abhinavp's picture
Training in progress, epoch 0, checkpoint
2d14ca5 verified
raw
history blame
56.1 kB
{
"best_metric": 3.9528167247772217,
"best_model_checkpoint": "/mmfs1/gscratch/stf/abhinavp/corpus-filtering/outputs/npi-sim-ques/transformer/2/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": 11.0359,
"step": 1
},
{
"epoch": 0.0,
"learning_rate": 4.999161405248948e-05,
"loss": 6.8165,
"step": 512
},
{
"epoch": 0.0,
"learning_rate": 4.998322810497896e-05,
"loss": 6.186,
"step": 1024
},
{
"epoch": 0.0,
"learning_rate": 4.997484215746844e-05,
"loss": 5.9717,
"step": 1536
},
{
"epoch": 0.0,
"learning_rate": 4.996645620995792e-05,
"loss": 5.8095,
"step": 2048
},
{
"epoch": 0.0,
"learning_rate": 4.99580702624474e-05,
"loss": 5.7099,
"step": 2560
},
{
"epoch": 0.0,
"learning_rate": 4.994968431493688e-05,
"loss": 5.603,
"step": 3072
},
{
"epoch": 0.0,
"learning_rate": 4.994129836742636e-05,
"loss": 5.5345,
"step": 3584
},
{
"epoch": 0.0,
"learning_rate": 4.993291241991584e-05,
"loss": 5.4553,
"step": 4096
},
{
"epoch": 0.0,
"learning_rate": 4.992452647240532e-05,
"loss": 5.407,
"step": 4608
},
{
"epoch": 0.0,
"learning_rate": 4.99161405248948e-05,
"loss": 5.3469,
"step": 5120
},
{
"epoch": 0.0,
"learning_rate": 4.990775457738428e-05,
"loss": 5.3257,
"step": 5632
},
{
"epoch": 0.0,
"learning_rate": 4.989938500867749e-05,
"loss": 5.273,
"step": 6144
},
{
"epoch": 0.0,
"learning_rate": 4.989099906116697e-05,
"loss": 5.2071,
"step": 6656
},
{
"epoch": 0.0,
"learning_rate": 4.988261311365645e-05,
"loss": 5.1798,
"step": 7168
},
{
"epoch": 0.0,
"learning_rate": 4.987422716614593e-05,
"loss": 5.1367,
"step": 7680
},
{
"epoch": 0.0,
"learning_rate": 4.986584121863541e-05,
"loss": 5.1191,
"step": 8192
},
{
"epoch": 0.0,
"learning_rate": 4.985745527112489e-05,
"loss": 5.0748,
"step": 8704
},
{
"epoch": 0.0,
"learning_rate": 4.984906932361437e-05,
"loss": 5.045,
"step": 9216
},
{
"epoch": 0.0,
"learning_rate": 4.984068337610385e-05,
"loss": 5.0215,
"step": 9728
},
{
"epoch": 0.0,
"learning_rate": 4.983229742859333e-05,
"loss": 5.0009,
"step": 10240
},
{
"epoch": 0.0,
"learning_rate": 4.982391148108281e-05,
"loss": 4.9793,
"step": 10752
},
{
"epoch": 0.0,
"learning_rate": 4.9815541912376026e-05,
"loss": 4.9573,
"step": 11264
},
{
"epoch": 0.0,
"learning_rate": 4.9807155964865506e-05,
"loss": 4.9386,
"step": 11776
},
{
"epoch": 0.0,
"learning_rate": 4.9798770017354986e-05,
"loss": 4.9166,
"step": 12288
},
{
"epoch": 0.0,
"learning_rate": 4.9790384069844466e-05,
"loss": 4.895,
"step": 12800
},
{
"epoch": 0.0,
"learning_rate": 4.9782014501137675e-05,
"loss": 4.8676,
"step": 13312
},
{
"epoch": 0.0,
"learning_rate": 4.9773628553627155e-05,
"loss": 4.8534,
"step": 13824
},
{
"epoch": 0.0,
"learning_rate": 4.9765242606116635e-05,
"loss": 4.831,
"step": 14336
},
{
"epoch": 0.0,
"learning_rate": 4.9756856658606115e-05,
"loss": 4.8227,
"step": 14848
},
{
"epoch": 0.01,
"learning_rate": 4.9748487089899324e-05,
"loss": 4.8075,
"step": 15360
},
{
"epoch": 0.01,
"learning_rate": 4.9740101142388804e-05,
"loss": 4.7907,
"step": 15872
},
{
"epoch": 0.01,
"learning_rate": 4.9731715194878284e-05,
"loss": 4.7807,
"step": 16384
},
{
"epoch": 0.01,
"learning_rate": 4.97233456261715e-05,
"loss": 4.7604,
"step": 16896
},
{
"epoch": 0.01,
"learning_rate": 4.971495967866098e-05,
"loss": 4.7578,
"step": 17408
},
{
"epoch": 0.01,
"learning_rate": 4.970657373115046e-05,
"loss": 4.7533,
"step": 17920
},
{
"epoch": 0.01,
"learning_rate": 4.969818778363994e-05,
"loss": 4.7343,
"step": 18432
},
{
"epoch": 0.01,
"learning_rate": 4.968980183612942e-05,
"loss": 4.7003,
"step": 18944
},
{
"epoch": 0.01,
"learning_rate": 4.96814158886189e-05,
"loss": 4.7084,
"step": 19456
},
{
"epoch": 0.01,
"learning_rate": 4.967302994110837e-05,
"loss": 4.6815,
"step": 19968
},
{
"epoch": 0.01,
"learning_rate": 4.966466037240159e-05,
"loss": 4.6757,
"step": 20480
},
{
"epoch": 0.01,
"learning_rate": 4.965627442489107e-05,
"loss": 4.6741,
"step": 20992
},
{
"epoch": 0.01,
"learning_rate": 4.964788847738054e-05,
"loss": 4.6657,
"step": 21504
},
{
"epoch": 0.01,
"learning_rate": 4.963950252987002e-05,
"loss": 4.6488,
"step": 22016
},
{
"epoch": 0.01,
"learning_rate": 4.96311165823595e-05,
"loss": 4.6506,
"step": 22528
},
{
"epoch": 0.01,
"learning_rate": 4.962274701365272e-05,
"loss": 4.6409,
"step": 23040
},
{
"epoch": 0.01,
"learning_rate": 4.96143610661422e-05,
"loss": 4.6324,
"step": 23552
},
{
"epoch": 0.01,
"learning_rate": 4.960597511863168e-05,
"loss": 4.6245,
"step": 24064
},
{
"epoch": 0.01,
"learning_rate": 4.959758917112116e-05,
"loss": 4.6017,
"step": 24576
},
{
"epoch": 0.01,
"learning_rate": 4.9589219602414374e-05,
"loss": 4.5961,
"step": 25088
},
{
"epoch": 0.01,
"learning_rate": 4.958083365490385e-05,
"loss": 4.6023,
"step": 25600
},
{
"epoch": 0.01,
"learning_rate": 4.957244770739333e-05,
"loss": 4.5866,
"step": 26112
},
{
"epoch": 0.01,
"learning_rate": 4.956406175988281e-05,
"loss": 4.572,
"step": 26624
},
{
"epoch": 0.01,
"learning_rate": 4.9555692191176016e-05,
"loss": 4.559,
"step": 27136
},
{
"epoch": 0.01,
"learning_rate": 4.9547306243665496e-05,
"loss": 4.5644,
"step": 27648
},
{
"epoch": 0.01,
"learning_rate": 4.9538920296154976e-05,
"loss": 4.5467,
"step": 28160
},
{
"epoch": 0.01,
"learning_rate": 4.9530534348644456e-05,
"loss": 4.5631,
"step": 28672
},
{
"epoch": 0.01,
"learning_rate": 4.9522148401133936e-05,
"loss": 4.5354,
"step": 29184
},
{
"epoch": 0.01,
"learning_rate": 4.951377883242715e-05,
"loss": 4.5416,
"step": 29696
},
{
"epoch": 0.01,
"learning_rate": 4.950539288491663e-05,
"loss": 4.5333,
"step": 30208
},
{
"epoch": 0.01,
"learning_rate": 4.949700693740611e-05,
"loss": 4.5096,
"step": 30720
},
{
"epoch": 0.01,
"learning_rate": 4.948862098989559e-05,
"loss": 4.5184,
"step": 31232
},
{
"epoch": 0.01,
"learning_rate": 4.94802514211888e-05,
"loss": 4.5061,
"step": 31744
},
{
"epoch": 0.01,
"learning_rate": 4.947186547367828e-05,
"loss": 4.4972,
"step": 32256
},
{
"epoch": 0.01,
"learning_rate": 4.946347952616776e-05,
"loss": 4.4925,
"step": 32768
},
{
"epoch": 0.01,
"learning_rate": 4.945509357865724e-05,
"loss": 4.5083,
"step": 33280
},
{
"epoch": 0.01,
"learning_rate": 4.944672400995045e-05,
"loss": 4.486,
"step": 33792
},
{
"epoch": 0.01,
"learning_rate": 4.943833806243993e-05,
"loss": 4.4748,
"step": 34304
},
{
"epoch": 0.01,
"learning_rate": 4.942995211492941e-05,
"loss": 4.4609,
"step": 34816
},
{
"epoch": 0.01,
"learning_rate": 4.942156616741889e-05,
"loss": 4.4629,
"step": 35328
},
{
"epoch": 0.01,
"learning_rate": 4.9413196598712105e-05,
"loss": 4.4775,
"step": 35840
},
{
"epoch": 0.01,
"learning_rate": 4.9404810651201585e-05,
"loss": 4.4766,
"step": 36352
},
{
"epoch": 0.01,
"learning_rate": 4.9396424703691065e-05,
"loss": 4.4508,
"step": 36864
},
{
"epoch": 0.01,
"learning_rate": 4.9388038756180545e-05,
"loss": 4.4684,
"step": 37376
},
{
"epoch": 0.01,
"learning_rate": 4.9379669187473754e-05,
"loss": 4.4616,
"step": 37888
},
{
"epoch": 0.01,
"learning_rate": 4.9371283239963234e-05,
"loss": 4.4479,
"step": 38400
},
{
"epoch": 0.01,
"learning_rate": 4.936291367125644e-05,
"loss": 4.4345,
"step": 38912
},
{
"epoch": 0.01,
"learning_rate": 4.935452772374592e-05,
"loss": 4.4393,
"step": 39424
},
{
"epoch": 0.01,
"learning_rate": 4.93461417762354e-05,
"loss": 4.4274,
"step": 39936
},
{
"epoch": 0.01,
"learning_rate": 4.933775582872488e-05,
"loss": 4.4218,
"step": 40448
},
{
"epoch": 0.01,
"learning_rate": 4.932936988121436e-05,
"loss": 4.4224,
"step": 40960
},
{
"epoch": 0.01,
"learning_rate": 4.932098393370384e-05,
"loss": 4.4184,
"step": 41472
},
{
"epoch": 0.01,
"learning_rate": 4.931259798619332e-05,
"loss": 4.4214,
"step": 41984
},
{
"epoch": 0.01,
"learning_rate": 4.930421203868281e-05,
"loss": 4.4173,
"step": 42496
},
{
"epoch": 0.01,
"learning_rate": 4.929584246997602e-05,
"loss": 4.3987,
"step": 43008
},
{
"epoch": 0.01,
"learning_rate": 4.92874565224655e-05,
"loss": 4.4037,
"step": 43520
},
{
"epoch": 0.01,
"learning_rate": 4.927907057495498e-05,
"loss": 4.4031,
"step": 44032
},
{
"epoch": 0.01,
"learning_rate": 4.927068462744446e-05,
"loss": 4.3945,
"step": 44544
},
{
"epoch": 0.01,
"learning_rate": 4.926231505873767e-05,
"loss": 4.3897,
"step": 45056
},
{
"epoch": 0.01,
"learning_rate": 4.925394549003088e-05,
"loss": 4.3871,
"step": 45568
},
{
"epoch": 0.02,
"learning_rate": 4.924555954252036e-05,
"loss": 4.3879,
"step": 46080
},
{
"epoch": 0.02,
"learning_rate": 4.923717359500984e-05,
"loss": 4.3767,
"step": 46592
},
{
"epoch": 0.02,
"learning_rate": 4.922878764749932e-05,
"loss": 4.377,
"step": 47104
},
{
"epoch": 0.02,
"learning_rate": 4.92204016999888e-05,
"loss": 4.3698,
"step": 47616
},
{
"epoch": 0.02,
"learning_rate": 4.921201575247828e-05,
"loss": 4.3742,
"step": 48128
},
{
"epoch": 0.02,
"learning_rate": 4.920362980496776e-05,
"loss": 4.3688,
"step": 48640
},
{
"epoch": 0.02,
"learning_rate": 4.919524385745724e-05,
"loss": 4.3636,
"step": 49152
},
{
"epoch": 0.02,
"learning_rate": 4.918687428875045e-05,
"loss": 4.3482,
"step": 49664
},
{
"epoch": 0.02,
"learning_rate": 4.917848834123993e-05,
"loss": 4.3523,
"step": 50176
},
{
"epoch": 0.02,
"learning_rate": 4.917010239372941e-05,
"loss": 4.3551,
"step": 50688
},
{
"epoch": 0.02,
"learning_rate": 4.916173282502262e-05,
"loss": 4.3619,
"step": 51200
},
{
"epoch": 0.02,
"learning_rate": 4.91533468775121e-05,
"loss": 4.347,
"step": 51712
},
{
"epoch": 0.02,
"learning_rate": 4.914497730880531e-05,
"loss": 4.3359,
"step": 52224
},
{
"epoch": 0.02,
"learning_rate": 4.913659136129479e-05,
"loss": 4.3323,
"step": 52736
},
{
"epoch": 0.02,
"learning_rate": 4.912820541378427e-05,
"loss": 4.3405,
"step": 53248
},
{
"epoch": 0.02,
"learning_rate": 4.911981946627375e-05,
"loss": 4.3108,
"step": 53760
},
{
"epoch": 0.02,
"learning_rate": 4.9111449897566966e-05,
"loss": 4.3209,
"step": 54272
},
{
"epoch": 0.02,
"learning_rate": 4.9103063950056446e-05,
"loss": 4.3224,
"step": 54784
},
{
"epoch": 0.02,
"learning_rate": 4.9094678002545926e-05,
"loss": 4.3213,
"step": 55296
},
{
"epoch": 0.02,
"learning_rate": 4.9086292055035406e-05,
"loss": 4.314,
"step": 55808
},
{
"epoch": 0.02,
"learning_rate": 4.9077906107524886e-05,
"loss": 4.3208,
"step": 56320
},
{
"epoch": 0.02,
"learning_rate": 4.9069520160014366e-05,
"loss": 4.301,
"step": 56832
},
{
"epoch": 0.02,
"learning_rate": 4.9061134212503846e-05,
"loss": 4.3273,
"step": 57344
},
{
"epoch": 0.02,
"learning_rate": 4.9052764643797055e-05,
"loss": 4.3071,
"step": 57856
},
{
"epoch": 0.02,
"learning_rate": 4.9044378696286535e-05,
"loss": 4.3105,
"step": 58368
},
{
"epoch": 0.02,
"learning_rate": 4.9035992748776015e-05,
"loss": 4.312,
"step": 58880
},
{
"epoch": 0.02,
"learning_rate": 4.9027606801265495e-05,
"loss": 4.3224,
"step": 59392
},
{
"epoch": 0.02,
"learning_rate": 4.9019237232558704e-05,
"loss": 4.3001,
"step": 59904
},
{
"epoch": 0.02,
"learning_rate": 4.9010851285048184e-05,
"loss": 4.2933,
"step": 60416
},
{
"epoch": 0.02,
"learning_rate": 4.900246533753767e-05,
"loss": 4.2986,
"step": 60928
},
{
"epoch": 0.02,
"learning_rate": 4.899407939002715e-05,
"loss": 4.2866,
"step": 61440
},
{
"epoch": 0.02,
"learning_rate": 4.8985693442516624e-05,
"loss": 4.2941,
"step": 61952
},
{
"epoch": 0.02,
"learning_rate": 4.8977307495006104e-05,
"loss": 4.2914,
"step": 62464
},
{
"epoch": 0.02,
"learning_rate": 4.8968921547495584e-05,
"loss": 4.2795,
"step": 62976
},
{
"epoch": 0.02,
"learning_rate": 4.89605519787888e-05,
"loss": 4.2982,
"step": 63488
},
{
"epoch": 0.02,
"learning_rate": 4.895216603127827e-05,
"loss": 4.279,
"step": 64000
},
{
"epoch": 0.02,
"learning_rate": 4.894378008376775e-05,
"loss": 4.2733,
"step": 64512
},
{
"epoch": 0.02,
"learning_rate": 4.893539413625723e-05,
"loss": 4.2736,
"step": 65024
},
{
"epoch": 0.02,
"learning_rate": 4.892702456755045e-05,
"loss": 4.2781,
"step": 65536
},
{
"epoch": 0.02,
"learning_rate": 4.891863862003992e-05,
"loss": 4.2633,
"step": 66048
},
{
"epoch": 0.02,
"learning_rate": 4.891025267252941e-05,
"loss": 4.2804,
"step": 66560
},
{
"epoch": 0.02,
"learning_rate": 4.890186672501889e-05,
"loss": 4.2735,
"step": 67072
},
{
"epoch": 0.02,
"learning_rate": 4.889348077750837e-05,
"loss": 4.2599,
"step": 67584
},
{
"epoch": 0.02,
"learning_rate": 4.888509482999785e-05,
"loss": 4.2521,
"step": 68096
},
{
"epoch": 0.02,
"learning_rate": 4.887670888248733e-05,
"loss": 4.2633,
"step": 68608
},
{
"epoch": 0.02,
"learning_rate": 4.886832293497681e-05,
"loss": 4.2586,
"step": 69120
},
{
"epoch": 0.02,
"learning_rate": 4.885995336627002e-05,
"loss": 4.2639,
"step": 69632
},
{
"epoch": 0.02,
"learning_rate": 4.8851583797563226e-05,
"loss": 4.2581,
"step": 70144
},
{
"epoch": 0.02,
"learning_rate": 4.8843197850052706e-05,
"loss": 4.2592,
"step": 70656
},
{
"epoch": 0.02,
"learning_rate": 4.8834811902542186e-05,
"loss": 4.2561,
"step": 71168
},
{
"epoch": 0.02,
"learning_rate": 4.8826425955031666e-05,
"loss": 4.2511,
"step": 71680
},
{
"epoch": 0.02,
"learning_rate": 4.8818040007521146e-05,
"loss": 4.2427,
"step": 72192
},
{
"epoch": 0.02,
"learning_rate": 4.8809654060010626e-05,
"loss": 4.2408,
"step": 72704
},
{
"epoch": 0.02,
"learning_rate": 4.8801268112500106e-05,
"loss": 4.2472,
"step": 73216
},
{
"epoch": 0.02,
"learning_rate": 4.879289854379332e-05,
"loss": 4.2262,
"step": 73728
},
{
"epoch": 0.02,
"learning_rate": 4.87845125962828e-05,
"loss": 4.2471,
"step": 74240
},
{
"epoch": 0.02,
"learning_rate": 4.877612664877228e-05,
"loss": 4.2419,
"step": 74752
},
{
"epoch": 0.02,
"learning_rate": 4.876774070126176e-05,
"loss": 4.2341,
"step": 75264
},
{
"epoch": 0.02,
"learning_rate": 4.875935475375124e-05,
"loss": 4.2354,
"step": 75776
},
{
"epoch": 0.02,
"learning_rate": 4.875098518504445e-05,
"loss": 4.236,
"step": 76288
},
{
"epoch": 0.03,
"eval_loss": 4.196593284606934,
"eval_runtime": 303.8622,
"eval_samples_per_second": 1255.803,
"eval_steps_per_second": 39.245,
"step": 76320
},
{
"epoch": 1.0,
"learning_rate": 4.874259923753393e-05,
"loss": 4.2105,
"step": 76800
},
{
"epoch": 1.0,
"learning_rate": 4.873421329002341e-05,
"loss": 4.2181,
"step": 77312
},
{
"epoch": 1.0,
"learning_rate": 4.872582734251289e-05,
"loss": 4.2297,
"step": 77824
},
{
"epoch": 1.0,
"learning_rate": 4.871744139500237e-05,
"loss": 4.2154,
"step": 78336
},
{
"epoch": 1.0,
"learning_rate": 4.870905544749185e-05,
"loss": 4.2198,
"step": 78848
},
{
"epoch": 1.0,
"learning_rate": 4.870066949998133e-05,
"loss": 4.2098,
"step": 79360
},
{
"epoch": 1.0,
"learning_rate": 4.869228355247081e-05,
"loss": 4.2076,
"step": 79872
},
{
"epoch": 1.0,
"learning_rate": 4.8683913983764027e-05,
"loss": 4.2033,
"step": 80384
},
{
"epoch": 1.0,
"learning_rate": 4.8675528036253507e-05,
"loss": 4.2113,
"step": 80896
},
{
"epoch": 1.0,
"learning_rate": 4.8667142088742986e-05,
"loss": 4.1989,
"step": 81408
},
{
"epoch": 1.0,
"learning_rate": 4.8658772520036196e-05,
"loss": 4.2193,
"step": 81920
},
{
"epoch": 1.0,
"learning_rate": 4.8650386572525676e-05,
"loss": 4.2135,
"step": 82432
},
{
"epoch": 1.0,
"learning_rate": 4.8642017003818885e-05,
"loss": 4.1948,
"step": 82944
},
{
"epoch": 1.0,
"learning_rate": 4.8633631056308365e-05,
"loss": 4.1907,
"step": 83456
},
{
"epoch": 1.0,
"learning_rate": 4.8625245108797845e-05,
"loss": 4.1882,
"step": 83968
},
{
"epoch": 1.0,
"learning_rate": 4.8616859161287324e-05,
"loss": 4.1889,
"step": 84480
},
{
"epoch": 1.0,
"learning_rate": 4.8608473213776804e-05,
"loss": 4.1836,
"step": 84992
},
{
"epoch": 1.0,
"learning_rate": 4.860008726626628e-05,
"loss": 4.1849,
"step": 85504
},
{
"epoch": 1.0,
"learning_rate": 4.8591701318755764e-05,
"loss": 4.1817,
"step": 86016
},
{
"epoch": 1.0,
"learning_rate": 4.8583315371245244e-05,
"loss": 4.1943,
"step": 86528
},
{
"epoch": 1.0,
"learning_rate": 4.8574929423734724e-05,
"loss": 4.1845,
"step": 87040
},
{
"epoch": 1.0,
"learning_rate": 4.8566543476224204e-05,
"loss": 4.1868,
"step": 87552
},
{
"epoch": 1.0,
"learning_rate": 4.8558157528713684e-05,
"loss": 4.1822,
"step": 88064
},
{
"epoch": 1.0,
"learning_rate": 4.854978796000689e-05,
"loss": 4.1915,
"step": 88576
},
{
"epoch": 1.0,
"learning_rate": 4.854140201249637e-05,
"loss": 4.1678,
"step": 89088
},
{
"epoch": 1.0,
"learning_rate": 4.853301606498585e-05,
"loss": 4.1723,
"step": 89600
},
{
"epoch": 1.0,
"learning_rate": 4.852463011747533e-05,
"loss": 4.1654,
"step": 90112
},
{
"epoch": 1.0,
"learning_rate": 4.851624416996481e-05,
"loss": 4.1651,
"step": 90624
},
{
"epoch": 1.0,
"learning_rate": 4.850785822245429e-05,
"loss": 4.1644,
"step": 91136
},
{
"epoch": 1.01,
"learning_rate": 4.849947227494377e-05,
"loss": 4.1618,
"step": 91648
},
{
"epoch": 1.01,
"learning_rate": 4.849110270623698e-05,
"loss": 4.1655,
"step": 92160
},
{
"epoch": 1.01,
"learning_rate": 4.848271675872646e-05,
"loss": 4.1649,
"step": 92672
},
{
"epoch": 1.01,
"learning_rate": 4.847433081121595e-05,
"loss": 4.1546,
"step": 93184
},
{
"epoch": 1.01,
"learning_rate": 4.846594486370543e-05,
"loss": 4.1626,
"step": 93696
},
{
"epoch": 1.01,
"learning_rate": 4.845755891619491e-05,
"loss": 4.1692,
"step": 94208
},
{
"epoch": 1.01,
"learning_rate": 4.844917296868439e-05,
"loss": 4.1573,
"step": 94720
},
{
"epoch": 1.01,
"learning_rate": 4.84408033999776e-05,
"loss": 4.1417,
"step": 95232
},
{
"epoch": 1.01,
"learning_rate": 4.843241745246708e-05,
"loss": 4.1519,
"step": 95744
},
{
"epoch": 1.01,
"learning_rate": 4.842403150495656e-05,
"loss": 4.1342,
"step": 96256
},
{
"epoch": 1.01,
"learning_rate": 4.841564555744604e-05,
"loss": 4.1433,
"step": 96768
},
{
"epoch": 1.01,
"learning_rate": 4.840727598873925e-05,
"loss": 4.1439,
"step": 97280
},
{
"epoch": 1.01,
"learning_rate": 4.839889004122873e-05,
"loss": 4.1507,
"step": 97792
},
{
"epoch": 1.01,
"learning_rate": 4.839050409371821e-05,
"loss": 4.1381,
"step": 98304
},
{
"epoch": 1.01,
"learning_rate": 4.8382118146207687e-05,
"loss": 4.1448,
"step": 98816
},
{
"epoch": 1.01,
"learning_rate": 4.8373732198697167e-05,
"loss": 4.1465,
"step": 99328
},
{
"epoch": 1.01,
"learning_rate": 4.8365346251186646e-05,
"loss": 4.1402,
"step": 99840
},
{
"epoch": 1.01,
"learning_rate": 4.835697668247986e-05,
"loss": 4.1409,
"step": 100352
},
{
"epoch": 1.01,
"learning_rate": 4.834859073496934e-05,
"loss": 4.1243,
"step": 100864
},
{
"epoch": 1.01,
"learning_rate": 4.834020478745882e-05,
"loss": 4.1308,
"step": 101376
},
{
"epoch": 1.01,
"learning_rate": 4.8331818839948295e-05,
"loss": 4.1387,
"step": 101888
},
{
"epoch": 1.01,
"learning_rate": 4.832344927124151e-05,
"loss": 4.1301,
"step": 102400
},
{
"epoch": 1.01,
"learning_rate": 4.831506332373099e-05,
"loss": 4.1244,
"step": 102912
},
{
"epoch": 1.01,
"learning_rate": 4.830667737622047e-05,
"loss": 4.1106,
"step": 103424
},
{
"epoch": 1.01,
"learning_rate": 4.8298291428709944e-05,
"loss": 4.1225,
"step": 103936
},
{
"epoch": 1.01,
"learning_rate": 4.8289905481199424e-05,
"loss": 4.1126,
"step": 104448
},
{
"epoch": 1.01,
"learning_rate": 4.828153591249264e-05,
"loss": 4.1275,
"step": 104960
},
{
"epoch": 1.01,
"learning_rate": 4.827314996498212e-05,
"loss": 4.1139,
"step": 105472
},
{
"epoch": 1.01,
"learning_rate": 4.82647640174716e-05,
"loss": 4.1225,
"step": 105984
},
{
"epoch": 1.01,
"learning_rate": 4.825637806996108e-05,
"loss": 4.1241,
"step": 106496
},
{
"epoch": 1.01,
"learning_rate": 4.824799212245056e-05,
"loss": 4.1007,
"step": 107008
},
{
"epoch": 1.01,
"learning_rate": 4.823960617494004e-05,
"loss": 4.1135,
"step": 107520
},
{
"epoch": 1.01,
"learning_rate": 4.823123660623325e-05,
"loss": 4.1113,
"step": 108032
},
{
"epoch": 1.01,
"learning_rate": 4.822285065872273e-05,
"loss": 4.1029,
"step": 108544
},
{
"epoch": 1.01,
"learning_rate": 4.821446471121221e-05,
"loss": 4.103,
"step": 109056
},
{
"epoch": 1.01,
"learning_rate": 4.820607876370169e-05,
"loss": 4.1189,
"step": 109568
},
{
"epoch": 1.01,
"learning_rate": 4.81977091949949e-05,
"loss": 4.1068,
"step": 110080
},
{
"epoch": 1.01,
"learning_rate": 4.818932324748438e-05,
"loss": 4.0955,
"step": 110592
},
{
"epoch": 1.01,
"learning_rate": 4.818093729997386e-05,
"loss": 4.0883,
"step": 111104
},
{
"epoch": 1.01,
"learning_rate": 4.817255135246334e-05,
"loss": 4.0963,
"step": 111616
},
{
"epoch": 1.01,
"learning_rate": 4.816416540495282e-05,
"loss": 4.1091,
"step": 112128
},
{
"epoch": 1.01,
"learning_rate": 4.8155795836246034e-05,
"loss": 4.1151,
"step": 112640
},
{
"epoch": 1.01,
"learning_rate": 4.8147409888735514e-05,
"loss": 4.0981,
"step": 113152
},
{
"epoch": 1.01,
"learning_rate": 4.8139023941224994e-05,
"loss": 4.1119,
"step": 113664
},
{
"epoch": 1.01,
"learning_rate": 4.8130637993714474e-05,
"loss": 4.1068,
"step": 114176
},
{
"epoch": 1.01,
"learning_rate": 4.812226842500768e-05,
"loss": 4.1002,
"step": 114688
},
{
"epoch": 1.01,
"learning_rate": 4.811388247749716e-05,
"loss": 4.0866,
"step": 115200
},
{
"epoch": 1.01,
"learning_rate": 4.810549652998664e-05,
"loss": 4.1014,
"step": 115712
},
{
"epoch": 1.01,
"learning_rate": 4.809711058247612e-05,
"loss": 4.0911,
"step": 116224
},
{
"epoch": 1.01,
"learning_rate": 4.80887246349656e-05,
"loss": 4.0864,
"step": 116736
},
{
"epoch": 1.01,
"learning_rate": 4.808035506625881e-05,
"loss": 4.0903,
"step": 117248
},
{
"epoch": 1.01,
"learning_rate": 4.807196911874829e-05,
"loss": 4.086,
"step": 117760
},
{
"epoch": 1.01,
"learning_rate": 4.806358317123777e-05,
"loss": 4.1012,
"step": 118272
},
{
"epoch": 1.01,
"learning_rate": 4.805519722372726e-05,
"loss": 4.0875,
"step": 118784
},
{
"epoch": 1.01,
"learning_rate": 4.804682765502047e-05,
"loss": 4.0809,
"step": 119296
},
{
"epoch": 1.01,
"learning_rate": 4.803844170750995e-05,
"loss": 4.0845,
"step": 119808
},
{
"epoch": 1.01,
"learning_rate": 4.803005575999943e-05,
"loss": 4.0873,
"step": 120320
},
{
"epoch": 1.01,
"learning_rate": 4.802166981248891e-05,
"loss": 4.0829,
"step": 120832
},
{
"epoch": 1.01,
"learning_rate": 4.8013300243782116e-05,
"loss": 4.0827,
"step": 121344
},
{
"epoch": 1.01,
"learning_rate": 4.8004914296271596e-05,
"loss": 4.0754,
"step": 121856
},
{
"epoch": 1.02,
"learning_rate": 4.7996528348761076e-05,
"loss": 4.0884,
"step": 122368
},
{
"epoch": 1.02,
"learning_rate": 4.7988142401250556e-05,
"loss": 4.0739,
"step": 122880
},
{
"epoch": 1.02,
"learning_rate": 4.7979756453740036e-05,
"loss": 4.0752,
"step": 123392
},
{
"epoch": 1.02,
"learning_rate": 4.7971386885033245e-05,
"loss": 4.0696,
"step": 123904
},
{
"epoch": 1.02,
"learning_rate": 4.7963000937522725e-05,
"loss": 4.0816,
"step": 124416
},
{
"epoch": 1.02,
"learning_rate": 4.795461499001221e-05,
"loss": 4.0798,
"step": 124928
},
{
"epoch": 1.02,
"learning_rate": 4.794622904250169e-05,
"loss": 4.0726,
"step": 125440
},
{
"epoch": 1.02,
"learning_rate": 4.79378594737949e-05,
"loss": 4.0614,
"step": 125952
},
{
"epoch": 1.02,
"learning_rate": 4.792947352628438e-05,
"loss": 4.0705,
"step": 126464
},
{
"epoch": 1.02,
"learning_rate": 4.792108757877386e-05,
"loss": 4.0709,
"step": 126976
},
{
"epoch": 1.02,
"learning_rate": 4.791270163126334e-05,
"loss": 4.08,
"step": 127488
},
{
"epoch": 1.02,
"learning_rate": 4.790433206255655e-05,
"loss": 4.071,
"step": 128000
},
{
"epoch": 1.02,
"learning_rate": 4.789594611504603e-05,
"loss": 4.0596,
"step": 128512
},
{
"epoch": 1.02,
"learning_rate": 4.788756016753551e-05,
"loss": 4.0555,
"step": 129024
},
{
"epoch": 1.02,
"learning_rate": 4.787917422002499e-05,
"loss": 4.07,
"step": 129536
},
{
"epoch": 1.02,
"learning_rate": 4.78708046513182e-05,
"loss": 4.0389,
"step": 130048
},
{
"epoch": 1.02,
"learning_rate": 4.786241870380768e-05,
"loss": 4.0563,
"step": 130560
},
{
"epoch": 1.02,
"learning_rate": 4.7854032756297166e-05,
"loss": 4.0582,
"step": 131072
},
{
"epoch": 1.02,
"learning_rate": 4.7845646808786646e-05,
"loss": 4.0573,
"step": 131584
},
{
"epoch": 1.02,
"learning_rate": 4.7837277240079855e-05,
"loss": 4.0474,
"step": 132096
},
{
"epoch": 1.02,
"learning_rate": 4.7828891292569335e-05,
"loss": 4.0569,
"step": 132608
},
{
"epoch": 1.02,
"learning_rate": 4.7820505345058815e-05,
"loss": 4.0486,
"step": 133120
},
{
"epoch": 1.02,
"learning_rate": 4.7812119397548295e-05,
"loss": 4.0676,
"step": 133632
},
{
"epoch": 1.02,
"learning_rate": 4.7803749828841504e-05,
"loss": 4.0525,
"step": 134144
},
{
"epoch": 1.02,
"learning_rate": 4.7795363881330984e-05,
"loss": 4.0522,
"step": 134656
},
{
"epoch": 1.02,
"learning_rate": 4.7786977933820464e-05,
"loss": 4.0669,
"step": 135168
},
{
"epoch": 1.02,
"learning_rate": 4.7778591986309944e-05,
"loss": 4.0723,
"step": 135680
},
{
"epoch": 1.02,
"learning_rate": 4.777022241760315e-05,
"loss": 4.0501,
"step": 136192
},
{
"epoch": 1.02,
"learning_rate": 4.776183647009263e-05,
"loss": 4.0466,
"step": 136704
},
{
"epoch": 1.02,
"learning_rate": 4.775345052258212e-05,
"loss": 4.0565,
"step": 137216
},
{
"epoch": 1.02,
"learning_rate": 4.77450645750716e-05,
"loss": 4.0435,
"step": 137728
},
{
"epoch": 1.02,
"learning_rate": 4.773667862756108e-05,
"loss": 4.048,
"step": 138240
},
{
"epoch": 1.02,
"learning_rate": 4.772829268005055e-05,
"loss": 4.0537,
"step": 138752
},
{
"epoch": 1.02,
"learning_rate": 4.771990673254003e-05,
"loss": 4.04,
"step": 139264
},
{
"epoch": 1.02,
"learning_rate": 4.771152078502951e-05,
"loss": 4.0594,
"step": 139776
},
{
"epoch": 1.02,
"learning_rate": 4.770315121632272e-05,
"loss": 4.0414,
"step": 140288
},
{
"epoch": 1.02,
"learning_rate": 4.76947652688122e-05,
"loss": 4.0441,
"step": 140800
},
{
"epoch": 1.02,
"learning_rate": 4.768637932130168e-05,
"loss": 4.0373,
"step": 141312
},
{
"epoch": 1.02,
"learning_rate": 4.767799337379116e-05,
"loss": 4.0441,
"step": 141824
},
{
"epoch": 1.02,
"learning_rate": 4.766962380508437e-05,
"loss": 4.0351,
"step": 142336
},
{
"epoch": 1.02,
"learning_rate": 4.766123785757386e-05,
"loss": 4.0494,
"step": 142848
},
{
"epoch": 1.02,
"learning_rate": 4.765285191006334e-05,
"loss": 4.0468,
"step": 143360
},
{
"epoch": 1.02,
"learning_rate": 4.764446596255282e-05,
"loss": 4.0372,
"step": 143872
},
{
"epoch": 1.02,
"learning_rate": 4.7636096393846026e-05,
"loss": 4.0204,
"step": 144384
},
{
"epoch": 1.02,
"learning_rate": 4.7627710446335506e-05,
"loss": 4.0444,
"step": 144896
},
{
"epoch": 1.02,
"learning_rate": 4.7619324498824986e-05,
"loss": 4.0351,
"step": 145408
},
{
"epoch": 1.02,
"learning_rate": 4.7610938551314466e-05,
"loss": 4.043,
"step": 145920
},
{
"epoch": 1.02,
"learning_rate": 4.7602568982607675e-05,
"loss": 4.0389,
"step": 146432
},
{
"epoch": 1.02,
"learning_rate": 4.7594183035097155e-05,
"loss": 4.0388,
"step": 146944
},
{
"epoch": 1.02,
"learning_rate": 4.7585797087586635e-05,
"loss": 4.0361,
"step": 147456
},
{
"epoch": 1.02,
"learning_rate": 4.7577411140076115e-05,
"loss": 4.0412,
"step": 147968
},
{
"epoch": 1.02,
"learning_rate": 4.7569041571369324e-05,
"loss": 4.0249,
"step": 148480
},
{
"epoch": 1.02,
"learning_rate": 4.756065562385881e-05,
"loss": 4.0255,
"step": 148992
},
{
"epoch": 1.02,
"learning_rate": 4.755226967634829e-05,
"loss": 4.0353,
"step": 149504
},
{
"epoch": 1.02,
"learning_rate": 4.754388372883777e-05,
"loss": 4.0188,
"step": 150016
},
{
"epoch": 1.02,
"learning_rate": 4.753549778132725e-05,
"loss": 4.0347,
"step": 150528
},
{
"epoch": 1.02,
"learning_rate": 4.752712821262046e-05,
"loss": 4.038,
"step": 151040
},
{
"epoch": 1.02,
"learning_rate": 4.751874226510994e-05,
"loss": 4.0235,
"step": 151552
},
{
"epoch": 1.02,
"learning_rate": 4.751035631759942e-05,
"loss": 4.0254,
"step": 152064
},
{
"epoch": 1.02,
"learning_rate": 4.75019703700889e-05,
"loss": 4.0278,
"step": 152576
},
{
"epoch": 1.03,
"eval_loss": 4.027783393859863,
"eval_runtime": 307.6267,
"eval_samples_per_second": 1240.435,
"eval_steps_per_second": 38.765,
"step": 152640
},
{
"epoch": 0.0,
"learning_rate": 4.749360080138211e-05,
"loss": 4.0083,
"step": 153088
},
{
"epoch": 0.0,
"learning_rate": 4.748521485387159e-05,
"loss": 4.0149,
"step": 153600
},
{
"epoch": 0.0,
"learning_rate": 4.747682890636107e-05,
"loss": 4.0269,
"step": 154112
},
{
"epoch": 0.0,
"learning_rate": 4.746844295885055e-05,
"loss": 4.0182,
"step": 154624
},
{
"epoch": 0.0,
"learning_rate": 4.746005701134003e-05,
"loss": 4.0257,
"step": 155136
},
{
"epoch": 0.0,
"learning_rate": 4.745167106382951e-05,
"loss": 4.0084,
"step": 155648
},
{
"epoch": 0.0,
"learning_rate": 4.744328511631899e-05,
"loss": 4.018,
"step": 156160
},
{
"epoch": 0.0,
"learning_rate": 4.7434899168808475e-05,
"loss": 4.0034,
"step": 156672
},
{
"epoch": 0.0,
"learning_rate": 4.7426513221297955e-05,
"loss": 4.0167,
"step": 157184
},
{
"epoch": 0.0,
"learning_rate": 4.7418127273787435e-05,
"loss": 4.0105,
"step": 157696
},
{
"epoch": 0.0,
"learning_rate": 4.740974132627691e-05,
"loss": 4.0209,
"step": 158208
},
{
"epoch": 0.0,
"learning_rate": 4.7401371757570124e-05,
"loss": 4.0232,
"step": 158720
},
{
"epoch": 0.0,
"learning_rate": 4.739300218886333e-05,
"loss": 4.004,
"step": 159232
},
{
"epoch": 0.0,
"learning_rate": 4.738461624135281e-05,
"loss": 4.0047,
"step": 159744
},
{
"epoch": 0.0,
"learning_rate": 4.737623029384229e-05,
"loss": 4.0038,
"step": 160256
},
{
"epoch": 0.0,
"learning_rate": 4.736784434633177e-05,
"loss": 4.0013,
"step": 160768
},
{
"epoch": 0.0,
"learning_rate": 4.735945839882125e-05,
"loss": 3.9968,
"step": 161280
},
{
"epoch": 0.0,
"learning_rate": 4.735107245131073e-05,
"loss": 4.002,
"step": 161792
},
{
"epoch": 0.0,
"learning_rate": 4.734268650380021e-05,
"loss": 3.9966,
"step": 162304
},
{
"epoch": 0.0,
"learning_rate": 4.733431693509343e-05,
"loss": 4.0205,
"step": 162816
},
{
"epoch": 0.0,
"learning_rate": 4.732593098758291e-05,
"loss": 3.9975,
"step": 163328
},
{
"epoch": 0.0,
"learning_rate": 4.731754504007238e-05,
"loss": 4.0121,
"step": 163840
},
{
"epoch": 0.0,
"learning_rate": 4.730915909256186e-05,
"loss": 4.0047,
"step": 164352
},
{
"epoch": 0.0,
"learning_rate": 4.730077314505134e-05,
"loss": 4.0125,
"step": 164864
},
{
"epoch": 0.0,
"learning_rate": 4.729240357634456e-05,
"loss": 3.9869,
"step": 165376
},
{
"epoch": 0.0,
"learning_rate": 4.728401762883403e-05,
"loss": 3.9983,
"step": 165888
},
{
"epoch": 0.0,
"learning_rate": 4.727563168132351e-05,
"loss": 3.9879,
"step": 166400
},
{
"epoch": 0.0,
"learning_rate": 4.726724573381299e-05,
"loss": 3.9961,
"step": 166912
},
{
"epoch": 0.0,
"learning_rate": 4.725887616510621e-05,
"loss": 3.9889,
"step": 167424
},
{
"epoch": 0.01,
"learning_rate": 4.725049021759568e-05,
"loss": 3.9893,
"step": 167936
},
{
"epoch": 0.01,
"learning_rate": 4.7242120648888896e-05,
"loss": 3.9987,
"step": 168448
},
{
"epoch": 0.01,
"learning_rate": 4.723373470137838e-05,
"loss": 3.995,
"step": 168960
},
{
"epoch": 0.01,
"learning_rate": 4.7225348753867856e-05,
"loss": 3.9836,
"step": 169472
},
{
"epoch": 0.01,
"learning_rate": 4.7216962806357336e-05,
"loss": 3.9972,
"step": 169984
},
{
"epoch": 0.01,
"learning_rate": 4.7208576858846816e-05,
"loss": 3.9942,
"step": 170496
},
{
"epoch": 0.01,
"learning_rate": 4.7200190911336296e-05,
"loss": 3.9912,
"step": 171008
},
{
"epoch": 0.01,
"learning_rate": 4.7191804963825775e-05,
"loss": 3.9792,
"step": 171520
},
{
"epoch": 0.01,
"learning_rate": 4.7183419016315255e-05,
"loss": 3.9879,
"step": 172032
},
{
"epoch": 0.01,
"learning_rate": 4.7175049447608465e-05,
"loss": 3.9694,
"step": 172544
},
{
"epoch": 0.01,
"learning_rate": 4.7166663500097945e-05,
"loss": 3.983,
"step": 173056
},
{
"epoch": 0.01,
"learning_rate": 4.7158277552587424e-05,
"loss": 3.979,
"step": 173568
},
{
"epoch": 0.01,
"learning_rate": 4.7149891605076904e-05,
"loss": 3.9889,
"step": 174080
},
{
"epoch": 0.01,
"learning_rate": 4.714152203637012e-05,
"loss": 3.9799,
"step": 174592
},
{
"epoch": 0.01,
"learning_rate": 4.71331360888596e-05,
"loss": 3.9833,
"step": 175104
},
{
"epoch": 0.01,
"learning_rate": 4.712475014134908e-05,
"loss": 3.9877,
"step": 175616
},
{
"epoch": 0.01,
"learning_rate": 4.711636419383856e-05,
"loss": 3.9834,
"step": 176128
},
{
"epoch": 0.01,
"learning_rate": 4.710799462513177e-05,
"loss": 3.9853,
"step": 176640
},
{
"epoch": 0.01,
"learning_rate": 4.709960867762125e-05,
"loss": 3.9664,
"step": 177152
},
{
"epoch": 0.01,
"learning_rate": 4.709122273011073e-05,
"loss": 3.9715,
"step": 177664
},
{
"epoch": 0.01,
"learning_rate": 4.708283678260021e-05,
"loss": 3.9838,
"step": 178176
},
{
"epoch": 0.01,
"learning_rate": 4.707446721389342e-05,
"loss": 3.9754,
"step": 178688
},
{
"epoch": 0.01,
"learning_rate": 4.70660812663829e-05,
"loss": 3.9695,
"step": 179200
},
{
"epoch": 0.01,
"learning_rate": 4.705769531887238e-05,
"loss": 3.9615,
"step": 179712
},
{
"epoch": 0.01,
"learning_rate": 4.704930937136186e-05,
"loss": 3.9709,
"step": 180224
},
{
"epoch": 0.01,
"learning_rate": 4.7040939802655074e-05,
"loss": 3.9584,
"step": 180736
},
{
"epoch": 0.01,
"learning_rate": 4.7032553855144554e-05,
"loss": 3.9758,
"step": 181248
},
{
"epoch": 0.01,
"learning_rate": 4.7024167907634034e-05,
"loss": 3.9638,
"step": 181760
},
{
"epoch": 0.01,
"learning_rate": 4.7015781960123514e-05,
"loss": 3.9768,
"step": 182272
},
{
"epoch": 0.01,
"learning_rate": 4.700741239141672e-05,
"loss": 3.9729,
"step": 182784
},
{
"epoch": 0.01,
"learning_rate": 4.69990264439062e-05,
"loss": 3.954,
"step": 183296
},
{
"epoch": 0.01,
"learning_rate": 4.699064049639568e-05,
"loss": 3.9636,
"step": 183808
},
{
"epoch": 0.01,
"learning_rate": 4.698225454888516e-05,
"loss": 3.9637,
"step": 184320
},
{
"epoch": 0.01,
"learning_rate": 4.697388498017837e-05,
"loss": 3.9554,
"step": 184832
},
{
"epoch": 0.01,
"learning_rate": 4.696549903266785e-05,
"loss": 3.9621,
"step": 185344
},
{
"epoch": 0.01,
"learning_rate": 4.695711308515733e-05,
"loss": 3.9682,
"step": 185856
},
{
"epoch": 0.01,
"learning_rate": 4.694872713764681e-05,
"loss": 3.9651,
"step": 186368
},
{
"epoch": 0.01,
"learning_rate": 4.694035756894003e-05,
"loss": 3.9543,
"step": 186880
},
{
"epoch": 0.01,
"learning_rate": 4.693197162142951e-05,
"loss": 3.9452,
"step": 187392
},
{
"epoch": 0.01,
"learning_rate": 4.692358567391899e-05,
"loss": 3.9518,
"step": 187904
},
{
"epoch": 0.01,
"learning_rate": 4.691519972640847e-05,
"loss": 3.965,
"step": 188416
},
{
"epoch": 0.01,
"learning_rate": 4.6906830157701677e-05,
"loss": 3.9735,
"step": 188928
},
{
"epoch": 0.01,
"learning_rate": 4.6898444210191157e-05,
"loss": 3.9562,
"step": 189440
},
{
"epoch": 0.01,
"learning_rate": 4.6890058262680636e-05,
"loss": 3.9698,
"step": 189952
},
{
"epoch": 0.01,
"learning_rate": 4.6881672315170116e-05,
"loss": 3.9671,
"step": 190464
},
{
"epoch": 0.01,
"learning_rate": 4.6873302746463326e-05,
"loss": 3.9644,
"step": 190976
},
{
"epoch": 0.01,
"learning_rate": 4.6864916798952806e-05,
"loss": 3.9412,
"step": 191488
},
{
"epoch": 0.01,
"learning_rate": 4.6856530851442285e-05,
"loss": 3.966,
"step": 192000
},
{
"epoch": 0.01,
"learning_rate": 4.6848144903931765e-05,
"loss": 3.9566,
"step": 192512
},
{
"epoch": 0.01,
"learning_rate": 4.683977533522498e-05,
"loss": 3.9482,
"step": 193024
},
{
"epoch": 0.01,
"learning_rate": 4.683138938771446e-05,
"loss": 3.9575,
"step": 193536
},
{
"epoch": 0.01,
"learning_rate": 4.682300344020394e-05,
"loss": 3.9516,
"step": 194048
},
{
"epoch": 0.01,
"learning_rate": 4.681461749269342e-05,
"loss": 3.963,
"step": 194560
},
{
"epoch": 0.01,
"learning_rate": 4.680624792398663e-05,
"loss": 3.9546,
"step": 195072
},
{
"epoch": 0.01,
"learning_rate": 4.679786197647611e-05,
"loss": 3.9449,
"step": 195584
},
{
"epoch": 0.01,
"learning_rate": 4.678947602896559e-05,
"loss": 3.9506,
"step": 196096
},
{
"epoch": 0.01,
"learning_rate": 4.67811064602588e-05,
"loss": 3.9517,
"step": 196608
},
{
"epoch": 0.01,
"learning_rate": 4.677272051274828e-05,
"loss": 3.9526,
"step": 197120
},
{
"epoch": 0.01,
"learning_rate": 4.676433456523776e-05,
"loss": 3.9477,
"step": 197632
},
{
"epoch": 0.01,
"learning_rate": 4.675594861772724e-05,
"loss": 3.9453,
"step": 198144
},
{
"epoch": 0.02,
"learning_rate": 4.674757904902045e-05,
"loss": 3.9536,
"step": 198656
},
{
"epoch": 0.02,
"learning_rate": 4.6739193101509935e-05,
"loss": 3.9423,
"step": 199168
},
{
"epoch": 0.02,
"learning_rate": 4.6730807153999415e-05,
"loss": 3.9461,
"step": 199680
},
{
"epoch": 0.02,
"learning_rate": 4.6722421206488895e-05,
"loss": 3.9402,
"step": 200192
},
{
"epoch": 0.02,
"learning_rate": 4.6714051637782104e-05,
"loss": 3.9551,
"step": 200704
},
{
"epoch": 0.02,
"learning_rate": 4.6705665690271584e-05,
"loss": 3.9501,
"step": 201216
},
{
"epoch": 0.02,
"learning_rate": 4.6697279742761064e-05,
"loss": 3.9437,
"step": 201728
},
{
"epoch": 0.02,
"learning_rate": 4.668891017405427e-05,
"loss": 3.9372,
"step": 202240
},
{
"epoch": 0.02,
"learning_rate": 4.668052422654375e-05,
"loss": 3.9472,
"step": 202752
},
{
"epoch": 0.02,
"learning_rate": 4.667213827903323e-05,
"loss": 3.9373,
"step": 203264
},
{
"epoch": 0.02,
"learning_rate": 4.666375233152271e-05,
"loss": 3.9566,
"step": 203776
},
{
"epoch": 0.02,
"learning_rate": 4.665538276281592e-05,
"loss": 3.943,
"step": 204288
},
{
"epoch": 0.02,
"learning_rate": 4.66469968153054e-05,
"loss": 3.9373,
"step": 204800
},
{
"epoch": 0.02,
"learning_rate": 4.663861086779488e-05,
"loss": 3.9306,
"step": 205312
},
{
"epoch": 0.02,
"learning_rate": 4.663022492028437e-05,
"loss": 3.9426,
"step": 205824
},
{
"epoch": 0.02,
"learning_rate": 4.662183897277385e-05,
"loss": 3.9207,
"step": 206336
},
{
"epoch": 0.02,
"learning_rate": 4.661345302526333e-05,
"loss": 3.9332,
"step": 206848
},
{
"epoch": 0.02,
"learning_rate": 4.660508345655654e-05,
"loss": 3.9365,
"step": 207360
},
{
"epoch": 0.02,
"learning_rate": 4.659669750904602e-05,
"loss": 3.9353,
"step": 207872
},
{
"epoch": 0.02,
"learning_rate": 4.65883115615355e-05,
"loss": 3.9223,
"step": 208384
},
{
"epoch": 0.02,
"learning_rate": 4.657992561402498e-05,
"loss": 3.9398,
"step": 208896
},
{
"epoch": 0.02,
"learning_rate": 4.657153966651446e-05,
"loss": 3.9228,
"step": 209408
},
{
"epoch": 0.02,
"learning_rate": 4.656315371900393e-05,
"loss": 3.9471,
"step": 209920
},
{
"epoch": 0.02,
"learning_rate": 4.6554784150297146e-05,
"loss": 3.9325,
"step": 210432
},
{
"epoch": 0.02,
"learning_rate": 4.6546398202786626e-05,
"loss": 3.9303,
"step": 210944
},
{
"epoch": 0.02,
"learning_rate": 4.6538012255276106e-05,
"loss": 3.947,
"step": 211456
},
{
"epoch": 0.02,
"learning_rate": 4.6529626307765586e-05,
"loss": 3.954,
"step": 211968
},
{
"epoch": 0.02,
"learning_rate": 4.65212567390588e-05,
"loss": 3.9324,
"step": 212480
},
{
"epoch": 0.02,
"learning_rate": 4.651287079154828e-05,
"loss": 3.9255,
"step": 212992
},
{
"epoch": 0.02,
"learning_rate": 4.650448484403776e-05,
"loss": 3.9367,
"step": 213504
},
{
"epoch": 0.02,
"learning_rate": 4.6496098896527235e-05,
"loss": 3.9278,
"step": 214016
},
{
"epoch": 0.02,
"learning_rate": 4.6487712949016715e-05,
"loss": 3.9283,
"step": 214528
},
{
"epoch": 0.02,
"learning_rate": 4.6479327001506195e-05,
"loss": 3.9415,
"step": 215040
},
{
"epoch": 0.02,
"learning_rate": 4.6470941053995675e-05,
"loss": 3.9189,
"step": 215552
},
{
"epoch": 0.02,
"learning_rate": 4.6462571485288884e-05,
"loss": 3.9416,
"step": 216064
},
{
"epoch": 0.02,
"learning_rate": 4.6454185537778364e-05,
"loss": 3.9289,
"step": 216576
},
{
"epoch": 0.02,
"learning_rate": 4.6445799590267844e-05,
"loss": 3.9288,
"step": 217088
},
{
"epoch": 0.02,
"learning_rate": 4.6437413642757324e-05,
"loss": 3.9205,
"step": 217600
},
{
"epoch": 0.02,
"learning_rate": 4.6429027695246804e-05,
"loss": 3.9303,
"step": 218112
},
{
"epoch": 0.02,
"learning_rate": 4.642064174773629e-05,
"loss": 3.9183,
"step": 218624
},
{
"epoch": 0.02,
"learning_rate": 4.641225580022577e-05,
"loss": 3.939,
"step": 219136
},
{
"epoch": 0.02,
"learning_rate": 4.640386985271525e-05,
"loss": 3.929,
"step": 219648
},
{
"epoch": 0.02,
"learning_rate": 4.639550028400846e-05,
"loss": 3.9255,
"step": 220160
},
{
"epoch": 0.02,
"learning_rate": 4.638711433649794e-05,
"loss": 3.9086,
"step": 220672
},
{
"epoch": 0.02,
"learning_rate": 4.637874476779115e-05,
"loss": 3.9314,
"step": 221184
},
{
"epoch": 0.02,
"learning_rate": 4.637035882028063e-05,
"loss": 3.9257,
"step": 221696
},
{
"epoch": 0.02,
"learning_rate": 4.636197287277011e-05,
"loss": 3.9276,
"step": 222208
},
{
"epoch": 0.02,
"learning_rate": 4.635358692525959e-05,
"loss": 3.9233,
"step": 222720
},
{
"epoch": 0.02,
"learning_rate": 4.634520097774907e-05,
"loss": 3.9315,
"step": 223232
},
{
"epoch": 0.02,
"learning_rate": 4.633681503023855e-05,
"loss": 3.9261,
"step": 223744
},
{
"epoch": 0.02,
"learning_rate": 4.632842908272803e-05,
"loss": 3.9328,
"step": 224256
},
{
"epoch": 0.02,
"learning_rate": 4.632004313521751e-05,
"loss": 3.9137,
"step": 224768
},
{
"epoch": 0.02,
"learning_rate": 4.6311673566510724e-05,
"loss": 3.9152,
"step": 225280
},
{
"epoch": 0.02,
"learning_rate": 4.6303287619000204e-05,
"loss": 3.9235,
"step": 225792
},
{
"epoch": 0.02,
"learning_rate": 4.6294901671489684e-05,
"loss": 3.9082,
"step": 226304
},
{
"epoch": 0.02,
"learning_rate": 4.6286532102782893e-05,
"loss": 3.9244,
"step": 226816
},
{
"epoch": 0.02,
"learning_rate": 4.6278146155272373e-05,
"loss": 3.9311,
"step": 227328
},
{
"epoch": 0.02,
"learning_rate": 4.626976020776185e-05,
"loss": 3.9165,
"step": 227840
},
{
"epoch": 0.02,
"learning_rate": 4.626137426025133e-05,
"loss": 3.9142,
"step": 228352
},
{
"epoch": 0.02,
"learning_rate": 4.625300469154454e-05,
"loss": 3.9233,
"step": 228864
},
{
"epoch": 0.03,
"eval_loss": 3.9528167247772217,
"eval_runtime": 303.7,
"eval_samples_per_second": 1256.473,
"eval_steps_per_second": 39.266,
"step": 228960
}
],
"logging_steps": 512,
"max_steps": 3052726,
"num_train_epochs": 9223372036854775807,
"save_steps": 10,
"total_flos": 1.6046213234352538e+17,
"trial_name": null,
"trial_params": null
}