2022-08-12 03:33:09 +00:00
|
|
|
import pytest
|
2022-10-19 04:53:06 +00:00
|
|
|
import torch
|
2022-08-12 03:33:09 +00:00
|
|
|
import torch.nn.functional as F
|
2022-10-19 04:53:06 +00:00
|
|
|
|
|
|
|
import colossalai
|
2022-08-12 03:33:09 +00:00
|
|
|
from colossalai.device.device_mesh import DeviceMesh
|
|
|
|
from colossalai.nn._ops._utils import gather_forward_split_backward
|
2022-10-19 04:53:06 +00:00
|
|
|
from colossalai.tensor import ColoParameter, ColoTensor, ProcessGroup
|
|
|
|
from colossalai.tensor.sharding_spec import ShardingSpec
|
2023-04-06 06:51:35 +00:00
|
|
|
from colossalai.testing import rerun_if_address_is_in_use, spawn
|
2022-08-12 03:33:09 +00:00
|
|
|
|
|
|
|
|
|
|
|
def run_dist(rank, world_size, port):
|
|
|
|
config = {}
|
|
|
|
colossalai.launch(config=config, rank=rank, world_size=world_size, host='localhost', port=port, backend='nccl')
|
|
|
|
|
|
|
|
# create mlp vars
|
2022-10-19 04:53:06 +00:00
|
|
|
x = ColoTensor.from_torch_tensor(torch.rand(4, 4, 8, requires_grad=True)).cuda()
|
2022-08-12 03:33:09 +00:00
|
|
|
w = ColoParameter.from_torch_tensor(torch.rand(16, 8, requires_grad=True)).cuda()
|
|
|
|
b = ColoParameter.from_torch_tensor(torch.rand(16, requires_grad=True)).cuda()
|
|
|
|
|
|
|
|
# run normal forward
|
|
|
|
out = F.linear(x, w, b)
|
|
|
|
|
|
|
|
# create mesh meta
|
|
|
|
# the mesh is in the following topo
|
|
|
|
# [[0, 1],
|
|
|
|
# [2, 3]]
|
2023-06-09 01:41:27 +00:00
|
|
|
physical_mesh_id = torch.arange(0, 4).reshape(2, 2)
|
2022-08-12 03:33:09 +00:00
|
|
|
mesh_shape = (2, 2)
|
|
|
|
device_mesh = DeviceMesh(physical_mesh_id, mesh_shape)
|
|
|
|
row_id = rank // 2
|
|
|
|
column_id = rank % 2
|
|
|
|
|
|
|
|
# create pg
|
|
|
|
row_process_group = None
|
|
|
|
col_process_group = None
|
|
|
|
row_to_ranks = {0: [0, 1], 1: [2, 3]}
|
|
|
|
col_to_ranks = {0: [0, 2], 1: [1, 3]}
|
|
|
|
|
|
|
|
for idx in range(2):
|
|
|
|
# row ranks
|
|
|
|
row_ranks = row_to_ranks[idx]
|
|
|
|
row_pg = ProcessGroup(ranks=row_ranks, tp_degree=2)
|
|
|
|
|
|
|
|
# col ranks
|
|
|
|
col_ranks = col_to_ranks[idx]
|
|
|
|
col_pg = ProcessGroup(ranks=col_ranks, tp_degree=2)
|
|
|
|
|
|
|
|
if rank in row_ranks:
|
|
|
|
row_process_group = row_pg
|
|
|
|
|
|
|
|
if rank in col_ranks:
|
|
|
|
col_process_group = col_pg
|
|
|
|
|
|
|
|
########################
|
|
|
|
# RRR x RS0 -> RRS0 #
|
|
|
|
########################
|
|
|
|
# w will be transposed in F.linear
|
|
|
|
x_replica = x.detach().clone()
|
|
|
|
w_shard = torch.chunk(w.detach().clone(), chunks=2, dim=0)[row_id]
|
|
|
|
b_shard = torch.chunk(b.detach().clone(), chunks=2, dim=0)[row_id]
|
|
|
|
|
|
|
|
# adding sharding spec
|
|
|
|
x_replica.sharding_spec = ShardingSpec(device_mesh, x.shape, dim_partition_dict={})
|
|
|
|
w_shard.sharding_spec = ShardingSpec(device_mesh, w.shape, dim_partition_dict={0: [0]})
|
|
|
|
b_shard.sharding_spec = ShardingSpec(device_mesh, b.shape, dim_partition_dict={0: [0]})
|
|
|
|
|
|
|
|
# check sharding spec
|
|
|
|
assert str(x_replica.sharding_spec.sharding_sequence) == "[R, R, R]"
|
|
|
|
assert str(w_shard.sharding_spec.sharding_sequence) == "[S0, R]"
|
|
|
|
assert str(b_shard.sharding_spec.sharding_sequence) == "[S0]"
|
|
|
|
|
|
|
|
w_shard.pg_axis0 = col_process_group
|
|
|
|
w_shard.pg_axis1 = row_process_group
|
|
|
|
|
|
|
|
out_shard = F.linear(x_replica, w_shard, b_shard)
|
|
|
|
assert str(out_shard.sharding_spec.sharding_sequence) == "[R, R, S0]"
|
|
|
|
|
|
|
|
# each row only has a mini-batch
|
|
|
|
expected_out_shard = torch.chunk(out, chunks=2, dim=2)[row_id]
|
|
|
|
assert torch.allclose(out_shard, expected_out_shard)
|
|
|
|
|
|
|
|
########################
|
|
|
|
# S0RR x RS1 -> S0RS1 #
|
|
|
|
########################
|
|
|
|
# w will be transposed in F.linear
|
|
|
|
x_shard = torch.chunk(x.detach().clone(), chunks=2, dim=0)[row_id]
|
|
|
|
w_shard = torch.chunk(w.detach().clone(), chunks=2, dim=0)[column_id]
|
|
|
|
b_shard = torch.chunk(b.detach().clone(), chunks=2, dim=0)[column_id]
|
|
|
|
|
|
|
|
# adding sharding spec
|
|
|
|
x_shard.sharding_spec = ShardingSpec(device_mesh, x.shape, dim_partition_dict={0: [0]})
|
|
|
|
w_shard.sharding_spec = ShardingSpec(device_mesh, w.shape, dim_partition_dict={0: [1]})
|
|
|
|
b_shard.sharding_spec = ShardingSpec(device_mesh, b.shape, dim_partition_dict={0: [1]})
|
|
|
|
|
|
|
|
# check sharding spec
|
|
|
|
assert str(x_shard.sharding_spec.sharding_sequence) == "[S0, R, R]"
|
|
|
|
assert str(w_shard.sharding_spec.sharding_sequence) == "[S1, R]"
|
|
|
|
assert str(b_shard.sharding_spec.sharding_sequence) == "[S1]"
|
|
|
|
|
|
|
|
w_shard.pg_axis0 = col_process_group
|
|
|
|
w_shard.pg_axis1 = row_process_group
|
|
|
|
|
|
|
|
out_shard = F.linear(x_shard, w_shard, b_shard)
|
|
|
|
|
|
|
|
# each row only has a mini-batch
|
|
|
|
expected_out_shard = torch.chunk(out, chunks=2, dim=0)[row_id]
|
|
|
|
expected_out_shard = torch.chunk(expected_out_shard, chunks=2, dim=2)[column_id]
|
|
|
|
assert torch.allclose(out_shard, expected_out_shard)
|
|
|
|
|
|
|
|
########################
|
|
|
|
# S0RS1 x S1R -> S0RR #
|
|
|
|
########################
|
|
|
|
# w will be transposed in F.linear
|
|
|
|
x_shard = torch.chunk(x.clone(), chunks=2, dim=0)[row_id]
|
|
|
|
x_shard = torch.chunk(x_shard, chunks=2, dim=2)[column_id]
|
|
|
|
w_shard = torch.chunk(w.clone(), chunks=2, dim=1)[column_id]
|
|
|
|
b_replica = b.clone()
|
|
|
|
|
|
|
|
# adding sharding spec
|
|
|
|
x_shard.sharding_spec = ShardingSpec(device_mesh, x.shape, dim_partition_dict={0: [0], 2: [1]})
|
|
|
|
w_shard.sharding_spec = ShardingSpec(device_mesh, w.shape, dim_partition_dict={1: [1]})
|
|
|
|
b_replica.sharding_spec = ShardingSpec(device_mesh, b.shape, dim_partition_dict={})
|
|
|
|
|
|
|
|
# check sharding spec
|
|
|
|
assert str(x_shard.sharding_spec.sharding_sequence) == "[S0, R, S1]"
|
|
|
|
assert str(w_shard.sharding_spec.sharding_sequence) == "[R, S1]"
|
|
|
|
assert str(b_replica.sharding_spec.sharding_sequence) == "[R]"
|
|
|
|
|
|
|
|
w_shard.pg_axis0 = col_process_group
|
|
|
|
w_shard.pg_axis1 = row_process_group
|
|
|
|
|
|
|
|
out_shard = F.linear(x_shard, w_shard, b_replica)
|
|
|
|
|
|
|
|
# each row only has a mini-batch
|
|
|
|
expected_out_shard = torch.chunk(out, chunks=2, dim=0)[row_id]
|
|
|
|
assert torch.allclose(out_shard, expected_out_shard)
|
|
|
|
|
|
|
|
########################
|
|
|
|
# RRS0 x S0R -> RRR #
|
|
|
|
########################
|
|
|
|
# w will be transposed in F.linear
|
|
|
|
x_shard = torch.chunk(x.clone(), chunks=2, dim=2)[row_id]
|
|
|
|
w_shard = torch.chunk(w.clone(), chunks=2, dim=1)[row_id]
|
|
|
|
b_replica = b.clone()
|
|
|
|
|
|
|
|
# adding sharding spec
|
|
|
|
x_shard.sharding_spec = ShardingSpec(device_mesh, x.shape, dim_partition_dict={2: [0]})
|
|
|
|
w_shard.sharding_spec = ShardingSpec(device_mesh, w.shape, dim_partition_dict={1: [0]})
|
|
|
|
b_replica.sharding_spec = ShardingSpec(device_mesh, b.shape, dim_partition_dict={})
|
|
|
|
|
|
|
|
# check sharding spec
|
|
|
|
assert str(x_shard.sharding_spec.sharding_sequence) == "[R, R, S0]"
|
|
|
|
assert str(w_shard.sharding_spec.sharding_sequence) == "[R, S0]"
|
|
|
|
assert str(b_replica.sharding_spec.sharding_sequence) == "[R]"
|
|
|
|
|
|
|
|
w_shard.pg_axis0 = col_process_group
|
|
|
|
w_shard.pg_axis1 = row_process_group
|
|
|
|
|
|
|
|
out_shard = F.linear(x_shard, w_shard, b_replica)
|
|
|
|
|
|
|
|
# each row only has a mini-batch
|
|
|
|
expected_out_shard = out
|
|
|
|
assert torch.allclose(out_shard, expected_out_shard)
|
|
|
|
|
|
|
|
########################
|
|
|
|
# RS0S1 x S1R -> RS0R #
|
|
|
|
########################
|
|
|
|
# w will be transposed in F.linear
|
|
|
|
x_shard = torch.chunk(x.clone(), chunks=2, dim=1)[row_id]
|
|
|
|
x_shard = torch.chunk(x_shard, chunks=2, dim=2)[column_id]
|
|
|
|
w_shard = torch.chunk(w.clone(), chunks=2, dim=1)[column_id]
|
|
|
|
b_replica = b.clone()
|
|
|
|
|
|
|
|
# adding sharding spec
|
|
|
|
x_shard.sharding_spec = ShardingSpec(device_mesh, x.shape, dim_partition_dict={1: [0], 2: [1]})
|
|
|
|
w_shard.sharding_spec = ShardingSpec(device_mesh, w.shape, dim_partition_dict={1: [1]})
|
|
|
|
b_replica.sharding_spec = ShardingSpec(device_mesh, b.shape, dim_partition_dict={})
|
|
|
|
|
|
|
|
# check sharding spec
|
|
|
|
assert str(x_shard.sharding_spec.sharding_sequence) == "[R, S0, S1]"
|
|
|
|
assert str(w_shard.sharding_spec.sharding_sequence) == "[R, S1]"
|
|
|
|
assert str(b_replica.sharding_spec.sharding_sequence) == "[R]"
|
|
|
|
|
|
|
|
w_shard.pg_axis0 = col_process_group
|
|
|
|
w_shard.pg_axis1 = row_process_group
|
|
|
|
|
|
|
|
out_shard = F.linear(x_shard, w_shard, b_replica)
|
|
|
|
|
|
|
|
# each row only has a mini-batch
|
|
|
|
expected_out_shard = torch.chunk(out, chunks=2, dim=1)[row_id]
|
|
|
|
assert torch.allclose(out_shard, expected_out_shard)
|
|
|
|
|
|
|
|
########################
|
|
|
|
# RRS0 x S0S1 -> RRS1 #
|
|
|
|
########################
|
|
|
|
# w will be transposed in F.linear
|
|
|
|
x_shard = torch.chunk(x.clone(), chunks=2, dim=2)[row_id]
|
|
|
|
w_shard = torch.chunk(w.clone(), chunks=2, dim=1)[row_id]
|
|
|
|
w_shard = torch.chunk(w_shard, chunks=2, dim=0)[column_id]
|
|
|
|
b_shard = torch.chunk(b.clone(), chunks=2, dim=0)[column_id]
|
|
|
|
|
|
|
|
# adding sharding spec
|
|
|
|
x_shard.sharding_spec = ShardingSpec(device_mesh, x.shape, dim_partition_dict={2: [0]})
|
|
|
|
w_shard.sharding_spec = ShardingSpec(device_mesh, w.shape, dim_partition_dict={0: [1], 1: [0]})
|
|
|
|
b_shard.sharding_spec = ShardingSpec(device_mesh, b.shape, dim_partition_dict={0: [1]})
|
|
|
|
|
|
|
|
# check sharding spec
|
|
|
|
assert str(x_shard.sharding_spec.sharding_sequence) == "[R, R, S0]"
|
|
|
|
assert str(w_shard.sharding_spec.sharding_sequence) == "[S1, S0]"
|
|
|
|
assert str(b_shard.sharding_spec.sharding_sequence) == "[S1]"
|
|
|
|
|
|
|
|
w_shard.pg_axis0 = col_process_group
|
|
|
|
w_shard.pg_axis1 = row_process_group
|
|
|
|
|
|
|
|
out_shard = F.linear(x_shard, w_shard, b_shard)
|
|
|
|
|
|
|
|
# each row only has a mini-batch
|
|
|
|
expected_out_shard = torch.chunk(out, chunks=2, dim=2)[column_id]
|
|
|
|
assert torch.allclose(out_shard, expected_out_shard)
|
|
|
|
|
|
|
|
|
|
|
|
@pytest.mark.dist
|
|
|
|
@pytest.mark.parametrize('world_size', [4])
|
|
|
|
@rerun_if_address_is_in_use()
|
|
|
|
def test_sharded_mlp(world_size):
|
2023-04-06 06:51:35 +00:00
|
|
|
spawn(run_dist, world_size)
|
2022-08-12 03:33:09 +00:00
|
|
|
|
|
|
|
|
|
|
|
if __name__ == '__main__':
|
|
|
|
test_sharded_mlp(4)
|