ColossalAI/colossalai/nn/layer/parallel_1d/layers.py

1041 lines
49 KiB
Python
Raw Normal View History

2021-10-28 16:21:23 +00:00
#!/usr/bin/env python
# -*- encoding: utf-8 -*-
Develop/experiments (#59) * Add gradient accumulation, fix lr scheduler * fix FP16 optimizer and adapted torch amp with tensor parallel (#18) * fixed bugs in compatibility between torch amp and tensor parallel and performed some minor fixes * fixed trainer * Revert "fixed trainer" This reverts commit 2e0b0b76990e8d4e337add483d878c0f61cf5097. * improved consistency between trainer, engine and schedule (#23) Co-authored-by: 1SAA <c2h214748@gmail.com> * Split conv2d, class token, positional embedding in 2d, Fix random number in ddp Fix convergence in cifar10, Imagenet1000 * Integrate 1d tensor parallel in Colossal-AI (#39) * fixed 1D and 2D convergence (#38) * optimized 2D operations * fixed 1D ViT convergence problem * Feature/ddp (#49) * remove redundancy func in setup (#19) (#20) * use env to control the language of doc (#24) (#25) * Support TP-compatible Torch AMP and Update trainer API (#27) * Add gradient accumulation, fix lr scheduler * fix FP16 optimizer and adapted torch amp with tensor parallel (#18) * fixed bugs in compatibility between torch amp and tensor parallel and performed some minor fixes * fixed trainer * Revert "fixed trainer" This reverts commit 2e0b0b76990e8d4e337add483d878c0f61cf5097. * improved consistency between trainer, engine and schedule (#23) Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: ver217 <lhx0217@gmail.com> * add an example of ViT-B/16 and remove w_norm clipping in LAMB (#29) * add explanation for ViT example (#35) (#36) * support torch ddp * fix loss accumulation * add log for ddp * change seed * modify timing hook Co-authored-by: Frank Lee <somerlee.9@gmail.com> Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: binmakeswell <binmakeswell@gmail.com> * Feature/pipeline (#40) * remove redundancy func in setup (#19) (#20) * use env to control the language of doc (#24) (#25) * Support TP-compatible Torch AMP and Update trainer API (#27) * Add gradient accumulation, fix lr scheduler * fix FP16 optimizer and adapted torch amp with tensor parallel (#18) * fixed bugs in compatibility between torch amp and tensor parallel and performed some minor fixes * fixed trainer * Revert "fixed trainer" This reverts commit 2e0b0b76990e8d4e337add483d878c0f61cf5097. * improved consistency between trainer, engine and schedule (#23) Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: ver217 <lhx0217@gmail.com> * add an example of ViT-B/16 and remove w_norm clipping in LAMB (#29) * add explanation for ViT example (#35) (#36) * optimize communication of pipeline parallel * fix grad clip for pipeline Co-authored-by: Frank Lee <somerlee.9@gmail.com> Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: binmakeswell <binmakeswell@gmail.com> * optimized 3d layer to fix slow computation ; tested imagenet performance with 3d; reworked lr_scheduler config definition; fixed launch args; fixed some printing issues; simplified apis of 3d layers (#51) * Update 2.5d layer code to get a similar accuracy on imagenet-1k dataset * update api for better usability (#58) update api for better usability Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: ver217 <lhx0217@gmail.com> Co-authored-by: puck_WCR <46049915+WANG-CR@users.noreply.github.com> Co-authored-by: binmakeswell <binmakeswell@gmail.com> Co-authored-by: アマデウス <kurisusnowdeng@users.noreply.github.com> Co-authored-by: BoxiangW <45734921+BoxiangW@users.noreply.github.com>
2021-12-09 07:08:29 +00:00
import math
from collections import OrderedDict
from typing import Callable, Tuple
2021-10-28 16:21:23 +00:00
import torch
import torch.nn.functional as F
2022-11-08 08:22:23 +00:00
from torch import Tensor
from torch.nn.parameter import Parameter
from colossalai.communication import broadcast
from colossalai.context import ParallelMode, seed
2021-10-28 16:21:23 +00:00
from colossalai.core import global_context as gpc
from colossalai.global_variables import tensor_parallel_env as env
from colossalai.kernel import LayerNorm
from colossalai.nn import init as init
2021-10-28 16:21:23 +00:00
from colossalai.registry import LAYERS
2022-11-08 08:22:23 +00:00
from colossalai.utils.checkpointing import (
broadcast_state_dict,
gather_tensor_parallel_state_dict,
partition_tensor_parallel_state_dict,
)
from colossalai.utils.cuda import get_current_device
2022-11-08 08:22:23 +00:00
2021-10-28 16:21:23 +00:00
from ..base_layer import ParallelLayer
from ..colossalai_layer._utils import ColossalaiModule
from ..utils import divide, set_tensor_parallel_attribute_by_partition
2022-11-08 08:22:23 +00:00
from ..vanilla import VanillaLayerNorm, VanillaPatchEmbedding
2022-10-26 12:54:39 +00:00
from ._operation import linear_with_async_comm
2022-11-08 08:22:23 +00:00
from ._utils import (
gather_forward_split_backward,
get_parallel_input,
reduce_grad,
reduce_input,
set_parallel_input,
split_forward_gather_backward,
)
Fast_LN = None
try:
from apex.contrib.layer_norm.layer_norm import FastLayerNorm
Fast_LN = FastLayerNorm
except ImportError:
pass
@LAYERS.register_module
class Linear1D(ColossalaiModule):
2022-03-25 05:02:39 +00:00
r"""Linear layer for 1D parallelism.
Args:
in_features (int): size of each input sample.
out_features (int): size of each output sample.
bias (bool, optional): If set to ``False``, the layer will not learn an additive bias, defaults to ``True``.
dtype (:class:`torch.dtype`, optional): The dtype of parameters, defaults to None.
gather_output (bool, optional): Whether to call all-gather on output, defaults to False.
skip_bias_add (bool, optional): If set to ``True``, it will skip bias add for linear layer,
which is preserved for kernel fusion, defaults to False
weight_initializer (:class:`typing.Callable`, optional):
The initializer of weight, defaults to kaiming uniform initializer.
bias_initializer (:class:`typing.Callable`, optional):
The initializer of bias, defaults to xavier uniform initializer.
More details about ``initializer`` please refer to
`init <https://github.com/hpcaitech/ColossalAI/blob/main/colossalai/nn/init.py>`_.
"""
def __init__(self,
in_features: int,
out_features: int,
bias: bool = True,
dtype: torch.dtype = None,
gather_output: bool = False,
skip_bias_add: bool = False,
weight_initializer: Callable = init.kaiming_uniform_(a=math.sqrt(5)),
bias_initializer: Callable = init.xavier_uniform_(a=1, scale=1)):
parallel_input = get_parallel_input()
if not parallel_input and not gather_output:
layer = Linear1D_Col(in_features,
out_features,
bias=bias,
dtype=dtype,
skip_bias_add=skip_bias_add,
weight_initializer=weight_initializer,
bias_initializer=bias_initializer)
else:
layer = Linear1D_Row(in_features,
out_features,
bias=bias,
dtype=dtype,
parallel_input=parallel_input,
skip_bias_add=skip_bias_add,
weight_initializer=weight_initializer,
bias_initializer=bias_initializer)
super().__init__(layer)
@LAYERS.register_module
class LayerNorm1D(ColossalaiModule):
r"""
Layer Normalization for colossalai
Args:
normalized_shape (int): input shape from an expected input of size.
:math:`[* \times \text{normalized_shape}[0] \times \text{normalized_shape}[1]
\times \ldots \times \text{normalized_shape}[-1]]`
If a single integer is used, it is treated as a singleton list, and this module will
normalize over the last dimension which is expected to be of that specific size.
eps (float): a value added to the denominator for numerical stability, defaults to 1e-05.
bias (bool, optional): Whether to add a bias, defaults to ``True``.
dtype (:class:`torch.dtype`, optional): The dtype of parameters, defaults to None.
"""
2022-10-26 12:54:39 +00:00
_fast_ln_supported_sizes = [
1024, 1536, 2048, 2304, 3072, 3840, 4096, 5120, 6144, 8192, 10240, 12288, 12800, 15360, 16384, 18432, 20480,
24576, 25600, 30720, 32768, 40960, 49152, 65536
]
def __init__(self, normalized_shape: int, eps=1e-05, bias=True, dtype=None):
2022-11-08 08:22:23 +00:00
if Fast_LN is not None and normalized_shape in self._fast_ln_supported_sizes:
norm = Fast_LN(normalized_shape, eps=eps).to(dtype)
2022-10-26 12:54:39 +00:00
else:
2022-11-08 08:22:23 +00:00
norm = None
try:
from apex.normalization import FusedLayerNorm
norm = FusedLayerNorm(normalized_shape, eps=eps).to(dtype)
except ImportError:
norm = LayerNorm(normalized_shape, eps=eps).to(dtype)
super().__init__(norm)
def _load_from_state_dict(self, state_dict, prefix, *args):
local_state = OrderedDict()
weight_key = prefix + 'weight'
bias_key = prefix + 'bias'
if gpc.get_local_rank(ParallelMode.TENSOR) == 0:
# weight
weight = state_dict.pop(weight_key, None)
if weight is not None:
local_state[weight_key] = weight
# bias
bias = state_dict.pop(bias_key, None)
if bias is not None:
local_state[bias_key] = bias
local_state = broadcast_state_dict(local_state, ParallelMode.PARALLEL_1D)
super()._load_from_state_dict(local_state, prefix, *args)
def _save_to_state_dict(self, destination, prefix, keep_vars):
if gpc.get_local_rank(ParallelMode.TENSOR) == 0:
super()._save_to_state_dict(destination, prefix, keep_vars)
@LAYERS.register_module
class Classifier1D(ParallelLayer):
2022-03-25 05:02:39 +00:00
r"""RowLinear with given weight. Classifier of 1D parallelism.
Args:
in_features (int): size of each input sample.
num_classes (int): number of classes.
weight (:class:`torch.nn.Parameter`, optional): weight of the classifier, defaults to None.
bias (bool, optional): If set to ``False``, the layer will not learn an additive bias, defaults to ``True``.
dtype (:class:`torch.dtype`, optional): The dtype of parameters, defaults to None.
weight_initializer (:class:`typing.Callable`, optional):
The initializer of weight, defaults to kaiming uniform initializer.
bias_initializer (:class:`typing.Callable`, optional):
The initializer of bias, defaults to xavier uniform initializer.
More details about ``initializer`` please refer to
`init <https://github.com/hpcaitech/ColossalAI/blob/main/colossalai/nn/init.py>`_.
"""
def __init__(self,
in_features: int,
num_classes: int,
weight: Parameter = None,
bias: bool = True,
dtype: torch.dtype = None,
weight_initializer: Callable = init.kaiming_uniform_(a=math.sqrt(5)),
bias_initializer: Callable = init.xavier_uniform_(a=1, scale=1)):
super().__init__()
self.in_features = in_features
self.num_classes = num_classes
self.parallel_input = get_parallel_input()
# Divide the weight matrix along the last dimension.
self.input_size_per_partition = divide(in_features, gpc.tensor_parallel_size)
# Parameters.
# Initialize weight.
factory_kwargs = {'device': get_current_device(), 'dtype': dtype}
if weight is not None:
self.weight = weight
self.has_weight = False
else:
self.weight = Parameter(torch.empty(self.num_classes, self.input_size_per_partition, **factory_kwargs))
self.has_weight = True
if bias:
self.bias = Parameter(torch.empty(self.num_classes, **factory_kwargs))
else:
self.bias = None
with seed(ParallelMode.TENSOR):
self.reset_parameters(weight_initializer, bias_initializer)
self._set_tensor_parallel_attributes()
set_parallel_input(False)
env.vocab_parallel = False
def reset_parameters(self, weight_initializer, bias_initializer) -> None:
fan_in, fan_out = self.in_features, self.num_classes
if self.has_weight:
weight_initializer(self.weight, fan_in=fan_in, fan_out=fan_out)
if self.bias is not None:
bias_initializer(self.bias, fan_in=fan_in)
broadcast(self.bias, gpc.get_ranks_in_group(ParallelMode.PARALLEL_1D)[0], ParallelMode.PARALLEL_1D)
def _set_tensor_parallel_attributes(self):
if self.has_weight:
num_partition = gpc.get_world_size(ParallelMode.TENSOR)
set_tensor_parallel_attribute_by_partition(self.weight, num_partition)
def _load_from_global_state_dict(self, state_dict, prefix, *args):
local_state = OrderedDict()
weight_key = prefix + 'weight'
bias_key = prefix + 'bias'
if gpc.get_local_rank(ParallelMode.TENSOR) == 0:
# weight
if self.has_weight:
weight = state_dict.pop(weight_key, None)
if weight is not None:
local_state[weight_key] = weight
# bias
if self.bias is not None:
bias = state_dict.pop(bias_key, None)
if bias is not None:
local_state[bias_key] = bias
local_state = partition_tensor_parallel_state_dict(local_state,
ParallelMode.PARALLEL_1D,
dims={
weight_key: -1,
bias_key: 0
},
partition_states={
weight_key: True,
bias_key: False
})
super()._load_from_global_state_dict(local_state, prefix, *args)
def _save_to_global_state_dict(self, destination, prefix, keep_vars):
weight_key = prefix + 'weight'
bias_key = prefix + 'bias'
local_state = OrderedDict()
if self.has_weight:
local_state[weight_key] = self.weight
if self.bias is not None:
local_state[bias_key] = self.bias
local_state = gather_tensor_parallel_state_dict(local_state,
ParallelMode.PARALLEL_1D,
dims={
weight_key: -1,
bias_key: 0
},
partition_states={
weight_key: True,
bias_key: False
},
keep_vars=keep_vars)
destination.update(local_state)
def forward(self, input_: Tensor) -> Tensor:
# Set up backprop all-reduce.
if self.parallel_input:
2022-04-01 08:44:23 +00:00
assert input_.shape[-1] == self.weight.shape[-1], \
'Invalid shapes in Classifier1D forward: input={}, weight={}. Expected last dim of input {}.'.format(
2022-04-01 08:44:23 +00:00
input_.shape, self.weight.shape, self.weight.shape[-1])
input_ = input_
else:
2022-04-01 08:44:23 +00:00
assert divide(input_.shape[-1], gpc.tensor_parallel_size) == self.weight.shape[-1], \
'Invalid shapes in Classifier1D forward: input={}, weight={}. Expected last dim of input {}.'.format(
2022-04-01 08:44:23 +00:00
input_.shape, self.weight.shape, self.weight.shape[-1] * gpc.tensor_parallel_size)
input_ = split_forward_gather_backward(input_, ParallelMode.PARALLEL_1D, dim=-1)
output_parallel = F.linear(input_, self.weight)
output = reduce_input(output_parallel, ParallelMode.PARALLEL_1D)
if self.bias is not None:
output = output + self.bias
return output
2021-10-28 16:21:23 +00:00
@LAYERS.register_module
class VocabParallelClassifier1D(ParallelLayer):
2022-03-25 05:02:39 +00:00
r"""ColLinear with given weight. Classifier of 1D parallelism.
Args:
in_features (int): size of each input sample.
num_classes (int): number of classes.
weight (:class:`torch.nn.Parameter`, optional): weight of the classifier, defaults to None.
bias (bool, optional): If set to ``False``, the layer will not learn an additive bias, defaults to ``True``.
dtype (:class:`torch.dtype`, optional): The dtype of parameters, defaults to None.
weight_initializer (:class:`typing.Callable`, optional):
The initializer of weight, defaults to kaiming uniform initializer.
bias_initializer (:class:`typing.Callable`, optional):
The initializer of bias, defaults to xavier uniform initializer.
More details about ``initializer`` please refer to
`init <https://github.com/hpcaitech/ColossalAI/blob/main/colossalai/nn/init.py>`_.
"""
def __init__(self,
in_features: int,
num_classes: int,
weight: Parameter = None,
bias: bool = True,
dtype: torch.dtype = None,
gather_output: bool = False,
weight_initializer: Callable = init.kaiming_uniform_(a=math.sqrt(5)),
bias_initializer: Callable = init.xavier_uniform_(a=1, scale=1)):
super().__init__()
self.in_features = in_features
self.num_classes = num_classes
self.gather_output = gather_output
self.parallel_input = get_parallel_input()
# Divide the weight matrix along the last dimension.
self.num_classes_per_partition = divide(num_classes, gpc.tensor_parallel_size)
# Parameters.
# Initialize weight.
factory_kwargs = {'device': get_current_device(), 'dtype': dtype}
if weight is not None:
self.weight = weight
self.has_weight = False
else:
self.weight = Parameter(torch.empty(self.num_classes_per_partition, self.in_features, **factory_kwargs))
self.has_weight = True
if bias:
self.bias = Parameter(torch.empty(self.num_classes_per_partition, **factory_kwargs))
else:
self.bias = None
with seed(ParallelMode.TENSOR):
self.reset_parameters(weight_initializer, bias_initializer)
self._set_tensor_parallel_attributes()
set_parallel_input(False)
env.vocab_parallel = True
def reset_parameters(self, weight_initializer, bias_initializer) -> None:
fan_in, fan_out = self.in_features, self.num_classes
if self.has_weight:
weight_initializer(self.weight, fan_in=fan_in, fan_out=fan_out)
if self.bias is not None:
bias_initializer(self.bias, fan_in=fan_in)
def _set_tensor_parallel_attributes(self):
num_partition = gpc.get_world_size(ParallelMode.TENSOR)
if self.has_weight:
set_tensor_parallel_attribute_by_partition(self.weight, num_partition)
if self.bias is not None:
set_tensor_parallel_attribute_by_partition(self.bias, num_partition)
def _load_from_global_state_dict(self, state_dict, prefix, *args):
local_state = OrderedDict()
weight_key = prefix + 'weight'
bias_key = prefix + 'bias'
if gpc.get_local_rank(ParallelMode.TENSOR) == 0:
# weight
if self.has_weight:
weight = state_dict.pop(weight_key, None)
if weight is not None:
local_state[weight_key] = weight
# bias
if self.bias is not None:
bias = state_dict.pop(bias_key, None)
if bias is not None:
local_state[bias_key] = bias
local_state = partition_tensor_parallel_state_dict(local_state,
ParallelMode.PARALLEL_1D,
dims={
weight_key: 0,
bias_key: 0
},
partition_states={
weight_key: True,
bias_key: True
})
super()._load_from_global_state_dict(local_state, prefix, *args)
def _save_to_global_state_dict(self, destination, prefix, keep_vars):
weight_key = prefix + 'weight'
bias_key = prefix + 'bias'
local_state = OrderedDict()
if self.has_weight:
local_state[weight_key] = self.weight
if self.bias is not None:
local_state[bias_key] = self.bias
local_state = gather_tensor_parallel_state_dict(local_state,
ParallelMode.PARALLEL_1D,
dims={
weight_key: 0,
bias_key: 0
},
partition_states={
weight_key: True,
bias_key: True
},
keep_vars=keep_vars)
destination.update(local_state)
def forward(self, input_: Tensor) -> Tensor:
2022-04-01 08:44:23 +00:00
assert input_.shape[-1] == self.weight.shape[-1], \
'Invalid shapes in VocabParallelClassifier1D forward: input={}, weight={}. Expected last dim of input {}.'.format(
input_.shape, self.weight.shape, self.weight.shape[-1])
# Set up backprop all-reduce.
input_parallel = reduce_grad(input_, ParallelMode.PARALLEL_1D)
# Matrix multiply.
output_parallel = F.linear(input_parallel, self.weight, self.bias)
if self.gather_output:
# All-gather across the partitions.
output = gather_forward_split_backward(output_parallel, ParallelMode.PARALLEL_1D, dim=-1)
else:
output = output_parallel
return output
Develop/experiments (#59) * Add gradient accumulation, fix lr scheduler * fix FP16 optimizer and adapted torch amp with tensor parallel (#18) * fixed bugs in compatibility between torch amp and tensor parallel and performed some minor fixes * fixed trainer * Revert "fixed trainer" This reverts commit 2e0b0b76990e8d4e337add483d878c0f61cf5097. * improved consistency between trainer, engine and schedule (#23) Co-authored-by: 1SAA <c2h214748@gmail.com> * Split conv2d, class token, positional embedding in 2d, Fix random number in ddp Fix convergence in cifar10, Imagenet1000 * Integrate 1d tensor parallel in Colossal-AI (#39) * fixed 1D and 2D convergence (#38) * optimized 2D operations * fixed 1D ViT convergence problem * Feature/ddp (#49) * remove redundancy func in setup (#19) (#20) * use env to control the language of doc (#24) (#25) * Support TP-compatible Torch AMP and Update trainer API (#27) * Add gradient accumulation, fix lr scheduler * fix FP16 optimizer and adapted torch amp with tensor parallel (#18) * fixed bugs in compatibility between torch amp and tensor parallel and performed some minor fixes * fixed trainer * Revert "fixed trainer" This reverts commit 2e0b0b76990e8d4e337add483d878c0f61cf5097. * improved consistency between trainer, engine and schedule (#23) Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: ver217 <lhx0217@gmail.com> * add an example of ViT-B/16 and remove w_norm clipping in LAMB (#29) * add explanation for ViT example (#35) (#36) * support torch ddp * fix loss accumulation * add log for ddp * change seed * modify timing hook Co-authored-by: Frank Lee <somerlee.9@gmail.com> Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: binmakeswell <binmakeswell@gmail.com> * Feature/pipeline (#40) * remove redundancy func in setup (#19) (#20) * use env to control the language of doc (#24) (#25) * Support TP-compatible Torch AMP and Update trainer API (#27) * Add gradient accumulation, fix lr scheduler * fix FP16 optimizer and adapted torch amp with tensor parallel (#18) * fixed bugs in compatibility between torch amp and tensor parallel and performed some minor fixes * fixed trainer * Revert "fixed trainer" This reverts commit 2e0b0b76990e8d4e337add483d878c0f61cf5097. * improved consistency between trainer, engine and schedule (#23) Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: ver217 <lhx0217@gmail.com> * add an example of ViT-B/16 and remove w_norm clipping in LAMB (#29) * add explanation for ViT example (#35) (#36) * optimize communication of pipeline parallel * fix grad clip for pipeline Co-authored-by: Frank Lee <somerlee.9@gmail.com> Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: binmakeswell <binmakeswell@gmail.com> * optimized 3d layer to fix slow computation ; tested imagenet performance with 3d; reworked lr_scheduler config definition; fixed launch args; fixed some printing issues; simplified apis of 3d layers (#51) * Update 2.5d layer code to get a similar accuracy on imagenet-1k dataset * update api for better usability (#58) update api for better usability Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: ver217 <lhx0217@gmail.com> Co-authored-by: puck_WCR <46049915+WANG-CR@users.noreply.github.com> Co-authored-by: binmakeswell <binmakeswell@gmail.com> Co-authored-by: アマデウス <kurisusnowdeng@users.noreply.github.com> Co-authored-by: BoxiangW <45734921+BoxiangW@users.noreply.github.com>
2021-12-09 07:08:29 +00:00
@LAYERS.register_module
2021-10-28 16:21:23 +00:00
class Linear1D_Col(ParallelLayer):
2022-03-25 05:02:39 +00:00
r"""Linear layer with column parallelism.
2021-10-28 16:21:23 +00:00
The linear layer is defined as :math:`Y = XA + b`. A is parallelized along
its second dimension as :math:`A = [A_1, ..., A_p]`.
2022-03-25 05:02:39 +00:00
Args:
in_features (int): size of each input sample.
out_features (int): size of each output sample.
bias (bool, optional): If set to ``False``, the layer will not learn an additive bias, defaults to ``True``.
dtype (:class:`torch.dtype`, optional): The dtype of parameters, defaults to None.
gather_output (bool, optional): If true, call all-gather on output and make Y available
2021-10-28 16:21:23 +00:00
to all GPUs, otherwise, every GPU will have its output
which is :math:`Y_i = XA_i`, defaults to False
2022-03-25 05:02:39 +00:00
skip_bias_add (bool, optional): If set to ``True``, it will skip bias add for linear layer,
which is preserved for kernel fusion, defaults to Fals
weight_initializer (:class:`typing.Callable`, optional):
The initializer of weight, defaults to kaiming uniform initializer.
bias_initializer (:class:`typing.Callable`, optional):
The initializer of bias, defaults to xavier uniform initializer.
More details about ``initializer`` please refer to
`init <https://github.com/hpcaitech/ColossalAI/blob/main/colossalai/nn/init.py>`_.
2021-10-28 16:21:23 +00:00
"""
2021-10-28 16:21:23 +00:00
def __init__(self,
in_features: int,
out_features: int,
2021-10-28 16:21:23 +00:00
bias: bool = True,
dtype: torch.dtype = None,
Develop/experiments (#59) * Add gradient accumulation, fix lr scheduler * fix FP16 optimizer and adapted torch amp with tensor parallel (#18) * fixed bugs in compatibility between torch amp and tensor parallel and performed some minor fixes * fixed trainer * Revert "fixed trainer" This reverts commit 2e0b0b76990e8d4e337add483d878c0f61cf5097. * improved consistency between trainer, engine and schedule (#23) Co-authored-by: 1SAA <c2h214748@gmail.com> * Split conv2d, class token, positional embedding in 2d, Fix random number in ddp Fix convergence in cifar10, Imagenet1000 * Integrate 1d tensor parallel in Colossal-AI (#39) * fixed 1D and 2D convergence (#38) * optimized 2D operations * fixed 1D ViT convergence problem * Feature/ddp (#49) * remove redundancy func in setup (#19) (#20) * use env to control the language of doc (#24) (#25) * Support TP-compatible Torch AMP and Update trainer API (#27) * Add gradient accumulation, fix lr scheduler * fix FP16 optimizer and adapted torch amp with tensor parallel (#18) * fixed bugs in compatibility between torch amp and tensor parallel and performed some minor fixes * fixed trainer * Revert "fixed trainer" This reverts commit 2e0b0b76990e8d4e337add483d878c0f61cf5097. * improved consistency between trainer, engine and schedule (#23) Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: ver217 <lhx0217@gmail.com> * add an example of ViT-B/16 and remove w_norm clipping in LAMB (#29) * add explanation for ViT example (#35) (#36) * support torch ddp * fix loss accumulation * add log for ddp * change seed * modify timing hook Co-authored-by: Frank Lee <somerlee.9@gmail.com> Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: binmakeswell <binmakeswell@gmail.com> * Feature/pipeline (#40) * remove redundancy func in setup (#19) (#20) * use env to control the language of doc (#24) (#25) * Support TP-compatible Torch AMP and Update trainer API (#27) * Add gradient accumulation, fix lr scheduler * fix FP16 optimizer and adapted torch amp with tensor parallel (#18) * fixed bugs in compatibility between torch amp and tensor parallel and performed some minor fixes * fixed trainer * Revert "fixed trainer" This reverts commit 2e0b0b76990e8d4e337add483d878c0f61cf5097. * improved consistency between trainer, engine and schedule (#23) Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: ver217 <lhx0217@gmail.com> * add an example of ViT-B/16 and remove w_norm clipping in LAMB (#29) * add explanation for ViT example (#35) (#36) * optimize communication of pipeline parallel * fix grad clip for pipeline Co-authored-by: Frank Lee <somerlee.9@gmail.com> Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: binmakeswell <binmakeswell@gmail.com> * optimized 3d layer to fix slow computation ; tested imagenet performance with 3d; reworked lr_scheduler config definition; fixed launch args; fixed some printing issues; simplified apis of 3d layers (#51) * Update 2.5d layer code to get a similar accuracy on imagenet-1k dataset * update api for better usability (#58) update api for better usability Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: ver217 <lhx0217@gmail.com> Co-authored-by: puck_WCR <46049915+WANG-CR@users.noreply.github.com> Co-authored-by: binmakeswell <binmakeswell@gmail.com> Co-authored-by: アマデウス <kurisusnowdeng@users.noreply.github.com> Co-authored-by: BoxiangW <45734921+BoxiangW@users.noreply.github.com>
2021-12-09 07:08:29 +00:00
gather_output: bool = False,
skip_bias_add: bool = False,
weight_initializer: Callable = init.kaiming_uniform_(a=math.sqrt(5)),
bias_initializer: Callable = init.xavier_uniform_(a=1, scale=1)):
2021-10-28 16:21:23 +00:00
super().__init__()
# Keep input parameters
Develop/experiments (#59) * Add gradient accumulation, fix lr scheduler * fix FP16 optimizer and adapted torch amp with tensor parallel (#18) * fixed bugs in compatibility between torch amp and tensor parallel and performed some minor fixes * fixed trainer * Revert "fixed trainer" This reverts commit 2e0b0b76990e8d4e337add483d878c0f61cf5097. * improved consistency between trainer, engine and schedule (#23) Co-authored-by: 1SAA <c2h214748@gmail.com> * Split conv2d, class token, positional embedding in 2d, Fix random number in ddp Fix convergence in cifar10, Imagenet1000 * Integrate 1d tensor parallel in Colossal-AI (#39) * fixed 1D and 2D convergence (#38) * optimized 2D operations * fixed 1D ViT convergence problem * Feature/ddp (#49) * remove redundancy func in setup (#19) (#20) * use env to control the language of doc (#24) (#25) * Support TP-compatible Torch AMP and Update trainer API (#27) * Add gradient accumulation, fix lr scheduler * fix FP16 optimizer and adapted torch amp with tensor parallel (#18) * fixed bugs in compatibility between torch amp and tensor parallel and performed some minor fixes * fixed trainer * Revert "fixed trainer" This reverts commit 2e0b0b76990e8d4e337add483d878c0f61cf5097. * improved consistency between trainer, engine and schedule (#23) Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: ver217 <lhx0217@gmail.com> * add an example of ViT-B/16 and remove w_norm clipping in LAMB (#29) * add explanation for ViT example (#35) (#36) * support torch ddp * fix loss accumulation * add log for ddp * change seed * modify timing hook Co-authored-by: Frank Lee <somerlee.9@gmail.com> Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: binmakeswell <binmakeswell@gmail.com> * Feature/pipeline (#40) * remove redundancy func in setup (#19) (#20) * use env to control the language of doc (#24) (#25) * Support TP-compatible Torch AMP and Update trainer API (#27) * Add gradient accumulation, fix lr scheduler * fix FP16 optimizer and adapted torch amp with tensor parallel (#18) * fixed bugs in compatibility between torch amp and tensor parallel and performed some minor fixes * fixed trainer * Revert "fixed trainer" This reverts commit 2e0b0b76990e8d4e337add483d878c0f61cf5097. * improved consistency between trainer, engine and schedule (#23) Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: ver217 <lhx0217@gmail.com> * add an example of ViT-B/16 and remove w_norm clipping in LAMB (#29) * add explanation for ViT example (#35) (#36) * optimize communication of pipeline parallel * fix grad clip for pipeline Co-authored-by: Frank Lee <somerlee.9@gmail.com> Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: binmakeswell <binmakeswell@gmail.com> * optimized 3d layer to fix slow computation ; tested imagenet performance with 3d; reworked lr_scheduler config definition; fixed launch args; fixed some printing issues; simplified apis of 3d layers (#51) * Update 2.5d layer code to get a similar accuracy on imagenet-1k dataset * update api for better usability (#58) update api for better usability Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: ver217 <lhx0217@gmail.com> Co-authored-by: puck_WCR <46049915+WANG-CR@users.noreply.github.com> Co-authored-by: binmakeswell <binmakeswell@gmail.com> Co-authored-by: アマデウス <kurisusnowdeng@users.noreply.github.com> Co-authored-by: BoxiangW <45734921+BoxiangW@users.noreply.github.com>
2021-12-09 07:08:29 +00:00
self.in_features = in_features
self.out_features = out_features
2021-10-28 16:21:23 +00:00
self.gather_output = gather_output
Develop/experiments (#59) * Add gradient accumulation, fix lr scheduler * fix FP16 optimizer and adapted torch amp with tensor parallel (#18) * fixed bugs in compatibility between torch amp and tensor parallel and performed some minor fixes * fixed trainer * Revert "fixed trainer" This reverts commit 2e0b0b76990e8d4e337add483d878c0f61cf5097. * improved consistency between trainer, engine and schedule (#23) Co-authored-by: 1SAA <c2h214748@gmail.com> * Split conv2d, class token, positional embedding in 2d, Fix random number in ddp Fix convergence in cifar10, Imagenet1000 * Integrate 1d tensor parallel in Colossal-AI (#39) * fixed 1D and 2D convergence (#38) * optimized 2D operations * fixed 1D ViT convergence problem * Feature/ddp (#49) * remove redundancy func in setup (#19) (#20) * use env to control the language of doc (#24) (#25) * Support TP-compatible Torch AMP and Update trainer API (#27) * Add gradient accumulation, fix lr scheduler * fix FP16 optimizer and adapted torch amp with tensor parallel (#18) * fixed bugs in compatibility between torch amp and tensor parallel and performed some minor fixes * fixed trainer * Revert "fixed trainer" This reverts commit 2e0b0b76990e8d4e337add483d878c0f61cf5097. * improved consistency between trainer, engine and schedule (#23) Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: ver217 <lhx0217@gmail.com> * add an example of ViT-B/16 and remove w_norm clipping in LAMB (#29) * add explanation for ViT example (#35) (#36) * support torch ddp * fix loss accumulation * add log for ddp * change seed * modify timing hook Co-authored-by: Frank Lee <somerlee.9@gmail.com> Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: binmakeswell <binmakeswell@gmail.com> * Feature/pipeline (#40) * remove redundancy func in setup (#19) (#20) * use env to control the language of doc (#24) (#25) * Support TP-compatible Torch AMP and Update trainer API (#27) * Add gradient accumulation, fix lr scheduler * fix FP16 optimizer and adapted torch amp with tensor parallel (#18) * fixed bugs in compatibility between torch amp and tensor parallel and performed some minor fixes * fixed trainer * Revert "fixed trainer" This reverts commit 2e0b0b76990e8d4e337add483d878c0f61cf5097. * improved consistency between trainer, engine and schedule (#23) Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: ver217 <lhx0217@gmail.com> * add an example of ViT-B/16 and remove w_norm clipping in LAMB (#29) * add explanation for ViT example (#35) (#36) * optimize communication of pipeline parallel * fix grad clip for pipeline Co-authored-by: Frank Lee <somerlee.9@gmail.com> Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: binmakeswell <binmakeswell@gmail.com> * optimized 3d layer to fix slow computation ; tested imagenet performance with 3d; reworked lr_scheduler config definition; fixed launch args; fixed some printing issues; simplified apis of 3d layers (#51) * Update 2.5d layer code to get a similar accuracy on imagenet-1k dataset * update api for better usability (#58) update api for better usability Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: ver217 <lhx0217@gmail.com> Co-authored-by: puck_WCR <46049915+WANG-CR@users.noreply.github.com> Co-authored-by: binmakeswell <binmakeswell@gmail.com> Co-authored-by: アマデウス <kurisusnowdeng@users.noreply.github.com> Co-authored-by: BoxiangW <45734921+BoxiangW@users.noreply.github.com>
2021-12-09 07:08:29 +00:00
self.skip_bias_add = skip_bias_add
2021-10-28 16:21:23 +00:00
Develop/experiments (#59) * Add gradient accumulation, fix lr scheduler * fix FP16 optimizer and adapted torch amp with tensor parallel (#18) * fixed bugs in compatibility between torch amp and tensor parallel and performed some minor fixes * fixed trainer * Revert "fixed trainer" This reverts commit 2e0b0b76990e8d4e337add483d878c0f61cf5097. * improved consistency between trainer, engine and schedule (#23) Co-authored-by: 1SAA <c2h214748@gmail.com> * Split conv2d, class token, positional embedding in 2d, Fix random number in ddp Fix convergence in cifar10, Imagenet1000 * Integrate 1d tensor parallel in Colossal-AI (#39) * fixed 1D and 2D convergence (#38) * optimized 2D operations * fixed 1D ViT convergence problem * Feature/ddp (#49) * remove redundancy func in setup (#19) (#20) * use env to control the language of doc (#24) (#25) * Support TP-compatible Torch AMP and Update trainer API (#27) * Add gradient accumulation, fix lr scheduler * fix FP16 optimizer and adapted torch amp with tensor parallel (#18) * fixed bugs in compatibility between torch amp and tensor parallel and performed some minor fixes * fixed trainer * Revert "fixed trainer" This reverts commit 2e0b0b76990e8d4e337add483d878c0f61cf5097. * improved consistency between trainer, engine and schedule (#23) Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: ver217 <lhx0217@gmail.com> * add an example of ViT-B/16 and remove w_norm clipping in LAMB (#29) * add explanation for ViT example (#35) (#36) * support torch ddp * fix loss accumulation * add log for ddp * change seed * modify timing hook Co-authored-by: Frank Lee <somerlee.9@gmail.com> Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: binmakeswell <binmakeswell@gmail.com> * Feature/pipeline (#40) * remove redundancy func in setup (#19) (#20) * use env to control the language of doc (#24) (#25) * Support TP-compatible Torch AMP and Update trainer API (#27) * Add gradient accumulation, fix lr scheduler * fix FP16 optimizer and adapted torch amp with tensor parallel (#18) * fixed bugs in compatibility between torch amp and tensor parallel and performed some minor fixes * fixed trainer * Revert "fixed trainer" This reverts commit 2e0b0b76990e8d4e337add483d878c0f61cf5097. * improved consistency between trainer, engine and schedule (#23) Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: ver217 <lhx0217@gmail.com> * add an example of ViT-B/16 and remove w_norm clipping in LAMB (#29) * add explanation for ViT example (#35) (#36) * optimize communication of pipeline parallel * fix grad clip for pipeline Co-authored-by: Frank Lee <somerlee.9@gmail.com> Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: binmakeswell <binmakeswell@gmail.com> * optimized 3d layer to fix slow computation ; tested imagenet performance with 3d; reworked lr_scheduler config definition; fixed launch args; fixed some printing issues; simplified apis of 3d layers (#51) * Update 2.5d layer code to get a similar accuracy on imagenet-1k dataset * update api for better usability (#58) update api for better usability Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: ver217 <lhx0217@gmail.com> Co-authored-by: puck_WCR <46049915+WANG-CR@users.noreply.github.com> Co-authored-by: binmakeswell <binmakeswell@gmail.com> Co-authored-by: アマデウス <kurisusnowdeng@users.noreply.github.com> Co-authored-by: BoxiangW <45734921+BoxiangW@users.noreply.github.com>
2021-12-09 07:08:29 +00:00
if skip_bias_add and not bias:
raise ValueError('cannot skip bias addition if bias is None')
self.out_features_per_partition = divide(out_features, gpc.tensor_parallel_size)
2021-10-28 16:21:23 +00:00
# Parameters.
# Initialize weight.
factory_kwargs = {'device': get_current_device(), 'dtype': dtype}
self.weight = Parameter(torch.empty(self.out_features_per_partition, self.in_features, **factory_kwargs))
2021-10-28 16:21:23 +00:00
if bias:
self.bias = Parameter(torch.empty(self.out_features_per_partition, **factory_kwargs))
2021-10-28 16:21:23 +00:00
else:
self.bias = None
Develop/experiments (#59) * Add gradient accumulation, fix lr scheduler * fix FP16 optimizer and adapted torch amp with tensor parallel (#18) * fixed bugs in compatibility between torch amp and tensor parallel and performed some minor fixes * fixed trainer * Revert "fixed trainer" This reverts commit 2e0b0b76990e8d4e337add483d878c0f61cf5097. * improved consistency between trainer, engine and schedule (#23) Co-authored-by: 1SAA <c2h214748@gmail.com> * Split conv2d, class token, positional embedding in 2d, Fix random number in ddp Fix convergence in cifar10, Imagenet1000 * Integrate 1d tensor parallel in Colossal-AI (#39) * fixed 1D and 2D convergence (#38) * optimized 2D operations * fixed 1D ViT convergence problem * Feature/ddp (#49) * remove redundancy func in setup (#19) (#20) * use env to control the language of doc (#24) (#25) * Support TP-compatible Torch AMP and Update trainer API (#27) * Add gradient accumulation, fix lr scheduler * fix FP16 optimizer and adapted torch amp with tensor parallel (#18) * fixed bugs in compatibility between torch amp and tensor parallel and performed some minor fixes * fixed trainer * Revert "fixed trainer" This reverts commit 2e0b0b76990e8d4e337add483d878c0f61cf5097. * improved consistency between trainer, engine and schedule (#23) Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: ver217 <lhx0217@gmail.com> * add an example of ViT-B/16 and remove w_norm clipping in LAMB (#29) * add explanation for ViT example (#35) (#36) * support torch ddp * fix loss accumulation * add log for ddp * change seed * modify timing hook Co-authored-by: Frank Lee <somerlee.9@gmail.com> Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: binmakeswell <binmakeswell@gmail.com> * Feature/pipeline (#40) * remove redundancy func in setup (#19) (#20) * use env to control the language of doc (#24) (#25) * Support TP-compatible Torch AMP and Update trainer API (#27) * Add gradient accumulation, fix lr scheduler * fix FP16 optimizer and adapted torch amp with tensor parallel (#18) * fixed bugs in compatibility between torch amp and tensor parallel and performed some minor fixes * fixed trainer * Revert "fixed trainer" This reverts commit 2e0b0b76990e8d4e337add483d878c0f61cf5097. * improved consistency between trainer, engine and schedule (#23) Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: ver217 <lhx0217@gmail.com> * add an example of ViT-B/16 and remove w_norm clipping in LAMB (#29) * add explanation for ViT example (#35) (#36) * optimize communication of pipeline parallel * fix grad clip for pipeline Co-authored-by: Frank Lee <somerlee.9@gmail.com> Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: binmakeswell <binmakeswell@gmail.com> * optimized 3d layer to fix slow computation ; tested imagenet performance with 3d; reworked lr_scheduler config definition; fixed launch args; fixed some printing issues; simplified apis of 3d layers (#51) * Update 2.5d layer code to get a similar accuracy on imagenet-1k dataset * update api for better usability (#58) update api for better usability Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: ver217 <lhx0217@gmail.com> Co-authored-by: puck_WCR <46049915+WANG-CR@users.noreply.github.com> Co-authored-by: binmakeswell <binmakeswell@gmail.com> Co-authored-by: アマデウス <kurisusnowdeng@users.noreply.github.com> Co-authored-by: BoxiangW <45734921+BoxiangW@users.noreply.github.com>
2021-12-09 07:08:29 +00:00
with seed(ParallelMode.TENSOR):
self.reset_parameters(weight_initializer, bias_initializer)
Develop/experiments (#59) * Add gradient accumulation, fix lr scheduler * fix FP16 optimizer and adapted torch amp with tensor parallel (#18) * fixed bugs in compatibility between torch amp and tensor parallel and performed some minor fixes * fixed trainer * Revert "fixed trainer" This reverts commit 2e0b0b76990e8d4e337add483d878c0f61cf5097. * improved consistency between trainer, engine and schedule (#23) Co-authored-by: 1SAA <c2h214748@gmail.com> * Split conv2d, class token, positional embedding in 2d, Fix random number in ddp Fix convergence in cifar10, Imagenet1000 * Integrate 1d tensor parallel in Colossal-AI (#39) * fixed 1D and 2D convergence (#38) * optimized 2D operations * fixed 1D ViT convergence problem * Feature/ddp (#49) * remove redundancy func in setup (#19) (#20) * use env to control the language of doc (#24) (#25) * Support TP-compatible Torch AMP and Update trainer API (#27) * Add gradient accumulation, fix lr scheduler * fix FP16 optimizer and adapted torch amp with tensor parallel (#18) * fixed bugs in compatibility between torch amp and tensor parallel and performed some minor fixes * fixed trainer * Revert "fixed trainer" This reverts commit 2e0b0b76990e8d4e337add483d878c0f61cf5097. * improved consistency between trainer, engine and schedule (#23) Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: ver217 <lhx0217@gmail.com> * add an example of ViT-B/16 and remove w_norm clipping in LAMB (#29) * add explanation for ViT example (#35) (#36) * support torch ddp * fix loss accumulation * add log for ddp * change seed * modify timing hook Co-authored-by: Frank Lee <somerlee.9@gmail.com> Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: binmakeswell <binmakeswell@gmail.com> * Feature/pipeline (#40) * remove redundancy func in setup (#19) (#20) * use env to control the language of doc (#24) (#25) * Support TP-compatible Torch AMP and Update trainer API (#27) * Add gradient accumulation, fix lr scheduler * fix FP16 optimizer and adapted torch amp with tensor parallel (#18) * fixed bugs in compatibility between torch amp and tensor parallel and performed some minor fixes * fixed trainer * Revert "fixed trainer" This reverts commit 2e0b0b76990e8d4e337add483d878c0f61cf5097. * improved consistency between trainer, engine and schedule (#23) Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: ver217 <lhx0217@gmail.com> * add an example of ViT-B/16 and remove w_norm clipping in LAMB (#29) * add explanation for ViT example (#35) (#36) * optimize communication of pipeline parallel * fix grad clip for pipeline Co-authored-by: Frank Lee <somerlee.9@gmail.com> Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: binmakeswell <binmakeswell@gmail.com> * optimized 3d layer to fix slow computation ; tested imagenet performance with 3d; reworked lr_scheduler config definition; fixed launch args; fixed some printing issues; simplified apis of 3d layers (#51) * Update 2.5d layer code to get a similar accuracy on imagenet-1k dataset * update api for better usability (#58) update api for better usability Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: ver217 <lhx0217@gmail.com> Co-authored-by: puck_WCR <46049915+WANG-CR@users.noreply.github.com> Co-authored-by: binmakeswell <binmakeswell@gmail.com> Co-authored-by: アマデウス <kurisusnowdeng@users.noreply.github.com> Co-authored-by: BoxiangW <45734921+BoxiangW@users.noreply.github.com>
2021-12-09 07:08:29 +00:00
self._set_tensor_parallel_attributes()
2022-03-31 09:11:03 +00:00
is_parallel_output = not self.gather_output
set_parallel_input(is_parallel_output)
Develop/experiments (#59) * Add gradient accumulation, fix lr scheduler * fix FP16 optimizer and adapted torch amp with tensor parallel (#18) * fixed bugs in compatibility between torch amp and tensor parallel and performed some minor fixes * fixed trainer * Revert "fixed trainer" This reverts commit 2e0b0b76990e8d4e337add483d878c0f61cf5097. * improved consistency between trainer, engine and schedule (#23) Co-authored-by: 1SAA <c2h214748@gmail.com> * Split conv2d, class token, positional embedding in 2d, Fix random number in ddp Fix convergence in cifar10, Imagenet1000 * Integrate 1d tensor parallel in Colossal-AI (#39) * fixed 1D and 2D convergence (#38) * optimized 2D operations * fixed 1D ViT convergence problem * Feature/ddp (#49) * remove redundancy func in setup (#19) (#20) * use env to control the language of doc (#24) (#25) * Support TP-compatible Torch AMP and Update trainer API (#27) * Add gradient accumulation, fix lr scheduler * fix FP16 optimizer and adapted torch amp with tensor parallel (#18) * fixed bugs in compatibility between torch amp and tensor parallel and performed some minor fixes * fixed trainer * Revert "fixed trainer" This reverts commit 2e0b0b76990e8d4e337add483d878c0f61cf5097. * improved consistency between trainer, engine and schedule (#23) Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: ver217 <lhx0217@gmail.com> * add an example of ViT-B/16 and remove w_norm clipping in LAMB (#29) * add explanation for ViT example (#35) (#36) * support torch ddp * fix loss accumulation * add log for ddp * change seed * modify timing hook Co-authored-by: Frank Lee <somerlee.9@gmail.com> Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: binmakeswell <binmakeswell@gmail.com> * Feature/pipeline (#40) * remove redundancy func in setup (#19) (#20) * use env to control the language of doc (#24) (#25) * Support TP-compatible Torch AMP and Update trainer API (#27) * Add gradient accumulation, fix lr scheduler * fix FP16 optimizer and adapted torch amp with tensor parallel (#18) * fixed bugs in compatibility between torch amp and tensor parallel and performed some minor fixes * fixed trainer * Revert "fixed trainer" This reverts commit 2e0b0b76990e8d4e337add483d878c0f61cf5097. * improved consistency between trainer, engine and schedule (#23) Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: ver217 <lhx0217@gmail.com> * add an example of ViT-B/16 and remove w_norm clipping in LAMB (#29) * add explanation for ViT example (#35) (#36) * optimize communication of pipeline parallel * fix grad clip for pipeline Co-authored-by: Frank Lee <somerlee.9@gmail.com> Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: binmakeswell <binmakeswell@gmail.com> * optimized 3d layer to fix slow computation ; tested imagenet performance with 3d; reworked lr_scheduler config definition; fixed launch args; fixed some printing issues; simplified apis of 3d layers (#51) * Update 2.5d layer code to get a similar accuracy on imagenet-1k dataset * update api for better usability (#58) update api for better usability Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: ver217 <lhx0217@gmail.com> Co-authored-by: puck_WCR <46049915+WANG-CR@users.noreply.github.com> Co-authored-by: binmakeswell <binmakeswell@gmail.com> Co-authored-by: アマデウス <kurisusnowdeng@users.noreply.github.com> Co-authored-by: BoxiangW <45734921+BoxiangW@users.noreply.github.com>
2021-12-09 07:08:29 +00:00
def reset_parameters(self, weight_initializer, bias_initializer) -> None:
Develop/experiments (#59) * Add gradient accumulation, fix lr scheduler * fix FP16 optimizer and adapted torch amp with tensor parallel (#18) * fixed bugs in compatibility between torch amp and tensor parallel and performed some minor fixes * fixed trainer * Revert "fixed trainer" This reverts commit 2e0b0b76990e8d4e337add483d878c0f61cf5097. * improved consistency between trainer, engine and schedule (#23) Co-authored-by: 1SAA <c2h214748@gmail.com> * Split conv2d, class token, positional embedding in 2d, Fix random number in ddp Fix convergence in cifar10, Imagenet1000 * Integrate 1d tensor parallel in Colossal-AI (#39) * fixed 1D and 2D convergence (#38) * optimized 2D operations * fixed 1D ViT convergence problem * Feature/ddp (#49) * remove redundancy func in setup (#19) (#20) * use env to control the language of doc (#24) (#25) * Support TP-compatible Torch AMP and Update trainer API (#27) * Add gradient accumulation, fix lr scheduler * fix FP16 optimizer and adapted torch amp with tensor parallel (#18) * fixed bugs in compatibility between torch amp and tensor parallel and performed some minor fixes * fixed trainer * Revert "fixed trainer" This reverts commit 2e0b0b76990e8d4e337add483d878c0f61cf5097. * improved consistency between trainer, engine and schedule (#23) Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: ver217 <lhx0217@gmail.com> * add an example of ViT-B/16 and remove w_norm clipping in LAMB (#29) * add explanation for ViT example (#35) (#36) * support torch ddp * fix loss accumulation * add log for ddp * change seed * modify timing hook Co-authored-by: Frank Lee <somerlee.9@gmail.com> Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: binmakeswell <binmakeswell@gmail.com> * Feature/pipeline (#40) * remove redundancy func in setup (#19) (#20) * use env to control the language of doc (#24) (#25) * Support TP-compatible Torch AMP and Update trainer API (#27) * Add gradient accumulation, fix lr scheduler * fix FP16 optimizer and adapted torch amp with tensor parallel (#18) * fixed bugs in compatibility between torch amp and tensor parallel and performed some minor fixes * fixed trainer * Revert "fixed trainer" This reverts commit 2e0b0b76990e8d4e337add483d878c0f61cf5097. * improved consistency between trainer, engine and schedule (#23) Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: ver217 <lhx0217@gmail.com> * add an example of ViT-B/16 and remove w_norm clipping in LAMB (#29) * add explanation for ViT example (#35) (#36) * optimize communication of pipeline parallel * fix grad clip for pipeline Co-authored-by: Frank Lee <somerlee.9@gmail.com> Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: binmakeswell <binmakeswell@gmail.com> * optimized 3d layer to fix slow computation ; tested imagenet performance with 3d; reworked lr_scheduler config definition; fixed launch args; fixed some printing issues; simplified apis of 3d layers (#51) * Update 2.5d layer code to get a similar accuracy on imagenet-1k dataset * update api for better usability (#58) update api for better usability Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: ver217 <lhx0217@gmail.com> Co-authored-by: puck_WCR <46049915+WANG-CR@users.noreply.github.com> Co-authored-by: binmakeswell <binmakeswell@gmail.com> Co-authored-by: アマデウス <kurisusnowdeng@users.noreply.github.com> Co-authored-by: BoxiangW <45734921+BoxiangW@users.noreply.github.com>
2021-12-09 07:08:29 +00:00
fan_in, fan_out = self.in_features, self.out_features
weight_initializer(self.weight, fan_in=fan_in, fan_out=fan_out)
Develop/experiments (#59) * Add gradient accumulation, fix lr scheduler * fix FP16 optimizer and adapted torch amp with tensor parallel (#18) * fixed bugs in compatibility between torch amp and tensor parallel and performed some minor fixes * fixed trainer * Revert "fixed trainer" This reverts commit 2e0b0b76990e8d4e337add483d878c0f61cf5097. * improved consistency between trainer, engine and schedule (#23) Co-authored-by: 1SAA <c2h214748@gmail.com> * Split conv2d, class token, positional embedding in 2d, Fix random number in ddp Fix convergence in cifar10, Imagenet1000 * Integrate 1d tensor parallel in Colossal-AI (#39) * fixed 1D and 2D convergence (#38) * optimized 2D operations * fixed 1D ViT convergence problem * Feature/ddp (#49) * remove redundancy func in setup (#19) (#20) * use env to control the language of doc (#24) (#25) * Support TP-compatible Torch AMP and Update trainer API (#27) * Add gradient accumulation, fix lr scheduler * fix FP16 optimizer and adapted torch amp with tensor parallel (#18) * fixed bugs in compatibility between torch amp and tensor parallel and performed some minor fixes * fixed trainer * Revert "fixed trainer" This reverts commit 2e0b0b76990e8d4e337add483d878c0f61cf5097. * improved consistency between trainer, engine and schedule (#23) Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: ver217 <lhx0217@gmail.com> * add an example of ViT-B/16 and remove w_norm clipping in LAMB (#29) * add explanation for ViT example (#35) (#36) * support torch ddp * fix loss accumulation * add log for ddp * change seed * modify timing hook Co-authored-by: Frank Lee <somerlee.9@gmail.com> Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: binmakeswell <binmakeswell@gmail.com> * Feature/pipeline (#40) * remove redundancy func in setup (#19) (#20) * use env to control the language of doc (#24) (#25) * Support TP-compatible Torch AMP and Update trainer API (#27) * Add gradient accumulation, fix lr scheduler * fix FP16 optimizer and adapted torch amp with tensor parallel (#18) * fixed bugs in compatibility between torch amp and tensor parallel and performed some minor fixes * fixed trainer * Revert "fixed trainer" This reverts commit 2e0b0b76990e8d4e337add483d878c0f61cf5097. * improved consistency between trainer, engine and schedule (#23) Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: ver217 <lhx0217@gmail.com> * add an example of ViT-B/16 and remove w_norm clipping in LAMB (#29) * add explanation for ViT example (#35) (#36) * optimize communication of pipeline parallel * fix grad clip for pipeline Co-authored-by: Frank Lee <somerlee.9@gmail.com> Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: binmakeswell <binmakeswell@gmail.com> * optimized 3d layer to fix slow computation ; tested imagenet performance with 3d; reworked lr_scheduler config definition; fixed launch args; fixed some printing issues; simplified apis of 3d layers (#51) * Update 2.5d layer code to get a similar accuracy on imagenet-1k dataset * update api for better usability (#58) update api for better usability Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: ver217 <lhx0217@gmail.com> Co-authored-by: puck_WCR <46049915+WANG-CR@users.noreply.github.com> Co-authored-by: binmakeswell <binmakeswell@gmail.com> Co-authored-by: アマデウス <kurisusnowdeng@users.noreply.github.com> Co-authored-by: BoxiangW <45734921+BoxiangW@users.noreply.github.com>
2021-12-09 07:08:29 +00:00
if self.bias is not None:
bias_initializer(self.bias, fan_in=fan_in)
Develop/experiments (#59) * Add gradient accumulation, fix lr scheduler * fix FP16 optimizer and adapted torch amp with tensor parallel (#18) * fixed bugs in compatibility between torch amp and tensor parallel and performed some minor fixes * fixed trainer * Revert "fixed trainer" This reverts commit 2e0b0b76990e8d4e337add483d878c0f61cf5097. * improved consistency between trainer, engine and schedule (#23) Co-authored-by: 1SAA <c2h214748@gmail.com> * Split conv2d, class token, positional embedding in 2d, Fix random number in ddp Fix convergence in cifar10, Imagenet1000 * Integrate 1d tensor parallel in Colossal-AI (#39) * fixed 1D and 2D convergence (#38) * optimized 2D operations * fixed 1D ViT convergence problem * Feature/ddp (#49) * remove redundancy func in setup (#19) (#20) * use env to control the language of doc (#24) (#25) * Support TP-compatible Torch AMP and Update trainer API (#27) * Add gradient accumulation, fix lr scheduler * fix FP16 optimizer and adapted torch amp with tensor parallel (#18) * fixed bugs in compatibility between torch amp and tensor parallel and performed some minor fixes * fixed trainer * Revert "fixed trainer" This reverts commit 2e0b0b76990e8d4e337add483d878c0f61cf5097. * improved consistency between trainer, engine and schedule (#23) Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: ver217 <lhx0217@gmail.com> * add an example of ViT-B/16 and remove w_norm clipping in LAMB (#29) * add explanation for ViT example (#35) (#36) * support torch ddp * fix loss accumulation * add log for ddp * change seed * modify timing hook Co-authored-by: Frank Lee <somerlee.9@gmail.com> Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: binmakeswell <binmakeswell@gmail.com> * Feature/pipeline (#40) * remove redundancy func in setup (#19) (#20) * use env to control the language of doc (#24) (#25) * Support TP-compatible Torch AMP and Update trainer API (#27) * Add gradient accumulation, fix lr scheduler * fix FP16 optimizer and adapted torch amp with tensor parallel (#18) * fixed bugs in compatibility between torch amp and tensor parallel and performed some minor fixes * fixed trainer * Revert "fixed trainer" This reverts commit 2e0b0b76990e8d4e337add483d878c0f61cf5097. * improved consistency between trainer, engine and schedule (#23) Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: ver217 <lhx0217@gmail.com> * add an example of ViT-B/16 and remove w_norm clipping in LAMB (#29) * add explanation for ViT example (#35) (#36) * optimize communication of pipeline parallel * fix grad clip for pipeline Co-authored-by: Frank Lee <somerlee.9@gmail.com> Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: binmakeswell <binmakeswell@gmail.com> * optimized 3d layer to fix slow computation ; tested imagenet performance with 3d; reworked lr_scheduler config definition; fixed launch args; fixed some printing issues; simplified apis of 3d layers (#51) * Update 2.5d layer code to get a similar accuracy on imagenet-1k dataset * update api for better usability (#58) update api for better usability Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: ver217 <lhx0217@gmail.com> Co-authored-by: puck_WCR <46049915+WANG-CR@users.noreply.github.com> Co-authored-by: binmakeswell <binmakeswell@gmail.com> Co-authored-by: アマデウス <kurisusnowdeng@users.noreply.github.com> Co-authored-by: BoxiangW <45734921+BoxiangW@users.noreply.github.com>
2021-12-09 07:08:29 +00:00
def _set_tensor_parallel_attributes(self):
num_partition = gpc.get_world_size(ParallelMode.TENSOR)
set_tensor_parallel_attribute_by_partition(self.weight, num_partition)
if self.bias is not None:
set_tensor_parallel_attribute_by_partition(self.bias, num_partition)
2021-10-28 16:21:23 +00:00
def _load_from_global_state_dict(self, state_dict, prefix, *args):
local_state = OrderedDict()
weight_key = prefix + 'weight'
bias_key = prefix + 'bias'
if gpc.get_local_rank(ParallelMode.TENSOR) == 0:
# weight
weight = state_dict.pop(weight_key, None)
if weight is not None:
local_state[weight_key] = weight
# bias
if self.bias is not None:
bias = state_dict.pop(bias_key, None)
if bias is not None:
local_state[bias_key] = bias
local_state = partition_tensor_parallel_state_dict(local_state,
ParallelMode.PARALLEL_1D,
dims={
weight_key: 0,
bias_key: 0
},
partition_states={
weight_key: True,
bias_key: True
})
super()._load_from_global_state_dict(local_state, prefix, *args)
def _save_to_global_state_dict(self, destination, prefix, keep_vars):
weight_key = prefix + 'weight'
bias_key = prefix + 'bias'
local_state = OrderedDict({weight_key: self.weight})
if self.bias is not None:
local_state[bias_key] = self.bias
local_state = gather_tensor_parallel_state_dict(local_state,
ParallelMode.PARALLEL_1D,
dims={
weight_key: 0,
bias_key: 0
},
partition_states={
weight_key: True,
bias_key: True
},
keep_vars=keep_vars)
destination.update(local_state)
2021-10-28 16:21:23 +00:00
def forward(self, input_: Tensor) -> Tuple[Tensor, Tensor]:
2022-04-01 08:44:23 +00:00
assert input_.shape[-1] == self.weight.shape[-1], \
'Invalid shapes in Linear1D_Col forward: input={}, weight={}. Expected last dim of input {}.'.format(
input_.shape, self.weight.shape, self.weight.shape[-1])
2021-10-28 16:21:23 +00:00
# Set up backprop all-reduce.
2022-10-26 12:54:39 +00:00
# input_parallel = reduce_grad(input_, ParallelMode.PARALLEL_1D)
input_parallel = input_
2021-10-28 16:21:23 +00:00
# Matrix multiply.
bias = self.bias if not self.skip_bias_add else None
2022-10-26 12:54:39 +00:00
# output_parallel = F.linear(input_parallel, self.weight, bias)
output_parallel = linear_with_async_comm(input_parallel, self.weight, bias, ParallelMode.PARALLEL_1D, True)
2021-10-28 16:21:23 +00:00
if self.gather_output:
# All-gather across the partitions.
output = gather_forward_split_backward(output_parallel, ParallelMode.PARALLEL_1D, dim=-1)
2021-10-28 16:21:23 +00:00
else:
output = output_parallel
2022-04-01 08:44:23 +00:00
2021-10-28 16:21:23 +00:00
if self.skip_bias_add:
return output, self.bias
else:
return output
@LAYERS.register_module
class Linear1D_Row(ParallelLayer):
2022-03-25 05:02:39 +00:00
r""" Linear layer with row parallelism
Args:
in_features (int): size of each input sample.
out_features (int): size of each output sample.
bias (bool, optional): If set to ``False``, the layer will not learn an additive bias, defaults to ``True``.
dtype (:class:`torch.dtype`, optional): The dtype of parameters, defaults to None.
parallel_input (bool, optional): If set to ``True``, it's assumed that the input is split, defaults to False.
skip_bias_add (bool, optional): If set to ``True``, it will skip bias add for linear layer,
which is preserved for kernel fusion, defaults to Fals
weight_initializer (:class:`typing.Callable`, optional):
The initializer of weight, defaults to kaiming uniform initializer.
bias_initializer (:class:`typing.Callable`, optional):
The initializer of bias, defaults to xavier uniform initializer.
More details about ``initializer`` please refer to
`init <https://github.com/hpcaitech/ColossalAI/blob/main/colossalai/nn/init.py>`_.
2021-10-28 16:21:23 +00:00
"""
2021-10-28 16:21:23 +00:00
def __init__(self,
in_features: int,
out_features: int,
bias: bool = True,
dtype: torch.dtype = None,
parallel_input: bool = True,
Develop/experiments (#59) * Add gradient accumulation, fix lr scheduler * fix FP16 optimizer and adapted torch amp with tensor parallel (#18) * fixed bugs in compatibility between torch amp and tensor parallel and performed some minor fixes * fixed trainer * Revert "fixed trainer" This reverts commit 2e0b0b76990e8d4e337add483d878c0f61cf5097. * improved consistency between trainer, engine and schedule (#23) Co-authored-by: 1SAA <c2h214748@gmail.com> * Split conv2d, class token, positional embedding in 2d, Fix random number in ddp Fix convergence in cifar10, Imagenet1000 * Integrate 1d tensor parallel in Colossal-AI (#39) * fixed 1D and 2D convergence (#38) * optimized 2D operations * fixed 1D ViT convergence problem * Feature/ddp (#49) * remove redundancy func in setup (#19) (#20) * use env to control the language of doc (#24) (#25) * Support TP-compatible Torch AMP and Update trainer API (#27) * Add gradient accumulation, fix lr scheduler * fix FP16 optimizer and adapted torch amp with tensor parallel (#18) * fixed bugs in compatibility between torch amp and tensor parallel and performed some minor fixes * fixed trainer * Revert "fixed trainer" This reverts commit 2e0b0b76990e8d4e337add483d878c0f61cf5097. * improved consistency between trainer, engine and schedule (#23) Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: ver217 <lhx0217@gmail.com> * add an example of ViT-B/16 and remove w_norm clipping in LAMB (#29) * add explanation for ViT example (#35) (#36) * support torch ddp * fix loss accumulation * add log for ddp * change seed * modify timing hook Co-authored-by: Frank Lee <somerlee.9@gmail.com> Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: binmakeswell <binmakeswell@gmail.com> * Feature/pipeline (#40) * remove redundancy func in setup (#19) (#20) * use env to control the language of doc (#24) (#25) * Support TP-compatible Torch AMP and Update trainer API (#27) * Add gradient accumulation, fix lr scheduler * fix FP16 optimizer and adapted torch amp with tensor parallel (#18) * fixed bugs in compatibility between torch amp and tensor parallel and performed some minor fixes * fixed trainer * Revert "fixed trainer" This reverts commit 2e0b0b76990e8d4e337add483d878c0f61cf5097. * improved consistency between trainer, engine and schedule (#23) Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: ver217 <lhx0217@gmail.com> * add an example of ViT-B/16 and remove w_norm clipping in LAMB (#29) * add explanation for ViT example (#35) (#36) * optimize communication of pipeline parallel * fix grad clip for pipeline Co-authored-by: Frank Lee <somerlee.9@gmail.com> Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: binmakeswell <binmakeswell@gmail.com> * optimized 3d layer to fix slow computation ; tested imagenet performance with 3d; reworked lr_scheduler config definition; fixed launch args; fixed some printing issues; simplified apis of 3d layers (#51) * Update 2.5d layer code to get a similar accuracy on imagenet-1k dataset * update api for better usability (#58) update api for better usability Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: ver217 <lhx0217@gmail.com> Co-authored-by: puck_WCR <46049915+WANG-CR@users.noreply.github.com> Co-authored-by: binmakeswell <binmakeswell@gmail.com> Co-authored-by: アマデウス <kurisusnowdeng@users.noreply.github.com> Co-authored-by: BoxiangW <45734921+BoxiangW@users.noreply.github.com>
2021-12-09 07:08:29 +00:00
skip_bias_add: bool = False,
weight_initializer: Callable = init.kaiming_uniform_(a=math.sqrt(5)),
bias_initializer: Callable = init.xavier_uniform_(a=1, scale=1),
stream_chunk_num: int = 1):
2021-10-28 16:21:23 +00:00
super().__init__()
self.stream_chunk_num = stream_chunk_num
2021-10-28 16:21:23 +00:00
# Keep input parameters
self.in_features = in_features
self.out_features = out_features
self.parallel_input = parallel_input
Develop/experiments (#59) * Add gradient accumulation, fix lr scheduler * fix FP16 optimizer and adapted torch amp with tensor parallel (#18) * fixed bugs in compatibility between torch amp and tensor parallel and performed some minor fixes * fixed trainer * Revert "fixed trainer" This reverts commit 2e0b0b76990e8d4e337add483d878c0f61cf5097. * improved consistency between trainer, engine and schedule (#23) Co-authored-by: 1SAA <c2h214748@gmail.com> * Split conv2d, class token, positional embedding in 2d, Fix random number in ddp Fix convergence in cifar10, Imagenet1000 * Integrate 1d tensor parallel in Colossal-AI (#39) * fixed 1D and 2D convergence (#38) * optimized 2D operations * fixed 1D ViT convergence problem * Feature/ddp (#49) * remove redundancy func in setup (#19) (#20) * use env to control the language of doc (#24) (#25) * Support TP-compatible Torch AMP and Update trainer API (#27) * Add gradient accumulation, fix lr scheduler * fix FP16 optimizer and adapted torch amp with tensor parallel (#18) * fixed bugs in compatibility between torch amp and tensor parallel and performed some minor fixes * fixed trainer * Revert "fixed trainer" This reverts commit 2e0b0b76990e8d4e337add483d878c0f61cf5097. * improved consistency between trainer, engine and schedule (#23) Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: ver217 <lhx0217@gmail.com> * add an example of ViT-B/16 and remove w_norm clipping in LAMB (#29) * add explanation for ViT example (#35) (#36) * support torch ddp * fix loss accumulation * add log for ddp * change seed * modify timing hook Co-authored-by: Frank Lee <somerlee.9@gmail.com> Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: binmakeswell <binmakeswell@gmail.com> * Feature/pipeline (#40) * remove redundancy func in setup (#19) (#20) * use env to control the language of doc (#24) (#25) * Support TP-compatible Torch AMP and Update trainer API (#27) * Add gradient accumulation, fix lr scheduler * fix FP16 optimizer and adapted torch amp with tensor parallel (#18) * fixed bugs in compatibility between torch amp and tensor parallel and performed some minor fixes * fixed trainer * Revert "fixed trainer" This reverts commit 2e0b0b76990e8d4e337add483d878c0f61cf5097. * improved consistency between trainer, engine and schedule (#23) Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: ver217 <lhx0217@gmail.com> * add an example of ViT-B/16 and remove w_norm clipping in LAMB (#29) * add explanation for ViT example (#35) (#36) * optimize communication of pipeline parallel * fix grad clip for pipeline Co-authored-by: Frank Lee <somerlee.9@gmail.com> Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: binmakeswell <binmakeswell@gmail.com> * optimized 3d layer to fix slow computation ; tested imagenet performance with 3d; reworked lr_scheduler config definition; fixed launch args; fixed some printing issues; simplified apis of 3d layers (#51) * Update 2.5d layer code to get a similar accuracy on imagenet-1k dataset * update api for better usability (#58) update api for better usability Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: ver217 <lhx0217@gmail.com> Co-authored-by: puck_WCR <46049915+WANG-CR@users.noreply.github.com> Co-authored-by: binmakeswell <binmakeswell@gmail.com> Co-authored-by: アマデウス <kurisusnowdeng@users.noreply.github.com> Co-authored-by: BoxiangW <45734921+BoxiangW@users.noreply.github.com>
2021-12-09 07:08:29 +00:00
self.skip_bias_add = skip_bias_add
if skip_bias_add and not bias:
raise ValueError('cannot skip bias addition if bias is None')
2021-10-28 16:21:23 +00:00
# Divide the weight matrix along the last dimension.
Develop/experiments (#59) * Add gradient accumulation, fix lr scheduler * fix FP16 optimizer and adapted torch amp with tensor parallel (#18) * fixed bugs in compatibility between torch amp and tensor parallel and performed some minor fixes * fixed trainer * Revert "fixed trainer" This reverts commit 2e0b0b76990e8d4e337add483d878c0f61cf5097. * improved consistency between trainer, engine and schedule (#23) Co-authored-by: 1SAA <c2h214748@gmail.com> * Split conv2d, class token, positional embedding in 2d, Fix random number in ddp Fix convergence in cifar10, Imagenet1000 * Integrate 1d tensor parallel in Colossal-AI (#39) * fixed 1D and 2D convergence (#38) * optimized 2D operations * fixed 1D ViT convergence problem * Feature/ddp (#49) * remove redundancy func in setup (#19) (#20) * use env to control the language of doc (#24) (#25) * Support TP-compatible Torch AMP and Update trainer API (#27) * Add gradient accumulation, fix lr scheduler * fix FP16 optimizer and adapted torch amp with tensor parallel (#18) * fixed bugs in compatibility between torch amp and tensor parallel and performed some minor fixes * fixed trainer * Revert "fixed trainer" This reverts commit 2e0b0b76990e8d4e337add483d878c0f61cf5097. * improved consistency between trainer, engine and schedule (#23) Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: ver217 <lhx0217@gmail.com> * add an example of ViT-B/16 and remove w_norm clipping in LAMB (#29) * add explanation for ViT example (#35) (#36) * support torch ddp * fix loss accumulation * add log for ddp * change seed * modify timing hook Co-authored-by: Frank Lee <somerlee.9@gmail.com> Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: binmakeswell <binmakeswell@gmail.com> * Feature/pipeline (#40) * remove redundancy func in setup (#19) (#20) * use env to control the language of doc (#24) (#25) * Support TP-compatible Torch AMP and Update trainer API (#27) * Add gradient accumulation, fix lr scheduler * fix FP16 optimizer and adapted torch amp with tensor parallel (#18) * fixed bugs in compatibility between torch amp and tensor parallel and performed some minor fixes * fixed trainer * Revert "fixed trainer" This reverts commit 2e0b0b76990e8d4e337add483d878c0f61cf5097. * improved consistency between trainer, engine and schedule (#23) Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: ver217 <lhx0217@gmail.com> * add an example of ViT-B/16 and remove w_norm clipping in LAMB (#29) * add explanation for ViT example (#35) (#36) * optimize communication of pipeline parallel * fix grad clip for pipeline Co-authored-by: Frank Lee <somerlee.9@gmail.com> Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: binmakeswell <binmakeswell@gmail.com> * optimized 3d layer to fix slow computation ; tested imagenet performance with 3d; reworked lr_scheduler config definition; fixed launch args; fixed some printing issues; simplified apis of 3d layers (#51) * Update 2.5d layer code to get a similar accuracy on imagenet-1k dataset * update api for better usability (#58) update api for better usability Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: ver217 <lhx0217@gmail.com> Co-authored-by: puck_WCR <46049915+WANG-CR@users.noreply.github.com> Co-authored-by: binmakeswell <binmakeswell@gmail.com> Co-authored-by: アマデウス <kurisusnowdeng@users.noreply.github.com> Co-authored-by: BoxiangW <45734921+BoxiangW@users.noreply.github.com>
2021-12-09 07:08:29 +00:00
self.input_size_per_partition = divide(in_features, gpc.tensor_parallel_size)
2021-10-28 16:21:23 +00:00
# Parameters.
# Initialize weight.
factory_kwargs = {'device': get_current_device(), 'dtype': dtype}
self.weight = Parameter(torch.empty(self.out_features, self.input_size_per_partition, **factory_kwargs))
2021-10-28 16:21:23 +00:00
if self.stream_chunk_num > 1:
# TODO() work for inference only
self.chunk_weight()
2021-10-28 16:21:23 +00:00
if bias:
self.bias = Parameter(torch.empty(self.out_features, **factory_kwargs))
2021-10-28 16:21:23 +00:00
else:
self.bias = None
Develop/experiments (#59) * Add gradient accumulation, fix lr scheduler * fix FP16 optimizer and adapted torch amp with tensor parallel (#18) * fixed bugs in compatibility between torch amp and tensor parallel and performed some minor fixes * fixed trainer * Revert "fixed trainer" This reverts commit 2e0b0b76990e8d4e337add483d878c0f61cf5097. * improved consistency between trainer, engine and schedule (#23) Co-authored-by: 1SAA <c2h214748@gmail.com> * Split conv2d, class token, positional embedding in 2d, Fix random number in ddp Fix convergence in cifar10, Imagenet1000 * Integrate 1d tensor parallel in Colossal-AI (#39) * fixed 1D and 2D convergence (#38) * optimized 2D operations * fixed 1D ViT convergence problem * Feature/ddp (#49) * remove redundancy func in setup (#19) (#20) * use env to control the language of doc (#24) (#25) * Support TP-compatible Torch AMP and Update trainer API (#27) * Add gradient accumulation, fix lr scheduler * fix FP16 optimizer and adapted torch amp with tensor parallel (#18) * fixed bugs in compatibility between torch amp and tensor parallel and performed some minor fixes * fixed trainer * Revert "fixed trainer" This reverts commit 2e0b0b76990e8d4e337add483d878c0f61cf5097. * improved consistency between trainer, engine and schedule (#23) Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: ver217 <lhx0217@gmail.com> * add an example of ViT-B/16 and remove w_norm clipping in LAMB (#29) * add explanation for ViT example (#35) (#36) * support torch ddp * fix loss accumulation * add log for ddp * change seed * modify timing hook Co-authored-by: Frank Lee <somerlee.9@gmail.com> Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: binmakeswell <binmakeswell@gmail.com> * Feature/pipeline (#40) * remove redundancy func in setup (#19) (#20) * use env to control the language of doc (#24) (#25) * Support TP-compatible Torch AMP and Update trainer API (#27) * Add gradient accumulation, fix lr scheduler * fix FP16 optimizer and adapted torch amp with tensor parallel (#18) * fixed bugs in compatibility between torch amp and tensor parallel and performed some minor fixes * fixed trainer * Revert "fixed trainer" This reverts commit 2e0b0b76990e8d4e337add483d878c0f61cf5097. * improved consistency between trainer, engine and schedule (#23) Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: ver217 <lhx0217@gmail.com> * add an example of ViT-B/16 and remove w_norm clipping in LAMB (#29) * add explanation for ViT example (#35) (#36) * optimize communication of pipeline parallel * fix grad clip for pipeline Co-authored-by: Frank Lee <somerlee.9@gmail.com> Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: binmakeswell <binmakeswell@gmail.com> * optimized 3d layer to fix slow computation ; tested imagenet performance with 3d; reworked lr_scheduler config definition; fixed launch args; fixed some printing issues; simplified apis of 3d layers (#51) * Update 2.5d layer code to get a similar accuracy on imagenet-1k dataset * update api for better usability (#58) update api for better usability Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: ver217 <lhx0217@gmail.com> Co-authored-by: puck_WCR <46049915+WANG-CR@users.noreply.github.com> Co-authored-by: binmakeswell <binmakeswell@gmail.com> Co-authored-by: アマデウス <kurisusnowdeng@users.noreply.github.com> Co-authored-by: BoxiangW <45734921+BoxiangW@users.noreply.github.com>
2021-12-09 07:08:29 +00:00
with seed(ParallelMode.TENSOR):
self.reset_parameters(weight_initializer, bias_initializer)
Develop/experiments (#59) * Add gradient accumulation, fix lr scheduler * fix FP16 optimizer and adapted torch amp with tensor parallel (#18) * fixed bugs in compatibility between torch amp and tensor parallel and performed some minor fixes * fixed trainer * Revert "fixed trainer" This reverts commit 2e0b0b76990e8d4e337add483d878c0f61cf5097. * improved consistency between trainer, engine and schedule (#23) Co-authored-by: 1SAA <c2h214748@gmail.com> * Split conv2d, class token, positional embedding in 2d, Fix random number in ddp Fix convergence in cifar10, Imagenet1000 * Integrate 1d tensor parallel in Colossal-AI (#39) * fixed 1D and 2D convergence (#38) * optimized 2D operations * fixed 1D ViT convergence problem * Feature/ddp (#49) * remove redundancy func in setup (#19) (#20) * use env to control the language of doc (#24) (#25) * Support TP-compatible Torch AMP and Update trainer API (#27) * Add gradient accumulation, fix lr scheduler * fix FP16 optimizer and adapted torch amp with tensor parallel (#18) * fixed bugs in compatibility between torch amp and tensor parallel and performed some minor fixes * fixed trainer * Revert "fixed trainer" This reverts commit 2e0b0b76990e8d4e337add483d878c0f61cf5097. * improved consistency between trainer, engine and schedule (#23) Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: ver217 <lhx0217@gmail.com> * add an example of ViT-B/16 and remove w_norm clipping in LAMB (#29) * add explanation for ViT example (#35) (#36) * support torch ddp * fix loss accumulation * add log for ddp * change seed * modify timing hook Co-authored-by: Frank Lee <somerlee.9@gmail.com> Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: binmakeswell <binmakeswell@gmail.com> * Feature/pipeline (#40) * remove redundancy func in setup (#19) (#20) * use env to control the language of doc (#24) (#25) * Support TP-compatible Torch AMP and Update trainer API (#27) * Add gradient accumulation, fix lr scheduler * fix FP16 optimizer and adapted torch amp with tensor parallel (#18) * fixed bugs in compatibility between torch amp and tensor parallel and performed some minor fixes * fixed trainer * Revert "fixed trainer" This reverts commit 2e0b0b76990e8d4e337add483d878c0f61cf5097. * improved consistency between trainer, engine and schedule (#23) Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: ver217 <lhx0217@gmail.com> * add an example of ViT-B/16 and remove w_norm clipping in LAMB (#29) * add explanation for ViT example (#35) (#36) * optimize communication of pipeline parallel * fix grad clip for pipeline Co-authored-by: Frank Lee <somerlee.9@gmail.com> Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: binmakeswell <binmakeswell@gmail.com> * optimized 3d layer to fix slow computation ; tested imagenet performance with 3d; reworked lr_scheduler config definition; fixed launch args; fixed some printing issues; simplified apis of 3d layers (#51) * Update 2.5d layer code to get a similar accuracy on imagenet-1k dataset * update api for better usability (#58) update api for better usability Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: ver217 <lhx0217@gmail.com> Co-authored-by: puck_WCR <46049915+WANG-CR@users.noreply.github.com> Co-authored-by: binmakeswell <binmakeswell@gmail.com> Co-authored-by: アマデウス <kurisusnowdeng@users.noreply.github.com> Co-authored-by: BoxiangW <45734921+BoxiangW@users.noreply.github.com>
2021-12-09 07:08:29 +00:00
self._set_tensor_parallel_attributes()
set_parallel_input(False)
Develop/experiments (#59) * Add gradient accumulation, fix lr scheduler * fix FP16 optimizer and adapted torch amp with tensor parallel (#18) * fixed bugs in compatibility between torch amp and tensor parallel and performed some minor fixes * fixed trainer * Revert "fixed trainer" This reverts commit 2e0b0b76990e8d4e337add483d878c0f61cf5097. * improved consistency between trainer, engine and schedule (#23) Co-authored-by: 1SAA <c2h214748@gmail.com> * Split conv2d, class token, positional embedding in 2d, Fix random number in ddp Fix convergence in cifar10, Imagenet1000 * Integrate 1d tensor parallel in Colossal-AI (#39) * fixed 1D and 2D convergence (#38) * optimized 2D operations * fixed 1D ViT convergence problem * Feature/ddp (#49) * remove redundancy func in setup (#19) (#20) * use env to control the language of doc (#24) (#25) * Support TP-compatible Torch AMP and Update trainer API (#27) * Add gradient accumulation, fix lr scheduler * fix FP16 optimizer and adapted torch amp with tensor parallel (#18) * fixed bugs in compatibility between torch amp and tensor parallel and performed some minor fixes * fixed trainer * Revert "fixed trainer" This reverts commit 2e0b0b76990e8d4e337add483d878c0f61cf5097. * improved consistency between trainer, engine and schedule (#23) Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: ver217 <lhx0217@gmail.com> * add an example of ViT-B/16 and remove w_norm clipping in LAMB (#29) * add explanation for ViT example (#35) (#36) * support torch ddp * fix loss accumulation * add log for ddp * change seed * modify timing hook Co-authored-by: Frank Lee <somerlee.9@gmail.com> Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: binmakeswell <binmakeswell@gmail.com> * Feature/pipeline (#40) * remove redundancy func in setup (#19) (#20) * use env to control the language of doc (#24) (#25) * Support TP-compatible Torch AMP and Update trainer API (#27) * Add gradient accumulation, fix lr scheduler * fix FP16 optimizer and adapted torch amp with tensor parallel (#18) * fixed bugs in compatibility between torch amp and tensor parallel and performed some minor fixes * fixed trainer * Revert "fixed trainer" This reverts commit 2e0b0b76990e8d4e337add483d878c0f61cf5097. * improved consistency between trainer, engine and schedule (#23) Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: ver217 <lhx0217@gmail.com> * add an example of ViT-B/16 and remove w_norm clipping in LAMB (#29) * add explanation for ViT example (#35) (#36) * optimize communication of pipeline parallel * fix grad clip for pipeline Co-authored-by: Frank Lee <somerlee.9@gmail.com> Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: binmakeswell <binmakeswell@gmail.com> * optimized 3d layer to fix slow computation ; tested imagenet performance with 3d; reworked lr_scheduler config definition; fixed launch args; fixed some printing issues; simplified apis of 3d layers (#51) * Update 2.5d layer code to get a similar accuracy on imagenet-1k dataset * update api for better usability (#58) update api for better usability Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: ver217 <lhx0217@gmail.com> Co-authored-by: puck_WCR <46049915+WANG-CR@users.noreply.github.com> Co-authored-by: binmakeswell <binmakeswell@gmail.com> Co-authored-by: アマデウス <kurisusnowdeng@users.noreply.github.com> Co-authored-by: BoxiangW <45734921+BoxiangW@users.noreply.github.com>
2021-12-09 07:08:29 +00:00
def chunk_weight(self):
self.weight_list = torch.chunk(self.weight, self.stream_chunk_num, dim=0)
def reset_parameters(self, weight_initializer, bias_initializer) -> None:
Develop/experiments (#59) * Add gradient accumulation, fix lr scheduler * fix FP16 optimizer and adapted torch amp with tensor parallel (#18) * fixed bugs in compatibility between torch amp and tensor parallel and performed some minor fixes * fixed trainer * Revert "fixed trainer" This reverts commit 2e0b0b76990e8d4e337add483d878c0f61cf5097. * improved consistency between trainer, engine and schedule (#23) Co-authored-by: 1SAA <c2h214748@gmail.com> * Split conv2d, class token, positional embedding in 2d, Fix random number in ddp Fix convergence in cifar10, Imagenet1000 * Integrate 1d tensor parallel in Colossal-AI (#39) * fixed 1D and 2D convergence (#38) * optimized 2D operations * fixed 1D ViT convergence problem * Feature/ddp (#49) * remove redundancy func in setup (#19) (#20) * use env to control the language of doc (#24) (#25) * Support TP-compatible Torch AMP and Update trainer API (#27) * Add gradient accumulation, fix lr scheduler * fix FP16 optimizer and adapted torch amp with tensor parallel (#18) * fixed bugs in compatibility between torch amp and tensor parallel and performed some minor fixes * fixed trainer * Revert "fixed trainer" This reverts commit 2e0b0b76990e8d4e337add483d878c0f61cf5097. * improved consistency between trainer, engine and schedule (#23) Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: ver217 <lhx0217@gmail.com> * add an example of ViT-B/16 and remove w_norm clipping in LAMB (#29) * add explanation for ViT example (#35) (#36) * support torch ddp * fix loss accumulation * add log for ddp * change seed * modify timing hook Co-authored-by: Frank Lee <somerlee.9@gmail.com> Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: binmakeswell <binmakeswell@gmail.com> * Feature/pipeline (#40) * remove redundancy func in setup (#19) (#20) * use env to control the language of doc (#24) (#25) * Support TP-compatible Torch AMP and Update trainer API (#27) * Add gradient accumulation, fix lr scheduler * fix FP16 optimizer and adapted torch amp with tensor parallel (#18) * fixed bugs in compatibility between torch amp and tensor parallel and performed some minor fixes * fixed trainer * Revert "fixed trainer" This reverts commit 2e0b0b76990e8d4e337add483d878c0f61cf5097. * improved consistency between trainer, engine and schedule (#23) Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: ver217 <lhx0217@gmail.com> * add an example of ViT-B/16 and remove w_norm clipping in LAMB (#29) * add explanation for ViT example (#35) (#36) * optimize communication of pipeline parallel * fix grad clip for pipeline Co-authored-by: Frank Lee <somerlee.9@gmail.com> Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: binmakeswell <binmakeswell@gmail.com> * optimized 3d layer to fix slow computation ; tested imagenet performance with 3d; reworked lr_scheduler config definition; fixed launch args; fixed some printing issues; simplified apis of 3d layers (#51) * Update 2.5d layer code to get a similar accuracy on imagenet-1k dataset * update api for better usability (#58) update api for better usability Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: ver217 <lhx0217@gmail.com> Co-authored-by: puck_WCR <46049915+WANG-CR@users.noreply.github.com> Co-authored-by: binmakeswell <binmakeswell@gmail.com> Co-authored-by: アマデウス <kurisusnowdeng@users.noreply.github.com> Co-authored-by: BoxiangW <45734921+BoxiangW@users.noreply.github.com>
2021-12-09 07:08:29 +00:00
fan_in, fan_out = self.in_features, self.out_features
weight_initializer(self.weight, fan_in=fan_in, fan_out=fan_out)
Develop/experiments (#59) * Add gradient accumulation, fix lr scheduler * fix FP16 optimizer and adapted torch amp with tensor parallel (#18) * fixed bugs in compatibility between torch amp and tensor parallel and performed some minor fixes * fixed trainer * Revert "fixed trainer" This reverts commit 2e0b0b76990e8d4e337add483d878c0f61cf5097. * improved consistency between trainer, engine and schedule (#23) Co-authored-by: 1SAA <c2h214748@gmail.com> * Split conv2d, class token, positional embedding in 2d, Fix random number in ddp Fix convergence in cifar10, Imagenet1000 * Integrate 1d tensor parallel in Colossal-AI (#39) * fixed 1D and 2D convergence (#38) * optimized 2D operations * fixed 1D ViT convergence problem * Feature/ddp (#49) * remove redundancy func in setup (#19) (#20) * use env to control the language of doc (#24) (#25) * Support TP-compatible Torch AMP and Update trainer API (#27) * Add gradient accumulation, fix lr scheduler * fix FP16 optimizer and adapted torch amp with tensor parallel (#18) * fixed bugs in compatibility between torch amp and tensor parallel and performed some minor fixes * fixed trainer * Revert "fixed trainer" This reverts commit 2e0b0b76990e8d4e337add483d878c0f61cf5097. * improved consistency between trainer, engine and schedule (#23) Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: ver217 <lhx0217@gmail.com> * add an example of ViT-B/16 and remove w_norm clipping in LAMB (#29) * add explanation for ViT example (#35) (#36) * support torch ddp * fix loss accumulation * add log for ddp * change seed * modify timing hook Co-authored-by: Frank Lee <somerlee.9@gmail.com> Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: binmakeswell <binmakeswell@gmail.com> * Feature/pipeline (#40) * remove redundancy func in setup (#19) (#20) * use env to control the language of doc (#24) (#25) * Support TP-compatible Torch AMP and Update trainer API (#27) * Add gradient accumulation, fix lr scheduler * fix FP16 optimizer and adapted torch amp with tensor parallel (#18) * fixed bugs in compatibility between torch amp and tensor parallel and performed some minor fixes * fixed trainer * Revert "fixed trainer" This reverts commit 2e0b0b76990e8d4e337add483d878c0f61cf5097. * improved consistency between trainer, engine and schedule (#23) Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: ver217 <lhx0217@gmail.com> * add an example of ViT-B/16 and remove w_norm clipping in LAMB (#29) * add explanation for ViT example (#35) (#36) * optimize communication of pipeline parallel * fix grad clip for pipeline Co-authored-by: Frank Lee <somerlee.9@gmail.com> Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: binmakeswell <binmakeswell@gmail.com> * optimized 3d layer to fix slow computation ; tested imagenet performance with 3d; reworked lr_scheduler config definition; fixed launch args; fixed some printing issues; simplified apis of 3d layers (#51) * Update 2.5d layer code to get a similar accuracy on imagenet-1k dataset * update api for better usability (#58) update api for better usability Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: ver217 <lhx0217@gmail.com> Co-authored-by: puck_WCR <46049915+WANG-CR@users.noreply.github.com> Co-authored-by: binmakeswell <binmakeswell@gmail.com> Co-authored-by: アマデウス <kurisusnowdeng@users.noreply.github.com> Co-authored-by: BoxiangW <45734921+BoxiangW@users.noreply.github.com>
2021-12-09 07:08:29 +00:00
if self.bias is not None:
bias_initializer(self.bias, fan_in=fan_in)
broadcast(self.bias, gpc.get_ranks_in_group(ParallelMode.PARALLEL_1D)[0], ParallelMode.PARALLEL_1D)
Develop/experiments (#59) * Add gradient accumulation, fix lr scheduler * fix FP16 optimizer and adapted torch amp with tensor parallel (#18) * fixed bugs in compatibility between torch amp and tensor parallel and performed some minor fixes * fixed trainer * Revert "fixed trainer" This reverts commit 2e0b0b76990e8d4e337add483d878c0f61cf5097. * improved consistency between trainer, engine and schedule (#23) Co-authored-by: 1SAA <c2h214748@gmail.com> * Split conv2d, class token, positional embedding in 2d, Fix random number in ddp Fix convergence in cifar10, Imagenet1000 * Integrate 1d tensor parallel in Colossal-AI (#39) * fixed 1D and 2D convergence (#38) * optimized 2D operations * fixed 1D ViT convergence problem * Feature/ddp (#49) * remove redundancy func in setup (#19) (#20) * use env to control the language of doc (#24) (#25) * Support TP-compatible Torch AMP and Update trainer API (#27) * Add gradient accumulation, fix lr scheduler * fix FP16 optimizer and adapted torch amp with tensor parallel (#18) * fixed bugs in compatibility between torch amp and tensor parallel and performed some minor fixes * fixed trainer * Revert "fixed trainer" This reverts commit 2e0b0b76990e8d4e337add483d878c0f61cf5097. * improved consistency between trainer, engine and schedule (#23) Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: ver217 <lhx0217@gmail.com> * add an example of ViT-B/16 and remove w_norm clipping in LAMB (#29) * add explanation for ViT example (#35) (#36) * support torch ddp * fix loss accumulation * add log for ddp * change seed * modify timing hook Co-authored-by: Frank Lee <somerlee.9@gmail.com> Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: binmakeswell <binmakeswell@gmail.com> * Feature/pipeline (#40) * remove redundancy func in setup (#19) (#20) * use env to control the language of doc (#24) (#25) * Support TP-compatible Torch AMP and Update trainer API (#27) * Add gradient accumulation, fix lr scheduler * fix FP16 optimizer and adapted torch amp with tensor parallel (#18) * fixed bugs in compatibility between torch amp and tensor parallel and performed some minor fixes * fixed trainer * Revert "fixed trainer" This reverts commit 2e0b0b76990e8d4e337add483d878c0f61cf5097. * improved consistency between trainer, engine and schedule (#23) Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: ver217 <lhx0217@gmail.com> * add an example of ViT-B/16 and remove w_norm clipping in LAMB (#29) * add explanation for ViT example (#35) (#36) * optimize communication of pipeline parallel * fix grad clip for pipeline Co-authored-by: Frank Lee <somerlee.9@gmail.com> Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: binmakeswell <binmakeswell@gmail.com> * optimized 3d layer to fix slow computation ; tested imagenet performance with 3d; reworked lr_scheduler config definition; fixed launch args; fixed some printing issues; simplified apis of 3d layers (#51) * Update 2.5d layer code to get a similar accuracy on imagenet-1k dataset * update api for better usability (#58) update api for better usability Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: ver217 <lhx0217@gmail.com> Co-authored-by: puck_WCR <46049915+WANG-CR@users.noreply.github.com> Co-authored-by: binmakeswell <binmakeswell@gmail.com> Co-authored-by: アマデウス <kurisusnowdeng@users.noreply.github.com> Co-authored-by: BoxiangW <45734921+BoxiangW@users.noreply.github.com>
2021-12-09 07:08:29 +00:00
def _set_tensor_parallel_attributes(self):
num_partition = gpc.get_world_size(ParallelMode.TENSOR)
set_tensor_parallel_attribute_by_partition(self.weight, num_partition)
2021-10-28 16:21:23 +00:00
def _load_from_global_state_dict(self, state_dict, prefix, *args):
local_state = OrderedDict()
weight_key = prefix + 'weight'
bias_key = prefix + 'bias'
if gpc.get_local_rank(ParallelMode.TENSOR) == 0:
# weight
weight = state_dict.pop(weight_key, None)
if weight is not None:
local_state[weight_key] = weight
# bias
if self.bias is not None:
bias = state_dict.pop(bias_key, None)
if bias is not None:
local_state[bias_key] = bias
local_state = partition_tensor_parallel_state_dict(local_state,
ParallelMode.PARALLEL_1D,
dims={
weight_key: -1,
bias_key: 0
},
partition_states={
weight_key: True,
bias_key: False
})
super()._load_from_global_state_dict(local_state, prefix, *args)
def _save_to_global_state_dict(self, destination, prefix, keep_vars):
weight_key = prefix + 'weight'
bias_key = prefix + 'bias'
local_state = OrderedDict({weight_key: self.weight})
if self.bias is not None:
local_state[bias_key] = self.bias
local_state = gather_tensor_parallel_state_dict(local_state,
ParallelMode.PARALLEL_1D,
dims={
weight_key: -1,
bias_key: 0
},
partition_states={
weight_key: True,
bias_key: False
},
keep_vars=keep_vars)
destination.update(local_state)
2021-10-28 16:21:23 +00:00
def forward(self, input_: Tensor) -> Tensor:
# Set up backprop all-reduce.
if self.parallel_input:
2022-04-01 08:44:23 +00:00
assert input_.shape[-1] == self.weight.shape[-1], \
'Invalid shapes in Linear1D_Row forward: input={}, weight={}. Expected last dim of input {}.'.format(
2022-04-01 08:44:23 +00:00
input_.shape, self.weight.shape, self.weight.shape[-1])
2021-10-28 16:21:23 +00:00
input_ = input_
else:
2022-04-01 08:44:23 +00:00
assert divide(input_.shape[-1], gpc.tensor_parallel_size) == self.weight.shape[-1], \
'Invalid shapes in Linear1D_Row forward: input={}, weight={}. Expected last dim of input {}.'.format(
2022-04-01 08:44:23 +00:00
input_.shape, self.weight.shape, self.weight.shape[-1] * gpc.tensor_parallel_size)
input_ = split_forward_gather_backward(input_, ParallelMode.PARALLEL_1D, dim=-1)
2021-10-28 16:21:23 +00:00
if self.stream_chunk_num > 1:
if self.training:
raise RuntimeError("use stream_chunk_num=1 in Linear1D_Row for training!")
with torch.no_grad():
output_parallel_list = [None for i in range(self.stream_chunk_num)]
handle_list = []
for i in range(self.stream_chunk_num):
output_parallel_list[i] = F.linear(input_, self.weight_list[i])
handle = torch.distributed.all_reduce(output_parallel_list[i],
group=gpc.get_group(ParallelMode.PARALLEL_1D),
async_op=True)
handle_list.append(handle)
# output_parallel_list[i] = reduce_input(output_parallel_list[i], ParallelMode.PARALLEL_1D)
for handle in handle_list:
handle.wait()
output = torch.cat(output_parallel_list, dim=-1)
else:
output_parallel = F.linear(input_, self.weight)
# output_parallel = linear_with_async_comm(input_, self.weight, None, ParallelMode.PARALLEL_1D, False)
output = reduce_input(output_parallel, ParallelMode.PARALLEL_1D)
2021-10-28 16:21:23 +00:00
if not self.skip_bias_add:
if self.bias is not None:
output = output + self.bias
Develop/experiments (#59) * Add gradient accumulation, fix lr scheduler * fix FP16 optimizer and adapted torch amp with tensor parallel (#18) * fixed bugs in compatibility between torch amp and tensor parallel and performed some minor fixes * fixed trainer * Revert "fixed trainer" This reverts commit 2e0b0b76990e8d4e337add483d878c0f61cf5097. * improved consistency between trainer, engine and schedule (#23) Co-authored-by: 1SAA <c2h214748@gmail.com> * Split conv2d, class token, positional embedding in 2d, Fix random number in ddp Fix convergence in cifar10, Imagenet1000 * Integrate 1d tensor parallel in Colossal-AI (#39) * fixed 1D and 2D convergence (#38) * optimized 2D operations * fixed 1D ViT convergence problem * Feature/ddp (#49) * remove redundancy func in setup (#19) (#20) * use env to control the language of doc (#24) (#25) * Support TP-compatible Torch AMP and Update trainer API (#27) * Add gradient accumulation, fix lr scheduler * fix FP16 optimizer and adapted torch amp with tensor parallel (#18) * fixed bugs in compatibility between torch amp and tensor parallel and performed some minor fixes * fixed trainer * Revert "fixed trainer" This reverts commit 2e0b0b76990e8d4e337add483d878c0f61cf5097. * improved consistency between trainer, engine and schedule (#23) Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: ver217 <lhx0217@gmail.com> * add an example of ViT-B/16 and remove w_norm clipping in LAMB (#29) * add explanation for ViT example (#35) (#36) * support torch ddp * fix loss accumulation * add log for ddp * change seed * modify timing hook Co-authored-by: Frank Lee <somerlee.9@gmail.com> Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: binmakeswell <binmakeswell@gmail.com> * Feature/pipeline (#40) * remove redundancy func in setup (#19) (#20) * use env to control the language of doc (#24) (#25) * Support TP-compatible Torch AMP and Update trainer API (#27) * Add gradient accumulation, fix lr scheduler * fix FP16 optimizer and adapted torch amp with tensor parallel (#18) * fixed bugs in compatibility between torch amp and tensor parallel and performed some minor fixes * fixed trainer * Revert "fixed trainer" This reverts commit 2e0b0b76990e8d4e337add483d878c0f61cf5097. * improved consistency between trainer, engine and schedule (#23) Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: ver217 <lhx0217@gmail.com> * add an example of ViT-B/16 and remove w_norm clipping in LAMB (#29) * add explanation for ViT example (#35) (#36) * optimize communication of pipeline parallel * fix grad clip for pipeline Co-authored-by: Frank Lee <somerlee.9@gmail.com> Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: binmakeswell <binmakeswell@gmail.com> * optimized 3d layer to fix slow computation ; tested imagenet performance with 3d; reworked lr_scheduler config definition; fixed launch args; fixed some printing issues; simplified apis of 3d layers (#51) * Update 2.5d layer code to get a similar accuracy on imagenet-1k dataset * update api for better usability (#58) update api for better usability Co-authored-by: 1SAA <c2h214748@gmail.com> Co-authored-by: ver217 <lhx0217@gmail.com> Co-authored-by: puck_WCR <46049915+WANG-CR@users.noreply.github.com> Co-authored-by: binmakeswell <binmakeswell@gmail.com> Co-authored-by: アマデウス <kurisusnowdeng@users.noreply.github.com> Co-authored-by: BoxiangW <45734921+BoxiangW@users.noreply.github.com>
2021-12-09 07:08:29 +00:00
return output
else:
return output, self.bias
@LAYERS.register_module
class Embedding1D(ParallelLayer):
2022-03-25 05:02:39 +00:00
r"""Embedding for 1D parallelism.
Args:
num_embeddings (int): number of embeddings.
embedding_dim (int): dimension of embedding.
padding_idx (int, optional): If specified, the entries at padding_idx do not contribute to the gradient;
therefore, the embedding vector at padding_idx is not updated during training,
i.e. it remains as a fixed pad, defaults to None.
dtype (:class:`torch.dtype`, optional): The dtype of parameters, defaults to None.
weight_initializer (:class:`typing.Callable`, optional):
he initializer of weight, defaults to normal initializer.
The ``args`` and ``kwargs`` used in :class:`torch.nn.functional.embedding` should contain:
::
max_norm (float, optional): If given, each embedding vector with norm larger than max_norm is
renormalized to have norm max_norm. Note: this will modify weight in-place.
norm_type (float, optional): The p of the p-norm to compute for the max_norm option. Default 2.
scale_grad_by_freq (bool, optional): If given, this will scale gradients by the inverse
of frequency of the words in the mini-batch. Default False.
sparse (bool, optional): If True, gradient w.r.t. weight will be a sparse tensor. Default False.
More details about ``args`` and ``kwargs`` could be found in
`Embedding <https://pytorch.org/docs/stable/generated/torch.nn.functional.embedding.html#torch.nn.functional.embedding>`_.
More details about ``initializer`` please refer to
`init <https://github.com/hpcaitech/ColossalAI/blob/main/colossalai/nn/init.py>`_
"""
def __init__(self,
num_embeddings: int,
embedding_dim: int,
padding_idx: int = None,
dtype: torch.dtype = None,
weight_initializer: Callable = init.normal_(),
*args,
**kwargs):
super().__init__()
self.num_embeddings = num_embeddings
self.embed_dim = embedding_dim
embed_dim_per_partition = divide(embedding_dim, gpc.tensor_parallel_size)
self.padding_idx = padding_idx
self.embed_args = args
self.embed_kwargs = kwargs
self.weight = Parameter(
torch.empty((num_embeddings, embed_dim_per_partition), device=get_current_device(), dtype=dtype))
self.reset_parameters(weight_initializer)
self._set_tensor_parallel_attributes()
set_parallel_input(False)
def _set_tensor_parallel_attributes(self):
set_tensor_parallel_attribute_by_partition(self.weight, gpc.tensor_parallel_size)
def reset_parameters(self, weight_initializer) -> None:
with seed(ParallelMode.TENSOR):
fan_in, fan_out = self.num_embeddings, self.embed_dim
weight_initializer(self.weight, fan_in=fan_in, fan_out=fan_out)
self._fill_padding_idx_with_zero()
def _fill_padding_idx_with_zero(self) -> None:
if self.padding_idx is not None:
with torch.no_grad():
self.weight[self.padding_idx].fill_(0)
def _load_from_global_state_dict(self, state_dict, prefix, *args):
local_state = OrderedDict()
weight_key = prefix + 'weight'
if gpc.get_local_rank(ParallelMode.TENSOR) == 0:
# weight
weight = state_dict.pop(weight_key, None)
if weight is not None:
local_state[weight_key] = weight
local_state = partition_tensor_parallel_state_dict(local_state,
ParallelMode.PARALLEL_1D,
dims={weight_key: -1},
partition_states={weight_key: True})
super()._load_from_global_state_dict(local_state, prefix, *args)
def _save_to_global_state_dict(self, destination, prefix, keep_vars):
weight_key = prefix + 'weight'
local_state = OrderedDict({weight_key: self.weight})
local_state = gather_tensor_parallel_state_dict(local_state,
ParallelMode.PARALLEL_1D,
dims={weight_key: -1},
partition_states={weight_key: True},
keep_vars=keep_vars)
destination.update(local_state)
def forward(self, input_: Tensor) -> Tensor:
output_parallel = F.embedding(input_, self.weight, self.padding_idx, *self.embed_args, **self.embed_kwargs)
output = gather_forward_split_backward(output_parallel, ParallelMode.PARALLEL_1D, dim=-1)
return output
@LAYERS.register_module
class VocabParallelEmbedding1D(ParallelLayer):
2022-03-25 05:02:39 +00:00
r"""Embedding parallelized in the vocabulary dimension.
Args:
num_embeddings (int): number of embeddings.
embedding_dim (int): dimension of embedding.
padding_idx (int, optional): If specified, the entries at padding_idx do not contribute to the gradient;
therefore, the embedding vector at padding_idx is not updated during training,
i.e. it remains as a fixed pad, defaults to None.
dtype (:class:`torch.dtype`, optional): The dtype of parameters, defaults to None.
weight_initializer (:class:`typing.Callable`, optional):
he initializer of weight, defaults to normal initializer.
The ``args`` and ``kwargs`` used in :class:``torch.nn.functional.embedding`` should contain:
::
max_norm (float, optional): If given, each embedding vector with norm larger than max_norm is
renormalized to have norm max_norm. Note: this will modify weight in-place.
norm_type (float, optional): The p of the p-norm to compute for the max_norm option. Default 2.
scale_grad_by_freq (bool, optional): If given, this will scale gradients by the inverse
of frequency of the words in the mini-batch. Default False.
sparse (bool, optional): If True, gradient w.r.t. weight will be a sparse tensor. Default False.
More details about ``args`` and ``kwargs`` could be found in
`Embedding <https://pytorch.org/docs/stable/generated/torch.nn.functional.embedding.html#torch.nn.functional.embedding>`_.
More details about initializer please refer to
`init <https://github.com/hpcaitech/ColossalAI/blob/main/colossalai/nn/init.py>`_.
"""
def __init__(self,
num_embeddings: int,
embedding_dim: int,
padding_idx: int = None,
dtype: torch.dtype = None,
weight_initializer: Callable = init.normal_(),
*args,
**kwargs):
super().__init__()
self.num_embeddings = num_embeddings
self.embed_dim = embedding_dim
self.padding_idx = padding_idx
self.embed_args = args
self.embed_kwargs = kwargs
tensor_parallel_size = gpc.get_world_size(ParallelMode.PARALLEL_1D)
tensor_parallel_rank = gpc.get_local_rank(ParallelMode.PARALLEL_1D)
self.num_embeddings_per_partition = divide(num_embeddings, tensor_parallel_size)
self.vocab_start_index = tensor_parallel_rank * self.num_embeddings_per_partition
self.vocab_end_index = self.vocab_start_index + self.num_embeddings_per_partition
self.weight = Parameter(
torch.empty((self.num_embeddings_per_partition, self.embed_dim), device=get_current_device(), dtype=dtype))
self.reset_parameters(weight_initializer)
self._set_tensor_parallel_attributes()
set_parallel_input(False)
env.vocab_parallel = True
def _set_tensor_parallel_attributes(self):
set_tensor_parallel_attribute_by_partition(self.weight, gpc.tensor_parallel_size)
def reset_parameters(self, weight_initializer) -> None:
with seed(ParallelMode.TENSOR):
fan_in, fan_out = self.num_embeddings, self.embed_dim
weight_initializer(self.weight, fan_in=fan_in, fan_out=fan_out)
self._fill_padding_idx_with_zero()
def _fill_padding_idx_with_zero(self) -> None:
if self.padding_idx is not None and \
self.padding_idx >= self.vocab_start_index and self.padding_idx < self.vocab_end_index:
with torch.no_grad():
self.weight[self.padding_idx - self.vocab_start_index].fill_(0)
def _load_from_global_state_dict(self, state_dict, prefix, *args):
local_state = OrderedDict()
weight_key = prefix + 'weight'
if gpc.get_local_rank(ParallelMode.TENSOR) == 0:
# weight
weight = state_dict.pop(weight_key, None)
if weight is not None:
local_state[weight_key] = weight
local_state = partition_tensor_parallel_state_dict(local_state,
ParallelMode.PARALLEL_1D,
dims={weight_key: 0},
partition_states={weight_key: True})
super()._load_from_global_state_dict(local_state, prefix, *args)
def _save_to_global_state_dict(self, destination, prefix, keep_vars):
weight_key = prefix + 'weight'
local_state = OrderedDict({weight_key: self.weight})
local_state = gather_tensor_parallel_state_dict(local_state,
ParallelMode.PARALLEL_1D,
dims={weight_key: 0},
partition_states={weight_key: True},
keep_vars=keep_vars)
destination.update(local_state)
def forward(self, input_: Tensor) -> Tensor:
# Build the mask.
input_mask = (input_ < self.vocab_start_index) | (input_ >= self.vocab_end_index)
# Mask the input.
masked_input = input_.clone() - self.vocab_start_index
masked_input[input_mask] = 0
output_parallel = F.embedding(masked_input, self.weight, self.padding_idx, *self.embed_args,
**self.embed_kwargs)
# Mask the output embedding.
output_parallel[input_mask, :] = 0.
# Reduce across all the model parallel GPUs.
output = reduce_input(output_parallel, ParallelMode.PARALLEL_1D)
return output
@LAYERS.register_module
class Dropout1D(ParallelLayer):
2022-03-25 05:02:39 +00:00
"""Dropout layer of 1D parallelism.
2022-03-25 05:02:39 +00:00
Args:
p (float, optional): probability of an element to be zeroed, defaults 0.5.
inplace (bool, optional): whether to do dropout in-place, default to be False.
"""
def __init__(self, p: float = 0.5, inplace: bool = False):
super().__init__()
self.parallel_input = get_parallel_input()
self.p = p
self.inplace = inplace
def forward(self, input_: Tensor) -> Tensor:
if self.parallel_input:
with seed(ParallelMode.TENSOR):
output = F.dropout(input_, self.p, self.training, self.inplace)
else:
output = F.dropout(input_, self.p, self.training, self.inplace)
return output
@LAYERS.register_module
class PatchEmbedding1D(ColossalaiModule):
"""
2D Image to Patch Embedding
:param img_size: image size
:type img_size: int
:param patch_size: patch size
:type patch_size: int
:param in_chans: number of channels of input image
:type in_chans: int
:param embed_size: size of embedding
:type embed_size: int
:param dtype: The dtype of parameters, defaults to None
:type dtype: torch.dtype, optional
:param flatten: whether to flatten output tensor, defaults to True
:type flatten: bool, optional
:param weight_initializer: The intializer of weight, defaults to kaiming uniform initializer
:type weight_initializer: typing.Callable, optional
:param bias_initializer: The intializer of bias, defaults to xavier uniform initializer
:type bias_initializer: typing.Callable, optional
:param position_embed_initializer: The intializer of position embedding, defaults to zero
:type position_embed_initializer: typing.Callable, optional
"""
def __init__(self,
img_size: int,
patch_size: int,
in_chans: int,
embed_size: int,
dtype: torch.dtype = None,
flatten: bool = True,
weight_initializer: Callable = init.kaiming_uniform_(a=math.sqrt(5)),
bias_initializer: Callable = init.xavier_uniform_(a=1, scale=1),
position_embed_initializer: Callable = init.zeros_()):
embed = VanillaPatchEmbedding(img_size,
patch_size,
in_chans,
embed_size,
dtype=dtype,
flatten=flatten,
weight_initializer=weight_initializer,
bias_initializer=bias_initializer,
position_embed_initializer=position_embed_initializer)
super().__init__(embed)
def _load_from_state_dict(self, state_dict, prefix, *args):
local_state = OrderedDict()
param_keys = [prefix + 'weight', prefix + 'bias', prefix + 'cls_token', prefix + 'pos_embed']
if gpc.get_local_rank(ParallelMode.TENSOR) == 0:
for key in param_keys:
param = state_dict.pop(key, None)
if param is not None:
local_state[key] = param
local_state = broadcast_state_dict(local_state, ParallelMode.PARALLEL_1D)
super()._load_from_state_dict(local_state, prefix, *args)
def _save_to_state_dict(self, destination, prefix, keep_vars):
if gpc.get_local_rank(ParallelMode.TENSOR) == 0:
super()._save_to_state_dict(destination, prefix, keep_vars)