/usr/local/lib/python3.6/site-packages/transformers
Edit: /usr/local/lib/python3.6/site-packages/transformers/activations.py (6612B)
# Copyright 2020 The HuggingFace Team. All rights reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
import math
import torch
from packaging import version
from torch import Tensor, nn
from .utils import logging
logger = logging.get_logger(__name__)
class NewGELUActivation(nn.Module):
"""
Implementation of the GELU activation function currently in Google BERT repo (identical to OpenAI GPT). Also see
the Gaussian Error Linear Units paper: https://arxiv.org/abs/1606.08415
"""
def forward(self, input: Tensor) -> Tensor:
return 0.5 * input * (1.0 + torch.tanh(math.sqrt(2.0 / math.pi) * (input + 0.044715 * torch.pow(input, 3.0))))
class GELUActivation(nn.Module):
"""
Original Implementation of the GELU activation function in Google BERT repo when initially created. For
information: OpenAI GPT's GELU is slightly different (and gives slightly different results): 0.5 * x * (1 +
torch.tanh(math.sqrt(2 / math.pi) * (x + 0.044715 * torch.pow(x, 3)))) This is now written in C in nn.functional
Also see the Gaussian Error Linear Units paper: https://arxiv.org/abs/1606.08415
"""
def __init__(self, use_gelu_python: bool = False):
super().__init__()
if version.parse(torch.__version__) < version.parse("1.4") or use_gelu_python:
self.act = self._gelu_python
else:
self.act = nn.functional.gelu
def _gelu_python(self, input: Tensor) -> Tensor:
return input * 0.5 * (1.0 + torch.erf(input / math.sqrt(2.0)))
def forward(self, input: Tensor) -> Tensor:
return self.act(input)
class FastGELUActivation(nn.Module):
"""
Applies GELU approximation that is slower than QuickGELU but more accurate. See: https://github.com/hendrycks/GELUs
"""
def forward(self, input: Tensor) -> Tensor:
return 0.5 * input * (1.0 + torch.tanh(input * 0.7978845608 * (1.0 + 0.044715 * input * input)))
class QuickGELUActivation(nn.Module):
"""
Applies GELU approximation that is fast but somewhat inaccurate. See: https://github.com/hendrycks/GELUs
"""
def forward(self, input: Tensor) -> Tensor:
return input * torch.sigmoid(1.702 * input)
class ClippedGELUActivation(nn.Module):
"""
Clip the range of possible GeLU outputs between [min, max]. This is especially useful for quantization purpose, as
it allows mapping negatives values in the GeLU spectrum. For more information on this trick, please refer to
https://arxiv.org/abs/2004.09602.
Gaussian Error Linear Unit. Original Implementation of the gelu activation function in Google Bert repo when
initially created.
For information: OpenAI GPT's gelu is slightly different (and gives slightly different results): 0.5 * x * (1 +
torch.tanh(math.sqrt(2 / math.pi) * (x + 0.044715 * torch.pow(x, 3)))). See https://arxiv.org/abs/1606.08415
"""
def __init__(self, min: float, max: float):
if min > max:
raise ValueError(f"min should be < max (got min: {min}, max: {max})")
super().__init__()
self.min = min
self.max = max
def forward(self, x: Tensor) -> Tensor:
return torch.clip(gelu(x), self.min, self.max)
class SiLUActivation(nn.Module):
"""
See Gaussian Error Linear Units (Hendrycks et al., https://arxiv.org/abs/1606.08415) where the SiLU (Sigmoid Linear
Unit) was originally introduced and coined, and see Sigmoid-Weighted Linear Units for Neural Network Function
Approximation in Reinforcement Learning (Elfwing et al., https://arxiv.org/abs/1702.03118) and Swish: a Self-Gated
Activation Function (Ramachandran et al., https://arxiv.org/abs/1710.05941v1) where the SiLU was experimented with
later.
"""
def __init__(self):
super().__init__()
if version.parse(torch.__version__) < version.parse("1.7"):
self.act = self._silu_python
else:
self.act = nn.functional.silu
def _silu_python(self, input: Tensor) -> Tensor:
return input * torch.sigmoid(input)
def forward(self, input: Tensor) -> Tensor:
return self.act(input)
class MishActivation(nn.Module):
"""
See Mish: A Self-Regularized Non-Monotonic Activation Function (Misra., https://arxiv.org/abs/1908.08681). Also
visit the official repository for the paper: https://github.com/digantamisra98/Mish
"""
def __init__(self):
super().__init__()
if version.parse(torch.__version__) < version.parse("1.9"):
self.act = self._mish_python
else:
self.act = nn.functional.mish
def _mish_python(self, input: Tensor) -> Tensor:
return input * torch.tanh(nn.functional.softplus(input))
def forward(self, input: Tensor) -> Tensor:
return self.act(input)
class LinearActivation(nn.Module):
"""
Applies the linear activation function, i.e. forwarding input directly to output.
"""
def forward(self, input: Tensor) -> Tensor:
return input
ACT2FN = {
"relu": nn.ReLU(),
"silu": SiLUActivation(),
"swish": SiLUActivation(),
"gelu": GELUActivation(),
"tanh": nn.Tanh(),
"gelu_python": GELUActivation(use_gelu_python=True),
"gelu_new": NewGELUActivation(),
"gelu_fast": FastGELUActivation(),
"quick_gelu": QuickGELUActivation(),
"gelu_10": ClippedGELUActivation(-10, 10),
"mish": MishActivation(),
"linear": LinearActivation(),
"sigmoid": nn.Sigmoid(),
}
def get_activation(activation_string):
if activation_string in ACT2FN:
return ACT2FN[activation_string]
else:
raise KeyError(f"function {activation_string} not found in ACT2FN mapping {list(ACT2FN.keys())}")
# For backwards compatibility with: from activations import gelu_python
gelu_python = get_activation("gelu_python")
gelu_new = get_activation("gelu_new")
gelu = get_activation("gelu")
gelu_fast = get_activation("gelu_fast")
quick_gelu = get_activation("quick_gelu")
silu = get_activation("silu")
mish = get_activation("mish")
linear_act = get_activation("linear")