Upload folder using huggingface_hub

config.json

@@ -6,19 +6,24 @@
   "addition_embed_type": null,
   "addition_embed_type_num_heads": 64,
   "addition_time_embed_dim": null,
-  "attention_head_dim": 8,
+  "attention_head_dim": [
+    10,
+    16,
+    20,
+    20
+  ],
   "attention_type": "default",
   "block_out_channels": [
-    320,
     640,
+    1024,
     1280,
     1280
   ],
   "center_input_sample": false,
   "class_embed_type": null,
   "class_embeddings_concat": false,
-  "conv_in_kernel": 3,
-  "conv_out_kernel": 3,
+  "conv_in_kernel": 1,
+  "conv_out_kernel": 1,
   "cross_attention_dim": 768,
   "cross_attention_norm": null,
   "down_block_types": [
@@ -35,7 +40,7 @@
   "flip_sin_to_cos": true,
   "freq_shift": 0,
   "in_channels": 128,
-  "layers_per_block": 2,
+  "layers_per_block": 1,
   "mid_block_only_cross_attention": null,
   "mid_block_scale_factor": 1.0,
   "mid_block_type": "UNetMidBlock2DCrossAttn",
@@ -57,10 +62,10 @@
   "time_embedding_type": "positional",
   "timestep_post_act": null,
   "transformer_layers_per_block": [
-    2,
-    2,
-    3,
-    3
+    0,
+    6,
+    6,
+    6
   ],
   "up_block_types": [
     "UpBlock2D",

diffusion_pytorch_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:d3cc9613b04319a9725e28e725fdeb803c3d34cf078329214b4b2184c646186c
-size 5166216656
+oid sha256:e9f7c10295759b2616d71a79a6f00998f5b3139e09d8ad1775e70adb35f892ee
+size 6689101296