---
base_model: stabilityai/stable-diffusion-2-1
license: openrail++
model_creator: stabilityai
model_name: stable-diffusion-2-1
quantized_by: Second State Inc.
tags:
- stable-diffusion
- text-to-image
---

<!-- header start -->
<!-- 200823 -->
<div style="width: auto; margin-left: auto; margin-right: auto">
<img src="https://github.com/LlamaEdge/LlamaEdge/raw/dev/assets/logo.svg" style="width: 100%; min-width: 400px; display: block; margin: auto;">
</div>
<hr style="margin-top: 1.0em; margin-bottom: 1.0em;">
<!-- header end -->

# stable-diffusion-2-1-GGUF

## Original Model

[stabilityai/stable-diffusion-2-1](https://huggingface.co/stabilityai/stable-diffusion-2-1)

## Run with `sd-api-server`

Go to the [sd-api-server](https://github.com/LlamaEdge/sd-api-server/blob/main/README.md) repository for more information.

<!-- - LlamaEdge version: [v0.12.2](https://github.com/LlamaEdge/LlamaEdge/releases/tag/0.12.2) and above

- Prompt template

  - Prompt type: `chatml`

  - Prompt string

    ```text
    <|im_start|>system
    {system_message}<|im_end|>
    <|im_start|>user
    {prompt}<|im_end|>
    <|im_start|>assistant
    ```

- Context size: `4096`

- Run as LlamaEdge service

  ```bash
  wasmedge --dir .:. --nn-preload default:GGML:AUTO:stablelm-2-12b-chat-Q5_K_M.gguf \
    llama-api-server.wasm \
    --prompt-template chatml \
    --ctx-size 4096 \
    --model-name stablelm-2-12b-chat
  ```

- Run as LlamaEdge command app

  ```bash
  wasmedge --dir .:. \
    --nn-preload default:GGML:AUTO:stablelm-2-12b-chat-Q5_K_M.gguf \
    llama-chat.wasm \
    --prompt-template chatml \
    --ctx-size 4096
  ``` -->

## Quantized GGUF Models

Using formats of different precisions will yield results of varying quality.

| f32  | f16  |q8_0  |q5_0  |q5_1  |q4_0  |q4_1  |
| ----  |----  |----  |----  |----  |----  |----  |
| ![](https://github.com/leejet/stable-diffusion.cpp/blob/master/assets/f32.png?raw=true) |![](https://github.com/leejet/stable-diffusion.cpp/blob/master/assets/f16.png?raw=true) |![](https://github.com/leejet/stable-diffusion.cpp/blob/master/assets/q8_0.png?raw=true) |![](https://github.com/leejet/stable-diffusion.cpp/blob/master/assets/q5_0.png?raw=true) |![](https://github.com/leejet/stable-diffusion.cpp/blob/master/assets/q5_1.png?raw=true) |![](https://github.com/leejet/stable-diffusion.cpp/blob/master/assets/q4_0.png?raw=true) |![](https://github.com/leejet/stable-diffusion.cpp/blob/master/assets/q4_1.png?raw=true) |

| Name | Quant method | Bits | Size | Use case |
| ---- | ---- | ---- | ---- | ----- |
| [v2-1_768-nonema-pruned-Q4_0.gguf](https://huggingface.co/second-state/stable-diffusion-2-1-GGUF/blob/main/v2-1_768-nonema-pruned-Q4_0.gguf) | Q4_0 | 2 | 1.70 GB | |
| [v2-1_768-nonema-pruned-Q4_1.gguf](https://huggingface.co/second-state/stable-diffusion-2-1-GGUF/blob/main/v2-1_768-nonema-pruned-Q4_1.gguf) | Q4_1 | 3 | 1.74 GB | |
| [v2-1_768-nonema-pruned-Q5_0.gguf](https://huggingface.co/second-state/stable-diffusion-2-1-GGUF/blob/main/v2-1_768-nonema-pruned-Q5_0.gguf) | Q5_0 | 3 | 1.78 GB | |
| [v2-1_768-nonema-pruned-Q5_1.gguf](https://huggingface.co/second-state/stable-diffusion-2-1-GGUF/blob/main/v2-1_768-nonema-pruned-Q5_1.gguf) | Q5_1 | 3 | 1.82 GB | |
| [v2-1_768-nonema-pruned-Q8_0.gguf](https://huggingface.co/second-state/stable-diffusion-2-1-GGUF/blob/main/v2-1_768-nonema-pruned-Q8_0.gguf) | Q8_0 | 4 | 2.01 GB | |
| [v2-1_768-nonema-pruned-f16.gguf](https://huggingface.co/second-state/stable-diffusion-2-1-GGUF/blob/main/v2-1_768-nonema-pruned-f16.gguf)   | f16  | 4 | 2.61 GB | |
| [v2-1_768-nonema-pruned-f32.gguf](https://huggingface.co/second-state/stable-diffusion-2-1-GGUF/blob/main/v2-1_768-nonema-pruned-f32.gguf)   | f32  | 4 | 5.21 GB | |