Cofounder / Head of AI at P-1 AI

I'm an ex Research Engineer at Google DeepMind & Microsoft. I run the 'The AI Epiphany' community. I'm also a proud father of 16 H100s (generously sponsored byHyperstack). And many more GPUs now at P-1 AI. :)

More on P-1 AI:
*Exclusive: What If AI could design a jet engine—or even a starship? Google DeepMind and Airbus veterans just raised $23M with an eye on that future
*P-1 AI Comes Out of Stealth, Aims to Build Engineering AGI for Physical Systems

Highlights:
* I'm one of 3 main devs on thellm.c project led byAndrej Karpathy

Most recent OSS contributions:

• llm.c - training LLMs in C/CUDA :)
• airoboros - synthetic instruction following data generation framework

My older recent projects:

• AI/tech jobs platformCracked Engineers
  
• Huberman Transcripts - transcripts, chapter/episode level summaries to help you learn faster (had to sunset it due to alegal notice)
• YugoGPT - trained SOTA 7B LLM for Croatian, Bosnian, Serbian, Montenegrin languages
  
• SlovenianGPT - trained SOTA 7B LLM for Slovenian language
  
• YugoChat - talk to YugoGPT
  
• First Serbian LLM eval
  
• Open-NLLB - Replicating Meta's "no language left behind" machine translation (MT) project
• Jarvis for images
• YouTube AI Buddy (removed from Chrome store)

PinnedLoading

1. llm.cllm.cPublic

   Forked fromkarpathy/llm.c

   LLM training in simple, raw C/CUDA

   Cuda 6

2. Open-NLLBOpen-NLLBPublic

   Effort to open-source NLLB checkpoints.

   Python 452 42

3. pytorch-GATpytorch-GATPublic

   My implementation of the original GAT paper (Veličković et al.). I've additionally included the playground.py file for visualizing the Cora dataset, GAT embeddings, an attention mechanism, and entr…

   Jupyter Notebook 2.6k 341

4. pytorch-original-transformerpytorch-original-transformerPublic

   My implementation of the original transformer model (Vaswani et al.). I've additionally included the playground.py file for visualizing otherwise seemingly hard concepts. Currently included IWSLT p…

   Jupyter Notebook 1k 177

5. get-started-with-JAXget-started-with-JAXPublic

   The purpose of this repo is to make it easy to get started with JAX, Flax, and Haiku. It contains my "Machine Learning with JAX" series of tutorials (YouTube videos and Jupyter Notebooks) as well a…

   Jupyter Notebook 742 108

6. pytorch-neural-style-transferpytorch-neural-style-transferPublic

   Reconstruction of the original paper on neural style transfer (Gatys et al.). I've additionally included reconstruction scripts which allow you to reconstruct only the content or the style of the i…

   Python 419 86