Timeline for How does Bert masked language modelling task make sense if half the time the next sentence is wrong context in the sequence passed through the encoder
Current License: CC BY-SA 4.0
3 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Nov 15, 2023 at 5:44 | vote | accept | figs_and_nuts | ||
| Nov 15, 2023 at 5:44 | comment | added | figs_and_nuts | I understood the purpose of different losses however, I was confused about the MLM loss also being generated from the sequences where the next sentence was not an actual entailment. But, that part you have answered in the later half of your answer. Going through RoBERTa paper was also revealing. Thank you for your time and knowledge :) | |
| Nov 14, 2023 at 8:30 | history | answered | noe | CC BY-SA 4.0 |