higgsfield-ai / higgsfield

Open-source GPU cluster orchestrator and PyTorch-based framework for training LLMs with billions to trillions of parameters.

FluidInference / FluidAudio

Swift SDK running open-source speech models (ASR, TTS, VAD, speaker diarization) locally on Apple devices via CoreML and the Neural Engine.