Skip to content

Instantly share code, notes, and snippets.

View RamonYeung's full-sized avatar
🚀
Working on a rocket ticket !

杨海宏 RamonYeung

🚀
Working on a rocket ticket !
  • MIT, The Alibaba DAMO Academy
  • Hangzhou, China
View GitHub Profile
''' Script for downloading all GLUE data.
Note: for legal reasons, we are unable to host MRPC.
You can either use the version hosted by the SentEval team, which is already tokenized,
or you can download the original data from (https://download.microsoft.com/download/D/4/6/D46FF87A-F6B9-4252-AA8B-3604ED519838/MSRParaphraseCorpus.msi) and extract the data from it manually.
For Windows users, you can run the .msi file. For Mac and Linux users, consider an external library such as 'cabextract' (see below for an example).
You should then rename and place specific files in a folder (see below for an example).
mkdir MRPC
cabextract MSRParaphraseCorpus.msi -d MRPC
@RamonYeung
RamonYeung / loading_pretrained_models
Created July 12, 2018 07:46
PyTorch Loading Pre-trained Models
# 1. Directly Load a Pre-trained Model
# https://github.com/pytorch/vision/tree/master/torchvision/models
import torchvision.models as models
resnet50 = models.resnet50(pretrained=True)
# or
model = models.resnet50(pretrained=False)
# Maybe you want to modify the last fc layer?
resnet.fc = nn.Linear(2048, 2)
@RamonYeung
RamonYeung / BPE
Created June 27, 2019 08:33 — forked from ranihorev/BPE
Byte Pair Encoding example (Source: Sennrich et al. - https://arxiv.org/abs/1508.07909)
import re, collections
def get_stats(vocab):
pairs = collections.defaultdict(int)
for word, freq in vocab.items():
symbols = word.split()
for i in range(len(symbols)-1):
pairs[symbols[i],symbols[i+1]] += freq
return pairs
from graphviz import Digraph
import torch
from torch.autograd import Variable, Function
def iter_graph(root, callback):
queue = [root]
seen = set()
while queue:
fn = queue.pop()
if fn in seen:
# encoding: utf-8
import os
import pygame
font_file = '/System/Library/Fonts/PingFang.ttc'
chinese_dir = 'chinese'
if not os.path.exists(chinese_dir):
os.mkdir(chinese_dir)

tmux cheatsheet

As configured in my dotfiles.

start new:

tmux

start new with session name:

@RamonYeung
RamonYeung / tmux-cheatsheet.markdown
Created April 5, 2019 03:38 — forked from MohamedAlaa/tmux-cheatsheet.markdown
tmux shortcuts & cheatsheet

tmux shortcuts & cheatsheet

start new:

tmux

start new with session name:

tmux new -s myname
@RamonYeung
RamonYeung / pad_packed_demo.py
Created February 16, 2019 17:09 — forked from Tushar-N/pad_packed_demo.py
How to use pad_packed_sequence in pytorch
import torch
import torch.nn as nn
from torch.nn.utils.rnn import pack_padded_sequence, pad_packed_sequence
seqs = ['gigantic_string','tiny_str','medium_str']
# make <pad> idx 0
vocab = ['<pad>'] + sorted(set(''.join(seqs)))
# make model
@RamonYeung
RamonYeung / simplify_chinese_characters.py
Created July 4, 2018 04:20
Demo for Converting Traditional Chinese Characters to Simplified Chinese Characters.
# copy and paste two modules.
# https://raw.githubusercontent.com/skydark/nstools/master/zhtools/langconv.py
# https://raw.githubusercontent.com/skydark/nstools/master/zhtools/zh_wiki.py
from langconv import *
def Traditional2Simplified(sentence):
sentence = Converter('zh-hans').convert(sentence)
return sentence
# coding: utf-8
import logging
import re
from collections import Counter
import numpy as np
import torch
from sklearn.datasets import fetch_20newsgroups
from torch.autograd import Variable