Ankur Goyal
commited on
Commit
·
9d33a03
1
Parent(s):
dcab013
Improve titles
Browse files
README.md
CHANGED
|
@@ -18,8 +18,13 @@ invoices as well as both [SQuAD2.0](https://huggingface.co/datasets/squad_v2) an
|
|
| 18 |
Unlike other QA models, which can only extract consecutive tokens (because they predict the start and end of a sequence), this model can predict longer-range, non-consecutive sequences with an additional
|
| 19 |
classifier head. For example, QA models often encounter this failure mode:
|
| 20 |
|
|
|
|
|
|
|
| 21 |
|
| 22 |
|
|
|
|
|
|
|
|
|
|
| 23 |
However this model is able to predict non-consecutive tokens and therefore the address correctly:
|
| 24 |
|
| 25 |
|
|
|
|
| 18 |
Unlike other QA models, which can only extract consecutive tokens (because they predict the start and end of a sequence), this model can predict longer-range, non-consecutive sequences with an additional
|
| 19 |
classifier head. For example, QA models often encounter this failure mode:
|
| 20 |
|
| 21 |
+
### Before
|
| 22 |
+
|
| 23 |
|
| 24 |
|
| 25 |
+
|
| 26 |
+
### After
|
| 27 |
+
|
| 28 |
However this model is able to predict non-consecutive tokens and therefore the address correctly:
|
| 29 |
|
| 30 |
|