nolanliou commented on a change in pull request #6546:
URL: https://github.com/apache/incubator-tvm/pull/6546#discussion_r494123685



##########
File path: python/tvm/relay/frontend/pytorch.py
##########
@@ -3244,7 +3267,8 @@ def get_all_op_names(graph):
     return set(node.kind() for node in nodes)
 
 
-def from_pytorch(script_module, input_shapes, custom_convert_map=None, 
default_dtype="float32"):
+def from_pytorch(script_module, input_shapes, custom_convert_map=None, 
default_dtype="float32",
+                 input_types=None):

Review comment:
       Just for forward compatibility. If not, one argument `input_info` maybe 
the best choice.

##########
File path: python/tvm/relay/frontend/pytorch.py
##########
@@ -3254,10 +3259,14 @@ def from_pytorch(script_module, input_shapes, 
custom_convert_map=None, default_d
         TorchScripted PyTorch graph
         Note: We currently only support traces (ie: torch.jit.trace(model, 
input))
 
-    input_shapes : List of tuples of input name and input dimensions
-        Graph level input shape list
+    input_infos: List of tuples of input name and (input dimensions, input 
types)
+        Graph level input shape and type list
         The same input names need to be used for deployment, so choose easy to
         remember names (such as: input0, input1)
+        e.g. [('input0', ((1, 2), 'int')), ('input1', ((3, 4), 'float'))]
+        supported data types: ['half', 'float', 'double', 'bool', 'char', 
+                               'byte', 'short', 'int', 'long',
+                               'quint8', 'qint8', 'qint32', 'str']

Review comment:
       
https://github.com/apache/incubator-tvm/blob/master/python/tvm/relay/frontend/pytorch.py#L2416
 from here

##########
File path: python/tvm/relay/frontend/pytorch.py
##########
@@ -3254,10 +3259,14 @@ def from_pytorch(script_module, input_shapes, 
custom_convert_map=None, default_d
         TorchScripted PyTorch graph
         Note: We currently only support traces (ie: torch.jit.trace(model, 
input))
 
-    input_shapes : List of tuples of input name and input dimensions
-        Graph level input shape list
+    input_infos: List of tuples of input name and (input dimensions, input 
types)
+        Graph level input shape and type list
         The same input names need to be used for deployment, so choose easy to
         remember names (such as: input0, input1)
+        e.g. [('input0', ((1, 2), 'int')), ('input1', ((3, 4), 'float'))]
+        supported data types: ['half', 'float', 'double', 'bool', 'char', 
+                               'byte', 'short', 'int', 'long',
+                               'quint8', 'qint8', 'qint32', 'str']

Review comment:
       remove the comment




----------------------------------------------------------------
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.

For queries about this service, please contact Infrastructure at:
[email protected]


Reply via email to