manujosephv

dataset = get_dataset(args.datasource, regenerate=False)
prediction_length = dataset.metadata.prediction_length
freq = dataset.metadata.freq
cardinality = ast.literal_eval(dataset.metadata.feat_static_cat[0].cardinality)
train_ds = dataset.train
test_ds = dataset.test
trainer = Trainer(ctx=mx.context.gpu() if is_gpu&args.use_cuda else mx.context.cpu(), 
                  batch_size=args.batch_size,
                  learning_rate=args.learning_rate,
                  epochs=20,

manujosephv

usage: deeprenewal [-h] [--use-cuda USE_CUDA]
                   [--datasource {retail_dataset}]
                   [--regenerate-datasource REGENERATE_DATASOURCE]
                   [--model-save-dir MODEL_SAVE_DIR]
                   [--point-forecast {median,mean}]
                   [--calculate-spec CALCULATE_SPEC]
                   [--batch_size BATCH_SIZE]
                   [--learning-rate LEARNING_RATE]
                   [--max-epochs MAX_EPOCHS]
                   [--number-of-batches-per-epoch NUMBER_OF_BATCHES_PER_EPOCH]