]> git.djapps.eu Git - pkg/ggml/sources/llama.cpp/commit
overview / pkg / ggml / sources / llama.cpp / commit
summary | shortlog | log | commit | commitdiff | tree
(parent: 1466621)
gguf-py : support lazy tensor splitting (#12809)
authorcompilade <redacted>
Tue, 8 Apr 2025 07:03:07 +0000 (03:03 -0400)
committerGitHub <redacted>
Tue, 8 Apr 2025 07:03:07 +0000 (09:03 +0200)
commita226bc7a9ac50551f9f113808de0f0046837f188
tree0b2b77b3f95deaa7df4542565a6ea8df6fbcf1a9tree
parent1466621e738779eefe1bb672e17dc55d63d166bbcommit | diff
gguf-py : support lazy tensor splitting (#12809)

* gguf-py : support lazy tensor splitting

Splitting usually involves returning tuples of tensors,
which need to be handled properly to avoid early eager evaluation.

* gguf-py : fix flake8 lint
gguf-py/gguf/lazy.py diff | blob | history
Packaging of ggml-org/llama.cpp