Build uploaded using `kernels`.

build/torch210-cxx11-cpu-x86_64-linux/__init__.py

@@ -1,14 +0,0 @@
-from . import layers
-
-from ._ops import ops
-
-
-def apply_rms_norm(input, weight, eps):
-    return ops.apply_rms_norm(
-            input,
-            weight,
-            eps,
-    )
-
-__all__ = ["layers", "apply_rms_norm"]
-

build/torch210-cxx11-cpu-x86_64-linux/_ops.py

@@ -1,9 +0,0 @@
-import torch
-from . import _rmsnorm_fb26d8c
-ops = torch.ops._rmsnorm_fb26d8c
-
-def add_op_namespace_prefix(op_name: str):
-    """
-    Prefix op by namespace.
-    """
-    return f"_rmsnorm_fb26d8c::{op_name}"

build/torch210-cxx11-cpu-x86_64-linux/_rmsnorm_fb26d8c.abi3.so

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:f8a1c744d46b5b0b6455825653741008b06242630ae9946f0205ac2c055dbc7e
-size 326352

build/torch210-cxx11-cpu-x86_64-linux/layers.py

@@ -1,36 +0,0 @@
-import torch
-from ._ops import ops
-
-class RMSNorm(torch.nn.Module):
-    """
-    RMSNorm module that uses the optimized LigerRMSNormFunction.
-    
-    Args:
-        hidden_size (int): The size of the hidden dimension.
-        eps (float, optional): The epsilon value for numerical stability. Defaults to 1e-6.
-        offset (float, optional): Offset value to shift the weight tensor. Defaults to 0.0.
-        casting_mode (str, optional): The casting mode to use. Defaults to "llama".
-        in_place (bool, optional): Whether to modify dY in-place to store dX during backward. Defaults to True.
-    """
-    
-
-    weight: torch.Tensor
-    variance_epsilon: float
-    
-    def forward(self, hidden_states):
-        """
-        Apply RMS normalization to the input tensor.
-        
-        Args:
-            hidden_states (torch.Tensor): Input tensor of shape (B, T, H) or (BxT, H)
-            
-        Returns:
-            torch.Tensor: Normalized tensor of the same shape as input
-        """
-        return ops.apply_rms_norm(
-            hidden_states,
-            self.weight, 
-            self.variance_epsilon,
-        )
-    
-__all__ = ["RMSNorm"]

build/torch210-cxx11-cpu-x86_64-linux/metadata.json

@@ -1 +0,0 @@
-{"python-depends":[]}

build/torch210-cxx11-cpu-x86_64-linux/rmsnorm/__init__.py

@@ -1,26 +0,0 @@
-import ctypes
-import sys
-
-import importlib
-from pathlib import Path
-from types import ModuleType
-
-def _import_from_path(file_path: Path) -> ModuleType:
-    # We cannot use the module name as-is, after adding it to `sys.modules`,
-    # it would also be used for other imports. So, we make a module name that
-    # depends on the path for it to be unique using the hex-encoded hash of
-    # the path.
-    path_hash = "{:x}".format(ctypes.c_size_t(hash(file_path.absolute())).value)
-    module_name = path_hash
-    spec = importlib.util.spec_from_file_location(module_name, file_path)
-    if spec is None:
-        raise ImportError(f"Cannot load spec for {module_name} from {file_path}")
-    module = importlib.util.module_from_spec(spec)
-    if module is None:
-        raise ImportError(f"Cannot load module {module_name} from spec")
-    sys.modules[module_name] = module
-    spec.loader.exec_module(module)  # type: ignore
-    return module
-
-
-globals().update(vars(_import_from_path(Path(__file__).parent.parent / "__init__.py")))

build/torch210-cxx11-xpu20253-x86_64-linux/__init__.py

@@ -1,14 +0,0 @@
-from . import layers
-
-from ._ops import ops
-
-
-def apply_rms_norm(input, weight, eps):
-    return ops.apply_rms_norm(
-            input,
-            weight,
-            eps,
-    )
-
-__all__ = ["layers", "apply_rms_norm"]
-

build/torch210-cxx11-xpu20253-x86_64-linux/_ops.py

@@ -1,9 +0,0 @@
-import torch
-from . import _rmsnorm_fb26d8c
-ops = torch.ops._rmsnorm_fb26d8c
-
-def add_op_namespace_prefix(op_name: str):
-    """
-    Prefix op by namespace.
-    """
-    return f"_rmsnorm_fb26d8c::{op_name}"

build/torch210-cxx11-xpu20253-x86_64-linux/_rmsnorm_fb26d8c.abi3.so

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:4be94737423cc4d02f4be83f38144614d71ccd8672d96699f0b10136dd541847
-size 104941392

build/torch210-cxx11-xpu20253-x86_64-linux/layers.py

@@ -1,36 +0,0 @@
-import torch
-from ._ops import ops
-
-class RMSNorm(torch.nn.Module):
-    """
-    RMSNorm module that uses the optimized LigerRMSNormFunction.
-    
-    Args:
-        hidden_size (int): The size of the hidden dimension.
-        eps (float, optional): The epsilon value for numerical stability. Defaults to 1e-6.
-        offset (float, optional): Offset value to shift the weight tensor. Defaults to 0.0.
-        casting_mode (str, optional): The casting mode to use. Defaults to "llama".
-        in_place (bool, optional): Whether to modify dY in-place to store dX during backward. Defaults to True.
-    """
-    
-
-    weight: torch.Tensor
-    variance_epsilon: float
-    
-    def forward(self, hidden_states):
-        """
-        Apply RMS normalization to the input tensor.
-        
-        Args:
-            hidden_states (torch.Tensor): Input tensor of shape (B, T, H) or (BxT, H)
-            
-        Returns:
-            torch.Tensor: Normalized tensor of the same shape as input
-        """
-        return ops.apply_rms_norm(
-            hidden_states,
-            self.weight, 
-            self.variance_epsilon,
-        )
-    
-__all__ = ["RMSNorm"]

build/torch210-cxx11-xpu20253-x86_64-linux/metadata.json

@@ -1 +0,0 @@
-{"python-depends":[]}

build/torch210-cxx11-xpu20253-x86_64-linux/rmsnorm/__init__.py

@@ -1,26 +0,0 @@
-import ctypes
-import sys
-
-import importlib
-from pathlib import Path
-from types import ModuleType
-
-def _import_from_path(file_path: Path) -> ModuleType:
-    # We cannot use the module name as-is, after adding it to `sys.modules`,
-    # it would also be used for other imports. So, we make a module name that
-    # depends on the path for it to be unique using the hex-encoded hash of
-    # the path.
-    path_hash = "{:x}".format(ctypes.c_size_t(hash(file_path.absolute())).value)
-    module_name = path_hash
-    spec = importlib.util.spec_from_file_location(module_name, file_path)
-    if spec is None:
-        raise ImportError(f"Cannot load spec for {module_name} from {file_path}")
-    module = importlib.util.module_from_spec(spec)
-    if module is None:
-        raise ImportError(f"Cannot load module {module_name} from spec")
-    sys.modules[module_name] = module
-    spec.loader.exec_module(module)  # type: ignore
-    return module
-
-
-globals().update(vars(_import_from_path(Path(__file__).parent.parent / "__init__.py")))

build/torch28-cxx11-cpu-x86_64-linux/__init__.py

@@ -1,14 +0,0 @@
-from . import layers
-
-from ._ops import ops
-
-
-def apply_rms_norm(input, weight, eps):
-    return ops.apply_rms_norm(
-            input,
-            weight,
-            eps,
-    )
-
-__all__ = ["layers", "apply_rms_norm"]
-

build/torch28-cxx11-cpu-x86_64-linux/_ops.py

@@ -1,9 +0,0 @@
-import torch
-from . import _rmsnorm_fb26d8c
-ops = torch.ops._rmsnorm_fb26d8c
-
-def add_op_namespace_prefix(op_name: str):
-    """
-    Prefix op by namespace.
-    """
-    return f"_rmsnorm_fb26d8c::{op_name}"

build/torch28-cxx11-cpu-x86_64-linux/_rmsnorm_fb26d8c.abi3.so

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:055fc9c5e82e48e503963bac3da30001e128774d8d9a333680b8aacab0650644
-size 324616

build/torch28-cxx11-cpu-x86_64-linux/layers.py

@@ -1,36 +0,0 @@
-import torch
-from ._ops import ops
-
-class RMSNorm(torch.nn.Module):
-    """
-    RMSNorm module that uses the optimized LigerRMSNormFunction.
-    
-    Args:
-        hidden_size (int): The size of the hidden dimension.
-        eps (float, optional): The epsilon value for numerical stability. Defaults to 1e-6.
-        offset (float, optional): Offset value to shift the weight tensor. Defaults to 0.0.
-        casting_mode (str, optional): The casting mode to use. Defaults to "llama".
-        in_place (bool, optional): Whether to modify dY in-place to store dX during backward. Defaults to True.
-    """
-    
-
-    weight: torch.Tensor
-    variance_epsilon: float
-    
-    def forward(self, hidden_states):
-        """
-        Apply RMS normalization to the input tensor.
-        
-        Args:
-            hidden_states (torch.Tensor): Input tensor of shape (B, T, H) or (BxT, H)
-            
-        Returns:
-            torch.Tensor: Normalized tensor of the same shape as input
-        """
-        return ops.apply_rms_norm(
-            hidden_states,
-            self.weight, 
-            self.variance_epsilon,
-        )
-    
-__all__ = ["RMSNorm"]

build/torch28-cxx11-cpu-x86_64-linux/metadata.json

@@ -1 +0,0 @@
-{"python-depends":[]}

build/torch28-cxx11-cpu-x86_64-linux/rmsnorm/__init__.py

@@ -1,26 +0,0 @@
-import ctypes
-import sys
-
-import importlib
-from pathlib import Path
-from types import ModuleType
-
-def _import_from_path(file_path: Path) -> ModuleType:
-    # We cannot use the module name as-is, after adding it to `sys.modules`,
-    # it would also be used for other imports. So, we make a module name that
-    # depends on the path for it to be unique using the hex-encoded hash of
-    # the path.
-    path_hash = "{:x}".format(ctypes.c_size_t(hash(file_path.absolute())).value)
-    module_name = path_hash
-    spec = importlib.util.spec_from_file_location(module_name, file_path)
-    if spec is None:
-        raise ImportError(f"Cannot load spec for {module_name} from {file_path}")
-    module = importlib.util.module_from_spec(spec)
-    if module is None:
-        raise ImportError(f"Cannot load module {module_name} from spec")
-    sys.modules[module_name] = module
-    spec.loader.exec_module(module)  # type: ignore
-    return module
-
-
-globals().update(vars(_import_from_path(Path(__file__).parent.parent / "__init__.py")))

build/torch28-cxx11-xpu20251-x86_64-linux/__init__.py

@@ -1,14 +0,0 @@
-from . import layers
-
-from ._ops import ops
-
-
-def apply_rms_norm(input, weight, eps):
-    return ops.apply_rms_norm(
-            input,
-            weight,
-            eps,
-    )
-
-__all__ = ["layers", "apply_rms_norm"]
-

build/torch28-cxx11-xpu20251-x86_64-linux/_ops.py

@@ -1,9 +0,0 @@
-import torch
-from . import _rmsnorm_fb26d8c
-ops = torch.ops._rmsnorm_fb26d8c
-
-def add_op_namespace_prefix(op_name: str):
-    """
-    Prefix op by namespace.
-    """
-    return f"_rmsnorm_fb26d8c::{op_name}"

build/torch28-cxx11-xpu20251-x86_64-linux/_rmsnorm_fb26d8c.abi3.so

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:c0ba9e0355977f76b16f6346377026ffde2977c613ee9b5633083d6f95f4e07c
-size 103861336

build/torch28-cxx11-xpu20251-x86_64-linux/layers.py

@@ -1,36 +0,0 @@
-import torch
-from ._ops import ops
-
-class RMSNorm(torch.nn.Module):
-    """
-    RMSNorm module that uses the optimized LigerRMSNormFunction.
-    
-    Args:
-        hidden_size (int): The size of the hidden dimension.
-        eps (float, optional): The epsilon value for numerical stability. Defaults to 1e-6.
-        offset (float, optional): Offset value to shift the weight tensor. Defaults to 0.0.
-        casting_mode (str, optional): The casting mode to use. Defaults to "llama".
-        in_place (bool, optional): Whether to modify dY in-place to store dX during backward. Defaults to True.
-    """
-    
-
-    weight: torch.Tensor
-    variance_epsilon: float
-    
-    def forward(self, hidden_states):
-        """
-        Apply RMS normalization to the input tensor.
-        
-        Args:
-            hidden_states (torch.Tensor): Input tensor of shape (B, T, H) or (BxT, H)
-            
-        Returns:
-            torch.Tensor: Normalized tensor of the same shape as input
-        """
-        return ops.apply_rms_norm(
-            hidden_states,
-            self.weight, 
-            self.variance_epsilon,
-        )
-    
-__all__ = ["RMSNorm"]

build/torch28-cxx11-xpu20251-x86_64-linux/metadata.json

@@ -1 +0,0 @@
-{"python-depends":[]}

build/torch28-cxx11-xpu20251-x86_64-linux/rmsnorm/__init__.py

@@ -1,26 +0,0 @@
-import ctypes
-import sys
-
-import importlib
-from pathlib import Path
-from types import ModuleType
-
-def _import_from_path(file_path: Path) -> ModuleType:
-    # We cannot use the module name as-is, after adding it to `sys.modules`,
-    # it would also be used for other imports. So, we make a module name that
-    # depends on the path for it to be unique using the hex-encoded hash of
-    # the path.
-    path_hash = "{:x}".format(ctypes.c_size_t(hash(file_path.absolute())).value)
-    module_name = path_hash
-    spec = importlib.util.spec_from_file_location(module_name, file_path)
-    if spec is None:
-        raise ImportError(f"Cannot load spec for {module_name} from {file_path}")
-    module = importlib.util.module_from_spec(spec)
-    if module is None:
-        raise ImportError(f"Cannot load module {module_name} from spec")
-    sys.modules[module_name] = module
-    spec.loader.exec_module(module)  # type: ignore
-    return module
-
-
-globals().update(vars(_import_from_path(Path(__file__).parent.parent / "__init__.py")))

build/torch29-cxx11-cpu-x86_64-linux/__init__.py

@@ -1,14 +0,0 @@
-from . import layers
-
-from ._ops import ops
-
-
-def apply_rms_norm(input, weight, eps):
-    return ops.apply_rms_norm(
-            input,
-            weight,
-            eps,
-    )
-
-__all__ = ["layers", "apply_rms_norm"]
-

build/torch29-cxx11-cpu-x86_64-linux/_ops.py

@@ -1,9 +0,0 @@
-import torch
-from . import _rmsnorm_fb26d8c
-ops = torch.ops._rmsnorm_fb26d8c
-
-def add_op_namespace_prefix(op_name: str):
-    """
-    Prefix op by namespace.
-    """
-    return f"_rmsnorm_fb26d8c::{op_name}"

build/torch29-cxx11-cpu-x86_64-linux/_rmsnorm_fb26d8c.abi3.so

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:97de49bd6f5edb8a54394a123362f30026a119e4f8ccf796884f108c343ec562
-size 324592

build/torch29-cxx11-cpu-x86_64-linux/layers.py

@@ -1,36 +0,0 @@
-import torch
-from ._ops import ops
-
-class RMSNorm(torch.nn.Module):
-    """
-    RMSNorm module that uses the optimized LigerRMSNormFunction.
-    
-    Args:
-        hidden_size (int): The size of the hidden dimension.
-        eps (float, optional): The epsilon value for numerical stability. Defaults to 1e-6.
-        offset (float, optional): Offset value to shift the weight tensor. Defaults to 0.0.
-        casting_mode (str, optional): The casting mode to use. Defaults to "llama".
-        in_place (bool, optional): Whether to modify dY in-place to store dX during backward. Defaults to True.
-    """
-    
-
-    weight: torch.Tensor
-    variance_epsilon: float
-    
-    def forward(self, hidden_states):
-        """
-        Apply RMS normalization to the input tensor.
-        
-        Args:
-            hidden_states (torch.Tensor): Input tensor of shape (B, T, H) or (BxT, H)
-            
-        Returns:
-            torch.Tensor: Normalized tensor of the same shape as input
-        """
-        return ops.apply_rms_norm(
-            hidden_states,
-            self.weight, 
-            self.variance_epsilon,
-        )
-    
-__all__ = ["RMSNorm"]

build/torch29-cxx11-cpu-x86_64-linux/metadata.json

@@ -1 +0,0 @@
-{"python-depends":[]}

build/torch29-cxx11-cpu-x86_64-linux/rmsnorm/__init__.py

@@ -1,26 +0,0 @@
-import ctypes
-import sys
-
-import importlib
-from pathlib import Path
-from types import ModuleType
-
-def _import_from_path(file_path: Path) -> ModuleType:
-    # We cannot use the module name as-is, after adding it to `sys.modules`,
-    # it would also be used for other imports. So, we make a module name that
-    # depends on the path for it to be unique using the hex-encoded hash of
-    # the path.
-    path_hash = "{:x}".format(ctypes.c_size_t(hash(file_path.absolute())).value)
-    module_name = path_hash
-    spec = importlib.util.spec_from_file_location(module_name, file_path)
-    if spec is None:
-        raise ImportError(f"Cannot load spec for {module_name} from {file_path}")
-    module = importlib.util.module_from_spec(spec)
-    if module is None:
-        raise ImportError(f"Cannot load module {module_name} from spec")
-    sys.modules[module_name] = module
-    spec.loader.exec_module(module)  # type: ignore
-    return module
-
-
-globals().update(vars(_import_from_path(Path(__file__).parent.parent / "__init__.py")))

build/torch29-cxx11-xpu20252-x86_64-linux/__init__.py

@@ -1,14 +0,0 @@
-from . import layers
-
-from ._ops import ops
-
-
-def apply_rms_norm(input, weight, eps):
-    return ops.apply_rms_norm(
-            input,
-            weight,
-            eps,
-    )
-
-__all__ = ["layers", "apply_rms_norm"]
-

build/torch29-cxx11-xpu20252-x86_64-linux/_ops.py

@@ -1,9 +0,0 @@
-import torch
-from . import _rmsnorm_fb26d8c
-ops = torch.ops._rmsnorm_fb26d8c
-
-def add_op_namespace_prefix(op_name: str):
-    """
-    Prefix op by namespace.
-    """
-    return f"_rmsnorm_fb26d8c::{op_name}"

build/torch29-cxx11-xpu20252-x86_64-linux/_rmsnorm_fb26d8c.abi3.so

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:47dcb713294a6eca6d920f2e9aba27be280d75ac2d356845232008210d1df17a
-size 102340240

build/torch29-cxx11-xpu20252-x86_64-linux/layers.py

@@ -1,36 +0,0 @@
-import torch
-from ._ops import ops
-
-class RMSNorm(torch.nn.Module):
-    """
-    RMSNorm module that uses the optimized LigerRMSNormFunction.
-    
-    Args:
-        hidden_size (int): The size of the hidden dimension.
-        eps (float, optional): The epsilon value for numerical stability. Defaults to 1e-6.
-        offset (float, optional): Offset value to shift the weight tensor. Defaults to 0.0.
-        casting_mode (str, optional): The casting mode to use. Defaults to "llama".
-        in_place (bool, optional): Whether to modify dY in-place to store dX during backward. Defaults to True.
-    """
-    
-
-    weight: torch.Tensor
-    variance_epsilon: float
-    
-    def forward(self, hidden_states):
-        """
-        Apply RMS normalization to the input tensor.
-        
-        Args:
-            hidden_states (torch.Tensor): Input tensor of shape (B, T, H) or (BxT, H)
-            
-        Returns:
-            torch.Tensor: Normalized tensor of the same shape as input
-        """
-        return ops.apply_rms_norm(
-            hidden_states,
-            self.weight, 
-            self.variance_epsilon,
-        )
-    
-__all__ = ["RMSNorm"]

build/torch29-cxx11-xpu20252-x86_64-linux/metadata.json

@@ -1 +0,0 @@
-{"python-depends":[]}

build/torch29-cxx11-xpu20252-x86_64-linux/rmsnorm/__init__.py

@@ -1,26 +0,0 @@
-import ctypes
-import sys
-
-import importlib
-from pathlib import Path
-from types import ModuleType
-
-def _import_from_path(file_path: Path) -> ModuleType:
-    # We cannot use the module name as-is, after adding it to `sys.modules`,
-    # it would also be used for other imports. So, we make a module name that
-    # depends on the path for it to be unique using the hex-encoded hash of
-    # the path.
-    path_hash = "{:x}".format(ctypes.c_size_t(hash(file_path.absolute())).value)
-    module_name = path_hash
-    spec = importlib.util.spec_from_file_location(module_name, file_path)
-    if spec is None:
-        raise ImportError(f"Cannot load spec for {module_name} from {file_path}")
-    module = importlib.util.module_from_spec(spec)
-    if module is None:
-        raise ImportError(f"Cannot load module {module_name} from spec")
-    sys.modules[module_name] = module
-    spec.loader.exec_module(module)  # type: ignore
-    return module
-
-
-globals().update(vars(_import_from_path(Path(__file__).parent.parent / "__init__.py")))