Show HN: Boole-1.7B, a 1-bit LLM that outperforms Bonsai-1.7B

Category: ai-ml

Tags: 1-bit-quantization, llm, qwen3, llama-cpp, vllm

Score: 7.3/10 (Innovation: 7, Technical: 8, Documentation: 8, Utility: 6)

Boole-1.7B is a 1-bit quantized language model derived from Qwen3-1.7B, reducing weight storage to 247 MB while retaining 85.9% of the original's benchmark performance. It provides runtimes for llama.cpp and vLLM, making on-device inference practical for low-resource environments.

Target audience: AI/ML engineers, edge deployment developers

Repository: https://precisionnumerics.ai/blog/boole-1p7b · Python · Apache-2.0 · 4 stars

View on Hacker News