---
title: wd-1.5-beta2_unCLIP
canonical_url: "https://www.modelscope.cn/models/AI-ModelScope/wd-1.5-beta2_unCLIP"
md_url: "https://www.modelscope.cn/models/AI-ModelScope/wd-1.5-beta2_unCLIP.md"
repository: AI-ModelScope/wd-1.5-beta2_unCLIP
last_updated: 2024-03-21
pipeline_tag: text-generation
tasks:
  - text-generation
parameters: 11.4B
tensor_type:
  - F16
  - I32
  - F32
library_name:
  - safetensors
  - pytorch
frameworks:
  - pytorch
downloads: 4288
stars: 0
---

# wd-1.5-beta2_unCLIP

> wd-1.5-beta2_unCLIP - AI-ModelScope 在 ModelScope 开源的模型。https://huggingface.co/waifu-diffusion/wd-1-5-beta2/tree/main/checkpoints

- **Repository**: AI-ModelScope/wd-1.5-beta2_unCLIP
- **Tasks**: text-generation
- **Parameters**: 11.4B
- **Downloads**: 4288
- **Stars**: 0
- **Last updated**: 2024-03-21

Source: https://www.modelscope.cn/models/AI-ModelScope/wd-1.5-beta2_unCLIP

---

Obtained by merging 

https://huggingface.co/waifu-diffusion/wd-1-5-beta2/tree/main/checkpoints

with 

https://huggingface.co/stabilityai/stable-diffusion-2-1-unclip/tree/main


The exact recipe for the wd-1-5-beta2-aesthetic-unclip-h-fp32.safetensors is: (sd21-unclip-h.ckpt - v2-1_768-ema-pruned.ckpt) + wd-1-5-beta2-aesthetic-fp32.safetensors

Then I put those new text encoder and unet weights in the unCLIP checkpoint.


For how to use this in ComfyUI and for some information on what unCLIP is see: https://comfyanonymous.github.io/ComfyUI_examples/unclip/