Logo Lanfrica

bdu-birhanu/LID_TCD

Domain:

natural language processing

Record type:

dataset
Creator:
bdu
Host:
Dataset and benchmarks for Language identification and content detection for low-resource languages # LID_TCD ## Dataset for Ethiopian language identification and topic classification - This datset consists of 22,624 texts labled for two tasks: - Language identification: this task is used to identify the lanaguage a give text written in. - Topic classification: this task is also useful to classify the topics of a given text according to its content. ## To run the code with Terminal use the following info. ``` # Load and Pre-process data python preprocess.py # Train python train.py # Test and results python test.py ``` ## Some issues to know 1. The test environment is - Python 3.5.2 - Keras 2.3.1 - tensorflow 2.1.0 =======