daviibrt's picture
Update spaCy pipeline
7a5dc31 verified
raw
history blame
No virus
4.83 kB
{
"lang":"en",
"name":"ner_bionlp13cg_md",
"version":"0.5.3",
"description":"Spacy Models for Biomedical Text.",
"author":"Allen Institute for Artificial Intelligence",
"email":"ai2-info@allenai.org",
"url":"https://allenai.github.io/SciSpaCy/",
"license":"CC BY-SA 3.0",
"spacy_version":">=3.6.1,<3.7.0",
"spacy_git_version":"458bc5f45",
"vectors":{
"width":200,
"vectors":50000,
"keys":4087446,
"name":"en_core_sci_md.vectors"
},
"labels":{
"tok2vec":[
],
"tagger":[
"$",
"''",
",",
"-LRB-",
"-RRB-",
".",
":",
"ADD",
"AFX",
"CC",
"CD",
"DT",
"EX",
"FW",
"HYPH",
"IN",
"JJ",
"JJR",
"JJS",
"LS",
"MD",
"NFP",
"NN",
"NNP",
"NNPS",
"NNS",
"PDT",
"POS",
"PRP",
"PRP$",
"RB",
"RBR",
"RBS",
"RP",
"SYM",
"TO",
"UH",
"VB",
"VBD",
"VBG",
"VBN",
"VBP",
"VBZ",
"WDT",
"WP",
"WP$",
"WRB",
"XX",
"``"
],
"attribute_ruler":[
],
"lemmatizer":[
],
"parser":[
"ROOT",
"acl",
"acl:relcl",
"acomp",
"advcl",
"advmod",
"amod",
"amod@nmod",
"appos",
"attr",
"aux",
"auxpass",
"case",
"cc",
"cc:preconj",
"ccomp",
"compound",
"compound:prt",
"conj",
"cop",
"csubj",
"dative",
"dep",
"det",
"det:predet",
"dobj",
"expl",
"intj",
"mark",
"meta",
"mwe",
"neg",
"nmod",
"nmod:npmod",
"nmod:poss",
"nmod:tmod",
"nsubj",
"nsubjpass",
"nummod",
"parataxis",
"pcomp",
"pobj",
"preconj",
"predet",
"prep",
"punct",
"quantmod",
"xcomp"
],
"ner":[
"AMINO_ACID",
"ANATOMICAL_SYSTEM",
"CANCER",
"CELL",
"CELLULAR_COMPONENT",
"DEVELOPING_ANATOMICAL_STRUCTURE",
"GENE_OR_GENE_PRODUCT",
"IMMATERIAL_ANATOMICAL_ENTITY",
"MULTI_TISSUE_STRUCTURE",
"ORGAN",
"ORGANISM",
"ORGANISM_SUBDIVISION",
"ORGANISM_SUBSTANCE",
"PATHOLOGICAL_FORMATION",
"SIMPLE_CHEMICAL",
"TISSUE"
]
},
"pipeline":[
"tok2vec",
"tagger",
"attribute_ruler",
"lemmatizer",
"parser",
"ner"
],
"components":[
"tok2vec",
"tagger",
"attribute_ruler",
"lemmatizer",
"parser",
"ner"
],
"disabled":[
],
"performance":{
"tag_acc":0.0,
"lemma_acc":0.0,
"dep_uas":0.0,
"dep_las":0.0,
"dep_las_per_type":0.0,
"sents_p":0.0,
"sents_r":0.0,
"sents_f":0.0,
"ents_f":0.7808239261,
"ents_p":0.7980221756,
"ents_r":0.7643513203,
"ents_per_type":{
"SIMPLE_CHEMICAL":{
"p":0.7573333333,
"r":0.6439909297,
"f":0.6960784314
},
"ORGANISM":{
"p":0.8175675676,
"r":0.8432055749,
"f":0.8301886792
},
"MULTI_TISSUE_STRUCTURE":{
"p":0.6422764228,
"r":0.5895522388,
"f":0.6147859922
},
"CANCER":{
"p":0.7788235294,
"r":0.8193069307,
"f":0.7985524729
},
"GENE_OR_GENE_PRODUCT":{
"p":0.8547750592,
"r":0.8155120482,
"f":0.8346820809
},
"CELL":{
"p":0.7831325301,
"r":0.7707509881,
"f":0.7768924303
},
"ORGANISM_SUBSTANCE":{
"p":0.7619047619,
"r":0.4848484848,
"f":0.5925925926
},
"ORGANISM_SUBDIVISION":{
"p":0.75,
"r":0.5,
"f":0.6
},
"TISSUE":{
"p":0.6543209877,
"r":0.6708860759,
"f":0.6625
},
"ORGAN":{
"p":0.8088235294,
"r":0.8088235294,
"f":0.8088235294
},
"CELLULAR_COMPONENT":{
"p":0.7244897959,
"r":0.7802197802,
"f":0.7513227513
},
"IMMATERIAL_ANATOMICAL_ENTITY":{
"p":0.9090909091,
"r":0.5555555556,
"f":0.6896551724
},
"AMINO_ACID":{
"p":0.6315789474,
"r":0.7741935484,
"f":0.6956521739
},
"PATHOLOGICAL_FORMATION":{
"p":0.652173913,
"r":0.3409090909,
"f":0.447761194
},
"ANATOMICAL_SYSTEM":{
"p":1.0,
"r":0.6666666667,
"f":0.8
},
"DEVELOPING_ANATOMICAL_STRUCTURE":{
"p":0.6666666667,
"r":0.4,
"f":0.5
}
},
"ner_loss":5887.003366243
},
"sources":[
"BIONLP13CG",
"OntoNotes 5",
"Common Crawl",
"GENIA 1.0"
],
"requirements":[
]
}