Hey everybody! I hope you find the setup video helpful. If you run into issues feel free to comment and hopefully myself and others can help you with your problem. I have a lot planned for new features (and bug fixes) so check back often for Chatbot UI updates! Also - expect an official hosted version in the coming days for simple use. This is my take on a great open-source AI chat app. I hope you like it, but if you don't the code is 100% open-source with an MIT license, so feel free to modify it and use it however you want for any purpose :)
Training models unfortunately is beyond my expertise. I've played around with fine-tuning a bit, but I'm probably not qualified enough to give you a great answer. Generally speaking, I think your use-case from what I can understand sounds like a great use-case for a RAG type application and you likely don't need to train a full model.@ciberpr0tecci0nparatodos182
Thank you for this amazing Chatbot-UI! I downloaded it and everythin worked out so far! I just have one question, I built an Assistant on the OpenAI API Site and I want to acces it through the chatbot-UI, but I can't realy figure out how. On the Assistant Page, it is not schown and when I click on create there is nowhere to paste the name of the assistant. Am I doing something wrong? Thanks in advance :)
Hello! I'm glad you find this work amazing! I'm from Taiwan, and I would like to ask for your permission to create a Chinese version tutorial. Since it involves your copyright, I believe I need your consent. I will credit you as the author and include your GitHub link.
Really wanted to try v2 of chatbot ui but docker isn't working, keeps stopping, and the scoop is not recognized as a command, I heard this was going to be insanely easy to install but too much headaches for someone like me
First of all congratulations for the project, its awesome. I'm looking forward to the video explaining how to deploy it in our own hosting in the web. Will you still do this tutorial here on the channel?
great chat ai , some wonderful features, I am running with LM Studio on windows it works great but if I use an Llm like Llava which has vision it would be great to allow the uploading of an image for discussion etc , keep up the good work Pho
Thanks for your video, I've successfully run this project in my local device. However, I found a problem that this project offers only the API keys for openAI and Google etc. What should I do if I want to use another online model that are not prepared in this project ? I'll appreciate it a lot if you could do me a favor. Looking forward to your reply!
Can you please explain the process for github codespaces too? the problem I am facing is that the scoop bucket cannot be used on the remote terminal. when I ran npx supabase start, it did install some tables into my supabase but some of them are missing. I wonder if it is due to the fact that i did not install supabase with scoop. Also, as the codespace itself is a container, I could not use docker in it. this can also be the reason for missing tables.
damn. this was sick. Just a question - i want to use my fine tuned llm on this chat-bot ui. so i am assuming i would not be able to use ollama ? what options do i have ? can i use the legacy version and instead of open_AI_API, can i add my deployed model API which i got from sagemaker?
Hi! When I get into the local host url This error appears "Error: Command failed: ollama list "ollama" no se reconoce como un comando interno o externo, programa o archivo por lotes ejecutable." I'm usign windows, someone can help me? :(
Nothing happens when i enter npm install, "npm : The term 'npm' is not recognized as the name of a cmdlet, function, script file, or operable program. Check the spelling of the name, or if a path was included, verify that the path is correct and try again."
@realmckaywrigley - works perfectly locally running on my mac and using azure OpenAI. Any tips how to easily deploy it on Azure? Also how to change "chatbot" name what is visible?
I am so clueless with all this, lol I am confused with step 1. I guess I will wait for the offical hosted version because I am a literal boomer when it comes to these things. But thanks for a get program.
Hey everybody! I hope you find the setup video helpful. If you run into issues feel free to comment and hopefully myself and others can help you with your problem. I have a lot planned for new features (and bug fixes) so check back often for Chatbot UI updates! Also - expect an official hosted version in the coming days for simple use. This is my take on a great open-source AI chat app. I hope you like it, but if you don't the code is 100% open-source with an MIT license, so feel free to modify it and use it however you want for any purpose :)
Training models unfortunately is beyond my expertise. I've played around with fine-tuning a bit, but I'm probably not qualified enough to give you a great answer. Generally speaking, I think your use-case from what I can understand sounds like a great use-case for a RAG type application and you likely don't need to train a full model.@ciberpr0tecci0nparatodos182
Think a true NAS scale setup would be helpful. Or have it put into truecharts so easily setup
Obrigado amigo.
Thank you for this amazing Chatbot-UI! I downloaded it and everythin worked out so far! I just have one question, I built an Assistant on the OpenAI API Site and I want to acces it through the chatbot-UI, but I can't realy figure out how. On the Assistant Page, it is not schown and when I click on create there is nowhere to paste the name of the assistant. Am I doing something wrong? Thanks in advance :)
Hello! I'm glad you find this work amazing!
I'm from Taiwan, and I would like to ask for your permission to create a Chinese version tutorial. Since it involves your copyright, I believe I need your consent. I will credit you as the author and include your GitHub link.
Thanks for all the work on this UI and the starter video! Dope AF. This interface is the gold standard *chefs kiss*
We need more people like you - build cool shit that is freely accessible and actually help most people, not just developers and hackers.
Really wanted to try v2 of chatbot ui but docker isn't working, keeps stopping, and the scoop is not recognized as a command, I heard this was going to be insanely easy to install but too much headaches for someone like me
Scoop needs to be installed, docker needs some configuration on *some* computers. Do some googling the guy did a great job in my opinion.
Can‘ wait for the second part. I still use a modified version of the former legacy Chatbot-ui on Vercel and Supabase. 🤗
First of all congratulations for the project, its awesome.
I'm looking forward to the video explaining how to deploy it in our own hosting in the web.
Will you still do this tutorial here on the channel?
great chat ai , some wonderful features, I am running with LM Studio on windows it works great but if I use an Llm like Llava which has vision it would be great to allow the uploading of an image for discussion etc , keep up the good work Pho
Dude. This is an excellent tool!
Thanks for the work
I can't express my happiness after successful run!! I really loved it ❤❤❤ , Thank you is very small word for your hard work , lots of love bro ❤❤❤
i followed it but i cannot chat because it says i dont have gpt4 and gpt 3.5.What are the other models that can be used for it.
Thanks for your video, I've successfully run this project in my local device. However, I found a problem that this project offers only the API keys for openAI and Google etc. What should I do if I want to use another online model that are not prepared in this project ? I'll appreciate it a lot if you could do me a favor. Looking forward to your reply!
Do you have any specific instructions if you want to host the front end in an Azure App Service?
It looks like the most recent version requires a paid-up OpenRouter API key for some reason.
What is embeddings provider in the advanced settings?
Awesome. What would you recommend running locally when it comes to code assistance?
Is there anyway to build a search function that can search through all of your chat history contents and not just the titles?
Thanks so much for sharing this! Works great! Next step is to deploy it to my Azure portal.
Can you please explain the process for github codespaces too? the problem I am facing is that the scoop bucket cannot be used on the remote terminal. when I ran npx supabase start, it did install some tables into my supabase but some of them are missing. I wonder if it is due to the fact that i did not install supabase with scoop.
Also, as the codespace itself is a container, I could not use docker in it. this can also be the reason for missing tables.
Amazing project! Works seamlessly for me locally! 👏
thanks, Mckay, good buddy, good chatbot UI
Hmm you changed an SQL file but doesn't run a migration.. does supabase start get the diff and apply or...?
damn. this was sick. Just a question - i want to use my fine tuned llm on this chat-bot ui.
so i am assuming i would not be able to use ollama ?
what options do i have ?
can i use the legacy version and instead of open_AI_API, can i add my deployed model API which i got from sagemaker?
Hello Thanks.. how to install that on openmediavault portainer "brew" command don't work, i don't find on hub docker supabase CLI te deploy it. Thanks
Thank you for the amazing work on this, managed to get it installed and works like a charm. Is there a reason for the 10mb limit on the files?
Hey Man, Step 2. "rpm install" returns "zsh: command not found: npm". What I'm missing?
part 2 please!! with audio dictation would be great
Hello! Thank you for this video. I was wondering how can I connect my own backend with this UI.
This is an awesome setup !
Is it possible to run ollama on an external cloud gpu server such as Lambdalabs?
Great Job Mckay!
I tried running on vercel and couldn't get past login screen. I will be waiting for the cloud video.
Thanks for sharing.
I can help. Can you explain what u see after the login screen
Works good so far, but how to add open source llms except with ollama?
Hi!
When I get into the local host url
This error appears
"Error: Command failed: ollama list
"ollama" no se reconoce como un comando interno o externo,
programa o archivo por lotes ejecutable."
I'm usign windows, someone can help me? :(
I cannot see the assistance and data in the sidebar. How to enable it.
Is there a way to get it to create images with dalle? or is this text only
If I want to use Railway, it's the same process? Or there's aeasy way?
Is there a way to run your own model (not ollama)? from huggingface?
If I want to download it using my Kali VM does it work?
error:
supabase_imgproxy_chatbot-ui-main container is not ready: unhealthy
tips: the remaining memory is 245MB.
what about google singup in signup screen
Great job mate.
Thank you for this!
Where can I find Linux instructions?
Nothing happens when i enter npm install, "npm : The term 'npm' is not recognized as the name of a cmdlet, function, script file, or operable program. Check the
spelling of the name, or if a path was included, verify that the path is correct and try again."
You have to install node js in your system
@realmckaywrigley - works perfectly locally running on my mac and using azure OpenAI. Any tips how to easily deploy it on Azure?
Also how to change "chatbot" name what is visible?
can it run chatgpt 4o model?
Where is 2nd video
How to install it on the server? What to pay attention to
I am so clueless with all this, lol I am confused with step 1. I guess I will wait for the offical hosted version because I am a literal boomer when it comes to these things. But thanks for a get program.
right there with ya.
this works for windows users what about brew?
There are Windows commands in the readme.
I just wish i had a more powerful GPU to run mistral locally
as far as i can tell, these instructions will only work on a mac.