fix(ui): remove experimental webgpu tab and switch semantic cache back to stable numpy arrays

feat(webgpu): add actionable UI error message for enabling chrome webgpu flags

fix(webgpu): bump transformers.js to v3.8.1 to support lfm2 architecture

feat: add experimental WebGPU Liquid AI LFM prototype

feat: precompute semantic embeddings to drastically reduce startup time and prevent Gradio timeouts

fix: resolve massive duplicate array block in nursing_sections.json

fix: resolve remaining merge conflicts across all affected files

fix: remove duplicate code block from app.py caused by merge conflict

fix: resolve merge conflict in requirements.txt

fix: remove merge conflict marker from README to restore HF YAML config

feat: complete local embedding search with i-dot-ai HF model