token

app.py

@@ -2,7 +2,7 @@ import torch
 from transformers import AutoTokenizer, AutoModelForCausalLM, BitsAndBytesConfig
 import gradio as gr
 from peft import PeftModel
-import os
+import spaces  # Ensure spaces is imported
 
 # Define the base model ID
 base_model_id = "meta-llama/Llama-2-13b-hf"
@@ -38,9 +38,10 @@ def formatting_func(job_description):
     text = f"### The job description: {job_description}\n ### The skills: "
     return text
 
+@spaces.GPU  # Decorate the function to ensure it uses GPU
 def generate_skills(job_description):
     formatted_text = formatting_func(job_description)
-    model_input = tokenizer(formatted_text, return_tensors="pt").to("cuda")
+    model_input = tokenizer(formatted_text, return_tensors="pt").to("cuda")  # Use CUDA for GPU support
 
     ft_model.eval()
     with torch.no_grad():