[ { "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" } ]