Enable rewrites for quantized input models

If the input model for rewriting is quantized:

- Record de-quantized TFRecords
  - enable writing de-quantized calibration data for the training
  - re-generate augmented training data, if needed
- Use quantization-aware training (QAT) to train the replacement models
- Check if replacement model is quantized:
  If source model is quantized, we make sure rewrite's output model
  is quantized too. Right now, only int8 is supported so raising
  an error if any other datatype is present in the output.

Resolves: MLIA-907, MLIA-908, MLIA-927
Signed-off-by: Benjamin Klimczak <benjamin.klimczak@arm.com>
Change-Id: Icb4070a9e6f1fdb5ce36120d73823986e89ac955

1 files changed, 9 insertions, 7 deletions

diff --git a/src/mlia/nn/rewrite/core/graph_edit/cut.py b/src/mlia/nn/rewrite/core/graph_edit/cut.py
index 13a5268..53d5389 100644
--- a/src/mlia/nn/rewrite/core/graph_edit/cut.py
+++ b/src/mlia/nn/rewrite/core/graph_edit/cut.py
@@ -1,9 +1,11 @@
 # SPDX-FileCopyrightText: Copyright 2023, Arm Limited and/or its affiliates.
 # SPDX-License-Identifier: Apache-2.0
 """Cut module."""
+from __future__ import annotations
+
 import os
 from collections import defaultdict
-from typing import Optional
+from pathlib import Path
 
 import tensorflow as tf
 from tensorflow.lite.python.schema_py_generated import ModelT
@@ -25,8 +27,8 @@ def tensors_by_name(subgraph: SubGraphT, names: list) -> list:
 
 def cut_subgraph(
     subgraph: SubGraphT,
-    input_tensor_names: Optional[list],
-    output_tensor_names: Optional[list],
+    input_tensor_names: list | None,
+    output_tensor_names: list | None,
 ) -> None:
     """Change the global inputs and outputs of a graph to the provided named tensors."""
     if input_tensor_names is not None:
@@ -131,11 +133,11 @@ def filter_relabel(src_subgraph: SubGraphT, relabel_filter: set) -> tuple:
 
 
 def cut_model(
-    model_file: str,
-    input_names: Optional[list],
-    output_names: Optional[list],
+    model_file: str | Path,
+    input_names: list | None,
+    output_names: list | None,
     subgraph_index: int,
-    output_file: str,
+    output_file: str | Path,
 ) -> None:
     """Cut subgraphs and simplify a given model."""
     model = load_fb(model_file)