ComfyUI Node
Release Memory
A ComfyUI node in Info-Prompt-Toolkit/Utility with 5 inputs and 1 output.
Release Memory
- passthrough
- then
◄generation_runtimetrue►
◄sam3_runtimetrue►
◄tagger_runtimetrue►
◄gc_cuda_cleanuptrue►
CategoryInfo-Prompt-Toolkit/Utility
Inputs (5)
| Name | Type | Default | Description |
|---|---|---|---|
| generation_runtime | BOOLEAN | true | Release generation runtime references, including sampler model, CLIP, and VAE runtimes |
| sam3_runtime | BOOLEAN | true | Release the SAM3 Prompt To Mask runtime cache |
| tagger_runtime | BOOLEAN | true | Release the PixAI Tagger and OppaiOracle Tagger runtime caches |
| gc_cuda_cleanup | BOOLEAN | true | Run Python garbage collection, then ask ComfyUI and torch to clear available CUDA cache memory |
| passthroughopt | COMFY_MATCHTYPE_V3 | Connect this input to run workflow memory release; the value is returned unchanged |
Outputs (1)
| Name | Type | Description |
|---|---|---|
| then | COMFY_MATCHTYPE_V3 | — |