Efficiency about v6 vision-rwkv #60
Description
I wonder if you adapt any experiment on v6 vision-rwkv efficiency.
In my down stream work, it seems that the v6 model perform worse than v4 model.
GPU: 4090
{{ message }}
I wonder if you adapt any experiment on v6 vision-rwkv efficiency.
In my down stream work, it seems that the v6 model perform worse than v4 model.
GPU: 4090