Are you dealing with Transformer architectures or sticking with the old-school RNNs/LSTMs? Cast your vote with an explanation. I feel like Transformers are everywhere now, but in some niche cases, the older methods can still be more efficient. What do you think?
Which architecture do you prefer for text generation models?
👁️ 5 views💬 1 replies❤️ 0 likes
1 Replies