Pinned
One year and half after starting the first draft of the first chapter, look what arrived in the mail!
Sylvain Gugger
1,295 posts
Sylvain Gugger
@GuggerSylvain
Machine Learning at Jane Street. Previously at @huggingface and @fastdotai Co-author of github.com/fastai/fastbook He/him
- Finished the final review of the book today (glad this is over 😅). It should go to the printer next week!
- There are a lot of models in the transformers🤗 repo. Feeling lost? I know I was, so I made a little high-level summary of the differences between each model. huggingface.co/transformers/s…
- Want to run your PyTorch training loop on multi-GPUs or TPUs without using an abstract class you can't control or tweak easily? Try out 🤗 Accelerate!
- So excited to share what I've been working on this past month. Not enough RAM to load the model? No problem. Not enough GPU RAM to host the model? No problem. As long as it fits on your hard drive, we'll make it run.Last week @metaai publicly released huge LMs, with up to ☄️30B parameters. Great win for Open-Source🎉 These checkpoints are now in 🤗transformers! But how to use such big checkpoints? Introducing Accelerate and ⚡️BIG MODEL INFERENCE⚡️ Load & USE the 30B model in colab (!)⬇️
- If you at the PyTorch conference, come by the Jane Street booth and say hi!
- Today is my last day at @fastdotai. It has been a lot of fun working there for two years, and I'm pretty proud of what I helped accomplish: being a core contributor of two versions of the library, helping with two iterations of the MOOC and co-writing a book with @jeremyphoward
- Yesterday was my last day at Hugging Face. The past three years have been exhilarating and I am very proud of what the team has accomplished during that time! Taking a bit of a break with opensource full time (though I will still contribute to Transformers and Accelerate)
- Training a transformer model for text classification has never been easier. Pick a model checkpoint from the 🤗Transformers library, a dataset from the dataset library and fine-tune your model on the task with the built-in Trainer!
- During all this time I was struggling with git, deleting and re-cloning every other week, there was a tool with a nice UI to handle all the stuff I need and no one told me?
- We had an awesome talk at Jane Street from the amazing @cHHillee on scaling ML systems to and I just realized the recording is now online:
