Description

Needle 2 is a 45M-parameter model designed for tool calling, device use, and structured extraction on resource-constrained hardware like ESP32-S3 and Raspberry Pi 5. It utilizes a novel Simple Attention Network architecture featuring Hadamard MLPs and engram memory to minimize arithmetic load, fitting a single 14MB binary into 28MB of session RAM via training-aware 2-bit quantization.

Descriptions, tags, and model credits may be AI-generated or inferred from public sources and can be incomplete or wrong. Learn more.