diff --git a/pytorch/data.py b/pytorch/data.py
new file mode 100644
index 0000000..d1e186e
--- /dev/null
+++ b/pytorch/data.py
@@ -0,0 +1,87 @@
+#!/usr/bin/env python
+# -*- coding: utf-8 -*-
+"""
+@Author: Yue Wang
+@Contact: yuewangx@mit.edu
+@File: data.py
+@Time: 2018/10/13 6:21 PM
+"""
+
+
+import os
+import sys
+import glob
+import h5py
+import numpy as np
+from torch.utils.data import Dataset
+
+
+def download():
+    BASE_DIR = os.path.dirname(os.path.abspath(__file__))
+    DATA_DIR = os.path.join(BASE_DIR, 'data')
+    if not os.path.exists(DATA_DIR):
+        os.mkdir(DATA_DIR)
+    if not os.path.exists(os.path.join(DATA_DIR, 'modelnet40_ply_hdf5_2048')):
+        www = 'https://shapenet.cs.stanford.edu/media/modelnet40_ply_hdf5_2048.zip'
+        zipfile = os.path.basename(www)
+        os.system('wget %s; unzip %s' % (www, zipfile))
+        os.system('mv %s %s' % (zipfile[:-4], DATA_DIR))
+        os.system('rm %s' % (zipfile))
+
+
+def load_data(partition):
+    download()
+    BASE_DIR = os.path.dirname(os.path.abspath(__file__))
+    DATA_DIR = os.path.join(BASE_DIR, 'data')
+    all_data = []
+    all_label = []
+    for h5_name in glob.glob(os.path.join(DATA_DIR, 'modelnet40_ply_hdf5_2048', 'ply_data_%s*.h5'%partition)):
+        f = h5py.File(h5_name)
+        data = f['data'][:].astype('float32')
+        label = f['label'][:].astype('int64')
+        f.close()
+        all_data.append(data)
+        all_label.append(label)
+    all_data = np.concatenate(all_data, axis=0)
+    all_label = np.concatenate(all_label, axis=0)
+    return all_data, all_label
+
+
+def translate_pointcloud(pointcloud):
+    xyz1 = np.random.uniform(low=2./3., high=3./2., size=[3])
+    xyz2 = np.random.uniform(low=-0.2, high=0.2, size=[3])
+       
+    translated_pointcloud = np.add(np.multiply(pointcloud, xyz1), xyz2).astype('float32')
+    return translated_pointcloud
+
+
+def jitter_pointcloud(pointcloud, sigma=0.01, clip=0.02):
+    N, C = pointcloud.shape
+    pointcloud += np.clip(sigma * np.random.randn(N, C), -1*clip, clip)
+    return pointcloud
+
+
+class ModelNet40(Dataset):
+    def __init__(self, num_points, partition='train'):
+        self.data, self.label = load_data(partition)
+        self.num_points = num_points
+        self.partition = partition        
+
+    def __getitem__(self, item):
+        pointcloud = self.data[item][:self.num_points]
+        label = self.label[item]
+        if self.partition == 'train':
+            pointcloud = translate_pointcloud(pointcloud)
+            np.random.shuffle(pointcloud)
+        return pointcloud, label
+
+    def __len__(self):
+        return self.data.shape[0]
+
+
+if __name__ == '__main__':
+    train = ModelNet40(1024)
+    test = ModelNet40(1024, 'test')
+    for data, label in train:
+        print(data.shape)
+        print(label.shape)
diff --git a/pytorch/main.py b/pytorch/main.py
new file mode 100644
index 0000000..3f45f26
--- /dev/null
+++ b/pytorch/main.py
@@ -0,0 +1,227 @@
+#!/usr/bin/env python
+# -*- coding: utf-8 -*-
+"""
+@Author: Yue Wang
+@Contact: yuewangx@mit.edu
+@File: main.py
+@Time: 2018/10/13 10:39 PM
+"""
+
+
+from __future__ import print_function
+import os
+import argparse
+import torch
+import torch.nn as nn
+import torch.nn.functional as F
+import torch.optim as optim
+from torch.optim.lr_scheduler import CosineAnnealingLR
+from data import ModelNet40
+from model import PointNet, DGCNN
+import numpy as np
+from torch.utils.data import DataLoader
+from util import cal_loss, IOStream
+import sklearn.metrics as metrics
+
+
+def _init_():
+    if not os.path.exists('checkpoints'):
+        os.makedirs('checkpoints')
+    if not os.path.exists('checkpoints/'+args.exp_name):
+        os.makedirs('checkpoints/'+args.exp_name)
+    if not os.path.exists('checkpoints/'+args.exp_name+'/'+'models'):
+        os.makedirs('checkpoints/'+args.exp_name+'/'+'models')
+    os.system('cp main.py checkpoints'+'/'+args.exp_name+'/'+'main.py.backup')
+    os.system('cp model.py checkpoints' + '/' + args.exp_name + '/' + 'model.py.backup')
+    os.system('cp util.py checkpoints' + '/' + args.exp_name + '/' + 'util.py.backup')
+    os.system('cp data.py checkpoints' + '/' + args.exp_name + '/' + 'data.py.backup')
+
+def train(args, io):
+    train_loader = DataLoader(ModelNet40(partition='train', num_points=args.num_points), num_workers=8,
+                              batch_size=args.batch_size, shuffle=True, drop_last=True)
+    test_loader = DataLoader(ModelNet40(partition='test', num_points=args.num_points), num_workers=8,
+                             batch_size=args.test_batch_size, shuffle=True, drop_last=False)
+
+    device = torch.device("cuda" if args.cuda else "cpu")
+
+    #Try to load models
+    if args.model == 'pointnet':
+        model = PointNet(args).to(device)
+    elif args.model == 'dgcnn':
+        model = DGCNN(args).to(device)
+    else:
+        raise Exception("Not implemented")
+    print(str(model))
+
+    model = nn.DataParallel(model)
+    print("Let's use", torch.cuda.device_count(), "GPUs!")
+
+    if args.use_sgd:
+        print("Use SGD")
+        opt = optim.SGD(model.parameters(), lr=args.lr*100, momentum=args.momentum, weight_decay=1e-4)
+    else:
+        print("Use Adam")
+        opt = optim.Adam(model.parameters(), lr=args.lr, weight_decay=1e-4)
+
+    scheduler = CosineAnnealingLR(opt, args.epochs, eta_min=args.lr)
+    
+    criterion = cal_loss
+
+    best_test_acc = 0
+    for epoch in range(args.epochs):
+        scheduler.step()
+        ####################
+        # Train
+        ####################
+        train_loss = 0.0
+        count = 0.0
+        model.train()
+        train_pred = []
+        train_true = []
+        for data, label in train_loader:
+            data, label = data.to(device), label.to(device).squeeze()
+            data = data.permute(0, 2, 1)
+            batch_size = data.size()[0]
+            opt.zero_grad()
+            logits = model(data)
+            loss = criterion(logits, label)
+            loss.backward()
+            opt.step()
+            preds = logits.max(dim=1)[1]
+            count += batch_size
+            train_loss += loss.item() * batch_size
+            train_true.append(label.cpu().numpy())
+            train_pred.append(preds.detach().cpu().numpy())
+        train_true = np.concatenate(train_true)
+        train_pred = np.concatenate(train_pred)
+        outstr = 'Train %d, loss: %.6f, train acc: %.6f, train avg acc: %.6f' % (epoch,
+                                                                                 train_loss*1.0/count,
+                                                                                 metrics.accuracy_score(
+                                                                                     train_true, train_pred),
+                                                                                 metrics.balanced_accuracy_score(
+                                                                                     train_true, train_pred))
+        io.cprint(outstr)
+
+        ####################
+        # Test
+        ####################
+        test_loss = 0.0
+        count = 0.0
+        model.eval()
+        test_pred = []
+        test_true = []
+        for data, label in test_loader:
+            data, label = data.to(device), label.to(device).squeeze()
+            data = data.permute(0, 2, 1)
+            batch_size = data.size()[0]
+            logits = model(data)
+            loss = criterion(logits, label)
+            preds = logits.max(dim=1)[1]
+            count += batch_size
+            test_loss += loss.item() * batch_size
+            test_true.append(label.cpu().numpy())
+            test_pred.append(preds.detach().cpu().numpy())
+        test_true = np.concatenate(test_true)
+        test_pred = np.concatenate(test_pred)
+        test_acc = metrics.accuracy_score(test_true, test_pred)
+        avg_per_class_acc = metrics.balanced_accuracy_score(test_true, test_pred)
+        outstr = 'Test %d, loss: %.6f, test acc: %.6f, test avg acc: %.6f' % (epoch,
+                                                                              test_loss*1.0/count,
+                                                                              test_acc,
+                                                                              avg_per_class_acc)
+        io.cprint(outstr)
+        if test_acc >= best_test_acc:
+            best_test_acc = test_acc
+            torch.save(model.state_dict(), 'checkpoints/%s/models/model.t7' % args.exp_name)
+
+
+def test(args, io):
+    test_loader = DataLoader(ModelNet40(partition='test', num_points=args.num_points),
+                             batch_size=args.test_batch_size, shuffle=True, drop_last=False)
+
+    device = torch.device("cuda" if args.cuda else "cpu")
+
+    #Try to load models
+    model = DGCNN(args).to(device)
+    model = nn.DataParallel(model)
+    model.load_state_dict(torch.load(args.model_path))
+    model = model.eval()
+    test_acc = 0.0
+    count = 0.0
+    test_true = []
+    test_pred = []
+    for data, label in test_loader:
+
+        data, label = data.to(device), label.to(device).squeeze()
+        data = data.permute(0, 2, 1)
+        batch_size = data.size()[0]
+        logits = model(data)
+        preds = logits.max(dim=1)[1]
+        test_true.append(label.cpu().numpy())
+        test_pred.append(preds.detach().cpu().numpy())
+    test_true = np.concatenate(test_true)
+    test_pred = np.concatenate(test_pred)
+    test_acc = metrics.accuracy_score(test_true, test_pred)
+    avg_per_class_acc = metrics.balanced_accuracy_score(test_true, test_pred)
+    outstr = 'Test :: test acc: %.6f, test avg acc: %.6f'%(test_acc, avg_per_class_acc)
+    io.cprint(outstr)
+
+
+if __name__ == "__main__":
+    # Training settings
+    parser = argparse.ArgumentParser(description='Point Cloud Recognition')
+    parser.add_argument('--exp_name', type=str, default='exp', metavar='N',
+                        help='Name of the experiment')
+    parser.add_argument('--model', type=str, default='dgcnn', metavar='N',
+                        choices=['pointnet', 'dgcnn'],
+                        help='Model to use, [pointnet, dgcnn]')
+    parser.add_argument('--dataset', type=str, default='modelnet40', metavar='N',
+                        choices=['modelnet40'])
+    parser.add_argument('--batch_size', type=int, default=32, metavar='batch_size',
+                        help='Size of batch)')
+    parser.add_argument('--test_batch_size', type=int, default=16, metavar='batch_size',
+                        help='Size of batch)')
+    parser.add_argument('--epochs', type=int, default=250, metavar='N',
+                        help='number of episode to train ')
+    parser.add_argument('--use_sgd', type=bool, default=True,
+                        help='Use SGD')
+    parser.add_argument('--lr', type=float, default=0.001, metavar='LR',
+                        help='learning rate (default: 0.001, 0.1 if using sgd)')
+    parser.add_argument('--momentum', type=float, default=0.9, metavar='M',
+                        help='SGD momentum (default: 0.9)')
+    parser.add_argument('--no_cuda', type=bool, default=False,
+                        help='enables CUDA training')
+    parser.add_argument('--seed', type=int, default=1, metavar='S',
+                        help='random seed (default: 1)')
+    parser.add_argument('--eval', type=bool,  default=False,
+                        help='evaluate the model')
+    parser.add_argument('--num_points', type=int, default=1024,
+                        help='num of points to use')
+    parser.add_argument('--dropout', type=float, default=0.5,
+                        help='dropout rate')
+    parser.add_argument('--emb_dims', type=int, default=1024, metavar='N',
+                        help='Dimension of embeddings')
+    parser.add_argument('--k', type=int, default=20, metavar='N',
+                        help='Num of nearest neighbors to use')
+    parser.add_argument('--model_path', type=str, default='', metavar='N',
+                        help='Pretrained model path')
+    args = parser.parse_args()
+
+    _init_()
+
+    io = IOStream('checkpoints/' + args.exp_name + '/run.log')
+    io.cprint(str(args))
+
+    args.cuda = not args.no_cuda and torch.cuda.is_available()
+    torch.manual_seed(args.seed)
+    if args.cuda:
+        io.cprint(
+            'Using GPU : ' + str(torch.cuda.current_device()) + ' from ' + str(torch.cuda.device_count()) + ' devices')
+        torch.cuda.manual_seed(args.seed)
+    else:
+        io.cprint('Using CPU')
+
+    if not args.eval:
+        train(args, io)
+    else:
+        test(args, io)
diff --git a/pytorch/model.py b/pytorch/model.py
new file mode 100644
index 0000000..b718591
--- /dev/null
+++ b/pytorch/model.py
@@ -0,0 +1,153 @@
+#!/usr/bin/env python
+# -*- coding: utf-8 -*-
+"""
+@Author: Yue Wang
+@Contact: yuewangx@mit.edu
+@File: model.py
+@Time: 2018/10/13 6:35 PM
+"""
+
+
+import os
+import sys
+import copy
+import math
+import numpy as np
+import torch
+import torch.nn as nn
+import torch.nn.functional as F
+
+
+def knn(x, k):
+    inner = -2*torch.matmul(x.transpose(2, 1), x)
+    xx = torch.sum(x**2, dim=1, keepdim=True)
+    pairwise_distance = -xx - inner - xx.transpose(2, 1)
+ 
+    idx = pairwise_distance.topk(k=k, dim=-1)[1]   # (batch_size, num_points, k)
+    return idx
+
+
+def get_graph_feature(x, k=20, idx=None):
+    batch_size = x.size(0)
+    num_points = x.size(2)
+    x = x.view(batch_size, -1, num_points)
+    if idx is None:
+        idx = knn(x, k=k)   # (batch_size, num_points, k)
+    device = torch.device('cuda')
+
+    idx_base = torch.arange(0, batch_size, device=device).view(-1, 1, 1)*num_points
+
+    idx = idx + idx_base
+
+    idx = idx.view(-1)
+ 
+    _, num_dims, _ = x.size()
+
+    x = x.transpose(2, 1).contiguous()   # (batch_size, num_points, num_dims)  -> (batch_size*num_points, num_dims) #   batch_size * num_points * k + range(0, batch_size*num_points)
+    feature = x.view(batch_size*num_points, -1)[idx, :]
+    feature = feature.view(batch_size, num_points, k, num_dims) 
+    x = x.view(batch_size, num_points, 1, num_dims).repeat(1, 1, k, 1)
+    
+    feature = torch.cat((feature, x), dim=3).permute(0, 3, 1, 2)
+  
+    return feature
+
+
+class PointNet(nn.Module):
+    def __init__(self, args, output_channels=40):
+        super(PointNet, self).__init__()
+        self.args = args
+        self.conv1 = nn.Conv1d(3, 64, kernel_size=1, bias=False)
+        self.conv2 = nn.Conv1d(64, 64, kernel_size=1, bias=False)
+        self.conv3 = nn.Conv1d(64, 64, kernel_size=1, bias=False)
+        self.conv4 = nn.Conv1d(64, 128, kernel_size=1, bias=False)
+        self.conv5 = nn.Conv1d(128, args.emb_dims, kernel_size=1, bias=False)
+        self.bn1 = nn.BatchNorm1d(64)
+        self.bn2 = nn.BatchNorm1d(64)
+        self.bn3 = nn.BatchNorm1d(64)
+        self.bn4 = nn.BatchNorm1d(128)
+        self.bn5 = nn.BatchNorm1d(args.emb_dims)
+        self.linear1 = nn.Linear(args.emb_dims, 512, bias=False)
+        self.bn6 = nn.BatchNorm1d(512)
+        self.dp1 = nn.Dropout()
+        self.linear2 = nn.Linear(512, output_channels)
+
+    def forward(self, x):
+        x = F.relu(self.bn1(self.conv1(x)))
+        x = F.relu(self.bn2(self.conv2(x)))
+        x = F.relu(self.bn3(self.conv3(x)))
+        x = F.relu(self.bn4(self.conv4(x)))
+        x = F.relu(self.bn5(self.conv5(x)))
+        x = F.adaptive_max_pool1d(x, 1).squeeze()
+        x = F.relu(self.bn6(self.linear1(x)))
+        x = self.dp1(x)
+        x = self.linear2(x)
+        return x
+
+
+class DGCNN(nn.Module):
+    def __init__(self, args, output_channels=40):
+        super(DGCNN, self).__init__()
+        self.args = args
+        self.k = args.k
+        
+        self.bn1 = nn.BatchNorm2d(64)
+        self.bn2 = nn.BatchNorm2d(64)
+        self.bn3 = nn.BatchNorm2d(128)
+        self.bn4 = nn.BatchNorm2d(256)
+        self.bn5 = nn.BatchNorm1d(args.emb_dims)
+
+        self.conv1 = nn.Sequential(nn.Conv2d(6, 64, kernel_size=1, bias=False),
+                                   self.bn1,
+                                   nn.LeakyReLU(negative_slope=0.2))
+        self.conv2 = nn.Sequential(nn.Conv2d(64*2, 64, kernel_size=1, bias=False),
+                                   self.bn2,
+                                   nn.LeakyReLU(negative_slope=0.2))
+        self.conv3 = nn.Sequential(nn.Conv2d(64*2, 128, kernel_size=1, bias=False),
+                                   self.bn3,
+                                   nn.LeakyReLU(negative_slope=0.2))
+        self.conv4 = nn.Sequential(nn.Conv2d(128*2, 256, kernel_size=1, bias=False),
+                                   self.bn4,
+                                   nn.LeakyReLU(negative_slope=0.2))
+        self.conv5 = nn.Sequential(nn.Conv1d(512, args.emb_dims, kernel_size=1, bias=False),
+                                   self.bn5,
+                                   nn.LeakyReLU(negative_slope=0.2))
+        self.linear1 = nn.Linear(args.emb_dims*2, 512, bias=False)
+        self.bn6 = nn.BatchNorm1d(512)
+        self.dp1 = nn.Dropout(p=args.dropout)
+        self.linear2 = nn.Linear(512, 256)
+        self.bn7 = nn.BatchNorm1d(256)
+        self.dp2 = nn.Dropout(p=args.dropout)
+        self.linear3 = nn.Linear(256, output_channels)
+
+    def forward(self, x):
+        batch_size = x.size(0)
+        x = get_graph_feature(x, k=self.k)
+        x = self.conv1(x)
+        x1 = x.max(dim=-1, keepdim=False)[0]
+
+        x = get_graph_feature(x1, k=self.k)
+        x = self.conv2(x)
+        x2 = x.max(dim=-1, keepdim=False)[0]
+
+        x = get_graph_feature(x2, k=self.k)
+        x = self.conv3(x)
+        x3 = x.max(dim=-1, keepdim=False)[0]
+
+        x = get_graph_feature(x3, k=self.k)
+        x = self.conv4(x)
+        x4 = x.max(dim=-1, keepdim=False)[0]
+
+        x = torch.cat((x1, x2, x3, x4), dim=1)
+
+        x = self.conv5(x)
+        x1 = F.adaptive_max_pool1d(x, 1).view(batch_size, -1)
+        x2 = F.adaptive_avg_pool1d(x, 1).view(batch_size, -1)
+        x = torch.cat((x1, x2), 1)
+
+        x = F.leaky_relu(self.bn6(self.linear1(x)), negative_slope=0.2)
+        x = self.dp1(x)
+        x = F.leaky_relu(self.bn7(self.linear2(x)), negative_slope=0.2)
+        x = self.dp2(x)
+        x = self.linear3(x)
+        return x
diff --git a/pytorch/pretrained/model.1024.t7 b/pytorch/pretrained/model.1024.t7
new file mode 100644
index 0000000..4e99f49
Binary files /dev/null and b/pytorch/pretrained/model.1024.t7 differ
diff --git a/pytorch/pretrained/model.2048.t7 b/pytorch/pretrained/model.2048.t7
new file mode 100644
index 0000000..3f615ec
Binary files /dev/null and b/pytorch/pretrained/model.2048.t7 differ
diff --git a/pytorch/util.py b/pytorch/util.py
new file mode 100644
index 0000000..6875176
--- /dev/null
+++ b/pytorch/util.py
@@ -0,0 +1,46 @@
+#!/usr/bin/env python
+# -*- coding: utf-8 -*-
+"""
+@Author: Yue Wang
+@Contact: yuewangx@mit.edu
+@File: util
+@Time: 4/5/19 3:47 PM
+"""
+
+
+import numpy as np
+import torch
+import torch.nn.functional as F
+
+
+def cal_loss(pred, gold, smoothing=True):
+    ''' Calculate cross entropy loss, apply label smoothing if needed. '''
+
+    gold = gold.contiguous().view(-1)
+
+    if smoothing:
+        eps = 0.2
+        n_class = pred.size(1)
+
+        one_hot = torch.zeros_like(pred).scatter(1, gold.view(-1, 1), 1)
+        one_hot = one_hot * (1 - eps) + (1 - one_hot) * eps / (n_class - 1)
+        log_prb = F.log_softmax(pred, dim=1)
+
+        loss = -(one_hot * log_prb).sum(dim=1).mean()
+    else:
+        loss = F.cross_entropy(pred, gold, reduction='mean')
+
+    return loss
+
+
+class IOStream():
+    def __init__(self, path):
+        self.f = open(path, 'a')
+
+    def cprint(self, text):
+        print(text)
+        self.f.write(text+'\n')
+        self.f.flush()
+
+    def close(self):
+        self.f.close()