train_all_pa.sh

#!/bin/bash

#!/bin/bash

# Source the conda configuration script to make the 'conda' command availabl

# Activate the conda environment
conda activate ~/miniconda3/envs/OpenDPD



# Arguments

# Global Settings
dataset_name=DPA_200MHz
accelerator=cpu
devices=0

# Hyperparameters
seed=0
n_epochs=100
frame_length=50
frame_stride=1
loss_type=l2
opt_type=adamw
batch_size=64
batch_size_eval=256
lr_schedule=1
lr=1e-3
lr_end=1e-6
decay_factor=0.5
patience=10

#########################
# Train
#########################
seed=(0 1 2 3 4)

# PA Model
PA_backbone=(gru vdlstm rvtdcnn gmp)
PA_hidden_size=(11 8 12 8)
PA_num_layers=(1 1 1 1)

for i_seed in "${seed[@]}"; do
    for ((i=0; i<${#PA_backbone[@]}; i++)); do
        # Train PA
        step=train_pa
        python main.py --dataset_name "$dataset_name" --seed "$i_seed" --step "$step"\
          --accelerator "$accelerator" --devices "$devices"\
          --PA_backbone "${PA_backbone[$i]}" --PA_hidden_size "${PA_hidden_size[$i]}" --PA_num_layers "${PA_num_layers[$i]}"\
          --frame_length "$frame_length" --frame_stride "$frame_stride" --loss_type "$loss_type" --opt_type "$opt_type"\
          --batch_size "$batch_size" --batch_size_eval "$batch_size_eval" --n_epochs "$n_epochs" --lr_schedule "$lr_schedule"\
          --lr "$lr" --lr_end "$lr_end" --decay_factor "$decay_factor" --patience "$patience" || exit 1;
    done
done