Instructions to use timduck8/hmv10 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Diffusers
How to use timduck8/hmv10 with Diffusers:
pip install -U diffusers transformers accelerate
import torch from diffusers import DiffusionPipeline # switch to "mps" for apple devices pipe = DiffusionPipeline.from_pretrained("timduck8/hmv10", torch_dtype=torch.bfloat16, device_map="cuda") prompt = "Astronaut in a jungle, cold color palette, muted colors, detailed, 8k" image = pipe(prompt).images[0] - Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- Draw Things
- DiffusionBee
Safetensors 转 Diffusers 格式说明
快速开始
1. 安装依赖
pip install -r requirements.txt
2. 执行转换
python convert_to_diffusers.py
转换后的目录结构
转换完成后,会在当前目录生成 novaAsianXL-diffusers/ 文件夹,包含完整的 Diffusers 格式:
novaAsianXL-diffusers/
├── model_index.json ← 模型索引配置
├── scheduler/
│ └── scheduler_config.json ← 调度器配置
├── text_encoder/
│ ├── config.json
│ └── model.safetensors ← 文本编码器权重
├── text_encoder_2/
│ ├── config.json
│ └── model.safetensors ← 第二个文本编码器
├── tokenizer/ ← 分词器配置
├── tokenizer_2/ ← 第二个分词器
├── unet/
│ ├── config.json ← UNet 配置
│ └── diffusion_pytorch_model.safetensors
├── vae/
│ ├── config.json ← VAE 配置
│ └── diffusion_pytorch_model.safetensors
└── ...
上传到 fal.ai
转换完成后,可以直接将 novaAsianXL-diffusers/ 整个文件夹上传到 Hugging Face,然后在 fal.ai 中引用。
上传到 Hugging Face(可选)
from huggingface_hub import HfApi
api = HfApi()
api.upload_folder(
folder_path="./novaAsianXL-diffusers",
repo_id="your-username/novaAsianXL",
repo_type="model"
)
自定义参数
如需修改输入/输出路径,编辑 convert_to_diffusers.py 中的:
INPUT_FILE = "./novaAsianXL_illustriousV50.safetensors" # 输入文件
OUTPUT_DIR = "./novaAsianXL-diffusers" # 输出目录
注意事项
- 转换过程需要较大内存(至少 16GB RAM 推荐)
- 生成的文件夹大小约为原文件的 1.5-2 倍
- 确保有足够的磁盘空间(至少 20GB)