]> git.djapps.eu Git - pkg/ggml/sources/llama.cpp/commit
overview / pkg / ggml / sources / llama.cpp / commit
summary | shortlog | log | commit | commitdiff | tree
(parent: 510b5c2)
opencl: refactor backend initilization (#23318)
authorlhez <redacted>
Wed, 20 May 2026 16:57:36 +0000 (09:57 -0700)
committerGitHub <redacted>
Wed, 20 May 2026 16:57:36 +0000 (09:57 -0700)
commit3a6db741a8189a45260536581f4ebb0a7f051f3c
tree75f8877d78cb25b1d7338318d37f2b3f1f682201tree
parent510b5c2a35652390c71327ecb29c2fb14bfe0e8ccommit | diff
opencl: refactor backend initilization (#23318)

* opencl: refactor initialization

* opencl: refactor GPU identification

* opencl: rename for consistency

* opencl: cache global mem size in dev_ctx

* opencl: adjust log level

* opencl: load argsort and flash_attn kernels in supports_op

* argsort kernel must be built for supports_op for querying the max
  workgroups
* flash_attn kernel has many variants, only load them when needed
ggml/src/ggml-opencl/ggml-opencl.cpp diff | blob | history
Packaging of ggml-org/llama.cpp