Update demos/musicgen_app.py

demos/musicgen_app.py

@@ -18,6 +18,7 @@ from tempfile import NamedTemporaryFile
 import time
 import typing as tp
 import warnings
+import base64
 
 from einops import rearrange
 import torch
@@ -209,9 +210,22 @@ def predict_full(secret_token, model, model_path, decoder, text, melody, duratio
         [text], [melody], duration, progress=True,
         top_k=topk, top_p=topp, temperature=temperature, cfg_coef=cfg_coef,
         gradio_progress=progress)
+
+    wav_path = wavs[0]
     if USE_DIFFUSION:
-        return wavs[1]
-    return wavs[0]
+        wav_path = wavs[1]
+
+
+    wav_base64 = ""
+    
+    # Read the content of the video file and encode it to base64
+    with open(wav_path, "rb") as wav_file:
+        wav_base64 = base64.b64encode(wav_file.read()).decode('utf-8')
+
+    # Prepend the appropriate data URI header with MIME type
+    wav_base64_data_uri = 'data:audio/wav;base64,' + wav_base64
+    
+    return wav_base64_data_uri
 
 def toggle_audio_src(choice):
     if choice == "mic":
@@ -274,8 +288,8 @@ def ui_full(launch_kwargs):
                     temperature = gr.Number(label="Temperature", value=1.0, interactive=True)
                     cfg_coef = gr.Number(label="Classifier Free Guidance", value=3.0, interactive=True)
             with gr.Column():
-                audio_output = gr.Audio(label="Generated Music (wav)", type='filepath')
-                audio_diffusion = gr.Audio(label="MultiBand Diffusion Decoder (wav)", type='filepath')
+                audio_output = gr.Audio(label="Generated Music (wav)")
+                audio_diffusion = gr.Textbox(label="MultiBand Diffusion Decoder (wav)")
         submit.click(toggle_diffusion, decoder, [audio_diffusion], queue=False,
                      show_progress=False).then(predict_full, inputs=[secret_token, model, model_path, decoder, text, melody, duration, topk, topp,
                                                                      temperature, cfg_coef],