fix: Updated nomic ai with mean pooling

Author: hh-space-invader

fastembed/text/pooled_embedding.py

@@ -18,17 +18,6 @@
         },
         "model_file": "onnx/model.onnx",
     },
-    {
-        "model": "nomic-ai/nomic-embed-text-v1.5-Q",
-        "dim": 768,
-        "description": "Text embeddings, Multimodal (text, image), English, 8192 input tokens truncation, Prefixes for queries/documents: necessary, 2024 year.",
-        "license": "apache-2.0",
-        "size_in_GB": 0.13,
-        "sources": {
-            "hf": "nomic-ai/nomic-embed-text-v1.5",
-        },
-        "model_file": "onnx/model_quantized.onnx",
-    },
     {
         "model": "nomic-ai/nomic-embed-text-v1",
         "dim": 768,

fastembed/text/pooled_normalized_embedding.py

@@ -9,6 +9,17 @@
 from fastembed.text.pooled_embedding import PooledEmbedding
 
 supported_pooled_normalized_models = [
+    {
+        "model": "nomic-ai/nomic-embed-text-v1.5-Q",
+        "dim": 768,
+        "description": "Text embeddings, Multimodal (text, image), English, 8192 input tokens truncation, Prefixes for queries/documents: necessary, 2024 year.",
+        "license": "apache-2.0",
+        "size_in_GB": 0.13,
+        "sources": {
+            "hf": "nomic-ai/nomic-embed-text-v1.5",
+        },
+        "model_file": "onnx/model_quantized.onnx",
+    },
     {
         "model": "sentence-transformers/all-MiniLM-L6-v2",
         "dim": 384,

fastembed/text/text_embedding.py

@@ -1,3 +1,4 @@
+import warnings
 from typing import Any, Iterable, Optional, Sequence, Type, Union
 
 import numpy as np
@@ -62,6 +63,14 @@ def __init__(
         **kwargs,
     ):
         super().__init__(model_name, cache_dir, threads, **kwargs)
+        if model_name == "nomic-ai/nomic-embed-text-v1.5-Q":
+            warnings.warn(
+                "The model 'nomic-ai/nomic-embed-text-v1.5-Q' has been updated to include a mean pooling layer. "
+                "Please ensure your usage aligns with the new functionality. "
+                "Support for the previous version without mean pooling will be removed as of version 0.5.2.",
+                UserWarning,
+                stacklevel=2,
+            )
         for EMBEDDING_MODEL_TYPE in self.EMBEDDINGS_REGISTRY:
             supported_models = EMBEDDING_MODEL_TYPE.list_supported_models()
             if any(model_name.lower() == model["model"].lower() for model in supported_models):

tests/test_text_onnx_embeddings.py

@@ -48,7 +48,7 @@
         [-0.15407836, -0.03053198, -3.9138033, 0.1910364, 0.13224715]
     ),
     "nomic-ai/nomic-embed-text-v1.5-Q": np.array(
-        [-0.12525563, 0.38030425, -3.961622, 0.04176439, -0.0758301]
+        [0.0035279, 0.0162738, -0.1893193, 0.0194878, -0.0119418]
     ),
     "thenlper/gte-large": np.array(
         [-0.01920587, 0.00113156, -0.00708992, -0.00632304, -0.04025577]