Use dim_t instead of size_t (pytorch#3913)

Summary:
Fix build
Documentation:

[Optional Fixes #issue]
Pull Request resolved: pytorch#3913

Test Plan: Please see a detailed explanation of how to fill out the fields in the relevant sections in PULL_REQUEST.md.

Differential Revision: D19085746

Pulled By: jackm321

fbshipit-source-id: d8a464963100af40d51957928f858820f917c10d

torch_glow/src/PyTorchModelLoader.cpp

@@ -2016,8 +2016,7 @@ Error PyTorchModelLoader::loadReshape(const torch::jit::Node *ptNode) {
   }
 
   return addValueMapping(
-      outputs[0],
-      F_.createReshape("reshape", input, castVector<size_t>(shape)));
+      outputs[0], F_.createReshape("reshape", input, castVector<dim_t>(shape)));
 }
 
 Error PyTorchModelLoader::loadRelu(const torch::jit::Node *ptNode) {