mk_core_news_sm / meta.json
osanseviero's picture
Update spaCy pipeline
5b81c91
raw
history blame
7.35 kB
{
"lang":"mk",
"name":"core_news_sm",
"version":"3.2.0",
"description":"Macedonian pipeline optimized for CPU. Components: tok2vec, morphologizer, parser, senter, ner, attribute_ruler, lemmatizer.",
"author":"Explosion",
"email":"contact@explosion.ai",
"url":"https://explosion.ai",
"license":"CC BY-SA 4.0",
"spacy_version":">=3.2.0,<3.3.0",
"spacy_git_version":"bb26550e2",
"vectors":{
"width":0,
"vectors":0,
"keys":0,
"name":null
},
"labels":{
"morphologizer":[
"POS=PROPN",
"POS=AUX",
"POS=ADJ",
"POS=NOUN",
"POS=ADP",
"POS=PUNCT",
"POS=CONJ",
"POS=NUM",
"POS=VERB",
"POS=PRON",
"POS=ADV",
"POS=SCONJ",
"POS=PART",
"POS=SYM",
"POS=X",
"_",
"POS=INTJ"
],
"parser":[
"ROOT",
"advmod",
"att",
"aux",
"cc",
"dep",
"det",
"dobj",
"iobj",
"neg",
"nsubj",
"pobj",
"poss",
"pozm",
"pozv",
"prep",
"punct",
"relcl"
],
"senter":[
"I",
"S"
],
"attribute_ruler":[
],
"lemmatizer":[
],
"ner":[
"CARDINAL",
"DATE",
"EVENT",
"FAC",
"GPE",
"LANGUAGE",
"LAW",
"LOC",
"MONEY",
"NORP",
"ORDINAL",
"ORG",
"PERCENT",
"PERSON",
"PRODUCT",
"QUANTITY",
"TIME",
"WORK_OF_ART"
]
},
"pipeline":[
"morphologizer",
"parser",
"attribute_ruler",
"lemmatizer",
"ner"
],
"components":[
"morphologizer",
"parser",
"senter",
"attribute_ruler",
"lemmatizer",
"ner"
],
"disabled":[
"senter"
],
"performance":{
"token_acc":1.0,
"token_p":1.0,
"token_r":1.0,
"token_f":1.0,
"sents_p":0.6710526316,
"sents_r":0.6623376623,
"sents_f":0.6666666667,
"ents_p":0.7299912049,
"ents_r":0.7063829787,
"ents_f":0.7179930796,
"ents_per_type":{
"ORG":{
"p":0.5636363636,
"r":0.6326530612,
"f":0.5961538462
},
"GPE":{
"p":0.8344671202,
"r":0.8679245283,
"f":0.850867052
},
"LOC":{
"p":0.7142857143,
"r":0.4597701149,
"f":0.5594405594
},
"QUANTITY":{
"p":0.6444444444,
"r":0.7073170732,
"f":0.6744186047
},
"CARDINAL":{
"p":0.7142857143,
"r":0.6914893617,
"f":0.7027027027
},
"DATE":{
"p":0.7162162162,
"r":0.7464788732,
"f":0.7310344828
},
"NORP":{
"p":0.4905660377,
"r":0.4,
"f":0.4406779661
},
"PERSON":{
"p":0.7333333333,
"r":0.7333333333,
"f":0.7333333333
},
"ORDINAL":{
"p":0.5,
"r":0.3636363636,
"f":0.4210526316
},
"MONEY":{
"p":0.5,
"r":0.5,
"f":0.5
},
"PERCENT":{
"p":0.8421052632,
"r":1.0,
"f":0.9142857143
},
"WORK_OF_ART":{
"p":0.5333333333,
"r":0.3902439024,
"f":0.4507042254
},
"LANGUAGE":{
"p":0.0,
"r":0.0,
"f":0.0
},
"TIME":{
"p":1.0,
"r":0.6666666667,
"f":0.8
},
"FAC":{
"p":0.25,
"r":0.1,
"f":0.1428571429
},
"EVENT":{
"p":0.5238095238,
"r":0.6470588235,
"f":0.5789473684
},
"LAW":{
"p":0.25,
"r":0.3333333333,
"f":0.2857142857
},
"PRODUCT":{
"p":0.0,
"r":0.0,
"f":0.0
}
},
"speed":2245.3508517398,
"pos_acc":0.9163744268,
"dep_uas":0.6307541626,
"dep_las":0.4760039177,
"dep_las_per_type":{
"nsubj":{
"p":0.475,
"r":0.5,
"f":0.4871794872
},
"root":{
"p":0.6184210526,
"r":0.6714285714,
"f":0.6438356164
},
"cc":{
"p":0.7368421053,
"r":0.5,
"f":0.5957446809
},
"relcl":{
"p":0.4,
"r":0.3076923077,
"f":0.347826087
},
"pozm":{
"p":1.0,
"r":0.3636363636,
"f":0.5333333333
},
"poss":{
"p":0.25,
"r":0.2,
"f":0.2222222222
},
"aux":{
"p":0.5526315789,
"r":0.6363636364,
"f":0.5915492958
},
"prep":{
"p":0.6984126984,
"r":0.7333333333,
"f":0.7154471545
},
"iobj":{
"p":0.1,
"r":0.125,
"f":0.1111111111
},
"pozv":{
"p":0.2857142857,
"r":0.1333333333,
"f":0.1818181818
},
"quantmod":{
"p":0.0,
"r":0.0,
"f":0.0
},
"att":{
"p":0.7179487179,
"r":0.5384615385,
"f":0.6153846154
},
"det":{
"p":0.0,
"r":0.0,
"f":0.0
},
"num":{
"p":0.0,
"r":0.0,
"f":0.0
},
"dep":{
"p":0.0151515152,
"r":0.3333333333,
"f":0.0289855072
},
"dobj":{
"p":0.393442623,
"r":0.4,
"f":0.3966942149
},
"ppdo":{
"p":0.8,
"r":0.5333333333,
"f":0.64
},
"neg":{
"p":0.6,
"r":0.5454545455,
"f":0.5714285714
},
"pobj":{
"p":0.375,
"r":0.46875,
"f":0.4166666667
},
"mwe":{
"p":0.0,
"r":0.0,
"f":0.0
},
"ppio":{
"p":0.0,
"r":0.0,
"f":0.0
},
"appos":{
"p":0.0,
"r":0.0,
"f":0.0
},
"advmod":{
"p":0.0,
"r":0.0,
"f":0.0
},
"advcl":{
"p":0.0,
"r":0.0,
"f":0.0
},
"number":{
"p":0.0,
"r":0.0,
"f":0.0
},
"amod":{
"p":0.0,
"r":0.0,
"f":0.0
},
"_":{
"p":0.0,
"r":0.0,
"f":0.0
},
"acl":{
"p":0.0,
"r":0.0,
"f":0.0
},
"pozn":{
"p":0.0,
"r":0.0,
"f":0.0
},
"pozk":{
"p":0.0,
"r":0.0,
"f":0.0
}
}
},
"sources":[
{
"name":"Macedonian Corpus",
"url":"https://blog.netcetera.com/macedonian-spacy-f3c85484777f",
"license":"CC BY-SA 4.0",
"author":"Damjan Zlatinov, Melanija Gerasimovska, Borijan Georgievski, Marija Todosovska"
},
{
"name":"Macedonian Corpus",
"url":"https://blog.netcetera.com/macedonian-spacy-f3c85484777f",
"license":"CC BY-SA 4.0",
"author":"Damjan Zlatinov, Melanija Gerasimovska, Borijan Georgievski, Marija Todosovska"
},
{
"name":"Macedonian Corpus",
"url":"https://blog.netcetera.com/macedonian-spacy-f3c85484777f",
"license":"CC BY-SA 4.0",
"author":"Damjan Zlatinov, Melanija Gerasimovska, Borijan Georgievski, Marija Todosovska"
},
{
"name":"spaCy lookups data",
"author":"Explosion",
"url":"https://github.com/explosion/spacy-lookups-data",
"license":"MIT"
}
],
"requirements":[
]
}