Deprecate xformers.components (fairinternal/xformers#1210)

__original_commit__ = fairinternal/xformers@01c08fb

xformers/components/__init__.py

@@ -4,6 +4,7 @@
 # LICENSE file in the root directory of this source tree.
 
 
+import warnings
 from dataclasses import fields
 from pathlib import Path
 from typing import Any, Dict, Union
@@ -24,6 +25,14 @@
 from .residual import Residual  # noqa
 from .residual import ResidualNormStyle  # noqa
 
+warnings.warn(
+    "xformers.components is deprecated and is not maintained anymore. "
+    "It might be removed in a future version of xFormers ",
+    FutureWarning,
+    stacklevel=2,
+)
+
+
 # automatically import any Python files in the directory
 import_all_modules(str(Path(__file__).parent), "xformers.components")
 

xformers/components/activations.py

@@ -10,6 +10,8 @@
 import torch
 from torch import nn
 
+from xformers._deprecation_warning import deprecated_function
+
 
 class Activation(str, Enum):
     SquaredReLU = "squared_relu"
@@ -24,6 +26,7 @@ class Activation(str, Enum):
 class SquaredReLU(nn.Module):
     def __init__(self) -> None:
         super().__init__()
+        deprecated_function(self)
 
     def forward(self, x: torch.Tensor) -> torch.Tensor:
         x_ = torch.nn.functional.relu(x)
@@ -33,6 +36,7 @@ def forward(self, x: torch.Tensor) -> torch.Tensor:
 class StarReLU(nn.Module):
     def __init__(self) -> None:
         super().__init__()
+        deprecated_function(self)
 
     def forward(self, x: torch.Tensor) -> torch.Tensor:
         x_ = torch.nn.functional.relu(x)
@@ -43,6 +47,7 @@ class SmeLU(nn.Module):
     def __init__(self, beta: float = 2.0) -> None:
         super().__init__()
         self.beta = beta
+        deprecated_function(self)
 
     def forward(self, x: torch.Tensor) -> torch.Tensor:
         relu = torch.where(

xformers/components/attention/base.py

@@ -11,6 +11,7 @@
 import torch
 import torch.nn as nn
 
+from xformers._deprecation_warning import deprecated_function
 from xformers.components.attention import AttentionMask
 
 
@@ -36,6 +37,7 @@ class Attention(nn.Module, metaclass=ABCMeta):
     @abstractmethod
     def __init__(self, dropout: Optional[float] = None, *args, **kwargs):
         super().__init__()
+        deprecated_function(self)
 
         # Requires the inputs to be projected
         self.requires_input_projection = True

xformers/components/feedforward/base.py

@@ -10,6 +10,7 @@
 
 import torch.nn as nn
 
+from xformers._deprecation_warning import deprecated_function
 from xformers.components import Activation
 
 Self = TypeVar("Self", bound="Feedforward")
@@ -35,6 +36,7 @@ def __init__(
         **kwargs,
     ):
         super().__init__()
+        deprecated_function(self)
 
         # This feedforward requires a CUDA accelerator
         self.requires_cuda = False

xformers/components/input_projection.py

@@ -14,6 +14,8 @@
 import torch
 from torch import nn
 
+from xformers._deprecation_warning import deprecated_function
+
 logger = logging.getLogger("xformers")
 
 
@@ -38,6 +40,7 @@ def __init__(
     ):
 
         super().__init__()
+        deprecated_function(self)
 
         self.out_features = query_proj_params.out_features
 

xformers/components/multi_head_dispatch.py

@@ -12,6 +12,7 @@
 import torch.nn as nn
 from torch.nn.init import constant_
 
+from xformers._deprecation_warning import deprecated_function
 from xformers.components.attention import Attention
 from xformers.components.input_projection import InputProjection, InputProjectionConfig
 from xformers.components.positional_embedding import RotaryEmbedding
@@ -90,6 +91,7 @@ def __init__(
         **kwargs,
     ):
         super().__init__()
+        deprecated_function(self)
 
         if isinstance(bias, bool):
             logger.warning(

xformers/components/patch_embedding.py

@@ -9,6 +9,8 @@
 
 import torch
 
+from xformers._deprecation_warning import deprecated_function
+
 
 class PoolType(str, Enum):
     Conv2D = "CONV_2D"
@@ -39,6 +41,7 @@ class PatchEmbeddingConfig:
 class ConditionalReshape(torch.nn.Module):
     def __init__(self) -> None:
         super().__init__()
+        deprecated_function(self)
 
     def forward(self, x):
         if x.ndim == 3:
@@ -54,6 +57,7 @@ def forward(self, x):
 class PatchToSequence(torch.nn.Module):
     def __init__(self) -> None:
         super().__init__()
+        deprecated_function(self)
 
     def forward(self, x):
         return x.flatten(2, 3).transpose(1, 2).contiguous()  # B HW C

xformers/components/positional_embedding/base.py

@@ -10,6 +10,8 @@
 
 import torch.nn as nn
 
+from xformers._deprecation_warning import deprecated_function
+
 Self = TypeVar("Self", bound="PositionEmbedding")
 
 
@@ -24,6 +26,7 @@ class PositionEmbedding(nn.Module, metaclass=ABCMeta):
     @abstractmethod
     def __init__(self, *args, **kwargs) -> None:
         super().__init__()
+        deprecated_function(self)
 
     @classmethod
     def from_config(cls: Type[Self], config: PositionEmbeddingConfig) -> Self:

xformers/components/residual.py

@@ -11,6 +11,8 @@
 import torch
 import torch.nn as nn
 
+from xformers._deprecation_warning import deprecated_function
+
 
 class ResidualNormStyle(str, Enum):
     """Support different residual path and norm styles.
@@ -34,6 +36,7 @@ def get_normalization_layer(normalization_type: NormalizationType):
     class Skip(nn.Module):
         def __init__(self, *_, **__) -> None:
             super().__init__()
+            deprecated_function(self)
 
         def forward(self, x: torch.Tensor, **_):
             return x
@@ -64,6 +67,7 @@ class Residual(nn.Module, RequiresWrappedInputs):
 
     def __init__(self, layer: nn.Module, scale: Optional[float] = None):
         super().__init__()
+        deprecated_function(self)
         self.layer = layer
         self.scale = scale
 
@@ -97,6 +101,7 @@ def __init__(
     ):
 
         super().__init__()
+        deprecated_function(self)
         self.norm = get_normalization_layer(normalization)(d_norm)
 
         self.sublayer = sublayer
@@ -132,6 +137,7 @@ def __init__(
         use_triton: bool = True,
     ):
         super().__init__()
+        deprecated_function(self)
         self.norm = get_normalization_layer(normalization)(d_norm)
 
         self.sublayer = sublayer

xformers/components/reversible.py

@@ -11,6 +11,7 @@
 from torch.autograd.function import Function
 from torch.utils.checkpoint import get_device_states, set_device_states
 
+from xformers._deprecation_warning import deprecated_function
 from xformers.components import RequiresWrappedInputs
 
 # CREDITS: Code adapted from
@@ -23,6 +24,7 @@
 class Deterministic(nn.Module):
     def __init__(self, net: nn.Module):
         super().__init__()
+        deprecated_function(self)
         self.net = net
         self.cpu_state: torch.Tensor = torch.get_rng_state()
         self.cuda_in_fwd: bool = False
@@ -146,6 +148,7 @@ def backward(
 class ReversibleSequence(nn.Module):
     def __init__(self, blocks: nn.ModuleList):
         super().__init__()
+        deprecated_function(self)
 
         # pyre-fixme[23]: Unable to unpack `torch.nn.Module` into 2 values.
         self.blocks = nn.ModuleList([ReversibleBlock(f, g) for f, g in blocks])

xformers/components/simplicial_embedding.py

@@ -8,6 +8,8 @@
 
 import torch
 
+from xformers._deprecation_warning import deprecated_function
+
 Self = TypeVar("Self", bound="SimplicialEmbedding")
 
 
@@ -32,6 +34,7 @@ class SimplicialEmbedding(torch.nn.Module):
 
     def __init__(self, L: int, temperature: Optional[float] = None) -> None:
         super().__init__()
+        deprecated_function(self)
         self.L = L
         self.temperature = temperature