This is a minimal implementation of Boris Dayma's [DALL·E Mini](https://github.com/borisdayma/dalle-mini) in PyTorch. It has been stripped to the bare essentials necessary for doing inference. The only third party dependencies are numpy and torch.
It currently takes **7.4 seconds** to generate an image with DALL·E Mega on a standard GPU runtime in Colab
It currently takes **7.4 seconds** to generate an image with DALL·E Mega on a standard GPU runtime in Colab.
The flax model, and the code for coverting it to torch can be found [here](https://github.com/kuprel/min-dalle-flax).
The flax model and the code for coverting it to torch can be found [here](https://github.com/kuprel/min-dalle-flax).
## Install
@ -32,9 +32,10 @@ $ python image_from_text.py --text='court sketch of godzilla on trial' --mega
```
![Godzilla Trial](examples/godzilla_on_trial.png)
### Python
To load a model once and generate multiple times, first initialize `MinDalleTorch`
To load a model once and generate multiple times, first initialize `MinDalleTorch`.
```python
from min_dalle import MinDalleTorch
@ -46,19 +47,19 @@ model = MinDalleTorch(
)
```
The required models will be downloaded to `models_root` if they are not already there. After the model has loaded, call `generate_image` with some text and a seed as many times as you want.
```python
image = model.generate_image("a comfy chair that looks like an avocado")