Merge pull request #199 from yt605155624/add_pwg_vctk

add vctk fastspeech2 and pwg

examples/GANVocoder/parallelwave_gan/baker/README.md

@@ -5,7 +5,7 @@ This example contains code used to train a [parallel wavegan](http://arxiv.org/a
 Download CSMSC from the [official website](https://www.data-baker.com/data/index/source) and extract it to `~/datasets`. Then the dataset is in directory `~/datasets/BZNSYP`.
 
 ### Get MFA results for silence trim
-We use [MFA](https://github.com/MontrealCorpusTools/Montreal-Forced-Aligner) results to  cut silence in the edge of audio. (Also, you can set `trim_silence` to `true` in `conf/default.yaml` as an alternative.）  
+We use [MFA](https://github.com/MontrealCorpusTools/Montreal-Forced-Aligner) results to  cut silence in the edge of audio.
 You can download from here [baker_alignment_tone.tar.gz](https://paddlespeech.bj.bcebos.com/MFA/BZNSYP/with_tone/baker_alignment_tone.tar.gz), or train your own MFA model reference to  [use_mfa example](https://github.com/PaddlePaddle/Parakeet/tree/develop/examples/use_mfa) of our repo.
 
 ### Preprocess the dataset

examples/GANVocoder/parallelwave_gan/ljspeech/README.md

@@ -8,7 +8,7 @@ This example contains code used to train a [parallel wavegan](http://arxiv.org/a
 Download LJSpeech-1.1 from the [official website](https://keithito.com/LJ-Speech-Dataset/).
 
 ### Get MFA results for silence trim
-We use [MFA](https://github.com/MontrealCorpusTools/Montreal-Forced-Aligner) results to  cut silence in the edge of audio. (Also, you can set `trim_silence` to `true` in `conf/default.yaml` as an alternative.）  
+We use [MFA](https://github.com/MontrealCorpusTools/Montreal-Forced-Aligner) results to  cut silence in the edge of audio.
 You can download from here [ljspeech_alignment.tar.gz](https://paddlespeech.bj.bcebos.com/MFA/LJSpeech-1.1/ljspeech_alignment.tar.gz), or train your own MFA model reference to  [use_mfa example](https://github.com/PaddlePaddle/Parakeet/tree/develop/examples/use_mfa) of our repo.
 
 ### Preprocess the dataset

examples/GANVocoder/parallelwave_gan/vctk/README.md

@@ -0,0 +1,128 @@
+# Parallel WaveGAN with VCTK
+This example contains code used to train a [parallel wavegan](http://arxiv.org/abs/1910.11480) model with [VCTK](https://datashare.ed.ac.uk/handle/10283/3443).
+## Preprocess the dataset
+### Download and Extract the datasaet
+Download VCTK-0.92  from the [official website](https://datashare.ed.ac.uk/handle/10283/3443) and extract it to `~/datasets`. Then the dataset is in directory `~/datasets/VCTK-Corpus-0.92`.
+
+### Get MFA results for silence trim
+We use [MFA](https://github.com/MontrealCorpusTools/Montreal-Forced-Aligner) results to  cut silence in the edge of audio.
+You can download from here [vctk_alignment.tar.gz](https://paddlespeech.bj.bcebos.com/MFA/VCTK-Corpus-0.92/vctk_alignment.tar.gz), or train your own MFA model reference to  [use_mfa example](https://github.com/PaddlePaddle/Parakeet/tree/develop/examples/use_mfa) of our repo.
+ps: we remove three speakers in VCTK-0.92 (see [reorganize_vctk.py](https://github.com/PaddlePaddle/Parakeet/tree/develop/examples/use_mfa/local/reorganize_vctk.py)):
+1. `p315`, because no txt for it.
+2. `p280` and `p362`, because no *_mic2.flac (which is better than *_mic1.flac) for  them.
+
+### Preprocess the dataset
+Assume the path to the dataset is `~/datasets/VCTK-Corpus-0.92`.
+Assume the path to the MFA result of VCTK is `./vctk_alignment`.
+Run the command below to preprocess the dataset.
+```bash
+./preprocess.sh
+```
+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
+├── test
+│   ├── norm
+│   └── raw
+└── train
+    ├── norm
+    ├── raw
+    └── feats_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 log magnitude of mel spectrogram of each utterances, while the norm folder contains normalized spectrogram. The statistics used to normalize the spectrogram is computed from the training set, which is located in `dump/train/feats_stats.npy`.
+
+Also there is a `metadata.jsonl` in each subfolder. It is a table-like file which contains id and paths to spectrogam of each utterance.
+
+## Train the model
+
+`./run.sh` calls `../train.py`.
+```bash
+./run.sh
+```
+Here's the complete help message.
+
+```text
+usage: train.py [-h] [--config CONFIG] [--train-metadata TRAIN_METADATA]
+                [--dev-metadata DEV_METADATA] [--output-dir OUTPUT_DIR]
+                [--device DEVICE] [--nprocs NPROCS] [--verbose VERBOSE]
+                [--batch-size BATCH_SIZE] [--max-iter MAX_ITER]
+                [--run-benchmark RUN_BENCHMARK]
+                [--profiler_options PROFILER_OPTIONS]
+
+Train a ParallelWaveGAN model.
+
+optional arguments:
+  -h, --help            show this help message and exit
+  --config CONFIG       config file to overwrite default config.
+  --train-metadata TRAIN_METADATA
+                        training data.
+  --dev-metadata DEV_METADATA
+                        dev data.
+  --output-dir OUTPUT_DIR
+                        output dir.
+  --device DEVICE       device type to use.
+  --nprocs NPROCS       number of processes.
+  --verbose VERBOSE     verbose.
+
+benchmark:
+  arguments related to benchmark.
+
+  --batch-size BATCH_SIZE
+                        batch size.
+  --max-iter MAX_ITER   train max steps.
+  --run-benchmark RUN_BENCHMARK
+                        runing benchmark or not, if True, use the --batch-size
+                        and --max-iter.
+  --profiler_options PROFILER_OPTIONS
+                        The option of profiler, which should be in format
+                        "key1=value1;key2=value2;key3=value3".
+```
+
+1. `--config` is a config file in yaml format to overwrite the default config, which can be found at `conf/default.yaml`.
+2. `--train-metadata` and `--dev-metadata` should be the metadata file in the normalized subfolder of `train` and `dev` in the `dump` folder.
+3. `--output-dir` is the directory to save the results of the experiment. Checkpoints are save in `checkpoints/` inside this directory.
+4. `--device` is the type of the device to run the experiment, 'cpu' or 'gpu' are supported.
+5. `--nprocs` is the number of processes to run in parallel, note that nprocs > 1 is only supported when `--device` is 'gpu'.
+
+## Pretrained Models
+
+
+## Synthesize
+
+`synthesize.sh` calls `../synthesize.py `, which can synthesize waveform from `metadata.jsonl`.
+```bash
+./synthesize.sh
+```
+```text
+usage: synthesize.py [-h] [--config CONFIG] [--checkpoint CHECKPOINT]
+                     [--test-metadata TEST_METADATA] [--output-dir OUTPUT_DIR]
+                     [--device DEVICE] [--verbose VERBOSE]
+
+Synthesize with parallel wavegan.
+
+optional arguments:
+  -h, --help            show this help message and exit
+  --config CONFIG       parallel wavegan config file.
+  --checkpoint CHECKPOINT
+                        snapshot to load.
+  --test-metadata TEST_METADATA
+                        dev data.
+  --output-dir OUTPUT_DIR
+                        output dir.
+  --device DEVICE       device to run.
+  --verbose VERBOSE     verbose.
+```
+
+1. `--config` parallel wavegan config file. You should use the same config with which the model is trained.
+2. `--checkpoint` is the checkpoint to load. Pick one of the checkpoints from `checkpoints` inside the training output directory. If you use the pretrained model, use the `pwg_snapshot_iter_400000.pdz`.
+3. `--test-metadata` is the metadata of the test dataset. Use the `metadata.jsonl` in the `dev/norm` subfolder from the processed directory.
+4. `--output-dir` is the directory to save the synthesized audio files.
+5. `--device` is the type of device to run synthesis, 'cpu' and 'gpu' are supported.
+
+## Acknowledgement
+We adapted some code from https://github.com/kan-bayashi/ParallelWaveGAN.

examples/GANVocoder/parallelwave_gan/vctk/conf/default.yaml

@@ -0,0 +1,115 @@
+# This is the hyperparameter configuration file for Parallel WaveGAN.
+# Please make sure this is adjusted for the VCTK corpus. If you want to
+# apply to the other dataset, you might need to carefully change some parameters.
+# This configuration requires 12 GB GPU memory and takes ~3 days on RTX TITAN.
+
+###########################################################
+#                FEATURE EXTRACTION SETTING               #
+###########################################################
+fs: 24000                # Sampling rate.
+n_fft: 2048              # FFT size. (in samples)
+n_shift: 300             # Hop size. (in samples)
+win_length: 1200         # Window length. (in samples)
+                         # If set to null, it will be the same as fft_size.
+window: "hann"           # Window function.
+n_mels: 80               # Number of mel basis.
+fmin: 80                 # Minimum freq in mel basis calculation. (Hz)
+fmax: 7600               # Maximum frequency in mel basis calculation. (Hz)
+
+###########################################################
+#         GENERATOR NETWORK ARCHITECTURE SETTING          #
+###########################################################
+generator_params:
+    in_channels: 1        # Number of input channels.
+    out_channels: 1       # Number of output channels.
+    kernel_size: 3        # Kernel size of dilated convolution.
+    layers: 30            # Number of residual block layers.
+    stacks: 3             # Number of stacks i.e., dilation cycles.
+    residual_channels: 64 # Number of channels in residual conv.
+    gate_channels: 128    # Number of channels in gated conv.
+    skip_channels: 64     # Number of channels in skip conv.
+    aux_channels: 80      # Number of channels for auxiliary feature conv.
+                          # Must be the same as num_mels.
+    aux_context_window: 2 # Context window size for auxiliary feature.
+                          # If set to 2, previous 2 and future 2 frames will be considered.
+    dropout: 0.0          # Dropout rate. 0.0 means no dropout applied.
+    use_weight_norm: true # Whether to use weight norm.
+                          # If set to true, it will be applied to all of the conv layers.
+    upsample_scales: [4, 5, 3, 5]     # Upsampling scales. Prodcut of these must be the same as hop size.
+
+###########################################################
+#       DISCRIMINATOR NETWORK ARCHITECTURE SETTING        #
+###########################################################
+discriminator_params:
+    in_channels: 1        # Number of input channels.
+    out_channels: 1       # Number of output channels.
+    kernel_size: 3        # Number of output channels.
+    layers: 10            # Number of conv layers.
+    conv_channels: 64     # Number of chnn layers.
+    bias: true            # Whether to use bias parameter in conv.
+    use_weight_norm: true # Whether to use weight norm.
+                          # If set to true, it will be applied to all of the conv layers.
+    nonlinear_activation: "LeakyReLU" # Nonlinear function after each conv.
+    nonlinear_activation_params:      # Nonlinear function parameters
+        negative_slope: 0.2           # Alpha in LeakyReLU.
+
+###########################################################
+#                   STFT LOSS SETTING                     #
+###########################################################
+stft_loss_params:
+    fft_sizes: [1024, 2048, 512]  # List of FFT size for STFT-based loss.
+    hop_sizes: [120, 240, 50]     # List of hop size for STFT-based loss
+    win_lengths: [600, 1200, 240] # List of window length for STFT-based loss.
+    window: "hann"                # Window function for STFT-based loss
+
+###########################################################
+#               ADVERSARIAL LOSS SETTING                  #
+###########################################################
+lambda_adv: 4.0  # Loss balancing coefficient.
+
+###########################################################
+#                  DATA LOADER SETTING                    #
+###########################################################
+batch_size: 8              # Batch size.
+batch_max_steps: 24000     # Length of each audio in batch. Make sure dividable by hop_size.
+pin_memory: true           # Whether to pin memory in Pytorch DataLoader.
+num_workers: 4             # Number of workers in Pytorch DataLoader.
+remove_short_samples: true # Whether to remove samples the length of which are less than batch_max_steps.
+allow_cache: true          # Whether to allow cache in dataset. If true, it requires cpu memory.
+
+###########################################################
+#             OPTIMIZER & SCHEDULER SETTING               #
+###########################################################
+generator_optimizer_params:
+    epsilon: 1.0e-6            # Generator's epsilon.
+    weight_decay: 0.0      # Generator's weight decay coefficient.
+generator_scheduler_params:
+    learning_rate: 0.0001             # Generator's learning rate.
+    step_size: 200000      # Generator's scheduler step size.
+    gamma: 0.5             # Generator's scheduler gamma.
+                           # At each step size, lr will be multiplied by this parameter.
+generator_grad_norm: 10    # Generator's gradient norm.
+discriminator_optimizer_params:
+    epsilon: 1.0e-6            # Discriminator's epsilon.
+    weight_decay: 0.0      # Discriminator's weight decay coefficient.
+discriminator_scheduler_params:
+    learning_rate: 0.00005            # Discriminator's learning rate. 
+    step_size: 200000      # Discriminator's scheduler step size.
+    gamma: 0.5             # Discriminator's scheduler gamma.
+                           # At each step size, lr will be multiplied by this parameter.
+discriminator_grad_norm: 1 # Discriminator's gradient norm.
+
+###########################################################
+#                    INTERVAL SETTING                     #
+###########################################################
+discriminator_train_start_steps: 100000 # Number of steps to start to train discriminator.
+train_max_steps: 1000000                # Number of training steps.
+save_interval_steps: 5000               # Interval steps to save checkpoint.
+eval_interval_steps: 1000               # Interval steps to evaluate the network.
+
+###########################################################
+#                     OTHER SETTING                       #
+###########################################################
+num_save_intermediate_results: 4  # Number of results to be saved as intermediate results.
+num_snapshots: 10                 # max number of snapshots to keep while training
+seed: 42                          # random seed for paddle, random, and np.random

examples/GANVocoder/parallelwave_gan/vctk/preprocess.sh

@@ -0,0 +1,54 @@
+#!/bin/bash
+
+stage=0
+stop_stage=100
+
+export MAIN_ROOT=`realpath ${PWD}/../../../../`
+
+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=./vctk_alignment \
+        --output=durations.txt \
+        --config=conf/default.yaml
+fi
+
+if [ ${stage} -le 1 ] && [ ${stop_stage} -ge 1 ]; then
+    echo "Extract features ..."
+    python3 ../../preprocess.py \
+        --rootdir=~/datasets/VCTK-Corpus-0.92/ \
+        --dataset=vctk \
+        --dumpdir=dump \
+        --dur-file=durations.txt \
+        --config=conf/default.yaml \
+        --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 ../../normalize.py \
+        --metadata=dump/train/raw/metadata.jsonl \
+        --dumpdir=dump/train/norm \
+        --stats=dump/train/feats_stats.npy
+    python3 ../../normalize.py \
+        --metadata=dump/dev/raw/metadata.jsonl \
+        --dumpdir=dump/dev/norm \
+        --stats=dump/train/feats_stats.npy
+
+    python3 ../../normalize.py \
+        --metadata=dump/test/raw/metadata.jsonl \
+        --dumpdir=dump/test/norm \
+        --stats=dump/train/feats_stats.npy
+fi

examples/GANVocoder/parallelwave_gan/vctk/run.sh

@@ -0,0 +1,10 @@
+#!/bin/bash
+
+FLAGS_cudnn_exhaustive_search=true \
+FLAGS_conv_workspace_size_limit=4000 \
+python ../train.py \
+    --train-metadata=dump/train/norm/metadata.jsonl \
+    --dev-metadata=dump/dev/norm/metadata.jsonl \
+    --config=conf/default.yaml \
+    --output-dir=exp/default \
+    --nprocs=1

examples/GANVocoder/parallelwave_gan/vctk/synthesize.sh

@@ -0,0 +1,7 @@
+#!/bin/bash
+
+python3 ../synthesize.py \
+  --config=conf/default.yaml \
+  --checkpoint=exp/default/checkpoints/snapshot_iter_35000.pdz_bak\
+  --test-metadata=dump/test/norm/metadata.jsonl \
+  --output-dir=exp/default/test