| | --- |
| | library_name: diffusers |
| | base_model: segmind/Segmind-Vega |
| | tags: |
| | - lora |
| | - text-to-image |
| | license: apache-2.0 |
| | inference: false |
| | --- |
| | # Segmind-VegaRT - Latent Consistency Model (LCM) LoRA of Segmind-Vega |
| |
|
| | Try real-time inference here **[VegaRT demo⚡](https://www.segmind.com/segmind-vega-rt)** |
| |
|
| | API for **[Segmind-VegaRT](https://www.segmind.com/models/segmind-vega-rt-v1/api)** |
| |
|
| | <video controls autoplay src="https://cdn-uploads.huggingface.co/production/uploads/62039c2d91d53938a643317d/WacXd5DqP5hx8iEGTPt16.mp4"></video> |
| |
|
| | Segmind-VegaRT a distilled consistency adapter for [Segmind-Vega](https://huggingface.co/segmind/Segmind-Vega) that allows |
| | to reduce the number of inference steps to only between **2 - 8 steps**. |
| |
|
| | Latent Consistency Model (LCM) LoRA was proposed in [LCM-LoRA: A universal Stable-Diffusion Acceleration Module](https://arxiv.org/abs/2311.05556) |
| | by *Simian Luo, Yiqin Tan, Suraj Patil, Daniel Gu et al.* |
| |
|
| | # Image comparison (Segmind-VegaRT vs SDXL-Turbo) |
| |
|
| |  |
| |
|
| |  |
| |
|
| |  |
| |
|
| | # Speed comparison (Segmind-VegaRT vs SDXL-Turbo) on A100 80GB |
| |
|
| |  |
| |
|
| | | Model | Params / M | |
| | |----------------------------------------------------------------------------|------------| |
| | | [lcm-lora-sdv1-5](https://huggingface.co/latent-consistency/lcm-lora-sdv1-5) | 67.5 | |
| | | [**Segmind-VegaRT**](https://huggingface.co/segmind/Segmind-VegaRT) | **119** | |
| | | [lcm-lora-sdxl](https://huggingface.co/latent-consistency/lcm-lora-sdxl) | 197 | |
| |
|
| | ## Usage |
| |
|
| | LCM-LoRA is supported in 🤗 Hugging Face Diffusers library from version v0.23.0 onwards. To run the model, first |
| | install the latest version of the Diffusers library as well as `peft`, `accelerate` and `transformers`. |
| | audio dataset from the Hugging Face Hub: |
| |
|
| | ```bash |
| | pip install --upgrade pip |
| | pip install --upgrade diffusers transformers accelerate peft |
| | ``` |
| |
|
| | ### Text-to-Image |
| |
|
| | Let's load the base model `segmind/Segmind-Vega` first. Next, the scheduler needs to be changed to [`LCMScheduler`](https://huggingface.co/docs/diffusers/v0.22.3/en/api/schedulers/lcm#diffusers.LCMScheduler) and we can reduce the number of inference steps to just 2 to 8 steps. |
| | Please make sure to either disable `guidance_scale` or use values between 1.0 and 2.0. |
| |
|
| | ```python |
| | import torch |
| | from diffusers import LCMScheduler, AutoPipelineForText2Image |
| | |
| | model_id = "segmind/Segmind-Vega" |
| | adapter_id = "segmind/Segmind-VegaRT" |
| | |
| | pipe = AutoPipelineForText2Image.from_pretrained(model_id, torch_dtype=torch.float16, variant="fp16") |
| | pipe.scheduler = LCMScheduler.from_config(pipe.scheduler.config) |
| | pipe.to("cuda") |
| | |
| | # load and fuse lcm lora |
| | pipe.load_lora_weights(adapter_id) |
| | pipe.fuse_lora() |
| | |
| | |
| | prompt = "Self-portrait oil painting, a beautiful cyborg with golden hair, 8k" |
| | |
| | # disable guidance_scale by passing 0 |
| | image = pipe(prompt=prompt, num_inference_steps=4, guidance_scale=0).images[0] |
| | ``` |
