train.sh

#!/bin/bash

. ./path.sh || exit 1;
export CUDA_VISIBLE_DEVICES=0

########## Train BiLSTM oneshot VC model ##########
# python main.py --config ./conf/bilstm_ppg2mel_vctk_libri_oneshotvc.yaml \
#                --oneshotvc \
#                --bilstm
###################################################

########## Train Seq2seq oneshot VC model ###########
# python main.py --config ./conf/seq2seq_mol_ppg2mel_vctk_libri_oneshotvc.yaml \
#                --oneshotvc \
###################################################

########## Train Own oneshot VC model ##########
# python main.py --config ./conf/own.yaml \
#                --ownarc \
#                --bilstm

########## Train Transformer VC model ##########

# Config 1:
# Speaker Embed = atinput
# pitch = no
python main.py --config /mnt/data1/waris/repo/transformer-vc/conf/transformer_vc_ppg2mel_outspkdloss_inp_conct.yaml \
               --name=transformer-vc \
               --seed=2 \
               --transvcsplinpconc
#
# Status: Running
#

# Config 2:
# Speaker Embed = atinput
# pitch = yes
# python main.py --config /mnt/data1/waris/repo/transformer-vc/conf/transformer_vc_ppg2mel_outspkdloss_inp_conct_with_pitch.yaml \
#                --name=transformer-vc-config2 \
#                --seed=2 \
#                --transvcsplinpconc
#
# Status: Waiting to run
#

# Config 3:
# Speaker Embed = after_encoder
# pitch = yes
# python main.py --config /mnt/data1/waris/repo/transformer-vc/conf/transformer_vc_ppg2mel_outspkdloss_attn_conct_with_pitch.yaml \
#                --name=transformer-vc-config3 \
#                --seed=2 \
#                --transvcsplinpconc
#
# Status: Waiting to run
#