pyvene.models.gpt_neox.modelings_intervenable_gpt_neox.create_gpt_neox

pyvene.models.gpt_neox.modelings_intervenable_gpt_neox.create_gpt_neox#

create_gpt_neox(name='EleutherAI/pythia-70m', cache_dir=None)[source]#

Creates a GPT2 model, config, and tokenizer from the given name and revision