Isuckatcode

Joined 4 years ago

Karma

10

Submissions

23

Member since

Jul 14

View on Hacker News

Recent submissions

Comment · 6 months ago

Just throwing a hypothetical in case we hit an algorithmic plateau of three nines of safety and have to start optimizing other parts of the stack. No idea is silly imo :)

Ask HN: Hypothetical to robotics/autonomy engineers working on FSD

3 Isuckatcode 6 months ago 2 comments

Comment · 9 months ago

You beat me to it :D love the podcast. Too bad they stopped uploading from a while ago. You got any suggestions of podcasts in similar space? Nothing ever could scratch the itch like Damn Interesting

Comment · 1 year ago

Cool project! I wish there was a way to play Twitch streams on it. Maybe it’s something I’ll look into next weekend

Comment · 1 year ago

Man , I just wanted to enjoy celebrating Diwali with my family but been up from 3am trying to recover our services. There goes some quality time

Comment · 1 year ago

Stop trying to promote slop on HN. Under "How does it work" you just put "CVNextLevel is a tool that helps you get imediate feedback on how to improve your CV." 1. fix the grammar 2.That did not give us any insight. If this was an one night attempt of an cash grab AI wrapper, it was a bad one.

Comment · 2 years ago

Oh, I am sorry. I somehow read that as humane pin.

Comment · 2 years ago

I heard that humane doesn’t let users transfer ownership to the people [1]. They are willing to give a full refund for it

[1] https://x.com/snazzylabs/status/1838770446801735798?s=46

Comment · 2 years ago

I assume this would require a jail broken iPhone ? Or is it using accessibility feature to control the device ?

Comment · 2 years ago

I wish the password manager app allowed you to set a custom password to open the app. The iPhone password is something that people often share easily with those around them. With just the passcode, a person with bad intentions could easily gain access to all the passwords

Comment · 2 years ago

>By fine-tuning only the adapter layers, the original parameters of the base pre-trained model remain unchanged, preserving the general knowledge of the model while tailoring the adapter layers to support specific tasks.

From a ML noob (me) understanding of this, does this mean that the final matrix is regularly fine tuned instead of fine tuning the main model ? Is this similar to how chatGPT now remembers memory[1] ?

[1] https://help.openai.com/en/articles/8590148-memory-faq

Comment · 2 years ago

Do you recommend any good running shoes with velcro

Comment · 2 years ago

Can they recover the booster back to shore for an investigation?

Comment · 2 years ago

Here [1] is a reference to the token/sec of Llama 3 on different apple hardware. You can evaluate if this is an acceptable performance for your agents. I would assume the token/sec would be much lower if the LLM agent is running along the side as the game would also be using a portion of the CPU and GPU. I think this is something that you need to test out on your own to determine its usability.

You can also look into lower parameter models (3B for example) to determine if the balance between accuracy and performance fits under your usecase.

>Is there a way to reliably package these models with existing games and make them run locally? This would virtually make inference free right?

I don't have any knowledge on game dev so I can comment on this but yes, packaging it locally would make the inference free.

[1] https://github.com/ggerganov/llama.cpp/discussions/4167

Comment · 2 years ago

I was able to successfully run Llama 3 8B, mistral 7B, phi and other 7B models using Ollama [1] on my M1 MacBook Air.

[1] https://ollama.com