Tacotron2 retrain (#4103)

* fix yaml Signed-off-by: treacker <[email protected]> * Fix for new TTSDataset class Signed-off-by: treacker <[email protected]> * added wandb logging Signed-off-by: treacker <[email protected]> * added wandb logging Signed-off-by: treacker <[email protected]> * fix numpy version Signed-off-by: treacker <[email protected]> * fix numpy version Signed-off-by: treacker <[email protected]> * inference fix Signed-off-by: treacker <[email protected]> * removed old code Signed-off-by: treacker <[email protected]> * updated parser logic Signed-off-by: treacker <[email protected]> * reverted version update Signed-off-by: treacker <[email protected]> * refactored parser logic Signed-off-by: treacker <[email protected]> * Updated Jenkinsfile Signed-off-by: treacker <[email protected]> * Refactored tutorial for Tacotron2 Signed-off-by: treacker <[email protected]> * Made backward compatibility Signed-off-by: treacker <[email protected]> * Made backward compatibility Signed-off-by: treacker <[email protected]> * Update Jenkinsfile Signed-off-by: treacker <[email protected]> * Update tacotron.yaml Signed-off-by: treacker <[email protected]> * Refactoring Signed-off-by: treacker <[email protected]> * cleaned up TN/ ITN doc (#4119) * cleaned up TN/ ITN doc Signed-off-by: Yang Zhang <[email protected]> * fix typo Signed-off-by: Yang Zhang <[email protected]> * fix image Signed-off-by: Yang Zhang <[email protected]> * fix image Signed-off-by: Yang Zhang <[email protected]> Signed-off-by: treacker <[email protected]> * Check implicit grad acc in GLUE dataset building (#4123) * Check implicit grad acc in GLUE dataset building Signed-off-by: MaximumEntropy <[email protected]> * Fix jenkins test for GLUE/XNLI Signed-off-by: MaximumEntropy <[email protected]> Signed-off-by: treacker <[email protected]> * Refactoring Signed-off-by: treacker <[email protected]> * Refactoring Signed-off-by: treacker <[email protected]> * Fixed jenkins Signed-off-by: treacker <[email protected]> * Refactoring Signed-off-by: treacker <[email protected]> * Refactoring Signed-off-by: treacker <[email protected]> * Refactoring Signed-off-by: treacker <[email protected]> Co-authored-by: Yang Zhang <[email protected]> Co-authored-by: Sandeep Subramanian <[email protected]>
NVIDIA · May 11, 2022 · 30db4d4 · 30db4d4
1 parent df33239
commit 30db4d4
Show file tree

Hide file tree

Showing 5 changed files with 640 additions and 455 deletions.
diff --git a/Jenkinsfile b/Jenkinsfile
@@ -2983,7 +2983,6 @@ pipeline {
         }
       }
       parallel {
-        // TODO(Oktai15): update it in 1.8.0 version
         stage('Tacotron 2') {
           steps {
             sh 'python examples/tts/tacotron2.py \
@@ -2993,13 +2992,18 @@ pipeline {
             trainer.accelerator="gpu" \
             +trainer.limit_train_batches=1 +trainer.limit_val_batches=1 trainer.max_epochs=1 \
             trainer.strategy=null \
-            model.train_ds.dataloader_params.batch_size=4 \
-            model.validation_ds.dataloader_params.batch_size=4 \
             model.decoder.decoder_rnn_dim=256 \
             model.decoder.attention_rnn_dim=1024 \
             model.decoder.prenet_dim=128 \
             model.postnet.postnet_n_convolutions=3 \
-            ~trainer.check_val_every_n_epoch'
+            model.train_ds.dataloader_params.batch_size=4 \
+            model.train_ds.dataloader_params.num_workers=1 \
+            model.validation_ds.dataloader_params.batch_size=4 \
+            model.validation_ds.dataloader_params.num_workers=1 \
+            ~model.text_normalizer \
+            ~model.text_normalizer_call_kwargs \
+            ~trainer.check_val_every_n_epoch \
+            '
           }
         }
         stage('WaveGlow') {

diff --git a/examples/tts/conf/tacotron2.yaml b/examples/tts/conf/tacotron2.yaml
@@ -1,81 +1,136 @@
-# TODO(Oktai15): update this config in 1.8.0 version
+# This config contains the default values for training Tacotron2 model on LJSpeech dataset.
+# If you want to train model on other dataset, you can change config values according to your dataset.
+# Most dataset-specific arguments are in the head of the config file, see below.
 
 name: Tacotron2
-sample_rate: 22050
-# <PAD>, <BOS>, <EOS> will be added by the tacotron2.py script
-labels: [' ', '!', '"', "'", '(', ')', ',', '-', '.', ':', ';', '?', 'A', 'B', 'C', 'D', 'E', 'F', 'G', 'H',
-        'I', 'J', 'K', 'L', 'M', 'N', 'O', 'P', 'Q', 'R', 'S', 'T', 'U', 'V', 'W', 'X', 'Y', 'Z', '[', ']',
-        'a', 'b', 'c', 'd', 'e', 'f', 'g', 'h', 'i', 'j', 'k', 'l', 'm', 'n', 'o', 'p', 'q', 'r', 's', 't',
-        'u', 'v', 'w', 'x', 'y', 'z']
-n_fft: 1024
-n_mels: 80
-fmax: 8000
-n_stride: 256
-pad_value: -11.52
+
 train_dataset: ???
 validation_datasets: ???
+sup_data_path: null
+sup_data_types: null
+
+phoneme_dict_path: "scripts/tts_dataset_files/cmudict-0.7b_nv22.01"
+heteronyms_path: "scripts/tts_dataset_files/heteronyms-030921"
+whitelist_path: "nemo_text_processing/text_normalization/en/data/whitelist_lj_speech.tsv"
+
+
 
 model:
-  labels: ${labels}
+  pitch_fmin: 65.40639132514966
+  pitch_fmax: 2093.004522404789
+
+  sample_rate: 22050
+  n_mel_channels: 80
+  n_window_size: 1024
+  n_window_stride: 256
+  n_fft: 1024
+  lowfreq: 0
+  highfreq: 8000
+  window: hann
+  pad_value: -11.52
+
+
+  text_normalizer:
+    _target_: nemo_text_processing.text_normalization.normalize.Normalizer
+    lang: en
+    input_case: cased
+    whitelist: ${whitelist_path}
+
+  text_normalizer_call_kwargs:
+    verbose: false
+    punct_pre_process: true
+    punct_post_process: true
+
+  text_tokenizer:
+    _target_: nemo.collections.tts.torch.tts_tokenizers.EnglishPhonemesTokenizer
+    punct: true
+    stresses: true
+    chars: true
+    apostrophe: true
+    pad_with_space: true
+    g2p:
+      _target_: nemo.collections.tts.torch.g2ps.EnglishG2p
+      phoneme_dict: ${phoneme_dict_path}
+      heteronyms: ${heteronyms_path}
+
   train_ds:
     dataset:
-      _target_: "nemo.collections.asr.data.audio_to_text.AudioToCharDataset"
+      _target_: "nemo.collections.tts.torch.data.TTSDataset"
       manifest_filepath: ${train_dataset}
+      sample_rate: ${model.sample_rate}
+      sup_data_path: ${sup_data_path}
+      sup_data_types: ${sup_data_types}
+      n_fft: ${model.n_fft}
+      win_length: ${model.n_window_size}
+      hop_length: ${model.n_window_stride}
+      window: ${model.window}
+      n_mels: ${model.n_mel_channels}
+      lowfreq: ${model.lowfreq}
+      highfreq: ${model.highfreq}
       max_duration: null
       min_duration: 0.1
-      trim: false
-      int_values: false
-      normalize: true
-      sample_rate: ${sample_rate}
-      # bos_id: 66
-      # eos_id: 67
-      # pad_id: 68  These parameters are added automatically in Tacotron2
+      ignore_file: null
+      trim: False
+      pitch_fmin: ${model.pitch_fmin}
+      pitch_fmax: ${model.pitch_fmax}
     dataloader_params:
       drop_last: false
       shuffle: true
       batch_size: 48
       num_workers: 4
-
-
+      pin_memory: false
+  
   validation_ds:
     dataset:
-      _target_: "nemo.collections.asr.data.audio_to_text.AudioToCharDataset"
-      manifest_filepath: ${validation_datasets}
+      _target_: "nemo.collections.tts.torch.data.TTSDataset"
+      manifest_filepath: ${train_dataset}
+      sample_rate: ${model.sample_rate}
+      sup_data_path: ${sup_data_path}
+      sup_data_types: ${sup_data_types}
+      n_fft: ${model.n_fft}
+      win_length: ${model.n_window_size}
+      hop_length: ${model.n_window_stride}
+      window: ${model.window}
+      n_mels: ${model.n_mel_channels}
+      lowfreq: ${model.lowfreq}
+      highfreq: ${model.highfreq}
       max_duration: null
       min_duration: 0.1
-      int_values: false
-      normalize: true
-      sample_rate: ${sample_rate}
-      trim: false
-      # bos_id: 66
-      # eos_id: 67
-      # pad_id: 68  These parameters are added automatically in Tacotron2
+      ignore_file: null
+      trim: False
+      pitch_fmin: ${model.pitch_fmin}
+      pitch_fmax: ${model.pitch_fmax}
     dataloader_params:
       drop_last: false
       shuffle: false
-      batch_size: 48
+      batch_size: 24
       num_workers: 8
+      pin_memory: false
 
   preprocessor:
     _target_: nemo.collections.asr.parts.preprocessing.features.FilterbankFeatures
-    dither: 0.0
-    nfilt: ${n_mels}
-    frame_splicing: 1
-    highfreq: ${fmax}
+    nfilt: ${model.n_mel_channels}
+    highfreq: ${model.highfreq}
     log: true
     log_zero_guard_type: clamp
     log_zero_guard_value: 1e-05
-    lowfreq: 0
-    mag_power: 1.0
-    n_fft: ${n_fft}
-    n_window_size: 1024
-    n_window_stride: ${n_stride}
-    normalize: null
+    lowfreq: ${model.lowfreq}
+    n_fft: ${model.n_fft}
+    n_window_size: ${model.n_window_size}
+    n_window_stride: ${model.n_window_stride}
     pad_to: 16
-    pad_value: ${pad_value}
+    pad_value: ${model.pad_value}
+    sample_rate: ${model.sample_rate}
+    window: ${model.window}
+    normalize: null
     preemph: null
-    sample_rate: ${sample_rate}
-    window: hann
+    dither: 0.0
+    frame_splicing: 1
+    stft_conv: false
+    nb_augmentation_prob : 0
+    mag_power: 1.0
+    exact_pad: true
+    use_grads: false
 
   encoder:
     _target_: nemo.collections.tts.modules.tacotron2.Encoder
@@ -90,7 +145,7 @@ model:
     gate_threshold: 0.5
     max_decoder_steps: 1000
     n_frames_per_step: 1  # currently only 1 is supported
-    n_mel_channels: ${n_mels}
+    n_mel_channels: ${model.n_mel_channels}
     p_attention_dropout: 0.1
     p_decoder_dropout: 0.1
     prenet_dim: 256
@@ -105,7 +160,7 @@ model:
 
   postnet:
     _target_: nemo.collections.tts.modules.tacotron2.Postnet
-    n_mel_channels: ${n_mels}
+    n_mel_channels: ${model.n_mel_channels}
     p_dropout: 0.5
     postnet_embedding_dim: 512
     postnet_kernel_size: 5
@@ -132,11 +187,15 @@ trainer:
   enable_checkpointing: False  # Provided by exp_manager
   logger: False  # Provided by exp_manager
   gradient_clip_val: 1.0
-  log_every_n_steps: 200
-  check_val_every_n_epoch: 25
+  log_every_n_steps: 60
+  check_val_every_n_epoch: 2
+
 
 exp_manager:
   exp_dir: null
   name: ${name}
-  create_tensorboard_logger: True
-  create_checkpoint_callback: True
+  create_tensorboard_logger: true
+  create_checkpoint_callback: true
+  checkpoint_callback_params:
+    monitor: val_loss
+    mode: min
diff --git a/nemo/collections/tts/helpers/helpers.py b/nemo/collections/tts/helpers/helpers.py
@@ -56,6 +56,12 @@
 
 from nemo.utils import logging
 
+HAVE_WANDB = True
+try:
+    import wandb
+except ModuleNotFoundError:
+    HAVE_WANDB = False
+
 try:
     from pytorch_lightning.utilities import rank_zero_only
 except ModuleNotFoundError:
@@ -284,6 +290,7 @@ def tacotron2_log_to_tb_func(
             step,
             dataformats="HWC",
         )
+
         if add_audio:
             filterbank = librosa.filters.mel(sr=sr, n_fft=n_fft, n_mels=n_mels, fmax=fmax)
             log_mel = mel_postnet[0].data.cpu().numpy().T
@@ -299,6 +306,67 @@ def tacotron2_log_to_tb_func(
             swriter.add_audio(f"audio/{tag}_target", audio / max(np.abs(audio)), step, sample_rate=sr)
 
 
+def tacotron2_log_to_wandb_func(
+    swriter,
+    tensors,
+    step,
+    tag="train",
+    log_images=False,
+    log_images_freq=1,
+    add_audio=True,
+    griffin_lim_mag_scale=1024,
+    griffin_lim_power=1.2,
+    sr=22050,
+    n_fft=1024,
+    n_mels=80,
+    fmax=8000,
+):
+    _, spec_target, mel_postnet, gate, gate_target, alignments = tensors
+    if not HAVE_WANDB:
+        return
+    if log_images and step % log_images_freq == 0:
+        alignments = []
+        specs = []
+        gates = []
+        alignments += [
+            wandb.Image(plot_alignment_to_numpy(alignments[0].data.cpu().numpy().T), caption=f"{tag}_alignment",)
+        ]
+        alignments += [
+            wandb.Image(plot_spectrogram_to_numpy(spec_target[0].data.cpu().numpy()), caption=f"{tag}_mel_target",),
+            wandb.Image(plot_spectrogram_to_numpy(mel_postnet[0].data.cpu().numpy()), caption=f"{tag}_mel_predicted",),
+        ]
+        gates += [
+            wandb.Image(
+                plot_gate_outputs_to_numpy(
+                    gate_target[0].data.cpu().numpy(), torch.sigmoid(gate[0]).data.cpu().numpy(),
+                ),
+                caption=f"{tag}_gate",
+            )
+        ]
+
+        swriter.log({"specs": specs, "alignments": alignments, "gates": gates})
+
+        if add_audio:
+            audios = []
+            filterbank = librosa.filters.mel(sr=sr, n_fft=n_fft, n_mels=n_mels, fmax=fmax)
+            log_mel = mel_postnet[0].data.cpu().numpy().T
+            mel = np.exp(log_mel)
+            magnitude = np.dot(mel, filterbank) * griffin_lim_mag_scale
+            audio_pred = griffin_lim(magnitude.T ** griffin_lim_power)
+
+            log_mel = spec_target[0].data.cpu().numpy().T
+            mel = np.exp(log_mel)
+            magnitude = np.dot(mel, filterbank) * griffin_lim_mag_scale
+            audio_true = griffin_lim(magnitude.T ** griffin_lim_power)
+
+            audios += [
+                wandb.Audio(audio_true / max(np.abs(audio_true)), caption=f"{tag}_wav_target", sample_rate=sr,),
+                wandb.Audio(audio_pred / max(np.abs(audio_pred)), caption=f"{tag}_wav_predicted", sample_rate=sr,),
+            ]
+
+            swriter.log({"audios": audios})
+
+
 def plot_alignment_to_numpy(alignment, info=None):
     fig, ax = plt.subplots(figsize=(6, 4))
     im = ax.imshow(alignment, aspect='auto', origin='lower', interpolation='none')