sleepy/Gemma4Pad
1
0
Fork
You've already forked Gemma4Pad
 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions
No description
18 commits 1 branch 0 tags 195 KiB
  • Swift 98.8%
  • Shell 1.2%
Find a file
Exact
Repository files (latest commit first)
Filename Latest commit message Latest commit date
Kaloyan Nikolov f18f942e1a context slider, model pin, streamTick diagnostic, LazyVStack->VStack
2026-05-22 22:28:56 +02:00
Gemma4Pad context slider, model pin, streamTick diagnostic, LazyVStack->VStack 2026-05-22 22:28:56 +02:00
Gemma4Pad.xcodeproj feat: add app icon from downloaded image 2026-05-22 18:56:06 +02:00
.gitignore [app] Full Gemma4Pad iPadOS implementation (#4) 2026-05-22 12:59:11 +02:00
AGENTS.md docs: add AGENTS.md and ONBOARDING.md for future agents 2026-05-22 16:16:26 +02:00
ONBOARDING.md docs: add AGENTS.md and ONBOARDING.md for future agents 2026-05-22 16:16:26 +02:00
project.yml fix: disable extended-virtual-addressing entitlement for personal team 2026-05-22 18:52:52 +02:00
README.md initial: project scaffold 2026-05-22 12:17:56 +02:00
setup.sh [app] Full Gemma4Pad iPadOS implementation (#4) 2026-05-22 12:59:11 +02:00

README.md

Gemma4Pad

A minimal iPad app for locally running Gemma 4 E2B and E4B models on iPadOS.

Built with SwiftUI, MLX Swift, and the gemma-4-swift-mlx inference library.

Features

  • Text, image, and audio input
  • On-device inference via MLX (no cloud dependency)
  • Multiple chat sessions with conversation switching
  • Model selection (E2B / E4B) with quantization options (Q4 / Q8)
  • Web search integration via configurable SearXNG endpoint
  • Editable system prompt
  • Light/dark mode

Requirements

  • iPadOS 18.0+
  • Apple Silicon (M1+ or A16+)
  • Xcode 16+

Setup

  1. Open Gemma4Pad.xcodeproj in Xcode
  2. Select your iPad as the run destination
  3. Build and run