]> git.djapps.eu Git - pkg/ggml/sources/llama.cpp/commit
overview / pkg / ggml / sources / llama.cpp / commit
summary | shortlog | log | commit | commitdiff | tree
(parent: 88c46cb)
llama : add `gemma` model (#5631)
authorpostmasters <redacted>
Wed, 21 Feb 2024 13:08:22 +0000 (05:08 -0800)
committerGitHub <redacted>
Wed, 21 Feb 2024 13:08:22 +0000 (15:08 +0200)
commit580111d42b3b6ad0a390bfb267d6e3077506eb31
tree9eed0a46aacfa10e586ea478c191106f11e59febtree
parent88c46cbdac05cebd936511b1d3c74112e721615fcommit | diff
llama : add `gemma` model (#5631)

There are couple things in this architecture:

1. Shared input and output embedding parameters.
2. Key length and value length are not derived from `n_embd`.

More information about the models can be found at
https://ai.google.dev/gemma. GGUFs can be downloaded from
https://huggingface.co/google.
README.md diff | blob | history
gguf-py/gguf/constants.py diff | blob | history
llama.cpp diff | blob | history
Packaging of ggml-org/llama.cpp