phishmaster

Significantly different results (and WRONG) inference when GPU is enabled.

32

I am running llama_cpp version 0.2.68 on Ubuntu 22.04LTS under conda environment. Attached are two Jupyter notebooks with ONLY one line changed (use CPU vs GPU). As you can see...

Crashing with "IndexError: index 200 is out of bounds for axis 0 with size 200"

9

Installed from conda environment with pip **Version: '0.2.69'** The code is as follow: ``` llm = Llama( model_path="/data/codelama-2024-02/CodeLlama-7b-Python/ggml-model-f16.gguf", seed=1023, # Uncomment to set a specific seed n_ctx=200, # Uncomment to...