ajders commited on
Commit
d66d1c6
1 Parent(s): 2a4e7c0

model documentation (#3)

Browse files

- model documentation (884c81352a0a584a461f16f9f2ae9e09b80f8a21)
- model documentation (99e7ee749b8d76d12554fc0414232903561ce47d)

Files changed (1) hide show
  1. README.md +156 -17
README.md CHANGED
@@ -1,39 +1,178 @@
 
1
  ---
2
  language:
3
  - da
 
4
  tags:
5
  - bert
6
  - pytorch
7
  - sentiment
8
  - polarity
9
- license: cc-by-sa-4.0
10
- datasets:
11
- - Twitter Sentiment
12
- - Europarl Sentiment
13
  metrics:
14
  - f1
15
  widget:
16
  - text: Det er super godt
17
  ---
18
 
19
- # Danish BERT Tone for sentiment polarity detection
20
-
21
- The BERT Tone model detects sentiment polarity (positive, neutral or negative) in Danish texts.
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
22
  It has been finetuned on the pretrained [Danish BERT](https://github.com/certainlyio/nordic_bert) model by BotXO.
23
-
24
- See the [DaNLP documentation](https://danlp-alexandra.readthedocs.io/en/latest/docs/tasks/sentiment_analysis.html#bert-tone) for more details.
25
-
26
-
27
- Here is how to use the model:
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
28
 
29
  ```python
30
  from transformers import BertTokenizer, BertForSequenceClassification
31
-
32
  model = BertForSequenceClassification.from_pretrained("DaNLP/da-bert-tone-sentiment-polarity")
33
  tokenizer = BertTokenizer.from_pretrained("DaNLP/da-bert-tone-sentiment-polarity")
34
  ```
35
-
36
- ## Training data
37
-
38
- The data used for training come from the [Twitter Sentiment](https://danlp-alexandra.readthedocs.io/en/latest/docs/datasets.html#twitsent) and [EuroParl sentiment 2](https://danlp-alexandra.readthedocs.io/en/latest/docs/datasets.html#europarl-sentiment2) datasets.
39
 
 
1
+
2
  ---
3
  language:
4
  - da
5
+ license: cc-by-sa-4.0
6
  tags:
7
  - bert
8
  - pytorch
9
  - sentiment
10
  - polarity
 
 
 
 
11
  metrics:
12
  - f1
13
  widget:
14
  - text: Det er super godt
15
  ---
16
 
17
+ # Model Card for Danish BERT
18
+ Danish BERT Tone for sentiment polarity detection
19
+
20
+
21
+
22
+ # Model Details
23
+
24
+ ## Model Description
25
+
26
+ The BERT Tone model detects sentiment polarity (positive, neutral or negative) in Danish texts. It has been finetuned on the pretrained Danish BERT model by BotXO.
27
+
28
+ - **Developed by:** DaNLP
29
+ - **Shared by [Optional]:** Hugging Face
30
+ - **Model type:** Text Classification
31
+ - **Language(s) (NLP):** Danish (da)
32
+ - **License:** cc-by-sa-4.0
33
+ - **Related Models:** More information needed
34
+ - **Parent Model:** BERT
35
+ - **Resources for more information:**
36
+ - [GitHub Repo](https://github.com/certainlyio/nordic_bert)
37
+ - [Associated Documentation](https://danlp-alexandra.readthedocs.io/en/latest/docs/tasks/sentiment_analysis.html#bert-tone)
38
+
39
+
40
+ # Uses
41
+
42
+ ## Direct Use
43
+
44
+ This model can be used for text classification
45
+
46
+
47
+ ## Downstream Use [Optional]
48
+
49
+
50
+ More information needed.
51
+
52
+
53
+ ## Out-of-Scope Use
54
+
55
+ The model should not be used to intentionally create hostile or alienating environments for people.
56
+
57
+ # Bias, Risks, and Limitations
58
+
59
+
60
+ Significant research has explored bias and fairness issues with language models (see, e.g., [Sheng et al. (2021)](https://aclanthology.org/2021.acl-long.330.pdf) and [Bender et al. (2021)](https://dl.acm.org/doi/pdf/10.1145/3442188.3445922)). Predictions generated by the model may include disturbing and harmful stereotypes across protected classes; identity characteristics; and sensitive, social, and occupational groups.
61
+
62
+
63
+ ## Recommendations
64
+
65
+
66
+ Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.
67
+
68
+
69
+ # Training Details
70
+
71
+ ## Training Data
72
+
73
+ The data used for training come from the [Twitter Sentiment](https://danlp-alexandra.readthedocs.io/en/latest/docs/datasets.html#twitsent) and [EuroParl sentiment 2](https://danlp-alexandra.readthedocs.io/en/latest/docs/datasets.html#europarl-sentiment2) datasets.
74
+
75
+ ## Training Procedure
76
+
77
+ ### Preprocessing
78
+
79
  It has been finetuned on the pretrained [Danish BERT](https://github.com/certainlyio/nordic_bert) model by BotXO.
80
+
81
+ ### Speeds, Sizes, Times
82
+ More information needed.
83
+
84
+ # Evaluation
85
+
86
+
87
+ ## Testing Data, Factors & Metrics
88
+
89
+ ### Testing Data
90
+
91
+ More information needed.
92
+
93
+ ### Factors
94
+
95
+
96
+
97
+ ### Metrics
98
+
99
+ F1
100
+
101
+ ## Results
102
+
103
+ More information needed.
104
+
105
+ # Model Examination
106
+
107
+ More information needed.
108
+
109
+ # Environmental Impact
110
+
111
+
112
+ Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).
113
+
114
+ - **Hardware Type:** More information needed.
115
+ - **Hours used:** More information needed.
116
+ - **Cloud Provider:** More information needed.
117
+ - **Compute Region:** More information needed.
118
+ - **Carbon Emitted:** More information needed.
119
+
120
+ # Technical Specifications [optional]
121
+
122
+ ## Model Architecture and Objective
123
+
124
+ More information needed.
125
+
126
+ ## Compute Infrastructure
127
+
128
+ More information needed.
129
+
130
+ ### Hardware
131
+
132
+ More information needed.
133
+
134
+ ### Software
135
+
136
+ More information needed.
137
+
138
+ # Citation
139
+
140
+ **BibTeX:**
141
+
142
+ More information needed.
143
+
144
+ **APA:**
145
+
146
+ More information needed.
147
+
148
+ # Glossary [optional]
149
+
150
+ More information needed.
151
+
152
+ # More Information [optional]
153
+
154
+ More information needed.
155
+
156
+ # Model Card Authors [optional]
157
+
158
+ DaNLP in collaboration with Ezi Ozoani and the Hugging Face team
159
+
160
+ # Model Card Contact
161
+
162
+ More information needed.
163
+
164
+ # How to Get Started with the Model
165
+
166
+ Use the code below to get started with the model.
167
+ <details>
168
+ <summary> Click to expand </summary>
169
 
170
  ```python
171
  from transformers import BertTokenizer, BertForSequenceClassification
172
+
173
  model = BertForSequenceClassification.from_pretrained("DaNLP/da-bert-tone-sentiment-polarity")
174
  tokenizer = BertTokenizer.from_pretrained("DaNLP/da-bert-tone-sentiment-polarity")
175
  ```
176
+ </details>
177
+
 
 
178