# train model CUDA_VISIBLE_DEVICES=0 \ python -u synthesis.py \ --max_len=600 \ --transformer_step=160000 \ --vocoder_step=90000 \ --use_gpu=1 \ --checkpoint_path='./checkpoint' \ --log_dir='./log' \ --sample_path='./sample' \ --config_path='config/synthesis.yaml' \ if [ $? -ne 0 ]; then echo "Failed in training!" exit 1 fi exit 0