metadata
dataset_info:
features:
- name: image
dtype: image
- name: imgid
dtype: string
- name: filename
dtype: string
- name: captions_en
sequence: string
- name: captions_tr
sequence: string
splits:
- name: train
num_bytes: 829019413
num_examples: 6000
- name: test
num_bytes: 137254728
num_examples: 1000
- name: validation
num_bytes: 138403935
num_examples: 1000
download_size: 1104541860
dataset_size: 1104678076
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*
- split: validation
path: data/validation-*
task_categories:
- image-to-text
- translation
- sentence-similarity
language:
- tr
- en
tags:
- image captioning
- image-to-text
pretty_name: flickr8k-tr-mt
size_categories:
- 1K<n<10K
This dataset is a machine-translated version of the flickr8k into Turkish using the facebook/nllb-200-3.3B translation model.