abhinavp's picture
Training in progress, epoch 0, checkpoint
fd0b6b6 verified
{
"best_metric": 4.190673351287842,
"best_model_checkpoint": "/mmfs1/gscratch/stf/abhinavp/corpus-filtering/outputs/full2/transformer/2/checkpoints/checkpoint-76320",
"epoch": 0.025000606015738065,
"eval_steps": 10,
"global_step": 76320,
"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": 10.9422,
"step": 1
},
{
"epoch": 0.0,
"learning_rate": 4.999161405248948e-05,
"loss": 6.8087,
"step": 512
},
{
"epoch": 0.0,
"learning_rate": 4.998322810497896e-05,
"loss": 6.1703,
"step": 1024
},
{
"epoch": 0.0,
"learning_rate": 4.997484215746844e-05,
"loss": 5.9608,
"step": 1536
},
{
"epoch": 0.0,
"learning_rate": 4.996645620995792e-05,
"loss": 5.7996,
"step": 2048
},
{
"epoch": 0.0,
"learning_rate": 4.99580702624474e-05,
"loss": 5.6916,
"step": 2560
},
{
"epoch": 0.0,
"learning_rate": 4.994968431493688e-05,
"loss": 5.5901,
"step": 3072
},
{
"epoch": 0.0,
"learning_rate": 4.994129836742636e-05,
"loss": 5.523,
"step": 3584
},
{
"epoch": 0.0,
"learning_rate": 4.993291241991584e-05,
"loss": 5.4358,
"step": 4096
},
{
"epoch": 0.0,
"learning_rate": 4.992452647240532e-05,
"loss": 5.3874,
"step": 4608
},
{
"epoch": 0.0,
"learning_rate": 4.99161405248948e-05,
"loss": 5.3369,
"step": 5120
},
{
"epoch": 0.0,
"learning_rate": 4.990775457738428e-05,
"loss": 5.314,
"step": 5632
},
{
"epoch": 0.0,
"learning_rate": 4.989938500867749e-05,
"loss": 5.2569,
"step": 6144
},
{
"epoch": 0.0,
"learning_rate": 4.989099906116697e-05,
"loss": 5.1946,
"step": 6656
},
{
"epoch": 0.0,
"learning_rate": 4.988261311365645e-05,
"loss": 5.1689,
"step": 7168
},
{
"epoch": 0.0,
"learning_rate": 4.987422716614593e-05,
"loss": 5.1321,
"step": 7680
},
{
"epoch": 0.0,
"learning_rate": 4.986584121863541e-05,
"loss": 5.1015,
"step": 8192
},
{
"epoch": 0.0,
"learning_rate": 4.985745527112489e-05,
"loss": 5.0742,
"step": 8704
},
{
"epoch": 0.0,
"learning_rate": 4.984906932361437e-05,
"loss": 5.0484,
"step": 9216
},
{
"epoch": 0.0,
"learning_rate": 4.984068337610385e-05,
"loss": 5.0067,
"step": 9728
},
{
"epoch": 0.0,
"learning_rate": 4.983229742859333e-05,
"loss": 5.0017,
"step": 10240
},
{
"epoch": 0.0,
"learning_rate": 4.9823927859886547e-05,
"loss": 4.9672,
"step": 10752
},
{
"epoch": 0.0,
"learning_rate": 4.9815541912376026e-05,
"loss": 4.9451,
"step": 11264
},
{
"epoch": 0.0,
"learning_rate": 4.9807155964865506e-05,
"loss": 4.9183,
"step": 11776
},
{
"epoch": 0.0,
"learning_rate": 4.9798770017354986e-05,
"loss": 4.9163,
"step": 12288
},
{
"epoch": 0.0,
"learning_rate": 4.9790400448648195e-05,
"loss": 4.8894,
"step": 12800
},
{
"epoch": 0.0,
"learning_rate": 4.9782014501137675e-05,
"loss": 4.8614,
"step": 13312
},
{
"epoch": 0.0,
"learning_rate": 4.9773628553627155e-05,
"loss": 4.8452,
"step": 13824
},
{
"epoch": 0.0,
"learning_rate": 4.9765242606116635e-05,
"loss": 4.8264,
"step": 14336
},
{
"epoch": 0.0,
"learning_rate": 4.9756873037409844e-05,
"loss": 4.8155,
"step": 14848
},
{
"epoch": 0.01,
"learning_rate": 4.9748487089899324e-05,
"loss": 4.7933,
"step": 15360
},
{
"epoch": 0.01,
"learning_rate": 4.9740101142388804e-05,
"loss": 4.7857,
"step": 15872
},
{
"epoch": 0.01,
"learning_rate": 4.9731715194878284e-05,
"loss": 4.7723,
"step": 16384
},
{
"epoch": 0.01,
"learning_rate": 4.9723329247367764e-05,
"loss": 4.7592,
"step": 16896
},
{
"epoch": 0.01,
"learning_rate": 4.971495967866098e-05,
"loss": 4.7553,
"step": 17408
},
{
"epoch": 0.01,
"learning_rate": 4.970657373115046e-05,
"loss": 4.7416,
"step": 17920
},
{
"epoch": 0.01,
"learning_rate": 4.969820416244367e-05,
"loss": 4.7268,
"step": 18432
},
{
"epoch": 0.01,
"learning_rate": 4.968981821493315e-05,
"loss": 4.7002,
"step": 18944
},
{
"epoch": 0.01,
"learning_rate": 4.968143226742263e-05,
"loss": 4.6953,
"step": 19456
},
{
"epoch": 0.01,
"learning_rate": 4.967304631991211e-05,
"loss": 4.6778,
"step": 19968
},
{
"epoch": 0.01,
"learning_rate": 4.966466037240159e-05,
"loss": 4.6755,
"step": 20480
},
{
"epoch": 0.01,
"learning_rate": 4.965627442489107e-05,
"loss": 4.66,
"step": 20992
},
{
"epoch": 0.01,
"learning_rate": 4.964788847738054e-05,
"loss": 4.6553,
"step": 21504
},
{
"epoch": 0.01,
"learning_rate": 4.963950252987002e-05,
"loss": 4.6391,
"step": 22016
},
{
"epoch": 0.01,
"learning_rate": 4.96311165823595e-05,
"loss": 4.6446,
"step": 22528
},
{
"epoch": 0.01,
"learning_rate": 4.962274701365272e-05,
"loss": 4.628,
"step": 23040
},
{
"epoch": 0.01,
"learning_rate": 4.96143610661422e-05,
"loss": 4.6295,
"step": 23552
},
{
"epoch": 0.01,
"learning_rate": 4.960597511863168e-05,
"loss": 4.6205,
"step": 24064
},
{
"epoch": 0.01,
"learning_rate": 4.959758917112116e-05,
"loss": 4.5924,
"step": 24576
},
{
"epoch": 0.01,
"learning_rate": 4.9589219602414374e-05,
"loss": 4.5849,
"step": 25088
},
{
"epoch": 0.01,
"learning_rate": 4.958083365490385e-05,
"loss": 4.5966,
"step": 25600
},
{
"epoch": 0.01,
"learning_rate": 4.957244770739333e-05,
"loss": 4.5827,
"step": 26112
},
{
"epoch": 0.01,
"learning_rate": 4.956406175988281e-05,
"loss": 4.5687,
"step": 26624
},
{
"epoch": 0.01,
"learning_rate": 4.9555692191176016e-05,
"loss": 4.5534,
"step": 27136
},
{
"epoch": 0.01,
"learning_rate": 4.954732262246923e-05,
"loss": 4.5586,
"step": 27648
},
{
"epoch": 0.01,
"learning_rate": 4.953893667495871e-05,
"loss": 4.541,
"step": 28160
},
{
"epoch": 0.01,
"learning_rate": 4.953055072744819e-05,
"loss": 4.5533,
"step": 28672
},
{
"epoch": 0.01,
"learning_rate": 4.952216477993767e-05,
"loss": 4.5209,
"step": 29184
},
{
"epoch": 0.01,
"learning_rate": 4.951377883242715e-05,
"loss": 4.5315,
"step": 29696
},
{
"epoch": 0.01,
"learning_rate": 4.950539288491663e-05,
"loss": 4.5306,
"step": 30208
},
{
"epoch": 0.01,
"learning_rate": 4.949700693740611e-05,
"loss": 4.5065,
"step": 30720
},
{
"epoch": 0.01,
"learning_rate": 4.948862098989559e-05,
"loss": 4.5231,
"step": 31232
},
{
"epoch": 0.01,
"learning_rate": 4.94802514211888e-05,
"loss": 4.4931,
"step": 31744
},
{
"epoch": 0.01,
"learning_rate": 4.947186547367828e-05,
"loss": 4.4943,
"step": 32256
},
{
"epoch": 0.01,
"learning_rate": 4.946347952616776e-05,
"loss": 4.4831,
"step": 32768
},
{
"epoch": 0.01,
"learning_rate": 4.945509357865724e-05,
"loss": 4.499,
"step": 33280
},
{
"epoch": 0.01,
"learning_rate": 4.944670763114672e-05,
"loss": 4.4698,
"step": 33792
},
{
"epoch": 0.01,
"learning_rate": 4.943833806243993e-05,
"loss": 4.4754,
"step": 34304
},
{
"epoch": 0.01,
"learning_rate": 4.942995211492941e-05,
"loss": 4.4598,
"step": 34816
},
{
"epoch": 0.01,
"learning_rate": 4.942156616741889e-05,
"loss": 4.4567,
"step": 35328
},
{
"epoch": 0.01,
"learning_rate": 4.941318021990837e-05,
"loss": 4.4699,
"step": 35840
},
{
"epoch": 0.01,
"learning_rate": 4.9404810651201585e-05,
"loss": 4.4622,
"step": 36352
},
{
"epoch": 0.01,
"learning_rate": 4.9396424703691065e-05,
"loss": 4.4519,
"step": 36864
},
{
"epoch": 0.01,
"learning_rate": 4.9388038756180545e-05,
"loss": 4.4515,
"step": 37376
},
{
"epoch": 0.01,
"learning_rate": 4.9379652808670025e-05,
"loss": 4.4598,
"step": 37888
},
{
"epoch": 0.01,
"learning_rate": 4.9371283239963234e-05,
"loss": 4.4314,
"step": 38400
},
{
"epoch": 0.01,
"learning_rate": 4.936291367125644e-05,
"loss": 4.4479,
"step": 38912
},
{
"epoch": 0.01,
"learning_rate": 4.935452772374592e-05,
"loss": 4.4221,
"step": 39424
},
{
"epoch": 0.01,
"learning_rate": 4.93461417762354e-05,
"loss": 4.4205,
"step": 39936
},
{
"epoch": 0.01,
"learning_rate": 4.933775582872488e-05,
"loss": 4.4192,
"step": 40448
},
{
"epoch": 0.01,
"learning_rate": 4.932936988121436e-05,
"loss": 4.4218,
"step": 40960
},
{
"epoch": 0.01,
"learning_rate": 4.932098393370384e-05,
"loss": 4.4141,
"step": 41472
},
{
"epoch": 0.01,
"learning_rate": 4.931259798619332e-05,
"loss": 4.4203,
"step": 41984
},
{
"epoch": 0.01,
"learning_rate": 4.930422841748654e-05,
"loss": 4.405,
"step": 42496
},
{
"epoch": 0.01,
"learning_rate": 4.929584246997602e-05,
"loss": 4.3855,
"step": 43008
},
{
"epoch": 0.01,
"learning_rate": 4.92874565224655e-05,
"loss": 4.3888,
"step": 43520
},
{
"epoch": 0.01,
"learning_rate": 4.927907057495498e-05,
"loss": 4.3976,
"step": 44032
},
{
"epoch": 0.01,
"learning_rate": 4.927068462744446e-05,
"loss": 4.3846,
"step": 44544
},
{
"epoch": 0.01,
"learning_rate": 4.92623314375414e-05,
"loss": 4.384,
"step": 45056
},
{
"epoch": 0.01,
"learning_rate": 4.925394549003088e-05,
"loss": 4.3887,
"step": 45568
},
{
"epoch": 0.02,
"learning_rate": 4.924555954252036e-05,
"loss": 4.3706,
"step": 46080
},
{
"epoch": 0.02,
"learning_rate": 4.923717359500984e-05,
"loss": 4.3766,
"step": 46592
},
{
"epoch": 0.02,
"learning_rate": 4.9228804026303046e-05,
"loss": 4.3697,
"step": 47104
},
{
"epoch": 0.02,
"learning_rate": 4.922041807879253e-05,
"loss": 4.3729,
"step": 47616
},
{
"epoch": 0.02,
"learning_rate": 4.921203213128201e-05,
"loss": 4.3717,
"step": 48128
},
{
"epoch": 0.02,
"learning_rate": 4.920364618377149e-05,
"loss": 4.3673,
"step": 48640
},
{
"epoch": 0.02,
"learning_rate": 4.919526023626097e-05,
"loss": 4.3456,
"step": 49152
},
{
"epoch": 0.02,
"learning_rate": 4.918687428875045e-05,
"loss": 4.3532,
"step": 49664
},
{
"epoch": 0.02,
"learning_rate": 4.917848834123993e-05,
"loss": 4.3469,
"step": 50176
},
{
"epoch": 0.02,
"learning_rate": 4.917010239372941e-05,
"loss": 4.3496,
"step": 50688
},
{
"epoch": 0.02,
"learning_rate": 4.916171644621889e-05,
"loss": 4.3527,
"step": 51200
},
{
"epoch": 0.02,
"learning_rate": 4.915333049870837e-05,
"loss": 4.3321,
"step": 51712
},
{
"epoch": 0.02,
"learning_rate": 4.914494455119785e-05,
"loss": 4.3376,
"step": 52224
},
{
"epoch": 0.02,
"learning_rate": 4.9136558603687325e-05,
"loss": 4.3388,
"step": 52736
},
{
"epoch": 0.02,
"learning_rate": 4.912818903498054e-05,
"loss": 4.331,
"step": 53248
},
{
"epoch": 0.02,
"learning_rate": 4.911980308747002e-05,
"loss": 4.3203,
"step": 53760
},
{
"epoch": 0.02,
"learning_rate": 4.911143351876323e-05,
"loss": 4.3241,
"step": 54272
},
{
"epoch": 0.02,
"learning_rate": 4.910304757125272e-05,
"loss": 4.3132,
"step": 54784
},
{
"epoch": 0.02,
"learning_rate": 4.90946616237422e-05,
"loss": 4.3117,
"step": 55296
},
{
"epoch": 0.02,
"learning_rate": 4.908627567623168e-05,
"loss": 4.3103,
"step": 55808
},
{
"epoch": 0.02,
"learning_rate": 4.907788972872115e-05,
"loss": 4.3093,
"step": 56320
},
{
"epoch": 0.02,
"learning_rate": 4.906950378121063e-05,
"loss": 4.3054,
"step": 56832
},
{
"epoch": 0.02,
"learning_rate": 4.906111783370011e-05,
"loss": 4.3098,
"step": 57344
},
{
"epoch": 0.02,
"learning_rate": 4.905273188618959e-05,
"loss": 4.3007,
"step": 57856
},
{
"epoch": 0.02,
"learning_rate": 4.90443623174828e-05,
"loss": 4.3037,
"step": 58368
},
{
"epoch": 0.02,
"learning_rate": 4.903597636997228e-05,
"loss": 4.3106,
"step": 58880
},
{
"epoch": 0.02,
"learning_rate": 4.9027606801265495e-05,
"loss": 4.3055,
"step": 59392
},
{
"epoch": 0.02,
"learning_rate": 4.9019220853754975e-05,
"loss": 4.3007,
"step": 59904
},
{
"epoch": 0.02,
"learning_rate": 4.9010834906244455e-05,
"loss": 4.2906,
"step": 60416
},
{
"epoch": 0.02,
"learning_rate": 4.9002448958733935e-05,
"loss": 4.2898,
"step": 60928
},
{
"epoch": 0.02,
"learning_rate": 4.8994063011223415e-05,
"loss": 4.2885,
"step": 61440
},
{
"epoch": 0.02,
"learning_rate": 4.8985677063712895e-05,
"loss": 4.276,
"step": 61952
},
{
"epoch": 0.02,
"learning_rate": 4.8977307495006104e-05,
"loss": 4.2818,
"step": 62464
},
{
"epoch": 0.02,
"learning_rate": 4.8968921547495584e-05,
"loss": 4.2757,
"step": 62976
},
{
"epoch": 0.02,
"learning_rate": 4.8960535599985064e-05,
"loss": 4.2805,
"step": 63488
},
{
"epoch": 0.02,
"learning_rate": 4.8952149652474544e-05,
"loss": 4.2717,
"step": 64000
},
{
"epoch": 0.02,
"learning_rate": 4.8943763704964024e-05,
"loss": 4.2796,
"step": 64512
},
{
"epoch": 0.02,
"learning_rate": 4.8935377757453504e-05,
"loss": 4.2732,
"step": 65024
},
{
"epoch": 0.02,
"learning_rate": 4.8926991809942984e-05,
"loss": 4.2653,
"step": 65536
},
{
"epoch": 0.02,
"learning_rate": 4.891862224123619e-05,
"loss": 4.2604,
"step": 66048
},
{
"epoch": 0.02,
"learning_rate": 4.891023629372567e-05,
"loss": 4.2761,
"step": 66560
},
{
"epoch": 0.02,
"learning_rate": 4.890185034621515e-05,
"loss": 4.2671,
"step": 67072
},
{
"epoch": 0.02,
"learning_rate": 4.889346439870464e-05,
"loss": 4.2593,
"step": 67584
},
{
"epoch": 0.02,
"learning_rate": 4.888507845119412e-05,
"loss": 4.261,
"step": 68096
},
{
"epoch": 0.02,
"learning_rate": 4.887670888248733e-05,
"loss": 4.2641,
"step": 68608
},
{
"epoch": 0.02,
"learning_rate": 4.886832293497681e-05,
"loss": 4.2588,
"step": 69120
},
{
"epoch": 0.02,
"learning_rate": 4.885993698746629e-05,
"loss": 4.2614,
"step": 69632
},
{
"epoch": 0.02,
"learning_rate": 4.885155103995577e-05,
"loss": 4.2507,
"step": 70144
},
{
"epoch": 0.02,
"learning_rate": 4.884316509244525e-05,
"loss": 4.2497,
"step": 70656
},
{
"epoch": 0.02,
"learning_rate": 4.883477914493473e-05,
"loss": 4.2491,
"step": 71168
},
{
"epoch": 0.02,
"learning_rate": 4.882639319742421e-05,
"loss": 4.2409,
"step": 71680
},
{
"epoch": 0.02,
"learning_rate": 4.881800724991369e-05,
"loss": 4.2378,
"step": 72192
},
{
"epoch": 0.02,
"learning_rate": 4.88096376812069e-05,
"loss": 4.2372,
"step": 72704
},
{
"epoch": 0.02,
"learning_rate": 4.880125173369638e-05,
"loss": 4.2402,
"step": 73216
},
{
"epoch": 0.02,
"learning_rate": 4.879286578618586e-05,
"loss": 4.231,
"step": 73728
},
{
"epoch": 0.02,
"learning_rate": 4.878449621747907e-05,
"loss": 4.2402,
"step": 74240
},
{
"epoch": 0.02,
"learning_rate": 4.877611026996855e-05,
"loss": 4.2315,
"step": 74752
},
{
"epoch": 0.02,
"learning_rate": 4.876772432245803e-05,
"loss": 4.2336,
"step": 75264
},
{
"epoch": 0.02,
"learning_rate": 4.875933837494751e-05,
"loss": 4.2379,
"step": 75776
},
{
"epoch": 0.02,
"learning_rate": 4.8750952427436986e-05,
"loss": 4.22,
"step": 76288
},
{
"epoch": 0.03,
"eval_loss": 4.190673351287842,
"eval_runtime": 313.8813,
"eval_samples_per_second": 1215.718,
"eval_steps_per_second": 37.992,
"step": 76320
}
],
"logging_steps": 512,
"max_steps": 3052726,
"num_train_epochs": 9223372036854775807,
"save_steps": 10,
"total_flos": 5.311679277613363e+16,
"trial_name": null,
"trial_params": null
}