Run benchmarks with shuffle

AB_environments/AB_sample.conda.yaml

@@ -55,6 +55,6 @@ dependencies:
     # conda. You may alternatively point to your own git fork (but make sure you
     # sync'ed tags!)
     # Read README.md for troubleshooting.
-    - git+https://github.com/phofl/dask@no_culling
+    - git+https://github.com/phofl/dask@array-shuffle-take
     - git+https://github.com/dask/distributed
     - git+https://github.com/dask/dask-expr

AB_environments/config.yaml

@@ -8,7 +8,7 @@ repeat: 4
 # Set to true to automatically create a verbatim copy of AB_baseline and then compare
 # the two in the A/B tests. Set to false to save some money if you are already confident
 # that the 'repeat' setting is high enough.
-test_null_hypothesis: true
+test_null_hypothesis: false
 
 # Test directories, test files, or individual tests to run.
 targets:
@@ -51,7 +51,7 @@ h2o_datasets:
 # never actually reached with real paying users.
 max_parallel:
   # Number of parallel A/B test jobs per branch.
-  ci_jobs: 5
+  ci_jobs: 8
   # Number of parallel test_*.py modules per A/B test job.
   # Each module typically spawns one Coiled cluster at a time.
   # Set to 1 to disable pytest-xdist.