Gemma Tuner Multimodal vs GitNexus

Gemma Tuner Multimodal is an open-source fine-tuning toolkit for Google's Gemma 4 and Gemma 3n models that runs entirely on Apple Silicon using PyTorch with Metal Performance Shaders (MPS) backend — no NVIDIA GPU or cloud infrastructure required. It supports LoRA training on multimodal inputs: audio, images, and text simultaneously, using local CSV files or streamed from Google Cloud Storage or BigQuery. The tool targets the growing segment of developers who own M-series Macs but have been locked out of fine-tuning workflows that assume CUDA availability. Gemma 4's architecture is particularly well-suited to this use case: its 4B multimodal variant (designed for on-device deployment) trains efficiently on M3 Max and M4 Pro hardware within the available unified memory constraints. Primary use cases include medical transcription fine-tuning (audio → text with clinical terminology), visual QA systems (image + text → structured response), and private on-device pipelines where cloud API calls are prohibited by compliance requirements. The project fills a specific niche that Google's own fine-tuning documentation doesn't cover well for Apple hardware.

GitNexus builds a client-side knowledge graph of any GitHub repository or ZIP file, giving AI coding agents genuine architectural awareness. The browser-based UI runs entirely in WebAssembly — no server, no data upload — and renders an interactive dependency graph you can explore and query via a built-in Graph RAG agent. The CLI mode launches an MCP server that connects directly to Claude Code, Cursor, Codex, and Windsurf. Once connected, agents can run blast radius analysis before making changes, do hybrid semantic + structural search across the codebase, trace dependency chains, and auto-generate or update CLAUDE.md configuration files. The underlying graph is built using a combination of AST parsing and embedding-based similarity. The project exploded on GitHub Trending on April 8, 2026 — picking up over 1,100 stars in a single day to reach nearly 25,000 total. It addresses a real pain point: AI coding agents frequently break things because they lack a global model of the codebase structure. GitNexus bridges that gap without sending your code anywhere.