Change organization of code

app.py

@@ -1,241 +1,20 @@
-import os
 import gradio as gr
-from pathlib import Path
 
-os.system("python3 -m pip install -e .")
-
-import functools
-import os
-
-import numpy as np
-import tensorflow.compat.v2 as tf
-from pydub import AudioSegment
-
-import functools
-import gin
-import jax
-import librosa
 import note_seq
-import seqio
-import t5
-import t5x
-
-from mt3 import metrics_utils
-from mt3 import models
-from mt3 import network
-from mt3 import note_sequences
-from mt3 import preprocessors
-from mt3 import spectrograms
-from mt3 import vocabularies
-
 
 import nest_asyncio
 nest_asyncio.apply()
 
+from inferencemodel import InferenceModel
+
 SAMPLE_RATE = 16000
 SF2_PATH = 'SGM-v2.01-Sal-Guit-Bass-V1.3.sf2'
 
 def upload_audio(audio, sample_rate):
+  
   return note_seq.audio_io.wav_data_to_samples_librosa(
     audio, sample_rate=sample_rate)
 
-
-
-class InferenceModel(object):
-  """Wrapper of T5X model for music transcription."""
-
-  def __init__(self, checkpoint_path, model_type='mt3'):
-
-    # Model Constants.
-    if model_type == 'ismir2021':
-      num_velocity_bins = 127
-      self.encoding_spec = note_sequences.NoteEncodingSpec
-      self.inputs_length = 512
-    elif model_type == 'mt3':
-      num_velocity_bins = 1
-      self.encoding_spec = note_sequences.NoteEncodingWithTiesSpec
-      self.inputs_length = 256
-    else:
-      raise ValueError('unknown model_type: %s' % model_type)
-
-    gin_files = ['/home/user/app/mt3/gin/model.gin',
-                 '/home/user/app/mt3/gin/mt3.gin']
-
-    self.batch_size = 8
-    self.outputs_length = 1024
-    self.sequence_length = {'inputs': self.inputs_length, 
-                            'targets': self.outputs_length}
-
-    self.partitioner = t5x.partitioning.PjitPartitioner(
-        model_parallel_submesh=(1, 1, 1, 1), num_partitions=1)
-
-    # Build Codecs and Vocabularies.
-    self.spectrogram_config = spectrograms.SpectrogramConfig()
-    self.codec = vocabularies.build_codec(
-        vocab_config=vocabularies.VocabularyConfig(
-            num_velocity_bins=num_velocity_bins))
-    self.vocabulary = vocabularies.vocabulary_from_codec(self.codec)
-    self.output_features = {
-        'inputs': seqio.ContinuousFeature(dtype=tf.float32, rank=2),
-        'targets': seqio.Feature(vocabulary=self.vocabulary),
-    }
-
-    # Create a T5X model.
-    self._parse_gin(gin_files)
-    self.model = self._load_model()
-
-    # Restore from checkpoint.
-    self.restore_from_checkpoint(checkpoint_path)
-
-  @property
-  def input_shapes(self):
-    return {
-          'encoder_input_tokens': (self.batch_size, self.inputs_length),
-          'decoder_input_tokens': (self.batch_size, self.outputs_length)
-    }
-
-  def _parse_gin(self, gin_files):
-    """Parse gin files used to train the model."""
-    gin_bindings = [
-        'from __gin__ import dynamic_registration',
-        'from mt3 import vocabularies',
-        'VOCAB_CONFIG=@vocabularies.VocabularyConfig()',
-        'vocabularies.VocabularyConfig.num_velocity_bins=%NUM_VELOCITY_BINS'
-    ]
-    with gin.unlock_config():
-      gin.parse_config_files_and_bindings(
-          gin_files, gin_bindings, finalize_config=False)
-
-  def _load_model(self):
-    """Load up a T5X `Model` after parsing training gin config."""
-    model_config = gin.get_configurable(network.T5Config)()
-    module = network.Transformer(config=model_config)
-    return models.ContinuousInputsEncoderDecoderModel(
-        module=module,
-        input_vocabulary=self.output_features['inputs'].vocabulary,
-        output_vocabulary=self.output_features['targets'].vocabulary,
-        optimizer_def=t5x.adafactor.Adafactor(decay_rate=0.8, step_offset=0),
-        input_depth=spectrograms.input_depth(self.spectrogram_config))
-
-
-  def restore_from_checkpoint(self, checkpoint_path):
-    """Restore training state from checkpoint, resets self._predict_fn()."""
-    train_state_initializer = t5x.utils.TrainStateInitializer(
-      optimizer_def=self.model.optimizer_def,
-      init_fn=self.model.get_initial_variables,
-      input_shapes=self.input_shapes,
-      partitioner=self.partitioner)
-
-    restore_checkpoint_cfg = t5x.utils.RestoreCheckpointConfig(
-        path=checkpoint_path, mode='specific', dtype='float32')
-
-    train_state_axes = train_state_initializer.train_state_axes
-    self._predict_fn = self._get_predict_fn(train_state_axes)
-    self._train_state = train_state_initializer.from_checkpoint_or_scratch(
-        [restore_checkpoint_cfg], init_rng=jax.random.PRNGKey(0))
-
-  @functools.lru_cache()
-  def _get_predict_fn(self, train_state_axes):
-    """Generate a partitioned prediction function for decoding."""
-    def partial_predict_fn(params, batch, decode_rng):
-      return self.model.predict_batch_with_aux(
-          params, batch, decoder_params={'decode_rng': None})
-    return self.partitioner.partition(
-        partial_predict_fn,
-        in_axis_resources=(
-            train_state_axes.params,
-            t5x.partitioning.PartitionSpec('data',), None),
-        out_axis_resources=t5x.partitioning.PartitionSpec('data',)
-    )
-
-  def predict_tokens(self, batch, seed=0):
-    """Predict tokens from preprocessed dataset batch."""
-    prediction, _ = self._predict_fn(
-        self._train_state.params, batch, jax.random.PRNGKey(seed))
-    return self.vocabulary.decode_tf(prediction).numpy()
-
-  def __call__(self, audio):
-    """Infer note sequence from audio samples.
-    
-    Args:
-      audio: 1-d numpy array of audio samples (16kHz) for a single example.
-    Returns:
-      A note_sequence of the transcribed audio.
-    """
-    ds = self.audio_to_dataset(audio)
-    ds = self.preprocess(ds)
-
-    model_ds = self.model.FEATURE_CONVERTER_CLS(pack=False)(
-        ds, task_feature_lengths=self.sequence_length)
-    model_ds = model_ds.batch(self.batch_size)
-
-    inferences = (tokens for batch in model_ds.as_numpy_iterator()
-                  for tokens in self.predict_tokens(batch))
-
-    predictions = []
-    for example, tokens in zip(ds.as_numpy_iterator(), inferences):
-      predictions.append(self.postprocess(tokens, example))
-
-    result = metrics_utils.event_predictions_to_ns(
-        predictions, codec=self.codec, encoding_spec=self.encoding_spec)
-    return result['est_ns']
-
-  def audio_to_dataset(self, audio):
-    """Create a TF Dataset of spectrograms from input audio."""
-    frames, frame_times = self._audio_to_frames(audio)
-    return tf.data.Dataset.from_tensors({
-        'inputs': frames,
-        'input_times': frame_times,
-    })
-
-  def _audio_to_frames(self, audio):
-    """Compute spectrogram frames from audio."""
-    frame_size = self.spectrogram_config.hop_width
-    padding = [0, frame_size - len(audio) % frame_size]
-    audio = np.pad(audio, padding, mode='constant')
-    frames = spectrograms.split_audio(audio, self.spectrogram_config)
-    num_frames = len(audio) // frame_size
-    times = np.arange(num_frames) / self.spectrogram_config.frames_per_second
-    return frames, times
-
-  def preprocess(self, ds):
-    pp_chain = [
-        functools.partial(
-            t5.data.preprocessors.split_tokens_to_inputs_length,
-            sequence_length=self.sequence_length,
-            output_features=self.output_features,
-            feature_key='inputs',
-            additional_feature_keys=['input_times']),
-        # Cache occurs here during training.
-        preprocessors.add_dummy_targets,
-        functools.partial(
-            preprocessors.compute_spectrograms,
-            spectrogram_config=self.spectrogram_config)
-    ]
-    for pp in pp_chain:
-      ds = pp(ds)
-    return ds
-
-  def postprocess(self, tokens, example):
-    tokens = self._trim_eos(tokens)
-    start_time = example['input_times'][0]
-    # Round down to nearest symbolic token step.
-    start_time -= start_time % (1 / self.codec.steps_per_second)
-    return {
-        'est_tokens': tokens,
-        'start_time': start_time,
-        # Internal MT3 code expects raw inputs, not used here.
-        'raw_inputs': []
-    }
-
-  @staticmethod
-  def _trim_eos(tokens):
-    tokens = np.array(tokens, np.int32)
-    if vocabularies.DECODED_EOS_ID in tokens:
-      tokens = tokens[:np.argmax(tokens == vocabularies.DECODED_EOS_ID)]
-    return tokens
-
-
 # Start inference model
 inference_model = InferenceModel('/home/user/app/checkpoints/mt3/', 'mt3')
 
@@ -267,7 +46,4 @@ gr.Interface(
     description=description,
     article=article,
     examples=examples,
-    allow_flagging=False,
-    allow_screenshot=False,
-    enable_queue=True
     ).launch()

inferencemodel.py

@@ -0,0 +1,222 @@
+import os
+
+os.system("python3 -m pip install -e .")
+
+import functools
+import os
+
+import numpy as np
+import tensorflow.compat.v2 as tf
+
+import functools
+import gin
+import jax
+import seqio
+import t5
+import t5x
+
+from mt3 import metrics_utils
+from mt3 import models
+from mt3 import network
+from mt3 import note_sequences
+from mt3 import preprocessors
+from mt3 import spectrograms
+from mt3 import vocabularies
+
+
+import nest_asyncio
+nest_asyncio.apply()
+
+class InferenceModel(object):
+  """Wrapper of T5X model for music transcription."""
+
+  def __init__(self, checkpoint_path, model_type='mt3'):
+
+    # Model Constants.
+    if model_type == 'ismir2021':
+      num_velocity_bins = 127
+      self.encoding_spec = note_sequences.NoteEncodingSpec
+      self.inputs_length = 512
+    elif model_type == 'mt3':
+      num_velocity_bins = 1
+      self.encoding_spec = note_sequences.NoteEncodingWithTiesSpec
+      self.inputs_length = 256
+    else:
+      raise ValueError('unknown model_type: %s' % model_type)
+
+    gin_files = ['/home/user/app/mt3/gin/model.gin',
+                 '/home/user/app/mt3/gin/mt3.gin']
+
+    self.batch_size = 8
+    self.outputs_length = 1024
+    self.sequence_length = {'inputs': self.inputs_length, 
+                            'targets': self.outputs_length}
+
+    self.partitioner = t5x.partitioning.PjitPartitioner(
+        model_parallel_submesh=(1, 1, 1, 1), num_partitions=1)
+
+    # Build Codecs and Vocabularies.
+    self.spectrogram_config = spectrograms.SpectrogramConfig()
+    self.codec = vocabularies.build_codec(
+        vocab_config=vocabularies.VocabularyConfig(
+            num_velocity_bins=num_velocity_bins))
+    self.vocabulary = vocabularies.vocabulary_from_codec(self.codec)
+    self.output_features = {
+        'inputs': seqio.ContinuousFeature(dtype=tf.float32, rank=2),
+        'targets': seqio.Feature(vocabulary=self.vocabulary),
+    }
+
+    # Create a T5X model.
+    self._parse_gin(gin_files)
+    self.model = self._load_model()
+
+    # Restore from checkpoint.
+    self.restore_from_checkpoint(checkpoint_path)
+
+  @property
+  def input_shapes(self):
+    return {
+          'encoder_input_tokens': (self.batch_size, self.inputs_length),
+          'decoder_input_tokens': (self.batch_size, self.outputs_length)
+    }
+
+  def _parse_gin(self, gin_files):
+    """Parse gin files used to train the model."""
+    gin_bindings = [
+        'from __gin__ import dynamic_registration',
+        'from mt3 import vocabularies',
+        'VOCAB_CONFIG=@vocabularies.VocabularyConfig()',
+        'vocabularies.VocabularyConfig.num_velocity_bins=%NUM_VELOCITY_BINS'
+    ]
+    with gin.unlock_config():
+      gin.parse_config_files_and_bindings(
+          gin_files, gin_bindings, finalize_config=False)
+
+  def _load_model(self):
+    """Load up a T5X `Model` after parsing training gin config."""
+    model_config = gin.get_configurable(network.T5Config)()
+    module = network.Transformer(config=model_config)
+    return models.ContinuousInputsEncoderDecoderModel(
+        module=module,
+        input_vocabulary=self.output_features['inputs'].vocabulary,
+        output_vocabulary=self.output_features['targets'].vocabulary,
+        optimizer_def=t5x.adafactor.Adafactor(decay_rate=0.8, step_offset=0),
+        input_depth=spectrograms.input_depth(self.spectrogram_config))
+
+
+  def restore_from_checkpoint(self, checkpoint_path):
+    """Restore training state from checkpoint, resets self._predict_fn()."""
+    train_state_initializer = t5x.utils.TrainStateInitializer(
+      optimizer_def=self.model.optimizer_def,
+      init_fn=self.model.get_initial_variables,
+      input_shapes=self.input_shapes,
+      partitioner=self.partitioner)
+
+    restore_checkpoint_cfg = t5x.utils.RestoreCheckpointConfig(
+        path=checkpoint_path, mode='specific', dtype='float32')
+
+    train_state_axes = train_state_initializer.train_state_axes
+    self._predict_fn = self._get_predict_fn(train_state_axes)
+    self._train_state = train_state_initializer.from_checkpoint_or_scratch(
+        [restore_checkpoint_cfg], init_rng=jax.random.PRNGKey(0))
+
+  @functools.lru_cache()
+  def _get_predict_fn(self, train_state_axes):
+    """Generate a partitioned prediction function for decoding."""
+    def partial_predict_fn(params, batch, decode_rng):
+      return self.model.predict_batch_with_aux(
+          params, batch, decoder_params={'decode_rng': None})
+    return self.partitioner.partition(
+        partial_predict_fn,
+        in_axis_resources=(
+            train_state_axes.params,
+            t5x.partitioning.PartitionSpec('data',), None),
+        out_axis_resources=t5x.partitioning.PartitionSpec('data',)
+    )
+
+  def predict_tokens(self, batch, seed=0):
+    """Predict tokens from preprocessed dataset batch."""
+    prediction, _ = self._predict_fn(
+        self._train_state.params, batch, jax.random.PRNGKey(seed))
+    return self.vocabulary.decode_tf(prediction).numpy()
+
+  def __call__(self, audio):
+    """Infer note sequence from audio samples.
+    
+    Args:
+      audio: 1-d numpy array of audio samples (16kHz) for a single example.
+    Returns:
+      A note_sequence of the transcribed audio.
+    """
+    ds = self.audio_to_dataset(audio)
+    ds = self.preprocess(ds)
+
+    model_ds = self.model.FEATURE_CONVERTER_CLS(pack=False)(
+        ds, task_feature_lengths=self.sequence_length)
+    model_ds = model_ds.batch(self.batch_size)
+
+    inferences = (tokens for batch in model_ds.as_numpy_iterator()
+                  for tokens in self.predict_tokens(batch))
+
+    predictions = []
+    for example, tokens in zip(ds.as_numpy_iterator(), inferences):
+      predictions.append(self.postprocess(tokens, example))
+
+    result = metrics_utils.event_predictions_to_ns(
+        predictions, codec=self.codec, encoding_spec=self.encoding_spec)
+    return result['est_ns']
+
+  def audio_to_dataset(self, audio):
+    """Create a TF Dataset of spectrograms from input audio."""
+    frames, frame_times = self._audio_to_frames(audio)
+    return tf.data.Dataset.from_tensors({
+        'inputs': frames,
+        'input_times': frame_times,
+    })
+
+  def _audio_to_frames(self, audio):
+    """Compute spectrogram frames from audio."""
+    frame_size = self.spectrogram_config.hop_width
+    padding = [0, frame_size - len(audio) % frame_size]
+    audio = np.pad(audio, padding, mode='constant')
+    frames = spectrograms.split_audio(audio, self.spectrogram_config)
+    num_frames = len(audio) // frame_size
+    times = np.arange(num_frames) / self.spectrogram_config.frames_per_second
+    return frames, times
+
+  def preprocess(self, ds):
+    pp_chain = [
+        functools.partial(
+            t5.data.preprocessors.split_tokens_to_inputs_length,
+            sequence_length=self.sequence_length,
+            output_features=self.output_features,
+            feature_key='inputs',
+            additional_feature_keys=['input_times']),
+        # Cache occurs here during training.
+        preprocessors.add_dummy_targets,
+        functools.partial(
+            preprocessors.compute_spectrograms,
+            spectrogram_config=self.spectrogram_config)
+    ]
+    for pp in pp_chain:
+      ds = pp(ds)
+    return ds
+
+  def postprocess(self, tokens, example):
+    tokens = self._trim_eos(tokens)
+    start_time = example['input_times'][0]
+    # Round down to nearest symbolic token step.
+    start_time -= start_time % (1 / self.codec.steps_per_second)
+    return {
+        'est_tokens': tokens,
+        'start_time': start_time,
+        # Internal MT3 code expects raw inputs, not used here.
+        'raw_inputs': []
+    }
+
+  @staticmethod
+  def _trim_eos(tokens):
+    tokens = np.array(tokens, np.int32)
+    if vocabularies.DECODED_EOS_ID in tokens:
+      tokens = tokens[:np.argmax(tokens == vocabularies.DECODED_EOS_ID)]
+    return tokens

requirements.txt

@@ -7,5 +7,4 @@ jax[cpu]==0.3.15 -f https://storage.googleapis.com/jax-releases/jax_releases.htm
 # pin CLU for python 3.7 compatibility
 clu==0.0.7
 # pin Orbax to use Checkpointer
-orbax==0.0.2
-pydub
+orbax==0.0.2