Ollama: How To Create Custom Models From HuggingFace ( GGUF )

[@SIR_Studios786]

gguf can be run directly with ollama, how to convert other formats to be converted to use with ollama
?

[@SantK1208]

Thanks for sharing it, could you please create a video where we can do Q&A with local documents using Ollma models

[@SantK1208]

Kindly do it , it’s really needed

[@datasciencebasics]

Please check other videos in this channel, here u go ->
Chat With Documents Using ChainLit, LangChain, Ollama & Mistral 🧠
ua-cam.com/video/2IL0Sd3neWc/v-deo.html

[@albuquerqueroger]

Parabéns, conteúdo excelente! Obrigado por compartilhar.

[@ph3ll3r]

It would be interesting to see if you could Open-webUI instead of "Chat UI using ChainLit, LangChain, Ollama and Gemma"

[@BiranchiNarayanNayak]

Excellent tutorial !!

[@datasciencebasics]

Thanks !!

[@amiranvarov]

Hey mate, very good video with clear explanation. Do u mind to share what terminal app you are using? seems like very convenient with all that autocompletion and some hints

[@datasciencebasics]

I am using Warp but you can watch this video for making terminal better
HOW To Make Your Mac Terminal Amazing
ua-cam.com/video/ycapVWVl98M/v-deo.html

[@ScrantonStrangler19]

Good explanation! Is there a list of model architectures that are supported by Ollama?

[@datasciencebasics]

I don’t think there is any architecture for Ollama, you can download any models and use with Ollama. Give a try !!

[@build.aiagents]

Phenomenal

[@nat.amato-10]

How can I edit the Modelfile, so that it includes within it a context, a personality, or a precise way of answering questions?

[@mixp1x]

Is there any way we can use Exl2 on Ollama?

[@hackedbyBLAGH]

Thank you

[@datasciencebasics]

You are welcome !!

[@Pedro-mq7kt]

Nice video,
Can you make a video changing the system-prompt, temperature
If its possible.
Thankyou

[@datasciencebasics]

Its possible, just provide those parameters in Modelfile.

[@SantK1208]

Could you please use llama index with hugging face

[@121Gamerscom]

how do i pimp my chainlit logo to mine to change it please

[@zamanganji1262]

I want to finetune the llama 3 but I need to crate the special_tokens_map.json as follows:
{
"bos_token": {
"content": "",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false
},
"eos_token": {
"content": "",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false
},
"pad_token": {
"content": "",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false
}
}
How can I do?
Moreover I want to have a ollama run the model to have a chat with the model.

[@terry-bn7bh]

good, how to create the Modefile on windows?😁

[@datasciencebasics]

it should work in the same way, hive a try !!

[@terry-bn7bh]

how to create the Modefile on windows?

[@AnudeepKolluri]

Just open vscode and create a file called Modelfile (not sure about capitalization) and insert the content into it. It doesnt need any extension

[@adhyanmishra0197]

hello sir can we deploy this model

[@datasciencebasics]

Yes you can !!

[@adhyanmishra0197]

@@datasciencebasics Sir if possible please can u create one tutorial for deploying this model

[@datasciencebasics]

hello, deploying a model is use case specific. It can be deployed locally, in different cloud services , etc. Please refer to other videos in my channel for help.

[@AlanCarrOnlineHypnosis]

Far more hassle than it's worth. Just use any other app that uses normal GGUF files, like normal people.

[@datasciencebasics]

Thanks for the comment. There are many ways to do and its just a preference which one to use🙂

[@IamalwaysOK]

Excellent tutorial !!

[@datasciencebasics]

Glad you liked it!