Add files via upload

controlnet/README.md

@@ -0,0 +1,15 @@
+## Depth-Conditioned ControlNet based on Depth Anything V2
+
+We use [Diffusers](https://github.com/huggingface/diffusers/tree/main) to re-train a better depth-conditioned ControlNet based on our Depth Anything.
+
+Please download our [config file](./config.json) and [pre-trained weights](https://huggingface.co/MackinationsAi/Depth-Anything-V2_Safetensors/blob/main/depth_anything_v2_vitl.safetensors), then follow the [instructions](https://github.com/huggingface/diffusers/tree/main/examples/controlnet) in Diffusers for inference. 
+
+## Depth-to-Image Synthesis
+
+![demo2](../assets/controlnet_demo1.png)
+![demo1](../assets/controlnet_demo2.png)
+
+
+## Video Editing
+
+Please refer to our [project page](https://depth-anything-v2.github.io/). We use [MagicEdit](https://github.com/magic-research/magic-edit) to show demos of video editing based on depth information.

controlnet/config.json

@@ -0,0 +1,51 @@
+{
+  "_class_name": "ControlNetModel",
+  "_diffusers_version": "0.26.0.dev0",
+  "act_fn": "silu",
+  "addition_embed_type": null,
+  "addition_embed_type_num_heads": 64,
+  "addition_time_embed_dim": null,
+  "attention_head_dim": 8,
+  "block_out_channels": [
+    320,
+    640,
+    1280,
+    1280
+  ],
+  "class_embed_type": null,
+  "conditioning_channels": 3,
+  "conditioning_embedding_out_channels": [
+    16,
+    32,
+    96,
+    256
+  ],
+  "controlnet_conditioning_channel_order": "rgb",
+  "cross_attention_dim": 768,
+  "down_block_types": [
+    "CrossAttnDownBlock2D",
+    "CrossAttnDownBlock2D",
+    "CrossAttnDownBlock2D",
+    "DownBlock2D"
+  ],
+  "downsample_padding": 1,
+  "encoder_hid_dim": null,
+  "encoder_hid_dim_type": null,
+  "flip_sin_to_cos": true,
+  "freq_shift": 0,
+  "global_pool_conditions": false,
+  "in_channels": 4,
+  "layers_per_block": 2,
+  "mid_block_scale_factor": 1,
+  "mid_block_type": "UNetMidBlock2DCrossAttn",
+  "norm_eps": 1e-05,
+  "norm_num_groups": 32,
+  "num_attention_heads": null,
+  "num_class_embeds": null,
+  "only_cross_attention": false,
+  "projection_class_embeddings_input_dim": null,
+  "resnet_time_scale_shift": "default",
+  "transformer_layers_per_block": 1,
+  "upcast_attention": false,
+  "use_linear_projection": false
+}