title | app_file | sdk | colorFrom | colorTo | sdk_version |
---|---|---|---|---|---|
TonAI-Creative |
app.py |
gradio |
red |
green |
5.0.2 |
- Prompt: Enter the text prompt to generate your desired output. It should be less than 77 words.
- Width (and Height): Specify the width (and height) of the generated image.
- Fast Inference: Model will be quantized to FP16 (half precision) to boost the speed but the picture quality will be decreased a bit
- Negative Prompt: Text prompts that instruct the AI model that it should not include certain elements in its generated images.
- CFG Scale: A parameter that controls how much the image generation process follows the text prompt. The higher the value, the more the image sticks to a given text input.
- Inference Step: The number of steps the model takes to generate the output.
- Mode: Choose the style of image you want to generate.
- LoRA safetensor File: Upload a Low-Rank Adaptation (LoRA) safetensor file to fine-tune the model. You can seek the LoRA weight on the Internet. You can also adjust the
- Sampler: Method of generating data in a specific way. Set its value as
Default
to use the default config of pipeline
- GPU: NVIDIA GTX 1050Ti or equivalent
- RAM: 8 GB
- CPU: Intel i5 or equivalent
- Storage: 50 GB of free space
- Install Anaconda
- Create a new conda environment named 'tonai' with Python 3.10:
conda create -n tonai python=3.10
- Activate the environment:
conda activate tonai
- Ensure you have
requirement.txt
file in your working directory. - Install the required packages using pip:
pip install -r requirement.txt
For running on local machine, from your Terminal (or CMD) run the command
python app.py
And your Web UI app will run on local URL: http://127.0.0.1:7860
To deploy your app on your own server, refer This Blog