[Model][VLM] Support multi-images inputs for Phi-3-vision models (#7783)

tests/models/test_phi3v.py

@@ -21,6 +21,7 @@ HF_IMAGE_PROMPTS = IMAGE_ASSETS.prompts({
     "cherry_blossom":
     "<|user|>\n<|image_1|>\nWhat is the season?<|end|>\n<|assistant|>\n",
 })
+HF_MULTIIMAGE_IMAGE_PROMPT = "<|user|>\n<|image_1|>\n<|image_2|>\nDescribe these images.<|end|>\n<|assistant|>\n"  # noqa: E501
 
 models = ["microsoft/Phi-3.5-vision-instruct"]
 
@@ -184,3 +185,113 @@ def test_regression_7840(hf_runner, vllm_runner, image_assets, model,
         num_logprobs=10,
         tensor_parallel_size=1,
     )
+
+
+def run_multi_image_test(
+    hf_runner: Type[HfRunner],
+    vllm_runner: Type[VllmRunner],
+    images: List[Image.Image],
+    model: str,
+    *,
+    size_factors: List[float],
+    dtype: str,
+    max_tokens: int,
+    num_logprobs: int,
+    tensor_parallel_size: int,
+    distributed_executor_backend: Optional[str] = None,
+):
+    """Inference result should be the same between hf and vllm.
+
+    All the image fixtures for the test is under tests/images.
+    For huggingface runner, we provide the PIL images as input.
+    For vllm runner, we provide MultiModalDataDict objects 
+    and corresponding MultiModalConfig as input.
+    Note, the text input is also adjusted to abide by vllm contract.
+    The text output is sanitized to be able to compare with hf.
+    """
+
+    inputs_per_case = [
+        ([HF_MULTIIMAGE_IMAGE_PROMPT for _ in size_factors],
+         [[rescale_image_size(image, factor) for image in images]
+          for factor in size_factors])
+    ]
+
+    # NOTE: take care of the order. run vLLM first, and then run HF.
+    # vLLM needs a fresh new process without cuda initialization.
+    # if we run HF first, the cuda initialization will be done and it
+    # will hurt multiprocessing backend with fork method (the default method).
+
+    # max_model_len should be greater than image_feature_size
+    with vllm_runner(model,
+                     max_model_len=4096,
+                     max_num_seqs=1,
+                     limit_mm_per_prompt={"image": len(images)},
+                     dtype=dtype,
+                     tensor_parallel_size=tensor_parallel_size,
+                     distributed_executor_backend=distributed_executor_backend,
+                     enforce_eager=True) as vllm_model:
+        vllm_outputs_per_case = [
+            vllm_model.generate_greedy_logprobs(prompts,
+                                                max_tokens,
+                                                num_logprobs=num_logprobs,
+                                                images=images)
+            for prompts, images in inputs_per_case
+        ]
+
+    hf_model_kwargs = {"_attn_implementation": "eager"}
+    with hf_runner(model, dtype=dtype,
+                   model_kwargs=hf_model_kwargs) as hf_model:
+        eos_token_id = hf_model.processor.tokenizer.eos_token_id
+        hf_outputs_per_case = [
+            hf_model.generate_greedy_logprobs_limit(prompts,
+                                                    max_tokens,
+                                                    num_logprobs=num_logprobs,
+                                                    images=images,
+                                                    eos_token_id=eos_token_id)
+            for prompts, images in inputs_per_case
+        ]
+
+    for hf_outputs, vllm_outputs in zip(hf_outputs_per_case,
+                                        vllm_outputs_per_case):
+        check_logprobs_close(
+            outputs_0_lst=hf_outputs,
+            outputs_1_lst=[
+                vllm_to_hf_output(vllm_output, model)
+                for vllm_output in vllm_outputs
+            ],
+            name_0="hf",
+            name_1="vllm",
+        )
+
+
+@pytest.mark.parametrize("model", models)
+@pytest.mark.parametrize(
+    "size_factors",
+    [
+        # No image
+        [],
+        # Single-scale
+        [1.0],
+        # Single-scale, batched
+        [1.0, 1.0, 1.0],
+        # Multi-scale
+        [0.25, 0.5, 1.0],
+    ],
+)
+@pytest.mark.parametrize("dtype", [target_dtype])
+@pytest.mark.parametrize("max_tokens", [128])
+@pytest.mark.parametrize("num_logprobs", [5])
+def test_multi_images_models(hf_runner, vllm_runner, image_assets, model,
+                             size_factors, dtype: str, max_tokens: int,
+                             num_logprobs: int) -> None:
+    run_multi_image_test(
+        hf_runner,
+        vllm_runner,
+        [asset.pil_image for asset in image_assets],
+        model,
+        size_factors=size_factors,
+        dtype=dtype,
+        max_tokens=max_tokens,
+        num_logprobs=num_logprobs,
+        tensor_parallel_size=1,
+    )

vllm/model_executor/models/phi3v.py

@@ -13,6 +13,7 @@
 # WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
 # See the License for the specific language governing permissions and
 # limitations under the License.
+import itertools
 import re
 from functools import lru_cache
 from typing import (Any, Dict, Iterable, List, Literal, Mapping, Optional,
@@ -37,11 +38,11 @@ from vllm.model_executor.models.clip import CLIPVisionModel
 from vllm.model_executor.models.llama import LlamaModel
 from vllm.model_executor.sampling_metadata import SamplingMetadata
 from vllm.multimodal import MULTIMODAL_REGISTRY
-from vllm.multimodal.utils import cached_get_tokenizer
+from vllm.multimodal.utils import cached_get_tokenizer, repeat_and_pad_token
 from vllm.sequence import IntermediateTensors, SamplerOutput
+from vllm.utils import is_list_of
 
-from .clip import (dummy_image_for_clip, dummy_seq_data_for_clip,
-                   input_processor_for_clip)
+from .clip import dummy_image_for_clip, dummy_seq_data_for_clip
 from .interfaces import SupportsMultiModal
 from .utils import merge_multimodal_embeddings
 
@@ -400,9 +401,20 @@ def input_processor_for_phi3v(ctx: InputContext, llm_inputs: LLMInputs):
     image_data = multi_modal_data["image"]
     if isinstance(image_data, Image.Image):
         w, h = image_data.size
-        image_feature_size = get_phi3v_image_feature_size(hf_config,
+        image_feature_size = [
+            get_phi3v_image_feature_size(hf_config,
                                          input_width=w,
                                          input_height=h)
+        ]
+        image_data = [image_data]
+    elif is_list_of(image_data, Image.Image):
+        image_feature_size = []
+        for image in image_data:
+            w, h = image.size
+            image_feature_size.append(
+                get_phi3v_image_feature_size(hf_config,
+                                             input_width=w,
+                                             input_height=h))
     elif isinstance(image_data, torch.Tensor):
         image_feature_size = image_data.shape[0]
     else:
@@ -410,45 +422,61 @@ def input_processor_for_phi3v(ctx: InputContext, llm_inputs: LLMInputs):
 
     prompt = llm_inputs.get("prompt")
     if prompt is None:
+        image_idx = []
         new_prompt = None
     else:
+        image_idx = sorted(map(int, re.findall(r"<\|image_(\d+)\|>+", prompt)))
         if prompt.count("<|image|>") > 0:
             logger.warning("Please follow the prompt format that is "
                            "documented on HuggingFace which does not involve "
                            "repeating <|image|> tokens.")
-        elif len(re.findall(r"(<\|image_\d+\|>)+", prompt)) > 1:
-            logger.warning("Multiple image input is not supported yet, "
-                           "so any extra image tokens will be treated "
-                           "as plain text.")
-
+        elif (num_image_tags := len(image_idx)) > 1:
+            assert num_image_tags == len(
+                image_data), "The count of image_placeholder not match image's"
         new_prompt = prompt
 
-    prompt_token_ids = llm_inputs["prompt_token_ids"]
-    image_1_token_ids = _get_image_placeholder_token_ids(model_config, idx=1)
+    prompt_token_ids = llm_inputs["prompt_token_ids"].copy()
+
+    # masked place_holder with image token id
+    for idx in image_idx:
+        image_token_ids = _get_image_placeholder_token_ids(model_config,
+                                                           idx=idx)
+        for i in range(len(prompt_token_ids) - len(image_token_ids) + 1):
+            if prompt_token_ids[i:i + len(image_token_ids)] == image_token_ids:
+                prompt_token_ids[i:i + len(image_token_ids)] = [
+                    _IMAGE_TOKEN_ID
+                ] * len(image_token_ids)
+                break
 
-    new_token_ids: List[int] = []
-    for i in range(len(prompt_token_ids) - len(image_1_token_ids) + 1):
-        if prompt_token_ids[i:i + len(image_1_token_ids)] == image_1_token_ids:
-            new_token_ids.append(_IMAGE_TOKEN_ID)
+    # merge consecutive tag ids
+    merged_token_ids: List[int] = []
+    for is_placeholder, token_ids in itertools.groupby(
+            prompt_token_ids, lambda x: x == _IMAGE_TOKEN_ID):
+        if is_placeholder:
+            merged_token_ids.append(_IMAGE_TOKEN_ID)
+        else:
+            merged_token_ids.extend(list(token_ids))
 
-            # No need to further scan the list since we only replace once
-            new_token_ids.extend(prompt_token_ids[i + len(image_1_token_ids):])
-            break
+    # TODO: Move this to utils or integrate with clip.
+    new_token_ids: List[int] = []
+    placeholder_idx = 0
+    while merged_token_ids:
+        token_id = merged_token_ids.pop(0)
+        if token_id == _IMAGE_TOKEN_ID:
+            new_token_ids.extend(
+                repeat_and_pad_token(
+                    _IMAGE_TOKEN_ID,
+                    repeat_count=image_feature_size[placeholder_idx],
+                ))
+            placeholder_idx += 1
         else:
-            new_token_ids.append(prompt_token_ids[i])
+            new_token_ids.append(token_id)
 
     # NOTE: Create a defensive copy of the original inputs
     llm_inputs = LLMInputs(prompt_token_ids=new_token_ids,
                            prompt=new_prompt,
                            multi_modal_data=multi_modal_data)
-
-    return input_processor_for_clip(
-        model_config,
-        CLIP_VIT_LARGE_PATCH14_336_CONFIG,
-        llm_inputs,
-        image_token_id=_IMAGE_TOKEN_ID,
-        image_feature_size_override=image_feature_size,
-    )
+    return llm_inputs
 
 
 @MULTIMODAL_REGISTRY.register_image_input_mapper()