djovak commited on
Commit
84e0e0c
1 Parent(s): b560103

added tokenizer on top of base model

Browse files
README.md ADDED
@@ -0,0 +1,250 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ language: en
3
+ license: cc-by-4.0
4
+ datasets:
5
+ - squad_v2
6
+ model-index:
7
+ - name: deepset/roberta-base-squad2
8
+ results:
9
+ - task:
10
+ type: question-answering
11
+ name: Question Answering
12
+ dataset:
13
+ name: squad_v2
14
+ type: squad_v2
15
+ config: squad_v2
16
+ split: validation
17
+ metrics:
18
+ - type: exact_match
19
+ value: 79.9309
20
+ name: Exact Match
21
+ verified: true
22
+ verifyToken: eyJhbGciOiJFZERTQSIsInR5cCI6IkpXVCJ9.eyJoYXNoIjoiMDhhNjg5YzNiZGQ1YTIyYTAwZGUwOWEzZTRiYzdjM2QzYjA3ZTUxNDM1NjE1MTUyMjE1MGY1YzEzMjRjYzVjYiIsInZlcnNpb24iOjF9.EH5JJo8EEFwU7osPz3s7qanw_tigeCFhCXjSfyN0Y1nWVnSfulSxIk_DbAEI5iE80V4EKLyp5-mYFodWvL2KDA
23
+ - type: f1
24
+ value: 82.9501
25
+ name: F1
26
+ verified: true
27
+ verifyToken: eyJhbGciOiJFZERTQSIsInR5cCI6IkpXVCJ9.eyJoYXNoIjoiMjk5ZDYwOGQyNjNkMWI0OTE4YzRmOTlkY2JjNjQ0YTZkNTMzMzNkYTA0MDFmNmI3NjA3NjNlMjhiMDQ2ZjJjNSIsInZlcnNpb24iOjF9.DDm0LNTkdLbGsue58bg1aH_s67KfbcmkvL-6ZiI2s8IoxhHJMSf29H_uV2YLyevwx900t-MwTVOW3qfFnMMEAQ
28
+ - type: total
29
+ value: 11869
30
+ name: total
31
+ verified: true
32
+ verifyToken: eyJhbGciOiJFZERTQSIsInR5cCI6IkpXVCJ9.eyJoYXNoIjoiMGFkMmI2ODM0NmY5NGNkNmUxYWViOWYxZDNkY2EzYWFmOWI4N2VhYzY5MGEzMTVhOTU4Zjc4YWViOGNjOWJjMCIsInZlcnNpb24iOjF9.fexrU1icJK5_MiifBtZWkeUvpmFISqBLDXSQJ8E6UnrRof-7cU0s4tX_dIsauHWtUpIHMPZCf5dlMWQKXZuAAA
33
+ - task:
34
+ type: question-answering
35
+ name: Question Answering
36
+ dataset:
37
+ name: squad
38
+ type: squad
39
+ config: plain_text
40
+ split: validation
41
+ metrics:
42
+ - type: exact_match
43
+ value: 85.289
44
+ name: Exact Match
45
+ - type: f1
46
+ value: 91.841
47
+ name: F1
48
+ - task:
49
+ type: question-answering
50
+ name: Question Answering
51
+ dataset:
52
+ name: adversarial_qa
53
+ type: adversarial_qa
54
+ config: adversarialQA
55
+ split: validation
56
+ metrics:
57
+ - type: exact_match
58
+ value: 29.500
59
+ name: Exact Match
60
+ - type: f1
61
+ value: 40.367
62
+ name: F1
63
+ - task:
64
+ type: question-answering
65
+ name: Question Answering
66
+ dataset:
67
+ name: squad_adversarial
68
+ type: squad_adversarial
69
+ config: AddOneSent
70
+ split: validation
71
+ metrics:
72
+ - type: exact_match
73
+ value: 78.567
74
+ name: Exact Match
75
+ - type: f1
76
+ value: 84.469
77
+ name: F1
78
+ - task:
79
+ type: question-answering
80
+ name: Question Answering
81
+ dataset:
82
+ name: squadshifts amazon
83
+ type: squadshifts
84
+ config: amazon
85
+ split: test
86
+ metrics:
87
+ - type: exact_match
88
+ value: 69.924
89
+ name: Exact Match
90
+ - type: f1
91
+ value: 83.284
92
+ name: F1
93
+ - task:
94
+ type: question-answering
95
+ name: Question Answering
96
+ dataset:
97
+ name: squadshifts new_wiki
98
+ type: squadshifts
99
+ config: new_wiki
100
+ split: test
101
+ metrics:
102
+ - type: exact_match
103
+ value: 81.204
104
+ name: Exact Match
105
+ - type: f1
106
+ value: 90.595
107
+ name: F1
108
+ - task:
109
+ type: question-answering
110
+ name: Question Answering
111
+ dataset:
112
+ name: squadshifts nyt
113
+ type: squadshifts
114
+ config: nyt
115
+ split: test
116
+ metrics:
117
+ - type: exact_match
118
+ value: 82.931
119
+ name: Exact Match
120
+ - type: f1
121
+ value: 90.756
122
+ name: F1
123
+ - task:
124
+ type: question-answering
125
+ name: Question Answering
126
+ dataset:
127
+ name: squadshifts reddit
128
+ type: squadshifts
129
+ config: reddit
130
+ split: test
131
+ metrics:
132
+ - type: exact_match
133
+ value: 71.550
134
+ name: Exact Match
135
+ - type: f1
136
+ value: 82.939
137
+ name: F1
138
+ ---
139
+
140
+ # roberta-base for QA
141
+
142
+ This is the [roberta-base](https://huggingface.co/roberta-base) 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.
143
+
144
+
145
+ ## Overview
146
+ **Language model:** roberta-base
147
+ **Language:** English
148
+ **Downstream-task:** Extractive QA
149
+ **Training data:** SQuAD 2.0
150
+ **Eval data:** SQuAD 2.0
151
+ **Code:** See [an example QA pipeline on Haystack](https://haystack.deepset.ai/tutorials/first-qa-system)
152
+ **Infrastructure**: 4x Tesla v100
153
+
154
+ ## Hyperparameters
155
+
156
+ ```
157
+ batch_size = 96
158
+ n_epochs = 2
159
+ base_LM_model = "roberta-base"
160
+ max_seq_len = 386
161
+ learning_rate = 3e-5
162
+ lr_schedule = LinearWarmup
163
+ warmup_proportion = 0.2
164
+ doc_stride=128
165
+ max_query_length=64
166
+ ```
167
+
168
+ ## Using a distilled model instead
169
+ Please note that we have also released a distilled version of this model called [deepset/tinyroberta-squad2](https://huggingface.co/deepset/tinyroberta-squad2). The distilled model has a comparable prediction quality and runs at twice the speed of the base model.
170
+
171
+ ## Usage
172
+
173
+ ### In Haystack
174
+ 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/):
175
+ ```python
176
+ reader = FARMReader(model_name_or_path="deepset/roberta-base-squad2")
177
+ # or
178
+ reader = TransformersReader(model_name_or_path="deepset/roberta-base-squad2",tokenizer="deepset/roberta-base-squad2")
179
+ ```
180
+ 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)
181
+
182
+ ### In Transformers
183
+ ```python
184
+ from transformers import AutoModelForQuestionAnswering, AutoTokenizer, pipeline
185
+
186
+ model_name = "deepset/roberta-base-squad2"
187
+
188
+ # a) Get predictions
189
+ nlp = pipeline('question-answering', model=model_name, tokenizer=model_name)
190
+ QA_input = {
191
+ 'question': 'Why is model conversion important?',
192
+ 'context': 'The option to convert models between FARM and transformers gives freedom to the user and let people easily switch between frameworks.'
193
+ }
194
+ res = nlp(QA_input)
195
+
196
+ # b) Load model & tokenizer
197
+ model = AutoModelForQuestionAnswering.from_pretrained(model_name)
198
+ tokenizer = AutoTokenizer.from_pretrained(model_name)
199
+ ```
200
+
201
+ ## Performance
202
+ Evaluated on the SQuAD 2.0 dev set with the [official eval script](https://worksheets.codalab.org/rest/bundles/0x6b567e1cf2e041ec80d7098f031c5c9e/contents/blob/).
203
+
204
+ ```
205
+ "exact": 79.87029394424324,
206
+ "f1": 82.91251169582613,
207
+
208
+ "total": 11873,
209
+ "HasAns_exact": 77.93522267206478,
210
+ "HasAns_f1": 84.02838248389763,
211
+ "HasAns_total": 5928,
212
+ "NoAns_exact": 81.79983179142137,
213
+ "NoAns_f1": 81.79983179142137,
214
+ "NoAns_total": 5945
215
+ ```
216
+
217
+ ## Authors
218
+ **Branden Chan:** branden.chan@deepset.ai
219
+ **Timo Möller:** timo.moeller@deepset.ai
220
+ **Malte Pietsch:** malte.pietsch@deepset.ai
221
+ **Tanay Soni:** tanay.soni@deepset.ai
222
+
223
+ ## About us
224
+
225
+ <div class="grid lg:grid-cols-2 gap-x-4 gap-y-3">
226
+ <div class="w-full h-40 object-cover mb-2 rounded-lg flex items-center justify-center">
227
+ <img alt="" src="https://raw.githubusercontent.com/deepset-ai/.github/main/deepset-logo-colored.png" class="w-40"/>
228
+ </div>
229
+ <div class="w-full h-40 object-cover mb-2 rounded-lg flex items-center justify-center">
230
+ <img alt="" src="https://raw.githubusercontent.com/deepset-ai/.github/main/haystack-logo-colored.png" class="w-40"/>
231
+ </div>
232
+ </div>
233
+
234
+ [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.
235
+
236
+
237
+ Some of our other work:
238
+ - [Distilled roberta-base-squad2 (aka "tinyroberta-squad2")]([https://huggingface.co/deepset/tinyroberta-squad2)
239
+ - [German BERT (aka "bert-base-german-cased")](https://deepset.ai/german-bert)
240
+ - [GermanQuAD and GermanDPR datasets and models (aka "gelectra-base-germanquad", "gbert-base-germandpr")](https://deepset.ai/germanquad)
241
+
242
+ ## Get in touch and join the Haystack community
243
+
244
+ <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://docs.haystack.deepset.ai">Documentation</a></strong>.
245
+
246
+ We also have a <strong><a class="h-7" href="https://haystack.deepset.ai/community">Discord community open to everyone!</a></strong></p>
247
+
248
+ [Twitter](https://twitter.com/deepset_ai) | [LinkedIn](https://www.linkedin.com/company/deepset-ai/) | [Discord](https://haystack.deepset.ai/community) | [GitHub Discussions](https://github.com/deepset-ai/haystack/discussions) | [Website](https://deepset.ai)
249
+
250
+ By the way: [we're hiring!](http://www.deepset.ai/jobs)
config.json ADDED
@@ -0,0 +1,24 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "RobertaForQuestionAnswering"
4
+ ],
5
+ "attention_probs_dropout_prob": 0.1,
6
+ "bos_token_id": 0,
7
+ "eos_token_id": 2,
8
+ "gradient_checkpointing": false,
9
+ "hidden_act": "gelu",
10
+ "hidden_dropout_prob": 0.1,
11
+ "hidden_size": 768,
12
+ "initializer_range": 0.02,
13
+ "intermediate_size": 3072,
14
+ "language": "english",
15
+ "layer_norm_eps": 1e-05,
16
+ "max_position_embeddings": 514,
17
+ "model_type": "roberta",
18
+ "name": "Roberta",
19
+ "num_attention_heads": 12,
20
+ "num_hidden_layers": 12,
21
+ "pad_token_id": 1,
22
+ "type_vocab_size": 1,
23
+ "vocab_size": 50265
24
+ }
flax_model.msgpack ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6a8d759d881d9c1b39dbf4ee451fb8a8c2d43ccbd180218863a54ffd9b4d2447
3
+ size 496233457
merges.txt ADDED
The diff for this file is too large to render. See raw diff
 
model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ac5db66fdcfecb400345d09787b71009d60805ef9883451071669cf951b5e2c7
3
+ size 496254442
pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e0b64ccefc1bcb569b604baea27eb873e5482fdf6eb3ceff1fb5368397db5aed
3
+ size 496313727
rust_model.ot ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5a16ed126bbc8c4cf794406bac0c7946f62d0f175c02dc54d77a00a6255597ed
3
+ size 498638704
special_tokens_map.json ADDED
@@ -0,0 +1,51 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token": {
3
+ "content": "<s>",
4
+ "lstrip": false,
5
+ "normalized": true,
6
+ "rstrip": false,
7
+ "single_word": false
8
+ },
9
+ "cls_token": {
10
+ "content": "<s>",
11
+ "lstrip": false,
12
+ "normalized": true,
13
+ "rstrip": false,
14
+ "single_word": false
15
+ },
16
+ "eos_token": {
17
+ "content": "</s>",
18
+ "lstrip": false,
19
+ "normalized": true,
20
+ "rstrip": false,
21
+ "single_word": false
22
+ },
23
+ "mask_token": {
24
+ "content": "<mask>",
25
+ "lstrip": true,
26
+ "normalized": true,
27
+ "rstrip": false,
28
+ "single_word": false
29
+ },
30
+ "pad_token": {
31
+ "content": "<pad>",
32
+ "lstrip": false,
33
+ "normalized": true,
34
+ "rstrip": false,
35
+ "single_word": false
36
+ },
37
+ "sep_token": {
38
+ "content": "</s>",
39
+ "lstrip": false,
40
+ "normalized": true,
41
+ "rstrip": false,
42
+ "single_word": false
43
+ },
44
+ "unk_token": {
45
+ "content": "<unk>",
46
+ "lstrip": false,
47
+ "normalized": true,
48
+ "rstrip": false,
49
+ "single_word": false
50
+ }
51
+ }
tf_model.h5 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9b672dd16f09f6f805d407800278e60217b9d7c040df1dde5098765a40cdc88a
3
+ size 496513256
tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
tokenizer_config.json ADDED
@@ -0,0 +1,59 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_prefix_space": false,
3
+ "added_tokens_decoder": {
4
+ "0": {
5
+ "content": "<s>",
6
+ "lstrip": false,
7
+ "normalized": true,
8
+ "rstrip": false,
9
+ "single_word": false,
10
+ "special": true
11
+ },
12
+ "1": {
13
+ "content": "<pad>",
14
+ "lstrip": false,
15
+ "normalized": true,
16
+ "rstrip": false,
17
+ "single_word": false,
18
+ "special": true
19
+ },
20
+ "2": {
21
+ "content": "</s>",
22
+ "lstrip": false,
23
+ "normalized": true,
24
+ "rstrip": false,
25
+ "single_word": false,
26
+ "special": true
27
+ },
28
+ "3": {
29
+ "content": "<unk>",
30
+ "lstrip": false,
31
+ "normalized": true,
32
+ "rstrip": false,
33
+ "single_word": false,
34
+ "special": true
35
+ },
36
+ "50264": {
37
+ "content": "<mask>",
38
+ "lstrip": true,
39
+ "normalized": true,
40
+ "rstrip": false,
41
+ "single_word": false,
42
+ "special": true
43
+ }
44
+ },
45
+ "bos_token": "<s>",
46
+ "clean_up_tokenization_spaces": true,
47
+ "cls_token": "<s>",
48
+ "do_lower_case": false,
49
+ "eos_token": "</s>",
50
+ "errors": "replace",
51
+ "full_tokenizer_file": null,
52
+ "mask_token": "<mask>",
53
+ "model_max_length": 512,
54
+ "pad_token": "<pad>",
55
+ "sep_token": "</s>",
56
+ "tokenizer_class": "RobertaTokenizer",
57
+ "trim_offsets": true,
58
+ "unk_token": "<unk>"
59
+ }
vocab.json ADDED
The diff for this file is too large to render. See raw diff