Swahili Translation Exposure v1
Domain:
natural language processingRecord type:
dataset
Creator:
nil
Host:
nileagi/swahili-translation-exposure-v1
This dataset contains aligned English-Swahili sentence pairs prepared for translation pretraining/exposure use cases.
Total pairs: 229,312
Format: parallel text in two files with line-by-line alignment
Languages: English (en) and Swahili (sw)
eng.txt: English sentences (one sentence per line)