Can we have some more popular benchmarks
#8
by
rombodawg
- opened
Ive never even heard of the benchmarks you are using. Except for the first one (IFEval)
Here is a good list of benchmarks to use for your models.
GLM-4-32B has better instruction following than qwq 32b, it's the strongest open source model I've ever used for instruction following, and the ability to write code is okay, it's a pretty good model