Commit
•
4b2e331
1
Parent(s):
9007626
Update README.md
Browse files
README.md
CHANGED
@@ -11,36 +11,14 @@ should probably proofread and complete it, then remove this comment. -->
|
|
11 |
|
12 |
# flan-t5-xl-t2s
|
13 |
|
14 |
-
This model was trained from scratch on an unknown dataset.
|
15 |
-
|
16 |
## Model description
|
17 |
|
18 |
-
|
19 |
-
|
20 |
-
|
21 |
-
|
22 |
-
|
23 |
-
|
24 |
-
|
25 |
-
|
26 |
-
|
27 |
-
|
28 |
-
## Training procedure
|
29 |
-
|
30 |
-
### Training hyperparameters
|
31 |
-
|
32 |
-
The following hyperparameters were used during training:
|
33 |
-
- learning_rate: 5e-05
|
34 |
-
- train_batch_size: 64
|
35 |
-
- eval_batch_size: 64
|
36 |
-
- seed: 42
|
37 |
-
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
38 |
-
- lr_scheduler_type: linear
|
39 |
-
- num_epochs: 3.0
|
40 |
-
|
41 |
-
### Framework versions
|
42 |
-
|
43 |
-
- Transformers 4.26.0
|
44 |
-
- Pytorch 2.0.0+cu117
|
45 |
-
- Datasets 2.6.1
|
46 |
-
- Tokenizers 0.13.2
|
|
|
11 |
|
12 |
# flan-t5-xl-t2s
|
13 |
|
|
|
|
|
14 |
## Model description
|
15 |
|
16 |
+
This FLAN-T5-XL model converts a given English question, conditioned on a DB schema, to the corresponding input query. Input format:
|
17 |
+
```
|
18 |
+
semantic parse: <question> | <db_id> | <schema_without_content>
|
19 |
+
```
|
20 |
+
Where `<schema_without_content>` is of the form `<table1>: <col1>, <col2>, ... | <table2>: ...`. The output format is
|
21 |
+
```
|
22 |
+
<db_id> | <query>
|
23 |
+
```
|
24 |
+
This model is one of the baselines (FLAN-T5-XL) we compared to in our paper.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|