github/ColossalAI
1
0
Fork 0
mirror of https://github.com/hpcaitech/ColossalAI.git synced 2025-11-02 14:59:32 +00:00
Code Issues Packages Projects Releases Wiki Activity

[test] fixed rerun_on_exception and adapted test cases (#487)

Browse Source
This commit is contained in:
Frank Lee
2022-03-25 17:25:12 +08:00
committed by GitHub
parent 4d322b79da
commit 3601b2bad0
31 changed files with 143 additions and 135 deletions
Download Patch File Download Diff File Expand all files Collapse all files

3
tests/test_zero_data_parallel/test_shard_param.py
View File

@@ -10,6 +10,7 @@ from colossalai.utils import free_port
from colossalai.zero.shard_utils import (BucketTensorShardStrategy, TensorShardStrategy)
from colossalai.zero.sharded_param import ShardedTensor
from colossalai.zero.sharded_param.sharded_param import ShardedParamV2
from colossalai.testing import rerun_on_exception
from tests.test_zero_data_parallel.common import CONFIG, allclose
@@ -35,6 +36,7 @@ def _run_shard_tensor(rank, world_size, port):
@pytest.mark.dist
@pytest.mark.parametrize("world_size", [1, 2])
@rerun_on_exception(exception_type=mp.ProcessRaisedException, pattern=".*Address already in use.*")
def test_shard_tensor(world_size):
run_func = partial(_run_shard_tensor, world_size=world_size, port=free_port())
mp.spawn(run_func, nprocs=world_size)
@@ -55,6 +57,7 @@ def _run_shard_param_v2(rank, world_size, port):
@pytest.mark.dist
@pytest.mark.parametrize("world_size", [1, 2])
@rerun_on_exception(exception_type=mp.ProcessRaisedException, pattern=".*Address already in use.*")
def test_shard_param_v2(world_size):
run_func = partial(_run_shard_param_v2, world_size=world_size, port=free_port())
mp.spawn(run_func, nprocs=world_size)