[Tutorial, QNN] Add tutorial for loading quantized PyTorch model (#5321)

* add pytorch tutorial code and doc stub

* add more docs

* formatting, more docs

* typo fix

* try make sphinx happy

* add performance section

* type and nit fix

* format fix

docs/dev/relay_pass_infra.rst

@@ -612,7 +612,7 @@ sequential pass example could be like the following to enable IR dumping for
     seq = tvm.transform.Sequential([
         relay.transform.InferType(),
         relay.transform.FoldConstant(),
-        relay.transform.PrintIR(),
+        transform.PrintIR(),
         relay.transform.EliminateCommonSubexpr(),
         relay.transform.AlterOpLayout()
     ])

tutorials/frontend/from_pytorch.py

@@ -88,8 +88,8 @@ img = np.expand_dims(img, 0)
 ######################################################################
 # Import the graph to Relay
 # -------------------------
-# Convert PyTorch graph to Relay graph.
-input_name = 'input0'  # only one input, set it to this name
+# Convert PyTorch graph to Relay graph. The input name can be arbitrary.
+input_name = 'input0'
 shape_list = [(input_name, img.shape)]
 mod, params = relay.frontend.from_pytorch(scripted_model,
                                           shape_list)