Merge branch 'main' of https://github.com/parthmannan/fx_graph_converter

 into main

README.md

@@ -12,6 +12,7 @@ torch.compile(net, backend='nvprims_nvfuser')
 ```
 I am using `nvprims_nvfuser` backend here to get FX Graph with higher level semantics. You can use TorchInductor (default backend) and you'll get a graph with lower level ops. For example, Softmax would be broken down into primitive ops.
 Though I don't anticipate any issues, this code has not been tested with TorchInductor.
+
 3. Install pydot package
 ```
 pip install pydot
@@ -82,4 +83,4 @@ edge_shapes = [nx_g.edges[x]['shape'] for x in nx.edge_bfs(nx_g, source=source_n
 
 2. Graph drawing of the extracted graph
 Example below
-![Example Graph][examples/fx_graph_extracted_id_count(1).png]
+![Example Graph](examples/fx_graph_extracted_id_count(1).png)