From be06873198cd544f963b4e2f0f0877b23ab5861e Mon Sep 17 00:00:00 2001
From: Yohann Prigent <prigent.yohann@gmail.com>
Date: Wed, 19 Aug 2026 16:55:59 +0200
Subject: [PATCH] [Bugfix] compressed-tensors: restore int8 grouped WNA16 MoE
 support (#52002)

Signed-off-by: Yohann Prigent <prigent.yohann@gmail.com>
Co-authored-by: Misha Goin <mgoin64@gmail.com>
---
 .../compressed_tensors_moe/compressed_tensors_moe_wna16.py       | 1 -
 1 file changed, 1 deletion(-)

diff --git a/vllm/model_executor/layers/quantization/compressed_tensors/compressed_tensors_moe/compressed_tensors_moe_wna16.py b/vllm/model_executor/layers/quantization/compressed_tensors/compressed_tensors_moe/compressed_tensors_moe_wna16.py
index bd7ad4234a..35894b3224 100644
--- a/vllm/model_executor/layers/quantization/compressed_tensors/compressed_tensors_moe/compressed_tensors_moe_wna16.py
+++ b/vllm/model_executor/layers/quantization/compressed_tensors/compressed_tensors_moe/compressed_tensors_moe_wna16.py
@@ -82,7 +82,6 @@ class CompressedTensorsWNA16MoEMethod(CompressedTensorsMoEMethod):
             else:
                 scale = kInt4StaticGroupScale
         elif self.num_bits == 8:
-            assert self.group_size == -1
             scale = kInt8StaticGroupScale
         else:
             raise ValueError(