Upload folder using huggingface_hub

README.md

@@ -0,0 +1,27 @@
+# Core ML Stable Diffusion Mixed-Bit Palettization Resources
+
+This repository contains resources related to the use of [mixed-bit palettization techniques for Stable Diffusion](https://github.com/apple/ml-stable-diffusion#mbp). As explained in the original repo, this is an advanced compression technique that applies palettization using a varying number of bits per layer to keep quality above the desired threshold (as measured by the PSNR with respect to the original model). This can achieve strong space savings without resorting to learned quantization tables during fine-tuning.
+
+The repository contains the following resources:
+
+- [`recipes/`](tree/main/recipes). JSON files with results from the analysis phase [(Step 1 of the instructions)](https://github.com/apple/ml-stable-diffusion#mbp) on three popular Stable Diffusion models:
+    * Stable Diffusion 1.5
+    * Stable Diffusion 2.1 base
+    * Stable Diffusion XL 1.0 base
+
+You can apply any of the available recipes on the appropriate model to experiment and compare results.
+
+- [`unet-mbp-sdxl-1-base/`](tree/main/unet-mbp-sdxl-1-base). Three pre-applied recipes on the UNet of Stable Diffusion XL 1.0 base:
+
+| Recipe                       | Effective bits | UNet size (GB) | Size reduction |
+|------------------------------|----------------|----------------|----------------|
+| recipe_3_41_bit_mixedpalette | 3.41           | 1.1            | 77%            |
+| recipe_4_50_bit_mixedpalette | 4.50           | 1.4            | 71%            |
+| recipe_6_55_bit_mixedpalette | 6.55           | 2.0            | 58%            |
+
+- [`coreml-stable-diffusion-xl-base_mbp_4_50_palettized`]. A full pipeline of a Core ML version of Stable Diffusion XL 1.0 base, with the `recipe_4_50_bit_mixedpalette` recipe pre-applied to the UNet. You can download it to use in your own Core ML apps.
+
+- [`coreml-stable-diffusion-xl-base_mbp_4_50_palettized.zip`]. An archived version of the same pipeline, for use with [Hugging Face demo app](https://github.com/huggingface/swift-coreml-diffusers) and other third party tools.
+
+This repository was prepared by Apple and Hugging Face in July 2023, from experiments conducted on July 2023 using public beta versions of iOS 17.0, iPadOS 17.0 and macOS 14.0.
+

coreml-stable-diffusion-xl-base_mbp_4_50_palettized.zip

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:64f0b7ee4cbb803bbaf83d4eef69b2868e0da864d8fd7baec988eb8155f92ced
+size 3356775856

coreml-stable-diffusion-xl-base_mbp_4_50_palettized/compiled/TextEncoder.mlmodelc/analytics/coremldata.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f732919fa370a1b7b09ec2b227539269b6543149a2b0dbae95cc4cf350e4b697
+size 207

coreml-stable-diffusion-xl-base_mbp_4_50_palettized/compiled/TextEncoder.mlmodelc/coremldata.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9ff26866d8d8fbb4e53a0628f8aab5f7edf1b3ec763a96e6812c8f7fbf4c9827
+size 825

coreml-stable-diffusion-xl-base_mbp_4_50_palettized/compiled/TextEncoder.mlmodelc/metadata.json

@@ -0,0 +1,82 @@
+[
+  {
+    "shortDescription" : "Stable Diffusion generates images conditioned on text and\/or other images as input through the diffusion process. Please refer to https:\/\/arxiv.org\/abs\/2112.10752 for details.",
+    "metadataOutputVersion" : "3.0",
+    "outputSchema" : [
+      {
+        "hasShapeFlexibility" : "0",
+        "isOptional" : "0",
+        "dataType" : "Float32",
+        "formattedType" : "MultiArray (Float32)",
+        "shortDescription" : "Hidden states after the encoder layers",
+        "shape" : "[]",
+        "name" : "hidden_embeds",
+        "type" : "MultiArray"
+      },
+      {
+        "hasShapeFlexibility" : "0",
+        "isOptional" : "0",
+        "dataType" : "Float32",
+        "formattedType" : "MultiArray (Float32)",
+        "shortDescription" : "The version of the `last_hidden_state` output after pooling",
+        "shape" : "[]",
+        "name" : "pooled_outputs",
+        "type" : "MultiArray"
+      }
+    ],
+    "version" : "diffusers\/stable-diffusion-xl-base-1.0",
+    "modelParameters" : [
+
+    ],
+    "author" : "Please refer to the Model Card available at huggingface.co\/diffusers\/stable-diffusion-xl-base-1.0",
+    "specificationVersion" : 7,
+    "storagePrecision" : "Float16",
+    "license" : "OpenRAIL (https:\/\/huggingface.co\/spaces\/CompVis\/stable-diffusion-license)",
+    "mlProgramOperationTypeHistogram" : {
+      "Ios16.cast" : 3,
+      "Ios16.mul" : 36,
+      "Ios16.layerNorm" : 25,
+      "Stack" : 1,
+      "Transpose" : 60,
+      "Ios16.sigmoid" : 12,
+      "Ios16.linear" : 72,
+      "Ios16.add" : 37,
+      "Ios16.matmul" : 24,
+      "Ios16.softmax" : 12,
+      "Ios16.gatherNd" : 1,
+      "Ios16.gather" : 1,
+      "Ios16.reshape" : 120,
+      "Ios16.reduceArgmax" : 1
+    },
+    "computePrecision" : "Mixed (Float16, Float32, Int32)",
+    "isUpdatable" : "0",
+    "availability" : {
+      "macOS" : "13.0",
+      "tvOS" : "16.0",
+      "watchOS" : "9.0",
+      "iOS" : "16.0",
+      "macCatalyst" : "16.0"
+    },
+    "modelType" : {
+      "name" : "MLModelType_mlProgram"
+    },
+    "inputSchema" : [
+      {
+        "hasShapeFlexibility" : "0",
+        "isOptional" : "0",
+        "dataType" : "Float32",
+        "formattedType" : "MultiArray (Float32 1 × 77)",
+        "shortDescription" : "The token ids that represent the input text",
+        "shape" : "[1, 77]",
+        "name" : "input_ids",
+        "type" : "MultiArray"
+      }
+    ],
+    "userDefinedMetadata" : {
+      "com.github.apple.coremltools.version" : "7.0b1",
+      "com.github.apple.coremltools.source" : "torch==2.0.1+cu117"
+    },
+    "generatedClassName" : "Stable_Diffusion_version_diffusers_stable_diffusion_xl_base_1_0_text_encoder",
+    "method" : "predict"
+  }
+]

coreml-stable-diffusion-xl-base_mbp_4_50_palettized/compiled/TextEncoder.mlmodelc/weights/weight.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a86533724aadf50c8f5539592d440887a484f60002d7967505c69c7faf4d7797
+size 246145536

coreml-stable-diffusion-xl-base_mbp_4_50_palettized/compiled/TextEncoder2.mlmodelc/analytics/coremldata.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:eef66b480388714bb62f2f0f2f97a8953e44acbb00b25f9a9fd63c759f4f0e83
+size 207

coreml-stable-diffusion-xl-base_mbp_4_50_palettized/compiled/TextEncoder2.mlmodelc/coremldata.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9ff26866d8d8fbb4e53a0628f8aab5f7edf1b3ec763a96e6812c8f7fbf4c9827
+size 825

coreml-stable-diffusion-xl-base_mbp_4_50_palettized/compiled/TextEncoder2.mlmodelc/metadata.json

@@ -0,0 +1,82 @@
+[
+  {
+    "shortDescription" : "Stable Diffusion generates images conditioned on text and\/or other images as input through the diffusion process. Please refer to https:\/\/arxiv.org\/abs\/2112.10752 for details.",
+    "metadataOutputVersion" : "3.0",
+    "outputSchema" : [
+      {
+        "hasShapeFlexibility" : "0",
+        "isOptional" : "0",
+        "dataType" : "Float32",
+        "formattedType" : "MultiArray (Float32)",
+        "shortDescription" : "Hidden states after the encoder layers",
+        "shape" : "[]",
+        "name" : "hidden_embeds",
+        "type" : "MultiArray"
+      },
+      {
+        "hasShapeFlexibility" : "0",
+        "isOptional" : "0",
+        "dataType" : "Float32",
+        "formattedType" : "MultiArray (Float32)",
+        "shortDescription" : "The version of the `last_hidden_state` output after pooling",
+        "shape" : "[]",
+        "name" : "pooled_outputs",
+        "type" : "MultiArray"
+      }
+    ],
+    "version" : "diffusers\/stable-diffusion-xl-base-1.0",
+    "modelParameters" : [
+
+    ],
+    "author" : "Please refer to the Model Card available at huggingface.co\/diffusers\/stable-diffusion-xl-base-1.0",
+    "specificationVersion" : 7,
+    "storagePrecision" : "Float16",
+    "license" : "OpenRAIL (https:\/\/huggingface.co\/spaces\/CompVis\/stable-diffusion-license)",
+    "mlProgramOperationTypeHistogram" : {
+      "Ios16.cast" : 3,
+      "Ios16.mul" : 32,
+      "Ios16.layerNorm" : 65,
+      "Stack" : 1,
+      "Transpose" : 160,
+      "Ios16.linear" : 193,
+      "Ios16.add" : 97,
+      "Ios16.matmul" : 64,
+      "Ios16.gelu" : 32,
+      "Ios16.softmax" : 32,
+      "Ios16.gatherNd" : 1,
+      "Ios16.gather" : 1,
+      "Ios16.reshape" : 320,
+      "Ios16.reduceArgmax" : 1
+    },
+    "computePrecision" : "Mixed (Float16, Float32, Int32)",
+    "isUpdatable" : "0",
+    "availability" : {
+      "macOS" : "13.0",
+      "tvOS" : "16.0",
+      "watchOS" : "9.0",
+      "iOS" : "16.0",
+      "macCatalyst" : "16.0"
+    },
+    "modelType" : {
+      "name" : "MLModelType_mlProgram"
+    },
+    "inputSchema" : [
+      {
+        "hasShapeFlexibility" : "0",
+        "isOptional" : "0",
+        "dataType" : "Float32",
+        "formattedType" : "MultiArray (Float32 1 × 77)",
+        "shortDescription" : "The token ids that represent the input text",
+        "shape" : "[1, 77]",
+        "name" : "input_ids",
+        "type" : "MultiArray"
+      }
+    ],
+    "userDefinedMetadata" : {
+      "com.github.apple.coremltools.version" : "7.0b1",
+      "com.github.apple.coremltools.source" : "torch==2.0.1+cu117"
+    },
+    "generatedClassName" : "Stable_Diffusion_version_diffusers_stable_diffusion_xl_base_1_0_text_encoder_2",
+    "method" : "predict"
+  }
+]

coreml-stable-diffusion-xl-base_mbp_4_50_palettized/compiled/TextEncoder2.mlmodelc/weights/weight.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8bd1fc0bcce11cff685648387b0060e0b6ecfce6c34e580e1ae904cae5903363
+size 1389367424

coreml-stable-diffusion-xl-base_mbp_4_50_palettized/compiled/Unet.mlmodelc/analytics/coremldata.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:48f726f7f29091a23eb434bd4febf1f83733c95db26b912dfa671cdccad5d874
+size 243

coreml-stable-diffusion-xl-base_mbp_4_50_palettized/compiled/Unet.mlmodelc/coremldata.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8aea3887ffdc8e059925f3981259e1cd3227b827e5f91edff613c73ac0ea16f6
+size 1338

coreml-stable-diffusion-xl-base_mbp_4_50_palettized/compiled/Unet.mlmodelc/metadata.json

@@ -0,0 +1,124 @@
+[
+  {
+    "shortDescription" : "Stable Diffusion generates images conditioned on text or other images as input through the diffusion process. Please refer to https:\/\/arxiv.org\/abs\/2112.10752 for details.",
+    "metadataOutputVersion" : "3.0",
+    "outputSchema" : [
+      {
+        "hasShapeFlexibility" : "0",
+        "isOptional" : "0",
+        "dataType" : "Float32",
+        "formattedType" : "MultiArray (Float32)",
+        "shortDescription" : "Same shape and dtype as the `sample` input. The predicted noise to facilitate the reverse diffusion (denoising) process",
+        "shape" : "[]",
+        "name" : "noise_pred",
+        "type" : "MultiArray"
+      }
+    ],
+    "version" : "diffusers\/stable-diffusion-xl-base-1.0",
+    "modelParameters" : [
+
+    ],
+    "author" : "Please refer to the Model Card available at huggingface.co\/diffusers\/stable-diffusion-xl-base-1.0",
+    "specificationVersion" : 7,
+    "storagePrecision" : "Mixed (Float16, Palettized (1 bits), Palettized (2 bits), Palettized (4 bits), Palettized (6 bits), Palettized (8 bits))",
+    "license" : "OpenRAIL (https:\/\/huggingface.co\/spaces\/CompVis\/stable-diffusion-license)",
+    "mlProgramOperationTypeHistogram" : {
+      "UpsampleNearestNeighbor" : 2,
+      "Ios16.reduceMean" : 512,
+      "Ios16.sin" : 2,
+      "Ios16.softmax" : 140,
+      "Split" : 70,
+      "Ios16.add" : 722,
+      "Concat" : 14,
+      "Ios16.realDiv" : 46,
+      "Ios16.square" : 46,
+      "ExpandDims" : 6,
+      "Ios16.sub" : 256,
+      "Ios16.cast" : 1,
+      "Ios16.conv" : 794,
+      "Ios16.constexprLutToDense" : 775,
+      "Ios16.gelu" : 70,
+      "Ios16.matmul" : 280,
+      "Ios16.batchNorm" : 46,
+      "Ios16.reshape" : 675,
+      "Ios16.rsqrt" : 210,
+      "Ios16.silu" : 38,
+      "Ios16.sqrt" : 46,
+      "Ios16.mul" : 842,
+      "Ios16.cos" : 2,
+      "SliceByIndex" : 4
+    },
+    "computePrecision" : "Mixed (Float16, Float32, Int32)",
+    "isUpdatable" : "0",
+    "availability" : {
+      "macOS" : "13.0",
+      "tvOS" : "16.0",
+      "visionOS" : "1.0",
+      "watchOS" : "9.0",
+      "iOS" : "16.0",
+      "macCatalyst" : "16.0"
+    },
+    "modelType" : {
+      "name" : "MLModelType_mlProgram"
+    },
+    "inputSchema" : [
+      {
+        "hasShapeFlexibility" : "0",
+        "isOptional" : "0",
+        "dataType" : "Float16",
+        "formattedType" : "MultiArray (Float16 2 × 4 × 128 × 128)",
+        "shortDescription" : "The low resolution latent feature maps being denoised through reverse diffusion",
+        "shape" : "[2, 4, 128, 128]",
+        "name" : "sample",
+        "type" : "MultiArray"
+      },
+      {
+        "hasShapeFlexibility" : "0",
+        "isOptional" : "0",
+        "dataType" : "Float16",
+        "formattedType" : "MultiArray (Float16 2)",
+        "shortDescription" : "A value emitted by the associated scheduler object to condition the model on a given noise schedule",
+        "shape" : "[2]",
+        "name" : "timestep",
+        "type" : "MultiArray"
+      },
+      {
+        "hasShapeFlexibility" : "0",
+        "isOptional" : "0",
+        "dataType" : "Float16",
+        "formattedType" : "MultiArray (Float16 2 × 2048 × 1 × 77)",
+        "shortDescription" : "Output embeddings from the associated text_encoder model to condition to generated image on text. A maximum of 77 tokens (~40 words) are allowed. Longer text is truncated. Shorter text does not reduce computation.",
+        "shape" : "[2, 2048, 1, 77]",
+        "name" : "encoder_hidden_states",
+        "type" : "MultiArray"
+      },
+      {
+        "hasShapeFlexibility" : "0",
+        "isOptional" : "0",
+        "dataType" : "Float16",
+        "formattedType" : "MultiArray (Float16 12)",
+        "shortDescription" : "",
+        "shape" : "[12]",
+        "name" : "time_ids",
+        "type" : "MultiArray"
+      },
+      {
+        "hasShapeFlexibility" : "0",
+        "isOptional" : "0",
+        "dataType" : "Float16",
+        "formattedType" : "MultiArray (Float16 2 × 1280)",
+        "shortDescription" : "",
+        "shape" : "[2, 1280]",
+        "name" : "text_embeds",
+        "type" : "MultiArray"
+      }
+    ],
+    "userDefinedMetadata" : {
+      "com.github.apple.coremltools.version" : "7.0b1",
+      "com.github.apple.coremltools.source" : "torch==2.0.1+cu117",
+      "com.github.apple.ml-stable-diffusion.version" : "1.0.0"
+    },
+    "generatedClassName" : "recipe_4_50_bit_mixedpalette",
+    "method" : "predict"
+  }
+]

coreml-stable-diffusion-xl-base_mbp_4_50_palettized/compiled/Unet.mlmodelc/weights/weight.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d9219e9fcaa60ff76a31ea84658b485264b50f66147c058e18a7cfd456bb0f3a
+size 1450049728

coreml-stable-diffusion-xl-base_mbp_4_50_palettized/compiled/VAEDecoder.mlmodelc/analytics/coremldata.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bac2854504d7f6bfdc5645982dd965cca9cc8c12b9fdd2493cf50cd583684cc2
+size 207

coreml-stable-diffusion-xl-base_mbp_4_50_palettized/compiled/VAEDecoder.mlmodelc/coremldata.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:766fce2585587fe93a1d32e09bc4d63ad45335ea62239f11b193e481b5888258
+size 773

coreml-stable-diffusion-xl-base_mbp_4_50_palettized/compiled/VAEDecoder.mlmodelc/metadata.json

@@ -0,0 +1,74 @@
+[
+  {
+    "shortDescription" : "Stable Diffusion generates images conditioned on text and\/or other images as input through the diffusion process. Please refer to https:\/\/arxiv.org\/abs\/2112.10752 for details.",
+    "metadataOutputVersion" : "3.0",
+    "outputSchema" : [
+      {
+        "hasShapeFlexibility" : "0",
+        "isOptional" : "0",
+        "dataType" : "Float32",
+        "formattedType" : "MultiArray (Float32)",
+        "shortDescription" : "Generated image normalized to range [-1, 1]",
+        "shape" : "[]",
+        "name" : "image",
+        "type" : "MultiArray"
+      }
+    ],
+    "version" : "diffusers\/stable-diffusion-xl-base-1.0",
+    "modelParameters" : [
+
+    ],
+    "author" : "Please refer to the Model Card available at huggingface.co\/diffusers\/stable-diffusion-xl-base-1.0",
+    "specificationVersion" : 7,
+    "storagePrecision" : "Float32",
+    "license" : "OpenRAIL (https:\/\/huggingface.co\/spaces\/CompVis\/stable-diffusion-license)",
+    "mlProgramOperationTypeHistogram" : {
+      "Ios16.mul" : 2,
+      "Ios16.sqrt" : 30,
+      "Ios16.sub" : 30,
+      "Transpose" : 6,
+      "UpsampleNearestNeighbor" : 3,
+      "Ios16.conv" : 36,
+      "Ios16.add" : 46,
+      "Ios16.linear" : 4,
+      "Ios16.matmul" : 2,
+      "Ios16.realDiv" : 30,
+      "Ios16.reduceMean" : 60,
+      "Ios16.softmax" : 1,
+      "Ios16.batchNorm" : 29,
+      "Ios16.square" : 30,
+      "Ios16.reshape" : 65,
+      "Ios16.silu" : 29
+    },
+    "computePrecision" : "Mixed (Float32, Int32)",
+    "isUpdatable" : "0",
+    "availability" : {
+      "macOS" : "13.0",
+      "tvOS" : "16.0",
+      "watchOS" : "9.0",
+      "iOS" : "16.0",
+      "macCatalyst" : "16.0"
+    },
+    "modelType" : {
+      "name" : "MLModelType_mlProgram"
+    },
+    "inputSchema" : [
+      {
+        "hasShapeFlexibility" : "0",
+        "isOptional" : "0",
+        "dataType" : "Float32",
+        "formattedType" : "MultiArray (Float32 1 × 4 × 128 × 128)",
+        "shortDescription" : "The denoised latent embeddings from the unet model after the last step of reverse diffusion",
+        "shape" : "[1, 4, 128, 128]",
+        "name" : "z",
+        "type" : "MultiArray"
+      }
+    ],
+    "userDefinedMetadata" : {
+      "com.github.apple.coremltools.version" : "7.0b1",
+      "com.github.apple.coremltools.source" : "torch==2.0.1"
+    },
+    "generatedClassName" : "Stable_Diffusion_version_diffusers_stable_diffusion_xl_base_1_0_vae_decoder",
+    "method" : "predict"
+  }
+]

coreml-stable-diffusion-xl-base_mbp_4_50_palettized/compiled/VAEDecoder.mlmodelc/weights/weight.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7ade814d6037fb5ba892963be1596c8e37852f96c399101401831f8c07e64bd2
+size 197977216

coreml-stable-diffusion-xl-base_mbp_4_50_palettized/compiled/VAEEncoder.mlmodelc/analytics/coremldata.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0e44297c358d64101602d9abfe4d6c9fb96ddb3b120f84fbb74001aa4312cf93
+size 207

coreml-stable-diffusion-xl-base_mbp_4_50_palettized/compiled/VAEEncoder.mlmodelc/coremldata.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7269e365034d061c3ad035d9a5b7c38864d7db71bee6fb7545c97a4942a865f1
+size 783

coreml-stable-diffusion-xl-base_mbp_4_50_palettized/compiled/VAEEncoder.mlmodelc/metadata.json

@@ -0,0 +1,75 @@
+[
+  {
+    "shortDescription" : "Stable Diffusion generates images conditioned on text and\/or other images as input through the diffusion process. Please refer to https:\/\/arxiv.org\/abs\/2112.10752 for details.",
+    "metadataOutputVersion" : "3.0",
+    "outputSchema" : [
+      {
+        "hasShapeFlexibility" : "0",
+        "isOptional" : "0",
+        "dataType" : "Float32",
+        "formattedType" : "MultiArray (Float32)",
+        "shortDescription" : "The latent embeddings from the unet model from the input image.",
+        "shape" : "[]",
+        "name" : "latent",
+        "type" : "MultiArray"
+      }
+    ],
+    "version" : "diffusers\/stable-diffusion-xl-base-1.0",
+    "modelParameters" : [
+
+    ],
+    "author" : "Please refer to the Model Card available at huggingface.co\/diffusers\/stable-diffusion-xl-base-1.0",
+    "specificationVersion" : 7,
+    "storagePrecision" : "Float16",
+    "license" : "OpenRAIL (https:\/\/huggingface.co\/spaces\/CompVis\/stable-diffusion-license)",
+    "mlProgramOperationTypeHistogram" : {
+      "Pad" : 3,
+      "Ios16.cast" : 1,
+      "Ios16.mul" : 2,
+      "Ios16.sqrt" : 22,
+      "Ios16.sub" : 22,
+      "Transpose" : 6,
+      "Ios16.conv" : 28,
+      "Ios16.add" : 34,
+      "Ios16.linear" : 4,
+      "Ios16.matmul" : 2,
+      "Ios16.realDiv" : 22,
+      "Ios16.reduceMean" : 44,
+      "Ios16.softmax" : 1,
+      "Ios16.batchNorm" : 21,
+      "Ios16.square" : 22,
+      "Ios16.reshape" : 49,
+      "Ios16.silu" : 21
+    },
+    "computePrecision" : "Mixed (Float16, Float32, Int32)",
+    "isUpdatable" : "0",
+    "availability" : {
+      "macOS" : "13.0",
+      "tvOS" : "16.0",
+      "watchOS" : "9.0",
+      "iOS" : "16.0",
+      "macCatalyst" : "16.0"
+    },
+    "modelType" : {
+      "name" : "MLModelType_mlProgram"
+    },
+    "inputSchema" : [
+      {
+        "hasShapeFlexibility" : "0",
+        "isOptional" : "0",
+        "dataType" : "Float16",
+        "formattedType" : "MultiArray (Float16 1 × 3 × 1024 × 1024)",
+        "shortDescription" : "The input image to base the initial latents on normalized to range [-1, 1]",
+        "shape" : "[1, 3, 1024, 1024]",
+        "name" : "z",
+        "type" : "MultiArray"
+      }
+    ],
+    "userDefinedMetadata" : {
+      "com.github.apple.coremltools.version" : "7.0b1",
+      "com.github.apple.coremltools.source" : "torch==2.0.1+cu117"
+    },
+    "generatedClassName" : "Stable_Diffusion_version_diffusers_stable_diffusion_xl_base_1_0_vae_encoder",
+    "method" : "predict"
+  }
+]

coreml-stable-diffusion-xl-base_mbp_4_50_palettized/compiled/VAEEncoder.mlmodelc/weights/weight.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:329f708df0bae1990a1886007b5ae56cfd9a44e7091e8f822db907a9fc411858
+size 68338112

coreml-stable-diffusion-xl-base_mbp_4_50_palettized/packages/Stable_Diffusion_version_stabilityai_stable-diffusion-xl-base-1.0_text_encoder.mlpackage/Data/com.apple.CoreML/FeatureDescriptions.json

@@ -0,0 +1,18 @@
+{
+  "Outputs" : {
+    "pooled_outputs" : {
+      "MLFeatureShortDescription" : "The version of the `last_hidden_state` output after pooling"
+    },
+    "hidden_embeds" : {
+      "MLFeatureShortDescription" : "Hidden states after the encoder layers"
+    }
+  },
+  "Inputs" : {
+    "input_ids" : {
+      "MLFeatureShortDescription" : "The token ids that represent the input text"
+    }
+  },
+  "TrainingInputs" : {
+
+  }
+}

coreml-stable-diffusion-xl-base_mbp_4_50_palettized/packages/Stable_Diffusion_version_stabilityai_stable-diffusion-xl-base-1.0_text_encoder.mlpackage/Data/com.apple.CoreML/Metadata.json

@@ -0,0 +1,10 @@
+{
+  "MLModelVersionStringKey" : "stabilityai\/stable-diffusion-xl-base-1.0",
+  "MLModelDescriptionKey" : "Stable Diffusion generates images conditioned on text and\/or other images as input through the diffusion process. Please refer to https:\/\/arxiv.org\/abs\/2112.10752 for details.",
+  "MLModelCreatorDefinedKey" : {
+    "com.github.apple.coremltools.version" : "7.0b1",
+    "com.github.apple.coremltools.source" : "torch==2.0.1+cu117"
+  },
+  "MLModelAuthorKey" : "Please refer to the Model Card available at huggingface.co\/stabilityai\/stable-diffusion-xl-base-1.0",
+  "MLModelLicenseKey" : "OpenRAIL++-M (https:\/\/huggingface.co\/stabilityai\/stable-diffusion-xl-base-1.0\/blob\/main\/LICENSE.md)"
+}

coreml-stable-diffusion-xl-base_mbp_4_50_palettized/packages/Stable_Diffusion_version_stabilityai_stable-diffusion-xl-base-1.0_text_encoder.mlpackage/Data/com.apple.CoreML/model.mlmodel

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d7fc64f553e0f6be677899106f40ed3bf85eceb3d0b55f90022252cb8dff1d4a
+size 161160

coreml-stable-diffusion-xl-base_mbp_4_50_palettized/packages/Stable_Diffusion_version_stabilityai_stable-diffusion-xl-base-1.0_text_encoder.mlpackage/Data/com.apple.CoreML/weights/weight.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a86533724aadf50c8f5539592d440887a484f60002d7967505c69c7faf4d7797
+size 246145536

coreml-stable-diffusion-xl-base_mbp_4_50_palettized/packages/Stable_Diffusion_version_stabilityai_stable-diffusion-xl-base-1.0_text_encoder.mlpackage/Manifest.json

@@ -0,0 +1,30 @@
+{
+    "fileFormatVersion": "1.0.0",
+    "itemInfoEntries": {
+        "2E437C75-2D29-4906-9BBF-06B4F60AF608": {
+            "author": "com.apple.CoreML",
+            "description": "External Metadata Overlay",
+            "name": "Metadata.json",
+            "path": "com.apple.CoreML/Metadata.json"
+        },
+        "3d765fc2-d834-4477-9cf7-818901242816": {
+            "author": "com.apple.CoreML",
+            "description": "CoreML Model Weights",
+            "name": "weights",
+            "path": "com.apple.CoreML/weights"
+        },
+        "58CF0F86-0A4C-4478-B878-FB21979F2094": {
+            "author": "com.apple.CoreML",
+            "description": "External FeatureDescription Overlay",
+            "name": "FeatureDescriptions.json",
+            "path": "com.apple.CoreML/FeatureDescriptions.json"
+        },
+        "5fc727d7-449d-4885-8c48-b3b711dfed7b": {
+            "author": "com.apple.CoreML",
+            "description": "CoreML Model Specification",
+            "name": "model.mlmodel",
+            "path": "com.apple.CoreML/model.mlmodel"
+        }
+    },
+    "rootModelIdentifier": "5fc727d7-449d-4885-8c48-b3b711dfed7b"
+}

coreml-stable-diffusion-xl-base_mbp_4_50_palettized/packages/Stable_Diffusion_version_stabilityai_stable-diffusion-xl-base-1.0_text_encoder_2.mlpackage/Data/com.apple.CoreML/FeatureDescriptions.json

@@ -0,0 +1,18 @@
+{
+  "Outputs" : {
+    "pooled_outputs" : {
+      "MLFeatureShortDescription" : "The version of the `last_hidden_state` output after pooling"
+    },
+    "hidden_embeds" : {
+      "MLFeatureShortDescription" : "Hidden states after the encoder layers"
+    }
+  },
+  "Inputs" : {
+    "input_ids" : {
+      "MLFeatureShortDescription" : "The token ids that represent the input text"
+    }
+  },
+  "TrainingInputs" : {
+
+  }
+}

coreml-stable-diffusion-xl-base_mbp_4_50_palettized/packages/Stable_Diffusion_version_stabilityai_stable-diffusion-xl-base-1.0_text_encoder_2.mlpackage/Data/com.apple.CoreML/Metadata.json

@@ -0,0 +1,10 @@
+{
+  "MLModelVersionStringKey" : "stabilityai\/stable-diffusion-xl-base-1.0",
+  "MLModelDescriptionKey" : "Stable Diffusion generates images conditioned on text and\/or other images as input through the diffusion process. Please refer to https:\/\/arxiv.org\/abs\/2112.10752 for details.",
+  "MLModelCreatorDefinedKey" : {
+    "com.github.apple.coremltools.version" : "7.0b1",
+    "com.github.apple.coremltools.source" : "torch==2.0.1+cu117"
+  },
+  "MLModelAuthorKey" : "Please refer to the Model Card available at huggingface.co\/stabilityai\/stable-diffusion-xl-base-1.0",
+  "MLModelLicenseKey" : "OpenRAIL++-M (https:\/\/huggingface.co\/stabilityai\/stable-diffusion-xl-base-1.0\/blob\/main\/LICENSE.md)"
+}

coreml-stable-diffusion-xl-base_mbp_4_50_palettized/packages/Stable_Diffusion_version_stabilityai_stable-diffusion-xl-base-1.0_text_encoder_2.mlpackage/Data/com.apple.CoreML/model.mlmodel

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1cb832076a7e67ff890e274281c5bf83c3be6743b8214d0c6df95b88fce6711f
+size 418221

coreml-stable-diffusion-xl-base_mbp_4_50_palettized/packages/Stable_Diffusion_version_stabilityai_stable-diffusion-xl-base-1.0_text_encoder_2.mlpackage/Data/com.apple.CoreML/weights/weight.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8bd1fc0bcce11cff685648387b0060e0b6ecfce6c34e580e1ae904cae5903363
+size 1389367424

coreml-stable-diffusion-xl-base_mbp_4_50_palettized/packages/Stable_Diffusion_version_stabilityai_stable-diffusion-xl-base-1.0_text_encoder_2.mlpackage/Manifest.json

@@ -0,0 +1,30 @@
+{
+    "fileFormatVersion": "1.0.0",
+    "itemInfoEntries": {
+        "1003fcc7-51c2-44b1-bfd2-71eaffb11598": {
+            "author": "com.apple.CoreML",
+            "description": "CoreML Model Weights",
+            "name": "weights",
+            "path": "com.apple.CoreML/weights"
+        },
+        "5bff403e-2a80-4472-b7fe-4c3438e76e85": {
+            "author": "com.apple.CoreML",
+            "description": "CoreML Model Specification",
+            "name": "model.mlmodel",
+            "path": "com.apple.CoreML/model.mlmodel"
+        },
+        "66E76242-93D0-4D4D-8D64-697552EC6C38": {
+            "author": "com.apple.CoreML",
+            "description": "External FeatureDescription Overlay",
+            "name": "FeatureDescriptions.json",
+            "path": "com.apple.CoreML/FeatureDescriptions.json"
+        },
+        "8DF8F909-251F-4BFD-BFA3-F72860B2F27D": {
+            "author": "com.apple.CoreML",
+            "description": "External Metadata Overlay",
+            "name": "Metadata.json",
+            "path": "com.apple.CoreML/Metadata.json"
+        }
+    },
+    "rootModelIdentifier": "5bff403e-2a80-4472-b7fe-4c3438e76e85"
+}

coreml-stable-diffusion-xl-base_mbp_4_50_palettized/packages/Stable_Diffusion_version_stabilityai_stable-diffusion-xl-base-1.0_unet.mlpackage/Data/com.apple.CoreML/model.mlmodel

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d32aa25518ec6d20563c61f0cc8f0b9ecca6d31665e4a5345a4b44d4c7706004
+size 2135914

coreml-stable-diffusion-xl-base_mbp_4_50_palettized/packages/Stable_Diffusion_version_stabilityai_stable-diffusion-xl-base-1.0_unet.mlpackage/Data/com.apple.CoreML/weights/weight.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d9219e9fcaa60ff76a31ea84658b485264b50f66147c058e18a7cfd456bb0f3a
+size 1450049728

coreml-stable-diffusion-xl-base_mbp_4_50_palettized/packages/Stable_Diffusion_version_stabilityai_stable-diffusion-xl-base-1.0_unet.mlpackage/Manifest.json

@@ -0,0 +1,18 @@
+{
+    "fileFormatVersion": "1.0.0",
+    "itemInfoEntries": {
+        "4aab2a70-8353-41d1-81e0-1761fcebde0f": {
+            "author": "com.apple.CoreML",
+            "description": "CoreML Model Weights",
+            "name": "weights",
+            "path": "com.apple.CoreML/weights"
+        },
+        "58e673b4-eef0-4992-a74f-f42d8ec45a72": {
+            "author": "com.apple.CoreML",
+            "description": "CoreML Model Specification",
+            "name": "model.mlmodel",
+            "path": "com.apple.CoreML/model.mlmodel"
+        }
+    },
+    "rootModelIdentifier": "58e673b4-eef0-4992-a74f-f42d8ec45a72"
+}

coreml-stable-diffusion-xl-base_mbp_4_50_palettized/packages/Stable_Diffusion_version_stabilityai_stable-diffusion-xl-base-1.0_vae_decoder.mlpackage/Data/com.apple.CoreML/FeatureDescriptions.json

@@ -0,0 +1,15 @@
+{
+  "Outputs" : {
+    "image" : {
+      "MLFeatureShortDescription" : "Generated image normalized to range [-1, 1]"
+    }
+  },
+  "Inputs" : {
+    "z" : {
+      "MLFeatureShortDescription" : "The denoised latent embeddings from the unet model after the last step of reverse diffusion"
+    }
+  },
+  "TrainingInputs" : {
+
+  }
+}

coreml-stable-diffusion-xl-base_mbp_4_50_palettized/packages/Stable_Diffusion_version_stabilityai_stable-diffusion-xl-base-1.0_vae_decoder.mlpackage/Data/com.apple.CoreML/Metadata.json

@@ -0,0 +1,10 @@
+{
+  "MLModelVersionStringKey" : "stabilityai\/stable-diffusion-xl-base-1.0",
+  "MLModelDescriptionKey" : "Stable Diffusion generates images conditioned on text and\/or other images as input through the diffusion process. Please refer to https:\/\/arxiv.org\/abs\/2112.10752 for details.",
+  "MLModelCreatorDefinedKey" : {
+    "com.github.apple.coremltools.version" : "7.0b1",
+    "com.github.apple.coremltools.source" : "torch==2.0.1"
+  },
+  "MLModelAuthorKey" : "Please refer to the Model Card available at huggingface.co\/stabilityai\/stable-diffusion-xl-base-1.0",
+  "MLModelLicenseKey" : "OpenRAIL++-M (https:\/\/huggingface.co\/stabilityai\/stable-diffusion-xl-base-1.0\/blob\/main\/LICENSE.md)"
+}

coreml-stable-diffusion-xl-base_mbp_4_50_palettized/packages/Stable_Diffusion_version_stabilityai_stable-diffusion-xl-base-1.0_vae_decoder.mlpackage/Data/com.apple.CoreML/model.mlmodel

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fb66e41edf6678acc5df408ffb294157df24996f810f649d953d2997df2f40cc
+size 143700

coreml-stable-diffusion-xl-base_mbp_4_50_palettized/packages/Stable_Diffusion_version_stabilityai_stable-diffusion-xl-base-1.0_vae_decoder.mlpackage/Data/com.apple.CoreML/weights/weight.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7ade814d6037fb5ba892963be1596c8e37852f96c399101401831f8c07e64bd2
+size 197977216

coreml-stable-diffusion-xl-base_mbp_4_50_palettized/packages/Stable_Diffusion_version_stabilityai_stable-diffusion-xl-base-1.0_vae_decoder.mlpackage/Manifest.json

@@ -0,0 +1,30 @@
+{
+    "fileFormatVersion": "1.0.0",
+    "itemInfoEntries": {
+        "0C4B4A84-E58C-4B55-9232-7AD156C36DF1": {
+            "author": "com.apple.CoreML",
+            "description": "CoreML Model Specification",
+            "name": "model.mlmodel",
+            "path": "com.apple.CoreML/model.mlmodel"
+        },
+        "63939630-D849-41F2-B7BB-9439F4215E25": {
+            "author": "com.apple.CoreML",
+            "description": "External FeatureDescription Overlay",
+            "name": "FeatureDescriptions.json",
+            "path": "com.apple.CoreML/FeatureDescriptions.json"
+        },
+        "920AF9A5-7F1B-4655-9367-780D11D0CACC": {
+            "author": "com.apple.CoreML",
+            "description": "External Metadata Overlay",
+            "name": "Metadata.json",
+            "path": "com.apple.CoreML/Metadata.json"
+        },
+        "FF129861-DA4E-4A61-A51B-4E991CC3DAEA": {
+            "author": "com.apple.CoreML",
+            "description": "CoreML Model Weights",
+            "name": "weights",
+            "path": "com.apple.CoreML/weights"
+        }
+    },
+    "rootModelIdentifier": "0C4B4A84-E58C-4B55-9232-7AD156C36DF1"
+}

coreml-stable-diffusion-xl-base_mbp_4_50_palettized/packages/Stable_Diffusion_version_stabilityai_stable-diffusion-xl-base-1.0_vae_encoder.mlpackage/Data/com.apple.CoreML/FeatureDescriptions.json

@@ -0,0 +1,15 @@
+{
+  "Outputs" : {
+    "latent" : {
+      "MLFeatureShortDescription" : "The latent embeddings from the unet model from the input image."
+    }
+  },
+  "Inputs" : {
+    "z" : {
+      "MLFeatureShortDescription" : "The input image to base the initial latents on normalized to range [-1, 1]"
+    }
+  },
+  "TrainingInputs" : {
+
+  }
+}

coreml-stable-diffusion-xl-base_mbp_4_50_palettized/packages/Stable_Diffusion_version_stabilityai_stable-diffusion-xl-base-1.0_vae_encoder.mlpackage/Data/com.apple.CoreML/Metadata.json

@@ -0,0 +1,10 @@
+{
+  "MLModelVersionStringKey" : "stabilityai\/stable-diffusion-xl-base-1.0",
+  "MLModelDescriptionKey" : "Stable Diffusion generates images conditioned on text and\/or other images as input through the diffusion process. Please refer to https:\/\/arxiv.org\/abs\/2112.10752 for details.",
+  "MLModelCreatorDefinedKey" : {
+    "com.github.apple.coremltools.version" : "7.0b1",
+    "com.github.apple.coremltools.source" : "torch==2.0.1+cu117"
+  },
+  "MLModelAuthorKey" : "Please refer to the Model Card available at huggingface.co\/stabilityai\/stable-diffusion-xl-base-1.0",
+  "MLModelLicenseKey" : "OpenRAIL++-M (https:\/\/huggingface.co\/stabilityai\/stable-diffusion-xl-base-1.0\/blob\/main\/LICENSE.md)"
+}

coreml-stable-diffusion-xl-base_mbp_4_50_palettized/packages/Stable_Diffusion_version_stabilityai_stable-diffusion-xl-base-1.0_vae_encoder.mlpackage/Data/com.apple.CoreML/model.mlmodel

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ee32c2ed8e0d8dc35e513bd328fda885af82ec3d252f50037cc70fc5d966860d
+size 119397