Killer_Tree @ Killer_Tree @sh.itjust.works

Posts

0

Comments

77

Joined

2 yr. ago

Exactly! If you only want to use a Large Language Model (LLM) to run your own local chatbot, then using a quantized version will dramatically improve speed and performance. It also allows consumer hardware to run larger models which would otherwise be prohibitively resource intensive.

You can, and it's easier than you might think! Check out a platform like Oobabooga and find a nice 4-bit quantized LLM of a flavor you prefer. Check out TheBloke on hugging face, they quantized a ton of great LLMs.

ICQ... Now there's a name I haven't heard in a long time. I used to use the crap out of that service, but I can't for the life of me recall what killed it anymore than I can recall migrating from AIM.

Give https://lemmy.dbzer0.com/c/piracy a try!

Edit: Lemmy link - !piracy@lemmy.dbzer0.com

You can disagree without being rude you absolute muppet.

Merry solstice!

I respectfully disagree - it's very easy to contract comment threads you have no interest in (at least on my client if I long-press a comment, it hides the comment and all responses), but I sometimes enjoy reading through an actual discussion two or three people have in a comment chain. They may be few and far between, but that's the nature of an open forum.

Don't forget the "fix" Sony offered after they were initally caught was an even MORE invasive rootkit. 🫠

Good bot!

In this context, I would love to see more audits!

I spent many hours while home sick just lying down and watching Johnny Castaway. I've had his hum stuck in my head... forever.

Look at me - I'm the farmer Maggot now!

I spent more time at Reddit then I did in the K-12 education system. I have a lot of good memories and learned a lot from both, but I have no desire to return to either.

I don't use JIRA myself, but after a brief search it looks like you can use the JIRA Python library or pull data directly from the REST api fairly easily. I know there are several libraries and methods to interact with Excel using Python, so I am fairly confident your user case would be doable with some scripting.

The userbase of Python is so large that for almost anything you might want to do it's likely someone else has already worked out a solution and created a library for it. For everything else, you can make your own solution and share it for the next person with that same problem.

I would argue it's worth having at least a passing knowledge of how Python works. It is a very simple yet powerful language that is used for a lot of applications.

Personally, I've utilized it at work to process data for reports that I would otherwise be doing in spreadsheets by hand. If you learn how to import .csv files, manipulate rows of data, and export back to a new .csv file then you will probably eventually find a use in any office you end up working at.

As a hobby, if you have any interest in AI art or AI large language model projects then knowing some basic python will be a huge help. Most of the open-source projects and their extensions use Python, and there are many times I've tried to use a GitHub tool but gotten an error. Knowing Python, I am able to track down and fix small issues about 80% of the time, which feels pretty cool.

Finally, even if you don't get much/any use out of Python, it's probably worth learning just so you understand how scripts, imported libraries, and basic programming logic works. Just having that baseline understanding will make you look like a rockstar when dealing with a companies proprietary software in many office settings.

It's no surprise that anyone with enough intelligence to have the clearance to handle this case wouldn't want anything to do with it. The writing is on the wall (and in the ballroom, in the bathroom, in the poolhouse...)

I imagine there will be many duplicates of several subs over the next while. That means Highlander rules will eventually go into effect!