Add MusicGen Melody (#28819)

* first modeling code

* make repository

* still WIP

* update model

* add tests

* add latest change

* clean docstrings and copied from

* update docstrings md and readme

* correct chroma function

* correct copied from and remove unreleated test

* add doc to toctree

* correct imports

* add convert script to notdoctested

* Add suggestion from Sanchit

Co-authored-by: Sanchit Gandhi <93869735+sanchit-gandhi@users.noreply.github.com>

* correct get_uncoditional_inputs docstrings

* modify README according to SANCHIT feedback

* add chroma to audio utils

* clean librosa and torchaudio hard dependencies

* fix FE

* refactor audio decoder -> audio encoder for consistency with previous musicgen

* refactor conditional -> encoder

* modify sampling rate logics

* modify license at the beginning

* refactor all_self_attns->all_attentions

* remove ignore copy from causallm generate

* add copied from for from_sub_models

* fix make copies

* add warning if audio is truncated

* add copied from where relevant

* remove artefact

* fix convert script

* fix torchaudio and FE

* modify chroma method according to feedback-> better naming

* refactor input_values->input_features

* refactor input_values->input_features and fix import fe

* add input_features to docstrigs

* correct inputs_embeds logics

* remove dtype conversion

* refactor _prepare_conditional_hidden_states_kwargs_for_generation ->_prepare_encoder_hidden_states_kwargs_for_generation

* change warning for chroma length

* Update src/transformers/models/musicgen_melody/convert_musicgen_melody_transformers.py

Co-authored-by: Sanchit Gandhi <93869735+sanchit-gandhi@users.noreply.github.com>

* change way to save wav, using soundfile

* correct docs and change to soundfile

* fix import

* fix init proj layers

* remove line breaks from md

* fix issue with docstrings

* add FE suggestions

* improve is in logics and remove useless imports

* remove custom from_pretrained

* simplify docstring code

* add suggestions for modeling tests

* make style

* update converting script with sanity check

* remove encoder attention mask from conditional generation

* replace musicgen melody checkpoints with official orga

* rename ylacombe->facebook in checkpoints

* fix copies

* remove unecessary warning

* add shape in code docstrings

* add files to slow doc tests

* fix md bug and add md to not_tested

* make fix-copies

* fix hidden states test and batching

---------

Co-authored-by: Sanchit Gandhi <93869735+sanchit-gandhi@users.noreply.github.com>

tests/utils/test_audio_utils.py

@@ -20,6 +20,7 @@ import pytest
 
 from transformers.audio_utils import (
     amplitude_to_db,
+    chroma_filter_bank,
     hertz_to_mel,
     mel_filter_bank,
     mel_to_hertz,
@@ -27,6 +28,11 @@ from transformers.audio_utils import (
     spectrogram,
     window_function,
 )
+from transformers.testing_utils import is_librosa_available, require_librosa
+
+
+if is_librosa_available():
+    from librosa.filters import chroma
 
 
 class AudioUtilsFunctionTester(unittest.TestCase):
@@ -755,3 +761,57 @@ class AudioUtilsFunctionTester(unittest.TestCase):
             amplitude_to_db(spectrogram, min_value=0.0)
         with pytest.raises(ValueError):
             amplitude_to_db(spectrogram, db_range=-80)
+
+    @require_librosa
+    def test_chroma_equivalence(self):
+        num_frequency_bins = 25
+        num_chroma = 6
+        sampling_rate = 24000
+
+        # test default parameters
+        original_chroma = chroma(sr=sampling_rate, n_chroma=num_chroma, n_fft=num_frequency_bins)
+        utils_chroma = chroma_filter_bank(
+            num_frequency_bins=num_frequency_bins, num_chroma=num_chroma, sampling_rate=sampling_rate
+        )
+
+        self.assertTrue(np.allclose(original_chroma, utils_chroma))
+
+        # test no weighting_parameters
+        original_chroma = chroma(sr=sampling_rate, n_chroma=num_chroma, n_fft=num_frequency_bins, octwidth=None)
+        utils_chroma = chroma_filter_bank(
+            num_frequency_bins=num_frequency_bins,
+            num_chroma=num_chroma,
+            sampling_rate=sampling_rate,
+            weighting_parameters=None,
+        )
+
+        self.assertTrue(np.allclose(original_chroma, utils_chroma))
+
+        # test with L1 norm
+        original_chroma = chroma(sr=sampling_rate, n_chroma=num_chroma, n_fft=num_frequency_bins, norm=1.0)
+        utils_chroma = chroma_filter_bank(
+            num_frequency_bins=num_frequency_bins, num_chroma=num_chroma, sampling_rate=sampling_rate, power=1.0
+        )
+
+        self.assertTrue(np.allclose(original_chroma, utils_chroma))
+
+        # test starting at 'A' chroma, power = None, tuning = 0, different weighting_parameters
+        original_chroma = chroma(
+            sr=sampling_rate,
+            n_chroma=num_chroma,
+            n_fft=num_frequency_bins,
+            norm=None,
+            base_c=None,
+            octwidth=1.0,
+            ctroct=4.0,
+        )
+        utils_chroma = chroma_filter_bank(
+            num_frequency_bins=num_frequency_bins,
+            num_chroma=num_chroma,
+            sampling_rate=sampling_rate,
+            power=None,
+            start_at_c_chroma=False,
+            weighting_parameters=(4.0, 1.0),
+        )
+
+        self.assertTrue(np.allclose(original_chroma, utils_chroma))