Hello,
First, thank you for sharing this excellent work. After reading your paper, I noticed that both AIM and DFN-5B CLIP were trained on similar training data. I'm interested in discussing and comparing the performance of these two models. This would be very helpful for understanding and selecting the appropriate model for various applications.
Thank you again for your contribution to the field!
Hello, First, thank you for sharing this excellent work. After reading your paper, I noticed that both AIM and DFN-5B CLIP were trained on similar training data. I'm interested in discussing and comparing the performance of these two models. This would be very helpful for understanding and selecting the appropriate model for various applications. Thank you again for your contribution to the field!