Skip to content
hpu-gaudi2

Enabled configurable auto Tensor Parallelism (TP) for the inference of diverse models #1520

Sign in to view logs

Enabled configurable auto Tensor Parallelism (TP) for the inference of diverse models

Enabled configurable auto Tensor Parallelism (TP) for the inference of diverse models #1520

Re-run triggered January 21, 2025 03:34
@delockdelock
#6553
gyou2021:configurable_autoTP
Status Success
Total duration 1h 48m 52s
Artifacts –

hpu-gaudi2.yml

on: pull_request
unit-tests
59m 9s
unit-tests
Fit to window
Zoom out
Zoom in