howard-haowen

import pandas as pd
from sklearn.model_selection import train_test_split
from sklearn.metrics import accuracy_score, confusion_matrix
from sklearn.metrics import classification_report
from sklearn.naive_bayes import MultinomialNB
from sklearn.pipeline import Pipeline
from sklearn.feature_extraction.text import TfidfTransformer
from sklearn.feature_extraction.text import CountVectorizer, TfidfVectorizer

#Are the tags which we are using currently for tagging the sentences

howard-haowen

!pip install transformers 

# Taken from https://huggingface.co/transformers/model_doc/auto.html
#=====Ways to initiate a tokenizer and embedding model=====
### Three ways to initiate a tokenizer ###
from transformers import AutoTokenizer

# From official models hosted on HuggingFace (just the model name)
tokenizer = AutoTokenizer.from_pretrained('bert-base-chinese')

howard-haowen

howard-haowen

# Method 1
# 須先下載wqy-microhei.ttc，因中文顯示需做特殊處理
zhfont = matplotlib.font_manager.FontProperties(fname='/Users/youngmihuang/Downloads/wqy-microhei.ttc')

==================================
# Method 2

# install Korean font
!apt install fonts-nanum

howard-haowen

howard-haowen

# Reference source of the codes: https://github.com/ckiplab/ckiptagger. 

# Intall dependencies
## !pip install -U ckiptagger[gdown]
## !pip install -U ckiptagger[tfgpu] # for Tensorflow GPU

from ckiptagger import data_utils, WS, POS, NER

# Download model files to ./data.zip (2GB) and extract to ./data/
## data_utils.download_data_url("./") # iis-ckip (Option 1)

howard-haowen

#!pip install zhon

import re
import zhon.hanzi as hanzi
from string import punctuation as en_punc
from string import ascii_letters as roman_letters

zh_punc = hanzi.punctuation
punc_set = set(zh_punc).union(set(en_punc)) #puncs in both English and Chinese
punc_list = list(punc_set)

howard-haowen

import io
import pandas as pd
from sklearn.model_selection import train_test_split

# fasttext.train_unsupervised for word embeddings
def CreateTxt(df, text_col="text"):
  # Each cell in "text_col" contains a string of a single tokenized document, with tokens seperated by spaces  
  df[text_col].to_csv('output.txt', sep='\n', index=False)

def CreateVecAndMeta():

howard-haowen

# Reference: kstathou/acl-search-engine

# !pip install faiss-cpu --no-cache
# !pip install sentence_transformers

import faiss
import numpy as np
import pandas as pd 
import pickle
import torch

howard-haowen

# Taken from https://clay-atlas.com/blog/2020/06/30/pytorch-如何使用-hugging-face-所提供的-transformers-以-bert-為例/

# coding: utf-8
import torch
from transformers import AutoTokenizer, AutoModel
from keras.preprocessing.sequence import pad_sequences

# Tokenizer and Bert Model
MODEL_NAME = '' # e.g. 'bert-base-chinese'
tokenizer = AutoTokenizer.from_pretrained(MODEL_NAME)