From 6fe1f90ee77a2cbc23087a9dd451440ac09b876e Mon Sep 17 00:00:00 2001
From: auphelia <jakobapk@web.de>
Date: Mon, 15 Jun 2020 10:53:03 +0100
Subject: [PATCH] [Util] Update fct to validate quant values (update execution
context)
---
src/finn/util/basic.py | 38 ++++++++++++++++++++++----------------
1 file changed, 22 insertions(+), 16 deletions(-)
diff --git a/src/finn/util/basic.py b/src/finn/util/basic.py
index 08d92b388..c4154aad4 100644
--- a/src/finn/util/basic.py
+++ b/src/finn/util/basic.py
@@ -265,12 +265,16 @@ def calculate_signed_dot_prod_range(dt_a, dt_b, len):
return (min_prod, max_prod)
-def update_execution_context(model, node, execution_context):
- for inp in node.input:
- dtype = model.get_tensor_datatype(inp)
- current_values = execution_context[inp]
+def validate_quant_values(model, node_tensors, execution_context, check_values=False):
+ for tensor in node_tensors:
+ dtype = model.get_tensor_datatype(tensor)
+ # if FLOAT32 skip to next tensor
+ if dtype == DataType.FLOAT32:
+ break
+ current_values = execution_context[tensor]
updated_values = current_values
has_to_be_rounded = False
+ # TODO: vectorize with numpy
for value in np.nditer(current_values):
if not dtype.allowed(value):
has_to_be_rounded = True
@@ -280,26 +284,28 @@ def update_execution_context(model, node, execution_context):
warnings.warn(
"The values of tensor {} can't be represented "
"with the set finn datatype ({}), they will be rounded to match the "
- "finn datatype.".format(inp, dtype)
+ "finn datatype.".format(tensor, dtype)
)
# check if rounded values are not too far from original values
max_error = max(np.abs(current_values - updated_values).flatten())
if max_error <= 1e-4:
- # check again if values can now be represented with set finn datatype
- for value in np.nditer(updated_values):
- if not dtype.allowed(value):
- raise Exception(
- """Values can't be represented with set
- finn datatype ({}) for input {}""".format(
- dtype, inp
+ if check_values is True:
+ # check again if values can now be represented with set finn datatype
+ # TODO: vectorize with numpy
+ for value in np.nditer(updated_values):
+ if not dtype.allowed(value):
+ raise Exception(
+ """Values can't be represented with set
+ finn datatype ({}) for input {}""".format(
+ dtype, tensor
+ )
)
- )
- execution_context[inp] = updated_values
+ execution_context[tensor] = updated_values
else:
raise Exception(
- """Values can't be rounded to match set finn
+ """Rounding error is too high to match set finn
datatype ({}) for input {}""".format(
- dtype, inp
+ dtype, tensor
)
)
return execution_context
--
GitLab