TagForge Files

[!IMPORTANT] Manual Update Required for v1.x Users The auto-update engine has been completely rewritten for better security. If you are upgrading from v1.x, the in-app updater may fail or crash. Please download v2.0 manually from the Releases page. Future updates (v2.0+) will work automatically.

🚀 New Features

• 👁️ Vision Intelligence:
  • VLM Support: Full support for Vision-Language Models (e.g., Llama 3.2 Vision, Qwen-VL). Attach images into the prompt generator for analysis.
  • Smart Detection: Automatically detects vision capabilities based on model metadata. Includes a Manual Override toggle for custom or unrecognized models.
  • Auto-Prompts: Context-aware injection of system prompts (e.g., "Describe every aspect...") when images are attached or vision personas are active.
• 🎭 Persona System 2.0:
  • Retired legacy defaults in favor of 4 high-performance presets: Natural Prompts, Tag Prompts, Natural Vision, and Tag Vision.
• 📜 Advanced Logging:
  • Complete overhaul of the console with collapsible details, log level coloring, and "One-Click Copy" functionality for easier debugging.

🛡️ Security & Stability

• Crash Reporting: Added automatic generation of detailed crash reports in the working directory for critical failures.
• Generation Reliability: Implemented "Stuck Prevention" logic to auto-recover if a model hangs during inference.
• Cross-Platform: Enhanced release workflows to ensure stricter compatibility checks for Linux and macOS.

🛠 Improvements

• Workflow: Optimized spacing constants for a cleaner layout on high-DPI screens.

🐛 Bug Fixes

• Fixed critical instability in the previous auto-update logic. (See note above).
• Fixed a false positive where "Update Available" would trigger on fresh installations.
• Resolved file locking conflicts that occasionally caused updates to fail.