Upload 2 files

vae/config.json

@@ -2,38 +2,54 @@
   "_class_name": "AutoencoderKLLTX2Video",
   "_diffusers_version": "0.37.0.dev0",
   "block_out_channels": [
+    16,
+    16,
     32,
-    64
+    32
   ],
   "decoder_block_out_channels": [
+    8,
+    16,
     32
   ],
   "decoder_causal": false,
   "decoder_inject_noise": [
+    false,
+    false,
     false,
     false
   ],
   "decoder_layers_per_block": [
+    1,
+    1,
     1,
     1
   ],
   "decoder_spatial_padding_mode": "reflect",
   "decoder_spatio_temporal_scaling": [
+    true,
+    true,
     true
   ],
   "down_block_types": [
+    "LTX2VideoDownBlock3D",
+    "LTX2VideoDownBlock3D",
     "LTX2VideoDownBlock3D",
     "LTX2VideoDownBlock3D"
   ],
   "downsample_type": [
     "spatial",
-    "temporal"
+    "temporal",
+    "spatiotemporal",
+    "spatiotemporal"
   ],
   "encoder_causal": true,
   "encoder_spatial_padding_mode": "zeros",
   "in_channels": 3,
   "latent_channels": 4,
   "layers_per_block": [
+    1,
+    1,
     1,
     1,
     1
@@ -45,15 +61,21 @@
   "scaling_factor": 1.0,
   "spatial_compression_ratio": 32,
   "spatio_temporal_scaling": [
+    true,
+    true,
     true,
     true
   ],
   "temporal_compression_ratio": 8,
   "timestep_conditioning": false,
   "upsample_factor": [
+    2,
+    2,
     2
   ],
   "upsample_residual": [
+    true,
+    true,
     true
   ]
 }

vae/diffusion_pytorch_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:3e6671690f82856448442ec433c4037a5691d03c2985a06558b0d59c2b55f69c
-size 1052218
+oid sha256:0291ee2b34c08e6bd93aab7a4cbc63cc0e41dd10e7e721e4e1f6cbd8e574d292
+size 777842