picolm_flutter 0.0.2

On-device LLM inference engine for Flutter. Run LLaMA-architecture models via Dart FFI — no cloud, no internet, no API keys.

0.0.2

• Updated example app with streaming UI and ChatML template formatting.
• Fixed heavy operations (model load/generation) running on UI thread — now reliably using background Isolates.
• Added logo and usage instructions to README.md.

0.0.1

• Added PicoLM class with load(), generate(), and generateFull().
• Background isolate support for load() and generate().
• Support for GGUF model files.
• JSON grammar mode (setJsonMode).
• Streaming token output.
• iOS, Android, and macOS platform support.