Upload folder using huggingface_hub

.gitattributes

@@ -33,3 +33,5 @@ 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
+model/embeddinggemma-300m_float32_static.aimodel/main.mlirb filter=lfs diff=lfs merge=lfs -text
+model/tokenizer/tokenizer.json filter=lfs diff=lfs merge=lfs -text

README.md

@@ -0,0 +1,56 @@
+---
+license: gemma
+tags:
+  - coreai
+  - sentence-similarity
+  - feature-extraction
+  - apple-silicon
+  - on-device
+---
+
+# EmbeddingGemma 300m — Core AI export
+
+[google/embeddinggemma-300m](https://huggingface.co/google/embeddinggemma-300m) as a
+single static Core AI graph: the full sentence-transformers pipeline (transformer →
+mean pooling → dense projection → L2 normalize) runs in-graph, so one call returns a
+normalized 768-d embedding. On-device semantic search / RAG for macOS 27 / iOS 27 beta.
+
+Runs out of the box with [CoreAIKit](https://github.com/john-rocky/coreai-kit)'s
+`TextEmbedder`:
+
+```swift
+let embedder = try await TextEmbedder()   // downloads this repo
+let doc = try await embedder.embed(document: "Tokyo is the capital of Japan.")
+let query = try await embedder.embed(query: "what is the capital of Japan")
+let score = TextEmbedder.cosineSimilarity(doc, query)
+```
+
+Retrieval prompt prefixes (`task: search result | query: ` / `title: none | text: `)
+are applied automatically by `TextEmbedder`.
+
+## Bundle layout
+
+```
+model/
+├── embeddinggemma-300m_float32_static.aimodel
+├── tokenizer/            (HF tokenizer files)
+└── reference.json        (torch reference cosines used by the parity test)
+```
+
+## Graph contract
+
+| | name | shape | dtype |
+|---|---|---|---|
+| input | `input_ids` | [1, 256] | int32 (pad id 0, mask 0 over padding) |
+| input | `attention_mask` | [1, 256] | int32 |
+| output | `embedding` | [1, 768] | fp32, L2-normalized |
+
+Precision: fp32. Cross-runtime parity vs the torch SentenceTransformer pipeline is
+exact to 6 decimal places (see reference.json). fp16 variants (full cast AND
+mixed-precision autocast) produce NaN embeddings on-device — Gemma3 activations
+overflow half precision — so fp32 is shipped; a smaller int8 variant is future work.
+
+## License
+
+Gemma Terms of Use (see the upstream model card). Conversion script: this repo's
+sibling, based on apple/coreai-models' recipe patterns (BSD-3-Clause).

model/embeddinggemma-300m_float32_static.aimodel/main.hash

@@ -0,0 +1 @@
+��,��[V��)j'ՙv�Yˌ�o���v%./

model/embeddinggemma-300m_float32_static.aimodel/main.mlirb

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9af02ca8fc5b1a56bbf1290b6a020327d59976b059cb8cfd6f9cb4e176252e2f
+size 1243973215

model/embeddinggemma-300m_float32_static.aimodel/metadata.json

@@ -0,0 +1,7 @@
+{
+  "assetVersion" : "2.0",
+  "license" : "Gemma",
+  "description" : "EmbeddingGemma-300m text embedding model (mean pooling + dense projection, L2-normalized 768-d). Source: https:\/\/huggingface.co\/google\/embeddinggemma-300m",
+  "creationDate" : "20260612T003758Z",
+  "author" : "Google DeepMind"
+}

model/reference.json

@@ -0,0 +1,47 @@
+{
+  "model": "google/embeddinggemma-300m",
+  "seq_len": 256,
+  "dtype": "float32",
+  "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: "
+  },
+  "texts": {
+    "query_bike": {
+      "kind": "query",
+      "text": "red bicycle parked at the beach"
+    },
+    "query_capital": {
+      "kind": "query",
+      "text": "what is the capital of Japan"
+    },
+    "doc_bike": {
+      "kind": "document",
+      "text": "A crimson bike leaning against a palm tree by the sea."
+    },
+    "doc_tokyo": {
+      "kind": "document",
+      "text": "Tokyo is the capital and largest city of Japan."
+    }
+  },
+  "cosines": {
+    "query_bike|query_capital": 0.14591985940933228,
+    "doc_bike|query_bike": 0.6273255348205566,
+    "doc_bike|query_capital": 0.038044270128011703,
+    "doc_bike|doc_tokyo": 0.2074926644563675,
+    "doc_tokyo|query_bike": 0.08935076743364334,
+    "doc_tokyo|query_capital": 0.5779669284820557
+  }
+}

model/tokenizer/tokenizer.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bcb89366c4787d27671d021cdc006defcd1cb53b02b04213ae8dd47c009c7f71
+size 33385272

model/tokenizer/tokenizer_config.json

@@ -0,0 +1,25 @@
+{
+  "backend": "tokenizers",
+  "boi_token": "<start_of_image>",
+  "bos_token": "<bos>",
+  "clean_up_tokenization_spaces": false,
+  "eoi_token": "<end_of_image>",
+  "eos_token": "<eos>",
+  "image_token": "<image_soft_token>",
+  "is_local": false,
+  "local_files_only": false,
+  "mask_token": "<mask>",
+  "model_max_length": 2048,
+  "model_specific_special_tokens": {
+    "boi_token": "<start_of_image>",
+    "eoi_token": "<end_of_image>",
+    "image_token": "<image_soft_token>"
+  },
+  "pad_token": "<pad>",
+  "padding_side": "right",
+  "sp_model_kwargs": null,
+  "spaces_between_special_tokens": false,
+  "tokenizer_class": "GemmaTokenizer",
+  "unk_token": "<unk>",
+  "use_default_system_prompt": false
+}