[Spec] Enable FlashInfer autotune for spec draft (#29595)

python/sglang/test/kits/attention_unittest/attention_methods/dense_attention.py

@@ -276,6 +276,7 @@ class TinyModelConfig:
         self.is_encoder_decoder = False
         self.is_multimodal = False
         self.is_generation = True
+        self.quantization = None
         self.is_hybrid_swa = sliding_window_size is not None
         self.is_local_attention_model = sliding_window_size is not None
         self.attention_chunk_size = None

python/sglang/test/kits/attention_unittest/attention_methods/dsa_attention.py

@@ -242,6 +242,7 @@ class TinyDSAModelConfig:
         self.is_encoder_decoder = False
         self.is_multimodal = False
         self.is_generation = True
+        self.quantization = None
         self.is_hybrid_swa = False
         self.attention_chunk_size = None
         self.sliding_window_size = None

python/sglang/test/kits/attention_unittest/attention_methods/dsv4_attention.py

@@ -265,6 +265,7 @@ class TinyDSV4ModelConfig:
         self.is_encoder_decoder = False
         self.is_multimodal = False
         self.is_generation = True
+        self.quantization = None
         self.is_hybrid_swa = False
         self.is_local_attention_model = False
         self.attention_chunk_size = None

python/sglang/test/kits/attention_unittest/attention_methods/dual_chunk_attention.py

@@ -274,6 +274,7 @@ class TinyDualChunkModelConfig:
         self.is_encoder_decoder = False
         self.is_multimodal = False
         self.is_generation = True
+        self.quantization = None
         self.is_hybrid_swa = False
         self.attention_chunk_size = None
         self.sliding_window_size = None

python/sglang/test/kits/attention_unittest/attention_methods/gdn_attention.py

@@ -178,6 +178,7 @@ class TinyGDNModelConfig:
         self.is_encoder_decoder = False
         self.is_multimodal = False
         self.is_generation = True
+        self.quantization = None
         self.is_hybrid_swa = False
         self.is_local_attention_model = False
         self.attention_chunk_size = None

python/sglang/test/kits/attention_unittest/attention_methods/kda_attention.py

@@ -184,6 +184,7 @@ class TinyKDAModelConfig:
         self.is_encoder_decoder = False
         self.is_multimodal = False
         self.is_generation = True
+        self.quantization = None
         self.is_hybrid_swa = False
         self.is_local_attention_model = False
         self.attention_chunk_size = None

python/sglang/test/kits/attention_unittest/attention_methods/lightning_attention.py

@@ -187,6 +187,7 @@ class TinyLightningModelConfig:
         self.is_encoder_decoder = False
         self.is_multimodal = False
         self.is_generation = True
+        self.quantization = None
         self.is_hybrid_swa = False
         self.is_local_attention_model = False
         self.attention_chunk_size = None

python/sglang/test/kits/attention_unittest/attention_methods/mamba2_attention.py

@@ -275,6 +275,7 @@ class TinyMamba2ModelConfig:
         self.is_encoder_decoder = False
         self.is_multimodal = False
         self.is_generation = True
+        self.quantization = None
         self.is_hybrid_swa = False
         self.is_local_attention_model = False
         self.attention_chunk_size = None

python/sglang/test/kits/attention_unittest/attention_methods/mla_attention.py

@@ -178,6 +178,7 @@ class TinyMLAModelConfig:
         self.is_encoder_decoder = False
         self.is_multimodal = False
         self.is_generation = True
+        self.quantization = None
         self.is_hybrid_swa = False
         self.is_local_attention_model = False
         self.attention_chunk_size = None