File size: 18,853 Bytes
fd7c5f8 feb8f14 7209842 adcb6a8 fd7c5f8 feb8f14 ecb092b fd7c5f8 ecb092b fd7c5f8 ecb092b efb51ab ecb092b feb8f14 ecb092b 953203e ecb092b 5ccdfd6 da66b3b ecb092b fd7c5f8 ecb092b fd7c5f8 953203e fd7c5f8 efb51ab fd7c5f8 953203e fd7c5f8 ecb092b fd7c5f8 feb8f14 fd7c5f8 feb8f14 fd7c5f8 feb8f14 fd7c5f8 feb8f14 fd7c5f8 05bf0d3 a0d344e 05bf0d3 a0d344e 05bf0d3 efb51ab 1d40b26 feb8f14 c9a26fa feb8f14 c9a26fa feb8f14 efb51ab 05bf0d3 feb8f14 c9a26fa feb8f14 c9a26fa feb8f14 da66b3b feb8f14 da66b3b fd7c5f8 feb8f14 fd7c5f8 feb8f14 fd7c5f8 efb51ab fd7c5f8 da66b3b fd7c5f8 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 |
import gradio as gr
import os
import logging
from typing import List, Dict, Tuple
import numpy as np
from analyzer import combine_repo_files_for_llm, handle_load_repository
from hf_utils import download_filtered_space_files
# Setup logger
logger = logging.getLogger(__name__)
class SimpleVectorStore:
"""Simple in-memory vector store for repository chunks."""
def __init__(self):
self.chunks = []
self.embeddings = []
self.chunk_metadata = []
self.model = None
def _get_embedding_model(self):
"""Lazy load the embedding model."""
if self.model is None:
try:
from sentence_transformers import SentenceTransformer
self.model = SentenceTransformer('all-MiniLM-L6-v2') # Lightweight, fast model
logger.info("Loaded SentenceTransformer model for vectorization")
except ImportError:
logger.error("sentence-transformers not installed. Install with: pip install sentence-transformers")
raise ImportError("sentence-transformers package is required for vectorization")
return self.model
def add_chunks(self, chunks: List[str], metadata: List[Dict] = None):
"""Add text chunks and create embeddings."""
try:
model = self._get_embedding_model()
embeddings = model.encode(chunks, convert_to_tensor=False)
self.chunks.extend(chunks)
self.embeddings.extend(embeddings)
self.chunk_metadata.extend(metadata or [{} for _ in chunks])
logger.info(f"Added {len(chunks)} chunks to vector store")
except Exception as e:
logger.error(f"Error adding chunks to vector store: {e}")
def search(self, query: str, top_k: int = 3) -> List[Tuple[str, float, Dict]]:
"""Search for similar chunks using cosine similarity."""
if not self.chunks or not self.embeddings:
return []
try:
model = self._get_embedding_model()
query_embedding = model.encode([query], convert_to_tensor=False)[0]
# Calculate cosine similarities
similarities = []
for i, chunk_embedding in enumerate(self.embeddings):
similarity = np.dot(query_embedding, chunk_embedding) / (
np.linalg.norm(query_embedding) * np.linalg.norm(chunk_embedding)
)
similarities.append((self.chunks[i], similarity, self.chunk_metadata[i]))
# Sort by similarity and return top_k
similarities.sort(key=lambda x: x[1], reverse=True)
return similarities[:top_k]
except Exception as e:
logger.error(f"Error searching vector store: {e}")
return []
def clear(self):
"""Clear all stored data."""
self.chunks = []
self.embeddings = []
self.chunk_metadata = []
def get_stats(self) -> Dict:
"""Get statistics about the vector store."""
return {
'total_chunks': len(self.chunks),
'total_embeddings': len(self.embeddings),
'model_loaded': self.model is not None
}
# Global vector store instance
vector_store = SimpleVectorStore()
def vectorize_repository_content(repo_content: str, repo_id: str, chunk_size: int = 500) -> bool:
"""
Vectorize repository content by splitting into chunks and creating embeddings.
Args:
repo_content: The combined repository content
repo_id: Repository identifier
chunk_size: Number of lines per chunk
Returns:
bool: True if vectorization was successful
"""
try:
# Clear previous data
vector_store.clear()
lines = repo_content.split('\n')
chunks = []
metadata = []
# Split into chunks with overlap for better context
overlap = 50 # lines of overlap between chunks
for i in range(0, len(lines), chunk_size - overlap):
chunk_lines = lines[i:i + chunk_size]
chunk_text = '\n'.join(chunk_lines)
if chunk_text.strip(): # Only add non-empty chunks
chunks.append(chunk_text)
metadata.append({
'repo_id': repo_id,
'chunk_index': len(chunks) - 1,
'start_line': i,
'end_line': min(i + chunk_size, len(lines))
})
# Add chunks to vector store
vector_store.add_chunks(chunks, metadata)
logger.info(f"Successfully vectorized {len(chunks)} chunks for repository {repo_id}")
return True
except Exception as e:
logger.error(f"Error vectorizing repository content: {e}")
return False
def create_repo_explorer_tab() -> Tuple[Dict[str, gr.components.Component], Dict[str, gr.State]]:
"""
Creates the Repo Explorer tab content and returns the component references and state variables.
"""
# State variables for repo explorer
states = {
"repo_context_summary": gr.State(""),
"current_repo_id": gr.State("")
}
gr.Markdown("### ποΈ Deep Dive into a Specific Repository")
with gr.Row():
with gr.Column(scale=2):
repo_explorer_input = gr.Textbox(
label="π Repository ID",
placeholder="microsoft/DialoGPT-medium",
info="Enter a Hugging Face repository ID to explore"
)
with gr.Column(scale=1):
load_repo_btn = gr.Button("π Load Repository", variant="primary", size="lg")
with gr.Row():
visit_hf_link = gr.HTML(
value="",
label="π Repository Link",
visible=False
)
with gr.Row():
repo_status_display = gr.Textbox(
label="π Repository Status",
interactive=False,
lines=4,
info="Current repository loading status and vectorization info"
)
with gr.Row():
with gr.Column(scale=2):
repo_chatbot = gr.Chatbot(
label="π€ Repository Assistant",
height=400,
type="messages",
avatar_images=(
"https://cdn-icons-png.flaticon.com/512/149/149071.png",
"https://huggingface.co/datasets/huggingface/brand-assets/resolve/main/hf-logo.png"
),
show_copy_button=True,
value=[] # Start empty - welcome message will appear only after repo is loaded
)
with gr.Row():
repo_msg_input = gr.Textbox(
label="π Ask about this repository",
placeholder="What does this repository do? How do I use it?",
lines=1,
scale=4,
info="Ask anything about the loaded repository"
)
repo_send_btn = gr.Button("π€ Send", variant="primary", scale=1)
# with gr.Column(scale=1):
# # Repository content preview
# repo_content_display = gr.Textbox(
# label="π Repository Content Preview",
# lines=20,
# show_copy_button=True,
# interactive=False,
# info="Overview of the loaded repository structure and content"
# )
# Component references
components = {
"repo_explorer_input": repo_explorer_input,
"load_repo_btn": load_repo_btn,
"visit_hf_link": visit_hf_link,
"repo_status_display": repo_status_display,
"repo_chatbot": repo_chatbot,
"repo_msg_input": repo_msg_input,
"repo_send_btn": repo_send_btn,
# "repo_content_display": repo_content_display
}
return components, states
def handle_repo_user_message(user_message: str, history: List[Dict[str, str]], repo_context_summary: str, repo_id: str) -> Tuple[List[Dict[str, str]], str]:
"""Handle user messages in the repo-specific chatbot."""
if not repo_context_summary.strip():
return history, ""
# Initialize with repository-specific welcome message if empty
if not history:
welcome_msg = f"Hello! I'm your assistant for the '{repo_id}' repository. I have analyzed all the files and created a comprehensive understanding of this repository. I'm ready to answer any questions about its functionality, usage, architecture, and more. What would you like to know?"
history = [{"role": "assistant", "content": welcome_msg}]
if user_message:
history.append({"role": "user", "content": user_message})
return history, ""
def handle_repo_bot_response(history: List[Dict[str, str]], repo_context_summary: str, repo_id: str) -> List[Dict[str, str]]:
"""Generate bot response for repo-specific questions using comprehensive context and vector search."""
if not history or history[-1]["role"] != "user" or not repo_context_summary.strip():
return history
user_message = history[-1]["content"]
# Use vector search to find relevant chunks
relevant_chunks = vector_store.search(user_message, top_k=3)
# Build enhanced context using vector search results
vector_context = ""
if relevant_chunks:
vector_context = "\n\n=== MOST RELEVANT CODE SECTIONS ===\n"
for i, (chunk, similarity, metadata) in enumerate(relevant_chunks):
chunk_id = metadata.get('chunk_index', i)
start_line = metadata.get('start_line', 'unknown')
end_line = metadata.get('end_line', 'unknown')
vector_context += f"\n--- Relevant Section {i+1} (similarity: {similarity:.3f}, lines {start_line}-{end_line}) ---\n{chunk}\n"
# Create a specialized prompt using both comprehensive context and vector search results
repo_system_prompt = f"""You are an expert assistant for the Hugging Face repository '{repo_id}'.
You have comprehensive knowledge about this repository based on detailed analysis of all its files and components.
Use the following comprehensive analysis to answer user questions accurately and helpfully:
{repo_context_summary}
{vector_context}
Instructions:
- Answer questions clearly and conversationally about this specific repository
- Reference specific components, functions, or features when relevant
- Provide practical guidance on installation, usage, and implementation
- If asked about code details, refer to the analysis above and the relevant code sections
- Use the most relevant code sections to provide specific examples and implementation details
- Be helpful and informative while staying focused on this repository
- If something isn't covered in the analysis, acknowledge the limitation
Answer the user's question based on your comprehensive knowledge of this repository."""
try:
from openai import OpenAI
client = OpenAI(api_key=os.getenv("modal_api"))
client.base_url = os.getenv("base_url")
response = client.chat.completions.create(
model="Orion-zhen/Qwen2.5-Coder-7B-Instruct-AWQ",
messages=[
{"role": "system", "content": repo_system_prompt},
{"role": "user", "content": user_message}
],
max_tokens=1024,
temperature=0.7
)
bot_response = response.choices[0].message.content
history.append({"role": "assistant", "content": bot_response})
except Exception as e:
logger.error(f"Error generating repo bot response: {e}")
error_response = f"I apologize, but I encountered an error while processing your question: {e}"
history.append({"role": "assistant", "content": error_response})
return history
def get_huggingface_url(repo_id: str) -> str:
"""Generate the Hugging Face Spaces URL for a repository."""
if not repo_id.strip():
return ""
return f"https://huggingface.co/spaces/{repo_id}"
def generate_repo_link_html(repo_id: str) -> str:
"""Generate HTML with clickable link for the repository."""
if not repo_id or not repo_id.strip():
return ""
clean_repo_id = str(repo_id).strip()
hf_url = f"https://huggingface.co/spaces/{clean_repo_id}"
html_link = f'''
<div style="margin: 10px 0; padding: 15px; background: rgba(255, 255, 255, 0.1); border-radius: 12px; backdrop-filter: blur(10px); text-align: center;">
<a href="{hf_url}" target="_blank" style="display: inline-block; padding: 12px 24px; background: linear-gradient(45deg, #667eea, #764ba2); color: white; text-decoration: none; border-radius: 8px; font-weight: 600; font-size: 16px; transition: all 0.3s ease; box-shadow: 0 4px 12px rgba(0,0,0,0.2);">
π Visit {clean_repo_id} on Hugging Face
</a>
</div>
'''
return html_link
def handle_load_repository_with_vectorization(repo_id: str) -> Tuple[str, str, gr.HTML]:
"""Load repository and create both context summary and vector embeddings."""
if not repo_id.strip():
return "Status: Please enter a repository ID.", "", gr.update(value="", visible=False)
try:
logger.info(f"Loading repository with vectorization: {repo_id}")
# Download and process the repository (existing logic)
try:
download_filtered_space_files(repo_id, local_dir="repo_files", file_extensions=['.py', '.md', '.txt'])
combined_text_path = combine_repo_files_for_llm()
except Exception as e:
logger.error(f"Error downloading repository {repo_id}: {e}")
error_status = f"β Error downloading repository: {e}"
return error_status, "", gr.update(value="", visible=False)
# Read the combined content
with open(combined_text_path, "r", encoding="utf-8") as f:
repo_content = f.read()
# Create vectorized representation
vectorization_success = vectorize_repository_content(repo_content, repo_id)
# Get the original context summary
from analyzer import create_repo_context_summary
context_summary = create_repo_context_summary(repo_content, repo_id)
# Update status message
if vectorization_success:
status = f"β
Repository '{repo_id}' loaded successfully!\nπ Files processed and ready for exploration.\nπ Vector embeddings created for semantic search.\nπ¬ You can now ask questions about this repository."
else:
status = f"β
Repository '{repo_id}' loaded successfully!\nπ Files processed and ready for exploration.\nβ οΈ Vectorization failed - using text-only analysis.\nπ¬ You can now ask questions about this repository."
# Generate the HTML link for the repository
repo_link_html = generate_repo_link_html(repo_id)
logger.info(f"Repository {repo_id} loaded and processed successfully")
return status, context_summary, gr.update(value=repo_link_html, visible=True)
except Exception as e:
logger.error(f"Error loading repository {repo_id}: {e}")
error_status = f"β Error loading repository: {e}"
return error_status, "", gr.update(value="", visible=False)
def initialize_repo_chatbot(repo_status: str, repo_id: str, repo_context_summary: str) -> List[Dict[str, str]]:
"""Initialize the repository chatbot with a welcome message after successful repo loading."""
# Only initialize if repository was loaded successfully
if repo_context_summary.strip() and "successfully" in repo_status.lower():
# Check if vectorization was successful
vectorization_status = "π **Enhanced with vector search** for finding relevant code sections" if "Vector embeddings created" in repo_status else "π **Text-based analysis** (vector search unavailable)"
welcome_msg = f"π Welcome! I've successfully analyzed the **{repo_id}** repository.\n\nπ§ **I now have comprehensive knowledge of:**\nβ’ All files and code structure\nβ’ Key features and capabilities\nβ’ Installation and usage instructions\nβ’ Architecture and implementation details\nβ’ Dependencies and requirements\n\n{vectorization_status}\n\nπ¬ **Ask me anything about this repository!** \nFor example:\nβ’ \"What does this repository do?\"\nβ’ \"How do I install and use it?\"\nβ’ \"What are the main components?\"\nβ’ \"Show me usage examples\"\n\nWhat would you like to know? π€"
return [{"role": "assistant", "content": welcome_msg}]
else:
# Keep chatbot empty if loading failed
return []
def setup_repo_explorer_events(components: Dict[str, gr.components.Component], states: Dict[str, gr.State]):
"""Setup event handlers for the repo explorer components."""
# Load repository event with vectorization
components["load_repo_btn"].click(
fn=handle_load_repository_with_vectorization,
inputs=[components["repo_explorer_input"]],
outputs=[components["repo_status_display"], states["repo_context_summary"], components["visit_hf_link"]]
).then(
fn=lambda repo_id: repo_id,
inputs=[components["repo_explorer_input"]],
outputs=[states["current_repo_id"]]
).then(
fn=initialize_repo_chatbot,
inputs=[components["repo_status_display"], states["current_repo_id"], states["repo_context_summary"]],
outputs=[components["repo_chatbot"]]
)
# Chat message submission events
components["repo_msg_input"].submit(
fn=handle_repo_user_message,
inputs=[components["repo_msg_input"], components["repo_chatbot"], states["repo_context_summary"], states["current_repo_id"]],
outputs=[components["repo_chatbot"], components["repo_msg_input"]]
).then(
fn=handle_repo_bot_response,
inputs=[components["repo_chatbot"], states["repo_context_summary"], states["current_repo_id"]],
outputs=[components["repo_chatbot"]]
)
components["repo_send_btn"].click(
fn=handle_repo_user_message,
inputs=[components["repo_msg_input"], components["repo_chatbot"], states["repo_context_summary"], states["current_repo_id"]],
outputs=[components["repo_chatbot"], components["repo_msg_input"]]
).then(
fn=handle_repo_bot_response,
inputs=[components["repo_chatbot"], states["repo_context_summary"], states["current_repo_id"]],
outputs=[components["repo_chatbot"]]
) |