Japanese multi-task CNN trained on UD-Japanese BCCWJ r2.8 + GSK2014-A(2019) + transformers-ud-japanese-electra–base. Components: transformer, parser, atteribute_ruler, ner, morphologizer, compound_splitter, bunsetu_recognizer.
Feature | Description |
— | — |
Name | ja_ginza_bert_large |
Version | 5.3.0 |
spaCy | >=3.8.16,<4.0.0 |
Default Pipeline | transformer, parser, ner, morphologizer, compound_splitter, bunsetu_recognizer |
Components | transformer, parser, attribute_ruler, ner, morphologizer, compound_splitter, bunsetu_recognizer |
Vectors | 0 keys, 0 unique vectors (0 dimensions) |
Sources | [UD_Japanese-BCCWJ r2.8](https://github.com/UniversalDependencies/UD_Japanese-BCCWJ) (Asahara, M., Kanayama, H., Tanaka, T., Miyao, Y., Uematsu, S., Mori, S., Matsumoto, Y., Omura, M., & Murawaki, Y.)<br>[GSK2014-A(2019)](https://www.gsk.or.jp/catalog/gsk2014-a/) (Tokyo Institute of Technology)<br>[SudachiDict_core](https://github.com/WorksApplications/SudachiDict) (Works Applications Enterprise Co., Ltd.)<br>[cl-tohoku/bert-large-japanese-v2](https://huggingface.co/cl-tohoku/bert-large-japanese-v2) (Tohoku University) |
License | MIT License |
Author | [Megagon Labs Tokyo.](https://github.com/megagonlabs/ginza) |
### Label Scheme
<details>
<summary>View label scheme (241 labels for 3 components)</summary>
Component | Labels |
— | — |
`parser` | ROOT, acl, acl_bunsetu, advcl, advcl_bunsetu, advmod, advmod_bunsetu, amod, amod_bunsetu, aux, aux_bunsetu, case, case_bunsetu, cc, cc_bunsetu, ccomp_bunsetu, compound, compound_bunsetu, cop, csubj_bunsetu, dep, dep_bunsetu, det_bunsetu, discourse_bunsetu, dislocated_bunsetu, fixed, mark, nmod, nmod_bunsetu, nsubj_bunsetu, nummod, obj_bunsetu, obl_bunsetu, punct, punct_bunsetu |
`ner` | Academic, Age, Aircraft, Airport, Amphibia, Amusement_Park, Animal_Disease, Animal_Part, Archaeological_Place_Other, Art_Other, Astral_Body_Other, Award, Bay, Bird, Book, Bridge, Broadcast_Program, Cabinet, Calorie, Canal, Car, Car_Stop, Character, City, Class, Clothing, Color_Other, Company, Company_Group, Compound, Conference, Constellation, Continental_Region, Corporation_Other, Country, Countx_Other, County, Culture, Date, Day_Of_Week, Disease_Other, Dish, Doctrine_Method_Other, Domestic_Region, Drug, Earthquake, Element, Email, Era, Ethnic_Group_Other, Event_Other, Facility_Other, Facility_Part, Family, Fish, Flora, Flora_Part, Food_Other, Frequency, Fungus, GOE_Other, GPE_Other, Game, Geological_Region_Other, God, Government, ID_Number, Incident_Other, Insect, Intensity, International_Organization, Island, Lake, Language_Other, Latitude_Longtitude, Law, Line_Other, Living_Thing_Other, Living_Thing_Part_Other, Location_Other, Magazine, Mammal, Material, Measurement_Other, Military, Mineral, Mollusc_Arthropod, Money, Money_Form, Mountain, Movement, Movie, Multiplication, Museum, Music, N_Animal, N_Country, N_Event, N_Facility, N_Flora, N_Location_Other, N_Natural_Object_Other, N_Organization, N_Person, N_Product, Name_Other, National_Language, Nationality, Natural_Disaster, Natural_Object_Other, Natural_Phenomenon_Other, Nature_Color, Newspaper, Numex_Other, Occasion_Other, Offense, Ordinal_Number, Organization_Other, Park, Percent, Period_Day, Period_Month, Period_Time, Period_Week, Period_Year, Periodx_Other, Person, Phone_Number, Physical_Extent, Plan, Planet, Point, Political_Organization_Other, Political_Party, Port, Position_Vocation, Postal_Address, Printing_Other, Pro_Sports_Organization, Product_Other, Province, Public_Institution, Railroad, Rank, Region_Other, Religion, Religious_Festival, Reptile, Research_Institute, River, Road, Rule_Other, School, School_Age, Sea, Ship, Show, Show_Organization, Spa, Space, Spaceship, Speed, Sport, Sports_Facility, Sports_League, Sports_Organization_Other, Station, Style, Temperature, Theater, Theory, Time, Time_Top_Other, Timex_Other, Title_Other, Train, Treaty, Tumulus, Tunnel, URL, Unit_Other, Vehicle_Other, Volume, War, Water_Route, Weapon, Weight, Worship_Place, Zoo |
`morphologizer` | POS=PUNCT, POS=NUM, POS=NOUN, POS=ADP, POS=AUX, POS=VERB, POS=CCONJ, POS=PART, POS=SCONJ, POS=SYM, POS=ADJ, POS=DET, POS=PRON, POS=PROPN, POS=ADV, POS=X, POS=INTJ |
</details>
### Accuracy
Type | Score |
— | — |
DEP_UAS | 94.48 |
DEP_LAS | 93.35 |
SENTS_P | 89.73 |
SENTS_R | 90.44 |
SENTS_F | 90.08 |
ENTS_F | 71.12 |
ENTS_P | 71.73 |
ENTS_R | 70.52 |
POS_ACC | 98.67 |
MORPH_MICRO_F | 0.00 |
MORPH_PER_FEAT | 0.00 |
TAG_ACC | 0.00 |
TRANSFORMER_LOSS | 86315.36 |
PARSER_LOSS | 946823.28 |
NER_LOSS | 3432.99 |
MORPHOLOGIZER_LOSS | 2551.99 |
1 maintainer
Metadata
1 maintainer
Release files for ja-ginza-bert-large 5.3.0
For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.
Built distribution (wheel)
| File | Interpreter | ABI | Platform | Actions Reset |
|---|---|---|---|---|
| ja_ginza_bert_large-5.3.0-py3-none-any.whl | Python 3 | none | any | Details for ja_ginza_bert_large-5.3.0-py3-none-any.whl |
Release files / ja_ginza_bert_large-5.3.0-py3-none-any.whl
| Download URL | ja_ginza_bert_large-5.3.0-py3-none-any.whl |
|---|---|
| Size | 2.9 MB |
| Tags | Python 3 |
|
SHA-256 checksum How to use checksums |
9114483c50613cce78be75499e994a3f50181867e89f7b67a5e8439c70e33268
|
|
BLAKE2b-256 checksum How to use checksums |
fbf4c77d6720c5e0037dd75c4dc1585442f49994a7b90cc6b638677c0c276451
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
twine/7.0.0 CPython/3.13.15
|
Security for ja-ginza-bert-large 5.3.0
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
No provenance found
None of the files in this release were uploaded with provenance metadata.
A note on package security
PyPI is a repository for distributing packages, not a service for auditing them. While third-party security researchers and our staff do their best to detect and remove malicious packages, PyPI cannot guarantee the quality, security or suitability of any given package.
To learn more about protecting yourself from malicious actors, visit our package security documentation.
