Add files using upload-large-folder tool

.gitattributes

@@ -81,3 +81,4 @@ tuning-competition-baseline/.venv/lib/python3.11/site-packages/pip/_vendor/distl
 tuning-competition-baseline/.venv/lib/python3.11/site-packages/nvidia/cuda_runtime/lib/libcudart.so.11.0 filter=lfs diff=lfs merge=lfs -text
 tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/_inductor/__pycache__/cudagraph_trees.cpython-311.pyc filter=lfs diff=lfs merge=lfs -text
 tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/nn/parallel/__pycache__/distributed.cpython-311.pyc filter=lfs diff=lfs merge=lfs -text
+tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/linalg/__pycache__/__init__.cpython-311.pyc filter=lfs diff=lfs merge=lfs -text

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_aminmax_compositeexplicitautograd_dispatch.h

@@ -0,0 +1,26 @@
+#pragma once
+// @generated by torchgen/gen.py from DispatchKeyFunction.h
+
+// NB: The implementing C++ file is RegisterDispatchKey.cpp
+
+// The only #includes we need are for custom classes that have defaults in the C++ API
+#include <c10/core/MemoryFormat.h>
+#include <c10/core/Scalar.h>
+#include <ATen/core/Reduction.h>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+
+namespace compositeexplicitautograd {
+
+TORCH_API ::std::tuple<at::Tensor &,at::Tensor &> _aminmax_out(at::Tensor & out0, at::Tensor & out1, const at::Tensor & self);
+TORCH_API ::std::tuple<at::Tensor &,at::Tensor &> _aminmax_outf(const at::Tensor & self, at::Tensor & out0, at::Tensor & out1);
+TORCH_API ::std::tuple<at::Tensor &,at::Tensor &> _aminmax_out(at::Tensor & out0, at::Tensor & out1, const at::Tensor & self, int64_t dim, bool keepdim=false);
+TORCH_API ::std::tuple<at::Tensor &,at::Tensor &> _aminmax_outf(const at::Tensor & self, int64_t dim, bool keepdim, at::Tensor & out0, at::Tensor & out1);
+
+} // namespace compositeexplicitautograd
+} // namespace at

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_efficientzerotensor_cuda_dispatch.h

@@ -0,0 +1,26 @@
+#pragma once
+// @generated by torchgen/gen.py from DispatchKeyFunction.h
+
+// NB: The implementing C++ file is RegisterDispatchKey.cpp
+
+// The only #includes we need are for custom classes that have defaults in the C++ API
+#include <c10/core/MemoryFormat.h>
+#include <c10/core/Scalar.h>
+#include <ATen/core/Reduction.h>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+
+namespace cuda {
+
+TORCH_API at::Tensor _efficientzerotensor(at::IntArrayRef size, at::TensorOptions options={});
+TORCH_API at::Tensor _efficientzerotensor(at::IntArrayRef size, c10::optional<at::ScalarType> dtype, c10::optional<at::Layout> layout, c10::optional<at::Device> device, c10::optional<bool> pin_memory);
+TORCH_API at::Tensor _efficientzerotensor_symint(c10::SymIntArrayRef size, at::TensorOptions options={});
+TORCH_API at::Tensor _efficientzerotensor_symint(c10::SymIntArrayRef size, c10::optional<at::ScalarType> dtype, c10::optional<at::Layout> layout, c10::optional<at::Device> device, c10::optional<bool> pin_memory);
+
+} // namespace cuda
+} // namespace at

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_fft_c2r_ops.h

@@ -0,0 +1,39 @@
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <tuple>
+#include <vector>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+namespace _ops {
+
+
+struct TORCH_API _fft_c2r {
+  using schema = at::Tensor (const at::Tensor &, at::IntArrayRef, int64_t, c10::SymInt);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(name, "aten::_fft_c2r")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "_fft_c2r(Tensor self, int[] dim, int normalization, SymInt last_dim_size) -> Tensor")
+  static at::Tensor call(const at::Tensor & self, at::IntArrayRef dim, int64_t normalization, c10::SymInt last_dim_size);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, at::IntArrayRef dim, int64_t normalization, c10::SymInt last_dim_size);
+};
+
+struct TORCH_API _fft_c2r_out {
+  using schema = at::Tensor & (const at::Tensor &, at::IntArrayRef, int64_t, c10::SymInt, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(name, "aten::_fft_c2r")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "out")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "_fft_c2r.out(Tensor self, int[] dim, int normalization, SymInt last_dim_size, *, Tensor(a!) out) -> Tensor(a!)")
+  static at::Tensor & call(const at::Tensor & self, at::IntArrayRef dim, int64_t normalization, c10::SymInt last_dim_size, at::Tensor & out);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, at::IntArrayRef dim, int64_t normalization, c10::SymInt last_dim_size, at::Tensor & out);
+};
+
+}} // namespace at::_ops

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_foreach_addcdiv_ops.h

@@ -0,0 +1,116 @@
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <tuple>
+#include <vector>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+namespace _ops {
+
+
+struct TORCH_API _foreach_addcdiv_Scalar {
+  using schema = ::std::vector<at::Tensor> (at::TensorList, at::TensorList, at::TensorList, const at::Scalar &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(name, "aten::_foreach_addcdiv")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "Scalar")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "_foreach_addcdiv.Scalar(Tensor[] self, Tensor[] tensor1, Tensor[] tensor2, Scalar value=1) -> Tensor[]")
+  static ::std::vector<at::Tensor> call(at::TensorList self, at::TensorList tensor1, at::TensorList tensor2, const at::Scalar & value);
+  static ::std::vector<at::Tensor> redispatch(c10::DispatchKeySet dispatchKeySet, at::TensorList self, at::TensorList tensor1, at::TensorList tensor2, const at::Scalar & value);
+};
+
+struct TORCH_API _foreach_addcdiv_ScalarList {
+  using schema = ::std::vector<at::Tensor> (at::TensorList, at::TensorList, at::TensorList, at::ArrayRef<at::Scalar>);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(name, "aten::_foreach_addcdiv")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "ScalarList")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "_foreach_addcdiv.ScalarList(Tensor[] self, Tensor[] tensor1, Tensor[] tensor2, Scalar[] scalars) -> Tensor[]")
+  static ::std::vector<at::Tensor> call(at::TensorList self, at::TensorList tensor1, at::TensorList tensor2, at::ArrayRef<at::Scalar> scalars);
+  static ::std::vector<at::Tensor> redispatch(c10::DispatchKeySet dispatchKeySet, at::TensorList self, at::TensorList tensor1, at::TensorList tensor2, at::ArrayRef<at::Scalar> scalars);
+};
+
+struct TORCH_API _foreach_addcdiv_Tensor {
+  using schema = ::std::vector<at::Tensor> (at::TensorList, at::TensorList, at::TensorList, const at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(name, "aten::_foreach_addcdiv")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "Tensor")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "_foreach_addcdiv.Tensor(Tensor[] self, Tensor[] tensor1, Tensor[] tensor2, Tensor scalars) -> Tensor[]")
+  static ::std::vector<at::Tensor> call(at::TensorList self, at::TensorList tensor1, at::TensorList tensor2, const at::Tensor & scalars);
+  static ::std::vector<at::Tensor> redispatch(c10::DispatchKeySet dispatchKeySet, at::TensorList self, at::TensorList tensor1, at::TensorList tensor2, const at::Tensor & scalars);
+};
+
+struct TORCH_API _foreach_addcdiv__Scalar {
+  using schema = void (at::TensorList, at::TensorList, at::TensorList, const at::Scalar &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(name, "aten::_foreach_addcdiv_")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "Scalar")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "_foreach_addcdiv_.Scalar(Tensor(a!)[] self, Tensor[] tensor1, Tensor[] tensor2, Scalar value=1) -> ()")
+  static void call(at::TensorList self, at::TensorList tensor1, at::TensorList tensor2, const at::Scalar & value);
+  static void redispatch(c10::DispatchKeySet dispatchKeySet, at::TensorList self, at::TensorList tensor1, at::TensorList tensor2, const at::Scalar & value);
+};
+
+struct TORCH_API _foreach_addcdiv__ScalarList {
+  using schema = void (at::TensorList, at::TensorList, at::TensorList, at::ArrayRef<at::Scalar>);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(name, "aten::_foreach_addcdiv_")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "ScalarList")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "_foreach_addcdiv_.ScalarList(Tensor(a!)[] self, Tensor[] tensor1, Tensor[] tensor2, Scalar[] scalars) -> ()")
+  static void call(at::TensorList self, at::TensorList tensor1, at::TensorList tensor2, at::ArrayRef<at::Scalar> scalars);
+  static void redispatch(c10::DispatchKeySet dispatchKeySet, at::TensorList self, at::TensorList tensor1, at::TensorList tensor2, at::ArrayRef<at::Scalar> scalars);
+};
+
+struct TORCH_API _foreach_addcdiv__Tensor {
+  using schema = void (at::TensorList, at::TensorList, at::TensorList, const at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(name, "aten::_foreach_addcdiv_")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "Tensor")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "_foreach_addcdiv_.Tensor(Tensor(a!)[] self, Tensor[] tensor1, Tensor[] tensor2, Tensor scalars) -> ()")
+  static void call(at::TensorList self, at::TensorList tensor1, at::TensorList tensor2, const at::Tensor & scalars);
+  static void redispatch(c10::DispatchKeySet dispatchKeySet, at::TensorList self, at::TensorList tensor1, at::TensorList tensor2, const at::Tensor & scalars);
+};
+
+struct TORCH_API _foreach_addcdiv_Scalar_out {
+  using schema = void (at::TensorList, at::TensorList, at::TensorList, const at::Scalar &, at::TensorList);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(name, "aten::_foreach_addcdiv")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "Scalar_out")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "_foreach_addcdiv.Scalar_out(Tensor[] self, Tensor[] tensor1, Tensor[] tensor2, Scalar value=1, *, Tensor(a!)[] out) -> ()")
+  static void call(at::TensorList self, at::TensorList tensor1, at::TensorList tensor2, const at::Scalar & value, at::TensorList out);
+  static void redispatch(c10::DispatchKeySet dispatchKeySet, at::TensorList self, at::TensorList tensor1, at::TensorList tensor2, const at::Scalar & value, at::TensorList out);
+};
+
+struct TORCH_API _foreach_addcdiv_ScalarList_out {
+  using schema = void (at::TensorList, at::TensorList, at::TensorList, at::ArrayRef<at::Scalar>, at::TensorList);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(name, "aten::_foreach_addcdiv")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "ScalarList_out")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "_foreach_addcdiv.ScalarList_out(Tensor[] self, Tensor[] tensor1, Tensor[] tensor2, Scalar[] scalars, *, Tensor(a!)[] out) -> ()")
+  static void call(at::TensorList self, at::TensorList tensor1, at::TensorList tensor2, at::ArrayRef<at::Scalar> scalars, at::TensorList out);
+  static void redispatch(c10::DispatchKeySet dispatchKeySet, at::TensorList self, at::TensorList tensor1, at::TensorList tensor2, at::ArrayRef<at::Scalar> scalars, at::TensorList out);
+};
+
+struct TORCH_API _foreach_addcdiv_Tensor_out {
+  using schema = void (at::TensorList, at::TensorList, at::TensorList, const at::Tensor &, at::TensorList);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(name, "aten::_foreach_addcdiv")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "Tensor_out")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "_foreach_addcdiv.Tensor_out(Tensor[] self, Tensor[] tensor1, Tensor[] tensor2, Tensor scalars, *, Tensor(a!)[] out) -> ()")
+  static void call(at::TensorList self, at::TensorList tensor1, at::TensorList tensor2, const at::Tensor & scalars, at::TensorList out);
+  static void redispatch(c10::DispatchKeySet dispatchKeySet, at::TensorList self, at::TensorList tensor1, at::TensorList tensor2, const at::Tensor & scalars, at::TensorList out);
+};
+
+}} // namespace at::_ops

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_foreach_copy.h

@@ -0,0 +1,44 @@
+#pragma once
+
+// @generated by torchgen/gen.py from Function.h
+
+#include <ATen/Context.h>
+#include <ATen/DeviceGuard.h>
+#include <ATen/TensorUtils.h>
+#include <ATen/TracerMode.h>
+#include <ATen/core/Generator.h>
+#include <ATen/core/Reduction.h>
+#include <ATen/core/Tensor.h>
+#include <c10/core/Scalar.h>
+#include <c10/core/Storage.h>
+#include <c10/core/TensorOptions.h>
+#include <c10/util/Deprecated.h>
+#include <c10/util/Optional.h>
+
+
+
+#include <ATen/ops/_foreach_copy_ops.h>
+
+namespace at {
+
+
+// aten::_foreach_copy_(Tensor(a!)[] self, Tensor[] src, bool non_blocking=False) -> ()
+inline void _foreach_copy_(at::TensorList self, at::TensorList src, bool non_blocking=false) {
+    return at::_ops::_foreach_copy_::call(self, src, non_blocking);
+}
+
+// aten::_foreach_copy.out(Tensor[] self, Tensor[] src, bool non_blocking=False, *, Tensor(a!)[] out) -> ()
+inline void _foreach_copy_out(at::TensorList out, at::TensorList self, at::TensorList src, bool non_blocking=false) {
+    return at::_ops::_foreach_copy_out::call(self, src, non_blocking, out);
+}
+// aten::_foreach_copy.out(Tensor[] self, Tensor[] src, bool non_blocking=False, *, Tensor(a!)[] out) -> ()
+inline void _foreach_copy_outf(at::TensorList self, at::TensorList src, bool non_blocking, at::TensorList out) {
+    return at::_ops::_foreach_copy_out::call(self, src, non_blocking, out);
+}
+
+// aten::_foreach_copy(Tensor[] self, Tensor[] src, bool non_blocking=False) -> Tensor[] self_out
+inline ::std::vector<at::Tensor> _foreach_copy(at::TensorList self, at::TensorList src, bool non_blocking=false) {
+    return at::_ops::_foreach_copy::call(self, src, non_blocking);
+}
+
+}

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_foreach_norm_cuda_dispatch.h

@@ -0,0 +1,23 @@
+#pragma once
+// @generated by torchgen/gen.py from DispatchKeyFunction.h
+
+// NB: The implementing C++ file is RegisterDispatchKey.cpp
+
+// The only #includes we need are for custom classes that have defaults in the C++ API
+#include <c10/core/MemoryFormat.h>
+#include <c10/core/Scalar.h>
+#include <ATen/core/Reduction.h>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+
+namespace cuda {
+
+TORCH_API ::std::vector<at::Tensor> _foreach_norm(at::TensorList self, const at::Scalar & ord=2);
+
+} // namespace cuda
+} // namespace at

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_int_mm.h

@@ -0,0 +1,39 @@
+#pragma once
+
+// @generated by torchgen/gen.py from Function.h
+
+#include <ATen/Context.h>
+#include <ATen/DeviceGuard.h>
+#include <ATen/TensorUtils.h>
+#include <ATen/TracerMode.h>
+#include <ATen/core/Generator.h>
+#include <ATen/core/Reduction.h>
+#include <ATen/core/Tensor.h>
+#include <c10/core/Scalar.h>
+#include <c10/core/Storage.h>
+#include <c10/core/TensorOptions.h>
+#include <c10/util/Deprecated.h>
+#include <c10/util/Optional.h>
+
+
+
+#include <ATen/ops/_int_mm_ops.h>
+
+namespace at {
+
+
+// aten::_int_mm(Tensor self, Tensor mat2) -> Tensor
+inline at::Tensor _int_mm(const at::Tensor & self, const at::Tensor & mat2) {
+    return at::_ops::_int_mm::call(self, mat2);
+}
+
+// aten::_int_mm.out(Tensor self, Tensor mat2, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & _int_mm_out(at::Tensor & out, const at::Tensor & self, const at::Tensor & mat2) {
+    return at::_ops::_int_mm_out::call(self, mat2, out);
+}
+// aten::_int_mm.out(Tensor self, Tensor mat2, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & _int_mm_outf(const at::Tensor & self, const at::Tensor & mat2, at::Tensor & out) {
+    return at::_ops::_int_mm_out::call(self, mat2, out);
+}
+
+}

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_nested_get_lengths_ops.h

@@ -0,0 +1,28 @@
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <tuple>
+#include <vector>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+namespace _ops {
+
+
+struct TORCH_API _nested_get_lengths {
+  using schema = at::Tensor (const at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(name, "aten::_nested_get_lengths")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "_nested_get_lengths(Tensor self) -> Tensor")
+  static at::Tensor call(const at::Tensor & self);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self);
+};
+
+}} // namespace at::_ops

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_prelu_kernel_backward_native.h

@@ -0,0 +1,22 @@
+#pragma once
+
+// @generated by torchgen/gen.py from NativeFunction.h
+
+#include <c10/core/Scalar.h>
+#include <c10/core/Storage.h>
+#include <c10/core/TensorOptions.h>
+#include <c10/util/Deprecated.h>
+#include <c10/util/Optional.h>
+#include <c10/core/QScheme.h>
+#include <ATen/core/Reduction.h>
+#include <ATen/core/Tensor.h>
+#include <tuple>
+#include <vector>
+
+
+namespace at {
+namespace native {
+TORCH_API ::std::tuple<at::Tensor,at::Tensor> _prelu_kernel_backward(const at::Tensor & grad_output, const at::Tensor & self, const at::Tensor & weight);
+TORCH_API ::std::tuple<at::Tensor,at::Tensor> mkldnn_prelu_backward(const at::Tensor & grad_output, const at::Tensor & self, const at::Tensor & weight);
+} // namespace native
+} // namespace at

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_remove_batch_dim_native.h

@@ -0,0 +1,21 @@
+#pragma once
+
+// @generated by torchgen/gen.py from NativeFunction.h
+
+#include <c10/core/Scalar.h>
+#include <c10/core/Storage.h>
+#include <c10/core/TensorOptions.h>
+#include <c10/util/Deprecated.h>
+#include <c10/util/Optional.h>
+#include <c10/core/QScheme.h>
+#include <ATen/core/Reduction.h>
+#include <ATen/core/Tensor.h>
+#include <tuple>
+#include <vector>
+
+
+namespace at {
+namespace native {
+TORCH_API at::Tensor _remove_batch_dim(const at::Tensor & self, int64_t level, int64_t batch_size, int64_t out_dim);
+} // namespace native
+} // namespace at

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_softmax_backward_data_meta.h

@@ -0,0 +1,27 @@
+#pragma once
+
+// @generated by torchgen/gen.py from NativeMetaFunction.h
+
+#include <c10/core/Scalar.h>
+#include <c10/core/Storage.h>
+#include <c10/core/TensorOptions.h>
+#include <c10/util/Deprecated.h>
+#include <c10/util/Optional.h>
+#include <c10/core/QScheme.h>
+#include <ATen/core/Reduction.h>
+#include <ATen/TensorIterator.h>
+#include <ATen/TensorMeta.h>
+#include <tuple>
+#include <vector>
+
+namespace at {
+namespace meta {
+
+struct TORCH_API structured__softmax_backward_data : public at::impl::MetaBase {
+    
+    
+    void meta(const at::Tensor & grad_output, const at::Tensor & output, int64_t dim, at::ScalarType input_dtype);
+};
+
+} // namespace native
+} // namespace at

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_sparse_bsr_tensor_unsafe_native.h

@@ -0,0 +1,21 @@
+#pragma once
+
+// @generated by torchgen/gen.py from NativeFunction.h
+
+#include <c10/core/Scalar.h>
+#include <c10/core/Storage.h>
+#include <c10/core/TensorOptions.h>
+#include <c10/util/Deprecated.h>
+#include <c10/util/Optional.h>
+#include <c10/core/QScheme.h>
+#include <ATen/core/Reduction.h>
+#include <ATen/core/Tensor.h>
+#include <tuple>
+#include <vector>
+
+
+namespace at {
+namespace native {
+TORCH_API at::Tensor _sparse_bsr_tensor_unsafe(const at::Tensor & crow_indices, const at::Tensor & col_indices, const at::Tensor & values, at::IntArrayRef size, c10::optional<at::ScalarType> dtype={}, c10::optional<at::Layout> layout={}, c10::optional<at::Device> device={}, c10::optional<bool> pin_memory={});
+} // namespace native
+} // namespace at

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_upsample_nearest_exact1d_ops.h

@@ -0,0 +1,50 @@
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <tuple>
+#include <vector>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+namespace _ops {
+
+
+struct TORCH_API _upsample_nearest_exact1d_vec {
+  using schema = at::Tensor (const at::Tensor &, at::OptionalSymIntArrayRef, c10::optional<at::ArrayRef<double>>);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(name, "aten::_upsample_nearest_exact1d")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "vec")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "_upsample_nearest_exact1d.vec(Tensor input, SymInt[]? output_size, float[]? scale_factors) -> Tensor")
+  static at::Tensor call(const at::Tensor & input, at::OptionalSymIntArrayRef output_size, c10::optional<at::ArrayRef<double>> scale_factors);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & input, at::OptionalSymIntArrayRef output_size, c10::optional<at::ArrayRef<double>> scale_factors);
+};
+
+struct TORCH_API _upsample_nearest_exact1d_out {
+  using schema = at::Tensor & (const at::Tensor &, c10::SymIntArrayRef, c10::optional<double>, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(name, "aten::_upsample_nearest_exact1d")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "out")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "_upsample_nearest_exact1d.out(Tensor self, SymInt[1] output_size, float? scales=None, *, Tensor(a!) out) -> Tensor(a!)")
+  static at::Tensor & call(const at::Tensor & self, c10::SymIntArrayRef output_size, c10::optional<double> scales, at::Tensor & out);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, c10::SymIntArrayRef output_size, c10::optional<double> scales, at::Tensor & out);
+};
+
+struct TORCH_API _upsample_nearest_exact1d {
+  using schema = at::Tensor (const at::Tensor &, c10::SymIntArrayRef, c10::optional<double>);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(name, "aten::_upsample_nearest_exact1d")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "_upsample_nearest_exact1d(Tensor self, SymInt[1] output_size, float? scales=None) -> Tensor")
+  static at::Tensor call(const at::Tensor & self, c10::SymIntArrayRef output_size, c10::optional<double> scales);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, c10::SymIntArrayRef output_size, c10::optional<double> scales);
+};
+
+}} // namespace at::_ops

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_validate_sparse_csr_tensor_args_native.h

@@ -0,0 +1,21 @@
+#pragma once
+
+// @generated by torchgen/gen.py from NativeFunction.h
+
+#include <c10/core/Scalar.h>
+#include <c10/core/Storage.h>
+#include <c10/core/TensorOptions.h>
+#include <c10/util/Deprecated.h>
+#include <c10/util/Optional.h>
+#include <c10/core/QScheme.h>
+#include <ATen/core/Reduction.h>
+#include <ATen/core/Tensor.h>
+#include <tuple>
+#include <vector>
+
+
+namespace at {
+namespace native {
+TORCH_API void _validate_sparse_csr_tensor_args(const at::Tensor & crow_indices, const at::Tensor & col_indices, const at::Tensor & values, at::IntArrayRef size);
+} // namespace native
+} // namespace at

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/adaptive_avg_pool1d_compositeimplicitautograd_dispatch.h

@@ -0,0 +1,23 @@
+#pragma once
+// @generated by torchgen/gen.py from DispatchKeyFunction.h
+
+// NB: The implementing C++ file is RegisterDispatchKey.cpp
+
+// The only #includes we need are for custom classes that have defaults in the C++ API
+#include <c10/core/MemoryFormat.h>
+#include <c10/core/Scalar.h>
+#include <ATen/core/Reduction.h>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+
+namespace compositeimplicitautograd {
+
+TORCH_API at::Tensor adaptive_avg_pool1d(const at::Tensor & self, at::IntArrayRef output_size);
+
+} // namespace compositeimplicitautograd
+} // namespace at

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/addmm_meta_dispatch.h

@@ -0,0 +1,26 @@
+#pragma once
+// @generated by torchgen/gen.py from DispatchKeyFunction.h
+
+// NB: The implementing C++ file is RegisterDispatchKey.cpp
+
+// The only #includes we need are for custom classes that have defaults in the C++ API
+#include <c10/core/MemoryFormat.h>
+#include <c10/core/Scalar.h>
+#include <ATen/core/Reduction.h>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+
+namespace meta {
+
+TORCH_API at::Tensor addmm(const at::Tensor & self, const at::Tensor & mat1, const at::Tensor & mat2, const at::Scalar & beta=1, const at::Scalar & alpha=1);
+TORCH_API at::Tensor & addmm_out(at::Tensor & out, const at::Tensor & self, const at::Tensor & mat1, const at::Tensor & mat2, const at::Scalar & beta=1, const at::Scalar & alpha=1);
+TORCH_API at::Tensor & addmm_outf(const at::Tensor & self, const at::Tensor & mat1, const at::Tensor & mat2, const at::Scalar & beta, const at::Scalar & alpha, at::Tensor & out);
+TORCH_API at::Tensor & addmm_(at::Tensor & self, const at::Tensor & mat1, const at::Tensor & mat2, const at::Scalar & beta=1, const at::Scalar & alpha=1);
+
+} // namespace meta
+} // namespace at

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/amax_compositeexplicitautogradnonfunctional_dispatch.h

@@ -0,0 +1,23 @@
+#pragma once
+// @generated by torchgen/gen.py from DispatchKeyFunction.h
+
+// NB: The implementing C++ file is RegisterDispatchKey.cpp
+
+// The only #includes we need are for custom classes that have defaults in the C++ API
+#include <c10/core/MemoryFormat.h>
+#include <c10/core/Scalar.h>
+#include <ATen/core/Reduction.h>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+
+namespace compositeexplicitautogradnonfunctional {
+
+TORCH_API at::Tensor amax(const at::Tensor & self, at::IntArrayRef dim={}, bool keepdim=false);
+
+} // namespace compositeexplicitautogradnonfunctional
+} // namespace at

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/aminmax_cpu_dispatch.h

@@ -0,0 +1,25 @@
+#pragma once
+// @generated by torchgen/gen.py from DispatchKeyFunction.h
+
+// NB: The implementing C++ file is RegisterDispatchKey.cpp
+
+// The only #includes we need are for custom classes that have defaults in the C++ API
+#include <c10/core/MemoryFormat.h>
+#include <c10/core/Scalar.h>
+#include <ATen/core/Reduction.h>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+
+namespace cpu {
+
+TORCH_API ::std::tuple<at::Tensor,at::Tensor> aminmax(const at::Tensor & self, c10::optional<int64_t> dim=c10::nullopt, bool keepdim=false);
+TORCH_API ::std::tuple<at::Tensor &,at::Tensor &> aminmax_out(at::Tensor & min, at::Tensor & max, const at::Tensor & self, c10::optional<int64_t> dim=c10::nullopt, bool keepdim=false);
+TORCH_API ::std::tuple<at::Tensor &,at::Tensor &> aminmax_outf(const at::Tensor & self, c10::optional<int64_t> dim, bool keepdim, at::Tensor & min, at::Tensor & max);
+
+} // namespace cpu
+} // namespace at

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/any_cpu_dispatch.h

@@ -0,0 +1,31 @@
+#pragma once
+// @generated by torchgen/gen.py from DispatchKeyFunction.h
+
+// NB: The implementing C++ file is RegisterDispatchKey.cpp
+
+// The only #includes we need are for custom classes that have defaults in the C++ API
+#include <c10/core/MemoryFormat.h>
+#include <c10/core/Scalar.h>
+#include <ATen/core/Reduction.h>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+
+namespace cpu {
+
+TORCH_API at::Tensor any(const at::Tensor & self, int64_t dim, bool keepdim=false);
+TORCH_API at::Tensor & any_out(at::Tensor & out, const at::Tensor & self, int64_t dim, bool keepdim=false);
+TORCH_API at::Tensor & any_outf(const at::Tensor & self, int64_t dim, bool keepdim, at::Tensor & out);
+TORCH_API at::Tensor any(const at::Tensor & self, at::OptionalIntArrayRef dim, bool keepdim=false);
+TORCH_API at::Tensor & any_out(at::Tensor & out, const at::Tensor & self, at::OptionalIntArrayRef dim, bool keepdim=false);
+TORCH_API at::Tensor & any_outf(const at::Tensor & self, at::OptionalIntArrayRef dim, bool keepdim, at::Tensor & out);
+TORCH_API at::Tensor any(const at::Tensor & self);
+TORCH_API at::Tensor & any_out(at::Tensor & out, const at::Tensor & self);
+TORCH_API at::Tensor & any_outf(const at::Tensor & self, at::Tensor & out);
+
+} // namespace cpu
+} // namespace at

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/arccosh_native.h

@@ -0,0 +1,23 @@
+#pragma once
+
+// @generated by torchgen/gen.py from NativeFunction.h
+
+#include <c10/core/Scalar.h>
+#include <c10/core/Storage.h>
+#include <c10/core/TensorOptions.h>
+#include <c10/util/Deprecated.h>
+#include <c10/util/Optional.h>
+#include <c10/core/QScheme.h>
+#include <ATen/core/Reduction.h>
+#include <ATen/core/Tensor.h>
+#include <tuple>
+#include <vector>
+
+
+namespace at {
+namespace native {
+TORCH_API at::Tensor arccosh(const at::Tensor & self);
+TORCH_API at::Tensor & arccosh_out(const at::Tensor & self, at::Tensor & out);
+TORCH_API at::Tensor & arccosh_(at::Tensor & self);
+} // namespace native
+} // namespace at

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/bitwise_not_compositeexplicitautogradnonfunctional_dispatch.h

@@ -0,0 +1,24 @@
+#pragma once
+// @generated by torchgen/gen.py from DispatchKeyFunction.h
+
+// NB: The implementing C++ file is RegisterDispatchKey.cpp
+
+// The only #includes we need are for custom classes that have defaults in the C++ API
+#include <c10/core/MemoryFormat.h>
+#include <c10/core/Scalar.h>
+#include <ATen/core/Reduction.h>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+
+namespace compositeexplicitautogradnonfunctional {
+
+TORCH_API at::Tensor bitwise_not(const at::Tensor & self);
+TORCH_API at::Tensor & bitwise_not_(at::Tensor & self);
+
+} // namespace compositeexplicitautogradnonfunctional
+} // namespace at

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/bmm_meta_dispatch.h

@@ -0,0 +1,25 @@
+#pragma once
+// @generated by torchgen/gen.py from DispatchKeyFunction.h
+
+// NB: The implementing C++ file is RegisterDispatchKey.cpp
+
+// The only #includes we need are for custom classes that have defaults in the C++ API
+#include <c10/core/MemoryFormat.h>
+#include <c10/core/Scalar.h>
+#include <ATen/core/Reduction.h>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+
+namespace meta {
+
+TORCH_API at::Tensor bmm(const at::Tensor & self, const at::Tensor & mat2);
+TORCH_API at::Tensor & bmm_out(at::Tensor & out, const at::Tensor & self, const at::Tensor & mat2);
+TORCH_API at::Tensor & bmm_outf(const at::Tensor & self, const at::Tensor & mat2, at::Tensor & out);
+
+} // namespace meta
+} // namespace at

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/conv_transpose1d_compositeimplicitautograd_dispatch.h

@@ -0,0 +1,24 @@
+#pragma once
+// @generated by torchgen/gen.py from DispatchKeyFunction.h
+
+// NB: The implementing C++ file is RegisterDispatchKey.cpp
+
+// The only #includes we need are for custom classes that have defaults in the C++ API
+#include <c10/core/MemoryFormat.h>
+#include <c10/core/Scalar.h>
+#include <ATen/core/Reduction.h>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+
+namespace compositeimplicitautograd {
+
+TORCH_API at::Tensor conv_transpose1d(const at::Tensor & input, const at::Tensor & weight, const c10::optional<at::Tensor> & bias={}, at::IntArrayRef stride=1, at::IntArrayRef padding=0, at::IntArrayRef output_padding=0, int64_t groups=1, at::IntArrayRef dilation=1);
+TORCH_API at::Tensor conv_transpose1d_symint(const at::Tensor & input, const at::Tensor & weight, const c10::optional<at::Tensor> & bias={}, c10::SymIntArrayRef stride=c10::SymInt(1), c10::SymIntArrayRef padding=c10::SymInt(0), c10::SymIntArrayRef output_padding=c10::SymInt(0), c10::SymInt groups=1, c10::SymIntArrayRef dilation=c10::SymInt(1));
+
+} // namespace compositeimplicitautograd
+} // namespace at

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/convolution_backward_overrideable_ops.h

@@ -0,0 +1,39 @@
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <tuple>
+#include <vector>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+namespace _ops {
+
+
+struct TORCH_API convolution_backward_overrideable {
+  using schema = ::std::tuple<at::Tensor,at::Tensor,at::Tensor> (const at::Tensor &, const at::Tensor &, const at::Tensor &, c10::SymIntArrayRef, c10::SymIntArrayRef, c10::SymIntArrayRef, bool, c10::SymIntArrayRef, c10::SymInt, ::std::array<bool,3>);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(name, "aten::convolution_backward_overrideable")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "convolution_backward_overrideable(Tensor grad_output, Tensor input, Tensor weight, SymInt[] stride, SymInt[] padding, SymInt[] dilation, bool transposed, SymInt[] output_padding, SymInt groups, bool[3] output_mask) -> (Tensor grad_input, Tensor grad_weight, Tensor grad_bias)")
+  static ::std::tuple<at::Tensor,at::Tensor,at::Tensor> call(const at::Tensor & grad_output, const at::Tensor & input, const at::Tensor & weight, c10::SymIntArrayRef stride, c10::SymIntArrayRef padding, c10::SymIntArrayRef dilation, bool transposed, c10::SymIntArrayRef output_padding, c10::SymInt groups, ::std::array<bool,3> output_mask);
+  static ::std::tuple<at::Tensor,at::Tensor,at::Tensor> redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & grad_output, const at::Tensor & input, const at::Tensor & weight, c10::SymIntArrayRef stride, c10::SymIntArrayRef padding, c10::SymIntArrayRef dilation, bool transposed, c10::SymIntArrayRef output_padding, c10::SymInt groups, ::std::array<bool,3> output_mask);
+};
+
+struct TORCH_API convolution_backward_overrideable_out {
+  using schema = ::std::tuple<at::Tensor &,at::Tensor &,at::Tensor &> (const at::Tensor &, const at::Tensor &, const at::Tensor &, c10::SymIntArrayRef, c10::SymIntArrayRef, c10::SymIntArrayRef, bool, c10::SymIntArrayRef, c10::SymInt, ::std::array<bool,3>, at::Tensor &, at::Tensor &, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(name, "aten::convolution_backward_overrideable")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "out")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "convolution_backward_overrideable.out(Tensor grad_output, Tensor input, Tensor weight, SymInt[] stride, SymInt[] padding, SymInt[] dilation, bool transposed, SymInt[] output_padding, SymInt groups, bool[3] output_mask, *, Tensor(a!) out0, Tensor(b!) out1, Tensor(c!) out2) -> (Tensor(a!), Tensor(b!), Tensor(c!))")
+  static ::std::tuple<at::Tensor &,at::Tensor &,at::Tensor &> call(const at::Tensor & grad_output, const at::Tensor & input, const at::Tensor & weight, c10::SymIntArrayRef stride, c10::SymIntArrayRef padding, c10::SymIntArrayRef dilation, bool transposed, c10::SymIntArrayRef output_padding, c10::SymInt groups, ::std::array<bool,3> output_mask, at::Tensor & out0, at::Tensor & out1, at::Tensor & out2);
+  static ::std::tuple<at::Tensor &,at::Tensor &,at::Tensor &> redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & grad_output, const at::Tensor & input, const at::Tensor & weight, c10::SymIntArrayRef stride, c10::SymIntArrayRef padding, c10::SymIntArrayRef dilation, bool transposed, c10::SymIntArrayRef output_padding, c10::SymInt groups, ::std::array<bool,3> output_mask, at::Tensor & out0, at::Tensor & out1, at::Tensor & out2);
+};
+
+}} // namespace at::_ops

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/convolution_overrideable_native.h

@@ -0,0 +1,22 @@
+#pragma once
+
+// @generated by torchgen/gen.py from NativeFunction.h
+
+#include <c10/core/Scalar.h>
+#include <c10/core/Storage.h>
+#include <c10/core/TensorOptions.h>
+#include <c10/util/Deprecated.h>
+#include <c10/util/Optional.h>
+#include <c10/core/QScheme.h>
+#include <ATen/core/Reduction.h>
+#include <ATen/core/Tensor.h>
+#include <tuple>
+#include <vector>
+
+
+namespace at {
+namespace native {
+TORCH_API at::Tensor convolution_overrideable(const at::Tensor & input, const at::Tensor & weight, const c10::optional<at::Tensor> & bias, at::IntArrayRef stride, at::IntArrayRef padding, at::IntArrayRef dilation, bool transposed, at::IntArrayRef output_padding, int64_t groups);
+TORCH_API at::Tensor & convolution_overrideable_out_symint(const at::Tensor & input, const at::Tensor & weight, const c10::optional<at::Tensor> & bias, c10::SymIntArrayRef stride, c10::SymIntArrayRef padding, c10::SymIntArrayRef dilation, bool transposed, c10::SymIntArrayRef output_padding, c10::SymInt groups, at::Tensor & out);
+} // namespace native
+} // namespace at

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/cudnn_is_acceptable.h

@@ -0,0 +1,30 @@
+#pragma once
+
+// @generated by torchgen/gen.py from Function.h
+
+#include <ATen/Context.h>
+#include <ATen/DeviceGuard.h>
+#include <ATen/TensorUtils.h>
+#include <ATen/TracerMode.h>
+#include <ATen/core/Generator.h>
+#include <ATen/core/Reduction.h>
+#include <ATen/core/Tensor.h>
+#include <c10/core/Scalar.h>
+#include <c10/core/Storage.h>
+#include <c10/core/TensorOptions.h>
+#include <c10/util/Deprecated.h>
+#include <c10/util/Optional.h>
+
+
+
+#include <ATen/ops/cudnn_is_acceptable_ops.h>
+
+namespace at {
+
+
+// aten::cudnn_is_acceptable(Tensor self) -> bool
+inline bool cudnn_is_acceptable(const at::Tensor & self) {
+    return at::_ops::cudnn_is_acceptable::call(self);
+}
+
+}

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/dense_dim_cpu_dispatch.h

@@ -0,0 +1,23 @@
+#pragma once
+// @generated by torchgen/gen.py from DispatchKeyFunction.h
+
+// NB: The implementing C++ file is RegisterDispatchKey.cpp
+
+// The only #includes we need are for custom classes that have defaults in the C++ API
+#include <c10/core/MemoryFormat.h>
+#include <c10/core/Scalar.h>
+#include <ATen/core/Reduction.h>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+
+namespace cpu {
+
+TORCH_API int64_t dense_dim(const at::Tensor & self);
+
+} // namespace cpu
+} // namespace at

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/div_meta.h

@@ -0,0 +1,32 @@
+#pragma once
+
+// @generated by torchgen/gen.py from NativeMetaFunction.h
+
+#include <c10/core/Scalar.h>
+#include <c10/core/Storage.h>
+#include <c10/core/TensorOptions.h>
+#include <c10/util/Deprecated.h>
+#include <c10/util/Optional.h>
+#include <c10/core/QScheme.h>
+#include <ATen/core/Reduction.h>
+#include <ATen/TensorIterator.h>
+#include <ATen/TensorMeta.h>
+#include <tuple>
+#include <vector>
+
+namespace at {
+namespace meta {
+
+struct TORCH_API structured_div_Tensor : public TensorIteratorBase {
+    
+    
+    void meta(const at::Tensor & self, const at::Tensor & other);
+};
+struct TORCH_API structured_div_Tensor_mode : public TensorIteratorBase {
+    
+    
+    void meta(const at::Tensor & self, const at::Tensor & other, c10::optional<c10::string_view> rounding_mode);
+};
+
+} // namespace native
+} // namespace at

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/dstack_native.h

@@ -0,0 +1,22 @@
+#pragma once
+
+// @generated by torchgen/gen.py from NativeFunction.h
+
+#include <c10/core/Scalar.h>
+#include <c10/core/Storage.h>
+#include <c10/core/TensorOptions.h>
+#include <c10/util/Deprecated.h>
+#include <c10/util/Optional.h>
+#include <c10/core/QScheme.h>
+#include <ATen/core/Reduction.h>
+#include <ATen/core/Tensor.h>
+#include <tuple>
+#include <vector>
+
+
+namespace at {
+namespace native {
+TORCH_API at::Tensor dstack(at::TensorList tensors);
+TORCH_API at::Tensor & dstack_out(at::TensorList tensors, at::Tensor & out);
+} // namespace native
+} // namespace at

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/embedding_renorm.h

@@ -0,0 +1,44 @@
+#pragma once
+
+// @generated by torchgen/gen.py from Function.h
+
+#include <ATen/Context.h>
+#include <ATen/DeviceGuard.h>
+#include <ATen/TensorUtils.h>
+#include <ATen/TracerMode.h>
+#include <ATen/core/Generator.h>
+#include <ATen/core/Reduction.h>
+#include <ATen/core/Tensor.h>
+#include <c10/core/Scalar.h>
+#include <c10/core/Storage.h>
+#include <c10/core/TensorOptions.h>
+#include <c10/util/Deprecated.h>
+#include <c10/util/Optional.h>
+
+
+
+#include <ATen/ops/embedding_renorm_ops.h>
+
+namespace at {
+
+
+// aten::embedding_renorm_(Tensor(a!) self, Tensor indices, float max_norm, float norm_type) -> Tensor(a!)
+inline at::Tensor & embedding_renorm_(at::Tensor & self, const at::Tensor & indices, double max_norm, double norm_type) {
+    return at::_ops::embedding_renorm_::call(self, indices, max_norm, norm_type);
+}
+
+// aten::embedding_renorm.out(Tensor self, Tensor indices, float max_norm, float norm_type, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & embedding_renorm_out(at::Tensor & out, const at::Tensor & self, const at::Tensor & indices, double max_norm, double norm_type) {
+    return at::_ops::embedding_renorm_out::call(self, indices, max_norm, norm_type, out);
+}
+// aten::embedding_renorm.out(Tensor self, Tensor indices, float max_norm, float norm_type, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & embedding_renorm_outf(const at::Tensor & self, const at::Tensor & indices, double max_norm, double norm_type, at::Tensor & out) {
+    return at::_ops::embedding_renorm_out::call(self, indices, max_norm, norm_type, out);
+}
+
+// aten::embedding_renorm(Tensor self, Tensor indices, float max_norm, float norm_type) -> Tensor
+inline at::Tensor embedding_renorm(const at::Tensor & self, const at::Tensor & indices, double max_norm, double norm_type) {
+    return at::_ops::embedding_renorm::call(self, indices, max_norm, norm_type);
+}
+
+}

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/exp2_meta_dispatch.h

@@ -0,0 +1,26 @@
+#pragma once
+// @generated by torchgen/gen.py from DispatchKeyFunction.h
+
+// NB: The implementing C++ file is RegisterDispatchKey.cpp
+
+// The only #includes we need are for custom classes that have defaults in the C++ API
+#include <c10/core/MemoryFormat.h>
+#include <c10/core/Scalar.h>
+#include <ATen/core/Reduction.h>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+
+namespace meta {
+
+TORCH_API at::Tensor exp2(const at::Tensor & self);
+TORCH_API at::Tensor & exp2_out(at::Tensor & out, const at::Tensor & self);
+TORCH_API at::Tensor & exp2_outf(const at::Tensor & self, at::Tensor & out);
+TORCH_API at::Tensor & exp2_(at::Tensor & self);
+
+} // namespace meta
+} // namespace at

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/gcd_native.h

@@ -0,0 +1,23 @@
+#pragma once
+
+// @generated by torchgen/gen.py from NativeFunction.h
+
+#include <c10/core/Scalar.h>
+#include <c10/core/Storage.h>
+#include <c10/core/TensorOptions.h>
+#include <c10/util/Deprecated.h>
+#include <c10/util/Optional.h>
+#include <c10/core/QScheme.h>
+#include <ATen/core/Reduction.h>
+#include <ATen/core/Tensor.h>
+#include <tuple>
+#include <vector>
+#include <ATen/ops/gcd_meta.h>
+
+namespace at {
+namespace native {
+struct TORCH_API structured_gcd_out : public at::meta::structured_gcd {
+void impl(const at::Tensor & self, const at::Tensor & other, const at::Tensor & out);
+};
+} // namespace native
+} // namespace at

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/hamming_window_native.h

@@ -0,0 +1,28 @@
+#pragma once
+
+// @generated by torchgen/gen.py from NativeFunction.h
+
+#include <c10/core/Scalar.h>
+#include <c10/core/Storage.h>
+#include <c10/core/TensorOptions.h>
+#include <c10/util/Deprecated.h>
+#include <c10/util/Optional.h>
+#include <c10/core/QScheme.h>
+#include <ATen/core/Reduction.h>
+#include <ATen/core/Tensor.h>
+#include <tuple>
+#include <vector>
+
+
+namespace at {
+namespace native {
+TORCH_API at::Tensor hamming_window(int64_t window_length, c10::optional<at::ScalarType> dtype={}, c10::optional<at::Layout> layout={}, c10::optional<at::Device> device={}, c10::optional<bool> pin_memory={});
+TORCH_API at::Tensor & hamming_window_out(int64_t window_length, at::Tensor & out);
+TORCH_API at::Tensor hamming_window(int64_t window_length, bool periodic, c10::optional<at::ScalarType> dtype={}, c10::optional<at::Layout> layout={}, c10::optional<at::Device> device={}, c10::optional<bool> pin_memory={});
+TORCH_API at::Tensor & hamming_window_periodic_out(int64_t window_length, bool periodic, at::Tensor & out);
+TORCH_API at::Tensor hamming_window(int64_t window_length, bool periodic, double alpha, c10::optional<at::ScalarType> dtype={}, c10::optional<at::Layout> layout={}, c10::optional<at::Device> device={}, c10::optional<bool> pin_memory={});
+TORCH_API at::Tensor & hamming_window_periodic_alpha_out(int64_t window_length, bool periodic, double alpha, at::Tensor & out);
+TORCH_API at::Tensor hamming_window(int64_t window_length, bool periodic, double alpha, double beta, c10::optional<at::ScalarType> dtype={}, c10::optional<at::Layout> layout={}, c10::optional<at::Device> device={}, c10::optional<bool> pin_memory={});
+TORCH_API at::Tensor & hamming_window_periodic_alpha_beta_out(int64_t window_length, bool periodic, double alpha, double beta, at::Tensor & out);
+} // namespace native
+} // namespace at

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/histc.h

@@ -0,0 +1,39 @@
+#pragma once
+
+// @generated by torchgen/gen.py from Function.h
+
+#include <ATen/Context.h>
+#include <ATen/DeviceGuard.h>
+#include <ATen/TensorUtils.h>
+#include <ATen/TracerMode.h>
+#include <ATen/core/Generator.h>
+#include <ATen/core/Reduction.h>
+#include <ATen/core/Tensor.h>
+#include <c10/core/Scalar.h>
+#include <c10/core/Storage.h>
+#include <c10/core/TensorOptions.h>
+#include <c10/util/Deprecated.h>
+#include <c10/util/Optional.h>
+
+
+
+#include <ATen/ops/histc_ops.h>
+
+namespace at {
+
+
+// aten::histc.out(Tensor self, int bins=100, Scalar min=0, Scalar max=0, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & histc_out(at::Tensor & out, const at::Tensor & self, int64_t bins=100, const at::Scalar & min=0, const at::Scalar & max=0) {
+    return at::_ops::histc_out::call(self, bins, min, max, out);
+}
+// aten::histc.out(Tensor self, int bins=100, Scalar min=0, Scalar max=0, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & histc_outf(const at::Tensor & self, int64_t bins, const at::Scalar & min, const at::Scalar & max, at::Tensor & out) {
+    return at::_ops::histc_out::call(self, bins, min, max, out);
+}
+
+// aten::histc(Tensor self, int bins=100, Scalar min=0, Scalar max=0) -> Tensor
+inline at::Tensor histc(const at::Tensor & self, int64_t bins=100, const at::Scalar & min=0, const at::Scalar & max=0) {
+    return at::_ops::histc::call(self, bins, min, max);
+}
+
+}

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/im2col_cpu_dispatch.h

@@ -0,0 +1,25 @@
+#pragma once
+// @generated by torchgen/gen.py from DispatchKeyFunction.h
+
+// NB: The implementing C++ file is RegisterDispatchKey.cpp
+
+// The only #includes we need are for custom classes that have defaults in the C++ API
+#include <c10/core/MemoryFormat.h>
+#include <c10/core/Scalar.h>
+#include <ATen/core/Reduction.h>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+
+namespace cpu {
+
+TORCH_API at::Tensor im2col(const at::Tensor & self, at::IntArrayRef kernel_size, at::IntArrayRef dilation, at::IntArrayRef padding, at::IntArrayRef stride);
+TORCH_API at::Tensor & im2col_out(at::Tensor & out, const at::Tensor & self, at::IntArrayRef kernel_size, at::IntArrayRef dilation, at::IntArrayRef padding, at::IntArrayRef stride);
+TORCH_API at::Tensor & im2col_outf(const at::Tensor & self, at::IntArrayRef kernel_size, at::IntArrayRef dilation, at::IntArrayRef padding, at::IntArrayRef stride, at::Tensor & out);
+
+} // namespace cpu
+} // namespace at

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/index_native.h

@@ -0,0 +1,24 @@
+#pragma once
+
+// @generated by torchgen/gen.py from NativeFunction.h
+
+#include <c10/core/Scalar.h>
+#include <c10/core/Storage.h>
+#include <c10/core/TensorOptions.h>
+#include <c10/util/Deprecated.h>
+#include <c10/util/Optional.h>
+#include <c10/core/QScheme.h>
+#include <ATen/core/Reduction.h>
+#include <ATen/core/Tensor.h>
+#include <tuple>
+#include <vector>
+#include <ATen/ops/index_meta.h>
+
+namespace at {
+namespace native {
+struct TORCH_API structured_index_out : public at::meta::structured_index_Tensor {
+void impl(const at::Tensor & self, at::DimVector sizes, at::DimVector strides, const at::Tensor & out);
+};
+TORCH_API at::Tensor quantized_index(const at::Tensor & self, const c10::List<c10::optional<at::Tensor>> & indices);
+} // namespace native
+} // namespace at

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/index_select_cuda_dispatch.h

@@ -0,0 +1,25 @@
+#pragma once
+// @generated by torchgen/gen.py from DispatchKeyFunction.h
+
+// NB: The implementing C++ file is RegisterDispatchKey.cpp
+
+// The only #includes we need are for custom classes that have defaults in the C++ API
+#include <c10/core/MemoryFormat.h>
+#include <c10/core/Scalar.h>
+#include <ATen/core/Reduction.h>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+
+namespace cuda {
+
+TORCH_API at::Tensor index_select(const at::Tensor & self, int64_t dim, const at::Tensor & index);
+TORCH_API at::Tensor & index_select_out(at::Tensor & out, const at::Tensor & self, int64_t dim, const at::Tensor & index);
+TORCH_API at::Tensor & index_select_outf(const at::Tensor & self, int64_t dim, const at::Tensor & index, at::Tensor & out);
+
+} // namespace cuda
+} // namespace at

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/indices_native.h

@@ -0,0 +1,22 @@
+#pragma once
+
+// @generated by torchgen/gen.py from NativeFunction.h
+
+#include <c10/core/Scalar.h>
+#include <c10/core/Storage.h>
+#include <c10/core/TensorOptions.h>
+#include <c10/util/Deprecated.h>
+#include <c10/util/Optional.h>
+#include <c10/core/QScheme.h>
+#include <ATen/core/Reduction.h>
+#include <ATen/core/Tensor.h>
+#include <tuple>
+#include <vector>
+
+
+namespace at {
+namespace native {
+TORCH_API at::Tensor indices_default(const at::Tensor & self);
+TORCH_API at::Tensor indices_sparse(const at::Tensor & self);
+} // namespace native
+} // namespace at

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/int_repr_compositeexplicitautograd_dispatch.h

@@ -0,0 +1,24 @@
+#pragma once
+// @generated by torchgen/gen.py from DispatchKeyFunction.h
+
+// NB: The implementing C++ file is RegisterDispatchKey.cpp
+
+// The only #includes we need are for custom classes that have defaults in the C++ API
+#include <c10/core/MemoryFormat.h>
+#include <c10/core/Scalar.h>
+#include <ATen/core/Reduction.h>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+
+namespace compositeexplicitautograd {
+
+TORCH_API at::Tensor & int_repr_out(at::Tensor & out, const at::Tensor & self);
+TORCH_API at::Tensor & int_repr_outf(const at::Tensor & self, at::Tensor & out);
+
+} // namespace compositeexplicitautograd
+} // namespace at

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/is_complex.h

@@ -0,0 +1,30 @@
+#pragma once
+
+// @generated by torchgen/gen.py from Function.h
+
+#include <ATen/Context.h>
+#include <ATen/DeviceGuard.h>
+#include <ATen/TensorUtils.h>
+#include <ATen/TracerMode.h>
+#include <ATen/core/Generator.h>
+#include <ATen/core/Reduction.h>
+#include <ATen/core/Tensor.h>
+#include <c10/core/Scalar.h>
+#include <c10/core/Storage.h>
+#include <c10/core/TensorOptions.h>
+#include <c10/util/Deprecated.h>
+#include <c10/util/Optional.h>
+
+
+
+#include <ATen/ops/is_complex_ops.h>
+
+namespace at {
+
+
+// aten::is_complex(Tensor self) -> bool
+inline bool __dispatch_is_complex(const at::Tensor & self) {
+    return at::_ops::is_complex::call(self);
+}
+
+}

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/istft_compositeimplicitautograd_dispatch.h

@@ -0,0 +1,23 @@
+#pragma once
+// @generated by torchgen/gen.py from DispatchKeyFunction.h
+
+// NB: The implementing C++ file is RegisterDispatchKey.cpp
+
+// The only #includes we need are for custom classes that have defaults in the C++ API
+#include <c10/core/MemoryFormat.h>
+#include <c10/core/Scalar.h>
+#include <ATen/core/Reduction.h>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+
+namespace compositeimplicitautograd {
+
+TORCH_API at::Tensor istft(const at::Tensor & self, int64_t n_fft, c10::optional<int64_t> hop_length=c10::nullopt, c10::optional<int64_t> win_length=c10::nullopt, const c10::optional<at::Tensor> & window={}, bool center=true, bool normalized=false, c10::optional<bool> onesided=c10::nullopt, c10::optional<int64_t> length=c10::nullopt, bool return_complex=false);
+
+} // namespace compositeimplicitautograd
+} // namespace at

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/linalg_householder_product_cuda_dispatch.h

@@ -0,0 +1,25 @@
+#pragma once
+// @generated by torchgen/gen.py from DispatchKeyFunction.h
+
+// NB: The implementing C++ file is RegisterDispatchKey.cpp
+
+// The only #includes we need are for custom classes that have defaults in the C++ API
+#include <c10/core/MemoryFormat.h>
+#include <c10/core/Scalar.h>
+#include <ATen/core/Reduction.h>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+
+namespace cuda {
+
+TORCH_API at::Tensor linalg_householder_product(const at::Tensor & input, const at::Tensor & tau);
+TORCH_API at::Tensor & linalg_householder_product_out(at::Tensor & out, const at::Tensor & input, const at::Tensor & tau);
+TORCH_API at::Tensor & linalg_householder_product_outf(const at::Tensor & input, const at::Tensor & tau, at::Tensor & out);
+
+} // namespace cuda
+} // namespace at

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/log2_meta.h

@@ -0,0 +1,27 @@
+#pragma once
+
+// @generated by torchgen/gen.py from NativeMetaFunction.h
+
+#include <c10/core/Scalar.h>
+#include <c10/core/Storage.h>
+#include <c10/core/TensorOptions.h>
+#include <c10/util/Deprecated.h>
+#include <c10/util/Optional.h>
+#include <c10/core/QScheme.h>
+#include <ATen/core/Reduction.h>
+#include <ATen/TensorIterator.h>
+#include <ATen/TensorMeta.h>
+#include <tuple>
+#include <vector>
+
+namespace at {
+namespace meta {
+
+struct TORCH_API structured_log2 : public TensorIteratorBase {
+    
+    
+    void meta(const at::Tensor & self);
+};
+
+} // namespace native
+} // namespace at

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/masked_scatter_backward.h

@@ -0,0 +1,47 @@
+#pragma once
+
+// @generated by torchgen/gen.py from Function.h
+
+#include <ATen/Context.h>
+#include <ATen/DeviceGuard.h>
+#include <ATen/TensorUtils.h>
+#include <ATen/TracerMode.h>
+#include <ATen/core/Generator.h>
+#include <ATen/core/Reduction.h>
+#include <ATen/core/Tensor.h>
+#include <c10/core/Scalar.h>
+#include <c10/core/Storage.h>
+#include <c10/core/TensorOptions.h>
+#include <c10/util/Deprecated.h>
+#include <c10/util/Optional.h>
+
+
+
+#include <ATen/ops/masked_scatter_backward_ops.h>
+
+namespace at {
+
+
+// aten::masked_scatter_backward(Tensor grad_output, Tensor mask, SymInt[] sizes) -> Tensor
+inline at::Tensor masked_scatter_backward(const at::Tensor & grad_output, const at::Tensor & mask, at::IntArrayRef sizes) {
+    return at::_ops::masked_scatter_backward::call(grad_output, mask, c10::fromIntArrayRefSlow(sizes));
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same<T, int64_t>::value>>
+  at::Tensor masked_scatter_backward(const at::Tensor & grad_output, const at::Tensor & mask, at::IntArrayRef sizes) {
+    return at::_ops::masked_scatter_backward::call(grad_output, mask, c10::fromIntArrayRefSlow(sizes));
+  }
+}
+
+// aten::masked_scatter_backward(Tensor grad_output, Tensor mask, SymInt[] sizes) -> Tensor
+inline at::Tensor masked_scatter_backward_symint(const at::Tensor & grad_output, const at::Tensor & mask, c10::SymIntArrayRef sizes) {
+    return at::_ops::masked_scatter_backward::call(grad_output, mask, sizes);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same<T, c10::SymInt>::value>>
+  at::Tensor masked_scatter_backward(const at::Tensor & grad_output, const at::Tensor & mask, c10::SymIntArrayRef sizes) {
+    return at::_ops::masked_scatter_backward::call(grad_output, mask, sizes);
+  }
+}
+
+}

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/max_pool2d.h

@@ -0,0 +1,30 @@
+#pragma once
+
+// @generated by torchgen/gen.py from Function.h
+
+#include <ATen/Context.h>
+#include <ATen/DeviceGuard.h>
+#include <ATen/TensorUtils.h>
+#include <ATen/TracerMode.h>
+#include <ATen/core/Generator.h>
+#include <ATen/core/Reduction.h>
+#include <ATen/core/Tensor.h>
+#include <c10/core/Scalar.h>
+#include <c10/core/Storage.h>
+#include <c10/core/TensorOptions.h>
+#include <c10/util/Deprecated.h>
+#include <c10/util/Optional.h>
+
+
+
+#include <ATen/ops/max_pool2d_ops.h>
+
+namespace at {
+
+
+// aten::max_pool2d(Tensor self, int[2] kernel_size, int[2] stride=[], int[2] padding=0, int[2] dilation=1, bool ceil_mode=False) -> Tensor
+inline at::Tensor max_pool2d(const at::Tensor & self, at::IntArrayRef kernel_size, at::IntArrayRef stride={}, at::IntArrayRef padding=0, at::IntArrayRef dilation=1, bool ceil_mode=false) {
+    return at::_ops::max_pool2d::call(self, kernel_size, stride, padding, dilation, ceil_mode);
+}
+
+}

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/median_compositeexplicitautograd_dispatch.h

@@ -0,0 +1,25 @@
+#pragma once
+// @generated by torchgen/gen.py from DispatchKeyFunction.h
+
+// NB: The implementing C++ file is RegisterDispatchKey.cpp
+
+// The only #includes we need are for custom classes that have defaults in the C++ API
+#include <c10/core/MemoryFormat.h>
+#include <c10/core/Scalar.h>
+#include <ATen/core/Reduction.h>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+
+namespace compositeexplicitautograd {
+
+TORCH_API at::Tensor & median_out(at::Tensor & out, const at::Tensor & self);
+TORCH_API at::Tensor & median_outf(const at::Tensor & self, at::Tensor & out);
+TORCH_API ::std::tuple<at::Tensor,at::Tensor> median(const at::Tensor & self, int64_t dim, bool keepdim=false);
+
+} // namespace compositeexplicitautograd
+} // namespace at

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/miopen_rnn_cuda_dispatch.h

@@ -0,0 +1,23 @@
+#pragma once
+// @generated by torchgen/gen.py from DispatchKeyFunction.h
+
+// NB: The implementing C++ file is RegisterDispatchKey.cpp
+
+// The only #includes we need are for custom classes that have defaults in the C++ API
+#include <c10/core/MemoryFormat.h>
+#include <c10/core/Scalar.h>
+#include <ATen/core/Reduction.h>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+
+namespace cuda {
+
+TORCH_API ::std::tuple<at::Tensor,at::Tensor,at::Tensor,at::Tensor,at::Tensor> miopen_rnn(const at::Tensor & input, at::TensorList weight, int64_t weight_stride0, const at::Tensor & hx, const c10::optional<at::Tensor> & cx, int64_t mode, int64_t hidden_size, int64_t num_layers, bool batch_first, double dropout, bool train, bool bidirectional, at::IntArrayRef batch_sizes, const c10::optional<at::Tensor> & dropout_state);
+
+} // namespace cuda
+} // namespace at

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/mkldnn_linear_backward_ops.h

@@ -0,0 +1,39 @@
+#pragma once
+
+// @generated by torchgen/gen.py from Operator.h
+
+#include <tuple>
+#include <vector>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+namespace _ops {
+
+
+struct TORCH_API mkldnn_linear_backward {
+  using schema = ::std::tuple<at::Tensor,at::Tensor,at::Tensor> (const at::Tensor &, const at::Tensor &, const at::Tensor &, ::std::array<bool,3>);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(name, "aten::mkldnn_linear_backward")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "mkldnn_linear_backward(Tensor self, Tensor grad_output, Tensor weight, bool[3] output_mask) -> (Tensor, Tensor, Tensor)")
+  static ::std::tuple<at::Tensor,at::Tensor,at::Tensor> call(const at::Tensor & self, const at::Tensor & grad_output, const at::Tensor & weight, ::std::array<bool,3> output_mask);
+  static ::std::tuple<at::Tensor,at::Tensor,at::Tensor> redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, const at::Tensor & grad_output, const at::Tensor & weight, ::std::array<bool,3> output_mask);
+};
+
+struct TORCH_API mkldnn_linear_backward_out {
+  using schema = ::std::tuple<at::Tensor &,at::Tensor &,at::Tensor &> (const at::Tensor &, const at::Tensor &, const at::Tensor &, ::std::array<bool,3>, at::Tensor &, at::Tensor &, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(name, "aten::mkldnn_linear_backward")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "out")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "mkldnn_linear_backward.out(Tensor self, Tensor grad_output, Tensor weight, bool[3] output_mask, *, Tensor(a!) out0, Tensor(b!) out1, Tensor(c!) out2) -> (Tensor(a!), Tensor(b!), Tensor(c!))")
+  static ::std::tuple<at::Tensor &,at::Tensor &,at::Tensor &> call(const at::Tensor & self, const at::Tensor & grad_output, const at::Tensor & weight, ::std::array<bool,3> output_mask, at::Tensor & out0, at::Tensor & out1, at::Tensor & out2);
+  static ::std::tuple<at::Tensor &,at::Tensor &,at::Tensor &> redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, const at::Tensor & grad_output, const at::Tensor & weight, ::std::array<bool,3> output_mask, at::Tensor & out0, at::Tensor & out1, at::Tensor & out2);
+};
+
+}} // namespace at::_ops

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/mvlgamma_native.h

@@ -0,0 +1,23 @@
+#pragma once
+
+// @generated by torchgen/gen.py from NativeFunction.h
+
+#include <c10/core/Scalar.h>
+#include <c10/core/Storage.h>
+#include <c10/core/TensorOptions.h>
+#include <c10/util/Deprecated.h>
+#include <c10/util/Optional.h>
+#include <c10/core/QScheme.h>
+#include <ATen/core/Reduction.h>
+#include <ATen/core/Tensor.h>
+#include <tuple>
+#include <vector>
+
+
+namespace at {
+namespace native {
+TORCH_API at::Tensor mvlgamma(const at::Tensor & self, int64_t p);
+TORCH_API at::Tensor & mvlgamma_(at::Tensor & self, int64_t p);
+TORCH_API at::Tensor & mvlgamma_out(const at::Tensor & self, int64_t p, at::Tensor & out);
+} // namespace native
+} // namespace at