Interview request: genAI evaluation & documentation
Hi! We are researchers from Carnegie Mellon University conducting a study on the evaluation and documentation practices of generative AI developers. Given the popularity and success of your model, we're particularly interested in learning from your team's experiences.
Our study aims to:
- Understand current practices in Gen AI model evaluation and reporting
- Identify challenges faced by developers in these areas
- Explore potential improvements in evaluation and documentation processes
We're seeking participants with hands-on experience in these aspects of Gen AI development. Would any members of your team be interested in participating in a (compensated) interview to share their insights?
For more details about the study and to express interest, please visit our recruitment page: https://forms.gle/fbn4734YxrRg6mkBA
I didn't actually make this model, this is Google's model that I quantized. You should likely reach out to them for any real information and experiences:
Hi! Thanks for your response! We noticed you have a fair few models on hugging face. If there are any generative models that you feel you did significant development/eval/documentation on, could you direct us to them?
Ah I didn't actually "make" any of these, the original authors upload the full model weights, then I download them, compress them, and re-upload for the general population :)