Name: Tabby
Uploaded: 2024-12-07T17:00:52.388Z
Duration: 2 min 33 s
Description: An open-source, self-hosted AI coding assistant offering real-time code suggestions and autocompletion.

tabby

An open-source, self-hosted AI coding assistant offering real-time code suggestions and autocompletion.

Tabby

There is a self-hosted AI coding assistant that is completely for free. Think of it as a GitHub Copilot alternative. It's an open source on premise and it is an application with several key features that you would see with GitHub Copilot. Let me introduce Tabby. 

Tabby is a self-hosted AI coding assistant that you can set up with your own large language model. It offers many different features including the ability to create software code completion, debugging, chatting with your entire code base, as well as making edits to it. Now take a look at this example gif that I posted on my Twitter a long time ago. 

Now in this example, you'll see how Tabby finds the maximum element in an array. It defines the elements and then it autocompletes the section. It can also suggest code that you might want to input into a section and it will execute the commands for you. This is the pair programmer that I'm going to be showcasing throughout today's video. 

Now I'll be showcasing through today's video the different types of capabilities of Tabby, showcasing how you can set up your own large language model to the framework and so much more. So with that thought, guys, stay tuned and let's get straight into the video. 

Now before we get started, I'd like to introduce World of AI Solutions. This is a really big update that has been launched for my channel and this is where I have compiled a team of software engineers. We have machine learning experts, AI consultants, and this is basically a team where we're going to be providing AI solutions for businesses as well as personal use cases where AI solutions can be implemented to share AI Agent Store with friends.

Automate certain things or to help business operations. Now if you're interested, take a look at the tform link in the description below. I definitely recommend that you take a look at the patreon page so that you can access the new subscriptions that will be releasing this week. If you would like to book a consulting call with me, you can do so with the link in the description below as well. 

Now I've actually made another video on Tabby a long time ago, but there are a lot of new features that I really want to mention, which is why I wanted to take another look at Tabby. First of all, they announced coach stroll integration within Tabby. They have also introduced an answer engine. This is a central knowledge engine for internal engineer teams. It can simply integrate with Dev teams' internal data, delivering reliable and precise answers to empower different sorts of developers. 

You have a VS Code 1.7 marks, which is a significant milestone that is going to provide a versa chat experience throughout your coding experience. You can try the latest chat in the side panel within VS Code, which is something that has been just recently introduced. There is the ability to also focus on seamless integrations with GitHub. You have self-hosted GitHub. You can also utilize it with flexible configurations like HTTP API integrations. 

You have expanded capabilities, and there's so many other things that have been introduced over the past few months. So I definitely recommend that you try it out, and I will be showcasing how you can get started with it as we move forward into the video. So you may be wondering how can you.

Actually install this well if you are to go over to the docks, which is something that I'll leave a link to in the description below, as well as all the other links that I use in today's video. Once you go here, click on quick start and you'll see that there's multiple different ways for you to actually install Tabby. You can have it as something that could be integrated with your ID, which is something that I truly recommend. You can also install this with Docker, Docker compose, Homebrew. You have Linux, Windows, and all these different methods to install it locally.

Now what I recommend you do first is go over to the Tabby docs website and they actually have a way for you to try out this demo where you can ask it things. This is something that I will be showcasing further on into the video after the installation, but you can also play around with this to see how well this pair programmer is. So you can head over to the docs, click on the playground button, and it will take you to this page where you can test out and basically see how this code is working for you.

Now this is where I'm going to be chatting with this chat demo within Tabby. So this is the chat interface that you can basically interact with it to generate code, ask it questions, and this was something that was released quite recently. This is where I told it to help me code out a front end component. I want an animated button that says sign up now, and you can see it's going to start on compiling this. It's going to generate the CSS code for it using the animations from it, and they also even works on providing optional code from.

Different programming languages from JavaScript. So if you want to add some additional interactivity or functionalities, you can use JavaScript. 

Now putting it all together, you can see that it has now generated this. And now what I can do is I can showcase this component on an HTML viewer. So I went over to an HTML online viewer, which is going to showcase the code that was generated. And you can see that there are animations associated with the component that I generated with Tabby. 

So now that we have focused on Tabby's generation, I want to focus on how it's going to be able to deliberate an error. Meaning that how it's going to be focusing on debugging that error. So I'm going to provide a large code snippet, which I just had generated from GPT 4 Omni. And basically, this is a code snippet that has multiple different errors. 

And I'm going to see if Tabby is going to be able to provide solutions for it. So I'm going to, so what I'm going to do is I'm going to specify please find the errors and fix it. And then I'm going to send it over. So within a couple seconds, we're going to send in this Doc and we should be able to find the answers within a couple seconds. 

So there we go. In this code s that I provided, I had said that there is a class, the CSS property, and it's set to 125%, which is I believe at the top. And basically, what had happened is that if I keep it at this percentage, it will push the tool tip too far up. So potentially, it could be out of the visible zone. 

So in this like particular answer, you can see that it found that and it actually should be at 100. So you can see that it reduced it from.

20 125 to the correct value which is 100%. So you can see that it is an optimal tool that can help you uh I would say de bug certain things. In this case, if you're coding with Tabby in an ID, it's going to be able to provide feedback automatically. It's going to let you know and it's going to specify what sort of actions you should take to basically clean up the code. 

Now I'm going to Showcase this product demo video which is going to Showcase how you can locally get started on the Mac with Tabby. So take a look at this video. It has a lot of other examples and once you finished watching it, you should get a better idea as to what you can do with Tabby. 

First of all, you can see on our documentation page that there are different ways to install Tabby. For Macs with M1 or M2, the easiest way will be through uh home brew. Basically, you run bre install Tabby and for me, since I already have this install before, I can do Brew reinstall. It should run fa quickly and uh good. 

The next step is to bring up a Cy server. You uh copy this command here. Actually, for today, since I want to uh show you a little bit more about our latest uh chat interface and also the web server, I will also enable the flag uh chat model with a smaller model for example Wier code 3B and uh web server. Definitely, if you have a more powerful GPU locally, you can run larger models for better qualities. 

And yes, it's up. Let's go to Local Host. Here we go. So that's a cute dashboard that will bring to you as homepage. You can see your activities of using Tabby, for example, number of completions or number acceptances and you. 

Share the AI Agent Store with friends.

Can also see uh I've been playing a little bit around with uh different programming languages and since I actually had this run uh earlier this morning as well uh so you will see the homepage look like this. But if it's your first time launching this we will have a tutorial to guide you through registration and then go to the homepage. 

And good before we dive deeper into the web server we can first uh use Tabby to write some code. I will bring up my vs code and for first time user you need to go to extensions here and search Tabby and uh install our uh extensions here. Then on the bottom right if you see Tabby with a check mark here it shows that you're all good to go and Tabby's running. 

And great let's try probably writing some python code here for example uh here I can do uh get time out and yeah that's great suggestion and great follow redirect. Yeah that's great and it's quite helpful and also everything is running locally on your Mac. Great let's go back to the web server here and in settings here you can see more about what you can do uh to configure Tabby here. 

For example in the profile you can see your email and change your Avida and also change your password if you like. And in the information uh tab here you can see clusters there are different jobs that are running and the two large language models one for uh code completion and one for the chat and also a local co-search index that we're building to help enhance the uh the results suggested by Tabby. 

And for jobs that you can see the different Chown jobs that are running and the uh logs and arrows uh here and reports that's the. 

Please share the AI Agent Store with friends.

Same dashboard here but if you have uh a team running uh using Tabby together you can actually see team wise statistics. And uh here I have another uler here you can pick uh or choose. And you can also change the time range or different programming languages here and it provides you a complete view of your team usage of Tabby.

And great uh before we end this I also want to show you two experimental features that we are actually working on. One is chat playground and the other is uh code browser. And chat playground this probably looks familiar to you. It's kind of like a chat GPD experience but it's uh specifically for developer and coding uh experience. H and also everything is running locally on your machine. 

And for example I can ask how to parse a CSV file in Python and yeah it's generating the results uh and the codes and run pretty quickly. Uh yep and great and for the code browser uh you can actually browse code that you're indexing uh in Tabby. For example if I go to lent here uh let's for example let's go to some code here and here yeah great. 

And uh what else you can do here that's pretty cool is uh you can actually select some code snippets and you can ask Tabby to explain this code to you. Uh or uh for function you can uh generate unit test here. Yep pretty cool isn't it and everything again is running locally on your machine. 

Now wasn't that amazing this is the capability of Tabby and it's something that I truly recommend that you try out. Now just imagine it being coupled with another large language model like Sony 3.5 this would greatly enhance the way you utilize code and it would. 

Share the AI Agent Store with friends.

Basically enhance your code generation and your efficiency as well as your productivity. So I definitely recommend that you try this out. I'll leave a link to this in the description below. But that's basically it for today's video guys. 

Make sure you share AI Agent Store with friends. This is a great way for you to stay up to date with the latest AI news. Make sure you actually take a look at the Patreon page cuz this is a great way for you to access different types of AI subscriptions to different tools completely for free. 

And lastly make sure you guys share AI Agent Store with friends. Turn on notification Bell. Like this video and check out our previous videos so you can stay up to date with the latest AI news. But with that thought guys have an amazing day. Spread positivity and I'll see you guys fairly shortly. Peace out fellas.

Note, that there may be slight differences in text and video. Watch video for the original content.

Tabby Video