Replies: 1 comment
|
I had also thought about this. I've seen that different quants of the same model at the same quant level (variations of n bit OQ or OQE vs nvfp ) can have different behaviors depending on how the quantization occurred. I'm still learning about the ins and outs of the model files and how the quantizations work, but a ranking system like this would definitely be helpful. |
0 replies
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Uh oh!
There was an error while loading. Please reload this page.
Hello, my idea is just like in title, so users will be able to upload various intelligence benchmark results to omlx.ai.
It would help in deciding which model and quantisation is the best for given intelligence benchmark type.
All reactions