|
--- |
|
license: mit |
|
thumbnail: https://huggingface.co/front/thumbnails/facebook.png |
|
inference: false |
|
--- |
|
|
|
# XGLM-564M |
|
|
|
XGLM-564M is a multilingual autoregressive language model (with 564 million parameters) trained on a balanced corpus of a diverse set of 30 languages totaling 500 billion sub-tokens. It was introduced in the paper [Few-shot Learning with Multilingual Language Models](https://arxiv.org/abs/2112.10668) by Xi Victoria Lin\*, Todor Mihaylov, Mikel Artetxe, Tianlu Wang, Shuohui Chen, Daniel Simig, Myle Ott, Naman Goyal, Shruti Bhosale, Jingfei Du, Ramakanth Pasunuru, Sam Shleifer, Punit Singh Koura, Vishrav Chaudhary, Brian O'Horo, Jeff Wang, Luke Zettlemoyer, Zornitsa Kozareva, Mona Diab, Veselin Stoyanov, Xian Li\* (\*Equal Contribution). The original implementation was released in [this repository](https://github.com/pytorch/fairseq/tree/main/examples/xglm). |
|
|
|
## Training Data Statistics |
|
|
|
The training data statistics of XGLM-564M is shown in the table below. |
|
|
|
| ISO-639-1| family | name | # tokens | ratio | ratio w/ lowRes upsampling | |
|
|:--------|:-----------------|:------------------------|-------------:|------------:|-------------:| |
|
| en | Indo-European | English | 803526736124 | 0.489906 | 0.3259 | |
|
| ru | Indo-European | Russian | 147791898098 | 0.0901079 | 0.0602 | |
|
| zh | Sino-Tibetan | Chinese | 132770494630 | 0.0809494 | 0.0483 | |
|
| de | Indo-European | German | 89223707856 | 0.0543992 | 0.0363 | |
|
| es | Indo-European | Spanish | 87303083105 | 0.0532282 | 0.0353 | |
|
| fr | Indo-European | French | 77419639775 | 0.0472023 | 0.0313 | |
|
| ja | Japonic | Japanese | 66054364513 | 0.040273 | 0.0269 | |
|
| it | Indo-European | Italian | 41930465338 | 0.0255648 | 0.0171 | |
|
| pt | Indo-European | Portuguese | 36586032444 | 0.0223063 | 0.0297 | |
|
| el | Indo-European | Greek (modern) | 28762166159 | 0.0175361 | 0.0233 | |
|
| ko | Koreanic | Korean | 20002244535 | 0.0121953 | 0.0811 | |
|
| fi | Uralic | Finnish | 16804309722 | 0.0102455 | 0.0681 | |
|
| id | Austronesian | Indonesian | 15423541953 | 0.00940365 | 0.0125 | |
|
| tr | Turkic | Turkish | 12413166065 | 0.00756824 | 0.0101 | |
|
| ar | Afro-Asiatic | Arabic | 12248607345 | 0.00746791 | 0.0099 | |
|
| vi | Austroasiatic | Vietnamese | 11199121869 | 0.00682804 | 0.0091 | |
|
| th | Tai–Kadai | Thai | 10842172807 | 0.00661041 | 0.044 | |
|
| bg | Indo-European | Bulgarian | 9703797869 | 0.00591635 | 0.0393 | |
|
| ca | Indo-European | Catalan | 7075834775 | 0.0043141 | 0.0287 | |
|
| hi | Indo-European | Hindi | 3448390110 | 0.00210246 | 0.014 | |
|
| et | Uralic | Estonian | 3286873851 | 0.00200399 | 0.0133 | |
|
| bn | Indo-European | Bengali, Bangla | 1627447450 | 0.000992245 | 0.0066 | |
|
| ta | Dravidian | Tamil | 1476973397 | 0.000900502 | 0.006 | |
|
| ur | Indo-European | Urdu | 1351891969 | 0.000824241 | 0.0055 | |
|
| sw | Niger–Congo | Swahili | 907516139 | 0.000553307 | 0.0037 | |
|
| te | Dravidian | Telugu | 689316485 | 0.000420272 | 0.0028 | |
|
| eu | Language isolate | Basque | 105304423 | 6.42035e-05 | 0.0043 | |
|
| my | Sino-Tibetan | Burmese | 101358331 | 6.17976e-05 | 0.003 | |
|
| ht | Creole | Haitian, Haitian Creole | 86584697 | 5.27902e-05 | 0.0035 | |
|
| qu | Quechuan | Quechua | 3236108 | 1.97304e-06 | 0.0001 | |
|
|
|
## Model card |
|
|
|
For intended usage of the model, please refer to the [model card](https://github.com/pytorch/fairseq/blob/main/examples/xglm/model_card.md) released by the XGLM-564M development team. |