Please bump llama.cpp version to b3472 to have llama 3.1 support

[leszekhanusz]

llama.cpp release b3472 added support for llama 3.1.

Could you please bump the llama.cpp version?

[RandomInternetPreson]

The last piece of the puzzle. Textgeneration webui uses this library, and because of the rope factors are added as a tensor, an error is returned saying it received the incorrect number of tensors. When trying to load a model converted to gguf using the latest llama.cpp code.

[RandomInternetPreson]

It's happening ❤️

[leszekhanusz]

It has been done in release 0.2.84. Thanks!