Merge pull request #4 from akatz-ai/akatz-dev

Update with workflows, assets, and readme.

.github/workflows/publish.yml

@@ -0,0 +1,30 @@
+name: Publish to Comfy registry
+on:
+  workflow_dispatch:
+  push:
+    branches:
+      - main
+      - master
+    paths:
+      - "pyproject.toml"
+  pull_request:
+    branches:
+      - main
+    types:
+      - closed  # Only trigger when the pull request is closed, which includes when it is merged
+
+
+jobs:
+  publish-node:
+    name: Publish Custom Node to registry
+    runs-on: ubuntu-latest
+    # if this is a forked repository. Skipping the workflow.
+    if: github.event.repository.fork == false && github.event.pull_request.merged == true
+    steps:
+      - name: Check out code
+        uses: actions/checkout@v4
+      - name: Publish Custom Node
+        uses: Comfy-Org/publish-node-action@main
+        with:
+          ## Add your own personal access token to your Github Repository secrets and reference it here.
+          personal_access_token: ${{ secrets.REGISTRY_ACCESS_TOKEN }}

.gitignore

@@ -2,5 +2,4 @@ venv/
 **/*.pyc
 checkpoint/
 outputs/
-.vscode/
-assets/
+.vscode/

README.md

@@ -2,6 +2,20 @@
 
 **Animate portraits with an input video and a reference image using X-Portrait in ComfyUI.**
 
+[Original X-Portrait Repo](https://github.com/bytedance/X-Portrait)
+
+<img src="assets/x-portrait-nodes.png" alt="x-portrait nodes" width="400"/>
+
+- source_image: the reference image for generation, should be square and max 512x512.
+- driving_video: the driving video containing a face, should match the resolution of the source_image and max 512x512.
+- seed: controls current generation seed
+- ddim_steps: how many samples (steps) for generating the output. Diminishing returns after 15.
+- best_frame: which frame of the driving_video most closely matches the source_image.
+
+## ⭐ Example Workflow (in examples/)
+
+![Example workflow 1](assets/x-portrait-workflow.png)
+
 ## 🔧 Installation and Usage
 
 1. ComfyUI Manager:

core/test_xportrait.py

@@ -375,15 +375,8 @@ def visualize_mm(args, name, batch_data, infer_model, nSample, local_image_dir,
     output_img = torch.permute(output_img, [0, 2, 3, 1])
 
     output_img = output_img.data.cpu().numpy()
-    print(f"Output image shape: {output_img.shape}")
-    print(f"Output image data peek: {output_img[:5, :5, :]}")
-    # output_img = img_as_ubyte(output_img)
     output_img = torch.from_numpy(output_img)
-    print(f"Output image shape: {output_img.shape}")
     output_img = output_img[:,:,:512]
-    print(f"Output image shape: {output_img.shape}")
-    # print(f"Output image data peek: {output_img[:5, :5, :]}")
-    # imageio.mimsave(output_path, output_img[:,:,:512], fps=batch_data['fps'], quality=10, pixelformat='yuv420p', codec='libx264')
 
     # After processing, clear any large temporary variables
     del cond_img_cat, pose_cond_list, local_pose_cond_list