bolt. new with olama which means you able to create a full-fledged application completely local on your computer for free if you don't know what is bolts. new this is a framework which we can use to create an end to-end application where you can use any programming language to create an application you just give a prompt then based on that it's able to create an application it is able to create files and you are able to edit those files directly in the browser but the issue here is that you need to use Cloe by default but if you can use olama you are able to run this whole application for free and we have a solution for that we have a folked version created by Co meden which has all these options we will also see if it's 7 billion parameter model is capable of creating a full-fledged application so in this video I'm going to take you through step by step how you can install this application locally on your computer 100% private and local then set up Ola and finally you are able to create an application with just using natural language but before that I regularly create videos in regards to Artificial Intelligence on my YouTube channel so do subscribe and click the Bell icon to stay tuned make sure you click the like button so this video can be helpful for many others like here so in your terminal G clone and the full path to that URL and then click enter I will provide all the code and the commands in the description below so you can just copy and paste it so after this click enter now navigate to that folder now step number two how to set up Ola Lama so go to ola.
com and download ol after that ol pull quen 2. 5 coder and then click enter now it's downloading the model next we need to create a model file so create model file now I'm going to open the file even in that folder if you see here is the model file I'm going to open it next inside the file I'm placing from quen 2. 5 coder 7B and the parameter that's the context length that is 32,000 because this is required to run bolt.
new locally on your computer because by default all these models come with less context length so we need to increase that so after this save it and now coming back to our terminal here ol create hyphen f model file then we are providing quen 2. 5 and I'm just naming it as large 7B and then click enter so this you can name whatever you like just for our reference and then click enter this will automatically create the model which you can see by by just typing ol list so here you can see quen 2. 5 L 7B created 8 seconds ago so we're going to use this model now final step we need to run the application for this purpose I'm going to use Docker so go to Docker and download Docker on your computer then in your terminal Docker compos profile development up and then click enter now you can see it automatically downloads the required packages and you can see the application is currently running in this URL now I'm going to click this URL and open it and here is our bolt.
new running locally on your computer you can see that here I'm just running in Local Host 5173 so after this you can choose whatever models you like but in our case I'm going to use oama click olama then you have multiple models to choose from so we changed the context length for quen 2. 5 large here so I'm going to choose that now I'm ready to prompt the application so I'm going to ask it to create a too app in react using Tailwind now it's going through the step by step open this preview window if you don't increase the context length you won't see this window so you can see clearly it's creating all the required packages installing all the packages then creating all the files then creating code inside the files which you can browse here that is super cool and finally it's running this command and here is the visual presentation I don't see anything here I'm going to prompt again saying I am seeing a empty screen so now based on this message it's giving me some suggestion on what to do but this is not ideal I want the model to go and edit the file for me so that is the key difference between a close Source model and a local model because local model are limited in regards to what they can do and also the context length limits its capability you have Cloe which contains 200,000 context length and chat GPT 128,000 context length this model only 32,000 context length so you can see the difference and that is the key difference why this model is not able to understand the context and able to edit the file when I prompt it I'm sure that if I ask you to create small small application it should be able to create but I'm sure it won't be able to create a large application in ol there are multiple models which you can try the one which I used is quen 2. 57 billion parameter model so this might work if you try 32 billion parameter 72 billion parameter model which you can completely download and run locally on your computer so I will leave that to you and you let me know in the comments below what do you think about this and did it work for you now we learned how we can integrate olama with bolt.
new and run locally on your computer considering you already like this bolt. new I've already created a step-by-step tutorial if you want to try that on a hosted platform there are some easy steps which I have mentioned that in this video which I'll put that here and I highly recommend for you to watch that to understand clearly the potential in regards to bolt.
