Add usage of int8-mixed-bf16 quantization with X86InductorQuantizer (#2668)

* Add the example code for int8-mixed-bf16 quantization in X86Inductor Quantizer
* Specify the quantized operators like input output running precision
* highlight the usage same as regular BF16 Autocast

Author: leslie-fang-intel

Diff for: prototype_source/pt2e_quant_ptq_x86_inductor.rst

@@ -165,11 +165,37 @@ After we get the quantized model, we will further lower it to the inductor backe
 
 ::
 
-    optimized_model = torch.compile(converted_model)
+    with torch.no_grad():
+        optimized_model = torch.compile(converted_model)
+
+        # Running some benchmark
+        optimized_model(*example_inputs)
 
-    # Running some benchmark
-    optimized_model(*example_inputs)
+In a more advanced scenario, int8-mixed-bf16 quantization comes into play. In this instance,
+a Convolution or GEMM operator produces BFloat16 output data type instead of Float32 in the absence
+of a subsequent quantization node. Subsequently, the BFloat16 tensor seamlessly propagates through
+subsequent pointwise operators, effectively minimizing memory usage and potentially enhancing performance.
+The utilization of this feature mirrors that of regular BFloat16 Autocast, as simple as wrapping the
+script within the BFloat16 Autocast context.
+
+::
 
+    with torch.autocast(device_type="cpu", dtype=torch.bfloat16, enabled=True), torch.no_grad():
+        # Turn on Autocast to use int8-mixed-bf16 quantization. After lowering into Inductor CPP Backend,
+        # For operators such as QConvolution and QLinear:
+        # * The input data type is consistently defined as int8, attributable to the presence of a pair
+            of quantization and dequantization nodes inserted at the input.
+        # * The computation precision remains at int8.
+        # * The output data type may vary, being either int8 or BFloat16, contingent on the presence
+        #   of a pair of quantization and dequantization nodes at the output.
+        # For non-quantizable pointwise operators, the data type will be inherited from the previous node,
+        # potentially resulting in a data type of BFloat16 in this scenario.
+        # For quantizable pointwise operators such as QMaxpool2D, it continues to operate with the int8
+        # data type for both input and output.
+        optimized_model = torch.compile(converted_model)
+
+        # Running some benchmark
+        optimized_model(*example_inputs)
 
 Put all these codes together, we will have the toy example code.
 Please note that since the Inductor ``freeze`` feature does not turn on by default yet, run your example code with ``TORCHINDUCTOR_FREEZING=1``.