A high-throughput and memory-efficient inference and serving engine for LLMs
Fix docs build on `main` (#45536)
Signed-off-by: Harry Mellor <19981378+hmellor@users.noreply.github.com>
Harry Mellor committed
Jun 13, 2026 at 16:53 UTC
b3f0a0a0df76dda92ec4b2c9335f77e84adad911
2 files changed
+1
-2
vllm/model_executor/layers/fused_moe/experts/cpu_moe.py
+1
-2
@@ -351,8 +351,7 @@ def prepare_int4_moe_layer_for_cpu(
351
If None, synthetic zeros are created for symmetric quant.
352
353
Returns:
354
- (blocked_w13, blocked_w2, blocked_s13, blocked_s2,
355
- blocked_z13, blocked_z2)
354
+ (blocked_w13, blocked_w2, blocked_s13, blocked_s2, blocked_z13, blocked_z2)
355
"""
356
E = w13_packed.size(0)
357