Update main.py

main.py

@@ -1,22 +1,97 @@
-from fastapi import FastAPI, File, UploadFile
-from pathlib import Path
+import torch
+from transformers import AutoModelForSpeechSeq2Seq, AutoProcessor, pipeline
+from datasets import load_dataset
+from googletrans import Translator
+from fastapi import FastAPI, File, UploadFile, HTTPException
+from fastapi.responses import JSONResponse
 
 app = FastAPI()
 
-# Endpoint untuk mengunggah file
-@app.post("/uploadfile/")
-async def upload_file(file: UploadFile = File(...)):
-    # Tentukan direktori tempat file akan disimpan
-    # Pastikan direktori ini ada dan memiliki izin yang tepat di server kamu
-    save_directory = Path("/home/user")
-    save_directory.mkdir(parents=True, exist_ok=True)
+device = "cuda:0" if torch.cuda.is_available() else "cpu"
+torch_dtype = torch.float16 if torch.cuda.is_available() else torch.float32
 
-    # Menentukan path lengkap tempat file akan disimpan
-    file_location = save_directory / file.filename
+model_id = "openai/whisper-large-v3"
 
-    # Baca konten file dan tulis ke lokasi yang ditentukan
-    with open(file_location, "wb") as saved_file:
-        content = await file.read()  # async read
-        saved_file.write(content)
+model = AutoModelForSpeechSeq2Seq.from_pretrained(model_id, torch_dtype=torch_dtype, low_cpu_mem_usage=True, use_safetensors=True)
+model.to(device)
+
+processor = AutoProcessor.from_pretrained(model_id)
+
+pipe = pipeline(
+    "automatic-speech-recognition",
+    model=model,
+    tokenizer=processor.tokenizer,
+    feature_extractor=processor.feature_extractor,
+    max_new_tokens=256,
+    chunk_length_s=30,
+    batch_size=16,
+    return_timestamps=True,
+    torch_dtype=torch_dtype,
+    device=device,
+)
+
+dataset = load_dataset("distil-whisper/librispeech_long", "clean", split="validation")
+
+@app.post("/voice_recognition")
+async def process_audio(file: UploadFile = File(...)):
+    try:
+        # File
+        file_path = f"/home/user/{file.filename}"
+        file_path.mkdir(parents=True, exist_ok=True)
+        with open(file_path, "wb") as f:
+            f.write(file.file.read())
+
+        # JP
+        original = pipe(file_path)
+        original_version = original["text"]
+
+        # EN
+        result = pipe(file_path, generate_kwargs={"task": "translate"})
+        hasil = result["text"]
+
+        # ID
+        detect = detect_google(hasil)
+        id_ver = translate_google(hasil, f"{detect}", "ID")
+
+        # Additional modifications
+        id_ver = modify_text(id_ver)
+
+        return JSONResponse(content={"response": {"jp_text": original_version, "en_text": hasil, "id_text": id_ver}}, status_code=200)
+
+    except Exception as e:
+        return HTTPException(status_code=500, detail=f"Error: {e}")
+
+def detect_google(text):
+    try:
+        translator = Translator()
+        detected_lang = translator.detect(text)
+        return detected_lang.lang.upper()
+    except Exception as e:
+        print(f"Error detect: {e}")
+        return None
+
+def translate_google(text, source, target):
+    try:
+        translator = Translator()
+        translated_text = translator.translate(text, src=source, dest=target)
+        return translated_text.text
+    except Exception as e:
+        print(f"Error translate: {e}")
+        return None
+
+def modify_text(text):
+    # Additional modifications, case-sensitive
+    replacements = {
+        "Tuan": "Master",
+        "tuan": "Master",
+        "Guru": "Master",
+        "guru": "Master",
+        "Monica": "Monika",
+        "monica": "Monika",
+    }
+
+    for original, replacement in replacements.items():
+        text = text.replace(original, replacement)
+
+    return text
 
-    return {"info": "File saved successfully", "filename": file.filename}