Upscale

Dockerfile

@@ -0,0 +1,11 @@
+FROM python:3.10.9
+
+WORKDIR /code
+
+COPY ./requirements.txt /code/requirements.txt
+
+RUN pip install --no-cache-dir --upgrade -r /code/requirements.txt
+
+COPY . .
+
+CMD ["uvicorn", "main:app", "--host", "0.0.0.0", "--port", "7860"]

main.py

@@ -0,0 +1,96 @@
+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
+from pyngrok import ngrok
+
+app = FastAPI()
+
+device = "cuda:0" if torch.cuda.is_available() else "cpu"
+torch_dtype = torch.float16 if torch.cuda.is_available() else torch.float32
+
+model_id = "openai/whisper-large-v3"
+
+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"{file.filename}"
+        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

requirements.txt

@@ -0,0 +1,4 @@
+fastapi
+uvicorn
+googletrans==4.0.0rc1
+--upgrade git+https://github.com/huggingface/transformers.git accelerate datasets[audio]