🚀 reddy-v3

reddy-v3 是一个基于 black-forest-labs/FLUX.1-dev 的标准 PEFT LoRA 模型。该模型可用于文本到图像的生成任务，通过特定的训练和验证设置，能够生成具有特定风格和质量的图像。

🚀 快速开始

本项目是基于 black-forest-labs/FLUX.1-dev 派生的标准 PEFT LoRA 模型。以下是使用该模型进行推理的基本步骤：

import torch
from diffusers import DiffusionPipeline

model_id = 'black-forest-labs/FLUX.1-dev'
adapter_id = 'Unmapped2895/reddy-v3'
pipeline = DiffusionPipeline.from_pretrained(model_id, torch_dtype=torch.bfloat16) # 直接以 bf16 格式加载
pipeline.load_lora_weights(adapter_id)

prompt = "Realistic wide shot photo of woman posing in a luxurious satin lingerie set, featuring a plunging bra, delicate thong and a classic garter belt with black stockings. The satin lingerie shimmers softly in the light, and the cut emphasizes both sophistication and a hint of allure. The lingerie is detailed with fine lace edges, highlighting her alluring figure. She elegantly styled hair as if getting ready for a formal event. The photo has a cinematic quality with rays of light and dramatic play of shadow and light"

## 可选：对模型进行量化以节省显存。
## 注意：模型在训练期间已进行量化，因此建议在推理时也进行相同操作。
from optimum.quanto import quantize, freeze, qint8
quantize(pipeline.transformer, weights=qint8)
freeze(pipeline.transformer)

pipeline.to('cuda' if torch.cuda.is_available() else 'mps' if torch.backends.mps.is_available() else 'cpu') # 管道已处于目标精度级别
model_output = pipeline(
    prompt=prompt,
    num_inference_steps=20,
    generator=torch.Generator(device='cuda' if torch.cuda.is_available() else 'mps' if torch.backends.mps.is_available() else 'cpu').manual_seed(42),
    width=832,
    height=1216,
    guidance_scale=3.5,
).images[0]

model_output.save("output.png", format="PNG")

💡 使用建议

• 文本编码器在训练过程中未进行训练，推理时可复用基础模型的文本编码器。
• 验证设置不一定与训练设置相同，使用时需注意。

✨ 主要特性

• 基于 black-forest-labs/FLUX.1-dev 基础模型派生，具有特定的图像生成能力。
• 提供了详细的训练和验证设置，方便用户了解模型的训练过程。
• 支持推理时的模型量化，可节省显存。

📦 安装指南

文档中未提及具体安装命令，暂不提供安装指南。

💻 使用示例

基础用法

import torch
from diffusers import DiffusionPipeline

model_id = 'black-forest-labs/FLUX.1-dev'
adapter_id = 'Unmapped2895/reddy-v3'
pipeline = DiffusionPipeline.from_pretrained(model_id, torch_dtype=torch.bfloat16) # 直接以 bf16 格式加载
pipeline.load_lora_weights(adapter_id)

prompt = "Realistic wide shot photo of woman posing in a luxurious satin lingerie set, featuring a plunging bra, delicate thong and a classic garter belt with black stockings. The satin lingerie shimmers softly in the light, and the cut emphasizes both sophistication and a hint of allure. The lingerie is detailed with fine lace edges, highlighting her alluring figure. She elegantly styled hair as if getting ready for a formal event. The photo has a cinematic quality with rays of light and dramatic play of shadow and light"

pipeline.to('cuda' if torch.cuda.is_available() else 'mps' if torch.backends.mps.is_available() else 'cpu') # 管道已处于目标精度级别
model_output = pipeline(
    prompt=prompt,
    num_inference_steps=20,
    generator=torch.Generator(device='cuda' if torch.cuda.is_available() else 'mps' if torch.backends.mps.is_available() else 'cpu').manual_seed(42),
    width=832,
    height=1216,
    guidance_scale=3.5,
).images[0]

model_output.save("output.png", format="PNG")

高级用法

import torch
from diffusers import DiffusionPipeline

model_id = 'black-forest-labs/FLUX.1-dev'
adapter_id = 'Unmapped2895/reddy-v3'
pipeline = DiffusionPipeline.from_pretrained(model_id, torch_dtype=torch.bfloat16) # 直接以 bf16 格式加载
pipeline.load_lora_weights(adapter_id)

prompt = "Realistic wide shot photo of woman posing in a luxurious satin lingerie set, featuring a plunging bra, delicate thong and a classic garter belt with black stockings. The satin lingerie shimmers softly in the light, and the cut emphasizes both sophistication and a hint of allure. The lingerie is detailed with fine lace edges, highlighting her alluring figure. She elegantly styled hair as if getting ready for a formal event. The photo has a cinematic quality with rays of light and dramatic play of shadow and light"

## 可选：对模型进行量化以节省显存。
## 注意：模型在训练期间已进行量化，因此建议在推理时也进行相同操作。
from optimum.quanto import quantize, freeze, qint8
quantize(pipeline.transformer, weights=qint8)
freeze(pipeline.transformer)

pipeline.to('cuda' if torch.cuda.is_available() else 'mps' if torch.backends.mps.is_available() else 'cpu') # 管道已处于目标精度级别
model_output = pipeline(
    prompt=prompt,
    num_inference_steps=20,
    generator=torch.Generator(device='cuda' if torch.cuda.is_available() else 'mps' if torch.backends.mps.is_available() else 'cpu').manual_seed(42),
    width=832,
    height=1216,
    guidance_scale=3.5,
).images[0]

model_output.save("output.png", format="PNG")

📚 详细文档

验证设置

• CFG：3.5
• CFG Rescale：0.0
• Steps：20
• Sampler：FlowMatchEulerDiscreteScheduler
• Seed：42
• Resolution：832x1216
• Skip-layer guidance：无

注意：验证设置不一定与训练设置相同。

你可以在以下图库中找到一些示例图像：