Update app.py

app.py

@@ -30,7 +30,7 @@ if os.environ["HUGGINGFACEHUB_API_TOKEN"] == "default-token":
 # Model and embedding options
 LLM_MODELS = {
     "High Accuracy (Mixtral-8x7B)": "mistralai/Mixtral-8x7B-Instruct-v0.1",
-    "Balanced (Gemma-2-9B)": "google/gemma-2-9b-it",
+    "Balanced (Gemma-2-2B)": "google/gemma-2-2b-it",
     "Lightweight (Mistral-7B)": "mistralai/Mistral-7B-Instruct-v0.2"
 }
 
@@ -188,7 +188,7 @@ def initialize_qa_chain(llm_model, temperature):
         if "503" in str(e):
             return f"Error: Hugging Face API temporarily unavailable for {llm_model}. Try 'High Accuracy (Mixtral-8x7B)' or wait and retry.", None
         elif "403" in str(e):
-            return f"Error: Access denied for {llm_model}. Check your HF token permissions or upgrade to a Pro account for larger models.", None
+            return f"Error: Access denied for {llm_model}. Free-tier API limits models >10GB. Try 'High Accuracy (Mixtral-8x7B)' or upgrade to Pro at https://huggingface.co/settings/billing.", None
         return f"Error initializing QA chain: {str(e)}.", None
     except Exception as e:
         logger.error(f"Error initializing QA chain for {llm_model}: {str(e)}")
@@ -220,7 +220,7 @@ def answer_question(question, llm_model, embedding_model, temperature, chunk_siz
         if "503" in str(e):
             return f"Error: Hugging Face API temporarily unavailable for {llm_model}. Try 'High Accuracy (Mixtral-8x7B)' or wait and retry.", chat_history
         elif "403" in str(e):
-            return f"Error: Access denied for {llm_model}. Check your HF token permissions or upgrade to a Pro account for larger models.", chat_history
+            return f"Error: Access denied for {llm_model}. Free-tier API limits models >10GB. Try 'High Accuracy (Mixtral-8x7B)' or upgrade to Pro at https://huggingface.co/settings/billing.", chat_history
         return f"Error answering question: {str(e)}", chat_history
     except Exception as e:
         logger.error(f"Error answering question: {str(e)}")
@@ -301,7 +301,6 @@ with gr.Blocks(theme=gr.themes.Soft(), title="DocTalk: Document Q&A Chatbot") as
         inputs=[llm_model, temperature],
         outputs=[status, chat_display]
     )
-    question里的
     question.submit(
         fn=answer_question,
         inputs=[question, llm_model, embedding_model, temperature, chunk_size, chunk_overlap],