From 8dd8b6ed78a33dfec9edb0ff85fcd069cb7e045d Mon Sep 17 00:00:00 2001 From: Yejing Lai Date: Thu, 18 Jun 2026 10:16:20 +0800 Subject: [PATCH] [XPU] Fix FP8 block-scaled scheme selection on non-CUDA platforms (#43958) Signed-off-by: Lai, Yejing Co-authored-by: Kunshang Ji --- tests/quantization/test_compressed_tensors.py | 1 + .../quantization/compressed_tensors/compressed_tensors.py | 2 +- 2 files changed, 2 insertions(+), 1 deletion(-) diff --git a/tests/quantization/test_compressed_tensors.py b/tests/quantization/test_compressed_tensors.py index 0ca3df7e912..2620b679b6e 100644 --- a/tests/quantization/test_compressed_tensors.py +++ b/tests/quantization/test_compressed_tensors.py @@ -480,6 +480,7 @@ def test_compressed_tensors_fp8_block_enabled(vllm_runner): assert input_quant_op._forward_method in ( input_quant_op.forward_cuda, input_quant_op.forward_hip, + input_quant_op.forward_xpu, ) llm.apply_model(check_model) diff --git a/vllm/model_executor/layers/quantization/compressed_tensors/compressed_tensors.py b/vllm/model_executor/layers/quantization/compressed_tensors/compressed_tensors.py index 2231b2ca9af..229112739a4 100644 --- a/vllm/model_executor/layers/quantization/compressed_tensors/compressed_tensors.py +++ b/vllm/model_executor/layers/quantization/compressed_tensors/compressed_tensors.py @@ -396,7 +396,7 @@ class CompressedTensorsConfig(QuantizationConfig): ) return supported else: - return False + return not match_exact @staticmethod def _is_nvfp4_format(quant_args: QuantizationArgs):