Add Llama4 support (#5092)

Co-authored-by: Cheng Wan <cwan39@gatech.edu>
Co-authored-by: fzyzcjy <ch271828n@outlook.com>
Co-authored-by: ispobock <ispobaoke@163.com>

python/sglang/srt/configs/model_config.py

@@ -65,6 +65,9 @@ class ModelConfig:
             **kwargs,
         )
         self.hf_text_config = get_hf_text_config(self.hf_config)
+        self.attention_chunk_size = getattr(
+            self.hf_text_config, "attention_chunk_size", None
+        )
 
         # Check model type
         self.is_generation = is_generation_model(
@@ -467,6 +470,7 @@ multimodal_model_archs = [
     "Gemma3ForConditionalGeneration",
     "Grok1VForCausalLM",
     "Grok1AForCausalLM",
+    # TODO: add multimodal support for "Llama4ForConditionalGeneration",
     "LlavaLlamaForCausalLM",
     "LlavaMistralForCausalLM",
     "LlavaQwenForCausalLM",