]> git.djapps.eu Git - pkg/ggml/sources/llama.cpp/commit
overview / pkg / ggml / sources / llama.cpp / commit
summary | shortlog | log | commit | commitdiff | tree
(parent: 7f9753f)
llama : remove shards weight file support (#2000)
authorHoward Su <redacted>
Wed, 28 Jun 2023 17:13:02 +0000 (10:13 -0700)
committerGitHub <redacted>
Wed, 28 Jun 2023 17:13:02 +0000 (20:13 +0300)
commitb922bc351b69770cec2d35d2aa50fa052b95ca93
treeab14ba0af7274ddc5822fb5d110a7d0c953daa43tree
parent7f9753fa1263c4eded9a3de19778562f0e1093d7commit | diff
llama : remove shards weight file support (#2000)

* Remove multiple shards

* Remove multiple file loaders

* Remove llama_load_tensor_shard class

* Simplify load logic

* Remove dead code guess_n_parts function

* Remove vocab_only from constructor of llama_model_loader

* Remove alignment_prevents_mmap which is not more needed.

* Remove useless check
llama.cpp diff | blob | history
Packaging of ggml-org/llama.cpp