remove llama model

app.py

@@ -73,54 +73,53 @@ model_basename = "llama-2-13b-chat.ggmlv3.q5_1.bin"  # The model is in bin forma
 
 # Download the model file
 print('downloading llama model...')
-model_path_llama = hf_hub_download(repo_id=model_name_or_path, filename=model_basename, force_download=True, local_dir="./llama-model")
+# model_path_llama = hf_hub_download(repo_id=model_name_or_path, filename=model_basename, force_download=True, local_dir="./llama_model")
 print('finished download...')
 # Initialize the Llama model with appropriate settings for GPU
-lcpp_llm = Llama(
-    model_path=model_path_llama,
-    n_batch=512,  # Batch size for processing; adjust as per your VRAM capacity
-)
-
-def generate_email_response(email_prompt):
-    # Check input received by the function
-    print("Received prompt:", email_prompt)
-
-    # Determine if the input is a shorthand command or an actual email
-    if 'email to' in email_prompt.lower():
-        # Assume it's a shorthand command, format appropriately
-        formatted_prompt = f'''
-        Email received: "{email_prompt}"
-        Respond to this email, ensuring a professional tone, providing a concise update, and addressing any potential concerns the sender might have.
-        Response:
-        '''
-    else:
-        # Assume it's direct email content
-        formatted_prompt = f'''
-        Email received: "{email_prompt}"
-        Respond to this email, ensuring a professional tone, providing a concise update, and addressing any potential concerns the sender might have.
-        Response:
-        '''
-
-    # Generate response using Llama-2 model
-    try:
-        response = lcpp_llm(
-            prompt=formatted_prompt,
-            max_tokens=256,
-            temperature=0.5,
-            top_p=0.95,
-            repeat_penalty=1.2,
-            top_k=150,
-            echo=True
-        )
-        generated_response = response["choices"][0]["text"]
-        # Remove the input part from the output if it is included
-        if formatted_prompt in generated_response:
-            generated_response = generated_response.replace(formatted_prompt, '').strip()
-        print("Generated response:", generated_response)
-        return generated_response
-    except Exception as e:
-        print("Error in response generation:", str(e))
-        return "Failed to generate response, please check the console for errors."
+# lcpp_llm = Llama(
+#     model_path=model_path_llama,
+# )
+
+# def generate_email_response(email_prompt):
+#     # Check input received by the function
+#     print("Received prompt:", email_prompt)
+
+#     # Determine if the input is a shorthand command or an actual email
+#     if 'email to' in email_prompt.lower():
+#         # Assume it's a shorthand command, format appropriately
+#         formatted_prompt = f'''
+#         Email received: "{email_prompt}"
+#         Respond to this email, ensuring a professional tone, providing a concise update, and addressing any potential concerns the sender might have.
+#         Response:
+#         '''
+#     else:
+#         # Assume it's direct email content
+#         formatted_prompt = f'''
+#         Email received: "{email_prompt}"
+#         Respond to this email, ensuring a professional tone, providing a concise update, and addressing any potential concerns the sender might have.
+#         Response:
+#         '''
+
+#     # Generate response using Llama-2 model
+#     try:
+#         response = lcpp_llm(
+#             prompt=formatted_prompt,
+#             max_tokens=256,
+#             temperature=0.5,
+#             top_p=0.95,
+#             repeat_penalty=1.2,
+#             top_k=150,
+#             echo=True
+#         )
+#         generated_response = response["choices"][0]["text"]
+#         # Remove the input part from the output if it is included
+#         if formatted_prompt in generated_response:
+#             generated_response = generated_response.replace(formatted_prompt, '').strip()
+#         print("Generated response:", generated_response)
+#         return generated_response
+#     except Exception as e:
+#         print("Error in response generation:", str(e))
+#         return "Failed to generate response, please check the console for errors."
     
 def classify_sentiment(text):
     # Encode the text using the tokenizer
@@ -215,15 +214,15 @@ iface_ner = gr.Interface(
     title="NER Analysis",
     description="Performs Named Entity Recognition using spaCy and Transformer models."
 )
-iface_response = gr.Interface(
-    fn=generate_email_response,
-    inputs=gr.Textbox(lines=10, placeholder="Enter the email prompt..."),
-    outputs=gr.Textbox(label="Generated Email Response"),
-    title="Email Response Generator",
-    description="Generate email responses using Llama-2 model."
-)
+# iface_response = gr.Interface(
+#     fn=generate_email_response,
+#     inputs=gr.Textbox(lines=10, placeholder="Enter the email prompt..."),
+#     outputs=gr.Textbox(label="Generated Email Response"),
+#     title="Email Response Generator",
+#     description="Generate email responses using Llama-2 model."
+# )
 
 # Using tabs to organize the interfaces
-tabs = gr.TabbedInterface([iface_category, iface_sentiment,iface_summary,iface_ner,iface_response], ["Category", "Sentiment"," Summary","NER","Response Generator"], css=custom_css)
+tabs = gr.TabbedInterface([iface_category, iface_sentiment,iface_summary,iface_ner], ["Category", "Sentiment"," Summary","NER"], css=custom_css)
 tabs.launch(share=True)