tkonolige commented on a change in pull request #8975:
URL: https://github.com/apache/tvm/pull/8975#discussion_r706310416
##########
File path: docs/dev/debugger.rst
##########
@@ -140,7 +140,7 @@ How to use Debugger?
::
from tvm.contrib.debugger import debug_executor as graph_executor
- m = graph_executor.create(graph, lib, dev, dump_root="/tmp/tvmdbg")
+ m = graph_executor.create(graph, lib, dev, dump_root="/tmp/tvmdbg",
iters_num=10, repeat=1)
Review comment:
Could you add min_repeat_ms here.
##########
File path: docs/dev/debugger.rst
##########
@@ -166,6 +166,8 @@ How to use Debugger?
The outputs are dumped to a temporary folder in ``/tmp`` folder or the
folder specified while creating the runtime.
+iters_num is number of runs for measurement of average performance time for
each op.
+repeat is number of iterations group. As a result, the number of outputs (see
below) will be equal to repeat
Review comment:
Could you add min_repeat_ms here.
##########
File path: python/tvm/contrib/debugger/debug_executor.py
##########
@@ -50,6 +50,18 @@ def create(graph_json_str, libmod, device, dump_root=None):
dump_root : str
To select which folder the outputs should be kept.
None will make a temp folder in /tmp/tvmdbg<rand_string> and does the
dumping
+ number : int
+ Number of times to run the inner loop of the timing code. This inner
loop is run in
+ between the timer starting and stopping. In order to amortize any
timing overhead,
+ `number` should be increased when the runtime of the function is small
(less than a 1/10
+ of a millisecond).
+ repeat : int
+ Number of times to run the outer loop of the timing code (see above).
The output will
Review comment:
```suggestion
Number of times to run the outer loop of the timing code. The output
will
```
##########
File path: python/tvm/contrib/debugger/debug_executor.py
##########
@@ -93,9 +105,26 @@ class GraphModuleDebug(graph_executor.GraphModule):
dump_root : str
To select which folder the outputs should be kept.
None will make a temp folder in /tmp/tvmdbg<rand_string> and does the
dumping
+ number : int
+ Number of times to run the inner loop of the timing code. This inner
loop is run in
+ between the timer starting and stopping. In order to amortize any
timing overhead,
+ `number` should be increased when the runtime of the function is small
(less than a 1/10
+ of a millisecond).
+ repeat : int
+ Number of times to run the outer loop of the timing code (see above).
The output will
Review comment:
```suggestion
Number of times to run the outer loop of the timing code. The output
will
```
##########
File path: docs/dev/debugger.rst
##########
@@ -155,7 +155,7 @@ How to use Debugger?
::
lib = tvm.runtime.load_module("network.so")
- m = graph_executor.create(lib["get_graph_json"](), lib, dev,
dump_root="/tmp/tvmdbg")
+ m = graph_executor.create(lib["get_graph_json"](), lib, dev,
dump_root="/tmp/tvmdbg", iters_num=10, repeat=1)
Review comment:
Could you add min_repeat_ms here.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
