Update on "[ET-VK][AOT] Enable exporting Q8 Quantized Linear + Convolution"

As title. Introduce fusion patterns to enable fusing quantized convolution and linear graph patterns into a custom op.

## Changes

Introduce the concept of using custom pattern detection functions to detect graph patterns rather than solely relying on SubgraphMatcher. The issue with SubgraphMatcher is that a large number of graph patterns may need to be exported to obtain variants for different combinations of decompositions/quantization workflows. Having a custom detection function improves maintainability.

Implement detection + replacement functions for quantized linear and quantized conv2d.

Differential Revision: [D81323425](https://our.internmc.facebook.com/intern/diff/D81323425/)

[ghstack-poisoned]

Author: ssjia

backends/vulkan/custom_ops_lib.py

@@ -408,7 +408,6 @@ def conv2d_q8ta_q8csw(
 
     # Reshape to original 4D format (OC, IC, H, W)
     qweights_4d = qweights_transposed.view(OC, IC, H, W)
-    print(qweights_4d.shape)
 
     # Dequantize weights
     weights = torch.ops.quantized_decomposed.dequantize_per_channel(
@@ -420,8 +419,6 @@ def conv2d_q8ta_q8csw(
         127,
         torch.int8,
     )
-    print(weights.shape)
-    print(x.shape)
 
     # Perform convolution
     out = torch.nn.functional.conv2d(

backends/vulkan/test/test_vulkan_delegate.py

@@ -104,14 +104,6 @@ def quantize_and_lower_module(
 
     program = export(program, sample_inputs, dynamic_shapes=dynamic_shapes)
 
-    print(program.graph_module.graph)
-    # test = to_edge(
-    #     program,
-    #     compile_config=EdgeCompileConfig(_check_ir_validity=False),
-    # )
-    # print(program.exported_program().graph_module)
-    # raise Exception("stop")
-
     edge_program = to_edge_transform_and_lower(
         program,
         compile_config=edge_compile_config,