abhinavp's picture
Training in progress, epoch 0, checkpoint
ed5b737 verified
raw
history blame
56.1 kB
{
"best_metric": 3.9674437046051025,
"best_model_checkpoint": "/mmfs1/gscratch/stf/abhinavp/corpus-filtering/outputs/superlative-quantifier/transformer/1/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": 11.1003,
"step": 1
},
{
"epoch": 0.0,
"learning_rate": 4.999161405248948e-05,
"loss": 6.8296,
"step": 512
},
{
"epoch": 0.0,
"learning_rate": 4.998322810497896e-05,
"loss": 6.1836,
"step": 1024
},
{
"epoch": 0.0,
"learning_rate": 4.997484215746844e-05,
"loss": 5.9784,
"step": 1536
},
{
"epoch": 0.0,
"learning_rate": 4.996645620995792e-05,
"loss": 5.8165,
"step": 2048
},
{
"epoch": 0.0,
"learning_rate": 4.99580702624474e-05,
"loss": 5.7061,
"step": 2560
},
{
"epoch": 0.0,
"learning_rate": 4.994968431493688e-05,
"loss": 5.6062,
"step": 3072
},
{
"epoch": 0.0,
"learning_rate": 4.994129836742636e-05,
"loss": 5.5289,
"step": 3584
},
{
"epoch": 0.0,
"learning_rate": 4.993291241991584e-05,
"loss": 5.4503,
"step": 4096
},
{
"epoch": 0.0,
"learning_rate": 4.992452647240532e-05,
"loss": 5.4104,
"step": 4608
},
{
"epoch": 0.0,
"learning_rate": 4.99161405248948e-05,
"loss": 5.3559,
"step": 5120
},
{
"epoch": 0.0,
"learning_rate": 4.990775457738428e-05,
"loss": 5.303,
"step": 5632
},
{
"epoch": 0.0,
"learning_rate": 4.989936862987376e-05,
"loss": 5.2642,
"step": 6144
},
{
"epoch": 0.0,
"learning_rate": 4.989099906116697e-05,
"loss": 5.2163,
"step": 6656
},
{
"epoch": 0.0,
"learning_rate": 4.988261311365645e-05,
"loss": 5.1754,
"step": 7168
},
{
"epoch": 0.0,
"learning_rate": 4.987422716614593e-05,
"loss": 5.1456,
"step": 7680
},
{
"epoch": 0.0,
"learning_rate": 4.986584121863541e-05,
"loss": 5.106,
"step": 8192
},
{
"epoch": 0.0,
"learning_rate": 4.985745527112489e-05,
"loss": 5.0844,
"step": 8704
},
{
"epoch": 0.0,
"learning_rate": 4.984906932361437e-05,
"loss": 5.0474,
"step": 9216
},
{
"epoch": 0.0,
"learning_rate": 4.984068337610385e-05,
"loss": 5.0323,
"step": 9728
},
{
"epoch": 0.0,
"learning_rate": 4.983229742859333e-05,
"loss": 5.013,
"step": 10240
},
{
"epoch": 0.0,
"learning_rate": 4.982391148108281e-05,
"loss": 4.9782,
"step": 10752
},
{
"epoch": 0.0,
"learning_rate": 4.981552553357229e-05,
"loss": 4.9525,
"step": 11264
},
{
"epoch": 0.0,
"learning_rate": 4.980713958606178e-05,
"loss": 4.9316,
"step": 11776
},
{
"epoch": 0.0,
"learning_rate": 4.9798770017354986e-05,
"loss": 4.9197,
"step": 12288
},
{
"epoch": 0.0,
"learning_rate": 4.9790384069844466e-05,
"loss": 4.8934,
"step": 12800
},
{
"epoch": 0.0,
"learning_rate": 4.9781998122333946e-05,
"loss": 4.8686,
"step": 13312
},
{
"epoch": 0.0,
"learning_rate": 4.9773612174823426e-05,
"loss": 4.8494,
"step": 13824
},
{
"epoch": 0.0,
"learning_rate": 4.9765242606116635e-05,
"loss": 4.8529,
"step": 14336
},
{
"epoch": 0.0,
"learning_rate": 4.9756856658606115e-05,
"loss": 4.8158,
"step": 14848
},
{
"epoch": 0.01,
"learning_rate": 4.9748470711095595e-05,
"loss": 4.8077,
"step": 15360
},
{
"epoch": 0.01,
"learning_rate": 4.974008476358507e-05,
"loss": 4.8032,
"step": 15872
},
{
"epoch": 0.01,
"learning_rate": 4.973169881607455e-05,
"loss": 4.7852,
"step": 16384
},
{
"epoch": 0.01,
"learning_rate": 4.9723329247367764e-05,
"loss": 4.7698,
"step": 16896
},
{
"epoch": 0.01,
"learning_rate": 4.9714943299857244e-05,
"loss": 4.7685,
"step": 17408
},
{
"epoch": 0.01,
"learning_rate": 4.9706557352346724e-05,
"loss": 4.7497,
"step": 17920
},
{
"epoch": 0.01,
"learning_rate": 4.9698171404836204e-05,
"loss": 4.7428,
"step": 18432
},
{
"epoch": 0.01,
"learning_rate": 4.968980183612942e-05,
"loss": 4.709,
"step": 18944
},
{
"epoch": 0.01,
"learning_rate": 4.96814158886189e-05,
"loss": 4.7086,
"step": 19456
},
{
"epoch": 0.01,
"learning_rate": 4.967302994110837e-05,
"loss": 4.688,
"step": 19968
},
{
"epoch": 0.01,
"learning_rate": 4.966464399359785e-05,
"loss": 4.6821,
"step": 20480
},
{
"epoch": 0.01,
"learning_rate": 4.965627442489107e-05,
"loss": 4.6766,
"step": 20992
},
{
"epoch": 0.01,
"learning_rate": 4.964788847738054e-05,
"loss": 4.6693,
"step": 21504
},
{
"epoch": 0.01,
"learning_rate": 4.963950252987002e-05,
"loss": 4.6512,
"step": 22016
},
{
"epoch": 0.01,
"learning_rate": 4.96311165823595e-05,
"loss": 4.6535,
"step": 22528
},
{
"epoch": 0.01,
"learning_rate": 4.962274701365272e-05,
"loss": 4.6372,
"step": 23040
},
{
"epoch": 0.01,
"learning_rate": 4.96143610661422e-05,
"loss": 4.6374,
"step": 23552
},
{
"epoch": 0.01,
"learning_rate": 4.960597511863168e-05,
"loss": 4.6382,
"step": 24064
},
{
"epoch": 0.01,
"learning_rate": 4.959758917112116e-05,
"loss": 4.6073,
"step": 24576
},
{
"epoch": 0.01,
"learning_rate": 4.958920322361064e-05,
"loss": 4.5984,
"step": 25088
},
{
"epoch": 0.01,
"learning_rate": 4.958083365490385e-05,
"loss": 4.6065,
"step": 25600
},
{
"epoch": 0.01,
"learning_rate": 4.957244770739333e-05,
"loss": 4.6053,
"step": 26112
},
{
"epoch": 0.01,
"learning_rate": 4.956406175988281e-05,
"loss": 4.579,
"step": 26624
},
{
"epoch": 0.01,
"learning_rate": 4.955567581237229e-05,
"loss": 4.573,
"step": 27136
},
{
"epoch": 0.01,
"learning_rate": 4.9547306243665496e-05,
"loss": 4.5726,
"step": 27648
},
{
"epoch": 0.01,
"learning_rate": 4.9538920296154976e-05,
"loss": 4.5515,
"step": 28160
},
{
"epoch": 0.01,
"learning_rate": 4.9530534348644456e-05,
"loss": 4.5627,
"step": 28672
},
{
"epoch": 0.01,
"learning_rate": 4.9522148401133936e-05,
"loss": 4.5348,
"step": 29184
},
{
"epoch": 0.01,
"learning_rate": 4.951377883242715e-05,
"loss": 4.544,
"step": 29696
},
{
"epoch": 0.01,
"learning_rate": 4.950539288491663e-05,
"loss": 4.5368,
"step": 30208
},
{
"epoch": 0.01,
"learning_rate": 4.949700693740611e-05,
"loss": 4.5342,
"step": 30720
},
{
"epoch": 0.01,
"learning_rate": 4.948862098989559e-05,
"loss": 4.5083,
"step": 31232
},
{
"epoch": 0.01,
"learning_rate": 4.94802514211888e-05,
"loss": 4.5253,
"step": 31744
},
{
"epoch": 0.01,
"learning_rate": 4.947186547367828e-05,
"loss": 4.4963,
"step": 32256
},
{
"epoch": 0.01,
"learning_rate": 4.946347952616776e-05,
"loss": 4.5065,
"step": 32768
},
{
"epoch": 0.01,
"learning_rate": 4.945509357865724e-05,
"loss": 4.5056,
"step": 33280
},
{
"epoch": 0.01,
"learning_rate": 4.944672400995045e-05,
"loss": 4.4981,
"step": 33792
},
{
"epoch": 0.01,
"learning_rate": 4.943833806243993e-05,
"loss": 4.4787,
"step": 34304
},
{
"epoch": 0.01,
"learning_rate": 4.942995211492941e-05,
"loss": 4.4711,
"step": 34816
},
{
"epoch": 0.01,
"learning_rate": 4.9421582546222625e-05,
"loss": 4.4663,
"step": 35328
},
{
"epoch": 0.01,
"learning_rate": 4.9413196598712105e-05,
"loss": 4.4795,
"step": 35840
},
{
"epoch": 0.01,
"learning_rate": 4.9404810651201585e-05,
"loss": 4.4731,
"step": 36352
},
{
"epoch": 0.01,
"learning_rate": 4.9396424703691065e-05,
"loss": 4.4557,
"step": 36864
},
{
"epoch": 0.01,
"learning_rate": 4.9388038756180545e-05,
"loss": 4.4672,
"step": 37376
},
{
"epoch": 0.01,
"learning_rate": 4.9379652808670025e-05,
"loss": 4.4715,
"step": 37888
},
{
"epoch": 0.01,
"learning_rate": 4.9371266861159505e-05,
"loss": 4.4527,
"step": 38400
},
{
"epoch": 0.01,
"learning_rate": 4.9362897292452714e-05,
"loss": 4.4494,
"step": 38912
},
{
"epoch": 0.01,
"learning_rate": 4.9354511344942194e-05,
"loss": 4.4393,
"step": 39424
},
{
"epoch": 0.01,
"learning_rate": 4.9346125397431674e-05,
"loss": 4.4402,
"step": 39936
},
{
"epoch": 0.01,
"learning_rate": 4.9337739449921154e-05,
"loss": 4.4214,
"step": 40448
},
{
"epoch": 0.01,
"learning_rate": 4.9329353502410634e-05,
"loss": 4.4257,
"step": 40960
},
{
"epoch": 0.01,
"learning_rate": 4.9320967554900114e-05,
"loss": 4.4212,
"step": 41472
},
{
"epoch": 0.01,
"learning_rate": 4.9312581607389594e-05,
"loss": 4.4317,
"step": 41984
},
{
"epoch": 0.01,
"learning_rate": 4.9304195659879074e-05,
"loss": 4.4201,
"step": 42496
},
{
"epoch": 0.01,
"learning_rate": 4.929582609117229e-05,
"loss": 4.3996,
"step": 43008
},
{
"epoch": 0.01,
"learning_rate": 4.928744014366177e-05,
"loss": 4.4058,
"step": 43520
},
{
"epoch": 0.01,
"learning_rate": 4.927905419615125e-05,
"loss": 4.4032,
"step": 44032
},
{
"epoch": 0.01,
"learning_rate": 4.927066824864073e-05,
"loss": 4.4062,
"step": 44544
},
{
"epoch": 0.01,
"learning_rate": 4.926229867993394e-05,
"loss": 4.4015,
"step": 45056
},
{
"epoch": 0.01,
"learning_rate": 4.925392911122715e-05,
"loss": 4.3903,
"step": 45568
},
{
"epoch": 0.02,
"learning_rate": 4.924554316371663e-05,
"loss": 4.3891,
"step": 46080
},
{
"epoch": 0.02,
"learning_rate": 4.923715721620611e-05,
"loss": 4.3796,
"step": 46592
},
{
"epoch": 0.02,
"learning_rate": 4.922877126869559e-05,
"loss": 4.3708,
"step": 47104
},
{
"epoch": 0.02,
"learning_rate": 4.922038532118507e-05,
"loss": 4.3634,
"step": 47616
},
{
"epoch": 0.02,
"learning_rate": 4.921201575247828e-05,
"loss": 4.3685,
"step": 48128
},
{
"epoch": 0.02,
"learning_rate": 4.920362980496776e-05,
"loss": 4.3767,
"step": 48640
},
{
"epoch": 0.02,
"learning_rate": 4.919524385745724e-05,
"loss": 4.3679,
"step": 49152
},
{
"epoch": 0.02,
"learning_rate": 4.918685790994672e-05,
"loss": 4.3539,
"step": 49664
},
{
"epoch": 0.02,
"learning_rate": 4.91784719624362e-05,
"loss": 4.3549,
"step": 50176
},
{
"epoch": 0.02,
"learning_rate": 4.9170086014925676e-05,
"loss": 4.3587,
"step": 50688
},
{
"epoch": 0.02,
"learning_rate": 4.9161700067415156e-05,
"loss": 4.3593,
"step": 51200
},
{
"epoch": 0.02,
"learning_rate": 4.9153314119904636e-05,
"loss": 4.3512,
"step": 51712
},
{
"epoch": 0.02,
"learning_rate": 4.914494455119785e-05,
"loss": 4.3462,
"step": 52224
},
{
"epoch": 0.02,
"learning_rate": 4.9136558603687325e-05,
"loss": 4.3392,
"step": 52736
},
{
"epoch": 0.02,
"learning_rate": 4.9128172656176805e-05,
"loss": 4.3465,
"step": 53248
},
{
"epoch": 0.02,
"learning_rate": 4.9119786708666285e-05,
"loss": 4.3256,
"step": 53760
},
{
"epoch": 0.02,
"learning_rate": 4.91114171399595e-05,
"loss": 4.3305,
"step": 54272
},
{
"epoch": 0.02,
"learning_rate": 4.910303119244898e-05,
"loss": 4.3347,
"step": 54784
},
{
"epoch": 0.02,
"learning_rate": 4.909464524493846e-05,
"loss": 4.3233,
"step": 55296
},
{
"epoch": 0.02,
"learning_rate": 4.908625929742794e-05,
"loss": 4.3217,
"step": 55808
},
{
"epoch": 0.02,
"learning_rate": 4.907787334991742e-05,
"loss": 4.3218,
"step": 56320
},
{
"epoch": 0.02,
"learning_rate": 4.90694874024069e-05,
"loss": 4.3144,
"step": 56832
},
{
"epoch": 0.02,
"learning_rate": 4.906110145489638e-05,
"loss": 4.3178,
"step": 57344
},
{
"epoch": 0.02,
"learning_rate": 4.905273188618959e-05,
"loss": 4.3152,
"step": 57856
},
{
"epoch": 0.02,
"learning_rate": 4.904434593867907e-05,
"loss": 4.3092,
"step": 58368
},
{
"epoch": 0.02,
"learning_rate": 4.903595999116855e-05,
"loss": 4.3145,
"step": 58880
},
{
"epoch": 0.02,
"learning_rate": 4.902757404365803e-05,
"loss": 4.3129,
"step": 59392
},
{
"epoch": 0.02,
"learning_rate": 4.9019220853754975e-05,
"loss": 4.3061,
"step": 59904
},
{
"epoch": 0.02,
"learning_rate": 4.9010834906244455e-05,
"loss": 4.3092,
"step": 60416
},
{
"epoch": 0.02,
"learning_rate": 4.9002448958733935e-05,
"loss": 4.2981,
"step": 60928
},
{
"epoch": 0.02,
"learning_rate": 4.8994063011223415e-05,
"loss": 4.2975,
"step": 61440
},
{
"epoch": 0.02,
"learning_rate": 4.8985677063712895e-05,
"loss": 4.2895,
"step": 61952
},
{
"epoch": 0.02,
"learning_rate": 4.8977291116202375e-05,
"loss": 4.2871,
"step": 62464
},
{
"epoch": 0.02,
"learning_rate": 4.8968905168691855e-05,
"loss": 4.2876,
"step": 62976
},
{
"epoch": 0.02,
"learning_rate": 4.8960519221181335e-05,
"loss": 4.2874,
"step": 63488
},
{
"epoch": 0.02,
"learning_rate": 4.8952133273670815e-05,
"loss": 4.2847,
"step": 64000
},
{
"epoch": 0.02,
"learning_rate": 4.8943747326160294e-05,
"loss": 4.285,
"step": 64512
},
{
"epoch": 0.02,
"learning_rate": 4.8935361378649774e-05,
"loss": 4.2828,
"step": 65024
},
{
"epoch": 0.02,
"learning_rate": 4.8926991809942984e-05,
"loss": 4.2691,
"step": 65536
},
{
"epoch": 0.02,
"learning_rate": 4.8918605862432463e-05,
"loss": 4.2813,
"step": 66048
},
{
"epoch": 0.02,
"learning_rate": 4.8910219914921943e-05,
"loss": 4.2801,
"step": 66560
},
{
"epoch": 0.02,
"learning_rate": 4.8901833967411423e-05,
"loss": 4.2609,
"step": 67072
},
{
"epoch": 0.02,
"learning_rate": 4.88934480199009e-05,
"loss": 4.2524,
"step": 67584
},
{
"epoch": 0.02,
"learning_rate": 4.888506207239038e-05,
"loss": 4.2603,
"step": 68096
},
{
"epoch": 0.02,
"learning_rate": 4.887667612487986e-05,
"loss": 4.2736,
"step": 68608
},
{
"epoch": 0.02,
"learning_rate": 4.886830655617308e-05,
"loss": 4.274,
"step": 69120
},
{
"epoch": 0.02,
"learning_rate": 4.885993698746629e-05,
"loss": 4.2599,
"step": 69632
},
{
"epoch": 0.02,
"learning_rate": 4.885155103995577e-05,
"loss": 4.2577,
"step": 70144
},
{
"epoch": 0.02,
"learning_rate": 4.884316509244525e-05,
"loss": 4.2631,
"step": 70656
},
{
"epoch": 0.02,
"learning_rate": 4.883477914493473e-05,
"loss": 4.2638,
"step": 71168
},
{
"epoch": 0.02,
"learning_rate": 4.882639319742421e-05,
"loss": 4.2547,
"step": 71680
},
{
"epoch": 0.02,
"learning_rate": 4.881800724991369e-05,
"loss": 4.2453,
"step": 72192
},
{
"epoch": 0.02,
"learning_rate": 4.880962130240316e-05,
"loss": 4.2491,
"step": 72704
},
{
"epoch": 0.02,
"learning_rate": 4.880123535489264e-05,
"loss": 4.2436,
"step": 73216
},
{
"epoch": 0.02,
"learning_rate": 4.8792882164989586e-05,
"loss": 4.2411,
"step": 73728
},
{
"epoch": 0.02,
"learning_rate": 4.878449621747907e-05,
"loss": 4.2501,
"step": 74240
},
{
"epoch": 0.02,
"learning_rate": 4.877611026996855e-05,
"loss": 4.2359,
"step": 74752
},
{
"epoch": 0.02,
"learning_rate": 4.876772432245803e-05,
"loss": 4.247,
"step": 75264
},
{
"epoch": 0.02,
"learning_rate": 4.875933837494751e-05,
"loss": 4.2419,
"step": 75776
},
{
"epoch": 0.02,
"learning_rate": 4.8750952427436986e-05,
"loss": 4.2257,
"step": 76288
},
{
"epoch": 0.03,
"eval_loss": 4.211981773376465,
"eval_runtime": 315.8607,
"eval_samples_per_second": 1208.099,
"eval_steps_per_second": 37.754,
"step": 76320
},
{
"epoch": 1.0,
"learning_rate": 4.8742566479926466e-05,
"loss": 4.2306,
"step": 76800
},
{
"epoch": 1.0,
"learning_rate": 4.8734180532415946e-05,
"loss": 4.221,
"step": 77312
},
{
"epoch": 1.0,
"learning_rate": 4.872581096370916e-05,
"loss": 4.2333,
"step": 77824
},
{
"epoch": 1.0,
"learning_rate": 4.8717425016198635e-05,
"loss": 4.2255,
"step": 78336
},
{
"epoch": 1.0,
"learning_rate": 4.8709039068688115e-05,
"loss": 4.2291,
"step": 78848
},
{
"epoch": 1.0,
"learning_rate": 4.8700653121177595e-05,
"loss": 4.2122,
"step": 79360
},
{
"epoch": 1.0,
"learning_rate": 4.869228355247081e-05,
"loss": 4.2091,
"step": 79872
},
{
"epoch": 1.0,
"learning_rate": 4.868389760496029e-05,
"loss": 4.1968,
"step": 80384
},
{
"epoch": 1.0,
"learning_rate": 4.867551165744977e-05,
"loss": 4.2262,
"step": 80896
},
{
"epoch": 1.0,
"learning_rate": 4.866712570993925e-05,
"loss": 4.214,
"step": 81408
},
{
"epoch": 1.0,
"learning_rate": 4.865875614123246e-05,
"loss": 4.2074,
"step": 81920
},
{
"epoch": 1.0,
"learning_rate": 4.865037019372194e-05,
"loss": 4.2154,
"step": 82432
},
{
"epoch": 1.0,
"learning_rate": 4.8642017003818885e-05,
"loss": 4.1999,
"step": 82944
},
{
"epoch": 1.0,
"learning_rate": 4.8633631056308365e-05,
"loss": 4.1962,
"step": 83456
},
{
"epoch": 1.0,
"learning_rate": 4.8625245108797845e-05,
"loss": 4.1935,
"step": 83968
},
{
"epoch": 1.0,
"learning_rate": 4.8616859161287324e-05,
"loss": 4.1923,
"step": 84480
},
{
"epoch": 1.0,
"learning_rate": 4.8608473213776804e-05,
"loss": 4.1917,
"step": 84992
},
{
"epoch": 1.0,
"learning_rate": 4.860008726626628e-05,
"loss": 4.1862,
"step": 85504
},
{
"epoch": 1.0,
"learning_rate": 4.8591701318755764e-05,
"loss": 4.1967,
"step": 86016
},
{
"epoch": 1.0,
"learning_rate": 4.8583315371245244e-05,
"loss": 4.2026,
"step": 86528
},
{
"epoch": 1.0,
"learning_rate": 4.8574929423734724e-05,
"loss": 4.1884,
"step": 87040
},
{
"epoch": 1.0,
"learning_rate": 4.8566543476224204e-05,
"loss": 4.1801,
"step": 87552
},
{
"epoch": 1.0,
"learning_rate": 4.8558157528713684e-05,
"loss": 4.1814,
"step": 88064
},
{
"epoch": 1.0,
"learning_rate": 4.8549771581203164e-05,
"loss": 4.1847,
"step": 88576
},
{
"epoch": 1.0,
"learning_rate": 4.8541385633692644e-05,
"loss": 4.1776,
"step": 89088
},
{
"epoch": 1.0,
"learning_rate": 4.8532999686182124e-05,
"loss": 4.1679,
"step": 89600
},
{
"epoch": 1.0,
"learning_rate": 4.8524613738671604e-05,
"loss": 4.1686,
"step": 90112
},
{
"epoch": 1.0,
"learning_rate": 4.851624416996481e-05,
"loss": 4.1801,
"step": 90624
},
{
"epoch": 1.0,
"learning_rate": 4.850785822245429e-05,
"loss": 4.1566,
"step": 91136
},
{
"epoch": 1.01,
"learning_rate": 4.849947227494377e-05,
"loss": 4.1632,
"step": 91648
},
{
"epoch": 1.01,
"learning_rate": 4.849110270623698e-05,
"loss": 4.1695,
"step": 92160
},
{
"epoch": 1.01,
"learning_rate": 4.848271675872646e-05,
"loss": 4.1681,
"step": 92672
},
{
"epoch": 1.01,
"learning_rate": 4.847433081121595e-05,
"loss": 4.1703,
"step": 93184
},
{
"epoch": 1.01,
"learning_rate": 4.846594486370543e-05,
"loss": 4.166,
"step": 93696
},
{
"epoch": 1.01,
"learning_rate": 4.845757529499864e-05,
"loss": 4.1611,
"step": 94208
},
{
"epoch": 1.01,
"learning_rate": 4.844918934748812e-05,
"loss": 4.1687,
"step": 94720
},
{
"epoch": 1.01,
"learning_rate": 4.84408033999776e-05,
"loss": 4.1455,
"step": 95232
},
{
"epoch": 1.01,
"learning_rate": 4.843241745246708e-05,
"loss": 4.1553,
"step": 95744
},
{
"epoch": 1.01,
"learning_rate": 4.842403150495656e-05,
"loss": 4.1403,
"step": 96256
},
{
"epoch": 1.01,
"learning_rate": 4.841564555744604e-05,
"loss": 4.1464,
"step": 96768
},
{
"epoch": 1.01,
"learning_rate": 4.840725960993552e-05,
"loss": 4.1482,
"step": 97280
},
{
"epoch": 1.01,
"learning_rate": 4.8398873662425e-05,
"loss": 4.1504,
"step": 97792
},
{
"epoch": 1.01,
"learning_rate": 4.839050409371821e-05,
"loss": 4.146,
"step": 98304
},
{
"epoch": 1.01,
"learning_rate": 4.8382118146207687e-05,
"loss": 4.1412,
"step": 98816
},
{
"epoch": 1.01,
"learning_rate": 4.8373732198697167e-05,
"loss": 4.1413,
"step": 99328
},
{
"epoch": 1.01,
"learning_rate": 4.8365346251186646e-05,
"loss": 4.1456,
"step": 99840
},
{
"epoch": 1.01,
"learning_rate": 4.835697668247986e-05,
"loss": 4.1462,
"step": 100352
},
{
"epoch": 1.01,
"learning_rate": 4.834859073496934e-05,
"loss": 4.1324,
"step": 100864
},
{
"epoch": 1.01,
"learning_rate": 4.834020478745882e-05,
"loss": 4.1268,
"step": 101376
},
{
"epoch": 1.01,
"learning_rate": 4.8331818839948295e-05,
"loss": 4.1392,
"step": 101888
},
{
"epoch": 1.01,
"learning_rate": 4.8323432892437775e-05,
"loss": 4.1468,
"step": 102400
},
{
"epoch": 1.01,
"learning_rate": 4.831506332373099e-05,
"loss": 4.1279,
"step": 102912
},
{
"epoch": 1.01,
"learning_rate": 4.830667737622047e-05,
"loss": 4.1305,
"step": 103424
},
{
"epoch": 1.01,
"learning_rate": 4.8298291428709944e-05,
"loss": 4.127,
"step": 103936
},
{
"epoch": 1.01,
"learning_rate": 4.8289905481199424e-05,
"loss": 4.12,
"step": 104448
},
{
"epoch": 1.01,
"learning_rate": 4.8281519533688904e-05,
"loss": 4.1256,
"step": 104960
},
{
"epoch": 1.01,
"learning_rate": 4.8273133586178384e-05,
"loss": 4.1128,
"step": 105472
},
{
"epoch": 1.01,
"learning_rate": 4.82647640174716e-05,
"loss": 4.1257,
"step": 105984
},
{
"epoch": 1.01,
"learning_rate": 4.825637806996108e-05,
"loss": 4.1253,
"step": 106496
},
{
"epoch": 1.01,
"learning_rate": 4.824799212245056e-05,
"loss": 4.1213,
"step": 107008
},
{
"epoch": 1.01,
"learning_rate": 4.823960617494004e-05,
"loss": 4.1033,
"step": 107520
},
{
"epoch": 1.01,
"learning_rate": 4.823123660623325e-05,
"loss": 4.1249,
"step": 108032
},
{
"epoch": 1.01,
"learning_rate": 4.822285065872273e-05,
"loss": 4.1008,
"step": 108544
},
{
"epoch": 1.01,
"learning_rate": 4.821446471121221e-05,
"loss": 4.116,
"step": 109056
},
{
"epoch": 1.01,
"learning_rate": 4.820607876370169e-05,
"loss": 4.1178,
"step": 109568
},
{
"epoch": 1.01,
"learning_rate": 4.819769281619117e-05,
"loss": 4.1168,
"step": 110080
},
{
"epoch": 1.01,
"learning_rate": 4.818930686868065e-05,
"loss": 4.1002,
"step": 110592
},
{
"epoch": 1.01,
"learning_rate": 4.818093729997386e-05,
"loss": 4.0985,
"step": 111104
},
{
"epoch": 1.01,
"learning_rate": 4.817255135246334e-05,
"loss": 4.0987,
"step": 111616
},
{
"epoch": 1.01,
"learning_rate": 4.816416540495282e-05,
"loss": 4.107,
"step": 112128
},
{
"epoch": 1.01,
"learning_rate": 4.8155779457442305e-05,
"loss": 4.1078,
"step": 112640
},
{
"epoch": 1.01,
"learning_rate": 4.8147409888735514e-05,
"loss": 4.1021,
"step": 113152
},
{
"epoch": 1.01,
"learning_rate": 4.8139023941224994e-05,
"loss": 4.11,
"step": 113664
},
{
"epoch": 1.01,
"learning_rate": 4.8130637993714474e-05,
"loss": 4.1112,
"step": 114176
},
{
"epoch": 1.01,
"learning_rate": 4.8122252046203954e-05,
"loss": 4.1073,
"step": 114688
},
{
"epoch": 1.01,
"learning_rate": 4.811388247749716e-05,
"loss": 4.0949,
"step": 115200
},
{
"epoch": 1.01,
"learning_rate": 4.810549652998664e-05,
"loss": 4.1037,
"step": 115712
},
{
"epoch": 1.01,
"learning_rate": 4.809711058247612e-05,
"loss": 4.1023,
"step": 116224
},
{
"epoch": 1.01,
"learning_rate": 4.80887246349656e-05,
"loss": 4.0827,
"step": 116736
},
{
"epoch": 1.01,
"learning_rate": 4.808035506625881e-05,
"loss": 4.1013,
"step": 117248
},
{
"epoch": 1.01,
"learning_rate": 4.807196911874829e-05,
"loss": 4.0881,
"step": 117760
},
{
"epoch": 1.01,
"learning_rate": 4.806358317123777e-05,
"loss": 4.1087,
"step": 118272
},
{
"epoch": 1.01,
"learning_rate": 4.805519722372726e-05,
"loss": 4.0957,
"step": 118784
},
{
"epoch": 1.01,
"learning_rate": 4.804681127621674e-05,
"loss": 4.0748,
"step": 119296
},
{
"epoch": 1.01,
"learning_rate": 4.803844170750995e-05,
"loss": 4.0919,
"step": 119808
},
{
"epoch": 1.01,
"learning_rate": 4.803005575999943e-05,
"loss": 4.0827,
"step": 120320
},
{
"epoch": 1.01,
"learning_rate": 4.802166981248891e-05,
"loss": 4.0954,
"step": 120832
},
{
"epoch": 1.01,
"learning_rate": 4.801328386497839e-05,
"loss": 4.0874,
"step": 121344
},
{
"epoch": 1.01,
"learning_rate": 4.8004930675075325e-05,
"loss": 4.0816,
"step": 121856
},
{
"epoch": 1.02,
"learning_rate": 4.7996544727564805e-05,
"loss": 4.0854,
"step": 122368
},
{
"epoch": 1.02,
"learning_rate": 4.7988158780054285e-05,
"loss": 4.075,
"step": 122880
},
{
"epoch": 1.02,
"learning_rate": 4.7979772832543765e-05,
"loss": 4.0701,
"step": 123392
},
{
"epoch": 1.02,
"learning_rate": 4.7971386885033245e-05,
"loss": 4.0688,
"step": 123904
},
{
"epoch": 1.02,
"learning_rate": 4.7963000937522725e-05,
"loss": 4.0751,
"step": 124416
},
{
"epoch": 1.02,
"learning_rate": 4.795463136881594e-05,
"loss": 4.0853,
"step": 124928
},
{
"epoch": 1.02,
"learning_rate": 4.794624542130542e-05,
"loss": 4.0764,
"step": 125440
},
{
"epoch": 1.02,
"learning_rate": 4.79378594737949e-05,
"loss": 4.0665,
"step": 125952
},
{
"epoch": 1.02,
"learning_rate": 4.792947352628438e-05,
"loss": 4.0682,
"step": 126464
},
{
"epoch": 1.02,
"learning_rate": 4.792108757877386e-05,
"loss": 4.0729,
"step": 126976
},
{
"epoch": 1.02,
"learning_rate": 4.791271801006707e-05,
"loss": 4.0791,
"step": 127488
},
{
"epoch": 1.02,
"learning_rate": 4.790433206255655e-05,
"loss": 4.0726,
"step": 128000
},
{
"epoch": 1.02,
"learning_rate": 4.789594611504603e-05,
"loss": 4.069,
"step": 128512
},
{
"epoch": 1.02,
"learning_rate": 4.788756016753551e-05,
"loss": 4.067,
"step": 129024
},
{
"epoch": 1.02,
"learning_rate": 4.787917422002499e-05,
"loss": 4.0742,
"step": 129536
},
{
"epoch": 1.02,
"learning_rate": 4.787078827251447e-05,
"loss": 4.0518,
"step": 130048
},
{
"epoch": 1.02,
"learning_rate": 4.786240232500395e-05,
"loss": 4.064,
"step": 130560
},
{
"epoch": 1.02,
"learning_rate": 4.7854032756297166e-05,
"loss": 4.0679,
"step": 131072
},
{
"epoch": 1.02,
"learning_rate": 4.7845646808786646e-05,
"loss": 4.0584,
"step": 131584
},
{
"epoch": 1.02,
"learning_rate": 4.7837260861276126e-05,
"loss": 4.0555,
"step": 132096
},
{
"epoch": 1.02,
"learning_rate": 4.7828874913765606e-05,
"loss": 4.0569,
"step": 132608
},
{
"epoch": 1.02,
"learning_rate": 4.782048896625508e-05,
"loss": 4.0581,
"step": 133120
},
{
"epoch": 1.02,
"learning_rate": 4.781210301874456e-05,
"loss": 4.0567,
"step": 133632
},
{
"epoch": 1.02,
"learning_rate": 4.780371707123404e-05,
"loss": 4.0594,
"step": 134144
},
{
"epoch": 1.02,
"learning_rate": 4.779533112372352e-05,
"loss": 4.0538,
"step": 134656
},
{
"epoch": 1.02,
"learning_rate": 4.778696155501673e-05,
"loss": 4.067,
"step": 135168
},
{
"epoch": 1.02,
"learning_rate": 4.777857560750621e-05,
"loss": 4.0622,
"step": 135680
},
{
"epoch": 1.02,
"learning_rate": 4.7770206038799423e-05,
"loss": 4.0573,
"step": 136192
},
{
"epoch": 1.02,
"learning_rate": 4.7761820091288903e-05,
"loss": 4.0591,
"step": 136704
},
{
"epoch": 1.02,
"learning_rate": 4.7753434143778383e-05,
"loss": 4.0552,
"step": 137216
},
{
"epoch": 1.02,
"learning_rate": 4.774504819626786e-05,
"loss": 4.0523,
"step": 137728
},
{
"epoch": 1.02,
"learning_rate": 4.773666224875734e-05,
"loss": 4.045,
"step": 138240
},
{
"epoch": 1.02,
"learning_rate": 4.772827630124682e-05,
"loss": 4.0436,
"step": 138752
},
{
"epoch": 1.02,
"learning_rate": 4.77198903537363e-05,
"loss": 4.0504,
"step": 139264
},
{
"epoch": 1.02,
"learning_rate": 4.771150440622578e-05,
"loss": 4.052,
"step": 139776
},
{
"epoch": 1.02,
"learning_rate": 4.770313483751899e-05,
"loss": 4.0475,
"step": 140288
},
{
"epoch": 1.02,
"learning_rate": 4.769474889000847e-05,
"loss": 4.0499,
"step": 140800
},
{
"epoch": 1.02,
"learning_rate": 4.768636294249795e-05,
"loss": 4.044,
"step": 141312
},
{
"epoch": 1.02,
"learning_rate": 4.767797699498743e-05,
"loss": 4.044,
"step": 141824
},
{
"epoch": 1.02,
"learning_rate": 4.766960742628064e-05,
"loss": 4.0457,
"step": 142336
},
{
"epoch": 1.02,
"learning_rate": 4.766122147877012e-05,
"loss": 4.0516,
"step": 142848
},
{
"epoch": 1.02,
"learning_rate": 4.76528355312596e-05,
"loss": 4.0364,
"step": 143360
},
{
"epoch": 1.02,
"learning_rate": 4.764444958374909e-05,
"loss": 4.0262,
"step": 143872
},
{
"epoch": 1.02,
"learning_rate": 4.76360800150423e-05,
"loss": 4.0324,
"step": 144384
},
{
"epoch": 1.02,
"learning_rate": 4.762769406753178e-05,
"loss": 4.049,
"step": 144896
},
{
"epoch": 1.02,
"learning_rate": 4.761930812002126e-05,
"loss": 4.0478,
"step": 145408
},
{
"epoch": 1.02,
"learning_rate": 4.761092217251074e-05,
"loss": 4.041,
"step": 145920
},
{
"epoch": 1.02,
"learning_rate": 4.7602552603803946e-05,
"loss": 4.036,
"step": 146432
},
{
"epoch": 1.02,
"learning_rate": 4.7594166656293426e-05,
"loss": 4.047,
"step": 146944
},
{
"epoch": 1.02,
"learning_rate": 4.7585780708782906e-05,
"loss": 4.0425,
"step": 147456
},
{
"epoch": 1.02,
"learning_rate": 4.7577394761272386e-05,
"loss": 4.0403,
"step": 147968
},
{
"epoch": 1.02,
"learning_rate": 4.7569025192565595e-05,
"loss": 4.034,
"step": 148480
},
{
"epoch": 1.02,
"learning_rate": 4.7560639245055075e-05,
"loss": 4.0336,
"step": 148992
},
{
"epoch": 1.02,
"learning_rate": 4.7552253297544555e-05,
"loss": 4.0346,
"step": 149504
},
{
"epoch": 1.02,
"learning_rate": 4.754388372883777e-05,
"loss": 4.0248,
"step": 150016
},
{
"epoch": 1.02,
"learning_rate": 4.753549778132725e-05,
"loss": 4.0429,
"step": 150528
},
{
"epoch": 1.02,
"learning_rate": 4.752711183381673e-05,
"loss": 4.0259,
"step": 151040
},
{
"epoch": 1.02,
"learning_rate": 4.751872588630621e-05,
"loss": 4.0365,
"step": 151552
},
{
"epoch": 1.02,
"learning_rate": 4.751033993879569e-05,
"loss": 4.0339,
"step": 152064
},
{
"epoch": 1.02,
"learning_rate": 4.750195399128517e-05,
"loss": 4.0221,
"step": 152576
},
{
"epoch": 1.03,
"eval_loss": 4.041590690612793,
"eval_runtime": 316.7556,
"eval_samples_per_second": 1204.686,
"eval_steps_per_second": 37.647,
"step": 152640
},
{
"epoch": 0.0,
"learning_rate": 4.749356804377465e-05,
"loss": 4.0258,
"step": 153088
},
{
"epoch": 0.0,
"learning_rate": 4.748518209626413e-05,
"loss": 4.0189,
"step": 153600
},
{
"epoch": 0.0,
"learning_rate": 4.747679614875361e-05,
"loss": 4.0291,
"step": 154112
},
{
"epoch": 0.0,
"learning_rate": 4.7468410201243084e-05,
"loss": 4.0247,
"step": 154624
},
{
"epoch": 0.0,
"learning_rate": 4.7460024253732563e-05,
"loss": 4.0321,
"step": 155136
},
{
"epoch": 0.0,
"learning_rate": 4.7451638306222043e-05,
"loss": 4.0141,
"step": 155648
},
{
"epoch": 0.0,
"learning_rate": 4.744325235871152e-05,
"loss": 4.014,
"step": 156160
},
{
"epoch": 0.0,
"learning_rate": 4.7434866411201e-05,
"loss": 4.0013,
"step": 156672
},
{
"epoch": 0.0,
"learning_rate": 4.742648046369049e-05,
"loss": 4.0347,
"step": 157184
},
{
"epoch": 0.0,
"learning_rate": 4.741809451617997e-05,
"loss": 4.0215,
"step": 157696
},
{
"epoch": 0.0,
"learning_rate": 4.740970856866945e-05,
"loss": 4.0117,
"step": 158208
},
{
"epoch": 0.0,
"learning_rate": 4.740133899996266e-05,
"loss": 4.0264,
"step": 158720
},
{
"epoch": 0.0,
"learning_rate": 4.7392985810059604e-05,
"loss": 4.0078,
"step": 159232
},
{
"epoch": 0.0,
"learning_rate": 4.7384599862549084e-05,
"loss": 4.0119,
"step": 159744
},
{
"epoch": 0.0,
"learning_rate": 4.737621391503856e-05,
"loss": 4.0022,
"step": 160256
},
{
"epoch": 0.0,
"learning_rate": 4.736782796752804e-05,
"loss": 4.004,
"step": 160768
},
{
"epoch": 0.0,
"learning_rate": 4.735944202001752e-05,
"loss": 4.0101,
"step": 161280
},
{
"epoch": 0.0,
"learning_rate": 4.7351056072507e-05,
"loss": 3.9997,
"step": 161792
},
{
"epoch": 0.0,
"learning_rate": 4.734267012499648e-05,
"loss": 4.0154,
"step": 162304
},
{
"epoch": 0.0,
"learning_rate": 4.733428417748596e-05,
"loss": 4.02,
"step": 162816
},
{
"epoch": 0.0,
"learning_rate": 4.7325898229975444e-05,
"loss": 4.008,
"step": 163328
},
{
"epoch": 0.0,
"learning_rate": 4.7317512282464924e-05,
"loss": 4.0041,
"step": 163840
},
{
"epoch": 0.0,
"learning_rate": 4.7309126334954404e-05,
"loss": 4.0035,
"step": 164352
},
{
"epoch": 0.0,
"learning_rate": 4.730075676624761e-05,
"loss": 4.0057,
"step": 164864
},
{
"epoch": 0.0,
"learning_rate": 4.729237081873709e-05,
"loss": 3.9958,
"step": 165376
},
{
"epoch": 0.0,
"learning_rate": 4.728398487122657e-05,
"loss": 3.9974,
"step": 165888
},
{
"epoch": 0.0,
"learning_rate": 4.727559892371605e-05,
"loss": 3.9919,
"step": 166400
},
{
"epoch": 0.0,
"learning_rate": 4.726722935500926e-05,
"loss": 4.0083,
"step": 166912
},
{
"epoch": 0.0,
"learning_rate": 4.725884340749874e-05,
"loss": 3.9805,
"step": 167424
},
{
"epoch": 0.01,
"learning_rate": 4.725045745998822e-05,
"loss": 3.9932,
"step": 167936
},
{
"epoch": 0.01,
"learning_rate": 4.72420715124777e-05,
"loss": 3.9989,
"step": 168448
},
{
"epoch": 0.01,
"learning_rate": 4.723370194377091e-05,
"loss": 4.0013,
"step": 168960
},
{
"epoch": 0.01,
"learning_rate": 4.72253159962604e-05,
"loss": 3.9997,
"step": 169472
},
{
"epoch": 0.01,
"learning_rate": 4.721693004874988e-05,
"loss": 3.9979,
"step": 169984
},
{
"epoch": 0.01,
"learning_rate": 4.720854410123936e-05,
"loss": 3.9925,
"step": 170496
},
{
"epoch": 0.01,
"learning_rate": 4.7200174532532566e-05,
"loss": 4.0012,
"step": 171008
},
{
"epoch": 0.01,
"learning_rate": 4.7191788585022046e-05,
"loss": 3.9846,
"step": 171520
},
{
"epoch": 0.01,
"learning_rate": 4.7183402637511526e-05,
"loss": 3.9898,
"step": 172032
},
{
"epoch": 0.01,
"learning_rate": 4.7175016690001006e-05,
"loss": 3.9738,
"step": 172544
},
{
"epoch": 0.01,
"learning_rate": 4.7166647121294215e-05,
"loss": 3.9844,
"step": 173056
},
{
"epoch": 0.01,
"learning_rate": 4.7158261173783695e-05,
"loss": 3.9808,
"step": 173568
},
{
"epoch": 0.01,
"learning_rate": 4.7149875226273175e-05,
"loss": 3.9931,
"step": 174080
},
{
"epoch": 0.01,
"learning_rate": 4.7141489278762655e-05,
"loss": 3.9848,
"step": 174592
},
{
"epoch": 0.01,
"learning_rate": 4.7133119710055864e-05,
"loss": 3.9826,
"step": 175104
},
{
"epoch": 0.01,
"learning_rate": 4.712473376254535e-05,
"loss": 3.9852,
"step": 175616
},
{
"epoch": 0.01,
"learning_rate": 4.711634781503483e-05,
"loss": 3.987,
"step": 176128
},
{
"epoch": 0.01,
"learning_rate": 4.710796186752431e-05,
"loss": 3.9863,
"step": 176640
},
{
"epoch": 0.01,
"learning_rate": 4.709959229881752e-05,
"loss": 3.9737,
"step": 177152
},
{
"epoch": 0.01,
"learning_rate": 4.7091206351307e-05,
"loss": 3.9678,
"step": 177664
},
{
"epoch": 0.01,
"learning_rate": 4.708282040379648e-05,
"loss": 3.9875,
"step": 178176
},
{
"epoch": 0.01,
"learning_rate": 4.707443445628596e-05,
"loss": 3.9886,
"step": 178688
},
{
"epoch": 0.01,
"learning_rate": 4.706606488757917e-05,
"loss": 3.9766,
"step": 179200
},
{
"epoch": 0.01,
"learning_rate": 4.705767894006865e-05,
"loss": 3.9762,
"step": 179712
},
{
"epoch": 0.01,
"learning_rate": 4.704929299255813e-05,
"loss": 3.9714,
"step": 180224
},
{
"epoch": 0.01,
"learning_rate": 4.704090704504761e-05,
"loss": 3.9687,
"step": 180736
},
{
"epoch": 0.01,
"learning_rate": 4.703253747634082e-05,
"loss": 3.9757,
"step": 181248
},
{
"epoch": 0.01,
"learning_rate": 4.7024151528830305e-05,
"loss": 3.9645,
"step": 181760
},
{
"epoch": 0.01,
"learning_rate": 4.7015765581319785e-05,
"loss": 3.9763,
"step": 182272
},
{
"epoch": 0.01,
"learning_rate": 4.7007379633809265e-05,
"loss": 3.9748,
"step": 182784
},
{
"epoch": 0.01,
"learning_rate": 4.6999010065102474e-05,
"loss": 3.9701,
"step": 183296
},
{
"epoch": 0.01,
"learning_rate": 4.6990624117591954e-05,
"loss": 3.9576,
"step": 183808
},
{
"epoch": 0.01,
"learning_rate": 4.6982238170081434e-05,
"loss": 3.9787,
"step": 184320
},
{
"epoch": 0.01,
"learning_rate": 4.6973852222570914e-05,
"loss": 3.9556,
"step": 184832
},
{
"epoch": 0.01,
"learning_rate": 4.696548265386412e-05,
"loss": 3.9687,
"step": 185344
},
{
"epoch": 0.01,
"learning_rate": 4.69570967063536e-05,
"loss": 3.968,
"step": 185856
},
{
"epoch": 0.01,
"learning_rate": 4.694871075884308e-05,
"loss": 3.9744,
"step": 186368
},
{
"epoch": 0.01,
"learning_rate": 4.694032481133256e-05,
"loss": 3.9552,
"step": 186880
},
{
"epoch": 0.01,
"learning_rate": 4.693195524262577e-05,
"loss": 3.9602,
"step": 187392
},
{
"epoch": 0.01,
"learning_rate": 4.692356929511526e-05,
"loss": 3.9513,
"step": 187904
},
{
"epoch": 0.01,
"learning_rate": 4.691518334760474e-05,
"loss": 3.9662,
"step": 188416
},
{
"epoch": 0.01,
"learning_rate": 4.690679740009422e-05,
"loss": 3.9674,
"step": 188928
},
{
"epoch": 0.01,
"learning_rate": 4.689842783138743e-05,
"loss": 3.964,
"step": 189440
},
{
"epoch": 0.01,
"learning_rate": 4.689004188387691e-05,
"loss": 3.9687,
"step": 189952
},
{
"epoch": 0.01,
"learning_rate": 4.688165593636639e-05,
"loss": 3.969,
"step": 190464
},
{
"epoch": 0.01,
"learning_rate": 4.687326998885587e-05,
"loss": 3.9731,
"step": 190976
},
{
"epoch": 0.01,
"learning_rate": 4.6864900420149076e-05,
"loss": 3.9532,
"step": 191488
},
{
"epoch": 0.01,
"learning_rate": 4.6856514472638556e-05,
"loss": 3.9644,
"step": 192000
},
{
"epoch": 0.01,
"learning_rate": 4.6848128525128036e-05,
"loss": 3.9694,
"step": 192512
},
{
"epoch": 0.01,
"learning_rate": 4.683974257761751e-05,
"loss": 3.9466,
"step": 193024
},
{
"epoch": 0.01,
"learning_rate": 4.6831373008910725e-05,
"loss": 3.963,
"step": 193536
},
{
"epoch": 0.01,
"learning_rate": 4.6822987061400205e-05,
"loss": 3.9502,
"step": 194048
},
{
"epoch": 0.01,
"learning_rate": 4.681460111388969e-05,
"loss": 3.9744,
"step": 194560
},
{
"epoch": 0.01,
"learning_rate": 4.6806215166379165e-05,
"loss": 3.9608,
"step": 195072
},
{
"epoch": 0.01,
"learning_rate": 4.679784559767238e-05,
"loss": 3.9401,
"step": 195584
},
{
"epoch": 0.01,
"learning_rate": 4.678945965016186e-05,
"loss": 3.954,
"step": 196096
},
{
"epoch": 0.01,
"learning_rate": 4.678107370265134e-05,
"loss": 3.9531,
"step": 196608
},
{
"epoch": 0.01,
"learning_rate": 4.6772687755140814e-05,
"loss": 3.9614,
"step": 197120
},
{
"epoch": 0.01,
"learning_rate": 4.676431818643403e-05,
"loss": 3.9556,
"step": 197632
},
{
"epoch": 0.01,
"learning_rate": 4.675594861772724e-05,
"loss": 3.9511,
"step": 198144
},
{
"epoch": 0.02,
"learning_rate": 4.674756267021672e-05,
"loss": 3.9521,
"step": 198656
},
{
"epoch": 0.02,
"learning_rate": 4.67391767227062e-05,
"loss": 3.9479,
"step": 199168
},
{
"epoch": 0.02,
"learning_rate": 4.673079077519568e-05,
"loss": 3.9417,
"step": 199680
},
{
"epoch": 0.02,
"learning_rate": 4.672240482768516e-05,
"loss": 3.9355,
"step": 200192
},
{
"epoch": 0.02,
"learning_rate": 4.671401888017464e-05,
"loss": 3.9484,
"step": 200704
},
{
"epoch": 0.02,
"learning_rate": 4.6705649311467855e-05,
"loss": 3.9567,
"step": 201216
},
{
"epoch": 0.02,
"learning_rate": 4.6697263363957335e-05,
"loss": 3.9493,
"step": 201728
},
{
"epoch": 0.02,
"learning_rate": 4.6688877416446815e-05,
"loss": 3.9379,
"step": 202240
},
{
"epoch": 0.02,
"learning_rate": 4.668049146893629e-05,
"loss": 3.9409,
"step": 202752
},
{
"epoch": 0.02,
"learning_rate": 4.667210552142577e-05,
"loss": 3.9428,
"step": 203264
},
{
"epoch": 0.02,
"learning_rate": 4.6663735952718984e-05,
"loss": 3.9549,
"step": 203776
},
{
"epoch": 0.02,
"learning_rate": 4.665535000520846e-05,
"loss": 3.9465,
"step": 204288
},
{
"epoch": 0.02,
"learning_rate": 4.664696405769794e-05,
"loss": 3.9442,
"step": 204800
},
{
"epoch": 0.02,
"learning_rate": 4.663857811018742e-05,
"loss": 3.9378,
"step": 205312
},
{
"epoch": 0.02,
"learning_rate": 4.66301921626769e-05,
"loss": 3.9495,
"step": 205824
},
{
"epoch": 0.02,
"learning_rate": 4.6621806215166383e-05,
"loss": 3.9261,
"step": 206336
},
{
"epoch": 0.02,
"learning_rate": 4.6613420267655863e-05,
"loss": 3.9418,
"step": 206848
},
{
"epoch": 0.02,
"learning_rate": 4.660505069894907e-05,
"loss": 3.9397,
"step": 207360
},
{
"epoch": 0.02,
"learning_rate": 4.659666475143855e-05,
"loss": 3.9383,
"step": 207872
},
{
"epoch": 0.02,
"learning_rate": 4.658827880392803e-05,
"loss": 3.9304,
"step": 208384
},
{
"epoch": 0.02,
"learning_rate": 4.657989285641751e-05,
"loss": 3.9368,
"step": 208896
},
{
"epoch": 0.02,
"learning_rate": 4.657150690890699e-05,
"loss": 3.9367,
"step": 209408
},
{
"epoch": 0.02,
"learning_rate": 4.65631373402002e-05,
"loss": 3.9358,
"step": 209920
},
{
"epoch": 0.02,
"learning_rate": 4.655475139268968e-05,
"loss": 3.9395,
"step": 210432
},
{
"epoch": 0.02,
"learning_rate": 4.654636544517916e-05,
"loss": 3.9353,
"step": 210944
},
{
"epoch": 0.02,
"learning_rate": 4.653797949766864e-05,
"loss": 3.9445,
"step": 211456
},
{
"epoch": 0.02,
"learning_rate": 4.652959355015812e-05,
"loss": 3.9434,
"step": 211968
},
{
"epoch": 0.02,
"learning_rate": 4.652122398145134e-05,
"loss": 3.9406,
"step": 212480
},
{
"epoch": 0.02,
"learning_rate": 4.651283803394082e-05,
"loss": 3.9399,
"step": 212992
},
{
"epoch": 0.02,
"learning_rate": 4.65044520864303e-05,
"loss": 3.9367,
"step": 213504
},
{
"epoch": 0.02,
"learning_rate": 4.649606613891978e-05,
"loss": 3.9302,
"step": 214016
},
{
"epoch": 0.02,
"learning_rate": 4.648768019140926e-05,
"loss": 3.9288,
"step": 214528
},
{
"epoch": 0.02,
"learning_rate": 4.647929424389874e-05,
"loss": 3.9305,
"step": 215040
},
{
"epoch": 0.02,
"learning_rate": 4.647090829638822e-05,
"loss": 3.9285,
"step": 215552
},
{
"epoch": 0.02,
"learning_rate": 4.64625223488777e-05,
"loss": 3.9401,
"step": 216064
},
{
"epoch": 0.02,
"learning_rate": 4.6454152780170906e-05,
"loss": 3.929,
"step": 216576
},
{
"epoch": 0.02,
"learning_rate": 4.6445766832660386e-05,
"loss": 3.9352,
"step": 217088
},
{
"epoch": 0.02,
"learning_rate": 4.6437380885149866e-05,
"loss": 3.9317,
"step": 217600
},
{
"epoch": 0.02,
"learning_rate": 4.6429011316443075e-05,
"loss": 3.9271,
"step": 218112
},
{
"epoch": 0.02,
"learning_rate": 4.6420625368932555e-05,
"loss": 3.9268,
"step": 218624
},
{
"epoch": 0.02,
"learning_rate": 4.6412239421422035e-05,
"loss": 3.9405,
"step": 219136
},
{
"epoch": 0.02,
"learning_rate": 4.640385347391152e-05,
"loss": 3.922,
"step": 219648
},
{
"epoch": 0.02,
"learning_rate": 4.639548390520473e-05,
"loss": 3.9194,
"step": 220160
},
{
"epoch": 0.02,
"learning_rate": 4.638709795769421e-05,
"loss": 3.9142,
"step": 220672
},
{
"epoch": 0.02,
"learning_rate": 4.637871201018369e-05,
"loss": 3.934,
"step": 221184
},
{
"epoch": 0.02,
"learning_rate": 4.63703424414769e-05,
"loss": 3.937,
"step": 221696
},
{
"epoch": 0.02,
"learning_rate": 4.636195649396638e-05,
"loss": 3.934,
"step": 222208
},
{
"epoch": 0.02,
"learning_rate": 4.635357054645586e-05,
"loss": 3.9234,
"step": 222720
},
{
"epoch": 0.02,
"learning_rate": 4.634518459894534e-05,
"loss": 3.9363,
"step": 223232
},
{
"epoch": 0.02,
"learning_rate": 4.633681503023855e-05,
"loss": 3.9261,
"step": 223744
},
{
"epoch": 0.02,
"learning_rate": 4.632842908272803e-05,
"loss": 3.9311,
"step": 224256
},
{
"epoch": 0.02,
"learning_rate": 4.632004313521751e-05,
"loss": 3.9215,
"step": 224768
},
{
"epoch": 0.02,
"learning_rate": 4.631165718770699e-05,
"loss": 3.9266,
"step": 225280
},
{
"epoch": 0.02,
"learning_rate": 4.6303271240196475e-05,
"loss": 3.9236,
"step": 225792
},
{
"epoch": 0.02,
"learning_rate": 4.6294901671489684e-05,
"loss": 3.9129,
"step": 226304
},
{
"epoch": 0.02,
"learning_rate": 4.6286515723979164e-05,
"loss": 3.9341,
"step": 226816
},
{
"epoch": 0.02,
"learning_rate": 4.6278129776468644e-05,
"loss": 3.9224,
"step": 227328
},
{
"epoch": 0.02,
"learning_rate": 4.626974382895812e-05,
"loss": 3.9248,
"step": 227840
},
{
"epoch": 0.02,
"learning_rate": 4.62613578814476e-05,
"loss": 3.9253,
"step": 228352
},
{
"epoch": 0.02,
"learning_rate": 4.625297193393708e-05,
"loss": 3.9131,
"step": 228864
},
{
"epoch": 0.03,
"eval_loss": 3.9674437046051025,
"eval_runtime": 304.1285,
"eval_samples_per_second": 1254.703,
"eval_steps_per_second": 39.21,
"step": 228960
}
],
"logging_steps": 512,
"max_steps": 3052726,
"num_train_epochs": 9223372036854775807,
"save_steps": 10,
"total_flos": 1.5293807645947085e+17,
"trial_name": null,
"trial_params": null
}