ddpm butterflies 128

Introduction

The DDPM-BUTTERFLIES-128 model is a diffusion model trained using the 🤗 Diffusers library. It is designed for generating images of butterflies, based on the dataset located at /content/drive/MyDrive/image_and_text.

Architecture

This model utilizes the diffusion model architecture, which iteratively refines a sample through a series of denoising steps, guided by the Diffusers library.

Training

The model was trained with the following hyperparameters:

• Learning Rate: 0.0001
• Training Batch Size: 16
• Evaluation Batch Size: 16
• Gradient Accumulation Steps: 1
• Optimizer: AdamW
• Learning Rate Scheduler: None
• Learning Rate Warmup Steps: 500
• Mixed Precision: FP16

The training process captured metrics that are available for review via TensorBoard logs.

Guide: Running Locally

To run this model locally, follow these basic steps:

1. Clone the Repository: Get the model files from the Hugging Face repository.
2. Install Dependencies: Ensure you have the 🤗 Diffusers library installed.
3. Prepare Dataset: Ensure access to the dataset at /content/drive/MyDrive/image_and_text.
4. Execute Model: Run the model with your preferred configuration.

For optimal performance, it is recommended to use cloud GPUs such as those provided by AWS, Google Cloud, or Azure.

License

This model is licensed under the Apache-2.0 License, allowing for both personal and commercial use, with attribution required.