abhinavp's picture
Training in progress, epoch 1, checkpoint
c4f0671 verified
{
"best_metric": 3.8919215202331543,
"best_model_checkpoint": "/mmfs1/gscratch/stf/abhinavp/corpus-filtering/outputs/passive/transformer/3/checkpoints/checkpoint-381600",
"epoch": 1.0250006060157382,
"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": 10.9682,
"step": 1
},
{
"epoch": 0.0,
"learning_rate": 4.999161405248948e-05,
"loss": 6.8285,
"step": 512
},
{
"epoch": 0.0,
"learning_rate": 4.998322810497896e-05,
"loss": 6.1802,
"step": 1024
},
{
"epoch": 0.0,
"learning_rate": 4.997484215746844e-05,
"loss": 5.9721,
"step": 1536
},
{
"epoch": 0.0,
"learning_rate": 4.996645620995792e-05,
"loss": 5.8073,
"step": 2048
},
{
"epoch": 0.0,
"learning_rate": 4.99580702624474e-05,
"loss": 5.6887,
"step": 2560
},
{
"epoch": 0.0,
"learning_rate": 4.994968431493688e-05,
"loss": 5.602,
"step": 3072
},
{
"epoch": 0.0,
"learning_rate": 4.994129836742636e-05,
"loss": 5.5256,
"step": 3584
},
{
"epoch": 0.0,
"learning_rate": 4.993291241991584e-05,
"loss": 5.444,
"step": 4096
},
{
"epoch": 0.0,
"learning_rate": 4.992452647240532e-05,
"loss": 5.3966,
"step": 4608
},
{
"epoch": 0.0,
"learning_rate": 4.99161405248948e-05,
"loss": 5.3464,
"step": 5120
},
{
"epoch": 0.0,
"learning_rate": 4.990775457738428e-05,
"loss": 5.3064,
"step": 5632
},
{
"epoch": 0.0,
"learning_rate": 4.989936862987376e-05,
"loss": 5.2596,
"step": 6144
},
{
"epoch": 0.0,
"learning_rate": 4.989099906116697e-05,
"loss": 5.211,
"step": 6656
},
{
"epoch": 0.0,
"learning_rate": 4.988261311365645e-05,
"loss": 5.1732,
"step": 7168
},
{
"epoch": 0.0,
"learning_rate": 4.987422716614593e-05,
"loss": 5.1263,
"step": 7680
},
{
"epoch": 0.0,
"learning_rate": 4.986584121863541e-05,
"loss": 5.1123,
"step": 8192
},
{
"epoch": 0.0,
"learning_rate": 4.985745527112489e-05,
"loss": 5.0786,
"step": 8704
},
{
"epoch": 0.0,
"learning_rate": 4.984906932361437e-05,
"loss": 5.0487,
"step": 9216
},
{
"epoch": 0.0,
"learning_rate": 4.984068337610385e-05,
"loss": 5.0205,
"step": 9728
},
{
"epoch": 0.0,
"learning_rate": 4.983229742859333e-05,
"loss": 5.0124,
"step": 10240
},
{
"epoch": 0.0,
"learning_rate": 4.9823927859886547e-05,
"loss": 4.9787,
"step": 10752
},
{
"epoch": 0.0,
"learning_rate": 4.9815541912376026e-05,
"loss": 4.9548,
"step": 11264
},
{
"epoch": 0.0,
"learning_rate": 4.9807155964865506e-05,
"loss": 4.9274,
"step": 11776
},
{
"epoch": 0.0,
"learning_rate": 4.9798770017354986e-05,
"loss": 4.9246,
"step": 12288
},
{
"epoch": 0.0,
"learning_rate": 4.9790400448648195e-05,
"loss": 4.8876,
"step": 12800
},
{
"epoch": 0.0,
"learning_rate": 4.9782014501137675e-05,
"loss": 4.8755,
"step": 13312
},
{
"epoch": 0.0,
"learning_rate": 4.9773628553627155e-05,
"loss": 4.8465,
"step": 13824
},
{
"epoch": 0.0,
"learning_rate": 4.9765242606116635e-05,
"loss": 4.837,
"step": 14336
},
{
"epoch": 0.0,
"learning_rate": 4.9756856658606115e-05,
"loss": 4.8225,
"step": 14848
},
{
"epoch": 0.01,
"learning_rate": 4.9748487089899324e-05,
"loss": 4.8048,
"step": 15360
},
{
"epoch": 0.01,
"learning_rate": 4.9740101142388804e-05,
"loss": 4.7969,
"step": 15872
},
{
"epoch": 0.01,
"learning_rate": 4.9731715194878284e-05,
"loss": 4.7842,
"step": 16384
},
{
"epoch": 0.01,
"learning_rate": 4.97233456261715e-05,
"loss": 4.7668,
"step": 16896
},
{
"epoch": 0.01,
"learning_rate": 4.971495967866098e-05,
"loss": 4.7669,
"step": 17408
},
{
"epoch": 0.01,
"learning_rate": 4.970657373115046e-05,
"loss": 4.7453,
"step": 17920
},
{
"epoch": 0.01,
"learning_rate": 4.969818778363994e-05,
"loss": 4.7414,
"step": 18432
},
{
"epoch": 0.01,
"learning_rate": 4.968981821493315e-05,
"loss": 4.7015,
"step": 18944
},
{
"epoch": 0.01,
"learning_rate": 4.968143226742263e-05,
"loss": 4.7119,
"step": 19456
},
{
"epoch": 0.01,
"learning_rate": 4.967304631991211e-05,
"loss": 4.6761,
"step": 19968
},
{
"epoch": 0.01,
"learning_rate": 4.966466037240159e-05,
"loss": 4.6873,
"step": 20480
},
{
"epoch": 0.01,
"learning_rate": 4.965627442489107e-05,
"loss": 4.6693,
"step": 20992
},
{
"epoch": 0.01,
"learning_rate": 4.964788847738054e-05,
"loss": 4.6668,
"step": 21504
},
{
"epoch": 0.01,
"learning_rate": 4.963950252987002e-05,
"loss": 4.6434,
"step": 22016
},
{
"epoch": 0.01,
"learning_rate": 4.96311165823595e-05,
"loss": 4.6552,
"step": 22528
},
{
"epoch": 0.01,
"learning_rate": 4.962274701365272e-05,
"loss": 4.6406,
"step": 23040
},
{
"epoch": 0.01,
"learning_rate": 4.96143610661422e-05,
"loss": 4.635,
"step": 23552
},
{
"epoch": 0.01,
"learning_rate": 4.960597511863168e-05,
"loss": 4.6289,
"step": 24064
},
{
"epoch": 0.01,
"learning_rate": 4.959758917112116e-05,
"loss": 4.5955,
"step": 24576
},
{
"epoch": 0.01,
"learning_rate": 4.9589219602414374e-05,
"loss": 4.6008,
"step": 25088
},
{
"epoch": 0.01,
"learning_rate": 4.958083365490385e-05,
"loss": 4.601,
"step": 25600
},
{
"epoch": 0.01,
"learning_rate": 4.957244770739333e-05,
"loss": 4.5904,
"step": 26112
},
{
"epoch": 0.01,
"learning_rate": 4.956406175988281e-05,
"loss": 4.5836,
"step": 26624
},
{
"epoch": 0.01,
"learning_rate": 4.9555692191176016e-05,
"loss": 4.5643,
"step": 27136
},
{
"epoch": 0.01,
"learning_rate": 4.9547306243665496e-05,
"loss": 4.5609,
"step": 27648
},
{
"epoch": 0.01,
"learning_rate": 4.9538920296154976e-05,
"loss": 4.5427,
"step": 28160
},
{
"epoch": 0.01,
"learning_rate": 4.9530534348644456e-05,
"loss": 4.5666,
"step": 28672
},
{
"epoch": 0.01,
"learning_rate": 4.9522148401133936e-05,
"loss": 4.5208,
"step": 29184
},
{
"epoch": 0.01,
"learning_rate": 4.951377883242715e-05,
"loss": 4.5443,
"step": 29696
},
{
"epoch": 0.01,
"learning_rate": 4.950539288491663e-05,
"loss": 4.5361,
"step": 30208
},
{
"epoch": 0.01,
"learning_rate": 4.949700693740611e-05,
"loss": 4.515,
"step": 30720
},
{
"epoch": 0.01,
"learning_rate": 4.948862098989559e-05,
"loss": 4.5151,
"step": 31232
},
{
"epoch": 0.01,
"learning_rate": 4.94802514211888e-05,
"loss": 4.5053,
"step": 31744
},
{
"epoch": 0.01,
"learning_rate": 4.947186547367828e-05,
"loss": 4.4986,
"step": 32256
},
{
"epoch": 0.01,
"learning_rate": 4.946347952616776e-05,
"loss": 4.4909,
"step": 32768
},
{
"epoch": 0.01,
"learning_rate": 4.945509357865724e-05,
"loss": 4.502,
"step": 33280
},
{
"epoch": 0.01,
"learning_rate": 4.944672400995045e-05,
"loss": 4.4797,
"step": 33792
},
{
"epoch": 0.01,
"learning_rate": 4.943833806243993e-05,
"loss": 4.4753,
"step": 34304
},
{
"epoch": 0.01,
"learning_rate": 4.942995211492941e-05,
"loss": 4.4667,
"step": 34816
},
{
"epoch": 0.01,
"learning_rate": 4.9421582546222625e-05,
"loss": 4.47,
"step": 35328
},
{
"epoch": 0.01,
"learning_rate": 4.9413196598712105e-05,
"loss": 4.4777,
"step": 35840
},
{
"epoch": 0.01,
"learning_rate": 4.9404810651201585e-05,
"loss": 4.4724,
"step": 36352
},
{
"epoch": 0.01,
"learning_rate": 4.9396424703691065e-05,
"loss": 4.4491,
"step": 36864
},
{
"epoch": 0.01,
"learning_rate": 4.9388038756180545e-05,
"loss": 4.463,
"step": 37376
},
{
"epoch": 0.01,
"learning_rate": 4.9379652808670025e-05,
"loss": 4.4631,
"step": 37888
},
{
"epoch": 0.01,
"learning_rate": 4.9371266861159505e-05,
"loss": 4.4346,
"step": 38400
},
{
"epoch": 0.01,
"learning_rate": 4.9362897292452714e-05,
"loss": 4.4424,
"step": 38912
},
{
"epoch": 0.01,
"learning_rate": 4.9354511344942194e-05,
"loss": 4.4329,
"step": 39424
},
{
"epoch": 0.01,
"learning_rate": 4.9346125397431674e-05,
"loss": 4.4289,
"step": 39936
},
{
"epoch": 0.01,
"learning_rate": 4.9337739449921154e-05,
"loss": 4.4189,
"step": 40448
},
{
"epoch": 0.01,
"learning_rate": 4.9329353502410634e-05,
"loss": 4.4222,
"step": 40960
},
{
"epoch": 0.01,
"learning_rate": 4.9320967554900114e-05,
"loss": 4.4249,
"step": 41472
},
{
"epoch": 0.01,
"learning_rate": 4.9312581607389594e-05,
"loss": 4.4282,
"step": 41984
},
{
"epoch": 0.01,
"learning_rate": 4.930421203868281e-05,
"loss": 4.4054,
"step": 42496
},
{
"epoch": 0.01,
"learning_rate": 4.929582609117229e-05,
"loss": 4.3833,
"step": 43008
},
{
"epoch": 0.01,
"learning_rate": 4.928744014366177e-05,
"loss": 4.3976,
"step": 43520
},
{
"epoch": 0.01,
"learning_rate": 4.927905419615125e-05,
"loss": 4.3984,
"step": 44032
},
{
"epoch": 0.01,
"learning_rate": 4.927066824864073e-05,
"loss": 4.3971,
"step": 44544
},
{
"epoch": 0.01,
"learning_rate": 4.926229867993394e-05,
"loss": 4.385,
"step": 45056
},
{
"epoch": 0.01,
"learning_rate": 4.925392911122715e-05,
"loss": 4.3881,
"step": 45568
},
{
"epoch": 0.02,
"learning_rate": 4.924554316371663e-05,
"loss": 4.384,
"step": 46080
},
{
"epoch": 0.02,
"learning_rate": 4.923715721620611e-05,
"loss": 4.3802,
"step": 46592
},
{
"epoch": 0.02,
"learning_rate": 4.922878764749932e-05,
"loss": 4.3679,
"step": 47104
},
{
"epoch": 0.02,
"learning_rate": 4.92204016999888e-05,
"loss": 4.3737,
"step": 47616
},
{
"epoch": 0.02,
"learning_rate": 4.921201575247828e-05,
"loss": 4.3608,
"step": 48128
},
{
"epoch": 0.02,
"learning_rate": 4.920362980496776e-05,
"loss": 4.3748,
"step": 48640
},
{
"epoch": 0.02,
"learning_rate": 4.919524385745724e-05,
"loss": 4.3612,
"step": 49152
},
{
"epoch": 0.02,
"learning_rate": 4.918685790994672e-05,
"loss": 4.3415,
"step": 49664
},
{
"epoch": 0.02,
"learning_rate": 4.91784719624362e-05,
"loss": 4.3547,
"step": 50176
},
{
"epoch": 0.02,
"learning_rate": 4.9170086014925676e-05,
"loss": 4.3551,
"step": 50688
},
{
"epoch": 0.02,
"learning_rate": 4.9161700067415156e-05,
"loss": 4.3584,
"step": 51200
},
{
"epoch": 0.02,
"learning_rate": 4.9153314119904636e-05,
"loss": 4.3424,
"step": 51712
},
{
"epoch": 0.02,
"learning_rate": 4.914494455119785e-05,
"loss": 4.3422,
"step": 52224
},
{
"epoch": 0.02,
"learning_rate": 4.9136558603687325e-05,
"loss": 4.3276,
"step": 52736
},
{
"epoch": 0.02,
"learning_rate": 4.9128172656176805e-05,
"loss": 4.3364,
"step": 53248
},
{
"epoch": 0.02,
"learning_rate": 4.9119786708666285e-05,
"loss": 4.3174,
"step": 53760
},
{
"epoch": 0.02,
"learning_rate": 4.91114171399595e-05,
"loss": 4.327,
"step": 54272
},
{
"epoch": 0.02,
"learning_rate": 4.910303119244898e-05,
"loss": 4.3231,
"step": 54784
},
{
"epoch": 0.02,
"learning_rate": 4.909464524493846e-05,
"loss": 4.3223,
"step": 55296
},
{
"epoch": 0.02,
"learning_rate": 4.908625929742794e-05,
"loss": 4.3139,
"step": 55808
},
{
"epoch": 0.02,
"learning_rate": 4.907787334991742e-05,
"loss": 4.3161,
"step": 56320
},
{
"epoch": 0.02,
"learning_rate": 4.90694874024069e-05,
"loss": 4.3074,
"step": 56832
},
{
"epoch": 0.02,
"learning_rate": 4.906110145489638e-05,
"loss": 4.3155,
"step": 57344
},
{
"epoch": 0.02,
"learning_rate": 4.905273188618959e-05,
"loss": 4.3062,
"step": 57856
},
{
"epoch": 0.02,
"learning_rate": 4.904434593867907e-05,
"loss": 4.3011,
"step": 58368
},
{
"epoch": 0.02,
"learning_rate": 4.903595999116855e-05,
"loss": 4.3154,
"step": 58880
},
{
"epoch": 0.02,
"learning_rate": 4.902757404365803e-05,
"loss": 4.3056,
"step": 59392
},
{
"epoch": 0.02,
"learning_rate": 4.9019220853754975e-05,
"loss": 4.3019,
"step": 59904
},
{
"epoch": 0.02,
"learning_rate": 4.9010834906244455e-05,
"loss": 4.2895,
"step": 60416
},
{
"epoch": 0.02,
"learning_rate": 4.9002448958733935e-05,
"loss": 4.2959,
"step": 60928
},
{
"epoch": 0.02,
"learning_rate": 4.8994063011223415e-05,
"loss": 4.2879,
"step": 61440
},
{
"epoch": 0.02,
"learning_rate": 4.8985677063712895e-05,
"loss": 4.2875,
"step": 61952
},
{
"epoch": 0.02,
"learning_rate": 4.8977291116202375e-05,
"loss": 4.2958,
"step": 62464
},
{
"epoch": 0.02,
"learning_rate": 4.8968905168691855e-05,
"loss": 4.2793,
"step": 62976
},
{
"epoch": 0.02,
"learning_rate": 4.8960519221181335e-05,
"loss": 4.2875,
"step": 63488
},
{
"epoch": 0.02,
"learning_rate": 4.8952133273670815e-05,
"loss": 4.2677,
"step": 64000
},
{
"epoch": 0.02,
"learning_rate": 4.8943763704964024e-05,
"loss": 4.2711,
"step": 64512
},
{
"epoch": 0.02,
"learning_rate": 4.8935377757453504e-05,
"loss": 4.277,
"step": 65024
},
{
"epoch": 0.02,
"learning_rate": 4.8926991809942984e-05,
"loss": 4.2687,
"step": 65536
},
{
"epoch": 0.02,
"learning_rate": 4.8918605862432463e-05,
"loss": 4.26,
"step": 66048
},
{
"epoch": 0.02,
"learning_rate": 4.8910219914921943e-05,
"loss": 4.2858,
"step": 66560
},
{
"epoch": 0.02,
"learning_rate": 4.8901833967411423e-05,
"loss": 4.2652,
"step": 67072
},
{
"epoch": 0.02,
"learning_rate": 4.889346439870464e-05,
"loss": 4.2599,
"step": 67584
},
{
"epoch": 0.02,
"learning_rate": 4.888507845119412e-05,
"loss": 4.2603,
"step": 68096
},
{
"epoch": 0.02,
"learning_rate": 4.88766925036836e-05,
"loss": 4.2606,
"step": 68608
},
{
"epoch": 0.02,
"learning_rate": 4.886830655617308e-05,
"loss": 4.2616,
"step": 69120
},
{
"epoch": 0.02,
"learning_rate": 4.885992060866256e-05,
"loss": 4.2584,
"step": 69632
},
{
"epoch": 0.02,
"learning_rate": 4.885155103995577e-05,
"loss": 4.2534,
"step": 70144
},
{
"epoch": 0.02,
"learning_rate": 4.884316509244525e-05,
"loss": 4.2615,
"step": 70656
},
{
"epoch": 0.02,
"learning_rate": 4.883477914493473e-05,
"loss": 4.2583,
"step": 71168
},
{
"epoch": 0.02,
"learning_rate": 4.882639319742421e-05,
"loss": 4.2527,
"step": 71680
},
{
"epoch": 0.02,
"learning_rate": 4.881802362871742e-05,
"loss": 4.2425,
"step": 72192
},
{
"epoch": 0.02,
"learning_rate": 4.88096376812069e-05,
"loss": 4.2498,
"step": 72704
},
{
"epoch": 0.02,
"learning_rate": 4.880125173369638e-05,
"loss": 4.2415,
"step": 73216
},
{
"epoch": 0.02,
"learning_rate": 4.879286578618586e-05,
"loss": 4.2428,
"step": 73728
},
{
"epoch": 0.02,
"learning_rate": 4.878449621747907e-05,
"loss": 4.2393,
"step": 74240
},
{
"epoch": 0.02,
"learning_rate": 4.877611026996855e-05,
"loss": 4.2362,
"step": 74752
},
{
"epoch": 0.02,
"learning_rate": 4.876772432245803e-05,
"loss": 4.2407,
"step": 75264
},
{
"epoch": 0.02,
"learning_rate": 4.875933837494751e-05,
"loss": 4.2284,
"step": 75776
},
{
"epoch": 0.02,
"learning_rate": 4.8750952427436986e-05,
"loss": 4.2257,
"step": 76288
},
{
"epoch": 0.03,
"eval_loss": 4.201012134552002,
"eval_runtime": 303.7486,
"eval_samples_per_second": 1256.272,
"eval_steps_per_second": 39.259,
"step": 76320
},
{
"epoch": 1.0,
"learning_rate": 4.87425828587302e-05,
"loss": 4.2093,
"step": 76800
},
{
"epoch": 1.0,
"learning_rate": 4.873419691121968e-05,
"loss": 4.212,
"step": 77312
},
{
"epoch": 1.0,
"learning_rate": 4.872581096370916e-05,
"loss": 4.2394,
"step": 77824
},
{
"epoch": 1.0,
"learning_rate": 4.8717425016198635e-05,
"loss": 4.2251,
"step": 78336
},
{
"epoch": 1.0,
"learning_rate": 4.8709039068688115e-05,
"loss": 4.2166,
"step": 78848
},
{
"epoch": 1.0,
"learning_rate": 4.8700653121177595e-05,
"loss": 4.213,
"step": 79360
},
{
"epoch": 1.0,
"learning_rate": 4.8692267173667075e-05,
"loss": 4.2111,
"step": 79872
},
{
"epoch": 1.0,
"learning_rate": 4.868389760496029e-05,
"loss": 4.1995,
"step": 80384
},
{
"epoch": 1.0,
"learning_rate": 4.867551165744977e-05,
"loss": 4.2115,
"step": 80896
},
{
"epoch": 1.0,
"learning_rate": 4.866712570993925e-05,
"loss": 4.2041,
"step": 81408
},
{
"epoch": 1.0,
"learning_rate": 4.865873976242873e-05,
"loss": 4.2084,
"step": 81920
},
{
"epoch": 1.0,
"learning_rate": 4.865035381491821e-05,
"loss": 4.2099,
"step": 82432
},
{
"epoch": 1.0,
"learning_rate": 4.8642000625015155e-05,
"loss": 4.1943,
"step": 82944
},
{
"epoch": 1.0,
"learning_rate": 4.8633614677504635e-05,
"loss": 4.1905,
"step": 83456
},
{
"epoch": 1.0,
"learning_rate": 4.862522872999411e-05,
"loss": 4.1824,
"step": 83968
},
{
"epoch": 1.0,
"learning_rate": 4.861684278248359e-05,
"loss": 4.1817,
"step": 84480
},
{
"epoch": 1.0,
"learning_rate": 4.860845683497307e-05,
"loss": 4.1913,
"step": 84992
},
{
"epoch": 1.0,
"learning_rate": 4.860007088746255e-05,
"loss": 4.1798,
"step": 85504
},
{
"epoch": 1.0,
"learning_rate": 4.859168493995203e-05,
"loss": 4.1845,
"step": 86016
},
{
"epoch": 1.0,
"learning_rate": 4.858329899244151e-05,
"loss": 4.2032,
"step": 86528
},
{
"epoch": 1.0,
"learning_rate": 4.8574929423734724e-05,
"loss": 4.1847,
"step": 87040
},
{
"epoch": 1.0,
"learning_rate": 4.8566543476224204e-05,
"loss": 4.1848,
"step": 87552
},
{
"epoch": 1.0,
"learning_rate": 4.8558157528713684e-05,
"loss": 4.1765,
"step": 88064
},
{
"epoch": 1.0,
"learning_rate": 4.8549771581203164e-05,
"loss": 4.1899,
"step": 88576
},
{
"epoch": 1.0,
"learning_rate": 4.8541385633692644e-05,
"loss": 4.1686,
"step": 89088
},
{
"epoch": 1.0,
"learning_rate": 4.8532999686182124e-05,
"loss": 4.1726,
"step": 89600
},
{
"epoch": 1.0,
"learning_rate": 4.852463011747533e-05,
"loss": 4.1645,
"step": 90112
},
{
"epoch": 1.0,
"learning_rate": 4.851624416996481e-05,
"loss": 4.1692,
"step": 90624
},
{
"epoch": 1.0,
"learning_rate": 4.850785822245429e-05,
"loss": 4.1626,
"step": 91136
},
{
"epoch": 1.01,
"learning_rate": 4.849947227494377e-05,
"loss": 4.1589,
"step": 91648
},
{
"epoch": 1.01,
"learning_rate": 4.849110270623698e-05,
"loss": 4.1634,
"step": 92160
},
{
"epoch": 1.01,
"learning_rate": 4.848271675872646e-05,
"loss": 4.1716,
"step": 92672
},
{
"epoch": 1.01,
"learning_rate": 4.847433081121595e-05,
"loss": 4.161,
"step": 93184
},
{
"epoch": 1.01,
"learning_rate": 4.846594486370543e-05,
"loss": 4.1628,
"step": 93696
},
{
"epoch": 1.01,
"learning_rate": 4.845757529499864e-05,
"loss": 4.1566,
"step": 94208
},
{
"epoch": 1.01,
"learning_rate": 4.844918934748812e-05,
"loss": 4.1642,
"step": 94720
},
{
"epoch": 1.01,
"learning_rate": 4.84408033999776e-05,
"loss": 4.1399,
"step": 95232
},
{
"epoch": 1.01,
"learning_rate": 4.843241745246708e-05,
"loss": 4.1549,
"step": 95744
},
{
"epoch": 1.01,
"learning_rate": 4.842404788376029e-05,
"loss": 4.1328,
"step": 96256
},
{
"epoch": 1.01,
"learning_rate": 4.841566193624977e-05,
"loss": 4.154,
"step": 96768
},
{
"epoch": 1.01,
"learning_rate": 4.840727598873925e-05,
"loss": 4.1379,
"step": 97280
},
{
"epoch": 1.01,
"learning_rate": 4.839889004122873e-05,
"loss": 4.1512,
"step": 97792
},
{
"epoch": 1.01,
"learning_rate": 4.8390520472521936e-05,
"loss": 4.1336,
"step": 98304
},
{
"epoch": 1.01,
"learning_rate": 4.8382134525011416e-05,
"loss": 4.1464,
"step": 98816
},
{
"epoch": 1.01,
"learning_rate": 4.83737485775009e-05,
"loss": 4.145,
"step": 99328
},
{
"epoch": 1.01,
"learning_rate": 4.836536262999038e-05,
"loss": 4.1423,
"step": 99840
},
{
"epoch": 1.01,
"learning_rate": 4.835699306128359e-05,
"loss": 4.148,
"step": 100352
},
{
"epoch": 1.01,
"learning_rate": 4.834860711377307e-05,
"loss": 4.1187,
"step": 100864
},
{
"epoch": 1.01,
"learning_rate": 4.834022116626255e-05,
"loss": 4.1267,
"step": 101376
},
{
"epoch": 1.01,
"learning_rate": 4.833183521875203e-05,
"loss": 4.1385,
"step": 101888
},
{
"epoch": 1.01,
"learning_rate": 4.832346565004524e-05,
"loss": 4.1349,
"step": 102400
},
{
"epoch": 1.01,
"learning_rate": 4.831507970253472e-05,
"loss": 4.1285,
"step": 102912
},
{
"epoch": 1.01,
"learning_rate": 4.83066937550242e-05,
"loss": 4.1231,
"step": 103424
},
{
"epoch": 1.01,
"learning_rate": 4.829830780751368e-05,
"loss": 4.1193,
"step": 103936
},
{
"epoch": 1.01,
"learning_rate": 4.828992186000316e-05,
"loss": 4.1113,
"step": 104448
},
{
"epoch": 1.01,
"learning_rate": 4.828155229129637e-05,
"loss": 4.1263,
"step": 104960
},
{
"epoch": 1.01,
"learning_rate": 4.8273166343785856e-05,
"loss": 4.1052,
"step": 105472
},
{
"epoch": 1.01,
"learning_rate": 4.8264780396275336e-05,
"loss": 4.1174,
"step": 105984
},
{
"epoch": 1.01,
"learning_rate": 4.8256394448764816e-05,
"loss": 4.127,
"step": 106496
},
{
"epoch": 1.01,
"learning_rate": 4.8248024880058025e-05,
"loss": 4.1069,
"step": 107008
},
{
"epoch": 1.01,
"learning_rate": 4.8239638932547505e-05,
"loss": 4.1089,
"step": 107520
},
{
"epoch": 1.01,
"learning_rate": 4.8231252985036985e-05,
"loss": 4.1078,
"step": 108032
},
{
"epoch": 1.01,
"learning_rate": 4.8222867037526465e-05,
"loss": 4.0985,
"step": 108544
},
{
"epoch": 1.01,
"learning_rate": 4.8214497468819674e-05,
"loss": 4.0997,
"step": 109056
},
{
"epoch": 1.01,
"learning_rate": 4.8206111521309154e-05,
"loss": 4.1168,
"step": 109568
},
{
"epoch": 1.01,
"learning_rate": 4.8197725573798634e-05,
"loss": 4.1009,
"step": 110080
},
{
"epoch": 1.01,
"learning_rate": 4.8189339626288114e-05,
"loss": 4.0966,
"step": 110592
},
{
"epoch": 1.01,
"learning_rate": 4.818097005758132e-05,
"loss": 4.0955,
"step": 111104
},
{
"epoch": 1.01,
"learning_rate": 4.817260048887454e-05,
"loss": 4.0962,
"step": 111616
},
{
"epoch": 1.01,
"learning_rate": 4.816421454136402e-05,
"loss": 4.1103,
"step": 112128
},
{
"epoch": 1.01,
"learning_rate": 4.81558285938535e-05,
"loss": 4.1037,
"step": 112640
},
{
"epoch": 1.01,
"learning_rate": 4.814744264634298e-05,
"loss": 4.0957,
"step": 113152
},
{
"epoch": 1.01,
"learning_rate": 4.813905669883246e-05,
"loss": 4.1041,
"step": 113664
},
{
"epoch": 1.01,
"learning_rate": 4.813067075132194e-05,
"loss": 4.1063,
"step": 114176
},
{
"epoch": 1.01,
"learning_rate": 4.812228480381141e-05,
"loss": 4.0869,
"step": 114688
},
{
"epoch": 1.01,
"learning_rate": 4.811391523510463e-05,
"loss": 4.0955,
"step": 115200
},
{
"epoch": 1.01,
"learning_rate": 4.810552928759411e-05,
"loss": 4.0929,
"step": 115712
},
{
"epoch": 1.01,
"learning_rate": 4.809714334008359e-05,
"loss": 4.09,
"step": 116224
},
{
"epoch": 1.01,
"learning_rate": 4.808875739257306e-05,
"loss": 4.0846,
"step": 116736
},
{
"epoch": 1.01,
"learning_rate": 4.808037144506255e-05,
"loss": 4.0879,
"step": 117248
},
{
"epoch": 1.01,
"learning_rate": 4.807198549755203e-05,
"loss": 4.0966,
"step": 117760
},
{
"epoch": 1.01,
"learning_rate": 4.806359955004151e-05,
"loss": 4.0952,
"step": 118272
},
{
"epoch": 1.01,
"learning_rate": 4.805521360253099e-05,
"loss": 4.0852,
"step": 118784
},
{
"epoch": 1.01,
"learning_rate": 4.8046844033824197e-05,
"loss": 4.0617,
"step": 119296
},
{
"epoch": 1.01,
"learning_rate": 4.8038458086313677e-05,
"loss": 4.0779,
"step": 119808
},
{
"epoch": 1.01,
"learning_rate": 4.8030072138803156e-05,
"loss": 4.084,
"step": 120320
},
{
"epoch": 1.01,
"learning_rate": 4.8021686191292636e-05,
"loss": 4.0829,
"step": 120832
},
{
"epoch": 1.01,
"learning_rate": 4.8013316622585846e-05,
"loss": 4.0779,
"step": 121344
},
{
"epoch": 1.01,
"learning_rate": 4.8004930675075325e-05,
"loss": 4.0743,
"step": 121856
},
{
"epoch": 1.02,
"learning_rate": 4.7996561106368535e-05,
"loss": 4.0822,
"step": 122368
},
{
"epoch": 1.02,
"learning_rate": 4.7988175158858015e-05,
"loss": 4.0728,
"step": 122880
},
{
"epoch": 1.02,
"learning_rate": 4.7979789211347494e-05,
"loss": 4.0667,
"step": 123392
},
{
"epoch": 1.02,
"learning_rate": 4.797140326383698e-05,
"loss": 4.0812,
"step": 123904
},
{
"epoch": 1.02,
"learning_rate": 4.796301731632646e-05,
"loss": 4.0629,
"step": 124416
},
{
"epoch": 1.02,
"learning_rate": 4.795464774761967e-05,
"loss": 4.0842,
"step": 124928
},
{
"epoch": 1.02,
"learning_rate": 4.794626180010915e-05,
"loss": 4.0749,
"step": 125440
},
{
"epoch": 1.02,
"learning_rate": 4.793787585259863e-05,
"loss": 4.0538,
"step": 125952
},
{
"epoch": 1.02,
"learning_rate": 4.792948990508811e-05,
"loss": 4.0688,
"step": 126464
},
{
"epoch": 1.02,
"learning_rate": 4.792110395757759e-05,
"loss": 4.066,
"step": 126976
},
{
"epoch": 1.02,
"learning_rate": 4.791271801006707e-05,
"loss": 4.0773,
"step": 127488
},
{
"epoch": 1.02,
"learning_rate": 4.790433206255655e-05,
"loss": 4.066,
"step": 128000
},
{
"epoch": 1.02,
"learning_rate": 4.789596249384976e-05,
"loss": 4.0617,
"step": 128512
},
{
"epoch": 1.02,
"learning_rate": 4.788757654633924e-05,
"loss": 4.0532,
"step": 129024
},
{
"epoch": 1.02,
"learning_rate": 4.787919059882872e-05,
"loss": 4.0692,
"step": 129536
},
{
"epoch": 1.02,
"learning_rate": 4.78708046513182e-05,
"loss": 4.0462,
"step": 130048
},
{
"epoch": 1.02,
"learning_rate": 4.7862435082611415e-05,
"loss": 4.0551,
"step": 130560
},
{
"epoch": 1.02,
"learning_rate": 4.7854049135100895e-05,
"loss": 4.0587,
"step": 131072
},
{
"epoch": 1.02,
"learning_rate": 4.7845663187590375e-05,
"loss": 4.0566,
"step": 131584
},
{
"epoch": 1.02,
"learning_rate": 4.7837277240079855e-05,
"loss": 4.0457,
"step": 132096
},
{
"epoch": 1.02,
"learning_rate": 4.7828891292569335e-05,
"loss": 4.0567,
"step": 132608
},
{
"epoch": 1.02,
"learning_rate": 4.7820505345058815e-05,
"loss": 4.0501,
"step": 133120
},
{
"epoch": 1.02,
"learning_rate": 4.7812119397548295e-05,
"loss": 4.0542,
"step": 133632
},
{
"epoch": 1.02,
"learning_rate": 4.7803733450037775e-05,
"loss": 4.0505,
"step": 134144
},
{
"epoch": 1.02,
"learning_rate": 4.7795363881330984e-05,
"loss": 4.045,
"step": 134656
},
{
"epoch": 1.02,
"learning_rate": 4.7786977933820464e-05,
"loss": 4.0711,
"step": 135168
},
{
"epoch": 1.02,
"learning_rate": 4.7778591986309944e-05,
"loss": 4.0561,
"step": 135680
},
{
"epoch": 1.02,
"learning_rate": 4.777022241760315e-05,
"loss": 4.0541,
"step": 136192
},
{
"epoch": 1.02,
"learning_rate": 4.776183647009263e-05,
"loss": 4.0402,
"step": 136704
},
{
"epoch": 1.02,
"learning_rate": 4.775345052258212e-05,
"loss": 4.0471,
"step": 137216
},
{
"epoch": 1.02,
"learning_rate": 4.77450645750716e-05,
"loss": 4.0475,
"step": 137728
},
{
"epoch": 1.02,
"learning_rate": 4.773667862756108e-05,
"loss": 4.0393,
"step": 138240
},
{
"epoch": 1.02,
"learning_rate": 4.772829268005055e-05,
"loss": 4.056,
"step": 138752
},
{
"epoch": 1.02,
"learning_rate": 4.771990673254003e-05,
"loss": 4.0438,
"step": 139264
},
{
"epoch": 1.02,
"learning_rate": 4.771152078502951e-05,
"loss": 4.0422,
"step": 139776
},
{
"epoch": 1.02,
"learning_rate": 4.770315121632272e-05,
"loss": 4.0379,
"step": 140288
},
{
"epoch": 1.02,
"learning_rate": 4.76947652688122e-05,
"loss": 4.038,
"step": 140800
},
{
"epoch": 1.02,
"learning_rate": 4.768637932130168e-05,
"loss": 4.0406,
"step": 141312
},
{
"epoch": 1.02,
"learning_rate": 4.76780097525949e-05,
"loss": 4.0395,
"step": 141824
},
{
"epoch": 1.02,
"learning_rate": 4.766962380508437e-05,
"loss": 4.028,
"step": 142336
},
{
"epoch": 1.02,
"learning_rate": 4.766123785757386e-05,
"loss": 4.0562,
"step": 142848
},
{
"epoch": 1.02,
"learning_rate": 4.765285191006334e-05,
"loss": 4.0437,
"step": 143360
},
{
"epoch": 1.02,
"learning_rate": 4.764448234135655e-05,
"loss": 4.0334,
"step": 143872
},
{
"epoch": 1.02,
"learning_rate": 4.7636096393846026e-05,
"loss": 4.0327,
"step": 144384
},
{
"epoch": 1.02,
"learning_rate": 4.7627710446335506e-05,
"loss": 4.0304,
"step": 144896
},
{
"epoch": 1.02,
"learning_rate": 4.7619324498824986e-05,
"loss": 4.0427,
"step": 145408
},
{
"epoch": 1.02,
"learning_rate": 4.7610938551314466e-05,
"loss": 4.0376,
"step": 145920
},
{
"epoch": 1.02,
"learning_rate": 4.7602552603803946e-05,
"loss": 4.0331,
"step": 146432
},
{
"epoch": 1.02,
"learning_rate": 4.7594166656293426e-05,
"loss": 4.0437,
"step": 146944
},
{
"epoch": 1.02,
"learning_rate": 4.7585780708782906e-05,
"loss": 4.035,
"step": 147456
},
{
"epoch": 1.02,
"learning_rate": 4.7577411140076115e-05,
"loss": 4.0394,
"step": 147968
},
{
"epoch": 1.02,
"learning_rate": 4.7569025192565595e-05,
"loss": 4.028,
"step": 148480
},
{
"epoch": 1.02,
"learning_rate": 4.7560639245055075e-05,
"loss": 4.0362,
"step": 148992
},
{
"epoch": 1.02,
"learning_rate": 4.7552253297544555e-05,
"loss": 4.03,
"step": 149504
},
{
"epoch": 1.02,
"learning_rate": 4.754388372883777e-05,
"loss": 4.0278,
"step": 150016
},
{
"epoch": 1.02,
"learning_rate": 4.753549778132725e-05,
"loss": 4.0289,
"step": 150528
},
{
"epoch": 1.02,
"learning_rate": 4.752711183381673e-05,
"loss": 4.0292,
"step": 151040
},
{
"epoch": 1.02,
"learning_rate": 4.751872588630621e-05,
"loss": 4.0313,
"step": 151552
},
{
"epoch": 1.02,
"learning_rate": 4.751035631759942e-05,
"loss": 4.0182,
"step": 152064
},
{
"epoch": 1.02,
"learning_rate": 4.75019703700889e-05,
"loss": 4.0198,
"step": 152576
},
{
"epoch": 1.03,
"eval_loss": 4.033823490142822,
"eval_runtime": 303.4895,
"eval_samples_per_second": 1257.345,
"eval_steps_per_second": 39.293,
"step": 152640
},
{
"epoch": 0.0,
"learning_rate": 4.749358442257838e-05,
"loss": 4.0055,
"step": 153088
},
{
"epoch": 0.0,
"learning_rate": 4.748519847506786e-05,
"loss": 4.0083,
"step": 153600
},
{
"epoch": 0.0,
"learning_rate": 4.747681252755734e-05,
"loss": 4.036,
"step": 154112
},
{
"epoch": 0.0,
"learning_rate": 4.746842658004682e-05,
"loss": 4.0282,
"step": 154624
},
{
"epoch": 0.0,
"learning_rate": 4.74600406325363e-05,
"loss": 4.0193,
"step": 155136
},
{
"epoch": 0.0,
"learning_rate": 4.745165468502578e-05,
"loss": 4.0142,
"step": 155648
},
{
"epoch": 0.0,
"learning_rate": 4.744326873751526e-05,
"loss": 4.016,
"step": 156160
},
{
"epoch": 0.0,
"learning_rate": 4.743488279000474e-05,
"loss": 4.0034,
"step": 156672
},
{
"epoch": 0.0,
"learning_rate": 4.742649684249422e-05,
"loss": 4.0211,
"step": 157184
},
{
"epoch": 0.0,
"learning_rate": 4.74181108949837e-05,
"loss": 4.014,
"step": 157696
},
{
"epoch": 0.0,
"learning_rate": 4.740974132627691e-05,
"loss": 4.0138,
"step": 158208
},
{
"epoch": 0.0,
"learning_rate": 4.740135537876639e-05,
"loss": 4.0163,
"step": 158720
},
{
"epoch": 0.0,
"learning_rate": 4.739296943125587e-05,
"loss": 4.0035,
"step": 159232
},
{
"epoch": 0.0,
"learning_rate": 4.738458348374535e-05,
"loss": 4.0043,
"step": 159744
},
{
"epoch": 0.0,
"learning_rate": 4.737619753623483e-05,
"loss": 3.9992,
"step": 160256
},
{
"epoch": 0.0,
"learning_rate": 4.736782796752804e-05,
"loss": 3.9966,
"step": 160768
},
{
"epoch": 0.0,
"learning_rate": 4.735944202001752e-05,
"loss": 4.0025,
"step": 161280
},
{
"epoch": 0.0,
"learning_rate": 4.7351056072507e-05,
"loss": 3.9968,
"step": 161792
},
{
"epoch": 0.0,
"learning_rate": 4.734267012499648e-05,
"loss": 4.0015,
"step": 162304
},
{
"epoch": 0.0,
"learning_rate": 4.733428417748596e-05,
"loss": 4.026,
"step": 162816
},
{
"epoch": 0.0,
"learning_rate": 4.7325898229975444e-05,
"loss": 4.0016,
"step": 163328
},
{
"epoch": 0.0,
"learning_rate": 4.731752866126865e-05,
"loss": 4.0044,
"step": 163840
},
{
"epoch": 0.0,
"learning_rate": 4.730914271375813e-05,
"loss": 4.0024,
"step": 164352
},
{
"epoch": 0.0,
"learning_rate": 4.730075676624761e-05,
"loss": 4.0096,
"step": 164864
},
{
"epoch": 0.0,
"learning_rate": 4.729237081873709e-05,
"loss": 3.9884,
"step": 165376
},
{
"epoch": 0.0,
"learning_rate": 4.728398487122657e-05,
"loss": 4.0,
"step": 165888
},
{
"epoch": 0.0,
"learning_rate": 4.727559892371605e-05,
"loss": 3.9917,
"step": 166400
},
{
"epoch": 0.0,
"learning_rate": 4.726721297620553e-05,
"loss": 3.9923,
"step": 166912
},
{
"epoch": 0.0,
"learning_rate": 4.725882702869501e-05,
"loss": 3.9871,
"step": 167424
},
{
"epoch": 0.01,
"learning_rate": 4.725045745998822e-05,
"loss": 3.9856,
"step": 167936
},
{
"epoch": 0.01,
"learning_rate": 4.72420715124777e-05,
"loss": 3.9911,
"step": 168448
},
{
"epoch": 0.01,
"learning_rate": 4.723370194377091e-05,
"loss": 4.0011,
"step": 168960
},
{
"epoch": 0.01,
"learning_rate": 4.72253159962604e-05,
"loss": 3.9922,
"step": 169472
},
{
"epoch": 0.01,
"learning_rate": 4.721693004874988e-05,
"loss": 3.9983,
"step": 169984
},
{
"epoch": 0.01,
"learning_rate": 4.720854410123936e-05,
"loss": 3.9828,
"step": 170496
},
{
"epoch": 0.01,
"learning_rate": 4.720015815372884e-05,
"loss": 4.0005,
"step": 171008
},
{
"epoch": 0.01,
"learning_rate": 4.7191788585022046e-05,
"loss": 3.9725,
"step": 171520
},
{
"epoch": 0.01,
"learning_rate": 4.7183402637511526e-05,
"loss": 3.9922,
"step": 172032
},
{
"epoch": 0.01,
"learning_rate": 4.7175016690001006e-05,
"loss": 3.9683,
"step": 172544
},
{
"epoch": 0.01,
"learning_rate": 4.7166630742490486e-05,
"loss": 3.9884,
"step": 173056
},
{
"epoch": 0.01,
"learning_rate": 4.7158244794979966e-05,
"loss": 3.9749,
"step": 173568
},
{
"epoch": 0.01,
"learning_rate": 4.7149858847469446e-05,
"loss": 3.9912,
"step": 174080
},
{
"epoch": 0.01,
"learning_rate": 4.714147289995892e-05,
"loss": 3.9764,
"step": 174592
},
{
"epoch": 0.01,
"learning_rate": 4.7133103331252135e-05,
"loss": 3.9818,
"step": 175104
},
{
"epoch": 0.01,
"learning_rate": 4.7124717383741615e-05,
"loss": 3.9854,
"step": 175616
},
{
"epoch": 0.01,
"learning_rate": 4.7116331436231095e-05,
"loss": 3.9864,
"step": 176128
},
{
"epoch": 0.01,
"learning_rate": 4.7107945488720575e-05,
"loss": 3.9847,
"step": 176640
},
{
"epoch": 0.01,
"learning_rate": 4.7099559541210055e-05,
"loss": 3.9662,
"step": 177152
},
{
"epoch": 0.01,
"learning_rate": 4.709118997250327e-05,
"loss": 3.9665,
"step": 177664
},
{
"epoch": 0.01,
"learning_rate": 4.7082804024992744e-05,
"loss": 3.9839,
"step": 178176
},
{
"epoch": 0.01,
"learning_rate": 4.7074418077482224e-05,
"loss": 3.9804,
"step": 178688
},
{
"epoch": 0.01,
"learning_rate": 4.7066032129971704e-05,
"loss": 3.9737,
"step": 179200
},
{
"epoch": 0.01,
"learning_rate": 4.7057646182461184e-05,
"loss": 3.9727,
"step": 179712
},
{
"epoch": 0.01,
"learning_rate": 4.7049260234950664e-05,
"loss": 3.9693,
"step": 180224
},
{
"epoch": 0.01,
"learning_rate": 4.704089066624387e-05,
"loss": 3.9563,
"step": 180736
},
{
"epoch": 0.01,
"learning_rate": 4.703250471873335e-05,
"loss": 3.9724,
"step": 181248
},
{
"epoch": 0.01,
"learning_rate": 4.702411877122283e-05,
"loss": 3.956,
"step": 181760
},
{
"epoch": 0.01,
"learning_rate": 4.701574920251605e-05,
"loss": 3.9698,
"step": 182272
},
{
"epoch": 0.01,
"learning_rate": 4.700736325500553e-05,
"loss": 3.9783,
"step": 182784
},
{
"epoch": 0.01,
"learning_rate": 4.699897730749501e-05,
"loss": 3.9577,
"step": 183296
},
{
"epoch": 0.01,
"learning_rate": 4.699059135998449e-05,
"loss": 3.9611,
"step": 183808
},
{
"epoch": 0.01,
"learning_rate": 4.69822217912777e-05,
"loss": 3.9615,
"step": 184320
},
{
"epoch": 0.01,
"learning_rate": 4.697383584376718e-05,
"loss": 3.9515,
"step": 184832
},
{
"epoch": 0.01,
"learning_rate": 4.696544989625666e-05,
"loss": 3.9501,
"step": 185344
},
{
"epoch": 0.01,
"learning_rate": 4.695706394874614e-05,
"loss": 3.972,
"step": 185856
},
{
"epoch": 0.01,
"learning_rate": 4.694867800123562e-05,
"loss": 3.9586,
"step": 186368
},
{
"epoch": 0.01,
"learning_rate": 4.69402920537251e-05,
"loss": 3.9531,
"step": 186880
},
{
"epoch": 0.01,
"learning_rate": 4.6931922485018307e-05,
"loss": 3.952,
"step": 187392
},
{
"epoch": 0.01,
"learning_rate": 4.6923536537507787e-05,
"loss": 3.9552,
"step": 187904
},
{
"epoch": 0.01,
"learning_rate": 4.6915150589997267e-05,
"loss": 3.965,
"step": 188416
},
{
"epoch": 0.01,
"learning_rate": 4.690676464248675e-05,
"loss": 3.9638,
"step": 188928
},
{
"epoch": 0.01,
"learning_rate": 4.689837869497623e-05,
"loss": 3.9522,
"step": 189440
},
{
"epoch": 0.01,
"learning_rate": 4.689000912626944e-05,
"loss": 3.9646,
"step": 189952
},
{
"epoch": 0.01,
"learning_rate": 4.688162317875892e-05,
"loss": 3.9616,
"step": 190464
},
{
"epoch": 0.01,
"learning_rate": 4.68732372312484e-05,
"loss": 3.9502,
"step": 190976
},
{
"epoch": 0.01,
"learning_rate": 4.686485128373788e-05,
"loss": 3.9564,
"step": 191488
},
{
"epoch": 0.01,
"learning_rate": 4.685648171503109e-05,
"loss": 3.9596,
"step": 192000
},
{
"epoch": 0.01,
"learning_rate": 4.684809576752057e-05,
"loss": 3.952,
"step": 192512
},
{
"epoch": 0.01,
"learning_rate": 4.683970982001005e-05,
"loss": 3.9489,
"step": 193024
},
{
"epoch": 0.01,
"learning_rate": 4.683132387249953e-05,
"loss": 3.9504,
"step": 193536
},
{
"epoch": 0.01,
"learning_rate": 4.682293792498901e-05,
"loss": 3.9592,
"step": 194048
},
{
"epoch": 0.01,
"learning_rate": 4.681456835628222e-05,
"loss": 3.9611,
"step": 194560
},
{
"epoch": 0.01,
"learning_rate": 4.680618240877171e-05,
"loss": 3.9456,
"step": 195072
},
{
"epoch": 0.01,
"learning_rate": 4.679779646126119e-05,
"loss": 3.9334,
"step": 195584
},
{
"epoch": 0.01,
"learning_rate": 4.678941051375067e-05,
"loss": 3.9395,
"step": 196096
},
{
"epoch": 0.01,
"learning_rate": 4.678102456624015e-05,
"loss": 3.9484,
"step": 196608
},
{
"epoch": 0.01,
"learning_rate": 4.6772654997533356e-05,
"loss": 3.9513,
"step": 197120
},
{
"epoch": 0.01,
"learning_rate": 4.6764269050022836e-05,
"loss": 3.9432,
"step": 197632
},
{
"epoch": 0.01,
"learning_rate": 4.6755883102512316e-05,
"loss": 3.9409,
"step": 198144
},
{
"epoch": 0.02,
"learning_rate": 4.6747497155001796e-05,
"loss": 3.955,
"step": 198656
},
{
"epoch": 0.02,
"learning_rate": 4.6739127586295005e-05,
"loss": 3.9381,
"step": 199168
},
{
"epoch": 0.02,
"learning_rate": 4.6730741638784485e-05,
"loss": 3.9386,
"step": 199680
},
{
"epoch": 0.02,
"learning_rate": 4.6722355691273965e-05,
"loss": 3.9489,
"step": 200192
},
{
"epoch": 0.02,
"learning_rate": 4.6713969743763445e-05,
"loss": 3.9368,
"step": 200704
},
{
"epoch": 0.02,
"learning_rate": 4.670560017505666e-05,
"loss": 3.9556,
"step": 201216
},
{
"epoch": 0.02,
"learning_rate": 4.669721422754614e-05,
"loss": 3.9436,
"step": 201728
},
{
"epoch": 0.02,
"learning_rate": 4.668882828003562e-05,
"loss": 3.9272,
"step": 202240
},
{
"epoch": 0.02,
"learning_rate": 4.66804423325251e-05,
"loss": 3.9413,
"step": 202752
},
{
"epoch": 0.02,
"learning_rate": 4.667205638501458e-05,
"loss": 3.9366,
"step": 203264
},
{
"epoch": 0.02,
"learning_rate": 4.6663670437504054e-05,
"loss": 3.9531,
"step": 203776
},
{
"epoch": 0.02,
"learning_rate": 4.6655284489993534e-05,
"loss": 3.9426,
"step": 204288
},
{
"epoch": 0.02,
"learning_rate": 4.6646898542483013e-05,
"loss": 3.9337,
"step": 204800
},
{
"epoch": 0.02,
"learning_rate": 4.663852897377623e-05,
"loss": 3.9248,
"step": 205312
},
{
"epoch": 0.02,
"learning_rate": 4.66301430262657e-05,
"loss": 3.9436,
"step": 205824
},
{
"epoch": 0.02,
"learning_rate": 4.662175707875518e-05,
"loss": 3.9237,
"step": 206336
},
{
"epoch": 0.02,
"learning_rate": 4.661337113124466e-05,
"loss": 3.9341,
"step": 206848
},
{
"epoch": 0.02,
"learning_rate": 4.660500156253788e-05,
"loss": 3.9346,
"step": 207360
},
{
"epoch": 0.02,
"learning_rate": 4.659661561502736e-05,
"loss": 3.933,
"step": 207872
},
{
"epoch": 0.02,
"learning_rate": 4.658822966751684e-05,
"loss": 3.922,
"step": 208384
},
{
"epoch": 0.02,
"learning_rate": 4.657984372000632e-05,
"loss": 3.9415,
"step": 208896
},
{
"epoch": 0.02,
"learning_rate": 4.657147415129953e-05,
"loss": 3.9261,
"step": 209408
},
{
"epoch": 0.02,
"learning_rate": 4.656308820378901e-05,
"loss": 3.9305,
"step": 209920
},
{
"epoch": 0.02,
"learning_rate": 4.655470225627849e-05,
"loss": 3.9348,
"step": 210432
},
{
"epoch": 0.02,
"learning_rate": 4.654631630876797e-05,
"loss": 3.9262,
"step": 210944
},
{
"epoch": 0.02,
"learning_rate": 4.6537946740061176e-05,
"loss": 3.9471,
"step": 211456
},
{
"epoch": 0.02,
"learning_rate": 4.6529560792550656e-05,
"loss": 3.9367,
"step": 211968
},
{
"epoch": 0.02,
"learning_rate": 4.6521174845040136e-05,
"loss": 3.9348,
"step": 212480
},
{
"epoch": 0.02,
"learning_rate": 4.6512788897529616e-05,
"loss": 3.9233,
"step": 212992
},
{
"epoch": 0.02,
"learning_rate": 4.6504402950019096e-05,
"loss": 3.9292,
"step": 213504
},
{
"epoch": 0.02,
"learning_rate": 4.649603338131231e-05,
"loss": 3.9314,
"step": 214016
},
{
"epoch": 0.02,
"learning_rate": 4.648764743380179e-05,
"loss": 3.9184,
"step": 214528
},
{
"epoch": 0.02,
"learning_rate": 4.647926148629127e-05,
"loss": 3.9345,
"step": 215040
},
{
"epoch": 0.02,
"learning_rate": 4.647087553878075e-05,
"loss": 3.9289,
"step": 215552
},
{
"epoch": 0.02,
"learning_rate": 4.646250597007396e-05,
"loss": 3.9263,
"step": 216064
},
{
"epoch": 0.02,
"learning_rate": 4.645412002256344e-05,
"loss": 3.9265,
"step": 216576
},
{
"epoch": 0.02,
"learning_rate": 4.644573407505292e-05,
"loss": 3.9247,
"step": 217088
},
{
"epoch": 0.02,
"learning_rate": 4.64373481275424e-05,
"loss": 3.9177,
"step": 217600
},
{
"epoch": 0.02,
"learning_rate": 4.642897855883561e-05,
"loss": 3.9297,
"step": 218112
},
{
"epoch": 0.02,
"learning_rate": 4.642059261132509e-05,
"loss": 3.9158,
"step": 218624
},
{
"epoch": 0.02,
"learning_rate": 4.641220666381457e-05,
"loss": 3.9389,
"step": 219136
},
{
"epoch": 0.02,
"learning_rate": 4.640382071630405e-05,
"loss": 3.9346,
"step": 219648
},
{
"epoch": 0.02,
"learning_rate": 4.6395451147597266e-05,
"loss": 3.9187,
"step": 220160
},
{
"epoch": 0.02,
"learning_rate": 4.6387065200086746e-05,
"loss": 3.9192,
"step": 220672
},
{
"epoch": 0.02,
"learning_rate": 4.6378679252576226e-05,
"loss": 3.9207,
"step": 221184
},
{
"epoch": 0.02,
"learning_rate": 4.6370293305065705e-05,
"loss": 3.9275,
"step": 221696
},
{
"epoch": 0.02,
"learning_rate": 4.6361907357555185e-05,
"loss": 3.9238,
"step": 222208
},
{
"epoch": 0.02,
"learning_rate": 4.6353537788848395e-05,
"loss": 3.9199,
"step": 222720
},
{
"epoch": 0.02,
"learning_rate": 4.6345151841337874e-05,
"loss": 3.9295,
"step": 223232
},
{
"epoch": 0.02,
"learning_rate": 4.6336765893827354e-05,
"loss": 3.9239,
"step": 223744
},
{
"epoch": 0.02,
"learning_rate": 4.6328379946316834e-05,
"loss": 3.9293,
"step": 224256
},
{
"epoch": 0.02,
"learning_rate": 4.6320010377610043e-05,
"loss": 3.9155,
"step": 224768
},
{
"epoch": 0.02,
"learning_rate": 4.6311624430099523e-05,
"loss": 3.9286,
"step": 225280
},
{
"epoch": 0.02,
"learning_rate": 4.6303238482589003e-05,
"loss": 3.9208,
"step": 225792
},
{
"epoch": 0.02,
"learning_rate": 4.629485253507848e-05,
"loss": 3.9145,
"step": 226304
},
{
"epoch": 0.02,
"learning_rate": 4.62864829663717e-05,
"loss": 3.923,
"step": 226816
},
{
"epoch": 0.02,
"learning_rate": 4.627809701886118e-05,
"loss": 3.9231,
"step": 227328
},
{
"epoch": 0.02,
"learning_rate": 4.626971107135066e-05,
"loss": 3.9192,
"step": 227840
},
{
"epoch": 0.02,
"learning_rate": 4.626132512384014e-05,
"loss": 3.9073,
"step": 228352
},
{
"epoch": 0.02,
"learning_rate": 4.625295555513335e-05,
"loss": 3.9095,
"step": 228864
},
{
"epoch": 0.03,
"eval_loss": 3.958364486694336,
"eval_runtime": 308.0893,
"eval_samples_per_second": 1238.573,
"eval_steps_per_second": 38.706,
"step": 228960
},
{
"epoch": 0.0,
"learning_rate": 4.624456960762283e-05,
"loss": 3.9009,
"step": 229376
},
{
"epoch": 0.0,
"learning_rate": 4.623618366011231e-05,
"loss": 3.9006,
"step": 229888
},
{
"epoch": 0.0,
"learning_rate": 4.622779771260179e-05,
"loss": 3.9274,
"step": 230400
},
{
"epoch": 0.0,
"learning_rate": 4.621941176509127e-05,
"loss": 3.922,
"step": 230912
},
{
"epoch": 0.0,
"learning_rate": 4.621102581758075e-05,
"loss": 3.9205,
"step": 231424
},
{
"epoch": 0.0,
"learning_rate": 4.620263987007023e-05,
"loss": 3.9009,
"step": 231936
},
{
"epoch": 0.0,
"learning_rate": 4.619425392255971e-05,
"loss": 3.9157,
"step": 232448
},
{
"epoch": 0.0,
"learning_rate": 4.618586797504919e-05,
"loss": 3.8953,
"step": 232960
},
{
"epoch": 0.0,
"learning_rate": 4.6177498406342404e-05,
"loss": 3.9188,
"step": 233472
},
{
"epoch": 0.0,
"learning_rate": 4.6169112458831884e-05,
"loss": 3.9127,
"step": 233984
},
{
"epoch": 0.0,
"learning_rate": 4.6160726511321364e-05,
"loss": 3.9063,
"step": 234496
},
{
"epoch": 0.0,
"learning_rate": 4.615234056381084e-05,
"loss": 3.917,
"step": 235008
},
{
"epoch": 0.0,
"learning_rate": 4.614398737390778e-05,
"loss": 3.8986,
"step": 235520
},
{
"epoch": 0.0,
"learning_rate": 4.613560142639726e-05,
"loss": 3.9049,
"step": 236032
},
{
"epoch": 0.0,
"learning_rate": 4.612721547888674e-05,
"loss": 3.8927,
"step": 236544
},
{
"epoch": 0.0,
"learning_rate": 4.611882953137622e-05,
"loss": 3.9002,
"step": 237056
},
{
"epoch": 0.0,
"learning_rate": 4.61104435838657e-05,
"loss": 3.8979,
"step": 237568
},
{
"epoch": 0.0,
"learning_rate": 4.610205763635518e-05,
"loss": 3.8989,
"step": 238080
},
{
"epoch": 0.0,
"learning_rate": 4.609367168884466e-05,
"loss": 3.9001,
"step": 238592
},
{
"epoch": 0.0,
"learning_rate": 4.608528574133414e-05,
"loss": 3.9214,
"step": 239104
},
{
"epoch": 0.0,
"learning_rate": 4.607691617262736e-05,
"loss": 3.9005,
"step": 239616
},
{
"epoch": 0.0,
"learning_rate": 4.606853022511684e-05,
"loss": 3.9064,
"step": 240128
},
{
"epoch": 0.0,
"learning_rate": 4.606014427760631e-05,
"loss": 3.9019,
"step": 240640
},
{
"epoch": 0.0,
"learning_rate": 4.605175833009579e-05,
"loss": 3.9089,
"step": 241152
},
{
"epoch": 0.0,
"learning_rate": 4.6043388761389006e-05,
"loss": 3.8895,
"step": 241664
},
{
"epoch": 0.0,
"learning_rate": 4.603500281387848e-05,
"loss": 3.9013,
"step": 242176
},
{
"epoch": 0.0,
"learning_rate": 4.602661686636796e-05,
"loss": 3.8945,
"step": 242688
},
{
"epoch": 0.0,
"learning_rate": 4.601823091885744e-05,
"loss": 3.8923,
"step": 243200
},
{
"epoch": 0.0,
"learning_rate": 4.6009861350150655e-05,
"loss": 3.8884,
"step": 243712
},
{
"epoch": 0.01,
"learning_rate": 4.600147540264013e-05,
"loss": 3.8947,
"step": 244224
},
{
"epoch": 0.01,
"learning_rate": 4.5993089455129615e-05,
"loss": 3.8919,
"step": 244736
},
{
"epoch": 0.01,
"learning_rate": 4.598471988642283e-05,
"loss": 3.9031,
"step": 245248
},
{
"epoch": 0.01,
"learning_rate": 4.597633393891231e-05,
"loss": 3.8958,
"step": 245760
},
{
"epoch": 0.01,
"learning_rate": 4.5967947991401784e-05,
"loss": 3.8984,
"step": 246272
},
{
"epoch": 0.01,
"learning_rate": 4.5959562043891264e-05,
"loss": 3.8892,
"step": 246784
},
{
"epoch": 0.01,
"learning_rate": 4.5951176096380744e-05,
"loss": 3.9014,
"step": 247296
},
{
"epoch": 0.01,
"learning_rate": 4.5942790148870224e-05,
"loss": 3.8851,
"step": 247808
},
{
"epoch": 0.01,
"learning_rate": 4.5934404201359704e-05,
"loss": 3.894,
"step": 248320
},
{
"epoch": 0.01,
"learning_rate": 4.5926018253849184e-05,
"loss": 3.8727,
"step": 248832
},
{
"epoch": 0.01,
"learning_rate": 4.591764868514239e-05,
"loss": 3.8911,
"step": 249344
},
{
"epoch": 0.01,
"learning_rate": 4.590926273763187e-05,
"loss": 3.8851,
"step": 249856
},
{
"epoch": 0.01,
"learning_rate": 4.590087679012135e-05,
"loss": 3.9006,
"step": 250368
},
{
"epoch": 0.01,
"learning_rate": 4.589249084261083e-05,
"loss": 3.8841,
"step": 250880
},
{
"epoch": 0.01,
"learning_rate": 4.588412127390405e-05,
"loss": 3.889,
"step": 251392
},
{
"epoch": 0.01,
"learning_rate": 4.587573532639353e-05,
"loss": 3.8912,
"step": 251904
},
{
"epoch": 0.01,
"learning_rate": 4.586734937888301e-05,
"loss": 3.8971,
"step": 252416
},
{
"epoch": 0.01,
"learning_rate": 4.585896343137249e-05,
"loss": 3.8875,
"step": 252928
},
{
"epoch": 0.01,
"learning_rate": 4.58505938626657e-05,
"loss": 3.8763,
"step": 253440
},
{
"epoch": 0.01,
"learning_rate": 4.584220791515518e-05,
"loss": 3.8701,
"step": 253952
},
{
"epoch": 0.01,
"learning_rate": 4.583382196764466e-05,
"loss": 3.8915,
"step": 254464
},
{
"epoch": 0.01,
"learning_rate": 4.582543602013414e-05,
"loss": 3.8903,
"step": 254976
},
{
"epoch": 0.01,
"learning_rate": 4.581706645142735e-05,
"loss": 3.8832,
"step": 255488
},
{
"epoch": 0.01,
"learning_rate": 4.580868050391683e-05,
"loss": 3.8758,
"step": 256000
},
{
"epoch": 0.01,
"learning_rate": 4.580029455640631e-05,
"loss": 3.8817,
"step": 256512
},
{
"epoch": 0.01,
"learning_rate": 4.579190860889579e-05,
"loss": 3.8644,
"step": 257024
},
{
"epoch": 0.01,
"learning_rate": 4.5783539040189e-05,
"loss": 3.8856,
"step": 257536
},
{
"epoch": 0.01,
"learning_rate": 4.577515309267848e-05,
"loss": 3.8654,
"step": 258048
},
{
"epoch": 0.01,
"learning_rate": 4.576676714516796e-05,
"loss": 3.8802,
"step": 258560
},
{
"epoch": 0.01,
"learning_rate": 4.575838119765744e-05,
"loss": 3.8855,
"step": 259072
},
{
"epoch": 0.01,
"learning_rate": 4.575001162895065e-05,
"loss": 3.8745,
"step": 259584
},
{
"epoch": 0.01,
"learning_rate": 4.574162568144013e-05,
"loss": 3.868,
"step": 260096
},
{
"epoch": 0.01,
"learning_rate": 4.573323973392961e-05,
"loss": 3.8754,
"step": 260608
},
{
"epoch": 0.01,
"learning_rate": 4.572485378641909e-05,
"loss": 3.8644,
"step": 261120
},
{
"epoch": 0.01,
"learning_rate": 4.57164842177123e-05,
"loss": 3.8654,
"step": 261632
},
{
"epoch": 0.01,
"learning_rate": 4.570809827020178e-05,
"loss": 3.8805,
"step": 262144
},
{
"epoch": 0.01,
"learning_rate": 4.569971232269126e-05,
"loss": 3.8733,
"step": 262656
},
{
"epoch": 0.01,
"learning_rate": 4.569132637518074e-05,
"loss": 3.8654,
"step": 263168
},
{
"epoch": 0.01,
"learning_rate": 4.5682956806473956e-05,
"loss": 3.8624,
"step": 263680
},
{
"epoch": 0.01,
"learning_rate": 4.5674570858963436e-05,
"loss": 3.8703,
"step": 264192
},
{
"epoch": 0.01,
"learning_rate": 4.5666184911452916e-05,
"loss": 3.8726,
"step": 264704
},
{
"epoch": 0.01,
"learning_rate": 4.5657798963942396e-05,
"loss": 3.878,
"step": 265216
},
{
"epoch": 0.01,
"learning_rate": 4.5649429395235605e-05,
"loss": 3.87,
"step": 265728
},
{
"epoch": 0.01,
"learning_rate": 4.5641043447725085e-05,
"loss": 3.877,
"step": 266240
},
{
"epoch": 0.01,
"learning_rate": 4.5632657500214565e-05,
"loss": 3.8792,
"step": 266752
},
{
"epoch": 0.01,
"learning_rate": 4.5624271552704045e-05,
"loss": 3.8647,
"step": 267264
},
{
"epoch": 0.01,
"learning_rate": 4.5615901983997254e-05,
"loss": 3.867,
"step": 267776
},
{
"epoch": 0.01,
"learning_rate": 4.5607516036486734e-05,
"loss": 3.8757,
"step": 268288
},
{
"epoch": 0.01,
"learning_rate": 4.5599130088976214e-05,
"loss": 3.8708,
"step": 268800
},
{
"epoch": 0.01,
"learning_rate": 4.5590744141465694e-05,
"loss": 3.8653,
"step": 269312
},
{
"epoch": 0.01,
"learning_rate": 4.558237457275891e-05,
"loss": 3.8655,
"step": 269824
},
{
"epoch": 0.01,
"learning_rate": 4.557398862524839e-05,
"loss": 3.8727,
"step": 270336
},
{
"epoch": 0.01,
"learning_rate": 4.556560267773787e-05,
"loss": 3.8771,
"step": 270848
},
{
"epoch": 0.01,
"learning_rate": 4.555721673022735e-05,
"loss": 3.865,
"step": 271360
},
{
"epoch": 0.01,
"learning_rate": 4.554884716152056e-05,
"loss": 3.8519,
"step": 271872
},
{
"epoch": 0.01,
"learning_rate": 4.554046121401004e-05,
"loss": 3.8522,
"step": 272384
},
{
"epoch": 0.01,
"learning_rate": 4.553207526649952e-05,
"loss": 3.8631,
"step": 272896
},
{
"epoch": 0.01,
"learning_rate": 4.5523689318989e-05,
"loss": 3.8703,
"step": 273408
},
{
"epoch": 0.01,
"learning_rate": 4.551531975028221e-05,
"loss": 3.864,
"step": 273920
},
{
"epoch": 0.01,
"learning_rate": 4.550693380277169e-05,
"loss": 3.853,
"step": 274432
},
{
"epoch": 0.02,
"learning_rate": 4.549854785526117e-05,
"loss": 3.8691,
"step": 274944
},
{
"epoch": 0.02,
"learning_rate": 4.549016190775065e-05,
"loss": 3.8615,
"step": 275456
},
{
"epoch": 0.02,
"learning_rate": 4.5481792339043864e-05,
"loss": 3.8576,
"step": 275968
},
{
"epoch": 0.02,
"learning_rate": 4.5473406391533343e-05,
"loss": 3.8616,
"step": 276480
},
{
"epoch": 0.02,
"learning_rate": 4.5465020444022823e-05,
"loss": 3.8646,
"step": 276992
},
{
"epoch": 0.02,
"learning_rate": 4.545665087531603e-05,
"loss": 3.8748,
"step": 277504
},
{
"epoch": 0.02,
"learning_rate": 4.544826492780551e-05,
"loss": 3.8585,
"step": 278016
},
{
"epoch": 0.02,
"learning_rate": 4.543987898029499e-05,
"loss": 3.8438,
"step": 278528
},
{
"epoch": 0.02,
"learning_rate": 4.543149303278447e-05,
"loss": 3.8647,
"step": 279040
},
{
"epoch": 0.02,
"learning_rate": 4.542310708527395e-05,
"loss": 3.8557,
"step": 279552
},
{
"epoch": 0.02,
"learning_rate": 4.541472113776343e-05,
"loss": 3.8717,
"step": 280064
},
{
"epoch": 0.02,
"learning_rate": 4.540633519025291e-05,
"loss": 3.8661,
"step": 280576
},
{
"epoch": 0.02,
"learning_rate": 4.5397949242742385e-05,
"loss": 3.8532,
"step": 281088
},
{
"epoch": 0.02,
"learning_rate": 4.53895796740356e-05,
"loss": 3.8483,
"step": 281600
},
{
"epoch": 0.02,
"learning_rate": 4.538119372652508e-05,
"loss": 3.8652,
"step": 282112
},
{
"epoch": 0.02,
"learning_rate": 4.537280777901456e-05,
"loss": 3.844,
"step": 282624
},
{
"epoch": 0.02,
"learning_rate": 4.536442183150404e-05,
"loss": 3.8558,
"step": 283136
},
{
"epoch": 0.02,
"learning_rate": 4.535605226279726e-05,
"loss": 3.8538,
"step": 283648
},
{
"epoch": 0.02,
"learning_rate": 4.534766631528674e-05,
"loss": 3.8582,
"step": 284160
},
{
"epoch": 0.02,
"learning_rate": 4.5339296746579946e-05,
"loss": 3.8429,
"step": 284672
},
{
"epoch": 0.02,
"learning_rate": 4.5330910799069426e-05,
"loss": 3.8622,
"step": 285184
},
{
"epoch": 0.02,
"learning_rate": 4.5322524851558906e-05,
"loss": 3.8472,
"step": 285696
},
{
"epoch": 0.02,
"learning_rate": 4.5314138904048386e-05,
"loss": 3.851,
"step": 286208
},
{
"epoch": 0.02,
"learning_rate": 4.530575295653786e-05,
"loss": 3.8603,
"step": 286720
},
{
"epoch": 0.02,
"learning_rate": 4.529736700902734e-05,
"loss": 3.8482,
"step": 287232
},
{
"epoch": 0.02,
"learning_rate": 4.528898106151682e-05,
"loss": 3.8696,
"step": 287744
},
{
"epoch": 0.02,
"learning_rate": 4.52805951140063e-05,
"loss": 3.8591,
"step": 288256
},
{
"epoch": 0.02,
"learning_rate": 4.5272225545299515e-05,
"loss": 3.8558,
"step": 288768
},
{
"epoch": 0.02,
"learning_rate": 4.5263839597788995e-05,
"loss": 3.8464,
"step": 289280
},
{
"epoch": 0.02,
"learning_rate": 4.5255453650278475e-05,
"loss": 3.8535,
"step": 289792
},
{
"epoch": 0.02,
"learning_rate": 4.5247067702767955e-05,
"loss": 3.851,
"step": 290304
},
{
"epoch": 0.02,
"learning_rate": 4.5238698134061164e-05,
"loss": 3.8446,
"step": 290816
},
{
"epoch": 0.02,
"learning_rate": 4.5230312186550644e-05,
"loss": 3.8566,
"step": 291328
},
{
"epoch": 0.02,
"learning_rate": 4.5221926239040124e-05,
"loss": 3.8533,
"step": 291840
},
{
"epoch": 0.02,
"learning_rate": 4.5213540291529604e-05,
"loss": 3.8516,
"step": 292352
},
{
"epoch": 0.02,
"learning_rate": 4.520517072282281e-05,
"loss": 3.8483,
"step": 292864
},
{
"epoch": 0.02,
"learning_rate": 4.519678477531229e-05,
"loss": 3.8513,
"step": 293376
},
{
"epoch": 0.02,
"learning_rate": 4.518839882780177e-05,
"loss": 3.8458,
"step": 293888
},
{
"epoch": 0.02,
"learning_rate": 4.518001288029125e-05,
"loss": 3.8545,
"step": 294400
},
{
"epoch": 0.02,
"learning_rate": 4.517164331158447e-05,
"loss": 3.8404,
"step": 294912
},
{
"epoch": 0.02,
"learning_rate": 4.516325736407395e-05,
"loss": 3.8604,
"step": 295424
},
{
"epoch": 0.02,
"learning_rate": 4.515487141656343e-05,
"loss": 3.8547,
"step": 295936
},
{
"epoch": 0.02,
"learning_rate": 4.514648546905291e-05,
"loss": 3.8517,
"step": 296448
},
{
"epoch": 0.02,
"learning_rate": 4.513811590034612e-05,
"loss": 3.8432,
"step": 296960
},
{
"epoch": 0.02,
"learning_rate": 4.51297299528356e-05,
"loss": 3.8477,
"step": 297472
},
{
"epoch": 0.02,
"learning_rate": 4.512134400532508e-05,
"loss": 3.8525,
"step": 297984
},
{
"epoch": 0.02,
"learning_rate": 4.511295805781456e-05,
"loss": 3.8517,
"step": 298496
},
{
"epoch": 0.02,
"learning_rate": 4.5104588489107766e-05,
"loss": 3.8426,
"step": 299008
},
{
"epoch": 0.02,
"learning_rate": 4.5096202541597246e-05,
"loss": 3.8537,
"step": 299520
},
{
"epoch": 0.02,
"learning_rate": 4.5087816594086726e-05,
"loss": 3.8497,
"step": 300032
},
{
"epoch": 0.02,
"learning_rate": 4.5079430646576206e-05,
"loss": 3.8594,
"step": 300544
},
{
"epoch": 0.02,
"learning_rate": 4.507106107786942e-05,
"loss": 3.8437,
"step": 301056
},
{
"epoch": 0.02,
"learning_rate": 4.50626751303589e-05,
"loss": 3.8558,
"step": 301568
},
{
"epoch": 0.02,
"learning_rate": 4.505428918284838e-05,
"loss": 3.85,
"step": 302080
},
{
"epoch": 0.02,
"learning_rate": 4.504590323533786e-05,
"loss": 3.8413,
"step": 302592
},
{
"epoch": 0.02,
"learning_rate": 4.503753366663107e-05,
"loss": 3.8464,
"step": 303104
},
{
"epoch": 0.02,
"learning_rate": 4.502914771912055e-05,
"loss": 3.8553,
"step": 303616
},
{
"epoch": 0.02,
"learning_rate": 4.502076177161003e-05,
"loss": 3.8447,
"step": 304128
},
{
"epoch": 0.02,
"learning_rate": 4.501237582409951e-05,
"loss": 3.8318,
"step": 304640
},
{
"epoch": 0.02,
"learning_rate": 4.500400625539272e-05,
"loss": 3.8455,
"step": 305152
},
{
"epoch": 0.03,
"eval_loss": 3.9177167415618896,
"eval_runtime": 304.5017,
"eval_samples_per_second": 1253.166,
"eval_steps_per_second": 39.162,
"step": 305280
},
{
"epoch": 1.0,
"learning_rate": 4.49956203078822e-05,
"loss": 3.8347,
"step": 305664
},
{
"epoch": 1.0,
"learning_rate": 4.498723436037168e-05,
"loss": 3.829,
"step": 306176
},
{
"epoch": 1.0,
"learning_rate": 4.497884841286116e-05,
"loss": 3.8525,
"step": 306688
},
{
"epoch": 1.0,
"learning_rate": 4.4970478844154376e-05,
"loss": 3.852,
"step": 307200
},
{
"epoch": 1.0,
"learning_rate": 4.4962092896643856e-05,
"loss": 3.8496,
"step": 307712
},
{
"epoch": 1.0,
"learning_rate": 4.4953706949133336e-05,
"loss": 3.8359,
"step": 308224
},
{
"epoch": 1.0,
"learning_rate": 4.4945321001622816e-05,
"loss": 3.841,
"step": 308736
},
{
"epoch": 1.0,
"learning_rate": 4.4936951432916025e-05,
"loss": 3.8221,
"step": 309248
},
{
"epoch": 1.0,
"learning_rate": 4.4928565485405505e-05,
"loss": 3.8475,
"step": 309760
},
{
"epoch": 1.0,
"learning_rate": 4.4920179537894985e-05,
"loss": 3.8439,
"step": 310272
},
{
"epoch": 1.0,
"learning_rate": 4.4911793590384465e-05,
"loss": 3.8342,
"step": 310784
},
{
"epoch": 1.0,
"learning_rate": 4.4903424021677674e-05,
"loss": 3.8437,
"step": 311296
},
{
"epoch": 1.0,
"learning_rate": 4.489505445297088e-05,
"loss": 3.8301,
"step": 311808
},
{
"epoch": 1.0,
"learning_rate": 4.488666850546037e-05,
"loss": 3.8352,
"step": 312320
},
{
"epoch": 1.0,
"learning_rate": 4.487828255794985e-05,
"loss": 3.8254,
"step": 312832
},
{
"epoch": 1.0,
"learning_rate": 4.486989661043933e-05,
"loss": 3.8265,
"step": 313344
},
{
"epoch": 1.0,
"learning_rate": 4.486151066292881e-05,
"loss": 3.8286,
"step": 313856
},
{
"epoch": 1.0,
"learning_rate": 4.485312471541829e-05,
"loss": 3.8306,
"step": 314368
},
{
"epoch": 1.0,
"learning_rate": 4.484473876790777e-05,
"loss": 3.8324,
"step": 314880
},
{
"epoch": 1.0,
"learning_rate": 4.483635282039725e-05,
"loss": 3.85,
"step": 315392
},
{
"epoch": 1.0,
"learning_rate": 4.482798325169046e-05,
"loss": 3.8349,
"step": 315904
},
{
"epoch": 1.0,
"learning_rate": 4.481959730417994e-05,
"loss": 3.8407,
"step": 316416
},
{
"epoch": 1.0,
"learning_rate": 4.481121135666942e-05,
"loss": 3.8338,
"step": 316928
},
{
"epoch": 1.0,
"learning_rate": 4.48028254091589e-05,
"loss": 3.8376,
"step": 317440
},
{
"epoch": 1.0,
"learning_rate": 4.479445584045211e-05,
"loss": 3.8247,
"step": 317952
},
{
"epoch": 1.0,
"learning_rate": 4.478606989294159e-05,
"loss": 3.8328,
"step": 318464
},
{
"epoch": 1.0,
"learning_rate": 4.477768394543107e-05,
"loss": 3.8309,
"step": 318976
},
{
"epoch": 1.0,
"learning_rate": 4.476931437672428e-05,
"loss": 3.8226,
"step": 319488
},
{
"epoch": 1.0,
"learning_rate": 4.476092842921376e-05,
"loss": 3.8198,
"step": 320000
},
{
"epoch": 1.0,
"learning_rate": 4.475254248170324e-05,
"loss": 3.8279,
"step": 320512
},
{
"epoch": 1.01,
"learning_rate": 4.474415653419272e-05,
"loss": 3.8276,
"step": 321024
},
{
"epoch": 1.01,
"learning_rate": 4.473578696548593e-05,
"loss": 3.8368,
"step": 321536
},
{
"epoch": 1.01,
"learning_rate": 4.472740101797541e-05,
"loss": 3.8303,
"step": 322048
},
{
"epoch": 1.01,
"learning_rate": 4.471901507046489e-05,
"loss": 3.8322,
"step": 322560
},
{
"epoch": 1.01,
"learning_rate": 4.471062912295437e-05,
"loss": 3.821,
"step": 323072
},
{
"epoch": 1.01,
"learning_rate": 4.470224317544385e-05,
"loss": 3.8353,
"step": 323584
},
{
"epoch": 1.01,
"learning_rate": 4.469385722793333e-05,
"loss": 3.8202,
"step": 324096
},
{
"epoch": 1.01,
"learning_rate": 4.468547128042281e-05,
"loss": 3.825,
"step": 324608
},
{
"epoch": 1.01,
"learning_rate": 4.467710171171602e-05,
"loss": 3.8095,
"step": 325120
},
{
"epoch": 1.01,
"learning_rate": 4.46687157642055e-05,
"loss": 3.8239,
"step": 325632
},
{
"epoch": 1.01,
"learning_rate": 4.466032981669499e-05,
"loss": 3.8213,
"step": 326144
},
{
"epoch": 1.01,
"learning_rate": 4.465194386918447e-05,
"loss": 3.8336,
"step": 326656
},
{
"epoch": 1.01,
"learning_rate": 4.464357430047768e-05,
"loss": 3.8191,
"step": 327168
},
{
"epoch": 1.01,
"learning_rate": 4.463518835296716e-05,
"loss": 3.8255,
"step": 327680
},
{
"epoch": 1.01,
"learning_rate": 4.462680240545664e-05,
"loss": 3.8246,
"step": 328192
},
{
"epoch": 1.01,
"learning_rate": 4.461841645794611e-05,
"loss": 3.8304,
"step": 328704
},
{
"epoch": 1.01,
"learning_rate": 4.4610046889239326e-05,
"loss": 3.8242,
"step": 329216
},
{
"epoch": 1.01,
"learning_rate": 4.4601660941728806e-05,
"loss": 3.8141,
"step": 329728
},
{
"epoch": 1.01,
"learning_rate": 4.4593274994218286e-05,
"loss": 3.803,
"step": 330240
},
{
"epoch": 1.01,
"learning_rate": 4.458488904670776e-05,
"loss": 3.8292,
"step": 330752
},
{
"epoch": 1.01,
"learning_rate": 4.4576519478000975e-05,
"loss": 3.8279,
"step": 331264
},
{
"epoch": 1.01,
"learning_rate": 4.4568133530490455e-05,
"loss": 3.818,
"step": 331776
},
{
"epoch": 1.01,
"learning_rate": 4.455974758297994e-05,
"loss": 3.8167,
"step": 332288
},
{
"epoch": 1.01,
"learning_rate": 4.4551361635469415e-05,
"loss": 3.8174,
"step": 332800
},
{
"epoch": 1.01,
"learning_rate": 4.454299206676263e-05,
"loss": 3.7997,
"step": 333312
},
{
"epoch": 1.01,
"learning_rate": 4.453460611925211e-05,
"loss": 3.8161,
"step": 333824
},
{
"epoch": 1.01,
"learning_rate": 4.4526220171741584e-05,
"loss": 3.8054,
"step": 334336
},
{
"epoch": 1.01,
"learning_rate": 4.4517834224231064e-05,
"loss": 3.8161,
"step": 334848
},
{
"epoch": 1.01,
"learning_rate": 4.450946465552428e-05,
"loss": 3.8253,
"step": 335360
},
{
"epoch": 1.01,
"learning_rate": 4.450107870801376e-05,
"loss": 3.8112,
"step": 335872
},
{
"epoch": 1.01,
"learning_rate": 4.449269276050323e-05,
"loss": 3.8054,
"step": 336384
},
{
"epoch": 1.01,
"learning_rate": 4.448430681299271e-05,
"loss": 3.8184,
"step": 336896
},
{
"epoch": 1.01,
"learning_rate": 4.447593724428593e-05,
"loss": 3.7982,
"step": 337408
},
{
"epoch": 1.01,
"learning_rate": 4.446755129677541e-05,
"loss": 3.8067,
"step": 337920
},
{
"epoch": 1.01,
"learning_rate": 4.445916534926489e-05,
"loss": 3.8165,
"step": 338432
},
{
"epoch": 1.01,
"learning_rate": 4.445077940175437e-05,
"loss": 3.8129,
"step": 338944
},
{
"epoch": 1.01,
"learning_rate": 4.4442409833047584e-05,
"loss": 3.8027,
"step": 339456
},
{
"epoch": 1.01,
"learning_rate": 4.443402388553706e-05,
"loss": 3.8042,
"step": 339968
},
{
"epoch": 1.01,
"learning_rate": 4.442563793802654e-05,
"loss": 3.8049,
"step": 340480
},
{
"epoch": 1.01,
"learning_rate": 4.441725199051602e-05,
"loss": 3.8119,
"step": 340992
},
{
"epoch": 1.01,
"learning_rate": 4.440888242180923e-05,
"loss": 3.8126,
"step": 341504
},
{
"epoch": 1.01,
"learning_rate": 4.4400496474298706e-05,
"loss": 3.8083,
"step": 342016
},
{
"epoch": 1.01,
"learning_rate": 4.4392110526788186e-05,
"loss": 3.8171,
"step": 342528
},
{
"epoch": 1.01,
"learning_rate": 4.4383724579277666e-05,
"loss": 3.8182,
"step": 343040
},
{
"epoch": 1.01,
"learning_rate": 4.437535501057088e-05,
"loss": 3.8067,
"step": 343552
},
{
"epoch": 1.01,
"learning_rate": 4.436696906306036e-05,
"loss": 3.8021,
"step": 344064
},
{
"epoch": 1.01,
"learning_rate": 4.435858311554984e-05,
"loss": 3.8216,
"step": 344576
},
{
"epoch": 1.01,
"learning_rate": 4.435019716803932e-05,
"loss": 3.8096,
"step": 345088
},
{
"epoch": 1.01,
"learning_rate": 4.434182759933253e-05,
"loss": 3.8013,
"step": 345600
},
{
"epoch": 1.01,
"learning_rate": 4.433344165182201e-05,
"loss": 3.8057,
"step": 346112
},
{
"epoch": 1.01,
"learning_rate": 4.432505570431149e-05,
"loss": 3.8112,
"step": 346624
},
{
"epoch": 1.01,
"learning_rate": 4.431666975680097e-05,
"loss": 3.8183,
"step": 347136
},
{
"epoch": 1.01,
"learning_rate": 4.430830018809418e-05,
"loss": 3.8004,
"step": 347648
},
{
"epoch": 1.01,
"learning_rate": 4.429991424058366e-05,
"loss": 3.7915,
"step": 348160
},
{
"epoch": 1.01,
"learning_rate": 4.429152829307314e-05,
"loss": 3.7993,
"step": 348672
},
{
"epoch": 1.01,
"learning_rate": 4.428314234556262e-05,
"loss": 3.7989,
"step": 349184
},
{
"epoch": 1.01,
"learning_rate": 4.4274772776855836e-05,
"loss": 3.8097,
"step": 349696
},
{
"epoch": 1.01,
"learning_rate": 4.4266386829345316e-05,
"loss": 3.8058,
"step": 350208
},
{
"epoch": 1.01,
"learning_rate": 4.4258000881834796e-05,
"loss": 3.7923,
"step": 350720
},
{
"epoch": 1.02,
"learning_rate": 4.4249614934324276e-05,
"loss": 3.8097,
"step": 351232
},
{
"epoch": 1.02,
"learning_rate": 4.4241245365617485e-05,
"loss": 3.8002,
"step": 351744
},
{
"epoch": 1.02,
"learning_rate": 4.4232859418106965e-05,
"loss": 3.8028,
"step": 352256
},
{
"epoch": 1.02,
"learning_rate": 4.4224473470596445e-05,
"loss": 3.7983,
"step": 352768
},
{
"epoch": 1.02,
"learning_rate": 4.4216103901889654e-05,
"loss": 3.8038,
"step": 353280
},
{
"epoch": 1.02,
"learning_rate": 4.420773433318287e-05,
"loss": 3.8182,
"step": 353792
},
{
"epoch": 1.02,
"learning_rate": 4.419934838567235e-05,
"loss": 3.804,
"step": 354304
},
{
"epoch": 1.02,
"learning_rate": 4.419096243816182e-05,
"loss": 3.7824,
"step": 354816
},
{
"epoch": 1.02,
"learning_rate": 4.418257649065131e-05,
"loss": 3.8053,
"step": 355328
},
{
"epoch": 1.02,
"learning_rate": 4.417419054314079e-05,
"loss": 3.8006,
"step": 355840
},
{
"epoch": 1.02,
"learning_rate": 4.416580459563027e-05,
"loss": 3.8071,
"step": 356352
},
{
"epoch": 1.02,
"learning_rate": 4.415741864811975e-05,
"loss": 3.8087,
"step": 356864
},
{
"epoch": 1.02,
"learning_rate": 4.414903270060923e-05,
"loss": 3.796,
"step": 357376
},
{
"epoch": 1.02,
"learning_rate": 4.414066313190244e-05,
"loss": 3.7913,
"step": 357888
},
{
"epoch": 1.02,
"learning_rate": 4.413227718439192e-05,
"loss": 3.803,
"step": 358400
},
{
"epoch": 1.02,
"learning_rate": 4.41238912368814e-05,
"loss": 3.7834,
"step": 358912
},
{
"epoch": 1.02,
"learning_rate": 4.411550528937088e-05,
"loss": 3.7995,
"step": 359424
},
{
"epoch": 1.02,
"learning_rate": 4.410713572066409e-05,
"loss": 3.7942,
"step": 359936
},
{
"epoch": 1.02,
"learning_rate": 4.409874977315357e-05,
"loss": 3.8019,
"step": 360448
},
{
"epoch": 1.02,
"learning_rate": 4.4090380204446776e-05,
"loss": 3.7867,
"step": 360960
},
{
"epoch": 1.02,
"learning_rate": 4.408199425693626e-05,
"loss": 3.8041,
"step": 361472
},
{
"epoch": 1.02,
"learning_rate": 4.407360830942574e-05,
"loss": 3.7893,
"step": 361984
},
{
"epoch": 1.02,
"learning_rate": 4.406522236191522e-05,
"loss": 3.7925,
"step": 362496
},
{
"epoch": 1.02,
"learning_rate": 4.40568364144047e-05,
"loss": 3.8045,
"step": 363008
},
{
"epoch": 1.02,
"learning_rate": 4.404845046689418e-05,
"loss": 3.7946,
"step": 363520
},
{
"epoch": 1.02,
"learning_rate": 4.404006451938366e-05,
"loss": 3.8088,
"step": 364032
},
{
"epoch": 1.02,
"learning_rate": 4.403167857187314e-05,
"loss": 3.8025,
"step": 364544
},
{
"epoch": 1.02,
"learning_rate": 4.402330900316635e-05,
"loss": 3.7943,
"step": 365056
},
{
"epoch": 1.02,
"learning_rate": 4.401492305565583e-05,
"loss": 3.798,
"step": 365568
},
{
"epoch": 1.02,
"learning_rate": 4.400653710814531e-05,
"loss": 3.7918,
"step": 366080
},
{
"epoch": 1.02,
"learning_rate": 4.399815116063479e-05,
"loss": 3.798,
"step": 366592
},
{
"epoch": 1.02,
"learning_rate": 4.398976521312427e-05,
"loss": 3.7894,
"step": 367104
},
{
"epoch": 1.02,
"learning_rate": 4.398137926561375e-05,
"loss": 3.8021,
"step": 367616
},
{
"epoch": 1.02,
"learning_rate": 4.397299331810323e-05,
"loss": 3.7931,
"step": 368128
},
{
"epoch": 1.02,
"learning_rate": 4.396460737059271e-05,
"loss": 3.7907,
"step": 368640
},
{
"epoch": 1.02,
"learning_rate": 4.395623780188593e-05,
"loss": 3.7907,
"step": 369152
},
{
"epoch": 1.02,
"learning_rate": 4.394785185437541e-05,
"loss": 3.7971,
"step": 369664
},
{
"epoch": 1.02,
"learning_rate": 4.393946590686489e-05,
"loss": 3.7887,
"step": 370176
},
{
"epoch": 1.02,
"learning_rate": 4.393107995935437e-05,
"loss": 3.799,
"step": 370688
},
{
"epoch": 1.02,
"learning_rate": 4.3922710390647576e-05,
"loss": 3.7866,
"step": 371200
},
{
"epoch": 1.02,
"learning_rate": 4.3914324443137056e-05,
"loss": 3.8041,
"step": 371712
},
{
"epoch": 1.02,
"learning_rate": 4.3905938495626536e-05,
"loss": 3.7955,
"step": 372224
},
{
"epoch": 1.02,
"learning_rate": 4.3897552548116016e-05,
"loss": 3.7963,
"step": 372736
},
{
"epoch": 1.02,
"learning_rate": 4.3889182979409225e-05,
"loss": 3.7914,
"step": 373248
},
{
"epoch": 1.02,
"learning_rate": 4.3880797031898705e-05,
"loss": 3.79,
"step": 373760
},
{
"epoch": 1.02,
"learning_rate": 4.3872411084388185e-05,
"loss": 3.8022,
"step": 374272
},
{
"epoch": 1.02,
"learning_rate": 4.3864041515681394e-05,
"loss": 3.7929,
"step": 374784
},
{
"epoch": 1.02,
"learning_rate": 4.385565556817088e-05,
"loss": 3.7853,
"step": 375296
},
{
"epoch": 1.02,
"learning_rate": 4.384726962066036e-05,
"loss": 3.8003,
"step": 375808
},
{
"epoch": 1.02,
"learning_rate": 4.383888367314984e-05,
"loss": 3.7965,
"step": 376320
},
{
"epoch": 1.02,
"learning_rate": 4.383051410444305e-05,
"loss": 3.8073,
"step": 376832
},
{
"epoch": 1.02,
"learning_rate": 4.382212815693253e-05,
"loss": 3.7918,
"step": 377344
},
{
"epoch": 1.02,
"learning_rate": 4.381374220942201e-05,
"loss": 3.7994,
"step": 377856
},
{
"epoch": 1.02,
"learning_rate": 4.380535626191148e-05,
"loss": 3.7939,
"step": 378368
},
{
"epoch": 1.02,
"learning_rate": 4.37969866932047e-05,
"loss": 3.7869,
"step": 378880
},
{
"epoch": 1.02,
"learning_rate": 4.378860074569418e-05,
"loss": 3.7945,
"step": 379392
},
{
"epoch": 1.02,
"learning_rate": 4.378021479818366e-05,
"loss": 3.7967,
"step": 379904
},
{
"epoch": 1.02,
"learning_rate": 4.377182885067313e-05,
"loss": 3.7933,
"step": 380416
},
{
"epoch": 1.02,
"learning_rate": 4.376344290316262e-05,
"loss": 3.7784,
"step": 380928
},
{
"epoch": 1.02,
"learning_rate": 4.37550569556521e-05,
"loss": 3.7916,
"step": 381440
},
{
"epoch": 1.03,
"eval_loss": 3.8919215202331543,
"eval_runtime": 304.4828,
"eval_samples_per_second": 1253.243,
"eval_steps_per_second": 39.165,
"step": 381600
}
],
"logging_steps": 512,
"max_steps": 3052726,
"num_train_epochs": 9223372036854775807,
"save_steps": 10,
"total_flos": 2.656222362952704e+17,
"trial_name": null,
"trial_params": null
}