t5-small-coqa / trainer_state.json
rooftopcoder's picture
End of training
80ce11b
raw
history blame
5.47 kB
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 3.0,
"global_step": 20373,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.07,
"learning_rate": 4.882079900826349e-05,
"loss": 1.4754,
"step": 500
},
{
"epoch": 0.15,
"learning_rate": 4.539443723136697e-05,
"loss": 1.3547,
"step": 1000
},
{
"epoch": 0.22,
"learning_rate": 4.004414420573958e-05,
"loss": 1.3184,
"step": 1500
},
{
"epoch": 0.29,
"learning_rate": 3.3274645598733375e-05,
"loss": 1.3174,
"step": 2000
},
{
"epoch": 0.37,
"learning_rate": 2.5724549368023616e-05,
"loss": 1.2904,
"step": 2500
},
{
"epoch": 0.44,
"learning_rate": 1.8106101990647016e-05,
"loss": 1.2746,
"step": 3000
},
{
"epoch": 0.52,
"learning_rate": 1.1137997922835163e-05,
"loss": 1.2553,
"step": 3500
},
{
"epoch": 0.59,
"learning_rate": 5.477580782771013e-06,
"loss": 1.2449,
"step": 4000
},
{
"epoch": 0.66,
"learning_rate": 1.6588321308710408e-06,
"loss": 1.2664,
"step": 4500
},
{
"epoch": 0.74,
"learning_rate": 4.1997742936272785e-08,
"loss": 1.2583,
"step": 5000
},
{
"epoch": 0.81,
"learning_rate": 7.796034360648269e-07,
"loss": 1.2692,
"step": 5500
},
{
"epoch": 0.88,
"learning_rate": 3.802066381068883e-06,
"loss": 1.2549,
"step": 6000
},
{
"epoch": 0.96,
"learning_rate": 8.8242592737696e-06,
"loss": 1.2621,
"step": 6500
},
{
"epoch": 1.03,
"learning_rate": 1.5372408126985777e-05,
"loss": 1.2511,
"step": 7000
},
{
"epoch": 1.1,
"learning_rate": 2.282878625097739e-05,
"loss": 1.2055,
"step": 7500
},
{
"epoch": 1.18,
"learning_rate": 3.0489988167458534e-05,
"loss": 1.2219,
"step": 8000
},
{
"epoch": 1.25,
"learning_rate": 3.76332861246046e-05,
"loss": 1.2093,
"step": 8500
},
{
"epoch": 1.33,
"learning_rate": 4.3584809399588655e-05,
"loss": 1.2115,
"step": 9000
},
{
"epoch": 1.4,
"learning_rate": 4.778311462055443e-05,
"loss": 1.2122,
"step": 9500
},
{
"epoch": 1.47,
"learning_rate": 4.983215013308785e-05,
"loss": 1.1913,
"step": 10000
},
{
"epoch": 1.55,
"learning_rate": 4.953861796051032e-05,
"loss": 1.1973,
"step": 10500
},
{
"epoch": 1.62,
"learning_rate": 4.6930208777142614e-05,
"loss": 1.2112,
"step": 11000
},
{
"epoch": 1.69,
"learning_rate": 4.225298967865532e-05,
"loss": 1.1787,
"step": 11500
},
{
"epoch": 1.77,
"learning_rate": 3.59481911770088e-05,
"loss": 1.2036,
"step": 12000
},
{
"epoch": 1.84,
"learning_rate": 2.8610583243870337e-05,
"loss": 1.1818,
"step": 12500
},
{
"epoch": 1.91,
"learning_rate": 2.09323670433783e-05,
"loss": 1.1674,
"step": 13000
},
{
"epoch": 1.99,
"learning_rate": 1.363787538820378e-05,
"loss": 1.1715,
"step": 13500
},
{
"epoch": 2.06,
"learning_rate": 7.4152420218663585e-06,
"loss": 1.122,
"step": 14000
},
{
"epoch": 2.14,
"learning_rate": 2.8514857793098454e-06,
"loss": 1.0968,
"step": 14500
},
{
"epoch": 2.21,
"learning_rate": 3.771335315155861e-07,
"loss": 1.1392,
"step": 15000
},
{
"epoch": 2.28,
"learning_rate": 2.2560596844390935e-07,
"loss": 1.1046,
"step": 15500
},
{
"epoch": 2.36,
"learning_rate": 2.4111976063067728e-06,
"loss": 1.15,
"step": 16000
},
{
"epoch": 2.43,
"learning_rate": 6.727728298952271e-06,
"loss": 1.1218,
"step": 16500
},
{
"epoch": 2.5,
"learning_rate": 1.2767993464490152e-05,
"loss": 1.1315,
"step": 17000
},
{
"epoch": 2.58,
"learning_rate": 1.9962178169036043e-05,
"loss": 1.119,
"step": 17500
},
{
"epoch": 2.65,
"learning_rate": 2.7631611233523096e-05,
"loss": 1.1307,
"step": 18000
},
{
"epoch": 2.72,
"learning_rate": 3.505278841189536e-05,
"loss": 1.144,
"step": 18500
},
{
"epoch": 2.8,
"learning_rate": 4.152562494506565e-05,
"loss": 1.1165,
"step": 19000
},
{
"epoch": 2.87,
"learning_rate": 4.643949881229301e-05,
"loss": 1.1387,
"step": 19500
},
{
"epoch": 2.95,
"learning_rate": 4.933085441857718e-05,
"loss": 1.1241,
"step": 20000
},
{
"epoch": 3.0,
"step": 20373,
"total_flos": 5.063263464534835e+16,
"train_loss": 1.2059516395565477,
"train_runtime": 8228.1791,
"train_samples_per_second": 39.613,
"train_steps_per_second": 2.476
}
],
"max_steps": 20373,
"num_train_epochs": 3,
"total_flos": 5.063263464534835e+16,
"trial_name": null,
"trial_params": null
}