Thomas Lemberger commited on
Commit ·
2595e1f
1
Parent(s): 90be2bd
dataset card
Browse files- README.md +40 -11
- dataset_infos.json +284 -1
README.md
CHANGED
|
@@ -1,9 +1,24 @@
|
|
| 1 |
---
|
| 2 |
-
|
| 3 |
-
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 4 |
---
|
| 5 |
|
| 6 |
-
# Dataset Card for
|
| 7 |
|
| 8 |
## Table of Contents
|
| 9 |
- [Dataset Card for [Dataset Name]](#dataset-card-for-dataset-name)
|
|
@@ -37,15 +52,16 @@ YAML tags:
|
|
| 37 |
|
| 38 |
## Dataset Description
|
| 39 |
|
| 40 |
-
- **Homepage:**
|
| 41 |
-
- **Repository:**
|
| 42 |
- **Paper:**
|
| 43 |
- **Leaderboard:**
|
| 44 |
-
- **Point of Contact:**
|
|
|
|
| 45 |
|
| 46 |
### Dataset Summary
|
| 47 |
|
| 48 |
-
|
| 49 |
|
| 50 |
### Supported Tasks and Leaderboards
|
| 51 |
|
|
@@ -80,13 +96,26 @@ English
|
|
| 80 |
|
| 81 |
### Data Fields
|
| 82 |
|
| 83 |
-
`
|
| 84 |
-
`
|
| 85 |
-
`
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 86 |
|
| 87 |
### Data Splits
|
| 88 |
|
| 89 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 90 |
|
| 91 |
## Dataset Creation
|
| 92 |
|
|
|
|
| 1 |
---
|
| 2 |
+
annotations_creators:
|
| 3 |
+
- machine-generated
|
| 4 |
+
language_creators:
|
| 5 |
+
- expert-generated
|
| 6 |
+
languages:
|
| 7 |
+
- en
|
| 8 |
+
licenses:
|
| 9 |
+
- cc-by-4.0
|
| 10 |
+
multilinguality:
|
| 11 |
+
- monolingual
|
| 12 |
+
size_categories:
|
| 13 |
+
- n>1M
|
| 14 |
+
source_datasets: []
|
| 15 |
+
task_categories:
|
| 16 |
+
- sequence-modeling
|
| 17 |
+
task_ids:
|
| 18 |
+
- language-modeling
|
| 19 |
---
|
| 20 |
|
| 21 |
+
# Dataset Card for BioLang
|
| 22 |
|
| 23 |
## Table of Contents
|
| 24 |
- [Dataset Card for [Dataset Name]](#dataset-card-for-dataset-name)
|
|
|
|
| 52 |
|
| 53 |
## Dataset Description
|
| 54 |
|
| 55 |
+
- **Homepage:** https://sourcedata.embo.org
|
| 56 |
+
- **Repository:** https://github.com/source-data/soda-roberta
|
| 57 |
- **Paper:**
|
| 58 |
- **Leaderboard:**
|
| 59 |
+
- **Point of Contact:** thomas.lemberger@embo.org
|
| 60 |
+
- **Download Size:** 5_299_878_661
|
| 61 |
|
| 62 |
### Dataset Summary
|
| 63 |
|
| 64 |
+
BioLang is a dataset is based on abstracts from the open access section of EuropePubMed Central to train language models in the domain of biology. The dataset can be used for random masked language modeling or for language modeling using only specific part-of-speech maksing. More details on generation and use of the dataset at https://github.com/source-data/soda-roberta .
|
| 65 |
|
| 66 |
### Supported Tasks and Leaderboards
|
| 67 |
|
|
|
|
| 96 |
|
| 97 |
### Data Fields
|
| 98 |
|
| 99 |
+
`MLM`:
|
| 100 |
+
- `input_ids`: a `list` of `int32` features.
|
| 101 |
+
- `special_tokens_mask`: a `list` of `int8` features.
|
| 102 |
+
|
| 103 |
+
`DET`, `VERB`, `SMALL`:
|
| 104 |
+
- `input_ids`: a `list` of `int32` features.
|
| 105 |
+
- `tag_mask`: a `list` of `int8` features.
|
| 106 |
+
|
| 107 |
|
| 108 |
### Data Splits
|
| 109 |
|
| 110 |
+
- train:
|
| 111 |
+
- features: ['input_ids', 'special_tokens_mask'],
|
| 112 |
+
- num_rows: 12_005_390
|
| 113 |
+
- test:
|
| 114 |
+
- features: ['input_ids', 'special_tokens_mask'],
|
| 115 |
+
- num_rows: 37_112
|
| 116 |
+
- validation:
|
| 117 |
+
- features: ['input_ids', 'special_tokens_mask'],
|
| 118 |
+
- num_rows: 36_713
|
| 119 |
|
| 120 |
## Dataset Creation
|
| 121 |
|
dataset_infos.json
CHANGED
|
@@ -1 +1,284 @@
|
|
| 1 |
-
{
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"MLM": {
|
| 3 |
+
"description": "This dataset is based on abstracts from the open access section of EuropePubMed Central to train language models in the domain of biology. \n",
|
| 4 |
+
"citation": "@Unpublished{\n huggingface: dataset,\n title = {biolang},\n authors={Thomas Lemberger, EMBO},\n year={2021}\n}\n",
|
| 5 |
+
"homepage": "https://europepmc.org/downloads/openaccess",
|
| 6 |
+
"license": "CC BY 4.0",
|
| 7 |
+
"features": {
|
| 8 |
+
"input_ids": {
|
| 9 |
+
"feature": {
|
| 10 |
+
"dtype": "int32",
|
| 11 |
+
"id": null,
|
| 12 |
+
"_type": "Value"
|
| 13 |
+
},
|
| 14 |
+
"length": -1,
|
| 15 |
+
"id": null,
|
| 16 |
+
"_type": "Sequence"
|
| 17 |
+
},
|
| 18 |
+
"special_tokens_mask": {
|
| 19 |
+
"feature": {
|
| 20 |
+
"dtype": "int8",
|
| 21 |
+
"id": null,
|
| 22 |
+
"_type": "Value"
|
| 23 |
+
},
|
| 24 |
+
"length": -1,
|
| 25 |
+
"id": null,
|
| 26 |
+
"_type": "Sequence"
|
| 27 |
+
}
|
| 28 |
+
},
|
| 29 |
+
"post_processed": null,
|
| 30 |
+
"supervised_keys": {
|
| 31 |
+
"input": "input_ids",
|
| 32 |
+
"output": "pos_mask"
|
| 33 |
+
},
|
| 34 |
+
"builder_name": "bio_lang",
|
| 35 |
+
"config_name": "MLM",
|
| 36 |
+
"version": {
|
| 37 |
+
"version_str": "0.0.1",
|
| 38 |
+
"description": null,
|
| 39 |
+
"major": 0,
|
| 40 |
+
"minor": 0,
|
| 41 |
+
"patch": 1
|
| 42 |
+
},
|
| 43 |
+
"splits": {
|
| 44 |
+
"train": {
|
| 45 |
+
"name": "train",
|
| 46 |
+
"num_bytes": 11906898263,
|
| 47 |
+
"num_examples": 12005390,
|
| 48 |
+
"dataset_name": "bio_lang"
|
| 49 |
+
},
|
| 50 |
+
"test": {
|
| 51 |
+
"name": "test",
|
| 52 |
+
"num_bytes": 36513458,
|
| 53 |
+
"num_examples": 37112,
|
| 54 |
+
"dataset_name": "bio_lang"
|
| 55 |
+
},
|
| 56 |
+
"validation": {
|
| 57 |
+
"name": "validation",
|
| 58 |
+
"num_bytes": 36248011,
|
| 59 |
+
"num_examples": 36713,
|
| 60 |
+
"dataset_name": "bio_lang"
|
| 61 |
+
}
|
| 62 |
+
},
|
| 63 |
+
"download_checksums": {
|
| 64 |
+
"https://huggingface.co/datasets/EMBO/biolang/resolve/main/oapmc_abstracts_figs.zip": {
|
| 65 |
+
"num_bytes": 5299878661,
|
| 66 |
+
"checksum": "babe2e036065d73ca3361c96ce4c73aafd820aefaea1afd34bc5d1ac8e7760a5"
|
| 67 |
+
}
|
| 68 |
+
},
|
| 69 |
+
"download_size": 5299878661,
|
| 70 |
+
"post_processing_size": null,
|
| 71 |
+
"dataset_size": 11979659732,
|
| 72 |
+
"size_in_bytes": 17279538393
|
| 73 |
+
},
|
| 74 |
+
"DET": {
|
| 75 |
+
"description": "This dataset is based on abstracts from the open access section of EuropePubMed Central to train language models in the domain of biology. \n",
|
| 76 |
+
"citation": "@Unpublished{\n huggingface: dataset,\n title = {biolang},\n authors={Thomas Lemberger, EMBO},\n year={2021}\n}\n",
|
| 77 |
+
"homepage": "https://europepmc.org/downloads/openaccess",
|
| 78 |
+
"license": "CC BY 4.0",
|
| 79 |
+
"features": {
|
| 80 |
+
"input_ids": {
|
| 81 |
+
"feature": {
|
| 82 |
+
"dtype": "int32",
|
| 83 |
+
"id": null,
|
| 84 |
+
"_type": "Value"
|
| 85 |
+
},
|
| 86 |
+
"length": -1,
|
| 87 |
+
"id": null,
|
| 88 |
+
"_type": "Sequence"
|
| 89 |
+
},
|
| 90 |
+
"tag_mask": {
|
| 91 |
+
"feature": {
|
| 92 |
+
"dtype": "int8",
|
| 93 |
+
"id": null,
|
| 94 |
+
"_type": "Value"
|
| 95 |
+
},
|
| 96 |
+
"length": -1,
|
| 97 |
+
"id": null,
|
| 98 |
+
"_type": "Sequence"
|
| 99 |
+
}
|
| 100 |
+
},
|
| 101 |
+
"post_processed": null,
|
| 102 |
+
"supervised_keys": {
|
| 103 |
+
"input": "input_ids",
|
| 104 |
+
"output": "pos_mask"
|
| 105 |
+
},
|
| 106 |
+
"builder_name": "bio_lang",
|
| 107 |
+
"config_name": "DET",
|
| 108 |
+
"version": {
|
| 109 |
+
"version_str": "0.0.1",
|
| 110 |
+
"description": null,
|
| 111 |
+
"major": 0,
|
| 112 |
+
"minor": 0,
|
| 113 |
+
"patch": 1
|
| 114 |
+
},
|
| 115 |
+
"splits": {
|
| 116 |
+
"train": {
|
| 117 |
+
"name": "train",
|
| 118 |
+
"num_bytes": 11906898263,
|
| 119 |
+
"num_examples": 12005390,
|
| 120 |
+
"dataset_name": "bio_lang"
|
| 121 |
+
},
|
| 122 |
+
"test": {
|
| 123 |
+
"name": "test",
|
| 124 |
+
"num_bytes": 36513458,
|
| 125 |
+
"num_examples": 37112,
|
| 126 |
+
"dataset_name": "bio_lang"
|
| 127 |
+
},
|
| 128 |
+
"validation": {
|
| 129 |
+
"name": "validation",
|
| 130 |
+
"num_bytes": 36248011,
|
| 131 |
+
"num_examples": 36713,
|
| 132 |
+
"dataset_name": "bio_lang"
|
| 133 |
+
}
|
| 134 |
+
},
|
| 135 |
+
"download_checksums": {
|
| 136 |
+
"https://huggingface.co/datasets/EMBO/biolang/resolve/main/oapmc_abstracts_figs.zip": {
|
| 137 |
+
"num_bytes": 5299878661,
|
| 138 |
+
"checksum": "babe2e036065d73ca3361c96ce4c73aafd820aefaea1afd34bc5d1ac8e7760a5"
|
| 139 |
+
}
|
| 140 |
+
},
|
| 141 |
+
"download_size": 5299878661,
|
| 142 |
+
"post_processing_size": null,
|
| 143 |
+
"dataset_size": 11979659732,
|
| 144 |
+
"size_in_bytes": 17279538393
|
| 145 |
+
},
|
| 146 |
+
"VERB": {
|
| 147 |
+
"description": "This dataset is based on abstracts from the open access section of EuropePubMed Central to train language models in the domain of biology. \n",
|
| 148 |
+
"citation": "@Unpublished{\n huggingface: dataset,\n title = {biolang},\n authors={Thomas Lemberger, EMBO},\n year={2021}\n}\n",
|
| 149 |
+
"homepage": "https://europepmc.org/downloads/openaccess",
|
| 150 |
+
"license": "CC BY 4.0",
|
| 151 |
+
"features": {
|
| 152 |
+
"input_ids": {
|
| 153 |
+
"feature": {
|
| 154 |
+
"dtype": "int32",
|
| 155 |
+
"id": null,
|
| 156 |
+
"_type": "Value"
|
| 157 |
+
},
|
| 158 |
+
"length": -1,
|
| 159 |
+
"id": null,
|
| 160 |
+
"_type": "Sequence"
|
| 161 |
+
},
|
| 162 |
+
"tag_mask": {
|
| 163 |
+
"feature": {
|
| 164 |
+
"dtype": "int8",
|
| 165 |
+
"id": null,
|
| 166 |
+
"_type": "Value"
|
| 167 |
+
},
|
| 168 |
+
"length": -1,
|
| 169 |
+
"id": null,
|
| 170 |
+
"_type": "Sequence"
|
| 171 |
+
}
|
| 172 |
+
},
|
| 173 |
+
"post_processed": null,
|
| 174 |
+
"supervised_keys": {
|
| 175 |
+
"input": "input_ids",
|
| 176 |
+
"output": "pos_mask"
|
| 177 |
+
},
|
| 178 |
+
"builder_name": "bio_lang",
|
| 179 |
+
"config_name": "VERB",
|
| 180 |
+
"version": {
|
| 181 |
+
"version_str": "0.0.1",
|
| 182 |
+
"description": null,
|
| 183 |
+
"major": 0,
|
| 184 |
+
"minor": 0,
|
| 185 |
+
"patch": 1
|
| 186 |
+
},
|
| 187 |
+
"splits": {
|
| 188 |
+
"train": {
|
| 189 |
+
"name": "train",
|
| 190 |
+
"num_bytes": 11906898263,
|
| 191 |
+
"num_examples": 12005390,
|
| 192 |
+
"dataset_name": "bio_lang"
|
| 193 |
+
},
|
| 194 |
+
"test": {
|
| 195 |
+
"name": "test",
|
| 196 |
+
"num_bytes": 36513458,
|
| 197 |
+
"num_examples": 37112,
|
| 198 |
+
"dataset_name": "bio_lang"
|
| 199 |
+
},
|
| 200 |
+
"validation": {
|
| 201 |
+
"name": "validation",
|
| 202 |
+
"num_bytes": 36248011,
|
| 203 |
+
"num_examples": 36713,
|
| 204 |
+
"dataset_name": "bio_lang"
|
| 205 |
+
}
|
| 206 |
+
},
|
| 207 |
+
"download_checksums": {
|
| 208 |
+
"https://huggingface.co/datasets/EMBO/biolang/resolve/main/oapmc_abstracts_figs.zip": {
|
| 209 |
+
"num_bytes": 5299878661,
|
| 210 |
+
"checksum": "babe2e036065d73ca3361c96ce4c73aafd820aefaea1afd34bc5d1ac8e7760a5"
|
| 211 |
+
}
|
| 212 |
+
},
|
| 213 |
+
"download_size": 5299878661,
|
| 214 |
+
"post_processing_size": null,
|
| 215 |
+
"dataset_size": 11979659732,
|
| 216 |
+
"size_in_bytes": 17279538393
|
| 217 |
+
},
|
| 218 |
+
"SMALL": {
|
| 219 |
+
"description": "This dataset is based on abstracts from the open access section of EuropePubMed Central to train language models in the domain of biology. \n",
|
| 220 |
+
"citation": "@Unpublished{\n huggingface: dataset,\n title = {biolang},\n authors={Thomas Lemberger, EMBO},\n year={2021}\n}\n",
|
| 221 |
+
"homepage": "https://europepmc.org/downloads/openaccess",
|
| 222 |
+
"license": "CC BY 4.0",
|
| 223 |
+
"features": {
|
| 224 |
+
"input_ids": {
|
| 225 |
+
"feature": {
|
| 226 |
+
"dtype": "int32",
|
| 227 |
+
"id": null,
|
| 228 |
+
"_type": "Value"
|
| 229 |
+
},
|
| 230 |
+
"length": -1,
|
| 231 |
+
"id": null,
|
| 232 |
+
"_type": "Sequence"
|
| 233 |
+
},
|
| 234 |
+
"tag_mask": {
|
| 235 |
+
"feature": {
|
| 236 |
+
"dtype": "int8",
|
| 237 |
+
"id": null,
|
| 238 |
+
"_type": "Value"
|
| 239 |
+
},
|
| 240 |
+
"length": -1,
|
| 241 |
+
"id": null,
|
| 242 |
+
"_type": "Sequence"
|
| 243 |
+
}
|
| 244 |
+
},
|
| 245 |
+
"post_processed": null,
|
| 246 |
+
"supervised_keys": {
|
| 247 |
+
"input": "input_ids",
|
| 248 |
+
"output": "pos_mask"
|
| 249 |
+
},
|
| 250 |
+
"builder_name": "bio_lang",
|
| 251 |
+
"config_name": "SMALL",
|
| 252 |
+
"version": "0.0.1",
|
| 253 |
+
"splits": {
|
| 254 |
+
"train": {
|
| 255 |
+
"name": "train",
|
| 256 |
+
"num_bytes": 11906898263,
|
| 257 |
+
"num_examples": 12005390,
|
| 258 |
+
"dataset_name": "bio_lang"
|
| 259 |
+
},
|
| 260 |
+
"test": {
|
| 261 |
+
"name": "test",
|
| 262 |
+
"num_bytes": 36513458,
|
| 263 |
+
"num_examples": 37112,
|
| 264 |
+
"dataset_name": "bio_lang"
|
| 265 |
+
},
|
| 266 |
+
"validation": {
|
| 267 |
+
"name": "validation",
|
| 268 |
+
"num_bytes": 36248011,
|
| 269 |
+
"num_examples": 36713,
|
| 270 |
+
"dataset_name": "bio_lang"
|
| 271 |
+
}
|
| 272 |
+
},
|
| 273 |
+
"download_checksums": {
|
| 274 |
+
"https://huggingface.co/datasets/EMBO/biolang/resolve/main/oapmc_abstracts_figs.zip": {
|
| 275 |
+
"num_bytes": 5299878661,
|
| 276 |
+
"checksum": "babe2e036065d73ca3361c96ce4c73aafd820aefaea1afd34bc5d1ac8e7760a5"
|
| 277 |
+
}
|
| 278 |
+
},
|
| 279 |
+
"download_size": 5299878661,
|
| 280 |
+
"post_processing_size": null,
|
| 281 |
+
"dataset_size": 11979659732,
|
| 282 |
+
"size_in_bytes": 17279538393
|
| 283 |
+
}
|
| 284 |
+
}
|