bytedance / decoupleQ

A quantization algorithm for LLM
Apache License 2.0
94 stars 5 forks source link

add llama w2 infer demo #2

Closed MyPandaShaoxiang closed 3 months ago

ChuanhongLi commented 3 months ago

您好,我切换到了这个pr,执行 git submodule update --init 时,碰到如下问题: error: The following untracked working tree files would be overwritten by checkout:

error: The following untracked working tree files would be overwritten by checkout:
.github/ISSUE_TEMPLATE/bug_report.md
.github/ISSUE_TEMPLATE/config.yml
.github/ISSUE_TEMPLATE/documentation_request.md
.github/ISSUE_TEMPLATE/feature_request.md
.github/ISSUE_TEMPLATE/submit_question.md
.github/workflows/labeler.yml
.github/workflows/new-issues-to-triage-projects.yml
.github/workflows/stale.yml
.gitignore
.gitmodules
CHANGELOG.md
CITATION.cff
CMakeLists.txt
CONTRIBUTORS.md
CUDA.cmake
Doxyfile
LICENSE.txt
PUBLICATIONS.md
README.md
bin2hex.cmake
cmake/CTestTestfile.configure.cmake
cmake/CTestTestfile.test.configure.cmake
cmake/NvidiaCutlassConfig.cmake.in
cmake/NvidiaCutlassPackageConfig.cmake
cmake/googletest.cmake
cmake/nop.cu
cmake/version_extended.h.in
cuBLAS.cmake
cuDNN.cmake
docs/_config.yml
docs/aligned__buffer_8h.html
docs/aligned__buffer_8h__dep__incl.md5
docs/aligned__buffer_8h__incl.md5
docs/aligned__buffer_8h_source.html
docs/annotated.html
docs/arch_2mma_8h.html
docs/arch_2mma_8h__dep__incl.md5
docs/arch_2mma_8h__incl.md5
docs/arch_2mma_8h_source.html
docs/arch_2mma__sm50_8h.html
docs/arch_2mma__sm50_8h__dep__incl.md5
docs/arch_2mma__sm50_8h__incl.md5
docs/arch_2mma__sm50_8h_source.html
docs/arch_2mma__sm60_8h.html
docs/arch_2mma__sm60_8h__dep__incl.md5
docs/arch_2mma__sm60_8h__incl.md5
docs/arch_2mma__sm60_8h_source.html
docs/arch_2mma__sm61_8h.html
docs/arch_2mma__sm61_8h__dep__incl.md5
docs/arch_2mma__sm61_8h__incl.md5
docs/arch_2mma__sm61_8h_source.html
docs/arch_8h.html
docs/arch_8h__dep__incl.md5
docs/arch_8h_source.html
docs/array_8h.html
docs/array_8h__incl.md5
docs/array_8h_source.html
docs/array__subbyte_8h.html
docs/array__subbyte_8h__dep__incl.md5
docs/array__subbyte_8h__incl.md5
docs/array__subbyte_8h_source.html
docs/batched__reduction_8h.html
docs/batched__reduction_8h__dep__incl.md5
docs/batched__reduction_8h__incl.md5
docs/batched__reduction_8h_source.html
docs/batched__reduction__traits_8h.html
docs/batched__reduction__traits_8h__incl.md5
docs/batched__reduction__traits_8h_source.html
docs/bc_s.png
docs/bdwn.png
docs/classcutlass_1_1AlignedArray.html
docs/classcutlass_1_1AlignedArray__coll__graph.md5
docs/classcutlass_1_1AlignedArray__inherit__graph.md5
docs/classcutlass_1_1Array_3_01T_00_01N_00_01false_01_4-members.html
docs/classcutlass_1_1Array_3_01T_00_01N_00_01false_01_4.html
docs/classcutlass_1_1Array_3_01T_00_01N_00_01false_01_4_1_1const__iterator-members.html
docs/classcutlass_1_1Array_3_01T_00_01N_00_01false_01_4_1_1const__iterator.html
docs/classcutlass_1_1Array_3_01T_00_01N_00_01false_01_4_1_1const__reference-members.html
docs/classcutlass_1_1Array_3_01T_00_01N_00_01false_01_4_1_1const__reference.html
docs/classcutlass_1_1Array_3_01T_00_01N_00_01false_01_4_1_1const__reverse__iterator-members.html
docs/classcutlass_1_1Array_3_01T_00_01N_00_01false_01_4_1_1const__reverse__iterator.html
docs/classcutlass_1_1Array_3_01T_00_01N_00_01false_01_4_1_1iterator-members.html
docs/classcutlass_1_1Array_3_01T_00_01N_00_01false_01_4_1_1iterator.html
docs/classcutlass_1_1Array_3_01T_00_01N_00_01false_01_4_1_1reference-members.html
docs/classcutlass_1_1Array_3_01T_00_01N_00_01false_01_4_1_1reference.html
docs/classcutlass_1_1Array_3_01T_00_01N_00_01false_01_4_1_1reverse__iterator-members.html
docs/classcutlass_1_1Array_3_01T_00_01N_00_01false_01_4_1_1reverse__iterator.html
docs/classcutlass_1_1Array_3_01T_00_01N_00_01true_01_4-members.html
docs/classcutlass_1_1Array_3_01T_00_01N_00_01true_01_4.html
docs/classcutlass_1_1Array_3_01T_00_01N_00_01true_01_4_1_1const__iterator-members.html
docs/classcutlass_1_1Array_3_01T_00_01N_00_01true_01_4_1_1const__iterator.html
docs/classcutlass_1_1Array_3_01T_00_01N_00_01true_01_4_1_1const__reverse__iterator-members.html
docs/classcutlass_1_1Array_3_01T_00_01N_00_01true_01_4_1_1const__reverse__iterator.html
docs/classcutlass_1_1Array_3_01T_00_01N_00_01true_01_4_1_1iterator-members.html
docs/classcutlass_1_1Array_3_01T_00_01N_00_01true_01_4_1_1iterator.html
docs/classcut
Aborting
Unable to checkout '579af9606d998c329e38e98329342376932cb429' in submodule path 'dependencies/cutlass'

这个得怎么解决一下吗?谢谢!