Fudan-FUXI commited on
Commit
fdb3e6b
1 Parent(s): 9d9cbd9

Upload 3 files

Browse files
transformer/diffusion_pytorch_model.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e47eb1a659a6e9df29fb5fc8d9a5f1c1f5e95c998a23a1498559a4f7e3ce659e
3
+ size 10863441186
vae/config.yaml ADDED
@@ -0,0 +1,51 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ embed_dim: 8
2
+ lossconfig:
3
+ type: LPIPS
4
+ params:
5
+ kl_weight: 1.0e-06
6
+ perceptual_weight: 1.0
7
+ ddconfig:
8
+ double_z: true
9
+ z_channels: 8
10
+ t_frames: 17
11
+ in_channels: 3
12
+ out_ch: 3
13
+ ch: 128
14
+ ch_mult:
15
+ - 1
16
+ - 2
17
+ - 4
18
+ - 4
19
+ spatial_stride:
20
+ - 2
21
+ - 2
22
+ - 2
23
+ - 1
24
+ temporal_stride:
25
+ - 2
26
+ - 2
27
+ - 1
28
+ - 1
29
+ num_res_blocks: 4
30
+ attn_resolutions: []
31
+ dropout: 0.0
32
+ attn_type: causual_spatial_temporal
33
+ revise_norm_bug: True
34
+
35
+ fps_ds: 8
36
+ resolution_video: 256
37
+ resolution_image: 256
38
+ model_name: CausualVAEVideo
39
+ precision: bf16
40
+ keep_aspect_ratio: 'False'
41
+
42
+
43
+ #scaling_factor_size_128_video: 0.21585
44
+ #scaling_factor_size_256_video: 0.22970
45
+ #scaling_factor_size_512_video: 0.21397
46
+ scaling_factor_video: 0.22970
47
+
48
+ #scaling_factor_size_128_image: 0.21585
49
+ #scaling_factor_size_256_image: 0.26691
50
+ #scaling_factor_size_512_image: 0.270743
51
+ scaling_factor_image: 0.26691
vae/diffusion_pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2a3dbd352da3509d447fd3fb7b1e5df53334ef4571dff2d87e41940afa2203e6
3
+ size 1575768318