[Frontend] Resettle pooling entrypoints (#29634)

Signed-off-by: wang.yuqi <[email protected]>

Author: noooop

.github/CODEOWNERS

@@ -149,6 +149,7 @@ mkdocs.yaml @hmellor
 /examples/*/pooling/ @noooop
 /tests/models/*/pooling* @noooop
 /tests/entrypoints/pooling @noooop
+/vllm/entrypoints/pooling @aarnphm @chaunceyjiang @noooop
 /vllm/config/pooler.py @noooop
 /vllm/pooling_params.py @noooop
 /vllm/model_executor/layers/pooler.py @noooop

docs/design/io_processor_plugins.md

@@ -77,7 +77,7 @@ The `parse_request` method is used for validating the user prompt and converting
 The `pre_process*` methods take the validated plugin input to generate vLLM's model prompts for regular inference.
 The `post_process*` methods take `PoolingRequestOutput` objects as input and generate a custom plugin output.
 The `validate_or_generate_params` method is used for validating with the plugin any `SamplingParameters`/`PoolingParameters` received with the user request, or to generate new ones if none are specified. The function always returns the validated/generated parameters.
-The `output_to_response` method is used only for online serving and converts the plugin output to the `IOProcessorResponse` type that is then returned by the API Server. The implementation of the `/pooling` serving endpoint is available here [vllm/entrypoints/openai/serving_pooling.py](../../vllm/entrypoints/openai/serving_pooling.py).
+The `output_to_response` method is used only for online serving and converts the plugin output to the `IOProcessorResponse` type that is then returned by the API Server. The implementation of the `/pooling` serving endpoint is available here [vllm/entrypoints/openai/serving_pooling.py](../../vllm/entrypoints/pooling/pooling/serving.py).
 
 An example implementation of a plugin that enables generating geotiff images with the PrithviGeospatialMAE model is available [here](https://github.com/IBM/terratorch/tree/main/terratorch/vllm/plugins/segmentation). Please, also refer to our online ([examples/online_serving/pooling/prithvi_geospatial_mae.py](../../examples/online_serving/pooling/prithvi_geospatial_mae.py)) and offline ([examples/offline_inference/pooling/prithvi_geospatial_mae_io_processor.py](../../examples/offline_inference/pooling/prithvi_geospatial_mae_io_processor.py)) inference examples.
 

docs/serving/openai_compatible_server.md

@@ -351,15 +351,15 @@ The following extra parameters are supported by default:
 ??? code
 
     ```python
-    --8<-- "vllm/entrypoints/openai/protocol.py:embedding-extra-params"
+    --8<-- "vllm/entrypoints/pooling/embed/protocol.py:embedding-extra-params"
     ```
 
 For chat-like input (i.e. if `messages` is passed), these extra parameters are supported instead:
 
 ??? code
 
     ```python
-    --8<-- "vllm/entrypoints/openai/protocol.py:chat-embedding-extra-params"
+    --8<-- "vllm/entrypoints/pooling/embed/protocol.py:chat-embedding-extra-params"
     ```
 
 ### Transcriptions API
@@ -629,7 +629,7 @@ The following [pooling parameters][vllm.PoolingParams] are supported.
 The following extra parameters are supported:
 
 ```python
---8<-- "vllm/entrypoints/openai/protocol.py:classification-extra-params"
+--8<-- "vllm/entrypoints/pooling/classify/protocol.py:classification-extra-params"
 ```
 
 ### Score API
@@ -834,7 +834,7 @@ The following [pooling parameters][vllm.PoolingParams] are supported.
 The following extra parameters are supported:
 
 ```python
---8<-- "vllm/entrypoints/openai/protocol.py:score-extra-params"
+--8<-- "vllm/entrypoints/pooling/score/protocol.py:score-extra-params"
 ```
 
 ### Re-rank API
@@ -915,7 +915,7 @@ The following [pooling parameters][vllm.PoolingParams] are supported.
 The following extra parameters are supported:
 
 ```python
---8<-- "vllm/entrypoints/openai/protocol.py:rerank-extra-params"
+--8<-- "vllm/entrypoints/pooling/score/protocol.py:rerank-extra-params"
 ```
 
 ## Ray Serve LLM

tests/entrypoints/openai/test_run_batch.py

@@ -7,7 +7,7 @@
 
 import pytest
 
-from vllm.entrypoints.openai.protocol import BatchRequestOutput
+from vllm.entrypoints.openai.run_batch import BatchRequestOutput
 
 MODEL_NAME = "hmellor/tiny-random-LlamaForCausalLM"
 

tests/entrypoints/pooling/classify/test_online.py

@@ -7,7 +7,8 @@
 import torch.nn.functional as F
 
 from tests.utils import RemoteOpenAIServer
-from vllm.entrypoints.openai.protocol import ClassificationResponse, PoolingResponse
+from vllm.entrypoints.pooling.classify.protocol import ClassificationResponse
+from vllm.entrypoints.pooling.pooling.protocol import PoolingResponse
 
 MODEL_NAME = "jason9693/Qwen2.5-1.5B-apeach"
 DTYPE = "float32"  # Use float32 to avoid NaN issue

tests/entrypoints/pooling/classify/test_online_vision.py

@@ -7,7 +7,7 @@
 import requests
 
 from tests.utils import RemoteOpenAIServer
-from vllm.entrypoints.openai.protocol import ClassificationResponse
+from vllm.entrypoints.pooling.classify.protocol import ClassificationResponse
 
 VLM_MODEL_NAME = "muziyongshixin/Qwen2.5-VL-7B-for-VideoCls"
 MAXIMUM_VIDEOS = 1

tests/entrypoints/pooling/embed/test_online.py

@@ -15,10 +15,8 @@
 from tests.models.language.pooling.embed_utils import run_embedding_correctness_test
 from tests.models.utils import check_embeddings_close
 from tests.utils import RemoteOpenAIServer
-from vllm.entrypoints.openai.protocol import (
-    EmbeddingResponse,
-    PoolingResponse,
-)
+from vllm.entrypoints.pooling.embed.protocol import EmbeddingResponse
+from vllm.entrypoints.pooling.pooling.protocol import PoolingResponse
 from vllm.platforms import current_platform
 from vllm.transformers_utils.tokenizer import get_tokenizer
 from vllm.utils.serial_utils import (

tests/entrypoints/pooling/embed/test_online_dimensions.py

@@ -11,7 +11,7 @@
 from tests.models.language.pooling.embed_utils import run_embedding_correctness_test
 from tests.models.utils import EmbedModelInfo
 from tests.utils import RemoteOpenAIServer
-from vllm.entrypoints.openai.protocol import EmbeddingResponse
+from vllm.entrypoints.pooling.embed.protocol import EmbeddingResponse
 from vllm.platforms import current_platform
 
 if current_platform.is_rocm():

tests/entrypoints/pooling/embed/test_online_long_text.py

@@ -15,7 +15,7 @@
 import pytest_asyncio
 
 from tests.utils import RemoteOpenAIServer
-from vllm.entrypoints.openai.protocol import EmbeddingResponse
+from vllm.entrypoints.pooling.embed.protocol import EmbeddingResponse
 from vllm.platforms import current_platform
 
 if current_platform.is_rocm():

tests/entrypoints/pooling/embed/test_online_vision.py

@@ -8,7 +8,7 @@
 from transformers import AutoProcessor
 
 from tests.utils import VLLM_PATH, RemoteOpenAIServer
-from vllm.entrypoints.openai.protocol import EmbeddingResponse
+from vllm.entrypoints.pooling.embed.protocol import EmbeddingResponse
 from vllm.multimodal.utils import encode_image_base64, fetch_image
 
 MODEL_NAME = "TIGER-Lab/VLM2Vec-Full"