fix: fix colpali preprocessing, add examples to readme

README.md

@@ -63,6 +63,23 @@ embeddings = list(model.embed(documents))
 
 ```
 
+Dense text embedding can also be extended with models which are not in the list of supported models.
+
+```python
+from fastembed import TextEmbedding
+from fastembed.common.model_description import PoolingType, ModelSource
+
+TextEmbedding.add_custom_model(
+    model="intfloat/multilingual-e5-small",
+    pooling=PoolingType.MEAN,
+    normalization=True,
+    sources=ModelSource(hf="intfloat/multilingual-e5-small"),  # can be used with an `url` to load files from a private storage
+    dim=384,
+    model_file="onnx/model.onnx",  # can be used to load an already supported model with another optimization or quantization, e.g. onnx/model_O4.onnx
+)
+model = TextEmbedding(model_name="intfloat/multilingual-e5-small")
+embeddings = list(model.embed(documents))
+```
 
 
 ### 🔱 Sparse text embeddings
@@ -137,6 +154,27 @@ embeddings = list(model.embed(images))
 # ]
 ```
 
+### Late interaction multimodal models (ColPali)
+
+```python
+from fastembed import LateInteractionMultimodalEmbedding
+
+doc_images = [
+    "./path/to/qdrant_pdf_doc_1_screenshot.jpg",
+    "./path/to/colpali_pdf_doc_2_screenshot.jpg",
+]
+
+query = "What is Qdrant?"
+
+model = LateInteractionMultimodalEmbedding(model_name="Qdrant/colpali-v1.3-fp16")
+doc_images_embeddings = list(model.embed_image(doc_images))
+# shape (2, 1030, 128)
+# [array([[-0.03353882, -0.02090454, ..., -0.15576172, -0.07678223]], dtype=float32)]
+query_embedding = model.embed_text(query)
+# shape (1, 20, 128)
+# [array([[-0.00218201,  0.14758301, ...,  -0.02207947,  0.16833496]], dtype=float32)]
+```
+
 ### 🔄 Rerankers
 ```python
 from fastembed.rerank.cross_encoder import TextCrossEncoder

fastembed/late_interaction_multimodal/colpali.py

@@ -197,12 +197,11 @@ def _preprocess_onnx_image_input(
         Returns:
             Dict[str, NumpyArray]: ONNX input with text placeholders.
         """
-
         onnx_input["input_ids"] = np.array(
-            [self.EMPTY_TEXT_PLACEHOLDER for _ in onnx_input["input_ids"]]
+            [self.EMPTY_TEXT_PLACEHOLDER for _ in onnx_input["pixel_values"]]
         )
         onnx_input["attention_mask"] = np.array(
-            [self.EVEN_ATTENTION_MASK for _ in onnx_input["input_ids"]]
+            [self.EVEN_ATTENTION_MASK for _ in onnx_input["pixel_values"]]
         )
         return onnx_input
 

fastembed/late_interaction_multimodal/onnx_multimodal_model.py

@@ -73,8 +73,8 @@ def _load_onnx_model(
             cuda=cuda,
             device_id=device_id,
         )
-        assert self.tokenizer is not None
         self.tokenizer, self.special_token_to_id = load_tokenizer(model_dir=model_dir)
+        assert self.tokenizer is not None
         self.processor = load_preprocessor(model_dir=model_dir)
 
     def load_onnx_model(self) -> None:
@@ -159,10 +159,6 @@ def _embed_documents(
             for batch in pool.ordered_map(iter_batch(documents, batch_size), **params):
                 yield from self._post_process_onnx_text_output(batch)  # type: ignore
 
-    def _build_onnx_image_input(self, encoded: NumpyArray) -> dict[str, NumpyArray]:
-        input_name = self.model.get_inputs()[0].name  # type: ignore[union-attr]
-        return {input_name: encoded}
-
     def onnx_embed_image(self, images: list[ImageInput], **kwargs: Any) -> OnnxOutputContext:
         with contextlib.ExitStack():
             image_files = [
@@ -171,7 +167,7 @@ def onnx_embed_image(self, images: list[ImageInput], **kwargs: Any) -> OnnxOutpu
             ]
             assert self.processor is not None, "Processor is not initialized"
             encoded = np.array(self.processor(image_files))
-        onnx_input = self._build_onnx_image_input(encoded)
+        onnx_input = {"pixel_values": encoded}
         onnx_input = self._preprocess_onnx_image_input(onnx_input, **kwargs)
         model_output = self.model.run(None, onnx_input)  # type: ignore[union-attr]
         embeddings = model_output[0].reshape(len(images), -1)