Add files using upload-large-folder tool

LICENSE.md

@@ -0,0 +1,174 @@
+# Fair AI Public License 1.0-SD
+
+Published by the Freedom of Development Project.
+
+This “SD variant” license contains a Prohibited Uses section designed to be
+compatible with Stable Diffusion’s license. Because of that section, this is
+not a free software license. Unless you are releasing a derivative of a Stable
+Diffusion model, it is not recommended that you use this license.
+
+This license comes with special requirements if you intend to allow users to
+access this software over a network. See Notices for more information.
+
+## Purpose
+
+This license gives everyone as much permission to work with this software as
+possible, while protecting contributors from liability, protecting the freedom
+of end users, and reducing harm.
+
+## Definitions
+
+In this license, “model” refers to machine learning model weights, biases,
+parameters, optimizer states, and any byproducts of a training or pretraining
+process, whether in the form of checkpoints or any other form.
+
+The term “derived model” refers to any model based on this model.
+
+The term “software” also refers to any model along with documentation or other
+resources provided with the software.
+
+The term “source code” refers to the preferred form of making modifications to
+software. It also includes any models, if applicable, but it does not include
+any datasets used to train a model.
+
+To “modify” also means to perform any training on a model or to combine a model
+with another model.
+
+## Acceptance
+
+In order to receive this license, you must agree to its rules. The rules of this
+license are both obligations under that agreement and conditions to your
+license. You must not do anything with this software that triggers a rule that
+you cannot or will not follow. If you do not agree, then you cannot use this
+software in any way.
+
+## Copyright
+
+Each contributor licenses you to do everything with this software that would
+otherwise infringe that contributor’s copyright in it.
+
+## Freedom
+
+Neither this software nor any work that is combined with this software will be
+considered a technological protection measure under the WIPO Copyright Treaty
+or any similar law. Reverse engineering of this software and of any work that
+is combined with this software is always allowed.
+
+## Notices
+
+You must ensure that everyone who gets a copy of any part of this software from
+you, with or without changes, also gets the text of this license along with the
+corresponding source code.
+
+If you modify this software and allow users to interact with it through a
+computer network, you must ensure they have a reasonable way to receive the
+corresponding source code from you, whether that is via a download link or a
+prominent written offer. As a special case, if you are only allowing users to
+interact with a derived model, then you may choose to provide a download link
+or written offer only for the derived model.
+
+This software, all source code, and all modifications must be provided under
+this license or another license that allows everything this license allows.
+Note that this does not give you permission to change the license for this
+software.
+
+## Excuse
+
+If anyone notifies you in writing that you have not complied with Notices, you
+can keep your license by taking all practical steps to comply within 30 days
+after the notice. If you do not do so, your license ends immediately.
+
+## Output
+
+The output of this software is not covered by this license, and no contributor
+claims any rights to it.
+
+## Patent
+
+Each contributor licenses you to do everything with this software that would
+otherwise infringe any patent claims they can license or become able to
+license.
+
+## Reliability
+
+No contributor can revoke this license.
+
+## Alternatives
+
+You can also use any non-model parts of this software under the terms of the
+GNU AGPL 3.0, or any later version of that license. If you do, No Harm and No
+Liability still apply.
+
+## Revisions
+
+The Freedom of Development Project may publish revised or new versions of the
+Fair AI Public License. Those new versions will be similar in spirit to this
+license.
+
+Unless a contributor specifies otherwise, you have the option of following the
+terms of any later version of this license. Your choice to follow a later
+version of the license will not impose additional obligations on any
+contributor. Even if you do choose to follow a later version, the restrictions
+of Prohibited Uses will still apply.
+
+## Survival
+
+The provisions of No Harm and No Liability survive the end of your license.
+
+## No Harm
+
+You agree that no contributor’s conduct in the creation of this software has
+caused you any harm. As far as the law allows, you give up your right to pursue
+any kind of legal claim against any contributor for actions related the
+creation of this software, even if those actions broke a previous agreement.
+
+Additionally, you agree not to use this model for harmful purposes, as listed
+in Prohibited Uses. These restrictions do not apply to non-model parts of this
+software.
+
+## No Liability
+
+As far as the law allows, this software comes as is, without any warranty or
+condition, and no contributor will be liable to anyone for any damages related
+to this software or this license, under any kind of legal claim.
+
+## Prohibited Uses
+
+You may not use this model or any derived model for the following:
+
+- In any way that violates any applicable national, federal, state, local or
+  international law or regulation;
+- For the purpose of exploiting, harming or attempting to exploit or harm
+  minors in any way;
+- To generate or disseminate verifiably false information and/or content with
+  the purpose of harming others;
+- To generate or disseminate personal identifiable information that can be
+  used to harm an individual;
+- To defame, disparage or otherwise harass others;
+- For fully automated decision making that adversely impacts an individual’s
+  legal rights or otherwise creates or modifies a binding, enforceable
+  obligation;
+- For any use intended to or which has the effect of discriminating against or
+  harming individuals or groups based on online or offline social behavior or
+  known or predicted personal or personality characteristics;
+- To exploit any of the vulnerabilities of a specific group of persons based
+  on their age, social, physical or mental characteristics, in order to
+  materially distort the behavior of a person pertaining to that group in a
+  manner that causes or is likely to cause that person or another person
+  physical or psychological harm;
+- For any use intended to or which has the effect of discriminating against
+  individuals or groups based on legally protected characteristics or
+  categories;
+- To provide medical advice and medical results interpretation;
+- To generate or disseminate information for the purpose to be used for
+  administration of justice, law enforcement, immigration or asylum processes,
+  such as predicting an individual will commit fraud/crime commitment (e.g. by
+  text profiling, drawing causal relationships between assertions made in
+  documents, indiscriminate and arbitrarily-targeted use).
+
+Copyright © 2023 The Freedom of Development Project
+
+Verbatim copying and distribution of this entire article are permitted
+worldwide, without royalty, in any medium, provided this notice is preserved.
+
+Source: https://freedevproject.org/faipl-1.0-sd/

PROVENANCE.json

@@ -0,0 +1,116 @@
+{
+  "schemaVersion": 1,
+  "model": {
+    "name": "WAI Illustrious SDXL V17",
+    "base": "Illustrious SDXL",
+    "workflow": [
+      "text-to-image",
+      "image-to-image",
+      "face-detail"
+    ]
+  },
+  "officialRelease": {
+    "author": "WAI0731",
+    "civitaiModelId": 827184,
+    "civitaiVersionId": 2883731,
+    "url": "https://civitai.com/models/827184?modelVersionId=2883731"
+  },
+  "source": {
+    "repository": "LyliaEngine/waiIllustriousSDXL_v170",
+    "revision": "5ef4e2da7173a160ad04aebcaa2fdcd6d20ed792",
+    "file": "waiIllustriousSDXL_v170.safetensors",
+    "sizeBytes": 6938040682,
+    "sha256": "f116b0c78ff441467b0cdc8f1936e1ed18ea31e9997c7b132b1b8db533f0bd04",
+    "role": "Pinned checkpoint transport mirror"
+  },
+  "configuration": {
+    "repository": "frankjoshua/waiIllustriousSDXL_v170",
+    "revision": "9303ce49345822823717889e3677b6ffd43fc6a9",
+    "vaeScalingFactor": 0.13025,
+    "unetInputChannels": 4
+  },
+  "license": {
+    "officialRelease": "Fair AI Public License 1.0-SD",
+    "includedLicense": "Fair AI Public License 1.0-SD",
+    "licenseURL": "https://freedevproject.org/faipl-1.0-sd/",
+    "note": "The official Civitai release controls attribution and licensing for this conversion. Inconsistent metadata on a transport mirror does not replace that declared license."
+  },
+  "textEncoderRepair": {
+    "reason": "The source checkpoint contains non-finite values only in the unused final CLIP-L transformer layer. SDXL conditioning uses the preceding hidden state, and pooled conditioning comes from TextEncoder2.",
+    "replacementRepository": "OnomaAIResearch/Illustrious-xl-early-release-v0",
+    "replacementRevision": "dca0dac303e6dc4b0c31d8001bc685b89b5d0204",
+    "replacementFile": "text_encoder/model.safetensors",
+    "replacementSizeBytes": 246144152,
+    "replacementSha256": "f8480c7d1b3ee29a5a9922d5c088dfbfc6d5b94c8d7312892eacfe7a880052cc",
+    "affectedLayer": "text_model.encoder.layers.11",
+    "replacedNonFiniteValueCount": 7077888,
+    "scope": "Only non-finite values in the unused final CLIP-L layer were replaced; all conditioning layers used by inference remain from WAI Illustrious SDXL V17."
+  },
+  "conversion": {
+    "appleMLStableDiffusionRevision": "e12202c1f6405b83918b58a5d097cd61e3e1f702",
+    "python": "3.9.6",
+    "coremltools": "8.0",
+    "diffusers": "0.30.2",
+    "torch": "2.4.0",
+    "transformers": "4.44.2",
+    "huggingfaceHub": "0.24.6",
+    "safetensors": "0.4.5",
+    "minimumDeploymentTarget": "iOS 17",
+    "attentionImplementation": "SPLIT_EINSUM",
+    "computeUnits": "CPU_AND_NE conversion; GPU VAE runtime",
+    "resolution": [
+      1024,
+      1024
+    ],
+    "unet": {
+      "quantizationBits": 6,
+      "split": true,
+      "batchSize": 2,
+      "inputShape": [
+        2,
+        4,
+        128,
+        128
+      ],
+      "outputShape": [
+        2,
+        4,
+        128,
+        128
+      ]
+    },
+    "textEncoder": {
+      "precision": "FP16",
+      "source": "WAI Illustrious SDXL V17 with documented unused-layer repair"
+    },
+    "textEncoder2": {
+      "quantizationBits": 8,
+      "source": "WAI Illustrious SDXL V17"
+    },
+    "vae": {
+      "precision": "Float32 Core ML",
+      "source": "WAI Illustrious SDXL V17",
+      "decoderIncluded": true,
+      "encoderIncludedForImageToImageAndFaceDetail": true,
+      "scalingFactor": 0.13025
+    }
+  },
+  "componentWeightBytes": {
+    "TextEncoder.mlmodelc": 246145536,
+    "TextEncoder2.mlmodelc": 758591232,
+    "UnetChunk1.mlmodelc": 968067264,
+    "UnetChunk2.mlmodelc": 959821632,
+    "VAEDecoder.mlmodelc": 197977216,
+    "VAEEncoder.mlmodelc": 136668992
+  },
+  "validation": {
+    "checkpointTensorShapesVerified": true,
+    "allUsedCheckpointTensorsFinite": true,
+    "compiledComponentShapesVerified": true,
+    "allCompiledComponentsBelowOneGiB": true,
+    "fullPipelineGeneration": false,
+    "compiledComponentPredictionsFinite": false,
+    "iPhoneRuntimeValidation": "pending",
+    "note": "macOS execution is not used as a substitute for the final iPhone Core ML runtime test."
+  }
+}

README.md

@@ -0,0 +1,68 @@
+---
+license: other
+license_name: fair-ai-public-license-1.0-sd
+license_link: https://freedevproject.org/faipl-1.0-sd/
+base_model: OnomaAIResearch/Illustrious-xl-early-release-v0
+pipeline_tag: text-to-image
+tags:
+- coreml
+- ios
+- stable-diffusion-xl
+- illustrious
+- anime
+- localmuse
+---
+
+# WAI Illustrious SDXL V17 — Core ML for LocalMuse
+
+A production-oriented Core ML conversion of WAI Illustrious SDXL V17 for
+private, on-device 1024×1024 generation in LocalMuse.
+
+## Runtime profile
+
+- Native, non-distilled Illustrious SDXL at 1024×1024.
+- 15–30 diffusion steps; LocalMuse defaults to 25.
+- DPM-Solver++ Karras, CFG 5–7, batch size one.
+- Recommended positive quality tags:
+  `masterpiece, best quality, amazing quality`.
+- Two 6-bit split UNet components, each below the iOS 1 GiB component limit.
+- Model CLIP-L at FP16 and OpenCLIP-bigG at 8-bit.
+- Checkpoint VAE decoder and encoder for Text to Image, Transform, and
+  LocalMuse Face Detail.
+- Requires an iPhone or iPad in the nominal 8 GiB memory class.
+
+The UNet and second text-encoder palettization reduce model storage. They do
+not turn V17 into a Lightning, LCM, Turbo, or other reduced-step model.
+
+## Source and reproducibility
+
+Official release: [WAI Illustrious SDXL on Civitai](https://civitai.com/models/827184?modelVersionId=2883731),
+model ID `827184`, version ID `2883731`.
+
+Conversion source checkpoint:
+[`LyliaEngine/waiIllustriousSDXL_v170`](https://huggingface.co/LyliaEngine/waiIllustriousSDXL_v170/tree/5ef4e2da7173a160ad04aebcaa2fdcd6d20ed792),
+revision `5ef4e2da7173a160ad04aebcaa2fdcd6d20ed792`, file
+`waiIllustriousSDXL_v170.safetensors`, SHA-256
+`f116b0c78ff441467b0cdc8f1936e1ed18ea31e9997c7b132b1b8db533f0bd04`.
+
+The checkpoint contains non-finite values only in CLIP-L's unused final
+transformer layer. The converter replaces only those values with the matching
+pinned Illustrious base-model layer. SDXL prompt conditioning uses the
+preceding hidden state, while pooled conditioning is produced by the exact WAI
+second encoder. Exact inputs, revisions, hashes, conversion versions,
+component sizes, and validation status are recorded in `PROVENANCE.json`.
+
+## License and attribution
+
+WAI Illustrious SDXL is by WAI0731. The official Civitai release declares the
+[Fair AI Public License 1.0-SD](https://freedevproject.org/faipl-1.0-sd/).
+The complete license text is included in `LICENSE.md`; its use restrictions
+continue to apply to this converted artifact. The Hugging Face source is used
+as a pinned transport mirror and its inconsistent repository metadata does not
+replace the license declared by the official release.
+
+## Validation status
+
+Checkpoint shapes, all tensors used by inference, compiled component shapes,
+and component size limits were verified. The final iPhone runtime test remains
+pending and is intentionally not represented as completed in provenance.

TextEncoder.mlmodelc/weights/weight.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5af12b89dc9ffaa1b95d2ae03529e87fe50912a7ca603ee64dafa70fa8be6058
+size 246145536

TextEncoder2.mlmodelc/analytics/coremldata.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:55690a4bae17f1d00cbd81faa69687cf51e5fe41ecebf41486edd633f6a8e55f
+size 243

TextEncoder2.mlmodelc/weights/weight.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:619c44fce4a0050eb15dbdc1e19ab8351df019f66f81a0bf5ae7b98791f33ea8
+size 758591232

UnetChunk1.mlmodelc/analytics/coremldata.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3f98cd5ea61d5c8930e8dcd0fe4108232eecf59eb856e5ffc0b875439ba1216f
+size 243

UnetChunk1.mlmodelc/coremldata.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:df9fc6af09556e03da4d6a6bf2538be42aab919f23150bdf93939db64ad192bd
+size 669

UnetChunk1.mlmodelc/weights/weight.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cb22b8707a8304342511fc64f024feab2c40c0cd7a23a924a85ab640841a1901
+size 968067264

UnetChunk2.mlmodelc/weights/weight.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d32ec9f25f31482bfae9d92b93ae7fddfd715d05f306253ce78ccb933c3dce6d
+size 959821632

VAEDecoder.mlmodelc/analytics/coremldata.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fc23f3efa60729038c380d54e2326463e5b565454381615d0d6f95cb0f39846a
+size 243

VAEDecoder.mlmodelc/coremldata.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:99d760358450e8bd1904cbb564ed01dcdbc1933562e5a39492d34e8f755e7d86
+size 860

VAEDecoder.mlmodelc/weights/weight.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7631d6b747e772c3da7be0860765f8893776fac46160132cce7d2ba15f2e3948
+size 197977216

VAEEncoder.mlmodelc/metadata.json

@@ -0,0 +1,79 @@
+[
+  {
+    "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 1 × 8 × 128 × 128)",
+        "shortDescription" : "The latent embeddings from the unet model from the input image.",
+        "shape" : "[1, 8, 128, 128]",
+        "name" : "latent",
+        "type" : "MultiArray"
+      }
+    ],
+    "version" : "LyliaEngine\/waiIllustriousSDXL_v170",
+    "modelParameters" : [
+
+    ],
+    "author" : "Please refer to the Model Card available at huggingface.co\/LyliaEngine\/waiIllustriousSDXL_v170",
+    "specificationVersion" : 8,
+    "storagePrecision" : "Float32",
+    "license" : "OpenRAIL++-M (https:\/\/huggingface.co\/stabilityai\/stable-diffusion-xl-base-1.0\/blob\/main\/LICENSE.md)",
+    "mlProgramOperationTypeHistogram" : {
+      "Pad" : 3,
+      "Ios17.mul" : 2,
+      "Ios17.sqrt" : 22,
+      "Ios17.square" : 22,
+      "Ios17.transpose" : 6,
+      "Ios17.sub" : 22,
+      "Ios17.linear" : 4,
+      "Ios17.conv" : 28,
+      "Ios17.matmul" : 2,
+      "Ios17.add" : 34,
+      "Ios17.realDiv" : 22,
+      "Ios16.reduceMean" : 44,
+      "Ios17.batchNorm" : 21,
+      "Ios16.softmax" : 1,
+      "Ios17.reshape" : 49,
+      "Ios16.silu" : 21
+    },
+    "computePrecision" : "Mixed (Float32, Int32)",
+    "stateSchema" : [
+
+    ],
+    "isUpdatable" : "0",
+    "availability" : {
+      "macOS" : "14.0",
+      "tvOS" : "17.0",
+      "visionOS" : "1.0",
+      "watchOS" : "10.0",
+      "iOS" : "17.0",
+      "macCatalyst" : "17.0"
+    },
+    "modelType" : {
+      "name" : "MLModelType_mlProgram"
+    },
+    "inputSchema" : [
+      {
+        "hasShapeFlexibility" : "0",
+        "isOptional" : "0",
+        "dataType" : "Float32",
+        "formattedType" : "MultiArray (Float32 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" : "x",
+        "type" : "MultiArray"
+      }
+    ],
+    "userDefinedMetadata" : {
+      "com.github.apple.coremltools.source_dialect" : "TorchScript",
+      "com.github.apple.coremltools.source" : "torch==2.4.0",
+      "com.github.apple.coremltools.version" : "8.0"
+    },
+    "generatedClassName" : "Stable_Diffusion_version_LyliaEngine_waiIllustriousSDXL_v170_vae_encoder",
+    "method" : "predict"
+  }
+]