TheBloke

/

Vicuna-33B-1-3-SuperHOT-8K-GPTQ

like 27

Hello, The Bloke

SuperHOT for 7B model has been released: https://huggingface.co/kaiokendev/superhot-7b-8k-no-rlhf-test

And I need orca_mini_7B-GPTQ with 8k context size, and I am not sure whether SuperHOT work on orca_mini_7B (As SuperHOT is a LORA for Llama while orca_mini_7B is Open-Llama)

If it work, could you help me create orca_mini_7B-SuperHOT-8K-GPTQ?

Thank you!