Upload causal LTX2 VAE
Browse files
ltx2_causal_vae_0516/config.json
ADDED
|
@@ -0,0 +1,82 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_class_name": "AutoencoderKLCausalLTX2Video",
|
| 3 |
+
"_diffusers_version": "0.38.0.dev0",
|
| 4 |
+
"block_out_channels": [
|
| 5 |
+
256,
|
| 6 |
+
512,
|
| 7 |
+
1024,
|
| 8 |
+
2048
|
| 9 |
+
],
|
| 10 |
+
"decoder_block_out_channels": [
|
| 11 |
+
256,
|
| 12 |
+
512,
|
| 13 |
+
1024
|
| 14 |
+
],
|
| 15 |
+
"decoder_causal": true,
|
| 16 |
+
"decoder_inject_noise": [
|
| 17 |
+
false,
|
| 18 |
+
false,
|
| 19 |
+
false,
|
| 20 |
+
false
|
| 21 |
+
],
|
| 22 |
+
"decoder_layers_per_block": [
|
| 23 |
+
5,
|
| 24 |
+
5,
|
| 25 |
+
5,
|
| 26 |
+
5
|
| 27 |
+
],
|
| 28 |
+
"decoder_spatial_padding_mode": "reflect",
|
| 29 |
+
"decoder_spatio_temporal_scaling": [
|
| 30 |
+
true,
|
| 31 |
+
true,
|
| 32 |
+
true
|
| 33 |
+
],
|
| 34 |
+
"decoder_upsample_type": null,
|
| 35 |
+
"down_block_types": [
|
| 36 |
+
"LTX2VideoDownBlock3D",
|
| 37 |
+
"LTX2VideoDownBlock3D",
|
| 38 |
+
"LTX2VideoDownBlock3D",
|
| 39 |
+
"LTX2VideoDownBlock3D"
|
| 40 |
+
],
|
| 41 |
+
"downsample_type": [
|
| 42 |
+
"spatial",
|
| 43 |
+
"temporal",
|
| 44 |
+
"spatiotemporal",
|
| 45 |
+
"spatiotemporal"
|
| 46 |
+
],
|
| 47 |
+
"encoder_causal": true,
|
| 48 |
+
"encoder_spatial_padding_mode": "zeros",
|
| 49 |
+
"in_channels": 3,
|
| 50 |
+
"latent_channels": 128,
|
| 51 |
+
"layers_per_block": [
|
| 52 |
+
4,
|
| 53 |
+
6,
|
| 54 |
+
6,
|
| 55 |
+
2,
|
| 56 |
+
2
|
| 57 |
+
],
|
| 58 |
+
"out_channels": 3,
|
| 59 |
+
"patch_size": 4,
|
| 60 |
+
"patch_size_t": 1,
|
| 61 |
+
"resnet_norm_eps": 1e-06,
|
| 62 |
+
"scaling_factor": 1.0,
|
| 63 |
+
"spatial_compression_ratio": 32,
|
| 64 |
+
"spatio_temporal_scaling": [
|
| 65 |
+
true,
|
| 66 |
+
true,
|
| 67 |
+
true,
|
| 68 |
+
true
|
| 69 |
+
],
|
| 70 |
+
"temporal_compression_ratio": 8,
|
| 71 |
+
"timestep_conditioning": false,
|
| 72 |
+
"upsample_factor": [
|
| 73 |
+
2,
|
| 74 |
+
2,
|
| 75 |
+
2
|
| 76 |
+
],
|
| 77 |
+
"upsample_residual": [
|
| 78 |
+
true,
|
| 79 |
+
true,
|
| 80 |
+
true
|
| 81 |
+
]
|
| 82 |
+
}
|
ltx2_causal_vae_0516/diffusion_pytorch_model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3694526c1da964fcf1ebccb70edb09895bf9e0a723b3ac48e0b5adf5bf97cc4a
|
| 3 |
+
size 4889942004
|