{
"best_metric": 0.7931726907630522,
"best_model_checkpoint": "models/german_xlm_xnli/checkpoint-49088",
"epoch": 5.0,
"global_step": 61360,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.04,
"learning_rate": 1.983702737940026e-05,
"loss": 1.0212,
"step": 500
},
{
"epoch": 0.08,
"learning_rate": 1.9674054758800523e-05,
"loss": 0.8564,
"step": 1000
},
{
"epoch": 0.12,
"learning_rate": 1.9511082138200782e-05,
"loss": 0.7866,
"step": 1500
},
{
"epoch": 0.16,
"learning_rate": 1.9348109517601044e-05,
"loss": 0.7406,
"step": 2000
},
{
"epoch": 0.2,
"learning_rate": 1.9185136897001307e-05,
"loss": 0.7337,
"step": 2500
},
{
"epoch": 0.24,
"learning_rate": 1.9022164276401566e-05,
"loss": 0.7218,
"step": 3000
},
{
"epoch": 0.29,
"learning_rate": 1.8859191655801828e-05,
"loss": 0.7036,
"step": 3500
},
{
"epoch": 0.33,
"learning_rate": 1.8696219035202087e-05,
"loss": 0.6906,
"step": 4000
},
{
"epoch": 0.37,
"learning_rate": 1.853324641460235e-05,
"loss": 0.6761,
"step": 4500
},
{
"epoch": 0.41,
"learning_rate": 1.837027379400261e-05,
"loss": 0.67,
"step": 5000
},
{
"epoch": 0.45,
"learning_rate": 1.820730117340287e-05,
"loss": 0.6608,
"step": 5500
},
{
"epoch": 0.49,
"learning_rate": 1.804432855280313e-05,
"loss": 0.6617,
"step": 6000
},
{
"epoch": 0.53,
"learning_rate": 1.788135593220339e-05,
"loss": 0.6531,
"step": 6500
},
{
"epoch": 0.57,
"learning_rate": 1.771838331160365e-05,
"loss": 0.6474,
"step": 7000
},
{
"epoch": 0.61,
"learning_rate": 1.7555410691003914e-05,
"loss": 0.6439,
"step": 7500
},
{
"epoch": 0.65,
"learning_rate": 1.7392438070404173e-05,
"loss": 0.6378,
"step": 8000
},
{
"epoch": 0.69,
"learning_rate": 1.7229465449804435e-05,
"loss": 0.6356,
"step": 8500
},
{
"epoch": 0.73,
"learning_rate": 1.7066492829204694e-05,
"loss": 0.626,
"step": 9000
},
{
"epoch": 0.77,
"learning_rate": 1.6903520208604957e-05,
"loss": 0.6297,
"step": 9500
},
{
"epoch": 0.81,
"learning_rate": 1.6740547588005215e-05,
"loss": 0.6189,
"step": 10000
},
{
"epoch": 0.86,
"learning_rate": 1.6577574967405478e-05,
"loss": 0.6198,
"step": 10500
},
{
"epoch": 0.9,
"learning_rate": 1.6414602346805737e-05,
"loss": 0.6216,
"step": 11000
},
{
"epoch": 0.94,
"learning_rate": 1.6251629726206e-05,
"loss": 0.6068,
"step": 11500
},
{
"epoch": 0.98,
"learning_rate": 1.608865710560626e-05,
"loss": 0.6159,
"step": 12000
},
{
"epoch": 1.0,
"eval_accuracy": 0.7726907630522089,
"eval_loss": 0.5801302790641785,
"eval_runtime": 6.0805,
"eval_samples_per_second": 409.504,
"eval_steps_per_second": 51.311,
"step": 12272
},
{
"epoch": 1.02,
"learning_rate": 1.592568448500652e-05,
"loss": 0.5841,
"step": 12500
},
{
"epoch": 1.06,
"learning_rate": 1.576271186440678e-05,
"loss": 0.549,
"step": 13000
},
{
"epoch": 1.1,
"learning_rate": 1.5599739243807042e-05,
"loss": 0.5519,
"step": 13500
},
{
"epoch": 1.14,
"learning_rate": 1.54367666232073e-05,
"loss": 0.5413,
"step": 14000
},
{
"epoch": 1.18,
"learning_rate": 1.5273794002607563e-05,
"loss": 0.5308,
"step": 14500
},
{
"epoch": 1.22,
"learning_rate": 1.5110821382007822e-05,
"loss": 0.5367,
"step": 15000
},
{
"epoch": 1.26,
"learning_rate": 1.4947848761408083e-05,
"loss": 0.5362,
"step": 15500
},
{
"epoch": 1.3,
"learning_rate": 1.4784876140808346e-05,
"loss": 0.5269,
"step": 16000
},
{
"epoch": 1.34,
"learning_rate": 1.4621903520208606e-05,
"loss": 0.539,
"step": 16500
},
{
"epoch": 1.39,
"learning_rate": 1.4458930899608867e-05,
"loss": 0.5434,
"step": 17000
},
{
"epoch": 1.43,
"learning_rate": 1.4295958279009128e-05,
"loss": 0.5269,
"step": 17500
},
{
"epoch": 1.47,
"learning_rate": 1.4132985658409388e-05,
"loss": 0.5381,
"step": 18000
},
{
"epoch": 1.51,
"learning_rate": 1.3970013037809649e-05,
"loss": 0.5343,
"step": 18500
},
{
"epoch": 1.55,
"learning_rate": 1.380704041720991e-05,
"loss": 0.5434,
"step": 19000
},
{
"epoch": 1.59,
"learning_rate": 1.364406779661017e-05,
"loss": 0.5389,
"step": 19500
},
{
"epoch": 1.63,
"learning_rate": 1.3481095176010431e-05,
"loss": 0.5372,
"step": 20000
},
{
"epoch": 1.67,
"learning_rate": 1.3318122555410693e-05,
"loss": 0.5339,
"step": 20500
},
{
"epoch": 1.71,
"learning_rate": 1.3155149934810954e-05,
"loss": 0.5342,
"step": 21000
},
{
"epoch": 1.75,
"learning_rate": 1.2992177314211213e-05,
"loss": 0.5327,
"step": 21500
},
{
"epoch": 1.79,
"learning_rate": 1.2829204693611474e-05,
"loss": 0.5301,
"step": 22000
},
{
"epoch": 1.83,
"learning_rate": 1.2666232073011735e-05,
"loss": 0.5316,
"step": 22500
},
{
"epoch": 1.87,
"learning_rate": 1.2503259452411995e-05,
"loss": 0.5402,
"step": 23000
},
{
"epoch": 1.91,
"learning_rate": 1.2340286831812256e-05,
"loss": 0.5196,
"step": 23500
},
{
"epoch": 1.96,
"learning_rate": 1.2177314211212517e-05,
"loss": 0.5329,
"step": 24000
},
{
"epoch": 2.0,
"learning_rate": 1.2014341590612777e-05,
"loss": 0.5257,
"step": 24500
},
{
"epoch": 2.0,
"eval_accuracy": 0.7895582329317269,
"eval_loss": 0.5412057638168335,
"eval_runtime": 6.3368,
"eval_samples_per_second": 392.945,
"eval_steps_per_second": 49.236,
"step": 24544
},
{
"epoch": 2.04,
"learning_rate": 1.185136897001304e-05,
"loss": 0.4563,
"step": 25000
},
{
"epoch": 2.08,
"learning_rate": 1.16883963494133e-05,
"loss": 0.4525,
"step": 25500
},
{
"epoch": 2.12,
"learning_rate": 1.1525423728813561e-05,
"loss": 0.4436,
"step": 26000
},
{
"epoch": 2.16,
"learning_rate": 1.1362451108213822e-05,
"loss": 0.4471,
"step": 26500
},
{
"epoch": 2.2,
"learning_rate": 1.1199478487614082e-05,
"loss": 0.4483,
"step": 27000
},
{
"epoch": 2.24,
"learning_rate": 1.1036505867014341e-05,
"loss": 0.4446,
"step": 27500
},
{
"epoch": 2.28,
"learning_rate": 1.0873533246414602e-05,
"loss": 0.4583,
"step": 28000
},
{
"epoch": 2.32,
"learning_rate": 1.0710560625814863e-05,
"loss": 0.4487,
"step": 28500
},
{
"epoch": 2.36,
"learning_rate": 1.0547588005215125e-05,
"loss": 0.4484,
"step": 29000
},
{
"epoch": 2.4,
"learning_rate": 1.0384615384615386e-05,
"loss": 0.4559,
"step": 29500
},
{
"epoch": 2.44,
"learning_rate": 1.0221642764015647e-05,
"loss": 0.4503,
"step": 30000
},
{
"epoch": 2.49,
"learning_rate": 1.0058670143415907e-05,
"loss": 0.4634,
"step": 30500
},
{
"epoch": 2.53,
"learning_rate": 9.895697522816168e-06,
"loss": 0.4516,
"step": 31000
},
{
"epoch": 2.57,
"learning_rate": 9.732724902216429e-06,
"loss": 0.4506,
"step": 31500
},
{
"epoch": 2.61,
"learning_rate": 9.56975228161669e-06,
"loss": 0.4487,
"step": 32000
},
{
"epoch": 2.65,
"learning_rate": 9.40677966101695e-06,
"loss": 0.4355,
"step": 32500
},
{
"epoch": 2.69,
"learning_rate": 9.24380704041721e-06,
"loss": 0.4523,
"step": 33000
},
{
"epoch": 2.73,
"learning_rate": 9.080834419817471e-06,
"loss": 0.4449,
"step": 33500
},
{
"epoch": 2.77,
"learning_rate": 8.917861799217732e-06,
"loss": 0.4453,
"step": 34000
},
{
"epoch": 2.81,
"learning_rate": 8.754889178617993e-06,
"loss": 0.4564,
"step": 34500
},
{
"epoch": 2.85,
"learning_rate": 8.591916558018254e-06,
"loss": 0.4545,
"step": 35000
},
{
"epoch": 2.89,
"learning_rate": 8.428943937418514e-06,
"loss": 0.4487,
"step": 35500
},
{
"epoch": 2.93,
"learning_rate": 8.265971316818775e-06,
"loss": 0.4521,
"step": 36000
},
{
"epoch": 2.97,
"learning_rate": 8.102998696219036e-06,
"loss": 0.4468,
"step": 36500
},
{
"epoch": 3.0,
"eval_accuracy": 0.7863453815261044,
"eval_loss": 0.5777254700660706,
"eval_runtime": 5.2465,
"eval_samples_per_second": 474.598,
"eval_steps_per_second": 59.468,
"step": 36816
},
{
"epoch": 3.01,
"learning_rate": 7.940026075619296e-06,
"loss": 0.4161,
"step": 37000
},
{
"epoch": 3.06,
"learning_rate": 7.777053455019557e-06,
"loss": 0.3732,
"step": 37500
},
{
"epoch": 3.1,
"learning_rate": 7.614080834419818e-06,
"loss": 0.3717,
"step": 38000
},
{
"epoch": 3.14,
"learning_rate": 7.451108213820078e-06,
"loss": 0.3946,
"step": 38500
},
{
"epoch": 3.18,
"learning_rate": 7.288135593220339e-06,
"loss": 0.3709,
"step": 39000
},
{
"epoch": 3.22,
"learning_rate": 7.125162972620601e-06,
"loss": 0.3795,
"step": 39500
},
{
"epoch": 3.26,
"learning_rate": 6.962190352020861e-06,
"loss": 0.3777,
"step": 40000
},
{
"epoch": 3.3,
"learning_rate": 6.799217731421122e-06,
"loss": 0.3741,
"step": 40500
},
{
"epoch": 3.34,
"learning_rate": 6.636245110821382e-06,
"loss": 0.3768,
"step": 41000
},
{
"epoch": 3.38,
"learning_rate": 6.473272490221643e-06,
"loss": 0.3818,
"step": 41500
},
{
"epoch": 3.42,
"learning_rate": 6.310299869621904e-06,
"loss": 0.3756,
"step": 42000
},
{
"epoch": 3.46,
"learning_rate": 6.147327249022165e-06,
"loss": 0.3834,
"step": 42500
},
{
"epoch": 3.5,
"learning_rate": 5.9843546284224255e-06,
"loss": 0.3795,
"step": 43000
},
{
"epoch": 3.54,
"learning_rate": 5.821382007822687e-06,
"loss": 0.3765,
"step": 43500
},
{
"epoch": 3.59,
"learning_rate": 5.658409387222948e-06,
"loss": 0.3851,
"step": 44000
},
{
"epoch": 3.63,
"learning_rate": 5.4954367666232076e-06,
"loss": 0.3787,
"step": 44500
},
{
"epoch": 3.67,
"learning_rate": 5.332464146023468e-06,
"loss": 0.3829,
"step": 45000
},
{
"epoch": 3.71,
"learning_rate": 5.169491525423729e-06,
"loss": 0.3774,
"step": 45500
},
{
"epoch": 3.75,
"learning_rate": 5.0065189048239905e-06,
"loss": 0.372,
"step": 46000
},
{
"epoch": 3.79,
"learning_rate": 4.843546284224251e-06,
"loss": 0.3754,
"step": 46500
},
{
"epoch": 3.83,
"learning_rate": 4.680573663624511e-06,
"loss": 0.3846,
"step": 47000
},
{
"epoch": 3.87,
"learning_rate": 4.5176010430247726e-06,
"loss": 0.365,
"step": 47500
},
{
"epoch": 3.91,
"learning_rate": 4.354628422425033e-06,
"loss": 0.3647,
"step": 48000
},
{
"epoch": 3.95,
"learning_rate": 4.191655801825294e-06,
"loss": 0.3779,
"step": 48500
},
{
"epoch": 3.99,
"learning_rate": 4.028683181225555e-06,
"loss": 0.3729,
"step": 49000
},
{
"epoch": 4.0,
"eval_accuracy": 0.7931726907630522,
"eval_loss": 0.6138126850128174,
"eval_runtime": 5.1705,
"eval_samples_per_second": 481.577,
"eval_steps_per_second": 60.342,
"step": 49088
},
{
"epoch": 4.03,
"learning_rate": 3.865710560625815e-06,
"loss": 0.3345,
"step": 49500
},
{
"epoch": 4.07,
"learning_rate": 3.702737940026076e-06,
"loss": 0.3222,
"step": 50000
},
{
"epoch": 4.12,
"learning_rate": 3.5397653194263363e-06,
"loss": 0.3269,
"step": 50500
},
{
"epoch": 4.16,
"learning_rate": 3.3767926988265974e-06,
"loss": 0.3154,
"step": 51000
},
{
"epoch": 4.2,
"learning_rate": 3.213820078226858e-06,
"loss": 0.3249,
"step": 51500
},
{
"epoch": 4.24,
"learning_rate": 3.0508474576271192e-06,
"loss": 0.3175,
"step": 52000
},
{
"epoch": 4.28,
"learning_rate": 2.8878748370273795e-06,
"loss": 0.3279,
"step": 52500
},
{
"epoch": 4.32,
"learning_rate": 2.7249022164276406e-06,
"loss": 0.313,
"step": 53000
},
{
"epoch": 4.36,
"learning_rate": 2.5619295958279013e-06,
"loss": 0.3219,
"step": 53500
},
{
"epoch": 4.4,
"learning_rate": 2.398956975228162e-06,
"loss": 0.3202,
"step": 54000
},
{
"epoch": 4.44,
"learning_rate": 2.2359843546284227e-06,
"loss": 0.3203,
"step": 54500
},
{
"epoch": 4.48,
"learning_rate": 2.0730117340286834e-06,
"loss": 0.3285,
"step": 55000
},
{
"epoch": 4.52,
"learning_rate": 1.910039113428944e-06,
"loss": 0.3179,
"step": 55500
},
{
"epoch": 4.56,
"learning_rate": 1.7470664928292048e-06,
"loss": 0.3195,
"step": 56000
},
{
"epoch": 4.6,
"learning_rate": 1.5840938722294655e-06,
"loss": 0.3281,
"step": 56500
},
{
"epoch": 4.64,
"learning_rate": 1.4211212516297262e-06,
"loss": 0.3242,
"step": 57000
},
{
"epoch": 4.69,
"learning_rate": 1.258148631029987e-06,
"loss": 0.3108,
"step": 57500
},
{
"epoch": 4.73,
"learning_rate": 1.0951760104302478e-06,
"loss": 0.315,
"step": 58000
},
{
"epoch": 4.77,
"learning_rate": 9.322033898305086e-07,
"loss": 0.3174,
"step": 58500
},
{
"epoch": 4.81,
"learning_rate": 7.692307692307694e-07,
"loss": 0.3138,
"step": 59000
},
{
"epoch": 4.85,
"learning_rate": 6.0625814863103e-07,
"loss": 0.3254,
"step": 59500
},
{
"epoch": 4.89,
"learning_rate": 4.432855280312908e-07,
"loss": 0.322,
"step": 60000
},
{
"epoch": 4.93,
"learning_rate": 2.803129074315515e-07,
"loss": 0.3147,
"step": 60500
},
{
"epoch": 4.97,
"learning_rate": 1.1734028683181226e-07,
"loss": 0.3267,
"step": 61000
},
{
"epoch": 5.0,
"eval_accuracy": 0.7847389558232932,
"eval_loss": 0.6564955115318298,
"eval_runtime": 5.1501,
"eval_samples_per_second": 483.486,
"eval_steps_per_second": 60.581,
"step": 61360
},
{
"epoch": 5.0,
"step": 61360,
"total_flos": 1.2915645671077632e+17,
"train_loss": 0.4736778818799433,
"train_runtime": 12286.2871,
"train_samples_per_second": 159.813,
"train_steps_per_second": 4.994
}
],
"max_steps": 61360,
"num_train_epochs": 5,
"total_flos": 1.2915645671077632e+17,
"trial_name": null,
"trial_params": null
}