llama.cpp
71564139 - Support multiple GPUs (split mode) on SYCL backend (#5806)

Commit

1 year ago

Support multiple GPUs (split mode) on SYCL backend (#5806) * suport multiple cards: split-mode - layer|row * rm warning * rebase with master, support tow new OPs, close feature for -sm=row, fix for unit test * update news * fix merge error * update according to review comments

References

#5806 - Support multiple GPUs (split mode) on SYCL backend

Author

NeoZhangJianyu

Parents

9bf297a0

Files8

README-sycl.md
common
- common.cpp
examples
- llama-bench
  - llama-bench.cpp
- sycl
  - ls-sycl-device.cpp
  - run-llama2.sh
ggml-sycl.cpp
ggml-sycl.h
llama.cpp

llama.cpp 71564139 - Support multiple GPUs (split mode) on SYCL backend (#5806)

llama.cpp
71564139 - Support multiple GPUs (split mode) on SYCL backend (#5806)