https://huggingface.co/Akicou/Hy3-REAM-100B
https://huggingface.co/Akicou/Hy3-REAM-100B
Hello I'd be grateful if you could quantize this! Appreciate your contributions to HF!
It's queued!
You can check for progress at http://hf.tst.eu/status.html or regularly check the model
summary page at https://hf.tst.eu/model#Hy3-REAM-100B-GGUF for quants to appear.
It's queued!
You can check for progress at http://hf.tst.eu/status.html or regularly check the model
summary page at https://hf.tst.eu/model#Hy3-REAM-100B-GGUF for quants to appear.
Hello Richard. Its my fault i didn't dig deeper into any other GGUF model repos to see if it used a custom llama cpp build.
Would it be possible to use the same build and patches mentioned in this shell script? https://huggingface.co/AngelSlim/Hy3-GGUF/blob/main/setup_hyv3_llama.sh
as i understand it:
Clones llama.cpp
Checks out a specific verified commit
Applies two patch sets:
01: adds the hy_v3 architecture
02: adds MTP (multi-token prediction) speculative decoding and the Hy3 parser
Builds that patched version.
no. we dont do custom forks/modifications, sorry. Only main llamacpp, otherwise too much user confusion. To be honest with you, some users have confusion with main llamacpp somehow, so imagine the chaos with different modifications that may or may not work with main llamacpp...
Oh alright. Was worth asking. Imma close this.