YAML Metadata Warning:empty or missing yaml metadata in repo card
Check out the documentation for more information.
๐ง Conversation Memory System
A self-contained, local-first conversation memory system that replaces the fraudulent MemPalace project with real working code.
Why not MemPalace?
MemPalace is a scam repository โ no actual source code, fake 22K+ stars, empty PyPI package. This system is the real deal: SQLite + sentence-transformers, zero external API dependencies.
Features
- ๐พ SQLite persistence โ conversations stored locally
- ๐ Text search โ find any past conversation
- ๐งฎ Vector embeddings โ semantic search via sentence-transformers (optional)
- ๐ค Export โ JSON or Markdown output
- ๐ฅ๏ธ CLI โ command-line interface for search/retrieve
- ๐ Zero external APIs โ works offline
Quick Start
pip install sentence-transformers
python memory_system.py
CLI Usage
# Search memories
python memory_cli.py search "ZeroGPU"
# List all threads
python memory_cli.py threads
# Show a specific thread
python memory_cli.py show cydonia-space-setup-20260430-215601
# Export all memories
python memory_cli.py export --all
Python API
from memory_system import ConversationMemory
memory = ConversationMemory(user_id="scottzilla")
# Save a message
memory.save_message("user", "Hello!", thread_id="my-thread")
# Save a full conversation
memory.save_conversation([
{"role": "user", "content": "Hi"},
{"role": "assistant", "content": "Hello!"}
], thread_id="my-thread")
# Search
results = memory.search("model loading")
# Get thread history
messages = memory.get_thread("my-thread")
# Export
memory.export_to_markdown("conversation.md", "my-thread")
Files
| File | Purpose |
|---|---|
memory_system.py |
Core memory class |
memory_cli.py |
Command-line interface |
Saved Conversations
This repo includes the exported conversation about fixing the Cydonia-24B Space:
cydonia-space-setup-20260430-215601.jsoncydonia-space-setup-20260430-215601.md
Inference Providers NEW
This model isn't deployed by any Inference Provider. ๐ Ask for provider support