Tensor_Build | Devpost

Tensor_Build Logo
Block Code Area where you will train your model!
Landing page for Website where you will find the AI Training Model

Inspiration

All of us had started with block coding. As starters, we were faced with tough times, especially with the introduction of AI. That is why, to help beginner learners who are learning AI, we have built Tensor_Build.

What it does

Tensor_Build is a beginner-friendly interface for people starting their journey into AI development. It is something similar to scratch, where users drag blocks to code and train AI. This way, users can train their AI using simple inputs.

How we built it

We first created a Python library that condenses normally used code to train AI into functions. We then integrated it into the front-end by calling a function for every block in the workspace. This will create a series of AI training functions to run, which will result in a fully functional AI.

Challenges we ran into

At one point, the code repeatedly failed upon the attempt to compile it. We stayed until 4:00am in attempts to fix this issue. One issue had led to another, and it became a sleepless night for all of us. Eventually, we realized that it was a problem with one of the file names. We then refactored all of the references, and it worked.

Accomplishments that we're proud of

One of the accomplishments we are proud of is the launch page. The launch page took a lot of innovation from the team as well as a lot of hard work from some of our team members. It incorporated many design elements as well as animations, which made it a challenge for all the members that were on our team. We were eventually able to make a good launch page with the efforts of everyone. Another thing was the backend for the code blocks. The functions appended to the code blocks were incredibly difficult to accomplish, and even when we understood the logic, the implementation was still very difficult, so when we finished the code, it was one of the greatest feelings we had in this hackathon.

What we learned

All of us were able to gain valuable experience and more in-depth knowledge about AI training as well as the Flask framework. Through many workshops, we were also able to obtain valuable knowledge about design and many different languages. Some of us were less knowledgable about AI, and used this opportunity as a great learning experience.

What's next for Tensor_Build

For Tensor_Build, we wanted to give the user a plethora of selections in order to tailor the experience to the skills of each user. With these aspirations, we believe that the next step for Tensor_Build is to give the user a selection for an intermediate skill level interface as well as improve upon our beginner-friendly version.