Update docs

2026-09-09 11:16:00 +08:00 · 2021-09-30 14:47:56 +00:00
parent f904dd4828
commit 6d3b2d3cdd
3 changed files with 310 additions and 84 deletions
@@ -23,63 +23,104 @@ each line.

 ### Pure Python Way

-```python
-import os
+1. Define `train.py`.

-# GlowTTSConfig: all model related values for training, validating and testing.
-from TTS.tts.configs import GlowTTSConfig
+    ```python
+    import os

-# BaseDatasetConfig: defines name, formatter and path of the dataset.
-from TTS.tts.configs import BaseDatasetConfig
+    # GlowTTSConfig: all model related values for training, validating and testing.
+    from TTS.tts.configs import GlowTTSConfig

-# init_training: Initialize and setup the training environment.
-# Trainer: Where the ✨️ happens.
-# TrainingArgs: Defines the set of arguments of the Trainer.
-from TTS.trainer import init_training, Trainer, TrainingArgs
+    # BaseDatasetConfig: defines name, formatter and path of the dataset.
+    from TTS.tts.configs import BaseDatasetConfig

-# we use the same path as this script as our training folder.
-output_path = os.path.dirname(os.path.abspath(__file__))
+    # init_training: Initialize and setup the training environment.
+    # Trainer: Where the ✨️ happens.
+    # TrainingArgs: Defines the set of arguments of the Trainer.
+    from TTS.trainer import init_training, Trainer, TrainingArgs

-# set LJSpeech as our target dataset and define its path so that the Trainer knows what data formatter it needs.
-dataset_config = BaseDatasetConfig(name="ljspeech", meta_file_train="metadata.csv", path=os.path.join(output_path, "../LJSpeech-1.1/"))
+    # we use the same path as this script as our training folder.
+    output_path = os.path.dirname(os.path.abspath(__file__))

-# Configure the model. Every config class inherits the BaseTTSConfig to have all the fields defined for the Trainer.
-config = GlowTTSConfig(
-    batch_size=32,
-    eval_batch_size=16,
-    num_loader_workers=4,
-    num_eval_loader_workers=4,
-    run_eval=True,
-    test_delay_epochs=-1,
-    epochs=1000,
-    text_cleaner="english_cleaners",
-    use_phonemes=False,
-    phoneme_language="en-us",
-    phoneme_cache_path=os.path.join(output_path, "phoneme_cache"),
-    print_step=25,
-    print_eval=True,
-    mixed_precision=False,
-    output_path=output_path,
-    datasets=[dataset_config]
-)
+    # set LJSpeech as our target dataset and define its path so that the Trainer knows what data formatter it needs.
+    dataset_config = BaseDatasetConfig(name="ljspeech", meta_file_train="metadata.csv", path=os.path.join(output_path, "../LJSpeech-1.1/"))

-# Take the config and the default Trainer arguments, setup the training environment and override the existing
-# config values from the terminal. So you can do the following.
-# >>> python train.py --coqpit.batch_size 128
-args, config, output_path, _, _, _= init_training(TrainingArgs(), config)
+    # Configure the model. Every config class inherits the BaseTTSConfig to have all the fields defined for the Trainer.
+    config = GlowTTSConfig(
+        batch_size=32,
+        eval_batch_size=16,
+        num_loader_workers=4,
+        num_eval_loader_workers=4,
+        run_eval=True,
+        test_delay_epochs=-1,
+        epochs=1000,
+        text_cleaner="english_cleaners",
+        use_phonemes=False,
+        phoneme_language="en-us",
+        phoneme_cache_path=os.path.join(output_path, "phoneme_cache"),
+        print_step=25,
+        print_eval=True,
+        mixed_precision=False,
+        output_path=output_path,
+        datasets=[dataset_config]
+    )

-# Initiate the Trainer.
-# Trainer provides a generic API to train all the 🐸TTS models with all its perks like mixed-precision training,
-# distributed training etc.
-trainer = Trainer(args, config, output_path)
+    # initialize the audio processor used for feature extraction and audio I/O.
+    # It is mainly used by the dataloader and the training loggers.
+    ap = AudioProcessor(**config.audio.to_dict())

-# And kick it 🚀
-trainer.fit()
-```
+    # load a list of training samples
+    # Each sample is a list of ```[text, audio_file_path, speaker_name]```
+    train_samples, eval_samples = load_tts_samples(dataset_config, eval_split=True)
+
+    # initialize the model
+    # Models only takes the config object as input.
+    model = GlowTTS(config)
+
+    # Initiate the Trainer.
+    # Trainer provides a generic API to train all the 🐸TTS models with all its perks like mixed-precision training,
+    # distributed training etc.
+    trainer = Trainer(
+        TrainingArgs(),
+        config,
+        output_path,
+        model=model,
+        train_samples=train_samples,
+        eval_samples=eval_samples,
+        training_assets={"audio_processor": ap},
+    )
+
+    # And kick it 🚀
+    trainer.fit()
+    ```
+
+2. Run the script.
+
+    ```bash
+    CUDA_VISIBLE_DEVICES=0 python train.py
+    ```
+
+    - Continue a previous run.
+
+        ```bash
+        CUDA_VISIBLE_DEVICES=0 python train.py --continue_path path/to/previous/run/folder/
+        ```
+
+    - Fine-tune a model.
+
+        ```bash
+        CUDA_VISIBLE_DEVICES=0 python train.py --restore_path path/to/model/checkpoint.pth.tar
+        ```
+
+    - Run multi-gpu training.
+
+        ```bash
+        CUDA_VISIBLE_DEVICES=0,1,2 python TTS/bin/distribute.py --script train.py
+        ```

 ### CLI Way

-We still support running training from CLI like in the old days. The same training can be started as follows.
+We still support running training from CLI like in the old days. The same training run can also be started as follows.

 1. Define your `config.json`

@@ -111,45 +152,63 @@ We still support running training from CLI like in the old days. The same traini
    $ CUDA_VISIBLE_DEVICES="0" python TTS/bin/train_tts.py --config_path config.json
    ```

-
-
 ## Training a `vocoder` Model

 ```python
 import os

+from TTS.trainer import Trainer, TrainingArgs
+from TTS.utils.audio import AudioProcessor
 from TTS.vocoder.configs import HifiganConfig
-from TTS.trainer import init_training, Trainer, TrainingArgs
-
+from TTS.vocoder.datasets.preprocess import load_wav_data
+from TTS.vocoder.models.gan import GAN

 output_path = os.path.dirname(os.path.abspath(__file__))
+
 config = HifiganConfig(
    batch_size=32,
    eval_batch_size=16,
    num_loader_workers=4,
    num_eval_loader_workers=4,
    run_eval=True,
-    test_delay_epochs=-1,
+    test_delay_epochs=5,
    epochs=1000,
    seq_len=8192,
    pad_short=2000,
    use_noise_augment=True,
    eval_split_size=10,
    print_step=25,
-    print_eval=True,
+    print_eval=False,
    mixed_precision=False,
    lr_gen=1e-4,
    lr_disc=1e-4,
-    # `vocoder` only needs a data path and they read recursively all the `.wav` files underneath.
    data_path=os.path.join(output_path, "../LJSpeech-1.1/wavs/"),
    output_path=output_path,
 )
-args, config, output_path, _, c_logger, tb_logger = init_training(TrainingArgs(), config)
-trainer = Trainer(args, config, output_path, c_logger, tb_logger)
+
+# init audio processor
+ap = AudioProcessor(**config.audio.to_dict())
+
+# load training samples
+eval_samples, train_samples = load_wav_data(config.data_path, config.eval_split_size)
+
+# init model
+model = GAN(config)
+
+# init the trainer and 🚀
+trainer = Trainer(
+    TrainingArgs(),
+    config,
+    output_path,
+    model=model,
+    train_samples=train_samples,
+    eval_samples=eval_samples,
+    training_assets={"audio_processor": ap},
+)
 trainer.fit()
 ```

-❗️ Note that you can also start the training run from CLI as the `tts` model above.
+❗️ Note that you can also use ```train_vocoder.py``` as the ```tts``` models above.

 ## Synthesizing Speech