Has any of you tried to run fastai on PlaidML? The promise is that it works with non-nVidia GPUs. Slower speed is not of a great issue for learning if it helps you to save $$ on buying expensive cards/cloud time.
I have just stumbled upon it in this reddit thread: https://www.reddit.com/r/MachineLearning/comments/8gxwfi/d_recommended_cloud_computing_services_for/, so I don’t have any personal experience with it yet.
PlaidML - A platform for making deep learning work everywhere.
PlaidML is the easiest, fastest way to learn and deploy deep learning on any device, especially those running macOS or Windows:
- Fastest: PlaidML is often 10x faster (or more) than popular platforms (like TensorFlow CPU) because it supports all GPUs, independent of make and model.
- PlaidML accelerates deep learning on AMD, Intel, NVIDIA, ARM, and embedded GPUs.
- Easiest: PlaidML is simple to install and supports Keras, ONNX
- Free: PlaidML is completely open source and doesn’t rely on any vendor libraries with proprietary and restrictive licenses.
… follow the link to read the rest of the document on plaidml :