[V1] Support Mistral3 in V1 (#15950)

Signed-off-by: mgoin <[email protected]>

Author: mgoin

docs/source/models/supported_models.md

@@ -888,7 +888,7 @@ See [this page](#generative-models) for more information on how to use generativ
   * `mistralai/Mistral-Small-3.1-24B-Instruct-2503`, etc.
   *
   * ✅︎
-  *
+  * ✅︎
 - * `MllamaForConditionalGeneration`
   * Llama 3.2
   * T + I<sup>+</sup>

vllm/model_executor/models/mistral3.py

@@ -31,12 +31,12 @@
 from vllm.multimodal.profiling import BaseDummyInputsBuilder, ProcessorInputs
 from vllm.sequence import IntermediateTensors
 
-from .interfaces import (MultiModalEmbeddings, SupportsMultiModal, SupportsPP,
-                         SupportsV0Only)
+from .interfaces import MultiModalEmbeddings, SupportsMultiModal, SupportsPP
 from .pixtral import PixtralHFEncoderInfo, PixtralHFVisionModel
 from .utils import (AutoWeightsLoader, flatten_bn, init_vllm_registered_model,
                     maybe_prefix, merge_multimodal_embeddings)
-from .vision import get_vision_encoder_info, select_patch_features
+from .vision import (get_vision_encoder_info, scatter_patch_features,
+                     select_patch_features)
 
 
 class Mistral3ImagePixelInputs(TypedDict):
@@ -425,7 +425,7 @@ def init_vision_tower_for_llava(
     info=_build_mistral3_info,
     dummy_inputs=Mistral3DummyInputsBuilder)
 class Mistral3ForConditionalGeneration(nn.Module, SupportsMultiModal,
-                                       SupportsPP, SupportsV0Only):
+                                       SupportsPP):
 
     packed_modules_mapping = {
         "qkv_proj": ["q_proj", "k_proj", "v_proj"],
@@ -518,7 +518,7 @@ def _parse_and_validate_image_input(
         return Mistral3ImagePixelInputs(
             type="pixel_values_pixtral",
             pixel_values=flatten_bn(pixel_values),
-            embed_is_patch=embed_is_patch,
+            embed_is_patch=flatten_bn(embed_is_patch),
         )
 
     def _process_image_input(
@@ -557,7 +557,10 @@ def get_multimodal_embeddings(
 
         vision_embeddings = self._process_image_input(image_input)
 
-        return vision_embeddings
+        return scatter_patch_features(
+            vision_embeddings,
+            image_input["embed_is_patch"],
+        )
 
     def get_input_embeddings(
         self,