dl-backtrace 0.0.24__tar.gz → 0.0.25__tar.gz

{dl_backtrace-0.0.24 → dl_backtrace-0.0.25}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.2
 Name: dl_backtrace
-Version: 0.0.24
+Version: 0.0.25
 Summary: A python SDK for Deep Learning Backtrace
 Home-page: https://xai.arya.ai/docs/introduction
 License: MIT

{dl_backtrace-0.0.24 → dl_backtrace-0.0.25}/dl_backtrace/pytorch_backtrace/backtrace/backtrace.py

@@ -475,23 +475,26 @@ class Backtrace(object):
                         all_wt[start_layer], all_out[child_nodes[0]][0]
                     )
                     all_wt[child_nodes[0]] += temp_wt
-                elif (
-                        model_resource[1][start_layer]["class"] == "AdaptiveAvgPool2d"
-                ):
+                elif model_resource[1][start_layer]["class"] == "AdaptiveAvgPool2d":
                     temp_wt = UP.calculate_wt_gavgpool(
                         all_wt[start_layer], all_out[child_nodes[0]][0]
                     )
                     all_wt[child_nodes[0]] += temp_wt.T
                 elif model_resource[1][start_layer]["class"] == "MaxPool2d":
                     l1 = model_resource[0][start_layer]
+                    pad1 = l1.padding
+                    strides1 = l1.stride
+
                     temp_wt = UP.calculate_wt_maxpool(
-                        all_wt[start_layer], all_out[child_nodes[0]][0], (l1.kernel_size, l1.kernel_size)
+                        all_wt[start_layer], all_out[child_nodes[0]][0], (l1.kernel_size, l1.kernel_size),pad1,strides1
                     )
                     all_wt[child_nodes[0]] += temp_wt.T
                 elif model_resource[1][start_layer]["class"] == "AvgPool2d":
                     l1 = model_resource[0][start_layer]
+                    pad1 = l1.padding
+                    strides1 = l1.stride
                     temp_wt = UP.calculate_wt_avgpool(
-                        all_wt[start_layer], all_out[child_nodes[0]][0], (l1.kernel_size, l1.kernel_size)
+                        all_wt[start_layer], all_out[child_nodes[0]][0], (l1.kernel_size, l1.kernel_size),pad1,strides1
                     )
                     all_wt[child_nodes[0]] += temp_wt.T
                 elif model_resource[1][start_layer]["class"] == "MaxPool1d":

{dl_backtrace-0.0.24 → dl_backtrace-0.0.25}/dl_backtrace/version.py

@@ -1,8 +1,13 @@
-# file generated by setuptools_scm
+# file generated by setuptools-scm
 # don't change, don't track in version control
+
+__all__ = ["__version__", "__version_tuple__", "version", "version_tuple"]
+
 TYPE_CHECKING = False
 if TYPE_CHECKING:
-    from typing import Tuple, Union
+    from typing import Tuple
+    from typing import Union
+
     VERSION_TUPLE = Tuple[Union[int, str], ...]
 else:
     VERSION_TUPLE = object
@@ -12,5 +17,5 @@ __version__: str
 __version_tuple__: VERSION_TUPLE
 version_tuple: VERSION_TUPLE
 
-__version__ = version = '0.0.24'
-__version_tuple__ = version_tuple = (0, 0, 24)
+__version__ = version = '0.0.25'
+__version_tuple__ = version_tuple = (0, 0, 25)

{dl_backtrace-0.0.24 → dl_backtrace-0.0.25}/dl_backtrace.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.2
 Name: dl_backtrace
-Version: 0.0.24
+Version: 0.0.25
 Summary: A python SDK for Deep Learning Backtrace
 Home-page: https://xai.arya.ai/docs/introduction
 License: MIT

{dl_backtrace-0.0.24 → dl_backtrace-0.0.25}/dl_backtrace/pytorch_backtrace/backtrace/utils/prop.py

@@ -715,9 +715,9 @@ def calculate_wt_avgpool(wts, inp, pool_size, padding, strides):
                     paddings[1][0]:(paddings[1][0]+inp.shape[1]),:]
     return out_ds
 def calculate_wt_gavgpool(wts, inp):
+    channels = wts.shape[0]
     wts=wts.T
     inp=inp.T
-    channels = wts.shape[0]
     wt_mat = np.zeros_like(inp)
     for c in range(channels):
         wt = wts[c]