Spaces:
Running
on
A10G
Running
on
A10G
File size: 3,515 Bytes
5238467 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 |
"""
Copyright (c) Meta Platforms, Inc. and affiliates.
All rights reserved.
This source code is licensed under the license found in the
LICENSE file in the root directory of this source tree.
"""
from tempfile import NamedTemporaryFile
import torch
import gradio as gr
from audiocraft.data.audio_utils import convert_audio
from audiocraft.data.audio import audio_write
from hf_loading import get_pretrained
MODEL = None
def load_model():
print("Loading model")
return get_pretrained("melody")
def predict(texts, melodies):
global MODEL
if MODEL is None:
MODEL = load_model()
duration = 12
MODEL.set_generation_params(duration=duration)
print(texts, melodies)
processed_melodies = []
target_sr = 32000
target_ac = 1
for melody in melodies:
if melody is None:
processed_melodies.append(None)
else:
sr, melody = melody[0], torch.from_numpy(melody[1]).to(MODEL.device).float().t()
if melody.dim() == 1:
melody = melody[None]
melody = melody[..., :int(sr * duration)]
melody = convert_audio(melody, sr, target_sr, target_ac)
processed_melodies.append(melody)
outputs = MODEL.generate_with_chroma(
descriptions=texts,
melody_wavs=processed_melodies,
melody_sample_rate=target_sr,
progress=False
)
outputs = outputs.detach().cpu().float()
out_files = []
for output in outputs:
with NamedTemporaryFile("wb", suffix=".wav", delete=False) as file:
audio_write(file.name, output, MODEL.sample_rate, strategy="loudness", add_suffix=False)
out_files.append([file.name])
return out_files
with gr.Blocks() as demo:
gr.Markdown(
"""
# MusicGen
This is the demo for MusicGen, a simple and controllable model for music generation
presented at: "Simple and Controllable Music Generation".
Enter the description of the music you want and an optional audio used for melody conditioning.
This will generate a 12s extract with the `melody` model. For generating longer sequences
(up to 30 seconds), use the Colab demo or your own GPU.
See [github.com/facebookresearch/audiocraft](https://github.com/facebookresearch/audiocraft)
for more details.
"""
)
with gr.Row():
with gr.Column():
with gr.Row():
text = gr.Text(label="Input Text", interactive=True)
melody = gr.Audio(source="upload", type="numpy", label="Melody Condition (optional)", interactive=True)
with gr.Row():
submit = gr.Button("Submit")
with gr.Column():
output = gr.Audio(label="Generated Music", type="filepath", format="wav")
submit.click(predict, inputs=[text, melody], outputs=[output], batch=True, max_batch_size=12)
gr.Examples(
fn=predict,
examples=[
[
"An 80s driving pop song with heavy drums and synth pads in the background",
"./assets/bach.mp3",
],
[
"90s rock song with electric guitar and heavy drums",
None,
],
[
"a light and cheerly EDM track, with syncopated drums, aery pads, and strong emotions",
"./assets/bach.mp3",
]
],
inputs=[text, melody],
outputs=[output]
)
demo.launch()
|