e12202c1f6
Merge pull request #368 from Aintor/main
...
Fix for Mask Patch Failure and Quantization Issues in Latest `transformers` Versions
2025-07-03 11:01:02 -07:00
e17915a9cb
Add --min-deployment-target argument for configurable iOS/macOS deployment targets
...
- Added --min-deployment-target argument with default value macOS13 for backward compatibility
- Supports macOS13, macOS14, macOS15, iOS16, iOS17, iOS18 targets
- Graceful fallback for newer targets not available in older coremltools versions
- Enables targeting iOS 18 for advanced quantization features while maintaining compatibility
- Includes comprehensive help text explaining usage and compatibility requirements
2025-06-14 08:27:06 -07:00
e5d960c41a
Update README.md for converting-models-to-coreml reference
...
Fix typo that does not anchor the links correctly
2025-01-23 08:28:43 -08:00
f8e36141a9
Add support for mixed precision activation quantization for UNet model ( #365 )
2024-11-15 16:05:22 -08:00
50bb4c43d0
Update torch2coreml.py
2024-11-12 11:19:16 +08:00
7e3df29361
Update torch2coreml.py
2024-11-12 10:36:17 +08:00
4ef83f13d0
Update torch2coreml.py
2024-11-12 10:31:26 +08:00
200f255cf4
Merge pull request #367 from TobyRoseman/swift-serial-unet-predictions
...
In Swift, allow serial uNet predictions
2024-11-11 09:01:34 -08:00
244703046f
Update torch2coreml.py
2024-11-11 00:13:05 +08:00
bdbe459eaf
In Swift, allow serial unet predictions
2024-11-08 09:18:39 -08:00
bef26ae4ff
Merge pull request #364 from TobyRoseman/add-unet-sequential-option
...
Add Option for Sequential Unet Predictions
2024-10-23 16:40:45 -07:00
7505f91208
Add option for sequential unet predictions
2024-10-23 14:23:19 -07:00
6365319010
fix: Remove duplicate logger instantiation
2024-10-15 19:41:54 -07:00
6775c2c01a
Allow not using classifier free guidance for inference in Swift code
2024-10-08 12:06:39 -07:00
1c194dafb4
Merge pull request #360 from TobyRoseman/unet-batch-1-or-2
...
Allow not using classifier free guidance
2024-10-04 13:57:20 -07:00
e497107793
Allow not using classifier free guidance
2024-10-04 11:10:18 -07:00
cab24e9a51
Merge pull request #357 from TobyRoseman/dev
...
Fix Cartoonish Output for non-XL model
2024-09-18 15:23:47 -07:00
3e04bcecf3
On macOS 15+, set fast prediction optimization hint for the unet.
2024-09-18 09:26:45 -07:00
32be4af031
Pin versions for important depedencies
2024-09-17 13:47:02 -07:00
aa0ec6e569
Do not force zeros embeddings for empty prompt, unless using XL
2024-09-17 13:21:33 -07:00
73a4fca864
Merge pull request #355 from junpeiz/patch-1
...
Add diffusionkit to installation dependency
2024-08-28 10:25:37 -07:00
5d5f155512
Add diffusionkit into setup.py
...
Adds diffusionkit to package installation dependency to avoid erroring out when converting UNet.
2024-08-26 15:29:13 -07:00
70cfcce955
Merge pull request #354 from jakesabathia2/eng/PR-henry-refactor-chunk-script
...
Refactor the model chunking script to work with all version of coremltools
2024-08-23 11:39:52 -07:00
c6884bf3b5
add some logging
2024-08-23 11:31:46 -07:00
3fc577cc41
fix typo
2024-08-22 16:20:24 -07:00
f7b2618c0d
refactor the model chunking script to work with all version of coremltools
2024-08-22 14:24:46 -07:00
cf16df8207
Merge pull request #348 from lamtrinhdev/Update_README.md
...
Correct wording in README.md
2024-08-11 09:47:48 -07:00
cf4e1e8d6e
Update README.md
...
Correct wording.
2024-08-11 20:07:36 +07:00
c891f4357b
Add SD3 Pipeline ( #329 )
...
* Add SD3 Pipeline
Co-authored-by: atiorh <atiorh@users.noreply.github.com >
Co-authored-by: arda-argmax <arda-argmax@users.noreply.github.com >
* Use swift-transformers for tokenization
* Use diffusionkit converters in torch2coreml
* Documentation and cleanup
* Add model link
* Consolidate batch prediction logic
* Remove DecoderSD3.swift and consolidate logic into Decoder.swift
* Remove DiffusionKit MLX inference reference from README
---------
Co-authored-by: atiorh <atiorh@users.noreply.github.com >
Co-authored-by: arda-argmax <arda-argmax@users.noreply.github.com >
Co-authored-by: atila <atiorh@icloud.com >
2024-07-22 21:16:50 -07:00
5a170d29cf
Update pull_request_template.md after changing to MIT license
2024-05-03 18:03:21 -07:00
c5be2739d0
Update CONTRIBUTING.md to reflect the MIT license
2024-05-03 17:59:39 -07:00
5d8cdbb92d
Update LICENSE.md to MIT
...
Changing repository license to MIT
2024-05-03 17:58:28 -07:00
f7d2f5e9fb
Update swift-argument-parser to latest minor version
...
This will also make ml-stable-diffusion more tolerant with regard to which swift-argument-parser version it needs, resulting in fewer version conflicts in Xcode projects.
2024-03-12 12:38:44 -07:00
d456a972cd
Merge pull request #311 from TobyRoseman/mbp-new-api
...
Applying mixed bit compression using new optimize API
2024-01-23 16:42:12 -08:00
ec09fe064f
code review feedback
2024-01-23 16:41:35 -08:00
a9b84c658e
Update Documentation
2024-01-23 08:32:22 -08:00
9a3d463397
Applying mixed bit compression using new optimize API
2024-01-17 15:43:08 -08:00
7449ce46a4
Update CGImage+vImage.swift
2023-11-14 10:35:27 -08:00
5c62fab073
Use sRGB colorspace
2023-11-14 10:35:27 -08:00
3a0aeba5e5
Merge pull request #296 from TobyRoseman/compiled_models
...
Compiled Models in Python
2023-11-07 13:18:01 -08:00
46c94e8a86
Link to Core ML documentation about why compiled models are faster
2023-11-07 12:46:13 -08:00
4c54204bb0
Compiled Models in Python
2023-11-07 12:13:31 -08:00
bea04420b5
Merge pull request #287 from TobyRoseman/minor-fix
...
Fix pipeline.py for Python 3.8
2023-10-24 08:26:53 -07:00
d159b9254c
Fix pipeline.py for Python 3.8
2023-10-23 10:27:56 -07:00
567dbf2757
fixes #281
2023-10-20 22:45:43 -07:00
f4466f0b73
fix error ( #279 )
2023-10-10 12:15:08 -07:00
7fe1bf23b9
Cast safety_checker to float32 ( #276 )
2023-10-06 20:54:39 -07:00
11ba3a0965
Add custom_vae_version in MBP pre-analysis
2023-09-30 19:44:01 -07:00
96c044e07b
Add iPhone 12 Pro Benchmark
2023-09-28 14:34:09 -07:00
081a8cb2cc
v1.1
2023-09-28 07:26:41 -07:00
Aseem Wadhwa