You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.

93 lines
2.2 KiB

export CUDA_VISIBLE_DEVICES=0
# Yearly SMAPE: 13.366, MASE: 2.999, OWA: 0.786 Patience 5
python -u short_term_forecast.py \
--task_name "short_term_forecast" \
--dataset_name "m4Benchmark" \
--model "Model" \
--root_path "./datasets/m4/" \
--data_path "m4" \
--seasonal_patterns "Yearly" \
--batch_size 8 \
--lradj "type2" \
--learning_rate 0.0002 \
--forward_layers 3 \
--stride 4 \
--seed 2025 \
# Monthly SMAPE: 12.608, MASE: 0.925, OWA: 0.872
python -u short_term_forecast.py \
--task_name "short_term_forecast" \
--dataset_name "m4Benchmark" \
--model "Model" \
--root_path "./datasets/m4/" \
--data_path "m4" \
--seasonal_patterns "Monthly" \
--batch_size 32 \
--lradj "cosine" \
--learning_rate 0.0001 \
--forward_layers 3 \
--stride 1 \
--seed 2025 \
# Quarterly
python -u short_term_forecast.py \
--task_name "short_term_forecast" \
--dataset_name "m4Benchmark" \
--model "Model" \
--root_path "./datasets/m4/" \
--data_path "m4" \
--seasonal_patterns "Quarterly" \
--batch_size 32 \
--lradj "cosine" \
--learning_rate 0.0001 \
--forward_layers 3 \
--stride 1 \
--seed 2025 \
# Others SMAPE: 4.941 MASE: 3.327 OWA: 1.045
python -u short_term_forecast.py \
--task_name "short_term_forecast" \
--dataset_name "m4Benchmark" \
--model "Model" \
--root_path "./datasets/m4/" \
--data_path "m4" \
--seasonal_patterns "Daily" \
--batch_size 32 \
--lradj "cosine" \
--learning_rate 0.0002 \
--forward_layers 3 \
--stride 1 \
--seed 2025 \
python -u fine_tuning.py \
--task_name "short_term_forecast" \
--dataset_name "m4Benchmark" \
--model "Model" \
--root_path "./datasets/m4/" \
--data_path "m4" \
--seasonal_patterns "Weekly" \
--batch_size 32 \
--lradj "cosine" \
--learning_rate 0.0002 \
--forward_layers 3 \
--stride 1 \
--seed 2025 \
python -u short_term_forecast.py \
--task_name "short_term_forecast" \
--dataset_name "m4Benchmark" \
--model "Model" \
--root_path "./datasets/m4/" \
--data_path "m4" \
--seasonal_patterns "Hourly" \
--batch_size 32 \
--lradj "cosine" \
--learning_rate 0.0002 \
--forward_layers 3 \
--stride 1 \
--seed 2025 \