Imagine how big the startup ecosystem for LLMs would be in 2021’s macroeconomic environment.
I’ve always wondered about this. When I went to Europe there were several days I walked 13+ miles. Probably can count on one hand how many times I’ve done that in the US.
Man eats less than 400 calories for breakfast while in Italy. Man walks 15,000 steps while on holiday. Man loses 3 pounds. Man blames it on “poisonous food in the US”. Peak Twitter right here.
types slow me down which is why i take extra time to meticulously litter all of my python code with type hints that dont actually do anything
Are you even working if you’re not 6.
This Section 174 is baffling. All software R&D costs now have to be amortized over 5 years if done within the country, and an insane 15 years outside. Means you end up with phantom profits that don't match expenses, so big tax bills. Terrible for US software industry.
Hey chat, do we think this government hiring program will be successful?
The recent AI executive order calls for a technical talent surge into government. To help with this, OPM just authorized a government-wide direct hire authority for AI roles. Using flexible hiring authorities is going to be essential for building state capacity in this area.
I wasn’t going to buy it, but then I saw it had Meyer lemons for only 1.5M!
"Never attribute to malice that which is adequately explained by stupidity." I found out there’s a word for this today, Hanlon’s Razor.
Interesting. When teachers are poorly compensated, education quality will decline.
If the in-laws are asking about your education choices this Christmas - ("Your kids are going to be weird if they don't go to public school!") Here are 9 stats you should share with them 👇
Legal question: Do states have the authority to adjudicate violations of the Federal Constitution?
Excellent upgrade for every Python developer. I've been using the following three libraries for years: • flake8 - Linting • black - Formatting • isort - Sorting imports Many people recommended a replacement: "ruff." I installed it, and I've been playing with it for a couple…
Excited about recent improvements of our NeurIPS Spotlight paper, now even faster with ⚡️⚡️multiprocessing⚡️⚡️! We now get 2x speedup on as low as 50-step DDIM, and 4x speedup on 200-step DDIM! The first version of our paper showed good results, but we wanted even better. -…
Diffusion models are slow to sample from. Many methods propose to sample using *fewer* steps, but this can hurt sample quality. We introduce ParaDiGMS, a new method to speed up diffusion models by 2-4x while using the *same* number of steps! arxiv.org/abs/2305.16317
Student prior, normal likelihood y ~ Normal(mu,1) mu ~ Student(2,10,1) Now the likelihood dominates - it's thin tails are very skeptical of the prior, but the prior's thick tails not so surprised by the likelihood
United States Trends
- 1. #UFC309 77,9 B posts
- 2. #MissUniverse 254 B posts
- 3. Beck 16,3 B posts
- 4. Oregon 27 B posts
- 5. Dinamarca 5.334 posts
- 6. Brian Kelly 9.994 posts
- 7. Jim Miller 2.820 posts
- 8. Mac Miller N/A
- 9. Dalton Knecht 5.162 posts
- 10. #AEWCollision 13,6 B posts
- 11. Nigeria 226 B posts
- 12. Wisconsin 45,5 B posts
- 13. Inside the NBA 5.821 posts
- 14. Tennessee 41,9 B posts
- 15. Martinez 70,8 B posts
- 16. #DevilWearsHope 7.849 posts
- 17. Tailandia 3.715 posts
- 18. Feds 37,1 B posts
- 19. Chile 188 B posts
- 20. Nebraska 12,8 B posts
Something went wrong.
Something went wrong.