ud.py

import numpy
from transformers import TokenClassificationPipeline

class UniversalDependenciesPipeline(TokenClassificationPipeline):
  def __init__(self,**kwargs):
    super().__init__(**kwargs)
    x=self.model.config.label2id
    self.root=numpy.full((len(x)),-numpy.inf)
    self.arc=numpy.full((len(x)),-numpy.inf)
    for k,v in x.items():
      if k.endswith("|[root]"):
        self.root[v]=0
      elif k.endswith("]"):
        self.arc[v]=0
  def _forward(self,model_inputs):
    import torch
    v=model_inputs["input_ids"][0].tolist()
    if len(v)<91:
      x=[True]*(len(v)-2)
    else:
      with torch.no_grad():
        e=self.model(input_ids=torch.tensor([v]).to(self.device))
      m=e.logits[0].cpu().numpy()
      e=numpy.exp(m-numpy.max(m,axis=-1,keepdims=True))
      z=e/e.sum(axis=-1,keepdims=True)
      k=numpy.argmax(m,axis=1).tolist()
      x=[not self.model.config.id2label[p].split("|")[0].endswith(".") for p in k[1:-1]]
      w=(sum([1 for b in x if b])+1)*(len(x)+1)+1
      for i in numpy.argsort([z[i+1,k[i+1]] for i in range(len(x))]):
        if w+len(x)>8191:
          break
        if not x[i]:
          x[i]=True
          w+=len(x)+1
    ids=list(v)
    for i in range(len(x)):
      if x[i]:
        ids+=v[1:i+1]+[self.tokenizer.mask_token_id]+v[i+2:]
    with torch.no_grad():
      e=self.model(input_ids=torch.tensor([ids]).to(self.device))
    return {"logits":e.logits,"thin_out":x,**model_inputs}
  def check_model_type(self,supported_models):
    pass
  def postprocess(self,model_outputs,**kwargs):
    if "logits" not in model_outputs:
      return "".join(self.postprocess(x,**kwargs) for x in model_outputs)
    m=model_outputs["logits"][0].cpu().numpy()
    x=model_outputs["thin_out"]
    e=numpy.full((len(x),len(x),m.shape[-1]),m.min())
    k=len(x)+2
    for i in range(len(x)):
      if x[i]:
        for j in range(len(x)):
          if i==j:
            e[i,i]=m[k]+self.root
          else:
            e[i,j]=m[k]+self.arc
          k+=1
        k+=1
    g=self.model.config.label2id["X.|[goeswith]"]
    m,r=numpy.max(e,axis=2),numpy.tri(e.shape[0])
    for i in range(e.shape[0]):
      for j in range(i+2,e.shape[1]):
        r[i,j]=1
        if numpy.argmax(e[i,j-1])==g:
          if numpy.argmax(m[:,j-1])==i:
            r[i,j]=r[i,j-1]
    e[:,:,g]+=numpy.where(r==0,0,-numpy.inf)
    m,p=numpy.max(e,axis=2),numpy.argmax(e,axis=2)
    h=self.chu_liu_edmonds(m)
    z=[i for i,j in enumerate(h) if i==j]
    if len(z)>1:
      k,h=z[numpy.argmax(m[z,z])],numpy.min(m)-numpy.max(m)
      m[:,z]+=[[0 if j in z and (i!=j or i==k) else h for i in z] for j in range(m.shape[0])]
      h=self.chu_liu_edmonds(m)
    v=[(s,e) for s,e in model_outputs["offset_mapping"][0].tolist() if s<e]
    q=[self.model.config.id2label[p[j,i]].split("|") for i,j in enumerate(h)]
    if "aggregation_strategy" in kwargs and kwargs["aggregation_strategy"]!="none":
      for i,j in reversed(list(enumerate(q[1:],1))):
        if j[-1]=="[goeswith]" and set([t[-1] for t in q[h[i]+1:i+1]])=={"[goeswith]"}:
          h=[b if i>b else b-1 for a,b in enumerate(h) if i!=a]
          v[i-1]=(v[i-1][0],v.pop(i)[1])
          q.pop(i)
        elif v[i-1][1]>v[i][0]:
          h=[b if i>b else b-1 for a,b in enumerate(h) if i!=a]
          v[i-1]=(v[i-1][0],v.pop(i)[1])
          q.pop(i)
    t=model_outputs["sentence"].replace("\n"," ")
    for i,(s,e) in reversed(list(enumerate(v))):
      d=t[s:e]
      j=len(d)-len(d.lstrip())
      if j>0:
        d=d.lstrip()
        v[i]=(v[i][0]+j,v[i][1])
      j=len(d)-len(d.rstrip())
      if j>0:
        d=d.rstrip()
        v[i]=(v[i][0],v[i][1]-j)
      if d.strip()=="":
        h=[b if i>b else b-1 for a,b in enumerate(h) if i!=a]
        v.pop(i)
        q.pop(i)
    u="# text = "+t+"\n"
    for i,(s,e) in enumerate(v):
      u+="\t".join([str(i+1),t[s:e],t[s:e],q[i][0].replace(".",""),"_","_" if len(q[i])<3 else "|".join(q[i][1:-1]),str(0 if h[i]==i else h[i]+1),q[i][-1][1:-1],"_","_" if i+1<len(v) and e<v[i+1][0] else "SpaceAfter=No"])+"\n"
    return u+"\n"
  def chu_liu_edmonds(self,matrix):
    import numpy
    h=numpy.argmax(matrix,axis=0)
    x=[-1 if i==j else j for i,j in enumerate(h)]
    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]]:
      y=[]
      while x!=y:
        y=list(x)
        for i,j in enumerate(x):
          x[i]=b(x,i,j)
      if max(x)<0:
        return h
    y,x=[i for i,j in enumerate(x) if j==max(x)],[i for i,j in enumerate(x) if j<max(x)]
    z=matrix-numpy.max(matrix,axis=0)
    m=numpy.block([[z[x,:][:,x],numpy.max(z[x,:][:,y],axis=1).reshape(len(x),1)],[numpy.max(z[y,:][:,x],axis=0),numpy.max(z[y,y])]])
    k=[j if i==len(x) else x[j] if j<len(x) else y[numpy.argmax(z[y,x[i]])] for i,j in enumerate(self.chu_liu_edmonds(m))]
    h=[j if i in y else k[x.index(i)] for i,j in enumerate(h)]
    i=y[numpy.argmax(z[x[k[-1]],y] if k[-1]<len(x) else z[y,y])]
    h[i]=x[k[-1]] if k[-1]<len(x) else i
    return h