Doesn't work with new ggml versions

[tarunchand]

Failed to load new ggml versions - bad magic

[Yiyi-philosophy]

Yes, I meet a same question in my machine:

llama_model_load: loading model from '../ggml-model-q4_0.bin' - please wait ...
llama_model_load: invalid model file '../ggml-model-q4_0.bin' (bad magic)
main: failed to load model from '../ggml-model-q4_0.bin'

May be we need a script to transfer to the old ggml version.
By the way, llama.cpp is changing so fast ... Why so many developers get involve in it ?

[HanClinto]

The root version of llama.cpp now includes options for loading Alpaca models running in an interactive manner.
https://github.com/ggerganov/llama.cpp#instruction-mode-with-alpaca

I'm not sure that it's worth maintaining this as a separate fork anymore. Are there any improvements from alpaca.cpp that we should port back into the main trunk of llama.cpp?