stanlochten
commited on
Commit
•
4cba82b
1
Parent(s):
34edea6
Update README.md
Browse files
README.md
CHANGED
@@ -1,3 +1,17 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
T5-base model fine-tuned for question generation from knowledge graphs. Can be used to generate questions from linearized knowledge graphs, meaning graphs in the form of its all its triples listed in the following format:
|
2 |
|
3 |
`<A> answer node(s) <H> head <R> relation <T> tail <H> head <R> relation <T> tail ... etc ...`,
|
@@ -23,4 +37,6 @@ print('Decoding...')
|
|
23 |
preds = tokenizer.batch_decode(y_hats, skip_special_tokens=True, clean_up_tokenization_spaces=True)
|
24 |
```
|
25 |
|
26 |
-
Good luck!
|
|
|
|
|
|
1 |
+
---
|
2 |
+
license: openrail
|
3 |
+
datasets:
|
4 |
+
- web_questions
|
5 |
+
language:
|
6 |
+
- en
|
7 |
+
metrics:
|
8 |
+
- bleu
|
9 |
+
- bertscore
|
10 |
+
library_name: transformers
|
11 |
+
tags:
|
12 |
+
- knowledge_graphs
|
13 |
+
- question_generation
|
14 |
+
---
|
15 |
T5-base model fine-tuned for question generation from knowledge graphs. Can be used to generate questions from linearized knowledge graphs, meaning graphs in the form of its all its triples listed in the following format:
|
16 |
|
17 |
`<A> answer node(s) <H> head <R> relation <T> tail <H> head <R> relation <T> tail ... etc ...`,
|
|
|
37 |
preds = tokenizer.batch_decode(y_hats, skip_special_tokens=True, clean_up_tokenization_spaces=True)
|
38 |
```
|
39 |
|
40 |
+
Good luck!
|
41 |
+
|
42 |
+
[Associated research report](https://dspace.uba.uva.nl/server/api/core/bitstreams/fee95174-b7d4-4cd8-8545-f7ec8ab29e2d/content)
|