abhinavp's picture
Training in progress, epoch 0, checkpoint
9df2979 verified
{
"best_metric": 3.95322847366333,
"best_model_checkpoint": "/mmfs1/gscratch/stf/abhinavp/corpus-filtering/outputs/existential-there-quantifier/transformer/0/checkpoints/checkpoint-228960",
"epoch": 0.025000606015738065,
"eval_steps": 10,
"global_step": 228960,
"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": 10.9724,
"step": 1
},
{
"epoch": 0.0,
"learning_rate": 4.999161405248948e-05,
"loss": 6.8197,
"step": 512
},
{
"epoch": 0.0,
"learning_rate": 4.998322810497896e-05,
"loss": 6.1971,
"step": 1024
},
{
"epoch": 0.0,
"learning_rate": 4.997484215746844e-05,
"loss": 5.9867,
"step": 1536
},
{
"epoch": 0.0,
"learning_rate": 4.996645620995792e-05,
"loss": 5.8145,
"step": 2048
},
{
"epoch": 0.0,
"learning_rate": 4.99580702624474e-05,
"loss": 5.7061,
"step": 2560
},
{
"epoch": 0.0,
"learning_rate": 4.994968431493688e-05,
"loss": 5.602,
"step": 3072
},
{
"epoch": 0.0,
"learning_rate": 4.994129836742636e-05,
"loss": 5.5384,
"step": 3584
},
{
"epoch": 0.0,
"learning_rate": 4.993291241991584e-05,
"loss": 5.4496,
"step": 4096
},
{
"epoch": 0.0,
"learning_rate": 4.992452647240532e-05,
"loss": 5.4067,
"step": 4608
},
{
"epoch": 0.0,
"learning_rate": 4.99161405248948e-05,
"loss": 5.3559,
"step": 5120
},
{
"epoch": 0.0,
"learning_rate": 4.990775457738428e-05,
"loss": 5.3164,
"step": 5632
},
{
"epoch": 0.0,
"learning_rate": 4.989936862987376e-05,
"loss": 5.2621,
"step": 6144
},
{
"epoch": 0.0,
"learning_rate": 4.989099906116697e-05,
"loss": 5.2155,
"step": 6656
},
{
"epoch": 0.0,
"learning_rate": 4.988261311365645e-05,
"loss": 5.174,
"step": 7168
},
{
"epoch": 0.0,
"learning_rate": 4.987422716614593e-05,
"loss": 5.1412,
"step": 7680
},
{
"epoch": 0.0,
"learning_rate": 4.986584121863541e-05,
"loss": 5.1128,
"step": 8192
},
{
"epoch": 0.0,
"learning_rate": 4.985745527112489e-05,
"loss": 5.0964,
"step": 8704
},
{
"epoch": 0.0,
"learning_rate": 4.984906932361437e-05,
"loss": 5.0555,
"step": 9216
},
{
"epoch": 0.0,
"learning_rate": 4.984068337610385e-05,
"loss": 5.0346,
"step": 9728
},
{
"epoch": 0.0,
"learning_rate": 4.983229742859333e-05,
"loss": 5.0226,
"step": 10240
},
{
"epoch": 0.0,
"learning_rate": 4.9823927859886547e-05,
"loss": 4.9766,
"step": 10752
},
{
"epoch": 0.0,
"learning_rate": 4.9815541912376026e-05,
"loss": 4.9523,
"step": 11264
},
{
"epoch": 0.0,
"learning_rate": 4.9807155964865506e-05,
"loss": 4.935,
"step": 11776
},
{
"epoch": 0.0,
"learning_rate": 4.9798770017354986e-05,
"loss": 4.9275,
"step": 12288
},
{
"epoch": 0.0,
"learning_rate": 4.9790400448648195e-05,
"loss": 4.8925,
"step": 12800
},
{
"epoch": 0.0,
"learning_rate": 4.9782030879941405e-05,
"loss": 4.8725,
"step": 13312
},
{
"epoch": 0.0,
"learning_rate": 4.9773644932430885e-05,
"loss": 4.866,
"step": 13824
},
{
"epoch": 0.0,
"learning_rate": 4.9765258984920364e-05,
"loss": 4.843,
"step": 14336
},
{
"epoch": 0.0,
"learning_rate": 4.9756873037409844e-05,
"loss": 4.8167,
"step": 14848
},
{
"epoch": 0.01,
"learning_rate": 4.9748487089899324e-05,
"loss": 4.8006,
"step": 15360
},
{
"epoch": 0.01,
"learning_rate": 4.9740101142388804e-05,
"loss": 4.7969,
"step": 15872
},
{
"epoch": 0.01,
"learning_rate": 4.9731715194878284e-05,
"loss": 4.7839,
"step": 16384
},
{
"epoch": 0.01,
"learning_rate": 4.9723329247367764e-05,
"loss": 4.7793,
"step": 16896
},
{
"epoch": 0.01,
"learning_rate": 4.971495967866098e-05,
"loss": 4.7565,
"step": 17408
},
{
"epoch": 0.01,
"learning_rate": 4.970659010995419e-05,
"loss": 4.74,
"step": 17920
},
{
"epoch": 0.01,
"learning_rate": 4.969820416244367e-05,
"loss": 4.7426,
"step": 18432
},
{
"epoch": 0.01,
"learning_rate": 4.968981821493315e-05,
"loss": 4.7062,
"step": 18944
},
{
"epoch": 0.01,
"learning_rate": 4.968143226742263e-05,
"loss": 4.7007,
"step": 19456
},
{
"epoch": 0.01,
"learning_rate": 4.967304631991211e-05,
"loss": 4.6878,
"step": 19968
},
{
"epoch": 0.01,
"learning_rate": 4.966466037240159e-05,
"loss": 4.6875,
"step": 20480
},
{
"epoch": 0.01,
"learning_rate": 4.965627442489107e-05,
"loss": 4.6647,
"step": 20992
},
{
"epoch": 0.01,
"learning_rate": 4.964788847738054e-05,
"loss": 4.6759,
"step": 21504
},
{
"epoch": 0.01,
"learning_rate": 4.963951890867376e-05,
"loss": 4.6454,
"step": 22016
},
{
"epoch": 0.01,
"learning_rate": 4.963113296116324e-05,
"loss": 4.6496,
"step": 22528
},
{
"epoch": 0.01,
"learning_rate": 4.962274701365272e-05,
"loss": 4.639,
"step": 23040
},
{
"epoch": 0.01,
"learning_rate": 4.96143610661422e-05,
"loss": 4.6313,
"step": 23552
},
{
"epoch": 0.01,
"learning_rate": 4.960597511863168e-05,
"loss": 4.6288,
"step": 24064
},
{
"epoch": 0.01,
"learning_rate": 4.959758917112116e-05,
"loss": 4.603,
"step": 24576
},
{
"epoch": 0.01,
"learning_rate": 4.958920322361064e-05,
"loss": 4.5906,
"step": 25088
},
{
"epoch": 0.01,
"learning_rate": 4.958083365490385e-05,
"loss": 4.6048,
"step": 25600
},
{
"epoch": 0.01,
"learning_rate": 4.957244770739333e-05,
"loss": 4.5908,
"step": 26112
},
{
"epoch": 0.01,
"learning_rate": 4.956406175988281e-05,
"loss": 4.5719,
"step": 26624
},
{
"epoch": 0.01,
"learning_rate": 4.955567581237229e-05,
"loss": 4.5619,
"step": 27136
},
{
"epoch": 0.01,
"learning_rate": 4.954732262246923e-05,
"loss": 4.5677,
"step": 27648
},
{
"epoch": 0.01,
"learning_rate": 4.953893667495871e-05,
"loss": 4.5508,
"step": 28160
},
{
"epoch": 0.01,
"learning_rate": 4.953055072744819e-05,
"loss": 4.5626,
"step": 28672
},
{
"epoch": 0.01,
"learning_rate": 4.952216477993767e-05,
"loss": 4.5345,
"step": 29184
},
{
"epoch": 0.01,
"learning_rate": 4.951377883242715e-05,
"loss": 4.5351,
"step": 29696
},
{
"epoch": 0.01,
"learning_rate": 4.950539288491663e-05,
"loss": 4.5383,
"step": 30208
},
{
"epoch": 0.01,
"learning_rate": 4.949700693740611e-05,
"loss": 4.5198,
"step": 30720
},
{
"epoch": 0.01,
"learning_rate": 4.948862098989559e-05,
"loss": 4.5017,
"step": 31232
},
{
"epoch": 0.01,
"learning_rate": 4.94802514211888e-05,
"loss": 4.5069,
"step": 31744
},
{
"epoch": 0.01,
"learning_rate": 4.947186547367828e-05,
"loss": 4.4903,
"step": 32256
},
{
"epoch": 0.01,
"learning_rate": 4.946347952616776e-05,
"loss": 4.4914,
"step": 32768
},
{
"epoch": 0.01,
"learning_rate": 4.945509357865724e-05,
"loss": 4.4997,
"step": 33280
},
{
"epoch": 0.01,
"learning_rate": 4.944670763114672e-05,
"loss": 4.496,
"step": 33792
},
{
"epoch": 0.01,
"learning_rate": 4.943833806243993e-05,
"loss": 4.4789,
"step": 34304
},
{
"epoch": 0.01,
"learning_rate": 4.942995211492941e-05,
"loss": 4.4666,
"step": 34816
},
{
"epoch": 0.01,
"learning_rate": 4.942156616741889e-05,
"loss": 4.461,
"step": 35328
},
{
"epoch": 0.01,
"learning_rate": 4.941318021990837e-05,
"loss": 4.4722,
"step": 35840
},
{
"epoch": 0.01,
"learning_rate": 4.9404810651201585e-05,
"loss": 4.4692,
"step": 36352
},
{
"epoch": 0.01,
"learning_rate": 4.9396441082494794e-05,
"loss": 4.4595,
"step": 36864
},
{
"epoch": 0.01,
"learning_rate": 4.9388055134984274e-05,
"loss": 4.4581,
"step": 37376
},
{
"epoch": 0.01,
"learning_rate": 4.9379669187473754e-05,
"loss": 4.4619,
"step": 37888
},
{
"epoch": 0.01,
"learning_rate": 4.9371283239963234e-05,
"loss": 4.4532,
"step": 38400
},
{
"epoch": 0.01,
"learning_rate": 4.936291367125644e-05,
"loss": 4.4286,
"step": 38912
},
{
"epoch": 0.01,
"learning_rate": 4.935452772374592e-05,
"loss": 4.4398,
"step": 39424
},
{
"epoch": 0.01,
"learning_rate": 4.93461417762354e-05,
"loss": 4.4356,
"step": 39936
},
{
"epoch": 0.01,
"learning_rate": 4.933775582872488e-05,
"loss": 4.4201,
"step": 40448
},
{
"epoch": 0.01,
"learning_rate": 4.932936988121436e-05,
"loss": 4.4321,
"step": 40960
},
{
"epoch": 0.01,
"learning_rate": 4.932098393370384e-05,
"loss": 4.4225,
"step": 41472
},
{
"epoch": 0.01,
"learning_rate": 4.931259798619332e-05,
"loss": 4.4283,
"step": 41984
},
{
"epoch": 0.01,
"learning_rate": 4.930421203868281e-05,
"loss": 4.4179,
"step": 42496
},
{
"epoch": 0.01,
"learning_rate": 4.929584246997602e-05,
"loss": 4.3931,
"step": 43008
},
{
"epoch": 0.01,
"learning_rate": 4.92874565224655e-05,
"loss": 4.4021,
"step": 43520
},
{
"epoch": 0.01,
"learning_rate": 4.927907057495498e-05,
"loss": 4.4009,
"step": 44032
},
{
"epoch": 0.01,
"learning_rate": 4.927068462744446e-05,
"loss": 4.4022,
"step": 44544
},
{
"epoch": 0.01,
"learning_rate": 4.926231505873767e-05,
"loss": 4.3939,
"step": 45056
},
{
"epoch": 0.01,
"learning_rate": 4.925394549003088e-05,
"loss": 4.3847,
"step": 45568
},
{
"epoch": 0.02,
"learning_rate": 4.924555954252036e-05,
"loss": 4.3882,
"step": 46080
},
{
"epoch": 0.02,
"learning_rate": 4.923717359500984e-05,
"loss": 4.377,
"step": 46592
},
{
"epoch": 0.02,
"learning_rate": 4.9228804026303046e-05,
"loss": 4.3692,
"step": 47104
},
{
"epoch": 0.02,
"learning_rate": 4.922041807879253e-05,
"loss": 4.3721,
"step": 47616
},
{
"epoch": 0.02,
"learning_rate": 4.921203213128201e-05,
"loss": 4.367,
"step": 48128
},
{
"epoch": 0.02,
"learning_rate": 4.920364618377149e-05,
"loss": 4.3785,
"step": 48640
},
{
"epoch": 0.02,
"learning_rate": 4.919526023626097e-05,
"loss": 4.3566,
"step": 49152
},
{
"epoch": 0.02,
"learning_rate": 4.918687428875045e-05,
"loss": 4.3506,
"step": 49664
},
{
"epoch": 0.02,
"learning_rate": 4.917848834123993e-05,
"loss": 4.3508,
"step": 50176
},
{
"epoch": 0.02,
"learning_rate": 4.917010239372941e-05,
"loss": 4.3582,
"step": 50688
},
{
"epoch": 0.02,
"learning_rate": 4.916171644621889e-05,
"loss": 4.3556,
"step": 51200
},
{
"epoch": 0.02,
"learning_rate": 4.915333049870837e-05,
"loss": 4.3411,
"step": 51712
},
{
"epoch": 0.02,
"learning_rate": 4.914494455119785e-05,
"loss": 4.3415,
"step": 52224
},
{
"epoch": 0.02,
"learning_rate": 4.9136558603687325e-05,
"loss": 4.3235,
"step": 52736
},
{
"epoch": 0.02,
"learning_rate": 4.912818903498054e-05,
"loss": 4.3386,
"step": 53248
},
{
"epoch": 0.02,
"learning_rate": 4.911980308747002e-05,
"loss": 4.3242,
"step": 53760
},
{
"epoch": 0.02,
"learning_rate": 4.91114171399595e-05,
"loss": 4.3226,
"step": 54272
},
{
"epoch": 0.02,
"learning_rate": 4.910304757125272e-05,
"loss": 4.3261,
"step": 54784
},
{
"epoch": 0.02,
"learning_rate": 4.90946616237422e-05,
"loss": 4.3257,
"step": 55296
},
{
"epoch": 0.02,
"learning_rate": 4.908627567623168e-05,
"loss": 4.3194,
"step": 55808
},
{
"epoch": 0.02,
"learning_rate": 4.907788972872115e-05,
"loss": 4.3175,
"step": 56320
},
{
"epoch": 0.02,
"learning_rate": 4.906950378121063e-05,
"loss": 4.3089,
"step": 56832
},
{
"epoch": 0.02,
"learning_rate": 4.906111783370011e-05,
"loss": 4.3152,
"step": 57344
},
{
"epoch": 0.02,
"learning_rate": 4.9052748264993326e-05,
"loss": 4.303,
"step": 57856
},
{
"epoch": 0.02,
"learning_rate": 4.90443623174828e-05,
"loss": 4.3045,
"step": 58368
},
{
"epoch": 0.02,
"learning_rate": 4.903597636997228e-05,
"loss": 4.31,
"step": 58880
},
{
"epoch": 0.02,
"learning_rate": 4.902759042246176e-05,
"loss": 4.3061,
"step": 59392
},
{
"epoch": 0.02,
"learning_rate": 4.9019220853754975e-05,
"loss": 4.3042,
"step": 59904
},
{
"epoch": 0.02,
"learning_rate": 4.9010834906244455e-05,
"loss": 4.2941,
"step": 60416
},
{
"epoch": 0.02,
"learning_rate": 4.9002448958733935e-05,
"loss": 4.2988,
"step": 60928
},
{
"epoch": 0.02,
"learning_rate": 4.8994063011223415e-05,
"loss": 4.295,
"step": 61440
},
{
"epoch": 0.02,
"learning_rate": 4.8985677063712895e-05,
"loss": 4.2806,
"step": 61952
},
{
"epoch": 0.02,
"learning_rate": 4.8977291116202375e-05,
"loss": 4.2983,
"step": 62464
},
{
"epoch": 0.02,
"learning_rate": 4.8968905168691855e-05,
"loss": 4.2833,
"step": 62976
},
{
"epoch": 0.02,
"learning_rate": 4.8960535599985064e-05,
"loss": 4.2914,
"step": 63488
},
{
"epoch": 0.02,
"learning_rate": 4.8952149652474544e-05,
"loss": 4.2725,
"step": 64000
},
{
"epoch": 0.02,
"learning_rate": 4.8943763704964024e-05,
"loss": 4.2682,
"step": 64512
},
{
"epoch": 0.02,
"learning_rate": 4.8935377757453504e-05,
"loss": 4.2657,
"step": 65024
},
{
"epoch": 0.02,
"learning_rate": 4.892700818874671e-05,
"loss": 4.269,
"step": 65536
},
{
"epoch": 0.02,
"learning_rate": 4.891862224123619e-05,
"loss": 4.2602,
"step": 66048
},
{
"epoch": 0.02,
"learning_rate": 4.891023629372567e-05,
"loss": 4.2767,
"step": 66560
},
{
"epoch": 0.02,
"learning_rate": 4.890185034621515e-05,
"loss": 4.2728,
"step": 67072
},
{
"epoch": 0.02,
"learning_rate": 4.889348077750837e-05,
"loss": 4.2573,
"step": 67584
},
{
"epoch": 0.02,
"learning_rate": 4.888509482999785e-05,
"loss": 4.2604,
"step": 68096
},
{
"epoch": 0.02,
"learning_rate": 4.887670888248733e-05,
"loss": 4.2654,
"step": 68608
},
{
"epoch": 0.02,
"learning_rate": 4.886832293497681e-05,
"loss": 4.2618,
"step": 69120
},
{
"epoch": 0.02,
"learning_rate": 4.885995336627002e-05,
"loss": 4.2654,
"step": 69632
},
{
"epoch": 0.02,
"learning_rate": 4.8851583797563226e-05,
"loss": 4.2617,
"step": 70144
},
{
"epoch": 0.02,
"learning_rate": 4.8843197850052706e-05,
"loss": 4.2483,
"step": 70656
},
{
"epoch": 0.02,
"learning_rate": 4.8834811902542186e-05,
"loss": 4.253,
"step": 71168
},
{
"epoch": 0.02,
"learning_rate": 4.8826425955031666e-05,
"loss": 4.2487,
"step": 71680
},
{
"epoch": 0.02,
"learning_rate": 4.8818040007521146e-05,
"loss": 4.2405,
"step": 72192
},
{
"epoch": 0.02,
"learning_rate": 4.8809654060010626e-05,
"loss": 4.252,
"step": 72704
},
{
"epoch": 0.02,
"learning_rate": 4.8801268112500106e-05,
"loss": 4.2381,
"step": 73216
},
{
"epoch": 0.02,
"learning_rate": 4.879289854379332e-05,
"loss": 4.2283,
"step": 73728
},
{
"epoch": 0.02,
"learning_rate": 4.87845125962828e-05,
"loss": 4.2508,
"step": 74240
},
{
"epoch": 0.02,
"learning_rate": 4.877612664877228e-05,
"loss": 4.2314,
"step": 74752
},
{
"epoch": 0.02,
"learning_rate": 4.876774070126176e-05,
"loss": 4.2305,
"step": 75264
},
{
"epoch": 0.02,
"learning_rate": 4.875935475375124e-05,
"loss": 4.232,
"step": 75776
},
{
"epoch": 0.02,
"learning_rate": 4.875098518504445e-05,
"loss": 4.226,
"step": 76288
},
{
"epoch": 0.03,
"eval_loss": 4.196994781494141,
"eval_runtime": 302.917,
"eval_samples_per_second": 1259.721,
"eval_steps_per_second": 39.367,
"step": 76320
},
{
"epoch": 1.0,
"learning_rate": 4.874259923753393e-05,
"loss": 4.2093,
"step": 76800
},
{
"epoch": 1.0,
"learning_rate": 4.873421329002341e-05,
"loss": 4.2203,
"step": 77312
},
{
"epoch": 1.0,
"learning_rate": 4.872582734251289e-05,
"loss": 4.2368,
"step": 77824
},
{
"epoch": 1.0,
"learning_rate": 4.871744139500237e-05,
"loss": 4.2146,
"step": 78336
},
{
"epoch": 1.0,
"learning_rate": 4.870905544749185e-05,
"loss": 4.228,
"step": 78848
},
{
"epoch": 1.0,
"learning_rate": 4.870066949998133e-05,
"loss": 4.2021,
"step": 79360
},
{
"epoch": 1.0,
"learning_rate": 4.869228355247081e-05,
"loss": 4.2157,
"step": 79872
},
{
"epoch": 1.0,
"learning_rate": 4.8683913983764027e-05,
"loss": 4.1985,
"step": 80384
},
{
"epoch": 1.0,
"learning_rate": 4.8675544415057236e-05,
"loss": 4.2096,
"step": 80896
},
{
"epoch": 1.0,
"learning_rate": 4.8667158467546716e-05,
"loss": 4.207,
"step": 81408
},
{
"epoch": 1.0,
"learning_rate": 4.8658772520036196e-05,
"loss": 4.2098,
"step": 81920
},
{
"epoch": 1.0,
"learning_rate": 4.8650386572525676e-05,
"loss": 4.2068,
"step": 82432
},
{
"epoch": 1.0,
"learning_rate": 4.8642017003818885e-05,
"loss": 4.1941,
"step": 82944
},
{
"epoch": 1.0,
"learning_rate": 4.8633631056308365e-05,
"loss": 4.1845,
"step": 83456
},
{
"epoch": 1.0,
"learning_rate": 4.8625245108797845e-05,
"loss": 4.1857,
"step": 83968
},
{
"epoch": 1.0,
"learning_rate": 4.8616859161287324e-05,
"loss": 4.1886,
"step": 84480
},
{
"epoch": 1.0,
"learning_rate": 4.8608473213776804e-05,
"loss": 4.1981,
"step": 84992
},
{
"epoch": 1.0,
"learning_rate": 4.860008726626628e-05,
"loss": 4.1824,
"step": 85504
},
{
"epoch": 1.0,
"learning_rate": 4.8591701318755764e-05,
"loss": 4.1919,
"step": 86016
},
{
"epoch": 1.0,
"learning_rate": 4.8583315371245244e-05,
"loss": 4.2042,
"step": 86528
},
{
"epoch": 1.0,
"learning_rate": 4.8574929423734724e-05,
"loss": 4.1811,
"step": 87040
},
{
"epoch": 1.0,
"learning_rate": 4.8566543476224204e-05,
"loss": 4.1744,
"step": 87552
},
{
"epoch": 1.0,
"learning_rate": 4.8558157528713684e-05,
"loss": 4.1791,
"step": 88064
},
{
"epoch": 1.0,
"learning_rate": 4.854978796000689e-05,
"loss": 4.1891,
"step": 88576
},
{
"epoch": 1.0,
"learning_rate": 4.854141839130011e-05,
"loss": 4.1677,
"step": 89088
},
{
"epoch": 1.0,
"learning_rate": 4.853303244378958e-05,
"loss": 4.1732,
"step": 89600
},
{
"epoch": 1.0,
"learning_rate": 4.852464649627906e-05,
"loss": 4.1738,
"step": 90112
},
{
"epoch": 1.0,
"learning_rate": 4.851626054876854e-05,
"loss": 4.1692,
"step": 90624
},
{
"epoch": 1.0,
"learning_rate": 4.850787460125802e-05,
"loss": 4.1521,
"step": 91136
},
{
"epoch": 1.01,
"learning_rate": 4.84994886537475e-05,
"loss": 4.1598,
"step": 91648
},
{
"epoch": 1.01,
"learning_rate": 4.849110270623698e-05,
"loss": 4.1568,
"step": 92160
},
{
"epoch": 1.01,
"learning_rate": 4.84827331375302e-05,
"loss": 4.1657,
"step": 92672
},
{
"epoch": 1.01,
"learning_rate": 4.847434719001968e-05,
"loss": 4.1728,
"step": 93184
},
{
"epoch": 1.01,
"learning_rate": 4.846596124250916e-05,
"loss": 4.1563,
"step": 93696
},
{
"epoch": 1.01,
"learning_rate": 4.845757529499864e-05,
"loss": 4.152,
"step": 94208
},
{
"epoch": 1.01,
"learning_rate": 4.844918934748812e-05,
"loss": 4.1663,
"step": 94720
},
{
"epoch": 1.01,
"learning_rate": 4.84408033999776e-05,
"loss": 4.1395,
"step": 95232
},
{
"epoch": 1.01,
"learning_rate": 4.843241745246708e-05,
"loss": 4.147,
"step": 95744
},
{
"epoch": 1.01,
"learning_rate": 4.842404788376029e-05,
"loss": 4.1375,
"step": 96256
},
{
"epoch": 1.01,
"learning_rate": 4.841566193624977e-05,
"loss": 4.1461,
"step": 96768
},
{
"epoch": 1.01,
"learning_rate": 4.840727598873925e-05,
"loss": 4.1359,
"step": 97280
},
{
"epoch": 1.01,
"learning_rate": 4.839889004122873e-05,
"loss": 4.154,
"step": 97792
},
{
"epoch": 1.01,
"learning_rate": 4.839053685132567e-05,
"loss": 4.1396,
"step": 98304
},
{
"epoch": 1.01,
"learning_rate": 4.838215090381515e-05,
"loss": 4.141,
"step": 98816
},
{
"epoch": 1.01,
"learning_rate": 4.837376495630463e-05,
"loss": 4.1374,
"step": 99328
},
{
"epoch": 1.01,
"learning_rate": 4.836537900879411e-05,
"loss": 4.1418,
"step": 99840
},
{
"epoch": 1.01,
"learning_rate": 4.835699306128359e-05,
"loss": 4.1357,
"step": 100352
},
{
"epoch": 1.01,
"learning_rate": 4.834860711377307e-05,
"loss": 4.1263,
"step": 100864
},
{
"epoch": 1.01,
"learning_rate": 4.834022116626255e-05,
"loss": 4.1193,
"step": 101376
},
{
"epoch": 1.01,
"learning_rate": 4.833185159755576e-05,
"loss": 4.1403,
"step": 101888
},
{
"epoch": 1.01,
"learning_rate": 4.832346565004524e-05,
"loss": 4.1305,
"step": 102400
},
{
"epoch": 1.01,
"learning_rate": 4.831507970253472e-05,
"loss": 4.1168,
"step": 102912
},
{
"epoch": 1.01,
"learning_rate": 4.83066937550242e-05,
"loss": 4.1176,
"step": 103424
},
{
"epoch": 1.01,
"learning_rate": 4.829832418631741e-05,
"loss": 4.1254,
"step": 103936
},
{
"epoch": 1.01,
"learning_rate": 4.828993823880689e-05,
"loss": 4.1149,
"step": 104448
},
{
"epoch": 1.01,
"learning_rate": 4.828155229129637e-05,
"loss": 4.1303,
"step": 104960
},
{
"epoch": 1.01,
"learning_rate": 4.8273166343785856e-05,
"loss": 4.1121,
"step": 105472
},
{
"epoch": 1.01,
"learning_rate": 4.8264780396275336e-05,
"loss": 4.1158,
"step": 105984
},
{
"epoch": 1.01,
"learning_rate": 4.8256394448764816e-05,
"loss": 4.1273,
"step": 106496
},
{
"epoch": 1.01,
"learning_rate": 4.8248008501254296e-05,
"loss": 4.1039,
"step": 107008
},
{
"epoch": 1.01,
"learning_rate": 4.823962255374377e-05,
"loss": 4.1011,
"step": 107520
},
{
"epoch": 1.01,
"learning_rate": 4.8231269363840714e-05,
"loss": 4.1109,
"step": 108032
},
{
"epoch": 1.01,
"learning_rate": 4.8222883416330194e-05,
"loss": 4.0937,
"step": 108544
},
{
"epoch": 1.01,
"learning_rate": 4.8214497468819674e-05,
"loss": 4.0992,
"step": 109056
},
{
"epoch": 1.01,
"learning_rate": 4.8206111521309154e-05,
"loss": 4.1106,
"step": 109568
},
{
"epoch": 1.01,
"learning_rate": 4.8197725573798634e-05,
"loss": 4.1156,
"step": 110080
},
{
"epoch": 1.01,
"learning_rate": 4.8189339626288114e-05,
"loss": 4.1018,
"step": 110592
},
{
"epoch": 1.01,
"learning_rate": 4.8180953678777594e-05,
"loss": 4.095,
"step": 111104
},
{
"epoch": 1.01,
"learning_rate": 4.817258411007081e-05,
"loss": 4.0873,
"step": 111616
},
{
"epoch": 1.01,
"learning_rate": 4.816419816256029e-05,
"loss": 4.1074,
"step": 112128
},
{
"epoch": 1.01,
"learning_rate": 4.815581221504977e-05,
"loss": 4.1042,
"step": 112640
},
{
"epoch": 1.01,
"learning_rate": 4.814742626753924e-05,
"loss": 4.1011,
"step": 113152
},
{
"epoch": 1.01,
"learning_rate": 4.813904032002872e-05,
"loss": 4.102,
"step": 113664
},
{
"epoch": 1.01,
"learning_rate": 4.81306543725182e-05,
"loss": 4.1042,
"step": 114176
},
{
"epoch": 1.01,
"learning_rate": 4.812228480381141e-05,
"loss": 4.1054,
"step": 114688
},
{
"epoch": 1.01,
"learning_rate": 4.811391523510463e-05,
"loss": 4.0832,
"step": 115200
},
{
"epoch": 1.01,
"learning_rate": 4.810552928759411e-05,
"loss": 4.097,
"step": 115712
},
{
"epoch": 1.01,
"learning_rate": 4.809714334008359e-05,
"loss": 4.0956,
"step": 116224
},
{
"epoch": 1.01,
"learning_rate": 4.808875739257306e-05,
"loss": 4.0809,
"step": 116736
},
{
"epoch": 1.01,
"learning_rate": 4.808037144506255e-05,
"loss": 4.1024,
"step": 117248
},
{
"epoch": 1.01,
"learning_rate": 4.807198549755203e-05,
"loss": 4.0954,
"step": 117760
},
{
"epoch": 1.01,
"learning_rate": 4.806359955004151e-05,
"loss": 4.0991,
"step": 118272
},
{
"epoch": 1.01,
"learning_rate": 4.805521360253099e-05,
"loss": 4.0967,
"step": 118784
},
{
"epoch": 1.01,
"learning_rate": 4.804682765502047e-05,
"loss": 4.0722,
"step": 119296
},
{
"epoch": 1.01,
"learning_rate": 4.8038458086313677e-05,
"loss": 4.0854,
"step": 119808
},
{
"epoch": 1.01,
"learning_rate": 4.8030072138803156e-05,
"loss": 4.0835,
"step": 120320
},
{
"epoch": 1.01,
"learning_rate": 4.8021686191292636e-05,
"loss": 4.0887,
"step": 120832
},
{
"epoch": 1.01,
"learning_rate": 4.8013300243782116e-05,
"loss": 4.0834,
"step": 121344
},
{
"epoch": 1.01,
"learning_rate": 4.8004914296271596e-05,
"loss": 4.0763,
"step": 121856
},
{
"epoch": 1.02,
"learning_rate": 4.7996544727564805e-05,
"loss": 4.0851,
"step": 122368
},
{
"epoch": 1.02,
"learning_rate": 4.7988158780054285e-05,
"loss": 4.079,
"step": 122880
},
{
"epoch": 1.02,
"learning_rate": 4.7979772832543765e-05,
"loss": 4.0645,
"step": 123392
},
{
"epoch": 1.02,
"learning_rate": 4.7971386885033245e-05,
"loss": 4.0736,
"step": 123904
},
{
"epoch": 1.02,
"learning_rate": 4.796301731632646e-05,
"loss": 4.0775,
"step": 124416
},
{
"epoch": 1.02,
"learning_rate": 4.795463136881594e-05,
"loss": 4.0875,
"step": 124928
},
{
"epoch": 1.02,
"learning_rate": 4.794624542130542e-05,
"loss": 4.0692,
"step": 125440
},
{
"epoch": 1.02,
"learning_rate": 4.79378594737949e-05,
"loss": 4.0572,
"step": 125952
},
{
"epoch": 1.02,
"learning_rate": 4.792948990508811e-05,
"loss": 4.0666,
"step": 126464
},
{
"epoch": 1.02,
"learning_rate": 4.792110395757759e-05,
"loss": 4.0753,
"step": 126976
},
{
"epoch": 1.02,
"learning_rate": 4.791271801006707e-05,
"loss": 4.0736,
"step": 127488
},
{
"epoch": 1.02,
"learning_rate": 4.790433206255655e-05,
"loss": 4.066,
"step": 128000
},
{
"epoch": 1.02,
"learning_rate": 4.789594611504603e-05,
"loss": 4.0626,
"step": 128512
},
{
"epoch": 1.02,
"learning_rate": 4.788757654633924e-05,
"loss": 4.05,
"step": 129024
},
{
"epoch": 1.02,
"learning_rate": 4.787919059882872e-05,
"loss": 4.0668,
"step": 129536
},
{
"epoch": 1.02,
"learning_rate": 4.78708046513182e-05,
"loss": 4.0523,
"step": 130048
},
{
"epoch": 1.02,
"learning_rate": 4.786241870380768e-05,
"loss": 4.0554,
"step": 130560
},
{
"epoch": 1.02,
"learning_rate": 4.7854049135100895e-05,
"loss": 4.0599,
"step": 131072
},
{
"epoch": 1.02,
"learning_rate": 4.7845663187590375e-05,
"loss": 4.0607,
"step": 131584
},
{
"epoch": 1.02,
"learning_rate": 4.7837277240079855e-05,
"loss": 4.0545,
"step": 132096
},
{
"epoch": 1.02,
"learning_rate": 4.7828891292569335e-05,
"loss": 4.0589,
"step": 132608
},
{
"epoch": 1.02,
"learning_rate": 4.7820521723862544e-05,
"loss": 4.0476,
"step": 133120
},
{
"epoch": 1.02,
"learning_rate": 4.7812135776352024e-05,
"loss": 4.055,
"step": 133632
},
{
"epoch": 1.02,
"learning_rate": 4.7803749828841504e-05,
"loss": 4.0474,
"step": 134144
},
{
"epoch": 1.02,
"learning_rate": 4.7795363881330984e-05,
"loss": 4.0475,
"step": 134656
},
{
"epoch": 1.02,
"learning_rate": 4.778699431262419e-05,
"loss": 4.0621,
"step": 135168
},
{
"epoch": 1.02,
"learning_rate": 4.777860836511367e-05,
"loss": 4.0589,
"step": 135680
},
{
"epoch": 1.02,
"learning_rate": 4.777023879640689e-05,
"loss": 4.0541,
"step": 136192
},
{
"epoch": 1.02,
"learning_rate": 4.776185284889637e-05,
"loss": 4.0463,
"step": 136704
},
{
"epoch": 1.02,
"learning_rate": 4.775346690138585e-05,
"loss": 4.0527,
"step": 137216
},
{
"epoch": 1.02,
"learning_rate": 4.774508095387533e-05,
"loss": 4.05,
"step": 137728
},
{
"epoch": 1.02,
"learning_rate": 4.773669500636481e-05,
"loss": 4.0386,
"step": 138240
},
{
"epoch": 1.02,
"learning_rate": 4.772830905885429e-05,
"loss": 4.0567,
"step": 138752
},
{
"epoch": 1.02,
"learning_rate": 4.771992311134377e-05,
"loss": 4.0424,
"step": 139264
},
{
"epoch": 1.02,
"learning_rate": 4.771153716383325e-05,
"loss": 4.0522,
"step": 139776
},
{
"epoch": 1.02,
"learning_rate": 4.770315121632272e-05,
"loss": 4.0349,
"step": 140288
},
{
"epoch": 1.02,
"learning_rate": 4.769478164761594e-05,
"loss": 4.0361,
"step": 140800
},
{
"epoch": 1.02,
"learning_rate": 4.768639570010542e-05,
"loss": 4.0329,
"step": 141312
},
{
"epoch": 1.02,
"learning_rate": 4.76780097525949e-05,
"loss": 4.0348,
"step": 141824
},
{
"epoch": 1.02,
"learning_rate": 4.766962380508437e-05,
"loss": 4.0329,
"step": 142336
},
{
"epoch": 1.02,
"learning_rate": 4.7661254236377586e-05,
"loss": 4.0465,
"step": 142848
},
{
"epoch": 1.02,
"learning_rate": 4.765286828886707e-05,
"loss": 4.0493,
"step": 143360
},
{
"epoch": 1.02,
"learning_rate": 4.764448234135655e-05,
"loss": 4.0357,
"step": 143872
},
{
"epoch": 1.02,
"learning_rate": 4.7636096393846026e-05,
"loss": 4.0287,
"step": 144384
},
{
"epoch": 1.02,
"learning_rate": 4.762772682513924e-05,
"loss": 4.0413,
"step": 144896
},
{
"epoch": 1.02,
"learning_rate": 4.761934087762872e-05,
"loss": 4.0392,
"step": 145408
},
{
"epoch": 1.02,
"learning_rate": 4.7610954930118195e-05,
"loss": 4.0446,
"step": 145920
},
{
"epoch": 1.02,
"learning_rate": 4.7602568982607675e-05,
"loss": 4.0401,
"step": 146432
},
{
"epoch": 1.02,
"learning_rate": 4.759419941390089e-05,
"loss": 4.0354,
"step": 146944
},
{
"epoch": 1.02,
"learning_rate": 4.758581346639037e-05,
"loss": 4.0269,
"step": 147456
},
{
"epoch": 1.02,
"learning_rate": 4.7577427518879844e-05,
"loss": 4.0371,
"step": 147968
},
{
"epoch": 1.02,
"learning_rate": 4.7569041571369324e-05,
"loss": 4.0276,
"step": 148480
},
{
"epoch": 1.02,
"learning_rate": 4.756067200266254e-05,
"loss": 4.0324,
"step": 148992
},
{
"epoch": 1.02,
"learning_rate": 4.755228605515202e-05,
"loss": 4.0299,
"step": 149504
},
{
"epoch": 1.02,
"learning_rate": 4.75439001076415e-05,
"loss": 4.0205,
"step": 150016
},
{
"epoch": 1.02,
"learning_rate": 4.753551416013098e-05,
"loss": 4.0417,
"step": 150528
},
{
"epoch": 1.02,
"learning_rate": 4.7527144591424196e-05,
"loss": 4.024,
"step": 151040
},
{
"epoch": 1.02,
"learning_rate": 4.751875864391367e-05,
"loss": 4.0196,
"step": 151552
},
{
"epoch": 1.02,
"learning_rate": 4.751037269640315e-05,
"loss": 4.023,
"step": 152064
},
{
"epoch": 1.02,
"learning_rate": 4.750198674889263e-05,
"loss": 4.0196,
"step": 152576
},
{
"epoch": 1.03,
"eval_loss": 4.027214527130127,
"eval_runtime": 303.1649,
"eval_samples_per_second": 1258.691,
"eval_steps_per_second": 39.335,
"step": 152640
},
{
"epoch": 0.0,
"learning_rate": 4.749360080138211e-05,
"loss": 4.0042,
"step": 153088
},
{
"epoch": 0.0,
"learning_rate": 4.748521485387159e-05,
"loss": 4.0197,
"step": 153600
},
{
"epoch": 0.0,
"learning_rate": 4.747682890636107e-05,
"loss": 4.0336,
"step": 154112
},
{
"epoch": 0.0,
"learning_rate": 4.746844295885055e-05,
"loss": 4.0161,
"step": 154624
},
{
"epoch": 0.0,
"learning_rate": 4.746005701134003e-05,
"loss": 4.028,
"step": 155136
},
{
"epoch": 0.0,
"learning_rate": 4.745167106382951e-05,
"loss": 4.0074,
"step": 155648
},
{
"epoch": 0.0,
"learning_rate": 4.744328511631899e-05,
"loss": 4.0208,
"step": 156160
},
{
"epoch": 0.0,
"learning_rate": 4.7434899168808475e-05,
"loss": 4.0005,
"step": 156672
},
{
"epoch": 0.0,
"learning_rate": 4.7426529600101684e-05,
"loss": 4.018,
"step": 157184
},
{
"epoch": 0.0,
"learning_rate": 4.7418143652591164e-05,
"loss": 4.0151,
"step": 157696
},
{
"epoch": 0.0,
"learning_rate": 4.7409757705080644e-05,
"loss": 4.0146,
"step": 158208
},
{
"epoch": 0.0,
"learning_rate": 4.7401371757570124e-05,
"loss": 4.021,
"step": 158720
},
{
"epoch": 0.0,
"learning_rate": 4.739300218886333e-05,
"loss": 4.0038,
"step": 159232
},
{
"epoch": 0.0,
"learning_rate": 4.738461624135281e-05,
"loss": 3.9984,
"step": 159744
},
{
"epoch": 0.0,
"learning_rate": 4.737623029384229e-05,
"loss": 3.9967,
"step": 160256
},
{
"epoch": 0.0,
"learning_rate": 4.736784434633177e-05,
"loss": 4.007,
"step": 160768
},
{
"epoch": 0.0,
"learning_rate": 4.735945839882125e-05,
"loss": 4.0104,
"step": 161280
},
{
"epoch": 0.0,
"learning_rate": 4.735107245131073e-05,
"loss": 4.0018,
"step": 161792
},
{
"epoch": 0.0,
"learning_rate": 4.734268650380021e-05,
"loss": 4.0057,
"step": 162304
},
{
"epoch": 0.0,
"learning_rate": 4.733430055628969e-05,
"loss": 4.0241,
"step": 162816
},
{
"epoch": 0.0,
"learning_rate": 4.732591460877917e-05,
"loss": 3.9987,
"step": 163328
},
{
"epoch": 0.0,
"learning_rate": 4.731752866126865e-05,
"loss": 3.9953,
"step": 163840
},
{
"epoch": 0.0,
"learning_rate": 4.730914271375813e-05,
"loss": 4.0014,
"step": 164352
},
{
"epoch": 0.0,
"learning_rate": 4.730075676624761e-05,
"loss": 4.0098,
"step": 164864
},
{
"epoch": 0.0,
"learning_rate": 4.729238719754082e-05,
"loss": 3.9925,
"step": 165376
},
{
"epoch": 0.0,
"learning_rate": 4.72840012500303e-05,
"loss": 3.9959,
"step": 165888
},
{
"epoch": 0.0,
"learning_rate": 4.727561530251978e-05,
"loss": 3.9986,
"step": 166400
},
{
"epoch": 0.0,
"learning_rate": 4.726722935500926e-05,
"loss": 3.9939,
"step": 166912
},
{
"epoch": 0.0,
"learning_rate": 4.725885978630247e-05,
"loss": 3.9803,
"step": 167424
},
{
"epoch": 0.01,
"learning_rate": 4.725047383879195e-05,
"loss": 3.9881,
"step": 167936
},
{
"epoch": 0.01,
"learning_rate": 4.724208789128143e-05,
"loss": 3.9904,
"step": 168448
},
{
"epoch": 0.01,
"learning_rate": 4.723370194377091e-05,
"loss": 3.9926,
"step": 168960
},
{
"epoch": 0.01,
"learning_rate": 4.7225332375064127e-05,
"loss": 4.0063,
"step": 169472
},
{
"epoch": 0.01,
"learning_rate": 4.7216946427553606e-05,
"loss": 3.9861,
"step": 169984
},
{
"epoch": 0.01,
"learning_rate": 4.7208560480043086e-05,
"loss": 3.9857,
"step": 170496
},
{
"epoch": 0.01,
"learning_rate": 4.7200174532532566e-05,
"loss": 3.9961,
"step": 171008
},
{
"epoch": 0.01,
"learning_rate": 4.7191804963825775e-05,
"loss": 3.9803,
"step": 171520
},
{
"epoch": 0.01,
"learning_rate": 4.7183419016315255e-05,
"loss": 3.9794,
"step": 172032
},
{
"epoch": 0.01,
"learning_rate": 4.7175033068804735e-05,
"loss": 3.9735,
"step": 172544
},
{
"epoch": 0.01,
"learning_rate": 4.7166663500097945e-05,
"loss": 3.9848,
"step": 173056
},
{
"epoch": 0.01,
"learning_rate": 4.7158277552587424e-05,
"loss": 3.9726,
"step": 173568
},
{
"epoch": 0.01,
"learning_rate": 4.7149891605076904e-05,
"loss": 3.9975,
"step": 174080
},
{
"epoch": 0.01,
"learning_rate": 4.7141505657566384e-05,
"loss": 3.9762,
"step": 174592
},
{
"epoch": 0.01,
"learning_rate": 4.7133119710055864e-05,
"loss": 3.9859,
"step": 175104
},
{
"epoch": 0.01,
"learning_rate": 4.712475014134908e-05,
"loss": 3.9798,
"step": 175616
},
{
"epoch": 0.01,
"learning_rate": 4.711636419383856e-05,
"loss": 3.9775,
"step": 176128
},
{
"epoch": 0.01,
"learning_rate": 4.710797824632804e-05,
"loss": 3.9793,
"step": 176640
},
{
"epoch": 0.01,
"learning_rate": 4.709959229881752e-05,
"loss": 3.9671,
"step": 177152
},
{
"epoch": 0.01,
"learning_rate": 4.709122273011073e-05,
"loss": 3.9628,
"step": 177664
},
{
"epoch": 0.01,
"learning_rate": 4.708283678260021e-05,
"loss": 3.9857,
"step": 178176
},
{
"epoch": 0.01,
"learning_rate": 4.707445083508969e-05,
"loss": 3.975,
"step": 178688
},
{
"epoch": 0.01,
"learning_rate": 4.70660812663829e-05,
"loss": 3.9646,
"step": 179200
},
{
"epoch": 0.01,
"learning_rate": 4.705769531887238e-05,
"loss": 3.9687,
"step": 179712
},
{
"epoch": 0.01,
"learning_rate": 4.704930937136186e-05,
"loss": 3.9757,
"step": 180224
},
{
"epoch": 0.01,
"learning_rate": 4.704092342385134e-05,
"loss": 3.9565,
"step": 180736
},
{
"epoch": 0.01,
"learning_rate": 4.7032553855144554e-05,
"loss": 3.9775,
"step": 181248
},
{
"epoch": 0.01,
"learning_rate": 4.7024167907634034e-05,
"loss": 3.9635,
"step": 181760
},
{
"epoch": 0.01,
"learning_rate": 4.7015781960123514e-05,
"loss": 3.9642,
"step": 182272
},
{
"epoch": 0.01,
"learning_rate": 4.7007396012612994e-05,
"loss": 3.981,
"step": 182784
},
{
"epoch": 0.01,
"learning_rate": 4.69990264439062e-05,
"loss": 3.9576,
"step": 183296
},
{
"epoch": 0.01,
"learning_rate": 4.699064049639568e-05,
"loss": 3.9537,
"step": 183808
},
{
"epoch": 0.01,
"learning_rate": 4.698225454888516e-05,
"loss": 3.9664,
"step": 184320
},
{
"epoch": 0.01,
"learning_rate": 4.697386860137464e-05,
"loss": 3.9505,
"step": 184832
},
{
"epoch": 0.01,
"learning_rate": 4.696549903266785e-05,
"loss": 3.9542,
"step": 185344
},
{
"epoch": 0.01,
"learning_rate": 4.695711308515733e-05,
"loss": 3.967,
"step": 185856
},
{
"epoch": 0.01,
"learning_rate": 4.694872713764681e-05,
"loss": 3.9693,
"step": 186368
},
{
"epoch": 0.01,
"learning_rate": 4.694034119013629e-05,
"loss": 3.9532,
"step": 186880
},
{
"epoch": 0.01,
"learning_rate": 4.693197162142951e-05,
"loss": 3.9554,
"step": 187392
},
{
"epoch": 0.01,
"learning_rate": 4.692360205272272e-05,
"loss": 3.943,
"step": 187904
},
{
"epoch": 0.01,
"learning_rate": 4.69152161052122e-05,
"loss": 3.9659,
"step": 188416
},
{
"epoch": 0.01,
"learning_rate": 4.6906830157701677e-05,
"loss": 3.9595,
"step": 188928
},
{
"epoch": 0.01,
"learning_rate": 4.6898444210191157e-05,
"loss": 3.9637,
"step": 189440
},
{
"epoch": 0.01,
"learning_rate": 4.6890058262680636e-05,
"loss": 3.9612,
"step": 189952
},
{
"epoch": 0.01,
"learning_rate": 4.6881672315170116e-05,
"loss": 3.9646,
"step": 190464
},
{
"epoch": 0.01,
"learning_rate": 4.6873286367659596e-05,
"loss": 3.9702,
"step": 190976
},
{
"epoch": 0.01,
"learning_rate": 4.6864916798952806e-05,
"loss": 3.945,
"step": 191488
},
{
"epoch": 0.01,
"learning_rate": 4.6856530851442285e-05,
"loss": 3.9567,
"step": 192000
},
{
"epoch": 0.01,
"learning_rate": 4.6848144903931765e-05,
"loss": 3.9618,
"step": 192512
},
{
"epoch": 0.01,
"learning_rate": 4.6839758956421245e-05,
"loss": 3.9447,
"step": 193024
},
{
"epoch": 0.01,
"learning_rate": 4.6831373008910725e-05,
"loss": 3.9652,
"step": 193536
},
{
"epoch": 0.01,
"learning_rate": 4.6822987061400205e-05,
"loss": 3.9597,
"step": 194048
},
{
"epoch": 0.01,
"learning_rate": 4.681460111388969e-05,
"loss": 3.9623,
"step": 194560
},
{
"epoch": 0.01,
"learning_rate": 4.6806215166379165e-05,
"loss": 3.9621,
"step": 195072
},
{
"epoch": 0.01,
"learning_rate": 4.679784559767238e-05,
"loss": 3.9386,
"step": 195584
},
{
"epoch": 0.01,
"learning_rate": 4.678945965016186e-05,
"loss": 3.9526,
"step": 196096
},
{
"epoch": 0.01,
"learning_rate": 4.678107370265134e-05,
"loss": 3.95,
"step": 196608
},
{
"epoch": 0.01,
"learning_rate": 4.6772687755140814e-05,
"loss": 3.9573,
"step": 197120
},
{
"epoch": 0.01,
"learning_rate": 4.676431818643403e-05,
"loss": 3.948,
"step": 197632
},
{
"epoch": 0.01,
"learning_rate": 4.675593223892351e-05,
"loss": 3.9478,
"step": 198144
},
{
"epoch": 0.02,
"learning_rate": 4.674754629141298e-05,
"loss": 3.9526,
"step": 198656
},
{
"epoch": 0.02,
"learning_rate": 4.673916034390246e-05,
"loss": 3.9485,
"step": 199168
},
{
"epoch": 0.02,
"learning_rate": 4.6730807153999415e-05,
"loss": 3.9359,
"step": 199680
},
{
"epoch": 0.02,
"learning_rate": 4.6722421206488895e-05,
"loss": 3.9448,
"step": 200192
},
{
"epoch": 0.02,
"learning_rate": 4.6714035258978375e-05,
"loss": 3.9502,
"step": 200704
},
{
"epoch": 0.02,
"learning_rate": 4.6705649311467855e-05,
"loss": 3.9569,
"step": 201216
},
{
"epoch": 0.02,
"learning_rate": 4.6697263363957335e-05,
"loss": 3.9457,
"step": 201728
},
{
"epoch": 0.02,
"learning_rate": 4.6688877416446815e-05,
"loss": 3.9271,
"step": 202240
},
{
"epoch": 0.02,
"learning_rate": 4.668049146893629e-05,
"loss": 3.9401,
"step": 202752
},
{
"epoch": 0.02,
"learning_rate": 4.667210552142577e-05,
"loss": 3.9477,
"step": 203264
},
{
"epoch": 0.02,
"learning_rate": 4.6663735952718984e-05,
"loss": 3.945,
"step": 203776
},
{
"epoch": 0.02,
"learning_rate": 4.665535000520846e-05,
"loss": 3.9413,
"step": 204288
},
{
"epoch": 0.02,
"learning_rate": 4.664696405769794e-05,
"loss": 3.9372,
"step": 204800
},
{
"epoch": 0.02,
"learning_rate": 4.663857811018742e-05,
"loss": 3.9262,
"step": 205312
},
{
"epoch": 0.02,
"learning_rate": 4.663020854148063e-05,
"loss": 3.9466,
"step": 205824
},
{
"epoch": 0.02,
"learning_rate": 4.662182259397011e-05,
"loss": 3.9217,
"step": 206336
},
{
"epoch": 0.02,
"learning_rate": 4.661343664645959e-05,
"loss": 3.9349,
"step": 206848
},
{
"epoch": 0.02,
"learning_rate": 4.660506707775281e-05,
"loss": 3.9328,
"step": 207360
},
{
"epoch": 0.02,
"learning_rate": 4.659668113024229e-05,
"loss": 3.9447,
"step": 207872
},
{
"epoch": 0.02,
"learning_rate": 4.658829518273176e-05,
"loss": 3.9283,
"step": 208384
},
{
"epoch": 0.02,
"learning_rate": 4.657990923522124e-05,
"loss": 3.9403,
"step": 208896
},
{
"epoch": 0.02,
"learning_rate": 4.657152328771072e-05,
"loss": 3.9289,
"step": 209408
},
{
"epoch": 0.02,
"learning_rate": 4.65631373402002e-05,
"loss": 3.9298,
"step": 209920
},
{
"epoch": 0.02,
"learning_rate": 4.655476777149341e-05,
"loss": 3.9339,
"step": 210432
},
{
"epoch": 0.02,
"learning_rate": 4.654638182398289e-05,
"loss": 3.9242,
"step": 210944
},
{
"epoch": 0.02,
"learning_rate": 4.653799587647237e-05,
"loss": 3.9428,
"step": 211456
},
{
"epoch": 0.02,
"learning_rate": 4.652960992896185e-05,
"loss": 3.9407,
"step": 211968
},
{
"epoch": 0.02,
"learning_rate": 4.6521240360255066e-05,
"loss": 3.9353,
"step": 212480
},
{
"epoch": 0.02,
"learning_rate": 4.6512854412744546e-05,
"loss": 3.9286,
"step": 212992
},
{
"epoch": 0.02,
"learning_rate": 4.6504468465234026e-05,
"loss": 3.9327,
"step": 213504
},
{
"epoch": 0.02,
"learning_rate": 4.6496082517723506e-05,
"loss": 3.9375,
"step": 214016
},
{
"epoch": 0.02,
"learning_rate": 4.6487696570212986e-05,
"loss": 3.9172,
"step": 214528
},
{
"epoch": 0.02,
"learning_rate": 4.6479310622702466e-05,
"loss": 3.9391,
"step": 215040
},
{
"epoch": 0.02,
"learning_rate": 4.6470924675191946e-05,
"loss": 3.9274,
"step": 215552
},
{
"epoch": 0.02,
"learning_rate": 4.6462538727681426e-05,
"loss": 3.9341,
"step": 216064
},
{
"epoch": 0.02,
"learning_rate": 4.6454169158974635e-05,
"loss": 3.9219,
"step": 216576
},
{
"epoch": 0.02,
"learning_rate": 4.6445783211464115e-05,
"loss": 3.9249,
"step": 217088
},
{
"epoch": 0.02,
"learning_rate": 4.6437397263953595e-05,
"loss": 3.9144,
"step": 217600
},
{
"epoch": 0.02,
"learning_rate": 4.6429011316443075e-05,
"loss": 3.9218,
"step": 218112
},
{
"epoch": 0.02,
"learning_rate": 4.642064174773629e-05,
"loss": 3.917,
"step": 218624
},
{
"epoch": 0.02,
"learning_rate": 4.641225580022577e-05,
"loss": 3.9322,
"step": 219136
},
{
"epoch": 0.02,
"learning_rate": 4.640386985271525e-05,
"loss": 3.9367,
"step": 219648
},
{
"epoch": 0.02,
"learning_rate": 4.639548390520473e-05,
"loss": 3.925,
"step": 220160
},
{
"epoch": 0.02,
"learning_rate": 4.638711433649794e-05,
"loss": 3.9161,
"step": 220672
},
{
"epoch": 0.02,
"learning_rate": 4.637872838898742e-05,
"loss": 3.9263,
"step": 221184
},
{
"epoch": 0.02,
"learning_rate": 4.63703424414769e-05,
"loss": 3.9276,
"step": 221696
},
{
"epoch": 0.02,
"learning_rate": 4.636195649396638e-05,
"loss": 3.9333,
"step": 222208
},
{
"epoch": 0.02,
"learning_rate": 4.635358692525959e-05,
"loss": 3.9284,
"step": 222720
},
{
"epoch": 0.02,
"learning_rate": 4.634520097774907e-05,
"loss": 3.9196,
"step": 223232
},
{
"epoch": 0.02,
"learning_rate": 4.633681503023855e-05,
"loss": 3.9243,
"step": 223744
},
{
"epoch": 0.02,
"learning_rate": 4.632842908272803e-05,
"loss": 3.9288,
"step": 224256
},
{
"epoch": 0.02,
"learning_rate": 4.6320059514021244e-05,
"loss": 3.9154,
"step": 224768
},
{
"epoch": 0.02,
"learning_rate": 4.6311673566510724e-05,
"loss": 3.9234,
"step": 225280
},
{
"epoch": 0.02,
"learning_rate": 4.6303287619000204e-05,
"loss": 3.919,
"step": 225792
},
{
"epoch": 0.02,
"learning_rate": 4.6294901671489684e-05,
"loss": 3.9117,
"step": 226304
},
{
"epoch": 0.02,
"learning_rate": 4.6286532102782893e-05,
"loss": 3.9328,
"step": 226816
},
{
"epoch": 0.02,
"learning_rate": 4.6278146155272373e-05,
"loss": 3.9178,
"step": 227328
},
{
"epoch": 0.02,
"learning_rate": 4.626976020776185e-05,
"loss": 3.9105,
"step": 227840
},
{
"epoch": 0.02,
"learning_rate": 4.626137426025133e-05,
"loss": 3.9151,
"step": 228352
},
{
"epoch": 0.02,
"learning_rate": 4.625298831274081e-05,
"loss": 3.9103,
"step": 228864
},
{
"epoch": 0.03,
"eval_loss": 3.95322847366333,
"eval_runtime": 303.7283,
"eval_samples_per_second": 1256.357,
"eval_steps_per_second": 39.262,
"step": 228960
}
],
"logging_steps": 512,
"max_steps": 3052726,
"num_train_epochs": 9223372036854775807,
"save_steps": 10,
"total_flos": 1.5869192680625357e+17,
"trial_name": null,
"trial_params": null
}