Use int8 quantizer in the OSS flow (#6166)

Summary:
Pull Request resolved: #6166

As titled. This change add the ability to supply a qconfig to the `CadenceQuantizer`, and uses `int8` instead of `uint8` in `export_model` as per Cadence request. Going forward, `int8` should be the primary 8-bit type.

Reviewed By: dulinriley

Differential Revision: D64209639

fbshipit-source-id: c8bb385aa75cdeb0cfb92217f4cdc5335a10a3b9

Author: mcremon-meta

backends/cadence/aot/export_example.py

@@ -9,6 +9,8 @@
 import logging
 import tempfile
 
+import torch
+
 from executorch.backends.cadence.aot.ops_registrations import *  # noqa
 from typing import Any, Tuple
 
@@ -17,18 +19,42 @@
     export_to_cadence_edge_executorch,
     fuse_pt2,
 )
+
 from executorch.backends.cadence.aot.quantizer.quantizer import CadenceQuantizer
 from executorch.backends.cadence.runtime import runtime
 from executorch.backends.cadence.runtime.executor import BundledProgramManager
 from executorch.exir import ExecutorchProgramManager
 from torch import nn
+from torch.ao.quantization.observer import HistogramObserver, MinMaxObserver
+from torch.ao.quantization.quantizer.xnnpack_quantizer_utils import (
+    QuantizationConfig,
+    QuantizationSpec,
+)
 
 from .utils import save_bpte_program, save_pte_program
 
 
 FORMAT = "[%(levelname)s %(asctime)s %(filename)s:%(lineno)s] %(message)s"
 logging.basicConfig(level=logging.INFO, format=FORMAT)
 
+act_qspec = QuantizationSpec(
+    dtype=torch.int8,
+    quant_min=-128,
+    quant_max=127,
+    qscheme=torch.per_tensor_affine,
+    is_dynamic=False,
+    observer_or_fake_quant_ctr=HistogramObserver.with_args(eps=2**-12),
+)
+
+wgt_qspec = QuantizationSpec(
+    dtype=torch.int8,
+    quant_min=-128,
+    quant_max=127,
+    qscheme=torch.per_tensor_affine,
+    is_dynamic=False,
+    observer_or_fake_quant_ctr=MinMaxObserver,
+)
+
 
 def export_model(
     model: nn.Module,
@@ -39,8 +65,15 @@ def export_model(
     working_dir = tempfile.mkdtemp(dir="/tmp")
     logging.debug(f"Created work directory {working_dir}")
 
+    qconfig = QuantizationConfig(
+        act_qspec,
+        act_qspec,
+        wgt_qspec,
+        None,
+    )
+
     # Instantiate the quantizer
-    quantizer = CadenceQuantizer()
+    quantizer = CadenceQuantizer(qconfig)
 
     # Convert the model
     converted_model = convert_pt2(model, example_inputs, quantizer)

backends/cadence/aot/quantizer/quantizer.py

@@ -141,13 +141,20 @@ def get_supported_operators(cls) -> List[OperatorConfig]:
 
 
 class CadenceQuantizer(ComposableQuantizer):
-    def __init__(self) -> None:
-        static_qconfig = QuantizationConfig(
-            act_qspec,
-            act_qspec,
-            wgt_qspec,
-            None,
+    def __init__(
+        self, quantization_config: Optional[QuantizationConfig] = None
+    ) -> None:
+        static_qconfig = (
+            QuantizationConfig(
+                act_qspec,
+                act_qspec,
+                wgt_qspec,
+                None,
+            )
+            if not quantization_config
+            else quantization_config
         )
+
         super().__init__(
             [
                 CadenceAtenQuantizer(AddmmPattern(), static_qconfig),