diffsynth 2.1.10__tar.gz → 2.1.12__tar.gz

{diffsynth-2.1.10 → diffsynth-2.1.12}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: diffsynth
-Version: 2.1.10
+Version: 2.1.12
 Summary: Enjoy the magic of Diffusion models!
 Author: ModelScope Team
 License-Expression: Apache-2.0
@@ -114,6 +114,8 @@ See also:
 
 > Currently, the development personnel of this project are limited, with most of the work handled by [Artiprocher](https://github.com/Artiprocher) and [mi804](https://github.com/mi804). Therefore, the progress of new feature development will be relatively slow, and the speed of responding to and resolving issues is limited. We apologize for this and ask developers to understand.
 
+- **October 21, 2026** We now support Fibo 2.0, including unified text-to-image and image editing inference, masked and multi-reference editing, text rendering, low VRAM inference, and full/LoRA training for both text-to-image and image editing on [Fibo-2.0-dev](https://www.modelscope.cn/models/briaai/Fibo-2.0-dev). See the [documentation](/docs/en/Model_Details/Fibo.md) and [example code](/examples/fibo/).
+
 - **October 20, 2026** We now support MOSS-Image, including inference and training for [MOSS-Image-Base](https://modelscope.cn/models/openmoss/MOSS-Image-Base) and [MOSS-Image-SFT](https://modelscope.cn/models/openmoss/MOSS-Image-SFT). In addition, we provide an [EntroPack low-bit quantized model](https://www.modelscope.cn/models/DiffSynth-Studio/MOSS-Image-EntroPack), which can run the 3.5-bit quantized model with as little as 4 GB of VRAM; and an [accelerated model](https://modelscope.cn/models/DiffSynth-Studio/MOSS-Image-SFT-TurboLoRA) distillation-accelerated via [OPD training](https://diffsynth-studio-doc.readthedocs.io/en/latest/Training/Distillation.html), supporting 8-step ultra-fast inference.
 
 - **October 16, 2026** flux-3-image-klein open-sourced, welcome a new member to the image model family! Support includes text-to-image generation, multi-reference editing, layout control, low VRAM inference, and Base training capabilities. For details, please refer to the [documentation](/docs/en/Model_Details/FLUX.3.md) and [example code](/examples/flux3/). We quantized the model with [EntroPack](https://github.com/modelscope/entropack), providing pre-quantized models in 3-bit, 4-bit, and 6-bit versions at [DiffSynth-Studio/flux-3-image-klein-entropack](https://www.modelscope.cn/models/DiffSynth-Studio/flux-3-image-klein-entropack).
@@ -124,7 +126,7 @@ See also:
 
   * Project Page: https://github.com/modelscope/entropack
   * Technical Report: https://arxiv.org/abs/2609.34185
-  * Pre-quantized Models: https://www.modelscope.cn/models/DiffSynth-Studio/EntroPackPreQuants
+  * Pre-quantized Models: https://www.modelscope.cn/collections/DiffSynth-Studio/EntroPack-Quantized-Models
   * Example Code: [Z-Image-Turbo](/examples/z_image/model_inference_low_vram/Z-Image-Turbo-EntroPack.py), [Qwen-Image-2.1](/examples/qwen_image_21/model_inference_low_vram/Qwen-Image-2.1-EntroPack.py), [MiniMax-H3](/examples/minimax_h3/model_inference_low_vram/MiniMax-H3-EntroPack.py), [LoRA Training Example](/examples/z_image/model_training/special/quant_training/Z-Image-Turbo-entropack_lossy_quant.sh)
 
 - **September 20, 2026** Qwen-Image-2.1 was open-sourced. The model supports text-to-image generation, image editing, and RGBA output with a transparency channel. DiffSynth-Studio now provides full support for it. For details, please refer to the [documentation](/docs/zh/Model_Details/Qwen-Image-2.1.md) and [example code](/examples/qwen_image_21/).
@@ -402,6 +404,8 @@ Quick start: experience popular and the latest models:
 
 | Architecture | Model ID | Inference | Low VRAM Inference | Full Training | Validation After Full Training | LoRA Training | Validation After LoRA Training |
 |-|-|-|-|-|-|-|-|
+| Fibo 2.0 | [briaai/Fibo-2.0-dev](https://www.modelscope.cn/models/briaai/Fibo-2.0-dev) | [code](/examples/fibo/model_inference/Fibo-2.0-dev.py) | [code](/examples/fibo/model_inference_low_vram/Fibo-2.0-dev.py) | [code](/examples/fibo/model_training/full/Fibo-2.0-dev-T2I.sh) | [code](/examples/fibo/model_training/validate_full/Fibo-2.0-dev-T2I.py) | [code](/examples/fibo/model_training/lora/Fibo-2.0-dev-T2I.sh) | [code](/examples/fibo/model_training/validate_lora/Fibo-2.0-dev-T2I.py) |
+| Fibo 2.0 | [briaai/Fibo-2.0-edit-turbo](https://www.modelscope.cn/models/briaai/Fibo-2.0-edit-turbo) | [code](/examples/fibo/model_inference/Fibo-2.0-edit-turbo.py) | [code](/examples/fibo/model_inference_low_vram/Fibo-2.0-edit-turbo.py) | - | - | - | - |
 | Qwen-Image-2.1 | [Qwen/Qwen-Image-2.1](https://www.modelscope.cn/models/Qwen/Qwen-Image-2.1) | [code](/examples/qwen_image_21/model_inference/Qwen-Image-2.1.py) | [code](/examples/qwen_image_21/model_inference_low_vram/Qwen-Image-2.1.py) | [code](/examples/qwen_image_21/model_training/full/Qwen-Image-2.1.sh) | [code](/examples/qwen_image_21/model_training/validate_full/Qwen-Image-2.1.py) | [code](/examples/qwen_image_21/model_training/lora/Qwen-Image-2.1.sh) | [code](/examples/qwen_image_21/model_training/validate_lora/Qwen-Image-2.1.py) |
 | Qwen-Image-2.1 | [Qwen/Qwen-Image-2.1-Turbo](https://www.modelscope.cn/models/Qwen/Qwen-Image-2.1-Turbo) | [code](/examples/qwen_image_21/model_inference/Qwen-Image-2.1-Turbo.py) | [code](/examples/qwen_image_21/model_inference_low_vram/Qwen-Image-2.1-Turbo.py) | [code](/examples/qwen_image_21/model_training/full/Qwen-Image-2.1-Turbo.sh) | [code](/examples/qwen_image_21/model_training/validate_full/Qwen-Image-2.1-Turbo.py) | [code](/examples/qwen_image_21/model_training/lora/Qwen-Image-2.1-Turbo.sh) | [code](/examples/qwen_image_21/model_training/validate_lora/Qwen-Image-2.1-Turbo.py) |
 | MiniMax-H3 | [MiniMax/MiniMax-H3: FL2VA](https://www.modelscope.cn/models/MiniMax/MiniMax-H3) | [code](/examples/minimax_h3/model_inference/MiniMax-H3-FL2VA.py) | [code](/examples/minimax_h3/model_inference_low_vram/MiniMax-H3-FL2VA.py) | [code](/examples/minimax_h3/model_training/full/MiniMax-H3-FL2VA.sh) | [code](/examples/minimax_h3/model_training/validate_full/MiniMax-H3-FL2VA.py) | [code](/examples/minimax_h3/model_training/lora/MiniMax-H3-FL2VA.sh) | [code](/examples/minimax_h3/model_training/validate_lora/MiniMax-H3-FL2VA.py) |
@@ -414,6 +418,7 @@ Quick start: experience popular and the latest models:
 Model overview:
 
 - Image generation
+    - Fibo 2.0: [Documentation](https://diffsynth-studio-doc.readthedocs.io/en/latest/Model_Details/Fibo.html), [Example code](/examples/fibo/)
     - MOSS-Image: [Documentation](https://diffsynth-studio-doc.readthedocs.io/en/latest/Model_Details/MOSS-Image.html), [Example code](/examples/moss_image/)
     - FLUX.3: [Documentation](https://diffsynth-studio-doc.readthedocs.io/en/latest/Model_Details/FLUX.3.html), [Example code](/examples/flux3/)
     - Qwen-Image-2.1: [Documentation](https://diffsynth-studio-doc.readthedocs.io/en/latest/Model_Details/Qwen-Image-2.1.html), [Example code](/examples/qwen_image_21/)
@@ -702,6 +707,8 @@ https://github.com/Artiprocher/DiffSynth-Studio/assets/35051019/59fb2f7b-8de0-44
 
 | Architecture | Model ID | Inference | Low VRAM Inference | Full Training | Validation After Full Training | LoRA Training | Validation After LoRA Training |
 |-|-|-|-|-|-|-|-|
+| Fibo 2.0 | [briaai/Fibo-2.0-dev](https://www.modelscope.cn/models/briaai/Fibo-2.0-dev) | [code](/examples/fibo/model_inference/Fibo-2.0-dev.py) | [code](/examples/fibo/model_inference_low_vram/Fibo-2.0-dev.py) | [code](/examples/fibo/model_training/full/Fibo-2.0-dev-T2I.sh) | [code](/examples/fibo/model_training/validate_full/Fibo-2.0-dev-T2I.py) | [code](/examples/fibo/model_training/lora/Fibo-2.0-dev-T2I.sh) | [code](/examples/fibo/model_training/validate_lora/Fibo-2.0-dev-T2I.py) |
+| Fibo 2.0 | [briaai/Fibo-2.0-edit-turbo](https://www.modelscope.cn/models/briaai/Fibo-2.0-edit-turbo) | [code](/examples/fibo/model_inference/Fibo-2.0-edit-turbo.py) | [code](/examples/fibo/model_inference_low_vram/Fibo-2.0-edit-turbo.py) | - | - | - | - |
 | MOSS-Image | [openmoss/MOSS-Image-Base](https://www.modelscope.cn/models/openmoss/MOSS-Image-Base) | [code](/examples/moss_image/model_inference/MOSS-Image-Base.py) | [code](/examples/moss_image/model_inference_low_vram/MOSS-Image-Base.py) | [code](/examples/moss_image/model_training/full/MOSS-Image-Base.sh) | [code](/examples/moss_image/model_training/validate_full/MOSS-Image-Base.py) | [code](/examples/moss_image/model_training/lora/MOSS-Image-Base.sh) | [code](/examples/moss_image/model_training/validate_lora/MOSS-Image-Base.py) |
 | MOSS-Image | [openmoss/MOSS-Image-SFT](https://www.modelscope.cn/models/openmoss/MOSS-Image-SFT) | [code](/examples/moss_image/model_inference/MOSS-Image-SFT.py) | [code](/examples/moss_image/model_inference_low_vram/MOSS-Image-SFT.py) | [code](/examples/moss_image/model_training/full/MOSS-Image-SFT.sh) | [code](/examples/moss_image/model_training/validate_full/MOSS-Image-SFT.py) | [code](/examples/moss_image/model_training/lora/MOSS-Image-SFT.sh) | [code](/examples/moss_image/model_training/validate_lora/MOSS-Image-SFT.py) |
 | MOSS-Image | [DiffSynth-Studio/MOSS-Image-EntroPack](https://www.modelscope.cn/models/DiffSynth-Studio/MOSS-Image-EntroPack) | [code](/examples/moss_image/model_inference/MOSS-Image-EntroPack.py) | [code](/examples/moss_image/model_inference_low_vram/MOSS-Image-EntroPack.py) | - | - | [code](/examples/moss_image/model_training/lora/MOSS-Image-entropack.sh) | [code](/examples/moss_image/model_training/validate_lora/MOSS-Image-entropack.py) |

{diffsynth-2.1.10 → diffsynth-2.1.12}/README.md

@@ -41,6 +41,8 @@ See also:
 
 > Currently, the development personnel of this project are limited, with most of the work handled by [Artiprocher](https://github.com/Artiprocher) and [mi804](https://github.com/mi804). Therefore, the progress of new feature development will be relatively slow, and the speed of responding to and resolving issues is limited. We apologize for this and ask developers to understand.
 
+- **October 21, 2026** We now support Fibo 2.0, including unified text-to-image and image editing inference, masked and multi-reference editing, text rendering, low VRAM inference, and full/LoRA training for both text-to-image and image editing on [Fibo-2.0-dev](https://www.modelscope.cn/models/briaai/Fibo-2.0-dev). See the [documentation](/docs/en/Model_Details/Fibo.md) and [example code](/examples/fibo/).
+
 - **October 20, 2026** We now support MOSS-Image, including inference and training for [MOSS-Image-Base](https://modelscope.cn/models/openmoss/MOSS-Image-Base) and [MOSS-Image-SFT](https://modelscope.cn/models/openmoss/MOSS-Image-SFT). In addition, we provide an [EntroPack low-bit quantized model](https://www.modelscope.cn/models/DiffSynth-Studio/MOSS-Image-EntroPack), which can run the 3.5-bit quantized model with as little as 4 GB of VRAM; and an [accelerated model](https://modelscope.cn/models/DiffSynth-Studio/MOSS-Image-SFT-TurboLoRA) distillation-accelerated via [OPD training](https://diffsynth-studio-doc.readthedocs.io/en/latest/Training/Distillation.html), supporting 8-step ultra-fast inference.
 
 - **October 16, 2026** flux-3-image-klein open-sourced, welcome a new member to the image model family! Support includes text-to-image generation, multi-reference editing, layout control, low VRAM inference, and Base training capabilities. For details, please refer to the [documentation](/docs/en/Model_Details/FLUX.3.md) and [example code](/examples/flux3/). We quantized the model with [EntroPack](https://github.com/modelscope/entropack), providing pre-quantized models in 3-bit, 4-bit, and 6-bit versions at [DiffSynth-Studio/flux-3-image-klein-entropack](https://www.modelscope.cn/models/DiffSynth-Studio/flux-3-image-klein-entropack).
@@ -51,7 +53,7 @@ See also:
 
   * Project Page: https://github.com/modelscope/entropack
   * Technical Report: https://arxiv.org/abs/2609.34185
-  * Pre-quantized Models: https://www.modelscope.cn/models/DiffSynth-Studio/EntroPackPreQuants
+  * Pre-quantized Models: https://www.modelscope.cn/collections/DiffSynth-Studio/EntroPack-Quantized-Models
   * Example Code: [Z-Image-Turbo](/examples/z_image/model_inference_low_vram/Z-Image-Turbo-EntroPack.py), [Qwen-Image-2.1](/examples/qwen_image_21/model_inference_low_vram/Qwen-Image-2.1-EntroPack.py), [MiniMax-H3](/examples/minimax_h3/model_inference_low_vram/MiniMax-H3-EntroPack.py), [LoRA Training Example](/examples/z_image/model_training/special/quant_training/Z-Image-Turbo-entropack_lossy_quant.sh)
 
 - **September 20, 2026** Qwen-Image-2.1 was open-sourced. The model supports text-to-image generation, image editing, and RGBA output with a transparency channel. DiffSynth-Studio now provides full support for it. For details, please refer to the [documentation](/docs/zh/Model_Details/Qwen-Image-2.1.md) and [example code](/examples/qwen_image_21/).
@@ -329,6 +331,8 @@ Quick start: experience popular and the latest models:
 
 | Architecture | Model ID | Inference | Low VRAM Inference | Full Training | Validation After Full Training | LoRA Training | Validation After LoRA Training |
 |-|-|-|-|-|-|-|-|
+| Fibo 2.0 | [briaai/Fibo-2.0-dev](https://www.modelscope.cn/models/briaai/Fibo-2.0-dev) | [code](/examples/fibo/model_inference/Fibo-2.0-dev.py) | [code](/examples/fibo/model_inference_low_vram/Fibo-2.0-dev.py) | [code](/examples/fibo/model_training/full/Fibo-2.0-dev-T2I.sh) | [code](/examples/fibo/model_training/validate_full/Fibo-2.0-dev-T2I.py) | [code](/examples/fibo/model_training/lora/Fibo-2.0-dev-T2I.sh) | [code](/examples/fibo/model_training/validate_lora/Fibo-2.0-dev-T2I.py) |
+| Fibo 2.0 | [briaai/Fibo-2.0-edit-turbo](https://www.modelscope.cn/models/briaai/Fibo-2.0-edit-turbo) | [code](/examples/fibo/model_inference/Fibo-2.0-edit-turbo.py) | [code](/examples/fibo/model_inference_low_vram/Fibo-2.0-edit-turbo.py) | - | - | - | - |
 | Qwen-Image-2.1 | [Qwen/Qwen-Image-2.1](https://www.modelscope.cn/models/Qwen/Qwen-Image-2.1) | [code](/examples/qwen_image_21/model_inference/Qwen-Image-2.1.py) | [code](/examples/qwen_image_21/model_inference_low_vram/Qwen-Image-2.1.py) | [code](/examples/qwen_image_21/model_training/full/Qwen-Image-2.1.sh) | [code](/examples/qwen_image_21/model_training/validate_full/Qwen-Image-2.1.py) | [code](/examples/qwen_image_21/model_training/lora/Qwen-Image-2.1.sh) | [code](/examples/qwen_image_21/model_training/validate_lora/Qwen-Image-2.1.py) |
 | Qwen-Image-2.1 | [Qwen/Qwen-Image-2.1-Turbo](https://www.modelscope.cn/models/Qwen/Qwen-Image-2.1-Turbo) | [code](/examples/qwen_image_21/model_inference/Qwen-Image-2.1-Turbo.py) | [code](/examples/qwen_image_21/model_inference_low_vram/Qwen-Image-2.1-Turbo.py) | [code](/examples/qwen_image_21/model_training/full/Qwen-Image-2.1-Turbo.sh) | [code](/examples/qwen_image_21/model_training/validate_full/Qwen-Image-2.1-Turbo.py) | [code](/examples/qwen_image_21/model_training/lora/Qwen-Image-2.1-Turbo.sh) | [code](/examples/qwen_image_21/model_training/validate_lora/Qwen-Image-2.1-Turbo.py) |
 | MiniMax-H3 | [MiniMax/MiniMax-H3: FL2VA](https://www.modelscope.cn/models/MiniMax/MiniMax-H3) | [code](/examples/minimax_h3/model_inference/MiniMax-H3-FL2VA.py) | [code](/examples/minimax_h3/model_inference_low_vram/MiniMax-H3-FL2VA.py) | [code](/examples/minimax_h3/model_training/full/MiniMax-H3-FL2VA.sh) | [code](/examples/minimax_h3/model_training/validate_full/MiniMax-H3-FL2VA.py) | [code](/examples/minimax_h3/model_training/lora/MiniMax-H3-FL2VA.sh) | [code](/examples/minimax_h3/model_training/validate_lora/MiniMax-H3-FL2VA.py) |
@@ -341,6 +345,7 @@ Quick start: experience popular and the latest models:
 Model overview:
 
 - Image generation
+    - Fibo 2.0: [Documentation](https://diffsynth-studio-doc.readthedocs.io/en/latest/Model_Details/Fibo.html), [Example code](/examples/fibo/)
     - MOSS-Image: [Documentation](https://diffsynth-studio-doc.readthedocs.io/en/latest/Model_Details/MOSS-Image.html), [Example code](/examples/moss_image/)
     - FLUX.3: [Documentation](https://diffsynth-studio-doc.readthedocs.io/en/latest/Model_Details/FLUX.3.html), [Example code](/examples/flux3/)
     - Qwen-Image-2.1: [Documentation](https://diffsynth-studio-doc.readthedocs.io/en/latest/Model_Details/Qwen-Image-2.1.html), [Example code](/examples/qwen_image_21/)
@@ -629,6 +634,8 @@ https://github.com/Artiprocher/DiffSynth-Studio/assets/35051019/59fb2f7b-8de0-44
 
 | Architecture | Model ID | Inference | Low VRAM Inference | Full Training | Validation After Full Training | LoRA Training | Validation After LoRA Training |
 |-|-|-|-|-|-|-|-|
+| Fibo 2.0 | [briaai/Fibo-2.0-dev](https://www.modelscope.cn/models/briaai/Fibo-2.0-dev) | [code](/examples/fibo/model_inference/Fibo-2.0-dev.py) | [code](/examples/fibo/model_inference_low_vram/Fibo-2.0-dev.py) | [code](/examples/fibo/model_training/full/Fibo-2.0-dev-T2I.sh) | [code](/examples/fibo/model_training/validate_full/Fibo-2.0-dev-T2I.py) | [code](/examples/fibo/model_training/lora/Fibo-2.0-dev-T2I.sh) | [code](/examples/fibo/model_training/validate_lora/Fibo-2.0-dev-T2I.py) |
+| Fibo 2.0 | [briaai/Fibo-2.0-edit-turbo](https://www.modelscope.cn/models/briaai/Fibo-2.0-edit-turbo) | [code](/examples/fibo/model_inference/Fibo-2.0-edit-turbo.py) | [code](/examples/fibo/model_inference_low_vram/Fibo-2.0-edit-turbo.py) | - | - | - | - |
 | MOSS-Image | [openmoss/MOSS-Image-Base](https://www.modelscope.cn/models/openmoss/MOSS-Image-Base) | [code](/examples/moss_image/model_inference/MOSS-Image-Base.py) | [code](/examples/moss_image/model_inference_low_vram/MOSS-Image-Base.py) | [code](/examples/moss_image/model_training/full/MOSS-Image-Base.sh) | [code](/examples/moss_image/model_training/validate_full/MOSS-Image-Base.py) | [code](/examples/moss_image/model_training/lora/MOSS-Image-Base.sh) | [code](/examples/moss_image/model_training/validate_lora/MOSS-Image-Base.py) |
 | MOSS-Image | [openmoss/MOSS-Image-SFT](https://www.modelscope.cn/models/openmoss/MOSS-Image-SFT) | [code](/examples/moss_image/model_inference/MOSS-Image-SFT.py) | [code](/examples/moss_image/model_inference_low_vram/MOSS-Image-SFT.py) | [code](/examples/moss_image/model_training/full/MOSS-Image-SFT.sh) | [code](/examples/moss_image/model_training/validate_full/MOSS-Image-SFT.py) | [code](/examples/moss_image/model_training/lora/MOSS-Image-SFT.sh) | [code](/examples/moss_image/model_training/validate_lora/MOSS-Image-SFT.py) |
 | MOSS-Image | [DiffSynth-Studio/MOSS-Image-EntroPack](https://www.modelscope.cn/models/DiffSynth-Studio/MOSS-Image-EntroPack) | [code](/examples/moss_image/model_inference/MOSS-Image-EntroPack.py) | [code](/examples/moss_image/model_inference_low_vram/MOSS-Image-EntroPack.py) | - | - | [code](/examples/moss_image/model_training/lora/MOSS-Image-entropack.sh) | [code](/examples/moss_image/model_training/validate_lora/MOSS-Image-entropack.py) |

{diffsynth-2.1.10 → diffsynth-2.1.12}/diffsynth/configs/model_configs.py

@@ -1843,9 +1843,19 @@ flux3_series = [
     },
 ]
 
+fibo_series = [
+    {
+        # Example: ModelConfig(model_id="briaai/Fibo-2.0-dev", origin_file_pattern="transformer/*.safetensors")
+        # Example: ModelConfig(model_id="briaai/Fibo-2.0-edit-turbo", origin_file_pattern="transformer/*.safetensors")
+        "model_hash": "39f09f95705fd6d9b20670f1923d3c12",
+        "model_name": "fibo_dit",
+        "model_class": "diffsynth.models.fibo_dit.FiboDiT",
+    },
+]
+
 MODEL_CONFIGS = (
     stable_diffusion_xl_series + stable_diffusion_series + qwen_image_series + qwen_video_edit_series + wan_series + flux_series + flux2_series + ernie_image_series
     + z_image_series + ltx2_series + anima_series + mova_series + joyai_image_series + boogu_image_series + ace_step_series + hidream_o1_image_series
     + image_metrics_series + ideogram4_series + krea2_series + lingbot_video_series + minimax_h3_series + minimax_music3_series + sensenova_u1_series
-    + yue2_series + qwen_image_21_series + flux3_series + moss_image_series
+    + yue2_series + qwen_image_21_series + flux3_series + moss_image_series + fibo_series
 )

{diffsynth-2.1.10 → diffsynth-2.1.12}/diffsynth/configs/vram_management_module_maps.py

@@ -256,6 +256,14 @@ VRAM_MANAGEMENT_MODULE_MAPS = {
         "torch.nn.Linear": "diffsynth.core.vram.layers.AutoWrappedLinear",
         "torch.nn.Conv2d": "diffsynth.core.vram.layers.AutoWrappedModule",
         "torch.nn.GroupNorm": "diffsynth.core.vram.layers.AutoWrappedModule",
+        "diffsynth.models.flux2_vae.Flux2VAELatentNorm": "diffsynth.core.vram.layers.AutoWrappedModule",
+    },
+    "diffsynth.models.fibo_dit.FiboDiT": {
+        "torch.nn.Linear": "diffsynth.core.vram.layers.AutoWrappedLinear",
+        "diffsynth.models.general_modules.RMSNorm": "diffsynth.core.vram.layers.AutoWrappedModule",
+        "torch.nn.LayerNorm": "diffsynth.core.vram.layers.AutoWrappedModule",
+        "diffsynth.models.fibo_dit.FiboPerceiverResampler": "diffsynth.core.vram.layers.AutoWrappedModule",
+        "diffsynth.models.fibo_dit.FiboTransformerBlock": "diffsynth.core.vram.layers.AutoWrappedModule",
     },
     "diffsynth.models.z_image_text_encoder.ZImageTextEncoder": {
         "torch.nn.Linear": "diffsynth.core.vram.layers.AutoWrappedLinear",

{diffsynth-2.1.10 → diffsynth-2.1.12}/diffsynth/core/quant/backends/entropack.py

@@ -1,4 +1,4 @@
-from dataclasses import dataclass
+from dataclasses import dataclass, fields
 
 import torch
 
@@ -45,15 +45,20 @@ if ENTROPACK_AVAILABLE:
     @dataclass
     class EntroPackINT8CodedConfig(BackendConfig, ep.LatticeRANSConfig):
         linear_cls = ep.CompressedINT8Linear
+        use_convrot: bool = False
+        convrot_groupsize: int = 256
 
     @dataclass
     class EntroPackINT8RawConfig(BackendConfig, ep.RawConfig):
         linear_cls = ep.CompressedINT8Linear
+        use_convrot: bool = False
+        convrot_groupsize: int = 256
 
 
-def _graph_break_quantized_linear(module):
-    module.forward = torch._dynamo.disable(module.forward)
-    return module
+def _patch_compile_graph_breaks():
+    for cls in (ep.CompressedLinear, ep.CompressedFP8Linear, ep.CompressedINT8Linear):
+        if not getattr(cls.forward, "_torchdynamo_disable", False):
+            cls.forward = torch._dynamo.disable(cls.forward)
 
 
 @register_quant_backend("entropack")
@@ -70,6 +75,7 @@ class EntroPackQuantBackend(QuantBackend):
                 '`pip install "entropack[cuda13]"` (use `cuda12` for a CUDA 12 environment) '
                 'or `pip install "diffsynth[quant]"`.'
             ) from _ENTROPACK_IMPORT_ERROR
+        _patch_compile_graph_breaks()
 
     def capabilities(self):
         return {
@@ -85,20 +91,44 @@ class EntroPackQuantBackend(QuantBackend):
     def unflatten_state_dict(self, state_dict, metadata):
         return state_dict
 
+    def _compression_config_cls(self):
+        for cls in type(self.config).__mro__:
+            if cls.__module__.startswith("entropack") and issubclass(cls, ep.CompressionConfig):
+                return cls
+        return None
+
+    def _compression_config(self):
+        cls = self._compression_config_cls()
+        if cls is None:
+            return self.config
+        names = {field.name for field in fields(cls) if field.init}
+        return cls(**{name: getattr(self.config, name) for name in names})
+
+    def linear_spec_configs(self):
+        cls = self._compression_config_cls()
+        if cls is None:
+            return {}
+        compression = {field.name for field in fields(cls)}
+        return {
+            field.name: getattr(self.config, field.name)
+            for field in fields(self.config)
+            if field.init and field.name not in compression
+        }
+
     def create_quantized_linear(self, linear, compute_device=None, model_device=None):
         linear.requires_grad_(False)
         if compute_device is not None:
             linear = linear.to(device=compute_device)
-        quantized = self.config.linear_cls.from_linear(linear, config=self.config)
-        quantized = quantized if model_device is None else quantized.to(device=model_device)
-        return _graph_break_quantized_linear(quantized)
+        quantized = self.config.linear_cls.from_linear(
+            linear, config=self._compression_config(), **self.linear_spec_configs(),
+        )
+        return quantized if model_device is None else quantized.to(device=model_device)
 
     def create_quantized_linear_shell(self, linear, compute_dtype):
-        shell = self.config.linear_cls(
+        return self.config.linear_cls(
             linear.in_features, linear.out_features, bias=linear.bias is not None,
-            dtype=compute_dtype, config=self.config,
+            dtype=compute_dtype, config=self._compression_config(), **self.linear_spec_configs(),
         )
-        return _graph_break_quantized_linear(shell)
 
     def dequantize_to_linear(self, module, compute_dtype, compute_device=None, model_device=None):
         if compute_device is not None:

{diffsynth-2.1.10 → diffsynth-2.1.12}/diffsynth/core/vram/layers.py

@@ -246,10 +246,33 @@ class AutoWrappedNonRecurseModule(AutoWrappedModule):
         if self.disk_offload:
             self.required_params = [name for name, _ in self.module.named_parameters(recurse=False)]
             
+    def _copy_with_shared_children(self, module):
+        # Child wrappers own their DiskMap and devices; share them instead of copying.
+        shared = {id(child): child for child in module.modules() if isinstance(child, AutoTorchModule)}
+        return copy.deepcopy(module, shared), set(shared)
+
+    def _detach_shared_children(self, module, shared_ids):
+        detached = []
+        for parent in list(module.modules()):
+            if isinstance(parent, AutoTorchModule):
+                continue
+            original = parent._modules
+            filtered = type(original)(
+                (name, child) for name, child in original.items() if id(child) not in shared_ids
+            )
+            if len(filtered) != len(original):
+                parent._modules = filtered
+                detached.append((parent, original))
+        return detached
+
+    def _restore_shared_children(self, detached):
+        for parent, original in detached:
+            parent._modules = original
+
     @torch._dynamo.disable
     def load_from_disk(self, torch_dtype, device, copy_module=False):
         if copy_module:
-            module = copy.deepcopy(self.module)
+            module, shared_ids = self._copy_with_shared_children(self.module)
         else:
             module = self.module
         state_dict = {}
@@ -257,7 +280,14 @@ class AutoWrappedNonRecurseModule(AutoWrappedModule):
             param = self.disk_map[self.param_name(name)]
             param = param.to(dtype=torch_dtype, device=device)
             state_dict[name] = param
-        module.load_state_dict(state_dict, assign=True, strict=False)
+        if copy_module:
+            detached = self._detach_shared_children(module, shared_ids)
+            try:
+                module.load_state_dict(state_dict, assign=True, strict=False)
+            finally:
+                self._restore_shared_children(detached)
+        else:
+            module.load_state_dict(state_dict, assign=True, strict=False)
         return module
     
     def offload_to_disk(self, model: torch.nn.Module):
@@ -265,8 +295,13 @@ class AutoWrappedNonRecurseModule(AutoWrappedModule):
             getattr(self, name).to("meta")
     
     def cast_to(self, module, dtype, device):
-        # Parameter casting is implemented in the model architecture.
-        return module
+        result, shared_ids = self._copy_with_shared_children(module)
+        detached = self._detach_shared_children(result, shared_ids)
+        try:
+            result = result.to(dtype=dtype, device=device)
+        finally:
+            self._restore_shared_children(detached)
+        return result
     
     def __getattr__(self, name):
         if name in self.__dict__ or name == "module":
@@ -531,6 +566,7 @@ class AutoWrappedQuantizedModule(torch.nn.Linear, AutoTorchModule, LoRAHotLoadMi
                 self.module.to(device=self.onload_device)
             self.state = 1
 
+    @torch._dynamo.disable
     def preparing(self):
         if self.state != 2:
             if self.disk_offload and self.preparing_device != "disk" and self.onload_device == "disk":
@@ -539,6 +575,7 @@ class AutoWrappedQuantizedModule(torch.nn.Linear, AutoTorchModule, LoRAHotLoadMi
                 self.module.to(device=self.preparing_device)
             self.state = 2
 
+    @torch._dynamo.disable
     def computation_module(self):
         device = self.preparing_device if self.state == 2 else self.onload_device
         if device == self.computation_device:

{diffsynth-2.1.10 → diffsynth-2.1.12}/diffsynth/diffusion/flow_match.py

@@ -5,7 +5,7 @@ from typing_extensions import Literal
 
 class FlowMatchScheduler():
 
-    def __init__(self, template: Literal["FLUX.1", "Wan", "Qwen-Image", "FLUX.2", "FLUX.3", "Z-Image", "LTX-2", "Qwen-Image-Lightning", "ERNIE-Image", "MOSS-Image", "ACE-Step", "Ideogram4", "Krea-2", "Boogu", "MiniMax-H3", "MiniMax-Music3", "LingBot-Video", "SenseNova-U1", "YuE2"] = "FLUX.1"):
+    def __init__(self, template: Literal["FLUX.1", "Wan", "Qwen-Image", "FLUX.2", "FLUX.3", "Z-Image", "LTX-2", "Qwen-Image-Lightning", "ERNIE-Image", "MOSS-Image", "ACE-Step", "Ideogram4", "Krea-2", "Boogu", "MiniMax-H3", "MiniMax-Music3", "LingBot-Video", "SenseNova-U1", "YuE2", "Fibo"] = "FLUX.1"):
         self.set_timesteps_fn = {
             "FLUX.1": FlowMatchScheduler.set_timesteps_flux,
             "Wan": FlowMatchScheduler.set_timesteps_wan,
@@ -27,6 +27,7 @@ class FlowMatchScheduler():
             "LingBot-Video": FlowMatchScheduler.set_timesteps_lingbot_video,
             "SenseNova-U1": FlowMatchScheduler.set_timesteps_sensenova_u1,
             "YuE2": FlowMatchScheduler.set_timesteps_yue2,
+            "Fibo": FlowMatchScheduler.set_timesteps_fibo,
         }.get(template, FlowMatchScheduler.set_timesteps_flux)
         self.num_train_timesteps = 1000
 
@@ -168,6 +169,22 @@ class FlowMatchScheduler():
         timesteps = sigmas * num_train_timesteps
         return sigmas, timesteps
 
+    @staticmethod
+    def set_timesteps_fibo(num_inference_steps=30, denoising_strength=1.0, dynamic_shift_len=None):
+        sigma_min = 1 / num_inference_steps
+        sigma_max = 1.0
+        sigma_start = sigma_min + (sigma_max - sigma_min) * denoising_strength
+        sigmas = torch.linspace(sigma_start, sigma_min, num_inference_steps)
+        if dynamic_shift_len is None:
+            mu = 0.5
+        else:
+            mu = FlowMatchScheduler._calculate_shift_qwen_image(
+                dynamic_shift_len, base_seq_len=256, max_seq_len=4096, base_shift=0.5, max_shift=1.15
+            )
+        sigmas = math.exp(mu) / (math.exp(mu) + (1 / sigmas - 1))
+        timesteps = sigmas * 1000
+        return sigmas, timesteps
+
     @staticmethod
     def set_timesteps_flux3(num_inference_steps=16, denoising_strength=1.0, alpha=6.93, **kwargs):
         u = torch.linspace(denoising_strength, 0.0, num_inference_steps + 1)[:-1]
@@ -188,7 +205,7 @@ class FlowMatchScheduler():
         return sigmas, timesteps
 
     @staticmethod
-    def set_timesteps_moss_image(num_inference_steps=50, denoising_strength=1.0, shift=5.0, training_shift=1.0):
+    def set_timesteps_moss_image(num_inference_steps=50, denoising_strength=1.0, shift=1.0, training_shift=1.0):
         num_train_timesteps = 1000
 
         def grid(count, shift, start=1.0):