4 ms·
What's the advantage of using this over something like Huggingface Transformers, possibly with the MPS backend?
by miki123211 3y ago
What's the advantage of using this over something like Huggingface Transformers, possibly with the MPS backend?
- pshc 3y ago"MLX examples demonstrate how to run CoreNet models efficiently on Apple Silicon. Please find further information in the README.md file within the corresponding example directory." > mlx_example/clip: ... an example to convert CoreNet's CLIP model implementation to MLX's CLIP example with some customized modification. - FP16 Base variant: 60% speedup over PyTorch - FP16 Huge variant: 12% speedup > mlx_example/open_elm: ... an MLX port of OpenELM model trained with CoreNet. MLX is an Apple deep learning framework similar in spirit to PyTorch, which is optimized for Apple Silicon based hardware. Seems like an advantage is extra speedups thanks to specialization for Apple Silicon. This might be the most power-efficient DNN training framework (for small models) out there. But we won't really know until someone benchmarks it.
- HarHarVeryFunny 3y agoOpenELM (ELM = Efficient Language Models) has an unfortunate name clash with another LLM-related open source project. https://github.com/CarperAI/OpenELM https://github.com/CarperAI/OpenELM (ELM = Evolution through Large Models)
- jaimex2 3y agoNothing, its basically pytorch with an Apple logo.
- upbeat_general 3y agoThe implementation seems to be pretty clean and modular here where transformers (and diffusers) isn’t, unless you take their modules standalone. This repo has a lot of handy utilities but also a bunch of clean implementations of common models, metrics, etc. In other words, this is more for writing new models rather than inference.