Referensi:
XNLI
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/XNLI')
- Keterangan :
The Cross-lingual Natural Language Inference (XNLI) corpus is a crowd-sourced collection of 5,000 test and
2,500 dev pairs for the MultiNLI corpus. The pairs are annotated with textual entailment and translated into
14 languages: French, Spanish, German, Greek, Bulgarian, Russian, Turkish, Arabic, Vietnamese, Thai, Chinese,
Hindi, Swahili and Urdu. This results in 112.5k annotated pairs. Each premise can be associated with the
corresponding hypothesis in the 15 languages, summing up to more than 1.5M combinations. The corpus is made to
evaluate how to perform inference in any language (including low-resources ones like Swahili or Urdu) when only
English NLI data is available at training time. One solution is cross-lingual sentence encoding, for which XNLI
is an evaluation benchmark.
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'test' | 75150 |
'validation' | 37350 |
- Fitur :
{
"language": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"sentence1": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"sentence2": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"gold_label": {
"dtype": "string",
"id": null,
"_type": "Value"
}
}
tydiqa
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/tydiqa')
- Keterangan :
Gold passage task (GoldP): Given a passage that is guaranteed to contain the
answer, predict the single contiguous span of characters that answers the question. This is more similar to
existing reading comprehension datasets (as opposed to the information-seeking task outlined above).
This task is constructed with two goals in mind: (1) more directly comparing with prior work and (2) providing
a simplified way for researchers to use TyDi QA by providing compatibility with existing code for SQuAD 1.1,
XQuAD, and MLQA. Toward these goals, the gold passage task differs from the primary task in several ways:
only the gold answer passage is provided rather than the entire Wikipedia article;
unanswerable questions have been discarded, similar to MLQA and XQuAD;
we evaluate with the SQuAD 1.1 metrics like XQuAD; and
Thai and Japanese are removed since the lack of whitespace breaks some tools.
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'train' | 49881 |
'validation' | 5077 |
- Fitur :
{
"id": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"title": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"context": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"question": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"answers": {
"feature": {
"answer_start": {
"dtype": "int32",
"id": null,
"_type": "Value"
},
"text": {
"dtype": "string",
"id": null,
"_type": "Value"
}
},
"length": -1,
"id": null,
"_type": "Sequence"
}
}
Pasukan
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/SQuAD')
- Keterangan :
Stanford Question Answering Dataset (SQuAD) is a reading comprehension dataset, consisting of questions posed by crowdworkers on a set of Wikipedia articles, where the answer to every question is a segment of text, or span, from the corresponding reading passage, or the question might be unanswerable.
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'train' | 87599 |
'validation' | 10570 |
- Fitur :
{
"id": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"title": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"context": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"question": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"answers": {
"feature": {
"answer_start": {
"dtype": "int32",
"id": null,
"_type": "Value"
},
"text": {
"dtype": "string",
"id": null,
"_type": "Value"
}
},
"length": -1,
"id": null,
"_type": "Sequence"
}
}
PAN-X.af
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/PAN-X.af')
- Keterangan :
The WikiANN dataset (Pan et al. 2017) is a dataset with NER annotations for PER, ORG and LOC. It has been
constructed using the linked entities in Wikipedia pages for 282 different languages including Danish. The dataset
can be loaded with the DaNLP package:
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'test' | 1000 |
'train' | 5000 |
'validation' | 1000 |
- Fitur :
{
"tokens": {
"feature": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"length": -1,
"id": null,
"_type": "Sequence"
},
"ner_tags": {
"feature": {
"num_classes": 7,
"names": [
"O",
"B-PER",
"I-PER",
"B-ORG",
"I-ORG",
"B-LOC",
"I-LOC"
],
"names_file": null,
"id": null,
"_type": "ClassLabel"
},
"length": -1,
"id": null,
"_type": "Sequence"
},
"langs": {
"feature": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"length": -1,
"id": null,
"_type": "Sequence"
}
}
PAN-X.ar
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/PAN-X.ar')
- Keterangan :
The WikiANN dataset (Pan et al. 2017) is a dataset with NER annotations for PER, ORG and LOC. It has been
constructed using the linked entities in Wikipedia pages for 282 different languages including Danish. The dataset
can be loaded with the DaNLP package:
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'test' | 10.000 |
'train' | 20.000 |
'validation' | 10.000 |
- Fitur :
{
"tokens": {
"feature": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"length": -1,
"id": null,
"_type": "Sequence"
},
"ner_tags": {
"feature": {
"num_classes": 7,
"names": [
"O",
"B-PER",
"I-PER",
"B-ORG",
"I-ORG",
"B-LOC",
"I-LOC"
],
"names_file": null,
"id": null,
"_type": "ClassLabel"
},
"length": -1,
"id": null,
"_type": "Sequence"
},
"langs": {
"feature": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"length": -1,
"id": null,
"_type": "Sequence"
}
}
PAN-X.bg
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/PAN-X.bg')
- Keterangan :
The WikiANN dataset (Pan et al. 2017) is a dataset with NER annotations for PER, ORG and LOC. It has been
constructed using the linked entities in Wikipedia pages for 282 different languages including Danish. The dataset
can be loaded with the DaNLP package:
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'test' | 10.000 |
'train' | 20.000 |
'validation' | 10.000 |
- Fitur :
{
"tokens": {
"feature": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"length": -1,
"id": null,
"_type": "Sequence"
},
"ner_tags": {
"feature": {
"num_classes": 7,
"names": [
"O",
"B-PER",
"I-PER",
"B-ORG",
"I-ORG",
"B-LOC",
"I-LOC"
],
"names_file": null,
"id": null,
"_type": "ClassLabel"
},
"length": -1,
"id": null,
"_type": "Sequence"
},
"langs": {
"feature": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"length": -1,
"id": null,
"_type": "Sequence"
}
}
PAN-X.bn
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/PAN-X.bn')
- Keterangan :
The WikiANN dataset (Pan et al. 2017) is a dataset with NER annotations for PER, ORG and LOC. It has been
constructed using the linked entities in Wikipedia pages for 282 different languages including Danish. The dataset
can be loaded with the DaNLP package:
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'test' | 1000 |
'train' | 10.000 |
'validation' | 1000 |
- Fitur :
{
"tokens": {
"feature": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"length": -1,
"id": null,
"_type": "Sequence"
},
"ner_tags": {
"feature": {
"num_classes": 7,
"names": [
"O",
"B-PER",
"I-PER",
"B-ORG",
"I-ORG",
"B-LOC",
"I-LOC"
],
"names_file": null,
"id": null,
"_type": "ClassLabel"
},
"length": -1,
"id": null,
"_type": "Sequence"
},
"langs": {
"feature": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"length": -1,
"id": null,
"_type": "Sequence"
}
}
PAN-X.de
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/PAN-X.de')
- Keterangan :
The WikiANN dataset (Pan et al. 2017) is a dataset with NER annotations for PER, ORG and LOC. It has been
constructed using the linked entities in Wikipedia pages for 282 different languages including Danish. The dataset
can be loaded with the DaNLP package:
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'test' | 10.000 |
'train' | 20.000 |
'validation' | 10.000 |
- Fitur :
{
"tokens": {
"feature": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"length": -1,
"id": null,
"_type": "Sequence"
},
"ner_tags": {
"feature": {
"num_classes": 7,
"names": [
"O",
"B-PER",
"I-PER",
"B-ORG",
"I-ORG",
"B-LOC",
"I-LOC"
],
"names_file": null,
"id": null,
"_type": "ClassLabel"
},
"length": -1,
"id": null,
"_type": "Sequence"
},
"langs": {
"feature": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"length": -1,
"id": null,
"_type": "Sequence"
}
}
PAN-X.el
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/PAN-X.el')
- Keterangan :
The WikiANN dataset (Pan et al. 2017) is a dataset with NER annotations for PER, ORG and LOC. It has been
constructed using the linked entities in Wikipedia pages for 282 different languages including Danish. The dataset
can be loaded with the DaNLP package:
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'test' | 10.000 |
'train' | 20.000 |
'validation' | 10.000 |
- Fitur :
{
"tokens": {
"feature": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"length": -1,
"id": null,
"_type": "Sequence"
},
"ner_tags": {
"feature": {
"num_classes": 7,
"names": [
"O",
"B-PER",
"I-PER",
"B-ORG",
"I-ORG",
"B-LOC",
"I-LOC"
],
"names_file": null,
"id": null,
"_type": "ClassLabel"
},
"length": -1,
"id": null,
"_type": "Sequence"
},
"langs": {
"feature": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"length": -1,
"id": null,
"_type": "Sequence"
}
}
PAN-X.en
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/PAN-X.en')
- Keterangan :
The WikiANN dataset (Pan et al. 2017) is a dataset with NER annotations for PER, ORG and LOC. It has been
constructed using the linked entities in Wikipedia pages for 282 different languages including Danish. The dataset
can be loaded with the DaNLP package:
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'test' | 10.000 |
'train' | 20.000 |
'validation' | 10.000 |
- Fitur :
{
"tokens": {
"feature": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"length": -1,
"id": null,
"_type": "Sequence"
},
"ner_tags": {
"feature": {
"num_classes": 7,
"names": [
"O",
"B-PER",
"I-PER",
"B-ORG",
"I-ORG",
"B-LOC",
"I-LOC"
],
"names_file": null,
"id": null,
"_type": "ClassLabel"
},
"length": -1,
"id": null,
"_type": "Sequence"
},
"langs": {
"feature": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"length": -1,
"id": null,
"_type": "Sequence"
}
}
PAN-X.es
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/PAN-X.es')
- Keterangan :
The WikiANN dataset (Pan et al. 2017) is a dataset with NER annotations for PER, ORG and LOC. It has been
constructed using the linked entities in Wikipedia pages for 282 different languages including Danish. The dataset
can be loaded with the DaNLP package:
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'test' | 10.000 |
'train' | 20.000 |
'validation' | 10.000 |
- Fitur :
{
"tokens": {
"feature": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"length": -1,
"id": null,
"_type": "Sequence"
},
"ner_tags": {
"feature": {
"num_classes": 7,
"names": [
"O",
"B-PER",
"I-PER",
"B-ORG",
"I-ORG",
"B-LOC",
"I-LOC"
],
"names_file": null,
"id": null,
"_type": "ClassLabel"
},
"length": -1,
"id": null,
"_type": "Sequence"
},
"langs": {
"feature": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"length": -1,
"id": null,
"_type": "Sequence"
}
}
PAN-X.et
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/PAN-X.et')
- Keterangan :
The WikiANN dataset (Pan et al. 2017) is a dataset with NER annotations for PER, ORG and LOC. It has been
constructed using the linked entities in Wikipedia pages for 282 different languages including Danish. The dataset
can be loaded with the DaNLP package:
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'test' | 10.000 |
'train' | 15000 |
'validation' | 10.000 |
- Fitur :
{
"tokens": {
"feature": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"length": -1,
"id": null,
"_type": "Sequence"
},
"ner_tags": {
"feature": {
"num_classes": 7,
"names": [
"O",
"B-PER",
"I-PER",
"B-ORG",
"I-ORG",
"B-LOC",
"I-LOC"
],
"names_file": null,
"id": null,
"_type": "ClassLabel"
},
"length": -1,
"id": null,
"_type": "Sequence"
},
"langs": {
"feature": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"length": -1,
"id": null,
"_type": "Sequence"
}
}
PAN-X.eu
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/PAN-X.eu')
- Keterangan :
The WikiANN dataset (Pan et al. 2017) is a dataset with NER annotations for PER, ORG and LOC. It has been
constructed using the linked entities in Wikipedia pages for 282 different languages including Danish. The dataset
can be loaded with the DaNLP package:
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'test' | 10.000 |
'train' | 10.000 |
'validation' | 10.000 |
- Fitur :
{
"tokens": {
"feature": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"length": -1,
"id": null,
"_type": "Sequence"
},
"ner_tags": {
"feature": {
"num_classes": 7,
"names": [
"O",
"B-PER",
"I-PER",
"B-ORG",
"I-ORG",
"B-LOC",
"I-LOC"
],
"names_file": null,
"id": null,
"_type": "ClassLabel"
},
"length": -1,
"id": null,
"_type": "Sequence"
},
"langs": {
"feature": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"length": -1,
"id": null,
"_type": "Sequence"
}
}
PAN-X.fa
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/PAN-X.fa')
- Keterangan :
The WikiANN dataset (Pan et al. 2017) is a dataset with NER annotations for PER, ORG and LOC. It has been
constructed using the linked entities in Wikipedia pages for 282 different languages including Danish. The dataset
can be loaded with the DaNLP package:
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'test' | 10.000 |
'train' | 20.000 |
'validation' | 10.000 |
- Fitur :
{
"tokens": {
"feature": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"length": -1,
"id": null,
"_type": "Sequence"
},
"ner_tags": {
"feature": {
"num_classes": 7,
"names": [
"O",
"B-PER",
"I-PER",
"B-ORG",
"I-ORG",
"B-LOC",
"I-LOC"
],
"names_file": null,
"id": null,
"_type": "ClassLabel"
},
"length": -1,
"id": null,
"_type": "Sequence"
},
"langs": {
"feature": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"length": -1,
"id": null,
"_type": "Sequence"
}
}
PAN-X.fi
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/PAN-X.fi')
- Keterangan :
The WikiANN dataset (Pan et al. 2017) is a dataset with NER annotations for PER, ORG and LOC. It has been
constructed using the linked entities in Wikipedia pages for 282 different languages including Danish. The dataset
can be loaded with the DaNLP package:
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'test' | 10.000 |
'train' | 20.000 |
'validation' | 10.000 |
- Fitur :
{
"tokens": {
"feature": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"length": -1,
"id": null,
"_type": "Sequence"
},
"ner_tags": {
"feature": {
"num_classes": 7,
"names": [
"O",
"B-PER",
"I-PER",
"B-ORG",
"I-ORG",
"B-LOC",
"I-LOC"
],
"names_file": null,
"id": null,
"_type": "ClassLabel"
},
"length": -1,
"id": null,
"_type": "Sequence"
},
"langs": {
"feature": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"length": -1,
"id": null,
"_type": "Sequence"
}
}
PAN-X.fr
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/PAN-X.fr')
- Keterangan :
The WikiANN dataset (Pan et al. 2017) is a dataset with NER annotations for PER, ORG and LOC. It has been
constructed using the linked entities in Wikipedia pages for 282 different languages including Danish. The dataset
can be loaded with the DaNLP package:
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'test' | 10.000 |
'train' | 20.000 |
'validation' | 10.000 |
- Fitur :
{
"tokens": {
"feature": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"length": -1,
"id": null,
"_type": "Sequence"
},
"ner_tags": {
"feature": {
"num_classes": 7,
"names": [
"O",
"B-PER",
"I-PER",
"B-ORG",
"I-ORG",
"B-LOC",
"I-LOC"
],
"names_file": null,
"id": null,
"_type": "ClassLabel"
},
"length": -1,
"id": null,
"_type": "Sequence"
},
"langs": {
"feature": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"length": -1,
"id": null,
"_type": "Sequence"
}
}
PAN-X.he
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/PAN-X.he')
- Keterangan :
The WikiANN dataset (Pan et al. 2017) is a dataset with NER annotations for PER, ORG and LOC. It has been
constructed using the linked entities in Wikipedia pages for 282 different languages including Danish. The dataset
can be loaded with the DaNLP package:
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'test' | 10.000 |
'train' | 20.000 |
'validation' | 10.000 |
- Fitur :
{
"tokens": {
"feature": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"length": -1,
"id": null,
"_type": "Sequence"
},
"ner_tags": {
"feature": {
"num_classes": 7,
"names": [
"O",
"B-PER",
"I-PER",
"B-ORG",
"I-ORG",
"B-LOC",
"I-LOC"
],
"names_file": null,
"id": null,
"_type": "ClassLabel"
},
"length": -1,
"id": null,
"_type": "Sequence"
},
"langs": {
"feature": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"length": -1,
"id": null,
"_type": "Sequence"
}
}
PAN-X.hai
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/PAN-X.hi')
- Keterangan :
The WikiANN dataset (Pan et al. 2017) is a dataset with NER annotations for PER, ORG and LOC. It has been
constructed using the linked entities in Wikipedia pages for 282 different languages including Danish. The dataset
can be loaded with the DaNLP package:
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'test' | 1000 |
'train' | 5000 |
'validation' | 1000 |
- Fitur :
{
"tokens": {
"feature": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"length": -1,
"id": null,
"_type": "Sequence"
},
"ner_tags": {
"feature": {
"num_classes": 7,
"names": [
"O",
"B-PER",
"I-PER",
"B-ORG",
"I-ORG",
"B-LOC",
"I-LOC"
],
"names_file": null,
"id": null,
"_type": "ClassLabel"
},
"length": -1,
"id": null,
"_type": "Sequence"
},
"langs": {
"feature": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"length": -1,
"id": null,
"_type": "Sequence"
}
}
PAN-X.hu
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/PAN-X.hu')
- Keterangan :
The WikiANN dataset (Pan et al. 2017) is a dataset with NER annotations for PER, ORG and LOC. It has been
constructed using the linked entities in Wikipedia pages for 282 different languages including Danish. The dataset
can be loaded with the DaNLP package:
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'test' | 10.000 |
'train' | 20.000 |
'validation' | 10.000 |
- Fitur :
{
"tokens": {
"feature": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"length": -1,
"id": null,
"_type": "Sequence"
},
"ner_tags": {
"feature": {
"num_classes": 7,
"names": [
"O",
"B-PER",
"I-PER",
"B-ORG",
"I-ORG",
"B-LOC",
"I-LOC"
],
"names_file": null,
"id": null,
"_type": "ClassLabel"
},
"length": -1,
"id": null,
"_type": "Sequence"
},
"langs": {
"feature": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"length": -1,
"id": null,
"_type": "Sequence"
}
}
PAN-X.id
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/PAN-X.id')
- Keterangan :
The WikiANN dataset (Pan et al. 2017) is a dataset with NER annotations for PER, ORG and LOC. It has been
constructed using the linked entities in Wikipedia pages for 282 different languages including Danish. The dataset
can be loaded with the DaNLP package:
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'test' | 10.000 |
'train' | 20.000 |
'validation' | 10.000 |
- Fitur :
{
"tokens": {
"feature": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"length": -1,
"id": null,
"_type": "Sequence"
},
"ner_tags": {
"feature": {
"num_classes": 7,
"names": [
"O",
"B-PER",
"I-PER",
"B-ORG",
"I-ORG",
"B-LOC",
"I-LOC"
],
"names_file": null,
"id": null,
"_type": "ClassLabel"
},
"length": -1,
"id": null,
"_type": "Sequence"
},
"langs": {
"feature": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"length": -1,
"id": null,
"_type": "Sequence"
}
}
PAN-X.it
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/PAN-X.it')
- Keterangan :
The WikiANN dataset (Pan et al. 2017) is a dataset with NER annotations for PER, ORG and LOC. It has been
constructed using the linked entities in Wikipedia pages for 282 different languages including Danish. The dataset
can be loaded with the DaNLP package:
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'test' | 10.000 |
'train' | 20.000 |
'validation' | 10.000 |
- Fitur :
{
"tokens": {
"feature": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"length": -1,
"id": null,
"_type": "Sequence"
},
"ner_tags": {
"feature": {
"num_classes": 7,
"names": [
"O",
"B-PER",
"I-PER",
"B-ORG",
"I-ORG",
"B-LOC",
"I-LOC"
],
"names_file": null,
"id": null,
"_type": "ClassLabel"
},
"length": -1,
"id": null,
"_type": "Sequence"
},
"langs": {
"feature": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"length": -1,
"id": null,
"_type": "Sequence"
}
}
PAN-X.ja
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/PAN-X.ja')
- Keterangan :
The WikiANN dataset (Pan et al. 2017) is a dataset with NER annotations for PER, ORG and LOC. It has been
constructed using the linked entities in Wikipedia pages for 282 different languages including Danish. The dataset
can be loaded with the DaNLP package:
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'test' | 10.000 |
'train' | 20.000 |
'validation' | 10.000 |
- Fitur :
{
"tokens": {
"feature": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"length": -1,
"id": null,
"_type": "Sequence"
},
"ner_tags": {
"feature": {
"num_classes": 7,
"names": [
"O",
"B-PER",
"I-PER",
"B-ORG",
"I-ORG",
"B-LOC",
"I-LOC"
],
"names_file": null,
"id": null,
"_type": "ClassLabel"
},
"length": -1,
"id": null,
"_type": "Sequence"
},
"langs": {
"feature": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"length": -1,
"id": null,
"_type": "Sequence"
}
}
PAN-X.jv
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/PAN-X.jv')
- Keterangan :
The WikiANN dataset (Pan et al. 2017) is a dataset with NER annotations for PER, ORG and LOC. It has been
constructed using the linked entities in Wikipedia pages for 282 different languages including Danish. The dataset
can be loaded with the DaNLP package:
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'test' | 100 |
'train' | 100 |
'validation' | 100 |
- Fitur :
{
"tokens": {
"feature": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"length": -1,
"id": null,
"_type": "Sequence"
},
"ner_tags": {
"feature": {
"num_classes": 7,
"names": [
"O",
"B-PER",
"I-PER",
"B-ORG",
"I-ORG",
"B-LOC",
"I-LOC"
],
"names_file": null,
"id": null,
"_type": "ClassLabel"
},
"length": -1,
"id": null,
"_type": "Sequence"
},
"langs": {
"feature": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"length": -1,
"id": null,
"_type": "Sequence"
}
}
PAN-X.ka
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/PAN-X.ka')
- Keterangan :
The WikiANN dataset (Pan et al. 2017) is a dataset with NER annotations for PER, ORG and LOC. It has been
constructed using the linked entities in Wikipedia pages for 282 different languages including Danish. The dataset
can be loaded with the DaNLP package:
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'test' | 10.000 |
'train' | 10.000 |
'validation' | 10.000 |
- Fitur :
{
"tokens": {
"feature": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"length": -1,
"id": null,
"_type": "Sequence"
},
"ner_tags": {
"feature": {
"num_classes": 7,
"names": [
"O",
"B-PER",
"I-PER",
"B-ORG",
"I-ORG",
"B-LOC",
"I-LOC"
],
"names_file": null,
"id": null,
"_type": "ClassLabel"
},
"length": -1,
"id": null,
"_type": "Sequence"
},
"langs": {
"feature": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"length": -1,
"id": null,
"_type": "Sequence"
}
}
PAN-X.kk
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/PAN-X.kk')
- Keterangan :
The WikiANN dataset (Pan et al. 2017) is a dataset with NER annotations for PER, ORG and LOC. It has been
constructed using the linked entities in Wikipedia pages for 282 different languages including Danish. The dataset
can be loaded with the DaNLP package:
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'test' | 1000 |
'train' | 1000 |
'validation' | 1000 |
- Fitur :
{
"tokens": {
"feature": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"length": -1,
"id": null,
"_type": "Sequence"
},
"ner_tags": {
"feature": {
"num_classes": 7,
"names": [
"O",
"B-PER",
"I-PER",
"B-ORG",
"I-ORG",
"B-LOC",
"I-LOC"
],
"names_file": null,
"id": null,
"_type": "ClassLabel"
},
"length": -1,
"id": null,
"_type": "Sequence"
},
"langs": {
"feature": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"length": -1,
"id": null,
"_type": "Sequence"
}
}
PAN-X.ko
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/PAN-X.ko')
- Keterangan :
The WikiANN dataset (Pan et al. 2017) is a dataset with NER annotations for PER, ORG and LOC. It has been
constructed using the linked entities in Wikipedia pages for 282 different languages including Danish. The dataset
can be loaded with the DaNLP package:
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'test' | 10.000 |
'train' | 20.000 |
'validation' | 10.000 |
- Fitur :
{
"tokens": {
"feature": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"length": -1,
"id": null,
"_type": "Sequence"
},
"ner_tags": {
"feature": {
"num_classes": 7,
"names": [
"O",
"B-PER",
"I-PER",
"B-ORG",
"I-ORG",
"B-LOC",
"I-LOC"
],
"names_file": null,
"id": null,
"_type": "ClassLabel"
},
"length": -1,
"id": null,
"_type": "Sequence"
},
"langs": {
"feature": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"length": -1,
"id": null,
"_type": "Sequence"
}
}
PAN-X.ml
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/PAN-X.ml')
- Keterangan :
The WikiANN dataset (Pan et al. 2017) is a dataset with NER annotations for PER, ORG and LOC. It has been
constructed using the linked entities in Wikipedia pages for 282 different languages including Danish. The dataset
can be loaded with the DaNLP package:
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'test' | 1000 |
'train' | 10.000 |
'validation' | 1000 |
- Fitur :
{
"tokens": {
"feature": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"length": -1,
"id": null,
"_type": "Sequence"
},
"ner_tags": {
"feature": {
"num_classes": 7,
"names": [
"O",
"B-PER",
"I-PER",
"B-ORG",
"I-ORG",
"B-LOC",
"I-LOC"
],
"names_file": null,
"id": null,
"_type": "ClassLabel"
},
"length": -1,
"id": null,
"_type": "Sequence"
},
"langs": {
"feature": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"length": -1,
"id": null,
"_type": "Sequence"
}
}
PAN-X.mr
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/PAN-X.mr')
- Keterangan :
The WikiANN dataset (Pan et al. 2017) is a dataset with NER annotations for PER, ORG and LOC. It has been
constructed using the linked entities in Wikipedia pages for 282 different languages including Danish. The dataset
can be loaded with the DaNLP package:
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'test' | 1000 |
'train' | 5000 |
'validation' | 1000 |
- Fitur :
{
"tokens": {
"feature": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"length": -1,
"id": null,
"_type": "Sequence"
},
"ner_tags": {
"feature": {
"num_classes": 7,
"names": [
"O",
"B-PER",
"I-PER",
"B-ORG",
"I-ORG",
"B-LOC",
"I-LOC"
],
"names_file": null,
"id": null,
"_type": "ClassLabel"
},
"length": -1,
"id": null,
"_type": "Sequence"
},
"langs": {
"feature": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"length": -1,
"id": null,
"_type": "Sequence"
}
}
PAN-X.ms
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/PAN-X.ms')
- Keterangan :
The WikiANN dataset (Pan et al. 2017) is a dataset with NER annotations for PER, ORG and LOC. It has been
constructed using the linked entities in Wikipedia pages for 282 different languages including Danish. The dataset
can be loaded with the DaNLP package:
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'test' | 1000 |
'train' | 20.000 |
'validation' | 1000 |
- Fitur :
{
"tokens": {
"feature": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"length": -1,
"id": null,
"_type": "Sequence"
},
"ner_tags": {
"feature": {
"num_classes": 7,
"names": [
"O",
"B-PER",
"I-PER",
"B-ORG",
"I-ORG",
"B-LOC",
"I-LOC"
],
"names_file": null,
"id": null,
"_type": "ClassLabel"
},
"length": -1,
"id": null,
"_type": "Sequence"
},
"langs": {
"feature": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"length": -1,
"id": null,
"_type": "Sequence"
}
}
PAN-X.my
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/PAN-X.my')
- Keterangan :
The WikiANN dataset (Pan et al. 2017) is a dataset with NER annotations for PER, ORG and LOC. It has been
constructed using the linked entities in Wikipedia pages for 282 different languages including Danish. The dataset
can be loaded with the DaNLP package:
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'test' | 100 |
'train' | 100 |
'validation' | 100 |
- Fitur :
{
"tokens": {
"feature": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"length": -1,
"id": null,
"_type": "Sequence"
},
"ner_tags": {
"feature": {
"num_classes": 7,
"names": [
"O",
"B-PER",
"I-PER",
"B-ORG",
"I-ORG",
"B-LOC",
"I-LOC"
],
"names_file": null,
"id": null,
"_type": "ClassLabel"
},
"length": -1,
"id": null,
"_type": "Sequence"
},
"langs": {
"feature": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"length": -1,
"id": null,
"_type": "Sequence"
}
}
PAN-X.nl
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/PAN-X.nl')
- Keterangan :
The WikiANN dataset (Pan et al. 2017) is a dataset with NER annotations for PER, ORG and LOC. It has been
constructed using the linked entities in Wikipedia pages for 282 different languages including Danish. The dataset
can be loaded with the DaNLP package:
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'test' | 10.000 |
'train' | 20.000 |
'validation' | 10.000 |
- Fitur :
{
"tokens": {
"feature": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"length": -1,
"id": null,
"_type": "Sequence"
},
"ner_tags": {
"feature": {
"num_classes": 7,
"names": [
"O",
"B-PER",
"I-PER",
"B-ORG",
"I-ORG",
"B-LOC",
"I-LOC"
],
"names_file": null,
"id": null,
"_type": "ClassLabel"
},
"length": -1,
"id": null,
"_type": "Sequence"
},
"langs": {
"feature": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"length": -1,
"id": null,
"_type": "Sequence"
}
}
PAN-X.pt
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/PAN-X.pt')
- Keterangan :
The WikiANN dataset (Pan et al. 2017) is a dataset with NER annotations for PER, ORG and LOC. It has been
constructed using the linked entities in Wikipedia pages for 282 different languages including Danish. The dataset
can be loaded with the DaNLP package:
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'test' | 10.000 |
'train' | 20.000 |
'validation' | 10.000 |
- Fitur :
{
"tokens": {
"feature": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"length": -1,
"id": null,
"_type": "Sequence"
},
"ner_tags": {
"feature": {
"num_classes": 7,
"names": [
"O",
"B-PER",
"I-PER",
"B-ORG",
"I-ORG",
"B-LOC",
"I-LOC"
],
"names_file": null,
"id": null,
"_type": "ClassLabel"
},
"length": -1,
"id": null,
"_type": "Sequence"
},
"langs": {
"feature": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"length": -1,
"id": null,
"_type": "Sequence"
}
}
PAN-X.ru
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/PAN-X.ru')
- Keterangan :
The WikiANN dataset (Pan et al. 2017) is a dataset with NER annotations for PER, ORG and LOC. It has been
constructed using the linked entities in Wikipedia pages for 282 different languages including Danish. The dataset
can be loaded with the DaNLP package:
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'test' | 10.000 |
'train' | 20.000 |
'validation' | 10.000 |
- Fitur :
{
"tokens": {
"feature": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"length": -1,
"id": null,
"_type": "Sequence"
},
"ner_tags": {
"feature": {
"num_classes": 7,
"names": [
"O",
"B-PER",
"I-PER",
"B-ORG",
"I-ORG",
"B-LOC",
"I-LOC"
],
"names_file": null,
"id": null,
"_type": "ClassLabel"
},
"length": -1,
"id": null,
"_type": "Sequence"
},
"langs": {
"feature": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"length": -1,
"id": null,
"_type": "Sequence"
}
}
PAN-X.sw
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/PAN-X.sw')
- Keterangan :
The WikiANN dataset (Pan et al. 2017) is a dataset with NER annotations for PER, ORG and LOC. It has been
constructed using the linked entities in Wikipedia pages for 282 different languages including Danish. The dataset
can be loaded with the DaNLP package:
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'test' | 1000 |
'train' | 1000 |
'validation' | 1000 |
- Fitur :
{
"tokens": {
"feature": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"length": -1,
"id": null,
"_type": "Sequence"
},
"ner_tags": {
"feature": {
"num_classes": 7,
"names": [
"O",
"B-PER",
"I-PER",
"B-ORG",
"I-ORG",
"B-LOC",
"I-LOC"
],
"names_file": null,
"id": null,
"_type": "ClassLabel"
},
"length": -1,
"id": null,
"_type": "Sequence"
},
"langs": {
"feature": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"length": -1,
"id": null,
"_type": "Sequence"
}
}
PAN-X.ta
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/PAN-X.ta')
- Keterangan :
The WikiANN dataset (Pan et al. 2017) is a dataset with NER annotations for PER, ORG and LOC. It has been
constructed using the linked entities in Wikipedia pages for 282 different languages including Danish. The dataset
can be loaded with the DaNLP package:
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'test' | 1000 |
'train' | 15000 |
'validation' | 1000 |
- Fitur :
{
"tokens": {
"feature": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"length": -1,
"id": null,
"_type": "Sequence"
},
"ner_tags": {
"feature": {
"num_classes": 7,
"names": [
"O",
"B-PER",
"I-PER",
"B-ORG",
"I-ORG",
"B-LOC",
"I-LOC"
],
"names_file": null,
"id": null,
"_type": "ClassLabel"
},
"length": -1,
"id": null,
"_type": "Sequence"
},
"langs": {
"feature": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"length": -1,
"id": null,
"_type": "Sequence"
}
}
PAN-X.te
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/PAN-X.te')
- Keterangan :
The WikiANN dataset (Pan et al. 2017) is a dataset with NER annotations for PER, ORG and LOC. It has been
constructed using the linked entities in Wikipedia pages for 282 different languages including Danish. The dataset
can be loaded with the DaNLP package:
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'test' | 1000 |
'train' | 1000 |
'validation' | 1000 |
- Fitur :
{
"tokens": {
"feature": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"length": -1,
"id": null,
"_type": "Sequence"
},
"ner_tags": {
"feature": {
"num_classes": 7,
"names": [
"O",
"B-PER",
"I-PER",
"B-ORG",
"I-ORG",
"B-LOC",
"I-LOC"
],
"names_file": null,
"id": null,
"_type": "ClassLabel"
},
"length": -1,
"id": null,
"_type": "Sequence"
},
"langs": {
"feature": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"length": -1,
"id": null,
"_type": "Sequence"
}
}
PAN-X.th
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/PAN-X.th')
- Keterangan :
The WikiANN dataset (Pan et al. 2017) is a dataset with NER annotations for PER, ORG and LOC. It has been
constructed using the linked entities in Wikipedia pages for 282 different languages including Danish. The dataset
can be loaded with the DaNLP package:
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'test' | 10.000 |
'train' | 20.000 |
'validation' | 10.000 |
- Fitur :
{
"tokens": {
"feature": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"length": -1,
"id": null,
"_type": "Sequence"
},
"ner_tags": {
"feature": {
"num_classes": 7,
"names": [
"O",
"B-PER",
"I-PER",
"B-ORG",
"I-ORG",
"B-LOC",
"I-LOC"
],
"names_file": null,
"id": null,
"_type": "ClassLabel"
},
"length": -1,
"id": null,
"_type": "Sequence"
},
"langs": {
"feature": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"length": -1,
"id": null,
"_type": "Sequence"
}
}
PAN-X.tl
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/PAN-X.tl')
- Keterangan :
The WikiANN dataset (Pan et al. 2017) is a dataset with NER annotations for PER, ORG and LOC. It has been
constructed using the linked entities in Wikipedia pages for 282 different languages including Danish. The dataset
can be loaded with the DaNLP package:
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'test' | 1000 |
'train' | 10.000 |
'validation' | 1000 |
- Fitur :
{
"tokens": {
"feature": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"length": -1,
"id": null,
"_type": "Sequence"
},
"ner_tags": {
"feature": {
"num_classes": 7,
"names": [
"O",
"B-PER",
"I-PER",
"B-ORG",
"I-ORG",
"B-LOC",
"I-LOC"
],
"names_file": null,
"id": null,
"_type": "ClassLabel"
},
"length": -1,
"id": null,
"_type": "Sequence"
},
"langs": {
"feature": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"length": -1,
"id": null,
"_type": "Sequence"
}
}
PAN-X.tr
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/PAN-X.tr')
- Keterangan :
The WikiANN dataset (Pan et al. 2017) is a dataset with NER annotations for PER, ORG and LOC. It has been
constructed using the linked entities in Wikipedia pages for 282 different languages including Danish. The dataset
can be loaded with the DaNLP package:
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'test' | 10.000 |
'train' | 20.000 |
'validation' | 10.000 |
- Fitur :
{
"tokens": {
"feature": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"length": -1,
"id": null,
"_type": "Sequence"
},
"ner_tags": {
"feature": {
"num_classes": 7,
"names": [
"O",
"B-PER",
"I-PER",
"B-ORG",
"I-ORG",
"B-LOC",
"I-LOC"
],
"names_file": null,
"id": null,
"_type": "ClassLabel"
},
"length": -1,
"id": null,
"_type": "Sequence"
},
"langs": {
"feature": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"length": -1,
"id": null,
"_type": "Sequence"
}
}
PAN-X.ur
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/PAN-X.ur')
- Keterangan :
The WikiANN dataset (Pan et al. 2017) is a dataset with NER annotations for PER, ORG and LOC. It has been
constructed using the linked entities in Wikipedia pages for 282 different languages including Danish. The dataset
can be loaded with the DaNLP package:
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'test' | 1000 |
'train' | 20.000 |
'validation' | 1000 |
- Fitur :
{
"tokens": {
"feature": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"length": -1,
"id": null,
"_type": "Sequence"
},
"ner_tags": {
"feature": {
"num_classes": 7,
"names": [
"O",
"B-PER",
"I-PER",
"B-ORG",
"I-ORG",
"B-LOC",
"I-LOC"
],
"names_file": null,
"id": null,
"_type": "ClassLabel"
},
"length": -1,
"id": null,
"_type": "Sequence"
},
"langs": {
"feature": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"length": -1,
"id": null,
"_type": "Sequence"
}
}
PAN-X.vi
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/PAN-X.vi')
- Keterangan :
The WikiANN dataset (Pan et al. 2017) is a dataset with NER annotations for PER, ORG and LOC. It has been
constructed using the linked entities in Wikipedia pages for 282 different languages including Danish. The dataset
can be loaded with the DaNLP package:
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'test' | 10.000 |
'train' | 20.000 |
'validation' | 10.000 |
- Fitur :
{
"tokens": {
"feature": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"length": -1,
"id": null,
"_type": "Sequence"
},
"ner_tags": {
"feature": {
"num_classes": 7,
"names": [
"O",
"B-PER",
"I-PER",
"B-ORG",
"I-ORG",
"B-LOC",
"I-LOC"
],
"names_file": null,
"id": null,
"_type": "ClassLabel"
},
"length": -1,
"id": null,
"_type": "Sequence"
},
"langs": {
"feature": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"length": -1,
"id": null,
"_type": "Sequence"
}
}
PAN-X.yo
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/PAN-X.yo')
- Keterangan :
The WikiANN dataset (Pan et al. 2017) is a dataset with NER annotations for PER, ORG and LOC. It has been
constructed using the linked entities in Wikipedia pages for 282 different languages including Danish. The dataset
can be loaded with the DaNLP package:
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'test' | 100 |
'train' | 100 |
'validation' | 100 |
- Fitur :
{
"tokens": {
"feature": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"length": -1,
"id": null,
"_type": "Sequence"
},
"ner_tags": {
"feature": {
"num_classes": 7,
"names": [
"O",
"B-PER",
"I-PER",
"B-ORG",
"I-ORG",
"B-LOC",
"I-LOC"
],
"names_file": null,
"id": null,
"_type": "ClassLabel"
},
"length": -1,
"id": null,
"_type": "Sequence"
},
"langs": {
"feature": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"length": -1,
"id": null,
"_type": "Sequence"
}
}
PAN-X.zh
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/PAN-X.zh')
- Keterangan :
The WikiANN dataset (Pan et al. 2017) is a dataset with NER annotations for PER, ORG and LOC. It has been
constructed using the linked entities in Wikipedia pages for 282 different languages including Danish. The dataset
can be loaded with the DaNLP package:
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'test' | 10.000 |
'train' | 20.000 |
'validation' | 10.000 |
- Fitur :
{
"tokens": {
"feature": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"length": -1,
"id": null,
"_type": "Sequence"
},
"ner_tags": {
"feature": {
"num_classes": 7,
"names": [
"O",
"B-PER",
"I-PER",
"B-ORG",
"I-ORG",
"B-LOC",
"I-LOC"
],
"names_file": null,
"id": null,
"_type": "ClassLabel"
},
"length": -1,
"id": null,
"_type": "Sequence"
},
"langs": {
"feature": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"length": -1,
"id": null,
"_type": "Sequence"
}
}
MLQA.ar.ar
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/MLQA.ar.ar')
- Keterangan :
MLQA (MultiLingual Question Answering) is a benchmark dataset for evaluating cross-lingual question answering performance.
MLQA consists of over 5K extractive QA instances (12K in English) in SQuAD format in seven languages - English, Arabic,
German, Spanish, Hindi, Vietnamese and Simplified Chinese. MLQA is highly parallel, with QA instances parallel between
4 different languages on average.
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'test' | 5335 |
'validation' | 517 |
- Fitur :
{
"id": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"title": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"context": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"question": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"answers": {
"feature": {
"answer_start": {
"dtype": "int32",
"id": null,
"_type": "Value"
},
"text": {
"dtype": "string",
"id": null,
"_type": "Value"
}
},
"length": -1,
"id": null,
"_type": "Sequence"
}
}
MLQA.ar.de
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/MLQA.ar.de')
- Keterangan :
MLQA (MultiLingual Question Answering) is a benchmark dataset for evaluating cross-lingual question answering performance.
MLQA consists of over 5K extractive QA instances (12K in English) in SQuAD format in seven languages - English, Arabic,
German, Spanish, Hindi, Vietnamese and Simplified Chinese. MLQA is highly parallel, with QA instances parallel between
4 different languages on average.
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'test' | 1649 |
'validation' | 207 |
- Fitur :
{
"id": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"title": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"context": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"question": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"answers": {
"feature": {
"answer_start": {
"dtype": "int32",
"id": null,
"_type": "Value"
},
"text": {
"dtype": "string",
"id": null,
"_type": "Value"
}
},
"length": -1,
"id": null,
"_type": "Sequence"
}
}
MLQA.ar.vi
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/MLQA.ar.vi')
- Keterangan :
MLQA (MultiLingual Question Answering) is a benchmark dataset for evaluating cross-lingual question answering performance.
MLQA consists of over 5K extractive QA instances (12K in English) in SQuAD format in seven languages - English, Arabic,
German, Spanish, Hindi, Vietnamese and Simplified Chinese. MLQA is highly parallel, with QA instances parallel between
4 different languages on average.
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'test' | 2047 |
'validation' | 163 |
- Fitur :
{
"id": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"title": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"context": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"question": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"answers": {
"feature": {
"answer_start": {
"dtype": "int32",
"id": null,
"_type": "Value"
},
"text": {
"dtype": "string",
"id": null,
"_type": "Value"
}
},
"length": -1,
"id": null,
"_type": "Sequence"
}
}
MLQA.ar.zh
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/MLQA.ar.zh')
- Keterangan :
MLQA (MultiLingual Question Answering) is a benchmark dataset for evaluating cross-lingual question answering performance.
MLQA consists of over 5K extractive QA instances (12K in English) in SQuAD format in seven languages - English, Arabic,
German, Spanish, Hindi, Vietnamese and Simplified Chinese. MLQA is highly parallel, with QA instances parallel between
4 different languages on average.
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'test' | 1912 |
'validation' | 188 |
- Fitur :
{
"id": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"title": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"context": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"question": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"answers": {
"feature": {
"answer_start": {
"dtype": "int32",
"id": null,
"_type": "Value"
},
"text": {
"dtype": "string",
"id": null,
"_type": "Value"
}
},
"length": -1,
"id": null,
"_type": "Sequence"
}
}
MLQA.ar.en
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/MLQA.ar.en')
- Keterangan :
MLQA (MultiLingual Question Answering) is a benchmark dataset for evaluating cross-lingual question answering performance.
MLQA consists of over 5K extractive QA instances (12K in English) in SQuAD format in seven languages - English, Arabic,
German, Spanish, Hindi, Vietnamese and Simplified Chinese. MLQA is highly parallel, with QA instances parallel between
4 different languages on average.
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'test' | 5335 |
'validation' | 517 |
- Fitur :
{
"id": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"title": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"context": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"question": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"answers": {
"feature": {
"answer_start": {
"dtype": "int32",
"id": null,
"_type": "Value"
},
"text": {
"dtype": "string",
"id": null,
"_type": "Value"
}
},
"length": -1,
"id": null,
"_type": "Sequence"
}
}
MLQA.ar.es
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/MLQA.ar.es')
- Keterangan :
MLQA (MultiLingual Question Answering) is a benchmark dataset for evaluating cross-lingual question answering performance.
MLQA consists of over 5K extractive QA instances (12K in English) in SQuAD format in seven languages - English, Arabic,
German, Spanish, Hindi, Vietnamese and Simplified Chinese. MLQA is highly parallel, with QA instances parallel between
4 different languages on average.
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'test' | 1978 |
'validation' | 161 |
- Fitur :
{
"id": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"title": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"context": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"question": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"answers": {
"feature": {
"answer_start": {
"dtype": "int32",
"id": null,
"_type": "Value"
},
"text": {
"dtype": "string",
"id": null,
"_type": "Value"
}
},
"length": -1,
"id": null,
"_type": "Sequence"
}
}
MLQA.ar.hi
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/MLQA.ar.hi')
- Keterangan :
MLQA (MultiLingual Question Answering) is a benchmark dataset for evaluating cross-lingual question answering performance.
MLQA consists of over 5K extractive QA instances (12K in English) in SQuAD format in seven languages - English, Arabic,
German, Spanish, Hindi, Vietnamese and Simplified Chinese. MLQA is highly parallel, with QA instances parallel between
4 different languages on average.
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'test' | 1831 |
'validation' | 186 |
- Fitur :
{
"id": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"title": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"context": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"question": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"answers": {
"feature": {
"answer_start": {
"dtype": "int32",
"id": null,
"_type": "Value"
},
"text": {
"dtype": "string",
"id": null,
"_type": "Value"
}
},
"length": -1,
"id": null,
"_type": "Sequence"
}
}
MLQA.de.ar
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/MLQA.de.ar')
- Keterangan :
MLQA (MultiLingual Question Answering) is a benchmark dataset for evaluating cross-lingual question answering performance.
MLQA consists of over 5K extractive QA instances (12K in English) in SQuAD format in seven languages - English, Arabic,
German, Spanish, Hindi, Vietnamese and Simplified Chinese. MLQA is highly parallel, with QA instances parallel between
4 different languages on average.
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'test' | 1649 |
'validation' | 207 |
- Fitur :
{
"id": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"title": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"context": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"question": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"answers": {
"feature": {
"answer_start": {
"dtype": "int32",
"id": null,
"_type": "Value"
},
"text": {
"dtype": "string",
"id": null,
"_type": "Value"
}
},
"length": -1,
"id": null,
"_type": "Sequence"
}
}
MLQA.de.de
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/MLQA.de.de')
- Keterangan :
MLQA (MultiLingual Question Answering) is a benchmark dataset for evaluating cross-lingual question answering performance.
MLQA consists of over 5K extractive QA instances (12K in English) in SQuAD format in seven languages - English, Arabic,
German, Spanish, Hindi, Vietnamese and Simplified Chinese. MLQA is highly parallel, with QA instances parallel between
4 different languages on average.
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'test' | 4517 |
'validation' | 512 |
- Fitur :
{
"id": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"title": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"context": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"question": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"answers": {
"feature": {
"answer_start": {
"dtype": "int32",
"id": null,
"_type": "Value"
},
"text": {
"dtype": "string",
"id": null,
"_type": "Value"
}
},
"length": -1,
"id": null,
"_type": "Sequence"
}
}
MLQA.de.vi
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/MLQA.de.vi')
- Keterangan :
MLQA (MultiLingual Question Answering) is a benchmark dataset for evaluating cross-lingual question answering performance.
MLQA consists of over 5K extractive QA instances (12K in English) in SQuAD format in seven languages - English, Arabic,
German, Spanish, Hindi, Vietnamese and Simplified Chinese. MLQA is highly parallel, with QA instances parallel between
4 different languages on average.
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'test' | 1675 |
'validation' | 182 |
- Fitur :
{
"id": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"title": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"context": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"question": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"answers": {
"feature": {
"answer_start": {
"dtype": "int32",
"id": null,
"_type": "Value"
},
"text": {
"dtype": "string",
"id": null,
"_type": "Value"
}
},
"length": -1,
"id": null,
"_type": "Sequence"
}
}
MLQA.de.zh
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/MLQA.de.zh')
- Keterangan :
MLQA (MultiLingual Question Answering) is a benchmark dataset for evaluating cross-lingual question answering performance.
MLQA consists of over 5K extractive QA instances (12K in English) in SQuAD format in seven languages - English, Arabic,
German, Spanish, Hindi, Vietnamese and Simplified Chinese. MLQA is highly parallel, with QA instances parallel between
4 different languages on average.
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'test' | 1621 |
'validation' | 190 |
- Fitur :
{
"id": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"title": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"context": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"question": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"answers": {
"feature": {
"answer_start": {
"dtype": "int32",
"id": null,
"_type": "Value"
},
"text": {
"dtype": "string",
"id": null,
"_type": "Value"
}
},
"length": -1,
"id": null,
"_type": "Sequence"
}
}
MLQA.de.en
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/MLQA.de.en')
- Keterangan :
MLQA (MultiLingual Question Answering) is a benchmark dataset for evaluating cross-lingual question answering performance.
MLQA consists of over 5K extractive QA instances (12K in English) in SQuAD format in seven languages - English, Arabic,
German, Spanish, Hindi, Vietnamese and Simplified Chinese. MLQA is highly parallel, with QA instances parallel between
4 different languages on average.
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'test' | 4517 |
'validation' | 512 |
- Fitur :
{
"id": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"title": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"context": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"question": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"answers": {
"feature": {
"answer_start": {
"dtype": "int32",
"id": null,
"_type": "Value"
},
"text": {
"dtype": "string",
"id": null,
"_type": "Value"
}
},
"length": -1,
"id": null,
"_type": "Sequence"
}
}
MLQA.de.es
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/MLQA.de.es')
- Keterangan :
MLQA (MultiLingual Question Answering) is a benchmark dataset for evaluating cross-lingual question answering performance.
MLQA consists of over 5K extractive QA instances (12K in English) in SQuAD format in seven languages - English, Arabic,
German, Spanish, Hindi, Vietnamese and Simplified Chinese. MLQA is highly parallel, with QA instances parallel between
4 different languages on average.
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'test' | 1776 |
'validation' | 196 |
- Fitur :
{
"id": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"title": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"context": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"question": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"answers": {
"feature": {
"answer_start": {
"dtype": "int32",
"id": null,
"_type": "Value"
},
"text": {
"dtype": "string",
"id": null,
"_type": "Value"
}
},
"length": -1,
"id": null,
"_type": "Sequence"
}
}
MLQA.de.hi
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/MLQA.de.hi')
- Keterangan :
MLQA (MultiLingual Question Answering) is a benchmark dataset for evaluating cross-lingual question answering performance.
MLQA consists of over 5K extractive QA instances (12K in English) in SQuAD format in seven languages - English, Arabic,
German, Spanish, Hindi, Vietnamese and Simplified Chinese. MLQA is highly parallel, with QA instances parallel between
4 different languages on average.
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'test' | 1430 |
'validation' | 163 |
- Fitur :
{
"id": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"title": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"context": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"question": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"answers": {
"feature": {
"answer_start": {
"dtype": "int32",
"id": null,
"_type": "Value"
},
"text": {
"dtype": "string",
"id": null,
"_type": "Value"
}
},
"length": -1,
"id": null,
"_type": "Sequence"
}
}
MLQA.vi.ar
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/MLQA.vi.ar')
- Keterangan :
MLQA (MultiLingual Question Answering) is a benchmark dataset for evaluating cross-lingual question answering performance.
MLQA consists of over 5K extractive QA instances (12K in English) in SQuAD format in seven languages - English, Arabic,
German, Spanish, Hindi, Vietnamese and Simplified Chinese. MLQA is highly parallel, with QA instances parallel between
4 different languages on average.
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'test' | 2047 |
'validation' | 163 |
- Fitur :
{
"id": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"title": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"context": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"question": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"answers": {
"feature": {
"answer_start": {
"dtype": "int32",
"id": null,
"_type": "Value"
},
"text": {
"dtype": "string",
"id": null,
"_type": "Value"
}
},
"length": -1,
"id": null,
"_type": "Sequence"
}
}
MLQA.vi.de
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/MLQA.vi.de')
- Keterangan :
MLQA (MultiLingual Question Answering) is a benchmark dataset for evaluating cross-lingual question answering performance.
MLQA consists of over 5K extractive QA instances (12K in English) in SQuAD format in seven languages - English, Arabic,
German, Spanish, Hindi, Vietnamese and Simplified Chinese. MLQA is highly parallel, with QA instances parallel between
4 different languages on average.
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'test' | 1675 |
'validation' | 182 |
- Fitur :
{
"id": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"title": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"context": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"question": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"answers": {
"feature": {
"answer_start": {
"dtype": "int32",
"id": null,
"_type": "Value"
},
"text": {
"dtype": "string",
"id": null,
"_type": "Value"
}
},
"length": -1,
"id": null,
"_type": "Sequence"
}
}
MLQA.vi.vi
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/MLQA.vi.vi')
- Keterangan :
MLQA (MultiLingual Question Answering) is a benchmark dataset for evaluating cross-lingual question answering performance.
MLQA consists of over 5K extractive QA instances (12K in English) in SQuAD format in seven languages - English, Arabic,
German, Spanish, Hindi, Vietnamese and Simplified Chinese. MLQA is highly parallel, with QA instances parallel between
4 different languages on average.
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'test' | 5495 |
'validation' | 511 |
- Fitur :
{
"id": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"title": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"context": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"question": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"answers": {
"feature": {
"answer_start": {
"dtype": "int32",
"id": null,
"_type": "Value"
},
"text": {
"dtype": "string",
"id": null,
"_type": "Value"
}
},
"length": -1,
"id": null,
"_type": "Sequence"
}
}
MLQA.vi.zh
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/MLQA.vi.zh')
- Keterangan :
MLQA (MultiLingual Question Answering) is a benchmark dataset for evaluating cross-lingual question answering performance.
MLQA consists of over 5K extractive QA instances (12K in English) in SQuAD format in seven languages - English, Arabic,
German, Spanish, Hindi, Vietnamese and Simplified Chinese. MLQA is highly parallel, with QA instances parallel between
4 different languages on average.
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'test' | 1943 |
'validation' | 184 |
- Fitur :
{
"id": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"title": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"context": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"question": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"answers": {
"feature": {
"answer_start": {
"dtype": "int32",
"id": null,
"_type": "Value"
},
"text": {
"dtype": "string",
"id": null,
"_type": "Value"
}
},
"length": -1,
"id": null,
"_type": "Sequence"
}
}
MLQA.vi.en
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/MLQA.vi.en')
- Keterangan :
MLQA (MultiLingual Question Answering) is a benchmark dataset for evaluating cross-lingual question answering performance.
MLQA consists of over 5K extractive QA instances (12K in English) in SQuAD format in seven languages - English, Arabic,
German, Spanish, Hindi, Vietnamese and Simplified Chinese. MLQA is highly parallel, with QA instances parallel between
4 different languages on average.
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'test' | 5495 |
'validation' | 511 |
- Fitur :
{
"id": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"title": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"context": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"question": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"answers": {
"feature": {
"answer_start": {
"dtype": "int32",
"id": null,
"_type": "Value"
},
"text": {
"dtype": "string",
"id": null,
"_type": "Value"
}
},
"length": -1,
"id": null,
"_type": "Sequence"
}
}
MLQA.vi.es
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/MLQA.vi.es')
- Keterangan :
MLQA (MultiLingual Question Answering) is a benchmark dataset for evaluating cross-lingual question answering performance.
MLQA consists of over 5K extractive QA instances (12K in English) in SQuAD format in seven languages - English, Arabic,
German, Spanish, Hindi, Vietnamese and Simplified Chinese. MLQA is highly parallel, with QA instances parallel between
4 different languages on average.
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'test' | 2018 |
'validation' | 189 |
- Fitur :
{
"id": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"title": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"context": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"question": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"answers": {
"feature": {
"answer_start": {
"dtype": "int32",
"id": null,
"_type": "Value"
},
"text": {
"dtype": "string",
"id": null,
"_type": "Value"
}
},
"length": -1,
"id": null,
"_type": "Sequence"
}
}
MLQA.vi.hi
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/MLQA.vi.hi')
- Keterangan :
MLQA (MultiLingual Question Answering) is a benchmark dataset for evaluating cross-lingual question answering performance.
MLQA consists of over 5K extractive QA instances (12K in English) in SQuAD format in seven languages - English, Arabic,
German, Spanish, Hindi, Vietnamese and Simplified Chinese. MLQA is highly parallel, with QA instances parallel between
4 different languages on average.
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'test' | 1947 |
'validation' | 177 |
- Fitur :
{
"id": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"title": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"context": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"question": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"answers": {
"feature": {
"answer_start": {
"dtype": "int32",
"id": null,
"_type": "Value"
},
"text": {
"dtype": "string",
"id": null,
"_type": "Value"
}
},
"length": -1,
"id": null,
"_type": "Sequence"
}
}
MLQA.zh.ar
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/MLQA.zh.ar')
- Keterangan :
MLQA (MultiLingual Question Answering) is a benchmark dataset for evaluating cross-lingual question answering performance.
MLQA consists of over 5K extractive QA instances (12K in English) in SQuAD format in seven languages - English, Arabic,
German, Spanish, Hindi, Vietnamese and Simplified Chinese. MLQA is highly parallel, with QA instances parallel between
4 different languages on average.
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'test' | 1912 |
'validation' | 188 |
- Fitur :
{
"id": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"title": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"context": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"question": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"answers": {
"feature": {
"answer_start": {
"dtype": "int32",
"id": null,
"_type": "Value"
},
"text": {
"dtype": "string",
"id": null,
"_type": "Value"
}
},
"length": -1,
"id": null,
"_type": "Sequence"
}
}
MLQA.zh.de
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/MLQA.zh.de')
- Keterangan :
MLQA (MultiLingual Question Answering) is a benchmark dataset for evaluating cross-lingual question answering performance.
MLQA consists of over 5K extractive QA instances (12K in English) in SQuAD format in seven languages - English, Arabic,
German, Spanish, Hindi, Vietnamese and Simplified Chinese. MLQA is highly parallel, with QA instances parallel between
4 different languages on average.
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'test' | 1621 |
'validation' | 190 |
- Fitur :
{
"id": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"title": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"context": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"question": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"answers": {
"feature": {
"answer_start": {
"dtype": "int32",
"id": null,
"_type": "Value"
},
"text": {
"dtype": "string",
"id": null,
"_type": "Value"
}
},
"length": -1,
"id": null,
"_type": "Sequence"
}
}
MLQA.zh.vi
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/MLQA.zh.vi')
- Keterangan :
MLQA (MultiLingual Question Answering) is a benchmark dataset for evaluating cross-lingual question answering performance.
MLQA consists of over 5K extractive QA instances (12K in English) in SQuAD format in seven languages - English, Arabic,
German, Spanish, Hindi, Vietnamese and Simplified Chinese. MLQA is highly parallel, with QA instances parallel between
4 different languages on average.
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'test' | 1943 |
'validation' | 184 |
- Fitur :
{
"id": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"title": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"context": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"question": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"answers": {
"feature": {
"answer_start": {
"dtype": "int32",
"id": null,
"_type": "Value"
},
"text": {
"dtype": "string",
"id": null,
"_type": "Value"
}
},
"length": -1,
"id": null,
"_type": "Sequence"
}
}
MLQA.zh.zh
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/MLQA.zh.zh')
- Keterangan :
MLQA (MultiLingual Question Answering) is a benchmark dataset for evaluating cross-lingual question answering performance.
MLQA consists of over 5K extractive QA instances (12K in English) in SQuAD format in seven languages - English, Arabic,
German, Spanish, Hindi, Vietnamese and Simplified Chinese. MLQA is highly parallel, with QA instances parallel between
4 different languages on average.
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'test' | 5137 |
'validation' | 504 |
- Fitur :
{
"id": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"title": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"context": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"question": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"answers": {
"feature": {
"answer_start": {
"dtype": "int32",
"id": null,
"_type": "Value"
},
"text": {
"dtype": "string",
"id": null,
"_type": "Value"
}
},
"length": -1,
"id": null,
"_type": "Sequence"
}
}
MLQA.zh.en
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/MLQA.zh.en')
- Keterangan :
MLQA (MultiLingual Question Answering) is a benchmark dataset for evaluating cross-lingual question answering performance.
MLQA consists of over 5K extractive QA instances (12K in English) in SQuAD format in seven languages - English, Arabic,
German, Spanish, Hindi, Vietnamese and Simplified Chinese. MLQA is highly parallel, with QA instances parallel between
4 different languages on average.
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'test' | 5137 |
'validation' | 504 |
- Fitur :
{
"id": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"title": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"context": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"question": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"answers": {
"feature": {
"answer_start": {
"dtype": "int32",
"id": null,
"_type": "Value"
},
"text": {
"dtype": "string",
"id": null,
"_type": "Value"
}
},
"length": -1,
"id": null,
"_type": "Sequence"
}
}
MLQA.zh.es
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/MLQA.zh.es')
- Keterangan :
MLQA (MultiLingual Question Answering) is a benchmark dataset for evaluating cross-lingual question answering performance.
MLQA consists of over 5K extractive QA instances (12K in English) in SQuAD format in seven languages - English, Arabic,
German, Spanish, Hindi, Vietnamese and Simplified Chinese. MLQA is highly parallel, with QA instances parallel between
4 different languages on average.
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'test' | 1947 |
'validation' | 161 |
- Fitur :
{
"id": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"title": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"context": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"question": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"answers": {
"feature": {
"answer_start": {
"dtype": "int32",
"id": null,
"_type": "Value"
},
"text": {
"dtype": "string",
"id": null,
"_type": "Value"
}
},
"length": -1,
"id": null,
"_type": "Sequence"
}
}
MLQA.zh.hai
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/MLQA.zh.hi')
- Keterangan :
MLQA (MultiLingual Question Answering) is a benchmark dataset for evaluating cross-lingual question answering performance.
MLQA consists of over 5K extractive QA instances (12K in English) in SQuAD format in seven languages - English, Arabic,
German, Spanish, Hindi, Vietnamese and Simplified Chinese. MLQA is highly parallel, with QA instances parallel between
4 different languages on average.
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'test' | 1767 |
'validation' | 189 |
- Fitur :
{
"id": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"title": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"context": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"question": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"answers": {
"feature": {
"answer_start": {
"dtype": "int32",
"id": null,
"_type": "Value"
},
"text": {
"dtype": "string",
"id": null,
"_type": "Value"
}
},
"length": -1,
"id": null,
"_type": "Sequence"
}
}
MLQA.en.ar
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/MLQA.en.ar')
- Keterangan :
MLQA (MultiLingual Question Answering) is a benchmark dataset for evaluating cross-lingual question answering performance.
MLQA consists of over 5K extractive QA instances (12K in English) in SQuAD format in seven languages - English, Arabic,
German, Spanish, Hindi, Vietnamese and Simplified Chinese. MLQA is highly parallel, with QA instances parallel between
4 different languages on average.
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'test' | 5335 |
'validation' | 517 |
- Fitur :
{
"id": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"title": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"context": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"question": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"answers": {
"feature": {
"answer_start": {
"dtype": "int32",
"id": null,
"_type": "Value"
},
"text": {
"dtype": "string",
"id": null,
"_type": "Value"
}
},
"length": -1,
"id": null,
"_type": "Sequence"
}
}
MLQA.en.de
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/MLQA.en.de')
- Keterangan :
MLQA (MultiLingual Question Answering) is a benchmark dataset for evaluating cross-lingual question answering performance.
MLQA consists of over 5K extractive QA instances (12K in English) in SQuAD format in seven languages - English, Arabic,
German, Spanish, Hindi, Vietnamese and Simplified Chinese. MLQA is highly parallel, with QA instances parallel between
4 different languages on average.
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'test' | 4517 |
'validation' | 512 |
- Fitur :
{
"id": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"title": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"context": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"question": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"answers": {
"feature": {
"answer_start": {
"dtype": "int32",
"id": null,
"_type": "Value"
},
"text": {
"dtype": "string",
"id": null,
"_type": "Value"
}
},
"length": -1,
"id": null,
"_type": "Sequence"
}
}
MLQA.en.vi
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/MLQA.en.vi')
- Keterangan :
MLQA (MultiLingual Question Answering) is a benchmark dataset for evaluating cross-lingual question answering performance.
MLQA consists of over 5K extractive QA instances (12K in English) in SQuAD format in seven languages - English, Arabic,
German, Spanish, Hindi, Vietnamese and Simplified Chinese. MLQA is highly parallel, with QA instances parallel between
4 different languages on average.
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'test' | 5495 |
'validation' | 511 |
- Fitur :
{
"id": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"title": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"context": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"question": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"answers": {
"feature": {
"answer_start": {
"dtype": "int32",
"id": null,
"_type": "Value"
},
"text": {
"dtype": "string",
"id": null,
"_type": "Value"
}
},
"length": -1,
"id": null,
"_type": "Sequence"
}
}
MLQA.en.zh
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/MLQA.en.zh')
- Keterangan :
MLQA (MultiLingual Question Answering) is a benchmark dataset for evaluating cross-lingual question answering performance.
MLQA consists of over 5K extractive QA instances (12K in English) in SQuAD format in seven languages - English, Arabic,
German, Spanish, Hindi, Vietnamese and Simplified Chinese. MLQA is highly parallel, with QA instances parallel between
4 different languages on average.
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'test' | 5137 |
'validation' | 504 |
- Fitur :
{
"id": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"title": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"context": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"question": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"answers": {
"feature": {
"answer_start": {
"dtype": "int32",
"id": null,
"_type": "Value"
},
"text": {
"dtype": "string",
"id": null,
"_type": "Value"
}
},
"length": -1,
"id": null,
"_type": "Sequence"
}
}
MLQA.en.en
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/MLQA.en.en')
- Keterangan :
MLQA (MultiLingual Question Answering) is a benchmark dataset for evaluating cross-lingual question answering performance.
MLQA consists of over 5K extractive QA instances (12K in English) in SQuAD format in seven languages - English, Arabic,
German, Spanish, Hindi, Vietnamese and Simplified Chinese. MLQA is highly parallel, with QA instances parallel between
4 different languages on average.
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'test' | 11590 |
'validation' | 1148 |
- Fitur :
{
"id": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"title": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"context": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"question": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"answers": {
"feature": {
"answer_start": {
"dtype": "int32",
"id": null,
"_type": "Value"
},
"text": {
"dtype": "string",
"id": null,
"_type": "Value"
}
},
"length": -1,
"id": null,
"_type": "Sequence"
}
}
MLQA.en.es
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/MLQA.en.es')
- Keterangan :
MLQA (MultiLingual Question Answering) is a benchmark dataset for evaluating cross-lingual question answering performance.
MLQA consists of over 5K extractive QA instances (12K in English) in SQuAD format in seven languages - English, Arabic,
German, Spanish, Hindi, Vietnamese and Simplified Chinese. MLQA is highly parallel, with QA instances parallel between
4 different languages on average.
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'test' | 5253 |
'validation' | 500 |
- Fitur :
{
"id": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"title": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"context": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"question": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"answers": {
"feature": {
"answer_start": {
"dtype": "int32",
"id": null,
"_type": "Value"
},
"text": {
"dtype": "string",
"id": null,
"_type": "Value"
}
},
"length": -1,
"id": null,
"_type": "Sequence"
}
}
MLQA.en.hi
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/MLQA.en.hi')
- Keterangan :
MLQA (MultiLingual Question Answering) is a benchmark dataset for evaluating cross-lingual question answering performance.
MLQA consists of over 5K extractive QA instances (12K in English) in SQuAD format in seven languages - English, Arabic,
German, Spanish, Hindi, Vietnamese and Simplified Chinese. MLQA is highly parallel, with QA instances parallel between
4 different languages on average.
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'test' | 4918 |
'validation' | 507 |
- Fitur :
{
"id": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"title": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"context": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"question": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"answers": {
"feature": {
"answer_start": {
"dtype": "int32",
"id": null,
"_type": "Value"
},
"text": {
"dtype": "string",
"id": null,
"_type": "Value"
}
},
"length": -1,
"id": null,
"_type": "Sequence"
}
}
MLQA.es.ar
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/MLQA.es.ar')
- Keterangan :
MLQA (MultiLingual Question Answering) is a benchmark dataset for evaluating cross-lingual question answering performance.
MLQA consists of over 5K extractive QA instances (12K in English) in SQuAD format in seven languages - English, Arabic,
German, Spanish, Hindi, Vietnamese and Simplified Chinese. MLQA is highly parallel, with QA instances parallel between
4 different languages on average.
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'test' | 1978 |
'validation' | 161 |
- Fitur :
{
"id": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"title": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"context": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"question": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"answers": {
"feature": {
"answer_start": {
"dtype": "int32",
"id": null,
"_type": "Value"
},
"text": {
"dtype": "string",
"id": null,
"_type": "Value"
}
},
"length": -1,
"id": null,
"_type": "Sequence"
}
}
MLQA.es.de
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/MLQA.es.de')
- Keterangan :
MLQA (MultiLingual Question Answering) is a benchmark dataset for evaluating cross-lingual question answering performance.
MLQA consists of over 5K extractive QA instances (12K in English) in SQuAD format in seven languages - English, Arabic,
German, Spanish, Hindi, Vietnamese and Simplified Chinese. MLQA is highly parallel, with QA instances parallel between
4 different languages on average.
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'test' | 1776 |
'validation' | 196 |
- Fitur :
{
"id": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"title": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"context": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"question": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"answers": {
"feature": {
"answer_start": {
"dtype": "int32",
"id": null,
"_type": "Value"
},
"text": {
"dtype": "string",
"id": null,
"_type": "Value"
}
},
"length": -1,
"id": null,
"_type": "Sequence"
}
}
MLQA.es.vi
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/MLQA.es.vi')
- Keterangan :
MLQA (MultiLingual Question Answering) is a benchmark dataset for evaluating cross-lingual question answering performance.
MLQA consists of over 5K extractive QA instances (12K in English) in SQuAD format in seven languages - English, Arabic,
German, Spanish, Hindi, Vietnamese and Simplified Chinese. MLQA is highly parallel, with QA instances parallel between
4 different languages on average.
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'test' | 2018 |
'validation' | 189 |
- Fitur :
{
"id": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"title": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"context": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"question": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"answers": {
"feature": {
"answer_start": {
"dtype": "int32",
"id": null,
"_type": "Value"
},
"text": {
"dtype": "string",
"id": null,
"_type": "Value"
}
},
"length": -1,
"id": null,
"_type": "Sequence"
}
}
MLQA.es.zh
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/MLQA.es.zh')
- Keterangan :
MLQA (MultiLingual Question Answering) is a benchmark dataset for evaluating cross-lingual question answering performance.
MLQA consists of over 5K extractive QA instances (12K in English) in SQuAD format in seven languages - English, Arabic,
German, Spanish, Hindi, Vietnamese and Simplified Chinese. MLQA is highly parallel, with QA instances parallel between
4 different languages on average.
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'test' | 1947 |
'validation' | 161 |
- Fitur :
{
"id": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"title": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"context": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"question": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"answers": {
"feature": {
"answer_start": {
"dtype": "int32",
"id": null,
"_type": "Value"
},
"text": {
"dtype": "string",
"id": null,
"_type": "Value"
}
},
"length": -1,
"id": null,
"_type": "Sequence"
}
}
MLQA.es.en
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/MLQA.es.en')
- Keterangan :
MLQA (MultiLingual Question Answering) is a benchmark dataset for evaluating cross-lingual question answering performance.
MLQA consists of over 5K extractive QA instances (12K in English) in SQuAD format in seven languages - English, Arabic,
German, Spanish, Hindi, Vietnamese and Simplified Chinese. MLQA is highly parallel, with QA instances parallel between
4 different languages on average.
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'test' | 5253 |
'validation' | 500 |
- Fitur :
{
"id": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"title": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"context": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"question": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"answers": {
"feature": {
"answer_start": {
"dtype": "int32",
"id": null,
"_type": "Value"
},
"text": {
"dtype": "string",
"id": null,
"_type": "Value"
}
},
"length": -1,
"id": null,
"_type": "Sequence"
}
}
MLQA.es.es
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/MLQA.es.es')
- Keterangan :
MLQA (MultiLingual Question Answering) is a benchmark dataset for evaluating cross-lingual question answering performance.
MLQA consists of over 5K extractive QA instances (12K in English) in SQuAD format in seven languages - English, Arabic,
German, Spanish, Hindi, Vietnamese and Simplified Chinese. MLQA is highly parallel, with QA instances parallel between
4 different languages on average.
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'test' | 5253 |
'validation' | 500 |
- Fitur :
{
"id": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"title": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"context": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"question": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"answers": {
"feature": {
"answer_start": {
"dtype": "int32",
"id": null,
"_type": "Value"
},
"text": {
"dtype": "string",
"id": null,
"_type": "Value"
}
},
"length": -1,
"id": null,
"_type": "Sequence"
}
}
MLQA.e.hi
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/MLQA.es.hi')
- Keterangan :
MLQA (MultiLingual Question Answering) is a benchmark dataset for evaluating cross-lingual question answering performance.
MLQA consists of over 5K extractive QA instances (12K in English) in SQuAD format in seven languages - English, Arabic,
German, Spanish, Hindi, Vietnamese and Simplified Chinese. MLQA is highly parallel, with QA instances parallel between
4 different languages on average.
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'test' | 1723 |
'validation' | 187 |
- Fitur :
{
"id": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"title": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"context": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"question": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"answers": {
"feature": {
"answer_start": {
"dtype": "int32",
"id": null,
"_type": "Value"
},
"text": {
"dtype": "string",
"id": null,
"_type": "Value"
}
},
"length": -1,
"id": null,
"_type": "Sequence"
}
}
MLQA.hi.ar
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/MLQA.hi.ar')
- Keterangan :
MLQA (MultiLingual Question Answering) is a benchmark dataset for evaluating cross-lingual question answering performance.
MLQA consists of over 5K extractive QA instances (12K in English) in SQuAD format in seven languages - English, Arabic,
German, Spanish, Hindi, Vietnamese and Simplified Chinese. MLQA is highly parallel, with QA instances parallel between
4 different languages on average.
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'test' | 1831 |
'validation' | 186 |
- Fitur :
{
"id": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"title": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"context": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"question": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"answers": {
"feature": {
"answer_start": {
"dtype": "int32",
"id": null,
"_type": "Value"
},
"text": {
"dtype": "string",
"id": null,
"_type": "Value"
}
},
"length": -1,
"id": null,
"_type": "Sequence"
}
}
MLQA.h.de
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/MLQA.hi.de')
- Keterangan :
MLQA (MultiLingual Question Answering) is a benchmark dataset for evaluating cross-lingual question answering performance.
MLQA consists of over 5K extractive QA instances (12K in English) in SQuAD format in seven languages - English, Arabic,
German, Spanish, Hindi, Vietnamese and Simplified Chinese. MLQA is highly parallel, with QA instances parallel between
4 different languages on average.
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'test' | 1430 |
'validation' | 163 |
- Fitur :
{
"id": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"title": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"context": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"question": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"answers": {
"feature": {
"answer_start": {
"dtype": "int32",
"id": null,
"_type": "Value"
},
"text": {
"dtype": "string",
"id": null,
"_type": "Value"
}
},
"length": -1,
"id": null,
"_type": "Sequence"
}
}
MLQA.hi.vi
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/MLQA.hi.vi')
- Keterangan :
MLQA (MultiLingual Question Answering) is a benchmark dataset for evaluating cross-lingual question answering performance.
MLQA consists of over 5K extractive QA instances (12K in English) in SQuAD format in seven languages - English, Arabic,
German, Spanish, Hindi, Vietnamese and Simplified Chinese. MLQA is highly parallel, with QA instances parallel between
4 different languages on average.
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'test' | 1947 |
'validation' | 177 |
- Fitur :
{
"id": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"title": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"context": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"question": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"answers": {
"feature": {
"answer_start": {
"dtype": "int32",
"id": null,
"_type": "Value"
},
"text": {
"dtype": "string",
"id": null,
"_type": "Value"
}
},
"length": -1,
"id": null,
"_type": "Sequence"
}
}
MLQA.hi.zh
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/MLQA.hi.zh')
- Keterangan :
MLQA (MultiLingual Question Answering) is a benchmark dataset for evaluating cross-lingual question answering performance.
MLQA consists of over 5K extractive QA instances (12K in English) in SQuAD format in seven languages - English, Arabic,
German, Spanish, Hindi, Vietnamese and Simplified Chinese. MLQA is highly parallel, with QA instances parallel between
4 different languages on average.
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'test' | 1767 |
'validation' | 189 |
- Fitur :
{
"id": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"title": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"context": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"question": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"answers": {
"feature": {
"answer_start": {
"dtype": "int32",
"id": null,
"_type": "Value"
},
"text": {
"dtype": "string",
"id": null,
"_type": "Value"
}
},
"length": -1,
"id": null,
"_type": "Sequence"
}
}
MLQA.hi.en
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/MLQA.hi.en')
- Keterangan :
MLQA (MultiLingual Question Answering) is a benchmark dataset for evaluating cross-lingual question answering performance.
MLQA consists of over 5K extractive QA instances (12K in English) in SQuAD format in seven languages - English, Arabic,
German, Spanish, Hindi, Vietnamese and Simplified Chinese. MLQA is highly parallel, with QA instances parallel between
4 different languages on average.
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'test' | 4918 |
'validation' | 507 |
- Fitur :
{
"id": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"title": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"context": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"question": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"answers": {
"feature": {
"answer_start": {
"dtype": "int32",
"id": null,
"_type": "Value"
},
"text": {
"dtype": "string",
"id": null,
"_type": "Value"
}
},
"length": -1,
"id": null,
"_type": "Sequence"
}
}
MLQA.hai
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/MLQA.hi.es')
- Keterangan :
MLQA (MultiLingual Question Answering) is a benchmark dataset for evaluating cross-lingual question answering performance.
MLQA consists of over 5K extractive QA instances (12K in English) in SQuAD format in seven languages - English, Arabic,
German, Spanish, Hindi, Vietnamese and Simplified Chinese. MLQA is highly parallel, with QA instances parallel between
4 different languages on average.
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'test' | 1723 |
'validation' | 187 |
- Fitur :
{
"id": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"title": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"context": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"question": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"answers": {
"feature": {
"answer_start": {
"dtype": "int32",
"id": null,
"_type": "Value"
},
"text": {
"dtype": "string",
"id": null,
"_type": "Value"
}
},
"length": -1,
"id": null,
"_type": "Sequence"
}
}
MLQA.hi.hi
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/MLQA.hi.hi')
- Keterangan :
MLQA (MultiLingual Question Answering) is a benchmark dataset for evaluating cross-lingual question answering performance.
MLQA consists of over 5K extractive QA instances (12K in English) in SQuAD format in seven languages - English, Arabic,
German, Spanish, Hindi, Vietnamese and Simplified Chinese. MLQA is highly parallel, with QA instances parallel between
4 different languages on average.
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'test' | 4918 |
'validation' | 507 |
- Fitur :
{
"id": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"title": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"context": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"question": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"answers": {
"feature": {
"answer_start": {
"dtype": "int32",
"id": null,
"_type": "Value"
},
"text": {
"dtype": "string",
"id": null,
"_type": "Value"
}
},
"length": -1,
"id": null,
"_type": "Sequence"
}
}
XQuAD.ar
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/XQuAD.ar')
- Keterangan :
XQuAD (Cross-lingual Question Answering Dataset) is a benchmark dataset for evaluating cross-lingual question
answering performance. The dataset consists of a subset of 240 paragraphs and 1190 question-answer pairs from
the development set of SQuAD v1.1 (Rajpurkar et al., 2016) together with their professional translations into
ten languages: Spanish, German, Greek, Russian, Turkish, Arabic, Vietnamese, Thai, Chinese, and Hindi. Consequently,
the dataset is entirely parallel across 11 languages.
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'validation' | 1190 |
- Fitur :
{
"id": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"context": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"question": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"answers": {
"feature": {
"answer_start": {
"dtype": "int32",
"id": null,
"_type": "Value"
},
"text": {
"dtype": "string",
"id": null,
"_type": "Value"
}
},
"length": -1,
"id": null,
"_type": "Sequence"
}
}
XQuAD.de
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/XQuAD.de')
- Keterangan :
XQuAD (Cross-lingual Question Answering Dataset) is a benchmark dataset for evaluating cross-lingual question
answering performance. The dataset consists of a subset of 240 paragraphs and 1190 question-answer pairs from
the development set of SQuAD v1.1 (Rajpurkar et al., 2016) together with their professional translations into
ten languages: Spanish, German, Greek, Russian, Turkish, Arabic, Vietnamese, Thai, Chinese, and Hindi. Consequently,
the dataset is entirely parallel across 11 languages.
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'validation' | 1190 |
- Fitur :
{
"id": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"context": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"question": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"answers": {
"feature": {
"answer_start": {
"dtype": "int32",
"id": null,
"_type": "Value"
},
"text": {
"dtype": "string",
"id": null,
"_type": "Value"
}
},
"length": -1,
"id": null,
"_type": "Sequence"
}
}
XQuAD.vi
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/XQuAD.vi')
- Keterangan :
XQuAD (Cross-lingual Question Answering Dataset) is a benchmark dataset for evaluating cross-lingual question
answering performance. The dataset consists of a subset of 240 paragraphs and 1190 question-answer pairs from
the development set of SQuAD v1.1 (Rajpurkar et al., 2016) together with their professional translations into
ten languages: Spanish, German, Greek, Russian, Turkish, Arabic, Vietnamese, Thai, Chinese, and Hindi. Consequently,
the dataset is entirely parallel across 11 languages.
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'validation' | 1190 |
- Fitur :
{
"id": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"context": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"question": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"answers": {
"feature": {
"answer_start": {
"dtype": "int32",
"id": null,
"_type": "Value"
},
"text": {
"dtype": "string",
"id": null,
"_type": "Value"
}
},
"length": -1,
"id": null,
"_type": "Sequence"
}
}
XQuAD.zh
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/XQuAD.zh')
- Keterangan :
XQuAD (Cross-lingual Question Answering Dataset) is a benchmark dataset for evaluating cross-lingual question
answering performance. The dataset consists of a subset of 240 paragraphs and 1190 question-answer pairs from
the development set of SQuAD v1.1 (Rajpurkar et al., 2016) together with their professional translations into
ten languages: Spanish, German, Greek, Russian, Turkish, Arabic, Vietnamese, Thai, Chinese, and Hindi. Consequently,
the dataset is entirely parallel across 11 languages.
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'validation' | 1190 |
- Fitur :
{
"id": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"context": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"question": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"answers": {
"feature": {
"answer_start": {
"dtype": "int32",
"id": null,
"_type": "Value"
},
"text": {
"dtype": "string",
"id": null,
"_type": "Value"
}
},
"length": -1,
"id": null,
"_type": "Sequence"
}
}
XQuAD.en
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/XQuAD.en')
- Keterangan :
XQuAD (Cross-lingual Question Answering Dataset) is a benchmark dataset for evaluating cross-lingual question
answering performance. The dataset consists of a subset of 240 paragraphs and 1190 question-answer pairs from
the development set of SQuAD v1.1 (Rajpurkar et al., 2016) together with their professional translations into
ten languages: Spanish, German, Greek, Russian, Turkish, Arabic, Vietnamese, Thai, Chinese, and Hindi. Consequently,
the dataset is entirely parallel across 11 languages.
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'validation' | 1190 |
- Fitur :
{
"id": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"context": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"question": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"answers": {
"feature": {
"answer_start": {
"dtype": "int32",
"id": null,
"_type": "Value"
},
"text": {
"dtype": "string",
"id": null,
"_type": "Value"
}
},
"length": -1,
"id": null,
"_type": "Sequence"
}
}
XQuAD.es
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/XQuAD.es')
- Keterangan :
XQuAD (Cross-lingual Question Answering Dataset) is a benchmark dataset for evaluating cross-lingual question
answering performance. The dataset consists of a subset of 240 paragraphs and 1190 question-answer pairs from
the development set of SQuAD v1.1 (Rajpurkar et al., 2016) together with their professional translations into
ten languages: Spanish, German, Greek, Russian, Turkish, Arabic, Vietnamese, Thai, Chinese, and Hindi. Consequently,
the dataset is entirely parallel across 11 languages.
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'validation' | 1190 |
- Fitur :
{
"id": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"context": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"question": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"answers": {
"feature": {
"answer_start": {
"dtype": "int32",
"id": null,
"_type": "Value"
},
"text": {
"dtype": "string",
"id": null,
"_type": "Value"
}
},
"length": -1,
"id": null,
"_type": "Sequence"
}
}
XQuAD.hai
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/XQuAD.hi')
- Keterangan :
XQuAD (Cross-lingual Question Answering Dataset) is a benchmark dataset for evaluating cross-lingual question
answering performance. The dataset consists of a subset of 240 paragraphs and 1190 question-answer pairs from
the development set of SQuAD v1.1 (Rajpurkar et al., 2016) together with their professional translations into
ten languages: Spanish, German, Greek, Russian, Turkish, Arabic, Vietnamese, Thai, Chinese, and Hindi. Consequently,
the dataset is entirely parallel across 11 languages.
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'validation' | 1190 |
- Fitur :
{
"id": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"context": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"question": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"answers": {
"feature": {
"answer_start": {
"dtype": "int32",
"id": null,
"_type": "Value"
},
"text": {
"dtype": "string",
"id": null,
"_type": "Value"
}
},
"length": -1,
"id": null,
"_type": "Sequence"
}
}
XQuAD.el
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/XQuAD.el')
- Keterangan :
XQuAD (Cross-lingual Question Answering Dataset) is a benchmark dataset for evaluating cross-lingual question
answering performance. The dataset consists of a subset of 240 paragraphs and 1190 question-answer pairs from
the development set of SQuAD v1.1 (Rajpurkar et al., 2016) together with their professional translations into
ten languages: Spanish, German, Greek, Russian, Turkish, Arabic, Vietnamese, Thai, Chinese, and Hindi. Consequently,
the dataset is entirely parallel across 11 languages.
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'validation' | 1190 |
- Fitur :
{
"id": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"context": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"question": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"answers": {
"feature": {
"answer_start": {
"dtype": "int32",
"id": null,
"_type": "Value"
},
"text": {
"dtype": "string",
"id": null,
"_type": "Value"
}
},
"length": -1,
"id": null,
"_type": "Sequence"
}
}
XQuAD.ru
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/XQuAD.ru')
- Keterangan :
XQuAD (Cross-lingual Question Answering Dataset) is a benchmark dataset for evaluating cross-lingual question
answering performance. The dataset consists of a subset of 240 paragraphs and 1190 question-answer pairs from
the development set of SQuAD v1.1 (Rajpurkar et al., 2016) together with their professional translations into
ten languages: Spanish, German, Greek, Russian, Turkish, Arabic, Vietnamese, Thai, Chinese, and Hindi. Consequently,
the dataset is entirely parallel across 11 languages.
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'validation' | 1190 |
- Fitur :
{
"id": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"context": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"question": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"answers": {
"feature": {
"answer_start": {
"dtype": "int32",
"id": null,
"_type": "Value"
},
"text": {
"dtype": "string",
"id": null,
"_type": "Value"
}
},
"length": -1,
"id": null,
"_type": "Sequence"
}
}
XQuAD.th
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/XQuAD.th')
- Keterangan :
XQuAD (Cross-lingual Question Answering Dataset) is a benchmark dataset for evaluating cross-lingual question
answering performance. The dataset consists of a subset of 240 paragraphs and 1190 question-answer pairs from
the development set of SQuAD v1.1 (Rajpurkar et al., 2016) together with their professional translations into
ten languages: Spanish, German, Greek, Russian, Turkish, Arabic, Vietnamese, Thai, Chinese, and Hindi. Consequently,
the dataset is entirely parallel across 11 languages.
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'validation' | 1190 |
- Fitur :
{
"id": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"context": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"question": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"answers": {
"feature": {
"answer_start": {
"dtype": "int32",
"id": null,
"_type": "Value"
},
"text": {
"dtype": "string",
"id": null,
"_type": "Value"
}
},
"length": -1,
"id": null,
"_type": "Sequence"
}
}
XQuAD.tr
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/XQuAD.tr')
- Keterangan :
XQuAD (Cross-lingual Question Answering Dataset) is a benchmark dataset for evaluating cross-lingual question
answering performance. The dataset consists of a subset of 240 paragraphs and 1190 question-answer pairs from
the development set of SQuAD v1.1 (Rajpurkar et al., 2016) together with their professional translations into
ten languages: Spanish, German, Greek, Russian, Turkish, Arabic, Vietnamese, Thai, Chinese, and Hindi. Consequently,
the dataset is entirely parallel across 11 languages.
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'validation' | 1190 |
- Fitur :
{
"id": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"context": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"question": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"answers": {
"feature": {
"answer_start": {
"dtype": "int32",
"id": null,
"_type": "Value"
},
"text": {
"dtype": "string",
"id": null,
"_type": "Value"
}
},
"length": -1,
"id": null,
"_type": "Sequence"
}
}
bucc18.de
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/bucc18.de')
- Keterangan :
Building and Using Comparable Corpora
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'test' | 9580 |
'validation' | 1038 |
- Fitur :
{
"source_sentence": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"target_sentence": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"source_lang": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"target_lang": {
"dtype": "string",
"id": null,
"_type": "Value"
}
}
bucc18.fr
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/bucc18.fr')
- Keterangan :
Building and Using Comparable Corpora
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'test' | 9086 |
'validation' | 929 |
- Fitur :
{
"source_sentence": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"target_sentence": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"source_lang": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"target_lang": {
"dtype": "string",
"id": null,
"_type": "Value"
}
}
bucc18.zh
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/bucc18.zh')
- Keterangan :
Building and Using Comparable Corpora
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'test' | 1899 |
'validation' | 257 |
- Fitur :
{
"source_sentence": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"target_sentence": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"source_lang": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"target_lang": {
"dtype": "string",
"id": null,
"_type": "Value"
}
}
bucc18.ru
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/bucc18.ru')
- Keterangan :
Building and Using Comparable Corpora
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'test' | 14435 |
'validation' | 2374 |
- Fitur :
{
"source_sentence": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"target_sentence": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"source_lang": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"target_lang": {
"dtype": "string",
"id": null,
"_type": "Value"
}
}
PAWS-X.de
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/PAWS-X.de')
- Keterangan :
This dataset contains 23,659 human translated PAWS evaluation pairs and 296,406 machine translated training
pairs in six typologically distinct languages: French, Spanish, German, Chinese, Japanese, and Korean. All
translated pairs are sourced from examples in PAWS-Wiki.
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'test' | 2000 |
'train' | 49380 |
'validation' | 2000 |
- Fitur :
{
"sentence1": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"sentence2": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"label": {
"dtype": "string",
"id": null,
"_type": "Value"
}
}
PAWS-X.en
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/PAWS-X.en')
- Keterangan :
This dataset contains 23,659 human translated PAWS evaluation pairs and 296,406 machine translated training
pairs in six typologically distinct languages: French, Spanish, German, Chinese, Japanese, and Korean. All
translated pairs are sourced from examples in PAWS-Wiki.
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'test' | 2000 |
'train' | 49175 |
'validation' | 2000 |
- Fitur :
{
"sentence1": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"sentence2": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"label": {
"dtype": "string",
"id": null,
"_type": "Value"
}
}
PAWS-X.es
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/PAWS-X.es')
- Keterangan :
This dataset contains 23,659 human translated PAWS evaluation pairs and 296,406 machine translated training
pairs in six typologically distinct languages: French, Spanish, German, Chinese, Japanese, and Korean. All
translated pairs are sourced from examples in PAWS-Wiki.
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'test' | 2000 |
'train' | 49401 |
'validation' | 1961 |
- Fitur :
{
"sentence1": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"sentence2": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"label": {
"dtype": "string",
"id": null,
"_type": "Value"
}
}
PAWS-X.fr
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/PAWS-X.fr')
- Keterangan :
This dataset contains 23,659 human translated PAWS evaluation pairs and 296,406 machine translated training
pairs in six typologically distinct languages: French, Spanish, German, Chinese, Japanese, and Korean. All
translated pairs are sourced from examples in PAWS-Wiki.
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'test' | 2000 |
'train' | 49399 |
'validation' | 1988 |
- Fitur :
{
"sentence1": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"sentence2": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"label": {
"dtype": "string",
"id": null,
"_type": "Value"
}
}
PAWS-X.ja
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/PAWS-X.ja')
- Keterangan :
This dataset contains 23,659 human translated PAWS evaluation pairs and 296,406 machine translated training
pairs in six typologically distinct languages: French, Spanish, German, Chinese, Japanese, and Korean. All
translated pairs are sourced from examples in PAWS-Wiki.
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'test' | 2000 |
'train' | 49401 |
'validation' | 2000 |
- Fitur :
{
"sentence1": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"sentence2": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"label": {
"dtype": "string",
"id": null,
"_type": "Value"
}
}
PAWS-X.ko
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/PAWS-X.ko')
- Keterangan :
This dataset contains 23,659 human translated PAWS evaluation pairs and 296,406 machine translated training
pairs in six typologically distinct languages: French, Spanish, German, Chinese, Japanese, and Korean. All
translated pairs are sourced from examples in PAWS-Wiki.
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'test' | 1999 |
'train' | 49164 |
'validation' | 2000 |
- Fitur :
{
"sentence1": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"sentence2": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"label": {
"dtype": "string",
"id": null,
"_type": "Value"
}
}
PAWS-X.zh
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/PAWS-X.zh')
- Keterangan :
This dataset contains 23,659 human translated PAWS evaluation pairs and 296,406 machine translated training
pairs in six typologically distinct languages: French, Spanish, German, Chinese, Japanese, and Korean. All
translated pairs are sourced from examples in PAWS-Wiki.
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'test' | 2000 |
'train' | 49401 |
'validation' | 2000 |
- Fitur :
{
"sentence1": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"sentence2": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"label": {
"dtype": "string",
"id": null,
"_type": "Value"
}
}
tatoeba.afr
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/tatoeba.afr')
- Keterangan :
his data is extracted from the Tatoeba corpus, dated Saturday 2018/11/17.
For each languages, we have selected 1000 English sentences and their translations, if available. Please check
this paper for a description of the languages, their families and scripts as well as baseline results.
Please note that the English sentences are not identical for all language pairs. This means that the results are
not directly comparable across languages. In particular, the sentences tend to have less variety for several
low-resource languages, e.g. "Tom needed water", "Tom needs water", "Tom is getting water", ...
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'validation' | 1000 |
- Fitur :
{
"source_sentence": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"target_sentence": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"source_lang": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"target_lang": {
"dtype": "string",
"id": null,
"_type": "Value"
}
}
tatoeba.ara
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/tatoeba.ara')
- Keterangan :
his data is extracted from the Tatoeba corpus, dated Saturday 2018/11/17.
For each languages, we have selected 1000 English sentences and their translations, if available. Please check
this paper for a description of the languages, their families and scripts as well as baseline results.
Please note that the English sentences are not identical for all language pairs. This means that the results are
not directly comparable across languages. In particular, the sentences tend to have less variety for several
low-resource languages, e.g. "Tom needed water", "Tom needs water", "Tom is getting water", ...
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'validation' | 1000 |
- Fitur :
{
"source_sentence": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"target_sentence": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"source_lang": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"target_lang": {
"dtype": "string",
"id": null,
"_type": "Value"
}
}
tatoeba.ben
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/tatoeba.ben')
- Keterangan :
his data is extracted from the Tatoeba corpus, dated Saturday 2018/11/17.
For each languages, we have selected 1000 English sentences and their translations, if available. Please check
this paper for a description of the languages, their families and scripts as well as baseline results.
Please note that the English sentences are not identical for all language pairs. This means that the results are
not directly comparable across languages. In particular, the sentences tend to have less variety for several
low-resource languages, e.g. "Tom needed water", "Tom needs water", "Tom is getting water", ...
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'validation' | 1000 |
- Fitur :
{
"source_sentence": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"target_sentence": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"source_lang": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"target_lang": {
"dtype": "string",
"id": null,
"_type": "Value"
}
}
tatoeba.bul
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/tatoeba.bul')
- Keterangan :
his data is extracted from the Tatoeba corpus, dated Saturday 2018/11/17.
For each languages, we have selected 1000 English sentences and their translations, if available. Please check
this paper for a description of the languages, their families and scripts as well as baseline results.
Please note that the English sentences are not identical for all language pairs. This means that the results are
not directly comparable across languages. In particular, the sentences tend to have less variety for several
low-resource languages, e.g. "Tom needed water", "Tom needs water", "Tom is getting water", ...
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'validation' | 1000 |
- Fitur :
{
"source_sentence": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"target_sentence": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"source_lang": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"target_lang": {
"dtype": "string",
"id": null,
"_type": "Value"
}
}
tatoeba.deu
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/tatoeba.deu')
- Keterangan :
his data is extracted from the Tatoeba corpus, dated Saturday 2018/11/17.
For each languages, we have selected 1000 English sentences and their translations, if available. Please check
this paper for a description of the languages, their families and scripts as well as baseline results.
Please note that the English sentences are not identical for all language pairs. This means that the results are
not directly comparable across languages. In particular, the sentences tend to have less variety for several
low-resource languages, e.g. "Tom needed water", "Tom needs water", "Tom is getting water", ...
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'validation' | 1000 |
- Fitur :
{
"source_sentence": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"target_sentence": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"source_lang": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"target_lang": {
"dtype": "string",
"id": null,
"_type": "Value"
}
}
tatoeba.cmn
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/tatoeba.cmn')
- Keterangan :
his data is extracted from the Tatoeba corpus, dated Saturday 2018/11/17.
For each languages, we have selected 1000 English sentences and their translations, if available. Please check
this paper for a description of the languages, their families and scripts as well as baseline results.
Please note that the English sentences are not identical for all language pairs. This means that the results are
not directly comparable across languages. In particular, the sentences tend to have less variety for several
low-resource languages, e.g. "Tom needed water", "Tom needs water", "Tom is getting water", ...
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'validation' | 1000 |
- Fitur :
{
"source_sentence": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"target_sentence": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"source_lang": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"target_lang": {
"dtype": "string",
"id": null,
"_type": "Value"
}
}
tatoeba.ell
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/tatoeba.ell')
- Keterangan :
his data is extracted from the Tatoeba corpus, dated Saturday 2018/11/17.
For each languages, we have selected 1000 English sentences and their translations, if available. Please check
this paper for a description of the languages, their families and scripts as well as baseline results.
Please note that the English sentences are not identical for all language pairs. This means that the results are
not directly comparable across languages. In particular, the sentences tend to have less variety for several
low-resource languages, e.g. "Tom needed water", "Tom needs water", "Tom is getting water", ...
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'validation' | 1000 |
- Fitur :
{
"source_sentence": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"target_sentence": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"source_lang": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"target_lang": {
"dtype": "string",
"id": null,
"_type": "Value"
}
}
tatoeba.est
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/tatoeba.est')
- Keterangan :
his data is extracted from the Tatoeba corpus, dated Saturday 2018/11/17.
For each languages, we have selected 1000 English sentences and their translations, if available. Please check
this paper for a description of the languages, their families and scripts as well as baseline results.
Please note that the English sentences are not identical for all language pairs. This means that the results are
not directly comparable across languages. In particular, the sentences tend to have less variety for several
low-resource languages, e.g. "Tom needed water", "Tom needs water", "Tom is getting water", ...
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'validation' | 1000 |
- Fitur :
{
"source_sentence": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"target_sentence": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"source_lang": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"target_lang": {
"dtype": "string",
"id": null,
"_type": "Value"
}
}
tatoeba.eus
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/tatoeba.eus')
- Keterangan :
his data is extracted from the Tatoeba corpus, dated Saturday 2018/11/17.
For each languages, we have selected 1000 English sentences and their translations, if available. Please check
this paper for a description of the languages, their families and scripts as well as baseline results.
Please note that the English sentences are not identical for all language pairs. This means that the results are
not directly comparable across languages. In particular, the sentences tend to have less variety for several
low-resource languages, e.g. "Tom needed water", "Tom needs water", "Tom is getting water", ...
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'validation' | 1000 |
- Fitur :
{
"source_sentence": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"target_sentence": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"source_lang": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"target_lang": {
"dtype": "string",
"id": null,
"_type": "Value"
}
}
tatoeba.fin
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/tatoeba.fin')
- Keterangan :
his data is extracted from the Tatoeba corpus, dated Saturday 2018/11/17.
For each languages, we have selected 1000 English sentences and their translations, if available. Please check
this paper for a description of the languages, their families and scripts as well as baseline results.
Please note that the English sentences are not identical for all language pairs. This means that the results are
not directly comparable across languages. In particular, the sentences tend to have less variety for several
low-resource languages, e.g. "Tom needed water", "Tom needs water", "Tom is getting water", ...
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'validation' | 1000 |
- Fitur :
{
"source_sentence": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"target_sentence": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"source_lang": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"target_lang": {
"dtype": "string",
"id": null,
"_type": "Value"
}
}
tatoeba.fra
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/tatoeba.fra')
- Keterangan :
his data is extracted from the Tatoeba corpus, dated Saturday 2018/11/17.
For each languages, we have selected 1000 English sentences and their translations, if available. Please check
this paper for a description of the languages, their families and scripts as well as baseline results.
Please note that the English sentences are not identical for all language pairs. This means that the results are
not directly comparable across languages. In particular, the sentences tend to have less variety for several
low-resource languages, e.g. "Tom needed water", "Tom needs water", "Tom is getting water", ...
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'validation' | 1000 |
- Fitur :
{
"source_sentence": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"target_sentence": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"source_lang": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"target_lang": {
"dtype": "string",
"id": null,
"_type": "Value"
}
}
tatoeba.heb
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/tatoeba.heb')
- Keterangan :
his data is extracted from the Tatoeba corpus, dated Saturday 2018/11/17.
For each languages, we have selected 1000 English sentences and their translations, if available. Please check
this paper for a description of the languages, their families and scripts as well as baseline results.
Please note that the English sentences are not identical for all language pairs. This means that the results are
not directly comparable across languages. In particular, the sentences tend to have less variety for several
low-resource languages, e.g. "Tom needed water", "Tom needs water", "Tom is getting water", ...
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'validation' | 1000 |
- Fitur :
{
"source_sentence": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"target_sentence": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"source_lang": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"target_lang": {
"dtype": "string",
"id": null,
"_type": "Value"
}
}
tatoeba.hin
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/tatoeba.hin')
- Keterangan :
his data is extracted from the Tatoeba corpus, dated Saturday 2018/11/17.
For each languages, we have selected 1000 English sentences and their translations, if available. Please check
this paper for a description of the languages, their families and scripts as well as baseline results.
Please note that the English sentences are not identical for all language pairs. This means that the results are
not directly comparable across languages. In particular, the sentences tend to have less variety for several
low-resource languages, e.g. "Tom needed water", "Tom needs water", "Tom is getting water", ...
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'validation' | 1000 |
- Fitur :
{
"source_sentence": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"target_sentence": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"source_lang": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"target_lang": {
"dtype": "string",
"id": null,
"_type": "Value"
}
}
tatoeba.hun
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/tatoeba.hun')
- Keterangan :
his data is extracted from the Tatoeba corpus, dated Saturday 2018/11/17.
For each languages, we have selected 1000 English sentences and their translations, if available. Please check
this paper for a description of the languages, their families and scripts as well as baseline results.
Please note that the English sentences are not identical for all language pairs. This means that the results are
not directly comparable across languages. In particular, the sentences tend to have less variety for several
low-resource languages, e.g. "Tom needed water", "Tom needs water", "Tom is getting water", ...
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'validation' | 1000 |
- Fitur :
{
"source_sentence": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"target_sentence": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"source_lang": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"target_lang": {
"dtype": "string",
"id": null,
"_type": "Value"
}
}
tatoeba.ind
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/tatoeba.ind')
- Keterangan :
his data is extracted from the Tatoeba corpus, dated Saturday 2018/11/17.
For each languages, we have selected 1000 English sentences and their translations, if available. Please check
this paper for a description of the languages, their families and scripts as well as baseline results.
Please note that the English sentences are not identical for all language pairs. This means that the results are
not directly comparable across languages. In particular, the sentences tend to have less variety for several
low-resource languages, e.g. "Tom needed water", "Tom needs water", "Tom is getting water", ...
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'validation' | 1000 |
- Fitur :
{
"source_sentence": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"target_sentence": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"source_lang": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"target_lang": {
"dtype": "string",
"id": null,
"_type": "Value"
}
}
tatoeba.ita
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/tatoeba.ita')
- Keterangan :
his data is extracted from the Tatoeba corpus, dated Saturday 2018/11/17.
For each languages, we have selected 1000 English sentences and their translations, if available. Please check
this paper for a description of the languages, their families and scripts as well as baseline results.
Please note that the English sentences are not identical for all language pairs. This means that the results are
not directly comparable across languages. In particular, the sentences tend to have less variety for several
low-resource languages, e.g. "Tom needed water", "Tom needs water", "Tom is getting water", ...
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'validation' | 1000 |
- Fitur :
{
"source_sentence": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"target_sentence": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"source_lang": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"target_lang": {
"dtype": "string",
"id": null,
"_type": "Value"
}
}
tatoeba.jav
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/tatoeba.jav')
- Keterangan :
his data is extracted from the Tatoeba corpus, dated Saturday 2018/11/17.
For each languages, we have selected 1000 English sentences and their translations, if available. Please check
this paper for a description of the languages, their families and scripts as well as baseline results.
Please note that the English sentences are not identical for all language pairs. This means that the results are
not directly comparable across languages. In particular, the sentences tend to have less variety for several
low-resource languages, e.g. "Tom needed water", "Tom needs water", "Tom is getting water", ...
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'validation' | 205 |
- Fitur :
{
"source_sentence": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"target_sentence": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"source_lang": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"target_lang": {
"dtype": "string",
"id": null,
"_type": "Value"
}
}
tatoeba.jpn
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/tatoeba.jpn')
- Keterangan :
his data is extracted from the Tatoeba corpus, dated Saturday 2018/11/17.
For each languages, we have selected 1000 English sentences and their translations, if available. Please check
this paper for a description of the languages, their families and scripts as well as baseline results.
Please note that the English sentences are not identical for all language pairs. This means that the results are
not directly comparable across languages. In particular, the sentences tend to have less variety for several
low-resource languages, e.g. "Tom needed water", "Tom needs water", "Tom is getting water", ...
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'validation' | 1000 |
- Fitur :
{
"source_sentence": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"target_sentence": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"source_lang": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"target_lang": {
"dtype": "string",
"id": null,
"_type": "Value"
}
}
tatoeba.kat
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/tatoeba.kat')
- Keterangan :
his data is extracted from the Tatoeba corpus, dated Saturday 2018/11/17.
For each languages, we have selected 1000 English sentences and their translations, if available. Please check
this paper for a description of the languages, their families and scripts as well as baseline results.
Please note that the English sentences are not identical for all language pairs. This means that the results are
not directly comparable across languages. In particular, the sentences tend to have less variety for several
low-resource languages, e.g. "Tom needed water", "Tom needs water", "Tom is getting water", ...
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'validation' | 746 |
- Fitur :
{
"source_sentence": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"target_sentence": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"source_lang": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"target_lang": {
"dtype": "string",
"id": null,
"_type": "Value"
}
}
tatoeba.kaz
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/tatoeba.kaz')
- Keterangan :
his data is extracted from the Tatoeba corpus, dated Saturday 2018/11/17.
For each languages, we have selected 1000 English sentences and their translations, if available. Please check
this paper for a description of the languages, their families and scripts as well as baseline results.
Please note that the English sentences are not identical for all language pairs. This means that the results are
not directly comparable across languages. In particular, the sentences tend to have less variety for several
low-resource languages, e.g. "Tom needed water", "Tom needs water", "Tom is getting water", ...
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'validation' | 575 |
- Fitur :
{
"source_sentence": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"target_sentence": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"source_lang": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"target_lang": {
"dtype": "string",
"id": null,
"_type": "Value"
}
}
tatoeba.kor
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/tatoeba.kor')
- Keterangan :
his data is extracted from the Tatoeba corpus, dated Saturday 2018/11/17.
For each languages, we have selected 1000 English sentences and their translations, if available. Please check
this paper for a description of the languages, their families and scripts as well as baseline results.
Please note that the English sentences are not identical for all language pairs. This means that the results are
not directly comparable across languages. In particular, the sentences tend to have less variety for several
low-resource languages, e.g. "Tom needed water", "Tom needs water", "Tom is getting water", ...
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'validation' | 1000 |
- Fitur :
{
"source_sentence": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"target_sentence": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"source_lang": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"target_lang": {
"dtype": "string",
"id": null,
"_type": "Value"
}
}
tatoeba.mal
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/tatoeba.mal')
- Keterangan :
his data is extracted from the Tatoeba corpus, dated Saturday 2018/11/17.
For each languages, we have selected 1000 English sentences and their translations, if available. Please check
this paper for a description of the languages, their families and scripts as well as baseline results.
Please note that the English sentences are not identical for all language pairs. This means that the results are
not directly comparable across languages. In particular, the sentences tend to have less variety for several
low-resource languages, e.g. "Tom needed water", "Tom needs water", "Tom is getting water", ...
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'validation' | 687 |
- Fitur :
{
"source_sentence": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"target_sentence": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"source_lang": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"target_lang": {
"dtype": "string",
"id": null,
"_type": "Value"
}
}
tatoeba.mar
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/tatoeba.mar')
- Keterangan :
his data is extracted from the Tatoeba corpus, dated Saturday 2018/11/17.
For each languages, we have selected 1000 English sentences and their translations, if available. Please check
this paper for a description of the languages, their families and scripts as well as baseline results.
Please note that the English sentences are not identical for all language pairs. This means that the results are
not directly comparable across languages. In particular, the sentences tend to have less variety for several
low-resource languages, e.g. "Tom needed water", "Tom needs water", "Tom is getting water", ...
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'validation' | 1000 |
- Fitur :
{
"source_sentence": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"target_sentence": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"source_lang": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"target_lang": {
"dtype": "string",
"id": null,
"_type": "Value"
}
}
tatoeba.nld
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/tatoeba.nld')
- Keterangan :
his data is extracted from the Tatoeba corpus, dated Saturday 2018/11/17.
For each languages, we have selected 1000 English sentences and their translations, if available. Please check
this paper for a description of the languages, their families and scripts as well as baseline results.
Please note that the English sentences are not identical for all language pairs. This means that the results are
not directly comparable across languages. In particular, the sentences tend to have less variety for several
low-resource languages, e.g. "Tom needed water", "Tom needs water", "Tom is getting water", ...
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'validation' | 1000 |
- Fitur :
{
"source_sentence": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"target_sentence": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"source_lang": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"target_lang": {
"dtype": "string",
"id": null,
"_type": "Value"
}
}
tatoeba.pes
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/tatoeba.pes')
- Keterangan :
his data is extracted from the Tatoeba corpus, dated Saturday 2018/11/17.
For each languages, we have selected 1000 English sentences and their translations, if available. Please check
this paper for a description of the languages, their families and scripts as well as baseline results.
Please note that the English sentences are not identical for all language pairs. This means that the results are
not directly comparable across languages. In particular, the sentences tend to have less variety for several
low-resource languages, e.g. "Tom needed water", "Tom needs water", "Tom is getting water", ...
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'validation' | 1000 |
- Fitur :
{
"source_sentence": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"target_sentence": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"source_lang": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"target_lang": {
"dtype": "string",
"id": null,
"_type": "Value"
}
}
tatoeba.por
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/tatoeba.por')
- Keterangan :
his data is extracted from the Tatoeba corpus, dated Saturday 2018/11/17.
For each languages, we have selected 1000 English sentences and their translations, if available. Please check
this paper for a description of the languages, their families and scripts as well as baseline results.
Please note that the English sentences are not identical for all language pairs. This means that the results are
not directly comparable across languages. In particular, the sentences tend to have less variety for several
low-resource languages, e.g. "Tom needed water", "Tom needs water", "Tom is getting water", ...
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'validation' | 1000 |
- Fitur :
{
"source_sentence": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"target_sentence": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"source_lang": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"target_lang": {
"dtype": "string",
"id": null,
"_type": "Value"
}
}
tatoeba.rus
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/tatoeba.rus')
- Keterangan :
his data is extracted from the Tatoeba corpus, dated Saturday 2018/11/17.
For each languages, we have selected 1000 English sentences and their translations, if available. Please check
this paper for a description of the languages, their families and scripts as well as baseline results.
Please note that the English sentences are not identical for all language pairs. This means that the results are
not directly comparable across languages. In particular, the sentences tend to have less variety for several
low-resource languages, e.g. "Tom needed water", "Tom needs water", "Tom is getting water", ...
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'validation' | 1000 |
- Fitur :
{
"source_sentence": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"target_sentence": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"source_lang": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"target_lang": {
"dtype": "string",
"id": null,
"_type": "Value"
}
}
tatoeba.spa
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/tatoeba.spa')
- Keterangan :
his data is extracted from the Tatoeba corpus, dated Saturday 2018/11/17.
For each languages, we have selected 1000 English sentences and their translations, if available. Please check
this paper for a description of the languages, their families and scripts as well as baseline results.
Please note that the English sentences are not identical for all language pairs. This means that the results are
not directly comparable across languages. In particular, the sentences tend to have less variety for several
low-resource languages, e.g. "Tom needed water", "Tom needs water", "Tom is getting water", ...
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'validation' | 1000 |
- Fitur :
{
"source_sentence": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"target_sentence": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"source_lang": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"target_lang": {
"dtype": "string",
"id": null,
"_type": "Value"
}
}
tatoeba.swh
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/tatoeba.swh')
- Keterangan :
his data is extracted from the Tatoeba corpus, dated Saturday 2018/11/17.
For each languages, we have selected 1000 English sentences and their translations, if available. Please check
this paper for a description of the languages, their families and scripts as well as baseline results.
Please note that the English sentences are not identical for all language pairs. This means that the results are
not directly comparable across languages. In particular, the sentences tend to have less variety for several
low-resource languages, e.g. "Tom needed water", "Tom needs water", "Tom is getting water", ...
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'validation' | 390 |
- Fitur :
{
"source_sentence": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"target_sentence": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"source_lang": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"target_lang": {
"dtype": "string",
"id": null,
"_type": "Value"
}
}
tatoeba.tam
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/tatoeba.tam')
- Keterangan :
his data is extracted from the Tatoeba corpus, dated Saturday 2018/11/17.
For each languages, we have selected 1000 English sentences and their translations, if available. Please check
this paper for a description of the languages, their families and scripts as well as baseline results.
Please note that the English sentences are not identical for all language pairs. This means that the results are
not directly comparable across languages. In particular, the sentences tend to have less variety for several
low-resource languages, e.g. "Tom needed water", "Tom needs water", "Tom is getting water", ...
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'validation' | 307 |
- Fitur :
{
"source_sentence": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"target_sentence": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"source_lang": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"target_lang": {
"dtype": "string",
"id": null,
"_type": "Value"
}
}
tatoeba.tel
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/tatoeba.tel')
- Keterangan :
his data is extracted from the Tatoeba corpus, dated Saturday 2018/11/17.
For each languages, we have selected 1000 English sentences and their translations, if available. Please check
this paper for a description of the languages, their families and scripts as well as baseline results.
Please note that the English sentences are not identical for all language pairs. This means that the results are
not directly comparable across languages. In particular, the sentences tend to have less variety for several
low-resource languages, e.g. "Tom needed water", "Tom needs water", "Tom is getting water", ...
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'validation' | 234 |
- Fitur :
{
"source_sentence": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"target_sentence": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"source_lang": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"target_lang": {
"dtype": "string",
"id": null,
"_type": "Value"
}
}
tatoeba.tgl
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/tatoeba.tgl')
- Keterangan :
his data is extracted from the Tatoeba corpus, dated Saturday 2018/11/17.
For each languages, we have selected 1000 English sentences and their translations, if available. Please check
this paper for a description of the languages, their families and scripts as well as baseline results.
Please note that the English sentences are not identical for all language pairs. This means that the results are
not directly comparable across languages. In particular, the sentences tend to have less variety for several
low-resource languages, e.g. "Tom needed water", "Tom needs water", "Tom is getting water", ...
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'validation' | 1000 |
- Fitur :
{
"source_sentence": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"target_sentence": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"source_lang": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"target_lang": {
"dtype": "string",
"id": null,
"_type": "Value"
}
}
tatoeba.tha
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/tatoeba.tha')
- Keterangan :
his data is extracted from the Tatoeba corpus, dated Saturday 2018/11/17.
For each languages, we have selected 1000 English sentences and their translations, if available. Please check
this paper for a description of the languages, their families and scripts as well as baseline results.
Please note that the English sentences are not identical for all language pairs. This means that the results are
not directly comparable across languages. In particular, the sentences tend to have less variety for several
low-resource languages, e.g. "Tom needed water", "Tom needs water", "Tom is getting water", ...
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'validation' | 548 |
- Fitur :
{
"source_sentence": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"target_sentence": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"source_lang": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"target_lang": {
"dtype": "string",
"id": null,
"_type": "Value"
}
}
tatoeba.tur
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/tatoeba.tur')
- Keterangan :
his data is extracted from the Tatoeba corpus, dated Saturday 2018/11/17.
For each languages, we have selected 1000 English sentences and their translations, if available. Please check
this paper for a description of the languages, their families and scripts as well as baseline results.
Please note that the English sentences are not identical for all language pairs. This means that the results are
not directly comparable across languages. In particular, the sentences tend to have less variety for several
low-resource languages, e.g. "Tom needed water", "Tom needs water", "Tom is getting water", ...
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'validation' | 1000 |
- Fitur :
{
"source_sentence": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"target_sentence": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"source_lang": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"target_lang": {
"dtype": "string",
"id": null,
"_type": "Value"
}
}
tatoeba.urd
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/tatoeba.urd')
- Keterangan :
his data is extracted from the Tatoeba corpus, dated Saturday 2018/11/17.
For each languages, we have selected 1000 English sentences and their translations, if available. Please check
this paper for a description of the languages, their families and scripts as well as baseline results.
Please note that the English sentences are not identical for all language pairs. This means that the results are
not directly comparable across languages. In particular, the sentences tend to have less variety for several
low-resource languages, e.g. "Tom needed water", "Tom needs water", "Tom is getting water", ...
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'validation' | 1000 |
- Fitur :
{
"source_sentence": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"target_sentence": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"source_lang": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"target_lang": {
"dtype": "string",
"id": null,
"_type": "Value"
}
}
tatoeba.vie
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/tatoeba.vie')
- Keterangan :
his data is extracted from the Tatoeba corpus, dated Saturday 2018/11/17.
For each languages, we have selected 1000 English sentences and their translations, if available. Please check
this paper for a description of the languages, their families and scripts as well as baseline results.
Please note that the English sentences are not identical for all language pairs. This means that the results are
not directly comparable across languages. In particular, the sentences tend to have less variety for several
low-resource languages, e.g. "Tom needed water", "Tom needs water", "Tom is getting water", ...
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'validation' | 1000 |
- Fitur :
{
"source_sentence": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"target_sentence": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"source_lang": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"target_lang": {
"dtype": "string",
"id": null,
"_type": "Value"
}
}
udpos.Orang Afrika
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/udpos.Afrikaans')
- Keterangan :
Universal Dependencies (UD) is a framework for consistent annotation of grammar (parts of speech, morphological
features, and syntactic dependencies) across different human languages. UD is an open community effort with over 200
contributors producing more than 100 treebanks in over 70 languages. If you’re new to UD, you should start by reading
the first part of the Short Introduction and then browsing the annotation guidelines.
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'test' | 425 |
'train' | 1315 |
'validation' | 194 |
- Fitur :
{
"tokens": {
"feature": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"length": -1,
"id": null,
"_type": "Sequence"
},
"pos_tags": {
"feature": {
"num_classes": 17,
"names": [
"ADJ",
"ADP",
"ADV",
"AUX",
"CCONJ",
"DET",
"INTJ",
"NOUN",
"NUM",
"PART",
"PRON",
"PROPN",
"PUNCT",
"SCONJ",
"SYM",
"VERB",
"X"
],
"names_file": null,
"id": null,
"_type": "ClassLabel"
},
"length": -1,
"id": null,
"_type": "Sequence"
}
}
udpos.Arab
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/udpos.Arabic')
- Keterangan :
Universal Dependencies (UD) is a framework for consistent annotation of grammar (parts of speech, morphological
features, and syntactic dependencies) across different human languages. UD is an open community effort with over 200
contributors producing more than 100 treebanks in over 70 languages. If you’re new to UD, you should start by reading
the first part of the Short Introduction and then browsing the annotation guidelines.
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'test' | 1680 |
'train' | 6075 |
'validation' | 909 |
- Fitur :
{
"tokens": {
"feature": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"length": -1,
"id": null,
"_type": "Sequence"
},
"pos_tags": {
"feature": {
"num_classes": 17,
"names": [
"ADJ",
"ADP",
"ADV",
"AUX",
"CCONJ",
"DET",
"INTJ",
"NOUN",
"NUM",
"PART",
"PRON",
"PROPN",
"PUNCT",
"SCONJ",
"SYM",
"VERB",
"X"
],
"names_file": null,
"id": null,
"_type": "ClassLabel"
},
"length": -1,
"id": null,
"_type": "Sequence"
}
}
udpos.Basque
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/udpos.Basque')
- Keterangan :
Universal Dependencies (UD) is a framework for consistent annotation of grammar (parts of speech, morphological
features, and syntactic dependencies) across different human languages. UD is an open community effort with over 200
contributors producing more than 100 treebanks in over 70 languages. If you’re new to UD, you should start by reading
the first part of the Short Introduction and then browsing the annotation guidelines.
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'test' | 1799 |
'train' | 5396 |
'validation' | 1798 |
- Fitur :
{
"tokens": {
"feature": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"length": -1,
"id": null,
"_type": "Sequence"
},
"pos_tags": {
"feature": {
"num_classes": 17,
"names": [
"ADJ",
"ADP",
"ADV",
"AUX",
"CCONJ",
"DET",
"INTJ",
"NOUN",
"NUM",
"PART",
"PRON",
"PROPN",
"PUNCT",
"SCONJ",
"SYM",
"VERB",
"X"
],
"names_file": null,
"id": null,
"_type": "ClassLabel"
},
"length": -1,
"id": null,
"_type": "Sequence"
}
}
udpos.Bulgaria
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/udpos.Bulgarian')
- Keterangan :
Universal Dependencies (UD) is a framework for consistent annotation of grammar (parts of speech, morphological
features, and syntactic dependencies) across different human languages. UD is an open community effort with over 200
contributors producing more than 100 treebanks in over 70 languages. If you’re new to UD, you should start by reading
the first part of the Short Introduction and then browsing the annotation guidelines.
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'test' | 1116 |
'train' | 8907 |
'validation' | 1115 |
- Fitur :
{
"tokens": {
"feature": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"length": -1,
"id": null,
"_type": "Sequence"
},
"pos_tags": {
"feature": {
"num_classes": 17,
"names": [
"ADJ",
"ADP",
"ADV",
"AUX",
"CCONJ",
"DET",
"INTJ",
"NOUN",
"NUM",
"PART",
"PRON",
"PROPN",
"PUNCT",
"SCONJ",
"SYM",
"VERB",
"X"
],
"names_file": null,
"id": null,
"_type": "ClassLabel"
},
"length": -1,
"id": null,
"_type": "Sequence"
}
}
udpos.Belanda
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/udpos.Dutch')
- Keterangan :
Universal Dependencies (UD) is a framework for consistent annotation of grammar (parts of speech, morphological
features, and syntactic dependencies) across different human languages. UD is an open community effort with over 200
contributors producing more than 100 treebanks in over 70 languages. If you’re new to UD, you should start by reading
the first part of the Short Introduction and then browsing the annotation guidelines.
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'test' | 1471 |
'train' | 18051 |
'validation' | 1394 |
- Fitur :
{
"tokens": {
"feature": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"length": -1,
"id": null,
"_type": "Sequence"
},
"pos_tags": {
"feature": {
"num_classes": 17,
"names": [
"ADJ",
"ADP",
"ADV",
"AUX",
"CCONJ",
"DET",
"INTJ",
"NOUN",
"NUM",
"PART",
"PRON",
"PROPN",
"PUNCT",
"SCONJ",
"SYM",
"VERB",
"X"
],
"names_file": null,
"id": null,
"_type": "ClassLabel"
},
"length": -1,
"id": null,
"_type": "Sequence"
}
}
udpos.Bahasa Inggris
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/udpos.English')
- Keterangan :
Universal Dependencies (UD) is a framework for consistent annotation of grammar (parts of speech, morphological
features, and syntactic dependencies) across different human languages. UD is an open community effort with over 200
contributors producing more than 100 treebanks in over 70 languages. If you’re new to UD, you should start by reading
the first part of the Short Introduction and then browsing the annotation guidelines.
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'test' | 5440 |
'train' | 21253 |
'validation' | 3974 |
- Fitur :
{
"tokens": {
"feature": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"length": -1,
"id": null,
"_type": "Sequence"
},
"pos_tags": {
"feature": {
"num_classes": 17,
"names": [
"ADJ",
"ADP",
"ADV",
"AUX",
"CCONJ",
"DET",
"INTJ",
"NOUN",
"NUM",
"PART",
"PRON",
"PROPN",
"PUNCT",
"SCONJ",
"SYM",
"VERB",
"X"
],
"names_file": null,
"id": null,
"_type": "ClassLabel"
},
"length": -1,
"id": null,
"_type": "Sequence"
}
}
udpos.Estonia
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/udpos.Estonian')
- Keterangan :
Universal Dependencies (UD) is a framework for consistent annotation of grammar (parts of speech, morphological
features, and syntactic dependencies) across different human languages. UD is an open community effort with over 200
contributors producing more than 100 treebanks in over 70 languages. If you’re new to UD, you should start by reading
the first part of the Short Introduction and then browsing the annotation guidelines.
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'test' | 3760 |
'train' | 25749 |
'validation' | 3125 |
- Fitur :
{
"tokens": {
"feature": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"length": -1,
"id": null,
"_type": "Sequence"
},
"pos_tags": {
"feature": {
"num_classes": 17,
"names": [
"ADJ",
"ADP",
"ADV",
"AUX",
"CCONJ",
"DET",
"INTJ",
"NOUN",
"NUM",
"PART",
"PRON",
"PROPN",
"PUNCT",
"SCONJ",
"SYM",
"VERB",
"X"
],
"names_file": null,
"id": null,
"_type": "ClassLabel"
},
"length": -1,
"id": null,
"_type": "Sequence"
}
}
udpos.Finlandia
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/udpos.Finnish')
- Keterangan :
Universal Dependencies (UD) is a framework for consistent annotation of grammar (parts of speech, morphological
features, and syntactic dependencies) across different human languages. UD is an open community effort with over 200
contributors producing more than 100 treebanks in over 70 languages. If you’re new to UD, you should start by reading
the first part of the Short Introduction and then browsing the annotation guidelines.
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'test' | 4422 |
'train' | 27198 |
'validation' | 3239 |
- Fitur :
{
"tokens": {
"feature": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"length": -1,
"id": null,
"_type": "Sequence"
},
"pos_tags": {
"feature": {
"num_classes": 17,
"names": [
"ADJ",
"ADP",
"ADV",
"AUX",
"CCONJ",
"DET",
"INTJ",
"NOUN",
"NUM",
"PART",
"PRON",
"PROPN",
"PUNCT",
"SCONJ",
"SYM",
"VERB",
"X"
],
"names_file": null,
"id": null,
"_type": "ClassLabel"
},
"length": -1,
"id": null,
"_type": "Sequence"
}
}
udpos.Perancis
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/udpos.French')
- Keterangan :
Universal Dependencies (UD) is a framework for consistent annotation of grammar (parts of speech, morphological
features, and syntactic dependencies) across different human languages. UD is an open community effort with over 200
contributors producing more than 100 treebanks in over 70 languages. If you’re new to UD, you should start by reading
the first part of the Short Introduction and then browsing the annotation guidelines.
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'test' | 9465 |
'train' | 47308 |
'validation' | 5979 |
- Fitur :
{
"tokens": {
"feature": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"length": -1,
"id": null,
"_type": "Sequence"
},
"pos_tags": {
"feature": {
"num_classes": 17,
"names": [
"ADJ",
"ADP",
"ADV",
"AUX",
"CCONJ",
"DET",
"INTJ",
"NOUN",
"NUM",
"PART",
"PRON",
"PROPN",
"PUNCT",
"SCONJ",
"SYM",
"VERB",
"X"
],
"names_file": null,
"id": null,
"_type": "ClassLabel"
},
"length": -1,
"id": null,
"_type": "Sequence"
}
}
udpos.Jerman
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/udpos.German')
- Keterangan :
Universal Dependencies (UD) is a framework for consistent annotation of grammar (parts of speech, morphological
features, and syntactic dependencies) across different human languages. UD is an open community effort with over 200
contributors producing more than 100 treebanks in over 70 languages. If you’re new to UD, you should start by reading
the first part of the Short Introduction and then browsing the annotation guidelines.
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'test' | 22458 |
'train' | 166849 |
'validation' | 19233 |
- Fitur :
{
"tokens": {
"feature": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"length": -1,
"id": null,
"_type": "Sequence"
},
"pos_tags": {
"feature": {
"num_classes": 17,
"names": [
"ADJ",
"ADP",
"ADV",
"AUX",
"CCONJ",
"DET",
"INTJ",
"NOUN",
"NUM",
"PART",
"PRON",
"PROPN",
"PUNCT",
"SCONJ",
"SYM",
"VERB",
"X"
],
"names_file": null,
"id": null,
"_type": "ClassLabel"
},
"length": -1,
"id": null,
"_type": "Sequence"
}
}
udpos.Yunani
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/udpos.Greek')
- Keterangan :
Universal Dependencies (UD) is a framework for consistent annotation of grammar (parts of speech, morphological
features, and syntactic dependencies) across different human languages. UD is an open community effort with over 200
contributors producing more than 100 treebanks in over 70 languages. If you’re new to UD, you should start by reading
the first part of the Short Introduction and then browsing the annotation guidelines.
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'test' | 2809 |
'train' | 28152 |
'validation' | 2559 |
- Fitur :
{
"tokens": {
"feature": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"length": -1,
"id": null,
"_type": "Sequence"
},
"pos_tags": {
"feature": {
"num_classes": 17,
"names": [
"ADJ",
"ADP",
"ADV",
"AUX",
"CCONJ",
"DET",
"INTJ",
"NOUN",
"NUM",
"PART",
"PRON",
"PROPN",
"PUNCT",
"SCONJ",
"SYM",
"VERB",
"X"
],
"names_file": null,
"id": null,
"_type": "ClassLabel"
},
"length": -1,
"id": null,
"_type": "Sequence"
}
}
udpos.Ibrani
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/udpos.Hebrew')
- Keterangan :
Universal Dependencies (UD) is a framework for consistent annotation of grammar (parts of speech, morphological
features, and syntactic dependencies) across different human languages. UD is an open community effort with over 200
contributors producing more than 100 treebanks in over 70 languages. If you’re new to UD, you should start by reading
the first part of the Short Introduction and then browsing the annotation guidelines.
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'test' | 491 |
'train' | 5241 |
'validation' | 484 |
- Fitur :
{
"tokens": {
"feature": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"length": -1,
"id": null,
"_type": "Sequence"
},
"pos_tags": {
"feature": {
"num_classes": 17,
"names": [
"ADJ",
"ADP",
"ADV",
"AUX",
"CCONJ",
"DET",
"INTJ",
"NOUN",
"NUM",
"PART",
"PRON",
"PROPN",
"PUNCT",
"SCONJ",
"SYM",
"VERB",
"X"
],
"names_file": null,
"id": null,
"_type": "ClassLabel"
},
"length": -1,
"id": null,
"_type": "Sequence"
}
}
udpos.Hindi
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/udpos.Hindi')
- Keterangan :
Universal Dependencies (UD) is a framework for consistent annotation of grammar (parts of speech, morphological
features, and syntactic dependencies) across different human languages. UD is an open community effort with over 200
contributors producing more than 100 treebanks in over 70 languages. If you’re new to UD, you should start by reading
the first part of the Short Introduction and then browsing the annotation guidelines.
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'test' | 2684 |
'train' | 13304 |
'validation' | 1659 |
- Fitur :
{
"tokens": {
"feature": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"length": -1,
"id": null,
"_type": "Sequence"
},
"pos_tags": {
"feature": {
"num_classes": 17,
"names": [
"ADJ",
"ADP",
"ADV",
"AUX",
"CCONJ",
"DET",
"INTJ",
"NOUN",
"NUM",
"PART",
"PRON",
"PROPN",
"PUNCT",
"SCONJ",
"SYM",
"VERB",
"X"
],
"names_file": null,
"id": null,
"_type": "ClassLabel"
},
"length": -1,
"id": null,
"_type": "Sequence"
}
}
udpos.Hongaria
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/udpos.Hungarian')
- Keterangan :
Universal Dependencies (UD) is a framework for consistent annotation of grammar (parts of speech, morphological
features, and syntactic dependencies) across different human languages. UD is an open community effort with over 200
contributors producing more than 100 treebanks in over 70 languages. If you’re new to UD, you should start by reading
the first part of the Short Introduction and then browsing the annotation guidelines.
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'test' | 449 |
'train' | 910 |
'validation' | 441 |
- Fitur :
{
"tokens": {
"feature": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"length": -1,
"id": null,
"_type": "Sequence"
},
"pos_tags": {
"feature": {
"num_classes": 17,
"names": [
"ADJ",
"ADP",
"ADV",
"AUX",
"CCONJ",
"DET",
"INTJ",
"NOUN",
"NUM",
"PART",
"PRON",
"PROPN",
"PUNCT",
"SCONJ",
"SYM",
"VERB",
"X"
],
"names_file": null,
"id": null,
"_type": "ClassLabel"
},
"length": -1,
"id": null,
"_type": "Sequence"
}
}
udpos.Bahasa Indonesia
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/udpos.Indonesian')
- Keterangan :
Universal Dependencies (UD) is a framework for consistent annotation of grammar (parts of speech, morphological
features, and syntactic dependencies) across different human languages. UD is an open community effort with over 200
contributors producing more than 100 treebanks in over 70 languages. If you’re new to UD, you should start by reading
the first part of the Short Introduction and then browsing the annotation guidelines.
The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark is a benchmark for the evaluation of
the cross-lingual generalization ability of pre-trained multilingual models. It covers 40 typologically diverse languages
(spanning 12 language families) and includes nine tasks that collectively require reasoning about different levels of
syntax and semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks,
and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil
(spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the
Niger-Congo languages Swahili and Yoruba, spoken in Africa.
- Lisensi : Tidak ada lisensi yang diketahui
- Versi : 1.0.0
- Perpecahan :
Membelah | Contoh |
---|---|
'test' | 1557 |
'train' | 4477 |
'validation' | 559 |
- Fitur :
{
"tokens": {
"feature": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"length": -1,
"id": null,
"_type": "Sequence"
},
"pos_tags": {
"feature": {
"num_classes": 17,
"names": [
"ADJ",
"ADP",
"ADV",
"AUX",
"CCONJ",
"DET",
"INTJ",
"NOUN",
"NUM",
"PART",
"PRON",
"PROPN",
"PUNCT",
"SCONJ",
"SYM",
"VERB",
"X"
],
"names_file": null,
"id": null,
"_type": "ClassLabel"
},
"length": -1,
"id": null,
"_type": "Sequence"
}
}
udpos.Italia
Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:
ds = tfds.load('huggingface:xtreme/udpos.Italian')
- Keterangan :
Universal Dependencies (UD) is a framework for consistent annotation of grammar (parts of speech, morphological
features, and syntactic dependencies) across different human languages. UD is an open community effort with over 200
contributors producing more than 100 treebanks in over 70 languages.