Z-Image-Turbo with student+adapter text encoder
Browse files- .gitattributes +11 -0
- README.md +200 -0
- assets/DMDR.webp +3 -0
- assets/Z-Image-Gallery.pdf +3 -0
- assets/architecture.webp +3 -0
- assets/decoupled-dmd.webp +3 -0
- assets/leaderboard.png +3 -0
- assets/leaderboard.webp +0 -0
- assets/reasoning.png +3 -0
- assets/showcase.jpg +3 -0
- assets/showcase_editing.png +3 -0
- assets/showcase_realistic.png +3 -0
- assets/showcase_rendering.png +3 -0
- custom_adapter/adapter_config.json +20 -0
- custom_adapter/adapter_model.safetensors +3 -0
- model_index.json +24 -0
- scheduler/scheduler_config.json +7 -0
- student_adapter_build_meta.json +26 -0
- text_encoder/__init__.py +4 -0
- text_encoder/config.json +141 -0
- text_encoder/configuration_student_adapter.py +30 -0
- text_encoder/model.safetensors +3 -0
- text_encoder/modeling_student_adapter.py +137 -0
- tokenizer/merges.txt +0 -0
- tokenizer/tokenizer.json +3 -0
- tokenizer/tokenizer_config.json +239 -0
- tokenizer/vocab.json +0 -0
- transformer/config.json +31 -0
- transformer/diffusion_pytorch_model-00001-of-00003.safetensors +3 -0
- transformer/diffusion_pytorch_model-00002-of-00003.safetensors +3 -0
- transformer/diffusion_pytorch_model-00003-of-00003.safetensors +3 -0
- transformer/diffusion_pytorch_model.safetensors.index.json +528 -0
- vae/config.json +38 -0
- vae/diffusion_pytorch_model.safetensors +3 -0
.gitattributes
CHANGED
|
@@ -33,3 +33,14 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
|
|
| 33 |
*.zip filter=lfs diff=lfs merge=lfs -text
|
| 34 |
*.zst filter=lfs diff=lfs merge=lfs -text
|
| 35 |
*tfevents* filter=lfs diff=lfs merge=lfs -text
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 33 |
*.zip filter=lfs diff=lfs merge=lfs -text
|
| 34 |
*.zst filter=lfs diff=lfs merge=lfs -text
|
| 35 |
*tfevents* filter=lfs diff=lfs merge=lfs -text
|
| 36 |
+
tokenizer/tokenizer.json filter=lfs diff=lfs merge=lfs -text
|
| 37 |
+
assets/DMDR.webp filter=lfs diff=lfs merge=lfs -text
|
| 38 |
+
assets/architecture.webp filter=lfs diff=lfs merge=lfs -text
|
| 39 |
+
assets/decoupled-dmd.webp filter=lfs diff=lfs merge=lfs -text
|
| 40 |
+
assets/reasoning.png filter=lfs diff=lfs merge=lfs -text
|
| 41 |
+
assets/showcase.jpg filter=lfs diff=lfs merge=lfs -text
|
| 42 |
+
assets/showcase_editing.png filter=lfs diff=lfs merge=lfs -text
|
| 43 |
+
assets/showcase_realistic.png filter=lfs diff=lfs merge=lfs -text
|
| 44 |
+
assets/showcase_rendering.png filter=lfs diff=lfs merge=lfs -text
|
| 45 |
+
assets/Z-Image-Gallery.pdf filter=lfs diff=lfs merge=lfs -text
|
| 46 |
+
assets/leaderboard.png filter=lfs diff=lfs merge=lfs -text
|
README.md
ADDED
|
@@ -0,0 +1,200 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: apache-2.0
|
| 3 |
+
language:
|
| 4 |
+
- en
|
| 5 |
+
pipeline_tag: text-to-image
|
| 6 |
+
library_name: diffusers
|
| 7 |
+
---
|
| 8 |
+
|
| 9 |
+
|
| 10 |
+
<h1 align="center">⚡️- Image<br><sub><sup>An Efficient Image Generation Foundation Model with Single-Stream Diffusion Transformer</sup></sub></h1>
|
| 11 |
+
|
| 12 |
+
<div align="center">
|
| 13 |
+
|
| 14 |
+
[](https://tongyi-mai.github.io/Z-Image-blog/) 
|
| 15 |
+
[](https://github.com/Tongyi-MAI/Z-Image) 
|
| 16 |
+
[](https://huggingface.co/Tongyi-MAI/Z-Image-Turbo) 
|
| 17 |
+
[](https://huggingface.co/spaces/Tongyi-MAI/Z-Image-Turbo) 
|
| 18 |
+
[](https://huggingface.co/spaces/akhaliq/Z-Image-Turbo) 
|
| 19 |
+
[](https://www.modelscope.cn/models/Tongyi-MAI/Z-Image-Turbo) 
|
| 20 |
+
[](https://www.modelscope.cn/aigc/imageGeneration?tab=advanced&versionId=469191&modelType=Checkpoint&sdVersion=Z_IMAGE_TURBO&modelUrl=modelscope%3A%2F%2FTongyi-MAI%2FZ-Image-Turbo%3Frevision%3Dmaster) 
|
| 21 |
+
[](assets/Z-Image-Gallery.pdf) 
|
| 22 |
+
[](https://modelscope.cn/studios/Tongyi-MAI/Z-Image-Gallery/summary) 
|
| 23 |
+
<a href="https://arxiv.org/abs/2511.22699" target="_blank"><img src="https://img.shields.io/badge/Report-b5212f.svg?logo=arxiv" height="21px"></a>
|
| 24 |
+
|
| 25 |
+
|
| 26 |
+
Welcome to the official repository for the Z-Image（造相）project!
|
| 27 |
+
|
| 28 |
+
</div>
|
| 29 |
+
|
| 30 |
+
|
| 31 |
+
|
| 32 |
+
## ✨ Z-Image
|
| 33 |
+
|
| 34 |
+
Z-Image is a powerful and highly efficient image generation model family with **6B** parameters. Currently there are four variants:
|
| 35 |
+
|
| 36 |
+
- 🚀 **Z-Image-Turbo** – A distilled version of Z-Image that matches or exceeds leading competitors with only **8 NFEs** (Number of Function Evaluations). It offers **⚡️sub-second inference latency⚡️** on enterprise-grade H800 GPUs and fits comfortably within **16G VRAM consumer devices**. It excels in photorealistic image generation, bilingual text rendering (English & Chinese), and robust instruction adherence.
|
| 37 |
+
|
| 38 |
+
- 🎨 **Z-Image** – The foundation model behind Z-Image-Turbo. Z-Image focuses on **high-quality generation**, **rich aesthetics**, **strong diversity**, and **controllability**, well-suited for creative generation, **fine-tuning**, and downstream development. It supports a wide range of artistic styles, effective negative prompting, and high diversity across identities, poses, compositions, and layouts.
|
| 39 |
+
|
| 40 |
+
- 🧱 **Z-Image-Omni-Base** – The versatile foundation model capable of both **generation and editing tasks**. By releasing this checkpoint, we aim to unlock the full potential for community-driven fine-tuning and custom development, providing the most "raw" and diverse starting point for the open-source community.
|
| 41 |
+
|
| 42 |
+
- ✍️ **Z-Image-Edit** – A variant fine-tuned on Z-Image specifically for image editing tasks. It supports creative image-to-image generation with impressive instruction-following capabilities, allowing for precise edits based on natural language prompts.
|
| 43 |
+
|
| 44 |
+
### 📥 Model Zoo
|
| 45 |
+
|
| 46 |
+
| Model | Pre-Training | SFT | RL | Step | CFG | Task | Visual Quality | Diversity | Fine-Tunability | Hugging Face | ModelScope |
|
| 47 |
+
| :--- | :---: | :---: | :---: | :---: | :---: | :---: | :---: | :---: | :---: | :---: | :---: |
|
| 48 |
+
| **Z-Image-Omni-Base** | ✅ | ❌ | ❌ | 50 | ✅ | Gen. / Editing | Medium | High | Easy | *To be released* | *To be released* |
|
| 49 |
+
| **Z-Image** | ✅ | ✅ | ❌ | 50 | ✅ | Gen. | High | Medium | Easy | [](https://huggingface.co/Tongyi-MAI/Z-Image) <br> [](https://huggingface.co/spaces/Tongyi-MAI/Z-Image) | [](https://www.modelscope.cn/models/Tongyi-MAI/Z-Image) <br> [](https://www.modelscope.cn/aigc/imageGeneration?tab=advanced&versionId=569345&modelType=Checkpoint&sdVersion=Z_IMAGE&modelUrl=modelscope%3A%2F%2FTongyi-MAI%2FZ-Image%3Frevision%3Dmaster) |
|
| 50 |
+
| **Z-Image-Turbo** | ✅ | ✅ | ✅ | 8 | ❌ | Gen. | Very High | Low | N/A | [](https://huggingface.co/Tongyi-MAI/Z-Image-Turbo) <br> [](https://huggingface.co/spaces/Tongyi-MAI/Z-Image-Turbo) | [](https://www.modelscope.cn/models/Tongyi-MAI/Z-Image-Turbo) <br> [](https://www.modelscope.cn/aigc/imageGeneration?tab=advanced&versionId=469191&modelType=Checkpoint&sdVersion=Z_IMAGE_TURBO&modelUrl=modelscope%3A%2F%2FTongyi-MAI%2FZ-Image-Turbo%3Frevision%3Dmaster) |
|
| 51 |
+
| **Z-Image-Edit** | ✅ | ✅ | ❌ | 50 | ✅ | Editing | High | Medium | Easy | *To be released* | *To be released* | | *To be released* |
|
| 52 |
+
|
| 53 |
+
### 🖼️ Showcase
|
| 54 |
+
|
| 55 |
+
📸 **Photorealistic Quality**: **Z-Image-Turbo** delivers strong photorealistic image generation while maintaining excellent aesthetic quality.
|
| 56 |
+
|
| 57 |
+
|
| 58 |
+
|
| 59 |
+
📖 **Accurate Bilingual Text Rendering**: **Z-Image-Turbo** excels at accurately rendering complex Chinese and English text.
|
| 60 |
+
|
| 61 |
+
|
| 62 |
+
|
| 63 |
+
💡 **Prompt Enhancing & Reasoning**: Prompt Enhancer empowers the model with reasoning capabilities, enabling it to transcend surface-level descriptions and tap into underlying world knowledge.
|
| 64 |
+
|
| 65 |
+
|
| 66 |
+
|
| 67 |
+
🧠 **Creative Image Editing**: **Z-Image-Edit** shows a strong understanding of bilingual editing instructions, enabling imaginative and flexible image transformations.
|
| 68 |
+
|
| 69 |
+
|
| 70 |
+
|
| 71 |
+
### 🏗️ Model Architecture
|
| 72 |
+
We adopt a **Scalable Single-Stream DiT** (S3-DiT) architecture. In this setup, text, visual semantic tokens, and image VAE tokens are concatenated at the sequence level to serve as a unified input stream, maximizing parameter efficiency compared to dual-stream approaches.
|
| 73 |
+
|
| 74 |
+
|
| 75 |
+
|
| 76 |
+
### 📈 Performance
|
| 77 |
+
According to the Elo-based Human Preference Evaluation (on [*Alibaba AI Arena*](https://aiarena.alibaba-inc.com/corpora/arena/leaderboard?arenaType=T2I)), Z-Image-Turbo shows highly competitive performance against other leading models, while achieving state-of-the-art results among open-source models.
|
| 78 |
+
|
| 79 |
+
<p align="center">
|
| 80 |
+
<a href="https://aiarena.alibaba-inc.com/corpora/arena/leaderboard?arenaType=T2I">
|
| 81 |
+
<img src="assets/leaderboard.png" alt="Z-Image Elo Rating on AI Arena"/><br />
|
| 82 |
+
<span style="font-size:1.05em; cursor:pointer; text-decoration:underline;"> Click to view the full leaderboard</span>
|
| 83 |
+
</a>
|
| 84 |
+
</p>
|
| 85 |
+
|
| 86 |
+
### 🚀 Quick Start
|
| 87 |
+
Install the latest version of diffusers, use the following command:
|
| 88 |
+
<details>
|
| 89 |
+
<summary><sup>Click here for details for why you need to install diffusers from source</sup></summary>
|
| 90 |
+
|
| 91 |
+
We have submitted two pull requests ([#12703](https://github.com/huggingface/diffusers/pull/12703) and [#12715](https://github.com/huggingface/diffusers/pull/12715)) to the 🤗 diffusers repository to add support for Z-Image. Both PRs have been merged into the latest official diffusers release.
|
| 92 |
+
Therefore, you need to install diffusers from source for the latest features and Z-Image support.
|
| 93 |
+
|
| 94 |
+
</details>
|
| 95 |
+
|
| 96 |
+
```bash
|
| 97 |
+
pip install git+https://github.com/huggingface/diffusers
|
| 98 |
+
```
|
| 99 |
+
|
| 100 |
+
```python
|
| 101 |
+
import torch
|
| 102 |
+
from diffusers import ZImagePipeline
|
| 103 |
+
|
| 104 |
+
# 1. Load the pipeline
|
| 105 |
+
# Use bfloat16 for optimal performance on supported GPUs
|
| 106 |
+
pipe = ZImagePipeline.from_pretrained(
|
| 107 |
+
"Tongyi-MAI/Z-Image-Turbo",
|
| 108 |
+
torch_dtype=torch.bfloat16,
|
| 109 |
+
low_cpu_mem_usage=False,
|
| 110 |
+
)
|
| 111 |
+
pipe.to("cuda")
|
| 112 |
+
|
| 113 |
+
# [Optional] Attention Backend
|
| 114 |
+
# Diffusers uses SDPA by default. Switch to Flash Attention for better efficiency if supported:
|
| 115 |
+
# pipe.transformer.set_attention_backend("flash") # Enable Flash-Attention-2
|
| 116 |
+
# pipe.transformer.set_attention_backend("_flash_3") # Enable Flash-Attention-3
|
| 117 |
+
|
| 118 |
+
# [Optional] Model Compilation
|
| 119 |
+
# Compiling the DiT model accelerates inference, but the first run will take longer to compile.
|
| 120 |
+
# pipe.transformer.compile()
|
| 121 |
+
|
| 122 |
+
# [Optional] CPU Offloading
|
| 123 |
+
# Enable CPU offloading for memory-constrained devices.
|
| 124 |
+
# pipe.enable_model_cpu_offload()
|
| 125 |
+
|
| 126 |
+
prompt = "Young Chinese woman in red Hanfu, intricate embroidery. Impeccable makeup, red floral forehead pattern. Elaborate high bun, golden phoenix headdress, red flowers, beads. Holds round folding fan with lady, trees, bird. Neon lightning-bolt lamp (⚡️), bright yellow glow, above extended left palm. Soft-lit outdoor night background, silhouetted tiered pagoda (西安大雁塔), blurred colorful distant lights."
|
| 127 |
+
|
| 128 |
+
# 2. Generate Image
|
| 129 |
+
image = pipe(
|
| 130 |
+
prompt=prompt,
|
| 131 |
+
height=1024,
|
| 132 |
+
width=1024,
|
| 133 |
+
num_inference_steps=9, # This actually results in 8 DiT forwards
|
| 134 |
+
guidance_scale=0.0, # Guidance should be 0 for the Turbo models
|
| 135 |
+
generator=torch.Generator("cuda").manual_seed(42),
|
| 136 |
+
).images[0]
|
| 137 |
+
|
| 138 |
+
image.save("example.png")
|
| 139 |
+
```
|
| 140 |
+
|
| 141 |
+
## 🔬 Decoupled-DMD: The Acceleration Magic Behind Z-Image
|
| 142 |
+
|
| 143 |
+
[](https://arxiv.org/abs/2511.22677)
|
| 144 |
+
|
| 145 |
+
Decoupled-DMD is the core few-step distillation algorithm that empowers the 8-step Z-Image model.
|
| 146 |
+
|
| 147 |
+
Our core insight in Decoupled-DMD is that the success of existing DMD (Distributaion Matching Distillation) methods is the result of two independent, collaborating mechanisms:
|
| 148 |
+
|
| 149 |
+
- **CFG Augmentation (CA)**: The primary **engine** 🚀 driving the distillation process, a factor largely overlooked in previous work.
|
| 150 |
+
- **Distribution Matching (DM)**: Acts more as a **regularizer** ⚖️, ensuring the stability and quality of the generated output.
|
| 151 |
+
|
| 152 |
+
By recognizing and decoupling these two mechanisms, we were able to study and optimize them in isolation. This ultimately motivated us to develop an improved distillation process that significantly enhances the performance of few-step generation.
|
| 153 |
+
|
| 154 |
+
|
| 155 |
+
|
| 156 |
+
## 🤖 DMDR: Fusing DMD with Reinforcement Learning
|
| 157 |
+
|
| 158 |
+
[](https://arxiv.org/abs/2511.13649)
|
| 159 |
+
|
| 160 |
+
Building upon the strong foundation of Decoupled-DMD, our 8-step Z-Image model has already demonstrated exceptional capabilities. To achieve further improvements in terms of semantic alignment, aesthetic quality, and structural coherence—while producing images with richer high-frequency details—we present **DMDR**.
|
| 161 |
+
|
| 162 |
+
Our core insight behind DMDR is that Reinforcement Learning (RL) and Distribution Matching Distillation (DMD) can be synergistically integrated during the post-training of few-step models. We demonstrate that:
|
| 163 |
+
|
| 164 |
+
- **RL Unlocks the Performance of DMD** 🚀
|
| 165 |
+
- **DMD Effectively Regularizes RL** ⚖️
|
| 166 |
+
|
| 167 |
+
|
| 168 |
+
|
| 169 |
+
## ⏬ Download
|
| 170 |
+
```bash
|
| 171 |
+
pip install -U huggingface_hub
|
| 172 |
+
HF_XET_HIGH_PERFORMANCE=1 hf download Tongyi-MAI/Z-Image-Turbo
|
| 173 |
+
```
|
| 174 |
+
|
| 175 |
+
## 📜 Citation
|
| 176 |
+
|
| 177 |
+
If you find our work useful in your research, please consider citing:
|
| 178 |
+
|
| 179 |
+
```bibtex
|
| 180 |
+
@article{team2025zimage,
|
| 181 |
+
title={Z-Image: An Efficient Image Generation Foundation Model with Single-Stream Diffusion Transformer},
|
| 182 |
+
author={Z-Image Team},
|
| 183 |
+
journal={arXiv preprint arXiv:2511.22699},
|
| 184 |
+
year={2025}
|
| 185 |
+
}
|
| 186 |
+
|
| 187 |
+
@article{liu2025decoupled,
|
| 188 |
+
title={Decoupled DMD: CFG Augmentation as the Spear, Distribution Matching as the Shield},
|
| 189 |
+
author={Dongyang Liu and Peng Gao and David Liu and Ruoyi Du and Zhen Li and Qilong Wu and Xin Jin and Sihan Cao and Shifeng Zhang and Hongsheng Li and Steven Hoi},
|
| 190 |
+
journal={arXiv preprint arXiv:2511.22677},
|
| 191 |
+
year={2025}
|
| 192 |
+
}
|
| 193 |
+
|
| 194 |
+
@article{jiang2025distribution,
|
| 195 |
+
title={Distribution Matching Distillation Meets Reinforcement Learning},
|
| 196 |
+
author={Jiang, Dengyang and Liu, Dongyang and Wang, Zanyi and Wu, Qilong and Jin, Xin and Liu, David and Li, Zhen and Wang, Mengmeng and Gao, Peng and Yang, Harry},
|
| 197 |
+
journal={arXiv preprint arXiv:2511.13649},
|
| 198 |
+
year={2025}
|
| 199 |
+
}
|
| 200 |
+
```
|
assets/DMDR.webp
ADDED
|
Git LFS Details
|
assets/Z-Image-Gallery.pdf
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6f9895b3246d2547bac74bbe0be975da500eaae93f2cad4248ad3281786b1ac6
|
| 3 |
+
size 15767436
|
assets/architecture.webp
ADDED
|
Git LFS Details
|
assets/decoupled-dmd.webp
ADDED
|
Git LFS Details
|
assets/leaderboard.png
ADDED
|
Git LFS Details
|
assets/leaderboard.webp
ADDED
|
assets/reasoning.png
ADDED
|
Git LFS Details
|
assets/showcase.jpg
ADDED
|
Git LFS Details
|
assets/showcase_editing.png
ADDED
|
Git LFS Details
|
assets/showcase_realistic.png
ADDED
|
Git LFS Details
|
assets/showcase_rendering.png
ADDED
|
Git LFS Details
|
custom_adapter/adapter_config.json
ADDED
|
@@ -0,0 +1,20 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"base_student_model": "Qwen/Qwen3-1.7B",
|
| 3 |
+
"teacher_model": "Tongyi-MAI/Z-Image-Turbo",
|
| 4 |
+
"zimage_source": "/content/hf_models_unzipped/Z-Image-Turbo",
|
| 5 |
+
"student_hidden_size": 2048,
|
| 6 |
+
"teacher_hidden_size": 2560,
|
| 7 |
+
"adapter_dim": 1024,
|
| 8 |
+
"adapter_heads": 8,
|
| 9 |
+
"adapter_blocks": 2,
|
| 10 |
+
"adapter_ff_mult": 4,
|
| 11 |
+
"adapter_dropout": 0.1,
|
| 12 |
+
"hs_tap_index": -2,
|
| 13 |
+
"layer_mapping": [
|
| 14 |
+
[
|
| 15 |
+
-2,
|
| 16 |
+
-2
|
| 17 |
+
]
|
| 18 |
+
],
|
| 19 |
+
"zimage_enable_thinking": true
|
| 20 |
+
}
|
custom_adapter/adapter_model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a6e198d801f05cd3d4b16232035b4603d48c957429564697340c9e2e8c73827f
|
| 3 |
+
size 128070720
|
model_index.json
ADDED
|
@@ -0,0 +1,24 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_class_name": "ZImagePipeline",
|
| 3 |
+
"_diffusers_version": "0.36.0.dev0",
|
| 4 |
+
"scheduler": [
|
| 5 |
+
"diffusers",
|
| 6 |
+
"FlowMatchEulerDiscreteScheduler"
|
| 7 |
+
],
|
| 8 |
+
"text_encoder": [
|
| 9 |
+
"transformers",
|
| 10 |
+
"Qwen3Model"
|
| 11 |
+
],
|
| 12 |
+
"tokenizer": [
|
| 13 |
+
"transformers",
|
| 14 |
+
"Qwen2Tokenizer"
|
| 15 |
+
],
|
| 16 |
+
"transformer": [
|
| 17 |
+
"diffusers",
|
| 18 |
+
"ZImageTransformer2DModel"
|
| 19 |
+
],
|
| 20 |
+
"vae": [
|
| 21 |
+
"diffusers",
|
| 22 |
+
"AutoencoderKL"
|
| 23 |
+
]
|
| 24 |
+
}
|
scheduler/scheduler_config.json
ADDED
|
@@ -0,0 +1,7 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_class_name": "FlowMatchEulerDiscreteScheduler",
|
| 3 |
+
"_diffusers_version": "0.36.0.dev0",
|
| 4 |
+
"num_train_timesteps": 1000,
|
| 5 |
+
"use_dynamic_shifting": false,
|
| 6 |
+
"shift": 3.0
|
| 7 |
+
}
|
student_adapter_build_meta.json
ADDED
|
@@ -0,0 +1,26 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"base_repo": "Tongyi-MAI/Z-Image-Turbo",
|
| 3 |
+
"student_source": "/content/drive/MyDrive/zimage_distillkit_contract_qwen17b/run_20260414_092303/phase2",
|
| 4 |
+
"adapter_source": "/content/drive/MyDrive/zimage_distillkit_contract_qwen17b/run_20260414_092303/phase2/custom_adapter",
|
| 5 |
+
"adapter_config": {
|
| 6 |
+
"base_student_model": "Qwen/Qwen3-1.7B",
|
| 7 |
+
"teacher_model": "Tongyi-MAI/Z-Image-Turbo",
|
| 8 |
+
"zimage_source": "/content/hf_models_unzipped/Z-Image-Turbo",
|
| 9 |
+
"student_hidden_size": 2048,
|
| 10 |
+
"teacher_hidden_size": 2560,
|
| 11 |
+
"adapter_dim": 1024,
|
| 12 |
+
"adapter_heads": 8,
|
| 13 |
+
"adapter_blocks": 2,
|
| 14 |
+
"adapter_ff_mult": 4,
|
| 15 |
+
"adapter_dropout": 0.1,
|
| 16 |
+
"hs_tap_index": -2,
|
| 17 |
+
"layer_mapping": [
|
| 18 |
+
[
|
| 19 |
+
-2,
|
| 20 |
+
-2
|
| 21 |
+
]
|
| 22 |
+
],
|
| 23 |
+
"zimage_enable_thinking": true
|
| 24 |
+
},
|
| 25 |
+
"created_at_unix": 1778152795.3211486
|
| 26 |
+
}
|
text_encoder/__init__.py
ADDED
|
@@ -0,0 +1,4 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
# Student Adapter Text Encoder — drop-in replacement for Qwen3 text encoder
|
| 2 |
+
# Supports ZImagePipeline via trust_remote_code=True auto_map.
|
| 3 |
+
from .configuration_student_adapter import StudentAdapterConfig
|
| 4 |
+
from .modeling_student_adapter import StudentAdapterTextEncoder
|
text_encoder/config.json
ADDED
|
@@ -0,0 +1,141 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"adapter_blocks": 2,
|
| 3 |
+
"adapter_dim": 1024,
|
| 4 |
+
"adapter_dropout": 0.1,
|
| 5 |
+
"adapter_ff_mult": 4,
|
| 6 |
+
"adapter_heads": 8,
|
| 7 |
+
"architectures": [
|
| 8 |
+
"StudentAdapterTextEncoder"
|
| 9 |
+
],
|
| 10 |
+
"dtype": "bfloat16",
|
| 11 |
+
"hs_tap_index": -2,
|
| 12 |
+
"model_type": "zimage_student_adapter",
|
| 13 |
+
"student_config_dict": {
|
| 14 |
+
"_name_or_path": "/content/drive/MyDrive/zimage_distillkit_contract_qwen17b/run_20260414_092303/phase2",
|
| 15 |
+
"add_cross_attention": false,
|
| 16 |
+
"architectures": [
|
| 17 |
+
"Qwen3ForCausalLM"
|
| 18 |
+
],
|
| 19 |
+
"attention_bias": false,
|
| 20 |
+
"attention_dropout": 0.0,
|
| 21 |
+
"bad_words_ids": null,
|
| 22 |
+
"begin_suppress_tokens": null,
|
| 23 |
+
"bos_token_id": null,
|
| 24 |
+
"chunk_size_feed_forward": 0,
|
| 25 |
+
"cross_attention_hidden_size": null,
|
| 26 |
+
"decoder_start_token_id": null,
|
| 27 |
+
"diversity_penalty": 0.0,
|
| 28 |
+
"do_sample": false,
|
| 29 |
+
"dtype": "bfloat16",
|
| 30 |
+
"early_stopping": false,
|
| 31 |
+
"encoder_no_repeat_ngram_size": 0,
|
| 32 |
+
"eos_token_id": 151645,
|
| 33 |
+
"exponential_decay_length_penalty": null,
|
| 34 |
+
"finetuning_task": null,
|
| 35 |
+
"forced_bos_token_id": null,
|
| 36 |
+
"forced_eos_token_id": null,
|
| 37 |
+
"head_dim": 128,
|
| 38 |
+
"hidden_act": "silu",
|
| 39 |
+
"hidden_size": 2048,
|
| 40 |
+
"id2label": {
|
| 41 |
+
"0": "LABEL_0",
|
| 42 |
+
"1": "LABEL_1"
|
| 43 |
+
},
|
| 44 |
+
"initializer_range": 0.02,
|
| 45 |
+
"intermediate_size": 6144,
|
| 46 |
+
"is_decoder": false,
|
| 47 |
+
"is_encoder_decoder": false,
|
| 48 |
+
"label2id": {
|
| 49 |
+
"LABEL_0": 0,
|
| 50 |
+
"LABEL_1": 1
|
| 51 |
+
},
|
| 52 |
+
"layer_types": [
|
| 53 |
+
"full_attention",
|
| 54 |
+
"full_attention",
|
| 55 |
+
"full_attention",
|
| 56 |
+
"full_attention",
|
| 57 |
+
"full_attention",
|
| 58 |
+
"full_attention",
|
| 59 |
+
"full_attention",
|
| 60 |
+
"full_attention",
|
| 61 |
+
"full_attention",
|
| 62 |
+
"full_attention",
|
| 63 |
+
"full_attention",
|
| 64 |
+
"full_attention",
|
| 65 |
+
"full_attention",
|
| 66 |
+
"full_attention",
|
| 67 |
+
"full_attention",
|
| 68 |
+
"full_attention",
|
| 69 |
+
"full_attention",
|
| 70 |
+
"full_attention",
|
| 71 |
+
"full_attention",
|
| 72 |
+
"full_attention",
|
| 73 |
+
"full_attention",
|
| 74 |
+
"full_attention",
|
| 75 |
+
"full_attention",
|
| 76 |
+
"full_attention",
|
| 77 |
+
"full_attention",
|
| 78 |
+
"full_attention",
|
| 79 |
+
"full_attention",
|
| 80 |
+
"full_attention"
|
| 81 |
+
],
|
| 82 |
+
"length_penalty": 1.0,
|
| 83 |
+
"max_length": 20,
|
| 84 |
+
"max_position_embeddings": 40960,
|
| 85 |
+
"max_window_layers": 28,
|
| 86 |
+
"min_length": 0,
|
| 87 |
+
"model_type": "qwen3",
|
| 88 |
+
"no_repeat_ngram_size": 0,
|
| 89 |
+
"num_attention_heads": 16,
|
| 90 |
+
"num_beam_groups": 1,
|
| 91 |
+
"num_beams": 1,
|
| 92 |
+
"num_hidden_layers": 28,
|
| 93 |
+
"num_key_value_heads": 8,
|
| 94 |
+
"num_return_sequences": 1,
|
| 95 |
+
"output_attentions": false,
|
| 96 |
+
"output_hidden_states": false,
|
| 97 |
+
"output_scores": false,
|
| 98 |
+
"pad_token_id": 151643,
|
| 99 |
+
"prefix": null,
|
| 100 |
+
"problem_type": null,
|
| 101 |
+
"pruned_heads": {},
|
| 102 |
+
"remove_invalid_values": false,
|
| 103 |
+
"repetition_penalty": 1.0,
|
| 104 |
+
"return_dict": true,
|
| 105 |
+
"return_dict_in_generate": false,
|
| 106 |
+
"rms_norm_eps": 1e-06,
|
| 107 |
+
"rope_parameters": {
|
| 108 |
+
"rope_theta": 1000000,
|
| 109 |
+
"rope_type": "default"
|
| 110 |
+
},
|
| 111 |
+
"rope_scaling": null,
|
| 112 |
+
"rope_theta": 10000.0,
|
| 113 |
+
"sep_token_id": null,
|
| 114 |
+
"sliding_window": null,
|
| 115 |
+
"suppress_tokens": null,
|
| 116 |
+
"task_specific_params": null,
|
| 117 |
+
"temperature": 1.0,
|
| 118 |
+
"tf_legacy_loss": false,
|
| 119 |
+
"tie_encoder_decoder": false,
|
| 120 |
+
"tie_word_embeddings": true,
|
| 121 |
+
"tokenizer_class": null,
|
| 122 |
+
"top_k": 50,
|
| 123 |
+
"top_p": 1.0,
|
| 124 |
+
"torchscript": false,
|
| 125 |
+
"transformers_version": "4.57.6",
|
| 126 |
+
"typical_p": 1.0,
|
| 127 |
+
"use_bfloat16": false,
|
| 128 |
+
"use_cache": false,
|
| 129 |
+
"use_sliding_window": false,
|
| 130 |
+
"vocab_size": 151936
|
| 131 |
+
},
|
| 132 |
+
"student_hidden_size": 2048,
|
| 133 |
+
"student_model_type": "qwen3",
|
| 134 |
+
"teacher_hidden_size": 2560,
|
| 135 |
+
"transformers_version": "4.57.6",
|
| 136 |
+
"auto_map": {
|
| 137 |
+
"AutoConfig": "configuration_student_adapter.StudentAdapterConfig",
|
| 138 |
+
"AutoModel": "modeling_student_adapter.StudentAdapterTextEncoder",
|
| 139 |
+
"AutoModelForCausalLM": "modeling_student_adapter.StudentAdapterTextEncoder"
|
| 140 |
+
}
|
| 141 |
+
}
|
text_encoder/configuration_student_adapter.py
ADDED
|
@@ -0,0 +1,30 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
from transformers import PretrainedConfig
|
| 2 |
+
|
| 3 |
+
class StudentAdapterConfig(PretrainedConfig):
|
| 4 |
+
model_type = "zimage_student_adapter"
|
| 5 |
+
|
| 6 |
+
def __init__(
|
| 7 |
+
self,
|
| 8 |
+
student_config_dict=None,
|
| 9 |
+
student_model_type=None,
|
| 10 |
+
hs_tap_index=-2,
|
| 11 |
+
adapter_dim=1024,
|
| 12 |
+
adapter_heads=8,
|
| 13 |
+
adapter_blocks=2,
|
| 14 |
+
adapter_ff_mult=4,
|
| 15 |
+
adapter_dropout=0.1,
|
| 16 |
+
teacher_hidden_size=None,
|
| 17 |
+
student_hidden_size=None,
|
| 18 |
+
**kwargs,
|
| 19 |
+
):
|
| 20 |
+
super().__init__(**kwargs)
|
| 21 |
+
self.student_config_dict = student_config_dict or {}
|
| 22 |
+
self.student_model_type = student_model_type
|
| 23 |
+
self.hs_tap_index = int(hs_tap_index)
|
| 24 |
+
self.adapter_dim = int(adapter_dim)
|
| 25 |
+
self.adapter_heads = int(adapter_heads)
|
| 26 |
+
self.adapter_blocks = int(adapter_blocks)
|
| 27 |
+
self.adapter_ff_mult = int(adapter_ff_mult)
|
| 28 |
+
self.adapter_dropout = float(adapter_dropout)
|
| 29 |
+
self.teacher_hidden_size = teacher_hidden_size
|
| 30 |
+
self.student_hidden_size = student_hidden_size
|
text_encoder/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1526ed816759565aa04aa5a33a847c6652ed6b765addeb5f7d8eaaf46d96ff15
|
| 3 |
+
size 3569259104
|
text_encoder/modeling_student_adapter.py
ADDED
|
@@ -0,0 +1,137 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
import torch
|
| 2 |
+
import torch.nn as nn
|
| 3 |
+
from transformers import AutoConfig, AutoModelForCausalLM, PreTrainedModel
|
| 4 |
+
from transformers.modeling_outputs import BaseModelOutputWithPast
|
| 5 |
+
from .configuration_student_adapter import StudentAdapterConfig
|
| 6 |
+
|
| 7 |
+
|
| 8 |
+
class XAttnBlock(nn.Module):
|
| 9 |
+
def __init__(self, dim, heads, ff_mult=4, dropout=0.1):
|
| 10 |
+
super().__init__()
|
| 11 |
+
self.norm_q = nn.LayerNorm(dim)
|
| 12 |
+
self.norm_kv = nn.LayerNorm(dim)
|
| 13 |
+
self.attn = nn.MultiheadAttention(dim, heads, dropout=dropout, batch_first=True)
|
| 14 |
+
self.norm_ff = nn.LayerNorm(dim)
|
| 15 |
+
self.ff = nn.Sequential(
|
| 16 |
+
nn.Linear(dim, dim * ff_mult),
|
| 17 |
+
nn.GELU(),
|
| 18 |
+
nn.Dropout(dropout),
|
| 19 |
+
nn.Linear(dim * ff_mult, dim),
|
| 20 |
+
nn.Dropout(dropout),
|
| 21 |
+
)
|
| 22 |
+
|
| 23 |
+
def forward(self, q, kv, key_padding_mask=None):
|
| 24 |
+
q = q + self.attn(
|
| 25 |
+
self.norm_q(q),
|
| 26 |
+
self.norm_kv(kv),
|
| 27 |
+
self.norm_kv(kv),
|
| 28 |
+
key_padding_mask=key_padding_mask,
|
| 29 |
+
need_weights=False,
|
| 30 |
+
)[0]
|
| 31 |
+
q = q + self.ff(self.norm_ff(q))
|
| 32 |
+
return q
|
| 33 |
+
|
| 34 |
+
|
| 35 |
+
class Adapter(nn.Module):
|
| 36 |
+
def __init__(self, s_dim, t_dim, dim=1024, heads=8, blocks=2, ff_mult=4, dropout=0.1):
|
| 37 |
+
super().__init__()
|
| 38 |
+
self.q_proj = nn.Linear(s_dim, dim)
|
| 39 |
+
self.kv_proj = nn.Linear(s_dim, dim)
|
| 40 |
+
self.blocks = nn.ModuleList([
|
| 41 |
+
XAttnBlock(dim, heads, ff_mult=ff_mult, dropout=dropout)
|
| 42 |
+
for _ in range(blocks)
|
| 43 |
+
])
|
| 44 |
+
self.proj_out = nn.Linear(dim, t_dim)
|
| 45 |
+
|
| 46 |
+
def forward(self, student_hs, mask):
|
| 47 |
+
q = self.q_proj(student_hs)
|
| 48 |
+
kv = self.kv_proj(student_hs)
|
| 49 |
+
key_padding_mask = ~mask.bool()
|
| 50 |
+
for block in self.blocks:
|
| 51 |
+
q = block(q, kv, key_padding_mask=key_padding_mask)
|
| 52 |
+
out = self.proj_out(q)
|
| 53 |
+
out = out.masked_fill(~mask[..., None].bool(), 0)
|
| 54 |
+
return out
|
| 55 |
+
|
| 56 |
+
|
| 57 |
+
class StudentAdapterTextEncoder(PreTrainedModel):
|
| 58 |
+
config_class = StudentAdapterConfig
|
| 59 |
+
base_model_prefix = "student"
|
| 60 |
+
|
| 61 |
+
def __init__(self, config: StudentAdapterConfig):
|
| 62 |
+
super().__init__(config)
|
| 63 |
+
student_cfg_dict = dict(config.student_config_dict or {})
|
| 64 |
+
if not student_cfg_dict:
|
| 65 |
+
raise ValueError("StudentAdapterConfig.student_config_dict is required")
|
| 66 |
+
|
| 67 |
+
model_type = student_cfg_dict.get("model_type") or config.student_model_type
|
| 68 |
+
if model_type is None:
|
| 69 |
+
raise ValueError("Missing student model_type")
|
| 70 |
+
|
| 71 |
+
cfg_kwargs = dict(student_cfg_dict)
|
| 72 |
+
cfg_kwargs.pop("model_type", None)
|
| 73 |
+
student_cfg = AutoConfig.for_model(model_type, **cfg_kwargs)
|
| 74 |
+
self.student = AutoModelForCausalLM.from_config(student_cfg, trust_remote_code=True)
|
| 75 |
+
|
| 76 |
+
s_dim = int(getattr(self.student.config, "hidden_size", config.student_hidden_size))
|
| 77 |
+
t_dim = int(config.teacher_hidden_size)
|
| 78 |
+
self.adapter = Adapter(
|
| 79 |
+
s_dim=s_dim,
|
| 80 |
+
t_dim=t_dim,
|
| 81 |
+
dim=config.adapter_dim,
|
| 82 |
+
heads=config.adapter_heads,
|
| 83 |
+
blocks=config.adapter_blocks,
|
| 84 |
+
ff_mult=config.adapter_ff_mult,
|
| 85 |
+
dropout=config.adapter_dropout,
|
| 86 |
+
)
|
| 87 |
+
self.hs_tap_index = int(config.hs_tap_index)
|
| 88 |
+
self.post_init()
|
| 89 |
+
|
| 90 |
+
def _extract_hs(self, outputs, idx: int):
|
| 91 |
+
hs = outputs.hidden_states
|
| 92 |
+
if hs is None:
|
| 93 |
+
raise RuntimeError("Student output_hidden_states is required")
|
| 94 |
+
if not (-len(hs) <= idx < len(hs)):
|
| 95 |
+
raise IndexError(f"hidden-state index {idx} out of range for len={len(hs)}")
|
| 96 |
+
return hs[idx]
|
| 97 |
+
|
| 98 |
+
def forward(self, input_ids=None, attention_mask=None, output_hidden_states=True, return_dict=True, **kwargs):
|
| 99 |
+
if input_ids is None:
|
| 100 |
+
raise ValueError("input_ids is required")
|
| 101 |
+
if attention_mask is None:
|
| 102 |
+
attention_mask = torch.ones_like(input_ids, dtype=torch.long)
|
| 103 |
+
# Qwen3 student model expects long dtype; pipeline may pass bool masks
|
| 104 |
+
if attention_mask.dtype == torch.bool:
|
| 105 |
+
attention_mask = attention_mask.long()
|
| 106 |
+
|
| 107 |
+
out = self.student(
|
| 108 |
+
input_ids=input_ids,
|
| 109 |
+
attention_mask=attention_mask,
|
| 110 |
+
output_hidden_states=True,
|
| 111 |
+
return_dict=True,
|
| 112 |
+
**kwargs,
|
| 113 |
+
)
|
| 114 |
+
|
| 115 |
+
hs_list = list(out.hidden_states)
|
| 116 |
+
s_hs = self._extract_hs(out, self.hs_tap_index)
|
| 117 |
+
|
| 118 |
+
ad_dtype = next(self.adapter.parameters()).dtype
|
| 119 |
+
if s_hs.dtype != ad_dtype:
|
| 120 |
+
s_hs = s_hs.to(ad_dtype)
|
| 121 |
+
|
| 122 |
+
adapted = self.adapter(s_hs, attention_mask)
|
| 123 |
+
|
| 124 |
+
if len(hs_list) >= 2:
|
| 125 |
+
hs_list[-2] = adapted
|
| 126 |
+
else:
|
| 127 |
+
hs_list.append(adapted)
|
| 128 |
+
|
| 129 |
+
if not return_dict:
|
| 130 |
+
return (adapted, None, tuple(hs_list), None)
|
| 131 |
+
|
| 132 |
+
return BaseModelOutputWithPast(
|
| 133 |
+
last_hidden_state=adapted,
|
| 134 |
+
past_key_values=None,
|
| 135 |
+
hidden_states=tuple(hs_list),
|
| 136 |
+
attentions=None,
|
| 137 |
+
)
|
tokenizer/merges.txt
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
tokenizer/tokenizer.json
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:aeb13307a71acd8fe81861d94ad54ab689df773318809eed3cbe794b4492dae4
|
| 3 |
+
size 11422654
|
tokenizer/tokenizer_config.json
ADDED
|
@@ -0,0 +1,239 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"add_bos_token": false,
|
| 3 |
+
"add_prefix_space": false,
|
| 4 |
+
"added_tokens_decoder": {
|
| 5 |
+
"151643": {
|
| 6 |
+
"content": "<|endoftext|>",
|
| 7 |
+
"lstrip": false,
|
| 8 |
+
"normalized": false,
|
| 9 |
+
"rstrip": false,
|
| 10 |
+
"single_word": false,
|
| 11 |
+
"special": true
|
| 12 |
+
},
|
| 13 |
+
"151644": {
|
| 14 |
+
"content": "<|im_start|>",
|
| 15 |
+
"lstrip": false,
|
| 16 |
+
"normalized": false,
|
| 17 |
+
"rstrip": false,
|
| 18 |
+
"single_word": false,
|
| 19 |
+
"special": true
|
| 20 |
+
},
|
| 21 |
+
"151645": {
|
| 22 |
+
"content": "<|im_end|>",
|
| 23 |
+
"lstrip": false,
|
| 24 |
+
"normalized": false,
|
| 25 |
+
"rstrip": false,
|
| 26 |
+
"single_word": false,
|
| 27 |
+
"special": true
|
| 28 |
+
},
|
| 29 |
+
"151646": {
|
| 30 |
+
"content": "<|object_ref_start|>",
|
| 31 |
+
"lstrip": false,
|
| 32 |
+
"normalized": false,
|
| 33 |
+
"rstrip": false,
|
| 34 |
+
"single_word": false,
|
| 35 |
+
"special": true
|
| 36 |
+
},
|
| 37 |
+
"151647": {
|
| 38 |
+
"content": "<|object_ref_end|>",
|
| 39 |
+
"lstrip": false,
|
| 40 |
+
"normalized": false,
|
| 41 |
+
"rstrip": false,
|
| 42 |
+
"single_word": false,
|
| 43 |
+
"special": true
|
| 44 |
+
},
|
| 45 |
+
"151648": {
|
| 46 |
+
"content": "<|box_start|>",
|
| 47 |
+
"lstrip": false,
|
| 48 |
+
"normalized": false,
|
| 49 |
+
"rstrip": false,
|
| 50 |
+
"single_word": false,
|
| 51 |
+
"special": true
|
| 52 |
+
},
|
| 53 |
+
"151649": {
|
| 54 |
+
"content": "<|box_end|>",
|
| 55 |
+
"lstrip": false,
|
| 56 |
+
"normalized": false,
|
| 57 |
+
"rstrip": false,
|
| 58 |
+
"single_word": false,
|
| 59 |
+
"special": true
|
| 60 |
+
},
|
| 61 |
+
"151650": {
|
| 62 |
+
"content": "<|quad_start|>",
|
| 63 |
+
"lstrip": false,
|
| 64 |
+
"normalized": false,
|
| 65 |
+
"rstrip": false,
|
| 66 |
+
"single_word": false,
|
| 67 |
+
"special": true
|
| 68 |
+
},
|
| 69 |
+
"151651": {
|
| 70 |
+
"content": "<|quad_end|>",
|
| 71 |
+
"lstrip": false,
|
| 72 |
+
"normalized": false,
|
| 73 |
+
"rstrip": false,
|
| 74 |
+
"single_word": false,
|
| 75 |
+
"special": true
|
| 76 |
+
},
|
| 77 |
+
"151652": {
|
| 78 |
+
"content": "<|vision_start|>",
|
| 79 |
+
"lstrip": false,
|
| 80 |
+
"normalized": false,
|
| 81 |
+
"rstrip": false,
|
| 82 |
+
"single_word": false,
|
| 83 |
+
"special": true
|
| 84 |
+
},
|
| 85 |
+
"151653": {
|
| 86 |
+
"content": "<|vision_end|>",
|
| 87 |
+
"lstrip": false,
|
| 88 |
+
"normalized": false,
|
| 89 |
+
"rstrip": false,
|
| 90 |
+
"single_word": false,
|
| 91 |
+
"special": true
|
| 92 |
+
},
|
| 93 |
+
"151654": {
|
| 94 |
+
"content": "<|vision_pad|>",
|
| 95 |
+
"lstrip": false,
|
| 96 |
+
"normalized": false,
|
| 97 |
+
"rstrip": false,
|
| 98 |
+
"single_word": false,
|
| 99 |
+
"special": true
|
| 100 |
+
},
|
| 101 |
+
"151655": {
|
| 102 |
+
"content": "<|image_pad|>",
|
| 103 |
+
"lstrip": false,
|
| 104 |
+
"normalized": false,
|
| 105 |
+
"rstrip": false,
|
| 106 |
+
"single_word": false,
|
| 107 |
+
"special": true
|
| 108 |
+
},
|
| 109 |
+
"151656": {
|
| 110 |
+
"content": "<|video_pad|>",
|
| 111 |
+
"lstrip": false,
|
| 112 |
+
"normalized": false,
|
| 113 |
+
"rstrip": false,
|
| 114 |
+
"single_word": false,
|
| 115 |
+
"special": true
|
| 116 |
+
},
|
| 117 |
+
"151657": {
|
| 118 |
+
"content": "<tool_call>",
|
| 119 |
+
"lstrip": false,
|
| 120 |
+
"normalized": false,
|
| 121 |
+
"rstrip": false,
|
| 122 |
+
"single_word": false,
|
| 123 |
+
"special": false
|
| 124 |
+
},
|
| 125 |
+
"151658": {
|
| 126 |
+
"content": "</tool_call>",
|
| 127 |
+
"lstrip": false,
|
| 128 |
+
"normalized": false,
|
| 129 |
+
"rstrip": false,
|
| 130 |
+
"single_word": false,
|
| 131 |
+
"special": false
|
| 132 |
+
},
|
| 133 |
+
"151659": {
|
| 134 |
+
"content": "<|fim_prefix|>",
|
| 135 |
+
"lstrip": false,
|
| 136 |
+
"normalized": false,
|
| 137 |
+
"rstrip": false,
|
| 138 |
+
"single_word": false,
|
| 139 |
+
"special": false
|
| 140 |
+
},
|
| 141 |
+
"151660": {
|
| 142 |
+
"content": "<|fim_middle|>",
|
| 143 |
+
"lstrip": false,
|
| 144 |
+
"normalized": false,
|
| 145 |
+
"rstrip": false,
|
| 146 |
+
"single_word": false,
|
| 147 |
+
"special": false
|
| 148 |
+
},
|
| 149 |
+
"151661": {
|
| 150 |
+
"content": "<|fim_suffix|>",
|
| 151 |
+
"lstrip": false,
|
| 152 |
+
"normalized": false,
|
| 153 |
+
"rstrip": false,
|
| 154 |
+
"single_word": false,
|
| 155 |
+
"special": false
|
| 156 |
+
},
|
| 157 |
+
"151662": {
|
| 158 |
+
"content": "<|fim_pad|>",
|
| 159 |
+
"lstrip": false,
|
| 160 |
+
"normalized": false,
|
| 161 |
+
"rstrip": false,
|
| 162 |
+
"single_word": false,
|
| 163 |
+
"special": false
|
| 164 |
+
},
|
| 165 |
+
"151663": {
|
| 166 |
+
"content": "<|repo_name|>",
|
| 167 |
+
"lstrip": false,
|
| 168 |
+
"normalized": false,
|
| 169 |
+
"rstrip": false,
|
| 170 |
+
"single_word": false,
|
| 171 |
+
"special": false
|
| 172 |
+
},
|
| 173 |
+
"151664": {
|
| 174 |
+
"content": "<|file_sep|>",
|
| 175 |
+
"lstrip": false,
|
| 176 |
+
"normalized": false,
|
| 177 |
+
"rstrip": false,
|
| 178 |
+
"single_word": false,
|
| 179 |
+
"special": false
|
| 180 |
+
},
|
| 181 |
+
"151665": {
|
| 182 |
+
"content": "<tool_response>",
|
| 183 |
+
"lstrip": false,
|
| 184 |
+
"normalized": false,
|
| 185 |
+
"rstrip": false,
|
| 186 |
+
"single_word": false,
|
| 187 |
+
"special": false
|
| 188 |
+
},
|
| 189 |
+
"151666": {
|
| 190 |
+
"content": "</tool_response>",
|
| 191 |
+
"lstrip": false,
|
| 192 |
+
"normalized": false,
|
| 193 |
+
"rstrip": false,
|
| 194 |
+
"single_word": false,
|
| 195 |
+
"special": false
|
| 196 |
+
},
|
| 197 |
+
"151667": {
|
| 198 |
+
"content": "<think>",
|
| 199 |
+
"lstrip": false,
|
| 200 |
+
"normalized": false,
|
| 201 |
+
"rstrip": false,
|
| 202 |
+
"single_word": false,
|
| 203 |
+
"special": false
|
| 204 |
+
},
|
| 205 |
+
"151668": {
|
| 206 |
+
"content": "</think>",
|
| 207 |
+
"lstrip": false,
|
| 208 |
+
"normalized": false,
|
| 209 |
+
"rstrip": false,
|
| 210 |
+
"single_word": false,
|
| 211 |
+
"special": false
|
| 212 |
+
}
|
| 213 |
+
},
|
| 214 |
+
"additional_special_tokens": [
|
| 215 |
+
"<|im_start|>",
|
| 216 |
+
"<|im_end|>",
|
| 217 |
+
"<|object_ref_start|>",
|
| 218 |
+
"<|object_ref_end|>",
|
| 219 |
+
"<|box_start|>",
|
| 220 |
+
"<|box_end|>",
|
| 221 |
+
"<|quad_start|>",
|
| 222 |
+
"<|quad_end|>",
|
| 223 |
+
"<|vision_start|>",
|
| 224 |
+
"<|vision_end|>",
|
| 225 |
+
"<|vision_pad|>",
|
| 226 |
+
"<|image_pad|>",
|
| 227 |
+
"<|video_pad|>"
|
| 228 |
+
],
|
| 229 |
+
"bos_token": null,
|
| 230 |
+
"chat_template": "{%- if tools %}\n {{- '<|im_start|>system\\n' }}\n {%- if messages[0].role == 'system' %}\n {{- messages[0].content + '\\n\\n' }}\n {%- endif %}\n {{- \"# Tools\\n\\nYou may call one or more functions to assist with the user query.\\n\\nYou are provided with function signatures within <tools></tools> XML tags:\\n<tools>\" }}\n {%- for tool in tools %}\n {{- \"\\n\" }}\n {{- tool | tojson }}\n {%- endfor %}\n {{- \"\\n</tools>\\n\\nFor each function call, return a json object with function name and arguments within <tool_call></tool_call> XML tags:\\n<tool_call>\\n{\\\"name\\\": <function-name>, \\\"arguments\\\": <args-json-object>}\\n</tool_call><|im_end|>\\n\" }}\n{%- else %}\n {%- if messages[0].role == 'system' %}\n {{- '<|im_start|>system\\n' + messages[0].content + '<|im_end|>\\n' }}\n {%- endif %}\n{%- endif %}\n{%- set ns = namespace(multi_step_tool=true, last_query_index=messages|length - 1) %}\n{%- for message in messages[::-1] %}\n {%- set index = (messages|length - 1) - loop.index0 %}\n {%- if ns.multi_step_tool and message.role == \"user\" and message.content is string and not(message.content.startswith('<tool_response>') and message.content.endswith('</tool_response>')) %}\n {%- set ns.multi_step_tool = false %}\n {%- set ns.last_query_index = index %}\n {%- endif %}\n{%- endfor %}\n{%- for message in messages %}\n {%- if message.content is string %}\n {%- set content = message.content %}\n {%- else %}\n {%- set content = '' %}\n {%- endif %}\n {%- if (message.role == \"user\") or (message.role == \"system\" and not loop.first) %}\n {{- '<|im_start|>' + message.role + '\\n' + content + '<|im_end|>' + '\\n' }}\n {%- elif message.role == \"assistant\" %}\n {%- set reasoning_content = '' %}\n {%- if message.reasoning_content is string %}\n {%- set reasoning_content = message.reasoning_content %}\n {%- else %}\n {%- if '</think>' in content %}\n {%- set reasoning_content = content.split('</think>')[0].rstrip('\\n').split('<think>')[-1].lstrip('\\n') %}\n {%- set content = content.split('</think>')[-1].lstrip('\\n') %}\n {%- endif %}\n {%- endif %}\n {%- if loop.index0 > ns.last_query_index %}\n {%- if loop.last or (not loop.last and reasoning_content) %}\n {{- '<|im_start|>' + message.role + '\\n<think>\\n' + reasoning_content.strip('\\n') + '\\n</think>\\n\\n' + content.lstrip('\\n') }}\n {%- else %}\n {{- '<|im_start|>' + message.role + '\\n' + content }}\n {%- endif %}\n {%- else %}\n {{- '<|im_start|>' + message.role + '\\n' + content }}\n {%- endif %}\n {%- if message.tool_calls %}\n {%- for tool_call in message.tool_calls %}\n {%- if (loop.first and content) or (not loop.first) %}\n {{- '\\n' }}\n {%- endif %}\n {%- if tool_call.function %}\n {%- set tool_call = tool_call.function %}\n {%- endif %}\n {{- '<tool_call>\\n{\"name\": \"' }}\n {{- tool_call.name }}\n {{- '\", \"arguments\": ' }}\n {%- if tool_call.arguments is string %}\n {{- tool_call.arguments }}\n {%- else %}\n {{- tool_call.arguments | tojson }}\n {%- endif %}\n {{- '}\\n</tool_call>' }}\n {%- endfor %}\n {%- endif %}\n {{- '<|im_end|>\\n' }}\n {%- elif message.role == \"tool\" %}\n {%- if loop.first or (messages[loop.index0 - 1].role != \"tool\") %}\n {{- '<|im_start|>user' }}\n {%- endif %}\n {{- '\\n<tool_response>\\n' }}\n {{- content }}\n {{- '\\n</tool_response>' }}\n {%- if loop.last or (messages[loop.index0 + 1].role != \"tool\") %}\n {{- '<|im_end|>\\n' }}\n {%- endif %}\n {%- endif %}\n{%- endfor %}\n{%- if add_generation_prompt %}\n {{- '<|im_start|>assistant\\n' }}\n {%- if enable_thinking is defined and enable_thinking is false %}\n {{- '<think>\\n\\n</think>\\n\\n' }}\n {%- endif %}\n{%- endif %}",
|
| 231 |
+
"clean_up_tokenization_spaces": false,
|
| 232 |
+
"eos_token": "<|im_end|>",
|
| 233 |
+
"errors": "replace",
|
| 234 |
+
"model_max_length": 131072,
|
| 235 |
+
"pad_token": "<|endoftext|>",
|
| 236 |
+
"split_special_tokens": false,
|
| 237 |
+
"tokenizer_class": "Qwen2Tokenizer",
|
| 238 |
+
"unk_token": null
|
| 239 |
+
}
|
tokenizer/vocab.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
transformer/config.json
ADDED
|
@@ -0,0 +1,31 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_class_name": "ZImageTransformer2DModel",
|
| 3 |
+
"_diffusers_version": "0.36.0.dev0",
|
| 4 |
+
"all_f_patch_size": [
|
| 5 |
+
1
|
| 6 |
+
],
|
| 7 |
+
"all_patch_size": [
|
| 8 |
+
2
|
| 9 |
+
],
|
| 10 |
+
"axes_dims": [
|
| 11 |
+
32,
|
| 12 |
+
48,
|
| 13 |
+
48
|
| 14 |
+
],
|
| 15 |
+
"axes_lens": [
|
| 16 |
+
1536,
|
| 17 |
+
512,
|
| 18 |
+
512
|
| 19 |
+
],
|
| 20 |
+
"cap_feat_dim": 2560,
|
| 21 |
+
"dim": 3840,
|
| 22 |
+
"in_channels": 16,
|
| 23 |
+
"n_heads": 30,
|
| 24 |
+
"n_kv_heads": 30,
|
| 25 |
+
"n_layers": 30,
|
| 26 |
+
"n_refiner_layers": 2,
|
| 27 |
+
"norm_eps": 1e-05,
|
| 28 |
+
"qk_norm": true,
|
| 29 |
+
"rope_theta": 256.0,
|
| 30 |
+
"t_scale": 1000.0
|
| 31 |
+
}
|
transformer/diffusion_pytorch_model-00001-of-00003.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:95facd593e2549e8252acb571c653d57f7ddb7f1060d4e81712f152555a88804
|
| 3 |
+
size 9973693184
|
transformer/diffusion_pytorch_model-00002-of-00003.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a4bbe43ee184a1fb5af4b412d27555f532893bdc3165b1149e304ed82b5d7015
|
| 3 |
+
size 9973714824
|
transformer/diffusion_pytorch_model-00003-of-00003.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:aba4e37a590e63210878160a718d916d80398f4e1f78ab6c9b2b2a00d92769fa
|
| 3 |
+
size 4672282880
|
transformer/diffusion_pytorch_model.safetensors.index.json
ADDED
|
@@ -0,0 +1,528 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"metadata": {
|
| 3 |
+
"total_size": 24619634944
|
| 4 |
+
},
|
| 5 |
+
"weight_map": {
|
| 6 |
+
"all_final_layer.2-1.adaLN_modulation.1.bias": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 7 |
+
"all_final_layer.2-1.adaLN_modulation.1.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 8 |
+
"all_final_layer.2-1.linear.bias": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 9 |
+
"all_final_layer.2-1.linear.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 10 |
+
"all_x_embedder.2-1.bias": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 11 |
+
"all_x_embedder.2-1.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 12 |
+
"cap_embedder.0.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 13 |
+
"cap_embedder.1.bias": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 14 |
+
"cap_embedder.1.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 15 |
+
"cap_pad_token": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 16 |
+
"context_refiner.0.attention.norm_k.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 17 |
+
"context_refiner.0.attention.norm_q.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 18 |
+
"context_refiner.0.attention.to_k.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 19 |
+
"context_refiner.0.attention.to_out.0.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 20 |
+
"context_refiner.0.attention.to_q.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 21 |
+
"context_refiner.0.attention.to_v.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 22 |
+
"context_refiner.0.attention_norm1.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 23 |
+
"context_refiner.0.attention_norm2.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 24 |
+
"context_refiner.0.feed_forward.w1.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 25 |
+
"context_refiner.0.feed_forward.w2.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 26 |
+
"context_refiner.0.feed_forward.w3.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 27 |
+
"context_refiner.0.ffn_norm1.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 28 |
+
"context_refiner.0.ffn_norm2.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 29 |
+
"context_refiner.1.attention.norm_k.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 30 |
+
"context_refiner.1.attention.norm_q.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 31 |
+
"context_refiner.1.attention.to_k.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 32 |
+
"context_refiner.1.attention.to_out.0.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 33 |
+
"context_refiner.1.attention.to_q.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 34 |
+
"context_refiner.1.attention.to_v.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 35 |
+
"context_refiner.1.attention_norm1.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 36 |
+
"context_refiner.1.attention_norm2.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 37 |
+
"context_refiner.1.feed_forward.w1.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 38 |
+
"context_refiner.1.feed_forward.w2.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 39 |
+
"context_refiner.1.feed_forward.w3.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 40 |
+
"context_refiner.1.ffn_norm1.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 41 |
+
"context_refiner.1.ffn_norm2.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 42 |
+
"layers.0.adaLN_modulation.0.bias": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 43 |
+
"layers.0.adaLN_modulation.0.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 44 |
+
"layers.0.attention.norm_k.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 45 |
+
"layers.0.attention.norm_q.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 46 |
+
"layers.0.attention.to_k.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 47 |
+
"layers.0.attention.to_out.0.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 48 |
+
"layers.0.attention.to_q.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 49 |
+
"layers.0.attention.to_v.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 50 |
+
"layers.0.attention_norm1.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 51 |
+
"layers.0.attention_norm2.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 52 |
+
"layers.0.feed_forward.w1.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 53 |
+
"layers.0.feed_forward.w2.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 54 |
+
"layers.0.feed_forward.w3.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 55 |
+
"layers.0.ffn_norm1.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 56 |
+
"layers.0.ffn_norm2.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 57 |
+
"layers.1.adaLN_modulation.0.bias": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 58 |
+
"layers.1.adaLN_modulation.0.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 59 |
+
"layers.1.attention.norm_k.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 60 |
+
"layers.1.attention.norm_q.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 61 |
+
"layers.1.attention.to_k.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 62 |
+
"layers.1.attention.to_out.0.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 63 |
+
"layers.1.attention.to_q.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 64 |
+
"layers.1.attention.to_v.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 65 |
+
"layers.1.attention_norm1.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 66 |
+
"layers.1.attention_norm2.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 67 |
+
"layers.1.feed_forward.w1.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 68 |
+
"layers.1.feed_forward.w2.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 69 |
+
"layers.1.feed_forward.w3.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 70 |
+
"layers.1.ffn_norm1.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 71 |
+
"layers.1.ffn_norm2.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 72 |
+
"layers.10.adaLN_modulation.0.bias": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 73 |
+
"layers.10.adaLN_modulation.0.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 74 |
+
"layers.10.attention.norm_k.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 75 |
+
"layers.10.attention.norm_q.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 76 |
+
"layers.10.attention.to_k.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 77 |
+
"layers.10.attention.to_out.0.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 78 |
+
"layers.10.attention.to_q.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 79 |
+
"layers.10.attention.to_v.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 80 |
+
"layers.10.attention_norm1.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 81 |
+
"layers.10.attention_norm2.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 82 |
+
"layers.10.feed_forward.w1.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 83 |
+
"layers.10.feed_forward.w2.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 84 |
+
"layers.10.feed_forward.w3.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 85 |
+
"layers.10.ffn_norm1.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 86 |
+
"layers.10.ffn_norm2.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 87 |
+
"layers.11.adaLN_modulation.0.bias": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 88 |
+
"layers.11.adaLN_modulation.0.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 89 |
+
"layers.11.attention.norm_k.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 90 |
+
"layers.11.attention.norm_q.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 91 |
+
"layers.11.attention.to_k.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 92 |
+
"layers.11.attention.to_out.0.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 93 |
+
"layers.11.attention.to_q.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 94 |
+
"layers.11.attention.to_v.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 95 |
+
"layers.11.attention_norm1.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 96 |
+
"layers.11.attention_norm2.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 97 |
+
"layers.11.feed_forward.w1.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 98 |
+
"layers.11.feed_forward.w2.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 99 |
+
"layers.11.feed_forward.w3.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 100 |
+
"layers.11.ffn_norm1.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 101 |
+
"layers.11.ffn_norm2.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 102 |
+
"layers.12.adaLN_modulation.0.bias": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 103 |
+
"layers.12.adaLN_modulation.0.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 104 |
+
"layers.12.attention.norm_k.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 105 |
+
"layers.12.attention.norm_q.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 106 |
+
"layers.12.attention.to_k.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 107 |
+
"layers.12.attention.to_out.0.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 108 |
+
"layers.12.attention.to_q.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 109 |
+
"layers.12.attention.to_v.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 110 |
+
"layers.12.attention_norm1.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 111 |
+
"layers.12.attention_norm2.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 112 |
+
"layers.12.feed_forward.w1.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 113 |
+
"layers.12.feed_forward.w2.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 114 |
+
"layers.12.feed_forward.w3.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 115 |
+
"layers.12.ffn_norm1.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 116 |
+
"layers.12.ffn_norm2.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 117 |
+
"layers.13.adaLN_modulation.0.bias": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 118 |
+
"layers.13.adaLN_modulation.0.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 119 |
+
"layers.13.attention.norm_k.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 120 |
+
"layers.13.attention.norm_q.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 121 |
+
"layers.13.attention.to_k.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 122 |
+
"layers.13.attention.to_out.0.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 123 |
+
"layers.13.attention.to_q.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 124 |
+
"layers.13.attention.to_v.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 125 |
+
"layers.13.attention_norm1.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 126 |
+
"layers.13.attention_norm2.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 127 |
+
"layers.13.feed_forward.w1.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 128 |
+
"layers.13.feed_forward.w2.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 129 |
+
"layers.13.feed_forward.w3.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 130 |
+
"layers.13.ffn_norm1.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 131 |
+
"layers.13.ffn_norm2.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 132 |
+
"layers.14.adaLN_modulation.0.bias": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 133 |
+
"layers.14.adaLN_modulation.0.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 134 |
+
"layers.14.attention.norm_k.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 135 |
+
"layers.14.attention.norm_q.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 136 |
+
"layers.14.attention.to_k.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 137 |
+
"layers.14.attention.to_out.0.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 138 |
+
"layers.14.attention.to_q.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 139 |
+
"layers.14.attention.to_v.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 140 |
+
"layers.14.attention_norm1.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 141 |
+
"layers.14.attention_norm2.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 142 |
+
"layers.14.feed_forward.w1.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 143 |
+
"layers.14.feed_forward.w2.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 144 |
+
"layers.14.feed_forward.w3.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 145 |
+
"layers.14.ffn_norm1.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 146 |
+
"layers.14.ffn_norm2.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 147 |
+
"layers.15.adaLN_modulation.0.bias": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 148 |
+
"layers.15.adaLN_modulation.0.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 149 |
+
"layers.15.attention.norm_k.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 150 |
+
"layers.15.attention.norm_q.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 151 |
+
"layers.15.attention.to_k.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 152 |
+
"layers.15.attention.to_out.0.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 153 |
+
"layers.15.attention.to_q.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 154 |
+
"layers.15.attention.to_v.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 155 |
+
"layers.15.attention_norm1.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 156 |
+
"layers.15.attention_norm2.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 157 |
+
"layers.15.feed_forward.w1.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 158 |
+
"layers.15.feed_forward.w2.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 159 |
+
"layers.15.feed_forward.w3.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 160 |
+
"layers.15.ffn_norm1.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 161 |
+
"layers.15.ffn_norm2.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 162 |
+
"layers.16.adaLN_modulation.0.bias": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 163 |
+
"layers.16.adaLN_modulation.0.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 164 |
+
"layers.16.attention.norm_k.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 165 |
+
"layers.16.attention.norm_q.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 166 |
+
"layers.16.attention.to_k.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 167 |
+
"layers.16.attention.to_out.0.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 168 |
+
"layers.16.attention.to_q.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 169 |
+
"layers.16.attention.to_v.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 170 |
+
"layers.16.attention_norm1.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 171 |
+
"layers.16.attention_norm2.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 172 |
+
"layers.16.feed_forward.w1.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 173 |
+
"layers.16.feed_forward.w2.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 174 |
+
"layers.16.feed_forward.w3.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 175 |
+
"layers.16.ffn_norm1.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 176 |
+
"layers.16.ffn_norm2.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 177 |
+
"layers.17.adaLN_modulation.0.bias": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 178 |
+
"layers.17.adaLN_modulation.0.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 179 |
+
"layers.17.attention.norm_k.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 180 |
+
"layers.17.attention.norm_q.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 181 |
+
"layers.17.attention.to_k.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 182 |
+
"layers.17.attention.to_out.0.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 183 |
+
"layers.17.attention.to_q.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 184 |
+
"layers.17.attention.to_v.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 185 |
+
"layers.17.attention_norm1.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 186 |
+
"layers.17.attention_norm2.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 187 |
+
"layers.17.feed_forward.w1.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 188 |
+
"layers.17.feed_forward.w2.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 189 |
+
"layers.17.feed_forward.w3.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 190 |
+
"layers.17.ffn_norm1.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 191 |
+
"layers.17.ffn_norm2.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 192 |
+
"layers.18.adaLN_modulation.0.bias": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 193 |
+
"layers.18.adaLN_modulation.0.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 194 |
+
"layers.18.attention.norm_k.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 195 |
+
"layers.18.attention.norm_q.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 196 |
+
"layers.18.attention.to_k.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 197 |
+
"layers.18.attention.to_out.0.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 198 |
+
"layers.18.attention.to_q.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 199 |
+
"layers.18.attention.to_v.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 200 |
+
"layers.18.attention_norm1.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 201 |
+
"layers.18.attention_norm2.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 202 |
+
"layers.18.feed_forward.w1.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 203 |
+
"layers.18.feed_forward.w2.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 204 |
+
"layers.18.feed_forward.w3.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 205 |
+
"layers.18.ffn_norm1.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 206 |
+
"layers.18.ffn_norm2.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 207 |
+
"layers.19.adaLN_modulation.0.bias": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 208 |
+
"layers.19.adaLN_modulation.0.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 209 |
+
"layers.19.attention.norm_k.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 210 |
+
"layers.19.attention.norm_q.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 211 |
+
"layers.19.attention.to_k.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 212 |
+
"layers.19.attention.to_out.0.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 213 |
+
"layers.19.attention.to_q.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 214 |
+
"layers.19.attention.to_v.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 215 |
+
"layers.19.attention_norm1.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 216 |
+
"layers.19.attention_norm2.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 217 |
+
"layers.19.feed_forward.w1.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 218 |
+
"layers.19.feed_forward.w2.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 219 |
+
"layers.19.feed_forward.w3.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 220 |
+
"layers.19.ffn_norm1.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 221 |
+
"layers.19.ffn_norm2.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 222 |
+
"layers.2.adaLN_modulation.0.bias": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 223 |
+
"layers.2.adaLN_modulation.0.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 224 |
+
"layers.2.attention.norm_k.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 225 |
+
"layers.2.attention.norm_q.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 226 |
+
"layers.2.attention.to_k.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 227 |
+
"layers.2.attention.to_out.0.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 228 |
+
"layers.2.attention.to_q.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 229 |
+
"layers.2.attention.to_v.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 230 |
+
"layers.2.attention_norm1.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 231 |
+
"layers.2.attention_norm2.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 232 |
+
"layers.2.feed_forward.w1.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 233 |
+
"layers.2.feed_forward.w2.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 234 |
+
"layers.2.feed_forward.w3.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 235 |
+
"layers.2.ffn_norm1.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 236 |
+
"layers.2.ffn_norm2.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 237 |
+
"layers.20.adaLN_modulation.0.bias": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 238 |
+
"layers.20.adaLN_modulation.0.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 239 |
+
"layers.20.attention.norm_k.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 240 |
+
"layers.20.attention.norm_q.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 241 |
+
"layers.20.attention.to_k.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 242 |
+
"layers.20.attention.to_out.0.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 243 |
+
"layers.20.attention.to_q.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 244 |
+
"layers.20.attention.to_v.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 245 |
+
"layers.20.attention_norm1.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 246 |
+
"layers.20.attention_norm2.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 247 |
+
"layers.20.feed_forward.w1.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 248 |
+
"layers.20.feed_forward.w2.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 249 |
+
"layers.20.feed_forward.w3.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 250 |
+
"layers.20.ffn_norm1.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 251 |
+
"layers.20.ffn_norm2.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 252 |
+
"layers.21.adaLN_modulation.0.bias": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 253 |
+
"layers.21.adaLN_modulation.0.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 254 |
+
"layers.21.attention.norm_k.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 255 |
+
"layers.21.attention.norm_q.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 256 |
+
"layers.21.attention.to_k.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 257 |
+
"layers.21.attention.to_out.0.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 258 |
+
"layers.21.attention.to_q.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 259 |
+
"layers.21.attention.to_v.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 260 |
+
"layers.21.attention_norm1.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 261 |
+
"layers.21.attention_norm2.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 262 |
+
"layers.21.feed_forward.w1.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 263 |
+
"layers.21.feed_forward.w2.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 264 |
+
"layers.21.feed_forward.w3.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 265 |
+
"layers.21.ffn_norm1.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 266 |
+
"layers.21.ffn_norm2.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 267 |
+
"layers.22.adaLN_modulation.0.bias": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 268 |
+
"layers.22.adaLN_modulation.0.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 269 |
+
"layers.22.attention.norm_k.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 270 |
+
"layers.22.attention.norm_q.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 271 |
+
"layers.22.attention.to_k.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 272 |
+
"layers.22.attention.to_out.0.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 273 |
+
"layers.22.attention.to_q.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 274 |
+
"layers.22.attention.to_v.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 275 |
+
"layers.22.attention_norm1.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 276 |
+
"layers.22.attention_norm2.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 277 |
+
"layers.22.feed_forward.w1.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 278 |
+
"layers.22.feed_forward.w2.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 279 |
+
"layers.22.feed_forward.w3.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 280 |
+
"layers.22.ffn_norm1.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 281 |
+
"layers.22.ffn_norm2.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 282 |
+
"layers.23.adaLN_modulation.0.bias": "diffusion_pytorch_model-00003-of-00003.safetensors",
|
| 283 |
+
"layers.23.adaLN_modulation.0.weight": "diffusion_pytorch_model-00003-of-00003.safetensors",
|
| 284 |
+
"layers.23.attention.norm_k.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 285 |
+
"layers.23.attention.norm_q.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 286 |
+
"layers.23.attention.to_k.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 287 |
+
"layers.23.attention.to_out.0.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 288 |
+
"layers.23.attention.to_q.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 289 |
+
"layers.23.attention.to_v.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 290 |
+
"layers.23.attention_norm1.weight": "diffusion_pytorch_model-00003-of-00003.safetensors",
|
| 291 |
+
"layers.23.attention_norm2.weight": "diffusion_pytorch_model-00003-of-00003.safetensors",
|
| 292 |
+
"layers.23.feed_forward.w1.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 293 |
+
"layers.23.feed_forward.w2.weight": "diffusion_pytorch_model-00003-of-00003.safetensors",
|
| 294 |
+
"layers.23.feed_forward.w3.weight": "diffusion_pytorch_model-00003-of-00003.safetensors",
|
| 295 |
+
"layers.23.ffn_norm1.weight": "diffusion_pytorch_model-00003-of-00003.safetensors",
|
| 296 |
+
"layers.23.ffn_norm2.weight": "diffusion_pytorch_model-00003-of-00003.safetensors",
|
| 297 |
+
"layers.24.adaLN_modulation.0.bias": "diffusion_pytorch_model-00003-of-00003.safetensors",
|
| 298 |
+
"layers.24.adaLN_modulation.0.weight": "diffusion_pytorch_model-00003-of-00003.safetensors",
|
| 299 |
+
"layers.24.attention.norm_k.weight": "diffusion_pytorch_model-00003-of-00003.safetensors",
|
| 300 |
+
"layers.24.attention.norm_q.weight": "diffusion_pytorch_model-00003-of-00003.safetensors",
|
| 301 |
+
"layers.24.attention.to_k.weight": "diffusion_pytorch_model-00003-of-00003.safetensors",
|
| 302 |
+
"layers.24.attention.to_out.0.weight": "diffusion_pytorch_model-00003-of-00003.safetensors",
|
| 303 |
+
"layers.24.attention.to_q.weight": "diffusion_pytorch_model-00003-of-00003.safetensors",
|
| 304 |
+
"layers.24.attention.to_v.weight": "diffusion_pytorch_model-00003-of-00003.safetensors",
|
| 305 |
+
"layers.24.attention_norm1.weight": "diffusion_pytorch_model-00003-of-00003.safetensors",
|
| 306 |
+
"layers.24.attention_norm2.weight": "diffusion_pytorch_model-00003-of-00003.safetensors",
|
| 307 |
+
"layers.24.feed_forward.w1.weight": "diffusion_pytorch_model-00003-of-00003.safetensors",
|
| 308 |
+
"layers.24.feed_forward.w2.weight": "diffusion_pytorch_model-00003-of-00003.safetensors",
|
| 309 |
+
"layers.24.feed_forward.w3.weight": "diffusion_pytorch_model-00003-of-00003.safetensors",
|
| 310 |
+
"layers.24.ffn_norm1.weight": "diffusion_pytorch_model-00003-of-00003.safetensors",
|
| 311 |
+
"layers.24.ffn_norm2.weight": "diffusion_pytorch_model-00003-of-00003.safetensors",
|
| 312 |
+
"layers.25.adaLN_modulation.0.bias": "diffusion_pytorch_model-00003-of-00003.safetensors",
|
| 313 |
+
"layers.25.adaLN_modulation.0.weight": "diffusion_pytorch_model-00003-of-00003.safetensors",
|
| 314 |
+
"layers.25.attention.norm_k.weight": "diffusion_pytorch_model-00003-of-00003.safetensors",
|
| 315 |
+
"layers.25.attention.norm_q.weight": "diffusion_pytorch_model-00003-of-00003.safetensors",
|
| 316 |
+
"layers.25.attention.to_k.weight": "diffusion_pytorch_model-00003-of-00003.safetensors",
|
| 317 |
+
"layers.25.attention.to_out.0.weight": "diffusion_pytorch_model-00003-of-00003.safetensors",
|
| 318 |
+
"layers.25.attention.to_q.weight": "diffusion_pytorch_model-00003-of-00003.safetensors",
|
| 319 |
+
"layers.25.attention.to_v.weight": "diffusion_pytorch_model-00003-of-00003.safetensors",
|
| 320 |
+
"layers.25.attention_norm1.weight": "diffusion_pytorch_model-00003-of-00003.safetensors",
|
| 321 |
+
"layers.25.attention_norm2.weight": "diffusion_pytorch_model-00003-of-00003.safetensors",
|
| 322 |
+
"layers.25.feed_forward.w1.weight": "diffusion_pytorch_model-00003-of-00003.safetensors",
|
| 323 |
+
"layers.25.feed_forward.w2.weight": "diffusion_pytorch_model-00003-of-00003.safetensors",
|
| 324 |
+
"layers.25.feed_forward.w3.weight": "diffusion_pytorch_model-00003-of-00003.safetensors",
|
| 325 |
+
"layers.25.ffn_norm1.weight": "diffusion_pytorch_model-00003-of-00003.safetensors",
|
| 326 |
+
"layers.25.ffn_norm2.weight": "diffusion_pytorch_model-00003-of-00003.safetensors",
|
| 327 |
+
"layers.26.adaLN_modulation.0.bias": "diffusion_pytorch_model-00003-of-00003.safetensors",
|
| 328 |
+
"layers.26.adaLN_modulation.0.weight": "diffusion_pytorch_model-00003-of-00003.safetensors",
|
| 329 |
+
"layers.26.attention.norm_k.weight": "diffusion_pytorch_model-00003-of-00003.safetensors",
|
| 330 |
+
"layers.26.attention.norm_q.weight": "diffusion_pytorch_model-00003-of-00003.safetensors",
|
| 331 |
+
"layers.26.attention.to_k.weight": "diffusion_pytorch_model-00003-of-00003.safetensors",
|
| 332 |
+
"layers.26.attention.to_out.0.weight": "diffusion_pytorch_model-00003-of-00003.safetensors",
|
| 333 |
+
"layers.26.attention.to_q.weight": "diffusion_pytorch_model-00003-of-00003.safetensors",
|
| 334 |
+
"layers.26.attention.to_v.weight": "diffusion_pytorch_model-00003-of-00003.safetensors",
|
| 335 |
+
"layers.26.attention_norm1.weight": "diffusion_pytorch_model-00003-of-00003.safetensors",
|
| 336 |
+
"layers.26.attention_norm2.weight": "diffusion_pytorch_model-00003-of-00003.safetensors",
|
| 337 |
+
"layers.26.feed_forward.w1.weight": "diffusion_pytorch_model-00003-of-00003.safetensors",
|
| 338 |
+
"layers.26.feed_forward.w2.weight": "diffusion_pytorch_model-00003-of-00003.safetensors",
|
| 339 |
+
"layers.26.feed_forward.w3.weight": "diffusion_pytorch_model-00003-of-00003.safetensors",
|
| 340 |
+
"layers.26.ffn_norm1.weight": "diffusion_pytorch_model-00003-of-00003.safetensors",
|
| 341 |
+
"layers.26.ffn_norm2.weight": "diffusion_pytorch_model-00003-of-00003.safetensors",
|
| 342 |
+
"layers.27.adaLN_modulation.0.bias": "diffusion_pytorch_model-00003-of-00003.safetensors",
|
| 343 |
+
"layers.27.adaLN_modulation.0.weight": "diffusion_pytorch_model-00003-of-00003.safetensors",
|
| 344 |
+
"layers.27.attention.norm_k.weight": "diffusion_pytorch_model-00003-of-00003.safetensors",
|
| 345 |
+
"layers.27.attention.norm_q.weight": "diffusion_pytorch_model-00003-of-00003.safetensors",
|
| 346 |
+
"layers.27.attention.to_k.weight": "diffusion_pytorch_model-00003-of-00003.safetensors",
|
| 347 |
+
"layers.27.attention.to_out.0.weight": "diffusion_pytorch_model-00003-of-00003.safetensors",
|
| 348 |
+
"layers.27.attention.to_q.weight": "diffusion_pytorch_model-00003-of-00003.safetensors",
|
| 349 |
+
"layers.27.attention.to_v.weight": "diffusion_pytorch_model-00003-of-00003.safetensors",
|
| 350 |
+
"layers.27.attention_norm1.weight": "diffusion_pytorch_model-00003-of-00003.safetensors",
|
| 351 |
+
"layers.27.attention_norm2.weight": "diffusion_pytorch_model-00003-of-00003.safetensors",
|
| 352 |
+
"layers.27.feed_forward.w1.weight": "diffusion_pytorch_model-00003-of-00003.safetensors",
|
| 353 |
+
"layers.27.feed_forward.w2.weight": "diffusion_pytorch_model-00003-of-00003.safetensors",
|
| 354 |
+
"layers.27.feed_forward.w3.weight": "diffusion_pytorch_model-00003-of-00003.safetensors",
|
| 355 |
+
"layers.27.ffn_norm1.weight": "diffusion_pytorch_model-00003-of-00003.safetensors",
|
| 356 |
+
"layers.27.ffn_norm2.weight": "diffusion_pytorch_model-00003-of-00003.safetensors",
|
| 357 |
+
"layers.28.adaLN_modulation.0.bias": "diffusion_pytorch_model-00003-of-00003.safetensors",
|
| 358 |
+
"layers.28.adaLN_modulation.0.weight": "diffusion_pytorch_model-00003-of-00003.safetensors",
|
| 359 |
+
"layers.28.attention.norm_k.weight": "diffusion_pytorch_model-00003-of-00003.safetensors",
|
| 360 |
+
"layers.28.attention.norm_q.weight": "diffusion_pytorch_model-00003-of-00003.safetensors",
|
| 361 |
+
"layers.28.attention.to_k.weight": "diffusion_pytorch_model-00003-of-00003.safetensors",
|
| 362 |
+
"layers.28.attention.to_out.0.weight": "diffusion_pytorch_model-00003-of-00003.safetensors",
|
| 363 |
+
"layers.28.attention.to_q.weight": "diffusion_pytorch_model-00003-of-00003.safetensors",
|
| 364 |
+
"layers.28.attention.to_v.weight": "diffusion_pytorch_model-00003-of-00003.safetensors",
|
| 365 |
+
"layers.28.attention_norm1.weight": "diffusion_pytorch_model-00003-of-00003.safetensors",
|
| 366 |
+
"layers.28.attention_norm2.weight": "diffusion_pytorch_model-00003-of-00003.safetensors",
|
| 367 |
+
"layers.28.feed_forward.w1.weight": "diffusion_pytorch_model-00003-of-00003.safetensors",
|
| 368 |
+
"layers.28.feed_forward.w2.weight": "diffusion_pytorch_model-00003-of-00003.safetensors",
|
| 369 |
+
"layers.28.feed_forward.w3.weight": "diffusion_pytorch_model-00003-of-00003.safetensors",
|
| 370 |
+
"layers.28.ffn_norm1.weight": "diffusion_pytorch_model-00003-of-00003.safetensors",
|
| 371 |
+
"layers.28.ffn_norm2.weight": "diffusion_pytorch_model-00003-of-00003.safetensors",
|
| 372 |
+
"layers.29.adaLN_modulation.0.bias": "diffusion_pytorch_model-00003-of-00003.safetensors",
|
| 373 |
+
"layers.29.adaLN_modulation.0.weight": "diffusion_pytorch_model-00003-of-00003.safetensors",
|
| 374 |
+
"layers.29.attention.norm_k.weight": "diffusion_pytorch_model-00003-of-00003.safetensors",
|
| 375 |
+
"layers.29.attention.norm_q.weight": "diffusion_pytorch_model-00003-of-00003.safetensors",
|
| 376 |
+
"layers.29.attention.to_k.weight": "diffusion_pytorch_model-00003-of-00003.safetensors",
|
| 377 |
+
"layers.29.attention.to_out.0.weight": "diffusion_pytorch_model-00003-of-00003.safetensors",
|
| 378 |
+
"layers.29.attention.to_q.weight": "diffusion_pytorch_model-00003-of-00003.safetensors",
|
| 379 |
+
"layers.29.attention.to_v.weight": "diffusion_pytorch_model-00003-of-00003.safetensors",
|
| 380 |
+
"layers.29.attention_norm1.weight": "diffusion_pytorch_model-00003-of-00003.safetensors",
|
| 381 |
+
"layers.29.attention_norm2.weight": "diffusion_pytorch_model-00003-of-00003.safetensors",
|
| 382 |
+
"layers.29.feed_forward.w1.weight": "diffusion_pytorch_model-00003-of-00003.safetensors",
|
| 383 |
+
"layers.29.feed_forward.w2.weight": "diffusion_pytorch_model-00003-of-00003.safetensors",
|
| 384 |
+
"layers.29.feed_forward.w3.weight": "diffusion_pytorch_model-00003-of-00003.safetensors",
|
| 385 |
+
"layers.29.ffn_norm1.weight": "diffusion_pytorch_model-00003-of-00003.safetensors",
|
| 386 |
+
"layers.29.ffn_norm2.weight": "diffusion_pytorch_model-00003-of-00003.safetensors",
|
| 387 |
+
"layers.3.adaLN_modulation.0.bias": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 388 |
+
"layers.3.adaLN_modulation.0.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 389 |
+
"layers.3.attention.norm_k.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 390 |
+
"layers.3.attention.norm_q.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 391 |
+
"layers.3.attention.to_k.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 392 |
+
"layers.3.attention.to_out.0.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 393 |
+
"layers.3.attention.to_q.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 394 |
+
"layers.3.attention.to_v.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 395 |
+
"layers.3.attention_norm1.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 396 |
+
"layers.3.attention_norm2.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 397 |
+
"layers.3.feed_forward.w1.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 398 |
+
"layers.3.feed_forward.w2.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 399 |
+
"layers.3.feed_forward.w3.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 400 |
+
"layers.3.ffn_norm1.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 401 |
+
"layers.3.ffn_norm2.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 402 |
+
"layers.4.adaLN_modulation.0.bias": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 403 |
+
"layers.4.adaLN_modulation.0.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 404 |
+
"layers.4.attention.norm_k.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 405 |
+
"layers.4.attention.norm_q.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 406 |
+
"layers.4.attention.to_k.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 407 |
+
"layers.4.attention.to_out.0.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 408 |
+
"layers.4.attention.to_q.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 409 |
+
"layers.4.attention.to_v.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 410 |
+
"layers.4.attention_norm1.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 411 |
+
"layers.4.attention_norm2.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 412 |
+
"layers.4.feed_forward.w1.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 413 |
+
"layers.4.feed_forward.w2.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 414 |
+
"layers.4.feed_forward.w3.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 415 |
+
"layers.4.ffn_norm1.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 416 |
+
"layers.4.ffn_norm2.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 417 |
+
"layers.5.adaLN_modulation.0.bias": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 418 |
+
"layers.5.adaLN_modulation.0.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 419 |
+
"layers.5.attention.norm_k.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 420 |
+
"layers.5.attention.norm_q.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 421 |
+
"layers.5.attention.to_k.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 422 |
+
"layers.5.attention.to_out.0.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 423 |
+
"layers.5.attention.to_q.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 424 |
+
"layers.5.attention.to_v.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 425 |
+
"layers.5.attention_norm1.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 426 |
+
"layers.5.attention_norm2.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 427 |
+
"layers.5.feed_forward.w1.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 428 |
+
"layers.5.feed_forward.w2.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 429 |
+
"layers.5.feed_forward.w3.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 430 |
+
"layers.5.ffn_norm1.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 431 |
+
"layers.5.ffn_norm2.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 432 |
+
"layers.6.adaLN_modulation.0.bias": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 433 |
+
"layers.6.adaLN_modulation.0.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 434 |
+
"layers.6.attention.norm_k.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 435 |
+
"layers.6.attention.norm_q.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 436 |
+
"layers.6.attention.to_k.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 437 |
+
"layers.6.attention.to_out.0.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 438 |
+
"layers.6.attention.to_q.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 439 |
+
"layers.6.attention.to_v.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 440 |
+
"layers.6.attention_norm1.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 441 |
+
"layers.6.attention_norm2.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 442 |
+
"layers.6.feed_forward.w1.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 443 |
+
"layers.6.feed_forward.w2.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 444 |
+
"layers.6.feed_forward.w3.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 445 |
+
"layers.6.ffn_norm1.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 446 |
+
"layers.6.ffn_norm2.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 447 |
+
"layers.7.adaLN_modulation.0.bias": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 448 |
+
"layers.7.adaLN_modulation.0.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 449 |
+
"layers.7.attention.norm_k.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 450 |
+
"layers.7.attention.norm_q.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 451 |
+
"layers.7.attention.to_k.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 452 |
+
"layers.7.attention.to_out.0.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 453 |
+
"layers.7.attention.to_q.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 454 |
+
"layers.7.attention.to_v.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 455 |
+
"layers.7.attention_norm1.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 456 |
+
"layers.7.attention_norm2.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 457 |
+
"layers.7.feed_forward.w1.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 458 |
+
"layers.7.feed_forward.w2.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 459 |
+
"layers.7.feed_forward.w3.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 460 |
+
"layers.7.ffn_norm1.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 461 |
+
"layers.7.ffn_norm2.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 462 |
+
"layers.8.adaLN_modulation.0.bias": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 463 |
+
"layers.8.adaLN_modulation.0.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 464 |
+
"layers.8.attention.norm_k.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 465 |
+
"layers.8.attention.norm_q.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 466 |
+
"layers.8.attention.to_k.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 467 |
+
"layers.8.attention.to_out.0.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 468 |
+
"layers.8.attention.to_q.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 469 |
+
"layers.8.attention.to_v.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 470 |
+
"layers.8.attention_norm1.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 471 |
+
"layers.8.attention_norm2.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 472 |
+
"layers.8.feed_forward.w1.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 473 |
+
"layers.8.feed_forward.w2.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 474 |
+
"layers.8.feed_forward.w3.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 475 |
+
"layers.8.ffn_norm1.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 476 |
+
"layers.8.ffn_norm2.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 477 |
+
"layers.9.adaLN_modulation.0.bias": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 478 |
+
"layers.9.adaLN_modulation.0.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 479 |
+
"layers.9.attention.norm_k.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 480 |
+
"layers.9.attention.norm_q.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 481 |
+
"layers.9.attention.to_k.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 482 |
+
"layers.9.attention.to_out.0.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 483 |
+
"layers.9.attention.to_q.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 484 |
+
"layers.9.attention.to_v.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 485 |
+
"layers.9.attention_norm1.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 486 |
+
"layers.9.attention_norm2.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 487 |
+
"layers.9.feed_forward.w1.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 488 |
+
"layers.9.feed_forward.w2.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 489 |
+
"layers.9.feed_forward.w3.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 490 |
+
"layers.9.ffn_norm1.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 491 |
+
"layers.9.ffn_norm2.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 492 |
+
"noise_refiner.0.adaLN_modulation.0.bias": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 493 |
+
"noise_refiner.0.adaLN_modulation.0.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 494 |
+
"noise_refiner.0.attention.norm_k.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 495 |
+
"noise_refiner.0.attention.norm_q.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 496 |
+
"noise_refiner.0.attention.to_k.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 497 |
+
"noise_refiner.0.attention.to_out.0.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 498 |
+
"noise_refiner.0.attention.to_q.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 499 |
+
"noise_refiner.0.attention.to_v.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 500 |
+
"noise_refiner.0.attention_norm1.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 501 |
+
"noise_refiner.0.attention_norm2.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 502 |
+
"noise_refiner.0.feed_forward.w1.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 503 |
+
"noise_refiner.0.feed_forward.w2.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 504 |
+
"noise_refiner.0.feed_forward.w3.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 505 |
+
"noise_refiner.0.ffn_norm1.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 506 |
+
"noise_refiner.0.ffn_norm2.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 507 |
+
"noise_refiner.1.adaLN_modulation.0.bias": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 508 |
+
"noise_refiner.1.adaLN_modulation.0.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 509 |
+
"noise_refiner.1.attention.norm_k.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 510 |
+
"noise_refiner.1.attention.norm_q.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 511 |
+
"noise_refiner.1.attention.to_k.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 512 |
+
"noise_refiner.1.attention.to_out.0.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 513 |
+
"noise_refiner.1.attention.to_q.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 514 |
+
"noise_refiner.1.attention.to_v.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 515 |
+
"noise_refiner.1.attention_norm1.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 516 |
+
"noise_refiner.1.attention_norm2.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 517 |
+
"noise_refiner.1.feed_forward.w1.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 518 |
+
"noise_refiner.1.feed_forward.w2.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 519 |
+
"noise_refiner.1.feed_forward.w3.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 520 |
+
"noise_refiner.1.ffn_norm1.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 521 |
+
"noise_refiner.1.ffn_norm2.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 522 |
+
"t_embedder.mlp.0.bias": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 523 |
+
"t_embedder.mlp.0.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 524 |
+
"t_embedder.mlp.2.bias": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 525 |
+
"t_embedder.mlp.2.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 526 |
+
"x_pad_token": "diffusion_pytorch_model-00001-of-00003.safetensors"
|
| 527 |
+
}
|
| 528 |
+
}
|
vae/config.json
ADDED
|
@@ -0,0 +1,38 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_class_name": "AutoencoderKL",
|
| 3 |
+
"_diffusers_version": "0.36.0.dev0",
|
| 4 |
+
"_name_or_path": "flux-dev",
|
| 5 |
+
"act_fn": "silu",
|
| 6 |
+
"block_out_channels": [
|
| 7 |
+
128,
|
| 8 |
+
256,
|
| 9 |
+
512,
|
| 10 |
+
512
|
| 11 |
+
],
|
| 12 |
+
"down_block_types": [
|
| 13 |
+
"DownEncoderBlock2D",
|
| 14 |
+
"DownEncoderBlock2D",
|
| 15 |
+
"DownEncoderBlock2D",
|
| 16 |
+
"DownEncoderBlock2D"
|
| 17 |
+
],
|
| 18 |
+
"force_upcast": true,
|
| 19 |
+
"in_channels": 3,
|
| 20 |
+
"latent_channels": 16,
|
| 21 |
+
"latents_mean": null,
|
| 22 |
+
"latents_std": null,
|
| 23 |
+
"layers_per_block": 2,
|
| 24 |
+
"mid_block_add_attention": true,
|
| 25 |
+
"norm_num_groups": 32,
|
| 26 |
+
"out_channels": 3,
|
| 27 |
+
"sample_size": 1024,
|
| 28 |
+
"scaling_factor": 0.3611,
|
| 29 |
+
"shift_factor": 0.1159,
|
| 30 |
+
"up_block_types": [
|
| 31 |
+
"UpDecoderBlock2D",
|
| 32 |
+
"UpDecoderBlock2D",
|
| 33 |
+
"UpDecoderBlock2D",
|
| 34 |
+
"UpDecoderBlock2D"
|
| 35 |
+
],
|
| 36 |
+
"use_post_quant_conv": false,
|
| 37 |
+
"use_quant_conv": false
|
| 38 |
+
}
|
vae/diffusion_pytorch_model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f5b59a26851551b67ae1fe58d32e76486e1e812def4696a4bea97f16604d40a3
|
| 3 |
+
size 167666902
|