Base model
#1
by
Ayush342
- opened
What is the base model for this?
This are GGUF quants of https://huggingface.co/abideen/Bitnet-Llama-70M as mentioned in the model card. To me it looks like Bitnet-Llama-70M was trained from scratch and is not based of any model as it is of architecture LlamaForCausalLM. More accurately one could say that while the model was likely trained from scratch with some files likely originating form NousResearch/Llama-2-7b-hf. If you want to be sure please just ask the original author.
maybe it should have used TC quants. let's make it a testcase for quant overrides
sigh. doesn't convert anymore. or maybe never converted and nobody has found out yet.