tldextract==3.4.0 elasticsearch==8.6.2 matplotlib==3.7.0 seaborn==0.12.2 tqdm==4.64.1 transformers==4.26.1 pycld2==0.41 spacytextblob==4.0.0 spacy==3.5.4 pandas==2.1.1 tokenizers==0.13.2 h5py==3.8.0 uniseg datasets wandb better_profanity smart_open