Merge pull request #689 from PINTO0309/fix_flatten_2dim

Addressed the issue of missing conversions when multi-dimensional flattening is performed and the batch size of the first dimension is an undefined dimension.

README.md

@@ -299,15 +299,15 @@ Video speed is adjusted approximately 50 times slower than actual speed.
   docker run --rm -it \
   -v `pwd`:/workdir \
   -w /workdir \
-  ghcr.io/pinto0309/onnx2tf:1.25.9
+  ghcr.io/pinto0309/onnx2tf:1.25.10
 
   or
 
   # Authentication is not required for pulls from Docker Hub.
   docker run --rm -it \
   -v `pwd`:/workdir \
   -w /workdir \
-  docker.io/pinto0309/onnx2tf:1.25.9
+  docker.io/pinto0309/onnx2tf:1.25.10
 
   or
 

onnx2tf/__init__.py

@@ -1,3 +1,3 @@
 from onnx2tf.onnx2tf import convert, main
 
-__version__ = '1.25.9'
+__version__ = '1.25.10'

onnx2tf/ops/Flatten.py

@@ -87,8 +87,20 @@ def make_node(
         cal_shape = (1, -1)
     elif axis >= input_tensor_rank:
         cal_shape = (-1, 1)
-    elif graph_node_output.shape is not None and len(graph_node_output.shape) == 2 and axis == input_tensor_rank - 1:
-        cal_shape = (1, -1)
+    elif graph_node_output.shape is not None \
+        and len(graph_node_output.shape) == 2 \
+        and axis == input_tensor_rank - 1 \
+        and not isinstance(graph_node_output.shape[0], str):
+        cal_shape = (graph_node_output.shape[0], -1)
+    elif graph_node_output.shape is not None \
+        and len(graph_node_output.shape) == 2 \
+        and axis == input_tensor_rank - 1 \
+        and isinstance(graph_node_output.shape[0], str):
+        try:
+            dim_prod = int(np.prod(graph_node_output.shape[1:]))
+            cal_shape = (-1, dim_prod)
+        except:
+            cal_shape = (1, -1)
     elif input_tensor_rank >= 2 \
         and input_tensor_shape[0] is None \
         and len([idx for idx in input_tensor_shape[1:] if idx is not None]) == input_tensor_rank - 1 \