metadata
dataset_info:
features:
- name: info
dtype: string
- name: keyword
dtype: string
- name: action
dtype: string
- name: current_frame
dtype: image
- name: previous_frame_1
dtype: image
- name: previous_frame_2
dtype: image
- name: previous_frame_3
dtype: image
- name: previous_frame_4
dtype: image
- name: previous_frame_5
dtype: image
- name: previous_frame_6
dtype: image
- name: previous_frame_7
dtype: image
- name: previous_frame_8
dtype: image
- name: previous_frame_9
dtype: image
- name: previous_frame_10
dtype: image
configs:
- config_name: default
data_files:
- split: train
path: data/*.parquet
This is the dataset of the diffusion dream dataset. The dataset contains the following columns:
- info: A string describing the action taken in the frame
- keyword: A string describing the keyword of the action
- action: A string describing the action taken in the frame
- current_frame: The current frame of the video
- previous_frame_1: The frame before the current frame
- previous_frame_2: The frame before the previous frame
- previous_frame_3: The frame before the previous frame
- previous_frame_4: The frame before the previous frame
- previous_frame_5: The frame before the previous frame
- previous_frame_6: The frame before the previous frame
- previous_frame_7: The frame before the previous frame
- previous_frame_8: The frame before the previous frame
- previous_frame_9: The frame before the previous frame
- previous_frame_10: The frame before the previous frame
The dataset is used to train a model to predict the next frame in a video.