5 commits

Author	SHA1	Message	Date
matthewdeng	e3a096f412	[air] add bulk ingest benchmarks (#26618)	2022-07-15 22:01:23 -07:00
Richard Liaw	5ad4e75831	[air] Add initial benchmark section (#26608)	2022-07-15 15:33:48 -07:00
xwjiang2010	a241e6a0f5	[air] Add xgboost release test for silver tier(10-node case). (#26460) ... Co-authored-by: Antoni Baum <antoni.baum@protonmail.com> Co-authored-by: Richard Liaw <rliaw@berkeley.edu>	2022-07-15 13:21:10 -07:00
Kai Fricke	213a96e239	[air/benchmarks] Add distributed Tensorflow benchmarks (CPU only) (#26519) ... Following up from #26436, this PR adds a distributed benchmark test for Tensorflow FashionMNIST training. It compares training with Ray AIR with training with vanilla PyTorch. Signed-off-by: Kai Fricke <kai@anyscale.com>	2022-07-14 22:08:43 +01:00
Kai Fricke	cf75cf7232	[air] Add AIR distributed training benchmark for Torch FashionMNIST (#26436) ... This PR adds a distributed benchmark test for Pytorch MNIST training. It compares training with Ray AIR with training with vanilla PyTorch. In both cases, the same training loop is used. For Ray AIR, we use a TorchTrainer with 4 CPU workers. For vanilla PyTorch, we upload a training script and kick it off (using Ray tasks) in subprocesses on each node. In both cases, we collect the end to end runtime. Signed-off-by: Kai Fricke <kai@anyscale.com>	2022-07-13 10:53:24 +01:00