Update app.py (#10)

- Update app.py (3d42cfc2a940e31fd6bb2983beee88b3d3bc0464)

app.py

@@ -2,7 +2,7 @@ import torch
 import librosa
 import soundfile as sf
 import io
-from fastapi import FastAPI, File, UploadFile
+from fastapi import FastAPI, File, UploadFile, Request
 from fastapi.responses import JSONResponse
 from transformers import Wav2Vec2ForCTC, Wav2Vec2Processor
 
@@ -41,7 +41,7 @@ except Exception as e:
 
 # --- 3. Define the Transcription Endpoint ---
 @app.post("/transcribe/")
-async def transcribe_audio(audio_file: UploadFile = File(...)):
+async def transcribe(audio_file: UploadFile = File(...)):
     if not model or not processor:
         return JSONResponse(status_code=503, content={"error": "Model is not loaded."})
 
@@ -71,6 +71,38 @@ async def transcribe_audio(audio_file: UploadFile = File(...)):
     except Exception as e:
         print(f"Error during transcription: {str(e)}")
         return JSONResponse(status_code=500, content={"error": f"An error occurred: {str(e)}"})
+    
+@app.post("/transcribe_audio/")
+async def transcribe_audio(request: Request):
+    if not model or not processor:
+        return JSONResponse(status_code=503, content={"error": "Model is not loaded."})
+
+    try:
+        contents = await request.body()
+        
+        audio_data, original_sr = sf.read(io.BytesIO(contents))
+        if audio_data.ndim > 1:
+            audio_data = audio_data.mean(axis=1)
+
+        resampled_audio = librosa.resample(y=audio_data, orig_sr=original_sr, target_sr=16000)
+
+        inputs = processor(resampled_audio, sampling_rate=16000, return_tensors="pt", padding=True)
+        
+        # <-- CHANGED: Move the input tensors to the same device as the model
+        inputs = inputs.to(device)
+
+        with torch.no_grad():
+            logits = model(**inputs).logits
+        
+        predicted_ids = torch.argmax(logits, dim=-1)
+        transcription = processor.batch_decode(predicted_ids)[0]
+        
+        print(f"Transcription complete: {transcription}")
+        return {"transcription": transcription}
+
+    except Exception as e:
+        print(f"Error during transcription: {str(e)}")
+        return JSONResponse(status_code=500, content={"error": f"An error occurred: {str(e)}"})
 
 # --- 4. Root Endpoint for Health Check ---
 @app.get("/")