Create app.py

app.py

@@ -0,0 +1,59 @@
+import torch
+from fastapi import FastAPI
+from pydantic import BaseModel
+from transformers import AutoTokenizer, AutoModelForCausalLM, pipeline
+import uvicorn
+
+app = FastAPI()
+
+MODEL_ID = "Qwen/Qwen1.5-1.8B-Chat"
+
+tokenizer = AutoTokenizer.from_pretrained(MODEL_ID)
+model = AutoModelForCausalLM.from_pretrained(
+    MODEL_ID,
+    torch_dtype=torch.bfloat16,
+    device_map="auto",
+)
+
+pipe = pipeline(
+    "text-generation",
+    model=model,
+    tokenizer=tokenizer,
+)
+
+class TranslationRequest(BaseModel):
+    text: str
+    target_lang: str
+
+@app.post("/translate")
+async def translate(request: TranslationRequest):
+    korean_text = request.text
+    target_lang = request.target_lang
+
+    if target_lang == 'english':
+        prompt = f"Translate the following Korean sentence into natural, everyday English. Provide only the translated sentence, without any additional explanations or quotation marks.\n\nKorean: \"{korean_text}\"\n\nEnglish:"
+    elif target_lang == 'japanese':
+        prompt = f"Translate the following Korean sentence into natural, everyday Japanese. Provide only the translated sentence, without any additional explanations or quotation marks.\n\nKorean: \"{korean_text}\"\n\nJapanese:"
+    else:
+        return {"error": "Invalid target language"}
+
+    messages = [
+        {"role": "user", "content": prompt}
+    ]
+
+    outputs = pipe(
+        messages,
+        max_new_tokens=150,
+        do_sample=True,
+        temperature=0.7,
+        top_k=50,
+    )
+
+    generated_text = outputs[0]["generated_text"]
+    translated_text = generated_text.split("assistant\n")[-1].strip()
+
+    return {"translated_text": translated_text}
+
+@app.get("/")
+def read_root():
+    return {"message": "Translation API is running"}