2023-06-20 03:45:16 +00:00
|
|
|
#!/usr/bin/env python
|
|
|
|
# -*- encoding: utf-8 -*-
|
|
|
|
|
|
|
|
from typing import Callable, List, Union
|
|
|
|
|
|
|
|
import torch
|
|
|
|
import torch.distributed as dist
|
|
|
|
import torch.nn as nn
|
|
|
|
import torch.nn.functional as F
|
|
|
|
from torch import Tensor
|
|
|
|
from torch.distributed import ProcessGroup
|
|
|
|
from torch.nn.parameter import Parameter
|
|
|
|
|
|
|
|
from colossalai.nn import init as init
|
|
|
|
from colossalai.nn.layer.utils import divide
|
2023-06-22 03:42:11 +00:00
|
|
|
from colossalai.tensor.d_tensor.api import shard_colwise, shard_rowwise, sharded_tensor_to_param
|
2023-06-20 03:45:16 +00:00
|
|
|
|
2023-06-23 08:07:09 +00:00
|
|
|
from ._operation import gather_forward_split_backward, reduce_forward
|
2023-06-21 06:30:06 +00:00
|
|
|
from .parallel_module import ParallelModule
|
2023-06-20 03:45:16 +00:00
|
|
|
from .utils import create_randomizer_with_offset
|
|
|
|
|
2023-06-21 06:30:06 +00:00
|
|
|
__all__ = ['Embedding1D', 'VocabParallelEmbedding1D']
|
2023-06-20 03:45:16 +00:00
|
|
|
|
|
|
|
|
2023-06-21 06:30:06 +00:00
|
|
|
class Embedding1D(ParallelModule):
|
|
|
|
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,
|
|
|
|
device: torch.device = None,
|
|
|
|
process_group: ProcessGroup = None,
|
|
|
|
gather_output: bool = True,
|
|
|
|
weight_initializer: Callable = init.normal_(),
|
|
|
|
*args,
|
|
|
|
**kwargs):
|
|
|
|
super().__init__()
|
|
|
|
|
|
|
|
self.num_embeddings = num_embeddings
|
|
|
|
self.embedding_dim = embedding_dim
|
|
|
|
self.process_group = process_group
|
|
|
|
|
|
|
|
self.padding_idx = padding_idx
|
|
|
|
self.embed_args = args
|
|
|
|
self.embed_kwargs = kwargs
|
|
|
|
self.gather_output = gather_output
|
|
|
|
|
2023-06-22 03:42:11 +00:00
|
|
|
# Parameters.
|
|
|
|
factory_kwargs = {'device': device, 'dtype': dtype}
|
|
|
|
weight = torch.empty((num_embeddings, self.embedding_dim), **factory_kwargs)
|
|
|
|
sharded_weight = shard_colwise(weight, process_group)
|
|
|
|
self.weight = sharded_tensor_to_param(sharded_weight)
|
2023-06-21 06:30:06 +00:00
|
|
|
|
|
|
|
# offset the seed with randomizer index and rank
|
|
|
|
seed = torch.random.initial_seed()
|
|
|
|
self.randomizer = create_randomizer_with_offset(seed, process_group=self.process_group)
|
|
|
|
|
|
|
|
with self.randomizer.fork_rng(enable_cpu=True):
|
|
|
|
self.reset_parameters(weight_initializer)
|
|
|
|
|
|
|
|
@staticmethod
|
|
|
|
def from_native_module(module: nn.Embedding,
|
|
|
|
process_group: Union[ProcessGroup, List[ProcessGroup]] = None,
|
|
|
|
*args,
|
|
|
|
**kwargs) -> "Embedding1D":
|
|
|
|
r"""
|
|
|
|
Build a 1D parallelized Embedding from a native nn.Embedding module.
|
|
|
|
"""
|
|
|
|
# get the attributes
|
|
|
|
num_embedding = module.num_embeddings
|
|
|
|
embedding_dim = module.embedding_dim
|
|
|
|
padding_idx = module.padding_idx
|
|
|
|
max_norm = module.max_norm
|
|
|
|
norm_type = module.norm_type
|
|
|
|
scale_grad_by_freq = module.scale_grad_by_freq
|
|
|
|
sparse = module.sparse
|
|
|
|
dtype = module.weight.dtype
|
|
|
|
device = module.weight.device
|
|
|
|
|
|
|
|
# sparse is not support yet
|
|
|
|
if sparse:
|
|
|
|
raise NotImplementedError("The Embedding1D module does not support sparse embedding yet.")
|
|
|
|
|
|
|
|
embedding = Embedding1D(num_embeddings=num_embedding,
|
|
|
|
embedding_dim=embedding_dim,
|
|
|
|
padding_idx=padding_idx,
|
|
|
|
process_group=process_group,
|
|
|
|
dtype=dtype,
|
|
|
|
device=device,
|
|
|
|
max_norm=max_norm,
|
|
|
|
norm_type=norm_type,
|
|
|
|
scale_grad_by_freq=scale_grad_by_freq,
|
|
|
|
sparse=sparse,
|
|
|
|
*args,
|
|
|
|
**kwargs)
|
|
|
|
|
|
|
|
# copy the weight
|
|
|
|
with torch.no_grad():
|
|
|
|
sharded_weight = shard_colwise(module.weight.data, process_group)
|
|
|
|
embedding.weight.copy_(sharded_weight)
|
|
|
|
|
|
|
|
return embedding
|
|
|
|
|
|
|
|
def reset_parameters(self, weight_initializer) -> None:
|
|
|
|
fan_in, fan_out = self.num_embeddings, self.embedding_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 forward(self, input_: Tensor) -> Tensor:
|
|
|
|
output_parallel = F.embedding(input_, self.weight, self.padding_idx, *self.embed_args, **self.embed_kwargs)
|
|
|
|
|
|
|
|
if self.gather_output:
|
|
|
|
output = gather_forward_split_backward(output_parallel, dim=-1, process_group=self.process_group)
|
|
|
|
return output
|
|
|
|
else:
|
|
|
|
return output_parallel
|
|
|
|
|
|
|
|
|
|
|
|
class VocabParallelEmbedding1D(ParallelModule):
|
2023-06-20 03:45:16 +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,
|
|
|
|
device: torch.device = None,
|
|
|
|
process_group: ProcessGroup = None,
|
|
|
|
weight_initializer: Callable = init.normal_(),
|
|
|
|
*args,
|
|
|
|
**kwargs):
|
|
|
|
super().__init__()
|
|
|
|
self.num_embeddings = num_embeddings
|
2023-06-22 03:42:11 +00:00
|
|
|
self.embedding_dim = embedding_dim
|
2023-06-20 03:45:16 +00:00
|
|
|
self.padding_idx = padding_idx
|
|
|
|
self.embed_args = args
|
|
|
|
self.embed_kwargs = kwargs
|
|
|
|
self.process_group = process_group
|
|
|
|
|
|
|
|
tensor_parallel_size = dist.get_world_size(group=process_group)
|
|
|
|
tensor_parallel_rank = dist.get_rank(group=process_group)
|
|
|
|
|
|
|
|
self.num_embeddings_per_partition = divide(num_embeddings, tensor_parallel_size)
|
|
|
|
self.num_embeddings = self.num_embeddings_per_partition
|
|
|
|
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
|
|
|
|
|
2023-06-22 03:42:11 +00:00
|
|
|
# parameter
|
|
|
|
factory_kwargs = {'device': device, 'dtype': dtype}
|
|
|
|
weight = torch.empty((num_embeddings, self.embedding_dim), **factory_kwargs)
|
|
|
|
sharded_weight = shard_rowwise(weight, process_group)
|
|
|
|
self.weight = sharded_tensor_to_param(sharded_weight)
|
2023-06-20 03:45:16 +00:00
|
|
|
|
|
|
|
# offset the seed with randomizer index and rank
|
|
|
|
seed = torch.random.initial_seed()
|
|
|
|
self.randomizer = create_randomizer_with_offset(seed, process_group=self.process_group)
|
2023-06-21 06:30:06 +00:00
|
|
|
self.reset_parameters(weight_initializer)
|
2023-06-20 03:45:16 +00:00
|
|
|
|
|
|
|
@staticmethod
|
|
|
|
def from_native_module(module: nn.Embedding, process_group: Union[ProcessGroup, List[ProcessGroup]], *args,
|
|
|
|
**kwargs) -> ParallelModule:
|
|
|
|
r"""
|
|
|
|
Convert a native pytorch embedding module to a parallel module.
|
|
|
|
"""
|
|
|
|
# get the origin attributes
|
|
|
|
num_embeddings = module.num_embeddings
|
|
|
|
embedding_dim = module.embedding_dim
|
|
|
|
padding_idx = module.padding_idx
|
|
|
|
device = module.weight.device
|
|
|
|
|
|
|
|
# ensure only one process group is used
|
|
|
|
if isinstance(process_group, (list, tuple)):
|
|
|
|
assert len(process_group) == 1, \
|
|
|
|
f'Expected only one process group, got {len(process_group)}.'
|
|
|
|
process_group = process_group[0]
|
|
|
|
|
|
|
|
# create the parallel module
|
|
|
|
vocab_embedding_1d = VocabParallelEmbedding1D(num_embeddings=num_embeddings,
|
|
|
|
embedding_dim=embedding_dim,
|
|
|
|
padding_idx=padding_idx,
|
|
|
|
device=device,
|
|
|
|
process_group=process_group,
|
|
|
|
*args,
|
|
|
|
**kwargs)
|
|
|
|
with torch.no_grad():
|
|
|
|
# shard and slice the weight along the vocabulary(num_embeddings) dimension
|
|
|
|
# the shape of the weight is (num_embeddings, embedding_dim)
|
|
|
|
shard_weight = shard_rowwise(module.weight.data, process_group)
|
|
|
|
vocab_embedding_1d.weight.data.copy_(shard_weight)
|
|
|
|
|
|
|
|
return vocab_embedding_1d
|
|
|
|
|
|
|
|
def reset_parameters(self, weight_initializer) -> None:
|
2023-06-21 06:30:06 +00:00
|
|
|
with self.randomizer.fork_rng(enable_cpu=True):
|
2023-06-22 03:42:11 +00:00
|
|
|
fan_in, fan_out = self.num_embeddings, self.embedding_dim
|
2023-06-20 03:45:16 +00:00
|
|
|
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 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.
|
2023-06-23 08:07:09 +00:00
|
|
|
output = reduce_forward(output_parallel, self.process_group)
|
2023-06-20 03:45:16 +00:00
|
|
|
return output
|