vllm.collect_env ¶
Functions:
-
get_cudnn_version–Return a list of libcudnn.so; it's hard to tell which one is being used.
-
get_intel_graphics_compiler_version–Return Intel Graphics Compiler (IGC) version.
-
get_level_zero_driver_version–Return Level Zero driver version.
-
get_level_zero_loader_version–Return Level Zero loader runtime version.
-
get_oneapi_ccl_version–Return oneAPI Collective Communications Library (oneCCL) version.
-
get_oneapi_compiler_version–Return Intel oneAPI DPC++/C++ Compiler version via icpx.
-
get_pip_packages–Return
pip listoutput. Note: will also find conda-installed pytorch and numpy packages. -
get_rocm_version–Returns the ROCm version if available, otherwise 'N/A'.
-
get_sycl_version–Return SYCL/DPC++ compiler build version.
-
run–Return (return-code, stdout, stderr).
-
run_and_parse_first_match–Run command using run_lambda, returns the first regex match if it exists.
-
run_and_read_all–Run command using run_lambda; reads and returns entire output if rc is 0.
get_cudnn_version(run_lambda) ¶
Return a list of libcudnn.so; it's hard to tell which one is being used.
Source code in vllm/collect_env.py
get_intel_graphics_compiler_version(run_lambda) ¶
get_level_zero_driver_version(run_lambda) ¶
get_level_zero_loader_version(run_lambda) ¶
get_oneapi_ccl_version(run_lambda) ¶
get_oneapi_compiler_version(run_lambda) ¶
Return Intel oneAPI DPC++/C++ Compiler version via icpx.
get_pip_packages(run_lambda, patterns=None) ¶
Return pip list output. Note: will also find conda-installed pytorch and numpy packages.
Source code in vllm/collect_env.py
get_rocm_version(run_lambda) ¶
get_sycl_version(run_lambda) ¶
run(command) ¶
Return (return-code, stdout, stderr).
Source code in vllm/collect_env.py
run_and_parse_first_match(run_lambda, command, regex) ¶
Run command using run_lambda, returns the first regex match if it exists.
Source code in vllm/collect_env.py
run_and_read_all(run_lambda, command) ¶
Run command using run_lambda; reads and returns entire output if rc is 0.