Generate answer from embedding vectors

[leewoosub]

Hi, I'm not familiar with llama-cpp-python (actually not familiar with cpp) but I have to use gguf model for my project.

I want to generate answer from pre-computed embedding vecotrs(torch.Tensor) with size (1, n_tokens, 4096), not from query text.
(Just like inputs_embeds argument of generate() function of transformers model)

Is this feature already implemented? If not, please anyone help me where should I begin.