francoisfleuret

3 ms·

1. 1.▲
   by francoisfleuret 2y ago

   "there is no understanding, only prediction" I have no idea what this means.
2. 2.▲
   by francoisfleuret 2y ago

   This is 300M parameters model (1/1300th of the big llama-3) trained with 5M frames with 12 days of a GTX4090. This is what a big tech company was doing in 2015. The same stuff at industrial scale à la large LLMs would be absolutely min
3. 3.▲
   Diffusion for World Modeling(diamond-wm.github.io)

   476 points by francoisfleuret 2y ago|218 comments
4. 4.▲
   The Little Book of Deep Learning v1.2 (May 19th, 2024)(fleuret.org)

   1 points by francoisfleuret 2y ago|1 comments
5. 5.▲
   by francoisfleuret 2y ago

   This is a minor update of the "Little Book of Deep Learning", a phone-formatted free book for people with a STEM background that happened to be quite popular (580k+ download in a year). This update adds in particular a chapter abo