Seeking advice on open-source llm selection

Hello everyone,

I’m currently embarking on a project to develop AI-driven courses tailored for high school students. A crucial decision I’m facing is choosing between an open-source model, like Llama2, and a proprietary model, such as GPT-4 through the OpenAI API. This decision is proving to be a major challenge, as each option offers distinct advantages and limitations.

I’m particularly interested in hearing from anyone who has experience using open-source models in their projects. Have you seen successful implementations that could serve as a benchmark? If so, what are these projects or businesses? Insights into successful cases would greatly reinforce my belief that leveraging open-source technology could be the most promising path forward for my project.

I appreciate any examples or advice you could share that would help illuminate the viability of open-source models in educational applications.

Thank you!

Hello,
Open-source models are great, the only downside is that you might need to invest in hardware as some are very computationally intensive, and for project scalability, you will need some technical skills.
You can check some projects here: Spaces - Hugging Face
All the best!

Regards
Isaak

1 Like