Correct UNet model

model_index.json

@@ -15,7 +15,7 @@
   ],
   "unet": [
     "diffusers",
-    "UNet2DModel"
+    "UNet2DConditionModel"
   ],
   "vae": [
     "diffusers",

unet/config.json

@@ -1,8 +1,7 @@
 {
-  "_class_name": "UNet2DModel",
+  "_class_name": "UNet2DConditionModel",
   "_diffusers_version": "0.13.0.dev0",
   "act_fn": "silu",
-  "add_attention": true,
   "attention_head_dim": 8,
   "block_out_channels": [
     128,
@@ -12,6 +11,8 @@
   ],
   "center_input_sample": false,
   "class_embed_type": null,
+  "conv_in_kernel": 3,
+  "conv_out_kernel": 3,
   "cross_attention_dim": true,
   "down_block_types": [
     "DownBlock2D",
@@ -20,6 +21,7 @@
     "CrossAttnDownBlock2D"
   ],
   "downsample_padding": 1,
+  "dual_cross_attention": false,
   "extra_film_condition_dim": 512,
   "extra_film_use_concat": true,
   "flip_sin_to_cos": true,
@@ -27,17 +29,23 @@
   "in_channels": 8,
   "layers_per_block": 2,
   "mid_block_scale_factor": 1,
+  "mid_block_type": "UNetMidBlock2DCrossAttn",
   "norm_eps": 1e-05,
   "norm_num_groups": 32,
   "num_class_embeds": null,
+  "only_cross_attention": false,
   "out_channels": 8,
   "resnet_time_scale_shift": "default",
   "sample_size": 128,
+  "time_cond_proj_dim": null,
   "time_embedding_type": "positional",
+  "timestep_post_act": null,
   "up_block_types": [
     "CrossAttnUpBlock2D",
     "CrossAttnUpBlock2D",
     "CrossAttnUpBlock2D",
     "UpBlock2D"
-  ]
+  ],
+  "upcast_attention": false,
+  "use_linear_projection": false
 }

unet/diffusion_pytorch_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:51f71fa4e1233882e2a5cd2cd7ed4cf64d11d59a6c8c0b7a09358c58f36ef9f3
+oid sha256:5bf3ccd62553e214c86a63d7808d21923f17a0c8411e97d8cdfbbb8bc44814d1
 size 740429981