pere commited on
Commit
dcda64c
1 Parent(s): 2c1585e

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +8 -27
README.md CHANGED
@@ -48,32 +48,14 @@ task_ids:
48
  - **Point of Contact:** [Per Erik Solberg](mailto:per.solberg@nb.no)
49
 
50
  The Norwegian Parliament Speech Corpus (NPSC) is a corpus for training a Norwegian ASR (Automatic Speech Recognition) models.
 
51
 
52
  ## How to Use
53
  ```python
 
54
  from datasets import load_dataset
55
- data = load_dataset("NbAiLab/NPSC", streaming=True)
56
  ```
57
- ## Download Data
58
- If you do not want to use the HuggingFace Dataset-library for training, or if you want to do additional pre-processing, it is also possible to download the files locally.
59
- ```bash
60
- # Clone the training set
61
- git clone https://huggingface.co/datasets/NbAiLab/NPSC
62
-
63
- # Create one large training file of all shards without unpacking
64
- cat NPSC/data/train*.gz > onefile.json.gz
65
- ```
66
-
67
- <details>
68
- <summary>List of all the files.</summary>
69
-
70
-
71
- * [eval](https://huggingface.co/datasets/NbAiLab/NPSC/resolve/main/data/eval.json.gz)
72
- * [test](https://huggingface.co/datasets/NbAiLab/NPSC/resolve/main/data/test.json.gz)
73
- * [train](https://huggingface.co/datasets/NbAiLab/NPSC/resolve/main/data/train.json.gz)
74
-
75
-
76
- </details>
77
 
78
  ### Dataset Summary
79
  The NPSC dataset contains json lines with language training data. Here is an example json line:
@@ -119,7 +101,7 @@ The NPSC dataset contains json lines with language training data. Here is an exa
119
 
120
  ### Dataset Creation
121
  We are providing a **train** and a **validation** split. The standard size of the validation is a single 1GB file, while train is sharded in 1GB chunks.
122
- All files are gzipped.
123
 
124
  Build date: 22012022
125
 
@@ -142,21 +124,20 @@ The procedure for the dataset creation is described in detail in our paper.
142
  ## Considerations for Using the Data
143
  This corpus contains speech data and is allowed to be used outside the National Library of Norway for speech recognition technology purposes.
144
 
145
- ### Discussion of Biases
146
- Please refer to our paper.
147
-
148
  ### Dataset Curators
149
- [Freddy Wetjen](mailto:Freddy.wetjen@nb.no) and [Andre Kaasen](mailto:andre.kasen@nb.no)
150
 
151
  ### Licensing Information
152
  Licensed for use outside the National Library of Norway.
153
 
154
  ## License
155
- CC-ZERO(https://creativecommons.org/publicdomain/zero/1.0/)
156
 
157
  ### Citation Information
158
  We are preparing an article with detailed information about this corpus. Until it is published, please cite out paper discussing the first version of this corpus:
159
  ```
 
 
160
  @inproceedings{kummervold-etal-2021-operationalizing,
161
  title = {Operationalizing a National Digital Library: The Case for a {N}orwegian Transformer Model},
162
  author = {Kummervold, Per E and
 
48
  - **Point of Contact:** [Per Erik Solberg](mailto:per.solberg@nb.no)
49
 
50
  The Norwegian Parliament Speech Corpus (NPSC) is a corpus for training a Norwegian ASR (Automatic Speech Recognition) models.
51
+ **!!!NEDS TO BE UPDATED!!! We need a real description here. About one paragraph summing up that is on the main web page, and telling how this dataset is the same but different - ie it is in a streaming format...**
52
 
53
  ## How to Use
54
  ```python
55
+ # Loads the 16K Bokmål corpus in streaming mode
56
  from datasets import load_dataset
57
+ data = load_dataset("NbAiLab/NPSC", config="16K_mp3_bokmaal", streaming=True)
58
  ```
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
59
 
60
  ### Dataset Summary
61
  The NPSC dataset contains json lines with language training data. Here is an example json line:
 
101
 
102
  ### Dataset Creation
103
  We are providing a **train** and a **validation** split. The standard size of the validation is a single 1GB file, while train is sharded in 1GB chunks.
104
+ All files are gzipped. There is also a **test** split available for the dataset but this is hidden. Please contact [Per Erik Solberg](mailto:per.erik.solberg@nb.no) for access to the test set.
105
 
106
  Build date: 22012022
107
 
 
124
  ## Considerations for Using the Data
125
  This corpus contains speech data and is allowed to be used outside the National Library of Norway for speech recognition technology purposes.
126
 
 
 
 
127
  ### Dataset Curators
128
+ [Freddy Wetjen](mailto:Freddy.wetjen@nb.no) and [Andre Kaasen](mailto:andre.kasen@nb.no) **!!!NEDS TO BE UPDATED!!! Underline the origin of the data and the project. Add Javier de la Rosa and Per Egil Kummervold to the people having worked on this!!!**
129
 
130
  ### Licensing Information
131
  Licensed for use outside the National Library of Norway.
132
 
133
  ## License
134
+ The dataset is released under the [CC-ZERO-license](https://creativecommons.org/publicdomain/zero/1.0/). The curation of the data
135
 
136
  ### Citation Information
137
  We are preparing an article with detailed information about this corpus. Until it is published, please cite out paper discussing the first version of this corpus:
138
  ```
139
+
140
+ **!!!NEDS TO BE UPDATED!!!**
141
  @inproceedings{kummervold-etal-2021-operationalizing,
142
  title = {Operationalizing a National Digital Library: The Case for a {N}orwegian Transformer Model},
143
  author = {Kummervold, Per E and