Update README.md
Browse files
README.md
CHANGED
@@ -1,3 +1,34 @@
|
|
1 |
---
|
2 |
license: cc-by-sa-3.0
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
3 |
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
---
|
2 |
license: cc-by-sa-3.0
|
3 |
+
datasets:
|
4 |
+
- kilt_tasks
|
5 |
+
metrics:
|
6 |
+
- exact_match
|
7 |
+
pipeline_tag: text-generation
|
8 |
+
model-index:
|
9 |
+
- name: results
|
10 |
+
results:
|
11 |
+
- task:
|
12 |
+
name: Question Answering
|
13 |
+
type: text-generation
|
14 |
+
dataset:
|
15 |
+
name: NQ KILT
|
16 |
+
type: kilt_tasks
|
17 |
+
args: nq
|
18 |
+
metrics:
|
19 |
+
- name: Exact Macth
|
20 |
+
type: exact_match
|
21 |
+
value: 59.01
|
22 |
+
language:
|
23 |
+
- en
|
24 |
---
|
25 |
+
|
26 |
+
# Fusion-In-Decoder Base on Natural Questions
|
27 |
+
|
28 |
+
This trained model is based on the Fusion-In-Decoder model, and trained on the Natural Questions dataset.
|
29 |
+
|
30 |
+
# Model Details
|
31 |
+
|
32 |
+
Model is based on Fusion-In-Decoder, which in turn is based on the t5-large checkpoint. For training, we utilized text retrieval for each query, which provides a collection of relevant passages for it.
|
33 |
+
|
34 |
+
We note that the passages were retrieved using a corpus based on Wikipeida.
|