Compare commits

14 Commits
v1.0 ... master

Author SHA1 Message Date
263c1a9cc4 Update README.md 2025-02-08 05:46:51 +00:00
cffca352f9 Update README.md 2025-02-08 05:41:27 +00:00
29e820ce55 Update README.md 2025-02-08 05:41:01 +00:00
41eb07dbff cover images 2025-01-22 06:11:46 +00:00
a5d1015ef9 cover images 2025-01-22 06:11:36 +00:00
9e4b0888b0 cover images 2025-01-22 06:11:22 +00:00
a0cb3b6de3 cover images 2025-01-22 06:11:04 +00:00
8339047e58 cover images 2025-01-22 06:10:52 +00:00
a5e99f37f9 cover images 2025-01-22 06:10:38 +00:00
335e37d4f0 Update README.md 2025-01-22 06:07:41 +00:00
487cf4d578 Update README.md 2025-01-22 06:07:38 +00:00
126cc2cbe8 Update README.md 2025-01-22 06:06:26 +00:00
19b36b07fb update 2025-01-15 12:25:22 +08:00
12760301c6 封面图 2025-01-06 14:09:43 +00:00
9 changed files with 123 additions and 40 deletions

163
README.md
View File

@ -1,7 +1,4 @@
--- ---
base_model: AI-ModelScope/FLUX.1-dev
cover_images:
- _cover_images_/cover.png
frameworks: frameworks:
- Pytorch - Pytorch
license: Apache License 2.0 license: Apache License 2.0
@ -12,47 +9,133 @@ tasks:
- text-to-image-synthesis - text-to-image-synthesis
vision_foundation: FLUX_1 vision_foundation: FLUX_1
#model-type: base_model:
##如 gpt、phi、llama、chatglm、baichuan 等 - black-forest-labs/FLUX.1-dev
#- gpt
#domain:
##如 nlp、cv、audio、multi-modal
#- nlp
#language:
##语言代码列表 https://help.aliyun.com/document_detail/215387.html?spm=a2c4g.11186623.0.0.9f8d7467kni6Aa
#- cn
#metrics:
##如 CIDEr、Blue、ROUGE 等
#- CIDEr
#tags:
##各种自定义,包括 pretrained、fine-tuned、instruction-tuned、RL-tuned 等训练方法和其他
#- pretrained
#tools:
##如 vllm、fastchat、llamacpp、AdaSeq 等
#- vllm
--- ---
### 当前模型的贡献者未提供更加详细的模型介绍。模型文件和权重,可浏览“模型文件”页面获取。
#### 您可以通过如下git clone命令或者ModelScope SDK来下载模型
SDK下载 麦橘超然 MajicFlus 是一款基于 [flux.dev](https://www.modelscope.cn/models/black-forest-labs/FLUX.1-dev) 微调融合的模型,专注于高质量人像生成,尤其擅长表现 亚洲女性 的细腻与美感。模型以 **唯美、写实、易用** 为核心特色,能够通过简单的提示词生成优质效果,同时对复杂提示词也有出色的响应能力。
```bash
#安装ModelScope ## 模型特点
pip install modelscope - 卓越的人像生成能力: 优化了在不同光影条件下的表现,确保人像在各种构图中的 面部细节 和 肢体完整性。
- 广泛的适用性: 除了人像生成外,模型在生成 非人生物 和 场景 时也有显著改进,适应更多创作需求。
- 简单易用: 用户无需复杂的提示词即可生成高质量作品,同时支持更长提示词的精细控制。
## 社区适配
MajicFlus 模型在发布的同时,多位社区成员基于模型制作的 LoRA 也将一同发布,进一步扩展了模型的功能与表现力。这些 LoRA 为用户提供了更多样化的创作可能性,使模型能够适应更多特定场景和风格需求。
## 弱点
- MajicFlus 并非为生成 NSFW 内容而设计。然而,如果有需要,可以使用相关 LoRA 来实现此类目的。
- MajicFlus 的存在是为了解决国际社区中模型缺乏亚洲代表性的问题。如果您希望生成非亚洲种族的图像,请考虑使用其他高质量模型。
- 由于该模型是个微调融合模型对社区大部分的lora都是不完美兼容的需要降低权重至0.5以下。推荐使用带有majicFlus标志的矩阵模型搜索关键字majic并筛选f1模型就可以看到全部现在已有超过50款风格各异的优质模型。
## 参数推荐 Parameter
- Steps: 20~30
- Distilled CFG Scale: 3.5
- CFG : 1
- Diffusion in Low Bits: float8-e4m3fn
- Sampling: Euler + simple/beta (for general)DPM2M + SGM uniform (for skin texture)DEIS + DDIM uniform (for casual realistic look)
- Vae: flux vae
- Clip: clip_l.safetensors and t5xxl_fp8_e4m3fn.safetensors
## 生图
本模型可通过[AIGC专区生图](https://www.modelscope.cn/aigc/imageGeneration?tab=advanced&versionId=14497&modelType=Checkpoint&sdVersion=FLUX_1&modelUrl=modelscope%3A%2F%2FMAILAND%2Fmajicflus_v1%3Frevision%3Dv1.0)直接在线使用。也通过ModelScope API-Inference使用见本页面右侧
如果要下载模型到本地进行推理生图需要结合原始FLUX模型的vae等模块可以使用DiffSynth-Studio封装好的生图pipeline。
### 安装DiffSynth
``` ```
pip install diffsynth -U
```
### 推理生图
#### 量化推理(需要至少 14G 显存)
推荐方式,能在较小显存下,实现无损生图。需要较大的内存来支持模型不同模块的逐个加载。
```python ```python
#SDK模型下载 import torch
from modelscope import snapshot_download from modelscope import snapshot_download
model_dir = snapshot_download('merjic/majicflus_v1') from diffsynth import ModelManager, FluxImagePipeline
```
Git下载 # 下载模型
``` snapshot_download(
#Git模型下载 model_id="MAILAND/majicflus_v1",
git clone https://www.modelscope.cn/merjic/majicflus_v1.git allow_file_pattern="majicflus_v134.safetensors",
cache_dir="models"
)
snapshot_download(
model_id="black-forest-labs/FLUX.1-dev",
allow_file_pattern=["ae.safetensors", "text_encoder/model.safetensors", "text_encoder_2/*"],
cache_dir="models"
)
# 设置推理计算精度为 bfloat16
model_manager = ModelManager(torch_dtype=torch.bfloat16)
# 以 float8 精度加载 DiT 部分
model_manager.load_models(
["models/MAILAND/majicflus_v1/majicflus_v134.safetensors"],
torch_dtype=torch.float8_e4m3fn,
device="cpu"
)
# 以 bfloat16 精度加载两个 Text Encoder 和 VAE
model_manager.load_models(
[
"models/black-forest-labs/FLUX.1-dev/text_encoder/model.safetensors",
"models/black-forest-labs/FLUX.1-dev/text_encoder_2",
"models/black-forest-labs/FLUX.1-dev/ae.safetensors",
],
torch_dtype=torch.bfloat16,
device="cpu"
)
# 开启量化与显存管理
pipe = FluxImagePipeline.from_model_manager(model_manager, device="cuda")
pipe.enable_cpu_offload()
pipe.dit.quantize()
# 生图!
image = pipe(prompt="a beautiful girl", seed=0)
image.save("image.jpg")
``` ```
<p style="color: lightgrey;">如果您是本模型的贡献者,我们邀请您根据<a href="https://modelscope.cn/docs/ModelScope%E6%A8%A1%E5%9E%8B%E6%8E%A5%E5%85%A5%E6%B5%81%E7%A8%8B%E6%A6%82%E8%A7%88" style="color: lightgrey; text-decoration: underline;">模型贡献文档</a>,及时完善模型卡片内容。</p> #### 原生精度推理(需要至少 40G 显存)
```python
import torch
from modelscope import snapshot_download
from diffsynth import ModelManager, FluxImagePipeline
# 下载模型
snapshot_download(
model_id="MAILAND/majicflus_v1",
allow_file_pattern="majicflus_v134.safetensors",
cache_dir="models"
)
snapshot_download(
model_id="black-forest-labs/FLUX.1-dev",
allow_file_pattern=["ae.safetensors", "text_encoder/model.safetensors", "text_encoder_2/*"],
cache_dir="models"
)
# 加载模型
model_manager = ModelManager(
torch_dtype=torch.bfloat16,
device="cuda",
file_path_list=[
"models/black-forest-labs/FLUX.1-dev/text_encoder/model.safetensors",
"models/black-forest-labs/FLUX.1-dev/text_encoder_2",
"models/black-forest-labs/FLUX.1-dev/ae.safetensors",
"models/MAILAND/majicflus_v1/majicflus_v134.safetensors",
]
)
pipe = FluxImagePipeline.from_model_manager(model_manager, device="cuda")
# 生图!
image = pipe(prompt="a beautiful girl", seed=0)
image.save("image.jpg")
```

Binary file not shown.

Before

Width:  |  Height:  |  Size: 15 MiB

BIN
_cover_images_/cover.png Normal file

Binary file not shown.

After

Width:  |  Height:  |  Size: 1.1 MiB

BIN
_cover_images_/cover2.png Normal file

Binary file not shown.

After

Width:  |  Height:  |  Size: 1.2 MiB

BIN
_cover_images_/cover3.webp Normal file

Binary file not shown.

After

Width:  |  Height:  |  Size: 26 KiB

BIN
_cover_images_/cover4.webp Normal file

Binary file not shown.

After

Width:  |  Height:  |  Size: 33 KiB

BIN
_cover_images_/cover5.webp Normal file

Binary file not shown.

After

Width:  |  Height:  |  Size: 74 KiB

BIN
_cover_images_/cover6.webp Normal file

Binary file not shown.

After

Width:  |  Height:  |  Size: 32 KiB

BIN
_cover_images_/cover7.webp Normal file

Binary file not shown.

After

Width:  |  Height:  |  Size: 22 KiB