mirror of https://github.com/hpcaitech/ColossalAI
[Tensor] add hybrid device demo and fix bugs (#1059)
parent
b167258b6a
commit
df9dcbbff6
|
@ -51,11 +51,17 @@ class ColoDDP(torch.nn.Module):
|
||||||
free_storage(empty_grad)
|
free_storage(empty_grad)
|
||||||
if self.dp_world_size > 1:
|
if self.dp_world_size > 1:
|
||||||
grad = grad / self.dp_world_size
|
grad = grad / self.dp_world_size
|
||||||
|
if grad.device.type != "cpu":
|
||||||
self.comm_stream.wait_stream(torch.cuda.current_stream())
|
self.comm_stream.wait_stream(torch.cuda.current_stream())
|
||||||
with torch.cuda.stream(self.comm_stream):
|
with torch.cuda.stream(self.comm_stream):
|
||||||
dist.all_reduce(grad, group=gpc.get_group(ParallelMode.DATA))
|
group = gpc.get_group(ParallelMode.DATA)
|
||||||
|
dist.all_reduce(grad, group=group)
|
||||||
ColoDDP._save_grad(p, grad)
|
ColoDDP._save_grad(p, grad)
|
||||||
grad.record_stream(self.comm_stream)
|
grad.record_stream(self.comm_stream)
|
||||||
|
else:
|
||||||
|
group = gpc.get_cpu_group(ParallelMode.DATA)
|
||||||
|
dist.all_reduce(grad, group=group)
|
||||||
|
ColoDDP._save_grad(p, grad)
|
||||||
else:
|
else:
|
||||||
ColoDDP._save_grad(p, grad)
|
ColoDDP._save_grad(p, grad)
|
||||||
return empty_grad
|
return empty_grad
|
||||||
|
|
|
@ -12,12 +12,16 @@ def register_colo_module(module_type: type, colo_module: ColoModule):
|
||||||
|
|
||||||
def is_colo_module(module: torch.nn.Module):
|
def is_colo_module(module: torch.nn.Module):
|
||||||
global _COLOSSAL_MODULES
|
global _COLOSSAL_MODULES
|
||||||
return type(module) in _COLOSSAL_MODULES
|
for module_type in _COLOSSAL_MODULES.keys():
|
||||||
|
if isinstance(type(module), module_type):
|
||||||
|
return True
|
||||||
|
return False
|
||||||
|
|
||||||
def get_colo_module(module: torch.nn.Module):
|
def get_colo_module(module: torch.nn.Module):
|
||||||
global _COLOSSAL_MODULES
|
global _COLOSSAL_MODULES
|
||||||
if is_colo_module(module):
|
if is_colo_module(module):
|
||||||
colo_module = _COLOSSAL_MODULES[type(module)]
|
for module_type, colo_module in _COLOSSAL_MODULES.items():
|
||||||
|
if isinstance(type(module), module_type):
|
||||||
return colo_module
|
return colo_module
|
||||||
else:
|
else:
|
||||||
return None
|
return None
|
||||||
|
|
|
@ -92,4 +92,5 @@ class ColoInitContext(InsertPostInitMethodToModuleSubClasses):
|
||||||
setattr(submodule, param_name, colo_param)
|
setattr(submodule, param_name, colo_param)
|
||||||
colo_param.shared_param_modules.append(submodule)
|
colo_param.shared_param_modules.append(submodule)
|
||||||
|
|
||||||
|
module.to(self._device)
|
||||||
ColoModulize(module)
|
ColoModulize(module)
|
||||||
|
|
|
@ -101,4 +101,4 @@ def test_gpt(world_size, use_ddp):
|
||||||
|
|
||||||
|
|
||||||
if __name__ == '__main__':
|
if __name__ == '__main__':
|
||||||
test_gpt(4)
|
test_gpt(4, False)
|
||||||
|
|
|
@ -0,0 +1,75 @@
|
||||||
|
from colossalai.utils import free_port, ColoInitContext, get_current_device
|
||||||
|
from colossalai.testing import rerun_if_address_is_in_use
|
||||||
|
from colossalai.tensor import TensorSpec, ComputePattern, ParallelAction, init_colo_module
|
||||||
|
from functools import partial
|
||||||
|
from colossalai.core import global_context as gpc
|
||||||
|
from colossalai.context import ParallelMode
|
||||||
|
from colossalai.nn.parallel import ColoDDP
|
||||||
|
|
||||||
|
import colossalai
|
||||||
|
import torch
|
||||||
|
import torch.multiprocessing as mp
|
||||||
|
import pytest
|
||||||
|
|
||||||
|
class Net(torch.nn.Module):
|
||||||
|
def __init__(self):
|
||||||
|
super(Net, self).__init__()
|
||||||
|
self.embed = torch.nn.Embedding(20, 4)
|
||||||
|
self.proj = torch.nn.Linear(4, 8)
|
||||||
|
|
||||||
|
def forward(self, x):
|
||||||
|
# move input to cpu and restore output
|
||||||
|
current_dev = x.device
|
||||||
|
x = x.to('cpu')
|
||||||
|
x = self.embed(x)
|
||||||
|
x = x.to(current_dev)
|
||||||
|
|
||||||
|
x = self.proj(x)
|
||||||
|
return x
|
||||||
|
|
||||||
|
def run_hybrid_device(use_ddp):
|
||||||
|
with ColoInitContext(device=get_current_device()):
|
||||||
|
model = Net()
|
||||||
|
|
||||||
|
real_model = model
|
||||||
|
if use_ddp:
|
||||||
|
model = ColoDDP(model)
|
||||||
|
real_model = model.module
|
||||||
|
|
||||||
|
|
||||||
|
print(f'embedding weight size: {real_model.embed.weight.size()} | device: {real_model.embed.weight.device}')
|
||||||
|
#print(f'linear weight size: {real_model.proj.weight.size()} | device: {real_model.proj.weight.device}')
|
||||||
|
parallel_action = ParallelAction(ComputePattern.TP1D)
|
||||||
|
init_colo_module(model, parallel_action, recursive=True, mode='col')
|
||||||
|
|
||||||
|
# use cpu gloo to handle embedding
|
||||||
|
real_model.embed.to('cpu')
|
||||||
|
gloo_group_tp = gpc.get_cpu_group(ParallelMode.PARALLEL_1D)
|
||||||
|
real_model.embed.weight.spec.dist_spec.process_group = gloo_group_tp
|
||||||
|
|
||||||
|
print(f'embedding weight size: {real_model.embed.weight.size()} | new device: {real_model.embed.weight.device}')
|
||||||
|
#print(f'linear weight size: {real_model.proj.weight.size()} | new device: {real_model.proj.weight.device}')
|
||||||
|
|
||||||
|
data = torch.randint(low=0, high=20, size=(16,), device=get_current_device())
|
||||||
|
out = model(data)
|
||||||
|
out.sum().backward()
|
||||||
|
|
||||||
|
def run_dist(rank, world_size, port, use_ddp):
|
||||||
|
if use_ddp and world_size == 1:
|
||||||
|
return
|
||||||
|
tp_world_size = world_size // 2 if use_ddp else world_size
|
||||||
|
config = dict(parallel=dict(tensor=dict(mode="1d", size=tp_world_size),))
|
||||||
|
colossalai.launch(config=config, rank=rank, world_size=world_size, host='localhost', port=port, backend='nccl')
|
||||||
|
run_hybrid_device(use_ddp)
|
||||||
|
|
||||||
|
@pytest.mark.dist
|
||||||
|
@pytest.mark.parametrize('world_size', [1, 4])
|
||||||
|
@pytest.mark.parametrize('use_ddp', [False, True])
|
||||||
|
@rerun_if_address_is_in_use()
|
||||||
|
# Working for simulate the embedding(CPU DP+TP) -> nn(GPU DP+TP)
|
||||||
|
def _test_hybrid_device(world_size, use_ddp):
|
||||||
|
run_func = partial(run_dist, world_size=world_size, port=free_port(), use_ddp=use_ddp)
|
||||||
|
mp.spawn(run_func, nprocs=world_size)
|
||||||
|
|
||||||
|
if __name__ == '__main__':
|
||||||
|
_test_hybrid_device(1, False)
|
Loading…
Reference in New Issue