Docs.rs

High-performance LLM fine-tuning framework for Apple Silicon

Training loops and optimization for PMetal

HuggingFace Hub integration for PMetal

OpenAI-compatible inference server for PMetal

LoRA and QLoRA implementations for PMetal

Neural vocoder (BigVGAN) for text-to-speech synthesis

LLM model architectures for PMetal

Model merging toolkit for PMetal - SLERP, TIES, DARE, and more

Knowledge distillation toolkit for PMetal - KL divergence, hidden state alignment, and offline distillation

MLX backend implementation for PMetal LLM fine-tuning

Manifold-Constrained Hyper-Connections (mHC) for stable deep network training

GGUF file format implementation for PMetal

Metal GPU compute kernels for PMetal - FlashAttention and optimized ML primitives

Distributed training backend for PMetal (Home Clusters)

Dataset handling and preprocessing for PMetal