Safetensors
llama

Config / model type could probably just be `llama` / `LlamaForCausalLM`

#2
by llllvvuu - opened

Since this is not using MoE, it does not need to use deepseek config or custom code. Could be simplified to llama for better/easier support.

Fixed, thanks!

llllvvuu changed discussion status to closed

Sign up or log in to comment