embeddinggemma-300m Core ML bundle (ANE encoder) — production embedding model

3fa12f0b97b8afe23264f76800afe14af4615ca5/encoder.mlmodelc/analytics/coremldata.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8a1c9388b1b67fce1271a45d02c569e573b7cf348d7c81a0b6874788a1faf51b
+size 243

3fa12f0b97b8afe23264f76800afe14af4615ca5/encoder.mlmodelc/coremldata.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bf423b8bb375f2cc20abe29e19c28c22684b773e8b17fd2ec58f41187d1bab7e
+size 408

3fa12f0b97b8afe23264f76800afe14af4615ca5/encoder.mlmodelc/metadata.json

@@ -0,0 +1,95 @@
+[
+  {
+    "metadataOutputVersion" : "3.0",
+    "storagePrecision" : "Mixed (Float16, Int8)",
+    "outputSchema" : [
+      {
+        "hasShapeFlexibility" : "0",
+        "isOptional" : "0",
+        "dataType" : "Float16",
+        "formattedType" : "MultiArray (Float16 1 × 768)",
+        "shortDescription" : "",
+        "shape" : "[1, 768]",
+        "name" : "embedding",
+        "type" : "MultiArray"
+      }
+    ],
+    "modelParameters" : [
+
+    ],
+    "specificationVersion" : 10,
+    "mlProgramOperationTypeHistogram" : {
+      "Ios18.softmax" : 24,
+      "Ios19.mul" : 487,
+      "Ios18.matmul" : 48,
+      "Ios19.maximum" : 2,
+      "Ios18.rsqrt" : 1,
+      "Ios19.squeeze" : 24,
+      "Ios16.reduceMax" : 1,
+      "Tile" : 49,
+      "Ios18.gather" : 1,
+      "Ios16.reduceSum" : 3,
+      "Ios19.add" : 123,
+      "Ios18.layerNorm" : 145,
+      "Ios19.concat" : 193,
+      "Ios19.constexprBlockwiseShiftScale" : 175,
+      "Ios19.transpose" : 216,
+      "Ios19.realDiv" : 2,
+      "Ios19.expandDims" : 99,
+      "Ios18.conv" : 170,
+      "Ios18.gelu" : 24,
+      "Ios19.split" : 193,
+      "Ios19.sub" : 1,
+      "Ios19.greaterEqual" : 2,
+      "Ios19.select" : 2,
+      "Ios18.abs" : 1,
+      "Ios19.reshape" : 146
+    },
+    "computePrecision" : "Mixed (Float16, Float32, Int32)",
+    "isUpdatable" : "0",
+    "stateSchema" : [
+
+    ],
+    "availability" : {
+      "macOS" : "16.0",
+      "tvOS" : "19.0",
+      "visionOS" : "3.0",
+      "watchOS" : "12.0",
+      "iOS" : "19.0",
+      "macCatalyst" : "19.0"
+    },
+    "modelType" : {
+      "name" : "MLModelType_mlProgram"
+    },
+    "userDefinedMetadata" : {
+      "com.github.apple.coremltools.conversion_date" : "2026-04-23",
+      "com.github.apple.coremltools.source" : "torch==2.8.0",
+      "com.github.apple.coremltools.version" : "9.0",
+      "com.github.apple.coremltools.source_dialect" : "TorchScript"
+    },
+    "inputSchema" : [
+      {
+        "hasShapeFlexibility" : "0",
+        "isOptional" : "0",
+        "dataType" : "Int32",
+        "formattedType" : "MultiArray (Int32 1 × 128)",
+        "shortDescription" : "",
+        "shape" : "[1, 128]",
+        "name" : "input_ids",
+        "type" : "MultiArray"
+      },
+      {
+        "hasShapeFlexibility" : "0",
+        "isOptional" : "0",
+        "dataType" : "Float16",
+        "formattedType" : "MultiArray (Float16 1 × 128)",
+        "shortDescription" : "",
+        "shape" : "[1, 128]",
+        "name" : "attention_mask",
+        "type" : "MultiArray"
+      }
+    ],
+    "generatedClassName" : "encoder",
+    "method" : "predict"
+  }
+]

3fa12f0b97b8afe23264f76800afe14af4615ca5/encoder.mlmodelc/weights/weight.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f81f60eb283f2c1479100beb27bbadfa7efc4576c6511ca3d13751e7680e950e
+size 308616576

3fa12f0b97b8afe23264f76800afe14af4615ca5/model_config.json

@@ -0,0 +1,98 @@
+{
+  "model_name": "embeddinggemma-300m",
+  "architecture": "gemma3-embedding",
+  "tokenizer_repo": "google/embeddinggemma-300m",
+  "parts": {
+    "encoder": "encoder.mlpackage"
+  },
+  "io_contract": {
+    "inputs": {
+      "input_ids": {
+        "shape": [
+          1,
+          128
+        ],
+        "dtype": "int32"
+      },
+      "attention_mask": {
+        "shape": [
+          1,
+          128
+        ],
+        "dtype": "fp16",
+        "doc": "1.0 for valid tokens, 0.0 for pad"
+      }
+    },
+    "outputs": {
+      "embedding": {
+        "shape": [
+          1,
+          768
+        ],
+        "dtype": "fp16",
+        "doc": "L2-normalized; Matryoshka-truncate the last dim"
+      }
+    }
+  },
+  "hidden_size": 768,
+  "num_hidden_layers": 24,
+  "num_attention_heads": 3,
+  "num_key_value_heads": 1,
+  "head_dim": 256,
+  "intermediate_size": 1152,
+  "vocab_size": 262144,
+  "max_seq_len": 128,
+  "sliding_window": 512,
+  "sliding_window_pattern": 6,
+  "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"
+  ],
+  "rms_norm_eps": 1e-06,
+  "rope_theta_global": 1000000.0,
+  "rope_theta_local": 10000.0,
+  "embed_dim": 768,
+  "dense_intermediate_dim": 3072,
+  "pooling": "mean",
+  "normalize": "l2",
+  "matryoshka_dims": [
+    768,
+    512,
+    256,
+    128
+  ],
+  "task_prefixes": {
+    "retrieval_query": "task: search result | query: ",
+    "retrieval_document": "title: none | text: ",
+    "classification": "task: classification | query: ",
+    "clustering": "task: clustering | query: ",
+    "similarity": "task: sentence similarity | query: ",
+    "code_retrieval": "task: code retrieval | query: ",
+    "question_answering": "task: question answering | query: ",
+    "fact_verification": "task: fact checking | query: "
+  },
+  "quantization": "int8",
+  "compute_units": "CPU_AND_NE"
+}