Add files using upload-large-folder tool

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_adaptive_avg_pool3d_cuda_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 cuda {
+
+TORCH_API at::Tensor _adaptive_avg_pool3d(const at::Tensor & self, at::IntArrayRef output_size);
+TORCH_API at::Tensor _adaptive_avg_pool3d_symint(const at::Tensor & self, c10::SymIntArrayRef output_size);
+
+} // namespace cuda
+} // namespace at

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_autocast_to_full_precision_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 _autocast_to_full_precision(const at::Tensor & self, bool cuda_enabled, bool cpu_enabled);
+} // namespace native
+} // namespace at

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_cast_Long.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/_cast_Long_ops.h>
+
+namespace at {
+
+
+// aten::_cast_Long(Tensor self, bool non_blocking=False) -> Tensor
+inline at::Tensor _cast_Long(const at::Tensor & self, bool non_blocking=false) {
+    return at::_ops::_cast_Long::call(self, non_blocking);
+}
+
+}

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_ctc_loss_backward_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 at::Tensor _ctc_loss_backward(const at::Tensor & grad, const at::Tensor & log_probs, const at::Tensor & targets, at::IntArrayRef input_lengths, at::IntArrayRef target_lengths, const at::Tensor & neg_log_likelihood, const at::Tensor & log_alpha, int64_t blank, bool zero_infinity=false);
+TORCH_API at::Tensor _ctc_loss_backward(const at::Tensor & grad, const at::Tensor & log_probs, const at::Tensor & targets, const at::Tensor & input_lengths, const at::Tensor & target_lengths, const at::Tensor & neg_log_likelihood, const at::Tensor & log_alpha, int64_t blank, bool zero_infinity=false);
+
+} // namespace cpu
+} // namespace at

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

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

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

@@ -0,0 +1,25 @@
+#pragma once
+// @generated by torchgen/gen.py from DispatchKeyFunction.h
+
+// NB: The implementing C++ file is RegisterDispatchKey.cpp
+
+// The only #includes we need are for custom classes that have defaults in the C++ API
+#include <c10/core/MemoryFormat.h>
+#include <c10/core/Scalar.h>
+#include <ATen/core/Reduction.h>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+
+namespace compositeexplicitautograd {
+
+TORCH_API at::Tensor _euclidean_dist(const at::Tensor & x1, const at::Tensor & x2);
+TORCH_API at::Tensor & _euclidean_dist_out(at::Tensor & out, const at::Tensor & x1, const at::Tensor & x2);
+TORCH_API at::Tensor & _euclidean_dist_outf(const at::Tensor & x1, const at::Tensor & x2, at::Tensor & out);
+
+} // namespace compositeexplicitautograd
+} // namespace at

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_fft_r2c.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/_fft_r2c_ops.h>
+
+namespace at {
+
+
+// aten::_fft_r2c(Tensor self, int[] dim, int normalization, bool onesided) -> Tensor
+inline at::Tensor _fft_r2c(const at::Tensor & self, at::IntArrayRef dim, int64_t normalization, bool onesided) {
+    return at::_ops::_fft_r2c::call(self, dim, normalization, onesided);
+}
+
+// aten::_fft_r2c.out(Tensor self, int[] dim, int normalization, bool onesided, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & _fft_r2c_out(at::Tensor & out, const at::Tensor & self, at::IntArrayRef dim, int64_t normalization, bool onesided) {
+    return at::_ops::_fft_r2c_out::call(self, dim, normalization, onesided, out);
+}
+// aten::_fft_r2c.out(Tensor self, int[] dim, int normalization, bool onesided, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & _fft_r2c_outf(const at::Tensor & self, at::IntArrayRef dim, int64_t normalization, bool onesided, at::Tensor & out) {
+    return at::_ops::_fft_r2c_out::call(self, dim, normalization, onesided, out);
+}
+
+}

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

@@ -0,0 +1,82 @@
+#pragma once
+
+// @generated by torchgen/gen.py from Function.h
+
+#include <ATen/Context.h>
+#include <ATen/DeviceGuard.h>
+#include <ATen/TensorUtils.h>
+#include <ATen/TracerMode.h>
+#include <ATen/core/Generator.h>
+#include <ATen/core/Reduction.h>
+#include <ATen/core/Tensor.h>
+#include <c10/core/Scalar.h>
+#include <c10/core/Storage.h>
+#include <c10/core/TensorOptions.h>
+#include <c10/util/Deprecated.h>
+#include <c10/util/Optional.h>
+
+
+
+#include <ATen/ops/_foreach_addcdiv_ops.h>
+
+namespace at {
+
+
+// aten::_foreach_addcdiv.Scalar(Tensor[] self, Tensor[] tensor1, Tensor[] tensor2, Scalar value=1) -> Tensor[]
+inline ::std::vector<at::Tensor> _foreach_addcdiv(at::TensorList self, at::TensorList tensor1, at::TensorList tensor2, const at::Scalar & value=1) {
+    return at::_ops::_foreach_addcdiv_Scalar::call(self, tensor1, tensor2, value);
+}
+
+// aten::_foreach_addcdiv.ScalarList(Tensor[] self, Tensor[] tensor1, Tensor[] tensor2, Scalar[] scalars) -> Tensor[]
+inline ::std::vector<at::Tensor> _foreach_addcdiv(at::TensorList self, at::TensorList tensor1, at::TensorList tensor2, at::ArrayRef<at::Scalar> scalars) {
+    return at::_ops::_foreach_addcdiv_ScalarList::call(self, tensor1, tensor2, scalars);
+}
+
+// aten::_foreach_addcdiv.Tensor(Tensor[] self, Tensor[] tensor1, Tensor[] tensor2, Tensor scalars) -> Tensor[]
+inline ::std::vector<at::Tensor> _foreach_addcdiv(at::TensorList self, at::TensorList tensor1, at::TensorList tensor2, const at::Tensor & scalars) {
+    return at::_ops::_foreach_addcdiv_Tensor::call(self, tensor1, tensor2, scalars);
+}
+
+// aten::_foreach_addcdiv_.Scalar(Tensor(a!)[] self, Tensor[] tensor1, Tensor[] tensor2, Scalar value=1) -> ()
+inline void _foreach_addcdiv_(at::TensorList self, at::TensorList tensor1, at::TensorList tensor2, const at::Scalar & value=1) {
+    return at::_ops::_foreach_addcdiv__Scalar::call(self, tensor1, tensor2, value);
+}
+
+// aten::_foreach_addcdiv_.ScalarList(Tensor(a!)[] self, Tensor[] tensor1, Tensor[] tensor2, Scalar[] scalars) -> ()
+inline void _foreach_addcdiv_(at::TensorList self, at::TensorList tensor1, at::TensorList tensor2, at::ArrayRef<at::Scalar> scalars) {
+    return at::_ops::_foreach_addcdiv__ScalarList::call(self, tensor1, tensor2, scalars);
+}
+
+// aten::_foreach_addcdiv_.Tensor(Tensor(a!)[] self, Tensor[] tensor1, Tensor[] tensor2, Tensor scalars) -> ()
+inline void _foreach_addcdiv_(at::TensorList self, at::TensorList tensor1, at::TensorList tensor2, const at::Tensor & scalars) {
+    return at::_ops::_foreach_addcdiv__Tensor::call(self, tensor1, tensor2, scalars);
+}
+
+// aten::_foreach_addcdiv.Scalar_out(Tensor[] self, Tensor[] tensor1, Tensor[] tensor2, Scalar value=1, *, Tensor(a!)[] out) -> ()
+inline void _foreach_addcdiv_out(at::TensorList out, at::TensorList self, at::TensorList tensor1, at::TensorList tensor2, const at::Scalar & value=1) {
+    return at::_ops::_foreach_addcdiv_Scalar_out::call(self, tensor1, tensor2, value, out);
+}
+// aten::_foreach_addcdiv.Scalar_out(Tensor[] self, Tensor[] tensor1, Tensor[] tensor2, Scalar value=1, *, Tensor(a!)[] out) -> ()
+inline void _foreach_addcdiv_outf(at::TensorList self, at::TensorList tensor1, at::TensorList tensor2, const at::Scalar & value, at::TensorList out) {
+    return at::_ops::_foreach_addcdiv_Scalar_out::call(self, tensor1, tensor2, value, out);
+}
+
+// aten::_foreach_addcdiv.ScalarList_out(Tensor[] self, Tensor[] tensor1, Tensor[] tensor2, Scalar[] scalars, *, Tensor(a!)[] out) -> ()
+inline void _foreach_addcdiv_out(at::TensorList out, at::TensorList self, at::TensorList tensor1, at::TensorList tensor2, at::ArrayRef<at::Scalar> scalars) {
+    return at::_ops::_foreach_addcdiv_ScalarList_out::call(self, tensor1, tensor2, scalars, out);
+}
+// aten::_foreach_addcdiv.ScalarList_out(Tensor[] self, Tensor[] tensor1, Tensor[] tensor2, Scalar[] scalars, *, Tensor(a!)[] out) -> ()
+inline void _foreach_addcdiv_outf(at::TensorList self, at::TensorList tensor1, at::TensorList tensor2, at::ArrayRef<at::Scalar> scalars, at::TensorList out) {
+    return at::_ops::_foreach_addcdiv_ScalarList_out::call(self, tensor1, tensor2, scalars, out);
+}
+
+// aten::_foreach_addcdiv.Tensor_out(Tensor[] self, Tensor[] tensor1, Tensor[] tensor2, Tensor scalars, *, Tensor(a!)[] out) -> ()
+inline void _foreach_addcdiv_out(at::TensorList out, at::TensorList self, at::TensorList tensor1, at::TensorList tensor2, const at::Tensor & scalars) {
+    return at::_ops::_foreach_addcdiv_Tensor_out::call(self, tensor1, tensor2, scalars, out);
+}
+// aten::_foreach_addcdiv.Tensor_out(Tensor[] self, Tensor[] tensor1, Tensor[] tensor2, Tensor scalars, *, Tensor(a!)[] out) -> ()
+inline void _foreach_addcdiv_outf(at::TensorList self, at::TensorList tensor1, at::TensorList tensor2, const at::Tensor & scalars, at::TensorList out) {
+    return at::_ops::_foreach_addcdiv_Tensor_out::call(self, tensor1, tensor2, scalars, out);
+}
+
+}

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_foreach_lgamma_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 ::std::vector<at::Tensor> _foreach_lgamma(at::TensorList self);
+TORCH_API void _foreach_lgamma_(at::TensorList self);
+
+} // namespace cpu
+} // namespace at

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_foreach_round_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_round_out(at::TensorList self, at::TensorList out);
+TORCH_API ::std::vector<at::Tensor> foreach_tensor_round_slow(at::TensorList self);
+TORCH_API void foreach_tensor_round_slow_(at::TensorList self);
+TORCH_API ::std::vector<at::Tensor> foreach_tensor_round_cuda(at::TensorList self);
+TORCH_API void foreach_tensor_round_cuda_(at::TensorList self);
+} // namespace native
+} // namespace at

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_fused_moving_avg_obs_fq_helper_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> _fused_moving_avg_obs_fq_helper(const at::Tensor & self, const at::Tensor & observer_on, const at::Tensor & fake_quant_on, at::Tensor & running_min, at::Tensor & running_max, at::Tensor & scale, at::Tensor & zero_point, double averaging_const, int64_t quant_min, int64_t quant_max, int64_t ch_axis, bool per_row_fake_quant=false, bool symmetric_quant=false);
+
+} // namespace cpu
+} // namespace at

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_histogramdd_from_bin_tensors_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 & _histogramdd_from_bin_tensors_out(at::Tensor & out, const at::Tensor & self, at::TensorList bins, const c10::optional<at::Tensor> & weight={}, bool density=false);
+TORCH_API at::Tensor & _histogramdd_from_bin_tensors_outf(const at::Tensor & self, at::TensorList bins, const c10::optional<at::Tensor> & weight, bool density, at::Tensor & out);
+
+} // namespace compositeexplicitautograd
+} // namespace at

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_int_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 _int_mm {
+  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::_int_mm")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "_int_mm(Tensor self, Tensor mat2) -> Tensor")
+  static at::Tensor call(const at::Tensor & self, const at::Tensor & mat2);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, const at::Tensor & mat2);
+};
+
+struct TORCH_API _int_mm_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::_int_mm")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "out")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "_int_mm.out(Tensor self, Tensor mat2, *, Tensor(a!) out) -> Tensor(a!)")
+  static at::Tensor & call(const at::Tensor & self, const at::Tensor & mat2, at::Tensor & out);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, const at::Tensor & mat2, at::Tensor & out);
+};
+
+}} // namespace at::_ops

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

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_linalg_solve_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_solve_ex_ops.h>
+
+namespace at {
+
+
+// aten::_linalg_solve_ex(Tensor A, Tensor B, *, bool left=True, bool check_errors=False) -> (Tensor result, Tensor LU, Tensor pivots, Tensor info)
+inline ::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) {
+    return at::_ops::_linalg_solve_ex::call(A, B, left, check_errors);
+}
+
+// aten::_linalg_solve_ex.result(Tensor A, Tensor B, *, bool left=True, bool check_errors=False, Tensor(a!) result, Tensor(b!) LU, Tensor(c!) pivots, Tensor(d!) info) -> (Tensor(a!) result, Tensor(b!) LU, Tensor(c!) pivots, Tensor(d!) info)
+inline ::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) {
+    return at::_ops::_linalg_solve_ex_result::call(A, B, left, check_errors, result, LU, pivots, info);
+}
+// aten::_linalg_solve_ex.result(Tensor A, Tensor B, *, bool left=True, bool check_errors=False, Tensor(a!) result, Tensor(b!) LU, Tensor(c!) pivots, Tensor(d!) info) -> (Tensor(a!) result, Tensor(b!) LU, Tensor(c!) pivots, Tensor(d!) info)
+inline ::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) {
+    return at::_ops::_linalg_solve_ex_result::call(A, B, left, check_errors, result, LU, pivots, info);
+}
+
+}

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

@@ -0,0 +1,22 @@
+#pragma once
+
+// @generated by torchgen/gen.py from NativeFunction.h
+
+#include <c10/core/Scalar.h>
+#include <c10/core/Storage.h>
+#include <c10/core/TensorOptions.h>
+#include <c10/util/Deprecated.h>
+#include <c10/util/Optional.h>
+#include <c10/core/QScheme.h>
+#include <ATen/core/Reduction.h>
+#include <ATen/core/Tensor.h>
+#include <tuple>
+#include <vector>
+
+
+namespace at {
+namespace native {
+TORCH_API at::Tensor & _segment_reduce_backward_out(const at::Tensor & grad, const at::Tensor & output, const at::Tensor & data, c10::string_view reduce, const c10::optional<at::Tensor> & lengths, const c10::optional<at::Tensor> & offsets, int64_t axis, const c10::optional<at::Scalar> & initial, at::Tensor & out);
+TORCH_API at::Tensor _segment_reduce_backward_kernel(const at::Tensor & grad, const at::Tensor & output, const at::Tensor & data, c10::string_view reduce, const c10::optional<at::Tensor> & lengths={}, const c10::optional<at::Tensor> & offsets={}, int64_t axis=0, const c10::optional<at::Scalar> & initial=c10::nullopt);
+} // namespace native
+} // namespace at

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

@@ -0,0 +1,61 @@
+#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_log_softmax_int {
+  using schema = at::Tensor (const at::Tensor &, int64_t, 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_log_softmax")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "int")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "_sparse_log_softmax.int(Tensor self, int dim, ScalarType? dtype=None) -> Tensor")
+  static at::Tensor call(const at::Tensor & self, int64_t dim, c10::optional<at::ScalarType> dtype);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, int64_t dim, c10::optional<at::ScalarType> dtype);
+};
+
+struct TORCH_API _sparse_log_softmax_Dimname {
+  using schema = at::Tensor (const at::Tensor &, at::Dimname, 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_log_softmax")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "Dimname")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "_sparse_log_softmax.Dimname(Tensor self, Dimname dim, *, ScalarType? dtype=None) -> Tensor")
+  static at::Tensor call(const at::Tensor & self, at::Dimname dim, c10::optional<at::ScalarType> dtype);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, at::Dimname dim, c10::optional<at::ScalarType> dtype);
+};
+
+struct TORCH_API _sparse_log_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::_sparse_log_softmax")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "_sparse_log_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 _sparse_log_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::_sparse_log_softmax")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "out")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "_sparse_log_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_softmax.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/_sparse_softmax_ops.h>
+
+namespace at {
+
+
+// aten::_sparse_softmax.int(Tensor self, int dim, ScalarType? dtype=None) -> Tensor
+inline at::Tensor _sparse_softmax(const at::Tensor & self, int64_t dim, c10::optional<at::ScalarType> dtype=c10::nullopt) {
+    return at::_ops::_sparse_softmax_int::call(self, dim, dtype);
+}
+
+// aten::_sparse_softmax.Dimname(Tensor self, Dimname dim, *, ScalarType? dtype=None) -> Tensor
+inline at::Tensor _sparse_softmax(const at::Tensor & self, at::Dimname dim, c10::optional<at::ScalarType> dtype=c10::nullopt) {
+    return at::_ops::_sparse_softmax_Dimname::call(self, dim, dtype);
+}
+
+// aten::_sparse_softmax(Tensor self, int dim, bool half_to_float) -> Tensor
+inline at::Tensor _sparse_softmax(const at::Tensor & self, int64_t dim, bool half_to_float) {
+    return at::_ops::_sparse_softmax::call(self, dim, half_to_float);
+}
+
+// aten::_sparse_softmax.out(Tensor self, int dim, bool half_to_float, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & _sparse_softmax_out(at::Tensor & out, const at::Tensor & self, int64_t dim, bool half_to_float) {
+    return at::_ops::_sparse_softmax_out::call(self, dim, half_to_float, out);
+}
+// aten::_sparse_softmax.out(Tensor self, int dim, bool half_to_float, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & _sparse_softmax_outf(const at::Tensor & self, int64_t dim, bool half_to_float, at::Tensor & out) {
+    return at::_ops::_sparse_softmax_out::call(self, dim, half_to_float, out);
+}
+
+}

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

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

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

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

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/_weight_int4pack_mm_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 _weight_int4pack_mm {
+  using schema = at::Tensor (const at::Tensor &, const at::Tensor &, int64_t, 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::_weight_int4pack_mm")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "_weight_int4pack_mm(Tensor self, Tensor mat2, int qGroupSize, Tensor qScaleAndZeros) -> Tensor")
+  static at::Tensor call(const at::Tensor & self, const at::Tensor & mat2, int64_t qGroupSize, const at::Tensor & qScaleAndZeros);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, const at::Tensor & mat2, int64_t qGroupSize, const at::Tensor & qScaleAndZeros);
+};
+
+}} // namespace at::_ops

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

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

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/absolute.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/absolute_ops.h>
+
+namespace at {
+
+
+// aten::absolute(Tensor self) -> Tensor
+inline at::Tensor absolute(const at::Tensor & self) {
+    return at::_ops::absolute::call(self);
+}
+
+// aten::absolute.out(Tensor self, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & absolute_out(at::Tensor & out, const at::Tensor & self) {
+    return at::_ops::absolute_out::call(self, out);
+}
+// aten::absolute.out(Tensor self, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & absolute_outf(const at::Tensor & self, at::Tensor & out) {
+    return at::_ops::absolute_out::call(self, out);
+}
+
+}

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

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

@@ -0,0 +1,43 @@
+#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/add_meta.h>
+
+namespace at {
+namespace native {
+struct TORCH_API structured_ufunc_add_CPU : public at::meta::structured_add_Tensor {
+void impl(const at::Tensor & self, const at::Tensor & other, const at::Scalar & alpha, const at::Tensor & out);
+};
+struct TORCH_API structured_ufunc_add_CUDA : public at::meta::structured_add_Tensor {
+void impl(const at::Tensor & self, const at::Tensor & other, const at::Scalar & alpha, const at::Tensor & out);
+};
+TORCH_API at::Tensor NestedTensor_add_Tensor(const at::Tensor & self, const at::Tensor & other, const at::Scalar & alpha=1);
+TORCH_API at::Tensor & NestedTensor_add__Tensor(at::Tensor & self, const at::Tensor & other, const at::Scalar & alpha=1);
+TORCH_API at::Tensor add_sparse(const at::Tensor & self, const at::Tensor & other, const at::Scalar & alpha=1);
+TORCH_API at::Tensor & add_out_sparse_cpu(const at::Tensor & self, const at::Tensor & other, const at::Scalar & alpha, at::Tensor & out);
+TORCH_API at::Tensor & add_sparse_(at::Tensor & self, const at::Tensor & other, const at::Scalar & alpha=1);
+TORCH_API at::Tensor & add_out_sparse_cuda(const at::Tensor & self, const at::Tensor & other, const at::Scalar & alpha, at::Tensor & out);
+TORCH_API at::Tensor add_sparse_csr(const at::Tensor & self, const at::Tensor & other, const at::Scalar & alpha=1);
+TORCH_API at::Tensor & add_out_sparse_compressed_cpu(const at::Tensor & self, const at::Tensor & other, const at::Scalar & alpha, at::Tensor & out);
+TORCH_API at::Tensor & add_sparse_csr_(at::Tensor & self, const at::Tensor & other, const at::Scalar & alpha=1);
+TORCH_API at::Tensor & add_out_sparse_compressed_cuda(const at::Tensor & self, const at::Tensor & other, const at::Scalar & alpha, at::Tensor & out);
+TORCH_API at::Tensor mkldnn_add(const at::Tensor & self, const at::Tensor & other, const at::Scalar & alpha=1);
+TORCH_API at::Tensor & mkldnn_add_out(const at::Tensor & self, const at::Tensor & other, const at::Scalar & alpha, at::Tensor & out);
+TORCH_API at::Tensor & mkldnn_add_(at::Tensor & self, const at::Tensor & other, const at::Scalar & alpha=1);
+TORCH_API at::Tensor add_zerotensor(const at::Tensor & self, const at::Tensor & other, const at::Scalar & alpha=1);
+TORCH_API at::Tensor add(const at::Tensor & self, const at::Scalar & other, const at::Scalar & alpha=1);
+TORCH_API at::Tensor & add_Scalar_out(const at::Tensor & self, const at::Scalar & other, const at::Scalar & alpha, at::Tensor & out);
+TORCH_API at::Tensor & add_(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/affine_grid_generator_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 at::Tensor affine_grid_generator_backward(const at::Tensor & grad, at::IntArrayRef size, bool align_corners);
+} // namespace native
+} // namespace at

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/avg_pool2d_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/avg_pool2d_backward_ops.h>
+
+namespace at {
+
+
+// aten::avg_pool2d_backward.grad_input(Tensor grad_output, Tensor self, int[2] kernel_size, int[2] stride, int[2] padding, bool ceil_mode, bool count_include_pad, int? divisor_override, *, Tensor(a!) grad_input) -> Tensor(a!)
+inline at::Tensor & avg_pool2d_backward_out(at::Tensor & grad_input, const at::Tensor & grad_output, 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) {
+    return at::_ops::avg_pool2d_backward_grad_input::call(grad_output, self, kernel_size, stride, padding, ceil_mode, count_include_pad, divisor_override, grad_input);
+}
+// aten::avg_pool2d_backward.grad_input(Tensor grad_output, Tensor self, int[2] kernel_size, int[2] stride, int[2] padding, bool ceil_mode, bool count_include_pad, int? divisor_override, *, Tensor(a!) grad_input) -> Tensor(a!)
+inline at::Tensor & avg_pool2d_backward_outf(const at::Tensor & grad_output, 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 & grad_input) {
+    return at::_ops::avg_pool2d_backward_grad_input::call(grad_output, self, kernel_size, stride, padding, ceil_mode, count_include_pad, divisor_override, grad_input);
+}
+
+// aten::avg_pool2d_backward(Tensor grad_output, Tensor self, int[2] kernel_size, int[2] stride, int[2] padding, bool ceil_mode, bool count_include_pad, int? divisor_override) -> Tensor
+inline at::Tensor avg_pool2d_backward(const at::Tensor & grad_output, 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) {
+    return at::_ops::avg_pool2d_backward::call(grad_output, self, kernel_size, stride, padding, ceil_mode, count_include_pad, divisor_override);
+}
+
+}

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/batch_norm_gather_stats_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 &> batch_norm_gather_stats_out(const at::Tensor & input, const at::Tensor & mean, const at::Tensor & invstd, const c10::optional<at::Tensor> & running_mean, const c10::optional<at::Tensor> & running_var, double momentum, double eps, int64_t count, at::Tensor & out0, at::Tensor & out1);
+TORCH_API ::std::tuple<at::Tensor,at::Tensor> batch_norm_gather_stats_cuda(const at::Tensor & input, const at::Tensor & mean, const at::Tensor & invstd, const c10::optional<at::Tensor> & running_mean, const c10::optional<at::Tensor> & running_var, double momentum, double eps, int64_t count);
+} // namespace native
+} // namespace at

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/blackman_window_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 blackman_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 & blackman_window_out(int64_t window_length, at::Tensor & out);
+TORCH_API at::Tensor blackman_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 & blackman_window_periodic_out(int64_t window_length, bool periodic, at::Tensor & out);
+} // namespace native
+} // namespace at

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/channel_shuffle_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 at::Tensor channel_shuffle(const at::Tensor & self, int64_t groups);
+TORCH_API at::Tensor channel_shuffle_symint(const at::Tensor & self, c10::SymInt groups);
+
+} // namespace cpu
+} // namespace at

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

@@ -0,0 +1,26 @@
+#pragma once
+// @generated by torchgen/gen.py from DispatchKeyFunction.h
+
+// NB: The implementing C++ file is RegisterDispatchKey.cpp
+
+// The only #includes we need are for custom classes that have defaults in the C++ API
+#include <c10/core/MemoryFormat.h>
+#include <c10/core/Scalar.h>
+#include <ATen/core/Reduction.h>
+
+// Forward declarations of any types needed in the operator signatures.
+// We can't directly include these classes because it will cause circular include dependencies.
+// This file is included by TensorBody.h, which defines the Tensor class.
+#include <ATen/core/ATen_fwd.h>
+
+namespace at {
+
+namespace compositeexplicitautograd {
+
+TORCH_API at::Tensor & cudnn_convolution_add_relu_out(at::Tensor & out, const at::Tensor & self, const at::Tensor & weight, const at::Tensor & z, const c10::optional<at::Scalar> & alpha, const c10::optional<at::Tensor> & bias, at::IntArrayRef stride, at::IntArrayRef padding, at::IntArrayRef dilation, int64_t groups);
+TORCH_API at::Tensor & cudnn_convolution_add_relu_outf(const at::Tensor & self, const at::Tensor & weight, const at::Tensor & z, const c10::optional<at::Scalar> & alpha, const c10::optional<at::Tensor> & bias, at::IntArrayRef stride, at::IntArrayRef padding, at::IntArrayRef dilation, int64_t groups, at::Tensor & out);
+TORCH_API at::Tensor & cudnn_convolution_add_relu_symint_out(at::Tensor & out, const at::Tensor & self, const at::Tensor & weight, const at::Tensor & z, const c10::optional<at::Scalar> & alpha, const c10::optional<at::Tensor> & bias, c10::SymIntArrayRef stride, c10::SymIntArrayRef padding, c10::SymIntArrayRef dilation, c10::SymInt groups);
+TORCH_API at::Tensor & cudnn_convolution_add_relu_symint_outf(const at::Tensor & self, const at::Tensor & weight, const at::Tensor & z, const c10::optional<at::Scalar> & alpha, const c10::optional<at::Tensor> & bias, c10::SymIntArrayRef stride, c10::SymIntArrayRef padding, c10::SymIntArrayRef dilation, c10::SymInt groups, at::Tensor & out);
+
+} // namespace compositeexplicitautograd
+} // namespace at

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/cummax_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> cummax(const at::Tensor & self, int64_t dim);
+TORCH_API ::std::tuple<at::Tensor &,at::Tensor &> cummax_out(at::Tensor & values, at::Tensor & indices, const at::Tensor & self, int64_t dim);
+TORCH_API ::std::tuple<at::Tensor &,at::Tensor &> cummax_outf(const at::Tensor & self, int64_t dim, at::Tensor & values, at::Tensor & indices);
+
+} // namespace compositeexplicitautograd
+} // namespace at

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/dense_dim_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 dense_dim {
+  using schema = int64_t (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::dense_dim")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "dense_dim(Tensor self) -> int")
+  static int64_t call(const at::Tensor & self);
+  static int64_t redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self);
+};
+
+}} // namespace at::_ops

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/einsum_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 einsum {
+  using schema = at::Tensor (c10::string_view, at::TensorList, at::OptionalIntArrayRef);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(name, "aten::einsum")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "einsum(str equation, Tensor[] tensors, *, int[]? path=None) -> Tensor")
+  static at::Tensor call(c10::string_view equation, at::TensorList tensors, at::OptionalIntArrayRef path);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, c10::string_view equation, at::TensorList tensors, at::OptionalIntArrayRef path);
+};
+
+}} // namespace at::_ops

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/elu_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 elu_backward(const at::Tensor & grad_output, const at::Scalar & alpha, const at::Scalar & scale, const at::Scalar & input_scale, bool is_result, const at::Tensor & self_or_result);
+TORCH_API at::Tensor & elu_backward_out(at::Tensor & grad_input, const at::Tensor & grad_output, const at::Scalar & alpha, const at::Scalar & scale, const at::Scalar & input_scale, bool is_result, const at::Tensor & self_or_result);
+TORCH_API at::Tensor & elu_backward_outf(const at::Tensor & grad_output, const at::Scalar & alpha, const at::Scalar & scale, const at::Scalar & input_scale, bool is_result, const at::Tensor & self_or_result, at::Tensor & grad_input);
+
+} // namespace cpu
+} // namespace at

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/fake_quantize_per_tensor_affine_cachemask_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 ::std::tuple<at::Tensor &,at::Tensor &> fake_quantize_per_tensor_affine_cachemask_out(at::Tensor & out0, at::Tensor & out1, const at::Tensor & self, double scale, int64_t zero_point, int64_t quant_min, int64_t quant_max);
+TORCH_API ::std::tuple<at::Tensor &,at::Tensor &> fake_quantize_per_tensor_affine_cachemask_outf(const at::Tensor & self, double scale, int64_t zero_point, int64_t quant_min, int64_t quant_max, at::Tensor & out0, at::Tensor & out1);
+
+} // namespace compositeexplicitautograd
+} // namespace at

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/fbgemm_linear_fp16_weight_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 fbgemm_linear_fp16_weight {
+  using schema = at::Tensor (const 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::fbgemm_linear_fp16_weight")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "fbgemm_linear_fp16_weight(Tensor input, Tensor packed_weight, Tensor bias) -> Tensor")
+  static at::Tensor call(const at::Tensor & input, const at::Tensor & packed_weight, const at::Tensor & bias);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & input, const at::Tensor & packed_weight, const at::Tensor & bias);
+};
+
+}} // namespace at::_ops

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/fbgemm_linear_int8_weight_fp32_activation_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 fbgemm_linear_int8_weight_fp32_activation {
+  using schema = at::Tensor (const at::Tensor &, const at::Tensor &, const at::Tensor &, const at::Tensor &, const at::Scalar &, const at::Scalar &, 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::fbgemm_linear_int8_weight_fp32_activation")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "fbgemm_linear_int8_weight_fp32_activation(Tensor input, Tensor weight, Tensor packed, Tensor col_offsets, Scalar weight_scale, Scalar weight_zero_point, Tensor bias) -> Tensor")
+  static at::Tensor call(const at::Tensor & input, const at::Tensor & weight, const at::Tensor & packed, const at::Tensor & col_offsets, const at::Scalar & weight_scale, const at::Scalar & weight_zero_point, const at::Tensor & bias);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & input, const at::Tensor & weight, const at::Tensor & packed, const at::Tensor & col_offsets, const at::Scalar & weight_scale, const at::Scalar & weight_zero_point, const at::Tensor & bias);
+};
+
+}} // namespace at::_ops

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/fbgemm_pack_gemm_matrix_fp16_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 fbgemm_pack_gemm_matrix_fp16(const at::Tensor & input);
+} // namespace native
+} // namespace at

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/ger_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 ger(const at::Tensor & self, const at::Tensor & vec2);
+TORCH_API at::Tensor & ger_out(const at::Tensor & self, const at::Tensor & vec2, at::Tensor & out);
+} // namespace native
+} // namespace at

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

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

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/le.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/le_ops.h>
+
+namespace at {
+
+
+// aten::le.Scalar_out(Tensor self, Scalar other, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & le_out(at::Tensor & out, const at::Tensor & self, const at::Scalar & other) {
+    return at::_ops::le_Scalar_out::call(self, other, out);
+}
+// aten::le.Scalar_out(Tensor self, Scalar other, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & le_outf(const at::Tensor & self, const at::Scalar & other, at::Tensor & out) {
+    return at::_ops::le_Scalar_out::call(self, other, out);
+}
+
+// aten::le.Scalar(Tensor self, Scalar other) -> Tensor
+inline at::Tensor le(const at::Tensor & self, const at::Scalar & other) {
+    return at::_ops::le_Scalar::call(self, other);
+}
+
+// aten::le.Tensor_out(Tensor self, Tensor other, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & le_out(at::Tensor & out, const at::Tensor & self, const at::Tensor & other) {
+    return at::_ops::le_Tensor_out::call(self, other, out);
+}
+// aten::le.Tensor_out(Tensor self, Tensor other, *, Tensor(a!) out) -> Tensor(a!)
+inline at::Tensor & le_outf(const at::Tensor & self, const at::Tensor & other, at::Tensor & out) {
+    return at::_ops::le_Tensor_out::call(self, other, out);
+}
+
+// aten::le.Tensor(Tensor self, Tensor other) -> Tensor
+inline at::Tensor le(const at::Tensor & self, const at::Tensor & other) {
+    return at::_ops::le_Tensor::call(self, other);
+}
+
+}

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/linalg_qr_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_qr : public at::impl::MetaBase {
+    
+    
+    void meta(const at::Tensor & A, c10::string_view mode);
+};
+
+} // namespace native
+} // namespace at

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/linalg_tensorinv_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 linalg_tensorinv {
+  using schema = 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::linalg_tensorinv")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "linalg_tensorinv(Tensor self, int ind=2) -> Tensor")
+  static at::Tensor call(const at::Tensor & self, int64_t ind);
+  static at::Tensor redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, int64_t ind);
+};
+
+struct TORCH_API linalg_tensorinv_out {
+  using schema = at::Tensor & (const at::Tensor &, int64_t, at::Tensor &);
+  using ptr_schema = schema*;
+  // See Note [static constexpr char* members for windows NVCC]
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(name, "aten::linalg_tensorinv")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(overload_name, "out")
+  STATIC_CONSTEXPR_STR_INL_EXCEPT_WIN_CUDA(schema_str, "linalg_tensorinv.out(Tensor self, int ind=2, *, Tensor(a!) out) -> Tensor(a!)")
+  static at::Tensor & call(const at::Tensor & self, int64_t ind, at::Tensor & out);
+  static at::Tensor & redispatch(c10::DispatchKeySet dispatchKeySet, const at::Tensor & self, int64_t ind, at::Tensor & out);
+};
+
+}} // namespace at::_ops

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/log_sigmoid_backward_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 log_sigmoid_backward_cpu(const at::Tensor & grad_output, const at::Tensor & self, const at::Tensor & buffer);
+TORCH_API at::Tensor & log_sigmoid_backward_cpu_out(const at::Tensor & grad_output, const at::Tensor & self, const at::Tensor & buffer, at::Tensor & grad_input);
+TORCH_API at::Tensor log_sigmoid_backward_cuda(const at::Tensor & grad_output, const at::Tensor & self, const at::Tensor & buffer);
+TORCH_API at::Tensor & log_sigmoid_backward_cuda_out(const at::Tensor & grad_output, const at::Tensor & self, const at::Tensor & buffer, at::Tensor & grad_input);
+} // namespace native
+} // namespace at

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/lt_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 lt(const at::Tensor & self, const at::Scalar & other);
+TORCH_API at::Tensor & lt_out(at::Tensor & out, const at::Tensor & self, const at::Scalar & other);
+TORCH_API at::Tensor & lt_outf(const at::Tensor & self, const at::Scalar & other, at::Tensor & out);
+TORCH_API at::Tensor & lt_(at::Tensor & self, const at::Scalar & other);
+TORCH_API at::Tensor lt(const at::Tensor & self, const at::Tensor & other);
+TORCH_API at::Tensor & lt_out(at::Tensor & out, const at::Tensor & self, const at::Tensor & other);
+TORCH_API at::Tensor & lt_outf(const at::Tensor & self, const at::Tensor & other, at::Tensor & out);
+TORCH_API at::Tensor & lt_(at::Tensor & self, const at::Tensor & other);
+
+} // namespace meta
+} // namespace at

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/lu_unpack.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/lu_unpack_ops.h>
+
+namespace at {
+
+
+// aten::lu_unpack(Tensor LU_data, Tensor LU_pivots, bool unpack_data=True, bool unpack_pivots=True) -> (Tensor P, Tensor L, Tensor U)
+inline ::std::tuple<at::Tensor,at::Tensor,at::Tensor> lu_unpack(const at::Tensor & LU_data, const at::Tensor & LU_pivots, bool unpack_data=true, bool unpack_pivots=true) {
+    return at::_ops::lu_unpack::call(LU_data, LU_pivots, unpack_data, unpack_pivots);
+}
+
+// aten::lu_unpack.out(Tensor LU_data, Tensor LU_pivots, bool unpack_data=True, bool unpack_pivots=True, *, Tensor(a!) P, Tensor(b!) L, Tensor(c!) U) -> (Tensor(a!) P, Tensor(b!) L, Tensor(c!) U)
+inline ::std::tuple<at::Tensor &,at::Tensor &,at::Tensor &> lu_unpack_out(at::Tensor & P, at::Tensor & L, at::Tensor & U, const at::Tensor & LU_data, const at::Tensor & LU_pivots, bool unpack_data=true, bool unpack_pivots=true) {
+    return at::_ops::lu_unpack_out::call(LU_data, LU_pivots, unpack_data, unpack_pivots, P, L, U);
+}
+// aten::lu_unpack.out(Tensor LU_data, Tensor LU_pivots, bool unpack_data=True, bool unpack_pivots=True, *, Tensor(a!) P, Tensor(b!) L, Tensor(c!) U) -> (Tensor(a!) P, Tensor(b!) L, Tensor(c!) U)
+inline ::std::tuple<at::Tensor &,at::Tensor &,at::Tensor &> lu_unpack_outf(const at::Tensor & LU_data, const at::Tensor & LU_pivots, bool unpack_data, bool unpack_pivots, at::Tensor & P, at::Tensor & L, at::Tensor & U) {
+    return at::_ops::lu_unpack_out::call(LU_data, LU_pivots, unpack_data, unpack_pivots, P, L, U);
+}
+
+}

tuning-competition-baseline/.venv/lib/python3.11/site-packages/torch/include/ATen/ops/mean_compositeimplicitautograd_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 compositeimplicitautograd {
+
+TORCH_API at::Tensor mean(const at::Tensor & self, at::DimnameList dim, bool keepdim=false, c10::optional<at::ScalarType> dtype=c10::nullopt);
+TORCH_API at::Tensor & mean_out(at::Tensor & out, const at::Tensor & self, at::DimnameList dim, bool keepdim=false, c10::optional<at::ScalarType> dtype=c10::nullopt);
+TORCH_API at::Tensor & mean_outf(const at::Tensor & self, at::DimnameList dim, bool keepdim, c10::optional<at::ScalarType> dtype, at::Tensor & out);
+
+} // namespace compositeimplicitautograd
+} // namespace at

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