Update app.py

app.py

@@ -3,26 +3,28 @@ from transformers import AutoModelForCausalLM, AutoTokenizer, pipeline
 import torch
 import os
 
+# === Model ID and Token ===
 model_id = "TrabbyPatty/mistral-7b-instruct-finetuned-flashcards"
-hf_token = os.getenv("alluse")  # retrieve from Space secrets
+hf_token = os.getenv("alluse")  # Hugging Face token from Space secrets
 
-tokenizer = AutoTokenizer.from_pretrained(model_id)
+# === Load tokenizer & model with authentication ===
+tokenizer = AutoTokenizer.from_pretrained(model_id, token=hf_token)
 model = AutoModelForCausalLM.from_pretrained(
     model_id,
-    device_map="cpu",     # force CPU
-    torch_dtype="float32" # safest for CPU
+    device_map="cpu",         # force CPU
+    torch_dtype=torch.float32, # safest for CPU
+    token=hf_token
 )
 
-
+# === Create pipeline ===
 pipe = pipeline(
     "text-generation",
     model=model,
     tokenizer=tokenizer,
-    torch_dtype=torch.float16,
-    device_map="auto"
+    torch_dtype=torch.float32,
+    device_map="cpu"
 )
 
-
 # === SYSTEM MESSAGE ===
 SYSTEM_MESSAGE = """<<SYS>>
 You are a strict flashcard generator.
@@ -31,8 +33,8 @@ You are a strict flashcard generator.
 - Always follow the requested format exactly.
 <</SYS>>"""
 
+# === Generation function ===
 def generate(user_input, max_new_tokens=800, temperature=0.5):
-    # Wrap input with system instruction + prompt template
     prompt = (
         f"<s>[INST] {SYSTEM_MESSAGE}\n\n"
         f"Create a variety of study aids with 10 items each, strictly using only the information provided.\n\n"
@@ -47,7 +49,7 @@ def generate(user_input, max_new_tokens=800, temperature=0.5):
     )
     return output[0]["generated_text"]
 
-# Gradio UI
+# === Gradio UI ===
 demo = gr.Interface(
     fn=generate,
     inputs=[