feat: add qwen3 support (microsoft#6528)

## Why are these changes needed?

Add ollama qwen 3 support

Author: mirpo

python/packages/autogen-ext/src/autogen_ext/models/ollama/_model_info.py

@@ -258,6 +258,13 @@
         "family": ModelFamily.UNKNOWN,
         "structured_output": True,
     },
+    "qwen3": {
+        "vision": False,
+        "function_calling": True,
+        "json_output": True,
+        "family": ModelFamily.UNKNOWN,
+        "structured_output": True,
+    },
     "snowflake-arctic-embed": {
         "vision": False,
         "function_calling": False,
@@ -351,6 +358,15 @@
     "qwen2.5-coder:0.5b": 32768,
     "qwen2.5-coder:1.5b": 32768,
     "qwen2.5-coder:3b": 32768,
+    "qwen3": 40960,
+    "qwen3:0.6b": 40960,
+    "qwen3:1.7b": 40960,
+    "qwen3:4b": 40960,
+    "qwen3:8b": 40960,
+    "qwen3:14b": 40960,
+    "qwen3:30b": 40960,
+    "qwen3:32b": 40960,
+    "qwen3:235b": 40960,
     "snowflake-arctic-embed": 512,
     "starcoder2": 16384,
     "tinyllama": 2048,

python/packages/autogen-ext/tests/models/test_ollama_chat_completion_client.py

@@ -570,7 +570,7 @@ class ResponseType(BaseModel):
 
 
 @pytest.mark.asyncio
-@pytest.mark.parametrize("model", ["qwen2.5:0.5b", "llama3.2:1b"])
+@pytest.mark.parametrize("model", ["qwen2.5:0.5b", "llama3.2:1b", "qwen3:0.6b"])
 async def test_ollama_create_tools(model: str, ollama_client: OllamaChatCompletionClient) -> None:
     def add(x: int, y: int) -> str:
         return str(x + y)
@@ -653,7 +653,7 @@ def add(x: int, y: int) -> str:
 
 
 @pytest.mark.asyncio
-@pytest.mark.parametrize("model", ["qwen2.5:0.5b", "llama3.2:1b"])
+@pytest.mark.parametrize("model", ["qwen2.5:0.5b", "llama3.2:1b", "qwen3:0.6b"])
 async def test_ollama_create_stream_tools(model: str, ollama_client: OllamaChatCompletionClient) -> None:
     def add(x: int, y: int) -> str:
         return str(x + y)