abhinavp's picture
Training in progress, epoch 1, checkpoint
21ebd8f verified
raw
history blame
37.6 kB
{
"best_metric": 4.029199600219727,
"best_model_checkpoint": "/mmfs1/gscratch/stf/abhinavp/corpus-filtering/outputs/existential-there-quantifier/transformer/2/checkpoints/checkpoint-152640",
"epoch": 1.0250006060157382,
"eval_steps": 10,
"global_step": 152640,
"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": 11.016,
"step": 1
},
{
"epoch": 0.0,
"learning_rate": 4.999161405248948e-05,
"loss": 6.8135,
"step": 512
},
{
"epoch": 0.0,
"learning_rate": 4.998322810497896e-05,
"loss": 6.1935,
"step": 1024
},
{
"epoch": 0.0,
"learning_rate": 4.997484215746844e-05,
"loss": 5.989,
"step": 1536
},
{
"epoch": 0.0,
"learning_rate": 4.996645620995792e-05,
"loss": 5.8212,
"step": 2048
},
{
"epoch": 0.0,
"learning_rate": 4.99580702624474e-05,
"loss": 5.7126,
"step": 2560
},
{
"epoch": 0.0,
"learning_rate": 4.994968431493688e-05,
"loss": 5.6089,
"step": 3072
},
{
"epoch": 0.0,
"learning_rate": 4.994129836742636e-05,
"loss": 5.5426,
"step": 3584
},
{
"epoch": 0.0,
"learning_rate": 4.993291241991584e-05,
"loss": 5.4525,
"step": 4096
},
{
"epoch": 0.0,
"learning_rate": 4.992452647240532e-05,
"loss": 5.408,
"step": 4608
},
{
"epoch": 0.0,
"learning_rate": 4.99161405248948e-05,
"loss": 5.357,
"step": 5120
},
{
"epoch": 0.0,
"learning_rate": 4.990775457738428e-05,
"loss": 5.3155,
"step": 5632
},
{
"epoch": 0.0,
"learning_rate": 4.989936862987376e-05,
"loss": 5.2618,
"step": 6144
},
{
"epoch": 0.0,
"learning_rate": 4.989099906116697e-05,
"loss": 5.2164,
"step": 6656
},
{
"epoch": 0.0,
"learning_rate": 4.988261311365645e-05,
"loss": 5.1741,
"step": 7168
},
{
"epoch": 0.0,
"learning_rate": 4.987422716614593e-05,
"loss": 5.1411,
"step": 7680
},
{
"epoch": 0.0,
"learning_rate": 4.986584121863541e-05,
"loss": 5.1112,
"step": 8192
},
{
"epoch": 0.0,
"learning_rate": 4.985745527112489e-05,
"loss": 5.0959,
"step": 8704
},
{
"epoch": 0.0,
"learning_rate": 4.984906932361437e-05,
"loss": 5.0536,
"step": 9216
},
{
"epoch": 0.0,
"learning_rate": 4.984068337610385e-05,
"loss": 5.0335,
"step": 9728
},
{
"epoch": 0.0,
"learning_rate": 4.983229742859333e-05,
"loss": 5.0219,
"step": 10240
},
{
"epoch": 0.0,
"learning_rate": 4.9823927859886547e-05,
"loss": 4.9772,
"step": 10752
},
{
"epoch": 0.0,
"learning_rate": 4.9815541912376026e-05,
"loss": 4.9516,
"step": 11264
},
{
"epoch": 0.0,
"learning_rate": 4.9807155964865506e-05,
"loss": 4.9374,
"step": 11776
},
{
"epoch": 0.0,
"learning_rate": 4.9798770017354986e-05,
"loss": 4.9288,
"step": 12288
},
{
"epoch": 0.0,
"learning_rate": 4.9790400448648195e-05,
"loss": 4.8946,
"step": 12800
},
{
"epoch": 0.0,
"learning_rate": 4.9782030879941405e-05,
"loss": 4.8738,
"step": 13312
},
{
"epoch": 0.0,
"learning_rate": 4.9773644932430885e-05,
"loss": 4.8655,
"step": 13824
},
{
"epoch": 0.0,
"learning_rate": 4.9765258984920364e-05,
"loss": 4.8441,
"step": 14336
},
{
"epoch": 0.0,
"learning_rate": 4.9756873037409844e-05,
"loss": 4.8186,
"step": 14848
},
{
"epoch": 0.01,
"learning_rate": 4.9748487089899324e-05,
"loss": 4.8018,
"step": 15360
},
{
"epoch": 0.01,
"learning_rate": 4.9740101142388804e-05,
"loss": 4.7983,
"step": 15872
},
{
"epoch": 0.01,
"learning_rate": 4.9731715194878284e-05,
"loss": 4.7853,
"step": 16384
},
{
"epoch": 0.01,
"learning_rate": 4.9723329247367764e-05,
"loss": 4.7813,
"step": 16896
},
{
"epoch": 0.01,
"learning_rate": 4.9714943299857244e-05,
"loss": 4.7579,
"step": 17408
},
{
"epoch": 0.01,
"learning_rate": 4.970657373115046e-05,
"loss": 4.7416,
"step": 17920
},
{
"epoch": 0.01,
"learning_rate": 4.969818778363994e-05,
"loss": 4.7452,
"step": 18432
},
{
"epoch": 0.01,
"learning_rate": 4.968981821493315e-05,
"loss": 4.7076,
"step": 18944
},
{
"epoch": 0.01,
"learning_rate": 4.968143226742263e-05,
"loss": 4.7035,
"step": 19456
},
{
"epoch": 0.01,
"learning_rate": 4.967304631991211e-05,
"loss": 4.6897,
"step": 19968
},
{
"epoch": 0.01,
"learning_rate": 4.966466037240159e-05,
"loss": 4.6887,
"step": 20480
},
{
"epoch": 0.01,
"learning_rate": 4.965627442489107e-05,
"loss": 4.6667,
"step": 20992
},
{
"epoch": 0.01,
"learning_rate": 4.964788847738054e-05,
"loss": 4.6773,
"step": 21504
},
{
"epoch": 0.01,
"learning_rate": 4.963950252987002e-05,
"loss": 4.6477,
"step": 22016
},
{
"epoch": 0.01,
"learning_rate": 4.96311165823595e-05,
"loss": 4.6527,
"step": 22528
},
{
"epoch": 0.01,
"learning_rate": 4.962274701365272e-05,
"loss": 4.6424,
"step": 23040
},
{
"epoch": 0.01,
"learning_rate": 4.96143610661422e-05,
"loss": 4.6343,
"step": 23552
},
{
"epoch": 0.01,
"learning_rate": 4.960597511863168e-05,
"loss": 4.6304,
"step": 24064
},
{
"epoch": 0.01,
"learning_rate": 4.959758917112116e-05,
"loss": 4.6062,
"step": 24576
},
{
"epoch": 0.01,
"learning_rate": 4.9589219602414374e-05,
"loss": 4.5929,
"step": 25088
},
{
"epoch": 0.01,
"learning_rate": 4.958085003370758e-05,
"loss": 4.6059,
"step": 25600
},
{
"epoch": 0.01,
"learning_rate": 4.957246408619706e-05,
"loss": 4.5919,
"step": 26112
},
{
"epoch": 0.01,
"learning_rate": 4.956407813868654e-05,
"loss": 4.574,
"step": 26624
},
{
"epoch": 0.01,
"learning_rate": 4.9555692191176016e-05,
"loss": 4.5635,
"step": 27136
},
{
"epoch": 0.01,
"learning_rate": 4.9547306243665496e-05,
"loss": 4.5697,
"step": 27648
},
{
"epoch": 0.01,
"learning_rate": 4.9538920296154976e-05,
"loss": 4.5521,
"step": 28160
},
{
"epoch": 0.01,
"learning_rate": 4.9530534348644456e-05,
"loss": 4.5653,
"step": 28672
},
{
"epoch": 0.01,
"learning_rate": 4.952216477993767e-05,
"loss": 4.5385,
"step": 29184
},
{
"epoch": 0.01,
"learning_rate": 4.951377883242715e-05,
"loss": 4.5381,
"step": 29696
},
{
"epoch": 0.01,
"learning_rate": 4.950539288491663e-05,
"loss": 4.5403,
"step": 30208
},
{
"epoch": 0.01,
"learning_rate": 4.949700693740611e-05,
"loss": 4.5218,
"step": 30720
},
{
"epoch": 0.01,
"learning_rate": 4.948863736869932e-05,
"loss": 4.505,
"step": 31232
},
{
"epoch": 0.01,
"learning_rate": 4.94802514211888e-05,
"loss": 4.5098,
"step": 31744
},
{
"epoch": 0.01,
"learning_rate": 4.947186547367828e-05,
"loss": 4.4938,
"step": 32256
},
{
"epoch": 0.01,
"learning_rate": 4.946347952616776e-05,
"loss": 4.4941,
"step": 32768
},
{
"epoch": 0.01,
"learning_rate": 4.945509357865724e-05,
"loss": 4.5034,
"step": 33280
},
{
"epoch": 0.01,
"learning_rate": 4.944670763114672e-05,
"loss": 4.4972,
"step": 33792
},
{
"epoch": 0.01,
"learning_rate": 4.94383216836362e-05,
"loss": 4.4825,
"step": 34304
},
{
"epoch": 0.01,
"learning_rate": 4.942993573612568e-05,
"loss": 4.471,
"step": 34816
},
{
"epoch": 0.01,
"learning_rate": 4.942156616741889e-05,
"loss": 4.4648,
"step": 35328
},
{
"epoch": 0.01,
"learning_rate": 4.9413196598712105e-05,
"loss": 4.4748,
"step": 35840
},
{
"epoch": 0.01,
"learning_rate": 4.9404810651201585e-05,
"loss": 4.4702,
"step": 36352
},
{
"epoch": 0.01,
"learning_rate": 4.9396424703691065e-05,
"loss": 4.4627,
"step": 36864
},
{
"epoch": 0.01,
"learning_rate": 4.9388038756180545e-05,
"loss": 4.4621,
"step": 37376
},
{
"epoch": 0.01,
"learning_rate": 4.9379652808670025e-05,
"loss": 4.4649,
"step": 37888
},
{
"epoch": 0.01,
"learning_rate": 4.9371266861159505e-05,
"loss": 4.4566,
"step": 38400
},
{
"epoch": 0.01,
"learning_rate": 4.9362897292452714e-05,
"loss": 4.4309,
"step": 38912
},
{
"epoch": 0.01,
"learning_rate": 4.9354511344942194e-05,
"loss": 4.4426,
"step": 39424
},
{
"epoch": 0.01,
"learning_rate": 4.9346125397431674e-05,
"loss": 4.4366,
"step": 39936
},
{
"epoch": 0.01,
"learning_rate": 4.9337739449921154e-05,
"loss": 4.4238,
"step": 40448
},
{
"epoch": 0.01,
"learning_rate": 4.9329353502410634e-05,
"loss": 4.435,
"step": 40960
},
{
"epoch": 0.01,
"learning_rate": 4.9320967554900114e-05,
"loss": 4.424,
"step": 41472
},
{
"epoch": 0.01,
"learning_rate": 4.9312581607389594e-05,
"loss": 4.4328,
"step": 41984
},
{
"epoch": 0.01,
"learning_rate": 4.9304195659879074e-05,
"loss": 4.421,
"step": 42496
},
{
"epoch": 0.01,
"learning_rate": 4.929584246997602e-05,
"loss": 4.3936,
"step": 43008
},
{
"epoch": 0.01,
"learning_rate": 4.92874565224655e-05,
"loss": 4.405,
"step": 43520
},
{
"epoch": 0.01,
"learning_rate": 4.927907057495498e-05,
"loss": 4.4037,
"step": 44032
},
{
"epoch": 0.01,
"learning_rate": 4.927068462744446e-05,
"loss": 4.4055,
"step": 44544
},
{
"epoch": 0.01,
"learning_rate": 4.926229867993394e-05,
"loss": 4.3969,
"step": 45056
},
{
"epoch": 0.01,
"learning_rate": 4.925392911122715e-05,
"loss": 4.39,
"step": 45568
},
{
"epoch": 0.02,
"learning_rate": 4.924554316371663e-05,
"loss": 4.3905,
"step": 46080
},
{
"epoch": 0.02,
"learning_rate": 4.923715721620611e-05,
"loss": 4.3802,
"step": 46592
},
{
"epoch": 0.02,
"learning_rate": 4.922877126869559e-05,
"loss": 4.3719,
"step": 47104
},
{
"epoch": 0.02,
"learning_rate": 4.922038532118507e-05,
"loss": 4.3742,
"step": 47616
},
{
"epoch": 0.02,
"learning_rate": 4.921199937367455e-05,
"loss": 4.371,
"step": 48128
},
{
"epoch": 0.02,
"learning_rate": 4.920362980496776e-05,
"loss": 4.3809,
"step": 48640
},
{
"epoch": 0.02,
"learning_rate": 4.919524385745724e-05,
"loss": 4.3577,
"step": 49152
},
{
"epoch": 0.02,
"learning_rate": 4.918685790994672e-05,
"loss": 4.3551,
"step": 49664
},
{
"epoch": 0.02,
"learning_rate": 4.91784719624362e-05,
"loss": 4.3533,
"step": 50176
},
{
"epoch": 0.02,
"learning_rate": 4.9170086014925676e-05,
"loss": 4.3604,
"step": 50688
},
{
"epoch": 0.02,
"learning_rate": 4.9161700067415156e-05,
"loss": 4.3592,
"step": 51200
},
{
"epoch": 0.02,
"learning_rate": 4.915333049870837e-05,
"loss": 4.344,
"step": 51712
},
{
"epoch": 0.02,
"learning_rate": 4.914494455119785e-05,
"loss": 4.3449,
"step": 52224
},
{
"epoch": 0.02,
"learning_rate": 4.9136558603687325e-05,
"loss": 4.3266,
"step": 52736
},
{
"epoch": 0.02,
"learning_rate": 4.9128172656176805e-05,
"loss": 4.3403,
"step": 53248
},
{
"epoch": 0.02,
"learning_rate": 4.911980308747002e-05,
"loss": 4.3264,
"step": 53760
},
{
"epoch": 0.02,
"learning_rate": 4.91114171399595e-05,
"loss": 4.3239,
"step": 54272
},
{
"epoch": 0.02,
"learning_rate": 4.910304757125272e-05,
"loss": 4.3297,
"step": 54784
},
{
"epoch": 0.02,
"learning_rate": 4.90946616237422e-05,
"loss": 4.3276,
"step": 55296
},
{
"epoch": 0.02,
"learning_rate": 4.908627567623168e-05,
"loss": 4.3201,
"step": 55808
},
{
"epoch": 0.02,
"learning_rate": 4.907788972872115e-05,
"loss": 4.3192,
"step": 56320
},
{
"epoch": 0.02,
"learning_rate": 4.906950378121063e-05,
"loss": 4.31,
"step": 56832
},
{
"epoch": 0.02,
"learning_rate": 4.906111783370011e-05,
"loss": 4.3188,
"step": 57344
},
{
"epoch": 0.02,
"learning_rate": 4.905273188618959e-05,
"loss": 4.3059,
"step": 57856
},
{
"epoch": 0.02,
"learning_rate": 4.904434593867907e-05,
"loss": 4.3059,
"step": 58368
},
{
"epoch": 0.02,
"learning_rate": 4.903597636997228e-05,
"loss": 4.3119,
"step": 58880
},
{
"epoch": 0.02,
"learning_rate": 4.902759042246176e-05,
"loss": 4.3078,
"step": 59392
},
{
"epoch": 0.02,
"learning_rate": 4.9019220853754975e-05,
"loss": 4.3053,
"step": 59904
},
{
"epoch": 0.02,
"learning_rate": 4.9010834906244455e-05,
"loss": 4.297,
"step": 60416
},
{
"epoch": 0.02,
"learning_rate": 4.9002448958733935e-05,
"loss": 4.3023,
"step": 60928
},
{
"epoch": 0.02,
"learning_rate": 4.8994063011223415e-05,
"loss": 4.2978,
"step": 61440
},
{
"epoch": 0.02,
"learning_rate": 4.8985677063712895e-05,
"loss": 4.2825,
"step": 61952
},
{
"epoch": 0.02,
"learning_rate": 4.8977291116202375e-05,
"loss": 4.3009,
"step": 62464
},
{
"epoch": 0.02,
"learning_rate": 4.8968905168691855e-05,
"loss": 4.2869,
"step": 62976
},
{
"epoch": 0.02,
"learning_rate": 4.8960519221181335e-05,
"loss": 4.2934,
"step": 63488
},
{
"epoch": 0.02,
"learning_rate": 4.8952149652474544e-05,
"loss": 4.275,
"step": 64000
},
{
"epoch": 0.02,
"learning_rate": 4.8943763704964024e-05,
"loss": 4.2702,
"step": 64512
},
{
"epoch": 0.02,
"learning_rate": 4.8935377757453504e-05,
"loss": 4.2688,
"step": 65024
},
{
"epoch": 0.02,
"learning_rate": 4.8926991809942984e-05,
"loss": 4.2702,
"step": 65536
},
{
"epoch": 0.02,
"learning_rate": 4.891862224123619e-05,
"loss": 4.2636,
"step": 66048
},
{
"epoch": 0.02,
"learning_rate": 4.891023629372567e-05,
"loss": 4.2789,
"step": 66560
},
{
"epoch": 0.02,
"learning_rate": 4.890185034621515e-05,
"loss": 4.2755,
"step": 67072
},
{
"epoch": 0.02,
"learning_rate": 4.889346439870464e-05,
"loss": 4.2598,
"step": 67584
},
{
"epoch": 0.02,
"learning_rate": 4.888507845119412e-05,
"loss": 4.2641,
"step": 68096
},
{
"epoch": 0.02,
"learning_rate": 4.887670888248733e-05,
"loss": 4.2669,
"step": 68608
},
{
"epoch": 0.02,
"learning_rate": 4.886832293497681e-05,
"loss": 4.2629,
"step": 69120
},
{
"epoch": 0.02,
"learning_rate": 4.885993698746629e-05,
"loss": 4.2667,
"step": 69632
},
{
"epoch": 0.02,
"learning_rate": 4.885155103995577e-05,
"loss": 4.2648,
"step": 70144
},
{
"epoch": 0.02,
"learning_rate": 4.884318147124898e-05,
"loss": 4.2509,
"step": 70656
},
{
"epoch": 0.02,
"learning_rate": 4.883479552373846e-05,
"loss": 4.2567,
"step": 71168
},
{
"epoch": 0.02,
"learning_rate": 4.882640957622794e-05,
"loss": 4.252,
"step": 71680
},
{
"epoch": 0.02,
"learning_rate": 4.881802362871742e-05,
"loss": 4.2433,
"step": 72192
},
{
"epoch": 0.02,
"learning_rate": 4.8809654060010626e-05,
"loss": 4.2538,
"step": 72704
},
{
"epoch": 0.02,
"learning_rate": 4.8801268112500106e-05,
"loss": 4.2426,
"step": 73216
},
{
"epoch": 0.02,
"learning_rate": 4.8792882164989586e-05,
"loss": 4.2303,
"step": 73728
},
{
"epoch": 0.02,
"learning_rate": 4.878449621747907e-05,
"loss": 4.2548,
"step": 74240
},
{
"epoch": 0.02,
"learning_rate": 4.877612664877228e-05,
"loss": 4.2328,
"step": 74752
},
{
"epoch": 0.02,
"learning_rate": 4.876774070126176e-05,
"loss": 4.2315,
"step": 75264
},
{
"epoch": 0.02,
"learning_rate": 4.875935475375124e-05,
"loss": 4.2346,
"step": 75776
},
{
"epoch": 0.02,
"learning_rate": 4.875096880624072e-05,
"loss": 4.2284,
"step": 76288
},
{
"epoch": 0.03,
"eval_loss": 4.198453903198242,
"eval_runtime": 303.027,
"eval_samples_per_second": 1259.264,
"eval_steps_per_second": 39.353,
"step": 76320
},
{
"epoch": 1.0,
"learning_rate": 4.874259923753393e-05,
"loss": 4.2117,
"step": 76800
},
{
"epoch": 1.0,
"learning_rate": 4.873421329002341e-05,
"loss": 4.2218,
"step": 77312
},
{
"epoch": 1.0,
"learning_rate": 4.872582734251289e-05,
"loss": 4.2385,
"step": 77824
},
{
"epoch": 1.0,
"learning_rate": 4.871744139500237e-05,
"loss": 4.2172,
"step": 78336
},
{
"epoch": 1.0,
"learning_rate": 4.870907182629558e-05,
"loss": 4.2315,
"step": 78848
},
{
"epoch": 1.0,
"learning_rate": 4.870068587878506e-05,
"loss": 4.2041,
"step": 79360
},
{
"epoch": 1.0,
"learning_rate": 4.869229993127454e-05,
"loss": 4.2185,
"step": 79872
},
{
"epoch": 1.0,
"learning_rate": 4.8683913983764027e-05,
"loss": 4.2012,
"step": 80384
},
{
"epoch": 1.0,
"learning_rate": 4.8675560793860965e-05,
"loss": 4.2125,
"step": 80896
},
{
"epoch": 1.0,
"learning_rate": 4.8667174846350445e-05,
"loss": 4.2101,
"step": 81408
},
{
"epoch": 1.0,
"learning_rate": 4.8658788898839925e-05,
"loss": 4.2129,
"step": 81920
},
{
"epoch": 1.0,
"learning_rate": 4.8650402951329405e-05,
"loss": 4.2092,
"step": 82432
},
{
"epoch": 1.0,
"learning_rate": 4.8642017003818885e-05,
"loss": 4.1947,
"step": 82944
},
{
"epoch": 1.0,
"learning_rate": 4.8633631056308365e-05,
"loss": 4.1867,
"step": 83456
},
{
"epoch": 1.0,
"learning_rate": 4.8625245108797845e-05,
"loss": 4.1886,
"step": 83968
},
{
"epoch": 1.0,
"learning_rate": 4.8616859161287324e-05,
"loss": 4.1906,
"step": 84480
},
{
"epoch": 1.0,
"learning_rate": 4.8608489592580534e-05,
"loss": 4.2008,
"step": 84992
},
{
"epoch": 1.0,
"learning_rate": 4.8600103645070014e-05,
"loss": 4.1851,
"step": 85504
},
{
"epoch": 1.0,
"learning_rate": 4.8591717697559493e-05,
"loss": 4.1933,
"step": 86016
},
{
"epoch": 1.0,
"learning_rate": 4.858333175004898e-05,
"loss": 4.2079,
"step": 86528
},
{
"epoch": 1.0,
"learning_rate": 4.857496218134219e-05,
"loss": 4.1831,
"step": 87040
},
{
"epoch": 1.0,
"learning_rate": 4.856657623383167e-05,
"loss": 4.1781,
"step": 87552
},
{
"epoch": 1.0,
"learning_rate": 4.855819028632115e-05,
"loss": 4.1814,
"step": 88064
},
{
"epoch": 1.0,
"learning_rate": 4.854980433881063e-05,
"loss": 4.1889,
"step": 88576
},
{
"epoch": 1.0,
"learning_rate": 4.854143477010384e-05,
"loss": 4.1717,
"step": 89088
},
{
"epoch": 1.0,
"learning_rate": 4.853304882259332e-05,
"loss": 4.175,
"step": 89600
},
{
"epoch": 1.0,
"learning_rate": 4.85246628750828e-05,
"loss": 4.1748,
"step": 90112
},
{
"epoch": 1.0,
"learning_rate": 4.851627692757228e-05,
"loss": 4.1707,
"step": 90624
},
{
"epoch": 1.0,
"learning_rate": 4.850790735886549e-05,
"loss": 4.1557,
"step": 91136
},
{
"epoch": 1.01,
"learning_rate": 4.849952141135497e-05,
"loss": 4.161,
"step": 91648
},
{
"epoch": 1.01,
"learning_rate": 4.849113546384445e-05,
"loss": 4.1598,
"step": 92160
},
{
"epoch": 1.01,
"learning_rate": 4.8482749516333934e-05,
"loss": 4.17,
"step": 92672
},
{
"epoch": 1.01,
"learning_rate": 4.847437994762714e-05,
"loss": 4.1745,
"step": 93184
},
{
"epoch": 1.01,
"learning_rate": 4.846599400011662e-05,
"loss": 4.1584,
"step": 93696
},
{
"epoch": 1.01,
"learning_rate": 4.84576080526061e-05,
"loss": 4.154,
"step": 94208
},
{
"epoch": 1.01,
"learning_rate": 4.844922210509558e-05,
"loss": 4.1671,
"step": 94720
},
{
"epoch": 1.01,
"learning_rate": 4.844085253638879e-05,
"loss": 4.1422,
"step": 95232
},
{
"epoch": 1.01,
"learning_rate": 4.843246658887827e-05,
"loss": 4.1501,
"step": 95744
},
{
"epoch": 1.01,
"learning_rate": 4.842408064136775e-05,
"loss": 4.1418,
"step": 96256
},
{
"epoch": 1.01,
"learning_rate": 4.8415694693857225e-05,
"loss": 4.148,
"step": 96768
},
{
"epoch": 1.01,
"learning_rate": 4.840732512515044e-05,
"loss": 4.1397,
"step": 97280
},
{
"epoch": 1.01,
"learning_rate": 4.839893917763992e-05,
"loss": 4.1563,
"step": 97792
},
{
"epoch": 1.01,
"learning_rate": 4.83905532301294e-05,
"loss": 4.143,
"step": 98304
},
{
"epoch": 1.01,
"learning_rate": 4.838216728261888e-05,
"loss": 4.1412,
"step": 98816
},
{
"epoch": 1.01,
"learning_rate": 4.83737977139121e-05,
"loss": 4.1408,
"step": 99328
},
{
"epoch": 1.01,
"learning_rate": 4.836541176640158e-05,
"loss": 4.1442,
"step": 99840
},
{
"epoch": 1.01,
"learning_rate": 4.8357025818891057e-05,
"loss": 4.1371,
"step": 100352
},
{
"epoch": 1.01,
"learning_rate": 4.834863987138053e-05,
"loss": 4.1277,
"step": 100864
},
{
"epoch": 1.01,
"learning_rate": 4.8340270302673746e-05,
"loss": 4.1198,
"step": 101376
},
{
"epoch": 1.01,
"learning_rate": 4.8331884355163226e-05,
"loss": 4.1429,
"step": 101888
},
{
"epoch": 1.01,
"learning_rate": 4.83234984076527e-05,
"loss": 4.1316,
"step": 102400
},
{
"epoch": 1.01,
"learning_rate": 4.831511246014218e-05,
"loss": 4.1192,
"step": 102912
},
{
"epoch": 1.01,
"learning_rate": 4.8306742891435395e-05,
"loss": 4.1205,
"step": 103424
},
{
"epoch": 1.01,
"learning_rate": 4.8298356943924875e-05,
"loss": 4.1282,
"step": 103936
},
{
"epoch": 1.01,
"learning_rate": 4.8289970996414354e-05,
"loss": 4.1148,
"step": 104448
},
{
"epoch": 1.01,
"learning_rate": 4.8281585048903834e-05,
"loss": 4.1332,
"step": 104960
},
{
"epoch": 1.01,
"learning_rate": 4.827321548019705e-05,
"loss": 4.1158,
"step": 105472
},
{
"epoch": 1.01,
"learning_rate": 4.826482953268653e-05,
"loss": 4.1182,
"step": 105984
},
{
"epoch": 1.01,
"learning_rate": 4.8256443585176003e-05,
"loss": 4.1291,
"step": 106496
},
{
"epoch": 1.01,
"learning_rate": 4.8248057637665483e-05,
"loss": 4.1052,
"step": 107008
},
{
"epoch": 1.01,
"learning_rate": 4.82396880689587e-05,
"loss": 4.1042,
"step": 107520
},
{
"epoch": 1.01,
"learning_rate": 4.823130212144817e-05,
"loss": 4.1127,
"step": 108032
},
{
"epoch": 1.01,
"learning_rate": 4.822291617393765e-05,
"loss": 4.0963,
"step": 108544
},
{
"epoch": 1.01,
"learning_rate": 4.821453022642713e-05,
"loss": 4.1005,
"step": 109056
},
{
"epoch": 1.01,
"learning_rate": 4.820616065772035e-05,
"loss": 4.1137,
"step": 109568
},
{
"epoch": 1.01,
"learning_rate": 4.819777471020983e-05,
"loss": 4.1182,
"step": 110080
},
{
"epoch": 1.01,
"learning_rate": 4.818938876269931e-05,
"loss": 4.1043,
"step": 110592
},
{
"epoch": 1.01,
"learning_rate": 4.818100281518879e-05,
"loss": 4.0986,
"step": 111104
},
{
"epoch": 1.01,
"learning_rate": 4.817264962528573e-05,
"loss": 4.0891,
"step": 111616
},
{
"epoch": 1.01,
"learning_rate": 4.816426367777521e-05,
"loss": 4.1087,
"step": 112128
},
{
"epoch": 1.01,
"learning_rate": 4.815587773026469e-05,
"loss": 4.1055,
"step": 112640
},
{
"epoch": 1.01,
"learning_rate": 4.814749178275417e-05,
"loss": 4.1048,
"step": 113152
},
{
"epoch": 1.01,
"learning_rate": 4.8139105835243646e-05,
"loss": 4.1056,
"step": 113664
},
{
"epoch": 1.01,
"learning_rate": 4.8130719887733126e-05,
"loss": 4.1076,
"step": 114176
},
{
"epoch": 1.01,
"learning_rate": 4.812235031902634e-05,
"loss": 4.1069,
"step": 114688
},
{
"epoch": 1.01,
"learning_rate": 4.811396437151582e-05,
"loss": 4.0849,
"step": 115200
},
{
"epoch": 1.01,
"learning_rate": 4.81055784240053e-05,
"loss": 4.0975,
"step": 115712
},
{
"epoch": 1.01,
"learning_rate": 4.809719247649478e-05,
"loss": 4.0992,
"step": 116224
},
{
"epoch": 1.01,
"learning_rate": 4.808880652898426e-05,
"loss": 4.0835,
"step": 116736
},
{
"epoch": 1.01,
"learning_rate": 4.808042058147374e-05,
"loss": 4.1053,
"step": 117248
},
{
"epoch": 1.01,
"learning_rate": 4.807205101276695e-05,
"loss": 4.0973,
"step": 117760
},
{
"epoch": 1.01,
"learning_rate": 4.806366506525643e-05,
"loss": 4.1022,
"step": 118272
},
{
"epoch": 1.01,
"learning_rate": 4.805527911774591e-05,
"loss": 4.0972,
"step": 118784
},
{
"epoch": 1.01,
"learning_rate": 4.804689317023539e-05,
"loss": 4.0753,
"step": 119296
},
{
"epoch": 1.01,
"learning_rate": 4.803850722272487e-05,
"loss": 4.0879,
"step": 119808
},
{
"epoch": 1.01,
"learning_rate": 4.803012127521435e-05,
"loss": 4.087,
"step": 120320
},
{
"epoch": 1.01,
"learning_rate": 4.802173532770383e-05,
"loss": 4.091,
"step": 120832
},
{
"epoch": 1.01,
"learning_rate": 4.801334938019331e-05,
"loss": 4.0865,
"step": 121344
},
{
"epoch": 1.01,
"learning_rate": 4.800497981148652e-05,
"loss": 4.0781,
"step": 121856
},
{
"epoch": 1.02,
"learning_rate": 4.7996593863976e-05,
"loss": 4.0868,
"step": 122368
},
{
"epoch": 1.02,
"learning_rate": 4.7988207916465486e-05,
"loss": 4.0824,
"step": 122880
},
{
"epoch": 1.02,
"learning_rate": 4.7979821968954966e-05,
"loss": 4.0668,
"step": 123392
},
{
"epoch": 1.02,
"learning_rate": 4.7971436021444446e-05,
"loss": 4.0761,
"step": 123904
},
{
"epoch": 1.02,
"learning_rate": 4.7963050073933926e-05,
"loss": 4.0812,
"step": 124416
},
{
"epoch": 1.02,
"learning_rate": 4.7954680505227135e-05,
"loss": 4.0876,
"step": 124928
},
{
"epoch": 1.02,
"learning_rate": 4.7946294557716615e-05,
"loss": 4.0705,
"step": 125440
},
{
"epoch": 1.02,
"learning_rate": 4.7937908610206095e-05,
"loss": 4.0605,
"step": 125952
},
{
"epoch": 1.02,
"learning_rate": 4.7929522662695575e-05,
"loss": 4.0688,
"step": 126464
},
{
"epoch": 1.02,
"learning_rate": 4.7921153093988784e-05,
"loss": 4.0757,
"step": 126976
},
{
"epoch": 1.02,
"learning_rate": 4.7912767146478264e-05,
"loss": 4.0754,
"step": 127488
},
{
"epoch": 1.02,
"learning_rate": 4.7904381198967744e-05,
"loss": 4.0679,
"step": 128000
},
{
"epoch": 1.02,
"learning_rate": 4.7895995251457224e-05,
"loss": 4.0637,
"step": 128512
},
{
"epoch": 1.02,
"learning_rate": 4.7887609303946704e-05,
"loss": 4.0532,
"step": 129024
},
{
"epoch": 1.02,
"learning_rate": 4.7879223356436184e-05,
"loss": 4.067,
"step": 129536
},
{
"epoch": 1.02,
"learning_rate": 4.78708537877294e-05,
"loss": 4.0539,
"step": 130048
},
{
"epoch": 1.02,
"learning_rate": 4.786246784021888e-05,
"loss": 4.0584,
"step": 130560
},
{
"epoch": 1.02,
"learning_rate": 4.785408189270836e-05,
"loss": 4.063,
"step": 131072
},
{
"epoch": 1.02,
"learning_rate": 4.784569594519783e-05,
"loss": 4.0623,
"step": 131584
},
{
"epoch": 1.02,
"learning_rate": 4.783730999768731e-05,
"loss": 4.0563,
"step": 132096
},
{
"epoch": 1.02,
"learning_rate": 4.782892405017679e-05,
"loss": 4.0602,
"step": 132608
},
{
"epoch": 1.02,
"learning_rate": 4.782053810266627e-05,
"loss": 4.0509,
"step": 133120
},
{
"epoch": 1.02,
"learning_rate": 4.781216853395948e-05,
"loss": 4.057,
"step": 133632
},
{
"epoch": 1.02,
"learning_rate": 4.780378258644896e-05,
"loss": 4.0504,
"step": 134144
},
{
"epoch": 1.02,
"learning_rate": 4.779539663893844e-05,
"loss": 4.0485,
"step": 134656
},
{
"epoch": 1.02,
"learning_rate": 4.778701069142792e-05,
"loss": 4.0614,
"step": 135168
},
{
"epoch": 1.02,
"learning_rate": 4.77786247439174e-05,
"loss": 4.0609,
"step": 135680
},
{
"epoch": 1.02,
"learning_rate": 4.777025517521062e-05,
"loss": 4.0558,
"step": 136192
},
{
"epoch": 1.02,
"learning_rate": 4.77618692277001e-05,
"loss": 4.0505,
"step": 136704
},
{
"epoch": 1.02,
"learning_rate": 4.775348328018958e-05,
"loss": 4.0545,
"step": 137216
},
{
"epoch": 1.02,
"learning_rate": 4.774509733267906e-05,
"loss": 4.0531,
"step": 137728
},
{
"epoch": 1.02,
"learning_rate": 4.773671138516854e-05,
"loss": 4.0391,
"step": 138240
},
{
"epoch": 1.02,
"learning_rate": 4.7728341816461747e-05,
"loss": 4.0586,
"step": 138752
},
{
"epoch": 1.02,
"learning_rate": 4.7719955868951227e-05,
"loss": 4.0459,
"step": 139264
},
{
"epoch": 1.02,
"learning_rate": 4.7711569921440707e-05,
"loss": 4.0534,
"step": 139776
},
{
"epoch": 1.02,
"learning_rate": 4.7703183973930186e-05,
"loss": 4.0375,
"step": 140288
},
{
"epoch": 1.02,
"learning_rate": 4.7694814405223396e-05,
"loss": 4.0379,
"step": 140800
},
{
"epoch": 1.02,
"learning_rate": 4.7686428457712876e-05,
"loss": 4.0351,
"step": 141312
},
{
"epoch": 1.02,
"learning_rate": 4.7678042510202355e-05,
"loss": 4.0353,
"step": 141824
},
{
"epoch": 1.02,
"learning_rate": 4.766965656269184e-05,
"loss": 4.0345,
"step": 142336
},
{
"epoch": 1.02,
"learning_rate": 4.766128699398505e-05,
"loss": 4.0466,
"step": 142848
},
{
"epoch": 1.02,
"learning_rate": 4.765290104647453e-05,
"loss": 4.0516,
"step": 143360
},
{
"epoch": 1.02,
"learning_rate": 4.764451509896401e-05,
"loss": 4.0379,
"step": 143872
},
{
"epoch": 1.02,
"learning_rate": 4.763612915145349e-05,
"loss": 4.0318,
"step": 144384
},
{
"epoch": 1.02,
"learning_rate": 4.762774320394297e-05,
"loss": 4.0423,
"step": 144896
},
{
"epoch": 1.02,
"learning_rate": 4.761937363523618e-05,
"loss": 4.0417,
"step": 145408
},
{
"epoch": 1.02,
"learning_rate": 4.761098768772566e-05,
"loss": 4.0467,
"step": 145920
},
{
"epoch": 1.02,
"learning_rate": 4.760260174021514e-05,
"loss": 4.0402,
"step": 146432
},
{
"epoch": 1.02,
"learning_rate": 4.759421579270462e-05,
"loss": 4.0369,
"step": 146944
},
{
"epoch": 1.02,
"learning_rate": 4.75858298451941e-05,
"loss": 4.0288,
"step": 147456
},
{
"epoch": 1.02,
"learning_rate": 4.757744389768358e-05,
"loss": 4.0403,
"step": 147968
},
{
"epoch": 1.02,
"learning_rate": 4.7569074328976796e-05,
"loss": 4.0298,
"step": 148480
},
{
"epoch": 1.02,
"learning_rate": 4.7560688381466276e-05,
"loss": 4.0334,
"step": 148992
},
{
"epoch": 1.02,
"learning_rate": 4.7552302433955756e-05,
"loss": 4.0316,
"step": 149504
},
{
"epoch": 1.02,
"learning_rate": 4.7543916486445236e-05,
"loss": 4.0212,
"step": 150016
},
{
"epoch": 1.02,
"learning_rate": 4.7535546917738445e-05,
"loss": 4.0453,
"step": 150528
},
{
"epoch": 1.02,
"learning_rate": 4.7527160970227925e-05,
"loss": 4.0253,
"step": 151040
},
{
"epoch": 1.02,
"learning_rate": 4.7518775022717405e-05,
"loss": 4.0204,
"step": 151552
},
{
"epoch": 1.02,
"learning_rate": 4.7510389075206885e-05,
"loss": 4.0248,
"step": 152064
},
{
"epoch": 1.02,
"learning_rate": 4.7502019506500094e-05,
"loss": 4.0216,
"step": 152576
},
{
"epoch": 1.03,
"eval_loss": 4.029199600219727,
"eval_runtime": 302.4487,
"eval_samples_per_second": 1261.672,
"eval_steps_per_second": 39.428,
"step": 152640
}
],
"logging_steps": 512,
"max_steps": 3052726,
"num_train_epochs": 9223372036854775807,
"save_steps": 10,
"total_flos": 1.0579461787083571e+17,
"trial_name": null,
"trial_params": null
}