testing deleting fast api

main.py

@@ -9,15 +9,12 @@ import torch
 import soundfile as sf
 from demucs.apply import apply_model
 from demucs.pretrained import DEFAULT_MODEL, get_model
-from fastapi import FastAPI, UploadFile, File
+import gradio as gr
 from huggingface_hub import hf_hub_download, list_repo_files
-from starlette.responses import StreamingResponse
 
 from so_vits_svc_fork.hparams import HParams
 from so_vits_svc_fork.inference.core import Svc
 
-app = FastAPI()
-
 ###################################################################
 # REPLACE THESE VALUES TO CHANGE THE MODEL REPO/CKPT NAME/SETTINGS
 ###################################################################
@@ -102,36 +99,20 @@ def predict(
     return model.target_sample, out
 
 
-@app.post("/voice_cloning/")
-async def voice_cloning(
-    speaker: str,
-    file: UploadFile = File(...),
-    transpose: int = 0,
-    auto_predict_f0: bool = False,
-    cluster_infer_ratio: float = default_cluster_infer_ratio,
-    noise_scale: float = 0.4,
-    f0_method: str = default_f0_method,
-):
-    # Process the audio file
-    audio_bytes = await file.read()
-    with io.BytesIO(audio_bytes) as audio_io:
-        sample_rate, audio_data = predict(
-            speaker,
-            audio_io,
-            transpose=transpose,
-            auto_predict_f0=auto_predict_f0,
-            cluster_infer_ratio=cluster_infer_ratio,
-            noise_scale=noise_scale,
-            f0_method=f0_method
-        )
-        # Create a BytesIO object to hold the audio data
-        audio_byte_stream = io.BytesIO()
-        # Write the audio data to the BytesIO object
-        sf.write(audio_byte_stream, audio_data, sample_rate, format="wav")
-        # Create a StreamingResponse to return the audio
-        return StreamingResponse(io.BytesIO(audio_byte_stream.getvalue()), media_type="audio/wav")
+def voice_cloning(speaker, audio):
+    sample_rate, audio_data = predict(speaker, audio)
+    return audio_data, sample_rate
+
+
+# Configure the Gradio interface
+inputs = [
+    gr.inputs.Dropdown(choices=speakers, label="Speaker"),
+    gr.inputs.Audio(label="Audio")
+]
+
+outputs = gr.outputs.Audio(label="Cloned Audio")
 
+iface = gr.Interface(fn=voice_cloning, inputs=inputs, outputs=outputs)
 
 if __name__ == "__main__":
-    import uvicorn
-    uvicorn.run(app, host="0.0.0.0", port=8000)
+    iface.launch()