LiuhanChen commited on
Commit
f2b657e
·
verified ·
1 Parent(s): 7fed6be

Add files using upload-large-folder tool

Browse files
Files changed (45) hide show
  1. .gitignore +11 -0
  2. =4.10 +0 -0
  3. README.md +1 -0
  4. causalvideovae/dataset/__init__.py +65 -0
  5. causalvideovae/dataset/feature_datasets.py +213 -0
  6. causalvideovae/dataset/t2v_datasets.py +203 -0
  7. causalvideovae/dataset/transform.py +573 -0
  8. causalvideovae/eval/RAFT/chairs_split.txt +22872 -0
  9. causalvideovae/eval/RAFT/core/corr.py +91 -0
  10. causalvideovae/eval/RAFT/core/utils/frame_utils.py +137 -0
  11. causalvideovae/eval/RAFT/core/utils/utils.py +82 -0
  12. causalvideovae/eval/RAFT/demo.py +75 -0
  13. causalvideovae/eval/RAFT/download_models.sh +3 -0
  14. causalvideovae/eval/RAFT/train_mixed.sh +6 -0
  15. causalvideovae/eval/cal_mse.py +93 -0
  16. causalvideovae/eval/flolpips/flolpips.py +308 -0
  17. causalvideovae/eval/script/cal_clip_score.sh +23 -0
  18. causalvideovae/eval/script/cal_fvd.sh +9 -0
  19. causalvideovae/eval/script/cal_psnr.sh +9 -0
  20. causalvideovae/eval/script/cal_ssim.sh +8 -0
  21. causalvideovae/model/causal_vae/__init__.py +29 -0
  22. causalvideovae/model/causal_vae/modeling_causalvae.py +610 -0
  23. causalvideovae/model/configuration_videobase.py +44 -0
  24. causalvideovae/model/losses/__init__.py +1 -0
  25. causalvideovae/model/losses/discriminator.py +181 -0
  26. causalvideovae/model/modeling_videobase.py +53 -0
  27. causalvideovae/model/modules/__init__.py +6 -0
  28. causalvideovae/model/modules/conv.py +142 -0
  29. causalvideovae/model/modules/quant.py +100 -0
  30. causalvideovae/model/modules/resnet_block.py +218 -0
  31. causalvideovae/model/modules/updownsample.py +370 -0
  32. causalvideovae/model/refiner/modeling_refiner.py +490 -0
  33. causalvideovae/model/trainer_videobase.py +26 -0
  34. causalvideovae/model/utils/scheduler_utils.py +7 -0
  35. causalvideovae/utils/downloader.py +18 -0
  36. causalvideovae/utils/taming_download.py +145 -0
  37. scripts/cal_expense.py +347 -0
  38. scripts/config_deepspeed.yaml +20 -0
  39. scripts/convert_weight.py +74 -0
  40. scripts/cv_vae_gen_video.sh +30 -0
  41. scripts/expand_weight.py +40 -0
  42. scripts/rec_TATS_vae.py +310 -0
  43. scripts/rec_causalvideo_vae.py +324 -0
  44. scripts/rec_nus_vae.py +302 -0
  45. scripts/rec_svd_vae.py +301 -0
.gitignore ADDED
@@ -0,0 +1,11 @@
 
 
 
 
 
 
 
 
 
 
 
 
1
+ **/__pycache__
2
+ *.pt*
3
+ models
4
+ taming
5
+ wandb/
6
+ *.png
7
+ outputs.png
8
+ test.ipynb
9
+ results/
10
+ scripts/expand_weight.ipynb
11
+ *.mp4
=4.10 ADDED
File without changes
README.md ADDED
@@ -0,0 +1 @@
 
 
1
+ # Causal Video VAE
causalvideovae/dataset/__init__.py ADDED
@@ -0,0 +1,65 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ from torchvision.transforms import Compose
2
+ from transformers import AutoTokenizer
3
+
4
+ from .feature_datasets import T2V_Feature_dataset, T2V_T5_Feature_dataset
5
+ from torchvision import transforms
6
+ from torchvision.transforms import Lambda
7
+
8
+ from .t2v_datasets import T2V_dataset
9
+ from .transform import ToTensorVideo, TemporalRandomCrop, RandomHorizontalFlipVideo, CenterCropResizeVideo, LongSideResizeVideo, SpatialStrideCropVideo
10
+
11
+
12
+ ae_norm = {
13
+ 'CausalVAEModel_4x8x8': Lambda(lambda x: 2. * x - 1.),
14
+ 'CausalVQVAEModel_4x4x4': Lambda(lambda x: x - 0.5),
15
+ 'CausalVQVAEModel_4x8x8': Lambda(lambda x: x - 0.5),
16
+ 'VQVAEModel_4x4x4': Lambda(lambda x: x - 0.5),
17
+ 'VQVAEModel_4x8x8': Lambda(lambda x: x - 0.5),
18
+ "bair_stride4x2x2": Lambda(lambda x: x - 0.5),
19
+ "ucf101_stride4x4x4": Lambda(lambda x: x - 0.5),
20
+ "kinetics_stride4x4x4": Lambda(lambda x: x - 0.5),
21
+ "kinetics_stride2x4x4": Lambda(lambda x: x - 0.5),
22
+ 'stabilityai/sd-vae-ft-mse': transforms.Normalize(mean=[0.5, 0.5, 0.5], std=[0.5, 0.5, 0.5], inplace=True),
23
+ 'stabilityai/sd-vae-ft-ema': transforms.Normalize(mean=[0.5, 0.5, 0.5], std=[0.5, 0.5, 0.5], inplace=True),
24
+ 'vqgan_imagenet_f16_1024': Lambda(lambda x: 2. * x - 1.),
25
+ 'vqgan_imagenet_f16_16384': Lambda(lambda x: 2. * x - 1.),
26
+ 'vqgan_gumbel_f8': Lambda(lambda x: 2. * x - 1.),
27
+
28
+ }
29
+ ae_denorm = {
30
+ 'CausalVAEModel_4x8x8': lambda x: (x + 1.) / 2.,
31
+ 'CausalVQVAEModel_4x4x4': lambda x: x + 0.5,
32
+ 'CausalVQVAEModel_4x8x8': lambda x: x + 0.5,
33
+ 'VQVAEModel_4x4x4': lambda x: x + 0.5,
34
+ 'VQVAEModel_4x8x8': lambda x: x + 0.5,
35
+ "bair_stride4x2x2": lambda x: x + 0.5,
36
+ "ucf101_stride4x4x4": lambda x: x + 0.5,
37
+ "kinetics_stride4x4x4": lambda x: x + 0.5,
38
+ "kinetics_stride2x4x4": lambda x: x + 0.5,
39
+ 'stabilityai/sd-vae-ft-mse': lambda x: 0.5 * x + 0.5,
40
+ 'stabilityai/sd-vae-ft-ema': lambda x: 0.5 * x + 0.5,
41
+ 'vqgan_imagenet_f16_1024': lambda x: (x + 1.) / 2.,
42
+ 'vqgan_imagenet_f16_16384': lambda x: (x + 1.) / 2.,
43
+ 'vqgan_gumbel_f8': lambda x: (x + 1.) / 2.,
44
+ }
45
+
46
+ def getdataset(args):
47
+ temporal_sample = TemporalRandomCrop(args.num_frames * args.sample_rate) # 16 x
48
+ norm_fun = ae_norm[args.ae]
49
+ if args.dataset == 't2v':
50
+ if args.multi_scale:
51
+ resize = [
52
+ LongSideResizeVideo(args.max_image_size, skip_low_resolution=True),
53
+ SpatialStrideCropVideo(args.stride)
54
+ ]
55
+ else:
56
+ resize = [CenterCropResizeVideo(args.max_image_size), ]
57
+ transform = transforms.Compose([
58
+ ToTensorVideo(),
59
+ *resize,
60
+ # RandomHorizontalFlipVideo(p=0.5), # in case their caption have position decription
61
+ norm_fun
62
+ ])
63
+ tokenizer = AutoTokenizer.from_pretrained(args.text_encoder_name, cache_dir=args.cache_dir)
64
+ return T2V_dataset(args, transform=transform, temporal_sample=temporal_sample, tokenizer=tokenizer)
65
+ raise NotImplementedError(args.dataset)
causalvideovae/dataset/feature_datasets.py ADDED
@@ -0,0 +1,213 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ import json
2
+ import os
3
+ import torch
4
+ import random
5
+ import torch.utils.data as data
6
+
7
+ import numpy as np
8
+ from glob import glob
9
+ from PIL import Image
10
+ from torch.utils.data import Dataset
11
+ from tqdm import tqdm
12
+
13
+ from .transform import center_crop, RandomCropVideo
14
+ from ..utils.dataset_utils import DecordInit
15
+
16
+
17
+ class T2V_Feature_dataset(Dataset):
18
+ def __init__(self, args, temporal_sample):
19
+
20
+ self.video_folder = args.video_folder
21
+ self.num_frames = args.video_length
22
+ self.temporal_sample = temporal_sample
23
+
24
+ print('Building dataset...')
25
+ if os.path.exists('samples_430k.json'):
26
+ with open('samples_430k.json', 'r') as f:
27
+ self.samples = json.load(f)
28
+ else:
29
+ self.samples = self._make_dataset()
30
+ with open('samples_430k.json', 'w') as f:
31
+ json.dump(self.samples, f, indent=2)
32
+
33
+ self.use_image_num = args.use_image_num
34
+ self.use_img_from_vid = args.use_img_from_vid
35
+ if self.use_image_num != 0 and not self.use_img_from_vid:
36
+ self.img_cap_list = self.get_img_cap_list()
37
+
38
+ def _make_dataset(self):
39
+ all_mp4 = list(glob(os.path.join(self.video_folder, '**', '*.mp4'), recursive=True))
40
+ # all_mp4 = all_mp4[:1000]
41
+ samples = []
42
+ for i in tqdm(all_mp4):
43
+ video_id = os.path.basename(i).split('.')[0]
44
+ ae = os.path.split(i)[0].replace('data_split_tt', 'lb_causalvideovae444_feature')
45
+ ae = os.path.join(ae, f'{video_id}_causalvideovae444.npy')
46
+ if not os.path.exists(ae):
47
+ continue
48
+
49
+ t5 = os.path.split(i)[0].replace('data_split_tt', 'lb_t5_feature')
50
+ cond_list = []
51
+ cond_llava = os.path.join(t5, f'{video_id}_t5_llava_fea.npy')
52
+ mask_llava = os.path.join(t5, f'{video_id}_t5_llava_mask.npy')
53
+ if os.path.exists(cond_llava) and os.path.exists(mask_llava):
54
+ llava = dict(cond=cond_llava, mask=mask_llava)
55
+ cond_list.append(llava)
56
+ cond_sharegpt4v = os.path.join(t5, f'{video_id}_t5_sharegpt4v_fea.npy')
57
+ mask_sharegpt4v = os.path.join(t5, f'{video_id}_t5_sharegpt4v_mask.npy')
58
+ if os.path.exists(cond_sharegpt4v) and os.path.exists(mask_sharegpt4v):
59
+ sharegpt4v = dict(cond=cond_sharegpt4v, mask=mask_sharegpt4v)
60
+ cond_list.append(sharegpt4v)
61
+ if len(cond_list) > 0:
62
+ sample = dict(ae=ae, t5=cond_list)
63
+ samples.append(sample)
64
+ return samples
65
+
66
+ def __len__(self):
67
+ return len(self.samples)
68
+
69
+ def __getitem__(self, idx):
70
+ # try:
71
+ sample = self.samples[idx]
72
+ ae, t5 = sample['ae'], sample['t5']
73
+ t5 = random.choice(t5)
74
+ video_origin = np.load(ae)[0] # C T H W
75
+ _, total_frames, _, _ = video_origin.shape
76
+ # Sampling video frames
77
+ start_frame_ind, end_frame_ind = self.temporal_sample(total_frames)
78
+ assert end_frame_ind - start_frame_ind >= self.num_frames
79
+ select_video_idx = np.linspace(start_frame_ind, end_frame_ind - 1, num=self.num_frames, dtype=int) # start, stop, num=50
80
+ # print('select_video_idx', total_frames, select_video_idx)
81
+ video = video_origin[:, select_video_idx] # C num_frames H W
82
+ video = torch.from_numpy(video)
83
+
84
+ cond = torch.from_numpy(np.load(t5['cond']))[0] # L
85
+ cond_mask = torch.from_numpy(np.load(t5['mask']))[0] # L D
86
+
87
+ if self.use_image_num != 0 and self.use_img_from_vid:
88
+ select_image_idx = np.random.randint(0, total_frames, self.use_image_num)
89
+ # print('select_image_idx', total_frames, self.use_image_num, select_image_idx)
90
+ images = video_origin[:, select_image_idx] # c, num_img, h, w
91
+ images = torch.from_numpy(images)
92
+ video = torch.cat([video, images], dim=1) # c, num_frame+num_img, h, w
93
+ cond = torch.stack([cond] * (1+self.use_image_num)) # 1+self.use_image_num, l
94
+ cond_mask = torch.stack([cond_mask] * (1+self.use_image_num)) # 1+self.use_image_num, l
95
+ elif self.use_image_num != 0 and not self.use_img_from_vid:
96
+ images, captions = self.img_cap_list[idx]
97
+ raise NotImplementedError
98
+ else:
99
+ pass
100
+
101
+ return video, cond, cond_mask
102
+ # except Exception as e:
103
+ # print(f'Error with {e}, {sample}')
104
+ # return self.__getitem__(random.randint(0, self.__len__() - 1))
105
+
106
+ def get_img_cap_list(self):
107
+ raise NotImplementedError
108
+
109
+
110
+
111
+
112
+ class T2V_T5_Feature_dataset(Dataset):
113
+ def __init__(self, args, transform, temporal_sample):
114
+
115
+ self.video_folder = args.video_folder
116
+ self.num_frames = args.num_frames
117
+ self.transform = transform
118
+ self.temporal_sample = temporal_sample
119
+ self.v_decoder = DecordInit()
120
+
121
+ print('Building dataset...')
122
+ if os.path.exists('samples_430k.json'):
123
+ with open('samples_430k.json', 'r') as f:
124
+ self.samples = json.load(f)
125
+ self.samples = [dict(ae=i['ae'].replace('lb_causalvideovae444_feature', 'data_split_1024').replace('_causalvideovae444.npy', '.mp4'), t5=i['t5']) for i in self.samples]
126
+ else:
127
+ self.samples = self._make_dataset()
128
+ with open('samples_430k.json', 'w') as f:
129
+ json.dump(self.samples, f, indent=2)
130
+
131
+ self.use_image_num = args.use_image_num
132
+ self.use_img_from_vid = args.use_img_from_vid
133
+ if self.use_image_num != 0 and not self.use_img_from_vid:
134
+ self.img_cap_list = self.get_img_cap_list()
135
+
136
+ def _make_dataset(self):
137
+ all_mp4 = list(glob(os.path.join(self.video_folder, '**', '*.mp4'), recursive=True))
138
+ # all_mp4 = all_mp4[:1000]
139
+ samples = []
140
+ for i in tqdm(all_mp4):
141
+ video_id = os.path.basename(i).split('.')[0]
142
+ # ae = os.path.split(i)[0].replace('data_split', 'lb_causalvideovae444_feature')
143
+ # ae = os.path.join(ae, f'{video_id}_causalvideovae444.npy')
144
+ ae = i
145
+ if not os.path.exists(ae):
146
+ continue
147
+
148
+ t5 = os.path.split(i)[0].replace('data_split_1024', 'lb_t5_feature')
149
+ cond_list = []
150
+ cond_llava = os.path.join(t5, f'{video_id}_t5_llava_fea.npy')
151
+ mask_llava = os.path.join(t5, f'{video_id}_t5_llava_mask.npy')
152
+ if os.path.exists(cond_llava) and os.path.exists(mask_llava):
153
+ llava = dict(cond=cond_llava, mask=mask_llava)
154
+ cond_list.append(llava)
155
+ cond_sharegpt4v = os.path.join(t5, f'{video_id}_t5_sharegpt4v_fea.npy')
156
+ mask_sharegpt4v = os.path.join(t5, f'{video_id}_t5_sharegpt4v_mask.npy')
157
+ if os.path.exists(cond_sharegpt4v) and os.path.exists(mask_sharegpt4v):
158
+ sharegpt4v = dict(cond=cond_sharegpt4v, mask=mask_sharegpt4v)
159
+ cond_list.append(sharegpt4v)
160
+ if len(cond_list) > 0:
161
+ sample = dict(ae=ae, t5=cond_list)
162
+ samples.append(sample)
163
+ return samples
164
+
165
+ def __len__(self):
166
+ return len(self.samples)
167
+
168
+ def __getitem__(self, idx):
169
+ try:
170
+ sample = self.samples[idx]
171
+ ae, t5 = sample['ae'], sample['t5']
172
+ t5 = random.choice(t5)
173
+
174
+ video = self.decord_read(ae)
175
+ video = self.transform(video) # T C H W -> T C H W
176
+ video = video.transpose(0, 1) # T C H W -> C T H W
177
+ total_frames = video.shape[1]
178
+ cond = torch.from_numpy(np.load(t5['cond']))[0] # L
179
+ cond_mask = torch.from_numpy(np.load(t5['mask']))[0] # L D
180
+
181
+ if self.use_image_num != 0 and self.use_img_from_vid:
182
+ select_image_idx = np.random.randint(0, total_frames, self.use_image_num)
183
+ # print('select_image_idx', total_frames, self.use_image_num, select_image_idx)
184
+ images = video.numpy()[:, select_image_idx] # c, num_img, h, w
185
+ images = torch.from_numpy(images)
186
+ video = torch.cat([video, images], dim=1) # c, num_frame+num_img, h, w
187
+ cond = torch.stack([cond] * (1+self.use_image_num)) # 1+self.use_image_num, l
188
+ cond_mask = torch.stack([cond_mask] * (1+self.use_image_num)) # 1+self.use_image_num, l
189
+ elif self.use_image_num != 0 and not self.use_img_from_vid:
190
+ images, captions = self.img_cap_list[idx]
191
+ raise NotImplementedError
192
+ else:
193
+ pass
194
+
195
+ return video, cond, cond_mask
196
+ except Exception as e:
197
+ print(f'Error with {e}, {sample}')
198
+ return self.__getitem__(random.randint(0, self.__len__() - 1))
199
+
200
+ def decord_read(self, path):
201
+ decord_vr = self.v_decoder(path)
202
+ total_frames = len(decord_vr)
203
+ # Sampling video frames
204
+ start_frame_ind, end_frame_ind = self.temporal_sample(total_frames)
205
+ # assert end_frame_ind - start_frame_ind >= self.num_frames
206
+ frame_indice = np.linspace(start_frame_ind, end_frame_ind - 1, self.num_frames, dtype=int)
207
+ video_data = decord_vr.get_batch(frame_indice).asnumpy()
208
+ video_data = torch.from_numpy(video_data)
209
+ video_data = video_data.permute(0, 3, 1, 2) # (T, H, W, C) -> (T C H W)
210
+ return video_data
211
+
212
+ def get_img_cap_list(self):
213
+ raise NotImplementedError
causalvideovae/dataset/t2v_datasets.py ADDED
@@ -0,0 +1,203 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ import json
2
+ import os, io, csv, math, random
3
+ import numpy as np
4
+ import torchvision
5
+ from einops import rearrange
6
+ from decord import VideoReader
7
+ from os.path import join as opj
8
+ import gc
9
+ import torch
10
+ import torchvision.transforms as transforms
11
+ from torch.utils.data.dataset import Dataset
12
+ from tqdm import tqdm
13
+ from PIL import Image
14
+
15
+ from ..utils.dataset_utils import DecordInit
16
+ from ..utils.utils import text_preprocessing
17
+
18
+
19
+ def random_video_noise(t, c, h, w):
20
+ vid = torch.rand(t, c, h, w) * 255.0
21
+ vid = vid.to(torch.uint8)
22
+ return vid
23
+
24
+ class T2V_dataset(Dataset):
25
+ def __init__(self, args, transform, temporal_sample, tokenizer):
26
+ self.image_data = args.image_data
27
+ self.video_data = args.video_data
28
+ self.num_frames = args.num_frames
29
+ self.transform = transform
30
+ self.temporal_sample = temporal_sample
31
+ self.tokenizer = tokenizer
32
+ self.model_max_length = args.model_max_length
33
+ self.v_decoder = DecordInit()
34
+
35
+ self.vid_cap_list = self.get_vid_cap_list()
36
+
37
+ self.use_image_num = args.use_image_num
38
+ self.use_img_from_vid = args.use_img_from_vid
39
+ if self.use_image_num != 0 and not self.use_img_from_vid:
40
+ self.img_cap_list = self.get_img_cap_list()
41
+
42
+
43
+ def __len__(self):
44
+ return len(self.vid_cap_list)
45
+
46
+ def __getitem__(self, idx):
47
+ try:
48
+ # import ipdb;ipdb.set_trace()
49
+ video_data = self.get_video(idx)
50
+ image_data = {}
51
+ if self.use_image_num != 0 and self.use_img_from_vid:
52
+ image_data = self.get_image_from_video(video_data)
53
+ elif self.use_image_num != 0 and not self.use_img_from_vid:
54
+ image_data = self.get_image(idx)
55
+ else:
56
+ raise NotImplementedError
57
+ gc.collect()
58
+ return dict(video_data=video_data, image_data=image_data)
59
+ except Exception as e:
60
+ # print(f'Error with {e}, {self.vid_cap_list[idx]}')
61
+ if os.path.exists(self.vid_cap_list[idx]['path']) and '_resize_1080p' in self.vid_cap_list[idx]['path']:
62
+ os.remove(self.vid_cap_list[idx]['path'])
63
+ print('remove:', self.vid_cap_list[idx]['path'])
64
+ return self.__getitem__(random.randint(0, self.__len__() - 1))
65
+
66
+ def get_video(self, idx):
67
+ # video = random.choice([random_video_noise(65, 3, 720, 360) * 255, random_video_noise(65, 3, 1024, 1024), random_video_noise(65, 3, 360, 720)])
68
+ # # print('random shape', video.shape)
69
+ # input_ids = torch.ones(1, 120).to(torch.long).squeeze(0)
70
+ # cond_mask = torch.cat([torch.ones(1, 60).to(torch.long), torch.ones(1, 60).to(torch.long)], dim=1).squeeze(0)
71
+
72
+ video_path = self.vid_cap_list[idx]['path']
73
+ frame_idx = self.vid_cap_list[idx]['frame_idx']
74
+ #print('before decord')
75
+ video = self.decord_read(video_path, frame_idx)
76
+ # video = self.tv_read(video_path, frame_idx)
77
+ #print('after decord')
78
+ video = self.transform(video) # T C H W -> T C H W
79
+ # del raw_video
80
+ # gc.collect()
81
+ # video = torch.rand(65, 3, 512, 512)
82
+ #print('after transform')
83
+ video = video.transpose(0, 1) # T C H W -> C T H W
84
+ text = self.vid_cap_list[idx]['cap']
85
+
86
+ text = text_preprocessing(text)
87
+ text_tokens_and_mask = self.tokenizer(
88
+ text,
89
+ max_length=self.model_max_length,
90
+ padding='max_length',
91
+ truncation=True,
92
+ return_attention_mask=True,
93
+ add_special_tokens=True,
94
+ return_tensors='pt'
95
+ )
96
+ input_ids = text_tokens_and_mask['input_ids']
97
+ cond_mask = text_tokens_and_mask['attention_mask']
98
+ return dict(video=video, input_ids=input_ids, cond_mask=cond_mask)
99
+
100
+ def get_image_from_video(self, video_data):
101
+ select_image_idx = np.linspace(0, self.num_frames-1, self.use_image_num, dtype=int)
102
+ assert self.num_frames >= self.use_image_num
103
+ image = [video_data['video'][:, i:i+1] for i in select_image_idx] # num_img [c, 1, h, w]
104
+ input_ids = video_data['input_ids'].repeat(self.use_image_num, 1) # self.use_image_num, l
105
+ cond_mask = video_data['cond_mask'].repeat(self.use_image_num, 1) # self.use_image_num, l
106
+ return dict(image=image, input_ids=input_ids, cond_mask=cond_mask)
107
+
108
+ def get_image(self, idx):
109
+ idx = idx % len(self.img_cap_list) # out of range
110
+ image_data = self.img_cap_list[idx] # [{'path': path, 'cap': cap}, ...]
111
+
112
+ image = [Image.open(i['path']).convert('RGB') for i in image_data] # num_img [h, w, c]
113
+ image = [torch.from_numpy(np.array(i)) for i in image] # num_img [h, w, c]
114
+ image = [rearrange(i, 'h w c -> c h w').unsqueeze(0) for i in image] # num_img [1 c h w]
115
+ image = [self.transform(i) for i in image] # num_img [1 C H W] -> num_img [1 C H W]
116
+ image = [i.transpose(0, 1) for i in image] # num_img [1 C H W] -> num_img [C 1 H W]
117
+
118
+ caps = [i['cap'] for i in image_data]
119
+ text = [text_preprocessing(cap) for cap in caps]
120
+ input_ids, cond_mask = [], []
121
+ for t in text:
122
+ text_tokens_and_mask = self.tokenizer(
123
+ t,
124
+ max_length=self.model_max_length,
125
+ padding='max_length',
126
+ truncation=True,
127
+ return_attention_mask=True,
128
+ add_special_tokens=True,
129
+ return_tensors='pt'
130
+ )
131
+ input_ids.append(text_tokens_and_mask['input_ids'])
132
+ cond_mask.append(text_tokens_and_mask['attention_mask'])
133
+ input_ids = torch.cat(input_ids) # self.use_image_num, l
134
+ cond_mask = torch.cat(cond_mask) # self.use_image_num, l
135
+ return dict(image=image, input_ids=input_ids, cond_mask=cond_mask)
136
+
137
+ def tv_read(self, path, frame_idx=None):
138
+ vframes, aframes, info = torchvision.io.read_video(filename=path, pts_unit='sec', output_format='TCHW')
139
+ total_frames = len(vframes)
140
+ if frame_idx is None:
141
+ start_frame_ind, end_frame_ind = self.temporal_sample(total_frames)
142
+ else:
143
+ start_frame_ind, end_frame_ind = frame_idx.split(':')
144
+ start_frame_ind, end_frame_ind = int(start_frame_ind), int(end_frame_ind)
145
+ # assert end_frame_ind - start_frame_ind >= self.num_frames
146
+ frame_indice = np.linspace(start_frame_ind, end_frame_ind - 1, self.num_frames, dtype=int)
147
+ # frame_indice = np.linspace(0, 63, self.num_frames, dtype=int)
148
+
149
+ video = vframes[frame_indice] # (T, C, H, W)
150
+
151
+ return video
152
+
153
+ def decord_read(self, path, frame_idx=None):
154
+ decord_vr = self.v_decoder(path)
155
+ total_frames = len(decord_vr)
156
+ # Sampling video frames
157
+ if frame_idx is None:
158
+ start_frame_ind, end_frame_ind = self.temporal_sample(total_frames)
159
+ else:
160
+ start_frame_ind, end_frame_ind = frame_idx.split(':')
161
+ start_frame_ind, end_frame_ind = int(start_frame_ind), int(end_frame_ind)
162
+ start_frame_ind, end_frame_ind = int(start_frame_ind), int(start_frame_ind) + self.num_frames
163
+ # assert end_frame_ind - start_frame_ind >= self.num_frames
164
+ frame_indice = np.linspace(start_frame_ind, end_frame_ind - 1, self.num_frames, dtype=int)
165
+ # frame_indice = np.linspace(0, 63, self.num_frames, dtype=int)
166
+
167
+ video_data = decord_vr.get_batch(frame_indice).asnumpy()
168
+ video_data = torch.from_numpy(video_data)
169
+ video_data = video_data.permute(0, 3, 1, 2) # (T, H, W, C) -> (T C H W)
170
+ return video_data
171
+
172
+
173
+ def get_vid_cap_list(self):
174
+ vid_cap_lists = []
175
+ with open(self.video_data, 'r') as f:
176
+ folder_anno = [i.strip().split(',') for i in f.readlines() if len(i.strip()) > 0]
177
+ # print(folder_anno)
178
+ for folder, anno in folder_anno:
179
+ with open(anno, 'r') as f:
180
+ vid_cap_list = json.load(f)
181
+ print(f'Building {anno}...')
182
+ for i in tqdm(range(len(vid_cap_list))):
183
+ path = opj(folder, vid_cap_list[i]['path'])
184
+ if os.path.exists(path.replace('.mp4', '_resize_1080p.mp4')):
185
+ path = path.replace('.mp4', '_resize_1080p.mp4')
186
+ vid_cap_list[i]['path'] = path
187
+
188
+ vid_cap_lists += vid_cap_list
189
+ return vid_cap_lists
190
+
191
+ def get_img_cap_list(self):
192
+ img_cap_lists = []
193
+ with open(self.image_data, 'r') as f:
194
+ folder_anno = [i.strip().split(',') for i in f.readlines() if len(i.strip()) > 0]
195
+ for folder, anno in folder_anno:
196
+ with open(anno, 'r') as f:
197
+ img_cap_list = json.load(f)
198
+ print(f'Building {anno}...')
199
+ for i in tqdm(range(len(img_cap_list))):
200
+ img_cap_list[i]['path'] = opj(folder, img_cap_list[i]['path'])
201
+ img_cap_lists += img_cap_list
202
+ img_cap_lists = [img_cap_lists[i: i+self.use_image_num] for i in range(0, len(img_cap_lists), self.use_image_num)]
203
+ return img_cap_lists[:-1] # drop last to avoid error length
causalvideovae/dataset/transform.py ADDED
@@ -0,0 +1,573 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ import torch
2
+ import random
3
+ import numbers
4
+ from torchvision.transforms import RandomCrop, RandomResizedCrop
5
+
6
+
7
+ def _is_tensor_video_clip(clip):
8
+ if not torch.is_tensor(clip):
9
+ raise TypeError("clip should be Tensor. Got %s" % type(clip))
10
+
11
+ if not clip.ndimension() == 4:
12
+ raise ValueError("clip should be 4D. Got %dD" % clip.dim())
13
+
14
+ return True
15
+
16
+
17
+ def center_crop_arr(pil_image, image_size):
18
+ """
19
+ Center cropping implementation from ADM.
20
+ https://github.com/openai/guided-diffusion/blob/8fb3ad9197f16bbc40620447b2742e13458d2831/guided_diffusion/image_datasets.py#L126
21
+ """
22
+ while min(*pil_image.size) >= 2 * image_size:
23
+ pil_image = pil_image.resize(
24
+ tuple(x // 2 for x in pil_image.size), resample=Image.BOX
25
+ )
26
+
27
+ scale = image_size / min(*pil_image.size)
28
+ pil_image = pil_image.resize(
29
+ tuple(round(x * scale) for x in pil_image.size), resample=Image.BICUBIC
30
+ )
31
+
32
+ arr = np.array(pil_image)
33
+ crop_y = (arr.shape[0] - image_size) // 2
34
+ crop_x = (arr.shape[1] - image_size) // 2
35
+ return Image.fromarray(arr[crop_y: crop_y + image_size, crop_x: crop_x + image_size])
36
+
37
+
38
+ def crop(clip, i, j, h, w):
39
+ """
40
+ Args:
41
+ clip (torch.tensor): Video clip to be cropped. Size is (T, C, H, W)
42
+ """
43
+ if len(clip.size()) != 4:
44
+ raise ValueError("clip should be a 4D tensor")
45
+ return clip[..., i: i + h, j: j + w]
46
+
47
+
48
+ def resize(clip, target_size, interpolation_mode):
49
+ if len(target_size) != 2:
50
+ raise ValueError(f"target size should be tuple (height, width), instead got {target_size}")
51
+ return torch.nn.functional.interpolate(clip, size=target_size, mode=interpolation_mode, align_corners=True, antialias=True)
52
+
53
+
54
+ def resize_scale(clip, target_size, interpolation_mode):
55
+ if len(target_size) != 2:
56
+ raise ValueError(f"target size should be tuple (height, width), instead got {target_size}")
57
+ H, W = clip.size(-2), clip.size(-1)
58
+ scale_ = target_size[0] / min(H, W)
59
+ return torch.nn.functional.interpolate(clip, scale_factor=scale_, mode=interpolation_mode, align_corners=True, antialias=True)
60
+
61
+
62
+ def resized_crop(clip, i, j, h, w, size, interpolation_mode="bilinear"):
63
+ """
64
+ Do spatial cropping and resizing to the video clip
65
+ Args:
66
+ clip (torch.tensor): Video clip to be cropped. Size is (T, C, H, W)
67
+ i (int): i in (i,j) i.e coordinates of the upper left corner.
68
+ j (int): j in (i,j) i.e coordinates of the upper left corner.
69
+ h (int): Height of the cropped region.
70
+ w (int): Width of the cropped region.
71
+ size (tuple(int, int)): height and width of resized clip
72
+ Returns:
73
+ clip (torch.tensor): Resized and cropped clip. Size is (T, C, H, W)
74
+ """
75
+ if not _is_tensor_video_clip(clip):
76
+ raise ValueError("clip should be a 4D torch.tensor")
77
+ clip = crop(clip, i, j, h, w)
78
+ clip = resize(clip, size, interpolation_mode)
79
+ return clip
80
+
81
+
82
+ def center_crop(clip, crop_size):
83
+ if not _is_tensor_video_clip(clip):
84
+ raise ValueError("clip should be a 4D torch.tensor")
85
+ h, w = clip.size(-2), clip.size(-1)
86
+ th, tw = crop_size
87
+ if h < th or w < tw:
88
+ raise ValueError("height and width must be no smaller than crop_size")
89
+
90
+ i = int(round((h - th) / 2.0))
91
+ j = int(round((w - tw) / 2.0))
92
+ return crop(clip, i, j, th, tw)
93
+
94
+
95
+ def center_crop_using_short_edge(clip):
96
+ if not _is_tensor_video_clip(clip):
97
+ raise ValueError("clip should be a 4D torch.tensor")
98
+ h, w = clip.size(-2), clip.size(-1)
99
+ if h < w:
100
+ th, tw = h, h
101
+ i = 0
102
+ j = int(round((w - tw) / 2.0))
103
+ else:
104
+ th, tw = w, w
105
+ i = int(round((h - th) / 2.0))
106
+ j = 0
107
+ return crop(clip, i, j, th, tw)
108
+
109
+
110
+ def random_shift_crop(clip):
111
+ '''
112
+ Slide along the long edge, with the short edge as crop size
113
+ '''
114
+ if not _is_tensor_video_clip(clip):
115
+ raise ValueError("clip should be a 4D torch.tensor")
116
+ h, w = clip.size(-2), clip.size(-1)
117
+
118
+ if h <= w:
119
+ long_edge = w
120
+ short_edge = h
121
+ else:
122
+ long_edge = h
123
+ short_edge = w
124
+
125
+ th, tw = short_edge, short_edge
126
+
127
+ i = torch.randint(0, h - th + 1, size=(1,)).item()
128
+ j = torch.randint(0, w - tw + 1, size=(1,)).item()
129
+ return crop(clip, i, j, th, tw)
130
+
131
+
132
+ def to_tensor(clip):
133
+ """
134
+ Convert tensor data type from uint8 to float, divide value by 255.0 and
135
+ permute the dimensions of clip tensor
136
+ Args:
137
+ clip (torch.tensor, dtype=torch.uint8): Size is (T, C, H, W)
138
+ Return:
139
+ clip (torch.tensor, dtype=torch.float): Size is (T, C, H, W)
140
+ """
141
+ _is_tensor_video_clip(clip)
142
+ if not clip.dtype == torch.uint8:
143
+ raise TypeError("clip tensor should have data type uint8. Got %s" % str(clip.dtype))
144
+ # return clip.float().permute(3, 0, 1, 2) / 255.0
145
+ return clip.float() / 255.0
146
+
147
+
148
+ def normalize(clip, mean, std, inplace=False):
149
+ """
150
+ Args:
151
+ clip (torch.tensor): Video clip to be normalized. Size is (T, C, H, W)
152
+ mean (tuple): pixel RGB mean. Size is (3)
153
+ std (tuple): pixel standard deviation. Size is (3)
154
+ Returns:
155
+ normalized clip (torch.tensor): Size is (T, C, H, W)
156
+ """
157
+ if not _is_tensor_video_clip(clip):
158
+ raise ValueError("clip should be a 4D torch.tensor")
159
+ if not inplace:
160
+ clip = clip.clone()
161
+ mean = torch.as_tensor(mean, dtype=clip.dtype, device=clip.device)
162
+ # print(mean)
163
+ std = torch.as_tensor(std, dtype=clip.dtype, device=clip.device)
164
+ clip.sub_(mean[:, None, None, None]).div_(std[:, None, None, None])
165
+ return clip
166
+
167
+
168
+ def hflip(clip):
169
+ """
170
+ Args:
171
+ clip (torch.tensor): Video clip to be normalized. Size is (T, C, H, W)
172
+ Returns:
173
+ flipped clip (torch.tensor): Size is (T, C, H, W)
174
+ """
175
+ if not _is_tensor_video_clip(clip):
176
+ raise ValueError("clip should be a 4D torch.tensor")
177
+ return clip.flip(-1)
178
+
179
+
180
+ class RandomCropVideo:
181
+ def __init__(self, size):
182
+ if isinstance(size, numbers.Number):
183
+ self.size = (int(size), int(size))
184
+ else:
185
+ self.size = size
186
+
187
+ def __call__(self, clip):
188
+ """
189
+ Args:
190
+ clip (torch.tensor): Video clip to be cropped. Size is (T, C, H, W)
191
+ Returns:
192
+ torch.tensor: randomly cropped video clip.
193
+ size is (T, C, OH, OW)
194
+ """
195
+ i, j, h, w = self.get_params(clip)
196
+ return crop(clip, i, j, h, w)
197
+
198
+ def get_params(self, clip):
199
+ h, w = clip.shape[-2:]
200
+ th, tw = self.size
201
+
202
+ if h < th or w < tw:
203
+ raise ValueError(f"Required crop size {(th, tw)} is larger than input image size {(h, w)}")
204
+
205
+ if w == tw and h == th:
206
+ return 0, 0, h, w
207
+
208
+ i = torch.randint(0, h - th + 1, size=(1,)).item()
209
+ j = torch.randint(0, w - tw + 1, size=(1,)).item()
210
+
211
+ return i, j, th, tw
212
+
213
+ def __repr__(self) -> str:
214
+ return f"{self.__class__.__name__}(size={self.size})"
215
+
216
+
217
+ class SpatialStrideCropVideo:
218
+ def __init__(self, stride):
219
+ self.stride = stride
220
+
221
+ def __call__(self, clip):
222
+ """
223
+ Args:
224
+ clip (torch.tensor): Video clip to be cropped. Size is (T, C, H, W)
225
+ Returns:
226
+ torch.tensor: cropped video clip by stride.
227
+ size is (T, C, OH, OW)
228
+ """
229
+ i, j, h, w = self.get_params(clip)
230
+ return crop(clip, i, j, h, w)
231
+
232
+ def get_params(self, clip):
233
+ h, w = clip.shape[-2:]
234
+
235
+ th, tw = h // self.stride * self.stride, w // self.stride * self.stride
236
+
237
+ return 0, 0, th, tw # from top-left
238
+
239
+ def __repr__(self) -> str:
240
+ return f"{self.__class__.__name__}(size={self.size})"
241
+
242
+ class LongSideResizeVideo:
243
+ '''
244
+ First use the long side,
245
+ then resize to the specified size
246
+ '''
247
+
248
+ def __init__(
249
+ self,
250
+ size,
251
+ skip_low_resolution=False,
252
+ interpolation_mode="bilinear",
253
+ ):
254
+ self.size = size
255
+ self.skip_low_resolution = skip_low_resolution
256
+ self.interpolation_mode = interpolation_mode
257
+
258
+ def __call__(self, clip):
259
+ """
260
+ Args:
261
+ clip (torch.tensor): Video clip to be cropped. Size is (T, C, H, W)
262
+ Returns:
263
+ torch.tensor: scale resized video clip.
264
+ size is (T, C, 512, *) or (T, C, *, 512)
265
+ """
266
+ _, _, h, w = clip.shape
267
+ if self.skip_low_resolution and max(h, w) <= self.size:
268
+ return clip
269
+ if h > w:
270
+ w = int(w * self.size / h)
271
+ h = self.size
272
+ else:
273
+ h = int(h * self.size / w)
274
+ w = self.size
275
+ resize_clip = resize(clip, target_size=(h, w),
276
+ interpolation_mode=self.interpolation_mode)
277
+ return resize_clip
278
+
279
+ def __repr__(self) -> str:
280
+ return f"{self.__class__.__name__}(size={self.size}, interpolation_mode={self.interpolation_mode}"
281
+
282
+ class CenterCropResizeVideo:
283
+ '''
284
+ First use the short side for cropping length,
285
+ center crop video, then resize to the specified size
286
+ '''
287
+
288
+ def __init__(
289
+ self,
290
+ size,
291
+ interpolation_mode="bilinear",
292
+ ):
293
+ if isinstance(size, tuple):
294
+ if len(size) != 2:
295
+ raise ValueError(f"size should be tuple (height, width), instead got {size}")
296
+ self.size = size
297
+ else:
298
+ self.size = (size, size)
299
+
300
+ self.interpolation_mode = interpolation_mode
301
+
302
+ def __call__(self, clip):
303
+ """
304
+ Args:
305
+ clip (torch.tensor): Video clip to be cropped. Size is (T, C, H, W)
306
+ Returns:
307
+ torch.tensor: scale resized / center cropped video clip.
308
+ size is (T, C, crop_size, crop_size)
309
+ """
310
+ clip_center_crop = center_crop_using_short_edge(clip)
311
+ clip_center_crop_resize = resize(clip_center_crop, target_size=self.size,
312
+ interpolation_mode=self.interpolation_mode)
313
+ return clip_center_crop_resize
314
+
315
+ def __repr__(self) -> str:
316
+ return f"{self.__class__.__name__}(size={self.size}, interpolation_mode={self.interpolation_mode}"
317
+
318
+
319
+ class UCFCenterCropVideo:
320
+ '''
321
+ First scale to the specified size in equal proportion to the short edge,
322
+ then center cropping
323
+ '''
324
+
325
+ def __init__(
326
+ self,
327
+ size,
328
+ interpolation_mode="bilinear",
329
+ ):
330
+ if isinstance(size, tuple):
331
+ if len(size) != 2:
332
+ raise ValueError(f"size should be tuple (height, width), instead got {size}")
333
+ self.size = size
334
+ else:
335
+ self.size = (size, size)
336
+
337
+ self.interpolation_mode = interpolation_mode
338
+
339
+ def __call__(self, clip):
340
+ """
341
+ Args:
342
+ clip (torch.tensor): Video clip to be cropped. Size is (T, C, H, W)
343
+ Returns:
344
+ torch.tensor: scale resized / center cropped video clip.
345
+ size is (T, C, crop_size, crop_size)
346
+ """
347
+ clip_resize = resize_scale(clip=clip, target_size=self.size, interpolation_mode=self.interpolation_mode)
348
+ clip_center_crop = center_crop(clip_resize, self.size)
349
+ return clip_center_crop
350
+
351
+ def __repr__(self) -> str:
352
+ return f"{self.__class__.__name__}(size={self.size}, interpolation_mode={self.interpolation_mode}"
353
+
354
+
355
+ class KineticsRandomCropResizeVideo:
356
+ '''
357
+ Slide along the long edge, with the short edge as crop size. And resie to the desired size.
358
+ '''
359
+
360
+ def __init__(
361
+ self,
362
+ size,
363
+ interpolation_mode="bilinear",
364
+ ):
365
+ if isinstance(size, tuple):
366
+ if len(size) != 2:
367
+ raise ValueError(f"size should be tuple (height, width), instead got {size}")
368
+ self.size = size
369
+ else:
370
+ self.size = (size, size)
371
+
372
+ self.interpolation_mode = interpolation_mode
373
+
374
+ def __call__(self, clip):
375
+ clip_random_crop = random_shift_crop(clip)
376
+ clip_resize = resize(clip_random_crop, self.size, self.interpolation_mode)
377
+ return clip_resize
378
+
379
+
380
+ class CenterCropVideo:
381
+ def __init__(
382
+ self,
383
+ size,
384
+ interpolation_mode="bilinear",
385
+ ):
386
+ if isinstance(size, tuple):
387
+ if len(size) != 2:
388
+ raise ValueError(f"size should be tuple (height, width), instead got {size}")
389
+ self.size = size
390
+ else:
391
+ self.size = (size, size)
392
+
393
+ self.interpolation_mode = interpolation_mode
394
+
395
+ def __call__(self, clip):
396
+ """
397
+ Args:
398
+ clip (torch.tensor): Video clip to be cropped. Size is (T, C, H, W)
399
+ Returns:
400
+ torch.tensor: center cropped video clip.
401
+ size is (T, C, crop_size, crop_size)
402
+ """
403
+ clip_center_crop = center_crop(clip, self.size)
404
+ return clip_center_crop
405
+
406
+ def __repr__(self) -> str:
407
+ return f"{self.__class__.__name__}(size={self.size}, interpolation_mode={self.interpolation_mode}"
408
+
409
+
410
+ class NormalizeVideo:
411
+ """
412
+ Normalize the video clip by mean subtraction and division by standard deviation
413
+ Args:
414
+ mean (3-tuple): pixel RGB mean
415
+ std (3-tuple): pixel RGB standard deviation
416
+ inplace (boolean): whether do in-place normalization
417
+ """
418
+
419
+ def __init__(self, mean, std, inplace=False):
420
+ self.mean = mean
421
+ self.std = std
422
+ self.inplace = inplace
423
+
424
+ def __call__(self, clip):
425
+ """
426
+ Args:
427
+ clip (torch.tensor): video clip must be normalized. Size is (C, T, H, W)
428
+ """
429
+ return normalize(clip, self.mean, self.std, self.inplace)
430
+
431
+ def __repr__(self) -> str:
432
+ return f"{self.__class__.__name__}(mean={self.mean}, std={self.std}, inplace={self.inplace})"
433
+
434
+
435
+ class ToTensorVideo:
436
+ """
437
+ Convert tensor data type from uint8 to float, divide value by 255.0 and
438
+ permute the dimensions of clip tensor
439
+ """
440
+
441
+ def __init__(self):
442
+ pass
443
+
444
+ def __call__(self, clip):
445
+ """
446
+ Args:
447
+ clip (torch.tensor, dtype=torch.uint8): Size is (T, C, H, W)
448
+ Return:
449
+ clip (torch.tensor, dtype=torch.float): Size is (T, C, H, W)
450
+ """
451
+ return to_tensor(clip)
452
+
453
+ def __repr__(self) -> str:
454
+ return self.__class__.__name__
455
+
456
+
457
+ class RandomHorizontalFlipVideo:
458
+ """
459
+ Flip the video clip along the horizontal direction with a given probability
460
+ Args:
461
+ p (float): probability of the clip being flipped. Default value is 0.5
462
+ """
463
+
464
+ def __init__(self, p=0.5):
465
+ self.p = p
466
+
467
+ def __call__(self, clip):
468
+ """
469
+ Args:
470
+ clip (torch.tensor): Size is (T, C, H, W)
471
+ Return:
472
+ clip (torch.tensor): Size is (T, C, H, W)
473
+ """
474
+ if random.random() < self.p:
475
+ clip = hflip(clip)
476
+ return clip
477
+
478
+ def __repr__(self) -> str:
479
+ return f"{self.__class__.__name__}(p={self.p})"
480
+
481
+
482
+ # ------------------------------------------------------------
483
+ # --------------------- Sampling ---------------------------
484
+ # ------------------------------------------------------------
485
+ class TemporalRandomCrop(object):
486
+ """Temporally crop the given frame indices at a random location.
487
+
488
+ Args:
489
+ size (int): Desired length of frames will be seen in the model.
490
+ """
491
+
492
+ def __init__(self, size):
493
+ self.size = size
494
+
495
+ def __call__(self, total_frames):
496
+ rand_end = max(0, total_frames - self.size - 1)
497
+ begin_index = random.randint(0, rand_end)
498
+ end_index = min(begin_index + self.size, total_frames)
499
+ return begin_index, end_index
500
+
501
+ class DynamicSampleDuration(object):
502
+ """Temporally crop the given frame indices at a random location.
503
+
504
+ Args:
505
+ size (int): Desired length of frames will be seen in the model.
506
+ """
507
+
508
+ def __init__(self, t_stride, extra_1):
509
+ self.t_stride = t_stride
510
+ self.extra_1 = extra_1
511
+
512
+ def __call__(self, t, h, w):
513
+ if self.extra_1:
514
+ t = t - 1
515
+ truncate_t_list = list(range(t+1))[t//2:][::self.t_stride] # need half at least
516
+ truncate_t = random.choice(truncate_t_list)
517
+ if self.extra_1:
518
+ truncate_t = truncate_t + 1
519
+ return 0, truncate_t
520
+
521
+ if __name__ == '__main__':
522
+ from torchvision import transforms
523
+ import torchvision.io as io
524
+ import numpy as np
525
+ from torchvision.utils import save_image
526
+ import os
527
+
528
+ vframes, aframes, info = io.read_video(
529
+ filename='./v_Archery_g01_c03.avi',
530
+ pts_unit='sec',
531
+ output_format='TCHW'
532
+ )
533
+
534
+ trans = transforms.Compose([
535
+ ToTensorVideo(),
536
+ RandomHorizontalFlipVideo(),
537
+ UCFCenterCropVideo(512),
538
+ # NormalizeVideo(mean=[0.5, 0.5, 0.5], std=[0.5, 0.5, 0.5], inplace=True),
539
+ transforms.Normalize(mean=[0.5, 0.5, 0.5], std=[0.5, 0.5, 0.5], inplace=True)
540
+ ])
541
+
542
+ target_video_len = 32
543
+ frame_interval = 1
544
+ total_frames = len(vframes)
545
+ print(total_frames)
546
+
547
+ temporal_sample = TemporalRandomCrop(target_video_len * frame_interval)
548
+
549
+ # Sampling video frames
550
+ start_frame_ind, end_frame_ind = temporal_sample(total_frames)
551
+ # print(start_frame_ind)
552
+ # print(end_frame_ind)
553
+ assert end_frame_ind - start_frame_ind >= target_video_len
554
+ frame_indice = np.linspace(start_frame_ind, end_frame_ind - 1, target_video_len, dtype=int)
555
+ print(frame_indice)
556
+
557
+ select_vframes = vframes[frame_indice]
558
+ print(select_vframes.shape)
559
+ print(select_vframes.dtype)
560
+
561
+ select_vframes_trans = trans(select_vframes)
562
+ print(select_vframes_trans.shape)
563
+ print(select_vframes_trans.dtype)
564
+
565
+ select_vframes_trans_int = ((select_vframes_trans * 0.5 + 0.5) * 255).to(dtype=torch.uint8)
566
+ print(select_vframes_trans_int.dtype)
567
+ print(select_vframes_trans_int.permute(0, 2, 3, 1).shape)
568
+
569
+ io.write_video('./test.avi', select_vframes_trans_int.permute(0, 2, 3, 1), fps=8)
570
+
571
+ for i in range(target_video_len):
572
+ save_image(select_vframes_trans[i], os.path.join('./test000', '%04d.png' % i), normalize=True,
573
+ value_range=(-1, 1))
causalvideovae/eval/RAFT/chairs_split.txt ADDED
@@ -0,0 +1,22872 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ 1
2
+ 1
3
+ 1
4
+ 1
5
+ 1
6
+ 2
7
+ 1
8
+ 1
9
+ 1
10
+ 1
11
+ 1
12
+ 1
13
+ 1
14
+ 1
15
+ 1
16
+ 1
17
+ 1
18
+ 2
19
+ 1
20
+ 1
21
+ 1
22
+ 1
23
+ 1
24
+ 1
25
+ 1
26
+ 1
27
+ 1
28
+ 1
29
+ 1
30
+ 1
31
+ 1
32
+ 1
33
+ 1
34
+ 1
35
+ 1
36
+ 1
37
+ 1
38
+ 1
39
+ 1
40
+ 1
41
+ 1
42
+ 1
43
+ 2
44
+ 1
45
+ 1
46
+ 2
47
+ 1
48
+ 1
49
+ 1
50
+ 1
51
+ 1
52
+ 1
53
+ 1
54
+ 1
55
+ 1
56
+ 1
57
+ 1
58
+ 1
59
+ 2
60
+ 1
61
+ 1
62
+ 1
63
+ 2
64
+ 1
65
+ 1
66
+ 1
67
+ 1
68
+ 1
69
+ 1
70
+ 1
71
+ 1
72
+ 1
73
+ 1
74
+ 1
75
+ 1
76
+ 1
77
+ 1
78
+ 1
79
+ 1
80
+ 1
81
+ 1
82
+ 1
83
+ 1
84
+ 1
85
+ 1
86
+ 1
87
+ 1
88
+ 1
89
+ 1
90
+ 1
91
+ 1
92
+ 1
93
+ 1
94
+ 1
95
+ 1
96
+ 1
97
+ 2
98
+ 1
99
+ 1
100
+ 1
101
+ 1
102
+ 1
103
+ 1
104
+ 1
105
+ 1
106
+ 1
107
+ 1
108
+ 1
109
+ 1
110
+ 1
111
+ 1
112
+ 2
113
+ 1
114
+ 1
115
+ 1
116
+ 1
117
+ 1
118
+ 2
119
+ 1
120
+ 1
121
+ 2
122
+ 2
123
+ 1
124
+ 1
125
+ 1
126
+ 1
127
+ 1
128
+ 1
129
+ 1
130
+ 1
131
+ 1
132
+ 2
133
+ 2
134
+ 1
135
+ 1
136
+ 1
137
+ 1
138
+ 1
139
+ 1
140
+ 1
141
+ 1
142
+ 1
143
+ 1
144
+ 1
145
+ 1
146
+ 1
147
+ 1
148
+ 1
149
+ 1
150
+ 1
151
+ 1
152
+ 1
153
+ 2
154
+ 1
155
+ 1
156
+ 1
157
+ 1
158
+ 1
159
+ 1
160
+ 1
161
+ 2
162
+ 1
163
+ 1
164
+ 1
165
+ 1
166
+ 1
167
+ 1
168
+ 1
169
+ 1
170
+ 1
171
+ 1
172
+ 1
173
+ 1
174
+ 1
175
+ 1
176
+ 1
177
+ 1
178
+ 1
179
+ 1
180
+ 1
181
+ 1
182
+ 1
183
+ 1
184
+ 1
185
+ 1
186
+ 1
187
+ 1
188
+ 1
189
+ 1
190
+ 1
191
+ 1
192
+ 1
193
+ 1
194
+ 1
195
+ 1
196
+ 1
197
+ 1
198
+ 1
199
+ 1
200
+ 1
201
+ 1
202
+ 1
203
+ 1
204
+ 1
205
+ 1
206
+ 1
207
+ 1
208
+ 1
209
+ 1
210
+ 1
211
+ 1
212
+ 1
213
+ 1
214
+ 1
215
+ 1
216
+ 1
217
+ 1
218
+ 1
219
+ 1
220
+ 1
221
+ 1
222
+ 1
223
+ 1
224
+ 1
225
+ 1
226
+ 1
227
+ 1
228
+ 1
229
+ 1
230
+ 1
231
+ 1
232
+ 1
233
+ 1
234
+ 1
235
+ 1
236
+ 1
237
+ 1
238
+ 1
239
+ 1
240
+ 1
241
+ 1
242
+ 1
243
+ 1
244
+ 1
245
+ 1
246
+ 1
247
+ 1
248
+ 1
249
+ 2
250
+ 1
251
+ 1
252
+ 1
253
+ 1
254
+ 1
255
+ 1
256
+ 1
257
+ 1
258
+ 1
259
+ 1
260
+ 1
261
+ 1
262
+ 1
263
+ 1
264
+ 2
265
+ 2
266
+ 1
267
+ 1
268
+ 1
269
+ 1
270
+ 1
271
+ 1
272
+ 1
273
+ 1
274
+ 1
275
+ 1
276
+ 1
277
+ 1
278
+ 1
279
+ 1
280
+ 1
281
+ 1
282
+ 1
283
+ 1
284
+ 1
285
+ 1
286
+ 1
287
+ 1
288
+ 1
289
+ 1
290
+ 1
291
+ 1
292
+ 2
293
+ 1
294
+ 2
295
+ 1
296
+ 2
297
+ 1
298
+ 1
299
+ 1
300
+ 2
301
+ 1
302
+ 1
303
+ 1
304
+ 1
305
+ 1
306
+ 1
307
+ 1
308
+ 1
309
+ 1
310
+ 1
311
+ 1
312
+ 1
313
+ 1
314
+ 1
315
+ 1
316
+ 1
317
+ 2
318
+ 1
319
+ 1
320
+ 1
321
+ 2
322
+ 1
323
+ 1
324
+ 1
325
+ 1
326
+ 1
327
+ 1
328
+ 1
329
+ 1
330
+ 1
331
+ 1
332
+ 1
333
+ 1
334
+ 1
335
+ 1
336
+ 1
337
+ 2
338
+ 2
339
+ 1
340
+ 1
341
+ 1
342
+ 1
343
+ 1
344
+ 2
345
+ 1
346
+ 1
347
+ 1
348
+ 1
349
+ 1
350
+ 1
351
+ 1
352
+ 1
353
+ 1
354
+ 1
355
+ 1
356
+ 1
357
+ 1
358
+ 1
359
+ 2
360
+ 1
361
+ 1
362
+ 1
363
+ 1
364
+ 1
365
+ 1
366
+ 1
367
+ 1
368
+ 1
369
+ 1
370
+ 1
371
+ 1
372
+ 1
373
+ 1
374
+ 1
375
+ 1
376
+ 1
377
+ 1
378
+ 1
379
+ 1
380
+ 1
381
+ 1
382
+ 1
383
+ 1
384
+ 1
385
+ 1
386
+ 1
387
+ 1
388
+ 1
389
+ 1
390
+ 1
391
+ 1
392
+ 1
393
+ 1
394
+ 1
395
+ 1
396
+ 1
397
+ 1
398
+ 1
399
+ 1
400
+ 2
401
+ 1
402
+ 2
403
+ 1
404
+ 1
405
+ 1
406
+ 1
407
+ 1
408
+ 1
409
+ 1
410
+ 1
411
+ 1
412
+ 1
413
+ 1
414
+ 1
415
+ 1
416
+ 1
417
+ 1
418
+ 1
419
+ 1
420
+ 1
421
+ 1
422
+ 1
423
+ 1
424
+ 1
425
+ 1
426
+ 1
427
+ 1
428
+ 1
429
+ 1
430
+ 2
431
+ 1
432
+ 1
433
+ 1
434
+ 1
435
+ 1
436
+ 1
437
+ 1
438
+ 1
439
+ 2
440
+ 1
441
+ 1
442
+ 1
443
+ 1
444
+ 1
445
+ 1
446
+ 1
447
+ 1
448
+ 1
449
+ 1
450
+ 1
451
+ 1
452
+ 1
453
+ 1
454
+ 1
455
+ 1
456
+ 1
457
+ 1
458
+ 1
459
+ 1
460
+ 1
461
+ 1
462
+ 1
463
+ 1
464
+ 1
465
+ 1
466
+ 1
467
+ 1
468
+ 1
469
+ 2
470
+ 1
471
+ 1
472
+ 1
473
+ 1
474
+ 1
475
+ 1
476
+ 1
477
+ 2
478
+ 1
479
+ 1
480
+ 1
481
+ 1
482
+ 1
483
+ 1
484
+ 1
485
+ 1
486
+ 1
487
+ 1
488
+ 1
489
+ 1
490
+ 1
491
+ 1
492
+ 1
493
+ 1
494
+ 1
495
+ 2
496
+ 1
497
+ 1
498
+ 1
499
+ 1
500
+ 1
501
+ 1
502
+ 1
503
+ 1
504
+ 1
505
+ 1
506
+ 1
507
+ 1
508
+ 1
509
+ 1
510
+ 2
511
+ 1
512
+ 1
513
+ 1
514
+ 1
515
+ 1
516
+ 1
517
+ 1
518
+ 1
519
+ 1
520
+ 1
521
+ 1
522
+ 1
523
+ 1
524
+ 1
525
+ 1
526
+ 1
527
+ 1
528
+ 1
529
+ 2
530
+ 1
531
+ 1
532
+ 2
533
+ 1
534
+ 1
535
+ 1
536
+ 1
537
+ 1
538
+ 1
539
+ 1
540
+ 1
541
+ 1
542
+ 1
543
+ 1
544
+ 1
545
+ 1
546
+ 1
547
+ 1
548
+ 1
549
+ 1
550
+ 1
551
+ 1
552
+ 1
553
+ 1
554
+ 1
555
+ 1
556
+ 1
557
+ 1
558
+ 1
559
+ 1
560
+ 1
561
+ 1
562
+ 1
563
+ 1
564
+ 1
565
+ 1
566
+ 1
567
+ 1
568
+ 1
569
+ 1
570
+ 1
571
+ 1
572
+ 1
573
+ 2
574
+ 1
575
+ 1
576
+ 1
577
+ 1
578
+ 1
579
+ 1
580
+ 1
581
+ 1
582
+ 2
583
+ 1
584
+ 2
585
+ 1
586
+ 1
587
+ 1
588
+ 1
589
+ 2
590
+ 1
591
+ 1
592
+ 1
593
+ 1
594
+ 2
595
+ 1
596
+ 1
597
+ 1
598
+ 1
599
+ 1
600
+ 1
601
+ 1
602
+ 1
603
+ 1
604
+ 1
605
+ 1
606
+ 1
607
+ 1
608
+ 1
609
+ 1
610
+ 1
611
+ 1
612
+ 1
613
+ 1
614
+ 1
615
+ 1
616
+ 1
617
+ 1
618
+ 1
619
+ 1
620
+ 1
621
+ 1
622
+ 1
623
+ 1
624
+ 1
625
+ 1
626
+ 1
627
+ 1
628
+ 1
629
+ 1
630
+ 1
631
+ 1
632
+ 1
633
+ 1
634
+ 1
635
+ 1
636
+ 1
637
+ 1
638
+ 1
639
+ 1
640
+ 1
641
+ 1
642
+ 1
643
+ 1
644
+ 1
645
+ 1
646
+ 1
647
+ 1
648
+ 1
649
+ 1
650
+ 1
651
+ 1
652
+ 1
653
+ 1
654
+ 1
655
+ 1
656
+ 1
657
+ 1
658
+ 1
659
+ 1
660
+ 1
661
+ 1
662
+ 1
663
+ 1
664
+ 1
665
+ 1
666
+ 1
667
+ 1
668
+ 1
669
+ 1
670
+ 1
671
+ 1
672
+ 1
673
+ 1
674
+ 1
675
+ 1
676
+ 1
677
+ 1
678
+ 1
679
+ 1
680
+ 1
681
+ 1
682
+ 2
683
+ 1
684
+ 1
685
+ 1
686
+ 1
687
+ 1
688
+ 1
689
+ 2
690
+ 1
691
+ 1
692
+ 1
693
+ 1
694
+ 1
695
+ 1
696
+ 1
697
+ 2
698
+ 1
699
+ 1
700
+ 1
701
+ 1
702
+ 1
703
+ 1
704
+ 1
705
+ 1
706
+ 1
707
+ 1
708
+ 1
709
+ 1
710
+ 1
711
+ 1
712
+ 1
713
+ 1
714
+ 1
715
+ 2
716
+ 1
717
+ 1
718
+ 1
719
+ 1
720
+ 1
721
+ 1
722
+ 1
723
+ 1
724
+ 1
725
+ 1
726
+ 1
727
+ 1
728
+ 1
729
+ 1
730
+ 1
731
+ 1
732
+ 1
733
+ 1
734
+ 1
735
+ 1
736
+ 1
737
+ 1
738
+ 1
739
+ 1
740
+ 1
741
+ 1
742
+ 1
743
+ 1
744
+ 1
745
+ 1
746
+ 1
747
+ 1
748
+ 1
749
+ 1
750
+ 1
751
+ 1
752
+ 1
753
+ 1
754
+ 1
755
+ 1
756
+ 1
757
+ 1
758
+ 1
759
+ 1
760
+ 1
761
+ 1
762
+ 1
763
+ 1
764
+ 1
765
+ 1
766
+ 1
767
+ 1
768
+ 2
769
+ 1
770
+ 1
771
+ 1
772
+ 1
773
+ 1
774
+ 1
775
+ 1
776
+ 1
777
+ 1
778
+ 1
779
+ 1
780
+ 1
781
+ 1
782
+ 1
783
+ 1
784
+ 1
785
+ 1
786
+ 1
787
+ 2
788
+ 1
789
+ 1
790
+ 1
791
+ 1
792
+ 1
793
+ 1
794
+ 1
795
+ 1
796
+ 1
797
+ 1
798
+ 1
799
+ 1
800
+ 1
801
+ 1
802
+ 1
803
+ 1
804
+ 1
805
+ 1
806
+ 1
807
+ 1
808
+ 1
809
+ 1
810
+ 1
811
+ 2
812
+ 1
813
+ 1
814
+ 1
815
+ 1
816
+ 1
817
+ 1
818
+ 1
819
+ 1
820
+ 1
821
+ 1
822
+ 1
823
+ 1
824
+ 1
825
+ 1
826
+ 2
827
+ 1
828
+ 1
829
+ 1
830
+ 1
831
+ 1
832
+ 1
833
+ 1
834
+ 1
835
+ 1
836
+ 1
837
+ 2
838
+ 1
839
+ 1
840
+ 1
841
+ 1
842
+ 2
843
+ 1
844
+ 1
845
+ 1
846
+ 1
847
+ 1
848
+ 1
849
+ 1
850
+ 1
851
+ 1
852
+ 1
853
+ 1
854
+ 1
855
+ 1
856
+ 1
857
+ 1
858
+ 1
859
+ 1
860
+ 1
861
+ 1
862
+ 1
863
+ 1
864
+ 1
865
+ 1
866
+ 1
867
+ 1
868
+ 1
869
+ 1
870
+ 1
871
+ 1
872
+ 1
873
+ 1
874
+ 1
875
+ 1
876
+ 1
877
+ 1
878
+ 1
879
+ 1
880
+ 1
881
+ 1
882
+ 1
883
+ 1
884
+ 2
885
+ 1
886
+ 1
887
+ 1
888
+ 1
889
+ 1
890
+ 1
891
+ 1
892
+ 1
893
+ 1
894
+ 1
895
+ 1
896
+ 1
897
+ 1
898
+ 1
899
+ 1
900
+ 1
901
+ 1
902
+ 1
903
+ 1
904
+ 1
905
+ 1
906
+ 1
907
+ 1
908
+ 1
909
+ 1
910
+ 1
911
+ 1
912
+ 1
913
+ 1
914
+ 1
915
+ 1
916
+ 1
917
+ 1
918
+ 2
919
+ 1
920
+ 1
921
+ 1
922
+ 1
923
+ 1
924
+ 1
925
+ 1
926
+ 1
927
+ 1
928
+ 1
929
+ 1
930
+ 1
931
+ 1
932
+ 1
933
+ 1
934
+ 1
935
+ 1
936
+ 1
937
+ 1
938
+ 2
939
+ 1
940
+ 1
941
+ 1
942
+ 1
943
+ 2
944
+ 1
945
+ 1
946
+ 1
947
+ 1
948
+ 1
949
+ 1
950
+ 1
951
+ 1
952
+ 1
953
+ 1
954
+ 1
955
+ 1
956
+ 1
957
+ 1
958
+ 1
959
+ 1
960
+ 1
961
+ 1
962
+ 1
963
+ 1
964
+ 1
965
+ 1
966
+ 1
967
+ 1
968
+ 1
969
+ 1
970
+ 1
971
+ 2
972
+ 1
973
+ 1
974
+ 1
975
+ 2
976
+ 1
977
+ 1
978
+ 1
979
+ 1
980
+ 1
981
+ 2
982
+ 1
983
+ 1
984
+ 1
985
+ 1
986
+ 1
987
+ 1
988
+ 1
989
+ 1
990
+ 1
991
+ 1
992
+ 1
993
+ 1
994
+ 1
995
+ 1
996
+ 1
997
+ 1
998
+ 1
999
+ 1
1000
+ 1
1001
+ 1
1002
+ 1
1003
+ 1
1004
+ 1
1005
+ 1
1006
+ 1
1007
+ 1
1008
+ 1
1009
+ 1
1010
+ 1
1011
+ 1
1012
+ 1
1013
+ 1
1014
+ 1
1015
+ 1
1016
+ 1
1017
+ 2
1018
+ 1
1019
+ 1
1020
+ 1
1021
+ 1
1022
+ 1
1023
+ 1
1024
+ 1
1025
+ 1
1026
+ 1
1027
+ 1
1028
+ 1
1029
+ 1
1030
+ 1
1031
+ 1
1032
+ 1
1033
+ 1
1034
+ 1
1035
+ 1
1036
+ 1
1037
+ 1
1038
+ 1
1039
+ 1
1040
+ 1
1041
+ 1
1042
+ 1
1043
+ 1
1044
+ 2
1045
+ 1
1046
+ 1
1047
+ 1
1048
+ 1
1049
+ 1
1050
+ 1
1051
+ 1
1052
+ 1
1053
+ 1
1054
+ 1
1055
+ 1
1056
+ 1
1057
+ 1
1058
+ 1
1059
+ 1
1060
+ 1
1061
+ 1
1062
+ 1
1063
+ 1
1064
+ 1
1065
+ 2
1066
+ 1
1067
+ 1
1068
+ 1
1069
+ 1
1070
+ 1
1071
+ 1
1072
+ 1
1073
+ 1
1074
+ 1
1075
+ 1
1076
+ 1
1077
+ 1
1078
+ 1
1079
+ 1
1080
+ 1
1081
+ 1
1082
+ 1
1083
+ 1
1084
+ 1
1085
+ 1
1086
+ 1
1087
+ 1
1088
+ 1
1089
+ 1
1090
+ 1
1091
+ 1
1092
+ 1
1093
+ 1
1094
+ 1
1095
+ 1
1096
+ 1
1097
+ 1
1098
+ 1
1099
+ 1
1100
+ 1
1101
+ 1
1102
+ 1
1103
+ 1
1104
+ 1
1105
+ 1
1106
+ 1
1107
+ 1
1108
+ 1
1109
+ 1
1110
+ 1
1111
+ 1
1112
+ 1
1113
+ 1
1114
+ 1
1115
+ 1
1116
+ 1
1117
+ 1
1118
+ 1
1119
+ 2
1120
+ 1
1121
+ 1
1122
+ 2
1123
+ 1
1124
+ 1
1125
+ 1
1126
+ 1
1127
+ 1
1128
+ 1
1129
+ 1
1130
+ 1
1131
+ 1
1132
+ 1
1133
+ 1
1134
+ 2
1135
+ 1
1136
+ 1
1137
+ 1
1138
+ 1
1139
+ 1
1140
+ 1
1141
+ 1
1142
+ 1
1143
+ 1
1144
+ 1
1145
+ 1
1146
+ 1
1147
+ 1
1148
+ 1
1149
+ 1
1150
+ 1
1151
+ 1
1152
+ 1
1153
+ 1
1154
+ 2
1155
+ 1
1156
+ 2
1157
+ 1
1158
+ 1
1159
+ 2
1160
+ 2
1161
+ 1
1162
+ 1
1163
+ 1
1164
+ 1
1165
+ 1
1166
+ 1
1167
+ 1
1168
+ 1
1169
+ 1
1170
+ 1
1171
+ 1
1172
+ 1
1173
+ 1
1174
+ 2
1175
+ 1
1176
+ 1
1177
+ 1
1178
+ 1
1179
+ 1
1180
+ 1
1181
+ 1
1182
+ 1
1183
+ 1
1184
+ 1
1185
+ 1
1186
+ 1
1187
+ 1
1188
+ 2
1189
+ 1
1190
+ 1
1191
+ 1
1192
+ 1
1193
+ 1
1194
+ 1
1195
+ 1
1196
+ 1
1197
+ 1
1198
+ 1
1199
+ 1
1200
+ 1
1201
+ 1
1202
+ 1
1203
+ 1
1204
+ 1
1205
+ 1
1206
+ 1
1207
+ 1
1208
+ 1
1209
+ 1
1210
+ 1
1211
+ 1
1212
+ 1
1213
+ 1
1214
+ 1
1215
+ 1
1216
+ 1
1217
+ 1
1218
+ 1
1219
+ 1
1220
+ 2
1221
+ 1
1222
+ 1
1223
+ 1
1224
+ 1
1225
+ 1
1226
+ 1
1227
+ 1
1228
+ 1
1229
+ 1
1230
+ 1
1231
+ 1
1232
+ 1
1233
+ 1
1234
+ 1
1235
+ 1
1236
+ 1
1237
+ 1
1238
+ 2
1239
+ 2
1240
+ 1
1241
+ 1
1242
+ 1
1243
+ 1
1244
+ 1
1245
+ 1
1246
+ 1
1247
+ 1
1248
+ 1
1249
+ 1
1250
+ 1
1251
+ 1
1252
+ 1
1253
+ 1
1254
+ 1
1255
+ 1
1256
+ 1
1257
+ 1
1258
+ 1
1259
+ 1
1260
+ 2
1261
+ 1
1262
+ 1
1263
+ 1
1264
+ 1
1265
+ 1
1266
+ 1
1267
+ 2
1268
+ 1
1269
+ 1
1270
+ 1
1271
+ 1
1272
+ 1
1273
+ 1
1274
+ 1
1275
+ 1
1276
+ 1
1277
+ 1
1278
+ 1
1279
+ 2
1280
+ 1
1281
+ 1
1282
+ 1
1283
+ 1
1284
+ 1
1285
+ 1
1286
+ 1
1287
+ 1
1288
+ 1
1289
+ 1
1290
+ 1
1291
+ 1
1292
+ 1
1293
+ 1
1294
+ 1
1295
+ 1
1296
+ 1
1297
+ 2
1298
+ 1
1299
+ 1
1300
+ 1
1301
+ 1
1302
+ 1
1303
+ 1
1304
+ 1
1305
+ 1
1306
+ 1
1307
+ 1
1308
+ 1
1309
+ 1
1310
+ 1
1311
+ 1
1312
+ 1
1313
+ 1
1314
+ 1
1315
+ 1
1316
+ 1
1317
+ 1
1318
+ 1
1319
+ 1
1320
+ 1
1321
+ 1
1322
+ 1
1323
+ 1
1324
+ 1
1325
+ 1
1326
+ 1
1327
+ 1
1328
+ 1
1329
+ 1
1330
+ 1
1331
+ 1
1332
+ 1
1333
+ 1
1334
+ 1
1335
+ 1
1336
+ 1
1337
+ 1
1338
+ 1
1339
+ 1
1340
+ 1
1341
+ 1
1342
+ 1
1343
+ 1
1344
+ 1
1345
+ 1
1346
+ 1
1347
+ 1
1348
+ 1
1349
+ 1
1350
+ 1
1351
+ 1
1352
+ 1
1353
+ 1
1354
+ 1
1355
+ 2
1356
+ 1
1357
+ 1
1358
+ 1
1359
+ 1
1360
+ 1
1361
+ 1
1362
+ 1
1363
+ 1
1364
+ 1
1365
+ 1
1366
+ 1
1367
+ 1
1368
+ 1
1369
+ 1
1370
+ 1
1371
+ 1
1372
+ 1
1373
+ 1
1374
+ 1
1375
+ 1
1376
+ 1
1377
+ 1
1378
+ 1
1379
+ 2
1380
+ 1
1381
+ 1
1382
+ 1
1383
+ 1
1384
+ 1
1385
+ 1
1386
+ 1
1387
+ 1
1388
+ 2
1389
+ 1
1390
+ 1
1391
+ 1
1392
+ 1
1393
+ 1
1394
+ 1
1395
+ 1
1396
+ 1
1397
+ 1
1398
+ 1
1399
+ 1
1400
+ 1
1401
+ 1
1402
+ 1
1403
+ 1
1404
+ 1
1405
+ 1
1406
+ 1
1407
+ 1
1408
+ 1
1409
+ 1
1410
+ 1
1411
+ 1
1412
+ 1
1413
+ 1
1414
+ 1
1415
+ 1
1416
+ 1
1417
+ 1
1418
+ 1
1419
+ 1
1420
+ 1
1421
+ 1
1422
+ 1
1423
+ 1
1424
+ 1
1425
+ 1
1426
+ 1
1427
+ 1
1428
+ 1
1429
+ 1
1430
+ 1
1431
+ 1
1432
+ 1
1433
+ 1
1434
+ 1
1435
+ 1
1436
+ 1
1437
+ 1
1438
+ 1
1439
+ 1
1440
+ 1
1441
+ 1
1442
+ 1
1443
+ 1
1444
+ 1
1445
+ 1
1446
+ 1
1447
+ 1
1448
+ 1
1449
+ 1
1450
+ 1
1451
+ 1
1452
+ 1
1453
+ 1
1454
+ 1
1455
+ 1
1456
+ 1
1457
+ 1
1458
+ 1
1459
+ 1
1460
+ 1
1461
+ 1
1462
+ 1
1463
+ 1
1464
+ 1
1465
+ 1
1466
+ 1
1467
+ 1
1468
+ 1
1469
+ 1
1470
+ 1
1471
+ 1
1472
+ 1
1473
+ 1
1474
+ 1
1475
+ 1
1476
+ 1
1477
+ 1
1478
+ 1
1479
+ 1
1480
+ 1
1481
+ 1
1482
+ 1
1483
+ 1
1484
+ 1
1485
+ 1
1486
+ 1
1487
+ 1
1488
+ 1
1489
+ 1
1490
+ 1
1491
+ 1
1492
+ 1
1493
+ 1
1494
+ 1
1495
+ 2
1496
+ 1
1497
+ 1
1498
+ 1
1499
+ 1
1500
+ 1
1501
+ 1
1502
+ 1
1503
+ 1
1504
+ 1
1505
+ 1
1506
+ 1
1507
+ 1
1508
+ 1
1509
+ 2
1510
+ 1
1511
+ 1
1512
+ 1
1513
+ 1
1514
+ 1
1515
+ 1
1516
+ 1
1517
+ 1
1518
+ 1
1519
+ 2
1520
+ 1
1521
+ 1
1522
+ 1
1523
+ 1
1524
+ 1
1525
+ 1
1526
+ 1
1527
+ 1
1528
+ 1
1529
+ 1
1530
+ 1
1531
+ 1
1532
+ 1
1533
+ 1
1534
+ 1
1535
+ 1
1536
+ 1
1537
+ 1
1538
+ 1
1539
+ 1
1540
+ 1
1541
+ 1
1542
+ 1
1543
+ 1
1544
+ 1
1545
+ 1
1546
+ 1
1547
+ 1
1548
+ 1
1549
+ 1
1550
+ 1
1551
+ 1
1552
+ 1
1553
+ 1
1554
+ 1
1555
+ 1
1556
+ 1
1557
+ 1
1558
+ 1
1559
+ 1
1560
+ 1
1561
+ 1
1562
+ 1
1563
+ 1
1564
+ 1
1565
+ 1
1566
+ 1
1567
+ 1
1568
+ 1
1569
+ 1
1570
+ 1
1571
+ 1
1572
+ 1
1573
+ 1
1574
+ 1
1575
+ 2
1576
+ 1
1577
+ 1
1578
+ 1
1579
+ 1
1580
+ 1
1581
+ 1
1582
+ 1
1583
+ 1
1584
+ 1
1585
+ 1
1586
+ 1
1587
+ 1
1588
+ 1
1589
+ 1
1590
+ 1
1591
+ 1
1592
+ 1
1593
+ 1
1594
+ 1
1595
+ 1
1596
+ 1
1597
+ 1
1598
+ 1
1599
+ 1
1600
+ 1
1601
+ 1
1602
+ 2
1603
+ 1
1604
+ 1
1605
+ 1
1606
+ 1
1607
+ 1
1608
+ 1
1609
+ 1
1610
+ 1
1611
+ 1
1612
+ 1
1613
+ 1
1614
+ 1
1615
+ 2
1616
+ 1
1617
+ 1
1618
+ 1
1619
+ 1
1620
+ 1
1621
+ 1
1622
+ 1
1623
+ 1
1624
+ 1
1625
+ 1
1626
+ 1
1627
+ 1
1628
+ 1
1629
+ 1
1630
+ 1
1631
+ 1
1632
+ 1
1633
+ 1
1634
+ 1
1635
+ 1
1636
+ 1
1637
+ 1
1638
+ 1
1639
+ 1
1640
+ 1
1641
+ 1
1642
+ 1
1643
+ 1
1644
+ 1
1645
+ 1
1646
+ 1
1647
+ 1
1648
+ 1
1649
+ 1
1650
+ 1
1651
+ 1
1652
+ 1
1653
+ 1
1654
+ 1
1655
+ 1
1656
+ 1
1657
+ 1
1658
+ 1
1659
+ 1
1660
+ 1
1661
+ 1
1662
+ 1
1663
+ 1
1664
+ 1
1665
+ 1
1666
+ 1
1667
+ 1
1668
+ 1
1669
+ 2
1670
+ 1
1671
+ 1
1672
+ 1
1673
+ 1
1674
+ 2
1675
+ 1
1676
+ 1
1677
+ 1
1678
+ 1
1679
+ 1
1680
+ 1
1681
+ 1
1682
+ 1
1683
+ 1
1684
+ 1
1685
+ 1
1686
+ 1
1687
+ 1
1688
+ 1
1689
+ 1
1690
+ 1
1691
+ 1
1692
+ 1
1693
+ 1
1694
+ 1
1695
+ 1
1696
+ 1
1697
+ 1
1698
+ 1
1699
+ 1
1700
+ 2
1701
+ 1
1702
+ 1
1703
+ 1
1704
+ 1
1705
+ 1
1706
+ 1
1707
+ 1
1708
+ 1
1709
+ 1
1710
+ 1
1711
+ 1
1712
+ 1
1713
+ 2
1714
+ 1
1715
+ 2
1716
+ 1
1717
+ 1
1718
+ 1
1719
+ 1
1720
+ 1
1721
+ 1
1722
+ 1
1723
+ 1
1724
+ 1
1725
+ 1
1726
+ 1
1727
+ 1
1728
+ 1
1729
+ 1
1730
+ 1
1731
+ 1
1732
+ 1
1733
+ 1
1734
+ 1
1735
+ 1
1736
+ 1
1737
+ 1
1738
+ 2
1739
+ 1
1740
+ 1
1741
+ 1
1742
+ 1
1743
+ 1
1744
+ 1
1745
+ 1
1746
+ 1
1747
+ 1
1748
+ 1
1749
+ 1
1750
+ 1
1751
+ 1
1752
+ 1
1753
+ 1
1754
+ 1
1755
+ 1
1756
+ 1
1757
+ 1
1758
+ 1
1759
+ 1
1760
+ 1
1761
+ 1
1762
+ 1
1763
+ 1
1764
+ 1
1765
+ 1
1766
+ 1
1767
+ 1
1768
+ 1
1769
+ 1
1770
+ 1
1771
+ 1
1772
+ 1
1773
+ 1
1774
+ 1
1775
+ 1
1776
+ 1
1777
+ 1
1778
+ 1
1779
+ 1
1780
+ 1
1781
+ 1
1782
+ 1
1783
+ 1
1784
+ 1
1785
+ 1
1786
+ 1
1787
+ 1
1788
+ 1
1789
+ 1
1790
+ 1
1791
+ 1
1792
+ 1
1793
+ 1
1794
+ 1
1795
+ 1
1796
+ 1
1797
+ 1
1798
+ 1
1799
+ 1
1800
+ 1
1801
+ 1
1802
+ 1
1803
+ 1
1804
+ 1
1805
+ 1
1806
+ 1
1807
+ 1
1808
+ 1
1809
+ 1
1810
+ 1
1811
+ 1
1812
+ 1
1813
+ 1
1814
+ 1
1815
+ 1
1816
+ 1
1817
+ 1
1818
+ 1
1819
+ 1
1820
+ 1
1821
+ 1
1822
+ 1
1823
+ 1
1824
+ 1
1825
+ 1
1826
+ 1
1827
+ 1
1828
+ 1
1829
+ 1
1830
+ 1
1831
+ 1
1832
+ 1
1833
+ 1
1834
+ 1
1835
+ 1
1836
+ 1
1837
+ 1
1838
+ 1
1839
+ 1
1840
+ 1
1841
+ 1
1842
+ 2
1843
+ 1
1844
+ 1
1845
+ 1
1846
+ 1
1847
+ 1
1848
+ 1
1849
+ 1
1850
+ 1
1851
+ 1
1852
+ 1
1853
+ 1
1854
+ 1
1855
+ 1
1856
+ 1
1857
+ 1
1858
+ 1
1859
+ 1
1860
+ 1
1861
+ 1
1862
+ 1
1863
+ 1
1864
+ 1
1865
+ 1
1866
+ 1
1867
+ 1
1868
+ 1
1869
+ 1
1870
+ 1
1871
+ 1
1872
+ 1
1873
+ 2
1874
+ 1
1875
+ 1
1876
+ 1
1877
+ 1
1878
+ 1
1879
+ 1
1880
+ 2
1881
+ 1
1882
+ 1
1883
+ 1
1884
+ 1
1885
+ 1
1886
+ 1
1887
+ 1
1888
+ 1
1889
+ 1
1890
+ 1
1891
+ 1
1892
+ 1
1893
+ 1
1894
+ 1
1895
+ 1
1896
+ 1
1897
+ 1
1898
+ 1
1899
+ 1
1900
+ 1
1901
+ 1
1902
+ 2
1903
+ 1
1904
+ 1
1905
+ 1
1906
+ 1
1907
+ 1
1908
+ 1
1909
+ 1
1910
+ 1
1911
+ 1
1912
+ 1
1913
+ 1
1914
+ 1
1915
+ 1
1916
+ 1
1917
+ 1
1918
+ 1
1919
+ 1
1920
+ 1
1921
+ 1
1922
+ 2
1923
+ 1
1924
+ 1
1925
+ 1
1926
+ 1
1927
+ 1
1928
+ 1
1929
+ 1
1930
+ 1
1931
+ 1
1932
+ 1
1933
+ 1
1934
+ 1
1935
+ 2
1936
+ 1
1937
+ 1
1938
+ 1
1939
+ 1
1940
+ 1
1941
+ 1
1942
+ 1
1943
+ 1
1944
+ 1
1945
+ 1
1946
+ 1
1947
+ 1
1948
+ 1
1949
+ 1
1950
+ 1
1951
+ 1
1952
+ 1
1953
+ 1
1954
+ 1
1955
+ 1
1956
+ 1
1957
+ 1
1958
+ 1
1959
+ 1
1960
+ 1
1961
+ 1
1962
+ 2
1963
+ 1
1964
+ 1
1965
+ 1
1966
+ 1
1967
+ 1
1968
+ 2
1969
+ 1
1970
+ 1
1971
+ 1
1972
+ 1
1973
+ 1
1974
+ 1
1975
+ 1
1976
+ 1
1977
+ 1
1978
+ 1
1979
+ 2
1980
+ 1
1981
+ 1
1982
+ 1
1983
+ 1
1984
+ 1
1985
+ 1
1986
+ 1
1987
+ 1
1988
+ 1
1989
+ 1
1990
+ 1
1991
+ 1
1992
+ 1
1993
+ 1
1994
+ 1
1995
+ 1
1996
+ 1
1997
+ 1
1998
+ 1
1999
+ 1
2000
+ 1
2001
+ 1
2002
+ 1
2003
+ 1
2004
+ 1
2005
+ 1
2006
+ 1
2007
+ 1
2008
+ 1
2009
+ 1
2010
+ 1
2011
+ 1
2012
+ 1
2013
+ 1
2014
+ 1
2015
+ 1
2016
+ 1
2017
+ 1
2018
+ 1
2019
+ 2
2020
+ 1
2021
+ 1
2022
+ 1
2023
+ 1
2024
+ 1
2025
+ 1
2026
+ 1
2027
+ 1
2028
+ 1
2029
+ 1
2030
+ 1
2031
+ 2
2032
+ 1
2033
+ 1
2034
+ 1
2035
+ 1
2036
+ 1
2037
+ 1
2038
+ 1
2039
+ 1
2040
+ 2
2041
+ 1
2042
+ 1
2043
+ 1
2044
+ 2
2045
+ 1
2046
+ 1
2047
+ 1
2048
+ 1
2049
+ 1
2050
+ 1
2051
+ 1
2052
+ 1
2053
+ 1
2054
+ 1
2055
+ 1
2056
+ 1
2057
+ 1
2058
+ 1
2059
+ 1
2060
+ 1
2061
+ 1
2062
+ 2
2063
+ 1
2064
+ 1
2065
+ 1
2066
+ 1
2067
+ 1
2068
+ 1
2069
+ 1
2070
+ 1
2071
+ 1
2072
+ 1
2073
+ 1
2074
+ 1
2075
+ 1
2076
+ 1
2077
+ 1
2078
+ 1
2079
+ 1
2080
+ 1
2081
+ 1
2082
+ 1
2083
+ 1
2084
+ 1
2085
+ 1
2086
+ 1
2087
+ 1
2088
+ 1
2089
+ 1
2090
+ 1
2091
+ 1
2092
+ 1
2093
+ 1
2094
+ 1
2095
+ 1
2096
+ 1
2097
+ 1
2098
+ 1
2099
+ 1
2100
+ 1
2101
+ 1
2102
+ 1
2103
+ 1
2104
+ 1
2105
+ 1
2106
+ 1
2107
+ 1
2108
+ 1
2109
+ 1
2110
+ 1
2111
+ 1
2112
+ 1
2113
+ 1
2114
+ 2
2115
+ 1
2116
+ 1
2117
+ 1
2118
+ 1
2119
+ 1
2120
+ 1
2121
+ 1
2122
+ 1
2123
+ 1
2124
+ 1
2125
+ 1
2126
+ 1
2127
+ 1
2128
+ 1
2129
+ 1
2130
+ 1
2131
+ 1
2132
+ 1
2133
+ 1
2134
+ 1
2135
+ 1
2136
+ 1
2137
+ 1
2138
+ 1
2139
+ 1
2140
+ 1
2141
+ 1
2142
+ 1
2143
+ 1
2144
+ 1
2145
+ 1
2146
+ 1
2147
+ 1
2148
+ 1
2149
+ 1
2150
+ 1
2151
+ 1
2152
+ 1
2153
+ 1
2154
+ 1
2155
+ 1
2156
+ 1
2157
+ 1
2158
+ 1
2159
+ 1
2160
+ 1
2161
+ 1
2162
+ 1
2163
+ 1
2164
+ 1
2165
+ 1
2166
+ 1
2167
+ 1
2168
+ 1
2169
+ 1
2170
+ 1
2171
+ 1
2172
+ 1
2173
+ 1
2174
+ 1
2175
+ 1
2176
+ 1
2177
+ 1
2178
+ 1
2179
+ 1
2180
+ 1
2181
+ 1
2182
+ 1
2183
+ 1
2184
+ 1
2185
+ 1
2186
+ 1
2187
+ 1
2188
+ 1
2189
+ 1
2190
+ 1
2191
+ 1
2192
+ 1
2193
+ 1
2194
+ 1
2195
+ 1
2196
+ 1
2197
+ 1
2198
+ 1
2199
+ 1
2200
+ 1
2201
+ 1
2202
+ 1
2203
+ 1
2204
+ 1
2205
+ 2
2206
+ 1
2207
+ 1
2208
+ 1
2209
+ 1
2210
+ 1
2211
+ 1
2212
+ 1
2213
+ 1
2214
+ 1
2215
+ 1
2216
+ 1
2217
+ 2
2218
+ 1
2219
+ 1
2220
+ 1
2221
+ 1
2222
+ 1
2223
+ 1
2224
+ 1
2225
+ 1
2226
+ 1
2227
+ 1
2228
+ 1
2229
+ 1
2230
+ 1
2231
+ 1
2232
+ 1
2233
+ 1
2234
+ 1
2235
+ 1
2236
+ 1
2237
+ 2
2238
+ 1
2239
+ 1
2240
+ 1
2241
+ 1
2242
+ 1
2243
+ 1
2244
+ 1
2245
+ 1
2246
+ 1
2247
+ 1
2248
+ 1
2249
+ 1
2250
+ 1
2251
+ 2
2252
+ 1
2253
+ 1
2254
+ 1
2255
+ 1
2256
+ 1
2257
+ 1
2258
+ 1
2259
+ 1
2260
+ 1
2261
+ 1
2262
+ 1
2263
+ 1
2264
+ 1
2265
+ 1
2266
+ 1
2267
+ 1
2268
+ 1
2269
+ 1
2270
+ 1
2271
+ 1
2272
+ 1
2273
+ 1
2274
+ 1
2275
+ 2
2276
+ 1
2277
+ 1
2278
+ 1
2279
+ 1
2280
+ 1
2281
+ 1
2282
+ 1
2283
+ 1
2284
+ 1
2285
+ 1
2286
+ 1
2287
+ 1
2288
+ 1
2289
+ 1
2290
+ 1
2291
+ 1
2292
+ 1
2293
+ 2
2294
+ 1
2295
+ 1
2296
+ 1
2297
+ 1
2298
+ 1
2299
+ 1
2300
+ 1
2301
+ 1
2302
+ 1
2303
+ 1
2304
+ 1
2305
+ 1
2306
+ 1
2307
+ 1
2308
+ 1
2309
+ 1
2310
+ 1
2311
+ 2
2312
+ 1
2313
+ 1
2314
+ 1
2315
+ 1
2316
+ 1
2317
+ 1
2318
+ 1
2319
+ 1
2320
+ 1
2321
+ 1
2322
+ 1
2323
+ 1
2324
+ 1
2325
+ 1
2326
+ 1
2327
+ 1
2328
+ 1
2329
+ 1
2330
+ 1
2331
+ 1
2332
+ 1
2333
+ 1
2334
+ 1
2335
+ 1
2336
+ 1
2337
+ 1
2338
+ 1
2339
+ 1
2340
+ 1
2341
+ 1
2342
+ 1
2343
+ 2
2344
+ 1
2345
+ 1
2346
+ 1
2347
+ 1
2348
+ 1
2349
+ 1
2350
+ 1
2351
+ 1
2352
+ 1
2353
+ 1
2354
+ 1
2355
+ 1
2356
+ 1
2357
+ 1
2358
+ 1
2359
+ 1
2360
+ 2
2361
+ 1
2362
+ 1
2363
+ 1
2364
+ 1
2365
+ 1
2366
+ 1
2367
+ 1
2368
+ 1
2369
+ 1
2370
+ 1
2371
+ 1
2372
+ 1
2373
+ 1
2374
+ 1
2375
+ 2
2376
+ 1
2377
+ 1
2378
+ 1
2379
+ 1
2380
+ 1
2381
+ 1
2382
+ 1
2383
+ 2
2384
+ 1
2385
+ 1
2386
+ 1
2387
+ 1
2388
+ 1
2389
+ 1
2390
+ 1
2391
+ 1
2392
+ 1
2393
+ 1
2394
+ 1
2395
+ 1
2396
+ 1
2397
+ 1
2398
+ 1
2399
+ 1
2400
+ 2
2401
+ 1
2402
+ 1
2403
+ 1
2404
+ 1
2405
+ 1
2406
+ 1
2407
+ 1
2408
+ 1
2409
+ 1
2410
+ 1
2411
+ 1
2412
+ 1
2413
+ 1
2414
+ 1
2415
+ 1
2416
+ 2
2417
+ 1
2418
+ 1
2419
+ 1
2420
+ 2
2421
+ 1
2422
+ 1
2423
+ 1
2424
+ 1
2425
+ 1
2426
+ 1
2427
+ 1
2428
+ 1
2429
+ 1
2430
+ 1
2431
+ 1
2432
+ 1
2433
+ 1
2434
+ 1
2435
+ 1
2436
+ 1
2437
+ 1
2438
+ 1
2439
+ 1
2440
+ 1
2441
+ 1
2442
+ 1
2443
+ 1
2444
+ 1
2445
+ 1
2446
+ 1
2447
+ 1
2448
+ 1
2449
+ 1
2450
+ 1
2451
+ 1
2452
+ 1
2453
+ 1
2454
+ 1
2455
+ 1
2456
+ 1
2457
+ 1
2458
+ 1
2459
+ 1
2460
+ 1
2461
+ 1
2462
+ 1
2463
+ 1
2464
+ 1
2465
+ 1
2466
+ 1
2467
+ 1
2468
+ 1
2469
+ 1
2470
+ 1
2471
+ 1
2472
+ 1
2473
+ 1
2474
+ 1
2475
+ 1
2476
+ 1
2477
+ 1
2478
+ 1
2479
+ 1
2480
+ 1
2481
+ 1
2482
+ 1
2483
+ 1
2484
+ 2
2485
+ 1
2486
+ 1
2487
+ 1
2488
+ 1
2489
+ 1
2490
+ 1
2491
+ 1
2492
+ 1
2493
+ 1
2494
+ 1
2495
+ 1
2496
+ 1
2497
+ 1
2498
+ 1
2499
+ 1
2500
+ 1
2501
+ 1
2502
+ 1
2503
+ 2
2504
+ 1
2505
+ 2
2506
+ 1
2507
+ 1
2508
+ 1
2509
+ 1
2510
+ 1
2511
+ 1
2512
+ 1
2513
+ 1
2514
+ 1
2515
+ 1
2516
+ 1
2517
+ 1
2518
+ 1
2519
+ 1
2520
+ 1
2521
+ 1
2522
+ 1
2523
+ 1
2524
+ 1
2525
+ 1
2526
+ 1
2527
+ 1
2528
+ 1
2529
+ 1
2530
+ 1
2531
+ 1
2532
+ 1
2533
+ 1
2534
+ 1
2535
+ 1
2536
+ 1
2537
+ 1
2538
+ 1
2539
+ 1
2540
+ 1
2541
+ 1
2542
+ 1
2543
+ 1
2544
+ 1
2545
+ 1
2546
+ 1
2547
+ 1
2548
+ 1
2549
+ 1
2550
+ 1
2551
+ 1
2552
+ 1
2553
+ 1
2554
+ 1
2555
+ 1
2556
+ 1
2557
+ 1
2558
+ 1
2559
+ 1
2560
+ 1
2561
+ 1
2562
+ 1
2563
+ 1
2564
+ 1
2565
+ 1
2566
+ 1
2567
+ 1
2568
+ 1
2569
+ 1
2570
+ 1
2571
+ 1
2572
+ 1
2573
+ 1
2574
+ 1
2575
+ 1
2576
+ 1
2577
+ 2
2578
+ 1
2579
+ 1
2580
+ 1
2581
+ 1
2582
+ 1
2583
+ 1
2584
+ 1
2585
+ 1
2586
+ 1
2587
+ 1
2588
+ 1
2589
+ 1
2590
+ 2
2591
+ 2
2592
+ 1
2593
+ 1
2594
+ 1
2595
+ 1
2596
+ 1
2597
+ 1
2598
+ 1
2599
+ 1
2600
+ 1
2601
+ 1
2602
+ 1
2603
+ 1
2604
+ 1
2605
+ 1
2606
+ 1
2607
+ 1
2608
+ 1
2609
+ 1
2610
+ 1
2611
+ 1
2612
+ 1
2613
+ 1
2614
+ 1
2615
+ 1
2616
+ 1
2617
+ 1
2618
+ 1
2619
+ 1
2620
+ 1
2621
+ 1
2622
+ 1
2623
+ 2
2624
+ 1
2625
+ 2
2626
+ 1
2627
+ 1
2628
+ 1
2629
+ 1
2630
+ 1
2631
+ 1
2632
+ 1
2633
+ 1
2634
+ 1
2635
+ 1
2636
+ 1
2637
+ 2
2638
+ 1
2639
+ 1
2640
+ 1
2641
+ 1
2642
+ 1
2643
+ 1
2644
+ 1
2645
+ 1
2646
+ 1
2647
+ 1
2648
+ 1
2649
+ 1
2650
+ 1
2651
+ 1
2652
+ 2
2653
+ 1
2654
+ 1
2655
+ 1
2656
+ 2
2657
+ 1
2658
+ 1
2659
+ 2
2660
+ 2
2661
+ 1
2662
+ 1
2663
+ 1
2664
+ 1
2665
+ 2
2666
+ 1
2667
+ 1
2668
+ 1
2669
+ 1
2670
+ 1
2671
+ 1
2672
+ 1
2673
+ 2
2674
+ 1
2675
+ 1
2676
+ 1
2677
+ 1
2678
+ 1
2679
+ 1
2680
+ 1
2681
+ 1
2682
+ 1
2683
+ 1
2684
+ 1
2685
+ 1
2686
+ 1
2687
+ 1
2688
+ 1
2689
+ 1
2690
+ 1
2691
+ 1
2692
+ 1
2693
+ 1
2694
+ 1
2695
+ 1
2696
+ 1
2697
+ 1
2698
+ 1
2699
+ 1
2700
+ 1
2701
+ 1
2702
+ 1
2703
+ 1
2704
+ 1
2705
+ 1
2706
+ 1
2707
+ 2
2708
+ 2
2709
+ 1
2710
+ 2
2711
+ 1
2712
+ 1
2713
+ 1
2714
+ 1
2715
+ 1
2716
+ 1
2717
+ 1
2718
+ 1
2719
+ 1
2720
+ 1
2721
+ 1
2722
+ 1
2723
+ 1
2724
+ 1
2725
+ 1
2726
+ 2
2727
+ 1
2728
+ 1
2729
+ 1
2730
+ 1
2731
+ 1
2732
+ 1
2733
+ 2
2734
+ 1
2735
+ 1
2736
+ 1
2737
+ 1
2738
+ 1
2739
+ 1
2740
+ 1
2741
+ 1
2742
+ 1
2743
+ 1
2744
+ 1
2745
+ 1
2746
+ 1
2747
+ 1
2748
+ 1
2749
+ 1
2750
+ 1
2751
+ 1
2752
+ 1
2753
+ 1
2754
+ 1
2755
+ 1
2756
+ 1
2757
+ 1
2758
+ 1
2759
+ 1
2760
+ 1
2761
+ 1
2762
+ 2
2763
+ 1
2764
+ 1
2765
+ 1
2766
+ 1
2767
+ 1
2768
+ 1
2769
+ 1
2770
+ 1
2771
+ 1
2772
+ 1
2773
+ 1
2774
+ 1
2775
+ 1
2776
+ 1
2777
+ 1
2778
+ 1
2779
+ 1
2780
+ 1
2781
+ 1
2782
+ 1
2783
+ 1
2784
+ 1
2785
+ 1
2786
+ 1
2787
+ 1
2788
+ 1
2789
+ 1
2790
+ 1
2791
+ 1
2792
+ 1
2793
+ 1
2794
+ 1
2795
+ 1
2796
+ 1
2797
+ 1
2798
+ 1
2799
+ 1
2800
+ 1
2801
+ 1
2802
+ 1
2803
+ 1
2804
+ 1
2805
+ 1
2806
+ 1
2807
+ 1
2808
+ 1
2809
+ 1
2810
+ 1
2811
+ 1
2812
+ 1
2813
+ 1
2814
+ 1
2815
+ 1
2816
+ 1
2817
+ 1
2818
+ 1
2819
+ 1
2820
+ 1
2821
+ 1
2822
+ 1
2823
+ 1
2824
+ 1
2825
+ 1
2826
+ 1
2827
+ 1
2828
+ 2
2829
+ 1
2830
+ 1
2831
+ 1
2832
+ 1
2833
+ 1
2834
+ 1
2835
+ 1
2836
+ 1
2837
+ 1
2838
+ 1
2839
+ 1
2840
+ 1
2841
+ 1
2842
+ 1
2843
+ 1
2844
+ 1
2845
+ 1
2846
+ 1
2847
+ 1
2848
+ 1
2849
+ 1
2850
+ 1
2851
+ 1
2852
+ 1
2853
+ 1
2854
+ 1
2855
+ 1
2856
+ 1
2857
+ 1
2858
+ 1
2859
+ 1
2860
+ 1
2861
+ 1
2862
+ 1
2863
+ 1
2864
+ 1
2865
+ 2
2866
+ 1
2867
+ 2
2868
+ 1
2869
+ 1
2870
+ 1
2871
+ 1
2872
+ 1
2873
+ 1
2874
+ 1
2875
+ 1
2876
+ 1
2877
+ 1
2878
+ 1
2879
+ 1
2880
+ 1
2881
+ 1
2882
+ 1
2883
+ 1
2884
+ 1
2885
+ 1
2886
+ 1
2887
+ 1
2888
+ 1
2889
+ 1
2890
+ 1
2891
+ 1
2892
+ 1
2893
+ 1
2894
+ 1
2895
+ 1
2896
+ 1
2897
+ 1
2898
+ 1
2899
+ 1
2900
+ 1
2901
+ 1
2902
+ 1
2903
+ 1
2904
+ 1
2905
+ 1
2906
+ 2
2907
+ 1
2908
+ 1
2909
+ 1
2910
+ 1
2911
+ 1
2912
+ 1
2913
+ 1
2914
+ 1
2915
+ 1
2916
+ 1
2917
+ 1
2918
+ 1
2919
+ 1
2920
+ 1
2921
+ 1
2922
+ 1
2923
+ 2
2924
+ 1
2925
+ 1
2926
+ 1
2927
+ 1
2928
+ 1
2929
+ 1
2930
+ 2
2931
+ 1
2932
+ 1
2933
+ 1
2934
+ 1
2935
+ 1
2936
+ 1
2937
+ 1
2938
+ 1
2939
+ 1
2940
+ 1
2941
+ 1
2942
+ 1
2943
+ 1
2944
+ 1
2945
+ 1
2946
+ 1
2947
+ 1
2948
+ 1
2949
+ 1
2950
+ 1
2951
+ 1
2952
+ 1
2953
+ 1
2954
+ 1
2955
+ 1
2956
+ 1
2957
+ 1
2958
+ 1
2959
+ 1
2960
+ 1
2961
+ 1
2962
+ 1
2963
+ 1
2964
+ 1
2965
+ 1
2966
+ 1
2967
+ 2
2968
+ 1
2969
+ 1
2970
+ 1
2971
+ 1
2972
+ 1
2973
+ 2
2974
+ 1
2975
+ 1
2976
+ 1
2977
+ 1
2978
+ 1
2979
+ 1
2980
+ 1
2981
+ 1
2982
+ 1
2983
+ 1
2984
+ 1
2985
+ 1
2986
+ 1
2987
+ 1
2988
+ 1
2989
+ 1
2990
+ 1
2991
+ 1
2992
+ 1
2993
+ 1
2994
+ 2
2995
+ 1
2996
+ 1
2997
+ 1
2998
+ 1
2999
+ 1
3000
+ 1
3001
+ 1
3002
+ 1
3003
+ 1
3004
+ 1
3005
+ 1
3006
+ 1
3007
+ 1
3008
+ 1
3009
+ 1
3010
+ 1
3011
+ 2
3012
+ 1
3013
+ 1
3014
+ 1
3015
+ 1
3016
+ 1
3017
+ 1
3018
+ 1
3019
+ 1
3020
+ 1
3021
+ 1
3022
+ 1
3023
+ 1
3024
+ 1
3025
+ 1
3026
+ 2
3027
+ 1
3028
+ 1
3029
+ 1
3030
+ 1
3031
+ 1
3032
+ 2
3033
+ 1
3034
+ 1
3035
+ 1
3036
+ 1
3037
+ 1
3038
+ 1
3039
+ 1
3040
+ 1
3041
+ 2
3042
+ 2
3043
+ 1
3044
+ 1
3045
+ 1
3046
+ 1
3047
+ 1
3048
+ 1
3049
+ 1
3050
+ 1
3051
+ 1
3052
+ 1
3053
+ 1
3054
+ 1
3055
+ 1
3056
+ 1
3057
+ 1
3058
+ 1
3059
+ 1
3060
+ 1
3061
+ 1
3062
+ 1
3063
+ 1
3064
+ 1
3065
+ 1
3066
+ 1
3067
+ 1
3068
+ 1
3069
+ 1
3070
+ 1
3071
+ 2
3072
+ 1
3073
+ 1
3074
+ 1
3075
+ 1
3076
+ 1
3077
+ 1
3078
+ 1
3079
+ 1
3080
+ 1
3081
+ 1
3082
+ 1
3083
+ 1
3084
+ 1
3085
+ 1
3086
+ 1
3087
+ 1
3088
+ 1
3089
+ 1
3090
+ 1
3091
+ 1
3092
+ 1
3093
+ 1
3094
+ 1
3095
+ 1
3096
+ 1
3097
+ 1
3098
+ 1
3099
+ 1
3100
+ 1
3101
+ 1
3102
+ 1
3103
+ 1
3104
+ 1
3105
+ 1
3106
+ 1
3107
+ 1
3108
+ 1
3109
+ 1
3110
+ 1
3111
+ 1
3112
+ 1
3113
+ 1
3114
+ 2
3115
+ 1
3116
+ 1
3117
+ 1
3118
+ 1
3119
+ 1
3120
+ 1
3121
+ 1
3122
+ 1
3123
+ 1
3124
+ 1
3125
+ 2
3126
+ 1
3127
+ 1
3128
+ 1
3129
+ 1
3130
+ 2
3131
+ 1
3132
+ 1
3133
+ 1
3134
+ 1
3135
+ 1
3136
+ 1
3137
+ 1
3138
+ 2
3139
+ 1
3140
+ 1
3141
+ 1
3142
+ 2
3143
+ 1
3144
+ 1
3145
+ 1
3146
+ 1
3147
+ 1
3148
+ 1
3149
+ 1
3150
+ 1
3151
+ 1
3152
+ 1
3153
+ 1
3154
+ 1
3155
+ 1
3156
+ 1
3157
+ 1
3158
+ 2
3159
+ 1
3160
+ 1
3161
+ 1
3162
+ 1
3163
+ 1
3164
+ 1
3165
+ 1
3166
+ 1
3167
+ 1
3168
+ 1
3169
+ 1
3170
+ 1
3171
+ 1
3172
+ 1
3173
+ 1
3174
+ 1
3175
+ 1
3176
+ 1
3177
+ 1
3178
+ 1
3179
+ 1
3180
+ 1
3181
+ 1
3182
+ 1
3183
+ 1
3184
+ 2
3185
+ 1
3186
+ 1
3187
+ 1
3188
+ 1
3189
+ 1
3190
+ 1
3191
+ 1
3192
+ 1
3193
+ 1
3194
+ 1
3195
+ 1
3196
+ 1
3197
+ 1
3198
+ 1
3199
+ 1
3200
+ 1
3201
+ 1
3202
+ 1
3203
+ 1
3204
+ 1
3205
+ 1
3206
+ 1
3207
+ 2
3208
+ 1
3209
+ 1
3210
+ 1
3211
+ 1
3212
+ 1
3213
+ 1
3214
+ 1
3215
+ 1
3216
+ 1
3217
+ 1
3218
+ 1
3219
+ 1
3220
+ 2
3221
+ 1
3222
+ 1
3223
+ 1
3224
+ 1
3225
+ 1
3226
+ 1
3227
+ 1
3228
+ 1
3229
+ 1
3230
+ 1
3231
+ 1
3232
+ 1
3233
+ 1
3234
+ 1
3235
+ 1
3236
+ 1
3237
+ 1
3238
+ 1
3239
+ 1
3240
+ 1
3241
+ 1
3242
+ 1
3243
+ 1
3244
+ 1
3245
+ 1
3246
+ 1
3247
+ 1
3248
+ 2
3249
+ 1
3250
+ 1
3251
+ 1
3252
+ 1
3253
+ 1
3254
+ 2
3255
+ 1
3256
+ 1
3257
+ 1
3258
+ 1
3259
+ 1
3260
+ 1
3261
+ 1
3262
+ 1
3263
+ 1
3264
+ 1
3265
+ 1
3266
+ 1
3267
+ 1
3268
+ 1
3269
+ 1
3270
+ 1
3271
+ 1
3272
+ 1
3273
+ 2
3274
+ 1
3275
+ 1
3276
+ 1
3277
+ 2
3278
+ 1
3279
+ 1
3280
+ 1
3281
+ 1
3282
+ 1
3283
+ 1
3284
+ 1
3285
+ 1
3286
+ 1
3287
+ 1
3288
+ 1
3289
+ 1
3290
+ 1
3291
+ 1
3292
+ 1
3293
+ 1
3294
+ 1
3295
+ 1
3296
+ 1
3297
+ 1
3298
+ 1
3299
+ 1
3300
+ 1
3301
+ 1
3302
+ 1
3303
+ 1
3304
+ 1
3305
+ 1
3306
+ 1
3307
+ 1
3308
+ 1
3309
+ 1
3310
+ 1
3311
+ 1
3312
+ 1
3313
+ 1
3314
+ 1
3315
+ 1
3316
+ 1
3317
+ 1
3318
+ 1
3319
+ 1
3320
+ 1
3321
+ 1
3322
+ 2
3323
+ 1
3324
+ 1
3325
+ 1
3326
+ 1
3327
+ 1
3328
+ 1
3329
+ 2
3330
+ 1
3331
+ 1
3332
+ 1
3333
+ 1
3334
+ 2
3335
+ 1
3336
+ 1
3337
+ 1
3338
+ 1
3339
+ 2
3340
+ 1
3341
+ 1
3342
+ 2
3343
+ 1
3344
+ 1
3345
+ 1
3346
+ 1
3347
+ 2
3348
+ 1
3349
+ 1
3350
+ 1
3351
+ 1
3352
+ 2
3353
+ 1
3354
+ 1
3355
+ 1
3356
+ 1
3357
+ 1
3358
+ 1
3359
+ 1
3360
+ 1
3361
+ 1
3362
+ 1
3363
+ 1
3364
+ 1
3365
+ 1
3366
+ 1
3367
+ 1
3368
+ 1
3369
+ 1
3370
+ 1
3371
+ 1
3372
+ 1
3373
+ 1
3374
+ 1
3375
+ 1
3376
+ 1
3377
+ 1
3378
+ 1
3379
+ 1
3380
+ 1
3381
+ 1
3382
+ 1
3383
+ 1
3384
+ 1
3385
+ 1
3386
+ 1
3387
+ 1
3388
+ 1
3389
+ 1
3390
+ 1
3391
+ 1
3392
+ 1
3393
+ 1
3394
+ 1
3395
+ 1
3396
+ 1
3397
+ 2
3398
+ 1
3399
+ 1
3400
+ 1
3401
+ 1
3402
+ 1
3403
+ 1
3404
+ 1
3405
+ 1
3406
+ 1
3407
+ 1
3408
+ 1
3409
+ 1
3410
+ 1
3411
+ 1
3412
+ 1
3413
+ 1
3414
+ 1
3415
+ 1
3416
+ 1
3417
+ 1
3418
+ 1
3419
+ 1
3420
+ 2
3421
+ 1
3422
+ 1
3423
+ 1
3424
+ 1
3425
+ 1
3426
+ 1
3427
+ 1
3428
+ 1
3429
+ 1
3430
+ 1
3431
+ 2
3432
+ 1
3433
+ 1
3434
+ 2
3435
+ 1
3436
+ 1
3437
+ 1
3438
+ 1
3439
+ 1
3440
+ 1
3441
+ 1
3442
+ 1
3443
+ 1
3444
+ 1
3445
+ 1
3446
+ 1
3447
+ 1
3448
+ 1
3449
+ 2
3450
+ 1
3451
+ 1
3452
+ 1
3453
+ 1
3454
+ 1
3455
+ 1
3456
+ 2
3457
+ 1
3458
+ 1
3459
+ 1
3460
+ 1
3461
+ 1
3462
+ 1
3463
+ 1
3464
+ 2
3465
+ 1
3466
+ 1
3467
+ 1
3468
+ 1
3469
+ 1
3470
+ 1
3471
+ 1
3472
+ 1
3473
+ 1
3474
+ 1
3475
+ 1
3476
+ 1
3477
+ 1
3478
+ 1
3479
+ 1
3480
+ 1
3481
+ 1
3482
+ 1
3483
+ 1
3484
+ 1
3485
+ 1
3486
+ 1
3487
+ 1
3488
+ 1
3489
+ 1
3490
+ 1
3491
+ 1
3492
+ 1
3493
+ 1
3494
+ 1
3495
+ 1
3496
+ 1
3497
+ 1
3498
+ 1
3499
+ 1
3500
+ 1
3501
+ 1
3502
+ 1
3503
+ 1
3504
+ 2
3505
+ 1
3506
+ 1
3507
+ 1
3508
+ 1
3509
+ 1
3510
+ 1
3511
+ 1
3512
+ 1
3513
+ 1
3514
+ 1
3515
+ 1
3516
+ 1
3517
+ 1
3518
+ 1
3519
+ 1
3520
+ 1
3521
+ 1
3522
+ 1
3523
+ 1
3524
+ 1
3525
+ 1
3526
+ 1
3527
+ 2
3528
+ 1
3529
+ 1
3530
+ 2
3531
+ 1
3532
+ 1
3533
+ 1
3534
+ 1
3535
+ 1
3536
+ 1
3537
+ 1
3538
+ 2
3539
+ 1
3540
+ 1
3541
+ 1
3542
+ 1
3543
+ 1
3544
+ 1
3545
+ 1
3546
+ 1
3547
+ 1
3548
+ 1
3549
+ 1
3550
+ 1
3551
+ 1
3552
+ 1
3553
+ 1
3554
+ 1
3555
+ 1
3556
+ 2
3557
+ 1
3558
+ 1
3559
+ 1
3560
+ 1
3561
+ 1
3562
+ 1
3563
+ 1
3564
+ 1
3565
+ 1
3566
+ 1
3567
+ 1
3568
+ 1
3569
+ 1
3570
+ 1
3571
+ 1
3572
+ 1
3573
+ 1
3574
+ 1
3575
+ 1
3576
+ 1
3577
+ 1
3578
+ 2
3579
+ 1
3580
+ 1
3581
+ 1
3582
+ 1
3583
+ 1
3584
+ 1
3585
+ 2
3586
+ 1
3587
+ 1
3588
+ 1
3589
+ 1
3590
+ 1
3591
+ 1
3592
+ 2
3593
+ 1
3594
+ 1
3595
+ 2
3596
+ 1
3597
+ 1
3598
+ 2
3599
+ 1
3600
+ 1
3601
+ 1
3602
+ 1
3603
+ 1
3604
+ 2
3605
+ 1
3606
+ 1
3607
+ 1
3608
+ 1
3609
+ 1
3610
+ 1
3611
+ 1
3612
+ 1
3613
+ 1
3614
+ 2
3615
+ 1
3616
+ 2
3617
+ 1
3618
+ 1
3619
+ 1
3620
+ 1
3621
+ 1
3622
+ 1
3623
+ 1
3624
+ 1
3625
+ 1
3626
+ 1
3627
+ 1
3628
+ 1
3629
+ 1
3630
+ 1
3631
+ 1
3632
+ 1
3633
+ 1
3634
+ 1
3635
+ 1
3636
+ 1
3637
+ 1
3638
+ 1
3639
+ 1
3640
+ 1
3641
+ 1
3642
+ 1
3643
+ 1
3644
+ 1
3645
+ 1
3646
+ 1
3647
+ 1
3648
+ 1
3649
+ 1
3650
+ 1
3651
+ 1
3652
+ 1
3653
+ 1
3654
+ 1
3655
+ 1
3656
+ 1
3657
+ 1
3658
+ 1
3659
+ 1
3660
+ 1
3661
+ 1
3662
+ 1
3663
+ 1
3664
+ 1
3665
+ 1
3666
+ 1
3667
+ 1
3668
+ 1
3669
+ 1
3670
+ 1
3671
+ 2
3672
+ 1
3673
+ 1
3674
+ 1
3675
+ 1
3676
+ 1
3677
+ 2
3678
+ 1
3679
+ 2
3680
+ 1
3681
+ 1
3682
+ 1
3683
+ 1
3684
+ 1
3685
+ 1
3686
+ 1
3687
+ 1
3688
+ 1
3689
+ 1
3690
+ 1
3691
+ 1
3692
+ 1
3693
+ 1
3694
+ 1
3695
+ 1
3696
+ 1
3697
+ 1
3698
+ 2
3699
+ 1
3700
+ 1
3701
+ 1
3702
+ 1
3703
+ 1
3704
+ 1
3705
+ 1
3706
+ 1
3707
+ 1
3708
+ 1
3709
+ 1
3710
+ 1
3711
+ 1
3712
+ 1
3713
+ 1
3714
+ 1
3715
+ 1
3716
+ 1
3717
+ 1
3718
+ 1
3719
+ 1
3720
+ 1
3721
+ 1
3722
+ 1
3723
+ 1
3724
+ 2
3725
+ 1
3726
+ 1
3727
+ 1
3728
+ 1
3729
+ 2
3730
+ 1
3731
+ 1
3732
+ 1
3733
+ 1
3734
+ 1
3735
+ 2
3736
+ 1
3737
+ 1
3738
+ 1
3739
+ 1
3740
+ 1
3741
+ 1
3742
+ 1
3743
+ 1
3744
+ 1
3745
+ 1
3746
+ 2
3747
+ 1
3748
+ 1
3749
+ 1
3750
+ 1
3751
+ 2
3752
+ 1
3753
+ 2
3754
+ 1
3755
+ 1
3756
+ 1
3757
+ 1
3758
+ 1
3759
+ 1
3760
+ 1
3761
+ 1
3762
+ 1
3763
+ 1
3764
+ 1
3765
+ 1
3766
+ 1
3767
+ 1
3768
+ 1
3769
+ 1
3770
+ 1
3771
+ 1
3772
+ 1
3773
+ 1
3774
+ 1
3775
+ 1
3776
+ 1
3777
+ 1
3778
+ 1
3779
+ 1
3780
+ 2
3781
+ 1
3782
+ 1
3783
+ 2
3784
+ 1
3785
+ 1
3786
+ 1
3787
+ 1
3788
+ 1
3789
+ 1
3790
+ 1
3791
+ 1
3792
+ 1
3793
+ 1
3794
+ 1
3795
+ 1
3796
+ 1
3797
+ 1
3798
+ 1
3799
+ 1
3800
+ 1
3801
+ 1
3802
+ 1
3803
+ 1
3804
+ 1
3805
+ 1
3806
+ 1
3807
+ 1
3808
+ 1
3809
+ 1
3810
+ 1
3811
+ 1
3812
+ 1
3813
+ 1
3814
+ 2
3815
+ 1
3816
+ 1
3817
+ 1
3818
+ 2
3819
+ 1
3820
+ 2
3821
+ 1
3822
+ 1
3823
+ 1
3824
+ 1
3825
+ 1
3826
+ 1
3827
+ 1
3828
+ 1
3829
+ 1
3830
+ 1
3831
+ 1
3832
+ 1
3833
+ 1
3834
+ 1
3835
+ 1
3836
+ 1
3837
+ 1
3838
+ 1
3839
+ 1
3840
+ 1
3841
+ 1
3842
+ 1
3843
+ 1
3844
+ 1
3845
+ 1
3846
+ 1
3847
+ 1
3848
+ 1
3849
+ 1
3850
+ 1
3851
+ 1
3852
+ 1
3853
+ 1
3854
+ 1
3855
+ 2
3856
+ 1
3857
+ 1
3858
+ 1
3859
+ 1
3860
+ 1
3861
+ 1
3862
+ 1
3863
+ 1
3864
+ 1
3865
+ 1
3866
+ 1
3867
+ 1
3868
+ 1
3869
+ 1
3870
+ 1
3871
+ 1
3872
+ 1
3873
+ 1
3874
+ 1
3875
+ 1
3876
+ 1
3877
+ 1
3878
+ 1
3879
+ 1
3880
+ 1
3881
+ 1
3882
+ 1
3883
+ 1
3884
+ 1
3885
+ 1
3886
+ 2
3887
+ 1
3888
+ 1
3889
+ 1
3890
+ 1
3891
+ 1
3892
+ 1
3893
+ 1
3894
+ 1
3895
+ 1
3896
+ 1
3897
+ 1
3898
+ 1
3899
+ 1
3900
+ 1
3901
+ 1
3902
+ 1
3903
+ 1
3904
+ 1
3905
+ 1
3906
+ 1
3907
+ 1
3908
+ 1
3909
+ 1
3910
+ 1
3911
+ 1
3912
+ 1
3913
+ 1
3914
+ 1
3915
+ 1
3916
+ 1
3917
+ 1
3918
+ 1
3919
+ 1
3920
+ 1
3921
+ 1
3922
+ 1
3923
+ 1
3924
+ 1
3925
+ 1
3926
+ 1
3927
+ 1
3928
+ 1
3929
+ 1
3930
+ 1
3931
+ 1
3932
+ 1
3933
+ 1
3934
+ 1
3935
+ 1
3936
+ 1
3937
+ 1
3938
+ 1
3939
+ 1
3940
+ 1
3941
+ 1
3942
+ 1
3943
+ 1
3944
+ 1
3945
+ 2
3946
+ 1
3947
+ 1
3948
+ 2
3949
+ 1
3950
+ 1
3951
+ 1
3952
+ 1
3953
+ 1
3954
+ 1
3955
+ 1
3956
+ 1
3957
+ 1
3958
+ 1
3959
+ 1
3960
+ 1
3961
+ 1
3962
+ 1
3963
+ 1
3964
+ 1
3965
+ 1
3966
+ 1
3967
+ 1
3968
+ 1
3969
+ 1
3970
+ 1
3971
+ 2
3972
+ 1
3973
+ 1
3974
+ 1
3975
+ 1
3976
+ 1
3977
+ 1
3978
+ 1
3979
+ 1
3980
+ 1
3981
+ 1
3982
+ 1
3983
+ 1
3984
+ 1
3985
+ 1
3986
+ 2
3987
+ 1
3988
+ 1
3989
+ 1
3990
+ 1
3991
+ 1
3992
+ 1
3993
+ 1
3994
+ 1
3995
+ 1
3996
+ 1
3997
+ 1
3998
+ 1
3999
+ 1
4000
+ 1
4001
+ 1
4002
+ 1
4003
+ 1
4004
+ 1
4005
+ 1
4006
+ 1
4007
+ 1
4008
+ 1
4009
+ 1
4010
+ 1
4011
+ 1
4012
+ 2
4013
+ 1
4014
+ 1
4015
+ 1
4016
+ 1
4017
+ 1
4018
+ 1
4019
+ 1
4020
+ 1
4021
+ 1
4022
+ 1
4023
+ 2
4024
+ 1
4025
+ 1
4026
+ 1
4027
+ 1
4028
+ 1
4029
+ 1
4030
+ 1
4031
+ 1
4032
+ 1
4033
+ 1
4034
+ 1
4035
+ 1
4036
+ 1
4037
+ 1
4038
+ 1
4039
+ 1
4040
+ 1
4041
+ 1
4042
+ 1
4043
+ 1
4044
+ 1
4045
+ 1
4046
+ 1
4047
+ 1
4048
+ 1
4049
+ 1
4050
+ 1
4051
+ 1
4052
+ 1
4053
+ 1
4054
+ 1
4055
+ 1
4056
+ 1
4057
+ 1
4058
+ 1
4059
+ 1
4060
+ 1
4061
+ 1
4062
+ 1
4063
+ 1
4064
+ 1
4065
+ 1
4066
+ 1
4067
+ 1
4068
+ 1
4069
+ 1
4070
+ 1
4071
+ 1
4072
+ 2
4073
+ 1
4074
+ 1
4075
+ 1
4076
+ 2
4077
+ 1
4078
+ 1
4079
+ 1
4080
+ 1
4081
+ 1
4082
+ 1
4083
+ 1
4084
+ 1
4085
+ 1
4086
+ 1
4087
+ 1
4088
+ 1
4089
+ 1
4090
+ 1
4091
+ 1
4092
+ 1
4093
+ 1
4094
+ 1
4095
+ 1
4096
+ 1
4097
+ 1
4098
+ 1
4099
+ 1
4100
+ 1
4101
+ 1
4102
+ 1
4103
+ 1
4104
+ 1
4105
+ 1
4106
+ 1
4107
+ 1
4108
+ 1
4109
+ 1
4110
+ 1
4111
+ 1
4112
+ 1
4113
+ 1
4114
+ 1
4115
+ 1
4116
+ 1
4117
+ 1
4118
+ 1
4119
+ 1
4120
+ 1
4121
+ 1
4122
+ 1
4123
+ 1
4124
+ 1
4125
+ 1
4126
+ 1
4127
+ 1
4128
+ 1
4129
+ 1
4130
+ 1
4131
+ 1
4132
+ 1
4133
+ 2
4134
+ 1
4135
+ 1
4136
+ 1
4137
+ 1
4138
+ 1
4139
+ 1
4140
+ 1
4141
+ 1
4142
+ 1
4143
+ 1
4144
+ 1
4145
+ 1
4146
+ 1
4147
+ 1
4148
+ 1
4149
+ 1
4150
+ 1
4151
+ 1
4152
+ 1
4153
+ 1
4154
+ 1
4155
+ 1
4156
+ 1
4157
+ 1
4158
+ 1
4159
+ 2
4160
+ 1
4161
+ 1
4162
+ 1
4163
+ 1
4164
+ 1
4165
+ 1
4166
+ 1
4167
+ 1
4168
+ 2
4169
+ 1
4170
+ 1
4171
+ 1
4172
+ 1
4173
+ 1
4174
+ 1
4175
+ 1
4176
+ 1
4177
+ 1
4178
+ 1
4179
+ 1
4180
+ 1
4181
+ 1
4182
+ 1
4183
+ 1
4184
+ 1
4185
+ 1
4186
+ 1
4187
+ 1
4188
+ 1
4189
+ 1
4190
+ 1
4191
+ 2
4192
+ 1
4193
+ 1
4194
+ 1
4195
+ 2
4196
+ 1
4197
+ 1
4198
+ 1
4199
+ 1
4200
+ 1
4201
+ 1
4202
+ 1
4203
+ 1
4204
+ 1
4205
+ 1
4206
+ 1
4207
+ 1
4208
+ 2
4209
+ 1
4210
+ 1
4211
+ 1
4212
+ 1
4213
+ 1
4214
+ 1
4215
+ 1
4216
+ 1
4217
+ 1
4218
+ 1
4219
+ 1
4220
+ 1
4221
+ 1
4222
+ 1
4223
+ 1
4224
+ 1
4225
+ 1
4226
+ 1
4227
+ 1
4228
+ 1
4229
+ 1
4230
+ 1
4231
+ 1
4232
+ 1
4233
+ 1
4234
+ 1
4235
+ 1
4236
+ 1
4237
+ 1
4238
+ 1
4239
+ 1
4240
+ 1
4241
+ 1
4242
+ 1
4243
+ 1
4244
+ 1
4245
+ 1
4246
+ 1
4247
+ 2
4248
+ 1
4249
+ 1
4250
+ 2
4251
+ 1
4252
+ 1
4253
+ 1
4254
+ 1
4255
+ 1
4256
+ 1
4257
+ 1
4258
+ 1
4259
+ 1
4260
+ 1
4261
+ 1
4262
+ 1
4263
+ 1
4264
+ 1
4265
+ 1
4266
+ 1
4267
+ 1
4268
+ 1
4269
+ 1
4270
+ 1
4271
+ 1
4272
+ 1
4273
+ 1
4274
+ 1
4275
+ 1
4276
+ 1
4277
+ 1
4278
+ 1
4279
+ 1
4280
+ 1
4281
+ 1
4282
+ 1
4283
+ 1
4284
+ 1
4285
+ 1
4286
+ 1
4287
+ 1
4288
+ 1
4289
+ 1
4290
+ 1
4291
+ 1
4292
+ 1
4293
+ 1
4294
+ 1
4295
+ 1
4296
+ 1
4297
+ 1
4298
+ 1
4299
+ 2
4300
+ 1
4301
+ 1
4302
+ 1
4303
+ 1
4304
+ 1
4305
+ 1
4306
+ 1
4307
+ 1
4308
+ 2
4309
+ 1
4310
+ 1
4311
+ 1
4312
+ 1
4313
+ 1
4314
+ 1
4315
+ 1
4316
+ 1
4317
+ 1
4318
+ 2
4319
+ 2
4320
+ 2
4321
+ 2
4322
+ 1
4323
+ 1
4324
+ 1
4325
+ 1
4326
+ 1
4327
+ 1
4328
+ 1
4329
+ 1
4330
+ 1
4331
+ 1
4332
+ 1
4333
+ 1
4334
+ 1
4335
+ 1
4336
+ 1
4337
+ 1
4338
+ 1
4339
+ 1
4340
+ 1
4341
+ 1
4342
+ 1
4343
+ 1
4344
+ 1
4345
+ 1
4346
+ 1
4347
+ 1
4348
+ 1
4349
+ 1
4350
+ 1
4351
+ 1
4352
+ 1
4353
+ 1
4354
+ 1
4355
+ 1
4356
+ 1
4357
+ 1
4358
+ 1
4359
+ 1
4360
+ 1
4361
+ 1
4362
+ 1
4363
+ 1
4364
+ 1
4365
+ 1
4366
+ 1
4367
+ 1
4368
+ 1
4369
+ 1
4370
+ 1
4371
+ 1
4372
+ 1
4373
+ 1
4374
+ 1
4375
+ 1
4376
+ 1
4377
+ 1
4378
+ 1
4379
+ 1
4380
+ 1
4381
+ 1
4382
+ 1
4383
+ 2
4384
+ 1
4385
+ 1
4386
+ 1
4387
+ 1
4388
+ 1
4389
+ 1
4390
+ 1
4391
+ 1
4392
+ 1
4393
+ 1
4394
+ 1
4395
+ 1
4396
+ 1
4397
+ 1
4398
+ 1
4399
+ 1
4400
+ 2
4401
+ 1
4402
+ 2
4403
+ 1
4404
+ 1
4405
+ 1
4406
+ 1
4407
+ 1
4408
+ 2
4409
+ 1
4410
+ 1
4411
+ 1
4412
+ 1
4413
+ 1
4414
+ 1
4415
+ 1
4416
+ 1
4417
+ 2
4418
+ 1
4419
+ 1
4420
+ 1
4421
+ 1
4422
+ 1
4423
+ 1
4424
+ 2
4425
+ 1
4426
+ 1
4427
+ 1
4428
+ 1
4429
+ 1
4430
+ 1
4431
+ 1
4432
+ 1
4433
+ 1
4434
+ 1
4435
+ 1
4436
+ 1
4437
+ 1
4438
+ 1
4439
+ 1
4440
+ 1
4441
+ 1
4442
+ 1
4443
+ 1
4444
+ 1
4445
+ 1
4446
+ 1
4447
+ 1
4448
+ 1
4449
+ 1
4450
+ 1
4451
+ 1
4452
+ 1
4453
+ 1
4454
+ 1
4455
+ 1
4456
+ 1
4457
+ 1
4458
+ 1
4459
+ 1
4460
+ 1
4461
+ 1
4462
+ 1
4463
+ 1
4464
+ 1
4465
+ 1
4466
+ 1
4467
+ 1
4468
+ 1
4469
+ 1
4470
+ 1
4471
+ 1
4472
+ 1
4473
+ 1
4474
+ 1
4475
+ 1
4476
+ 1
4477
+ 1
4478
+ 1
4479
+ 1
4480
+ 1
4481
+ 1
4482
+ 1
4483
+ 1
4484
+ 1
4485
+ 2
4486
+ 1
4487
+ 1
4488
+ 1
4489
+ 1
4490
+ 1
4491
+ 1
4492
+ 2
4493
+ 1
4494
+ 2
4495
+ 1
4496
+ 1
4497
+ 1
4498
+ 1
4499
+ 1
4500
+ 1
4501
+ 1
4502
+ 1
4503
+ 1
4504
+ 1
4505
+ 1
4506
+ 1
4507
+ 1
4508
+ 1
4509
+ 1
4510
+ 1
4511
+ 1
4512
+ 1
4513
+ 1
4514
+ 1
4515
+ 1
4516
+ 1
4517
+ 1
4518
+ 2
4519
+ 1
4520
+ 1
4521
+ 1
4522
+ 1
4523
+ 1
4524
+ 1
4525
+ 1
4526
+ 2
4527
+ 1
4528
+ 1
4529
+ 1
4530
+ 1
4531
+ 1
4532
+ 1
4533
+ 1
4534
+ 1
4535
+ 1
4536
+ 1
4537
+ 1
4538
+ 1
4539
+ 2
4540
+ 1
4541
+ 1
4542
+ 1
4543
+ 1
4544
+ 1
4545
+ 1
4546
+ 1
4547
+ 1
4548
+ 1
4549
+ 1
4550
+ 1
4551
+ 1
4552
+ 1
4553
+ 1
4554
+ 1
4555
+ 1
4556
+ 1
4557
+ 1
4558
+ 1
4559
+ 1
4560
+ 1
4561
+ 1
4562
+ 1
4563
+ 1
4564
+ 1
4565
+ 1
4566
+ 1
4567
+ 1
4568
+ 1
4569
+ 1
4570
+ 1
4571
+ 1
4572
+ 1
4573
+ 1
4574
+ 1
4575
+ 1
4576
+ 1
4577
+ 1
4578
+ 1
4579
+ 2
4580
+ 1
4581
+ 1
4582
+ 1
4583
+ 1
4584
+ 1
4585
+ 1
4586
+ 1
4587
+ 1
4588
+ 1
4589
+ 1
4590
+ 1
4591
+ 1
4592
+ 1
4593
+ 1
4594
+ 1
4595
+ 1
4596
+ 1
4597
+ 1
4598
+ 1
4599
+ 1
4600
+ 1
4601
+ 1
4602
+ 1
4603
+ 1
4604
+ 1
4605
+ 1
4606
+ 1
4607
+ 2
4608
+ 1
4609
+ 1
4610
+ 2
4611
+ 1
4612
+ 1
4613
+ 1
4614
+ 1
4615
+ 1
4616
+ 1
4617
+ 1
4618
+ 1
4619
+ 1
4620
+ 1
4621
+ 2
4622
+ 1
4623
+ 1
4624
+ 2
4625
+ 1
4626
+ 1
4627
+ 1
4628
+ 1
4629
+ 1
4630
+ 1
4631
+ 1
4632
+ 1
4633
+ 1
4634
+ 1
4635
+ 1
4636
+ 1
4637
+ 1
4638
+ 2
4639
+ 1
4640
+ 1
4641
+ 1
4642
+ 1
4643
+ 1
4644
+ 1
4645
+ 1
4646
+ 1
4647
+ 2
4648
+ 1
4649
+ 1
4650
+ 1
4651
+ 1
4652
+ 1
4653
+ 1
4654
+ 1
4655
+ 1
4656
+ 1
4657
+ 1
4658
+ 1
4659
+ 1
4660
+ 1
4661
+ 1
4662
+ 1
4663
+ 2
4664
+ 1
4665
+ 1
4666
+ 1
4667
+ 1
4668
+ 1
4669
+ 2
4670
+ 1
4671
+ 1
4672
+ 1
4673
+ 1
4674
+ 1
4675
+ 1
4676
+ 1
4677
+ 1
4678
+ 1
4679
+ 1
4680
+ 1
4681
+ 1
4682
+ 1
4683
+ 1
4684
+ 1
4685
+ 1
4686
+ 1
4687
+ 1
4688
+ 1
4689
+ 1
4690
+ 1
4691
+ 1
4692
+ 1
4693
+ 1
4694
+ 1
4695
+ 1
4696
+ 1
4697
+ 1
4698
+ 1
4699
+ 1
4700
+ 1
4701
+ 1
4702
+ 1
4703
+ 1
4704
+ 1
4705
+ 1
4706
+ 1
4707
+ 1
4708
+ 1
4709
+ 1
4710
+ 1
4711
+ 1
4712
+ 1
4713
+ 1
4714
+ 1
4715
+ 1
4716
+ 1
4717
+ 2
4718
+ 1
4719
+ 1
4720
+ 1
4721
+ 1
4722
+ 1
4723
+ 1
4724
+ 1
4725
+ 1
4726
+ 1
4727
+ 1
4728
+ 1
4729
+ 1
4730
+ 1
4731
+ 1
4732
+ 1
4733
+ 1
4734
+ 1
4735
+ 1
4736
+ 1
4737
+ 1
4738
+ 1
4739
+ 1
4740
+ 2
4741
+ 1
4742
+ 1
4743
+ 1
4744
+ 1
4745
+ 1
4746
+ 1
4747
+ 1
4748
+ 2
4749
+ 1
4750
+ 1
4751
+ 1
4752
+ 1
4753
+ 1
4754
+ 1
4755
+ 1
4756
+ 1
4757
+ 1
4758
+ 1
4759
+ 1
4760
+ 1
4761
+ 1
4762
+ 1
4763
+ 1
4764
+ 1
4765
+ 1
4766
+ 1
4767
+ 1
4768
+ 1
4769
+ 1
4770
+ 1
4771
+ 2
4772
+ 1
4773
+ 1
4774
+ 1
4775
+ 2
4776
+ 1
4777
+ 2
4778
+ 1
4779
+ 1
4780
+ 1
4781
+ 1
4782
+ 1
4783
+ 1
4784
+ 1
4785
+ 1
4786
+ 2
4787
+ 1
4788
+ 1
4789
+ 1
4790
+ 1
4791
+ 1
4792
+ 1
4793
+ 1
4794
+ 1
4795
+ 1
4796
+ 1
4797
+ 1
4798
+ 1
4799
+ 1
4800
+ 1
4801
+ 2
4802
+ 1
4803
+ 1
4804
+ 1
4805
+ 1
4806
+ 1
4807
+ 1
4808
+ 1
4809
+ 1
4810
+ 1
4811
+ 1
4812
+ 1
4813
+ 1
4814
+ 1
4815
+ 1
4816
+ 1
4817
+ 1
4818
+ 1
4819
+ 1
4820
+ 1
4821
+ 1
4822
+ 1
4823
+ 1
4824
+ 1
4825
+ 1
4826
+ 1
4827
+ 1
4828
+ 1
4829
+ 1
4830
+ 1
4831
+ 1
4832
+ 1
4833
+ 1
4834
+ 1
4835
+ 1
4836
+ 1
4837
+ 1
4838
+ 1
4839
+ 1
4840
+ 1
4841
+ 1
4842
+ 1
4843
+ 1
4844
+ 1
4845
+ 1
4846
+ 2
4847
+ 1
4848
+ 1
4849
+ 1
4850
+ 1
4851
+ 1
4852
+ 1
4853
+ 1
4854
+ 1
4855
+ 1
4856
+ 1
4857
+ 1
4858
+ 1
4859
+ 1
4860
+ 1
4861
+ 1
4862
+ 1
4863
+ 1
4864
+ 2
4865
+ 1
4866
+ 1
4867
+ 1
4868
+ 1
4869
+ 1
4870
+ 1
4871
+ 1
4872
+ 1
4873
+ 1
4874
+ 1
4875
+ 1
4876
+ 1
4877
+ 1
4878
+ 1
4879
+ 1
4880
+ 1
4881
+ 1
4882
+ 1
4883
+ 1
4884
+ 1
4885
+ 1
4886
+ 1
4887
+ 1
4888
+ 1
4889
+ 1
4890
+ 1
4891
+ 1
4892
+ 2
4893
+ 1
4894
+ 1
4895
+ 1
4896
+ 1
4897
+ 1
4898
+ 1
4899
+ 1
4900
+ 1
4901
+ 1
4902
+ 1
4903
+ 1
4904
+ 1
4905
+ 2
4906
+ 1
4907
+ 1
4908
+ 1
4909
+ 1
4910
+ 1
4911
+ 1
4912
+ 1
4913
+ 1
4914
+ 1
4915
+ 1
4916
+ 1
4917
+ 1
4918
+ 1
4919
+ 1
4920
+ 1
4921
+ 1
4922
+ 1
4923
+ 2
4924
+ 1
4925
+ 1
4926
+ 2
4927
+ 1
4928
+ 1
4929
+ 1
4930
+ 1
4931
+ 1
4932
+ 1
4933
+ 1
4934
+ 1
4935
+ 1
4936
+ 1
4937
+ 1
4938
+ 1
4939
+ 1
4940
+ 1
4941
+ 1
4942
+ 1
4943
+ 1
4944
+ 1
4945
+ 1
4946
+ 1
4947
+ 1
4948
+ 1
4949
+ 1
4950
+ 1
4951
+ 1
4952
+ 1
4953
+ 1
4954
+ 1
4955
+ 1
4956
+ 1
4957
+ 2
4958
+ 1
4959
+ 1
4960
+ 1
4961
+ 1
4962
+ 1
4963
+ 1
4964
+ 2
4965
+ 2
4966
+ 1
4967
+ 1
4968
+ 1
4969
+ 1
4970
+ 1
4971
+ 1
4972
+ 1
4973
+ 1
4974
+ 1
4975
+ 1
4976
+ 1
4977
+ 1
4978
+ 1
4979
+ 1
4980
+ 1
4981
+ 1
4982
+ 1
4983
+ 1
4984
+ 1
4985
+ 1
4986
+ 1
4987
+ 1
4988
+ 1
4989
+ 1
4990
+ 1
4991
+ 1
4992
+ 1
4993
+ 1
4994
+ 1
4995
+ 2
4996
+ 1
4997
+ 1
4998
+ 1
4999
+ 1
5000
+ 1
5001
+ 1
5002
+ 1
5003
+ 1
5004
+ 1
5005
+ 1
5006
+ 1
5007
+ 1
5008
+ 1
5009
+ 1
5010
+ 1
5011
+ 1
5012
+ 2
5013
+ 1
5014
+ 1
5015
+ 1
5016
+ 1
5017
+ 1
5018
+ 1
5019
+ 1
5020
+ 2
5021
+ 1
5022
+ 1
5023
+ 1
5024
+ 1
5025
+ 1
5026
+ 1
5027
+ 1
5028
+ 1
5029
+ 1
5030
+ 1
5031
+ 1
5032
+ 1
5033
+ 1
5034
+ 1
5035
+ 1
5036
+ 1
5037
+ 2
5038
+ 1
5039
+ 2
5040
+ 1
5041
+ 1
5042
+ 2
5043
+ 1
5044
+ 1
5045
+ 1
5046
+ 1
5047
+ 1
5048
+ 1
5049
+ 1
5050
+ 1
5051
+ 1
5052
+ 1
5053
+ 1
5054
+ 1
5055
+ 1
5056
+ 2
5057
+ 1
5058
+ 1
5059
+ 1
5060
+ 1
5061
+ 1
5062
+ 1
5063
+ 1
5064
+ 1
5065
+ 1
5066
+ 1
5067
+ 1
5068
+ 1
5069
+ 1
5070
+ 1
5071
+ 1
5072
+ 1
5073
+ 1
5074
+ 1
5075
+ 1
5076
+ 1
5077
+ 1
5078
+ 1
5079
+ 1
5080
+ 1
5081
+ 1
5082
+ 1
5083
+ 1
5084
+ 1
5085
+ 1
5086
+ 1
5087
+ 1
5088
+ 1
5089
+ 1
5090
+ 1
5091
+ 1
5092
+ 1
5093
+ 1
5094
+ 1
5095
+ 1
5096
+ 1
5097
+ 1
5098
+ 1
5099
+ 1
5100
+ 1
5101
+ 1
5102
+ 1
5103
+ 1
5104
+ 1
5105
+ 1
5106
+ 1
5107
+ 1
5108
+ 1
5109
+ 1
5110
+ 1
5111
+ 1
5112
+ 1
5113
+ 1
5114
+ 1
5115
+ 1
5116
+ 1
5117
+ 1
5118
+ 1
5119
+ 2
5120
+ 1
5121
+ 1
5122
+ 1
5123
+ 2
5124
+ 1
5125
+ 1
5126
+ 1
5127
+ 1
5128
+ 1
5129
+ 1
5130
+ 1
5131
+ 2
5132
+ 1
5133
+ 1
5134
+ 1
5135
+ 1
5136
+ 1
5137
+ 1
5138
+ 1
5139
+ 1
5140
+ 1
5141
+ 1
5142
+ 1
5143
+ 1
5144
+ 1
5145
+ 1
5146
+ 1
5147
+ 1
5148
+ 1
5149
+ 1
5150
+ 1
5151
+ 1
5152
+ 1
5153
+ 1
5154
+ 1
5155
+ 1
5156
+ 1
5157
+ 1
5158
+ 1
5159
+ 1
5160
+ 1
5161
+ 1
5162
+ 1
5163
+ 2
5164
+ 1
5165
+ 2
5166
+ 1
5167
+ 1
5168
+ 1
5169
+ 1
5170
+ 1
5171
+ 1
5172
+ 1
5173
+ 1
5174
+ 1
5175
+ 1
5176
+ 1
5177
+ 1
5178
+ 1
5179
+ 2
5180
+ 1
5181
+ 1
5182
+ 1
5183
+ 1
5184
+ 1
5185
+ 1
5186
+ 1
5187
+ 1
5188
+ 1
5189
+ 1
5190
+ 1
5191
+ 1
5192
+ 1
5193
+ 1
5194
+ 1
5195
+ 1
5196
+ 1
5197
+ 2
5198
+ 1
5199
+ 1
5200
+ 1
5201
+ 1
5202
+ 1
5203
+ 1
5204
+ 1
5205
+ 1
5206
+ 1
5207
+ 1
5208
+ 1
5209
+ 1
5210
+ 1
5211
+ 1
5212
+ 1
5213
+ 1
5214
+ 1
5215
+ 1
5216
+ 1
5217
+ 1
5218
+ 1
5219
+ 1
5220
+ 1
5221
+ 1
5222
+ 1
5223
+ 1
5224
+ 1
5225
+ 1
5226
+ 1
5227
+ 1
5228
+ 2
5229
+ 1
5230
+ 1
5231
+ 1
5232
+ 1
5233
+ 1
5234
+ 1
5235
+ 1
5236
+ 1
5237
+ 1
5238
+ 1
5239
+ 1
5240
+ 1
5241
+ 1
5242
+ 1
5243
+ 1
5244
+ 1
5245
+ 1
5246
+ 1
5247
+ 1
5248
+ 1
5249
+ 1
5250
+ 1
5251
+ 1
5252
+ 1
5253
+ 1
5254
+ 1
5255
+ 1
5256
+ 1
5257
+ 1
5258
+ 1
5259
+ 1
5260
+ 1
5261
+ 1
5262
+ 1
5263
+ 1
5264
+ 1
5265
+ 1
5266
+ 1
5267
+ 2
5268
+ 1
5269
+ 1
5270
+ 1
5271
+ 2
5272
+ 1
5273
+ 1
5274
+ 2
5275
+ 1
5276
+ 1
5277
+ 1
5278
+ 1
5279
+ 1
5280
+ 2
5281
+ 1
5282
+ 1
5283
+ 1
5284
+ 1
5285
+ 1
5286
+ 1
5287
+ 1
5288
+ 1
5289
+ 1
5290
+ 1
5291
+ 1
5292
+ 1
5293
+ 1
5294
+ 1
5295
+ 1
5296
+ 1
5297
+ 1
5298
+ 1
5299
+ 1
5300
+ 2
5301
+ 1
5302
+ 1
5303
+ 1
5304
+ 1
5305
+ 1
5306
+ 1
5307
+ 1
5308
+ 1
5309
+ 1
5310
+ 1
5311
+ 2
5312
+ 1
5313
+ 1
5314
+ 1
5315
+ 2
5316
+ 1
5317
+ 1
5318
+ 1
5319
+ 1
5320
+ 1
5321
+ 1
5322
+ 1
5323
+ 1
5324
+ 1
5325
+ 1
5326
+ 1
5327
+ 1
5328
+ 1
5329
+ 1
5330
+ 1
5331
+ 1
5332
+ 1
5333
+ 1
5334
+ 1
5335
+ 1
5336
+ 1
5337
+ 1
5338
+ 1
5339
+ 1
5340
+ 1
5341
+ 1
5342
+ 1
5343
+ 1
5344
+ 1
5345
+ 1
5346
+ 1
5347
+ 1
5348
+ 1
5349
+ 1
5350
+ 1
5351
+ 1
5352
+ 1
5353
+ 1
5354
+ 1
5355
+ 1
5356
+ 1
5357
+ 1
5358
+ 1
5359
+ 1
5360
+ 1
5361
+ 1
5362
+ 1
5363
+ 1
5364
+ 2
5365
+ 1
5366
+ 1
5367
+ 1
5368
+ 1
5369
+ 1
5370
+ 1
5371
+ 1
5372
+ 1
5373
+ 1
5374
+ 1
5375
+ 1
5376
+ 2
5377
+ 1
5378
+ 1
5379
+ 1
5380
+ 1
5381
+ 1
5382
+ 1
5383
+ 1
5384
+ 1
5385
+ 2
5386
+ 1
5387
+ 1
5388
+ 1
5389
+ 1
5390
+ 1
5391
+ 1
5392
+ 1
5393
+ 1
5394
+ 2
5395
+ 1
5396
+ 1
5397
+ 1
5398
+ 1
5399
+ 1
5400
+ 1
5401
+ 1
5402
+ 1
5403
+ 1
5404
+ 1
5405
+ 1
5406
+ 1
5407
+ 1
5408
+ 1
5409
+ 1
5410
+ 1
5411
+ 1
5412
+ 1
5413
+ 1
5414
+ 1
5415
+ 2
5416
+ 1
5417
+ 1
5418
+ 2
5419
+ 1
5420
+ 1
5421
+ 1
5422
+ 1
5423
+ 1
5424
+ 1
5425
+ 1
5426
+ 1
5427
+ 1
5428
+ 1
5429
+ 1
5430
+ 1
5431
+ 1
5432
+ 1
5433
+ 1
5434
+ 2
5435
+ 1
5436
+ 1
5437
+ 1
5438
+ 1
5439
+ 1
5440
+ 1
5441
+ 1
5442
+ 1
5443
+ 1
5444
+ 1
5445
+ 1
5446
+ 1
5447
+ 1
5448
+ 1
5449
+ 2
5450
+ 1
5451
+ 1
5452
+ 1
5453
+ 1
5454
+ 1
5455
+ 1
5456
+ 1
5457
+ 1
5458
+ 1
5459
+ 1
5460
+ 1
5461
+ 1
5462
+ 1
5463
+ 1
5464
+ 1
5465
+ 1
5466
+ 1
5467
+ 1
5468
+ 1
5469
+ 1
5470
+ 1
5471
+ 1
5472
+ 1
5473
+ 1
5474
+ 1
5475
+ 1
5476
+ 1
5477
+ 1
5478
+ 1
5479
+ 1
5480
+ 1
5481
+ 1
5482
+ 1
5483
+ 1
5484
+ 1
5485
+ 1
5486
+ 1
5487
+ 1
5488
+ 1
5489
+ 1
5490
+ 1
5491
+ 1
5492
+ 1
5493
+ 1
5494
+ 1
5495
+ 2
5496
+ 1
5497
+ 1
5498
+ 1
5499
+ 1
5500
+ 1
5501
+ 1
5502
+ 1
5503
+ 1
5504
+ 1
5505
+ 1
5506
+ 2
5507
+ 1
5508
+ 1
5509
+ 1
5510
+ 2
5511
+ 1
5512
+ 1
5513
+ 1
5514
+ 1
5515
+ 1
5516
+ 1
5517
+ 1
5518
+ 1
5519
+ 1
5520
+ 1
5521
+ 1
5522
+ 1
5523
+ 1
5524
+ 1
5525
+ 1
5526
+ 2
5527
+ 1
5528
+ 1
5529
+ 1
5530
+ 1
5531
+ 1
5532
+ 1
5533
+ 1
5534
+ 1
5535
+ 1
5536
+ 1
5537
+ 1
5538
+ 1
5539
+ 1
5540
+ 1
5541
+ 1
5542
+ 1
5543
+ 1
5544
+ 1
5545
+ 1
5546
+ 1
5547
+ 1
5548
+ 1
5549
+ 1
5550
+ 1
5551
+ 1
5552
+ 1
5553
+ 1
5554
+ 1
5555
+ 1
5556
+ 1
5557
+ 1
5558
+ 1
5559
+ 1
5560
+ 1
5561
+ 1
5562
+ 1
5563
+ 1
5564
+ 1
5565
+ 1
5566
+ 1
5567
+ 2
5568
+ 1
5569
+ 1
5570
+ 1
5571
+ 1
5572
+ 1
5573
+ 1
5574
+ 1
5575
+ 1
5576
+ 1
5577
+ 1
5578
+ 1
5579
+ 1
5580
+ 1
5581
+ 1
5582
+ 2
5583
+ 1
5584
+ 1
5585
+ 1
5586
+ 1
5587
+ 1
5588
+ 1
5589
+ 1
5590
+ 1
5591
+ 1
5592
+ 1
5593
+ 1
5594
+ 1
5595
+ 1
5596
+ 1
5597
+ 1
5598
+ 1
5599
+ 1
5600
+ 1
5601
+ 1
5602
+ 1
5603
+ 2
5604
+ 1
5605
+ 1
5606
+ 1
5607
+ 1
5608
+ 1
5609
+ 1
5610
+ 2
5611
+ 1
5612
+ 1
5613
+ 1
5614
+ 1
5615
+ 1
5616
+ 1
5617
+ 1
5618
+ 1
5619
+ 1
5620
+ 1
5621
+ 2
5622
+ 1
5623
+ 1
5624
+ 1
5625
+ 1
5626
+ 1
5627
+ 1
5628
+ 1
5629
+ 1
5630
+ 1
5631
+ 1
5632
+ 1
5633
+ 1
5634
+ 1
5635
+ 1
5636
+ 1
5637
+ 1
5638
+ 1
5639
+ 1
5640
+ 1
5641
+ 1
5642
+ 1
5643
+ 1
5644
+ 1
5645
+ 1
5646
+ 1
5647
+ 1
5648
+ 1
5649
+ 1
5650
+ 1
5651
+ 1
5652
+ 1
5653
+ 1
5654
+ 2
5655
+ 1
5656
+ 1
5657
+ 1
5658
+ 1
5659
+ 1
5660
+ 1
5661
+ 1
5662
+ 1
5663
+ 1
5664
+ 1
5665
+ 1
5666
+ 1
5667
+ 1
5668
+ 1
5669
+ 1
5670
+ 1
5671
+ 2
5672
+ 1
5673
+ 1
5674
+ 1
5675
+ 1
5676
+ 1
5677
+ 1
5678
+ 1
5679
+ 2
5680
+ 1
5681
+ 1
5682
+ 1
5683
+ 1
5684
+ 1
5685
+ 1
5686
+ 1
5687
+ 1
5688
+ 1
5689
+ 1
5690
+ 1
5691
+ 2
5692
+ 1
5693
+ 1
5694
+ 1
5695
+ 1
5696
+ 1
5697
+ 1
5698
+ 1
5699
+ 1
5700
+ 1
5701
+ 2
5702
+ 1
5703
+ 1
5704
+ 2
5705
+ 1
5706
+ 1
5707
+ 1
5708
+ 1
5709
+ 1
5710
+ 1
5711
+ 1
5712
+ 1
5713
+ 1
5714
+ 1
5715
+ 1
5716
+ 1
5717
+ 1
5718
+ 1
5719
+ 1
5720
+ 1
5721
+ 1
5722
+ 1
5723
+ 1
5724
+ 1
5725
+ 2
5726
+ 1
5727
+ 1
5728
+ 1
5729
+ 1
5730
+ 1
5731
+ 1
5732
+ 1
5733
+ 1
5734
+ 1
5735
+ 1
5736
+ 1
5737
+ 1
5738
+ 1
5739
+ 1
5740
+ 1
5741
+ 1
5742
+ 1
5743
+ 1
5744
+ 1
5745
+ 1
5746
+ 1
5747
+ 1
5748
+ 1
5749
+ 1
5750
+ 1
5751
+ 1
5752
+ 1
5753
+ 2
5754
+ 1
5755
+ 1
5756
+ 1
5757
+ 1
5758
+ 1
5759
+ 1
5760
+ 1
5761
+ 1
5762
+ 1
5763
+ 1
5764
+ 1
5765
+ 1
5766
+ 2
5767
+ 1
5768
+ 1
5769
+ 1
5770
+ 1
5771
+ 1
5772
+ 1
5773
+ 1
5774
+ 1
5775
+ 1
5776
+ 1
5777
+ 1
5778
+ 1
5779
+ 1
5780
+ 1
5781
+ 1
5782
+ 1
5783
+ 1
5784
+ 1
5785
+ 1
5786
+ 1
5787
+ 1
5788
+ 1
5789
+ 1
5790
+ 1
5791
+ 1
5792
+ 1
5793
+ 1
5794
+ 1
5795
+ 1
5796
+ 1
5797
+ 1
5798
+ 1
5799
+ 1
5800
+ 1
5801
+ 1
5802
+ 1
5803
+ 1
5804
+ 2
5805
+ 1
5806
+ 1
5807
+ 1
5808
+ 1
5809
+ 1
5810
+ 1
5811
+ 1
5812
+ 2
5813
+ 1
5814
+ 1
5815
+ 1
5816
+ 1
5817
+ 1
5818
+ 1
5819
+ 1
5820
+ 1
5821
+ 1
5822
+ 1
5823
+ 1
5824
+ 1
5825
+ 1
5826
+ 1
5827
+ 1
5828
+ 1
5829
+ 1
5830
+ 1
5831
+ 1
5832
+ 1
5833
+ 1
5834
+ 1
5835
+ 1
5836
+ 1
5837
+ 1
5838
+ 1
5839
+ 1
5840
+ 1
5841
+ 1
5842
+ 1
5843
+ 1
5844
+ 1
5845
+ 1
5846
+ 1
5847
+ 1
5848
+ 1
5849
+ 1
5850
+ 1
5851
+ 1
5852
+ 1
5853
+ 1
5854
+ 1
5855
+ 1
5856
+ 1
5857
+ 1
5858
+ 1
5859
+ 1
5860
+ 1
5861
+ 2
5862
+ 1
5863
+ 1
5864
+ 1
5865
+ 1
5866
+ 1
5867
+ 1
5868
+ 1
5869
+ 1
5870
+ 1
5871
+ 1
5872
+ 1
5873
+ 1
5874
+ 1
5875
+ 1
5876
+ 1
5877
+ 1
5878
+ 1
5879
+ 1
5880
+ 1
5881
+ 1
5882
+ 2
5883
+ 1
5884
+ 1
5885
+ 1
5886
+ 1
5887
+ 1
5888
+ 1
5889
+ 1
5890
+ 1
5891
+ 1
5892
+ 1
5893
+ 1
5894
+ 1
5895
+ 1
5896
+ 2
5897
+ 1
5898
+ 1
5899
+ 1
5900
+ 1
5901
+ 1
5902
+ 1
5903
+ 1
5904
+ 1
5905
+ 1
5906
+ 1
5907
+ 1
5908
+ 1
5909
+ 1
5910
+ 1
5911
+ 1
5912
+ 1
5913
+ 2
5914
+ 1
5915
+ 1
5916
+ 2
5917
+ 1
5918
+ 1
5919
+ 1
5920
+ 1
5921
+ 1
5922
+ 1
5923
+ 1
5924
+ 1
5925
+ 1
5926
+ 1
5927
+ 1
5928
+ 1
5929
+ 1
5930
+ 1
5931
+ 1
5932
+ 1
5933
+ 1
5934
+ 1
5935
+ 1
5936
+ 1
5937
+ 1
5938
+ 1
5939
+ 1
5940
+ 1
5941
+ 2
5942
+ 1
5943
+ 1
5944
+ 1
5945
+ 1
5946
+ 1
5947
+ 1
5948
+ 1
5949
+ 1
5950
+ 1
5951
+ 1
5952
+ 1
5953
+ 2
5954
+ 1
5955
+ 1
5956
+ 1
5957
+ 1
5958
+ 1
5959
+ 1
5960
+ 1
5961
+ 1
5962
+ 1
5963
+ 1
5964
+ 1
5965
+ 1
5966
+ 1
5967
+ 2
5968
+ 1
5969
+ 1
5970
+ 1
5971
+ 1
5972
+ 1
5973
+ 1
5974
+ 1
5975
+ 1
5976
+ 1
5977
+ 1
5978
+ 2
5979
+ 1
5980
+ 1
5981
+ 1
5982
+ 1
5983
+ 1
5984
+ 1
5985
+ 1
5986
+ 1
5987
+ 1
5988
+ 1
5989
+ 2
5990
+ 1
5991
+ 1
5992
+ 1
5993
+ 1
5994
+ 1
5995
+ 1
5996
+ 1
5997
+ 1
5998
+ 1
5999
+ 1
6000
+ 1
6001
+ 1
6002
+ 1
6003
+ 1
6004
+ 1
6005
+ 1
6006
+ 1
6007
+ 1
6008
+ 2
6009
+ 1
6010
+ 1
6011
+ 1
6012
+ 1
6013
+ 1
6014
+ 1
6015
+ 1
6016
+ 1
6017
+ 1
6018
+ 1
6019
+ 1
6020
+ 1
6021
+ 1
6022
+ 1
6023
+ 1
6024
+ 1
6025
+ 1
6026
+ 1
6027
+ 1
6028
+ 1
6029
+ 1
6030
+ 1
6031
+ 1
6032
+ 1
6033
+ 1
6034
+ 1
6035
+ 1
6036
+ 1
6037
+ 1
6038
+ 2
6039
+ 1
6040
+ 1
6041
+ 1
6042
+ 1
6043
+ 1
6044
+ 1
6045
+ 1
6046
+ 1
6047
+ 1
6048
+ 1
6049
+ 1
6050
+ 1
6051
+ 1
6052
+ 1
6053
+ 1
6054
+ 1
6055
+ 1
6056
+ 1
6057
+ 1
6058
+ 1
6059
+ 1
6060
+ 1
6061
+ 1
6062
+ 2
6063
+ 1
6064
+ 1
6065
+ 1
6066
+ 1
6067
+ 1
6068
+ 1
6069
+ 1
6070
+ 2
6071
+ 1
6072
+ 1
6073
+ 1
6074
+ 1
6075
+ 1
6076
+ 1
6077
+ 1
6078
+ 1
6079
+ 1
6080
+ 1
6081
+ 2
6082
+ 1
6083
+ 1
6084
+ 1
6085
+ 1
6086
+ 1
6087
+ 1
6088
+ 1
6089
+ 1
6090
+ 1
6091
+ 1
6092
+ 1
6093
+ 1
6094
+ 1
6095
+ 1
6096
+ 1
6097
+ 1
6098
+ 1
6099
+ 1
6100
+ 1
6101
+ 1
6102
+ 1
6103
+ 1
6104
+ 1
6105
+ 1
6106
+ 1
6107
+ 1
6108
+ 1
6109
+ 1
6110
+ 1
6111
+ 1
6112
+ 2
6113
+ 1
6114
+ 1
6115
+ 1
6116
+ 1
6117
+ 1
6118
+ 1
6119
+ 1
6120
+ 1
6121
+ 1
6122
+ 1
6123
+ 1
6124
+ 1
6125
+ 1
6126
+ 1
6127
+ 1
6128
+ 2
6129
+ 1
6130
+ 1
6131
+ 1
6132
+ 1
6133
+ 1
6134
+ 1
6135
+ 1
6136
+ 1
6137
+ 1
6138
+ 1
6139
+ 1
6140
+ 1
6141
+ 1
6142
+ 1
6143
+ 1
6144
+ 1
6145
+ 1
6146
+ 1
6147
+ 2
6148
+ 1
6149
+ 1
6150
+ 1
6151
+ 1
6152
+ 1
6153
+ 1
6154
+ 1
6155
+ 1
6156
+ 1
6157
+ 1
6158
+ 1
6159
+ 1
6160
+ 1
6161
+ 1
6162
+ 2
6163
+ 1
6164
+ 1
6165
+ 1
6166
+ 1
6167
+ 2
6168
+ 1
6169
+ 2
6170
+ 1
6171
+ 1
6172
+ 1
6173
+ 1
6174
+ 1
6175
+ 1
6176
+ 1
6177
+ 1
6178
+ 1
6179
+ 2
6180
+ 1
6181
+ 1
6182
+ 1
6183
+ 2
6184
+ 1
6185
+ 1
6186
+ 1
6187
+ 1
6188
+ 1
6189
+ 1
6190
+ 1
6191
+ 2
6192
+ 1
6193
+ 1
6194
+ 1
6195
+ 1
6196
+ 1
6197
+ 1
6198
+ 1
6199
+ 1
6200
+ 1
6201
+ 1
6202
+ 1
6203
+ 1
6204
+ 1
6205
+ 1
6206
+ 1
6207
+ 1
6208
+ 1
6209
+ 1
6210
+ 1
6211
+ 1
6212
+ 1
6213
+ 1
6214
+ 1
6215
+ 1
6216
+ 1
6217
+ 1
6218
+ 1
6219
+ 1
6220
+ 1
6221
+ 2
6222
+ 1
6223
+ 1
6224
+ 1
6225
+ 1
6226
+ 1
6227
+ 1
6228
+ 1
6229
+ 1
6230
+ 1
6231
+ 1
6232
+ 1
6233
+ 1
6234
+ 1
6235
+ 1
6236
+ 2
6237
+ 1
6238
+ 1
6239
+ 1
6240
+ 1
6241
+ 1
6242
+ 1
6243
+ 1
6244
+ 1
6245
+ 1
6246
+ 1
6247
+ 1
6248
+ 1
6249
+ 1
6250
+ 1
6251
+ 1
6252
+ 1
6253
+ 1
6254
+ 2
6255
+ 1
6256
+ 1
6257
+ 1
6258
+ 1
6259
+ 1
6260
+ 1
6261
+ 1
6262
+ 1
6263
+ 1
6264
+ 1
6265
+ 1
6266
+ 1
6267
+ 1
6268
+ 1
6269
+ 1
6270
+ 1
6271
+ 2
6272
+ 1
6273
+ 1
6274
+ 1
6275
+ 1
6276
+ 1
6277
+ 1
6278
+ 1
6279
+ 1
6280
+ 1
6281
+ 1
6282
+ 1
6283
+ 1
6284
+ 1
6285
+ 1
6286
+ 1
6287
+ 1
6288
+ 1
6289
+ 1
6290
+ 1
6291
+ 1
6292
+ 1
6293
+ 1
6294
+ 1
6295
+ 1
6296
+ 1
6297
+ 1
6298
+ 1
6299
+ 1
6300
+ 1
6301
+ 1
6302
+ 1
6303
+ 1
6304
+ 1
6305
+ 1
6306
+ 1
6307
+ 1
6308
+ 1
6309
+ 1
6310
+ 1
6311
+ 1
6312
+ 1
6313
+ 1
6314
+ 1
6315
+ 1
6316
+ 1
6317
+ 1
6318
+ 1
6319
+ 1
6320
+ 1
6321
+ 1
6322
+ 1
6323
+ 1
6324
+ 1
6325
+ 1
6326
+ 1
6327
+ 1
6328
+ 1
6329
+ 1
6330
+ 1
6331
+ 1
6332
+ 1
6333
+ 1
6334
+ 1
6335
+ 1
6336
+ 1
6337
+ 1
6338
+ 1
6339
+ 1
6340
+ 1
6341
+ 1
6342
+ 1
6343
+ 1
6344
+ 2
6345
+ 1
6346
+ 1
6347
+ 1
6348
+ 1
6349
+ 1
6350
+ 1
6351
+ 1
6352
+ 1
6353
+ 1
6354
+ 1
6355
+ 1
6356
+ 1
6357
+ 1
6358
+ 1
6359
+ 1
6360
+ 1
6361
+ 1
6362
+ 1
6363
+ 1
6364
+ 1
6365
+ 1
6366
+ 1
6367
+ 1
6368
+ 1
6369
+ 1
6370
+ 1
6371
+ 1
6372
+ 1
6373
+ 2
6374
+ 1
6375
+ 1
6376
+ 1
6377
+ 1
6378
+ 1
6379
+ 1
6380
+ 2
6381
+ 1
6382
+ 1
6383
+ 1
6384
+ 1
6385
+ 1
6386
+ 1
6387
+ 1
6388
+ 1
6389
+ 1
6390
+ 1
6391
+ 1
6392
+ 1
6393
+ 1
6394
+ 1
6395
+ 1
6396
+ 1
6397
+ 1
6398
+ 1
6399
+ 1
6400
+ 1
6401
+ 1
6402
+ 1
6403
+ 1
6404
+ 1
6405
+ 1
6406
+ 1
6407
+ 1
6408
+ 1
6409
+ 1
6410
+ 1
6411
+ 2
6412
+ 2
6413
+ 1
6414
+ 1
6415
+ 1
6416
+ 1
6417
+ 1
6418
+ 1
6419
+ 1
6420
+ 1
6421
+ 1
6422
+ 1
6423
+ 1
6424
+ 1
6425
+ 1
6426
+ 1
6427
+ 1
6428
+ 1
6429
+ 1
6430
+ 1
6431
+ 1
6432
+ 1
6433
+ 1
6434
+ 1
6435
+ 1
6436
+ 1
6437
+ 1
6438
+ 1
6439
+ 1
6440
+ 1
6441
+ 1
6442
+ 1
6443
+ 2
6444
+ 1
6445
+ 1
6446
+ 1
6447
+ 1
6448
+ 1
6449
+ 1
6450
+ 1
6451
+ 1
6452
+ 1
6453
+ 1
6454
+ 2
6455
+ 1
6456
+ 1
6457
+ 1
6458
+ 1
6459
+ 1
6460
+ 1
6461
+ 1
6462
+ 1
6463
+ 1
6464
+ 1
6465
+ 1
6466
+ 1
6467
+ 1
6468
+ 1
6469
+ 1
6470
+ 1
6471
+ 1
6472
+ 1
6473
+ 1
6474
+ 1
6475
+ 1
6476
+ 1
6477
+ 1
6478
+ 1
6479
+ 1
6480
+ 1
6481
+ 1
6482
+ 2
6483
+ 1
6484
+ 1
6485
+ 1
6486
+ 1
6487
+ 1
6488
+ 1
6489
+ 1
6490
+ 1
6491
+ 1
6492
+ 1
6493
+ 1
6494
+ 1
6495
+ 1
6496
+ 1
6497
+ 1
6498
+ 1
6499
+ 2
6500
+ 1
6501
+ 2
6502
+ 1
6503
+ 1
6504
+ 1
6505
+ 1
6506
+ 1
6507
+ 1
6508
+ 1
6509
+ 1
6510
+ 2
6511
+ 1
6512
+ 1
6513
+ 1
6514
+ 1
6515
+ 1
6516
+ 1
6517
+ 1
6518
+ 1
6519
+ 1
6520
+ 1
6521
+ 1
6522
+ 1
6523
+ 1
6524
+ 1
6525
+ 1
6526
+ 1
6527
+ 1
6528
+ 1
6529
+ 1
6530
+ 1
6531
+ 1
6532
+ 1
6533
+ 2
6534
+ 1
6535
+ 1
6536
+ 1
6537
+ 1
6538
+ 1
6539
+ 1
6540
+ 1
6541
+ 1
6542
+ 2
6543
+ 1
6544
+ 2
6545
+ 1
6546
+ 1
6547
+ 1
6548
+ 1
6549
+ 1
6550
+ 1
6551
+ 1
6552
+ 1
6553
+ 1
6554
+ 1
6555
+ 1
6556
+ 1
6557
+ 1
6558
+ 1
6559
+ 1
6560
+ 1
6561
+ 2
6562
+ 1
6563
+ 1
6564
+ 1
6565
+ 1
6566
+ 1
6567
+ 1
6568
+ 1
6569
+ 1
6570
+ 1
6571
+ 1
6572
+ 1
6573
+ 1
6574
+ 1
6575
+ 1
6576
+ 1
6577
+ 2
6578
+ 1
6579
+ 1
6580
+ 1
6581
+ 2
6582
+ 1
6583
+ 1
6584
+ 1
6585
+ 1
6586
+ 1
6587
+ 1
6588
+ 1
6589
+ 1
6590
+ 1
6591
+ 1
6592
+ 1
6593
+ 1
6594
+ 1
6595
+ 2
6596
+ 2
6597
+ 1
6598
+ 1
6599
+ 1
6600
+ 1
6601
+ 1
6602
+ 1
6603
+ 1
6604
+ 1
6605
+ 1
6606
+ 1
6607
+ 1
6608
+ 1
6609
+ 1
6610
+ 2
6611
+ 1
6612
+ 1
6613
+ 1
6614
+ 1
6615
+ 1
6616
+ 1
6617
+ 1
6618
+ 1
6619
+ 1
6620
+ 1
6621
+ 1
6622
+ 1
6623
+ 1
6624
+ 1
6625
+ 1
6626
+ 2
6627
+ 1
6628
+ 1
6629
+ 1
6630
+ 2
6631
+ 1
6632
+ 1
6633
+ 1
6634
+ 1
6635
+ 1
6636
+ 1
6637
+ 1
6638
+ 1
6639
+ 1
6640
+ 1
6641
+ 1
6642
+ 1
6643
+ 1
6644
+ 1
6645
+ 2
6646
+ 1
6647
+ 1
6648
+ 1
6649
+ 1
6650
+ 1
6651
+ 1
6652
+ 1
6653
+ 1
6654
+ 1
6655
+ 1
6656
+ 1
6657
+ 1
6658
+ 1
6659
+ 2
6660
+ 1
6661
+ 1
6662
+ 1
6663
+ 1
6664
+ 1
6665
+ 1
6666
+ 1
6667
+ 1
6668
+ 1
6669
+ 1
6670
+ 1
6671
+ 1
6672
+ 1
6673
+ 1
6674
+ 2
6675
+ 1
6676
+ 1
6677
+ 1
6678
+ 1
6679
+ 1
6680
+ 1
6681
+ 2
6682
+ 1
6683
+ 1
6684
+ 1
6685
+ 1
6686
+ 1
6687
+ 1
6688
+ 1
6689
+ 1
6690
+ 1
6691
+ 1
6692
+ 1
6693
+ 1
6694
+ 1
6695
+ 1
6696
+ 1
6697
+ 1
6698
+ 1
6699
+ 2
6700
+ 2
6701
+ 1
6702
+ 1
6703
+ 2
6704
+ 1
6705
+ 1
6706
+ 2
6707
+ 1
6708
+ 1
6709
+ 1
6710
+ 1
6711
+ 1
6712
+ 1
6713
+ 1
6714
+ 1
6715
+ 1
6716
+ 1
6717
+ 1
6718
+ 1
6719
+ 1
6720
+ 1
6721
+ 1
6722
+ 1
6723
+ 1
6724
+ 1
6725
+ 1
6726
+ 1
6727
+ 1
6728
+ 1
6729
+ 1
6730
+ 1
6731
+ 1
6732
+ 1
6733
+ 1
6734
+ 1
6735
+ 1
6736
+ 1
6737
+ 1
6738
+ 1
6739
+ 1
6740
+ 1
6741
+ 1
6742
+ 2
6743
+ 1
6744
+ 1
6745
+ 1
6746
+ 1
6747
+ 1
6748
+ 1
6749
+ 1
6750
+ 1
6751
+ 1
6752
+ 1
6753
+ 1
6754
+ 1
6755
+ 1
6756
+ 1
6757
+ 1
6758
+ 1
6759
+ 1
6760
+ 2
6761
+ 1
6762
+ 1
6763
+ 1
6764
+ 1
6765
+ 1
6766
+ 1
6767
+ 1
6768
+ 1
6769
+ 1
6770
+ 1
6771
+ 1
6772
+ 1
6773
+ 1
6774
+ 1
6775
+ 1
6776
+ 1
6777
+ 1
6778
+ 1
6779
+ 1
6780
+ 1
6781
+ 1
6782
+ 1
6783
+ 1
6784
+ 1
6785
+ 1
6786
+ 2
6787
+ 1
6788
+ 1
6789
+ 1
6790
+ 1
6791
+ 1
6792
+ 1
6793
+ 2
6794
+ 1
6795
+ 2
6796
+ 1
6797
+ 1
6798
+ 1
6799
+ 1
6800
+ 1
6801
+ 1
6802
+ 1
6803
+ 1
6804
+ 1
6805
+ 1
6806
+ 1
6807
+ 1
6808
+ 1
6809
+ 1
6810
+ 2
6811
+ 2
6812
+ 1
6813
+ 1
6814
+ 1
6815
+ 1
6816
+ 1
6817
+ 1
6818
+ 1
6819
+ 1
6820
+ 1
6821
+ 1
6822
+ 1
6823
+ 1
6824
+ 1
6825
+ 1
6826
+ 1
6827
+ 1
6828
+ 1
6829
+ 1
6830
+ 1
6831
+ 2
6832
+ 1
6833
+ 1
6834
+ 1
6835
+ 1
6836
+ 1
6837
+ 1
6838
+ 1
6839
+ 2
6840
+ 1
6841
+ 1
6842
+ 1
6843
+ 1
6844
+ 1
6845
+ 1
6846
+ 1
6847
+ 1
6848
+ 1
6849
+ 1
6850
+ 1
6851
+ 1
6852
+ 1
6853
+ 1
6854
+ 1
6855
+ 1
6856
+ 1
6857
+ 1
6858
+ 1
6859
+ 1
6860
+ 1
6861
+ 1
6862
+ 1
6863
+ 1
6864
+ 1
6865
+ 1
6866
+ 1
6867
+ 1
6868
+ 1
6869
+ 1
6870
+ 2
6871
+ 1
6872
+ 2
6873
+ 1
6874
+ 1
6875
+ 1
6876
+ 1
6877
+ 1
6878
+ 1
6879
+ 1
6880
+ 1
6881
+ 1
6882
+ 1
6883
+ 1
6884
+ 1
6885
+ 1
6886
+ 1
6887
+ 1
6888
+ 1
6889
+ 1
6890
+ 2
6891
+ 1
6892
+ 1
6893
+ 1
6894
+ 1
6895
+ 1
6896
+ 1
6897
+ 1
6898
+ 1
6899
+ 1
6900
+ 1
6901
+ 1
6902
+ 1
6903
+ 1
6904
+ 1
6905
+ 1
6906
+ 1
6907
+ 1
6908
+ 1
6909
+ 1
6910
+ 1
6911
+ 1
6912
+ 1
6913
+ 1
6914
+ 1
6915
+ 1
6916
+ 1
6917
+ 1
6918
+ 1
6919
+ 1
6920
+ 1
6921
+ 1
6922
+ 1
6923
+ 1
6924
+ 1
6925
+ 1
6926
+ 2
6927
+ 1
6928
+ 1
6929
+ 1
6930
+ 1
6931
+ 1
6932
+ 1
6933
+ 1
6934
+ 1
6935
+ 1
6936
+ 1
6937
+ 1
6938
+ 1
6939
+ 1
6940
+ 1
6941
+ 1
6942
+ 1
6943
+ 1
6944
+ 1
6945
+ 1
6946
+ 1
6947
+ 1
6948
+ 1
6949
+ 1
6950
+ 1
6951
+ 1
6952
+ 1
6953
+ 1
6954
+ 1
6955
+ 1
6956
+ 1
6957
+ 1
6958
+ 1
6959
+ 1
6960
+ 1
6961
+ 1
6962
+ 1
6963
+ 1
6964
+ 1
6965
+ 1
6966
+ 1
6967
+ 1
6968
+ 1
6969
+ 1
6970
+ 1
6971
+ 1
6972
+ 1
6973
+ 1
6974
+ 1
6975
+ 1
6976
+ 1
6977
+ 1
6978
+ 1
6979
+ 1
6980
+ 1
6981
+ 1
6982
+ 1
6983
+ 1
6984
+ 1
6985
+ 1
6986
+ 1
6987
+ 1
6988
+ 1
6989
+ 1
6990
+ 1
6991
+ 1
6992
+ 1
6993
+ 1
6994
+ 1
6995
+ 1
6996
+ 2
6997
+ 1
6998
+ 1
6999
+ 1
7000
+ 1
7001
+ 1
7002
+ 1
7003
+ 1
7004
+ 2
7005
+ 1
7006
+ 1
7007
+ 1
7008
+ 1
7009
+ 1
7010
+ 1
7011
+ 1
7012
+ 1
7013
+ 1
7014
+ 1
7015
+ 1
7016
+ 1
7017
+ 1
7018
+ 1
7019
+ 1
7020
+ 1
7021
+ 1
7022
+ 1
7023
+ 1
7024
+ 1
7025
+ 1
7026
+ 1
7027
+ 2
7028
+ 1
7029
+ 1
7030
+ 2
7031
+ 1
7032
+ 1
7033
+ 1
7034
+ 1
7035
+ 1
7036
+ 1
7037
+ 1
7038
+ 1
7039
+ 1
7040
+ 1
7041
+ 1
7042
+ 1
7043
+ 1
7044
+ 1
7045
+ 1
7046
+ 1
7047
+ 1
7048
+ 1
7049
+ 1
7050
+ 1
7051
+ 1
7052
+ 1
7053
+ 1
7054
+ 1
7055
+ 1
7056
+ 1
7057
+ 1
7058
+ 1
7059
+ 1
7060
+ 1
7061
+ 1
7062
+ 1
7063
+ 1
7064
+ 1
7065
+ 1
7066
+ 1
7067
+ 1
7068
+ 1
7069
+ 1
7070
+ 1
7071
+ 1
7072
+ 1
7073
+ 1
7074
+ 1
7075
+ 1
7076
+ 1
7077
+ 1
7078
+ 1
7079
+ 1
7080
+ 1
7081
+ 2
7082
+ 1
7083
+ 2
7084
+ 1
7085
+ 1
7086
+ 1
7087
+ 1
7088
+ 1
7089
+ 1
7090
+ 1
7091
+ 1
7092
+ 1
7093
+ 1
7094
+ 1
7095
+ 1
7096
+ 1
7097
+ 1
7098
+ 2
7099
+ 1
7100
+ 1
7101
+ 1
7102
+ 1
7103
+ 2
7104
+ 1
7105
+ 1
7106
+ 1
7107
+ 1
7108
+ 1
7109
+ 1
7110
+ 1
7111
+ 1
7112
+ 1
7113
+ 1
7114
+ 1
7115
+ 1
7116
+ 1
7117
+ 2
7118
+ 1
7119
+ 1
7120
+ 1
7121
+ 1
7122
+ 1
7123
+ 1
7124
+ 1
7125
+ 1
7126
+ 1
7127
+ 1
7128
+ 1
7129
+ 1
7130
+ 1
7131
+ 1
7132
+ 1
7133
+ 1
7134
+ 1
7135
+ 1
7136
+ 1
7137
+ 1
7138
+ 1
7139
+ 1
7140
+ 1
7141
+ 1
7142
+ 1
7143
+ 1
7144
+ 1
7145
+ 1
7146
+ 1
7147
+ 1
7148
+ 1
7149
+ 1
7150
+ 1
7151
+ 1
7152
+ 1
7153
+ 1
7154
+ 1
7155
+ 1
7156
+ 1
7157
+ 1
7158
+ 1
7159
+ 1
7160
+ 1
7161
+ 1
7162
+ 1
7163
+ 1
7164
+ 1
7165
+ 1
7166
+ 2
7167
+ 1
7168
+ 1
7169
+ 1
7170
+ 1
7171
+ 1
7172
+ 1
7173
+ 1
7174
+ 1
7175
+ 1
7176
+ 1
7177
+ 1
7178
+ 1
7179
+ 1
7180
+ 1
7181
+ 1
7182
+ 1
7183
+ 1
7184
+ 1
7185
+ 1
7186
+ 1
7187
+ 1
7188
+ 1
7189
+ 1
7190
+ 1
7191
+ 1
7192
+ 1
7193
+ 1
7194
+ 1
7195
+ 1
7196
+ 1
7197
+ 1
7198
+ 1
7199
+ 1
7200
+ 1
7201
+ 2
7202
+ 1
7203
+ 1
7204
+ 1
7205
+ 1
7206
+ 1
7207
+ 1
7208
+ 1
7209
+ 1
7210
+ 1
7211
+ 1
7212
+ 1
7213
+ 1
7214
+ 1
7215
+ 1
7216
+ 1
7217
+ 1
7218
+ 1
7219
+ 1
7220
+ 1
7221
+ 1
7222
+ 1
7223
+ 1
7224
+ 1
7225
+ 1
7226
+ 1
7227
+ 1
7228
+ 1
7229
+ 1
7230
+ 1
7231
+ 1
7232
+ 1
7233
+ 2
7234
+ 1
7235
+ 1
7236
+ 1
7237
+ 1
7238
+ 1
7239
+ 1
7240
+ 1
7241
+ 1
7242
+ 1
7243
+ 1
7244
+ 1
7245
+ 1
7246
+ 1
7247
+ 1
7248
+ 1
7249
+ 1
7250
+ 1
7251
+ 1
7252
+ 1
7253
+ 1
7254
+ 1
7255
+ 1
7256
+ 1
7257
+ 1
7258
+ 1
7259
+ 1
7260
+ 1
7261
+ 1
7262
+ 1
7263
+ 1
7264
+ 1
7265
+ 1
7266
+ 1
7267
+ 1
7268
+ 1
7269
+ 1
7270
+ 1
7271
+ 1
7272
+ 2
7273
+ 1
7274
+ 1
7275
+ 1
7276
+ 1
7277
+ 1
7278
+ 1
7279
+ 1
7280
+ 1
7281
+ 1
7282
+ 1
7283
+ 2
7284
+ 1
7285
+ 1
7286
+ 1
7287
+ 1
7288
+ 1
7289
+ 1
7290
+ 1
7291
+ 1
7292
+ 1
7293
+ 1
7294
+ 1
7295
+ 1
7296
+ 1
7297
+ 1
7298
+ 1
7299
+ 1
7300
+ 1
7301
+ 1
7302
+ 1
7303
+ 1
7304
+ 1
7305
+ 1
7306
+ 1
7307
+ 1
7308
+ 1
7309
+ 1
7310
+ 1
7311
+ 1
7312
+ 1
7313
+ 1
7314
+ 1
7315
+ 1
7316
+ 1
7317
+ 1
7318
+ 1
7319
+ 1
7320
+ 1
7321
+ 1
7322
+ 1
7323
+ 1
7324
+ 1
7325
+ 2
7326
+ 1
7327
+ 1
7328
+ 1
7329
+ 1
7330
+ 1
7331
+ 1
7332
+ 1
7333
+ 1
7334
+ 2
7335
+ 1
7336
+ 2
7337
+ 1
7338
+ 1
7339
+ 1
7340
+ 1
7341
+ 1
7342
+ 1
7343
+ 1
7344
+ 1
7345
+ 1
7346
+ 1
7347
+ 1
7348
+ 1
7349
+ 1
7350
+ 1
7351
+ 1
7352
+ 1
7353
+ 1
7354
+ 1
7355
+ 1
7356
+ 1
7357
+ 1
7358
+ 1
7359
+ 1
7360
+ 1
7361
+ 1
7362
+ 1
7363
+ 1
7364
+ 1
7365
+ 1
7366
+ 1
7367
+ 1
7368
+ 1
7369
+ 1
7370
+ 1
7371
+ 1
7372
+ 1
7373
+ 2
7374
+ 1
7375
+ 1
7376
+ 1
7377
+ 1
7378
+ 1
7379
+ 1
7380
+ 1
7381
+ 1
7382
+ 1
7383
+ 1
7384
+ 1
7385
+ 1
7386
+ 1
7387
+ 1
7388
+ 2
7389
+ 1
7390
+ 1
7391
+ 1
7392
+ 1
7393
+ 1
7394
+ 1
7395
+ 1
7396
+ 1
7397
+ 1
7398
+ 1
7399
+ 1
7400
+ 1
7401
+ 1
7402
+ 1
7403
+ 1
7404
+ 1
7405
+ 1
7406
+ 1
7407
+ 1
7408
+ 2
7409
+ 1
7410
+ 1
7411
+ 1
7412
+ 1
7413
+ 1
7414
+ 1
7415
+ 1
7416
+ 1
7417
+ 1
7418
+ 1
7419
+ 1
7420
+ 1
7421
+ 1
7422
+ 1
7423
+ 1
7424
+ 1
7425
+ 1
7426
+ 1
7427
+ 1
7428
+ 1
7429
+ 1
7430
+ 1
7431
+ 1
7432
+ 1
7433
+ 1
7434
+ 1
7435
+ 1
7436
+ 1
7437
+ 1
7438
+ 1
7439
+ 1
7440
+ 1
7441
+ 1
7442
+ 1
7443
+ 1
7444
+ 1
7445
+ 1
7446
+ 1
7447
+ 1
7448
+ 1
7449
+ 1
7450
+ 1
7451
+ 1
7452
+ 1
7453
+ 1
7454
+ 1
7455
+ 1
7456
+ 1
7457
+ 1
7458
+ 1
7459
+ 1
7460
+ 1
7461
+ 1
7462
+ 1
7463
+ 1
7464
+ 1
7465
+ 1
7466
+ 1
7467
+ 1
7468
+ 1
7469
+ 1
7470
+ 1
7471
+ 1
7472
+ 1
7473
+ 2
7474
+ 1
7475
+ 2
7476
+ 1
7477
+ 1
7478
+ 1
7479
+ 1
7480
+ 1
7481
+ 1
7482
+ 1
7483
+ 2
7484
+ 1
7485
+ 1
7486
+ 1
7487
+ 1
7488
+ 1
7489
+ 1
7490
+ 2
7491
+ 1
7492
+ 1
7493
+ 1
7494
+ 1
7495
+ 1
7496
+ 1
7497
+ 1
7498
+ 1
7499
+ 1
7500
+ 2
7501
+ 1
7502
+ 1
7503
+ 1
7504
+ 1
7505
+ 1
7506
+ 1
7507
+ 1
7508
+ 1
7509
+ 1
7510
+ 1
7511
+ 1
7512
+ 1
7513
+ 1
7514
+ 1
7515
+ 1
7516
+ 1
7517
+ 2
7518
+ 1
7519
+ 1
7520
+ 1
7521
+ 1
7522
+ 1
7523
+ 1
7524
+ 1
7525
+ 1
7526
+ 1
7527
+ 1
7528
+ 1
7529
+ 1
7530
+ 1
7531
+ 1
7532
+ 1
7533
+ 1
7534
+ 2
7535
+ 1
7536
+ 1
7537
+ 2
7538
+ 1
7539
+ 1
7540
+ 1
7541
+ 1
7542
+ 1
7543
+ 1
7544
+ 1
7545
+ 1
7546
+ 1
7547
+ 1
7548
+ 1
7549
+ 1
7550
+ 1
7551
+ 1
7552
+ 1
7553
+ 1
7554
+ 1
7555
+ 1
7556
+ 1
7557
+ 1
7558
+ 1
7559
+ 1
7560
+ 1
7561
+ 1
7562
+ 1
7563
+ 1
7564
+ 1
7565
+ 1
7566
+ 1
7567
+ 2
7568
+ 1
7569
+ 1
7570
+ 1
7571
+ 1
7572
+ 1
7573
+ 1
7574
+ 1
7575
+ 1
7576
+ 1
7577
+ 1
7578
+ 1
7579
+ 1
7580
+ 1
7581
+ 1
7582
+ 1
7583
+ 1
7584
+ 1
7585
+ 1
7586
+ 1
7587
+ 1
7588
+ 1
7589
+ 1
7590
+ 1
7591
+ 1
7592
+ 1
7593
+ 1
7594
+ 1
7595
+ 1
7596
+ 1
7597
+ 1
7598
+ 1
7599
+ 1
7600
+ 1
7601
+ 1
7602
+ 1
7603
+ 1
7604
+ 1
7605
+ 1
7606
+ 1
7607
+ 1
7608
+ 1
7609
+ 1
7610
+ 1
7611
+ 1
7612
+ 1
7613
+ 1
7614
+ 1
7615
+ 1
7616
+ 1
7617
+ 1
7618
+ 1
7619
+ 1
7620
+ 1
7621
+ 2
7622
+ 1
7623
+ 1
7624
+ 1
7625
+ 1
7626
+ 1
7627
+ 1
7628
+ 1
7629
+ 1
7630
+ 1
7631
+ 1
7632
+ 1
7633
+ 1
7634
+ 1
7635
+ 1
7636
+ 1
7637
+ 1
7638
+ 1
7639
+ 1
7640
+ 1
7641
+ 1
7642
+ 1
7643
+ 1
7644
+ 1
7645
+ 1
7646
+ 1
7647
+ 1
7648
+ 1
7649
+ 1
7650
+ 1
7651
+ 1
7652
+ 1
7653
+ 1
7654
+ 1
7655
+ 2
7656
+ 1
7657
+ 1
7658
+ 1
7659
+ 1
7660
+ 1
7661
+ 1
7662
+ 1
7663
+ 1
7664
+ 1
7665
+ 1
7666
+ 1
7667
+ 1
7668
+ 1
7669
+ 1
7670
+ 1
7671
+ 1
7672
+ 1
7673
+ 1
7674
+ 1
7675
+ 1
7676
+ 1
7677
+ 1
7678
+ 1
7679
+ 1
7680
+ 1
7681
+ 1
7682
+ 1
7683
+ 1
7684
+ 1
7685
+ 1
7686
+ 1
7687
+ 1
7688
+ 1
7689
+ 1
7690
+ 1
7691
+ 1
7692
+ 2
7693
+ 1
7694
+ 1
7695
+ 1
7696
+ 1
7697
+ 1
7698
+ 1
7699
+ 1
7700
+ 1
7701
+ 1
7702
+ 1
7703
+ 1
7704
+ 1
7705
+ 2
7706
+ 1
7707
+ 1
7708
+ 1
7709
+ 1
7710
+ 1
7711
+ 1
7712
+ 1
7713
+ 1
7714
+ 1
7715
+ 1
7716
+ 1
7717
+ 1
7718
+ 1
7719
+ 1
7720
+ 1
7721
+ 1
7722
+ 1
7723
+ 2
7724
+ 1
7725
+ 1
7726
+ 1
7727
+ 1
7728
+ 1
7729
+ 1
7730
+ 1
7731
+ 1
7732
+ 1
7733
+ 1
7734
+ 1
7735
+ 1
7736
+ 1
7737
+ 1
7738
+ 1
7739
+ 1
7740
+ 1
7741
+ 1
7742
+ 1
7743
+ 1
7744
+ 1
7745
+ 1
7746
+ 1
7747
+ 2
7748
+ 1
7749
+ 1
7750
+ 1
7751
+ 2
7752
+ 1
7753
+ 1
7754
+ 1
7755
+ 1
7756
+ 1
7757
+ 1
7758
+ 1
7759
+ 1
7760
+ 1
7761
+ 1
7762
+ 1
7763
+ 1
7764
+ 1
7765
+ 1
7766
+ 1
7767
+ 1
7768
+ 1
7769
+ 1
7770
+ 1
7771
+ 1
7772
+ 1
7773
+ 1
7774
+ 2
7775
+ 1
7776
+ 1
7777
+ 1
7778
+ 1
7779
+ 1
7780
+ 1
7781
+ 1
7782
+ 1
7783
+ 1
7784
+ 1
7785
+ 1
7786
+ 1
7787
+ 1
7788
+ 1
7789
+ 1
7790
+ 1
7791
+ 1
7792
+ 1
7793
+ 1
7794
+ 1
7795
+ 1
7796
+ 1
7797
+ 1
7798
+ 1
7799
+ 1
7800
+ 1
7801
+ 1
7802
+ 1
7803
+ 1
7804
+ 1
7805
+ 1
7806
+ 1
7807
+ 2
7808
+ 1
7809
+ 1
7810
+ 1
7811
+ 1
7812
+ 1
7813
+ 1
7814
+ 1
7815
+ 1
7816
+ 1
7817
+ 1
7818
+ 1
7819
+ 1
7820
+ 1
7821
+ 1
7822
+ 2
7823
+ 1
7824
+ 1
7825
+ 1
7826
+ 1
7827
+ 1
7828
+ 2
7829
+ 1
7830
+ 1
7831
+ 1
7832
+ 1
7833
+ 1
7834
+ 1
7835
+ 1
7836
+ 1
7837
+ 1
7838
+ 1
7839
+ 1
7840
+ 1
7841
+ 1
7842
+ 1
7843
+ 1
7844
+ 1
7845
+ 1
7846
+ 1
7847
+ 1
7848
+ 1
7849
+ 1
7850
+ 1
7851
+ 1
7852
+ 2
7853
+ 1
7854
+ 1
7855
+ 1
7856
+ 1
7857
+ 1
7858
+ 1
7859
+ 1
7860
+ 1
7861
+ 1
7862
+ 1
7863
+ 1
7864
+ 1
7865
+ 1
7866
+ 1
7867
+ 1
7868
+ 1
7869
+ 1
7870
+ 1
7871
+ 1
7872
+ 1
7873
+ 1
7874
+ 2
7875
+ 1
7876
+ 1
7877
+ 1
7878
+ 1
7879
+ 1
7880
+ 1
7881
+ 2
7882
+ 1
7883
+ 1
7884
+ 1
7885
+ 2
7886
+ 1
7887
+ 1
7888
+ 1
7889
+ 1
7890
+ 1
7891
+ 1
7892
+ 1
7893
+ 1
7894
+ 1
7895
+ 1
7896
+ 1
7897
+ 1
7898
+ 1
7899
+ 1
7900
+ 1
7901
+ 1
7902
+ 1
7903
+ 1
7904
+ 1
7905
+ 2
7906
+ 1
7907
+ 1
7908
+ 1
7909
+ 1
7910
+ 1
7911
+ 1
7912
+ 1
7913
+ 2
7914
+ 1
7915
+ 1
7916
+ 1
7917
+ 1
7918
+ 1
7919
+ 1
7920
+ 1
7921
+ 1
7922
+ 1
7923
+ 1
7924
+ 1
7925
+ 1
7926
+ 1
7927
+ 1
7928
+ 1
7929
+ 1
7930
+ 1
7931
+ 1
7932
+ 1
7933
+ 1
7934
+ 1
7935
+ 1
7936
+ 1
7937
+ 1
7938
+ 1
7939
+ 1
7940
+ 1
7941
+ 1
7942
+ 1
7943
+ 1
7944
+ 1
7945
+ 1
7946
+ 1
7947
+ 1
7948
+ 1
7949
+ 2
7950
+ 1
7951
+ 1
7952
+ 1
7953
+ 1
7954
+ 1
7955
+ 1
7956
+ 1
7957
+ 1
7958
+ 1
7959
+ 1
7960
+ 1
7961
+ 1
7962
+ 1
7963
+ 1
7964
+ 1
7965
+ 2
7966
+ 2
7967
+ 1
7968
+ 1
7969
+ 1
7970
+ 1
7971
+ 1
7972
+ 1
7973
+ 1
7974
+ 1
7975
+ 1
7976
+ 1
7977
+ 1
7978
+ 1
7979
+ 1
7980
+ 1
7981
+ 1
7982
+ 1
7983
+ 1
7984
+ 1
7985
+ 2
7986
+ 1
7987
+ 1
7988
+ 1
7989
+ 1
7990
+ 2
7991
+ 1
7992
+ 1
7993
+ 2
7994
+ 1
7995
+ 1
7996
+ 1
7997
+ 1
7998
+ 1
7999
+ 1
8000
+ 1
8001
+ 1
8002
+ 1
8003
+ 1
8004
+ 1
8005
+ 1
8006
+ 1
8007
+ 1
8008
+ 1
8009
+ 1
8010
+ 1
8011
+ 1
8012
+ 1
8013
+ 1
8014
+ 1
8015
+ 1
8016
+ 1
8017
+ 1
8018
+ 1
8019
+ 1
8020
+ 1
8021
+ 1
8022
+ 1
8023
+ 1
8024
+ 1
8025
+ 1
8026
+ 1
8027
+ 1
8028
+ 1
8029
+ 1
8030
+ 1
8031
+ 1
8032
+ 1
8033
+ 1
8034
+ 1
8035
+ 1
8036
+ 2
8037
+ 1
8038
+ 1
8039
+ 1
8040
+ 1
8041
+ 1
8042
+ 1
8043
+ 1
8044
+ 1
8045
+ 1
8046
+ 1
8047
+ 1
8048
+ 1
8049
+ 1
8050
+ 1
8051
+ 2
8052
+ 1
8053
+ 1
8054
+ 1
8055
+ 1
8056
+ 1
8057
+ 1
8058
+ 1
8059
+ 1
8060
+ 1
8061
+ 1
8062
+ 1
8063
+ 1
8064
+ 1
8065
+ 1
8066
+ 1
8067
+ 1
8068
+ 1
8069
+ 1
8070
+ 1
8071
+ 1
8072
+ 1
8073
+ 1
8074
+ 1
8075
+ 2
8076
+ 1
8077
+ 1
8078
+ 1
8079
+ 1
8080
+ 1
8081
+ 1
8082
+ 1
8083
+ 1
8084
+ 1
8085
+ 1
8086
+ 1
8087
+ 1
8088
+ 1
8089
+ 1
8090
+ 1
8091
+ 1
8092
+ 2
8093
+ 1
8094
+ 1
8095
+ 2
8096
+ 1
8097
+ 1
8098
+ 1
8099
+ 1
8100
+ 1
8101
+ 1
8102
+ 1
8103
+ 1
8104
+ 1
8105
+ 1
8106
+ 1
8107
+ 1
8108
+ 1
8109
+ 1
8110
+ 1
8111
+ 1
8112
+ 1
8113
+ 1
8114
+ 2
8115
+ 1
8116
+ 1
8117
+ 2
8118
+ 1
8119
+ 1
8120
+ 1
8121
+ 1
8122
+ 1
8123
+ 1
8124
+ 1
8125
+ 1
8126
+ 1
8127
+ 1
8128
+ 1
8129
+ 1
8130
+ 1
8131
+ 1
8132
+ 1
8133
+ 1
8134
+ 1
8135
+ 1
8136
+ 1
8137
+ 1
8138
+ 1
8139
+ 1
8140
+ 1
8141
+ 1
8142
+ 1
8143
+ 1
8144
+ 1
8145
+ 1
8146
+ 1
8147
+ 1
8148
+ 1
8149
+ 1
8150
+ 1
8151
+ 1
8152
+ 2
8153
+ 1
8154
+ 1
8155
+ 1
8156
+ 1
8157
+ 1
8158
+ 1
8159
+ 1
8160
+ 2
8161
+ 1
8162
+ 1
8163
+ 1
8164
+ 1
8165
+ 1
8166
+ 1
8167
+ 1
8168
+ 1
8169
+ 1
8170
+ 1
8171
+ 1
8172
+ 2
8173
+ 1
8174
+ 1
8175
+ 1
8176
+ 1
8177
+ 1
8178
+ 1
8179
+ 1
8180
+ 2
8181
+ 1
8182
+ 1
8183
+ 1
8184
+ 1
8185
+ 1
8186
+ 1
8187
+ 1
8188
+ 1
8189
+ 1
8190
+ 1
8191
+ 1
8192
+ 1
8193
+ 1
8194
+ 1
8195
+ 2
8196
+ 2
8197
+ 1
8198
+ 1
8199
+ 1
8200
+ 1
8201
+ 1
8202
+ 1
8203
+ 1
8204
+ 1
8205
+ 1
8206
+ 1
8207
+ 1
8208
+ 1
8209
+ 1
8210
+ 1
8211
+ 1
8212
+ 1
8213
+ 1
8214
+ 1
8215
+ 1
8216
+ 1
8217
+ 1
8218
+ 1
8219
+ 1
8220
+ 1
8221
+ 1
8222
+ 1
8223
+ 1
8224
+ 1
8225
+ 1
8226
+ 1
8227
+ 1
8228
+ 1
8229
+ 1
8230
+ 1
8231
+ 1
8232
+ 1
8233
+ 1
8234
+ 1
8235
+ 1
8236
+ 1
8237
+ 1
8238
+ 1
8239
+ 1
8240
+ 2
8241
+ 1
8242
+ 1
8243
+ 1
8244
+ 1
8245
+ 1
8246
+ 1
8247
+ 1
8248
+ 1
8249
+ 1
8250
+ 1
8251
+ 1
8252
+ 1
8253
+ 1
8254
+ 1
8255
+ 1
8256
+ 1
8257
+ 1
8258
+ 1
8259
+ 1
8260
+ 1
8261
+ 1
8262
+ 1
8263
+ 1
8264
+ 2
8265
+ 1
8266
+ 1
8267
+ 1
8268
+ 1
8269
+ 1
8270
+ 1
8271
+ 1
8272
+ 1
8273
+ 1
8274
+ 1
8275
+ 1
8276
+ 1
8277
+ 1
8278
+ 1
8279
+ 1
8280
+ 1
8281
+ 1
8282
+ 1
8283
+ 1
8284
+ 1
8285
+ 1
8286
+ 1
8287
+ 1
8288
+ 1
8289
+ 1
8290
+ 1
8291
+ 2
8292
+ 1
8293
+ 1
8294
+ 1
8295
+ 1
8296
+ 2
8297
+ 1
8298
+ 1
8299
+ 1
8300
+ 1
8301
+ 1
8302
+ 1
8303
+ 1
8304
+ 1
8305
+ 1
8306
+ 1
8307
+ 1
8308
+ 1
8309
+ 1
8310
+ 1
8311
+ 1
8312
+ 1
8313
+ 2
8314
+ 1
8315
+ 1
8316
+ 1
8317
+ 1
8318
+ 1
8319
+ 1
8320
+ 1
8321
+ 1
8322
+ 1
8323
+ 1
8324
+ 1
8325
+ 1
8326
+ 1
8327
+ 1
8328
+ 1
8329
+ 1
8330
+ 1
8331
+ 1
8332
+ 1
8333
+ 1
8334
+ 1
8335
+ 1
8336
+ 1
8337
+ 1
8338
+ 1
8339
+ 1
8340
+ 1
8341
+ 1
8342
+ 1
8343
+ 1
8344
+ 1
8345
+ 1
8346
+ 1
8347
+ 1
8348
+ 1
8349
+ 1
8350
+ 1
8351
+ 1
8352
+ 1
8353
+ 1
8354
+ 1
8355
+ 1
8356
+ 1
8357
+ 1
8358
+ 1
8359
+ 1
8360
+ 1
8361
+ 1
8362
+ 1
8363
+ 1
8364
+ 1
8365
+ 1
8366
+ 1
8367
+ 1
8368
+ 2
8369
+ 1
8370
+ 1
8371
+ 1
8372
+ 1
8373
+ 1
8374
+ 1
8375
+ 2
8376
+ 1
8377
+ 1
8378
+ 1
8379
+ 1
8380
+ 1
8381
+ 1
8382
+ 1
8383
+ 1
8384
+ 1
8385
+ 1
8386
+ 1
8387
+ 1
8388
+ 2
8389
+ 1
8390
+ 1
8391
+ 1
8392
+ 1
8393
+ 1
8394
+ 1
8395
+ 1
8396
+ 1
8397
+ 1
8398
+ 1
8399
+ 1
8400
+ 1
8401
+ 1
8402
+ 1
8403
+ 1
8404
+ 1
8405
+ 1
8406
+ 1
8407
+ 1
8408
+ 2
8409
+ 1
8410
+ 1
8411
+ 1
8412
+ 1
8413
+ 1
8414
+ 1
8415
+ 1
8416
+ 1
8417
+ 1
8418
+ 1
8419
+ 1
8420
+ 1
8421
+ 1
8422
+ 1
8423
+ 1
8424
+ 1
8425
+ 1
8426
+ 1
8427
+ 1
8428
+ 1
8429
+ 1
8430
+ 1
8431
+ 1
8432
+ 1
8433
+ 1
8434
+ 1
8435
+ 1
8436
+ 1
8437
+ 1
8438
+ 2
8439
+ 1
8440
+ 2
8441
+ 1
8442
+ 1
8443
+ 1
8444
+ 1
8445
+ 1
8446
+ 1
8447
+ 1
8448
+ 1
8449
+ 1
8450
+ 1
8451
+ 1
8452
+ 1
8453
+ 1
8454
+ 1
8455
+ 1
8456
+ 1
8457
+ 1
8458
+ 1
8459
+ 1
8460
+ 1
8461
+ 1
8462
+ 1
8463
+ 1
8464
+ 1
8465
+ 1
8466
+ 1
8467
+ 1
8468
+ 1
8469
+ 1
8470
+ 1
8471
+ 1
8472
+ 1
8473
+ 1
8474
+ 1
8475
+ 1
8476
+ 1
8477
+ 1
8478
+ 1
8479
+ 1
8480
+ 1
8481
+ 1
8482
+ 1
8483
+ 1
8484
+ 1
8485
+ 1
8486
+ 1
8487
+ 1
8488
+ 1
8489
+ 1
8490
+ 1
8491
+ 1
8492
+ 1
8493
+ 1
8494
+ 1
8495
+ 1
8496
+ 1
8497
+ 1
8498
+ 1
8499
+ 1
8500
+ 1
8501
+ 1
8502
+ 1
8503
+ 1
8504
+ 1
8505
+ 1
8506
+ 1
8507
+ 1
8508
+ 1
8509
+ 1
8510
+ 1
8511
+ 1
8512
+ 1
8513
+ 1
8514
+ 1
8515
+ 1
8516
+ 1
8517
+ 1
8518
+ 1
8519
+ 2
8520
+ 1
8521
+ 1
8522
+ 1
8523
+ 1
8524
+ 1
8525
+ 1
8526
+ 1
8527
+ 1
8528
+ 1
8529
+ 1
8530
+ 1
8531
+ 1
8532
+ 1
8533
+ 1
8534
+ 1
8535
+ 1
8536
+ 1
8537
+ 1
8538
+ 1
8539
+ 1
8540
+ 1
8541
+ 1
8542
+ 1
8543
+ 1
8544
+ 1
8545
+ 1
8546
+ 1
8547
+ 1
8548
+ 1
8549
+ 1
8550
+ 1
8551
+ 1
8552
+ 1
8553
+ 1
8554
+ 1
8555
+ 1
8556
+ 1
8557
+ 2
8558
+ 1
8559
+ 1
8560
+ 1
8561
+ 1
8562
+ 1
8563
+ 1
8564
+ 1
8565
+ 1
8566
+ 1
8567
+ 1
8568
+ 1
8569
+ 1
8570
+ 1
8571
+ 1
8572
+ 1
8573
+ 1
8574
+ 1
8575
+ 1
8576
+ 1
8577
+ 1
8578
+ 1
8579
+ 1
8580
+ 1
8581
+ 1
8582
+ 1
8583
+ 1
8584
+ 1
8585
+ 1
8586
+ 1
8587
+ 1
8588
+ 1
8589
+ 2
8590
+ 1
8591
+ 1
8592
+ 1
8593
+ 1
8594
+ 1
8595
+ 1
8596
+ 1
8597
+ 1
8598
+ 2
8599
+ 1
8600
+ 1
8601
+ 1
8602
+ 2
8603
+ 1
8604
+ 1
8605
+ 1
8606
+ 1
8607
+ 1
8608
+ 1
8609
+ 1
8610
+ 1
8611
+ 1
8612
+ 1
8613
+ 1
8614
+ 1
8615
+ 1
8616
+ 1
8617
+ 1
8618
+ 1
8619
+ 1
8620
+ 1
8621
+ 1
8622
+ 1
8623
+ 1
8624
+ 1
8625
+ 1
8626
+ 1
8627
+ 1
8628
+ 1
8629
+ 1
8630
+ 1
8631
+ 1
8632
+ 1
8633
+ 1
8634
+ 1
8635
+ 1
8636
+ 1
8637
+ 1
8638
+ 1
8639
+ 1
8640
+ 1
8641
+ 1
8642
+ 1
8643
+ 1
8644
+ 1
8645
+ 1
8646
+ 1
8647
+ 1
8648
+ 1
8649
+ 1
8650
+ 1
8651
+ 1
8652
+ 2
8653
+ 1
8654
+ 1
8655
+ 1
8656
+ 1
8657
+ 1
8658
+ 2
8659
+ 1
8660
+ 1
8661
+ 1
8662
+ 1
8663
+ 1
8664
+ 1
8665
+ 1
8666
+ 1
8667
+ 1
8668
+ 1
8669
+ 1
8670
+ 1
8671
+ 1
8672
+ 1
8673
+ 1
8674
+ 1
8675
+ 1
8676
+ 1
8677
+ 1
8678
+ 1
8679
+ 1
8680
+ 1
8681
+ 1
8682
+ 1
8683
+ 1
8684
+ 1
8685
+ 1
8686
+ 1
8687
+ 1
8688
+ 1
8689
+ 1
8690
+ 1
8691
+ 1
8692
+ 1
8693
+ 1
8694
+ 1
8695
+ 1
8696
+ 1
8697
+ 1
8698
+ 1
8699
+ 1
8700
+ 1
8701
+ 1
8702
+ 1
8703
+ 1
8704
+ 1
8705
+ 1
8706
+ 1
8707
+ 1
8708
+ 1
8709
+ 1
8710
+ 1
8711
+ 1
8712
+ 1
8713
+ 1
8714
+ 1
8715
+ 1
8716
+ 1
8717
+ 1
8718
+ 1
8719
+ 1
8720
+ 1
8721
+ 1
8722
+ 1
8723
+ 1
8724
+ 2
8725
+ 1
8726
+ 1
8727
+ 1
8728
+ 1
8729
+ 1
8730
+ 1
8731
+ 1
8732
+ 1
8733
+ 1
8734
+ 1
8735
+ 1
8736
+ 1
8737
+ 1
8738
+ 1
8739
+ 1
8740
+ 1
8741
+ 1
8742
+ 1
8743
+ 1
8744
+ 1
8745
+ 1
8746
+ 1
8747
+ 1
8748
+ 1
8749
+ 1
8750
+ 1
8751
+ 1
8752
+ 1
8753
+ 1
8754
+ 1
8755
+ 1
8756
+ 1
8757
+ 1
8758
+ 1
8759
+ 1
8760
+ 2
8761
+ 1
8762
+ 1
8763
+ 1
8764
+ 2
8765
+ 1
8766
+ 1
8767
+ 1
8768
+ 1
8769
+ 1
8770
+ 1
8771
+ 1
8772
+ 1
8773
+ 1
8774
+ 1
8775
+ 1
8776
+ 1
8777
+ 1
8778
+ 1
8779
+ 1
8780
+ 1
8781
+ 1
8782
+ 1
8783
+ 1
8784
+ 1
8785
+ 1
8786
+ 2
8787
+ 1
8788
+ 1
8789
+ 1
8790
+ 1
8791
+ 1
8792
+ 1
8793
+ 1
8794
+ 1
8795
+ 1
8796
+ 1
8797
+ 1
8798
+ 1
8799
+ 1
8800
+ 1
8801
+ 1
8802
+ 1
8803
+ 2
8804
+ 1
8805
+ 1
8806
+ 1
8807
+ 1
8808
+ 1
8809
+ 1
8810
+ 1
8811
+ 1
8812
+ 1
8813
+ 1
8814
+ 2
8815
+ 1
8816
+ 1
8817
+ 1
8818
+ 1
8819
+ 1
8820
+ 1
8821
+ 1
8822
+ 1
8823
+ 1
8824
+ 1
8825
+ 1
8826
+ 1
8827
+ 2
8828
+ 1
8829
+ 1
8830
+ 1
8831
+ 1
8832
+ 1
8833
+ 1
8834
+ 1
8835
+ 1
8836
+ 1
8837
+ 1
8838
+ 1
8839
+ 1
8840
+ 1
8841
+ 1
8842
+ 1
8843
+ 1
8844
+ 1
8845
+ 1
8846
+ 1
8847
+ 1
8848
+ 1
8849
+ 1
8850
+ 1
8851
+ 1
8852
+ 1
8853
+ 1
8854
+ 1
8855
+ 2
8856
+ 1
8857
+ 2
8858
+ 1
8859
+ 1
8860
+ 1
8861
+ 1
8862
+ 1
8863
+ 1
8864
+ 1
8865
+ 1
8866
+ 1
8867
+ 2
8868
+ 1
8869
+ 1
8870
+ 1
8871
+ 1
8872
+ 1
8873
+ 1
8874
+ 1
8875
+ 1
8876
+ 1
8877
+ 1
8878
+ 1
8879
+ 1
8880
+ 1
8881
+ 1
8882
+ 1
8883
+ 1
8884
+ 1
8885
+ 1
8886
+ 1
8887
+ 1
8888
+ 1
8889
+ 1
8890
+ 1
8891
+ 1
8892
+ 1
8893
+ 1
8894
+ 1
8895
+ 1
8896
+ 1
8897
+ 1
8898
+ 1
8899
+ 1
8900
+ 1
8901
+ 1
8902
+ 1
8903
+ 1
8904
+ 1
8905
+ 1
8906
+ 1
8907
+ 1
8908
+ 1
8909
+ 1
8910
+ 1
8911
+ 1
8912
+ 1
8913
+ 1
8914
+ 1
8915
+ 1
8916
+ 1
8917
+ 1
8918
+ 1
8919
+ 2
8920
+ 1
8921
+ 1
8922
+ 1
8923
+ 2
8924
+ 2
8925
+ 1
8926
+ 1
8927
+ 1
8928
+ 1
8929
+ 1
8930
+ 1
8931
+ 1
8932
+ 1
8933
+ 2
8934
+ 1
8935
+ 1
8936
+ 1
8937
+ 1
8938
+ 1
8939
+ 1
8940
+ 1
8941
+ 1
8942
+ 1
8943
+ 1
8944
+ 1
8945
+ 1
8946
+ 1
8947
+ 1
8948
+ 1
8949
+ 1
8950
+ 1
8951
+ 1
8952
+ 1
8953
+ 1
8954
+ 1
8955
+ 1
8956
+ 1
8957
+ 1
8958
+ 1
8959
+ 2
8960
+ 1
8961
+ 1
8962
+ 1
8963
+ 1
8964
+ 1
8965
+ 1
8966
+ 1
8967
+ 1
8968
+ 2
8969
+ 1
8970
+ 1
8971
+ 1
8972
+ 1
8973
+ 1
8974
+ 1
8975
+ 1
8976
+ 1
8977
+ 1
8978
+ 1
8979
+ 1
8980
+ 1
8981
+ 1
8982
+ 1
8983
+ 1
8984
+ 1
8985
+ 1
8986
+ 1
8987
+ 1
8988
+ 1
8989
+ 1
8990
+ 1
8991
+ 1
8992
+ 1
8993
+ 1
8994
+ 1
8995
+ 1
8996
+ 1
8997
+ 1
8998
+ 1
8999
+ 1
9000
+ 1
9001
+ 1
9002
+ 1
9003
+ 1
9004
+ 2
9005
+ 1
9006
+ 1
9007
+ 1
9008
+ 1
9009
+ 1
9010
+ 1
9011
+ 1
9012
+ 1
9013
+ 1
9014
+ 1
9015
+ 1
9016
+ 1
9017
+ 1
9018
+ 1
9019
+ 2
9020
+ 1
9021
+ 1
9022
+ 1
9023
+ 1
9024
+ 1
9025
+ 1
9026
+ 1
9027
+ 1
9028
+ 1
9029
+ 1
9030
+ 1
9031
+ 1
9032
+ 1
9033
+ 1
9034
+ 1
9035
+ 1
9036
+ 1
9037
+ 1
9038
+ 1
9039
+ 1
9040
+ 1
9041
+ 1
9042
+ 1
9043
+ 1
9044
+ 1
9045
+ 1
9046
+ 1
9047
+ 1
9048
+ 1
9049
+ 1
9050
+ 1
9051
+ 1
9052
+ 1
9053
+ 1
9054
+ 1
9055
+ 1
9056
+ 1
9057
+ 1
9058
+ 1
9059
+ 1
9060
+ 1
9061
+ 1
9062
+ 1
9063
+ 1
9064
+ 1
9065
+ 1
9066
+ 1
9067
+ 1
9068
+ 1
9069
+ 1
9070
+ 1
9071
+ 1
9072
+ 1
9073
+ 1
9074
+ 1
9075
+ 1
9076
+ 1
9077
+ 1
9078
+ 1
9079
+ 2
9080
+ 1
9081
+ 1
9082
+ 1
9083
+ 1
9084
+ 1
9085
+ 1
9086
+ 1
9087
+ 1
9088
+ 1
9089
+ 1
9090
+ 1
9091
+ 1
9092
+ 1
9093
+ 1
9094
+ 1
9095
+ 1
9096
+ 2
9097
+ 1
9098
+ 1
9099
+ 1
9100
+ 1
9101
+ 1
9102
+ 1
9103
+ 1
9104
+ 1
9105
+ 2
9106
+ 1
9107
+ 1
9108
+ 1
9109
+ 1
9110
+ 1
9111
+ 1
9112
+ 1
9113
+ 2
9114
+ 1
9115
+ 1
9116
+ 1
9117
+ 1
9118
+ 1
9119
+ 1
9120
+ 1
9121
+ 1
9122
+ 1
9123
+ 1
9124
+ 1
9125
+ 1
9126
+ 1
9127
+ 1
9128
+ 1
9129
+ 1
9130
+ 2
9131
+ 1
9132
+ 1
9133
+ 1
9134
+ 1
9135
+ 1
9136
+ 1
9137
+ 1
9138
+ 1
9139
+ 1
9140
+ 1
9141
+ 1
9142
+ 1
9143
+ 1
9144
+ 1
9145
+ 1
9146
+ 1
9147
+ 1
9148
+ 2
9149
+ 1
9150
+ 1
9151
+ 1
9152
+ 1
9153
+ 1
9154
+ 1
9155
+ 1
9156
+ 1
9157
+ 1
9158
+ 1
9159
+ 1
9160
+ 1
9161
+ 1
9162
+ 1
9163
+ 1
9164
+ 1
9165
+ 1
9166
+ 1
9167
+ 1
9168
+ 1
9169
+ 1
9170
+ 1
9171
+ 2
9172
+ 2
9173
+ 1
9174
+ 1
9175
+ 1
9176
+ 1
9177
+ 1
9178
+ 1
9179
+ 1
9180
+ 1
9181
+ 1
9182
+ 1
9183
+ 1
9184
+ 1
9185
+ 1
9186
+ 1
9187
+ 1
9188
+ 1
9189
+ 1
9190
+ 1
9191
+ 1
9192
+ 1
9193
+ 1
9194
+ 1
9195
+ 1
9196
+ 1
9197
+ 1
9198
+ 2
9199
+ 1
9200
+ 1
9201
+ 2
9202
+ 1
9203
+ 1
9204
+ 1
9205
+ 1
9206
+ 1
9207
+ 1
9208
+ 1
9209
+ 1
9210
+ 1
9211
+ 1
9212
+ 1
9213
+ 1
9214
+ 1
9215
+ 1
9216
+ 1
9217
+ 1
9218
+ 1
9219
+ 1
9220
+ 1
9221
+ 1
9222
+ 1
9223
+ 1
9224
+ 1
9225
+ 1
9226
+ 1
9227
+ 1
9228
+ 1
9229
+ 1
9230
+ 1
9231
+ 1
9232
+ 1
9233
+ 1
9234
+ 1
9235
+ 1
9236
+ 1
9237
+ 1
9238
+ 1
9239
+ 1
9240
+ 1
9241
+ 1
9242
+ 1
9243
+ 1
9244
+ 1
9245
+ 1
9246
+ 1
9247
+ 1
9248
+ 1
9249
+ 1
9250
+ 2
9251
+ 1
9252
+ 1
9253
+ 1
9254
+ 2
9255
+ 1
9256
+ 1
9257
+ 1
9258
+ 1
9259
+ 1
9260
+ 1
9261
+ 1
9262
+ 1
9263
+ 1
9264
+ 1
9265
+ 1
9266
+ 1
9267
+ 1
9268
+ 1
9269
+ 1
9270
+ 1
9271
+ 2
9272
+ 1
9273
+ 1
9274
+ 1
9275
+ 1
9276
+ 1
9277
+ 1
9278
+ 1
9279
+ 1
9280
+ 1
9281
+ 1
9282
+ 1
9283
+ 2
9284
+ 1
9285
+ 1
9286
+ 1
9287
+ 1
9288
+ 1
9289
+ 2
9290
+ 1
9291
+ 1
9292
+ 1
9293
+ 1
9294
+ 1
9295
+ 1
9296
+ 2
9297
+ 1
9298
+ 1
9299
+ 1
9300
+ 1
9301
+ 1
9302
+ 1
9303
+ 1
9304
+ 1
9305
+ 1
9306
+ 1
9307
+ 1
9308
+ 1
9309
+ 1
9310
+ 1
9311
+ 1
9312
+ 1
9313
+ 1
9314
+ 1
9315
+ 1
9316
+ 1
9317
+ 1
9318
+ 1
9319
+ 1
9320
+ 1
9321
+ 1
9322
+ 2
9323
+ 1
9324
+ 2
9325
+ 2
9326
+ 1
9327
+ 1
9328
+ 1
9329
+ 1
9330
+ 1
9331
+ 1
9332
+ 1
9333
+ 1
9334
+ 1
9335
+ 1
9336
+ 1
9337
+ 1
9338
+ 1
9339
+ 1
9340
+ 1
9341
+ 1
9342
+ 1
9343
+ 1
9344
+ 1
9345
+ 1
9346
+ 1
9347
+ 1
9348
+ 2
9349
+ 1
9350
+ 1
9351
+ 1
9352
+ 1
9353
+ 1
9354
+ 1
9355
+ 1
9356
+ 1
9357
+ 1
9358
+ 1
9359
+ 1
9360
+ 1
9361
+ 1
9362
+ 1
9363
+ 1
9364
+ 1
9365
+ 1
9366
+ 1
9367
+ 1
9368
+ 1
9369
+ 1
9370
+ 1
9371
+ 1
9372
+ 1
9373
+ 1
9374
+ 1
9375
+ 1
9376
+ 1
9377
+ 1
9378
+ 1
9379
+ 1
9380
+ 1
9381
+ 1
9382
+ 1
9383
+ 1
9384
+ 1
9385
+ 1
9386
+ 1
9387
+ 1
9388
+ 1
9389
+ 1
9390
+ 1
9391
+ 1
9392
+ 1
9393
+ 1
9394
+ 1
9395
+ 1
9396
+ 1
9397
+ 1
9398
+ 1
9399
+ 1
9400
+ 2
9401
+ 1
9402
+ 1
9403
+ 1
9404
+ 2
9405
+ 1
9406
+ 1
9407
+ 1
9408
+ 1
9409
+ 1
9410
+ 1
9411
+ 1
9412
+ 1
9413
+ 1
9414
+ 1
9415
+ 1
9416
+ 1
9417
+ 1
9418
+ 2
9419
+ 1
9420
+ 1
9421
+ 1
9422
+ 1
9423
+ 1
9424
+ 1
9425
+ 1
9426
+ 1
9427
+ 2
9428
+ 2
9429
+ 1
9430
+ 1
9431
+ 1
9432
+ 1
9433
+ 1
9434
+ 1
9435
+ 1
9436
+ 1
9437
+ 1
9438
+ 1
9439
+ 1
9440
+ 2
9441
+ 1
9442
+ 1
9443
+ 1
9444
+ 1
9445
+ 1
9446
+ 1
9447
+ 1
9448
+ 1
9449
+ 1
9450
+ 1
9451
+ 1
9452
+ 1
9453
+ 1
9454
+ 1
9455
+ 1
9456
+ 1
9457
+ 1
9458
+ 1
9459
+ 1
9460
+ 1
9461
+ 1
9462
+ 1
9463
+ 1
9464
+ 1
9465
+ 1
9466
+ 1
9467
+ 1
9468
+ 1
9469
+ 2
9470
+ 1
9471
+ 1
9472
+ 1
9473
+ 1
9474
+ 1
9475
+ 1
9476
+ 1
9477
+ 1
9478
+ 1
9479
+ 1
9480
+ 1
9481
+ 1
9482
+ 1
9483
+ 1
9484
+ 1
9485
+ 1
9486
+ 1
9487
+ 2
9488
+ 1
9489
+ 1
9490
+ 1
9491
+ 1
9492
+ 1
9493
+ 1
9494
+ 1
9495
+ 1
9496
+ 1
9497
+ 2
9498
+ 1
9499
+ 1
9500
+ 1
9501
+ 1
9502
+ 1
9503
+ 1
9504
+ 1
9505
+ 1
9506
+ 1
9507
+ 1
9508
+ 1
9509
+ 1
9510
+ 1
9511
+ 1
9512
+ 2
9513
+ 1
9514
+ 1
9515
+ 1
9516
+ 1
9517
+ 2
9518
+ 1
9519
+ 2
9520
+ 1
9521
+ 1
9522
+ 1
9523
+ 1
9524
+ 1
9525
+ 1
9526
+ 1
9527
+ 1
9528
+ 1
9529
+ 1
9530
+ 2
9531
+ 1
9532
+ 1
9533
+ 1
9534
+ 1
9535
+ 1
9536
+ 1
9537
+ 1
9538
+ 1
9539
+ 1
9540
+ 1
9541
+ 1
9542
+ 1
9543
+ 1
9544
+ 1
9545
+ 1
9546
+ 1
9547
+ 1
9548
+ 1
9549
+ 1
9550
+ 1
9551
+ 1
9552
+ 1
9553
+ 1
9554
+ 1
9555
+ 1
9556
+ 1
9557
+ 1
9558
+ 2
9559
+ 1
9560
+ 1
9561
+ 1
9562
+ 1
9563
+ 1
9564
+ 2
9565
+ 2
9566
+ 1
9567
+ 1
9568
+ 1
9569
+ 1
9570
+ 1
9571
+ 1
9572
+ 1
9573
+ 1
9574
+ 1
9575
+ 1
9576
+ 1
9577
+ 1
9578
+ 1
9579
+ 1
9580
+ 1
9581
+ 1
9582
+ 1
9583
+ 1
9584
+ 1
9585
+ 2
9586
+ 1
9587
+ 2
9588
+ 1
9589
+ 1
9590
+ 1
9591
+ 1
9592
+ 2
9593
+ 1
9594
+ 1
9595
+ 1
9596
+ 1
9597
+ 1
9598
+ 1
9599
+ 1
9600
+ 2
9601
+ 2
9602
+ 2
9603
+ 1
9604
+ 1
9605
+ 1
9606
+ 1
9607
+ 1
9608
+ 1
9609
+ 1
9610
+ 1
9611
+ 1
9612
+ 1
9613
+ 1
9614
+ 1
9615
+ 1
9616
+ 1
9617
+ 1
9618
+ 1
9619
+ 1
9620
+ 1
9621
+ 1
9622
+ 1
9623
+ 1
9624
+ 1
9625
+ 1
9626
+ 1
9627
+ 1
9628
+ 1
9629
+ 1
9630
+ 1
9631
+ 1
9632
+ 1
9633
+ 2
9634
+ 1
9635
+ 1
9636
+ 1
9637
+ 1
9638
+ 1
9639
+ 1
9640
+ 1
9641
+ 1
9642
+ 1
9643
+ 1
9644
+ 1
9645
+ 1
9646
+ 1
9647
+ 1
9648
+ 1
9649
+ 1
9650
+ 1
9651
+ 1
9652
+ 1
9653
+ 1
9654
+ 1
9655
+ 2
9656
+ 1
9657
+ 1
9658
+ 1
9659
+ 1
9660
+ 1
9661
+ 1
9662
+ 1
9663
+ 1
9664
+ 1
9665
+ 1
9666
+ 1
9667
+ 1
9668
+ 2
9669
+ 1
9670
+ 1
9671
+ 1
9672
+ 1
9673
+ 1
9674
+ 1
9675
+ 1
9676
+ 1
9677
+ 1
9678
+ 1
9679
+ 2
9680
+ 1
9681
+ 1
9682
+ 1
9683
+ 1
9684
+ 1
9685
+ 1
9686
+ 1
9687
+ 1
9688
+ 1
9689
+ 1
9690
+ 1
9691
+ 1
9692
+ 1
9693
+ 1
9694
+ 1
9695
+ 1
9696
+ 1
9697
+ 2
9698
+ 1
9699
+ 1
9700
+ 1
9701
+ 1
9702
+ 1
9703
+ 1
9704
+ 1
9705
+ 1
9706
+ 1
9707
+ 1
9708
+ 1
9709
+ 1
9710
+ 1
9711
+ 1
9712
+ 1
9713
+ 1
9714
+ 1
9715
+ 1
9716
+ 1
9717
+ 2
9718
+ 1
9719
+ 1
9720
+ 1
9721
+ 1
9722
+ 1
9723
+ 1
9724
+ 2
9725
+ 1
9726
+ 1
9727
+ 1
9728
+ 1
9729
+ 1
9730
+ 1
9731
+ 1
9732
+ 1
9733
+ 1
9734
+ 1
9735
+ 1
9736
+ 1
9737
+ 1
9738
+ 1
9739
+ 1
9740
+ 1
9741
+ 2
9742
+ 1
9743
+ 1
9744
+ 1
9745
+ 1
9746
+ 1
9747
+ 1
9748
+ 1
9749
+ 1
9750
+ 1
9751
+ 1
9752
+ 1
9753
+ 1
9754
+ 1
9755
+ 1
9756
+ 1
9757
+ 1
9758
+ 1
9759
+ 1
9760
+ 1
9761
+ 1
9762
+ 1
9763
+ 1
9764
+ 1
9765
+ 1
9766
+ 1
9767
+ 1
9768
+ 1
9769
+ 1
9770
+ 1
9771
+ 1
9772
+ 1
9773
+ 1
9774
+ 1
9775
+ 1
9776
+ 1
9777
+ 1
9778
+ 1
9779
+ 1
9780
+ 1
9781
+ 1
9782
+ 1
9783
+ 1
9784
+ 1
9785
+ 1
9786
+ 1
9787
+ 1
9788
+ 1
9789
+ 1
9790
+ 1
9791
+ 1
9792
+ 1
9793
+ 1
9794
+ 1
9795
+ 1
9796
+ 1
9797
+ 1
9798
+ 1
9799
+ 1
9800
+ 1
9801
+ 1
9802
+ 1
9803
+ 1
9804
+ 1
9805
+ 1
9806
+ 1
9807
+ 1
9808
+ 1
9809
+ 1
9810
+ 1
9811
+ 1
9812
+ 1
9813
+ 1
9814
+ 1
9815
+ 1
9816
+ 1
9817
+ 1
9818
+ 1
9819
+ 1
9820
+ 1
9821
+ 2
9822
+ 1
9823
+ 1
9824
+ 1
9825
+ 2
9826
+ 2
9827
+ 1
9828
+ 1
9829
+ 2
9830
+ 1
9831
+ 1
9832
+ 1
9833
+ 1
9834
+ 1
9835
+ 1
9836
+ 1
9837
+ 1
9838
+ 1
9839
+ 1
9840
+ 1
9841
+ 1
9842
+ 1
9843
+ 1
9844
+ 1
9845
+ 1
9846
+ 1
9847
+ 1
9848
+ 1
9849
+ 1
9850
+ 1
9851
+ 1
9852
+ 1
9853
+ 1
9854
+ 1
9855
+ 1
9856
+ 1
9857
+ 1
9858
+ 1
9859
+ 1
9860
+ 1
9861
+ 1
9862
+ 1
9863
+ 1
9864
+ 2
9865
+ 1
9866
+ 1
9867
+ 2
9868
+ 1
9869
+ 2
9870
+ 1
9871
+ 1
9872
+ 1
9873
+ 1
9874
+ 1
9875
+ 1
9876
+ 1
9877
+ 1
9878
+ 1
9879
+ 1
9880
+ 1
9881
+ 1
9882
+ 1
9883
+ 1
9884
+ 1
9885
+ 1
9886
+ 1
9887
+ 1
9888
+ 1
9889
+ 1
9890
+ 2
9891
+ 1
9892
+ 1
9893
+ 1
9894
+ 1
9895
+ 1
9896
+ 1
9897
+ 1
9898
+ 1
9899
+ 1
9900
+ 1
9901
+ 1
9902
+ 1
9903
+ 1
9904
+ 1
9905
+ 1
9906
+ 1
9907
+ 1
9908
+ 1
9909
+ 1
9910
+ 1
9911
+ 1
9912
+ 1
9913
+ 1
9914
+ 1
9915
+ 1
9916
+ 1
9917
+ 1
9918
+ 1
9919
+ 1
9920
+ 1
9921
+ 1
9922
+ 1
9923
+ 1
9924
+ 1
9925
+ 1
9926
+ 1
9927
+ 1
9928
+ 1
9929
+ 1
9930
+ 2
9931
+ 1
9932
+ 1
9933
+ 1
9934
+ 1
9935
+ 1
9936
+ 1
9937
+ 1
9938
+ 1
9939
+ 2
9940
+ 1
9941
+ 1
9942
+ 1
9943
+ 1
9944
+ 1
9945
+ 1
9946
+ 1
9947
+ 1
9948
+ 1
9949
+ 1
9950
+ 1
9951
+ 1
9952
+ 1
9953
+ 1
9954
+ 2
9955
+ 1
9956
+ 1
9957
+ 1
9958
+ 1
9959
+ 1
9960
+ 1
9961
+ 1
9962
+ 1
9963
+ 1
9964
+ 1
9965
+ 1
9966
+ 1
9967
+ 1
9968
+ 2
9969
+ 1
9970
+ 1
9971
+ 1
9972
+ 1
9973
+ 1
9974
+ 1
9975
+ 1
9976
+ 1
9977
+ 1
9978
+ 1
9979
+ 1
9980
+ 1
9981
+ 1
9982
+ 1
9983
+ 1
9984
+ 1
9985
+ 1
9986
+ 1
9987
+ 1
9988
+ 1
9989
+ 1
9990
+ 1
9991
+ 1
9992
+ 1
9993
+ 1
9994
+ 1
9995
+ 1
9996
+ 1
9997
+ 1
9998
+ 1
9999
+ 1
10000
+ 1
10001
+ 1
10002
+ 1
10003
+ 1
10004
+ 1
10005
+ 1
10006
+ 1
10007
+ 1
10008
+ 1
10009
+ 1
10010
+ 1
10011
+ 1
10012
+ 1
10013
+ 1
10014
+ 1
10015
+ 1
10016
+ 1
10017
+ 1
10018
+ 1
10019
+ 1
10020
+ 2
10021
+ 2
10022
+ 1
10023
+ 1
10024
+ 1
10025
+ 1
10026
+ 2
10027
+ 1
10028
+ 1
10029
+ 1
10030
+ 1
10031
+ 1
10032
+ 1
10033
+ 1
10034
+ 1
10035
+ 1
10036
+ 1
10037
+ 1
10038
+ 1
10039
+ 1
10040
+ 1
10041
+ 1
10042
+ 1
10043
+ 1
10044
+ 1
10045
+ 1
10046
+ 1
10047
+ 1
10048
+ 1
10049
+ 1
10050
+ 1
10051
+ 1
10052
+ 1
10053
+ 1
10054
+ 1
10055
+ 1
10056
+ 1
10057
+ 1
10058
+ 1
10059
+ 1
10060
+ 2
10061
+ 1
10062
+ 1
10063
+ 1
10064
+ 1
10065
+ 1
10066
+ 1
10067
+ 1
10068
+ 1
10069
+ 1
10070
+ 1
10071
+ 1
10072
+ 1
10073
+ 1
10074
+ 1
10075
+ 1
10076
+ 1
10077
+ 1
10078
+ 1
10079
+ 1
10080
+ 1
10081
+ 1
10082
+ 1
10083
+ 1
10084
+ 1
10085
+ 1
10086
+ 1
10087
+ 1
10088
+ 1
10089
+ 1
10090
+ 1
10091
+ 1
10092
+ 1
10093
+ 1
10094
+ 1
10095
+ 1
10096
+ 1
10097
+ 1
10098
+ 1
10099
+ 1
10100
+ 1
10101
+ 1
10102
+ 1
10103
+ 1
10104
+ 1
10105
+ 1
10106
+ 1
10107
+ 1
10108
+ 1
10109
+ 1
10110
+ 1
10111
+ 1
10112
+ 2
10113
+ 1
10114
+ 1
10115
+ 1
10116
+ 1
10117
+ 1
10118
+ 1
10119
+ 2
10120
+ 1
10121
+ 1
10122
+ 1
10123
+ 1
10124
+ 1
10125
+ 1
10126
+ 2
10127
+ 1
10128
+ 1
10129
+ 1
10130
+ 1
10131
+ 1
10132
+ 1
10133
+ 1
10134
+ 1
10135
+ 1
10136
+ 1
10137
+ 1
10138
+ 1
10139
+ 1
10140
+ 1
10141
+ 1
10142
+ 1
10143
+ 1
10144
+ 1
10145
+ 1
10146
+ 1
10147
+ 1
10148
+ 1
10149
+ 1
10150
+ 1
10151
+ 1
10152
+ 1
10153
+ 1
10154
+ 1
10155
+ 1
10156
+ 1
10157
+ 1
10158
+ 1
10159
+ 1
10160
+ 1
10161
+ 1
10162
+ 1
10163
+ 1
10164
+ 1
10165
+ 1
10166
+ 1
10167
+ 1
10168
+ 1
10169
+ 1
10170
+ 1
10171
+ 1
10172
+ 1
10173
+ 1
10174
+ 1
10175
+ 2
10176
+ 1
10177
+ 1
10178
+ 1
10179
+ 1
10180
+ 1
10181
+ 1
10182
+ 1
10183
+ 1
10184
+ 1
10185
+ 1
10186
+ 1
10187
+ 1
10188
+ 1
10189
+ 1
10190
+ 1
10191
+ 1
10192
+ 1
10193
+ 1
10194
+ 1
10195
+ 2
10196
+ 1
10197
+ 1
10198
+ 1
10199
+ 1
10200
+ 1
10201
+ 1
10202
+ 2
10203
+ 2
10204
+ 1
10205
+ 1
10206
+ 1
10207
+ 1
10208
+ 1
10209
+ 1
10210
+ 1
10211
+ 1
10212
+ 1
10213
+ 1
10214
+ 1
10215
+ 1
10216
+ 1
10217
+ 1
10218
+ 1
10219
+ 1
10220
+ 1
10221
+ 2
10222
+ 2
10223
+ 1
10224
+ 1
10225
+ 1
10226
+ 1
10227
+ 2
10228
+ 1
10229
+ 1
10230
+ 1
10231
+ 1
10232
+ 1
10233
+ 1
10234
+ 1
10235
+ 1
10236
+ 1
10237
+ 1
10238
+ 1
10239
+ 1
10240
+ 1
10241
+ 1
10242
+ 1
10243
+ 2
10244
+ 1
10245
+ 1
10246
+ 1
10247
+ 1
10248
+ 1
10249
+ 1
10250
+ 1
10251
+ 2
10252
+ 1
10253
+ 1
10254
+ 1
10255
+ 1
10256
+ 1
10257
+ 1
10258
+ 1
10259
+ 1
10260
+ 1
10261
+ 1
10262
+ 1
10263
+ 1
10264
+ 1
10265
+ 1
10266
+ 1
10267
+ 1
10268
+ 1
10269
+ 1
10270
+ 1
10271
+ 1
10272
+ 1
10273
+ 1
10274
+ 1
10275
+ 1
10276
+ 1
10277
+ 2
10278
+ 1
10279
+ 1
10280
+ 1
10281
+ 1
10282
+ 1
10283
+ 1
10284
+ 1
10285
+ 1
10286
+ 1
10287
+ 1
10288
+ 1
10289
+ 1
10290
+ 1
10291
+ 1
10292
+ 1
10293
+ 1
10294
+ 1
10295
+ 1
10296
+ 2
10297
+ 1
10298
+ 1
10299
+ 1
10300
+ 1
10301
+ 1
10302
+ 1
10303
+ 2
10304
+ 1
10305
+ 1
10306
+ 2
10307
+ 1
10308
+ 1
10309
+ 1
10310
+ 1
10311
+ 1
10312
+ 1
10313
+ 1
10314
+ 1
10315
+ 1
10316
+ 1
10317
+ 1
10318
+ 1
10319
+ 1
10320
+ 1
10321
+ 1
10322
+ 1
10323
+ 1
10324
+ 1
10325
+ 1
10326
+ 1
10327
+ 1
10328
+ 2
10329
+ 1
10330
+ 1
10331
+ 1
10332
+ 1
10333
+ 1
10334
+ 1
10335
+ 1
10336
+ 1
10337
+ 1
10338
+ 1
10339
+ 1
10340
+ 1
10341
+ 1
10342
+ 1
10343
+ 1
10344
+ 1
10345
+ 1
10346
+ 1
10347
+ 1
10348
+ 1
10349
+ 1
10350
+ 1
10351
+ 1
10352
+ 2
10353
+ 1
10354
+ 1
10355
+ 1
10356
+ 1
10357
+ 1
10358
+ 1
10359
+ 1
10360
+ 1
10361
+ 2
10362
+ 1
10363
+ 1
10364
+ 1
10365
+ 1
10366
+ 1
10367
+ 1
10368
+ 1
10369
+ 1
10370
+ 2
10371
+ 1
10372
+ 1
10373
+ 1
10374
+ 1
10375
+ 1
10376
+ 1
10377
+ 1
10378
+ 1
10379
+ 1
10380
+ 1
10381
+ 1
10382
+ 1
10383
+ 1
10384
+ 1
10385
+ 1
10386
+ 1
10387
+ 1
10388
+ 1
10389
+ 1
10390
+ 1
10391
+ 1
10392
+ 1
10393
+ 1
10394
+ 2
10395
+ 1
10396
+ 1
10397
+ 1
10398
+ 1
10399
+ 1
10400
+ 1
10401
+ 1
10402
+ 1
10403
+ 1
10404
+ 1
10405
+ 1
10406
+ 1
10407
+ 1
10408
+ 2
10409
+ 1
10410
+ 1
10411
+ 1
10412
+ 1
10413
+ 1
10414
+ 1
10415
+ 1
10416
+ 1
10417
+ 1
10418
+ 1
10419
+ 1
10420
+ 1
10421
+ 1
10422
+ 1
10423
+ 1
10424
+ 1
10425
+ 1
10426
+ 1
10427
+ 1
10428
+ 1
10429
+ 1
10430
+ 1
10431
+ 1
10432
+ 1
10433
+ 1
10434
+ 1
10435
+ 1
10436
+ 1
10437
+ 1
10438
+ 1
10439
+ 2
10440
+ 1
10441
+ 1
10442
+ 1
10443
+ 1
10444
+ 1
10445
+ 1
10446
+ 1
10447
+ 1
10448
+ 1
10449
+ 1
10450
+ 1
10451
+ 1
10452
+ 1
10453
+ 1
10454
+ 1
10455
+ 1
10456
+ 2
10457
+ 1
10458
+ 1
10459
+ 1
10460
+ 1
10461
+ 1
10462
+ 1
10463
+ 1
10464
+ 2
10465
+ 1
10466
+ 2
10467
+ 1
10468
+ 1
10469
+ 1
10470
+ 1
10471
+ 2
10472
+ 1
10473
+ 1
10474
+ 1
10475
+ 1
10476
+ 1
10477
+ 1
10478
+ 1
10479
+ 2
10480
+ 1
10481
+ 1
10482
+ 1
10483
+ 1
10484
+ 1
10485
+ 1
10486
+ 1
10487
+ 1
10488
+ 1
10489
+ 1
10490
+ 1
10491
+ 1
10492
+ 1
10493
+ 1
10494
+ 1
10495
+ 1
10496
+ 1
10497
+ 1
10498
+ 1
10499
+ 1
10500
+ 1
10501
+ 1
10502
+ 1
10503
+ 1
10504
+ 2
10505
+ 1
10506
+ 1
10507
+ 1
10508
+ 1
10509
+ 2
10510
+ 2
10511
+ 1
10512
+ 1
10513
+ 1
10514
+ 1
10515
+ 1
10516
+ 1
10517
+ 1
10518
+ 1
10519
+ 1
10520
+ 1
10521
+ 1
10522
+ 1
10523
+ 1
10524
+ 1
10525
+ 1
10526
+ 1
10527
+ 1
10528
+ 1
10529
+ 1
10530
+ 1
10531
+ 1
10532
+ 1
10533
+ 1
10534
+ 1
10535
+ 1
10536
+ 1
10537
+ 1
10538
+ 1
10539
+ 1
10540
+ 1
10541
+ 1
10542
+ 1
10543
+ 1
10544
+ 1
10545
+ 1
10546
+ 1
10547
+ 1
10548
+ 1
10549
+ 1
10550
+ 1
10551
+ 1
10552
+ 1
10553
+ 1
10554
+ 1
10555
+ 1
10556
+ 1
10557
+ 1
10558
+ 1
10559
+ 1
10560
+ 1
10561
+ 1
10562
+ 1
10563
+ 1
10564
+ 1
10565
+ 1
10566
+ 1
10567
+ 1
10568
+ 1
10569
+ 1
10570
+ 1
10571
+ 1
10572
+ 1
10573
+ 1
10574
+ 1
10575
+ 1
10576
+ 1
10577
+ 1
10578
+ 1
10579
+ 1
10580
+ 1
10581
+ 1
10582
+ 1
10583
+ 1
10584
+ 1
10585
+ 1
10586
+ 1
10587
+ 1
10588
+ 1
10589
+ 1
10590
+ 1
10591
+ 1
10592
+ 1
10593
+ 1
10594
+ 1
10595
+ 1
10596
+ 1
10597
+ 1
10598
+ 1
10599
+ 1
10600
+ 1
10601
+ 1
10602
+ 1
10603
+ 1
10604
+ 1
10605
+ 1
10606
+ 1
10607
+ 1
10608
+ 1
10609
+ 1
10610
+ 1
10611
+ 1
10612
+ 1
10613
+ 1
10614
+ 1
10615
+ 1
10616
+ 1
10617
+ 1
10618
+ 1
10619
+ 1
10620
+ 1
10621
+ 1
10622
+ 1
10623
+ 1
10624
+ 1
10625
+ 1
10626
+ 1
10627
+ 1
10628
+ 1
10629
+ 1
10630
+ 1
10631
+ 1
10632
+ 1
10633
+ 1
10634
+ 1
10635
+ 1
10636
+ 1
10637
+ 1
10638
+ 1
10639
+ 1
10640
+ 1
10641
+ 1
10642
+ 1
10643
+ 1
10644
+ 1
10645
+ 1
10646
+ 1
10647
+ 1
10648
+ 1
10649
+ 1
10650
+ 1
10651
+ 1
10652
+ 1
10653
+ 1
10654
+ 1
10655
+ 1
10656
+ 1
10657
+ 1
10658
+ 1
10659
+ 1
10660
+ 1
10661
+ 1
10662
+ 1
10663
+ 1
10664
+ 1
10665
+ 1
10666
+ 1
10667
+ 1
10668
+ 1
10669
+ 1
10670
+ 1
10671
+ 1
10672
+ 1
10673
+ 1
10674
+ 1
10675
+ 1
10676
+ 1
10677
+ 1
10678
+ 1
10679
+ 1
10680
+ 1
10681
+ 1
10682
+ 1
10683
+ 1
10684
+ 1
10685
+ 1
10686
+ 1
10687
+ 1
10688
+ 1
10689
+ 1
10690
+ 1
10691
+ 1
10692
+ 1
10693
+ 1
10694
+ 1
10695
+ 1
10696
+ 1
10697
+ 1
10698
+ 1
10699
+ 1
10700
+ 1
10701
+ 1
10702
+ 1
10703
+ 1
10704
+ 1
10705
+ 1
10706
+ 1
10707
+ 1
10708
+ 1
10709
+ 1
10710
+ 1
10711
+ 1
10712
+ 1
10713
+ 1
10714
+ 1
10715
+ 1
10716
+ 1
10717
+ 1
10718
+ 1
10719
+ 1
10720
+ 1
10721
+ 1
10722
+ 1
10723
+ 1
10724
+ 1
10725
+ 1
10726
+ 1
10727
+ 1
10728
+ 1
10729
+ 1
10730
+ 1
10731
+ 1
10732
+ 1
10733
+ 1
10734
+ 1
10735
+ 1
10736
+ 1
10737
+ 1
10738
+ 1
10739
+ 1
10740
+ 1
10741
+ 1
10742
+ 1
10743
+ 1
10744
+ 1
10745
+ 1
10746
+ 1
10747
+ 1
10748
+ 1
10749
+ 1
10750
+ 1
10751
+ 1
10752
+ 1
10753
+ 1
10754
+ 1
10755
+ 1
10756
+ 1
10757
+ 1
10758
+ 1
10759
+ 1
10760
+ 1
10761
+ 1
10762
+ 1
10763
+ 1
10764
+ 1
10765
+ 1
10766
+ 1
10767
+ 1
10768
+ 1
10769
+ 1
10770
+ 1
10771
+ 1
10772
+ 1
10773
+ 1
10774
+ 1
10775
+ 1
10776
+ 1
10777
+ 1
10778
+ 1
10779
+ 1
10780
+ 1
10781
+ 1
10782
+ 1
10783
+ 1
10784
+ 1
10785
+ 1
10786
+ 1
10787
+ 1
10788
+ 1
10789
+ 1
10790
+ 1
10791
+ 1
10792
+ 1
10793
+ 1
10794
+ 1
10795
+ 1
10796
+ 1
10797
+ 1
10798
+ 1
10799
+ 1
10800
+ 1
10801
+ 1
10802
+ 1
10803
+ 1
10804
+ 1
10805
+ 1
10806
+ 1
10807
+ 1
10808
+ 1
10809
+ 1
10810
+ 2
10811
+ 1
10812
+ 1
10813
+ 1
10814
+ 1
10815
+ 1
10816
+ 1
10817
+ 1
10818
+ 1
10819
+ 1
10820
+ 1
10821
+ 1
10822
+ 1
10823
+ 1
10824
+ 1
10825
+ 1
10826
+ 1
10827
+ 1
10828
+ 1
10829
+ 1
10830
+ 1
10831
+ 1
10832
+ 1
10833
+ 1
10834
+ 1
10835
+ 1
10836
+ 1
10837
+ 1
10838
+ 1
10839
+ 1
10840
+ 1
10841
+ 1
10842
+ 1
10843
+ 1
10844
+ 1
10845
+ 1
10846
+ 1
10847
+ 1
10848
+ 1
10849
+ 1
10850
+ 1
10851
+ 1
10852
+ 1
10853
+ 1
10854
+ 1
10855
+ 1
10856
+ 1
10857
+ 1
10858
+ 1
10859
+ 1
10860
+ 1
10861
+ 1
10862
+ 1
10863
+ 1
10864
+ 1
10865
+ 1
10866
+ 1
10867
+ 1
10868
+ 1
10869
+ 1
10870
+ 1
10871
+ 1
10872
+ 1
10873
+ 1
10874
+ 1
10875
+ 1
10876
+ 1
10877
+ 1
10878
+ 1
10879
+ 1
10880
+ 1
10881
+ 1
10882
+ 1
10883
+ 1
10884
+ 1
10885
+ 1
10886
+ 1
10887
+ 1
10888
+ 1
10889
+ 1
10890
+ 1
10891
+ 1
10892
+ 1
10893
+ 1
10894
+ 1
10895
+ 1
10896
+ 1
10897
+ 1
10898
+ 1
10899
+ 1
10900
+ 1
10901
+ 1
10902
+ 1
10903
+ 1
10904
+ 1
10905
+ 1
10906
+ 1
10907
+ 1
10908
+ 1
10909
+ 1
10910
+ 1
10911
+ 1
10912
+ 1
10913
+ 1
10914
+ 1
10915
+ 1
10916
+ 1
10917
+ 1
10918
+ 1
10919
+ 1
10920
+ 1
10921
+ 1
10922
+ 1
10923
+ 1
10924
+ 1
10925
+ 1
10926
+ 1
10927
+ 1
10928
+ 1
10929
+ 1
10930
+ 1
10931
+ 1
10932
+ 1
10933
+ 1
10934
+ 1
10935
+ 1
10936
+ 1
10937
+ 1
10938
+ 1
10939
+ 1
10940
+ 1
10941
+ 1
10942
+ 1
10943
+ 1
10944
+ 1
10945
+ 1
10946
+ 1
10947
+ 1
10948
+ 1
10949
+ 1
10950
+ 1
10951
+ 1
10952
+ 1
10953
+ 1
10954
+ 1
10955
+ 1
10956
+ 1
10957
+ 1
10958
+ 1
10959
+ 1
10960
+ 1
10961
+ 1
10962
+ 1
10963
+ 1
10964
+ 1
10965
+ 1
10966
+ 1
10967
+ 1
10968
+ 1
10969
+ 1
10970
+ 1
10971
+ 1
10972
+ 1
10973
+ 1
10974
+ 1
10975
+ 1
10976
+ 1
10977
+ 1
10978
+ 1
10979
+ 1
10980
+ 1
10981
+ 1
10982
+ 1
10983
+ 1
10984
+ 1
10985
+ 1
10986
+ 1
10987
+ 1
10988
+ 1
10989
+ 1
10990
+ 1
10991
+ 1
10992
+ 1
10993
+ 1
10994
+ 1
10995
+ 1
10996
+ 1
10997
+ 1
10998
+ 1
10999
+ 1
11000
+ 1
11001
+ 1
11002
+ 1
11003
+ 1
11004
+ 1
11005
+ 1
11006
+ 1
11007
+ 1
11008
+ 1
11009
+ 1
11010
+ 1
11011
+ 1
11012
+ 1
11013
+ 1
11014
+ 1
11015
+ 1
11016
+ 1
11017
+ 1
11018
+ 1
11019
+ 1
11020
+ 1
11021
+ 1
11022
+ 1
11023
+ 1
11024
+ 1
11025
+ 1
11026
+ 1
11027
+ 1
11028
+ 1
11029
+ 1
11030
+ 1
11031
+ 1
11032
+ 1
11033
+ 1
11034
+ 1
11035
+ 1
11036
+ 1
11037
+ 1
11038
+ 1
11039
+ 1
11040
+ 1
11041
+ 1
11042
+ 1
11043
+ 1
11044
+ 1
11045
+ 1
11046
+ 1
11047
+ 1
11048
+ 1
11049
+ 1
11050
+ 1
11051
+ 1
11052
+ 1
11053
+ 1
11054
+ 1
11055
+ 1
11056
+ 1
11057
+ 1
11058
+ 1
11059
+ 1
11060
+ 1
11061
+ 1
11062
+ 1
11063
+ 1
11064
+ 1
11065
+ 1
11066
+ 1
11067
+ 1
11068
+ 1
11069
+ 1
11070
+ 1
11071
+ 1
11072
+ 1
11073
+ 1
11074
+ 1
11075
+ 1
11076
+ 1
11077
+ 1
11078
+ 1
11079
+ 1
11080
+ 1
11081
+ 2
11082
+ 1
11083
+ 1
11084
+ 1
11085
+ 1
11086
+ 1
11087
+ 1
11088
+ 1
11089
+ 1
11090
+ 1
11091
+ 1
11092
+ 1
11093
+ 1
11094
+ 1
11095
+ 1
11096
+ 1
11097
+ 1
11098
+ 1
11099
+ 1
11100
+ 1
11101
+ 1
11102
+ 1
11103
+ 1
11104
+ 1
11105
+ 1
11106
+ 1
11107
+ 1
11108
+ 1
11109
+ 1
11110
+ 1
11111
+ 1
11112
+ 1
11113
+ 1
11114
+ 1
11115
+ 1
11116
+ 1
11117
+ 1
11118
+ 1
11119
+ 1
11120
+ 1
11121
+ 1
11122
+ 1
11123
+ 1
11124
+ 1
11125
+ 1
11126
+ 1
11127
+ 1
11128
+ 1
11129
+ 1
11130
+ 1
11131
+ 1
11132
+ 1
11133
+ 1
11134
+ 1
11135
+ 1
11136
+ 1
11137
+ 1
11138
+ 1
11139
+ 1
11140
+ 1
11141
+ 1
11142
+ 1
11143
+ 1
11144
+ 1
11145
+ 1
11146
+ 1
11147
+ 1
11148
+ 1
11149
+ 1
11150
+ 1
11151
+ 1
11152
+ 1
11153
+ 1
11154
+ 1
11155
+ 1
11156
+ 1
11157
+ 1
11158
+ 1
11159
+ 1
11160
+ 1
11161
+ 1
11162
+ 1
11163
+ 1
11164
+ 1
11165
+ 1
11166
+ 1
11167
+ 1
11168
+ 1
11169
+ 1
11170
+ 1
11171
+ 1
11172
+ 1
11173
+ 1
11174
+ 1
11175
+ 1
11176
+ 1
11177
+ 1
11178
+ 1
11179
+ 1
11180
+ 1
11181
+ 1
11182
+ 1
11183
+ 1
11184
+ 1
11185
+ 1
11186
+ 1
11187
+ 1
11188
+ 1
11189
+ 1
11190
+ 1
11191
+ 1
11192
+ 1
11193
+ 1
11194
+ 1
11195
+ 1
11196
+ 1
11197
+ 1
11198
+ 1
11199
+ 1
11200
+ 1
11201
+ 1
11202
+ 1
11203
+ 1
11204
+ 1
11205
+ 1
11206
+ 1
11207
+ 1
11208
+ 1
11209
+ 1
11210
+ 1
11211
+ 1
11212
+ 1
11213
+ 1
11214
+ 1
11215
+ 1
11216
+ 1
11217
+ 1
11218
+ 1
11219
+ 1
11220
+ 1
11221
+ 1
11222
+ 1
11223
+ 1
11224
+ 1
11225
+ 1
11226
+ 1
11227
+ 1
11228
+ 1
11229
+ 1
11230
+ 1
11231
+ 1
11232
+ 1
11233
+ 1
11234
+ 1
11235
+ 1
11236
+ 1
11237
+ 1
11238
+ 1
11239
+ 1
11240
+ 1
11241
+ 1
11242
+ 1
11243
+ 1
11244
+ 1
11245
+ 1
11246
+ 1
11247
+ 1
11248
+ 1
11249
+ 1
11250
+ 1
11251
+ 1
11252
+ 1
11253
+ 1
11254
+ 1
11255
+ 1
11256
+ 1
11257
+ 1
11258
+ 1
11259
+ 1
11260
+ 1
11261
+ 1
11262
+ 1
11263
+ 1
11264
+ 1
11265
+ 1
11266
+ 1
11267
+ 1
11268
+ 1
11269
+ 1
11270
+ 1
11271
+ 1
11272
+ 1
11273
+ 1
11274
+ 1
11275
+ 1
11276
+ 1
11277
+ 1
11278
+ 1
11279
+ 1
11280
+ 1
11281
+ 1
11282
+ 1
11283
+ 1
11284
+ 1
11285
+ 1
11286
+ 1
11287
+ 1
11288
+ 1
11289
+ 1
11290
+ 1
11291
+ 1
11292
+ 1
11293
+ 1
11294
+ 1
11295
+ 1
11296
+ 1
11297
+ 1
11298
+ 1
11299
+ 1
11300
+ 1
11301
+ 1
11302
+ 1
11303
+ 1
11304
+ 1
11305
+ 1
11306
+ 1
11307
+ 1
11308
+ 1
11309
+ 1
11310
+ 1
11311
+ 1
11312
+ 1
11313
+ 1
11314
+ 1
11315
+ 1
11316
+ 1
11317
+ 1
11318
+ 1
11319
+ 1
11320
+ 1
11321
+ 1
11322
+ 1
11323
+ 1
11324
+ 1
11325
+ 1
11326
+ 1
11327
+ 1
11328
+ 1
11329
+ 1
11330
+ 1
11331
+ 1
11332
+ 2
11333
+ 1
11334
+ 1
11335
+ 1
11336
+ 1
11337
+ 1
11338
+ 1
11339
+ 1
11340
+ 1
11341
+ 1
11342
+ 1
11343
+ 1
11344
+ 1
11345
+ 1
11346
+ 1
11347
+ 1
11348
+ 1
11349
+ 1
11350
+ 1
11351
+ 1
11352
+ 1
11353
+ 1
11354
+ 1
11355
+ 1
11356
+ 1
11357
+ 1
11358
+ 1
11359
+ 1
11360
+ 1
11361
+ 1
11362
+ 1
11363
+ 1
11364
+ 1
11365
+ 1
11366
+ 1
11367
+ 1
11368
+ 1
11369
+ 1
11370
+ 1
11371
+ 1
11372
+ 1
11373
+ 1
11374
+ 1
11375
+ 1
11376
+ 1
11377
+ 1
11378
+ 1
11379
+ 1
11380
+ 1
11381
+ 1
11382
+ 1
11383
+ 1
11384
+ 1
11385
+ 1
11386
+ 1
11387
+ 1
11388
+ 1
11389
+ 1
11390
+ 1
11391
+ 1
11392
+ 1
11393
+ 1
11394
+ 1
11395
+ 1
11396
+ 1
11397
+ 1
11398
+ 1
11399
+ 1
11400
+ 1
11401
+ 1
11402
+ 1
11403
+ 1
11404
+ 1
11405
+ 1
11406
+ 1
11407
+ 1
11408
+ 1
11409
+ 1
11410
+ 1
11411
+ 1
11412
+ 1
11413
+ 1
11414
+ 1
11415
+ 1
11416
+ 1
11417
+ 1
11418
+ 1
11419
+ 1
11420
+ 1
11421
+ 1
11422
+ 1
11423
+ 1
11424
+ 1
11425
+ 1
11426
+ 1
11427
+ 1
11428
+ 1
11429
+ 1
11430
+ 1
11431
+ 1
11432
+ 1
11433
+ 1
11434
+ 1
11435
+ 1
11436
+ 1
11437
+ 1
11438
+ 1
11439
+ 1
11440
+ 1
11441
+ 1
11442
+ 1
11443
+ 1
11444
+ 1
11445
+ 1
11446
+ 1
11447
+ 1
11448
+ 1
11449
+ 1
11450
+ 1
11451
+ 1
11452
+ 1
11453
+ 1
11454
+ 1
11455
+ 1
11456
+ 1
11457
+ 1
11458
+ 1
11459
+ 1
11460
+ 1
11461
+ 1
11462
+ 1
11463
+ 1
11464
+ 1
11465
+ 1
11466
+ 1
11467
+ 1
11468
+ 1
11469
+ 1
11470
+ 1
11471
+ 1
11472
+ 1
11473
+ 1
11474
+ 1
11475
+ 1
11476
+ 1
11477
+ 1
11478
+ 1
11479
+ 1
11480
+ 1
11481
+ 1
11482
+ 1
11483
+ 1
11484
+ 1
11485
+ 1
11486
+ 1
11487
+ 1
11488
+ 1
11489
+ 1
11490
+ 1
11491
+ 1
11492
+ 1
11493
+ 1
11494
+ 1
11495
+ 1
11496
+ 1
11497
+ 1
11498
+ 1
11499
+ 1
11500
+ 1
11501
+ 1
11502
+ 1
11503
+ 1
11504
+ 1
11505
+ 1
11506
+ 1
11507
+ 1
11508
+ 1
11509
+ 1
11510
+ 1
11511
+ 1
11512
+ 1
11513
+ 1
11514
+ 1
11515
+ 1
11516
+ 1
11517
+ 1
11518
+ 1
11519
+ 1
11520
+ 1
11521
+ 1
11522
+ 1
11523
+ 1
11524
+ 1
11525
+ 1
11526
+ 1
11527
+ 1
11528
+ 1
11529
+ 1
11530
+ 1
11531
+ 1
11532
+ 1
11533
+ 1
11534
+ 1
11535
+ 1
11536
+ 1
11537
+ 1
11538
+ 1
11539
+ 1
11540
+ 1
11541
+ 1
11542
+ 1
11543
+ 1
11544
+ 1
11545
+ 1
11546
+ 1
11547
+ 1
11548
+ 1
11549
+ 1
11550
+ 1
11551
+ 1
11552
+ 1
11553
+ 1
11554
+ 1
11555
+ 1
11556
+ 1
11557
+ 1
11558
+ 1
11559
+ 1
11560
+ 1
11561
+ 1
11562
+ 1
11563
+ 1
11564
+ 1
11565
+ 1
11566
+ 1
11567
+ 1
11568
+ 1
11569
+ 1
11570
+ 1
11571
+ 1
11572
+ 1
11573
+ 1
11574
+ 1
11575
+ 1
11576
+ 1
11577
+ 1
11578
+ 1
11579
+ 1
11580
+ 1
11581
+ 1
11582
+ 1
11583
+ 1
11584
+ 1
11585
+ 1
11586
+ 1
11587
+ 1
11588
+ 1
11589
+ 1
11590
+ 1
11591
+ 1
11592
+ 1
11593
+ 1
11594
+ 1
11595
+ 1
11596
+ 1
11597
+ 1
11598
+ 1
11599
+ 1
11600
+ 1
11601
+ 1
11602
+ 1
11603
+ 1
11604
+ 1
11605
+ 1
11606
+ 1
11607
+ 1
11608
+ 2
11609
+ 1
11610
+ 1
11611
+ 2
11612
+ 1
11613
+ 1
11614
+ 1
11615
+ 1
11616
+ 1
11617
+ 1
11618
+ 1
11619
+ 1
11620
+ 1
11621
+ 1
11622
+ 1
11623
+ 1
11624
+ 1
11625
+ 1
11626
+ 1
11627
+ 1
11628
+ 1
11629
+ 1
11630
+ 1
11631
+ 1
11632
+ 1
11633
+ 1
11634
+ 1
11635
+ 1
11636
+ 1
11637
+ 1
11638
+ 1
11639
+ 1
11640
+ 1
11641
+ 1
11642
+ 1
11643
+ 1
11644
+ 1
11645
+ 1
11646
+ 1
11647
+ 1
11648
+ 1
11649
+ 1
11650
+ 1
11651
+ 1
11652
+ 1
11653
+ 1
11654
+ 1
11655
+ 1
11656
+ 1
11657
+ 1
11658
+ 1
11659
+ 1
11660
+ 1
11661
+ 1
11662
+ 1
11663
+ 1
11664
+ 1
11665
+ 1
11666
+ 1
11667
+ 1
11668
+ 1
11669
+ 1
11670
+ 1
11671
+ 1
11672
+ 1
11673
+ 1
11674
+ 1
11675
+ 1
11676
+ 1
11677
+ 1
11678
+ 1
11679
+ 1
11680
+ 1
11681
+ 1
11682
+ 1
11683
+ 1
11684
+ 1
11685
+ 1
11686
+ 1
11687
+ 1
11688
+ 1
11689
+ 1
11690
+ 1
11691
+ 1
11692
+ 1
11693
+ 1
11694
+ 1
11695
+ 1
11696
+ 1
11697
+ 1
11698
+ 1
11699
+ 1
11700
+ 1
11701
+ 1
11702
+ 1
11703
+ 1
11704
+ 1
11705
+ 1
11706
+ 1
11707
+ 1
11708
+ 1
11709
+ 1
11710
+ 1
11711
+ 1
11712
+ 1
11713
+ 1
11714
+ 1
11715
+ 1
11716
+ 1
11717
+ 1
11718
+ 1
11719
+ 1
11720
+ 1
11721
+ 1
11722
+ 1
11723
+ 1
11724
+ 1
11725
+ 1
11726
+ 1
11727
+ 1
11728
+ 1
11729
+ 1
11730
+ 1
11731
+ 1
11732
+ 1
11733
+ 1
11734
+ 1
11735
+ 1
11736
+ 1
11737
+ 1
11738
+ 1
11739
+ 1
11740
+ 1
11741
+ 1
11742
+ 1
11743
+ 1
11744
+ 1
11745
+ 1
11746
+ 1
11747
+ 1
11748
+ 1
11749
+ 1
11750
+ 1
11751
+ 1
11752
+ 1
11753
+ 1
11754
+ 1
11755
+ 1
11756
+ 1
11757
+ 1
11758
+ 1
11759
+ 1
11760
+ 1
11761
+ 1
11762
+ 1
11763
+ 1
11764
+ 1
11765
+ 1
11766
+ 1
11767
+ 1
11768
+ 1
11769
+ 1
11770
+ 1
11771
+ 1
11772
+ 1
11773
+ 1
11774
+ 1
11775
+ 1
11776
+ 1
11777
+ 1
11778
+ 1
11779
+ 1
11780
+ 1
11781
+ 1
11782
+ 1
11783
+ 1
11784
+ 1
11785
+ 1
11786
+ 1
11787
+ 1
11788
+ 1
11789
+ 1
11790
+ 1
11791
+ 1
11792
+ 1
11793
+ 1
11794
+ 1
11795
+ 1
11796
+ 1
11797
+ 1
11798
+ 1
11799
+ 1
11800
+ 1
11801
+ 1
11802
+ 1
11803
+ 1
11804
+ 1
11805
+ 1
11806
+ 1
11807
+ 1
11808
+ 1
11809
+ 1
11810
+ 1
11811
+ 1
11812
+ 1
11813
+ 1
11814
+ 1
11815
+ 1
11816
+ 1
11817
+ 1
11818
+ 1
11819
+ 1
11820
+ 1
11821
+ 1
11822
+ 1
11823
+ 1
11824
+ 1
11825
+ 1
11826
+ 1
11827
+ 1
11828
+ 1
11829
+ 1
11830
+ 1
11831
+ 1
11832
+ 1
11833
+ 1
11834
+ 1
11835
+ 1
11836
+ 1
11837
+ 1
11838
+ 1
11839
+ 1
11840
+ 1
11841
+ 1
11842
+ 1
11843
+ 1
11844
+ 1
11845
+ 1
11846
+ 1
11847
+ 1
11848
+ 1
11849
+ 1
11850
+ 1
11851
+ 1
11852
+ 1
11853
+ 1
11854
+ 1
11855
+ 1
11856
+ 1
11857
+ 1
11858
+ 1
11859
+ 1
11860
+ 1
11861
+ 1
11862
+ 1
11863
+ 1
11864
+ 1
11865
+ 2
11866
+ 1
11867
+ 1
11868
+ 1
11869
+ 1
11870
+ 1
11871
+ 1
11872
+ 1
11873
+ 1
11874
+ 1
11875
+ 1
11876
+ 1
11877
+ 1
11878
+ 1
11879
+ 1
11880
+ 1
11881
+ 1
11882
+ 1
11883
+ 1
11884
+ 1
11885
+ 1
11886
+ 1
11887
+ 1
11888
+ 1
11889
+ 1
11890
+ 1
11891
+ 1
11892
+ 1
11893
+ 1
11894
+ 1
11895
+ 1
11896
+ 1
11897
+ 1
11898
+ 1
11899
+ 1
11900
+ 1
11901
+ 1
11902
+ 1
11903
+ 1
11904
+ 1
11905
+ 1
11906
+ 1
11907
+ 1
11908
+ 1
11909
+ 1
11910
+ 1
11911
+ 1
11912
+ 1
11913
+ 1
11914
+ 1
11915
+ 1
11916
+ 1
11917
+ 1
11918
+ 1
11919
+ 1
11920
+ 1
11921
+ 1
11922
+ 1
11923
+ 1
11924
+ 1
11925
+ 1
11926
+ 1
11927
+ 1
11928
+ 1
11929
+ 1
11930
+ 1
11931
+ 1
11932
+ 1
11933
+ 1
11934
+ 1
11935
+ 1
11936
+ 1
11937
+ 1
11938
+ 1
11939
+ 1
11940
+ 1
11941
+ 1
11942
+ 1
11943
+ 1
11944
+ 1
11945
+ 1
11946
+ 1
11947
+ 1
11948
+ 1
11949
+ 1
11950
+ 1
11951
+ 1
11952
+ 1
11953
+ 1
11954
+ 1
11955
+ 1
11956
+ 1
11957
+ 1
11958
+ 1
11959
+ 1
11960
+ 1
11961
+ 1
11962
+ 1
11963
+ 1
11964
+ 1
11965
+ 1
11966
+ 1
11967
+ 1
11968
+ 1
11969
+ 1
11970
+ 1
11971
+ 1
11972
+ 1
11973
+ 1
11974
+ 1
11975
+ 1
11976
+ 1
11977
+ 1
11978
+ 1
11979
+ 1
11980
+ 1
11981
+ 1
11982
+ 1
11983
+ 1
11984
+ 1
11985
+ 1
11986
+ 1
11987
+ 1
11988
+ 1
11989
+ 1
11990
+ 1
11991
+ 1
11992
+ 1
11993
+ 1
11994
+ 1
11995
+ 1
11996
+ 1
11997
+ 1
11998
+ 1
11999
+ 1
12000
+ 1
12001
+ 1
12002
+ 1
12003
+ 1
12004
+ 1
12005
+ 1
12006
+ 1
12007
+ 1
12008
+ 1
12009
+ 1
12010
+ 1
12011
+ 1
12012
+ 1
12013
+ 1
12014
+ 1
12015
+ 1
12016
+ 1
12017
+ 1
12018
+ 1
12019
+ 1
12020
+ 1
12021
+ 1
12022
+ 1
12023
+ 1
12024
+ 1
12025
+ 1
12026
+ 1
12027
+ 1
12028
+ 1
12029
+ 1
12030
+ 1
12031
+ 1
12032
+ 1
12033
+ 1
12034
+ 1
12035
+ 1
12036
+ 1
12037
+ 1
12038
+ 1
12039
+ 1
12040
+ 1
12041
+ 1
12042
+ 1
12043
+ 1
12044
+ 1
12045
+ 1
12046
+ 1
12047
+ 1
12048
+ 1
12049
+ 1
12050
+ 1
12051
+ 1
12052
+ 1
12053
+ 1
12054
+ 1
12055
+ 1
12056
+ 1
12057
+ 1
12058
+ 1
12059
+ 1
12060
+ 1
12061
+ 1
12062
+ 1
12063
+ 1
12064
+ 1
12065
+ 1
12066
+ 1
12067
+ 1
12068
+ 1
12069
+ 1
12070
+ 1
12071
+ 1
12072
+ 1
12073
+ 1
12074
+ 1
12075
+ 1
12076
+ 1
12077
+ 1
12078
+ 1
12079
+ 1
12080
+ 1
12081
+ 1
12082
+ 1
12083
+ 1
12084
+ 1
12085
+ 1
12086
+ 1
12087
+ 1
12088
+ 1
12089
+ 1
12090
+ 1
12091
+ 1
12092
+ 1
12093
+ 1
12094
+ 1
12095
+ 1
12096
+ 1
12097
+ 1
12098
+ 1
12099
+ 1
12100
+ 1
12101
+ 1
12102
+ 1
12103
+ 1
12104
+ 1
12105
+ 1
12106
+ 1
12107
+ 1
12108
+ 1
12109
+ 1
12110
+ 1
12111
+ 1
12112
+ 1
12113
+ 1
12114
+ 1
12115
+ 1
12116
+ 1
12117
+ 1
12118
+ 1
12119
+ 1
12120
+ 1
12121
+ 1
12122
+ 1
12123
+ 1
12124
+ 1
12125
+ 1
12126
+ 1
12127
+ 1
12128
+ 1
12129
+ 1
12130
+ 1
12131
+ 1
12132
+ 1
12133
+ 1
12134
+ 1
12135
+ 1
12136
+ 1
12137
+ 1
12138
+ 1
12139
+ 1
12140
+ 1
12141
+ 1
12142
+ 1
12143
+ 1
12144
+ 1
12145
+ 1
12146
+ 1
12147
+ 1
12148
+ 1
12149
+ 1
12150
+ 1
12151
+ 1
12152
+ 1
12153
+ 1
12154
+ 1
12155
+ 1
12156
+ 1
12157
+ 1
12158
+ 1
12159
+ 1
12160
+ 1
12161
+ 1
12162
+ 1
12163
+ 1
12164
+ 1
12165
+ 1
12166
+ 1
12167
+ 1
12168
+ 1
12169
+ 1
12170
+ 1
12171
+ 1
12172
+ 1
12173
+ 1
12174
+ 1
12175
+ 1
12176
+ 1
12177
+ 1
12178
+ 1
12179
+ 1
12180
+ 1
12181
+ 1
12182
+ 1
12183
+ 1
12184
+ 1
12185
+ 1
12186
+ 1
12187
+ 1
12188
+ 1
12189
+ 1
12190
+ 1
12191
+ 1
12192
+ 1
12193
+ 1
12194
+ 1
12195
+ 1
12196
+ 1
12197
+ 1
12198
+ 1
12199
+ 1
12200
+ 1
12201
+ 1
12202
+ 1
12203
+ 1
12204
+ 1
12205
+ 1
12206
+ 1
12207
+ 1
12208
+ 1
12209
+ 1
12210
+ 1
12211
+ 1
12212
+ 1
12213
+ 1
12214
+ 1
12215
+ 1
12216
+ 1
12217
+ 1
12218
+ 1
12219
+ 1
12220
+ 1
12221
+ 1
12222
+ 1
12223
+ 1
12224
+ 1
12225
+ 1
12226
+ 1
12227
+ 1
12228
+ 1
12229
+ 1
12230
+ 1
12231
+ 1
12232
+ 1
12233
+ 1
12234
+ 1
12235
+ 1
12236
+ 1
12237
+ 1
12238
+ 1
12239
+ 1
12240
+ 1
12241
+ 1
12242
+ 1
12243
+ 1
12244
+ 1
12245
+ 1
12246
+ 1
12247
+ 1
12248
+ 1
12249
+ 1
12250
+ 1
12251
+ 1
12252
+ 1
12253
+ 1
12254
+ 1
12255
+ 1
12256
+ 1
12257
+ 1
12258
+ 1
12259
+ 1
12260
+ 1
12261
+ 1
12262
+ 1
12263
+ 1
12264
+ 1
12265
+ 1
12266
+ 1
12267
+ 1
12268
+ 1
12269
+ 1
12270
+ 1
12271
+ 1
12272
+ 1
12273
+ 1
12274
+ 1
12275
+ 1
12276
+ 1
12277
+ 1
12278
+ 1
12279
+ 1
12280
+ 1
12281
+ 1
12282
+ 1
12283
+ 1
12284
+ 1
12285
+ 1
12286
+ 1
12287
+ 1
12288
+ 1
12289
+ 1
12290
+ 1
12291
+ 1
12292
+ 1
12293
+ 1
12294
+ 1
12295
+ 1
12296
+ 1
12297
+ 1
12298
+ 1
12299
+ 1
12300
+ 1
12301
+ 1
12302
+ 1
12303
+ 1
12304
+ 1
12305
+ 1
12306
+ 1
12307
+ 1
12308
+ 1
12309
+ 1
12310
+ 1
12311
+ 1
12312
+ 1
12313
+ 1
12314
+ 1
12315
+ 1
12316
+ 1
12317
+ 1
12318
+ 1
12319
+ 1
12320
+ 1
12321
+ 1
12322
+ 1
12323
+ 1
12324
+ 1
12325
+ 1
12326
+ 1
12327
+ 1
12328
+ 1
12329
+ 1
12330
+ 1
12331
+ 1
12332
+ 1
12333
+ 1
12334
+ 1
12335
+ 1
12336
+ 1
12337
+ 1
12338
+ 1
12339
+ 1
12340
+ 1
12341
+ 1
12342
+ 1
12343
+ 1
12344
+ 1
12345
+ 1
12346
+ 1
12347
+ 1
12348
+ 1
12349
+ 1
12350
+ 1
12351
+ 1
12352
+ 1
12353
+ 1
12354
+ 1
12355
+ 1
12356
+ 1
12357
+ 1
12358
+ 1
12359
+ 1
12360
+ 1
12361
+ 1
12362
+ 1
12363
+ 1
12364
+ 1
12365
+ 1
12366
+ 1
12367
+ 1
12368
+ 1
12369
+ 1
12370
+ 1
12371
+ 1
12372
+ 1
12373
+ 1
12374
+ 1
12375
+ 1
12376
+ 1
12377
+ 1
12378
+ 1
12379
+ 1
12380
+ 1
12381
+ 1
12382
+ 1
12383
+ 1
12384
+ 1
12385
+ 1
12386
+ 1
12387
+ 1
12388
+ 1
12389
+ 1
12390
+ 1
12391
+ 2
12392
+ 1
12393
+ 1
12394
+ 2
12395
+ 1
12396
+ 1
12397
+ 2
12398
+ 1
12399
+ 1
12400
+ 2
12401
+ 1
12402
+ 1
12403
+ 1
12404
+ 1
12405
+ 1
12406
+ 1
12407
+ 1
12408
+ 1
12409
+ 1
12410
+ 1
12411
+ 1
12412
+ 1
12413
+ 1
12414
+ 1
12415
+ 1
12416
+ 1
12417
+ 1
12418
+ 1
12419
+ 1
12420
+ 1
12421
+ 1
12422
+ 1
12423
+ 1
12424
+ 1
12425
+ 1
12426
+ 1
12427
+ 1
12428
+ 1
12429
+ 1
12430
+ 1
12431
+ 1
12432
+ 1
12433
+ 1
12434
+ 1
12435
+ 1
12436
+ 1
12437
+ 1
12438
+ 1
12439
+ 1
12440
+ 1
12441
+ 1
12442
+ 1
12443
+ 1
12444
+ 1
12445
+ 1
12446
+ 1
12447
+ 1
12448
+ 1
12449
+ 1
12450
+ 1
12451
+ 1
12452
+ 1
12453
+ 1
12454
+ 1
12455
+ 1
12456
+ 1
12457
+ 1
12458
+ 1
12459
+ 1
12460
+ 1
12461
+ 1
12462
+ 1
12463
+ 1
12464
+ 1
12465
+ 1
12466
+ 1
12467
+ 1
12468
+ 1
12469
+ 1
12470
+ 1
12471
+ 1
12472
+ 1
12473
+ 1
12474
+ 1
12475
+ 1
12476
+ 1
12477
+ 1
12478
+ 1
12479
+ 1
12480
+ 1
12481
+ 1
12482
+ 1
12483
+ 1
12484
+ 1
12485
+ 1
12486
+ 1
12487
+ 1
12488
+ 1
12489
+ 1
12490
+ 1
12491
+ 1
12492
+ 1
12493
+ 1
12494
+ 1
12495
+ 1
12496
+ 1
12497
+ 1
12498
+ 1
12499
+ 1
12500
+ 1
12501
+ 1
12502
+ 1
12503
+ 1
12504
+ 1
12505
+ 1
12506
+ 1
12507
+ 1
12508
+ 1
12509
+ 1
12510
+ 1
12511
+ 1
12512
+ 1
12513
+ 1
12514
+ 1
12515
+ 1
12516
+ 1
12517
+ 1
12518
+ 1
12519
+ 1
12520
+ 1
12521
+ 1
12522
+ 1
12523
+ 1
12524
+ 1
12525
+ 1
12526
+ 1
12527
+ 1
12528
+ 1
12529
+ 1
12530
+ 1
12531
+ 1
12532
+ 1
12533
+ 1
12534
+ 1
12535
+ 1
12536
+ 1
12537
+ 1
12538
+ 1
12539
+ 1
12540
+ 1
12541
+ 1
12542
+ 1
12543
+ 1
12544
+ 1
12545
+ 1
12546
+ 1
12547
+ 1
12548
+ 1
12549
+ 1
12550
+ 1
12551
+ 1
12552
+ 1
12553
+ 1
12554
+ 1
12555
+ 1
12556
+ 1
12557
+ 1
12558
+ 1
12559
+ 1
12560
+ 1
12561
+ 1
12562
+ 1
12563
+ 1
12564
+ 1
12565
+ 1
12566
+ 1
12567
+ 1
12568
+ 1
12569
+ 1
12570
+ 1
12571
+ 1
12572
+ 1
12573
+ 1
12574
+ 1
12575
+ 1
12576
+ 1
12577
+ 1
12578
+ 1
12579
+ 1
12580
+ 1
12581
+ 1
12582
+ 1
12583
+ 1
12584
+ 1
12585
+ 1
12586
+ 1
12587
+ 1
12588
+ 1
12589
+ 1
12590
+ 1
12591
+ 1
12592
+ 1
12593
+ 1
12594
+ 1
12595
+ 1
12596
+ 1
12597
+ 1
12598
+ 1
12599
+ 1
12600
+ 1
12601
+ 1
12602
+ 1
12603
+ 1
12604
+ 1
12605
+ 1
12606
+ 1
12607
+ 1
12608
+ 1
12609
+ 1
12610
+ 1
12611
+ 1
12612
+ 1
12613
+ 1
12614
+ 1
12615
+ 1
12616
+ 1
12617
+ 1
12618
+ 1
12619
+ 1
12620
+ 1
12621
+ 1
12622
+ 1
12623
+ 1
12624
+ 1
12625
+ 1
12626
+ 1
12627
+ 1
12628
+ 1
12629
+ 1
12630
+ 1
12631
+ 1
12632
+ 1
12633
+ 1
12634
+ 1
12635
+ 1
12636
+ 1
12637
+ 1
12638
+ 1
12639
+ 1
12640
+ 1
12641
+ 1
12642
+ 1
12643
+ 1
12644
+ 1
12645
+ 1
12646
+ 1
12647
+ 1
12648
+ 1
12649
+ 1
12650
+ 1
12651
+ 1
12652
+ 1
12653
+ 1
12654
+ 1
12655
+ 1
12656
+ 1
12657
+ 1
12658
+ 1
12659
+ 1
12660
+ 1
12661
+ 1
12662
+ 1
12663
+ 1
12664
+ 1
12665
+ 1
12666
+ 1
12667
+ 1
12668
+ 1
12669
+ 1
12670
+ 1
12671
+ 1
12672
+ 2
12673
+ 1
12674
+ 1
12675
+ 1
12676
+ 1
12677
+ 1
12678
+ 1
12679
+ 1
12680
+ 1
12681
+ 1
12682
+ 1
12683
+ 1
12684
+ 1
12685
+ 1
12686
+ 1
12687
+ 1
12688
+ 1
12689
+ 1
12690
+ 1
12691
+ 1
12692
+ 1
12693
+ 1
12694
+ 1
12695
+ 1
12696
+ 1
12697
+ 1
12698
+ 1
12699
+ 1
12700
+ 1
12701
+ 1
12702
+ 1
12703
+ 1
12704
+ 1
12705
+ 1
12706
+ 1
12707
+ 1
12708
+ 1
12709
+ 1
12710
+ 1
12711
+ 1
12712
+ 1
12713
+ 1
12714
+ 1
12715
+ 1
12716
+ 1
12717
+ 1
12718
+ 1
12719
+ 1
12720
+ 1
12721
+ 1
12722
+ 1
12723
+ 1
12724
+ 1
12725
+ 1
12726
+ 1
12727
+ 1
12728
+ 1
12729
+ 1
12730
+ 1
12731
+ 1
12732
+ 1
12733
+ 1
12734
+ 1
12735
+ 1
12736
+ 1
12737
+ 1
12738
+ 1
12739
+ 1
12740
+ 1
12741
+ 1
12742
+ 1
12743
+ 1
12744
+ 1
12745
+ 1
12746
+ 1
12747
+ 1
12748
+ 1
12749
+ 1
12750
+ 1
12751
+ 1
12752
+ 1
12753
+ 1
12754
+ 1
12755
+ 1
12756
+ 1
12757
+ 1
12758
+ 1
12759
+ 1
12760
+ 1
12761
+ 1
12762
+ 1
12763
+ 1
12764
+ 1
12765
+ 1
12766
+ 1
12767
+ 1
12768
+ 1
12769
+ 1
12770
+ 1
12771
+ 1
12772
+ 1
12773
+ 1
12774
+ 1
12775
+ 1
12776
+ 1
12777
+ 1
12778
+ 1
12779
+ 1
12780
+ 1
12781
+ 1
12782
+ 1
12783
+ 1
12784
+ 1
12785
+ 1
12786
+ 1
12787
+ 1
12788
+ 1
12789
+ 1
12790
+ 1
12791
+ 1
12792
+ 1
12793
+ 1
12794
+ 1
12795
+ 1
12796
+ 1
12797
+ 1
12798
+ 1
12799
+ 1
12800
+ 1
12801
+ 1
12802
+ 1
12803
+ 1
12804
+ 1
12805
+ 1
12806
+ 1
12807
+ 1
12808
+ 1
12809
+ 1
12810
+ 1
12811
+ 1
12812
+ 1
12813
+ 1
12814
+ 1
12815
+ 1
12816
+ 1
12817
+ 1
12818
+ 1
12819
+ 1
12820
+ 1
12821
+ 1
12822
+ 1
12823
+ 1
12824
+ 1
12825
+ 1
12826
+ 1
12827
+ 1
12828
+ 1
12829
+ 1
12830
+ 1
12831
+ 1
12832
+ 1
12833
+ 1
12834
+ 1
12835
+ 1
12836
+ 1
12837
+ 1
12838
+ 1
12839
+ 1
12840
+ 1
12841
+ 1
12842
+ 1
12843
+ 1
12844
+ 1
12845
+ 1
12846
+ 1
12847
+ 1
12848
+ 1
12849
+ 1
12850
+ 1
12851
+ 1
12852
+ 1
12853
+ 1
12854
+ 1
12855
+ 1
12856
+ 1
12857
+ 1
12858
+ 1
12859
+ 1
12860
+ 1
12861
+ 1
12862
+ 1
12863
+ 1
12864
+ 1
12865
+ 1
12866
+ 1
12867
+ 1
12868
+ 1
12869
+ 1
12870
+ 1
12871
+ 1
12872
+ 1
12873
+ 1
12874
+ 1
12875
+ 1
12876
+ 1
12877
+ 1
12878
+ 1
12879
+ 1
12880
+ 1
12881
+ 1
12882
+ 1
12883
+ 1
12884
+ 1
12885
+ 1
12886
+ 1
12887
+ 1
12888
+ 1
12889
+ 1
12890
+ 1
12891
+ 1
12892
+ 1
12893
+ 1
12894
+ 1
12895
+ 1
12896
+ 1
12897
+ 1
12898
+ 1
12899
+ 1
12900
+ 1
12901
+ 1
12902
+ 1
12903
+ 1
12904
+ 1
12905
+ 1
12906
+ 1
12907
+ 1
12908
+ 1
12909
+ 1
12910
+ 1
12911
+ 1
12912
+ 1
12913
+ 1
12914
+ 1
12915
+ 1
12916
+ 1
12917
+ 1
12918
+ 1
12919
+ 1
12920
+ 1
12921
+ 1
12922
+ 2
12923
+ 1
12924
+ 1
12925
+ 1
12926
+ 1
12927
+ 1
12928
+ 1
12929
+ 1
12930
+ 1
12931
+ 2
12932
+ 1
12933
+ 1
12934
+ 1
12935
+ 1
12936
+ 1
12937
+ 1
12938
+ 1
12939
+ 1
12940
+ 1
12941
+ 1
12942
+ 1
12943
+ 1
12944
+ 1
12945
+ 1
12946
+ 1
12947
+ 1
12948
+ 1
12949
+ 1
12950
+ 1
12951
+ 1
12952
+ 1
12953
+ 1
12954
+ 1
12955
+ 1
12956
+ 1
12957
+ 1
12958
+ 1
12959
+ 1
12960
+ 1
12961
+ 1
12962
+ 1
12963
+ 1
12964
+ 1
12965
+ 1
12966
+ 1
12967
+ 1
12968
+ 1
12969
+ 1
12970
+ 1
12971
+ 1
12972
+ 1
12973
+ 1
12974
+ 1
12975
+ 1
12976
+ 1
12977
+ 1
12978
+ 1
12979
+ 1
12980
+ 1
12981
+ 1
12982
+ 1
12983
+ 1
12984
+ 1
12985
+ 1
12986
+ 1
12987
+ 1
12988
+ 1
12989
+ 1
12990
+ 1
12991
+ 1
12992
+ 1
12993
+ 1
12994
+ 1
12995
+ 1
12996
+ 1
12997
+ 1
12998
+ 1
12999
+ 1
13000
+ 1
13001
+ 1
13002
+ 1
13003
+ 1
13004
+ 1
13005
+ 1
13006
+ 1
13007
+ 1
13008
+ 1
13009
+ 1
13010
+ 1
13011
+ 1
13012
+ 1
13013
+ 1
13014
+ 1
13015
+ 1
13016
+ 1
13017
+ 1
13018
+ 1
13019
+ 1
13020
+ 1
13021
+ 1
13022
+ 1
13023
+ 1
13024
+ 1
13025
+ 1
13026
+ 1
13027
+ 1
13028
+ 1
13029
+ 1
13030
+ 1
13031
+ 1
13032
+ 1
13033
+ 1
13034
+ 1
13035
+ 1
13036
+ 1
13037
+ 1
13038
+ 1
13039
+ 1
13040
+ 1
13041
+ 1
13042
+ 1
13043
+ 1
13044
+ 1
13045
+ 1
13046
+ 1
13047
+ 1
13048
+ 1
13049
+ 1
13050
+ 1
13051
+ 1
13052
+ 1
13053
+ 1
13054
+ 1
13055
+ 1
13056
+ 1
13057
+ 1
13058
+ 1
13059
+ 1
13060
+ 1
13061
+ 1
13062
+ 1
13063
+ 1
13064
+ 1
13065
+ 1
13066
+ 1
13067
+ 1
13068
+ 1
13069
+ 1
13070
+ 1
13071
+ 1
13072
+ 1
13073
+ 1
13074
+ 1
13075
+ 1
13076
+ 1
13077
+ 1
13078
+ 1
13079
+ 1
13080
+ 1
13081
+ 1
13082
+ 1
13083
+ 1
13084
+ 1
13085
+ 1
13086
+ 1
13087
+ 1
13088
+ 1
13089
+ 1
13090
+ 1
13091
+ 1
13092
+ 1
13093
+ 1
13094
+ 1
13095
+ 1
13096
+ 1
13097
+ 1
13098
+ 1
13099
+ 1
13100
+ 1
13101
+ 1
13102
+ 1
13103
+ 1
13104
+ 1
13105
+ 1
13106
+ 1
13107
+ 1
13108
+ 1
13109
+ 1
13110
+ 1
13111
+ 1
13112
+ 1
13113
+ 1
13114
+ 1
13115
+ 1
13116
+ 1
13117
+ 1
13118
+ 1
13119
+ 1
13120
+ 1
13121
+ 1
13122
+ 1
13123
+ 1
13124
+ 1
13125
+ 1
13126
+ 1
13127
+ 1
13128
+ 1
13129
+ 1
13130
+ 1
13131
+ 1
13132
+ 1
13133
+ 1
13134
+ 1
13135
+ 1
13136
+ 1
13137
+ 1
13138
+ 1
13139
+ 1
13140
+ 1
13141
+ 1
13142
+ 1
13143
+ 1
13144
+ 1
13145
+ 1
13146
+ 1
13147
+ 1
13148
+ 1
13149
+ 1
13150
+ 1
13151
+ 1
13152
+ 1
13153
+ 1
13154
+ 1
13155
+ 1
13156
+ 1
13157
+ 1
13158
+ 1
13159
+ 1
13160
+ 1
13161
+ 1
13162
+ 1
13163
+ 1
13164
+ 1
13165
+ 1
13166
+ 1
13167
+ 1
13168
+ 1
13169
+ 1
13170
+ 1
13171
+ 1
13172
+ 1
13173
+ 1
13174
+ 1
13175
+ 1
13176
+ 1
13177
+ 1
13178
+ 1
13179
+ 2
13180
+ 1
13181
+ 1
13182
+ 1
13183
+ 1
13184
+ 1
13185
+ 1
13186
+ 1
13187
+ 1
13188
+ 1
13189
+ 1
13190
+ 1
13191
+ 1
13192
+ 1
13193
+ 1
13194
+ 1
13195
+ 1
13196
+ 1
13197
+ 1
13198
+ 1
13199
+ 1
13200
+ 1
13201
+ 1
13202
+ 1
13203
+ 1
13204
+ 1
13205
+ 1
13206
+ 1
13207
+ 1
13208
+ 1
13209
+ 1
13210
+ 1
13211
+ 1
13212
+ 1
13213
+ 1
13214
+ 1
13215
+ 1
13216
+ 1
13217
+ 1
13218
+ 1
13219
+ 1
13220
+ 1
13221
+ 1
13222
+ 1
13223
+ 1
13224
+ 1
13225
+ 1
13226
+ 1
13227
+ 1
13228
+ 1
13229
+ 1
13230
+ 1
13231
+ 1
13232
+ 1
13233
+ 1
13234
+ 1
13235
+ 1
13236
+ 1
13237
+ 1
13238
+ 1
13239
+ 1
13240
+ 1
13241
+ 1
13242
+ 1
13243
+ 1
13244
+ 1
13245
+ 1
13246
+ 1
13247
+ 1
13248
+ 1
13249
+ 1
13250
+ 1
13251
+ 1
13252
+ 1
13253
+ 1
13254
+ 1
13255
+ 1
13256
+ 1
13257
+ 1
13258
+ 1
13259
+ 1
13260
+ 1
13261
+ 1
13262
+ 1
13263
+ 1
13264
+ 1
13265
+ 1
13266
+ 1
13267
+ 1
13268
+ 1
13269
+ 1
13270
+ 1
13271
+ 1
13272
+ 1
13273
+ 1
13274
+ 1
13275
+ 1
13276
+ 1
13277
+ 1
13278
+ 1
13279
+ 1
13280
+ 1
13281
+ 1
13282
+ 1
13283
+ 1
13284
+ 1
13285
+ 1
13286
+ 1
13287
+ 1
13288
+ 1
13289
+ 1
13290
+ 1
13291
+ 1
13292
+ 1
13293
+ 1
13294
+ 1
13295
+ 1
13296
+ 1
13297
+ 1
13298
+ 1
13299
+ 1
13300
+ 1
13301
+ 1
13302
+ 1
13303
+ 1
13304
+ 1
13305
+ 1
13306
+ 1
13307
+ 1
13308
+ 1
13309
+ 1
13310
+ 1
13311
+ 1
13312
+ 1
13313
+ 1
13314
+ 1
13315
+ 1
13316
+ 1
13317
+ 1
13318
+ 1
13319
+ 1
13320
+ 1
13321
+ 1
13322
+ 1
13323
+ 1
13324
+ 1
13325
+ 1
13326
+ 1
13327
+ 1
13328
+ 1
13329
+ 1
13330
+ 1
13331
+ 1
13332
+ 1
13333
+ 1
13334
+ 1
13335
+ 1
13336
+ 1
13337
+ 1
13338
+ 1
13339
+ 1
13340
+ 1
13341
+ 1
13342
+ 1
13343
+ 1
13344
+ 1
13345
+ 1
13346
+ 1
13347
+ 1
13348
+ 1
13349
+ 1
13350
+ 1
13351
+ 1
13352
+ 1
13353
+ 1
13354
+ 1
13355
+ 1
13356
+ 1
13357
+ 1
13358
+ 1
13359
+ 1
13360
+ 1
13361
+ 1
13362
+ 1
13363
+ 1
13364
+ 1
13365
+ 1
13366
+ 1
13367
+ 1
13368
+ 1
13369
+ 1
13370
+ 1
13371
+ 1
13372
+ 1
13373
+ 1
13374
+ 1
13375
+ 1
13376
+ 1
13377
+ 1
13378
+ 1
13379
+ 1
13380
+ 1
13381
+ 1
13382
+ 1
13383
+ 1
13384
+ 1
13385
+ 1
13386
+ 1
13387
+ 1
13388
+ 1
13389
+ 1
13390
+ 1
13391
+ 1
13392
+ 1
13393
+ 1
13394
+ 1
13395
+ 1
13396
+ 1
13397
+ 1
13398
+ 1
13399
+ 1
13400
+ 1
13401
+ 1
13402
+ 1
13403
+ 1
13404
+ 1
13405
+ 1
13406
+ 1
13407
+ 1
13408
+ 1
13409
+ 1
13410
+ 1
13411
+ 1
13412
+ 1
13413
+ 1
13414
+ 1
13415
+ 1
13416
+ 1
13417
+ 1
13418
+ 1
13419
+ 1
13420
+ 1
13421
+ 1
13422
+ 1
13423
+ 1
13424
+ 1
13425
+ 1
13426
+ 1
13427
+ 1
13428
+ 1
13429
+ 1
13430
+ 1
13431
+ 1
13432
+ 1
13433
+ 1
13434
+ 1
13435
+ 1
13436
+ 1
13437
+ 1
13438
+ 1
13439
+ 1
13440
+ 1
13441
+ 1
13442
+ 1
13443
+ 1
13444
+ 1
13445
+ 1
13446
+ 1
13447
+ 1
13448
+ 1
13449
+ 1
13450
+ 1
13451
+ 1
13452
+ 1
13453
+ 1
13454
+ 2
13455
+ 1
13456
+ 1
13457
+ 1
13458
+ 1
13459
+ 1
13460
+ 1
13461
+ 1
13462
+ 1
13463
+ 1
13464
+ 1
13465
+ 1
13466
+ 1
13467
+ 1
13468
+ 1
13469
+ 1
13470
+ 1
13471
+ 1
13472
+ 1
13473
+ 1
13474
+ 1
13475
+ 1
13476
+ 1
13477
+ 1
13478
+ 1
13479
+ 1
13480
+ 1
13481
+ 1
13482
+ 1
13483
+ 1
13484
+ 1
13485
+ 1
13486
+ 1
13487
+ 1
13488
+ 1
13489
+ 1
13490
+ 1
13491
+ 1
13492
+ 1
13493
+ 1
13494
+ 1
13495
+ 1
13496
+ 1
13497
+ 1
13498
+ 1
13499
+ 1
13500
+ 1
13501
+ 1
13502
+ 1
13503
+ 1
13504
+ 1
13505
+ 1
13506
+ 1
13507
+ 1
13508
+ 1
13509
+ 1
13510
+ 1
13511
+ 1
13512
+ 1
13513
+ 1
13514
+ 1
13515
+ 1
13516
+ 1
13517
+ 1
13518
+ 1
13519
+ 1
13520
+ 1
13521
+ 1
13522
+ 1
13523
+ 1
13524
+ 1
13525
+ 1
13526
+ 1
13527
+ 1
13528
+ 1
13529
+ 1
13530
+ 1
13531
+ 1
13532
+ 1
13533
+ 1
13534
+ 1
13535
+ 1
13536
+ 1
13537
+ 1
13538
+ 1
13539
+ 1
13540
+ 1
13541
+ 1
13542
+ 1
13543
+ 1
13544
+ 1
13545
+ 1
13546
+ 1
13547
+ 1
13548
+ 1
13549
+ 1
13550
+ 1
13551
+ 1
13552
+ 1
13553
+ 1
13554
+ 1
13555
+ 1
13556
+ 1
13557
+ 1
13558
+ 1
13559
+ 1
13560
+ 1
13561
+ 1
13562
+ 1
13563
+ 1
13564
+ 1
13565
+ 1
13566
+ 1
13567
+ 1
13568
+ 1
13569
+ 1
13570
+ 1
13571
+ 1
13572
+ 1
13573
+ 1
13574
+ 1
13575
+ 1
13576
+ 1
13577
+ 1
13578
+ 1
13579
+ 1
13580
+ 1
13581
+ 1
13582
+ 1
13583
+ 1
13584
+ 1
13585
+ 1
13586
+ 1
13587
+ 1
13588
+ 1
13589
+ 1
13590
+ 1
13591
+ 1
13592
+ 1
13593
+ 1
13594
+ 1
13595
+ 1
13596
+ 1
13597
+ 1
13598
+ 1
13599
+ 1
13600
+ 1
13601
+ 1
13602
+ 1
13603
+ 1
13604
+ 1
13605
+ 1
13606
+ 1
13607
+ 1
13608
+ 1
13609
+ 1
13610
+ 1
13611
+ 1
13612
+ 1
13613
+ 1
13614
+ 1
13615
+ 1
13616
+ 1
13617
+ 1
13618
+ 1
13619
+ 1
13620
+ 1
13621
+ 1
13622
+ 1
13623
+ 1
13624
+ 1
13625
+ 1
13626
+ 1
13627
+ 1
13628
+ 1
13629
+ 1
13630
+ 1
13631
+ 1
13632
+ 1
13633
+ 1
13634
+ 1
13635
+ 1
13636
+ 1
13637
+ 1
13638
+ 1
13639
+ 1
13640
+ 1
13641
+ 1
13642
+ 1
13643
+ 1
13644
+ 1
13645
+ 1
13646
+ 1
13647
+ 1
13648
+ 1
13649
+ 1
13650
+ 1
13651
+ 1
13652
+ 1
13653
+ 1
13654
+ 1
13655
+ 1
13656
+ 1
13657
+ 1
13658
+ 1
13659
+ 1
13660
+ 1
13661
+ 1
13662
+ 1
13663
+ 1
13664
+ 1
13665
+ 1
13666
+ 1
13667
+ 1
13668
+ 1
13669
+ 1
13670
+ 1
13671
+ 1
13672
+ 1
13673
+ 1
13674
+ 1
13675
+ 1
13676
+ 1
13677
+ 1
13678
+ 1
13679
+ 1
13680
+ 1
13681
+ 1
13682
+ 1
13683
+ 1
13684
+ 1
13685
+ 1
13686
+ 1
13687
+ 1
13688
+ 1
13689
+ 1
13690
+ 1
13691
+ 1
13692
+ 1
13693
+ 1
13694
+ 1
13695
+ 1
13696
+ 1
13697
+ 1
13698
+ 1
13699
+ 1
13700
+ 1
13701
+ 1
13702
+ 1
13703
+ 1
13704
+ 1
13705
+ 1
13706
+ 1
13707
+ 1
13708
+ 1
13709
+ 1
13710
+ 1
13711
+ 1
13712
+ 1
13713
+ 1
13714
+ 1
13715
+ 1
13716
+ 1
13717
+ 1
13718
+ 2
13719
+ 1
13720
+ 1
13721
+ 1
13722
+ 1
13723
+ 1
13724
+ 1
13725
+ 1
13726
+ 1
13727
+ 1
13728
+ 1
13729
+ 1
13730
+ 1
13731
+ 1
13732
+ 1
13733
+ 1
13734
+ 1
13735
+ 1
13736
+ 1
13737
+ 1
13738
+ 1
13739
+ 1
13740
+ 1
13741
+ 1
13742
+ 1
13743
+ 1
13744
+ 1
13745
+ 1
13746
+ 1
13747
+ 1
13748
+ 1
13749
+ 1
13750
+ 1
13751
+ 1
13752
+ 1
13753
+ 1
13754
+ 1
13755
+ 1
13756
+ 1
13757
+ 1
13758
+ 1
13759
+ 1
13760
+ 1
13761
+ 1
13762
+ 1
13763
+ 1
13764
+ 1
13765
+ 1
13766
+ 1
13767
+ 1
13768
+ 1
13769
+ 1
13770
+ 1
13771
+ 1
13772
+ 1
13773
+ 1
13774
+ 1
13775
+ 1
13776
+ 1
13777
+ 1
13778
+ 1
13779
+ 1
13780
+ 1
13781
+ 1
13782
+ 1
13783
+ 1
13784
+ 1
13785
+ 1
13786
+ 1
13787
+ 1
13788
+ 1
13789
+ 1
13790
+ 1
13791
+ 1
13792
+ 1
13793
+ 1
13794
+ 1
13795
+ 1
13796
+ 1
13797
+ 1
13798
+ 1
13799
+ 1
13800
+ 1
13801
+ 1
13802
+ 1
13803
+ 1
13804
+ 1
13805
+ 1
13806
+ 1
13807
+ 1
13808
+ 1
13809
+ 1
13810
+ 1
13811
+ 1
13812
+ 1
13813
+ 1
13814
+ 1
13815
+ 1
13816
+ 1
13817
+ 1
13818
+ 1
13819
+ 1
13820
+ 1
13821
+ 1
13822
+ 1
13823
+ 1
13824
+ 1
13825
+ 1
13826
+ 1
13827
+ 1
13828
+ 1
13829
+ 1
13830
+ 1
13831
+ 1
13832
+ 1
13833
+ 1
13834
+ 1
13835
+ 1
13836
+ 1
13837
+ 1
13838
+ 1
13839
+ 1
13840
+ 1
13841
+ 1
13842
+ 1
13843
+ 1
13844
+ 1
13845
+ 1
13846
+ 1
13847
+ 1
13848
+ 1
13849
+ 1
13850
+ 1
13851
+ 1
13852
+ 1
13853
+ 1
13854
+ 1
13855
+ 1
13856
+ 1
13857
+ 1
13858
+ 1
13859
+ 1
13860
+ 1
13861
+ 1
13862
+ 1
13863
+ 1
13864
+ 1
13865
+ 1
13866
+ 1
13867
+ 1
13868
+ 1
13869
+ 1
13870
+ 1
13871
+ 1
13872
+ 1
13873
+ 1
13874
+ 1
13875
+ 1
13876
+ 1
13877
+ 1
13878
+ 1
13879
+ 1
13880
+ 1
13881
+ 1
13882
+ 1
13883
+ 1
13884
+ 1
13885
+ 1
13886
+ 1
13887
+ 1
13888
+ 1
13889
+ 1
13890
+ 1
13891
+ 1
13892
+ 1
13893
+ 1
13894
+ 1
13895
+ 1
13896
+ 1
13897
+ 1
13898
+ 1
13899
+ 1
13900
+ 1
13901
+ 1
13902
+ 1
13903
+ 1
13904
+ 1
13905
+ 1
13906
+ 1
13907
+ 1
13908
+ 1
13909
+ 1
13910
+ 1
13911
+ 1
13912
+ 1
13913
+ 1
13914
+ 1
13915
+ 1
13916
+ 1
13917
+ 1
13918
+ 1
13919
+ 1
13920
+ 1
13921
+ 1
13922
+ 1
13923
+ 1
13924
+ 1
13925
+ 1
13926
+ 1
13927
+ 1
13928
+ 1
13929
+ 1
13930
+ 1
13931
+ 1
13932
+ 1
13933
+ 1
13934
+ 1
13935
+ 1
13936
+ 1
13937
+ 1
13938
+ 1
13939
+ 1
13940
+ 1
13941
+ 1
13942
+ 1
13943
+ 1
13944
+ 1
13945
+ 1
13946
+ 1
13947
+ 1
13948
+ 1
13949
+ 1
13950
+ 1
13951
+ 1
13952
+ 1
13953
+ 1
13954
+ 1
13955
+ 1
13956
+ 1
13957
+ 1
13958
+ 1
13959
+ 1
13960
+ 1
13961
+ 1
13962
+ 1
13963
+ 1
13964
+ 1
13965
+ 1
13966
+ 1
13967
+ 1
13968
+ 1
13969
+ 1
13970
+ 1
13971
+ 1
13972
+ 1
13973
+ 1
13974
+ 1
13975
+ 1
13976
+ 1
13977
+ 1
13978
+ 1
13979
+ 1
13980
+ 1
13981
+ 1
13982
+ 1
13983
+ 1
13984
+ 1
13985
+ 1
13986
+ 1
13987
+ 1
13988
+ 1
13989
+ 1
13990
+ 1
13991
+ 1
13992
+ 1
13993
+ 1
13994
+ 1
13995
+ 1
13996
+ 1
13997
+ 1
13998
+ 1
13999
+ 1
14000
+ 1
14001
+ 1
14002
+ 1
14003
+ 1
14004
+ 1
14005
+ 1
14006
+ 1
14007
+ 1
14008
+ 1
14009
+ 1
14010
+ 1
14011
+ 1
14012
+ 1
14013
+ 1
14014
+ 1
14015
+ 1
14016
+ 1
14017
+ 1
14018
+ 1
14019
+ 1
14020
+ 1
14021
+ 1
14022
+ 1
14023
+ 1
14024
+ 1
14025
+ 1
14026
+ 1
14027
+ 1
14028
+ 1
14029
+ 1
14030
+ 1
14031
+ 1
14032
+ 1
14033
+ 1
14034
+ 1
14035
+ 1
14036
+ 1
14037
+ 1
14038
+ 1
14039
+ 1
14040
+ 1
14041
+ 1
14042
+ 1
14043
+ 1
14044
+ 1
14045
+ 1
14046
+ 1
14047
+ 1
14048
+ 1
14049
+ 1
14050
+ 1
14051
+ 1
14052
+ 1
14053
+ 1
14054
+ 1
14055
+ 1
14056
+ 1
14057
+ 1
14058
+ 1
14059
+ 1
14060
+ 1
14061
+ 1
14062
+ 1
14063
+ 1
14064
+ 1
14065
+ 1
14066
+ 1
14067
+ 1
14068
+ 1
14069
+ 1
14070
+ 1
14071
+ 1
14072
+ 1
14073
+ 1
14074
+ 1
14075
+ 1
14076
+ 1
14077
+ 1
14078
+ 1
14079
+ 1
14080
+ 1
14081
+ 1
14082
+ 1
14083
+ 1
14084
+ 1
14085
+ 1
14086
+ 1
14087
+ 1
14088
+ 1
14089
+ 1
14090
+ 1
14091
+ 1
14092
+ 1
14093
+ 1
14094
+ 1
14095
+ 1
14096
+ 1
14097
+ 1
14098
+ 1
14099
+ 1
14100
+ 1
14101
+ 1
14102
+ 1
14103
+ 1
14104
+ 1
14105
+ 1
14106
+ 1
14107
+ 1
14108
+ 1
14109
+ 1
14110
+ 1
14111
+ 1
14112
+ 1
14113
+ 1
14114
+ 1
14115
+ 1
14116
+ 1
14117
+ 1
14118
+ 1
14119
+ 1
14120
+ 1
14121
+ 1
14122
+ 1
14123
+ 1
14124
+ 1
14125
+ 1
14126
+ 1
14127
+ 1
14128
+ 1
14129
+ 1
14130
+ 1
14131
+ 1
14132
+ 1
14133
+ 1
14134
+ 1
14135
+ 1
14136
+ 1
14137
+ 1
14138
+ 1
14139
+ 1
14140
+ 1
14141
+ 1
14142
+ 1
14143
+ 1
14144
+ 1
14145
+ 1
14146
+ 1
14147
+ 1
14148
+ 1
14149
+ 1
14150
+ 1
14151
+ 1
14152
+ 1
14153
+ 1
14154
+ 1
14155
+ 1
14156
+ 1
14157
+ 1
14158
+ 1
14159
+ 1
14160
+ 1
14161
+ 1
14162
+ 1
14163
+ 1
14164
+ 1
14165
+ 1
14166
+ 1
14167
+ 1
14168
+ 1
14169
+ 1
14170
+ 1
14171
+ 1
14172
+ 1
14173
+ 1
14174
+ 1
14175
+ 1
14176
+ 1
14177
+ 1
14178
+ 1
14179
+ 1
14180
+ 1
14181
+ 1
14182
+ 1
14183
+ 1
14184
+ 1
14185
+ 1
14186
+ 1
14187
+ 1
14188
+ 1
14189
+ 1
14190
+ 1
14191
+ 1
14192
+ 1
14193
+ 1
14194
+ 1
14195
+ 1
14196
+ 1
14197
+ 1
14198
+ 1
14199
+ 1
14200
+ 1
14201
+ 1
14202
+ 1
14203
+ 1
14204
+ 1
14205
+ 1
14206
+ 1
14207
+ 1
14208
+ 1
14209
+ 1
14210
+ 1
14211
+ 1
14212
+ 1
14213
+ 1
14214
+ 1
14215
+ 1
14216
+ 1
14217
+ 1
14218
+ 1
14219
+ 1
14220
+ 1
14221
+ 1
14222
+ 1
14223
+ 1
14224
+ 1
14225
+ 1
14226
+ 1
14227
+ 1
14228
+ 1
14229
+ 1
14230
+ 1
14231
+ 1
14232
+ 1
14233
+ 1
14234
+ 1
14235
+ 1
14236
+ 1
14237
+ 1
14238
+ 1
14239
+ 1
14240
+ 1
14241
+ 1
14242
+ 1
14243
+ 1
14244
+ 1
14245
+ 1
14246
+ 1
14247
+ 1
14248
+ 1
14249
+ 1
14250
+ 1
14251
+ 1
14252
+ 1
14253
+ 1
14254
+ 1
14255
+ 1
14256
+ 1
14257
+ 1
14258
+ 1
14259
+ 1
14260
+ 1
14261
+ 1
14262
+ 1
14263
+ 1
14264
+ 1
14265
+ 1
14266
+ 1
14267
+ 1
14268
+ 1
14269
+ 1
14270
+ 1
14271
+ 1
14272
+ 1
14273
+ 1
14274
+ 1
14275
+ 1
14276
+ 1
14277
+ 1
14278
+ 1
14279
+ 1
14280
+ 1
14281
+ 1
14282
+ 1
14283
+ 1
14284
+ 1
14285
+ 1
14286
+ 1
14287
+ 1
14288
+ 1
14289
+ 1
14290
+ 1
14291
+ 1
14292
+ 1
14293
+ 1
14294
+ 1
14295
+ 1
14296
+ 1
14297
+ 1
14298
+ 1
14299
+ 1
14300
+ 1
14301
+ 1
14302
+ 1
14303
+ 1
14304
+ 1
14305
+ 1
14306
+ 1
14307
+ 1
14308
+ 1
14309
+ 1
14310
+ 1
14311
+ 1
14312
+ 1
14313
+ 1
14314
+ 1
14315
+ 1
14316
+ 1
14317
+ 1
14318
+ 1
14319
+ 1
14320
+ 1
14321
+ 1
14322
+ 1
14323
+ 1
14324
+ 1
14325
+ 1
14326
+ 1
14327
+ 1
14328
+ 1
14329
+ 1
14330
+ 1
14331
+ 1
14332
+ 1
14333
+ 1
14334
+ 1
14335
+ 1
14336
+ 1
14337
+ 1
14338
+ 1
14339
+ 1
14340
+ 1
14341
+ 1
14342
+ 1
14343
+ 1
14344
+ 1
14345
+ 1
14346
+ 1
14347
+ 1
14348
+ 1
14349
+ 1
14350
+ 1
14351
+ 1
14352
+ 1
14353
+ 1
14354
+ 1
14355
+ 1
14356
+ 1
14357
+ 1
14358
+ 1
14359
+ 1
14360
+ 1
14361
+ 1
14362
+ 1
14363
+ 1
14364
+ 1
14365
+ 1
14366
+ 1
14367
+ 1
14368
+ 1
14369
+ 1
14370
+ 1
14371
+ 1
14372
+ 1
14373
+ 1
14374
+ 1
14375
+ 1
14376
+ 1
14377
+ 1
14378
+ 1
14379
+ 1
14380
+ 1
14381
+ 1
14382
+ 1
14383
+ 1
14384
+ 1
14385
+ 1
14386
+ 1
14387
+ 1
14388
+ 1
14389
+ 1
14390
+ 1
14391
+ 1
14392
+ 1
14393
+ 1
14394
+ 1
14395
+ 1
14396
+ 1
14397
+ 1
14398
+ 1
14399
+ 1
14400
+ 1
14401
+ 1
14402
+ 1
14403
+ 1
14404
+ 1
14405
+ 1
14406
+ 1
14407
+ 1
14408
+ 1
14409
+ 1
14410
+ 1
14411
+ 1
14412
+ 1
14413
+ 1
14414
+ 1
14415
+ 1
14416
+ 1
14417
+ 1
14418
+ 1
14419
+ 1
14420
+ 1
14421
+ 1
14422
+ 1
14423
+ 1
14424
+ 1
14425
+ 1
14426
+ 1
14427
+ 1
14428
+ 1
14429
+ 1
14430
+ 1
14431
+ 1
14432
+ 1
14433
+ 1
14434
+ 1
14435
+ 1
14436
+ 1
14437
+ 1
14438
+ 1
14439
+ 1
14440
+ 1
14441
+ 1
14442
+ 1
14443
+ 1
14444
+ 1
14445
+ 1
14446
+ 1
14447
+ 1
14448
+ 1
14449
+ 1
14450
+ 1
14451
+ 1
14452
+ 1
14453
+ 1
14454
+ 1
14455
+ 1
14456
+ 1
14457
+ 1
14458
+ 1
14459
+ 1
14460
+ 1
14461
+ 1
14462
+ 1
14463
+ 1
14464
+ 1
14465
+ 1
14466
+ 1
14467
+ 1
14468
+ 1
14469
+ 1
14470
+ 1
14471
+ 1
14472
+ 1
14473
+ 1
14474
+ 1
14475
+ 1
14476
+ 1
14477
+ 1
14478
+ 1
14479
+ 1
14480
+ 1
14481
+ 1
14482
+ 1
14483
+ 1
14484
+ 1
14485
+ 1
14486
+ 1
14487
+ 1
14488
+ 1
14489
+ 1
14490
+ 1
14491
+ 1
14492
+ 1
14493
+ 1
14494
+ 1
14495
+ 1
14496
+ 1
14497
+ 1
14498
+ 1
14499
+ 1
14500
+ 2
14501
+ 1
14502
+ 1
14503
+ 1
14504
+ 1
14505
+ 1
14506
+ 1
14507
+ 1
14508
+ 1
14509
+ 1
14510
+ 1
14511
+ 1
14512
+ 1
14513
+ 1
14514
+ 1
14515
+ 1
14516
+ 1
14517
+ 1
14518
+ 2
14519
+ 1
14520
+ 1
14521
+ 1
14522
+ 1
14523
+ 1
14524
+ 1
14525
+ 1
14526
+ 1
14527
+ 1
14528
+ 1
14529
+ 1
14530
+ 1
14531
+ 1
14532
+ 1
14533
+ 1
14534
+ 1
14535
+ 1
14536
+ 1
14537
+ 1
14538
+ 1
14539
+ 1
14540
+ 1
14541
+ 1
14542
+ 1
14543
+ 1
14544
+ 1
14545
+ 1
14546
+ 1
14547
+ 1
14548
+ 1
14549
+ 1
14550
+ 1
14551
+ 1
14552
+ 1
14553
+ 1
14554
+ 1
14555
+ 1
14556
+ 1
14557
+ 1
14558
+ 1
14559
+ 1
14560
+ 1
14561
+ 1
14562
+ 1
14563
+ 1
14564
+ 1
14565
+ 1
14566
+ 1
14567
+ 1
14568
+ 1
14569
+ 1
14570
+ 1
14571
+ 1
14572
+ 1
14573
+ 1
14574
+ 1
14575
+ 1
14576
+ 1
14577
+ 1
14578
+ 1
14579
+ 1
14580
+ 1
14581
+ 1
14582
+ 1
14583
+ 1
14584
+ 1
14585
+ 1
14586
+ 1
14587
+ 1
14588
+ 1
14589
+ 1
14590
+ 1
14591
+ 1
14592
+ 1
14593
+ 1
14594
+ 1
14595
+ 1
14596
+ 1
14597
+ 1
14598
+ 1
14599
+ 1
14600
+ 1
14601
+ 1
14602
+ 1
14603
+ 1
14604
+ 1
14605
+ 1
14606
+ 1
14607
+ 1
14608
+ 1
14609
+ 1
14610
+ 1
14611
+ 1
14612
+ 1
14613
+ 1
14614
+ 1
14615
+ 1
14616
+ 1
14617
+ 1
14618
+ 1
14619
+ 1
14620
+ 1
14621
+ 1
14622
+ 1
14623
+ 1
14624
+ 1
14625
+ 1
14626
+ 1
14627
+ 1
14628
+ 1
14629
+ 1
14630
+ 1
14631
+ 1
14632
+ 1
14633
+ 1
14634
+ 1
14635
+ 1
14636
+ 1
14637
+ 1
14638
+ 1
14639
+ 1
14640
+ 1
14641
+ 1
14642
+ 1
14643
+ 1
14644
+ 1
14645
+ 1
14646
+ 1
14647
+ 1
14648
+ 1
14649
+ 1
14650
+ 1
14651
+ 1
14652
+ 1
14653
+ 1
14654
+ 1
14655
+ 1
14656
+ 1
14657
+ 1
14658
+ 1
14659
+ 1
14660
+ 1
14661
+ 1
14662
+ 1
14663
+ 1
14664
+ 1
14665
+ 1
14666
+ 1
14667
+ 1
14668
+ 1
14669
+ 1
14670
+ 1
14671
+ 1
14672
+ 1
14673
+ 1
14674
+ 1
14675
+ 1
14676
+ 1
14677
+ 1
14678
+ 1
14679
+ 1
14680
+ 1
14681
+ 1
14682
+ 1
14683
+ 1
14684
+ 1
14685
+ 1
14686
+ 1
14687
+ 1
14688
+ 1
14689
+ 1
14690
+ 1
14691
+ 1
14692
+ 1
14693
+ 1
14694
+ 1
14695
+ 1
14696
+ 1
14697
+ 1
14698
+ 1
14699
+ 1
14700
+ 1
14701
+ 1
14702
+ 1
14703
+ 1
14704
+ 1
14705
+ 1
14706
+ 1
14707
+ 1
14708
+ 1
14709
+ 1
14710
+ 1
14711
+ 1
14712
+ 1
14713
+ 1
14714
+ 1
14715
+ 1
14716
+ 1
14717
+ 1
14718
+ 1
14719
+ 1
14720
+ 1
14721
+ 1
14722
+ 1
14723
+ 1
14724
+ 1
14725
+ 1
14726
+ 1
14727
+ 1
14728
+ 1
14729
+ 1
14730
+ 1
14731
+ 1
14732
+ 1
14733
+ 1
14734
+ 1
14735
+ 1
14736
+ 1
14737
+ 1
14738
+ 1
14739
+ 1
14740
+ 1
14741
+ 1
14742
+ 1
14743
+ 1
14744
+ 1
14745
+ 1
14746
+ 1
14747
+ 1
14748
+ 1
14749
+ 1
14750
+ 1
14751
+ 1
14752
+ 1
14753
+ 1
14754
+ 1
14755
+ 1
14756
+ 1
14757
+ 1
14758
+ 1
14759
+ 1
14760
+ 1
14761
+ 1
14762
+ 1
14763
+ 1
14764
+ 1
14765
+ 1
14766
+ 1
14767
+ 1
14768
+ 1
14769
+ 1
14770
+ 1
14771
+ 1
14772
+ 1
14773
+ 1
14774
+ 1
14775
+ 1
14776
+ 2
14777
+ 1
14778
+ 1
14779
+ 1
14780
+ 1
14781
+ 1
14782
+ 1
14783
+ 1
14784
+ 1
14785
+ 1
14786
+ 1
14787
+ 1
14788
+ 1
14789
+ 1
14790
+ 1
14791
+ 1
14792
+ 1
14793
+ 1
14794
+ 1
14795
+ 1
14796
+ 1
14797
+ 1
14798
+ 1
14799
+ 1
14800
+ 1
14801
+ 1
14802
+ 1
14803
+ 1
14804
+ 1
14805
+ 1
14806
+ 1
14807
+ 1
14808
+ 1
14809
+ 1
14810
+ 1
14811
+ 1
14812
+ 1
14813
+ 1
14814
+ 1
14815
+ 1
14816
+ 1
14817
+ 1
14818
+ 1
14819
+ 1
14820
+ 1
14821
+ 1
14822
+ 1
14823
+ 1
14824
+ 1
14825
+ 1
14826
+ 1
14827
+ 1
14828
+ 1
14829
+ 1
14830
+ 1
14831
+ 1
14832
+ 1
14833
+ 1
14834
+ 1
14835
+ 1
14836
+ 1
14837
+ 1
14838
+ 1
14839
+ 1
14840
+ 1
14841
+ 1
14842
+ 1
14843
+ 1
14844
+ 1
14845
+ 1
14846
+ 1
14847
+ 1
14848
+ 1
14849
+ 1
14850
+ 1
14851
+ 1
14852
+ 1
14853
+ 1
14854
+ 1
14855
+ 1
14856
+ 1
14857
+ 1
14858
+ 1
14859
+ 1
14860
+ 1
14861
+ 1
14862
+ 1
14863
+ 1
14864
+ 1
14865
+ 1
14866
+ 1
14867
+ 1
14868
+ 1
14869
+ 1
14870
+ 1
14871
+ 1
14872
+ 1
14873
+ 1
14874
+ 1
14875
+ 1
14876
+ 1
14877
+ 1
14878
+ 1
14879
+ 1
14880
+ 1
14881
+ 1
14882
+ 1
14883
+ 1
14884
+ 1
14885
+ 1
14886
+ 1
14887
+ 1
14888
+ 1
14889
+ 1
14890
+ 1
14891
+ 1
14892
+ 1
14893
+ 1
14894
+ 1
14895
+ 1
14896
+ 1
14897
+ 1
14898
+ 1
14899
+ 1
14900
+ 1
14901
+ 1
14902
+ 1
14903
+ 1
14904
+ 1
14905
+ 1
14906
+ 1
14907
+ 1
14908
+ 1
14909
+ 1
14910
+ 1
14911
+ 1
14912
+ 1
14913
+ 1
14914
+ 1
14915
+ 1
14916
+ 1
14917
+ 1
14918
+ 1
14919
+ 1
14920
+ 1
14921
+ 1
14922
+ 1
14923
+ 1
14924
+ 1
14925
+ 1
14926
+ 1
14927
+ 1
14928
+ 1
14929
+ 1
14930
+ 1
14931
+ 1
14932
+ 1
14933
+ 1
14934
+ 1
14935
+ 1
14936
+ 1
14937
+ 1
14938
+ 1
14939
+ 1
14940
+ 1
14941
+ 1
14942
+ 1
14943
+ 1
14944
+ 1
14945
+ 1
14946
+ 1
14947
+ 1
14948
+ 1
14949
+ 1
14950
+ 1
14951
+ 1
14952
+ 1
14953
+ 1
14954
+ 1
14955
+ 1
14956
+ 1
14957
+ 1
14958
+ 1
14959
+ 1
14960
+ 1
14961
+ 1
14962
+ 1
14963
+ 1
14964
+ 1
14965
+ 1
14966
+ 1
14967
+ 1
14968
+ 1
14969
+ 1
14970
+ 1
14971
+ 1
14972
+ 1
14973
+ 1
14974
+ 1
14975
+ 1
14976
+ 1
14977
+ 1
14978
+ 1
14979
+ 1
14980
+ 1
14981
+ 1
14982
+ 1
14983
+ 1
14984
+ 1
14985
+ 1
14986
+ 1
14987
+ 1
14988
+ 1
14989
+ 1
14990
+ 1
14991
+ 1
14992
+ 1
14993
+ 1
14994
+ 1
14995
+ 1
14996
+ 1
14997
+ 1
14998
+ 1
14999
+ 1
15000
+ 1
15001
+ 1
15002
+ 1
15003
+ 1
15004
+ 1
15005
+ 1
15006
+ 1
15007
+ 1
15008
+ 1
15009
+ 1
15010
+ 1
15011
+ 1
15012
+ 1
15013
+ 1
15014
+ 1
15015
+ 1
15016
+ 1
15017
+ 1
15018
+ 1
15019
+ 1
15020
+ 1
15021
+ 1
15022
+ 1
15023
+ 1
15024
+ 1
15025
+ 1
15026
+ 1
15027
+ 1
15028
+ 1
15029
+ 1
15030
+ 1
15031
+ 1
15032
+ 1
15033
+ 1
15034
+ 1
15035
+ 1
15036
+ 1
15037
+ 1
15038
+ 1
15039
+ 1
15040
+ 1
15041
+ 1
15042
+ 1
15043
+ 1
15044
+ 1
15045
+ 1
15046
+ 1
15047
+ 1
15048
+ 1
15049
+ 1
15050
+ 1
15051
+ 1
15052
+ 1
15053
+ 1
15054
+ 1
15055
+ 1
15056
+ 1
15057
+ 1
15058
+ 1
15059
+ 1
15060
+ 1
15061
+ 1
15062
+ 1
15063
+ 1
15064
+ 1
15065
+ 1
15066
+ 1
15067
+ 1
15068
+ 1
15069
+ 1
15070
+ 1
15071
+ 1
15072
+ 1
15073
+ 1
15074
+ 1
15075
+ 1
15076
+ 1
15077
+ 1
15078
+ 1
15079
+ 1
15080
+ 1
15081
+ 1
15082
+ 1
15083
+ 1
15084
+ 1
15085
+ 1
15086
+ 1
15087
+ 1
15088
+ 1
15089
+ 1
15090
+ 1
15091
+ 1
15092
+ 1
15093
+ 1
15094
+ 1
15095
+ 1
15096
+ 1
15097
+ 1
15098
+ 1
15099
+ 1
15100
+ 1
15101
+ 1
15102
+ 1
15103
+ 1
15104
+ 1
15105
+ 1
15106
+ 1
15107
+ 1
15108
+ 1
15109
+ 1
15110
+ 1
15111
+ 1
15112
+ 1
15113
+ 1
15114
+ 1
15115
+ 1
15116
+ 1
15117
+ 1
15118
+ 1
15119
+ 1
15120
+ 1
15121
+ 1
15122
+ 1
15123
+ 1
15124
+ 1
15125
+ 1
15126
+ 1
15127
+ 1
15128
+ 1
15129
+ 1
15130
+ 1
15131
+ 1
15132
+ 1
15133
+ 1
15134
+ 1
15135
+ 1
15136
+ 1
15137
+ 1
15138
+ 1
15139
+ 1
15140
+ 1
15141
+ 1
15142
+ 1
15143
+ 1
15144
+ 1
15145
+ 1
15146
+ 1
15147
+ 1
15148
+ 1
15149
+ 1
15150
+ 1
15151
+ 1
15152
+ 1
15153
+ 1
15154
+ 1
15155
+ 1
15156
+ 1
15157
+ 1
15158
+ 1
15159
+ 1
15160
+ 1
15161
+ 1
15162
+ 1
15163
+ 1
15164
+ 1
15165
+ 1
15166
+ 1
15167
+ 1
15168
+ 1
15169
+ 1
15170
+ 1
15171
+ 1
15172
+ 1
15173
+ 1
15174
+ 1
15175
+ 1
15176
+ 1
15177
+ 1
15178
+ 1
15179
+ 1
15180
+ 1
15181
+ 1
15182
+ 1
15183
+ 1
15184
+ 1
15185
+ 1
15186
+ 1
15187
+ 1
15188
+ 1
15189
+ 1
15190
+ 1
15191
+ 1
15192
+ 1
15193
+ 1
15194
+ 1
15195
+ 1
15196
+ 1
15197
+ 1
15198
+ 1
15199
+ 1
15200
+ 1
15201
+ 1
15202
+ 1
15203
+ 1
15204
+ 1
15205
+ 1
15206
+ 1
15207
+ 1
15208
+ 1
15209
+ 1
15210
+ 1
15211
+ 1
15212
+ 1
15213
+ 1
15214
+ 1
15215
+ 1
15216
+ 1
15217
+ 1
15218
+ 1
15219
+ 1
15220
+ 1
15221
+ 1
15222
+ 1
15223
+ 1
15224
+ 1
15225
+ 1
15226
+ 1
15227
+ 1
15228
+ 1
15229
+ 1
15230
+ 1
15231
+ 1
15232
+ 1
15233
+ 1
15234
+ 1
15235
+ 1
15236
+ 1
15237
+ 1
15238
+ 1
15239
+ 1
15240
+ 1
15241
+ 1
15242
+ 1
15243
+ 1
15244
+ 1
15245
+ 1
15246
+ 1
15247
+ 1
15248
+ 1
15249
+ 1
15250
+ 1
15251
+ 1
15252
+ 1
15253
+ 1
15254
+ 1
15255
+ 1
15256
+ 1
15257
+ 1
15258
+ 1
15259
+ 1
15260
+ 1
15261
+ 1
15262
+ 1
15263
+ 1
15264
+ 1
15265
+ 1
15266
+ 1
15267
+ 1
15268
+ 1
15269
+ 1
15270
+ 1
15271
+ 1
15272
+ 1
15273
+ 1
15274
+ 1
15275
+ 1
15276
+ 1
15277
+ 1
15278
+ 1
15279
+ 1
15280
+ 1
15281
+ 1
15282
+ 1
15283
+ 1
15284
+ 1
15285
+ 1
15286
+ 1
15287
+ 1
15288
+ 1
15289
+ 1
15290
+ 1
15291
+ 1
15292
+ 1
15293
+ 1
15294
+ 1
15295
+ 1
15296
+ 1
15297
+ 1
15298
+ 2
15299
+ 1
15300
+ 1
15301
+ 1
15302
+ 1
15303
+ 1
15304
+ 1
15305
+ 1
15306
+ 1
15307
+ 1
15308
+ 1
15309
+ 1
15310
+ 1
15311
+ 1
15312
+ 1
15313
+ 1
15314
+ 1
15315
+ 1
15316
+ 1
15317
+ 1
15318
+ 1
15319
+ 1
15320
+ 1
15321
+ 1
15322
+ 1
15323
+ 1
15324
+ 1
15325
+ 1
15326
+ 1
15327
+ 1
15328
+ 1
15329
+ 1
15330
+ 1
15331
+ 1
15332
+ 1
15333
+ 1
15334
+ 1
15335
+ 1
15336
+ 1
15337
+ 1
15338
+ 1
15339
+ 1
15340
+ 1
15341
+ 1
15342
+ 1
15343
+ 1
15344
+ 1
15345
+ 1
15346
+ 1
15347
+ 1
15348
+ 1
15349
+ 1
15350
+ 1
15351
+ 1
15352
+ 1
15353
+ 1
15354
+ 1
15355
+ 1
15356
+ 1
15357
+ 1
15358
+ 1
15359
+ 1
15360
+ 1
15361
+ 1
15362
+ 1
15363
+ 1
15364
+ 1
15365
+ 1
15366
+ 1
15367
+ 1
15368
+ 1
15369
+ 1
15370
+ 1
15371
+ 1
15372
+ 1
15373
+ 1
15374
+ 1
15375
+ 1
15376
+ 1
15377
+ 1
15378
+ 1
15379
+ 1
15380
+ 1
15381
+ 1
15382
+ 1
15383
+ 1
15384
+ 1
15385
+ 1
15386
+ 1
15387
+ 1
15388
+ 1
15389
+ 1
15390
+ 1
15391
+ 1
15392
+ 1
15393
+ 1
15394
+ 1
15395
+ 1
15396
+ 1
15397
+ 1
15398
+ 1
15399
+ 1
15400
+ 1
15401
+ 1
15402
+ 1
15403
+ 1
15404
+ 1
15405
+ 1
15406
+ 1
15407
+ 1
15408
+ 1
15409
+ 1
15410
+ 1
15411
+ 1
15412
+ 1
15413
+ 1
15414
+ 1
15415
+ 1
15416
+ 1
15417
+ 1
15418
+ 1
15419
+ 1
15420
+ 1
15421
+ 1
15422
+ 1
15423
+ 1
15424
+ 1
15425
+ 1
15426
+ 1
15427
+ 1
15428
+ 1
15429
+ 1
15430
+ 1
15431
+ 1
15432
+ 1
15433
+ 1
15434
+ 1
15435
+ 1
15436
+ 1
15437
+ 1
15438
+ 1
15439
+ 1
15440
+ 1
15441
+ 1
15442
+ 1
15443
+ 1
15444
+ 1
15445
+ 1
15446
+ 1
15447
+ 1
15448
+ 1
15449
+ 1
15450
+ 1
15451
+ 1
15452
+ 1
15453
+ 1
15454
+ 1
15455
+ 1
15456
+ 1
15457
+ 1
15458
+ 1
15459
+ 1
15460
+ 1
15461
+ 1
15462
+ 1
15463
+ 1
15464
+ 1
15465
+ 1
15466
+ 1
15467
+ 1
15468
+ 1
15469
+ 1
15470
+ 1
15471
+ 1
15472
+ 1
15473
+ 1
15474
+ 1
15475
+ 1
15476
+ 1
15477
+ 1
15478
+ 1
15479
+ 1
15480
+ 1
15481
+ 1
15482
+ 1
15483
+ 1
15484
+ 1
15485
+ 1
15486
+ 1
15487
+ 1
15488
+ 1
15489
+ 1
15490
+ 1
15491
+ 1
15492
+ 1
15493
+ 1
15494
+ 1
15495
+ 1
15496
+ 1
15497
+ 1
15498
+ 1
15499
+ 1
15500
+ 1
15501
+ 1
15502
+ 1
15503
+ 1
15504
+ 1
15505
+ 1
15506
+ 1
15507
+ 1
15508
+ 1
15509
+ 1
15510
+ 1
15511
+ 1
15512
+ 1
15513
+ 1
15514
+ 1
15515
+ 1
15516
+ 1
15517
+ 1
15518
+ 1
15519
+ 1
15520
+ 1
15521
+ 1
15522
+ 1
15523
+ 1
15524
+ 1
15525
+ 1
15526
+ 1
15527
+ 1
15528
+ 1
15529
+ 1
15530
+ 1
15531
+ 1
15532
+ 1
15533
+ 1
15534
+ 1
15535
+ 1
15536
+ 1
15537
+ 1
15538
+ 1
15539
+ 1
15540
+ 1
15541
+ 1
15542
+ 1
15543
+ 1
15544
+ 1
15545
+ 1
15546
+ 1
15547
+ 1
15548
+ 1
15549
+ 1
15550
+ 1
15551
+ 1
15552
+ 1
15553
+ 1
15554
+ 1
15555
+ 1
15556
+ 1
15557
+ 2
15558
+ 1
15559
+ 1
15560
+ 1
15561
+ 1
15562
+ 1
15563
+ 1
15564
+ 1
15565
+ 1
15566
+ 1
15567
+ 1
15568
+ 1
15569
+ 1
15570
+ 1
15571
+ 1
15572
+ 1
15573
+ 1
15574
+ 1
15575
+ 1
15576
+ 1
15577
+ 1
15578
+ 1
15579
+ 1
15580
+ 1
15581
+ 1
15582
+ 1
15583
+ 1
15584
+ 1
15585
+ 1
15586
+ 1
15587
+ 1
15588
+ 1
15589
+ 1
15590
+ 1
15591
+ 1
15592
+ 1
15593
+ 1
15594
+ 1
15595
+ 1
15596
+ 1
15597
+ 1
15598
+ 1
15599
+ 1
15600
+ 1
15601
+ 1
15602
+ 1
15603
+ 1
15604
+ 1
15605
+ 1
15606
+ 1
15607
+ 1
15608
+ 1
15609
+ 1
15610
+ 1
15611
+ 1
15612
+ 1
15613
+ 1
15614
+ 1
15615
+ 1
15616
+ 1
15617
+ 1
15618
+ 1
15619
+ 1
15620
+ 1
15621
+ 1
15622
+ 1
15623
+ 1
15624
+ 1
15625
+ 1
15626
+ 1
15627
+ 1
15628
+ 1
15629
+ 1
15630
+ 1
15631
+ 1
15632
+ 1
15633
+ 1
15634
+ 1
15635
+ 1
15636
+ 1
15637
+ 1
15638
+ 1
15639
+ 1
15640
+ 1
15641
+ 1
15642
+ 1
15643
+ 1
15644
+ 1
15645
+ 1
15646
+ 1
15647
+ 1
15648
+ 1
15649
+ 1
15650
+ 1
15651
+ 1
15652
+ 1
15653
+ 1
15654
+ 1
15655
+ 1
15656
+ 1
15657
+ 1
15658
+ 1
15659
+ 1
15660
+ 1
15661
+ 1
15662
+ 1
15663
+ 1
15664
+ 1
15665
+ 1
15666
+ 1
15667
+ 1
15668
+ 1
15669
+ 1
15670
+ 1
15671
+ 1
15672
+ 1
15673
+ 1
15674
+ 1
15675
+ 1
15676
+ 1
15677
+ 1
15678
+ 1
15679
+ 1
15680
+ 1
15681
+ 1
15682
+ 1
15683
+ 1
15684
+ 1
15685
+ 1
15686
+ 1
15687
+ 1
15688
+ 1
15689
+ 1
15690
+ 1
15691
+ 1
15692
+ 1
15693
+ 1
15694
+ 1
15695
+ 1
15696
+ 1
15697
+ 1
15698
+ 1
15699
+ 1
15700
+ 1
15701
+ 1
15702
+ 1
15703
+ 1
15704
+ 1
15705
+ 1
15706
+ 1
15707
+ 1
15708
+ 1
15709
+ 1
15710
+ 1
15711
+ 1
15712
+ 1
15713
+ 1
15714
+ 1
15715
+ 1
15716
+ 1
15717
+ 1
15718
+ 1
15719
+ 1
15720
+ 1
15721
+ 1
15722
+ 1
15723
+ 1
15724
+ 1
15725
+ 1
15726
+ 1
15727
+ 1
15728
+ 1
15729
+ 1
15730
+ 1
15731
+ 1
15732
+ 1
15733
+ 1
15734
+ 1
15735
+ 1
15736
+ 1
15737
+ 1
15738
+ 1
15739
+ 1
15740
+ 1
15741
+ 1
15742
+ 1
15743
+ 1
15744
+ 1
15745
+ 1
15746
+ 1
15747
+ 1
15748
+ 1
15749
+ 1
15750
+ 1
15751
+ 1
15752
+ 1
15753
+ 1
15754
+ 1
15755
+ 1
15756
+ 1
15757
+ 1
15758
+ 1
15759
+ 1
15760
+ 1
15761
+ 1
15762
+ 1
15763
+ 1
15764
+ 1
15765
+ 1
15766
+ 1
15767
+ 1
15768
+ 1
15769
+ 1
15770
+ 1
15771
+ 1
15772
+ 1
15773
+ 1
15774
+ 1
15775
+ 1
15776
+ 1
15777
+ 1
15778
+ 1
15779
+ 1
15780
+ 1
15781
+ 1
15782
+ 1
15783
+ 1
15784
+ 1
15785
+ 1
15786
+ 1
15787
+ 1
15788
+ 1
15789
+ 1
15790
+ 1
15791
+ 1
15792
+ 1
15793
+ 1
15794
+ 1
15795
+ 1
15796
+ 1
15797
+ 1
15798
+ 1
15799
+ 1
15800
+ 1
15801
+ 1
15802
+ 1
15803
+ 1
15804
+ 1
15805
+ 1
15806
+ 1
15807
+ 1
15808
+ 1
15809
+ 1
15810
+ 1
15811
+ 1
15812
+ 1
15813
+ 1
15814
+ 1
15815
+ 1
15816
+ 1
15817
+ 1
15818
+ 1
15819
+ 1
15820
+ 1
15821
+ 1
15822
+ 1
15823
+ 1
15824
+ 1
15825
+ 1
15826
+ 1
15827
+ 1
15828
+ 1
15829
+ 1
15830
+ 1
15831
+ 1
15832
+ 1
15833
+ 1
15834
+ 1
15835
+ 2
15836
+ 1
15837
+ 1
15838
+ 1
15839
+ 1
15840
+ 2
15841
+ 1
15842
+ 1
15843
+ 1
15844
+ 1
15845
+ 1
15846
+ 1
15847
+ 1
15848
+ 1
15849
+ 1
15850
+ 1
15851
+ 1
15852
+ 1
15853
+ 1
15854
+ 1
15855
+ 1
15856
+ 1
15857
+ 1
15858
+ 1
15859
+ 1
15860
+ 1
15861
+ 1
15862
+ 1
15863
+ 1
15864
+ 1
15865
+ 1
15866
+ 1
15867
+ 1
15868
+ 1
15869
+ 1
15870
+ 1
15871
+ 1
15872
+ 1
15873
+ 1
15874
+ 1
15875
+ 1
15876
+ 1
15877
+ 1
15878
+ 1
15879
+ 1
15880
+ 1
15881
+ 1
15882
+ 1
15883
+ 1
15884
+ 1
15885
+ 1
15886
+ 1
15887
+ 1
15888
+ 1
15889
+ 1
15890
+ 1
15891
+ 1
15892
+ 1
15893
+ 1
15894
+ 1
15895
+ 1
15896
+ 1
15897
+ 1
15898
+ 1
15899
+ 1
15900
+ 1
15901
+ 1
15902
+ 1
15903
+ 1
15904
+ 1
15905
+ 1
15906
+ 1
15907
+ 1
15908
+ 1
15909
+ 1
15910
+ 1
15911
+ 1
15912
+ 1
15913
+ 1
15914
+ 1
15915
+ 1
15916
+ 1
15917
+ 1
15918
+ 1
15919
+ 1
15920
+ 1
15921
+ 1
15922
+ 1
15923
+ 1
15924
+ 1
15925
+ 1
15926
+ 1
15927
+ 1
15928
+ 1
15929
+ 1
15930
+ 1
15931
+ 1
15932
+ 1
15933
+ 1
15934
+ 1
15935
+ 1
15936
+ 1
15937
+ 1
15938
+ 1
15939
+ 1
15940
+ 1
15941
+ 1
15942
+ 1
15943
+ 1
15944
+ 1
15945
+ 1
15946
+ 1
15947
+ 1
15948
+ 1
15949
+ 1
15950
+ 1
15951
+ 1
15952
+ 1
15953
+ 1
15954
+ 1
15955
+ 1
15956
+ 1
15957
+ 1
15958
+ 1
15959
+ 1
15960
+ 1
15961
+ 1
15962
+ 1
15963
+ 1
15964
+ 1
15965
+ 1
15966
+ 1
15967
+ 1
15968
+ 1
15969
+ 1
15970
+ 1
15971
+ 1
15972
+ 1
15973
+ 1
15974
+ 1
15975
+ 1
15976
+ 1
15977
+ 1
15978
+ 1
15979
+ 1
15980
+ 1
15981
+ 1
15982
+ 1
15983
+ 1
15984
+ 1
15985
+ 1
15986
+ 1
15987
+ 1
15988
+ 1
15989
+ 1
15990
+ 1
15991
+ 1
15992
+ 1
15993
+ 1
15994
+ 1
15995
+ 1
15996
+ 1
15997
+ 1
15998
+ 1
15999
+ 1
16000
+ 1
16001
+ 1
16002
+ 1
16003
+ 1
16004
+ 1
16005
+ 1
16006
+ 1
16007
+ 1
16008
+ 1
16009
+ 1
16010
+ 1
16011
+ 1
16012
+ 1
16013
+ 1
16014
+ 1
16015
+ 1
16016
+ 1
16017
+ 1
16018
+ 1
16019
+ 1
16020
+ 1
16021
+ 1
16022
+ 1
16023
+ 1
16024
+ 1
16025
+ 1
16026
+ 1
16027
+ 1
16028
+ 1
16029
+ 1
16030
+ 1
16031
+ 1
16032
+ 1
16033
+ 1
16034
+ 1
16035
+ 1
16036
+ 1
16037
+ 1
16038
+ 1
16039
+ 1
16040
+ 1
16041
+ 1
16042
+ 1
16043
+ 1
16044
+ 1
16045
+ 1
16046
+ 1
16047
+ 1
16048
+ 1
16049
+ 1
16050
+ 1
16051
+ 1
16052
+ 1
16053
+ 1
16054
+ 1
16055
+ 1
16056
+ 1
16057
+ 1
16058
+ 1
16059
+ 1
16060
+ 1
16061
+ 1
16062
+ 1
16063
+ 1
16064
+ 1
16065
+ 1
16066
+ 1
16067
+ 1
16068
+ 1
16069
+ 1
16070
+ 1
16071
+ 1
16072
+ 1
16073
+ 1
16074
+ 1
16075
+ 1
16076
+ 1
16077
+ 1
16078
+ 1
16079
+ 1
16080
+ 1
16081
+ 1
16082
+ 1
16083
+ 1
16084
+ 1
16085
+ 1
16086
+ 1
16087
+ 1
16088
+ 1
16089
+ 1
16090
+ 1
16091
+ 1
16092
+ 1
16093
+ 1
16094
+ 1
16095
+ 1
16096
+ 1
16097
+ 1
16098
+ 1
16099
+ 1
16100
+ 1
16101
+ 1
16102
+ 1
16103
+ 1
16104
+ 1
16105
+ 1
16106
+ 1
16107
+ 1
16108
+ 1
16109
+ 1
16110
+ 1
16111
+ 1
16112
+ 1
16113
+ 1
16114
+ 1
16115
+ 1
16116
+ 1
16117
+ 1
16118
+ 1
16119
+ 1
16120
+ 1
16121
+ 1
16122
+ 1
16123
+ 1
16124
+ 1
16125
+ 1
16126
+ 1
16127
+ 2
16128
+ 2
16129
+ 1
16130
+ 1
16131
+ 1
16132
+ 1
16133
+ 1
16134
+ 1
16135
+ 1
16136
+ 1
16137
+ 1
16138
+ 1
16139
+ 1
16140
+ 1
16141
+ 1
16142
+ 1
16143
+ 1
16144
+ 1
16145
+ 1
16146
+ 1
16147
+ 1
16148
+ 1
16149
+ 1
16150
+ 1
16151
+ 1
16152
+ 1
16153
+ 1
16154
+ 1
16155
+ 1
16156
+ 1
16157
+ 1
16158
+ 1
16159
+ 1
16160
+ 1
16161
+ 1
16162
+ 1
16163
+ 1
16164
+ 1
16165
+ 1
16166
+ 1
16167
+ 1
16168
+ 1
16169
+ 1
16170
+ 1
16171
+ 1
16172
+ 1
16173
+ 1
16174
+ 1
16175
+ 1
16176
+ 1
16177
+ 1
16178
+ 1
16179
+ 1
16180
+ 1
16181
+ 1
16182
+ 1
16183
+ 1
16184
+ 1
16185
+ 1
16186
+ 1
16187
+ 1
16188
+ 1
16189
+ 1
16190
+ 1
16191
+ 1
16192
+ 1
16193
+ 1
16194
+ 1
16195
+ 1
16196
+ 1
16197
+ 1
16198
+ 1
16199
+ 1
16200
+ 1
16201
+ 1
16202
+ 1
16203
+ 1
16204
+ 1
16205
+ 1
16206
+ 1
16207
+ 1
16208
+ 1
16209
+ 1
16210
+ 1
16211
+ 1
16212
+ 1
16213
+ 1
16214
+ 1
16215
+ 1
16216
+ 1
16217
+ 1
16218
+ 1
16219
+ 1
16220
+ 1
16221
+ 1
16222
+ 1
16223
+ 1
16224
+ 1
16225
+ 1
16226
+ 1
16227
+ 1
16228
+ 1
16229
+ 1
16230
+ 1
16231
+ 1
16232
+ 1
16233
+ 1
16234
+ 1
16235
+ 1
16236
+ 1
16237
+ 1
16238
+ 1
16239
+ 1
16240
+ 1
16241
+ 1
16242
+ 1
16243
+ 1
16244
+ 1
16245
+ 1
16246
+ 1
16247
+ 1
16248
+ 1
16249
+ 1
16250
+ 1
16251
+ 1
16252
+ 1
16253
+ 1
16254
+ 1
16255
+ 1
16256
+ 1
16257
+ 1
16258
+ 1
16259
+ 1
16260
+ 1
16261
+ 1
16262
+ 1
16263
+ 1
16264
+ 1
16265
+ 1
16266
+ 1
16267
+ 1
16268
+ 1
16269
+ 1
16270
+ 1
16271
+ 1
16272
+ 1
16273
+ 1
16274
+ 1
16275
+ 1
16276
+ 1
16277
+ 1
16278
+ 1
16279
+ 1
16280
+ 1
16281
+ 1
16282
+ 1
16283
+ 1
16284
+ 1
16285
+ 1
16286
+ 1
16287
+ 1
16288
+ 1
16289
+ 1
16290
+ 1
16291
+ 1
16292
+ 1
16293
+ 1
16294
+ 1
16295
+ 1
16296
+ 1
16297
+ 1
16298
+ 1
16299
+ 1
16300
+ 1
16301
+ 1
16302
+ 1
16303
+ 1
16304
+ 1
16305
+ 1
16306
+ 1
16307
+ 1
16308
+ 1
16309
+ 1
16310
+ 1
16311
+ 1
16312
+ 1
16313
+ 1
16314
+ 1
16315
+ 1
16316
+ 1
16317
+ 1
16318
+ 1
16319
+ 1
16320
+ 1
16321
+ 1
16322
+ 1
16323
+ 1
16324
+ 1
16325
+ 1
16326
+ 1
16327
+ 1
16328
+ 1
16329
+ 1
16330
+ 1
16331
+ 1
16332
+ 1
16333
+ 1
16334
+ 1
16335
+ 1
16336
+ 1
16337
+ 1
16338
+ 1
16339
+ 1
16340
+ 1
16341
+ 1
16342
+ 1
16343
+ 1
16344
+ 1
16345
+ 1
16346
+ 1
16347
+ 1
16348
+ 1
16349
+ 1
16350
+ 1
16351
+ 1
16352
+ 1
16353
+ 1
16354
+ 1
16355
+ 1
16356
+ 1
16357
+ 1
16358
+ 1
16359
+ 1
16360
+ 1
16361
+ 1
16362
+ 1
16363
+ 1
16364
+ 1
16365
+ 1
16366
+ 1
16367
+ 1
16368
+ 1
16369
+ 1
16370
+ 1
16371
+ 1
16372
+ 1
16373
+ 1
16374
+ 1
16375
+ 1
16376
+ 1
16377
+ 1
16378
+ 1
16379
+ 1
16380
+ 1
16381
+ 1
16382
+ 1
16383
+ 1
16384
+ 1
16385
+ 1
16386
+ 1
16387
+ 2
16388
+ 1
16389
+ 1
16390
+ 1
16391
+ 1
16392
+ 1
16393
+ 1
16394
+ 1
16395
+ 1
16396
+ 1
16397
+ 1
16398
+ 1
16399
+ 1
16400
+ 1
16401
+ 1
16402
+ 1
16403
+ 1
16404
+ 1
16405
+ 1
16406
+ 1
16407
+ 1
16408
+ 1
16409
+ 1
16410
+ 1
16411
+ 1
16412
+ 1
16413
+ 1
16414
+ 1
16415
+ 1
16416
+ 1
16417
+ 1
16418
+ 1
16419
+ 1
16420
+ 1
16421
+ 1
16422
+ 1
16423
+ 1
16424
+ 1
16425
+ 1
16426
+ 1
16427
+ 1
16428
+ 1
16429
+ 1
16430
+ 1
16431
+ 1
16432
+ 1
16433
+ 1
16434
+ 1
16435
+ 1
16436
+ 1
16437
+ 1
16438
+ 1
16439
+ 1
16440
+ 1
16441
+ 1
16442
+ 1
16443
+ 1
16444
+ 1
16445
+ 1
16446
+ 1
16447
+ 1
16448
+ 1
16449
+ 1
16450
+ 1
16451
+ 1
16452
+ 1
16453
+ 1
16454
+ 1
16455
+ 1
16456
+ 1
16457
+ 1
16458
+ 1
16459
+ 1
16460
+ 1
16461
+ 1
16462
+ 1
16463
+ 1
16464
+ 1
16465
+ 1
16466
+ 1
16467
+ 1
16468
+ 1
16469
+ 1
16470
+ 1
16471
+ 1
16472
+ 1
16473
+ 1
16474
+ 1
16475
+ 1
16476
+ 1
16477
+ 1
16478
+ 1
16479
+ 1
16480
+ 1
16481
+ 1
16482
+ 1
16483
+ 1
16484
+ 1
16485
+ 1
16486
+ 1
16487
+ 1
16488
+ 1
16489
+ 1
16490
+ 1
16491
+ 1
16492
+ 1
16493
+ 1
16494
+ 1
16495
+ 1
16496
+ 1
16497
+ 1
16498
+ 1
16499
+ 1
16500
+ 1
16501
+ 1
16502
+ 1
16503
+ 1
16504
+ 1
16505
+ 1
16506
+ 1
16507
+ 1
16508
+ 1
16509
+ 1
16510
+ 1
16511
+ 1
16512
+ 1
16513
+ 1
16514
+ 1
16515
+ 1
16516
+ 1
16517
+ 1
16518
+ 1
16519
+ 1
16520
+ 1
16521
+ 1
16522
+ 1
16523
+ 1
16524
+ 1
16525
+ 1
16526
+ 1
16527
+ 1
16528
+ 1
16529
+ 1
16530
+ 1
16531
+ 1
16532
+ 1
16533
+ 1
16534
+ 1
16535
+ 1
16536
+ 1
16537
+ 1
16538
+ 1
16539
+ 1
16540
+ 1
16541
+ 1
16542
+ 1
16543
+ 1
16544
+ 1
16545
+ 1
16546
+ 1
16547
+ 1
16548
+ 1
16549
+ 1
16550
+ 1
16551
+ 1
16552
+ 1
16553
+ 1
16554
+ 1
16555
+ 1
16556
+ 1
16557
+ 1
16558
+ 1
16559
+ 1
16560
+ 1
16561
+ 1
16562
+ 1
16563
+ 1
16564
+ 1
16565
+ 1
16566
+ 1
16567
+ 1
16568
+ 1
16569
+ 1
16570
+ 1
16571
+ 1
16572
+ 1
16573
+ 1
16574
+ 1
16575
+ 1
16576
+ 1
16577
+ 1
16578
+ 1
16579
+ 1
16580
+ 1
16581
+ 1
16582
+ 1
16583
+ 1
16584
+ 1
16585
+ 1
16586
+ 1
16587
+ 1
16588
+ 1
16589
+ 1
16590
+ 1
16591
+ 1
16592
+ 1
16593
+ 1
16594
+ 1
16595
+ 1
16596
+ 1
16597
+ 1
16598
+ 1
16599
+ 1
16600
+ 1
16601
+ 1
16602
+ 1
16603
+ 1
16604
+ 1
16605
+ 1
16606
+ 1
16607
+ 1
16608
+ 1
16609
+ 1
16610
+ 1
16611
+ 1
16612
+ 1
16613
+ 1
16614
+ 1
16615
+ 1
16616
+ 1
16617
+ 1
16618
+ 1
16619
+ 1
16620
+ 1
16621
+ 1
16622
+ 1
16623
+ 1
16624
+ 1
16625
+ 1
16626
+ 1
16627
+ 1
16628
+ 1
16629
+ 1
16630
+ 1
16631
+ 1
16632
+ 1
16633
+ 1
16634
+ 2
16635
+ 1
16636
+ 1
16637
+ 1
16638
+ 1
16639
+ 1
16640
+ 1
16641
+ 1
16642
+ 1
16643
+ 1
16644
+ 1
16645
+ 2
16646
+ 1
16647
+ 1
16648
+ 1
16649
+ 1
16650
+ 1
16651
+ 1
16652
+ 2
16653
+ 1
16654
+ 1
16655
+ 1
16656
+ 1
16657
+ 1
16658
+ 1
16659
+ 1
16660
+ 1
16661
+ 1
16662
+ 1
16663
+ 1
16664
+ 1
16665
+ 1
16666
+ 1
16667
+ 1
16668
+ 1
16669
+ 1
16670
+ 1
16671
+ 1
16672
+ 1
16673
+ 1
16674
+ 1
16675
+ 1
16676
+ 1
16677
+ 1
16678
+ 1
16679
+ 1
16680
+ 1
16681
+ 1
16682
+ 1
16683
+ 1
16684
+ 1
16685
+ 1
16686
+ 1
16687
+ 1
16688
+ 1
16689
+ 1
16690
+ 1
16691
+ 1
16692
+ 1
16693
+ 1
16694
+ 1
16695
+ 1
16696
+ 1
16697
+ 1
16698
+ 1
16699
+ 1
16700
+ 1
16701
+ 1
16702
+ 1
16703
+ 1
16704
+ 1
16705
+ 1
16706
+ 1
16707
+ 1
16708
+ 1
16709
+ 1
16710
+ 1
16711
+ 1
16712
+ 1
16713
+ 1
16714
+ 1
16715
+ 1
16716
+ 1
16717
+ 1
16718
+ 1
16719
+ 1
16720
+ 1
16721
+ 1
16722
+ 1
16723
+ 1
16724
+ 1
16725
+ 1
16726
+ 1
16727
+ 1
16728
+ 1
16729
+ 1
16730
+ 1
16731
+ 1
16732
+ 1
16733
+ 1
16734
+ 1
16735
+ 1
16736
+ 1
16737
+ 1
16738
+ 1
16739
+ 1
16740
+ 1
16741
+ 1
16742
+ 1
16743
+ 1
16744
+ 1
16745
+ 1
16746
+ 1
16747
+ 1
16748
+ 1
16749
+ 1
16750
+ 1
16751
+ 1
16752
+ 1
16753
+ 1
16754
+ 1
16755
+ 1
16756
+ 1
16757
+ 1
16758
+ 1
16759
+ 1
16760
+ 1
16761
+ 1
16762
+ 1
16763
+ 1
16764
+ 1
16765
+ 1
16766
+ 1
16767
+ 1
16768
+ 1
16769
+ 1
16770
+ 1
16771
+ 1
16772
+ 1
16773
+ 1
16774
+ 1
16775
+ 1
16776
+ 1
16777
+ 1
16778
+ 1
16779
+ 1
16780
+ 1
16781
+ 1
16782
+ 1
16783
+ 1
16784
+ 1
16785
+ 1
16786
+ 1
16787
+ 1
16788
+ 1
16789
+ 1
16790
+ 1
16791
+ 1
16792
+ 1
16793
+ 1
16794
+ 1
16795
+ 1
16796
+ 1
16797
+ 1
16798
+ 1
16799
+ 1
16800
+ 1
16801
+ 1
16802
+ 1
16803
+ 1
16804
+ 1
16805
+ 1
16806
+ 1
16807
+ 1
16808
+ 1
16809
+ 1
16810
+ 1
16811
+ 1
16812
+ 1
16813
+ 1
16814
+ 1
16815
+ 1
16816
+ 1
16817
+ 1
16818
+ 1
16819
+ 1
16820
+ 1
16821
+ 1
16822
+ 1
16823
+ 1
16824
+ 1
16825
+ 1
16826
+ 1
16827
+ 1
16828
+ 1
16829
+ 1
16830
+ 1
16831
+ 1
16832
+ 1
16833
+ 1
16834
+ 1
16835
+ 1
16836
+ 1
16837
+ 1
16838
+ 1
16839
+ 1
16840
+ 1
16841
+ 1
16842
+ 1
16843
+ 1
16844
+ 1
16845
+ 1
16846
+ 1
16847
+ 1
16848
+ 1
16849
+ 1
16850
+ 1
16851
+ 1
16852
+ 1
16853
+ 1
16854
+ 1
16855
+ 1
16856
+ 1
16857
+ 1
16858
+ 1
16859
+ 1
16860
+ 1
16861
+ 1
16862
+ 1
16863
+ 1
16864
+ 1
16865
+ 1
16866
+ 1
16867
+ 1
16868
+ 1
16869
+ 1
16870
+ 1
16871
+ 1
16872
+ 1
16873
+ 1
16874
+ 1
16875
+ 1
16876
+ 1
16877
+ 1
16878
+ 1
16879
+ 1
16880
+ 1
16881
+ 1
16882
+ 1
16883
+ 1
16884
+ 1
16885
+ 1
16886
+ 1
16887
+ 1
16888
+ 1
16889
+ 1
16890
+ 1
16891
+ 1
16892
+ 1
16893
+ 1
16894
+ 1
16895
+ 1
16896
+ 1
16897
+ 1
16898
+ 1
16899
+ 1
16900
+ 1
16901
+ 1
16902
+ 1
16903
+ 1
16904
+ 1
16905
+ 1
16906
+ 1
16907
+ 1
16908
+ 1
16909
+ 1
16910
+ 1
16911
+ 1
16912
+ 1
16913
+ 1
16914
+ 1
16915
+ 1
16916
+ 1
16917
+ 1
16918
+ 1
16919
+ 1
16920
+ 1
16921
+ 1
16922
+ 1
16923
+ 1
16924
+ 1
16925
+ 1
16926
+ 1
16927
+ 1
16928
+ 1
16929
+ 1
16930
+ 1
16931
+ 1
16932
+ 1
16933
+ 1
16934
+ 1
16935
+ 1
16936
+ 1
16937
+ 1
16938
+ 1
16939
+ 1
16940
+ 1
16941
+ 1
16942
+ 1
16943
+ 1
16944
+ 1
16945
+ 1
16946
+ 1
16947
+ 1
16948
+ 1
16949
+ 1
16950
+ 1
16951
+ 1
16952
+ 1
16953
+ 1
16954
+ 1
16955
+ 1
16956
+ 1
16957
+ 1
16958
+ 1
16959
+ 1
16960
+ 1
16961
+ 1
16962
+ 1
16963
+ 1
16964
+ 1
16965
+ 1
16966
+ 1
16967
+ 1
16968
+ 1
16969
+ 1
16970
+ 1
16971
+ 1
16972
+ 1
16973
+ 1
16974
+ 1
16975
+ 1
16976
+ 1
16977
+ 1
16978
+ 1
16979
+ 1
16980
+ 1
16981
+ 1
16982
+ 1
16983
+ 1
16984
+ 1
16985
+ 1
16986
+ 1
16987
+ 1
16988
+ 1
16989
+ 1
16990
+ 1
16991
+ 1
16992
+ 1
16993
+ 1
16994
+ 1
16995
+ 1
16996
+ 1
16997
+ 1
16998
+ 1
16999
+ 1
17000
+ 1
17001
+ 1
17002
+ 1
17003
+ 1
17004
+ 1
17005
+ 1
17006
+ 1
17007
+ 1
17008
+ 1
17009
+ 1
17010
+ 1
17011
+ 1
17012
+ 1
17013
+ 1
17014
+ 1
17015
+ 1
17016
+ 1
17017
+ 1
17018
+ 1
17019
+ 1
17020
+ 1
17021
+ 1
17022
+ 1
17023
+ 1
17024
+ 1
17025
+ 1
17026
+ 1
17027
+ 1
17028
+ 1
17029
+ 1
17030
+ 1
17031
+ 1
17032
+ 1
17033
+ 1
17034
+ 1
17035
+ 1
17036
+ 1
17037
+ 1
17038
+ 1
17039
+ 1
17040
+ 1
17041
+ 1
17042
+ 1
17043
+ 1
17044
+ 1
17045
+ 1
17046
+ 1
17047
+ 1
17048
+ 1
17049
+ 1
17050
+ 1
17051
+ 1
17052
+ 1
17053
+ 1
17054
+ 1
17055
+ 1
17056
+ 1
17057
+ 1
17058
+ 1
17059
+ 1
17060
+ 1
17061
+ 1
17062
+ 1
17063
+ 1
17064
+ 1
17065
+ 1
17066
+ 1
17067
+ 1
17068
+ 1
17069
+ 1
17070
+ 1
17071
+ 1
17072
+ 1
17073
+ 1
17074
+ 1
17075
+ 1
17076
+ 1
17077
+ 1
17078
+ 1
17079
+ 1
17080
+ 1
17081
+ 1
17082
+ 1
17083
+ 1
17084
+ 1
17085
+ 1
17086
+ 1
17087
+ 1
17088
+ 1
17089
+ 1
17090
+ 1
17091
+ 1
17092
+ 1
17093
+ 1
17094
+ 1
17095
+ 1
17096
+ 1
17097
+ 1
17098
+ 1
17099
+ 1
17100
+ 1
17101
+ 1
17102
+ 1
17103
+ 1
17104
+ 1
17105
+ 1
17106
+ 1
17107
+ 1
17108
+ 1
17109
+ 1
17110
+ 1
17111
+ 1
17112
+ 1
17113
+ 1
17114
+ 1
17115
+ 1
17116
+ 1
17117
+ 1
17118
+ 1
17119
+ 1
17120
+ 1
17121
+ 1
17122
+ 1
17123
+ 1
17124
+ 1
17125
+ 1
17126
+ 1
17127
+ 1
17128
+ 1
17129
+ 1
17130
+ 1
17131
+ 1
17132
+ 1
17133
+ 1
17134
+ 1
17135
+ 1
17136
+ 1
17137
+ 1
17138
+ 1
17139
+ 1
17140
+ 1
17141
+ 1
17142
+ 1
17143
+ 1
17144
+ 1
17145
+ 1
17146
+ 1
17147
+ 1
17148
+ 1
17149
+ 1
17150
+ 1
17151
+ 1
17152
+ 1
17153
+ 1
17154
+ 1
17155
+ 1
17156
+ 1
17157
+ 1
17158
+ 1
17159
+ 1
17160
+ 1
17161
+ 1
17162
+ 1
17163
+ 1
17164
+ 1
17165
+ 1
17166
+ 1
17167
+ 2
17168
+ 1
17169
+ 1
17170
+ 2
17171
+ 1
17172
+ 1
17173
+ 1
17174
+ 1
17175
+ 1
17176
+ 1
17177
+ 1
17178
+ 1
17179
+ 1
17180
+ 1
17181
+ 1
17182
+ 1
17183
+ 1
17184
+ 1
17185
+ 1
17186
+ 1
17187
+ 1
17188
+ 1
17189
+ 1
17190
+ 1
17191
+ 1
17192
+ 1
17193
+ 1
17194
+ 1
17195
+ 1
17196
+ 1
17197
+ 1
17198
+ 1
17199
+ 1
17200
+ 1
17201
+ 1
17202
+ 1
17203
+ 1
17204
+ 1
17205
+ 1
17206
+ 1
17207
+ 1
17208
+ 1
17209
+ 1
17210
+ 1
17211
+ 1
17212
+ 1
17213
+ 1
17214
+ 1
17215
+ 1
17216
+ 1
17217
+ 1
17218
+ 1
17219
+ 1
17220
+ 1
17221
+ 1
17222
+ 1
17223
+ 1
17224
+ 1
17225
+ 1
17226
+ 1
17227
+ 1
17228
+ 1
17229
+ 1
17230
+ 1
17231
+ 1
17232
+ 1
17233
+ 1
17234
+ 1
17235
+ 1
17236
+ 1
17237
+ 1
17238
+ 1
17239
+ 1
17240
+ 1
17241
+ 1
17242
+ 1
17243
+ 1
17244
+ 1
17245
+ 1
17246
+ 1
17247
+ 1
17248
+ 1
17249
+ 1
17250
+ 1
17251
+ 1
17252
+ 1
17253
+ 1
17254
+ 1
17255
+ 1
17256
+ 1
17257
+ 1
17258
+ 1
17259
+ 1
17260
+ 1
17261
+ 1
17262
+ 1
17263
+ 1
17264
+ 1
17265
+ 1
17266
+ 1
17267
+ 1
17268
+ 1
17269
+ 1
17270
+ 1
17271
+ 1
17272
+ 1
17273
+ 1
17274
+ 1
17275
+ 1
17276
+ 1
17277
+ 1
17278
+ 1
17279
+ 1
17280
+ 1
17281
+ 1
17282
+ 1
17283
+ 1
17284
+ 1
17285
+ 1
17286
+ 1
17287
+ 1
17288
+ 1
17289
+ 1
17290
+ 1
17291
+ 1
17292
+ 1
17293
+ 1
17294
+ 1
17295
+ 1
17296
+ 1
17297
+ 1
17298
+ 1
17299
+ 1
17300
+ 1
17301
+ 1
17302
+ 1
17303
+ 1
17304
+ 1
17305
+ 1
17306
+ 1
17307
+ 1
17308
+ 1
17309
+ 1
17310
+ 1
17311
+ 1
17312
+ 1
17313
+ 1
17314
+ 1
17315
+ 1
17316
+ 1
17317
+ 1
17318
+ 1
17319
+ 1
17320
+ 1
17321
+ 1
17322
+ 1
17323
+ 1
17324
+ 1
17325
+ 1
17326
+ 1
17327
+ 1
17328
+ 1
17329
+ 1
17330
+ 1
17331
+ 1
17332
+ 1
17333
+ 1
17334
+ 1
17335
+ 1
17336
+ 1
17337
+ 1
17338
+ 1
17339
+ 1
17340
+ 1
17341
+ 1
17342
+ 1
17343
+ 1
17344
+ 1
17345
+ 1
17346
+ 1
17347
+ 1
17348
+ 1
17349
+ 1
17350
+ 1
17351
+ 1
17352
+ 1
17353
+ 1
17354
+ 1
17355
+ 1
17356
+ 1
17357
+ 1
17358
+ 1
17359
+ 1
17360
+ 1
17361
+ 1
17362
+ 1
17363
+ 1
17364
+ 1
17365
+ 1
17366
+ 1
17367
+ 1
17368
+ 1
17369
+ 1
17370
+ 1
17371
+ 1
17372
+ 1
17373
+ 1
17374
+ 1
17375
+ 1
17376
+ 1
17377
+ 1
17378
+ 1
17379
+ 1
17380
+ 1
17381
+ 1
17382
+ 1
17383
+ 1
17384
+ 1
17385
+ 1
17386
+ 1
17387
+ 1
17388
+ 1
17389
+ 1
17390
+ 1
17391
+ 1
17392
+ 1
17393
+ 1
17394
+ 1
17395
+ 1
17396
+ 1
17397
+ 1
17398
+ 1
17399
+ 1
17400
+ 1
17401
+ 1
17402
+ 1
17403
+ 1
17404
+ 1
17405
+ 1
17406
+ 1
17407
+ 1
17408
+ 1
17409
+ 1
17410
+ 1
17411
+ 1
17412
+ 1
17413
+ 1
17414
+ 1
17415
+ 1
17416
+ 1
17417
+ 1
17418
+ 1
17419
+ 1
17420
+ 1
17421
+ 1
17422
+ 1
17423
+ 1
17424
+ 1
17425
+ 1
17426
+ 1
17427
+ 1
17428
+ 1
17429
+ 1
17430
+ 1
17431
+ 1
17432
+ 1
17433
+ 1
17434
+ 1
17435
+ 1
17436
+ 1
17437
+ 1
17438
+ 1
17439
+ 1
17440
+ 1
17441
+ 1
17442
+ 1
17443
+ 1
17444
+ 1
17445
+ 1
17446
+ 1
17447
+ 1
17448
+ 1
17449
+ 1
17450
+ 1
17451
+ 1
17452
+ 1
17453
+ 1
17454
+ 1
17455
+ 1
17456
+ 1
17457
+ 1
17458
+ 1
17459
+ 1
17460
+ 1
17461
+ 1
17462
+ 1
17463
+ 1
17464
+ 1
17465
+ 1
17466
+ 1
17467
+ 1
17468
+ 1
17469
+ 1
17470
+ 1
17471
+ 1
17472
+ 1
17473
+ 1
17474
+ 1
17475
+ 1
17476
+ 1
17477
+ 1
17478
+ 1
17479
+ 1
17480
+ 1
17481
+ 1
17482
+ 1
17483
+ 1
17484
+ 1
17485
+ 1
17486
+ 1
17487
+ 1
17488
+ 1
17489
+ 1
17490
+ 1
17491
+ 1
17492
+ 1
17493
+ 1
17494
+ 1
17495
+ 1
17496
+ 1
17497
+ 1
17498
+ 1
17499
+ 1
17500
+ 1
17501
+ 1
17502
+ 1
17503
+ 1
17504
+ 1
17505
+ 1
17506
+ 1
17507
+ 1
17508
+ 1
17509
+ 1
17510
+ 1
17511
+ 1
17512
+ 1
17513
+ 1
17514
+ 1
17515
+ 1
17516
+ 1
17517
+ 1
17518
+ 1
17519
+ 1
17520
+ 1
17521
+ 1
17522
+ 1
17523
+ 1
17524
+ 1
17525
+ 1
17526
+ 1
17527
+ 1
17528
+ 1
17529
+ 1
17530
+ 1
17531
+ 1
17532
+ 1
17533
+ 1
17534
+ 1
17535
+ 1
17536
+ 1
17537
+ 1
17538
+ 1
17539
+ 1
17540
+ 1
17541
+ 1
17542
+ 1
17543
+ 1
17544
+ 1
17545
+ 1
17546
+ 1
17547
+ 1
17548
+ 1
17549
+ 1
17550
+ 1
17551
+ 1
17552
+ 1
17553
+ 1
17554
+ 1
17555
+ 1
17556
+ 1
17557
+ 1
17558
+ 1
17559
+ 1
17560
+ 1
17561
+ 1
17562
+ 1
17563
+ 1
17564
+ 1
17565
+ 1
17566
+ 1
17567
+ 1
17568
+ 1
17569
+ 1
17570
+ 1
17571
+ 1
17572
+ 1
17573
+ 1
17574
+ 1
17575
+ 1
17576
+ 1
17577
+ 1
17578
+ 1
17579
+ 1
17580
+ 1
17581
+ 1
17582
+ 1
17583
+ 1
17584
+ 1
17585
+ 1
17586
+ 1
17587
+ 1
17588
+ 1
17589
+ 1
17590
+ 1
17591
+ 1
17592
+ 1
17593
+ 1
17594
+ 1
17595
+ 1
17596
+ 1
17597
+ 1
17598
+ 1
17599
+ 1
17600
+ 1
17601
+ 1
17602
+ 1
17603
+ 1
17604
+ 1
17605
+ 1
17606
+ 1
17607
+ 1
17608
+ 1
17609
+ 1
17610
+ 1
17611
+ 1
17612
+ 1
17613
+ 1
17614
+ 1
17615
+ 1
17616
+ 1
17617
+ 1
17618
+ 1
17619
+ 1
17620
+ 1
17621
+ 1
17622
+ 1
17623
+ 1
17624
+ 1
17625
+ 1
17626
+ 1
17627
+ 1
17628
+ 1
17629
+ 1
17630
+ 1
17631
+ 1
17632
+ 1
17633
+ 1
17634
+ 1
17635
+ 1
17636
+ 1
17637
+ 1
17638
+ 1
17639
+ 1
17640
+ 1
17641
+ 1
17642
+ 1
17643
+ 1
17644
+ 1
17645
+ 1
17646
+ 1
17647
+ 1
17648
+ 1
17649
+ 1
17650
+ 1
17651
+ 1
17652
+ 1
17653
+ 1
17654
+ 1
17655
+ 1
17656
+ 1
17657
+ 1
17658
+ 1
17659
+ 1
17660
+ 1
17661
+ 1
17662
+ 1
17663
+ 1
17664
+ 1
17665
+ 1
17666
+ 1
17667
+ 1
17668
+ 1
17669
+ 1
17670
+ 1
17671
+ 1
17672
+ 1
17673
+ 1
17674
+ 1
17675
+ 1
17676
+ 1
17677
+ 1
17678
+ 1
17679
+ 1
17680
+ 1
17681
+ 1
17682
+ 1
17683
+ 1
17684
+ 1
17685
+ 1
17686
+ 1
17687
+ 1
17688
+ 1
17689
+ 1
17690
+ 1
17691
+ 1
17692
+ 1
17693
+ 1
17694
+ 1
17695
+ 1
17696
+ 1
17697
+ 1
17698
+ 1
17699
+ 1
17700
+ 1
17701
+ 1
17702
+ 1
17703
+ 1
17704
+ 1
17705
+ 1
17706
+ 1
17707
+ 1
17708
+ 1
17709
+ 1
17710
+ 1
17711
+ 1
17712
+ 1
17713
+ 1
17714
+ 1
17715
+ 1
17716
+ 1
17717
+ 1
17718
+ 1
17719
+ 1
17720
+ 1
17721
+ 1
17722
+ 1
17723
+ 1
17724
+ 1
17725
+ 1
17726
+ 1
17727
+ 1
17728
+ 1
17729
+ 1
17730
+ 1
17731
+ 1
17732
+ 1
17733
+ 1
17734
+ 1
17735
+ 1
17736
+ 1
17737
+ 1
17738
+ 1
17739
+ 1
17740
+ 1
17741
+ 1
17742
+ 1
17743
+ 1
17744
+ 1
17745
+ 1
17746
+ 1
17747
+ 1
17748
+ 1
17749
+ 1
17750
+ 1
17751
+ 1
17752
+ 1
17753
+ 1
17754
+ 1
17755
+ 1
17756
+ 1
17757
+ 1
17758
+ 1
17759
+ 1
17760
+ 1
17761
+ 1
17762
+ 1
17763
+ 1
17764
+ 1
17765
+ 1
17766
+ 1
17767
+ 1
17768
+ 1
17769
+ 1
17770
+ 1
17771
+ 1
17772
+ 1
17773
+ 1
17774
+ 1
17775
+ 1
17776
+ 1
17777
+ 1
17778
+ 1
17779
+ 1
17780
+ 1
17781
+ 1
17782
+ 1
17783
+ 1
17784
+ 1
17785
+ 1
17786
+ 1
17787
+ 1
17788
+ 1
17789
+ 1
17790
+ 1
17791
+ 1
17792
+ 1
17793
+ 1
17794
+ 1
17795
+ 1
17796
+ 1
17797
+ 1
17798
+ 1
17799
+ 1
17800
+ 1
17801
+ 1
17802
+ 1
17803
+ 1
17804
+ 1
17805
+ 1
17806
+ 1
17807
+ 1
17808
+ 1
17809
+ 1
17810
+ 1
17811
+ 1
17812
+ 1
17813
+ 1
17814
+ 1
17815
+ 1
17816
+ 1
17817
+ 1
17818
+ 1
17819
+ 1
17820
+ 1
17821
+ 1
17822
+ 1
17823
+ 1
17824
+ 1
17825
+ 1
17826
+ 1
17827
+ 1
17828
+ 1
17829
+ 1
17830
+ 1
17831
+ 1
17832
+ 1
17833
+ 1
17834
+ 1
17835
+ 1
17836
+ 1
17837
+ 1
17838
+ 1
17839
+ 1
17840
+ 1
17841
+ 1
17842
+ 1
17843
+ 1
17844
+ 1
17845
+ 1
17846
+ 1
17847
+ 1
17848
+ 1
17849
+ 1
17850
+ 1
17851
+ 1
17852
+ 1
17853
+ 1
17854
+ 1
17855
+ 1
17856
+ 1
17857
+ 1
17858
+ 1
17859
+ 1
17860
+ 1
17861
+ 1
17862
+ 1
17863
+ 1
17864
+ 1
17865
+ 1
17866
+ 1
17867
+ 1
17868
+ 1
17869
+ 1
17870
+ 1
17871
+ 1
17872
+ 1
17873
+ 1
17874
+ 1
17875
+ 1
17876
+ 1
17877
+ 1
17878
+ 1
17879
+ 1
17880
+ 1
17881
+ 1
17882
+ 1
17883
+ 1
17884
+ 1
17885
+ 1
17886
+ 1
17887
+ 1
17888
+ 1
17889
+ 1
17890
+ 1
17891
+ 1
17892
+ 1
17893
+ 1
17894
+ 1
17895
+ 1
17896
+ 1
17897
+ 1
17898
+ 1
17899
+ 1
17900
+ 1
17901
+ 1
17902
+ 1
17903
+ 1
17904
+ 1
17905
+ 1
17906
+ 1
17907
+ 1
17908
+ 1
17909
+ 1
17910
+ 1
17911
+ 1
17912
+ 1
17913
+ 1
17914
+ 1
17915
+ 1
17916
+ 1
17917
+ 1
17918
+ 1
17919
+ 1
17920
+ 1
17921
+ 1
17922
+ 1
17923
+ 1
17924
+ 1
17925
+ 1
17926
+ 1
17927
+ 1
17928
+ 1
17929
+ 1
17930
+ 1
17931
+ 1
17932
+ 1
17933
+ 1
17934
+ 1
17935
+ 1
17936
+ 1
17937
+ 1
17938
+ 1
17939
+ 1
17940
+ 1
17941
+ 1
17942
+ 1
17943
+ 1
17944
+ 1
17945
+ 1
17946
+ 1
17947
+ 1
17948
+ 1
17949
+ 1
17950
+ 1
17951
+ 1
17952
+ 1
17953
+ 1
17954
+ 1
17955
+ 1
17956
+ 1
17957
+ 1
17958
+ 1
17959
+ 2
17960
+ 2
17961
+ 1
17962
+ 1
17963
+ 2
17964
+ 1
17965
+ 1
17966
+ 1
17967
+ 1
17968
+ 1
17969
+ 1
17970
+ 1
17971
+ 1
17972
+ 1
17973
+ 1
17974
+ 1
17975
+ 1
17976
+ 1
17977
+ 1
17978
+ 1
17979
+ 1
17980
+ 1
17981
+ 1
17982
+ 1
17983
+ 1
17984
+ 1
17985
+ 1
17986
+ 1
17987
+ 1
17988
+ 1
17989
+ 1
17990
+ 1
17991
+ 1
17992
+ 1
17993
+ 1
17994
+ 1
17995
+ 1
17996
+ 1
17997
+ 1
17998
+ 1
17999
+ 1
18000
+ 1
18001
+ 1
18002
+ 1
18003
+ 1
18004
+ 1
18005
+ 1
18006
+ 1
18007
+ 1
18008
+ 1
18009
+ 1
18010
+ 1
18011
+ 1
18012
+ 1
18013
+ 1
18014
+ 1
18015
+ 1
18016
+ 1
18017
+ 1
18018
+ 1
18019
+ 1
18020
+ 1
18021
+ 1
18022
+ 1
18023
+ 1
18024
+ 1
18025
+ 1
18026
+ 1
18027
+ 1
18028
+ 1
18029
+ 1
18030
+ 1
18031
+ 1
18032
+ 1
18033
+ 1
18034
+ 1
18035
+ 1
18036
+ 1
18037
+ 1
18038
+ 1
18039
+ 1
18040
+ 1
18041
+ 1
18042
+ 1
18043
+ 1
18044
+ 1
18045
+ 1
18046
+ 1
18047
+ 1
18048
+ 1
18049
+ 1
18050
+ 1
18051
+ 1
18052
+ 1
18053
+ 1
18054
+ 1
18055
+ 1
18056
+ 1
18057
+ 1
18058
+ 1
18059
+ 1
18060
+ 1
18061
+ 1
18062
+ 1
18063
+ 1
18064
+ 1
18065
+ 1
18066
+ 1
18067
+ 1
18068
+ 1
18069
+ 1
18070
+ 1
18071
+ 1
18072
+ 1
18073
+ 1
18074
+ 1
18075
+ 1
18076
+ 1
18077
+ 1
18078
+ 1
18079
+ 1
18080
+ 1
18081
+ 1
18082
+ 1
18083
+ 1
18084
+ 1
18085
+ 1
18086
+ 1
18087
+ 1
18088
+ 1
18089
+ 1
18090
+ 1
18091
+ 1
18092
+ 1
18093
+ 1
18094
+ 1
18095
+ 1
18096
+ 1
18097
+ 1
18098
+ 1
18099
+ 1
18100
+ 1
18101
+ 1
18102
+ 1
18103
+ 1
18104
+ 1
18105
+ 1
18106
+ 1
18107
+ 1
18108
+ 1
18109
+ 1
18110
+ 1
18111
+ 1
18112
+ 1
18113
+ 1
18114
+ 1
18115
+ 1
18116
+ 1
18117
+ 1
18118
+ 1
18119
+ 1
18120
+ 1
18121
+ 1
18122
+ 1
18123
+ 1
18124
+ 1
18125
+ 1
18126
+ 1
18127
+ 1
18128
+ 1
18129
+ 1
18130
+ 1
18131
+ 1
18132
+ 1
18133
+ 1
18134
+ 1
18135
+ 1
18136
+ 1
18137
+ 1
18138
+ 1
18139
+ 1
18140
+ 1
18141
+ 1
18142
+ 1
18143
+ 1
18144
+ 1
18145
+ 1
18146
+ 1
18147
+ 1
18148
+ 1
18149
+ 1
18150
+ 1
18151
+ 1
18152
+ 1
18153
+ 1
18154
+ 1
18155
+ 1
18156
+ 1
18157
+ 1
18158
+ 1
18159
+ 1
18160
+ 1
18161
+ 1
18162
+ 1
18163
+ 1
18164
+ 1
18165
+ 1
18166
+ 1
18167
+ 1
18168
+ 1
18169
+ 1
18170
+ 1
18171
+ 1
18172
+ 1
18173
+ 1
18174
+ 1
18175
+ 1
18176
+ 1
18177
+ 1
18178
+ 1
18179
+ 1
18180
+ 1
18181
+ 1
18182
+ 1
18183
+ 1
18184
+ 1
18185
+ 1
18186
+ 1
18187
+ 1
18188
+ 1
18189
+ 1
18190
+ 1
18191
+ 1
18192
+ 1
18193
+ 1
18194
+ 1
18195
+ 1
18196
+ 1
18197
+ 1
18198
+ 1
18199
+ 1
18200
+ 1
18201
+ 1
18202
+ 1
18203
+ 1
18204
+ 1
18205
+ 1
18206
+ 1
18207
+ 1
18208
+ 1
18209
+ 1
18210
+ 1
18211
+ 1
18212
+ 1
18213
+ 1
18214
+ 1
18215
+ 1
18216
+ 1
18217
+ 1
18218
+ 1
18219
+ 1
18220
+ 1
18221
+ 1
18222
+ 1
18223
+ 1
18224
+ 1
18225
+ 2
18226
+ 1
18227
+ 1
18228
+ 1
18229
+ 1
18230
+ 1
18231
+ 1
18232
+ 1
18233
+ 1
18234
+ 1
18235
+ 1
18236
+ 1
18237
+ 1
18238
+ 1
18239
+ 1
18240
+ 1
18241
+ 1
18242
+ 1
18243
+ 1
18244
+ 1
18245
+ 1
18246
+ 1
18247
+ 1
18248
+ 1
18249
+ 1
18250
+ 1
18251
+ 1
18252
+ 1
18253
+ 1
18254
+ 1
18255
+ 1
18256
+ 1
18257
+ 1
18258
+ 1
18259
+ 1
18260
+ 1
18261
+ 1
18262
+ 1
18263
+ 1
18264
+ 1
18265
+ 1
18266
+ 1
18267
+ 1
18268
+ 1
18269
+ 1
18270
+ 1
18271
+ 1
18272
+ 1
18273
+ 1
18274
+ 1
18275
+ 1
18276
+ 1
18277
+ 1
18278
+ 1
18279
+ 1
18280
+ 1
18281
+ 1
18282
+ 1
18283
+ 1
18284
+ 1
18285
+ 1
18286
+ 1
18287
+ 1
18288
+ 1
18289
+ 1
18290
+ 1
18291
+ 1
18292
+ 1
18293
+ 1
18294
+ 1
18295
+ 1
18296
+ 1
18297
+ 1
18298
+ 1
18299
+ 1
18300
+ 1
18301
+ 1
18302
+ 1
18303
+ 1
18304
+ 1
18305
+ 1
18306
+ 1
18307
+ 1
18308
+ 1
18309
+ 1
18310
+ 1
18311
+ 1
18312
+ 1
18313
+ 1
18314
+ 1
18315
+ 1
18316
+ 1
18317
+ 1
18318
+ 1
18319
+ 1
18320
+ 1
18321
+ 1
18322
+ 1
18323
+ 1
18324
+ 1
18325
+ 1
18326
+ 1
18327
+ 1
18328
+ 1
18329
+ 1
18330
+ 1
18331
+ 1
18332
+ 1
18333
+ 1
18334
+ 1
18335
+ 1
18336
+ 1
18337
+ 1
18338
+ 1
18339
+ 1
18340
+ 1
18341
+ 1
18342
+ 1
18343
+ 1
18344
+ 1
18345
+ 1
18346
+ 1
18347
+ 1
18348
+ 1
18349
+ 1
18350
+ 1
18351
+ 1
18352
+ 1
18353
+ 1
18354
+ 1
18355
+ 1
18356
+ 1
18357
+ 1
18358
+ 1
18359
+ 1
18360
+ 1
18361
+ 1
18362
+ 1
18363
+ 1
18364
+ 1
18365
+ 1
18366
+ 1
18367
+ 1
18368
+ 1
18369
+ 1
18370
+ 1
18371
+ 1
18372
+ 1
18373
+ 1
18374
+ 1
18375
+ 1
18376
+ 1
18377
+ 1
18378
+ 1
18379
+ 1
18380
+ 1
18381
+ 1
18382
+ 1
18383
+ 1
18384
+ 1
18385
+ 1
18386
+ 1
18387
+ 1
18388
+ 1
18389
+ 1
18390
+ 1
18391
+ 1
18392
+ 1
18393
+ 1
18394
+ 1
18395
+ 1
18396
+ 1
18397
+ 1
18398
+ 1
18399
+ 1
18400
+ 1
18401
+ 1
18402
+ 1
18403
+ 1
18404
+ 1
18405
+ 1
18406
+ 1
18407
+ 1
18408
+ 1
18409
+ 1
18410
+ 1
18411
+ 1
18412
+ 1
18413
+ 1
18414
+ 1
18415
+ 1
18416
+ 1
18417
+ 1
18418
+ 1
18419
+ 1
18420
+ 1
18421
+ 1
18422
+ 1
18423
+ 1
18424
+ 1
18425
+ 1
18426
+ 1
18427
+ 1
18428
+ 1
18429
+ 1
18430
+ 1
18431
+ 1
18432
+ 1
18433
+ 1
18434
+ 1
18435
+ 1
18436
+ 1
18437
+ 1
18438
+ 1
18439
+ 1
18440
+ 1
18441
+ 1
18442
+ 1
18443
+ 1
18444
+ 1
18445
+ 1
18446
+ 1
18447
+ 1
18448
+ 1
18449
+ 1
18450
+ 1
18451
+ 1
18452
+ 1
18453
+ 1
18454
+ 1
18455
+ 1
18456
+ 1
18457
+ 1
18458
+ 1
18459
+ 1
18460
+ 1
18461
+ 1
18462
+ 1
18463
+ 1
18464
+ 1
18465
+ 1
18466
+ 1
18467
+ 1
18468
+ 1
18469
+ 1
18470
+ 1
18471
+ 1
18472
+ 1
18473
+ 1
18474
+ 1
18475
+ 1
18476
+ 1
18477
+ 1
18478
+ 1
18479
+ 1
18480
+ 1
18481
+ 1
18482
+ 1
18483
+ 1
18484
+ 1
18485
+ 1
18486
+ 1
18487
+ 1
18488
+ 1
18489
+ 1
18490
+ 1
18491
+ 1
18492
+ 1
18493
+ 1
18494
+ 1
18495
+ 1
18496
+ 1
18497
+ 1
18498
+ 1
18499
+ 1
18500
+ 1
18501
+ 1
18502
+ 1
18503
+ 1
18504
+ 1
18505
+ 1
18506
+ 1
18507
+ 1
18508
+ 1
18509
+ 1
18510
+ 1
18511
+ 1
18512
+ 1
18513
+ 1
18514
+ 1
18515
+ 1
18516
+ 1
18517
+ 1
18518
+ 1
18519
+ 1
18520
+ 1
18521
+ 1
18522
+ 1
18523
+ 1
18524
+ 1
18525
+ 1
18526
+ 1
18527
+ 1
18528
+ 1
18529
+ 1
18530
+ 1
18531
+ 1
18532
+ 1
18533
+ 1
18534
+ 1
18535
+ 1
18536
+ 1
18537
+ 1
18538
+ 1
18539
+ 1
18540
+ 1
18541
+ 1
18542
+ 1
18543
+ 1
18544
+ 1
18545
+ 1
18546
+ 1
18547
+ 1
18548
+ 1
18549
+ 1
18550
+ 1
18551
+ 1
18552
+ 1
18553
+ 1
18554
+ 1
18555
+ 1
18556
+ 1
18557
+ 1
18558
+ 1
18559
+ 1
18560
+ 1
18561
+ 1
18562
+ 1
18563
+ 1
18564
+ 1
18565
+ 1
18566
+ 1
18567
+ 1
18568
+ 1
18569
+ 1
18570
+ 1
18571
+ 1
18572
+ 1
18573
+ 1
18574
+ 1
18575
+ 1
18576
+ 1
18577
+ 1
18578
+ 1
18579
+ 1
18580
+ 1
18581
+ 1
18582
+ 1
18583
+ 1
18584
+ 1
18585
+ 1
18586
+ 1
18587
+ 1
18588
+ 1
18589
+ 1
18590
+ 1
18591
+ 1
18592
+ 1
18593
+ 1
18594
+ 1
18595
+ 1
18596
+ 1
18597
+ 1
18598
+ 1
18599
+ 1
18600
+ 1
18601
+ 1
18602
+ 1
18603
+ 1
18604
+ 1
18605
+ 1
18606
+ 1
18607
+ 1
18608
+ 1
18609
+ 1
18610
+ 1
18611
+ 1
18612
+ 1
18613
+ 1
18614
+ 1
18615
+ 1
18616
+ 1
18617
+ 1
18618
+ 1
18619
+ 1
18620
+ 1
18621
+ 1
18622
+ 1
18623
+ 1
18624
+ 1
18625
+ 1
18626
+ 1
18627
+ 1
18628
+ 1
18629
+ 1
18630
+ 1
18631
+ 1
18632
+ 1
18633
+ 1
18634
+ 1
18635
+ 1
18636
+ 1
18637
+ 1
18638
+ 1
18639
+ 1
18640
+ 1
18641
+ 1
18642
+ 1
18643
+ 1
18644
+ 1
18645
+ 1
18646
+ 1
18647
+ 1
18648
+ 1
18649
+ 1
18650
+ 1
18651
+ 1
18652
+ 1
18653
+ 1
18654
+ 1
18655
+ 1
18656
+ 1
18657
+ 1
18658
+ 1
18659
+ 1
18660
+ 1
18661
+ 1
18662
+ 1
18663
+ 1
18664
+ 1
18665
+ 1
18666
+ 1
18667
+ 1
18668
+ 1
18669
+ 1
18670
+ 1
18671
+ 1
18672
+ 1
18673
+ 1
18674
+ 1
18675
+ 1
18676
+ 1
18677
+ 1
18678
+ 1
18679
+ 1
18680
+ 1
18681
+ 1
18682
+ 1
18683
+ 1
18684
+ 1
18685
+ 1
18686
+ 1
18687
+ 1
18688
+ 1
18689
+ 1
18690
+ 1
18691
+ 1
18692
+ 1
18693
+ 1
18694
+ 1
18695
+ 1
18696
+ 1
18697
+ 1
18698
+ 1
18699
+ 1
18700
+ 1
18701
+ 1
18702
+ 1
18703
+ 1
18704
+ 1
18705
+ 1
18706
+ 1
18707
+ 1
18708
+ 1
18709
+ 1
18710
+ 1
18711
+ 1
18712
+ 1
18713
+ 1
18714
+ 1
18715
+ 1
18716
+ 1
18717
+ 1
18718
+ 1
18719
+ 1
18720
+ 1
18721
+ 1
18722
+ 1
18723
+ 1
18724
+ 1
18725
+ 1
18726
+ 1
18727
+ 1
18728
+ 1
18729
+ 1
18730
+ 1
18731
+ 1
18732
+ 1
18733
+ 1
18734
+ 1
18735
+ 1
18736
+ 1
18737
+ 1
18738
+ 1
18739
+ 1
18740
+ 1
18741
+ 1
18742
+ 1
18743
+ 1
18744
+ 1
18745
+ 1
18746
+ 1
18747
+ 1
18748
+ 1
18749
+ 1
18750
+ 1
18751
+ 1
18752
+ 1
18753
+ 1
18754
+ 1
18755
+ 1
18756
+ 1
18757
+ 1
18758
+ 1
18759
+ 1
18760
+ 1
18761
+ 1
18762
+ 1
18763
+ 1
18764
+ 1
18765
+ 1
18766
+ 1
18767
+ 1
18768
+ 1
18769
+ 1
18770
+ 1
18771
+ 1
18772
+ 1
18773
+ 1
18774
+ 1
18775
+ 1
18776
+ 1
18777
+ 1
18778
+ 1
18779
+ 1
18780
+ 1
18781
+ 1
18782
+ 1
18783
+ 1
18784
+ 1
18785
+ 1
18786
+ 1
18787
+ 1
18788
+ 1
18789
+ 1
18790
+ 1
18791
+ 1
18792
+ 1
18793
+ 1
18794
+ 1
18795
+ 1
18796
+ 1
18797
+ 1
18798
+ 1
18799
+ 1
18800
+ 1
18801
+ 1
18802
+ 1
18803
+ 1
18804
+ 1
18805
+ 1
18806
+ 1
18807
+ 1
18808
+ 1
18809
+ 1
18810
+ 1
18811
+ 1
18812
+ 1
18813
+ 1
18814
+ 1
18815
+ 1
18816
+ 1
18817
+ 1
18818
+ 1
18819
+ 1
18820
+ 1
18821
+ 1
18822
+ 1
18823
+ 1
18824
+ 1
18825
+ 1
18826
+ 1
18827
+ 1
18828
+ 1
18829
+ 1
18830
+ 1
18831
+ 1
18832
+ 1
18833
+ 1
18834
+ 1
18835
+ 1
18836
+ 1
18837
+ 1
18838
+ 1
18839
+ 1
18840
+ 1
18841
+ 1
18842
+ 1
18843
+ 1
18844
+ 1
18845
+ 1
18846
+ 1
18847
+ 1
18848
+ 1
18849
+ 1
18850
+ 1
18851
+ 1
18852
+ 1
18853
+ 1
18854
+ 1
18855
+ 1
18856
+ 1
18857
+ 1
18858
+ 1
18859
+ 1
18860
+ 1
18861
+ 1
18862
+ 1
18863
+ 1
18864
+ 1
18865
+ 1
18866
+ 1
18867
+ 1
18868
+ 1
18869
+ 1
18870
+ 1
18871
+ 1
18872
+ 1
18873
+ 1
18874
+ 1
18875
+ 1
18876
+ 1
18877
+ 1
18878
+ 1
18879
+ 1
18880
+ 1
18881
+ 1
18882
+ 1
18883
+ 1
18884
+ 1
18885
+ 1
18886
+ 1
18887
+ 1
18888
+ 1
18889
+ 1
18890
+ 1
18891
+ 1
18892
+ 1
18893
+ 1
18894
+ 1
18895
+ 1
18896
+ 1
18897
+ 1
18898
+ 1
18899
+ 1
18900
+ 1
18901
+ 1
18902
+ 1
18903
+ 1
18904
+ 1
18905
+ 1
18906
+ 1
18907
+ 1
18908
+ 1
18909
+ 1
18910
+ 1
18911
+ 1
18912
+ 1
18913
+ 1
18914
+ 1
18915
+ 1
18916
+ 1
18917
+ 1
18918
+ 1
18919
+ 1
18920
+ 1
18921
+ 1
18922
+ 1
18923
+ 1
18924
+ 1
18925
+ 1
18926
+ 1
18927
+ 1
18928
+ 1
18929
+ 1
18930
+ 1
18931
+ 1
18932
+ 1
18933
+ 1
18934
+ 1
18935
+ 1
18936
+ 1
18937
+ 1
18938
+ 1
18939
+ 1
18940
+ 1
18941
+ 1
18942
+ 1
18943
+ 1
18944
+ 1
18945
+ 1
18946
+ 1
18947
+ 1
18948
+ 1
18949
+ 1
18950
+ 1
18951
+ 1
18952
+ 1
18953
+ 1
18954
+ 1
18955
+ 1
18956
+ 1
18957
+ 1
18958
+ 1
18959
+ 1
18960
+ 1
18961
+ 1
18962
+ 1
18963
+ 1
18964
+ 1
18965
+ 1
18966
+ 1
18967
+ 1
18968
+ 1
18969
+ 1
18970
+ 1
18971
+ 1
18972
+ 1
18973
+ 1
18974
+ 1
18975
+ 1
18976
+ 1
18977
+ 1
18978
+ 1
18979
+ 1
18980
+ 1
18981
+ 1
18982
+ 1
18983
+ 1
18984
+ 1
18985
+ 1
18986
+ 1
18987
+ 1
18988
+ 1
18989
+ 1
18990
+ 1
18991
+ 1
18992
+ 1
18993
+ 1
18994
+ 1
18995
+ 1
18996
+ 1
18997
+ 1
18998
+ 1
18999
+ 1
19000
+ 1
19001
+ 1
19002
+ 1
19003
+ 1
19004
+ 1
19005
+ 1
19006
+ 1
19007
+ 1
19008
+ 1
19009
+ 1
19010
+ 1
19011
+ 1
19012
+ 1
19013
+ 1
19014
+ 1
19015
+ 1
19016
+ 1
19017
+ 1
19018
+ 1
19019
+ 1
19020
+ 1
19021
+ 1
19022
+ 1
19023
+ 1
19024
+ 1
19025
+ 1
19026
+ 1
19027
+ 1
19028
+ 1
19029
+ 1
19030
+ 1
19031
+ 1
19032
+ 1
19033
+ 1
19034
+ 1
19035
+ 1
19036
+ 1
19037
+ 1
19038
+ 1
19039
+ 1
19040
+ 1
19041
+ 1
19042
+ 1
19043
+ 1
19044
+ 1
19045
+ 1
19046
+ 1
19047
+ 1
19048
+ 1
19049
+ 1
19050
+ 1
19051
+ 1
19052
+ 1
19053
+ 1
19054
+ 1
19055
+ 1
19056
+ 1
19057
+ 1
19058
+ 1
19059
+ 1
19060
+ 1
19061
+ 1
19062
+ 1
19063
+ 1
19064
+ 1
19065
+ 1
19066
+ 1
19067
+ 1
19068
+ 1
19069
+ 1
19070
+ 1
19071
+ 1
19072
+ 1
19073
+ 1
19074
+ 1
19075
+ 1
19076
+ 1
19077
+ 1
19078
+ 1
19079
+ 1
19080
+ 1
19081
+ 1
19082
+ 1
19083
+ 1
19084
+ 1
19085
+ 1
19086
+ 1
19087
+ 1
19088
+ 1
19089
+ 1
19090
+ 1
19091
+ 1
19092
+ 1
19093
+ 1
19094
+ 1
19095
+ 1
19096
+ 1
19097
+ 1
19098
+ 1
19099
+ 1
19100
+ 1
19101
+ 1
19102
+ 1
19103
+ 1
19104
+ 1
19105
+ 1
19106
+ 1
19107
+ 1
19108
+ 1
19109
+ 1
19110
+ 1
19111
+ 1
19112
+ 1
19113
+ 1
19114
+ 1
19115
+ 1
19116
+ 1
19117
+ 1
19118
+ 1
19119
+ 1
19120
+ 1
19121
+ 1
19122
+ 1
19123
+ 1
19124
+ 1
19125
+ 1
19126
+ 1
19127
+ 1
19128
+ 1
19129
+ 1
19130
+ 1
19131
+ 1
19132
+ 1
19133
+ 1
19134
+ 1
19135
+ 1
19136
+ 1
19137
+ 1
19138
+ 1
19139
+ 1
19140
+ 1
19141
+ 1
19142
+ 1
19143
+ 1
19144
+ 1
19145
+ 1
19146
+ 1
19147
+ 1
19148
+ 1
19149
+ 1
19150
+ 1
19151
+ 1
19152
+ 1
19153
+ 1
19154
+ 1
19155
+ 1
19156
+ 1
19157
+ 1
19158
+ 1
19159
+ 1
19160
+ 1
19161
+ 1
19162
+ 1
19163
+ 1
19164
+ 1
19165
+ 1
19166
+ 1
19167
+ 1
19168
+ 1
19169
+ 1
19170
+ 1
19171
+ 1
19172
+ 1
19173
+ 1
19174
+ 1
19175
+ 1
19176
+ 1
19177
+ 1
19178
+ 1
19179
+ 1
19180
+ 1
19181
+ 1
19182
+ 1
19183
+ 1
19184
+ 1
19185
+ 1
19186
+ 1
19187
+ 1
19188
+ 1
19189
+ 1
19190
+ 1
19191
+ 1
19192
+ 1
19193
+ 1
19194
+ 1
19195
+ 1
19196
+ 1
19197
+ 1
19198
+ 1
19199
+ 1
19200
+ 1
19201
+ 1
19202
+ 1
19203
+ 1
19204
+ 1
19205
+ 1
19206
+ 1
19207
+ 1
19208
+ 1
19209
+ 1
19210
+ 1
19211
+ 1
19212
+ 1
19213
+ 1
19214
+ 1
19215
+ 1
19216
+ 1
19217
+ 1
19218
+ 1
19219
+ 1
19220
+ 1
19221
+ 1
19222
+ 1
19223
+ 1
19224
+ 1
19225
+ 1
19226
+ 1
19227
+ 1
19228
+ 1
19229
+ 1
19230
+ 1
19231
+ 1
19232
+ 1
19233
+ 1
19234
+ 1
19235
+ 1
19236
+ 1
19237
+ 1
19238
+ 1
19239
+ 1
19240
+ 1
19241
+ 1
19242
+ 1
19243
+ 1
19244
+ 1
19245
+ 1
19246
+ 1
19247
+ 1
19248
+ 1
19249
+ 1
19250
+ 1
19251
+ 1
19252
+ 1
19253
+ 1
19254
+ 1
19255
+ 1
19256
+ 1
19257
+ 1
19258
+ 1
19259
+ 1
19260
+ 1
19261
+ 1
19262
+ 1
19263
+ 1
19264
+ 1
19265
+ 1
19266
+ 1
19267
+ 1
19268
+ 1
19269
+ 1
19270
+ 1
19271
+ 1
19272
+ 1
19273
+ 1
19274
+ 1
19275
+ 1
19276
+ 1
19277
+ 1
19278
+ 1
19279
+ 1
19280
+ 1
19281
+ 1
19282
+ 1
19283
+ 1
19284
+ 1
19285
+ 1
19286
+ 1
19287
+ 1
19288
+ 1
19289
+ 1
19290
+ 1
19291
+ 1
19292
+ 1
19293
+ 1
19294
+ 1
19295
+ 1
19296
+ 1
19297
+ 1
19298
+ 1
19299
+ 1
19300
+ 1
19301
+ 1
19302
+ 1
19303
+ 1
19304
+ 1
19305
+ 1
19306
+ 1
19307
+ 1
19308
+ 1
19309
+ 1
19310
+ 1
19311
+ 1
19312
+ 1
19313
+ 1
19314
+ 1
19315
+ 1
19316
+ 1
19317
+ 1
19318
+ 1
19319
+ 1
19320
+ 1
19321
+ 1
19322
+ 1
19323
+ 1
19324
+ 1
19325
+ 1
19326
+ 1
19327
+ 1
19328
+ 1
19329
+ 1
19330
+ 1
19331
+ 1
19332
+ 1
19333
+ 1
19334
+ 1
19335
+ 1
19336
+ 1
19337
+ 1
19338
+ 1
19339
+ 1
19340
+ 1
19341
+ 1
19342
+ 1
19343
+ 1
19344
+ 1
19345
+ 1
19346
+ 1
19347
+ 1
19348
+ 1
19349
+ 1
19350
+ 1
19351
+ 1
19352
+ 1
19353
+ 1
19354
+ 1
19355
+ 1
19356
+ 1
19357
+ 1
19358
+ 1
19359
+ 1
19360
+ 1
19361
+ 1
19362
+ 1
19363
+ 1
19364
+ 1
19365
+ 1
19366
+ 1
19367
+ 1
19368
+ 1
19369
+ 1
19370
+ 1
19371
+ 1
19372
+ 1
19373
+ 1
19374
+ 1
19375
+ 1
19376
+ 1
19377
+ 1
19378
+ 1
19379
+ 1
19380
+ 1
19381
+ 1
19382
+ 1
19383
+ 1
19384
+ 1
19385
+ 1
19386
+ 1
19387
+ 1
19388
+ 1
19389
+ 1
19390
+ 1
19391
+ 1
19392
+ 1
19393
+ 1
19394
+ 1
19395
+ 1
19396
+ 1
19397
+ 1
19398
+ 1
19399
+ 1
19400
+ 1
19401
+ 1
19402
+ 1
19403
+ 1
19404
+ 1
19405
+ 1
19406
+ 1
19407
+ 1
19408
+ 1
19409
+ 1
19410
+ 1
19411
+ 1
19412
+ 1
19413
+ 1
19414
+ 1
19415
+ 1
19416
+ 1
19417
+ 1
19418
+ 1
19419
+ 1
19420
+ 1
19421
+ 1
19422
+ 1
19423
+ 1
19424
+ 1
19425
+ 1
19426
+ 1
19427
+ 1
19428
+ 1
19429
+ 1
19430
+ 1
19431
+ 1
19432
+ 1
19433
+ 1
19434
+ 1
19435
+ 1
19436
+ 1
19437
+ 1
19438
+ 1
19439
+ 1
19440
+ 1
19441
+ 1
19442
+ 1
19443
+ 1
19444
+ 1
19445
+ 1
19446
+ 1
19447
+ 1
19448
+ 1
19449
+ 1
19450
+ 1
19451
+ 1
19452
+ 1
19453
+ 1
19454
+ 1
19455
+ 1
19456
+ 1
19457
+ 1
19458
+ 1
19459
+ 1
19460
+ 1
19461
+ 1
19462
+ 1
19463
+ 1
19464
+ 1
19465
+ 1
19466
+ 1
19467
+ 1
19468
+ 1
19469
+ 1
19470
+ 1
19471
+ 1
19472
+ 1
19473
+ 1
19474
+ 1
19475
+ 1
19476
+ 1
19477
+ 1
19478
+ 1
19479
+ 1
19480
+ 1
19481
+ 1
19482
+ 1
19483
+ 1
19484
+ 1
19485
+ 1
19486
+ 1
19487
+ 1
19488
+ 1
19489
+ 1
19490
+ 1
19491
+ 1
19492
+ 1
19493
+ 1
19494
+ 1
19495
+ 1
19496
+ 1
19497
+ 1
19498
+ 1
19499
+ 1
19500
+ 1
19501
+ 1
19502
+ 1
19503
+ 1
19504
+ 1
19505
+ 1
19506
+ 1
19507
+ 1
19508
+ 1
19509
+ 1
19510
+ 1
19511
+ 1
19512
+ 1
19513
+ 1
19514
+ 1
19515
+ 1
19516
+ 1
19517
+ 1
19518
+ 1
19519
+ 1
19520
+ 1
19521
+ 1
19522
+ 1
19523
+ 1
19524
+ 1
19525
+ 1
19526
+ 1
19527
+ 1
19528
+ 1
19529
+ 1
19530
+ 1
19531
+ 1
19532
+ 1
19533
+ 1
19534
+ 1
19535
+ 1
19536
+ 1
19537
+ 1
19538
+ 1
19539
+ 1
19540
+ 1
19541
+ 1
19542
+ 1
19543
+ 1
19544
+ 1
19545
+ 1
19546
+ 1
19547
+ 1
19548
+ 1
19549
+ 1
19550
+ 1
19551
+ 1
19552
+ 1
19553
+ 1
19554
+ 1
19555
+ 1
19556
+ 1
19557
+ 1
19558
+ 1
19559
+ 1
19560
+ 1
19561
+ 1
19562
+ 1
19563
+ 1
19564
+ 1
19565
+ 1
19566
+ 1
19567
+ 1
19568
+ 1
19569
+ 1
19570
+ 1
19571
+ 1
19572
+ 1
19573
+ 1
19574
+ 1
19575
+ 1
19576
+ 1
19577
+ 1
19578
+ 1
19579
+ 1
19580
+ 1
19581
+ 1
19582
+ 1
19583
+ 1
19584
+ 1
19585
+ 1
19586
+ 1
19587
+ 1
19588
+ 1
19589
+ 1
19590
+ 1
19591
+ 1
19592
+ 1
19593
+ 1
19594
+ 1
19595
+ 1
19596
+ 1
19597
+ 1
19598
+ 1
19599
+ 1
19600
+ 1
19601
+ 1
19602
+ 1
19603
+ 1
19604
+ 1
19605
+ 1
19606
+ 1
19607
+ 1
19608
+ 1
19609
+ 1
19610
+ 1
19611
+ 1
19612
+ 1
19613
+ 1
19614
+ 1
19615
+ 1
19616
+ 1
19617
+ 1
19618
+ 1
19619
+ 1
19620
+ 1
19621
+ 1
19622
+ 1
19623
+ 1
19624
+ 1
19625
+ 1
19626
+ 1
19627
+ 1
19628
+ 1
19629
+ 1
19630
+ 1
19631
+ 1
19632
+ 1
19633
+ 1
19634
+ 1
19635
+ 1
19636
+ 1
19637
+ 1
19638
+ 1
19639
+ 1
19640
+ 1
19641
+ 1
19642
+ 1
19643
+ 1
19644
+ 1
19645
+ 1
19646
+ 1
19647
+ 1
19648
+ 1
19649
+ 1
19650
+ 1
19651
+ 1
19652
+ 1
19653
+ 1
19654
+ 1
19655
+ 1
19656
+ 1
19657
+ 1
19658
+ 1
19659
+ 1
19660
+ 1
19661
+ 1
19662
+ 1
19663
+ 1
19664
+ 1
19665
+ 1
19666
+ 1
19667
+ 1
19668
+ 1
19669
+ 1
19670
+ 1
19671
+ 1
19672
+ 1
19673
+ 1
19674
+ 1
19675
+ 1
19676
+ 1
19677
+ 1
19678
+ 1
19679
+ 1
19680
+ 1
19681
+ 1
19682
+ 1
19683
+ 1
19684
+ 1
19685
+ 1
19686
+ 1
19687
+ 1
19688
+ 1
19689
+ 1
19690
+ 1
19691
+ 1
19692
+ 1
19693
+ 1
19694
+ 1
19695
+ 1
19696
+ 1
19697
+ 1
19698
+ 1
19699
+ 1
19700
+ 1
19701
+ 1
19702
+ 1
19703
+ 1
19704
+ 1
19705
+ 1
19706
+ 1
19707
+ 1
19708
+ 1
19709
+ 1
19710
+ 1
19711
+ 1
19712
+ 1
19713
+ 1
19714
+ 1
19715
+ 1
19716
+ 1
19717
+ 1
19718
+ 1
19719
+ 1
19720
+ 1
19721
+ 1
19722
+ 1
19723
+ 1
19724
+ 1
19725
+ 1
19726
+ 1
19727
+ 1
19728
+ 1
19729
+ 1
19730
+ 1
19731
+ 1
19732
+ 1
19733
+ 1
19734
+ 1
19735
+ 1
19736
+ 1
19737
+ 1
19738
+ 1
19739
+ 1
19740
+ 1
19741
+ 1
19742
+ 1
19743
+ 1
19744
+ 1
19745
+ 1
19746
+ 1
19747
+ 1
19748
+ 1
19749
+ 1
19750
+ 1
19751
+ 1
19752
+ 1
19753
+ 1
19754
+ 1
19755
+ 1
19756
+ 1
19757
+ 1
19758
+ 1
19759
+ 1
19760
+ 1
19761
+ 1
19762
+ 1
19763
+ 1
19764
+ 1
19765
+ 1
19766
+ 1
19767
+ 1
19768
+ 1
19769
+ 1
19770
+ 1
19771
+ 1
19772
+ 1
19773
+ 1
19774
+ 1
19775
+ 1
19776
+ 1
19777
+ 1
19778
+ 1
19779
+ 1
19780
+ 1
19781
+ 1
19782
+ 1
19783
+ 1
19784
+ 1
19785
+ 1
19786
+ 1
19787
+ 1
19788
+ 1
19789
+ 1
19790
+ 1
19791
+ 1
19792
+ 1
19793
+ 1
19794
+ 1
19795
+ 1
19796
+ 1
19797
+ 1
19798
+ 1
19799
+ 1
19800
+ 1
19801
+ 1
19802
+ 1
19803
+ 1
19804
+ 1
19805
+ 1
19806
+ 1
19807
+ 1
19808
+ 1
19809
+ 1
19810
+ 1
19811
+ 1
19812
+ 1
19813
+ 1
19814
+ 1
19815
+ 1
19816
+ 1
19817
+ 1
19818
+ 1
19819
+ 1
19820
+ 1
19821
+ 1
19822
+ 1
19823
+ 1
19824
+ 1
19825
+ 1
19826
+ 1
19827
+ 1
19828
+ 1
19829
+ 1
19830
+ 1
19831
+ 1
19832
+ 1
19833
+ 1
19834
+ 1
19835
+ 1
19836
+ 1
19837
+ 1
19838
+ 1
19839
+ 1
19840
+ 1
19841
+ 1
19842
+ 1
19843
+ 1
19844
+ 1
19845
+ 1
19846
+ 1
19847
+ 1
19848
+ 1
19849
+ 1
19850
+ 1
19851
+ 1
19852
+ 1
19853
+ 1
19854
+ 1
19855
+ 1
19856
+ 1
19857
+ 1
19858
+ 1
19859
+ 1
19860
+ 1
19861
+ 1
19862
+ 1
19863
+ 1
19864
+ 1
19865
+ 1
19866
+ 1
19867
+ 1
19868
+ 1
19869
+ 1
19870
+ 1
19871
+ 1
19872
+ 1
19873
+ 1
19874
+ 1
19875
+ 1
19876
+ 1
19877
+ 1
19878
+ 1
19879
+ 1
19880
+ 1
19881
+ 1
19882
+ 1
19883
+ 1
19884
+ 1
19885
+ 1
19886
+ 1
19887
+ 1
19888
+ 1
19889
+ 1
19890
+ 1
19891
+ 1
19892
+ 1
19893
+ 1
19894
+ 1
19895
+ 1
19896
+ 1
19897
+ 1
19898
+ 1
19899
+ 1
19900
+ 1
19901
+ 1
19902
+ 1
19903
+ 1
19904
+ 1
19905
+ 1
19906
+ 1
19907
+ 1
19908
+ 1
19909
+ 1
19910
+ 1
19911
+ 1
19912
+ 1
19913
+ 1
19914
+ 1
19915
+ 1
19916
+ 1
19917
+ 1
19918
+ 1
19919
+ 1
19920
+ 1
19921
+ 1
19922
+ 1
19923
+ 1
19924
+ 1
19925
+ 1
19926
+ 1
19927
+ 1
19928
+ 1
19929
+ 1
19930
+ 1
19931
+ 1
19932
+ 1
19933
+ 1
19934
+ 1
19935
+ 1
19936
+ 1
19937
+ 1
19938
+ 1
19939
+ 1
19940
+ 1
19941
+ 1
19942
+ 1
19943
+ 1
19944
+ 1
19945
+ 1
19946
+ 1
19947
+ 1
19948
+ 1
19949
+ 1
19950
+ 1
19951
+ 1
19952
+ 1
19953
+ 1
19954
+ 1
19955
+ 1
19956
+ 1
19957
+ 1
19958
+ 1
19959
+ 1
19960
+ 1
19961
+ 1
19962
+ 1
19963
+ 1
19964
+ 1
19965
+ 1
19966
+ 1
19967
+ 1
19968
+ 1
19969
+ 1
19970
+ 1
19971
+ 1
19972
+ 1
19973
+ 1
19974
+ 1
19975
+ 1
19976
+ 1
19977
+ 1
19978
+ 1
19979
+ 1
19980
+ 1
19981
+ 1
19982
+ 1
19983
+ 1
19984
+ 1
19985
+ 1
19986
+ 1
19987
+ 1
19988
+ 1
19989
+ 1
19990
+ 1
19991
+ 1
19992
+ 1
19993
+ 1
19994
+ 1
19995
+ 1
19996
+ 1
19997
+ 1
19998
+ 1
19999
+ 1
20000
+ 1
20001
+ 1
20002
+ 1
20003
+ 1
20004
+ 1
20005
+ 1
20006
+ 1
20007
+ 1
20008
+ 1
20009
+ 1
20010
+ 1
20011
+ 1
20012
+ 1
20013
+ 1
20014
+ 1
20015
+ 1
20016
+ 1
20017
+ 1
20018
+ 1
20019
+ 1
20020
+ 1
20021
+ 1
20022
+ 1
20023
+ 1
20024
+ 1
20025
+ 1
20026
+ 1
20027
+ 1
20028
+ 1
20029
+ 1
20030
+ 1
20031
+ 1
20032
+ 1
20033
+ 1
20034
+ 1
20035
+ 1
20036
+ 1
20037
+ 1
20038
+ 1
20039
+ 1
20040
+ 1
20041
+ 1
20042
+ 1
20043
+ 1
20044
+ 1
20045
+ 1
20046
+ 1
20047
+ 1
20048
+ 1
20049
+ 1
20050
+ 1
20051
+ 1
20052
+ 1
20053
+ 1
20054
+ 1
20055
+ 1
20056
+ 1
20057
+ 1
20058
+ 1
20059
+ 1
20060
+ 1
20061
+ 1
20062
+ 1
20063
+ 1
20064
+ 1
20065
+ 1
20066
+ 1
20067
+ 1
20068
+ 1
20069
+ 1
20070
+ 1
20071
+ 1
20072
+ 1
20073
+ 1
20074
+ 1
20075
+ 1
20076
+ 1
20077
+ 1
20078
+ 1
20079
+ 1
20080
+ 1
20081
+ 1
20082
+ 1
20083
+ 1
20084
+ 1
20085
+ 1
20086
+ 1
20087
+ 1
20088
+ 1
20089
+ 1
20090
+ 1
20091
+ 1
20092
+ 1
20093
+ 1
20094
+ 1
20095
+ 1
20096
+ 1
20097
+ 1
20098
+ 1
20099
+ 1
20100
+ 1
20101
+ 1
20102
+ 1
20103
+ 1
20104
+ 1
20105
+ 1
20106
+ 1
20107
+ 1
20108
+ 1
20109
+ 1
20110
+ 1
20111
+ 1
20112
+ 1
20113
+ 1
20114
+ 1
20115
+ 1
20116
+ 1
20117
+ 1
20118
+ 1
20119
+ 1
20120
+ 1
20121
+ 1
20122
+ 1
20123
+ 1
20124
+ 1
20125
+ 1
20126
+ 1
20127
+ 1
20128
+ 1
20129
+ 1
20130
+ 1
20131
+ 1
20132
+ 1
20133
+ 1
20134
+ 1
20135
+ 1
20136
+ 1
20137
+ 1
20138
+ 1
20139
+ 1
20140
+ 1
20141
+ 1
20142
+ 1
20143
+ 1
20144
+ 1
20145
+ 1
20146
+ 1
20147
+ 1
20148
+ 1
20149
+ 1
20150
+ 1
20151
+ 1
20152
+ 1
20153
+ 1
20154
+ 1
20155
+ 1
20156
+ 1
20157
+ 1
20158
+ 1
20159
+ 1
20160
+ 1
20161
+ 1
20162
+ 1
20163
+ 1
20164
+ 1
20165
+ 1
20166
+ 1
20167
+ 1
20168
+ 1
20169
+ 1
20170
+ 1
20171
+ 1
20172
+ 1
20173
+ 1
20174
+ 1
20175
+ 1
20176
+ 1
20177
+ 1
20178
+ 1
20179
+ 1
20180
+ 1
20181
+ 1
20182
+ 1
20183
+ 1
20184
+ 1
20185
+ 1
20186
+ 1
20187
+ 1
20188
+ 1
20189
+ 1
20190
+ 1
20191
+ 1
20192
+ 1
20193
+ 1
20194
+ 1
20195
+ 1
20196
+ 1
20197
+ 1
20198
+ 1
20199
+ 1
20200
+ 1
20201
+ 1
20202
+ 1
20203
+ 1
20204
+ 1
20205
+ 1
20206
+ 1
20207
+ 1
20208
+ 1
20209
+ 1
20210
+ 1
20211
+ 1
20212
+ 1
20213
+ 1
20214
+ 1
20215
+ 1
20216
+ 1
20217
+ 1
20218
+ 1
20219
+ 1
20220
+ 1
20221
+ 1
20222
+ 1
20223
+ 1
20224
+ 1
20225
+ 1
20226
+ 1
20227
+ 1
20228
+ 1
20229
+ 1
20230
+ 1
20231
+ 1
20232
+ 1
20233
+ 1
20234
+ 1
20235
+ 1
20236
+ 1
20237
+ 1
20238
+ 1
20239
+ 1
20240
+ 1
20241
+ 1
20242
+ 1
20243
+ 1
20244
+ 1
20245
+ 1
20246
+ 1
20247
+ 1
20248
+ 1
20249
+ 1
20250
+ 1
20251
+ 1
20252
+ 1
20253
+ 1
20254
+ 1
20255
+ 1
20256
+ 1
20257
+ 1
20258
+ 1
20259
+ 1
20260
+ 1
20261
+ 1
20262
+ 1
20263
+ 1
20264
+ 1
20265
+ 1
20266
+ 1
20267
+ 1
20268
+ 1
20269
+ 1
20270
+ 1
20271
+ 1
20272
+ 1
20273
+ 1
20274
+ 1
20275
+ 1
20276
+ 1
20277
+ 1
20278
+ 1
20279
+ 1
20280
+ 1
20281
+ 1
20282
+ 1
20283
+ 1
20284
+ 1
20285
+ 1
20286
+ 1
20287
+ 1
20288
+ 1
20289
+ 1
20290
+ 1
20291
+ 1
20292
+ 1
20293
+ 1
20294
+ 1
20295
+ 1
20296
+ 1
20297
+ 1
20298
+ 1
20299
+ 1
20300
+ 1
20301
+ 1
20302
+ 1
20303
+ 1
20304
+ 1
20305
+ 1
20306
+ 1
20307
+ 1
20308
+ 1
20309
+ 1
20310
+ 1
20311
+ 1
20312
+ 1
20313
+ 1
20314
+ 1
20315
+ 1
20316
+ 1
20317
+ 1
20318
+ 1
20319
+ 1
20320
+ 1
20321
+ 1
20322
+ 1
20323
+ 1
20324
+ 1
20325
+ 1
20326
+ 1
20327
+ 1
20328
+ 1
20329
+ 1
20330
+ 1
20331
+ 1
20332
+ 1
20333
+ 1
20334
+ 1
20335
+ 1
20336
+ 1
20337
+ 1
20338
+ 1
20339
+ 1
20340
+ 1
20341
+ 1
20342
+ 1
20343
+ 1
20344
+ 1
20345
+ 1
20346
+ 1
20347
+ 1
20348
+ 1
20349
+ 1
20350
+ 1
20351
+ 1
20352
+ 1
20353
+ 1
20354
+ 1
20355
+ 1
20356
+ 1
20357
+ 1
20358
+ 1
20359
+ 1
20360
+ 1
20361
+ 1
20362
+ 1
20363
+ 1
20364
+ 1
20365
+ 1
20366
+ 1
20367
+ 1
20368
+ 1
20369
+ 1
20370
+ 1
20371
+ 1
20372
+ 1
20373
+ 1
20374
+ 1
20375
+ 1
20376
+ 1
20377
+ 1
20378
+ 1
20379
+ 1
20380
+ 1
20381
+ 1
20382
+ 1
20383
+ 1
20384
+ 1
20385
+ 1
20386
+ 1
20387
+ 1
20388
+ 1
20389
+ 1
20390
+ 1
20391
+ 1
20392
+ 1
20393
+ 1
20394
+ 1
20395
+ 1
20396
+ 1
20397
+ 1
20398
+ 1
20399
+ 1
20400
+ 1
20401
+ 1
20402
+ 1
20403
+ 1
20404
+ 1
20405
+ 1
20406
+ 1
20407
+ 1
20408
+ 1
20409
+ 1
20410
+ 1
20411
+ 1
20412
+ 1
20413
+ 1
20414
+ 1
20415
+ 1
20416
+ 1
20417
+ 1
20418
+ 1
20419
+ 1
20420
+ 1
20421
+ 1
20422
+ 1
20423
+ 1
20424
+ 1
20425
+ 1
20426
+ 1
20427
+ 1
20428
+ 1
20429
+ 1
20430
+ 1
20431
+ 1
20432
+ 1
20433
+ 1
20434
+ 1
20435
+ 1
20436
+ 1
20437
+ 1
20438
+ 1
20439
+ 1
20440
+ 1
20441
+ 1
20442
+ 1
20443
+ 1
20444
+ 1
20445
+ 1
20446
+ 1
20447
+ 1
20448
+ 1
20449
+ 1
20450
+ 1
20451
+ 1
20452
+ 1
20453
+ 1
20454
+ 1
20455
+ 1
20456
+ 1
20457
+ 1
20458
+ 1
20459
+ 1
20460
+ 1
20461
+ 1
20462
+ 1
20463
+ 1
20464
+ 1
20465
+ 1
20466
+ 1
20467
+ 1
20468
+ 1
20469
+ 1
20470
+ 1
20471
+ 1
20472
+ 1
20473
+ 1
20474
+ 1
20475
+ 1
20476
+ 1
20477
+ 1
20478
+ 1
20479
+ 1
20480
+ 1
20481
+ 1
20482
+ 1
20483
+ 1
20484
+ 1
20485
+ 1
20486
+ 1
20487
+ 1
20488
+ 1
20489
+ 1
20490
+ 1
20491
+ 1
20492
+ 1
20493
+ 1
20494
+ 1
20495
+ 1
20496
+ 1
20497
+ 1
20498
+ 1
20499
+ 1
20500
+ 1
20501
+ 1
20502
+ 1
20503
+ 1
20504
+ 1
20505
+ 1
20506
+ 1
20507
+ 1
20508
+ 1
20509
+ 1
20510
+ 1
20511
+ 1
20512
+ 1
20513
+ 1
20514
+ 1
20515
+ 1
20516
+ 1
20517
+ 1
20518
+ 1
20519
+ 1
20520
+ 1
20521
+ 1
20522
+ 1
20523
+ 1
20524
+ 1
20525
+ 1
20526
+ 1
20527
+ 1
20528
+ 1
20529
+ 1
20530
+ 1
20531
+ 1
20532
+ 1
20533
+ 1
20534
+ 1
20535
+ 1
20536
+ 1
20537
+ 1
20538
+ 1
20539
+ 1
20540
+ 1
20541
+ 1
20542
+ 1
20543
+ 1
20544
+ 1
20545
+ 1
20546
+ 1
20547
+ 1
20548
+ 1
20549
+ 1
20550
+ 1
20551
+ 1
20552
+ 1
20553
+ 1
20554
+ 1
20555
+ 1
20556
+ 1
20557
+ 1
20558
+ 1
20559
+ 1
20560
+ 1
20561
+ 1
20562
+ 1
20563
+ 1
20564
+ 1
20565
+ 1
20566
+ 1
20567
+ 1
20568
+ 1
20569
+ 1
20570
+ 1
20571
+ 1
20572
+ 1
20573
+ 1
20574
+ 1
20575
+ 1
20576
+ 1
20577
+ 1
20578
+ 1
20579
+ 1
20580
+ 1
20581
+ 1
20582
+ 1
20583
+ 1
20584
+ 1
20585
+ 1
20586
+ 1
20587
+ 1
20588
+ 1
20589
+ 1
20590
+ 1
20591
+ 1
20592
+ 1
20593
+ 1
20594
+ 1
20595
+ 1
20596
+ 1
20597
+ 1
20598
+ 1
20599
+ 1
20600
+ 1
20601
+ 1
20602
+ 1
20603
+ 1
20604
+ 1
20605
+ 1
20606
+ 1
20607
+ 1
20608
+ 1
20609
+ 1
20610
+ 1
20611
+ 1
20612
+ 1
20613
+ 1
20614
+ 1
20615
+ 1
20616
+ 1
20617
+ 1
20618
+ 1
20619
+ 1
20620
+ 1
20621
+ 1
20622
+ 1
20623
+ 1
20624
+ 1
20625
+ 1
20626
+ 1
20627
+ 1
20628
+ 1
20629
+ 1
20630
+ 1
20631
+ 1
20632
+ 1
20633
+ 1
20634
+ 1
20635
+ 1
20636
+ 1
20637
+ 1
20638
+ 1
20639
+ 1
20640
+ 1
20641
+ 1
20642
+ 1
20643
+ 1
20644
+ 1
20645
+ 1
20646
+ 1
20647
+ 1
20648
+ 1
20649
+ 1
20650
+ 1
20651
+ 1
20652
+ 1
20653
+ 1
20654
+ 1
20655
+ 1
20656
+ 1
20657
+ 1
20658
+ 1
20659
+ 1
20660
+ 1
20661
+ 1
20662
+ 1
20663
+ 1
20664
+ 1
20665
+ 1
20666
+ 1
20667
+ 1
20668
+ 1
20669
+ 1
20670
+ 1
20671
+ 1
20672
+ 1
20673
+ 1
20674
+ 1
20675
+ 1
20676
+ 1
20677
+ 1
20678
+ 1
20679
+ 1
20680
+ 1
20681
+ 1
20682
+ 1
20683
+ 1
20684
+ 1
20685
+ 1
20686
+ 1
20687
+ 1
20688
+ 1
20689
+ 1
20690
+ 1
20691
+ 1
20692
+ 1
20693
+ 1
20694
+ 1
20695
+ 1
20696
+ 1
20697
+ 1
20698
+ 1
20699
+ 1
20700
+ 1
20701
+ 1
20702
+ 1
20703
+ 1
20704
+ 1
20705
+ 1
20706
+ 1
20707
+ 1
20708
+ 1
20709
+ 1
20710
+ 1
20711
+ 1
20712
+ 1
20713
+ 1
20714
+ 1
20715
+ 1
20716
+ 1
20717
+ 1
20718
+ 1
20719
+ 1
20720
+ 1
20721
+ 1
20722
+ 1
20723
+ 1
20724
+ 1
20725
+ 1
20726
+ 1
20727
+ 1
20728
+ 1
20729
+ 1
20730
+ 1
20731
+ 1
20732
+ 1
20733
+ 1
20734
+ 1
20735
+ 1
20736
+ 1
20737
+ 1
20738
+ 1
20739
+ 1
20740
+ 1
20741
+ 1
20742
+ 1
20743
+ 1
20744
+ 1
20745
+ 1
20746
+ 1
20747
+ 1
20748
+ 1
20749
+ 1
20750
+ 1
20751
+ 1
20752
+ 1
20753
+ 1
20754
+ 1
20755
+ 1
20756
+ 1
20757
+ 1
20758
+ 1
20759
+ 1
20760
+ 1
20761
+ 1
20762
+ 1
20763
+ 1
20764
+ 1
20765
+ 1
20766
+ 1
20767
+ 1
20768
+ 1
20769
+ 1
20770
+ 1
20771
+ 1
20772
+ 1
20773
+ 1
20774
+ 1
20775
+ 1
20776
+ 1
20777
+ 1
20778
+ 1
20779
+ 1
20780
+ 1
20781
+ 1
20782
+ 1
20783
+ 1
20784
+ 1
20785
+ 1
20786
+ 1
20787
+ 1
20788
+ 1
20789
+ 1
20790
+ 1
20791
+ 1
20792
+ 1
20793
+ 1
20794
+ 1
20795
+ 1
20796
+ 1
20797
+ 1
20798
+ 1
20799
+ 1
20800
+ 1
20801
+ 1
20802
+ 1
20803
+ 1
20804
+ 1
20805
+ 1
20806
+ 1
20807
+ 1
20808
+ 1
20809
+ 1
20810
+ 1
20811
+ 1
20812
+ 1
20813
+ 1
20814
+ 1
20815
+ 1
20816
+ 1
20817
+ 1
20818
+ 1
20819
+ 1
20820
+ 1
20821
+ 1
20822
+ 1
20823
+ 1
20824
+ 1
20825
+ 1
20826
+ 1
20827
+ 1
20828
+ 1
20829
+ 1
20830
+ 1
20831
+ 1
20832
+ 1
20833
+ 1
20834
+ 1
20835
+ 1
20836
+ 1
20837
+ 1
20838
+ 1
20839
+ 1
20840
+ 1
20841
+ 1
20842
+ 1
20843
+ 1
20844
+ 1
20845
+ 1
20846
+ 1
20847
+ 1
20848
+ 1
20849
+ 1
20850
+ 1
20851
+ 1
20852
+ 1
20853
+ 1
20854
+ 1
20855
+ 1
20856
+ 1
20857
+ 1
20858
+ 1
20859
+ 1
20860
+ 1
20861
+ 1
20862
+ 1
20863
+ 1
20864
+ 1
20865
+ 1
20866
+ 1
20867
+ 1
20868
+ 1
20869
+ 1
20870
+ 1
20871
+ 1
20872
+ 1
20873
+ 1
20874
+ 1
20875
+ 1
20876
+ 1
20877
+ 1
20878
+ 1
20879
+ 1
20880
+ 1
20881
+ 1
20882
+ 1
20883
+ 1
20884
+ 1
20885
+ 1
20886
+ 1
20887
+ 1
20888
+ 1
20889
+ 1
20890
+ 1
20891
+ 1
20892
+ 1
20893
+ 1
20894
+ 1
20895
+ 1
20896
+ 1
20897
+ 1
20898
+ 1
20899
+ 1
20900
+ 1
20901
+ 1
20902
+ 1
20903
+ 1
20904
+ 1
20905
+ 1
20906
+ 1
20907
+ 1
20908
+ 1
20909
+ 1
20910
+ 1
20911
+ 1
20912
+ 1
20913
+ 1
20914
+ 1
20915
+ 1
20916
+ 1
20917
+ 1
20918
+ 1
20919
+ 1
20920
+ 1
20921
+ 1
20922
+ 1
20923
+ 1
20924
+ 1
20925
+ 1
20926
+ 1
20927
+ 1
20928
+ 1
20929
+ 1
20930
+ 1
20931
+ 1
20932
+ 1
20933
+ 1
20934
+ 1
20935
+ 1
20936
+ 1
20937
+ 1
20938
+ 1
20939
+ 1
20940
+ 1
20941
+ 1
20942
+ 1
20943
+ 1
20944
+ 1
20945
+ 1
20946
+ 1
20947
+ 1
20948
+ 1
20949
+ 1
20950
+ 1
20951
+ 1
20952
+ 1
20953
+ 1
20954
+ 1
20955
+ 1
20956
+ 1
20957
+ 1
20958
+ 1
20959
+ 1
20960
+ 1
20961
+ 1
20962
+ 1
20963
+ 1
20964
+ 1
20965
+ 1
20966
+ 1
20967
+ 1
20968
+ 1
20969
+ 1
20970
+ 1
20971
+ 1
20972
+ 1
20973
+ 1
20974
+ 1
20975
+ 1
20976
+ 1
20977
+ 1
20978
+ 1
20979
+ 1
20980
+ 1
20981
+ 1
20982
+ 1
20983
+ 1
20984
+ 1
20985
+ 1
20986
+ 1
20987
+ 1
20988
+ 1
20989
+ 1
20990
+ 1
20991
+ 1
20992
+ 1
20993
+ 1
20994
+ 1
20995
+ 1
20996
+ 1
20997
+ 1
20998
+ 1
20999
+ 1
21000
+ 1
21001
+ 1
21002
+ 1
21003
+ 1
21004
+ 1
21005
+ 1
21006
+ 1
21007
+ 1
21008
+ 1
21009
+ 1
21010
+ 1
21011
+ 1
21012
+ 1
21013
+ 1
21014
+ 1
21015
+ 1
21016
+ 1
21017
+ 1
21018
+ 1
21019
+ 1
21020
+ 1
21021
+ 1
21022
+ 1
21023
+ 1
21024
+ 1
21025
+ 1
21026
+ 1
21027
+ 1
21028
+ 1
21029
+ 1
21030
+ 1
21031
+ 1
21032
+ 1
21033
+ 1
21034
+ 1
21035
+ 1
21036
+ 1
21037
+ 1
21038
+ 1
21039
+ 1
21040
+ 1
21041
+ 1
21042
+ 1
21043
+ 1
21044
+ 1
21045
+ 1
21046
+ 1
21047
+ 1
21048
+ 1
21049
+ 1
21050
+ 1
21051
+ 1
21052
+ 1
21053
+ 1
21054
+ 1
21055
+ 1
21056
+ 1
21057
+ 1
21058
+ 1
21059
+ 1
21060
+ 1
21061
+ 1
21062
+ 1
21063
+ 1
21064
+ 1
21065
+ 1
21066
+ 1
21067
+ 1
21068
+ 1
21069
+ 1
21070
+ 1
21071
+ 1
21072
+ 1
21073
+ 1
21074
+ 1
21075
+ 1
21076
+ 1
21077
+ 1
21078
+ 1
21079
+ 1
21080
+ 1
21081
+ 1
21082
+ 1
21083
+ 1
21084
+ 1
21085
+ 1
21086
+ 1
21087
+ 1
21088
+ 1
21089
+ 1
21090
+ 1
21091
+ 1
21092
+ 1
21093
+ 1
21094
+ 1
21095
+ 1
21096
+ 1
21097
+ 1
21098
+ 1
21099
+ 1
21100
+ 1
21101
+ 1
21102
+ 1
21103
+ 1
21104
+ 1
21105
+ 1
21106
+ 1
21107
+ 1
21108
+ 1
21109
+ 1
21110
+ 1
21111
+ 1
21112
+ 1
21113
+ 1
21114
+ 1
21115
+ 1
21116
+ 1
21117
+ 1
21118
+ 1
21119
+ 1
21120
+ 1
21121
+ 1
21122
+ 1
21123
+ 1
21124
+ 1
21125
+ 1
21126
+ 1
21127
+ 1
21128
+ 1
21129
+ 1
21130
+ 1
21131
+ 1
21132
+ 1
21133
+ 1
21134
+ 1
21135
+ 1
21136
+ 1
21137
+ 1
21138
+ 1
21139
+ 1
21140
+ 1
21141
+ 1
21142
+ 1
21143
+ 1
21144
+ 1
21145
+ 1
21146
+ 1
21147
+ 1
21148
+ 1
21149
+ 1
21150
+ 1
21151
+ 1
21152
+ 1
21153
+ 1
21154
+ 1
21155
+ 1
21156
+ 1
21157
+ 1
21158
+ 1
21159
+ 1
21160
+ 1
21161
+ 1
21162
+ 1
21163
+ 1
21164
+ 1
21165
+ 1
21166
+ 1
21167
+ 1
21168
+ 1
21169
+ 1
21170
+ 1
21171
+ 1
21172
+ 1
21173
+ 1
21174
+ 1
21175
+ 1
21176
+ 1
21177
+ 2
21178
+ 1
21179
+ 1
21180
+ 1
21181
+ 2
21182
+ 1
21183
+ 1
21184
+ 1
21185
+ 1
21186
+ 1
21187
+ 1
21188
+ 1
21189
+ 1
21190
+ 1
21191
+ 2
21192
+ 1
21193
+ 1
21194
+ 1
21195
+ 1
21196
+ 1
21197
+ 1
21198
+ 1
21199
+ 1
21200
+ 1
21201
+ 1
21202
+ 1
21203
+ 1
21204
+ 1
21205
+ 1
21206
+ 1
21207
+ 1
21208
+ 1
21209
+ 1
21210
+ 1
21211
+ 1
21212
+ 1
21213
+ 1
21214
+ 1
21215
+ 1
21216
+ 1
21217
+ 1
21218
+ 1
21219
+ 1
21220
+ 1
21221
+ 1
21222
+ 1
21223
+ 1
21224
+ 1
21225
+ 1
21226
+ 1
21227
+ 1
21228
+ 1
21229
+ 1
21230
+ 1
21231
+ 1
21232
+ 1
21233
+ 1
21234
+ 1
21235
+ 1
21236
+ 1
21237
+ 1
21238
+ 1
21239
+ 1
21240
+ 1
21241
+ 1
21242
+ 1
21243
+ 1
21244
+ 1
21245
+ 1
21246
+ 1
21247
+ 1
21248
+ 1
21249
+ 1
21250
+ 1
21251
+ 1
21252
+ 1
21253
+ 1
21254
+ 1
21255
+ 1
21256
+ 1
21257
+ 1
21258
+ 1
21259
+ 1
21260
+ 1
21261
+ 1
21262
+ 1
21263
+ 1
21264
+ 1
21265
+ 1
21266
+ 1
21267
+ 1
21268
+ 1
21269
+ 1
21270
+ 1
21271
+ 1
21272
+ 1
21273
+ 1
21274
+ 1
21275
+ 1
21276
+ 1
21277
+ 1
21278
+ 1
21279
+ 1
21280
+ 1
21281
+ 1
21282
+ 1
21283
+ 1
21284
+ 1
21285
+ 1
21286
+ 1
21287
+ 1
21288
+ 1
21289
+ 1
21290
+ 1
21291
+ 1
21292
+ 1
21293
+ 1
21294
+ 1
21295
+ 1
21296
+ 1
21297
+ 1
21298
+ 1
21299
+ 1
21300
+ 1
21301
+ 1
21302
+ 1
21303
+ 1
21304
+ 1
21305
+ 1
21306
+ 1
21307
+ 1
21308
+ 1
21309
+ 1
21310
+ 1
21311
+ 1
21312
+ 1
21313
+ 1
21314
+ 1
21315
+ 1
21316
+ 1
21317
+ 1
21318
+ 1
21319
+ 1
21320
+ 1
21321
+ 1
21322
+ 1
21323
+ 1
21324
+ 1
21325
+ 1
21326
+ 1
21327
+ 1
21328
+ 1
21329
+ 1
21330
+ 1
21331
+ 1
21332
+ 1
21333
+ 1
21334
+ 1
21335
+ 1
21336
+ 1
21337
+ 1
21338
+ 1
21339
+ 1
21340
+ 1
21341
+ 1
21342
+ 1
21343
+ 1
21344
+ 1
21345
+ 1
21346
+ 1
21347
+ 1
21348
+ 1
21349
+ 1
21350
+ 1
21351
+ 1
21352
+ 1
21353
+ 1
21354
+ 1
21355
+ 1
21356
+ 1
21357
+ 1
21358
+ 1
21359
+ 1
21360
+ 1
21361
+ 1
21362
+ 1
21363
+ 1
21364
+ 1
21365
+ 1
21366
+ 1
21367
+ 1
21368
+ 1
21369
+ 1
21370
+ 1
21371
+ 1
21372
+ 1
21373
+ 1
21374
+ 1
21375
+ 1
21376
+ 1
21377
+ 1
21378
+ 1
21379
+ 1
21380
+ 1
21381
+ 1
21382
+ 1
21383
+ 1
21384
+ 1
21385
+ 1
21386
+ 1
21387
+ 1
21388
+ 1
21389
+ 1
21390
+ 1
21391
+ 1
21392
+ 1
21393
+ 1
21394
+ 1
21395
+ 1
21396
+ 1
21397
+ 1
21398
+ 1
21399
+ 1
21400
+ 1
21401
+ 1
21402
+ 1
21403
+ 1
21404
+ 1
21405
+ 1
21406
+ 1
21407
+ 1
21408
+ 1
21409
+ 1
21410
+ 1
21411
+ 1
21412
+ 1
21413
+ 1
21414
+ 1
21415
+ 1
21416
+ 1
21417
+ 1
21418
+ 1
21419
+ 1
21420
+ 1
21421
+ 1
21422
+ 1
21423
+ 1
21424
+ 1
21425
+ 1
21426
+ 1
21427
+ 1
21428
+ 1
21429
+ 1
21430
+ 1
21431
+ 1
21432
+ 1
21433
+ 1
21434
+ 1
21435
+ 1
21436
+ 1
21437
+ 1
21438
+ 1
21439
+ 1
21440
+ 1
21441
+ 1
21442
+ 1
21443
+ 1
21444
+ 1
21445
+ 1
21446
+ 1
21447
+ 1
21448
+ 1
21449
+ 1
21450
+ 1
21451
+ 1
21452
+ 1
21453
+ 1
21454
+ 1
21455
+ 1
21456
+ 1
21457
+ 1
21458
+ 1
21459
+ 1
21460
+ 1
21461
+ 1
21462
+ 1
21463
+ 1
21464
+ 1
21465
+ 1
21466
+ 1
21467
+ 1
21468
+ 1
21469
+ 1
21470
+ 1
21471
+ 1
21472
+ 1
21473
+ 1
21474
+ 1
21475
+ 1
21476
+ 1
21477
+ 1
21478
+ 1
21479
+ 1
21480
+ 1
21481
+ 1
21482
+ 1
21483
+ 1
21484
+ 1
21485
+ 1
21486
+ 1
21487
+ 1
21488
+ 1
21489
+ 1
21490
+ 1
21491
+ 1
21492
+ 1
21493
+ 1
21494
+ 1
21495
+ 1
21496
+ 1
21497
+ 1
21498
+ 1
21499
+ 1
21500
+ 1
21501
+ 1
21502
+ 1
21503
+ 1
21504
+ 1
21505
+ 1
21506
+ 1
21507
+ 1
21508
+ 1
21509
+ 1
21510
+ 1
21511
+ 1
21512
+ 1
21513
+ 1
21514
+ 1
21515
+ 1
21516
+ 1
21517
+ 1
21518
+ 1
21519
+ 1
21520
+ 1
21521
+ 1
21522
+ 1
21523
+ 1
21524
+ 1
21525
+ 1
21526
+ 1
21527
+ 1
21528
+ 1
21529
+ 1
21530
+ 1
21531
+ 1
21532
+ 1
21533
+ 1
21534
+ 1
21535
+ 1
21536
+ 1
21537
+ 1
21538
+ 1
21539
+ 1
21540
+ 1
21541
+ 1
21542
+ 1
21543
+ 1
21544
+ 1
21545
+ 1
21546
+ 1
21547
+ 1
21548
+ 1
21549
+ 1
21550
+ 1
21551
+ 1
21552
+ 1
21553
+ 1
21554
+ 1
21555
+ 1
21556
+ 1
21557
+ 1
21558
+ 1
21559
+ 1
21560
+ 1
21561
+ 1
21562
+ 1
21563
+ 1
21564
+ 1
21565
+ 1
21566
+ 1
21567
+ 1
21568
+ 1
21569
+ 1
21570
+ 1
21571
+ 1
21572
+ 1
21573
+ 1
21574
+ 1
21575
+ 1
21576
+ 1
21577
+ 1
21578
+ 1
21579
+ 1
21580
+ 1
21581
+ 1
21582
+ 1
21583
+ 1
21584
+ 1
21585
+ 1
21586
+ 1
21587
+ 1
21588
+ 1
21589
+ 1
21590
+ 1
21591
+ 1
21592
+ 1
21593
+ 1
21594
+ 1
21595
+ 1
21596
+ 1
21597
+ 1
21598
+ 1
21599
+ 1
21600
+ 1
21601
+ 1
21602
+ 1
21603
+ 1
21604
+ 1
21605
+ 1
21606
+ 1
21607
+ 1
21608
+ 1
21609
+ 1
21610
+ 1
21611
+ 1
21612
+ 1
21613
+ 1
21614
+ 1
21615
+ 1
21616
+ 1
21617
+ 1
21618
+ 1
21619
+ 1
21620
+ 1
21621
+ 1
21622
+ 1
21623
+ 1
21624
+ 1
21625
+ 1
21626
+ 1
21627
+ 1
21628
+ 1
21629
+ 1
21630
+ 1
21631
+ 1
21632
+ 1
21633
+ 1
21634
+ 1
21635
+ 1
21636
+ 1
21637
+ 1
21638
+ 1
21639
+ 1
21640
+ 1
21641
+ 1
21642
+ 1
21643
+ 1
21644
+ 1
21645
+ 1
21646
+ 1
21647
+ 1
21648
+ 1
21649
+ 1
21650
+ 1
21651
+ 1
21652
+ 1
21653
+ 1
21654
+ 1
21655
+ 1
21656
+ 1
21657
+ 1
21658
+ 1
21659
+ 1
21660
+ 1
21661
+ 1
21662
+ 1
21663
+ 1
21664
+ 1
21665
+ 1
21666
+ 1
21667
+ 1
21668
+ 1
21669
+ 1
21670
+ 1
21671
+ 1
21672
+ 1
21673
+ 1
21674
+ 1
21675
+ 1
21676
+ 1
21677
+ 1
21678
+ 1
21679
+ 1
21680
+ 1
21681
+ 1
21682
+ 1
21683
+ 1
21684
+ 1
21685
+ 1
21686
+ 1
21687
+ 1
21688
+ 1
21689
+ 1
21690
+ 1
21691
+ 1
21692
+ 1
21693
+ 1
21694
+ 1
21695
+ 1
21696
+ 1
21697
+ 1
21698
+ 1
21699
+ 1
21700
+ 1
21701
+ 1
21702
+ 1
21703
+ 1
21704
+ 1
21705
+ 1
21706
+ 1
21707
+ 1
21708
+ 1
21709
+ 1
21710
+ 1
21711
+ 1
21712
+ 1
21713
+ 1
21714
+ 1
21715
+ 1
21716
+ 1
21717
+ 1
21718
+ 1
21719
+ 1
21720
+ 1
21721
+ 1
21722
+ 1
21723
+ 1
21724
+ 1
21725
+ 1
21726
+ 1
21727
+ 1
21728
+ 1
21729
+ 1
21730
+ 1
21731
+ 1
21732
+ 1
21733
+ 1
21734
+ 1
21735
+ 1
21736
+ 1
21737
+ 1
21738
+ 1
21739
+ 1
21740
+ 1
21741
+ 1
21742
+ 1
21743
+ 1
21744
+ 1
21745
+ 1
21746
+ 1
21747
+ 1
21748
+ 1
21749
+ 1
21750
+ 1
21751
+ 1
21752
+ 1
21753
+ 1
21754
+ 1
21755
+ 1
21756
+ 1
21757
+ 1
21758
+ 1
21759
+ 1
21760
+ 1
21761
+ 1
21762
+ 1
21763
+ 1
21764
+ 1
21765
+ 1
21766
+ 1
21767
+ 1
21768
+ 1
21769
+ 1
21770
+ 1
21771
+ 1
21772
+ 1
21773
+ 1
21774
+ 1
21775
+ 1
21776
+ 1
21777
+ 1
21778
+ 1
21779
+ 1
21780
+ 1
21781
+ 1
21782
+ 1
21783
+ 1
21784
+ 1
21785
+ 1
21786
+ 1
21787
+ 1
21788
+ 1
21789
+ 1
21790
+ 1
21791
+ 1
21792
+ 1
21793
+ 1
21794
+ 1
21795
+ 1
21796
+ 1
21797
+ 1
21798
+ 1
21799
+ 1
21800
+ 1
21801
+ 1
21802
+ 1
21803
+ 2
21804
+ 2
21805
+ 1
21806
+ 1
21807
+ 2
21808
+ 1
21809
+ 1
21810
+ 1
21811
+ 1
21812
+ 1
21813
+ 1
21814
+ 1
21815
+ 1
21816
+ 1
21817
+ 1
21818
+ 1
21819
+ 1
21820
+ 1
21821
+ 1
21822
+ 1
21823
+ 1
21824
+ 1
21825
+ 1
21826
+ 1
21827
+ 1
21828
+ 1
21829
+ 1
21830
+ 1
21831
+ 1
21832
+ 1
21833
+ 1
21834
+ 1
21835
+ 1
21836
+ 1
21837
+ 1
21838
+ 1
21839
+ 1
21840
+ 1
21841
+ 1
21842
+ 1
21843
+ 1
21844
+ 1
21845
+ 1
21846
+ 1
21847
+ 1
21848
+ 1
21849
+ 1
21850
+ 1
21851
+ 1
21852
+ 1
21853
+ 1
21854
+ 1
21855
+ 1
21856
+ 1
21857
+ 1
21858
+ 1
21859
+ 1
21860
+ 1
21861
+ 1
21862
+ 1
21863
+ 1
21864
+ 1
21865
+ 1
21866
+ 1
21867
+ 1
21868
+ 1
21869
+ 1
21870
+ 1
21871
+ 1
21872
+ 1
21873
+ 1
21874
+ 1
21875
+ 1
21876
+ 1
21877
+ 1
21878
+ 1
21879
+ 1
21880
+ 1
21881
+ 1
21882
+ 1
21883
+ 1
21884
+ 1
21885
+ 1
21886
+ 1
21887
+ 1
21888
+ 1
21889
+ 1
21890
+ 1
21891
+ 1
21892
+ 1
21893
+ 1
21894
+ 1
21895
+ 1
21896
+ 1
21897
+ 1
21898
+ 1
21899
+ 1
21900
+ 1
21901
+ 1
21902
+ 1
21903
+ 1
21904
+ 1
21905
+ 1
21906
+ 1
21907
+ 1
21908
+ 1
21909
+ 1
21910
+ 1
21911
+ 1
21912
+ 1
21913
+ 1
21914
+ 1
21915
+ 1
21916
+ 1
21917
+ 1
21918
+ 1
21919
+ 1
21920
+ 1
21921
+ 1
21922
+ 1
21923
+ 1
21924
+ 1
21925
+ 1
21926
+ 1
21927
+ 1
21928
+ 1
21929
+ 1
21930
+ 1
21931
+ 1
21932
+ 1
21933
+ 1
21934
+ 1
21935
+ 1
21936
+ 1
21937
+ 1
21938
+ 1
21939
+ 1
21940
+ 1
21941
+ 1
21942
+ 1
21943
+ 1
21944
+ 1
21945
+ 1
21946
+ 1
21947
+ 1
21948
+ 1
21949
+ 1
21950
+ 1
21951
+ 1
21952
+ 1
21953
+ 1
21954
+ 1
21955
+ 1
21956
+ 1
21957
+ 1
21958
+ 1
21959
+ 1
21960
+ 1
21961
+ 1
21962
+ 1
21963
+ 1
21964
+ 1
21965
+ 1
21966
+ 1
21967
+ 1
21968
+ 1
21969
+ 1
21970
+ 1
21971
+ 1
21972
+ 1
21973
+ 1
21974
+ 1
21975
+ 1
21976
+ 1
21977
+ 1
21978
+ 1
21979
+ 1
21980
+ 1
21981
+ 1
21982
+ 1
21983
+ 1
21984
+ 1
21985
+ 1
21986
+ 1
21987
+ 1
21988
+ 1
21989
+ 1
21990
+ 1
21991
+ 1
21992
+ 1
21993
+ 1
21994
+ 1
21995
+ 1
21996
+ 1
21997
+ 1
21998
+ 1
21999
+ 1
22000
+ 1
22001
+ 1
22002
+ 1
22003
+ 1
22004
+ 1
22005
+ 1
22006
+ 1
22007
+ 1
22008
+ 1
22009
+ 1
22010
+ 1
22011
+ 1
22012
+ 1
22013
+ 1
22014
+ 1
22015
+ 1
22016
+ 1
22017
+ 1
22018
+ 1
22019
+ 1
22020
+ 1
22021
+ 1
22022
+ 1
22023
+ 1
22024
+ 1
22025
+ 1
22026
+ 1
22027
+ 1
22028
+ 1
22029
+ 1
22030
+ 1
22031
+ 1
22032
+ 1
22033
+ 1
22034
+ 1
22035
+ 1
22036
+ 1
22037
+ 1
22038
+ 1
22039
+ 1
22040
+ 1
22041
+ 1
22042
+ 1
22043
+ 1
22044
+ 1
22045
+ 1
22046
+ 1
22047
+ 1
22048
+ 1
22049
+ 1
22050
+ 1
22051
+ 1
22052
+ 1
22053
+ 1
22054
+ 1
22055
+ 1
22056
+ 1
22057
+ 1
22058
+ 1
22059
+ 1
22060
+ 1
22061
+ 1
22062
+ 1
22063
+ 1
22064
+ 1
22065
+ 1
22066
+ 1
22067
+ 1
22068
+ 1
22069
+ 1
22070
+ 1
22071
+ 1
22072
+ 1
22073
+ 1
22074
+ 1
22075
+ 1
22076
+ 1
22077
+ 1
22078
+ 1
22079
+ 1
22080
+ 1
22081
+ 1
22082
+ 1
22083
+ 1
22084
+ 1
22085
+ 1
22086
+ 1
22087
+ 1
22088
+ 1
22089
+ 1
22090
+ 1
22091
+ 1
22092
+ 1
22093
+ 1
22094
+ 1
22095
+ 1
22096
+ 1
22097
+ 1
22098
+ 1
22099
+ 1
22100
+ 1
22101
+ 1
22102
+ 1
22103
+ 1
22104
+ 1
22105
+ 1
22106
+ 1
22107
+ 1
22108
+ 1
22109
+ 1
22110
+ 1
22111
+ 1
22112
+ 1
22113
+ 1
22114
+ 1
22115
+ 1
22116
+ 1
22117
+ 1
22118
+ 1
22119
+ 1
22120
+ 1
22121
+ 1
22122
+ 1
22123
+ 1
22124
+ 1
22125
+ 1
22126
+ 1
22127
+ 1
22128
+ 1
22129
+ 1
22130
+ 1
22131
+ 1
22132
+ 1
22133
+ 1
22134
+ 1
22135
+ 1
22136
+ 1
22137
+ 1
22138
+ 1
22139
+ 1
22140
+ 1
22141
+ 1
22142
+ 1
22143
+ 1
22144
+ 1
22145
+ 1
22146
+ 1
22147
+ 1
22148
+ 1
22149
+ 1
22150
+ 1
22151
+ 1
22152
+ 1
22153
+ 1
22154
+ 1
22155
+ 1
22156
+ 1
22157
+ 1
22158
+ 1
22159
+ 1
22160
+ 1
22161
+ 1
22162
+ 1
22163
+ 1
22164
+ 1
22165
+ 1
22166
+ 1
22167
+ 1
22168
+ 1
22169
+ 1
22170
+ 1
22171
+ 1
22172
+ 1
22173
+ 1
22174
+ 1
22175
+ 1
22176
+ 1
22177
+ 1
22178
+ 1
22179
+ 1
22180
+ 1
22181
+ 1
22182
+ 1
22183
+ 1
22184
+ 1
22185
+ 1
22186
+ 1
22187
+ 1
22188
+ 1
22189
+ 1
22190
+ 1
22191
+ 1
22192
+ 1
22193
+ 1
22194
+ 1
22195
+ 1
22196
+ 1
22197
+ 1
22198
+ 1
22199
+ 1
22200
+ 1
22201
+ 1
22202
+ 1
22203
+ 1
22204
+ 1
22205
+ 1
22206
+ 1
22207
+ 1
22208
+ 1
22209
+ 1
22210
+ 1
22211
+ 1
22212
+ 1
22213
+ 1
22214
+ 1
22215
+ 1
22216
+ 1
22217
+ 1
22218
+ 1
22219
+ 1
22220
+ 1
22221
+ 1
22222
+ 1
22223
+ 1
22224
+ 1
22225
+ 1
22226
+ 1
22227
+ 1
22228
+ 1
22229
+ 1
22230
+ 1
22231
+ 1
22232
+ 1
22233
+ 1
22234
+ 1
22235
+ 1
22236
+ 1
22237
+ 1
22238
+ 1
22239
+ 1
22240
+ 1
22241
+ 1
22242
+ 1
22243
+ 1
22244
+ 1
22245
+ 1
22246
+ 1
22247
+ 1
22248
+ 1
22249
+ 1
22250
+ 1
22251
+ 1
22252
+ 1
22253
+ 1
22254
+ 1
22255
+ 1
22256
+ 1
22257
+ 1
22258
+ 1
22259
+ 1
22260
+ 1
22261
+ 1
22262
+ 1
22263
+ 1
22264
+ 1
22265
+ 1
22266
+ 1
22267
+ 1
22268
+ 1
22269
+ 1
22270
+ 1
22271
+ 1
22272
+ 1
22273
+ 1
22274
+ 1
22275
+ 1
22276
+ 1
22277
+ 1
22278
+ 1
22279
+ 1
22280
+ 1
22281
+ 1
22282
+ 1
22283
+ 1
22284
+ 1
22285
+ 1
22286
+ 1
22287
+ 1
22288
+ 1
22289
+ 1
22290
+ 1
22291
+ 1
22292
+ 1
22293
+ 1
22294
+ 1
22295
+ 1
22296
+ 1
22297
+ 1
22298
+ 1
22299
+ 1
22300
+ 1
22301
+ 1
22302
+ 1
22303
+ 1
22304
+ 1
22305
+ 1
22306
+ 1
22307
+ 1
22308
+ 1
22309
+ 1
22310
+ 1
22311
+ 1
22312
+ 1
22313
+ 1
22314
+ 1
22315
+ 1
22316
+ 1
22317
+ 1
22318
+ 1
22319
+ 1
22320
+ 1
22321
+ 1
22322
+ 1
22323
+ 1
22324
+ 1
22325
+ 1
22326
+ 1
22327
+ 1
22328
+ 1
22329
+ 1
22330
+ 1
22331
+ 1
22332
+ 1
22333
+ 1
22334
+ 1
22335
+ 1
22336
+ 1
22337
+ 1
22338
+ 1
22339
+ 1
22340
+ 1
22341
+ 1
22342
+ 1
22343
+ 1
22344
+ 1
22345
+ 1
22346
+ 1
22347
+ 1
22348
+ 1
22349
+ 1
22350
+ 1
22351
+ 1
22352
+ 1
22353
+ 1
22354
+ 1
22355
+ 1
22356
+ 1
22357
+ 1
22358
+ 1
22359
+ 1
22360
+ 1
22361
+ 1
22362
+ 1
22363
+ 1
22364
+ 1
22365
+ 1
22366
+ 1
22367
+ 1
22368
+ 1
22369
+ 1
22370
+ 1
22371
+ 1
22372
+ 1
22373
+ 1
22374
+ 1
22375
+ 1
22376
+ 1
22377
+ 1
22378
+ 1
22379
+ 1
22380
+ 1
22381
+ 1
22382
+ 1
22383
+ 1
22384
+ 1
22385
+ 1
22386
+ 1
22387
+ 1
22388
+ 1
22389
+ 1
22390
+ 1
22391
+ 1
22392
+ 1
22393
+ 1
22394
+ 1
22395
+ 1
22396
+ 1
22397
+ 1
22398
+ 1
22399
+ 1
22400
+ 1
22401
+ 1
22402
+ 1
22403
+ 1
22404
+ 1
22405
+ 1
22406
+ 1
22407
+ 1
22408
+ 1
22409
+ 1
22410
+ 1
22411
+ 1
22412
+ 1
22413
+ 1
22414
+ 1
22415
+ 1
22416
+ 1
22417
+ 1
22418
+ 1
22419
+ 1
22420
+ 1
22421
+ 1
22422
+ 1
22423
+ 1
22424
+ 1
22425
+ 1
22426
+ 1
22427
+ 1
22428
+ 1
22429
+ 1
22430
+ 1
22431
+ 1
22432
+ 1
22433
+ 1
22434
+ 1
22435
+ 1
22436
+ 1
22437
+ 1
22438
+ 1
22439
+ 1
22440
+ 1
22441
+ 1
22442
+ 1
22443
+ 1
22444
+ 1
22445
+ 1
22446
+ 1
22447
+ 1
22448
+ 1
22449
+ 1
22450
+ 1
22451
+ 1
22452
+ 1
22453
+ 1
22454
+ 1
22455
+ 1
22456
+ 1
22457
+ 1
22458
+ 1
22459
+ 1
22460
+ 1
22461
+ 1
22462
+ 1
22463
+ 1
22464
+ 1
22465
+ 1
22466
+ 1
22467
+ 1
22468
+ 1
22469
+ 1
22470
+ 1
22471
+ 1
22472
+ 1
22473
+ 1
22474
+ 1
22475
+ 1
22476
+ 1
22477
+ 1
22478
+ 1
22479
+ 1
22480
+ 1
22481
+ 1
22482
+ 1
22483
+ 1
22484
+ 1
22485
+ 1
22486
+ 1
22487
+ 1
22488
+ 1
22489
+ 1
22490
+ 1
22491
+ 1
22492
+ 1
22493
+ 1
22494
+ 1
22495
+ 1
22496
+ 1
22497
+ 1
22498
+ 1
22499
+ 1
22500
+ 1
22501
+ 1
22502
+ 1
22503
+ 1
22504
+ 1
22505
+ 1
22506
+ 1
22507
+ 1
22508
+ 1
22509
+ 1
22510
+ 1
22511
+ 1
22512
+ 1
22513
+ 1
22514
+ 1
22515
+ 1
22516
+ 1
22517
+ 1
22518
+ 1
22519
+ 1
22520
+ 1
22521
+ 1
22522
+ 1
22523
+ 1
22524
+ 1
22525
+ 1
22526
+ 1
22527
+ 1
22528
+ 1
22529
+ 1
22530
+ 1
22531
+ 1
22532
+ 1
22533
+ 1
22534
+ 1
22535
+ 1
22536
+ 1
22537
+ 1
22538
+ 1
22539
+ 1
22540
+ 1
22541
+ 1
22542
+ 1
22543
+ 1
22544
+ 1
22545
+ 1
22546
+ 1
22547
+ 1
22548
+ 1
22549
+ 1
22550
+ 1
22551
+ 1
22552
+ 1
22553
+ 1
22554
+ 1
22555
+ 1
22556
+ 1
22557
+ 1
22558
+ 1
22559
+ 1
22560
+ 1
22561
+ 1
22562
+ 1
22563
+ 1
22564
+ 1
22565
+ 1
22566
+ 1
22567
+ 1
22568
+ 1
22569
+ 1
22570
+ 1
22571
+ 1
22572
+ 1
22573
+ 1
22574
+ 1
22575
+ 1
22576
+ 1
22577
+ 1
22578
+ 1
22579
+ 1
22580
+ 1
22581
+ 1
22582
+ 1
22583
+ 1
22584
+ 1
22585
+ 2
22586
+ 1
22587
+ 1
22588
+ 1
22589
+ 1
22590
+ 1
22591
+ 1
22592
+ 1
22593
+ 1
22594
+ 1
22595
+ 1
22596
+ 1
22597
+ 1
22598
+ 1
22599
+ 1
22600
+ 1
22601
+ 1
22602
+ 1
22603
+ 1
22604
+ 1
22605
+ 1
22606
+ 1
22607
+ 1
22608
+ 1
22609
+ 1
22610
+ 1
22611
+ 1
22612
+ 1
22613
+ 1
22614
+ 1
22615
+ 1
22616
+ 1
22617
+ 1
22618
+ 1
22619
+ 1
22620
+ 1
22621
+ 1
22622
+ 1
22623
+ 1
22624
+ 1
22625
+ 1
22626
+ 1
22627
+ 1
22628
+ 1
22629
+ 1
22630
+ 1
22631
+ 1
22632
+ 1
22633
+ 1
22634
+ 1
22635
+ 1
22636
+ 1
22637
+ 1
22638
+ 1
22639
+ 1
22640
+ 1
22641
+ 1
22642
+ 1
22643
+ 1
22644
+ 1
22645
+ 1
22646
+ 1
22647
+ 1
22648
+ 1
22649
+ 1
22650
+ 1
22651
+ 1
22652
+ 1
22653
+ 1
22654
+ 1
22655
+ 1
22656
+ 1
22657
+ 1
22658
+ 1
22659
+ 1
22660
+ 1
22661
+ 1
22662
+ 1
22663
+ 1
22664
+ 1
22665
+ 1
22666
+ 1
22667
+ 1
22668
+ 1
22669
+ 1
22670
+ 1
22671
+ 1
22672
+ 1
22673
+ 1
22674
+ 1
22675
+ 1
22676
+ 1
22677
+ 1
22678
+ 1
22679
+ 1
22680
+ 1
22681
+ 1
22682
+ 1
22683
+ 1
22684
+ 1
22685
+ 1
22686
+ 1
22687
+ 1
22688
+ 1
22689
+ 1
22690
+ 1
22691
+ 1
22692
+ 1
22693
+ 1
22694
+ 1
22695
+ 1
22696
+ 1
22697
+ 1
22698
+ 1
22699
+ 1
22700
+ 1
22701
+ 1
22702
+ 1
22703
+ 1
22704
+ 1
22705
+ 1
22706
+ 1
22707
+ 1
22708
+ 1
22709
+ 1
22710
+ 1
22711
+ 1
22712
+ 1
22713
+ 1
22714
+ 1
22715
+ 1
22716
+ 1
22717
+ 1
22718
+ 1
22719
+ 1
22720
+ 1
22721
+ 1
22722
+ 1
22723
+ 1
22724
+ 1
22725
+ 1
22726
+ 1
22727
+ 1
22728
+ 1
22729
+ 1
22730
+ 1
22731
+ 1
22732
+ 1
22733
+ 1
22734
+ 1
22735
+ 1
22736
+ 1
22737
+ 1
22738
+ 1
22739
+ 1
22740
+ 1
22741
+ 1
22742
+ 1
22743
+ 1
22744
+ 1
22745
+ 1
22746
+ 1
22747
+ 1
22748
+ 1
22749
+ 1
22750
+ 1
22751
+ 1
22752
+ 1
22753
+ 1
22754
+ 1
22755
+ 1
22756
+ 1
22757
+ 1
22758
+ 1
22759
+ 1
22760
+ 1
22761
+ 1
22762
+ 1
22763
+ 1
22764
+ 1
22765
+ 1
22766
+ 1
22767
+ 1
22768
+ 1
22769
+ 1
22770
+ 1
22771
+ 1
22772
+ 1
22773
+ 1
22774
+ 1
22775
+ 1
22776
+ 1
22777
+ 1
22778
+ 1
22779
+ 1
22780
+ 1
22781
+ 1
22782
+ 1
22783
+ 1
22784
+ 1
22785
+ 1
22786
+ 1
22787
+ 1
22788
+ 1
22789
+ 1
22790
+ 1
22791
+ 1
22792
+ 1
22793
+ 1
22794
+ 1
22795
+ 1
22796
+ 1
22797
+ 1
22798
+ 1
22799
+ 1
22800
+ 1
22801
+ 1
22802
+ 1
22803
+ 1
22804
+ 1
22805
+ 1
22806
+ 1
22807
+ 1
22808
+ 1
22809
+ 1
22810
+ 1
22811
+ 1
22812
+ 1
22813
+ 1
22814
+ 1
22815
+ 1
22816
+ 1
22817
+ 1
22818
+ 1
22819
+ 1
22820
+ 1
22821
+ 1
22822
+ 1
22823
+ 1
22824
+ 1
22825
+ 1
22826
+ 1
22827
+ 1
22828
+ 1
22829
+ 1
22830
+ 1
22831
+ 1
22832
+ 1
22833
+ 1
22834
+ 1
22835
+ 1
22836
+ 1
22837
+ 1
22838
+ 1
22839
+ 1
22840
+ 1
22841
+ 1
22842
+ 1
22843
+ 1
22844
+ 1
22845
+ 1
22846
+ 1
22847
+ 1
22848
+ 1
22849
+ 1
22850
+ 1
22851
+ 1
22852
+ 1
22853
+ 1
22854
+ 1
22855
+ 1
22856
+ 1
22857
+ 1
22858
+ 2
22859
+ 2
22860
+ 1
22861
+ 1
22862
+ 1
22863
+ 1
22864
+ 1
22865
+ 1
22866
+ 1
22867
+ 2
22868
+ 1
22869
+ 1
22870
+ 1
22871
+ 1
22872
+ 1
causalvideovae/eval/RAFT/core/corr.py ADDED
@@ -0,0 +1,91 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ import torch
2
+ import torch.nn.functional as F
3
+ from .utils.utils import bilinear_sampler, coords_grid
4
+
5
+ try:
6
+ import alt_cuda_corr
7
+ except:
8
+ # alt_cuda_corr is not compiled
9
+ pass
10
+
11
+
12
+ class CorrBlock:
13
+ def __init__(self, fmap1, fmap2, num_levels=4, radius=4):
14
+ self.num_levels = num_levels
15
+ self.radius = radius
16
+ self.corr_pyramid = []
17
+
18
+ # all pairs correlation
19
+ corr = CorrBlock.corr(fmap1, fmap2)
20
+
21
+ batch, h1, w1, dim, h2, w2 = corr.shape
22
+ corr = corr.reshape(batch*h1*w1, dim, h2, w2)
23
+
24
+ self.corr_pyramid.append(corr)
25
+ for i in range(self.num_levels-1):
26
+ corr = F.avg_pool2d(corr, 2, stride=2)
27
+ self.corr_pyramid.append(corr)
28
+
29
+ def __call__(self, coords):
30
+ r = self.radius
31
+ coords = coords.permute(0, 2, 3, 1)
32
+ batch, h1, w1, _ = coords.shape
33
+
34
+ out_pyramid = []
35
+ for i in range(self.num_levels):
36
+ corr = self.corr_pyramid[i]
37
+ dx = torch.linspace(-r, r, 2*r+1, device=coords.device)
38
+ dy = torch.linspace(-r, r, 2*r+1, device=coords.device)
39
+ delta = torch.stack(torch.meshgrid(dy, dx), axis=-1)
40
+
41
+ centroid_lvl = coords.reshape(batch*h1*w1, 1, 1, 2) / 2**i
42
+ delta_lvl = delta.view(1, 2*r+1, 2*r+1, 2)
43
+ coords_lvl = centroid_lvl + delta_lvl
44
+
45
+ corr = bilinear_sampler(corr, coords_lvl)
46
+ corr = corr.view(batch, h1, w1, -1)
47
+ out_pyramid.append(corr)
48
+
49
+ out = torch.cat(out_pyramid, dim=-1)
50
+ return out.permute(0, 3, 1, 2).contiguous().float()
51
+
52
+ @staticmethod
53
+ def corr(fmap1, fmap2):
54
+ batch, dim, ht, wd = fmap1.shape
55
+ fmap1 = fmap1.view(batch, dim, ht*wd)
56
+ fmap2 = fmap2.view(batch, dim, ht*wd)
57
+
58
+ corr = torch.matmul(fmap1.transpose(1,2), fmap2)
59
+ corr = corr.view(batch, ht, wd, 1, ht, wd)
60
+ return corr / torch.sqrt(torch.tensor(dim).float())
61
+
62
+
63
+ class AlternateCorrBlock:
64
+ def __init__(self, fmap1, fmap2, num_levels=4, radius=4):
65
+ self.num_levels = num_levels
66
+ self.radius = radius
67
+
68
+ self.pyramid = [(fmap1, fmap2)]
69
+ for i in range(self.num_levels):
70
+ fmap1 = F.avg_pool2d(fmap1, 2, stride=2)
71
+ fmap2 = F.avg_pool2d(fmap2, 2, stride=2)
72
+ self.pyramid.append((fmap1, fmap2))
73
+
74
+ def __call__(self, coords):
75
+ coords = coords.permute(0, 2, 3, 1)
76
+ B, H, W, _ = coords.shape
77
+ dim = self.pyramid[0][0].shape[1]
78
+
79
+ corr_list = []
80
+ for i in range(self.num_levels):
81
+ r = self.radius
82
+ fmap1_i = self.pyramid[0][0].permute(0, 2, 3, 1).contiguous()
83
+ fmap2_i = self.pyramid[i][1].permute(0, 2, 3, 1).contiguous()
84
+
85
+ coords_i = (coords / 2**i).reshape(B, 1, H, W, 2).contiguous()
86
+ corr, = alt_cuda_corr.forward(fmap1_i, fmap2_i, coords_i, r)
87
+ corr_list.append(corr.squeeze(1))
88
+
89
+ corr = torch.stack(corr_list, dim=1)
90
+ corr = corr.reshape(B, -1, H, W)
91
+ return corr / torch.sqrt(torch.tensor(dim).float())
causalvideovae/eval/RAFT/core/utils/frame_utils.py ADDED
@@ -0,0 +1,137 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ import numpy as np
2
+ from PIL import Image
3
+ from os.path import *
4
+ import re
5
+
6
+ import cv2
7
+ cv2.setNumThreads(0)
8
+ cv2.ocl.setUseOpenCL(False)
9
+
10
+ TAG_CHAR = np.array([202021.25], np.float32)
11
+
12
+ def readFlow(fn):
13
+ """ Read .flo file in Middlebury format"""
14
+ # Code adapted from:
15
+ # http://stackoverflow.com/questions/28013200/reading-middlebury-flow-files-with-python-bytes-array-numpy
16
+
17
+ # WARNING: this will work on little-endian architectures (eg Intel x86) only!
18
+ # print 'fn = %s'%(fn)
19
+ with open(fn, 'rb') as f:
20
+ magic = np.fromfile(f, np.float32, count=1)
21
+ if 202021.25 != magic:
22
+ print('Magic number incorrect. Invalid .flo file')
23
+ return None
24
+ else:
25
+ w = np.fromfile(f, np.int32, count=1)
26
+ h = np.fromfile(f, np.int32, count=1)
27
+ # print 'Reading %d x %d flo file\n' % (w, h)
28
+ data = np.fromfile(f, np.float32, count=2*int(w)*int(h))
29
+ # Reshape data into 3D array (columns, rows, bands)
30
+ # The reshape here is for visualization, the original code is (w,h,2)
31
+ return np.resize(data, (int(h), int(w), 2))
32
+
33
+ def readPFM(file):
34
+ file = open(file, 'rb')
35
+
36
+ color = None
37
+ width = None
38
+ height = None
39
+ scale = None
40
+ endian = None
41
+
42
+ header = file.readline().rstrip()
43
+ if header == b'PF':
44
+ color = True
45
+ elif header == b'Pf':
46
+ color = False
47
+ else:
48
+ raise Exception('Not a PFM file.')
49
+
50
+ dim_match = re.match(rb'^(\d+)\s(\d+)\s$', file.readline())
51
+ if dim_match:
52
+ width, height = map(int, dim_match.groups())
53
+ else:
54
+ raise Exception('Malformed PFM header.')
55
+
56
+ scale = float(file.readline().rstrip())
57
+ if scale < 0: # little-endian
58
+ endian = '<'
59
+ scale = -scale
60
+ else:
61
+ endian = '>' # big-endian
62
+
63
+ data = np.fromfile(file, endian + 'f')
64
+ shape = (height, width, 3) if color else (height, width)
65
+
66
+ data = np.reshape(data, shape)
67
+ data = np.flipud(data)
68
+ return data
69
+
70
+ def writeFlow(filename,uv,v=None):
71
+ """ Write optical flow to file.
72
+
73
+ If v is None, uv is assumed to contain both u and v channels,
74
+ stacked in depth.
75
+ Original code by Deqing Sun, adapted from Daniel Scharstein.
76
+ """
77
+ nBands = 2
78
+
79
+ if v is None:
80
+ assert(uv.ndim == 3)
81
+ assert(uv.shape[2] == 2)
82
+ u = uv[:,:,0]
83
+ v = uv[:,:,1]
84
+ else:
85
+ u = uv
86
+
87
+ assert(u.shape == v.shape)
88
+ height,width = u.shape
89
+ f = open(filename,'wb')
90
+ # write the header
91
+ f.write(TAG_CHAR)
92
+ np.array(width).astype(np.int32).tofile(f)
93
+ np.array(height).astype(np.int32).tofile(f)
94
+ # arrange into matrix form
95
+ tmp = np.zeros((height, width*nBands))
96
+ tmp[:,np.arange(width)*2] = u
97
+ tmp[:,np.arange(width)*2 + 1] = v
98
+ tmp.astype(np.float32).tofile(f)
99
+ f.close()
100
+
101
+
102
+ def readFlowKITTI(filename):
103
+ flow = cv2.imread(filename, cv2.IMREAD_ANYDEPTH|cv2.IMREAD_COLOR)
104
+ flow = flow[:,:,::-1].astype(np.float32)
105
+ flow, valid = flow[:, :, :2], flow[:, :, 2]
106
+ flow = (flow - 2**15) / 64.0
107
+ return flow, valid
108
+
109
+ def readDispKITTI(filename):
110
+ disp = cv2.imread(filename, cv2.IMREAD_ANYDEPTH) / 256.0
111
+ valid = disp > 0.0
112
+ flow = np.stack([-disp, np.zeros_like(disp)], -1)
113
+ return flow, valid
114
+
115
+
116
+ def writeFlowKITTI(filename, uv):
117
+ uv = 64.0 * uv + 2**15
118
+ valid = np.ones([uv.shape[0], uv.shape[1], 1])
119
+ uv = np.concatenate([uv, valid], axis=-1).astype(np.uint16)
120
+ cv2.imwrite(filename, uv[..., ::-1])
121
+
122
+
123
+ def read_gen(file_name, pil=False):
124
+ ext = splitext(file_name)[-1]
125
+ if ext == '.png' or ext == '.jpeg' or ext == '.ppm' or ext == '.jpg':
126
+ return Image.open(file_name)
127
+ elif ext == '.bin' or ext == '.raw':
128
+ return np.load(file_name)
129
+ elif ext == '.flo':
130
+ return readFlow(file_name).astype(np.float32)
131
+ elif ext == '.pfm':
132
+ flow = readPFM(file_name).astype(np.float32)
133
+ if len(flow.shape) == 2:
134
+ return flow
135
+ else:
136
+ return flow[:, :, :-1]
137
+ return []
causalvideovae/eval/RAFT/core/utils/utils.py ADDED
@@ -0,0 +1,82 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ import torch
2
+ import torch.nn.functional as F
3
+ import numpy as np
4
+ from scipy import interpolate
5
+
6
+
7
+ class InputPadder:
8
+ """ Pads images such that dimensions are divisible by 8 """
9
+ def __init__(self, dims, mode='sintel'):
10
+ self.ht, self.wd = dims[-2:]
11
+ pad_ht = (((self.ht // 8) + 1) * 8 - self.ht) % 8
12
+ pad_wd = (((self.wd // 8) + 1) * 8 - self.wd) % 8
13
+ if mode == 'sintel':
14
+ self._pad = [pad_wd//2, pad_wd - pad_wd//2, pad_ht//2, pad_ht - pad_ht//2]
15
+ else:
16
+ self._pad = [pad_wd//2, pad_wd - pad_wd//2, 0, pad_ht]
17
+
18
+ def pad(self, *inputs):
19
+ return [F.pad(x, self._pad, mode='replicate') for x in inputs]
20
+
21
+ def unpad(self,x):
22
+ ht, wd = x.shape[-2:]
23
+ c = [self._pad[2], ht-self._pad[3], self._pad[0], wd-self._pad[1]]
24
+ return x[..., c[0]:c[1], c[2]:c[3]]
25
+
26
+ def forward_interpolate(flow):
27
+ flow = flow.detach().cpu().numpy()
28
+ dx, dy = flow[0], flow[1]
29
+
30
+ ht, wd = dx.shape
31
+ x0, y0 = np.meshgrid(np.arange(wd), np.arange(ht))
32
+
33
+ x1 = x0 + dx
34
+ y1 = y0 + dy
35
+
36
+ x1 = x1.reshape(-1)
37
+ y1 = y1.reshape(-1)
38
+ dx = dx.reshape(-1)
39
+ dy = dy.reshape(-1)
40
+
41
+ valid = (x1 > 0) & (x1 < wd) & (y1 > 0) & (y1 < ht)
42
+ x1 = x1[valid]
43
+ y1 = y1[valid]
44
+ dx = dx[valid]
45
+ dy = dy[valid]
46
+
47
+ flow_x = interpolate.griddata(
48
+ (x1, y1), dx, (x0, y0), method='nearest', fill_value=0)
49
+
50
+ flow_y = interpolate.griddata(
51
+ (x1, y1), dy, (x0, y0), method='nearest', fill_value=0)
52
+
53
+ flow = np.stack([flow_x, flow_y], axis=0)
54
+ return torch.from_numpy(flow).float()
55
+
56
+
57
+ def bilinear_sampler(img, coords, mode='bilinear', mask=False):
58
+ """ Wrapper for grid_sample, uses pixel coordinates """
59
+ H, W = img.shape[-2:]
60
+ xgrid, ygrid = coords.split([1,1], dim=-1)
61
+ xgrid = 2*xgrid/(W-1) - 1
62
+ ygrid = 2*ygrid/(H-1) - 1
63
+
64
+ grid = torch.cat([xgrid, ygrid], dim=-1)
65
+ img = F.grid_sample(img, grid, align_corners=True)
66
+
67
+ if mask:
68
+ mask = (xgrid > -1) & (ygrid > -1) & (xgrid < 1) & (ygrid < 1)
69
+ return img, mask.float()
70
+
71
+ return img
72
+
73
+
74
+ def coords_grid(batch, ht, wd, device):
75
+ coords = torch.meshgrid(torch.arange(ht, device=device), torch.arange(wd, device=device))
76
+ coords = torch.stack(coords[::-1], dim=0).float()
77
+ return coords[None].repeat(batch, 1, 1, 1)
78
+
79
+
80
+ def upflow8(flow, mode='bilinear'):
81
+ new_size = (8 * flow.shape[2], 8 * flow.shape[3])
82
+ return 8 * F.interpolate(flow, size=new_size, mode=mode, align_corners=True)
causalvideovae/eval/RAFT/demo.py ADDED
@@ -0,0 +1,75 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ import sys
2
+ sys.path.append('core')
3
+
4
+ import argparse
5
+ import os
6
+ import cv2
7
+ import glob
8
+ import numpy as np
9
+ import torch
10
+ from PIL import Image
11
+
12
+ from raft import RAFT
13
+ from utils import flow_viz
14
+ from utils.utils import InputPadder
15
+
16
+
17
+
18
+ DEVICE = 'cuda'
19
+
20
+ def load_image(imfile):
21
+ img = np.array(Image.open(imfile)).astype(np.uint8)
22
+ img = torch.from_numpy(img).permute(2, 0, 1).float()
23
+ return img[None].to(DEVICE)
24
+
25
+
26
+ def viz(img, flo):
27
+ img = img[0].permute(1,2,0).cpu().numpy()
28
+ flo = flo[0].permute(1,2,0).cpu().numpy()
29
+
30
+ # map flow to rgb image
31
+ flo = flow_viz.flow_to_image(flo)
32
+ img_flo = np.concatenate([img, flo], axis=0)
33
+
34
+ # import matplotlib.pyplot as plt
35
+ # plt.imshow(img_flo / 255.0)
36
+ # plt.show()
37
+
38
+ cv2.imshow('image', img_flo[:, :, [2,1,0]]/255.0)
39
+ cv2.waitKey()
40
+
41
+
42
+ def demo(args):
43
+ model = torch.nn.DataParallel(RAFT(args))
44
+ model.load_state_dict(torch.load(args.model))
45
+
46
+ model = model.module
47
+ model.to(DEVICE)
48
+ model.eval()
49
+
50
+ with torch.no_grad():
51
+ images = glob.glob(os.path.join(args.path, '*.png')) + \
52
+ glob.glob(os.path.join(args.path, '*.jpg'))
53
+
54
+ images = sorted(images)
55
+ for imfile1, imfile2 in zip(images[:-1], images[1:]):
56
+ image1 = load_image(imfile1)
57
+ image2 = load_image(imfile2)
58
+
59
+ padder = InputPadder(image1.shape)
60
+ image1, image2 = padder.pad(image1, image2)
61
+
62
+ flow_low, flow_up = model(image1, image2, iters=20, test_mode=True)
63
+ viz(image1, flow_up)
64
+
65
+
66
+ if __name__ == '__main__':
67
+ parser = argparse.ArgumentParser()
68
+ parser.add_argument('--model', help="restore checkpoint")
69
+ parser.add_argument('--path', help="dataset for evaluation")
70
+ parser.add_argument('--small', action='store_true', help='use small model')
71
+ parser.add_argument('--mixed_precision', action='store_true', help='use mixed precision')
72
+ parser.add_argument('--alternate_corr', action='store_true', help='use efficent correlation implementation')
73
+ args = parser.parse_args()
74
+
75
+ demo(args)
causalvideovae/eval/RAFT/download_models.sh ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ #!/bin/bash
2
+ wget https://dl.dropboxusercontent.com/s/4j4z58wuv8o0mfz/models.zip
3
+ unzip models.zip
causalvideovae/eval/RAFT/train_mixed.sh ADDED
@@ -0,0 +1,6 @@
 
 
 
 
 
 
 
1
+ #!/bin/bash
2
+ mkdir -p checkpoints
3
+ python -u train.py --name raft-chairs --stage chairs --validation chairs --gpus 0 --num_steps 120000 --batch_size 8 --lr 0.00025 --image_size 368 496 --wdecay 0.0001 --mixed_precision
4
+ python -u train.py --name raft-things --stage things --validation sintel --restore_ckpt checkpoints/raft-chairs.pth --gpus 0 --num_steps 120000 --batch_size 5 --lr 0.0001 --image_size 400 720 --wdecay 0.0001 --mixed_precision
5
+ python -u train.py --name raft-sintel --stage sintel --validation sintel --restore_ckpt checkpoints/raft-things.pth --gpus 0 --num_steps 120000 --batch_size 5 --lr 0.0001 --image_size 368 768 --wdecay 0.00001 --gamma=0.85 --mixed_precision
6
+ python -u train.py --name raft-kitti --stage kitti --validation kitti --restore_ckpt checkpoints/raft-sintel.pth --gpus 0 --num_steps 50000 --batch_size 5 --lr 0.0001 --image_size 288 960 --wdecay 0.00001 --gamma=0.85 --mixed_precision
causalvideovae/eval/cal_mse.py ADDED
@@ -0,0 +1,93 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ import numpy as np
2
+ import torch
3
+ from tqdm import tqdm
4
+ import math
5
+
6
+ def img_psnr_cuda(img1, img2):
7
+ # [0,1]
8
+ # compute mse
9
+ # mse = np.mean((img1-img2)**2)
10
+ mse = torch.mean((img1 / 1.0 - img2 / 1.0) ** 2)
11
+ # compute psnr
12
+ if mse < 1e-10:
13
+ return 100
14
+ psnr = 20 * torch.log10(1 / torch.sqrt(mse))
15
+ return psnr
16
+
17
+
18
+ def img_mse(img1, img2):
19
+ # [0,1]
20
+ # compute mse
21
+ # mse = np.mean((img1-img2)**2)
22
+ mse = np.sqrt(np.mean((img1*255/ 1.0 - img2*255 / 1.0) ** 2))
23
+ return mse
24
+
25
+
26
+ def trans(x):
27
+ return x
28
+
29
+ def calculate_mse(videos1, videos2):
30
+ print("calculate_mse...")
31
+
32
+ # videos [batch_size, timestamps, channel, h, w]
33
+
34
+ assert videos1.shape == videos2.shape
35
+
36
+ videos1 = trans(videos1)
37
+ videos2 = trans(videos2)
38
+
39
+ mse_results = []
40
+
41
+ for video_num in tqdm(range(videos1.shape[0])):
42
+ # get a video
43
+ # video [timestamps, channel, h, w]
44
+ video1 = videos1[video_num]
45
+ video2 = videos2[video_num]
46
+
47
+ mse_results_of_a_video = []
48
+ for clip_timestamp in range(len(video1)):
49
+ # get a img
50
+ # img [timestamps[x], channel, h, w]
51
+ # img [channel, h, w] numpy
52
+
53
+ img1 = video1[clip_timestamp].numpy()
54
+ img2 = video2[clip_timestamp].numpy()
55
+
56
+ # calculate psnr of a video
57
+ mse_results_of_a_video.append(img_mse(img1, img2))
58
+
59
+ mse_results.append(mse_results_of_a_video)
60
+
61
+ mse_results = np.array(mse_results) # [batch_size, num_frames]
62
+ mse = {}
63
+ mse_std = {}
64
+
65
+ for clip_timestamp in range(len(video1)):
66
+ mse[clip_timestamp] = np.mean(mse_results[:,clip_timestamp])
67
+ mse_std[clip_timestamp] = np.std(mse_results[:,clip_timestamp])
68
+
69
+ result = {
70
+ "value": mse,
71
+ "value_std": mse_std,
72
+ "video_setting": video1.shape,
73
+ "video_setting_name": "time, channel, heigth, width",
74
+ }
75
+
76
+ return result
77
+
78
+ # test code / using example
79
+
80
+ def main():
81
+ NUMBER_OF_VIDEOS = 8
82
+ VIDEO_LENGTH = 50
83
+ CHANNEL = 3
84
+ SIZE = 64
85
+ videos1 = torch.zeros(NUMBER_OF_VIDEOS, VIDEO_LENGTH, CHANNEL, SIZE, SIZE, requires_grad=False)
86
+ videos2 = torch.zeros(NUMBER_OF_VIDEOS, VIDEO_LENGTH, CHANNEL, SIZE, SIZE, requires_grad=False)
87
+
88
+ import json
89
+ result = calculate_psnr(videos1, videos2)
90
+ print(json.dumps(result, indent=4))
91
+
92
+ if __name__ == "__main__":
93
+ main()
causalvideovae/eval/flolpips/flolpips.py ADDED
@@ -0,0 +1,308 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+
2
+ from __future__ import absolute_import
3
+ import os
4
+ import numpy as np
5
+ import torch
6
+ import torch.nn as nn
7
+ from torch.autograd import Variable
8
+ from .pretrained_networks import vgg16, alexnet, squeezenet
9
+ import torch.nn
10
+ import torch.nn.functional as F
11
+ import torchvision.transforms.functional as TF
12
+ import cv2
13
+
14
+ from .pwcnet import Network as PWCNet
15
+ from .utils import *
16
+
17
+ def spatial_average(in_tens, keepdim=True):
18
+ return in_tens.mean([2,3],keepdim=keepdim)
19
+
20
+ def mw_spatial_average(in_tens, flow, keepdim=True):
21
+ _,_,h,w = in_tens.shape
22
+ flow = F.interpolate(flow, (h,w), align_corners=False, mode='bilinear')
23
+ flow_mag = torch.sqrt(flow[:,0:1]**2 + flow[:,1:2]**2)
24
+ flow_mag = flow_mag / torch.sum(flow_mag, dim=[1,2,3], keepdim=True)
25
+ return torch.sum(in_tens*flow_mag, dim=[2,3],keepdim=keepdim)
26
+
27
+
28
+ def mtw_spatial_average(in_tens, flow, texture, keepdim=True):
29
+ _,_,h,w = in_tens.shape
30
+ flow = F.interpolate(flow, (h,w), align_corners=False, mode='bilinear')
31
+ texture = F.interpolate(texture, (h,w), align_corners=False, mode='bilinear')
32
+ flow_mag = torch.sqrt(flow[:,0:1]**2 + flow[:,1:2]**2)
33
+ flow_mag = (flow_mag - flow_mag.min()) / (flow_mag.max() - flow_mag.min()) + 1e-6
34
+ texture = (texture - texture.min()) / (texture.max() - texture.min()) + 1e-6
35
+ weight = flow_mag / texture
36
+ weight /= torch.sum(weight)
37
+ return torch.sum(in_tens*weight, dim=[2,3],keepdim=keepdim)
38
+
39
+
40
+
41
+ def m2w_spatial_average(in_tens, flow, keepdim=True):
42
+ _,_,h,w = in_tens.shape
43
+ flow = F.interpolate(flow, (h,w), align_corners=False, mode='bilinear')
44
+ flow_mag = flow[:,0:1]**2 + flow[:,1:2]**2 # B,1,H,W
45
+ flow_mag = flow_mag / torch.sum(flow_mag)
46
+ return torch.sum(in_tens*flow_mag, dim=[2,3],keepdim=keepdim)
47
+
48
+ def upsample(in_tens, out_HW=(64,64)): # assumes scale factor is same for H and W
49
+ in_H, in_W = in_tens.shape[2], in_tens.shape[3]
50
+ return nn.Upsample(size=out_HW, mode='bilinear', align_corners=False)(in_tens)
51
+
52
+ # Learned perceptual metric
53
+ class LPIPS(nn.Module):
54
+ def __init__(self, pretrained=True, net='alex', version='0.1', lpips=True, spatial=False,
55
+ pnet_rand=False, pnet_tune=False, use_dropout=True, model_path=None, eval_mode=True, verbose=False):
56
+ # lpips - [True] means with linear calibration on top of base network
57
+ # pretrained - [True] means load linear weights
58
+
59
+ super(LPIPS, self).__init__()
60
+ if(verbose):
61
+ print('Setting up [%s] perceptual loss: trunk [%s], v[%s], spatial [%s]'%
62
+ ('LPIPS' if lpips else 'baseline', net, version, 'on' if spatial else 'off'))
63
+
64
+ self.pnet_type = net
65
+ self.pnet_tune = pnet_tune
66
+ self.pnet_rand = pnet_rand
67
+ self.spatial = spatial
68
+ self.lpips = lpips # false means baseline of just averaging all layers
69
+ self.version = version
70
+ self.scaling_layer = ScalingLayer()
71
+
72
+ if(self.pnet_type in ['vgg','vgg16']):
73
+ net_type = vgg16
74
+ self.chns = [64,128,256,512,512]
75
+ elif(self.pnet_type=='alex'):
76
+ net_type = alexnet
77
+ self.chns = [64,192,384,256,256]
78
+ elif(self.pnet_type=='squeeze'):
79
+ net_type = squeezenet
80
+ self.chns = [64,128,256,384,384,512,512]
81
+ self.L = len(self.chns)
82
+
83
+ self.net = net_type(pretrained=not self.pnet_rand, requires_grad=self.pnet_tune)
84
+
85
+ if(lpips):
86
+ self.lin0 = NetLinLayer(self.chns[0], use_dropout=use_dropout)
87
+ self.lin1 = NetLinLayer(self.chns[1], use_dropout=use_dropout)
88
+ self.lin2 = NetLinLayer(self.chns[2], use_dropout=use_dropout)
89
+ self.lin3 = NetLinLayer(self.chns[3], use_dropout=use_dropout)
90
+ self.lin4 = NetLinLayer(self.chns[4], use_dropout=use_dropout)
91
+ self.lins = [self.lin0,self.lin1,self.lin2,self.lin3,self.lin4]
92
+ if(self.pnet_type=='squeeze'): # 7 layers for squeezenet
93
+ self.lin5 = NetLinLayer(self.chns[5], use_dropout=use_dropout)
94
+ self.lin6 = NetLinLayer(self.chns[6], use_dropout=use_dropout)
95
+ self.lins+=[self.lin5,self.lin6]
96
+ self.lins = nn.ModuleList(self.lins)
97
+
98
+ if(pretrained):
99
+ if(model_path is None):
100
+ import inspect
101
+ import os
102
+ model_path = os.path.abspath(os.path.join(inspect.getfile(self.__init__), '..', 'weights/v%s/%s.pth'%(version,net)))
103
+
104
+ if(verbose):
105
+ print('Loading model from: %s'%model_path)
106
+ self.load_state_dict(torch.load(model_path, map_location='cpu'), strict=False)
107
+
108
+ if(eval_mode):
109
+ self.eval()
110
+
111
+ def forward(self, in0, in1, retPerLayer=False, normalize=False):
112
+ if normalize: # turn on this flag if input is [0,1] so it can be adjusted to [-1, +1]
113
+ in0 = 2 * in0 - 1
114
+ in1 = 2 * in1 - 1
115
+
116
+ # v0.0 - original release had a bug, where input was not scaled
117
+ in0_input, in1_input = (self.scaling_layer(in0), self.scaling_layer(in1)) if self.version=='0.1' else (in0, in1)
118
+ outs0, outs1 = self.net.forward(in0_input), self.net.forward(in1_input)
119
+ feats0, feats1, diffs = {}, {}, {}
120
+
121
+ for kk in range(self.L):
122
+ feats0[kk], feats1[kk] = normalize_tensor(outs0[kk]), normalize_tensor(outs1[kk])
123
+ diffs[kk] = (feats0[kk]-feats1[kk])**2
124
+
125
+ if(self.lpips):
126
+ if(self.spatial):
127
+ res = [upsample(self.lins[kk](diffs[kk]), out_HW=in0.shape[2:]) for kk in range(self.L)]
128
+ else:
129
+ res = [spatial_average(self.lins[kk](diffs[kk]), keepdim=True) for kk in range(self.L)]
130
+ else:
131
+ if(self.spatial):
132
+ res = [upsample(diffs[kk].sum(dim=1,keepdim=True), out_HW=in0.shape[2:]) for kk in range(self.L)]
133
+ else:
134
+ res = [spatial_average(diffs[kk].sum(dim=1,keepdim=True), keepdim=True) for kk in range(self.L)]
135
+
136
+ # val = res[0]
137
+ # for l in range(1,self.L):
138
+ # val += res[l]
139
+ # print(val)
140
+
141
+ # a = spatial_average(self.lins[kk](diffs[kk]), keepdim=True)
142
+ # b = torch.max(self.lins[kk](feats0[kk]**2))
143
+ # for kk in range(self.L):
144
+ # a += spatial_average(self.lins[kk](diffs[kk]), keepdim=True)
145
+ # b = torch.max(b,torch.max(self.lins[kk](feats0[kk]**2)))
146
+ # a = a/self.L
147
+ # from IPython import embed
148
+ # embed()
149
+ # return 10*torch.log10(b/a)
150
+
151
+ # if(retPerLayer):
152
+ # return (val, res)
153
+ # else:
154
+ return torch.sum(torch.cat(res, 1), dim=(1,2,3), keepdims=False)
155
+
156
+
157
+ class ScalingLayer(nn.Module):
158
+ def __init__(self):
159
+ super(ScalingLayer, self).__init__()
160
+ self.register_buffer('shift', torch.Tensor([-.030,-.088,-.188])[None,:,None,None])
161
+ self.register_buffer('scale', torch.Tensor([.458,.448,.450])[None,:,None,None])
162
+
163
+ def forward(self, inp):
164
+ return (inp - self.shift) / self.scale
165
+
166
+
167
+ class NetLinLayer(nn.Module):
168
+ ''' A single linear layer which does a 1x1 conv '''
169
+ def __init__(self, chn_in, chn_out=1, use_dropout=False):
170
+ super(NetLinLayer, self).__init__()
171
+
172
+ layers = [nn.Dropout(),] if(use_dropout) else []
173
+ layers += [nn.Conv2d(chn_in, chn_out, 1, stride=1, padding=0, bias=False),]
174
+ self.model = nn.Sequential(*layers)
175
+
176
+ def forward(self, x):
177
+ return self.model(x)
178
+
179
+ class Dist2LogitLayer(nn.Module):
180
+ ''' takes 2 distances, puts through fc layers, spits out value between [0,1] (if use_sigmoid is True) '''
181
+ def __init__(self, chn_mid=32, use_sigmoid=True):
182
+ super(Dist2LogitLayer, self).__init__()
183
+
184
+ layers = [nn.Conv2d(5, chn_mid, 1, stride=1, padding=0, bias=True),]
185
+ layers += [nn.LeakyReLU(0.2,True),]
186
+ layers += [nn.Conv2d(chn_mid, chn_mid, 1, stride=1, padding=0, bias=True),]
187
+ layers += [nn.LeakyReLU(0.2,True),]
188
+ layers += [nn.Conv2d(chn_mid, 1, 1, stride=1, padding=0, bias=True),]
189
+ if(use_sigmoid):
190
+ layers += [nn.Sigmoid(),]
191
+ self.model = nn.Sequential(*layers)
192
+
193
+ def forward(self,d0,d1,eps=0.1):
194
+ return self.model.forward(torch.cat((d0,d1,d0-d1,d0/(d1+eps),d1/(d0+eps)),dim=1))
195
+
196
+ class BCERankingLoss(nn.Module):
197
+ def __init__(self, chn_mid=32):
198
+ super(BCERankingLoss, self).__init__()
199
+ self.net = Dist2LogitLayer(chn_mid=chn_mid)
200
+ # self.parameters = list(self.net.parameters())
201
+ self.loss = torch.nn.BCELoss()
202
+
203
+ def forward(self, d0, d1, judge):
204
+ per = (judge+1.)/2.
205
+ self.logit = self.net.forward(d0,d1)
206
+ return self.loss(self.logit, per)
207
+
208
+ # L2, DSSIM metrics
209
+ class FakeNet(nn.Module):
210
+ def __init__(self, use_gpu=True, colorspace='Lab'):
211
+ super(FakeNet, self).__init__()
212
+ self.use_gpu = use_gpu
213
+ self.colorspace = colorspace
214
+
215
+ class L2(FakeNet):
216
+ def forward(self, in0, in1, retPerLayer=None):
217
+ assert(in0.size()[0]==1) # currently only supports batchSize 1
218
+
219
+ if(self.colorspace=='RGB'):
220
+ (N,C,X,Y) = in0.size()
221
+ value = torch.mean(torch.mean(torch.mean((in0-in1)**2,dim=1).view(N,1,X,Y),dim=2).view(N,1,1,Y),dim=3).view(N)
222
+ return value
223
+ elif(self.colorspace=='Lab'):
224
+ value = l2(tensor2np(tensor2tensorlab(in0.data,to_norm=False)),
225
+ tensor2np(tensor2tensorlab(in1.data,to_norm=False)), range=100.).astype('float')
226
+ ret_var = Variable( torch.Tensor((value,) ) )
227
+ if(self.use_gpu):
228
+ ret_var = ret_var.cuda()
229
+ return ret_var
230
+
231
+ class DSSIM(FakeNet):
232
+
233
+ def forward(self, in0, in1, retPerLayer=None):
234
+ assert(in0.size()[0]==1) # currently only supports batchSize 1
235
+
236
+ if(self.colorspace=='RGB'):
237
+ value = dssim(1.*tensor2im(in0.data), 1.*tensor2im(in1.data), range=255.).astype('float')
238
+ elif(self.colorspace=='Lab'):
239
+ value = dssim(tensor2np(tensor2tensorlab(in0.data,to_norm=False)),
240
+ tensor2np(tensor2tensorlab(in1.data,to_norm=False)), range=100.).astype('float')
241
+ ret_var = Variable( torch.Tensor((value,) ) )
242
+ if(self.use_gpu):
243
+ ret_var = ret_var.cuda()
244
+ return ret_var
245
+
246
+ def print_network(net):
247
+ num_params = 0
248
+ for param in net.parameters():
249
+ num_params += param.numel()
250
+ print('Network',net)
251
+ print('Total number of parameters: %d' % num_params)
252
+
253
+
254
+ class FloLPIPS(LPIPS):
255
+ def __init__(self, pretrained=True, net='alex', version='0.1', lpips=True, spatial=False, pnet_rand=False, pnet_tune=False, use_dropout=True, model_path=None, eval_mode=True, verbose=False):
256
+ super(FloLPIPS, self).__init__(pretrained, net, version, lpips, spatial, pnet_rand, pnet_tune, use_dropout, model_path, eval_mode, verbose)
257
+
258
+ def forward(self, in0, in1, flow, retPerLayer=False, normalize=False):
259
+ if normalize: # turn on this flag if input is [0,1] so it can be adjusted to [-1, +1]
260
+ in0 = 2 * in0 - 1
261
+ in1 = 2 * in1 - 1
262
+
263
+ in0_input, in1_input = (self.scaling_layer(in0), self.scaling_layer(in1)) if self.version=='0.1' else (in0, in1)
264
+ outs0, outs1 = self.net.forward(in0_input), self.net.forward(in1_input)
265
+ feats0, feats1, diffs = {}, {}, {}
266
+
267
+ for kk in range(self.L):
268
+ feats0[kk], feats1[kk] = normalize_tensor(outs0[kk]), normalize_tensor(outs1[kk])
269
+ diffs[kk] = (feats0[kk]-feats1[kk])**2
270
+
271
+ res = [mw_spatial_average(self.lins[kk](diffs[kk]), flow, keepdim=True) for kk in range(self.L)]
272
+
273
+ return torch.sum(torch.cat(res, 1), dim=(1,2,3), keepdims=False)
274
+
275
+
276
+
277
+
278
+
279
+ class Flolpips(nn.Module):
280
+ def __init__(self):
281
+ super(Flolpips, self).__init__()
282
+ self.loss_fn = FloLPIPS(net='alex',version='0.1')
283
+ self.flownet = PWCNet()
284
+
285
+ @torch.no_grad()
286
+ def forward(self, I0, I1, frame_dis, frame_ref):
287
+ """
288
+ args:
289
+ I0: first frame of the triplet, shape: [B, C, H, W]
290
+ I1: third frame of the triplet, shape: [B, C, H, W]
291
+ frame_dis: prediction of the intermediate frame, shape: [B, C, H, W]
292
+ frame_ref: ground-truth of the intermediate frame, shape: [B, C, H, W]
293
+ """
294
+ assert I0.size() == I1.size() == frame_dis.size() == frame_ref.size(), \
295
+ "the 4 input tensors should have same size"
296
+
297
+ flow_ref = self.flownet(frame_ref, I0)
298
+ flow_dis = self.flownet(frame_dis, I0)
299
+ flow_diff = flow_ref - flow_dis
300
+ flolpips_wrt_I0 = self.loss_fn.forward(frame_ref, frame_dis, flow_diff, normalize=True)
301
+
302
+ flow_ref = self.flownet(frame_ref, I1)
303
+ flow_dis = self.flownet(frame_dis, I1)
304
+ flow_diff = flow_ref - flow_dis
305
+ flolpips_wrt_I1 = self.loss_fn.forward(frame_ref, frame_dis, flow_diff, normalize=True)
306
+
307
+ flolpips = (flolpips_wrt_I0 + flolpips_wrt_I1) / 2
308
+ return flolpips
causalvideovae/eval/script/cal_clip_score.sh ADDED
@@ -0,0 +1,23 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ # clip_score cross modality
2
+ python eval_clip_score.py \
3
+ --real_path path/to/image \
4
+ --generated_path path/to/text \
5
+ --batch-size 50 \
6
+ --device "cuda"
7
+
8
+ # clip_score within the same modality
9
+ python eval_clip_score.py \
10
+ --real_path path/to/textA \
11
+ --generated_path path/to/textB \
12
+ --real_flag txt \
13
+ --generated_flag txt \
14
+ --batch-size 50 \
15
+ --device "cuda"
16
+
17
+ python eval_clip_score.py \
18
+ --real_path path/to/imageA \
19
+ --generated_path path/to/imageB \
20
+ --real_flag img \
21
+ --generated_flag img \
22
+ --batch-size 50 \
23
+ --device "cuda"
causalvideovae/eval/script/cal_fvd.sh ADDED
@@ -0,0 +1,9 @@
 
 
 
 
 
 
 
 
 
 
1
+ python eval_common_metric.py \
2
+ --real_video_dir path/to/imageA\
3
+ --generated_video_dir path/to/imageB \
4
+ --batch_size 10 \
5
+ --crop_size 64 \
6
+ --num_frames 20 \
7
+ --device 'cuda' \
8
+ --metric 'fvd' \
9
+ --fvd_method 'styleganv'
causalvideovae/eval/script/cal_psnr.sh ADDED
@@ -0,0 +1,9 @@
 
 
 
 
 
 
 
 
 
 
1
+
2
+ python eval_common_metric.py \
3
+ --real_video_dir /data/xiaogeng_liu/data/video1 \
4
+ --generated_video_dir /data/xiaogeng_liu/data/video2 \
5
+ --batch_size 10 \
6
+ --num_frames 20 \
7
+ --crop_size 64 \
8
+ --device 'cuda' \
9
+ --metric 'psnr'
causalvideovae/eval/script/cal_ssim.sh ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ python eval_common_metric.py \
2
+ --real_video_dir /data/xiaogeng_liu/data/video1 \
3
+ --generated_video_dir /data/xiaogeng_liu/data/video2 \
4
+ --batch_size 10 \
5
+ --num_frames 20 \
6
+ --crop_size 64 \
7
+ --device 'cuda' \
8
+ --metric 'ssim'
causalvideovae/model/causal_vae/__init__.py ADDED
@@ -0,0 +1,29 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ from .modeling_causalvae import CausalVAEModel
2
+
3
+ from einops import rearrange
4
+ from torch import nn
5
+
6
+ class CausalVAEModelWrapper(nn.Module):
7
+ def __init__(self, model_path, subfolder=None, cache_dir=None, use_ema=False, **kwargs):
8
+ super(CausalVAEModelWrapper, self).__init__()
9
+ # if os.path.exists(ckpt):
10
+ # self.vae = CausalVAEModel.load_from_checkpoint(ckpt)
11
+ self.vae = CausalVAEModel.from_pretrained(model_path, subfolder=subfolder, cache_dir=cache_dir, **kwargs)
12
+ if use_ema:
13
+ self.vae.init_from_ema(model_path)
14
+ self.vae = self.vae.ema
15
+ def encode(self, x): # b c t h w
16
+ # x = self.vae.encode(x).sample()
17
+ x = self.vae.encode(x).sample().mul_(0.18215)
18
+ return x
19
+ def decode(self, x):
20
+ # x = self.vae.decode(x)
21
+ x = self.vae.decode(x / 0.18215)
22
+ x = rearrange(x, 'b c t h w -> b t c h w').contiguous()
23
+ return x
24
+
25
+ def dtype(self):
26
+ return self.vae.dtype
27
+ #
28
+ # def device(self):
29
+ # return self.vae.device
causalvideovae/model/causal_vae/modeling_causalvae.py ADDED
@@ -0,0 +1,610 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ from ..modeling_videobase import VideoBaseAE
2
+ from ..modules import Normalize
3
+ from ..modules.ops import nonlinearity
4
+ from typing import List, Tuple
5
+ import torch.nn as nn
6
+
7
+ from ..utils.module_utils import resolve_str_to_obj, Module
8
+ from ..utils.distrib_utils import DiagonalGaussianDistribution
9
+ from ..utils.scheduler_utils import cosine_scheduler
10
+ from ...utils.utils import custom_to_video
11
+
12
+ import torch
13
+ from diffusers.configuration_utils import register_to_config
14
+ from copy import deepcopy
15
+ import os
16
+ import glob
17
+
18
+ import numpy as np
19
+ from ...eval.cal_psnr import calculate_psnr
20
+ from decord import VideoReader, cpu
21
+ from pytorchvideo.transforms import ShortSideScale
22
+ from torchvision.io import read_video
23
+ from torchvision.transforms import Lambda, Compose
24
+ from torchvision.transforms._transforms_video import CenterCropVideo
25
+
26
+ class Encoder(nn.Module):
27
+ def __init__(
28
+ self,
29
+ z_channels: int,
30
+ hidden_size: int,
31
+ hidden_size_mult: Tuple[int] = (1, 2, 4, 4),
32
+ attn_resolutions: Tuple[int] = (16,),
33
+ conv_in: Module = "Conv2d",
34
+ conv_out: Module = "CasualConv3d",
35
+ attention: Module = "AttnBlock",
36
+ resnet_blocks: Tuple[Module] = (
37
+ "ResnetBlock2D",
38
+ "ResnetBlock2D",
39
+ "ResnetBlock2D",
40
+ "ResnetBlock3D",
41
+ ),
42
+ spatial_downsample: Tuple[Module] = (
43
+ "Downsample",
44
+ "Downsample",
45
+ "Downsample",
46
+ "",
47
+ ),
48
+ temporal_downsample: Tuple[Module] = ("", "", "TimeDownsampleRes2x", ""),
49
+ mid_resnet: Module = "ResnetBlock3D",
50
+ dropout: float = 0.0,
51
+ resolution: int = 256,
52
+ num_res_blocks: int = 2,
53
+ double_z: bool = True,
54
+ ) -> None:
55
+ super().__init__()
56
+ assert len(resnet_blocks) == len(hidden_size_mult), print(
57
+ hidden_size_mult, resnet_blocks
58
+ )
59
+ # ---- Config ----
60
+ self.num_resolutions = len(hidden_size_mult)
61
+ self.resolution = resolution
62
+ self.num_res_blocks = num_res_blocks
63
+
64
+ # ---- In ----
65
+ self.conv_in = resolve_str_to_obj(conv_in)(
66
+ 3, hidden_size, kernel_size=3, stride=1, padding=1
67
+ )
68
+
69
+ # ---- Downsample ----
70
+ curr_res = resolution
71
+ in_ch_mult = (1,) + tuple(hidden_size_mult)
72
+ self.in_ch_mult = in_ch_mult
73
+ self.down = nn.ModuleList()
74
+ for i_level in range(self.num_resolutions):
75
+ block = nn.ModuleList()
76
+ attn = nn.ModuleList()
77
+ block_in = hidden_size * in_ch_mult[i_level]
78
+ block_out = hidden_size * hidden_size_mult[i_level]
79
+ for i_block in range(self.num_res_blocks):
80
+ block.append(
81
+ resolve_str_to_obj(resnet_blocks[i_level])(
82
+ in_channels=block_in,
83
+ out_channels=block_out,
84
+ dropout=dropout,
85
+ )
86
+ )
87
+ block_in = block_out
88
+ if curr_res in attn_resolutions:
89
+ attn.append(resolve_str_to_obj(attention)(block_in))
90
+ down = nn.Module()
91
+ down.block = block
92
+ down.attn = attn
93
+ if spatial_downsample[i_level]:
94
+ down.downsample = resolve_str_to_obj(spatial_downsample[i_level])(
95
+ block_in, block_in
96
+ )
97
+ curr_res = curr_res // 2
98
+ if temporal_downsample[i_level]:
99
+ down.time_downsample = resolve_str_to_obj(temporal_downsample[i_level])(
100
+ block_in, block_in
101
+ )
102
+ self.down.append(down)
103
+
104
+ # ---- Mid ----
105
+ self.mid = nn.Module()
106
+ self.mid.block_1 = resolve_str_to_obj(mid_resnet)(
107
+ in_channels=block_in,
108
+ out_channels=block_in,
109
+ dropout=dropout,
110
+ )
111
+ self.mid.attn_1 = resolve_str_to_obj(attention)(block_in)
112
+ self.mid.block_2 = resolve_str_to_obj(mid_resnet)(
113
+ in_channels=block_in,
114
+ out_channels=block_in,
115
+ dropout=dropout,
116
+ )
117
+ # ---- Out ----
118
+ self.norm_out = Normalize(block_in)
119
+ self.conv_out = resolve_str_to_obj(conv_out)(
120
+ block_in,
121
+ 2 * z_channels if double_z else z_channels,
122
+ kernel_size=3,
123
+ stride=1,
124
+ padding=1,
125
+ )
126
+
127
+ def forward(self, x):
128
+ hs = [self.conv_in(x)]
129
+ for i_level in range(self.num_resolutions):
130
+ for i_block in range(self.num_res_blocks):
131
+ h = self.down[i_level].block[i_block](hs[-1])
132
+ if len(self.down[i_level].attn) > 0:
133
+ h = self.down[i_level].attn[i_block](h)
134
+ hs.append(h)
135
+ if hasattr(self.down[i_level], "downsample"):
136
+ hs.append(self.down[i_level].downsample(hs[-1]))
137
+ if hasattr(self.down[i_level], "time_downsample"):
138
+ hs_down = self.down[i_level].time_downsample(hs[-1])
139
+ hs.append(hs_down)
140
+
141
+ h = self.mid.block_1(h)
142
+ h = self.mid.attn_1(h)
143
+ h = self.mid.block_2(h)
144
+
145
+ h = h.permute(0, 2, 3, 4, 1)
146
+ h = self.norm_out(h)
147
+ h = h.permute(0, 4, 1, 2, 3)
148
+ h = nonlinearity(h)
149
+ h = self.conv_out(h)
150
+ return h
151
+
152
+
153
+ class Decoder(nn.Module):
154
+ def __init__(
155
+ self,
156
+ z_channels: int,
157
+ hidden_size: int,
158
+ hidden_size_mult: Tuple[int] = (1, 2, 4, 4),
159
+ attn_resolutions: Tuple[int] = (16,),
160
+ conv_in: Module = "Conv2d",
161
+ conv_out: Module = "CasualConv3d",
162
+ attention: Module = "AttnBlock",
163
+ resnet_blocks: Tuple[Module] = (
164
+ "ResnetBlock3D",
165
+ "ResnetBlock3D",
166
+ "ResnetBlock3D",
167
+ "ResnetBlock3D",
168
+ ),
169
+ spatial_upsample: Tuple[Module] = (
170
+ "",
171
+ "SpatialUpsample2x",
172
+ "SpatialUpsample2x",
173
+ "SpatialUpsample2x",
174
+ ),
175
+ temporal_upsample: Tuple[Module] = ("", "", "", "TimeUpsampleRes2x"),
176
+ mid_resnet: Module = "ResnetBlock3D",
177
+ dropout: float = 0.0,
178
+ resolution: int = 256,
179
+ num_res_blocks: int = 2,
180
+ ):
181
+ super().__init__()
182
+ # ---- Config ----
183
+ self.num_resolutions = len(hidden_size_mult)
184
+ self.resolution = resolution
185
+ self.num_res_blocks = num_res_blocks
186
+
187
+ # ---- In ----
188
+ block_in = hidden_size * hidden_size_mult[self.num_resolutions - 1]
189
+ curr_res = resolution // 2 ** (self.num_resolutions - 1)
190
+ self.conv_in = resolve_str_to_obj(conv_in)(
191
+ z_channels, block_in, kernel_size=3, padding=1
192
+ )
193
+
194
+ # ---- Mid ----
195
+ self.mid = nn.Module()
196
+ self.mid.block_1 = resolve_str_to_obj(mid_resnet)(
197
+ in_channels=block_in,
198
+ out_channels=block_in,
199
+ dropout=dropout,
200
+ )
201
+ self.mid.attn_1 = resolve_str_to_obj(attention)(block_in)
202
+ self.mid.block_2 = resolve_str_to_obj(mid_resnet)(
203
+ in_channels=block_in,
204
+ out_channels=block_in,
205
+ dropout=dropout,
206
+ )
207
+
208
+ # ---- Upsample ----
209
+ self.up = nn.ModuleList()
210
+ for i_level in reversed(range(self.num_resolutions)):
211
+ block = nn.ModuleList()
212
+ attn = nn.ModuleList()
213
+ block_out = hidden_size * hidden_size_mult[i_level]
214
+ for i_block in range(self.num_res_blocks + 1):
215
+ block.append(
216
+ resolve_str_to_obj(resnet_blocks[i_level])(
217
+ in_channels=block_in,
218
+ out_channels=block_out,
219
+ dropout=dropout,
220
+ )
221
+ )
222
+ block_in = block_out
223
+ if curr_res in attn_resolutions:
224
+ attn.append(resolve_str_to_obj(attention)(block_in))
225
+ up = nn.Module()
226
+ up.block = block
227
+ up.attn = attn
228
+ if spatial_upsample[i_level]:
229
+ up.upsample = resolve_str_to_obj(spatial_upsample[i_level])(
230
+ block_in, block_in
231
+ )
232
+ curr_res = curr_res * 2
233
+ if temporal_upsample[i_level]:
234
+ up.time_upsample = resolve_str_to_obj(temporal_upsample[i_level])(
235
+ block_in, block_in
236
+ )
237
+ self.up.insert(0, up)
238
+
239
+ # ---- Out ----
240
+ self.norm_out = Normalize(block_in)
241
+ self.conv_out = resolve_str_to_obj(conv_out)(
242
+ block_in, 3, kernel_size=3, padding=1
243
+ )
244
+
245
+ def forward(self, z):
246
+ h = self.conv_in(z)
247
+ h = self.mid.block_1(h)
248
+ h = self.mid.attn_1(h)
249
+ h = self.mid.block_2(h)
250
+
251
+ for i_level in reversed(range(self.num_resolutions)):
252
+ for i_block in range(self.num_res_blocks + 1):
253
+ h = self.up[i_level].block[i_block](h)
254
+ if len(self.up[i_level].attn) > 0:
255
+ h = self.up[i_level].attn[i_block](h)
256
+ if hasattr(self.up[i_level], "upsample"):
257
+ h = self.up[i_level].upsample(h)
258
+ if hasattr(self.up[i_level], "time_upsample"):
259
+ h = self.up[i_level].time_upsample(h)
260
+
261
+ h = h.permute(0, 2, 3, 4, 1)
262
+ h = self.norm_out(h)
263
+ h = h.permute(0, 4, 1, 2, 3)
264
+ h = nonlinearity(h)
265
+ h = self.conv_out(h)
266
+ return h
267
+
268
+
269
+ class CausalVAEModel(VideoBaseAE):
270
+
271
+ @register_to_config
272
+ def __init__(
273
+ self,
274
+ hidden_size: int = 128,
275
+ z_channels: int = 4,
276
+ hidden_size_mult: Tuple[int] = (1, 2, 4, 4),
277
+ attn_resolutions: Tuple[int] = [],
278
+ dropout: float = 0.0,
279
+ resolution: int = 256,
280
+ double_z: bool = True,
281
+ embed_dim: int = 4,
282
+ num_res_blocks: int = 2,
283
+ q_conv: str = "CausalConv3d",
284
+ encoder_conv_in: Module = "CausalConv3d",
285
+ encoder_conv_out: Module = "CausalConv3d",
286
+ encoder_attention: Module = "AttnBlock3D",
287
+ encoder_resnet_blocks: Tuple[Module] = (
288
+ "ResnetBlock3D",
289
+ "ResnetBlock3D",
290
+ "ResnetBlock3D",
291
+ "ResnetBlock3D",
292
+ ),
293
+ encoder_spatial_downsample: Tuple[Module] = (
294
+ "SpatialDownsample2x",
295
+ "SpatialDownsample2x",
296
+ "SpatialDownsample2x",
297
+ "",
298
+ ),
299
+ encoder_temporal_downsample: Tuple[Module] = (
300
+ "",
301
+ "TimeDownsample2x",
302
+ "TimeDownsample2x",
303
+ "",
304
+ ),
305
+ encoder_mid_resnet: Module = "ResnetBlock3D",
306
+ decoder_conv_in: Module = "CausalConv3d",
307
+ decoder_conv_out: Module = "CausalConv3d",
308
+ decoder_attention: Module = "AttnBlock3D",
309
+ decoder_resnet_blocks: Tuple[Module] = (
310
+ "ResnetBlock3D",
311
+ "ResnetBlock3D",
312
+ "ResnetBlock3D",
313
+ "ResnetBlock3D",
314
+ ),
315
+ decoder_spatial_upsample: Tuple[Module] = (
316
+ "",
317
+ "SpatialUpsample2x",
318
+ "SpatialUpsample2x",
319
+ "SpatialUpsample2x",
320
+ ),
321
+ decoder_temporal_upsample: Tuple[Module] = ("", "", "TimeUpsample2x", "TimeUpsample2x"),
322
+ decoder_mid_resnet: Module = "ResnetBlock3D",
323
+ use_quant_layer: bool = True
324
+ ) -> None:
325
+ super().__init__()
326
+
327
+ self.tile_sample_min_size = 512
328
+ self.tile_sample_min_size_t = 25
329
+ self.tile_latent_min_size = int(self.tile_sample_min_size / (2 ** (len(hidden_size_mult) - 1)))
330
+ t_down_ratio = [i for i in encoder_temporal_downsample if len(i) > 0]
331
+ self.tile_latent_min_size_t = int((self.tile_sample_min_size_t-1) / (4)) + 1
332
+ self.tile_overlap_factor = 0.125
333
+ self.use_tiling = False
334
+ self.use_quant_layer = use_quant_layer
335
+
336
+ self.encoder = Encoder(
337
+ z_channels=z_channels,
338
+ hidden_size=hidden_size,
339
+ hidden_size_mult=hidden_size_mult,
340
+ attn_resolutions=attn_resolutions,
341
+ conv_in=encoder_conv_in,
342
+ conv_out=encoder_conv_out,
343
+ attention=encoder_attention,
344
+ resnet_blocks=encoder_resnet_blocks,
345
+ spatial_downsample=encoder_spatial_downsample,
346
+ temporal_downsample=encoder_temporal_downsample,
347
+ mid_resnet=encoder_mid_resnet,
348
+ dropout=dropout,
349
+ resolution=resolution,
350
+ num_res_blocks=num_res_blocks,
351
+ double_z=double_z,
352
+ )
353
+
354
+ self.decoder = Decoder(
355
+ z_channels=z_channels,
356
+ hidden_size=hidden_size,
357
+ hidden_size_mult=hidden_size_mult,
358
+ attn_resolutions=attn_resolutions,
359
+ conv_in=decoder_conv_in,
360
+ conv_out=decoder_conv_out,
361
+ attention=decoder_attention,
362
+ resnet_blocks=decoder_resnet_blocks,
363
+ spatial_upsample=decoder_spatial_upsample,
364
+ temporal_upsample=decoder_temporal_upsample,
365
+ mid_resnet=decoder_mid_resnet,
366
+ dropout=dropout,
367
+ resolution=resolution,
368
+ num_res_blocks=num_res_blocks,
369
+ )
370
+ if self.use_quant_layer:
371
+ quant_conv_cls = resolve_str_to_obj(q_conv)
372
+ self.quant_conv = quant_conv_cls(2 * z_channels, 2 * embed_dim, 1)
373
+ self.post_quant_conv = quant_conv_cls(embed_dim, z_channels, 1)
374
+
375
+ def get_encoder(self):
376
+ if self.use_quant_layer:
377
+ return [self.quant_conv, self.encoder]
378
+ return [self.encoder]
379
+
380
+ def get_decoder(self):
381
+ if self.use_quant_layer:
382
+ return [self.post_quant_conv, self.decoder]
383
+ return [self.decoder]
384
+
385
+ def encode(self, x):
386
+ if self.use_tiling and (
387
+ x.shape[-1] > self.tile_sample_min_size
388
+ or x.shape[-2] > self.tile_sample_min_size
389
+ or x.shape[-3] > self.tile_sample_min_size_t
390
+ ):
391
+ return self.tiled_encode(x)
392
+ h = self.encoder(x)
393
+ if self.use_quant_layer:
394
+ h = self.quant_conv(h)
395
+ posterior = DiagonalGaussianDistribution(h)
396
+ return posterior
397
+
398
+ def decode(self, z):
399
+ if self.use_tiling and (
400
+ z.shape[-1] > self.tile_latent_min_size
401
+ or z.shape[-2] > self.tile_latent_min_size
402
+ or z.shape[-3] > self.tile_latent_min_size_t
403
+ ):
404
+ return self.tiled_decode(z)
405
+ if self.use_quant_layer:
406
+ z = self.post_quant_conv(z)
407
+ dec = self.decoder(z)
408
+ return dec
409
+
410
+ def forward(self, input, sample_posterior=True):
411
+ posterior = self.encode(input)
412
+ if sample_posterior:
413
+ z = posterior.sample()
414
+ else:
415
+ z = posterior.mode()
416
+ dec = self.decode(z)
417
+ return dec, posterior
418
+
419
+ def on_train_start(self):
420
+ self.ema = deepcopy(self) if self.save_ema==True else None
421
+
422
+ def get_last_layer(self):
423
+ if hasattr(self.decoder.conv_out, "conv"):
424
+ return self.decoder.conv_out.conv.weight
425
+ else:
426
+ return self.decoder.conv_out.weight
427
+
428
+ def blend_v(
429
+ self, a: torch.Tensor, b: torch.Tensor, blend_extent: int
430
+ ) -> torch.Tensor:
431
+ blend_extent = min(a.shape[3], b.shape[3], blend_extent)
432
+ for y in range(blend_extent):
433
+ b[:, :, :, y, :] = a[:, :, :, -blend_extent + y, :] * (
434
+ 1 - y / blend_extent
435
+ ) + b[:, :, :, y, :] * (y / blend_extent)
436
+ return b
437
+
438
+ def blend_h(
439
+ self, a: torch.Tensor, b: torch.Tensor, blend_extent: int
440
+ ) -> torch.Tensor:
441
+ blend_extent = min(a.shape[4], b.shape[4], blend_extent)
442
+ for x in range(blend_extent):
443
+ b[:, :, :, :, x] = a[:, :, :, :, -blend_extent + x] * (
444
+ 1 - x / blend_extent
445
+ ) + b[:, :, :, :, x] * (x / blend_extent)
446
+ return b
447
+
448
+ def tiled_encode(self, x):
449
+ t = x.shape[2]
450
+ t_chunk_idx = [i for i in range(0, t, self.tile_sample_min_size_t-1)]
451
+ if len(t_chunk_idx) == 1 and t_chunk_idx[0] == 0:
452
+ t_chunk_start_end = [[0, t]]
453
+ else:
454
+ t_chunk_start_end = [[t_chunk_idx[i], t_chunk_idx[i+1]+1] for i in range(len(t_chunk_idx)-1)]
455
+ if t_chunk_start_end[-1][-1] > t:
456
+ t_chunk_start_end[-1][-1] = t
457
+ elif t_chunk_start_end[-1][-1] < t:
458
+ last_start_end = [t_chunk_idx[-1], t]
459
+ t_chunk_start_end.append(last_start_end)
460
+ moments = []
461
+ for idx, (start, end) in enumerate(t_chunk_start_end):
462
+ chunk_x = x[:, :, start: end]
463
+ if idx != 0:
464
+ moment = self.tiled_encode2d(chunk_x, return_moments=True)[:, :, 1:]
465
+ else:
466
+ moment = self.tiled_encode2d(chunk_x, return_moments=True)
467
+ moments.append(moment)
468
+ moments = torch.cat(moments, dim=2)
469
+ posterior = DiagonalGaussianDistribution(moments)
470
+ return posterior
471
+
472
+ def tiled_decode(self, x):
473
+ t = x.shape[2]
474
+ t_chunk_idx = [i for i in range(0, t, self.tile_latent_min_size_t-1)]
475
+ if len(t_chunk_idx) == 1 and t_chunk_idx[0] == 0:
476
+ t_chunk_start_end = [[0, t]]
477
+ else:
478
+ t_chunk_start_end = [[t_chunk_idx[i], t_chunk_idx[i+1]+1] for i in range(len(t_chunk_idx)-1)]
479
+ if t_chunk_start_end[-1][-1] > t:
480
+ t_chunk_start_end[-1][-1] = t
481
+ elif t_chunk_start_end[-1][-1] < t:
482
+ last_start_end = [t_chunk_idx[-1], t]
483
+ t_chunk_start_end.append(last_start_end)
484
+ dec_ = []
485
+ for idx, (start, end) in enumerate(t_chunk_start_end):
486
+ chunk_x = x[:, :, start: end]
487
+ if idx != 0:
488
+ dec = self.tiled_decode2d(chunk_x)[:, :, 1:]
489
+ else:
490
+ dec = self.tiled_decode2d(chunk_x)
491
+ dec_.append(dec)
492
+ dec_ = torch.cat(dec_, dim=2)
493
+ return dec_
494
+
495
+ def tiled_encode2d(self, x, return_moments=False):
496
+ overlap_size = int(self.tile_sample_min_size * (1 - self.tile_overlap_factor))
497
+ blend_extent = int(self.tile_latent_min_size * self.tile_overlap_factor)
498
+ row_limit = self.tile_latent_min_size - blend_extent
499
+
500
+ # Split the image into 512x512 tiles and encode them separately.
501
+ rows = []
502
+ for i in range(0, x.shape[3], overlap_size):
503
+ row = []
504
+ for j in range(0, x.shape[4], overlap_size):
505
+ tile = x[
506
+ :,
507
+ :,
508
+ :,
509
+ i : i + self.tile_sample_min_size,
510
+ j : j + self.tile_sample_min_size,
511
+ ]
512
+ tile = self.encoder(tile)
513
+ if self.use_quant_layer:
514
+ tile = self.quant_conv(tile)
515
+ row.append(tile)
516
+ rows.append(row)
517
+ result_rows = []
518
+ for i, row in enumerate(rows):
519
+ result_row = []
520
+ for j, tile in enumerate(row):
521
+ # blend the above tile and the left tile
522
+ # to the current tile and add the current tile to the result row
523
+ if i > 0:
524
+ tile = self.blend_v(rows[i - 1][j], tile, blend_extent)
525
+ if j > 0:
526
+ tile = self.blend_h(row[j - 1], tile, blend_extent)
527
+ result_row.append(tile[:, :, :, :row_limit, :row_limit])
528
+ result_rows.append(torch.cat(result_row, dim=4))
529
+
530
+ moments = torch.cat(result_rows, dim=3)
531
+ posterior = DiagonalGaussianDistribution(moments)
532
+ if return_moments:
533
+ return moments
534
+ return posterior
535
+
536
+ def tiled_decode2d(self, z):
537
+
538
+ overlap_size = int(self.tile_latent_min_size * (1 - self.tile_overlap_factor))
539
+ blend_extent = int(self.tile_sample_min_size * self.tile_overlap_factor)
540
+ row_limit = self.tile_sample_min_size - blend_extent
541
+
542
+ # Split z into overlapping 64x64 tiles and decode them separately.
543
+ # The tiles have an overlap to avoid seams between tiles.
544
+ rows = []
545
+ for i in range(0, z.shape[3], overlap_size):
546
+ row = []
547
+ for j in range(0, z.shape[4], overlap_size):
548
+ tile = z[
549
+ :,
550
+ :,
551
+ :,
552
+ i : i + self.tile_latent_min_size,
553
+ j : j + self.tile_latent_min_size,
554
+ ]
555
+ if self.use_quant_layer:
556
+ tile = self.post_quant_conv(tile)
557
+ decoded = self.decoder(tile)
558
+ row.append(decoded)
559
+ rows.append(row)
560
+ result_rows = []
561
+ for i, row in enumerate(rows):
562
+ result_row = []
563
+ for j, tile in enumerate(row):
564
+ # blend the above tile and the left tile
565
+ # to the current tile and add the current tile to the result row
566
+ if i > 0:
567
+ tile = self.blend_v(rows[i - 1][j], tile, blend_extent)
568
+ if j > 0:
569
+ tile = self.blend_h(row[j - 1], tile, blend_extent)
570
+ result_row.append(tile[:, :, :, :row_limit, :row_limit])
571
+ result_rows.append(torch.cat(result_row, dim=4))
572
+
573
+ dec = torch.cat(result_rows, dim=3)
574
+ return dec
575
+
576
+ def enable_tiling(self, use_tiling: bool = True):
577
+ self.use_tiling = use_tiling
578
+
579
+ def disable_tiling(self):
580
+ self.enable_tiling(False)
581
+
582
+ def init_from_ckpt(self, path, ignore_keys=list()):
583
+ sd = torch.load(path, map_location="cpu")
584
+ print("init from " + path)
585
+
586
+ if "ema_state_dict" in sd and len(sd['ema_state_dict']) > 0 and os.environ.get("NOT_USE_EMA_MODEL", 0) == 0:
587
+ print("Load from ema model!")
588
+ sd = sd["ema_state_dict"]
589
+ sd = {key.replace("module.", ""): value for key, value in sd.items()}
590
+ elif "state_dict" in sd:
591
+ print("Load from normal model!")
592
+ if "gen_model" in sd["state_dict"]:
593
+ sd = sd["state_dict"]["gen_model"]
594
+ else:
595
+ sd = sd["state_dict"]
596
+
597
+ keys = list(sd.keys())
598
+
599
+ for k in keys:
600
+ for ik in ignore_keys:
601
+ if k.startswith(ik):
602
+ print("Deleting key {} from state_dict.".format(k))
603
+ del sd[k]
604
+
605
+ self.load_state_dict(sd, strict=True)
606
+
607
+
608
+
609
+
610
+
causalvideovae/model/configuration_videobase.py ADDED
@@ -0,0 +1,44 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ import json
2
+ import yaml
3
+ from typing import TypeVar, Dict, Any
4
+ from diffusers import ConfigMixin
5
+
6
+ T = TypeVar('T', bound='VideoBaseConfiguration')
7
+ class VideoBaseConfiguration(ConfigMixin):
8
+ config_name = "VideoBaseConfiguration"
9
+ _nested_config_fields: Dict[str, Any] = {}
10
+
11
+ def __init__(self, **kwargs):
12
+ pass
13
+
14
+ def to_dict(self) -> Dict[str, Any]:
15
+ d = {}
16
+ for key, value in vars(self).items():
17
+ if isinstance(value, VideoBaseConfiguration):
18
+ d[key] = value.to_dict() # Serialize nested VideoBaseConfiguration instances
19
+ elif isinstance(value, tuple):
20
+ d[key] = list(value)
21
+ else:
22
+ d[key] = value
23
+ return d
24
+
25
+ def to_yaml_file(self, yaml_path: str):
26
+ with open(yaml_path, 'w') as yaml_file:
27
+ yaml.dump(self.to_dict(), yaml_file, default_flow_style=False)
28
+
29
+ @classmethod
30
+ def load_from_yaml(cls: T, yaml_path: str) -> T:
31
+ with open(yaml_path, 'r') as yaml_file:
32
+ config_dict = yaml.safe_load(yaml_file)
33
+ for field, field_type in cls._nested_config_fields.items():
34
+ if field in config_dict:
35
+ config_dict[field] = field_type.load_from_dict(config_dict[field])
36
+ return cls(**config_dict)
37
+
38
+ @classmethod
39
+ def load_from_dict(cls: T, config_dict: Dict[str, Any]) -> T:
40
+ # Process nested configuration objects
41
+ for field, field_type in cls._nested_config_fields.items():
42
+ if field in config_dict:
43
+ config_dict[field] = field_type.load_from_dict(config_dict[field])
44
+ return cls(**config_dict)
causalvideovae/model/losses/__init__.py ADDED
@@ -0,0 +1 @@
 
 
1
+ from .perceptual_loss import SimpleLPIPS, LPIPSWithDiscriminator, LPIPSWithDiscriminator3D, LPIPSWithDiscriminator3Drefiner
causalvideovae/model/losses/discriminator.py ADDED
@@ -0,0 +1,181 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ import functools
2
+ import torch.nn as nn
3
+ from ..modules.normalize import ActNorm
4
+ from ..modules.conv import CausalConv3d
5
+ from einops import rearrange
6
+
7
+ def weights_init(m):
8
+ classname = m.__class__.__name__
9
+ if classname.find('Conv') != -1:
10
+ nn.init.normal_(m.weight.data, 0.0, 0.02)
11
+ elif classname.find('BatchNorm') != -1:
12
+ nn.init.normal_(m.weight.data, 1.0, 0.02)
13
+ nn.init.constant_(m.bias.data, 0)
14
+
15
+ def weights_init_conv(m):
16
+ if hasattr(m, 'conv'):
17
+ m = m.conv
18
+ classname = m.__class__.__name__
19
+ if classname.find('Conv') != -1:
20
+ nn.init.normal_(m.weight.data, 0.0, 0.02)
21
+ elif classname.find('BatchNorm') != -1:
22
+ nn.init.normal_(m.weight.data, 1.0, 0.02)
23
+ nn.init.constant_(m.bias.data, 0)
24
+
25
+ class NLayerDiscriminator(nn.Module):
26
+ """Defines a PatchGAN discriminator as in Pix2Pix
27
+ --> see https://github.com/junyanz/pytorch-CycleGAN-and-pix2pix/blob/master/models/networks.py
28
+ """
29
+ def __init__(self, input_nc=3, ndf=64, n_layers=3, use_actnorm=False):
30
+ """Construct a PatchGAN discriminator
31
+ Parameters:
32
+ input_nc (int) -- the number of channels in input images
33
+ ndf (int) -- the number of filters in the last conv layer
34
+ n_layers (int) -- the number of conv layers in the discriminator
35
+ norm_layer -- normalization layer
36
+ """
37
+ super(NLayerDiscriminator, self).__init__()
38
+ if not use_actnorm:
39
+ norm_layer = nn.BatchNorm2d
40
+ else:
41
+ norm_layer = ActNorm
42
+ if type(norm_layer) == functools.partial: # no need to use bias as BatchNorm2d has affine parameters
43
+ use_bias = norm_layer.func != nn.BatchNorm2d
44
+ else:
45
+ use_bias = norm_layer != nn.BatchNorm2d
46
+
47
+ kw = 4
48
+ padw = 1
49
+ sequence = [nn.Conv2d(input_nc, ndf, kernel_size=kw, stride=2, padding=padw), nn.LeakyReLU(0.2, True)]
50
+ nf_mult = 1
51
+ nf_mult_prev = 1
52
+ for n in range(1, n_layers): # gradually increase the number of filters
53
+ nf_mult_prev = nf_mult
54
+ nf_mult = min(2 ** n, 8)
55
+ sequence += [
56
+ nn.Conv2d(ndf * nf_mult_prev, ndf * nf_mult, kernel_size=kw, stride=2, padding=padw, bias=use_bias),
57
+ norm_layer(ndf * nf_mult),
58
+ nn.LeakyReLU(0.2, True)
59
+ ]
60
+
61
+ nf_mult_prev = nf_mult
62
+ nf_mult = min(2 ** n_layers, 8)
63
+ sequence += [
64
+ nn.Conv2d(ndf * nf_mult_prev, ndf * nf_mult, kernel_size=kw, stride=1, padding=padw, bias=use_bias),
65
+ norm_layer(ndf * nf_mult),
66
+ nn.LeakyReLU(0.2, True)
67
+ ]
68
+
69
+ sequence += [
70
+ nn.Conv2d(ndf * nf_mult, 1, kernel_size=kw, stride=1, padding=padw)] # output 1 channel prediction map
71
+ self.main = nn.Sequential(*sequence)
72
+
73
+ def forward(self, input):
74
+ """Standard forward."""
75
+ return self.main(input)
76
+
77
+ class NLayerDiscriminator3D(nn.Module):
78
+ """Defines a 3D PatchGAN discriminator as in Pix2Pix but for 3D inputs."""
79
+ def __init__(self, input_nc=1, ndf=64, n_layers=3, use_actnorm=False):
80
+ """
81
+ Construct a 3D PatchGAN discriminator
82
+
83
+ Parameters:
84
+ input_nc (int) -- the number of channels in input volumes
85
+ ndf (int) -- the number of filters in the last conv layer
86
+ n_layers (int) -- the number of conv layers in the discriminator
87
+ use_actnorm (bool) -- flag to use actnorm instead of batchnorm
88
+ """
89
+ super(NLayerDiscriminator3D, self).__init__()
90
+ if not use_actnorm:
91
+ norm_layer = nn.BatchNorm3d
92
+ else:
93
+ raise NotImplementedError("Not implemented.")
94
+ if type(norm_layer) == functools.partial:
95
+ use_bias = norm_layer.func != nn.BatchNorm3d
96
+ else:
97
+ use_bias = norm_layer != nn.BatchNorm3d
98
+
99
+ kw = 3
100
+ padw = 1
101
+ sequence = [nn.Conv3d(input_nc, ndf, kernel_size=kw, stride=2, padding=padw), nn.LeakyReLU(0.2, True)]
102
+ nf_mult = 1
103
+ nf_mult_prev = 1
104
+ for n in range(1, n_layers): # gradually increase the number of filters
105
+ nf_mult_prev = nf_mult
106
+ nf_mult = min(2 ** n, 8)
107
+ sequence += [
108
+ nn.Conv3d(ndf * nf_mult_prev, ndf * nf_mult, kernel_size=(kw, kw, kw), stride=(2 if n==1 else 1,2,2), padding=padw, bias=use_bias),
109
+ norm_layer(ndf * nf_mult),
110
+ nn.LeakyReLU(0.2, True)
111
+ ]
112
+
113
+ nf_mult_prev = nf_mult
114
+ nf_mult = min(2 ** n_layers, 8)
115
+ sequence += [
116
+ nn.Conv3d(ndf * nf_mult_prev, ndf * nf_mult, kernel_size=(kw, kw, kw), stride=1, padding=padw, bias=use_bias),
117
+ norm_layer(ndf * nf_mult),
118
+ nn.LeakyReLU(0.2, True)
119
+ ]
120
+
121
+ sequence += [nn.Conv3d(ndf * nf_mult, 1, kernel_size=kw, stride=1, padding=padw)] # output 1 channel prediction map
122
+ self.main = nn.Sequential(*sequence)
123
+
124
+ def forward(self, input):
125
+ """Standard forward."""
126
+ return self.main(input)
127
+
128
+
129
+
130
+
131
+
132
+ # class NLayerDiscriminator3D(nn.Module):
133
+ # """Defines a 3D PatchGAN discriminator as in Pix2Pix but for 3D inputs."""
134
+ # def __init__(self, input_nc=1, ndf=64, n_layers=3, use_actnorm=False):
135
+ # """
136
+ # Construct a 3D PatchGAN discriminator
137
+
138
+ # Parameters:
139
+ # input_nc (int) -- the number of channels in input volumes
140
+ # ndf (int) -- the number of filters in the last conv layer
141
+ # n_layers (int) -- the number of conv layers in the discriminator
142
+ # use_actnorm (bool) -- flag to use actnorm instead of batchnorm
143
+ # """
144
+ # super(NLayerDiscriminator3D, self).__init__()
145
+ # if not use_actnorm:
146
+ # norm_layer = nn.BatchNorm3d
147
+ # else:
148
+ # raise NotImplementedError("Not implemented.")
149
+ # if type(norm_layer) == functools.partial:
150
+ # use_bias = norm_layer.func != nn.BatchNorm3d
151
+ # else:
152
+ # use_bias = norm_layer != nn.BatchNorm3d
153
+
154
+ # kw = 4
155
+ # padw = 1
156
+ # sequence = [CausalConv3d(input_nc, ndf, kernel_size=kw, stride=2, padding=padw), nn.LeakyReLU(0.2, True)]
157
+ # nf_mult = 1
158
+ # nf_mult_prev = 1
159
+ # for n in range(1, n_layers): # gradually increase the number of filters
160
+ # nf_mult_prev = nf_mult
161
+ # nf_mult = min(2 ** n, 8)
162
+ # sequence += [
163
+ # CausalConv3d(ndf * nf_mult_prev, ndf * nf_mult, kernel_size=(kw, kw, kw), stride=(2 if n==1 else 1,2,2), padding=padw, bias=use_bias),
164
+ # norm_layer(ndf * nf_mult),
165
+ # nn.LeakyReLU(0.2, True)
166
+ # ]
167
+
168
+ # nf_mult_prev = nf_mult
169
+ # nf_mult = min(2 ** n_layers, 8)
170
+ # sequence += [
171
+ # CausalConv3d(ndf * nf_mult_prev, ndf * nf_mult, kernel_size=(kw, kw, kw), stride=1, padding=padw, bias=use_bias),
172
+ # norm_layer(ndf * nf_mult),
173
+ # nn.LeakyReLU(0.2, True)
174
+ # ]
175
+
176
+ # sequence += [CausalConv3d(ndf * nf_mult, 1, kernel_size=kw, stride=1, padding=padw)] # output 1 channel prediction map
177
+ # self.main = nn.Sequential(*sequence)
178
+
179
+ # def forward(self, input):
180
+ # """Standard forward."""
181
+ # return self.main(input)
causalvideovae/model/modeling_videobase.py ADDED
@@ -0,0 +1,53 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ import torch
2
+ from diffusers import ModelMixin, ConfigMixin
3
+ from torch import nn
4
+ import os
5
+ import json
6
+ import pytorch_lightning as pl
7
+ from diffusers.configuration_utils import ConfigMixin
8
+ from diffusers.models.modeling_utils import ModelMixin
9
+ from typing import Optional, Union
10
+ import glob
11
+
12
+
13
+ class VideoBaseAE(ModelMixin, ConfigMixin):
14
+ config_name = "config.json"
15
+
16
+ def __init__(self, *args, **kwargs) -> None:
17
+ super().__init__(*args, **kwargs)
18
+
19
+ def encode(self, x: torch.Tensor, *args, **kwargs):
20
+ pass
21
+
22
+ def decode(self, encoding: torch.Tensor, *args, **kwargs):
23
+ pass
24
+
25
+ @property
26
+ def num_training_steps(self) -> int:
27
+ """Total training steps inferred from datamodule and devices."""
28
+ if self.trainer.max_steps:
29
+ return self.trainer.max_steps
30
+
31
+ limit_batches = self.trainer.limit_train_batches
32
+ batches = len(self.train_dataloader())
33
+ batches = min(batches, limit_batches) if isinstance(limit_batches, int) else int(limit_batches * batches)
34
+
35
+ num_devices = max(1, self.trainer.num_gpus, self.trainer.num_processes)
36
+ if self.trainer.tpu_cores:
37
+ num_devices = max(num_devices, self.trainer.tpu_cores)
38
+
39
+ effective_accum = self.trainer.accumulate_grad_batches * num_devices
40
+ return (batches // effective_accum) * self.trainer.max_epochs
41
+
42
+ @classmethod
43
+ def from_pretrained(cls, pretrained_model_name_or_path: Optional[Union[str, os.PathLike]], **kwargs):
44
+ ckpt_files = glob.glob(os.path.join(pretrained_model_name_or_path, '*.ckpt'))
45
+ if ckpt_files:
46
+ # Adapt to checkpoint
47
+ last_ckpt_file = ckpt_files[-1]
48
+ config_file = os.path.join(pretrained_model_name_or_path, cls.config_name)
49
+ model = cls.from_config(config_file)
50
+ model.init_from_ckpt(last_ckpt_file)
51
+ return model
52
+ else:
53
+ return super().from_pretrained(pretrained_model_name_or_path, **kwargs)
causalvideovae/model/modules/__init__.py ADDED
@@ -0,0 +1,6 @@
 
 
 
 
 
 
 
1
+ from .block import Block
2
+ from .attention import *
3
+ from .conv import *
4
+ from .normalize import *
5
+ from .resnet_block import *
6
+ from .updownsample import *
causalvideovae/model/modules/conv.py ADDED
@@ -0,0 +1,142 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ import torch.nn as nn
2
+ from typing import Union, Tuple
3
+ import torch.nn.functional as F
4
+ import torch
5
+ from .block import Block
6
+ from .ops import cast_tuple
7
+ from einops import rearrange
8
+ from .ops import video_to_image
9
+ from torch.utils.checkpoint import checkpoint
10
+
11
+ class Conv2d(nn.Conv2d):
12
+ def __init__(
13
+ self,
14
+ in_channels: int,
15
+ out_channels: int,
16
+ kernel_size: Union[int, Tuple[int]] = 3,
17
+ stride: Union[int, Tuple[int]] = 1,
18
+ padding: Union[str, int, Tuple[int]] = 0,
19
+ dilation: Union[int, Tuple[int]] = 1,
20
+ groups: int = 1,
21
+ bias: bool = True,
22
+ padding_mode: str = "zeros",
23
+ device=None,
24
+ dtype=None,
25
+ ) -> None:
26
+ super().__init__(
27
+ in_channels,
28
+ out_channels,
29
+ kernel_size,
30
+ stride,
31
+ padding,
32
+ dilation,
33
+ groups,
34
+ bias,
35
+ padding_mode,
36
+ device,
37
+ dtype,
38
+ )
39
+
40
+ @video_to_image
41
+ def forward(self, x):
42
+ return super().forward(x)
43
+
44
+ class Conv3d(nn.Conv3d):
45
+ def __init__(
46
+ self,
47
+ in_channels: int,
48
+ out_channels: int,
49
+ kernel_size: Union[int, Tuple[int]] = 3,
50
+ stride: Union[int, Tuple[int]] = 1,
51
+ padding: Union[str, int, Tuple[int]] = 0,
52
+ dilation: Union[int, Tuple[int]] = 1,
53
+ groups: int = 1,
54
+ bias: bool = True,
55
+ padding_mode: str = "reflect",
56
+ device=None,
57
+ dtype=None,
58
+ ) -> None:
59
+ super().__init__(
60
+ in_channels,
61
+ out_channels,
62
+ kernel_size,
63
+ stride,
64
+ padding,
65
+ dilation,
66
+ groups,
67
+ bias,
68
+ padding_mode,
69
+ device,
70
+ dtype,
71
+ )
72
+
73
+ def forward(self, x):
74
+ return super().forward(x)
75
+
76
+ class CausalConv3d(nn.Module):
77
+ def __init__(
78
+ self, chan_in, chan_out, kernel_size: Union[int, Tuple[int, int, int]], init_method="random", **kwargs
79
+ ):
80
+ super().__init__()
81
+ self.kernel_size = cast_tuple(kernel_size, 3)
82
+ self.time_kernel_size = self.kernel_size[0]
83
+ self.chan_in = chan_in
84
+ self.chan_out = chan_out
85
+ stride = kwargs.pop("stride", 1)
86
+ self.padding = kwargs.pop("padding", 0)
87
+ self.padding = list(cast_tuple(self.padding, 3))
88
+ self.padding[0] = 0
89
+ stride = cast_tuple(stride, 3)
90
+ self.conv = nn.Conv3d(chan_in, chan_out, self.kernel_size, stride=stride, padding=self.padding)
91
+ self._init_weights(init_method)
92
+
93
+ def _init_weights(self, init_method):
94
+ ks = torch.tensor(self.kernel_size)
95
+ if init_method == "avg":
96
+ assert (
97
+ self.kernel_size[1] == 1 and self.kernel_size[2] == 1
98
+ ), "only support temporal up/down sample"
99
+ assert self.chan_in == self.chan_out, "chan_in must be equal to chan_out"
100
+ weight = torch.zeros((self.chan_out, self.chan_in, *self.kernel_size))
101
+
102
+ eyes = torch.concat(
103
+ [
104
+ torch.eye(self.chan_in).unsqueeze(-1) * 1/3,
105
+ torch.eye(self.chan_in).unsqueeze(-1) * 1/3,
106
+ torch.eye(self.chan_in).unsqueeze(-1) * 1/3,
107
+ ],
108
+ dim=-1,
109
+ )
110
+ weight[:, :, :, 0, 0] = eyes
111
+
112
+ self.conv.weight = nn.Parameter(
113
+ weight,
114
+ requires_grad=True,
115
+ )
116
+ elif init_method == "zero":
117
+ self.conv.weight = nn.Parameter(
118
+ torch.zeros((self.chan_out, self.chan_in, *self.kernel_size)),
119
+ requires_grad=True,
120
+ )
121
+ if self.conv.bias is not None:
122
+ nn.init.constant_(self.conv.bias, 0)
123
+
124
+ def forward(self, x):
125
+ # 1 + 16 16 as video, 1 as image
126
+ first_frame_pad = x[:, :, :1, :, :].repeat(
127
+ (1, 1, self.time_kernel_size - 1, 1, 1)
128
+ ) # b c t h w
129
+ x = torch.concatenate((first_frame_pad, x), dim=2) # 3 + 16
130
+ return self.conv(x)
131
+
132
+
133
+ class CausalConv3d_GC(CausalConv3d):
134
+ def __init__(self, chan_in, chan_out, kernel_size, init_method="random", **kwargs):
135
+ super().__init__(chan_in, chan_out, kernel_size, init_method, **kwargs)
136
+ def forward(self, x):
137
+ # 1 + 16 16 as video, 1 as image
138
+ first_frame_pad = x[:, :, :1, :, :].repeat(
139
+ (1, 1, self.time_kernel_size - 1, 1, 1)
140
+ ) # b c t h w
141
+ x = torch.concatenate((first_frame_pad, x), dim=2) # 3 + 16
142
+ return checkpoint(self.conv, x)
causalvideovae/model/modules/quant.py ADDED
@@ -0,0 +1,100 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ import torch
2
+ import torch.nn as nn
3
+ import torch.distributed as dist
4
+ import numpy as np
5
+ import torch.nn.functional as F
6
+ from .ops import shift_dim
7
+
8
+ class Codebook(nn.Module):
9
+ def __init__(self, n_codes, embedding_dim):
10
+ super().__init__()
11
+ self.register_buffer("embeddings", torch.randn(n_codes, embedding_dim))
12
+ self.register_buffer("N", torch.zeros(n_codes))
13
+ self.register_buffer("z_avg", self.embeddings.data.clone())
14
+
15
+ self.n_codes = n_codes
16
+ self.embedding_dim = embedding_dim
17
+ self._need_init = True
18
+
19
+ def _tile(self, x):
20
+ d, ew = x.shape
21
+ if d < self.n_codes:
22
+ n_repeats = (self.n_codes + d - 1) // d
23
+ std = 0.01 / np.sqrt(ew)
24
+ x = x.repeat(n_repeats, 1)
25
+ x = x + torch.randn_like(x) * std
26
+ return x
27
+
28
+ def _init_embeddings(self, z):
29
+ # z: [b, c, t, h, w]
30
+ self._need_init = False
31
+ flat_inputs = shift_dim(z, 1, -1).flatten(end_dim=-2)
32
+ y = self._tile(flat_inputs)
33
+
34
+ d = y.shape[0]
35
+ _k_rand = y[torch.randperm(y.shape[0])][: self.n_codes]
36
+ if dist.is_initialized():
37
+ dist.broadcast(_k_rand, 0)
38
+ self.embeddings.data.copy_(_k_rand)
39
+ self.z_avg.data.copy_(_k_rand)
40
+ self.N.data.copy_(torch.ones(self.n_codes))
41
+
42
+ def forward(self, z):
43
+ # z: [b, c, t, h, w]
44
+ if self._need_init and self.training:
45
+ self._init_embeddings(z)
46
+ flat_inputs = shift_dim(z, 1, -1).flatten(end_dim=-2)
47
+ distances = (
48
+ (flat_inputs**2).sum(dim=1, keepdim=True)
49
+ - 2 * flat_inputs @ self.embeddings.t()
50
+ + (self.embeddings.t() ** 2).sum(dim=0, keepdim=True)
51
+ )
52
+
53
+ encoding_indices = torch.argmin(distances, dim=1)
54
+ encode_onehot = F.one_hot(encoding_indices, self.n_codes).type_as(flat_inputs)
55
+ encoding_indices = encoding_indices.view(z.shape[0], *z.shape[2:])
56
+
57
+ embeddings = F.embedding(encoding_indices, self.embeddings)
58
+ embeddings = shift_dim(embeddings, -1, 1)
59
+
60
+ commitment_loss = 0.25 * F.mse_loss(z, embeddings.detach())
61
+
62
+ # EMA codebook update
63
+ if self.training:
64
+ n_total = encode_onehot.sum(dim=0)
65
+ encode_sum = flat_inputs.t() @ encode_onehot
66
+ if dist.is_initialized():
67
+ dist.all_reduce(n_total)
68
+ dist.all_reduce(encode_sum)
69
+
70
+ self.N.data.mul_(0.99).add_(n_total, alpha=0.01)
71
+ self.z_avg.data.mul_(0.99).add_(encode_sum.t(), alpha=0.01)
72
+
73
+ n = self.N.sum()
74
+ weights = (self.N + 1e-7) / (n + self.n_codes * 1e-7) * n
75
+ encode_normalized = self.z_avg / weights.unsqueeze(1)
76
+ self.embeddings.data.copy_(encode_normalized)
77
+
78
+ y = self._tile(flat_inputs)
79
+ _k_rand = y[torch.randperm(y.shape[0])][: self.n_codes]
80
+ if dist.is_initialized():
81
+ dist.broadcast(_k_rand, 0)
82
+
83
+ usage = (self.N.view(self.n_codes, 1) >= 1).float()
84
+ self.embeddings.data.mul_(usage).add_(_k_rand * (1 - usage))
85
+
86
+ embeddings_st = (embeddings - z).detach() + z
87
+
88
+ avg_probs = torch.mean(encode_onehot, dim=0)
89
+ perplexity = torch.exp(-torch.sum(avg_probs * torch.log(avg_probs + 1e-10)))
90
+
91
+ return dict(
92
+ embeddings=embeddings_st,
93
+ encodings=encoding_indices,
94
+ commitment_loss=commitment_loss,
95
+ perplexity=perplexity,
96
+ )
97
+
98
+ def dictionary_lookup(self, encodings):
99
+ embeddings = F.embedding(encodings, self.embeddings)
100
+ return embeddings
causalvideovae/model/modules/resnet_block.py ADDED
@@ -0,0 +1,218 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ import torch
2
+ import torch.nn as nn
3
+ from einops import rearrange, pack, unpack
4
+ from .normalize import Normalize
5
+ from .ops import nonlinearity, video_to_image
6
+ from .conv import CausalConv3d, CausalConv3d_GC
7
+ from .block import Block
8
+ from torch.utils.checkpoint import checkpoint
9
+
10
+ class ResnetBlock2D(Block):
11
+ def __init__(self, *, in_channels, out_channels=None, conv_shortcut=False,
12
+ dropout):
13
+ super().__init__()
14
+ self.in_channels = in_channels
15
+ self.out_channels = in_channels if out_channels is None else out_channels
16
+ self.use_conv_shortcut = conv_shortcut
17
+
18
+ self.norm1 = Normalize(in_channels)
19
+ self.conv1 = torch.nn.Conv2d(
20
+ in_channels, out_channels, kernel_size=3, stride=1, padding=1
21
+ )
22
+ self.norm2 = Normalize(out_channels)
23
+ self.dropout = torch.nn.Dropout(dropout)
24
+ self.conv2 = torch.nn.Conv2d(
25
+ out_channels, out_channels, kernel_size=3, stride=1, padding=1
26
+ )
27
+ if self.in_channels != self.out_channels:
28
+ if self.use_conv_shortcut:
29
+ self.conv_shortcut = torch.nn.Conv2d(
30
+ in_channels, out_channels, kernel_size=3, stride=1, padding=1
31
+ )
32
+ else:
33
+ self.nin_shortcut = torch.nn.Conv2d(
34
+ in_channels, out_channels, kernel_size=1, stride=1, padding=0
35
+ )
36
+
37
+ @video_to_image
38
+ def forward(self, x):
39
+ h = x
40
+ h = h.permute(0, 2, 3, 1)
41
+ h = self.norm1(h)
42
+ h = h.permute(0, 3, 1, 2)
43
+ h = nonlinearity(h)
44
+ h = self.conv1(h)
45
+ h = h.permute(0, 2, 3, 1)
46
+ h = self.norm2(h)
47
+ h = h.permute(0, 3, 1, 2)
48
+ h = nonlinearity(h)
49
+ h = self.dropout(h)
50
+ h = self.conv2(h)
51
+ if self.in_channels != self.out_channels:
52
+ if self.use_conv_shortcut:
53
+ x = self.conv_shortcut(x)
54
+ else:
55
+ x = self.nin_shortcut(x)
56
+ x = x + h
57
+ return x
58
+
59
+ class ResnetBlockNorm3D(Block):
60
+ def __init__(self, *, in_channels, out_channels=None, conv_shortcut=False,
61
+ dropout):
62
+ super().__init__()
63
+ self.in_channels = in_channels
64
+ self.out_channels = in_channels if out_channels is None else out_channels
65
+ self.use_conv_shortcut = conv_shortcut
66
+
67
+ self.norm1 = Normalize(in_channels)
68
+ self.conv1 = torch.nn.Conv3d(
69
+ in_channels, out_channels, kernel_size=3, stride=1, padding=1, padding_mode='reflect'
70
+ )
71
+ self.norm2 = Normalize(out_channels)
72
+ self.dropout = torch.nn.Dropout(dropout)
73
+ self.conv2 = torch.nn.Conv3d(
74
+ out_channels, out_channels, kernel_size=3, stride=1, padding=1, padding_mode='reflect'
75
+ )
76
+ if self.in_channels != self.out_channels:
77
+ if self.use_conv_shortcut:
78
+ self.conv_shortcut = torch.nn.Conv3d(
79
+ in_channels, out_channels, kernel_size=3, stride=1, padding=1
80
+ )
81
+ else:
82
+ self.nin_shortcut = torch.nn.Conv3d(
83
+ in_channels, out_channels, kernel_size=1, stride=1, padding=0
84
+ )
85
+
86
+ def forward(self, x):
87
+ h = x
88
+ h = self.norm1(h)
89
+ h = nonlinearity(h)
90
+ h = self.conv1(h)
91
+ h = self.norm2(h)
92
+ h = nonlinearity(h)
93
+ h = self.dropout(h)
94
+ h = self.conv2(h)
95
+ if self.in_channels != self.out_channels:
96
+ if self.use_conv_shortcut:
97
+ x = self.conv_shortcut(x)
98
+ else:
99
+ x = self.nin_shortcut(x)
100
+ x = x + h
101
+ return x
102
+
103
+ class ResnetBlock23D(Block):
104
+ def __init__(self, *, in_channels, out_channels=None, conv_shortcut=False, dropout):
105
+ super().__init__()
106
+ self.in_channels = in_channels
107
+ self.out_channels = in_channels if out_channels is None else out_channels
108
+ self.use_conv_shortcut = conv_shortcut
109
+
110
+ self.norm1 = Normalize(in_channels)
111
+ self.conv1 = CausalConv3d(in_channels, out_channels, 3, padding=1)
112
+ self.norm2 = Normalize(out_channels)
113
+ self.dropout = torch.nn.Dropout(dropout)
114
+ self.conv2 = nn.Conv2d(out_channels, out_channels, kernel_size=3, stride=1, padding=1)
115
+ if self.in_channels != self.out_channels:
116
+ if self.use_conv_shortcut:
117
+ self.conv_shortcut = CausalConv3d(in_channels, out_channels, 3, padding=1)
118
+ else:
119
+ self.nin_shortcut = CausalConv3d(in_channels, out_channels, 1, padding=0)
120
+
121
+ def forward(self, x):
122
+ h = x
123
+ h = self.norm1(h)
124
+ h = nonlinearity(h)
125
+ h = self.conv1(h)
126
+ h = self.norm2(h)
127
+ h = nonlinearity(h)
128
+ h = self.dropout(h)
129
+ if x.dim() == 5:
130
+ B, C, T, H, W = x.shape
131
+ h = rearrange(h, "b c t h w -> (b t) c h w")
132
+ h = self.conv2(h)
133
+ h = rearrange(h, "(b t) c h w -> b c t h w", b=B, t=T)
134
+ else:
135
+ h = self.conv2(h)
136
+
137
+ if self.in_channels != self.out_channels:
138
+ if self.use_conv_shortcut:
139
+ x = self.conv_shortcut(x)
140
+ else:
141
+ x = self.nin_shortcut(x)
142
+ return x + h
143
+
144
+ class ResnetBlock3D(Block):
145
+ def __init__(self, *, in_channels, out_channels=None, conv_shortcut=False, dropout):
146
+ super().__init__()
147
+ self.in_channels = in_channels
148
+ self.out_channels = in_channels if out_channels is None else out_channels
149
+ self.use_conv_shortcut = conv_shortcut
150
+
151
+ self.norm1 = Normalize(in_channels)
152
+ self.conv1 = CausalConv3d(in_channels, out_channels, 3, padding=1)
153
+ self.norm2 = Normalize(out_channels)
154
+ self.dropout = torch.nn.Dropout(dropout)
155
+ self.conv2 = CausalConv3d(out_channels, out_channels, 3, padding=1)
156
+ if self.in_channels != self.out_channels:
157
+ if self.use_conv_shortcut:
158
+ self.conv_shortcut = CausalConv3d(in_channels, out_channels, 3, padding=1)
159
+ else:
160
+ self.nin_shortcut = CausalConv3d(in_channels, out_channels, 1, padding=0)
161
+
162
+ def forward(self, x):
163
+ h = x
164
+ h = h.permute(0, 2, 3, 4, 1)
165
+ h = self.norm1(h)
166
+ h = h.permute(0, 4, 1, 2, 3)
167
+ h = nonlinearity(h)
168
+ h = self.conv1(h)
169
+ h = h.permute(0, 2, 3, 4, 1)
170
+ h = self.norm2(h)
171
+ h = h.permute(0, 4, 1, 2, 3)
172
+ h = nonlinearity(h)
173
+ h = self.dropout(h)
174
+ h = self.conv2(h)
175
+ if self.in_channels != self.out_channels:
176
+ if self.use_conv_shortcut:
177
+ x = self.conv_shortcut(x)
178
+ else:
179
+ x = self.nin_shortcut(x)
180
+ return x + h
181
+
182
+
183
+ class ResnetBlock3D_GC(Block):
184
+ def __init__(self, *, in_channels, out_channels=None, conv_shortcut=False, dropout):
185
+ super().__init__()
186
+ self.in_channels = in_channels
187
+ self.out_channels = in_channels if out_channels is None else out_channels
188
+ self.use_conv_shortcut = conv_shortcut
189
+
190
+ self.norm1 = Normalize(in_channels)
191
+ self.conv1 = CausalConv3d(in_channels, out_channels, 3, padding=1)
192
+ self.norm2 = Normalize(out_channels)
193
+ self.dropout = torch.nn.Dropout(dropout)
194
+ self.conv2 = CausalConv3d(out_channels, out_channels, 3, padding=1)
195
+ if self.in_channels != self.out_channels:
196
+ if self.use_conv_shortcut:
197
+ self.conv_shortcut = CausalConv3d(in_channels, out_channels, 3, padding=1)
198
+ else:
199
+ self.nin_shortcut = CausalConv3d(in_channels, out_channels, 1, padding=0)
200
+
201
+ def forward(self, x):
202
+ return checkpoint(self._forward, x, use_reentrant=True)
203
+
204
+ def _forward(self, x):
205
+ h = x
206
+ h = self.norm1(h)
207
+ h = nonlinearity(h)
208
+ h = self.conv1(h)
209
+ h = self.norm2(h)
210
+ h = nonlinearity(h)
211
+ h = self.dropout(h)
212
+ h = self.conv2(h)
213
+ if self.in_channels != self.out_channels:
214
+ if self.use_conv_shortcut:
215
+ x = self.conv_shortcut(x)
216
+ else:
217
+ x = self.nin_shortcut(x)
218
+ return x + h
causalvideovae/model/modules/updownsample.py ADDED
@@ -0,0 +1,370 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ from typing import Union, Tuple
2
+ import torch
3
+ import torch.nn as nn
4
+ import torch.nn.functional as F
5
+ from .resnet_block import ResnetBlock3D
6
+ from .attention import TemporalAttnBlock
7
+ from .normalize import Normalize
8
+ from .ops import cast_tuple, video_to_image
9
+ from .conv import CausalConv3d, CausalConv3d_GC
10
+ from einops import rearrange
11
+ from .block import Block
12
+
13
+ class Upsample(Block):
14
+ def __init__(self, in_channels, out_channels):
15
+ super().__init__()
16
+ self.with_conv = True
17
+ if self.with_conv:
18
+ self.conv = torch.nn.Conv2d(in_channels,
19
+ out_channels,
20
+ kernel_size=3,
21
+ stride=1,
22
+ padding=1)
23
+
24
+ @video_to_image
25
+ def forward(self, x):
26
+ x = torch.nn.functional.interpolate(x, scale_factor=2.0, mode="nearest")
27
+ if self.with_conv:
28
+ x = self.conv(x)
29
+ return x
30
+
31
+ class Downsample(Block):
32
+ def __init__(self, in_channels, out_channels, undown=False):
33
+ super().__init__()
34
+ self.with_conv = True
35
+ self.undown = undown
36
+ if self.with_conv:
37
+ # no asymmetric padding in torch conv, must do it ourselves
38
+ if self.undown:
39
+ self.conv = torch.nn.Conv2d(in_channels,
40
+ out_channels,
41
+ kernel_size=3,
42
+ stride=1,
43
+ padding=1)
44
+ else:
45
+ self.conv = torch.nn.Conv2d(in_channels,
46
+ out_channels,
47
+ kernel_size=3,
48
+ stride=2,
49
+ padding=0)
50
+ @video_to_image
51
+ def forward(self, x):
52
+ if self.with_conv:
53
+ if self.undown:
54
+ x = self.conv(x)
55
+ else:
56
+ pad = (0,1,0,1)
57
+ x = torch.nn.functional.pad(x, pad, mode="constant", value=0)
58
+ x = self.conv(x)
59
+ else:
60
+ x = torch.nn.functional.avg_pool2d(x, kernel_size=2, stride=2)
61
+ return x
62
+
63
+ class SpatialDownsample2x(Block):
64
+ def __init__(
65
+ self,
66
+ chan_in,
67
+ chan_out,
68
+ kernel_size: Union[int, Tuple[int]] = (3, 3),
69
+ stride: Union[int, Tuple[int]] = (2, 2),
70
+ **kwargs
71
+ ):
72
+ super().__init__()
73
+ kernel_size = cast_tuple(kernel_size, 2)
74
+ stride = cast_tuple(stride, 2)
75
+ self.chan_in = chan_in
76
+ self.chan_out = chan_out
77
+ self.kernel_size = kernel_size
78
+ self.conv = CausalConv3d(
79
+ self.chan_in,
80
+ self.chan_out,
81
+ (1,) + self.kernel_size,
82
+ stride=(1, ) + stride,
83
+ padding=0
84
+ )
85
+
86
+ def forward(self, x):
87
+ pad = (0,1,0,1,0,0)
88
+ x = torch.nn.functional.pad(x, pad, mode="constant", value=0)
89
+ x = self.conv(x)
90
+ return x
91
+
92
+ class SpatialUpsample2x_GC(Block):
93
+ def __init__(
94
+ self,
95
+ chan_in,
96
+ chan_out,
97
+ kernel_size: Union[int, Tuple[int]] = (3, 3),
98
+ stride: Union[int, Tuple[int]] = (1, 1),
99
+ unup=False,
100
+ ):
101
+ super().__init__()
102
+ self.chan_in = chan_in
103
+ self.chan_out = chan_out
104
+ self.kernel_size = kernel_size
105
+ self.unup = unup
106
+ self.conv = CausalConv3d_GC(
107
+ self.chan_in,
108
+ self.chan_out,
109
+ (1,) + self.kernel_size,
110
+ stride=(1, ) + stride,
111
+ padding=1
112
+ )
113
+
114
+ def forward(self, x):
115
+ if not self.unup:
116
+ t = x.shape[2]
117
+ x = rearrange(x, "b c t h w -> b (c t) h w")
118
+ x = F.interpolate(x, scale_factor=(2,2), mode="nearest")
119
+ x = rearrange(x, "b (c t) h w -> b c t h w", t=t)
120
+ x = self.conv(x)
121
+ return x
122
+
123
+
124
+ class SpatialUpsample2x(Block):
125
+ def __init__(
126
+ self,
127
+ chan_in,
128
+ chan_out,
129
+ kernel_size: Union[int, Tuple[int]] = (3, 3),
130
+ stride: Union[int, Tuple[int]] = (1, 1),
131
+ unup=False,
132
+ ):
133
+ super().__init__()
134
+ self.chan_in = chan_in
135
+ self.chan_out = chan_out
136
+ self.kernel_size = kernel_size
137
+ self.unup = unup
138
+ self.conv = CausalConv3d(
139
+ self.chan_in,
140
+ self.chan_out,
141
+ (1,) + self.kernel_size,
142
+ stride=(1, ) + stride,
143
+ padding=1
144
+ )
145
+
146
+ def forward(self, x):
147
+ if not self.unup:
148
+ t = x.shape[2]
149
+ x = rearrange(x, "b c t h w -> b (c t) h w")
150
+ x = F.interpolate(x, scale_factor=(2,2), mode="nearest")
151
+ x = rearrange(x, "b (c t) h w -> b c t h w", t=t)
152
+ x = self.conv(x)
153
+ return x
154
+
155
+ class SpatialNormUpsample2x(Block):
156
+ def __init__(self, in_channels, out_channels):
157
+ super().__init__()
158
+ self.conv = torch.nn.Conv3d(in_channels, out_channels, kernel_size=3, stride=1, padding=1, padding_mode='reflect')
159
+
160
+ def forward(self, x):
161
+ x = F.interpolate(x, scale_factor=(1,2,2), mode='nearest')
162
+ return self.conv(x)
163
+
164
+ class TimeDownsample2x(Block):
165
+ def __init__(
166
+ self,
167
+ chan_in,
168
+ chan_out,
169
+ kernel_size: int = 3
170
+ ):
171
+ super().__init__()
172
+ self.kernel_size = kernel_size
173
+ self.conv = nn.AvgPool3d((kernel_size,1,1), stride=(2,1,1))
174
+
175
+ def forward(self, x):
176
+ first_frame_pad = x[:, :, :1, :, :].repeat(
177
+ (1, 1, self.kernel_size - 1, 1, 1)
178
+ )
179
+ x = torch.concatenate((first_frame_pad, x), dim=2)
180
+ return self.conv(x)
181
+
182
+ class TimeUpsample2x(Block):
183
+ def __init__(
184
+ self,
185
+ chan_in,
186
+ chan_out
187
+ ):
188
+ super().__init__()
189
+ def forward(self, x):
190
+ if x.size(2) > 1:
191
+ x,x_= x[:,:,:1],x[:,:,1:]
192
+ x_= F.interpolate(x_, scale_factor=(2,1,1), mode='nearest')
193
+ x = torch.concat([x, x_], dim=2)
194
+ return x
195
+
196
+ class TimeDownsampleRes2x(nn.Module):
197
+ def __init__(
198
+ self,
199
+ in_channels,
200
+ out_channels,
201
+ kernel_size: int = 3,
202
+ mix_factor: float = 2.0,
203
+ ):
204
+ super().__init__()
205
+ self.kernel_size = cast_tuple(kernel_size, 3)
206
+ self.avg_pool = nn.AvgPool3d((kernel_size,1,1), stride=(2,1,1))
207
+ self.conv = nn.Conv3d(
208
+ in_channels, out_channels, self.kernel_size, stride=(2,1,1), padding=(0,1,1)
209
+ )
210
+ self.mix_factor = torch.nn.Parameter(torch.Tensor([mix_factor]))
211
+
212
+ def forward(self, x):
213
+ alpha = torch.sigmoid(self.mix_factor)
214
+ first_frame_pad = x[:, :, :1, :, :].repeat(
215
+ (1, 1, self.kernel_size[0] - 1, 1, 1)
216
+ )
217
+ x = torch.concatenate((first_frame_pad, x), dim=2)
218
+ return alpha * self.avg_pool(x) + (1 - alpha) * self.conv(x)
219
+
220
+ class TimeUpsampleRes2x(nn.Module):
221
+ def __init__(
222
+ self,
223
+ in_channels,
224
+ out_channels,
225
+ kernel_size: int = 3,
226
+ mix_factor: float = 2.0,
227
+ ):
228
+ super().__init__()
229
+ self.conv = CausalConv3d(
230
+ in_channels, out_channels, kernel_size, padding=1
231
+ )
232
+ self.mix_factor = torch.nn.Parameter(torch.Tensor([mix_factor]))
233
+
234
+ def forward(self, x):
235
+ alpha = torch.sigmoid(self.mix_factor)
236
+ if x.size(2) > 1:
237
+ x,x_= x[:,:,:1],x[:,:,1:]
238
+ x_= F.interpolate(x_, scale_factor=(2,1,1), mode='nearest')
239
+ x = torch.concat([x, x_], dim=2)
240
+ return alpha * x + (1-alpha) * self.conv(x)
241
+
242
+ class TimeDownsampleResAdv2x(nn.Module):
243
+ def __init__(
244
+ self,
245
+ in_channels,
246
+ out_channels,
247
+ kernel_size: int = 3,
248
+ mix_factor: float = 1.5,
249
+ ):
250
+ super().__init__()
251
+ self.kernel_size = cast_tuple(kernel_size, 3)
252
+ self.avg_pool = nn.AvgPool3d((kernel_size,1,1), stride=(2,1,1))
253
+ self.attn = TemporalAttnBlock(in_channels)
254
+ self.res = ResnetBlock3D(in_channels=in_channels, out_channels=in_channels, dropout=0.0)
255
+ self.conv = nn.Conv3d(
256
+ in_channels, out_channels, self.kernel_size, stride=(2,1,1), padding=(0,1,1)
257
+ )
258
+ self.mix_factor = torch.nn.Parameter(torch.Tensor([mix_factor]))
259
+
260
+ def forward(self, x):
261
+ first_frame_pad = x[:, :, :1, :, :].repeat(
262
+ (1, 1, self.kernel_size[0] - 1, 1, 1)
263
+ )
264
+ x = torch.concatenate((first_frame_pad, x), dim=2)
265
+ alpha = torch.sigmoid(self.mix_factor)
266
+ return alpha * self.avg_pool(x) + (1 - alpha) * self.conv(self.attn((self.res(x))))
267
+
268
+ class TimeUpsampleResAdv2x(nn.Module):
269
+ def __init__(
270
+ self,
271
+ in_channels,
272
+ out_channels,
273
+ kernel_size: int = 3,
274
+ mix_factor: float = 1.5,
275
+ ):
276
+ super().__init__()
277
+ self.res = ResnetBlock3D(in_channels=in_channels, out_channels=in_channels, dropout=0.0)
278
+ self.attn = TemporalAttnBlock(in_channels)
279
+ self.norm = Normalize(in_channels=in_channels)
280
+ self.conv = CausalConv3d(
281
+ in_channels, out_channels, kernel_size, padding=1
282
+ )
283
+ self.mix_factor = torch.nn.Parameter(torch.Tensor([mix_factor]))
284
+
285
+ def forward(self, x):
286
+ if x.size(2) > 1:
287
+ x,x_= x[:,:,:1],x[:,:,1:]
288
+ x_= F.interpolate(x_, scale_factor=(2,1,1), mode='nearest')
289
+ x = torch.concat([x, x_], dim=2)
290
+ alpha = torch.sigmoid(self.mix_factor)
291
+ return alpha * x + (1 - alpha) * self.conv(self.attn(self.res(x)))
292
+
293
+
294
+ class Spatial2xTime2x3DDownsample(Block):
295
+ def __init__(self, in_channels, out_channels):
296
+ super().__init__()
297
+ self.conv = CausalConv3d(in_channels, out_channels, kernel_size=3, padding=0, stride=2)
298
+
299
+ def forward(self, x):
300
+ pad = (0,1,0,1,0,0)
301
+ x = torch.nn.functional.pad(x, pad, mode="constant", value=0)
302
+ x = self.conv(x)
303
+ return x
304
+
305
+ class Spatial2xTime2x3DNormDownsample(Block):
306
+ def __init__(self, in_channels, out_channels):
307
+ super().__init__()
308
+ self.conv = torch.nn.Conv3d(in_channels,
309
+ out_channels,
310
+ kernel_size=3,
311
+ stride=2,
312
+ padding=0)
313
+
314
+ def forward(self, x):
315
+ pad = (0,1,0,1,0,1)
316
+ x = torch.nn.functional.pad(x, pad, mode="reflect")
317
+ x = self.conv(x)
318
+ return x
319
+
320
+ class Spatial2x3DDownsample(Block):
321
+ def __init__(self, in_channels, out_channels):
322
+ super().__init__()
323
+ self.conv = CausalConv3d(in_channels, out_channels, kernel_size=3, padding=0, stride=(1,2,2))
324
+
325
+ def forward(self, x):
326
+ pad = (0,1,0,1,0,0)
327
+ x = torch.nn.functional.pad(x, pad, mode="constant", value=0)
328
+ x = self.conv(x)
329
+ return x
330
+
331
+
332
+ class Spatial2x3DUpsample(Block):
333
+ def __init__(self, in_channels, out_channels):
334
+ super().__init__()
335
+ self.conv = CausalConv3d(in_channels, out_channels, kernel_size=3, padding=1)
336
+
337
+ def forward(self, x):
338
+ x = F.interpolate(x, scale_factor=(1,2,2), mode='nearest')
339
+ return self.conv(x)
340
+
341
+ class Spatial2x3DNormUpsample(Block):
342
+ def __init__(self, in_channels, out_channels):
343
+ super().__init__()
344
+ self.conv = torch.nn.Conv3d(in_channels, out_channels, kernel_size=3, stride=1, padding=1, padding_mode='reflect')
345
+
346
+ def forward(self, x):
347
+ x = F.interpolate(x, scale_factor=(1,2,2), mode='nearest')
348
+ return self.conv(x)
349
+
350
+ class Spatial2xTime2x3DUpsample(Block):
351
+ def __init__(self, in_channels, out_channels):
352
+ super().__init__()
353
+ self.conv = CausalConv3d(in_channels, out_channels, kernel_size=3, padding=1)
354
+ def forward(self, x):
355
+ if x.size(2) > 1:
356
+ x,x_= x[:,:,:1],x[:,:,1:]
357
+ x_= F.interpolate(x_, scale_factor=(2,2,2), mode='nearest')
358
+ x = F.interpolate(x, scale_factor=(1,2,2), mode='nearest')
359
+ x = torch.concat([x, x_], dim=2)
360
+ else:
361
+ x = F.interpolate(x, scale_factor=(1,2,2), mode='nearest')
362
+ return self.conv(x)
363
+
364
+ class Spatial2xTime2x3DNormUpsample(Block):
365
+ def __init__(self, in_channels, out_channels):
366
+ super().__init__()
367
+ self.conv = self.conv = torch.nn.Conv3d(in_channels, out_channels, kernel_size=3, stride=1, padding=1, padding_mode='reflect')
368
+ def forward(self, x):
369
+ x = F.interpolate(x, scale_factor=(2,2,2), mode='nearest')
370
+ return self.conv(x)
causalvideovae/model/refiner/modeling_refiner.py ADDED
@@ -0,0 +1,490 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ from ..modeling_videobase import VideoBaseAE
2
+ from ..modules import Normalize
3
+ from ..modules.ops import nonlinearity
4
+ from typing import List, Tuple
5
+ import torch.nn as nn
6
+
7
+ from ..utils.module_utils import resolve_str_to_obj, Module
8
+ from ..utils.distrib_utils import DiagonalGaussianDistribution
9
+ from ..utils.scheduler_utils import cosine_scheduler
10
+ from ...utils.utils import custom_to_video
11
+
12
+ import torch
13
+ from diffusers.configuration_utils import register_to_config
14
+ from copy import deepcopy
15
+ import os
16
+ import glob
17
+
18
+ import numpy as np
19
+ from ...eval.cal_psnr import calculate_psnr
20
+ from decord import VideoReader, cpu
21
+ from pytorchvideo.transforms import ShortSideScale
22
+ from torchvision.io import read_video
23
+ from torchvision.transforms import Lambda, Compose
24
+ from torchvision.transforms._transforms_video import CenterCropVideo
25
+
26
+ class Encoder(nn.Module):
27
+ def __init__(
28
+ self,
29
+ hidden_size: int,
30
+ hidden_size_mult: Tuple[int] = (1, 2, 4, 4),
31
+ attn_resolutions: Tuple[int] = (16,),
32
+ conv_in: Module = "Conv2d",
33
+ attention: Module = "AttnBlock",
34
+ resnet_blocks: Tuple[Module] = (
35
+ "ResnetBlock2D",
36
+ "ResnetBlock2D",
37
+ "ResnetBlock2D",
38
+ "ResnetBlock3D",
39
+ ),
40
+ spatial_downsample: Tuple[Module] = (
41
+ "Downsample",
42
+ "Downsample",
43
+ "Downsample",
44
+ "",
45
+ ),
46
+ dropout: float = 0.0,
47
+ resolution: int = 256,
48
+ num_res_blocks: int = 2,
49
+ ) -> None:
50
+ super().__init__()
51
+ assert len(resnet_blocks) == len(hidden_size_mult), print(
52
+ hidden_size_mult, resnet_blocks
53
+ )
54
+ # ---- Config ----
55
+ self.num_resolutions = len(hidden_size_mult)
56
+ self.resolution = resolution
57
+ self.num_res_blocks = num_res_blocks
58
+
59
+ # ---- In ----
60
+ self.conv_in = resolve_str_to_obj(conv_in)(
61
+ 3, hidden_size, kernel_size=3, stride=1, padding=1
62
+ )
63
+
64
+ # ---- Downsample ----
65
+ curr_res = resolution
66
+ in_ch_mult = (1,) + tuple(hidden_size_mult)
67
+ self.in_ch_mult = in_ch_mult
68
+ self.down = nn.ModuleList()
69
+ for i_level in range(self.num_resolutions):
70
+ block = nn.ModuleList()
71
+ attn = nn.ModuleList()
72
+ block_in = hidden_size * in_ch_mult[i_level]
73
+ block_out = hidden_size * hidden_size_mult[i_level]
74
+ for i_block in range(self.num_res_blocks):
75
+ block.append(
76
+ resolve_str_to_obj(resnet_blocks[i_level])(
77
+ in_channels=block_in,
78
+ out_channels=block_out,
79
+ dropout=dropout,
80
+ )
81
+ )
82
+ block_in = block_out
83
+ if curr_res in attn_resolutions:
84
+ attn.append(resolve_str_to_obj(attention)(block_in))
85
+ down = nn.Module()
86
+ down.block = block
87
+ down.attn = attn
88
+ if spatial_downsample[i_level]:
89
+ down.downsample = resolve_str_to_obj(spatial_downsample[i_level])(
90
+ block_in, block_in
91
+ )
92
+ curr_res = curr_res // 2
93
+ self.down.append(down)
94
+
95
+ def forward(self, x):
96
+ h = self.conv_in(x)
97
+ h_ = []
98
+ for i_level in range(self.num_resolutions):
99
+ for i_block in range(self.num_res_blocks):
100
+ h = self.down[i_level].block[i_block](h)
101
+ if len(self.down[i_level].attn) > 0:
102
+ h = self.down[i_level].attn[i_block](h)
103
+ if hasattr(self.down[i_level], "downsample"):
104
+ h_.append(h)
105
+ h = self.down[i_level].downsample(h)
106
+
107
+ return h, h_
108
+
109
+
110
+ class Decoder(nn.Module):
111
+ def __init__(
112
+ self,
113
+ hidden_size: int,
114
+ hidden_size_mult: Tuple[int] = (1, 2, 4, 4),
115
+ attn_resolutions: Tuple[int] = (16,),
116
+ conv_out: Module = "CasualConv3d",
117
+ attention: Module = "AttnBlock",
118
+ resnet_blocks: Tuple[Module] = (
119
+ "ResnetBlock3D",
120
+ "ResnetBlock3D",
121
+ "ResnetBlock3D",
122
+ "ResnetBlock3D",
123
+ ),
124
+ spatial_upsample: Tuple[Module] = (
125
+ "",
126
+ "SpatialUpsample2x",
127
+ "SpatialUpsample2x",
128
+ "SpatialUpsample2x",
129
+ ),
130
+ dropout: float = 0.0,
131
+ resolution: int = 256,
132
+ num_res_blocks: int = 2,
133
+ ):
134
+ super().__init__()
135
+ # ---- Config ----
136
+ self.num_resolutions = len(hidden_size_mult)
137
+ self.resolution = resolution
138
+ self.num_res_blocks = num_res_blocks
139
+
140
+ # ---- In ----
141
+ block_in = hidden_size * hidden_size_mult[self.num_resolutions - 1]
142
+ curr_res = resolution // 2 ** (self.num_resolutions - 1)
143
+
144
+ # ---- Upsample ----
145
+ self.up = nn.ModuleList()
146
+ for i_level in reversed(range(self.num_resolutions)):
147
+ block = nn.ModuleList()
148
+ attn = nn.ModuleList()
149
+ skip = nn.ModuleList()
150
+ block_out = hidden_size * hidden_size_mult[i_level]
151
+ for i_block in range(self.num_res_blocks):
152
+ block.append(
153
+ resolve_str_to_obj(resnet_blocks[i_level])(
154
+ in_channels=block_in,
155
+ out_channels=block_out,
156
+ dropout=dropout,
157
+ )
158
+ )
159
+ block_in = block_out
160
+ if curr_res in attn_resolutions:
161
+ attn.append(resolve_str_to_obj(attention)(block_in))
162
+ up = nn.Module()
163
+ up.block = block
164
+ up.attn = attn
165
+ up.skip = skip
166
+ if spatial_upsample[i_level]:
167
+ up.upsample = resolve_str_to_obj(spatial_upsample[i_level])(
168
+ block_in, block_in
169
+ )
170
+ up.skip = resolve_str_to_obj(conv_out)(block_in+hidden_size * hidden_size_mult[i_level-1],
171
+ block_in, kernel_size=3, padding=1)
172
+ curr_res = curr_res * 2
173
+ self.up.insert(0, up)
174
+
175
+ # ---- Out ----
176
+ self.norm_out = Normalize(block_in)
177
+ self.conv_out = resolve_str_to_obj(conv_out)(
178
+ block_in, 3, kernel_size=3, padding=1
179
+ )
180
+
181
+ def forward(self, h, h_):
182
+
183
+ for i_level in reversed(range(self.num_resolutions)):
184
+ for i_block in range(self.num_res_blocks):
185
+ h = self.up[i_level].block[i_block](h)
186
+ if len(self.up[i_level].attn) > 0:
187
+ h = self.up[i_level].attn[i_block](h)
188
+ if hasattr(self.up[i_level], "upsample"):
189
+ h = self.up[i_level].upsample(h)
190
+ h = torch.concat([h_[i_level-1], h], dim=1)
191
+ h = self.up[i_level].skip(h)
192
+
193
+ h = self.norm_out(h)
194
+ h = nonlinearity(h)
195
+ h = self.conv_out(h)
196
+ return h
197
+
198
+
199
+ class Refiner(VideoBaseAE):
200
+
201
+ @register_to_config
202
+ def __init__(
203
+ self,
204
+ hidden_size: int = 128,
205
+ hidden_size_mult: Tuple[int] = (1, 2, 4, 4),
206
+ attn_resolutions: Tuple[int] = [],
207
+ dropout: float = 0.0,
208
+ resolution: int = 256,
209
+ num_res_blocks: int = 2,
210
+ encoder_conv_in: Module = "CausalConv3d",
211
+ encoder_attention: Module = "AttnBlock3D",
212
+ encoder_resnet_blocks: Tuple[Module] = (
213
+ "ResnetBlock3D",
214
+ "ResnetBlock3D",
215
+ "ResnetBlock3D",
216
+ "ResnetBlock3D",
217
+ ),
218
+ encoder_spatial_downsample: Tuple[Module] = (
219
+ "SpatialDownsample2x",
220
+ "SpatialDownsample2x",
221
+ "SpatialDownsample2x",
222
+ "",
223
+ ),
224
+ decoder_conv_out: Module = "CausalConv3d",
225
+ decoder_attention: Module = "AttnBlock3D",
226
+ decoder_resnet_blocks: Tuple[Module] = (
227
+ "ResnetBlock3D",
228
+ "ResnetBlock3D",
229
+ "ResnetBlock3D",
230
+ "ResnetBlock3D",
231
+ ),
232
+ decoder_spatial_upsample: Tuple[Module] = (
233
+ "",
234
+ "SpatialUpsample2x",
235
+ "SpatialUpsample2x",
236
+ "SpatialUpsample2x",
237
+ ),
238
+ ) -> None:
239
+ super().__init__()
240
+
241
+ self.tile_sample_min_size = 256
242
+ self.tile_sample_min_size_t = 65
243
+ self.tile_latent_min_size = int(self.tile_sample_min_size / (2 ** (len(hidden_size_mult) - 1)))
244
+ self.tile_overlap_factor = 0.25
245
+ self.use_tiling = False
246
+
247
+ self.encoder = Encoder(
248
+ hidden_size=hidden_size,
249
+ hidden_size_mult=hidden_size_mult,
250
+ attn_resolutions=attn_resolutions,
251
+ conv_in=encoder_conv_in,
252
+ attention=encoder_attention,
253
+ resnet_blocks=encoder_resnet_blocks,
254
+ spatial_downsample=encoder_spatial_downsample,
255
+ dropout=dropout,
256
+ resolution=resolution,
257
+ num_res_blocks=num_res_blocks,
258
+ )
259
+
260
+ self.decoder = Decoder(
261
+ hidden_size=hidden_size,
262
+ hidden_size_mult=hidden_size_mult,
263
+ attn_resolutions=attn_resolutions,
264
+ conv_out=decoder_conv_out,
265
+ attention=decoder_attention,
266
+ resnet_blocks=decoder_resnet_blocks,
267
+ spatial_upsample=decoder_spatial_upsample,
268
+ dropout=dropout,
269
+ resolution=resolution,
270
+ num_res_blocks=num_res_blocks,
271
+ )
272
+
273
+ def get_encoder(self):
274
+ return [self.encoder]
275
+
276
+ def get_decoder(self):
277
+ return [self.decoder]
278
+
279
+ def encode(self, x):
280
+ if self.use_tiling and (
281
+ x.shape[-1] > self.tile_sample_min_size
282
+ or x.shape[-2] > self.tile_sample_min_size
283
+ or x.shape[-3] > self.tile_sample_min_size_t
284
+ ):
285
+ return self.tiled_encode(x)
286
+ enc = self.encoder(x)
287
+ return enc
288
+
289
+ def decode(self, z):
290
+ if self.use_tiling and (
291
+ z.shape[-1] > self.tile_latent_min_size
292
+ or z.shape[-2] > self.tile_latent_min_size
293
+ or z.shape[-3] > self.tile_latent_min_size_t
294
+ ):
295
+ return self.tiled_decode(z)
296
+ dec = self.decoder(z)
297
+ return dec
298
+
299
+ def forward(self, input):
300
+ enc, enc_ = self.encoder(input)
301
+ dec = self.decoder(enc, enc_)
302
+ return dec+input
303
+
304
+ def on_train_start(self):
305
+ self.ema = deepcopy(self) if self.save_ema==True else None
306
+
307
+ def get_last_layer(self):
308
+ if hasattr(self.decoder.conv_out, "conv"):
309
+ return self.decoder.conv_out.conv.weight
310
+ else:
311
+ return self.decoder.conv_out.weight
312
+
313
+ def blend_v(
314
+ self, a: torch.Tensor, b: torch.Tensor, blend_extent: int
315
+ ) -> torch.Tensor:
316
+ blend_extent = min(a.shape[3], b.shape[3], blend_extent)
317
+ for y in range(blend_extent):
318
+ b[:, :, :, y, :] = a[:, :, :, -blend_extent + y, :] * (
319
+ 1 - y / blend_extent
320
+ ) + b[:, :, :, y, :] * (y / blend_extent)
321
+ return b
322
+
323
+ def blend_h(
324
+ self, a: torch.Tensor, b: torch.Tensor, blend_extent: int
325
+ ) -> torch.Tensor:
326
+ blend_extent = min(a.shape[4], b.shape[4], blend_extent)
327
+ for x in range(blend_extent):
328
+ b[:, :, :, :, x] = a[:, :, :, :, -blend_extent + x] * (
329
+ 1 - x / blend_extent
330
+ ) + b[:, :, :, :, x] * (x / blend_extent)
331
+ return b
332
+
333
+ def tiled_encode(self, x):
334
+ t = x.shape[2]
335
+ t_chunk_idx = [i for i in range(0, t, self.tile_sample_min_size_t-1)]
336
+ if len(t_chunk_idx) == 1 and t_chunk_idx[0] == 0:
337
+ t_chunk_start_end = [[0, t]]
338
+ else:
339
+ t_chunk_start_end = [[t_chunk_idx[i], t_chunk_idx[i+1]+1] for i in range(len(t_chunk_idx)-1)]
340
+ if t_chunk_start_end[-1][-1] > t:
341
+ t_chunk_start_end[-1][-1] = t
342
+ elif t_chunk_start_end[-1][-1] < t:
343
+ last_start_end = [t_chunk_idx[-1], t]
344
+ t_chunk_start_end.append(last_start_end)
345
+ moments = []
346
+ for idx, (start, end) in enumerate(t_chunk_start_end):
347
+ chunk_x = x[:, :, start: end]
348
+ if idx != 0:
349
+ moment = self.tiled_encode2d(chunk_x, return_moments=True)[:, :, 1:]
350
+ else:
351
+ moment = self.tiled_encode2d(chunk_x, return_moments=True)
352
+ moments.append(moment)
353
+ moments = torch.cat(moments, dim=2)
354
+ return moments
355
+
356
+ def tiled_decode(self, x):
357
+ t = x.shape[2]
358
+ t_chunk_idx = [i for i in range(0, t, self.tile_latent_min_size_t-1)]
359
+ if len(t_chunk_idx) == 1 and t_chunk_idx[0] == 0:
360
+ t_chunk_start_end = [[0, t]]
361
+ else:
362
+ t_chunk_start_end = [[t_chunk_idx[i], t_chunk_idx[i+1]+1] for i in range(len(t_chunk_idx)-1)]
363
+ if t_chunk_start_end[-1][-1] > t:
364
+ t_chunk_start_end[-1][-1] = t
365
+ elif t_chunk_start_end[-1][-1] < t:
366
+ last_start_end = [t_chunk_idx[-1], t]
367
+ t_chunk_start_end.append(last_start_end)
368
+ dec_ = []
369
+ for idx, (start, end) in enumerate(t_chunk_start_end):
370
+ chunk_x = x[:, :, start: end]
371
+ if idx != 0:
372
+ dec = self.tiled_decode2d(chunk_x)[:, :, 1:]
373
+ else:
374
+ dec = self.tiled_decode2d(chunk_x)
375
+ dec_.append(dec)
376
+ dec_ = torch.cat(dec_, dim=2)
377
+ return dec_
378
+
379
+ def tiled_encode2d(self, x, return_moments=False):
380
+ overlap_size = int(self.tile_sample_min_size * (1 - self.tile_overlap_factor))
381
+ blend_extent = int(self.tile_latent_min_size * self.tile_overlap_factor)
382
+ row_limit = self.tile_latent_min_size - blend_extent
383
+
384
+ # Split the image into 512x512 tiles and encode them separately.
385
+ rows = []
386
+ for i in range(0, x.shape[3], overlap_size):
387
+ row = []
388
+ for j in range(0, x.shape[4], overlap_size):
389
+ tile = x[
390
+ :,
391
+ :,
392
+ :,
393
+ i : i + self.tile_sample_min_size,
394
+ j : j + self.tile_sample_min_size,
395
+ ]
396
+ tile = self.encoder(tile)
397
+ row.append(tile)
398
+ rows.append(row)
399
+ result_rows = []
400
+ for i, row in enumerate(rows):
401
+ result_row = []
402
+ for j, tile in enumerate(row):
403
+ # blend the above tile and the left tile
404
+ # to the current tile and add the current tile to the result row
405
+ if i > 0:
406
+ tile = self.blend_v(rows[i - 1][j], tile, blend_extent)
407
+ if j > 0:
408
+ tile = self.blend_h(row[j - 1], tile, blend_extent)
409
+ result_row.append(tile[:, :, :, :row_limit, :row_limit])
410
+ result_rows.append(torch.cat(result_row, dim=4))
411
+
412
+ moments = torch.cat(result_rows, dim=3)
413
+ posterior = DiagonalGaussianDistribution(moments)
414
+ return posterior
415
+
416
+ def tiled_decode2d(self, z):
417
+
418
+ overlap_size = int(self.tile_latent_min_size * (1 - self.tile_overlap_factor))
419
+ blend_extent = int(self.tile_sample_min_size * self.tile_overlap_factor)
420
+ row_limit = self.tile_sample_min_size - blend_extent
421
+
422
+ # Split z into overlapping 64x64 tiles and decode them separately.
423
+ # The tiles have an overlap to avoid seams between tiles.
424
+ rows = []
425
+ for i in range(0, z.shape[3], overlap_size):
426
+ row = []
427
+ for j in range(0, z.shape[4], overlap_size):
428
+ tile = z[
429
+ :,
430
+ :,
431
+ :,
432
+ i : i + self.tile_latent_min_size,
433
+ j : j + self.tile_latent_min_size,
434
+ ]
435
+ if self.use_quant_layer:
436
+ tile = self.post_quant_conv(tile)
437
+ decoded = self.decoder(tile)
438
+ row.append(decoded)
439
+ rows.append(row)
440
+ result_rows = []
441
+ for i, row in enumerate(rows):
442
+ result_row = []
443
+ for j, tile in enumerate(row):
444
+ # blend the above tile and the left tile
445
+ # to the current tile and add the current tile to the result row
446
+ if i > 0:
447
+ tile = self.blend_v(rows[i - 1][j], tile, blend_extent)
448
+ if j > 0:
449
+ tile = self.blend_h(row[j - 1], tile, blend_extent)
450
+ result_row.append(tile[:, :, :, :row_limit, :row_limit])
451
+ result_rows.append(torch.cat(result_row, dim=4))
452
+
453
+ dec = torch.cat(result_rows, dim=3)
454
+ return dec
455
+
456
+ def enable_tiling(self, use_tiling: bool = True):
457
+ self.use_tiling = use_tiling
458
+
459
+ def disable_tiling(self):
460
+ self.enable_tiling(False)
461
+
462
+ def init_from_ckpt(self, path, ignore_keys=list()):
463
+ sd = torch.load(path, map_location="cpu")
464
+ print("init from " + path)
465
+
466
+ if "ema_state_dict" in sd and len(sd['ema_state_dict']) > 0 and os.environ.get("NOT_USE_EMA_MODEL", 0) == 0:
467
+ print("Load from ema model!")
468
+ sd = sd["ema_state_dict"]
469
+ sd = {key.replace("module.", ""): value for key, value in sd.items()}
470
+ elif "state_dict" in sd:
471
+ print("Load from normal model!")
472
+ if "gen_model" in sd["state_dict"]:
473
+ sd = sd["state_dict"]["gen_model"]
474
+ else:
475
+ sd = sd["state_dict"]
476
+
477
+ keys = list(sd.keys())
478
+
479
+ for k in keys:
480
+ for ik in ignore_keys:
481
+ if k.startswith(ik):
482
+ print("Deleting key {} from state_dict.".format(k))
483
+ del sd[k]
484
+
485
+ self.load_state_dict(sd, strict=True)
486
+
487
+
488
+
489
+
490
+
causalvideovae/model/trainer_videobase.py ADDED
@@ -0,0 +1,26 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ from transformers import Trainer
2
+ import torch.nn.functional as F
3
+ from typing import Optional
4
+ import os
5
+ import torch
6
+ from transformers.utils import WEIGHTS_NAME
7
+ import json
8
+
9
+ class VideoBaseTrainer(Trainer):
10
+
11
+ def _save(self, output_dir: Optional[str] = None, state_dict=None):
12
+ output_dir = output_dir if output_dir is not None else self.args.output_dir
13
+ os.makedirs(output_dir, exist_ok=True)
14
+ if state_dict is None:
15
+ state_dict = self.model.state_dict()
16
+
17
+ # get model config
18
+ model_config = self.model.config.to_dict()
19
+
20
+ # add more information
21
+ model_config['model'] = self.model.__class__.__name__
22
+
23
+ with open(os.path.join(output_dir, "config.json"), "w") as file:
24
+ json.dump(self.model.config.to_dict(), file)
25
+ torch.save(state_dict, os.path.join(output_dir, WEIGHTS_NAME))
26
+ torch.save(self.args, os.path.join(output_dir, "training_args.bin"))
causalvideovae/model/utils/scheduler_utils.py ADDED
@@ -0,0 +1,7 @@
 
 
 
 
 
 
 
 
1
+ import torch
2
+
3
+ def cosine_scheduler(step, max_steps, value_base=1, value_end=0):
4
+ step = torch.tensor(step)
5
+ cosine_value = 0.5 * (1 + torch.cos(torch.pi * step / max_steps))
6
+ value = value_end + (value_base - value_end) * cosine_value
7
+ return value
causalvideovae/utils/downloader.py ADDED
@@ -0,0 +1,18 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ import gdown
2
+ import os
3
+
4
+ opensora_cache_home = os.path.expanduser(
5
+ os.getenv("OPENSORA_HOME", os.path.join("~/.cache", "opensora"))
6
+ )
7
+
8
+
9
+ def gdown_download(id, fname, cache_dir=None):
10
+ cache_dir = opensora_cache_home if not cache_dir else cache_dir
11
+
12
+ os.makedirs(cache_dir, exist_ok=True)
13
+ destination = os.path.join(cache_dir, fname)
14
+ if os.path.exists(destination):
15
+ return destination
16
+
17
+ gdown.download(id=id, output=destination, quiet=False)
18
+ return destination
causalvideovae/utils/taming_download.py ADDED
@@ -0,0 +1,145 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ """Modified from https://github.com/CompVis/taming-transformers.git"""
2
+
3
+ import os, hashlib
4
+ import requests
5
+ from tqdm import tqdm
6
+
7
+ URL_MAP = {
8
+ "vgg_lpips": "https://heibox.uni-heidelberg.de/f/607503859c864bc1b30b/?dl=1"
9
+ }
10
+
11
+ CKPT_MAP = {
12
+ "vgg_lpips": "vgg.pth"
13
+ }
14
+
15
+ MD5_MAP = {
16
+ "vgg_lpips": "d507d7349b931f0638a25a48a722f98a"
17
+ }
18
+
19
+
20
+ def download(url, local_path, chunk_size=1024):
21
+ os.makedirs(os.path.split(local_path)[0], exist_ok=True)
22
+ with requests.get(url, stream=True) as r:
23
+ total_size = int(r.headers.get("content-length", 0))
24
+ with tqdm(total=total_size, unit="B", unit_scale=True) as pbar:
25
+ with open(local_path, "wb") as f:
26
+ for data in r.iter_content(chunk_size=chunk_size):
27
+ if data:
28
+ f.write(data)
29
+ pbar.update(chunk_size)
30
+
31
+
32
+ def md5_hash(path):
33
+ with open(path, "rb") as f:
34
+ content = f.read()
35
+ return hashlib.md5(content).hexdigest()
36
+
37
+
38
+ def get_ckpt_path(name, root, check=False):
39
+ assert name in URL_MAP
40
+ path = os.path.join(root, CKPT_MAP[name])
41
+ if not os.path.exists(path) or (check and not md5_hash(path) == MD5_MAP[name]):
42
+ print("Downloading {} model from {} to {}".format(name, URL_MAP[name], path))
43
+ download(URL_MAP[name], path)
44
+ md5 = md5_hash(path)
45
+ assert md5 == MD5_MAP[name], md5
46
+ return path
47
+
48
+
49
+ class KeyNotFoundError(Exception):
50
+ def __init__(self, cause, keys=None, visited=None):
51
+ self.cause = cause
52
+ self.keys = keys
53
+ self.visited = visited
54
+ messages = list()
55
+ if keys is not None:
56
+ messages.append("Key not found: {}".format(keys))
57
+ if visited is not None:
58
+ messages.append("Visited: {}".format(visited))
59
+ messages.append("Cause:\n{}".format(cause))
60
+ message = "\n".join(messages)
61
+ super().__init__(message)
62
+
63
+
64
+ def retrieve(
65
+ list_or_dict, key, splitval="/", default=None, expand=True, pass_success=False
66
+ ):
67
+ """Given a nested list or dict return the desired value at key expanding
68
+ callable nodes if necessary and :attr:`expand` is ``True``. The expansion
69
+ is done in-place.
70
+
71
+ Parameters
72
+ ----------
73
+ list_or_dict : list or dict
74
+ Possibly nested list or dictionary.
75
+ key : str
76
+ key/to/value, path like string describing all keys necessary to
77
+ consider to get to the desired value. List indices can also be
78
+ passed here.
79
+ splitval : str
80
+ String that defines the delimiter between keys of the
81
+ different depth levels in `key`.
82
+ default : obj
83
+ Value returned if :attr:`key` is not found.
84
+ expand : bool
85
+ Whether to expand callable nodes on the path or not.
86
+
87
+ Returns
88
+ -------
89
+ The desired value or if :attr:`default` is not ``None`` and the
90
+ :attr:`key` is not found returns ``default``.
91
+
92
+ Raises
93
+ ------
94
+ Exception if ``key`` not in ``list_or_dict`` and :attr:`default` is
95
+ ``None``.
96
+ """
97
+
98
+ keys = key.split(splitval)
99
+
100
+ success = True
101
+ try:
102
+ visited = []
103
+ parent = None
104
+ last_key = None
105
+ for key in keys:
106
+ if callable(list_or_dict):
107
+ if not expand:
108
+ raise KeyNotFoundError(
109
+ ValueError(
110
+ "Trying to get past callable node with expand=False."
111
+ ),
112
+ keys=keys,
113
+ visited=visited,
114
+ )
115
+ list_or_dict = list_or_dict()
116
+ parent[last_key] = list_or_dict
117
+
118
+ last_key = key
119
+ parent = list_or_dict
120
+
121
+ try:
122
+ if isinstance(list_or_dict, dict):
123
+ list_or_dict = list_or_dict[key]
124
+ else:
125
+ list_or_dict = list_or_dict[int(key)]
126
+ except (KeyError, IndexError, ValueError) as e:
127
+ raise KeyNotFoundError(e, keys=keys, visited=visited)
128
+
129
+ visited += [key]
130
+ # final expansion of retrieved value
131
+ if expand and callable(list_or_dict):
132
+ list_or_dict = list_or_dict()
133
+ parent[last_key] = list_or_dict
134
+ except KeyNotFoundError as e:
135
+ if default is None:
136
+ raise e
137
+ else:
138
+ list_or_dict = default
139
+ success = False
140
+
141
+ if not pass_success:
142
+ return list_or_dict
143
+ else:
144
+ return list_or_dict, success
145
+
scripts/cal_expense.py ADDED
@@ -0,0 +1,347 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ import sys
2
+ sys.path.append(".")
3
+ from causalvideovae.model import CausalVAEModel
4
+ from CV_VAE.models.modeling_vae import CVVAEModel
5
+ from opensora.models.vae.vae import VideoAutoencoderPipeline
6
+ from opensora.registry import DATASETS, MODELS, build_module
7
+ from opensora.utils.config_utils import parse_configs
8
+ from diffusers.models import AutoencoderKL, AutoencoderKLTemporalDecoder
9
+ from tats import VQGAN
10
+ from tats.download import load_vqgan
11
+ from taming.models.vqgan import VQModel, GumbelVQ
12
+ import torch
13
+ from omegaconf import OmegaConf
14
+ import yaml
15
+ import argparse
16
+ from einops import rearrange
17
+ from causalvideovae.model.modules.normalize import Normalize
18
+ from causalvideovae.model.modules.block import Block
19
+ import time
20
+
21
+ def total_params(model):
22
+ total_params = sum(p.numel() for p in model.parameters())
23
+ total_params_in_millions = total_params / 1e6
24
+ return total_params_in_millions
25
+
26
+ device = torch.device('cuda')
27
+ data_type = torch.bfloat16
28
+ video_input = torch.randn(1, 3, 33, 256, 256).to(device).to(data_type)
29
+ image_input = torch.randn(33, 3, 256, 256).to(device).to(data_type)
30
+ num = 1000
31
+
32
+ """
33
+ #VQGAN
34
+ def load_config(config_path, display=False):
35
+ config = OmegaConf.load(config_path)
36
+ if display:
37
+ print(yaml.dump(OmegaConf.to_container(config)))
38
+ return config
39
+
40
+ def load_vqgan(config, ckpt_path=None, is_gumbel=False):
41
+ if is_gumbel:
42
+ model = GumbelVQ(**config.model.params)
43
+ else:
44
+ model = VQModel(**config.model.params)
45
+ if ckpt_path is not None:
46
+ sd = torch.load(ckpt_path, map_location="cpu")["state_dict"]
47
+ missing, unexpected = model.load_state_dict(sd, strict=False)
48
+ return model.eval()
49
+
50
+ vqgan_ckpt='/remote-home1/clh/taming-transformers/logs/vqgan_gumbel_f8/checkpoints/last.ckpt'
51
+ vqgan_config='/remote-home1/clh/taming-transformers/logs/vqgan_gumbel_f8/configs/model.yaml'
52
+ vqgan_config = load_config(vqgan_config, display=False)
53
+ vqgan = load_vqgan(vqgan_config, ckpt_path=vqgan_ckpt, is_gumbel=True).to(device).to(data_type).eval()
54
+ vqgan.requires_grad_(False)
55
+ print('VQGAN')
56
+ print(f"Generator:\t\t{total_params(vqgan) :.2f}M")
57
+ print(f"\t- Encoder:\t{total_params(vqgan.encoder) :.2f}M")
58
+ print(f"\t- Decoder:\t{total_params(vqgan.decoder) :.2f}M")
59
+ # 计算程序运行时间
60
+ start_time = time.time()
61
+ for i in range(num):
62
+ latents, _, [_, _, indices] = vqgan.encode(image_input)
63
+ end_time = time.time()
64
+ print(f"encode_time:{(end_time - start_time)/num :.3f}s")
65
+
66
+ start_time = time.time()
67
+ for i in range(num):
68
+ video_recon = vqgan.decode(latents.to(data_type))
69
+ end_time = time.time()
70
+ print(f"decode_time:{(end_time - start_time)/num :.3f}s")
71
+
72
+ start_time = time.time()
73
+ for i in range(num):
74
+ latents, _, [_, _, indices] = vqgan.encode(image_input)
75
+ video_recon = vqgan.decode(latents.to(data_type))
76
+ end_time = time.time()
77
+ print(f"rec_time:{(end_time - start_time)/num :.3f}s")
78
+
79
+
80
+
81
+ #TATS
82
+ tats_path = '/remote-home1/clh/TATS/vqgan_sky_128_488_epoch_12-step_29999-train.ckpt'
83
+ tats = VQGAN.load_from_checkpoint(tats_path).to(device).to(torch.float32).eval()
84
+ tats.requires_grad_(False)
85
+ print('TATS')
86
+ print(f"Generator:\t\t{total_params(tats) :.2f}M")
87
+ print(f"\t- Encoder:\t{total_params(tats.encoder):.2f}M")
88
+ print(f"\t- Decoder:\t{total_params(tats.decoder):.2f}M")
89
+ # 计算程序运行时间
90
+ start_time = time.time()
91
+ for i in range(num):
92
+ z = tats.pre_vq_conv(tats.encoder(video_input.to(torch.float32)))
93
+ vq_output = tats.codebook(z)
94
+ latents = vq_output['embeddings']
95
+ end_time = time.time()
96
+ print(f"encode_time:{(end_time - start_time)/num :.3f}s")
97
+
98
+ start_time = time.time()
99
+ for i in range(num):
100
+ video_recon = tats.decoder(tats.post_vq_conv(latents.to(torch.float32)))
101
+ end_time = time.time()
102
+ print(f"decode_time:{(end_time - start_time)/num :.3f}s")
103
+
104
+ start_time = time.time()
105
+ for i in range(num):
106
+ z = tats.pre_vq_conv(tats.encoder(video_input.to(torch.float32)))
107
+ vq_output = tats.codebook(z)
108
+ latents = vq_output['embeddings']
109
+ video_recon = tats.decoder(tats.post_vq_conv(latents.to(torch.float32)))
110
+ end_time = time.time()
111
+ print(f"rec_time:{(end_time - start_time)/num :.3f}s")
112
+
113
+
114
+ #SD2_1
115
+ sd2_1_path = '/remote-home1/clh/sd2_1'
116
+ sd2_1 = AutoencoderKL.from_pretrained(sd2_1_path).eval().to(device).to(data_type)
117
+ sd2_1.requires_grad_(False)
118
+ print('SD2_1')
119
+ print(f"Generator:\t\t{total_params(sd2_1) :.2f}M")
120
+ print(f"\t- Encoder:\t{total_params(sd2_1.encoder):.2f}M")
121
+ print(f"\t- Decoder:\t{total_params(sd2_1.decoder):.2f}M")
122
+ # 计算程序运行时间
123
+ start_time = time.time()
124
+ for i in range(num):
125
+ latents = sd2_1.encode(image_input)['latent_dist'].sample()
126
+ end_time = time.time()
127
+ print(f"encode_time:{(end_time - start_time)/num :.3f}s")
128
+
129
+ start_time = time.time()
130
+ for i in range(num):
131
+ video_recon = sd2_1.decode(latents.to(data_type))['sample']
132
+ end_time = time.time()
133
+ print(f"decode_time:{(end_time - start_time)/num :.3f}s")
134
+
135
+ start_time = time.time()
136
+ for i in range(num):
137
+ latents = sd2_1.encode(image_input)['latent_dist'].sample()
138
+ video_recon = sd2_1.decode(latents.to(data_type))['sample']
139
+ end_time = time.time()
140
+ print(f"rec_time:{(end_time - start_time)/num :.3f}s")
141
+
142
+ #SVD
143
+ svd_path = '/remote-home1/clh/svd/'
144
+ svd = AutoencoderKLTemporalDecoder.from_pretrained(svd_path).eval().to(device).to(data_type)
145
+ svd.requires_grad_(False)
146
+ print('SVD')
147
+ print(f"Generator:\t\t{total_params(svd):.2f}M")
148
+ print(f"\t- Encoder:\t{total_params(svd.encoder):.2f}M")
149
+ print(f"\t- Decoder:\t{total_params(svd.decoder):.2f}M")
150
+ # 计算程序运行时间
151
+ start_time = time.time()
152
+ for i in range(num):
153
+ latents = svd.encode(image_input)['latent_dist'].sample()
154
+ end_time = time.time()
155
+ print(f"encode_time:{(end_time - start_time)/num :.3f}s")
156
+
157
+ start_time = time.time()
158
+ for i in range(num):
159
+ video_recon = svd.decode(latents.to(data_type), num_frames=video_input.shape[2])['sample']
160
+ end_time = time.time()
161
+ print(f"decode_time:{(end_time - start_time)/num :.3f}s")
162
+
163
+ start_time = time.time()
164
+ for i in range(num):
165
+ latents = svd.encode(image_input)['latent_dist'].sample()
166
+ video_recon = svd.decode(latents.to(data_type), num_frames=video_input.shape[2])['sample']
167
+ end_time = time.time()
168
+ print(f"rec_time:{(end_time - start_time)/num :.3f}s")
169
+
170
+ #CV-VAE
171
+ cvvae_path = '/remote-home1/clh/CV-VAE/vae3d'
172
+ cvvae = CVVAEModel.from_pretrained(cvvae_path).eval().to(device).to(data_type)
173
+ cvvae.requires_grad_(False)
174
+ print('CV-VAE')
175
+ print(f"Generator:\t\t{total_params(cvvae):.2f}M")
176
+ print(f"\t- Encoder:\t{total_params(cvvae.encoder):.2f}M")
177
+ print(f"\t- Decoder:\t{total_params(cvvae.decoder):.2f}M")
178
+ # 计算程序运行时间
179
+ start_time = time.time()
180
+ for i in range(num):
181
+ latent = cvvae.encode(video_input).latent_dist.sample()
182
+ end_time = time.time()
183
+ print(f"encode_time:{(end_time - start_time)/num :.3f}s")
184
+
185
+ start_time = time.time()
186
+ for i in range(num):
187
+ video_recon = cvvae.decode(latent).sample
188
+ end_time = time.time()
189
+ print(f"decode_time:{(end_time - start_time)/num :.3f}s")
190
+
191
+ start_time = time.time()
192
+ for i in range(num):
193
+ latent = cvvae.encode(video_input).latent_dist.sample()
194
+ video_recon = cvvae.decode(latent).sample
195
+ end_time = time.time()
196
+ print(f"rec_time:{(end_time - start_time)/num :.3f}s")
197
+
198
+
199
+ #NUS-VAE
200
+ nusvae_path = '/remote-home1/clh/CV-VAE/vae3d'
201
+ parser = argparse.ArgumentParser()
202
+ parser.add_argument("--config", type=str, default="/remote-home1/clh/Causal-Video-VAE/opensora/video.py")
203
+ parser.add_argument("--ckpt", type=str, default="/remote-home1/clh/Open-Sora/OpenSora-VAE-v1.2")
204
+ args = parser.parse_args()
205
+ cfg = parse_configs(args, training=False)
206
+ nusvae = build_module(cfg.model, MODELS).eval().to(device).to(data_type)
207
+ nusvae.requires_grad_(False)
208
+ print('NUS-VAE')
209
+ print(f"Generator:\t\t{total_params(nusvae):.2f}M")
210
+ print(f"\t- Spatial_Encoder:\t{total_params(nusvae.spatial_vae.module.encoder):.2f}M")
211
+ print(f"\t- Temporal_Encoder:\t{total_params(nusvae.temporal_vae.encoder):.2f}M")
212
+ print(f"\t- Temporal_Decoder:\t{total_params(nusvae.temporal_vae.decoder):.2f}M")
213
+ print(f"\t- Spatial_Decoder:\t{total_params(nusvae.spatial_vae.module.decoder):.2f}M")
214
+ # 计算程序运行时间
215
+ start_time = time.time()
216
+ for i in range(num):
217
+ latents, posterior, x_z = nusvae.encode(video_input)
218
+ end_time = time.time()
219
+ print(f"encode_time:{(end_time - start_time)/num :.3f}s")
220
+
221
+ start_time = time.time()
222
+ for i in range(num):
223
+ video_recon, x_z_rec = nusvae.decode(latents, num_frames=video_input.size(2))
224
+ end_time = time.time()
225
+ print(f"decode_time:{(end_time - start_time)/num :.3f}s")
226
+
227
+ start_time = time.time()
228
+ for i in range(num):
229
+ latents, posterior, x_z = nusvae.encode(video_input)
230
+ video_recon, x_z_rec = nusvae.decode(latents, num_frames=video_input.size(2))
231
+ end_time = time.time()
232
+ print(f"rec_time:{(end_time - start_time)/num :.3f}s")
233
+ """
234
+
235
+ #ours1.2
236
+ ours1_2_vae_path = '/remote-home1/clh/models/23denc_3ddec_vae_pretrained_weight'
237
+ ours1_2_vae = CausalVAEModel.from_pretrained(ours1_2_vae_path).eval().to(device).to(data_type)
238
+ ours1_2_vae.requires_grad_(False)
239
+ print('open_sora_plan_vae_1_2')
240
+ print(f"Generator:\t\t{total_params(ours1_2_vae):.2f}M")
241
+ print(f"\t- Encoder:\t{total_params(ours1_2_vae.encoder):.2f}M")
242
+ print(f"\t- Decoder:\t{total_params(ours1_2_vae.decoder):.2f}M")
243
+ # 计算程序运行时间
244
+ start_time = time.time()
245
+ for i in range(num):
246
+ latents = ours1_2_vae.encode(video_input).sample().to(data_type)
247
+ end_time = time.time()
248
+ print(f"encode_time:{(end_time - start_time)/num :.3f}s")
249
+
250
+ start_time = time.time()
251
+ for i in range(num):
252
+ video_recon = ours1_2_vae.decode(latents)
253
+ end_time = time.time()
254
+ print(f"decode_time:{(end_time - start_time)/num :.3f}s")
255
+
256
+ start_time = time.time()
257
+ for i in range(num):
258
+ latents = ours1_2_vae.encode(video_input).sample().to(data_type)
259
+ video_recon = ours1_2_vae.decode(latents)
260
+ end_time = time.time()
261
+ print(f"rec_time:{(end_time - start_time)/num :.3f}s")
262
+
263
+
264
+ #23d
265
+ half3d_vae_path = '/remote-home1/clh/models/23d_vae_pretrained_weight'
266
+ half3d_vae = CausalVAEModel.from_pretrained(half3d_vae_path).eval().to(device).to(data_type)
267
+ half3d_vae.requires_grad_(False)
268
+ print('open_sora_plan_vae_half3d')
269
+ print(f"Generator:\t\t{total_params(half3d_vae):.2f}M")
270
+ print(f"\t- Encoder:\t{total_params(half3d_vae.encoder):.2f}M")
271
+ print(f"\t- Decoder:\t{total_params(half3d_vae.decoder):.2f}M")
272
+ # 计算程序运行时间
273
+ start_time = time.time()
274
+ for i in range(num):
275
+ latents = half3d_vae.encode(video_input).sample().to(data_type)
276
+ end_time = time.time()
277
+ print(f"encode_time:{(end_time - start_time)/num :.3f}s")
278
+
279
+ start_time = time.time()
280
+ for i in range(num):
281
+ video_recon = half3d_vae.decode(latents)
282
+ end_time = time.time()
283
+ print(f"decode_time:{(end_time - start_time)/num :.3f}s")
284
+
285
+ start_time = time.time()
286
+ for i in range(num):
287
+ latents = half3d_vae.encode(video_input).sample().to(data_type)
288
+ video_recon = half3d_vae.decode(latents)
289
+ end_time = time.time()
290
+ print(f"rec_time:{(end_time - start_time)/num :.3f}s")
291
+
292
+ #2and3d
293
+ mix23d_vae_path = '/remote-home1/clh/models/mix23d_vae_pretrained_weight'
294
+ mix23d_vae = CausalVAEModel.from_pretrained(mix23d_vae_path).eval().to(device).to(data_type)
295
+ mix23d_vae.requires_grad_(False)
296
+ print('open_sora_plan_vae_mix23d')
297
+ print(f"Generator:\t\t{total_params(mix23d_vae):.2f}M")
298
+ print(f"\t- Encoder:\t{total_params(mix23d_vae.encoder):.2f}M")
299
+ print(f"\t- Decoder:\t{total_params(mix23d_vae.decoder):.2f}M")
300
+ # 计算程序运行时间
301
+ start_time = time.time()
302
+ for i in range(num):
303
+ latents = mix23d_vae.encode(video_input).sample().to(data_type)
304
+ end_time = time.time()
305
+ print(f"encode_time:{(end_time - start_time)/num :.3f}s")
306
+
307
+ start_time = time.time()
308
+ for i in range(num):
309
+ video_recon = mix23d_vae.decode(latents)
310
+ end_time = time.time()
311
+ print(f"decode_time:{(end_time - start_time)/num :.3f}s")
312
+
313
+ start_time = time.time()
314
+ for i in range(num):
315
+ latents = mix23d_vae.encode(video_input).sample().to(data_type)
316
+ video_recon = mix23d_vae.decode(latents)
317
+ end_time = time.time()
318
+ print(f"rec_time:{(end_time - start_time)/num :.3f}s")
319
+
320
+ #full 3d
321
+ full3d_vae_path = '/remote-home1/clh/models/full3d_vae_pretrained_weight'
322
+ full3d_vae = CausalVAEModel.from_pretrained(full3d_vae_path).eval().to(device).to(data_type)
323
+ full3d_vae.requires_grad_(False)
324
+ print('open_sora_plan_vae_full3d')
325
+ print(f"Generator:\t\t{total_params(full3d_vae):.2f}M")
326
+ print(f"\t- Encoder:\t{total_params(full3d_vae.encoder):.2f}M")
327
+ print(f"\t- Decoder:\t{total_params(full3d_vae.decoder):.2f}M")
328
+ # 计算程序运行时间
329
+ start_time = time.time()
330
+ for i in range(num):
331
+ latents = full3d_vae.encode(video_input).sample().to(data_type)
332
+ end_time = time.time()
333
+ print(f"encode_time:{(end_time - start_time)/num :.3f}s")
334
+
335
+ start_time = time.time()
336
+ for i in range(num):
337
+ video_recon = full3d_vae.decode(latents)
338
+ end_time = time.time()
339
+ print(f"decode_time:{(end_time - start_time)/num :.3f}s")
340
+
341
+ start_time = time.time()
342
+ for i in range(num):
343
+ latents = full3d_vae.encode(video_input).sample().to(data_type)
344
+ video_recon = full3d_vae.decode(latents)
345
+ end_time = time.time()
346
+ print(f"rec_time:{(end_time - start_time)/num :.3f}s")
347
+
scripts/config_deepspeed.yaml ADDED
@@ -0,0 +1,20 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ compute_environment: LOCAL_MACHINE
2
+ debug: false
3
+ distributed_type: MULTI_GPU
4
+ # fsdp_config:
5
+ # fsdp_auto_wrap_policy: NO_WRAP
6
+ # fsdp_backward_prefetch_policy: BACKWARD_PRE
7
+ # fsdp_forward_prefetch: false
8
+ # fsdp_cpu_ram_efficient_loading: true
9
+ # fsdp_offload_params: false
10
+ # fsdp_sharding_strategy: FULL_SHARD
11
+ # fsdp_state_dict_type: SHARDED_STATE_DICT
12
+ # fsdp_sync_module_states: true
13
+ # fsdp_use_orig_params: true
14
+ downcast_bf16: 'no'
15
+ enable_cpu_affinity: false
16
+ gpu_ids: 1,2,3,4,5,6,7
17
+ machine_rank: 0
18
+ mixed_precision: bf16
19
+ num_machines: 1
20
+ num_processes: 7
scripts/convert_weight.py ADDED
@@ -0,0 +1,74 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ import torch
2
+ import sys
3
+ import re
4
+
5
+ import safetensors
6
+ sys.path.append(".")
7
+ from causalvideovae.model import CausalVAEModel
8
+
9
+ origin_ckpt_path = "/remote-home1/clh/models/sd2_1/vae-ft-mse-840000-ema-pruned.ckpt"
10
+ config_path = "/remote-home1/clh/models/sd2_1/config.json"
11
+ output_path = "/remote-home1/clh/models/norm3d_vae_pretrained_weight"
12
+ init_method = "tail"
13
+
14
+ model = CausalVAEModel.from_config(config_path)
15
+
16
+ if origin_ckpt_path.endswith('ckpt'):
17
+ ckpt = torch.load(origin_ckpt_path, map_location="cpu")['state_dict']
18
+ elif origin_ckpt_path.endswith('safetensors'):
19
+ ckpt = {}
20
+ with safetensors.safe_open(origin_ckpt_path, framework="pt") as file:
21
+ for k in file.keys():
22
+ ckpt[k] = file.get_tensor(k)
23
+ print("key", k)
24
+
25
+ for name, module in model.named_modules():
26
+ if "loss" in name:
27
+ continue
28
+
29
+ if isinstance(module, torch.nn.Conv3d):
30
+ in_channels = module.in_channels
31
+ out_channels = module.out_channels
32
+ kernel_size = module.kernel_size
33
+ old_name = re.sub(".conv$", "", name)
34
+ if old_name + ".weight" not in ckpt:
35
+ print(old_name + ".weight", "not found")
36
+ continue
37
+ if init_method == "tail":
38
+ shape_2d = ckpt[old_name + ".weight"].shape
39
+ new_weight = torch.zeros(*shape_2d)
40
+ new_weight = new_weight.unsqueeze(2).repeat(1, 1, kernel_size[0], 1, 1)
41
+ middle_idx = kernel_size[0] // 2
42
+ new_weight[:, :, -1, :, :] = ckpt[old_name + ".weight"]
43
+ new_bias = ckpt[old_name + ".bias"]
44
+ elif init_method == "avg":
45
+ new_weight = ckpt[old_name + ".weight"].unsqueeze(2)
46
+ new_weight = new_weight.repeat(1, 1, kernel_size[0], 1, 1) / kernel_size[0]
47
+ new_bias = ckpt[old_name + ".bias"]
48
+ assert new_weight.shape == module.weight.shape
49
+ module.weight.data = new_weight.cpu().float()
50
+ module.bias.data = new_bias.cpu().float()
51
+ elif isinstance(module, torch.nn.GroupNorm):
52
+ old_name = name
53
+ if old_name + ".weight" not in ckpt:
54
+ print(old_name + ".weight", "not found")
55
+ continue
56
+ new_weight = ckpt[old_name + ".weight"]
57
+ new_bias = ckpt[old_name + ".bias"]
58
+ module.weight.data = new_weight.cpu().float()
59
+ module.bias.data = new_bias.cpu().float()
60
+ elif isinstance(module, torch.nn.Conv2d):
61
+ in_channels = module.in_channels
62
+ out_channels = module.out_channels
63
+ kernel_size = module.kernel_size
64
+ old_name = name
65
+ if old_name + ".weight" not in ckpt:
66
+ print(old_name + ".weight", "not found")
67
+ continue
68
+ new_weight = ckpt[old_name + ".weight"]
69
+ new_bias = ckpt[old_name + ".bias"]
70
+ assert new_weight.shape == module.weight.shape
71
+ module.weight.data = new_weight.cpu().float()
72
+ module.bias.data = new_bias.cpu().float()
73
+
74
+ model.save_pretrained(output_path)
scripts/cv_vae_gen_video.sh ADDED
@@ -0,0 +1,30 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ export CUDA_VISIBLE_DEVICES=0,1,2,3,4,5,6,7
2
+ export NCCL_DEBUG=INFO
3
+ export NCCL_SOCKET_IFNAME=ibs11
4
+ export NCCL_IB_DISABLE=1
5
+ REAL_DATASET_DIR=/remote-home1/clh/dataset/panda70m_val
6
+ EXP_NAME=test_train
7
+ SAMPLE_RATE=1
8
+ NUM_FRAMES=33
9
+ RESOLUTION=256
10
+ SUBSET_SIZE=100
11
+ CKPT=/remote-home1/clh/CV-VAE/vae3d
12
+
13
+ torchrun \
14
+ --nnodes=1 --nproc_per_node=8 \
15
+ --rdzv_backend=c10d \
16
+ --rdzv_endpoint=localhost:29503 \
17
+ --master_addr=localhost \
18
+ --master_port=29600 \
19
+ scripts/rec_cv_vae.py \
20
+ --batch_size 8 \
21
+ --real_video_dir ${REAL_DATASET_DIR} \
22
+ --generated_video_dir /remote-home1/clh/gen/cvvae/panda70m \
23
+ --sample_fps 24 \
24
+ --sample_rate ${SAMPLE_RATE} \
25
+ --num_frames ${NUM_FRAMES} \
26
+ --resolution ${RESOLUTION} \
27
+ --crop_size ${RESOLUTION} \
28
+ --num_workers 8 \
29
+ --ckpt ${CKPT} \
30
+ --output_origin \
scripts/expand_weight.py ADDED
@@ -0,0 +1,40 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ import torch
2
+ import sys
3
+ import sys
4
+ sys.path.append(".")
5
+ from causalvideovae.model.causal_vae.modeling_causalvae import CausalVAEModel
6
+ from causalvideovae.model.modules import *
7
+
8
+ origin_path = "/remote-home1/lzj/causal-video-vae-github/results/test"
9
+ output_path = "models/latent8_3d"
10
+
11
+ print("Loading model!")
12
+ model = CausalVAEModel.from_pretrained(origin_path)
13
+ new_config = model.config.copy()
14
+ new_config['z_channels'] = 8
15
+ new_config['embed_dim'] = 8
16
+ reset_mix_factor = True
17
+ print("Building new model")
18
+ new_model = CausalVAEModel.from_config(new_config)
19
+
20
+ ckpt = new_model.state_dict()
21
+ old_ckpt = model.state_dict()
22
+
23
+ for name, parameter in new_model.named_parameters():
24
+ if name not in old_ckpt:
25
+ # ckpt[name] = torch.zeros_like(ckpt[name])
26
+ continue
27
+ shape1 = ckpt[name].shape
28
+ if sum(shape1) == 1:
29
+ if reset_mix_factor:
30
+ ckpt[name] = torch.tensor([0.])
31
+ continue
32
+ shape2 = old_ckpt[name].shape
33
+ slices = tuple(slice(0, s) for s in shape2)
34
+ mu = torch.mean(old_ckpt[name])
35
+ std = torch.std(old_ckpt[name])
36
+ ckpt[name] = torch.empty_like(ckpt[name]).normal_(mean=mu, std=std)
37
+ ckpt[name][slices] = old_ckpt[name]
38
+
39
+ new_model.load_state_dict(ckpt)
40
+ new_model.save_pretrained(output_path)
scripts/rec_TATS_vae.py ADDED
@@ -0,0 +1,310 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ import random
2
+ import argparse
3
+ import cv2
4
+ from tqdm import tqdm
5
+ import numpy as np
6
+ import numpy.typing as npt
7
+ import torch
8
+ import torch.distributed as dist
9
+ from torch.nn.parallel import DistributedDataParallel as DDP
10
+ from torch.utils.data import DataLoader, DistributedSampler, Subset
11
+ from decord import VideoReader, cpu
12
+ from torch.nn import functional as F
13
+ from pytorchvideo.transforms import ShortSideScale
14
+ from torchvision.transforms import Lambda, Compose
15
+ from torchvision.transforms._transforms_video import CenterCropVideo
16
+ import sys
17
+ from torch.utils.data import Dataset, DataLoader, Subset
18
+ import os
19
+ import glob
20
+ sys.path.append(".")
21
+ import torch.nn as nn
22
+ import yaml
23
+ from omegaconf import OmegaConf
24
+ from einops import rearrange
25
+ from tats import VQGAN
26
+ from tats.download import load_vqgan
27
+
28
+ def ddp_setup():
29
+ dist.init_process_group(backend="nccl")
30
+ torch.cuda.set_device(int(os.environ["LOCAL_RANK"]))
31
+
32
+ def array_to_video(
33
+ image_array: npt.NDArray, fps: float = 30.0, output_file: str = "output_video.mp4"
34
+ ) -> None:
35
+ height, width, channels = image_array[0].shape
36
+ fourcc = cv2.VideoWriter_fourcc(*"mp4v")
37
+ video_writer = cv2.VideoWriter(output_file, fourcc, float(fps), (width, height))
38
+
39
+ for image in image_array:
40
+ image_rgb = cv2.cvtColor(image, cv2.COLOR_RGB2BGR)
41
+ video_writer.write(image_rgb)
42
+
43
+ video_writer.release()
44
+
45
+
46
+ def custom_to_video(
47
+ x: torch.Tensor, fps: float = 2.0, output_file: str = "output_video.mp4"
48
+ ) -> None:
49
+ x = x.detach().cpu()
50
+ x = torch.clamp(x, -1, 1)
51
+ x = (x + 1) / 2
52
+ x = x.permute(1, 2, 3, 0).float().numpy()
53
+ x = (255 * x).astype(np.uint8)
54
+ array_to_video(x, fps=fps, output_file=output_file)
55
+ return
56
+
57
+
58
+ def read_video(video_path: str, num_frames: int, sample_rate: int) -> torch.Tensor:
59
+ decord_vr = VideoReader(video_path, ctx=cpu(0), num_threads=8)
60
+ total_frames = len(decord_vr)
61
+ sample_frames_len = sample_rate * num_frames
62
+
63
+ if total_frames > sample_frames_len:
64
+ s = 0
65
+ e = s + sample_frames_len
66
+ num_frames = num_frames
67
+ else:
68
+ s = 0
69
+ e = total_frames
70
+ num_frames = int(total_frames / sample_frames_len * num_frames)
71
+ print(
72
+ f"sample_frames_len {sample_frames_len}, only can sample {num_frames * sample_rate}",
73
+ video_path,
74
+ total_frames,
75
+ )
76
+
77
+ frame_id_list = np.linspace(s, e - 1, num_frames, dtype=int)
78
+ video_data = decord_vr.get_batch(frame_id_list).asnumpy()
79
+ video_data = torch.from_numpy(video_data)
80
+ video_data = video_data.permute(3, 0, 1, 2) # (T, H, W, C) -> (C, T, H, W)
81
+ return video_data
82
+
83
+
84
+ class RealVideoDataset(Dataset):
85
+ video_exts = ['avi', 'mp4', 'webm']
86
+
87
+ def __init__(
88
+ self,
89
+ real_video_dir,
90
+ num_frames,
91
+ sample_rate=1,
92
+ crop_size=None,
93
+ resolution=128,
94
+ ) -> None:
95
+ super().__init__()
96
+ self.real_video_files = self._combine_without_prefix(real_video_dir)
97
+ self.num_frames = num_frames
98
+ self.sample_rate = sample_rate
99
+ self.crop_size = crop_size
100
+ self.short_size = resolution
101
+
102
+ def __len__(self):
103
+ return len(self.real_video_files)
104
+
105
+ def __getitem__(self, index):
106
+ try:
107
+ if index >= len(self):
108
+ raise IndexError
109
+ real_video_file = self.real_video_files[index]
110
+ real_video_tensor = self._load_video(real_video_file)
111
+ video_name = os.path.basename(real_video_file)
112
+ except:
113
+ if index >= len(self):
114
+ raise IndexError
115
+ real_video_file = self.real_video_files[random.randint(1,index-1)]
116
+ real_video_tensor = self._load_video(real_video_file)
117
+ video_name = os.path.basename(real_video_file)
118
+ return {'video': real_video_tensor, 'file_name': video_name }
119
+
120
+ def _load_video(self, video_path):
121
+ num_frames = self.num_frames
122
+ sample_rate = self.sample_rate
123
+ decord_vr = VideoReader(video_path, ctx=cpu(0))
124
+ total_frames = len(decord_vr)
125
+ sample_frames_len = sample_rate * num_frames
126
+ s = 0
127
+ e = s + sample_frames_len
128
+ num_frames = num_frames
129
+ """
130
+ if total_frames > sample_frames_len:
131
+ s = 0
132
+ e = s + sample_frames_len
133
+ num_frames = num_frames
134
+
135
+ else:
136
+ s = 0
137
+ e = total_frames
138
+ num_frames = int(total_frames / sample_frames_len * num_frames)
139
+ print(
140
+ f"sample_frames_len {sample_frames_len}, only can sample {num_frames * sample_rate}",
141
+ video_path,
142
+ total_frames,
143
+ )
144
+ """
145
+ frame_id_list = np.linspace(s, e - 1, num_frames, dtype=int)
146
+ video_data = decord_vr.get_batch(frame_id_list).asnumpy()
147
+ video_data = torch.from_numpy(video_data)
148
+ video_data = video_data.permute(3, 0, 1, 2)
149
+ return _preprocess(
150
+ video_data, short_size=self.short_size, crop_size=self.crop_size
151
+ )
152
+
153
+ def _combine_without_prefix(self, folder_path):
154
+ samples = []
155
+ samples += sum([glob.glob(os.path.join(folder_path, '**', f'*.{ext}'), recursive=True)
156
+ for ext in self.video_exts], [])
157
+ samples.sort()
158
+ return samples
159
+
160
+ def resize(x, resolution):
161
+ height, width = x.shape[-2:]
162
+ aspect_ratio = width / height
163
+ if width <= height:
164
+ new_width = resolution
165
+ new_height = int(resolution / aspect_ratio)
166
+ else:
167
+ new_height = resolution
168
+ new_width = int(resolution * aspect_ratio)
169
+ resized_x = F.interpolate(x, size=(new_height, new_width), mode='bilinear', align_corners=True, antialias=True)
170
+ return resized_x
171
+
172
+ def _preprocess(video_data, short_size=128, crop_size=None):
173
+ transform = Compose(
174
+
175
+ [
176
+
177
+ Lambda(lambda x: ((x / 255.0) * 2 - 1)),
178
+ Lambda(lambda x: resize(x, short_size)),
179
+ (
180
+ CenterCropVideo(crop_size=crop_size)
181
+ if crop_size is not None
182
+ else Lambda(lambda x: x)
183
+ ),
184
+
185
+ ]
186
+
187
+ )
188
+ video_outputs = transform(video_data)
189
+ video_outputs = _format_video_shape(video_outputs)
190
+ return video_outputs
191
+
192
+
193
+ def _format_video_shape(video, time_compress=4, spatial_compress=8):
194
+ time = video.shape[1]
195
+ height = video.shape[2]
196
+ width = video.shape[3]
197
+ new_time = (
198
+ (time - (time - 1) % time_compress)
199
+ if (time - 1) % time_compress != 0
200
+ else time
201
+ )
202
+ new_height = (
203
+ (height - (height) % spatial_compress)
204
+ if height % spatial_compress != 0
205
+ else height
206
+ )
207
+ new_width = (
208
+ (width - (width) % spatial_compress) if width % spatial_compress != 0 else width
209
+ )
210
+ return video[:, :new_time, :new_height, :new_width]
211
+
212
+ def load_config(config_path, display=False):
213
+ config = OmegaConf.load(config_path)
214
+ if display:
215
+ print(yaml.dump(OmegaConf.to_container(config)))
216
+ return config
217
+
218
+ @torch.no_grad()
219
+ def main(args: argparse.Namespace):
220
+ real_video_dir = args.real_video_dir
221
+ generated_video_dir = args.generated_video_dir
222
+ ckpt = args.ckpt
223
+ sample_rate = args.sample_rate
224
+ resolution = args.resolution
225
+ crop_size = args.crop_size
226
+ num_frames = args.num_frames
227
+ sample_rate = args.sample_rate
228
+ sample_fps = args.sample_fps
229
+ batch_size = args.batch_size
230
+ num_workers = args.num_workers
231
+ subset_size = args.subset_size
232
+
233
+ if not os.path.exists(args.generated_video_dir):
234
+ os.makedirs(os.path.join(generated_video_dir, "vae_gen/"), exist_ok=True)
235
+
236
+ data_type = torch.float32
237
+
238
+ ddp_setup()
239
+ rank = int(os.environ["LOCAL_RANK"])
240
+
241
+ # ---- Load Model ----
242
+ vqgan = VQGAN.load_from_checkpoint(ckpt)
243
+ print(vqgan)
244
+ vqgan = vqgan.to(rank).to(data_type)
245
+ vqgan.eval()
246
+
247
+ # ---- Load Model ----
248
+
249
+ # ---- Prepare Dataset ----
250
+ dataset = RealVideoDataset(
251
+ real_video_dir=real_video_dir,
252
+ num_frames=num_frames,
253
+ sample_rate=sample_rate,
254
+ crop_size=crop_size,
255
+ resolution=resolution,
256
+ )
257
+
258
+ if subset_size:
259
+ indices = range(subset_size)
260
+ dataset = Subset(dataset, indices=indices)
261
+ ddp_sampler = DistributedSampler(dataset)
262
+ dataloader = DataLoader(
263
+ dataset, batch_size=batch_size, sampler=ddp_sampler ,pin_memory=True, num_workers=num_workers
264
+ )
265
+ # ---- Prepare Dataset
266
+
267
+ # ---- Inference ----
268
+ for batch in tqdm(dataloader):
269
+ x, file_names = batch['video'], batch['file_name']
270
+
271
+ x = x.to(rank).to(data_type) # b c t h w
272
+ z = vqgan.pre_vq_conv(vqgan.encoder(x))
273
+ vq_output = vqgan.codebook(z)
274
+ latents = vq_output['embeddings']
275
+ print(latents.shape)
276
+ video_recon = vqgan.decoder(vqgan.post_vq_conv(latents.to(data_type)))
277
+ for idx, video in enumerate(video_recon):
278
+ output_path = os.path.join(generated_video_dir, "vae_gen/", file_names[idx])
279
+ if args.output_origin:
280
+ os.makedirs(os.path.join(generated_video_dir, "origin/"), exist_ok=True)
281
+ origin_output_path = os.path.join(generated_video_dir, "origin/", file_names[idx])
282
+ custom_to_video(
283
+ x[idx], fps=sample_fps / sample_rate, output_file=origin_output_path
284
+ )
285
+ custom_to_video(
286
+ video, fps=sample_fps / sample_rate, output_file=output_path
287
+ )
288
+ # ---- Inference ----
289
+
290
+ if __name__ == "__main__":
291
+ parser = argparse.ArgumentParser()
292
+ parser.add_argument("--real_video_dir", type=str, default="")
293
+ parser.add_argument("--generated_video_dir", type=str, default="")
294
+ parser.add_argument("--ckpt", type=str, default="")
295
+ parser.add_argument("--sample_fps", type=int, default=30)
296
+ parser.add_argument("--resolution", type=int, default=336)
297
+ parser.add_argument("--crop_size", type=int, default=None)
298
+ parser.add_argument("--num_frames", type=int, default=17)
299
+ parser.add_argument("--sample_rate", type=int, default=1)
300
+ parser.add_argument("--batch_size", type=int, default=1)
301
+ parser.add_argument("--num_workers", type=int, default=8)
302
+ parser.add_argument("--subset_size", type=int, default=None)
303
+ parser.add_argument('--output_origin', action='store_true')
304
+ parser.add_argument("--config", type=str, default="")
305
+
306
+ ###VQGAN args
307
+ parser = VQGAN.add_model_specific_args(parser)
308
+
309
+ args = parser.parse_args()
310
+ main(args)
scripts/rec_causalvideo_vae.py ADDED
@@ -0,0 +1,324 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ import random
2
+ import argparse
3
+ import cv2
4
+ from tqdm import tqdm
5
+ import numpy as np
6
+ import numpy.typing as npt
7
+ import torch
8
+ import torch.distributed as dist
9
+ from torch.nn.parallel import DistributedDataParallel as DDP
10
+ from torch.utils.data import DataLoader, DistributedSampler, Subset
11
+ from decord import VideoReader, cpu
12
+ from torch.nn import functional as F
13
+ from pytorchvideo.transforms import ShortSideScale
14
+ from torchvision.transforms import Lambda, Compose
15
+ from torchvision.transforms._transforms_video import CenterCropVideo
16
+ import sys
17
+ from torch.utils.data import Dataset, DataLoader, Subset
18
+ import os
19
+ import glob
20
+ sys.path.append(".")
21
+ from causalvideovae.model import CausalVAEModel
22
+
23
+
24
+ def ddp_setup():
25
+ dist.init_process_group(backend="nccl")
26
+ torch.cuda.set_device(int(os.environ["LOCAL_RANK"]))
27
+
28
+ def array_to_video(
29
+ image_array: npt.NDArray, fps: float = 30.0, output_file: str = "output_video.mp4"
30
+ ) -> None:
31
+ height, width, channels = image_array[0].shape
32
+ fourcc = cv2.VideoWriter_fourcc(*"mp4v")
33
+ video_writer = cv2.VideoWriter(output_file, fourcc, float(fps), (width, height))
34
+ for image in image_array:
35
+ image_rgb = cv2.cvtColor(image, cv2.COLOR_RGB2BGR)
36
+ video_writer.write(image_rgb)
37
+
38
+ video_writer.release()
39
+
40
+
41
+ def custom_to_video(
42
+ x: torch.Tensor, fps: float = 2.0, output_file: str = "output_video.mp4"
43
+ ) -> None:
44
+ x = x.detach().cpu()
45
+ x = torch.clamp(x, -1, 1)
46
+ x = (x + 1) / 2
47
+ x = x.permute(1, 2, 3, 0).float().numpy()
48
+ x = (255 * x).astype(np.uint8)
49
+ array_to_video(x, fps=fps, output_file=output_file)
50
+ return
51
+
52
+
53
+ def read_video(video_path: str, num_frames: int, sample_rate: int) -> torch.Tensor:
54
+ decord_vr = VideoReader(video_path, ctx=cpu(0), num_threads=8)
55
+ total_frames = len(decord_vr)
56
+ sample_frames_len = sample_rate * num_frames
57
+
58
+ if total_frames > sample_frames_len:
59
+ s = 0
60
+ e = s + sample_frames_len
61
+ num_frames = num_frames
62
+ else:
63
+ s = 0
64
+ e = total_frames
65
+ num_frames = int(total_frames / sample_frames_len * num_frames)
66
+ print(
67
+ f"sample_frames_len {sample_frames_len}, only can sample {num_frames * sample_rate}",
68
+ video_path,
69
+ total_frames,
70
+ )
71
+
72
+ frame_id_list = np.linspace(s, e - 1, num_frames, dtype=int)
73
+ video_data = decord_vr.get_batch(frame_id_list).asnumpy()
74
+ video_data = torch.from_numpy(video_data)
75
+ video_data = video_data.permute(3, 0, 1, 2) # (T, H, W, C) -> (C, T, H, W)
76
+ return video_data
77
+
78
+
79
+ class RealVideoDataset(Dataset):
80
+ video_exts = ['avi', 'mp4', 'webm']
81
+
82
+ def __init__(
83
+ self,
84
+ real_video_dir,
85
+ num_frames,
86
+ sample_rate=1,
87
+ crop_size=None,
88
+ resolution=128,
89
+ ) -> None:
90
+ super().__init__()
91
+ self.real_video_files = self._combine_without_prefix(real_video_dir)
92
+ self.num_frames = num_frames
93
+ self.sample_rate = sample_rate
94
+ self.crop_size = crop_size
95
+ self.short_size = resolution
96
+
97
+ def __len__(self):
98
+ return len(self.real_video_files)
99
+
100
+ def __getitem__(self, index):
101
+ try:
102
+ if index >= len(self):
103
+ raise IndexError
104
+ real_video_file = self.real_video_files[index]
105
+ real_video_tensor = self._load_video(real_video_file)
106
+ video_name = os.path.basename(real_video_file)
107
+ except:
108
+ if index >= len(self):
109
+ raise IndexError
110
+ real_video_file = self.real_video_files[random.randint(1,index-1)]
111
+ real_video_tensor = self._load_video(real_video_file)
112
+ video_name = os.path.basename(real_video_file)
113
+ return {'video': real_video_tensor, 'file_name': video_name }
114
+
115
+ def _load_video(self, video_path):
116
+ num_frames = self.num_frames
117
+ sample_rate = self.sample_rate
118
+ decord_vr = VideoReader(video_path, ctx=cpu(0))
119
+ total_frames = len(decord_vr)
120
+ sample_frames_len = sample_rate * num_frames
121
+ s = 0
122
+ e = s + sample_frames_len
123
+ num_frames = num_frames
124
+ """
125
+ if total_frames > sample_frames_len:
126
+ s = 0
127
+ e = s + sample_frames_len
128
+ num_frames = num_frames
129
+
130
+ else:
131
+ s = 0
132
+ e = total_frames
133
+ num_frames = int(total_frames / sample_frames_len * num_frames)
134
+ print(
135
+ f"sample_frames_len {sample_frames_len}, only can sample {num_frames * sample_rate}",
136
+ video_path,
137
+ total_frames,
138
+ )
139
+ """
140
+ frame_id_list = np.linspace(s, e - 1, num_frames, dtype=int)
141
+ video_data = decord_vr.get_batch(frame_id_list).asnumpy()
142
+ video_data = torch.from_numpy(video_data)
143
+ video_data = video_data.permute(3, 0, 1, 2)
144
+ return _preprocess(
145
+ video_data, short_size=self.short_size, crop_size=self.crop_size
146
+ )
147
+
148
+ def _combine_without_prefix(self, folder_path):
149
+ samples = []
150
+ samples += sum([glob.glob(os.path.join(folder_path, '**', f'*.{ext}'), recursive=True)
151
+ for ext in self.video_exts], [])
152
+ samples.sort()
153
+ return samples
154
+
155
+ def resize(x, resolution):
156
+ height, width = x.shape[-2:]
157
+
158
+ aspect_ratio = width / height
159
+ if width <= height:
160
+ new_width = resolution
161
+ new_height = int(resolution / aspect_ratio)
162
+ else:
163
+ new_height = resolution
164
+ new_width = int(resolution * aspect_ratio)
165
+ resized_x = F.interpolate(x, size=(new_height, new_width), mode='bilinear', align_corners=True, antialias=True)
166
+ return resized_x
167
+
168
+ def _preprocess(video_data, short_size=128, crop_size=None):
169
+ transform = Compose(
170
+ [
171
+ Lambda(lambda x: ((x / 255.0) * 2 - 1)),
172
+
173
+ Lambda(lambda x: resize(x, short_size)),
174
+ (
175
+ CenterCropVideo(crop_size=crop_size)
176
+ if crop_size is not None
177
+ else Lambda(lambda x: x)
178
+ ),
179
+ ]
180
+
181
+ )
182
+ video_outputs = transform(video_data)
183
+ video_outputs = _format_video_shape(video_outputs)
184
+ return video_outputs
185
+
186
+
187
+ def _format_video_shape(video, time_compress=4, spatial_compress=8):
188
+ time = video.shape[1]
189
+ height = video.shape[2]
190
+ width = video.shape[3]
191
+ new_time = (
192
+ (time - (time - 1) % time_compress)
193
+ if (time - 1) % time_compress != 0
194
+ else time
195
+ )
196
+ new_height = (
197
+ (height - (height) % spatial_compress)
198
+ if height % spatial_compress != 0
199
+ else height
200
+ )
201
+ new_width = (
202
+ (width - (width) % spatial_compress) if width % spatial_compress != 0 else width
203
+ )
204
+ return video[:, :new_time, :new_height, :new_width]
205
+
206
+
207
+ @torch.no_grad()
208
+ def main(args: argparse.Namespace):
209
+ real_video_dir = args.real_video_dir
210
+ generated_video_dir = args.generated_video_dir
211
+ ckpt = args.ckpt
212
+ sample_rate = args.sample_rate
213
+ resolution = args.resolution
214
+ crop_size = args.crop_size
215
+ num_frames = args.num_frames
216
+ sample_rate = args.sample_rate
217
+ device = args.device
218
+ sample_fps = args.sample_fps
219
+ batch_size = args.batch_size
220
+ num_workers = args.num_workers
221
+ subset_size = args.subset_size
222
+
223
+ if not os.path.exists(args.generated_video_dir):
224
+ os.makedirs(os.path.join(generated_video_dir, "vae_gen/"), exist_ok=True)
225
+
226
+ data_type = torch.bfloat16
227
+
228
+ ddp_setup()
229
+ rank = int(os.environ["LOCAL_RANK"])
230
+
231
+ # ---- Load Model ----
232
+ device = args.device
233
+ vqvae = CausalVAEModel.from_pretrained(args.ckpt)
234
+ if args.change_decoder:
235
+ sd = torch.load(args.decoder_dir, map_location="cpu")
236
+
237
+ if "ema_state_dict" in sd and len(sd['ema_state_dict']) > 0 and os.environ.get("NOT_USE_EMA_MODEL", 0) == 0:
238
+ print("Load from ema model!")
239
+ sd = sd["ema_state_dict"]
240
+ sd = {key.replace("module.", ""): value for key, value in sd.items()}
241
+ elif "state_dict" in sd:
242
+ print("Load from normal model!")
243
+ if "gen_model" in sd["state_dict"]:
244
+ sd = sd["state_dict"]["gen_model"]
245
+ else:
246
+ sd = sd["state_dict"]
247
+ vqvae.load_state_dict(sd, strict=False)
248
+ #print(vqvae)
249
+ if args.enable_tiling:
250
+ vqvae.enable_tiling()
251
+ vqvae.tile_overlap_factor = args.tile_overlap_factor
252
+ vqvae = vqvae.to(rank).to(data_type)
253
+ vqvae.eval()
254
+
255
+ # ---- Load Model ----
256
+
257
+ # ---- Prepare Dataset ----
258
+ dataset = RealVideoDataset(
259
+ real_video_dir=real_video_dir,
260
+ num_frames=num_frames,
261
+ sample_rate=sample_rate,
262
+ crop_size=crop_size,
263
+ resolution=resolution,
264
+ )
265
+
266
+ if subset_size:
267
+ indices = range(subset_size)
268
+ dataset = Subset(dataset, indices=indices)
269
+ ddp_sampler = DistributedSampler(dataset)
270
+ dataloader = DataLoader(
271
+ dataset, batch_size=batch_size, sampler=ddp_sampler ,pin_memory=True, num_workers=num_workers
272
+ )
273
+ # ---- Prepare Dataset
274
+ # ---- Inference ----
275
+ all_latents = []
276
+ for batch in tqdm(dataloader):
277
+ x, file_names = batch['video'], batch['file_name']
278
+
279
+ x = x.to(device=device, dtype=data_type) # b c t h w
280
+ latents = vqvae.encode(x).sample().to(data_type)
281
+ all_latents.append(latents)
282
+ """
283
+ video_recon = vqvae.decode(latents)
284
+ for idx, video in enumerate(video_recon):
285
+ output_path = os.path.join(generated_video_dir, "vae_gen/", file_names[idx])
286
+ if args.output_origin:
287
+ os.makedirs(os.path.join(generated_video_dir, "origin/"), exist_ok=True)
288
+ origin_output_path = os.path.join(generated_video_dir, "origin/", file_names[idx])
289
+ custom_to_video(
290
+ x[idx], fps=sample_fps / sample_rate, output_file=origin_output_path
291
+ )
292
+ custom_to_video(
293
+ video, fps=sample_fps / sample_rate, output_file=output_path
294
+ )
295
+ """
296
+ all_latents_tensor = torch.cat(all_latents)
297
+ std = all_latents_tensor.std().item()
298
+ normalizer = 1 / std
299
+ print(f'{normalizer = }')
300
+ # ---- Inference ----
301
+
302
+ if __name__ == "__main__":
303
+ parser = argparse.ArgumentParser()
304
+ parser.add_argument("--real_video_dir", type=str, default="")
305
+ parser.add_argument("--generated_video_dir", type=str, default="")
306
+ parser.add_argument("--decoder_dir", type=str, default="")
307
+ parser.add_argument("--ckpt", type=str, default="")
308
+ parser.add_argument("--sample_fps", type=int, default=30)
309
+ parser.add_argument("--resolution", type=int, default=336)
310
+ parser.add_argument("--crop_size", type=int, default=None)
311
+ parser.add_argument("--num_frames", type=int, default=17)
312
+ parser.add_argument("--sample_rate", type=int, default=1)
313
+ parser.add_argument("--batch_size", type=int, default=1)
314
+ parser.add_argument("--num_workers", type=int, default=8)
315
+ parser.add_argument("--subset_size", type=int, default=None)
316
+ parser.add_argument("--tile_overlap_factor", type=float, default=0.25)
317
+ parser.add_argument('--enable_tiling', action='store_true')
318
+ parser.add_argument('--output_origin', action='store_true')
319
+ parser.add_argument('--change_decoder', action='store_true')
320
+ parser.add_argument("--device", type=str, default="cuda")
321
+
322
+ args = parser.parse_args()
323
+ main(args)
324
+
scripts/rec_nus_vae.py ADDED
@@ -0,0 +1,302 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ import random
2
+ import argparse
3
+ import cv2
4
+ from tqdm import tqdm
5
+ import numpy as np
6
+ import numpy.typing as npt
7
+ import torch
8
+ import torch.distributed as dist
9
+ from torch.nn.parallel import DistributedDataParallel as DDP
10
+ from torch.utils.data import DataLoader, DistributedSampler, Subset
11
+ from decord import VideoReader, cpu
12
+ from torch.nn import functional as F
13
+ from pytorchvideo.transforms import ShortSideScale
14
+ from torchvision.transforms import Lambda, Compose
15
+ from torchvision.transforms._transforms_video import CenterCropVideo
16
+ import sys
17
+ from torch.utils.data import Dataset, DataLoader, Subset
18
+ import os
19
+ import glob
20
+ sys.path.append(".")
21
+ import torch.nn as nn
22
+ import yaml
23
+ from omegaconf import OmegaConf
24
+ from einops import rearrange
25
+ from opensora.registry import MODELS, build_module
26
+ from opensora.utils.config_utils import parse_configs
27
+
28
+ def ddp_setup():
29
+ dist.init_process_group(backend="nccl")
30
+ torch.cuda.set_device(int(os.environ["LOCAL_RANK"]))
31
+
32
+ def array_to_video(
33
+ image_array: npt.NDArray, fps: float = 30.0, output_file: str = "output_video.mp4"
34
+ ) -> None:
35
+ height, width, channels = image_array[0].shape
36
+ fourcc = cv2.VideoWriter_fourcc(*"mp4v")
37
+ video_writer = cv2.VideoWriter(output_file, fourcc, float(fps), (width, height))
38
+
39
+ for image in image_array:
40
+ image_rgb = cv2.cvtColor(image, cv2.COLOR_RGB2BGR)
41
+ video_writer.write(image_rgb)
42
+
43
+ video_writer.release()
44
+
45
+
46
+ def custom_to_video(
47
+ x: torch.Tensor, fps: float = 2.0, output_file: str = "output_video.mp4"
48
+ ) -> None:
49
+ x = x.detach().cpu()
50
+ x = torch.clamp(x, -1, 1)
51
+ x = (x + 1) / 2
52
+ x = x.permute(1, 2, 3, 0).float().numpy()
53
+ x = (255 * x).astype(np.uint8)
54
+ array_to_video(x, fps=fps, output_file=output_file)
55
+ return
56
+
57
+
58
+ def read_video(video_path: str, num_frames: int, sample_rate: int) -> torch.Tensor:
59
+ decord_vr = VideoReader(video_path, ctx=cpu(0), num_threads=8)
60
+ total_frames = len(decord_vr)
61
+ sample_frames_len = sample_rate * num_frames
62
+
63
+ if total_frames > sample_frames_len:
64
+ s = 0
65
+ e = s + sample_frames_len
66
+ num_frames = num_frames
67
+ else:
68
+ s = 0
69
+ e = total_frames
70
+ num_frames = int(total_frames / sample_frames_len * num_frames)
71
+ print(
72
+ f"sample_frames_len {sample_frames_len}, only can sample {num_frames * sample_rate}",
73
+ video_path,
74
+ total_frames,
75
+ )
76
+
77
+ frame_id_list = np.linspace(s, e - 1, num_frames, dtype=int)
78
+ video_data = decord_vr.get_batch(frame_id_list).asnumpy()
79
+ video_data = torch.from_numpy(video_data)
80
+ video_data = video_data.permute(3, 0, 1, 2) # (T, H, W, C) -> (C, T, H, W)
81
+ return video_data
82
+
83
+
84
+ class RealVideoDataset(Dataset):
85
+ video_exts = ['avi', 'mp4', 'webm']
86
+
87
+ def __init__(
88
+ self,
89
+ real_video_dir,
90
+ num_frames,
91
+ sample_rate=1,
92
+ crop_size=None,
93
+ resolution=128,
94
+ ) -> None:
95
+ super().__init__()
96
+ self.real_video_files = self._combine_without_prefix(real_video_dir)
97
+ self.num_frames = num_frames
98
+ self.sample_rate = sample_rate
99
+ self.crop_size = crop_size
100
+ self.short_size = resolution
101
+
102
+ def __len__(self):
103
+ return len(self.real_video_files)
104
+
105
+ def __getitem__(self, index):
106
+ try:
107
+ if index >= len(self):
108
+ raise IndexError
109
+ real_video_file = self.real_video_files[index]
110
+ real_video_tensor = self._load_video(real_video_file)
111
+ video_name = os.path.basename(real_video_file)
112
+ except:
113
+ if index >= len(self):
114
+ raise IndexError
115
+ real_video_file = self.real_video_files[random.randint(1,index-1)]
116
+ real_video_tensor = self._load_video(real_video_file)
117
+ video_name = os.path.basename(real_video_file)
118
+ return {'video': real_video_tensor, 'file_name': video_name }
119
+
120
+ def _load_video(self, video_path):
121
+ num_frames = self.num_frames
122
+ sample_rate = self.sample_rate
123
+ decord_vr = VideoReader(video_path, ctx=cpu(0))
124
+ total_frames = len(decord_vr)
125
+ sample_frames_len = sample_rate * num_frames
126
+ s = 0
127
+ e = s + sample_frames_len
128
+ num_frames = num_frames
129
+ """
130
+ if total_frames > sample_frames_len:
131
+ s = 0
132
+ e = s + sample_frames_len
133
+ num_frames = num_frames
134
+
135
+ else:
136
+ s = 0
137
+ e = total_frames
138
+ num_frames = int(total_frames / sample_frames_len * num_frames)
139
+ print(
140
+ f"sample_frames_len {sample_frames_len}, only can sample {num_frames * sample_rate}",
141
+ video_path,
142
+ total_frames,
143
+ )
144
+ """
145
+ frame_id_list = np.linspace(s, e - 1, num_frames, dtype=int)
146
+ video_data = decord_vr.get_batch(frame_id_list).asnumpy()
147
+ video_data = torch.from_numpy(video_data)
148
+ video_data = video_data.permute(3, 0, 1, 2)
149
+ return _preprocess(
150
+ video_data, short_size=self.short_size, crop_size=self.crop_size
151
+ )
152
+
153
+ def _combine_without_prefix(self, folder_path):
154
+ samples = []
155
+ samples += sum([glob.glob(os.path.join(folder_path, '**', f'*.{ext}'), recursive=True)
156
+ for ext in self.video_exts], [])
157
+ samples.sort()
158
+ return samples
159
+
160
+ def resize(x, resolution):
161
+ height, width = x.shape[-2:]
162
+ aspect_ratio = width / height
163
+ if width <= height:
164
+ new_width = resolution
165
+ new_height = int(resolution / aspect_ratio)
166
+ else:
167
+ new_height = resolution
168
+ new_width = int(resolution * aspect_ratio)
169
+ resized_x = F.interpolate(x, size=(new_height, new_width), mode='bilinear', align_corners=True, antialias=True)
170
+ return resized_x
171
+
172
+ def _preprocess(video_data, short_size=128, crop_size=None):
173
+ transform = Compose(
174
+
175
+ [
176
+
177
+ Lambda(lambda x: ((x / 255.0) * 2 - 1)),
178
+ Lambda(lambda x: resize(x, short_size)),
179
+ (
180
+ CenterCropVideo(crop_size=crop_size)
181
+ if crop_size is not None
182
+ else Lambda(lambda x: x)
183
+ ),
184
+
185
+ ]
186
+
187
+ )
188
+ video_outputs = transform(video_data)
189
+ video_outputs = _format_video_shape(video_outputs)
190
+ return video_outputs
191
+
192
+
193
+ def _format_video_shape(video, time_compress=4, spatial_compress=8):
194
+ time = video.shape[1]
195
+ height = video.shape[2]
196
+ width = video.shape[3]
197
+ new_time = (
198
+ (time - (time - 1) % time_compress)
199
+ if (time - 1) % time_compress != 0
200
+ else time
201
+ )
202
+ new_height = (
203
+ (height - (height) % spatial_compress)
204
+ if height % spatial_compress != 0
205
+ else height
206
+ )
207
+ new_width = (
208
+ (width - (width) % spatial_compress) if width % spatial_compress != 0 else width
209
+ )
210
+ return video[:, :new_time, :new_height, :new_width]
211
+
212
+ @torch.no_grad()
213
+ def main(args: argparse.Namespace):
214
+ real_video_dir = args.real_video_dir
215
+ generated_video_dir = args.generated_video_dir
216
+ ckpt = args.ckpt
217
+ sample_rate = args.sample_rate
218
+ resolution = args.resolution
219
+ crop_size = args.crop_size
220
+ num_frames = args.num_frames
221
+ sample_rate = args.sample_rate
222
+ sample_fps = args.sample_fps
223
+ batch_size = args.batch_size
224
+ num_workers = args.num_workers
225
+ subset_size = args.subset_size
226
+
227
+ if not os.path.exists(args.generated_video_dir):
228
+ os.makedirs(os.path.join(generated_video_dir, "vae_gen/"), exist_ok=True)
229
+
230
+ data_type = torch.bfloat16
231
+
232
+ ddp_setup()
233
+ rank = int(os.environ["LOCAL_RANK"])
234
+
235
+ # ---- Load Model ----
236
+ cfg = parse_configs(args, training=False)
237
+ nus_vae = build_module(cfg.model, MODELS)
238
+ print(nus_vae)
239
+ nus_vae = nus_vae.to(rank).to(data_type)
240
+ nus_vae.eval()
241
+
242
+ # ---- Load Model ----
243
+
244
+ # ---- Prepare Dataset ----
245
+ dataset = RealVideoDataset(
246
+ real_video_dir=real_video_dir,
247
+ num_frames=num_frames,
248
+ sample_rate=sample_rate,
249
+ crop_size=crop_size,
250
+ resolution=resolution,
251
+ )
252
+
253
+ if subset_size:
254
+ indices = range(subset_size)
255
+ dataset = Subset(dataset, indices=indices)
256
+ ddp_sampler = DistributedSampler(dataset)
257
+ dataloader = DataLoader(
258
+ dataset, batch_size=batch_size, sampler=ddp_sampler ,pin_memory=True, num_workers=num_workers
259
+ )
260
+ # ---- Prepare Dataset
261
+
262
+ # ---- Inference ----
263
+ for batch in tqdm(dataloader):
264
+ x, file_names = batch['video'], batch['file_name']
265
+
266
+ x = x.to(rank).to(data_type) # b c t h w
267
+ latents, posterior, x_z = nus_vae.encode(x)
268
+ video_recon, x_z_rec = nus_vae.decode(latents, num_frames=x.size(2))
269
+
270
+ all_latents = []
271
+ for idx, video in enumerate(video_recon):
272
+ output_path = os.path.join(generated_video_dir, "vae_gen/", file_names[idx])
273
+ if args.output_origin:
274
+ os.makedirs(os.path.join(generated_video_dir, "origin/"), exist_ok=True)
275
+ origin_output_path = os.path.join(generated_video_dir, "origin/", file_names[idx])
276
+ custom_to_video(
277
+ x[idx], fps=sample_fps / sample_rate, output_file=origin_output_path
278
+ )
279
+ custom_to_video(
280
+ video, fps=sample_fps / sample_rate, output_file=output_path
281
+ )
282
+ # ---- Inference ----
283
+
284
+ if __name__ == "__main__":
285
+ parser = argparse.ArgumentParser()
286
+ parser.add_argument("--real_video_dir", type=str, default="")
287
+ parser.add_argument("--generated_video_dir", type=str, default="")
288
+ parser.add_argument("--ckpt", type=str, default="")
289
+ parser.add_argument("--sample_fps", type=int, default=30)
290
+ parser.add_argument("--resolution", type=int, default=336)
291
+ parser.add_argument("--crop_size", type=int, default=None)
292
+ parser.add_argument("--num_frames", type=int, default=17)
293
+ parser.add_argument("--sample_rate", type=int, default=1)
294
+ parser.add_argument("--batch_size", type=int, default=1)
295
+ parser.add_argument("--num_workers", type=int, default=8)
296
+ parser.add_argument("--subset_size", type=int, default=None)
297
+ parser.add_argument('--output_origin', action='store_true')
298
+ parser.add_argument("--config", type=str, default="")
299
+
300
+
301
+ args = parser.parse_args()
302
+ main(args)
scripts/rec_svd_vae.py ADDED
@@ -0,0 +1,301 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ import random
2
+ import argparse
3
+ import cv2
4
+ from tqdm import tqdm
5
+ import numpy as np
6
+ import numpy.typing as npt
7
+ import torch
8
+ import torch.distributed as dist
9
+ from torch.nn.parallel import DistributedDataParallel as DDP
10
+ from torch.utils.data import DataLoader, DistributedSampler, Subset
11
+ from decord import VideoReader, cpu
12
+ from torch.nn import functional as F
13
+ from pytorchvideo.transforms import ShortSideScale
14
+ from torchvision.transforms import Lambda, Compose
15
+ from torchvision.transforms._transforms_video import CenterCropVideo
16
+ import sys
17
+ from torch.utils.data import Dataset, DataLoader, Subset
18
+ import os
19
+ import glob
20
+ sys.path.append(".")
21
+ import torch.nn as nn
22
+ import yaml
23
+ from omegaconf import OmegaConf
24
+ from einops import rearrange
25
+ from diffusers.models import AutoencoderKLTemporalDecoder
26
+
27
+ def ddp_setup():
28
+ dist.init_process_group(backend="nccl")
29
+ torch.cuda.set_device(int(os.environ["LOCAL_RANK"]))
30
+
31
+ def array_to_video(
32
+ image_array: npt.NDArray, fps: float = 30.0, output_file: str = "output_video.mp4"
33
+ ) -> None:
34
+ height, width, channels = image_array[0].shape
35
+ fourcc = cv2.VideoWriter_fourcc(*"mp4v")
36
+ video_writer = cv2.VideoWriter(output_file, fourcc, float(fps), (width, height))
37
+
38
+ for image in image_array:
39
+ image_rgb = cv2.cvtColor(image, cv2.COLOR_RGB2BGR)
40
+ video_writer.write(image_rgb)
41
+
42
+ video_writer.release()
43
+
44
+
45
+ def custom_to_video(
46
+ x: torch.Tensor, fps: float = 2.0, output_file: str = "output_video.mp4"
47
+ ) -> None:
48
+ x = x.detach().cpu()
49
+ x = torch.clamp(x, -1, 1)
50
+ x = (x + 1) / 2
51
+ x = x.permute(1, 2, 3, 0).float().numpy()
52
+ x = (255 * x).astype(np.uint8)
53
+ array_to_video(x, fps=fps, output_file=output_file)
54
+ return
55
+
56
+
57
+ def read_video(video_path: str, num_frames: int, sample_rate: int) -> torch.Tensor:
58
+ decord_vr = VideoReader(video_path, ctx=cpu(0), num_threads=8)
59
+ total_frames = len(decord_vr)
60
+ sample_frames_len = sample_rate * num_frames
61
+
62
+ if total_frames > sample_frames_len:
63
+ s = 0
64
+ e = s + sample_frames_len
65
+ num_frames = num_frames
66
+ else:
67
+ s = 0
68
+ e = total_frames
69
+ num_frames = int(total_frames / sample_frames_len * num_frames)
70
+ print(
71
+ f"sample_frames_len {sample_frames_len}, only can sample {num_frames * sample_rate}",
72
+ video_path,
73
+ total_frames,
74
+ )
75
+
76
+ frame_id_list = np.linspace(s, e - 1, num_frames, dtype=int)
77
+ video_data = decord_vr.get_batch(frame_id_list).asnumpy()
78
+ video_data = torch.from_numpy(video_data)
79
+ video_data = video_data.permute(3, 0, 1, 2) # (T, H, W, C) -> (C, T, H, W)
80
+ return video_data
81
+
82
+
83
+ class RealVideoDataset(Dataset):
84
+ video_exts = ['avi', 'mp4', 'webm']
85
+
86
+ def __init__(
87
+ self,
88
+ real_video_dir,
89
+ num_frames,
90
+ sample_rate=1,
91
+ crop_size=None,
92
+ resolution=128,
93
+ ) -> None:
94
+ super().__init__()
95
+ self.real_video_files = self._combine_without_prefix(real_video_dir)
96
+ self.num_frames = num_frames
97
+ self.sample_rate = sample_rate
98
+ self.crop_size = crop_size
99
+ self.short_size = resolution
100
+
101
+ def __len__(self):
102
+ return len(self.real_video_files)
103
+
104
+ def __getitem__(self, index):
105
+ try:
106
+ if index >= len(self):
107
+ raise IndexError
108
+ real_video_file = self.real_video_files[index]
109
+ real_video_tensor = self._load_video(real_video_file)
110
+ video_name = os.path.basename(real_video_file)
111
+ except:
112
+ if index >= len(self):
113
+ raise IndexError
114
+ real_video_file = self.real_video_files[random.randint(1,index-1)]
115
+ real_video_tensor = self._load_video(real_video_file)
116
+ video_name = os.path.basename(real_video_file)
117
+ return {'video': real_video_tensor, 'file_name': video_name }
118
+
119
+ def _load_video(self, video_path):
120
+ num_frames = self.num_frames
121
+ sample_rate = self.sample_rate
122
+ decord_vr = VideoReader(video_path, ctx=cpu(0))
123
+ total_frames = len(decord_vr)
124
+ sample_frames_len = sample_rate * num_frames
125
+ s = 0
126
+ e = s + sample_frames_len
127
+ num_frames = num_frames
128
+ """
129
+ if total_frames > sample_frames_len:
130
+ s = 0
131
+ e = s + sample_frames_len
132
+ num_frames = num_frames
133
+
134
+ else:
135
+ s = 0
136
+ e = total_frames
137
+ num_frames = int(total_frames / sample_frames_len * num_frames)
138
+ print(
139
+ f"sample_frames_len {sample_frames_len}, only can sample {num_frames * sample_rate}",
140
+ video_path,
141
+ total_frames,
142
+ )
143
+ """
144
+ frame_id_list = np.linspace(s, e - 1, num_frames, dtype=int)
145
+ video_data = decord_vr.get_batch(frame_id_list).asnumpy()
146
+ video_data = torch.from_numpy(video_data)
147
+ video_data = video_data.permute(3, 0, 1, 2)
148
+ return _preprocess(
149
+ video_data, short_size=self.short_size, crop_size=self.crop_size
150
+ )
151
+
152
+ def _combine_without_prefix(self, folder_path):
153
+ samples = []
154
+ samples += sum([glob.glob(os.path.join(folder_path, '**', f'*.{ext}'), recursive=True)
155
+ for ext in self.video_exts], [])
156
+ samples.sort()
157
+ return samples
158
+
159
+ def resize(x, resolution):
160
+ height, width = x.shape[-2:]
161
+ aspect_ratio = width / height
162
+ if width <= height:
163
+ new_width = resolution
164
+ new_height = int(resolution / aspect_ratio)
165
+ else:
166
+ new_height = resolution
167
+ new_width = int(resolution * aspect_ratio)
168
+ resized_x = F.interpolate(x, size=(new_height, new_width), mode='bilinear', align_corners=True, antialias=True)
169
+ return resized_x
170
+
171
+ def _preprocess(video_data, short_size=128, crop_size=None):
172
+ transform = Compose(
173
+
174
+ [
175
+
176
+ Lambda(lambda x: ((x / 255.0) * 2 - 1)),
177
+ Lambda(lambda x: resize(x, short_size)),
178
+ (
179
+ CenterCropVideo(crop_size=crop_size)
180
+ if crop_size is not None
181
+ else Lambda(lambda x: x)
182
+ ),
183
+
184
+ ]
185
+
186
+ )
187
+ video_outputs = transform(video_data)
188
+ video_outputs = _format_video_shape(video_outputs)
189
+ return video_outputs
190
+
191
+
192
+ def _format_video_shape(video, time_compress=4, spatial_compress=8):
193
+ time = video.shape[1]
194
+ height = video.shape[2]
195
+ width = video.shape[3]
196
+ new_time = (
197
+ (time - (time - 1) % time_compress)
198
+ if (time - 1) % time_compress != 0
199
+ else time
200
+ )
201
+ new_height = (
202
+ (height - (height) % spatial_compress)
203
+ if height % spatial_compress != 0
204
+ else height
205
+ )
206
+ new_width = (
207
+ (width - (width) % spatial_compress) if width % spatial_compress != 0 else width
208
+ )
209
+ return video[:, :new_time, :new_height, :new_width]
210
+
211
+ @torch.no_grad()
212
+ def main(args: argparse.Namespace):
213
+ real_video_dir = args.real_video_dir
214
+ generated_video_dir = args.generated_video_dir
215
+ ckpt = args.ckpt
216
+ sample_rate = args.sample_rate
217
+ resolution = args.resolution
218
+ crop_size = args.crop_size
219
+ num_frames = args.num_frames
220
+ sample_rate = args.sample_rate
221
+ sample_fps = args.sample_fps
222
+ batch_size = args.batch_size
223
+ num_workers = args.num_workers
224
+ subset_size = args.subset_size
225
+
226
+ if not os.path.exists(args.generated_video_dir):
227
+ os.makedirs(os.path.join(generated_video_dir, "vae_gen/"), exist_ok=True)
228
+
229
+ data_type = torch.bfloat16
230
+
231
+ ddp_setup()
232
+ rank = int(os.environ["LOCAL_RANK"])
233
+
234
+ # ---- Load Model ----
235
+ svd_vae = AutoencoderKLTemporalDecoder.from_pretrained(ckpt)
236
+ print(svd_vae)
237
+ svd_vae = svd_vae.to(rank).to(data_type)
238
+ svd_vae.eval()
239
+
240
+ # ---- Load Model ----
241
+
242
+ # ---- Prepare Dataset ----
243
+ dataset = RealVideoDataset(
244
+ real_video_dir=real_video_dir,
245
+ num_frames=num_frames,
246
+ sample_rate=sample_rate,
247
+ crop_size=crop_size,
248
+ resolution=resolution,
249
+ )
250
+
251
+ if subset_size:
252
+ indices = range(subset_size)
253
+ dataset = Subset(dataset, indices=indices)
254
+ ddp_sampler = DistributedSampler(dataset)
255
+ dataloader = DataLoader(
256
+ dataset, batch_size=batch_size, sampler=ddp_sampler ,pin_memory=True, num_workers=num_workers
257
+ )
258
+ # ---- Prepare Dataset
259
+
260
+ # ---- Inference ----
261
+ for batch in tqdm(dataloader):
262
+ x, file_names = batch['video'], batch['file_name']
263
+
264
+ x = x.to(rank).to(data_type) # b c t h w
265
+ t = x.shape[2]
266
+ x = rearrange(x, "b c t h w -> (b t) c h w", t=t)
267
+ latents = svd_vae.encode(x)['latent_dist'].sample()
268
+ video_recon = svd_vae.decode(latents.to(data_type), num_frames=t)['sample']
269
+ x = rearrange(x, "(b t) c h w -> b c t h w", t=t)
270
+ for idx, video in enumerate(video_recon):
271
+ output_path = os.path.join(generated_video_dir, "vae_gen/", file_names[idx])
272
+ if args.output_origin:
273
+ os.makedirs(os.path.join(generated_video_dir, "origin/"), exist_ok=True)
274
+ origin_output_path = os.path.join(generated_video_dir, "origin/", file_names[idx])
275
+ custom_to_video(
276
+ x[idx], fps=sample_fps / sample_rate, output_file=origin_output_path
277
+ )
278
+ custom_to_video(
279
+ video, fps=sample_fps / sample_rate, output_file=output_path
280
+ )
281
+ # ---- Inference ----
282
+
283
+ if __name__ == "__main__":
284
+ parser = argparse.ArgumentParser()
285
+ parser.add_argument("--real_video_dir", type=str, default="")
286
+ parser.add_argument("--generated_video_dir", type=str, default="")
287
+ parser.add_argument("--ckpt", type=str, default="")
288
+ parser.add_argument("--sample_fps", type=int, default=30)
289
+ parser.add_argument("--resolution", type=int, default=336)
290
+ parser.add_argument("--crop_size", type=int, default=None)
291
+ parser.add_argument("--num_frames", type=int, default=17)
292
+ parser.add_argument("--sample_rate", type=int, default=1)
293
+ parser.add_argument("--batch_size", type=int, default=1)
294
+ parser.add_argument("--num_workers", type=int, default=8)
295
+ parser.add_argument("--subset_size", type=int, default=None)
296
+ parser.add_argument('--output_origin', action='store_true')
297
+ parser.add_argument("--config", type=str, default="")
298
+
299
+
300
+ args = parser.parse_args()
301
+ main(args)