v1.6.0

• Modify llm-yolo to support asynchronous inference
• Add yolo.boxV2 return structure to llm-yolo
• Modify llm-llm to support asynchronous inference
• Optimize inference speed for llm-llm
• Optimize image data output for llm-camera
• Add openai-api module to provide OpenAI API services
• Add qwen2.5-1.5B-Int4-ax630c Int4 quantized model
• Add qwen2.5-0.5B-Int4-ax630c Int4 quantized model

V1.5.0

• Added experimental support for Axera VIN in llm-camera.
• Added experimental support for BSON in llm-sys.
• Added new models to llm-llm.
• Optimized LLM post-processing algorithms.
• Fixed the issue of incorrect token truncation in LLM output.
• Updated the internVL2.5-MPO model.
• Added the Deepseek-r1-distill-qwen2.5-0.5b model.
• Added a new prefill 256 LLM model.
  Warning: This upgrade package requires installing lib-llm_1.6-m5stack1_arm64.deb first before installing other model packages.

v1.4.0: Merge pull request #4 from m5stack/dev

• Add whisper unit
• Add whisper-tiny model
• Add vad unit
• Add silero-vad model
• Add Qwen2.5-1.5B model
• Add depth_anything unit
• Add depth-anything model
• Optimize the initialization speed of the kws unit
• Optimize the initialization speed of the tts unit

LLM V1.3.0 Firmware Release Note

• add yolo unit.
• experimental add vlm unit.
• add camera unit.
• llama3.2-1B-prefill-ax630c llm mode.
• openbuddy-llama3.2-1B-ax630c llm mode.
• qwen2.5-coder-0.5B-ax630c llm mode.
• llm unit add tokenizer server.
• melotts unit fix english support.
• experimental add vlm unit internvl2-1B-ax630c mode.

rename StackFlow

v1.2.0

[update] compile tool link