Hugging Face's logo

danielpark
/
sentence-transformers-all-mini-lm-l6-v2

Sentence Similarity
sentence-transformers
PyTorch
English
bert
feature-extraction
text-embeddings-inference
Inference Endpoints
Model card Files Community
danielpark commited on
Commit
d97ecc6
•
1 Parent(s): 8c6ed75

Create README.md

Browse files
Files changed (1) hide show
  1. README.md +91 -0
README.md ADDED
@@ -0,0 +1,91 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ pipeline_tag: sentence-similarity
3
+ tags:
4
+ - sentence-transformers
5
+ - feature-extraction
6
+ - sentence-similarity
7
+ language: en
8
+ license: apache-2.0
9
+ datasets:
10
+ - s2orc
11
+ - flax-sentence-embeddings/stackexchange_xml
12
+ - ms_marco
13
+ - gooaq
14
+ - yahoo_answers_topics
15
+ - code_search_net
16
+ - search_qa
17
+ - eli5
18
+ - snli
19
+ - multi_nli
20
+ - wikihow
21
+ - natural_questions
22
+ - trivia_qa
23
+ - embedding-data/sentence-compression
24
+ - embedding-data/flickr30k-captions
25
+ - embedding-data/altlex
26
+ - embedding-data/simple-wiki
27
+ - embedding-data/QQP
28
+ - embedding-data/SPECTER
29
+ - embedding-data/PAQ_pairs
30
+ - embedding-data/WikiAnswers
31
+
32
+ ---
33
+
34
+
35
+ # Sentence Transformers
36
+ We are forking sentence-transformers/all-MiniLM-L6-v2 as it is similar to the targeting dataset and use case. For more details, please check the pre-trained model weight repository.
37
+ - https://huggingface.co/sentence-transformers/all-MiniLM-L6-v2
38
+ - Commit Hash: [7dbbc90392e2f80f3d3c277d6e90027e55de9125](https://huggingface.co/sentence-transformers/all-MiniLM-L6-v2/tree/7dbbc90392e2f80f3d3c277d6e90027e55de9125)
39
+
40
+
41
+
42
+
43
+ ### Fine-tuning
44
+ - Fine-tune the model using a contrastive objective.
45
+ - Compute the cosine similarity from each possible sentence pairs from the batch.
46
+ - Then apply the cross entropy loss by comparing with true pairs.
47
+
48
+ #### Hyper parameters
49
+
50
+ - Train the model during 100k steps using a batch size of 1024 (128 per TPU core).
51
+ - Use a learning rate warm up of 500.
52
+ - The sequence length was limited to 128 tokens.
53
+ - Used the AdamW optimizer with a 2e-5 learning rate.
54
+ - The full training script is accessible in this current repository: `train_script.py`.
55
+
56
+ #### Datasets
57
+ | Dataset | Paper | Number of training tuples |
58
+ |--------------------------------------------------------|:----------------------------------------:|:--------------------------:|
59
+ | [Reddit comments (2015-2018)](https://github.com/PolyAI-LDN/conversational-datasets/tree/master/reddit) | [paper](https://arxiv.org/abs/1904.06472) | 726,484,430 |
60
+ | [S2ORC](https://github.com/allenai/s2orc) Citation pairs (Abstracts) | [paper](https://aclanthology.org/2020.acl-main.447/) | 116,288,806 |
61
+ | [WikiAnswers](https://github.com/afader/oqa#wikianswers-corpus) Duplicate question pairs | [paper](https://doi.org/10.1145/2623330.2623677) | 77,427,422 |
62
+ | [PAQ](https://github.com/facebookresearch/PAQ) (Question, Answer) pairs | [paper](https://arxiv.org/abs/2102.07033) | 64,371,441 |
63
+ | [S2ORC](https://github.com/allenai/s2orc) Citation pairs (Titles) | [paper](https://aclanthology.org/2020.acl-main.447/) | 52,603,982 |
64
+ | [S2ORC](https://github.com/allenai/s2orc) (Title, Abstract) | [paper](https://aclanthology.org/2020.acl-main.447/) | 41,769,185 |
65
+ | [Stack Exchange](https://huggingface.co/datasets/flax-sentence-embeddings/stackexchange_xml) (Title, Body) pairs | - | 25,316,456 |
66
+ | [Stack Exchange](https://huggingface.co/datasets/flax-sentence-embeddings/stackexchange_xml) (Title+Body, Answer) pairs | - | 21,396,559 |
67
+ | [Stack Exchange](https://huggingface.co/datasets/flax-sentence-embeddings/stackexchange_xml) (Title, Answer) pairs | - | 21,396,559 |
68
+ | [MS MARCO](https://microsoft.github.io/msmarco/) triplets | [paper](https://doi.org/10.1145/3404835.3462804) | 9,144,553 |
69
+ | [GOOAQ: Open Question Answering with Diverse Answer Types](https://github.com/allenai/gooaq) | [paper](https://arxiv.org/pdf/2104.08727.pdf) | 3,012,496 |
70
+ | [Yahoo Answers](https://www.kaggle.com/soumikrakshit/yahoo-answers-dataset) (Title, Answer) | [paper](https://proceedings.neurips.cc/paper/2015/hash/250cf8b51c773f3f8dc8b4be867a9a02-Abstract.html) | 1,198,260 |
71
+ | [Code Search](https://huggingface.co/datasets/code_search_net) | - | 1,151,414 |
72
+ | [COCO](https://cocodataset.org/#home) Image captions | [paper](https://link.springer.com/chapter/10.1007%2F978-3-319-10602-1_48) | 828,395|
73
+ | [SPECTER](https://github.com/allenai/specter) citation triplets | [paper](https://doi.org/10.18653/v1/2020.acl-main.207) | 684,100 |
74
+ | [Yahoo Answers](https://www.kaggle.com/soumikrakshit/yahoo-answers-dataset) (Question, Answer) | [paper](https://proceedings.neurips.cc/paper/2015/hash/250cf8b51c773f3f8dc8b4be867a9a02-Abstract.html) | 681,164 |
75
+ | [Yahoo Answers](https://www.kaggle.com/soumikrakshit/yahoo-answers-dataset) (Title, Question) | [paper](https://proceedings.neurips.cc/paper/2015/hash/250cf8b51c773f3f8dc8b4be867a9a02-Abstract.html) | 659,896 |
76
+ | [SearchQA](https://huggingface.co/datasets/search_qa) | [paper](https://arxiv.org/abs/1704.05179) | 582,261 |
77
+ | [Eli5](https://huggingface.co/datasets/eli5) | [paper](https://doi.org/10.18653/v1/p19-1346) | 325,475 |
78
+ | [Flickr 30k](https://shannon.cs.illinois.edu/DenotationGraph/) | [paper](https://transacl.org/ojs/index.php/tacl/article/view/229/33) | 317,695 |
79
+ | [Stack Exchange](https://huggingface.co/datasets/flax-sentence-embeddings/stackexchange_xml) Duplicate questions (titles) | | 304,525 |
80
+ | AllNLI ([SNLI](https://nlp.stanford.edu/projects/snli/) and [MultiNLI](https://cims.nyu.edu/~sbowman/multinli/) | [paper SNLI](https://doi.org/10.18653/v1/d15-1075), [paper MultiNLI](https://doi.org/10.18653/v1/n18-1101) | 277,230 |
81
+ | [Stack Exchange](https://huggingface.co/datasets/flax-sentence-embeddings/stackexchange_xml) Duplicate questions (bodies) | | 250,519 |
82
+ | [Stack Exchange](https://huggingface.co/datasets/flax-sentence-embeddings/stackexchange_xml) Duplicate questions (titles+bodies) | | 250,460 |
83
+ | [Sentence Compression](https://github.com/google-research-datasets/sentence-compression) | [paper](https://www.aclweb.org/anthology/D13-1155/) | 180,000 |
84
+ | [Wikihow](https://github.com/pvl/wikihow_pairs_dataset) | [paper](https://arxiv.org/abs/1810.09305) | 128,542 |
85
+ | [Altlex](https://github.com/chridey/altlex/) | [paper](https://aclanthology.org/P16-1135.pdf) | 112,696 |
86
+ | [Quora Question Triplets](https://quoradata.quora.com/First-Quora-Dataset-Release-Question-Pairs) | - | 103,663 |
87
+ | [Simple Wikipedia](https://cs.pomona.edu/~dkauchak/simplification/) | [paper](https://www.aclweb.org/anthology/P11-2117/) | 102,225 |
88
+ | [Natural Questions (NQ)](https://ai.google.com/research/NaturalQuestions) | [paper](https://transacl.org/ojs/index.php/tacl/article/view/1455) | 100,231 |
89
+ | [SQuAD2.0](https://rajpurkar.github.io/SQuAD-explorer/) | [paper](https://aclanthology.org/P18-2124.pdf) | 87,599 |
90
+ | [TriviaQA](https://huggingface.co/datasets/trivia_qa) | - | 73,346 |
91
+ | **Total** | | **1,170,060,424** |