[Model] Add HyperCLOVAX-SEED-Think-32B vision-language model support (#31471)

Signed-off-by: effortprogrammer <yhjhoward7@gmail.com>
Co-authored-by: Cyrus Leung <tlleungac@connect.ust.hk>

tests/entrypoints/openai/test_realtime_validation.py

@@ -118,7 +118,7 @@ async def test_multi_chunk_streaming(
             # JIT compilation
             warmup_done = False
             while not warmup_done:
-                event = await receive_event(ws, timeout=360.0)
+                event = await receive_event(ws, timeout=600.0)
                 if event["type"] in ("transcription.done", "error"):
                     warmup_done = True
 

tests/entrypoints/test_chat_utils.py

@@ -1458,6 +1458,38 @@ def test_parse_chat_messages_context_text_format(
     assert mm_uuids is None
 
 
+def test_parse_chat_messages_openai_format_image_url(
+    phi3v_model_config,
+    image_url,
+):
+    content = [
+        {"type": "image_url", "image_url": {"url": image_url}},
+        {"type": "text", "text": "What's in the image?"},
+    ]
+    conversation, mm_data, mm_uuids = parse_chat_messages(
+        [
+            {
+                "role": "user",
+                "content": content,
+            }
+        ],
+        phi3v_model_config,
+        content_format="openai",
+    )
+
+    assert conversation == [
+        {
+            "role": "user",
+            "content": [
+                {"type": "image"},
+                {"type": "text", "text": "What's in the image?"},
+            ],
+        }
+    ]
+    _assert_mm_data_is_image_input(mm_data, 1)
+    _assert_mm_uuids(mm_uuids, 1, expected_uuids=[None])
+
+
 def test_parse_chat_messages_rejects_too_many_images_in_one_message(
     phi3v_model_config,
     image_url,

tests/models/registry.py

@@ -313,6 +313,10 @@ _TEXT_GENERATION_EXAMPLE_MODELS = {
     "HunYuanMoEV1ForCausalLM": _HfExamplesInfo(
         "tencent/Hunyuan-A13B-Instruct", trust_remote_code=True
     ),
+    "HyperCLOVAXForCausalLM": _HfExamplesInfo(
+        "naver-hyperclovax/HyperCLOVAX-SEED-Think-32B",
+        trust_remote_code=True,
+    ),
     "InternLMForCausalLM": _HfExamplesInfo(
         "internlm/internlm-chat-7b", trust_remote_code=True
     ),
@@ -793,6 +797,10 @@ _MULTIMODAL_EXAMPLE_MODELS = {
         "naver-hyperclovax/HyperCLOVAX-SEED-Vision-Instruct-3B",
         trust_remote_code=True,
     ),
+    "HCXVisionV2ForCausalLM": _HfExamplesInfo(
+        "naver-hyperclovax/HyperCLOVAX-SEED-Think-32B",
+        trust_remote_code=True,
+    ),
     "HunYuanVLForConditionalGeneration": _HfExamplesInfo(
         "tencent/HunyuanOCR",
         hf_overrides={"num_experts": 0},