Reminder to submit your nominations for PyTorch Contributor Awards 👇
PyTorch Contributor Awards are back 🔥💜 These awards honor individuals who have made significant contributions to the PyTorch ecosystem.This year, the nominating committee is excited to involve the community in the process. Suggest a nominee here: https://hubs.la/Q02CGbml0
🔥 7th Belgian Pytorch Meetup (1st of the year) is almost here! By now you know the drill ;) Pytorch talks, good people, food, drink. Our two speakers will have interesting thoughts to share with us: 1. Opening the black box: how to interpret a PyTorch model using Captum - Maxime Markov 2. A deep dive into mixture of experts - Cyril Maréchal Hesitate no more, it's time to sign up!
ML Tribe Lead at Devoteam G Cloud | Helping companies leverage AI & ML on Google Cloud ☁️
We are almost there 13/09 is the next Belgian Pytorch Meetup! We will focus on the integrations between Pytorch and Keras Core, share some concrete examples and go over caveats to consider! Next to that, as always, we have great food and lots of interesting people! https://lnkd.in/ePwS6Mig #pytorch #meetup #belgianpytorchmeetup
Google Developer Expert (GDE), AWS Community Builder, Senior Manager Data Science, Consultant, Trainer, Podcaster, Founder Malaysia R User Group, AI & ML Malaysia User Group
Check out what's new in PyTorch Documentation. Featuring 3 new tutorials and PyTorch 2.2 documentation + tutorials! #pytorch Check out here : https://lnkd.in/dehHw3Y2
The story behind EduSummarise, the mind maps creator 🧙♂️🔮 Two weeks ago, I got the opportunity to compete at TrojanHacks, a 24-hour hackathon competition held by USC ACM (uscacm.org), where I worked in a team of 4 on a project focusing on the Education pillar. We went from an exciting idea of a system that helps creating notes, mind maps, and personalized quizzes from video lectures for students to review lectures more effectively, to a very lame looking UI, and eventually to a GitHub repository showing only static images of LLM-generated mind maps on the presentation day. This happened because we couldn't fully commit to the project (2 of us got interview calls on the next day so of course they had to focus on what's more important, the other one tried to stay and work remotely, but I guess he was too upset to show up for the presentation, because we had nothing to present!). I almost gave up as well, but I still decided to present alone. I now for sure am proud of myself for being so brave! The judges and the audience, however, were really nice and supportive, and I am really thankful to have presented to them. Although there are tons to improve, here are some mind maps created by our EduSummarise🧙♂️🔮 framework, based on a video lecture on Lecture 3: File systems, from USC Computer Science Foundational data management class. Source code of EduSummarise🧙♂️🔮 can be found in our repository (https://lnkd.in/gbfP5K4b). All questions are welcome!
Introducing the PyTorch Community Blog 🎉 New today, we've added a new section on our website to highlight the amazing work being done by our community. See the latest ecosystem tools in today's blog: https://hubs.la/Q02w-GyP0 💜 To join our ecosystem, apply here: https://hubs.la/Q02w-JXj0
Catch Yamini Nimmagadda, Raymond Lo 👓🤖 and Myself at the #PyTorch Conference on October 16-17th in San Francisco! In our session(s), you’ll get a chance to learn about: - OpenVINO’s support for PyTorch 2.0 and how you can use OpenVINO as a backend through torch.compile - OpenVINO’s support for PyTorch models with the convert_model() API - How you can quantize and compress model weights using NNCF Looking forward to seeing you there! #iamintel #openvino #quantization #session #deeplearning
The PyTorch Conference schedule is now LIVE! 🥳💜 We have a great lineup of speakers planned, register to attend October 16-17 in San Francisco! Schedule 👉 https://hubs.la/Q01_4gYs0
Hello everyone! We're sorry to let you know that the "Hack-una Matata" machine learning hackathon is getting postponed. Some unexpected issues came up, and we want to make sure everything runs smoothly. We know you were excited, and we apologize for any inconvenience. We're working on figuring out a new date, and we'll let you know as soon as it's set. Thanks for your understanding, and we're eager to bring you a great hackathon soon! #gdsc #gdscisi #hackunamatata #machinelearning
Member of Research Group AIKE. University of Murcia. Alpha Testing Consultant at deeplearning.ai - Data Scientist Hosp Rafael Méndez
Quantization LLM course!
🤗 Just launched: Quantization in Depth, a short course built in collaboration with Hugging Face and taught by Marc Sun and Younes Belkada. Following up on our previous offering, Quantization Fundamentals, in this new course you’ll build your own linear quantizer from scratch, going beyond standard libraries such as Quanto. ⚡ Implement linear quantization, choosing between symmetric and asymmetric mode. ⚡ Try out per-tensor, per-channel, and per-group quantization granularities. ⚡ Build your own quantizer in PyTorch, to quantize any open source model's dense layers from 32 bits to 8 bits. Eager to go further into quantizing models? This course is the right next step! Enroll for free: https://hubs.la/Q02w8rTW0
PhD Candidate in Machine Learning | 3+ Years as Data Scientist & Machine Learning Engineer | Specializing in Large Language Models
I just finished Deep Learning AI's new course, "Quantization in Depth," which effectively lays out the fundamentals of quantization mathematically. This course should truly be called the real "Quantization Fundamentals." There's another very introductory course that serves as a perfect starting point. I'd suggest taking both courses as a combo for anyone interested in understanding quantization.
🤗 Just launched: Quantization in Depth, a short course built in collaboration with Hugging Face and taught by Marc Sun and Younes Belkada. Following up on our previous offering, Quantization Fundamentals, in this new course you’ll build your own linear quantizer from scratch, going beyond standard libraries such as Quanto. ⚡ Implement linear quantization, choosing between symmetric and asymmetric mode. ⚡ Try out per-tensor, per-channel, and per-group quantization granularities. ⚡ Build your own quantizer in PyTorch, to quantize any open source model's dense layers from 32 bits to 8 bits. Eager to go further into quantizing models? This course is the right next step! Enroll for free: https://hubs.la/Q02w8rTW0
After writing tens of thousands of lines of code, I've experienced thousands of painful and dreadful PyTorch errors. Now, I want to spare you the pain so I published a guide on how to debug PyTorch 50% faster! https://buff.ly/4bm0glx Let me know your thoughts!
