Skip to content

[Fix] Fix tvm.tir references in Tflite frontend #18940

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
mshr-h merged 1 commit into from
Mar 27, 2026
+4 −4
Merged

[Fix] Fix tvm.tir references in Tflite frontend #18940

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
8 changes: 4 additions & 4 deletions python/tvm/relax/frontend/tflite/tflite_frontend.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -597,8 +597,8 @@ def quantize(x):

# Get min/max of the output dtype. This will be used to ensure that clip a_min/a_max are not
# beyond the dtype range.
qmin = float(tvm.tir.op.min_value(dtype).value)
qmax = float(tvm.tir.op.max_value(dtype).value)
qmin = float(tvm.tirx.min_value(dtype).value)
qmax = float(tvm.tirx.max_value(dtype).value)

# The input expr is a quantized tensor with its scale and zero point. We calculate the
# suitable clip off points based on these scale and zero point.
Expand Down Expand Up @@ -1117,8 +1117,8 @@ def quantize(x):
# Get min/max of the input dtype. This will be used to ensure that
# clip a_min/a_max are not beyond the dtype range.
input_tensor_type_str = self.get_tensor_type_str(input_tensor.tensor.Type())
qmin = float(tvm.tir.op.min_value(input_tensor_type_str).value)
qmax = float(tvm.tir.op.max_value(input_tensor_type_str).value)
qmin = float(tvm.tirx.min_value(input_tensor_type_str).value)
qmax = float(tvm.tirx.max_value(input_tensor_type_str).value)

out = relax.op.clip(
in_expr, min=max(qmin, quantize(-1.0)), max=min(qmax, quantize(1.0))
Expand Down
Loading