Failed to load model, Wrong MAGIC in header

[TeqGin]

I quantized the llama 7b-chat model by llama.cpp, and get model ggml-model-q4_0.gguf. But llama.go seems not support the gguf version,
it shows the error:
`
[ERROR] Invalid model file '../llama.cpp/models/7B/ggml-model-q4_0.gguf'! Wrong MAGIC in header

[ ERROR ] Failed to load model "../llama.cpp/models/7B/ggml-model-q4_0.gguf"
`

[macie]

Llama.cpp project is in an active development, and from time to time it introduces breaking changes (in gguf format too). This project stopped its development around April 2023, so It probably isn't useful with today's models.