datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
mirea-tl-eda
RTU MIREA Telegram Channel EDA
The dataset contains 22 metrics, describing post user engagement, linguistic units features, readability, AI messages generation score, semantic topic label, and model labelling confidence for the whole available time period at the time of analysis (over 10k text messages).
Overview
The dataset is the result of EDA performed on the data from RTU MIREA Telegram channel, retreived via aiogram. The full code for the data processing… See the full description on the dataset page: https://huggingface.co/datasets/complicat9d/mirea-tl-eda.tl-mirea-messages-topic-classification
RTU MIREA Telegram Channel Messages Topic Classification
The dataset has been created to fine-tune mDeBERTa-v3-base-mnli-xnli model for text classification tasks on Telegram messages with 17 different topics specific to the target of classification.
Overview
This dataset has been parsed from the official Telegram channel of RTU MIREA, using aiogram, and manually labelled based on the topic of a given message.
Preprocessing
Preprocessing of the data included… See the full description on the dataset page: https://huggingface.co/datasets/complicat9d/tl-mirea-messages-topic-classification.mirea_memes