Commit
•
64fae56
1
Parent(s):
bdc45da
Upload 4 files
Browse files- model_index.json +24 -0
- transformer/transformer_config.json +19 -0
- vae/config.json +65 -0
- vae/diffusion_pytorch_model.safetensors +3 -0
model_index.json
ADDED
|
@@ -0,0 +1,24 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_class_name": "LTXPipeline",
|
| 3 |
+
"_diffusers_version": "0.32.0.dev0",
|
| 4 |
+
"scheduler": [
|
| 5 |
+
"diffusers",
|
| 6 |
+
"FlowMatchEulerDiscreteScheduler"
|
| 7 |
+
],
|
| 8 |
+
"text_encoder": [
|
| 9 |
+
"transformers",
|
| 10 |
+
"T5EncoderModel"
|
| 11 |
+
],
|
| 12 |
+
"tokenizer": [
|
| 13 |
+
"transformers",
|
| 14 |
+
"T5Tokenizer"
|
| 15 |
+
],
|
| 16 |
+
"transformer": [
|
| 17 |
+
"diffusers",
|
| 18 |
+
"LTXVideoTransformer3DModel"
|
| 19 |
+
],
|
| 20 |
+
"vae": [
|
| 21 |
+
"diffusers",
|
| 22 |
+
"AutoencoderKLLTXVideo"
|
| 23 |
+
]
|
| 24 |
+
}
|
transformer/transformer_config.json
ADDED
|
@@ -0,0 +1,19 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_class_name": "LTXVideoTransformer3DModel",
|
| 3 |
+
"_diffusers_version": "0.32.0.dev0",
|
| 4 |
+
"activation_fn": "gelu-approximate",
|
| 5 |
+
"attention_bias": true,
|
| 6 |
+
"attention_head_dim": 64,
|
| 7 |
+
"attention_out_bias": true,
|
| 8 |
+
"caption_channels": 4096,
|
| 9 |
+
"cross_attention_dim": 2048,
|
| 10 |
+
"in_channels": 128,
|
| 11 |
+
"norm_elementwise_affine": false,
|
| 12 |
+
"norm_eps": 1e-06,
|
| 13 |
+
"num_attention_heads": 32,
|
| 14 |
+
"num_layers": 28,
|
| 15 |
+
"out_channels": 128,
|
| 16 |
+
"patch_size": 1,
|
| 17 |
+
"patch_size_t": 1,
|
| 18 |
+
"qk_norm": "rms_norm_across_heads"
|
| 19 |
+
}
|
vae/config.json
ADDED
|
@@ -0,0 +1,65 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_class_name": "AutoencoderKLLTXVideo",
|
| 3 |
+
"_diffusers_version": "0.32.0.dev0",
|
| 4 |
+
"block_out_channels": [
|
| 5 |
+
128,
|
| 6 |
+
256,
|
| 7 |
+
512,
|
| 8 |
+
512
|
| 9 |
+
],
|
| 10 |
+
"decoder_block_out_channels": [
|
| 11 |
+
256,
|
| 12 |
+
512,
|
| 13 |
+
1024
|
| 14 |
+
],
|
| 15 |
+
"decoder_causal": false,
|
| 16 |
+
"decoder_inject_noise": [
|
| 17 |
+
true,
|
| 18 |
+
true,
|
| 19 |
+
true,
|
| 20 |
+
false
|
| 21 |
+
],
|
| 22 |
+
"decoder_layers_per_block": [
|
| 23 |
+
5,
|
| 24 |
+
6,
|
| 25 |
+
7,
|
| 26 |
+
8
|
| 27 |
+
],
|
| 28 |
+
"decoder_spatio_temporal_scaling": [
|
| 29 |
+
true,
|
| 30 |
+
true,
|
| 31 |
+
true
|
| 32 |
+
],
|
| 33 |
+
"encoder_causal": true,
|
| 34 |
+
"in_channels": 3,
|
| 35 |
+
"latent_channels": 128,
|
| 36 |
+
"layers_per_block": [
|
| 37 |
+
4,
|
| 38 |
+
3,
|
| 39 |
+
3,
|
| 40 |
+
3,
|
| 41 |
+
4
|
| 42 |
+
],
|
| 43 |
+
"out_channels": 3,
|
| 44 |
+
"patch_size": 4,
|
| 45 |
+
"patch_size_t": 1,
|
| 46 |
+
"resnet_norm_eps": 1e-06,
|
| 47 |
+
"scaling_factor": 1.0,
|
| 48 |
+
"spatio_temporal_scaling": [
|
| 49 |
+
true,
|
| 50 |
+
true,
|
| 51 |
+
true,
|
| 52 |
+
false
|
| 53 |
+
],
|
| 54 |
+
"timestep_conditioning": true,
|
| 55 |
+
"upsample_factor": [
|
| 56 |
+
2,
|
| 57 |
+
2,
|
| 58 |
+
2
|
| 59 |
+
],
|
| 60 |
+
"upsample_residual": [
|
| 61 |
+
true,
|
| 62 |
+
true,
|
| 63 |
+
true
|
| 64 |
+
]
|
| 65 |
+
}
|
vae/diffusion_pytorch_model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3a70d8d49fb2cc3698ffe9ed0e09fba5cd65d6a7d83fe89320ebc1d6fcc94536
|
| 3 |
+
size 1869989690
|