[Kernel] Delegate construction of FusedMoEQuantConfig to FusedMoEMethodBase subclasses (#22537)

Signed-off-by: Bill Nell <bnell@redhat.com>

[Kernel] Delegate construction of FusedMoEQuantConfig to FusedMoEMethodBase subclasses (#22537)
Signed-off-by: Bill Nell <bnell@redhat.com>
5963b98b · bnellnm · GitHub · e6585ddb · 5963b98b · 5963b98b
Unverified Commit 5963b98b authored Sep 17, 2025 by bnellnm Committed by GitHub Sep 17, 2025
8 changed files
--- a/vllm/model_executor/layers/quantization/utils/flashinfer_fp4_moe.py
+++ b/vllm/model_executor/layers/quantization/utils/flashinfer_fp4_moe.py
--- a/vllm/model_executor/layers/quantization/utils/flashinfer_utils.py
+++ b/vllm/model_executor/layers/quantization/utils/flashinfer_utils.py
--- a/vllm/model_executor/layers/quantization/utils/fp8_utils.py
+++ b/vllm/model_executor/layers/quantization/utils/fp8_utils.py
--- a/vllm/model_executor/models/bert_with_rope.py
+++ b/vllm/model_executor/models/bert_with_rope.py
--- a/vllm/model_executor/models/deepseek.py
+++ b/vllm/model_executor/models/deepseek.py
--- a/vllm/model_executor/models/minicpm.py
+++ b/vllm/model_executor/models/minicpm.py
--- a/vllm/model_executor/models/qwen3_moe.py
+++ b/vllm/model_executor/models/qwen3_moe.py
--- a/vllm/model_executor/warmup/deep_gemm_warmup.py
+++ b/vllm/model_executor/warmup/deep_gemm_warmup.py