Quantize aten.concat like aten.cat for Ethos-U

3l1 · web-flow · commit b3e11d35aea9 · 2026-06-30T05:47:10.000Z
Differential Revision: D110085289 Pull Request resolved: #20600
diff --git a/backends/arm/quantizer/quantization_annotator.py b/backends/arm/quantizer/quantization_annotator.py
@@ -859,6 +859,7 @@ def any_or_hardtanh_min_zero(n: Node):
         )
     elif node.target in (
         torch.ops.aten.cat.default,
+        torch.ops.aten.concat.default,
         torch.ops.aten.concatenate.default,
         torch.ops.aten.stack.default,
     ):

Original file line number	Diff line number	Diff line change
`@@ -859,6 +859,7 @@ def any_or_hardtanh_min_zero(n: Node):`
`859`	`859`	`)`
`860`	`860`	`elif node.target in (`
`861`	`861`	`torch.ops.aten.cat.default,`
	`862`	`+ torch.ops.aten.concat.default,`
`862`	`863`	`torch.ops.aten.concatenate.default,`
`863`	`864`	`torch.ops.aten.stack.default,`
`864`	`865`	`):`