I’m interested in hosting something like this, and I’d like to know experiences regarding this topic.
The main reason to host this for privacy reasons and also to integrate my own PKM data (markdown files, mainly).
Feel free to recommend me videos, articles, other Lemmy communities, etc.
Yes, mostly https://gpt4all.io/ only to find out that even the “uncensored” models are bullshit and won’t even provide you with a Windows XP Pro key. That’s kind of my benchmark for models nowadays. :P
Will it tell you how to make meth?
ollama + codellama works perfect, I use it from neovim with a plug-in called gen-nvim I think
deleted by creator
I’ve found it’s pretty good for translating between steps so to speak.
Converted some bash to python relatively quickly by giving it snippets and fixing errors as it made them.
I also had success generating an ansible playbook based on my own previously written install instructions for SillyTavern and llama.cpp.
I could do both of those tasks myself, but thar would be more difficult than having a mostly correct translation and fixing some errors.
It’s good for me because I’m piss poor at programming. In my defense, I’m not a programmer or even programmer adjacent. I do see how it wouldn’t be useful to a pro. It also has occasionally given me garbage advice that an expert would spot right away while I had to figure out in my own that it was ‘hallucinating’ again. There’s nothing better for learning than troubleshooting, though!
deleted by creator
I can absolutely see it getting useful for a pro. It’s already a better version of IDE templates. If you have to write boilerplate code this can already do that. It’s a huge time saver for the things you’d have to go look up to remember how to do and piece together yourself.
Example: today I wanted a quick way to serve my current working directory over HTTP so I could do some quick web work. I asked ChatGPT to write me a bash function I could stick in my profile to do this, and I told it to pick a random unused port. That would have taken me much longer had I went to lookup how to do that all. The only hint I gave it was to use the Python builtin module for serving http.
There’s a project called Tabby that your can host as a server on a machine that has a GPU, and has a VSCode extension that connects to the server.
The default model is called starcoder, and it’s the small version, 1B parameters. The downside is that it’s not super smart (but still an improvement over built in tools), but since it’s such a small model, I’m getting sub-second processing times.
deleted by creator
You should make sure you are running a model that fits in your vram, for me it runs faster than any online LLM I’ve tried.
I use koboldAI. It is local and open source
Checkout ollama.
There’s a lot of models you can pull from the official library.
Using ollama, you can also run external gguf models found on places like huggingface if you use a modelfile with something as simple as
echo "FROM ~/Documents/ollama/models/$model_filepath" >| ~/Documents/ollama/modelfiles/$model_name.modelfileI’m actively using ollama with docker to run llama2:13b model. It’s generally works fine but heavy on resources as expected.
deleted by creator
Recntly started using HuggingChat 🤗
Huggingchat for image generation is beautiful beautiful nightmare fuel.
I seriously love it.
I do image generation on AUTOMATIC1111
Really happy that i switched the text ai to something more opened that CloseAI
If you want to be able to use your models from everywhere sefurely, then koboldcpp on the ai horde is your best option. Super easy to set up
I haven’t tried any of them but I did just listen to a podcast the other week where they talk about LlamaGPT vs Ollama and other related tools. If you’re interested it’s episode 540: Uncensored AI on Linux by Linux Unplugged
“Uncensored” models are bullshit everything but uncensored. Just ask them for a Windows XP Pro key and you’ll see how uncensored they really are.
Last time I checked this, out of all the options available Serge was the simplest to host and use. Though you need a beefy computer to get fast and/or good responses.
Absolutely yes. You can try GPT4ALL which works on any decent CPU computer (the minimum I managed to run it with is a 2018 6 core 2.0ghz ARM64 processor) and has a lot of built in models. You can also import uncensored models (like the TheBloke ones on Huggingface ).
I also tried AutoGPT some times ago which is quite complex and cool.
I’ve played around with a few of them. I’ve found LM Studio the most robust and user friendly.
Mixtral is an amazing one that isn’t super slow or require incredible hardware foe a decent speed.
In general this guy has really good videos/tutorials for the latest tools.
Not with success but I’m using huggingface since a couple of days. You may want to have a look into it
