Skip to content

Commit ed63b12

Browse files
svekarssvekars
authored
Remove duplicate tables
1 parent 517e66f commit ed63b12

File tree

1 file changed

+2
-80
lines changed

1 file changed

+2
-80
lines changed

‎recipes_source/xeon_run_cpu.rst

Lines changed: 2 additions & 80 deletions
Original file line numberDiff line numberDiff line change
@@ -190,14 +190,6 @@ The argument list and usage guidance can be shown with the following command:
190190
191191
The command above has the following positional arguments:
192192

193-
+------------------+---------------------------------------------------------+
194-
| knob | help |
195-
+==================+=========================================================+
196-
| ``program`` | The full path of the program/script to be launched. |
197-
+------------------+---------------------------------------------------------+
198-
| ``program_args`` | All the arguments for the program/script to be launched.|
199-
+------------------+---------------------------------------------------------+
200-
201193
.. list-table::
202194
:widths: 25 50
203195
:header-rows: 1
@@ -213,20 +205,6 @@ Explanation of the options
213205

214206
The generic option settings (knobs) include the following:
215207

216-
+----------------------+------+---------------+-------------------------------------------------------------------------------------------------------------------------+
217-
| knob | type | default value | help |
218-
+======================+======+===============+=========================================================================================================================+
219-
| ``-h``, ``--help`` | | | Show the help message and exit. |
220-
+----------------------+------+---------------+-------------------------------------------------------------------------------------------------------------------------+
221-
| ``-m``, ``--module`` | | | Changes each process to interpret the launch script as a python module, executing with the same behavior as "python -m".|
222-
+----------------------+------+---------------+-------------------------------------------------------------------------------------------------------------------------+
223-
| ``--no-python`` | bool | False | Do not prepend the program with "python" - just exec it directly. Useful when the script is not a Python script. |
224-
+----------------------+------+---------------+-------------------------------------------------------------------------------------------------------------------------+
225-
| ``--log-path`` | str | '' | The log file directory. Default path is ``''``, which means disable logging to files. |
226-
+----------------------+------+---------------+-------------------------------------------------------------------------------------------------------------------------+
227-
| ``--log-file-prefix``| str | 'run' | log file name prefix. |
228-
+----------------------+------+---------------+-------------------------------------------------------------------------------------------------------------------------+
229-
230208
.. list-table::
231209
:widths: 25 10 15 50
232210
:header-rows: 1
@@ -257,18 +235,6 @@ The generic option settings (knobs) include the following:
257235

258236
Knobs for applying or disabling optimizations are:
259237

260-
+-----------------------------+------+---------------+--------------------------------------------------------------------------------------------------------------------------+
261-
| knob | type | default value | help |
262-
+=============================+======+===============+==========================================================================================================================+
263-
| ``--enable-tcmalloc`` | bool | False | Enable ``TCMalloc`` memory allocator. |
264-
+-----------------------------+------+---------------+--------------------------------------------------------------------------------------------------------------------------+
265-
| ``--enable-jemalloc`` | bool | False | Enable ``JeMalloc`` memory allocator. |
266-
+-----------------------------+------+---------------+--------------------------------------------------------------------------------------------------------------------------+
267-
| ``--use-default-allocator`` | bool | False | Use default memory allocator. Neither ``TCMalloc`` nor ``JeMalloc`` would be used. |
268-
+-----------------------------+------+---------------+--------------------------------------------------------------------------------------------------------------------------+
269-
| ``--disable-iomp`` | bool | False | By default, Intel(R) OpenMP lib will be used if installed. Setting this flag would disable the usage of Intel(R) OpenMP. |
270-
+-----------------------------+------+---------------+--------------------------------------------------------------------------------------------------------------------------+
271-
272238
.. list-table::
273239
:widths: 25 10 15 50
274240
:header-rows: 1
@@ -299,34 +265,6 @@ Knobs for applying or disabling optimizations are:
299265

300266
Knobs for controlling instance number and compute resource allocation are:
301267

302-
+-----------------------------+------+---------------+----------------------------------------------------------------------------------------------------------------------------------------------+
303-
| knob | type | default value | help |
304-
+=============================+======+===============+==============================================================================================================================================+
305-
| ``--ninstances`` | int | 0 | Number of instances. |
306-
+-----------------------------+------+---------------+----------------------------------------------------------------------------------------------------------------------------------------------+
307-
| ``--ncores-per-instance`` | int | 0 | Number of cores used by every instance. |
308-
+-----------------------------+------+---------------+----------------------------------------------------------------------------------------------------------------------------------------------+
309-
| ``--node-id`` | int | -1 | Node id for multi-instance, by default all nodes will be used. |
310-
+-----------------------------+------+---------------+----------------------------------------------------------------------------------------------------------------------------------------------+
311-
| ``--core-list`` | str | '' | Specify the core list as "core_id, core_id, ...." or core range as "core_id-core_id". By dafault all the cores will be used. |
312-
+-----------------------------+------+---------------+----------------------------------------------------------------------------------------------------------------------------------------------+
313-
| ``--use-logical-core`` | bool | False | By default only physical cores are used. Specify this flag to use logical cores. |
314-
+-----------------------------+------+---------------+----------------------------------------------------------------------------------------------------------------------------------------------+
315-
| ``--skip-cross-node-cores`` | bool | False | Prevent the workload to be executed on cores across NUMA nodes. |
316-
+-----------------------------+------+---------------+----------------------------------------------------------------------------------------------------------------------------------------------+
317-
| ``--rank`` | int | -1 | Specify instance index to assign ncores_per_instance for rank; otherwise ncores_per_instance will be assigned sequentially to the instances. |
318-
+-----------------------------+------+---------------+----------------------------------------------------------------------------------------------------------------------------------------------+
319-
| ``--multi-instance`` | bool | False | A quick set to invoke multiple instances of the workload on multi-socket CPU servers. |
320-
+-----------------------------+------+---------------+----------------------------------------------------------------------------------------------------------------------------------------------+
321-
| ``--latency-mode`` | bool | False | A quick set to invoke benchmarking with latency mode, in which all physical cores are used and 4 cores per instance. |
322-
+-----------------------------+------+---------------+----------------------------------------------------------------------------------------------------------------------------------------------+
323-
| ``--throughput-mode`` | bool | False | A quick set to invoke benchmarking with throughput mode, in which all physical cores are used and 1 numa node per instance. |
324-
+-----------------------------+------+---------------+----------------------------------------------------------------------------------------------------------------------------------------------+
325-
| ``--disable-numactl`` | bool | False | By default ``numactl`` command is used to control NUMA access. Setting this flag will disable it. |
326-
+-----------------------------+------+---------------+----------------------------------------------------------------------------------------------------------------------------------------------+
327-
| ``--disable-taskset`` | bool | False | Disable the usage of ``taskset`` command. |
328-
+-----------------------------+------+---------------+----------------------------------------------------------------------------------------------------------------------------------------------+
329-
330268
.. list-table::
331269
:widths: 25 10 15 50
332270
:header-rows: 1
@@ -387,26 +325,10 @@ Knobs for controlling instance number and compute resource allocation are:
387325

388326
Environment variables that will be set by this script include the following:
389327

390-
+------------------+-------------------------------------------------------------------------------------------------+
391-
| Environ Variable | Value |
392-
+==================+=================================================================================================+
393-
| LD_PRELOAD | Depending on knobs you set, <lib>/libiomp5.so, <lib>/libjemalloc.so, <lib>/libtcmalloc.so might |
394-
| | be appended to LD_PRELOAD. |
395-
+------------------+-------------------------------------------------------------------------------------------------+
396-
| KMP_AFFINITY | If libiomp5.so is preloaded, KMP_AFFINITY could be set to "granularity=fine,compact,1,0". |
397-
+------------------+-------------------------------------------------------------------------------------------------+
398-
| KMP_BLOCKTIME | If libiomp5.so is preloaded, KMP_BLOCKTIME is set to "1". |
399-
+------------------+-------------------------------------------------------------------------------------------------+
400-
| OMP_NUM_THREADS | value of ncores_per_instance |
401-
+------------------+-------------------------------------------------------------------------------------------------+
402-
| MALLOC_CONF | If libjemalloc.so is preloaded, MALLOC_CONF will be set to |
403-
| | "oversize_threshold:1,background_thread:true,metadata_thp:auto". |
404-
+------------------+-------------------------------------------------------------------------------------------------+
405-
406328
.. list-table::
407329
:widths: 25 50
408330
:header-rows: 1
409-
* - Environ Variable
331+
* - Environment Variable
410332
- Value
411333
* - LD_PRELOAD
412334
- Depending on knobs you set, <lib>/libiomp5.so, <lib>/libjemalloc.so, <lib>/libtcmalloc.so might be appended to LD_PRELOAD.
@@ -435,4 +357,4 @@ See also:
435357

436358
* `PyTorch Performance Tuning Guide <https://pytorch.org/tutorials/recipes/recipes/tuning_guide.html#cpu-specific-optimizations>`__
437359
* `PyTorch Multiprocessing Best Practices <https://pytorch.org/docs/stable/notes/multiprocessing.html#cpu-in-multiprocessing>`__
438-
* Grokking PyTorch Intel CPU performance: `Part 1 <https://pytorch.org/tutorials/intermediate/torchserve_with_ipex>`__ `Part 2 <https://pytorch.org/tutorials/intermediate/torchserve_with_ipex_2>`__
360+
* Grokking PyTorch Intel CPU performance: `Part 1 <https://pytorch.org/tutorials/intermediate/torchserve_with_ipex>`__ `Part 2 <https://pytorch.org/tutorials/intermediate/torchserve_with_ipex_2>`__

0 commit comments

Comments
 (0)