apple/ml-stable-diffusion issues and pull requests

#372 - Would it be possible to port this with some preprocessor macros to Intel Mac?

Issue - State: open - Opened by shyamalschandra 4 days ago

#370 - Removing Logic for Legacy Model Chunking

Pull Request - State: open - Opened by greshbasic 8 days ago

#369 - SD3.5 and Flux.1 on ANE !!!???

Issue - State: open - Opened by InformEthics 10 days ago

#368 - Fix for Mask Patch Failure and Quantization Issues in Latest `transformers` Versions

Pull Request - State: open - Opened by Aintor 16 days ago - 3 comments

#367 - In Swift, allow serial uNet predictions

Pull Request - State: closed - Opened by TobyRoseman 18 days ago

#366 - Issue downloading Stable Diffusion 3.5 Medium

Issue - State: open - Opened by mihok 27 days ago - 4 comments

#365 - Add support for mixed precision activation quantization for UNet model

Pull Request - State: closed - Opened by NehalBhandari 29 days ago - 1 comment

#364 - Add Option for Sequential Unet Predictions

Pull Request - State: closed - Opened by TobyRoseman about 1 month ago

#363 - Gray Image Output with cpuOnly Compute Units on iOS

Issue - State: open - Opened by rustui about 1 month ago

#362 - Add --min-deployment-target Argument to Set Deployment Target

Issue - State: open - Opened by swiftlysingh about 1 month ago

#361 - Allow not using classifier free guidance for inference in Swift code

Pull Request - State: closed - Opened by alejandro-isaza about 2 months ago

#360 - Allow not using classifier free guidance

Pull Request - State: closed - Opened by TobyRoseman about 2 months ago

#359 - Fix invalid mask implementation

Pull Request - State: open - Opened by KyuHwan00 2 months ago - 5 comments

#358 - Fix invalid mask implementation

Pull Request - State: closed - Opened by KyuHwan00 2 months ago

#357 - Fix Cartoonish Output for non-XL model

Pull Request - State: closed - Opened by TobyRoseman 2 months ago

#356 - Quanitzed Chunking

Issue - State: open - Opened by nighting0le01 2 months ago

#355 - Add diffusionkit to installation dependency

Pull Request - State: closed - Opened by junpeiz 3 months ago - 1 comment

#354 - Refactor the model chunking script to work with all version of coremltools

Pull Request - State: closed - Opened by jakesabathia2 3 months ago

#353 - Chunking quantized model leads to unequal Chunks

Issue - State: closed - Opened by nighting0le01 3 months ago - 2 comments

#352 - Pix2Pix assertion failed

Issue - State: open - Opened by kacperd 3 months ago - 1 comment

#351 - Why GPU and NPU give different results?

Issue - State: open - Opened by chn-lee-yumi 3 months ago - 2 comments

#350 - fix: Remove duplicate logger instantiation

Pull Request - State: closed - Opened by LaikaN57 3 months ago

#349 - Processing issue - Leaked semaphore objects - There appear to be %d

Issue - State: open - Opened by Retha23 4 months ago - 1 comment

#348 - Correct wording in README.md

Pull Request - State: closed - Opened by lamtrinhdev 4 months ago

#347 - Support - FLUX black-forest-labs/FLUX.1-schnell

Issue - State: open - Opened by mgierschdev 4 months ago - 1 comment

#346 - I want a new release. I can not find StableDiffusion3Pipeline in 1.1.1

Issue - State: open - Opened by LLLLLayer 4 months ago - 1 comment

#345 - ERROR - converting 'scaled_dot_product_attention' op (located at: 'text_encoder/text_model/encoder/0/self_attn'):

Issue - State: open - Opened by khanhvp2k 4 months ago - 2 comments

#344 - Example SwiftUI App

Issue - State: open - Opened by igor11191708 4 months ago

#343 - How can I change the image's size and aspect ratio?

Issue - State: open - Opened by caol64 4 months ago - 4 comments

#342 - fixes sd3-version argument needed for mixed-bit compression

Pull Request - State: open - Opened by c2p-cmd 4 months ago

#334 - MLFeatureDescription "time_ids" crash

Issue - State: open - Opened by vedlaicm 5 months ago

#333 - chunk unet does not work with coremltool>7.1

Issue - State: closed - Opened by nighting0le01 5 months ago - 1 comment

#332 - Bug fixes, Runtime & Performance optimization

Pull Request - State: open - Opened by DrHazemAli 5 months ago - 2 comments

#331 - ValueError: Input X contains infinity or a value too large for dtype('float64').

Issue - State: open - Opened by jaycoolslm 5 months ago - 1 comment

#330 - iOS App storage (documents and data) keep increasing

Issue - State: open - Opened by chn-lee-yumi 5 months ago - 4 comments

#329 - Add SD3 Pipeline

Pull Request - State: closed - Opened by ZachNagengast 6 months ago - 4 comments

#328 - Plz supprot Stable Cascade!

Issue - State: open - Opened by h3clikejava 6 months ago

#327 - Changed requirements versions in requiremnts.txt and setup.py files

Pull Request - State: closed - Opened by TheRealSalvo 6 months ago - 6 comments

#326 - Mixed Palettization of SD 1.5 LCM model

Issue - State: open - Opened by indoflaven 8 months ago - 4 comments

#325 - Input X contains infinity - Error when converting quantized SD-Turbo

Issue - State: closed - Opened by indoflaven 8 months ago - 3 comments

#324 - add LCMScheduler

Pull Request - State: closed - Opened by Jadeeeey 8 months ago

#323 - Custom Pipeline

Issue - State: open - Opened by alex-seville 8 months ago

#322 - Question about quantization in torch2coreml

Issue - State: closed - Opened by coleridge72 8 months ago

#321 - Question about chunking UNET

Issue - State: closed - Opened by coleridge72 8 months ago - 1 comment

#320 - webui.sh missing

Issue - State: closed - Opened by MikeyBeez 8 months ago - 1 comment

#319 - LCM scheduler

Issue - State: open - Opened by ThibaultCastells 9 months ago - 2 comments

#318 - Update swift-argument-parser to latest minor version

Pull Request - State: closed - Opened by godly-devotion 9 months ago

#317 - Roadmap for this branch

Issue - State: open - Opened by SpiraMira 9 months ago

#316 - Fix quantize-nbits flag

Pull Request - State: open - Opened by nosferatu500 9 months ago - 5 comments

#315 - CGImage extension method planarRGBShapedArray seems to leak memory

Issue - State: open - Opened by giedrius-stanevicius 9 months ago

#314 - Chunk Size for Einsum and Einsumv2 attention vs Resolution

Issue - State: closed - Opened by nighting0le01 10 months ago - 2 comments

#313 - Won't work on GPU on macOS 14.4 Beta 1 due to CoreML bug

Issue - State: closed - Opened by MenKuch 10 months ago - 3 comments

#312 - Torch2coreml multiprocessing

Issue - State: open - Opened by rbourgeat 10 months ago

#311 - Applying mixed bit compression using new optimize API

Pull Request - State: closed - Opened by TobyRoseman 10 months ago - 1 comment

#310 - When will training be implemented?

Issue - State: open - Opened by midescribba 10 months ago

#309 - Crashes on getting `currentImages`

Issue - State: open - Opened by sindresorhus 11 months ago

#308 - Add Euler Ancestral Discrete Scheduler

Pull Request - State: open - Opened by bofeizhu 11 months ago - 1 comment

#307 - How to modify or update the model weight in swift?

Issue - State: open - Opened by simon5u 11 months ago - 2 comments

#306 - EnumeratedShapes horrible performance

Issue - State: open - Opened by cchance27 11 months ago

#305 - SDXL Turbo support?

Issue - State: open - Opened by pj4533 12 months ago - 5 comments

#304 - SDXL produces black output images when given a starting image

Issue - State: closed - Opened by nosferatu500 12 months ago - 3 comments

#303 - Flag quantize-nbits returns error

Issue - State: closed - Opened by citruslab about 1 year ago - 9 comments

#302 - Adding multiple ControlNets has no effect

Issue - State: open - Opened by citruslab about 1 year ago - 1 comment

#301 - Use sRGB if available

Pull Request - State: closed - Opened by stuartjmoore about 1 year ago

#300 - bugfix: updated scaling factor for SDXL

Pull Request - State: open - Opened by lopez-hector about 1 year ago - 1 comment

#299 - Support async for generateImage and compatible with original interface.

Pull Request - State: closed - Opened by chiyoi about 1 year ago - 3 comments

#298 - Memory Crash on iPhone12 Pro Max

Issue - State: open - Opened by lamfan about 1 year ago - 10 comments

#297 - SDXL produces black output images when given a starting image

Issue - State: closed - Opened by operatesspaceships about 1 year ago - 3 comments

#296 - Compiled Models in Python

Pull Request - State: closed - Opened by TobyRoseman about 1 year ago

#295 - Memory „Leak“ when using cpu+gpu

Issue - State: open - Opened by MenKuch about 1 year ago - 6 comments

#294 - Colorspace

Issue - State: open - Opened by stuartjmoore about 1 year ago - 1 comment

#293 - mixed_bit_compression_pre_analysis: Weird recipe results

Issue - State: open - Opened by MenKuch about 1 year ago - 4 comments

#292 - loadResources took long time to load UNET

Issue - State: open - Opened by lamfan about 1 year ago - 6 comments

#291 - Memory Issues on 4 GB iOS/iPadOS devices

Issue - State: open - Opened by MenKuch about 1 year ago - 4 comments

#290 - Only one controlnet is supported

Issue - State: closed - Opened by qlmaster about 1 year ago

#289 - Conversion will be much slower on Python 3.9.6

Issue - State: closed - Opened by TimYao18 about 1 year ago - 3 comments

#288 - Model conversion failed

Issue - State: closed - Opened by TimYao18 about 1 year ago - 3 comments

#287 - Fix pipeline.py for Python 3.8

Pull Request - State: closed - Opened by TobyRoseman about 1 year ago - 1 comment

#286 - adapt segmind tiny model

Pull Request - State: closed - Opened by ZhouXiaolin about 1 year ago - 1 comment

#285 - Convert model: check-output-correctness met RuntimeError: mat1 and mat2 must have the same dtype

Issue - State: closed - Opened by TimYao18 about 1 year ago - 1 comment

#284 - result not as expected

Issue - State: open - Opened by lamfan about 1 year ago - 6 comments

#283 - [Performance]A17 PRO ANE has much more computing power than M2, but its stable diffusion performance is worse than M2?

Issue - State: open - Opened by AndreaChiChengdu about 1 year ago - 7 comments

#282 - Converting DreamShaper_8 results in invalid model

Issue - State: open - Opened by MenKuch about 1 year ago - 4 comments

#281 - Can not get predictions from baseline UNet

Issue - State: closed - Opened by TobyRoseman about 1 year ago

#280 - `Unable to compute the prediction using ML Program. It can be an invalid input data or broken/unsupported model.` on Swift image generation

Issue - State: closed - Opened by maciejtrybilo about 1 year ago - 2 comments

#279 - Fix missing parameters ('mixed_bit_compression_apply', ) & import pkg (mixed_bit_compression_pre_analysis) issues

Pull Request - State: closed - Opened by jiangdi0924 about 1 year ago

#278 - How to use with Easy Diffusion (or other GUI)?

Issue - State: closed - Opened by JasonS05 about 1 year ago - 1 comment

#277 - Update README.md with Python example

Pull Request - State: open - Opened by burningion about 1 year ago

#276 - Fix safety_checker dtype

Pull Request - State: closed - Opened by ZachNagengast about 1 year ago

#275 - Safety checker fails to trace

Issue - State: closed - Opened by TobyRoseman about 1 year ago - 2 comments

#274 - Why .mlpackage format no need the vocab.json & merges.txt file?

Issue - State: open - Opened by TimYao18 about 1 year ago - 3 comments

#273 - SDXL Refiner MBP pre-analysis

Issue - State: closed - Opened by GuiyeC about 1 year ago - 4 comments

#272 - Merge dicts with unpacking to support Python 3.8

Pull Request - State: open - Opened by 3c1u about 1 year ago

#271 - sdxl inference is too slow (0.02 step/sec) on m2 macbook air with macos14

Issue - State: open - Opened by 0ihsan about 1 year ago - 2 comments

GitHub / apple/ml-stable-diffusion issues and pull requests