Add files using upload-large-folder tool

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

@@ -0,0 +1,26 @@
+#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 add_relu(const at::Tensor & self, const at::Tensor & other, const at::Scalar & alpha=1);
+TORCH_API at::Tensor & add_relu_out(const at::Tensor & self, const at::Tensor & other, const at::Scalar & alpha, at::Tensor & out);
+TORCH_API at::Tensor & add_relu_(at::Tensor & self, const at::Tensor & other, const at::Scalar & alpha=1);
+TORCH_API at::Tensor & _add_relu_Scalar_out(const at::Tensor & self, const at::Scalar & other, const at::Scalar & alpha, at::Tensor & out);
+TORCH_API at::Tensor add_relu(const at::Tensor & self, const at::Scalar & other, const at::Scalar & alpha=1);
+TORCH_API at::Tensor & add_relu_(at::Tensor & self, const at::Scalar & other, const at::Scalar & alpha=1);
+} // namespace native
+} // namespace at

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_assert_async_cpu_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 cpu {
+
+TORCH_API void _assert_async(const at::Tensor & self);
+TORCH_API void _assert_async(const at::Tensor & self, c10::string_view assert_msg);
+
+} // namespace cpu
+} // namespace at

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_conv_depthwise2d_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 _conv_depthwise2d_out {
+  using schema = const at::Tensor & (const at::Tensor &, const at::Tensor &, c10::SymIntArrayRef, const c10::optional<at::Tensor> &, c10::SymIntArrayRef, c10::SymIntArrayRef, c10::SymIntArrayRef, 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::_conv_depthwise2d")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "out")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "_conv_depthwise2d.out(Tensor self, Tensor weight, SymInt[2] kernel_size, Tensor? bias, SymInt[2] stride, SymInt[2] padding, SymInt[2] dilation, *, Tensor(a!) out) -> Tensor(a!)")
+  static const at::Tensor & call(const at::Tensor & self, const at::Tensor & weight, c10::SymIntArrayRef kernel_size, const c10::optional<at::Tensor> & bias, c10::SymIntArrayRef stride, c10::SymIntArrayRef padding, c10::SymIntArrayRef dilation, const at::Tensor & out);
+  static const at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, const at::Tensor & weight, c10::SymIntArrayRef kernel_size, const c10::optional<at::Tensor> & bias, c10::SymIntArrayRef stride, c10::SymIntArrayRef padding, c10::SymIntArrayRef dilation, const at::Tensor & out);
+};
+
+struct TORCH_API _conv_depthwise2d {
+  using schema = at::Tensor (const at::Tensor &, const at::Tensor &, c10::SymIntArrayRef, const c10::optional<at::Tensor> &, c10::SymIntArrayRef, c10::SymIntArrayRef, c10::SymIntArrayRef);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(name, "aten::_conv_depthwise2d")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "_conv_depthwise2d(Tensor self, Tensor weight, SymInt[2] kernel_size, Tensor? bias, SymInt[2] stride, SymInt[2] padding, SymInt[2] dilation) -> Tensor")
+  static at::Tensor call(const at::Tensor & self, const at::Tensor & weight, c10::SymIntArrayRef kernel_size, const c10::optional<at::Tensor> & bias, c10::SymIntArrayRef stride, c10::SymIntArrayRef padding, c10::SymIntArrayRef dilation);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, const at::Tensor & weight, c10::SymIntArrayRef kernel_size, const c10::optional<at::Tensor> & bias, c10::SymIntArrayRef stride, c10::SymIntArrayRef padding, c10::SymIntArrayRef dilation);
+};
+
+}} // namespace at::_ops

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_copy_from_and_resize_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 & _copy_from_and_resize_out(at::Tensor & out, const at::Tensor & self, const at::Tensor & dst);
+TORCH_API at::Tensor & _copy_from_and_resize_outf(const at::Tensor & self, const at::Tensor & dst, at::Tensor & out);
+
+} // namespace compositeexplicitautograd
+} // namespace at

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

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_cufft_clear_plan_cache.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/_cufft_clear_plan_cache_ops.h>
+
+namespace at {
+
+
+// aten::_cufft_clear_plan_cache(DeviceIndex device_index) -> ()
+inline void _cufft_clear_plan_cache(at::DeviceIndex device_index) {
+    return at::_ops::_cufft_clear_plan_cache::call(device_index);
+}
+
+}

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_cufft_get_plan_cache_size_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 int64_t _cufft_get_plan_cache_size(at::DeviceIndex device_index);
+
+} // namespace compositeimplicitautograd
+} // namespace at

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_efficient_attention_backward_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 ::std::tuple<at::Tensor,at::Tensor,at::Tensor,at::Tensor> _efficient_attention_backward(const at::Tensor & grad_out_, const at::Tensor & query, const at::Tensor & key, const at::Tensor & value, const c10::optional<at::Tensor> & bias, const at::Tensor & out, const c10::optional<at::Tensor> & cu_seqlens_q, const c10::optional<at::Tensor> & cu_seqlens_k, int64_t max_seqlen_q, int64_t max_seqlen_k, const at::Tensor & logsumexp, double dropout_p, const at::Tensor & philox_seed, const at::Tensor & philox_offset, int64_t custom_mask_type, bool bias_requires_grad, c10::optional<double> scale=c10::nullopt, c10::optional<int64_t> num_splits_key=c10::nullopt);
+} // namespace native
+} // namespace at

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_fft_r2c_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 _fft_r2c_mkl(const at::Tensor & self, at::IntArrayRef dim, int64_t normalization, bool onesided);
+TORCH_API at::Tensor & _fft_r2c_mkl_out(const at::Tensor & self, at::IntArrayRef dim, int64_t normalization, bool onesided, at::Tensor & out);
+TORCH_API at::Tensor _fft_r2c_cufft(const at::Tensor & self, at::IntArrayRef dim, int64_t normalization, bool onesided);
+TORCH_API at::Tensor & _fft_r2c_cufft_out(const at::Tensor & self, at::IntArrayRef dim, int64_t normalization, bool onesided, at::Tensor & out);
+} // namespace native
+} // namespace at

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_foreach_addcmul_cuda_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 cuda {
+
+TORCH_API ::std::vector<at::Tensor> _foreach_addcmul(at::TensorList self, at::TensorList tensor1, at::TensorList tensor2, const at::Scalar & value=1);
+TORCH_API void _foreach_addcmul_(at::TensorList self, at::TensorList tensor1, at::TensorList tensor2, const at::Scalar & value=1);
+TORCH_API ::std::vector<at::Tensor> _foreach_addcmul(at::TensorList self, at::TensorList tensor1, at::TensorList tensor2, at::ArrayRef<at::Scalar> scalars);
+TORCH_API void _foreach_addcmul_(at::TensorList self, at::TensorList tensor1, at::TensorList tensor2, at::ArrayRef<at::Scalar> scalars);
+TORCH_API ::std::vector<at::Tensor> _foreach_addcmul(at::TensorList self, at::TensorList tensor1, at::TensorList tensor2, const at::Tensor & scalars);
+TORCH_API void _foreach_addcmul_(at::TensorList self, at::TensorList tensor1, at::TensorList tensor2, const at::Tensor & scalars);
+
+} // namespace cuda
+} // namespace at

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

@@ -0,0 +1,25 @@
+#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 _foreach_trunc_out(at::TensorList self, at::TensorList out);
+TORCH_API ::std::vector<at::Tensor> foreach_tensor_trunc_slow(at::TensorList self);
+TORCH_API void foreach_tensor_trunc_slow_(at::TensorList self);
+TORCH_API ::std::vector<at::Tensor> foreach_tensor_trunc_cuda(at::TensorList self);
+TORCH_API void foreach_tensor_trunc_cuda_(at::TensorList self);
+} // namespace native
+} // namespace at

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_functional_assert_async_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 at::Tensor _functional_assert_async(const at::Tensor & self, c10::string_view assert_msg, const at::Tensor & dep_token);
+
+} // namespace cpu
+} // namespace at

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

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_linalg_solve_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,at::Tensor,at::Tensor> _linalg_solve_ex(const at::Tensor & A, const at::Tensor & B, bool left=true, bool check_errors=false);
+TORCH_API ::std::tuple<at::Tensor &,at::Tensor &,at::Tensor &,at::Tensor &> _linalg_solve_ex_out(at::Tensor & result, at::Tensor & LU, at::Tensor & pivots, at::Tensor & info, const at::Tensor & A, const at::Tensor & B, bool left=true, bool check_errors=false);
+TORCH_API ::std::tuple<at::Tensor &,at::Tensor &,at::Tensor &,at::Tensor &> _linalg_solve_ex_outf(const at::Tensor & A, const at::Tensor & B, bool left, bool check_errors, at::Tensor & result, at::Tensor & LU, at::Tensor & pivots, at::Tensor & info);
+
+} // namespace meta
+} // namespace at

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_log_softmax_backward_data.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/_log_softmax_backward_data_ops.h>
+
+namespace at {
+
+
+// aten::_log_softmax_backward_data(Tensor grad_output, Tensor output, int dim, ScalarType input_dtype) -> Tensor
+inline at::Tensor _log_softmax_backward_data(const at::Tensor & grad_output, const at::Tensor & output, int64_t dim, at::ScalarType input_dtype) {
+    return at::_ops::_log_softmax_backward_data::call(grad_output, output, dim, input_dtype);
+}
+
+// aten::_log_softmax_backward_data.out(Tensor grad_output, Tensor output, int dim, ScalarType input_dtype, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & _log_softmax_backward_data_out(at::Tensor & out, const at::Tensor & grad_output, const at::Tensor & output, int64_t dim, at::ScalarType input_dtype) {
+    return at::_ops::_log_softmax_backward_data_out::call(grad_output, output, dim, input_dtype, out);
+}
+// aten::_log_softmax_backward_data.out(Tensor grad_output, Tensor output, int dim, ScalarType input_dtype, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & _log_softmax_backward_data_outf(const at::Tensor & grad_output, const at::Tensor & output, int64_t dim, at::ScalarType input_dtype, at::Tensor & out) {
+    return at::_ops::_log_softmax_backward_data_out::call(grad_output, output, dim, input_dtype, out);
+}
+
+}

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_nested_view_from_jagged.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_jagged_ops.h>
+
+namespace at {
+
+
+// aten::_nested_view_from_jagged(Tensor(a) self, Tensor offsets, Tensor dummy, Tensor? lengths=None, int ragged_idx=1) -> Tensor(a)
+inline at::Tensor _nested_view_from_jagged(const at::Tensor & self, const at::Tensor & offsets, const at::Tensor & dummy, const c10::optional<at::Tensor> & lengths={}, int64_t ragged_idx=1) {
+    return at::_ops::_nested_view_from_jagged::call(self, offsets, dummy, lengths, ragged_idx);
+}
+
+}

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_pin_memory_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 at::Tensor _pin_memory(const at::Tensor & self, c10::optional<at::Device> device=c10::nullopt);
+
+} // namespace cuda
+} // namespace at

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_scaled_mm_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 _scaled_mm {
+  using schema = ::std::tuple<at::Tensor,at::Tensor> (const at::Tensor &, const at::Tensor &, const c10::optional<at::Tensor> &, c10::optional<at::ScalarType>, const c10::optional<at::Tensor> &, const c10::optional<at::Tensor> &, const c10::optional<at::Tensor> &, bool);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(name, "aten::_scaled_mm")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "_scaled_mm(Tensor self, Tensor mat2, *, Tensor? bias=None, ScalarType? out_dtype=None, Tensor? scale_a=None, Tensor? scale_b=None, Tensor? scale_result=None, bool use_fast_accum=False) -> (Tensor, Tensor)")
+  static ::std::tuple<at::Tensor,at::Tensor> call(const at::Tensor & self, const at::Tensor & mat2, const c10::optional<at::Tensor> & bias, c10::optional<at::ScalarType> out_dtype, const c10::optional<at::Tensor> & scale_a, const c10::optional<at::Tensor> & scale_b, const c10::optional<at::Tensor> & scale_result, bool use_fast_accum);
+  static ::std::tuple<at::Tensor,at::Tensor> redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, const at::Tensor & mat2, const c10::optional<at::Tensor> & bias, c10::optional<at::ScalarType> out_dtype, const c10::optional<at::Tensor> & scale_a, const c10::optional<at::Tensor> & scale_b, const c10::optional<at::Tensor> & scale_result, bool use_fast_accum);
+};
+
+struct TORCH_API _scaled_mm_out {
+  using schema = ::std::tuple<at::Tensor &,at::Tensor &> (const at::Tensor &, const at::Tensor &, const c10::optional<at::Tensor> &, c10::optional<at::ScalarType>, const c10::optional<at::Tensor> &, const c10::optional<at::Tensor> &, const c10::optional<at::Tensor> &, bool, 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::_scaled_mm")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "out")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "_scaled_mm.out(Tensor self, Tensor mat2, *, Tensor? bias=None, ScalarType? out_dtype=None, Tensor? scale_a=None, Tensor? scale_b=None, Tensor? scale_result=None, bool use_fast_accum=False, Tensor(a!) out, Tensor(b!) out_amax) -> (Tensor(a!), Tensor(b!))")
+  static ::std::tuple<at::Tensor &,at::Tensor &> call(const at::Tensor & self, const at::Tensor & mat2, const c10::optional<at::Tensor> & bias, c10::optional<at::ScalarType> out_dtype, const c10::optional<at::Tensor> & scale_a, const c10::optional<at::Tensor> & scale_b, const c10::optional<at::Tensor> & scale_result, bool use_fast_accum, at::Tensor & out, at::Tensor & out_amax);
+  static ::std::tuple<at::Tensor &,at::Tensor &> redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, const at::Tensor & mat2, const c10::optional<at::Tensor> & bias, c10::optional<at::ScalarType> out_dtype, const c10::optional<at::Tensor> & scale_a, const c10::optional<at::Tensor> & scale_b, const c10::optional<at::Tensor> & scale_result, bool use_fast_accum, at::Tensor & out, at::Tensor & out_amax);
+};
+
+}} // namespace at::_ops

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_sobol_engine_ff_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 & _sobol_engine_ff_(at::Tensor & self, int64_t n, const at::Tensor & sobolstate, int64_t dimension, int64_t num_generated);
+
+} // namespace compositeimplicitautograd
+} // namespace at

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_sparse_addmm_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 _sparse_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 & _sparse_addmm_out(const at::Tensor & self, const at::Tensor & mat1, const at::Tensor & mat2, const at::Scalar & beta, const at::Scalar & alpha, at::Tensor & out);
+} // namespace native
+} // namespace at

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_sparse_broadcast_to_copy_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 & _sparse_broadcast_to_copy_out(at::Tensor & out, const at::Tensor & self, at::IntArrayRef size);
+TORCH_API at::Tensor & _sparse_broadcast_to_copy_outf(const at::Tensor & self, at::IntArrayRef size, at::Tensor & out);
+
+} // namespace compositeexplicitautograd
+} // namespace at

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_sparse_csr_tensor_unsafe_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 _sparse_csr_tensor_unsafe {
+  using schema = at::Tensor (const at::Tensor &, const at::Tensor &, const at::Tensor &, at::IntArrayRef, c10::optional<at::ScalarType>, c10::optional<at::Layout>, c10::optional<at::Device>, c10::optional<bool>);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(name, "aten::_sparse_csr_tensor_unsafe")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "_sparse_csr_tensor_unsafe(Tensor crow_indices, Tensor col_indices, Tensor values, int[] size, *, ScalarType? dtype=None, Layout? layout=None, Device? device=None, bool? pin_memory=None) -> Tensor")
+  static at::Tensor call(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);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, 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 at::_ops

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_upsample_bicubic2d_aa_native.h

@@ -0,0 +1,27 @@
+#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/_upsample_bicubic2d_aa_meta.h>
+
+namespace at {
+namespace native {
+TORCH_API at::Tensor _upsample_bicubic2d_aa(const at::Tensor & input, at::OptionalIntArrayRef output_size, bool align_corners, c10::optional<at::ArrayRef<double>> scale_factors);
+struct TORCH_API structured__upsample_bicubic2d_aa_out_cpu : public at::meta::structured__upsample_bicubic2d_aa {
+void impl(const at::Tensor & self, at::ArrayRef<int64_t> output_size, bool align_corners, c10::optional<double> scales_h, c10::optional<double> scales_w, const at::Tensor & out);
+};
+struct TORCH_API structured__upsample_bicubic2d_aa_out_cuda : public at::meta::structured__upsample_bicubic2d_aa {
+void impl(const at::Tensor & self, at::ArrayRef<int64_t> output_size, bool align_corners, c10::optional<double> scales_h, c10::optional<double> scales_w, const at::Tensor & out);
+};
+} // namespace native
+} // namespace at

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_upsample_nearest_exact1d_backward_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_exact1d_backward(const at::Tensor & grad_output, at::IntArrayRef output_size, at::IntArrayRef input_size, c10::optional<double> scales=c10::nullopt);
+TORCH_API at::Tensor _upsample_nearest_exact1d_backward_symint(const at::Tensor & grad_output, c10::SymIntArrayRef output_size, c10::SymIntArrayRef input_size, c10::optional<double> scales=c10::nullopt);
+TORCH_API at::Tensor & _upsample_nearest_exact1d_backward_out(at::Tensor & grad_input, const at::Tensor & grad_output, at::IntArrayRef output_size, at::IntArrayRef input_size, c10::optional<double> scales=c10::nullopt);
+TORCH_API at::Tensor & _upsample_nearest_exact1d_backward_outf(const at::Tensor & grad_output, at::IntArrayRef output_size, at::IntArrayRef input_size, c10::optional<double> scales, at::Tensor & grad_input);
+TORCH_API at::Tensor & _upsample_nearest_exact1d_backward_symint_out(at::Tensor & grad_input, const at::Tensor & grad_output, c10::SymIntArrayRef output_size, c10::SymIntArrayRef input_size, c10::optional<double> scales=c10::nullopt);
+TORCH_API at::Tensor & _upsample_nearest_exact1d_backward_symint_outf(const at::Tensor & grad_output, c10::SymIntArrayRef output_size, c10::SymIntArrayRef input_size, c10::optional<double> scales, at::Tensor & grad_input);
+
+} // namespace cpu
+} // namespace at

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_weight_norm_interface_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 ::std::tuple<at::Tensor,at::Tensor> _weight_norm_interface(const at::Tensor & v, const at::Tensor & g, int64_t dim=0);
+
+} // namespace cpu
+} // namespace at

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

@@ -0,0 +1,35 @@
+#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/addmm_meta.h>
+
+namespace at {
+namespace native {
+struct TORCH_API structured_addmm_out_cpu : public at::meta::structured_addmm {
+void impl(const at::Tensor & self, const at::Tensor & mat1, const at::Tensor & mat2, const at::Scalar & beta, const at::Scalar & alpha, const at::Tensor & out);
+};
+struct TORCH_API structured_addmm_out_cuda : public at::meta::structured_addmm {
+void impl(const at::Tensor & self, const at::Tensor & mat1, const at::Tensor & mat2, const at::Scalar & beta, const at::Scalar & alpha, const at::Tensor & out);
+};
+TORCH_API at::Tensor addmm_sparse_dense_cpu(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_sparse_dense_cpu(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 & s_addmm_sparse_dense_cpu_(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_sparse_dense_cuda(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_sparse_dense_cuda(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 & s_addmm_sparse_dense_cuda_(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_sparse_compressed_dense(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_sparse_compressed_cpu(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_out_sparse_compressed_cuda(const at::Tensor & self, const at::Tensor & mat1, const at::Tensor & mat2, const at::Scalar & beta, const at::Scalar & alpha, at::Tensor & out);
+} // namespace native
+} // namespace at

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

@@ -0,0 +1,49 @@
+#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/argsort_ops.h>
+
+namespace at {
+
+
+// aten::argsort(Tensor self, int dim=-1, bool descending=False) -> Tensor
+inline at::Tensor argsort(const at::Tensor & self, int64_t dim=-1, bool descending=false) {
+    return at::_ops::argsort::call(self, dim, descending);
+}
+
+// aten::argsort.stable(Tensor self, *, bool stable, int dim=-1, bool descending=False) -> Tensor
+inline at::Tensor argsort(const at::Tensor & self, bool stable, int64_t dim=-1, bool descending=false) {
+    return at::_ops::argsort_stable::call(self, stable, dim, descending);
+}
+
+// aten::argsort.dimname(Tensor self, Dimname dim, bool descending=False) -> Tensor
+inline at::Tensor argsort(const at::Tensor & self, at::Dimname dim, bool descending=false) {
+    return at::_ops::argsort_dimname::call(self, dim, descending);
+}
+
+// aten::argsort.stable_out(Tensor self, *, bool stable, int dim=-1, bool descending=False, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & argsort_out(at::Tensor & out, const at::Tensor & self, bool stable, int64_t dim=-1, bool descending=false) {
+    return at::_ops::argsort_stable_out::call(self, stable, dim, descending, out);
+}
+// aten::argsort.stable_out(Tensor self, *, bool stable, int dim=-1, bool descending=False, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & argsort_outf(const at::Tensor & self, bool stable, int64_t dim, bool descending, at::Tensor & out) {
+    return at::_ops::argsort_stable_out::call(self, stable, dim, descending, out);
+}
+
+}

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/asin_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 asin(const at::Tensor & self);
+TORCH_API at::Tensor & asin_out(at::Tensor & out, const at::Tensor & self);
+TORCH_API at::Tensor & asin_outf(const at::Tensor & self, at::Tensor & out);
+TORCH_API at::Tensor & asin_(at::Tensor & self);
+
+} // namespace cuda
+} // namespace at

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/atanh_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_atanh : 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/avg_pool2d_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 avg_pool2d(const at::Tensor & self, at::IntArrayRef kernel_size, at::IntArrayRef stride={}, at::IntArrayRef padding=0, bool ceil_mode=false, bool count_include_pad=true, c10::optional<int64_t> divisor_override=c10::nullopt);
+TORCH_API at::Tensor & avg_pool2d_out(at::Tensor & out, const at::Tensor & self, at::IntArrayRef kernel_size, at::IntArrayRef stride={}, at::IntArrayRef padding=0, bool ceil_mode=false, bool count_include_pad=true, c10::optional<int64_t> divisor_override=c10::nullopt);
+TORCH_API at::Tensor & avg_pool2d_outf(const at::Tensor & self, at::IntArrayRef kernel_size, at::IntArrayRef stride, at::IntArrayRef padding, bool ceil_mode, bool count_include_pad, c10::optional<int64_t> divisor_override, at::Tensor & out);
+
+} // namespace cuda
+} // namespace at

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/cartesian_prod_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 cartesian_prod {
+  using schema = 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::cartesian_prod")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "cartesian_prod(Tensor[] tensors) -> Tensor")
+  static at::Tensor call(at::TensorList tensors);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, at::TensorList tensors);
+};
+
+}} // namespace at::_ops

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/ccol_indices_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 ccol_indices {
+  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::ccol_indices")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "ccol_indices(Tensor(a) self) -> Tensor(a)")
+  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/clamp_max_compositeexplicitautogradnonfunctional_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 compositeexplicitautogradnonfunctional {
+
+TORCH_API at::Tensor clamp_max(const at::Tensor & self, const at::Scalar & max);
+TORCH_API at::Tensor & clamp_max_(at::Tensor & self, const at::Scalar & max);
+TORCH_API at::Tensor clamp_max(const at::Tensor & self, const at::Tensor & max);
+TORCH_API at::Tensor & clamp_max_(at::Tensor & self, const at::Tensor & max);
+
+} // namespace compositeexplicitautogradnonfunctional
+} // namespace at

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/clamp_meta_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 meta {
+
+TORCH_API at::Tensor clamp(const at::Tensor & self, const c10::optional<at::Scalar> & min, const c10::optional<at::Scalar> & max=c10::nullopt);
+TORCH_API at::Tensor & clamp_out(at::Tensor & out, const at::Tensor & self, const c10::optional<at::Scalar> & min, const c10::optional<at::Scalar> & max=c10::nullopt);
+TORCH_API at::Tensor & clamp_outf(const at::Tensor & self, const c10::optional<at::Scalar> & min, const c10::optional<at::Scalar> & max, at::Tensor & out);
+TORCH_API at::Tensor & clamp_(at::Tensor & self, const c10::optional<at::Scalar> & min, const c10::optional<at::Scalar> & max=c10::nullopt);
+TORCH_API at::Tensor clamp(const at::Tensor & self, const c10::optional<at::Tensor> & min={}, const c10::optional<at::Tensor> & max={});
+TORCH_API at::Tensor & clamp_out(at::Tensor & out, const at::Tensor & self, const c10::optional<at::Tensor> & min={}, const c10::optional<at::Tensor> & max={});
+TORCH_API at::Tensor & clamp_outf(const at::Tensor & self, const c10::optional<at::Tensor> & min, const c10::optional<at::Tensor> & max, at::Tensor & out);
+TORCH_API at::Tensor & clamp_(at::Tensor & self, const c10::optional<at::Tensor> & min={}, const c10::optional<at::Tensor> & max={});
+
+} // namespace meta
+} // namespace at

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/crow_indices_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 crow_indices {
+  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::crow_indices")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "crow_indices(Tensor(a) self) -> Tensor(a)")
+  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/cudnn_convolution_relu_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 & cudnn_convolution_relu_out_symint(const at::Tensor & self, const at::Tensor & weight, const c10::optional<at::Tensor> & bias, c10::SymIntArrayRef stride, c10::SymIntArrayRef padding, c10::SymIntArrayRef dilation, c10::SymInt groups, at::Tensor & out);
+TORCH_API at::Tensor cudnn_convolution_relu(const at::Tensor & self, const at::Tensor & weight, const c10::optional<at::Tensor> & bias, at::IntArrayRef stride, at::IntArrayRef padding, at::IntArrayRef dilation, int64_t groups);
+} // namespace native
+} // namespace at

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

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

@@ -0,0 +1,91 @@
+#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_ops.h>
+
+namespace at {
+
+
+// aten::embedding(Tensor weight, Tensor indices, SymInt padding_idx=-1, bool scale_grad_by_freq=False, bool sparse=False) -> Tensor
+inline at::Tensor embedding(const at::Tensor & weight, const at::Tensor & indices, int64_t padding_idx=-1, bool scale_grad_by_freq=false, bool sparse=false) {
+    return at::_ops::embedding::call(weight, indices, padding_idx, scale_grad_by_freq, sparse);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same<T, int64_t>::value>>
+  at::Tensor embedding(const at::Tensor & weight, const at::Tensor & indices, int64_t padding_idx=-1, bool scale_grad_by_freq=false, bool sparse=false) {
+    return at::_ops::embedding::call(weight, indices, padding_idx, scale_grad_by_freq, sparse);
+  }
+}
+
+// aten::embedding(Tensor weight, Tensor indices, SymInt padding_idx=-1, bool scale_grad_by_freq=False, bool sparse=False) -> Tensor
+inline at::Tensor embedding_symint(const at::Tensor & weight, const at::Tensor & indices, c10::SymInt padding_idx=-1, bool scale_grad_by_freq=false, bool sparse=false) {
+    return at::_ops::embedding::call(weight, indices, padding_idx, scale_grad_by_freq, sparse);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same<T, c10::SymInt>::value>>
+  at::Tensor embedding(const at::Tensor & weight, const at::Tensor & indices, c10::SymInt padding_idx=-1, bool scale_grad_by_freq=false, bool sparse=false) {
+    return at::_ops::embedding::call(weight, indices, padding_idx, scale_grad_by_freq, sparse);
+  }
+}
+
+// aten::embedding.out(Tensor weight, Tensor indices, SymInt padding_idx=-1, bool scale_grad_by_freq=False, bool sparse=False, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & embedding_out(at::Tensor & out, const at::Tensor & weight, const at::Tensor & indices, int64_t padding_idx=-1, bool scale_grad_by_freq=false, bool sparse=false) {
+    return at::_ops::embedding_out::call(weight, indices, padding_idx, scale_grad_by_freq, sparse, out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same<T, int64_t>::value>>
+  at::Tensor & embedding_out(at::Tensor & out, const at::Tensor & weight, const at::Tensor & indices, int64_t padding_idx=-1, bool scale_grad_by_freq=false, bool sparse=false) {
+    return at::_ops::embedding_out::call(weight, indices, padding_idx, scale_grad_by_freq, sparse, out);
+  }
+}
+
+// aten::embedding.out(Tensor weight, Tensor indices, SymInt padding_idx=-1, bool scale_grad_by_freq=False, bool sparse=False, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & embedding_outf(const at::Tensor & weight, const at::Tensor & indices, int64_t padding_idx, bool scale_grad_by_freq, bool sparse, at::Tensor & out) {
+    return at::_ops::embedding_out::call(weight, indices, padding_idx, scale_grad_by_freq, sparse, out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same<T, int64_t>::value>>
+  at::Tensor & embedding_outf(const at::Tensor & weight, const at::Tensor & indices, int64_t padding_idx, bool scale_grad_by_freq, bool sparse, at::Tensor & out) {
+    return at::_ops::embedding_out::call(weight, indices, padding_idx, scale_grad_by_freq, sparse, out);
+  }
+}
+
+// aten::embedding.out(Tensor weight, Tensor indices, SymInt padding_idx=-1, bool scale_grad_by_freq=False, bool sparse=False, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & embedding_symint_out(at::Tensor & out, const at::Tensor & weight, const at::Tensor & indices, c10::SymInt padding_idx=-1, bool scale_grad_by_freq=false, bool sparse=false) {
+    return at::_ops::embedding_out::call(weight, indices, padding_idx, scale_grad_by_freq, sparse, out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same<T, c10::SymInt>::value>>
+  at::Tensor & embedding_out(at::Tensor & out, const at::Tensor & weight, const at::Tensor & indices, c10::SymInt padding_idx=-1, bool scale_grad_by_freq=false, bool sparse=false) {
+    return at::_ops::embedding_out::call(weight, indices, padding_idx, scale_grad_by_freq, sparse, out);
+  }
+}
+
+// aten::embedding.out(Tensor weight, Tensor indices, SymInt padding_idx=-1, bool scale_grad_by_freq=False, bool sparse=False, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & embedding_symint_outf(const at::Tensor & weight, const at::Tensor & indices, c10::SymInt padding_idx, bool scale_grad_by_freq, bool sparse, at::Tensor & out) {
+    return at::_ops::embedding_out::call(weight, indices, padding_idx, scale_grad_by_freq, sparse, out);
+}
+namespace symint {
+  template <typename T, typename = std::enable_if_t<std::is_same<T, c10::SymInt>::value>>
+  at::Tensor & embedding_outf(const at::Tensor & weight, const at::Tensor & indices, c10::SymInt padding_idx, bool scale_grad_by_freq, bool sparse, at::Tensor & out) {
+    return at::_ops::embedding_out::call(weight, indices, padding_idx, scale_grad_by_freq, sparse, out);
+  }
+}
+
+}

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

@@ -0,0 +1,63 @@
+#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/fill_ops.h>
+
+namespace at {
+
+
+// aten::fill.Scalar(Tensor self, Scalar value) -> Tensor
+inline at::Tensor fill(const at::Tensor & self, const at::Scalar & value) {
+    return at::_ops::fill_Scalar::call(self, value);
+}
+
+// aten::fill.Tensor(Tensor self, Tensor value) -> Tensor
+inline at::Tensor fill(const at::Tensor & self, const at::Tensor & value) {
+    return at::_ops::fill_Tensor::call(self, value);
+}
+
+// aten::fill_.Scalar(Tensor(a!) self, Scalar value) -> Tensor(a!)
+inline at::Tensor & fill_(at::Tensor & self, const at::Scalar & value) {
+    return at::_ops::fill__Scalar::call(self, value);
+}
+
+// aten::fill_.Tensor(Tensor(a!) self, Tensor value) -> Tensor(a!)
+inline at::Tensor & fill_(at::Tensor & self, const at::Tensor & value) {
+    return at::_ops::fill__Tensor::call(self, value);
+}
+
+// aten::fill.Scalar_out(Tensor self, Scalar value, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & fill_out(at::Tensor & out, const at::Tensor & self, const at::Scalar & value) {
+    return at::_ops::fill_Scalar_out::call(self, value, out);
+}
+// aten::fill.Scalar_out(Tensor self, Scalar value, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & fill_outf(const at::Tensor & self, const at::Scalar & value, at::Tensor & out) {
+    return at::_ops::fill_Scalar_out::call(self, value, out);
+}
+
+// aten::fill.Tensor_out(Tensor self, Tensor value, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & fill_out(at::Tensor & out, const at::Tensor & self, const at::Tensor & value) {
+    return at::_ops::fill_Tensor_out::call(self, value, out);
+}
+// aten::fill.Tensor_out(Tensor self, Tensor value, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & fill_outf(const at::Tensor & self, const at::Tensor & value, at::Tensor & out) {
+    return at::_ops::fill_Tensor_out::call(self, value, out);
+}
+
+}

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/fmod_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 fmod(const at::Tensor & self, const at::Tensor & other);
+TORCH_API at::Tensor & fmod_out(at::Tensor & out, const at::Tensor & self, const at::Tensor & other);
+TORCH_API at::Tensor & fmod_outf(const at::Tensor & self, const at::Tensor & other, at::Tensor & out);
+TORCH_API at::Tensor & fmod_(at::Tensor & self, const at::Tensor & other);
+
+} // namespace meta
+} // namespace at

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/fractional_max_pool2d_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 fractional_max_pool2d_output {
+  using schema = ::std::tuple<at::Tensor &,at::Tensor &> (const at::Tensor &, at::IntArrayRef, at::IntArrayRef, const 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::fractional_max_pool2d")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "output")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "fractional_max_pool2d.output(Tensor self, int[2] kernel_size, int[2] output_size, Tensor random_samples, *, Tensor(a!) output, Tensor(b!) indices) -> (Tensor(a!), Tensor(b!))")
+  static ::std::tuple<at::Tensor &,at::Tensor &> call(const at::Tensor & self, at::IntArrayRef kernel_size, at::IntArrayRef output_size, const at::Tensor & random_samples, at::Tensor & output, at::Tensor & indices);
+  static ::std::tuple<at::Tensor &,at::Tensor &> redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, at::IntArrayRef kernel_size, at::IntArrayRef output_size, const at::Tensor & random_samples, at::Tensor & output, at::Tensor & indices);
+};
+
+struct TORCH_API fractional_max_pool2d {
+  using schema = ::std::tuple<at::Tensor,at::Tensor> (const at::Tensor &, at::IntArrayRef, at::IntArrayRef, 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::fractional_max_pool2d")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "fractional_max_pool2d(Tensor self, int[2] kernel_size, int[2] output_size, Tensor random_samples) -> (Tensor, Tensor)")
+  static ::std::tuple<at::Tensor,at::Tensor> call(const at::Tensor & self, at::IntArrayRef kernel_size, at::IntArrayRef output_size, const at::Tensor & random_samples);
+  static ::std::tuple<at::Tensor,at::Tensor> redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, at::IntArrayRef kernel_size, at::IntArrayRef output_size, const at::Tensor & random_samples);
+};
+
+}} // namespace at::_ops

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/ge_cuda_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 cuda {
+
+TORCH_API at::Tensor ge(const at::Tensor & self, const at::Scalar & other);
+TORCH_API at::Tensor & ge_out(at::Tensor & out, const at::Tensor & self, const at::Scalar & other);
+TORCH_API at::Tensor & ge_outf(const at::Tensor & self, const at::Scalar & other, at::Tensor & out);
+TORCH_API at::Tensor & ge_(at::Tensor & self, const at::Scalar & other);
+TORCH_API at::Tensor ge(const at::Tensor & self, const at::Tensor & other);
+TORCH_API at::Tensor & ge_out(at::Tensor & out, const at::Tensor & self, const at::Tensor & other);
+TORCH_API at::Tensor & ge_outf(const at::Tensor & self, const at::Tensor & other, at::Tensor & out);
+TORCH_API at::Tensor & ge_(at::Tensor & self, const at::Tensor & other);
+
+} // namespace cuda
+} // namespace at

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/gelu_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 gelu(const at::Tensor & self, c10::string_view approximate="none");
+TORCH_API at::Tensor & gelu_out(at::Tensor & out, const at::Tensor & self, c10::string_view approximate="none");
+TORCH_API at::Tensor & gelu_outf(const at::Tensor & self, c10::string_view approximate, at::Tensor & out);
+TORCH_API at::Tensor & gelu_(at::Tensor & self, c10::string_view approximate="none");
+
+} // namespace cuda
+} // namespace at

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/glu_backward_jvp_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 & glu_backward_jvp_out(const at::Tensor & grad_x, const at::Tensor & grad_glu, const at::Tensor & x, const at::Tensor & dgrad_glu, const at::Tensor & dx, int64_t dim, at::Tensor & out);
+TORCH_API at::Tensor glu_backward_jvp(const at::Tensor & grad_x, const at::Tensor & grad_glu, const at::Tensor & x, const at::Tensor & dgrad_glu, const at::Tensor & dx, int64_t dim);
+} // namespace native
+} // namespace at

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/grid_sampler_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 grid_sampler(const at::Tensor & input, const at::Tensor & grid, int64_t interpolation_mode, int64_t padding_mode, bool align_corners);
+} // namespace native
+} // namespace at

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/hardshrink_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 hardshrink_backward_grad_input {
+  using schema = at::Tensor & (const at::Tensor &, const at::Tensor &, const at::Scalar &, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(name, "aten::hardshrink_backward")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "grad_input")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "hardshrink_backward.grad_input(Tensor grad_out, Tensor self, Scalar lambd, *, Tensor(a!) grad_input) -> Tensor(a!)")
+  static at::Tensor & call(const at::Tensor & grad_out, const at::Tensor & self, const at::Scalar & lambd, at::Tensor & grad_input);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & grad_out, const at::Tensor & self, const at::Scalar & lambd, at::Tensor & grad_input);
+};
+
+struct TORCH_API hardshrink_backward {
+  using schema = at::Tensor (const at::Tensor &, const at::Tensor &, 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::hardshrink_backward")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "hardshrink_backward(Tensor grad_out, Tensor self, Scalar lambd) -> Tensor")
+  static at::Tensor call(const at::Tensor & grad_out, const at::Tensor & self, const at::Scalar & lambd);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & grad_out, const at::Tensor & self, const at::Scalar & lambd);
+};
+
+}} // namespace at::_ops

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/leaky_relu_backward_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 leaky_relu_backward(const at::Tensor & grad_output, const at::Tensor & self, const at::Scalar & negative_slope, bool self_is_result);
+TORCH_API at::Tensor & leaky_relu_backward_out(at::Tensor & grad_input, const at::Tensor & grad_output, const at::Tensor & self, const at::Scalar & negative_slope, bool self_is_result);
+TORCH_API at::Tensor & leaky_relu_backward_outf(const at::Tensor & grad_output, const at::Tensor & self, const at::Scalar & negative_slope, bool self_is_result, at::Tensor & grad_input);
+
+} // namespace cpu
+} // namespace at

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/less_equal_compositeimplicitautograd_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 compositeimplicitautograd {
+
+TORCH_API at::Tensor less_equal(const at::Tensor & self, const at::Scalar & other);
+TORCH_API at::Tensor & less_equal_out(at::Tensor & out, const at::Tensor & self, const at::Scalar & other);
+TORCH_API at::Tensor & less_equal_outf(const at::Tensor & self, const at::Scalar & other, at::Tensor & out);
+TORCH_API at::Tensor & less_equal_(at::Tensor & self, const at::Scalar & other);
+TORCH_API at::Tensor less_equal(const at::Tensor & self, const at::Tensor & other);
+TORCH_API at::Tensor & less_equal_out(at::Tensor & out, const at::Tensor & self, const at::Tensor & other);
+TORCH_API at::Tensor & less_equal_outf(const at::Tensor & self, const at::Tensor & other, at::Tensor & out);
+TORCH_API at::Tensor & less_equal_(at::Tensor & self, const at::Tensor & other);
+
+} // namespace compositeimplicitautograd
+} // namespace at

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/linalg_solve.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_solve_ops.h>
+
+namespace at {
+
+
+// aten::linalg_solve(Tensor A, Tensor B, *, bool left=True) -> Tensor
+inline at::Tensor linalg_solve(const at::Tensor & A, const at::Tensor & B, bool left=true) {
+    return at::_ops::linalg_solve::call(A, B, left);
+}
+
+// aten::linalg_solve.out(Tensor A, Tensor B, *, bool left=True, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & linalg_solve_out(at::Tensor & out, const at::Tensor & A, const at::Tensor & B, bool left=true) {
+    return at::_ops::linalg_solve_out::call(A, B, left, out);
+}
+// aten::linalg_solve.out(Tensor A, Tensor B, *, bool left=True, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & linalg_solve_outf(const at::Tensor & A, const at::Tensor & B, bool left, at::Tensor & out) {
+    return at::_ops::linalg_solve_out::call(A, B, left, out);
+}
+
+}

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