Stacker Nodes
Stacker Nodes are a type of node that make it very easy to chain inputs together and then apply them in one operation.
- Stacker nodes store lists of data but do not process data other than doing data validation.
- The nodes usually look like tables or lists and have the same input and outputs allowing them to be chained together to combine the dataset into a single output.
- They are generally paired with an 'Apply' node that processes the items in the table or list.
Example Stacker Node
In this example the stacker node is used to configure the models that will be merged in the CR Apply Model Merge node.
- The stacker node is also used to configure the model and clip ratios for each model.
- On/Off switches are included to allow easy configuration.
-
They are very extensible. If you need more items in your stack then just chain another stacker node of the same type. For example you can chain three CR LoRA Stack nodes to hold a list of 9 LoRAs.
-
Some stacker nodes may include a switch attribute that allows you to turn each item On/Off. This is very useful for retaining configurations in your workflow, and for rapidly switching configurations.
-
Stacker nodes work well with schedulers.
-
Stackers can also store local attributes that apply to all item in the stacker node.
-
They are compact and help reduce complexity in workflows. For example the CR Multi-Control Stack can be used used to replace six separate nodes.
-
Stack nodes do not need to be located lose to the point where they are applied in a workflow. They can be located in a control panel and connected to the apply node via a single connector. This facilitates modularization of workflows.
More information on stacker nodes can be found in this CivitAI article:
https://civitai.com/articles/1997/comfyui-guide-to-stacker-nodes