Molly Sophia
Molly Sophia
> Hey, I have already tested this method under Arrow OS 13, it works fine. To test the stability of this new method, I restarted my device for about 7...
This sounds interesting
Wow! Nice work What a pity it is that I don’t have mh2lm in hand to test
Update: added support for fla-hub's rwkv7 hf model format. (https://huggingface.co/fla-hub/rwkv7-1.5B-world)
> Just a heads up, this will likely take some time to merge - I want to finish #11213 first and then figure out how to fit RWKV in the...
> Great, keep a look at the #11213 PR. It's still very messy, but I hope it will soon start to make sense. I think maybe we can have this...
> I hope that we can merge this one and test new RWKV v7 models. 🤗 Sure! I'm rebasing the branch of this PR today.
Superseded by #12412 I think
Hi! I've just done the changes in rwkv part according to what's already in this PR. The code is here: https://github.com/MollySophia/llama.cpp/tree/molly/llama-kv-cache also made the graph building parts of existing rwkv...
> Pinging @MollySophia and @compilade if you could run some tests with this branch to check if the RWKV and Mamba models work correctly. > > Any suggestions for improving...