Update app.py

app.py

@@ -3,6 +3,8 @@ import gradio as gr
 from transformers import AutoModelForCausalLM, AutoTokenizer, BitsAndBytesConfig
 import torch
 import os
+
+# Load model
 bnb_config = BitsAndBytesConfig(load_in_4bit=True)
 model_name = "meta-llama/Meta-Llama-3-8B-Instruct"
 hf_token = os.environ.get("HF_TOKEN")
@@ -11,13 +13,15 @@ tokenizer = AutoTokenizer.from_pretrained(model_name, use_auth_token=hf_token)
 model = AutoModelForCausalLM.from_pretrained(
     model_name,
     quantization_config=bnb_config,
-    model.to("cpu"),
     use_auth_token=hf_token
 )
+model.to("cpu")  # move model to CPU (if not using GPU Space)
 
+# Prompt generator
 def generate_prompt(original, translation):
     return f"### Task: Machine Translation Quality Estimation\n\nSource: {original}\nTranslation: {translation}\n\nScore (0-1):"
 
+# Prediction function
 def predict_scores(file):
     df = pd.read_csv(file.name, sep="\t")
     scores = []
@@ -40,6 +44,7 @@ def predict_scores(file):
     df["predicted_score"] = scores
     return df
 
+# Gradio app
 iface = gr.Interface(
     fn=predict_scores,
     inputs=gr.File(label="Upload dev.tsv"),