Commit
•
14ecb48
1
Parent(s):
b1e0772
Update README.md
Browse files
README.md
CHANGED
@@ -10,10 +10,16 @@ license: mit
|
|
10 |
metrics:
|
11 |
- f1
|
12 |
widget:
|
13 |
-
- text:
|
14 |
-
|
15 |
-
|
16 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
17 |
---
|
18 |
|
19 |
# British Library Books Genre Detector
|
@@ -119,12 +125,11 @@ If you intend to use this model beyond initial experimentation, it is highly rec
|
|
119 |
|
120 |
|
121 |
## Training data
|
122 |
-
The training data for this model will soon be available from the British Libary Research Repository. This section will be updated once this dataset is made public.
|
123 |
|
124 |
-
The training data was created using the [Zooniverse platform](zooniverse.org/) and the annotations were done by cataloguers from the [British Library](https://www.bl.uk/). [Snorkel](https://github.com/snorkel-team/snorkel) was used to expand on this original training data through various labelling functions. As a result, some of the labels are *not* generated by a human. More information on the process of creating the annotations
|
125 |
|
126 |
## Training procedure
|
127 |
-
The model was trained using the [`blurr`](https://github.com/ohmeow/blurr) library. A notebook showing the training process
|
128 |
|
129 |
## Eval results
|
130 |
|
|
|
10 |
metrics:
|
11 |
- f1
|
12 |
widget:
|
13 |
+
- text: >-
|
14 |
+
Poems on various subjects. Whereto is prefixed a short essay on the
|
15 |
+
structure of English verse
|
16 |
+
- text: >-
|
17 |
+
Two Centuries of Soho: its institutions, firms, and amusements. By the
|
18 |
+
Clergy of St. Anne's, Soho, J. H. Cardwell ... H. B. Freeman ... G. C.
|
19 |
+
Wilton ... assisted by other contributors, etc
|
20 |
+
- text: The Adventures of Oliver Twist. [With plates.]
|
21 |
+
datasets:
|
22 |
+
- blbooksgenre
|
23 |
---
|
24 |
|
25 |
# British Library Books Genre Detector
|
|
|
125 |
|
126 |
|
127 |
## Training data
|
|
|
128 |
|
129 |
+
The training data was created using the [Zooniverse platform](zooniverse.org/) and the annotations were done by cataloguers from the [British Library](https://www.bl.uk/). [Snorkel](https://github.com/snorkel-team/snorkel) was used to expand on this original training data through various labelling functions. As a result, some of the labels are *not* generated by a human. More information on the process of creating the annotations can be found [here](https://github.com/Living-with-machines/genre-classification)
|
130 |
|
131 |
## Training procedure
|
132 |
+
The model was trained using the [`blurr`](https://github.com/ohmeow/blurr) library. A notebook showing the training process can be found in [Predicting Genre with Machine Learning](https://github.com/Living-with-machines/genre-classification).
|
133 |
|
134 |
## Eval results
|
135 |
|