Prior art
Hey fellas
Did you even search for dolphin?
I would appreciate if you would change the name of your model and your paper. As my model existed and was quite popular, from a year and a half ago, and a simple Google / hugging face search would have discovered it.
Thanks for the information, we have internal discussions and will get back to you soon
Thank you!
Dolphin is a generic name, and seems that the improvement in arch of this model is big, and it worth the name of an intelligent animal, rather than some mediocre finetunes. If you want this name, you should consider registering a trademark.
This is not a legal conflict - Dolphin is well established with a large number of users both commercial and non-commercial applications. Another model named Dolphin will be confusing to them.
I would like to express my gratitude to NexaAIDev, Eric Hartford and Cognitive Computations for their innovative contributions to the AI community. To avoid confusion and honor the unique strengths of both projects, I suggest considering a slight modification to the new model's name to better reflect its unique dual-decoder architecture that could still use Dolphin. Some options that highlight this include:
Hybrid-Dolphin
Dolphin-Duo
DuoCore-Dolphin
This small change can help maintain clarity while celebrating the distinct advances each project brings to the field.
I follow NexaAIDev projects, Eric Hartford's work, as well as Cognitive Computations and the naming was admittedly a
bit confusing to me at first (I thought Cognitive Computations had published it).
I know it is not a trivial process to update models and papers but at least this is a fresh release.
As a big fan of all involved parties I would like to see this come to a constructive conclusion and
maybe we can help set some precedents for cooperation and model naming moving forward.
Joseph Flowers
Hi, thanks, and we are considering to update the name to avoid the confusion :)
OK, loving the name change. Also goes well with the Octopus. Thanks again for all your hard work! Seriously, I was trying to figure out how to do something similar and kept failing.
Thank you very much! Love you guys!