Training LLMs for spam classification take 2: I added 14 experiments comparing different approaches: github.com/rasbt/LLMs-fro…
- which token to train
- which layers to train
- different model sizes
- LoRA
- unmasking
- and more!
Any additional experiments you'd like to see?

We have reached 100,000 users! This milestone is incredibly humbling. 😊
When we developed Lightning AI Studio, our goal was to create a modern platform for the current AI landscape. Basically to enable researchers like myself to interactively use VSCode from a CPU to multiple

It was a super fun podcast with Hugo Bowne-Anderson last week! We talked all about developing LLMs and LLM research for two hours!
There’s also a half-hour live demo training a GPT model for classification in there 😊

A suggestion for an effective 11-step LLM summer study plan:
1) Read* Chapters 1 and 2 on implementing the data loading pipeline (manning.com/books/build-a-… & github.com/rasbt/LLMs-fro…).
2) Watch Karpathy's video on training a BPE tokenizer from scratch (youtube.com/watch?v=zduSFx…).
3)

I also sat down this weekend to summarize the latest LLM releases. What a month! We had four major open LLM releases: Mixtral, Meta AI's Llama 3, Microsoft's Phi-3, and Apple's OpenELM. In my new article (magazine.sebastianraschka.com/p/how-good-are…), I review and discuss all four of these major

June 5, join Sebastian Raschka (@LightningAI) for the #ACMTechTalk , 'Understanding the LLM Development Cycle: Building, Training, and Finetuning.' Marlene Mhangami (@Microsoft), Vice Chair of the ACM Practitioner Board, will moderate.
Register (free) to attend: bit.ly/3wpzlpH

Hey, if you are at PyCon next week and are looking for a ~3h intro to PyTorch (incl. goodies like mixed-precision & multi-GPU training, and of course LLM finetuning), I'll be there ☺️ github.com/rasbt/pycon2024

I am so excited that xLSTM is out. LSTM is close to my heart - for more than 30 years now. With xLSTM we close the gap to existing state-of-the-art LLMs. With NXAI we have started to build our own European LLMs. I am very proud of my team. arxiv.org/abs/2405.04517

going live with Sebastian Raschka in 15 minutes for Vanishing Gradients Podcast come say 👋!

we're covering the LLM life cycle, implementing LLMs from scratch, developing LLM libraries, training LLMs on the cloud, the latest models, and more!

youtube.com/live/qL4JY6Y5p…

I spent the day playing with Lightning AI ⚡️ studio, to speed up my NeurIPS experiments, and it is totally awesome! It’s just like developing in VSCode on your laptop, except it runs on a free cloud CPU (with persistent state and fast startup), and you can just press a button to

Pumped for tomorrow's livestream w/ Sebastian Raschka -- we've been prepping some fun stuff for you all so register for free below 💫

lu.ma/build-llms-fro…

the LLM life cycle, implementing LLMs from scratch, developing LLM libraries, training LLMs on the cloud, the latest models, and

If you are looking for something to code & read this weekend, I uploaded a notebook to finetune a small GPT model to classify SPAM messages with ~96% accuracy: github.com/rasbt/LLMs-fro…
(Fun fact: it's small enough to train it on your laptop; ~5 min on my M3 MacBook Air!)

That moment when one of your heroes sends you a signed copy of their new book... Thank you Sebastian Raschka!!! I love your book! Easy to read and super relevant! TRIPLE BAM!