h2ogpt
h2ogpt copied to clipboard
error while running generate.py
I got this error while running python generate.py --base_model='llama' --prompt_type=wizard2 --score_model=None --langchain_mode='UserData' --user_path=user_path
ValueError: Model path does not exist: WizardLM-7B-uncensored.ggmlv3.q8_0.bin Exception ignored in: <function Llama.del at 0x00000214CF6BFD00> Traceback (most recent call last): File "D:\h2ogpt\venv\lib\site-packages\llama_cpp\llama.py", line 1431, in del if self.ctx is not None: AttributeError: 'Llama' object has no attribute 'ctx'
How to resolve this ? And also I want to use gpt4all using cpu so how can I do that?