Environment Variables

Specifies the number of parallel operator build processes during Ascend backend compilation.

Note: Only Ascend backend.

Specifies whether to save or load the cache of the graph compiled by front-end. The function is the same as the enable_compile_cache in MindSpore context.

Note: This environment variable has lower precedence than the context enable_compile_cache.

0: Disable the compile cache

1: Enable the compile cache

Enable debug function and generate the TBE instruction mapping file during Ascend backend compilation.

Note: Only Ascend backend.

The value of compiler op level should be one of [0, 1, 2, 3, 4].

0: Turn off op debug and delete op compile cache files

1: Turn on debug, generate the *.cce and *_loc.json

2: Turn on debug, generate the *.cce and *_loc.json files and turn off the compile optimization switch (The CCEC compiler option is set to -O0-g) at the same time

3: Turn off op debug

4: Turn off op debug, generate the *.cce and *_loc.json files, generate UB fusion calculation description files ({$kernel_name}_compute.json) for fusion ops

Turn off operator prebuild processes during Ascend backend compilation. The prebuild processing may fix the attr fusion_type of the operate, and then affect the operator fusion. If the performance of fusion operator can not meet the expectations, try to turn on this environment variable to verify if there is the performance problem of fusion operator.

Note: Only Ascend backend.

true: turn off prebuild

false: enable prebuild

Control options of graph kernel fusion, it can be used to open or close the graph kernel fusion, supports fine-tune of several optimizations in graph kernel fusion and supports dumping the fusion process, which is helpful in problems location and performance tuning.

Note: This environment variable will be deprecated and removed in a future version, use graph_kernel_flags in the context instead.

Refer to the value setting of graph_kernel_flags in mindspore/context.py

Note: The priority of environment variables is higher than context, that is, if both environment variables and context are set at the same time, only the settings in the environment variables will take effect.

Specifies the number of Ascend AI Processors to be called during deep learning.

Note: When the Ascend AI Processor is used, specified by user when a distributed case is executed.

Specifies the file to which a path points, including device_ip corresponding to multiple Ascend AI Processor device_id.

Note: When the Ascend AI Processor is used, specified by user when a distributed case is executed.

1：enables RDR

0：disables RDR

1：export data when training process terminates in exceptional scenario

2：export data when training process terminates in both exceptional scenario and normal scenario.

Default: 1.

1: logs are output to the screen

0: logs are output to a file

Default: 1

0-DEBUG

1-INFO

2-WARNING

3-ERROR

Default: 2

0-DEBUG

1-INFO

2-WARNING

3-ERROR

Default: 2.

0-DEBUG

1-INFO

2-WARNING

3-ERROR

SubModule: COMMON, MD, DEBUG, DEVICE, COMMON, IR...

true: enables pipeline tree optimization.

false: disables pipeline tree optimization.

1: enables Debugger.

0: disables Debugger.

1: enables memory overcommitment for nodes selected on Debugger.

0: disables memory overcommitment for nodes selected on Debugger.

0: task sinking mode.

1: non-task sinking mode.

1: enables fallback function

0: disables fallback function

Default: 1

0: display exception information related to model developers and framework developers

1: display exception information related to model developers

Default: 0