krotima1 commited on
Commit
2e0b72d
1 Parent(s): 813b970

feat: add readme.md

Browse files
Files changed (1) hide show
  1. README.md +82 -0
README.md CHANGED
@@ -1,3 +1,85 @@
1
  ---
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
2
  license: cc-by-sa-4.0
 
 
 
 
 
 
 
 
 
 
 
 
3
  ---
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
  ---
2
+ language:
3
+ - cs
4
+ - en
5
+ - de
6
+ - fr
7
+ - tu
8
+ - zh
9
+ - es
10
+ - ru
11
+ tags:
12
+ - Summarization
13
+ - abstractive summarization
14
+ - mt5-base
15
+ - Czech
16
+ - text2text generation
17
+ - text generation
18
  license: cc-by-sa-4.0
19
+ datasets:
20
+ - Multilingual_large_dataset_(multilarge)
21
+ - cnc/dm
22
+ - xsum
23
+ - mlsum
24
+ - cnewsum
25
+ - cnc
26
+ - sumeczech
27
+ metrics:
28
+ - rouge
29
+ - rougeraw
30
+ - MemesCS
31
  ---
32
+ # mbart25-multilingual-summarization-multilarge-cs
33
+ This model is a fine-tuned checkpoint of [google/mt5-base](https://huggingface.co/google/mt5-base) on the Multilingual large summarization dataset focused on Czech texts to produce multilingual summaries.
34
+ ## Task
35
+ The model deals with a multi-sentence summary in eight different languages. With the idea of adding other foreign language documents, and by having a considerable amount of Czech documents, we aimed to improve model summarization in the Czech language. Supported languages: ''cs': '<extra_id_0>', 'en': '<extra_id_1>','de': '<extra_id_2>', 'es': '<extra_id_3>', 'fr': '<extra_id_4>', 'ru': '<extra_id_5>', 'tu': '<extra_id_6>', 'zh': '<extra_id_7>'
36
+ ## Dataset
37
+ Multilingual large summarization dataset consists of 10 sub-datasets mainly based on news and daily mails. For the training, it was used the entire training set and 72% of the validation set.
38
+ ```
39
+ Train set: 3 464 563 docs
40
+ Validation set: 121 260 docs
41
+ ```
42
+ | Stats | fragment | | | avg document length | | avg summary length | | Documents |
43
+ |-------------|----------|---------------------|--------------------|--------|---------|--------|--------|--------|
44
+ | __dataset__ |__compression__ | __density__ | __coverage__ | __nsent__ | __nwords__ | __nsent__ | __nwords__ | __count__ |
45
+ | cnc | 7.388 | 0.303 | 0.088 | 16.121 | 316.912 | 3.272 | 46.805 | 750K |
46
+ | sumeczech | 11.769 | 0.471 | 0.115 | 27.857 | 415.711 | 2.765 | 38.644 | 1M |
47
+ | cnndm | 13.688 | 2.983 | 0.538 | 32.783 | 676.026 | 4.134 | 54.036 | 300K |
48
+ | xsum | 18.378 | 0.479 | 0.194 | 18.607 | 369.134 | 1.000 | 21.127 | 225K|
49
+ | mlsum/tu | 8.666 | 5.418 | 0.461 | 14.271 | 214.496 | 1.793 | 25.675 | 274K |
50
+ | mlsum/de | 24.741 | 8.235 | 0.469 | 32.544 | 539.653 | 1.951 | 23.077 | 243K|
51
+ | mlsum/fr | 24.388 | 2.688 | 0.424 | 24.533 | 612.080 | 1.320 | 26.93 | 425K |
52
+ | mlsum/es | 36.185 | 3.705 | 0.510 | 31.914 | 746.927 | 1.142 | 21.671 | 291K |
53
+ | mlsum/ru | 78.909 | 1.194 | 0.246 | 62.141 | 948.079 | 1.012 | 11.976 | 27K|
54
+ | cnewsum | 20.183 | 0.000 | 0.000 | 16.834 | 438.271 | 1.109 | 21.926 | 304K |
55
+ #### Tokenization
56
+ Truncation and padding were set to 512 tokens for the encoder (input text) and 128 for the decoder (summary).
57
+ ## Training
58
+ Trained based on cross-entropy loss.
59
+ ```
60
+ Time: 3 days 20 hours
61
+ Epochs: 1080K steps = 10 (from 10)
62
+ GPUs: 4x NVIDIA A100-SXM4-40GB
63
+ eloss: 2.462 - 1.797
64
+ tloss: 17.322 - 1.578
65
+ ```
66
+ ### ROUGE results per individual dataset test set:
67
+
68
+ | ROUGE | ROUGE-1 | | | ROUGE-2 | | | ROUGE-L | | |
69
+ |-----------|---------|---------|-----------|--------|--------|-----------|--------|--------|---------|
70
+ | |Precision | Recall | Fscore | Precision | Recall | Fscore | Precision | Recall | Fscore |
71
+ | cnc | 30.62 | 19.83 | 23.44 | 9.94 | 6.52 | 7.67 | 22.92 | 14.92 | 17.6 |
72
+ | sumeczech | 27.57 | 17.6 | 20.85 | 8.12 | 5.23 | 6.17 | 20.84 | 13.38 | 15.81 |
73
+ | cnndm | 43.83 | 37.73 | 39.34 | 20.81 | 17.82 | 18.6 | 31.8 | 27.42 | 28.55 |
74
+ | xsum | 41.63 | 30.54 | 34.56 | 16.13 | 11.76 | 13.33 | 33.65 | 24.74 | 27.97 |
75
+ | mlsum-tu- | 54.4 | 43.29 | 46.2 | 38.78 | 31.31 | 33.23 | 48.18 | 38.44 | 41 |
76
+ | mlsum-de | 47.94 | 44.14 | 45.11 | 36.42 | 35.24 | 35.42 | 44.43 | 41.42 | 42.16 |
77
+ | mlsum-fr | 35.26 | 25.96 | 28.98 | 16.72 | 12.35 | 13.75 | 28.06 | 20.75 | 23.12 |
78
+ | mlsum-es | 33.37 | 24.84 | 27.52 | 13.29 | 10.05 | 11.05 | 27.63 | 20.69 | 22.87 |
79
+ | mlsum-ru | 0.79 | 0.66 | 0.66 | 0.26 | 0.2 | 0.22 | 0.79 | 0.66 | 0.65 |
80
+ | cnewsum | 24.49 | 24.38 | 23.23 | 6.48 | 6.7 | 6.24 | 24.18 | 24.04 | 22.91 |
81
+
82
+ # USAGE
83
+ ```
84
+ soon
85
+ ```