abhinavp's picture
Training in progress, epoch 0, checkpoint
6f2780b verified
raw
history blame
37.5 kB
{
"best_metric": 4.075600624084473,
"best_model_checkpoint": "/mmfs1/gscratch/stf/abhinavp/corpus-filtering/outputs/pp-mod-subj/transformer/1/checkpoints/checkpoint-152638",
"epoch": 0.025000278439663435,
"eval_steps": 10,
"global_step": 152638,
"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": 11.0824,
"step": 1
},
{
"epoch": 0.0,
"learning_rate": 4.999161405248948e-05,
"loss": 6.8357,
"step": 512
},
{
"epoch": 0.0,
"learning_rate": 4.998322810497896e-05,
"loss": 6.1776,
"step": 1024
},
{
"epoch": 0.0,
"learning_rate": 4.997484215746844e-05,
"loss": 5.9577,
"step": 1536
},
{
"epoch": 0.0,
"learning_rate": 4.996645620995792e-05,
"loss": 5.8117,
"step": 2048
},
{
"epoch": 0.0,
"learning_rate": 4.99580702624474e-05,
"loss": 5.6911,
"step": 2560
},
{
"epoch": 0.0,
"learning_rate": 4.994968431493688e-05,
"loss": 5.6043,
"step": 3072
},
{
"epoch": 0.0,
"learning_rate": 4.994129836742636e-05,
"loss": 5.5303,
"step": 3584
},
{
"epoch": 0.0,
"learning_rate": 4.993291241991584e-05,
"loss": 5.4703,
"step": 4096
},
{
"epoch": 0.0,
"learning_rate": 4.992452647240532e-05,
"loss": 5.4005,
"step": 4608
},
{
"epoch": 0.0,
"learning_rate": 4.99161405248948e-05,
"loss": 5.354,
"step": 5120
},
{
"epoch": 0.0,
"learning_rate": 4.990775457738428e-05,
"loss": 5.3111,
"step": 5632
},
{
"epoch": 0.0,
"learning_rate": 4.989938500867749e-05,
"loss": 5.2777,
"step": 6144
},
{
"epoch": 0.0,
"learning_rate": 4.989099906116697e-05,
"loss": 5.2148,
"step": 6656
},
{
"epoch": 0.0,
"learning_rate": 4.988261311365645e-05,
"loss": 5.1736,
"step": 7168
},
{
"epoch": 0.0,
"learning_rate": 4.987422716614593e-05,
"loss": 5.1441,
"step": 7680
},
{
"epoch": 0.0,
"learning_rate": 4.986584121863541e-05,
"loss": 5.1033,
"step": 8192
},
{
"epoch": 0.0,
"learning_rate": 4.985745527112489e-05,
"loss": 5.0755,
"step": 8704
},
{
"epoch": 0.0,
"learning_rate": 4.984906932361437e-05,
"loss": 5.0564,
"step": 9216
},
{
"epoch": 0.0,
"learning_rate": 4.984068337610385e-05,
"loss": 5.0223,
"step": 9728
},
{
"epoch": 0.0,
"learning_rate": 4.983229742859333e-05,
"loss": 5.0076,
"step": 10240
},
{
"epoch": 0.0,
"learning_rate": 4.9823927859886547e-05,
"loss": 4.9809,
"step": 10752
},
{
"epoch": 0.0,
"learning_rate": 4.9815541912376026e-05,
"loss": 4.9507,
"step": 11264
},
{
"epoch": 0.0,
"learning_rate": 4.9807155964865506e-05,
"loss": 4.945,
"step": 11776
},
{
"epoch": 0.0,
"learning_rate": 4.9798770017354986e-05,
"loss": 4.9172,
"step": 12288
},
{
"epoch": 0.0,
"learning_rate": 4.9790400448648195e-05,
"loss": 4.8919,
"step": 12800
},
{
"epoch": 0.0,
"learning_rate": 4.9782014501137675e-05,
"loss": 4.8786,
"step": 13312
},
{
"epoch": 0.0,
"learning_rate": 4.9773628553627155e-05,
"loss": 4.8688,
"step": 13824
},
{
"epoch": 0.0,
"learning_rate": 4.9765242606116635e-05,
"loss": 4.8318,
"step": 14336
},
{
"epoch": 0.0,
"learning_rate": 4.9756856658606115e-05,
"loss": 4.8239,
"step": 14848
},
{
"epoch": 0.01,
"learning_rate": 4.9748487089899324e-05,
"loss": 4.8015,
"step": 15360
},
{
"epoch": 0.01,
"learning_rate": 4.9740101142388804e-05,
"loss": 4.7918,
"step": 15872
},
{
"epoch": 0.01,
"learning_rate": 4.9731715194878284e-05,
"loss": 4.7918,
"step": 16384
},
{
"epoch": 0.01,
"learning_rate": 4.9723329247367764e-05,
"loss": 4.7701,
"step": 16896
},
{
"epoch": 0.01,
"learning_rate": 4.9714943299857244e-05,
"loss": 4.7594,
"step": 17408
},
{
"epoch": 0.01,
"learning_rate": 4.9706557352346724e-05,
"loss": 4.7484,
"step": 17920
},
{
"epoch": 0.01,
"learning_rate": 4.969818778363994e-05,
"loss": 4.738,
"step": 18432
},
{
"epoch": 0.01,
"learning_rate": 4.968980183612942e-05,
"loss": 4.7262,
"step": 18944
},
{
"epoch": 0.01,
"learning_rate": 4.96814158886189e-05,
"loss": 4.7095,
"step": 19456
},
{
"epoch": 0.01,
"learning_rate": 4.967302994110837e-05,
"loss": 4.7031,
"step": 19968
},
{
"epoch": 0.01,
"learning_rate": 4.966464399359785e-05,
"loss": 4.6845,
"step": 20480
},
{
"epoch": 0.01,
"learning_rate": 4.965627442489107e-05,
"loss": 4.6799,
"step": 20992
},
{
"epoch": 0.01,
"learning_rate": 4.964788847738054e-05,
"loss": 4.6555,
"step": 21504
},
{
"epoch": 0.01,
"learning_rate": 4.963950252987002e-05,
"loss": 4.6641,
"step": 22016
},
{
"epoch": 0.01,
"learning_rate": 4.96311165823595e-05,
"loss": 4.6574,
"step": 22528
},
{
"epoch": 0.01,
"learning_rate": 4.962274701365272e-05,
"loss": 4.6453,
"step": 23040
},
{
"epoch": 0.01,
"learning_rate": 4.96143610661422e-05,
"loss": 4.6271,
"step": 23552
},
{
"epoch": 0.01,
"learning_rate": 4.960597511863168e-05,
"loss": 4.6255,
"step": 24064
},
{
"epoch": 0.01,
"learning_rate": 4.959758917112116e-05,
"loss": 4.6206,
"step": 24576
},
{
"epoch": 0.01,
"learning_rate": 4.958920322361064e-05,
"loss": 4.6036,
"step": 25088
},
{
"epoch": 0.01,
"learning_rate": 4.958083365490385e-05,
"loss": 4.5919,
"step": 25600
},
{
"epoch": 0.01,
"learning_rate": 4.957246408619706e-05,
"loss": 4.5862,
"step": 26112
},
{
"epoch": 0.01,
"learning_rate": 4.956407813868654e-05,
"loss": 4.5943,
"step": 26624
},
{
"epoch": 0.01,
"learning_rate": 4.9555692191176016e-05,
"loss": 4.5637,
"step": 27136
},
{
"epoch": 0.01,
"learning_rate": 4.9547306243665496e-05,
"loss": 4.5854,
"step": 27648
},
{
"epoch": 0.01,
"learning_rate": 4.9538920296154976e-05,
"loss": 4.548,
"step": 28160
},
{
"epoch": 0.01,
"learning_rate": 4.9530534348644456e-05,
"loss": 4.5559,
"step": 28672
},
{
"epoch": 0.01,
"learning_rate": 4.9522148401133936e-05,
"loss": 4.5481,
"step": 29184
},
{
"epoch": 0.01,
"learning_rate": 4.951376245362342e-05,
"loss": 4.5443,
"step": 29696
},
{
"epoch": 0.01,
"learning_rate": 4.95053765061129e-05,
"loss": 4.5348,
"step": 30208
},
{
"epoch": 0.01,
"learning_rate": 4.949700693740611e-05,
"loss": 4.5278,
"step": 30720
},
{
"epoch": 0.01,
"learning_rate": 4.948862098989559e-05,
"loss": 4.5124,
"step": 31232
},
{
"epoch": 0.01,
"learning_rate": 4.948023504238507e-05,
"loss": 4.5213,
"step": 31744
},
{
"epoch": 0.01,
"learning_rate": 4.947184909487455e-05,
"loss": 4.5041,
"step": 32256
},
{
"epoch": 0.01,
"learning_rate": 4.946347952616776e-05,
"loss": 4.5092,
"step": 32768
},
{
"epoch": 0.01,
"learning_rate": 4.945509357865724e-05,
"loss": 4.5016,
"step": 33280
},
{
"epoch": 0.01,
"learning_rate": 4.944670763114672e-05,
"loss": 4.4919,
"step": 33792
},
{
"epoch": 0.01,
"learning_rate": 4.94383216836362e-05,
"loss": 4.4755,
"step": 34304
},
{
"epoch": 0.01,
"learning_rate": 4.942995211492941e-05,
"loss": 4.488,
"step": 34816
},
{
"epoch": 0.01,
"learning_rate": 4.942156616741889e-05,
"loss": 4.4791,
"step": 35328
},
{
"epoch": 0.01,
"learning_rate": 4.941318021990837e-05,
"loss": 4.4651,
"step": 35840
},
{
"epoch": 0.01,
"learning_rate": 4.9404794272397856e-05,
"loss": 4.4686,
"step": 36352
},
{
"epoch": 0.01,
"learning_rate": 4.9396424703691065e-05,
"loss": 4.4595,
"step": 36864
},
{
"epoch": 0.01,
"learning_rate": 4.9388055134984274e-05,
"loss": 4.46,
"step": 37376
},
{
"epoch": 0.01,
"learning_rate": 4.9379669187473754e-05,
"loss": 4.4631,
"step": 37888
},
{
"epoch": 0.01,
"learning_rate": 4.9371283239963234e-05,
"loss": 4.4547,
"step": 38400
},
{
"epoch": 0.01,
"learning_rate": 4.9362897292452714e-05,
"loss": 4.4447,
"step": 38912
},
{
"epoch": 0.01,
"learning_rate": 4.9354511344942194e-05,
"loss": 4.4288,
"step": 39424
},
{
"epoch": 0.01,
"learning_rate": 4.9346125397431674e-05,
"loss": 4.4387,
"step": 39936
},
{
"epoch": 0.01,
"learning_rate": 4.9337739449921154e-05,
"loss": 4.4332,
"step": 40448
},
{
"epoch": 0.01,
"learning_rate": 4.9329353502410634e-05,
"loss": 4.4354,
"step": 40960
},
{
"epoch": 0.01,
"learning_rate": 4.932098393370384e-05,
"loss": 4.4298,
"step": 41472
},
{
"epoch": 0.01,
"learning_rate": 4.931259798619332e-05,
"loss": 4.4169,
"step": 41984
},
{
"epoch": 0.01,
"learning_rate": 4.930421203868281e-05,
"loss": 4.3997,
"step": 42496
},
{
"epoch": 0.01,
"learning_rate": 4.929582609117229e-05,
"loss": 4.4053,
"step": 43008
},
{
"epoch": 0.01,
"learning_rate": 4.92874565224655e-05,
"loss": 4.4058,
"step": 43520
},
{
"epoch": 0.01,
"learning_rate": 4.927907057495498e-05,
"loss": 4.4077,
"step": 44032
},
{
"epoch": 0.01,
"learning_rate": 4.927068462744446e-05,
"loss": 4.4014,
"step": 44544
},
{
"epoch": 0.01,
"learning_rate": 4.926229867993394e-05,
"loss": 4.3956,
"step": 45056
},
{
"epoch": 0.01,
"learning_rate": 4.925392911122715e-05,
"loss": 4.3852,
"step": 45568
},
{
"epoch": 0.02,
"learning_rate": 4.924555954252036e-05,
"loss": 4.3817,
"step": 46080
},
{
"epoch": 0.02,
"learning_rate": 4.923717359500984e-05,
"loss": 4.3799,
"step": 46592
},
{
"epoch": 0.02,
"learning_rate": 4.922878764749932e-05,
"loss": 4.3888,
"step": 47104
},
{
"epoch": 0.02,
"learning_rate": 4.922041807879253e-05,
"loss": 4.378,
"step": 47616
},
{
"epoch": 0.02,
"learning_rate": 4.921203213128201e-05,
"loss": 4.3555,
"step": 48128
},
{
"epoch": 0.02,
"learning_rate": 4.920364618377149e-05,
"loss": 4.3608,
"step": 48640
},
{
"epoch": 0.02,
"learning_rate": 4.919526023626097e-05,
"loss": 4.3656,
"step": 49152
},
{
"epoch": 0.02,
"learning_rate": 4.918687428875045e-05,
"loss": 4.368,
"step": 49664
},
{
"epoch": 0.02,
"learning_rate": 4.917848834123993e-05,
"loss": 4.3607,
"step": 50176
},
{
"epoch": 0.02,
"learning_rate": 4.917010239372941e-05,
"loss": 4.3556,
"step": 50688
},
{
"epoch": 0.02,
"learning_rate": 4.916171644621889e-05,
"loss": 4.3521,
"step": 51200
},
{
"epoch": 0.02,
"learning_rate": 4.915333049870837e-05,
"loss": 4.3469,
"step": 51712
},
{
"epoch": 0.02,
"learning_rate": 4.914494455119785e-05,
"loss": 4.3519,
"step": 52224
},
{
"epoch": 0.02,
"learning_rate": 4.913657498249106e-05,
"loss": 4.3502,
"step": 52736
},
{
"epoch": 0.02,
"learning_rate": 4.912820541378427e-05,
"loss": 4.3302,
"step": 53248
},
{
"epoch": 0.02,
"learning_rate": 4.911981946627375e-05,
"loss": 4.3418,
"step": 53760
},
{
"epoch": 0.02,
"learning_rate": 4.911143351876323e-05,
"loss": 4.3199,
"step": 54272
},
{
"epoch": 0.02,
"learning_rate": 4.910304757125272e-05,
"loss": 4.3278,
"step": 54784
},
{
"epoch": 0.02,
"learning_rate": 4.90946616237422e-05,
"loss": 4.3257,
"step": 55296
},
{
"epoch": 0.02,
"learning_rate": 4.908627567623168e-05,
"loss": 4.3211,
"step": 55808
},
{
"epoch": 0.02,
"learning_rate": 4.907788972872115e-05,
"loss": 4.3224,
"step": 56320
},
{
"epoch": 0.02,
"learning_rate": 4.906950378121063e-05,
"loss": 4.3097,
"step": 56832
},
{
"epoch": 0.02,
"learning_rate": 4.906111783370011e-05,
"loss": 4.3059,
"step": 57344
},
{
"epoch": 0.02,
"learning_rate": 4.905273188618959e-05,
"loss": 4.311,
"step": 57856
},
{
"epoch": 0.02,
"learning_rate": 4.904434593867907e-05,
"loss": 4.3248,
"step": 58368
},
{
"epoch": 0.02,
"learning_rate": 4.903595999116855e-05,
"loss": 4.2966,
"step": 58880
},
{
"epoch": 0.02,
"learning_rate": 4.9027606801265495e-05,
"loss": 4.2922,
"step": 59392
},
{
"epoch": 0.02,
"learning_rate": 4.9019220853754975e-05,
"loss": 4.2948,
"step": 59904
},
{
"epoch": 0.02,
"learning_rate": 4.9010834906244455e-05,
"loss": 4.3067,
"step": 60416
},
{
"epoch": 0.02,
"learning_rate": 4.9002448958733935e-05,
"loss": 4.2965,
"step": 60928
},
{
"epoch": 0.02,
"learning_rate": 4.899407939002715e-05,
"loss": 4.2979,
"step": 61440
},
{
"epoch": 0.02,
"learning_rate": 4.8985693442516624e-05,
"loss": 4.2916,
"step": 61952
},
{
"epoch": 0.02,
"learning_rate": 4.8977307495006104e-05,
"loss": 4.295,
"step": 62464
},
{
"epoch": 0.02,
"learning_rate": 4.8968921547495584e-05,
"loss": 4.2834,
"step": 62976
},
{
"epoch": 0.02,
"learning_rate": 4.8960535599985064e-05,
"loss": 4.2836,
"step": 63488
},
{
"epoch": 0.02,
"learning_rate": 4.8952149652474544e-05,
"loss": 4.28,
"step": 64000
},
{
"epoch": 0.02,
"learning_rate": 4.8943763704964024e-05,
"loss": 4.2881,
"step": 64512
},
{
"epoch": 0.02,
"learning_rate": 4.8935377757453504e-05,
"loss": 4.2679,
"step": 65024
},
{
"epoch": 0.02,
"learning_rate": 4.892700818874671e-05,
"loss": 4.279,
"step": 65536
},
{
"epoch": 0.02,
"learning_rate": 4.891862224123619e-05,
"loss": 4.2736,
"step": 66048
},
{
"epoch": 0.02,
"learning_rate": 4.891023629372567e-05,
"loss": 4.2632,
"step": 66560
},
{
"epoch": 0.02,
"learning_rate": 4.890185034621515e-05,
"loss": 4.2739,
"step": 67072
},
{
"epoch": 0.02,
"learning_rate": 4.889346439870464e-05,
"loss": 4.267,
"step": 67584
},
{
"epoch": 0.02,
"learning_rate": 4.888507845119412e-05,
"loss": 4.2717,
"step": 68096
},
{
"epoch": 0.02,
"learning_rate": 4.88766925036836e-05,
"loss": 4.2684,
"step": 68608
},
{
"epoch": 0.02,
"learning_rate": 4.886832293497681e-05,
"loss": 4.2514,
"step": 69120
},
{
"epoch": 0.02,
"learning_rate": 4.885993698746629e-05,
"loss": 4.2485,
"step": 69632
},
{
"epoch": 0.02,
"learning_rate": 4.885155103995577e-05,
"loss": 4.2536,
"step": 70144
},
{
"epoch": 0.02,
"learning_rate": 4.884316509244525e-05,
"loss": 4.2617,
"step": 70656
},
{
"epoch": 0.02,
"learning_rate": 4.883479552373846e-05,
"loss": 4.2626,
"step": 71168
},
{
"epoch": 0.02,
"learning_rate": 4.882640957622794e-05,
"loss": 4.2537,
"step": 71680
},
{
"epoch": 0.02,
"learning_rate": 4.8818040007521146e-05,
"loss": 4.2441,
"step": 72192
},
{
"epoch": 0.02,
"learning_rate": 4.8809654060010626e-05,
"loss": 4.2434,
"step": 72704
},
{
"epoch": 0.02,
"learning_rate": 4.8801268112500106e-05,
"loss": 4.2469,
"step": 73216
},
{
"epoch": 0.02,
"learning_rate": 4.8792882164989586e-05,
"loss": 4.2431,
"step": 73728
},
{
"epoch": 0.02,
"learning_rate": 4.878452897508653e-05,
"loss": 4.2396,
"step": 74240
},
{
"epoch": 0.02,
"learning_rate": 4.877614302757601e-05,
"loss": 4.2554,
"step": 74752
},
{
"epoch": 0.02,
"learning_rate": 4.876775708006549e-05,
"loss": 4.2401,
"step": 75264
},
{
"epoch": 0.02,
"learning_rate": 4.875937113255497e-05,
"loss": 4.2308,
"step": 75776
},
{
"epoch": 0.02,
"learning_rate": 4.875098518504445e-05,
"loss": 4.2284,
"step": 76288
},
{
"epoch": 0.03,
"eval_loss": 4.246391296386719,
"eval_runtime": 573.4963,
"eval_samples_per_second": 665.377,
"eval_steps_per_second": 20.794,
"step": 76319
},
{
"epoch": 0.0,
"learning_rate": 4.874259923753393e-05,
"loss": 4.237,
"step": 76800
},
{
"epoch": 0.0,
"learning_rate": 4.873421329002341e-05,
"loss": 4.2313,
"step": 77312
},
{
"epoch": 0.0,
"learning_rate": 4.872582734251289e-05,
"loss": 4.2205,
"step": 77824
},
{
"epoch": 0.0,
"learning_rate": 4.871744139500237e-05,
"loss": 4.22,
"step": 78336
},
{
"epoch": 0.0,
"learning_rate": 4.870905544749185e-05,
"loss": 4.2119,
"step": 78848
},
{
"epoch": 0.0,
"learning_rate": 4.870066949998133e-05,
"loss": 4.2105,
"step": 79360
},
{
"epoch": 0.0,
"learning_rate": 4.869228355247081e-05,
"loss": 4.2065,
"step": 79872
},
{
"epoch": 0.0,
"learning_rate": 4.868389760496029e-05,
"loss": 4.2064,
"step": 80384
},
{
"epoch": 0.0,
"learning_rate": 4.867551165744977e-05,
"loss": 4.2131,
"step": 80896
},
{
"epoch": 0.0,
"learning_rate": 4.866712570993925e-05,
"loss": 4.2106,
"step": 81408
},
{
"epoch": 0.0,
"learning_rate": 4.865873976242873e-05,
"loss": 4.2062,
"step": 81920
},
{
"epoch": 0.0,
"learning_rate": 4.865035381491821e-05,
"loss": 4.2099,
"step": 82432
},
{
"epoch": 0.0,
"learning_rate": 4.864198424621142e-05,
"loss": 4.2015,
"step": 82944
},
{
"epoch": 0.0,
"learning_rate": 4.86335982987009e-05,
"loss": 4.196,
"step": 83456
},
{
"epoch": 0.0,
"learning_rate": 4.862521235119038e-05,
"loss": 4.1918,
"step": 83968
},
{
"epoch": 0.0,
"learning_rate": 4.861682640367986e-05,
"loss": 4.169,
"step": 84480
},
{
"epoch": 0.0,
"learning_rate": 4.860845683497307e-05,
"loss": 4.1883,
"step": 84992
},
{
"epoch": 0.0,
"learning_rate": 4.860007088746255e-05,
"loss": 4.1877,
"step": 85504
},
{
"epoch": 0.0,
"learning_rate": 4.859168493995203e-05,
"loss": 4.1868,
"step": 86016
},
{
"epoch": 0.0,
"learning_rate": 4.858329899244151e-05,
"loss": 4.1985,
"step": 86528
},
{
"epoch": 0.0,
"learning_rate": 4.8574929423734724e-05,
"loss": 4.1812,
"step": 87040
},
{
"epoch": 0.0,
"learning_rate": 4.8566543476224204e-05,
"loss": 4.1841,
"step": 87552
},
{
"epoch": 0.0,
"learning_rate": 4.8558157528713684e-05,
"loss": 4.1819,
"step": 88064
},
{
"epoch": 0.0,
"learning_rate": 4.8549771581203164e-05,
"loss": 4.1796,
"step": 88576
},
{
"epoch": 0.0,
"learning_rate": 4.8541385633692644e-05,
"loss": 4.174,
"step": 89088
},
{
"epoch": 0.0,
"learning_rate": 4.853301606498585e-05,
"loss": 4.1703,
"step": 89600
},
{
"epoch": 0.0,
"learning_rate": 4.852463011747533e-05,
"loss": 4.1778,
"step": 90112
},
{
"epoch": 0.0,
"learning_rate": 4.851624416996481e-05,
"loss": 4.1616,
"step": 90624
},
{
"epoch": 0.0,
"learning_rate": 4.850785822245429e-05,
"loss": 4.1672,
"step": 91136
},
{
"epoch": 0.01,
"learning_rate": 4.849947227494377e-05,
"loss": 4.1552,
"step": 91648
},
{
"epoch": 0.01,
"learning_rate": 4.849110270623698e-05,
"loss": 4.1584,
"step": 92160
},
{
"epoch": 0.01,
"learning_rate": 4.848271675872646e-05,
"loss": 4.1713,
"step": 92672
},
{
"epoch": 0.01,
"learning_rate": 4.847433081121595e-05,
"loss": 4.1624,
"step": 93184
},
{
"epoch": 0.01,
"learning_rate": 4.846594486370543e-05,
"loss": 4.1619,
"step": 93696
},
{
"epoch": 0.01,
"learning_rate": 4.845755891619491e-05,
"loss": 4.1592,
"step": 94208
},
{
"epoch": 0.01,
"learning_rate": 4.844918934748812e-05,
"loss": 4.159,
"step": 94720
},
{
"epoch": 0.01,
"learning_rate": 4.84408033999776e-05,
"loss": 4.1585,
"step": 95232
},
{
"epoch": 0.01,
"learning_rate": 4.843241745246708e-05,
"loss": 4.1557,
"step": 95744
},
{
"epoch": 0.01,
"learning_rate": 4.842403150495656e-05,
"loss": 4.1541,
"step": 96256
},
{
"epoch": 0.01,
"learning_rate": 4.841566193624977e-05,
"loss": 4.143,
"step": 96768
},
{
"epoch": 0.01,
"learning_rate": 4.840727598873925e-05,
"loss": 4.1512,
"step": 97280
},
{
"epoch": 0.01,
"learning_rate": 4.839889004122873e-05,
"loss": 4.1342,
"step": 97792
},
{
"epoch": 0.01,
"learning_rate": 4.839050409371821e-05,
"loss": 4.1475,
"step": 98304
},
{
"epoch": 0.01,
"learning_rate": 4.8382134525011416e-05,
"loss": 4.1511,
"step": 98816
},
{
"epoch": 0.01,
"learning_rate": 4.83737485775009e-05,
"loss": 4.1465,
"step": 99328
},
{
"epoch": 0.01,
"learning_rate": 4.836536262999038e-05,
"loss": 4.1391,
"step": 99840
},
{
"epoch": 0.01,
"learning_rate": 4.835697668247986e-05,
"loss": 4.1404,
"step": 100352
},
{
"epoch": 0.01,
"learning_rate": 4.834860711377307e-05,
"loss": 4.1448,
"step": 100864
},
{
"epoch": 0.01,
"learning_rate": 4.834022116626255e-05,
"loss": 4.123,
"step": 101376
},
{
"epoch": 0.01,
"learning_rate": 4.833183521875203e-05,
"loss": 4.1278,
"step": 101888
},
{
"epoch": 0.01,
"learning_rate": 4.832344927124151e-05,
"loss": 4.1266,
"step": 102400
},
{
"epoch": 0.01,
"learning_rate": 4.831507970253472e-05,
"loss": 4.1417,
"step": 102912
},
{
"epoch": 0.01,
"learning_rate": 4.83066937550242e-05,
"loss": 4.1182,
"step": 103424
},
{
"epoch": 0.01,
"learning_rate": 4.829830780751368e-05,
"loss": 4.1401,
"step": 103936
},
{
"epoch": 0.01,
"learning_rate": 4.828992186000316e-05,
"loss": 4.1116,
"step": 104448
},
{
"epoch": 0.01,
"learning_rate": 4.828155229129637e-05,
"loss": 4.1243,
"step": 104960
},
{
"epoch": 0.01,
"learning_rate": 4.8273166343785856e-05,
"loss": 4.1217,
"step": 105472
},
{
"epoch": 0.01,
"learning_rate": 4.8264780396275336e-05,
"loss": 4.1271,
"step": 105984
},
{
"epoch": 0.01,
"learning_rate": 4.8256394448764816e-05,
"loss": 4.1183,
"step": 106496
},
{
"epoch": 0.01,
"learning_rate": 4.8248024880058025e-05,
"loss": 4.122,
"step": 107008
},
{
"epoch": 0.01,
"learning_rate": 4.8239638932547505e-05,
"loss": 4.1095,
"step": 107520
},
{
"epoch": 0.01,
"learning_rate": 4.8231252985036985e-05,
"loss": 4.1183,
"step": 108032
},
{
"epoch": 0.01,
"learning_rate": 4.8222867037526465e-05,
"loss": 4.1112,
"step": 108544
},
{
"epoch": 0.01,
"learning_rate": 4.8214497468819674e-05,
"loss": 4.1159,
"step": 109056
},
{
"epoch": 0.01,
"learning_rate": 4.8206111521309154e-05,
"loss": 4.1114,
"step": 109568
},
{
"epoch": 0.01,
"learning_rate": 4.8197725573798634e-05,
"loss": 4.1088,
"step": 110080
},
{
"epoch": 0.01,
"learning_rate": 4.8189339626288114e-05,
"loss": 4.0931,
"step": 110592
},
{
"epoch": 0.01,
"learning_rate": 4.818097005758132e-05,
"loss": 4.1164,
"step": 111104
},
{
"epoch": 0.01,
"learning_rate": 4.817258411007081e-05,
"loss": 4.1034,
"step": 111616
},
{
"epoch": 0.01,
"learning_rate": 4.816419816256029e-05,
"loss": 4.0988,
"step": 112128
},
{
"epoch": 0.01,
"learning_rate": 4.815581221504977e-05,
"loss": 4.1055,
"step": 112640
},
{
"epoch": 0.01,
"learning_rate": 4.814744264634298e-05,
"loss": 4.1006,
"step": 113152
},
{
"epoch": 0.01,
"learning_rate": 4.813905669883246e-05,
"loss": 4.1031,
"step": 113664
},
{
"epoch": 0.01,
"learning_rate": 4.813067075132194e-05,
"loss": 4.1119,
"step": 114176
},
{
"epoch": 0.01,
"learning_rate": 4.812228480381141e-05,
"loss": 4.106,
"step": 114688
},
{
"epoch": 0.01,
"learning_rate": 4.811391523510463e-05,
"loss": 4.0991,
"step": 115200
},
{
"epoch": 0.01,
"learning_rate": 4.810552928759411e-05,
"loss": 4.0882,
"step": 115712
},
{
"epoch": 0.01,
"learning_rate": 4.809714334008359e-05,
"loss": 4.096,
"step": 116224
},
{
"epoch": 0.01,
"learning_rate": 4.808875739257306e-05,
"loss": 4.0982,
"step": 116736
},
{
"epoch": 0.01,
"learning_rate": 4.808038782386628e-05,
"loss": 4.106,
"step": 117248
},
{
"epoch": 0.01,
"learning_rate": 4.807200187635576e-05,
"loss": 4.0956,
"step": 117760
},
{
"epoch": 0.01,
"learning_rate": 4.8063615928845243e-05,
"loss": 4.091,
"step": 118272
},
{
"epoch": 0.01,
"learning_rate": 4.8055229981334717e-05,
"loss": 4.0749,
"step": 118784
},
{
"epoch": 0.01,
"learning_rate": 4.804686041262793e-05,
"loss": 4.0824,
"step": 119296
},
{
"epoch": 0.01,
"learning_rate": 4.803847446511741e-05,
"loss": 4.0839,
"step": 119808
},
{
"epoch": 0.01,
"learning_rate": 4.8030088517606886e-05,
"loss": 4.0866,
"step": 120320
},
{
"epoch": 0.01,
"learning_rate": 4.8021702570096366e-05,
"loss": 4.0924,
"step": 120832
},
{
"epoch": 0.01,
"learning_rate": 4.801333300138958e-05,
"loss": 4.0848,
"step": 121344
},
{
"epoch": 0.01,
"learning_rate": 4.800494705387906e-05,
"loss": 4.077,
"step": 121856
},
{
"epoch": 0.02,
"learning_rate": 4.7996561106368535e-05,
"loss": 4.0826,
"step": 122368
},
{
"epoch": 0.02,
"learning_rate": 4.7988175158858015e-05,
"loss": 4.0748,
"step": 122880
},
{
"epoch": 0.02,
"learning_rate": 4.797980559015123e-05,
"loss": 4.0845,
"step": 123392
},
{
"epoch": 0.02,
"learning_rate": 4.797141964264071e-05,
"loss": 4.0818,
"step": 123904
},
{
"epoch": 0.02,
"learning_rate": 4.7963050073933926e-05,
"loss": 4.0634,
"step": 124416
},
{
"epoch": 0.02,
"learning_rate": 4.7954664126423406e-05,
"loss": 4.0645,
"step": 124928
},
{
"epoch": 0.02,
"learning_rate": 4.7946278178912886e-05,
"loss": 4.074,
"step": 125440
},
{
"epoch": 0.02,
"learning_rate": 4.793789223140236e-05,
"loss": 4.0835,
"step": 125952
},
{
"epoch": 0.02,
"learning_rate": 4.792950628389184e-05,
"loss": 4.0693,
"step": 126464
},
{
"epoch": 0.02,
"learning_rate": 4.792112033638132e-05,
"loss": 4.0763,
"step": 126976
},
{
"epoch": 0.02,
"learning_rate": 4.79127343888708e-05,
"loss": 4.0691,
"step": 127488
},
{
"epoch": 0.02,
"learning_rate": 4.790436482016401e-05,
"loss": 4.0655,
"step": 128000
},
{
"epoch": 0.02,
"learning_rate": 4.789597887265349e-05,
"loss": 4.0743,
"step": 128512
},
{
"epoch": 0.02,
"learning_rate": 4.788759292514297e-05,
"loss": 4.072,
"step": 129024
},
{
"epoch": 0.02,
"learning_rate": 4.7879223356436184e-05,
"loss": 4.0564,
"step": 129536
},
{
"epoch": 0.02,
"learning_rate": 4.7870837408925664e-05,
"loss": 4.071,
"step": 130048
},
{
"epoch": 0.02,
"learning_rate": 4.7862451461415144e-05,
"loss": 4.053,
"step": 130560
},
{
"epoch": 0.02,
"learning_rate": 4.7854065513904624e-05,
"loss": 4.0558,
"step": 131072
},
{
"epoch": 0.02,
"learning_rate": 4.7845679566394104e-05,
"loss": 4.0643,
"step": 131584
},
{
"epoch": 0.02,
"learning_rate": 4.7837293618883584e-05,
"loss": 4.0531,
"step": 132096
},
{
"epoch": 0.02,
"learning_rate": 4.7828907671373064e-05,
"loss": 4.0594,
"step": 132608
},
{
"epoch": 0.02,
"learning_rate": 4.7820521723862544e-05,
"loss": 4.0521,
"step": 133120
},
{
"epoch": 0.02,
"learning_rate": 4.781215215515575e-05,
"loss": 4.048,
"step": 133632
},
{
"epoch": 0.02,
"learning_rate": 4.780376620764523e-05,
"loss": 4.0531,
"step": 134144
},
{
"epoch": 0.02,
"learning_rate": 4.779538026013471e-05,
"loss": 4.0684,
"step": 134656
},
{
"epoch": 0.02,
"learning_rate": 4.778699431262419e-05,
"loss": 4.046,
"step": 135168
},
{
"epoch": 0.02,
"learning_rate": 4.77786247439174e-05,
"loss": 4.0471,
"step": 135680
},
{
"epoch": 0.02,
"learning_rate": 4.777023879640689e-05,
"loss": 4.0454,
"step": 136192
},
{
"epoch": 0.02,
"learning_rate": 4.776185284889637e-05,
"loss": 4.0536,
"step": 136704
},
{
"epoch": 0.02,
"learning_rate": 4.775346690138585e-05,
"loss": 4.0514,
"step": 137216
},
{
"epoch": 0.02,
"learning_rate": 4.774509733267906e-05,
"loss": 4.0505,
"step": 137728
},
{
"epoch": 0.02,
"learning_rate": 4.773671138516854e-05,
"loss": 4.0502,
"step": 138240
},
{
"epoch": 0.02,
"learning_rate": 4.772832543765802e-05,
"loss": 4.0523,
"step": 138752
},
{
"epoch": 0.02,
"learning_rate": 4.77199394901475e-05,
"loss": 4.0454,
"step": 139264
},
{
"epoch": 0.02,
"learning_rate": 4.771155354263698e-05,
"loss": 4.0409,
"step": 139776
},
{
"epoch": 0.02,
"learning_rate": 4.770316759512646e-05,
"loss": 4.0435,
"step": 140288
},
{
"epoch": 0.02,
"learning_rate": 4.769478164761594e-05,
"loss": 4.0533,
"step": 140800
},
{
"epoch": 0.02,
"learning_rate": 4.768639570010542e-05,
"loss": 4.0365,
"step": 141312
},
{
"epoch": 0.02,
"learning_rate": 4.7678026131398626e-05,
"loss": 4.0386,
"step": 141824
},
{
"epoch": 0.02,
"learning_rate": 4.7669640183888106e-05,
"loss": 4.0511,
"step": 142336
},
{
"epoch": 0.02,
"learning_rate": 4.7661254236377586e-05,
"loss": 4.0298,
"step": 142848
},
{
"epoch": 0.02,
"learning_rate": 4.765286828886707e-05,
"loss": 4.0439,
"step": 143360
},
{
"epoch": 0.02,
"learning_rate": 4.764449872016028e-05,
"loss": 4.0387,
"step": 143872
},
{
"epoch": 0.02,
"learning_rate": 4.763611277264976e-05,
"loss": 4.048,
"step": 144384
},
{
"epoch": 0.02,
"learning_rate": 4.762772682513924e-05,
"loss": 4.042,
"step": 144896
},
{
"epoch": 0.02,
"learning_rate": 4.761934087762872e-05,
"loss": 4.0284,
"step": 145408
},
{
"epoch": 0.02,
"learning_rate": 4.761097130892193e-05,
"loss": 4.0277,
"step": 145920
},
{
"epoch": 0.02,
"learning_rate": 4.760258536141141e-05,
"loss": 4.0317,
"step": 146432
},
{
"epoch": 0.02,
"learning_rate": 4.759419941390089e-05,
"loss": 4.0422,
"step": 146944
},
{
"epoch": 0.02,
"learning_rate": 4.758581346639037e-05,
"loss": 4.0448,
"step": 147456
},
{
"epoch": 0.02,
"learning_rate": 4.757744389768358e-05,
"loss": 4.0402,
"step": 147968
},
{
"epoch": 0.02,
"learning_rate": 4.756905795017306e-05,
"loss": 4.025,
"step": 148480
},
{
"epoch": 0.02,
"learning_rate": 4.756067200266254e-05,
"loss": 4.0297,
"step": 148992
},
{
"epoch": 0.02,
"learning_rate": 4.755228605515202e-05,
"loss": 4.0341,
"step": 149504
},
{
"epoch": 0.02,
"learning_rate": 4.7543916486445236e-05,
"loss": 4.0318,
"step": 150016
},
{
"epoch": 0.02,
"learning_rate": 4.7535546917738445e-05,
"loss": 4.0291,
"step": 150528
},
{
"epoch": 0.02,
"learning_rate": 4.7527160970227925e-05,
"loss": 4.0469,
"step": 151040
},
{
"epoch": 0.02,
"learning_rate": 4.7518775022717405e-05,
"loss": 4.0252,
"step": 151552
},
{
"epoch": 0.02,
"learning_rate": 4.7510389075206885e-05,
"loss": 4.0279,
"step": 152064
},
{
"epoch": 0.02,
"learning_rate": 4.7502003127696365e-05,
"loss": 4.022,
"step": 152576
},
{
"epoch": 0.03,
"eval_loss": 4.075600624084473,
"eval_runtime": 570.4712,
"eval_samples_per_second": 668.905,
"eval_steps_per_second": 20.904,
"step": 152638
}
],
"logging_steps": 512,
"max_steps": 3052726,
"num_train_epochs": 9223372036854775807,
"save_steps": 10,
"total_flos": 9.894919341649306e+16,
"trial_name": null,
"trial_params": null
}