FezzikTheGiant

13 ms·

1. 61.▲
   by FezzikTheGiant 3y ago

   I asked this another place in this thread, but curious to know how you plan to run and deploy this model. What's the cheapest or I guess most cost efficient way to do this without burning money. I'm a college student for reference
2. 62.▲
   by FezzikTheGiant 3y ago

   What's the cheapest way to run and deploy a fine-tuned model today? If I understand correctly I'll need to run cloud GPUs right
3. 63.▲
   Ask HN: How much does it cost to fine-tune an LLM? What services cam I use?

   1 points by FezzikTheGiant 3y ago|0 comments
4. 64.▲
   by FezzikTheGiant 3y ago

   I was just gonna ask this question and saw this at the top of Ask. Interested.
5. 65.▲
   by FezzikTheGiant 3y ago

   Thanks! As a CS student interested in learning more about this space how do you recommend I get started? I'm pretty early in my education so I kind of want to learn how to drive the car for now and learn how the engine works more forma
6. 66.▲
   by FezzikTheGiant 3y ago

   look forward to it. this is really cool
7. 67.▲
   by FezzikTheGiant 3y ago

   Appreciate the response. Really cool work!
8. 68.▲
   by FezzikTheGiant 3y ago

   Curious to know how you built this. Is it GPT-4 or a fine-tuned model. How much does it cost?
9. 69.▲
   by FezzikTheGiant 3y ago

   If you're open to it, it would be great if you could make a post explaining how you built this. Even if it's brief. Trying to learn more about this space and this looks pretty cool. And ofc, nice work!
10. 70.▲
    by FezzikTheGiant 3y ago

    This is really cool. Does anyone who knows this stuff better explain how it works on the backend? The founders mentioned 6 different models but how do they interact with each other so fast? There’s definitely an LLM in the back interpreting
11. 71.▲
    by FezzikTheGiant 3y ago

    yeah that would be pretty cool.
12. 72.▲
    by FezzikTheGiant 3y ago

    I think instead of building your own client for displaying the posts, if you linked it with substack/twitter and others it would be a useful tool.
13. 73.▲
    by FezzikTheGiant 3y ago

    I built something like this a couple years ago. very cool!
14. 74.▲
    by FezzikTheGiant 3y ago

    At Berkeley for most of my CS classes it is always pretty crowded, there is a long wait time to get to talk to a TA for 4-5 minutes.
15. 75.▲
    by FezzikTheGiant 3y ago

    This is how I approach things as well. I never got the appeal of office hours as TAs aren't going to be able to regurgitate lecture content for you, it only works if you have something extremely specific you need help with.