MichelBartelsDeepset commited on
Commit
de73c83
1 Parent(s): 7620389

Create README.md

Browse files
Files changed (1) hide show
  1. README.md +112 -0
README.md ADDED
@@ -0,0 +1,112 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ language: en
3
+ datasets:
4
+ - squad_v2
5
+ license: cc-by-4.0
6
+ tags:
7
+ - deberta
8
+ - deberta-v3
9
+ - deberta-v3-large
10
+ ---
11
+ # microsoft/deberta-v3-large for QA
12
+
13
+ This is the [deberta-v3-large](https://huggingface.co/microsoft/deberta-v3-large) model, fine-tuned using the [SQuAD2.0](https://huggingface.co/datasets/squad_v2) dataset. It's been trained on question-answer pairs, including unanswerable questions, for the task of Question Answering.
14
+
15
+
16
+ ## Overview
17
+ **Language model:** roberta-base
18
+ **Language:** English
19
+ **Downstream-task:** Extractive QA
20
+ **Training data:** SQuAD 2.0
21
+ **Eval data:** SQuAD 2.0
22
+ **Code:** See [an example QA pipeline on Haystack](https://haystack.deepset.ai/tutorials/first-qa-system)
23
+ **Infrastructure**: 1x NVIDIA A10G
24
+
25
+ ## Hyperparameters
26
+
27
+ ```
28
+ batch_size = 2
29
+ grad_acc_steps = 32
30
+ n_epochs = 6
31
+ base_LM_model = "microsoft/deberta-v3-large"
32
+ max_seq_len = 512
33
+ learning_rate = 7e-6
34
+ lr_schedule = LinearWarmup
35
+ warmup_proportion = 0.2
36
+ doc_stride=128
37
+ max_query_length=64
38
+ ```
39
+
40
+ ## Usage
41
+
42
+ ### In Haystack
43
+ Haystack is an NLP framework by deepset. You can use this model in a Haystack pipeline to do question answering at scale (over many documents). To load the model in [Haystack](https://github.com/deepset-ai/haystack/):
44
+ ```python
45
+ reader = FARMReader(model_name_or_path="deepset/roberta-base-squad2")
46
+ # or
47
+ reader = TransformersReader(model_name_or_path="deepset/roberta-base-squad2",tokenizer="deepset/roberta-base-squad2")
48
+ ```
49
+ For a complete example of ``roberta-base-squad2`` being used for Question Answering, check out the [Tutorials in Haystack Documentation](https://haystack.deepset.ai/tutorials/first-qa-system)
50
+
51
+ ### In Transformers
52
+ ```python
53
+ from transformers import AutoModelForQuestionAnswering, AutoTokenizer, pipeline
54
+
55
+ model_name = "deepset/deberta-v3-large-squad2"
56
+
57
+ # a) Get predictions
58
+ nlp = pipeline('question-answering', model=model_name, tokenizer=model_name)
59
+ QA_input = {
60
+ 'question': 'Why is model conversion important?',
61
+ 'context': 'The option to convert models between FARM and transformers gives freedom to the user and let people easily switch between frameworks.'
62
+ }
63
+ res = nlp(QA_input)
64
+
65
+ # b) Load model & tokenizer
66
+ model = AutoModelForQuestionAnswering.from_pretrained(model_name)
67
+ tokenizer = AutoTokenizer.from_pretrained(model_name)
68
+ ```
69
+
70
+ ## Performance
71
+ Evaluated on the SQuAD 2.0 dev set with the [official eval script](https://worksheets.codalab.org/rest/bundles/0x6b567e1cf2e041ec80d7098f031c5c9e/contents/blob/).
72
+
73
+ ```
74
+ "exact": 87.6105449338836,
75
+ "f1": 90.75307008866517,
76
+
77
+ "total": 11873,
78
+ "HasAns_exact": 84.37921727395411,
79
+ "HasAns_f1": 90.6732795483674,
80
+ "HasAns_total": 5928,
81
+ "NoAns_exact": 90.83263246425568,
82
+ "NoAns_f1": 90.83263246425568,
83
+ "NoAns_total": 5945
84
+ ```
85
+
86
+ ## About us
87
+ <div class="grid lg:grid-cols-2 gap-x-4 gap-y-3">
88
+ <div class="w-full h-40 object-cover mb-2 rounded-lg flex items-center justify-center">
89
+ <img alt="" src="https://huggingface.co/spaces/deepset/README/resolve/main/haystack-logo-colored.svg" class="w-40"/>
90
+ </div>
91
+ <div class="w-full h-40 object-cover mb-2 rounded-lg flex items-center justify-center">
92
+ <img alt="" src="https://huggingface.co/spaces/deepset/README/resolve/main/deepset-logo-colored.svg" class="w-40"/>
93
+ </div>
94
+ </div>
95
+
96
+ [deepset](http://deepset.ai/) is the company behind the open-source NLP framework [Haystack](https://haystack.deepset.ai/) which is designed to help you build production ready NLP systems that use: Question answering, summarization, ranking etc.
97
+
98
+
99
+ Some of our other work:
100
+ - [Distilled roberta-base-squad2 (aka "tinyroberta-squad2")]([https://huggingface.co/deepset/tinyroberta-squad2)
101
+ - [German BERT (aka "bert-base-german-cased")](https://deepset.ai/german-bert)
102
+ - [GermanQuAD and GermanDPR datasets and models (aka "gelectra-base-germanquad", "gbert-base-germandpr")](https://deepset.ai/germanquad)
103
+
104
+ ## Get in touch and join the Haystack community
105
+
106
+ <p>For more info on Haystack, visit our <strong><a href="https://github.com/deepset-ai/haystack">GitHub</a></strong> repo and <strong><a href="https://haystack.deepset.ai">Documentation</a></strong>.
107
+
108
+ We also have a <strong><a class="h-7" href="https://haystack.deepset.ai/community/join"><img alt="slack" class="h-7 inline-block m-0" style="margin: 0" src="https://huggingface.co/spaces/deepset/README/resolve/main/Slack_RGB.png"/>community open to everyone!</a></strong></p>
109
+
110
+ [Twitter](https://twitter.com/deepset_ai) | [LinkedIn](https://www.linkedin.com/company/deepset-ai/) | [Slack](https://haystack.deepset.ai/community/join) | [GitHub Discussions](https://github.com/deepset-ai/haystack/discussions) | [Website](https://deepset.ai)
111
+
112
+ By the way: [we're hiring!](http://www.deepset.ai/jobs)