Add files using upload-large-folder tool

.gitattributes

@@ -82,3 +82,4 @@ tuning-competition-baseline/.venv/lib/python3.11/site-packages/nvidia/cuda_runti
 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/nn/__pycache__/functional.cpython-311.pyc filter=lfs diff=lfs merge=lfs -text

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_cdist_backward.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/_cdist_backward_ops.h>
+
+namespace at {
+
+
+// aten::_cdist_backward(Tensor grad, Tensor x1, Tensor x2, float p, Tensor cdist) -> Tensor
+inline at::Tensor _cdist_backward(const at::Tensor & grad, const at::Tensor & x1, const at::Tensor & x2, double p, const at::Tensor & cdist) {
+    return at::_ops::_cdist_backward::call(grad, x1, x2, p, cdist);
+}
+
+// aten::_cdist_backward.out(Tensor grad, Tensor x1, Tensor x2, float p, Tensor cdist, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & _cdist_backward_out(at::Tensor & out, const at::Tensor & grad, const at::Tensor & x1, const at::Tensor & x2, double p, const at::Tensor & cdist) {
+    return at::_ops::_cdist_backward_out::call(grad, x1, x2, p, cdist, out);
+}
+// aten::_cdist_backward.out(Tensor grad, Tensor x1, Tensor x2, float p, Tensor cdist, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & _cdist_backward_outf(const at::Tensor & grad, const at::Tensor & x1, const at::Tensor & x2, double p, const at::Tensor & cdist, at::Tensor & out) {
+    return at::_ops::_cdist_backward_out::call(grad, x1, x2, p, cdist, out);
+}
+
+}

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_efficient_attention_forward_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,c10::SymInt,c10::SymInt> _efficient_attention_forward(const at::Tensor & query, const at::Tensor & key, const at::Tensor & value, const c10::optional<at::Tensor> & bias, const c10::optional<at::Tensor> & cu_seqlens_q, const c10::optional<at::Tensor> & cu_seqlens_k, c10::optional<int64_t> max_seqlen_q, c10::optional<int64_t> max_seqlen_k, double dropout_p, int64_t custom_mask_type, bool compute_log_sumexp=false, c10::optional<double> scale=c10::nullopt, const c10::optional<at::Tensor> & causal_diagonal={}, const c10::optional<at::Tensor> & seqlen_k={});
+
+} // namespace cuda
+} // namespace at

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

@@ -0,0 +1,30 @@
+#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::vector<at::Tensor> _foreach_mul(at::TensorList self, const at::Scalar & scalar);
+TORCH_API void _foreach_mul_(at::TensorList self, const at::Scalar & scalar);
+TORCH_API ::std::vector<at::Tensor> _foreach_mul(at::TensorList self, at::TensorList other);
+TORCH_API void _foreach_mul_(at::TensorList self, at::TensorList other);
+TORCH_API ::std::vector<at::Tensor> _foreach_mul(at::TensorList self, at::ArrayRef<at::Scalar> scalars);
+TORCH_API void _foreach_mul_(at::TensorList self, at::ArrayRef<at::Scalar> scalars);
+TORCH_API ::std::vector<at::Tensor> _foreach_mul(at::TensorList self, const at::Tensor & other);
+TORCH_API void _foreach_mul_(at::TensorList self, const at::Tensor & other);
+
+} // namespace cpu
+} // namespace at

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_has_compatible_shallow_copy_type_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 bool _has_compatible_shallow_copy_type(const at::Tensor & self, const at::Tensor & from);
+
+} // namespace compositeimplicitautograd
+} // namespace at

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_linalg_check_errors_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 _linalg_check_errors {
+  using schema = void (const at::Tensor &, c10::string_view, bool);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(name, "aten::_linalg_check_errors")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "_linalg_check_errors(Tensor info, str api_name, *, bool is_matrix) -> ()")
+  static void call(const at::Tensor & info, c10::string_view api_name, bool is_matrix);
+  static void redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & info, c10::string_view api_name, bool is_matrix);
+};
+
+}} // namespace at::_ops

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_linalg_det_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__linalg_det : public at::impl::MetaBase {
+    
+    
+    void meta(const at::Tensor & A);
+};
+
+} // namespace native
+} // namespace at

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_linalg_eigh_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/_linalg_eigh_meta.h>
+
+namespace at {
+namespace native {
+struct TORCH_API structured__linalg_eigh_out : public at::meta::structured__linalg_eigh {
+void impl(const at::Tensor & A, c10::string_view UPLO, bool compute_v, const at::Tensor & eigenvalues, const at::Tensor & eigenvectors);
+};
+} // namespace native
+} // namespace at

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_make_dual_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 _make_dual {
+  using schema = at::Tensor (const at::Tensor &, const at::Tensor &, int64_t);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(name, "aten::_make_dual")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "_make_dual(Tensor(a) primal, Tensor tangent, int level) -> Tensor(a)")
+  static at::Tensor call(const at::Tensor & primal, const at::Tensor & tangent, int64_t level);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & primal, const at::Tensor & tangent, int64_t level);
+};
+
+}} // namespace at::_ops

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_native_batch_norm_legit_no_training_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 ::std::tuple<at::Tensor,at::Tensor,at::Tensor> _native_batch_norm_legit_no_training(const at::Tensor & input, const c10::optional<at::Tensor> & weight, const c10::optional<at::Tensor> & bias, const at::Tensor & running_mean, const at::Tensor & running_var, double momentum, double eps);
+TORCH_API ::std::tuple<at::Tensor &,at::Tensor &,at::Tensor &> _native_batch_norm_legit_no_training_out(at::Tensor & out0, at::Tensor & out1, at::Tensor & out2, const at::Tensor & input, const c10::optional<at::Tensor> & weight, const c10::optional<at::Tensor> & bias, const at::Tensor & running_mean, const at::Tensor & running_var, double momentum, double eps);
+TORCH_API ::std::tuple<at::Tensor &,at::Tensor &,at::Tensor &> _native_batch_norm_legit_no_training_outf(const at::Tensor & input, const c10::optional<at::Tensor> & weight, const c10::optional<at::Tensor> & bias, const at::Tensor & running_mean, const at::Tensor & running_var, double momentum, double eps, at::Tensor & out0, at::Tensor & out1, at::Tensor & out2);
+
+} // namespace compositeexplicitautograd
+} // namespace at

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_nested_view_from_buffer.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/_nested_view_from_buffer_ops.h>
+
+namespace at {
+
+
+// aten::_nested_view_from_buffer(Tensor(a) self, Tensor nested_size, Tensor nested_strides, Tensor offsets) -> Tensor(a)
+inline at::Tensor _nested_view_from_buffer(const at::Tensor & self, const at::Tensor & nested_size, const at::Tensor & nested_strides, const at::Tensor & offsets) {
+    return at::_ops::_nested_view_from_buffer::call(self, nested_size, nested_strides, offsets);
+}
+
+}

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_remove_batch_dim_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 _remove_batch_dim(const at::Tensor & self, int64_t level, int64_t batch_size, int64_t out_dim);
+
+} // namespace compositeimplicitautograd
+} // namespace at

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_softmax_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 _softmax {
+  using schema = at::Tensor (const at::Tensor &, int64_t, bool);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(name, "aten::_softmax")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "_softmax(Tensor self, int dim, bool half_to_float) -> Tensor")
+  static at::Tensor call(const at::Tensor & self, int64_t dim, bool half_to_float);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, int64_t dim, bool half_to_float);
+};
+
+struct TORCH_API _softmax_out {
+  using schema = at::Tensor & (const at::Tensor &, int64_t, bool, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(name, "aten::_softmax")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "out")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "_softmax.out(Tensor self, int dim, bool half_to_float, *, Tensor(a!) out) -> Tensor(a!)")
+  static at::Tensor & call(const at::Tensor & self, int64_t dim, bool half_to_float, at::Tensor & out);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, int64_t dim, bool half_to_float, at::Tensor & out);
+};
+
+}} // namespace at::_ops

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_sparse_csr_prod_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 _sparse_csr_prod_dim_dtype {
+  using schema = at::Tensor (const at::Tensor &, at::IntArrayRef, bool, c10::optional<at::ScalarType>);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(name, "aten::_sparse_csr_prod")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "dim_dtype")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "_sparse_csr_prod.dim_dtype(Tensor self, int[1] dim, bool keepdim=False, *, ScalarType? dtype=None) -> Tensor")
+  static at::Tensor call(const at::Tensor & self, at::IntArrayRef dim, bool keepdim, c10::optional<at::ScalarType> dtype);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, at::IntArrayRef dim, bool keepdim, c10::optional<at::ScalarType> dtype);
+};
+
+struct TORCH_API _sparse_csr_prod_dim_dtype_out {
+  using schema = at::Tensor & (const at::Tensor &, at::IntArrayRef, bool, c10::optional<at::ScalarType>, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(name, "aten::_sparse_csr_prod")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "dim_dtype_out")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "_sparse_csr_prod.dim_dtype_out(Tensor self, int[1] dim, bool keepdim=False, *, ScalarType? dtype=None, Tensor(a!) out) -> Tensor(a!)")
+  static at::Tensor & call(const at::Tensor & self, at::IntArrayRef dim, bool keepdim, c10::optional<at::ScalarType> dtype, at::Tensor & out);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, at::IntArrayRef dim, bool keepdim, c10::optional<at::ScalarType> dtype, at::Tensor & out);
+};
+
+}} // namespace at::_ops

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_test_autograd_multiple_dispatch_view_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 _test_autograd_multiple_dispatch_view(const at::Tensor & self);
+} // namespace native
+} // namespace at

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_test_functorch_fallback_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 _test_functorch_fallback {
+  using schema = at::Tensor (const 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::_test_functorch_fallback")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "_test_functorch_fallback(Tensor self, Tensor other) -> Tensor")
+  static at::Tensor call(const at::Tensor & self, const at::Tensor & other);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, const at::Tensor & other);
+};
+
+struct TORCH_API _test_functorch_fallback_out {
+  using schema = at::Tensor & (const at::Tensor &, const 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::_test_functorch_fallback")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "out")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "_test_functorch_fallback.out(Tensor self, Tensor other, *, Tensor(a!) out) -> Tensor(a!)")
+  static at::Tensor & call(const at::Tensor & self, const at::Tensor & other, at::Tensor & out);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, const at::Tensor & other, at::Tensor & out);
+};
+
+}} // namespace at::_ops

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_to_sparse_bsr_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>
+
+
+namespace at {
+namespace native {
+TORCH_API at::Tensor & _to_sparse_bsr_out(const at::Tensor & self, at::IntArrayRef blocksize, c10::optional<int64_t> dense_dim, at::Tensor & out);
+TORCH_API at::Tensor dense_to_sparse_bsr(const at::Tensor & self, at::IntArrayRef blocksize, c10::optional<int64_t> dense_dim=c10::nullopt);
+TORCH_API at::Tensor coo_to_sparse_bsr(const at::Tensor & self, at::IntArrayRef blocksize, c10::optional<int64_t> dense_dim=c10::nullopt);
+TORCH_API at::Tensor sparse_compressed_to_sparse_bsr(const at::Tensor & self, at::IntArrayRef blocksize, c10::optional<int64_t> dense_dim=c10::nullopt);
+} // namespace native
+} // namespace at

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_upsample_nearest_exact1d_backward_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__upsample_nearest_exact1d_backward : public at::impl::MetaBase {
+    
+    
+    void meta(const at::Tensor & grad_output, at::ArrayRef<int64_t> output_size, at::ArrayRef<int64_t> input_size, c10::optional<double> scales);
+};
+
+} // namespace native
+} // namespace at

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_upsample_nearest_exact1d_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 _upsample_nearest_exact1d_backward_grad_input {
+  using schema = at::Tensor & (const at::Tensor &, c10::SymIntArrayRef, 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_backward")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "grad_input")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "_upsample_nearest_exact1d_backward.grad_input(Tensor grad_output, SymInt[1] output_size, SymInt[3] input_size, float? scales=None, *, Tensor(a!) grad_input) -> Tensor(a!)")
+  static at::Tensor & call(const at::Tensor & grad_output, c10::SymIntArrayRef output_size, c10::SymIntArrayRef input_size, c10::optional<double> scales, at::Tensor & grad_input);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & grad_output, c10::SymIntArrayRef output_size, c10::SymIntArrayRef input_size, c10::optional<double> scales, at::Tensor & grad_input);
+};
+
+struct TORCH_API _upsample_nearest_exact1d_backward {
+  using schema = at::Tensor (const at::Tensor &, c10::SymIntArrayRef, 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_backward")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "_upsample_nearest_exact1d_backward(Tensor grad_output, SymInt[1] output_size, SymInt[3] input_size, float? scales=None) -> Tensor")
+  static at::Tensor call(const at::Tensor & grad_output, c10::SymIntArrayRef output_size, c10::SymIntArrayRef input_size, c10::optional<double> scales);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & grad_output, c10::SymIntArrayRef output_size, c10::SymIntArrayRef input_size, c10::optional<double> scales);
+};
+
+}} // namespace at::_ops

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_upsample_nearest_exact2d.h

@@ -0,0 +1,113 @@
+#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/_upsample_nearest_exact2d_ops.h>
+
+namespace at {
+
+
+// aten::_upsample_nearest_exact2d.vec(Tensor input, SymInt[]? output_size, float[]? scale_factors) -> Tensor
+inline at::Tensor _upsample_nearest_exact2d(const at::Tensor & input, at::OptionalIntArrayRef output_size, c10::optional<at::ArrayRef<double>> scale_factors) {
+    return at::_ops::_upsample_nearest_exact2d_vec::call(input, output_size.has_value() ? c10::make_optional(c10::fromIntArrayRefSlow(*output_size)) : c10::nullopt, scale_factors);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same<T, int64_t>::value>>
+  at::Tensor _upsample_nearest_exact2d(const at::Tensor & input, at::OptionalIntArrayRef output_size, c10::optional<at::ArrayRef<double>> scale_factors) {
+    return at::_ops::_upsample_nearest_exact2d_vec::call(input, output_size.has_value() ? c10::make_optional(c10::fromIntArrayRefSlow(*output_size)) : c10::nullopt, scale_factors);
+  }
+}
+
+// aten::_upsample_nearest_exact2d.vec(Tensor input, SymInt[]? output_size, float[]? scale_factors) -> Tensor
+inline at::Tensor _upsample_nearest_exact2d_symint(const at::Tensor & input, at::OptionalSymIntArrayRef output_size, c10::optional<at::ArrayRef<double>> scale_factors) {
+    return at::_ops::_upsample_nearest_exact2d_vec::call(input, output_size, scale_factors);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same<T, c10::SymInt>::value>>
+  at::Tensor _upsample_nearest_exact2d(const at::Tensor & input, at::OptionalSymIntArrayRef output_size, c10::optional<at::ArrayRef<double>> scale_factors) {
+    return at::_ops::_upsample_nearest_exact2d_vec::call(input, output_size, scale_factors);
+  }
+}
+
+// aten::_upsample_nearest_exact2d.out(Tensor self, SymInt[2] output_size, float? scales_h=None, float? scales_w=None, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & _upsample_nearest_exact2d_out(at::Tensor & out, const at::Tensor & self, at::IntArrayRef output_size, c10::optional<double> scales_h=c10::nullopt, c10::optional<double> scales_w=c10::nullopt) {
+    return at::_ops::_upsample_nearest_exact2d_out::call(self, c10::fromIntArrayRefSlow(output_size), scales_h, scales_w, out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same<T, int64_t>::value>>
+  at::Tensor & _upsample_nearest_exact2d_out(at::Tensor & out, const at::Tensor & self, at::IntArrayRef output_size, c10::optional<double> scales_h=c10::nullopt, c10::optional<double> scales_w=c10::nullopt) {
+    return at::_ops::_upsample_nearest_exact2d_out::call(self, c10::fromIntArrayRefSlow(output_size), scales_h, scales_w, out);
+  }
+}
+
+// aten::_upsample_nearest_exact2d.out(Tensor self, SymInt[2] output_size, float? scales_h=None, float? scales_w=None, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & _upsample_nearest_exact2d_outf(const at::Tensor & self, at::IntArrayRef output_size, c10::optional<double> scales_h, c10::optional<double> scales_w, at::Tensor & out) {
+    return at::_ops::_upsample_nearest_exact2d_out::call(self, c10::fromIntArrayRefSlow(output_size), scales_h, scales_w, out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same<T, int64_t>::value>>
+  at::Tensor & _upsample_nearest_exact2d_outf(const at::Tensor & self, at::IntArrayRef output_size, c10::optional<double> scales_h, c10::optional<double> scales_w, at::Tensor & out) {
+    return at::_ops::_upsample_nearest_exact2d_out::call(self, c10::fromIntArrayRefSlow(output_size), scales_h, scales_w, out);
+  }
+}
+
+// aten::_upsample_nearest_exact2d.out(Tensor self, SymInt[2] output_size, float? scales_h=None, float? scales_w=None, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & _upsample_nearest_exact2d_symint_out(at::Tensor & out, const at::Tensor & self, c10::SymIntArrayRef output_size, c10::optional<double> scales_h=c10::nullopt, c10::optional<double> scales_w=c10::nullopt) {
+    return at::_ops::_upsample_nearest_exact2d_out::call(self, output_size, scales_h, scales_w, out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same<T, c10::SymInt>::value>>
+  at::Tensor & _upsample_nearest_exact2d_out(at::Tensor & out, const at::Tensor & self, c10::SymIntArrayRef output_size, c10::optional<double> scales_h=c10::nullopt, c10::optional<double> scales_w=c10::nullopt) {
+    return at::_ops::_upsample_nearest_exact2d_out::call(self, output_size, scales_h, scales_w, out);
+  }
+}
+
+// aten::_upsample_nearest_exact2d.out(Tensor self, SymInt[2] output_size, float? scales_h=None, float? scales_w=None, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & _upsample_nearest_exact2d_symint_outf(const at::Tensor & self, c10::SymIntArrayRef output_size, c10::optional<double> scales_h, c10::optional<double> scales_w, at::Tensor & out) {
+    return at::_ops::_upsample_nearest_exact2d_out::call(self, output_size, scales_h, scales_w, out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same<T, c10::SymInt>::value>>
+  at::Tensor & _upsample_nearest_exact2d_outf(const at::Tensor & self, c10::SymIntArrayRef output_size, c10::optional<double> scales_h, c10::optional<double> scales_w, at::Tensor & out) {
+    return at::_ops::_upsample_nearest_exact2d_out::call(self, output_size, scales_h, scales_w, out);
+  }
+}
+
+// aten::_upsample_nearest_exact2d(Tensor self, SymInt[2] output_size, float? scales_h=None, float? scales_w=None) -> Tensor
+inline at::Tensor _upsample_nearest_exact2d(const at::Tensor & self, at::IntArrayRef output_size, c10::optional<double> scales_h=c10::nullopt, c10::optional<double> scales_w=c10::nullopt) {
+    return at::_ops::_upsample_nearest_exact2d::call(self, c10::fromIntArrayRefSlow(output_size), scales_h, scales_w);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same<T, int64_t>::value>>
+  at::Tensor _upsample_nearest_exact2d(const at::Tensor & self, at::IntArrayRef output_size, c10::optional<double> scales_h=c10::nullopt, c10::optional<double> scales_w=c10::nullopt) {
+    return at::_ops::_upsample_nearest_exact2d::call(self, c10::fromIntArrayRefSlow(output_size), scales_h, scales_w);
+  }
+}
+
+// aten::_upsample_nearest_exact2d(Tensor self, SymInt[2] output_size, float? scales_h=None, float? scales_w=None) -> Tensor
+inline at::Tensor _upsample_nearest_exact2d_symint(const at::Tensor & self, c10::SymIntArrayRef output_size, c10::optional<double> scales_h=c10::nullopt, c10::optional<double> scales_w=c10::nullopt) {
+    return at::_ops::_upsample_nearest_exact2d::call(self, output_size, scales_h, scales_w);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same<T, c10::SymInt>::value>>
+  at::Tensor _upsample_nearest_exact2d(const at::Tensor & self, c10::SymIntArrayRef output_size, c10::optional<double> scales_h=c10::nullopt, c10::optional<double> scales_w=c10::nullopt) {
+    return at::_ops::_upsample_nearest_exact2d::call(self, output_size, scales_h, scales_w);
+  }
+}
+
+}

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_upsample_nearest_exact3d_cpu_dispatch.h

@@ -0,0 +1,28 @@
+#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 _upsample_nearest_exact3d(const at::Tensor & self, at::IntArrayRef output_size, c10::optional<double> scales_d=c10::nullopt, c10::optional<double> scales_h=c10::nullopt, c10::optional<double> scales_w=c10::nullopt);
+TORCH_API at::Tensor _upsample_nearest_exact3d_symint(const at::Tensor & self, c10::SymIntArrayRef output_size, c10::optional<double> scales_d=c10::nullopt, c10::optional<double> scales_h=c10::nullopt, c10::optional<double> scales_w=c10::nullopt);
+TORCH_API at::Tensor & _upsample_nearest_exact3d_out(at::Tensor & out, const at::Tensor & self, at::IntArrayRef output_size, c10::optional<double> scales_d=c10::nullopt, c10::optional<double> scales_h=c10::nullopt, c10::optional<double> scales_w=c10::nullopt);
+TORCH_API at::Tensor & _upsample_nearest_exact3d_outf(const at::Tensor & self, at::IntArrayRef output_size, c10::optional<double> scales_d, c10::optional<double> scales_h, c10::optional<double> scales_w, at::Tensor & out);
+TORCH_API at::Tensor & _upsample_nearest_exact3d_symint_out(at::Tensor & out, const at::Tensor & self, c10::SymIntArrayRef output_size, c10::optional<double> scales_d=c10::nullopt, c10::optional<double> scales_h=c10::nullopt, c10::optional<double> scales_w=c10::nullopt);
+TORCH_API at::Tensor & _upsample_nearest_exact3d_symint_outf(const at::Tensor & self, c10::SymIntArrayRef output_size, c10::optional<double> scales_d, c10::optional<double> scales_h, c10::optional<double> scales_w, at::Tensor & out);
+
+} // namespace cpu
+} // namespace at

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/arcsinh_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 arcsinh(const at::Tensor & self);
+TORCH_API at::Tensor & arcsinh_out(const at::Tensor & self, at::Tensor & out);
+TORCH_API at::Tensor & arcsinh_(at::Tensor & self);
+} // namespace native
+} // namespace at

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/arctan_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 arctan(const at::Tensor & self);
+TORCH_API at::Tensor & arctan_out(const at::Tensor & self, at::Tensor & out);
+TORCH_API at::Tensor & arctan_(at::Tensor & self);
+} // namespace native
+} // namespace at

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/argmin_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 argmin(const at::Tensor & self, c10::optional<int64_t> dim=c10::nullopt, bool keepdim=false);
+TORCH_API at::Tensor & argmin_out(at::Tensor & out, const at::Tensor & self, c10::optional<int64_t> dim=c10::nullopt, bool keepdim=false);
+TORCH_API at::Tensor & argmin_outf(const at::Tensor & self, c10::optional<int64_t> dim, bool keepdim, at::Tensor & out);
+
+} // namespace cpu
+} // namespace at

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/argmin_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 argmin(const at::Tensor & self, c10::optional<int64_t> dim=c10::nullopt, bool keepdim=false);
+TORCH_API at::Tensor & argmin_out(at::Tensor & out, const at::Tensor & self, c10::optional<int64_t> dim=c10::nullopt, bool keepdim=false);
+TORCH_API at::Tensor & argmin_outf(const at::Tensor & self, c10::optional<int64_t> dim, bool keepdim, at::Tensor & out);
+
+} // namespace cuda
+} // namespace at

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/asinh_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 asinh(const at::Tensor & self);
+TORCH_API at::Tensor & asinh_(at::Tensor & self);
+
+} // namespace compositeexplicitautogradnonfunctional
+} // namespace at

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/ceil_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 ceil(const at::Tensor & self);
+TORCH_API at::Tensor & ceil_out(at::Tensor & out, const at::Tensor & self);
+TORCH_API at::Tensor & ceil_outf(const at::Tensor & self, at::Tensor & out);
+TORCH_API at::Tensor & ceil_(at::Tensor & self);
+
+} // namespace meta
+} // namespace at

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/cholesky.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/cholesky_ops.h>
+
+namespace at {
+
+
+// aten::cholesky.out(Tensor self, bool upper=False, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & cholesky_out(at::Tensor & out, const at::Tensor & self, bool upper=false) {
+    return at::_ops::cholesky_out::call(self, upper, out);
+}
+// aten::cholesky.out(Tensor self, bool upper=False, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & cholesky_outf(const at::Tensor & self, bool upper, at::Tensor & out) {
+    return at::_ops::cholesky_out::call(self, upper, out);
+}
+
+// aten::cholesky(Tensor self, bool upper=False) -> Tensor
+inline at::Tensor cholesky(const at::Tensor & self, bool upper=false) {
+    return at::_ops::cholesky::call(self, upper);
+}
+
+}

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/cumprod_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 cumprod(const at::Tensor & self, int64_t dim, c10::optional<at::ScalarType> dtype=c10::nullopt);
+TORCH_API at::Tensor & cumprod_out(at::Tensor & out, const at::Tensor & self, int64_t dim, c10::optional<at::ScalarType> dtype=c10::nullopt);
+TORCH_API at::Tensor & cumprod_outf(const at::Tensor & self, int64_t dim, c10::optional<at::ScalarType> dtype, at::Tensor & out);
+TORCH_API at::Tensor & cumprod_(at::Tensor & self, int64_t dim, c10::optional<at::ScalarType> dtype=c10::nullopt);
+
+} // namespace cuda
+} // namespace at

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

@@ -0,0 +1,53 @@
+#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/eq_ops.h>
+
+namespace at {
+
+
+// aten::eq.Scalar_out(Tensor self, Scalar other, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & eq_out(at::Tensor & out, const at::Tensor & self, const at::Scalar & other) {
+    return at::_ops::eq_Scalar_out::call(self, other, out);
+}
+// aten::eq.Scalar_out(Tensor self, Scalar other, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & eq_outf(const at::Tensor & self, const at::Scalar & other, at::Tensor & out) {
+    return at::_ops::eq_Scalar_out::call(self, other, out);
+}
+
+// aten::eq.Scalar(Tensor self, Scalar other) -> Tensor
+inline at::Tensor eq(const at::Tensor & self, const at::Scalar & other) {
+    return at::_ops::eq_Scalar::call(self, other);
+}
+
+// aten::eq.Tensor_out(Tensor self, Tensor other, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & eq_out(at::Tensor & out, const at::Tensor & self, const at::Tensor & other) {
+    return at::_ops::eq_Tensor_out::call(self, other, out);
+}
+// aten::eq.Tensor_out(Tensor self, Tensor other, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & eq_outf(const at::Tensor & self, const at::Tensor & other, at::Tensor & out) {
+    return at::_ops::eq_Tensor_out::call(self, other, out);
+}
+
+// aten::eq.Tensor(Tensor self, Tensor other) -> Tensor
+inline at::Tensor eq(const at::Tensor & self, const at::Tensor & other) {
+    return at::_ops::eq_Tensor::call(self, other);
+}
+
+}

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/flatten_dense_tensors_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 flatten_dense_tensors(at::TensorList tensors);
+} // namespace native
+} // namespace at

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/glu_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 glu(const at::Tensor & self, int64_t dim=-1);
+TORCH_API at::Tensor & glu_out(at::Tensor & out, const at::Tensor & self, int64_t dim=-1);
+TORCH_API at::Tensor & glu_outf(const at::Tensor & self, int64_t dim, at::Tensor & out);
+
+} // namespace cuda
+} // namespace at

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

@@ -0,0 +1,38 @@
+#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 hamming_window(int64_t window_length, at::TensorOptions options={});
+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(at::Tensor & out, int64_t window_length);
+TORCH_API at::Tensor & hamming_window_outf(int64_t window_length, at::Tensor & out);
+TORCH_API at::Tensor hamming_window(int64_t window_length, bool periodic, at::TensorOptions options={});
+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_out(at::Tensor & out, int64_t window_length, bool periodic);
+TORCH_API at::Tensor & hamming_window_outf(int64_t window_length, bool periodic, at::Tensor & out);
+TORCH_API at::Tensor hamming_window(int64_t window_length, bool periodic, double alpha, at::TensorOptions options={});
+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_out(at::Tensor & out, int64_t window_length, bool periodic, double alpha);
+TORCH_API at::Tensor & hamming_window_outf(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, at::TensorOptions options={});
+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_out(at::Tensor & out, int64_t window_length, bool periodic, double alpha, double beta);
+TORCH_API at::Tensor & hamming_window_outf(int64_t window_length, bool periodic, double alpha, double beta, at::Tensor & out);
+
+} // namespace compositeexplicitautograd
+} // namespace at

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/index.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/index_ops.h>
+
+namespace at {
+
+
+// aten::index.Tensor(Tensor self, Tensor?[] indices) -> Tensor
+inline at::Tensor index(const at::Tensor & self, const c10::List<c10::optional<at::Tensor>> & indices) {
+    return at::_ops::index_Tensor::call(self, indices);
+}
+
+// aten::index.Tensor_out(Tensor self, Tensor?[] indices, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & index_out(at::Tensor & out, const at::Tensor & self, const c10::List<c10::optional<at::Tensor>> & indices) {
+    return at::_ops::index_Tensor_out::call(self, indices, out);
+}
+// aten::index.Tensor_out(Tensor self, Tensor?[] indices, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & index_outf(const at::Tensor & self, const c10::List<c10::optional<at::Tensor>> & indices, at::Tensor & out) {
+    return at::_ops::index_Tensor_out::call(self, indices, out);
+}
+
+}

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/is_nonzero_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 is_nonzero {
+  using schema = bool (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::is_nonzero")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "is_nonzero(Tensor self) -> bool")
+  static bool call(const at::Tensor & self);
+  static bool redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self);
+};
+
+}} // namespace at::_ops

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

@@ -0,0 +1,83 @@
+#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 isin_Tensor_Tensor_out {
+  using schema = at::Tensor & (const at::Tensor &, const at::Tensor &, bool, bool, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(name, "aten::isin")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "Tensor_Tensor_out")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "isin.Tensor_Tensor_out(Tensor elements, Tensor test_elements, *, bool assume_unique=False, bool invert=False, Tensor(a!) out) -> Tensor(a!)")
+  static at::Tensor & call(const at::Tensor & elements, const at::Tensor & test_elements, bool assume_unique, bool invert, at::Tensor & out);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & elements, const at::Tensor & test_elements, bool assume_unique, bool invert, at::Tensor & out);
+};
+
+struct TORCH_API isin_Tensor_Tensor {
+  using schema = at::Tensor (const at::Tensor &, const at::Tensor &, bool, bool);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(name, "aten::isin")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "Tensor_Tensor")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "isin.Tensor_Tensor(Tensor elements, Tensor test_elements, *, bool assume_unique=False, bool invert=False) -> Tensor")
+  static at::Tensor call(const at::Tensor & elements, const at::Tensor & test_elements, bool assume_unique, bool invert);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & elements, const at::Tensor & test_elements, bool assume_unique, bool invert);
+};
+
+struct TORCH_API isin_Tensor_Scalar_out {
+  using schema = at::Tensor & (const at::Tensor &, const at::Scalar &, bool, bool, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(name, "aten::isin")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "Tensor_Scalar_out")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "isin.Tensor_Scalar_out(Tensor elements, Scalar test_element, *, bool assume_unique=False, bool invert=False, Tensor(a!) out) -> Tensor(a!)")
+  static at::Tensor & call(const at::Tensor & elements, const at::Scalar & test_element, bool assume_unique, bool invert, at::Tensor & out);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & elements, const at::Scalar & test_element, bool assume_unique, bool invert, at::Tensor & out);
+};
+
+struct TORCH_API isin_Tensor_Scalar {
+  using schema = at::Tensor (const at::Tensor &, const at::Scalar &, bool, bool);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(name, "aten::isin")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "Tensor_Scalar")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "isin.Tensor_Scalar(Tensor elements, Scalar test_element, *, bool assume_unique=False, bool invert=False) -> Tensor")
+  static at::Tensor call(const at::Tensor & elements, const at::Scalar & test_element, bool assume_unique, bool invert);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & elements, const at::Scalar & test_element, bool assume_unique, bool invert);
+};
+
+struct TORCH_API isin_Scalar_Tensor_out {
+  using schema = at::Tensor & (const at::Scalar &, const at::Tensor &, bool, bool, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(name, "aten::isin")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "Scalar_Tensor_out")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "isin.Scalar_Tensor_out(Scalar element, Tensor test_elements, *, bool assume_unique=False, bool invert=False, Tensor(a!) out) -> Tensor(a!)")
+  static at::Tensor & call(const at::Scalar & element, const at::Tensor & test_elements, bool assume_unique, bool invert, at::Tensor & out);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, const at::Scalar & element, const at::Tensor & test_elements, bool assume_unique, bool invert, at::Tensor & out);
+};
+
+struct TORCH_API isin_Scalar_Tensor {
+  using schema = at::Tensor (const at::Scalar &, const at::Tensor &, bool, bool);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(name, "aten::isin")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "Scalar_Tensor")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "isin.Scalar_Tensor(Scalar element, Tensor test_elements, *, bool assume_unique=False, bool invert=False) -> Tensor")
+  static at::Tensor call(const at::Scalar & element, const at::Tensor & test_elements, bool assume_unique, bool invert);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Scalar & element, const at::Tensor & test_elements, bool assume_unique, bool invert);
+};
+
+}} // namespace at::_ops

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/isneginf_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 isneginf(const at::Tensor & self);
+
+} // namespace compositeexplicitautogradnonfunctional
+} // namespace at

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

@@ -0,0 +1,79 @@
+#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/kaiser_window_ops.h>
+
+namespace at {
+
+
+// aten::kaiser_window(int window_length, *, ScalarType? dtype=None, Layout? layout=None, Device? device=None, bool? pin_memory=None) -> Tensor
+inline at::Tensor kaiser_window(int64_t window_length, at::TensorOptions options={}) {
+    return at::_ops::kaiser_window::call(window_length, c10::optTypeMetaToScalarType(options.dtype_opt()), options.layout_opt(), options.device_opt(), options.pinned_memory_opt());
+}
+// aten::kaiser_window(int window_length, *, ScalarType? dtype=None, Layout? layout=None, Device? device=None, bool? pin_memory=None) -> Tensor
+inline at::Tensor kaiser_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) {
+    return at::_ops::kaiser_window::call(window_length, dtype, layout, device, pin_memory);
+}
+
+// aten::kaiser_window.periodic(int window_length, bool periodic, *, ScalarType? dtype=None, Layout? layout=None, Device? device=None, bool? pin_memory=None) -> Tensor
+inline at::Tensor kaiser_window(int64_t window_length, bool periodic, at::TensorOptions options={}) {
+    return at::_ops::kaiser_window_periodic::call(window_length, periodic, c10::optTypeMetaToScalarType(options.dtype_opt()), options.layout_opt(), options.device_opt(), options.pinned_memory_opt());
+}
+// aten::kaiser_window.periodic(int window_length, bool periodic, *, ScalarType? dtype=None, Layout? layout=None, Device? device=None, bool? pin_memory=None) -> Tensor
+inline at::Tensor kaiser_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) {
+    return at::_ops::kaiser_window_periodic::call(window_length, periodic, dtype, layout, device, pin_memory);
+}
+
+// aten::kaiser_window.beta(int window_length, bool periodic, float beta, *, ScalarType? dtype=None, Layout? layout=None, Device? device=None, bool? pin_memory=None) -> Tensor
+inline at::Tensor kaiser_window(int64_t window_length, bool periodic, double beta, at::TensorOptions options={}) {
+    return at::_ops::kaiser_window_beta::call(window_length, periodic, beta, c10::optTypeMetaToScalarType(options.dtype_opt()), options.layout_opt(), options.device_opt(), options.pinned_memory_opt());
+}
+// aten::kaiser_window.beta(int window_length, bool periodic, float beta, *, ScalarType? dtype=None, Layout? layout=None, Device? device=None, bool? pin_memory=None) -> Tensor
+inline at::Tensor kaiser_window(int64_t window_length, bool periodic, double beta, c10::optional<at::ScalarType> dtype, c10::optional<at::Layout> layout, c10::optional<at::Device> device, c10::optional<bool> pin_memory) {
+    return at::_ops::kaiser_window_beta::call(window_length, periodic, beta, dtype, layout, device, pin_memory);
+}
+
+// aten::kaiser_window.out(int window_length, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & kaiser_window_out(at::Tensor & out, int64_t window_length) {
+    return at::_ops::kaiser_window_out::call(window_length, out);
+}
+// aten::kaiser_window.out(int window_length, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & kaiser_window_outf(int64_t window_length, at::Tensor & out) {
+    return at::_ops::kaiser_window_out::call(window_length, out);
+}
+
+// aten::kaiser_window.periodic_out(int window_length, bool periodic, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & kaiser_window_out(at::Tensor & out, int64_t window_length, bool periodic) {
+    return at::_ops::kaiser_window_periodic_out::call(window_length, periodic, out);
+}
+// aten::kaiser_window.periodic_out(int window_length, bool periodic, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & kaiser_window_outf(int64_t window_length, bool periodic, at::Tensor & out) {
+    return at::_ops::kaiser_window_periodic_out::call(window_length, periodic, out);
+}
+
+// aten::kaiser_window.beta_out(int window_length, bool periodic, float beta, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & kaiser_window_out(at::Tensor & out, int64_t window_length, bool periodic, double beta) {
+    return at::_ops::kaiser_window_beta_out::call(window_length, periodic, beta, out);
+}
+// aten::kaiser_window.beta_out(int window_length, bool periodic, float beta, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & kaiser_window_outf(int64_t window_length, bool periodic, double beta, at::Tensor & out) {
+    return at::_ops::kaiser_window_beta_out::call(window_length, periodic, beta, out);
+}
+
+}

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/lcm_cpu_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 cpu {
+
+TORCH_API at::Tensor lcm(const at::Tensor & self, const at::Tensor & other);
+TORCH_API at::Tensor & lcm_out(at::Tensor & out, const at::Tensor & self, const at::Tensor & other);
+TORCH_API at::Tensor & lcm_outf(const at::Tensor & self, const at::Tensor & other, at::Tensor & out);
+TORCH_API at::Tensor & lcm_(at::Tensor & self, const at::Tensor & other);
+
+} // namespace cpu
+} // namespace at

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/lift_fresh_compositeexplicitautograd_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 compositeexplicitautograd {
+
+TORCH_API at::Tensor lift_fresh(const at::Tensor & self);
+
+} // namespace compositeexplicitautograd
+} // namespace at

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/linalg_cholesky_ex_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 ::std::tuple<at::Tensor,at::Tensor> linalg_cholesky_ex(const at::Tensor & self, bool upper=false, bool check_errors=false);
+TORCH_API ::std::tuple<at::Tensor &,at::Tensor &> linalg_cholesky_ex_out(at::Tensor & L, at::Tensor & info, const at::Tensor & self, bool upper=false, bool check_errors=false);
+TORCH_API ::std::tuple<at::Tensor &,at::Tensor &> linalg_cholesky_ex_outf(const at::Tensor & self, bool upper, bool check_errors, at::Tensor & L, at::Tensor & info);
+
+} // namespace meta
+} // namespace at

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/linalg_inv_ex.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/linalg_inv_ex_ops.h>
+
+namespace at {
+
+
+// aten::linalg_inv_ex(Tensor A, *, bool check_errors=False) -> (Tensor inverse, Tensor info)
+inline ::std::tuple<at::Tensor,at::Tensor> linalg_inv_ex(const at::Tensor & A, bool check_errors=false) {
+    return at::_ops::linalg_inv_ex::call(A, check_errors);
+}
+
+// aten::linalg_inv_ex.inverse(Tensor A, *, bool check_errors=False, Tensor(a!) inverse, Tensor(b!) info) -> (Tensor(a!) inverse, Tensor(b!) info)
+inline ::std::tuple<at::Tensor &,at::Tensor &> linalg_inv_ex_out(at::Tensor & inverse, at::Tensor & info, const at::Tensor & A, bool check_errors=false) {
+    return at::_ops::linalg_inv_ex_inverse::call(A, check_errors, inverse, info);
+}
+// aten::linalg_inv_ex.inverse(Tensor A, *, bool check_errors=False, Tensor(a!) inverse, Tensor(b!) info) -> (Tensor(a!) inverse, Tensor(b!) info)
+inline ::std::tuple<at::Tensor &,at::Tensor &> linalg_inv_ex_outf(const at::Tensor & A, bool check_errors, at::Tensor & inverse, at::Tensor & info) {
+    return at::_ops::linalg_inv_ex_inverse::call(A, check_errors, inverse, info);
+}
+
+}

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/linalg_slogdet_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> linalg_slogdet(const at::Tensor & A);
+TORCH_API ::std::tuple<at::Tensor &,at::Tensor &> linalg_slogdet_out(const at::Tensor & A, at::Tensor & sign, at::Tensor & logabsdet);
+} // namespace native
+} // namespace at

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/log_sigmoid_backward_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 log_sigmoid_backward(const at::Tensor & grad_output, const at::Tensor & self, const at::Tensor & buffer);
+TORCH_API at::Tensor & log_sigmoid_backward_out(at::Tensor & grad_input, const at::Tensor & grad_output, const at::Tensor & self, const at::Tensor & buffer);
+TORCH_API at::Tensor & log_sigmoid_backward_outf(const at::Tensor & grad_output, const at::Tensor & self, const at::Tensor & buffer, at::Tensor & grad_input);
+
+} // namespace cuda
+} // namespace at

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/log_sigmoid_forward_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 ::std::tuple<at::Tensor,at::Tensor> log_sigmoid_forward(const at::Tensor & self);
+TORCH_API ::std::tuple<at::Tensor &,at::Tensor &> log_sigmoid_forward_out(at::Tensor & output, at::Tensor & buffer, const at::Tensor & self);
+TORCH_API ::std::tuple<at::Tensor &,at::Tensor &> log_sigmoid_forward_outf(const at::Tensor & self, at::Tensor & output, at::Tensor & buffer);
+
+} // namespace cuda
+} // namespace at

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/lu_solve_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 lu_solve(const at::Tensor & self, const at::Tensor & LU_data, const at::Tensor & LU_pivots);
+TORCH_API at::Tensor & lu_solve_out(const at::Tensor & self, const at::Tensor & LU_data, const at::Tensor & LU_pivots, at::Tensor & out);
+} // namespace native
+} // namespace at

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/miopen_rnn_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 miopen_rnn {
+  using schema = ::std::tuple<at::Tensor,at::Tensor,at::Tensor,at::Tensor,at::Tensor> (const at::Tensor &, at::TensorList, int64_t, const at::Tensor &, const c10::optional<at::Tensor> &, int64_t, int64_t, int64_t, bool, double, bool, bool, at::IntArrayRef, const c10::optional<at::Tensor> &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(name, "aten::miopen_rnn")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "miopen_rnn(Tensor input, Tensor[] weight, int weight_stride0, Tensor hx, Tensor? cx, int mode, int hidden_size, int num_layers, bool batch_first, float dropout, bool train, bool bidirectional, int[] batch_sizes, Tensor? dropout_state) -> (Tensor, Tensor, Tensor, Tensor, Tensor)")
+  static ::std::tuple<at::Tensor,at::Tensor,at::Tensor,at::Tensor,at::Tensor> call(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);
+  static ::std::tuple<at::Tensor,at::Tensor,at::Tensor,at::Tensor,at::Tensor> redispatch(c10::DispatchKeySet dispatchKeySet, 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);
+};
+
+struct TORCH_API miopen_rnn_out {
+  using schema = ::std::tuple<at::Tensor &,at::Tensor &,at::Tensor &,at::Tensor &,at::Tensor &> (const at::Tensor &, at::TensorList, int64_t, const at::Tensor &, const c10::optional<at::Tensor> &, int64_t, int64_t, int64_t, bool, double, bool, bool, at::IntArrayRef, const c10::optional<at::Tensor> &, at::Tensor &, at::Tensor &, 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::miopen_rnn")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "out")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "miopen_rnn.out(Tensor input, Tensor[] weight, int weight_stride0, Tensor hx, Tensor? cx, int mode, int hidden_size, int num_layers, bool batch_first, float dropout, bool train, bool bidirectional, int[] batch_sizes, Tensor? dropout_state, *, Tensor(a!) out0, Tensor(b!) out1, Tensor(c!) out2, Tensor(d!) out3, Tensor(e!) out4) -> (Tensor(a!), Tensor(b!), Tensor(c!), Tensor(d!), Tensor(e!))")
+  static ::std::tuple<at::Tensor &,at::Tensor &,at::Tensor &,at::Tensor &,at::Tensor &> call(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, at::Tensor & out0, at::Tensor & out1, at::Tensor & out2, at::Tensor & out3, at::Tensor & out4);
+  static ::std::tuple<at::Tensor &,at::Tensor &,at::Tensor &,at::Tensor &,at::Tensor &> redispatch(c10::DispatchKeySet dispatchKeySet, 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, at::Tensor & out0, at::Tensor & out1, at::Tensor & out2, at::Tensor & out3, at::Tensor & out4);
+};
+
+}} // namespace at::_ops

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/mm_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 mm(const at::Tensor & self, const at::Tensor & mat2);
+TORCH_API at::Tensor & mm_out(at::Tensor & out, const at::Tensor & self, const at::Tensor & mat2);
+TORCH_API at::Tensor & mm_outf(const at::Tensor & self, const at::Tensor & mat2, at::Tensor & out);
+
+} // namespace cuda
+} // namespace at

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/mse_loss_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/mse_loss_meta.h>
+
+namespace at {
+namespace native {
+struct TORCH_API structured_mse_loss_out : public at::meta::structured_mse_loss {
+void impl(const at::Tensor & self, const at::Tensor & target, int64_t reduction, const at::Tensor & out);
+};
+} // namespace native
+} // namespace at

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/multilabel_margin_loss_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 multilabel_margin_loss(const at::Tensor & self, const at::Tensor & target, int64_t reduction=at::Reduction::Mean);
+TORCH_API at::Tensor & multilabel_margin_loss_out(const at::Tensor & self, const at::Tensor & target, int64_t reduction, at::Tensor & out);
+} // namespace native
+} // namespace at