]> git.djapps.eu Git - pkg/ggml/sources/llama.cpp/commit
overview / pkg / ggml / sources / llama.cpp / commit
summary | shortlog | log | commit | commitdiff | tree
(parent: 7bb36cc)
Fix more int overflow during quant (PPL/CUDA). (#6563)
authorDAN™ <redacted>
Sun, 28 Apr 2024 22:38:44 +0000 (18:38 -0400)
committerGitHub <redacted>
Sun, 28 Apr 2024 22:38:44 +0000 (00:38 +0200)
commite00b4a8f816ebc45b98a46e5f5231359b9a017e0
treef47ae2efca8cc0a5b82152b850ee8708fb500365tree
parent7bb36ccf91b8a2e92b182dd75624f1fd7cb205accommit | diff
Fix more int overflow during quant (PPL/CUDA). (#6563)

* Fix more int overflow during quant.

* Fix some more int overflow in softmax.

* Revert back to int64_t.
ggml-cuda/convert.cu diff | blob | history
ggml-cuda/softmax.cu diff | blob | history
Packaging of ggml-org/llama.cpp