forked from kubapok/en-ner-conll-2003
102 lines
3.5 KiB
Python
102 lines
3.5 KiB
Python
![]() |
# imports
|
||
|
import torch
|
||
|
import pandas as pd
|
||
|
import csv
|
||
|
from torchtext.vocab import Vocab
|
||
|
from collections import Counter
|
||
|
|
||
|
class NERModel(torch.nn.Module):
|
||
|
def __init__(self,):
|
||
|
super(NERModel, self).__init__()
|
||
|
self.emb = torch.nn.Embedding(23627,200)
|
||
|
self.fc1 = torch.nn.Linear(600,9)
|
||
|
def forward(self, x):
|
||
|
x = self.emb(x)
|
||
|
x = x.reshape(600)
|
||
|
x = self.fc1(x)
|
||
|
return x
|
||
|
|
||
|
class NeuralNetworkModel(torch.nn.Module):
|
||
|
def __init__(self, output_size):
|
||
|
super(NeuralNetworkModel, self).__init__()
|
||
|
self.fc1 = torch.nn.Linear(10_000, output_size)
|
||
|
self.softmax = torch.nn.Softmax(dim=0)
|
||
|
def forward(self, x):
|
||
|
x = self.fc1(x)
|
||
|
x = self.softmax(x)
|
||
|
return x
|
||
|
|
||
|
def build_vocab(dataset):
|
||
|
counter = Counter()
|
||
|
for document in dataset:
|
||
|
counter.update(document)
|
||
|
return Vocab(counter, specials=['<unk>', '<pad>', '<bos>', '<eos>'])
|
||
|
|
||
|
def data_process(dt):
|
||
|
return [torch.tensor([vocab['<bos>']] + [vocab[token] for token in document] + [vocab['<eos>']], dtype=torch.long)
|
||
|
for document in dt]
|
||
|
|
||
|
def labels_process(dt):
|
||
|
return [torch.tensor([0] + document + [0], dtype=torch.long) for document in dt]
|
||
|
|
||
|
|
||
|
LABELS = ['O','B-LOC', 'I-LOC','B-MISC', 'I-MISC', 'B-ORG', 'I-ORG', 'B-PER', 'I-PER']
|
||
|
train = pd.read_csv("./train/train.tsv.xz", error_bad_lines=False, compression='xz', sep='\t', header=None, quoting=csv.QUOTE_NONE)
|
||
|
dev = pd.read_csv('./dev-0/in.tsv', error_bad_lines=False, sep='\t', header=None, quoting=csv.QUOTE_NONE)
|
||
|
test = pd.read_csv('./test-A/in.tsv', error_bad_lines=False, sep='\t', header=None, quoting=csv.QUOTE_NONE)
|
||
|
|
||
|
tags = train[0].apply(lambda x: [LABELS.index(y) for y in x.split()])
|
||
|
tokens = train[1].apply(lambda x: x.split())
|
||
|
dev_tokens = dev[0].apply(lambda x: x.split())
|
||
|
test_tokens = dev[0].apply(lambda x: x.split())
|
||
|
|
||
|
vocab = build_vocab(tokens)
|
||
|
train_labels = labels_process(tags)
|
||
|
train_tokens_ids = data_process(tokens)
|
||
|
|
||
|
ner_model = NERModel()
|
||
|
nn_model = NeuralNetworkModel(len(train_tokens_ids))
|
||
|
criterion = torch.nn.CrossEntropyLoss()
|
||
|
optimizer = torch.optim.Adam(ner_model.parameters())
|
||
|
|
||
|
for epoch in range(2):
|
||
|
loss_score = 0
|
||
|
acc_score = 0
|
||
|
prec_score = 0
|
||
|
selected_items = 0
|
||
|
recall_score = 0
|
||
|
relevant_items = 0
|
||
|
items_total = 0
|
||
|
nn_model.train()
|
||
|
for i in range(100):
|
||
|
for j in range(1, len(train_labels[i]) - 1):
|
||
|
X = train_tokens_ids[i][j-1: j+2]
|
||
|
Y = train_labels[i][j: j+1]
|
||
|
Y_predictions = ner_model(X)
|
||
|
acc_score += int(torch.argmax(Y_predictions) == Y)
|
||
|
|
||
|
if torch.argmax(Y_predictions) != 0:
|
||
|
selected_items +=1
|
||
|
if torch.argmax(Y_predictions) != 0 and torch.argmax(Y_predictions) == Y.item():
|
||
|
prec_score += 1
|
||
|
|
||
|
if Y.item() != 0:
|
||
|
relevant_items +=1
|
||
|
if Y.item() != 0 and torch.argmax(Y_predictions) == Y.item():
|
||
|
recall_score += 1
|
||
|
items_total += 1
|
||
|
optimizer.zero_grad()
|
||
|
loss = criterion(Y_predictions.unsqueeze(0), Y)
|
||
|
loss.backward()
|
||
|
optimizer.step()
|
||
|
loss_score += loss.item()
|
||
|
|
||
|
precision = prec_score / selected_items
|
||
|
recall = recall_score / relevant_items
|
||
|
f1_score = (2*precision * recall) / (precision + recall)
|
||
|
print('epoch: ', epoch)
|
||
|
print('loss: ', loss_score / items_total)
|
||
|
print('acc: ', acc_score / items_total)
|
||
|
print('prec: ', precision)
|
||
|
print('recall: : ', recall)
|
||
|
print('f1: ', f1_score)
|