Upload folder using huggingface_hub

.gitattributes

@@ -33,3 +33,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zip filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
+tokenizer.json filter=lfs diff=lfs merge=lfs -text

1_Pooling/config.json

@@ -0,0 +1,10 @@
+{
+    "word_embedding_dimension": 768,
+    "pooling_mode_cls_token": false,
+    "pooling_mode_mean_tokens": true,
+    "pooling_mode_max_tokens": false,
+    "pooling_mode_mean_sqrt_len_tokens": false,
+    "pooling_mode_weightedmean_tokens": false,
+    "pooling_mode_lasttoken": false,
+    "include_prompt": true
+}

2_Dense/config.json

@@ -0,0 +1,6 @@
+{
+    "in_features": 768,
+    "out_features": 3072,
+    "bias": false,
+    "activation_function": "torch.nn.modules.linear.Identity"
+}

2_Dense/model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c327f2acb00149676ade24a75e11eb6ebbd367f9ee050267ba56829d2979f702
+size 9437272

3_Dense/config.json

@@ -0,0 +1,6 @@
+{
+    "in_features": 3072,
+    "out_features": 768,
+    "bias": false,
+    "activation_function": "torch.nn.modules.linear.Identity"
+}

3_Dense/model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ffb6cc5162e11e2ce6bc2367e121ee3bbbc4e82e1ee26826bd7573d4948d81b8
+size 9437272

README.md

@@ -0,0 +1,445 @@
+---
+license: gemma
+pipeline_tag: sentence-similarity
+library_name: sentence-transformers
+tags:
+- sentence-transformers
+- sentence-similarity
+- feature-extraction
+- text-embeddings-inference
+extra_gated_heading: Access EmbeddingGemma on Hugging Face
+extra_gated_prompt: To access EmbeddingGemma on Hugging Face, you’re required to review and
+  agree to Google’s usage license. To do this, please ensure you’re logged in to Hugging
+  Face and click below. Requests are processed immediately.
+extra_gated_button_content: Acknowledge license
+---
+
+# EmbeddingGemma model card
+
+**Model Page**: [EmbeddingGemma](https://ai.google.dev/gemma/docs/embeddinggemma)
+
+**Resources and Technical Documentation**:
+
+*   [Responsible Generative AI Toolkit](https://ai.google.dev/responsible)
+*   [EmbeddingGemma on Kaggle](https://www.kaggle.com/models/google/embeddinggemma/)
+*   [EmbeddingGemma on Vertex Model Garden](https://console.cloud.google.com/vertex-ai/publishers/google/model-garden/embeddinggemma)
+
+**Terms of Use**: [Terms](https://ai.google.dev/gemma/terms)
+
+**Authors**: Google DeepMind
+
+## Model Information
+
+### Description
+
+EmbeddingGemma is a 300M parameter, state-of-the-art for its size, open embedding model from Google, built from Gemma 3 (with T5Gemma initialization) and the same research and technology used to create Gemini models. EmbeddingGemma produces vector representations of text, making it well-suited for search and retrieval tasks, including classification, clustering, and semantic similarity search. This model was trained with data in 100+ spoken languages.
+
+The small size and on-device focus makes it possible to deploy in environments with limited resources such as mobile phones, laptops, or desktops, democratizing access to state of the art AI models and helping foster innovation for everyone.
+
+### Inputs and outputs
+
+-   **Input:**
+    -   Text string, such as a question, a prompt, or a document to be embedded
+    -   Maximum input context length of 2048 tokens
+
+-   **Output:**
+    -   Numerical vector representations of input text data
+    -   Output embedding dimension size of 768, with smaller options available (512, 256, or 128) via Matryoshka Representation Learning (MRL). MRL allows users to truncate the output embedding of size 768 to their desired size and then re-normalize for efficient and accurate representation.
+
+### Usage
+
+These model weights are designed to be used with [Sentence Transformers](https://www.SBERT.net), using the [Gemma 3](https://huggingface.co/docs/transformers/main/en/model_doc/gemma3) implementation from [Hugging Face Transformers](https://huggingface.co/docs/transformers/en/index) as the backbone.
+
+First install the Sentence Transformers library:
+
+```bash
+pip install -U sentence-transformers
+```
+
+Then you can load this model and run inference.
+
+```python
+from sentence_transformers import SentenceTransformer
+
+# Download from the 🤗 Hub
+model = SentenceTransformer("google/embeddinggemma-300m")
+
+# Run inference with queries and documents
+query = "Which planet is known as the Red Planet?"
+documents = [
+    "Venus is often called Earth's twin because of its similar size and proximity.",
+    "Mars, known for its reddish appearance, is often referred to as the Red Planet.",
+    "Jupiter, the largest planet in our solar system, has a prominent red spot.",
+    "Saturn, famous for its rings, is sometimes mistaken for the Red Planet."
+]
+query_embeddings = model.encode_query(query)
+document_embeddings = model.encode_document(documents)
+print(query_embeddings.shape, document_embeddings.shape)
+# (768,) (4, 768)
+
+# Compute similarities to determine a ranking
+similarities = model.similarity(query_embeddings, document_embeddings)
+print(similarities)
+# tensor([[0.3011, 0.6359, 0.4930, 0.4889]])
+```
+
+**NOTE**: EmbeddingGemma activations do not support `float16`. Please use `float32` or `bfloat16` as appropriate for your hardware.
+
+## Model Data
+
+### Training Dataset
+
+This model was trained on a dataset of text data that includes a wide variety of sources totaling approximately 320 billion tokens. Here are the key components:
+
+-   **Web Documents**: A diverse collection of web text ensures the model is exposed to a broad range of linguistic styles, topics, and vocabulary. The training dataset includes content in over 100 languages.
+-   **Code and Technical Documents**: Exposing the model to code and technical documentation helps it learn the structure and patterns of programming languages and specialized scientific content, which improves its understanding of code and technical questions.
+-   **Synthetic and Task-Specific Data**: Synthetically training data helps to teach the model specific skills. This includes curated data for tasks like information retrieval, classification, and sentiment analysis, which helps to fine-tune its performance for common embedding applications.
+
+The combination of these diverse data sources is crucial for training a powerful multilingual embedding model that can handle a wide variety of different tasks and data formats.
+
+### Data Preprocessing
+
+Here are the key data cleaning and filtering methods applied to the training data:
+
+-   CSAM Filtering: Rigorous CSAM (Child Sexual Abuse Material) filtering was applied at multiple stages in the data preparation process to ensure the exclusion of harmful and illegal content.
+-   Sensitive Data Filtering: As part of making Gemma pre-trained models safe and reliable, automated techniques were used to filter out certain personal information and other sensitive data from training sets.
+-   Additional methods: Filtering based on content quality and safety in line with [our policies](https://ai.google/static/documents/ai-responsibility-update-published-february-2025.pdf).
+
+## Model Development
+
+### Hardware
+
+EmbeddingGemma was trained using the latest generation of [Tensor Processing Unit (TPU)](https://cloud.google.com/tpu/docs/intro-to-tpu) hardware (TPUv5e), for more details refer to the [Gemma 3 model card](https://ai.google.dev/gemma/docs/core/model_card_3).
+
+### Software
+
+Training was done using [JAX](https://github.com/jax-ml/jax) and [ML Pathways](https://blog.google/technology/ai/introducing-pathways-next-generation-ai-architecture/). For more details refer to the [Gemma 3 model card](https://ai.google.dev/gemma/docs/core/model_card_3).
+
+## Evaluation
+
+### Benchmark Results
+
+The model was evaluated against a large collection of different datasets and metrics to cover different aspects of text understanding.
+
+#### Full Precision Checkpoint
+
+<table>
+  <thead>
+    <tr>
+      <th colspan="3"><strong>MTEB (Multilingual, v2)</strong></th>
+    </tr>
+  </thead>
+  <tbody>
+    <tr>
+      <td><strong>Dimensionality</strong></td>
+      <td><strong>Mean (Task)</strong></td>
+      <td><strong>Mean (TaskType)</strong></td>
+    </tr>
+    <tr>
+      <td>768d</td>
+      <td>61.15</td>
+      <td>54.31</td>
+    </tr>
+    <tr>
+      <td>512d</td>
+      <td>60.71</td>
+      <td>53.89</td>
+    </tr>
+    <tr>
+      <td>256d</td>
+      <td>59.68</td>
+      <td>53.01</td>
+    </tr>
+    <tr>
+      <td>128d</td>
+      <td>58.23</td>
+      <td>51.77</td>
+    </tr>
+  </tbody>
+</table>
+
+<table>
+  <thead>
+    <tr>
+      <th colspan="3"><strong>MTEB (English, v2)</strong></th>
+    </tr>
+  </thead>
+  <tbody>
+    <tr>
+      <td><strong>Dimensionality</strong></td>
+      <td><strong>Mean (Task)</strong></td>
+      <td><strong>Mean (TaskType)</strong></td>
+    </tr>
+    <tr>
+      <td>768d</td>
+      <td>68.36</td>
+      <td>64.15</td>
+    </tr>
+    <tr>
+      <td>512d</td>
+      <td>67.80</td>
+      <td>63.59</td>
+    </tr>
+    <tr>
+      <td>256d</td>
+      <td>66.89</td>
+      <td>62.94</td>
+    </tr>
+    <tr>
+      <td>128d</td>
+      <td>65.09</td>
+      <td>61.56</td>
+    </tr>
+  </tbody>
+</table>
+
+<table>
+  <thead>
+    <tr>
+      <th colspan="3"><strong>MTEB (Code, v1)</strong></th>
+    </tr>
+  </thead>
+  <tbody>
+    <tr>
+      <td><strong>Dimensionality</strong></td>
+      <td><strong>Mean (Task)</strong></td>
+      <td><strong>Mean (TaskType)</strong></td>
+    </tr>
+    <tr>
+      <td>768d</td>
+      <td>68.76</td>
+      <td>68.76</td>
+    </tr>
+    <tr>
+      <td>512d</td>
+      <td>68.48</td>
+      <td>68.48</td>
+    </tr>
+    <tr>
+      <td>256d</td>
+      <td>66.74</td>
+      <td>66.74</td>
+    </tr>
+    <tr>
+      <td>128d</td>
+      <td>62.96</td>
+      <td>62.96</td>
+    </tr>
+  </tbody>
+</table>
+
+#### QAT Checkpoints
+
+<table>
+  <thead>
+    <tr>
+      <th colspan="3"><strong>MTEB (Multilingual, v2)</strong></th>
+    </tr>
+  </thead>
+  <tbody>
+    <tr>
+      <td><strong>Quant config (dimensionality)</strong></td>
+      <td><strong>Mean (Task)</strong></td>
+      <td><strong>Mean (TaskType)</strong></td>
+    </tr>
+    <tr>
+      <td>Q4_0 (768d)</td>
+      <td>60.62</td>
+      <td>53.61</td>
+    </tr>
+    <tr>
+      <td>Q8_0 (768d)</td>
+      <td>60.93</td>
+      <td>53.95</td>
+    </tr>
+    <tr>
+      <td>Mixed Precision* (768d)</td>
+      <td>60.69</td>
+      <td>53.82</td>
+    </tr>
+  </tbody>
+</table>
+
+<table>
+  <thead>
+    <tr>
+      <th colspan="3"><strong>MTEB (English, v2)</strong></th>
+    </tr>
+  </thead>
+  <tbody>
+    <tr>
+      <td><strong>Quant config (dimensionality)</strong></td>
+      <td><strong>Mean (Task)</strong></td>
+      <td><strong>Mean (TaskType)</strong></td>
+    </tr>
+    <tr>
+      <td>Q4_0 (768d)</td>
+      <td>67.91</td>
+      <td>63.64</td>
+    </tr>
+    <tr>
+      <td>Q8_0 (768d)</td>
+      <td>68.13</td>
+      <td>63.85</td>
+    </tr>
+    <tr>
+      <td>Mixed Precision* (768d)</td>
+      <td>67.95</td>
+      <td>63.83</td>
+    </tr>
+  </tbody>
+</table>
+
+<table>
+  <thead>
+    <tr>
+      <th colspan="3"><strong>MTEB (Code, v1)</strong></th>
+    </tr>
+  </thead>
+  <tbody>
+    <tr>
+      <td><strong>Quant config (dimensionality)</strong></td>
+      <td><strong>Mean (Task)</strong></td>
+      <td><strong>Mean (TaskType)</strong></td>
+    </tr>
+    <tr>
+      <td>Q4_0 (768d)</td>
+      <td>67.99</td>
+      <td>67.99</td>
+    </tr>
+    <tr>
+      <td>Q8_0 (768d)</td>
+      <td>68.70</td>
+      <td>68.70</td>
+    </tr>
+    <tr>
+      <td>Mixed Precision* (768d)</td>
+      <td>68.03</td>
+      <td>68.03</td>
+    </tr>
+  </tbody>
+</table>
+
+Note: QAT models are evaluated after quantization
+
+\* Mixed Precision refers to per-channel quantization with int4 for embeddings, feedforward, and projection layers, and int8 for attention (e4_a8_f4_p4).
+
+### Prompt Instructions
+
+EmbeddingGemma can generate optimized embeddings for various use cases—such as document retrieval, question answering, and fact verification—or for specific input types—either a query or a document—using prompts that are prepended to the input strings.
+Query prompts follow the form `task: {task description} | query: ` where the task description varies by the use case, with the default task description being `search result`. Document-style prompts follow the form `title: {title | "none"} | text: ` where the title is either `none` (the default) or the actual title of the document. Note that providing a title, if available, will improve model performance for document prompts but may require manual formatting.
+
+Use the following prompts based on your use case and input data type. These may already be available in the EmbeddingGemma configuration in your modeling framework of choice.
+
+<table>
+  <thead>
+    <tr>
+      <th><br>
+<strong>Use Case (task type enum)</strong></th>
+      <th><br>
+<strong>Descriptions</strong></th>
+      <th><br>
+<strong>Recommended Prompt</strong></th>
+    </tr>
+  </thead>
+  <tbody>
+    <tr>
+      <td><br>
+Retrieval (Query)</td>
+      <td rowspan="4"><br>
+Used to generate embeddings that are optimized for document search or information retrieval</td>
+      <td><br>
+task: search result | query: {content}</td>
+    </tr>
+    <tr>
+      <td><br>
+Retrieval (Document)</td>
+      <td><br>
+title: {title | "none"} | text: {content}</td>
+    </tr>
+    <tr>
+      <td><br>
+Question Answering</td>
+      <td><br>
+task: question answering | query: {content}</td>
+    </tr>
+    <tr>
+      <td><br>
+Fact Verification</td>
+      <td><br>
+task: fact checking | query: {content}</td>
+    </tr>
+    <tr>
+      <td><br>
+Classification</td>
+      <td><br>
+Used to generate embeddings that are optimized to classify texts according to preset labels</td>
+      <td><br>
+task: classification | query: {content}</td>
+    </tr>
+    <tr>
+      <td><br>
+Clustering</td>
+      <td><br>
+Used to generate embeddings that are optimized to cluster texts based on their similarities</td>
+      <td><br>
+task: clustering | query: {content}</td>
+    </tr>
+    <tr>
+      <td><br>
+Semantic Similarity</td>
+      <td><br>
+Used to generate embeddings that are optimized to assess text similarity. This is not intended for retrieval use cases.</td>
+      <td><br>
+task: sentence similarity | query: {content}</td>
+    </tr>
+    <tr>
+      <td><br>
+Code Retrieval</td>
+      <td><br>
+Used to retrieve a code block based on a natural language query, such as <em>sort an array</em> or <em>reverse a linked list</em>. Embeddings of the code blocks are computed using retrieval_document.</td>
+      <td><br>
+task: code retrieval | query: {content}</td>
+    </tr>
+  </tbody>
+</table>
+
+## Usage and Limitations
+
+These models have certain limitations that users should be aware of.
+
+### Intended Usage
+
+Open embedding models have a wide range of applications across various industries and domains. The following list of potential uses is not comprehensive. The purpose of this list is to provide contextual information about the possible use-cases that the model creators considered as part of model training and development.
+
+-   **Semantic Similarity**: Embeddings optimized to assess text similarity, such as recommendation systems and duplicate detection
+-   **Classification**: Embeddings optimized to classify texts according to preset labels, such as sentiment analysis and spam detection
+-  **Clustering**: Embeddings optimized to cluster texts based on their similarities, such as document organization, market research, and anomaly detection
+-   **Retrieval**
+    -   **Document**: Embeddings optimized for document search, such as indexing articles, books, or web pages for search
+    -   **Query**: Embeddings optimized for general search queries, such as custom search
+    -   **Code Query**: Embeddings optimized for retrieval of code blocks based on natural language queries, such as code suggestions and search
+
+-   **Question Answering**: Embeddings for questions in a question-answering system, optimized for finding documents that answer the question, such as chatbox.
+-   **Fact Verification**: Embeddings for statements that need to be verified, optimized for retrieving documents that contain evidence supporting or refuting the statement, such as automated fact-checking systems.
+
+### Limitations
+
+-   Training Data
+    -   The quality and diversity of the training data significantly influence the model's capabilities. Biases or gaps in the training data can lead to limitations in the model's responses.
+    -   The scope of the training dataset determines the subject areas the model can handle effectively.
+
+-   Language Ambiguity and Nuance
+    -   Natural language is inherently complex. Models might struggle to grasp subtle nuances, sarcasm, or figurative language.
+
+### Ethical Considerations and Risks
+
+Risks identified and mitigations:
+
+-   **Perpetuation of biases**: It's encouraged to perform continuous monitoring (using evaluation metrics, human review) and the exploration of de-biasing techniques during model training, fine-tuning, and other use cases.
+-   **Misuse for malicious purposes**: Technical limitations and developer and end-user education can help mitigate against malicious applications of embeddings. Educational resources and reporting mechanisms for users to flag misuse are provided. Prohibited uses of Gemma models are outlined in the [Gemma Prohibited Use Policy](https://ai.google.dev/gemma/prohibited_use_policy).
+-   **Privacy violations**: Models were trained on data filtered for removal of certain personal information and other sensitive data. Developers are encouraged to adhere to privacy regulations with privacy-preserving techniques.
+
+### Benefits
+
+At the time of release, this family of models provides high-performance open embedding model implementations designed from the ground up for responsible AI development compared to similarly sized models. Using the benchmark evaluation metrics described in this document, these models have shown superior performance to other, comparably-sized open model alternatives.

added_tokens.json

@@ -0,0 +1,3 @@
+{
+  "<image_soft_token>": 262144
+}

config.json

@@ -0,0 +1,60 @@
+{
+  "_sliding_window_pattern": 6,
+  "architectures": [
+    "Gemma3TextModel"
+  ],
+  "attention_bias": false,
+  "attention_dropout": 0.0,
+  "attn_logit_softcapping": null,
+  "bos_token_id": 2,
+  "dtype": "float32",
+  "eos_token_id": 1,
+  "final_logit_softcapping": null,
+  "head_dim": 256,
+  "hidden_activation": "gelu_pytorch_tanh",
+  "hidden_size": 768,
+  "initializer_range": 0.02,
+  "intermediate_size": 1152,
+  "layer_types": [
+    "sliding_attention",
+    "sliding_attention",
+    "sliding_attention",
+    "sliding_attention",
+    "sliding_attention",
+    "full_attention",
+    "sliding_attention",
+    "sliding_attention",
+    "sliding_attention",
+    "sliding_attention",
+    "sliding_attention",
+    "full_attention",
+    "sliding_attention",
+    "sliding_attention",
+    "sliding_attention",
+    "sliding_attention",
+    "sliding_attention",
+    "full_attention",
+    "sliding_attention",
+    "sliding_attention",
+    "sliding_attention",
+    "sliding_attention",
+    "sliding_attention",
+    "full_attention"
+  ],
+  "max_position_embeddings": 2048,
+  "model_type": "gemma3_text",
+  "num_attention_heads": 3,
+  "num_hidden_layers": 24,
+  "num_key_value_heads": 1,
+  "pad_token_id": 0,
+  "query_pre_attn_scalar": 256,
+  "rms_norm_eps": 1e-06,
+  "rope_local_base_freq": 10000.0,
+  "rope_scaling": null,
+  "rope_theta": 1000000.0,
+  "sliding_window": 512,
+  "transformers_version": "4.57.0.dev0",
+  "use_bidirectional_attention": true,
+  "use_cache": true,
+  "vocab_size": 262144
+}

config_sentence_transformers.json

@@ -0,0 +1,26 @@
+{
+  "model_type": "SentenceTransformer",
+  "__version__": {
+    "sentence_transformers": "5.1.0",
+    "transformers": "4.57.0.dev0",
+    "pytorch": "2.8.0+cu128"
+  },
+  "prompts": {
+    "query": "task: search result | query: ",
+    "document": "title: none | text: ",
+    "BitextMining": "task: search result | query: ",
+    "Clustering": "task: clustering | query: ",
+    "Classification": "task: classification | query: ",
+    "InstructionRetrieval": "task: code retrieval | query: ",
+    "MultilabelClassification": "task: classification | query: ",
+    "PairClassification": "task: sentence similarity | query: ",
+    "Reranking": "task: search result | query: ",
+    "Retrieval": "task: search result | query: ",
+    "Retrieval-query": "task: search result | query: ",
+    "Retrieval-document": "title: none | text: ",
+    "STS": "task: sentence similarity | query: ",
+    "Summarization": "task: summarization | query: "
+  },
+  "default_prompt_name": null,
+  "similarity_fn_name": "cosine"
+}

generation_config.json

@@ -0,0 +1,7 @@
+{
+  "cache_implementation": "hybrid",
+  "do_sample": true,
+  "top_k": 64,
+  "top_p": 0.95,
+  "transformers_version": "4.57.0.dev0"
+}

model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cbf5a78393b6a033e0b8a63a57549964f7ed5c6fbeb4ba0694214f36123f2fd2
+size 1211486072

modules.json

@@ -0,0 +1,32 @@
+[
+  {
+    "idx": 0,
+    "name": "0",
+    "path": "",
+    "type": "sentence_transformers.models.Transformer"
+  },
+  {
+    "idx": 1,
+    "name": "1",
+    "path": "1_Pooling",
+    "type": "sentence_transformers.models.Pooling"
+  },
+  {
+    "idx": 2,
+    "name": "2",
+    "path": "2_Dense",
+    "type": "sentence_transformers.models.Dense"
+  },
+  {
+    "idx": 3,
+    "name": "3",
+    "path": "3_Dense",
+    "type": "sentence_transformers.models.Dense"
+  },
+  {
+    "idx": 4,
+    "name": "4",
+    "path": "4_Normalize",
+    "type": "sentence_transformers.models.Normalize"
+  }
+]

notebook.ipynb

@@ -0,0 +1,656 @@
+{
+  "cells": [
+    {
+      "cell_type": "markdown",
+      "metadata": {
+        "id": "-u7xRR3DeFXz"
+      },
+      "source": [
+        "##### Copyright 2025 Google LLC."
+      ]
+    },
+    {
+      "cell_type": "code",
+      "execution_count": null,
+      "metadata": {
+        "cellView": "form",
+        "id": "oed1Dh9SeIlD"
+      },
+      "outputs": [],
+      "source": [
+        "#@title Licensed under the Apache License, Version 2.0 (the \"License\");\n",
+        "# you may not use this file except in compliance with the License.\n",
+        "# You may obtain a copy of the License at\n",
+        "#\n",
+        "# https://www.apache.org/licenses/LICENSE-2.0\n",
+        "#\n",
+        "# Unless required by applicable law or agreed to in writing, software\n",
+        "# distributed under the License is distributed on an \"AS IS\" BASIS,\n",
+        "# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.\n",
+        "# See the License for the specific language governing permissions and\n",
+        "# limitations under the License."
+      ]
+    },
+    {
+      "cell_type": "markdown",
+      "metadata": {
+        "id": "UpJl85mfqdUB"
+      },
+      "source": [
+        "<table class=\"tfo-notebook-buttons\" align=\"left\">\n",
+        "  <td>\n",
+        "    <a target=\"_blank\" href=\"https://ai.google.dev/gemma/docs/embeddinggemma/inference-embeddinggemma-with-sentence-transformers\"><img src=\"https://ai.google.dev/static/site-assets/images/docs/notebook-site-button.png\" height=\"32\" width=\"32\" />View on ai.google.dev</a>\n",
+        "  </td>\n",
+        "  <td>\n",
+        "    <a target=\"_blank\" href=\"https://colab.research.google.com/github/google/generative-ai-docs/blob/main/site/en/gemma/docs/embeddinggemma/inference-embeddinggemma-with-sentence-transformers.ipynb\"\"><img src=\"https://www.tensorflow.org/images/colab_logo_32px.png\" />Run in Google Colab</a>\n",
+        "  </td>\n",
+        "  <td>\n",
+        "    <a target=\"_blank\" href=\"https://kaggle.com/kernels/welcome?src=https://github.com/google/generative-ai-docs/blob/main/site/en/gemma/docs/embeddinggemma/inference-embeddinggemma-with-sentence-transformers.ipynb\"><img src=\"https://www.kaggle.com/static/images/logos/kaggle-logo-transparent-300.png\" height=\"32\" width=\"70\"/>Run in Kaggle</a>\n",
+        "  </td>\n",
+        "  <td>\n",
+        "    <a target=\"_blank\" href=\"https://console.cloud.google.com/vertex-ai/workbench/deploy-notebook?download_url=https://raw.githubusercontent.com/google/generative-ai-docs/main/site/en/gemma/docs/embeddinggemma/inference-embeddinggemma-with-sentence-transformers.ipynb\"><img src=\"https://ai.google.dev/images/cloud-icon.svg\" width=\"40\" />Open in Vertex AI</a>\n",
+        "  </td>\n",
+        "  <td>\n",
+        "    <a target=\"_blank\" href=\"https://github.com/google/generative-ai-docs/blob/main/site/en/gemma/docs/embeddinggemma/inference-embeddinggemma-with-sentence-transformers.ipynb\"><img src=\"https://www.tensorflow.org/images/GitHub-Mark-32px.png\" />View source on GitHub</a>\n",
+        "  </td>\n",
+        "</table>"
+      ]
+    },
+    {
+      "cell_type": "markdown",
+      "metadata": {
+        "id": "Sq3lJyEiqqD-"
+      },
+      "source": [
+        "# Generate Embeddings with Sentence Transformers\n",
+        "\n",
+        "EmbeddingGemma is a lightweight, open embedding model designed for fast, high-quality retrieval on everyday devices like mobile phones. At only 308 million parameters, it's efficient enough to run advanced AI techniques, such as Retrieval Augmented Generation (RAG), directly on your local machine with no internet connection required.\n",
+        "\n",
+        "## Setup\n",
+        "\n",
+        "Before starting this tutorial, complete the following steps:\n",
+        "\n",
+        "* Get access to Gemma by logging into [Hugging Face](https://huggingface.co/google/embeddinggemma-300M) and selecting **Acknowledge license** for a Gemma model.\n",
+        "* Generate a Hugging Face [Access Token](https://huggingface.co/docs/hub/en/security-tokens#how-to-manage-user-access-token) and use it to login from Colab.\n",
+        "\n",
+        "This notebook will run on either CPU or GPU."
+      ]
+    },
+    {
+      "cell_type": "markdown",
+      "metadata": {
+        "id": "R3TOEqprq-X3"
+      },
+      "source": [
+        "### Install Python packages\n",
+        "\n",
+        "Install the libraries required for running the EmbeddingGemma model and generating embeddings. Sentence Transformers is a Python framework for text and image embeddings. For more information, see the [Sentence Transformers](https://www.sbert.net/) documentation."
+      ]
+    },
+    {
+      "cell_type": "code",
+      "execution_count": null,
+      "metadata": {
+        "id": "jZFuhT3nrHEK"
+      },
+      "outputs": [],
+      "source": [
+        "!pip install -U sentence-transformers git+https://github.com/huggingface/transformers@v4.56.0-Embedding-Gemma-preview"
+      ]
+    },
+    {
+      "cell_type": "markdown",
+      "metadata": {
+        "id": "O3ttIyfSA0Lj"
+      },
+      "source": [
+        "After you have accepted the license, you need a valid Hugging Face Token to access the model."
+      ]
+    },
+    {
+      "cell_type": "code",
+      "execution_count": null,
+      "metadata": {
+        "id": "WXK1Ev1Sq2iY"
+      },
+      "outputs": [],
+      "source": [
+        "# Login into Hugging Face Hub\n",
+        "from huggingface_hub import login\n",
+        "login()"
+      ]
+    },
+    {
+      "cell_type": "markdown",
+      "metadata": {
+        "id": "NUydcaDBrXDi"
+      },
+      "source": [
+        "### Load Model\n",
+        "\n",
+        "Use the `sentence-transformers` libraries to create an instance of a model class with EmbeddingGemma."
+      ]
+    },
+    {
+      "cell_type": "code",
+      "execution_count": null,
+      "metadata": {
+        "id": "mkpmqlU_rcOd",
+        "outputId": "f8458e59-9a6e-4a89-af83-ffdf391c323a"
+      },
+      "outputs": [
+        {
+          "name": "stdout",
+          "output_type": "stream",
+          "text": [
+            "Device: cuda:0\n",
+            "SentenceTransformer(\n",
+            "  (0): Transformer({'max_seq_length': 2048, 'do_lower_case': False, 'architecture': 'Gemma3TextModel'})\n",
+            "  (1): Pooling({'word_embedding_dimension': 768, 'pooling_mode_cls_token': False, 'pooling_mode_mean_tokens': True, 'pooling_mode_max_tokens': False, 'pooling_mode_mean_sqrt_len_tokens': False, 'pooling_mode_weightedmean_tokens': False, 'pooling_mode_lasttoken': False, 'include_prompt': True})\n",
+            "  (2): Dense({'in_features': 768, 'out_features': 3072, 'bias': False, 'activation_function': 'torch.nn.modules.linear.Identity'})\n",
+            "  (3): Dense({'in_features': 3072, 'out_features': 768, 'bias': False, 'activation_function': 'torch.nn.modules.linear.Identity'})\n",
+            "  (4): Normalize()\n",
+            ")\n",
+            "Total number of parameters in the model: 307581696\n"
+          ]
+        }
+      ],
+      "source": [
+        "import torch\n",
+        "from sentence_transformers import SentenceTransformer\n",
+        "\n",
+        "device = \"cuda\" if torch.cuda.is_available() else \"cpu\"\n",
+        "\n",
+        "model_id = \"google/embeddinggemma-300M\"\n",
+        "model = SentenceTransformer(model_id).to(device=device)\n",
+        "\n",
+        "print(f\"Device: {model.device}\")\n",
+        "print(model)\n",
+        "print(\"Total number of parameters in the model:\", sum([p.numel() for _, p in model.named_parameters()]))"
+      ]
+    },
+    {
+      "cell_type": "markdown",
+      "metadata": {
+        "id": "JxrZ8na0A7Hv"
+      },
+      "source": [
+        "## Generating Embedding\n",
+        "\n",
+        "An embedding is a numerical representation of text, like a word or sentence, that captures its semantic meaning. Essentially, it's a list of numbers (a vector) that allows computers to understand the relationships and context of words.\n",
+        "\n",
+        "Let's see how EmbeddingGemma would process three different words `[\"apple\", \"banana\", \"car\"]`.\n",
+        "\n",
+        "EmbeddingGemma has been trained on vast amounts of text and has learned the relationships between words and concepts."
+      ]
+    },
+    {
+      "cell_type": "code",
+      "execution_count": null,
+      "metadata": {
+        "id": "o0UK8UVAA9b7",
+        "outputId": "37c91847-57de-4a47-9c1a-0adffacd1867"
+      },
+      "outputs": [
+        {
+          "name": "stdout",
+          "output_type": "stream",
+          "text": [
+            "[[-0.18476306  0.00167681  0.03773484 ... -0.07996225 -0.02348064\n",
+            "   0.00976741]\n",
+            " [-0.21189538 -0.02657359  0.02513712 ... -0.08042689 -0.01999852\n",
+            "   0.00512146]\n",
+            " [-0.18924113 -0.02551468  0.04486253 ... -0.06377774 -0.03699806\n",
+            "   0.03973572]]\n",
+            "Embedding 1: (768,)\n",
+            "Embedding 2: (768,)\n",
+            "Embedding 3: (768,)\n"
+          ]
+        }
+      ],
+      "source": [
+        "words = [\"apple\", \"banana\", \"car\"]\n",
+        "\n",
+        "# Calculate embeddings by calling model.encode()\n",
+        "embeddings = model.encode(words)\n",
+        "\n",
+        "print(embeddings)\n",
+        "for idx, embedding in enumerate(embeddings):\n",
+        "    print(f\"Embedding {idx+1} (shape): {embedding.shape}\")"
+      ]
+    },
+    {
+      "cell_type": "markdown",
+      "metadata": {
+        "id": "inuWOAuMBAR7"
+      },
+      "source": [
+        "The model outpus a numerical vector for each sentence. The actual vectors are very long (768), but for simplicity, those are presented with a few dimensions.\n",
+        "\n",
+        "The key isn't the individual numbers themselves, but **the distance between the vectors**. If we were to plot these vectors in a multi-dimensional space, The vectors for `apple` and `banana` would be very close to each other. And the vector for `car` would be far away from the other two."
+      ]
+    },
+    {
+      "cell_type": "markdown",
+      "metadata": {
+        "id": "2oCpMMJUr4RT"
+      },
+      "source": [
+        "## Determining Similarity\n",
+        "\n",
+        "In this section, we use embeddings to determine how sementically similar different sentences are. Here we show examples with high, medieum, and low similarity scores.\n",
+        "\n",
+        "- High Similarity:\n",
+        "  - Sentence A: \"The chef prepared a delicious meal for the guests.\"\n",
+        "  - Sentence B: \"A tasty dinner was cooked by the chef for the visitors.\"\n",
+        "  - Reasoning: Both sentences describe the same event using different words and grammatical structures (active vs. passive voice). They convey the same core meaning.\n",
+        "\n",
+        "- Medium Similarity:\n",
+        "  - Sentence A: \"She is an expert in machine learning.\"\n",
+        "  - Sentence B: \"He has a deep interest in artificial intelligence.\"\n",
+        "  - Reasoning: The sentences are related as machine learning is a subfield of artificial intelligence. However, they talk about different people with different levels of engagement (expert vs. interest).\n",
+        "\n",
+        "- Low Similarity:\n",
+        "  - Sentence A: \"The weather in Tokyo is sunny today.\"\n",
+        "  - Sentence B: \"I need to buy groceries for the week.\"\n",
+        "  - Reasoning: The two sentences are on completely unrelated topics and share no semantic overlap."
+      ]
+    },
+    {
+      "cell_type": "code",
+      "execution_count": null,
+      "metadata": {
+        "id": "VeTEvnTyslyq",
+        "outputId": "b387529f-aad8-4150-e4f1-daef4f30cfc0"
+      },
+      "outputs": [
+        {
+          "name": "stdout",
+          "output_type": "stream",
+          "text": [
+            "🙋‍♂️\n",
+            "['The chef prepared a delicious meal for the guests.', 'A tasty dinner was cooked by the chef for the visitors.']\n",
+            "`-> 🤖 score:  0.8002148\n",
+            "🙋‍♂️\n",
+            "['She is an expert in machine learning.', 'He has a deep interest in artificial intelligence.']\n",
+            "`-> 🤖 score:  0.45417833\n",
+            "🙋‍♂️\n",
+            "['The weather in Tokyo is sunny today.', 'I need to buy groceries for the week.']\n",
+            "`-> 🤖 score:  0.22262995\n"
+          ]
+        }
+      ],
+      "source": [
+        "# The sentences to encode\n",
+        "sentence_high = [\n",
+        "    \"The chef prepared a delicious meal for the guests.\",\n",
+        "    \"A tasty dinner was cooked by the chef for the visitors.\"\n",
+        "]\n",
+        "sentence_medium = [\n",
+        "    \"She is an expert in machine learning.\",\n",
+        "    \"He has a deep interest in artificial intelligence.\"\n",
+        "]\n",
+        "sentence_low = [\n",
+        "    \"The weather in Tokyo is sunny today.\",\n",
+        "    \"I need to buy groceries for the week.\"\n",
+        "]\n",
+        "\n",
+        "for sentence in [sentence_high, sentence_medium, sentence_low]:\n",
+        "    print(\"🙋‍♂️\")\n",
+        "    print(sentence)\n",
+        "    embeddings = model.encode(sentence)\n",
+        "    similarities = model.similarity(embeddings[0], embeddings[1])\n",
+        "    print(\"`-> 🤖 score: \", similarities.numpy()[0][0])"
+      ]
+    },
+    {
+      "cell_type": "markdown",
+      "metadata": {
+        "id": "obfUiizULZE0"
+      },
+      "source": [
+        "### Using Prompts with EmbeddingGemma\n",
+        "\n",
+        "To generate the best embeddings with EmbeddingGemma, you should add an \"instructional prompt\" or \"task\" to the beginning of your input text. These prompts optimize the embeddings for specific tasks, such as document retrieval or question answering, and help the model distinguish between different input types, like a search query versus a document.\n",
+        "\n",
+        "#### How to Apply Prompts\n",
+        "\n",
+        "You can apply a prompt during inference in three ways.\n",
+        "\n",
+        "1.  **Using the `prompt` argument**<br>\n",
+        "    Pass the full prompt string directly to the `encode` method. This gives you precise control.\n",
+        "    ```python\n",
+        "    embeddings = model.encode(\n",
+        "        sentence,\n",
+        "        prompt=\"task: sentence similarity | query: \"\n",
+        "    )\n",
+        "    ```\n",
+        "2.  **Using the `prompt_name` argument**<br>\n",
+        "    Select a predefined prompt by its name. These prompts are loaded from the model's configuration or during its initialization.\n",
+        "    ```python\n",
+        "    embeddings = model.encode(sentence, prompt_name=\"STS\")\n",
+        "    ```\n",
+        "3.  **Using the Default Prompt**<br>\n",
+        "    If you don't specify either `prompt` or `prompt_name`, the system will automatically use the prompt set as `default_prompt_name`, if no default is set, then no prompt is applied.\n",
+        "    ```python\n",
+        "    embeddings = model.encode(sentence)\n",
+        "    ```\n"
+      ]
+    },
+    {
+      "cell_type": "code",
+      "execution_count": null,
+      "metadata": {
+        "id": "0p3qe3WDJV-I",
+        "outputId": "5fa2638e-e67b-479b-fba4-ca89a22cd10e"
+      },
+      "outputs": [
+        {
+          "name": "stdout",
+          "output_type": "stream",
+          "text": [
+            "Available tasks:\n",
+            " query: \"task: search result | query: \"\n",
+            " document: \"title: none | text: \"\n",
+            " BitextMining: \"task: search result | query: \"\n",
+            " Clustering: \"task: clustering | query: \"\n",
+            " Classification: \"task: classification | query: \"\n",
+            " InstructionRetrieval: \"task: code retrieval | query: \"\n",
+            " MultilabelClassification: \"task: classification | query: \"\n",
+            " PairClassification: \"task: sentence similarity | query: \"\n",
+            " Reranking: \"task: search result | query: \"\n",
+            " Retrieval: \"task: search result | query: \"\n",
+            " Retrieval-query: \"task: search result | query: \"\n",
+            " Retrieval-document: \"title: none | text: \"\n",
+            " STS: \"task: sentence similarity | query: \"\n",
+            " Summarization: \"task: summarization | query: \"\n",
+            "--------------------------------------------------------------------------------\n",
+            "🙋‍♂️\n",
+            "['The chef prepared a delicious meal for the guests.', 'A tasty dinner was cooked by the chef for the visitors.']\n",
+            "`-> 🤖 score:  0.9363755\n",
+            "🙋‍♂️\n",
+            "['She is an expert in machine learning.', 'He has a deep interest in artificial intelligence.']\n",
+            "`-> 🤖 score:  0.6425841\n",
+            "🙋‍♂️\n",
+            "['The weather in Tokyo is sunny today.', 'I need to buy groceries for the week.']\n",
+            "`-> 🤖 score:  0.38587403\n"
+          ]
+        }
+      ],
+      "source": [
+        "print(\"Available tasks:\")\n",
+        "for name, prefix in model.prompts.items():\n",
+        "    print(f\" {name}: \\\"{prefix}\\\"\")\n",
+        "print(\"-\"*80)\n",
+        "\n",
+        "for sentence in [sentence_high, sentence_medium, sentence_low]:\n",
+        "    print(\"🙋‍♂️\")\n",
+        "    print(sentence)\n",
+        "    embeddings = model.encode(sentence, prompt_name=\"STS\")\n",
+        "    similarities = model.similarity(embeddings[0], embeddings[1])\n",
+        "    print(\"`-> 🤖 score: \", similarities.numpy()[0][0])\n"
+      ]
+    },
+    {
+      "cell_type": "markdown",
+      "metadata": {
+        "id": "2YAqPXDctw2w"
+      },
+      "source": [
+        "#### Use Case: Retrieval-Augmented Generation (RAG)\n",
+        "\n",
+        "For RAG systems, use the following `prompt_name` values to create specialized embeddings for your queries and documents:\n",
+        "\n",
+        "* **For Queries:** Use `prompt_name=\"Retrieval-query\"`.<br>\n",
+        "    ```python\n",
+        "    query_embedding = model.encode(\n",
+        "        \"How do I use prompts with this model?\",\n",
+        "        prompt_name=\"Retrieval-query\"\n",
+        "    )\n",
+        "    ```\n",
+        "\n",
+        "* **For Documents:** Use `prompt_name=\"Retrieval-document\"`. To further improve document embeddings, you can also include a title by using the `prompt` argument directly:<br>\n",
+        "  * **With a title:**<br>\n",
+        "    ```python\n",
+        "    doc_embedding = model.encode(\n",
+        "        \"The document text...\",\n",
+        "        prompt=\"title: Using Prompts in RAG | text: \"\n",
+        "    )\n",
+        "    ```\n",
+        "  * **Without a title:**<br>\n",
+        "    ```python\n",
+        "    doc_embedding = model.encode(\n",
+        "        \"The document text...\",\n",
+        "        prompt=\"title: none | text: \"\n",
+        "    )\n",
+        "    ```\n",
+        "\n",
+        "#### Further Reading\n",
+        "\n",
+        "* For details on all available EmbeddingGemma prompts, see the [model card](http://ai.google.dev/gemma/docs/embeddinggemma/model_card#prompt_instructions).\n",
+        "* For general information on prompt templates, see the [Sentence Transformer documentation](https://sbert.net/examples/sentence_transformer/applications/computing-embeddings/README.html#prompt-templates).\n",
+        "* For a demo of RAG, see the [Simple RAG example](https://github.com/google-gemini/gemma-cookbook/blob/main/Gemma/%5BGemma_3%5DRAG_with_EmbeddingGemma.ipynb) in the Gemma Cookbook.\n"
+      ]
+    },
+    {
+      "cell_type": "markdown",
+      "metadata": {
+        "id": "aQh-QFAPsswb"
+      },
+      "source": [
+        "## Classification\n",
+        "\n",
+        "Classification is the task of assigning a piece of text to one or more predefined categories or labels. It's one of the most fundamental tasks in Natural Language Processing (NLP).\n",
+        "\n",
+        "A practical application of text classification is customer support ticket routing. This process automatically directs customer queries to the correct department, saving time and reducing manual work."
+      ]
+    },
+    {
+      "cell_type": "code",
+      "execution_count": null,
+      "metadata": {
+        "id": "C2Ufawl-tXvr",
+        "outputId": "347bd68c-dfee-470d-eef7-e3af5d096e91"
+      },
+      "outputs": [
+        {
+          "name": "stdout",
+          "output_type": "stream",
+          "text": [
+            "tensor([[0.4673, 0.5145, 0.3604],\n",
+            "        [0.4191, 0.5010, 0.5966]])\n",
+            "tensor([1, 2])\n",
+            "🙋‍♂️ Excuse me, the app freezes on the login screen. It won't work even when I try to reset my password. -> 🤖 Technical Support\n",
+            "🙋‍♂️ I would like to inquire about your enterprise plan pricing and features for a team of 50 people. -> 🤖 Sales Inquiry\n"
+          ]
+        }
+      ],
+      "source": [
+        "labels = [\"Billing Issue\", \"Technical Support\", \"Sales Inquiry\"]\n",
+        "\n",
+        "sentence = [\n",
+        "    \"Excuse me, the app freezes on the login screen. It won't work even when I try to reset my password.\",\n",
+        "    \"I would like to inquire about your enterprise plan pricing and features for a team of 50 people.\",\n",
+        "]\n",
+        "\n",
+        "# Calculate embeddings by calling model.encode()\n",
+        "label_embeddings = model.encode(labels, prompt_name=\"Classification\")\n",
+        "embeddings = model.encode(sentence, prompt_name=\"Classification\")\n",
+        "\n",
+        "# Calculate the embedding similarities\n",
+        "similarities = model.similarity(embeddings, label_embeddings)\n",
+        "print(similarities)\n",
+        "\n",
+        "idx = similarities.argmax(1)\n",
+        "print(idx)\n",
+        "\n",
+        "for example in sentence:\n",
+        "    print(\"🙋‍♂️\", example, \"-> 🤖\", labels[idx[sentence.index(example)]])"
+      ]
+    },
+    {
+      "cell_type": "markdown",
+      "metadata": {
+        "id": "IRUU2EIDPSmW"
+      },
+      "source": [
+        "## Matryoshka Representation Learning (MRL)\n",
+        "\n",
+        "EmbeddingGemma leverages MRL to provide multiple embedding sizes from one model. It's a clever training method that creates a single, high-quality embedding where the most important information is concentrated at the beginning of the vector.\n",
+        "\n",
+        "This means you can get a smaller but still very useful embedding by simply taking the first `N` dimensions of the full embedding. Using smaller, truncated embeddings is significantly cheaper to store and faster to process, but this efficiency comes at the cost of potential lower quality of embeddings. MRL gives you the power to choose the optimal balance between this speed and accuracy for your application's specific needs.\n",
+        "\n",
+        "Let's use three words `[\"apple\", \"banana\", \"car\"]` and create simplified embeddings to see how MRL works."
+      ]
+    },
+    {
+      "cell_type": "code",
+      "execution_count": null,
+      "metadata": {
+        "id": "B1q1F9I5PYSq",
+        "outputId": "a5b28e04-4783-4d79-ae82-3fac7e554a7a"
+      },
+      "outputs": [
+        {
+          "name": "stdout",
+          "output_type": "stream",
+          "text": [
+            "similarity function:  cosine\n",
+            "tensor([[0.7510, 0.6685]])\n",
+            "🙋‍♂️ apple vs. banana -> 🤖 score:  0.75102395\n",
+            "🙋‍♂️ apple vs. car -> 🤖 score:  0.6684626\n"
+          ]
+        }
+      ],
+      "source": [
+        "def check_word_similarities():\n",
+        "    # Calculate the embedding similarities\n",
+        "    print(\"similarity function: \", model.similarity_fn_name)\n",
+        "    similarities = model.similarity(embeddings[0], embeddings[1:])\n",
+        "    print(similarities)\n",
+        "\n",
+        "    for idx, word in enumerate(words[1:]):\n",
+        "        print(\"🙋‍♂️ apple vs.\", word, \"-> 🤖 score: \", similarities.numpy()[0][idx])\n",
+        "\n",
+        "# Calculate embeddings by calling model.encode()\n",
+        "embeddings = model.encode(words, prompt_name=\"STS\")\n",
+        "\n",
+        "check_word_similarities()"
+      ]
+    },
+    {
+      "cell_type": "markdown",
+      "metadata": {
+        "id": "_iv1xG0TPxkm"
+      },
+      "source": [
+        "Now, for a faster application, you don't need a new model. Simply **truncate** the full embeddings to the first **512 dimensions**. For optimal results, it is also recommended to set `normalize_embeddings=True`, which scales the vectors to a unit length of 1."
+      ]
+    },
+    {
+      "cell_type": "code",
+      "execution_count": null,
+      "metadata": {
+        "id": "9Ue4aWh8PzdL",
+        "outputId": "176dabd4-9d9c-4ce9-c7e5-472ba47ed55f"
+      },
+      "outputs": [
+        {
+          "name": "stdout",
+          "output_type": "stream",
+          "text": [
+            "Embedding 1: (512,)\n",
+            "Embedding 2: (512,)\n",
+            "Embedding 3: (512,)\n",
+            "--------------------------------------------------------------------------------\n",
+            "similarity function:  cosine\n",
+            "tensor([[0.7674, 0.7041]])\n",
+            "🙋‍♂️ apple vs. banana -> 🤖 score:  0.767427\n",
+            "🙋‍♂️ apple vs. car -> 🤖 score:  0.7040509\n"
+          ]
+        }
+      ],
+      "source": [
+        "embeddings = model.encode(words, truncate_dim=512, normalize_embeddings=True)\n",
+        "\n",
+        "for idx, embedding in enumerate(embeddings):\n",
+        "    print(f\"Embedding {idx+1}: {embedding.shape}\")\n",
+        "\n",
+        "print(\"-\"*80)\n",
+        "check_word_similarities()"
+      ]
+    },
+    {
+      "cell_type": "markdown",
+      "metadata": {
+        "id": "lgkmgzfVP24M"
+      },
+      "source": [
+        "In extremely constrained environments, you can further shorten the embeddings to just **256 dimensions**. You can also use the more efficient **dot-product** for similarity calculations instead of the standard **cosine** similarity."
+      ]
+    },
+    {
+      "cell_type": "code",
+      "execution_count": null,
+      "metadata": {
+        "id": "Gi4NlPv-P4RS",
+        "outputId": "656d8d6a-1e79-41be-f17a-cab136bf27ea"
+      },
+      "outputs": [
+        {
+          "name": "stdout",
+          "output_type": "stream",
+          "text": [
+            "Embedding 1: (256,)\n",
+            "Embedding 2: (256,)\n",
+            "Embedding 3: (256,)\n",
+            "--------------------------------------------------------------------------------\n",
+            "similarity function:  dot\n",
+            "tensor([[0.7855, 0.7382]])\n",
+            "🙋‍♂️ apple vs. banana -> 🤖 score:  0.7854644\n",
+            "🙋‍♂️ apple vs. car -> 🤖 score:  0.7382126\n"
+          ]
+        }
+      ],
+      "source": [
+        "model = SentenceTransformer(model_id, truncate_dim=256, similarity_fn_name=\"dot\").to(device=device)\n",
+        "embeddings = model.encode(words, prompt_name=\"STS\", normalize_embeddings=True)\n",
+        "\n",
+        "for idx, embedding in enumerate(embeddings):\n",
+        "    print(f\"Embedding {idx+1}: {embedding.shape}\")\n",
+        "\n",
+        "print(\"-\"*80)\n",
+        "check_word_similarities()"
+      ]
+    },
+    {
+      "cell_type": "markdown",
+      "metadata": {
+        "id": "RYr9uSI_t3fm"
+      },
+      "source": [
+        "## Summary and next steps\n",
+        "\n",
+        "You are now equipped to generate high-quality text embeddings using EmbeddingGemma and the Sentence Transformers library. Apply these skills to build powerful features like semantic similarity, text classification, and Retrieval-Augmented Generation (RAG) systems, and continue exploring what's possible with Gemma models.\n",
+        "\n",
+        "Check out the following docs next:\n",
+        "\n",
+        "* [Fine-tune EmbeddingGemma](https://ai.google.dev/gemma/docs/embeddinggemma/fine-tuning-embeddinggemma-with-sentence-transformers)\n",
+        "* [Simple RAG example](https://github.com/google-gemini/gemma-cookbook/blob/main/Gemma/%5BGemma_3%5DRAG_with_EmbeddingGemma.ipynb) in the Gemma Cookbook\n"
+      ]
+    }
+  ],
+  "metadata": {
+    "colab": {
+      "name": "inference-embeddinggemma-with-sentence-transformers.ipynb",
+      "provenance": [],
+      "toc_visible": true
+    },
+    "kernelspec": {
+      "display_name": "Python 3",
+      "name": "python3"
+    }
+  },
+  "nbformat": 4,
+  "nbformat_minor": 0
+}

sentence_bert_config.json

@@ -0,0 +1,4 @@
+{
+    "max_seq_length": 2048,
+    "do_lower_case": false
+}

special_tokens_map.json

@@ -0,0 +1,33 @@
+{
+  "boi_token": "<start_of_image>",
+  "bos_token": {
+    "content": "<bos>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "eoi_token": "<end_of_image>",
+  "eos_token": {
+    "content": "<eos>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "image_token": "<image_soft_token>",
+  "pad_token": {
+    "content": "<pad>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "unk_token": {
+    "content": "<unk>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  }
+}

tokenizer.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6852f8d561078cc0cebe70ca03c5bfdd0d60a45f9d2e0e1e4cc05b68e9ec329e
+size 33385008

tokenizer.model

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1299c11d7cf632ef3b4e11937501358ada021bbdf7c47638d13c0ee982f2e79c
+size 4689074