zac/llama-cpp-python-test2

1from llama_cpp import llama2from huggingface_hub import hf_hub_download #load from huggingfaces 3 4print("loading . . .")5 6 7llm = llama(model_path= hf_hub_download(repo_id="TheBloke/Vigogne-2-7B-Instruct-GGML", filename="vigogne-2-7b-instruct.ggmlv3.q4_1.bin"), n_ctx=2048) #download model from hf/ n_ctx=2048 for high ccontext length8 9print("model is loaded")