Update README.md
Browse files
README.md
CHANGED
@@ -10,40 +10,30 @@ widget:
|
|
10 |
example_title: "Programming"
|
11 |
---
|
12 |
|
13 |
-
|
14 |
-
should probably proofread and complete it, then remove this comment. -->
|
15 |
|
16 |
-
|
17 |
|
18 |
-
|
19 |
-
It achieves the following results on the evaluation set:
|
20 |
-
- eval_loss: 0.8474
|
21 |
-
- eval_rouge1: 38.6033
|
22 |
-
- eval_rouge2: 20.5952
|
23 |
-
- eval_rougeL: 36.4458
|
24 |
-
- eval_rougeLsum: 36.3202
|
25 |
-
- eval_gen_len: 15.257
|
26 |
-
- eval_runtime: 343.6547
|
27 |
-
- eval_samples_per_second: 2.91
|
28 |
-
- eval_steps_per_second: 0.364
|
29 |
-
- epoch: 0.31
|
30 |
-
- step: 2000
|
31 |
-
|
32 |
-
## Model description
|
33 |
-
|
34 |
-
More information needed
|
35 |
|
36 |
-
|
37 |
|
38 |
-
|
39 |
|
40 |
## Training and evaluation data
|
41 |
|
42 |
-
|
43 |
|
44 |
-
##
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
45 |
|
46 |
-
|
47 |
|
48 |
The following hyperparameters were used during training:
|
49 |
- learning_rate: 4e-05
|
|
|
10 |
example_title: "Programming"
|
11 |
---
|
12 |
|
13 |
+
# Model description
|
|
|
14 |
|
15 |
+
This model is [t5-base](https://huggingface.co/t5-base) fine-tuned on the [190k Medium Articles](https://www.kaggle.com/datasets/fabiochiusano/medium-articles) dataset for predicting article tags using the article textual content as input.
|
16 |
|
17 |
+
## Data cleaning
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
18 |
|
19 |
+
The dataset is composed of Medium articles and their tags. However, each Medium article can have at most five tags, therefore the author needs to choose what he/she believes are the best tags (mainly for SEO-related purposes). This means that an article with the "Python" tag may have not the "Programming Languages" tag, even though the first implies the latter.
|
20 |
|
21 |
+
To clean the dataset accounting for this problem, a hand-made taxonomy of about 1000 tags was built. Using the taxonomy, the tags of each articles have been augmented (e.g. an article with the "Python" tag will have the "Programming Languages" tag as well, as the taxonomy says that "Python" is part of "Programming Languages"). The taxonomy is not public, if you are interested in it please send an email at [email protected].
|
22 |
|
23 |
## Training and evaluation data
|
24 |
|
25 |
+
The model has been trained on a single epoch spanning about 50000 articles, evaluating on 1000 random articles not used during training.
|
26 |
|
27 |
+
## Evaluation results
|
28 |
+
|
29 |
+
- eval_loss: 0.8474
|
30 |
+
- eval_rouge1: 38.6033
|
31 |
+
- eval_rouge2: 20.5952
|
32 |
+
- eval_rougeL: 36.4458
|
33 |
+
- eval_rougeLsum: 36.3202
|
34 |
+
- eval_gen_len: 15.257 # average number of generated tokens
|
35 |
|
36 |
+
## Training hyperparameters
|
37 |
|
38 |
The following hyperparameters were used during training:
|
39 |
- learning_rate: 4e-05
|