yundai424 commited on
Commit
4913ad1
1 Parent(s): 3ea8f1a

Update README to remove blank line at top

Browse files

this blank line completely invalidates the markdown metadata schema. We have some internal code that checks at markdown metadata to get the license and the blank line will lead to markdown's `Meta` being absent

Files changed (1) hide show
  1. README.md +77 -78
README.md CHANGED
@@ -1,79 +1,78 @@
1
-
2
- ---
3
- language:
4
- - en
5
- tags:
6
- - aspect-based-sentiment-analysis
7
- - PyABSA
8
- license: mit
9
- datasets:
10
- - laptop14
11
- - restaurant14
12
- - restaurant16
13
- - ACL-Twitter
14
- - MAMS
15
- - Television
16
- - TShirt
17
- - Yelp
18
- metrics:
19
- - accuracy
20
- - macro-f1
21
- widget:
22
- - text: "[CLS] when tables opened up, the manager sat another party before us. [SEP] manager [SEP] "
23
- ---
24
-
25
- # Note
26
- This model is training with 30k+ ABSA samples, see [ABSADatasets](https://github.com/yangheng95/ABSADatasets). Yet the test sets are not included in pre-training, so you can use this model for training and benchmarking on common ABSA datasets, e.g., Laptop14, Rest14 datasets. (Except for the Rest15 dataset!)
27
-
28
- # DeBERTa for aspect-based sentiment analysis
29
- The `deberta-v3-large-absa` model for aspect-based sentiment analysis, trained with English datasets from [ABSADatasets](https://github.com/yangheng95/ABSADatasets).
30
-
31
- ## Training Model
32
- This model is trained based on the FAST-LCF-BERT model with `microsoft/deberta-v3-large`, which comes from [PyABSA](https://github.com/yangheng95/PyABSA).
33
- To track state-of-the-art models, please see [PyASBA](https://github.com/yangheng95/PyABSA).
34
-
35
- ## Usage
36
- ```python3
37
- from transformers import AutoTokenizer, AutoModelForSequenceClassification
38
-
39
- tokenizer = AutoTokenizer.from_pretrained("yangheng/deberta-v3-large-absa-v1.1")
40
-
41
- model = AutoModelForSequenceClassification.from_pretrained("yangheng/deberta-v3-large-absa-v1.1")
42
- ```
43
-
44
- ## Example in PyASBA
45
- An [example](https://github.com/yangheng95/PyABSA/blob/release/demos/aspect_polarity_classification/train_apc_multilingual.py) for using FAST-LCF-BERT in PyASBA datasets.
46
-
47
- ## Datasets
48
- This model is fine-tuned with 180k examples for the ABSA dataset (including augmented data). Training dataset files:
49
- ```
50
- loading: integrated_datasets/apc_datasets/SemEval/laptop14/Laptops_Train.xml.seg
51
- loading: integrated_datasets/apc_datasets/SemEval/restaurant14/Restaurants_Train.xml.seg
52
- loading: integrated_datasets/apc_datasets/SemEval/restaurant16/restaurant_train.raw
53
- loading: integrated_datasets/apc_datasets/ACL_Twitter/acl-14-short-data/train.raw
54
- loading: integrated_datasets/apc_datasets/MAMS/train.xml.dat
55
- loading: integrated_datasets/apc_datasets/Television/Television_Train.xml.seg
56
- loading: integrated_datasets/apc_datasets/TShirt/Menstshirt_Train.xml.seg
57
- loading: integrated_datasets/apc_datasets/Yelp/yelp.train.txt
58
-
59
- ```
60
- If you use this model in your research, please cite our paper:
61
- ```
62
- @article{YangZMT21,
63
- author = {Heng Yang and
64
- Biqing Zeng and
65
- Mayi Xu and
66
- Tianxing Wang},
67
- title = {Back to Reality: Leveraging Pattern-driven Modeling to Enable Affordable
68
- Sentiment Dependency Learning},
69
- journal = {CoRR},
70
- volume = {abs/2110.08604},
71
- year = {2021},
72
- url = {https://arxiv.org/abs/2110.08604},
73
- eprinttype = {arXiv},
74
- eprint = {2110.08604},
75
- timestamp = {Fri, 22 Oct 2021 13:33:09 +0200},
76
- biburl = {https://dblp.org/rec/journals/corr/abs-2110-08604.bib},
77
- bibsource = {dblp computer science bibliography, https://dblp.org}
78
- }
79
  ```
 
1
+ ---
2
+ language:
3
+ - en
4
+ tags:
5
+ - aspect-based-sentiment-analysis
6
+ - PyABSA
7
+ license: mit
8
+ datasets:
9
+ - laptop14
10
+ - restaurant14
11
+ - restaurant16
12
+ - ACL-Twitter
13
+ - MAMS
14
+ - Television
15
+ - TShirt
16
+ - Yelp
17
+ metrics:
18
+ - accuracy
19
+ - macro-f1
20
+ widget:
21
+ - text: "[CLS] when tables opened up, the manager sat another party before us. [SEP] manager [SEP] "
22
+ ---
23
+
24
+ # Note
25
+ This model is training with 30k+ ABSA samples, see [ABSADatasets](https://github.com/yangheng95/ABSADatasets). Yet the test sets are not included in pre-training, so you can use this model for training and benchmarking on common ABSA datasets, e.g., Laptop14, Rest14 datasets. (Except for the Rest15 dataset!)
26
+
27
+ # DeBERTa for aspect-based sentiment analysis
28
+ The `deberta-v3-large-absa` model for aspect-based sentiment analysis, trained with English datasets from [ABSADatasets](https://github.com/yangheng95/ABSADatasets).
29
+
30
+ ## Training Model
31
+ This model is trained based on the FAST-LCF-BERT model with `microsoft/deberta-v3-large`, which comes from [PyABSA](https://github.com/yangheng95/PyABSA).
32
+ To track state-of-the-art models, please see [PyASBA](https://github.com/yangheng95/PyABSA).
33
+
34
+ ## Usage
35
+ ```python3
36
+ from transformers import AutoTokenizer, AutoModelForSequenceClassification
37
+
38
+ tokenizer = AutoTokenizer.from_pretrained("yangheng/deberta-v3-large-absa-v1.1")
39
+
40
+ model = AutoModelForSequenceClassification.from_pretrained("yangheng/deberta-v3-large-absa-v1.1")
41
+ ```
42
+
43
+ ## Example in PyASBA
44
+ An [example](https://github.com/yangheng95/PyABSA/blob/release/demos/aspect_polarity_classification/train_apc_multilingual.py) for using FAST-LCF-BERT in PyASBA datasets.
45
+
46
+ ## Datasets
47
+ This model is fine-tuned with 180k examples for the ABSA dataset (including augmented data). Training dataset files:
48
+ ```
49
+ loading: integrated_datasets/apc_datasets/SemEval/laptop14/Laptops_Train.xml.seg
50
+ loading: integrated_datasets/apc_datasets/SemEval/restaurant14/Restaurants_Train.xml.seg
51
+ loading: integrated_datasets/apc_datasets/SemEval/restaurant16/restaurant_train.raw
52
+ loading: integrated_datasets/apc_datasets/ACL_Twitter/acl-14-short-data/train.raw
53
+ loading: integrated_datasets/apc_datasets/MAMS/train.xml.dat
54
+ loading: integrated_datasets/apc_datasets/Television/Television_Train.xml.seg
55
+ loading: integrated_datasets/apc_datasets/TShirt/Menstshirt_Train.xml.seg
56
+ loading: integrated_datasets/apc_datasets/Yelp/yelp.train.txt
57
+
58
+ ```
59
+ If you use this model in your research, please cite our paper:
60
+ ```
61
+ @article{YangZMT21,
62
+ author = {Heng Yang and
63
+ Biqing Zeng and
64
+ Mayi Xu and
65
+ Tianxing Wang},
66
+ title = {Back to Reality: Leveraging Pattern-driven Modeling to Enable Affordable
67
+ Sentiment Dependency Learning},
68
+ journal = {CoRR},
69
+ volume = {abs/2110.08604},
70
+ year = {2021},
71
+ url = {https://arxiv.org/abs/2110.08604},
72
+ eprinttype = {arXiv},
73
+ eprint = {2110.08604},
74
+ timestamp = {Fri, 22 Oct 2021 13:33:09 +0200},
75
+ biburl = {https://dblp.org/rec/journals/corr/abs-2110-08604.bib},
76
+ bibsource = {dblp computer science bibliography, https://dblp.org}
77
+ }
 
78
  ```