Gemma Tuner Multimodal vs VibeVoice

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.

VibeVoice is Microsoft's open-source family of voice AI models, comprising three specialized systems: a 7B-parameter ASR model that transcribes up to 60 minutes of audio in a single pass with speaker diarization and hotword support, a 1.5B TTS model that can synthesize up to 90 minutes of multi-speaker speech, and a lightweight 0.5B streaming TTS engine with ~300ms latency. All three are MIT licensed, published to Hugging Face, and come with Google Colab notebooks for quick experimentation. Under the hood, VibeVoice uses continuous speech tokenizers operating at an ultra-low 7.5 Hz frame rate, combining an LLM backbone for semantic understanding with a diffusion head for fine-grained acoustic detail. This architecture is designed to handle long-form audio without the chunking artifacts that plague most open-source speech models. The release is particularly notable for the indie builder community because the MIT license has no commercial restrictions baked into the model weights — though Microsoft does warn against production use without further testing and flags deepfake risks explicitly. With 45,000+ GitHub stars in under 48 hours, it's clear the community has been waiting for a serious open-weight voice stack that covers the full pipeline.