[MLBEDSW-297] Setup and run on Microsoft Windows

Various updates to make vela run and produce identical output on
Microsoft Windows.
* Fixed overflow errors
* Fixed compile warnings
* Avoid problematic numpy version
* Updated README.md

Signed-off-by: Fredrik Svedberg <Fredrik.Svedberg@arm.com>
Change-Id: Ie48c63a92a00c81b3247d07f05b75d881319ddbb
diff --git a/ethosu/tensor_allocator/tensor_allocatormodule.cpp b/ethosu/tensor_allocator/tensor_allocatormodule.cpp
index 79ee95a..02488ad 100644
--- a/ethosu/tensor_allocator/tensor_allocatormodule.cpp
+++ b/ethosu/tensor_allocator/tensor_allocatormodule.cpp
@@ -53,7 +53,7 @@
     }
 
     /* Unpack the length of the input integer list. */
-    int input_length = PyObject_Length (input_list_object);
+    int input_length = static_cast<int>(PyObject_Length (input_list_object));
     if (input_length < 0) {
         input_length = 0;
     }