monster/Qwen-Image-Layered-MRP-MLX
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Commit Graph

  • 9e6ace4cc3 Apply ruff formating (minor manual modifications to some files which are subsequently excluded) filipstrand 2024-09-21 15:39:52 +0200
  • 101711dcd4 Merge remote-tracking branch 'refs/remotes/origin/main' into intro-ruff-config filipstrand 2024-09-21 19:58:45 +0200
  • abd7893612 Fix hidden_states in single_transformer_blocks for controlnet transformer filipstrand 2024-09-21 14:36:09 +0200
  • a8d4300a86
         Merge pull request #57 from anthonywu/intro-makefile Filip Strand 2024-09-21 13:45:11 +0200
  • 557996049e Add newlines to fix markdown rendering filipstrand 2024-09-21 10:46:00 +0200
  • c170e89977 remove unnecessary imports filipstrand 2024-09-21 10:39:01 +0200
  • 3b3e232365
         Merge pull request #55 from anthonywu/easy-import-dx Filip Strand 2024-09-21 10:36:27 +0200
  • 5a7157e7e8 Simplify import paths for controlnet file filipstrand 2024-09-21 10:23:05 +0200
  • e007402cfa Merge remote-tracking branch 'refs/remotes/mflux/main' into easy-import-dx filipstrand 2024-09-21 10:33:36 +0200
  • cb785e49f1 introduce but not enforce .pre-commit-config.yaml Anthony Wu 2024-09-20 19:06:49 -0700
  • 77e5a2ef39 intro ruff config Anthony Wu 2024-09-20 10:26:10 -0700
  • d565807e5b update readme to match Anthony Wu 2024-09-19 12:48:06 -0700
  • 7cc2d4174d introduce Makefile with uv recommendation Anthony Wu 2024-09-19 12:37:22 -0700
  • 3daf4bc7de
         Merge pull request #50 from anthonywu/project-setup-recs Filip Strand 2024-09-20 12:20:50 +0200
  • c74279d394 pyproject.toml to be authoritative source of deps, soft introduce ruff linter/formatter Anthony Wu 2024-09-19 11:20:06 -0700
  • d7fe524c92
         Merge pull request #48 from Xuzzo/feature/add_controlnet Filip Strand 2024-09-17 23:50:34 +0200
  • 38c747b680 Extract model saving logic to separate class filipstrand 2024-09-17 07:13:59 +0200
  • a30dc4d5cc Rename image.py to generated_image.py filipstrand 2024-09-17 06:56:08 +0200
  • bce18747ff Fix formatting filipstrand 2024-09-17 06:53:56 +0200
  • fc8099e908 Move new controlnet module to its own file filipstrand 2024-09-17 06:50:02 +0200
  • ad6f4ca8af Create util class for controlnet filipstrand 2024-09-17 06:43:13 +0200
  • 78bbf3d8fb Fix formatting and some warnings filipstrand 2024-09-17 06:41:05 +0200
  • 25e375d45d Small additions and refactoring filipstrand 2024-09-17 06:35:06 +0200
  • 43618154d3 remove apparently unnecessary import of the src/mflux directory Anthony Wu 2024-09-16 15:04:30 -0700
  • 0812d6252c simplify import paths of frequently used items in scripts and demos Anthony Wu 2024-09-16 10:27:04 -0700
  • a2900c3b13 update generated image annotations Fabio 2024-09-15 21:54:17 +0200
  • dee452b3bb fixes from review Fabio 2024-09-15 21:37:29 +0200
  • a04d139ff5 further cleanup Fabio 2024-09-14 20:04:08 +0200
  • ff1086c06f add egg-info to gitignore Fabio 2024-09-14 20:02:06 +0200
  • 2a2a3fff63 delete egg-info Fabio 2024-09-14 20:01:33 +0200
  • 12a6386fca Merge branch 'main' into feature/add_controlnet Fabio 2024-09-14 19:58:51 +0200
  • 369bfc0b53 add cli, controlnet config and tidy up Fabio 2024-09-14 19:58:12 +0200
  • ca4e3f4c79 v.0.2.1 filipstrand 2024-09-14 14:46:07 +0200
  • 3dba1e38e1 controlnet - first working version Fabio 2024-09-14 14:37:30 +0200
  • 93e80ea724
         Merge pull request #46 from filipstrand/better-lora-support Filip Strand 2024-09-13 22:39:23 +0200
  • d2553f2951 Add supported lora format section to readme filipstrand 2024-09-13 22:20:14 +0200
  • db3e495be6 Remove restriction on lora weights filipstrand 2024-09-13 21:50:48 +0200
  • d10e6f4019 Add support for converting between LoRA formats filipstrand 2024-09-13 20:12:35 +0200
  • 1b9c645503 WIP controlnet Fabio 2024-09-13 00:28:43 +0200
  • 7d27f1c4c1 Adjust FF block only if it exists filipstrand 2024-09-12 20:43:06 +0200
  • 3739a6b914 Link to madroid 4-bit filipstrand 2024-09-10 18:57:34 +0200
  • fe0d01495f Suggest first setting up a venv when installing filipstrand 2024-09-09 18:52:43 +0200
  • 9d4027698c Default to latest mflux package filipstrand 2024-09-07 12:31:53 +0200
  • 0a90939bcc
         Merge pull request #32 from filipstrand/PyPI-release Filip Strand 2024-09-07 10:53:43 +0200
  • a2965a1833 Skip python-requires for now filipstrand 2024-09-06 22:51:27 +0200
  • 2510965027 Swap setup.py for pyproject.toml filipstrand 2024-09-06 22:46:50 +0200
  • 1769baa382 Define custom commands for image generation and model saving filipstrand 2024-09-06 22:28:53 +0200
  • 1567a01745 Rename run to generate filipstrand 2024-09-06 22:27:24 +0200
  • 2a7ca560a2 Update installation guide filipstrand 2024-09-06 22:15:06 +0200
  • c0d419777e Add quantize option to from_alias initializer filipstrand 2024-09-06 22:09:06 +0200
  • af73966f11 Fix system paths filipstrand 2024-09-06 21:44:54 +0200
  • b9ff93ea3f Fix imports filipstrand 2024-09-06 21:15:46 +0200
  • ea3362e8ab Revert "Temporarily allow .jpg and .png before renaming" filipstrand 2024-09-06 21:10:14 +0200
  • 00a556cbf2 Add setup.py filipstrand 2024-09-06 20:54:10 +0200
  • 23e44c768b Rename main.py to run.py filipstrand 2024-09-06 20:51:15 +0200
  • b8102f2a2f Move main.py and save.py one level down filipstrand 2024-09-06 20:50:53 +0200
  • 915f3c2950 Move flux.py to separate package filipstrand 2024-09-06 20:50:22 +0200
  • d2b38011ad Rename flux_1 to mflux filipstrand 2024-09-06 21:06:41 +0200
  • 893f6be9bb Temporarily allow .jpg and .png before renaming filipstrand 2024-09-06 21:04:50 +0200
  • 5b0bd4b7e7
         Merge pull request #27 from filipstrand/save-image-info Filip Strand 2024-09-06 20:23:28 +0200
  • 8e4e8f68d6 Fix example filipstrand 2024-09-06 20:20:11 +0200
  • 6fb51170de Also save LoRA metadata filipstrand 2024-09-06 20:08:53 +0200
  • 6d05f5b27e Fix saving bug filipstrand 2024-09-06 19:51:28 +0200
  • 1db28b7ffe Fix example filipstrand 2024-09-06 19:48:55 +0200
  • fab10cc046 Add proper image class and save metadata filipstrand 2024-09-06 19:19:05 +0200
  • 97dc64436e
         Merge pull request #31 from Xuzzo/fix/width_height_mixup Filip Strand 2024-09-06 10:02:46 +0200
  • 5431c9faa5
         Merge pull request #30 from elitexp/main Filip Strand 2024-09-05 23:28:33 +0200
  • 50c567c187 Fix link filipstrand 2024-09-05 23:26:06 +0200
  • 1d8714d3b1 Add some example runs filipstrand 2024-09-05 23:13:53 +0200
  • a0ee9a7a8a Rename apply-lora arg filipstrand 2024-09-05 21:26:48 +0200
  • 0a2f21af5a fix widht <-> height bug Fabio 2024-09-05 20:06:59 +0200
  • 96848e320f Add support for multiple LoRAs again filipstrand 2024-09-05 19:37:50 +0200
  • ff9f998638 Small fixes filipstrand 2024-09-04 23:16:02 +0200
  • d5f5f9607e
         Fix type Filip Strand 2024-09-04 23:14:38 +0200
  • f28fffe0cf Point out the .safetensors file again filipstrand 2024-09-04 22:42:29 +0200
  • a3567bddd7 Start supporting only single LoRA (and some other simplifications) filipstrand 2024-09-04 21:32:10 +0200
  • 6d35db5e9a Use mx.load instead of safe_open filipstrand 2024-09-04 21:05:19 +0200
  • 5ad5352481 Introduce weight util for shared methods filipstrand 2024-09-04 20:13:32 +0200
  • af41b08d90 Move LoRA related things to a LoraUtil (have to think about where to best place this later...) filipstrand 2024-09-04 19:53:24 +0200
  • 8983070728 Code style improvements and some restructuring filipstrand 2024-09-04 19:48:15 +0200
  • b4c27a6fab
         raise error if layer cannot be loaded Filip Strand 2024-09-04 19:43:32 +0200
  • 0fd03c7552
         @Xuzzo - Weight Adjustment as Original Transformer elitexp 2024-09-04 21:52:51 +0545
  • 8d8812675e feat: --apply-lora a.safetensors b.safetensors c.safetensor --lora-scales 0.5 1.0 0.5 Sujip Maharjan 2024-09-03 19:36:15 +0545
  • 03b929e732 feat: Add --lora-scale parameter to adjust the scaling of LoRA weights Sujip Maharjan 2024-09-03 18:39:01 +0545
  • 4dfb3aa892
         Clean up weight handler elitexp 2024-09-03 18:19:21 +0545
  • f82caa3298 Icrecream dependency removed, not required -- error Sujip Maharjan 2024-09-03 13:16:35 +0545
  • cd54ad137c Apply lora with --apply-lora <path-to-safetensors> Sujip Maharjan 2024-09-03 07:38:04 +0545
  • 94920896e8
         Merge pull request #24 from filipstrand/basic-quantization-support Filip Strand 2024-08-30 22:24:05 +0200
  • fdc4cb9892 Rewrite weight handling and add quantization support filipstrand 2024-08-30 14:04:58 +0200
  • 478fc4f8bc Update readme with new TODO list filipstrand 2024-08-23 11:54:21 +0200
  • bf26270695 Clarify where models are downloaded filipstrand 2024-08-23 10:27:45 +0200
  • 1254f3827c Update readme with new performance numbers filipstrand 2024-08-21 21:22:16 +0200
  • 58700ebed3 Update readme with new performance numbers filipstrand 2024-08-21 11:11:36 +0200
  • 580817afb6 Link to issue tracker for downloading the dev model filipstrand 2024-08-21 09:56:11 +0200
  • 336d8a17da Add missing guidance flag in readme description filipstrand 2024-08-21 09:10:23 +0200
  • 1f54b94175 Simplify: Remove latent creator class and inline logic filipstrand 2024-08-21 00:48:08 +0200
  • 45104ebf48 Fix width property filipstrand 2024-08-20 23:22:25 +0200
  • c6f0a17d34 Once again ignore png and jpg filipstrand 2024-08-20 22:56:50 +0200
  • d0970162a5 Fix images in readme filipstrand 2024-08-20 22:56:30 +0200
  • 04c2aaac26 Add helpful hints when configuring which model to use in example filipstrand 2024-08-20 22:46:31 +0200