Skip to main content

End Recording CUDAMemory History

KJNodes/experimental
EndRecordCUDAMemoryHistory

Records CUDA memory allocation history between start and end, saves to a file that can be analyzed here: https://docs.pytorch.org/memory_viz or with VisualizeCUDAMemoryHistory node

Example

JSON Example
{
  "class_type": "EndRecordCUDAMemoryHistory",
  "inputs": {
    "input": [
      "node_id",
      0
    ],
    "output_path": "https://example.com/path/to/file.bin"
  }
}

This example shows required inputs only. Connection values like ["node_id", 0] should reference actual node IDs from your workflow.

Inputs

NameTypeStatusConstraintsDefault
input*required--
output_pathSTRINGURL: Filerequired-"comfy_cuda_memory_history"

Outputs

IndexNameTypeIs ListConnection Reference
0input*No["{node_id}", 0]
1output_pathSTRINGNo["{node_id}", 1]
How to connect to these outputs

To connect another node's input to an output from this node, use the connection reference format:

["node_id", output_index]

Where node_id is the ID of this EndRecordCUDAMemoryHistory node in your workflow, and output_index is the index from the table above.

Example

If this node has ID "5" in your workflow:

  • input (*): ["5", 0]
  • output_path (STRING): ["5", 1]
Was this page helpful?