Update app.py

app.py

@@ -1,19 +1,16 @@
-from fastapi import FastAPI
-from pydantic import BaseModel
 from transformers import AutoModelForCausalLM, AutoTokenizer
 import torch
+import os
 
-app = FastAPI()
+model_name = "meta-llama/Llama-4-11B-Instruct"
 
-model_name = "meta-llama/Llama-4-11B"
-tokenizer = AutoTokenizer.from_pretrained(model_name)
-model = AutoModelForCausalLM.from_pretrained(model_name, torch_dtype=torch.float16, device_map="auto")
+# use the name of your secret, here "test"
+token = os.getenv("test")
 
-class InputText(BaseModel):
-    text: str
-
-@app.post("/predict")
-def predict(item: InputText):
-    inputs = tokenizer(item.text, return_tensors="pt").to(model.device)
-    outputs = model.generate(**inputs, max_new_tokens=200)
-    return {"result": tokenizer.decode(outputs[0], skip_special_tokens=True)}
+tokenizer = AutoTokenizer.from_pretrained(model_name, token=token)
+model = AutoModelForCausalLM.from_pretrained(
+    model_name,
+    torch_dtype=torch.float16,
+    device_map="auto",
+    token=token
+)