June 1st Flounder meeting - ML hack day
Russell Coker via linux-aus <[email protected]> Mon, 27 May 2024 00:34:54 +1000
| Newsgroups | gmane.linux.australia |
|---|---|
| Message-ID | <2197988.Icojqenx9y@cupcakke> |
https://flounder.linux.org.au/events/flounder-jun-2024-ml-hack-day/ June ML Hack Day. A day of learning about Machine Learning by trying stuff out. I have a LicheePi4A which has an NPU built in and a server with a NVidia card that I can give access to. But I encourage other people to try to have some hardware suitable for their own use, a minimum would be a 16G of DDR4 RAM (as a rough guide a system with DDR4 has adequate CPU performance) and 16G of free storage space. Hardware for ML is a really good thing, ideally a GPU with 8G of RAM or more but you can do some useful stuff with CPU if you are patient. Meeting will be at http://b.coker.com.au. No need to register just click on the link on the day. Meeting officially opens at 12pm Melbourne time (02:00UTC) on the 1st of June. A little earlier than usual, but if people from Perth miss the first hour they can join us for the later parts of the meeting as it’s not a formal meeting. Here are some links that are worth reviewing before the meeting (not required but recommended): https://www.tensorflow.org/lite – libraries for doing ML on small devices (Android, iOS, and Raspberry Pi). https://huggingface.co/ – HuggingFace AI community with models to download. https://github.com/EleutherAI/the-pile – The Pile of open training data. https://blends.debian.org/science/tasks/machine-learning – Debian Science machine learning packages. https://wiki.debian.org/MachineLearning/Tensorflow https://github.com/mitutitu16/Awesome-Mobile-Machine-Learning Links to useful resources. -- My Main Blog http://etbe.coker.com.au/ My Documents Blog http://doc.coker.com.au/ _______________________________________________ linux-aus mailing list [email protected] http://lists.linux.org.au/mailman/listinfo/linux-aus To unsubscribe from this list, send a blank email to [email protected]