Inspiration I was inspired by one LLM model trained to help within one textbook. And I wanted to make an analogue of such a bot, but for any topics and with the ability to load additional data.
What it does "Deep Seek education bot" was designed for use in adaptive learning systems. For more convenient and faster use, it was placed on the Huggingface platform. At the moment, only two training options are offered: 1. "Regular reasoning with a teacher"
- "Curriculum" when selected, a window appears for filling in data, on the basis of which the model will build an individual plan (in the future, the project will be improved and offer more extensive functionality. For now, we can say that this is a Demo version)
How we created it I found a suitable and relevant model on the platform. I created a Hugging Face space, cloned the repository and then worked in vs code. ## Problems we encountered There were huge problems with working with vs code files and huggingFace API. Spaces couldn’t find the API for the model and this slowed me down a lot
Achievements we are proud of I was able to participate in the Hackathon :)
What we learned I learned patience, situation analysis and project planning.
What awaits the educational bot Deep Seek It will undergo additional training on an additional database so that it can skillfully train users, and an update in training options awaits.
Built With
- pdfplumber
- python
- python-dotenv
Log in or sign up for Devpost to join the conversation.