Skip to content

Instantly share code, notes, and snippets.

View deeprenewal usage
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]
@manujosephv
manujosephv / deeprenewal notebook usage
Created Oct 13, 2020
how to use deeprenewal programatically
View deeprenewal notebook usage
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,