liangym
3 years ago
committed by
GitHub
commit
26f7644a48
@ -0,0 +1,3 @@
|
||||
process:
|
||||
# use raw audio
|
||||
- type: wav_process
|
||||
@ -0,0 +1,101 @@
|
||||
# Copyright (c) 2023 speechbrain Authors. All Rights Reserved.
|
||||
# Copyright (c) 2023 PaddlePaddle Authors. All Rights Reserved.
|
||||
#
|
||||
# Licensed under the Apache License, Version 2.0 (the "License");
|
||||
# you may not use this file except in compliance with the License.
|
||||
# You may obtain a copy of the License at
|
||||
#
|
||||
# http://www.apache.org/licenses/LICENSE-2.0
|
||||
#
|
||||
# Unless required by applicable law or agreed to in writing, software
|
||||
# distributed under the License is distributed on an "AS IS" BASIS,
|
||||
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
||||
# See the License for the specific language governing permissions and
|
||||
# limitations under the License.
|
||||
#
|
||||
# Modified from speechbrain 2023 (https://github.com/speechbrain/speechbrain/blob/develop/recipes/AISHELL-1/ASR/CTC/hparams/train_with_wav2vec.yaml)
|
||||
|
||||
# ############################################################################
|
||||
# Model: CTC-wav2vec2
|
||||
# Encoder: wav2vec2
|
||||
# Decoder: -
|
||||
# Tokens: Char
|
||||
# losses: CTC
|
||||
# Training: AISHELL-1
|
||||
# Authors: Yingzhi WANG 2022
|
||||
# ############################################################################
|
||||
|
||||
output_folder: !ref data
|
||||
cer_file: !ref <output_folder>/cer.txt
|
||||
save_folder: !ref <output_folder>/save
|
||||
train_log: !ref <output_folder>/train_log.txt
|
||||
|
||||
# Data files
|
||||
data_folder: data/aishell # e,g./path/to/aishell
|
||||
|
||||
skip_prep: False
|
||||
ckpt_interval_minutes: 15 # save checkpoint every N min
|
||||
train_data: !ref <output_folder>/train.csv
|
||||
valid_data: !ref <output_folder>/dev.csv
|
||||
test_data: !ref <output_folder>/test.csv
|
||||
|
||||
wav2vec2_hub: TencentGameMate/chinese-wav2vec2-large
|
||||
|
||||
# Training parameters
|
||||
number_of_epochs: 80
|
||||
lr: 1.0
|
||||
lr_wav2vec: 0.0001
|
||||
sorting: ascending
|
||||
auto_mix_prec: False
|
||||
sample_rate: 16000
|
||||
|
||||
# With data_parallel batch_size is split into N jobs
|
||||
# With DDP batch_size is multiplied by N jobs
|
||||
# Must be 8 per GPU to fit 32GB of VRAM
|
||||
batch_size: 5
|
||||
test_batch_size: 1 # need set to 1 when decoding
|
||||
|
||||
dynamic_batching: False
|
||||
dynamic_batch_sampler:
|
||||
feats_hop_size: 0.01
|
||||
max_batch_len: 15 # in terms of "duration" in annotations by default, second here
|
||||
left_bucket_len: 200 # old implementation attributs
|
||||
multiplier: 1.1 # old implementation attributs
|
||||
shuffle_ex: False # if true re-creates batches at each epoch shuffling examples.
|
||||
num_buckets: 10 # floor(log(max_batch_len/left_bucket_len, multiplier)) + 1
|
||||
batch_ordering: ascending
|
||||
|
||||
num_workers: 6
|
||||
|
||||
# Dataloader options
|
||||
train_dataloader_opts:
|
||||
batch_size: !ref <batch_size>
|
||||
num_workers: !ref <num_workers>
|
||||
valid_dataloader_opts:
|
||||
batch_size: !ref <test_batch_size>
|
||||
num_workers: !ref <num_workers>
|
||||
test_dataloader_opts:
|
||||
batch_size: !ref <test_batch_size>
|
||||
num_workers: !ref <num_workers>
|
||||
|
||||
wav2vec_output_dim: 1024
|
||||
dnn_neurons: 1024
|
||||
freeze_wav2vec: False
|
||||
dropout: 0.15
|
||||
|
||||
tokenizer: !apply:transformers.BertTokenizer.from_pretrained
|
||||
pretrained_model_name_or_path: bert-base-chinese
|
||||
# bert-base-chinese tokens length
|
||||
output_neurons: 21128
|
||||
|
||||
# Decoding parameters
|
||||
# Be sure that the bos and eos index match with the BPEs ones
|
||||
blank_index: 0
|
||||
|
||||
# AISHELL-1 has spaces between words in the transcripts,
|
||||
# which Chinese writing normally does not do.
|
||||
# If remove_spaces, spaces are removed
|
||||
# from the transcript before computing CER.
|
||||
# (e.g., 祝 可爱 的 你 —> 祝可爱的你)
|
||||
remove_spaces: True
|
||||
split_tokens: !apply:operator.not_ [!ref <remove_spaces>]
|
||||
@ -0,0 +1,4 @@
|
||||
decode_batch_size: 1
|
||||
error_rate_type: cer
|
||||
decoding_method: ctc_greedy_search # 'ctc_greedy_search', 'ctc_prefix_beam_search'
|
||||
beam_size: 10
|
||||
@ -0,0 +1,167 @@
|
||||
############################################
|
||||
# Network Architecture #
|
||||
############################################
|
||||
freeze_wav2vec2: False
|
||||
normalize_wav: True
|
||||
output_norm: True
|
||||
init_type: 'kaiming_uniform' # !Warning: need to convergence
|
||||
enc:
|
||||
input_shape: 1024
|
||||
dnn_blocks: 3
|
||||
dnn_neurons: 1024
|
||||
activation: True
|
||||
normalization: True
|
||||
dropout_rate: [0.15, 0.15, 0.0]
|
||||
ctc:
|
||||
enc_n_units: 1024
|
||||
blank_id: 0
|
||||
dropout_rate: 0.0
|
||||
|
||||
audio_augment:
|
||||
speeds: [90, 100, 110]
|
||||
|
||||
spec_augment:
|
||||
time_warp: True
|
||||
time_warp_window: 5
|
||||
time_warp_mode: bicubic
|
||||
freq_mask: True
|
||||
n_freq_mask: 2
|
||||
time_mask: True
|
||||
n_time_mask: 2
|
||||
replace_with_zero: False
|
||||
freq_mask_width: 30
|
||||
time_mask_width: 40
|
||||
wav2vec2_params_path: exp/wav2vec2/chinese-wav2vec2-large.pdparams
|
||||
|
||||
|
||||
############################################
|
||||
# Wav2Vec2.0 #
|
||||
############################################
|
||||
# vocab_size: 1000000
|
||||
hidden_size: 1024
|
||||
num_hidden_layers: 24
|
||||
num_attention_heads: 16
|
||||
intermediate_size: 4096
|
||||
hidden_act: gelu
|
||||
hidden_dropout: 0.1
|
||||
activation_dropout: 0.0
|
||||
attention_dropout: 0.1
|
||||
feat_proj_dropout: 0.1
|
||||
feat_quantizer_dropout: 0.0
|
||||
final_dropout: 0.0
|
||||
layerdrop: 0.1
|
||||
initializer_range: 0.02
|
||||
layer_norm_eps: 1e-5
|
||||
feat_extract_norm: layer
|
||||
feat_extract_activation: gelu
|
||||
conv_dim: [512, 512, 512, 512, 512, 512, 512]
|
||||
conv_stride: [5, 2, 2, 2, 2, 2, 2]
|
||||
conv_kernel: [10, 3, 3, 3, 3, 2, 2]
|
||||
conv_bias: True
|
||||
num_conv_pos_embeddings: 128
|
||||
num_conv_pos_embedding_groups: 16
|
||||
do_stable_layer_norm: True
|
||||
apply_spec_augment: False
|
||||
mask_channel_length: 10
|
||||
mask_channel_min_space: 1
|
||||
mask_channel_other: 0.0
|
||||
mask_channel_prob: 0.0
|
||||
mask_channel_selection: static
|
||||
mask_feature_length: 10
|
||||
mask_feature_min_masks: 0
|
||||
mask_feature_prob: 0.0
|
||||
mask_time_length: 10
|
||||
mask_time_min_masks: 2
|
||||
mask_time_min_space: 1
|
||||
mask_time_other: 0.0
|
||||
mask_time_prob: 0.075
|
||||
mask_time_selection: static
|
||||
num_codevectors_per_group: 320
|
||||
num_codevector_groups: 2
|
||||
contrastive_logits_temperature: 0.1
|
||||
num_negatives: 100
|
||||
codevector_dim: 256
|
||||
proj_codevector_dim: 256
|
||||
diversity_loss_weight: 0.1
|
||||
use_weighted_layer_sum: False
|
||||
# pad_token_id: 0
|
||||
# bos_token_id: 1
|
||||
# eos_token_id: 2
|
||||
add_adapter: False
|
||||
adapter_kernel_size: 3
|
||||
adapter_stride: 2
|
||||
num_adapter_layers: 3
|
||||
output_hidden_size: None
|
||||
|
||||
###########################################
|
||||
# Data #
|
||||
###########################################
|
||||
|
||||
train_manifest: data/manifest.train
|
||||
dev_manifest: data/manifest.dev
|
||||
test_manifest: data/manifest.test
|
||||
vocab_filepath: data/lang_char/vocab.txt
|
||||
|
||||
###########################################
|
||||
# Dataloader #
|
||||
###########################################
|
||||
|
||||
unit_type: 'char'
|
||||
mean_std_filepath:
|
||||
preprocess_config: conf/preprocess.yaml
|
||||
sortagrad: -1 # Feed samples from shortest to longest ; -1: enabled for all epochs, 0: disabled, other: enabled for 'other' epochs
|
||||
batch_size: 5 # Different batch_size may cause large differences in results
|
||||
maxlen_in: 51200000000 # if input length > maxlen-in batchsize is automatically reduced
|
||||
maxlen_out: 1500000 # if output length > maxlen-out batchsize is automatically reduced
|
||||
minibatches: 0 # for debug
|
||||
batch_count: auto
|
||||
batch_bins: 0
|
||||
batch_frames_in: 0
|
||||
batch_frames_out: 0
|
||||
batch_frames_inout: 0
|
||||
num_workers: 6
|
||||
subsampling_factor: 1
|
||||
num_encs: 1
|
||||
dist_sampler: True
|
||||
shortest_first: True
|
||||
return_lens_rate: True
|
||||
|
||||
###########################################
|
||||
# use speechbrain dataloader #
|
||||
###########################################
|
||||
use_sb_pipeline: True # whether use speechbrain pipeline. Default is True.
|
||||
sb_pipeline_conf: conf/train_with_wav2vec.yaml
|
||||
|
||||
###########################################
|
||||
# Training #
|
||||
###########################################
|
||||
n_epoch: 80
|
||||
accum_grad: 1
|
||||
global_grad_clip: 5.0
|
||||
|
||||
model_optim: adadelta
|
||||
model_optim_conf:
|
||||
lr: 1.0
|
||||
weight_decay: 0.0
|
||||
rho: 0.95
|
||||
epsilon: 1.0e-8
|
||||
|
||||
wav2vec2_optim: adam
|
||||
wav2vec2_optim_conf:
|
||||
lr: 0.0001
|
||||
weight_decay: 0.0
|
||||
|
||||
model_scheduler: newbobscheduler
|
||||
model_scheduler_conf:
|
||||
improvement_threshold: 0.0025
|
||||
annealing_factor: 0.8
|
||||
patient: 0
|
||||
wav2vec2_scheduler: newbobscheduler
|
||||
wav2vec2_scheduler_conf:
|
||||
improvement_threshold: 0.0025
|
||||
annealing_factor: 0.9
|
||||
patient: 0
|
||||
log_interval: 1
|
||||
checkpoint:
|
||||
kbest_n: 50
|
||||
latest_n: 5
|
||||
@ -0,0 +1,129 @@
|
||||
# Copyright (c) 2023 speechbrain Authors. All Rights Reserved.
|
||||
# Copyright (c) 2023 PaddlePaddle Authors. All Rights Reserved.
|
||||
#
|
||||
# Licensed under the Apache License, Version 2.0 (the "License");
|
||||
# you may not use this file except in compliance with the License.
|
||||
# You may obtain a copy of the License at
|
||||
#
|
||||
# http://www.apache.org/licenses/LICENSE-2.0
|
||||
#
|
||||
# Unless required by applicable law or agreed to in writing, software
|
||||
# distributed under the License is distributed on an "AS IS" BASIS,
|
||||
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
||||
# See the License for the specific language governing permissions and
|
||||
# limitations under the License.
|
||||
#
|
||||
# Modified from speechbrain 2023
|
||||
# (https://github.com/speechbrain/speechbrain/blob/develop/recipes/AISHELL-1/aishell_prepare.py)
|
||||
import argparse
|
||||
import csv
|
||||
import glob
|
||||
import logging
|
||||
import os
|
||||
|
||||
from paddlespeech.s2t.models.wav2vec2.io.dataio import read_audio
|
||||
|
||||
logger = logging.getLogger(__name__)
|
||||
|
||||
DATA_HOME = os.path.expanduser('~/.cache/paddle/dataset/speech')
|
||||
|
||||
parser = argparse.ArgumentParser(description=__doc__)
|
||||
parser.add_argument(
|
||||
"--data_folder",
|
||||
default=DATA_HOME + "/Aishell",
|
||||
type=str,
|
||||
help="Directory to save the dataset. (default: %(default)s)")
|
||||
parser.add_argument(
|
||||
"--save_folder",
|
||||
default="data/",
|
||||
type=str,
|
||||
help="Filepath prefix for output manifests. (default: %(default)s)")
|
||||
parser.add_argument(
|
||||
"--skip_prep",
|
||||
default=False,
|
||||
type=bool,
|
||||
help="If True, skip data preparation. (default: %(default)s)")
|
||||
args = parser.parse_args()
|
||||
|
||||
|
||||
def prepare_aishell(data_folder, save_folder, skip_prep=False):
|
||||
"""
|
||||
This function prepares the AISHELL-1 dataset.
|
||||
If the folder does not exist, the zip file will be extracted. If the zip file does not exist, it will be downloaded.
|
||||
data_folder : path to AISHELL-1 dataset.
|
||||
save_folder: path where to store the manifest csv files.
|
||||
skip_prep: If True, skip data preparation.
|
||||
"""
|
||||
if skip_prep:
|
||||
return
|
||||
|
||||
# Create filename-to-transcript dictionary
|
||||
filename2transcript = {}
|
||||
with open(
|
||||
os.path.join(data_folder,
|
||||
"data_aishell/transcript/aishell_transcript_v0.8.txt"),
|
||||
"r", ) as f:
|
||||
lines = f.readlines()
|
||||
for line in lines:
|
||||
key = line.split()[0]
|
||||
value = " ".join(line.split()[1:])
|
||||
filename2transcript[key] = value
|
||||
|
||||
splits = [
|
||||
"train",
|
||||
"dev",
|
||||
"test",
|
||||
]
|
||||
ID_start = 0 # needed to have a unique ID for each audio
|
||||
for split in splits:
|
||||
new_filename = os.path.join(save_folder, split) + ".csv"
|
||||
if os.path.exists(new_filename):
|
||||
continue
|
||||
logger.info("Preparing %s..." % new_filename)
|
||||
|
||||
csv_output = [["ID", "duration", "wav", "transcript"]]
|
||||
entry = []
|
||||
|
||||
all_wavs = glob.glob(
|
||||
os.path.join(data_folder, "data_aishell/wav") + "/" + split +
|
||||
"/*/*.wav")
|
||||
for i in range(len(all_wavs)):
|
||||
filename = all_wavs[i].split("/")[-1].split(".wav")[0]
|
||||
if filename not in filename2transcript:
|
||||
continue
|
||||
signal = read_audio(all_wavs[i])
|
||||
duration = signal.shape[0] / 16000
|
||||
transcript_ = filename2transcript[filename]
|
||||
csv_line = [
|
||||
ID_start + i,
|
||||
str(duration),
|
||||
all_wavs[i],
|
||||
transcript_,
|
||||
]
|
||||
entry.append(csv_line)
|
||||
|
||||
csv_output = csv_output + entry
|
||||
|
||||
with open(new_filename, mode="w") as csv_f:
|
||||
csv_writer = csv.writer(
|
||||
csv_f, delimiter=",", quotechar='"', quoting=csv.QUOTE_MINIMAL)
|
||||
for line in csv_output:
|
||||
csv_writer.writerow(line)
|
||||
|
||||
msg = "\t%s successfully created!" % (new_filename)
|
||||
logger.info(msg)
|
||||
|
||||
ID_start += len(all_wavs)
|
||||
|
||||
|
||||
def main():
|
||||
if args.data_folder.startswith('~'):
|
||||
args.data_folder = os.path.expanduser(args.data_folder)
|
||||
|
||||
prepare_aishell(args.data_folder, args.save_folder, skip_prep=False)
|
||||
|
||||
print("Data csv prepare done!")
|
||||
|
||||
|
||||
if __name__ == '__main__':
|
||||
main()
|
||||
@ -0,0 +1,101 @@
|
||||
#!/bin/bash
|
||||
|
||||
stage=-1
|
||||
stop_stage=-1
|
||||
dict_dir=data/lang_char
|
||||
|
||||
. ${MAIN_ROOT}/utils/parse_options.sh || exit -1;
|
||||
|
||||
mkdir -p data
|
||||
mkdir -p ${dict_dir}
|
||||
TARGET_DIR=${MAIN_ROOT}/dataset
|
||||
mkdir -p ${TARGET_DIR}
|
||||
|
||||
if [ ${stage} -le -1 ] && [ ${stop_stage} -ge -1 ]; then
|
||||
# download data, generate manifests
|
||||
python3 ${TARGET_DIR}/aishell/aishell.py \
|
||||
--manifest_prefix="data/manifest" \
|
||||
--target_dir="${TARGET_DIR}/aishell"
|
||||
|
||||
#generate csv file for speechbrain dataloader
|
||||
python3 local/aishell_prepare.py \
|
||||
--data_folder="${TARGET_DIR}/aishell" \
|
||||
--save_folder="data/"
|
||||
|
||||
|
||||
if [ $? -ne 0 ]; then
|
||||
echo "Prepare Aishell failed. Terminated."
|
||||
exit 1
|
||||
fi
|
||||
|
||||
for dataset in train dev test; do
|
||||
mv data/manifest.${dataset} data/manifest.${dataset}.raw
|
||||
done
|
||||
fi
|
||||
|
||||
if [ ${stage} -le 0 ] && [ ${stop_stage} -ge 0 ]; then
|
||||
# compute mean and stddev for normalizer
|
||||
num_workers=$(nproc)
|
||||
python3 ${MAIN_ROOT}/utils/compute_mean_std.py \
|
||||
--manifest_path="data/manifest.train.raw" \
|
||||
--spectrum_type="fbank" \
|
||||
--feat_dim=80 \
|
||||
--delta_delta=false \
|
||||
--stride_ms=10 \
|
||||
--window_ms=25 \
|
||||
--sample_rate=16000 \
|
||||
--use_dB_normalization=False \
|
||||
--num_samples=-1 \
|
||||
--num_workers=${num_workers} \
|
||||
--output_path="data/mean_std.json"
|
||||
|
||||
if [ $? -ne 0 ]; then
|
||||
echo "Compute mean and stddev failed. Terminated."
|
||||
exit 1
|
||||
fi
|
||||
fi
|
||||
|
||||
if [ ${stage} -le 1 ] && [ ${stop_stage} -ge 1 ]; then
|
||||
# download data, generate manifests
|
||||
# build vocabulary
|
||||
python3 ${MAIN_ROOT}/utils/build_vocab.py \
|
||||
--unit_type="char" \
|
||||
--count_threshold=0 \
|
||||
--vocab_path="${dict_dir}/vocab.txt" \
|
||||
--manifest_paths "data/manifest.train.raw"
|
||||
|
||||
if [ $? -ne 0 ]; then
|
||||
echo "Build vocabulary failed. Terminated."
|
||||
exit 1
|
||||
fi
|
||||
fi
|
||||
|
||||
if [ ${stage} -le 2 ] && [ ${stop_stage} -ge 2 ]; then
|
||||
# format manifest with tokenids, vocab size
|
||||
for dataset in train dev test; do
|
||||
{
|
||||
python3 ${MAIN_ROOT}/utils/format_data.py \
|
||||
--cmvn_path "data/mean_std.json" \
|
||||
--unit_type "char" \
|
||||
--vocab_path="${dict_dir}/vocab.txt" \
|
||||
--manifest_path="data/manifest.${dataset}.raw" \
|
||||
--output_path="data/manifest.${dataset}"
|
||||
|
||||
if [ $? -ne 0 ]; then
|
||||
echo "Formt mnaifest failed. Terminated."
|
||||
exit 1
|
||||
fi
|
||||
} &
|
||||
done
|
||||
wait
|
||||
fi
|
||||
echo "Aishell data preparation done."
|
||||
|
||||
if [ ${stage} -le 3 ] && [ ${stop_stage} -ge 3 ]; then
|
||||
mkdir -p exp/wav2vec2
|
||||
echo "Pretrained wav2vec2 model download"
|
||||
wget -P exp/wav2vec2 https://paddlespeech.bj.bcebos.com/wav2vec/chinese-wav2vec2-large.pdparams
|
||||
fi
|
||||
|
||||
exit 0
|
||||
|
||||
@ -0,0 +1,84 @@
|
||||
#!/bin/bash
|
||||
|
||||
set -e
|
||||
|
||||
ngpu=$(echo $CUDA_VISIBLE_DEVICES | awk -F "," '{print NF}')
|
||||
echo "using $ngpu gpus..."
|
||||
|
||||
expdir=exp
|
||||
datadir=data
|
||||
|
||||
train_set=train_960
|
||||
recog_set="test-clean test-other dev-clean dev-other"
|
||||
recog_set="test-clean"
|
||||
|
||||
config_path=$1
|
||||
decode_config_path=$2
|
||||
ckpt_prefix=$3
|
||||
|
||||
source ${MAIN_ROOT}/utils/parse_options.sh || exit 1;
|
||||
|
||||
# download language model
|
||||
#bash local/download_lm_en.sh
|
||||
#if [ $? -ne 0 ]; then
|
||||
# exit 1
|
||||
#fi
|
||||
|
||||
python3 utils/format_rsl.py \
|
||||
--origin_ref data/manifest.test.raw \
|
||||
--trans_ref data/manifest.test.text
|
||||
|
||||
|
||||
for type in ctc_greedy_search; do
|
||||
echo "decoding ${type}"
|
||||
batch_size=1
|
||||
python3 -u ${BIN_DIR}/test.py \
|
||||
--ngpu ${ngpu} \
|
||||
--config ${config_path} \
|
||||
--decode_cfg ${decode_config_path} \
|
||||
--result_file ${ckpt_prefix}.${type}.rsl \
|
||||
--checkpoint_path ${ckpt_prefix} \
|
||||
--opts decode.decoding_method ${type} \
|
||||
--opts decode.decode_batch_size ${batch_size}
|
||||
|
||||
if [ $? -ne 0 ]; then
|
||||
echo "Failed in evaluation!"
|
||||
exit 1
|
||||
fi
|
||||
python3 utils/format_rsl.py \
|
||||
--origin_hyp ${ckpt_prefix}.${type}.rsl \
|
||||
--trans_hyp ${ckpt_prefix}.${type}.rsl.text
|
||||
|
||||
python3 utils/compute-wer.py --char=1 --v=1 \
|
||||
data/manifest.test.text ${ckpt_prefix}.${type}.rsl.text > ${ckpt_prefix}.${type}.error
|
||||
echo "decoding ${type} done."
|
||||
done
|
||||
|
||||
for type in ctc_prefix_beam_search; do
|
||||
echo "decoding ${type}"
|
||||
batch_size=1
|
||||
python3 -u ${BIN_DIR}/test.py \
|
||||
--ngpu ${ngpu} \
|
||||
--config ${config_path} \
|
||||
--decode_cfg ${decode_config_path} \
|
||||
--result_file ${ckpt_prefix}.${type}.rsl \
|
||||
--checkpoint_path ${ckpt_prefix} \
|
||||
--opts decode.decoding_method ${type} \
|
||||
--opts decode.decode_batch_size ${batch_size}
|
||||
|
||||
if [ $? -ne 0 ]; then
|
||||
echo "Failed in evaluation!"
|
||||
exit 1
|
||||
fi
|
||||
python3 utils/format_rsl.py \
|
||||
--origin_hyp ${ckpt_prefix}.${type}.rsl \
|
||||
--trans_hyp ${ckpt_prefix}.${type}.rsl.text
|
||||
|
||||
python3 utils/compute-wer.py --char=1 --v=1 \
|
||||
data/manifest.test-clean.text ${ckpt_prefix}.${type}.rsl.text > ${ckpt_prefix}.${type}.error
|
||||
echo "decoding ${type} done."
|
||||
done
|
||||
|
||||
echo "Finished"
|
||||
|
||||
exit 0
|
||||
@ -0,0 +1,58 @@
|
||||
#!/bin/bash
|
||||
|
||||
if [ $# != 4 ];then
|
||||
echo "usage: ${0} config_path decode_config_path ckpt_path_prefix audio_file"
|
||||
exit -1
|
||||
fi
|
||||
|
||||
ngpu=$(echo $CUDA_VISIBLE_DEVICES | awk -F "," '{print NF}')
|
||||
echo "using $ngpu gpus..."
|
||||
|
||||
config_path=$1
|
||||
decode_config_path=$2
|
||||
ckpt_prefix=$3
|
||||
audio_file=$4
|
||||
|
||||
mkdir -p data
|
||||
wget -nc https://paddlespeech.bj.bcebos.com/datasets/single_wav/en/demo_002_en.wav -P data/
|
||||
if [ $? -ne 0 ]; then
|
||||
exit 1
|
||||
fi
|
||||
|
||||
if [ ! -f ${audio_file} ]; then
|
||||
echo "Plase input the right audio_file path"
|
||||
exit 1
|
||||
fi
|
||||
|
||||
chunk_mode=false
|
||||
if [[ ${config_path} =~ ^.*chunk_.*yaml$ ]];then
|
||||
chunk_mode=true
|
||||
fi
|
||||
|
||||
# download language model
|
||||
#bash local/download_lm_ch.sh
|
||||
#if [ $? -ne 0 ]; then
|
||||
# exit 1
|
||||
#fi
|
||||
|
||||
for type in ctc_greedy_search; do
|
||||
echo "decoding ${type}"
|
||||
batch_size=1
|
||||
output_dir=${ckpt_prefix}
|
||||
mkdir -p ${output_dir}
|
||||
python3 -u ${BIN_DIR}/test_wav.py \
|
||||
--ngpu ${ngpu} \
|
||||
--config ${config_path} \
|
||||
--decode_cfg ${decode_config_path} \
|
||||
--result_file ${output_dir}/${type}.rsl \
|
||||
--checkpoint_path ${ckpt_prefix} \
|
||||
--opts decode.decoding_method ${type} \
|
||||
--opts decode.decode_batch_size ${batch_size} \
|
||||
--audio_file ${audio_file}
|
||||
|
||||
if [ $? -ne 0 ]; then
|
||||
echo "Failed in evaluation!"
|
||||
exit 1
|
||||
fi
|
||||
done
|
||||
exit 0
|
||||
@ -0,0 +1,59 @@
|
||||
#!/bin/bash
|
||||
|
||||
if [ $# -lt 2 ] && [ $# -gt 3 ];then
|
||||
echo "usage: CUDA_VISIBLE_DEVICES=0 ${0} config_path ckpt_name ips(optional)"
|
||||
exit -1
|
||||
fi
|
||||
|
||||
ngpu=$(echo $CUDA_VISIBLE_DEVICES | awk -F "," '{print NF}')
|
||||
echo "using $ngpu gpus..."
|
||||
|
||||
config_path=$1
|
||||
ckpt_name=$2
|
||||
resume=$3
|
||||
ips=$4
|
||||
|
||||
if [ ! $ips ];then
|
||||
ips_config=
|
||||
else
|
||||
ips_config="--ips="${ips}
|
||||
fi
|
||||
|
||||
mkdir -p exp
|
||||
|
||||
# seed may break model convergence
|
||||
seed=2
|
||||
if [ ${seed} != 0 ]; then
|
||||
export FLAGS_cudnn_deterministic=True
|
||||
fi
|
||||
|
||||
# export FLAGS_cudnn_exhaustive_search=true
|
||||
# export FLAGS_conv_workspace_size_limit=4000
|
||||
# export FLAGS_allocator_strategy=naive_best_fit
|
||||
|
||||
if [ ${ngpu} == 0 ]; then
|
||||
python3 -u ${BIN_DIR}/train.py \
|
||||
--ngpu ${ngpu} \
|
||||
--config ${config_path} \
|
||||
--output exp/${ckpt_name} \
|
||||
--seed ${seed} \
|
||||
--resume ${resume}
|
||||
else
|
||||
python3 -m paddle.distributed.launch --log_dir=${ckpt_name} --gpus=${CUDA_VISIBLE_DEVICES} ${ips_config} ${BIN_DIR}/train.py \
|
||||
--ngpu ${ngpu} \
|
||||
--config ${config_path} \
|
||||
--output exp/${ckpt_name} \
|
||||
--seed ${seed} \
|
||||
--resume ${resume}
|
||||
fi
|
||||
|
||||
if [ ${seed} != 0 ]; then
|
||||
unset FLAGS_cudnn_deterministic
|
||||
fi
|
||||
|
||||
if [ $? -ne 0 ]; then
|
||||
echo "Failed in training!"
|
||||
exit 1
|
||||
fi
|
||||
|
||||
exit 0
|
||||
@ -0,0 +1,15 @@
|
||||
export MAIN_ROOT=`realpath ${PWD}/../../../`
|
||||
|
||||
export PATH=${MAIN_ROOT}:${MAIN_ROOT}/tools/sctk/bin:${PWD}/utils:${PATH}
|
||||
export LC_ALL=C
|
||||
|
||||
export PYTHONDONTWRITEBYTECODE=1
|
||||
# Use UTF-8 in Python to avoid UnicodeDecodeError when LC_ALL=C
|
||||
export PYTHONIOENCODING=UTF-8
|
||||
export PYTHONPATH=${MAIN_ROOT}:${PYTHONPATH}
|
||||
|
||||
export LD_LIBRARY_PATH=${LD_LIBRARY_PATH}:/usr/local/lib/
|
||||
|
||||
|
||||
MODEL=wav2vec2
|
||||
export BIN_DIR=${MAIN_ROOT}/paddlespeech/s2t/exps/${MODEL}/bin
|
||||
@ -0,0 +1,48 @@
|
||||
#!/bin/bash
|
||||
set -e
|
||||
|
||||
. ./path.sh || exit 1;
|
||||
. ./cmd.sh || exit 1;
|
||||
|
||||
gpus=0,1,2,3
|
||||
stage=0
|
||||
stop_stage=4
|
||||
conf_path=conf/wav2vec2ASR.yaml
|
||||
ips= #xx.xx.xx.xx,xx.xx.xx.xx
|
||||
decode_conf_path=conf/tuning/decode.yaml
|
||||
avg_num=1
|
||||
resume= # xx e.g. 30
|
||||
export FLAGS_cudnn_deterministic=1
|
||||
. ${MAIN_ROOT}/utils/parse_options.sh || exit 1;
|
||||
|
||||
audio_file=data/demo_002_en.wav
|
||||
|
||||
avg_ckpt=avg_${avg_num}
|
||||
ckpt=$(basename ${conf_path} | awk -F'.' '{print $1}')
|
||||
echo "checkpoint name ${ckpt}"git revert -v
|
||||
|
||||
if [ ${stage} -le 0 ] && [ ${stop_stage} -ge 0 ]; then
|
||||
# prepare data
|
||||
bash ./local/data.sh || exit -1
|
||||
fi
|
||||
|
||||
if [ ${stage} -le 1 ] && [ ${stop_stage} -ge 1 ]; then
|
||||
# train model, all `ckpt` under `exp` dir
|
||||
CUDA_VISIBLE_DEVICES=${gpus} ./local/train.sh ${conf_path} ${ckpt} ${resume} ${ips}
|
||||
fi
|
||||
|
||||
if [ ${stage} -le 2 ] && [ ${stop_stage} -ge 2 ]; then
|
||||
# avg n best model
|
||||
avg.sh last exp/${ckpt}/checkpoints ${avg_num}
|
||||
fi
|
||||
|
||||
if [ ${stage} -le 3 ] && [ ${stop_stage} -ge 3 ]; then
|
||||
# greedy search decoder
|
||||
CUDA_VISIBLE_DEVICES=0 ./local/test.sh ${conf_path} ${decode_conf_path} exp/${ckpt}/checkpoints/${avg_ckpt} || exit -1
|
||||
|
||||
fi
|
||||
|
||||
if [ ${stage} -le 4 ] && [ ${stop_stage} -ge 4 ]; then
|
||||
# test a single .wav file
|
||||
CUDA_VISIBLE_DEVICES=0 ./local/test_wav.sh ${conf_path} ${decode_conf_path} exp/${ckpt}/checkpoints/${avg_ckpt} ${audio_file} || exit -1
|
||||
fi
|
||||
@ -0,0 +1 @@
|
||||
../../../utils
|
||||
@ -1,13 +0,0 @@
|
||||
#!/bin/bash
|
||||
export MAIN_ROOT=`realpath ${PWD}/../../../`
|
||||
|
||||
export PATH=${MAIN_ROOT}:${MAIN_ROOT}/utils:${PATH}
|
||||
export LC_ALL=C
|
||||
|
||||
export PYTHONDONTWRITEBYTECODE=1
|
||||
# Use UTF-8 in Python to avoid UnicodeDecodeError when LC_ALL=C
|
||||
export PYTHONIOENCODING=UTF-8
|
||||
export PYTHONPATH=${MAIN_ROOT}:${PYTHONPATH}
|
||||
|
||||
MODEL=fastspeech2
|
||||
export BIN_DIR=${MAIN_ROOT}/paddlespeech/t2s/exps/${MODEL}
|
||||
@ -0,0 +1 @@
|
||||
../../csmsc/tts3/path.sh
|
||||
@ -1,13 +0,0 @@
|
||||
#!/bin/bash
|
||||
export MAIN_ROOT=`realpath ${PWD}/../../../`
|
||||
|
||||
export PATH=${MAIN_ROOT}:${MAIN_ROOT}/utils:${PATH}
|
||||
export LC_ALL=C
|
||||
|
||||
export PYTHONDONTWRITEBYTECODE=1
|
||||
# Use UTF-8 in Python to avoid UnicodeDecodeError when LC_ALL=C
|
||||
export PYTHONIOENCODING=UTF-8
|
||||
export PYTHONPATH=${MAIN_ROOT}:${PYTHONPATH}
|
||||
|
||||
MODEL=tacotron2
|
||||
export BIN_DIR=${MAIN_ROOT}/paddlespeech/t2s/exps/${MODEL}
|
||||
@ -0,0 +1 @@
|
||||
../../csmsc/tts0/path.sh
|
||||
@ -1,13 +0,0 @@
|
||||
#!/bin/bash
|
||||
|
||||
config_path=$1
|
||||
train_output_path=$2
|
||||
|
||||
python3 ${BIN_DIR}/train.py \
|
||||
--train-metadata=dump/train/norm/metadata.jsonl \
|
||||
--dev-metadata=dump/dev/norm/metadata.jsonl \
|
||||
--config=${config_path} \
|
||||
--output-dir=${train_output_path} \
|
||||
--ngpu=2 \
|
||||
--phones-dict=dump/phone_id_map.txt \
|
||||
--voice-cloning=True
|
||||
@ -0,0 +1 @@
|
||||
../../vc0/local/train.sh
|
||||
@ -1,13 +0,0 @@
|
||||
#!/bin/bash
|
||||
export MAIN_ROOT=`realpath ${PWD}/../../../`
|
||||
|
||||
export PATH=${MAIN_ROOT}:${MAIN_ROOT}/utils:${PATH}
|
||||
export LC_ALL=C
|
||||
|
||||
export PYTHONDONTWRITEBYTECODE=1
|
||||
# Use UTF-8 in Python to avoid UnicodeDecodeError when LC_ALL=C
|
||||
export PYTHONIOENCODING=UTF-8
|
||||
export PYTHONPATH=${MAIN_ROOT}:${PYTHONPATH}
|
||||
|
||||
MODEL=fastspeech2
|
||||
export BIN_DIR=${MAIN_ROOT}/paddlespeech/t2s/exps/${MODEL}
|
||||
@ -0,0 +1 @@
|
||||
../../csmsc/tts3/path.sh
|
||||
@ -1,20 +0,0 @@
|
||||
#!/bin/bash
|
||||
|
||||
config_path=$1
|
||||
train_output_path=$2
|
||||
ckpt_name=$3
|
||||
|
||||
python3 ${BIN_DIR}/../synthesize.py \
|
||||
--am=fastspeech2_aishell3 \
|
||||
--am_config=${config_path} \
|
||||
--am_ckpt=${train_output_path}/checkpoints/${ckpt_name} \
|
||||
--am_stat=dump/train/speech_stats.npy \
|
||||
--voc=pwgan_aishell3 \
|
||||
--voc_config=pwg_aishell3_ckpt_0.5/default.yaml \
|
||||
--voc_ckpt=pwg_aishell3_ckpt_0.5/snapshot_iter_1000000.pdz \
|
||||
--voc_stat=pwg_aishell3_ckpt_0.5/feats_stats.npy \
|
||||
--test_metadata=dump/test/norm/metadata.jsonl \
|
||||
--output_dir=${train_output_path}/test \
|
||||
--phones_dict=dump/phone_id_map.txt \
|
||||
--speaker_dict=dump/speaker_id_map.txt \
|
||||
--voice-cloning=True
|
||||
@ -0,0 +1 @@
|
||||
../../vc1/local/synthesize.sh
|
||||
@ -1,13 +0,0 @@
|
||||
#!/bin/bash
|
||||
|
||||
config_path=$1
|
||||
train_output_path=$2
|
||||
|
||||
python3 ${BIN_DIR}/train.py \
|
||||
--train-metadata=dump/train/norm/metadata.jsonl \
|
||||
--dev-metadata=dump/dev/norm/metadata.jsonl \
|
||||
--config=${config_path} \
|
||||
--output-dir=${train_output_path} \
|
||||
--ngpu=2 \
|
||||
--phones-dict=dump/phone_id_map.txt \
|
||||
--voice-cloning=True
|
||||
@ -0,0 +1 @@
|
||||
../../vc0/local/train.sh
|
||||
@ -1,13 +0,0 @@
|
||||
#!/bin/bash
|
||||
export MAIN_ROOT=`realpath ${PWD}/../../../`
|
||||
|
||||
export PATH=${MAIN_ROOT}:${MAIN_ROOT}/utils:${PATH}
|
||||
export LC_ALL=C
|
||||
|
||||
export PYTHONDONTWRITEBYTECODE=1
|
||||
# Use UTF-8 in Python to avoid UnicodeDecodeError when LC_ALL=C
|
||||
export PYTHONIOENCODING=UTF-8
|
||||
export PYTHONPATH=${MAIN_ROOT}:${PYTHONPATH}
|
||||
|
||||
MODEL=fastspeech2
|
||||
export BIN_DIR=${MAIN_ROOT}/paddlespeech/t2s/exps/${MODEL}
|
||||
@ -0,0 +1 @@
|
||||
../../csmsc/tts3/path.sh
|
||||
@ -1,14 +0,0 @@
|
||||
#!/bin/bash
|
||||
|
||||
config_path=$1
|
||||
train_output_path=$2
|
||||
ckpt_name=$3
|
||||
|
||||
FLAGS_allocator_strategy=naive_best_fit \
|
||||
FLAGS_fraction_of_gpu_memory_to_use=0.01 \
|
||||
python3 ${BIN_DIR}/../synthesize.py \
|
||||
--config=${config_path} \
|
||||
--checkpoint=${train_output_path}/checkpoints/${ckpt_name} \
|
||||
--test-metadata=dump/test/norm/metadata.jsonl \
|
||||
--output-dir=${train_output_path}/test \
|
||||
--generator-type=pwgan
|
||||
@ -0,0 +1 @@
|
||||
../../../csmsc/voc1/local/synthesize.sh
|
||||
@ -1,13 +0,0 @@
|
||||
#!/bin/bash
|
||||
|
||||
config_path=$1
|
||||
train_output_path=$2
|
||||
|
||||
FLAGS_cudnn_exhaustive_search=true \
|
||||
FLAGS_conv_workspace_size_limit=4000 \
|
||||
python ${BIN_DIR}/train.py \
|
||||
--train-metadata=dump/train/norm/metadata.jsonl \
|
||||
--dev-metadata=dump/dev/norm/metadata.jsonl \
|
||||
--config=${config_path} \
|
||||
--output-dir=${train_output_path} \
|
||||
--ngpu=1
|
||||
@ -0,0 +1 @@
|
||||
../../../csmsc/voc1/local/train.sh
|
||||
@ -1,13 +0,0 @@
|
||||
#!/bin/bash
|
||||
export MAIN_ROOT=`realpath ${PWD}/../../../`
|
||||
|
||||
export PATH=${MAIN_ROOT}:${MAIN_ROOT}/utils:${PATH}
|
||||
export LC_ALL=C
|
||||
|
||||
export PYTHONDONTWRITEBYTECODE=1
|
||||
# Use UTF-8 in Python to avoid UnicodeDecodeError when LC_ALL=C
|
||||
export PYTHONIOENCODING=UTF-8
|
||||
export PYTHONPATH=${MAIN_ROOT}:${PYTHONPATH}
|
||||
|
||||
MODEL=parallelwave_gan
|
||||
export BIN_DIR=${MAIN_ROOT}/paddlespeech/t2s/exps/gan_vocoder/${MODEL}
|
||||
@ -0,0 +1 @@
|
||||
../../csmsc/voc1/path.sh
|
||||
@ -1,55 +0,0 @@
|
||||
#!/bin/bash
|
||||
|
||||
stage=0
|
||||
stop_stage=100
|
||||
|
||||
config_path=$1
|
||||
|
||||
if [ ${stage} -le 0 ] && [ ${stop_stage} -ge 0 ]; then
|
||||
# get durations from MFA's result
|
||||
echo "Generate durations.txt from MFA results ..."
|
||||
python3 ${MAIN_ROOT}/utils/gen_duration_from_textgrid.py \
|
||||
--inputdir=./aishell3_alignment_tone \
|
||||
--output=durations.txt \
|
||||
--config=${config_path}
|
||||
fi
|
||||
|
||||
if [ ${stage} -le 1 ] && [ ${stop_stage} -ge 1 ]; then
|
||||
# extract features
|
||||
echo "Extract features ..."
|
||||
python3 ${BIN_DIR}/../preprocess.py \
|
||||
--rootdir=~/datasets/data_aishell3/ \
|
||||
--dataset=aishell3 \
|
||||
--dumpdir=dump \
|
||||
--dur-file=durations.txt \
|
||||
--config=${config_path} \
|
||||
--cut-sil=True \
|
||||
--num-cpu=20
|
||||
fi
|
||||
|
||||
if [ ${stage} -le 2 ] && [ ${stop_stage} -ge 2 ]; then
|
||||
# get features' stats(mean and std)
|
||||
echo "Get features' stats ..."
|
||||
python3 ${MAIN_ROOT}/utils/compute_statistics.py \
|
||||
--metadata=dump/train/raw/metadata.jsonl \
|
||||
--field-name="feats"
|
||||
fi
|
||||
|
||||
if [ ${stage} -le 3 ] && [ ${stop_stage} -ge 3 ]; then
|
||||
# normalize, dev and test should use train's stats
|
||||
echo "Normalize ..."
|
||||
|
||||
python3 ${BIN_DIR}/../normalize.py \
|
||||
--metadata=dump/train/raw/metadata.jsonl \
|
||||
--dumpdir=dump/train/norm \
|
||||
--stats=dump/train/feats_stats.npy
|
||||
python3 ${BIN_DIR}/../normalize.py \
|
||||
--metadata=dump/dev/raw/metadata.jsonl \
|
||||
--dumpdir=dump/dev/norm \
|
||||
--stats=dump/train/feats_stats.npy
|
||||
|
||||
python3 ${BIN_DIR}/../normalize.py \
|
||||
--metadata=dump/test/raw/metadata.jsonl \
|
||||
--dumpdir=dump/test/norm \
|
||||
--stats=dump/train/feats_stats.npy
|
||||
fi
|
||||
@ -0,0 +1 @@
|
||||
../../voc1/local/preprocess.sh
|
||||
@ -1,14 +0,0 @@
|
||||
#!/bin/bash
|
||||
|
||||
config_path=$1
|
||||
train_output_path=$2
|
||||
ckpt_name=$3
|
||||
|
||||
FLAGS_allocator_strategy=naive_best_fit \
|
||||
FLAGS_fraction_of_gpu_memory_to_use=0.01 \
|
||||
python3 ${BIN_DIR}/../synthesize.py \
|
||||
--config=${config_path} \
|
||||
--checkpoint=${train_output_path}/checkpoints/${ckpt_name} \
|
||||
--test-metadata=dump/test/norm/metadata.jsonl \
|
||||
--output-dir=${train_output_path}/test \
|
||||
--generator-type=hifigan
|
||||
@ -0,0 +1 @@
|
||||
../../../csmsc/voc5/local/synthesize.sh
|
||||
@ -1,13 +0,0 @@
|
||||
#!/bin/bash
|
||||
|
||||
config_path=$1
|
||||
train_output_path=$2
|
||||
|
||||
FLAGS_cudnn_exhaustive_search=true \
|
||||
FLAGS_conv_workspace_size_limit=4000 \
|
||||
python ${BIN_DIR}/train.py \
|
||||
--train-metadata=dump/train/norm/metadata.jsonl \
|
||||
--dev-metadata=dump/dev/norm/metadata.jsonl \
|
||||
--config=${config_path} \
|
||||
--output-dir=${train_output_path} \
|
||||
--ngpu=1
|
||||
@ -0,0 +1 @@
|
||||
../../../csmsc/voc1/local/train.sh
|
||||
@ -1,13 +0,0 @@
|
||||
#!/bin/bash
|
||||
export MAIN_ROOT=`realpath ${PWD}/../../../`
|
||||
|
||||
export PATH=${MAIN_ROOT}:${MAIN_ROOT}/utils:${PATH}
|
||||
export LC_ALL=C
|
||||
|
||||
export PYTHONDONTWRITEBYTECODE=1
|
||||
# Use UTF-8 in Python to avoid UnicodeDecodeError when LC_ALL=C
|
||||
export PYTHONIOENCODING=UTF-8
|
||||
export PYTHONPATH=${MAIN_ROOT}:${PYTHONPATH}
|
||||
|
||||
MODEL=hifigan
|
||||
export BIN_DIR=${MAIN_ROOT}/paddlespeech/t2s/exps/gan_vocoder/${MODEL}
|
||||
@ -0,0 +1 @@
|
||||
../../csmsc/voc5/path.sh
|
||||
@ -1,25 +0,0 @@
|
||||
#!/bin/bash
|
||||
|
||||
config_path=$1
|
||||
train_output_path=$2
|
||||
ckpt_name=$3
|
||||
|
||||
stage=0
|
||||
stop_stage=0
|
||||
|
||||
# hifigan
|
||||
if [ ${stage} -le 0 ] && [ ${stop_stage} -ge 0 ]; then
|
||||
FLAGS_allocator_strategy=naive_best_fit \
|
||||
FLAGS_fraction_of_gpu_memory_to_use=0.01 \
|
||||
python3 ${BIN_DIR}/synthesize.py \
|
||||
--erniesat_config=${config_path} \
|
||||
--erniesat_ckpt=${train_output_path}/checkpoints/${ckpt_name} \
|
||||
--erniesat_stat=dump/train/speech_stats.npy \
|
||||
--voc=hifigan_aishell3 \
|
||||
--voc_config=hifigan_aishell3_ckpt_0.2.0/default.yaml \
|
||||
--voc_ckpt=hifigan_aishell3_ckpt_0.2.0/snapshot_iter_2500000.pdz \
|
||||
--voc_stat=hifigan_aishell3_ckpt_0.2.0/feats_stats.npy \
|
||||
--test_metadata=dump/test/norm/metadata.jsonl \
|
||||
--output_dir=${train_output_path}/test \
|
||||
--phones_dict=dump/phone_id_map.txt
|
||||
fi
|
||||
@ -0,0 +1 @@
|
||||
../../../aishell3/ernie_sat/local/synthesize.sh
|
||||
@ -1,12 +0,0 @@
|
||||
#!/bin/bash
|
||||
|
||||
config_path=$1
|
||||
train_output_path=$2
|
||||
|
||||
python3 ${BIN_DIR}/train.py \
|
||||
--train-metadata=dump/train/norm/metadata.jsonl \
|
||||
--dev-metadata=dump/dev/norm/metadata.jsonl \
|
||||
--config=${config_path} \
|
||||
--output-dir=${train_output_path} \
|
||||
--ngpu=8 \
|
||||
--phones-dict=dump/phone_id_map.txt
|
||||
@ -0,0 +1 @@
|
||||
../../../aishell3/ernie_sat/local/train.sh
|
||||
@ -1,13 +0,0 @@
|
||||
#!/bin/bash
|
||||
export MAIN_ROOT=`realpath ${PWD}/../../../`
|
||||
|
||||
export PATH=${MAIN_ROOT}:${MAIN_ROOT}/utils:${PATH}
|
||||
export LC_ALL=C
|
||||
|
||||
export PYTHONDONTWRITEBYTECODE=1
|
||||
# Use UTF-8 in Python to avoid UnicodeDecodeError when LC_ALL=C
|
||||
export PYTHONIOENCODING=UTF-8
|
||||
export PYTHONPATH=${MAIN_ROOT}:${PYTHONPATH}
|
||||
|
||||
MODEL=ernie_sat
|
||||
export BIN_DIR=${MAIN_ROOT}/paddlespeech/t2s/exps/${MODEL}
|
||||
@ -0,0 +1 @@
|
||||
../../aishell3/ernie_sat/path.sh
|
||||
@ -0,0 +1,121 @@
|
||||
# FastSpeech2 with Cantonese language
|
||||
|
||||
## Dataset
|
||||
### Download and Extract
|
||||
If you don't have the Cantonese datasets mentioned above, please download and unzip [Guangzhou_Cantonese_Scripted_Speech_Corpus_Daily_Use_Sentence](https://magichub.com/datasets/guangzhou-cantonese-scripted-speech-corpus-daily-use-sentence/) and [Guangzhou_Cantonese_Scripted_Speech_Corpus_in_Vehicle](https://magichub.com/datasets/guangzhou-cantonese-scripted-speech-corpus-in-the-vehicle/) under `~/datasets/`.
|
||||
|
||||
To obtain better performance, please combine these two datasets together as follows:
|
||||
|
||||
```bash
|
||||
mkdir -p ~/datasets/canton_all/WAV
|
||||
cp -r ~/datasets/Guangzhou_Cantonese_Scripted_Speech_Corpus_Daily_Use_Sentence/WAV/* ~/datasets/canton_all/WAV
|
||||
cp -r ~/datasets/Guangzhou_Cantonese_Scripted_Speech_Corpus_in_Vehicle/WAV/* ~/datasets/canton_all/WAV
|
||||
```
|
||||
|
||||
After that, it should be look like:
|
||||
```
|
||||
~/datasets/canton_all
|
||||
│ └── WAV
|
||||
│ └──G0001
|
||||
│ └──G0002
|
||||
│ ...
|
||||
│ └──G0071
|
||||
│ └──G0072
|
||||
```
|
||||
|
||||
|
||||
### Get MFA Result and Extract
|
||||
We use [MFA1.x](https://github.com/MontrealCorpusTools/Montreal-Forced-Aligner) to get durations for canton_fastspeech2.
|
||||
You can train your MFA model reference to [canton_mfa example](https://github.com/PaddlePaddle/PaddleSpeech/tree/develop/examples/other/mfa) (use MFA1.x now) of our repo.
|
||||
We here provide the MFA results of these two datasets. [canton_alignment.zip](https://paddlespeech.bj.bcebos.com/MFA/Canton/canton_alignment.zip)
|
||||
|
||||
## Get Started
|
||||
Assume the path to the Cantonese MFA result of the two datsets mentioned above is `./canton_alignment`.
|
||||
Run the command below to
|
||||
1. **source path**.
|
||||
2. preprocess the dataset.
|
||||
3. train the model.
|
||||
4. synthesize wavs.
|
||||
- synthesize waveform from `metadata.jsonl`.
|
||||
- synthesize waveform from text file.
|
||||
```bash
|
||||
./run.sh
|
||||
```
|
||||
You can choose a range of stages you want to run, or set `stage` equal to `stop-stage` to use only one stage, for example, running the following command will only preprocess the dataset.
|
||||
```bash
|
||||
./run.sh --stage 0 --stop-stage 0
|
||||
```
|
||||
|
||||
### Data Preprocessing
|
||||
```bash
|
||||
./local/preprocess.sh ${conf_path}
|
||||
```
|
||||
When it is done. A `dump` folder is created in the current directory. The structure of the dump folder is listed below.
|
||||
```text
|
||||
dump
|
||||
├── dev
|
||||
│ ├── norm
|
||||
│ └── raw
|
||||
├── phone_id_map.txt
|
||||
├── speaker_id_map.txt
|
||||
├── test
|
||||
│ ├── norm
|
||||
│ └── raw
|
||||
└── train
|
||||
├── energy_stats.npy
|
||||
├── norm
|
||||
├── pitch_stats.npy
|
||||
├── raw
|
||||
└── speech_stats.npy
|
||||
```
|
||||
The dataset is split into 3 parts, namely `train`, `dev`, and` test`, each of which contains a `norm` and `raw` subfolder. The raw folder contains speech、pitch and energy features of each utterance, while the norm folder contains normalized ones. The statistics used to normalize features are computed from the training set, which is located in `dump/train/*_stats.npy`.
|
||||
|
||||
Also, there is a `metadata.jsonl` in each subfolder. It is a table-like file that contains phones, text_lengths, speech_lengths, durations, the path of speech features, the path of pitch features, a path of energy features, speaker, and id of each utterance.
|
||||
|
||||
### Training details can refer to the script of [examples/aishell3/tts3](../../aishell3/tts3).
|
||||
|
||||
## Pretrained Model
|
||||
Pretrained FastSpeech2 model with no silence in the edge of audios:
|
||||
- [fastspeech2_canton_ckpt_1.4.0.zip](https://paddlespeech.bj.bcebos.com/Parakeet/released_models/fastspeech2/fastspeech2_canton_ckpt_1.4.0.zip)
|
||||
|
||||
FastSpeech2 checkpoint contains files listed below.
|
||||
|
||||
```text
|
||||
fastspeech2_canton_ckpt_1.4.0
|
||||
├── default.yaml # default config used to train fastspeech2
|
||||
├── energy_stats.npy # statistics used to normalize energy when training fastspeech2
|
||||
├── phone_id_map.txt # phone vocabulary file when training fastspeech2
|
||||
├── pitch_stats.npy # statistics used to normalize pitch when training fastspeech2
|
||||
├── snapshot_iter_140000.pdz # model parameters and optimizer states
|
||||
├── speaker_id_map.txt # speaker id map file when training a multi-speaker fastspeech2
|
||||
└── speech_stats.npy # statistics used to normalize spectrogram when training fastspeech2
|
||||
```
|
||||
We use [parallel wavegan](https://github.com/PaddlePaddle/PaddleSpeech/tree/develop/examples/aishell3/voc1) as the neural vocoder.
|
||||
Download the pretrained parallel wavegan model from [pwg_aishell3_ckpt_0.5.zip](https://paddlespeech.bj.bcebos.com/Parakeet/released_models/pwgan/pwg_aishell3_ckpt_0.5.zip) and unzip it.
|
||||
```bash
|
||||
unzip pwg_aishell3_ckpt_0.5.zip
|
||||
```
|
||||
|
||||
You can use the following scripts to synthesize for `${BIN_DIR}/../sentences_canton.txt` using pretrained fastspeech2 and parallel wavegan models.
|
||||
```bash
|
||||
source path.sh
|
||||
|
||||
FLAGS_allocator_strategy=naive_best_fit \
|
||||
FLAGS_fraction_of_gpu_memory_to_use=0.01 \
|
||||
python3 ${BIN_DIR}/../synthesize_e2e.py \
|
||||
--am=fastspeech2_aishell3 \
|
||||
--am_config=fastspeech2_canton_ckpt_1.4.0/default.yaml \
|
||||
--am_ckpt=fastspeech2_canton_ckpt_1.4.0/snapshot_iter_140000.pdz \
|
||||
--am_stat=fastspeech2_canton_ckpt_1.4.0/speech_stats.npy \
|
||||
--voc=pwgan_aishell3 \
|
||||
--voc_config=pwg_aishell3_ckpt_0.5/default.yaml \
|
||||
--voc_ckpt=pwg_aishell3_ckpt_0.5/snapshot_iter_1000000.pdz \
|
||||
--voc_stat=pwg_aishell3_ckpt_0.5/feats_stats.npy \
|
||||
--lang=canton \
|
||||
--text=${BIN_DIR}/../sentences_canton.txt \
|
||||
--output_dir=exp/default/test_e2e \
|
||||
--phones_dict=fastspeech2_canton_ckpt_1.4.0/phone_id_map.txt \
|
||||
--speaker_dict=fastspeech2_canton_ckpt_1.4.0/speaker_id_map.txt \
|
||||
--spk_id=0 \
|
||||
--inference_dir=exp/default/inference
|
||||
```
|
||||
@ -0,0 +1,107 @@
|
||||
###########################################################
|
||||
# FEATURE EXTRACTION SETTING #
|
||||
###########################################################
|
||||
|
||||
fs: 24000 # sr
|
||||
n_fft: 2048 # FFT size (samples).
|
||||
n_shift: 300 # Hop size (samples). 12.5ms
|
||||
win_length: 1200 # Window length (samples). 50ms
|
||||
# If set to null, it will be the same as fft_size.
|
||||
window: "hann" # Window function.
|
||||
|
||||
# Only used for feats_type != raw
|
||||
|
||||
fmin: 80 # Minimum frequency of Mel basis.
|
||||
fmax: 7600 # Maximum frequency of Mel basis.
|
||||
n_mels: 80 # The number of mel basis.
|
||||
|
||||
# Only used for the model using pitch features (e.g. FastSpeech2)
|
||||
# The canton datasets we use are different from others like Databaker or LJSpeech,
|
||||
# we set it to 110 to avoid too many zero-pitch problem.
|
||||
# Reference: https://github.com/JeremyCCHsu/Python-Wrapper-for-World-Vocoder/issues/38
|
||||
f0min: 110 # Minimum f0 for pitch extraction.
|
||||
f0max: 400 # Maximum f0 for pitch extraction.
|
||||
|
||||
|
||||
###########################################################
|
||||
# DATA SETTING #
|
||||
###########################################################
|
||||
batch_size: 32
|
||||
num_workers: 2
|
||||
|
||||
|
||||
###########################################################
|
||||
# MODEL SETTING #
|
||||
###########################################################
|
||||
model:
|
||||
adim: 384 # attention dimension
|
||||
aheads: 2 # number of attention heads
|
||||
elayers: 4 # number of encoder layers
|
||||
eunits: 1536 # number of encoder ff units
|
||||
dlayers: 4 # number of decoder layers
|
||||
dunits: 1536 # number of decoder ff units
|
||||
positionwise_layer_type: conv1d # type of position-wise layer
|
||||
positionwise_conv_kernel_size: 3 # kernel size of position wise conv layer
|
||||
duration_predictor_layers: 2 # number of layers of duration predictor
|
||||
duration_predictor_chans: 256 # number of channels of duration predictor
|
||||
duration_predictor_kernel_size: 3 # filter size of duration predictor
|
||||
postnet_layers: 5 # number of layers of postnset
|
||||
postnet_filts: 5 # filter size of conv layers in postnet
|
||||
postnet_chans: 256 # number of channels of conv layers in postnet
|
||||
use_scaled_pos_enc: True # whether to use scaled positional encoding
|
||||
encoder_normalize_before: True # whether to perform layer normalization before the input
|
||||
decoder_normalize_before: True # whether to perform layer normalization before the input
|
||||
reduction_factor: 1 # reduction factor
|
||||
init_type: xavier_uniform # initialization type
|
||||
init_enc_alpha: 1.0 # initial value of alpha of encoder scaled position encoding
|
||||
init_dec_alpha: 1.0 # initial value of alpha of decoder scaled position encoding
|
||||
transformer_enc_dropout_rate: 0.2 # dropout rate for transformer encoder layer
|
||||
transformer_enc_positional_dropout_rate: 0.2 # dropout rate for transformer encoder positional encoding
|
||||
transformer_enc_attn_dropout_rate: 0.2 # dropout rate for transformer encoder attention layer
|
||||
transformer_dec_dropout_rate: 0.2 # dropout rate for transformer decoder layer
|
||||
transformer_dec_positional_dropout_rate: 0.2 # dropout rate for transformer decoder positional encoding
|
||||
transformer_dec_attn_dropout_rate: 0.2 # dropout rate for transformer decoder attention layer
|
||||
pitch_predictor_layers: 5 # number of conv layers in pitch predictor
|
||||
pitch_predictor_chans: 256 # number of channels of conv layers in pitch predictor
|
||||
pitch_predictor_kernel_size: 5 # kernel size of conv leyers in pitch predictor
|
||||
pitch_predictor_dropout: 0.5 # dropout rate in pitch predictor
|
||||
pitch_embed_kernel_size: 1 # kernel size of conv embedding layer for pitch
|
||||
pitch_embed_dropout: 0.0 # dropout rate after conv embedding layer for pitch
|
||||
stop_gradient_from_pitch_predictor: True # whether to stop the gradient from pitch predictor to encoder
|
||||
energy_predictor_layers: 2 # number of conv layers in energy predictor
|
||||
energy_predictor_chans: 256 # number of channels of conv layers in energy predictor
|
||||
energy_predictor_kernel_size: 3 # kernel size of conv leyers in energy predictor
|
||||
energy_predictor_dropout: 0.5 # dropout rate in energy predictor
|
||||
energy_embed_kernel_size: 1 # kernel size of conv embedding layer for energy
|
||||
energy_embed_dropout: 0.0 # dropout rate after conv embedding layer for energy
|
||||
stop_gradient_from_energy_predictor: False # whether to stop the gradient from energy predictor to encoder
|
||||
spk_embed_dim: 256 # speaker embedding dimension
|
||||
spk_embed_integration_type: concat # speaker embedding integration type
|
||||
|
||||
|
||||
|
||||
###########################################################
|
||||
# UPDATER SETTING #
|
||||
###########################################################
|
||||
updater:
|
||||
use_masking: True # whether to apply masking for padded part in loss calculation
|
||||
|
||||
|
||||
###########################################################
|
||||
# OPTIMIZER SETTING #
|
||||
###########################################################
|
||||
optimizer:
|
||||
optim: adam # optimizer type
|
||||
learning_rate: 0.001 # learning rate
|
||||
|
||||
###########################################################
|
||||
# TRAINING SETTING #
|
||||
###########################################################
|
||||
max_epoch: 1000
|
||||
num_snapshots: 5
|
||||
|
||||
|
||||
###########################################################
|
||||
# OTHER SETTING #
|
||||
###########################################################
|
||||
seed: 10086
|
||||
@ -0,0 +1,75 @@
|
||||
#!/bin/bash
|
||||
|
||||
stage=0
|
||||
stop_stage=100
|
||||
|
||||
config_path=$1
|
||||
|
||||
if [ ${stage} -le 0 ] && [ ${stop_stage} -ge 0 ]; then
|
||||
# get durations from MFA's result
|
||||
echo "Generate durations.txt from MFA results ..."
|
||||
python3 ${MAIN_ROOT}/utils/gen_duration_from_textgrid.py \
|
||||
--inputdir=./canton_alignment \
|
||||
--output durations.txt \
|
||||
--config=${config_path}
|
||||
fi
|
||||
|
||||
if [ ${stage} -le 1 ] && [ ${stop_stage} -ge 1 ]; then
|
||||
# extract features
|
||||
echo "Extract features ..."
|
||||
python3 ${BIN_DIR}/preprocess.py \
|
||||
--dataset=canton \
|
||||
--rootdir=~/datasets/canton_all \
|
||||
--dumpdir=dump \
|
||||
--dur-file=durations.txt \
|
||||
--config=${config_path} \
|
||||
--num-cpu=20 \
|
||||
--cut-sil=True
|
||||
fi
|
||||
|
||||
if [ ${stage} -le 2 ] && [ ${stop_stage} -ge 2 ]; then
|
||||
# get features' stats(mean and std)
|
||||
echo "Get features' stats ..."
|
||||
python3 ${MAIN_ROOT}/utils/compute_statistics.py \
|
||||
--metadata=dump/train/raw/metadata.jsonl \
|
||||
--field-name="speech"
|
||||
|
||||
python3 ${MAIN_ROOT}/utils/compute_statistics.py \
|
||||
--metadata=dump/train/raw/metadata.jsonl \
|
||||
--field-name="pitch"
|
||||
|
||||
python3 ${MAIN_ROOT}/utils/compute_statistics.py \
|
||||
--metadata=dump/train/raw/metadata.jsonl \
|
||||
--field-name="energy"
|
||||
fi
|
||||
|
||||
if [ ${stage} -le 3 ] && [ ${stop_stage} -ge 3 ]; then
|
||||
# normalize and covert phone/speaker to id, dev and test should use train's stats
|
||||
echo "Normalize ..."
|
||||
python3 ${BIN_DIR}/normalize.py \
|
||||
--metadata=dump/train/raw/metadata.jsonl \
|
||||
--dumpdir=dump/train/norm \
|
||||
--speech-stats=dump/train/speech_stats.npy \
|
||||
--pitch-stats=dump/train/pitch_stats.npy \
|
||||
--energy-stats=dump/train/energy_stats.npy \
|
||||
--phones-dict=dump/phone_id_map.txt \
|
||||
--speaker-dict=dump/speaker_id_map.txt
|
||||
|
||||
python3 ${BIN_DIR}/normalize.py \
|
||||
--metadata=dump/dev/raw/metadata.jsonl \
|
||||
--dumpdir=dump/dev/norm \
|
||||
--speech-stats=dump/train/speech_stats.npy \
|
||||
--pitch-stats=dump/train/pitch_stats.npy \
|
||||
--energy-stats=dump/train/energy_stats.npy \
|
||||
--phones-dict=dump/phone_id_map.txt \
|
||||
--speaker-dict=dump/speaker_id_map.txt
|
||||
|
||||
python3 ${BIN_DIR}/normalize.py \
|
||||
--metadata=dump/test/raw/metadata.jsonl \
|
||||
--dumpdir=dump/test/norm \
|
||||
--speech-stats=dump/train/speech_stats.npy \
|
||||
--pitch-stats=dump/train/pitch_stats.npy \
|
||||
--energy-stats=dump/train/energy_stats.npy \
|
||||
--phones-dict=dump/phone_id_map.txt \
|
||||
--speaker-dict=dump/speaker_id_map.txt
|
||||
fi
|
||||
@ -0,0 +1 @@
|
||||
../../../aishell3/tts3/local/synthesize.sh
|
||||
@ -0,0 +1,53 @@
|
||||
#!/bin/bash
|
||||
|
||||
config_path=$1
|
||||
train_output_path=$2
|
||||
ckpt_name=$3
|
||||
|
||||
stage=0
|
||||
stop_stage=0
|
||||
|
||||
# pwgan
|
||||
if [ ${stage} -le 0 ] && [ ${stop_stage} -ge 0 ]; then
|
||||
FLAGS_allocator_strategy=naive_best_fit \
|
||||
FLAGS_fraction_of_gpu_memory_to_use=0.01 \
|
||||
python3 ${BIN_DIR}/../synthesize_e2e.py \
|
||||
--am=fastspeech2_canton \
|
||||
--am_config=${config_path} \
|
||||
--am_ckpt=${train_output_path}/checkpoints/${ckpt_name} \
|
||||
--am_stat=dump/train/speech_stats.npy \
|
||||
--voc=pwgan_aishell3 \
|
||||
--voc_config=pwg_aishell3_ckpt_0.5/default.yaml \
|
||||
--voc_ckpt=pwg_aishell3_ckpt_0.5/snapshot_iter_1000000.pdz \
|
||||
--voc_stat=pwg_aishell3_ckpt_0.5/feats_stats.npy \
|
||||
--lang=canton \
|
||||
--text=${BIN_DIR}/../sentences_canton.txt \
|
||||
--output_dir=${train_output_path}/test_e2e \
|
||||
--phones_dict=dump/phone_id_map.txt \
|
||||
--speaker_dict=dump/speaker_id_map.txt \
|
||||
--spk_id=0 \
|
||||
--inference_dir=${train_output_path}/inference
|
||||
fi
|
||||
|
||||
# hifigan
|
||||
if [ ${stage} -le 1 ] && [ ${stop_stage} -ge 1 ]; then
|
||||
echo "in hifigan syn_e2e"
|
||||
FLAGS_allocator_strategy=naive_best_fit \
|
||||
FLAGS_fraction_of_gpu_memory_to_use=0.01 \
|
||||
python3 ${BIN_DIR}/../synthesize_e2e.py \
|
||||
--am=fastspeech2_canton \
|
||||
--am_config=${config_path} \
|
||||
--am_ckpt=${train_output_path}/checkpoints/${ckpt_name} \
|
||||
--am_stat=dump/train/speech_stats.npy \
|
||||
--voc=hifigan_aishell3 \
|
||||
--voc_config=hifigan_aishell3_ckpt_0.2.0/default.yaml \
|
||||
--voc_ckpt=hifigan_aishell3_ckpt_0.2.0/snapshot_iter_2500000.pdz \
|
||||
--voc_stat=hifigan_aishell3_ckpt_0.2.0/feats_stats.npy \
|
||||
--lang=canton \
|
||||
--text=${BIN_DIR}/../sentences_canton.txt \
|
||||
--output_dir=${train_output_path}/test_e2e \
|
||||
--phones_dict=dump/phone_id_map.txt \
|
||||
--speaker_dict=dump/speaker_id_map.txt \
|
||||
--spk_id=0 \
|
||||
--inference_dir=${train_output_path}/inference
|
||||
fi
|
||||
@ -0,0 +1 @@
|
||||
../../../aishell3/tts3/local/train.sh
|
||||
@ -0,0 +1 @@
|
||||
../../csmsc/tts3/path.sh
|
||||
@ -0,0 +1,38 @@
|
||||
#!/bin/bash
|
||||
|
||||
set -e
|
||||
source path.sh
|
||||
|
||||
gpus=0,1
|
||||
stage=0
|
||||
stop_stage=100
|
||||
|
||||
conf_path=conf/default.yaml
|
||||
train_output_path=exp/default
|
||||
|
||||
ckpt_name=snapshot_iter_140000.pdz
|
||||
|
||||
# with the following command, you can choose the stage range you want to run
|
||||
# such as `./run.sh --stage 0 --stop-stage 0`
|
||||
# this can not be mixed use with `$1`, `$2` ...
|
||||
source ${MAIN_ROOT}/utils/parse_options.sh || exit 1
|
||||
|
||||
if [ ${stage} -le 0 ] && [ ${stop_stage} -ge 0 ]; then
|
||||
# prepare data
|
||||
./local/preprocess.sh ${conf_path} || exit -1
|
||||
fi
|
||||
|
||||
if [ ${stage} -le 1 ] && [ ${stop_stage} -ge 1 ]; then
|
||||
# train model, all `ckpt` under `train_output_path/checkpoints/` dir
|
||||
CUDA_VISIBLE_DEVICES=${gpus} ./local/train.sh ${conf_path} ${train_output_path} || exit -1
|
||||
fi
|
||||
|
||||
if [ ${stage} -le 2 ] && [ ${stop_stage} -ge 2 ]; then
|
||||
# synthesize, vocoder is pwgan by default
|
||||
CUDA_VISIBLE_DEVICES=${gpus} ./local/synthesize.sh ${conf_path} ${train_output_path} ${ckpt_name} || exit -1
|
||||
fi
|
||||
|
||||
if [ ${stage} -le 3 ] && [ ${stop_stage} -ge 3 ]; then
|
||||
# synthesize_e2e, vocoder is pwgan by default
|
||||
CUDA_VISIBLE_DEVICES=${gpus} ./local/synthesize_e2e.sh ${conf_path} ${train_output_path} ${ckpt_name} || exit -1
|
||||
fi
|
||||
@ -1,64 +0,0 @@
|
||||
#!/bin/bash
|
||||
|
||||
source path.sh
|
||||
|
||||
gpus=0
|
||||
stage=0
|
||||
stop_stage=100
|
||||
|
||||
source ${MAIN_ROOT}/utils/parse_options.sh || exit 1
|
||||
|
||||
if [ ${stage} -le 0 ] && [ ${stop_stage} -ge 0 ]; then
|
||||
python3 ${MAIN_ROOT}/paddlespeech/t2s/exps/fastspeech2/gen_gta_mel.py \
|
||||
--fastspeech2-config=fastspeech2_nosil_baker_ckpt_0.4/default.yaml \
|
||||
--fastspeech2-checkpoint=fastspeech2_nosil_baker_ckpt_0.4/snapshot_iter_76000.pdz \
|
||||
--fastspeech2-stat=fastspeech2_nosil_baker_ckpt_0.4/speech_stats.npy \
|
||||
--dur-file=durations.txt \
|
||||
--output-dir=dump_finetune \
|
||||
--phones-dict=fastspeech2_nosil_baker_ckpt_0.4/phone_id_map.txt \
|
||||
--dataset=baker \
|
||||
--rootdir=~/datasets/BZNSYP/
|
||||
fi
|
||||
|
||||
if [ ${stage} -le 1 ] && [ ${stop_stage} -ge 1 ]; then
|
||||
python3 ${MAIN_ROOT}/utils/link_wav.py \
|
||||
--old-dump-dir=dump \
|
||||
--dump-dir=dump_finetune
|
||||
fi
|
||||
|
||||
if [ ${stage} -le 2 ] && [ ${stop_stage} -ge 2 ]; then
|
||||
# get features' stats(mean and std)
|
||||
echo "Get features' stats ..."
|
||||
cp dump/train/feats_stats.npy dump_finetune/train/
|
||||
fi
|
||||
|
||||
if [ ${stage} -le 3 ] && [ ${stop_stage} -ge 3 ]; then
|
||||
# normalize, dev and test should use train's stats
|
||||
echo "Normalize ..."
|
||||
|
||||
python3 ${BIN_DIR}/../normalize.py \
|
||||
--metadata=dump_finetune/train/raw/metadata.jsonl \
|
||||
--dumpdir=dump_finetune/train/norm \
|
||||
--stats=dump_finetune/train/feats_stats.npy
|
||||
python3 ${BIN_DIR}/../normalize.py \
|
||||
--metadata=dump_finetune/dev/raw/metadata.jsonl \
|
||||
--dumpdir=dump_finetune/dev/norm \
|
||||
--stats=dump_finetune/train/feats_stats.npy
|
||||
|
||||
python3 ${BIN_DIR}/../normalize.py \
|
||||
--metadata=dump_finetune/test/raw/metadata.jsonl \
|
||||
--dumpdir=dump_finetune/test/norm \
|
||||
--stats=dump_finetune/train/feats_stats.npy
|
||||
fi
|
||||
|
||||
if [ ${stage} -le 4 ] && [ ${stop_stage} -ge 4 ]; then
|
||||
CUDA_VISIBLE_DEVICES=${gpus} \
|
||||
FLAGS_cudnn_exhaustive_search=true \
|
||||
FLAGS_conv_workspace_size_limit=4000 \
|
||||
python ${BIN_DIR}/train.py \
|
||||
--train-metadata=dump_finetune/train/norm/metadata.jsonl \
|
||||
--dev-metadata=dump_finetune/dev/norm/metadata.jsonl \
|
||||
--config=conf/finetune.yaml \
|
||||
--output-dir=exp/finetune \
|
||||
--ngpu=1
|
||||
fi
|
||||
@ -0,0 +1 @@
|
||||
../voc5/finetune.sh
|
||||
@ -1,55 +0,0 @@
|
||||
#!/bin/bash
|
||||
|
||||
stage=0
|
||||
stop_stage=100
|
||||
|
||||
config_path=$1
|
||||
|
||||
if [ ${stage} -le 0 ] && [ ${stop_stage} -ge 0 ]; then
|
||||
# get durations from MFA's result
|
||||
echo "Generate durations.txt from MFA results ..."
|
||||
python3 ${MAIN_ROOT}/utils/gen_duration_from_textgrid.py \
|
||||
--inputdir=./baker_alignment_tone \
|
||||
--output=durations.txt \
|
||||
--config=${config_path}
|
||||
fi
|
||||
|
||||
if [ ${stage} -le 1 ] && [ ${stop_stage} -ge 1 ]; then
|
||||
# extract features
|
||||
echo "Extract features ..."
|
||||
python3 ${BIN_DIR}/../preprocess.py \
|
||||
--rootdir=~/datasets/BZNSYP/ \
|
||||
--dataset=baker \
|
||||
--dumpdir=dump \
|
||||
--dur-file=durations.txt \
|
||||
--config=${config_path} \
|
||||
--cut-sil=True \
|
||||
--num-cpu=20
|
||||
fi
|
||||
|
||||
if [ ${stage} -le 2 ] && [ ${stop_stage} -ge 2 ]; then
|
||||
# get features' stats(mean and std)
|
||||
echo "Get features' stats ..."
|
||||
python3 ${MAIN_ROOT}/utils/compute_statistics.py \
|
||||
--metadata=dump/train/raw/metadata.jsonl \
|
||||
--field-name="feats"
|
||||
fi
|
||||
|
||||
if [ ${stage} -le 3 ] && [ ${stop_stage} -ge 3 ]; then
|
||||
# normalize, dev and test should use train's stats
|
||||
echo "Normalize ..."
|
||||
|
||||
python3 ${BIN_DIR}/../normalize.py \
|
||||
--metadata=dump/train/raw/metadata.jsonl \
|
||||
--dumpdir=dump/train/norm \
|
||||
--stats=dump/train/feats_stats.npy
|
||||
python3 ${BIN_DIR}/../normalize.py \
|
||||
--metadata=dump/dev/raw/metadata.jsonl \
|
||||
--dumpdir=dump/dev/norm \
|
||||
--stats=dump/train/feats_stats.npy
|
||||
|
||||
python3 ${BIN_DIR}/../normalize.py \
|
||||
--metadata=dump/test/raw/metadata.jsonl \
|
||||
--dumpdir=dump/test/norm \
|
||||
--stats=dump/train/feats_stats.npy
|
||||
fi
|
||||
@ -0,0 +1 @@
|
||||
../../voc1/local/preprocess.sh
|
||||
@ -1,13 +0,0 @@
|
||||
#!/bin/bash
|
||||
|
||||
config_path=$1
|
||||
train_output_path=$2
|
||||
|
||||
FLAGS_cudnn_exhaustive_search=true \
|
||||
FLAGS_conv_workspace_size_limit=4000 \
|
||||
python ${BIN_DIR}/train.py \
|
||||
--train-metadata=dump/train/norm/metadata.jsonl \
|
||||
--dev-metadata=dump/dev/norm/metadata.jsonl \
|
||||
--config=${config_path} \
|
||||
--output-dir=${train_output_path} \
|
||||
--ngpu=1
|
||||
@ -0,0 +1 @@
|
||||
../../voc1/local/train.sh
|
||||
@ -1,55 +0,0 @@
|
||||
#!/bin/bash
|
||||
|
||||
stage=0
|
||||
stop_stage=100
|
||||
|
||||
config_path=$1
|
||||
|
||||
if [ ${stage} -le 0 ] && [ ${stop_stage} -ge 0 ]; then
|
||||
# get durations from MFA's result
|
||||
echo "Generate durations.txt from MFA results ..."
|
||||
python3 ${MAIN_ROOT}/utils/gen_duration_from_textgrid.py \
|
||||
--inputdir=./baker_alignment_tone \
|
||||
--output=durations.txt \
|
||||
--config=${config_path}
|
||||
fi
|
||||
|
||||
if [ ${stage} -le 1 ] && [ ${stop_stage} -ge 1 ]; then
|
||||
# extract features
|
||||
echo "Extract features ..."
|
||||
python3 ${BIN_DIR}/../preprocess.py \
|
||||
--rootdir=~/datasets/BZNSYP/ \
|
||||
--dataset=baker \
|
||||
--dumpdir=dump \
|
||||
--dur-file=durations.txt \
|
||||
--config=${config_path} \
|
||||
--cut-sil=True \
|
||||
--num-cpu=20
|
||||
fi
|
||||
|
||||
if [ ${stage} -le 2 ] && [ ${stop_stage} -ge 2 ]; then
|
||||
# get features' stats(mean and std)
|
||||
echo "Get features' stats ..."
|
||||
python3 ${MAIN_ROOT}/utils/compute_statistics.py \
|
||||
--metadata=dump/train/raw/metadata.jsonl \
|
||||
--field-name="feats"
|
||||
fi
|
||||
|
||||
if [ ${stage} -le 3 ] && [ ${stop_stage} -ge 3 ]; then
|
||||
# normalize, dev and test should use train's stats
|
||||
echo "Normalize ..."
|
||||
|
||||
python3 ${BIN_DIR}/../normalize.py \
|
||||
--metadata=dump/train/raw/metadata.jsonl \
|
||||
--dumpdir=dump/train/norm \
|
||||
--stats=dump/train/feats_stats.npy
|
||||
python3 ${BIN_DIR}/../normalize.py \
|
||||
--metadata=dump/dev/raw/metadata.jsonl \
|
||||
--dumpdir=dump/dev/norm \
|
||||
--stats=dump/train/feats_stats.npy
|
||||
|
||||
python3 ${BIN_DIR}/../normalize.py \
|
||||
--metadata=dump/test/raw/metadata.jsonl \
|
||||
--dumpdir=dump/test/norm \
|
||||
--stats=dump/train/feats_stats.npy
|
||||
fi
|
||||
@ -0,0 +1 @@
|
||||
../../voc1/local/preprocess.sh
|
||||
@ -1,13 +0,0 @@
|
||||
#!/bin/bash
|
||||
|
||||
config_path=$1
|
||||
train_output_path=$2
|
||||
|
||||
FLAGS_cudnn_exhaustive_search=true \
|
||||
FLAGS_conv_workspace_size_limit=4000 \
|
||||
python ${BIN_DIR}/train.py \
|
||||
--train-metadata=dump/train/norm/metadata.jsonl \
|
||||
--dev-metadata=dump/dev/norm/metadata.jsonl \
|
||||
--config=${config_path} \
|
||||
--output-dir=${train_output_path} \
|
||||
--ngpu=1
|
||||
@ -0,0 +1 @@
|
||||
../../voc1/local/train.sh
|
||||
@ -1,55 +0,0 @@
|
||||
#!/bin/bash
|
||||
|
||||
stage=0
|
||||
stop_stage=100
|
||||
|
||||
config_path=$1
|
||||
|
||||
if [ ${stage} -le 0 ] && [ ${stop_stage} -ge 0 ]; then
|
||||
# get durations from MFA's result
|
||||
echo "Generate durations.txt from MFA results ..."
|
||||
python3 ${MAIN_ROOT}/utils/gen_duration_from_textgrid.py \
|
||||
--inputdir=./baker_alignment_tone \
|
||||
--output=durations.txt \
|
||||
--config=${config_path}
|
||||
fi
|
||||
|
||||
if [ ${stage} -le 1 ] && [ ${stop_stage} -ge 1 ]; then
|
||||
# extract features
|
||||
echo "Extract features ..."
|
||||
python3 ${BIN_DIR}/../preprocess.py \
|
||||
--rootdir=~/datasets/BZNSYP/ \
|
||||
--dataset=baker \
|
||||
--dumpdir=dump \
|
||||
--dur-file=durations.txt \
|
||||
--config=${config_path} \
|
||||
--cut-sil=True \
|
||||
--num-cpu=20
|
||||
fi
|
||||
|
||||
if [ ${stage} -le 2 ] && [ ${stop_stage} -ge 2 ]; then
|
||||
# get features' stats(mean and std)
|
||||
echo "Get features' stats ..."
|
||||
python3 ${MAIN_ROOT}/utils/compute_statistics.py \
|
||||
--metadata=dump/train/raw/metadata.jsonl \
|
||||
--field-name="feats"
|
||||
fi
|
||||
|
||||
if [ ${stage} -le 3 ] && [ ${stop_stage} -ge 3 ]; then
|
||||
# normalize, dev and test should use train's stats
|
||||
echo "Normalize ..."
|
||||
|
||||
python3 ${BIN_DIR}/../normalize.py \
|
||||
--metadata=dump/train/raw/metadata.jsonl \
|
||||
--dumpdir=dump/train/norm \
|
||||
--stats=dump/train/feats_stats.npy
|
||||
python3 ${BIN_DIR}/../normalize.py \
|
||||
--metadata=dump/dev/raw/metadata.jsonl \
|
||||
--dumpdir=dump/dev/norm \
|
||||
--stats=dump/train/feats_stats.npy
|
||||
|
||||
python3 ${BIN_DIR}/../normalize.py \
|
||||
--metadata=dump/test/raw/metadata.jsonl \
|
||||
--dumpdir=dump/test/norm \
|
||||
--stats=dump/train/feats_stats.npy
|
||||
fi
|
||||
@ -0,0 +1 @@
|
||||
../../voc1/local/preprocess.sh
|
||||
@ -1,13 +0,0 @@
|
||||
#!/bin/bash
|
||||
|
||||
config_path=$1
|
||||
train_output_path=$2
|
||||
|
||||
FLAGS_cudnn_exhaustive_search=true \
|
||||
FLAGS_conv_workspace_size_limit=4000 \
|
||||
python ${BIN_DIR}/train.py \
|
||||
--train-metadata=dump/train/norm/metadata.jsonl \
|
||||
--dev-metadata=dump/dev/norm/metadata.jsonl \
|
||||
--config=${config_path} \
|
||||
--output-dir=${train_output_path} \
|
||||
--ngpu=1
|
||||
@ -0,0 +1 @@
|
||||
../../voc1/local/train.sh
|
||||
@ -1,13 +0,0 @@
|
||||
#!/bin/bash
|
||||
|
||||
config_path=$1
|
||||
train_output_path=$2
|
||||
|
||||
FLAGS_cudnn_exhaustive_search=true \
|
||||
FLAGS_conv_workspace_size_limit=4000 \
|
||||
python ${BIN_DIR}/train.py \
|
||||
--train-metadata=dump/train/norm/metadata.jsonl \
|
||||
--dev-metadata=dump/dev/norm/metadata.jsonl \
|
||||
--config=${config_path} \
|
||||
--output-dir=${train_output_path} \
|
||||
--ngpu=1
|
||||
@ -0,0 +1 @@
|
||||
../../voc1/local/train.sh
|
||||
@ -1,12 +0,0 @@
|
||||
#!/bin/bash
|
||||
|
||||
config_path=$1
|
||||
train_output_path=$2
|
||||
|
||||
python3 ${BIN_DIR}/train.py \
|
||||
--train-metadata=dump/train/norm/metadata.jsonl \
|
||||
--dev-metadata=dump/dev/norm/metadata.jsonl \
|
||||
--config=${config_path} \
|
||||
--output-dir=${train_output_path} \
|
||||
--ngpu=1 \
|
||||
--phones-dict=dump/phone_id_map.txt
|
||||
@ -0,0 +1 @@
|
||||
../../../csmsc/tts0/local/train.sh
|
||||
@ -1,13 +0,0 @@
|
||||
#!/bin/bash
|
||||
export MAIN_ROOT=`realpath ${PWD}/../../../`
|
||||
|
||||
export PATH=${MAIN_ROOT}:${MAIN_ROOT}/utils:${PATH}
|
||||
export LC_ALL=C
|
||||
|
||||
export PYTHONDONTWRITEBYTECODE=1
|
||||
# Use UTF-8 in Python to avoid UnicodeDecodeError when LC_ALL=C
|
||||
export PYTHONIOENCODING=UTF-8
|
||||
export PYTHONPATH=${MAIN_ROOT}:${PYTHONPATH}
|
||||
|
||||
MODEL=tacotron2
|
||||
export BIN_DIR=${MAIN_ROOT}/paddlespeech/t2s/exps/${MODEL}
|
||||
@ -0,0 +1 @@
|
||||
../../csmsc/tts0/path.sh
|
||||
@ -1,12 +0,0 @@
|
||||
#!/bin/bash
|
||||
|
||||
config_path=$1
|
||||
train_output_path=$2
|
||||
|
||||
python3 ${BIN_DIR}/train.py \
|
||||
--train-metadata=dump/train/norm/metadata.jsonl \
|
||||
--dev-metadata=dump/dev/norm/metadata.jsonl \
|
||||
--config=${config_path} \
|
||||
--output-dir=${train_output_path} \
|
||||
--ngpu=1 \
|
||||
--phones-dict=dump/phone_id_map.txt
|
||||
@ -0,0 +1 @@
|
||||
../../../csmsc/tts3/local/train.sh
|
||||
@ -1,13 +0,0 @@
|
||||
#!/bin/bash
|
||||
export MAIN_ROOT=`realpath ${PWD}/../../../`
|
||||
|
||||
export PATH=${MAIN_ROOT}:${MAIN_ROOT}/utils:${PATH}
|
||||
export LC_ALL=C
|
||||
|
||||
export PYTHONDONTWRITEBYTECODE=1
|
||||
# Use UTF-8 in Python to avoid UnicodeDecodeError when LC_ALL=C
|
||||
export PYTHONIOENCODING=UTF-8
|
||||
export PYTHONPATH=${MAIN_ROOT}:${PYTHONPATH}
|
||||
|
||||
MODEL=fastspeech2
|
||||
export BIN_DIR=${MAIN_ROOT}/paddlespeech/t2s/exps/${MODEL}
|
||||
@ -0,0 +1 @@
|
||||
../../csmsc/tts3/path.sh
|
||||
@ -1,14 +0,0 @@
|
||||
#!/bin/bash
|
||||
|
||||
config_path=$1
|
||||
train_output_path=$2
|
||||
ckpt_name=$3
|
||||
|
||||
FLAGS_allocator_strategy=naive_best_fit \
|
||||
FLAGS_fraction_of_gpu_memory_to_use=0.01 \
|
||||
python3 ${BIN_DIR}/../synthesize.py \
|
||||
--config=${config_path} \
|
||||
--checkpoint=${train_output_path}/checkpoints/${ckpt_name} \
|
||||
--test-metadata=dump/test/norm/metadata.jsonl \
|
||||
--output-dir=${train_output_path}/test \
|
||||
--generator-type=pwgan
|
||||
@ -0,0 +1 @@
|
||||
../../../csmsc/voc1/local/synthesize.sh
|
||||
@ -1,13 +0,0 @@
|
||||
#!/bin/bash
|
||||
|
||||
config_path=$1
|
||||
train_output_path=$2
|
||||
|
||||
FLAGS_cudnn_exhaustive_search=true \
|
||||
FLAGS_conv_workspace_size_limit=4000 \
|
||||
python ${BIN_DIR}/train.py \
|
||||
--train-metadata=dump/train/norm/metadata.jsonl \
|
||||
--dev-metadata=dump/dev/norm/metadata.jsonl \
|
||||
--config=${config_path} \
|
||||
--output-dir=${train_output_path} \
|
||||
--ngpu=1
|
||||
@ -0,0 +1 @@
|
||||
../../../csmsc/voc1/local/train.sh
|
||||
@ -1,13 +0,0 @@
|
||||
#!/bin/bash
|
||||
export MAIN_ROOT=`realpath ${PWD}/../../../`
|
||||
|
||||
export PATH=${MAIN_ROOT}:${MAIN_ROOT}/utils:${PATH}
|
||||
export LC_ALL=C
|
||||
|
||||
export PYTHONDONTWRITEBYTECODE=1
|
||||
# Use UTF-8 in Python to avoid UnicodeDecodeError when LC_ALL=C
|
||||
export PYTHONIOENCODING=UTF-8
|
||||
export PYTHONPATH=${MAIN_ROOT}:${PYTHONPATH}
|
||||
|
||||
MODEL=parallelwave_gan
|
||||
export BIN_DIR=${MAIN_ROOT}/paddlespeech/t2s/exps/gan_vocoder/${MODEL}
|
||||
@ -0,0 +1 @@
|
||||
../../csmsc/voc1/path.sh
|
||||
Some files were not shown because too many files have changed in this diff Show More
Loading…
Reference in new issue