Hugging Face's logo

Physis-AI
/
physisserve-ltx2p5_22b-same_step-ls-dev

Text-to-Video
Diffusers
Safetensors
LTX2Pipeline
physisserve
ltx-2.5
autoregressive
video-generation
same-step-kv
Model card Files
xet
 Community
Skyyyy0920 commited on
Commit
877fa76
·
verified ·
1 Parent(s): 6a8e351

Standard Diffusers layout: loadable by stock diffusers

Browse files
Files changed (5) hide show
  1. audio_vae/config.json +25 -0
  2. audio_vae/diffusion_pytorch_model.safetensors +3 -0
  3. model_index.json +4 -4
  4. vocoder/config.json +104 -0
  5. vocoder/diffusion_pytorch_model.safetensors +3 -0
audio_vae/config.json ADDED
@@ -0,0 +1,25 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_class_name": "AutoencoderKLLTX2Audio",
3
+ "_diffusers_version": "0.40.0.dev0",
4
+ "attn_resolutions": null,
5
+ "base_channels": 128,
6
+ "causality_axis": "height",
7
+ "ch_mult": [
8
+ 1,
9
+ 2,
10
+ 4
11
+ ],
12
+ "double_z": true,
13
+ "dropout": 0.0,
14
+ "in_channels": 2,
15
+ "is_causal": true,
16
+ "latent_channels": 8,
17
+ "mel_bins": 64,
18
+ "mel_hop_length": 160,
19
+ "mid_block_add_attention": false,
20
+ "norm_type": "pixel",
21
+ "num_res_blocks": 2,
22
+ "output_channels": 2,
23
+ "resolution": 256,
24
+ "sample_rate": 16000
25
+ }
audio_vae/diffusion_pytorch_model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:45dcb1c3cffc8c8b7710a04dd19500e481daff31f5ea8e8529f2cf697c439d12
3
+ size 106507972
model_index.json CHANGED
@@ -2,8 +2,8 @@
2
  "_class_name": "LTX2Pipeline",
3
  "_diffusers_version": "0.40.0.dev0",
4
  "audio_vae": [
5
- null,
6
- null
7
  ],
8
  "connectors": [
9
  "ltx2",
@@ -42,7 +42,7 @@
42
  "AutoencoderKLLTX2Video"
43
  ],
44
  "vocoder": [
45
- null,
46
- null
47
  ]
48
  }
 
2
  "_class_name": "LTX2Pipeline",
3
  "_diffusers_version": "0.40.0.dev0",
4
  "audio_vae": [
5
+ "diffusers",
6
+ "AutoencoderKLLTX2Audio"
7
  ],
8
  "connectors": [
9
  "ltx2",
 
42
  "AutoencoderKLLTX2Video"
43
  ],
44
  "vocoder": [
45
+ "ltx2",
46
+ "LTX2VocoderWithBWE"
47
  ]
48
  }
vocoder/config.json ADDED
@@ -0,0 +1,104 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_class_name": "LTX2VocoderWithBWE",
3
+ "_diffusers_version": "0.40.0.dev0",
4
+ "act_fn": "snakebeta",
5
+ "antialias": true,
6
+ "antialias_kernel_size": 12,
7
+ "antialias_ratio": 2,
8
+ "bwe_act_fn": "snakebeta",
9
+ "bwe_antialias": true,
10
+ "bwe_antialias_kernel_size": 12,
11
+ "bwe_antialias_ratio": 2,
12
+ "bwe_final_act_fn": null,
13
+ "bwe_final_bias": false,
14
+ "bwe_hidden_channels": 512,
15
+ "bwe_in_channels": 128,
16
+ "bwe_leaky_relu_negative_slope": 0.1,
17
+ "bwe_out_channels": 2,
18
+ "bwe_resnet_dilations": [
19
+ [
20
+ 1,
21
+ 3,
22
+ 5
23
+ ],
24
+ [
25
+ 1,
26
+ 3,
27
+ 5
28
+ ],
29
+ [
30
+ 1,
31
+ 3,
32
+ 5
33
+ ]
34
+ ],
35
+ "bwe_resnet_kernel_sizes": [
36
+ 3,
37
+ 7,
38
+ 11
39
+ ],
40
+ "bwe_upsample_factors": [
41
+ 6,
42
+ 5,
43
+ 2,
44
+ 2,
45
+ 2
46
+ ],
47
+ "bwe_upsample_kernel_sizes": [
48
+ 12,
49
+ 11,
50
+ 4,
51
+ 4,
52
+ 4
53
+ ],
54
+ "filter_length": 512,
55
+ "final_act_fn": null,
56
+ "final_bias": false,
57
+ "hidden_channels": 1536,
58
+ "hop_length": 80,
59
+ "in_channels": 128,
60
+ "input_sampling_rate": 16000,
61
+ "leaky_relu_negative_slope": 0.1,
62
+ "num_mel_channels": 64,
63
+ "out_channels": 2,
64
+ "output_sampling_rate": 48000,
65
+ "resnet_dilations": [
66
+ [
67
+ 1,
68
+ 3,
69
+ 5
70
+ ],
71
+ [
72
+ 1,
73
+ 3,
74
+ 5
75
+ ],
76
+ [
77
+ 1,
78
+ 3,
79
+ 5
80
+ ]
81
+ ],
82
+ "resnet_kernel_sizes": [
83
+ 3,
84
+ 7,
85
+ 11
86
+ ],
87
+ "upsample_factors": [
88
+ 5,
89
+ 2,
90
+ 2,
91
+ 2,
92
+ 2,
93
+ 2
94
+ ],
95
+ "upsample_kernel_sizes": [
96
+ 11,
97
+ 4,
98
+ 4,
99
+ 4,
100
+ 4,
101
+ 4
102
+ ],
103
+ "window_length": 512
104
+ }
vocoder/diffusion_pytorch_model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c99dde145a982974bc4c39608b3b929f232f0126a55d8326a47ad54f486ae23c
3
+ size 258305056