Add files via upload

- Adjust resolution when using videosrc
- Fix nodes for current comfyui version
This commit is contained in:
sakura1bgx
2024-10-30 22:54:26 +09:00
committed by GitHub
parent 53a6017ca2
commit 8b9902e1be
+18 -14
View File
@@ -1318,7 +1318,7 @@ class FlipStreamSource:
return {
"required": {
"vae": ("VAE",),
"default_width": ("INT", {"default": 512}),
"width": ("INT", {"default": 512}),
"height": ("INT", {"default": 512}),
},
}
@@ -1331,11 +1331,11 @@ class FlipStreamSource:
CATEGORY = "FlipStreamViewer"
@classmethod
def IS_CHANGED(cls, vae, default_width, height):
def IS_CHANGED(cls, vae, width, height):
state["height"] = height
return hash((default_width, height, param["frames"], param["videosrc"], param["videofst"], param["videoskp"], param["videostr"]))
return hash((width, height, param["frames"], param["videosrc"], param["videofst"], param["videoskp"], param["videostr"]))
def source(self, vae, default_width, height):
def source(self, vae, width, height):
videopath = str(Path("videosrc", param["videosrc"])) if param["videosrc"] else ""
frames = param["frames"]
image = None
@@ -1345,12 +1345,16 @@ class FlipStreamSource:
buf, _ = load_video(videopath, height, param["videofst"], param["videoskp"], frames)
if buf:
buf = [np.array(cv2.cvtColor(frame, cv2.COLOR_BGR2RGB), dtype=np.float32) / 255 for frame in buf]
image = torch.from_numpy(np.fromiter(buf, np.dtype((np.float32, buf[0].shape))))
image = torch.zeros([frames, height, width, 3])
image2 = torch.from_numpy(np.fromiter(buf, np.dtype((np.float32, buf[0].shape))))
x2 = image.shape[2] // 2
w2 = image2.shape[2] // 2
image[:, :, x2 - w2: x2 + w2] = image2
latent = vae.encode(image)
bypass = True
if image is None:
image = torch.zeros([frames, height, default_width, 3])
latent = torch.zeros([frames, 4, height // 8, default_width // 8], device=self.device)
image = torch.zeros([frames, height, width, 3])
latent = torch.zeros([frames, 4, height // 8, width // 8], device=self.device)
frames, h, w, _ = image.shape
images = [image[i:i + 1, ...] for i in range(image.shape[0])]
latents = [{"samples": latent[i:i + 1, ...]} for i in range(latent.shape[0])]
@@ -1387,10 +1391,10 @@ class FlipStreamPrompt:
def INPUT_TYPES(s):
return {
"required": {
"prompt": ("STRING", {"multiline": True}),
"batchPrompt": ("STRING", {"multiline": True}),
"appPrompt": ("STRING", {"multiline": True}),
"frames": ("INT",),
"prompt": ("STRING", {"multiline": True, "dynamicPrompts": True}),
"batchPrompt": ("STRING", {"multiline": True, "dynamicPrompts": True}),
"appPrompt": ("STRING", {"multiline": True, "dynamicPrompts": True}),
"frames": ("INT", {"default": 0}),
},
}
@@ -1417,7 +1421,7 @@ class FlipStreamOption:
def INPUT_TYPES(s):
return {
"required": {
"mode": ("STRING",),
"mode": ("STRING", {"default": ""}),
}
}
@@ -1464,8 +1468,8 @@ class FlipStreamViewer:
return {
"required": {
"tensor": ("IMAGE",),
"allowip": ("STRING",),
"wd14exc": ("STRING",),
"allowip": ("STRING", {"default": ""}),
"wd14exc": ("STRING", {"default": ""}),
"idle": ("FLOAT", {"default": 1.0}),
},
}