|
|
|
--- |
|
tags: |
|
- TensorRT |
|
- Text2Image |
|
- Stable Diffusion |
|
- Image2Image |
|
- SDA |
|
--- |
|
|
|
# andite/anything-v4.0 converted into TensorRT |
|
|
|
<a href="https://github.com/chavinlo/sda-node/"><img src="https://i.imgur.com/fQS926g.png"></a> |
|
|
|
Model converted from diffusers into TensorRT for accelerated inference up to 4x faster. |
|
|
|
For how to use the model check https://github.com/chavinlo/sda-node |
|
|
|
This model was automatically converted by SDA-node |
|
|
|
Compilation configuration: |
|
|
|
|
|
|
|
```json |
|
{ |
|
"_class_name": "StableDiffusionAccelerated_Base", |
|
"_sda_version": "0.1.2", |
|
"_trt_version": "8.5.1", |
|
"_cuda_version": "11.6", |
|
"_cudnn_version": "7.5", |
|
"_onnx2trt_version": "8.5.1", |
|
"UNET": { |
|
"precision": "fp16", |
|
"path": "engine/unet_fp16.plan" |
|
}, |
|
"CLIP": { |
|
"path": "engine/clip.plan" |
|
}, |
|
"DE_VAE": { |
|
"path": "engine/de_vae.plan" |
|
} |
|
} |
|
``` |
