Skip to content

[Feature] Add UCXX transport for fast node-to-node TensorDict transfers #2014

[Feature] Add UCXX transport for fast node-to-node TensorDict transfers

[Feature] Add UCXX transport for fast node-to-node TensorDict transfers #2014

Triggered via pull request March 2, 2026 18:11
@vmoensvmoens
synchronize #1611
ucxx
Status Cancelled
Total duration 3m 25s
Artifacts
generate-matrix  /  generate
4s
generate-matrix / generate
Matrix: pytorch/tensordict
Fit to window
Zoom out
Zoom in

Annotations

4 errors
pytorch/tensordict (pytorch/tensordict, tensordict) / build-wheel-py3_10-cpu-aarch64
Canceling since a higher priority waiting request for Build Aarch64 Linux Wheels-1611-pytorch/tensordict-false exists
pytorch/tensordict (pytorch/tensordict, tensordict) / upload / upload-wheel-py3_10-cpu-aarch64
Unable to download artifact(s): Artifact not found for name: pytorch_tensordict__3.10_cpu_aarch64 Please ensure that your artifact is not expired and the artifact was uploaded using a compatible version of toolkit/upload-artifact. For more information, visit the GitHub Artifacts FAQ: https://github.com/actions/toolkit/blob/main/packages/artifact/docs/faq.md
Build Aarch64 Linux Wheels
Canceling since a higher priority waiting request for Build Aarch64 Linux Wheels-1611-pytorch/tensordict-false exists