abhinavp's picture
Training in progress, epoch 0, checkpoint
2d09844 verified
{
"best_metric": 3.887401580810547,
"best_model_checkpoint": "/mmfs1/gscratch/stf/abhinavp/corpus-filtering/outputs/npi-sim-ques/transformer/0/checkpoints/checkpoint-381600",
"epoch": 0.025000606015738065,
"eval_steps": 10,
"global_step": 381600,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0,
"learning_rate": 4.999998362119627e-05,
"loss": 10.859,
"step": 1
},
{
"epoch": 0.0,
"learning_rate": 4.999161405248948e-05,
"loss": 6.8241,
"step": 512
},
{
"epoch": 0.0,
"learning_rate": 4.998322810497896e-05,
"loss": 6.1916,
"step": 1024
},
{
"epoch": 0.0,
"learning_rate": 4.997484215746844e-05,
"loss": 5.9753,
"step": 1536
},
{
"epoch": 0.0,
"learning_rate": 4.996645620995792e-05,
"loss": 5.8086,
"step": 2048
},
{
"epoch": 0.0,
"learning_rate": 4.99580702624474e-05,
"loss": 5.7055,
"step": 2560
},
{
"epoch": 0.0,
"learning_rate": 4.994968431493688e-05,
"loss": 5.5974,
"step": 3072
},
{
"epoch": 0.0,
"learning_rate": 4.994129836742636e-05,
"loss": 5.527,
"step": 3584
},
{
"epoch": 0.0,
"learning_rate": 4.993291241991584e-05,
"loss": 5.4481,
"step": 4096
},
{
"epoch": 0.0,
"learning_rate": 4.992452647240532e-05,
"loss": 5.3978,
"step": 4608
},
{
"epoch": 0.0,
"learning_rate": 4.99161405248948e-05,
"loss": 5.3376,
"step": 5120
},
{
"epoch": 0.0,
"learning_rate": 4.990775457738428e-05,
"loss": 5.3155,
"step": 5632
},
{
"epoch": 0.0,
"learning_rate": 4.989938500867749e-05,
"loss": 5.2639,
"step": 6144
},
{
"epoch": 0.0,
"learning_rate": 4.989099906116697e-05,
"loss": 5.1982,
"step": 6656
},
{
"epoch": 0.0,
"learning_rate": 4.988261311365645e-05,
"loss": 5.1727,
"step": 7168
},
{
"epoch": 0.0,
"learning_rate": 4.987422716614593e-05,
"loss": 5.1287,
"step": 7680
},
{
"epoch": 0.0,
"learning_rate": 4.986584121863541e-05,
"loss": 5.1115,
"step": 8192
},
{
"epoch": 0.0,
"learning_rate": 4.985745527112489e-05,
"loss": 5.0679,
"step": 8704
},
{
"epoch": 0.0,
"learning_rate": 4.984906932361437e-05,
"loss": 5.0395,
"step": 9216
},
{
"epoch": 0.0,
"learning_rate": 4.984068337610385e-05,
"loss": 5.0152,
"step": 9728
},
{
"epoch": 0.0,
"learning_rate": 4.983229742859333e-05,
"loss": 4.9959,
"step": 10240
},
{
"epoch": 0.0,
"learning_rate": 4.9823927859886547e-05,
"loss": 4.974,
"step": 10752
},
{
"epoch": 0.0,
"learning_rate": 4.9815541912376026e-05,
"loss": 4.9529,
"step": 11264
},
{
"epoch": 0.0,
"learning_rate": 4.9807155964865506e-05,
"loss": 4.9343,
"step": 11776
},
{
"epoch": 0.0,
"learning_rate": 4.9798770017354986e-05,
"loss": 4.9132,
"step": 12288
},
{
"epoch": 0.0,
"learning_rate": 4.9790384069844466e-05,
"loss": 4.8913,
"step": 12800
},
{
"epoch": 0.0,
"learning_rate": 4.9782014501137675e-05,
"loss": 4.8642,
"step": 13312
},
{
"epoch": 0.0,
"learning_rate": 4.9773628553627155e-05,
"loss": 4.85,
"step": 13824
},
{
"epoch": 0.0,
"learning_rate": 4.9765242606116635e-05,
"loss": 4.8278,
"step": 14336
},
{
"epoch": 0.0,
"learning_rate": 4.9756856658606115e-05,
"loss": 4.8214,
"step": 14848
},
{
"epoch": 0.01,
"learning_rate": 4.9748487089899324e-05,
"loss": 4.8035,
"step": 15360
},
{
"epoch": 0.01,
"learning_rate": 4.9740101142388804e-05,
"loss": 4.7879,
"step": 15872
},
{
"epoch": 0.01,
"learning_rate": 4.9731715194878284e-05,
"loss": 4.7779,
"step": 16384
},
{
"epoch": 0.01,
"learning_rate": 4.97233456261715e-05,
"loss": 4.7592,
"step": 16896
},
{
"epoch": 0.01,
"learning_rate": 4.971495967866098e-05,
"loss": 4.7561,
"step": 17408
},
{
"epoch": 0.01,
"learning_rate": 4.970657373115046e-05,
"loss": 4.7505,
"step": 17920
},
{
"epoch": 0.01,
"learning_rate": 4.969820416244367e-05,
"loss": 4.7318,
"step": 18432
},
{
"epoch": 0.01,
"learning_rate": 4.968981821493315e-05,
"loss": 4.6964,
"step": 18944
},
{
"epoch": 0.01,
"learning_rate": 4.968143226742263e-05,
"loss": 4.7076,
"step": 19456
},
{
"epoch": 0.01,
"learning_rate": 4.967304631991211e-05,
"loss": 4.6806,
"step": 19968
},
{
"epoch": 0.01,
"learning_rate": 4.966466037240159e-05,
"loss": 4.6745,
"step": 20480
},
{
"epoch": 0.01,
"learning_rate": 4.965627442489107e-05,
"loss": 4.6717,
"step": 20992
},
{
"epoch": 0.01,
"learning_rate": 4.964788847738054e-05,
"loss": 4.6646,
"step": 21504
},
{
"epoch": 0.01,
"learning_rate": 4.963950252987002e-05,
"loss": 4.6471,
"step": 22016
},
{
"epoch": 0.01,
"learning_rate": 4.96311165823595e-05,
"loss": 4.6479,
"step": 22528
},
{
"epoch": 0.01,
"learning_rate": 4.962273063484898e-05,
"loss": 4.6399,
"step": 23040
},
{
"epoch": 0.01,
"learning_rate": 4.961434468733847e-05,
"loss": 4.6312,
"step": 23552
},
{
"epoch": 0.01,
"learning_rate": 4.960595873982795e-05,
"loss": 4.6228,
"step": 24064
},
{
"epoch": 0.01,
"learning_rate": 4.959757279231743e-05,
"loss": 4.6012,
"step": 24576
},
{
"epoch": 0.01,
"learning_rate": 4.9589219602414374e-05,
"loss": 4.5941,
"step": 25088
},
{
"epoch": 0.01,
"learning_rate": 4.958083365490385e-05,
"loss": 4.5998,
"step": 25600
},
{
"epoch": 0.01,
"learning_rate": 4.957244770739333e-05,
"loss": 4.5853,
"step": 26112
},
{
"epoch": 0.01,
"learning_rate": 4.956406175988281e-05,
"loss": 4.5702,
"step": 26624
},
{
"epoch": 0.01,
"learning_rate": 4.955567581237229e-05,
"loss": 4.5583,
"step": 27136
},
{
"epoch": 0.01,
"learning_rate": 4.954728986486177e-05,
"loss": 4.5627,
"step": 27648
},
{
"epoch": 0.01,
"learning_rate": 4.953890391735125e-05,
"loss": 4.5466,
"step": 28160
},
{
"epoch": 0.01,
"learning_rate": 4.9530517969840727e-05,
"loss": 4.5598,
"step": 28672
},
{
"epoch": 0.01,
"learning_rate": 4.9522148401133936e-05,
"loss": 4.5335,
"step": 29184
},
{
"epoch": 0.01,
"learning_rate": 4.951376245362342e-05,
"loss": 4.5394,
"step": 29696
},
{
"epoch": 0.01,
"learning_rate": 4.95053765061129e-05,
"loss": 4.5325,
"step": 30208
},
{
"epoch": 0.01,
"learning_rate": 4.949699055860238e-05,
"loss": 4.5083,
"step": 30720
},
{
"epoch": 0.01,
"learning_rate": 4.948862098989559e-05,
"loss": 4.5159,
"step": 31232
},
{
"epoch": 0.01,
"learning_rate": 4.948023504238507e-05,
"loss": 4.5044,
"step": 31744
},
{
"epoch": 0.01,
"learning_rate": 4.947184909487455e-05,
"loss": 4.4941,
"step": 32256
},
{
"epoch": 0.01,
"learning_rate": 4.946346314736403e-05,
"loss": 4.4924,
"step": 32768
},
{
"epoch": 0.01,
"learning_rate": 4.945509357865724e-05,
"loss": 4.508,
"step": 33280
},
{
"epoch": 0.01,
"learning_rate": 4.944670763114672e-05,
"loss": 4.4837,
"step": 33792
},
{
"epoch": 0.01,
"learning_rate": 4.94383216836362e-05,
"loss": 4.4751,
"step": 34304
},
{
"epoch": 0.01,
"learning_rate": 4.942993573612568e-05,
"loss": 4.4607,
"step": 34816
},
{
"epoch": 0.01,
"learning_rate": 4.942154978861516e-05,
"loss": 4.4608,
"step": 35328
},
{
"epoch": 0.01,
"learning_rate": 4.941318021990837e-05,
"loss": 4.4767,
"step": 35840
},
{
"epoch": 0.01,
"learning_rate": 4.9404794272397856e-05,
"loss": 4.477,
"step": 36352
},
{
"epoch": 0.01,
"learning_rate": 4.9396408324887336e-05,
"loss": 4.4497,
"step": 36864
},
{
"epoch": 0.01,
"learning_rate": 4.9388022377376816e-05,
"loss": 4.4677,
"step": 37376
},
{
"epoch": 0.01,
"learning_rate": 4.9379652808670025e-05,
"loss": 4.4606,
"step": 37888
},
{
"epoch": 0.01,
"learning_rate": 4.9371266861159505e-05,
"loss": 4.4457,
"step": 38400
},
{
"epoch": 0.01,
"learning_rate": 4.9362897292452714e-05,
"loss": 4.433,
"step": 38912
},
{
"epoch": 0.01,
"learning_rate": 4.9354511344942194e-05,
"loss": 4.4378,
"step": 39424
},
{
"epoch": 0.01,
"learning_rate": 4.9346125397431674e-05,
"loss": 4.4264,
"step": 39936
},
{
"epoch": 0.01,
"learning_rate": 4.9337739449921154e-05,
"loss": 4.4213,
"step": 40448
},
{
"epoch": 0.01,
"learning_rate": 4.9329353502410634e-05,
"loss": 4.4215,
"step": 40960
},
{
"epoch": 0.01,
"learning_rate": 4.9320967554900114e-05,
"loss": 4.4169,
"step": 41472
},
{
"epoch": 0.01,
"learning_rate": 4.9312581607389594e-05,
"loss": 4.421,
"step": 41984
},
{
"epoch": 0.01,
"learning_rate": 4.9304195659879074e-05,
"loss": 4.4168,
"step": 42496
},
{
"epoch": 0.01,
"learning_rate": 4.929582609117229e-05,
"loss": 4.398,
"step": 43008
},
{
"epoch": 0.01,
"learning_rate": 4.928744014366177e-05,
"loss": 4.4032,
"step": 43520
},
{
"epoch": 0.01,
"learning_rate": 4.927905419615125e-05,
"loss": 4.4029,
"step": 44032
},
{
"epoch": 0.01,
"learning_rate": 4.927066824864073e-05,
"loss": 4.3942,
"step": 44544
},
{
"epoch": 0.01,
"learning_rate": 4.926229867993394e-05,
"loss": 4.3879,
"step": 45056
},
{
"epoch": 0.01,
"learning_rate": 4.925392911122715e-05,
"loss": 4.386,
"step": 45568
},
{
"epoch": 0.02,
"learning_rate": 4.924554316371663e-05,
"loss": 4.388,
"step": 46080
},
{
"epoch": 0.02,
"learning_rate": 4.923715721620611e-05,
"loss": 4.3754,
"step": 46592
},
{
"epoch": 0.02,
"learning_rate": 4.922877126869559e-05,
"loss": 4.3768,
"step": 47104
},
{
"epoch": 0.02,
"learning_rate": 4.922038532118507e-05,
"loss": 4.3698,
"step": 47616
},
{
"epoch": 0.02,
"learning_rate": 4.921199937367455e-05,
"loss": 4.3743,
"step": 48128
},
{
"epoch": 0.02,
"learning_rate": 4.920361342616403e-05,
"loss": 4.3679,
"step": 48640
},
{
"epoch": 0.02,
"learning_rate": 4.919522747865351e-05,
"loss": 4.3635,
"step": 49152
},
{
"epoch": 0.02,
"learning_rate": 4.918685790994672e-05,
"loss": 4.3462,
"step": 49664
},
{
"epoch": 0.02,
"learning_rate": 4.91784719624362e-05,
"loss": 4.3507,
"step": 50176
},
{
"epoch": 0.02,
"learning_rate": 4.9170086014925676e-05,
"loss": 4.3537,
"step": 50688
},
{
"epoch": 0.02,
"learning_rate": 4.9161700067415156e-05,
"loss": 4.3606,
"step": 51200
},
{
"epoch": 0.02,
"learning_rate": 4.915333049870837e-05,
"loss": 4.3463,
"step": 51712
},
{
"epoch": 0.02,
"learning_rate": 4.914496093000158e-05,
"loss": 4.3371,
"step": 52224
},
{
"epoch": 0.02,
"learning_rate": 4.913657498249106e-05,
"loss": 4.3322,
"step": 52736
},
{
"epoch": 0.02,
"learning_rate": 4.912818903498054e-05,
"loss": 4.3405,
"step": 53248
},
{
"epoch": 0.02,
"learning_rate": 4.911980308747002e-05,
"loss": 4.3093,
"step": 53760
},
{
"epoch": 0.02,
"learning_rate": 4.911143351876323e-05,
"loss": 4.3196,
"step": 54272
},
{
"epoch": 0.02,
"learning_rate": 4.910304757125272e-05,
"loss": 4.3223,
"step": 54784
},
{
"epoch": 0.02,
"learning_rate": 4.90946616237422e-05,
"loss": 4.3215,
"step": 55296
},
{
"epoch": 0.02,
"learning_rate": 4.908627567623168e-05,
"loss": 4.3127,
"step": 55808
},
{
"epoch": 0.02,
"learning_rate": 4.907788972872115e-05,
"loss": 4.3199,
"step": 56320
},
{
"epoch": 0.02,
"learning_rate": 4.906950378121063e-05,
"loss": 4.301,
"step": 56832
},
{
"epoch": 0.02,
"learning_rate": 4.906111783370011e-05,
"loss": 4.3264,
"step": 57344
},
{
"epoch": 0.02,
"learning_rate": 4.9052748264993326e-05,
"loss": 4.3061,
"step": 57856
},
{
"epoch": 0.02,
"learning_rate": 4.90443623174828e-05,
"loss": 4.3094,
"step": 58368
},
{
"epoch": 0.02,
"learning_rate": 4.903597636997228e-05,
"loss": 4.3134,
"step": 58880
},
{
"epoch": 0.02,
"learning_rate": 4.902759042246176e-05,
"loss": 4.3225,
"step": 59392
},
{
"epoch": 0.02,
"learning_rate": 4.9019220853754975e-05,
"loss": 4.2985,
"step": 59904
},
{
"epoch": 0.02,
"learning_rate": 4.9010834906244455e-05,
"loss": 4.2929,
"step": 60416
},
{
"epoch": 0.02,
"learning_rate": 4.9002448958733935e-05,
"loss": 4.2985,
"step": 60928
},
{
"epoch": 0.02,
"learning_rate": 4.8994063011223415e-05,
"loss": 4.2883,
"step": 61440
},
{
"epoch": 0.02,
"learning_rate": 4.8985677063712895e-05,
"loss": 4.294,
"step": 61952
},
{
"epoch": 0.02,
"learning_rate": 4.8977291116202375e-05,
"loss": 4.2895,
"step": 62464
},
{
"epoch": 0.02,
"learning_rate": 4.8968905168691855e-05,
"loss": 4.2783,
"step": 62976
},
{
"epoch": 0.02,
"learning_rate": 4.8960535599985064e-05,
"loss": 4.2954,
"step": 63488
},
{
"epoch": 0.02,
"learning_rate": 4.8952149652474544e-05,
"loss": 4.2775,
"step": 64000
},
{
"epoch": 0.02,
"learning_rate": 4.8943763704964024e-05,
"loss": 4.2749,
"step": 64512
},
{
"epoch": 0.02,
"learning_rate": 4.8935377757453504e-05,
"loss": 4.2722,
"step": 65024
},
{
"epoch": 0.02,
"learning_rate": 4.892700818874671e-05,
"loss": 4.2758,
"step": 65536
},
{
"epoch": 0.02,
"learning_rate": 4.891862224123619e-05,
"loss": 4.2638,
"step": 66048
},
{
"epoch": 0.02,
"learning_rate": 4.891023629372567e-05,
"loss": 4.2809,
"step": 66560
},
{
"epoch": 0.02,
"learning_rate": 4.890185034621515e-05,
"loss": 4.2717,
"step": 67072
},
{
"epoch": 0.02,
"learning_rate": 4.889346439870464e-05,
"loss": 4.26,
"step": 67584
},
{
"epoch": 0.02,
"learning_rate": 4.888507845119412e-05,
"loss": 4.2531,
"step": 68096
},
{
"epoch": 0.02,
"learning_rate": 4.887670888248733e-05,
"loss": 4.2636,
"step": 68608
},
{
"epoch": 0.02,
"learning_rate": 4.886832293497681e-05,
"loss": 4.2597,
"step": 69120
},
{
"epoch": 0.02,
"learning_rate": 4.885993698746629e-05,
"loss": 4.2631,
"step": 69632
},
{
"epoch": 0.02,
"learning_rate": 4.885155103995577e-05,
"loss": 4.2578,
"step": 70144
},
{
"epoch": 0.02,
"learning_rate": 4.884316509244525e-05,
"loss": 4.26,
"step": 70656
},
{
"epoch": 0.02,
"learning_rate": 4.883479552373846e-05,
"loss": 4.2572,
"step": 71168
},
{
"epoch": 0.02,
"learning_rate": 4.882640957622794e-05,
"loss": 4.2506,
"step": 71680
},
{
"epoch": 0.02,
"learning_rate": 4.881802362871742e-05,
"loss": 4.2419,
"step": 72192
},
{
"epoch": 0.02,
"learning_rate": 4.88096376812069e-05,
"loss": 4.2401,
"step": 72704
},
{
"epoch": 0.02,
"learning_rate": 4.880125173369638e-05,
"loss": 4.2462,
"step": 73216
},
{
"epoch": 0.02,
"learning_rate": 4.8792882164989586e-05,
"loss": 4.2255,
"step": 73728
},
{
"epoch": 0.02,
"learning_rate": 4.878449621747907e-05,
"loss": 4.247,
"step": 74240
},
{
"epoch": 0.02,
"learning_rate": 4.877611026996855e-05,
"loss": 4.2425,
"step": 74752
},
{
"epoch": 0.02,
"learning_rate": 4.876772432245803e-05,
"loss": 4.2343,
"step": 75264
},
{
"epoch": 0.02,
"learning_rate": 4.875933837494751e-05,
"loss": 4.236,
"step": 75776
},
{
"epoch": 0.02,
"learning_rate": 4.875096880624072e-05,
"loss": 4.2355,
"step": 76288
},
{
"epoch": 0.03,
"eval_loss": 4.195679664611816,
"eval_runtime": 306.1733,
"eval_samples_per_second": 1246.324,
"eval_steps_per_second": 38.949,
"step": 76320
},
{
"epoch": 1.0,
"learning_rate": 4.87425828587302e-05,
"loss": 4.2099,
"step": 76800
},
{
"epoch": 1.0,
"learning_rate": 4.873419691121968e-05,
"loss": 4.2183,
"step": 77312
},
{
"epoch": 1.0,
"learning_rate": 4.872581096370916e-05,
"loss": 4.2299,
"step": 77824
},
{
"epoch": 1.0,
"learning_rate": 4.871744139500237e-05,
"loss": 4.2151,
"step": 78336
},
{
"epoch": 1.0,
"learning_rate": 4.870905544749185e-05,
"loss": 4.2204,
"step": 78848
},
{
"epoch": 1.0,
"learning_rate": 4.870066949998133e-05,
"loss": 4.2106,
"step": 79360
},
{
"epoch": 1.0,
"learning_rate": 4.869228355247081e-05,
"loss": 4.2069,
"step": 79872
},
{
"epoch": 1.0,
"learning_rate": 4.868389760496029e-05,
"loss": 4.2028,
"step": 80384
},
{
"epoch": 1.0,
"learning_rate": 4.867551165744977e-05,
"loss": 4.2094,
"step": 80896
},
{
"epoch": 1.0,
"learning_rate": 4.8667142088742986e-05,
"loss": 4.1963,
"step": 81408
},
{
"epoch": 1.0,
"learning_rate": 4.865875614123246e-05,
"loss": 4.2179,
"step": 81920
},
{
"epoch": 1.0,
"learning_rate": 4.865037019372194e-05,
"loss": 4.2123,
"step": 82432
},
{
"epoch": 1.0,
"learning_rate": 4.8642000625015155e-05,
"loss": 4.1944,
"step": 82944
},
{
"epoch": 1.0,
"learning_rate": 4.8633614677504635e-05,
"loss": 4.1901,
"step": 83456
},
{
"epoch": 1.0,
"learning_rate": 4.862522872999411e-05,
"loss": 4.188,
"step": 83968
},
{
"epoch": 1.0,
"learning_rate": 4.861684278248359e-05,
"loss": 4.1877,
"step": 84480
},
{
"epoch": 1.0,
"learning_rate": 4.860845683497307e-05,
"loss": 4.184,
"step": 84992
},
{
"epoch": 1.0,
"learning_rate": 4.860007088746255e-05,
"loss": 4.1855,
"step": 85504
},
{
"epoch": 1.0,
"learning_rate": 4.859168493995203e-05,
"loss": 4.1782,
"step": 86016
},
{
"epoch": 1.0,
"learning_rate": 4.858329899244151e-05,
"loss": 4.1954,
"step": 86528
},
{
"epoch": 1.0,
"learning_rate": 4.8574929423734724e-05,
"loss": 4.1831,
"step": 87040
},
{
"epoch": 1.0,
"learning_rate": 4.8566543476224204e-05,
"loss": 4.1864,
"step": 87552
},
{
"epoch": 1.0,
"learning_rate": 4.8558157528713684e-05,
"loss": 4.1825,
"step": 88064
},
{
"epoch": 1.0,
"learning_rate": 4.8549771581203164e-05,
"loss": 4.1921,
"step": 88576
},
{
"epoch": 1.0,
"learning_rate": 4.854140201249637e-05,
"loss": 4.168,
"step": 89088
},
{
"epoch": 1.0,
"learning_rate": 4.853301606498585e-05,
"loss": 4.1713,
"step": 89600
},
{
"epoch": 1.0,
"learning_rate": 4.852463011747533e-05,
"loss": 4.1667,
"step": 90112
},
{
"epoch": 1.0,
"learning_rate": 4.851624416996481e-05,
"loss": 4.1644,
"step": 90624
},
{
"epoch": 1.0,
"learning_rate": 4.850785822245429e-05,
"loss": 4.1661,
"step": 91136
},
{
"epoch": 1.01,
"learning_rate": 4.84994886537475e-05,
"loss": 4.1617,
"step": 91648
},
{
"epoch": 1.01,
"learning_rate": 4.849111908504072e-05,
"loss": 4.1652,
"step": 92160
},
{
"epoch": 1.01,
"learning_rate": 4.84827331375302e-05,
"loss": 4.167,
"step": 92672
},
{
"epoch": 1.01,
"learning_rate": 4.847434719001968e-05,
"loss": 4.1554,
"step": 93184
},
{
"epoch": 1.01,
"learning_rate": 4.846596124250916e-05,
"loss": 4.1636,
"step": 93696
},
{
"epoch": 1.01,
"learning_rate": 4.845757529499864e-05,
"loss": 4.1702,
"step": 94208
},
{
"epoch": 1.01,
"learning_rate": 4.844918934748812e-05,
"loss": 4.1568,
"step": 94720
},
{
"epoch": 1.01,
"learning_rate": 4.84408033999776e-05,
"loss": 4.14,
"step": 95232
},
{
"epoch": 1.01,
"learning_rate": 4.843241745246708e-05,
"loss": 4.1541,
"step": 95744
},
{
"epoch": 1.01,
"learning_rate": 4.842404788376029e-05,
"loss": 4.1364,
"step": 96256
},
{
"epoch": 1.01,
"learning_rate": 4.841566193624977e-05,
"loss": 4.1425,
"step": 96768
},
{
"epoch": 1.01,
"learning_rate": 4.840727598873925e-05,
"loss": 4.1426,
"step": 97280
},
{
"epoch": 1.01,
"learning_rate": 4.839889004122873e-05,
"loss": 4.1511,
"step": 97792
},
{
"epoch": 1.01,
"learning_rate": 4.839050409371821e-05,
"loss": 4.1387,
"step": 98304
},
{
"epoch": 1.01,
"learning_rate": 4.8382134525011416e-05,
"loss": 4.144,
"step": 98816
},
{
"epoch": 1.01,
"learning_rate": 4.83737485775009e-05,
"loss": 4.1485,
"step": 99328
},
{
"epoch": 1.01,
"learning_rate": 4.836536262999038e-05,
"loss": 4.1401,
"step": 99840
},
{
"epoch": 1.01,
"learning_rate": 4.835697668247986e-05,
"loss": 4.1413,
"step": 100352
},
{
"epoch": 1.01,
"learning_rate": 4.834860711377307e-05,
"loss": 4.1238,
"step": 100864
},
{
"epoch": 1.01,
"learning_rate": 4.834022116626255e-05,
"loss": 4.1316,
"step": 101376
},
{
"epoch": 1.01,
"learning_rate": 4.833183521875203e-05,
"loss": 4.138,
"step": 101888
},
{
"epoch": 1.01,
"learning_rate": 4.832344927124151e-05,
"loss": 4.1298,
"step": 102400
},
{
"epoch": 1.01,
"learning_rate": 4.831507970253472e-05,
"loss": 4.1258,
"step": 102912
},
{
"epoch": 1.01,
"learning_rate": 4.83066937550242e-05,
"loss": 4.1113,
"step": 103424
},
{
"epoch": 1.01,
"learning_rate": 4.829830780751368e-05,
"loss": 4.1231,
"step": 103936
},
{
"epoch": 1.01,
"learning_rate": 4.828992186000316e-05,
"loss": 4.1128,
"step": 104448
},
{
"epoch": 1.01,
"learning_rate": 4.828155229129637e-05,
"loss": 4.1277,
"step": 104960
},
{
"epoch": 1.01,
"learning_rate": 4.8273166343785856e-05,
"loss": 4.1136,
"step": 105472
},
{
"epoch": 1.01,
"learning_rate": 4.8264780396275336e-05,
"loss": 4.1221,
"step": 105984
},
{
"epoch": 1.01,
"learning_rate": 4.8256394448764816e-05,
"loss": 4.1235,
"step": 106496
},
{
"epoch": 1.01,
"learning_rate": 4.8248008501254296e-05,
"loss": 4.0997,
"step": 107008
},
{
"epoch": 1.01,
"learning_rate": 4.8239638932547505e-05,
"loss": 4.1141,
"step": 107520
},
{
"epoch": 1.01,
"learning_rate": 4.8231252985036985e-05,
"loss": 4.111,
"step": 108032
},
{
"epoch": 1.01,
"learning_rate": 4.8222867037526465e-05,
"loss": 4.1008,
"step": 108544
},
{
"epoch": 1.01,
"learning_rate": 4.8214481090015945e-05,
"loss": 4.1014,
"step": 109056
},
{
"epoch": 1.01,
"learning_rate": 4.8206111521309154e-05,
"loss": 4.1191,
"step": 109568
},
{
"epoch": 1.01,
"learning_rate": 4.8197725573798634e-05,
"loss": 4.1066,
"step": 110080
},
{
"epoch": 1.01,
"learning_rate": 4.8189339626288114e-05,
"loss": 4.0947,
"step": 110592
},
{
"epoch": 1.01,
"learning_rate": 4.8180953678777594e-05,
"loss": 4.0892,
"step": 111104
},
{
"epoch": 1.01,
"learning_rate": 4.817258411007081e-05,
"loss": 4.0982,
"step": 111616
},
{
"epoch": 1.01,
"learning_rate": 4.816419816256029e-05,
"loss": 4.11,
"step": 112128
},
{
"epoch": 1.01,
"learning_rate": 4.815581221504977e-05,
"loss": 4.1146,
"step": 112640
},
{
"epoch": 1.01,
"learning_rate": 4.814742626753924e-05,
"loss": 4.0979,
"step": 113152
},
{
"epoch": 1.01,
"learning_rate": 4.813905669883246e-05,
"loss": 4.1118,
"step": 113664
},
{
"epoch": 1.01,
"learning_rate": 4.813067075132194e-05,
"loss": 4.1066,
"step": 114176
},
{
"epoch": 1.01,
"learning_rate": 4.812228480381141e-05,
"loss": 4.1017,
"step": 114688
},
{
"epoch": 1.01,
"learning_rate": 4.811389885630089e-05,
"loss": 4.0869,
"step": 115200
},
{
"epoch": 1.01,
"learning_rate": 4.810552928759411e-05,
"loss": 4.1021,
"step": 115712
},
{
"epoch": 1.01,
"learning_rate": 4.809714334008359e-05,
"loss": 4.0914,
"step": 116224
},
{
"epoch": 1.01,
"learning_rate": 4.808875739257306e-05,
"loss": 4.0878,
"step": 116736
},
{
"epoch": 1.01,
"learning_rate": 4.808037144506255e-05,
"loss": 4.0904,
"step": 117248
},
{
"epoch": 1.01,
"learning_rate": 4.807200187635576e-05,
"loss": 4.0867,
"step": 117760
},
{
"epoch": 1.01,
"learning_rate": 4.8063615928845243e-05,
"loss": 4.1003,
"step": 118272
},
{
"epoch": 1.01,
"learning_rate": 4.8055229981334717e-05,
"loss": 4.0886,
"step": 118784
},
{
"epoch": 1.01,
"learning_rate": 4.8046844033824197e-05,
"loss": 4.082,
"step": 119296
},
{
"epoch": 1.01,
"learning_rate": 4.803847446511741e-05,
"loss": 4.0839,
"step": 119808
},
{
"epoch": 1.01,
"learning_rate": 4.8030088517606886e-05,
"loss": 4.0873,
"step": 120320
},
{
"epoch": 1.01,
"learning_rate": 4.8021702570096366e-05,
"loss": 4.0813,
"step": 120832
},
{
"epoch": 1.01,
"learning_rate": 4.8013316622585846e-05,
"loss": 4.0828,
"step": 121344
},
{
"epoch": 1.01,
"learning_rate": 4.800494705387906e-05,
"loss": 4.0737,
"step": 121856
},
{
"epoch": 1.02,
"learning_rate": 4.7996561106368535e-05,
"loss": 4.0888,
"step": 122368
},
{
"epoch": 1.02,
"learning_rate": 4.7988175158858015e-05,
"loss": 4.0726,
"step": 122880
},
{
"epoch": 1.02,
"learning_rate": 4.7979789211347494e-05,
"loss": 4.075,
"step": 123392
},
{
"epoch": 1.02,
"learning_rate": 4.797140326383698e-05,
"loss": 4.0692,
"step": 123904
},
{
"epoch": 1.02,
"learning_rate": 4.796301731632646e-05,
"loss": 4.0834,
"step": 124416
},
{
"epoch": 1.02,
"learning_rate": 4.795463136881594e-05,
"loss": 4.0807,
"step": 124928
},
{
"epoch": 1.02,
"learning_rate": 4.794624542130542e-05,
"loss": 4.0738,
"step": 125440
},
{
"epoch": 1.02,
"learning_rate": 4.793787585259863e-05,
"loss": 4.0605,
"step": 125952
},
{
"epoch": 1.02,
"learning_rate": 4.792948990508811e-05,
"loss": 4.0691,
"step": 126464
},
{
"epoch": 1.02,
"learning_rate": 4.792110395757759e-05,
"loss": 4.0711,
"step": 126976
},
{
"epoch": 1.02,
"learning_rate": 4.791271801006707e-05,
"loss": 4.0783,
"step": 127488
},
{
"epoch": 1.02,
"learning_rate": 4.790434844136028e-05,
"loss": 4.0715,
"step": 128000
},
{
"epoch": 1.02,
"learning_rate": 4.789596249384976e-05,
"loss": 4.0619,
"step": 128512
},
{
"epoch": 1.02,
"learning_rate": 4.788757654633924e-05,
"loss": 4.0563,
"step": 129024
},
{
"epoch": 1.02,
"learning_rate": 4.787920697763245e-05,
"loss": 4.069,
"step": 129536
},
{
"epoch": 1.02,
"learning_rate": 4.7870821030121935e-05,
"loss": 4.0394,
"step": 130048
},
{
"epoch": 1.02,
"learning_rate": 4.7862451461415144e-05,
"loss": 4.0555,
"step": 130560
},
{
"epoch": 1.02,
"learning_rate": 4.7854065513904624e-05,
"loss": 4.058,
"step": 131072
},
{
"epoch": 1.02,
"learning_rate": 4.7845679566394104e-05,
"loss": 4.0572,
"step": 131584
},
{
"epoch": 1.02,
"learning_rate": 4.7837293618883584e-05,
"loss": 4.0478,
"step": 132096
},
{
"epoch": 1.02,
"learning_rate": 4.7828907671373064e-05,
"loss": 4.0576,
"step": 132608
},
{
"epoch": 1.02,
"learning_rate": 4.7820521723862544e-05,
"loss": 4.0495,
"step": 133120
},
{
"epoch": 1.02,
"learning_rate": 4.7812135776352024e-05,
"loss": 4.0667,
"step": 133632
},
{
"epoch": 1.02,
"learning_rate": 4.7803749828841504e-05,
"loss": 4.0516,
"step": 134144
},
{
"epoch": 1.02,
"learning_rate": 4.779538026013471e-05,
"loss": 4.0521,
"step": 134656
},
{
"epoch": 1.02,
"learning_rate": 4.778699431262419e-05,
"loss": 4.0681,
"step": 135168
},
{
"epoch": 1.02,
"learning_rate": 4.777860836511367e-05,
"loss": 4.0724,
"step": 135680
},
{
"epoch": 1.02,
"learning_rate": 4.777023879640689e-05,
"loss": 4.05,
"step": 136192
},
{
"epoch": 1.02,
"learning_rate": 4.776185284889637e-05,
"loss": 4.0476,
"step": 136704
},
{
"epoch": 1.02,
"learning_rate": 4.775346690138585e-05,
"loss": 4.0551,
"step": 137216
},
{
"epoch": 1.02,
"learning_rate": 4.774508095387533e-05,
"loss": 4.0442,
"step": 137728
},
{
"epoch": 1.02,
"learning_rate": 4.773669500636481e-05,
"loss": 4.0478,
"step": 138240
},
{
"epoch": 1.02,
"learning_rate": 4.772830905885429e-05,
"loss": 4.0527,
"step": 138752
},
{
"epoch": 1.02,
"learning_rate": 4.771992311134377e-05,
"loss": 4.0385,
"step": 139264
},
{
"epoch": 1.02,
"learning_rate": 4.771155354263698e-05,
"loss": 4.0584,
"step": 139776
},
{
"epoch": 1.02,
"learning_rate": 4.770316759512646e-05,
"loss": 4.0412,
"step": 140288
},
{
"epoch": 1.02,
"learning_rate": 4.769478164761594e-05,
"loss": 4.045,
"step": 140800
},
{
"epoch": 1.02,
"learning_rate": 4.768639570010542e-05,
"loss": 4.0367,
"step": 141312
},
{
"epoch": 1.02,
"learning_rate": 4.7678026131398626e-05,
"loss": 4.0429,
"step": 141824
},
{
"epoch": 1.02,
"learning_rate": 4.7669640183888106e-05,
"loss": 4.0373,
"step": 142336
},
{
"epoch": 1.02,
"learning_rate": 4.7661254236377586e-05,
"loss": 4.0515,
"step": 142848
},
{
"epoch": 1.02,
"learning_rate": 4.765286828886707e-05,
"loss": 4.0472,
"step": 143360
},
{
"epoch": 1.02,
"learning_rate": 4.764449872016028e-05,
"loss": 4.0362,
"step": 143872
},
{
"epoch": 1.02,
"learning_rate": 4.763611277264976e-05,
"loss": 4.0216,
"step": 144384
},
{
"epoch": 1.02,
"learning_rate": 4.762772682513924e-05,
"loss": 4.0452,
"step": 144896
},
{
"epoch": 1.02,
"learning_rate": 4.761934087762872e-05,
"loss": 4.036,
"step": 145408
},
{
"epoch": 1.02,
"learning_rate": 4.761097130892193e-05,
"loss": 4.0435,
"step": 145920
},
{
"epoch": 1.02,
"learning_rate": 4.760258536141141e-05,
"loss": 4.0388,
"step": 146432
},
{
"epoch": 1.02,
"learning_rate": 4.759419941390089e-05,
"loss": 4.0382,
"step": 146944
},
{
"epoch": 1.02,
"learning_rate": 4.758581346639037e-05,
"loss": 4.0391,
"step": 147456
},
{
"epoch": 1.02,
"learning_rate": 4.757744389768358e-05,
"loss": 4.0408,
"step": 147968
},
{
"epoch": 1.02,
"learning_rate": 4.756905795017306e-05,
"loss": 4.0272,
"step": 148480
},
{
"epoch": 1.02,
"learning_rate": 4.756067200266254e-05,
"loss": 4.0257,
"step": 148992
},
{
"epoch": 1.02,
"learning_rate": 4.755228605515202e-05,
"loss": 4.0354,
"step": 149504
},
{
"epoch": 1.02,
"learning_rate": 4.7543916486445236e-05,
"loss": 4.0181,
"step": 150016
},
{
"epoch": 1.02,
"learning_rate": 4.7535530538934716e-05,
"loss": 4.0338,
"step": 150528
},
{
"epoch": 1.02,
"learning_rate": 4.7527144591424196e-05,
"loss": 4.0401,
"step": 151040
},
{
"epoch": 1.02,
"learning_rate": 4.751875864391367e-05,
"loss": 4.0274,
"step": 151552
},
{
"epoch": 1.02,
"learning_rate": 4.7510389075206885e-05,
"loss": 4.025,
"step": 152064
},
{
"epoch": 1.02,
"learning_rate": 4.7502003127696365e-05,
"loss": 4.0283,
"step": 152576
},
{
"epoch": 1.03,
"eval_loss": 4.027114391326904,
"eval_runtime": 305.7714,
"eval_samples_per_second": 1247.962,
"eval_steps_per_second": 39.0,
"step": 152640
},
{
"epoch": 0.0,
"learning_rate": 4.7493617180185845e-05,
"loss": 4.0066,
"step": 153088
},
{
"epoch": 0.0,
"learning_rate": 4.748523123267532e-05,
"loss": 4.0157,
"step": 153600
},
{
"epoch": 0.0,
"learning_rate": 4.74768452851648e-05,
"loss": 4.0274,
"step": 154112
},
{
"epoch": 0.0,
"learning_rate": 4.746845933765428e-05,
"loss": 4.0189,
"step": 154624
},
{
"epoch": 0.0,
"learning_rate": 4.7460073390143764e-05,
"loss": 4.025,
"step": 155136
},
{
"epoch": 0.0,
"learning_rate": 4.7451687442633244e-05,
"loss": 4.0076,
"step": 155648
},
{
"epoch": 0.0,
"learning_rate": 4.7443301495122724e-05,
"loss": 4.0186,
"step": 156160
},
{
"epoch": 0.0,
"learning_rate": 4.7434915547612204e-05,
"loss": 4.0031,
"step": 156672
},
{
"epoch": 0.0,
"learning_rate": 4.7426529600101684e-05,
"loss": 4.0153,
"step": 157184
},
{
"epoch": 0.0,
"learning_rate": 4.7418143652591164e-05,
"loss": 4.0119,
"step": 157696
},
{
"epoch": 0.0,
"learning_rate": 4.740977408388437e-05,
"loss": 4.0205,
"step": 158208
},
{
"epoch": 0.0,
"learning_rate": 4.740138813637385e-05,
"loss": 4.0239,
"step": 158720
},
{
"epoch": 0.0,
"learning_rate": 4.739301856766706e-05,
"loss": 4.0048,
"step": 159232
},
{
"epoch": 0.0,
"learning_rate": 4.738463262015654e-05,
"loss": 4.0063,
"step": 159744
},
{
"epoch": 0.0,
"learning_rate": 4.737624667264602e-05,
"loss": 4.0051,
"step": 160256
},
{
"epoch": 0.0,
"learning_rate": 4.73678607251355e-05,
"loss": 4.0007,
"step": 160768
},
{
"epoch": 0.0,
"learning_rate": 4.735947477762498e-05,
"loss": 3.9988,
"step": 161280
},
{
"epoch": 0.0,
"learning_rate": 4.735108883011446e-05,
"loss": 4.0022,
"step": 161792
},
{
"epoch": 0.0,
"learning_rate": 4.734270288260394e-05,
"loss": 3.9963,
"step": 162304
},
{
"epoch": 0.0,
"learning_rate": 4.733431693509343e-05,
"loss": 4.0199,
"step": 162816
},
{
"epoch": 0.0,
"learning_rate": 4.732593098758291e-05,
"loss": 3.9983,
"step": 163328
},
{
"epoch": 0.0,
"learning_rate": 4.731754504007238e-05,
"loss": 4.013,
"step": 163840
},
{
"epoch": 0.0,
"learning_rate": 4.730915909256186e-05,
"loss": 4.0055,
"step": 164352
},
{
"epoch": 0.0,
"learning_rate": 4.730078952385508e-05,
"loss": 4.0122,
"step": 164864
},
{
"epoch": 0.0,
"learning_rate": 4.729240357634456e-05,
"loss": 3.9867,
"step": 165376
},
{
"epoch": 0.0,
"learning_rate": 4.728401762883403e-05,
"loss": 3.9994,
"step": 165888
},
{
"epoch": 0.0,
"learning_rate": 4.727563168132351e-05,
"loss": 3.9882,
"step": 166400
},
{
"epoch": 0.0,
"learning_rate": 4.726726211261673e-05,
"loss": 3.9957,
"step": 166912
},
{
"epoch": 0.0,
"learning_rate": 4.725887616510621e-05,
"loss": 3.9895,
"step": 167424
},
{
"epoch": 0.01,
"learning_rate": 4.725049021759568e-05,
"loss": 3.989,
"step": 167936
},
{
"epoch": 0.01,
"learning_rate": 4.7242120648888896e-05,
"loss": 4.0,
"step": 168448
},
{
"epoch": 0.01,
"learning_rate": 4.723373470137838e-05,
"loss": 3.9972,
"step": 168960
},
{
"epoch": 0.01,
"learning_rate": 4.7225348753867856e-05,
"loss": 3.9833,
"step": 169472
},
{
"epoch": 0.01,
"learning_rate": 4.7216962806357336e-05,
"loss": 3.9977,
"step": 169984
},
{
"epoch": 0.01,
"learning_rate": 4.7208576858846816e-05,
"loss": 3.9958,
"step": 170496
},
{
"epoch": 0.01,
"learning_rate": 4.7200190911336296e-05,
"loss": 3.9908,
"step": 171008
},
{
"epoch": 0.01,
"learning_rate": 4.7191804963825775e-05,
"loss": 3.9784,
"step": 171520
},
{
"epoch": 0.01,
"learning_rate": 4.7183419016315255e-05,
"loss": 3.989,
"step": 172032
},
{
"epoch": 0.01,
"learning_rate": 4.7175049447608465e-05,
"loss": 3.9701,
"step": 172544
},
{
"epoch": 0.01,
"learning_rate": 4.7166663500097945e-05,
"loss": 3.9844,
"step": 173056
},
{
"epoch": 0.01,
"learning_rate": 4.7158277552587424e-05,
"loss": 3.9794,
"step": 173568
},
{
"epoch": 0.01,
"learning_rate": 4.7149891605076904e-05,
"loss": 3.9896,
"step": 174080
},
{
"epoch": 0.01,
"learning_rate": 4.714152203637012e-05,
"loss": 3.9799,
"step": 174592
},
{
"epoch": 0.01,
"learning_rate": 4.71331360888596e-05,
"loss": 3.9836,
"step": 175104
},
{
"epoch": 0.01,
"learning_rate": 4.712476652015281e-05,
"loss": 3.9899,
"step": 175616
},
{
"epoch": 0.01,
"learning_rate": 4.711638057264229e-05,
"loss": 3.9846,
"step": 176128
},
{
"epoch": 0.01,
"learning_rate": 4.710799462513177e-05,
"loss": 3.9836,
"step": 176640
},
{
"epoch": 0.01,
"learning_rate": 4.709960867762125e-05,
"loss": 3.9665,
"step": 177152
},
{
"epoch": 0.01,
"learning_rate": 4.709122273011073e-05,
"loss": 3.9718,
"step": 177664
},
{
"epoch": 0.01,
"learning_rate": 4.708283678260021e-05,
"loss": 3.9847,
"step": 178176
},
{
"epoch": 0.01,
"learning_rate": 4.707445083508969e-05,
"loss": 3.9742,
"step": 178688
},
{
"epoch": 0.01,
"learning_rate": 4.706606488757917e-05,
"loss": 3.9691,
"step": 179200
},
{
"epoch": 0.01,
"learning_rate": 4.705769531887238e-05,
"loss": 3.9611,
"step": 179712
},
{
"epoch": 0.01,
"learning_rate": 4.704930937136186e-05,
"loss": 3.9743,
"step": 180224
},
{
"epoch": 0.01,
"learning_rate": 4.704092342385134e-05,
"loss": 3.9576,
"step": 180736
},
{
"epoch": 0.01,
"learning_rate": 4.703253747634082e-05,
"loss": 3.9771,
"step": 181248
},
{
"epoch": 0.01,
"learning_rate": 4.7024167907634034e-05,
"loss": 3.9641,
"step": 181760
},
{
"epoch": 0.01,
"learning_rate": 4.7015781960123514e-05,
"loss": 3.9754,
"step": 182272
},
{
"epoch": 0.01,
"learning_rate": 4.7007396012612994e-05,
"loss": 3.9722,
"step": 182784
},
{
"epoch": 0.01,
"learning_rate": 4.6999010065102474e-05,
"loss": 3.9547,
"step": 183296
},
{
"epoch": 0.01,
"learning_rate": 4.699064049639568e-05,
"loss": 3.9638,
"step": 183808
},
{
"epoch": 0.01,
"learning_rate": 4.698225454888516e-05,
"loss": 3.9653,
"step": 184320
},
{
"epoch": 0.01,
"learning_rate": 4.697386860137464e-05,
"loss": 3.9565,
"step": 184832
},
{
"epoch": 0.01,
"learning_rate": 4.696548265386412e-05,
"loss": 3.9618,
"step": 185344
},
{
"epoch": 0.01,
"learning_rate": 4.695711308515733e-05,
"loss": 3.9693,
"step": 185856
},
{
"epoch": 0.01,
"learning_rate": 4.694872713764681e-05,
"loss": 3.9652,
"step": 186368
},
{
"epoch": 0.01,
"learning_rate": 4.694034119013629e-05,
"loss": 3.9542,
"step": 186880
},
{
"epoch": 0.01,
"learning_rate": 4.693195524262577e-05,
"loss": 3.9449,
"step": 187392
},
{
"epoch": 0.01,
"learning_rate": 4.692358567391899e-05,
"loss": 3.9542,
"step": 187904
},
{
"epoch": 0.01,
"learning_rate": 4.691519972640847e-05,
"loss": 3.9666,
"step": 188416
},
{
"epoch": 0.01,
"learning_rate": 4.690681377889795e-05,
"loss": 3.9735,
"step": 188928
},
{
"epoch": 0.01,
"learning_rate": 4.6898444210191157e-05,
"loss": 3.9576,
"step": 189440
},
{
"epoch": 0.01,
"learning_rate": 4.6890074641484366e-05,
"loss": 3.9701,
"step": 189952
},
{
"epoch": 0.01,
"learning_rate": 4.6881688693973846e-05,
"loss": 3.968,
"step": 190464
},
{
"epoch": 0.01,
"learning_rate": 4.6873302746463326e-05,
"loss": 3.9671,
"step": 190976
},
{
"epoch": 0.01,
"learning_rate": 4.6864916798952806e-05,
"loss": 3.9428,
"step": 191488
},
{
"epoch": 0.01,
"learning_rate": 4.6856530851442285e-05,
"loss": 3.967,
"step": 192000
},
{
"epoch": 0.01,
"learning_rate": 4.6848144903931765e-05,
"loss": 3.9557,
"step": 192512
},
{
"epoch": 0.01,
"learning_rate": 4.6839758956421245e-05,
"loss": 3.9491,
"step": 193024
},
{
"epoch": 0.01,
"learning_rate": 4.683138938771446e-05,
"loss": 3.9573,
"step": 193536
},
{
"epoch": 0.01,
"learning_rate": 4.682300344020394e-05,
"loss": 3.951,
"step": 194048
},
{
"epoch": 0.01,
"learning_rate": 4.681461749269342e-05,
"loss": 3.9615,
"step": 194560
},
{
"epoch": 0.01,
"learning_rate": 4.68062315451829e-05,
"loss": 3.9561,
"step": 195072
},
{
"epoch": 0.01,
"learning_rate": 4.679784559767238e-05,
"loss": 3.9447,
"step": 195584
},
{
"epoch": 0.01,
"learning_rate": 4.678947602896559e-05,
"loss": 3.9512,
"step": 196096
},
{
"epoch": 0.01,
"learning_rate": 4.678109008145507e-05,
"loss": 3.9518,
"step": 196608
},
{
"epoch": 0.01,
"learning_rate": 4.677270413394455e-05,
"loss": 3.9501,
"step": 197120
},
{
"epoch": 0.01,
"learning_rate": 4.676431818643403e-05,
"loss": 3.9507,
"step": 197632
},
{
"epoch": 0.01,
"learning_rate": 4.675593223892351e-05,
"loss": 3.9444,
"step": 198144
},
{
"epoch": 0.02,
"learning_rate": 4.674756267021672e-05,
"loss": 3.9541,
"step": 198656
},
{
"epoch": 0.02,
"learning_rate": 4.67391767227062e-05,
"loss": 3.9436,
"step": 199168
},
{
"epoch": 0.02,
"learning_rate": 4.673079077519568e-05,
"loss": 3.9457,
"step": 199680
},
{
"epoch": 0.02,
"learning_rate": 4.672240482768516e-05,
"loss": 3.9401,
"step": 200192
},
{
"epoch": 0.02,
"learning_rate": 4.671401888017464e-05,
"loss": 3.9563,
"step": 200704
},
{
"epoch": 0.02,
"learning_rate": 4.670563293266412e-05,
"loss": 3.9513,
"step": 201216
},
{
"epoch": 0.02,
"learning_rate": 4.66972469851536e-05,
"loss": 3.9463,
"step": 201728
},
{
"epoch": 0.02,
"learning_rate": 4.6688877416446815e-05,
"loss": 3.9374,
"step": 202240
},
{
"epoch": 0.02,
"learning_rate": 4.668049146893629e-05,
"loss": 3.9461,
"step": 202752
},
{
"epoch": 0.02,
"learning_rate": 4.667210552142577e-05,
"loss": 3.9382,
"step": 203264
},
{
"epoch": 0.02,
"learning_rate": 4.666371957391525e-05,
"loss": 3.956,
"step": 203776
},
{
"epoch": 0.02,
"learning_rate": 4.665533362640473e-05,
"loss": 3.9429,
"step": 204288
},
{
"epoch": 0.02,
"learning_rate": 4.664696405769794e-05,
"loss": 3.9382,
"step": 204800
},
{
"epoch": 0.02,
"learning_rate": 4.663857811018742e-05,
"loss": 3.9318,
"step": 205312
},
{
"epoch": 0.02,
"learning_rate": 4.66301921626769e-05,
"loss": 3.9422,
"step": 205824
},
{
"epoch": 0.02,
"learning_rate": 4.6621806215166383e-05,
"loss": 3.9198,
"step": 206336
},
{
"epoch": 0.02,
"learning_rate": 4.6613420267655863e-05,
"loss": 3.9332,
"step": 206848
},
{
"epoch": 0.02,
"learning_rate": 4.660505069894907e-05,
"loss": 3.9349,
"step": 207360
},
{
"epoch": 0.02,
"learning_rate": 4.659666475143855e-05,
"loss": 3.9359,
"step": 207872
},
{
"epoch": 0.02,
"learning_rate": 4.658827880392803e-05,
"loss": 3.9236,
"step": 208384
},
{
"epoch": 0.02,
"learning_rate": 4.657989285641751e-05,
"loss": 3.9406,
"step": 208896
},
{
"epoch": 0.02,
"learning_rate": 4.657150690890699e-05,
"loss": 3.9254,
"step": 209408
},
{
"epoch": 0.02,
"learning_rate": 4.656312096139647e-05,
"loss": 3.9464,
"step": 209920
},
{
"epoch": 0.02,
"learning_rate": 4.655473501388595e-05,
"loss": 3.9331,
"step": 210432
},
{
"epoch": 0.02,
"learning_rate": 4.654636544517916e-05,
"loss": 3.9315,
"step": 210944
},
{
"epoch": 0.02,
"learning_rate": 4.653797949766864e-05,
"loss": 3.9478,
"step": 211456
},
{
"epoch": 0.02,
"learning_rate": 4.652959355015812e-05,
"loss": 3.9541,
"step": 211968
},
{
"epoch": 0.02,
"learning_rate": 4.652122398145134e-05,
"loss": 3.9336,
"step": 212480
},
{
"epoch": 0.02,
"learning_rate": 4.651283803394082e-05,
"loss": 3.9267,
"step": 212992
},
{
"epoch": 0.02,
"learning_rate": 4.65044520864303e-05,
"loss": 3.9343,
"step": 213504
},
{
"epoch": 0.02,
"learning_rate": 4.649606613891978e-05,
"loss": 3.9284,
"step": 214016
},
{
"epoch": 0.02,
"learning_rate": 4.648768019140926e-05,
"loss": 3.929,
"step": 214528
},
{
"epoch": 0.02,
"learning_rate": 4.647929424389874e-05,
"loss": 3.94,
"step": 215040
},
{
"epoch": 0.02,
"learning_rate": 4.647090829638822e-05,
"loss": 3.9199,
"step": 215552
},
{
"epoch": 0.02,
"learning_rate": 4.64625223488777e-05,
"loss": 3.9418,
"step": 216064
},
{
"epoch": 0.02,
"learning_rate": 4.6454152780170906e-05,
"loss": 3.9291,
"step": 216576
},
{
"epoch": 0.02,
"learning_rate": 4.6445766832660386e-05,
"loss": 3.9304,
"step": 217088
},
{
"epoch": 0.02,
"learning_rate": 4.6437380885149866e-05,
"loss": 3.9215,
"step": 217600
},
{
"epoch": 0.02,
"learning_rate": 4.6429011316443075e-05,
"loss": 3.9295,
"step": 218112
},
{
"epoch": 0.02,
"learning_rate": 4.6420625368932555e-05,
"loss": 3.9211,
"step": 218624
},
{
"epoch": 0.02,
"learning_rate": 4.6412239421422035e-05,
"loss": 3.9401,
"step": 219136
},
{
"epoch": 0.02,
"learning_rate": 4.640385347391152e-05,
"loss": 3.9307,
"step": 219648
},
{
"epoch": 0.02,
"learning_rate": 4.639548390520473e-05,
"loss": 3.9267,
"step": 220160
},
{
"epoch": 0.02,
"learning_rate": 4.638709795769421e-05,
"loss": 3.9092,
"step": 220672
},
{
"epoch": 0.02,
"learning_rate": 4.637871201018369e-05,
"loss": 3.9334,
"step": 221184
},
{
"epoch": 0.02,
"learning_rate": 4.637032606267317e-05,
"loss": 3.9268,
"step": 221696
},
{
"epoch": 0.02,
"learning_rate": 4.636194011516265e-05,
"loss": 3.9261,
"step": 222208
},
{
"epoch": 0.02,
"learning_rate": 4.6353554167652124e-05,
"loss": 3.9233,
"step": 222720
},
{
"epoch": 0.02,
"learning_rate": 4.6345168220141604e-05,
"loss": 3.9318,
"step": 223232
},
{
"epoch": 0.02,
"learning_rate": 4.633679865143482e-05,
"loss": 3.9291,
"step": 223744
},
{
"epoch": 0.02,
"learning_rate": 4.632841270392429e-05,
"loss": 3.9341,
"step": 224256
},
{
"epoch": 0.02,
"learning_rate": 4.632002675641377e-05,
"loss": 3.915,
"step": 224768
},
{
"epoch": 0.02,
"learning_rate": 4.631164080890326e-05,
"loss": 3.919,
"step": 225280
},
{
"epoch": 0.02,
"learning_rate": 4.630325486139274e-05,
"loss": 3.9251,
"step": 225792
},
{
"epoch": 0.02,
"learning_rate": 4.629488529268595e-05,
"loss": 3.9072,
"step": 226304
},
{
"epoch": 0.02,
"learning_rate": 4.628649934517543e-05,
"loss": 3.9237,
"step": 226816
},
{
"epoch": 0.02,
"learning_rate": 4.627811339766491e-05,
"loss": 3.9316,
"step": 227328
},
{
"epoch": 0.02,
"learning_rate": 4.626972745015439e-05,
"loss": 3.9194,
"step": 227840
},
{
"epoch": 0.02,
"learning_rate": 4.626134150264387e-05,
"loss": 3.9137,
"step": 228352
},
{
"epoch": 0.02,
"learning_rate": 4.625297193393708e-05,
"loss": 3.9226,
"step": 228864
},
{
"epoch": 0.03,
"eval_loss": 3.9522695541381836,
"eval_runtime": 305.3529,
"eval_samples_per_second": 1249.672,
"eval_steps_per_second": 39.053,
"step": 228960
},
{
"epoch": 1.0,
"learning_rate": 4.624458598642656e-05,
"loss": 3.9115,
"step": 229376
},
{
"epoch": 1.0,
"learning_rate": 4.623620003891604e-05,
"loss": 3.9093,
"step": 229888
},
{
"epoch": 1.0,
"learning_rate": 4.622781409140552e-05,
"loss": 3.9212,
"step": 230400
},
{
"epoch": 1.0,
"learning_rate": 4.6219428143895e-05,
"loss": 3.9106,
"step": 230912
},
{
"epoch": 1.0,
"learning_rate": 4.621105857518821e-05,
"loss": 3.9239,
"step": 231424
},
{
"epoch": 1.0,
"learning_rate": 4.620267262767769e-05,
"loss": 3.9023,
"step": 231936
},
{
"epoch": 1.0,
"learning_rate": 4.619428668016717e-05,
"loss": 3.9152,
"step": 232448
},
{
"epoch": 1.0,
"learning_rate": 4.618590073265665e-05,
"loss": 3.8988,
"step": 232960
},
{
"epoch": 1.0,
"learning_rate": 4.617753116394986e-05,
"loss": 3.9126,
"step": 233472
},
{
"epoch": 1.0,
"learning_rate": 4.616914521643934e-05,
"loss": 3.9062,
"step": 233984
},
{
"epoch": 1.0,
"learning_rate": 4.616075926892882e-05,
"loss": 3.9179,
"step": 234496
},
{
"epoch": 1.0,
"learning_rate": 4.61523733214183e-05,
"loss": 3.9188,
"step": 235008
},
{
"epoch": 1.0,
"learning_rate": 4.614402013151524e-05,
"loss": 3.9034,
"step": 235520
},
{
"epoch": 1.0,
"learning_rate": 4.613563418400472e-05,
"loss": 3.907,
"step": 236032
},
{
"epoch": 1.0,
"learning_rate": 4.61272482364942e-05,
"loss": 3.9025,
"step": 236544
},
{
"epoch": 1.0,
"learning_rate": 4.611886228898368e-05,
"loss": 3.9034,
"step": 237056
},
{
"epoch": 1.0,
"learning_rate": 4.611047634147316e-05,
"loss": 3.895,
"step": 237568
},
{
"epoch": 1.0,
"learning_rate": 4.610209039396265e-05,
"loss": 3.9025,
"step": 238080
},
{
"epoch": 1.0,
"learning_rate": 4.609370444645213e-05,
"loss": 3.8966,
"step": 238592
},
{
"epoch": 1.0,
"learning_rate": 4.6085318498941607e-05,
"loss": 3.9249,
"step": 239104
},
{
"epoch": 1.0,
"learning_rate": 4.6076932551431087e-05,
"loss": 3.9003,
"step": 239616
},
{
"epoch": 1.0,
"learning_rate": 4.6068562982724296e-05,
"loss": 3.9081,
"step": 240128
},
{
"epoch": 1.0,
"learning_rate": 4.6060177035213776e-05,
"loss": 3.9109,
"step": 240640
},
{
"epoch": 1.0,
"learning_rate": 4.6051791087703256e-05,
"loss": 3.9114,
"step": 241152
},
{
"epoch": 1.0,
"learning_rate": 4.6043405140192735e-05,
"loss": 3.8906,
"step": 241664
},
{
"epoch": 1.0,
"learning_rate": 4.6035019192682215e-05,
"loss": 3.9003,
"step": 242176
},
{
"epoch": 1.0,
"learning_rate": 4.6026649623975425e-05,
"loss": 3.8927,
"step": 242688
},
{
"epoch": 1.0,
"learning_rate": 4.6018263676464904e-05,
"loss": 3.8981,
"step": 243200
},
{
"epoch": 1.0,
"learning_rate": 4.6009877728954384e-05,
"loss": 3.8896,
"step": 243712
},
{
"epoch": 1.01,
"learning_rate": 4.6001491781443864e-05,
"loss": 3.8902,
"step": 244224
},
{
"epoch": 1.01,
"learning_rate": 4.5993105833933344e-05,
"loss": 3.9009,
"step": 244736
},
{
"epoch": 1.01,
"learning_rate": 4.598473626522656e-05,
"loss": 3.9028,
"step": 245248
},
{
"epoch": 1.01,
"learning_rate": 4.597635031771604e-05,
"loss": 3.8941,
"step": 245760
},
{
"epoch": 1.01,
"learning_rate": 4.596796437020552e-05,
"loss": 3.8994,
"step": 246272
},
{
"epoch": 1.01,
"learning_rate": 4.5959578422695e-05,
"loss": 3.9013,
"step": 246784
},
{
"epoch": 1.01,
"learning_rate": 4.595120885398821e-05,
"loss": 3.8952,
"step": 247296
},
{
"epoch": 1.01,
"learning_rate": 4.594282290647769e-05,
"loss": 3.8816,
"step": 247808
},
{
"epoch": 1.01,
"learning_rate": 4.593443695896717e-05,
"loss": 3.8956,
"step": 248320
},
{
"epoch": 1.01,
"learning_rate": 4.592605101145665e-05,
"loss": 3.8748,
"step": 248832
},
{
"epoch": 1.01,
"learning_rate": 4.591768144274986e-05,
"loss": 3.8915,
"step": 249344
},
{
"epoch": 1.01,
"learning_rate": 4.590929549523934e-05,
"loss": 3.885,
"step": 249856
},
{
"epoch": 1.01,
"learning_rate": 4.590090954772882e-05,
"loss": 3.8983,
"step": 250368
},
{
"epoch": 1.01,
"learning_rate": 4.58925236002183e-05,
"loss": 3.8947,
"step": 250880
},
{
"epoch": 1.01,
"learning_rate": 4.5884137652707785e-05,
"loss": 3.887,
"step": 251392
},
{
"epoch": 1.01,
"learning_rate": 4.5875768084000994e-05,
"loss": 3.8974,
"step": 251904
},
{
"epoch": 1.01,
"learning_rate": 4.5867382136490474e-05,
"loss": 3.8939,
"step": 252416
},
{
"epoch": 1.01,
"learning_rate": 4.5858996188979954e-05,
"loss": 3.8908,
"step": 252928
},
{
"epoch": 1.01,
"learning_rate": 4.585061024146943e-05,
"loss": 3.876,
"step": 253440
},
{
"epoch": 1.01,
"learning_rate": 4.584222429395891e-05,
"loss": 3.8793,
"step": 253952
},
{
"epoch": 1.01,
"learning_rate": 4.583385472525212e-05,
"loss": 3.8957,
"step": 254464
},
{
"epoch": 1.01,
"learning_rate": 4.58254687777416e-05,
"loss": 3.8792,
"step": 254976
},
{
"epoch": 1.01,
"learning_rate": 4.5817082830231076e-05,
"loss": 3.8774,
"step": 255488
},
{
"epoch": 1.01,
"learning_rate": 4.5808696882720556e-05,
"loss": 3.874,
"step": 256000
},
{
"epoch": 1.01,
"learning_rate": 4.5800310935210036e-05,
"loss": 3.8854,
"step": 256512
},
{
"epoch": 1.01,
"learning_rate": 4.579192498769952e-05,
"loss": 3.8706,
"step": 257024
},
{
"epoch": 1.01,
"learning_rate": 4.578355541899273e-05,
"loss": 3.8835,
"step": 257536
},
{
"epoch": 1.01,
"learning_rate": 4.577516947148221e-05,
"loss": 3.8784,
"step": 258048
},
{
"epoch": 1.01,
"learning_rate": 4.576678352397169e-05,
"loss": 3.8835,
"step": 258560
},
{
"epoch": 1.01,
"learning_rate": 4.575839757646117e-05,
"loss": 3.8843,
"step": 259072
},
{
"epoch": 1.01,
"learning_rate": 4.575001162895065e-05,
"loss": 3.8703,
"step": 259584
},
{
"epoch": 1.01,
"learning_rate": 4.574164206024386e-05,
"loss": 3.8712,
"step": 260096
},
{
"epoch": 1.01,
"learning_rate": 4.573325611273334e-05,
"loss": 3.8831,
"step": 260608
},
{
"epoch": 1.01,
"learning_rate": 4.572487016522282e-05,
"loss": 3.8665,
"step": 261120
},
{
"epoch": 1.01,
"learning_rate": 4.57164842177123e-05,
"loss": 3.874,
"step": 261632
},
{
"epoch": 1.01,
"learning_rate": 4.570809827020178e-05,
"loss": 3.8836,
"step": 262144
},
{
"epoch": 1.01,
"learning_rate": 4.569972870149499e-05,
"loss": 3.8823,
"step": 262656
},
{
"epoch": 1.01,
"learning_rate": 4.5691342753984476e-05,
"loss": 3.8634,
"step": 263168
},
{
"epoch": 1.01,
"learning_rate": 4.5682956806473956e-05,
"loss": 3.8613,
"step": 263680
},
{
"epoch": 1.01,
"learning_rate": 4.5674570858963436e-05,
"loss": 3.8674,
"step": 264192
},
{
"epoch": 1.01,
"learning_rate": 4.5666184911452916e-05,
"loss": 3.8771,
"step": 264704
},
{
"epoch": 1.01,
"learning_rate": 4.5657815342746125e-05,
"loss": 3.8853,
"step": 265216
},
{
"epoch": 1.01,
"learning_rate": 4.5649429395235605e-05,
"loss": 3.8726,
"step": 265728
},
{
"epoch": 1.01,
"learning_rate": 4.5641043447725085e-05,
"loss": 3.8836,
"step": 266240
},
{
"epoch": 1.01,
"learning_rate": 4.5632657500214565e-05,
"loss": 3.8841,
"step": 266752
},
{
"epoch": 1.01,
"learning_rate": 4.5624271552704045e-05,
"loss": 3.8826,
"step": 267264
},
{
"epoch": 1.01,
"learning_rate": 4.5615901983997254e-05,
"loss": 3.861,
"step": 267776
},
{
"epoch": 1.01,
"learning_rate": 4.5607516036486734e-05,
"loss": 3.8811,
"step": 268288
},
{
"epoch": 1.01,
"learning_rate": 4.5599130088976214e-05,
"loss": 3.8724,
"step": 268800
},
{
"epoch": 1.01,
"learning_rate": 4.5590744141465694e-05,
"loss": 3.8661,
"step": 269312
},
{
"epoch": 1.01,
"learning_rate": 4.558237457275891e-05,
"loss": 3.8733,
"step": 269824
},
{
"epoch": 1.01,
"learning_rate": 4.557398862524839e-05,
"loss": 3.8707,
"step": 270336
},
{
"epoch": 1.01,
"learning_rate": 4.556560267773787e-05,
"loss": 3.8752,
"step": 270848
},
{
"epoch": 1.01,
"learning_rate": 4.555721673022735e-05,
"loss": 3.8718,
"step": 271360
},
{
"epoch": 1.01,
"learning_rate": 4.554883078271683e-05,
"loss": 3.8636,
"step": 271872
},
{
"epoch": 1.01,
"learning_rate": 4.554046121401004e-05,
"loss": 3.8712,
"step": 272384
},
{
"epoch": 1.01,
"learning_rate": 4.553207526649952e-05,
"loss": 3.8663,
"step": 272896
},
{
"epoch": 1.01,
"learning_rate": 4.5523689318989e-05,
"loss": 3.8685,
"step": 273408
},
{
"epoch": 1.01,
"learning_rate": 4.551530337147848e-05,
"loss": 3.8672,
"step": 273920
},
{
"epoch": 1.01,
"learning_rate": 4.550693380277169e-05,
"loss": 3.8619,
"step": 274432
},
{
"epoch": 1.02,
"learning_rate": 4.549854785526117e-05,
"loss": 3.8737,
"step": 274944
},
{
"epoch": 1.02,
"learning_rate": 4.549016190775065e-05,
"loss": 3.8671,
"step": 275456
},
{
"epoch": 1.02,
"learning_rate": 4.548177596024013e-05,
"loss": 3.8606,
"step": 275968
},
{
"epoch": 1.02,
"learning_rate": 4.5473406391533343e-05,
"loss": 3.8606,
"step": 276480
},
{
"epoch": 1.02,
"learning_rate": 4.5465020444022823e-05,
"loss": 3.8783,
"step": 276992
},
{
"epoch": 1.02,
"learning_rate": 4.54566344965123e-05,
"loss": 3.8697,
"step": 277504
},
{
"epoch": 1.02,
"learning_rate": 4.544824854900178e-05,
"loss": 3.8641,
"step": 278016
},
{
"epoch": 1.02,
"learning_rate": 4.543987898029499e-05,
"loss": 3.8545,
"step": 278528
},
{
"epoch": 1.02,
"learning_rate": 4.543149303278447e-05,
"loss": 3.8637,
"step": 279040
},
{
"epoch": 1.02,
"learning_rate": 4.542310708527395e-05,
"loss": 3.8566,
"step": 279552
},
{
"epoch": 1.02,
"learning_rate": 4.541472113776343e-05,
"loss": 3.8783,
"step": 280064
},
{
"epoch": 1.02,
"learning_rate": 4.540633519025291e-05,
"loss": 3.8629,
"step": 280576
},
{
"epoch": 1.02,
"learning_rate": 4.539796562154612e-05,
"loss": 3.8604,
"step": 281088
},
{
"epoch": 1.02,
"learning_rate": 4.53895796740356e-05,
"loss": 3.8548,
"step": 281600
},
{
"epoch": 1.02,
"learning_rate": 4.538119372652508e-05,
"loss": 3.8647,
"step": 282112
},
{
"epoch": 1.02,
"learning_rate": 4.537280777901456e-05,
"loss": 3.8368,
"step": 282624
},
{
"epoch": 1.02,
"learning_rate": 4.536443821030778e-05,
"loss": 3.8566,
"step": 283136
},
{
"epoch": 1.02,
"learning_rate": 4.5356068641600986e-05,
"loss": 3.8553,
"step": 283648
},
{
"epoch": 1.02,
"learning_rate": 4.5347682694090466e-05,
"loss": 3.8591,
"step": 284160
},
{
"epoch": 1.02,
"learning_rate": 4.5339296746579946e-05,
"loss": 3.8439,
"step": 284672
},
{
"epoch": 1.02,
"learning_rate": 4.5330910799069426e-05,
"loss": 3.8626,
"step": 285184
},
{
"epoch": 1.02,
"learning_rate": 4.5322524851558906e-05,
"loss": 3.8481,
"step": 285696
},
{
"epoch": 1.02,
"learning_rate": 4.5314138904048386e-05,
"loss": 3.8657,
"step": 286208
},
{
"epoch": 1.02,
"learning_rate": 4.530575295653786e-05,
"loss": 3.8596,
"step": 286720
},
{
"epoch": 1.02,
"learning_rate": 4.529736700902734e-05,
"loss": 3.8512,
"step": 287232
},
{
"epoch": 1.02,
"learning_rate": 4.5288997440320555e-05,
"loss": 3.8684,
"step": 287744
},
{
"epoch": 1.02,
"learning_rate": 4.5280611492810035e-05,
"loss": 3.8769,
"step": 288256
},
{
"epoch": 1.02,
"learning_rate": 4.5272225545299515e-05,
"loss": 3.8581,
"step": 288768
},
{
"epoch": 1.02,
"learning_rate": 4.5263839597788995e-05,
"loss": 3.8522,
"step": 289280
},
{
"epoch": 1.02,
"learning_rate": 4.525547002908221e-05,
"loss": 3.858,
"step": 289792
},
{
"epoch": 1.02,
"learning_rate": 4.5247084081571684e-05,
"loss": 3.8492,
"step": 290304
},
{
"epoch": 1.02,
"learning_rate": 4.5238698134061164e-05,
"loss": 3.8518,
"step": 290816
},
{
"epoch": 1.02,
"learning_rate": 4.5230312186550644e-05,
"loss": 3.8648,
"step": 291328
},
{
"epoch": 1.02,
"learning_rate": 4.5221926239040124e-05,
"loss": 3.8473,
"step": 291840
},
{
"epoch": 1.02,
"learning_rate": 4.521355667033333e-05,
"loss": 3.8614,
"step": 292352
},
{
"epoch": 1.02,
"learning_rate": 4.520517072282281e-05,
"loss": 3.8543,
"step": 292864
},
{
"epoch": 1.02,
"learning_rate": 4.519678477531229e-05,
"loss": 3.8578,
"step": 293376
},
{
"epoch": 1.02,
"learning_rate": 4.518839882780177e-05,
"loss": 3.8394,
"step": 293888
},
{
"epoch": 1.02,
"learning_rate": 4.518002925909499e-05,
"loss": 3.8585,
"step": 294400
},
{
"epoch": 1.02,
"learning_rate": 4.517164331158447e-05,
"loss": 3.845,
"step": 294912
},
{
"epoch": 1.02,
"learning_rate": 4.516325736407395e-05,
"loss": 3.8624,
"step": 295424
},
{
"epoch": 1.02,
"learning_rate": 4.515487141656343e-05,
"loss": 3.8528,
"step": 295936
},
{
"epoch": 1.02,
"learning_rate": 4.514650184785664e-05,
"loss": 3.8583,
"step": 296448
},
{
"epoch": 1.02,
"learning_rate": 4.513811590034612e-05,
"loss": 3.8355,
"step": 296960
},
{
"epoch": 1.02,
"learning_rate": 4.51297299528356e-05,
"loss": 3.8614,
"step": 297472
},
{
"epoch": 1.02,
"learning_rate": 4.512134400532508e-05,
"loss": 3.85,
"step": 297984
},
{
"epoch": 1.02,
"learning_rate": 4.5112974436618287e-05,
"loss": 3.853,
"step": 298496
},
{
"epoch": 1.02,
"learning_rate": 4.5104588489107766e-05,
"loss": 3.85,
"step": 299008
},
{
"epoch": 1.02,
"learning_rate": 4.5096202541597246e-05,
"loss": 3.8594,
"step": 299520
},
{
"epoch": 1.02,
"learning_rate": 4.5087816594086726e-05,
"loss": 3.8563,
"step": 300032
},
{
"epoch": 1.02,
"learning_rate": 4.507944702537994e-05,
"loss": 3.858,
"step": 300544
},
{
"epoch": 1.02,
"learning_rate": 4.507106107786942e-05,
"loss": 3.842,
"step": 301056
},
{
"epoch": 1.02,
"learning_rate": 4.50626751303589e-05,
"loss": 3.848,
"step": 301568
},
{
"epoch": 1.02,
"learning_rate": 4.505428918284838e-05,
"loss": 3.8483,
"step": 302080
},
{
"epoch": 1.02,
"learning_rate": 4.504591961414159e-05,
"loss": 3.8381,
"step": 302592
},
{
"epoch": 1.02,
"learning_rate": 4.503753366663107e-05,
"loss": 3.85,
"step": 303104
},
{
"epoch": 1.02,
"learning_rate": 4.502914771912055e-05,
"loss": 3.8568,
"step": 303616
},
{
"epoch": 1.02,
"learning_rate": 4.502076177161003e-05,
"loss": 3.8492,
"step": 304128
},
{
"epoch": 1.02,
"learning_rate": 4.501239220290324e-05,
"loss": 3.8421,
"step": 304640
},
{
"epoch": 1.02,
"learning_rate": 4.500400625539272e-05,
"loss": 3.8525,
"step": 305152
},
{
"epoch": 1.03,
"eval_loss": 3.911241292953491,
"eval_runtime": 306.3641,
"eval_samples_per_second": 1245.547,
"eval_steps_per_second": 38.924,
"step": 305280
},
{
"epoch": 0.0,
"learning_rate": 4.49956203078822e-05,
"loss": 3.8319,
"step": 305664
},
{
"epoch": 0.0,
"learning_rate": 4.498723436037168e-05,
"loss": 3.8381,
"step": 306176
},
{
"epoch": 0.0,
"learning_rate": 4.497884841286116e-05,
"loss": 3.8475,
"step": 306688
},
{
"epoch": 0.0,
"learning_rate": 4.497046246535065e-05,
"loss": 3.8442,
"step": 307200
},
{
"epoch": 0.0,
"learning_rate": 4.496207651784013e-05,
"loss": 3.8552,
"step": 307712
},
{
"epoch": 0.0,
"learning_rate": 4.495369057032961e-05,
"loss": 3.8331,
"step": 308224
},
{
"epoch": 0.0,
"learning_rate": 4.4945304622819087e-05,
"loss": 3.8433,
"step": 308736
},
{
"epoch": 0.0,
"learning_rate": 4.4936918675308567e-05,
"loss": 3.829,
"step": 309248
},
{
"epoch": 0.0,
"learning_rate": 4.4928549106601776e-05,
"loss": 3.8446,
"step": 309760
},
{
"epoch": 0.0,
"learning_rate": 4.4920163159091256e-05,
"loss": 3.8374,
"step": 310272
},
{
"epoch": 0.0,
"learning_rate": 4.4911777211580736e-05,
"loss": 3.8412,
"step": 310784
},
{
"epoch": 0.0,
"learning_rate": 4.4903391264070216e-05,
"loss": 3.8502,
"step": 311296
},
{
"epoch": 0.0,
"learning_rate": 4.4895038074167154e-05,
"loss": 3.8302,
"step": 311808
},
{
"epoch": 0.0,
"learning_rate": 4.4886652126656634e-05,
"loss": 3.8402,
"step": 312320
},
{
"epoch": 0.0,
"learning_rate": 4.4878266179146114e-05,
"loss": 3.8316,
"step": 312832
},
{
"epoch": 0.0,
"learning_rate": 4.48698802316356e-05,
"loss": 3.8354,
"step": 313344
},
{
"epoch": 0.0,
"learning_rate": 4.486149428412508e-05,
"loss": 3.8241,
"step": 313856
},
{
"epoch": 0.0,
"learning_rate": 4.485310833661456e-05,
"loss": 3.8335,
"step": 314368
},
{
"epoch": 0.0,
"learning_rate": 4.484472238910404e-05,
"loss": 3.8309,
"step": 314880
},
{
"epoch": 0.0,
"learning_rate": 4.483633644159352e-05,
"loss": 3.8528,
"step": 315392
},
{
"epoch": 0.0,
"learning_rate": 4.482796687288673e-05,
"loss": 3.8306,
"step": 315904
},
{
"epoch": 0.0,
"learning_rate": 4.481958092537621e-05,
"loss": 3.8392,
"step": 316416
},
{
"epoch": 0.0,
"learning_rate": 4.481119497786569e-05,
"loss": 3.8439,
"step": 316928
},
{
"epoch": 0.0,
"learning_rate": 4.480280903035516e-05,
"loss": 3.8452,
"step": 317440
},
{
"epoch": 0.0,
"learning_rate": 4.479443946164838e-05,
"loss": 3.8225,
"step": 317952
},
{
"epoch": 0.0,
"learning_rate": 4.478605351413786e-05,
"loss": 3.8355,
"step": 318464
},
{
"epoch": 0.0,
"learning_rate": 4.477766756662734e-05,
"loss": 3.824,
"step": 318976
},
{
"epoch": 0.0,
"learning_rate": 4.476928161911682e-05,
"loss": 3.8337,
"step": 319488
},
{
"epoch": 0.0,
"learning_rate": 4.4760912050410034e-05,
"loss": 3.8208,
"step": 320000
},
{
"epoch": 0.0,
"learning_rate": 4.4752526102899514e-05,
"loss": 3.8259,
"step": 320512
},
{
"epoch": 0.01,
"learning_rate": 4.4744140155388994e-05,
"loss": 3.8349,
"step": 321024
},
{
"epoch": 0.01,
"learning_rate": 4.47357705866822e-05,
"loss": 3.8361,
"step": 321536
},
{
"epoch": 0.01,
"learning_rate": 4.472738463917168e-05,
"loss": 3.8263,
"step": 322048
},
{
"epoch": 0.01,
"learning_rate": 4.471899869166116e-05,
"loss": 3.8363,
"step": 322560
},
{
"epoch": 0.01,
"learning_rate": 4.4710612744150636e-05,
"loss": 3.8283,
"step": 323072
},
{
"epoch": 0.01,
"learning_rate": 4.4702226796640116e-05,
"loss": 3.835,
"step": 323584
},
{
"epoch": 0.01,
"learning_rate": 4.4693840849129596e-05,
"loss": 3.8163,
"step": 324096
},
{
"epoch": 0.01,
"learning_rate": 4.4685454901619076e-05,
"loss": 3.8283,
"step": 324608
},
{
"epoch": 0.01,
"learning_rate": 4.4677085332912285e-05,
"loss": 3.8144,
"step": 325120
},
{
"epoch": 0.01,
"learning_rate": 4.466869938540177e-05,
"loss": 3.8218,
"step": 325632
},
{
"epoch": 0.01,
"learning_rate": 4.466031343789125e-05,
"loss": 3.8193,
"step": 326144
},
{
"epoch": 0.01,
"learning_rate": 4.465192749038073e-05,
"loss": 3.833,
"step": 326656
},
{
"epoch": 0.01,
"learning_rate": 4.464355792167394e-05,
"loss": 3.8276,
"step": 327168
},
{
"epoch": 0.01,
"learning_rate": 4.463517197416342e-05,
"loss": 3.8273,
"step": 327680
},
{
"epoch": 0.01,
"learning_rate": 4.46267860266529e-05,
"loss": 3.8304,
"step": 328192
},
{
"epoch": 0.01,
"learning_rate": 4.461840007914238e-05,
"loss": 3.8298,
"step": 328704
},
{
"epoch": 0.01,
"learning_rate": 4.461003051043559e-05,
"loss": 3.825,
"step": 329216
},
{
"epoch": 0.01,
"learning_rate": 4.460164456292507e-05,
"loss": 3.8173,
"step": 329728
},
{
"epoch": 0.01,
"learning_rate": 4.459325861541455e-05,
"loss": 3.8084,
"step": 330240
},
{
"epoch": 0.01,
"learning_rate": 4.458487266790403e-05,
"loss": 3.8311,
"step": 330752
},
{
"epoch": 0.01,
"learning_rate": 4.457650309919724e-05,
"loss": 3.8191,
"step": 331264
},
{
"epoch": 0.01,
"learning_rate": 4.4568117151686725e-05,
"loss": 3.8138,
"step": 331776
},
{
"epoch": 0.01,
"learning_rate": 4.4559731204176205e-05,
"loss": 3.8147,
"step": 332288
},
{
"epoch": 0.01,
"learning_rate": 4.4551345256665685e-05,
"loss": 3.8211,
"step": 332800
},
{
"epoch": 0.01,
"learning_rate": 4.4542975687958895e-05,
"loss": 3.8067,
"step": 333312
},
{
"epoch": 0.01,
"learning_rate": 4.4534589740448374e-05,
"loss": 3.815,
"step": 333824
},
{
"epoch": 0.01,
"learning_rate": 4.4526203792937854e-05,
"loss": 3.8154,
"step": 334336
},
{
"epoch": 0.01,
"learning_rate": 4.4517817845427334e-05,
"loss": 3.815,
"step": 334848
},
{
"epoch": 0.01,
"learning_rate": 4.4509448276720543e-05,
"loss": 3.8267,
"step": 335360
},
{
"epoch": 0.01,
"learning_rate": 4.4501062329210023e-05,
"loss": 3.8071,
"step": 335872
},
{
"epoch": 0.01,
"learning_rate": 4.44926763816995e-05,
"loss": 3.8088,
"step": 336384
},
{
"epoch": 0.01,
"learning_rate": 4.448429043418898e-05,
"loss": 3.8219,
"step": 336896
},
{
"epoch": 0.01,
"learning_rate": 4.447592086548219e-05,
"loss": 3.8072,
"step": 337408
},
{
"epoch": 0.01,
"learning_rate": 4.446753491797168e-05,
"loss": 3.8104,
"step": 337920
},
{
"epoch": 0.01,
"learning_rate": 4.445914897046116e-05,
"loss": 3.8209,
"step": 338432
},
{
"epoch": 0.01,
"learning_rate": 4.445076302295064e-05,
"loss": 3.8202,
"step": 338944
},
{
"epoch": 0.01,
"learning_rate": 4.444239345424385e-05,
"loss": 3.8049,
"step": 339456
},
{
"epoch": 0.01,
"learning_rate": 4.443400750673333e-05,
"loss": 3.799,
"step": 339968
},
{
"epoch": 0.01,
"learning_rate": 4.442562155922281e-05,
"loss": 3.8015,
"step": 340480
},
{
"epoch": 0.01,
"learning_rate": 4.441723561171229e-05,
"loss": 3.815,
"step": 340992
},
{
"epoch": 0.01,
"learning_rate": 4.44088660430055e-05,
"loss": 3.8258,
"step": 341504
},
{
"epoch": 0.01,
"learning_rate": 4.440048009549498e-05,
"loss": 3.8103,
"step": 342016
},
{
"epoch": 0.01,
"learning_rate": 4.439209414798446e-05,
"loss": 3.8225,
"step": 342528
},
{
"epoch": 0.01,
"learning_rate": 4.438370820047394e-05,
"loss": 3.8201,
"step": 343040
},
{
"epoch": 0.01,
"learning_rate": 4.4375338631767146e-05,
"loss": 3.8249,
"step": 343552
},
{
"epoch": 0.01,
"learning_rate": 4.436695268425663e-05,
"loss": 3.8013,
"step": 344064
},
{
"epoch": 0.01,
"learning_rate": 4.435856673674611e-05,
"loss": 3.817,
"step": 344576
},
{
"epoch": 0.01,
"learning_rate": 4.435018078923559e-05,
"loss": 3.8096,
"step": 345088
},
{
"epoch": 0.01,
"learning_rate": 4.43418112205288e-05,
"loss": 3.8028,
"step": 345600
},
{
"epoch": 0.01,
"learning_rate": 4.433342527301828e-05,
"loss": 3.8178,
"step": 346112
},
{
"epoch": 0.01,
"learning_rate": 4.432503932550776e-05,
"loss": 3.8109,
"step": 346624
},
{
"epoch": 0.01,
"learning_rate": 4.431665337799724e-05,
"loss": 3.8089,
"step": 347136
},
{
"epoch": 0.01,
"learning_rate": 4.430828380929045e-05,
"loss": 3.8139,
"step": 347648
},
{
"epoch": 0.01,
"learning_rate": 4.429989786177993e-05,
"loss": 3.8079,
"step": 348160
},
{
"epoch": 0.01,
"learning_rate": 4.429151191426941e-05,
"loss": 3.8097,
"step": 348672
},
{
"epoch": 0.01,
"learning_rate": 4.428312596675889e-05,
"loss": 3.8063,
"step": 349184
},
{
"epoch": 0.01,
"learning_rate": 4.42747563980521e-05,
"loss": 3.8107,
"step": 349696
},
{
"epoch": 0.01,
"learning_rate": 4.4266370450541586e-05,
"loss": 3.804,
"step": 350208
},
{
"epoch": 0.01,
"learning_rate": 4.4257984503031066e-05,
"loss": 3.804,
"step": 350720
},
{
"epoch": 0.02,
"learning_rate": 4.4249614934324276e-05,
"loss": 3.8121,
"step": 351232
},
{
"epoch": 0.02,
"learning_rate": 4.4241228986813756e-05,
"loss": 3.8107,
"step": 351744
},
{
"epoch": 0.02,
"learning_rate": 4.4232843039303235e-05,
"loss": 3.7989,
"step": 352256
},
{
"epoch": 0.02,
"learning_rate": 4.4224457091792715e-05,
"loss": 3.8007,
"step": 352768
},
{
"epoch": 0.02,
"learning_rate": 4.4216071144282195e-05,
"loss": 3.8191,
"step": 353280
},
{
"epoch": 0.02,
"learning_rate": 4.4207685196771675e-05,
"loss": 3.8115,
"step": 353792
},
{
"epoch": 0.02,
"learning_rate": 4.4199299249261155e-05,
"loss": 3.8091,
"step": 354304
},
{
"epoch": 0.02,
"learning_rate": 4.4190913301750635e-05,
"loss": 3.7965,
"step": 354816
},
{
"epoch": 0.02,
"learning_rate": 4.4182543733043844e-05,
"loss": 3.8048,
"step": 355328
},
{
"epoch": 0.02,
"learning_rate": 4.4174157785533324e-05,
"loss": 3.8008,
"step": 355840
},
{
"epoch": 0.02,
"learning_rate": 4.4165771838022804e-05,
"loss": 3.8148,
"step": 356352
},
{
"epoch": 0.02,
"learning_rate": 4.4157385890512284e-05,
"loss": 3.8129,
"step": 356864
},
{
"epoch": 0.02,
"learning_rate": 4.41490163218055e-05,
"loss": 3.8011,
"step": 357376
},
{
"epoch": 0.02,
"learning_rate": 4.414063037429498e-05,
"loss": 3.7958,
"step": 357888
},
{
"epoch": 0.02,
"learning_rate": 4.413224442678446e-05,
"loss": 3.8085,
"step": 358400
},
{
"epoch": 0.02,
"learning_rate": 4.412385847927394e-05,
"loss": 3.7782,
"step": 358912
},
{
"epoch": 0.02,
"learning_rate": 4.411548891056715e-05,
"loss": 3.8026,
"step": 359424
},
{
"epoch": 0.02,
"learning_rate": 4.410710296305663e-05,
"loss": 3.7934,
"step": 359936
},
{
"epoch": 0.02,
"learning_rate": 4.409871701554611e-05,
"loss": 3.8012,
"step": 360448
},
{
"epoch": 0.02,
"learning_rate": 4.409033106803559e-05,
"loss": 3.7877,
"step": 360960
},
{
"epoch": 0.02,
"learning_rate": 4.40819614993288e-05,
"loss": 3.8055,
"step": 361472
},
{
"epoch": 0.02,
"learning_rate": 4.407357555181828e-05,
"loss": 3.7981,
"step": 361984
},
{
"epoch": 0.02,
"learning_rate": 4.406518960430776e-05,
"loss": 3.8002,
"step": 362496
},
{
"epoch": 0.02,
"learning_rate": 4.405680365679724e-05,
"loss": 3.8047,
"step": 363008
},
{
"epoch": 0.02,
"learning_rate": 4.4048434088090454e-05,
"loss": 3.7932,
"step": 363520
},
{
"epoch": 0.02,
"learning_rate": 4.4040048140579934e-05,
"loss": 3.8115,
"step": 364032
},
{
"epoch": 0.02,
"learning_rate": 4.4031662193069414e-05,
"loss": 3.8228,
"step": 364544
},
{
"epoch": 0.02,
"learning_rate": 4.4023276245558894e-05,
"loss": 3.8003,
"step": 365056
},
{
"epoch": 0.02,
"learning_rate": 4.40149066768521e-05,
"loss": 3.7939,
"step": 365568
},
{
"epoch": 0.02,
"learning_rate": 4.400652072934158e-05,
"loss": 3.8045,
"step": 366080
},
{
"epoch": 0.02,
"learning_rate": 4.399813478183106e-05,
"loss": 3.7977,
"step": 366592
},
{
"epoch": 0.02,
"learning_rate": 4.398974883432054e-05,
"loss": 3.7916,
"step": 367104
},
{
"epoch": 0.02,
"learning_rate": 4.398137926561375e-05,
"loss": 3.8108,
"step": 367616
},
{
"epoch": 0.02,
"learning_rate": 4.397299331810323e-05,
"loss": 3.7898,
"step": 368128
},
{
"epoch": 0.02,
"learning_rate": 4.396460737059271e-05,
"loss": 3.8039,
"step": 368640
},
{
"epoch": 0.02,
"learning_rate": 4.395623780188593e-05,
"loss": 3.8003,
"step": 369152
},
{
"epoch": 0.02,
"learning_rate": 4.394785185437541e-05,
"loss": 3.8014,
"step": 369664
},
{
"epoch": 0.02,
"learning_rate": 4.393946590686489e-05,
"loss": 3.7911,
"step": 370176
},
{
"epoch": 0.02,
"learning_rate": 4.393107995935437e-05,
"loss": 3.7952,
"step": 370688
},
{
"epoch": 0.02,
"learning_rate": 4.3922710390647576e-05,
"loss": 3.7939,
"step": 371200
},
{
"epoch": 0.02,
"learning_rate": 4.3914324443137056e-05,
"loss": 3.8071,
"step": 371712
},
{
"epoch": 0.02,
"learning_rate": 4.3905938495626536e-05,
"loss": 3.796,
"step": 372224
},
{
"epoch": 0.02,
"learning_rate": 4.3897552548116016e-05,
"loss": 3.8007,
"step": 372736
},
{
"epoch": 0.02,
"learning_rate": 4.3889182979409225e-05,
"loss": 3.7824,
"step": 373248
},
{
"epoch": 0.02,
"learning_rate": 4.3880797031898705e-05,
"loss": 3.8046,
"step": 373760
},
{
"epoch": 0.02,
"learning_rate": 4.3872411084388185e-05,
"loss": 3.7967,
"step": 374272
},
{
"epoch": 0.02,
"learning_rate": 4.3864025136877665e-05,
"loss": 3.8011,
"step": 374784
},
{
"epoch": 0.02,
"learning_rate": 4.385565556817088e-05,
"loss": 3.7939,
"step": 375296
},
{
"epoch": 0.02,
"learning_rate": 4.384726962066036e-05,
"loss": 3.8049,
"step": 375808
},
{
"epoch": 0.02,
"learning_rate": 4.383888367314984e-05,
"loss": 3.8004,
"step": 376320
},
{
"epoch": 0.02,
"learning_rate": 4.3830497725639314e-05,
"loss": 3.8074,
"step": 376832
},
{
"epoch": 0.02,
"learning_rate": 4.382212815693253e-05,
"loss": 3.7845,
"step": 377344
},
{
"epoch": 0.02,
"learning_rate": 4.381374220942201e-05,
"loss": 3.7958,
"step": 377856
},
{
"epoch": 0.02,
"learning_rate": 4.380535626191148e-05,
"loss": 3.7914,
"step": 378368
},
{
"epoch": 0.02,
"learning_rate": 4.379697031440096e-05,
"loss": 3.7859,
"step": 378880
},
{
"epoch": 0.02,
"learning_rate": 4.378860074569418e-05,
"loss": 3.7981,
"step": 379392
},
{
"epoch": 0.02,
"learning_rate": 4.378021479818366e-05,
"loss": 3.7987,
"step": 379904
},
{
"epoch": 0.02,
"learning_rate": 4.377182885067313e-05,
"loss": 3.8004,
"step": 380416
},
{
"epoch": 0.02,
"learning_rate": 4.376344290316262e-05,
"loss": 3.7805,
"step": 380928
},
{
"epoch": 0.02,
"learning_rate": 4.3755073334455835e-05,
"loss": 3.7987,
"step": 381440
},
{
"epoch": 0.03,
"eval_loss": 3.887401580810547,
"eval_runtime": 319.04,
"eval_samples_per_second": 1196.06,
"eval_steps_per_second": 37.378,
"step": 381600
}
],
"logging_steps": 512,
"max_steps": 3052726,
"num_train_epochs": 9223372036854775807,
"save_steps": 10,
"total_flos": 2.6743688723920896e+17,
"trial_name": null,
"trial_params": null
}