DataGen

• 

  DataGen - Home

Inspiration

Smaller Models are not really useful to do a lot of meaning tasks. How to fix it?

What it does

Framework to create auto generated synthetic datasets using Largers models

How we built it

Using Streamlit, powered by SambaNova APIs for bigger heavy lifting tasks, and AIML api for Vision tasks - verifying is the output is reasonable to train with!

Challenges we ran into

• Just putting it all together and time to train a reliable model ## Accomplishments that we're proud of This could be a very scalable framework reviving the actual use of 1B type models ## What we learned Hackathons make you think a lot and come up with ideas that you would do anything for! ## What's next for DataGen
  • Scale it and deploy it!

Built With

• aiml
• javascript
• python
• sambanova