PipeOwl-1.4-multilingual (Geometric Embedding)

• base = cosine similarity in embedding space
• Δfield = static scalar field bias

• O(n) over vocabulary.
• No attention.
• No transformer weights.

Changes in 1.4

• Embedding storage converted from FP32 to FP16
• Model size reduced from ~1.9GB → ~1.01GB
• Startup time improved from ~30s → ~2s
• Same scoring pipeline

Architecture

• Static embedding table (V × D)
• Aligned vocabulary index
• Optional scalar bias field
• Linear scoring
• Pluggable decoder stage
• Targeted for CPU environments and low-latency systems (e.g. IME).

Model Specs

Attribution

Quickstart

See full experimental notes here:

Example:

Repository Structure

Multilingual Vocabulary

• Chinese words
• English words
• Mathematical symbols
• Symbolic / byte fallback tokens

PipeOwl 是一個基於靜態語義場的幾何檢索系統。

• base = embedding cosine similarity
• delta = 靜態場偏移量
• α / β 為可調權重

LICENSE