From 5428315898e1656653c543a704ca61adb4e3c9df Mon Sep 17 00:00:00 2001
From: Yaman Umuroglu <yamanu@amd.com>
Date: Wed, 11 Jan 2023 12:29:54 +0300
Subject: [PATCH] [QONNX] make valid_predecessor_op_type a method, fallback to
QuantIdentityHandler
---
.../qonnx/qonnx_activation_handlers.py | 30 ++++++++++---------
.../qonnx/quant_act_to_multithreshold.py | 20 +++++++------
2 files changed, 27 insertions(+), 23 deletions(-)
diff --git a/src/finn/transformation/qonnx/qonnx_activation_handlers.py b/src/finn/transformation/qonnx/qonnx_activation_handlers.py
index a50a58507..9819086d8 100644
--- a/src/finn/transformation/qonnx/qonnx_activation_handlers.py
+++ b/src/finn/transformation/qonnx/qonnx_activation_handlers.py
@@ -52,9 +52,7 @@ class QuantActBaseHandler(ABC):
self._q_node = quant_node
self._q_index = quant_node_index
- @property
@classmethod
- @abstractmethod
def valid_predecessor_op_types(self):
"""Defines which op types the preceding node is allowed to have for
this type of activation.
@@ -284,9 +282,11 @@ class QuantReluHandler(QuantActBaseHandler):
"""Class for converting a quantized relu operation expressed in the QONNX
dialect to the FINN ONNX dialect."""
- valid_predecessor_op_types = [
- "Relu",
- ]
+ @classmethod
+ def valid_predecessor_op_types(self):
+ return [
+ "Relu",
+ ]
def _check_compatibility(self):
if self._q_node.op_type == "Quant":
@@ -391,15 +391,17 @@ class QuantIdentityHandler(QuantActBaseHandler):
these are equivalent to quantized identity activations.
"""
- valid_predecessor_op_types = [
- "BatchNormalization",
- "Sub",
- "Add",
- "Mul",
- "Div",
- "DebugMarker",
- None,
- ]
+ @classmethod
+ def valid_predecessor_op_types(self):
+ return [
+ "BatchNormalization",
+ "Sub",
+ "Add",
+ "Mul",
+ "Div",
+ "DebugMarker",
+ None,
+ ]
def _check_compatibility(self):
# Gather parameters to check
diff --git a/src/finn/transformation/qonnx/quant_act_to_multithreshold.py b/src/finn/transformation/qonnx/quant_act_to_multithreshold.py
index 77025ecdf..e0f893f35 100644
--- a/src/finn/transformation/qonnx/quant_act_to_multithreshold.py
+++ b/src/finn/transformation/qonnx/quant_act_to_multithreshold.py
@@ -30,7 +30,10 @@
import warnings
from qonnx.transformation.base import Transformation
-from finn.transformation.qonnx.qonnx_activation_handlers import QuantActBaseHandler
+from finn.transformation.qonnx.qonnx_activation_handlers import (
+ QuantActBaseHandler,
+ QuantIdentityHandler,
+)
def default_filter_function_generator(max_multithreshold_bit_width=8):
@@ -127,7 +130,7 @@ class ConvertQuantActToMultiThreshold(Transformation):
# Check for possible ambiguity in handler selection
valid_predecessors = []
for cls in QuantActBaseHandler.__subclasses__():
- valid_predecessors.extend(cls.valid_predecessor_op_types)
+ valid_predecessors.extend(cls.valid_predecessor_op_types())
if len(valid_predecessors) != len(set(valid_predecessors)):
raise RuntimeError(
"Two or more activation handlers declare the same "
@@ -138,16 +141,15 @@ class ConvertQuantActToMultiThreshold(Transformation):
# Try to find a fitting handler for this Quant activation node
for handler_cls in QuantActBaseHandler.__subclasses__():
- if predecessor_op_type in handler_cls.valid_predecessor_op_types:
+ if predecessor_op_type in handler_cls.valid_predecessor_op_types():
handler = handler_cls(model, n, node_ind)
break
else:
- raise ValueError(
- f"Quant nodes in the activation path and with predecessor "
- f"nodes of type {predecessor_op_type} are currently not "
- f"supported by FINN and can not be converted to "
- f"MultiThreshold nodes."
- )
+ # fall back to QuantIdentityHandler here
+ # it may still not work due to its particular restrictions,
+ # but better than just erroring out without trying
+ handler = QuantIdentityHandler(model, n, node_ind)
+
model = handler.replace_quant_node()
graph_modified = True
return (model, graph_modified)
--
GitLab