Sign in to view source links and access this dataset
Description
A dataset for training models to follow image edit instructions, created by timbrooks and last updated on March 2, 2023. It contains triplets of original images, textual edit prompts, and the resulting edited images. The dataset is intended for training generative models that can apply user-specified edits to images.
Use Cases
Training instruction-following image generation models based on the provided edit prompts and image pairs.
Fine-tuning diffusion or other generative models for controllable image editing based on the original and edited image examples.
Benchmarking model performance on following specific, user-provided image transformation instructions.
Strengths
Specifically designed for the concrete task of training instruction-following image models.
Provides aligned data triplets (original image, instruction, edited image) for supervised learning.
Limitations
Description metadata is limited; actual data quality requires manual inspection after download.
Row count, column details, and license information are unknown, which may limit suitability assessment.
Last updated 2023-03-02 11:19:16; freshness should be verified.
Provenance
Source
huggingface user timbrooks
Collection Method
Likely generated and filtered using CLIP models, as indicated by the dataset name.
Freshness
Last updated 2023-03-02 11:19:16.
License is unknown; users must verify permissions before commercial use.