rate": 1.9637210359192673e-05,
"loss": 0.9307,
"step": 1782
},
{
"epoch": 0.2270722893484248,
"grad_norm": 4.165843704331801,
"learning_rate": 1.9636659650474973e-05,
"loss": 0.9104,
"step": 1783
},
{
"epoch": 0.2271996434086314,
"grad_norm": 5.612109524539742,
"learning_rate": 1.9636108531824725e-05,
"loss": 1.0234,
"step": 1784
},
{
"epoch": 0.22732699746883805,
"grad_norm": 6.010012537582954,
"learning_rate": 1.9635557003265374e-05,
"loss": 0.8986,
"step": 1785
},
{
"epoch": 0.2274543515290447,
"grad_norm": 6.456680316007987,
"learning_rate": 1.9635005064820377e-05,
"loss": 0.9593,
"step": 1786
},
{
"epoch": 0.2275817055892513,
"grad_norm": 7.727566645127223,
"learning_rate": 1.9634452716513215e-05,
"loss": 1.0032,
"step": 1787
},
{
"epoch": 0.22770905964945795,
"grad_norm": 5.337484691160166,
"learning_rate": 1.9633899958367384e-05,
"loss": 0.9968,
"step": 1788
},
{
"epoch": 0.2278364137096646,
"grad_norm": 6.736345679884345,
"learning_rate": 1.9633346790406402e-05,
"loss": 0.9217,
"step": 1789
},
{
"epoch": 0.2279637677698712,
"grad_norm": 5.726731364218154,
"learning_rate": 1.9632793212653795e-05,
"loss": 0.9961,
"step": 1790
},
{
"epoch": 0.22809112183007785,
"grad_norm": 4.315122829558817,
"learning_rate": 1.9632239225133116e-05,
"loss": 0.9914,
"step": 1791
},
{
"epoch": 0.2282184758902845,
"grad_norm": 5.034835959132728,
"learning_rate": 1.9631684827867926e-05,
"loss": 1.0849,
"step": 1792
},
{
"epoch": 0.2283458299504911,
"grad_norm": 13.966661267215047,
"learning_rate": 1.9631130020881806e-05,
"loss": 1.0368,
"step": 1793
},
{
"epoch": 0.22847318401069774,
"grad_norm": 5.465619221025032,
"learning_rate": 1.963057480419837e-05,
"loss": 0.9769,
"step": 1794
},
{
"epoch": 0.22860053807090439,
"grad_norm": 4.252969714040979,
"learning_rate": 1.9630019177841224e-05,
"loss": 1.0516,
"step": 1795
},
{
"epoch": 0.228727892131111,
"grad_norm": 5.119678432678006,
"learning_rate": 1.9629463141834008e-05,
"loss": 0.9679,
"step": 1796
},
{
"epoch": 0.22885524619131764,
"grad_norm": 6.089616361413847,
"learning_rate": 1.9628906696200375e-05,
"loss": 1.0558,
"step": 1797
},
{
"epoch": 0.22898260025152428,
"grad_norm": 3.8992771016674848,
"learning_rate": 1.9628349840963997e-05,
"loss": 1.0169,
"step": 1798
},
{
"epoch": 0.2291099543117309,
"grad_norm": 4.4152406993673186,
"learning_rate": 1.9627792576148558e-05,
"loss": 0.9963,
"step": 1799
},
{
"epoch": 0.22923730837193754,
"grad_norm": 4.569440187853551,
"learning_rate": 1.9627234901777768e-05,
"loss": 1.0469,
"step": 1800
},
{
"epoch": 0.22936466243214418,
"grad_norm": 4.675970929345057,
"learning_rate": 1.9626676817875343e-05,
"loss": 0.9363,
"step": 1801
},
{
"epoch": 0.2294920164923508,
"grad_norm": 6.711610771396723,
"learning_rate": 1.9626118324465035e-05,
"loss": 0.9359,
"step": 1802
},
{
"epoch": 0.22961937055255743,
"grad_norm": 5.1523400480938175,
"learning_rate": 1.9625559421570587e-05,
"loss": 0.9964,
"step": 1803
},
{
"epoch": 0.22974672461276405,
"grad_norm": 4.124765562851582,
"learning_rate": 1.9625000109215787e-05,
"loss": 0.9782,
"step": 1804
},
{
"epoch": 0.2298740786729707,
"grad_norm": 5.117340877651308,
"learning_rate": 1.962444038742442e-05,
"loss": 0.9557,
"step": 1805
},
{
"epoch": 0.23000143273317733,
"grad_norm": 5.100902941188911,
"learning_rate": 1.96238802562203e-05,
"loss": 1.0152,
"step": 1806
},
{
"epoch": 0.23012878679338394,
"grad_norm": 7.678748631254456,
"learning_rate": 1.962331971562725e-05,
"loss": 0.9388,
"step": 1807
},
{
"epoch": 0.23025614085359059,
"grad_norm": 4.230898623055076,
"learning_rate": 1.9622758765669117e-05,
"loss": 1.0215,
"step": 1808
},
{
"epoch": 0.23038349491379723,
"grad_norm": 5.386940755826735,
"learning_rate": 1.9622197406369764e-05,
"loss": 1.0412,
"step": 1809
},
{
"epoch": 0.23051084897400384,
"grad_norm": 4.193235881463476,
"learning_rate": 1.962163563775307e-05,
"loss": 0.9781,
"step": 1810
},
{
"epoch": 0.23063820303421048,
"grad_norm": 8.132097636089389,
"learning_rate": 1.962107345984293e-05,
"loss": 0.9979,
"step": 1811
},
{
"epoch": 0.23076555709441712,
"grad_norm": 5.556965678651519,
"learning_rate": 1.962051087266326e-05,
"loss": 0.9099,
"step": 1812
},
{
"epoch": 0.23089291115462374,
"grad_norm": 6.530311918651575,
"learning_rate": 1.9619947876237996e-05,
"loss": 0.9762,
"step": 1813
},
{
"epoch": 0.23102026521483038,
"grad_norm": 5.109577139959268,
"learning_rate": 1.9619384470591082e-05,
"loss": 1.0315,
"step": 1814
},
{
"epoch": 0.23114761927503702,
"grad_norm": 5.855383406024149,
"learning_rate": 1.9618820655746488e-05,
"loss": 0.9857,
"step": 1815
},
{
"epoch": 0.23127497333524363,
"grad_norm": 4.980952029248604,
"learning_rate": 1.961825643172819e-05,
"loss": 1.132,
"step": 1816
},
{
"epoch": 0.23140232739545027,
"grad_norm": 3.5464613080863323,
"learning_rate": 1.96176917985602e-05,
"loss": 0.855,
"step": 1817
},
{
"epoch": 0.23152968145565692,
"grad_norm": 5.782907939000674,
"learning_rate": 1.9617126756266533e-05,
"loss": 1.062,
"step": 1818
},
{
"epoch": 0.23165703551586353,
"grad_norm": 5.962578743788539,
"learning_rate": 1.961656130487122e-05,
"loss": 0.9667,
"step": 1819
},
{
"epoch": 0.23178438957607017,
"grad_norm": 6.864858464548359,
"learning_rate": 1.9615995444398326e-05,
"loss": 1.0125,
"step": 1820
},
{
"epoch": 0.2319117436362768,
"grad_norm": 5.593194518399508,
"learning_rate": 1.9615429174871915e-05,
"loss": 0.9814,
"step": 1821
},
{
"epoch": 0.23203909769648343,
"grad_norm": 4.107295797592486,
"learning_rate": 1.961486249631607e-05,
"loss": 0.9323,
"step": 1822
},
{
"epoch": 0.23216645175669007,
"grad_norm": 6.355699531477541,
"learning_rate": 1.9614295408754908e-05,
"loss": 0.9149,
"step": 1823
},
{
"epoch": 0.2322938058168967,
"grad_norm": 6.169294604965106,
"learning_rate": 1.961372791221254e-05,
"loss": 1.0927,
"step": 1824
},
{
"epoch": 0.23242115987710332,
"grad_norm": 5.650601933916082,
"learning_rate": 1.961316000671312e-05,
"loss": 0.9797,
"step": 1825
},
{
"epoch": 0.23254851393730996,
"grad_norm": 5.9618915879234455,
"learning_rate": 1.9612591692280798e-05,
"loss": 0.9699,
"step": 1826
},
{
"epoch": 0.2326758679975166,
"grad_norm": 5.440465710283945,
"learning_rate": 1.961202296893975e-05,
"loss": 1.0496,
"step": 1827
},
{
"epoch": 0.23280322205772322,
"grad_norm": 4.8719408266955835,
"learning_rate": 1.961145383671417e-05,
"loss": 1.0317,
"step": 1828
},
{
"epoch": 0.23293057611792986,
"grad_norm": 6.786863473513039,
"learning_rate": 1.961088429562827e-05,
"loss": 1.0223,
"step": 1829
},
{
"epoch": 0.2330579301781365,
"grad_norm": 5.715195181543214,
"learning_rate": 1.9610314345706275e-05,
"loss": 1.091,
"step": 1830
},
{
"epoch": 0.23318528423834312,
"grad_norm": 5.537056154671483,
"learning_rate": 1.9609743986972427e-05,
"loss": 1.0621,
"step": 1831
},
{
"epoch": 0.23331263829854976,
"grad_norm": 5.431294901742344,
"learning_rate": 1.9609173219450998e-05,
"loss": 0.9379,
"step": 1832
},
{
"epoch": 0.2334399923587564,
"grad_norm": 4.635917562217512,
"learning_rate": 1.960860204316626e-05,
"loss": 1.034,
"step": 1833
},
{
"epoch": 0.233567346418963,
"grad_norm": 5.937687921059879,
"learning_rate": 1.960803045814251e-05,
"loss": 0.9789,
"step": 1834
},
{
"epoch": 0.23369470047916965,
"grad_norm": 4.805508377539422,
"learning_rate": 1.9607458464404065e-05,
"loss": 0.9514,
"step": 1835
},
{
"epoch": 0.2338220545393763,
"grad_norm": 5.356028944192273,
"learning_rate": 1.9606886061975258e-05,
"loss": 0.8895,
"step": 1836
},
{
"epoch": 0.2339494085995829,
"grad_norm": 4.521239278589203,
"learning_rate": 1.960631325088044e-05,
"loss": 1.1037,
"step": 1837
},
{
"epoch": 0.23407676265978955,
"grad_norm": 4.408012308259316,
"learning_rate": 1.9605740031143972e-05,
"loss": 0.9899,
"step": 1838
},
{
"epoch": 0.2342041167199962,
"grad_norm": 6.913754351801961,
"learning_rate": 1.9605166402790242e-05,
"loss": 0.8878,
"step": 1839
},
{
"epoch": 0.2343314707802028,
"grad_norm": 5.3691903587638405,
"learning_rate": 1.960459236584365e-05,
"loss": 0.9946,
"step": 1840
},
{
"epoch": 0.23445882484040945,
"grad_norm": 4.323790435116697,
"learning_rate": 1.9604017920328613e-05,
"loss": 0.944,
"step": 1841
},
{
"epoch": 0.2345861789006161,
"grad_norm": 5.879725077942632,
"learning_rate": 1.9603443066269575e-05,
"loss": 0.9666,
"step": 1842
},
{
"epoch": 0.2347135329608227,
"grad_norm": 6.673614632504906,
"learning_rate": 1.960286780369098e-05,
"loss": 0.9177,
"step": 1843
},
{
"epoch": 0.23484088702102934,
"grad_norm": 5.059346665131627,
"learning_rate": 1.96022921326173e-05,
"loss": 0.9912,
"step": 1844
},
{
"epoch": 0.23496824108123598,
"grad_norm": 5.094966438939124,
"learning_rate": 1.9601716053073034e-05,
"loss": 0.9824,
"step": 1845
},
{
"epoch": 0.2350955951414426,
"grad_norm": 4.978471566473348,
"learning_rate": 1.9601139565082677e-05,
"loss": 1.0614,
"step": 1846
},
{
"epoch": 0.23522294920164924,
"grad_norm": 6.046248529200191,
"learning_rate": 1.9600562668670756e-05,
"loss": 1.0039,
"step": 1847
},
{
"epoch": 0.23535030326185585,
"grad_norm": 4.59692065788742,
"learning_rate": 1.959998536386181e-05,
"loss": 0.997,
"step": 1848
},
{
"epoch": 0.2354776573220625,
"grad_norm": 4.304352956742825,
"learning_rate": 1.9599407650680397e-05,
"loss": 1.0137,
"step": 1849
},
{
"epoch": 0.23560501138226914,
"grad_norm": 5.236432118220893,
"learning_rate": 1.9598829529151096e-05,
"loss": 0.9861,
"step": 1850
},
{
"epoch": 0.23573236544247575,
"grad_norm": 4.803484113145805,
"learning_rate": 1.9598250999298495e-05,
"loss": 0.9934,
"step": 1851
},
{
"epoch": 0.2358597195026824,
"grad_norm": 4.582977535424973,
"learning_rate": 1.9597672061147207e-05,
"loss": 0.9304,
"step": 1852
},
{
"epoch": 0.23598707356288903,
"grad_norm": 4.685459150452763,
"learning_rate": 1.9597092714721858e-05,
"loss": 0.9713,
"step": 1853
},
{
"epoch": 0.23611442762309565,
"grad_norm": 4.903927781934567,
"learning_rate": 1.9596512960047092e-05,
"loss": 0.914,
"step": 1854
},
{
"epoch": 0.2362417816833023,
"grad_norm": 4.935888099725095,
"learning_rate": 1.9595932797147573e-05,
"loss": 0.9847,
"step": 1855
},
{
"epoch": 0.23636913574350893,
"grad_norm": 4.720197645025626,
"learning_rate": 1.959535222604798e-05,
"loss": 1.0138,
"step": 1856
},
{
"epoch": 0.23649648980371554,
"grad_norm": 7.660251365899786,
"learning_rate": 1.959477124677301e-05,
"loss": 0.928,
"step": 1857
},
{
"epoch": 0.23662384386392218,
"grad_norm": 6.028727107533511,
"learning_rate": 1.9594189859347376e-05,
"loss": 1.0456,
"step": 1858
},
{
"epoch": 0.23675119792412883,
"grad_norm": 5.285424382864863,
"learning_rate": 1.9593608063795808e-05,
"loss": 1.0047,
"step": 1859
},
{
"epoch": 0.23687855198433544,
"grad_norm": 4.413545885489942,
"learning_rate": 1.9593025860143058e-05,
"loss": 0.9942,
"step": 1860
},
{
"epoch": 0.23700590604454208,
"grad_norm": 6.422270899540137,
"learning_rate": 1.9592443248413896e-05,
"loss": 0.9535,
"step": 1861
},
{
"epoch": 0.23713326010474872,
"grad_norm": 4.863549110971738,
"learning_rate": 1.9591860228633093e-05,
"loss": 0.8156,
"step": 1862
},
{
"epoch": 0.23726061416495534,
"grad_norm": 5.639933789453035,
"learning_rate": 1.959127680082546e-05,
"loss": 0.9758,
"step": 1863
},
{
"epoch": 0.23738796822516198,
"grad_norm": 5.272758021907773,
"learning_rate": 1.9590692965015818e-05,
"loss": 0.9138,
"step": 1864
},
{
"epoch": 0.23751532228536862,
"grad_norm": 5.844045050205219,
"learning_rate": 1.9590108721228994e-05,
"loss": 1.0062,
"step": 1865
},
{
"epoch": 0.23764267634557523,
"grad_norm": 5.212810494302638,
"learning_rate": 1.958952406948985e-05,
"loss": 1.0041,
"step": 1866
},
{
"epoch": 0.23777003040578187,
"grad_norm": 4.192451169042947,
"learning_rate": 1.9588939009823246e-05,
"loss": 1.0265,
"step": 1867
},
{
"epoch": 0.23789738446598852,
"grad_norm": 7.087079512352341,
"learning_rate": 1.9588353542254076e-05,
"loss": 1.1435,
"step": 1868
},
{
"epoch": 0.23802473852619513,
"grad_norm": 4.905834010280595,
"learning_rate": 1.9587767666807245e-05,
"loss": 1.0076,
"step": 1869
},
{
"epoch": 0.23815209258640177,
"grad_norm": 5.806849861160403,
"learning_rate": 1.9587181383507678e-05,
"loss": 1.0408,
"step": 1870
},
{
"epoch": 0.2382794466466084,
"grad_norm": 4.947380706135362,
"learning_rate": 1.958659469238031e-05,
"loss": 0.9148,
"step": 1871
},
{
"epoch": 0.23840680070681503,
"grad_norm": 4.263103256718209,
"learning_rate": 1.9586007593450098e-05,
"loss": 0.8835,
"step": 1872
},
{
"epoch": 0.23853415476702167,
"grad_norm": 4.573667680364965,
"learning_rate": 1.958542008674202e-05,
"loss": 1.1058,
"step": 1873
},
{
"epoch": 0.2386615088272283,
"grad_norm": 4.741512948188774,
"learning_rate": 1.9584832172281064e-05,
"loss": 1.0385,
"step": 1874
},
{
"epoch": 0.23878886288743492,
"grad_norm": 5.806566388694644,
"learning_rate": 1.9584243850092246e-05,
"loss": 0.9899,
"step": 1875
},
{
"epoch": 0.23891621694764156,
"grad_norm": 7.7527082079816125,
"learning_rate": 1.9583655120200586e-05,
"loss": 1.1014,
"step": 1876
},
{
"epoch": 0.2390435710078482,
"grad_norm": 5.3338465245025795,
"learning_rate": 1.9583065982631128e-05,
"loss": 1.0165,
"step": 1877
},
{
"epoch": 0.23917092506805482,
"grad_norm": 4.761253878751353,
"learning_rate": 1.9582476437408937e-05,
"loss": 1.1034,
"step": 1878
},
{
"epoch": 0.23929827912826146,
"grad_norm": 6.169684144379374,
"learning_rate": 1.958188648455909e-05,
"loss": 0.9707,
"step": 1879
},
{
"epoch": 0.2394256331884681,
"grad_norm": 5.183876702549582,
"learning_rate": 1.9581296124106682e-05,
"loss": 0.9681,
"step": 1880
},
{
"epoch": 0.23955298724867471,
"grad_norm": 3.8059335661646267,
"learning_rate": 1.9580705356076826e-05,
"loss": 0.8875,
"step": 1881
},
{
"epoch": 0.23968034130888136,
"grad_norm": 4.421657597350017,
"learning_rate": 1.9580114180494655e-05,
"loss": 0.86,
"step": 1882
},
{
"epoch": 0.239807695369088,
"grad_norm": 6.634587673780053,
"learning_rate": 1.9579522597385315e-05,
"loss": 1.0404,
"step": 1883
},
{
"epoch": 0.2399350494292946,
"grad_norm": 4.316824251536347,
"learning_rate": 1.9578930606773975e-05,
"loss": 0.9788,
"step": 1884
},
{
"epoch": 0.24006240348950125,
"grad_norm": 4.444924175014784,
"learning_rate": 1.957833820868581e-05,
"loss": 1.0692,
"step": 1885
},
{
"epoch": 0.2401897575497079,
"grad_norm": 5.789228552780425,
"learning_rate": 1.9577745403146026e-05,
"loss": 1.043,
"step": 1886
},
{
"epoch": 0.2403171116099145,
"grad_norm": 5.8358917438801985,
"learning_rate": 1.9577152190179837e-05,
"loss": 0.9655,
"step": 1887
},
{
"epoch": 0.24044446567012115,
"grad_norm": 3.9383483457232096,
"learning_rate": 1.9576558569812484e-05,
"loss": 0.9138,
"step": 1888
},
{
"epoch": 0.2405718197303278,
"grad_norm": 4.634923238602164,
"learning_rate": 1.957596454206921e-05,
"loss": 1.0484,
"step": 1889
},
{
"epoch": 0.2406991737905344,
"grad_norm": 6.206385603515261,
"learning_rate": 1.9575370106975288e-05,
"loss": 0.9826,
"step": 1890
},
{
"epoch": 0.24082652785074105,
"grad_norm": 4.170206829403069,
"learning_rate": 1.9574775264556005e-05,
"loss": 0.9325,
"step": 1891
},
{
"epoch": 0.2409538819109477,
"grad_norm": 5.2007134608348675,
"learning_rate": 1.9574180014836668e-05,
"loss": 0.9554,
"step": 1892
},
{
"epoch": 0.2410812359711543,
"grad_norm": 6.073407615154317,
"learning_rate": 1.9573584357842592e-05,
"loss": 1.0859,
"step": 1893
},
{
"epoch": 0.24120859003136094,
"grad_norm": 5.51343611815531,
"learning_rate": 1.9572988293599124e-05,
"loss": 1.0487,
"step": 1894
},
{
"epoch": 0.24133594409156756,
"grad_norm": 5.406242238298257,
"learning_rate": 1.9572391822131606e-05,
"loss": 1.0153,
"step": 1895
},
{
"epoch": 0.2414632981517742,
"grad_norm": 3.6163605966280175,
"learning_rate": 1.9571794943465424e-05,
"loss": 0.9876,
"step": 1896
},
{
"epoch": 0.24159065221198084,
"grad_norm": 7.0188141436885605,
"learning_rate": 1.9571197657625967e-05,
"loss": 0.9274,
"step": 1897
},
{
"epoch": 0.24171800627218745,
"grad_norm": 5.502167881358521,
"learning_rate": 1.957059996463864e-05,
"loss": 1.1126,
"step": 1898
},
{
"epoch": 0.2418453603323941,
"grad_norm": 5.245404730091922,
"learning_rate": 1.9570001864528863e-05,
"loss": 0.9728,
"step": 1899
},
{
"epoch": 0.24197271439260074,
"grad_norm": 6.017843417046496,
"learning_rate": 1.956940335732209e-05,
"loss": 0.9613,
"step": 1900
},
{
"epoch": 0.24210006845280735,
"grad_norm": 5.594564708859503,
"learning_rate": 1.9568804443043774e-05,
"loss": 1.0798,
"step": 1901
},
{
"epoch": 0.242227422513014,
"grad_norm": 5.132520043156988,
"learning_rate": 1.956820512171939e-05,
"loss": 0.9427,
"step": 1902
},
{
"epoch": 0.24235477657322063,
"grad_norm": 6.049832352766647,
"learning_rate": 1.956760539337444e-05,
"loss": 1.0261,
"step": 1903
},
{
"epoch": 0.24248213063342725,
"grad_norm": 5.616791054605702,
"learning_rate": 1.9567005258034423e-05,
"loss": 0.9481,
"step": 1904
},
{
"epoch": 0.2426094846936339,
"grad_norm": 6.4642734241444035,
"learning_rate": 1.9566404715724884e-05,
"loss": 0.9793,
"step": 1905
},
{
"epoch": 0.24273683875384053,
"grad_norm": 6.313155640649643,
"learning_rate": 1.9565803766471355e-05,
"loss": 1.035,
"step": 1906
},
{
"epoch": 0.24286419281404714,
"grad_norm": 7.304047484077455,
"learning_rate": 1.9565202410299415e-05,
"loss": 1.0329,
"step": 1907
},
{
"epoch": 0.24299154687425378,
"grad_norm": 6.860041766248072,
"learning_rate": 1.9564600647234628e-05,
"loss": 0.9587,
"step": 1908
},
{
"epoch": 0.24311890093446042,
"grad_norm": 6.482658729007516,
"learning_rate": 1.9563998477302604e-05,
"loss": 0.9059,
"step": 1909
},
{
"epoch": 0.24324625499466704,
"grad_norm": 6.916423778022661,
"learning_rate": 1.9563395900528956e-05,
"loss": 0.8524,
"step": 1910
},
{
"epoch": 0.24337360905487368,
"grad_norm": 4.723823699658888,
"learning_rate": 1.956279291693931e-05,
"loss": 1.0333,
"step": 1911
},
{
"epoch": 0.24350096311508032,
"grad_norm": 8.42233790326866,
"learning_rate": 1.9562189526559333e-05,
"loss": 0.9848,
"step": 1912
},
{
"epoch": 0.24362831717528693,
"grad_norm": 4.949374635284291,
"learning_rate": 1.9561585729414675e-05,
"loss": 0.9388,
"step": 1913
},
{
"epoch": 0.24375567123549358,
"grad_norm": 5.121845297875669,
"learning_rate": 1.9560981525531027e-05,
"loss": 1.0829,
"step": 1914
},
{
"epoch": 0.24388302529570022,
"grad_norm": 4.631129559020309,
"learning_rate": 1.9560376914934098e-05,
"loss": 0.8969,
"step": 1915
},
{
"epoch": 0.24401037935590683,
"grad_norm": 5.316361403994358,
"learning_rate": 1.9559771897649592e-05,
"loss": 1.0067,
"step": 1916
},
{
"epoch": 0.24413773341611347,
"grad_norm": 6.211563635400579,
"learning_rate": 1.9559166473703265e-05,
"loss": 0.8899,
"step": 1917
},
{
"epoch": 0.24426508747632011,
"grad_norm": 6.407445284356433,
"learning_rate": 1.9558560643120855e-05,
"loss": 0.951,
"step": 1918
},
{
"epoch": 0.24439244153652673,
"grad_norm": 5.549064784254947,
"learning_rate": 1.9557954405928142e-05,
"loss": 0.967,
"step": 1919
},
{
"epoch": 0.24451979559673337,
"grad_norm": 5.021470320988695,
"learning_rate": 1.955734776215091e-05,
"loss": 1.0486,
"step": 1920
},
{
"epoch": 0.24464714965694,
"grad_norm": 6.492243026330688,
"learning_rate": 1.955674071181497e-05,
"loss": 0.9918,
"step": 1921
},
{
"epoch": 0.24477450371714662,
"grad_norm": 6.736913640652619,
"learning_rate": 1.955613325494614e-05,
"loss": 0.9308,
"step": 1922
},
{
"epoch": 0.24490185777735327,
"grad_norm": 5.582164850898567,
"learning_rate": 1.955552539157026e-05,
"loss": 0.9773,
"step": 1923
},
{
"epoch": 0.2450292118375599,
"grad_norm": 4.496928968496731,
"learning_rate": 1.9554917121713198e-05,
"loss": 1.0006,
"step": 1924
},
{
"epoch": 0.24515656589776652,
"grad_norm": 4.202497811302691,
"learning_rate": 1.955430844540082e-05,
"loss": 0.9419,
"step": 1925
},
{
"epoch": 0.24528391995797316,
"grad_norm": 7.195137013501995,
"learning_rate": 1.9553699362659016e-05,
"loss": 0.9848,
"step": 1926
},
{
"epoch": 0.2454112740181798,
"grad_norm": 5.762608511807093,
"learning_rate": 1.9553089873513702e-05,
"loss": 0.9671,
"step": 1927
},
{
"epoch": 0.24553862807838642,
"grad_norm": 6.201908174791442,
"learning_rate": 1.9552479977990802e-05,
"loss": 0.9007,
"step": 1928
},
{
"epoch": 0.24566598213859306,
"grad_norm": 6.22489587168181,
"learning_rate": 1.955186967611626e-05,
"loss": 0.9826,
"step": 1929
},
{
"epoch": 0.2457933361987997,
"grad_norm": 7.069124878674373,
"learning_rate": 1.955125896791604e-05,
"loss": 1.1846,
"step": 1930
},
{
"epoch": 0.2459206902590063,
"grad_norm": 5.212998927546745,
"learning_rate": 1.955064785341612e-05,
"loss": 0.8786,
"step": 1931
},
{
"epoch": 0.24604804431921296,
"grad_norm": 4.894650103929042,
"learning_rate": 1.9550036332642496e-05,
"loss": 0.9747,
"step": 1932
},
{
"epoch": 0.2461753983794196,
"grad_norm": 6.111738523392531,
"learning_rate": 1.954942440562118e-05,
"loss": 0.8886,
"step": 1933
},
{
"epoch": 0.2463027524396262,
"grad_norm": 5.085283416081644,
"learning_rate": 1.9548812072378208e-05,
"loss": 1.0079,
"step": 1934
},
{
"epoch": 0.24643010649983285,
"grad_norm": 7.121943838967712,
"learning_rate": 1.954819933293962e-05,
"loss": 1.0411,
"step": 1935
},
{
"epoch": 0.2465574605600395,
"grad_norm": 4.765921361494564,
"learning_rate": 1.954758618733148e-05,
"loss": 0.978,
"step": 1936
},
{
"epoch": 0.2466848146202461,
"grad_norm": 5.157511608796732,
"learning_rate": 1.954697263557988e-05,
"loss": 0.9591,
"step": 1937
},
{
"epoch": 0.24681216868045275,
"grad_norm": 4.500125035562163,
"learning_rate": 1.9546358677710917e-05,
"loss": 0.9751,
"step": 1938
},
{
"epoch": 0.24693952274065936,
"grad_norm": 5.388541794880216,
"learning_rate": 1.9545744313750706e-05,
"loss": 0.9856,
"step": 1939
},
{
"epoch": 0.247066876800866,
"grad_norm": 5.033008339265417,
"learning_rate": 1.954512954372538e-05,
"loss": 0.9379,
"step": 1940
},
{
"epoch": 0.24719423086107264,
"grad_norm": 5.048770702042888,
"learning_rate": 1.954451436766109e-05,
"loss": 1.0365,
"step": 1941
},
{
"epoch": 0.24732158492127926,
"grad_norm": 5.618927623754484,
"learning_rate": 1.954389878558401e-05,
"loss": 0.9645,
"step": 1942
},
{
"epoch": 0.2474489389814859,
"grad_norm": 6.407757539845989,
"learning_rate": 1.954328279752032e-05,
"loss": 1.1237,
"step": 1943
},
{
"epoch": 0.24757629304169254,
"grad_norm": 6.006274121858267,
"learning_rate": 1.9542666403496232e-05,
"loss": 1.0167,
"step": 1944
},
{
"epoch": 0.24770364710189915,
"grad_norm": 5.895080985693145,
"learning_rate": 1.954204960353796e-05,
"loss": 1.0035,
"step": 1945
},
{
"epoch": 0.2478310011621058,
"grad_norm": 4.438704206380312,
"learning_rate": 1.954143239767174e-05,
"loss": 0.972,
"step": 1946
},
{
"epoch": 0.24795835522231244,
"grad_norm": 4.267205808272698,
"learning_rate": 1.9540814785923832e-05,
"loss": 1.0307,
"step": 1947
},
{
"epoch": 0.24808570928251905,
"grad_norm": 5.4795965260572705,
"learning_rate": 1.954019676832051e-05,
"loss": 0.9458,
"step": 1948
},
{
"epoch": 0.2482130633427257,
"grad_norm": 5.286442839000025,
"learning_rate": 1.9539578344888057e-05,
"loss": 0.9832,
"step": 1949
},
{
"epoch": 0.24834041740293233,
"grad_norm": 5.362946768691932,
"learning_rate": 1.9538959515652786e-05,
"loss": 1.0164,
"step": 1950
},
{
"epoch": 0.24846777146313895,
"grad_norm": 5.207804457197322,
"learning_rate": 1.9538340280641018e-05,
"loss": 0.9356,
"step": 1951
},
{
"epoch": 0.2485951255233456,
"grad_norm": 6.356636376204565,
"learning_rate": 1.9537720639879096e-05,
"loss": 1.0118,
"step": 1952
},
{
"epoch": 0.24872247958355223,
"grad_norm": 4.992868296448806,
"learning_rate": 1.953710059339338e-05,
"loss": 1.0178,
"step": 1953
},
{
"epoch": 0.24884983364375884,
"grad_norm": 5.213643939691587,
"learning_rate": 1.9536480141210242e-05,
"loss": 0.9866,
"step": 1954
},
{
"epoch": 0.24897718770396549,
"grad_norm": 4.929728899266538,
"learning_rate": 1.953585928335608e-05,
"loss": 1.0418,
"step": 1955
},
{
"epoch": 0.24910454176417213,
"grad_norm": 6.206567841810802,
"learning_rate": 1.95352380198573e-05,
"loss": 1.0212,
"step": 1956
},
{
"epoch": 0.24923189582437874,
"grad_norm": 4.618441080749805,
"learning_rate": 1.9534616350740336e-05,
"loss": 1.0703,
"step": 1957
},
{
"epoch": 0.24935924988458538,
"grad_norm": 9.527918699338851,
"learning_rate": 1.953399427603163e-05,
"loss": 1.0498,
"step": 1958
},
{
"epoch": 0.24948660394479202,
"grad_norm": 4.475756507734566,
"learning_rate": 1.953337179575764e-05,
"loss": 0.9581,
"step": 1959
},
{
"epoch": 0.24961395800499864,
"grad_norm": 5.777107362895541,
"learning_rate": 1.953274890994485e-05,
"loss": 1.068,
"step": 1960
},
{
"epoch": 0.24974131206520528,
"grad_norm": 5.472723688857767,
"learning_rate": 1.953212561861976e-05,
"loss": 1.0794,
"step": 1961
},
{
"epoch": 0.24986866612541192,
"grad_norm": 4.63711363630185,
"learning_rate": 1.9531501921808877e-05,
"loss": 0.9387,
"step": 1962
},
{
"epoch": 0.24999602018561853,
"grad_norm": 5.787593845735321,
"learning_rate": 1.9530877819538736e-05,
"loss": 0.9188,
"step": 1963
},
{
"epoch": 0.25012337424582515,
"grad_norm": 5.269182491991631,
"learning_rate": 1.9530253311835884e-05,
"loss": 0.9789,
"step": 1964
},
{
"epoch": 0.2502507283060318,
"grad_norm": 4.672931462385088,
"learning_rate": 1.9529628398726892e-05,
"loss": 1.0085,
"step": 1965
},
{
"epoch": 0.25037808236623843,
"grad_norm": 5.314662358638423,
"learning_rate": 1.9529003080238337e-05,
"loss": 0.9308,
"step": 1966
},
{
"epoch": 0.25050543642644507,
"grad_norm": 4.658522083550714,
"learning_rate": 1.9528377356396825e-05,
"loss": 0.9868,
"step": 1967
},
{
"epoch": 0.2506327904866517,
"grad_norm": 4.732259850713794,
"learning_rate": 1.9527751227228964e-05,
"loss": 0.9688,
"step": 1968
},
{
"epoch": 0.25076014454685835,
"grad_norm": 4.549413437815234,
"learning_rate": 1.95271246927614e-05,
"loss": 0.9979,
"step": 1969
},
{
"epoch": 0.25088749860706494,
"grad_norm": 5.563897118149959,
"learning_rate": 1.9526497753020776e-05,
"loss": 0.9984,
"step": 1970
},
{
"epoch": 0.2510148526672716,
"grad_norm": 4.3159957207206885,
"learning_rate": 1.952587040803377e-05,
"loss": 1.0247,
"step": 1971
},
{
"epoch": 0.2511422067274782,
"grad_norm": 3.989694432769271,
"learning_rate": 1.9525242657827063e-05,
"loss": 0.981,
"step": 1972
},
{
"epoch": 0.25126956078768486,
"grad_norm": 4.9109895980746545,
"learning_rate": 1.9524614502427358e-05,
"loss": 1.0609,
"step": 1973
},
{
"epoch": 0.2513969148478915,
"grad_norm": 6.58293704357173,
"learning_rate": 1.9523985941861376e-05,
"loss": 0.9695,
"step": 1974
},
{
"epoch": 0.25152426890809815,
"grad_norm": 5.057107950233266,
"learning_rate": 1.952335697615586e-05,
"loss": 0.9499,
"step": 1975
},
{
"epoch": 0.25165162296830473,
"grad_norm": 5.667578717304745,
"learning_rate": 1.952272760533756e-05,
"loss": 1.0221,
"step": 1976
},
{
"epoch": 0.2517789770285114,
"grad_norm": 7.079981410977373,
"learning_rate": 1.9522097829433252e-05,
"loss": 0.9485,
"step": 1977
},
{
"epoch": 0.251906331088718,
"grad_norm": 5.338151642733225,
"learning_rate": 1.9521467648469728e-05,
"loss": 0.9187,
"step": 1978
},
{
"epoch": 0.25203368514892466,
"grad_norm": 5.3896529496096806,
"learning_rate": 1.9520837062473788e-05,
"loss": 0.9913,
"step": 1979
},
{
"epoch": 0.2521610392091313,
"grad_norm": 6.123413503278127,
"learning_rate": 1.9520206071472264e-05,
"loss": 0.9032,
"step": 1980
},
{
"epoch": 0.25228839326933794,
"grad_norm": 5.591531153636707,
"learning_rate": 1.9519574675491995e-05,
"loss": 0.9196,
"step": 1981
},
{
"epoch": 0.2524157473295445,
"grad_norm": 7.5871264295038205,
"learning_rate": 1.951894287455984e-05,
"loss": 0.9486,
"step": 1982
},
{
"epoch": 0.25254310138975117,
"grad_norm": 5.71256743209739,
"learning_rate": 1.951831066870267e-05,
"loss": 0.9122,
"step": 1983
},
{
"epoch": 0.2526704554499578,
"grad_norm": 5.781656446861163,
"learning_rate": 1.9517678057947385e-05,
"loss": 0.8943,
"step": 1984
},
{
"epoch": 0.25279780951016445,
"grad_norm": 5.738200478605293,
"learning_rate": 1.9517045042320893e-05,
"loss": 0.8755,
"step": 1985
},
{
"epoch": 0.2529251635703711,
"grad_norm": 6.186663946830765,
"learning_rate": 1.951641162185012e-05,
"loss": 1.0455,
"step": 1986
},
{
"epoch": 0.25305251763057773,
"grad_norm": 4.660019136315867,
"learning_rate": 1.9515777796562016e-05,
"loss": 0.9795,
"step": 1987
},
{
"epoch": 0.2531798716907843,
"grad_norm": 4.994360222611413,
"learning_rate": 1.951514356648354e-05,
"loss": 0.9546,
"step": 1988
},
{
"epoch": 0.25330722575099096,
"grad_norm": 5.91407604543528,
"learning_rate": 1.951450893164167e-05,
"loss": 0.99,
"step": 1989
},
{
"epoch": 0.2534345798111976,
"grad_norm": 5.697067429917377,
"learning_rate": 1.9513873892063403e-05,
"loss": 0.9586,
"step": 1990
},
{
"epoch": 0.25356193387140424,
"grad_norm": 5.019744037318567,
"learning_rate": 1.9513238447775757e-05,
"loss": 0.9406,
"step": 1991
},
{
"epoch": 0.2536892879316109,
"grad_norm": 5.232808621741993,
"learning_rate": 1.951260259880576e-05,
"loss": 0.9392,
"step": 1992
},
{
"epoch": 0.2538166419918175,
"grad_norm": 6.66558555738008,
"learning_rate": 1.9511966345180457e-05,
"loss": 0.9115,
"step": 1993
},
{
"epoch": 0.2539439960520241,
"grad_norm": 4.94620991499485,
"learning_rate": 1.9511329686926922e-05,
"loss": 1.0263,
"step": 1994
},
{
"epoch": 0.25407135011223075,
"grad_norm": 5.9253145158562175,
"learning_rate": 1.951069262407223e-05,
"loss": 1.0708,
"step": 1995
},
{
"epoch": 0.2541987041724374,
"grad_norm": 4.381045061615913,
"learning_rate": 1.9510055156643485e-05,
"loss": 0.8709,
"step": 1996
},
{
"epoch": 0.25432605823264404,
"grad_norm": 7.938926268266305,
"learning_rate": 1.95094172846678e-05,
"loss": 1.0671,
"step": 1997
},
{
"epoch": 0.2544534122928507,
"grad_norm": 7.953716143689362,
"learning_rate": 1.9508779008172314e-05,
"loss": 0.9639,
"step": 1998
},
{
"epoch": 0.2545807663530573,
"grad_norm": 7.371930016817276,
"learning_rate": 1.950814032718418e-05,
"loss": 0.9721,
"step": 1999
},
{
"epoch": 0.2547081204132639,
"grad_norm": 5.248665594633744,
"learning_rate": 1.9507501241730557e-05,
"loss": 0.9776,
"step": 2000
}
],
"logging_steps": 1.0,
"max_steps": 15704,
"num_input_tokens_seen": 0,
"num_train_epochs": 2,
"save_steps": 500,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 2.315219424473042e+22,
"train_batch_size": 2,
"trial_name": null,
"trial_params": null
}