Jindřich questions

Jindřich answers

• How is transformers loss calculated for blank token predictions?
• Restrict Vocab for BERT Encoder-Decoder Text Generation
• What's the difference between NLTK's BLEU score and SacreBLEU?
• Implementation details of positional encoding in transformer model?
• Is there a maximum sequence length for the output of a transformer?
• Unsupervised finetuning of BERT for embeddings only?
• How can i use BERT fo machine Translation?
• How to use forward() method instead of model.generate() for T5 model
• How do I translate using HuggingFace from Chinese to English?
• BERT sentence embeddings from transformers