Pedro Cuenca 27df66021d Remove mlpackage save workaround. (#240) ...

https://github.com/apple/coremltools/issues/1680 has been resolved in
coremltools 7.0b1, which is now required. The reason for addressing this
change now is that reloading the model before save may incur a
performance cost (specially on SDXL models) due to `ANECompilerService`
triggering and executing for a long time.

2023-08-24 22:16:28 +03:00

..

__init__.py

Initial commit

2022-11-30 13:41:05 -06:00

_version.py

WWDC23 update

2023-06-14 08:27:02 -07:00

attention.py

WWDC23 update

2023-06-14 08:27:02 -07:00

chunk_mlprogram.py

Initial commit

2022-11-30 13:41:05 -06:00

controlnet.py

Support ControlNet (#153)

2023-04-18 16:17:59 -07:00

coreml_model.py

Support ControlNet (#153)

2023-04-18 16:17:59 -07:00

layer_norm.py

Initial commit

2022-11-30 13:41:05 -06:00

mixed_bit_compression_apply.py

Mixed-bit Palettization

2023-07-27 09:07:44 -10:00

mixed_bit_compression_pre_analysis.py

Fix multiple issues in mixed bit compression pre analysis (#231)

2023-08-22 00:00:22 +03:00

multilingual_projection.py

WWDC23 update

2023-06-14 08:27:02 -07:00

pipeline.py

Fixes #160

2023-04-24 12:13:30 -07:00

torch2coreml.py

Remove mlpackage save workaround. (#240)

2023-08-24 22:16:28 +03:00

unet.py

SDXL and other improvements

2023-07-27 09:08:30 -10:00