Should you switch from BERT to ALBERT?
From the summaries you’ll find online, it sounds like ALBERT is both faster and more accurate than BERT--so we should probably switch over to ALBERT as our new go-to model for fine-tuning, right? In reality, if your plan is to try applying a pre-trained BERT model to your own NLP application, then you probably won’t find ALBERT to be any faster or more accurate...
In this video, we’ll make sense of this apparent contradiction, and I’ll cover the key take-aways from ALBERT, which is still a very important and fascinating model!
For a complete tutorial on ALBERT, I’ve published a ~25-page eBook, included in my "NLP Base Camp" membership here:
https://www.chrismccormick.ai/members...
Watch video Should you switch from BERT to ALBERT? online without registration, duration hours minute second in high quality. This video was added by user ChrisMcCormickAI 11 February 2020, don't forget to share it with your friends and acquaintances, it has been viewed on our site 14,883 once and liked it 510 people.