Urgent model support: support gemma-3-it (#4424)

python/sglang/srt/configs/model_config.py

@@ -391,9 +391,13 @@ def _get_and_verify_dtype(
         dtype = dtype.lower()
         if dtype == "auto":
             if config_dtype == torch.float32:
-                if config.model_type == "gemma2":
+                if config.model_type.startswith("gemma"):
+                    if config.model_type == "gemma":
+                        gemma_version = ""
+                    else:
+                        gemma_version = config.model_type[5]
                     logger.info(
-                        "For Gemma 2, we downcast float32 to bfloat16 instead "
+                        f"For Gemma {gemma_version}, we downcast float32 to bfloat16 instead "
                         "of float16 by default. Please specify `dtype` if you "
                         "want to use float16."
                     )
@@ -453,6 +457,7 @@ multimodal_model_archs = [
     "LlavaQwenForCausalLM",
     "LlavaMistralForCausalLM",
     "LlavaVidForCausalLM",
+    "Gemma3ForConditionalGeneration",
     "Grok1VForCausalLM",
     "Grok1AForCausalLM",
     "MllamaForConditionalGeneration",