Update app.py

app.py

@@ -3,12 +3,15 @@ import subprocess
 import random
 from huggingface_hub import InferenceClient
 import gradio as gr
-from safe_search import safe_search
+from safe_search import safe_search  # Make sure you have this function defined
 from i_search import google
 from i_search import i_search as i_s
 from datetime import datetime
 import logging
 import json
+import nltk  # Import nltk for the generate_text_chunked function
+
+nltk.download('punkt')  # Download the punkt tokenizer if you haven't already
 
 now = datetime.now()
 date_time_str = now.strftime("%Y-%m-%d %H:%M:%S")
@@ -112,15 +115,10 @@ def run_gpt(
     return resp
 
 def generate(prompt, history, agent_name=agents[0], sys_prompt="", temperature=0.7, max_new_tokens=2048, top_p=0.8, repetition_penalty=1.5, model="mistralai/Mixtral-8x7B-Instruct-v0.1"):
-    content = PREFIX.format(
-        date_time_str=date_time_str,
-        purpose=purpose,
-        safe_search=safe_search,
-    ) + prompt_template.format(**prompt_kwargs)
-    if VERBOSE:
-        logging.info(LOG_PROMPT.format(content))  # Log the prompt
-    
-    stream = client.text_generation(content, **generate_kwargs, stream=True, details=True, return_full_text=False)
+    # Use 'prompt' here instead of 'message'
+    formatted_prompt = format_prompt(prompt, history, max_history_turns=5)  # Truncated history
+    logging.info(f"Formatted Prompt: {formatted_prompt}")
+    stream = client.text_generation(formatted_prompt, temperature=temperature, max_new_tokens=max_new_tokens, top_p=top_p, repetition_penalty=repetition_penalty, stream=True, details=True, return_full_text=False)
     resp = ""
     for response in stream:
         resp += response.token.text
@@ -320,6 +318,20 @@ def generate(
         temperature = 1e-2
     top_p = float(top_p)
 
+    # Add the system prompt to the beginning of the prompt
+    formatted_prompt = f"{system_prompt} {prompt}"
+
+    # Use 'prompt' here instead of 'message'
+    formatted_prompt = format_prompt(formatted_prompt, history, max_history_turns=5)  # Truncated history
+    logging.info(f"Formatted Prompt: {formatted_prompt}")
+    stream = client.text_generation(formatted_prompt, temperature=temperature, max_new_tokens=max_new_tokens, top_p=top_p, repetition_penalty=repetition_penalty, stream=True, details=True, return_full_text=False)
+    resp = ""
+    for response in stream:
+        resp += response.token.text
+
+    if VERBOSE:
+        logging.info(LOG_RESPONSE.format(resp))  # Log the response
+    return resp
 
 
 def generate_text_chunked(input_text, model, generation_parameters, max_tokens_to_generate):
@@ -476,38 +488,32 @@ def project_explorer(path):
     Displays the file tree of a given path in a Streamlit app.
     """
     tree = get_file_tree(path)
-    display_file_tree(tree)
+    tree_str = json.dumps(tree, indent=4)  # Convert the tree to a string for display
+    return tree_str
 
 def chat_app_logic(message, history, purpose, agent_name, sys_prompt, temperature, max_new_tokens, top_p, repetition_penalty, model):
     # Your existing code here
     
     try:
-        # Attempt to join the generator output
+        # Pass 'message' as 'prompt'
         response = ''.join(generate(
             model=model,
-            message=message,
-            stream=True,
-            temperature=0.7,
-            max_tokens=1500
+            prompt=message,  # Use 'prompt' here
+            history=history,
+            agent_name=agent_name,
+            sys_prompt=sys_prompt,
+            temperature=temperature,
+            max_new_tokens=max_new_tokens,
+            top_p=top_p,
+            repetition_penalty=repetition_penalty,
         ))
     except TypeError:
-        # If joining fails, collect the output in a list
+        # ... (rest of the exception handling)
+
         response_parts = []
         for part in generate(
             model=model,
-            message=message,
-            stream=True,
-            temperature=0.7,
-            max_tokens=1500
-        ):
-            if isinstance(part, str):
-                response_parts.append(part)
-            elif isinstance(part, dict) and 'content' in part:
-                response_parts.append(part['content']),
-        
-        response = ''.join(response_parts,
-        # Run the model and get the response (convert generator to string)
-            prompt=message,
+            prompt=message,  # Use 'prompt' here
             history=history,
             agent_name=agent_name,
             sys_prompt=sys_prompt,
@@ -515,11 +521,17 @@ def chat_app_logic(message, history, purpose, agent_name, sys_prompt, temperatur
             max_new_tokens=max_new_tokens,
             top_p=top_p,
             repetition_penalty=repetition_penalty,
-            model=model # Pass the model argument here
-                          )
+        ):
+            if isinstance(part, str):
+                response_parts.append(part)
+            elif isinstance(part, dict) and 'content' in part:
+                response_parts.append(part['content'])
+
+        response = ''.join(response_parts)
         history.append((message, response))
         return history
 
+    history.append((message, response))
     return history
 
 def main():