chore(gallery): add EmbeddingGemma 2 variants

Add Q8_0 and BF16 builds for text embeddings through llama.cpp.
Document installation and retrieval task prefixes.

Assisted-by: Codex:gpt-6

docs/content/features/model-gallery.md

@@ -23,6 +23,36 @@ GPT and text generation models might have a license which is not permissive for
 
 ![output](https://github.com/mudler/LocalAI/assets/2420543/7b16676e-d5b1-4c97-89bd-9fa5065c21ad)
 
+## EmbeddingGemma 2 text embeddings
+
+[EmbeddingGemma 2](https://huggingface.co/google/embeddinggemma-2) is available
+for text embeddings through llama.cpp, with Q8_0 and BF16 builds.
+The entries use the 270M-parameter text backbone, return 768-dimensional vectors,
+and set an 8192-token context. They do not configure image, audio, or video input.
+
+Install with automatic variant selection:
+
+```bash
+local-ai models install embeddinggemma-2
+```
+
+To select a specific build, add `--variant embeddinggemma-2` for Q8_0
+or `--variant embeddinggemma-2-bf16` for BF16.
+
+Include the task prefix in each text input. For retrieval, use
+`task: search result | query: {query}` for queries and
+`title: {title} | text: {content}` for documents.
+Use `none` as the title when a document has no title.
+LocalAI does not add these prefixes automatically.
+
+For example, after installing the Q8_0 entry:
+
+```bash
+curl http://localhost:8080/v1/embeddings \
+  -H 'Content-Type: application/json' \
+  -d '{"model":"embeddinggemma-2","input":"task: search result | query: How does local inference work?"}'
+```
+
 ## How it works
 
 Open **Models** in the WebUI. It is the canonical page for a model's complete

gallery/index.yaml

@@ -34529,6 +34529,66 @@
     - filename: granite-embedding-125m-english-f16.gguf
       sha256: e2950cf0228514e0e81c6f0701a62a9e4763990ce660b4a3c0329cd6a4acd4b9
       uri: huggingface://bartowski/granite-embedding-125m-english-GGUF/granite-embedding-125m-english-f16.gguf
+- name: embeddinggemma-2
+  variants:
+    - model: embeddinggemma-2-bf16
+  url: github:mudler/LocalAI/gallery/virtual.yaml@master
+  urls:
+    - https://huggingface.co/google/embeddinggemma-2
+    - https://huggingface.co/ggml-org/embeddinggemma-2-GGUF
+  description: |
+    EmbeddingGemma 2 is Google's multilingual embedding model.
+    This Q8_0 GGUF serves the 270M-parameter text backbone through llama.cpp
+    and produces 768-dimensional embeddings for retrieval and semantic similarity.
+    This entry supports text input with an 8192-token context.
+    Use the model's task prefixes for queries and documents.
+  license: apache-2.0
+  tags:
+    - embeddings
+    - gguf
+    - cpu
+    - gpu
+  overrides:
+    backend: llama-cpp
+    embeddings: true
+    context_size: 8192
+    known_usecases:
+      - embeddings
+    parameters:
+      model: llama-cpp/models/embeddinggemma-2/embeddinggemma-2-Q8_0.gguf
+  files:
+    - filename: llama-cpp/models/embeddinggemma-2/embeddinggemma-2-Q8_0.gguf
+      sha256: 2188ac1deca4b77dffefd603c2776a9d76d9d74ec01841392982ebb840b09135
+      uri: https://huggingface.co/ggml-org/embeddinggemma-2-GGUF/resolve/bfcd298762cc34d0357ece5ebdd31791a3a374d8/embeddinggemma-2-Q8_0.gguf
+- name: embeddinggemma-2-bf16
+  url: github:mudler/LocalAI/gallery/virtual.yaml@master
+  urls:
+    - https://huggingface.co/google/embeddinggemma-2
+    - https://huggingface.co/ggml-org/embeddinggemma-2-GGUF
+  description: |
+    EmbeddingGemma 2 is Google's multilingual embedding model.
+    This BF16 GGUF serves the 270M-parameter text backbone through llama.cpp
+    and produces 768-dimensional embeddings for retrieval and semantic similarity.
+    This entry supports text input with an 8192-token context.
+    Use the model's task prefixes for queries and documents.
+  license: apache-2.0
+  tags:
+    - embeddings
+    - gguf
+    - cpu
+    - gpu
+  overrides:
+    backend: llama-cpp
+    embeddings: true
+    context_size: 8192
+    known_usecases:
+      - embeddings
+    parameters:
+      model: llama-cpp/models/embeddinggemma-2/embeddinggemma-2-BF16.gguf
+  files:
+    - filename: llama-cpp/models/embeddinggemma-2/embeddinggemma-2-BF16.gguf
+      sha256: 68bae29d62fb8c7d23e98d21fd4662753ddd636e6b62b8a70dfe059a9844f216
+      uri: https://huggingface.co/ggml-org/embeddinggemma-2-GGUF/resolve/bfcd298762cc34d0357ece5ebdd31791a3a374d8/embeddinggemma-2-BF16.gguf
 - name: embeddinggemma-300m
   url: github:mudler/LocalAI/gallery/virtual.yaml@master
   urls: