diff --git a/nodes.py b/nodes.py index f9b91b597..a8d27732e 100644 --- a/nodes.py +++ b/nodes.py @@ -54,7 +54,7 @@ class CLIPTextEncode: CATEGORY = "conditioning" TOOLTIPS = { "input": { - "text": "The text to be encoded", + "text": "The text to be encoded.", "clip": "The CLIP model used for encoding the text.", }, "output": ("A conditioning containing the embedded text used to guide the diffusion model.",) @@ -270,6 +270,13 @@ class VAEDecode: FUNCTION = "decode" CATEGORY = "latent" + TOOLTIPS = { + "input": { + "samples": "The latent to be decoded.", + "vae": "The VAE model used for decoding the latent.", + }, + "output": ("The decoded image.",) + } def decode(self, vae, samples): return (vae.decode(samples["samples"]), ) @@ -517,6 +524,12 @@ class CheckpointLoaderSimple: FUNCTION = "load_checkpoint" CATEGORY = "loaders" + TOOLTIPS = { + "input": { + "ckpt_name": "The name of the checkpoint (model) to load.", + }, + "output": ("The model used for denoising latents.", "The CLIP model used for encoding text prompts.", "The VAE model used for encoding and decoding images to and from latent space.") + } def load_checkpoint(self, ckpt_name): ckpt_path = folder_paths.get_full_path("checkpoints", ckpt_name) @@ -1046,6 +1059,14 @@ class EmptyLatentImage: FUNCTION = "generate" CATEGORY = "latent" + TOOLTIPS = { + "input": { + "width": "The width of the latent images in pixels.", + "height": "The height of the latent images in pixels.", + "batch_size": "The number of latent images in the batch.", + }, + "output": ("The empty latent image batch.",) + } def generate(self, width, height, batch_size=1): latent = torch.zeros([batch_size, 4, height // 8, width // 8], device=self.device) @@ -1383,6 +1404,21 @@ class KSampler: FUNCTION = "sample" CATEGORY = "sampling" + TOOLTIPS = { + "input": { + "model": "The model used for denoising the input latent.", + "seed": "The random seed used for creating the noise.", + "steps": "The number of steps used in the denoising process.", + "cfg": "The Classifier-Free Guidance scale balances creativity and adherence to the prompt. Higher values result in images more closely matching the prompt however too high values will negatively impact quality.", + "sampler_name": "The algorithm used when sampling, this can affect the quality, speed, and style of the generated output.", + "scheduler": "The scheduler controls how noise is gradually removed to form the image.", + "positive": "The conditioning describing the attributes you want to include in the image.", + "negative": "The conditioning describing the attributes you do not want to include in the image.", + "latent_image": "The latent image to denoise.", + "denoise": "The amount of denoising applied, lower values will maintain the structure of the initial image allowing for image to image sampling.", + }, + "output": ("The denoised latent.",) + } def sample(self, model, seed, steps, cfg, sampler_name, scheduler, positive, negative, latent_image, denoise=1.0): return common_ksampler(model, seed, steps, cfg, sampler_name, scheduler, positive, negative, latent_image, denoise=denoise) @@ -1442,6 +1478,12 @@ class SaveImage: OUTPUT_NODE = True CATEGORY = "image" + TOOLTIPS = { + "input": { + "images": "The images to save.", + "filename_prefix": "The prefix for the file to save. This may include formatting information such as %date:yyyy-MM-dd% or %Empty Latent Image.width% to include values from nodes.", + } + } def save_images(self, images, filename_prefix="ComfyUI", prompt=None, extra_pnginfo=None): filename_prefix += self.prefix_append