mjtechguy
mjtechguy on Hacker News
Former RedHatter here. This is a strong move on their part. I am definitely going to be checking this out.
Up until now I have been using Harvester (https://harvesterhci.io/) from Suse/Rancher and it’s based on similar concepts (KVM and kubevirt).
It works well but there are bugs and key features that are missing that would make it a true VMWare replacement.
Awesome. Thanks Price! Keep up the great work. I do love a good self hosted, well designed app.
The app looks, ok… looking forward to seeing it improve over time when it comes out of test flight.
I will say, the only reason why I haven’t moved my private comms with friend over to Zulip or RocketChat, and this is gonna sound petty, is because their mobile apps don’t let you insert gifs from the keyboard in your phone.
Using AI generated code will also make me a programmer I could never be. Been an infra and cloud guy, so this has been a game changer for me to actually create MVPs of things that have only ever existed in my head. I love it.
Cuz she maybe isn’t to everyone’s liking which is why we get a “choice”. I say “choice” cuz a 2 party system of handpicked elites isn’t really a choice, but I guess it’s better than a dictatorship.
Looks great! Is speaker diarization supported?
I see your point. How far do we go to protect the fools and is there infringement upon the non fools when everything has to be scoped down for the lowest common denominator?
No one is being forced to bet. The crime is not teaching people self sufficiency and good decision making. If heroin become suddenly legal I would not start using heroin. It’s a choice and it takes self control. It is no one’s responsibility but one’s own to make good decisions. “A fool and his money are easily parted” holds true here. Don’t be fools.
LLMs don’t currently “think”. So I assume a slow response is just a system load issue.
Nice. Does this support speaker diarization and output of subtitles to json format?
You win the internet today! Lolol
Would be interested to see what is in there. Luckily noone has posted the magnet link on Twitter.
Posted by
mjtechguy on Tuesday February 27, 2024 @11:24PM
from the show-hn dept.
https://github.com/mjtechguy/gemini-openai-api-proxy
Some other cool dudes made a serverless OpenAI API compatible proxy for Gemini here: https://github.com/PublicAffairs/openai-geminiI like that idea, but I don't like always relying on external vendors for my stuff. (Yes, ironic that I am using this to access an external vendor...).So I dockerized it so I can run it in my homelab. Now, maybe you can use it to do the same.
Very cool project. Can't wait to give it a try. I havent had much success with gpt-engineer, aider and others. I think this has the granularity for complex operations that is missing from other tools.
This is a really useful and well thought out app. Not without it's bugs, but Nazeem has been really responsive and cheerful about fixes. Curious if this is using RAG as well.
Any plans for a terraform provider module? Can’t use to many services that don’t support automation.
How is this different from a religion or a church. Seems like BS charges to me.
More likely that people have found better (often open source) frontends that interface with the OpenAI API, other commercial LLMs, or any of the excellent self hosted models via something like LocalAI. I know that is the case for me.
This looks pretty slick. Forgive me if this is obvious, but it says it loads “files”. What files can it load? Is there a list of file types that are supported?
The link to the conversation with ChatGPT is broken. Did you let chatgpt code that too? (JK). Interested to read the prompts.
I think having a bit of knowledge about the topic you are asking yields the best results. More than zero knowledge, but less than complete knowledge.
This is exactly what I thought when I read this.
Does it lean left? Or does the vocal majority lean left? Are we not hearing from the center and right for fear of ridicule, shadow banning, deplatforming, etc?