update

model_index.json

@@ -1,6 +1,7 @@
 {
-  "_class_name": "StableDiffusionPipeline",
-  "_diffusers_version": "0.2.2",
+  "_class_name": "TextToVideoSDPipeline",
+  "_diffusers_version": "0.21.0.dev0",
+  "_name_or_path": "gligen-1-4",
   "scheduler": [
     "diffusers",
     "PNDMScheduler"
@@ -14,11 +15,12 @@
     "CLIPTokenizer"
   ],
   "unet": [
-    "diffusers",
-    "UNet2DConditionModel"
+    "models.unet_3d_condition",
+    "UNet3DConditionModel"
   ],
   "vae": [
     "diffusers",
     "AutoencoderKL"
   ]
 }
+

unet/config.json

@@ -1,24 +1,24 @@
 {
-  "_class_name": "UNet2DConditionModel",
-  "_diffusers_version": "0.2.2",
+  "_class_name": "UNet3DConditionModel",
+  "_diffusers_version": "0.21.0.dev0",
+  "_name_or_path": "./",
   "act_fn": "silu",
-  "attention_head_dim": 8,
+  "attention_head_dim": 64,
+  "attention_type": "gated",
   "block_out_channels": [
     320,
     640,
     1280,
     1280
   ],
-  "center_input_sample": false,
   "cross_attention_dim": 768,
   "down_block_types": [
-    "CrossAttnDownBlock2D",
-    "CrossAttnDownBlock2D",
-    "CrossAttnDownBlock2D",
-    "DownBlock2D"
+    "CrossAttnDownBlock3D",
+    "CrossAttnDownBlock3D",
+    "CrossAttnDownBlock3D",
+    "DownBlock3D"
   ],
   "downsample_padding": 1,
-  "flip_sin_to_cos": true,
   "freq_shift": 0,
   "in_channels": 4,
   "layers_per_block": 2,
@@ -28,10 +28,10 @@
   "out_channels": 4,
   "sample_size": 64,
   "up_block_types": [
-    "UpBlock2D",
-    "CrossAttnUpBlock2D",
-    "CrossAttnUpBlock2D",
-    "CrossAttnUpBlock2D"
+    "UpBlock3D",
+    "CrossAttnUpBlock3D",
+    "CrossAttnUpBlock3D",
+    "CrossAttnUpBlock3D"
   ]
 }
 

unet/diffusion_pytorch_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:c7da0e21ba7ea50637bee26e81c220844defdf01aafca02b2c42ecdadb813de4
-size 3438354725
+oid sha256:268c57f44038f034daba3cc7519eec4b650031880d10828ca4e8fb65e78f093c
+size 3438318437