KoichiYasuoka
commited on
Commit
•
0d6db2b
1
Parent(s):
90dec1d
initial release
Browse files- README.md +76 -0
- config.json +1074 -0
- maker.py +55 -0
- pytorch_model.bin +3 -0
- sentencepiece.model +3 -0
- special_tokens_map.json +15 -0
- tokenizer.json +0 -0
- tokenizer_config.json +21 -0
- ud.py +63 -0
README.md
ADDED
@@ -0,0 +1,76 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
---
|
2 |
+
language:
|
3 |
+
- "cop"
|
4 |
+
tags:
|
5 |
+
- "coptic"
|
6 |
+
- "token-classification"
|
7 |
+
- "pos"
|
8 |
+
- "dependency-parsing"
|
9 |
+
datasets:
|
10 |
+
- "universal_dependencies"
|
11 |
+
license: "cc-by-sa-4.0"
|
12 |
+
pipeline_tag: "token-classification"
|
13 |
+
widget:
|
14 |
+
- text: "ⲧⲉⲛⲟⲩⲇⲉⲛ̄ⲟⲩⲟⲉⲓⲛϩ︤ⲙ︥ⲡϫⲟⲉⲓⲥ·"
|
15 |
+
---
|
16 |
+
|
17 |
+
# roberta-base-coptic-ud-goeswith
|
18 |
+
|
19 |
+
## Model Description
|
20 |
+
|
21 |
+
This is a RoBERTa model pre-trained on Coptic Scriptorium Corpora for POS-tagging and dependency-parsing (using `goeswith` for subwords), derived from [roberta-base-coptic](https://huggingface.co/KoichiYasuoka/roberta-base-coptic).
|
22 |
+
|
23 |
+
## How to Use
|
24 |
+
|
25 |
+
```py
|
26 |
+
class UDgoeswith(object):
|
27 |
+
def __init__(self,bert):
|
28 |
+
from transformers import AutoTokenizer,AutoModelForTokenClassification
|
29 |
+
self.tokenizer=AutoTokenizer.from_pretrained(bert)
|
30 |
+
self.model=AutoModelForTokenClassification.from_pretrained(bert)
|
31 |
+
def __call__(self,text):
|
32 |
+
import numpy,torch,ufal.chu_liu_edmonds
|
33 |
+
w=self.tokenizer(text,return_offsets_mapping=True)
|
34 |
+
v=w["input_ids"]
|
35 |
+
x=[v[0:i]+[self.tokenizer.mask_token_id]+v[i+1:]+[j] for i,j in enumerate(v[1:-1],1)]
|
36 |
+
with torch.no_grad():
|
37 |
+
e=self.model(input_ids=torch.tensor(x)).logits.numpy()[:,1:-2,:]
|
38 |
+
r=[1 if i==0 else -1 if j.endswith("|root") else 0 for i,j in sorted(self.model.config.id2label.items())]
|
39 |
+
e+=numpy.where(numpy.add.outer(numpy.identity(e.shape[0]),r)==0,0,numpy.nan)
|
40 |
+
g=self.model.config.label2id["X|_|goeswith"]
|
41 |
+
r=numpy.tri(e.shape[0])
|
42 |
+
for i in range(e.shape[0]):
|
43 |
+
for j in range(i+2,e.shape[1]):
|
44 |
+
r[i,j]=r[i,j-1] if numpy.nanargmax(e[i,j-1])==g else 1
|
45 |
+
e[:,:,g]+=numpy.where(r==0,0,numpy.nan)
|
46 |
+
m=numpy.full((e.shape[0]+1,e.shape[1]+1),numpy.nan)
|
47 |
+
m[1:,1:]=numpy.nanmax(e,axis=2).transpose()
|
48 |
+
p=numpy.zeros(m.shape)
|
49 |
+
p[1:,1:]=numpy.nanargmax(e,axis=2).transpose()
|
50 |
+
for i in range(1,m.shape[0]):
|
51 |
+
m[i,0],m[i,i],p[i,0]=m[i,i],numpy.nan,p[i,i]
|
52 |
+
h=ufal.chu_liu_edmonds.chu_liu_edmonds(m)[0]
|
53 |
+
if [0 for i in h if i==0]!=[0]:
|
54 |
+
m[:,0]+=numpy.where(m[:,0]==numpy.nanmax(m[[i for i,j in enumerate(h) if j==0],0]),0,numpy.nan)
|
55 |
+
m[[i for i,j in enumerate(h) if j==0]]+=[0 if i==0 or j==0 else numpy.nan for i,j in enumerate(h)]
|
56 |
+
h=ufal.chu_liu_edmonds.chu_liu_edmonds(m)[0]
|
57 |
+
u="# text = "+text+"\n"
|
58 |
+
v=[(s,e) for s,e in w["offset_mapping"] if s<e]
|
59 |
+
for i,(s,e) in enumerate(v,1):
|
60 |
+
q=self.model.config.id2label[p[i,h[i]]].split("|")
|
61 |
+
u+="\t".join([str(i),text[s:e],"_",q[0],"_","|".join(q[1:-1]),str(h[i]),q[-1],"_","_" if i<len(v) and e<v[i][0] else "SpaceAfter=No"])+"\n"
|
62 |
+
return u+"\n"
|
63 |
+
|
64 |
+
nlp=UDgoeswith("KoichiYasuoka/roberta-base-coptic-ud-goeswith")
|
65 |
+
print(nlp("ⲧⲉⲛⲟⲩⲇⲉⲛ̄ⲟⲩⲟⲉⲓⲛϩ︤ⲙ︥ⲡϫⲟⲉⲓⲥ·"))
|
66 |
+
```
|
67 |
+
|
68 |
+
with [ufal.chu-liu-edmonds](https://pypi.org/project/ufal.chu-liu-edmonds/).
|
69 |
+
Or without ufal.chu-liu-edmonds:
|
70 |
+
|
71 |
+
```
|
72 |
+
from transformers import pipeline
|
73 |
+
nlp=pipeline("universal-dependencies","KoichiYasuoka/roberta-base-coptic-ud-goeswith",trust_remote_code=True,aggregation_strategy="simple")
|
74 |
+
print(nlp("ⲧⲉⲛⲟⲩⲇⲉⲛ̄ⲟⲩⲟⲉⲓⲛϩ︤ⲙ︥ⲡϫⲟⲉⲓⲥ·"))
|
75 |
+
```
|
76 |
+
|
config.json
ADDED
@@ -0,0 +1,1074 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"architectures": [
|
3 |
+
"RobertaForTokenClassification"
|
4 |
+
],
|
5 |
+
"attention_probs_dropout_prob": 0.1,
|
6 |
+
"bos_token_id": 0,
|
7 |
+
"classifier_dropout": null,
|
8 |
+
"custom_pipelines": {
|
9 |
+
"universal-dependencies": {
|
10 |
+
"impl": "ud.UniversalDependenciesPipeline"
|
11 |
+
}
|
12 |
+
},
|
13 |
+
"eos_token_id": 2,
|
14 |
+
"hidden_act": "gelu",
|
15 |
+
"hidden_dropout_prob": 0.1,
|
16 |
+
"hidden_size": 768,
|
17 |
+
"id2label": {
|
18 |
+
"0": "-|_|dep",
|
19 |
+
"1": "ADJ|Foreign=Yes|amod",
|
20 |
+
"2": "ADJ|_|amod",
|
21 |
+
"3": "ADP|Foreign=Yes|case",
|
22 |
+
"4": "ADP|_|advmod",
|
23 |
+
"5": "ADP|_|appos",
|
24 |
+
"6": "ADP|_|case",
|
25 |
+
"7": "ADP|_|cc",
|
26 |
+
"8": "ADP|_|fixed",
|
27 |
+
"9": "ADP|_|nmod",
|
28 |
+
"10": "ADP|_|nsubj",
|
29 |
+
"11": "ADP|_|obj",
|
30 |
+
"12": "ADP|_|obl",
|
31 |
+
"13": "ADP|_|reparandum",
|
32 |
+
"14": "ADV|Definite=Def|PronType=Prs|Reflex=Yes|advmod",
|
33 |
+
"15": "ADV|Foreign=Yes|Polarity=Neg|advmod",
|
34 |
+
"16": "ADV|Foreign=Yes|advmod",
|
35 |
+
"17": "ADV|Foreign=Yes|mark",
|
36 |
+
"18": "ADV|Polarity=Neg|advcl",
|
37 |
+
"19": "ADV|Polarity=Neg|advmod",
|
38 |
+
"20": "ADV|PronType=Int|advmod",
|
39 |
+
"21": "ADV|PronType=Int|ccomp",
|
40 |
+
"22": "ADV|PronType=Int|conj",
|
41 |
+
"23": "ADV|PronType=Int|nmod",
|
42 |
+
"24": "ADV|PronType=Int|parataxis",
|
43 |
+
"25": "ADV|PronType=Int|root",
|
44 |
+
"26": "ADV|PronType=Tot|advmod",
|
45 |
+
"27": "ADV|_|acl:relcl",
|
46 |
+
"28": "ADV|_|advcl",
|
47 |
+
"29": "ADV|_|advmod",
|
48 |
+
"30": "ADV|_|case",
|
49 |
+
"31": "ADV|_|conj",
|
50 |
+
"32": "ADV|_|fixed",
|
51 |
+
"33": "ADV|_|mark",
|
52 |
+
"34": "ADV|_|nmod",
|
53 |
+
"35": "ADV|_|obl",
|
54 |
+
"36": "ADV|_|obl:npmod",
|
55 |
+
"37": "ADV|_|root",
|
56 |
+
"38": "ADV|_|xcomp",
|
57 |
+
"39": "AUX|Foreign=Yes|aux",
|
58 |
+
"40": "AUX|Polarity=Neg|aux",
|
59 |
+
"41": "AUX|VerbForm=Fin|aux",
|
60 |
+
"42": "AUX|_|aux",
|
61 |
+
"43": "AUX|_|orphan",
|
62 |
+
"44": "AUX|_|reparandum",
|
63 |
+
"45": "CCONJ|Foreign=Yes|Polarity=Neg|advmod",
|
64 |
+
"46": "CCONJ|Foreign=Yes|Polarity=Neg|cc",
|
65 |
+
"47": "CCONJ|Foreign=Yes|advmod",
|
66 |
+
"48": "CCONJ|Foreign=Yes|cc",
|
67 |
+
"49": "CCONJ|Foreign=Yes|discourse",
|
68 |
+
"50": "CCONJ|Foreign=Yes|fixed",
|
69 |
+
"51": "CCONJ|_|advmod",
|
70 |
+
"52": "CCONJ|_|case",
|
71 |
+
"53": "CCONJ|_|cc",
|
72 |
+
"54": "CCONJ|_|ccomp",
|
73 |
+
"55": "CCONJ|_|discourse",
|
74 |
+
"56": "CCONJ|_|fixed",
|
75 |
+
"57": "CCONJ|_|reparandum",
|
76 |
+
"58": "DET|Definite=Def|Foreign=Yes|Number=Plur|PronType=Dem|root",
|
77 |
+
"59": "DET|Definite=Def|Gender=Fem|Gender[psor]=Fem|Number=Sing|Number[psor]=Sing|Person=2|Poss=Yes|PronType=Prs|det",
|
78 |
+
"60": "DET|Definite=Def|Gender=Fem|Gender[psor]=Fem|Number=Sing|Number[psor]=Sing|Person=3|Poss=Yes|PronType=Prs|det",
|
79 |
+
"61": "DET|Definite=Def|Gender=Fem|Gender[psor]=Masc|Number=Sing|Number[psor]=Sing|Person=2|Poss=Yes|PronType=Prs|det",
|
80 |
+
"62": "DET|Definite=Def|Gender=Fem|Gender[psor]=Masc|Number=Sing|Number[psor]=Sing|Person=3|Poss=Yes|PronType=Prs|det",
|
81 |
+
"63": "DET|Definite=Def|Gender=Fem|Gender[psor]=Masc|Number=Sing|Number[psor]=Sing|Person=3|Poss=Yes|PronType=Prs|obl",
|
82 |
+
"64": "DET|Definite=Def|Gender=Fem|Number=Sing|Number[psor]=Plur|Person=1|Poss=Yes|PronType=Prs|det",
|
83 |
+
"65": "DET|Definite=Def|Gender=Fem|Number=Sing|Number[psor]=Plur|Person=2|Poss=Yes|PronType=Prs|det",
|
84 |
+
"66": "DET|Definite=Def|Gender=Fem|Number=Sing|Number[psor]=Plur|Person=3|Poss=Yes|PronType=Prs|det",
|
85 |
+
"67": "DET|Definite=Def|Gender=Fem|Number=Sing|Number[psor]=Sing|Person=1|Poss=Yes|PronType=Prs|det",
|
86 |
+
"68": "DET|Definite=Def|Gender=Fem|Number=Sing|Poss=Yes|PronType=Prs|acl:relcl",
|
87 |
+
"69": "DET|Definite=Def|Gender=Fem|Number=Sing|Poss=Yes|PronType=Prs|ccomp",
|
88 |
+
"70": "DET|Definite=Def|Gender=Fem|Number=Sing|Poss=Yes|PronType=Prs|conj",
|
89 |
+
"71": "DET|Definite=Def|Gender=Fem|Number=Sing|Poss=Yes|PronType=Prs|nmod",
|
90 |
+
"72": "DET|Definite=Def|Gender=Fem|Number=Sing|PronType=Art|appos",
|
91 |
+
"73": "DET|Definite=Def|Gender=Fem|Number=Sing|PronType=Art|ccomp",
|
92 |
+
"74": "DET|Definite=Def|Gender=Fem|Number=Sing|PronType=Art|conj",
|
93 |
+
"75": "DET|Definite=Def|Gender=Fem|Number=Sing|PronType=Art|det",
|
94 |
+
"76": "DET|Definite=Def|Gender=Fem|Number=Sing|PronType=Art|dislocated",
|
95 |
+
"77": "DET|Definite=Def|Gender=Fem|Number=Sing|PronType=Art|obj",
|
96 |
+
"78": "DET|Definite=Def|Gender=Fem|Number=Sing|PronType=Art|obl",
|
97 |
+
"79": "DET|Definite=Def|Gender=Fem|Number=Sing|PronType=Dem|acl:relcl",
|
98 |
+
"80": "DET|Definite=Def|Gender=Fem|Number=Sing|PronType=Dem|appos",
|
99 |
+
"81": "DET|Definite=Def|Gender=Fem|Number=Sing|PronType=Dem|ccomp",
|
100 |
+
"82": "DET|Definite=Def|Gender=Fem|Number=Sing|PronType=Dem|det",
|
101 |
+
"83": "DET|Definite=Def|Gender=Fem|Number=Sing|PronType=Dem|dislocated",
|
102 |
+
"84": "DET|Definite=Def|Gender=Fem|Number=Sing|PronType=Dem|nsubj",
|
103 |
+
"85": "DET|Definite=Def|Gender=Fem|Number=Sing|PronType=Dem|obl",
|
104 |
+
"86": "DET|Definite=Def|Gender=Fem|Number=Sing|PronType=Dem|parataxis",
|
105 |
+
"87": "DET|Definite=Def|Gender=Fem|Number=Sing|PronType=Dem|root",
|
106 |
+
"88": "DET|Definite=Def|Gender=Masc|Gender[psor]=Fem|Number=Sing|Number[psor]=Sing|Person=2|Poss=Yes|PronType=Prs|det",
|
107 |
+
"89": "DET|Definite=Def|Gender=Masc|Gender[psor]=Fem|Number=Sing|Number[psor]=Sing|Person=3|Poss=Yes|PronType=Prs|det",
|
108 |
+
"90": "DET|Definite=Def|Gender=Masc|Gender[psor]=Masc|Number=Sing|Number[psor]=Sing|Person=2|Poss=Yes|PronType=Prs|acl:relcl",
|
109 |
+
"91": "DET|Definite=Def|Gender=Masc|Gender[psor]=Masc|Number=Sing|Number[psor]=Sing|Person=2|Poss=Yes|PronType=Prs|appos",
|
110 |
+
"92": "DET|Definite=Def|Gender=Masc|Gender[psor]=Masc|Number=Sing|Number[psor]=Sing|Person=2|Poss=Yes|PronType=Prs|det",
|
111 |
+
"93": "DET|Definite=Def|Gender=Masc|Gender[psor]=Masc|Number=Sing|Number[psor]=Sing|Person=2|Poss=Yes|PronType=Prs|obj",
|
112 |
+
"94": "DET|Definite=Def|Gender=Masc|Gender[psor]=Masc|Number=Sing|Number[psor]=Sing|Person=3|Poss=Yes|PronType=Prs|acl:relcl",
|
113 |
+
"95": "DET|Definite=Def|Gender=Masc|Gender[psor]=Masc|Number=Sing|Number[psor]=Sing|Person=3|Poss=Yes|PronType=Prs|det",
|
114 |
+
"96": "DET|Definite=Def|Gender=Masc|Number=Sing|Number[psor]=Plur|Person=1|Poss=Yes|PronType=Prs|det",
|
115 |
+
"97": "DET|Definite=Def|Gender=Masc|Number=Sing|Number[psor]=Plur|Person=2|Poss=Yes|PronType=Prs|conj",
|
116 |
+
"98": "DET|Definite=Def|Gender=Masc|Number=Sing|Number[psor]=Plur|Person=2|Poss=Yes|PronType=Prs|det",
|
117 |
+
"99": "DET|Definite=Def|Gender=Masc|Number=Sing|Number[psor]=Plur|Person=2|Poss=Yes|PronType=Prs|obl",
|
118 |
+
"100": "DET|Definite=Def|Gender=Masc|Number=Sing|Number[psor]=Plur|Person=2|Poss=Yes|PronType=Prs|parataxis",
|
119 |
+
"101": "DET|Definite=Def|Gender=Masc|Number=Sing|Number[psor]=Plur|Person=2|Poss=Yes|PronType=Prs|root",
|
120 |
+
"102": "DET|Definite=Def|Gender=Masc|Number=Sing|Number[psor]=Plur|Person=3|Poss=Yes|PronType=Prs|det",
|
121 |
+
"103": "DET|Definite=Def|Gender=Masc|Number=Sing|Number[psor]=Sing|Person=1|Poss=Yes|PronType=Prs|acl:relcl",
|
122 |
+
"104": "DET|Definite=Def|Gender=Masc|Number=Sing|Number[psor]=Sing|Person=1|Poss=Yes|PronType=Prs|ccomp",
|
123 |
+
"105": "DET|Definite=Def|Gender=Masc|Number=Sing|Number[psor]=Sing|Person=1|Poss=Yes|PronType=Prs|det",
|
124 |
+
"106": "DET|Definite=Def|Gender=Masc|Number=Sing|Poss=Yes|PronType=Prs|acl:relcl",
|
125 |
+
"107": "DET|Definite=Def|Gender=Masc|Number=Sing|Poss=Yes|PronType=Prs|appos",
|
126 |
+
"108": "DET|Definite=Def|Gender=Masc|Number=Sing|Poss=Yes|PronType=Prs|ccomp",
|
127 |
+
"109": "DET|Definite=Def|Gender=Masc|Number=Sing|Poss=Yes|PronType=Prs|conj",
|
128 |
+
"110": "DET|Definite=Def|Gender=Masc|Number=Sing|Poss=Yes|PronType=Prs|nmod",
|
129 |
+
"111": "DET|Definite=Def|Gender=Masc|Number=Sing|Poss=Yes|PronType=Prs|nsubj",
|
130 |
+
"112": "DET|Definite=Def|Gender=Masc|Number=Sing|Poss=Yes|PronType=Prs|obj",
|
131 |
+
"113": "DET|Definite=Def|Gender=Masc|Number=Sing|Poss=Yes|PronType=Prs|parataxis",
|
132 |
+
"114": "DET|Definite=Def|Gender=Masc|Number=Sing|Poss=Yes|PronType=Prs|root",
|
133 |
+
"115": "DET|Definite=Def|Gender=Masc|Number=Sing|Poss=Yes|PronType=Prs|vocative",
|
134 |
+
"116": "DET|Definite=Def|Gender=Masc|Number=Sing|PronType=Art|acl:relcl",
|
135 |
+
"117": "DET|Definite=Def|Gender=Masc|Number=Sing|PronType=Art|advcl",
|
136 |
+
"118": "DET|Definite=Def|Gender=Masc|Number=Sing|PronType=Art|appos",
|
137 |
+
"119": "DET|Definite=Def|Gender=Masc|Number=Sing|PronType=Art|ccomp",
|
138 |
+
"120": "DET|Definite=Def|Gender=Masc|Number=Sing|PronType=Art|conj",
|
139 |
+
"121": "DET|Definite=Def|Gender=Masc|Number=Sing|PronType=Art|det",
|
140 |
+
"122": "DET|Definite=Def|Gender=Masc|Number=Sing|PronType=Art|dislocated",
|
141 |
+
"123": "DET|Definite=Def|Gender=Masc|Number=Sing|PronType=Art|fixed",
|
142 |
+
"124": "DET|Definite=Def|Gender=Masc|Number=Sing|PronType=Art|nmod",
|
143 |
+
"125": "DET|Definite=Def|Gender=Masc|Number=Sing|PronType=Art|nsubj",
|
144 |
+
"126": "DET|Definite=Def|Gender=Masc|Number=Sing|PronType=Art|obj",
|
145 |
+
"127": "DET|Definite=Def|Gender=Masc|Number=Sing|PronType=Art|obl",
|
146 |
+
"128": "DET|Definite=Def|Gender=Masc|Number=Sing|PronType=Art|parataxis",
|
147 |
+
"129": "DET|Definite=Def|Gender=Masc|Number=Sing|PronType=Art|root",
|
148 |
+
"130": "DET|Definite=Def|Gender=Masc|Number=Sing|PronType=Art|vocative",
|
149 |
+
"131": "DET|Definite=Def|Gender=Masc|Number=Sing|PronType=Dem|acl:relcl",
|
150 |
+
"132": "DET|Definite=Def|Gender=Masc|Number=Sing|PronType=Dem|advcl",
|
151 |
+
"133": "DET|Definite=Def|Gender=Masc|Number=Sing|PronType=Dem|appos",
|
152 |
+
"134": "DET|Definite=Def|Gender=Masc|Number=Sing|PronType=Dem|ccomp",
|
153 |
+
"135": "DET|Definite=Def|Gender=Masc|Number=Sing|PronType=Dem|conj",
|
154 |
+
"136": "DET|Definite=Def|Gender=Masc|Number=Sing|PronType=Dem|det",
|
155 |
+
"137": "DET|Definite=Def|Gender=Masc|Number=Sing|PronType=Dem|dislocated",
|
156 |
+
"138": "DET|Definite=Def|Gender=Masc|Number=Sing|PronType=Dem|iobj",
|
157 |
+
"139": "DET|Definite=Def|Gender=Masc|Number=Sing|PronType=Dem|nmod",
|
158 |
+
"140": "DET|Definite=Def|Gender=Masc|Number=Sing|PronType=Dem|nsubj",
|
159 |
+
"141": "DET|Definite=Def|Gender=Masc|Number=Sing|PronType=Dem|obj",
|
160 |
+
"142": "DET|Definite=Def|Gender=Masc|Number=Sing|PronType=Dem|obl",
|
161 |
+
"143": "DET|Definite=Def|Gender=Masc|Number=Sing|PronType=Dem|parataxis",
|
162 |
+
"144": "DET|Definite=Def|Gender=Masc|Number=Sing|PronType=Dem|root",
|
163 |
+
"145": "DET|Definite=Def|Gender=Masc|Number=Sing|PronType=Dem|xcomp",
|
164 |
+
"146": "DET|Definite=Def|Gender[psor]=Fem|Number=Plur|Number[psor]=Sing|Person=2|Poss=Yes|PronType=Prs|det",
|
165 |
+
"147": "DET|Definite=Def|Gender[psor]=Fem|Number=Plur|Number[psor]=Sing|Person=3|Poss=Yes|PronType=Prs|det",
|
166 |
+
"148": "DET|Definite=Def|Gender[psor]=Masc|Number=Plur|Number[psor]=Sing|Person=2|Poss=Yes|PronType=Prs|acl:relcl",
|
167 |
+
"149": "DET|Definite=Def|Gender[psor]=Masc|Number=Plur|Number[psor]=Sing|Person=2|Poss=Yes|PronType=Prs|det",
|
168 |
+
"150": "DET|Definite=Def|Gender[psor]=Masc|Number=Plur|Number[psor]=Sing|Person=3|Poss=Yes|PronType=Prs|det",
|
169 |
+
"151": "DET|Definite=Def|Number=Plur|Number[psor]=Plur|Person=1|Poss=Yes|PronType=Prs|det",
|
170 |
+
"152": "DET|Definite=Def|Number=Plur|Number[psor]=Plur|Person=2|Poss=Yes|PronType=Prs|det",
|
171 |
+
"153": "DET|Definite=Def|Number=Plur|Number[psor]=Plur|Person=3|Poss=Yes|PronType=Prs|det",
|
172 |
+
"154": "DET|Definite=Def|Number=Plur|Number[psor]=Sing|Person=1|Poss=Yes|PronType=Prs|conj",
|
173 |
+
"155": "DET|Definite=Def|Number=Plur|Number[psor]=Sing|Person=1|Poss=Yes|PronType=Prs|det",
|
174 |
+
"156": "DET|Definite=Def|Number=Plur|Poss=Yes|PronType=Prs|advcl",
|
175 |
+
"157": "DET|Definite=Def|Number=Plur|Poss=Yes|PronType=Prs|ccomp",
|
176 |
+
"158": "DET|Definite=Def|Number=Plur|Poss=Yes|PronType=Prs|conj",
|
177 |
+
"159": "DET|Definite=Def|Number=Plur|Poss=Yes|PronType=Prs|nmod",
|
178 |
+
"160": "DET|Definite=Def|Number=Plur|Poss=Yes|PronType=Prs|obj",
|
179 |
+
"161": "DET|Definite=Def|Number=Plur|Poss=Yes|PronType=Prs|obl",
|
180 |
+
"162": "DET|Definite=Def|Number=Plur|Poss=Yes|PronType=Prs|root",
|
181 |
+
"163": "DET|Definite=Def|Number=Plur|PronType=Art|acl:relcl",
|
182 |
+
"164": "DET|Definite=Def|Number=Plur|PronType=Art|advcl",
|
183 |
+
"165": "DET|Definite=Def|Number=Plur|PronType=Art|appos",
|
184 |
+
"166": "DET|Definite=Def|Number=Plur|PronType=Art|ccomp",
|
185 |
+
"167": "DET|Definite=Def|Number=Plur|PronType=Art|conj",
|
186 |
+
"168": "DET|Definite=Def|Number=Plur|PronType=Art|det",
|
187 |
+
"169": "DET|Definite=Def|Number=Plur|PronType=Art|discourse",
|
188 |
+
"170": "DET|Definite=Def|Number=Plur|PronType=Art|dislocated",
|
189 |
+
"171": "DET|Definite=Def|Number=Plur|PronType=Art|nmod",
|
190 |
+
"172": "DET|Definite=Def|Number=Plur|PronType=Art|nsubj",
|
191 |
+
"173": "DET|Definite=Def|Number=Plur|PronType=Art|obj",
|
192 |
+
"174": "DET|Definite=Def|Number=Plur|PronType=Art|obl",
|
193 |
+
"175": "DET|Definite=Def|Number=Plur|PronType=Art|root",
|
194 |
+
"176": "DET|Definite=Def|Number=Plur|PronType=Art|vocative",
|
195 |
+
"177": "DET|Definite=Def|Number=Plur|PronType=Art|xcomp",
|
196 |
+
"178": "DET|Definite=Def|Number=Plur|PronType=Dem|acl:relcl",
|
197 |
+
"179": "DET|Definite=Def|Number=Plur|PronType=Dem|appos",
|
198 |
+
"180": "DET|Definite=Def|Number=Plur|PronType=Dem|ccomp",
|
199 |
+
"181": "DET|Definite=Def|Number=Plur|PronType=Dem|det",
|
200 |
+
"182": "DET|Definite=Def|Number=Plur|PronType=Dem|dislocated",
|
201 |
+
"183": "DET|Definite=Def|Number=Plur|PronType=Dem|nmod",
|
202 |
+
"184": "DET|Definite=Def|Number=Plur|PronType=Dem|nsubj",
|
203 |
+
"185": "DET|Definite=Def|Number=Plur|PronType=Dem|obj",
|
204 |
+
"186": "DET|Definite=Def|Number=Plur|PronType=Dem|obl",
|
205 |
+
"187": "DET|Definite=Def|Number=Plur|PronType=Dem|parataxis",
|
206 |
+
"188": "DET|Definite=Def|Number=Plur|PronType=Dem|root",
|
207 |
+
"189": "DET|Definite=Ind|Number=Sing|PronType=Art|det",
|
208 |
+
"190": "DET|Definite=Ind|Number=Sing|PronType=Art|obj",
|
209 |
+
"191": "DET|Definite=Ind|Number=Sing|PronType=Art|obl",
|
210 |
+
"192": "DET|PronType=Art|advmod",
|
211 |
+
"193": "DET|PronType=Art|det",
|
212 |
+
"194": "DET|PronType=Art|nsubj",
|
213 |
+
"195": "DET|PronType=Art|obj",
|
214 |
+
"196": "DET|_|compound",
|
215 |
+
"197": "DET|_|conj",
|
216 |
+
"198": "DET|_|det",
|
217 |
+
"199": "DET|_|nmod",
|
218 |
+
"200": "DET|_|obl",
|
219 |
+
"201": "DET|_|reparandum",
|
220 |
+
"202": "NOUN|Foreign=Yes|acl:relcl",
|
221 |
+
"203": "NOUN|Foreign=Yes|advcl",
|
222 |
+
"204": "NOUN|Foreign=Yes|appos",
|
223 |
+
"205": "NOUN|Foreign=Yes|ccomp",
|
224 |
+
"206": "NOUN|Foreign=Yes|conj",
|
225 |
+
"207": "NOUN|Foreign=Yes|dislocated",
|
226 |
+
"208": "NOUN|Foreign=Yes|fixed",
|
227 |
+
"209": "NOUN|Foreign=Yes|nmod",
|
228 |
+
"210": "NOUN|Foreign=Yes|nsubj",
|
229 |
+
"211": "NOUN|Foreign=Yes|obj",
|
230 |
+
"212": "NOUN|Foreign=Yes|obl",
|
231 |
+
"213": "NOUN|Foreign=Yes|obl:npmod",
|
232 |
+
"214": "NOUN|Foreign=Yes|orphan",
|
233 |
+
"215": "NOUN|Foreign=Yes|parataxis",
|
234 |
+
"216": "NOUN|Foreign=Yes|reparandum",
|
235 |
+
"217": "NOUN|Foreign=Yes|root",
|
236 |
+
"218": "NOUN|Foreign=Yes|vocative",
|
237 |
+
"219": "NOUN|Foreign=Yes|xcomp",
|
238 |
+
"220": "NOUN|PronType=Rcp|acl:relcl",
|
239 |
+
"221": "NOUN|PronType=Rcp|advcl",
|
240 |
+
"222": "NOUN|PronType=Rcp|nmod",
|
241 |
+
"223": "NOUN|PronType=Rcp|obj",
|
242 |
+
"224": "NOUN|PronType=Rcp|obl",
|
243 |
+
"225": "NOUN|_|acl:relcl",
|
244 |
+
"226": "NOUN|_|advcl",
|
245 |
+
"227": "NOUN|_|appos",
|
246 |
+
"228": "NOUN|_|case",
|
247 |
+
"229": "NOUN|_|ccomp",
|
248 |
+
"230": "NOUN|_|compound",
|
249 |
+
"231": "NOUN|_|conj",
|
250 |
+
"232": "NOUN|_|csubj",
|
251 |
+
"233": "NOUN|_|dep",
|
252 |
+
"234": "NOUN|_|discourse",
|
253 |
+
"235": "NOUN|_|dislocated",
|
254 |
+
"236": "NOUN|_|fixed",
|
255 |
+
"237": "NOUN|_|iobj",
|
256 |
+
"238": "NOUN|_|nmod",
|
257 |
+
"239": "NOUN|_|nsubj",
|
258 |
+
"240": "NOUN|_|obj",
|
259 |
+
"241": "NOUN|_|obl",
|
260 |
+
"242": "NOUN|_|obl:npmod",
|
261 |
+
"243": "NOUN|_|orphan",
|
262 |
+
"244": "NOUN|_|parataxis",
|
263 |
+
"245": "NOUN|_|reparandum",
|
264 |
+
"246": "NOUN|_|root",
|
265 |
+
"247": "NOUN|_|vocative",
|
266 |
+
"248": "NOUN|_|xcomp",
|
267 |
+
"249": "NUM|Foreign=Yes|NumType=Card|nsubj",
|
268 |
+
"250": "NUM|Foreign=Yes|NumType=Card|reparandum",
|
269 |
+
"251": "NUM|NumType=Card|acl:relcl",
|
270 |
+
"252": "NUM|NumType=Card|ccomp",
|
271 |
+
"253": "NUM|NumType=Card|compound",
|
272 |
+
"254": "NUM|NumType=Card|conj",
|
273 |
+
"255": "NUM|NumType=Card|dislocated",
|
274 |
+
"256": "NUM|NumType=Card|nmod",
|
275 |
+
"257": "NUM|NumType=Card|nsubj",
|
276 |
+
"258": "NUM|NumType=Card|nummod",
|
277 |
+
"259": "NUM|NumType=Card|obj",
|
278 |
+
"260": "NUM|NumType=Card|obl",
|
279 |
+
"261": "NUM|NumType=Card|obl:npmod",
|
280 |
+
"262": "NUM|NumType=Card|orphan",
|
281 |
+
"263": "NUM|NumType=Card|parataxis",
|
282 |
+
"264": "NUM|NumType=Card|root",
|
283 |
+
"265": "NUM|NumType=Card|xcomp",
|
284 |
+
"266": "PART|Foreign=Yes|Polarity=Neg|cc",
|
285 |
+
"267": "PART|Foreign=Yes|advcl",
|
286 |
+
"268": "PART|Foreign=Yes|advmod",
|
287 |
+
"269": "PART|Foreign=Yes|ccomp",
|
288 |
+
"270": "PART|Foreign=Yes|discourse",
|
289 |
+
"271": "PART|Foreign=Yes|fixed",
|
290 |
+
"272": "PART|Foreign=Yes|mark",
|
291 |
+
"273": "PART|Foreign=Yes|orphan",
|
292 |
+
"274": "PART|Polarity=Neg|advcl",
|
293 |
+
"275": "PART|Polarity=Neg|ccomp",
|
294 |
+
"276": "PART|Polarity=Neg|conj",
|
295 |
+
"277": "PART|Polarity=Neg|discourse",
|
296 |
+
"278": "PART|_|advcl",
|
297 |
+
"279": "PART|_|advmod",
|
298 |
+
"280": "PART|_|case",
|
299 |
+
"281": "PART|_|ccomp",
|
300 |
+
"282": "PART|_|discourse",
|
301 |
+
"283": "PART|_|fixed",
|
302 |
+
"284": "PART|_|mark",
|
303 |
+
"285": "PART|_|parataxis",
|
304 |
+
"286": "PART|_|root",
|
305 |
+
"287": "PRON|Definite=Def|Foreign=Yes|Number=Plur|Person=3|PronType=Prs|nmod",
|
306 |
+
"288": "PRON|Definite=Def|Gender=Fem|Number=Sing|Person=2|Polarity=Neg|PronType=Prs|nsubj",
|
307 |
+
"289": "PRON|Definite=Def|Gender=Fem|Number=Sing|Person=2|Poss=Yes|PronType=Prs|Reflex=Yes|obl:npmod",
|
308 |
+
"290": "PRON|Definite=Def|Gender=Fem|Number=Sing|Person=2|PronType=Prs|acl:relcl",
|
309 |
+
"291": "PRON|Definite=Def|Gender=Fem|Number=Sing|Person=2|PronType=Prs|dislocated",
|
310 |
+
"292": "PRON|Definite=Def|Gender=Fem|Number=Sing|Person=2|PronType=Prs|iobj",
|
311 |
+
"293": "PRON|Definite=Def|Gender=Fem|Number=Sing|Person=2|PronType=Prs|nsubj",
|
312 |
+
"294": "PRON|Definite=Def|Gender=Fem|Number=Sing|Person=2|PronType=Prs|obj",
|
313 |
+
"295": "PRON|Definite=Def|Gender=Fem|Number=Sing|Person=2|PronType=Prs|obl",
|
314 |
+
"296": "PRON|Definite=Def|Gender=Fem|Number=Sing|Person=2|PronType=Prs|obl:npmod",
|
315 |
+
"297": "PRON|Definite=Def|Gender=Fem|Number=Sing|Person=2|PronType=Prs|vocative",
|
316 |
+
"298": "PRON|Definite=Def|Gender=Fem|Number=Sing|Person=3|PronType=Prs|acl:relcl",
|
317 |
+
"299": "PRON|Definite=Def|Gender=Fem|Number=Sing|Person=3|PronType=Prs|advcl",
|
318 |
+
"300": "PRON|Definite=Def|Gender=Fem|Number=Sing|Person=3|PronType=Prs|conj",
|
319 |
+
"301": "PRON|Definite=Def|Gender=Fem|Number=Sing|Person=3|PronType=Prs|dislocated",
|
320 |
+
"302": "PRON|Definite=Def|Gender=Fem|Number=Sing|Person=3|PronType=Prs|iobj",
|
321 |
+
"303": "PRON|Definite=Def|Gender=Fem|Number=Sing|Person=3|PronType=Prs|nmod",
|
322 |
+
"304": "PRON|Definite=Def|Gender=Fem|Number=Sing|Person=3|PronType=Prs|nsubj",
|
323 |
+
"305": "PRON|Definite=Def|Gender=Fem|Number=Sing|Person=3|PronType=Prs|obj",
|
324 |
+
"306": "PRON|Definite=Def|Gender=Fem|Number=Sing|Person=3|PronType=Prs|obl",
|
325 |
+
"307": "PRON|Definite=Def|Gender=Fem|Number=Sing|Person=3|PronType=Prs|obl:npmod",
|
326 |
+
"308": "PRON|Definite=Def|Gender=Masc|Number=Sing|Person=2|PronType=Prs|acl:relcl",
|
327 |
+
"309": "PRON|Definite=Def|Gender=Masc|Number=Sing|Person=2|PronType=Prs|advcl",
|
328 |
+
"310": "PRON|Definite=Def|Gender=Masc|Number=Sing|Person=2|PronType=Prs|ccomp",
|
329 |
+
"311": "PRON|Definite=Def|Gender=Masc|Number=Sing|Person=2|PronType=Prs|dislocated",
|
330 |
+
"312": "PRON|Definite=Def|Gender=Masc|Number=Sing|Person=2|PronType=Prs|iobj",
|
331 |
+
"313": "PRON|Definite=Def|Gender=Masc|Number=Sing|Person=2|PronType=Prs|nmod",
|
332 |
+
"314": "PRON|Definite=Def|Gender=Masc|Number=Sing|Person=2|PronType=Prs|nsubj",
|
333 |
+
"315": "PRON|Definite=Def|Gender=Masc|Number=Sing|Person=2|PronType=Prs|obj",
|
334 |
+
"316": "PRON|Definite=Def|Gender=Masc|Number=Sing|Person=2|PronType=Prs|obl",
|
335 |
+
"317": "PRON|Definite=Def|Gender=Masc|Number=Sing|Person=2|PronType=Prs|obl:npmod",
|
336 |
+
"318": "PRON|Definite=Def|Gender=Masc|Number=Sing|Person=2|PronType=Prs|orphan",
|
337 |
+
"319": "PRON|Definite=Def|Gender=Masc|Number=Sing|Person=2|PronType=Prs|parataxis",
|
338 |
+
"320": "PRON|Definite=Def|Gender=Masc|Number=Sing|Person=2|PronType=Prs|root",
|
339 |
+
"321": "PRON|Definite=Def|Gender=Masc|Number=Sing|Person=2|PronType=Prs|vocative",
|
340 |
+
"322": "PRON|Definite=Def|Gender=Masc|Number=Sing|Person=3|PronType=Prs|acl:relcl",
|
341 |
+
"323": "PRON|Definite=Def|Gender=Masc|Number=Sing|Person=3|PronType=Prs|advcl",
|
342 |
+
"324": "PRON|Definite=Def|Gender=Masc|Number=Sing|Person=3|PronType=Prs|ccomp",
|
343 |
+
"325": "PRON|Definite=Def|Gender=Masc|Number=Sing|Person=3|PronType=Prs|compound",
|
344 |
+
"326": "PRON|Definite=Def|Gender=Masc|Number=Sing|Person=3|PronType=Prs|dislocated",
|
345 |
+
"327": "PRON|Definite=Def|Gender=Masc|Number=Sing|Person=3|PronType=Prs|iobj",
|
346 |
+
"328": "PRON|Definite=Def|Gender=Masc|Number=Sing|Person=3|PronType=Prs|nmod",
|
347 |
+
"329": "PRON|Definite=Def|Gender=Masc|Number=Sing|Person=3|PronType=Prs|nsubj",
|
348 |
+
"330": "PRON|Definite=Def|Gender=Masc|Number=Sing|Person=3|PronType=Prs|obj",
|
349 |
+
"331": "PRON|Definite=Def|Gender=Masc|Number=Sing|Person=3|PronType=Prs|obl",
|
350 |
+
"332": "PRON|Definite=Def|Gender=Masc|Number=Sing|Person=3|PronType=Prs|obl:npmod",
|
351 |
+
"333": "PRON|Definite=Def|Gender=Masc|Number=Sing|Person=3|PronType=Prs|parataxis",
|
352 |
+
"334": "PRON|Definite=Def|Gender=Masc|Number=Sing|Person=3|PronType=Prs|root",
|
353 |
+
"335": "PRON|Definite=Def|Number=Plur|Person=1|PronType=Prs|acl:relcl",
|
354 |
+
"336": "PRON|Definite=Def|Number=Plur|Person=1|PronType=Prs|conj",
|
355 |
+
"337": "PRON|Definite=Def|Number=Plur|Person=1|PronType=Prs|dislocated",
|
356 |
+
"338": "PRON|Definite=Def|Number=Plur|Person=1|PronType=Prs|iobj",
|
357 |
+
"339": "PRON|Definite=Def|Number=Plur|Person=1|PronType=Prs|nmod",
|
358 |
+
"340": "PRON|Definite=Def|Number=Plur|Person=1|PronType=Prs|nsubj",
|
359 |
+
"341": "PRON|Definite=Def|Number=Plur|Person=1|PronType=Prs|obj",
|
360 |
+
"342": "PRON|Definite=Def|Number=Plur|Person=1|PronType=Prs|obl",
|
361 |
+
"343": "PRON|Definite=Def|Number=Plur|Person=1|PronType=Prs|obl:npmod",
|
362 |
+
"344": "PRON|Definite=Def|Number=Plur|Person=1|PronType=Prs|reparandum",
|
363 |
+
"345": "PRON|Definite=Def|Number=Plur|Person=1|PronType=Prs|root",
|
364 |
+
"346": "PRON|Definite=Def|Number=Plur|Person=2|PronType=Prs|acl:relcl",
|
365 |
+
"347": "PRON|Definite=Def|Number=Plur|Person=2|PronType=Prs|advcl",
|
366 |
+
"348": "PRON|Definite=Def|Number=Plur|Person=2|PronType=Prs|ccomp",
|
367 |
+
"349": "PRON|Definite=Def|Number=Plur|Person=2|PronType=Prs|conj",
|
368 |
+
"350": "PRON|Definite=Def|Number=Plur|Person=2|PronType=Prs|dislocated",
|
369 |
+
"351": "PRON|Definite=Def|Number=Plur|Person=2|PronType=Prs|iobj",
|
370 |
+
"352": "PRON|Definite=Def|Number=Plur|Person=2|PronType=Prs|nmod",
|
371 |
+
"353": "PRON|Definite=Def|Number=Plur|Person=2|PronType=Prs|nsubj",
|
372 |
+
"354": "PRON|Definite=Def|Number=Plur|Person=2|PronType=Prs|obj",
|
373 |
+
"355": "PRON|Definite=Def|Number=Plur|Person=2|PronType=Prs|obl",
|
374 |
+
"356": "PRON|Definite=Def|Number=Plur|Person=2|PronType=Prs|obl:npmod",
|
375 |
+
"357": "PRON|Definite=Def|Number=Plur|Person=2|PronType=Prs|vocative",
|
376 |
+
"358": "PRON|Definite=Def|Number=Plur|Person=3|PronType=Prs|acl:relcl",
|
377 |
+
"359": "PRON|Definite=Def|Number=Plur|Person=3|PronType=Prs|advcl",
|
378 |
+
"360": "PRON|Definite=Def|Number=Plur|Person=3|PronType=Prs|compound",
|
379 |
+
"361": "PRON|Definite=Def|Number=Plur|Person=3|PronType=Prs|conj",
|
380 |
+
"362": "PRON|Definite=Def|Number=Plur|Person=3|PronType=Prs|dislocated",
|
381 |
+
"363": "PRON|Definite=Def|Number=Plur|Person=3|PronType=Prs|iobj",
|
382 |
+
"364": "PRON|Definite=Def|Number=Plur|Person=3|PronType=Prs|nmod",
|
383 |
+
"365": "PRON|Definite=Def|Number=Plur|Person=3|PronType=Prs|nsubj",
|
384 |
+
"366": "PRON|Definite=Def|Number=Plur|Person=3|PronType=Prs|obj",
|
385 |
+
"367": "PRON|Definite=Def|Number=Plur|Person=3|PronType=Prs|obl",
|
386 |
+
"368": "PRON|Definite=Def|Number=Plur|Person=3|PronType=Prs|obl:npmod",
|
387 |
+
"369": "PRON|Definite=Def|Number=Plur|Person=3|PronType=Prs|orphan",
|
388 |
+
"370": "PRON|Definite=Def|Number=Plur|Person=3|PronType=Prs|root",
|
389 |
+
"371": "PRON|Definite=Def|Number=Sing|Person=1|PronType=Prs|acl:relcl",
|
390 |
+
"372": "PRON|Definite=Def|Number=Sing|Person=1|PronType=Prs|ccomp",
|
391 |
+
"373": "PRON|Definite=Def|Number=Sing|Person=1|PronType=Prs|conj",
|
392 |
+
"374": "PRON|Definite=Def|Number=Sing|Person=1|PronType=Prs|dislocated",
|
393 |
+
"375": "PRON|Definite=Def|Number=Sing|Person=1|PronType=Prs|iobj",
|
394 |
+
"376": "PRON|Definite=Def|Number=Sing|Person=1|PronType=Prs|nmod",
|
395 |
+
"377": "PRON|Definite=Def|Number=Sing|Person=1|PronType=Prs|nsubj",
|
396 |
+
"378": "PRON|Definite=Def|Number=Sing|Person=1|PronType=Prs|obj",
|
397 |
+
"379": "PRON|Definite=Def|Number=Sing|Person=1|PronType=Prs|obl",
|
398 |
+
"380": "PRON|Definite=Def|Number=Sing|Person=1|PronType=Prs|obl:npmod",
|
399 |
+
"381": "PRON|Definite=Def|Number=Sing|Person=1|PronType=Prs|parataxis",
|
400 |
+
"382": "PRON|Definite=Def|Number=Sing|Person=1|PronType=Prs|root",
|
401 |
+
"383": "PRON|Definite=Def|Poss=Yes|PronType=Prs|det",
|
402 |
+
"384": "PRON|Gender=Fem|Number=Sing|cop",
|
403 |
+
"385": "PRON|Gender=Masc|Number=Sing|cop",
|
404 |
+
"386": "PRON|Number=Plur|cop",
|
405 |
+
"387": "PRON|Number=Plur|reparandum",
|
406 |
+
"388": "PRON|PronType=Ind|det",
|
407 |
+
"389": "PRON|PronType=Int|advcl",
|
408 |
+
"390": "PRON|PronType=Int|ccomp",
|
409 |
+
"391": "PRON|PronType=Int|conj",
|
410 |
+
"392": "PRON|PronType=Int|nmod",
|
411 |
+
"393": "PRON|PronType=Int|nsubj",
|
412 |
+
"394": "PRON|PronType=Int|obj",
|
413 |
+
"395": "PRON|PronType=Int|obl",
|
414 |
+
"396": "PRON|PronType=Int|obl:npmod",
|
415 |
+
"397": "PRON|PronType=Int|parataxis",
|
416 |
+
"398": "PRON|PronType=Int|root",
|
417 |
+
"399": "PRON|VerbForm=Fin|acl:relcl",
|
418 |
+
"400": "PRON|VerbForm=Fin|xcomp",
|
419 |
+
"401": "PRON|_|det",
|
420 |
+
"402": "PRON|_|nsubj",
|
421 |
+
"403": "PRON|_|obj",
|
422 |
+
"404": "PRON|_|obl",
|
423 |
+
"405": "PRON|_|root",
|
424 |
+
"406": "PROPN|Foreign=Yes|acl:relcl",
|
425 |
+
"407": "PROPN|Foreign=Yes|advcl",
|
426 |
+
"408": "PROPN|Foreign=Yes|appos",
|
427 |
+
"409": "PROPN|Foreign=Yes|ccomp",
|
428 |
+
"410": "PROPN|Foreign=Yes|conj",
|
429 |
+
"411": "PROPN|Foreign=Yes|dislocated",
|
430 |
+
"412": "PROPN|Foreign=Yes|flat",
|
431 |
+
"413": "PROPN|Foreign=Yes|nmod",
|
432 |
+
"414": "PROPN|Foreign=Yes|nsubj",
|
433 |
+
"415": "PROPN|Foreign=Yes|obj",
|
434 |
+
"416": "PROPN|Foreign=Yes|obl",
|
435 |
+
"417": "PROPN|Foreign=Yes|orphan",
|
436 |
+
"418": "PROPN|Foreign=Yes|parataxis",
|
437 |
+
"419": "PROPN|Foreign=Yes|root",
|
438 |
+
"420": "PROPN|Foreign=Yes|vocative",
|
439 |
+
"421": "PROPN|Foreign=Yes|xcomp",
|
440 |
+
"422": "PROPN|_|appos",
|
441 |
+
"423": "PROPN|_|ccomp",
|
442 |
+
"424": "PROPN|_|compound",
|
443 |
+
"425": "PROPN|_|dislocated",
|
444 |
+
"426": "PROPN|_|flat",
|
445 |
+
"427": "PROPN|_|nmod",
|
446 |
+
"428": "PROPN|_|nsubj",
|
447 |
+
"429": "PROPN|_|obl",
|
448 |
+
"430": "PROPN|_|root",
|
449 |
+
"431": "PROPN|_|vocative",
|
450 |
+
"432": "PUNCT|_|punct",
|
451 |
+
"433": "SCONJ|Foreign=Yes|Polarity=Neg|mark",
|
452 |
+
"434": "SCONJ|Foreign=Yes|mark",
|
453 |
+
"435": "SCONJ|_|case",
|
454 |
+
"436": "SCONJ|_|mark",
|
455 |
+
"437": "SCONJ|_|reparandum",
|
456 |
+
"438": "VERB|Foreign=Yes|VerbForm=Fin|acl:relcl",
|
457 |
+
"439": "VERB|Foreign=Yes|VerbForm=Fin|advcl",
|
458 |
+
"440": "VERB|Foreign=Yes|VerbForm=Fin|ccomp",
|
459 |
+
"441": "VERB|Foreign=Yes|VerbForm=Fin|conj",
|
460 |
+
"442": "VERB|Foreign=Yes|VerbForm=Fin|csubj",
|
461 |
+
"443": "VERB|Foreign=Yes|VerbForm=Fin|nmod",
|
462 |
+
"444": "VERB|Foreign=Yes|VerbForm=Fin|obl",
|
463 |
+
"445": "VERB|Foreign=Yes|VerbForm=Fin|parataxis",
|
464 |
+
"446": "VERB|Foreign=Yes|VerbForm=Fin|root",
|
465 |
+
"447": "VERB|Foreign=Yes|VerbForm=Fin|xcomp",
|
466 |
+
"448": "VERB|Foreign=Yes|VerbForm=Inf|conj",
|
467 |
+
"449": "VERB|Foreign=Yes|VerbForm=Inf|csubj",
|
468 |
+
"450": "VERB|Foreign=Yes|VerbForm=Inf|obl",
|
469 |
+
"451": "VERB|Foreign=Yes|VerbForm=Inf|xcomp",
|
470 |
+
"452": "VERB|Foreign=Yes|advcl",
|
471 |
+
"453": "VERB|Foreign=Yes|ccomp",
|
472 |
+
"454": "VERB|Polarity=Neg|acl:relcl",
|
473 |
+
"455": "VERB|Polarity=Neg|advcl",
|
474 |
+
"456": "VERB|Polarity=Neg|ccomp",
|
475 |
+
"457": "VERB|Polarity=Neg|conj",
|
476 |
+
"458": "VERB|Polarity=Neg|csubj",
|
477 |
+
"459": "VERB|Polarity=Neg|orphan",
|
478 |
+
"460": "VERB|Polarity=Neg|parataxis",
|
479 |
+
"461": "VERB|Polarity=Neg|root",
|
480 |
+
"462": "VERB|VerbForm=Fin|acl",
|
481 |
+
"463": "VERB|VerbForm=Fin|acl:relcl",
|
482 |
+
"464": "VERB|VerbForm=Fin|advcl",
|
483 |
+
"465": "VERB|VerbForm=Fin|ccomp",
|
484 |
+
"466": "VERB|VerbForm=Fin|compound",
|
485 |
+
"467": "VERB|VerbForm=Fin|conj",
|
486 |
+
"468": "VERB|VerbForm=Fin|csubj",
|
487 |
+
"469": "VERB|VerbForm=Fin|dep",
|
488 |
+
"470": "VERB|VerbForm=Fin|discourse",
|
489 |
+
"471": "VERB|VerbForm=Fin|dislocated",
|
490 |
+
"472": "VERB|VerbForm=Fin|nmod",
|
491 |
+
"473": "VERB|VerbForm=Fin|nsubj",
|
492 |
+
"474": "VERB|VerbForm=Fin|obj",
|
493 |
+
"475": "VERB|VerbForm=Fin|obl",
|
494 |
+
"476": "VERB|VerbForm=Fin|orphan",
|
495 |
+
"477": "VERB|VerbForm=Fin|parataxis",
|
496 |
+
"478": "VERB|VerbForm=Fin|root",
|
497 |
+
"479": "VERB|VerbForm=Fin|xcomp",
|
498 |
+
"480": "VERB|VerbForm=Inf|acl",
|
499 |
+
"481": "VERB|VerbForm=Inf|advcl",
|
500 |
+
"482": "VERB|VerbForm=Inf|ccomp",
|
501 |
+
"483": "VERB|VerbForm=Inf|conj",
|
502 |
+
"484": "VERB|VerbForm=Inf|csubj",
|
503 |
+
"485": "VERB|VerbForm=Inf|dislocated",
|
504 |
+
"486": "VERB|VerbForm=Inf|obl",
|
505 |
+
"487": "VERB|VerbForm=Inf|root",
|
506 |
+
"488": "VERB|VerbForm=Inf|xcomp",
|
507 |
+
"489": "VERB|_|acl",
|
508 |
+
"490": "VERB|_|acl:relcl",
|
509 |
+
"491": "VERB|_|advcl",
|
510 |
+
"492": "VERB|_|ccomp",
|
511 |
+
"493": "VERB|_|conj",
|
512 |
+
"494": "VERB|_|csubj",
|
513 |
+
"495": "VERB|_|dislocated",
|
514 |
+
"496": "VERB|_|nmod",
|
515 |
+
"497": "VERB|_|obl",
|
516 |
+
"498": "VERB|_|parataxis",
|
517 |
+
"499": "VERB|_|root",
|
518 |
+
"500": "VERB|_|xcomp",
|
519 |
+
"501": "X|Foreign=Yes|Polarity=Neg|advmod",
|
520 |
+
"502": "X|Foreign=Yes|appos",
|
521 |
+
"503": "X|Foreign=Yes|ccomp",
|
522 |
+
"504": "X|Foreign=Yes|dep",
|
523 |
+
"505": "X|Foreign=Yes|fixed",
|
524 |
+
"506": "X|Foreign=Yes|obl",
|
525 |
+
"507": "X|Foreign=Yes|root",
|
526 |
+
"508": "X|Polarity=Neg|advmod",
|
527 |
+
"509": "X|Polarity=Neg|amod",
|
528 |
+
"510": "X|_|advcl",
|
529 |
+
"511": "X|_|conj",
|
530 |
+
"512": "X|_|dep",
|
531 |
+
"513": "X|_|goeswith",
|
532 |
+
"514": "X|_|nmod",
|
533 |
+
"515": "X|_|obl",
|
534 |
+
"516": "X|_|parataxis",
|
535 |
+
"517": "X|_|reparandum",
|
536 |
+
"518": "X|_|root"
|
537 |
+
},
|
538 |
+
"initializer_range": 0.02,
|
539 |
+
"intermediate_size": 3072,
|
540 |
+
"label2id": {
|
541 |
+
"-|_|dep": 0,
|
542 |
+
"ADJ|Foreign=Yes|amod": 1,
|
543 |
+
"ADJ|_|amod": 2,
|
544 |
+
"ADP|Foreign=Yes|case": 3,
|
545 |
+
"ADP|_|advmod": 4,
|
546 |
+
"ADP|_|appos": 5,
|
547 |
+
"ADP|_|case": 6,
|
548 |
+
"ADP|_|cc": 7,
|
549 |
+
"ADP|_|fixed": 8,
|
550 |
+
"ADP|_|nmod": 9,
|
551 |
+
"ADP|_|nsubj": 10,
|
552 |
+
"ADP|_|obj": 11,
|
553 |
+
"ADP|_|obl": 12,
|
554 |
+
"ADP|_|reparandum": 13,
|
555 |
+
"ADV|Definite=Def|PronType=Prs|Reflex=Yes|advmod": 14,
|
556 |
+
"ADV|Foreign=Yes|Polarity=Neg|advmod": 15,
|
557 |
+
"ADV|Foreign=Yes|advmod": 16,
|
558 |
+
"ADV|Foreign=Yes|mark": 17,
|
559 |
+
"ADV|Polarity=Neg|advcl": 18,
|
560 |
+
"ADV|Polarity=Neg|advmod": 19,
|
561 |
+
"ADV|PronType=Int|advmod": 20,
|
562 |
+
"ADV|PronType=Int|ccomp": 21,
|
563 |
+
"ADV|PronType=Int|conj": 22,
|
564 |
+
"ADV|PronType=Int|nmod": 23,
|
565 |
+
"ADV|PronType=Int|parataxis": 24,
|
566 |
+
"ADV|PronType=Int|root": 25,
|
567 |
+
"ADV|PronType=Tot|advmod": 26,
|
568 |
+
"ADV|_|acl:relcl": 27,
|
569 |
+
"ADV|_|advcl": 28,
|
570 |
+
"ADV|_|advmod": 29,
|
571 |
+
"ADV|_|case": 30,
|
572 |
+
"ADV|_|conj": 31,
|
573 |
+
"ADV|_|fixed": 32,
|
574 |
+
"ADV|_|mark": 33,
|
575 |
+
"ADV|_|nmod": 34,
|
576 |
+
"ADV|_|obl": 35,
|
577 |
+
"ADV|_|obl:npmod": 36,
|
578 |
+
"ADV|_|root": 37,
|
579 |
+
"ADV|_|xcomp": 38,
|
580 |
+
"AUX|Foreign=Yes|aux": 39,
|
581 |
+
"AUX|Polarity=Neg|aux": 40,
|
582 |
+
"AUX|VerbForm=Fin|aux": 41,
|
583 |
+
"AUX|_|aux": 42,
|
584 |
+
"AUX|_|orphan": 43,
|
585 |
+
"AUX|_|reparandum": 44,
|
586 |
+
"CCONJ|Foreign=Yes|Polarity=Neg|advmod": 45,
|
587 |
+
"CCONJ|Foreign=Yes|Polarity=Neg|cc": 46,
|
588 |
+
"CCONJ|Foreign=Yes|advmod": 47,
|
589 |
+
"CCONJ|Foreign=Yes|cc": 48,
|
590 |
+
"CCONJ|Foreign=Yes|discourse": 49,
|
591 |
+
"CCONJ|Foreign=Yes|fixed": 50,
|
592 |
+
"CCONJ|_|advmod": 51,
|
593 |
+
"CCONJ|_|case": 52,
|
594 |
+
"CCONJ|_|cc": 53,
|
595 |
+
"CCONJ|_|ccomp": 54,
|
596 |
+
"CCONJ|_|discourse": 55,
|
597 |
+
"CCONJ|_|fixed": 56,
|
598 |
+
"CCONJ|_|reparandum": 57,
|
599 |
+
"DET|Definite=Def|Foreign=Yes|Number=Plur|PronType=Dem|root": 58,
|
600 |
+
"DET|Definite=Def|Gender=Fem|Gender[psor]=Fem|Number=Sing|Number[psor]=Sing|Person=2|Poss=Yes|PronType=Prs|det": 59,
|
601 |
+
"DET|Definite=Def|Gender=Fem|Gender[psor]=Fem|Number=Sing|Number[psor]=Sing|Person=3|Poss=Yes|PronType=Prs|det": 60,
|
602 |
+
"DET|Definite=Def|Gender=Fem|Gender[psor]=Masc|Number=Sing|Number[psor]=Sing|Person=2|Poss=Yes|PronType=Prs|det": 61,
|
603 |
+
"DET|Definite=Def|Gender=Fem|Gender[psor]=Masc|Number=Sing|Number[psor]=Sing|Person=3|Poss=Yes|PronType=Prs|det": 62,
|
604 |
+
"DET|Definite=Def|Gender=Fem|Gender[psor]=Masc|Number=Sing|Number[psor]=Sing|Person=3|Poss=Yes|PronType=Prs|obl": 63,
|
605 |
+
"DET|Definite=Def|Gender=Fem|Number=Sing|Number[psor]=Plur|Person=1|Poss=Yes|PronType=Prs|det": 64,
|
606 |
+
"DET|Definite=Def|Gender=Fem|Number=Sing|Number[psor]=Plur|Person=2|Poss=Yes|PronType=Prs|det": 65,
|
607 |
+
"DET|Definite=Def|Gender=Fem|Number=Sing|Number[psor]=Plur|Person=3|Poss=Yes|PronType=Prs|det": 66,
|
608 |
+
"DET|Definite=Def|Gender=Fem|Number=Sing|Number[psor]=Sing|Person=1|Poss=Yes|PronType=Prs|det": 67,
|
609 |
+
"DET|Definite=Def|Gender=Fem|Number=Sing|Poss=Yes|PronType=Prs|acl:relcl": 68,
|
610 |
+
"DET|Definite=Def|Gender=Fem|Number=Sing|Poss=Yes|PronType=Prs|ccomp": 69,
|
611 |
+
"DET|Definite=Def|Gender=Fem|Number=Sing|Poss=Yes|PronType=Prs|conj": 70,
|
612 |
+
"DET|Definite=Def|Gender=Fem|Number=Sing|Poss=Yes|PronType=Prs|nmod": 71,
|
613 |
+
"DET|Definite=Def|Gender=Fem|Number=Sing|PronType=Art|appos": 72,
|
614 |
+
"DET|Definite=Def|Gender=Fem|Number=Sing|PronType=Art|ccomp": 73,
|
615 |
+
"DET|Definite=Def|Gender=Fem|Number=Sing|PronType=Art|conj": 74,
|
616 |
+
"DET|Definite=Def|Gender=Fem|Number=Sing|PronType=Art|det": 75,
|
617 |
+
"DET|Definite=Def|Gender=Fem|Number=Sing|PronType=Art|dislocated": 76,
|
618 |
+
"DET|Definite=Def|Gender=Fem|Number=Sing|PronType=Art|obj": 77,
|
619 |
+
"DET|Definite=Def|Gender=Fem|Number=Sing|PronType=Art|obl": 78,
|
620 |
+
"DET|Definite=Def|Gender=Fem|Number=Sing|PronType=Dem|acl:relcl": 79,
|
621 |
+
"DET|Definite=Def|Gender=Fem|Number=Sing|PronType=Dem|appos": 80,
|
622 |
+
"DET|Definite=Def|Gender=Fem|Number=Sing|PronType=Dem|ccomp": 81,
|
623 |
+
"DET|Definite=Def|Gender=Fem|Number=Sing|PronType=Dem|det": 82,
|
624 |
+
"DET|Definite=Def|Gender=Fem|Number=Sing|PronType=Dem|dislocated": 83,
|
625 |
+
"DET|Definite=Def|Gender=Fem|Number=Sing|PronType=Dem|nsubj": 84,
|
626 |
+
"DET|Definite=Def|Gender=Fem|Number=Sing|PronType=Dem|obl": 85,
|
627 |
+
"DET|Definite=Def|Gender=Fem|Number=Sing|PronType=Dem|parataxis": 86,
|
628 |
+
"DET|Definite=Def|Gender=Fem|Number=Sing|PronType=Dem|root": 87,
|
629 |
+
"DET|Definite=Def|Gender=Masc|Gender[psor]=Fem|Number=Sing|Number[psor]=Sing|Person=2|Poss=Yes|PronType=Prs|det": 88,
|
630 |
+
"DET|Definite=Def|Gender=Masc|Gender[psor]=Fem|Number=Sing|Number[psor]=Sing|Person=3|Poss=Yes|PronType=Prs|det": 89,
|
631 |
+
"DET|Definite=Def|Gender=Masc|Gender[psor]=Masc|Number=Sing|Number[psor]=Sing|Person=2|Poss=Yes|PronType=Prs|acl:relcl": 90,
|
632 |
+
"DET|Definite=Def|Gender=Masc|Gender[psor]=Masc|Number=Sing|Number[psor]=Sing|Person=2|Poss=Yes|PronType=Prs|appos": 91,
|
633 |
+
"DET|Definite=Def|Gender=Masc|Gender[psor]=Masc|Number=Sing|Number[psor]=Sing|Person=2|Poss=Yes|PronType=Prs|det": 92,
|
634 |
+
"DET|Definite=Def|Gender=Masc|Gender[psor]=Masc|Number=Sing|Number[psor]=Sing|Person=2|Poss=Yes|PronType=Prs|obj": 93,
|
635 |
+
"DET|Definite=Def|Gender=Masc|Gender[psor]=Masc|Number=Sing|Number[psor]=Sing|Person=3|Poss=Yes|PronType=Prs|acl:relcl": 94,
|
636 |
+
"DET|Definite=Def|Gender=Masc|Gender[psor]=Masc|Number=Sing|Number[psor]=Sing|Person=3|Poss=Yes|PronType=Prs|det": 95,
|
637 |
+
"DET|Definite=Def|Gender=Masc|Number=Sing|Number[psor]=Plur|Person=1|Poss=Yes|PronType=Prs|det": 96,
|
638 |
+
"DET|Definite=Def|Gender=Masc|Number=Sing|Number[psor]=Plur|Person=2|Poss=Yes|PronType=Prs|conj": 97,
|
639 |
+
"DET|Definite=Def|Gender=Masc|Number=Sing|Number[psor]=Plur|Person=2|Poss=Yes|PronType=Prs|det": 98,
|
640 |
+
"DET|Definite=Def|Gender=Masc|Number=Sing|Number[psor]=Plur|Person=2|Poss=Yes|PronType=Prs|obl": 99,
|
641 |
+
"DET|Definite=Def|Gender=Masc|Number=Sing|Number[psor]=Plur|Person=2|Poss=Yes|PronType=Prs|parataxis": 100,
|
642 |
+
"DET|Definite=Def|Gender=Masc|Number=Sing|Number[psor]=Plur|Person=2|Poss=Yes|PronType=Prs|root": 101,
|
643 |
+
"DET|Definite=Def|Gender=Masc|Number=Sing|Number[psor]=Plur|Person=3|Poss=Yes|PronType=Prs|det": 102,
|
644 |
+
"DET|Definite=Def|Gender=Masc|Number=Sing|Number[psor]=Sing|Person=1|Poss=Yes|PronType=Prs|acl:relcl": 103,
|
645 |
+
"DET|Definite=Def|Gender=Masc|Number=Sing|Number[psor]=Sing|Person=1|Poss=Yes|PronType=Prs|ccomp": 104,
|
646 |
+
"DET|Definite=Def|Gender=Masc|Number=Sing|Number[psor]=Sing|Person=1|Poss=Yes|PronType=Prs|det": 105,
|
647 |
+
"DET|Definite=Def|Gender=Masc|Number=Sing|Poss=Yes|PronType=Prs|acl:relcl": 106,
|
648 |
+
"DET|Definite=Def|Gender=Masc|Number=Sing|Poss=Yes|PronType=Prs|appos": 107,
|
649 |
+
"DET|Definite=Def|Gender=Masc|Number=Sing|Poss=Yes|PronType=Prs|ccomp": 108,
|
650 |
+
"DET|Definite=Def|Gender=Masc|Number=Sing|Poss=Yes|PronType=Prs|conj": 109,
|
651 |
+
"DET|Definite=Def|Gender=Masc|Number=Sing|Poss=Yes|PronType=Prs|nmod": 110,
|
652 |
+
"DET|Definite=Def|Gender=Masc|Number=Sing|Poss=Yes|PronType=Prs|nsubj": 111,
|
653 |
+
"DET|Definite=Def|Gender=Masc|Number=Sing|Poss=Yes|PronType=Prs|obj": 112,
|
654 |
+
"DET|Definite=Def|Gender=Masc|Number=Sing|Poss=Yes|PronType=Prs|parataxis": 113,
|
655 |
+
"DET|Definite=Def|Gender=Masc|Number=Sing|Poss=Yes|PronType=Prs|root": 114,
|
656 |
+
"DET|Definite=Def|Gender=Masc|Number=Sing|Poss=Yes|PronType=Prs|vocative": 115,
|
657 |
+
"DET|Definite=Def|Gender=Masc|Number=Sing|PronType=Art|acl:relcl": 116,
|
658 |
+
"DET|Definite=Def|Gender=Masc|Number=Sing|PronType=Art|advcl": 117,
|
659 |
+
"DET|Definite=Def|Gender=Masc|Number=Sing|PronType=Art|appos": 118,
|
660 |
+
"DET|Definite=Def|Gender=Masc|Number=Sing|PronType=Art|ccomp": 119,
|
661 |
+
"DET|Definite=Def|Gender=Masc|Number=Sing|PronType=Art|conj": 120,
|
662 |
+
"DET|Definite=Def|Gender=Masc|Number=Sing|PronType=Art|det": 121,
|
663 |
+
"DET|Definite=Def|Gender=Masc|Number=Sing|PronType=Art|dislocated": 122,
|
664 |
+
"DET|Definite=Def|Gender=Masc|Number=Sing|PronType=Art|fixed": 123,
|
665 |
+
"DET|Definite=Def|Gender=Masc|Number=Sing|PronType=Art|nmod": 124,
|
666 |
+
"DET|Definite=Def|Gender=Masc|Number=Sing|PronType=Art|nsubj": 125,
|
667 |
+
"DET|Definite=Def|Gender=Masc|Number=Sing|PronType=Art|obj": 126,
|
668 |
+
"DET|Definite=Def|Gender=Masc|Number=Sing|PronType=Art|obl": 127,
|
669 |
+
"DET|Definite=Def|Gender=Masc|Number=Sing|PronType=Art|parataxis": 128,
|
670 |
+
"DET|Definite=Def|Gender=Masc|Number=Sing|PronType=Art|root": 129,
|
671 |
+
"DET|Definite=Def|Gender=Masc|Number=Sing|PronType=Art|vocative": 130,
|
672 |
+
"DET|Definite=Def|Gender=Masc|Number=Sing|PronType=Dem|acl:relcl": 131,
|
673 |
+
"DET|Definite=Def|Gender=Masc|Number=Sing|PronType=Dem|advcl": 132,
|
674 |
+
"DET|Definite=Def|Gender=Masc|Number=Sing|PronType=Dem|appos": 133,
|
675 |
+
"DET|Definite=Def|Gender=Masc|Number=Sing|PronType=Dem|ccomp": 134,
|
676 |
+
"DET|Definite=Def|Gender=Masc|Number=Sing|PronType=Dem|conj": 135,
|
677 |
+
"DET|Definite=Def|Gender=Masc|Number=Sing|PronType=Dem|det": 136,
|
678 |
+
"DET|Definite=Def|Gender=Masc|Number=Sing|PronType=Dem|dislocated": 137,
|
679 |
+
"DET|Definite=Def|Gender=Masc|Number=Sing|PronType=Dem|iobj": 138,
|
680 |
+
"DET|Definite=Def|Gender=Masc|Number=Sing|PronType=Dem|nmod": 139,
|
681 |
+
"DET|Definite=Def|Gender=Masc|Number=Sing|PronType=Dem|nsubj": 140,
|
682 |
+
"DET|Definite=Def|Gender=Masc|Number=Sing|PronType=Dem|obj": 141,
|
683 |
+
"DET|Definite=Def|Gender=Masc|Number=Sing|PronType=Dem|obl": 142,
|
684 |
+
"DET|Definite=Def|Gender=Masc|Number=Sing|PronType=Dem|parataxis": 143,
|
685 |
+
"DET|Definite=Def|Gender=Masc|Number=Sing|PronType=Dem|root": 144,
|
686 |
+
"DET|Definite=Def|Gender=Masc|Number=Sing|PronType=Dem|xcomp": 145,
|
687 |
+
"DET|Definite=Def|Gender[psor]=Fem|Number=Plur|Number[psor]=Sing|Person=2|Poss=Yes|PronType=Prs|det": 146,
|
688 |
+
"DET|Definite=Def|Gender[psor]=Fem|Number=Plur|Number[psor]=Sing|Person=3|Poss=Yes|PronType=Prs|det": 147,
|
689 |
+
"DET|Definite=Def|Gender[psor]=Masc|Number=Plur|Number[psor]=Sing|Person=2|Poss=Yes|PronType=Prs|acl:relcl": 148,
|
690 |
+
"DET|Definite=Def|Gender[psor]=Masc|Number=Plur|Number[psor]=Sing|Person=2|Poss=Yes|PronType=Prs|det": 149,
|
691 |
+
"DET|Definite=Def|Gender[psor]=Masc|Number=Plur|Number[psor]=Sing|Person=3|Poss=Yes|PronType=Prs|det": 150,
|
692 |
+
"DET|Definite=Def|Number=Plur|Number[psor]=Plur|Person=1|Poss=Yes|PronType=Prs|det": 151,
|
693 |
+
"DET|Definite=Def|Number=Plur|Number[psor]=Plur|Person=2|Poss=Yes|PronType=Prs|det": 152,
|
694 |
+
"DET|Definite=Def|Number=Plur|Number[psor]=Plur|Person=3|Poss=Yes|PronType=Prs|det": 153,
|
695 |
+
"DET|Definite=Def|Number=Plur|Number[psor]=Sing|Person=1|Poss=Yes|PronType=Prs|conj": 154,
|
696 |
+
"DET|Definite=Def|Number=Plur|Number[psor]=Sing|Person=1|Poss=Yes|PronType=Prs|det": 155,
|
697 |
+
"DET|Definite=Def|Number=Plur|Poss=Yes|PronType=Prs|advcl": 156,
|
698 |
+
"DET|Definite=Def|Number=Plur|Poss=Yes|PronType=Prs|ccomp": 157,
|
699 |
+
"DET|Definite=Def|Number=Plur|Poss=Yes|PronType=Prs|conj": 158,
|
700 |
+
"DET|Definite=Def|Number=Plur|Poss=Yes|PronType=Prs|nmod": 159,
|
701 |
+
"DET|Definite=Def|Number=Plur|Poss=Yes|PronType=Prs|obj": 160,
|
702 |
+
"DET|Definite=Def|Number=Plur|Poss=Yes|PronType=Prs|obl": 161,
|
703 |
+
"DET|Definite=Def|Number=Plur|Poss=Yes|PronType=Prs|root": 162,
|
704 |
+
"DET|Definite=Def|Number=Plur|PronType=Art|acl:relcl": 163,
|
705 |
+
"DET|Definite=Def|Number=Plur|PronType=Art|advcl": 164,
|
706 |
+
"DET|Definite=Def|Number=Plur|PronType=Art|appos": 165,
|
707 |
+
"DET|Definite=Def|Number=Plur|PronType=Art|ccomp": 166,
|
708 |
+
"DET|Definite=Def|Number=Plur|PronType=Art|conj": 167,
|
709 |
+
"DET|Definite=Def|Number=Plur|PronType=Art|det": 168,
|
710 |
+
"DET|Definite=Def|Number=Plur|PronType=Art|discourse": 169,
|
711 |
+
"DET|Definite=Def|Number=Plur|PronType=Art|dislocated": 170,
|
712 |
+
"DET|Definite=Def|Number=Plur|PronType=Art|nmod": 171,
|
713 |
+
"DET|Definite=Def|Number=Plur|PronType=Art|nsubj": 172,
|
714 |
+
"DET|Definite=Def|Number=Plur|PronType=Art|obj": 173,
|
715 |
+
"DET|Definite=Def|Number=Plur|PronType=Art|obl": 174,
|
716 |
+
"DET|Definite=Def|Number=Plur|PronType=Art|root": 175,
|
717 |
+
"DET|Definite=Def|Number=Plur|PronType=Art|vocative": 176,
|
718 |
+
"DET|Definite=Def|Number=Plur|PronType=Art|xcomp": 177,
|
719 |
+
"DET|Definite=Def|Number=Plur|PronType=Dem|acl:relcl": 178,
|
720 |
+
"DET|Definite=Def|Number=Plur|PronType=Dem|appos": 179,
|
721 |
+
"DET|Definite=Def|Number=Plur|PronType=Dem|ccomp": 180,
|
722 |
+
"DET|Definite=Def|Number=Plur|PronType=Dem|det": 181,
|
723 |
+
"DET|Definite=Def|Number=Plur|PronType=Dem|dislocated": 182,
|
724 |
+
"DET|Definite=Def|Number=Plur|PronType=Dem|nmod": 183,
|
725 |
+
"DET|Definite=Def|Number=Plur|PronType=Dem|nsubj": 184,
|
726 |
+
"DET|Definite=Def|Number=Plur|PronType=Dem|obj": 185,
|
727 |
+
"DET|Definite=Def|Number=Plur|PronType=Dem|obl": 186,
|
728 |
+
"DET|Definite=Def|Number=Plur|PronType=Dem|parataxis": 187,
|
729 |
+
"DET|Definite=Def|Number=Plur|PronType=Dem|root": 188,
|
730 |
+
"DET|Definite=Ind|Number=Sing|PronType=Art|det": 189,
|
731 |
+
"DET|Definite=Ind|Number=Sing|PronType=Art|obj": 190,
|
732 |
+
"DET|Definite=Ind|Number=Sing|PronType=Art|obl": 191,
|
733 |
+
"DET|PronType=Art|advmod": 192,
|
734 |
+
"DET|PronType=Art|det": 193,
|
735 |
+
"DET|PronType=Art|nsubj": 194,
|
736 |
+
"DET|PronType=Art|obj": 195,
|
737 |
+
"DET|_|compound": 196,
|
738 |
+
"DET|_|conj": 197,
|
739 |
+
"DET|_|det": 198,
|
740 |
+
"DET|_|nmod": 199,
|
741 |
+
"DET|_|obl": 200,
|
742 |
+
"DET|_|reparandum": 201,
|
743 |
+
"NOUN|Foreign=Yes|acl:relcl": 202,
|
744 |
+
"NOUN|Foreign=Yes|advcl": 203,
|
745 |
+
"NOUN|Foreign=Yes|appos": 204,
|
746 |
+
"NOUN|Foreign=Yes|ccomp": 205,
|
747 |
+
"NOUN|Foreign=Yes|conj": 206,
|
748 |
+
"NOUN|Foreign=Yes|dislocated": 207,
|
749 |
+
"NOUN|Foreign=Yes|fixed": 208,
|
750 |
+
"NOUN|Foreign=Yes|nmod": 209,
|
751 |
+
"NOUN|Foreign=Yes|nsubj": 210,
|
752 |
+
"NOUN|Foreign=Yes|obj": 211,
|
753 |
+
"NOUN|Foreign=Yes|obl": 212,
|
754 |
+
"NOUN|Foreign=Yes|obl:npmod": 213,
|
755 |
+
"NOUN|Foreign=Yes|orphan": 214,
|
756 |
+
"NOUN|Foreign=Yes|parataxis": 215,
|
757 |
+
"NOUN|Foreign=Yes|reparandum": 216,
|
758 |
+
"NOUN|Foreign=Yes|root": 217,
|
759 |
+
"NOUN|Foreign=Yes|vocative": 218,
|
760 |
+
"NOUN|Foreign=Yes|xcomp": 219,
|
761 |
+
"NOUN|PronType=Rcp|acl:relcl": 220,
|
762 |
+
"NOUN|PronType=Rcp|advcl": 221,
|
763 |
+
"NOUN|PronType=Rcp|nmod": 222,
|
764 |
+
"NOUN|PronType=Rcp|obj": 223,
|
765 |
+
"NOUN|PronType=Rcp|obl": 224,
|
766 |
+
"NOUN|_|acl:relcl": 225,
|
767 |
+
"NOUN|_|advcl": 226,
|
768 |
+
"NOUN|_|appos": 227,
|
769 |
+
"NOUN|_|case": 228,
|
770 |
+
"NOUN|_|ccomp": 229,
|
771 |
+
"NOUN|_|compound": 230,
|
772 |
+
"NOUN|_|conj": 231,
|
773 |
+
"NOUN|_|csubj": 232,
|
774 |
+
"NOUN|_|dep": 233,
|
775 |
+
"NOUN|_|discourse": 234,
|
776 |
+
"NOUN|_|dislocated": 235,
|
777 |
+
"NOUN|_|fixed": 236,
|
778 |
+
"NOUN|_|iobj": 237,
|
779 |
+
"NOUN|_|nmod": 238,
|
780 |
+
"NOUN|_|nsubj": 239,
|
781 |
+
"NOUN|_|obj": 240,
|
782 |
+
"NOUN|_|obl": 241,
|
783 |
+
"NOUN|_|obl:npmod": 242,
|
784 |
+
"NOUN|_|orphan": 243,
|
785 |
+
"NOUN|_|parataxis": 244,
|
786 |
+
"NOUN|_|reparandum": 245,
|
787 |
+
"NOUN|_|root": 246,
|
788 |
+
"NOUN|_|vocative": 247,
|
789 |
+
"NOUN|_|xcomp": 248,
|
790 |
+
"NUM|Foreign=Yes|NumType=Card|nsubj": 249,
|
791 |
+
"NUM|Foreign=Yes|NumType=Card|reparandum": 250,
|
792 |
+
"NUM|NumType=Card|acl:relcl": 251,
|
793 |
+
"NUM|NumType=Card|ccomp": 252,
|
794 |
+
"NUM|NumType=Card|compound": 253,
|
795 |
+
"NUM|NumType=Card|conj": 254,
|
796 |
+
"NUM|NumType=Card|dislocated": 255,
|
797 |
+
"NUM|NumType=Card|nmod": 256,
|
798 |
+
"NUM|NumType=Card|nsubj": 257,
|
799 |
+
"NUM|NumType=Card|nummod": 258,
|
800 |
+
"NUM|NumType=Card|obj": 259,
|
801 |
+
"NUM|NumType=Card|obl": 260,
|
802 |
+
"NUM|NumType=Card|obl:npmod": 261,
|
803 |
+
"NUM|NumType=Card|orphan": 262,
|
804 |
+
"NUM|NumType=Card|parataxis": 263,
|
805 |
+
"NUM|NumType=Card|root": 264,
|
806 |
+
"NUM|NumType=Card|xcomp": 265,
|
807 |
+
"PART|Foreign=Yes|Polarity=Neg|cc": 266,
|
808 |
+
"PART|Foreign=Yes|advcl": 267,
|
809 |
+
"PART|Foreign=Yes|advmod": 268,
|
810 |
+
"PART|Foreign=Yes|ccomp": 269,
|
811 |
+
"PART|Foreign=Yes|discourse": 270,
|
812 |
+
"PART|Foreign=Yes|fixed": 271,
|
813 |
+
"PART|Foreign=Yes|mark": 272,
|
814 |
+
"PART|Foreign=Yes|orphan": 273,
|
815 |
+
"PART|Polarity=Neg|advcl": 274,
|
816 |
+
"PART|Polarity=Neg|ccomp": 275,
|
817 |
+
"PART|Polarity=Neg|conj": 276,
|
818 |
+
"PART|Polarity=Neg|discourse": 277,
|
819 |
+
"PART|_|advcl": 278,
|
820 |
+
"PART|_|advmod": 279,
|
821 |
+
"PART|_|case": 280,
|
822 |
+
"PART|_|ccomp": 281,
|
823 |
+
"PART|_|discourse": 282,
|
824 |
+
"PART|_|fixed": 283,
|
825 |
+
"PART|_|mark": 284,
|
826 |
+
"PART|_|parataxis": 285,
|
827 |
+
"PART|_|root": 286,
|
828 |
+
"PRON|Definite=Def|Foreign=Yes|Number=Plur|Person=3|PronType=Prs|nmod": 287,
|
829 |
+
"PRON|Definite=Def|Gender=Fem|Number=Sing|Person=2|Polarity=Neg|PronType=Prs|nsubj": 288,
|
830 |
+
"PRON|Definite=Def|Gender=Fem|Number=Sing|Person=2|Poss=Yes|PronType=Prs|Reflex=Yes|obl:npmod": 289,
|
831 |
+
"PRON|Definite=Def|Gender=Fem|Number=Sing|Person=2|PronType=Prs|acl:relcl": 290,
|
832 |
+
"PRON|Definite=Def|Gender=Fem|Number=Sing|Person=2|PronType=Prs|dislocated": 291,
|
833 |
+
"PRON|Definite=Def|Gender=Fem|Number=Sing|Person=2|PronType=Prs|iobj": 292,
|
834 |
+
"PRON|Definite=Def|Gender=Fem|Number=Sing|Person=2|PronType=Prs|nsubj": 293,
|
835 |
+
"PRON|Definite=Def|Gender=Fem|Number=Sing|Person=2|PronType=Prs|obj": 294,
|
836 |
+
"PRON|Definite=Def|Gender=Fem|Number=Sing|Person=2|PronType=Prs|obl": 295,
|
837 |
+
"PRON|Definite=Def|Gender=Fem|Number=Sing|Person=2|PronType=Prs|obl:npmod": 296,
|
838 |
+
"PRON|Definite=Def|Gender=Fem|Number=Sing|Person=2|PronType=Prs|vocative": 297,
|
839 |
+
"PRON|Definite=Def|Gender=Fem|Number=Sing|Person=3|PronType=Prs|acl:relcl": 298,
|
840 |
+
"PRON|Definite=Def|Gender=Fem|Number=Sing|Person=3|PronType=Prs|advcl": 299,
|
841 |
+
"PRON|Definite=Def|Gender=Fem|Number=Sing|Person=3|PronType=Prs|conj": 300,
|
842 |
+
"PRON|Definite=Def|Gender=Fem|Number=Sing|Person=3|PronType=Prs|dislocated": 301,
|
843 |
+
"PRON|Definite=Def|Gender=Fem|Number=Sing|Person=3|PronType=Prs|iobj": 302,
|
844 |
+
"PRON|Definite=Def|Gender=Fem|Number=Sing|Person=3|PronType=Prs|nmod": 303,
|
845 |
+
"PRON|Definite=Def|Gender=Fem|Number=Sing|Person=3|PronType=Prs|nsubj": 304,
|
846 |
+
"PRON|Definite=Def|Gender=Fem|Number=Sing|Person=3|PronType=Prs|obj": 305,
|
847 |
+
"PRON|Definite=Def|Gender=Fem|Number=Sing|Person=3|PronType=Prs|obl": 306,
|
848 |
+
"PRON|Definite=Def|Gender=Fem|Number=Sing|Person=3|PronType=Prs|obl:npmod": 307,
|
849 |
+
"PRON|Definite=Def|Gender=Masc|Number=Sing|Person=2|PronType=Prs|acl:relcl": 308,
|
850 |
+
"PRON|Definite=Def|Gender=Masc|Number=Sing|Person=2|PronType=Prs|advcl": 309,
|
851 |
+
"PRON|Definite=Def|Gender=Masc|Number=Sing|Person=2|PronType=Prs|ccomp": 310,
|
852 |
+
"PRON|Definite=Def|Gender=Masc|Number=Sing|Person=2|PronType=Prs|dislocated": 311,
|
853 |
+
"PRON|Definite=Def|Gender=Masc|Number=Sing|Person=2|PronType=Prs|iobj": 312,
|
854 |
+
"PRON|Definite=Def|Gender=Masc|Number=Sing|Person=2|PronType=Prs|nmod": 313,
|
855 |
+
"PRON|Definite=Def|Gender=Masc|Number=Sing|Person=2|PronType=Prs|nsubj": 314,
|
856 |
+
"PRON|Definite=Def|Gender=Masc|Number=Sing|Person=2|PronType=Prs|obj": 315,
|
857 |
+
"PRON|Definite=Def|Gender=Masc|Number=Sing|Person=2|PronType=Prs|obl": 316,
|
858 |
+
"PRON|Definite=Def|Gender=Masc|Number=Sing|Person=2|PronType=Prs|obl:npmod": 317,
|
859 |
+
"PRON|Definite=Def|Gender=Masc|Number=Sing|Person=2|PronType=Prs|orphan": 318,
|
860 |
+
"PRON|Definite=Def|Gender=Masc|Number=Sing|Person=2|PronType=Prs|parataxis": 319,
|
861 |
+
"PRON|Definite=Def|Gender=Masc|Number=Sing|Person=2|PronType=Prs|root": 320,
|
862 |
+
"PRON|Definite=Def|Gender=Masc|Number=Sing|Person=2|PronType=Prs|vocative": 321,
|
863 |
+
"PRON|Definite=Def|Gender=Masc|Number=Sing|Person=3|PronType=Prs|acl:relcl": 322,
|
864 |
+
"PRON|Definite=Def|Gender=Masc|Number=Sing|Person=3|PronType=Prs|advcl": 323,
|
865 |
+
"PRON|Definite=Def|Gender=Masc|Number=Sing|Person=3|PronType=Prs|ccomp": 324,
|
866 |
+
"PRON|Definite=Def|Gender=Masc|Number=Sing|Person=3|PronType=Prs|compound": 325,
|
867 |
+
"PRON|Definite=Def|Gender=Masc|Number=Sing|Person=3|PronType=Prs|dislocated": 326,
|
868 |
+
"PRON|Definite=Def|Gender=Masc|Number=Sing|Person=3|PronType=Prs|iobj": 327,
|
869 |
+
"PRON|Definite=Def|Gender=Masc|Number=Sing|Person=3|PronType=Prs|nmod": 328,
|
870 |
+
"PRON|Definite=Def|Gender=Masc|Number=Sing|Person=3|PronType=Prs|nsubj": 329,
|
871 |
+
"PRON|Definite=Def|Gender=Masc|Number=Sing|Person=3|PronType=Prs|obj": 330,
|
872 |
+
"PRON|Definite=Def|Gender=Masc|Number=Sing|Person=3|PronType=Prs|obl": 331,
|
873 |
+
"PRON|Definite=Def|Gender=Masc|Number=Sing|Person=3|PronType=Prs|obl:npmod": 332,
|
874 |
+
"PRON|Definite=Def|Gender=Masc|Number=Sing|Person=3|PronType=Prs|parataxis": 333,
|
875 |
+
"PRON|Definite=Def|Gender=Masc|Number=Sing|Person=3|PronType=Prs|root": 334,
|
876 |
+
"PRON|Definite=Def|Number=Plur|Person=1|PronType=Prs|acl:relcl": 335,
|
877 |
+
"PRON|Definite=Def|Number=Plur|Person=1|PronType=Prs|conj": 336,
|
878 |
+
"PRON|Definite=Def|Number=Plur|Person=1|PronType=Prs|dislocated": 337,
|
879 |
+
"PRON|Definite=Def|Number=Plur|Person=1|PronType=Prs|iobj": 338,
|
880 |
+
"PRON|Definite=Def|Number=Plur|Person=1|PronType=Prs|nmod": 339,
|
881 |
+
"PRON|Definite=Def|Number=Plur|Person=1|PronType=Prs|nsubj": 340,
|
882 |
+
"PRON|Definite=Def|Number=Plur|Person=1|PronType=Prs|obj": 341,
|
883 |
+
"PRON|Definite=Def|Number=Plur|Person=1|PronType=Prs|obl": 342,
|
884 |
+
"PRON|Definite=Def|Number=Plur|Person=1|PronType=Prs|obl:npmod": 343,
|
885 |
+
"PRON|Definite=Def|Number=Plur|Person=1|PronType=Prs|reparandum": 344,
|
886 |
+
"PRON|Definite=Def|Number=Plur|Person=1|PronType=Prs|root": 345,
|
887 |
+
"PRON|Definite=Def|Number=Plur|Person=2|PronType=Prs|acl:relcl": 346,
|
888 |
+
"PRON|Definite=Def|Number=Plur|Person=2|PronType=Prs|advcl": 347,
|
889 |
+
"PRON|Definite=Def|Number=Plur|Person=2|PronType=Prs|ccomp": 348,
|
890 |
+
"PRON|Definite=Def|Number=Plur|Person=2|PronType=Prs|conj": 349,
|
891 |
+
"PRON|Definite=Def|Number=Plur|Person=2|PronType=Prs|dislocated": 350,
|
892 |
+
"PRON|Definite=Def|Number=Plur|Person=2|PronType=Prs|iobj": 351,
|
893 |
+
"PRON|Definite=Def|Number=Plur|Person=2|PronType=Prs|nmod": 352,
|
894 |
+
"PRON|Definite=Def|Number=Plur|Person=2|PronType=Prs|nsubj": 353,
|
895 |
+
"PRON|Definite=Def|Number=Plur|Person=2|PronType=Prs|obj": 354,
|
896 |
+
"PRON|Definite=Def|Number=Plur|Person=2|PronType=Prs|obl": 355,
|
897 |
+
"PRON|Definite=Def|Number=Plur|Person=2|PronType=Prs|obl:npmod": 356,
|
898 |
+
"PRON|Definite=Def|Number=Plur|Person=2|PronType=Prs|vocative": 357,
|
899 |
+
"PRON|Definite=Def|Number=Plur|Person=3|PronType=Prs|acl:relcl": 358,
|
900 |
+
"PRON|Definite=Def|Number=Plur|Person=3|PronType=Prs|advcl": 359,
|
901 |
+
"PRON|Definite=Def|Number=Plur|Person=3|PronType=Prs|compound": 360,
|
902 |
+
"PRON|Definite=Def|Number=Plur|Person=3|PronType=Prs|conj": 361,
|
903 |
+
"PRON|Definite=Def|Number=Plur|Person=3|PronType=Prs|dislocated": 362,
|
904 |
+
"PRON|Definite=Def|Number=Plur|Person=3|PronType=Prs|iobj": 363,
|
905 |
+
"PRON|Definite=Def|Number=Plur|Person=3|PronType=Prs|nmod": 364,
|
906 |
+
"PRON|Definite=Def|Number=Plur|Person=3|PronType=Prs|nsubj": 365,
|
907 |
+
"PRON|Definite=Def|Number=Plur|Person=3|PronType=Prs|obj": 366,
|
908 |
+
"PRON|Definite=Def|Number=Plur|Person=3|PronType=Prs|obl": 367,
|
909 |
+
"PRON|Definite=Def|Number=Plur|Person=3|PronType=Prs|obl:npmod": 368,
|
910 |
+
"PRON|Definite=Def|Number=Plur|Person=3|PronType=Prs|orphan": 369,
|
911 |
+
"PRON|Definite=Def|Number=Plur|Person=3|PronType=Prs|root": 370,
|
912 |
+
"PRON|Definite=Def|Number=Sing|Person=1|PronType=Prs|acl:relcl": 371,
|
913 |
+
"PRON|Definite=Def|Number=Sing|Person=1|PronType=Prs|ccomp": 372,
|
914 |
+
"PRON|Definite=Def|Number=Sing|Person=1|PronType=Prs|conj": 373,
|
915 |
+
"PRON|Definite=Def|Number=Sing|Person=1|PronType=Prs|dislocated": 374,
|
916 |
+
"PRON|Definite=Def|Number=Sing|Person=1|PronType=Prs|iobj": 375,
|
917 |
+
"PRON|Definite=Def|Number=Sing|Person=1|PronType=Prs|nmod": 376,
|
918 |
+
"PRON|Definite=Def|Number=Sing|Person=1|PronType=Prs|nsubj": 377,
|
919 |
+
"PRON|Definite=Def|Number=Sing|Person=1|PronType=Prs|obj": 378,
|
920 |
+
"PRON|Definite=Def|Number=Sing|Person=1|PronType=Prs|obl": 379,
|
921 |
+
"PRON|Definite=Def|Number=Sing|Person=1|PronType=Prs|obl:npmod": 380,
|
922 |
+
"PRON|Definite=Def|Number=Sing|Person=1|PronType=Prs|parataxis": 381,
|
923 |
+
"PRON|Definite=Def|Number=Sing|Person=1|PronType=Prs|root": 382,
|
924 |
+
"PRON|Definite=Def|Poss=Yes|PronType=Prs|det": 383,
|
925 |
+
"PRON|Gender=Fem|Number=Sing|cop": 384,
|
926 |
+
"PRON|Gender=Masc|Number=Sing|cop": 385,
|
927 |
+
"PRON|Number=Plur|cop": 386,
|
928 |
+
"PRON|Number=Plur|reparandum": 387,
|
929 |
+
"PRON|PronType=Ind|det": 388,
|
930 |
+
"PRON|PronType=Int|advcl": 389,
|
931 |
+
"PRON|PronType=Int|ccomp": 390,
|
932 |
+
"PRON|PronType=Int|conj": 391,
|
933 |
+
"PRON|PronType=Int|nmod": 392,
|
934 |
+
"PRON|PronType=Int|nsubj": 393,
|
935 |
+
"PRON|PronType=Int|obj": 394,
|
936 |
+
"PRON|PronType=Int|obl": 395,
|
937 |
+
"PRON|PronType=Int|obl:npmod": 396,
|
938 |
+
"PRON|PronType=Int|parataxis": 397,
|
939 |
+
"PRON|PronType=Int|root": 398,
|
940 |
+
"PRON|VerbForm=Fin|acl:relcl": 399,
|
941 |
+
"PRON|VerbForm=Fin|xcomp": 400,
|
942 |
+
"PRON|_|det": 401,
|
943 |
+
"PRON|_|nsubj": 402,
|
944 |
+
"PRON|_|obj": 403,
|
945 |
+
"PRON|_|obl": 404,
|
946 |
+
"PRON|_|root": 405,
|
947 |
+
"PROPN|Foreign=Yes|acl:relcl": 406,
|
948 |
+
"PROPN|Foreign=Yes|advcl": 407,
|
949 |
+
"PROPN|Foreign=Yes|appos": 408,
|
950 |
+
"PROPN|Foreign=Yes|ccomp": 409,
|
951 |
+
"PROPN|Foreign=Yes|conj": 410,
|
952 |
+
"PROPN|Foreign=Yes|dislocated": 411,
|
953 |
+
"PROPN|Foreign=Yes|flat": 412,
|
954 |
+
"PROPN|Foreign=Yes|nmod": 413,
|
955 |
+
"PROPN|Foreign=Yes|nsubj": 414,
|
956 |
+
"PROPN|Foreign=Yes|obj": 415,
|
957 |
+
"PROPN|Foreign=Yes|obl": 416,
|
958 |
+
"PROPN|Foreign=Yes|orphan": 417,
|
959 |
+
"PROPN|Foreign=Yes|parataxis": 418,
|
960 |
+
"PROPN|Foreign=Yes|root": 419,
|
961 |
+
"PROPN|Foreign=Yes|vocative": 420,
|
962 |
+
"PROPN|Foreign=Yes|xcomp": 421,
|
963 |
+
"PROPN|_|appos": 422,
|
964 |
+
"PROPN|_|ccomp": 423,
|
965 |
+
"PROPN|_|compound": 424,
|
966 |
+
"PROPN|_|dislocated": 425,
|
967 |
+
"PROPN|_|flat": 426,
|
968 |
+
"PROPN|_|nmod": 427,
|
969 |
+
"PROPN|_|nsubj": 428,
|
970 |
+
"PROPN|_|obl": 429,
|
971 |
+
"PROPN|_|root": 430,
|
972 |
+
"PROPN|_|vocative": 431,
|
973 |
+
"PUNCT|_|punct": 432,
|
974 |
+
"SCONJ|Foreign=Yes|Polarity=Neg|mark": 433,
|
975 |
+
"SCONJ|Foreign=Yes|mark": 434,
|
976 |
+
"SCONJ|_|case": 435,
|
977 |
+
"SCONJ|_|mark": 436,
|
978 |
+
"SCONJ|_|reparandum": 437,
|
979 |
+
"VERB|Foreign=Yes|VerbForm=Fin|acl:relcl": 438,
|
980 |
+
"VERB|Foreign=Yes|VerbForm=Fin|advcl": 439,
|
981 |
+
"VERB|Foreign=Yes|VerbForm=Fin|ccomp": 440,
|
982 |
+
"VERB|Foreign=Yes|VerbForm=Fin|conj": 441,
|
983 |
+
"VERB|Foreign=Yes|VerbForm=Fin|csubj": 442,
|
984 |
+
"VERB|Foreign=Yes|VerbForm=Fin|nmod": 443,
|
985 |
+
"VERB|Foreign=Yes|VerbForm=Fin|obl": 444,
|
986 |
+
"VERB|Foreign=Yes|VerbForm=Fin|parataxis": 445,
|
987 |
+
"VERB|Foreign=Yes|VerbForm=Fin|root": 446,
|
988 |
+
"VERB|Foreign=Yes|VerbForm=Fin|xcomp": 447,
|
989 |
+
"VERB|Foreign=Yes|VerbForm=Inf|conj": 448,
|
990 |
+
"VERB|Foreign=Yes|VerbForm=Inf|csubj": 449,
|
991 |
+
"VERB|Foreign=Yes|VerbForm=Inf|obl": 450,
|
992 |
+
"VERB|Foreign=Yes|VerbForm=Inf|xcomp": 451,
|
993 |
+
"VERB|Foreign=Yes|advcl": 452,
|
994 |
+
"VERB|Foreign=Yes|ccomp": 453,
|
995 |
+
"VERB|Polarity=Neg|acl:relcl": 454,
|
996 |
+
"VERB|Polarity=Neg|advcl": 455,
|
997 |
+
"VERB|Polarity=Neg|ccomp": 456,
|
998 |
+
"VERB|Polarity=Neg|conj": 457,
|
999 |
+
"VERB|Polarity=Neg|csubj": 458,
|
1000 |
+
"VERB|Polarity=Neg|orphan": 459,
|
1001 |
+
"VERB|Polarity=Neg|parataxis": 460,
|
1002 |
+
"VERB|Polarity=Neg|root": 461,
|
1003 |
+
"VERB|VerbForm=Fin|acl": 462,
|
1004 |
+
"VERB|VerbForm=Fin|acl:relcl": 463,
|
1005 |
+
"VERB|VerbForm=Fin|advcl": 464,
|
1006 |
+
"VERB|VerbForm=Fin|ccomp": 465,
|
1007 |
+
"VERB|VerbForm=Fin|compound": 466,
|
1008 |
+
"VERB|VerbForm=Fin|conj": 467,
|
1009 |
+
"VERB|VerbForm=Fin|csubj": 468,
|
1010 |
+
"VERB|VerbForm=Fin|dep": 469,
|
1011 |
+
"VERB|VerbForm=Fin|discourse": 470,
|
1012 |
+
"VERB|VerbForm=Fin|dislocated": 471,
|
1013 |
+
"VERB|VerbForm=Fin|nmod": 472,
|
1014 |
+
"VERB|VerbForm=Fin|nsubj": 473,
|
1015 |
+
"VERB|VerbForm=Fin|obj": 474,
|
1016 |
+
"VERB|VerbForm=Fin|obl": 475,
|
1017 |
+
"VERB|VerbForm=Fin|orphan": 476,
|
1018 |
+
"VERB|VerbForm=Fin|parataxis": 477,
|
1019 |
+
"VERB|VerbForm=Fin|root": 478,
|
1020 |
+
"VERB|VerbForm=Fin|xcomp": 479,
|
1021 |
+
"VERB|VerbForm=Inf|acl": 480,
|
1022 |
+
"VERB|VerbForm=Inf|advcl": 481,
|
1023 |
+
"VERB|VerbForm=Inf|ccomp": 482,
|
1024 |
+
"VERB|VerbForm=Inf|conj": 483,
|
1025 |
+
"VERB|VerbForm=Inf|csubj": 484,
|
1026 |
+
"VERB|VerbForm=Inf|dislocated": 485,
|
1027 |
+
"VERB|VerbForm=Inf|obl": 486,
|
1028 |
+
"VERB|VerbForm=Inf|root": 487,
|
1029 |
+
"VERB|VerbForm=Inf|xcomp": 488,
|
1030 |
+
"VERB|_|acl": 489,
|
1031 |
+
"VERB|_|acl:relcl": 490,
|
1032 |
+
"VERB|_|advcl": 491,
|
1033 |
+
"VERB|_|ccomp": 492,
|
1034 |
+
"VERB|_|conj": 493,
|
1035 |
+
"VERB|_|csubj": 494,
|
1036 |
+
"VERB|_|dislocated": 495,
|
1037 |
+
"VERB|_|nmod": 496,
|
1038 |
+
"VERB|_|obl": 497,
|
1039 |
+
"VERB|_|parataxis": 498,
|
1040 |
+
"VERB|_|root": 499,
|
1041 |
+
"VERB|_|xcomp": 500,
|
1042 |
+
"X|Foreign=Yes|Polarity=Neg|advmod": 501,
|
1043 |
+
"X|Foreign=Yes|appos": 502,
|
1044 |
+
"X|Foreign=Yes|ccomp": 503,
|
1045 |
+
"X|Foreign=Yes|dep": 504,
|
1046 |
+
"X|Foreign=Yes|fixed": 505,
|
1047 |
+
"X|Foreign=Yes|obl": 506,
|
1048 |
+
"X|Foreign=Yes|root": 507,
|
1049 |
+
"X|Polarity=Neg|advmod": 508,
|
1050 |
+
"X|Polarity=Neg|amod": 509,
|
1051 |
+
"X|_|advcl": 510,
|
1052 |
+
"X|_|conj": 511,
|
1053 |
+
"X|_|dep": 512,
|
1054 |
+
"X|_|goeswith": 513,
|
1055 |
+
"X|_|nmod": 514,
|
1056 |
+
"X|_|obl": 515,
|
1057 |
+
"X|_|parataxis": 516,
|
1058 |
+
"X|_|reparandum": 517,
|
1059 |
+
"X|_|root": 518
|
1060 |
+
},
|
1061 |
+
"layer_norm_eps": 1e-12,
|
1062 |
+
"max_position_embeddings": 512,
|
1063 |
+
"model_type": "roberta",
|
1064 |
+
"num_attention_heads": 12,
|
1065 |
+
"num_hidden_layers": 12,
|
1066 |
+
"pad_token_id": 1,
|
1067 |
+
"position_embedding_type": "absolute",
|
1068 |
+
"tokenizer_class": "RemBertTokenizerFast",
|
1069 |
+
"torch_dtype": "float32",
|
1070 |
+
"transformers_version": "4.22.1",
|
1071 |
+
"type_vocab_size": 2,
|
1072 |
+
"use_cache": true,
|
1073 |
+
"vocab_size": 3005
|
1074 |
+
}
|
maker.py
ADDED
@@ -0,0 +1,55 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
#! /usr/bin/python3
|
2 |
+
src="KoichiYasuoka/roberta-base-coptic"
|
3 |
+
tgt="KoichiYasuoka/roberta-base-coptic-ud-goeswith"
|
4 |
+
url="https://github.com/UniversalDependencies/UD_Coptic-Scriptorium"
|
5 |
+
import os
|
6 |
+
d=os.path.basename(url)
|
7 |
+
os.system("test -d "+d+" || git clone --depth=1 "+url)
|
8 |
+
os.system("for F in train dev test ; do cp "+d+"/*-$F.conllu $F.conllu ; done")
|
9 |
+
class UDgoeswithDataset(object):
|
10 |
+
def __init__(self,conllu,tokenizer):
|
11 |
+
self.ids,self.tags,label=[],[],set()
|
12 |
+
with open(conllu,"r",encoding="utf-8") as r:
|
13 |
+
cls,sep,msk=tokenizer.cls_token_id,tokenizer.sep_token_id,tokenizer.mask_token_id
|
14 |
+
dep,c="-|_|dep",[]
|
15 |
+
for s in r:
|
16 |
+
t=s.split("\t")
|
17 |
+
if len(t)==10:
|
18 |
+
if t[0].isdecimal():
|
19 |
+
c.append(t)
|
20 |
+
elif c!=[]:
|
21 |
+
v=tokenizer([t[1] for t in c],add_special_tokens=False)["input_ids"]
|
22 |
+
for i in range(len(v)-1,-1,-1):
|
23 |
+
for j in range(1,len(v[i])):
|
24 |
+
c.insert(i+1,[c[i][0],"_","_","X","_","_",c[i][0],"goeswith","_","_"])
|
25 |
+
y=["0"]+[t[0] for t in c]
|
26 |
+
h=[i if t[6]=="0" else y.index(t[6]) for i,t in enumerate(c,1)]
|
27 |
+
p,v=[t[3]+"|"+t[5]+"|"+t[7] for t in c],sum(v,[])
|
28 |
+
self.ids.append([cls]+v+[sep])
|
29 |
+
self.tags.append([dep]+p+[dep])
|
30 |
+
label=set(sum([self.tags[-1],list(label)],[]))
|
31 |
+
for i,k in enumerate(v):
|
32 |
+
self.ids.append([cls]+v[0:i]+[msk]+v[i+1:]+[sep,k])
|
33 |
+
self.tags.append([dep]+[t if h[j]==i+1 else dep for j,t in enumerate(p)]+[dep,dep])
|
34 |
+
c=[]
|
35 |
+
self.label2id={l:i for i,l in enumerate(sorted(label))}
|
36 |
+
def __call__(*args):
|
37 |
+
label=set(sum([list(t.label2id) for t in args],[]))
|
38 |
+
lid={l:i for i,l in enumerate(sorted(label))}
|
39 |
+
for t in args:
|
40 |
+
t.label2id=lid
|
41 |
+
return lid
|
42 |
+
__len__=lambda self:len(self.ids)
|
43 |
+
__getitem__=lambda self,i:{"input_ids":self.ids[i],"labels":[self.label2id[t] for t in self.tags[i]]}
|
44 |
+
from transformers import AutoTokenizer,AutoConfig,AutoModelForTokenClassification,DataCollatorForTokenClassification,TrainingArguments,Trainer
|
45 |
+
tkz=AutoTokenizer.from_pretrained(src)
|
46 |
+
trainDS=UDgoeswithDataset("train.conllu",tkz)
|
47 |
+
devDS=UDgoeswithDataset("dev.conllu",tkz)
|
48 |
+
testDS=UDgoeswithDataset("test.conllu",tkz)
|
49 |
+
lid=trainDS(devDS,testDS)
|
50 |
+
cfg=AutoConfig.from_pretrained(src,num_labels=len(lid),label2id=lid,id2label={i:l for l,i in lid.items()})
|
51 |
+
arg=TrainingArguments(num_train_epochs=3,per_device_train_batch_size=32,output_dir="/tmp",overwrite_output_dir=True,save_total_limit=2,evaluation_strategy="epoch",learning_rate=5e-05,warmup_ratio=0.1)
|
52 |
+
trn=Trainer(args=arg,data_collator=DataCollatorForTokenClassification(tkz),model=AutoModelForTokenClassification.from_pretrained(src,config=cfg),train_dataset=trainDS,eval_dataset=devDS)
|
53 |
+
trn.train()
|
54 |
+
trn.save_model(tgt)
|
55 |
+
tkz.save_pretrained(tgt)
|
pytorch_model.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:4a2d77f44faa7c3328fc643829a9180a0697c416c8eb7feea19aadbc66c5bb0f
|
3 |
+
size 352701809
|
sentencepiece.model
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:01ba4719c80b6fe911b091a7c05124b64eeece964e09c058ef8f9805daca546b
|
3 |
+
size 1
|
special_tokens_map.json
ADDED
@@ -0,0 +1,15 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"bos_token": "[CLS]",
|
3 |
+
"cls_token": "[CLS]",
|
4 |
+
"eos_token": "[SEP]",
|
5 |
+
"mask_token": {
|
6 |
+
"content": "[MASK]",
|
7 |
+
"lstrip": true,
|
8 |
+
"normalized": true,
|
9 |
+
"rstrip": false,
|
10 |
+
"single_word": false
|
11 |
+
},
|
12 |
+
"pad_token": "[PAD]",
|
13 |
+
"sep_token": "[SEP]",
|
14 |
+
"unk_token": "[UNK]"
|
15 |
+
}
|
tokenizer.json
ADDED
The diff for this file is too large to render.
See raw diff
|
|
tokenizer_config.json
ADDED
@@ -0,0 +1,21 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"bos_token": "[CLS]",
|
3 |
+
"cls_token": "[CLS]",
|
4 |
+
"do_lower_case": false,
|
5 |
+
"eos_token": "[SEP]",
|
6 |
+
"keep_accents": false,
|
7 |
+
"mask_token": {
|
8 |
+
"__type": "AddedToken",
|
9 |
+
"content": "[MASK]",
|
10 |
+
"lstrip": true,
|
11 |
+
"normalized": true,
|
12 |
+
"rstrip": false,
|
13 |
+
"single_word": false
|
14 |
+
},
|
15 |
+
"model_max_length": 512,
|
16 |
+
"pad_token": "[PAD]",
|
17 |
+
"remove_space": true,
|
18 |
+
"sep_token": "[SEP]",
|
19 |
+
"tokenizer_class": "RemBertTokenizerFast",
|
20 |
+
"unk_token": "[UNK]"
|
21 |
+
}
|
ud.py
ADDED
@@ -0,0 +1,63 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
from transformers import TokenClassificationPipeline
|
2 |
+
|
3 |
+
class UniversalDependenciesPipeline(TokenClassificationPipeline):
|
4 |
+
def _forward(self,model_input):
|
5 |
+
import torch
|
6 |
+
v=model_input["input_ids"][0].tolist()
|
7 |
+
with torch.no_grad():
|
8 |
+
e=self.model(input_ids=torch.tensor([v[0:i]+[self.tokenizer.mask_token_id]+v[i+1:]+[j] for i,j in enumerate(v[1:-1],1)]))
|
9 |
+
return {"logits":e.logits[:,1:-2,:],**model_input}
|
10 |
+
def postprocess(self,model_output,**kwargs):
|
11 |
+
import numpy
|
12 |
+
e=model_output["logits"].numpy()
|
13 |
+
r=[1 if i==0 else -1 if j.endswith("|root") else 0 for i,j in sorted(self.model.config.id2label.items())]
|
14 |
+
e+=numpy.where(numpy.add.outer(numpy.identity(e.shape[0]),r)==0,0,numpy.nan)
|
15 |
+
g=self.model.config.label2id["X|_|goeswith"]
|
16 |
+
r=numpy.tri(e.shape[0])
|
17 |
+
for i in range(e.shape[0]):
|
18 |
+
for j in range(i+2,e.shape[1]):
|
19 |
+
r[i,j]=r[i,j-1] if numpy.nanargmax(e[i,j-1])==g else 1
|
20 |
+
e[:,:,g]+=numpy.where(r==0,0,numpy.nan)
|
21 |
+
m,p=numpy.nanmax(e,axis=2),numpy.nanargmax(e,axis=2)
|
22 |
+
h=self.chu_liu_edmonds(m)
|
23 |
+
z=[i for i,j in enumerate(h) if i==j]
|
24 |
+
if len(z)>1:
|
25 |
+
k=z[numpy.nanargmax(m[z,z])]
|
26 |
+
m[:,z]+=[[0 if j in z and (i!=j or i==k) else numpy.nan for i in z] for j in range(m.shape[0])]
|
27 |
+
h=self.chu_liu_edmonds(m)
|
28 |
+
v=[(s,e) for s,e in model_output["offset_mapping"][0].tolist() if s<e]
|
29 |
+
q=[self.model.config.id2label[p[j,i]].split("|") for i,j in enumerate(h)]
|
30 |
+
if "aggregation_strategy" in kwargs and kwargs["aggregation_strategy"]!="none":
|
31 |
+
for i,j in reversed(list(enumerate(q[1:],1))):
|
32 |
+
if j[-1]=="goeswith" and set([t[-1] for t in q[h[i]+1:i+1]])=={"goeswith"}:
|
33 |
+
h=[b if i>b else b-1 for a,b in enumerate(h) if i!=a]
|
34 |
+
v[i-1]=(v[i-1][0],v.pop(i)[1])
|
35 |
+
q.pop(i)
|
36 |
+
t=model_output["sentence"].replace("\n"," ")
|
37 |
+
u="# text = "+t+"\n"
|
38 |
+
for i,(s,e) in enumerate(v):
|
39 |
+
u+="\t".join([str(i+1),t[s:e],"_",q[i][0],"_","|".join(q[i][1:-1]),str(0 if h[i]==i else h[i]+1),q[i][-1],"_","_" if i+1<len(v) and e<v[i+1][0] else "SpaceAfter=No"])+"\n"
|
40 |
+
return u+"\n"
|
41 |
+
def chu_liu_edmonds(self,matrix):
|
42 |
+
import numpy
|
43 |
+
h=numpy.nanargmax(matrix,axis=0)
|
44 |
+
x=[-1 if i==j else j for i,j in enumerate(h)]
|
45 |
+
for b in [lambda x,i,j:-1 if i not in x else x[i],lambda x,i,j:-1 if j<0 else x[j]]:
|
46 |
+
y=[]
|
47 |
+
while x!=y:
|
48 |
+
y=list(x)
|
49 |
+
for i,j in enumerate(x):
|
50 |
+
x[i]=b(x,i,j)
|
51 |
+
if max(x)<0:
|
52 |
+
return h
|
53 |
+
y=[i for i,j in enumerate(x) if j==max(x)]
|
54 |
+
m=numpy.full((matrix.shape[0]+1,matrix.shape[1]+1),numpy.nan)
|
55 |
+
m[0:-1,0:-1]=z=matrix-numpy.nanmax(matrix,axis=0)
|
56 |
+
m[0:-1,-1],m[-1,0:-1],m[-1,-1]=numpy.nanmax(z[:,y],axis=1),numpy.nanmax(z[y,:],axis=0),numpy.nanmax(z[y,y])
|
57 |
+
m[y,:]=m[:,y]=numpy.nan
|
58 |
+
m[y,y]=0
|
59 |
+
k=self.chu_liu_edmonds(m)
|
60 |
+
j=y[numpy.nanargmax(z[k[-1],y] if k[-1]<z.shape[0] else z[y,y])]
|
61 |
+
i=k[-1] if k[-1]<z.shape[0] else j
|
62 |
+
z[0:i,j]=z[i+1:,j]=numpy.nan
|
63 |
+
return self.chu_liu_edmonds(z)
|