[polish] polish singleton and global context (#500)

colossalai/context/__init__.py

@@ -1,6 +1,6 @@
 from .config import Config, ConfigException
 from .parallel_context import ParallelContext
-from .moe_context import MoeContext
 from .parallel_mode import ParallelMode
+from .moe_context import MOE_CONTEXT
 from .process_group_initializer import *
 from .random import *

colossalai/context/moe_context.py

@@ -1,6 +1,9 @@
 import torch
 import torch.distributed as dist
-from .parallel_mode import ParallelMode
+
+from colossalai.context.parallel_mode import ParallelMode
+from colossalai.context.singleton_meta import SingletonMeta
+
 from typing import Tuple
 
 
@@ -56,17 +59,10 @@ class MoeParallelInfo:
                 self.dp_group = group
 
 
-class MoeContext:
+class MoeContext(metaclass=SingletonMeta):
     """MoE parallel context manager. This class manages different
     parallel groups in MoE context and MoE loss in training.
     """
-    __instance = None
-
-    @staticmethod
-    def get_instance():
-        if MoeContext.__instance is None:
-            MoeContext.__instance = MoeContext()
-        return MoeContext.__instance
 
     def __init__(self):
         self.world_size = 1
@@ -160,3 +156,6 @@ class MoeContext:
 
     def get_loss(self):
         return self.aux_loss
+
+
+MOE_CONTEXT = MoeContext()

colossalai/context/parallel_context.py

@@ -15,30 +15,16 @@ from colossalai.registry import DIST_GROUP_INITIALIZER
 
 from .parallel_mode import ParallelMode
 from .random import add_seed, get_seeds, set_mode
+from colossalai.context.singleton_meta import SingletonMeta
 
 
-class ParallelContext:
+class ParallelContext(metaclass=SingletonMeta):
     """This class provides interface functions for users to get the parallel context,
     such as the global rank, the local rank, the world size, etc. of each device.
 
     """
 
-    __instance = None
-
-    @staticmethod
-    def get_instance():
-        if ParallelContext.__instance is None:
-            ParallelContext()
-        return ParallelContext.__instance
-
     def __init__(self):
-        # create a singleton instance
-        if ParallelContext.__instance is not None:
-            raise Exception(
-                'ParallelContext is a singleton class, you should get the instance by colossalai.core.global_context')
-        else:
-            ParallelContext.__instance = self
-
         # distributed settings
         self._global_ranks = dict()
         self._local_ranks = dict()
@@ -510,3 +496,6 @@ class ParallelContext:
 
     def set_virtual_pipeline_parallel_rank(self, rank):
         self.virtual_pipeline_parallel_rank = rank
+
+
+global_context = ParallelContext()

colossalai/core.py

@@ -1,7 +1,4 @@
 #!/usr/bin/env python
 # -*- encoding: utf-8 -*-
 
-from colossalai.context import ParallelContext, MoeContext
-
-global_context = ParallelContext.get_instance()
-MOE_CONTEXT = MoeContext.get_instance()
+from colossalai.context.parallel_context import global_context

colossalai/engine/gradient_handler/_moe_gradient_handler.py

@@ -1,9 +1,10 @@
-from colossalai.core import global_context as gpc, MOE_CONTEXT
+from colossalai.core import global_context as gpc
 from colossalai.registry import GRADIENT_HANDLER
 from colossalai.utils.moe import get_moe_epsize_param_dict
 from ._base_gradient_handler import BaseGradientHandler
 from ...context.parallel_mode import ParallelMode
 from .utils import bucket_allreduce
+from colossalai.context.moe_context import MOE_CONTEXT
 
 
 @GRADIENT_HANDLER.register_module

colossalai/initialize.py

@@ -19,7 +19,9 @@ from colossalai.amp import AMP_TYPE, convert_to_amp
 from colossalai.amp.naive_amp import NaiveAMPModel
 from colossalai.builder.builder import build_gradient_handler
 from colossalai.context import Config, ConfigException, ParallelMode
-from colossalai.core import global_context as gpc, MOE_CONTEXT
+from colossalai.core import global_context as gpc
+
+from colossalai.context.moe_context import MOE_CONTEXT
 from colossalai.engine import Engine
 from colossalai.engine.ophooks import BaseOpHook
 from colossalai.logging import get_dist_logger

colossalai/nn/layer/moe/experts.py

@@ -4,7 +4,7 @@ import torch
 import torch.nn as nn
 from colossalai.context import ParallelMode, seed
 from colossalai.utils import get_current_device
-from colossalai.core import MOE_CONTEXT
+from colossalai.context.moe_context import MOE_CONTEXT
 from typing import Type
 
 

colossalai/nn/layer/moe/layers.py

@@ -4,7 +4,7 @@ import torch
 import torch.nn as nn
 import torch.nn.functional as F
 import torch.distributed as dist
-from colossalai.core import MOE_CONTEXT
+from colossalai.context.moe_context import MOE_CONTEXT
 from colossalai.utils import get_current_device
 from ._operation import COL_MOE_KERNEL_FLAG, AllToAll, AllGather, ReduceScatter, MoeDispatch, MoeCombine, moe_cumsum
 from .experts import MoeExperts, Experts

colossalai/nn/layer/moe/utils.py

@@ -1,6 +1,6 @@
 import torch
 from colossalai.utils import get_current_device
-from colossalai.core import MOE_CONTEXT
+from colossalai.context.moe_context import MOE_CONTEXT
 from .experts import FFNExperts, TPExperts
 
 

colossalai/nn/loss/loss_moe.py

@@ -1,7 +1,7 @@
 import torch.nn as nn
 from colossalai.registry import LOSSES
 from torch.nn.modules.loss import _Loss
-from colossalai.core import MOE_CONTEXT
+from colossalai.context.moe_context import MOE_CONTEXT
 
 
 @LOSSES.register_module

colossalai/utils/memory_tracer/model_data_memtracer.py

@@ -1,4 +1,4 @@
-from colossalai.utils.commons.singleton_meta import SingletonMeta
+from colossalai.context.singleton_meta import SingletonMeta
 from colossalai.utils.memory_tracer.commons import col_tensor_mem_usage
 import torch
 

colossalai/utils/moe.py

@@ -1,6 +1,7 @@
 import torch.nn as nn
 import torch.distributed as dist
-from colossalai.core import global_context as gpc, MOE_CONTEXT
+from colossalai.core import global_context as gpc
+from colossalai.context.moe_context import MOE_CONTEXT
 from colossalai.context import ParallelMode
 from .common import is_using_ddp
 from typing import Dict, List

model_zoo/moe/models.py

@@ -7,7 +7,7 @@ from colossalai.nn.layer import VanillaPatchEmbedding, VanillaClassifier, \
 from colossalai.nn.layer.moe import build_ffn_experts, MoeLayer, Top2Router, NormalNoiseGenerator, MoeModule
 from .util import moe_sa_args, moe_mlp_args
 from ..helper import TransformerLayer
-from colossalai.core import MOE_CONTEXT
+from colossalai.context.moe_context import MOE_CONTEXT
 from colossalai.utils import get_current_device
 from typing import List
 

tests/test_amp/test_naive_fp16.py

@@ -1,12 +1,15 @@
 import torch
-import colossalai
-import copy
-import pytest
 import torch.multiprocessing as mp
-from colossalai.amp import convert_to_naive_amp, convert_to_apex_amp
-from tests.components_to_test.registry import non_distributed_component_funcs
+
+import colossalai
 from colossalai.testing import assert_close_loose
 from colossalai.utils import free_port
+from colossalai.amp import convert_to_naive_amp, convert_to_apex_amp
+
+from tests.components_to_test.registry import non_distributed_component_funcs
+
+import copy
+import pytest
 from functools import partial
 
 

tests/test_moe/test_grad_handler.py

@@ -7,7 +7,7 @@ import torch.distributed as dist
 import colossalai
 from colossalai.utils import free_port, get_current_device
 from colossalai.nn.layer.moe import Top1Router, UniformNoiseGenerator, MoeLayer, Experts
-from colossalai.core import MOE_CONTEXT
+from colossalai.context.moe_context import MOE_CONTEXT
 from colossalai.utils.moe import sync_moe_model_param
 from colossalai.engine.gradient_handler import MoeGradientHandler
 from colossalai.testing import assert_equal_in_group

tests/test_moe/test_kernel.py

@@ -8,7 +8,7 @@ from colossalai.context import ParallelMode
 from colossalai.core import global_context as gpc
 from colossalai.utils import free_port, get_current_device
 from colossalai.nn.layer.moe import Top1Router, Top2Router, MoeLayer, Experts
-from colossalai.core import MOE_CONTEXT
+from colossalai.context.moe_context import MOE_CONTEXT
 
 BATCH_SIZE = 16
 NUM_EXPERTS = 4

tests/test_moe/test_moe_group.py

@@ -6,7 +6,7 @@ import torch.distributed as dist
 import colossalai
 from colossalai.utils import free_port, get_current_device
 from colossalai.nn.layer.moe import Experts
-from colossalai.core import MOE_CONTEXT
+from colossalai.context.moe_context import MOE_CONTEXT
 from colossalai.utils.moe import sync_moe_model_param
 from colossalai.testing import assert_equal_in_group