Benchmark code
#2
by amkkk - opened
Hi Jiunsong,
I hope you are doing well. Can you please share the benchmark code or script? I would like to test on my own end to see how well my own variants hold.
If it is not too much trouble, can you share the github repo for the benchmark if you have it or whatever else format you have it in, I will deeply appreciate it.
+1 , also the whole steps to get the main model abilitrated and unsencored without loosing much performance, thanks appreciate it in advance