How to Set up and Use PrivateGPT and LocalGPT

[@konstantinlozev2272]

These RAG solutions are basically chunking the contents into 2-3 paragraph long chunks, doing a semantic search and then putting out the results as a context for the chat function. It's normal that they do not perform well on summarisation.

[@kshitijdesai2402]

I have tried it before, what you are using are raw models, however large you use, it needs to be finetuned for the task that you wish to perform which is in this case, the answer that you are expecting, otherwise they don't seem to be practical enough as of now.

[@coder-bm2jb]

for localGPT i am getting this error while installing requirements.txt: note: This error originates from a subprocess, and is likely not a problem with pip. ERROR: Failed building wheel for chroma-hnswlib Failed to build chroma-hnswlib ERROR: ERROR: Failed to build installable wheels for some pyproject.toml based projects (chroma-hnswlib) please help to solve this.

[@rrzz1513]

What are differences between PrivateGPT and Local GPT?

[@jaujauhestogsauuasgotsehua9784]

privateGPT only uses cpu, but localGPT can use gpu. I BELIEVE, unless its updated. Cus then idk

[@PetarBanicevic-j8t]

"just install poetry" - do I have to integrate them or its done by install script of privateGPT?

[@vetonrushiti19]

does localgpt work in a ubuntu machine without nvidia gpu?

[@narendrakd]

Hi MAM, I want to know can i ingest multiple files and i have 1000 docuemts pdf,txt and docs & i want to create graphical user interface based chat bot,can i ? can you please guide me mam.

[@sru_0119]

have you done it? if so can u please help me

[@sivasangaranv6787]

It's very much useful to me for my studies. Thank you ma'am

[@Pure_Science_and_Technology]

What embedding model is being used?

[@John_Krone]

Perfect explainer. I have a mac also so this was golden. Thanks

[@AssemblyAI]

Glad it helped!

[@MuhammadDanyalKhan]

Does private GPT or localGPT allow stacking of model? i.e. giving query to model that is best suited to answer?

[@therohitdas_com]

Why did you skip the part where we need to build the llama.cpp with METAL support? It should have been included in this tutorial.

[@browny334]

how to do this?

[@therohitdas_com]

​@@browny334 go to the Installation section of the privateGPT and then search(cmd+f) OSX GPU support

[@manu_ovg]

Which Mac you have ? M1 or M2 ? Because this looks fast

[@Andreas-z5o]

ERROR: Could not find a version that satisfies the requirement autoawq (from versions: none) ERROR: No matching distribution found for autoawq why do I get this error for localGPT

[@verslemonde6208]

you may not need it and can comment it out if you don't want to build your own quantised models.

[@mochammadrevaldi1790]

halo thankyou for the tutorial. can you explain how to use posgreSQL mode for vectorstore database privateGPT?

[@kajalsharma09]

Hii can you please help me in this

[@justthestuff3324]

GPT4All does this now, right?

[@user-wr4yl7tx3w]

like to see privategpt with Ollama?

[@VaibhavPatil-rx7pc]

Excellent !🤗

[@kajalsharma09]

Hello have you successfully run this

[@kajalsharma09]

Hello, anyone please help me i am getting error again and again

[@Derouz]

If AI is so advanced, couldn't they create installers similar to those in Windows to make it easier to install these programs without having to manipulate code?

[@browny334]

localGPT does not work on Mac unfortunately

[@kaistriban]

Great video. Thanks. On Mac giving the command git clone I got a xcrun error and apparently is not an uncommon error. I solved it by reinstalling xcode giving the command in the terminal: xcode-select --install

[@DanielArnolf]

We call yaml extension "Yamel"

[@golyandro]

Teşekkürler, tam aradığım şey.

[@ozanpress]

Bu ablamiz Turk sanirim :)

[@dascoder7473]

great joB, thank you

[@StripedLime]

*Unstructured [pdf]* gives an error on macbook

[@otomakannioc8213]

I like LLMs as long as they come along in such sympathic manner as presented in this video ;-)

[@PetarBanicevic-j8t]

Its bit superficial guide, flat emotions... I can read readme myself. Yaml is read YAM-uhl.

[@ldandco]

The face of AssemblyAI is just gorgeous

[@pabloronald7362]

Thanks for the video. Interesting! Although bionicgpt looks much more robust for running files locally.

[@user_375a82]

I'm not subscribing at $20 a month when $5 is the right price. All these Corps are price fixing. £20 is absurdly high. Even Amazon Prime is only $5 a month.

[@jaybestnz]

God it's incredible how ungrateful people are. At this time, businesses are throwing billions of dollars into these platforms in the hope of maybe winning and maybe making money in the future. This tech just 5 years ago was via Watson, and would be $150k to have access to the licences for a fraction of this tech and power. If you are doing these tasks in a business setting you should be easily able to save hundreds of dollars by using this (eg a lawyer able to review files faster and more accurately). I strongly suggest if you don't like the price, learn coding and form a company that can do this, and then charge less. It's a free market.

[@ddhruvarora]

if you have mac M1 chip, then GPU is not necessary, the power of Mac M1

[@TheGooch111]

You asked if there was a quicker way to say Y-A-M-L (.yaml), it's like Camel, but with a Y.

[@AssemblyAI]

I was going to say that but doubted myself for a second thinking it would sound funny. :D Thank you though! - Mısra